Vous êtes sur la page 1sur 92

Insect Biochemistry and Molecular Biology

Volume 35, Issue 1, Pages 1-91 (January 2005)


1.

Sugar epitopes as potential universal disease transmission blocking targets
ARTICLE
Pages 1-10
Rhoel R. Dinglasan, J ess G. Valenzuela and Abdu F. Azad

2.

Two structurally different defensin genes, one of them encoding a novel
defensin isoform, are expressed in honeybee Apis mellifera ARTICLE
Pages 11-22
J aroslav Klaudiny, tefan Albert, Katarna Bachanov, J n Kopernick and J ozef
imth

3.

Cloning and characterization of chymotrypsin- and trypsin-like cDNAs from
the gut of the Hessian fly [Mayetiola destructor (say)] ARTICLE
Pages 23-32
Yu Cheng Zhu, Xiang Liu, Ashoka A. Maddur, Brenda Oppert and Ming-Shun
Chen

4.

Identification, cloning and expression of a Cry1Ab cadherin receptor from
European corn borer, Ostrinia nubilalis (Hbner) (Lepidoptera: Crambidae)
ARTICLE
Pages 33-40
Ronald D. Flannagan, Cao-Guo Yu, J ohn P. Mathis, Terry E. Meyer, Xiaomei Shi,
Herbert A.A. Siqueira and Blair D. Siegfried

5.

Cell size control by ovarian factors regulates juvenile hormone synthesis in
corpora allata of the cockroach, Diploptera punctata ARTICLE
Pages 41-50
Ling-Wen Chang, Chuan-Mei Tsai, De-Ming Yang and Ann-Shyn Chiang

6.

A fibroin secretion-deficient silkworm mutant, Nd-s
D
, provides an efficient
system for producing recombinant proteins ARTICLE
Pages 51-59
Satoshi Inoue, Toshio Kanda, Morikazu Imamura, Guo-Xing Quan, Katsura
Kojima, Hiromitsu Tanaka, Masahiro Tomita, Rika Hino, Katsutoshi Yoshizato,
Shigeki Mizuno and Toshiki Tamura

7.

Molecular cloning and functional characterization of a neuronal choline
transporter from Trichoplusia ni ARTICLE
Pages 61-72
Heather McLean, LouAnn Verellen, Stanley Caveney and Cam Donly

8.

Characterization of a silkworm thioredoxin peroxidase that is induced by
external temperature stimulus and viral infection ARTICLE
Pages 73-84
Kwang Sik Lee, Seong Ryul Kim, Nam Sook Park, Iksoo Kim, Pil Dong Kang,
Bong Hee Sohn, Kwang Ho Choi, Seok Woo Kang, Yeon Ho J e, Sang Mong Lee et
al.

9.

Profiling the proteome complement of the secretion from hypopharyngeal
gland of Africanized nurse-honeybees (Apis mellifera L.) SHORT
COMMUNICATION
Pages 85-91
Keity Souza Santos, Lucilene Delazari dos Santos, Maria Anita Mendes, Bibiana
Monson de Souza, Osmar Malaspina and Mario Sergio Palma
Copyright 2006 Elsevier Ltd. All rights reserved
Insect
Biochemistry
and
Molecular
Biology
Insect Biochemistry and Molecular Biology 35 (2005) 110
Sugar epitopes as potential universal disease transmission
blocking targets
Rhoel R. Dinglasan
a,
, Jesu s G. Valenzuela
b
, Abdu F. Azad
a
a
Department of Microbiology and Immunology, University of Maryland School of Medicine, 20 Penn Street, Baltimore, MD 21201, USA
b
Vector Molecular Biology Unit, Laboratory of Malaria and Vector Research, National Institutes of Allergy and Infectious Diseases,
National Institutes of Health, Rockville, MD 20852, USA
Received 2 June 2004; accepted 30 September 2004
Abstract
One promising method to prevent vector-borne diseases is through the use of transmission blocking vaccines (TBVs). However,
developing several anti-pathogen TBVs may be impractical. In this study, we have identied a conserved candidate carbohydrate
target in the midguts of several Arthropod vectors. A screen of the novel GlycoChip
s
glycan array found that the anti-carbohydrate
malaria transmission blocking monoclonal antibody (MG96) preferentially recognized D-mannose (a) and the type II lactosamine
disaccharide. The specicity for D-mannose was conrmed by competition ELISA using a-methyl mannoside as inhibitor. Con A,
which identies terminal mannose residues, did not inhibit MG96 reactivity with mosquito midgut lysates, suggesting that Con A
has differential recognition of this monosaccharide. However, the jack bean lectin, Jacalin, which recognizes D-mannose (a), D-
galactose (a=b) and the T antigen, not only displays a similar banding prole to that recognized by MG96 on immunoblot but was
also shown to effectively inhibit MG96. Wheat-germ agglutinin, which recognizes N-acetyllactosamine units, only partially inhibited
MG96 reactivity. This highlights the contribution of both glycan moieties to the MG96 epitope or glycotope. Enzyme
deglycosylation results suggest that MG96 recognizes a mannose a1-6 substitution on an O-linked oligosaccharide. Taken together,
the data suggest that MG96 recognizes a discontinuous glycotope composed of Mana1-6 proximal to Galb1-4GlcNAc-a-O-R
glycans on arthropod vector midguts. As such, these glycotopes may represent potential transmission blocking vaccine targets for a
wide range of vector-borne pathogens.
r 2004 Elsevier Ltd. All rights reserved.
Keywords: Carbohydrate; Transmission blocking vaccine; Mosquito; Midgut; Vector-borne disease
1. Introduction
To date, efforts at curbing diseases such as malaria,
dengue, African sleeping sickness and tick-borne fevers
have fallen short of their expected goals, and the
incidence of diseases that were not known to occur in
developed countries (e.g., West Nile) are rapidly on the
rise. In lieu of an effective anti-parasitic vaccine, vector
control seems to remain as the only promising control
method. Vector control programs, although simple to
initiate, are unfortunately difcult to sustain for many
developing countries, where the existing economic and
social conditions are far from optimal.
Although vaccines are still considered the best
approach to prevent the spread of vector-borne diseases
on the global scale, the notion of developing a vaccine
directly against every known human pathogen, although
ideal, is currently impractical. Another concept proposes
the development of a wide-spectrum vaccine that can
effectively decrease the transmission of several vector-
borne pathogens. The most suitable vaccine approach to
ARTICLE IN PRESS
www.elsevier.com/locate/ibmb
0965-1748/$ - see front matter r 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ibmb.2004.09.005

Corresponding author. Department of Molecular Microbiology


and Immunology, Johns Hopkins University Bloomberg School of
Public Health, 615 N. Wolfe Street, Room W4109, Baltimore, MD
21205, USA. Tel.: +443 287 0789; fax: +410 955 0105.
E-mail addresses: rdinglas@jhsph.edu (R.R. Dinglasan),
jvalenzuela@nih.gov (J.G. Valenzuela), aazad@umaryland.edu
(A.F. Azad).
meet this goal are transmission blocking vaccines
(TBVs), which are intended to block pathogen develop-
ment within the arthropod vectors (Carter, 2001). This
could be achieved by either targeting pathogen or
arthropod molecules as transmission blocking targets.
The problem lies with identifying an appropriate vaccine
candidate: the proverbial magic bullet. Convergent
evolution provides an opportunity to identify conserved
molecular targets resulting from evolved adaptations of
arthropod vectors to a dened niche: anthropophilic
hematophagy.
Carbohydrates have been shown to be essential
macromolecules for a plethora of cellular processes
such as signaling, structural support, protection and
trafcking (Roseman, 2001). Their critical role in
bacterial, parasitic and viral infectious disease models
further underscores their promise as vaccine targets
(Moe and Granoff, 2001; Schoeld et al., 2002;
Monzavi-Karbassi et al., 2003). In addition, studies
have outlined the critical role of carbohydrates in the
malaria parasite ookinete attachment and invasion
process and the ookineteoocyst transition stage within
the mosquito vector (Rudin and Hecker, 1989; Dessens
et al., 1999; Zieler et al., 1999). Thus, anti-carbohydrate
TBVs could stand up as the most viable system. The
proof of concept has been recently described for a
murine malaria model (Dinglasan et al., 2003). Mono-
clonal antibody (MAb) MG96 completely blocked
Plasmodium parasite development by targeting a carbo-
hydrate epitope or glycotope on mosquito midgut
microvilli. We report here, the identication of the
MG96 glycotope which is conserved across various
arthropod vectors. Thus, we propose targeting such
antigens in an effort to develop broad-based transmis-
sion blocking interventions against several vector-borne
pathogens.
2. Materials and methods
2.1. Biological material
The following arthropods were chosen since they are
the primary vectors of various human diseases. Triato-
ma pallidipennis midguts (male+female) were a kind gift
from Dr. Janine Ramsey at the Instituto Nacional de
Salud Publica in Cuernavaca, Mexico. Glossina morsi-
tans morsitans midguts (males only) were a kind gift
from Dr. Serap Aksoy at Yale University, New Haven,
CT. Ctenacephalides felis midguts (male and female)
came from a colony maintained at the University of
Maryland, Baltimore, originally from Flea Data, Inc.
(Freebill, NY). Dermacentor variabilis midguts
(females only) were from a colony maintained at Old
Dominion University. The dissected guts were trans-
ferred into Mammalian Protein Extraction Reagent,
M-PER (PIERCE, Rockford, IL), with the addition
of 2% ASCB8+ (Calbiochem, San Diego, CA)
and a 1:100 dilution of protease inhibitor cocktail
(Sigma, St. Louis, MO). To extract protein, the sample
was ground with a sterile pestle, subjected to repeated
freeze and thaw cycles and then spun down at
14,000 rpm at 4 1C. Protein concentration was deter-
mined by the bicinchoninic acid protein assay (PIERCE,
Rockford, IL).
2.2. Monoclonal antibody
The monoclonal antibody (MAb) MG96, IgG1
subclass, was produced at the Laboratory of Malaria
and Vector Research, NIAID, NIH, Bethesda, MD (I.
Fields and M. Shahabuddin, unpublished data). Briey,
6-week-old female BALB/c mice (Charles River La-
boratories, Wilmington, MA) were immunized by
subcutaneous injection with Aedes aegypti midgut
extracts in RIBI adjuvant (RIBI Immunochem Re-
search, Hamilton, MT). A boost was given 21 d after
initial immunization and a nal boost immunization was
given intravenously 14 d after the second boost. Gen-
eration of hybridoma cells followed the polyethylene
glycol method with selection in hypoxanthineaminop-
terinthymidine (HAT) medium. The resulting hybrido-
ma were screened for midgut specic antibodies by IFA
using paraformaldehyde xed, Triton X-100 permeabi-
lized Ae. aegypti midgut sections. The specicity to
mosquito midgut was conrmed using Western blot
analysis with midgut extracts and extracts of total
carcasses. The isotype of MAb MG96 was determined
using a mouse monoclonal typing kit (PIERCE, Rock-
ford, IL) following the instructions provided by the
manufacturer.
2.3. Immunoblot analysis
Lysates were separated by SDSpolyacrylamide gel
electrophoresis (PAGE) and transferred to polyvinyli-
dene diuoride (PVDF) membranes. Blots were blocked
with blocking buffer (1 PBS. 0.1% Tween-20, 0.1%
BSA) probed with MG96 antibodies, from ascites, at a
1:1000 dilution. Bound antibodies were detected by an
alkaline phosphatase-conjugated goat anti-mouse IgG
antibody (Invitrogen, Carlsbad, CA), followed by a
chemiluminescent reaction with the addition of CDP-
star substrate (Invitrogen, Carlsbad, CA). Immunoblot
analysis of midgut lysates with uorescein labeled lectins
were performed as above, with some modications:
blocking and washing buffer was composed of 1 Tris-
buffered saline+0.1% Tween-20 (TBST), and bound
lectins were detected by alkaline phosphatase-conju-
gated mouse-anti-FITC IgG antibody followed by
chromogenic detection with the addition of Western
Blue Substrate (Promega, Madison, WI). The lectins
ARTICLE IN PRESS
R.R. Dinglasan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 110 2
that were used in this study were Canavalia ensiformis
agglutinin (Con A), which recognizes terminal
mannose-a, a-glucose and GlcNAC (Becker et al.,
1976; Bouckaert et al., 1999); Jacalin (JAC) from
Atrocarpus integrifolia (jack bean), which recognizes
mannose (a), Gal (a), GalNAc (a) (Hagiwara et al.,
1988; Kabir, 1998; Bourne et al., 2002; Wu et al., 2003);
Triticum vulgaris or wheat-germ agglutinin (WGA),
which recognizes lactosamine type II, GlcNAc (b) and
sialic acid (Wu et al., 1998); and Sambuca nigra
agglutinin (SNA), which recognizes Neu5Aca1-6Gal
(b) (Shibuya et al., 1987).
2.4. Glycochip array
Arrays were purchased from Glycominds (Lod,
Israel) in a standard 96-well black 200 ml microwell
plate format. Each array is a panel of 46 mono- and
oligosaccharides covalently linked to a substrate matrix
by a exible linker. This enables and ensures that the
glycans maintain correctly dened orientations in
solution. Each glycan is spotted in duplicate including
two control wells: biotin and -OH. MAb MG96 (in 1
PBSBSA) was distributed to each well in 100 ml
volumes at a concentration of 234 ng/ml and allowed
to incubate for 1 h at 37 1C in a humidifying chamber.
The plates were washed four times in 1 PBS, 0.05%
Tween-20 and 0.5 M NaCl, with an additional 5 min
soak following the nal wash. Horseradish peroxidase
(HRP)-conjugated goat anti-mouse IgG antibody at
1:1000 dilution was then incubated in each well (100 ml
total volume) for 1 h at 37 1C in a humidifying chamber.
The plates were washed four times as before. Fluores-
cence detection of bound secondary antibody using the
Amplex Red ELISA Kit#1 (Molecular Probes, Eugene,
OR) was accomplished as per the manufacturers
instructions. Excitation was set at 530560 nm and
emission was detected at 590 nm using a VICTOR
Wallac 1420 multilabel reader (Perkin-Elmer, Wellesley,
MA). Statistical signicance was determined using the
Students t-test.
2.5. Competitive inhibition assays
Two sets of competitive inhibition ELISA assays were
performed using a panel of lectins and monosaccharides.
For lectin competition, black Dynex microuor 1
detection plates (Dynex Technologies, West Sussex,
UK) were coated with 10 mg/ml of Anopheles stephensi
midgut lysates. Positive controls were wells that were
spotted with serial dilutions of antibody (as standards).
The detection plates were blocked with TBST for 2 h at
25 1C. Serial dilutions of lectin were incubated in each
well for 1 h in a humidifying chamber at room
temperature. The plate was washed twice with TBST
prior to incubation of MG96 antibody (118 ng/ml) for
1 h at 25 1C in a humidier. Following incubation and
washing, HRP-conjugated goat anti-mouse IgG (H+L)
(Molecular Probes, Eugene, OR) was added,
incubated for 1 h in a humidifying chamber at room
temperature and then washed. The Amplex Red
substrate (Molecular Probes, Eugene, OR) was used
for uorescence detection. Excitation was set at
530560 nm and emission detected at 590 nm using a
VICTOR Wallac 1420 multi-label reader (Perkin-Elmer,
Wellesley, MA). For monosaccharide inhibition, serial
dilutions of D-glucose, a-methyl mannoside, D-galactose
or L-arabinose (Sigma, St. Louis, MO) were used as
inhibitors. MG96 (118 ng/ml) were incubated with the
monosaccharide dilutions in a microcentrifuge tube for
1 h on a shaker at room temperature. The solutions were
then transferred to the detection plate (described
above) and allowed to incubate for 1 h in a humidifying
chamber at room temperature. Following washing,
free antibody from the inhibition steps was measured
by the use of the Amplex Red substrate as described
previously.
2.6. Enzymatic deglycosylation
Midgut lysates (20 mg) were boiled at 95 1C for 10 min
and then treated with either 0.5 mU PNGase F (E.C.
3.5.1.52) or 1 mU of a1-(2,3) mannosidase (E.C.
3.2.1.77) and a1-6-mannosidase (E.C. 3.2.1.101) in
50 mM sodium citrate, 100 mg/ml BSA (New England
Biolabs, Beverly, MA). Control reactions were prepared
without the enzyme. The reactions proceeded overnight
in the respective deglycosylation buffers at 37 1C.
Deglycosylated proteins were run on SDSPAGE gels,
immunoblotted and probed with MG96 as described
previously.
3. Results and discussion
3.1. MG96 identies a conserved oligosaccharide in the
blood-fed midguts of other arthropod disease vectors
The ability of MG96 to recognize epitopes in the
midguts of hematophagous arthropods across different
sub-phyla was assayed by Western blot. Although each
of the respective arthropod midguts harbored a distinct
binding prole when probed with MG96, the MAb
recognized a consistent band migrating at 7580 kDa,
originally identied for Anopheles mosquitoes (Dingla-
san et al., 2003). The band intensity was found to be
most prominent for Dermacentor variabilis bloodfed
midguts and less intensely so for unfed tick guts. The
weakest signal for this band was observed for Ctenace-
phalides felis and Glossina morsitans morsitans bloodfed
guts. The greatest overall distribution of protein bands
was found for Triatoma pallidipennis bloodfed guts
ARTICLE IN PRESS
R.R. Dinglasan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 110 3
(Fig. 1). An Anopheles cell line, ASEIV, previously
suspected of being of gut origin (Kurtti and Munderloh,
1989; Shih et al., 1998; Fallon and Sun, 2001), did not
display the same banding prole as that of Anopheles
gambiae and An. stephensi midgut lysates (data not
shown). However, the 7580 kDa band was likewise
recognized.
The different banding proles observed across the
various vectors suggests that the glycotope is present on
a wide array of glycoproteins. The evolutionary and
functional relationships for these glycoproteins have yet
to be determined, especially since the conservation of the
glycotope alone cannot ascribe any functionality to the
protein backbone. It is known that the presence of this
oligosaccharide epitope is not sex-specic nor does it
necessarily have a role in blood feeding since it has been
shown to be present in adult male mosquitoes, larvae
and pupae (Dinglasan et al., 2003). Moreover, for some
of the vectors that were screened, hematophagy is not
limited to one sex alone. Additionally, band diversity
may correlate with the vectors ability to use alternate
nutritional sources. Ticks, eas and tsetse are obligatory
blood feeders and the relative simplicity of their banding
proles diverges from the proles observed for
mosquitoes and Triatomine bugs, which are optional
blood feeders (Beaty and Marquardt, 1996). Different
from other hemimetabolists, T. pallidipennis can take
both sugar and bloodmeals through their piercing
mouthparts at all stages of development. Triatomine
midgut epithelia may be more heavily glycosylated
due to the tissues acidic environment (Beaty and
Marquardt, 1996), in comparison to other hematopha-
gous arthropods.
3.2. MG96 recognizes a related set of mono- and
oligosaccharides on a 96-well glycan array
The glycan binding prole for MAb MG96 was
analyzed by applying a sample of the antibody (234 ng/
ml) to 46 mono- and oligosaccharides (Table 1)
expressed in condensed form, following the rules of
the new linear syntax for carbohydrates (Banin et al.,
2002). Although the antibody binding prole exhibited
wide cross-reactivity, only a few of the glycans showed
statistically signicant binding (po0:05) (Fig. 2). The
highest antibody binding titer was specic for mannose
moieties in a-linkage (100 ng/ml). b-Mannose residues
were not recognized by MG96. Interestingly, none of the
oligomannosides, Man (a1-3) Man (Table 1, #23), and
Man(a1-3)Man(a1-6)Man (b) (Table 1, #17), that were
present in the array were recognized. a-Glc, a-GlcNAc,
a-Gal and b-Gal monosaccharides were recognized as
well, albeit to a lesser extent.
The a-Gal and b-Gal moieties were recognized on
only two dened oligosaccharide structures presented on
the array: N-acetyllactosamine (Table 1, #12) and as the
dissacharide, Galb1-6Gal (b) (Table 1, #8), respectively.
Type II lactosamine units are present on N-glycans, O-
glycans and glycolipids. In addition to being core
structures for the ABO blood group antigens, lactosa-
mines were found to be primary components of mucins
covering epithelial cells on tissues found in both
vertebrates and invertebrates (Watkins, 1980; Kramerov
et al., 1996; Shen et al., 1999). The Galb1-6Gal (b)
substitution was rst described as a component of the
glycosphingolipid oligosaccharide GalSEGLx from the
adult Spirometra erinacei tapeworm (Kawakami et al.,
1996).
MG96 also showed recognition specicity for b-xylose
and a-L-rhamnose residues. b-Xyl is a component of
plant cell walls or xylan, and has been described to be
present in mammals and insects (Wilson, 2002). This
monosaccharide is important for the chain initiation of
proteoglycans, and recognition of this core moiety
suggests that without steric hindrance from proximal
oligosaccharide chains, MG96 can recognize Xyl-O-Ser
linkage on mosquito midgut microvilli. Evidence in
support of this has been described for Drosophila, where
anti-b1-2 xylose antibodies were found to be reactive
with neuronal proteins of the y (Varki et al., 1999). a-L-
Rham is an essential saccharide component of the O-
polysaccharide on bacterial cell walls for both GRAM-
positive and GRAM-negative bacteria. One study
showed that circulating anti-glycan antibodies from a
nave pool of individuals showed strong recognition of
a-L-Rham on a glycan array (Schwarz et al., 2003).
Therefore, it may not be uncommon for antibodies to be
generated against a-L-Rham, since bacterial contamina-
tion may have occurred during the preparation of the
immunogen.
ARTICLE IN PRESS
Fig. 1. Immunoblot analysis of blood-fed and unfed midgut lysates
from (1) Triatoma pallidipennis, blood-fed females, (2) Dermacentor
variabilis, unfed, (3) D. variabilis, blood-fed, (4) An. stephensi ASEIV
cell line 1, (5) Ctenacephalides felis, blood-fed females, (6) Glossina
morsitans morsitans, blood-fed males and (7) An. gambiae. MW ladder,
L, in kDa. 10% tris-glycine gel. Protein load: 10 mg/well.
R.R. Dinglasan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 110 4
The observation that MG96 has putative multivalent
interactions for several related monosaccharides helps
us further interpret our previous immunohistochemistry
results (Dinglasan et al., 2003). We had observed
intense, ubiquitous staining of the microvilli of mosqui-
to midgut cryosections. This suggested two possibilities.
First, that the carbohydrate epitope, or glycotope, may
be loosely dened, allowing different saccharide sub-
stitutions on core precursors and branching chains. A
loose t within the antibody combining site may allow
for wider, low-afnity binding of structurally similar
glycotopes on different microvillar glycoproteins, which
may only differ by one or more monosaccharides.
Second, that staining is a result of highly specic
recognition of a dened glycotope that is a common
component of both simple and complex N- and O-linked
glycans. It is unclear which of the two possibilities is
correct since the glycan array did not identify a single,
unambiguous oligosaccharide conguration.
The multivalent interactions with different monosac-
charides with varying specicities and afnities also shed
light on the mechanism by which MAb MG96 induces
transmission blocking immunity against Plasmodium
parasites in mosquitoes. MG96 cross-reactivity with
different glycans, which contributes to the complete
staining of the microvilli, may mask all potential
adhesion sites for invading Plasmodium ookinetes,
protecting the mosquito from parasite infection. How-
ever, it is also possible that only one subset of
oligosaccharides bearing the key sugar components,
e.g., mannose (a) or N-acetyllactosamine, and charac-
terized by high MAb binding afnity, are protective.
3.3. MG96 has binding specicity for mannose in a-
linkage
To conrm the binding specicity of MG96 for the
various monosaccharides identied from the glycan
array, binding afnity was indirectly measured by
competitive inhibition ELISA (Fig. 3a). a-Methyl-
mannoside competitively inhibited MG96 reactivity
with IC
50
values between 10 and 50 mM concentrations.
It has been shown that 200 mM of a-methyl-mannoside
can completely inhibit Con A binding to solid phase
antigen (Bouckaert et al., 1999). This suggests that
although both MG96 and Con A have recognition
specicity for this monosaccharide their respective
molecular interactions with a-mannose are quite differ-
ent. D-glucose, a stereoisomer of D-mannose, and D-
galactose, which were recognized on the glycan array,
failed to inhibit MG96 reactivity above that of the
control monosaccharide, L-arabinose, which is the
pentose precursor for both glucose and mannose.
It is difcult to determine from these data if binding
afnity alone is the best criterion for identifying
protective midgut epitopes. However, it is clear that
ARTICLE IN PRESS
Table 1
Mono- and oligosaccharides spotted on a 96-well GlycoChip
carbohydrate array
Glycan
no.
Glycan Linear syntax
1 Gal (a) Aa
a
2 Gal (b1-3) GlcNAc (b) Ab3GNb
3 L-Fucose (b) Fb
a
4 GlcNAc (Sulf) (b) GN(6S)b
5 Galacturonic Acid, GalA b) Lb
6 Neu5Ac (a2-3) Glc (b1-4)
GlcNAc (b)
NNa3Gb4GNb
7 Gal (a1-3) Fuc (a1-2) Gal b) Aa3(Fa2)Ab
8 Gal (b1-6) Gal (b) Ab6Ab
9 Glc (a) Ga
a
10 GlcNAc (a) GNa
a
11 Mannose (a) Ma
a
12 Neu5Ac (a1-3)Gal (b1-4)
GlcNac (b)
NNa3Ab4GNb
a
13 Glc (b1-4) Glc (b1-4) Glc (b1-
4) Glc (b1-4) Glc (b)
Gb4Gb4Gb4Gb4Gb-
pNP
14 GalNac (a) Ana
a
15 Glc (aGlc (a) Ga4Ga
16 GlcNac (b) GNb
a
17 Man (a1-3) Man (a1-6) Man
(b)
Ma3(Ma6)Mb
a
18 Neu5Ac (a2-6) Gal (b1-4)
GlcNAc (b)
NNa6Ab4GNb
19 Gal (b) Ab
a
20 GalNAc (a1-3) Fuc (a1-2) Gal
(b)
ANa3(Fa2)Ab
21 Glc (a1-4) Glc (b) Ga4Gb
22 GlcNAc (b1-4) Gal (b1-4) Glc
(b)
GNb4Ab4Gb
23 Man (a1-3) Man Ma3Ma
a
24 OH-blank Control I
25 Gal (b1-3) GlcNAc (b 1-6)
GalNAc (a)
Ab3(GNb6)Ana
26 GalNac (b ANb
a
27 Glc (b) Gb
a
28 GlcNAc (b 1-3) GalNAc (a) GNb3Ana
29 Mannose (b) Mb
a
30 L-Araf (a) Ra
31 Gal (b 1-3) GalNAc (a) Ab3Ana
a
32 Biotin Control II
33 Glc (b 1-3) Glc (b) Gb3Gb
34 GlcNAc (b1-4) GlcNAc (b) GNb4Gnb
a
35 Man (b 1-4) Glc (b) Mb4Gb
a
36 Glucoronic acid (b) Ub
37 Gal (b1-3) GlcNAc (b) Ab3GNb
a
38 L-Fucose (a) Fa
a
39 Glc (b 1-4) Glc (b) Gb4Gb
40 GlcNAc (b 1-6) GalNAc (a) GNb6Ana
41 Neu5Ac (a) NNa
42 Xylose (a) Xa
a
43 Gal (b 1-4) Glc (b) Ab4Gb
44 Fuc (a1-2) Gal (b) Fa2Ab
45 Glc (b 1-4) Glc (b 1-4) Glc (b) Gb4Gb4Gb
46 L-Rhamnose (a) Ha
47 Neu5Ac (a2-3)Gal (b 1-4)[Fuc
(a1-3)]GlcNac (b)
NNa3Ab4(Fa3)GNb
48 Xylose (b) Xb
a
a
Indicates glycans that have been characterized in insects. See text
for details and Seppo and Tiemeyer (2000).
R.R. Dinglasan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 110 5
mannose (a) residues are key target monosaccharides
and principal components of the MG96 epitope. The
remaining monosaccharides from the glycan array may
therefore constitute those moieties abutting mannose (a)
residues and contributing to the real glycotope, e.g., N-
acetyllactosamine, or antigens that are deeply buried
components of larger oligosaccharide structures and
only available for binding in the array. The question
remains, How are the key monosaccharides linked
together in sequence? Two methods can be used to
deduce the sugar linkages. The rst involves the use of
lectins to competitively inhibit MG96 reactivity with its
cognate epitope. Fine lectin specicity for distinct
oligosaccharide linkage congurations may help us
identify the MG96 epitope. The second method involves
conrming these congurations through the use of
enzyme deglycosidases with dened linkage specicities.
3.4. MG96 recognizes carbohydrate moieties in a lectin-
like manner
To determine the oligosaccharide structure specicity
of MG96, the following lectins were used: Canavalia
ensiformis agglutinin (Con A), which recognizes term-
inal Man (a), a-Glc and GlcNAc; Jacalin (JAC) from
Atrocarpus integrifolia (jack bean), which recognizes
Man (a), Gal (a), GalNac (a) and Galb1-3GalNAc (type
I lactosamine); wheat-germ agglutinin (WGA), which
recognizes GlcNAc (b), Neu5Ac (a) and Galb1-
4GlcNAc (a=b) (type II lactosamine); and Sambuca
nigra agglutinin (SNA), which recognizes Neu5Aca(1-
6)Gal(b). These lectins were chosen not only because
they share the same monosaccharide specicities
identied from the glycan array, but in addition they
have also been shown to block the chicken malaria
parasite, P. gallinaceum ookinete attachment to splayed
Aedes aegypti midguts ex vivo (Zieler et al., 2000).
Immunoblot analysis of An. stephensi midgut lysates
with Con A conrmed previous ndings of the
ubiquitous nature of mannose residues in the midgut
(Shen et al., 1999; Wilkins and Billingsley, 2001)
(Fig. 4a). SNA was originally found to be incapable of
recognizing microvilli lysates taken from An. stephensi
midguts on Western blots (Wilkins and Billingsley,
2001), although it had strongly stained the midgut
lumen of Ae. aegypti (Zieler et al., 2000). When used to
probe transblotted An. stephensi whole midgut lysates,
SNA showed slight reactivity, identifying a band that
migrated at approximately 80 kDa and, to an even lesser
extent, two more bands 4100 kDa. WGA recognized
three predominant bands on An. stephensi midgut
lysates at 150, 100 and around 40 kDa. The 150
and 40 kDa bands matched quite well with the
corresponding bands of comparable MW recognized
by MG96. JAC exhibited the most remarkable similarity
in terms of band recognition with MG96. JAC and
MG96 shared at least three predominant bands, at 150,
7580 and around 40 kDa.
To deduce the possible oligosaccharides and linkage
specicities of the MG96 glycotope, the lectins were
assessed in their ability to competitively inhibit MG96
reactivity by ELISA (Fig. 3b). In line with the
observation of banding prole similarity, JAC competi-
tively inhibited MG96 in its ability to recognize its
cognate epitope with an IC
50
100 ng=ml: WGA, on
the other hand, failed to inhibit MG96 reactivity above
80% at concentrations above 100 mg/ml. Interestingly,
Con A exhibited no inhibitory capacity.
Con A recognizes terminal mannose residues. Its
inability to inhibit MG96, which shows high specicity
ARTICLE IN PRESS
Fig. 2. Monoclonal antibody MG96 anti-glycan binding prole. Error bars represent 1 s.d around the mean. The asterisk (*) denotes statistical
signicance at po0:05:
R.R. Dinglasan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 110 6
for mannose residues, suggests that the recognition
events of these two proteins for the carbohydrate may be
quite different. Although it is a possibility that both
recognize different aspects of the same terminal man-
nose residues, this is an unlikely event. It has been
shown that complex formation of mannose with Con A
imparts a conformational change, thereby further
sequestering the residue (Bouckaert et al., 1999). It is
more likely that MG96 recognizes internal mannose
residues and that binding improves in the presence of
proximal, essential components, such as N-acetyllacto-
samine. Non-terminal mannose residues occur along
growing linear polymers, at the initiation of antennary
branches or directly O-linked to the polypeptide back-
bone (Varki et al., 1999). Mannose in a-linkage
commonly occurs at the initiation of antennary
branches on both N- and O-glycans, and the specicity
of the a-linkage of mannose residues recognized by
ARTICLE IN PRESS
0
10
20
30
40
50
60
70
80
90
100
400 100 25 12 6.25 1.5 0.4 0.1 0.025
Inhibitor (ug/ml)
%

I
n
h
i
b
i
t
i
o
n
%

I
n
h
i
b
i
t
i
o
n
Jacalin WGA ConA
0
10
20
30
40
50
60
70
80
90
100
1000 100 10 1 0.1 0.01 0.001 0.0001
Inhibitor (mM)
Mannose Glucose Arabinose Galactose
(a)
(b)
Fig. 3. Monosaccharide (a) and lectin inhibition (b) of MG96 reactivity for its cognate epitope on mosquito midguts.
R.R. Dinglasan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 110 7
MG96 is described further below. It is also possible that
the detection of MG96 may not be a result of
recognition of its cognate glycotope in the presence of
Con A, but is rather an artifact of the assay conditions.
Con A can bind to exposed mannose residues that are
conserved on immunoglobulin molecules (Wright and
Morrison, 1998). Therefore, MG96 presence in the
ELISA plate may be a result of Con A sequestration of
the antibody itself. However, a direct-binding ELISA
using FITC-labeled Con A as probe and either MG96 or
the HRP-conjugated secondary antibody as antigen
failed to show any reactivity above background. More-
over, we also showed that JAC, which recognizes several
residues present on IgG N-linked oligosaccharides
(Kabir et al., 1995), competitively inhibited MG96
recognition of its cognate glycotope. As such, these
facts would argue against a detection artifact of non-
specic lectin-mediated MG96 binding in our compe-
tition ELISAs.
JAC has been described to recognize Man (a), Gal (a),
GalNAc (a) as well as lactosamine type I and II
disaccharides with varying afnity (Sastry et al., 1986;
Bourne et al., 2002; Wu et al., 2003). WGA likewise
recognizes lactosamine units, however with greater
afnity for N-acetyllactosamine. However, GlcNAc
residues were found to be more predominant in Ae.
aegypti mosquito midgut glycoproteins (Rudin and
Hecker, 1989; Zieler et al., 1999), whereas GalNAc
residues predominate in An. stephensi midguts. Since
MG96 was originally raised against Ae. aegypti midgut,
the preferred identication of the GlcNAc saccharide
over GalNAc on the glycan array is not surprising.
However, this recognition may be of low afnity,
allowing for a loose-t binding of both GlcNAc and
GalNAc in a-linkage with other moieties.
3.5. MG96 recognizes mannose residues in a1-6 linkage
Treatment of An. stephensi midgut lysates with
mannosidase a1-(2,3) did not affect MG96 binding
(data not shown). However, treatment with mannosi-
dase a1-6 abolished recognition of the 80 kDa band
leaving detection of the other bands at 150 and 40 kDa
intact (Fig. 4b). PNGase F treatment produced a
secondary band 2025 kDa but in general, did not
alter the characteristic MG96 banding prole (Dingla-
san et al., 2003). These data suggest three scenarios: (i)
that the a1-(2,3,6) bonds of internal mannose residues
were cryptic in solution and only become available
following gel electrophoresis; (ii) partial recognition in
addition to loose-t binding of the epitope may occur,
thereby further confounding these results; and (iii) the
oligosaccharide is found only on O-linked glycans.
The likelihood of recognition of both types I and II
lactosamine structures in midgut lysates, compounded
with deglycosylation inefciencies, makes accurate
interpretation of the results difcult. The clearest picture
of the saccharide linkage is observed for the 80-kDa
O-linked glycoprotein which is presumably conserved
across different arthropod vectors. Note that our
previous attempts to remove O-linked sugars on An.
stephensi midguts using O-glycosidase solely removed
recognition of the lowest MW band identied by MG96
(Dinglasan et al., 2003). This was presumably due to
inappropriate bond specicity of the currently available
O-glycanase enzyme. Unlike the number of enzymes
that are available for N-link glycan analysis, O-link
specic enzymes are limited. The O-glycanase that is
available has a specicity restricted to Galb1-3GalNAc
disaccharides only (Dwek et al., 1993; Edge, 2003).
Consequently, oligosaccharides that are O-linked via
GlcNAc, mannose or xylose, for example, will be
resistant to enzymatic cleavage. Treatment of midgut
lysates with either PNGase A (Dinglasan et al., 2003) or
PNGase F lent support to the likelihood that the MG96
glycan epitopes are found on O-linked sugars.
The results of this study suggest that the MG96
recognition of its glycotope has three possible models.
The simplest model suggests that MG96 directly
recognizes the hypothesized oligosaccharide sequence,
Mana1-6Manb1-3Galb1-4GlcNAc, as part of a larger
oligosaccharide structure. Another model proposes a
more discontinuous glycotope, wherein the N-acetyllac-
tosamine unit and the mannose (a) residue, in a1-6
linkage to adjacent moieties, are on opposite branches of
a biantennary oligosaccharide. The last possibility is
that MG96 recognizes a cluster of several different
oligosaccharides producing a drastically discontinuous
glycotope. Such a model, termed the clustered sacchar-
ide patch (Varki, 1994) has been proposed for selectin
binding. Clearly, the solution of the crystal structure of
the antibodycarbohydrate complex will help determine
ARTICLE IN PRESS
Fig. 4. (a) Lectin-immunoblot analysis of An. stephensi midguts
(ASMG). Sambuca nigra agglutinin (SNA), wheat germ agglutinin
(WGA), Jacalin, JAC, and Canavalia ensiformis (Con A). Arrows
indicate bands which share similar mobilities with the primary bands
recognized by MG96 on ASMG lysates. (b) Enzymatic deglycosylation
of An. stephensi midguts mannosidase a1-6 and PNGase F. (+), with
enzyme, (), control group. Arrow designates the 80 kDa band.
R.R. Dinglasan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 110 8
which of these models provide the better t. It may also
be the case that any of these models of recognition are
possible at a given binding event, especially if the
looseness of t argument holds true.
4. Conclusion
We report on the identication of a conserved
glycotope across different arthropod disease vectors.
Although each pathogen utilizes different mechanisms
to attach and invade the arthropod vector host, the use
of carbohydrates as receptor molecules is universal. The
malaria model, as an example, accentuates the impor-
tance of this glycotope in Plasmodium parasite attach-
ment to the midgut of mosquitoes. A phenomenon
independently conrmed by the use of both antibodies
and lectins in extremely diverse assays. With some
optimism, the use of the MAb MG96 in the Anaplasma-
tick infection model is currently being investigated.
Undoubtedly, the results from these experiments will
help further dene the plausibility of the universal
transmission blocking vaccine strategy.
Acknowledgments
This study was supported by grants from the NIAID/
NIH.
References
Banin, E., Neuberger, Y., Altshuler, Y., Halevi, A., Inbar, O., Nir, D.,
Dukler, A., 2002. A novel linear code nomenclature for complex
carbohydrates. Trends Glycosci. Glycotechnol 14, 127137.
Beaty, B.J., Marquardt, W.C. (Eds.), 1996, The Biology of Disease
Vectors. University Press of Colorado, Niwot.
Becker, J.W., Reeke Jr., G.N., Cunningham, B.A., Edelman, G.M.,
1976. New evidence on the location of the saccharide-binding site
of concanavalin A. Nature 259, 406409.
Bouckaert, J., Hamelryck, T.W., Wyns, L., Loris, R., 1999. The crystal
structures of Man(alpha1-3)Man(alpha1-O)Me and Man(alpha1-
6)Man(alpha1-O)Me in complex with concanavalin A. J. Biol.
Chem. 274, 2918829195.
Bourne, Y., Astoul, C.H., Zamboni, V., Peumans, W.J., Menu-
Bouaouiche, L., Van Damme, E.J., Barre, A., Rouge, P., 2002.
Structural basis for the unusual carbohydrate-binding specicity of
jacalin towards galactose and mannose. Biochem. J. 364, 173180.
Carter, R., 2001. Transmission blocking malaria vaccines. Vaccine 19,
23092314.
Dessens, J.T., Beetsma, A.L., Dimopoulos, G., Wengelnik, K.,
Crisanti, A., Kafatos, F.C., Sinden, R.E., 1999. CTRP is essential
for mosquito infection by malaria ookinetes. Embo J. 18,
62216227.
Dinglasan, R.R., Fields, I., Shahabuddin, M., Azad, A.F., Sacci Jr.,
J.B., 2003. Monoclonal antibody MG96 completely blocks
Plasmodium yoelii development in Anopheles stephensi. Infect.
Immun. 71, 69957001.
Dwek, R.A., Edge, C.J., Harvery, D.J., Wormald, M.R., Parekh, R.B.,
1993. Analysis of glycoprotein-associated oligosaccharides. Annu.
Rev. Biochem. 62, 65100.
Edge, A.S.B., 2003. Deglycosylation of glycoproteins with triuor-
omethanesulfphonic acid: elucidation of molecular structure and
function. Biochem. J. 376, 339350.
Fallon, A.M., Sun, D., 2001. Exploration of mosquito immunity using
cells in culture. Insect Biochem. Mol. Biol. 31, 263278.
Hagiwara, K., Collet-Cassart, D., Kobayashi, K., Vaerman, J.P., 1988.
Jacalin: isolation, characterization, and inuence of various factors
on its interaction with human IgA1, as assessed by precipitation
and latex agglutination. Mol. Immunol. 25, 6983.
Kabir, S., 1998. Jacalin: a jackfruit (Artocarpus heterophyllus) seed-
derived lectin of versatile applications in immunobiological
research. J. Immunol. Methods 212, 193211.
Kabir, S., Ahmed, I.S., Daar, A.S., 1995. The binding of
jacalin with rabbit immunoglobulin G. Immunol. Invest. 24,
725735.
Kawakami, Y., Nakamura, K., Kojima, H., Suzuki, M., Inagaki, F.,
Suzuki, A., Ikuta, J., Uchida, A., Murata, Y., Tamai, Y., 1996. A
novel fucosyltetrahexosylceramide in plerocercoids of the parasite
Spirometra erinacei. Eur. J. Biochem. 239, 905911.
Kramerov, A.A., Arbatsky, N.P., Rozovsky, Y.M., Mikhaleva, E.A.,
Polesskaya, O.O., Gvozdev, V.A., Shibaev, V.N., 1996. Mucin-type
glycoprotein from Drosophila melanogaster embryonic cells:
characterization of carbohydrate component. FEBS Lett. 378,
213218.
Kurtti, T.J., Munderloh, U.G., 1989. Advances in the denition of
culture media for mosquito cells. In: Mitsusashi, J. (Ed.),
Invertebrate Cell System Applications. CRC Press, Inc., Boca
Raton, FL, pp. 2129.
Moe, G.R., Granoff, D.M., 2001. Molecular mimetics of Neisseria
meningitidis serogroup B polysaccharide. Intern. Rev. Immunol.
20, 201220.
Monzavi-Karbassi, B., Shamloo, S., Kieber-Emmons, M., Joushe-
ghany, F., Luo, P., Lin, K.Y., Cunto-Amesty, G., Weiner, D.B.,
Kieber-Emmons, T., 2003. Priming characteristics of peptide
mimotopes of carbohydrate antigens. Vaccine 21, 753760.
Roseman, S., 2001. Reections on glycobiology. J. Biol. Chem. 276,
4152741542.
Rudin, W., Hecker, H., 1989. Lectin-binding sites in the midgut of the
mosquitoes Anopheles stephensi Liston and Aedes aegypti L.
(Diptera: Culicidae). Parasitol. Res. 75, 268279.
Sastry, M.V., Banarjee, P., Patanjali, S.R., Swarny, M.J., Swarnalatha,
G.V., Surolia, A., 1986. Analysis of saccharide binding to
Artocaprus integrifolia lectin reveals specic recognition of T-
antigen (b-D-Gal (1-3)D-GalNAc). J. Biol. Chem. 261,
1172611733.
Schoeld, L., Hewitt, M.C., Evans, K., Siomos, M.A., Seeberger, P.H.,
2002. Synthetic GPI as a candidate anti-toxic vaccine in a model of
malaria. Nature 418, 785789.
Schwarz, M., Spector, L., Gargir, A., Shtevi, A., Gortler, M., Altstock,
R.T., Dukler, A.A., Dotan, N., 2003. A new kind of carbohydrate
array, its use for proling antiglycan antibodies, and the discovery
of a novel human cellulose-binding antibody. Glycobiology 13,
749754.
Seppo, A., Tiemeyer, M., 2000. Function and structure of Drosophila
glycans. Glycobiology 10, 751760.
Shen, Z., Dimopoulos, G., Kafatos, F.C., Jacobs-Lorena, M., 1999. A
cell surface mucin specically expressed in the midgut of the
malaria mosquito Anopheles gambiae. Proc. Natl. Acad. Sci. (USA)
96, 56105615.
Shibuya, N., Goldstein, I.J., Broekaert, W.F., Nsimba-Lubaki, M.,
Peeters, B., Peumans, W.J., 1987. The elderberry (Sambucus nigra
L.) bark lectin recognizes the Neu5Ac(alpha 2-6)Gal/GalNAc
sequence. J. Biol. Chem. 262, 15961601.
ARTICLE IN PRESS
R.R. Dinglasan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 110 9
Shih, K.M., Gerenday, A., Fallon, A.M., 1998. Culture of mosquito
cells in Eagles medium. In Vitro Cell. Dev. Biol.-Animal pp.
629630.
Varki, A., 1994. Selectin ligands. Proc. Natl. Acad. Sci. (USA) 91,
73907397.
Varki, A., Cummings, R., Esko, J., Freeze, H., Hart, G., Marth, J.
(Eds.), 1999, Essentials of Glycobiology. Cold Spring Harbor
Laboratory Press, Cold Spring Harbor.
Watkins, W.M., 1980. Biochemistry and Genetics of the ABO, Lewis,
and P blood group systems. Adv., Hum. Genet. 10 (1136),
379385.
Wilkins, S., Billingsley, P.F., 2001. Partial characterization of
oligosaccharides expressed on midgut microvillar glycoproteins of
the mosquito, Anopheles stephensi Liston. Insect Biochem. Mol.
Biol. 31, 937948.
Wilson, I.B., 2002. Functional characterization of Drosophila mela-
nogaster peptide O-xylosyltransferase, the key enzyme for
proteoglycan chain initiation and member of the core 2/I
N-acetylglucosaminyltransferase family. J. Biol. Chem. 277,
2120721212.
Wright, A., Morrison, S.L., 1998. Effect of C2-associated
carbohydrate structure on Ig effector function: studies
with chimeric mouse-human IgG1 antibodies in glycosylation
mutants of Chinese hamster ovary cells. J. Immunol. 160,
33933402.
Wu, A.M., Wu, J.H., Song, S.C., Tsai, M.S., Herp, A., 1998. Studies
on the binding of wheat germ agglutinin (Triticum vulgaris) to O-
glycans. FEBS Lett. 440, 315319.
Wu, A.M., Wu, J.H., Lin, L.H., Lin, S.H., Liu, J.H., 2003. Binding
prole of Artocarpus integrifolia agglutinin (Jacalin). Life Sci. 72,
22852302.
Zieler, H., Nawrocki, J.P., Shahabuddin, M., 1999. Plasmodium
gallinaceum ookinetes adhere specically to the midgut epithelium
of Aedes aegypti by interaction with a carbohydrate ligand. J. Exp.
Biol. 202 (Pt 5), 485495.
Zieler, H., Garon, C.F., Fischer, E.R., Shahabuddin, M., 2000.
A tubular network associated with the brush-border sur-
face of the Aedes aegypti midgut: implications for pathogen
transmission by mosquitoes. J. Exp. Biol. 203 (Pt 10),
15991611.
ARTICLE IN PRESS
R.R. Dinglasan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 110 10
Insect
Biochemistry
and
Molecular
Biology
Insect Biochemistry and Molecular Biology 35 (2005) 1122
Two structurally different defensin genes, one of them encoding a
novel defensin isoform, are expressed in honeybee Apis mellifera
Jaroslav Klaudiny
a,
, S

tefan Albert
b
, Katar na Bachanova
a
,
Ja n Kopernicky
c
, Jozef S

imu th
a
a
Laboratory of Genetic Engineering, Institute of Chemistry, Slovak Academy of Sciences, Dubravska cesta 9, 84538 Bratislava, Slovakia
b
Institute of Medical Radiation and Cell Research, Versbacherstr. 5, D-97078 Wurzburg, Germany
c
Institute of Apiculture, Research Institute of Animal Production, Gasperkova 599, 03308 Liptovsky Hradok, Slovakia
Received 30 July 2004; received in revised form 17 September 2004; accepted 29 September 2004
Abstract
Two defensins showing high mutual similarity have previously been characterized in honeybee Apis mellifera: royalisin, a peptide
isolated from the royal jelly, and defensin, found in the hemolymph of bacterially infected bees. Here we show that both these
peptides are encoded by the same polymorphic gene, which we termed defensin1. Besides this gene, we identied an additional
defensin gene coding for a novel honeybee defensin designated defensin2. The pre-pro-peptide sequence of defensin 2 was inferred
from its cDNA. Mature defensin 2 peptide shows 55.8% identity with defensin 1. Sequences of genomic loci of the two defensin
genes revealed their different structure. Defensin1 possesses an exonintron structure unique among arthropoda defensin genes. Its
second intron splits exactly the common structural module of defensins from a short amidated C-terminal extension found only in
hymenopteran defensins. Transcription of defensin genes in some nurse honeybees tissues was studied by RTPCR. Both defensins
are expressed in heads and thoraces. Defensin1 but not defensin2 mRNA was detected in hyphopharyngeal, mandibular and thoracic
salivary glands. Immune response elements were identied by computer analysis of the promoter regions of defensin genes. Their
different representation in these genes reects presumably observed tissue-specic expression of defensins.
r 2004 Elsevier Ltd. All rights reserved.
Keywords: Antimicrobial peptides; cDNA; Gene; Apis mellifera; Royal jelly; Insect immunity; Honeybee diseases; Mass spectrometry
1. Introduction
Antimicrobial peptides are key elements of insect
innate immunity (Boman, 1995; Bulet et al., 1999;
Hoffmann et al., 1999). Most of them are produced
upon infection or injury in the fat body or hemocytes
and secreted subsequently into the hemolymph. In some
insects, local expression of the peptides has also been
reported [e.g. cuticle (Brey et al., 1993), midgut and
salivary glands of blood-sucking insects (Lehane et al.,
1997; Dimopoulos et al., 1998; Lowenberger et al.,
1999a)].
Defensins comprise a widespread family of cystein-rich
cationic antimicrobial peptides that act against a variety
of microorganisms and constitute the primary defense
system of most organisms (Raj and Dentino, 2002).
Insect defensins are 3651-amino-acid-long peptides
possessing sequence similarity which is the basis of their
common structure comprising an amino-terminal loop,
an a-helix and two antiparallel b-strands stabilized by
three disulde bridges (Hanzawa et al., 1990; Bonmatin
et al., 1992; Cornet et al., 1995). They are active against a
broad spectrum of Gram-positive bacteria, although
activity against Gram-negative bacteria and fungi has
also been reported (Hetru et al., 1998; Yamauchi, 2001).
ARTICLE IN PRESS
www.elsevier.com/locate/ibmb
0965-1748/$ - see front matter r 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ibmb.2004.09.007

Corresponding author. Tel.: +421 2 59410203; fax:


+421 2 59410222.
E-mail address: chemjakl@savba.sk (J. Klaudiny).
Defensins kill bacteria by permeabilizing their cytoplas-
mic membrane (Cociancich et al., 1993).
Defensin isoforms differing by one to several amino
acids have been identied from several insect species
(Dimarcq et al., 1998; Lowenberger, 2001; Yamauchi,
2001; Lopez et al., 2003). In some insects the number of
genes encoding defensin isoforms was determined. A
single defensin gene exists in Drosophila melanogaster
(Dimarcq et al., 1994). Two defensin genes encoding
three defensin proteins A, B and C (A and B represent
allelic variants) with small differences in one or two
amino acid residues were identied in mosquito Aedes
aegypti (Lowenberger et al., 1999b; Lowenberger, 2001).
Rhodnius prolixus, an insect vector of Chagas disease,
possesses three genes in the genome that show relatively
large differences (Lopez et al., 2003).
Honeybees respond to infection by synthesizing
several peptides representing a broad spectrum of
antimicrobial activity. In the hemolymph of honeybees
infected with Escherichia coli, four different types of
antimicrobial cationic peptides were identied: apidae-
cins (Casteels et al., 1989), abaecin (Casteels et al.,
1990), hymenoptaecin (Casteels et al., 1993) and
defensin (Casteels-Josson et al., 1994). Defensin is
produced as the last of the peptides but its activity
persists for up to 2 weeks after infection (Casteels, 1998).
The hemolymph defensin and its pre-pro-peptide form
were characterized as a cDNA induced in abdomens of
infected bees (Casteels-Josson et al., 1994). Another
defensin of honeybee, named royalisin, was isolated
from honeybee royal jelly (RJ) and characterized at the
peptide level (Fujiwara et al., 1990). Both honeybee
defensins consist of 51 amino acids. They differ in one
amino acid and in the order of two other amino acids
(based on the comparison of sequences present in
databases). The honeybee defensins as well as bumble-
bee defensin (Rees et al., 1997) are amidated and have
an extra stretch of 11 amino acids at their C-terminus
(Bulet et al., 1999), unlike defensins of other insects.
We have described an antifungal activity of honeybee
royalisin (B likova et al., 2001) and its antibacterial
activity against honeybee pathogen Paenibacillus larvae
larvae, which is the cause of a serious disease of
honeybee larvae, American foulbrood (B likova et al.,
2001; Bachanova et al., 2002). It was also shown that
royal jellies from different colonies can differ in the
amount of antibacterial peptide(s) (Bachanova et al.,
2002). To date, it is still not clear whether defensin and
royalisin are encoded by different genes or by the same
gene.
To address this question, we characterized honeybee
defensin genes, their cDNAs and determined the
molecular mass of royalisin. This led to the identica-
tion of a novel royalisin/defensin variants and to the
conclusion that these are products of the same
polymorphic defensin1 gene. Furthermore, we identied
a new defensin2 gene encoding a novel defensin isoform.
The expression of both defensin genes in some tissues
was analyzed, and promoter regulatory sequences,
which could be participating in the regulation of the
expression of defensins, were identied.
2. Materials and methods
2.1. Biological samples
Honeybees Apis mellifera carnica and RJ were
collected from the apiary of the Institute of Apiculture
in Liptovsky Hradok and a private apiary in Valaska
(Slovak Republic), respectively. Nurse honeybees of
dened age (10 days) were obtained from a queenright
visually healthy colony by color labeling of freshly
emerged workers (aged 018 h). After removing from
the hive, 20 labeled honeybees were frozen directly in
liquid nitrogen and stored at 70 1C. From a group of
another 12 bees, hypopharyngeal, mandibular and
thoracic salivary glands were dissected and stored in
RNAlater solution at 20 1C. RJ was collected from the
queen cells containing 23-day-old larvae.
2.2. PCR amplications
The amplications of cDNA fragments were per-
formed with the total cDNA preparation used for the
construction of the cDNA library derived from heads of
8-day-old nurse honeybees (Apis mellifera carnica)
(Klaudiny et al., 1994). Genomic DNAs were isolated
from the thoraces of individual honeybees (Fondrk
et al., 1993). The primers for defensin1 were
derived from the honeybee defensin cDNA (Casteels-
Josson et al., 1994): Def1f1 (5
0
-ATTAATAT-
GAAAATCTATTTTATTGTCGG-3
0
), Def1f2 (5
0
-
ATTAATATGGTAACTTGTGACCTTCTCTCATT-
C-3
0
) and Def1r1 (5
0
-CATCGTTGAATCTTCAT-
AATGGCAC-3
0
) or from the sequence of defensin1
gene: Def1f3 (5
0
-CTGTTCTGTTCTACTTTACAC-3
0
)
and Def1r2 (5
0
-TAGCAACCATAGCCATGAAG-3
0
)
(Fig. 3). Underlined parts of primers represent the
overhangs added to primers for subcloning of the
products into expression vectors (unpublished data).
Defensin2 primers were derived from the genomic
sequence: Def2f1 (5
0
-GCTTTGTGCTGGGAAA-
GATG-3
0
), Def2f2 ACGATGGACCGATTTACGAG-
3
0
) and Def2r (5
0
-TCATCCGTTCTGCAAGAAGC-3
0
)
(Fig. 4). A complete ORF and a part encoding the
mature peptide of royalisin cDNA were amplied with
Pfu Turbo DNA polymerase (Stratagene). PCR ampli-
cations were done in 50 ml volumes using 7 ng of total
cDNA, Pfu polymerase buffer and 1.25 U of the
polymerase under the following conditions: 94 1C for
ARTICLE IN PRESS
J. Klaudiny et al. / Insect Biochemistry and Molecular Biology 35 (2005) 1122 12
5 min; 35 amplication cycles of 94 1C for 50 s; 60 1C for
50 s; 72 1C for 60 s and extension at 72 1C for 10 min.
Other PCRs were performed with AmpliTaq poly-
merase (Perkin Elmer) in 30 ml volume using 0.1 mg of
DNA or 4 ng of total cDNA by a modied hot start
procedure. After initial denaturation for 5 min at 94 1C
and subsequent cooling to 85 1C, the addition of 1 U of
Taq polymerase PCR was performed in a commercial
buffer containing 1.5 mM MgCl
2,
0.33 mM primers and
0.1 mM dNTPs. 35 cycles (40 s at 94 1C, 50 s at 53 1C and
120 s at 72 1C) were used, followed by 10 min incubation
at 72 1C. PCR products were analyzed by agarose gel
electrophoresis.
2.3. Cloning and sequencing of PCR products
PCR products were puried by agarose gel electro-
phoresis using the QIAEX II kit (QIAGEN) and ligated
into the SmaI site of pBluescript KS (+) (Stratagene)
and pGEM-T Easy (Promega), respectively. After
transformation of E. coli XL1-Blue or DH5a; the
plasmid DNAs were isolated by the QIAGEN plasmid
kit and sequenced by the cycle sequencing method
(Prism Ready Reaction BigDye Terminator kit, Perkin
Elmer) on an ABI 373A automatic sequencer. Some
PCR products were sequenced directly after electro-
phoretic separation and purication from the gel.
2.4. Preparation of royalisin from royal jelly
Acidic extraction of the RJ, electrophoresis in native
discontinuous acidic polyacrylamide gel and a bacterial
growth-inhibition assay on polyacrylamide gel employ-
ing honeybee larvae pathogen P. l. larvae as test bacteria
were performed as described (Bachanova et al., 2002).
The gel slice containing the antibacterial band was cut
off, placed in a dialysis tube with MWCO 1 kDa
(SpectraPor 6, SERVA) and lled with 0.1 M acetic
acid adjusted with KOH to pH 4.5. Electro-elution of
the peptides from the gel was performed in an
electrolytic solution (0.017 M acetic acid adjusted to
pH 4.5). Finally, the royalisin preparation was concen-
trated on a Microsep microconcentrator with MWCO
1 kDa (Pall Gelman Sciences).
2.5. Mass spectrometry
The mass spectrum was measured on a matrix-
assisted laser desorption ionization time of ight mass
spectrometer BIFLEX (Bruker-Franzen, Germany). A
saturated solution of 3, 5-dimethoxy-4-hydroxycinnamic
acid in 30% acetonitril, 0.2% TFA was used as the
matrix. 0.5 ml of the matrix solution was mixed with
0.5 ml of the sample on the target and the droplet was
allowed to dry at ambient temperature. For external
calibration, somatostatin ([M+H]
+
at m/z 1638.9) was
used as a standard peptide.
2.6. Database searches and assembly of genomic
sequences
MEGABLAST program was used for searches in the
database of A. mellifera genomic sequences at the
National Center for Biotechnology Information (Hon-
eybee Genome Project database at NCBI) (http://www.
ncbi.nlm.nih.gov/blast/mmtrace.html); BLAST pro-
gram for searches at the honeybee brain EST server
(http://titan.biotec.uiuc.edu/bee/honeybee_project.htm)
and at the server containing rst release of the Apis
mellifera genome (Amel_1.1) at Baylor College of
Medicine (http://www.hgsc.bcm.tmc.edu/projects/hon-
eybee). Overlapping sequence reads were edited manu-
ally using the trace archive at NCBI. The sequence of
the royalisin/defensin locus was assembled at a local
computer using the University of Wisconsin GCG
program package.
2.7. Other software used for sequence analyses
Transcription factor binding sites in the upstream
part of genes were identied using software
TFSEARCH (http://molsun1.cbrc.aist.go.jp/research/
db/TFSEARCH.html) with the threshold score=85,
Tfsitescan (http://www.ifti.org/cgi-bin/ifti/Tfsitescan.pl)
and PATCH (Transfac database 6.0 at http://www.gene-
regulation.com/, Wingender et al., 2000). Multiple
sequence alignments and phylogenic analyses were done
using CLUSTAL W (Higgins et al., 1996). Signal
peptidase cleavage sites were predicted using SIGNALP
3.0 (Bendtsen et al., 2004).
2.8. RTPCR analyses
Total RNAs from the heads and thoraces were
prepared by the guanidine thiocyanatephenol method
(Chomczynsky and Sacchi, 1987) after pulverization of
the tissues under liquid nitrogen. Obtained RNAs were
re-puried by RNA clean-up procedure using the
RNeasy Mini Kit (QIAGEN). Total RNA from glands
was extracted with the RNeasy Mini Kit after
ULTRATURRAX homogenization in denaturing so-
lution. The quality and the concentration (estimated
spectrophotometrically) of preparations were tested by
agarose gel electrophoresis. RTPCR assays were
performed with a one-tube, two-enzyme Access
RTPCR System (Promega) using 0.27 mg of RNAs in
25 ml volumes. Primers (0.4 mM) from the exon regions
giving similar amplication efciencies in PCR with
total cDNA and genomic DNA and containing two or
one intron, respectively, were used: Def1f1, Def1r1 for
defensin1 and Def2f2, Def2r for defensin2. Reverse
ARTICLE IN PRESS
J. Klaudiny et al. / Insect Biochemistry and Molecular Biology 35 (2005) 1122 13
transcription was made at 48 1C for 1 h. PCR after the
initial denaturation for 2 min at 94 1C was performed for
50 or 34 cycles (30 s at 94 1C, 60 s at 54 1C and 90 s at
69 1C) followed by 7 min at 69 1C. PCR products were
analyzed on 2.5% agarose gels using a 100 bp ladder
(Gibco) as a marker.
3. Results
3.1. Identication of the novel sequence variant of
honeybee royalisin
In order to explore whether royalisin and defensin are
products of two genes or allelic forms of the same gene,
we decided for the PCR approach. Two fragments of
defensin1 cDNA were amplied from the total cDNA of
nurse honeybee heads (Klaudiny et al., 1994) using
primers derived from the honeybee defensin cDNA
(Casteels-Josson et al., 1994). The PCR products were
cloned into pBluescript. Seven clones containing the
complete ORF and one clone encoding the mature
peptide were obtained. Their sequencing showed that all
clones carry inserts belonging to the identical cDNA.
The determined cDNA sequence (GenBank accession
number AY333923) differs slightly from the sequences
of honeybee defensin cDNA determined by Casteels-
Josson et al. (1994) and Lvov et al. (unpublished results,
GenBank accession number AJ308527). The sequence of
the pre-pro peptide inferred from the determined cDNA
sequence (termed Ro-K) contains one amino acid
substitution (Arg-Tyr) as compared to royalisin (Fuji-
wara et al., 1990, Fig. 1, Ro-F), another substitution
(Leu-His) as compared to Lvovs defensin (De-L) and
differs in the order of two amino acids (Gly-Val versus
Val-Gly) from defensin pre-pro-peptide (De-C, Fig. 1)
(Casteels-Josson et al., 1994).
Identication of the sole cDNA type among eight
characterized clones derived from two independent PCR
reactions using total cDNA of the heads (where RJ
producing glands reside) of the non-infected nurse
honeybees suggested that Ro-K represent the novel
variant of royalisin and not hemolymph defensin. To
conrm this, we characterized by MALDITOF mass
spectrometry the royalisin peptide of Apis mellifera
carnica, the same honeybee race as used for Ro-K
cDNA characterization. Royalisin was prepared by
preparative electrophoresis of an acidic extract of RJ
on the native acidic polyacrylamide gel combined with
the electro-elution of the peptides from the antibacterial
band (Figs. 2A, B). N-terminal sequencing showed
previously that royalisin is the major component of this
band (Bachanova et al., 2002). The molecular mass
(MW) of 5515.573 Da was determined for the royalisin
in this preparation (Fig. 3C). This value is about 10 Da
lower than that of the royalisin published by Fujiwara et
al. (1990) (5525.1 Da) and is in good agreement with the
mass of 5518.2 Da calculated for the mature amidated
Ro-K peptide. Two oxidized forms of the peptide were
identied in the preparation. We suppose that these
forms arose during extraction of RJ and/or during
native acidic electrophoresis. Mass spectrometry also
showed that the antibacterial band of RJ contains three
additional peptides with MWs of 45574783 (Fig. 2).
The results support the view that we characterized a
novel variant of royalisin and its cDNA.
ARTICLE IN PRESS
Fig. 1. Comparison of amino acid sequences of honeybee defensin1 variants. Pre-pro-peptide sequences of the honeybee defensin (De) and royalisin
(Ro) inferred from nucleotide sequences: Ro-K (this work), De-C (GenBank Accession No. U15955, Casteels-Josson et al., 1994), De-L (GenBank
Accession No. AJ308527, Lvov et al., unpublished). Ro-F, royalisin sequence determined by Fujiwara et al. (1990). Observed amino acid
substitutions are bold and underlined.
J. Klaudiny et al. / Insect Biochemistry and Molecular Biology 35 (2005) 1122 14
3.2. Search of honeybee sequence databases for royalisin
and defensin genes
High sequence similarities of royalisin and defensin
peptides and the identity of pre-pro regions of both
peptides suggested that these peptides might be the
variants encoded by a single polymorphic gene. There-
fore, searches for defensin-like sequences in honeybee
EST and genomic databases were performed with the
aim to determine the number of defensin genes and/or
polymorphism of the gene(s).
BLAST search of the honeybee brain EST database
[subtracted cDNA library prepared from 400 honeybee
brains from the same colony (Whiteld et al., 2002)]
revealed a single contig termed 1640 consisting of seven
independent sequence reads, which was identical with
the coding part of the royalisin cDNA determined here.
Search of the current release of honeybee genomic
sequencing (Amel_1.1) revealed several overlapping
sequence reads that could be assembled into a single
contig representing a single gene locus. The genomic
database represents the bulk of shotgun and nested BAC
sequence reads covering nearly 7-fold of the honeybee
genome. No further genomic fragments with obvious
homology to royalisin or defensin sequences, which
could represent another locus could be found in the
databases. However, another gene locus (GroupUn.243)
coding for a novel more distant defensin isoform was
found in the database. We designated the gene encoding
the novel defensin isoform as defensin2 and the one
encoding royalisin and hemolymph defensin as defensin1.
3.3. Honeybee defensin 1 gene and its polymorphism
The assembled defensin1 gene locus (Fig. 3) is 2012 bp
long (GenBank accession number AY496432). The gene
harbors two introns. The rst intron (571 bp long) is
located between bases 773 and 1345 and the second
intron (278 bp) locates between bases 1525 and 1804 in
the mature peptide coding region. Except for introns,
the sequence of the gene aligns perfectly with that of
royalisin cDNA characterized above.
The following analyses were done to assure that
defensin1 is a single locus. First, genomic DNAs of three
individual honeybees from different colonies were
amplied by PCR. The lengths of the amplied
fragments containing both introns (Def1f1Def1r1
primers) or the second intron (Def1f2Def1r1 primers)
corresponded with the sizes predicted from the genomic
sequence (1162 and 462 bp, respectively). Second, the
PCR product of one honeybee containing the second
intron was cloned and three clones were sequenced.
Obtained sequences were all identical with the as-
sembled defensin1 gene sequence. Third, a genomic
PCR fragment, covering the initial 755 bp of the
defensin1 containing the regulatory region of the gene
and the 5
0
untranslated region (UTR) of the defensin1
mRNA (Def1f3Def1r2 primers), was sequenced di-
rectly. The reads from both strands were homogeneous
and identical with the sequence of the assembled
defensin1 gene, except for a single overlapping peak
(CT) at the base 377, which indicates a single
nucleotide polymorphism at this site. Two further
ARTICLE IN PRESS
Fig. 2. Preparation and molecular mass determination of the royalisin (defensin1) peptide. (A), (B) Separation of RJ extract by native acidic
discontinuous polyacrylamide gel electrophoresis: (A) Growth inhibition zone visible on polyacrylamide gel after electrophoresis and growth of P. l.
larvae on the surface of the gel. (B) Parallel gel stained with Coomassie blue. Abthe antibacterial band containing royalisin and further unknown
peptides. (C) MALDITOF mass spectrum of the peptides isolated from the antibacterial band.
J. Klaudiny et al. / Insect Biochemistry and Molecular Biology 35 (2005) 1122 15
polymorphic sites in this part of the gene could be
deduced by comparison with the cDNA sequence of
hemolymph defensin (Casteels-Josson et al., 1994).
These and other polymorphic sites of the defensin1 gene
are listed in Table 1.
3.4. Honeybee defensin 2 gene and peptide
The sequence of defensin2 gene locus is presented in
Fig. 4. The gene is 1950 bp long and it contains one
intron (335 bp) located between bases 947 and 1283. It
was identied as a gene coding for a peptide with
homology to defensin 1 peptide. At the DNA level, the
highest homology is in the central defensin module
(56%). Otherwise, the genes have lower homology
upstream of TATA box, downstream of TATA box
up to the mature peptide processing site and also in the
3
0
non-coding regions. Both the exonintron structure
and the sequence of the pre-pro-peptide (Figs. 4 and 5)
were determined by sequencing of the cDNA fragments
amplied from the total cDNA of heads using primers
Def2f1, Def2r and Def2f2, Def2r. Except for the intron,
the sequence of defensin2 cDNA (Genbank accession
number AY588474) is identical with the defensin2 gene
sequence. There are four putative AATAAA sequences
in the defensin2 3
0
UTR, of which the rst does not seem
to function as a polyadenylation signal.
The encoded defensin 2 peptide (Fig. 5A) has a longer
pre-pro-peptide than defensin 1. The mature peptide
consists of 43 amino acids and does not contain a
ARTICLE IN PRESS
Fig. 3. Structure and translation of the A. mellifera defensin1 genomic locus. The sequence of 2012 bp of honeybee defensin 1 is shown. Only
boundary parts of the introns are presented (small bold letters). Putative TATA box and polyadenylation signal sequence AATAAA are in bold
characters. Upstream immune response elements located on sense (+) or antisense () strands are marked as follows: NF-kB consensus binding sites
are boxed, italics indicate a weak similarity. GATA binding sites are in bold underlined characters. Primers used for PCR amplications (without
overhangs for cloning) are marked by horizontal arrows above the sequence. Polymorphic nucleotides are highlighted. Processing sites (signal
peptidase site, mature peptide cleavage site, C-terminal amidation site) are marked by vertical arrows.
J. Klaudiny et al. / Insect Biochemistry and Molecular Biology 35 (2005) 1122 16
ARTICLE IN PRESS
Table 1
Polymorphic sites found in defensin1 gene by comparison of known corresponding genomic, cDNA and peptide sequences (see also Fig. 3)
Positions in gene sequence Substitutions or variants Amino acid substitutions Compared sequences References
337 T-C 5
0
upstream region Amplied genomic DNA This paper
634636 GTT-TTG 5
0
mRNA uncoding region Defensin cDNA Casteels-Josson et al. (1994)
1471 T-A L-H Genomic DNA fragment Lvov et al., Genbank
15071511 GGAGT-TTAGA GV-VG Defensin cDNA Casteels-Josson et al. (1994)
18321833 CG-TA
*
R-Y Royalisin peptide Fujiwara et al. (1990)
*
Nucleotide substitution was deduced on the basis of low degeneracy of the codon for amino acid tyrosine (Y) occurred in the royalisin peptide.
Fig. 4. Structure and translation of the defensin2 genomic locus of A. mellifera. The sequence of 1950 bp of honeybee defensin2 is shown. Boundary
sequences of the intron are marked by small bold letters. Putative TATA box and polyadenylation signal sequences are in bold characters. Upstream
immune response elements located on the sense (+) or antisense () strands are marked as follows: NF-kB consensus binding sites are labeled by
boxes, italics indicate a weak consensus. GATA binding sites are in bold underlined characters. SP1 regulatory elements are highlighted and
underlined. Primers used for PCR amplications are marked by horizontal arrows above the sequence. Putative signal peptidase and mature peptide
cleavage sites are marked by the vertical arrows.
J. Klaudiny et al. / Insect Biochemistry and Molecular Biology 35 (2005) 1122 17
ARTICLE IN PRESS
Fig. 5. Comparison of defensin sequences of A. mellifera with other insect defensins and their phylogeny. (A) Comparison of the pre- pro regions of
the two honeybee defensin isoforms. The sequences of pre-pro-regions have been aligned manually. The pro-peptide cleavage site of defensin 2 was
set arbitrarily, based on consensus processing site. Conserved residues are in bold characters. Black triangles mark positions of the introns. (B)
Alignment of mature insect defensins. The alignment was done using CLUSTAL W, edited manually and visualized using BOXSHADE. Preserved
residues are shaded black, conserved substitutions are gray. (C) Phylogeny of A. mellifera and other insect defensins. The phylogenic tree was
reconstructed from the aligned defensin sequences by the neighbor-joining method using mollusk defensin sequence as an outgroup. Results of the
branch-and-bound bootstrap analysis are shown above the nodes (only values above 65% are shown). Note the monophyly of the two mosquito
defensins (A. aegypti B, C), which is in direct contrast with the paraphyletic defensins of A. mellifera (bold), where defensin 1 forms a branch together
with other hymenopteran defensins (B. ignitus and B. pascuorum), whereas defensin 2 groups with beetle defensins (O. rhinoceros and A. dichotoma).
J. Klaudiny et al. / Insect Biochemistry and Molecular Biology 35 (2005) 1122 18
C-terminal extension found in defensin 1 and other
hymenopteran defensins (Fig. 5B). Its predicted MW
(4808.6 Da) excluded the fact that defensin 2 peptide
would be among the unknown peptides in the royalisin
preparation detected by mass spectrometry (Fig. 2).
Defensin 2 mature peptide shows 55.8% identity with
that of defensin 1. Comparison of both honeybee
defensins with other arthropodan defensins and their
phylogeny is shown in Figs. 5B and 5C.
3.5. Analyses of the regulatory sequences of defensin
genes
Upstream regulatory regions of both defensin genes
were screened for binding sites for transcription factors
known to be involved in insect immune response
(Dimarcq et al., 1994; Meister et al., 1994; Eggleston
et al., 2000). One consensus sequence for NF-kB was
identied in each gene in a similar position upstream of
TATA box (82 bp in defensin1 and 106 bp in defensin2)
(Figs. 3 and 4). Besides these sites, each gene contains
another less preserved NF-kB motif located more
upstream. Binding sites for two other transcription
factors were found in the genes. Eight GATA factor
binding sites were identied in defensin1 and only two in
defensin2 (one located downstream of TATA box). On
the other hand, two binding sites for the transcription
factor SP1 were identied in defensin2 but none in
defensin1. Thus, there are similarities as well as
differences between promoter regions of the two
honeybee defensin genes. This nding suggests differ-
ences in the regulation of these two genes.
3.6. Gene expression of honeybee defensins
Transcription patterns of the two defensins in some
body parts and tissues of nurse honeybees of dened age
taken from a visually healthy colony were determined by
RTPCR (Fig. 6). The analyses revealed that both
defensin isoforms are expressed in the heads and
thoraces. Transcriptional levels of defensin1 detected in
the heads and therein located hypopharyngeal and
mandibular glands were higher than those in thoraces
and thoracic salivary glands. No expression of defensin2
was observed in any of the three glands tested.
Interestingly, an inter-colony variation in defensin
expression was observed by analyses of RNA prepara-
tions of two different colonies. The level of defensin1
mRNA was higher in the colony H than Hc, whereas
defensin2 mRNA was more pronounced in Hc (Fig. 6).
4. Discussion
We have found that honeybee genome contains two
different defensin genes. One of them, defensin1, encodes
both known honeybee defensins: royalisin, a defensin
peptide of RJ, and defensin found in the hemolymph of
bacterially infected bees. All obtained data support the
view of a single defensin1 gene. The second gene,
defensin2, codes for a novel hitherto uncharacterized
defensin, whose precise expression pattern and its role in
honeybee physiology have to be further investigated.
The two defensin genes characterized here differ
signicantly in their lengths, overall buildup, intronex-
on structure (Figs. 3 and 4) and in the sequence of their
pre-pro regions (Fig. 5). In particular, the second intron
of defensin1, which is not present in defensin2, is the rst
intron found within the part coding for the mature
defensin peptide among arthropods. It splits exactly the
mature peptide module common to all arthropodan
defensins and a short amidated C-terminal extension
found only in hymenopteran defensins. This 11-amino-
acid-long peptide extension is supposed to adopt the a-
helix structure that is stabilized by C-terminal amidation
(Casteels, 1998). The precise role of this extension for
the function of the peptide is not clear. It is apparent
that the extension results from an exon-shufing event
(Patthy, 1999; Long, 2001) by which two exons
recombined to create an atypical defensin with the
extension. This event probably occurred during the
evolution of hymenoptera as defensins of two bumble-
bee species contain the same C-terminal extension.
Evolution of invertebrate defensins was reviewed
recently by Froy and Gurevitz (2003). These authors
suggested that in the course of arthropodan evolution,
ARTICLE IN PRESS
Fig. 6. Gene expression of honeybee defensins in different body parts
and gland tissues of A. mellifera. Total RNA was isolated from the
heads (H), hypopharyngeal glands (Hg), mandibular glands (Mg),
thoraces (T) and thoracic salivary glands (Tg) of nurse honeybees from
the same colony. For comparison, total RNA from the heads (Hc) of
honeybees is included here coming from the colony, from which total
cDNA used in the work for amplication and characterization of both
defensins was derived. RTPCR was performed with the primers
specic for defensins (see Section 2.8.) and RNAs mentioned above.
The RTPCR products of defensin1 and defensin2 mRNAs with the
sizes 320 and 492 bp, respectively, are shown.
J. Klaudiny et al. / Insect Biochemistry and Molecular Biology 35 (2005) 1122 19
an autonomous module represented by the exon
encoding the mature defensin underwent several exon
shufings and integrated downstream of unrelated
leader sequences. In line with the suggested mechanism,
defensin genes differ in the number of introns in gene
parts encoding pre-pro regions, have low sequence
homologies in the leader sequences, pro-regions and 3
0
UTR, but show sequence homology, similar biological
activity and structural relatedness in the sequences of
mature peptides. All these characteristics on both
genetic and protein levels also hold true for both
isoforms of honeybee defensin. The differences in the
exonintron structure of two honeybee defensin genes as
well as the extent of differences among sequences of
their mature peptides (Fig. 5) suggest that these genes,
although originating from the common defensin gene,
passed a longer period of independent evolution. This
view is also conrmed by the reconstruction of the
phylogeny of defensin genes, where, in contrast to
defensins of mosquito, honeybee defensins were para-
phyletic (with a high bootstrap support of 84%).
Defensin1 formed a phylum with two other hymenop-
teran defensins, whereas defensin2 grouped with beetle
defensins.
The presence of certain immune response elements
(Dimarcq et al., 1994; Meister et al., 1994; Eggleston et
al., 2000; Volkoff et al., 2003), rst of all the NF-kB
binding site, in upstream sequences of both defensin
genes indicates that the genes are inducible or up-
regulated by immune challenge. Inducible regulation of
defensin1 expression was shown to take place in the fat
body (Casteels-Josson et al., 1994). Recently defensin1
up-regulation in as yet unidentied tissue(s) of the rst
instar bee larvae infected orally with Paenibacillus l.
larvae has also been reported (Evans, 2004). On the
other hand, both genes differ in the occurrence of other
immune response elements. Defensin1 gene contains
eight GATA factor binding sites that were reported to
be involved in the interactions with NF-kB transcription
factor in fat body-specic expression (Abel et al., 1993;
Dittmer and Raikhel, 1997). In contrast, the defensin2
gene has only two GATA motives but it contains two
SP1 binding sites not identied in defensin1. SP1 belongs
to G-rich elements that are recognized by several
transcription factors which are involved in regulation
of the expression of housekeeping and many tissue-
specic genes (Lania et al., 1997; Suske, 1999). Different
occurrences of the above-mentioned regulatory elements
suggest differences in the tissue expression pattern,
which correlates with the results obtained from the
analyses of defensin expression.
The mRNA of defensin1 gene was located to both
head glands known to contribute to the production of
royal jelly proteinshyphopharyngeal and mandibular
glands (Lensky and Rakover, 1983). Transcription of
defensin1 in the head glands of nurse bees corresponds
to royalisin expression. This seems to be constitutive in
this group of bees, because royalisin was found in all
tested RJ samples collected from different colonies and
apiaries (Bachanova et al., 2002 and unpublished data).
The analyses also revealed that RJs from different
healthy colonies can differ in the antibacterial activity
and the amount of the antimicrobial peptide. One of the
reasons could be genetic variability among honeybees of
different colonies. The polymorphism of the honeybee
defensin1 gene reported here could play an important
role in modifying royalisin expression levels and its
antimicrobial and antifungal (B likova et al., 2001)
activity. This might be of importance for practical
beekeeping, where selecting naturally more resistant
colonies against microbial pathogens, especially to P. l.
larvae, becomes increasingly important due to wide-
spread antibiotic resistance (Evans, 2003).
Detection of defensin1 transcription in the thoracic
salivary glands suggests the presence of defensin 1 in
honeybee saliva. It can function to protect salivary
gland epithelium and also to protect comb cells, queen
and other bees, stored pollen bread, wax and propolis.
Defensin2 mRNA was detected in total RNAs of
nurse heads and thoraces, but not in any of the three
glands tested. Both low transcriptional level and the
presence of immune regulatory elements found in
promoter regions suggest that defensin2 expression is
infection-inducible or up-regulated in as yet unknown
cells or tissues. The expression of defensin2 in fat body
and/or hemocytes seems to be less probable while the
peptide was not identied among antibacterial peptides
in the hemolymph of infected honeybees (Casteels, 1998
and references therein).
The social lifestyle of honeybees in colonies consisting
of as many as ten thousands of individuals, which are
crowded in a small space conditioned to a constant
temperature of 35 1C, with frequent contacts and mutual
feeding of individuals is in fact an ideal milieu for the
rapid spreading of infectious diseases. This environment
certainly provides a challenge for the innate immune
defense to which defensins belong. The resistance of
honeybees and colonies to microbial and fungal patho-
gens depends on the genetic background of honeybees in
the colony and it could be modulated with age and role
of the honeybee in the colony. Therefore, more extensive
studies will be required to determine the precise function
of defensin 2, to obtain further knowledge about the
factors inuencing expression of defensin1 and about the
interplay of two defensins together with other anti-
bacterial peptides in honeybee colony health.
Acknowledgements
We thank Dr. R.M. Nitsch for providing us with the
opportunity to perform some experiments in his
ARTICLE IN PRESS
J. Klaudiny et al. / Insect Biochemistry and Molecular Biology 35 (2005) 1122 20
laboratory at ZMNH in Hamburg. We also thank
G. Pls kova and M. Krupcova for technical assistance
and D. Dianis ka, J. Za bojn k, A. Za bojn kova , E.
Droppova , V. Gajdos ova and T. C

erma kova for help in


preparing biological material. We acknowledge J.
Gadau, B. Malecova , D. Lee and an anonymous referee
for their valuable comments on the manuscript. Mass
spectrometry was measured at the Institute of Micro-
biology of ASCR in Prague. This work was supported
by the Scientic Grant Agency of the Ministry of
Education of Slovak Republic and the Slovak Academy
of Sciences VEGA 2/1053/21 and VEGA 2/4059/24.
References
Abel, T., Michelson, A.M., Maniatis, T., 1993. A. Drosophila GATA
family member that binds to Adh regulatory seguences is expressed
in the fat body. Development 119, 623633.
Bachanova , K., Klaudiny, J., Kopernicky , J., S

imu th, J., 2002.


Identication of honeybee peptide active against Paenibacillus
larvae larvae through bacterial growth-inhibition assay on poly-
acrylamide gel. Apidologie 33, 259269.
Bendtsen, J.D., Nielsen, H., von Heijne, G., Brunak, S., 2004.
Improved prediction of signal peptides: SignalP 3.0. J. Mol. Biol.
340, 783795.
B likova , K., Gusui, W., S

imu th, J., 2001. Isolation of a peptide


fraction from honeybee royal jelly as a potential antifoulbrood
factor. Apidologie 32, 275283.
Boman, H.G., 1995. Peptide antibiotics and their role in innate
immunity. Annu. Rev. Immunol. 13, 6192.
Bonmatin, J.M., Bonnat, J.L., Gallet, X., Vovelle, F., Ptak, M.,
Reichart, J.M., Hoffmann, J.A., Keppi, E., Legrain, M., Ach-
stetter, T., 1992. Two-dimensional
1
H NMR study of recombinant
insect defensin A in water: resonance assignments, secondary
structure and global folding. J. Biomol. NMR 2, 235256.
Brey, P.T., Lee, W., Yamakawa, M., Koizumi, Z., Perrot, S., Francois,
M., Ashida, M., 1993. Role of the integument in insect immunity:
epicuticular abrasion and induction of cecropin synthesis in
cuticular epithelial cells. Proc. Nat. Acad. Sci. USA 90, 62756279.
Bulet, P., Hetru, C., Dimarcq, J.L., Hoffmann, D., 1999. Antimicro-
bial peptides in insects; structure and function. Dev. Comp.
Immunol. 23, 329344.
Casteels, P., 1998. Immune response in Hymenoptera. In: Brey, P.T.,
Hultmark, D. (Eds.), Molecular Mechanisms of Immune Re-
sponses in Insect. Chapman & Hall, London, pp. 92110.
Casteels, P., Ampe, C., Jacobs, F., Vaek, M., Tempst, P., 1989.
Apidaecins: antibacterial peptides from honeybees. EMBO J. 8,
23872391.
Casteels, P., Ampe, C., Riviere, L., Damme, J.V., Elicone, C., Fleming,
M., Jacobs, F., Tempst, P., 1990. Isolation and characterization of
abaecin, a major antibacterial peptide in the honeybee (Apis
mellifera). Eur. J. Biochem. 187, 381386.
Casteels, P., Ampe, C., Jacobs, F., Tempst, P., 1993. Functional and
chemical characterization of hymenoptaecin, an antibacterial
polypeptide that is infection-inducible in the honeybee (Apis
mellifera). J. Biol. Chem. 268, 70447054.
Casteels-Josson, K., Zhang, W., Capaci, T., Casteels, P., Tempst, P.,
1994. Acute transcriptional response of the honeybee peptide-
antibiotics gene repertoire and required post-translational conver-
sion of the precursor structures. J. Biol. Chem. 269, 2856928575.
Chomczynsky, P., Sacchi, N., 1987. Single-step RNA isolation from
cultured cells or tissues. Anal. Biochem. 162, 156259.
Cociancich, S., Ghazi, A., Hetru, C., Hoffmann, J.A., Letellier, L.,
1993. Insect defensin, an inducible antibacterial peptide, forms
voltage-dependent channels in Micrococcus luteus. J. Biol. Chem.
268, 1923919245.
Cornet, B., Bonmatin, J.M., Hetru, C., Hoffmann, J.A., Ptak, M.,
Vovelle, F., 1995. Rened three-dimensional solution structure of
insect defensin A. Structure 3, 435448.
Dimarcq, J.L., Bulet, P., Hetru, C., Hoffmann, J., 1998. Cysteine-rich
antimicrobial peptides in invertebrates. Biopolymers 47, 465477.
Dimarcq, J.L., Hoffman, D., Meister, M., Bulet, P., Lanot, R.,
Reichhart, J.M., Hoffman, J.A., 1994. Characterization and
transcriptional proles of a Drosophila gene encoding an insect
defensin. Eur. J. Biochem. 221, 201209.
Dimopoulos, G., Seeley, D., Wolf, A., Kafatos, F.C., 1998. Malaria
infection of the mosquito Anopheles gambiae activates immunor-
esponsive genes during critical transition stages of the parasite life
cycle. EMBO J. 17, 61156123.
Dittmer, N.T., Raikhel, A.S., 1997. Analysis of the mosquito
lysosomal aspartic protease gene: an insect housekeeping gene
with fat body-enhanced expression. Insect Biochem. Mol. Biol. 27,
323335.
Eggleston, P., Lu, W., Zhao, Y., 2000. Genomic organization and
immune regulation of the defensin gene from the mosquito,
Anopheles gambiae. Insect Mol. Biol. 9, 481490.
Evans, J.D., 2003. Diverse origins of tetracycline resistance in the
honey bee bacterial pathogen Paenibacillus larvae. J. Inver. Pathol.
83, 4650.
Evans, J.D., 2004. Transcriptional immune responses by honey bee
larvae during invasion by the bacterial pathogen, Paenibacillus
larvae. J. Inver. Pathol. 85, 105111.
Fondrk, M.K., Page, R.E., Hunt, G.J., 1993. Paternity analysis of
worker honeybees using random amplied polymorphic DNA.
Naturwissenschaften 80, 226231.
Froy, O., Gurevitz, M., 2003. Arthropod and mollusk defensins-
evolution by exon-shufing. Trends Genet. 19, 684687.
Fujiwara, S., Imai, J., Fujiwara, M., Yaeshima, T., Kawashima, T.,
Kobayashi, K., 1990. A potent antibacterial protein in royal jelly.
J. Biol. Chem. 265, 1133311337.
Hanzawa, H., Shimada, I., Kuzuhara, T., Komano, H., Kohda, D.,
Inagaki, F., Natori, S., Arata, Y., 1990. 1 H nuclear magnetic
resonance study of the solution conformation of an antibacterial
protein, sapecin. FEBS Lett. 269, 413420.
Hetru, C., Hoffmann, D., Bulet, P., 1998. Antimicrobial peptides from
insects. In: Brey, P.T., Hultmark, D. (Eds.), Molecular Mechan-
isms of Immune Responses in Insects. Chapman & Hall, London,
pp. 4066.
Higgins, D.G., Thompson, J.D., Gibson, T.J., 1996. Using CLUSTAL
for multiple sequence alignments. Methods Enzymol. 266, 383402.
Hoffmann, J.A., Kafatos, F.C., Janawey, C.A., Ezekovitz, R.A.B.,
1999. Phylogenetic perspectives in innate immunity. Science 284,
13131318.
Klaudiny, J., Hanes, J., Kulifajova , J., Albert, S

., S

imu th, J., 1994.


Molecular cloning of two cDNAs from the head of the nurse honey
bee (Apis mellifera L.) for coding related proteins of royal jelly.
J. Apic. Res. 33, 105111.
Lania, L., Majello, B., De Luca, P., 1997. Transcriptional
regulation by the Sp family proteins. Int. J. Biochem. Cell Biol.
29, 13131323.
Lehane, M.J., Wu, D., Lehane, S.M., 1997. Midgut-specic immune
molecules are produced by the blood-sucking insect Stomoxys
calcitrans. Proc. Nat. Acad. Sci. USA 94, 1150211507.
Lensky, Y., Rakover, Y., 1983. Separate protein body compartments
of the worker honeybee (Apis mellifera L.). Comp. Biochem.
Physiol. 75, 607615.
Long, M., 2001. Evolution of novel genes. Curr. Opinion Genet. Dev.
11, 673680.
ARTICLE IN PRESS
J. Klaudiny et al. / Insect Biochemistry and Molecular Biology 35 (2005) 1122 21
Lopez, L., Morales, G., Ursic, R., Wolff, M., Lowneberger, C., 2003.
Isolation and characterization of a novel insect defensin from
Rhodnius prolixus, a vector of Chagas disease. Insect Biochem.
Mol. Biol. 33, 439447.
Lowenberger, C., 2001. Innate immune response of Aedes aegypti.
Insect Biochem. Mol. Biol. 31, 219229.
Lowenberger, C.A., Kamal, S., Chiles, J., Paskewitz, S., Bulet, P.,
Hoffmann, J.A., Christensen, B.M., 1999a. Mosquito-Plasmodium
interaction in response to immune activation of the vector. Exp.
Parasitol. 91, 5969.
Lowenberger, C.A., Smartt, C.T., Bulet, P., Ferdig, M.T., Severson,
D.W., Hoffman, J.A., Christensen, B.M., 1999b. Insect immunity:
molecular cloning, expression, and characterization of cDNAs and
genomic DANN encoding three isoforms of insect defensin in
Aedes aegypti. Insect Mol. Biol. 8, 107118.
Meister, M., Braun, A., Kappler, C., Reichhart, J.M., Hoffmann, J.A.,
1994. Insect immunity. A transgenic analysis in Drosophila denes
several functional domains in the diptericin promoter. EMBO J.
13, 59585966.
Patthy, L., 1999. Genome evolution and the evolution of exon-
shufinga review. Gene 238, 103114.
Raj, P.A., Dentino, A.R., 2002. Current status of defensins and their
role in innate and adaptive immunity. FEMS Microbiol. Lett. 206,
918.
Rees, J.A., Moniatte, M., Bulet, P., 1997. Novel antibacterial peptides
isolated from a European bumblebee, Bombus pascuorum (Hyme-
noptera, Apoidea). Insect Biochem. Mol. Biol. 27, 413422.
Suske, G., 1999. The Sp-family of transcription factors. Gene 238,
291300.
Volkoff, A.N., Rocher, J., d
0
Alencon, E., Bouton, M., Landais, I.,
Quesada-Moraga, E., Vey, A., Fournier, P., Mita, K., Devauchelle,
G., 2003. Characterization and transcriptional proles of three
Spodoptera frugiperda genes encoding cysteine-rich peptides. A new
class of defensin-like genes from lepidopteran insects? Gene 319,
4353.
Whiteld, C.W., Band, M.R., Bonaldo, M.F., Kumar, C.G., Liu, L.,
Pardinas, J.R., Robertson, H.M., Soares, M.B., Robinson, G.E.,
2002. Annotated expressed sequence tags and cDNA microarrays
for studies of brain and behavior in the honey bee. Genome Res.
12, 555566.
Wingender, E., Chen, X., Hehl, R., Karas, H., Liebich, I., Matys, V.,
Meinhardt, T., Reuter, I., Schacherer, F., 2000. TRANSFAC: an
integrated system for gene expression regulation. Nucleic Acids
Res. 28, 316319.
Yamauchi, H., 2001. Two novel insect defensins from larvae of the
cupreous chafer, Anomala cuprea: purication, amino acid
sequences and antibacterial activity. Insect Biochem. Mol. Biol.
32, 7578.
ARTICLE IN PRESS
J. Klaudiny et al. / Insect Biochemistry and Molecular Biology 35 (2005) 1122 22
Insect
Biochemistry
and
Molecular
Biology
Insect Biochemistry and Molecular Biology 35 (2005) 2332
Cloning and characterization of chymotrypsin- and trypsin-like
cDNAs from the gut of the Hessian y [Mayetiola destructor (say)]
$
Yu Cheng Zhu
a
, Xiang Liu
b
, Ashoka A. Maddur
b
, Brenda Oppert
b,c
, Ming-Shun Chen
b,d,
a
USDA-ARS-JWDSRC, PO Box 346/141 Exp Stn Rd, Stoneville, MS 38776, USA
b
Department of Entomology, USDA-ARS, Kansas State University, 123 Waters Hall, Manhattan, KS 66506, USA
c
USDA-ARS, Biological Research Unit, 1515 College Avenue, Manhattan, KS 66502, USA
d
USDA-ARS, Plant Science and Entomology Research Unit, 4008 Throckmorton Hall, Kansas State University, Manhattan, KS 66506, USA
Received 16 August 2004; received in revised form 26 September 2004; accepted 29 September 2004
Abstract
Fifteen unique cDNA clones encoding trypsin- or chymotrypsin-like proteins were cloned and characterized from a gut cDNA
library derived from Hessian y [Mayetiola destructor (Say)] larvae. Based on sequence similarities, the cDNAs were sorted into ve
gene groups, which were named MDP1 to MDP5. Two of the gene groups, MDP1 and MDP2, encoded chymotrypsin-like proteins;
the other three encoded putative trypsins. All deduced proteins have conserved His
87
, Asp
136
, and Ser
241
residues for the catalytic
triad and three pairs of cysteine residues for disulde bridge congurations. The substrate specicity determination residue at
position 235 was also conserved in the putative trypsins and chymotrypsins. In addition, all the deduced protein precursors had a
typical secretion signal peptide and activation peptide. Northern blot analysis revealed that all these gene groups were exclusively
expressed in the larval stage. The expression proles for each gene group differed signicantly in different ages of the larva, as well as
in different tissues. Protease activity analysis of gut extract, using specic inhibitors, demonstrated that serine proteases were the
major digestive enzymes in the gut of M. destructor larvae. Serine protease inhibitors inhibited as much as 90% proteolytic activities
of gut extract, whereas inhibitors specic to other proteases, including cysteine proteases, aspartic proteases, and metallo-proteases,
inhibited only 1024% of gut protease activity.
r 2004 Elsevier Ltd. All rights reserved.
Keywords: Hessian y; Mayetiola destructor; Wheat; Serine protease; Gut proteinases
1. Introduction
The Hessian y [Mayetiola destructor (Say)] is one
of the most destructive pests of wheat (Triticum aestivum
L.) (Hatchett et al., 1987; Buntin, 1999). The most
effective measure for controlling this insect pest is
through the release of resistant wheat cultivars (Ratcliffe
and Hatchett, 1997). As a consequence, most of the
research on M. destructor is related to hostplant
resistance. This includes the identication and introgres-
sion of hostplant resistance genes (Ratcliffe et al., 2003;
Martin-Sanchez et al., 2003; Williams et al., 2003),
molecular mapping of Avr genes (Rider et al., 2002;
Behura et al., 2004), study of the distribution of different
biotypes (Ratcliffe et al., 1994, 2000), and the char-
acterization of induced wheat genes after infestation
(Williams et al., 2002; Jang et al., 2003). All of the
resistance genes so far identied confer resistance
through antibiosis (Ratcliffe and Hatchett, 1997, Harris
et al., 2003). First instarlarvae feeding on resistant
plants die within 4 days without developing into the
second instar (Hatchett and Gallun, 1970; El Bouhssini
ARTICLE IN PRESS
www.elsevier.com/locate/ibmb
0965-1748/$ - see front matter r 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ibmb.2004.09.006
Abbreviations: kDa, kilo-Dalton; EST, Expressed Sequence Tag;
PCR, Polymerase Chain Reaction; ORF, Open Reading Frame
$
Mention of commercial or proprietary product does not constitute
endorsement by the USDA.

Corresponding author. Department of Entomology, USDA-ARS,


Kansas State University, 123 Waters Hall, Manhattan, KS 66506,
USA. Tel.: +7855324719; fax: +7855326232.
E-mail address: mchen@oznet.ksu.edu (M.-S. Chen).
et al., 1998). Molecular mechanisms for the antibiosis
are not yet known. Experimental results demonstrated
that there is no signicant difference in the amount of
food ingested from host plants between virulent and
avirulent larvae in the initial feeding (Gallun and
Langston, 1963). This observation raised the possibility
that the antibiosis might be caused by toxic molecules
produced in resistant plants. One of the likely targets for
such toxicity in the insect would be the gut, where food
digestion and detoxication take place (Terra and
Ferreira, 1994; Herrero et al., 2001). It is known that
plants synthesize various toxic molecules upon infesta-
tion, such as inhibitors to herbivores digestive enzymes
(Karban and Baldwin, 1997; Moura and Ryan, 2001).
Protease inhibitors from host plants have been found to
have a detrimental effect on insect development and are
being used as targets for bioengineering to generate
resistant plants (Murdock et al., 1988; Burgess et al.,
1994; Huang et al., 1997). Protease activity has been
detected in the gut extract from M. destructor larvae
(Shukle et al., 1985), but little is known about the
molecular and biochemical aspects of these enzymes.
To isolate protease genes that are expressed in the gut of
M. destructor larvae, we systematically analyzed the
genes expressed in the gut of the rst instarlarvae
following a transcriptomic approach. Here we report the
cloning and characterization of 15 trypsin- and chymo-
trypsin-like cDNAs identied from this analysis.
2. Materials and methods
2.1. Insects
Hessian y larvae used in this research were derived
from a laboratory colony. The insects were originally
collected from Ellis County, Kansas (Gagne and
Hatchett, 1989). Since then, the insects have been
maintained on susceptible wheat seedlings (Newton
or Karl 92) in the greenhouse. The majority (95%) of
the insects were biotype GP although biotypes A, B and
others were also found in low frequencies (Harris and
Rose, 1989).
2.2. cDNA library construction and sequencing
Two hundred guts were obtained by dissecting 3-day-
old larvae (rst instar) under a dissecting microscope.
The guts were immediately transferred into TRI
reagent
TM
(Molecular Research Center, Inc. Cincinnati,
OH). Total RNA was extracted from the tissue of the
gut using TRI reagent according the procedure provided
by the manufacturer. A cDNA library was constructed
from the RNA sample using a SMART
TM
cDNA
library construction kit from Clontech (Palo Alto, CA)
according to the protocol provided by the manufacturer
with one modication. Instead of using the original
phage vector, PCR fragments were cloned directly into a
plasmid using a TOPO TA cloning kit (Invitrogen,
Carlsbad, CA). Recombinant plasmid DNA was iso-
lated with a Qiagen BioRobot-3000 and sequenced using
an ABI 3700 DNA analyzer.
2.3. RNA isolation and Northern blot analysis
Total RNA was extracted from whole insects using
TRI reagent
TM
(see above). For Northern blot, equal
amounts (5 mg) of total RNA were separated on a 1.2%
agarose gel containing formaldehyde and blotted onto
GeneScreen membrane (Perkin Elmer Life Science Inc.,
Boston, MA). The membrane was baked at 80 1C for 2 h
to x RNAs onto the membrane. The membranes were
then hybridized separately to individual cDNA probes
labeled with
32
P-dCTP using the random labeling kit
from Stratagene (La Jolla, CA). Hybridization was
carried out overnight at 42 1C in a plastic bag containing
15 ml hybridization solution (10% dextran sulfate/1%
SDS/1 M NaCl, pH 8.0) as described by Chen et al.
(2004). After hybridization, the membranes were washed
twice with 2 SSC at room temperature for 30 min,
twice with 2 SSC plus 1%SDS at 65 1C for 30 min, and
twice with 0.1 SSC plus 1%SDS at room temperature
for 30 min. The membranes were then exposed to Kodak
SR-5 X-ray lm overnight.
2.4. Sequence analysis
Open-reading-frame (ORF) and sequence-similarity
analysis were performed through the website (http://
www.ncbi.nlm.nih.gov/) of the National Center for
Biotechnology Information (Bethesda, MD). Molecular
weight calculations and pI prediction of mature proteins
were carried out with the Compute pI/Mw tool (http://
us.expasy.org/tools/pi_tool.html). The ExPASy Proteo-
mics tools on the website (http://www.expasy.ch/tools/)
of the Swiss Institute of Bioinformatics were used to
process data of deduced protein sequences. ClustalW
(Thompson et al., 1994; gap weight=8, gap length
weight=2) was used in pairwise sequence comparison.
The BioEdit (Ver. 5.09; Hall, 1999) program was used to
conduct multiple-sequence alignment and phylogenetic
analysis.
2.5. Analysis of tissue-specic expression through RT-
PCR
RT-PCR was carried out to determine if specic
proteases were expressed in the gut or salivary glands.
Total RNA was separately prepared from dissected guts
or salivary glands. The RNA samples were treated with
RNase-free DNase-I (Promega, Madison, WI) to
remove potential DNA contamination. The RNA was
ARTICLE IN PRESS
Y.C. Zhu et al. / Insect Biochemistry and Molecular Biology 35 (2005) 2332 24
then reverse transcribed into cDNA using an oligo-dT
primer with SuperScript II reverse transcriptase (Invi-
trogen). PCR amplication was carried out for 45 cycles
as follows: 30 seconds (s) at 94 1C; 30 s at 55 1C; 60 s at
72 1C. DNA fragments from the PCR reactions were
separated on 1% agarose gels, stained with (0.5 mg/mL)
ethidium bromide. DNA bands were photographed with
a Kodak 290 digital camera, and band intensity was
determined using Kodak 1D image analysis software
(Version 3.54).
For primer design, we divided the protease cDNAs
into ve gene groups, according to their sequence
similarity (see the Results section). Primer pairs specic
to each gene group were synthesized (see primer
sequences in Fig. 3). The specicity of the primer pairs
to individual groups was tested experimentally using
individual plasmid DNA samples as templates. Primers
MDP1F+MDP1R could ank the corresponding
regions of all ve chymotrypsin (A-F) cDNAs in
the MDP1 gene group to produce 657-bp fragments.
Primers MDP2F+MDP2R were specic to the chymo-
trypsin cDNAs in the MDP2 gene group to produce
601-bp fragments. Primers MDP3F+MDP3R were
specic to the trypsin cDNAs in the MDP3 gene group
to produce 677-bp fragments. Primers MDP4F+
MDP4R were specic to trypsin cDNAs in the MDP4
gene group to produce 669-bp fragments. Primers
MDP5F+MDP5R were specic to the trypsin cDNAs
in the MDP5A gene group to produce 486-bp frag-
ments. As a control, a primer pair (MDrDF+MDrDR)
was synthesized according to a cDNA that encodes a
L27A-like ribosomal protein in the 60S subunit. The
RT-PCR products (420 bp) from this ribosomal protein
mRNA were used as an internal control to quantify the
amounts of the protease RNAs.
2.6. Protease activity assay
Midguts of rst instar Mayetiola destructor larvae
were dissected under a light microscope in cold
deionized water. The midgut was exposed by pulling
the head with a ne needle, and the salivary gland and
Malphigian tubules were removed. Midguts were
collected in deionized water, mixed through pipetting,
and pooled (20 guts per 20 ml water) in Eppendorf tubes
on ice. The extract was centrifuged at 15,000g for 5 min,
and the supernatant was aliquoted to 10 ml and frozen
at 20 1C.
Accordingly, inhibition studies were conducted by
measuring the hydrolysis of casein conjugated to a
uorescent probe, BODIPY-TR-X (Molecular Probes,
Eugene, OR) according to a method previously
described (Oppert et al., 1997). Inhibitors included
antipain, aprotinin, 4-amido-phenylmethylsulfonyl
uoride (APMSF), bestatin, chymostatin, ethyelenedia-
mine tetraacetic acid (EDTA), ethylene glycol-bis
(beta-aminoethyl-ether)-N,N,N
0
,N
0
-tetraacetate (EGTA),
transexposysuccinyl-L-leucylamido-(4-guanido)-butane
(E-64), soybean Kunitz trypsin inhibitor (SBTI),
soybean Bowman-Birk trypsin inhibitor (SBBTI), and
N-tosyl-L-phenyalanine chloromethyl ketone (TPCK).
These inhibitors were obtained either from Pierce
Chemical Company (Rockford, IL), Roche Applied
Science (Indianapolis, IN), or Sigma Chemical Com-
pany (St. Louis, MO). Stock solutions were prepared in
solvents per company specications. The starting con-
centrations of the inhibitors are provided in the legend
of Fig. 4. Protease inhibitors were added to universal
buffer (Oppert et al., 1997) of pH 7.9, which is the
optimal pH for proteolytic activity in mid-gut extract
from Hessian y larvae (data not shown). Inhibitors
were preincubated with the gut extract (0.8 gut
equivalents) for 10 min at 37 1C before the addition of
substrate. At time 0, 10 ml of a stock solution of 10 mg/ml
of uorescent-labeled casein (BODIPY-TR-X casein,
Molecular Probes, Eugene, OR) was added to each well
to initiate the reaction. Inhibition was measured after
4 h incubation at 37 1C using a uorescent microplate
reader (Fluoroskan Ascent FL, Labsystems, Thermo
Electron Corp., Milford, MA), with an excitation of
584 nm and emission of 620 nm. The percentage inhibi-
tion was calculated from incubations containing
the inhibitor divided by the control incubations without
the inhibitor, and multiplied by 100. The amount of the
inhibitor resulting in 50% inhibition (IC
50
) was calcu-
lated by linear regression from linear data points.
3. Results
In order to isolate genes that encode potential
digestive proteases, we sequenced 1014 random clones
from a gut cDNA library derived from rst instarlar-
vae. GenBank searching with the cDNA sequences
revealed 55 clones (or 5.4%) encoding various proteases.
Among these 55 clones, 27 coded for chymotrypsin- or
trypsin-like proteins, 19 coded for carboxypeptidases,
six for signal peptidases, two for cycteine proteases
(Cathepsin L), and one for lysosomal aspartic protease.
3.1. Chymotrypsin- and trypsin-like cDNAs
Of the 27 clones (all except one were full length) that
encoded chymotrypsin- or trypsin-like proteins, 15
cDNAs were unique clones that encode different
proteins. The GenBank-accession numbers for the
unique clones, lengths of the cDNAs and predicted
proteins, estimated molecular masses, pIs, putative
secretion signal peptides, and activation peptides are
listed in Table 1.
Sequence comparison indicated that these 15 cDNA
clones could be sorted into ve groups, which were
ARTICLE IN PRESS
Y.C. Zhu et al. / Insect Biochemistry and Molecular Biology 35 (2005) 2332 25
named MDP1, MDP2, MDP3, MDP4, and MDP5,
respectively. The MDP1 and MDP2 groups encode
chymotrypsin-like proteins whereas the other three
encode trypsin-like proteins. Members from the same
group share more than 95% sequence identity at both
nucleotide and amino acid levels (Table 2), except
chymotrypsin MDP1-C, which has only 8890% se-
quence identity compared with the other member in the
same group at the nucleotide level. The lower sequence
identity for chymotrypsin MDP1-C was caused by a
short 3
0
UTR showing 80 bp insertion/deletion (indel)
(data not shown). Excluding this indel, sequence identity
is also more than 95% between MDP1-C and other
members in this group. Members from different groups
share little sequence similarity at the nucleotide
level. No sequence alignments between members from
different groups could be found using BLASTN (data
not shown). Pairwise comparison of the nucleotide
sequences between members from different groups with
ClustalW (Gap open=10; Gap extension=0.1) again
revealed only low- (ranging from 47 to 62%) sequence
identity. At the amino acid level, there were conserva-
tions around regions that are common to serine
proteases between members from different groups
(Table 2 and see Section 3.2).
3.2. Structure of the deduced proteins
Sequence comparison of the fteen deduced protease
precursors is shown in Fig. 1. Forty-six residues are
conserved among all sequences. Most of the conserved
residues are located in the vicinity of the histidine and
ARTICLE IN PRESS
Table 1
Serine proteinase cDNAs and putative proteins from the gut of M. destructor
Clone name GenBank
number
cDNA (bp) Encoded
residues
Molecular
mass(kDa)
pI Signal peptide Activation
peptide
Putative
proteinase
MDP1A AY596471 1046 275 26.71 5.44 19 8 Chymotrypsin
MDP1B AY596472 1049 275 26.81 5.44 19 8 Chymotrypsin
MDP1C AY665658 953 275 26.78 5.44 19 8 Chymotrypsin
MDP1D AY665659 1045 275 26.78 5.55 19 8 Chymotrypsin
MDP1E AY665660 1068 275 26.81 5.44 19 8 Chymotrypsin
MDP1F AY665661 1048 275 26.78 5.44 19 8 Chymotrypsin
MDP2A AY596478 1052 269 26.24 8.66 21 10 Chymotrypsin
MDP2B AY596479 1052 269 26.25 8.66 19 12 Chymotrypsin
MDP3A AY596476 1116 268 26.11 8.41 17 14 Trypsin
MDP3B AY596475 1065 268 26.12 7.72 17 14 Trypsin
MDP3C AY665662 1121 268 26.09 7.72 17 14 Trypsin
MDP4A AY596477 1077 273 25.76 5.79 23 13 Trypsin
MDP4B AY596473 1075 273 25.71 5.79 23 13 Trypsin
MDP4C AY596474 1078 273 25.76 5.79 23 13 Trypsin
MDP5A AY669864 1147 263 25.67 5.75 17 13 Trypsin
Table 2
Pairwise comparison of cDNA and [putative protein] sequence identity (showing as % in upper-right corner) and phylogenetic distances (showing in
lower-left corner) among the 15 serine proteinase clones
Clones MDP1A MDP1B MDP1C MDP1D MDP1E MDP1F MDP2A MDP2B MDP3A MDP3B MDP3C MDP4A MDP4B MDP4D MDP5A
MDP1A 99[99] 90[99] 99[99] 97[99] 99[99] 48[25] 48[25] 51[28] 51[28] 51[28] 50[32] 50[31] 50[31] 49[25]
MDP1B 0.01 90[99] 99[99] 98[99] 99[99] 48[25] 48[25] 50[28] 51[28] 50[28] 50[31] 50[31] 50[31] 49[25]
MDP1C 0.02 0.01 90[99] 88[99] 90[99] 47[25] 47[25] 48[27] 49[28] 47[28] 50[31] 49[31] 50[31] 47[24]
MDP1D 0.02 0.01 0.01 97[99] 99[99] 48[25] 48[25] 50[28] 51[28] 50[28] 51[31] 51[31] 51[31] 49[25]
MDP1E 0.02 0.01 0.01 0.01 97[99] 48[25] 48[25] 50[28] 50[28] 50[28] 50[31] 50[31] 50[31] 49[25]
MDP1F 0.02 0.01 0.01 0.01 0.01 48[25] 48[25] 51[28] 51[28] 50[28] 51[31] 51[31] 51[31] 49[25]
MDP2A 2.65 2.66 2.67 2.66 2.68 2.66 98[99] 50[36] 51[37] 51[37] 52[40] 52[39] 52[40] 53[37]
MDP2B 2.67 2.67 2.68 2.67 2.69 2.67 0.01 51[36] 51[36] 51[36] 51[40] 52[39] 52[40] 53[38]
MDP3A 2.37 2.38 2.39 2.38 2.38 2.38 1.94 1.97 95[99] 99[99] 58[48] 58[48] 58[48] 62[60]
MDP3B 2.36 2.36 2.37 2.36 2.37 2.36 1.94 1.97 0.01 95[99] 58[48] 58[48] 58[48] 61[60]
MDP3C 2.36 2.36 2.37 2.36 2.37 2.36 1.94 1.97 0.02 0.01 57[48] 57[48] 57[48] 62[60]
MDP4A 2.17 2.17 2.18 2.17 2.18 2.17 1.77 1.79 1.40 1.38 1.38 99[99] 99[99] 57[49]
MDP4B 2.19 2.19 2.20 2.19 2.20 2.19 1.79 1.81 1.41 1.40 1.40 0.01 99[99] 57[49]
MDP4C 2.17 2.19 2.20 2.19 2.20 2.19 1.77 1.79 1.41 1.40 1.40 0.01 0.01 57[49]
MDP5A 2.50 2.51 2.53 2.50 2. 52 2.51 1.88 1.89 0.94 0.91 0.93 1.33 1.35 1.35
Y.C. Zhu et al. / Insect Biochemistry and Molecular Biology 35 (2005) 2332 26
serine active site, including a highly conserved N-
terminus (IVGG or VIGG preceded by an arginine or
lysine), which marks the N-termini of the active enzyme
(Jany and Haug, 1983; Wang et al., 1993). The three
critical residues, His
87
, Asp
136
, and Ser
241
(numbers are
assigned according to the multiple alignments) are all
conserved to form the catalytic triad that is character-
istic of serine proteases (Kraut, 1977; Wang et al., 1993;
Peterson et al., 1994). Other important residues,
including D
235
, G
265
, and G
276 (~)
, which dene the
substrate binding pocket, also are conserved in all
trypsin-like proteins encoded by the MDP3 to MDP5
ARTICLE IN PRESS
Fig. 1. Multiple alignment of the predicted amino acid sequences of the 15 proteinase-like precursors from M. destructor. Putative secretion signal
peptides were bold and underlined. Functionally important residues H, D, and S (active sites) were boxed. Cysteines corresponding to the sites of the
predicted disulde bridges were marked with arrow lines at the bottom. Trypsin specicity determinant residues were indicated with (~) on the top of
sequences. Identical residues among all fteen sequences are highlighted with gray background. Hyphens represent sequence alignment gaps.
Y.C. Zhu et al. / Insect Biochemistry and Molecular Biology 35 (2005) 2332 27
gene groups. The D
235
residue determines specicity in
both invertebrate and vertebrate trypsins by stabilizing
the Lys or Arg residue at the substrate cleavage site
through ionic interactions (Hedstrom et al., 1992; Wang
et al., 1993). The D
235
residue was replaced by either a
Gly or Ser residue in all the chymotrypsin-like proteins
encoded by the MDP1 or MDP2 gene groups.
Among all the deduced proteins, there are six cysteine
residues that are conserved at positions 72, 88, 206, 225,
237, and 269 (marked by arrows). These cysteine
residues are predicted to occur in disulde bridge
congurations in trypsins and chymotrypsins (Wang
et al., 1993).
3.3. Differential expression in different developmental
stages
To examine the expression proles of different
proteases in different developmental stages, Northern
blot analysis was performed with RNA samples derived
from larvae (at different ages), pupae, and adults. The
blots were examined separately for each gene group with
probes specic for each. As shown in Fig. 2, all of the
proteases were exclusively expressed in the larval stage.
No RNA could be detected in pupae or adults. Within
the larval stage, the RNA levels corresponding to each
gene group differed signicantly in larvae of different
ages. The MDP1 gene group (chymotrypsin MDP1-A to
MDP1-F) was expressed in all larval stages with the
maximum amount expressed in 4-day-old larvae. In
comparison, the MDP2 gene group (chymotrypsin
MDP2-A and MDP2-B) was mainly expressed in 4-
and 6-day-old larvae. The MDP3 gene group (trypsin
MDP3-A to MDP3-C) was expressed at a low level in 2-
days old larvae, signicantly increased in 4- and 6-days
old larvae, and then disappeared in 12-day-old larvae.
The expression for the MDP4 gene group (trypsin
MDP4-A to MDP4-C) was unique. It was highly
expressed in 0- (freshly hatched) and 12-day-old larvae,
but the expression level was signicantly less in 2- and
4-day-old larvae. The MDP5 gene group (trypsin
MDP5-A) was abundantly expressed in freshly hatched
larvae, but no RNA could be detected in larvae after
4 days.
3.4. Analysis of tissue-specic expression through RT-
PCR
RT-PCR was carried out to examine tissue-specic
expression of the ve gene groups in the gut and salivary
glands. All PCR reactions generated DNA fragments
with expected sizes. As shown in Fig. 3, the majority of
the protease groups were expressed in both the gut and
salivary glands. The only group that was specically
expressed in the gut was the trypsin MDP5 group.
3.5. Protease activity in gut extract
The isolation of trypsin- and chymotrypsin-like
cDNAs from a gut library, and subsequent detection
of transcripts in gut by RT-PCR, indicated that serine
proteinases were major digestive enzymes in M.
destructor larvae. To further determine the composition
of gut proteases, specic protease inhibitors were used to
categorize proteolytic activities contributed by different
digestive enzymes. From the data in Table 3 and Fig. 4,
serine proteases were the major digestive enzymes
because serine protease inhibitors inhibited as much as
90% of the proteolytic activities of gut extract.
Inhibitors specic to cysteine proteases, aspartic pro-
teases, or metallo-proteases were less effective, causing
only 1024% inhibition. One exception was the inhibitor
bestatin, which caused nearly 90% inhibition. Bestatin
ARTICLE IN PRESS
Fig. 2. Northern blot analysis. RNA samples were extracted from 0-
day (freshly hatched, nonfeeding), 2-day, 4-day, 6-day and 12-day-old
larvae, pupae (P), and adults (A), respectively, as indicated in the
gure. Probes specic to MDP1, MDP2, MDP3, MDP4, and MDP5
gene groups were used for blot analysis. Hybridization and washing
were performed as described in the Materials and Methods section.
The bottom panel of the gure shows the hybridization image of a
probe derived from an 18-S rRNA cDNA clone as a control for equal
loading.
Y.C. Zhu et al. / Insect Biochemistry and Molecular Biology 35 (2005) 2332 28
inhibits metalloproteases, aminopeptidase, and exopep-
tidases (Taylor et al., 1993).
4. Discussion
In this report, we have identied ve groups of cDNA
clones that encode chymotrypsin- or trypsin-like pro-
teins. Members from the same group encode similar
proteins. These group members could represent different
alleles since the cDNA library was made from multiple
insects. If that is true, these cDNAs could be converted
into molecular markers for genetic mapping. Different
cDNAs from the same group could also represent
different genes that arose by gene duplication. Indeed,
most of the genes encoding secreted proteins in the
salivary glands or gut of the Hessian y have multiple
copies, which are clustered within a short chromosome
region (Chen et al., 2004). Clustered organization of
trypsin genes has been reported in A. gambiae (Muller
et al., 1993). Different cDNAs in the same group from
M. destructor could be recent duplicates in the early
phase for diversication.
In contrast to highly conserved sequences within the
same group, members from different groups encode very
different proteins. Sequence identities between proteins
ARTICLE IN PRESS
Fig. 3. Analysis of tissue-specic expression through RT-PCR analysis. Primer names are given at bottom of corresponding lanes. M=100 bp
molecular markers.
Table 3
Effect of protease inhibitors on the hydrolysis of BODIPY conjugated casein by M. destructor midgut extract
Different class of inhibitors Concentration BODIPY-TR-X hydrolysis (% of control) IC
50
Serine protease inhibitors
Aprotinin 13.5 mM 19.72 2.38 mM
APMSF 0.5 mM 52.25 0.51 mM
Chymostatin 5.0 mM 16.97 82 mM
SBTI 0.03 mM 11.63 2.2 mM
SBBTI 1.5 mM 10.43 1.2 mM
TPCK 0.1 mM 18.41 12 mM
Cysteine protease inhibitors
Antipain 0.175 mM 76.01
E-64 0.1 mM 89.0
Aspartic protease inhibitor
Pepstatin 0.05 mM 80.93
Metallo protease inhibitors
Bestatin 2.8 mM 10.54 0.83 mM
EDTA 0.05 mM 135.0
EGTA 5.0 mM 90.59
Y.C. Zhu et al. / Insect Biochemistry and Molecular Biology 35 (2005) 2332 29
from different groups ranged from 24% to 60%
(Table 2). On the other hand, sequence identity ranged
from 44% to 51% between these proteins and serine
proteases from other insects (data not shown). The
greater homology between proteases from different
species (M. destructor and A. gambiae, for example)
than between proteins from the same species indicates
that these genes are orthologues that existed before the
divergence of the two species.
The deduced amino acid sequences of the 15 proteins
showed several structural features typical of serine
proteases (Kraut, 1977), including: (i) the amino acid
residues considered to determine trypsin or chymotryp-
sin substrate specicities; (ii) conserved histidine and
serine catalytic sites; (iii) the catalytic triad; and (iv) six
cysteine residues, all located at conserved positions. All
these features suggested that these cDNAs encode active
trypsins or chymotrypsins. All deduced proteins con-
tained at their N-termini a typical secretion signal
peptide consisting of 1723 amino acids (Table 2,
Fig. 1). Between the predicted signal peptides and the
mature enzyme sequences, all deduced proteins con-
tained a stretch of 814 amino acids characteristic of
putative activation peptides. In addition to the three
pairs of conserved cysteines that are believed to form
disulde bonds, additional cysteine residues were also
found in some of the deduced proteins. These extra
cysteine residues included one at position 170 in all
deduced trypsins and one at position 247 in the
chymotrypsin MDP2 group (Fig. 1). In addition, all
proteinase precursors, except for the trypsin MDP4
group, had at least one cysteine on the putative signal or
activation peptide. Extra cysteine residues have been
found in vertebrate, as well as in some insect serine
proteases (Colebatch et al., 2002; Jongsma et al., 1996;
Zhu and Baker, 1999; Zhu et al., 2003). It has been
speculated that free cysteine residues might be respon-
sible for the vulnerability of the enzymes to some typical
thio-reacting inhibitors (Jongsma et al., 1996; Gate-
house et al., 1997). The exact role and the pairing mode
for disulde bonding of these extra cysteines have yet to
be established.
Different insects use different types of proteases as
major digestive enzymes (Terra and Ferreira, 1994).
Digestive cysteine proteinases constitute the major gut
proteolytic activity in many coleopterans (Murdock
et al., 1987; Matsumoto et al., 1997). Proteinase
activity assay using various inhibitors demonstrated
that serine proteinases were the major digestive enzymes
in the gut of M. destructor larvae. As shown in Table 3
and Fig. 4, cysteine proteinase inhibitors reduced the
proteinase activity of gut extract only marginally
(10%), but serine proteinase inhibitors suppressed
gut proteinase activity as much as 90%. This observa-
tion is consistent with our transcriptomic analysis,
from which we have found only two identical cysteine
proteinase clones compared with 27 chymotrypsin
or trypsin clones. We are currently generating anti-
bodies which will be used for assays to establish the
connections between the detected serine protease activ-
ity in the gut and the specic protease genes we have
identied.
ARTICLE IN PRESS
0
10
20
30
40
50
60
70
80
90
100
110
0.0000010.00001 0.0001 0.001 0.01 0.1 10
Inhibitor Concentration (mM)
APMSF (0.5mM)
Aprotinin (0.0135 mM)
Bestatin (2.8mM)
Chymostatin (5.0mM)
SBTI (0.03 mM)
SBBTI (1.5mM)
TPCK (0.1mM)
1
P
e
r
c
e
n
t

o
f

C
o
n
t
r
o
l
Fig. 4. Casein hydrolysis by M. destructor larval gut extract in the presence of various protease inhibitors.
Y.C. Zhu et al. / Insect Biochemistry and Molecular Biology 35 (2005) 2332 30
Northern blot analysis revealed that all of the
proteases were exclusively expressed in the larval stage
(Fig. 2). Considering the fact that adults of this insect do
not feed, the exclusive expression of these genes in the
larval stage indicated a role in protein digestion of
ingested food. Although all of the protease genes were
expressed in the larval stage, the transcriptional control
of these genes seemed to be quite different. The
expression level of RNA corresponding to different
groups differed greatly in larvae at different ages. In
addition, the abundance of the transcripts detected by
probes specic to individual gene groups also differed in
different larval stages. Differential regulation of similar
proteases was also observed in other insects (Muller et
al., 1993), but the biological implication of differential
expression remains to be resolved. RT-PCR analysis
revealed that all of these gene groups except MDP5 were
expressed in both the gut and salivary glands (Fig. 3).
M. destructor larvae have sucking mouthparts and these
proteinases may also be injected into host plant tissue
for partial digestion before sucking the juice up (Hase-
man, 1930; Hatchett et al., 1990; Cohen, 1998). The
function of diversied serine proteinase genes and
differential expression of these genes in different
developmental stages remains to be characterized.
Further research will be also needed on the expression
of different proteinase genes in response to feeding and
alteration in host plants and on the development
of techniques to manipulate protein digestion in
M. destructor larvae.
Acknowledgements
This is Contribution No. 05-34-J from the Kansas
Agricultural Experiment Station, Manhattan, Kansas.
Hessian y voucher specimens (No. 150) are located in
the KSU Museum of Entomological and Prairie
Arthropod Research, Kansas State University, Man-
hattan, Kansas. The authors want to thank Drs. Gerald
R. Reeck and Srini Kambhampati for reviewing an
earlier version of the manuscript.
References
Behura, S., Valicente, F.H., Rider Jr., S.D., Chen, M.S., Jackson, S.,
Stuart, J.J., 2004. Recombination suppression in the Hessian y
revealed by a genetic map and linkage to avirulence. Genetics 167,
343355.
Buntin, G.D., 1999. Hessian y (Diptera: Cecidomyiidae) injury and
loss of winter wheat grain yield and quality. J. Econ. Entomol. 92
(5), 11901197.
Burgess, E.P.J., Main, C.A., Stevens, P.S., Christeller, J.T., Gatehouse,
A.M.R., Laing, W.A., 1994. Effects of protease inhibitor concen-
tration and combinations on the survival, growth and gut enzyme
activities of the black eld cricket, Teleogryllus commodus.
J. Insect Physiol. 40 (9), 803811.
Chen, M.S., Fellers, J.P., Stuart, J.J., Reese, J.C., Liu, X., 2004. A
group of related cDNAs encoding secreted proteins from Hessian
y [Mayetiola destructor (Say)] salivary glands. Insect Mol. Biol. 13
(1), 101108.
Cohen, A.C., 1998. Solid-to-liquid feeding: the inside(s) story of extra-
oral digestion in predaceous Arthropoda. Am. Entomol. 44,
103117.
Colebatch, G., Cooper, P., East, P., 2002. cDNA cloning of a salivary
chymotrypsin-like protease and the identication of six additional
cDNAs encoding putative digestive proteases from the green mirid,
Creontiades dilutus (Hemiptera: Miridae). Insect Biochem. Mol.
Biol. 32, 10651075.
El Bouhssini, M., Hatchett, J.H., Wilde, G.E., 1998. Survival of
Hessian y (Diptera: Cecidomyiidae) larvae on wheat cultivars
carrying different genes for antibiosis. J. Agric. Entomol. 15 (3),
183193.
Gagne, R.J., Hatchett, J.H., 1989. Instars of the Hessian y (Diptera:
Cecidomyiidae). Ann. Entomol. Soc. Am. 82 (1), 7379.
Gallun, R.L., Langston, R., 1963. Feeding habits of Hessian y larvae
on P32-labeled resistance and susceptible wheat seedlings. J. Econ.
Entomol. 56 (5), 702706.
Gatehouse, L.N., Shannon, A.L., Burgess, E.P.J., Christeller, J.T.,
1997. Characterization of major midgut proteinase cDNAs from
Helicoverpa armigera larvae and changes in gene expression in
response to four proteinase inhibitors in the diet. Insect Biochem.
Mol. Biol. 27, 929944.
Hall, T.A. (Ed.), 1999. BioEdit: a user-friendly biological sequence
alignment editor and analysis program for Windows 95/98/NT,
Nucl. Acids. Symp. Ser. 41, 9598.
Harris, M.O., Rose, S., 1989. Temporal changes in egglaying behavior
of the Hessian y. Entomol. Exp. Appl. 53 (1), 1729.
Harris, M.O., Stuart, J.J., Mohan, M., Nair, S., Lamb, R.J.,
Rohfritsch, O., 2003. Grasses and gall midges: plant defense and
insect adaptation. Ann. Rev. Entomol. 48, 549577.
Haseman, L., 1930. The Hessian y larva and its method of taking
food. J. Econ. Entomol. 23 (2), 316319.
Hatchett, J.H., Gallun, R.L., 1970. Genetics of the ability of Hessian
y, Mayetiola destructor, to survive on wheats having different
genes for resistance. Ann. Entomol. Soc. Am. 63 (5), 14001407.
Hatchett, J.H., Starks, K.J., Webster, J.A., 1987. Insect and mite pests
of wheat. In: Wheat and Wheat improvement. Agronomy
Monograph No. 13, pp. 625675.
Hatchett, J.H., Kreitner, G.L., Elzinga, R.J., 1990. Larval mouthparts
and feeding mechanism of the Hessian y (Diptera: Cecidomyii-
dae). Ann. Entomol. Soc. Am. 83 (6), 11371147.
Hedstrom, L., Szilagyi, L., Rutter, W.J., 1992. Converting trypsin to
chymotrypsin: the role of surface loops. Science 255, 12491253.
Herrero, S., Oppert, B., Ferre, J., 2001. Different mechanisms of
resistance to Bacillus thuringiensis toxins in the indianmeal moth.
Appl. Environ. Biol. 67 (3), 10851089.
Huang, Y., Nordeen, R.O., Di, M., Owens, L.D., McBeath, J.H., 1997.
Expression of an engineered cecropin gene cassette in transgenic
tobacco plants confers disease resistance to Pseudomonas syringae
pv. tabaci. Phytopathology 87 (5), 494499.
Jang, C.S., Kim, J.Y., Haam, J.W., Lee, M.S., Kim, D.S., Li, Y.W.,
Seo, Y.W., 2003. Expressed sequence tags from a wheat-
rye translocation line (2BS/2RL) infested by larvae of
Hessian y [Mayetiola destructor (Say)]. Plant Cell Rep. 22 (2),
150158.
Jany, K.D., Haug, H., 1983. Amino acid sequence of the chymotryptic
protease II from the larvae of the hornet, Vespa crabro. FEBS Lett.
158, 98102.
Jongsma, M.A., Peters, J., Stiekema, W.J., Bosch, D., 1996.
Characterization and partial purication of gut proteinase of
Spodoptera exigua Hu bner (Lepidoptera: Noctuidae). Insect
Biochem. Mol. Biol. 26, 185193.
ARTICLE IN PRESS
Y.C. Zhu et al. / Insect Biochemistry and Molecular Biology 35 (2005) 2332 31
Karban, R., Baldwin, I.T., 1997. In: Karban, Baldwin (Eds.), Induced
Response to Herbivory. University of Chicago Press, Chicago, IL,
pp. 108115.
Kraut, J., 1977. Serine proteases: structure and mechanism of catalysis.
Ann. Rev. Biochem. 46, 331358.
Martin-Sanchez, J.A., Gomez-Colmenarejo, M., Del Moral, J., Sin, E.,
Montes, M.J., Gonzalez-Belinchon, C., Lopez-Brana, I., Delibes,
A., 2003. A new Hessian y resistance gene (H30) transferred from
the wild grass Aegilops triuncialis to hexaploid wheat. Theor. Appl.
Genet. 106 (7), 12481255.
Matsumoto, I., Emori, Y., Abe, K., Arai, S., 1997. Characterization of
a gene family encoding cysteine proteinases of Sitophilus zeamais
(Maise weevil), and analysis of the protein distribution in various
tissues including alimentary tract and germ cells. J. Biochem. 121,
464476.
Moura, D.S., Ryan, C.A., 2001. Wound-inducible proteinase inhibi-
tors in pepper. Differential regulation upon wounding, systemin,
and methyl jasmonate. Plant Physiol. 126 (1), 289298.
Muller, H.-M., Crampton, J.M., Torre, A.D., Sinden, R., Crisanti, A.,
1993. Members of a trypsin gene family in Anopheles gambiae are
induced in the gut by blood meal. EMBO J. 12 (7), 28912900.
Murdock, L.L., Brookhart, G., Dunn, P.E., Foard, D.E., Kelley, S.,
Kitch, L., Shade, R.E., Shukle, R.H., Wolfson, J.L., 1987. Cysteine
digestive protenases in Coleoptera. Comp. Biochem. Physiol. 87B,
783787.
Murdock, L.L., Shade, R.E., Pomeroy, M.A., 1988. Effects of E-64, a
cysteine proteinase-inhibitor, on cowpea weevil growth, develop-
ment, and fecundity. Environ. Entomol. 17, 467469.
Oppert, B., Kramer, K.J., McGaughey, W.H., 1997. Rapid microplate
assay of proteinase mixtures. BioTechnology 23, 7072.
Peterson, A.M., Barillas-Mury, C.V., Wells, M.A., 1994. Sequence of
three cDNAs encoding an alkaline midgut trypsin from Manduca
sexta. Insect Biochem. Mol. Biol. 24, 463471.
Ratcliffe, R.H., Hatchett, J.H., 1997. Biology and genetics of the
Hessian y and resistance in wheat. In: Bondari, K. (Ed.), New
Developments in Entomology. Research Signpost, Scientic
Information Guild, Trivandurm, India. pp. 4756.
Ratcliffe, R.H., Safranski, G.G., Patterson, F.L., Ohm, H.W., Taylor,
P.L., 1994. Biotype status of Hessian y (Diptera: Cecidomyiidae)
populations from the eastern United States and their response to 14
Hessian y resistance genes. J. Econ. Entomol. 87 (4), 11131121.
Ratcliffe, R.H., Cambron, S.E., Flanders, K.L., Bosque-Perez, N.A.,
Clement, S.L., Ohm, H.W., 2000. Biotype composition of Hessian
y (Diptera: Cecidomyiidae) populations from the southeastern,
Midwestern, and northwestern United States and virulence to
resistance genes in wheat. J. Econ. Entomol. 93 (4), 13191328.
Ratcliffe, R.H., Ohm, H.W., Patterson, F.L., 2003. Breeding wheat for
Hessian y resistance. Plant Breed. Rev. 22, 221296.
Rider Jr., S.D., Sun, W., Ratcliffe, R.H., Stuart, J.J., 2002.
Chromosome landing near avirulence gene vH13 in the Hessian
y. Genome 45 (5), 812822.
Shukle, R.H., Murdock, L.L., Gallun, R.L., 1985. Identication and
partial characterization of a major gut proteinase from larvae of
the Hessian y, Mayetiola destructor (Say) (Deptera: Cecidomyii-
dae). Insect Biochem. 15 (1), 93101.
Taylor, A., Peltier, C.Z., Torre, F.J., Hakamian, N., 1993. Inhibition
of bovine lens leucine aminopeptidase by bestatin: number of
binding sites and slow binding of this inhibitor. Biochemistry 32
(3), 784790.
Terra, W.R., Ferreira, C., 1994. Insect digestive enzymes: properties,
compartmentalization and function. Comp. Biochem. Physiol.
109B (1), 162.
Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTAL W:
improving the sensitivity of progressive multiple sequence align-
ment through sequence weighting, positions-specic gap penalties
and weight matrix choice. Nucl. Acid Res. 22, 46734680.
Wang, S., Magoulas, C., Hickey, D.A., 1993. Isolation and character-
ization of a full-length trypsin-encoding cDNA clone from the
lepidopteran insect, Choristoneura fumiferana. Gene 136, 375376.
Williams, C.E., Collier, C.C., Nemacheck, J.A., Liang, C., Cambron,
S.E., 2002. A lectin-like wheat gene responds systemically to
attempted feeding by avirulent rst-instar Hessian y larvae.
J. Chem. Ecol. 28 (7), 14111428.
Williams, C.E., Collier, C.C., Sardesai, N., Ohm, H.W., Cambron,
S.E., 2003. Phenotypic assessment and mapped markers for H31, a
new wheat gene conferring resistance to Hessian y (Diptera:
Cecidomyiidae). Theor. Appl. Genet. 107 (8), 15161523.
Zhu, Y.C., Baker, J.E., 1999. Characterization of midgut trypsin-like
enzyme and molecular cloning of three trypsinogen cDNAs from
lesser grain borer, Rhyzopertha dominica (Coleoptera: Bostrichi-
dae). Insect Biochem. Mol. Biol. 29, 10531063.
Zhu, Y.C., Zeng, F., Oppert, B., 2003. Molecular cloning of
trypsin-like cDNAs and comparison of proteinase activities in
the salivary glands and gut of the tarnished plant but Lygus
lineolaris (Heteroptera: Miridae). Insect Biochem. Mol. Boil. 33,
889899.
ARTICLE IN PRESS
Y.C. Zhu et al. / Insect Biochemistry and Molecular Biology 35 (2005) 2332 32
Insect
Biochemistry
and
Molecular
Biology
Insect Biochemistry and Molecular Biology 35 (2005) 3340
Identication, cloning and expression of a Cry1Ab cadherin
receptor from European corn borer, Ostrinia nubilalis (Hu bner)
(Lepidoptera: Crambidae)
Ronald D. Flannagan
a
, Cao-Guo Yu
a
, John P. Mathis
a
, Terry E. Meyer
a
, Xiaomei Shi
a
,
Herbert A.A. Siqueira
b
, Blair D. Siegfried
b,
a
Pioneer Hi-Bred International Inc., 7301 NW 62nd Avenue, P.O. Box 85, Johnston, IA 50131-0085, USA
b
Department of Entomology, 202 Plant Industry Building, University of Nebraska, Lincoln, NE 68583-0816, USA
Received 8 July 2004; received in revised form 4 October 2004; accepted 7 October 2004
Abstract
Transgenic corn expressing the Cry1Ab toxin from Bacillus thuringiensis is highly toxic to European corn borer, Ostrinia nubilalis,
larvae. A putative Cry1Ab receptor (OnBt-R
1
) molecule was cloned and sequenced from a cDNA library prepared from midgut
tissue of O. nubilalis larvae. The 5.6 Kb gene is homologous with a number of cadherin genes identied as Cry1 binding proteins in
other lepidopterans. Brush border membrane vesicles were prepared using dissected midguts from late instars. A 220-kDa protein
was identied as a cadherin-like molecule, which bound to Cry1Ab toxin and cross-reacted with an anti-cadherin serum developed
from recombinant expression of a partial O. nubilalis cadherin peptide. Two additional proteins of smaller size cross-reacted with the
anti-cadherin serum indicating that Cry1Ab binds to multiple receptors or to different forms of the same protein. Spodoptera
frugiperda (SF9) cells transfected with the OnBt-R
1
gene were shown to express the receptor molecule which caused functional
susceptibility to Cry1Ab at concentrations as low as 0.1 mg/ml. These results in combination suggest strongly that a cadherin-like
protein acts as receptor and is involved with Cry1Ab toxicity in O. nubilalis.
r 2004 Elsevier Ltd. All rights reserved.
Keywords: Lepidoptera; Bacillus thuringiensis; Binding analysis; Cadherin-like protein; Midgut; Cry toxins
1. Introduction
Bacillus thuringiensis (Berliner) (Bt) is a Gram-
positive, spore-forming bacterium that produces crystal-
line inclusion bodies during sporulation that contain
insecticidal d-endotoxins. Lepidopteran insects are
particularly susceptible to Cry1 toxins, which bind
specically to midgut receptors and are highly toxic
after ingestion. Solubilization of the crystal releases a
130-kDa protoxin, which is activated by proteases in the
insect midgut to form the truncated 65-kDa toxin. The
target of the activated toxin is the apical (brush border)
membrane of larval midgut cells (Bravo et al., 1992).
Binding of the activated toxin to midgut-specic
receptors causes the toxin conformation change, which
allow its insertion and formation of ion channels or
pores in the midgut apical membrane, leading to
osmotic imbalance of the insect gut (Gill et al., 1992;
Knowles, 1994; Schnepf et al., 1998).
Cry1-binding proteins detected on ligand blots of
insect brush border membrane vesicles (BBMV) have
been identied as members of the aminopeptidase N and
cadherin families although the relative role of the two
putative receptor molecules in insects has yet to be
conclusively determined. A 210-kDa cadherin-like gly-
coprotein has been identied as a Cry1Ab binding
protein in BBMV prepared from the midguts of
Manduca sexta larvae (Vadlamudi et al., 1993, 1995).
ARTICLE IN PRESS
www.elsevier.com/locate/ibmb
0965-1748/$ - see front matter r 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ibmb.2004.10.001

Corresponding author. Tel.: +1 402 472 8714;


fax: +1 402 472 4687.
E-mail address: bsiegfried1@unl.edu (B.D. Siegfried).
Although initially detected with Cry1Ab, other toxins
such as Cry1Aa and Cry1Ac also bind the cadherin-like
protein. In Bombyx mori, a 175 kDa cadherin-like
protein was identied as a Cry1Aa binding
protein (Nagamatsu et al., 1998a, b). Gahan et al.
(2001) reported that Cry1Ac resistance in the tobacco
budworm, Heliothis virescens, was tightly linked to a
cadherin-encoding gene but not to genes encoding
aminopeptidases. More recently, Morin et al.
(2003) reported three different cadherin alleles from
the pink bollworm, Pectinophora gossypiella, linked with
resistance to Cry1Ac and survival on transgenic Bt
cotton.
An epitope involved in Cry toxinreceptor interac-
tions has been identied in a cadherin-like protein from
M. sexta (Gomez et al., 2001). Previously, Nagamatsu et
al. (1999) determined a region in the B. mori cadherin
BtR
175
which included the cadherin domain 9 and part
of the membrane proximal region (MPR) that bound to
Cry1Aa. Likewise, a region comprising the cadherin
domain 11 in M. sexta, also adjacent to the membrane-
proximal extracellular domain, was shown to bind
Cry1A toxins (Dorsch et al., 2002). Both binding
regions in these homologous proteins differed from that
reported by Gomez et al. (2001). In addition to these
studies, Hua et al. (2004) showed that both cadherin
domains 11 and 12 from BtR
1
are important for Cry1A
toxin binding, but that binding occurs rst to domain
12, which mediates the subsequent binding to domain
11.
Transgenic corn expressing the Cry1Ab toxin has
been deployed for control of the European corn borer,
Ostrinia nubilalis, and has become an important
component of corn production systems throughout the
US. Cry1Ab has been shown to recognize a single
population of binding sites on the brush border
epithelium of O. nubilalis larvae (Denolf et al., 1993).
Additionally, Hua et al. (2001) identied the presence of
both aminopeptidases and a cadherin protein from the
BBMVs, speculating that isoforms of both aminopepti-
dase and cadherin in the brush border membrane serve
as binding proteins. The present study reports the
cloning and expression of a cDNA that encodes a
cadherin-like protein (OnBt-R
1
) present in the midgut of
O. nubilalis larvae. The receptor binds the Cry1Ab
protein and is believed to be the major factor in
mediating Cry1Ab toxicity in this insect.
2. Material and methods
2.1. Construction of cDNA library
Total RNA was extracted from O. nubilalis midgut
tissue and used to create a lambda phage library. Briey,
total RNA was isolated from 4th-larval stage midgut
tissue using the Messenger RNA Isolation Kit (Strata-
gene). The rst strand cDNA was synthesized using
StrataScript RT (Stratagene) and a poly(dT) oligo: 5
0
-
GAGAGAGAGAGAGAGAGAGAACTAGTCTCG-
AGTTTTTTTTTTTTTTTTTT-3
0
. Second strand
synthesis was accomplished via nick translation
after the addition of DNA polymerase I. The termini
of the cDNA were blunted by adding pfu
DNA polymerase (Stratagene) and the following adap-
ters containing EcoR I cohesive termini were ligated to
the cDNA: 5
0
-OH-AATTCGGCACGAGG-3
0
and 5
0
-
CCTCGTGCCGp-3
0
. The cDNA was digested with
Xho I and size fractionated. Fractions that contained
fragments X1 kb in length were ligated to the UNI-ZAP
XR arms and packaged using the Gigapack III Gold
packaging system (Stratagene).
2.2. cDNA cloning of OnBT-R
1
Based on the cDNA sequence for the M. sexta
cadherin (Vadlamudi et al., 1995), a pair of degenerative
primers were synthesized by Sigma Genosys for use in
PCR reactions: 2A: 5
0
-CTTGGAATTCGAACAT/
GTCCA/GTGC and 4S: 5
0
-TTGTACACAG/CGCA/
TGGG/CATA/TTCCAC. PCR reactions were per-
formed by standard techniques (Sambrook et al., 1989)
using Pwo DNA polymerase (Roche). PCR products
were cloned into pCR-Blunt II TOPO (Invitrogen) and
sequenced using an ABI 3700 capillary electrophoresis
unit and uorescent dye termination chemistry (Foster
City, CA).
The 280 base pair cadherin fragment generated by
PCR was then used to screen the O. nubilalis midgut
cDNA library previously described. Oligonucleotide
probes were 3
0
end-labeled with [a-
32
P] dCTP using
Rediprime
TM
II DNA Labeling System (Amersham
Biosciences) in accordance with manufacturer recom-
mendations. Approximately 6 10
9
recombinants were
screened and several clones were found to hybridize to
the probe. Positive clones were subjected to subsequent
rounds of screening and plaque-puried. The cDNA
from positive clones was sequenced in both directions by
dideoxy chain termination.
2.3. Sequencing protocol
Sequencing reactions were performed using 1/8th
reactions of v3.1 BigDye dye terminator chemistry
(Applied Biosystems, Foster City, CA) in 20 ml reaction
volumes under the following conditions: 10 s melting at
96 1C, 5 s annealing at 50 1C, 4 min extension at 60 1C for
25 cycles, followed by a 4 1C hold. Reactions were
precipitated with 30 ml of 100% EtOH and resuspended
in 30 ml dH
2
O prior to loading on Applied Biosystems
(Foster City, CA) 3700 capillary electrophoresis auto-
mated DNA analyzers, using the Pop5 polymer and a
ARTICLE IN PRESS
R.D. Flannagan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 3340 34
run time of 6500 s. ABI base calls were reanalyzed using
Phred software to assign quality values.
Pairwise sequence analyses with other cadherin-like
proteins were performed with Genetics Computer
Group (GCG, version 10, Madison, WI), using the
GAP function and default settings (gapweight=8,
lengthweight=2). Multiple alignments were performed
using ClustalX (gapweight=15, lengthweight=0.30)
(Thompson et al., 1997), followed by a ne-tuning
alignment of block sequences.
2.4. Antibody production
A PCR fragment was generated containing the
nucleotides that encode amino acids 9581503 of the
O. nubilalis cadherin gene. This fragment was subse-
quently cloned into the pET28 expression vector
(Novagen) and transformed into BL21-De3 cells (In-
vitrogen). Transformed cells were grown at 37 1C to an
OD
600
of 0.8, induced with IPTG (1 mM) and harvested
after 14 h of growth at 16 1C. The suspension was
centrifuged at 10,000g for 15 min and the supernatant
was removed. The cell pellet was suspended in 1/25
volume lysis buffer (5 mM Imidazole, 500 mM NaCl,
20 mM TrisHCl pH 7.9); 1 mg/ml lysozyme; Complete
Protease Inhibitor (Roche). The protein was puried
according to the His-Bind Kit protocol (Novagen),
except Talon Metal Afnity Resin (Clontech) was
substituted for the Ni
+2
resin provided in the kit. The
puried protein was dialyzed into PBS and concentrated
to 1 mg/ml. The concentrated protein was loaded onto
an SDS-PAGE gel and silver stained according to the
SilverQuest protocol (Invitrogen). The purity of the
OnBt-R
1
fragment was estimated at 495% and 1 mg of
protein was provided to Strategic Biosolutions for
polyclonal antibody construction and protein A pur-
ication using standard procedures.
2.5. Brush border membrane vesicles (BBMV)
preparation
BBMV were prepared with gut tissue dissected from
ice-chilled 5th stage larvae. Larvae were excised so that
both the last three abdominal segments and the head
plus thorax were removed, and the midguts were pulled
gently from the carcass. A small glass culture tube was
rolled over the length of the gut to displace the gut
contents. Dissected gut tissue was transferred to a
centrifuge tube containing ice-cold MET buffer
[300 mM Mannitol, 17 mM TrisHCl [pH 7.5], 5 mM
EGTA, protease inhibitor (complete EDTA-free pro-
tease inhibitors, Roche), vigorously vortexed and briey
centrifuged for 5 min at 1000g to obtain the clean
midguts. Guts were either frozen at 80 1C or processed
immediately by the differential magnesium precipitation
method of Wolfersberger et al. (1987). Briey, gut
tissues were homogenized on ice in a tight-tting glass
Dounce homogenizer in ice-cold MET buffer (10%
weight/volume). The homogenate was diluted with an
equal volume of ice-cold 28 mM MgCl
2
, blended and
held on ice for 15 min before centrifugation. A low-
speed centrifugation (2500g for 15 min at 4 1C) was used
to pellet heavier cell debris, and the supernatant from
the initial centrifugation was further centrifuged at
30,000g for 30 min at 4 1C. The resulting pellet was
resuspended in MET buffer and centrifuged again at
30,000g for 30 min at 4 1C. The resulting pellet which
corresponded to the BBMV preparation was resus-
pended in HBS-N buffer (10 mM Hepes, pH 7.4,
150 mM NaCl), ash frozen in liquid nitrogen and
stored at 80 1C. The protein concentration of the
BBMV preparations was determined by the bicincho-
ninic acid method (Smith et al., 1985). Alkaline
phosphatase activity was used as a marker enzyme to
track purication of the BBMV preparation, and it was
812 times higher in BBMV preparations than in the
initial homogenates (data not shown).
2.6. Immunoblotting
Ligand blots of Cry1Ab binding to BBMV proteins
were performed using the chemiluminescence Western
Light
TM
kit (Tropix, Inc., Bedford, MA). Equal
amounts (80 mg) of BBMV protein were separated by
SDS-PAGE as described by Laemmli (1970), electro-
blotted onto polyvinylidene diuoride (PVDF) mem-
brane (Bio-Rad Inc., Hercules, CA) for 90 min by a
Mini Trans-Blot Electrophoretic transfer cell (Bio-Rad),
and blocked for 2 h at room temperature with phosphate
buffered saline (PBS) (pH 8.0) containing 5% non-fat
dry milk powder, 5% glycerol, 0.5% Tween-20. The
PVDF membrane was then incubated with activated
Cry1Ab (250 ng/ml) in blocking buffer overnight at 4 1C
and subsequently washed three times with blocking
buffer. The blot was then incubated with polyclonal
rabbit anti-Cry1Ab (1:2500) (provided by Monsanto
Co., St. Louis, MO), washed three times, then incubated
with goat anti-rabbit-AP (2nd antibody at 1:10,000) and
washed three times with blocking buffer. The PVDF
membrane was washed with assay buffer (Tropix Inc.,
Bedford, MA) and detection was performed with the
CDP-Star
s
chemiluminescence Kit (Tropix Inc.) using a
Fluor-S imager (Bio-rad).
Western blots for cadherin-like proteins were per-
formed as described above, except that the anti-OnBt-R
1
serum previously described (1:3000 at room temperature
for 1 h) was used. The membrane was washed as
described above, and an anti-rabbit-AP serum
(1:10,000 at room temperature for 1 h) was applied to
the PVDF membrane. After repeated washings with
assay buffer (Tropix Inc.), the membrane was incubated
ARTICLE IN PRESS
R.D. Flannagan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 3340 35
with CDP-Star for 5 min and the image captured as
described above.
2.7. Expression of OnBt-R
1
in Spodoptera frugiperda Sf9
cells
Spodoptera frugiperda (Sf9) cells (ATCC 1711-CRL)
were grown at 27 1C in Sf-900 II serum-free medium
(Invitrogen). Full-length OnBt-R
1
cDNA (5494 bp) was
cut from a pBluescript/Bt-R
1
construct and cloned into
plasmid pFastBac (Invitrogen) at different cloning sites
downstream from the polyhedrin promotor. The plas-
mid was recombined into the bacmid backbone via
DH101Bac E. coli cells in a BacBac Expression system
(Invitrogen). Bacmids (2 mg/100 ml Sf-900 medium)
that harbored OnBt-R
1
were mixed with CellFectin
ARTICLE IN PRESS
Fig. 1. Deduced amino acid sequence of O. nubilalis BT-R
1
. The protein sequence analysis was done by using the ISREC-ProleScan server (http://
hits.isb-sib.ch/cgi-bin/PFSCAN) and PROSITE (http://us.expasy.org/prosite/). The putative signal peptide sequence and TM spanning region are
underlined and boxed, respectively. Full-black arrows denote predicted putative N-glycosylation sites. CR1CR11 are cadherin repeats and MPR,
the membrane-proximal region. The bolded sequence at the C-terminal sequence represents the intracellular domain.
R.D. Flannagan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 3340 36
(6ml/100 ml Sf900 medium) (Invitrogen), and incubated
at room temperature for 30 min. The mixture was
diluted with 0.8 ml Sf-900 medium and cells (10
6
/ml in
35 mm well in culture plate) washed with Sf-900 medium
once. The Bacmid/cell mixture was added to the well
and incubated at room temperature for 5 h after which
the medium was removed and 2 ml Sf-900 medium
containing penicillin and streptomycin were added to
the well. Ligand blot with anti-OnBt-R
1
was used to
examine the expression 35 days after infection.
To determine if baculovirus infected cells were
susceptible to the Cry1Ab toxin, cells infected with
OnBt-R
1
-containing baculovirus were compared with
non-infected cells and with cells infected with baculo-
virus containing an empty vector. The constructs were
driven by a very late promoter and analyzed 35 days
post-infection. Cells were washed once with PBS, and
different concentrations of Cry1Ab were applied to the
cells in PBS at different time intervals. The cells were
examined by light microscopy to determine cell viability
in the presence of toxin.
3. Results
3.1. Identication and analysis of the O. nubilalis
cadherin-like protein
Degenerate primers designed from the M. sexta
cadherin Bt-R
1
(Vadlamudi et al., 1995) were used in
to amplify a 280 base pair fragment from O. nubilalis
midgut cDNA. Sequence analysis revealed the fragment
was a Bt-R
1
homologue, and it was subsequently used to
screen an O. nubilalis midgut cDNA library generating
several positive clones. The largest inserts contained
5328 bp with an open reading frame of 5151 bp
(Genbank Accession No. AX147201), and a polyadeny-
lation signal in the 3
0
-UTR. Upon sequence analysis of
the largest positive clones, all matched the protein
molecule sequence in Fig. 1 consisting of 1717 amino
acids which contains a transmembrane region (TM) of
23 amino acids. The extracellular domain comprises a
signal sequence (SP) of 22 amino acid residues, 11
cadherin repeats (CR), and a membrane-proximal
region (MPR) (Fig. 1). Fourteen putative N-glycosyla-
tion sites were identied in the protein sequence. The
protein cytosolic domain is composed of 126 amino acid
residues.
Phylogenetic analysis of the predicted OnBt-R
1
amino
acid sequence, based on alignment with other lepidopteran
cadherins identied as Cry1 receptors, revealed that this
receptor shares 6270% similarity and 5663% identity
(Table 1) with these other cadherin molecules. This
analysis indicated that the OnBt-R
1
is most closely related
to a cadherin from Chilo suppressalis (Family Pyralidae)
and P. gossypiella (Family Gelechiidae) (Fig. 2) and most
distant from the cadherin genes identied from ve
different noctuid species.
3.2. Immunoblots
Immunoblots were performed to identify potential
Cry1Ab binding proteins using both anti-Cry1Ab and
anti-OnBt-R
1
serum. The Cry1Ab toxin bound to three
proteins of approximately 220, 170, and 160 kD in size
(Fig. 3, Lane 2). The anti-cadherin detected bands of
similar molecular weight (Fig. 3, Lane 1) indicating that
the proteins binding the Cry1Ab toxin are cadherin-like
proteins. No bands were visible in the controls
performed without Cry1Ab and anti-cadherin serum
(data not shown).
3.3. Expression of OnBt-R
1
Sf9 cells infected with the recombinant DNA carrying
OnBt-R
1
cDNA produced a moderate amount a protein
of approximately 190 kDa that reacted specically with
anti-OnBt-R
1
serum (Fig. 4). The slightly smaller size
relative the major 220 kDa band observed in blots of
native BBMV protein is likely explained by differences
in post-translational modications that could affect
molecular weight of the expressed protein. There were
several smaller molecular weight proteins in the super-
natant obtained from cell lysates that cross-reacted with
the antiserum indicating possible degradation of the full-
length receptor or expression of peptide fragments (Fig.
4). Only the 190 kDa band was visible in membrane
fractions of the cell lysate indicating that full-length
OnBt-R1 was expressed predominately within the cell
membrane fraction. In control samples, there was a very
faint 190 KDa band indicating that Sf9 cells might
express small amount of a Bt-R
1
homologue.
ARTICLE IN PRESS
Table 1
Pairwise comparison of O. nubilalis cadherin-like protein to cadherins
from 10 other species of Lepidoptera
Species Similarity
(%)
Identity
(%)
Accession no.
Ostrinia nubilalis CAC 41165
Lymantria dispar 70.4 63.1 AAL 26896
Bombyx mori 70.6 64.6 BAA 77212
Manduca sexta 69.1 61.8 AAG 37912
Heliothis virescens 68.6 61.5 AAK 85198
Helicoverpa zea 68.1 60.7 CAC41166
Helicoverpa armigera 67.8 60.5 AAM 69351
Chilo suppressalis 66.4 59.4 AAM 78590
Pectinophora
gossypiella
66.3 59.0 AAP 30715
Spodoptera frugiperda 63.0 53.3 CAC41167
Agrotis ipsilon 62.6 56.3 Under
submission
R.D. Flannagan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 3340 37
Cells infected with pFast/Bac alone were not affected
by Cry1Ab (Fig. 5c and e). However, 30 min after
incubation with Cry1Ab, pFast/Bt-R1 infected cells
started swelling and lysing even at concentrations as low
as 0.1 mg/ml Cry1Ab (Fig. 5d). At higher magnication,
it was observed that the pFast/Bt-R1 infected cells were
enlarged and more granulated as compared to the
controls (data not shown). At higher concentrations
(5 mg/ml), cell lysis was clearly evident and the cell
contents (DNA-like material) were released (Fig. 5f).
However, control cells did not swell or lyse in the
presence of Cry1Ab even at elevated (5 mg/ml) concen-
trations.
4. Discussion
The relatively recent identication of receptors for Bt
toxins has provided a greater understanding of their
mechanism of toxicity. In addition to the aminopepti-
dase-like proteins that have been identied as receptors
for Bt toxins (Sato, 2003), cadherin-like proteins have
also been identied as Bt Cry1A receptors in a number
of different lepidopterans (Zhuang and Gill, 2003). In
the present work, a cadherin-like protein present in the
BBMVs of O. nubilalis was identied as a receptor for
the Cry1Ab toxin. These results conrm the involvement
of a cadherin-like protein that binds to the Cry1Ab
ARTICLE IN PRESS
Amino acid Substitutions (x100)
0
78.6
10 20 30 40 50 60 70
Bombyx mori
Manduca sexta
Lymantria dispar
Chilo suppressalis
Ostrinia nubilalis
Pectinophora gossypiella
Helicoverpa armigera
Helicoverpa zea
Heliothis virescens
Agrotis ipsilon
Spodoptera frugiperda
Fig. 2. Phylogenetic tree of aligned Lepidoptera cadherin-like proteins identied as Cry1 receptors. The tree was performed with MegAlign
(DNAStar). Species correspondent GenBank accession numbers are in Table 1.
Fig. 3. Binding of anti-cadherin (anti-OnBt-R1) (Lane 1) from O.
nubilalis and Cry1Ab toxin (Lane 2) to BBMV proteins.
Fig. 4. Western blot assay or Sf9 transfected cells: Sf9 cell were
transfected with pFastBac or pFastBac/BtR1. Five days after
transfection. Cells were lysed and centrifuged. The supernatants and
pellets were subjected to Western blot (see Material and methods). (1)
pFastBac/supernatant, (2) pFasBac/pellet, (3) protein marker, (4)
pFastBac/BtR1 supernatant, (5) pFastBac/BtR1 pellet, (6) non-
transfected supernatant, and (7) non-transfected cell pellet.
R.D. Flannagan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 3340 38
toxin and confers susceptibility to Sf9 cells transfected
with the OnBt-R1 gene. Results from immunoblot
analysis suggest that Cry1Ab binds to three putative
proteins from susceptible insects, and that the same
three bands cross-react with an anti-OnBt-R1 serum
suggesting that all three proteins belong to the cadherin-
family of proteins. Denolf et al. (1993) suggested that
Cry1Ab binds to two different gut receptors from O.
nubilalis, while more recent efforts (Hua et al., 2001)
suggested that Cry1Ab binds up to four different
receptors. Although both aminopeptidases and cadher-
ins have been identied as binding proteins for Cry1 Bt
toxins in lepidopteran midgut apical membranes, it
appears more likely that the Cry1Ab binding protein for
O. nubilalis is a cadherin-like protein. The 220-kDa
receptor identied in the present study is consistent with
results from Hua et al. (2001) who reported a similar
molecular weight protein from O. nubilalis midguts as a
cadherin-like protein.
Cadherins constitute a large family of transmembrane
glycoproteins responsible for cell adhesion and main-
tenance of the integrity of selective cellcell interactions
(Nollet et al., 2000). Although their specic functions in
insects have not been fully resolved, they seem to play an
important role in the binding of Cry1 toxins in a number
of different species (Vadlamudi et al., 1993; Nagamatsu
et al., 1999; Gahan et al., 2001; Hua et al., 2001). The O.
nubilalis cadherin-like protein shows a relative high
similarity and identity to other members of the cadherin
superfamily in insects. Eleven CRs are present in the O.
nubilalis cadherin and 14 N-glycosylation sites are
distributed along the extracellular domains. The rela-
tively high similarity to other lepidopteran cadherin-like
proteins indicates that the O. nubilalis cadherin-like
protein shares related structures, functions, and conse-
quently specicity for Cry1 Bt toxins.
Expression of the OnBt-R1 in Sf9 cells provided
strong evidence that this molecule not only binds
Cry1Ab but is responsible for its toxicity since
transfected cells were susceptible to toxin concentrations
as low as 0.1 mg/ml while control (untransfected) cells
exhibited no response. It has been suggested that Cry
toxins bind to a specic receptor and are then inserted
into the membrane to form a pore that alters membrane
permeability. The consequence is lysis of the epithelial
cells and death of the insect (Knowles, 1994). The
present results suggest that a cadherinlike protein from
O. nubilalis midgut tissue mediates both binding and
insertion into the cell membrane.
Identication of the Cry1Ab binding protein in O.
nubilalis is an important step in our understanding of
potential resistance mechanisms that might evolve for
transgenic corn which is currently comprised almost
exclusively by Cry1Ab expressing hybrids. Further
characterization of this receptor will facilitate our
understanding of possible mutations in the receptor
that could alter binding characteristics and confer
resistance. A number of recent studies have been
conducted to determine and characterize the binding
region for lepidopteran cadherins (Nagamatsu et al.,
ARTICLE IN PRESS
Fig. 5. Effect of Cry1Ab on pFastBac and OnBt-R
1
transformed-Sf9 cells. (a,c,e) Mock transformed Sf9 cells incubated with 0, 0.1 and 5 mg/ml of
Cry1Ab, respectively. (b,d,f) BtR1-transformed Sf9 cells incubated with 0, 0.1 and 5 mg/ml of Cry1Ab, respectively.
R.D. Flannagan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 3340 39
1999; Gomez et al., 2001; Dorsch et al., 2002). Hua et al.
(2004) determined that the cadherin domain 12 is critical
for Cry1Ab binding in M. sexta which was the minimum
region necessary to confer Drosophila S2 cell suscept-
ibility to Cry1Ab. Identication of the O. nubilalis
cadherin-like protein and epitope mapping studies will
eventually help to design molecular tools to detect
resistance associated with altered binding to cadherin
receptors. Additionally, future development of trans-
genic hybrids will be facilitated by the ability to predict
cross-resistance among different Bt toxins based on a
thorough characterization of receptor molecules. In
general, such information should facilitate the rational
implementation of management practices that employ
transgenic corn hybrids for controlling European corn
borer populations.
Acknowledgments
Support for this project was provided by Pioneer Hi-
Bred International, CAPES Coordenac-a o de Aperfeic-
oamento de Pessoal de N vel Superior (Brazil) and the
Consortium for Plant Biotechnology Research. Both
Cry1Ab toxin and a Cry1Ab antiserum were provided
by the Monsanto Company. This Paper is contribution
no. 14564 of the Journal Series of the Nebraska
Agricultural Experiment Station and Contribution no.
1184 of the Department of Entomology.
References
Bravo, A., Hendrickx, K., Jansens, S., Peferoen, M., 1992. Immuno-
cytochemical analysis of specic binding of Bacillus thuringiensis
insecticidal crystal proteins to lepidopteran and coleopteran midgut
membranes. J. Invertebr. Pathol. 60, 247253.
Denolf, P., Jansens, S., Peferoen, M., Degheele, D., Vanrie, J., 1993.
Two different Bacillus thuringiensis delta-endotoxin receptors in the
midgut brush-border membrane of the European corn corer,
Ostrinia nubilalis (Hubner) (Lepidoptera, Pyralidae). Appl. Envir-
on. Microbiol. 59, 18281837.
Dorsch, J.A., Candas, M., Griko, N.B., Maaty, W.S.A., Midboe,
E.G., Vadlamudi, R.K., Bulla Jr., L.A., 2002. Cry1A toxins of
Bacillus thuringiensis bind specically to a region adjacent to the
membrane-proximal extracellular domain of BT-R1 in Manduca
sexta: involvement of a cadherin in the entomopathogenicity of
Bacillus thuringiensis. Insect Biochem. Mol. Biol. 32, 10251036.
Gahan, L.J., Gould, F., Heckel, D.G., 2001. Identication of a gene
associated with Bt resistance in Heliothis virescens. Science 293,
857860.
Gill, S.S., Cowles, E.A., Pietrantonio, P.V., 1992. The mode of action
of Bacillus thuringiensis endotoxins. Annu. Rev. Entomol. 37,
615636.
Gomez, I., Oltean, D.I., Gill, S.S., Bravo, A., Soberon, M., 2001.
Mapping the epitope in cadherin-like receptors involved in Bacillus
thuringiensis Cry1A toxin interaction using phage display. J. Biol.
Chem. 276, 2890628912.
Hua, G., Masson, L., Jurat-Fuentes, J.L., Schwab, G., Adang, M.J.,
2001. Binding analyses of Bacillus thuringiensis Cry delta-endotox-
ins using brush border membrane vesicles of Ostrinia nubilalis.
Appl. Environ. Microbiol. 67, 872879.
Hua, G., Jurat-Fuentes, J.L., Adang, M.J., 2004. Bt-R1a extracellular
cadherin repeat 12 mediates Bacillus thuringiensis Cry1Ab binding
and cytotoxicity. J. Biol. Chem. 279, 2805128056.
Knowles, B.H., 1994. Mechanisms of action of Bacillus thuringiensis
insecticidal d-endotoxins. Adv. Insect Physiol. 24, 275308.
Laemmli, U.K., 1970. Cleavage of structural proteins during
the assembly of the head of bacteriophage T4. Nature 227,
680685.
Morin, S., Biggs, R.W., Sisterson, M.S., Shriver, L., Ellers-Kirk, C.,
Higginson, D., Holley, D., Gahan, L.J., Heckel, D.G., Carriere, Y.,
Dennehy, T.J., Brown, J.K., Tabashnik, B.E., 2003. Three cadherin
alleles associated with resistance to Bacillus thuringiensis in pink
bollworm. Proc. Natl. Acad. Sci. USA 100, 50045009.
Nagamatsu, Y., Toda, S., Koike, T., Miyoshi, Y., Shigematsu, S.,
Kogure, M., 1998a. Cloning, sequencing, and expression of the
Bombyx mori receptor for Bacillus thuringiensis insecticidal
CryIA(a) toxin. Biosci. Biotechnol. Biochem. 62, 727734.
Nagamatsu, Y., Toda, S., Yamaguchi, F., Ogo, M., Kogure, M.,
Nakamura, M., Shibata, Y., Katsumoto, T., 1998b. Identication
of Bombyx mori midgut receptor for Bacillus thuringiensis
insecticidal CryIA(a) toxin. Biosci. Biotechnol. Biochem. 62,
718726.
Nagamatsu, Y., Koike, T., Sasaki, K., Yoshimoto, A., Furukawa, Y.,
1999. The cadherin-like protein is essential to specicity determina-
tion and cytotoxic action of the Bacillus thuringiensis insecticidal
CryIAa toxin. FEBS Lett. 460, 385390.
Nollet, F., Kools, P., van Roy, F., 2000. Phylogenetic analysis
of the cadherin superfamily allows identication of six major
subfamilies besides several solitary members. J. Mol. Biol. 299,
551572.
Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning: A
Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold
Spring Harbor, NY.
Sato, R., 2003. Aminopeptidase N as a receptor for Bacillus
thuringiensis Cry toxins. In: Upadhyay, R.K. (Ed.), Advances in
Microbial Control of Insect Pests. Kluwer Academic/Plenum
Publishers, New York, NY, pp. 113.
Schnepf, E., Crickmore, N., Van Rie, J., Lereclus, D., Baum, J.,
Feitelson, J., Zeigler, D.R., Dean, D.H., 1998. Bacillus thuringiensis
and its pesticidal crystal proteins. Microbiol. Mol. Biol. Rev. 62,
775806.
Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.K., Gartner,
F.H., Provenzano, M.D., Fujimoto, E.K., Goeke, N.M., Olson,
B.J., Klenk, D.C., 1985. Measurement of protein using bicincho-
ninic acid. Anal. Biochem. 150, 7685.
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins,
D.G., 1997. The ClustalX windows interface: exible strategies for
multiple sequence alignment aided by quality analysis tools.
Nucleic Acids Res. 25, 48764882.
Vadlamudi, R.K., Ji, T.H., Bulla, L.A., 1993. A specic binding
protein from Manduca sexta for the insecticidal toxin of Bacillus
thuringiensis subsp. Berl. J. Biol. Chem. 268, 1233412340.
Vadlamudi, R.K., Weber, E., Ji, I., Ji, T.H., Bulla, L.A., 1995. Cloning
and expression of a receptor for an insecticidal toxin of Bacillus
thuringiensis. J. Biol. Chem. 270, 54905494.
Wolfersberger, M., Luthy, P., Maurer, A., Parenti, P., Sacchi, F.V.,
Giordana, B., Hanozet, G.M., 1987. Preparation and partial
characterization of amino acid transporting brush border mem-
brane vesicles from the larval midgut of the cabbage buttery Pieris
brassicae. Comp. Biochem. Physiol. A. 86, 301308.
Zhuang, M., Gill, S.S., 2003. Mode of action of Bacillus thuringiensis
toxins. In: Voss, G., Ramos, G. (Eds.), Chemistry of Crop
Protection: Progress and Prospects in Science and Regulation.
Wiley-VHC 410pp.
ARTICLE IN PRESS
R.D. Flannagan et al. / Insect Biochemistry and Molecular Biology 35 (2005) 3340 40
Insect
Biochemistry
and
Molecular
Biology
Insect Biochemistry and Molecular Biology 35 (2005) 4150
Cell size control by ovarian factors regulates juvenile hormone
synthesis in corpora allata of the cockroach, Diploptera punctata
Ling-Wen Chang, Chuan-Mei Tsai, De-Ming Yang, Ann-Shyn Chiang

Department of Life Science, Institute of Biotechnology, National Tsing Hua University, 101, Section 2 Kuang Fu Road, Hsinchu 30043, Taiwan
Received 19 July 2004; received in revised form 9 October 2004; accepted 11 October 2004
Abstract
The corpora allata synthesize and release juvenile hormone (JH) that in turn regulates insect growth, metamorphosis and
reproduction. In the corpus allatum (CA) of the female adult cockroach Diploptera punctata, cyclic rise and decline in JH synthesis
rates occur concurrently with cyclic growth and atrophy during an ovarian cycle. Here, we report that protein content decreases,
whereas Golgi population, lysosomal content and autophagic activities increase with decrease in CA cell size. Also, the
concentration of cyclic GMP (cGMP) is low in large cells and high in small cells. Results of treating CA with ovarian tissue suggest
that a putative peptidergic growth regulator released from mature ovaries acts directly on active CA cells and induces the elevation
of intracellular cGMP content. Consequently, elevated cGMP may inhibit protein synthesis or trigger massive and synchronous
autophagic activities, resulting in cell atrophy and reduction of protein content. As a result of the depletion of cellular machinery,
CA glands exhibit long-term depression in JH synthesis.
r 2004 Elsevier Ltd. All rights reserved.
Keywords: Cell size; Corpora allata; JH; Organ culture; Ovary; Cockroach
1. Introduction
The corpora allata are a pair of endocrine glands that
synthesize juvenile hormone (JH), which in turn reg-
ulates insect development, metamorphosis and reproduc-
tion. Corpus allatum (CA) activity rises and declines
several times during the life of an insect (Cassier, 1990).
For example, in adult females of both the viviparous
cockroach Diploptera punctata and the oviparous cock-
roach Blattella germanica, rates of JH synthesis increase
during oocyte growth, decrease before ovulation, remain
low through gestation and rise again after parturition
(Tobe and Stay, 1977; Szibbo and Tobe, 1981; Feyer-
eisen et al., 1981). CA activity may be regulated by a
rapid biochemical modulation of rate-limiting steps in
the process of JH synthesis and by a slow developmental
control of cellular machinery (Tobe and Pratt, 1976;
Feyereisen, 1985). The isolation of allatostatin and
allatotropin from insect brains and the identication of
glutamatergic nerve bers from the brain to the CA
provide evidence for the rst of these mechanisms (Stay
et al., 1994; Taylor et al., 1996; Chiang et al., 2002a). In
vitro JH production by CA is rapidly and reversibly
inhibited by allatostatin and stimulated by allatotropin.
Acute stimulation by glutamate-eliciting Ca
2+
inux via
NMDA and kainate subtype glutamate receptors also
results in the immediate increase in JH production
(Pszczolkowski et al., 1999; Chiang et al., 2002b).
Morphometric and ultrastructural studies demon-
strate that uctuations in JH production always
accompany changes in CA gland volume, cell size and
quantity of cellular components (Chiang et al., 1998;
Johnson et al., 1985, 1993). While many reports have
addressed short-term regulation, mechanisms control-
ling developmental plasticity of CA cells responsible for
long-term cycles of JH production are much less
understood. Here, we show that changes in Golgi
ARTICLE IN PRESS
www.elsevier.com/locate/ibmb
0965-1748/$ - see front matter r 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ibmb.2004.10.006

Corresponding author. Tel.: +886 3 5742760; fax: +886 3 5717237.


E-mail address: aschiang@life.nthu.edu.tw (A.-S. Chiang).
population, lysosomal content and autophagic activities
occur sequentially in CA cells and correlate precisely
with the variation in protein content, CA cell size and
amount of JH production during the ovarian cycle of
mated D. punctata adult females. Most importantly, we
also demonstrate that long-term repression of JH
synthesis is mainly due to atrophy of CA cells
modulated by ovarian regulation on changes in the
intracellular cyclic GMP (cGMP) content.
2. Materials and methods
2.1. Insects
The D. punctata colony was maintained as previously
described (Cheng and Chiang, 1995). Animal age in
the text refers to the age in days from imaginal ecdysis
or denervation. In all experiments, the basal oocyte
length was used as a measure of the physiological
condition of the mated female. The ranges used
were: day 2 (0.80.91 mm), day 4 (1.161.40 mm),
day 5 (1.411.56 mm), day 6 (1.571.66 mm), day 7
(1.67 mm-before ovulation), day 8 (ovulation). In
denerved cockroaches, unilateral transection of nervous
corporis allati-I was performed as described by Chiang
et al. (1998), except that day-1 adult virgin females were
operated on and the age was calculated from the
completion of surgery. The ranges of oocyte length used
were: day 2 (o0.85 mm), day 4 (0.851.23 mm), day 6
(1.231.54 mm), day 8 (1.54 mm-before ovulation).
2.2. Organ culture and cell dissociation
CA dissection and dissociation into single cells were
accomplished as previously described (Chiang et al.,
1989). Dissociated allatal cells were allowed to settle and
attach onto a poly-L-lysine coated slide for 30 min. For
organ culture, the modied L-15B culture medium
containing 5% fetal bovine serum was used (Holbrook
et al., 1997), with the exclusion of additional CaCl
2
. In all
experiments, each CA pair was split and cultured
separately in 300 ml L-15B for 3 days. One CA member
was incubated with ovaries, extract, 8-Br-cGMP (Sigma-
Aldrich Co., St. Louis, MO; B1381), IBMX (RBI,
Natick, MA; A-007) or Rp-8-pCPT-cGMPS (Biomol,
Plymouth Meeting, PA; CN-206), and the other member
derived from the same individual incubated without
adding the test factors. Two pairs of glands were used for
each test. All incubations were carried out in a humidied
chamber at 2770.5 1C with a constant gentle shaking.
2.3. Fluorescent labeling of Golgi complexes
The uorescent ceramide analogue, BODIPY-cera-
mide (Molecular Probes, Eugene, OR), accumulates
specically in the Golgi complexes of various cell types
(Pagano et al., 1991). To label Golgi complexes,
dissociated allatal cells were incubated in the modied
L-15B containing 10 mM BODIPY-ceramide for 30 min.
Subsequently, these cells were washed three times,
incubated in fresh modied L-15B medium for 30 min,
washed, and further incubated in the medium for 3 h. In
some experiments, the culture medium contained
brefeldin A (20 mg/ml) (Sigma; B7651). The resulting
disassembly of Golgi complexes (Donaldson et al., 1990)
validated the specicity of BODIPY-ceramide for
labeling of Golgi complexes in allatal cells. Live cells
mounted in the medium were immediately observed and
photographed under a Zeiss Axiophot microscope
equipped with epiuorescence optics. For each experi-
ment, at least six animals in the same physiological
condition were examined.
2.4. Neutral red labeling of lysosomes
Neutral red is a common lysosomal probe that stains
lysosomes by accumulation in the acidic compartment
of a living cell (DellAntone, 1979). Dissociated allatal
cells attaching to the slide were incubated in the
modied L-15B with 0.02% neutral red (Sigma) for
10 min, washed three times, further incubated in the
fresh medium without the labeling dye for 15 min,
washed once and sealed under a coverslip. Cells were
observed and photographed under a Zeiss Axiophot
microscope using differential interference-contrast
optics.
2.5. Enzyme assays
Thiamine pyrophosphatase (TPPase) activity in CA
cells was measured with thiamine pyrophosphate (TPP)
as substrate. We used phosphate-free cockroach saline
in the following assays in order to prevent phosphate
from interfering with the measurements of inorganic
phosphate. Four pairs of CA collected in phosphate-free
cockroach saline solution (158.6 mM NaCl, 17.6 mM
KCl, 3.86 mM NaHCO
3
, 15 mM glucose, pH 7.2,
osmolarity 360 mOsm) were used for each assay. Glands
were desheathed with 0.1% collagenase in the phos-
phate-free saline for 10 min, transferred into 16 ml of
deionized H
2
O in an Eppendorf tube, and dispersed into
tissue fragments with pipetting. Tissue fragments were
further pipetted after adding 16 ml of 0.5% Triton X-100
in the same saline. Immediately before measurement,
each sample solution was split in half. One half was
reacted with TPP substrate in 64 ml assay mixture
(50 mM Tris-HCl, 5 mM MgCl
2
, and 1 mM TPP),
whereas the other half was given the same amount of
assay mixture without TPP. A third sample of 16 ml of
saline solution without CA fragments was reacted with
TPP substrate in 64 ml assay mixture as another blank.
ARTICLE IN PRESS
L.-W. Chang et al. / Insect Biochemistry and Molecular Biology 35 (2005) 4150 42
After incubation for 60 min at 27 1C, the reaction was
stopped by adding 16 ml of 1 mM citric acid. Inorganic
phosphate (Pi) released was measured by using a rapid
colorimetric method (Heinonen and Lahti, 1981). Test
tubes containing 96 ml of sample solutions each received
704 ml of acetone-acid-molybdate solution (1:2:1 of 5 N
H
2
SO
4
:acetone:ammonium molybdate). The contents
were vortexed for 30 s to form bright yellow phospho-
molybdate prior to the addition of 70.4 ml of 1 mM citric
acid. After mixing again, the yellow color intensity was
measured with a Hitach U-2000 spectrophotometer at
355 nm in a 1 ml cuvette. A standard curve by using
different known amounts of Pi as samples was included
in the routine TPPase assay. All data were corrected by
the two blank incubations, one with no CA and the
other with no TPP in the enzyme assay.
Acid phosphatase activity was measured with p-
nitrophenol phosphate as substrate. Two pairs of
desheathed CA were transferred to 32 ml distilled water
containing 0.25% Triton X-100, sonicated in a water
bath for 5 min, and then stored in a 20 1C freezer until
used. Phosphatase activity was determined spectro-
photometrically (Xia et al., 2000).
2.6. Electron microscopy
Histochemical detection of TPPase and sample
preparation for electron microscopic observation was
as described previously (Cheng and Chiang, 1995; Yang
and Chiang, 1997). For each experiment, at least ve
animals in the same physiological condition were
examined.
2.7. Determination of cell sizes, rates of JH synthesis and
protein content
Enzymatically dissociated cells from freshly dissected
or cultured CA were randomly sampled in a hemocyto-
metric grid and the diameter of a cell served as the
measure of its size (Chiang et al., 1989). The mean cell
size of CA was determined from 60 cells in each
experiment. JH production by CA in vitro was
determined with a radiochemical assay as previously
described (Pszczolkowski et al., 1999). For determina-
tion of protein content, 810 CA glands were homo-
genized with a sonicator for 15 s. Protein content was
determined using the Bradford method (Biorad Labora-
tories, Hercules, CA).
2.8. Determination of cGMP content
Eight CA or four pairs of glands were lysed in 25 ml of
0.1% NaOH, and neutralized by the addition of 25 ml
0.025 N HCl. Fifty ml cold 12% trichloroacetic acid was
added to denature the sample protein. Samples were
centrifuged at 2000g for 15 min at 4 1C. The supernatant
was recovered and washed three times with 1 ml
water saturated diethyl ether. The aqueous extract
remaining was lyophilized. The enzyme immunoassays
for cGMP (Amersham Biosciences, Buckinghamshire,
UK; RPN226) were performed after the samples were
acetylated with acetic anhydride-triethylamine to en-
hance sensitivity.
2.9. Treatment of ovary extract
To examine whether ovary extract induced atrophy of
CA cells from day-4 mated females, 20 pairs of ovaries
from day-7 mated females were homogenized with a
dounce tissue grinder in 1 ml L15B medium and
centrifuged at 3000g for 20 min. The supernatant was
ltered through a 0.2 mm lter and stored in a 80 1C
freezer until used. The extract was treated with 0.5 mg/
ml trypsin (Sigma; T1426) for 2.5 h, and then trypsin
inhibitor (Sigma; T9003) was added (nal concentration
0.75 mg/ml) to determine whether the ovarian factor was
trypsin-sensitive. To examine whether the ovarian factor
was heat-sensitive, the extract was boiled for 10 min, and
then centrifuged at 3000g for 20 min.
To determine the effects of ovaries on changes in
cGMP content in the CA, 30 pairs of ovaries were
extracted in 1 ml cockroach saline (150 mM NaCl,
12 mM KCl, 10 mM CaCl
2
, 3 mM MgCl
2
, 4 mM Hepes
and 23 mM trehalose, pH7.2). Two pairs of CA were
used and pairs of glands were split in each cGMP
determination. Two glands were incubated in 240 ml
diluted ovarian extract (15 pairs/ml cockroach saline) or
10 mM Dippu allatostatin-7 (Dippu AST-7, allatostatin
I; Sigma; A9929) and 500 mM IBMX (RBI; A007) as
experiments; the others were incubated in cockroach
saline with only 500 mM IBMX as controls. cGMP
content in the CA glands was determined after 20 min
incubation.
3. Results
3.1. CA cell size and protein content uctuate together
during the rst ovarian cycle
Corpora allata underwent a cycle of cell size change
in mated adult females during the rst ovarian cycle
(Figs. 1A and 2A). In contrast, CA cells remained small
in virgin adult females unless nerve connections between
the brain and the CA were transected. By mimicking the
effects of mating, nerve severance allowed rapid CA cell
growth followed by synchronous cell atrophy. Uni-
lateral CA denervation did not affect development of
the other intact CA member remaining connected to the
brain. These results suggest that CA cell growth is under
direct brain neural inhibition that can be removed by
mating stimuli. Intriguingly, cell atrophy appeared to be
ARTICLE IN PRESS
L.-W. Chang et al. / Insect Biochemistry and Molecular Biology 35 (2005) 4150 43
regulated by humoral factors since it also occurred in
denervated CA.
Ultrastructural studies demonstrated that uctuation
in CA cell size was due to changes in the quantity of
cellular components (Chiang et al., 1998; Cheng and
Chiang, 1995; Yang and Chiang, 1997). To further
investigate mechanisms controlling CA cell plasticity,
changes in protein content were determined. Protein
content of CA glands uctuated concurrently with
changes in CA cell size (Fig. 1B). To calculate the
protein content per cell, the values of total cell number
were abstracted from Chiang et al. (1996). Gland
protein content was divided by the total cell number
of CA glands at different ages. The protein content per
cell doubled from days 0 to 6 and returned to the
original quantity by day 12 (Fig. 1C). Thus, cell
diameter appears to be a reliable index of CA cell size
and protein content.
3.2. Autophagic activities and Golgi population uctuate
together with changes in CA cell size
Cell size is related to the rates of protein synthesis and
degradation. In order to detect early timing of the switch
from cell growth to cell atrophy, we monitored
autophagic activities in CA cells during the rst ovarian
cycle. As expected, the occurrence of acidic organelles
and acid phosphatase activity increased with a decrease
in CA cell size (Fig. 2). Small neutral red bodies,
presumably primary lysosomes not yet involved in
cellular digestion, appeared in CA cells of both virgin
and mated females (Fig. 2A). Giant neutral red bodies,
presumably secondary lysosomes or autophagolyso-
somes that are involved in cellular digestion, appeared
in CA cells of mated females by day 6 (Fig. 2A, M6). In
the next few days, these giant autophagolysosomes
enlarged in size but were reduced in number with cell
size reduction (Fig. 2A, M6M12). Morphological
changes in neutral red bodies were consistent with
increased activity of acid phosphatase, a marker enzyme
of lysosomes (Fig. 2D). Acid phosphatase activity cycled
in CA glands of mated females peaked on day 6 and
remained at relatively high levels during the next
few days when giant autophagolysosomes appeared
(Figs. 2D and E). During this period of time, more than
two-thirds of CA cells contained giant autophagolyso-
somes suggesting massive and synchronous occurrence
of autophagic activities in CA glands (Fig. 2E). In CA
glands of virgin females, acid phosphatase activity
occurred at relatively low levels (Fig. 2D) and giant
autophagolysosomes were never observed (Fig. 2E).
Surprisingly, Golgi complexes also exhibited dramatic
changes in number and shape during the rst ovarian
cycle. In virgin females, CA cells contained only few
Golgi complexes (Fig. 2B, V0V12). When CA cells
grew in size after mating, the number of Golgi
complexes multiplied several fold (Fig. 2B, M2M6)
and then decreased abruptly with cell atrophy (Fig. 2B,
M8M12). To account for such a drastic change in
number, we speculated that Golgi complexes might be
involved in functions other than modication and
sorting of lysosomal enzymes required for cellular
digestion during cell atrophy. Histochemical tracing of
trans-Golgi cisternae with TPPase activities showed that
Golgi complexes were involved in a sequential process
of enclosing unwanted organelles during cell atrophy
(Fig. 2C). First, a Golgi complex made contact with
some membranous autophagic vacuoles via trans-
cisternae (Fig. 2C, a and b), wrapped these vacuoles
forming a whorl-like autophagosome (Fig. 2C, c and d),
and then fused with lysosomes forming a giant
autophagolysosome (Cheng and Chiang, 1995; Yang
and Chiang, 1997). While the later digestion process
might take several days, the wrapping phenomena of
Golgi complexes occurred only within a narrow time
ARTICLE IN PRESS
Fig. 1. Fluctuation of cell sizes, protein content and cGMP levels in
the CA glands during the rst ovarian cycle. (A) Changes in mean cell
diameter in relation to age in virgin, mated and denerved females. Each
value of mean cell diameter was determined from 60 cells of two CA
pairs (mated and virgin females) or two individual CA (one
denervated: one not in virgin females). (B) Changes in protein content
per CA gland. Each value was determined from four CA pairs or eight
individual CA. (C) Changes in protein content per CA cell in mated
females. Each datum was calculated from the value of protein content
per CA divided by total cell number per CA. (D) cGMP content per
milligram of CA protein. Each datum of cGMP content was
determined from four CA pairs or eight individual CA. Protein
content of CA was acquired from (B). Each point represents the
mean7SEM from 4 to 8 measurements.
L.-W. Chang et al. / Insect Biochemistry and Molecular Biology 35 (2005) 4150 44
window on day 5, one day before the rst appearance of
giant autophagolysosomes (Figs. 2A and B). These
morphological changes in Golgi complexes were con-
sistent with activity of TPPase, a marker enzyme of Golgi
complexes, in both mated and virgin females (Fig. 2F).
3.3. cGMP content increases when CA cell size decreases
The amount of cGMP in CA glands uctuates during
ovarian cycles (Tobe, 1990). By converting cGMP
content per CA to cGMP content per mg protein, we
found that cGMP content was inversely correlated to
changes in CA cell size (Fig. 1D). In virgin females,
cGMP content remained at relatively high levels when
CA cells were inactive and small. After females mated,
cGMP content decreased gradually and was lowest by
day 6 when CA cells grew to their maximum size.
Subsequently, cGMP content increased gradually when
CA cells underwent atrophy. Intriguingly, unilateral CA
denervation resulted in the decline of cGMP content in
both denerved and intact CA glands (Fig. 1D), although
the cell size of intact glands was unchanged. Thus, we
surmise that the elevated cGMP level was not a result of
cell atrophy. Based on these results, we hypothesize that
elevated cGMP content may induce CA cell atrophy.
Next, we examine factors inducing elevation of cGMP
content in CA cells.
3.4. Inuence of post-vitellogenic ovaries on atrophy of
CA cells
Ovarian factors synchronize changes in sizes of CA
cells in B. germanica and Supella longipalpa (Chiang et
al., 1991a; Chiang and Schal, 1994). In vitro CA culture
together with ovaries derived from mated females of
different ages for 3 days supported the direct action of
ovarian factors on CA cells (Table 1). Active CA glands
from day-4 mated females consisted of large CA cells
(average 17.3870.29 mm; Fig. 1A). When glands were
cultured in vitro, these cells continued to grow to an
even larger size in the absence of ovaries or in the
presence of previtellogenic ovaries from day-0 and day-2
mated females (Table 1). In contrast, CA cells ceased to
grow or decreased in size when cultured in the presence
of vitellogenic or post-vitellogenic ovaries from days 4
to 7 mated females. Mature ovaries from day-7 mated
females exerted strongest inhibition on CA cell growth
and were used for the following experiments. Unlike the
in vivo situation, the cells of day-4 CA cultured in the
absence of ovaries increased in diameter continuously
for at least six days (Fig. 3A). Therefore, in vivo atrophy
of CA cells in mated females (Fig. 1) was not a CA
autonomous program and required exogenous signals.
After incubation of CA together with a pair of mature
ovaries, the sizes of CA cells progressively decreased for
ARTICLE IN PRESS
Fig. 2. Lysosomal and Golgi activities during the rst ovarian cycle. (A) Occurrence of lysosomes. Living CA cells were labeled with neutral red for
acidic compartments. Each red dot represents a lysosome. (B) Occurrence of Golgi complexes. Living CA cells were labeled with the uorescent lipid
analogue BODIPY-ceramide for Golgi complexes. Each red dot represents a Golgi complex. Each picture in (A) and (B) shows a representative CA
cell derived from day-0 (V0), day-6 (V6) and day-12 (V12) virgin females and from day-2 to day-12 (M2M12) mated females. Bar, 10 mm. (C)
Ultrastructural tracing of Golgi apparatus with histochemical detection of trans-Golgi cisternae specic enzyme, TPPase, in the CA cells of day-5
mated females. Trans-Golgi cisternae (arrowhead) labeled with TPPase histochemical staining enclosed membranous autophagic vacuoles (arrow)
step by step (ad). ga: Golgi apparatus. Bar, 0.5 mm. (D) Acid phosphatase (AcP) activity indicated by nitrophenol (pNP) production using
p-nitrophenol phosphate as a substrate. (E) Occurrence of giant autophagolysosomes in CA cells. Red dots stained by neutral red with diameter
greater than 2 mm vacuoles were classied as giant autophagolysosomes. (F) TPPase activities indicated by the release of phosphate ion (Pi) using
TPP as substrate. Each point represents the mean7SEM from 4 to 8 measurements.
L.-W. Chang et al. / Insect Biochemistry and Molecular Biology 35 (2005) 4150 45
six days mimicking the atrophy phenomena in vivo
(Figs. 1 and 3).
Small cells of inactive CA glands derived from day-0
and day-12 mated females also increased size in vitro in
the absence of ovaries (Table 1 and Fig. 1A). Growth of
inactive cells ceased in the presence of mature ovaries.
Interestingly, culture medium preconditioned by mature
ovaries for two days prior to CA culture also inhibited
growth of CA cells (Table 1). These results suggest that
inhibitory growth regulators are released from mature
ovaries and act directly upon CA cells.
3.5. Characterization of the ovarian growth regulator
Ovary extracts exerted a dose-dependent inhibition on
CA cell growth (Fig. 3B). The effective component
appeared to be a peptide since the inhibitory effect was
abolished after trypsin digestion (Table 2). It might be a
small peptide because boiling for 10 min did not
completely abolish its inhibitory effect on CA cell
growth. Alternatively, the extracts might contain more
than one effective component. Although we have not
identied the factor, the ovary extract induced elevation
of cGMP content 2.8-fold compared with untreated
controls (Fig. 3C).
3.6. Cell size control by changing cGMP content
To understand the relationship between cGMP
content and cell atrophy, we directly manipulated
cGMP levels with drugs and examined growth of CA
cells and JH production. Our results showed a dose-
dependent response of size changes in CA cells to 8-Br-
cGMP, a cGMP analogue (Fig. 4A). CA glands from
day-4 mated females cultured for three days showed that
CA cell growth was most effectively inhibited by the
presence of 100 mM 8-Br-cGMP. Higher concentrations
of 8-Br-cGMP could not further inhibit cell growth.
Like ovary extracts, 8-Br-cGMP also effectively pre-
vented small and inactive CA cells from growth in vitro
(Fig. 4C). Treatment with IBMX, an inhibitor of
phosphodiesterases, also inhibited growth of CA cells
ARTICLE IN PRESS
Table 1
Effects of ovary on changes in CA cell sizes
Ovary CA Ovary +Ovary n Paired t-test
md0 md4 20.9670.48 19.9170.50 5 P 0.12
md2 md4 20.2870.27 19.8970.58 5 P 0.45
md4 md4 20.4370.19 17.1670.67 6 Po0.01
md6 md4 20.0270.38 15.3070.63 5 Po0.01
md7 md4 19.7570.54 14.4270.23 6 Po0.001
md7 md0 16.7870.52 13.3070.19 4 Po0.05
md7 md12 15.2570.41 12.6270.50 4 Po0.05
md7
a
md4 19.2670.50 15.6070.14 4 Po0.01
CA glands were derived from day-0 (md0), -4 (md4) and -12 (md12) mated females. In each experiment, a split pair test using two pairs of CA glands
was adopted. CA glands were cultured with or without a pair of ovaries for three days. Then the mean cell diameter (mm) of dissociated CA cells was
determined. Each datum represents the mean 7SEM with the number of measurements indicated.
a
CA glands were cultured in normal medium or preconditioned medium. Preconditioned medium was obtained by incubating eight pairs of ovaries
in 0.7 ml L-15B medium containing 5% serum for 48 h. The media were diluted to 2.1 ml for use.
Fig. 3. Ovarian effects on changes in cell size and cGMP content. (A)
Changes in cell diameter of CA glands cultured in vitro with or without
a pair of ovaries from day-7 mated females. (B) Dose response of cell
size changes to ovary extract. CA glands were derived from day-4
mated females. Cell diameters were measured after CA glands were
cultured for three days. Each point represents the mean7SEM from
ve experiments. (C) Effects of ovary extract (OE) and 10 mM Dippu
AST-7 (AST I) on cGMP content in the CA of day-5 mated females.
The level of cGMP was measured after CA glands were stimulated for
20 min (a). The level of cGMP was compared to that of control (b).
Each bar represents the mean7SEM from 9 to 11 measurements;
***, Po0.001.
L.-W. Chang et al. / Insect Biochemistry and Molecular Biology 35 (2005) 4150 46
(Fig. 4Ba). CA cells were signicantly smaller when
cultured in medium containing 200 mM IBMX or
100 mM 8-Br-cGMP for one day compared with the
controls. Since IBMX promotes both cGMP and
cAMP, we do not exclude the possibility that cAMP
may also participate in the process of CA cell atrophy.
To clarify whether ovarian extract caused cell atrophy
by elevating the cGMP level, we incubated CA glands in
medium containing ovarian extract with or without Rp-
8-pCPT-cGMPS, a cGMP antagonist. Fig. 4Bb shows
that Rp-8-pCPT-cGMPS inhibited the effects of 8-Br-
cGMP and ovarian extract on inducing cell atrophy (OE
vs. OE+Rp-8-pCPT-cGMPS and 8-Br-cGMP vs. 8-Br-
cGMP+Rp-8-pCPT-cGMPS, po0.001). Rp-8-pCPT-
cGMPS completely prevented 8-Br-cGMP from indu-
cing cell atrophy (8-Br-cGMP+Rp-8-pCPT-cGMPS vs.
none, p 0.74); however, the antagonist could only
partly rescue the effect of ovarian extract (OE+Rp-8-
pCPT-cGMPS vs. none, po0.05). Thus, ovarian factors
appear to regulate CA cell size through both cGMP-
dependent and cGMP-independent pathways.
Allatostatins are a family of small peptides released
by brain neurosecretory cells to inhibit JH synthesis
(Woodhead et al., 1989; Stay et al., 1992). Recently, it
has been shown that mature ovaries in D. punctata also
contain a large amount of allatostatin (Woodhead et al.,
2003). To examine whether the growth regulator in
ovaries was allatostatin, we measured cGMP levels in
CA glands after exposure to Dippu AST-7. Our results
indicated that the cGMP level in the CA glands was not
affected by the presence of Dippu AST-7 (Fig. 3C).
Thus, allatostatin was not the cause of the ovarian
induction of cGMP-dependent atrophy of CA cells.
3.7. Long-term arrest of JH production by cGMP
Protein content of CA glands cultured with 100 mM 8-
Br-cGMP for three days decreased approximately 30%
compared to controls (Table 3). Consistently, rates
of JH synthesis by CA glands cultured with 100 mM
ARTICLE IN PRESS
Table 2
Effects of trypsin and heat on the activity of the ovarian factor in inducing atrophy of CA cells
Treatment Control Ovary extract Paired t-test
Normal 20.2970.45 15.0670.34 Po0.001
Trypsin
a
19.5670.46 19.3270.21 P 0.53
Sham
b
19.9370.25 15.8770.43 Po0.001
Boiled
c
20.1570.40 16.8270.25 Po0.01
CA glands were cultured in medium containing extracts of 0.6 pairs of ovary equivalents per dish for three days before cell size measurements. Each
value represents the mean7SEM of cell diameter (mm) from six measurements.
a
Trypsin inhibitor was added after sample was treated with trypsin for 2.5 h.
b
Trypsin and trypsin inhibitor were added into sample simultaneously for 2.5 h.
c
The extract was boiled for 10 min, and then centrifuged at 3000g for 20 min.
Fig. 4. Effects of cGMP agonist and antagonist on changes in cell size.
(A) Dose response to 8-Br-cGMP. Degree of cell atrophy was
indicated by ratio of mean cell diameters between treatments and
controls. Each bar represents the mean7SEM from 6 to 9 measure-
ments. (B) Effects of IBMX (200 mM) (a) and cGMP antagonist (Rp-
8pCPT-cGMPS 500 mM) with ovarian extract or 100 mM 8-Br-cGMP
(b) on changes in cell size of mated day-4 CA cultured for 1821 and
4042 h, respectively. IBMX was added again after CA glands were
cultured for nine hours, and Rp-8-pCPT-cGMPS was added again at
the 12th and 24thh to sustain the effects of the drugs. Each bar
represents the mean 7SEM from 6 to 12 measurements; **, Po0.01;
***, Po0.001. (C) Effects of 100 mM 8-Br-cGMP on changes in cell
size of CA glands derived from mated day-2 (md2), day-4 (md4), day-
12 (md12) and virgin day-2 (vd2) cockroaches. fresh represents the
cell size of CA dissected freshly; control and 8-Br-cGMP
represent those of glands after cultured without and with 8-Br-cGMP
for 3 day, respectively. Each bar represents the mean7SEM from 4
measurements.
L.-W. Chang et al. / Insect Biochemistry and Molecular Biology 35 (2005) 4150 47
8-Br-cGMP for three days declined about 45% com-
pared to control glands (experiments vs. controls,
27.370.7 vs. 49.471.7 pmol/h/single CA) (Table 3).
These results suggest that in vivo atrophy of CA cells
and long-term low-JH production may be due to the
elevation of cGMP content when oocytes mature. On
the other hand, freshly dissected CA produced similar
amounts of JH, no matter whether incubated with or
without 8-Br-cGMP during the period of in vitro
radiochemical assay (48.675.3 vs. 45.773.4 pmol/h/
single CA, respectively). Thus, the inhibitory effects of
cGMP appear to be independent of rapid biochemical
modulation in the process of JH synthesis.
4. Discussion
4.1. Correlations between CA cell size and JH production
The maximum capacity of CA glands for JH
production appears determined by the amount of
cellular machinery available for hormone synthesis
(Chiang et al., 1991a, b). In virgin D. punctata adult
females, CA glands consist of small cells that have low
rates of JH synthesis. During oocyte maturation at days
04, these cells enlarge gradually and their rates of JH
production increase more than 5-fold; this process
follows mating of females or the disruption of the
brain-CA gland connection in virgins (Chiang et al.,
1998; Stay and Tobe, 1977). Following oocyte matura-
tion (days 58), CA cells undergo massive and
synchronous autophagy (Cheng and Chiang, 1995). As
a consequence, the cells shrink in size and JH produc-
tion is scaled back down (Chiang et al., 1998). In the
subsequent two months, the small size and arrested JH
production is maintained. The entire cycle is repeated
after parturition (Chiang et al., 1996).
Similar correlations between CA development and JH
production during ovarian cycles occur in at least two
other cockroach species, B. germanica and Supella
longipapla (Chiang and Schal, 1991, 1994). Three level
of CA regulatory signals have been proposed (Unnithan
et al., 1998), comprising cytological, constitutive and
dynamic responses. Presently, we report that intracel-
lular cGMP signaling regulates the plasticity of CA cells
(a cytological response). Also, while short-term expo-
sure to cGMP analogue appears not to affect JH
production in vitro, cell atrophy induced by long-term
exposure to high concentration of the analogue results
in signicant reduction of JH synthesis (Table 3, a
constitutive response). Thus, whether reduction in cell
size occurs in vivo or in vitro, the response shows a
strong positive correlation to low rates of JH synthesis.
4.2. How does cGMP affect cell size?
Atrophy of CA cells is a complicated process
involving inhibition of protein synthesis and sequential
autophagic activities. Autophagic digestion is a multi-
step process comprising elevation of lysosome and Golgi
activities, isolation of unwanted organelles with the
resultant formation of autophagosomes, engulfment of
autophagosomes by Golgi complex or fusion between
autophagosomes resulting in the formation of giant
autophagosomes, fusion between lysosomes and autop-
hagosomes to form autophagolysosomes and nally the
digestion of engulfed materials (Cheng and Chiang,
1995; Yang and Chiang, 1997; Fig. 2). Ultrastructural
observations conducted as part of the present study
showed that some autophagosomes occur in cells of CA
cultured with 8-Br-cGMP for one day (data not shown).
However, it is necessary to further examine quantita-
tively whether cGMP induces cell atrophy through
elevation of autophagic activity and/or reduction of
protein synthesis.
4.3. Regulation of CA activation
Mating or CA denervation induces a decrease in
cGMP content and cell growth in CA glands (Figs. 1A,
D). Unilateral CA denervation in virgin females results
in a decrease in cGMP content in both CA members,
although CA cell growth occurs only in the denerved
member. This result is consistent with the suggestion
that the decrease in cGMP and CA cell growth is
regulated by different mechanisms. It is not clear how
the denervation of CA affects the cGMP content of the
intact CA in the same individual, nor is it clear what
factors regulate CA cell growth. Since isolated CA
glands cultured in vitro exhibit cell growth (Table 1), we
ARTICLE IN PRESS
Table 3
Effects of 8-Br-cGMP on CA development and JH production
Short-term effects (Exp/ctrl) Long-term effects (Exp/ctrl)
JH production 101.976.7% (11) 55.471.12% (6)
Protein content 104.974.6% (7) 71.272.7% (4)
Direct effects of 8-Br-cGMP were determined by short-term incubation of freshly dissected CA derived from mated day-4 females in the presence
(experiment) or absence (control) of 8-Br-cGMP for 2 h. Long-term effects were determined after CA glands were cultured under the above
conditions for three days. A split pair test was adopted. Each datum represents the mean7SEM with the number of measurements indicated.
L.-W. Chang et al. / Insect Biochemistry and Molecular Biology 35 (2005) 4150 48
postulate that cGMP production in CA cells is under
continuous brain stimulation while cell growth is under
continuous brain inhibition in virgin females. It is
possible that low cGMP content may be a prerequisite,
but growth of CA cells also requires removal of the
inhibition from the brain.
4.4. Ovarian inhibition of growth of CA cells
We observed that active CA glands undergo cell
atrophy when the oocytes mature (Fig. 1). This is
consistent with the ovarian regulation of CA develop-
ment (Chiang et al., 1989; Rankin and Stay, 1984).
Furthermore, mechanisms controlling atrophy of CA
cells are independent of nerve connections, since cell
atrophy also occurs in denerved CA (Chiang et al., 1998;
Fig. 1A). In D. punctata, inactive CA glands of an adult
female transplanted into an adult male become active
and continue to synthesize JH at high rates (Stay et al.,
1980). But, if the CA glands are transplanted together
with nearly mature ovaries into males, they remain
inactive (Rankin and Stay, 1985). Based on these results
and our current ndings, we propose that the growth
and activity of CA cells are generally under brain
inhibition in virgin females. After mating or CA
denervation, CA cells are removed from brain inhibi-
tion, allowing them to grow larger and synthesize more
ovary growth-inducing JH. When ovaries mature, they
release peptidergic factor(s) to induce atrophy of CA
cells, resulting in the long-term arrest of JH production.
This scheme is entirely consistent with the cell size of CA
(Fig. 1) and the level of JH production cycle observed in
denerved virgin females.
How do the mature ovaries inhibit JH production in
vivo? In this report, we found that ovarian factor(s) may
indirectly inhibit JH production via the induction of
atrophy of CA cells. CA cells grew signicantly in size
when cultured in vitro, but they underwent atrophy
when cultured in medium containing mature ovaries or
ovarian extracts, or in medium pre-conditioned with
mature ovaries (Fig. 3, Table 1). In addition, this factor
may also control the expression of the CYP4C7 gene
encoding a cytochrome P450 terpenoid o-hydroxylase,
which is thought to metabolize JH and its precursor in
CA, causing the clearance of allatal JH and its
precursors at the end of the gonotrophic cycle (Suther-
land et al., 1998, 2000).
Trypsin sensitivity and partial heat-lability (Table 2)
suggest that the ovarian effector(s) may be a polypeptide
or a combination of peptide and non-peptide factors.
Furthermore, Rp-8-pCPT-cGMPS, an inhibitor of
cGMP-dependent protein kinase (PKG), prevented 8-
Br-cGMP and ovarian extract from inducing CA cell
atrophy. Thus, the ovarian effector may trigger the
membranous bound guanylyl cyclase to elevate cGMP
level, and then stimulate PKG to induce CA cell
atrophy. Allatostatins are short peptides released by
mature ovaries (Woodhead et al., 2003). We found
that Dippu AST-7 also induced atrophy of CA cells
(data unpublished); however, it appears to act via a
cGMP-independent mechanism since cGMP levels in
CA remained low in the presence of Dippu AST-7
(Fig. 3C; Cusson et al., 1992).
Finally, we conclude that mature ovaries release a
putative peptidergic growth regulator acting directly on
active CA cells and inducing the elevation of intracel-
lular cGMP contents. The cGMP signaling induces cell
atrophy and reduction in protein of the glands,
presumably resulting in depletion of cellular machinery,
and this causes long-term arrest of JH synthesis. Further
investigation will be required to purify the ovarian
factor and identify how cGMP signaling induces CA cell
atrophy.
Acknowledgments
We thank professor Babara Stay for critical reading
of the manuscript, and helpful suggestions. This work
was funded by grants from the National Science Council
(Taiwan), Brain Research Center of University System
of Taiwan, and Technology Development Program of
Ministry of Economy.
References
Cassier, P., 1990. Morphology, histology, and ultrastructure of JH-
producing glands in insects. In: Gupta, A.P. (Ed.), Morphogenetic
Hormone of Arthropods, vol. 2. Rutgers University Press, New
Brunswick, pp. 83194.
Cheng, H.-W., Chiang, A.-S., 1995. Autophagy and acid phosphatase
activity in the corpora allata of adult mated female of Diploptera
punctata. Cell and Tissue Research 281, 109117.
Chiang, A.-S., Schal, C., 1991. Correlation among corpus allatum
volume, cell size, and juvenile hormone biosynthesis in ovariecto-
mized adult Blattella germanica. Archives of Insect Biochemistry
and Physiology 18, 3744.
Chiang, A.-S., Schal, C., 1994. Cyclic volumetric changes in corpus
allatum cells in relation to juvenile hormone biosynthesis during
ovarian cycles in cockroaches. Archives of Insect Biochemistry and
Physiology 27, 5364.
Chiang, A.-S., Gadot, M., Schal, C., 1989. Morphometric analysis of
the corpus allatum cells in adult females of three cockroach species.
Molecular and Cellular Endocrinology 67, 179184.
Chiang, A.-S., Burns, E.L., Schal, C., 1991a. Ovarian regulation of
cyclic changes in size and activity of corpus allatum cells in
Blattella germanica. Journal of Insect Physiology 37, 907917.
Chiang, A.-S., Gadot, M., Burns, E.L., Schal, C., 1991b. Develop-
mental regulation of juvenile hormone synthesis: ovarian synchro-
nization of volumetric changes of corpus allatum cells in
cockroaches. Molecular and Cellular Endocrinology 75, 141147.
Chiang, A.-S., Tsai, W.H., Holbrook, G.L., Schal, C., 1996. Control of
cell proliferation in the corpora allata during the reproductive
cycle of the cockroach Diploptera punctata. Archives of Insect
Biochemistry and Physiology 32, 299313.
ARTICLE IN PRESS
L.-W. Chang et al. / Insect Biochemistry and Molecular Biology 35 (2005) 4150 49
Chiang, A.-S., Holbrook, G.L., Cheng, H.W., Schal, C., 1998.
Neural control of cell size in the corpora allata during the
reproductive cycle of the cockroach Diploptera punctata (Dictyop-
tera: Blaberidae). Invertebrate Reproduction & Development 33,
2534.
Chiang, A.-S., Lin, W.Y., Liu, H.P., Pszczolkowski, M.A., Fu, T.F.,
Chiu, S.L., Holbrook, G., 2002a. Insect NMDA receptors mediate
juvenile hormone biosynthesis. Proceedings of the National
Academy of Science of the United States of America 99, 3742.
Chiang, A.-S., Pszczolkowski, M.A., Liu, H.P., Lin, S.C., 2002b.
Ionotropic glutamate receptors mediate juvenile hormone synthesis
in the cockroach, Diploptera punctata. Insect Biochemistry and
Molecular Biology 32, 669678.
Cusson, M., Yagi, K.J., Guan, X.C., Tobe, S.S., 1992. Assessment of
the role of cyclic nucleotides in allatostatin-induced inhibition of
juvenile hormone biosynthesis in Diploptera punctata. Molecular
and Cellular Endocrinology 89, 121125.
Donaldson, J.G., Lippincott-Schwartz, J., Bloom, G.S., Kreis, T.E.,
Klausner, R.D., 1990. Dissociation of a 110-kD peripheral
membrane protein from the Golgi apparatus is an early event in
brefeldin A action. The Journal of Cell Biology 111, 22952306.
DellAntone, P., 1979. Evidence for an ATP-driven proton pump in
rat liver lysosomes by basic dyes uptake. Biochemical and
Biophysical Research Communications 86, 180189.
Feyereisen, R., 1985. Regulation of juvenile hormone titer: synthesis.
In: Kerkut, G.A., Gillert, L.I. (Eds.), Comprehensive Insect
Physiology, Biochemistry and Pharmacology, vol. 7. Pergamon
Press, Oxford, England, pp. 391429.
Feyereisen, R., Friedel, T., Tobe, S.S., 1981. Farnesoic acid stimula-
tion of C sub(16) juvenile hormone biosynthesis by corpora allata
of adult female Diploptera punctata. Insect Biochemistry 11,
401409.
Heinonen, J.K., Lahti, R.J., 1981. A new and convenient colorimetric
determination of inorganic orthophosphate and its application to
the assay of inorganic pyrophosphatase. Analytic Biochemistry
113, 313317.
Holbrook, G.L., Chiang, A.-S., Schal, C., 1997. Enhanced in vitro
juvenile hormone biosynthesis by cockroach corpora allata in an
arthropod cell culture medium. In Vitro Cellular & Developmental
BiologyAnimal 33, 452458.
Johnson, G.D., Stay, B., Rankin, S.M., 1985. Ultrastructure of
corpora allata of known activity during the vitellogenic cycle in the
cockroach Diploptera punctata. Cell and Tissue Research 239,
317327.
Johnson, G.D., Stay, B., Chan, K.K., 1993. Structure-activity
relationships in corpora allata of the cockroach Diploptera
punctata: roles of mating and the ovary. Cell and Tissue Research
274, 279293.
Pagano, R.E., Martin, O.C., Kang, H.C., Haugland, R.P., 1991. A
novel uorescent ceramide analogue for studying membrane trafc
in animal cells: accumulation at the Golgi apparatus results in
altered spectral properties of the sphingolipid precursor. The
Journal of Cell Biology 113, 12671279.
Pszczolkowski, M.A., Lee, W.S., Liu, H.P., Chiang, A.-S., 1999.
Glutamate-induced rise in cytosolic calcium concentration stimu-
lates in vitro rates of juvenile hormone biosynthesis in corpus
allatum of Diploptera punctata. Molecular and Cellular Endocri-
nology 158, 163171.
Rankin, S.M., Stay, B., 1984. The changing effect of the ovary on rates
of juvenile hormone synthesis in Diploptera punctata. General and
Comparative Endocrinology 54, 382388.
Rankin, S.M., Stay, B., 1985. Ovarian inhibition of juvenile hormone
synthesis in the viviparous cockroach, Diploptera punctata. General
and Comparative Endocrinology 59, 230237.
Stay, B., Tobe, S.S., 1977. Control of juvenile hormone biosynthesis
during the reproductive cycle of a viviparous cockroach. I.
Activation and inhibition of corpora allata. General and Com-
parative Endocrinology 33, 531540.
Stay, B., Friedel, T., Tobe, S.S., Mundall, E.C., 1980. Feedback
control of juvenile hormone synthesis in cockroaches: possible role
for ecdysterone. Science 207, 898900.
Stay, B., Chan, K.K., Woodhead, A.P., 1992. Allatostatin-immunor-
eactive neurons projecting to the corpora allata of adult Diploptera
punctata. Cell and Tissue Research 270, 1523.
Stay, B., Tobe, S.S., Bendena, W.G., 1994. Allatostatins: identica-
tion, primary structures, functions and distribution. Advances in
Insect Physiology 25, 267337.
Sutherland, T.D., Unnithan, G.C., Andersen, J.F., Evans, P.H.,
Murataliev, M.B., Szabo, L.Z., Mash, E.A., Bowers, W.S.,
Feyereisen, R., 1998. A cytochrome P450 terpenoid hydroxylase
linked to the suppression of insect juvenile hormone synthesis.
Proceedings of the National Academy of Science of the United
States of America 95, 1288412889.
Sutherland, T.D., Unnithan, G.C., Feyereisen, R., 2000. Terpenoid
omega-hydroxylase (CYP4C7) messenger RNA levels in the
corpora allata: a marker for ovarian control of juvenile hormone
synthesis in Diploptera punctata. Journal of Insect Physiology 46,
12191227.
Szibbo, C.M., Tobe, S.S., 1981. The mechanism of compensation in
juvenile hormone synthesis following unilateral allatectomy in
Diploptera punctata. Journal of Insect Physiology 27, 609613.
Taylor III, P.A., Bhatt, T.R., Horodyski, F.M., 1996. Molecular
characterization and expression analysis of Manduca sexta
allatotropin. European Journal of Biochemistry 239, 588596.
Tobe, S.S., 1990. Role of intracellular messengers in the regulation of
juvenile hormone biosynthesis in the cockroach, Diploptera
punctata. In: Epple, A., Scanes, C.G., Stetson, M.H. (Eds.),
Progress in Comparative Endocrinology. Wiley-Liss, Inc., New
York, pp. 174179.
Tobe, S.S., Pratt, G.E., 1976. Farnesenic acid stimulation of juvenile
hormone biosynthesis as an experimental probe in corpus allatum
physiology. In: Gilbert, L.I. (Ed.), The Juvenile Hormone. Plenum,
New York, pp. 147163.
Tobe, S.S., Stay, B., 1977. Corpus allatum activity in vitro during the
reproductive cycle of the viviparous cockroach, Diploptera punctata
(Eschscholtz). General and Comparative Endocrinology 31,
138147.
Unnithan, G.C., Sutherland, T.D., Cromey, D.W., Feyereisen, R.,
1998. A factor causing stable stimulation of juvenile hormone
synthesis by Diploptera punctata corpora allata in vitro. Journal of
Insect Physiology 11, 10271037.
Woodhead, A.P., Stay, B., Seidel, S.L., Khan, M.A., Tobe, S.S., 1989.
Primary structure of four allatostatins: neuropeptide inhibitors of
juvenile hormone synthesis. Proceedings of the National Academy
of Science of the United States of America 86, 59976001.
Woodhead, A.P., Thompson, M.E., Chan, K.K., Stay, B., 2003.
Allatostatin in ovaries, oviducts, and young embryos in the
cockroach Diploptera punctata. Journal of Insect Physiology 49,
11031114.
Xia, Y., Dean, P., Judge, A.J., Gillespie, J.P., Clarkson, J.M.,
Charnley, A.K., 2000. Acid phosphatases in the haemolymph of
the desert locust, Schistocerca gregaria, infected with the entomo-
pathogenic fungus Metarhizium anisopliae. Journal of Insect
Physiology 46, 12491257.
Yang, D.-M., Chiang, A.-S., 1997. Formation of a whorl-like
autophagosome by Golgi apparatus engulng a ribosome-contain-
ing vacuole in corpora allata of the cockroach Diploptera punctata.
Cell and Tissue Research 287, 385391.
ARTICLE IN PRESS
L.-W. Chang et al. / Insect Biochemistry and Molecular Biology 35 (2005) 4150 50
Insect
Biochemistry
and
Molecular
Biology
Insect Biochemistry and Molecular Biology 35 (2005) 5159
A broin secretion-decient silkworm mutant, Nd-s
D
, provides an
efcient system for producing recombinant proteins
Satoshi Inoue
a,1
, Toshio Kanda
a
, Morikazu Imamura
a,2
, Guo-Xing Quan
a,3
,
Katsura Kojima
a
, Hiromitsu Tanaka
a
, Masahiro Tomita
b
, Rika Hino
b
,
Katsutoshi Yoshizato
b
, Shigeki Mizuno
c
, Toshiki Tamura
a,
a
Insect Biotechnology and Sericology Department, National Institute of Agrobiological Sciences, 1-2 Owashi, Tsukuba, Ibaraki 305-8634, Japan
b
Hiroshima Tissue Regeneration Project, Hiroshima Prefecture Collaboration of Regional Entities for Advancement of Technological Excellence, Japan
Science and Technology Corporation, 3-10-32 Kagamiyama, Higashihiroshima, Hiroshima 739-0046, Japan
c
Department of Agricultural and Biological Chemistry, College of Bioresource Sciences, Nihon University, 1866 Kameino, Fujisawa 252-8510, Japan
Received 3 September 2004; received in revised form 12 October 2004; accepted 15 October 2004
Abstract
The silkworm Nd-s
D
mutant is silk broin-secretion decient. In the mutant, a disulde linkage between the heavy (H) and light
(L) chains, which is essential for the intracellular transport and secretion of broin, is not formed because of a partial deletion of the
L-chain gene. To utilize the inactivity of the mutant L-chain, we investigated the possibility of using the Nd-s
D
mutant for the
efcient production of recombinant proteins in the silkworm. A germ line transformation of the mutant with a normal L-chain-GFP
fusion gene was performed. In the transgenic mutant, normal development of the posterior silk gland (PSG) was restored and it
formed a normal cocoon. The biochemical analysis showed that the transgenic silkworms expressed the introduced gene in PSG
cells, produced a large amount of the recombinant protein, secreted it into the PSG lumen, and used it to construct the cocoon. The
molar ratio of silk proteins, H-chain:L-chain-GFP:brohexamerin, in the lumen and cocoon in the transgenic silkworm was 6:6:1,
and the nal product of the fusion gene formed about 10% of the cocoon silk. This indicates that the transgenic mutant silkworm
possesses the capacity to produce and secrete the recombinant proteins in a molar ratio equal to that of the broin H-chain,
contributing around half molecules of the total PSG silk proteins.
r 2004 Elsevier Ltd. All rights reserved.
Keywords: Intracellular transport; Transgenic; Recombinant protein; Bioreactor; PiggyBac; Fibroin; Silkworm; Bombyx; Nd-s mutant
1. Introduction
The use of a transgenic silkworm as a bioreactor for
the production of recombinant proteins has been
realized with the success of a germ line transformation
method using transposon piggyBac (Tamura et al.,
2000). The domesticated silkworm, Bombyx mori,
possesses efcient protein production organs called silk
glands, for making silk proteins. In this insect, more
than half of the dry weight of the cocoon consists of silk
proteins. Therefore, more than half of the nutrients
consumed by the larvae are converted into silk proteins.
Recently, it was reported that recombinant protein
could be produced in the silk gland of transgenic
silkworms (Tomita et al., 2003). However, the produc-
tion was not efcient. The major components of the
cocoon of the transgenic silkworm consisted of the
products of the native silk genes, suggesting that one of
ARTICLE IN PRESS
www.elsevier.com/locate/ibmb
0965-1748/$ - see front matter r 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ibmb.2004.10.002

Corresponding author. Tel./fax: +81 298 38 6091.


E-mail address: ttamura@affrc.go.jp (T. Tamura).
1
Present address: MRC Toxicology Unit, University of Leicester,
Leicester LE1 9HN, UK.
2
Present address: Prion Disease Research Center, National Institute
of Animal Health, Tsukuba, Ibaraki 305-0856, Japan.
3
Present address: Molecular Entomology, Great Lakes Forestry
Centre, Canadian Forest Service, Marie, Ontario, Canada P6A 5M7.
the critical factors for the production of recombinant
protein is related to the system of silk protein secretion.
The major components of cocoon silk in non-
transgenic normal silkworms are broin and sericin.
Fibroin is synthesized in the posterior silk gland (PSG)
and accumulates in the lumen of the middle silk gland
(MSG), where sericin is synthesized. Then, the broin is
secreted into the cocoon via the anterior silk gland
(ASG). The silk broin contains three polypeptides: a
350-kDa heavy chain (H-chain, Shimura, 1983), a 26-
kDa light chain (L-chain, Yamaguchi et al., 1989), and
brohexamerin (fhx), originally named P25(Inoue et al.,
2000). In the PSG, these proteins form a large protein
complex called an elementary unit of broin, which
consists of a 6:6:1 molar ratio of the H-chain:L-
chain:brohexamerin (Inoue et al., 2000). The L-chain
is linked with the H-chain by a disulde bond between
Cys-172 of the L-chain and Cys-c20 (twentieth residue
from the carboxy terminus) of the H-chain (Tanaka et
al., 1999b). One molecule of brohexamerin interacts
non-covalently with six sets of H-L heterodimers to
form the elementary unit in the endoplasmic reticulum
(ER) (Inoue et al., 2004), and the three N-linked
oligosaccharide chains of brohexamerin contribute to
maintaining the integrity of the elementary unit (Inoue
et al., 2000). This molecular complex is essential for
quality control in the ER (Reddy and Corley, 1998;
Ellgaard et al., 1999; Fassio and Sitia, 2002), enabling
the efcient intracellular transport and secretion of
signicant amounts of broin into the lumen (Inoue et
al., 2004).
The Nd-s
D
mutant is characterized by an immature
PSG and less than 1% of the normal secretion level of
broin, which leads to the production of a very thin,
naked-pupa cocoon that consists mostly of sericin. The
gene is mapped to the L-chain gene (b-L) locus on
chromosome 14 (Takei et al., 1984a). The mutation is
caused by a deletion downstream from exon III, which
causes the recombination of sequences in the third
intron with sequences in the far downstream region.
This recombination created a chimeric gene containing
the rst three exons of b-L and two new exons, IV
0
and
V
0
, from the far downstream region. This mutant
chimeric Nd-s
D
L-chain lacks Cys-172, which was
encoded by the original sixth exon, and consequently
cannot form a disulde linkage with the H-chain (Mori
et al., 1995). The HL disulde linkage contributes to
the efcient intracellular transport and secretion of silk
broin, which is analogous to IgM production (Sitia et
al., 1990), and the accumulation of L-chain-decient
broin in the ER might inhibit the development of PSG
cells. It has been reported that the normal L-chain
introduced into the Nd-s
D
mutant is secreted in the
lumen and this suggests that the mutant might be used
to increase the production and secretion of L-chain-
fused recombinant protein.
In this study, the L-chain-GFP fusion gene was
introduced into the Nd-s
D
mutant to assess its ability
to produce recombinant protein. The L-chain GFP
fusion gene was expressed in the PSG of the mutant
transgenic silkworm; large amounts of recombinant
protein were secreted into the lumen and spun into the
cocoon. The Nd-s
D
phenotype, which consisted of an
immature PSG and thin cocoon, was rescued dramati-
cally and the cocoon was almost indistinguishable from
those of normal silkworms. A single cocoon contained
about 8.6 mg dry weight (16.6 mmol) of L-chain-GFP
fusion protein, which corresponds to about 10% of the
total cocoon silk protein. The molar ratio of H-chain:L-
chain-GFP:brohexamerin in the secreted broin was
strictly 6:6:1, indicating that the formation of complexes
of the three proteins is critical, even with the fusion
protein.
2. Materials and methods
2.1. Silkworms
Fertilized eggs of broin-secreting wild-type B. mori
C108 and the broin-secretion-decient, naked-pupa
mutant of B. mori Nd-s
D
(with a mutation of the broin
L-chain gene, b-L) (Takei et al., 1984a; Mori et al.,
1995) were supplied by the Insect Genetics Laboratory
of the National Institute of Agrobiological Sciences,
Japan. The larvae were reared on an articial diet
(Nihon Nosan) at 25 1C.
2.2. Plasmid vectors
The helper plasmid vector, pHA3PIG (6.2 kb), is
described elsewhere (Tamura et al., 2000). The DsRed2
cDNA that was excised from the pDsRed2-1 vector
(Clontech) was used to replace EGFP cDNA in
pBac(3 P3-EGFPafm) (Horn and Wimmer, 2000),
giving rise to pBac(3 P3-DsRed2). The L-chain 5
0
-
anking region (from 600 to +34) (Kikuchi et al.,
1992), the L-chain cDNA sequence (from +34 to +767)
[2], and the L-chain 3
0
-anking region (from +13114 to
+13597) (Kikuchi et al., 1992) were amplied by PCR
from genomic DNA prepared from B. mori. EGFP
cDNA was excised from the pEGFP vector (Clontech).
All of the cDNA fragments were inserted into
pBac(3xP3-DsRed2) to produce pBac(3xP3-
DsRed2+L-chain-GFP) (Fig. 1A).
2.3. Transformation
Plasmid DNA was puried using the QIAGEN Midi
(Qiagen). To break embryonic diapause, eggs were
treated within 4 h after oviposition with 20% HCl for
1 h at 25 1C. A mixture of helper plasmid pHA3PIG
ARTICLE IN PRESS
S. Inoue et al. / Insect Biochemistry and Molecular Biology 35 (2005) 5159 52
(23 nl) [11] and vector DNA (0.2 mg/ml of each) in
0.5 mM phosphate buffer (pH 7.0) with 5 mM KCl was
injected into 476 eggs 4 h after the acid treatment.
Injected eggs were placed in a humidied box at 25 1C
until hatching. Hatched larvae were transferred to an
articial diet and grown to adulthood. G
0
moths were
mated within the same family, and 7-day G
1
embryos
were screened for DsRed2 expression in the stemmata
and nervous system tissues (Thomas et al., 2002) with a
Leika MZFL III uorescence microscope (Leika)
equipped with lter sets for DsRed. DsRed2 expression
in screened silkworms was conrmed in the compound
eyes of G
1
moths. Moths expressing DsRed2 were mated
with each other. The G
2
larvae that were most strongly
positive for DsRed2 were used as transgenic lines. Green
uorescent protein in the dissected silk gland and
cocoon was observed with a Leika MZFL III uores-
cence microscope (Leika) equipped with lter sets for
GFP2.
2.4. Southern blot analysis
Genomic DNA was extracted from G
2
moths
(Tamura et al., 2000). EcoRI-digested genomic DNA
(5 mg) was subjected to Southern blotting and hybridized
with the L-chain 3
0
anking region (from +887 to
+1354) from pFL18 (Yamaguchi et al., 1989) as a probe
for the normal L-chain gene.
2.5. Northern blot analysis
Total RNA from PSG or MSG at the 3rd day of the
5th instar was puried using ISOGEN (Nippongene).
Total RNA was subjected to Northern blotting and
hybridized with the probe as follows. Part of the L-chain
cDNA sequence (+1 to +357; corresponding to exons I
to III) that is common to both the normal and Nd-s
D
L-
chains was amplied from pFL18 by PCR (Yamaguchi
et al., 1989). Part of the GFP cDNA sequence (+1 to
+357) was amplied from pEGFP (Clontech) by PCR.
Part of the B. mori elongation factor a-I isoform cDNA
sequence (from +1 to +1682) was amplied from
pBmEF-1 a-1 by PCR (Kamiie et al., 1993).
2.6. Protein analysis
Preparation of broin samples from PSG lumen, PSG
tissues, and cocoons has been described elsewhere
(Tanaka et al., 1999a; Inoue et al., 2000). Western
blotting of broin samples was carried out, as described
previously (Inoue et al., 2000), with the following
antibodies: rabbit anti-H-chain or anti-L-chain poly-
clonal antibody (Inoue et al., 2000), mouse anti-fhx
polyclonal antibody (Inoue et al., 2000), or rabbit anti-
GFP (Invitrogen) antibodies.
2.7. Purication of L-chain-GFP
To purify the L-chain-GFP fusion protein, a broin
sample from the PSG lumen was subjected to 0.1%
SDS, 12.5% PAGE (SDS-PAGE); the L-chain-GFP
fusion protein was electro-eluted in 5 mM Tris, 38.4 mM
glycine. The purity of the L-chain-GFP fusion protein
was judged by silver staining and Western blotting.
2.8. Quantitative ELISA
Standard curves for the H-chain, L-chain, and
brohexamerin are described elsewhere (Inoue et al.,
2000). Standard curves for L-chain-GFP were made
using the puried L-chain-GFP protein; the protein was
dissolved in 20 mM TrisHCl (pH 8.0), 2 M urea, and
subjected to ELISA using rabbit anti-L-chain antibody.
Protein samples prepared from the PSG lumen or the
cocoon shell were dissolved in 20 mM TrisHCl (pH
8.0), 2 M urea, and subjected to quantitative ELISA as
described elsewhere [3].
3. Results
3.1. Construction of Nd-s
D
mutant transgenic silkworms
We constructed the transformation vector
pBac(3 P3-DsRed2+L-chain-GFP) (Fig. 1A) and
ARTICLE IN PRESS
Fig. 1. Restriction enzyme map of a piggyBac vector and genomic
DNA of the normal breed (C108) and Nd-s
D
. (A) Design of a piggyBac
vector for L-chain-GFP expression in PSG cells. The DsRed2 marker
gene construct is placed immediately downstream from the Pax-6
articial promoter sequences (3 P3), enabling its expression in
photoreceptor cells. The L-chain promoter is fused to the L-chain-
GFP coding sequence and L-chain 3
0
-untranslated sequences. Inserted
terminal repeats of piggyBac transposase (arrows) and EcoRI sites (E)
are indicated. (B) Physical maps of the normal (C108) and Nd-s
D
mutant L-chain genes. The normal gene (Fib-L) contains seven exons
(exon I-VII), whereas the chimeric mutant gene contains exons IIII of
Fib-L and exons IV
0
and V
0
derived from the far downstream region.
S. Inoue et al. / Insect Biochemistry and Molecular Biology 35 (2005) 5159 53
injected it with helper plasmid into 476 eggs of Nd-s
D
at
the preblastodermal stage. Fifteen transgenic silkworms
were veried by screening for DsRed2 emission in the
stemmata and nervous tissues of embryos or moths (Fig.
2B), as reported by Thomas et al. (2002) in the
transgenic line containing the 3XP3GFP insert, and
even in a silkworm with normal brown color stemmata
and compound eyes. These silkworms were mated with
each other to produce the LGL4 5 line, which was
characterized by a strong DsRed2 emission in the
stemma and complex eyes. To conrm the insertion of
the fusion gene and to identify the number of integrated
DNA molecules in the Nd-s
D
genome of the transgenic
line, the EcoRI-digested genomic DNA of wild type, Nd-
s
D
, and the LGL4 5 line were investigated by Southern
blot hybridization with a normal L-chain-specic probe
(Fig. 1B). The probe hybridized with one specic 8.7-kb
band in wild-type silkworms (Fig. 2B lane 1), no band in
Nd-s
D
mutants (Fig. 2B lane 2), as expected, and 9.8-,
7.8-, 5.6-, and 4.1-kb bands in the LGL4 5 line (Fig.
2B lane 3). This indicated that the LGL4 5 line was a
transgenic line containing the normal L-chain-GFP
fusion gene in at least four different loci of the original
Nd-s
D
genome.
3.2. Restoration of PSG development and broin
secretion in the transgenic line
The most prominent phenotypic features of the Nd-s
D
mutant silkworm are an immature PSG (Fig. 3A(b)) and
the production of a very thin cocoon (Fig. 3B(b))
relative to that of a normal silkworm (Figs. 3A(a), B(a)).
The phenotypes of the transgenic line LGL4 5 were
dramatically changed from those of Nd-s
D
with respect
to the extensive development of the PSG (Fig. 3A(c))
and the production of a thick cocoon (Fig. 3B(c)) similar
to that of the normal C108 (Figs. 3A(a), B(a)). The
mean weight of cocoons produced by transgenic line
LGL4 5 was 78.4 mg (n 50), which was much
heavier than that of Nd-s
D
(17.7 mg; n 50), but less
than half that of the normal C108 (187 mg; n 50).
These results indicate that the prominent phenotypes of
the Nd-s
D
mutant can be largely converted to those of
the normal broin-producing silkworm by transgenesis
with the normal L-chain-GFP fusion gene.
3.3. Expression and secretion of functional L-chain GFP
fusion protein in transgenic silkworms
The emission of green uorescence due to the
expression of the integrated GFP gene was observed in
the transgenic silkworm at the 5th day of the 5th instar,
when the secretion of broin reached the maximal level,
in the PSG cells and the lumen of whole silk glands
(from PSG, MSG to ASG) (Fig. 3A(c) and (c*)). The
uororescence was not detected in other tissues at whole
stages, suggesting that the fusion gene specically
expressed in PSG cells. Green uorescence was also
detected in the entire cocoon produced by the transgenic
silkworm (Fig. 3B(c)). These observations suggested
that the L-chain-GFP fusion protein was expressed from
the integrated fusion gene in the transgenic silkworm.
ARTICLE IN PRESS
Fig. 2. Morphological characters of transgenic Nd-s
D
mutant inserted
using the 3XP3DsRed2 marker gene with genomic Southern blotting.
(A) Expression of DsRed2 in the compound eyes of G
1
moths (upper)
and in the stemmata of a 7-day-old G
2
embryo (lower). Note that the
mutant possesses black stemmata and compound eyes. The marker is
still useful for detecting transgenic silkworm. Left panel; Nd-s
D
, right
panel; transgenic line. An arrowhead and arrows point to the ocelli and
abdominal nervous system, respectively. Scale bars: 0.5 mm. (B)
Southern blot of EcoRI-digested genomic DNA (5 mg each) from
C108 (lane 1), Nd-s
D
(lane 2), and the transgenic line (lane 3)
hybridized with the normal L-chain specic probe, as shown in
Fig. 1B.
S. Inoue et al. / Insect Biochemistry and Molecular Biology 35 (2005) 5159 54
In order to prove this, Northern blots of total RNA
preparations from PSG or MSG cells of the normal
breed C108, Nd-s
D
, or the transgenic line LGL4 5 were
probed with cDNA for the L-chain sequence derived
from exons I, II, and III (common to C108 and Nd-s
D
L-
chains) or cDNA for GFP. The L-chain cDNA probe
hybridized to 1.3- and 1.1-kb bands in the RNA
preparations from the PSGs of C108 and Nd-s
D
,
respectively (Fig. 4A, left panel, lanes 1 and 2). The
amount of Nd-s
D
L-chain mRNA was less than 5% of
that in C108. The L-chain cDNA probe did not
hybridize to RNA preparations from the MSG of either
C108 or Nd-s
D
(Fig. 4A, left panel, lanes 4 and 5). These
results are consistent with those reported previously
(Takei et al., 1987).
The L-chain cDNA probe hybridized to 2.1- and 1.1-
kb components in the RNA preparation from the PSG
of transgenic line LGL4 5 (Fig. 4A, left panel, lane 3).
ARTICLE IN PRESS
Fig. 3. GFP uorescence in the PSG and cocoon of the transgenic line.
(A) A pair of silk glands at the 5th day of the 5th instar larvae of C108
(a), Nd-s
D
(b), and the transgenic line (c) were observed under bright
eld (bright) and with the GFP uorescence system (GFP). The ASG,
MSG, and PSG of the transgenic line were observed under higher
magnication with the EGFP uorescence system (c*). ASG: anterior
silk glands, MSG: middle silk glands, PSG; posterior silk glands. (B)
Cocoons produced by C108 (a), Nd-s
D
(b), and the transgenic line (c)
were observed similarly.
Fig. 4. Detection of L-chain-GFP transcript and protein in the
transgenic line. (A) Northern blots of total RNA (5 mg/lane) from the
PSG (lanes 13) and MSG (lanes 46) of C108 (lanes 1 and 4), Nd-s
D
(lanes 2 and 5), and the transgenic line (lanes 3 and 6) hybridized with
the L-chain cDNA probe (exons IIII, 357 bp) common to both
normal and Nd-s
D
L-chain mRNA, GFP cDNA probe (357 bp), or the
cDNA probe for the B. mori elongation factor a-I isoform. (B) Western
blot of cocoon proteins of C108 (a), Nd-s
D
(b), and the transgenic line
(c). The PSG tissue (lanes 1 and 3) and the cocoon (lanes 2 and 4)
proteins were separated by SDS-PAGE without reduction (lanes 1 and
2) or under reducing conditions (lanes 3 and 4) and subjected to
Western blotting with anti-H-chain (left), anti-L-chain (middle), or
anti-GFP antibodies (right).
S. Inoue et al. / Insect Biochemistry and Molecular Biology 35 (2005) 5159 55
The 2.1-kb, but not the 1.3-kb, component hybridized
with the GFP cDNA probe (Fig. 4A, right panel, lane
3). The hybridizing RNA bands were interpreted as
normal L-chain mRNA (1.3 kb), Nd-s
D
L-chain mRNA
(1.1 kb), and normal L-chain-GFP fusion gene mRNA
(2.1 kb). Judging from the relatively strong intensity of
the 2.1-kb band, it appeared that the normal L-chain-
GFP fusion transgenes were actively transcribed in a
PSG-specic manner in the transgenic line LGL4 5.
To detect L-chain or L-chain-GFP fusion polypep-
tides, proteins from PSG tissue or from cocoon shells of
C108, Nd-s
D
, or LGL4 5 were subjected to Western
blotting using anti-H-chain, anti-L-chain, or anti-GFP
antibodies, before and after the reductive cleavage of
disulde bonds. In the normal C108 strain, the L-chain
was detected in two forms, either unbound or bound to
H-chain, in the PSG tissue sample (Fig. 4B(a), lanes 1
and 3), but only the H-chain-bound form was detected
in the cocoon sample (Fig. 4B(a), lanes 2 and 4), as
reported (Takei et al., 1987). In Nd-s
D
, the mutant L-
chain was detected in an unbound form in the PSG
tissue sample (Fig. 4B(b), lanes 1 and 3), but not in the
cocoon sample (Fig. 4B(b), lanes 2 and 4), as reported
(Takei et al., 1987). In the transgenic line, the anti-L-
chain antibody reacted with 52.5- and 27-kDa bands in
the PSG tissue sample after reducing the disulde bonds
(Fig. 4B(c), middle panel, lane 3), whereas the anti-GFP
antibody reacted with only the 52.5-kDa band under the
same conditions (Fig. 4B(c), right panel, lane 3). These
results strongly suggest that the 52.5-kDa band is the
normal L-chain-GFP fusion protein and that the 27-
kDa band is the Nd-s
D
mutant L-chain. As expected,
only the normal L-chain-GFP fusion protein was
detected in the cocoon protein sample in the reactions
with anti-L-chain antibody or anti-GFP antibody (Fig.
4B(c), middle and right panels, lane 4). The normal L-
chain-GFP fusion protein was associated with the H-
chain before reduction of the disulde bonds, as shown
by the reaction with both anti-L-chain and anti-GFP
antibodies (Fig. 4, middle and right panels, lane 2).
These results suggest that the normal L-chain-GFP
fusion genes, which were integrated into the genome of
the transgenic silkworm, are expressed as proteins that
behave like the normal L-chain and contribute to the
restoration of the phenotype with respect to high
secretion levels of broin and the production of nearly
normal cocoons.
3.4. Quantication of the normal L-chain-GFP fusion
protein in cocoons
The molar ratio of H-chain:L-chain:brohexamerin in
the elementary unit of broin produced by the normal-
level broin-producing breed of silkworms is 6:6:1
(Inoue et al., 2000). We used quantitative ELISA to
determine whether the normal L-chain-GFP fusion
protein expressed in the transgenic silkworm partici-
pated in the formation of an elementary unit with a
similar molar ratio.
For this purpose, the normal L-chain-GFP fusion
protein was puried from the broin secreted into the
PSG lumen of the transgenic silkworm, as described in
the Methods and materials. The purity of the protein
sample was examined by SDS-PAGE, followed by silver
staining or Western blotting. It behaved as a single 52.5-
kDa band (Fig. 5A(a)) and reacted with the anti-L-chain
or anti-GFP antibody, but not with the anti-H-chain or
anti-brohexamerin antibody (Fig. 5A(b)(e)). This
protein sample was then used to make an ELISA
standard curve for the L-chain-GFP fusion protein (Fig.
5B). ELISA standard curves for H-chain and brohex-
amerin were made, as described (Inoue et al., 2000). The
results of the quantitative ELISA for cocoon protein
samples are summarized in Table 1. The molar ratio of
H-chain:L-chain:brohexamerin was close to 6:6:1 for
the normal breed C108, 6:0:1 for Nd-s
D
, and close to
6:6:1 for the transgenic line LGL4 5. These results
ARTICLE IN PRESS
Fig. 5. Purication of L-chain-GFP and the production of a standard
curve for quantitative ELISA. (A) Purication of L-chain-GFP. A
broin sample (lane 1) from the transgenic line was separated by SDS-
PAGE without cleaving the disulde bonds, and L-chain-GFP (lane 2)
was recovered by electro-elution. The purity was judged by silver
staining (a), and Western blotting using anti-H-chain (b), anti-L-chain
(c), anti-brohexamerin (d), and anti-GFP (e) antibodies. (B) ELISA
for the puried L-chain-GFP. The puried L-chain-GFP was subjected
to ELISA using anti-L-chain antibody. The reactivity was monitored
by the reaction with horseradish peroxidase-conjugated anti-
rabbit IgG followed by the color-developing reaction (A
490
) with
o-phenylenediamine dihydrochloride.
S. Inoue et al. / Insect Biochemistry and Molecular Biology 35 (2005) 5159 56
suggest that the L-chain-GFP fusion proteins expressed
in the transgenic silkworm participated in the assembly
of an elementary unit of broin whose molecular
constitution was similar to that formed in the normal
breed, irrespective of the presence of Nd-s
D
mutant L-
chain proteins in the PSG cells. The content of L-chain-
GFP fusion protein was calculated to be about 110 mg
(212 nmol)/mg of the cocoon protein or about 8.6 mg
(16.6 mmol) per dried cocoon (78.4 mg) produced by the
transgenic silkworm.
4. Discussion
We demonstrated that the Nd-s
D
mutant, which
possesses degenerated PSG and expels only sericin in
the cocoon, is useful for producing a large amount of L-
chain fused recombinant proteins in the transgenic
silkworm. The amount of the L-chain-GFP fusion
proteins in our experiment exceeded 10% of the proteins
in the cocoon when the gene was introduced into the
mutant. In addition, the Nd-s
D
mutant was rescued by
the introduction of L-chain-GFP fusion protein gene,
suggesting that the protein production and secretion
machinery of the mutant PSG had recovered.
We used the fusion gene with a promoter from the
600-bp upstream region of the L-chain gene. The
expression of the inserted fusion gene was apparently
lower than that in the normal type assessed by Northern
blotting (Fig. 4A compare lanes 1 and 3). However, the
tissue-specicity of the inserted gene was maintained in
the transgenic silkworm, indicating that signals related
to the tissue-specic expression of the L-chain gene are
localized to this region. There are two possible explana-
tions for the low level of gene expression: (i) the inserted
gene disturbed a critical chromatin structure at the
insertion position, which affected its expression, or (ii)
the binding sites of some transcriptional activators were
not present in the promoter region used. For the latter
possibility, using a longer upstream region as the
promoter may increase the mRNA level. The broin
H- and L-chain genes and the brohexamerin gene share
the same transcriptional regulation system, and are
expressed in a coordinated manner in the PSG (Bello et
al., 1994), despite the fact that the three genes are
located on different chromosomes: FibH is on chromo-
some 25, b-L is on chromosome 14, and brohexamerin
is on chromosome 2. Otherwise, it might be better to use
the promoter of H-chain, or brohexamerin genes for
the L-chain, or the targeted gene expression system of
GAL4/UAS (Imamura et al., 2003). Reduced expression
of the introduced gene, as compared to the endogenous
gene, is also observed when the promoters of other genes
are used (Suzuki et al., 2003).
Another important factor is the secretion of the
recombinant proteins in PSG. The ER provides an
ARTICLE IN PRESS
T
a
b
l
e
1
M
o
l
a
r
r
a
t
i
o
s
o
f
H
-
c
h
a
i
n
,
L
-
c
h
a
i
n
o
r
L
-
c
h
a
i
n
-
G
F
P
,
a
n
d
f
h
x
i
n
t
h
e

b
r
o
i
n
o
f
c
o
c
o
o
n
s
B
r
e
e
d
a
A
4
9
0
n
g
/
p
r
o
t
e
i
n
(
1
0
0
m
g
)
p
m
o
l
M
o
l
a
r
r
a
t
i
o
H
-
c
h
a
i
n
L
-
c
h
a
i
n
o
r
L
-
c
h
a
i
n
-
G
F
P
F
h
x
H
-
c
h
a
i
n
L
-
c
h
a
i
n
o
r
L
-
c
h
a
i
n
-
G
F
P
f
h
x
H
-
c
h
a
i
n
L
-
c
h
a
i
n
o
r
L
-
c
h
a
i
n
-
G
F
P
f
h
x
H
-
c
h
a
i
n
:
L
-
c
h
a
i
n
o
r
L
-
c
h
a
i
n
-
G
F
P
:
f
h
x
C
1
0
8
0
.
5
1
7
0
.
3
5
4
0
.
1
9
6
6
.
3
1

1
0
4
.
5
.
7
5

1
0
3
1
.
1
2

1
0
3
2
.
2
1

1
0
4
2
.
2
3

1
0
4
3
.
7
3

1
0
3
5
.
9
2
7
0
.
0
4
:
5
.
9
8
7
0
.
0
5
:
1
N
d
-
s
D
0
.
4
0
1
N
D
0
.
4
2
1
2
4
5
N
D
3
7
0
N
D
1
1
.
1
6
.
3
1
7
0
.
0
7
:

:
1
T
r
a
n
s
g
e
n
i
c
l
i
n
e
0
.
6
2
2
0
.
2
7
4
0
.
1
8
7
.
5
8

1
0
4
1
.
1

1
0
4
1
.
0
4

1
0
3
2
.
1
7

1
0
4
2
.
1
4

1
0
4
3
.
4
7

1
0
3
6
.
2
0
7
0
.
0
7
:
6
.
1
7
7
0
.
0
3
:
1
n

6
t
i
m
e
s
;
7
,
S
D
;
N
D
,
n
o
t
d
e
t
e
c
t
e
d
.
a
V
a
l
u
e
s
o
b
t
a
i
n
s
b
y
E
L
I
S
A
.
S
a
m
p
l
e
s
(
a
l
l
i
n
1
0
0
m
l
)
a
s
s
a
y
e
d
w
e
r
e
1
0
n
g
(
f
o
r
H
-
c
h
a
i
n
)
,
5
0
n
g
(
f
o
r
L
-
c
h
a
i
n
)
a
n
d
1
2
5
n
g
(
f
o
r
f
h
x
)
p
r
o
t
e
i
n
s
f
r
o
m
t
h
e
c
o
c
o
o
n
s
o
f
C
1
0
8
a
n
d
t
h
e
t
r
a
n
s
g
e
n
i
c
l
i
n
e
;
2
m
g
(
f
o
r
H
-
c
h
a
i
n
)
,
1
0
m
g
f
o
r
L
-
c
h
a
i
n
,
1
0
0
m
g
f
o
r
f
h
x
p
r
o
t
e
i
n
s
f
r
o
m
N
d
-
s
D
.
S. Inoue et al. / Insect Biochemistry and Molecular Biology 35 (2005) 5159 57
environment and mechanisms to facilitate the proper
formation of newly synthesized proteins into higher
order conformations. The quality control function of the
ER (Hammond and Helenius, 1995) is critical to
ensuring that only properly folded proteins enter
secretory pathways. In the case of silk broin, a disulde
linkage between the H- and L-chains (Takei et al.,
1984b, 1987), the addition of three N-linked oligosac-
charide chains onto brohexamerin (Tanaka et al.,
1999a; Inoue et al., 2000), and the assembly of the
elementary unit (Inoue et al., 2000) all take place in the
ER (Inoue et al., 2004). These processes are important
for the efcient secretion of broin. The signicance of
the HL linkage for the efcient secretion of broin has
been suggested by genetic and biochemical studies of the
broin secretion decient Nd-s
D
mutant (Takei et al.,
1987; Mori et al., 1995; Tanaka et al., 1999b).
The application of the transgenic silkworm as a
bioreactor to produce recombinant proteins has several
advantages: (i) the cost of silkworm rearing is much
lower than the cost of rearing other animals; (ii) large
quantities of recombinant protein can be produced
(0.52 mmol broin per dried 200500 mg cocoon); (iii)
post-translational modications are similar to those of
mammalian proteins, except for oligosaccharide proces-
sing; (iv) the protein composition is relatively simple;
and (v) the risk of disease is lower.
Recently, Tomita et al. (2003) constructed a wild-type
transgenic silkworm containing the L-chain-GFP and L-
chain-human type III procollagen genes; the exogen-
ously expressed L-chain-GFP protein was secreted and
was found in the dried cocoon. However, the secretion
efciency of the fusion proteins in the wild-type
transgenic silkworm was very low. By contrast, the
Nd-s
D
mutant transgenic silkworm we established
showed much higher production of the recombinant
protein. The reason for this higher production by the
mutant is outlined in Fig. 6. The fusion protein
produced in the normal breed competes with the normal
L-chain in the process of formation of the HL linkage.
The endogenous normal L-chain has a greater afnity
for the H-chain than does the L-chain-GFP fusion
protein. Therefore, the wild-type transgenic silkworm
secreted relatively more endogenous normal L-chain
than the fusion protein. Actually, the molar ratio of
endogenous L-chain to L-chain-GFP fusion protein in
the cocoon of the transgenic normal breed is about 10:1,
concomitant with the accumulation of a relatively large
amount of fusion protein in PSG cells (S. Inoue et al.,
unpublished data). However, the mutant L-chain in the
transgenic Nd-s
D
mutant cannot compete with the
fusion protein. Therefore, only the fusion protein is
secreted into the lumen. Indeed, the fusion protein was
secreted efciently. In this study, a cocoon of the mutant
contained about 8.6 mg (16.6 mmol) of the functional
(green uorescence emitting) L-chain-GFP fusion pro-
tein (Table 1). The efcient secretion of L-chain-GFP
fusion protein provides the Nd-s
D
mutant transgenic line
with the advantage of higher protein yields than in the
wild-type transgenic silkworm. In addition, the L-chain-
GFP fusion protein is more easily puried from cocoons
of the Nd-s
D
mutant transgenic line, because the
cocoons lack the endogenous L-chain.
Our future work will focus on designing other
functional proteins and establishing an easy purication
system with high yields of the recombinant proteins
from cocoons. The tandem afnity purication (TAP)
system may prove to be a powerful tool for the
purication of functional foreign recombinant proteins
from cocoons (Rigaut et al., 1999).
Acknowledgments
We would like to thank Dr. Ernst A. Wimmer of
Universitat Bayeyth for kindly providing pBac(3 P3-
EGFPafm), Kazuko Seo and Hiroko Yamazaki for
technical assistance, and Dr. Michelle A. Hughes
(University of Leicester) for a critical reading of the
manuscript. This work was supported by the Ministry of
Agriculture, Forest, and Fisheries and by the Program
for the Promotion of Basic Research Activities for
Innovative Bioscience, Japan.
References
Bello, B., Horard, B., Couble, P., 1994. The selective expression of
silkprotein-encoding genes in Bombyx mori silk gland. Bull. Inst.
Pasteur 92, 81100.
Ellgaard, L., Molinari, M., Helenius, A., 1999. Setting the standards:
quality control in the secretory pathway. Science 286, 18821888.
ARTICLE IN PRESS
Fig. 6. Model of the intracellular transport of recombinant proteins
produced in the transgenic silkworms. H, broin H-chain; L, broin L-
chain; Lm, broin L-chain produced by Nd-s
D
mutant; fhx,
brohexamerin. The L-chain-GFP fusion protein (L+GFP) produced
in the PSG cells of the normal breed has lower afnity in the formation
of an SS linkage with the H-chain, compared with the normal L-
chain. Therefore, less of the fusion protein is secreted in the normal
breed. However, the fusion protein in the mutant does not compete
with the mutant L-chain. Therefore, only the fusion protein linked
with H-chain is secreted in the mutant.
S. Inoue et al. / Insect Biochemistry and Molecular Biology 35 (2005) 5159 58
Fassio, A., Sitia, R., 2002. Formation, isomerisation and reduction of
disulphide bonds during protein quality control in the endoplasmic
reticulum. Histochem. Cell Biol. 117, 151157.
Hammond, C., Helenius, A., 1995. Quality control in the secretory
pathway. Curr. Opin. Cell Biol. 7, 523529.
Horn, C., Wimmer, E.A., 2000. A versatile vector set for animal
transgenesis. Dev. Genes Evol. 210, 630637.
Imamura, M., Nakai, J., Inoue, S., Quan, G.-X., Kanda, T., Tamura,
T., 2003. Targeted gene expression using the Gal4/UAS system in
the silkworm Bombyx mori. Genetics 165, 13291340.
Inoue, S., Tanaka, K., Arisaka, F., Kimura, S., Ohtomo, K., Mizuno,
S., 2000. Silk broin of Bombyx mori is secreted, assembling a high
molecular mass elementary unit consisting of H-chain, L-chain,
and P25, with a 6:6:1 molar ratio. J. Biol. Chem. 275, 4051740528.
Inoue, S., Tanaka, K., Tanaka, H., Ohtomo, K., Kanda, T., Imamura,
M., Quan, G-X., Kojima, K., Yamashita, T., Nakajima, T., Taira,
H., Tamura, T., Mizuno, S., 2004. Assembly of the silk broin
elementary unit in endoplasmic reticulum and a role of L-chain for
protection of a-1,2-mannose residues in N-linked oligosaccharide
chains of brohexamerin/P25. Eur. J. Biochem. 271, 111.
Kamiie, K., Taira, H., Ooura, H., Kakuta, A., Matsumoto, S., Ejiri,
S., Katsumata, T., 1993. Nucleotide sequence of the cDNA
encoding silk gland elongation factor 1 alpha. Nucleic Acids Res
21, 742.
Kikuchi, Y., Mori, K., Suzuki, S., Yamaguchi, K., Mizuno, S., 1992.
Structure of the Bombyx mori broin light-chain-encoding gene:
upstream sequence elements common to the light and heavy chain.
Gene 110, 151158.
Mori, K., Tanaka, K., Kikuchi, Y., Waga, M., Waga, S., Mizuno, S.,
1995. Production of a chimeric broin light-chain polypeptide in a
broin secretion-decient naked pupa mutant of the silkworm
Bombyx mori. J. Mol. Biol. 251, 217228.
Reddy, P.S., Corley, R.B., 1998. Assembly, sorting, and exit of
oligomeric proteins from the endoplasmic reticulum. Bioessays 20,
546554.
Rigaut, G., Shevchenko, A., Rutz, B., Wilm, M., Mann, M., Seraphin,
B., 1999. A generic protein purication method for protein
complex characterization and proteome exploration. Nat. Biotech-
nol. 17, 10301032.
Shimura, K., 1983. Chemical composition and biosynthesis of silk
proteins. Experimentia 39, 455461.
Sitia, R., Neuberger, M., Alberini, C., Bet, P., Fra, A., Valetti, G.,
Williams, G., Milstein, C., 1990. Developmental regulation of IgM
secretion: the role of the carboxy-terminal cysteine. Cell 60,
781790.
Suzuki, M.G., Funaguma, S., Kanda, T., Tamura, T., Shimada, T.,
2003. Analysis of the biological functions of a doublesex
homologue in Bombyx mori. Dev. Genes 213, 345354.
Takei, F., Kimura, K., Mizuno, S., Yamamoto, T., Shimura, K.,
1984a. Genetic analysis of the Nd-s mutation in the silkworm. Jpn.
J. Genet. 59, 307313.
Takei, F., Oyama, F., Kimura, K., Hyodo, A., Mizuno, S., Shimura,
K., 1984b. Reduced level of secretion and absence of subunit
combination for the broin synthesized by a mutant silkworm,
Nd(2). J. Cell. Biol. 99, 20052010.
Takei, F., Kikuchi, Y., Kikuchi, A., Mizuno, S., Shimura, K., 1987.
Further evidence for importance of the subunit combination of silk
broin in its efcient secretion from the posterior silk gland cells. J.
Cell. Biol. 105, 175180.
Tamura, T., Thibert, C., Royer, C., Kanda, T., Abraham, E., Kamba,
M., Ko moto, N., Thomas, J.-L., Mauchamp, B., Chavancy, G.,
Shirk, P., Fraser, M., Prudhomme, J.-C., Couble, P., 2000. A
piggyBac-derived vector efciently promotes germ-line transforma-
tion in the silkworm Bombyx mori L. Nat. Biotechnol. 18,
8184.
Tanaka, K., Inoue, S., Mizuno, S., 1999a. Hydrophobic interaction of
P25, containing Asn-linked oligosaccharide chains, with the HL
complex of silk broin produced by Bombyx mori. Insect Biochem.
Mol. Biol. 29, 269276.
Tanaka, K., Kajiyama, N., Ishikura, K., Waga, S., Kikuchi, A.,
Ohtomo, K., Takagi, T., Mizuno, S., 1999b. Determination of the
site of disulde linkage between heavy and light chains of silk
broin produced by Bombyx mori. Biochim. Biophys. Acta 1432,
92103.
Thomas, J.L., Da Rocha, M., Besse, A., Mauchamp, B., Chavancy,
G., 2002. 3 P3-EGFP marker facilitates screening for transgenic
silkworm Bombyx mori L. from the embryonic stage onwards.
Insect Biochem. Mol. Biol. 32, 247253.
Tomita, M., Munetsuna, H., Sato, T., Adachi, T., Hino, R., Hayashi,
M., Shimizu, K., Nakamura, N., Tamura, T., Yoshizato, K., 2003.
Transgenic silkworms produce recombinant human type III
procollagen in cocoons. Nat. Biotechnol. 21, 5256.
Yamaguchi, K., Kikuchi, Y., Takagi, T., Kikuchi, A., Oyama, F.,
Shimura, K., Mizuno, S., 1989. Primary structure of the silk broin
light chain determined by cDNA sequencing and peptide analysis.
J. Mol. Biol. 210, 127139.
ARTICLE IN PRESS
S. Inoue et al. / Insect Biochemistry and Molecular Biology 35 (2005) 5159 59
Insect
Biochemistry
and
Molecular
Biology
Insect Biochemistry and Molecular Biology 35 (2005) 6172
Molecular cloning and functional characterization of a neuronal
choline transporter from Trichoplusia ni
Heather McLean
a
, LouAnn Verellen
b
, Stanley Caveney
a
, Cam Donly
b,
a
Department of Biology, The University of Western Ontario, London, Ont., Canada N6A 5B7
b
Southern Crop Protection and Food Research Centre, Agriculture and Agri-Food Canada, London, Ont., Canada N5V 4T3
Received 28 May 2004; received in revised form 20 October 2004; accepted 21 October 2004
Abstract
A cDNA encoding a high-afnity Na
+
-dependent choline transporter (TrnCHT) was isolated from the CNS of the cabbage
looper Trichoplusia ni using an RT-PCR-based approach. The deduced amino acid sequence of the CHT cDNA predicts a 594
amino acid protein of 64.74 kDa prior to glycosylation. TrnCHT has 80%, 79%, 76%, and 58% amino acid identity to putative
CHTs from Anopheles gambiae, Drosophila melanogaster and Apis mellifera, and a cloned CHT from Limulus polyphemus,
respectively. In situ hybridization of TrnCHT cRNA in whole-mount preparations of caterpillar CNS revealed that TrnCHT
mRNA is expressed by hundreds of presumably cholinergic neurons present in both the brain and cortex of all segmental ganglia.
Na
+
-dependent [
3
H]-choline uptake was induced in Sf9 cells in vitro following infection with a TrnCHT-expressing recombinant
baculovirus. Virally induced [
3
H]-choline uptake was found to approximately equal the endogenous rate of choline uptake in insect
cells, seen either after infection with a control virus or in TrnCHT-infected cells exposed to [
3
H]-choline in the absence of Na
+
. The
Na
+
-dependent component of [
3
H]-choline uptake by TrnCHT-infected cells was saturable with a K
m
for choline transport of
8.4 mM. Several compounds reported to be potent blockers of [
3
H]-choline uptake by cloned vertebrate choline transporters proved
to be relatively weak inhibitors of choline uptake by Sf9 cells expressing TrnCHT. Hemicholinium-3 (K
i
= 4:1 mM) and two
oxoquinuclidium analogues of choline, quireston-A (K
i
- 10 mM) and quireston (K
i
- 100 mM) inhibited 50% of control uptake
only at micromolar concentrations. The endogenous low-afnity Na
+
-independent uptake of [
3
H]-choline was also inhibited by
high micromolar concentrations of hemicholinium-3.
r 2004 Elsevier Ltd. All rights reserved.
Keywords: Choline; Transporter; Trichoplusia; Cloning; CNS; Neurons
1. Introduction
Choline has two critical roles in the functioning of the
nervous system, it serves as a precursor to the
neurotransmitter acetylcholine (ACh) and also as a
component of membrane phospholipids. Since choline is
not synthesized in neurons, it must be taken up by the
cells of the nervous system. Diffusion of such a charged
hydrophilic cation across membranes is inefcient,
necessitating the presence of transport mechanisms to
supply cellular needs. Two primary mechanisms exist
(Lockman and Allen, 2002). One consists of an
ubiquitous low-afnity and Na
+
-independent uptake
system that supplies choline for the synthesis of
phosphatidylcholine, an important membrane phospho-
lipid. Consequently, cells involved in lipid synthesis
have concentrative Na
+
-independent organic cationic
amine transporters (OCTs) with low-afnity in the
micromolar range for choline as a transport substrate
(Sinclair et al., 2000). The second consists of a high-
afnity, Na
+
-dependent system restricted to the pre-
synaptic nerve terminals of cholinergic neurons. Its role
is to supply choline for ACh synthesis by choline
acetyltransferase. This ACh is then pumped into
synaptic vesicles in cholinergic nerve terminals by a
ARTICLE IN PRESS
www.elsevier.com/locate/ibmb
0965-1748/$ - see front matter r 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ibmb.2004.10.005

Corresponding author. Tel.:+1 519 457 1470; fax: +1 519 457 3997.
E-mail address: donlyc@agr.gc.ca (C. Donly).
vesicular ACh transporter. The subsequent release of
ACh into the synaptic space, triggered by an action
potential, allows this neurotransmitter to interact with
ACh receptors in the post- and pre-synaptic nerve
membranes. Following its release, acetylcholine is
rapidly degraded into choline and acetate by the enzyme
acetylcholinesterase. The choline and acetate produced
are then taken up and recycled by cholinergic neurons as
precursors in new ACh synthesis.
ACh is the dominant excitatory neurotransmitter in
the insect CNS (Pitman, 1971; Gerschenfeld, 1973;
Callec, 1985). Cholinergic interneurons are found
throughout the insect CNS (reviewed in Sattelle, 1985;
Sattelle and Breer, 1990), as are neurons containing
ACh-membrane receptors (Matsuda et al., 2001). In
addition, many sensory neurons in the insect CNS are
cholinergic. Cholinergic chemoreceptors release ACh
from their synaptic terminals in the antennal lobe of the
brain (Waldrop and Hildebrand, 1989; Vickers et al.,
1998), as do cholinergic mechanoreceptor synapses in
the thoracic and abdominal ganglia (Callec, 1985;
Sattelle et al., 1985).
Sodium-dependent high-afnity choline uptake is a
more or less unique constituent of cholinergic synapses,
and is the rate limiting step in the synthesis of ACh in
the insect (Knipper and Breer, 1986; Breer and Knipper,
1990) and mammalian CNS (Kuhar and Murrin, 1978;
Jope, 1979). Synaptic choline transport in the insect
CNS was rst studied by Breer (1982) and Bermudez et
al. (1985). A high-afnity choline transport protein was
successfully isolated from locust synaptosomes and
puried with the use of monoclonal antibodies (Knipper
et al., 1989a, 1991). After reconstitution into liposomes,
this transporter induced a Na
+
-dependent accumula-
tion of choline that was sensitive to hemicholinium-3
(Knipper et al., 1989b), a hallmark feature of mamma-
lian high-afnity uptake systems. Although the authors
produced choline transporter (CHT)-specic antibodies,
the antibodies were not utilized to isolate a CHT cDNA
from the locust.
The cloning of an authentic CHT cDNA ultimately
proved to be more challenging than rst anticipated.
Initial attempts to clone CHTs using homology-based
PCR were founded on the premise that it was a member
of the Na
+
/Cl

-dependent family of neurotransmitter


transporters. This approach led to the isolation of rat
CHOT-1 (Mayser et al., 1992), a choline transporting
protein that was later found to function primarily as a
non-neuronal creatine transporter (Gonzalez and Uhl,
1994). Likewise, expression cloning efforts were also
ineffective with CHT due to high background and low
uptake levels. The rst functionally identiable CHT
was cloned from the nematode C. elegans with a strategy
utilizing bioinformatics as a means to select a manage-
able cohort of candidate sequences that could be
screened by functional expression in Xenopus oocytes
(Okuda et al., 2000). A single choline transporter was
thereby identied from all the potential Na
+
-dependent
transporters deduced from the worm genome. Subse-
quently, orthologues of this CHT sequence were cloned
and characterized in various other model organisms.
Using the same homology-based cloning strategy, we
have been able to clone a high-afnity choline trans-
porter cDNA from a pest insect species, the cabbage
looper, and express it in vitro to characterize the
transport properties of the encoded protein. This
research is motivated by the potential of neurotrans-
mitter transporters as novel pesticide targets for insect
control (Caveney and Donly, 2002).
2. Materials and methods
2.1. RNA isolation
Total RNA was isolated from different tissues
dissected from late instar T. ni larvae (epidermis, fat
body, brain, nerve cord, gonad, Malpighian tubules,
midgut, hindgut, rectum, silk gland and head) and adult
moths (optic lobes, ight muscle) using TRIzol Reagent
(Invitrogen, Burlington, ON). Total RNA isolated from
optic lobes was separated from the eye pigments by
passage through a Chroma Spin-100, DEPC-H
2
O
column (CLONTECH Laboratories, Inc., Palo Alto,
CA). The RNA recovered from the column was
dissolved in diethylpyrocarbonate (DEPC)-treated
water.
Poly(A
+
) RNA for cDNA synthesis was isolated
from total head RNA using a Poly(A) Quik oligo(dT)
push column (Stratagene, La Jolla, CA). Contaminating
genomic DNA was removed from the puried RNA by
treatment with RQ1 RNase-free DNase (Promega,
Madison, WI) for 30 min at 37 1C.
2.2. Reverse transcriptase - polymerase chain reaction
(RT-PCR) cloning strategy
Two degenerate primer pairs were designed from
conserved regions of known choline transporter proteins
to amplify a fragment of the cDNA encoding T. ni CHT.
The two primer pairs were positioned to make use of a
nested strategy, with the outside primers encoding
amino acids FTMTATWV (5259 in Fig. 1) and
FCIFVGLW (187194 in Fig. 1), and the inside primers
encoding YINGTAEA (6370 in Fig. 1) and
VAYTDVVQ (178185 in Fig. 1). A CHT cDNA
fragment (368 bp) was amplied from rst-strand cDNA
synthesized (cDNA Synthesis Kit, Stratagene) using
poly(A
+
) RNA isolated from the heads of late instar
caterpillars. The PCR mix contained 0.2 mM dNTPs,
2.5 mM MgCl
2
, 2 pmol/ml degenerate primers, 2.5 U
Platinum Taq DNA Polymerase (Invitrogen). The PCR
ARTICLE IN PRESS
H. McLean et al. / Insect Biochemistry and Molecular Biology 35 (2005) 6172 62
conditions were 94 1C for 2 min, followed by 35 cycles of
94 1C for 45 s, 58 1C for 45 s and 72 1C for 1 min,
followed by one 7 min hold at 72 1C.
The remaining 5
/
cDNA sequence was obtained
through a nested rapid amplication of cDNA ends
(RACE)-PCR approach using double stranded cDNA
(cDNA synthesis kit, Stratagene) ligated to the Strata-
gene PBK-CMV cloning vector as the template. Two
upstream facing primers were designed from the cDNA
fragment derived above, and used with two vector
specic primers to amplify a fragment representing the
5
/
end of the cDNA. The remaining 3
/
sequence was
obtained using multiple approaches. The same RACE-
PCR approach as outlined above produced a partial 3
/
sequence. Further extension of the 3
/
region was
achieved using inverse PCR to walk through the CHT
gene in the T. ni genome. Inverse PCR was performed on
T. ni genomic DNA using standard methods (Sambrook
and Russell, 2001). A nal RACE-PCR step using the
First-Choice RLM-RACE Kit (Ambion, Austin, TX)
was used to amplify a fragment containing the C-
terminal end of the ORF. The kit was used according to
the manufacturers instructions for 3
/
RACE-PCR.
Oligonucleotide primers were designed at the ends of
the composite sequence deduced above (nt 243 to
222 and 1786 to 1806 in Fig. 1) and used to perform
PCR to synthesize a cDNA containing the complete
ORF. Three separate amplications were performed
using Herculase DNA polymerase (Stratagene) with
random hexamer-primed cDNA from caterpillar heads
synthesized using Superscript II reverse transcriptase
(Invitrogen) as template. The products were cloned in
ARTICLE IN PRESS
Fig. 1. Nucleotide and deduced amino acid sequence of the cDNA encoding a T. ni choline transporter (TrnCHT). The rst nucleotide of the
translational start is designated as position 1 of the nucleotide sequence (numbered on the right) and the encoded methionine is position 1 of the
amino acid sequence (numbered in brackets). The translational start is shown boxed in bold and an upstream in-frame stop is bolded and underlined.
Primer sites at the 5
/
and 3
/
ends used to amplify the full sequence are shown in bold. The amino acids making up 13 predicted TMDs are underlined.
The amino acid residues upon which degenerate primers for nested PCR of the initial TrnCHT cDNA fragment were based are shown in bold.
Amino acids representing possible N-linked glycosylation sites on the extracellular domains are circled. Potential sites on the cytoplasmic domains for
phosphorylation by casein kinase II are enclosed in triangles, while a single site for protein kinase C is boxed. The sequence of the CHT cDNA shown
here has been entered in the GenBank/EMBL database under the accession number AY629593.
H. McLean et al. / Insect Biochemistry and Molecular Biology 35 (2005) 6172 63
pGEM-T Easy (Promega) for sequencing. A clone
produced from each of the amplications was sequenced
on both strands using dideoxy chain termination
sequencing (Applied Biosystems, Foster City, CA) and
the sequences compared to detect potential errors
produced by amplication.
2.3. RT-PCR expression analysis
First-strand cDNA was synthesized from 3 mg of total
RNA isolated from epidermis, fat body, brain, nerve
cord, optic lobes, gonad (testes), ight muscle, Mal-
pighian tubules, midgut, hindgut, rectum and silk gland,
respectively, using Superscript II Reverse Transcriptase
in the presence of 150 ng random primers (Invitrogen).
One microliter of each cDNA synthesized was used as
template to PCR amplify an internal CHT cDNA
fragment (primers were nt 210231 and 13981417,
Fig. 1). This pair of primers was selected so it would
enclose a splice junction from an intron, thus ensuring
that any products amplied from contaminating geno-
mic DNA in the RNA sample will be of a size different
from those resulting from cDNA. An RNA only
control was used to monitor for any PCR fragments
amplied from genomic DNA. The PCR was performed
using Platinum Taq DNA Polymerase (2.5 U) (Invitro-
gen) in the presence of 0.2 mM dNTPs, 2 mM MgCl
2
,
0.3 pmol primers per 50 ml reaction for 35 cycles of
denaturation at 94 1C for 30 s, annealing at 65 1C for 30 s
and elongation at 72 1C for 78 s. The denaturation step
of the rst cycle was 3 min long and the elongation step
of the last cycle was 7 min. The samples (10 ml) were
separated on a 1.2% agarose gel and transferred to
positively charged membrane (Roche, Laval, QC).
TrnCHT product was detected by hybridization with a
digoxigenin (DIG)-labelled DNA fragment covering the
amplied region, and then visualized by chemilumines-
cent detection of an anti-DIG antibody (Roche).
As an internal control for cDNA synthesis and PCR
amplication, a ubiquitously expressed transcript, gly-
ceraldehyde 3-phosphate dehydrogenase (G3PDH), was
amplied from all the cDNA samples. The primers were
designed from untranslated regions of the cDNA
encoding G3PDH (upstream: 5
/
CTTGTTTCTACA-
TAAATTTATTCC 3
/
and downstream: 5
/
AACAA-
CATTTATCTCTACACTGCTA 3
/
). PCR amplication
of the G3PDH fragment was performed under the same
conditions as those used for the TrnCHT fragment,
except the annealing temperature was 58 1C. The PCR
products were separated on 1.2% agarose and detected
by ethidium bromide staining.
2.4. In situ hybridization
A fragment of the TrnCHT cDNA encompassing
most of the putative ORF (nt 2101418 in Fig. 1) was
cloned in pGEM-T Easy vector and linearized using the
restriction enzymes PspOM I or Spe I (New England
Biolabs, Mississauga, ON). In vitro transcription was
performed using the Riboprobe in vitro Transcription
System (Promega), following the manufacturers direc-
tions with DIG-labelled ribonucleotides to generate
sense and anti-sense RNA probes (plasmid cut with
PspOM I used with SP6 to make anti-sense probe; Spe I
with T7 to make sense probe). To reduce the average
size of the resulting cRNAs the probes were incubated
with an equal volume of 120 mM Na
2
CO
3
/80 mM
NaHCO
3
at 60 1C for 15 min. Unincorporated nucleo-
tides were removed by passage through a Chroma Spin-
100 column (CLONTECH Laboratories). Brains plus
ventral nerve cords were dissected intact from late instar
cabbage looper caterpillars under PBS (130 mM NaCl,
70 mM Na
2
HPO
4
, 3 mM NaH
2
PO
4
, pH 7.4). Tissues
were xed in 4% paraformaldehyde in PBS for 23 h
and stored at 4 1C in PBST (0.3% Triton X-100 in PBS)
until use. In situ hybridizations were performed
essentially as described by Malutan et al. (2002) using
200400 ng/ml of labelled RNA for both sense and anti-
sense probes. DIG-labelled probe was detected
using alkaline phosphatase conjugated anti-digoxigenin
antibody (Roche) and visualized by incubation with
5-bromo-4-chloro-3-indolyl phosphate (BCIP)/ nitro-
blue tetrazolium (NBT) and 10 mM levamisol for
216 h. Mounted tissues were analyzed by bright-eld
microscopy.
2.5. Transient expression of TrnCHT in Sf9 cells
The TrnCHT-encoding fragment (243 to 1806 in
Fig. 1) was isolated from pGEM-T Easy, ligated into the
Bac-to-Bac transfer vector pFastBac 1 (Invitrogen) and
then transposed to bacmid as directed by the supplier.
The Sf9 (Spodoptera frugiperda) insect cell line used in
this study was maintained in monolayer culture at 27 1C
in Sf-900 II serum-free medium (Invitrogen) containing
20 mg/ml gentamicin (Invitrogen). Sf9 cells were trans-
fected with TrnCHT-recombinant bacmid using Cell-
Fectin reagent (Invitrogen). Medium containing
recombinant baculovirus was harvested 34 days after
cell transfection and the virus amplied in T25 asks of
Sf9 cells at 50% conuency. The viral titers in tertiary
amplications were estimated by plaque assay.
2.6. Transport assays
Assays were performed essentially as described by
Malutan et al. (2002). Sf9 cells were infected with
recombinant baculovirus at a multiplicity of infection
(MOI) of 0.1 and then incubated at 27 1C for 72 h
post-infection. The cells in each well were rst washed
for 1 h with Na
+
-containing saline (22.4 mM
MgSO
4
7H
2
O, 92.7 mM NaCl, 48.0 mM NaH
2
PO
4
,
ARTICLE IN PRESS
H. McLean et al. / Insect Biochemistry and Molecular Biology 35 (2005) 6172 64
20 mM K
+
gluconate, pH to 7.0 with KOH). The cells
were then incubated in 500 ml saline containing 5 ml of
1 mCi/ml [
3
H]-choline (specic activity 86 Ci/mmol;
NEN Life Science Products, Boston, MA) added to
give a nal concentration of 0.1 mM [
3
H]-choline. The
incubation period was set at 5 min, since the rate of [
3
H]-
choline uptake into infected Sf9 cells was found to be
linear for at least 6 min (data not shown). Choline
uptake was terminated by washing the cells several
times in Na
+
- free saline (11.2 mM MgCl
2
6H
2
O,
11.2 mM MgSO
4
7H
2
O, 7.3 mM KH
2
PO
4
, 55 mM
KCl, 53.5 mM N-methyl-D-glucamine
+
, 53.5 mM HCl,
76.8 mM sucrose, pH to 7.0 with KOH). The wells were
then air dried and the radiolabel accumulated by the
cells extracted in 500 ml 70% ethanol for 20 min. An
aliquot of the extract was added to Ready Safe
scintillation uid (Beckman Coulter, Fullerton, CA)
and counted in a Beckman LS-6500 scintillation
counter. The afnity of TrnCHT for choline was
determined by measuring [
3
H]-choline accumulation at
concentrations from 0.1 to 70 mM choline. [
3
H]-choline
was supplemented with unlabelled choline to give these
choline concentrations. The afnity of TrnCHT for
choline (K
m
choline
) and maximum rate of choline uptake
(V
max
) by virally infected cells were estimated through
EadieHofstee transformation of the [
3
H]-choline up-
take data. Cells incubated in Na
+
-free saline or infected
with a b-glucuronidase (GUS) recombinant virus were
used to correct the experimental data for endogenous
(background) accumulation of [
3
H]-choline by Sf9 cells.
The Na
+
dependence of choline uptake was assessed
using a series of salines in which equimolar NMG
+
replaced Na
+
. The ability of hemicholinium-3
(HC3, Sigma, St. Louis, MO), a competitive substrate
of Na
+
-dependent choline uptake by mammalian
CHTs, was tested over the concentration range
0.001200 mM. The concentration at which HC3 re-
duced the uptake of choline by 50% (IC
50
value) was
determined by non-linear regression analysis of Hill
plots (inhibitor concentration plotted against I=I
max
I
on a double logarithmic plot, with I = inhibition and
I
max
= maximum inhibition) using Microsoft Excel
2000. The TrnCHT inhibition constant (K
i
) for HC3
was derived from the IC
50
value using the Cheng and
Prusoff (1973) equation, IC50 = K
i
(1+[S]/K
m
choline
),
where [S] is the [
3
H]-choline concentration used experi-
mentally. The HC3 inhibition data given are the mean
values7SD obtained from at least four sets of Sf9 cells
infected with TrnCHT recombinant virus. Inhibition of
choline uptake by quireston and quireston A (Latoxan,
Valence, France), two reportedly potent blockers of
Na
+
-dependent choline uptake in mammals, was tested
over a concentration range of 0.1100 mM. Quireston
inhibition data were expressed as a percentage of control
uptake by CHT-infected Sf9 cells incubated in saline
lacking inhibitor. Cells were not pre-treated with either
HC3 or quireston inhibitors prior to exposure to [
3
H]-
choline.
3. Results
3.1. Cloning of a cDNA encoding a T. ni choline
transporter
A 368 bp cDNA fragment which is highly similar to
known high-afnity choline transporters was PCR
amplied from caterpillar head cDNA using degenerate
primers based on strongly conserved features of several
characterized choline transporter proteins. From this
fragment, unique primers were synthesized to be used in
combination with vector primers in nested RACE-PCR
to amplify the ends of the cDNA. The template for
RACE-PCR was double-stranded caterpillar head
cDNA (Stratagene) ligated to plasmid. The 5
/
RACE
strategy yielded an 900 bp product containing an in-
frame stop codon near the 5
/
terminus. When this
approach was used to amplify the remaining 3
/
cDNA
sequence the resulting product was incomplete, termi-
nating in an unspliced intron. To obtain the remaining
cDNA, the sequence was extended by walking through
genomic DNA using inverse PCR. Homology with
known choline transporters was used to identify exons
in the gene sequence as well as the probable stop codon.
Primers were then synthesized to allow the complete
ORF to be PCR amplied from rst-strand cDNA in a
single 2049 bp product.
The complete TrnCHT cDNA sequence obtained
contains a large ORF of 1785 bp encoding a potential
protein of 594 amino acids with a predicted molecular
weight of 64.740 kDa (DNAstar Inc., Madison, WI).
The start site for this ORF utilizes the rst ATG
downstream of an in-frame stop codon located at
position 30 in Fig. 1. The amino acid sequence
deduced using this ATG start codon is highly conserved
with other choline transporters, suggesting it is the
authentic start site. The amino acid sequence of
TrnCHT shows 7680% identity with putative insect
choline transporters obtained from genome databases
(GenBank) and 5060% identity with choline transpor-
ters cloned from non-insects (Caenorhabditis elegans,
Torpedo marmorata, Limulus polyphemus and mamma-
lian sequences). Phylogenetic comparison of these
sequences distinguishes a discrete clade of choline
transporters made up of all the insect CHTs currently
sequenced (Fig. 2).
The high-afnity choline transporters are members of
the Solute:Sodium Symporter (SSS) family of transpor-
ters (TC 2.A.21) according to the classication system of
Saier (2000). The SSS family corresponds to the
mammalian solute carrier group SLC5. Transporters
in this group catalyze solute: Na+ symport, taking up a
ARTICLE IN PRESS
H. McLean et al. / Insect Biochemistry and Molecular Biology 35 (2005) 6172 65
diverse range of substrate solutes including sugars,
amino acids, organic cations such as choline, nucleo-
sides, inositols, vitamins, urea or anions. Hydrophobi-
city analysis of the deduced amino acid sequence of
TrnCHT using TMpred (http://www.ch.embnet.org/
software/TMPRED_form.html) suggests the presence
of 13 transmembrane domains (TMDs). The presence of
a core of 13 TMDs is characteristic of all SSS
transporters, with an additional one or two TMDs
present in specic groups. The high-afnity choline
transporters all have 13 predicted TMDs (Okuda and
Haga, 2003). This membrane topology places the N-
terminus of the protein outside the cell. Two N-linked
glycosylation sites identied by scanning against the
PROSITE database (http://ca.expasy.org/cgi-bin/scan-
prosite) were located on extracellular loops EL 1 and EL
4. These positions are identical to the two sites found on
the human CHT 1 protein sequence (Okuda and Haga,
2003) and two of the three sites seen in the Limulus
choline transporter (Wang et al., 2001). Several potential
serine/threonine phosphorylation sites on the cytoplas-
mic domains of TrnCHT were also identied. A single
protein kinase C site is located on cytoplasmic loop CL
4 and casein kinase II sites located on CL 1, CL 2 and
CL 5 (Fig. 1).
3.2. Expression of TrnCHT mRNA
To evaluate the distribution of TrnCHT mRNA
expression in the insect, RT-PCR was used on total
RNA isolated from a variety of tissues (Fig. 3). Primers
were selected to yield an approximately 1200 bp product
which was visualized by hybridization with a DIG-
labelled probe followed by chemiluminescent detection.
Product was detected in the samples from the nervous
system (brain, nerve cord, and optic lobes) indicating
expression of TrnCHT mRNA is specic to the nervous
system. The primers were also selected to encompass an
intron in the TrnCHT gene such that products resulting
from amplication of contaminating genomic DNA
could be discerned. No such products were observed in
any sample or in the RNA only control. The detection
of a ubiquitously expressed gene (G3PDH) was used as
an internal control for RNA integrity and cDNA
synthesis. A 950 bp product representing the G3PDH
cDNA was detected in the samples from all of the tissues
(Fig. 3).
Cellular localization of the TrnCHT mRNA was
accomplished using whole-mount in situ hybridization.
DIG-labelled cRNA probes (sense and anti-sense) were
synthesized in vitro from a template containing 1208 bp
of the TrnCHT ORF and hybridized to central nervous
systems excised intact from last instar caterpillars.
Staining was observed with the anti-sense cRNA probes
ARTICLE IN PRESS
CaeCHT
TomCHT
hCHT
rCHT
1000
1000
LipCHT
813
1000
695
968
ApmCHT
DrmCHT
AngCHT
TrnCHT
Fig. 2. Phylogenetic analysis of Na
+
-dependent choline transporters.
The amino acid sequences of various known choline transporters and
putative sequences deduced from completed genomes were aligned
using ClustalX (1.81) and an unrooted tree calculated using the
neighbor joining method. Condence values for the derived tree were
determined by bootstrapping the dataset using 1000 replicates and a
generator seed value of 333 (ClustalX). The alignment was displayed
using Treeview (1.6.5). The accession numbers of the sequences taken
from the GenBank database are: human (hCHT; NP_068587), rat
(rCHT; NP_445973), horseshoe crab (LipCHT; AAG41055), nema-
tode (CaeCHT; NP_502539), fruit y (DrmCHT; NP_650743),
mosquito (AngCHT; EAA07459), electric ray (TomCHT;
CAD12727), and cabbage looper (TrnCHT; AY629593). The sequence
for the honey bee, ApmCHT, was assembled from release 1.1 of the
Apis mellifera genome available from the Human Genome Sequencing
Centre (Baylor College of Medicine).
Fig. 3. Tissue-specic expression of TrnCHT mRNA revealed by RT-
PCR analysis. Upper panel: Total RNA (3 ug) reverse-transcribed to
cDNA was used as target for PCR with TrnCHT specic primers to
identify the presence of the mRNA in various tissues. PCR products
were separated on agarose gels and transferred to membrane before
hybridization with DIG-labelled TrnCHT DNA. Products of the
expected size of 1208 bp were detected in each of the lanes representing
the nervous system (brain, nerve cord, and optic lobes). Tissue lanes
are as labelled, RNA only represents a sample of RNA used directly
as template in the PCR, and H
2
O represents a water control used in the
PCR. Lower panel: RT-PCR internal control detecting the ubiquitous
G3PDH transcript in the above samples. Two G3PDH-specic primers
were used to amplify a 950 bp band in all the cDNAs to verify the
integrity of the samples.
H. McLean et al. / Insect Biochemistry and Molecular Biology 35 (2005) 6172 66
in the brain and all ganglia of the ventral nerve cord,
while sense probe controls showed no specic staining
above background. Fig. 4 shows the caterpillar sub-
oesophageal ganglion (SOG) probed with TrnCHT
cRNA. Panel A of the gure is focussed on the dorsal
and panel B on the ventral aspect of the SOG hybridized
with the anti-sense probe, while panel C shows the
background staining in the SOG resulting from the
sense control probe. The TrnCHT-positive cell bodies
detected with the anti-sense probe were found to lie
within the cortical layer of the ganglion. Approximately
280 stained cells were present in the brain, 220 cells in
the SOG, 140 in each of the thoracic ganglia T1T3, 85
in the rst abdominal ganglion (A1), about 70 in each of
the other non-fused abdominal ganglia (A2A6), and
100 in the fused terminal ganglion (A7/A8). Little
evidence of strong leftright symmetry in the distribu-
tion of the TrnCHT RNA-positive cell bodies in the
ganglia was seen. Whereas clusters of cells were seen
reliably in certain quadrants of the ganglionic cortex, no
obvious sets of paired neurons were consistently seen in
the ganglia, although a few pairs of centrally located
large neurons were seen in the ventral cortex of most
abdominal ganglia, and were particularly evident in
ganglia A6 and A7/8. The frontal ganglion contained
three TrnCHT-positive cells. No expression was de-
tected in the corpora allata or corpora cardiaca. The
total number of cells expressing TrnCHT RNA in the
caterpillar brain plus ventral nerve cord is estimated to
be about 1450.
3.3. Functional characterization of TrnCHT
The total amount of [
3
H]-choline taken up by Sf9 cells
infected with TrnCHT-recombinant virus in the pre-
sence of Na
+
is concentration dependent and suggested
saturation kinetics (Fig. 5, solid circles). The rate of
choline uptake in such cells was at least twice that of
control cells infected with GUS-recombinant virus in the
presence of Na
+
(Fig. 5, solid squares). The rate of [
3
H]-
choline uptake by TrnCHT-expressing cells in the
absence of Na
+
(Fig. 5, open circles) was similar to
that seen in the control cells infected with GUS-
recombinant virus. The signicant Na
+
-independent
component of total [
3
H]-choline uptake indicated the
presence of an endogenous choline transport mechanism
reported to be present in many animal cells, including
insect cells such as the lepidopteran Sf9 and Sf21 cell
lines (pers. obs.). Consequently, the TrnCHT-specic
component of [
3
H]-choline uptake was analyzed (Fig. 5,
inset) by subtracting the Na
+
-independent component
from total uptake. The Na
+
-dependent afnity of
TrnCHT for choline was calculated to have a K
m
choline
of 8.471.5 mM over a V
max
range of 1338 pmol/well/
min (n = 6): The Na
+
requirement for TrnCHT activity
was analyzed in more detail at 11 Na
+
concentrations
between 0 and 140 mM (data not shown). No signicant
Na
+
-dependent uptake of choline above the endogen-
ous background uptake was seen until the external
[Na
+
] was raised to above 20 mM. At 40 mM and higher
[Na
+
], the rate of choline uptake climbed linearly to
reach 2.3 times the background rate at 140 mM. The
Na
+
-dependent component of choline uptake did not
show signs of saturation over this Na
+
concentration
range.
The pharmacological activity of three structural
analogues of choline, hemicholinium-3 (HC3) and two
oxoquinuclidium compounds, quireston and quireston
A, were assessed for their ability to inhibit the uptake of
[
3
H]-choline by TrnCHT. These quaternary nitrogen
compounds are routinely used to identify Na
+
-depen-
dent uptake of choline in mammalian neural tissues. At
ARTICLE IN PRESS
Fig. 4. CHT expression in the caterpillar suboesophageal ganglion. In situ hybridization with DIG-labelled CHT anti-sense cRNA identied the cell
bodies of many CHT-expressing neurons scattered throughout the cortex of the SOG (panels A and B). Panel A is focused in the plane of the dorsal
and dorsolateral region of the ganglion. CHT-positive neurons are seen as darkly stained areas of cytoplasm surrounding unstained nuclei. Few
neurons expressing CHT are found in the posterior region of the dorsal cortex (bottom), where ventrally located and out-of-focus CHT-positive
neurons can be seen. Panel B is in the ventral and ventrolateral plane of the ganglion, and brings these cells into focus. Panel C shows the lack of
specic staining produced with the sense cRNA control probe.
H. McLean et al. / Insect Biochemistry and Molecular Biology 35 (2005) 6172 67
sub-micromolar concentrations, HC3 failed to block
Na
+
-dependent [
3
H]-choline uptake by Sf9 cells expres-
sing TrnCHT. HC3 inhibited Na
+
-dependent choline
uptake at concentrations above 1 mM (Fig. 6). At these
concentrations the endogenous component of choline
uptake was also affected (Fig. 6). The inhibition of the
Na
+
-dependent (TrnCHT-specic) component of cho-
line uptake is the difference between the inhibition of
total choline uptake (Fig. 6, solid circles) and of the
Na
+
-independent endogenous uptake (Fig. 6, open
circles). After subtracting the endogenous component
from the total suppression of choline uptake by HC3, a
Hill plot (Fig. 6, inset) was used to estimate the IC
50
for
HC3 inhibition of the TrnCHT-dependent component
of choline uptake to be 4.4 mM. The corresponding K
i
for HC3 inhibition of the TrnCHT-dependent uptake
was calculated to be 4.1 mM. The choline analogues
quireston and quireston A were also tested for their
ability to disrupt [
3
H]-choline uptake by TrnCHT
(Fig. 7). Although there was a 2030% inhibition of
choline uptake with both compounds at concentrations
as low as 0.1 mM, quireston A inhibited 50% of the
TrnCHT-specic uptake at a concentration of 10 mM.
Quireston was less potent, needing a concentration of
100 mM to inhibit 50% of choline uptake.
4. Discussion
The TrnCHT cDNA predicts a 65 kDa protein with
considerable homology to high-afnity and Na
+
-
dependent choline transporters identied in mammals
ARTICLE IN PRESS
0
10
20
30
40
50
60
70
80
0 20 40 60
[choline] M
p
m
o
l

[
3
H
]
c
h
o
l
i
n
e

u
p
t
a
k
e
/
w
e
l
l
/
m
i
n
y = -8.4x + 20.5
0
5
10
15
20
25
30
0 1 2 3
V/S
V
Fig. 5. Kinetics of [
3
H]-choline uptake by insect cells infected with
TrnCHT-recombinant baculovirus in vitro. The rate of uptake of [
3
H]-
choline by Sf9 cells infected with TrnCHT virus is greater in the
presence of 140 mM Na
+
(+) than in the absence of Na
+
(J). The
relatively high rate of background (i.e. Na
+
-independent endogenous)
uptake of [
3
H]-choline by cells after baculoviral infection is also seen
after infection with a control (GUS)-recombinant virus in the presence
of 140 mM Na
+
(). Paired t-test analysis was used to assess the
statistical signicance of differences in choline uptake under the three
different conditions tested (Graphpad Prism software). TrnCHT-
expressing cells took up signicantly more choline in the presence of
Na
+
than in its absence (Po0:0001); as did TrnCHT-expressing cells
compared to GUS-expressing cells in the presence of Na
+
(Po0:0003):
There was a small but signicant difference in choline uptake by
TrnCHT-expressing cells in the absence of Na
+
compared to GUS-
expressing cells in the presence of Na
+
(Po0:0348): The kinetics of
Na
+
-dependent uptake became evident after adjusting for the back-
ground uptake of choline by the cells (see inset). The data are the
means and standard deviations of at least four experiments for each of
the three treatments. Inset: EadieHofstee plot of the Na
+
-dependent
component of [
3
H]-choline uptake after CHT-recombinant viral
infection of Sf9 cells. The average afnity (K
m
) of TrnCHT for choline
was 8:4 1:5 mM (n = 6 experiments) when expressed in these cells.
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.001 0.01 0.1 1 10 100 1000
[Hemicholinium-3] M
p
m
o
l

[
3
H
]
c
h
o
l
i
n
e

u
p
t
a
k
e
/
w
e
l
l
/
m
i
n
y = 0.23x
1.00
0.01
0.1
1
10
100
0.01 0.1 1 10 100
[HC-3]
I
/
I
m
a
x
-
I
Fig. 6. Hemicholinium-3 inhibition of [
3
H]-choline uptake by TrnCHT
in vitro. HC3 proved to be a moderately potent blocker both of the
Na
+
/TrnCHT- dependent and of the Na
+
-independent endogenous
component of choline uptake by Sf9 cells in vitro. A concentration of
0.1 mM choline was used in this experiment. The data shown are the
means and standard deviations of four or more experiments showing
HC3 inhibition of total choline uptake by the cells in the presence (+)
and absence (J) of Na
+
in the incubation medium. HC3 inhibition of
total choline uptake became evident at inhibitor concentrations above
1 mM, while that of the Na
+
-independent endogenous uptake at
concentrations above 10 mM. Inset: Hill plot used to calculate the
concentration of HC3 required to inhibit 50% of the TrnCHT-specic
component of [
3
H]-choline uptake seen in Sf9 cells after infection with
TrnCHT virus. The data shown are corrected for the inuence of HC3
on the endogenous component. The IC
50
for HC3 is determined from
the intercept of the slope at I=I
max
1 = 1 in the Hill plot. The
exponent (1.0) for the curve tted to the data suggests that TrnCHT
has a single binding site for HC3.
H. McLean et al. / Insect Biochemistry and Molecular Biology 35 (2005) 6172 68
and nematodes. The TrnCHT protein has potential sites
for N-linked glycosylation and phosphorylation similar
in position to those found in other authenticated CHTs.
Sequence comparisons with additional CHT-like se-
quences reveal a subfamily of candidate choline
transporters (Fig. 2) predicted to contain 13 TMDs in
the protein structure. This choline transporter subfamily
belongs to a broader family of Solute:Sodium Sympor-
ter (SSS) membrane transporters (Saier, 2000) that
drives the uptake of many organic solutes, including
sugars, amino acids, nucleosides and vitamins. The
choline transporters are unrelated to neurotransmitter
transporters, having apparently stemmed from an
ancient transport system primarily responsible for
nutrient uptake (Okuda and Haga, 2003).
In insects, Na
+
-dependent choline uptake is restricted
to cholinergic neurons and has not been reported
outside the CNS (Caveney and Donly, 2002). The
antennal lobe, the primary site of olfactory processing in
the insect brain, is particularly rich in cholinergic
activity (e.g. Manduca, Waldrop and Hildebrand,
1989; Homberg et al., 1995). Efferent bres from
antennal sensory neurons terminate in the glomerular
neuropile of the olfactory lobe, where they make
cholinergic synaptic contact with intrinsic and projec-
tion olfactory interneurons (Vickers et al., 1998). An
antibody raised against a neuronal choline transporter
puried from locust synaptosomes strongly labelled the
glomerular neuropile of the locust antennal lobe
(Knipper et al., 1989c). All three regions of the optic
tract (lamina, medulla and lobula) in the locust brain
also stained with this antibody, implying that some
sensory neurons and/or interneurons in these sensory
processing regions are cholinergic. Many identiable
cholinergic sensory neurons and interneurons are known
from the thoracic and abdominal ganglia. In the
cockroach Periplaneta americana, for example, afferent
bres from mechanoreceptors in the cerci that make
synaptic contact with the dendrites of giant interneurons
in the terminal abdominal ganglion are cholinergic
(Sattelle et al., 1985), as are the dorsal unpaired median
(DUM) interneurons that serially repeat themselves in
the abdominal ganglia (Buckingham et al., 1997).
In the caterpillar, TrnCHT is expressed in the brain
and segmental ganglia of the ventral nerve cord, and by
a few neurons in the frontal ganglion. Because CHT is a
marker protein of cholinergic neurons, its cellular
distribution records the number of cholinergic neurons
present in the nerve cord. In situ hybridization revealed
expression of TrnCHT mRNA in neurons scattered
throughout the cortical layer of the caterpillar brain and
segmental ganglia. This pattern of RNA expression
contrasts with that of neurons expressing mRNAs for
monoamine transporters (MATs) such as octopamine
(OAT, Malutan et al., 2002) dopamine (DAT, Po rzgen
et al., 2001; Gallant et al., 2003) and serotonin (SERT;
Demchyshyn et al., 1994). Neurons expressing MAT
RNAs in the insect CNS are few in number and occur in
a serially repeated fashion in the segmental ganglia. In
contrast, we estimate that the cortex of the cabbage
looper caterpillar brain and ventral nerve cord has
approximately 1450 TrnCHT-mRNA positive neurons.
TrnCHT-positive cell bodies are missing from the cortex
only where it is traversed by segmental nerves or
trachea. CHT mRNA expression was not detected in
the ganglionic neuropile or connectives. The TrnCHT
protein, on the other hand, is most likely localized to
cholinergic axons and synapses in the ganglionic
neuropile. Bermudez et al. (1985) showed by autoradio-
graphy that the majority of the axons of cockroach
neurons cultured in vitro were labelled after incubation
in [
3
H]-choline, but their somata remained unlabelled.
The antibody raised by Knipper et al. (1989c) against
the locust choline transporter failed to stain the cortex
and perineurial sheath of the mesothoracic ganglion or
the ganglionic connectives. It appears that CHT mRNA
is synthesized in the cell bodies of cholinergic neurons
whereas the bulk of the CHT protein accumulates at
cholinergic synapses in the ganglionic neuropile. The
presence of extrajunctional choline transporters and/or
glial choline transporters in insect ganglia has not been
demonstrated.
No evidence was found for the presence of TrnCHT
mRNA in peripheral tissues such as midgut and rectum
(Fig. 3), or in the neurosecretory/endocrine tissues such
as the corpus allatum and corpus cardiacum. Indeed,
while these organs are known to contain cholinergic
ARTICLE IN PRESS
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0.1 1 10 100 Na+ -free
drug concentration, M
p
e
r
c
e
n
t

c
o
n
t
r
o
l

u
p
t
a
k
e
Quireston
Quireston A
Fig. 7. Quireston analogues are weak inhibitors of Na
+
-dependent
uptake of [
3
H]-choline by TrnCHT-expressing Sf9 cells. Quireston A
was found to block [
3
H]-choline uptake by TrnCHT at concentrations
greater than 10 mM. Quireston had no convincing effect on choline
uptake below 100 mM. The endogenous level of [
3
H]-choline uptake
(cells exposed to choline in the absence of Na
+
and quireston
analogue) is shown on the right. The 2030% inhibition of choline
uptake at low concentrations of both analogues (left) is unexplained.
H. McLean et al. / Insect Biochemistry and Molecular Biology 35 (2005) 6172 69
nerve endings, the low external Na
+
/high external K
+
conditions present may preclude the activity of TrnCHT
protein, even were it to be expressed at peripheral nerve
terminals. We suspect, however, that in contrast to the
release of ACh by cholinergic terminals in the CNS, the
release of ACh by peripheral cholinergic terminals is not
followed by high afnity Na
+
-dependent re-uptake of
choline by the peripheral nerve endings. In binding to
metabotropic (muscarinic) ACh receptors on target
tissues outside the CNS, ACh released peripherally
appears to act more as a neurohormone/neurosecretion
than as a neurotransmitter in insects.
Functional studies of cells expressing recombinant
CHT in vitro have generally been unsuccessful, because
of a high background rate of Na
+
-independent choline
uptake. Transfected COS-7 cells expressing human CHT
in vitro, for example, show only a modest increase over
the endogenous rate of [
3
H]-choline uptake (Apparsun-
daram et al., 2000). Infection of Sf9 cells with TrnCHT-
recombinant baculovirus, however, doubled the rate of
[
3
H]-choline accumulation, with the saturable Na
+
-
dependent component of [
3
H]-choline being sufciently
above the endogenous rate in intact cells to make a
kinetic analysis of TrnCHT possible. The afnity of
TrnCHT for choline (K
m
= 8:4 mM) is lower than that
reported for high-afnity choline uptake described in
synaptosomes isolated from the locust CNS
(K
m
= 0:98 mM; Breer, 1982) and cockroach nerve cells
in vitro (K
m
= 0:5 mM; Bermudez et al., 1985) but close
to the range reported for high-afnity mammalian
CHTs (K
m
= 125 mM) (Okuda et al., 2000; Okuda
and Haga, 2000, 2003; Apparsundaram et al., 2000).
The afnity of TrnCHT for choline is considerably
higher than that of Na
+
-independent choline transpor-
ters reported from the locust (K
m
= 25 mM; Breer, 1982)
and mammalian CNS (K
m
= 302100 mM; Lockman
and Allen, 2002).
The ability of hemicholinium-3, in a concentration
range of 0.0010.1 mM, to selectively and strongly
inhibit [
3
H]-choline uptake (Okuda et al., 2000; Lock-
man and Allen, 2002) is a characteristic feature of
choline transporters belonging to the SSS family. HC3
has been used to identify high-afnity Na
+
-dependent
choline uptake systems in both mammalian (Lockman
and Allen, 2002) and arthropod neurons (Breer, 1982;
Knipper et al., 1989b; Ivy et al., 2001). TrnCHT,
although clearly an insect CHT homologue in the SSS
family, is relatively insensitive to HC3. The HC3
inhibition constant for TrnCHT was estimated at
4.1 mM. Also, analogues of choline reported to be
selective inhibitors of choline uptake in the mammalian
cerebral cortex (IC
50
- 0:5 mM; Sterling et al., 1986) are
weak inhibitors of [
3
H]-choline uptake by TrnCHT. The
experimental conditions used here to test the inhibitory
ability of HC3 and quireston analogues involved
exposing the TrnCHT expressing cells to [
3
H]-choline
and inhibitor simultaneously for 5 min. We suspect that
the rapid binding kinetics of choline to its binding site
on the TrnCHT protein, relative to those of HC3 and
the other structural analogues tested, partially obscured
the stronger inhibition that longer term incubation may
have revealed, although this remains to be shown
experimentally. Choline uptake by cultured cockroach
neurons is also relatively insensitive to inhibition by
HC3 (Bermudez et al., 1985). The apparent weak
inhibition of some insect CHTs by HC3 conicts with
published data on the binding kinetics of [
3
H]HC3 to
insect synaptosomes. Synaptosomes prepared from
locust ganglia are reported to have two HC3 binding
sites with equilibrium dissociation constants (K
d
s) of 3
and 69 nM (Knipper et al., 1991). The HC3 binding sites
of mammalian high-afnity CHTs have a K
d
in the
2 nM range (Apparsundaram et al., 2000; Okuda and
Haga, 2000).
Replacing Na
+
in the incubation saline with N-
methylglucamine
+
suppressed choline uptake by
TrnCHT-expressing Sf9 cells to levels seen in cells
expressing a control (GUS) construct. A near-exclusive
requirement for extracellular Na
+
to drive uptake is
seen in neurotransmitter transporters cloned from the
insect CNS (Caveney and Donly, 2002). The uptake of
[
3
H]-choline by TrnCHT occurred at Na
+
levels above
20 mM and failed to saturate at levels as high as
140 mM, suggesting that the extra-axonal synaptic space
in caterpillars has a Na
+
concentration in excess of this.
Li
+
is a poor substitute for Na
+
in the cation
requirement of insect choline transporters (Breer, 1983;
Bermudez et al., 1985). The uptake of [
3
H]-choline by
locust synaptosomes peaks in about 150 mM NaCl,
corresponding to a K
m
for Na
+
of between 40 and
60 mM (Breer, 1982). The synaptosomal transporter
appears to require K
+
on the inside of the vesicles in
order to take up choline (Breer, 1982), but it is not
known whether this is due to K
+
countertransport
during choline/Na
+
co-transport (i.e., is an electrogenic
component of the choline transporter). Elevating
external [K
+
] lowers [
3
H]-choline uptake in intact cells,
suggesting that choline uptake may be membrane
potential sensitive (Bermudez et al., 1985). Choline
transport by mammalian synaptosomes also has a
requirement for the external anion Cl

(Simon and
Kuhar, 1976; Okuda et al., 2000). Br

substitutes well
for Cl

, but I

does not, in driving choline uptake by rat


CHT1 (Okuda et al., 2000). We did not test the anion
dependency of choline uptake by TrnCHT, because our
ndings with other so-called Na
+
/Cl

-dependent trans-
porters cloned from the cabbage looper suggest that
there is limited anion selectivity (Gao et al., 1999;
Malutan et al., 2002; Gallant et al., 2003). In the locust,
nevertheless, the anions phosphate, isothiocyanate,
sulfate and acetate were unable to substitute for external
Cl

in driving [
3
H]-choline uptake (Breer, 1983).
ARTICLE IN PRESS
H. McLean et al. / Insect Biochemistry and Molecular Biology 35 (2005) 6172 70
The work here represents the rst functional char-
acterization of a cDNA encoding a protein needed for
the high-afnity and Na
+
-dependent uptake of choline
by cholinergic neurons in insects. TrnCHT retrieves
choline from the synaptic space and makes it available
for the synthesis of the neurotransmitter ACh in the
caterpillar CNS. A standard rationale to justify research
into the membrane proteins responsible for neurotrans-
mitter transport in neurons is that the information
obtained may be exploited to design strategies to disrupt
transporter activity. Blocking the synaptic transporter
of the monoamine serotonin (SERT), for example,
would cause the synaptic levels of this neurotransmitter
to rise, leading to at least a short-term and selective
over-stimulation of serotonin receptors in the nervous
system. Blocking the uptake of choline by CHT-
expressing neurons would have the opposite effect, in
that it would lead to a suppression of acetylcholine
synthesis by cholinergic neurons and cause a drop in
ACh levels in the synaptic space separating cholinergic
terminals from their target cells. The behavioral
consequences of blocking insect CHTs and the ensuing
suppression of choline biosynthesis in cholinergic
neurons remain to be evaluated in the context of insect
control strategies.
Acknowledgements
These studies were supported by the Matching
Investment Initiative Program of Agriculture and
Agri-Food Canada and by the Natural Sciences and
Engineering Research Council of Canada.
References
Apparsundaram, S., Ferguson, S.M., George, A.L., Blakely, R.D.,
2000. Molecular cloning of a human, hemicholinium-3-sensitive
choline transporter. Biochem. Biophys. Res. Comm. 276,
862867.
Bermudez, I., Lees, G., Middleton, C., Botham, R., Beadle, D.J., 1985.
Choline uptake by cultured neurons from the central nervous
system of embryonic cockroaches. Insect Biochem. 15, 427434.
Breer, H., 1982. Uptake of [NMe
3
H]-choline by synaptosomes from
the central nervous system of Locusta migratoria. J. Neurobiol. 13,
107117.
Breer, H., 1983. Choline transport by synaptosomal membrane vesicles
isolated from insect nervous tissue. FEBS Lett. 153, 345348.
Breer, H., Knipper, M., 1990. Regulation of high afnity choline
uptake. J. Neurobiol. 21, 269275.
Buckingham, S.S., Lapied, B., Le Corronc, H., Grolleau, F., Sattelle,
D.B., 1997. Imidacloprid actions on insect neuronal acetylcholine
receptors. J. Exp. Biol. 200, 26852692.
Callec, J.J., 1985. Synaptic transmission in the central nervous system.
In: Kerkut, G.A., Gilbert, L.I. (Eds.), Comprehensive Insect
Physiology, Biochemistry and Pharmacology, vol. 5. Pergamon
Press, Oxford, pp. 141179.
Caveney, S., Donly, B.C., 2002. Neurotransmitter transporters in the
insect nervous system. Adv. Insect Physiol. 29, 55149.
Cheng, Y.-C., Prusoff, W.H., 1973. Relationship between the
inhibition constant (K
i
) and the concentration of the inhibitor
which causes 50 percent inhibition (I
50
) of an enzymatic reaction.
Biochem. Pharmacol. 22, 30993108.
Demchyshyn, L.L., Pristupa, Z.B., Sugamori, K.S., Barker, E.L.,
Blakely, R.D., Wolfgang, W.J., Forte, M.A., Niznik, H.B., 1994.
Cloning, expression, and localization of a chloride-facilitated,
cocaine-sensitive serotonin transporter from Drosophila melanoga-
ster. Proc. Natl. Acad. Sci. USA 91, 51585162.
Gallant, P., Malutan, T., McLean, H., Verellen, L., Caveney, S.,
Donly, C., 2003. Functionally distinct dopamine and octopamine
transporters in the CNS of the cabbage looper moth. Eur.
J. Biochem. 270, 664674.
Gao, X., Caveney, S., Donly, C., 1999. Molecular cloning and
functional comparison of a GABA transporter cloned from the
CNS of the cabbage looper Trichoplusia ni. Insect Biochem. Mol.
Biol. 29, 609623.
Gerschenfeld, H.M., 1973. Chemical transmission in invertebrate
central nervous systems and neuromuscular junctions. Physiol.
Rev. 53, 1119.
Gonzalez, A.M., Uhl, G.R., 1994. Choline/orphan V8-2-1/creatine
transporter mRNA is expressed in nervous, renal and gastro-
intestinal systems. Mol. Brain Res. 23, 266270.
Homberg, U., Hoskins, S.G., Hildebrand, J.G., 1995. Distribution of
acetylcholinesterase activity in the deutocerebrum of the sphinx
moth Manduca sexta. Cell Tissue Res. 279, 249259.
Ivy, M.T., Newkirk, R.F., Karim, M.R., Mtshali, C.M.P., Townsel,
J.G., 2001. Hemicholinium-3 mustard reveals two populations of
cycling choline transporters in Limulus. Neuroscience 102, 96978.
Jope, R.S., 1979. High afnity choline transport and acetylCoA
production and their roles in the regulation of acetylcholine
synthesis. Brain Res. Rev. 1, 313344.
Knipper, M., Breer, H., 1986. Regulation of the high afnity choline
uptake in locust synaptosomes by adenosine triphosphate. Neu-
rosci. Lett. 72, 347351.
Knipper, M., Boekhoff, I., Breer, H., 1989a. Isolation and reconstitu-
tion of the high-afnity choline carrier. FEBS Lett. 245, 235237.
Knipper, M., Krieger, J., Breer, H., 1989b. Hemicholinium-3 binding
sites in the nervous tissue of insects. Neurochem. Int. 14, 211215.
Knipper, M., Strotmann, J., Ma dler, U., Kahle, C., Breer, H., 1989c.
Monoclonal antibodies against the high afnity choline transport
system. Neurochem. Int. 14, 217222.
Knipper, M., Kahle, C., Breer, H., 1991. Purication and reconstitu-
tion of the high afnity choline transporter. Biochim. Biophys.
Acta 1065, 107113.
Kuhar, M.J., Murrin, L.C., 1978. Sodium-dependent high afnity
choline uptake. J. Neurochem. 30, 1521.
Lockman, P.R., Allen, D.D., 2002. The transport of choline. Drug
Dev. Ind. Pharm. 28, 749771.
Malutan, T., McLean, H., Caveney, S., Donly, C., 2002. A high-
afnity octopamine transporter cloned from the central nervous
system of cabbage looper Trichoplusia ni. Insect Biochem. Mol.
Biol. 32, 343357.
Matsuda, K., Buckingham, S.D., Kleier, D., Rauh, J.J., Grauso, M.,
Sattelle, D.B., 2001. Neonicotinoids: insecticides acting on insect
nicotinic acetylcholine receptors. Trends Pharm. Sci. 22, 573580.
Mayser, W., Schloss, P., Betz, H., 1992. Primary structure and
functional expression of a choline transporter expressed in the rate
nervous system. FEBS Lett. 305, 3136.
Okuda, T., Haga, T., 2000. Functional characterization of the human
high-afnity choline transporter. FEBS Lett. 484, 9297.
Okuda, T., Haga, T., 2003. High-afnity choline transporter.
Neurochem. Res. 28, 483488.
Okuda, T., Haga, T., Kanai, Y., Endou, H., Ishihara, T., Katsura, I.,
2000. Identication and characterization of the high-afnity
choline transporter. Nat. Neurosci. 3, 120125.
ARTICLE IN PRESS
H. McLean et al. / Insect Biochemistry and Molecular Biology 35 (2005) 6172 71
Pitman, R., 1971. Transmitter substances in insects: a review. Comp.
Gen. Pharmacol. 2, 347371.
Po rzgen, P., Park, S.K., Hirsh, J., Sonders, M.S., Amara, S.G., 2001.
The antidepressant-sensitive dopamine transporter in Drosophila
melanogaster: A primordial carrier for catecholamines. Mol.
Pharmacol. 59, 8395.
Saier, M.H., 2000. A functional-phylogenetic classication system for
transmembrane solute transporters. Microbiol. Mol. Biol. Rev. 64,
354411.
Sambrook, J., Russell, D.W., 2001. Molecular Cloning: A Laboratory
Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor,
New York.
Sattelle, D., 1985. Acetylcholine receptors. In: Kerkut, G.A., Gilbert,
L.I. (Eds.), Comprehensive Insect Physiology, Biochemistry
and Pharmacology, vol. 11. Pergamon Press, Oxford, pp. 395434.
Sattelle, D.B., Breer, H., 1990. Cholinergic nerve terminals in the
central nervous system of insects. J. Neuroendocrinol. 2,
241256.
Sattelle, D.B., Harrow, I.D., Hue, B., Pelhate, M., Gepner, J.I., Hall,
L.M., 1985. a-Bungarotoxin blocks excitatory synaptic transmis-
sion between cercal sensory neurons and giant interneuron 2
of the cockroach Periplaneta americana. J. Exp. Biol. 107,
473489.
Simon, J.R., Kuhar, M.J., 1976. High afnity choline uptake: ionic
and energy requirements. J. Neurochem. 27, 9399.
Sinclair, C.J., Chi, K.D., Subramanian, V., Ward, K.L., Green, R.M.,
2000. Functional expression of a high afnity mammalian hepatic
choline/organic cation transporter. J. Lipid Res. 41, 18411848.
Sterling, G.H., Doukas, P.H., Ricciardi, F.J., Biedrzycka, D.W.,
ONiell, J.J., 1986. Inhibition of high-afnity choline uptake and
acetylcholine synthesis by quinuclidinyl and hemicholinium deri-
vatives. J. Neurochem. 14, 11701175.
Vickers, N.J., Christensen, T.A., Hildebrand, J.G., 1998. Integrating
behavior with neurobiology: odor-mediated moth ight and
olfactory discrimination by glomerular arrays. Integrative Biol. 1,
224230.
Waldrop, B., Hildebrand, J., 1989. Physiology and pharmacology of
acetylcholinergic response of interneurons in the antennal lobes of
the moth Manduca sexta. J. Comp. Physiol. A 164, 433441.
Wang, Y., Cao, Z., Newkirk, R.F., Ivy, M.T., Townsel, J.G., 2001.
Molecular cloning of a cDNA for a putative choline co-transporter
from Limulus CNS. Gene 268, 123131.
ARTICLE IN PRESS
H. McLean et al. / Insect Biochemistry and Molecular Biology 35 (2005) 6172 72
Insect
Biochemistry
and
Molecular
Biology
Insect Biochemistry and Molecular Biology 35 (2005) 7384
Characterization of a silkworm thioredoxin peroxidase that is induced
by external temperature stimulus and viral infection
Kwang Sik Lee
a
, Seong Ryul Kim
a
, Nam Sook Park
a
, Iksoo Kim
b
, Pil Dong Kang
b
,
Bong Hee Sohn
b
, Kwang Ho Choi
b
, Seok Woo Kang
b
, Yeon Ho Je
c
, Sang Mong Lee
d
,
Hung Dae Sohn
a
, Byung Rae Jin
a,
a
College of Natural Resources and Life Science, Dong-A University, Busan 604-714, Republic of Korea
b
Department of Agricultural Biology, National Institute of Agricultural Science and Technology, RDA, Suwon 441-100, Republic of Korea
c
School of Agricultural Biotechnology, Seoul National University, Seoul 151-742, Republic of Korea
d
Department of Sericultural and Entomological Biology, Miryang National University, Miryang 627-130, Republic of Korea
Received 24 March 2004; accepted 24 September 2004
Abstract
A thioredoxin peroxidase (TPx) that reduces H
2
O
2
was rstly characterized in the lepidopteran insect, silkworm Bombyx mori.
The B. mori TPx (BmTPx) cDNA contains an open reading frame of 585 bp encoding 195 amino acid residues and possesses two
cysteine residues that are characteristic of 2-Cys subgroup of peroxiredoxin family. The deduced amino acid sequence of the BmTPx
cDNA showed 78% identity to Drosophila melanogaster (DmTPx-1), 73% to Aedes aegypti (AaTPx), and 5448% to other insect 2-
Cys TPx. The cDNA encoding BmTPx was expressed as a 25-kDa polypeptide in baculovirus-infected insect Sf9 cells. The puried
recombinant BmTPx was shown to reduce H
2
O
2
in the presence of electrons donated by dithiothreitol and shown to be active in the
presence of thioredoxin as electron donor. Northern blot analysis revealed the presence of BmTPx transcripts in all tissues
examined. Western blot analysis showed the presence of the BmTPx in the fat body and midgut, but not in the hemolymph,
suggesting the BmTPx is not secretable. When H
2
O
2
was injected into body cavity of B. mori larva, BmTPx mRNA expression was
up-regulated in the fat body tissues. Interestingly, the expression levels of BmTPx enzyme in the fat body were particularly high
when B. mori larva was exposed at low (4 1C) and high (37 1C) temperatures or baculovirus infection, suggesting that the BmTPx
seems to play a protective role against oxidative stress caused by temperature stimuli and viral infection.
r 2004 Elsevier Ltd. All rights reserved.
Keywords: Antioxidant enzyme; Bombyx mori; Insect; Oxidative stress; Peroxiredoxin; Reactive oxygen species; Silkworm; Thioredoxin peroxidase
1. Introduction
Organisms living in aerobic environments require
defense mechanisms that prevent oxidative damage
caused by reactive oxygen species (ROS). ROS such as
the superoxide anion, hydrogen peroxide and the
hydroxyl radical are noted for their high reactivity and
resultant damage to proteins, lipid membranes, and
DNA (Halliwell and Gutterridge, 1989). To protect
against the toxicity of ROS, aerobic organisms have
evolved protective enzymatic systems. Among these,
thioredoxin peroxidase (TPx) is known to eliminate
H
2
O
2
and alkyl hydroperoxidases with use of a thiol-
reducing equivalent (Lim et al., 1993; Chae et al.,
1994a, b; Kang et al., 1998a).
The TPx family is a large family of antioxidant
proteins ubiquitously found in all living organisms, from
prokaryotes to eukaryotes (Chae et al., 1994b). Sacchar-
omyces cerevisiae cytosolic TPx I was the rst peroxir-
edoxin (Prx) isolated from an eukaryotic cell (Kim et al.,
1988). The response in the yeast cytosolic TPx I by
oxidative stress indicated that its gene, encoded by
ARTICLE IN PRESS
www.elsevier.com/locate/ibmb
0965-1748/$ - see front matter r 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ibmb.2004.09.008

Corresponding author. Tel.: +82 51 200 7594;


fax: +82 51 200 7594.
E-mail address: brjin@daunet.donga.ac.kr (B.R. Jin).
TSA1 (thiol-specic antioxidant), is transcriptionally
activated and highly expressed when cells are exposed to
H
2
O
2
(Demasi et al., 2001). The Prx family members
contain one or two conserved cysteines that correspond
to Cys47 or both Cys47 and Cys170 in the yeast TPx
(Chae et al., 1994ac). TPx reduces H
2
O
2
with the use of
electrons from thioredoxin and contains two essential
cysteines. Cys47 is the primary site of oxidation by
H
2
O
2
, and the oxidized Cys47 rapidly reacts with
Cys170 of the other subunit to form an intermolecular
disulde (Chae et al., 1994ac).
In insects, TPx genes have been isolated from
Drosophila melanogaster, Aedes aegypti and Apis melli-
fera have been isolated (Fang and Li, 2001; Radyuk et
al., 2001; Whiteld et al., 2002). Previously, known three
different TPxs in D. melanogaster have distinct patterns
of subcellular localization and expression (Radyuk et
al., 2001). The three 2-Cys TPxs were shown to reduce
H
2
O
2
in the presence of dithiothreitol and were also
shown to be active in the thioredoxin system. It is
supposed that TPxs play a particularly central role in the
enzymic removal of ROS (Radyuk et al., 2001, 2003). A
recent study on functional role of TPxs in D. melano-
gaster S2 cells showed higher resistance to oxidative
stress in the transfected cell line overexpressing TPx than
that underexpressing TPx (Radyuk et al., 2003).
Stressors, such as temperatures, have been reported to
act, at least in part, via oxidative stress-related mechan-
isms. Lethal heat shock stimulates polyamine oxidation
that generates hydrogen peroxide in mammalian cells
(Hariari et al., 1989). The oxidative stress plays a major
role in the lethal effect of heat in eukaryotes and
overexpression of antioxidant enzymes caused an
increase in thermotolerance (Davidson et al., 1996).
Lee and Park (1998) reported that the antioxidative
function of TPx facilitates the cells defense against heat
shock. Rauen et al. (1999) reported that the formation
of ROS is a key mediator of cold-induced apoptosis in
animal cells. On the other hand, neutrophils and
monocytes from HIV-infected patients spontaneously
produced increased amounts of H
2
O
2
and the produc-
tion was associated with changes in the expression of the
antiapoptotic/antioxidant compounds Bcl-2 and thior-
edoxin along the course of the disease (Elbim et al.,
2001). In Drosophila cells, amount of H
2
O
2
was lower in
the cells overexpressing TPx relative to the control and
the TPx enzymes have both benecial and detrimental
effects on cell viability under different stress conditions
such as heat and metals (Radyuk et al., 2003).
The purpose of the present study was to elucidate the
TPx gene and its function in the silkworm, Bombyx mori
as a model insect for lepidopteran insect. In this paper,
we reported the cDNA cloning, expression and char-
acterization of a silkworm TPx (BmTPx) for the rst
time. To gain insight into the physiological and
pathological roles of BmTPx in silkworm, furthermore,
we explored the transcriptional induction and transla-
tional activation of BmTPx in vivo by external low- and
high-temperature stimuli or B. mori nucleopolyhedrosis
viral infection.
2. Materials and methods
2.1. Animals
The larvae of the silkworm, B. mori, used in this study
were Chinese race Jam 108 supplied by Department of
Sericulture and Entomology, The National Institute of
Agricultural Science and Technology, Korea. Silkworms
were reared on fresh mulberry leaves at 25 1C, 6575%
of relative humidity, and 12 h light:12 h dark photo-
period as usual.
2.2. cDNA library screening, nucleotide sequencing and
data analysis
A cDNA library (Kim et al., 2003) constructed using
whole bodies of B. mori larvae was used in this study.
The clones harboring cDNA inserts were randomly
selected and sequenced to generate the expressed
sequence tags (ESTs) (Kim et al., 2003). The plasmid
DNA was extracted by Wizard mini-preparation kit
(Promega, Madison, WI). The nucleotide sequence was
determined by using a BigDyeTerminator cycle sequen-
cing kit in the automated DNA sequencer (model 310
Genetic Analyzer; Perkin-Elmer Applied Biosystems,
Foster City, CA). The sequences were compared using
the DNASIS and BLAST programs provided by the
NCBI (http://www.ncbi.nlm.nih.gov/BLAST). Gen-
Bank, EMBL and SwissProt databases were searched
for sequence homology using a BLAST algorithm
program. MacVector (ver. 6.5, Oxford Molecular Ltd.)
was used to align the amino acid sequences of TPx. With
the seven GenBank-registered TPx amino acid se-
quences, phylogenetic analysis was performed using
PAUP (Phylogenetic Analysis Using Parsimony) version
4.0 (Swofford, 2000). The accession numbers of the
sequences in the GenBank are as follows: B. mori
(AY438331; this study), D. melanogaster TPx-1
(Q9V3P0), D. melanogaster TPx-2 (AAF42986), D.
melanogaster TPx-3 (AAG41976), Aedes aegypti TPx
(AAL37254), Apis mellifera TPx (AAP93584), D.
melanogaster Prx-2540 (AAG47823), and D. melanoga-
ster Prx-6005 (AAG47822).
2.3. RNA isolation and northern blot analysis
The larvae of B. mori were dissected under the Stereo-
microscope (Zeiss, Jena, Germany), individual samples
such as fat body, midgut, silk gland, and epidermis were
harvested, and washed twice with PBS (140 mM NaCl,
ARTICLE IN PRESS
K.S. Lee et al. / Insect Biochemistry and Molecular Biology 35 (2005) 7384 74
27 mM KCl, 8 mM Na
2
HPO
4
, 1.5 mM KH
2
PO
4
, pH
7.4). Total RNA was isolated from the fat body, midgut,
silk gland, and epidermis of B. mori larvae by using the
Total RNA Extraction Kit (Promega). Total RNA
(10 mg/lane) from B. mori was denatured by glyoxalation
(McMaster and Carmichael, 1977), transferred onto a
nylon blotting membrane (Schleicher & Schuell, Dassel,
Germany) and hybridized at 42 1C with a probe in a
buffer containing 2 PIPES, 50% formamide, 1%
sodium dodecyl sulfate (SDS) and blocking agent
(Boehringer Mannheim, Mannheim, Germany). The
probe used to detect the TPx gene transcripts was
966 bp for BmTPx cDNA cloned in this study and
labeled with [a-
32
P] dCTP (Amersham, Arlington
Heights, IL) using the Prime-It II Random Primer
Labeling Kit (Stratagene, La Jolla, CA). After hybridi-
zation, the membrane lter was washed three times for
30 min each in 0.1% SDS and 0.2 SSC (1 SSC is
0.15 M NaCl and 0.015 M sodium citrate) at 65 1C, and
nally exposed to autoradiography lm.
2.4. Construction of baculovirus transfer vector
The 966 bp BmTPx cDNA from pBlueScript-AgCell
was subcloned between Sac I and Kpn I sites of
pBacPAK9 (Clontech, Palo Alto, CA) to produce
transfer vector pBacPAK9-BmTPx. In the transfer
vector, the TPx cDNA is under the control of the
polyhedrin promoter of Autographa californica nuclear
polyhedrosis virus (AcNPV).
2.5. Cell culture and virus
The Spodoptera frugiperda IPLB Sf21-AE (Vaughn
et al., 1977) clone 9 (Sf9) cells and B. mori (Bm5) cells (Je
et al., 2001) were maintained at 27 1C in TC100 medium
(GIBCO BRL LIFE Technologies, Gaithersburg, MD),
supplemented with 10% fetal bovine serum (FBS;
GIBCO BRL LIFE Technologies) as described by
standard methods (OReilly et al., 1992). Wild-type
AcNPV and recombinant AcNPV were propagated in
Sf9 cells. Wild-type B. mori nucleopolyhedrosis virus
(BmNPV) was propagated in Bm5 cells. The titer was
expressed as plaque forming units (PFU) per ml
(OReilly et al., 1992).
2.6. Construction of recombinant virus
Thirty-ve millimeter cell culture dishes were seeded
with 1.01.5 10
6
cells and incubated at 27 1C for 1 h to
allow cell attachment. One microgram of BacPAK6
viral DNA (Clontech), 5 mg of pBacPAK9-BmTPx in
20 mM HEPES buffer and sterile water to make a total
volume of 50 ml were mixed in a polystyrene tube. Fifty
microliters of 100 mg/ml Lipofectin
TM
(GIBCO BRL
LIFE Technologies) were gently mixed with the DNA
solution and the mixture was incubated at room
temperature for 30 min. The cells were washed twice
with 2 ml serum-free TC100 medium and fed with 1.5 ml
serum-free TC100 medium. The Lipofectin-DNA com-
plexes were added dropwise to the medium covering the
cells while the dish was gently swirled. After incubation
at 27 1C for 5 h, TC100 medium containing antibiotics
and 10% FBS was added to each dish and incubation at
27 1C was continued. At 5 days of postinfection (p.i.),
the supernatant was harvested, claried by centrifuga-
tion at 2000 rpm for 5 min, and stored at 4 1C.
Recombinant AcNPV was plaque puried on 6-well
plates seeded with 1.5 10
6
Sf9 cells as described by
OReilly et al. (1992).
2.7. SDS-polyacrylamide gel electrophoresis (PAGE)
Insect Sf9 cells were mock-infected or infected with
the wild-type AcNPV and recombinant AcNPV in a 35-
mm diameter dish (1 10
6
cells) at a multiplicity of
infection (MOI) of 5 PFU per cell. After incubation at
27 1C, cells were harvested at 3 days p.i. For SDS-PAGE
(Laemmli, 1970) of cell lysates, uninfected Sf9 cells and
cells infected with virus were washed twice with PBS and
mixed with protein sample buffer and boiled. The total
cellular lysates were subjected to 10% SDS-PAGE.
2.8. Purication of recombinant BmTPx
Insect Sf9 cells were infected with the recombinant
AcNPV expressing BmTPx at a MOI of 5 PFU per cell.
The culture supernatant was harvested at 5 days p.i. and
claried by centrifugation (10,000g) at 4 1C for 10 min.
The culture supernatant was supplemented with 1 M
ammonium sulfate and applied to gel permeation
chromatography on a Superdex 200 HR 10/30 column
(Pharmacia LKB) in 20 mM TrisHCl buffer (pH 8.0)
containing 0.1 mM PMSF with a ow rate of 0.5 ml/
min. Fractions containing recombinant BmTPx were
respectively assayed for enzyme activity. The recombi-
nant BmTPx-enriched fractions, identied by enzyme
activity, were subjected to anion exchange chromato-
graphy on a MonoQ HR 5/5 column (Pharmacia LKB)
in 20 mM TrisHCl buffer (pH 8.0) containing 0.1 mM
PMSF with a ow rate of 1 ml/min. Enzymes were
eluted in a linear segment sodium chloride gradient
(00.6 M). Fractions containing puried recombinant
BmTPx were, respectively, identied by SDS-PAGE and
enzyme activity.
2.9. Preparation of polyclonal antiserum and western blot
analysis
The puried recombinant BmTPx (about 10 mg) was
mixed with equal volume of Freunds complete adjuvant
(a total of 200 ml; Sigma, St. Louis, MO) and injected
ARTICLE IN PRESS
K.S. Lee et al. / Insect Biochemistry and Molecular Biology 35 (2005) 7384 75
into Balb/c mice, respectively. Three successive injec-
tions were performed with 1-week interval beginning a
week after the rst injection with antigens mixed with
equal volume of Freunds incomplete adjuvant (a total
of 200 ml, Sigma). Blood was collected 3 days after the
last injection and centrifuged at 13,000 rpm for 5 min.
The supernatant antibodies were stored at 70 1C until
use. For Western blot analysis, 10% SDS-PAGE was
performed as described above. Proteins were blotted to a
sheet of nitrocellulose transfer membrane (Schleicher &
Schuell) (Towbin et al., 1979). The blotting was
performed in transfer buffer (25 mM Tris and 192 mM
glycine in 20% methanol) at 30 volts overnight at 4 1C.
After blotting, the membrane was blocked by incubation
in 1% BSA solution for 2 h at room temperature. The
blocked membrane was incubated with antiserum
solution (1:1000 v/v) for 1 h at room temperature and
washed in TBST (10 mM TrisHCl, pH 8.0, 100 mM
NaCl, 0.05% Tween 20). Subsequently, the membrane
was incubated with anti-mouse IgG horseradish perox-
idase (HRP) conjugate and HRPstreptavidin complex.
After repeated washing, the membrane was incubated
with ECL detection reagents (Amersham Pharmacia
Biotech) and nally exposed to autoradiography lm.
2.10. Collection of hemolymph
Hemolymph was collected into cold test tubes by
cutting forelegs of 3-day-old 5th instar B. mori larva. A
few crystals of phenylthiourea were added to the tubes
to prevent melanization. The collected hemolymph was
centrifuged at 10,000g for 10 min to remove hemocytes
and cell debris, and the supernatant was stored at
70 1C until used.
2.11. In vitro peroxidase activity
Reduction of H
2
O
2
by puried recombinant BmTPx
has been estimated in the presence or absence of DTT by
monitoring decrease in OD at A
240
, as previously
described (Kang et al., 1998a,b; Radyuk et al., 2001).
H
2
O
2
degradation was monitored at 240 nm and the
differences in OD readings were plotted on the Y-axis.
The reactions were carried out with 10 mg of puried
BmTPx or 1.25 mg BSA in the absence or presence of
10 mM DTT or in the presence of DTT and the same
amount of thermo-inactivated BmTPx (boiled for 5 min
in 0.5% SDS). Heat-inactivated BmTPx and BSA were
used as negative controls.
To determine TPx activity, 10 mg of puried BmTPx
was added to 50 mM HepesNaOH buffer, pH 7.0,
containing 0.25 mM NADPH, 0.1 mM H
2
O
2
, 20 mg/ml
of thioredoxin, and 10 mg/ml of thioredoxin reductase.
Oxidation of NADPH as monitored by the decrease in
absorbance at 340 nm was used as an indirect measure of
peroxidase activity (Kang et al., 1998a,b; Radyuk et al.,
2001). The same amount of thermo-inactivated BmTPx
(boiled for 5 min in 0.5% SDS) served as a negative
control.
2.12. TPx expression in vivo after H
2
O
2
treatment
The 3-day-old 5th instar silkworm larva was injected
with 10 mM H
2
O
2
. After 5 h treatment, fat body cells
from B. mori larva treated with H
2
O
2
were harvested
and washed twice with PBS. Total RNA was isolated
from the fat body by using the Total RNA Extraction
Kit (Promega). Transcriptional induction of BmTPx
was analyzed by Northern blot hybridization as
described above.
2.13. TPx expression in vivo after temperature treatment
or viral infection
The 3-day-old 5th instar silkworm larvae were
exposed to 4 or 37 1C for 3 or 5 h, respectively, while
control was maintained at 25 1C. After treatment, fat
body cells from B. mori larvae were harvested and
washed twice with PBS. Total RNA and total cellular
lysates were isolated from the fat body by as described
above. Transcriptional and translational induction of
BmTPx to external temperature stimulus were analyzed
by Northern blot and Western blot as described above.
The 1-day-old 5th instar silkworm larvae were mock-
injected or injected with BmNPV of 2 10
5
PFU. At 1,
2 and 3 days p.i., fat body cells from B. mori larvae were
harvested and washed twice with PBS. Total RNA and
total cellular lysates were isolated from the fat body by
as described above. Transcriptional and translational
induction of BmTPx to viral infection were analyzed by
Northern blot and Western blot as described above.
Images of Northern blot and Western blot in this
study were analyzed with a computerized image analysis
system (Alpha Innotech Co., San Leandro, CA). Alpha
Imager 1220 (ver. 5.5) was used to aid the analyses. The
area of each band was calculated as the integrated
density value. Comparative studies of transcriptional
and translational induction of BmTPx by temperature
treatment or viral infection were performed with three
independent groups of silkworm larvae.
3. Results
3.1. cDNA cloning, sequencing and phylogenetic analysis
of BmTPx
In a search of B. mori ESTs (expressed sequence tags),
we identied a cDNA showing high homology with
previously reported TPx genes. The cDNA clone
including the full-length open reading frame (ORF)
was sequenced and characterized. The nucleotide and its
ARTICLE IN PRESS
K.S. Lee et al. / Insect Biochemistry and Molecular Biology 35 (2005) 7384 76
deduced amino acid sequences of the cDNA encoding
TPx are presented in Fig. 1A. The BmTPx cDNA is an
ORF of 585 bp encoding 195 amino acid residues.
A multiple sequence alignment of the deduced protein
sequence of BmTPx gene with available insect TPx
sequences indicates that BmTPx sequence is closely
related to D. melanogaster DmTPx-1 (Radyuk et al.,
2001) and Aedes aegypti AaTPx (Fang and Li, 2001),
belonging to 2-Cys TPx (Fig. 1B). The BmTPx also has
two conserved cysteines. The deduced protein sequence
of the BmTPx showed 78% and 73% protein sequence
identity to DmTPx-1 and AaTPx, respectively, while low
sequence identity was found with DmTPx-3 (48%
protein sequence identity) (Fig. 1C).
3.2. Expression, purication, and peroxidase activity of
recombinant BmTPx
To assess BmTPx gene, the 966 bp for BmTPx cDNA
was inserted into baculovirus transfer vector. Transfer
vector pBacPAK9-BmTPx was constructed by insertion
of BmTPx gene under the control of AcNPV polyhedrin
promoter of pBacPAK9 (data not shown). The baculo-
virus transfer vector was used to generate a recombinant
virus expressing BmTPx. Recombinant AcNPV, which
we have termed AcNPV-BmTPx, was produced in insect
Sf9 cells by cotransfection with wild-type AcNPV DNA
and the transfer vector.
To examine the expression of BmTPx by recombinant
virus in insect cells, the protein synthesis in Sf9 cells
infected with the recombinant virus was analyzed by
SDS-PAGE (Fig. 2A) and Western blot (Fig. 2B). The
recombinant BmTPx expressed by the BmTPx cDNA
was present as a single band of about 25 kDa polypep-
tide in the cells infected with the recombinant virus, but
not in the cells infected with the wild-type AcNPV or
mock-infected cells.
In order to characterize the recombinant BmTPx
expressed in baculovirus-infected insect cells, the re-
combinant BmTPx with a molecular mass of approxi-
mately 25 kDa was puried from the culture supernatant
using FPLC techniques. Fractions showing enzyme
activity from Superdex 200 HR column were applied
to anion exchange chromatography. The result showed
that peaks from MonoQ HR column were eluted and a
protein peak corresponded to the activity peak. The
puried recombinant BmTPx was identied as a single
band of 25 kDa by SDS-PAGE (Fig. 2C).
Thiol-dependent peroxidase activity of the puried
recombinant BmTPx was determined by directly mon-
itoring the decrease in H
2
O
2
concentration in the
presence of a thiol source DTT (Fig. 2D). The result
clearly revealed that the addition of 10 mg of the puried
recombinant BmTPx in the presence of DTT signi-
cantly increased the rate of H
2
O
2
degradation, whereas
the puried recombinant BmTPx in the absence of DTT
did not, indicating that BmTPx activity is thiol-
dependent. When BmTPx was thermally inactivated,
the peroxidase activity of BmTPx showed a substantial
drop to the level of without-DTT.
The puried recombinant BmTPx was assayed
indirectly for TPx activity by monitoring the decrease
in absorbance at 340 nm due to oxidation of NADPH in
the presence of peroxides (Fig. 2E). The result showed
that BmTPx clearly exhibited TPx activity and this
activity was lost upon thermal inactivation of the
protein.
3.3. TPx expression in silkworm tissues
To conrm the expression of BmTPx gene at
transcriptional level, Northern blot analysis was per-
formed using mRNA prepared from fat body, midgut,
silk gland and epidermis, respectively (Fig. 3A).
Hybridization signal was present in all tissues examined.
The presence of BmTPx in the larval fat body,
midgut extracts, and hemolymph was detected by
Western blot analysis using antiserum against recombi-
nant BmTPx (Fig. 3B). Western blot analysis showed
the presence of the BmTPx in the fat body and midgut
as a single band of about 25 kDa polypeptide, but not in
the hemolymph.
3.4. TPx expression in vivo after H
2
O
2
treatment
When H
2
O
2
was injected into body cavity of B. mori
larva, the transcript level of BmTPx was assessed in total
RNA isolated from the fat body tissues (Fig. 4).
Northern blot of mRNA was hybridized with BmTPX
cDNA probe cloned in this study. As expected, the
transcript level of BmTPx was dramatically increased in
the fat body tissues from B. mori larva treated with
H
2
O
2
, indicating that the BmTPx gene is up-regulated.
3.5. TPx expression in vivo after temperature treatment
To identify the induction of BmTPx gene to external
temperature stimulus, 5th instar B. mori larvae were
exposed to 4 or 37 1C for 3 or 5 h, respectively, while
controls were maintained at 25 1C. The induction of
BmTPx from fat body cells was analyzed by Northern
blot and Western blot. As shown in Fig. 5, the level of
BmTPx from fat body cells signicantly increased
during the exposure at low- and high-temperature
conditions, compared with control. The result indicated
that BmTPx is transcriptionally and translationally
induced by low-temperature shock as well as high-
temperature shock. In addition, BmTPx is signicantly
induced at high levels on the 5 h of each temperature
treatment.
ARTICLE IN PRESS
K.S. Lee et al. / Insect Biochemistry and Molecular Biology 35 (2005) 7384 77
ARTICLE IN PRESS
Fig. 1. cDNA and deduced protein sequence of the BmTPx. (A) The nucleotide and deduced protein sequence of the BmTPx cDNA. The start codon
of ATG is boxed and the termination codon is shown by asterisk. In the cDNA sequence, the polyadenylation sequence is underlined. This cDNA
sequence has been deposited in GenBank under accession number AY438331. (B) Alignment of the amino acid sequence of BmTPx with known
TPxs. Residues are numbered according to the aligned insect TPx sequences, and invariant residues are shaded black. Dots represent gaps introduced
to improve alignment. The conserved cysteine residues are shown by asterisk. The insect TPx sequences were taken from the following sources: D.
melanogaster TPx-1 (DmTPx-1; Q9V3P0), Aedes aegypti TPx (AaTPx; AAL37254), D. melanogaster TPx-2 (DmTPx-2; AAF42986), Apis mellifera
TPx (AmlTPx; AAP93584), and D. melanogaster TPx-3 (DmTPx-3; AAG41976). (C) Pairwise identities and similarities of the deduced amino acid
sequence of BmTPx among insect TPx sequences.
K.S. Lee et al. / Insect Biochemistry and Molecular Biology 35 (2005) 7384 78
3.6. TPx expression in vivo after viral infection
To assess the induction of BmTPx gene on viral
infection, furthermore, 5th instar B. mori larvae were
injected with BmNPV. At 1, 2 and 3 days p.i., total
RNA and cellular lysates were isolated from the fat
body cells of B. mori larvae. The induction of BmTPx
from fat body cells by baculovirus infection was
analyzed by Northern blot and Western blot. The result
showed that the level of BmTPx from fat body cells
signicantly increased in the infected larvae, compared
with control (Fig. 6), indicating that BmTPx is
transcriptionally and translationally induced by baculo-
virus infection.
4. Discussion
The antioxidative function of the TPxs, which
eliminate hydrogen and organic peroxides, using
ARTICLE IN PRESS
Fig. 2. Expression, purication and enzyme activity assay of the recombinant BmTPx in baculovirus-infected insect Sf9 cells and purication of its
recombinant BmTPx. Sf9 cells were mock-infected (lane 2) or infected with the wild-type AcNPV (lane 3) and the recombinant AcNPV (lane 4) at an
MOI of 5 PFU per cell. Cells were collected at 2 (lane 3) and 3 (lane 4) days p.i. Total cellular lysates were subjected to 12% SDS-PAGE (A),
electroblotted and incubated with antiserum to recombinant BmTPx (B). The arrow on the right of the panel indicates the 25 kDa recombinant
BmTPx polypeptide. Molecular weight standards were used as size marker (lane 1). (C) Purication of the recombinant BmTPx expressed in
baculovirus-infected insect cells. The recombinant BmTPx puried by using FPLC techniques was analyzed by 12% SDS-PAGE. The arrow on the
right of the panel indicates the puried recombinant BmTPx polypeptide of 25 kDa. (D) Hydrogen peroxide elimination by the recombinant BmTPx
expressed in baculovirus-infected insect Sf9 cells. H
2
O
2
degradation was monitored at 240 nm and the differences in OD readings were plotted on the
Y-axis. The reactions were carried out with 10 mg of each puried BmTPx or 1.25 mg BSA in the absence (DTT) or presence (+DTT) of 10 mM
DTT or in the presence of DTT and the same amount of thermo-inactivated BmTPx [boiled for 5 min in 0.5% SDS (+DTT, t)]. Heat-inactivated
BmTPx and BSA were used as negative controls. (E) Peroxidase activity of recombinant BmTPx in the presence of thioredoxin as electron donor.
NADPH oxidation was monitored at 340 nm in the presence of H
2
O
2
, thioredoxin, and thioredoxin reductase. Assay was performed with 10 mg of
BmTPx and the same amount of thermo-inactivated BmTPx [boiled for 5 min in 0.5% SDS (BmTPx, t)] served as a negative control.
K.S. Lee et al. / Insect Biochemistry and Molecular Biology 35 (2005) 7384 79
thioredoxin as the electron donor, has also been
demonstrated in insect species, especially for D.
melanogaster, and the TPxs play a particularly central
role in the enzymic removal of ROS (Radyuk et al.,
2001, 2003; Bauer et al., 2002). Therefore, TPxs are
believed to be present in insects for the protection
against the toxicity of ROS. The TPx genes in some
insect species including D. melanogaster, Aedes aegypti
and Apis mellifera have been isolated (Fang and Li,
2001; Radyuk et al., 2001; Whiteld et al., 2002).
Especially, ve distinct genes of the Prx family were
reported in D. melanogaster (Radyuk et al., 2001). A
recent work conrmed that TPxs can foster cytoprotec-
tion or cell death in response to different stressors
(Radyuk et al., 2003).
In this study, a novel TPx gene was cloned from the
larvae of the silkworm, B. mori as a model insect for
lepidopteran insect. This is the rst report about TPx
gene in lepidopteran insect species. The BmTPx cDNA
comprised a 585 bp encoding TPx of 195 amino acid
residues. Conservation of two cysteines and multiple
sequence alignment suggest that the BmTPx belongs to
2-Cys subgroup. Most Prx family members of the 2-Cys
subgroups possess peroxide reductase activity, capable
of using thioredoxin as an electron donor, and therefore
have been named TPxs (Chae et al., 1994b; Kang et al.,
1998a; Park et al., 2000; Zhou et al., 2000; Radyuk et al.,
2001). Among the ve Drosophila Prx, three of them
(TPx-1, TPx-2, and TPx-3) are 2-Cys type (Radyuk et
al., 2001) and BmTPx is closest to the DmTPx-1 (78%
protein sequence identity) and next to AaTPx (73%
protein sequence identity).
The BmTPx cDNA was expressed in the insect Sf9
cells and the recombinant BmTPx is detected as a single
band of approximately 25 kDa polypeptide in SDS-
PAGE. Furthermore, the puried recombinant BmTPx
showed peroxidase activity in the presence of a thiol
source DTT, but not in the absence of DTT, indicating
that BmTPx activity is thiol-dependent. We demon-
strated here that BmTPx has peroxidase activity in the
presence of the thioredoxin system (thioredoxin+thior-
edoxin reductase). This result indicates that BmTPx is
one of the TPx family members. Our results are in good
agreement with the previous result that three 2-Cys
DmTPxs were shown to reduce H
2
O
2
in the presence of
DTT and shown to have peroxidase activity in the
ARTICLE IN PRESS
H
2
O
2
Treatment
1
EtBr
Northern blot
2
+
Fig. 4. Induction of BmTPx by in vivo injection of H
2
O
2
. The 3-day-
old 5th instar silkworm larva was injected without (lane 1) or with
(lane 2) 10 mM H
2
O
2
per each larva. After 5 h-treatment, total RNA
was isolated from the fat body, respectively. The RNA was separated
by 1.0% formaldehyde agarose gel electrophoresis (lower panel),
transferred on to a nylon membrane, and hybridized with radiolabelled
966 bp BmTPx cDNA (upper panel). Transcripts are indicated on the
right side of the panel by arrow.
1 3 4
2 4
97.4
66
45
31
21.5
(kDa)
M
a
r
k
e
r
f
a
t

b
o
d
y
m
i
d
g
u
t
h
e
m
o
l
y
m
p
h
f
a
t

b
o
d
y
m
i
d
g
u
t
h
e
m
o
l
y
m
p
h
f
a
t

b
o
d
y
m
i
d
g
u
t
s
i
l
k

g
l
a
n
d
e
p
i
d
e
r
m
i
s
EtBr
Northern blot BmTPx
1 3 4 2
2
3
(A)
(B)
Fig. 3. BmTPx expression in silkworm tissues. (A) Northern blot
analysis of BmTPx. Total RNA was isolated from the fat body (lane
1), midgut (lane 2), silk gland (lane 3), and epidermis (lane 4),
respectively (upper panel). The RNA was separated by 1.0%
formaldehyde agarose gel electrophoresis, transferred on to a nylon
membrane, and hybridized with radiolabelled 966 bp BmTPx cDNA
(lower panel). Transcripts are indicated on the right side of the panel
by an open arrow. (B) Expression of BmTPx in B. mori larva. Total
cellular lysates were prepared from the fat body (lane 2) and midgut
(lane 3), respectively. The hemolymph (lane 4) was collected on the 3rd
day of 5th instar larva. The samples were subjected to 12% SDS-
PAGE (left panel), electroblotted and incubated with antiserum to
recombinant BmTPx (right panel). The arrow on the right of the panel
indicates the 25 kDa BmTPx polypeptide. Molecular weight standards
were used as size marker (lane 1).
K.S. Lee et al. / Insect Biochemistry and Molecular Biology 35 (2005) 7384 80
presence of the thioredoxin system (Radyuk et al.,
2001).
Previously, Radyuk et al. (2001) reported the identi-
cation of three different D. melanogaster TPxs, which
have distinct patterns of subcellular localization such as
cytosolic TPx (DmTPx-1), secretable TPx (DmTPx-2)
and mitochondrial TPx (DmTPx-3). Among the three
Drosophila TPxs, BmTPx was closest to the DmTPx-1, a
cytosolic TPx. In the present study, we observed the
presence of the BmTPx in the fat body and midgut, but
not in the hemolymph. Its absence in the hemolymph
suggests that the protein is not found in the extracellular
space and is therefore not secreted. However, it is still
unclear whether BmTPx is a cytosolic TPx or a
mitochondrial TPx. Thus, further investigation is
required to fully understand its detailed subcellular
localization.
Cytoplasmic TPx I could act as a housekeeping type
of peroxidase to regulate the intracellular ROS level
such as H
2
O
2
(Park et al., 2000). The expression of
BmTPx at transcriptional level indicated that the
BmTPx gene is expressed in all tissues examined,
suggesting some housekeeping/antioxidant role (Ra-
dyuk et al., 2001).
In addition, the transcript level of BmTPx was
dramatically increased in the fat body tissues of B. mori
larva injected with H
2
O
2
. This result demonstrates that
the role of BmTPx is directly related to the shielding of
cells against oxidative damage. As described by Demasi
et al. (2001), yeast cytosolic TPx-I is transcriptionally
ARTICLE IN PRESS
EtBr
Northern blot
3h 5h 3h 5h
4

C
2
5

C
3
7

C
1 4 5
3h 5h 3h 5h
4

C
2
5

C
3
7

C
Western blot
1 4 5
3h 5h 3h 5h
4C 25C 37C
0
50
100
150
200
250
300
350
3h 5h 3h 5h
4C 25C 37C
0
50
100
150
200
250
300
350
189
287
100
154
202
178
212
100
203
231
R
e
l
a
t
i
v
e

m
R
N
A

l
e
v
e
l
s

(
%
)
R
e
l
a
t
i
v
e

p
r
o
t
e
i
n

l
e
v
e
l
s

(
%
)
2 3
2 3
(A)
(B)
(C)
(D)
Fig. 5. Induction of BmTPx by external temperature stimulus. (A) Northern blot analysis of the BmTPx gene induced by external temperature
stimulus. The 5th instar silkworm larva was incubated at 4 1C (lanes 1 and 2) or 37 1C (lanes 4 and 5) for 3 h (lanes 1 and 4) or 5 h (lanes 2 and 5),
respectively. Control was indoor-reared at 25 1C (lane 3). Total RNA was isolated from the fat body of B. mori larva of each temperature treatment.
The RNA was separated by 1.0% formaldehyde agarose gel electrophoresis (upper panel), transferred on to a nylon membrane, and hybridized with
radiolabelled 966 bp BmTPx cDNA (lower panel). Transcripts are indicated on the right side of the panel by arrow. (B) Relative mRNA levels of
BmTPx induced by external temperature stimulus. The levels of BmTPx mRNA were normalized to the expression at the 25 1C. Bars represent
standard deviation. The data points reect the means of three independent experiments. (C) Western blot analysis of the BmTPx induced by external
temperature stimulus. The 5th instar silkworm larva was incubated at 4 1C (lanes 1 and 2) or 37 1C (lanes 4 and 5) for 3 h (lanes 1 and 4) or 5 h (lanes 2
and 5), respectively. Control was indoor-reared at 25 1C (lane 3). The protein samples were prepared from the fat body of B. mori larva of each
temperature treatment. The samples were subjected to 12% SDS-PAGE, electroblotted and incubated with antiserum to recombinant BmTPx. The
arrow on the right of the panel indicates the 25 kDa BmTPx polypeptide. (D) Relative protein levels of BmTPx induced by external temperature
stimulus. The levels of BmTPx were normalized to the expression at the 25 1C. Bars represent standard deviation. The data points reect the means of
three independent experiments.
K.S. Lee et al. / Insect Biochemistry and Molecular Biology 35 (2005) 7384 81
activated when cells were exposed to H
2
O
2
. On the other
hand, DmTPx-1 is a cytosolic protein, and overexpres-
sion of the gene in Drosophila S2 cells results in a higher
survival rate under peroxide exposure (Radyuk et al.,
2001).
Most living organisms are sensitive to sudden
temperature stress. A shift in temperature from a low
to an intermediate temperature induces the stress
response or heat-shock response (Lindquist, 1986). Heat
stress stimulates polyamine oxidation that generates
hydrogen peroxide in mammalian cells (Hariari et al.,
1989) and overexpression of antioxidant enzymes causes
an increase in thermotolerance (Davidson et al., 1996).
In addition, it has been reported that the formation of
ROS is a key mediator of cold-induced apoptosis in
animal cells (Rauen et al., 1999). In particular, a study
concluded that the antioxidative function of TPx
facilitates the cells defense against heat shock (Lee
and Park, 1998). Our results clearly revealed that the
expression level of BmTPx from fat body cells of
silkworm larvae signicantly increased during the
exposure at low- and high-temperature conditions,
demonstrating that BmTPx in silkworm larvae was
induced by cold stress as well as heat stress. These
results point to an important another role of BmTPx,
protection against oxidative damage caused by tem-
perature stimuli.
We have also demonstrated the induction of BmTPx
in silkworm larvae during viral infection. It is likely that
the induction of BmTPx is related to a protective role
against oxidative damage caused by viral infection.
Previously, TPx enzymes were shown to be able to
ARTICLE IN PRESS
Fig. 6. Induction of BmTPx by viral infection. (A) Northern blot analysis of the BmTPx gene induced by viral infection. The 1-day-old 5th instar
silkworm larvae were mock-injected (lower panel) or injected with BmNPV of 2 10
5
PFU (upper panel). At 1, 2 and 3 days p.i., fat body cells from
B. mori larvae were harvested. Total RNA was isolated from the fat body of B. mori larva of each temperature treatment. The RNA was separated by
1.0% formaldehyde agarose gel electrophoresis, transferred on to a nylon membrane, and hybridized with radiolabelled 966 bp BmTPx cDNA.
Transcripts are indicated on the right side of the panel by arrow. (B) Relative mRNA levels of BmTPx induced by viral infection. The levels of
BmTPx mRNA were normalized to the expression at the 1-day-old 5th instar silkworm larvae as a control. Open and shaded bars represent control
and viral infection, respectively. Bars represent standard deviation. The data points reect the means of three independent experiments. (C) Western
blot analysis of the BmTPx induced by viral infection. The 1-day-old 5th instar silkworm larvae were mock-injected (lower panel) or injected with
BmNPV of 2 10
5
PFU (upper panel). At 1, 2 and 3 days p.i., fat body cells from B. mori larvae were harvested. The protein samples were prepared
from the fat body of B. mori larva of each temperature treatment. The samples were subjected to 12% SDS-PAGE, electroblotted and incubated with
antiserum to recombinant BmTPx. The arrow on the right of the panel indicates the 25 kDa BmTPx polypeptide. (D) Relative protein levels of
BmTPx induced by viral infection. The levels of BmTPx were normalized to the expression at the 1-day-old 5th instar silkworm larvae as a control.
Open and shaded bars represent control and viral infection, respectively. Bars represent standard deviation. The data points reect the means of three
independent experiments.
K.S. Lee et al. / Insect Biochemistry and Molecular Biology 35 (2005) 7384 82
remove intracellular H
2
O
2
generated in response to
various extracellular stimuli (Kang et al., 1998a). In
mammalian cells, H
2
O
2
production is correlated with
viral load (Toro et al., 1998; Elbim et al., 2001) and
antioxidant enzyme decreased the formation of viral
antigens (Messaoudi et al., 2002). The increased ROS
production by viral infection was associated with
changes in the expression of the antiapoptotic/antiox-
idant compounds Bcl-2 and thioredoxin (Elbim et al.,
2001). Furthermore, DmTPx has shown to have an
important role for protection against direct oxidative
damage leading to cell necrosis and prevention against
progression towards apoptosis (Radyuk et al., 2003).
Based on their TPx activities toward H
2
O
2
production
and viral infection, our data suggest that BmTPx could
act as a housekeeping type of peroxidase to regulate the
intracellular H
2
O
2
by extracellular stimuli such as low
and high temperature and viral infection.
Acknowledgments
This work was supported by a grant from BioGreen21
Program, Rural Development Administration, Republic
of Korea.
References
Bauer, H., Kanzok, S.M., Schirmer, R.H., 2002. Thioredoxin-2 but
not thioredoxin-1 is a substrate of thioredoxin peroxidase-1 from
Drosophila melanogaster. J. Biol. Chem. 277, 1745717463.
Chae, H.Z., Chung, S.J., Rhee, S.G., 1994a. Thioredoxin-dependent
peroxide reductase from yeast. J. Biol. Chem. 269, 2767027678.
Chae, H.Z., Robison, K., Poole, L.B., Church, G., Storz, G., Rhee,
S.G., 1994b. Cloning and sequencing of thiol-specic antioxidant
from mammalian brain: alkyl hydroperoxide reductase and thiol-
specic antioxidant dene a large family of antioxidant enzymes.
Proc. Natl. Acad. Sci. USA 91, 70177021.
Chae, H.Z., Uhm, T.B., Rhee, S.G., 1994c. Dimerization of thiol-
specic antioxidant and the essential role of cysteine 47. Proc. Natl.
Acad. Sci. USA 91, 70227026.
Davidson, J.F., Whyte, B., Bissinger, P.H., Schiestl, R.H., 1996.
oxidative stress is involved in heat-induced cell death in Sacchar-
omyces cerevisiae. Proc. Natl. Acad. Sci. USA 93, 51165121.
Demasi, A.P., Pereira, G.A., Netto, L.E., 2001. Cytosolic thioredoxin
peroxidase I is essential for the antioxidant defense of yeast with
dysfunctional mitochondria. FEBS Lett. 509, 430434.
Elbim, C., Pillet, S., Prevost, M.H., Preira, A., Girard, P.M., Rogine,
N., Hakim, J., Israel, N., Gougerot-Pocidalo, M.A., 2001. The role
of phagocytes in HIV-related oxidative stress. J. Clin. Virol. 20,
99109.
Fang, J., Li, J., 2001. Isolation and characterization of Aedes aegypti
thioredoxin peroxidase that contains two conserved cysteines.
GenBank Accession Number AAL37254.
Halliwell, B., Gutterridge, J.M.C., 1989. Production against Radical
Damage: Systems with Problems. Free Radicals in Biology and
Medicine, second ed. Clarendon Press, Oxford.
Hariari, P.M., Fuller, D.J., Gerner, E.W., 1989. Heat shock stimulates
polyamine oxidation by two distinct mechanisms in mammalian
cell cultures. Int. J. Radiat. Oncol. Biol. Phys. 16, 451457.
Je, Y.H., Chang, J.H., Kim, M.H., Roh, J.Y., Jin, B.R., OReilly,
D.R., 2001. The use of defective Bombyx mori nucleopolyhedro-
virus genomes maintained in Escherichia coli for the rapid
generation of occlusion-positive and occlusion-negative expression
vectors. Biotechnol. Lett. 23, 18091817.
Kang, S.W., Baines, I.C., Rhee, S.G., 1998a. Characterization of a
mammalian peroxiredoxin that contains one conserved cysteine.
J. Biol. Chem. 273, 63036311.
Kang, S.W., Chae, H.Z., Seo, M.S., Kim, K., Baines, I.C., Rhee, S.G.,
1998b. Mammalian peroxiredoxin isoforms can reduce hydrogen
peroxide generated in response to growth factors and tumor
necrosis factor-a. J. Biol. Chem. 273, 62976302.
Kim, K., Kim, I.H., Lee, K.Y., Rhee, S.G., Stadtman, E.R., 1988. The
isolation and purication of a specic protector protein which
inhibits enzyme inactivation by a thiol/Fe(III)/O
2
mixed-function
oxidation system. J. Biol. Chem. 263, 47044711.
Kim, S.R., Lee, K.S., Kim, I., Kang, S.W., Nho, S.K., Sohn, H.D.,
Jin, B.R., 2003. Molecular cloning of a cDNA encoding putative
calreticulin from the silkworm, Bombyx mori. Int. J. Ind. Entomol.
6, 9397.
Laemmli, U.K., 1970. Cleavage of structural proteins during assembly
of the head of bacteriophage T4. Nature 227, 680685.
Lee, S.M., Park, J.W., 1998. Thermosensitive phenotype of yeast
mutant lacking thioredoxin peroxidase. Arch. Biochem. Biophys.
359, 99106.
Lim, Y.S., Cha, M.K., Uhm, T.B., Park, J.W., Kim, K., Kim, I.H.,
1993. Removals of hydrogen peroxide and hydroxyl radical by
thiol-specic antioxidant protein as a possible role in vivo.
Biochem. Biophys. Res. Commun. 192, 273280.
Lindquist, S., 1986. The heat-shock response. Annu. Rev. Biochem. 55,
11511191.
McMaster, G.K., Carmichael, G.G., 1977. Analysis of single- and
double-stranded nucleic acids on polyacrylamide and agarose gels
by using glyoxal and acridine orange. Proc. Natl. Acad. Sci. USA
74, 48354838.
Messaoudi, K.E., Verheyden, A.M., Thiry, L., Fourez, S., Tasiaux, N.,
Bollen, A., Moguilevsky, N., 2002. Human recombinant myeloper-
oxidase antiviral activity on cytomegalovirus. J. Med. Virol. 66,
218223.
OReilly, D.R., Miller, L.K., Luckow, V.A., 1992. Baculovirus
Expression Vectors: A Laboratory Manual. W.H. Freeman &
Co., New York.
Park, S.G., Cha, M.K., Jeong, W., Kim, I.H., 2000. Distinct
physiological functions of thiol peroxidase isoenzymes in Sacchar-
omyces cerevisiae. J. Biol. Chem. 275, 57235732.
Radyuk, S.N., Klichko, V.I., Spinola, B., Sohal, R.S., Orr, W.C., 2001.
The peroxiredoxin gene family in Drosophila melanogaster. Free
Radical Biol. Med. 31, 10901100.
Radyuk, S.N., Sohal, R.S., Orr, W.C., 2003. Thioredoxin peroxidases
can foster cytoprotection or cell death in response to different
stressors: over- and under-expression of thioredoxin peroxidase in
Drosophila cells. Biochem. J. 371, 743752.
Rauen, U., Polzar, B., Stephan, H., Mannherz, H.G., de Groot, H.,
1999. Cold-induced apoptosis in cultured hepatocytes and liver
endothelial cells: mediation by reactive oxygen species. FASEB J.
13, 155168.
Swofford, D.L., 2000. PAUP*. Phylogenetic Analysis Using Parsi-
mony (*and Other Methods), Version 4, Sinauer Sunderland, MA.
Toro, F., Conesa, A., Garcia, A., Bianco, N.E., De Sanctis, J.B., 1998.
Increased peroxide production by polymorphonuclear cells of
chronic hepatitis C virus-infected patients. Clin. Immunol.
Immunopathol. 88, 169175.
Towbin, H., Staehelin, T., Gordon, J., 1979. Electrophoretic transfer
of proteins from polyacrylamide gels to nitrocellulose sheets:
procedure and some applications. Proc. Natl. Acad. Sci. USA 76,
43504354.
ARTICLE IN PRESS
K.S. Lee et al. / Insect Biochemistry and Molecular Biology 35 (2005) 7384 83
Vaughn, J.L., Goodwin, R.H., Thompkins, G.J., McCawley, P., 1977.
The establishment of two insect cell lines from the insect
Spodoptera frugiperda (Lepidoptera: Noctuidae). In Vitro 13,
213217.
Whiteld, C.W., Band, M.R., Bonaldo, M.F., Kumar, C.G., Liu, L.,
Pardinas, J.R., Robertson, H.M., Soares, M.B., Robinson, G.E.,
2002. Annotated expressed sequence tags and cDNA microarrays
for studies of brain and behavior in the honey bee. Genome Res.
12, 555566.
Zhou, Y., Kok, K.H., Chun, A.C., Wong, C.M., Wu, H.W., Lin,
M.C., Fung, P.C., Kung, H., Jin, D.Y., 2000. Mouse peroxiredoxin
V is a thioredoxin peroxidase that inhibits p53-induced apoptosis.
Biochem. Biophys. Res. Commun. 268, 921927.
ARTICLE IN PRESS
K.S. Lee et al. / Insect Biochemistry and Molecular Biology 35 (2005) 7384 84
Insect
Biochemistry
and
Molecular
Biology
Insect Biochemistry and Molecular Biology 35 (2005) 8591
Short communication
Proling the proteome complement of the secretion from
hypopharyngeal gland of Africanized nurse-honeybees
(Apis mellifera L.)
Keity Souza Santos
a,b
, Lucilene Delazari dos Santos
a,b
, Maria Anita Mendes
a,b
,
Bibiana Monson de Souza
a,b
, Osmar Malaspina
a,b
, Mario Sergio Palma
a,b,
a
Center of Study of Social Insects (CEIS) Department of Biology, Institute of Biosciences, Sao Paulo State University (UNESP), 13506900 Rio Claro,
SP, Brazil
b
Institute of Immunological Investigations (CNPq/MCT);CAT/CEPID-FAPESP
Received 27 July 2004; received in revised form 5 October 2004; accepted 7 October 2004
Abstract
The protein complement of the secretion from hypopharyngeal gland of nurse-bees (Apis mellifera L.) was partially identied by
using a combination of 2D-PAGE, peptide sequencing by MALDI-PSD/MS and a protein engine identication tool applied to the
honeybee genome. The proteins identied were compared to those proteins already identied in the proteome complement of the
royal jelly of the honey bees. The 2-D gel electrophoresis demonstrated this protein complement is constituted of 61 different
polypepides, from which 34 were identied as follows: 27 proteins belonged to MRJPs family, 5 proteins were related to the
metabolism of carbohydrates and to the oxido-reduction metabolism of energetic substrates, 1 protein was related to the
accumulation of iron in honeybee bodies and 1 protein may be a regulator of MRJP-1 oligomerization. The proteins directly
involved with the carbohydrates and energetic metabolisms were: alpha glucosidase, glucose oxidase and alpha amylase, whose are
members of the same family of enzymes, catalyzing the hydrolysis of the glucosidic linkages of starch; alcohol dehydrogenase and
aldehyde dehydrogenase, whose are constituents of the energetic metabolism. The results of the present manuscript support the
hypothesis that the most of these proteins are produced in the hypoharyngeal gland of nurse-bees and secreted into the RJ.
r 2004 Elsevier Ltd. All rights reserved.
Keywords: Africanized Apis mellifera; Royal jelly; Peptide mass ngerprint; Proteome; MALDI-TOF
1. Introduction
The honeybee (Apis mellifera L.) is a social insect,
living in colonies constituted of different castes: a queen,
workers and drones (Lercker et al., 1982; Palma, 1992).
The age-dependent role is one of the most notable
features of the workers in these colonies (Ohashi et al.,
1999). Young workers, known as nurse-bees take care of
the brood, by synthesizing and secreting many compo-
nents of the royal jelly (RJ), while the older workers
usually forage for nectar, converting it into honey
(Robinson, 1987). This process is biologically regulated
and known as age polyethism, which is paralleled by
physiological changes in certain organs of the worker
honeybees (Ohashi et al., 1997).
The RJ is believed to be synthesized both by the
mandibular and hypopharyngeal glands of nurse-hon-
eybees (Knecht and Kaatz, 1990; Lensky and Rakover,
1983). The hypoharyngeal gland is well developed in the
nurse-bees, but shrinks in the older workers, to adapt
ARTICLE IN PRESS
www.elsevier.com/locate/ibmb
0965-1748/$ - see front matter r 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ibmb.2004.10.003

Corresponding author. Center of Study of Social Insects (CEIS)


Department of Biology, Institute of Biosciences, Sa o Paulo State
University (UNESP), 13506900 Rio Claro, SP, Brazil. Tel.:
+55 19 3526 4163; fax: +55 19 3534 8523.
E-mail address: mspalma@rc.unesp.br (M.S. Palma).
the insect to forage and honey production (Ohashi et al.,
1999). Thus, it may be assumed that the gland
synthesizes different proteins according to age polyeth-
ism (Kubo et al., 1996).
Taking into account that the secretion of this gland
constitutes an important part of the RJ composition, it
is necessary to identify the biochemical contributions of
the hypopharyngeal gland to the composition of the RJ,
in order to get a better understanding the role of this
gland for the brood care.
The RJ contains proteins, free amino acids, fatty
acids, sugars, vitamins, some minerals (Palma, 1992)
and constitutes the principal food of the queen
honeybees; also it is part of the initial diet of honeybee
larvae (Moritz and Southwick, 1992). The protein
content represent up to 12% (m/m) of freshly harvested
RJ and the major components seem to constitute
members of the most known family of proteins of this
secretion, the major royal jelly proteins (MRJP) (Knecht
and Kaatz, 1990; Lensky and Rakover, 1983); this
family of proteins represent about 82% of the total
protein content of RJ (Schimitzova et al., 1998).
The complexity of the protein complement composi-
tion of the RJ and the absence of fundamental
sequential data about the most of genes involved with
the expression of these proteins, make their biochemical
and physiological characterizations difcult, which in
turn makes the understanding of their functions not
fully known. In addition to this, there has been few
substantial biochemical analysis of the proteins synthe-
sized by the hypopharyngeal gland.
Four major proteins (50, 56, 57 and 64 kDa) were
demonstrated to be selectively synthesized by the
hypopharyngeal gland and secreted as RJ proteins
(Hanes and Simu th, 1992; Kubo et al., 1996). These
results constitute strong evidences that the some
proteins found in RJ are in fact products of the
hypopharyngeal gland secretion.
Recently, the basic composition of the RJ was
analyzed by a proteomic approach, which identied
the presence of several different forms of MRJP-1 to
MRJP-5 and glucose oxidase, presenting heterogeneities
in terms of molecular weights and isoelectric points
(Sato et al., 2004). This study offered a wide view of the
composition of the proteome complement of the RJ,
opening the possibility to develop a similar investigation
with the hypopharyngeal gland, in order to compare
both compositions considering a large number of
proteins. Thus, the aim of the present study was to
obtain a proteome prole from the secretion of
hypopharyngeal gland of nurse-bees (A. mellifera) by
using the honeybee (HB) genome databank as reference
for proteins identication and to compare the proteins
identied with those already described in RJ, in order to
get a better understanding about the contribution of this
gland to the composition of the RJ.
2. Material and methods
2.1. Insects and the collection of the secretion from
hypopharyngeal-gland
Africanized honeybees (A. mellifera L.) kept in the
apiary of the Institute of Biosciences of Sa o Paulo
State University, at Rio Claro, southeast Brazil,
were used. Newly emerged workers were marked (day-
0) on their thorax using paint and introduced into a
normal queenright colony; marked nurse bees
were collected 7 days later when they were feeding
brood. In order to extract proteins of the hypophar-
yngeal gland secretion the bees were anesthetized by
using carbonic gas and the hypopharyngeal glands were
dissected under a binocular microscope. The glands
(500 glands/ml) were washed in buffered saline solution
(10 mM Tris-HCl pH 7.4, containing 13 mM NaCl,
5 mM KCl and 1 mM CaCl
2
), containing a mixture of
protease inhibitors: 1 mM phenylmethylsulfonyl uor-
ide, 0.1 mg/ml pepstatin and 100 mg/ml leupeptin as
described elsewhere (Kubo et al., 1996). The washing
extracts were centrifuged at 2000 g during 15 min
at 4 1C.
The supernatant was collected and the debris dis-
carded; the supernatant was then ltered through a
Millex membrane (0.45 mm pore diameter) at 4 1C,
lyophilized and dissolved in 1 ml urea/thiourea buffer
(2 M thiourea, 7 M urea, 4 w/v DTT, 2% v/v carrier
ampholytes pH 310). Protein content was determined
by using the modied method of Bradford (Sedmak and
Groosber, 1977).
2.2. D-polyacrylamide gel electrophoresis (PAGE)
2-D PAGE was performed by using 13 cm gels and
IPG strips with immobilized pH gradients (IPG_Dalt);
for the rst dimension (IPG-IEF) IPG gels were
cast with pH gradients of 310 (Go rg et al., 1999,
2000). Samples (25 ml) were applied by cup-loading
close to the anode and the focusings were performed by
using the following voltage gradient: 500 V/1 h, 1000 V/
1 h and 8000 V/2 h permitting the accumulation of
16 000 Vh. The runnings were performed at 20 1C,
under a current of 0.05 mA with a maximal power of
5.0 W. After equilibration of the IPG strips with
SDS buffer were applied to vertical SDS-PAGE
gels [12.5% (w/v) polyacrylamide and 0.8%(w/v)
Bis(N,N
0
-methylenebisacrylamide)]. The second dimen-
sion runnings were performed according to the
following program: 5 W/gel during 30 min and 17 W/
gel during 5 h, at 10 1C. After the electrophoretic
separation the gels were stained with Coomassie
Brilliant Blue (CBB) and stored at room temperature
(21 1C).
ARTICLE IN PRESS
K.S. Santos et al. / Insect Biochemistry and Molecular Biology 35 (2005) 8591 86
2.3. Image acquisition
The 2-D gels were scanned using a transparency mode
scanner, connected to a PC system, at 24-bit red-green-
blue colors and 300 dpi resolution for documentation.
Images were analyzed using Image Master 2D Database
software version 4.02 (Amersham Biosciences, Uppsala,
Sweden).
2.4. Tryptic digestion
The CBB stained spots were excised and destained for
30 min using 100 ml acetonitrile (50%) and 25 mM
(NH
4
)HCO
3
pH 8 (50%); dried for 20 min with
acetonitrile and incubated with 30 ml of 5 mM DTT
[in 25 mM (NH
4
)HCO
3
] for 30 min at 60 1C. The gel
pieces were incubated for 30 min with 30 ml of 55 mM
iodoacetamide [in 25 mM (NH
4
)HCO
3
] for 30 min at
room temperature, washed two times with 100 ml of
25 mM (NH
4
)HCO
3
during 20 min, followed by dehy-
dration with acetonitrile. Finally the spots were dried at
401C for 30 min using a Speed-Vac system.
Trypsin solution was prepared by dissolving 20 mg of
the enzyme in 100 ml 1 mM HCl. Before use, 150 ml of
5 mM (NH
4
)HCO
3
were added to the trypsin solution
(nal concentration 12.5 ng/ml); 10 ml of this solution
was pipetted on each dried protein spot and incubated
for 45 min at 0 1C, the supernatant was discarded to
minimize auto digestion of trypsin. Then 20 ml of 5 mM
(NH
4
)HCO
3
were added and the sample was incubated
for 18 h at 37 1C under soft shaking. To extract the
peptide fragments from the tryptic digests 50 ml of 50%
(v/v) acetonitrile [containing 0.5% (v/v) TFA] were
added and incubated for 30 min at room temperature.
Thereafter, 50 ml of 50% (v/v) formic acid were added
and incubated for 20 min at room temperature. After
each step the supernatants were pooled together and
dried using a Speed-Vac system.
2.5. Sample preparation
The digested sample was mixed with a matrix solution
[10 mg/ml a-cyano-4-hydroxycinnamic acid in 50% (v/v)
acetonitrile+0.1% (v/v) TFA]; 0.4 ml of this solution
were pipetted on the MALDI slide sampler and air-
dried.
2.6. Mass spectrometry
The masses of the tryptic peptides were determined by
using MALDI-TOF mass spectrometry. Ettan z
2
MAL-
DI-TOF mass spectrometer (Amersham Biosciences,
Uppsala, Sweden) equipped with UV nitrogen laser
(334 nm) and harmonic reectron, was used in the
positive ion mode at 20 kV, adjusted to perform delayed
extraction and low mass rejection. Calibration was
performed with peptide samples (angiotensin II, ACTH
1-39). Spectra were acquired by accumulation of 600
single shots.
A switch to fragmentation analysis in MALDI
instrument allowed the sequencing of some tryptic
peptides. Tryptic peptides were derivatized with the
CAF-kit (Amersham Biosciences) according to the
manufacturer instructions. Peaks were selected from
complex spectra using the timed ion gate; each pulse of
the laser provided a Post Source Decay (PSD) spectrum
over the entire m/z range, without the need for data
stitching. The spectra were acquired using an accelerat-
ing potential of 20 kV, and adjusting the pulsed laser to
1 shot per second; 300 shots per spectrum were
accumulated
2.7. Database queries
The sequences of some tryptic peptides from each
digested protein were used to interrogate the protein
engine search Blast of Honeybee Genome at the
National Center for Biotechnology Information (NCBI)
home page in Internet (URL:http://www.ncbi.nlm.nih.-
gov/genome/seq/AmeBlast.html). The search para-
meters were carried out by using the Build proteins
databases and the blastp program. A protein was
regarded as identied when the E-value calculated by
the search algorithm was lower than 1 10
3
.
3. Results and discussion
The preliminary protein sequencing by MALDI-
TOF-PSD/MS and the access to the honeybee genome
database through the protein engine search used in the
present study revealed suitable for the identication of
several proteins from the secretion of the hypophar-
yngeal gland of nurse-bees. Fig. 1 shows the 2-D
electrophoresis pattern of proteins CBB stained from
the hypopharyngeal gland. A total of 61 proteins were
detected after background subtraction, from which 34
were identied and represented in the Table 1. Among
the identied proteins, 27 belong to the MRJPs family, 5
proteins were related to the metabolism of carbohy-
drates and energy production, 1 protein was related to
iron accumulation and 1 protein was identied as a
regulator of MRJP-1 oligomerization (Table 1). About
27 protein spots were detected but not identied since
their analysis resulted in poor spectral quality due to a
noise background and/or low protein amount to get
reliable sequencing data .
Workers perform the most of tasks in the honeybee
hive, except egg laying. The tasks performed by
individual workers are susceptible to age-dependent
changes, known as age polyethism (Moritz and
Southwick, 1992). Thus, nurse bees take care of the
ARTICLE IN PRESS
K.S. Santos et al. / Insect Biochemistry and Molecular Biology 35 (2005) 8591 87
brood by producing the RJ, while the forager bees
(generally older than 10 days post-eclosion) forage for
nectar processing it into honey (Ohashi et al., 1997).
Apparently, the hypopharyngeal glands seems to exist in
two distinct differentiation states, characterized by
different patterns of protein expression as function of
the age polyethism (Ohashi et al., 1997). Thus, there is a
period of time during which there is a transition between
the roles of nursing and foraging for the workers,
causing an overlapping of functions for the organ. As
consequence, in this period of time the hypopharyngeal
gland may present an overlapped pattern of proteins
expression.
It was previously demonstrated that the nurse-bee
hypopharyngeal gland is involved with the production
of three major groups of RJ proteins, while the
hypopharyngeal gland of the forager bees seem to be
involved with the production of an a-glucosidase
(70 kDa) (Kubo et al., 1996) and selectively it expresses
mRNAs for amylase (57 kDa) and glucose oxidase
(85 kDa) (Ohashi et al., 1997; Ohashi et al., 1999).
Among the identied proteins in the present investi-
gation some aspects are specially important and must be
emphasized:
Six different forms of MRJP-1 were identied with
MW values changing from 48,810 to 59,995 Da and with
pI values from 4.23 to 5.50 (spots 21, 23, 24, 36, 38 and
41; Fig. 1), as also reported in a previous publication
which demonstrated that MRJP-1 may presents differ-
ent variant forms with MW around 55,000 Da in the pI
range from 4.50 to 5.20 (Hanes and Simu th, 1992).
However, the NCBI protein databank (http://
www.ncbi.nlm.nih.gov/) reveals three hypothetical gly-
cosylation sites for this protein, which may be used to
explain the observed MW heterogeneity.
In the present investigation were identied six
different forms of MRJP-1 in the proteome complement
of the hypopharyngeal gland of nurse-bees, while Sato et
al. (2004) identied only two different forms of this
protein in the RJ of Africanized honeybees and one
form in European honeybees. The experimental ap-
proach used in the present investigation did not permit
to know if the most forms of MRJP-1 are retained,
degraded or even metabolically used by the nurse-bees.
Eight different forms of MRJP-2 were identied
(spots 25, 26, 27, 28, 29, 31, 32 and 37; Fig. 1) with
MW from 50,673 to 59,995 Da in the pI range from 4.92
to 7.02 in the proteome complement of the secretion
from the hypopharyngeal gland of nurse-bees, while 15
and 12 different forms of this protein were observed in
the RJ from Africanized and European honeybees,
respectively (Sato et al., 2004). Bilikova et al, (1999)
reported eight isoforms of this protein with MW values
around 49 kDa in pI range from 7.5 to 8.5 in RJ from
European honeybees. Exploiting the NCBI protein
databank (http://www.ncbi.nlm.nih.gov/) is possible to
identify two potential glycosylation sites, suggesting the
possible existence of different forms of MRJP-2,
presenting different degrees of glycosylation. It must
be emphasized that all the forms of this protein observed
in the secretion of hypopharyngeal gland were also
present in the RJ as described by Sato et al, (2004).
Thus, the higher number of forms observed in RJ may
be due to structural changes suffered by this protein,
such as proteolysis, glycosylation/deglycosylation oc-
curred during RJ storage.
Five different forms of MRJP-3 were identied in the
proteome complement of the hypopharyngeal glands of
nurse-bees (spots 4,5,6,7 and 8; Fig. 1), in the MW range
from 80,590 to 87,000 Da and pI values from 7.05 to
8.04 (Fig. 1 and Table 1), while 10 and 24 different
forms of this protein were reported in RJ from
Africanized and European honeybees, respectively (Sato
et al., 2004). Thus, taking into account that all forms
identied in the hypopharyngeal glands were also
present in the RJ and considering that the MRJP-3
may suffer partial proteolysis during the RJ storage
(Kimura et al., 1995), it may be speculated that this
protein is probably produced by the hypopharyngeal
gland, secreted into the RJ, in which it may suffer partial
degradation during the storage, originating the exceed-
ing number of forms observed in RJ.
No form of MRJP-4 was identied in the proteomic
analysis of the secretion of the hypopharyngeal gland,
by using the honeybee genome as reference for protein
identication; however the common sequence of a
tryptic peptide obtained from the spots 33 and 39 was
ARTICLE IN PRESS
Fig. 1. 2-D PAGE of the proteins complement of the hypopharyngeal
gland of nurse-bee, IPG 3-10, 12.5% T, showing all the CBB stained
proteins.
K.S. Santos et al. / Insect Biochemistry and Molecular Biology 35 (2005) 8591 88
observed in the C-terminal region of the protein RJP57-
2 (in NCBI protein database; accession number
Q17061), which is equivalent to MRJP-4. In the RJ of
Africanized and European honeybees were observed 5
and 2 different forms of this protein, respectively (Sato
et al., 2004).
Three different forms of MRJP-5 were identied in
the protein complement of the secretion from the
hypopharyngeal gland of nurse-bees, presenting MW
from 79,075 to 79,471 Da, and pI values from 6.34 to
6.80 (spots 9, 10 and 11; Fig. 1 and Table 1). In the RJ
from Africanized and European honeybees were identi-
ed respectively, 7 and 4 different forms of this protein
(Sato et al., 2004). All the spots corresponding to
MRJP-5 were also detected in the RJ of Africanized
honeybees with similar MW and pI values. In spite to
these multiple forms the honeybee genome has a single
copy of the gene of MRJP-5, which suggest that this
protein may be susceptible to post-translational mod-
ications.
A single form of the proteins MRJP-6, -7 and -8
(spots 40, 35 and 42, respectively; Fig. 1 and Table 1)
were identied in the present proteomic analysis in a
single form. According to Sato et al, (2004) the MRJP-6,
-7 and -8 were not observed in the proteome comple-
ment of the RJ; however, there are a series of protein
spots still not identied in RJ proteome, which could
correspond to these proteins.
The second largest group of the identied components
in RJ protein complement are those related to the
ARTICLE IN PRESS
Table 1
Proteins identied in the hypopharyngeal gland of nurse-bee
Spot number Sequences of tryptic fragments used in protein ID Protein identication in HB genome database Accession number E-value
21 VGDGGPLLQPYPDWSFAK Major royal jelly protein 1 XP_393380.1 2 10
6
23 VGDGGPLLQPYPDWSFAK Major royal jelly protein 1 XP_393380.1 2 10
6
24 VGDGGPLLQPYPDWSFAK Major royal jelly protein 1 XP_393380.1 2 10
6
36 VGDGGPLLQPYPDWSFAK Major royal jelly protein 1 XP_393380.1 2 10
6
38 VGDGGPLLQPYPDWSFAK Major royal jelly protein 1 XP_393380.1 2 10
6
41 VGDGGPLLQPYPDWSFAK Major royal jelly protein 1 XP_393380.1 2 10
6
25 SLNVIHEWKYFDYDFGSEER Major royal jelly protein 2 XP_393061.1 2 10
7
26 SLNVIHEWKYFDYDFGSEER Major royal jelly protein 2 XP_393061.1 2 10
7
27 SLNVIHEWKYFDYDFGSEER Major royal jelly protein 2 XP_393061.1 2 10
7
28 SLNVIHEWKYFDYDFGSEER Major royal jelly protein 2 XP_393061.1 2 10
7
29 SLNVIHEWKYFDYDFGSEER Major royal jelly protein 2 XP_393061.1 2 10
7
31 SLNVIHEWKYFDYDFGSEER Major royal jelly protein 2 XP_393061.1 2 10
7
32 SLNVIHEWKYFDYDFGSEER Major royal jelly protein 2 XP_393061.1 2 10
7
37 SLNVIHEWKYFDYDFGSEER Major royal jelly protein 2 XP_393061.1 2 10
7
4 WLLLVVCLGIACQDVTSAAVNHQRK Major royal jelly protein 3 XP_391893.1 2 10
9
5 WLLLVVCLGIACQDVTSAAVNHQRK Major royal jelly protein 3 XP_391893.1 2 10
9
6 WLLLVVCLGIACQDVTSAAVNHQRK Major royal jelly protein 3 XP_391893.1 2 10
9
7 WLLLVVCLGIACQDVTSAAVNHQRK Major royal jelly protein 3 XP_391893.1 2 10
9
8 WLLLVVCLGIACQDVTSAAVNHQRK Major royal jelly protein 3 XP_391893.1 2 10
9
33 LSSHTLNHNSDKMSDQQENLTLK n.i.
a

39 LSSHTLNHNSDKMSDQQENLTLK n.i.
a

9 LANSMNVIHEWKYLDYDFGSDER Major royal jelly protein 5 XP_396121.1 5 10
9
10 LANSMNVIHEWKYLDYDFGSDER Major royal jelly protein 5 XP_396121.1 5 10
9
11 LANSMNVIHEWKYLDYDFGSDER Major royal jelly protein 5 XP_396121.1 5 10
9
40 SSKNLEHSMNVIHEWK Major royal jelly protein 6 XP_393001.1 1 10
4
35 GDGLIVYQNSDDSFHR Major royal jelly protein 7 XP_393063.1 2 10
4
42 LWVLDNGISGETSVCPSQIVVFDLK Major royal jelly protein 8 XP_393049.1 2 10
9
20 PLPENLKEDL IVYQVYPRSF alpha- glucosidase XP_392790.1 1 10
6
17 GKNLGGTSSHNGMMYTR glucose oxidase XP_392386.1 4 10
5
30 ANTYNFDYPQVPYTVKNFHPR alpha-amylase XP_391983.1 2 10
8
50 HFVLAFSRSLARYYNNTGIR alcohol dehydrogenase XP_392596.1 2 10
6
16 RKNLKPLGQLVSLEMG aldehyde dehydrogenase XP_394614.1 2 10
3
53 AINDQINFELHASYIYLSMAYYFDR ferritin-like XP_392201.1 2 10
9
60 NISKIQIEPSFLKAIPNSTKIKMSK apisimin XP_393206.1 3 10
8
a
In spite this sequence did not permit the identication of the proteins in HB genome database, these sequences were observed in within the
primary sequence of the protein RJP57-2 (in NCBI protein database; accession number Q17061), which is equivalent to MRJP-4.
K.S. Santos et al. / Insect Biochemistry and Molecular Biology 35 (2005) 8591 89
metabolism of carbohydrates. Thus, the proteins identi-
ed in this group are:
Glucose-oxidase, alpha-glucosidase and alpha-amy-
lase were identied in the spots 17, 20 and 30,
respectively (Fig. 1 and Table 1). Sato et al, (2004)
identied ve different forms of glucose oxidase enzyme
in the RJ from Africanized honeybees, with MW and pI
values higher than those observed in the present
investigation.
Aldehyde dehydrogenase and alcohol dehydrogenase
were identied in the HB genome the spots 16 and 50
(Fig. 1 and Table 1). These enzymes are related to the
energetic metabolism, promoting the formation of
reduced coenzymes from their oxidated forms (NAD
+
or NADP
+
).
At rst these results suggest that the secretion from
hypoharyngeal gland of seven days old nurse-bees may
be starting to produce proteins related to the processing
of nectar into honey and also proteins related to the
catabolism of glucose to produce energya behavior
typical of workers bees, reecting the rst molecular
signs of the age-dependent role changes. However, still
must be considered that these enzymes may have
originated from the disruption of some cells of
hypoharyngeal glands during the manipulation of this
material, representing intracellular proteins (not exo-
crine products) which contaminated the samples, being
detected and identied together the secretion of the
gland.
A protein ferritin-like was identied in the spot 53
(Fig. 1), presenting pI 4.21 and MW: 24,767 Da (Table
1). Ferritin is an intracellular molecule that stores iron in
a soluble and non-toxic form, readily available for use.
The peptide component apisimin was identied in the
spot number 60 (Fig. 1 and Table 1) presenting a MW
value higher than the expected value from the HB
genome suggesting that it may be produced in a
precursor form. This peptide seems to be involved with
the oligomerization of MRJP-1 (Bilikova et al., 2002).
Apparently, this polypeptide may be produced by the
hypopharyngeal gland of nurse-bees and secreted into
the RJ, in spite it was not detected in the RJ by Sato et
al, (2004); however, these authors also observed a wide
heterogeneity of composition from one sample to
another one.
Among the 27 not identied proteins, some spots
seem to be characteristic of the hypopharyngeal gland,
since apparently they were not detected in the RJ by
Sato et al, (2004); these proteins correspond to the spots
1,2,3, 12,14, 15,18,19 and 22. The proteins of all these
spots generated a series spectral data which apparently
did not permit their identication, because the amount
of material was very reduced, resulting poor MALDI-
PSD/MS spectra.
In summary, the CBB-stained 2-D gel electrophoresis
demonstrated that the proteome complement of the
secretion from the hypopharyngeal gland of 7 days old
nurse-bees contains at least 61 different proteins, from
which 34 were identied through the use of peptide
sequencing by MALDI-PSD/MS combined with Blast
tool to search for the protein identication in the
honeybee genome. When the identied proteins were
compared to those previously identied in RJ, it
becomes clear that the MRJPs are produced in the
hypoharyngeal gland and secreted into the RJ, where
some of these proteins may suffer glycosylation and/or
partial proteolysis, generating new isoforms of these
proteins in stored RJ. It was also demonstrated that
some proteins characteristic of the hypopharyngeal
gland of worker bees are starting to be synthesized at
the seventh day of age, probably as consequence of age-
controlled biochemical and physiological changes to
prepare the transition from nurse to worker role of the
honeybees.
Although the most likeliest candidates in the database
to match the spots were found, these identications need
to be physiologically conrmed in order to clearly
understand the functional role of these proteins in queen
honeybee development.
Acknowledgments
This work was supported by the grants from FAPESP
and CNPq. M.A.M is Post-Doc fellow from FAPESP
(proc. 01/05060-4), KSS was fellow from CNPq. M.S.P.
(proc., 300337/2003-50) and O.M. .are researchers of the
Brazilian Council for Scientic and Technological
Development.
References
Bilikova , K., Klaudiny, J., Simu th, J., 1999. Characterization of the
basic major royal jelly protein MRJP2 of honeybee (Apis mellifera
L.) and its preparation by heterologous expession in. E. coli.
Biologia (Bratislava) 54, 733739.
Bilikova , K., Hanes, J., Nordhoff, E., Saenger, W., Klaudiny, J.,
Simuth, J., 2002. Apisimin, a new serine-valine-rich peptide from
honeybee (Apis mellifera L.) Royal Jelly: purication and
molecular characterization. FEBS Lett 528, 125129.
Go rg, A., Obermeier, C., Boguth, G., Weiss, W., 1999. Recent
developments in two-dimensional gel electrophoresis with immo-
bilized pH gradients: wide pH gradients up to pH 12, longer
separation distances and simplied procedures. Electrophoresis 20,
712717.
Go rg, A., Obermaier, C., Boguth, G., Harder, A., Scheibe, B.,
Wildgruber, R., Weiss, W., 2000. The current state of two-
dimensional electrophoresis with immobilized pH gradients.
Electrophoresis 21, 10371053.
Hanes, J., Simu th, J., 1992. Identication and partial characterization
of the major royal jelly protein of the honey bee (Apis mellifera L.).
J. Apic. Res. 31, 2226.
Kimura, M., Washino, N., Yonekura, M., 1995. N-linked sugar chains
of 350 kDa royal jelly glycoprotein. Biosci. Biotechnol. Biochem.
59, 507509.
ARTICLE IN PRESS
K.S. Santos et al. / Insect Biochemistry and Molecular Biology 35 (2005) 8591 90
Knecht, D., Kaatz, H.H., 1990. Patterns of larval food production by
hypopharyngeal glands in adult worker honeybees. Apidologie 21,
457468.
Kubo, T., Sasaki, M., Namura, J., Sasagawa, H., Ohashi, K.,
Takeuchi, H., Natori, S., 1996. Change in the expression of
hypopharingeal-gland proteins of the worker honeybees (Apis
mellifera L.) with the age and/or role. J. Biochem. 119, 291295.
Lensky, Y., Rakover, Y., 1983. Separate protein body compartments
of the workers honeybeee (Apis mellifera L.). Comp. Biochem.
Physiol. 75, 607615.
Lercker, G., Capella, P., Conte, L.S., Ruini, F., Giordani, G., 1982.
Components of royal jelly II. The lipid fraction, hydrocarbons and
sterols. J. Apic. Res. 21, 178184.
Moritz, R.F.A., Southwick, E.E., 1992. Bees a Superorganisms. An
Evolutionary reality. Springer, Berlin, Heidelberg.
Ohashi, K., Natori, S., Kubo, T., 1997. Change in the mode of gene
expression of the hypopharingeal gland cells with an age-dependent
role change of the worker honeybee Apis mellifera L. Eur. J.
Biochem. 249, 797802.
Ohashi, K., Natori, S., Kubo, T., 1999. Expression of amylase and
glucose oxidase in the hypopharyngeal gland with an age-
dependent role change of the worker honeybee (Apis mellifera
L.). Eur. J. Biochem. 265, 127133.
Palma, M.S., 1992. Composition of freshly harvested Brazilian royal
jelly identication of carbohydrates from the sugar fraction.
J. Apic. Res. 31, 4244.
Robinson, G.E., 1987. Hormonal regulation of age polyethism in the
honeybee, Apis mellifera. In: Menzel, R., Mercer, R. (Eds.),
Nurobiology and Behavior of Honeybees. Springer, Berlin, pp.
266279.
Sato, O., Kunikata, T., Kohno, K., Iwaki, K., Ikeda, M., Kurimoto,
M., 2004. Charcaterizaton of Royal Jelly Proteins in both
Africanized and European Honeybees (Apis mellifera) by Two
Dimensional Gel Electrophoresis. J. Agric. Food Chem. 52,
1520.
Schimitzova , J., Klaudiny, J., Albert, S., Schroder, W., Schreckengost,
W., Hanes, J., Ju dova , J., Simu th, J., 1998. A new family of major
royal jelly proteins of Honeybee Apis mellifera. L. Cell. Mol. Life
Sci. 54, 10201030.
Sedmak, J.J., Groosber, G.S.E., 1977. A rapid, sensitive and versatile
assay for protein using Coomassie Brilliant Blue G-250. Anal.
Biochem. 79, 544552.
ARTICLE IN PRESS
K.S. Santos et al. / Insect Biochemistry and Molecular Biology 35 (2005) 8591 91