JOURNAL OF BACTERIOLOGY, June 1988, p. 2659-2667 Vol. 170, No.

6
0021-9193/88/062659-09$02.00/0
Copyright X 1988, American Society for Microbiology

The Agrobacterium tumefaciens virE2 Gene Product Is a

Single-Stranded-DNA-Binding Protein That Associates with T-DNA
PETER J. CHRISTIE, JOHN E. WARD, STEPHEN C. WINANS, AND EUGENE W. NESTER*
Department of Microbiology, University of Washington, Seattle, Washington 98195
Received 26 January 1988/Accepted 17 March 1988

Agrobacterium tumefaciens transfers T-DNA into the plant genome by a process mediated by Ti plasmid-
encoded vir genes. Cleavage at T-DNA border sequences by the VirD endonuclease generates linear,
single-stranded T-DNA molecules. In the work described in this report, we used electrophoretic mobility shift
assays to show that the purified virE2 gene product binds to single-stranded DNA. VirE2 protein associates with
T-DNA as shown by immunoprecipitation studies with VirE2-specific antiserum. The VirE2 protein was
detected primarily in the cytoplasm, but also in the inner and outer membrane and periplasmic fractions.
Virulence of a virE2 mutant was restored by mixed infection with strains carrying an intact vir region, but not
with virA, virB, virD, virE, or virG mutants or chvA, chvB, or exoC mutants. We propose that the VirE2 protein
is involved in the processing of T-DNA and in T-strand protection during transfer to the plant cell.

Agrobacterium tumefaciens causes crown gall disease in
dicotyledenous plants by transferring a segment of DNA
(T-DNA) from a resident (Ti) tumor-inducing plasmid into
plant cells, where the DNA becomes covalently integrated
into the plant nuclear genome (for recent reviews, see
references 28a, 35, and 47; see also A. Powell and M. P.
Gordon, in S. Marcuo, ed., Biochemistry of Plants: a
Comprehensive Treatise, vol. II, in press). The T-DNA then
directs the synthesis of the plant growth hormones auxin and
cytokinin, resulting in unregulated proliferation of plant
tissue (35). Transfer of T-DNA is mediated by several
trans-acting virulence loci, including at least three chromo-
somal loci involved in the initial attachment of the bacterium
to the plant cell (8, 11, 12) and loci within the vir region, a ca.
35-kilobase (kb) region of the Ti plasmid outside of the
T-DNA (15, 27, 43). Mutations in six operons within the vir
region either abolish virulence (virA, virB, virD, and virG) or
result in a sharply attenuated virulence phenotype (virC and
virE) (21, 27, 43, 55).
Recently, workers in several laboratories have begun to
elucidate the roles of vir proteins in the transfer process.
Specific plant signals (4, 42) and the products of virA and
virG are required for induction of transcription of the other
vir operons (46, 52). Products of the first two genes of the
virD operon make up an endonuclease which recognizes and
cleaves at specific sites within 24-base-pair (bp) border
sequences that functionally define the boundaries of T-DNA
(23, 45, 49, 56). A strand-specific RNase protection assay (1,
56) and Southern hybridization studies (44, 45) have shown
that a product of the cleavage event is a linear, single-
stranded T-DNA molecule (T-strand) that corresponds to the
bottom strand of T-DNA. One model for T-DNA processing
is that initial border sequence cleavage by the virD endonu-
clease is followed by a replication event, which results in
replacement synthesis of the bottom strand of T-DNA and
the concomitant displacement of a T-strand molecule (1, 44).
However, the structure of T-DNA molecules transferred
into plant cells is not yet clearly established, since double-
stranded linear (23, 49) and circular (2, 29, 54) T-DNA
molecules also have been identified as products of the
cleavage reaction.
* Corresponding author.
2659
The roles of the other vir genes in the processing and
transfer of T-strands have not been elucidated. The virE
operon is of particular interest in that strains containing virE
mutations can be complemented during mixed infections by
using a helper strain of A. tumefaciens. Conceivably, the
virE product(s) synthesized in one bacterium are exported
and utilized during infection by a second strain (14, 39). The
operon consists of two open reading frames coding for
proteins of ca. 7 kilodaltons (virEl) and 60.5 kilodaltons
(virE2) (51), both of which are required for virulence (34).
The unusual complementation observed for virE mutants but
not for any other vir mutants suggests that an understanding
of the functions of the virE gene products might provide
insights into the mechanism of transfer of the T-DNA. In this
report we present evidence that the virE2 gene product is a
single-stranded-DNA-binding protein that associates with
T-DNA. The significance of this DNA-binding property is
discussed in relation to our current understanding of the
T-DNA transfer process.
MATERIALS AND METHODS
Bacterial strains, plasmids, and growth conditions. A348 is
A. tumefaciens A136 containing pTiA6NC (16). Mutations in
the vir region of pTiA6NC were generated by transposon
mutagenesis with the genetically engineered transposon
Tn3HoHol (43). Plasmids designated pTi contain
Tn3HoHol insertions introduced by marker exchange into
vir loci of pTiA6NC (43), while plasmids designated pSM
were constructed by cloning vir loci into cosmid pVK102
(28, 43). A derivative of A348 carrying an insertion in the
pinF (plant inducible) locus (pTi219) was used as the source
of VirE2 protein, since the virulence phenotype of this strain
is identical to that of A348 and the lacZ::pinF translational
fusion provides an assay for induction of the vir genes (43).
This strain also contained pToK9, which carries the 3' end of
virB and virG from pTiBo542, a region that enhances expres-
sion of the vir genes, cloned into pVK102 (24). Plasmid
pTi318 carries a Tn3HoHol insertion in virE2 that eliminates
gene expression, and pTi358 carries an insertion that results
in a lacZ::virE2 translational fusion (41, 43; see Fig. 1).
Plasmid pSW153 consists of a ca. 2.0-kb SphI-BamHI re-
striction fragment from pTiA6NC, encoding most of the
virEl coding sequence and the entire virE2 coding sequence,
2660 CHRISTIE ET AL.
cloned into the pUC19 polylinker site (51). This plasmid
contains a transcriptional fusion between the lacZ promoter
and virE2, resulting in isopropyl-,-D-thiogalactopyranoside
(IPTG)- inducible expression of virE2, and may contain a
lacZ::virEl translational fusion, although the fusion product
has not been detected (51). Plasmid pCGN394 contains the
virE operon cloned into pUC18, and pCGN1513 is a deriva-
tive of pCGN394 in which the virEl coding sequence was
disrupted by an in-frame deletion (34). Plasmid pNW34C-8-1
consists of the EcoRI 8 fragment (7.5 kb) from the T-DNA of
pTiA6NC cloned into pBR322 (16). Plasmid pJW211 consists
of the BamHI 14 fragment (5.14 kb) from the virB region of
pTiA6NC cloned into pMY1133, a derivative of pUC18 (J.
Ward, personal commnunication). A. tumefaciens strains
were maintained on AB minimal salts medium (10), and
Escherichia coli strains were maintained on LB medium (33)
with the addition of appropriate antibiotics as previously
described (56). A. tumefaciens vir gene expression was
induced by diluting a mid-exponential-phase cell culture to
an optical density at 550 nm of 0.1 in AB medium buffered at
pH 5.5 with 2-(N-morpholino)ethanesulfonic acid (MES),
followed by vigorous aeration at 28°C for 18 h in the
presence of acetosyringone (100 VxM) (42).
Protein analysis, immunological techniques, and cell frac-
tionation. Proteins were resolved by sodium dodecyl sulfate-
polyacrylamide gel electrophoresis (SDS-PAGE) through
10.0 and 12.5% gels and visualized by staining with Coomas-
sie brilliant blue G in 10% methanol-10% glacial acetic acid
(30). The relative amount of the virE2 gene product was
determined by scanning densitometry. Protein concentra-
tions were determined by the method of Bradford (5).
Antibodies to the overproduced VirE2 protein from E. coli
JM1O1(pSW153) were raised as follows. Extracts prepared
by French press disruption of IPTG-induced cells were
centrifuged at 30,000 x g for 30 min, which pelleted large
amounts of insoluble VirE2 protein. The pellet was washed
repeatedly with 1 M NaCl to solubilize and further remove
contaminating proteins from the VirE2-containing precipi-
tate. This insoluble material was boiled in electrophoresis
buffer and subjected to preparative-scale SDS-PAGE (30).
VirE2 protein then was eluted from the gels (19) and injected
into New Zealand White rabbits to raise VirE2-specific
antiserum. The VirE2-specific antiserum and alkaline phos-
phatase-conjugated anti-rabbit immunoglobulin G purchased
from Bio-Rad Laboratories were used to visualize VirE2
protein by immunostaining (48). Immune complexes were
collected from crude cell extracts by immunoprecipitation
with preimmune or VirE2-specific antiserum and protein
A-bearing Staphylococcus aureus cells (IgGsorb; The En-
zyme Center) essentially as described by Gaylo et al. (17).
Acetosyringone-induced cells from a 500-ml culture (optical
density at 550 nm of 0.8) were suspended in 5 ml of Z buffer
(25 mM HEPES [N-2-hydroxyethylpiperazine-N'-2-ethane-
sulfonic acid, potassium salt] 12.5 mM MgCl2, 1 mM dithio-
threitol, 20% [vol/vol] glycerol, 0.1% [vol/vol] Nonidet P-40)
(25) and disrupted by French press treatment. The crude
extract was centrifuged at 20,000 x g for 30 min, and the
resulting supernatant was centrifuged twice to remove insol-
uble cell debris. Then, 5 ,Ju of preimmune antiserum or
VirE2-specific antiserum was added to 500 pul of the super-
natant (representing material from ca. 7.5 x 1010 cells), and
the mixture was incubated on ice for 3 h. Five microliters of
a 10% (wt/vol) protein A-bearing Staphylococcus cell sus-
pension was added, and the reaction mixture was gently
shaken at 4°C for 3 h. Staphylococcus cells were pelleted by
centrifugation at 12,000 x g for 60 s in a Microfuge
J. BACTERIOL.
(Beckman Instruments, Inc.), washed five times with 500 ,ul
of NET buffer (50 mM Tris [pH 7.4], 150 mM NaCl, 5 mM
EDTA, 0.1% Triton X-100) (17), and suspended in 50 ,ul of
NET buffer. Samples (10 ,ul) were examined by immu-
nostaining (48) and Southern hybridization analysis (33).
Cells were separated into cytoplasmic, inner membrane, and
outer membrane fractions as described previously (31). The
fractionation quality was assessed by assaying for NADH
oxidase, an inner membrane marker (38), and malate dehy-
drogenase, a cytoplasmic marker (26), and by determining
A280 values. The periplasmic fraction was obtained by os-
motic shock as described by Maagd and Lugtenberg (32).
Purification of the virE2 gene product. The VirE2 protein
was purified from E. coli JM1O1(pSW153) by a purification
procedure similar to that used for other overproduced pro-
teins that precipitate in cells (6). Cells from a 2-liter culture
induced with IPTG were suspended in Z buffer and disrupted
by French press treatment. The suspension was centrifuged
at 10,000 x g for 30 min, and the resulting pellet was washed
four times in Z buffer containing 1 M NaCl and then
suspended in S ml of 5 M urea in Z buffer. Remaining
insoluble material was removed by centrifugation at 20,000
x g for 30 min. The supernatant then was dialyzed in Z
buffer or passed through a Sephadex G-25M column (Phar-
macia, Inc.) to remove the urea. Fractions which contained
VirE2 protein as determined by immunoblot analysis were
pooled and passed through a denatured calf thymus DNA-
cellulose column (Sigma Chemical Co.) with a 5-ml bed
volume. The column was washed with at least 10 column
volumes of Z buffer, and bound material then was step eluted
with 0.1, 0.5, 1.0, and 2.5 M NaCl in Z buffer. Chromatog-
raphy fractions were electrophoresed through SDS-poly-
acrylamide gels and examined by Coomassie blue staining
and immunostaining.
The VirE2 protein was purified from a 5-liter culture of
acetosyringone-induced A. tumefaciens A136(pTi2l9, pTo
K9) cells grown with vigorous aeration at 28°C for 18 h (final
optical density at 550 nm of 1.0). The cell pellet was
suspended in 50 ml of Z buffer, cells were broken in a French
press, and the resulting suspension was centrifuged at 20,000
x g for 1 h. The supernatant then was passed through a
denatured calf thymus DNA-cellulose column (bed volume,
10 ml), and adsorbed material was step eluted with NaCl in
Z buffer (see above). Fractions which contained VirE2
protein as measured by immunoblot analysis were pooled
and passed through an immunoaffinity column (bed volume,
10 ml) consisting of VirE2-specific antiserum coupled to
CnBr-activated Sepharose 4B beads (Pharmacia). The col-
umn was washed extensively with Z buffer, and adsorbed
material was step eluted with NaCl in Z buffer. VirE2-
containing fractions were concentrated, and salt was re-
moved with Centricon 10 microconcentrator filters (Amicon
Corp.).
DNA manipulations and preparation of oligonucleotides.
Plasmid DNA was isolated and manipulated by procedures
described by Maniatis et al. (33). Single-stranded oligonucle-
otides of 30 bases were synthesized by the phosphoramidate
method on a Biosearch-8600 DNA synthesizer. The eight
oligonucleotides used in these studies were as follows:
CAGGAAACAGCTATGATTGTACATCCTTCA and the
complementary sequence, TGAAGGATGTACAATCATA
GCTGTTTCCTG, which represent random sequences; GA
ATAAGTCGCTGTGTATGTTITlTTTGGTT and the com-
plementary sequence, AACCAAACAAACATACACAGCG
ACTTATTC, which contains the conserved overdrive se-
quence (underlined) shown previously to reside 17 base pairs
VOL. 170, 1988

from right border sequences of several Ti plasmids (40);

GCTCAAATTACAACGGTATATATCCTGCCA and the
complementary sequence, TGGCAGGATATATACCGTTG
TAATTTGAGC, which contains the 24-base sequence that
functionally define the right border B of T-DNA from the
octopine plasmid pTiA6NC (3, 53); a poly(dA-dG) sequence;
and a poly(dT-dC) sequence. Blunt-ended, double-stranded
molecules were generated by mixing equimolar amounts of
the corresponding upper and lower strands and incubating
them at 75°C for 10 min and then at 37°C for 30 min and
analyzed by electrophoresis through 20% acrylamide gels
(acrylamide/bisacrylamide ratio, 38:2). Oligonucleotides
were end labeled with T4 polynucleotide kinase as described
by Maniatis et al. (33).
Electrophoresis mobility shift assay. Binding assays were
conducted essentially as described by Hendrickson and
Schlief (20). Reaction mixtures consisted of 20 ,lI of Z buffer,
10 nM single- or double-stranded, end-labeled oligonucleo-
tides, and 100 nM VirE2 protein. Since the exposure of
VirE2 protein to urea (during purification from E. coli) or
high salt concentrations (during purification from A. tume-
faciens) may have reduced the ability of VirE2 to bind DNA,
a high ratio of protein to DNA molecules was used in these
initial studies. Experiments in progress are aimed at quanti-
tating the loss of DNA-binding activity resulting from dena-
turation of VirE2 protein during purification. The mixture
was placed on ice for 30 min, 2 ,ul of loading buffer (25%
glycerol, 0.1% bromophenol blue, 0.1% xylene-cyanol) was
added, and the DNA-protein complexes then were resolved
in a vertical 5 to 20% linear gradient of polyacrylamide
(acrylamide/bisacrylamide ratio, 38:2) in TBE buffer (90 mM
Tris base, 90 mM H2Bo3, 2.5 mM EDTA). Interestingly, in
these gels the double-stranded oligonucleotides migrate
slower than single-stranded oligonucleotides. The gel was
prerun for 3 h, samples were loaded with the current on, and
electrophoresis was carried out at 8 V/cm for 6 h. The gel
was exposed to X-ray film at -70°C with an intensifying
screen.
Virulence assays. Strains were tested for their virulence by
inoculating Kalanchoe diagremontiana leaves as described
by Garfinkel and Nester (15). For coinfection studies, a
_:5.St,-l2b43;iA+'F_
1), but also with monoclonal antibodies to ,B-galactosidase
(41).

S.
Purification of VirE2 protein from E. coli(pSW153) is
described in detail in Materials and Methods and is shown in
Fig. 1A and B. The overproduced VirE2 protein was detect-
able in total extracts from E. coli(pSW153) (Fig. 1A, lane 7).
Although a small amount of VirE2 protein, detectable by
Western immunoblot analysis, was present in the supema-
tant after centrifugation of extracts from French press-
disrupted cells, the majority of the protein was in the pellet
fraction (Fig. 1A, lane 8). Urea solubilization of the pellet
followed by single-stranded-DNA-cellulose column chroma-

_,.. _ s_
__

_
_
i_,
: ......

-A
1

..5*-f.t
.:X.

.4
v
i.

__

y
2

2
,
...

_
}

.z

_W
:: ...
3

.:: .:.

3
w

-
::
VirE2 ssDNA-BINDING PROTEIN

_
4

.1-

vw
5
5

4IM-1-0
0-1

6
6

&6&

7
7

8
8

40M
9

9
*

*
M

} 4 -66
-45

_0A

ss,aLa
-11

-111
-
2661

-24

- 18

- 15

-
A
1

66
45

culture of the virE2 mutant strain A136(pTi358) was mixed

on an AB minimal medium plate with each of the vir, chv,
and exo mutants listed in Table 1. Each test mixture was
inoculated on the left side of a Kalanchoe leaf, with the
corresponding "rescuing" strain inoculated on the right side
as a control. Experiments were performed in triplicate, and
at least two separate leaves were inoculated per strain.
RESULTS
Purification of the VirE2 protein from E. coli and A.
tumefaciens. Antibodies to the VirE2 protein were used to
monitor purification of the protein from both E. coli and A.
tumefaciens. Immunoblot analyses (Fig. 1B) revealed that
the antiserum, prepared as described in Materials and Meth-
ods, reacted specifically with the VirE2 protein from E.
coli(pSW153) (lane 7) and A. tumefaciens(pTi219) (lane 3)
and recognized no proteins from E. coli(pUC19) (vector
only) (lane 6) or A. tumefaciens(pTi318), which contains an
insertionally inactivated virE2 gene (see Materials and Meth-
ods) (lane 2). A ca. 156-kilodalton protein from A.
tumefaciens(pTi358) most probably represents a VirE2-,-
galactosidase fusion protein generated by a translational
fusion of lacZ (in Tn3HoHol) and virE2, since the protein
reacted not only with antibodies to the VirE2 protein (lane
- 24
-18
AS - 15
FIG. 1. Purification of VirE2 protein from A. tumefaciens
A136(pTi219, pToK9) and E. coli JM101(pSW153). (A) Coomassie
blue-stained SDS-12.5% polyacrylamide gel showing stages of the
purifications. (B) Corresponding immunoblot developed with anti-
serum raised against VirE2 protein (see Materials and Methods).
Lanes: 1, total protein from acetosyringone-induced A. tumefaciens
A136(pTi358); 2, total protein from acetosyringone-induced A136
(pTi318); 3, total protein from acetosyringone-induced A136(pTi219,
pToK9); 4, material in 1.5 M NaCl fractions from single-stranded
DNA-cellulose column chromatography of total protein from
A136(pTi219, pToK9); 5, material in pooled 1.0 and 2.5 M NaCl
fractions from immunoaffinity column chromatography of VirE2-
containing single-stranded-DNA-cellulose fractions (4 jig of pro-
tein); 6, total protein from IPTG-induced E. coli JM101(pUC19); 7,
total protein from IPTG-induced JM101(pSW153); 8, pelleted mate-
rial after centrifugation of crude extract obtained by French press
treatment of JM101(pSW153); 9, material in 1.0 M NaCl fraction
from single-stranded-DNA column chromatography of the urea-
solubilized pellet (4 ,ug of protein); 10, molecular mass standards,
with sizes in kilodaltons indicated at right.
2662 CHRISTIE ET AL.
tography resulted in a preparation that yielded a single band
when 4 ,ug of the protein was analyzed by SDS-PAGE and
Coomassie blue staining (lane 9).
The VirE2 protein also was purified from A. tumefaciens.
Preliminary experiments revealed that VirE2 protein was
retained after passage of crude extracts of acetosyringone-
induced A. tumefaciens A136(pTi219, pToK9) through de-
natured calf thymus DNA-cellulose columns. The column
was washed extensively (10 column volumes) with Z buffer
and then eluted in steps of 0.1, 0.2, 0.3, 0.4, 0.5, 1.0, 1.5, and
2.5 M NaCl in Z buffer. The majority of the protein eluted
with 1.0 and 1.5 M NaCl. However, SDS-PAGE analysis of
material in the 1.5 M NaCl fraction (Fig. 1A, lane 4) showed
that several contaminating proteins also were present in
these fractions. Therefore, the VirE2-containing single-
stranded-DNA column fractions were passed through an
immunoaffinity column containing antibodies reactive
against VirE2 (see Materials and Methods). The bound
VirE2 protein was eluted with 1.0 and 2.5 M NaCl. Material
in these fractions was concentrated through microconcentra-
tor filters (Amicon) and yielded a single band when 4 ,g of
protein was electrophoresed through SDS-polyacrylamide
gels (Fig. 1A, lane 5).
Electrophoresis mobility shift studies. The binding of puri-
fied VirE2 protein to eight single-stranded and four blunt-
ended, double-stranded oligonucleotides was measured by
an electrophoresis mobility shift assay (see Materials and
Methods). VirE2 protein purified from E. coli, as well as
from A. tumefaciens, retarded electrophoretic migration of
all of the single-stranded oligonucleotides without sequence
specificity, but did not retard any of the double-stranded
oligonucleotides tested. The altered migration of single-
stranded, but not of double-stranded oligonucleotides is
depicted in Fig. 2, in which the single-stranded oligonucleo-
tide sequence was CAGGAAACAGCTATGAAACACGTT
CTTCTT (odd-numbered lanes), with the complementary
sequence annealed to form the double-stranded molecule
(even-numbered lanes). A single DNA-protein complex was
identified near the top of the gel when the single-stranded
oligonucleotide was incubated with VirE2 protein purified
from either bacterial source (lanes 3 and 7). However, when
the double-stranded oligonucleotide was incubated with
VirE2 protein, no shift in electrophoretic migration was
detectable (lanes 4 and 8). The addition of double-stranded
oligonucleotides to a reaction mixture did not affect the
ability of VirE2 protein to associate with single-stranded
oligonucleotides (data not shown).
To confirm that the observed changes in migration were
due to the VirE2 protein, we also tested the association of
protein preparations from E. coli(pUC19) (vector only) and
from A. tumefaciens(pTi318) (virE2 mutant) with DNA. In
each case, the protein was isolated by the identical proce-
dure used to purify VirE2 from the corresponding VirE2-
producing host. Material in fractions collected after single-
stranded-DNA-cellulose chromatography of the urea-
solubilized pellet from E. coli(pUC19) (Fig. 2, lanes 5 and 6)
or in fractions collected after single-stranded-DNA-cellulose
and immunoaffinity column chromatography of A. tume-
faciens(pTi318) (lanes 9 and 10), did not retard either single-
or double-stranded oligonucleotides. These experiments
demonstrate that material from the two bacterial hosts that
could have copurified with VirE2 protein was not responsi-
ble for the observed retardation of DNA.
The ability of the VirE2 protein to bind to DNA might
have resulted from an interaction of VirE2 protein with a
copurified VirEl protein in the preparation from A. tumefa-
J. BACTERIOL.
1 2 3 4 5 6 7 8 9 10
E2::ssDNA -
*:
S...,:
FIG. 2. DNA-binding assay of purified VirE2 protein. Reaction
mixtures were electrophoresed through a 5 to 20% polyacrylamide
linear gradient gel to resolve free and protein-bound DNA, as
described in the text. Lanes 1, 3, 5, 7, and 9 show migration of
single-stranded DNA, and lanes 2, 4, 6, and 8 show migration of
double-stranded DNA. Lanes: 1 and 2, DNA-only controls; 3 and 4,
VirE2 protein purified from A. tumefaciens A136(pTi219, pToK9); 5
and 6, material from A136(pTi318), a virE2 mutant; 7 and 8, VirE2
protein purified from E. coli JM101(pSW153); 9 and 10, material
from E. coli JM101(pUC19), vector-only control. Gel-retarded
VirE2-single-stranded-DNA complexes appear in lanes 3 and 7
(E2::ssDNA).
ciens or with the LacZ::VirEl fusion product in the prepa-
ration from E. coli (see Materials and Methods). To address
this possibility, we assessed the DNA-binding property of
VirE2 protein produced by a strain of E. coli harboring a virE
operon in which the virEl coding sequence is disrupted by an
in-frame deletion (pCGN1513) (34). The VirE2 protein from
this strain bound to single-stranded-DNA-cellulose columns
and was eluted with the same salt concentrations as the
protein purified from E. coli(pSW153) and A. tumefa-
ciens(pTi219, pToK9) (data not shown). On the basis of
results of the above experiments, we conclude that the
VirE2 protein is a single-stranded-DNA-binding protein that
binds to any single-stranded DNA molecule. Further, this
binding is independent of the VirEl protein.
Localization of VirE2 protein. In order to localize the
product of virE2, proteins from acetosyringone-induced cells
of A. tumefaciens A348 were separated into cytoplasmic,
periplasmic, inner membrane, and outer membrane fractions
(see Materials and Methods). The A280 pattern demonstrat-
ing separation of the inner and outer membranes is shown in
Fig. 3A. The peak inner and outer membrane fractions had
buoyant densities of 1.132 and 1.235 g/cm3, respectively.
The inner membrane marker NADH oxidase was present
only in the inner membrane fractions (Fig. 3A). Malate
dehydrogenase activity, a cytoplasmic protein marker, was
present only in the cytoplasmic fraction (data not shown).
Fractions from the four cellular compartments were found to
have unique protein profiles when analyzed by SDS-PAGE
and stained with Coomassie blue (Fig. 3B), further suggest-
ing an efficient fractionation.
Proteins from an identical but unstained gel were trans-
ferred to nitrocellulose, and the VirE2 protein then was
identified by immunological staining with the VirE2-specific
antiserum (Fig. 3C). Interestingly, although VirE2 protein
was found primarily in the cytoplasmic protein fraction (lane
2), a small but significant amount was also present in the
 .e_*S=;s,g-:fli9
VOL. 170, 1988 VirE2 ssDNA-BINDING PROTEIN 2663

8
Jr
[ Ia

2
z

Fraction

2 3 4 5 6 M B 1 2 3 4 5 6 M C

_ - 214
__ - 214
- 111
- i11
w
ao - 66 - 66
n_. ....
_. . r'
n -.
4EU -
_m
s_ . . _
........
_

_
.....

e -45 j - 45
2
_I Ff: .:
_U:.
_ __.
I

w
-
- 24
- 24
-
- s-

FIG. 3. Immunoblotting analysis of fractionated A. tumefaciens A348 cellular components. Cells were induced with acetosyringone (100
,uM) for 18 h and then separated into cytoplasmic, inner and outer membrane, and periplasmic fractions, as described in Materials and
Methods. (A) A280 values ( ) and NADH oxidase activity (- -) of fractions from isopycnic sucrose density gradient centrifugation of cell
-

membranes. (B) Proteins from the various fractions electrophoresed through denaturing 10% polyacrylamide gels and stained with Coomassie
blue. (C) Proteins transferred to nitrocellulose and probed with VirE2 antisera. For panels B and C, lanes are as follows: 1, total cellular
protein; 2, cytoplasmic protein; 3, total membrane protein; 4, inner membrane protein; 5, outer membrane protein; 6, periplasmic protein.

inner membrane fraction (lane 4) and a low level was
detected in both the outer membrane and periplasmic frac-
tions (lanes 5 and 6 respectively). Strain A348 was fraction-
ated and analyzed several times as described above to
confirm these findings. Similar results also were obtained
when fractions from A. tumefaciens A136(pTi243) (a virB
mutant) (43) and LBA4404 (pTiA6NC deleted of T-DNA)
(37) were examined (data not shown).
Immunoprecipitation of VirE2 protein::T-DNA complexes.
Since both single-stranded T-DNA molecules and a single-
stranded-DNA-binding protein are synthesized following vir
gene induction, we looked for a possible association between
VirE2 protein and T-DNA. Crude cell extracts of A136
(pTi219) or A136(pTi318) were precipitated with preimmune
or VirE2-specific antiserum and protein A-bearing Staphy-
lococcus cells (see Materials and Methods). Immunoprecip-
itates were then subjected to SDS-PAGE and immuno-
stained with the VirE2-specific antiserum. VirE2-specific
antiserum precipitated the VirE2 protein from A136(pTi2l9)
crude extracts (Fig. 4A, lane 5). No material other than the
immunoreactive Staphylococcus protein A was detected
when the A136(pTi219) crude extracts were immunoprecip-
itated with protein A-bearing Staphylococcus cells only (lane
1) or with preimmune antiserum (lane 4). Similarly, no
material other than protein A was detected when A136
(pTi318) (virE2 mutant) crude extracts were immunoprecip-
itated with preimmune antiserum (lane 2) or with VirE2-
specific antiserum (lane 3).
The immunoprecipitates then were analyzed by Southern
hybridization for the presence of T-DNA. Samples (10 ,ul)
spotted onto nitrocellulose were probed with pNW34-C-81, a
plasmid carrying the BamHI 8 fragment (7.5 kb) from the
2664 CHRISTIE ET AL. J. BACTERIOL.

1 2 3 4 5 M A virEl (pTi341, pTi361) or virE2 (pTi358, pTi318) (Table 1).

LBA4404 did not restore virulence to any strain carrying a
- 214 mutation in any of the other vir loci (data not shown).
The vir mutants were tested for their ability to restore
* - 111 virulence to virE mutants. Strains with insertions in virA,
- 66 virB, virD, virE, or virG did not restore virulence to A136
(pTi358) (Table 1). However, strains with insertions in either
r--u~~~~~~~~~~~~~~~:, ....4':~ 45 virCI (pTi379, A1034) or virC2 (pTi365, pTi364) did restore
virulence to A136(pTi358). The same results were obtained
when virEl (pTi361) and other virE2 mutants (pTi318,
pTi341) were examined (data not shown). Interestingly,
- 24
TABLE 1. Restoration of virulence of virE2 mutant
A136(pTi358) by coinfection
Restoration of
S_ B Rescuing strain
(plasmid)a Dsrpin
Description
virulence on
Kalanchoe
leaves
FIG. 4. Immunoprecipitation of VirE2::T-DNA complexes.
Conditions for immunoprecipitation of material from crude extracts A136
of A136(pTi219, pToK9) and A136(pTi318) are described in Materi- A348(pTiA6NC) ++
als and Methods. (A) Electrophoresis of immunoprecipitated mate- LBA4404 pTiA6NC deleted of
rial through denaturing 12.5% acrylamide gels, followed by devel- T-DNA and Ti
opment of the immunoblot with VirE2-specific antiserum. (B) plasmid bacterial
Immunoprecipitated material spotted on nitrocellulose and probed conjugation genes
with radioactively labeled pNW34C-8-1 in Southern hybridization
analyses for the presence of T-DNA sequences. Lanes: 1, material LBA4404 x A136(pTi3l8) virE2 mutant ++
immunoprecipitated from A136(pTi219, pToK9) with protein A- LBA4404 x A136(pTi341) virE mutant ++
bearing Staphylococcus cells only; 2, material immunoprecipitated LBA4404 x A136(pTi361) virE mutant ++
from A136(pTi318) with preimmune antiserum; 3, material immuno- LBA4404 x A136(pTi358) virE2 mutant ++
precipitated from A136(pTi318) with VirE2-specific antiserum; 4,
material immunprecipitated from A136(pTi219, pToK9) with preim- A136(pTi226) Insertion in virA operon
mune antiserum; 5, material immunoprecipitated from A136(pTi219, A136(pTi237) Insertion in virA operon
pToK9) with VirE2-specific antiserum.
A136(pTi368) Insertion in virB operon
A136(pTi238) Insertion in virB operon
T-DNA region of pTiA6NC. The probe, made single A136(pTi243) Insertion in virB operon
stranded by being boiled, hybridized to DNA present in A136(pTi26) Insertion in virB operon
material immunoprecipitated by VirE2-specific antiserum A136(pTi241) Insertion in virB operon
from A136(pTi219) extracts (Fig. 4B). These findings A136(pTi27) Insertion in virB operon
strongly suggest that single-stranded T-DNA molecules were A136(pTi364) Insertion in virC operon ++
immunoprecipitated. Although it is conceivable that protein- A136(pTi365) Insertion in virC operon ++
bound double-stranded T-DNA molecules may also have A136(pTi379) Insertion in virC operon ++
been precipitated and transferred to nitrocellulose, the sin- A1034 Insertion in virC operon ++
gle-stranded DNA probe would not have hybridized to these
molecules, since they were not denatured prior to Southern A136(pTi355) Insertion in virD operon
hybridization. The possibility that the entire pTiA6NC plas- A136(pTi3ll) Insertion in virD operon
mid was precipitated by the VirE2-specific antiserum was A136(pTi328) Insertion in virD operon
tested by probing the immunoprecipitates with pJW211, A136(pTi3O4) Insertion in virD operon
which harbors the BamHI 14 fragment (5.14 kb) of the virB A136(pTi3O7) Insertion in virD operon
region of pTiA6NC (see Materials and Methods). This region A136(pTi358) Insertion in virE operon
of the Ti plasmid was not present in any of the immunopre- A136(pTi3l8) Insertion in virE operon
cipitates examined when analyzed by Southern hybridiza- A136(pTi341) Insertion in virE operon
tion (data not shown). Therefore, we conclude that antibod- A136(pTi361) Insertion in virE operon
ies to the VirE2 protein specifically precipitated a
DNA-protein complex that most probably consists of single- A136(pTil9) Insertion in virG operon
stranded T-DNA molecules and VirE2 protein. A136(pTi363) Insertion in virG operon
Restoration of A. tumefaciens A136(pTi358) virulence by
coinfection. We tested the observation first reported by Otten ME42(pTiA6NC) Insertion in chvA locus
et al. (39) that avirulent virE mutants regained virulence by ME66(pTiA6NC) Insertion in chvA locus
coinfection with strains harboring an intact vir region, but ME61(pTiA6NC) Insertion in chvB locus
not if the helper strains carried mutations in any of the vir ME103(pTiA6NC) Insertion in chvB locus
loci on the Ti plasmid. Test cultures were mixed, applied to
wound sites on Kalanchoe leaves, and examined for their A5729(pTiA6NC) Insertion in exoC locus
ability to incite tumor formation as described in Materials A5503(pTiA6NC) Insertion in exoC locus
and Methods. Strain LBA4404, which carries pTiA6NC with a All vir mutants, with the exception of A1034 (virC), were generated by
an intact vir region and a deleted T-DNA (37), restored Tn3HoHol insertions (43). Mutant A1034 and the chromosomal mutants
virulence to four strains carrying Tn3HoHol insertions in (chvA, chvB, and exoC) were generated by TnS insertions (8, 11, 12, 55).
VOL. 170, 1988
Otten et al. (39) reported that all vir loci from the octopine
plasmid pTiB6S3, including virC, must be intact for comple-
mentation by the rescuing strain.
We further extended the earlier studies of Otten by
examining strains carrying mutations in the chromosomal
virulence loci chvA, chvB, and exoC for their ability to
rescue virE mutants. Two strains with different mutations in
each locus were examined. None of the mutants restored
virulence to the virE mutants, as shown for A136(pTi358)
(Table 1). Since these loci mediate initial bacterium-plant
cell attachment processes, these findings suggest that com-
plementation requires that the rescuing strain be capable of
binding to plant cells.
DISCUSSION
Processing of T-DNA for transfer to plants most probably
is functionally analogous to processing of plasmid DNA for
interbacterial conjugative transfer. Both DNA transfer sys-
tems involve strand-specific nicking by a site-specific endo-
nuclease, 5'-to-3' directional unwinding of the nicked strand
and replacement synthesis of the displaced strand, and,
presumably, the transfer of a DNA-protein complex (1, 22,
44). Recently, Buchanan-Wollaston et al. (7) demonstrated
that plasmids containing mobilization functions (mob) and an
origin of transfer region (oril), the functional equivalents of
the VirD endonuclease and the T-DNA border sequences,
respectively, are transferred to plants by A. tumefaciens
strains carrying an intact vir region. Therefore, mob and oriT
functions required for interbacterial plasmid transfer appar-
ently can substitute for the analogous functions in the
T-DNA transfer system, providing more evidence for the
similarity of these DNA transfer systems.
The analogy between T-DNA transfer and interbacterial
conjugation is further strengthened by our identification of
the VirE2 single-stranded-DNA-binding protein. The ability
of VirE2 to bind to single-stranded DNA recently was also
implicated on the basis of gel retardation studies with crude
extracts from Agrobacterium mutants (17a; C. Geitl, Z.
Koukolikova-Nicola, and B. Hohn, personal communica-
tion). Our ability to coprecipitate VirE2 protein and free
T-DNA molecules provides strong evidence that this DNA-
protein association is a physiologically significant compo-
nent of the T-DNA processing reaction. Numerous bacterial
conjugative plasmids also encode single-stranded-DNA-
binding proteins (18, 50). The requirement of these proteins
in bacteriophage and bacterial DNA recombination and
replication and in stabilization and protection of newly
synthesized DNA is well documented (9). The plasmid-
encoded single-stranded-DNA-binding proteins, as well as
the VirE2 protein, are likely to provide analogous functions
during conjugative DNA synthesis. It should be noted that
virulence of A. tumefaciens is not completely abolished by a
virE2 mutation. This suggests either that T-strands may be
synthesized and transferred to plant cells at a very low
frequency in the absence of functions supplied by the VirE2
protein or that the product of another gene can compensate
for this loss of function. However, we have not been able to
demonstrate that the virE2 gene product complements the
ssb mutation in E. coli KL450 (18) (P. J. Christie, unpub-
lished data).
Although the available evidence strongly suggests that
T-DNA processing reactions are analogous to early events
associated with plasmid transfer (22, 50), the data do not rule
out the possibility that double-stranded T-DNA molecules
could excise precisely from the Ti plasmid by a mechanism
VirE2 ssDNA-BINDING PROTEIN 2665
analogous to conservative transposition or site-specific re-
combination (2, 23, 28a, 49). Since double-stranded linear
and circular T-DNA molecules have been isolated from
induced cells, it will be of considerable interest to discern
which of these intermediate molecules is biologically rele-
vant to the transfer process.
It is difficult to understand the complementation data of
Otten et al. (39) and the data reported in this paper if the
VirE2 protein functions only in the bacterial cell. Comple-
mentation in mixed infections most probably occurs either
extracellularly or within the plant cell, which suggests that
the VirE2 protein can be exported from the cell. The VirE2
protein localization studies presented here support this idea.
Although the majority of the protein was found in the
cytoplasmic fraction, both membrane fractions and the peri-
plasmic fraction also contained VirE2 protein. These find-
ings agree with those of Engstrom et al. (13), who recently
identified a protein the size of VirE2 that was present both in
cytoplasmic and total membrane fractions of induced cells,
but was absent in corresponding fractions from uninduced
cells and virE mutants.
Two lines of evidence suggest that export of VirE2 would
occur by a complex mechanism. First, the VirE2 protein
sequence as deduced from its DNA sequence does not
contain any signal sequences characteristic of secreted pro-
teins (36, 51). Second, rescue of a virE2 mutant in mixed
infections requires that the rescuing strain encode not only
the virE2 gene product, but also the products of the virA,
virB, virD, and virG operons. Since the products of virA and
virG are involved in transcriptional activation of the other vir
genes (46, 52), it is likely that products of the virB and virD
operons are involved directly in the helper phenomenon. We
recently analyzed DNA sequences of the virB (J. E. Ward,
D. E. Akiyoshi, D. Regier, A. Datta, M. P. Gordon, and
E. W. Nester, J. Biol. Chem., in press) and virD (41, 56)
operons and found that most of the 11 predicted virB gene
products, as well as the virD4 gene product, are likely to be
membrane associated. Interestingly, three strains carrying
virD4 mutations (strains 355, 307, and 328) failed to restore
virulence of A136(pTi358) in the coinfection studies (Table
1). Our analyses suggest that the virB and virD4 proteins
form a transmembrane structure through which T-DNA,
protein, or a T-DNA-protein complex might be exported.
VirE2 protein export from helper strains may require such a
transmembrane structure, which would explain why helper
strains require the additional vir products. Our mixed-infec-
tion experiments also showed that gene products of the virC
operon are not required for complementation. Recent evi-
dence from our laboratory indicates that these gene products
may function during initial T-DNA processing by enhancing
the virD endonuclease-mediated cleavage reaction (N. Toro,
unpublished data). Since the helper strains lack T-DNA,
only the products of the vir regulon necessary for virE2
protein synthesis and export would be required.
Our observation that the three chromosomal loci chvA,
chvB, and exoC involved in mediating attachment to plant
cells and tumor formation (8, 11, 12) are required for
complementation strongly suggests that the bacterium-plant
cell association is important in these mixed infections. It is
tempting to speculate that the restoration of virulence of a
virE2 mutant results from the independent transfer of VirE2
protein by helper strains and T-DNA by virE2 mutants into
the same plant cell, where a protein-DNA association is
established.
These results, taken together with results of our earlier
2666 CHRISTIE ET AL.
experiments, suggest the following model for T-DNA proc-
essing and transfer. The processing reaction is initiated by
cleavage of the bottom strand at T-DNA border sequences
by the virD endonuclease. Replication then proceeds from
one nicked border to the next and displaces the bottom
strand 5' to 3', thereby generating a linear single-stranded
molecule (1, 44, 45, 56). Recently, researchers in our labo-
ratory have determined that during the nicking event the
VirD2 protein binds to the 5' end of the T-strand (C. Young
and E. W. Nester, submitted for publication). We predict
that the VirE2 protein is involved in the replication-strand
displacement reaction and forms a stable association with
single-stranded T-DNA molecules. The structure of the
protein-T-DNA complex transferred to plant cells then
would consist of a linear single-stranded T-DNA molecule
capped at the 5' end by the VirD2 protein and coated along
its length by VirE2 protein. These proteins would protect T
strands from bacterium- or plant-encoded nucleases during
the entire infection process and conceivably could play a
role in the integration of T strands into the plant chromo-
some. This model predicts that a protein-T-DNA complex
would be detectable by electron microscopy in bacterial
and/or plant cells, and we are currently examining this
possibility.
ACKNOWLEDGMENTS
We thank K. E. McBride and V. C. Knauf for supplying bacterial
plasmids prior to publication and B. Traxler for sharing her excellent
scientific expertise. We also thank S. G. Porter for her imnportant
contributions in raising antiserum and in the localization studies,
Dale Parkhurst for synthesizing the oligonucleotides, L. Wong for
excellent technical assistance, and C. Young and H. B. Kaplan for
helpful discussions and critical reading of the manuscript.
This work was supported by Public Health Service grant 5 ROI
GM 32618-14 from the National Institutes of Health and National
Science Foundation grant DMB-8315826. P.J.C. and J.E.W. were
supported by American Cancer Society postdoctoral fellowships
PF-2937 and PF-2796, respectively. S.C.W. was supported by
Damon Runyon-Walter Winchell Cancer Fund fellowship DRG-800.
LITERATURE CITED
1. Albright, L. M., M. F. Yanofsky, B. Leroux, D. Ma, and E. W.
Nester. 1987. Processing of the T-DNA of Agrobacterium tume-
faciens generates border nicks and linear, single-stranded T-
DNA. J. Bacteriol. 169:1046-1055.
2. Alt-Moerbe, J., B. Rak, and J. Schroder. 1986. A 3.6-kb segment
from the vir region of Ti plasmids contains genes responsible for
border sequence-directed production of T region circles in E.
coli. EMBO J. 5:1129-1135.
3. Barker, R. F., K. B. Idler, D. V. Thompson, and J. D. Kemp.
1983. Nucleotide sequence of the T-DNA region from the
Agrobacterium tumefaciens octopine Ti plasmid pTi5955. Plant
Mol. Biol. 2:335-350.
4. Bolton, G. W., E. W. Nester, and M. P. Gordon. 1986. Plant
phenolic compounds induce expression of the Agrobacterium
tumefaciens loci needed for virulence. Science 232:983-985.
5. Bradford, M. 1976. A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing the
principle of protein-dye binding. Anal. Biochem. 72:248-254.
6. Bruist, M. F., and M. I. Simon. 1984. Phase variation and the
Hin protein: in vivo activity measurements, protein overproduc-
tion, and purification. J. Bacteriol. 159:71-79.
7. Buchanan-Wollaston, V., J. E. Passiatore, and F. Cannon. 1987.
The mob and oriT mobilization functions of a bacterial plasmid
promote its transfer to plants. Nature (London) 328:172-175.
8. Cangelosi, G. A., L. Hung, V. Puvanesarajah, G. Stacey, D. A.
Ozga, J. A. Leigh, and E. W. Nester. 1987. Common loci for
Agrobacterium tumefaciens and Rhizobium meliloti exopoly-
saccharide synthesis and their roles in plant interactions. J.
J. BACTERIOL.
Bacteriol. 169:2086-2091.
9. Chase, J. W., and K. R. Williams. 1986. Single-stranded DNA
binding proteins required for DNA replication. Annu. Rev.
Biochem. 55:103-136.
10. Chilton, M. D., T. C. Currier, S. K. Farrand, A. J. Bendich,
M. P. Gordon, and E. W. Nester. 1974. Agrobacterium DNA
and PS8 bacteriophage DNA not detected in crown gall tumors.
Proc. Natl. Acad. Sci. USA 71:3672-3676.
11. Douglas, C. J., R. J. Staneloni, R. A. Rubin, and E. W. Nester.
1985. Identification and genetic analysis of an Agrobacterium
tumefaciens chromosomal virulence region. J. Bacteriol. 161:
850-860.
12. Douglas C. J., W. Halperin, and E. W. Nester. 1982. Agrobac-
terium tumefaciens mutants affected in attachment to plant
cells. J. Bacteriol. 152:1265-1275.
13. Engstrom, P., P. Zambryski, M. Van Montagu, and S. Stachel.
1987. Characterization of Agrobacterium tumefaciens virulence
proteins induced by the plant factor acetosyringone. J. Mol.
Biol. 197:635-645.
14. Gardner, R. C., and V. C. Knauf. 1986. Transfer of Agrobac-
terium DNA to plants requires a T-DNA border but not the virE
locus. Science 231:725-727.
15. Garfinkel, D. J., and E. W. Nester. 1980. Agrobacterium tume-
faciens mutants affected in crown gall tumorigenesis and octo-
pine catabolism. J. Bacteriol. 144:732-743.
16. Garfinkel, D. J., R. B. Simpson, L. W. Ream, F. F. White, M. P.
Gordon, and E. W. Nester. 1981. Genetic analysis of crown gall:
fine structure map of the T-DNA by site-directed mutagenesis.
Cell 27:143-153.
17. Gaylo, P. J., N. Turjman, and D. Bastia. 1987. DnaA protein is
required for replication of the minimal replicon of the broad-
host-range plasmid RK2 in Escherichia coli. J. Bacteriol. 169:
4703-4709.
17a.Geitl, C., Z. Koukolikova-Nicola, and B. Hohn. 1987. Mobiliza-
tion of T-DNA from Agrobacterium to plant cells involves a
protein that binds single-stranded DNA. Proc. Natl. Acad. Sci.
USA 84:9006-9010.
18. Golub, E. I., and K. B. Low. 1985. Conjugative plasmids of
enteric bacteria from many different incompatibility groups
have similar genes for single-stranded DNA-binding proteins. J.
Bacteriol. 162:235-241.
19. Hager, D. A., and R. R. Burgess. 1980. Elution of proteins from
sodium dodecyl sulfate-polyacrylamide gels, removal of sodium
dodecyl sulfate, and renaturation of enzymatic activity; results
with sigma subunit of Escherichia coli RNA polymerase, wheat
germ DNA topoisomerase, and other enzymes. Anal. Biochem.
109:76-86.
20. Hendrickson, W., and R. F. Schlief. 1984. Regulation of the
Escherichia coli L-arabinose operon studied by gel electropho-
resis DNA binding assay. J. Mol. Biol. 174:611-628.
21. Horsch, R. B., H. J. Klee, S. Stachel, S. C. Winans, E. W.
Nester, S. G. Rogers, and R. T. Fraley. 1986. Analysis of
Agrobacterium tumefaciens virulence mutants in leaf discs.
Proc. Natl. Acad. Sci. USA 83:2571-2575.
22. Ippen-Ihler, K. A., and E. G. Minkley, Jr. 1986. The conjugation
system of F, the fertility factor of Escherichia coli. Annu. Rev.
Genet. 20:593-624.
23. Jayaswal, R. K., K. Veluthambi, S. B. Gelvin, and J. L.
Slightom. 1987. Double-stranded cleavage of T-DNA and gen-
eration of single-stranded T-DNA molecules in Escherichia coli
by a virD-encoded border-specific endonuclease from Agrobac-
terium tumefaciens. J. Bacteriol. 169:5035-5045.
24. Jin, S., T. Komari, M. Gordon, and E. W. Nester. 1987. Genes
responsible for the supervirulence phenotype of Agrobacterium
tumefaciens. J. Bacteriol. 169:4417-4425.
25. Kadonaga, J. T., and R. Tjian. 1986. Affinity purification of
sequence-specific DNA binding proteins. Proc. Natl. Acad. Sci.
USA 83:5889-5893.
26. Kitto, G. B. 1969. Intra- and extramitochondrial malate dehy-
drogenases from chicken and tuna heart. Methods Enzymol.
13:106-116.
27. Klee, H. J., F. F. White, V. N. Iyer, M. P. Gordon, and E. W.
Nester. 1983. Mutational analysis of the virulence region of an
VOL. 170, 1988
Agrobacterium tumefaciens Ti plasmid J. Bacteriol. 153:878-
883.
28. Knauf, V. C., and E. W. Nester. 1982. Wide host range cloning
vectors: a cosmid clone bank of an Agrobacterium Ti plasmid.
Plasmid 8:45-54.
28a.Koukolikova-Nicola, Z., L. Albright, and B. Hohn. 1987. Plant
DNA infectious agents, p. 110-148. In T. Hohn and J. Schell
(ed.), Plant gene research, vol. 4. Springer-Verlag, Vienna.
29. Koukolikova-Nicola, Z., R. D. Shillito, B. Hohn, K. Wang, M.
Van Montagu, and P. Zambryski. 1985. Involvement of circular
intermediates in the transfer of T-DNA from Agrobacterium
tumefaciens to plant cells. Nature (London) 313:191-1%.
30. Laemmli, U. K. 1970. Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature (London)
227:680-685.
31. Leroux, B., M. F. Yandfsky, S. C. Winans, J. E. Ward, S. F.
Ziegler, and E. W. Nester. 1987. Characterization of the virA
locus of Agrobacterium tumefaciens: a transcriptional regulator
and host range determinant. EMBO J. 6:849-856.
32. Maagd, R. A., and B. Lugtenberg. 1986. Fractionation of
Rhizobium leguminosarum cells into outer membrane, cytoplas-
mic membrane, periplasmic, and cytoplasmic components. J.
Bacteriol. 167:1083-1085.
33. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular
cloning: a laboratory manual. Cold Spring Harbor Laboratory,
Cold Spring Harbor, N.Y.
34. McBride, K. E., and V. C. Knauf. 1988. Genetic analysis of the
virE operon of the Agrobacterium Ti plasmid pTiA6. J. Bacte-
riol. 170:1430-1437.
35. Nester, E. W., M. P. Gordon, R. M. Amasino, and M. F.
Yanofsky. 1984. Crown gall: a molecular and physiological
analysis. Annu. Rev. Plant Physiol. 35:387-413.
36. Oliver, D. 1985. Protein secretion in Escherichia coli. Annu.
Rev. Microbiol. 39:615-648.
37. Ooms, G., P. J. J. Hooykaas, R. J. M. Van Veen, P. Van Beelen,
T. Regensburg-Tunik, and R. A. Schilperoort, 1982. Octopine
Ti-plasmid deletion mutants of Agrobacterium tumefaciens with
emphasis on the right side of the T-region. Plasmid 7:15-19.
38. Osborn, M. J., J. E. Gander, E. Parisi, and J. Carson. 1972.
Mechanism of assembly of the outer membrane of Salmonella
typhimurium. J. Biol. Chem. 247:3962-3972.
39. Otten, L., H. DeGreve, J. Leemans, R. Hain, P. Hooykaas, and
J. Schell. 1984. Restoration of virulence of vir region mutants of
Agrobacterium tumefaciens strain B53 by coinfection with
normal and mutant Agrobacterium strains. Mol. Gen. Genet.
195:159-163.
40. Peralta, E. G., R. Hellmiss, and R. Ream. 1986. Overdrive, a
T-DNA transmission enhancer on the A. tumefaciens tumour-
inducing plasmid. EMBO J. 5:1137-1142.
41. Porter, S. G., M. F. Yanofsky, and E. W. Nester. 1987. Molec-
ular characterization of the virD operon from Agrobacterium
tumefaciens. Nucleic Acids Res. 15:7503-7517.
42. Stachel, S. E., E. Messens, M. Van Montagu, and P. Zambryski.
1985. Identification of the signal molecules produced by
VirE2 ssDNA-BINDING PROTEIN 2667
wounded plant cells that activate T-DNA transfer in Agrobac-
terium tumefaciens. Nature (London) 318:624-629.
43. Stachel, S. E., and E. W. Nester. 1986. The genetic and tran-
scriptional organization of the vir region of the A6 Ti-plasmid of
Agrobacterium tumefaciens. EMBO J. 5:1445-1454.
44. Stachel, S. E., B. Timmerman, and P. Zambryski. 1986. Gener-
ation of single-stranded T-DNA molecules during the initial
stages of T-DNA transfer from Agrobacterium tumefaciens to
plant cells. Nature (London) 322:706-712.
45. Stachel, S. E., B. Timmerman, and P. Zambryski. 1987. Activa-
tion of Agrobacterium tumefaciens vir gene expression gener-
ates multiple single-stranded T-strand molecules from the pTiA6
T-region: requirement for 5' virD gene products. EMBO J. 6:
857-863.
46. Stachel, S. E., and P. C. Zambryski. 1986. virA and virG control
the plant-induced activation of the T-DNA transfer process of
Agrobacterium tumefaciens. Cell 46:325-333.
47. Stachel, S. E., and P. C. Zambryski. 1986. Agrobacterium
tumefaciens and the susceptible plant cell: a novel adaptation of
extracellular recognition and DNA conjugation. Cell 47:155-
157.
48. Towbin, H., and J. Gordon. 1984. Immunoblotting and dot
immunoblotting-current status and outlook. J. Immunol.
Methods 72:313-340.
49. Veluthambi, K., R. K. Jayaswal, and S. B. Gelvin. 1987. Viru-
lence genes A, G, and D mediate the double-stranded border
cleavage of T-DNA from the Agrobacterium Ti plasmid. Proc.
Natl. Acad. Sci. USA 84:1881-1885.
50. Willetts, N., and B. Wilkins. 1984. Processing of plasmid DNA
during bacterial conjugation. Microbiol. Rev. 48:24-41.
51. Winans, S. C., P. Allenza, S. E. Stachel, K. E. McBride, and
E. W. Nester. 1987. Characterization of the virE operon of the
Agrobacterium tumefaciens Ti plasmid pTiA6. Nucleic Acids
Res. 15:825-837.
52. Winans, S. C., P. R. Ebert, S. E. Stachel, M. P. Gordon, and
E. W. Nester.. 1986. A gene essential for Agrobacterium viru-
lence is homologous to a family of positive regulatory loci. Proc.
Natl. Acad. Sci. USA 83:8278-8282.
53. Yadav, N. S., J. van der Leyden, D. R. Bennett, W. M. Barnes,
and M.-D. Chilton. 1982. Short direct repeats flank the T-DNA
on a nopaline Ti plasmid. Proc. Natl. Acad. Sci. USA 79:6322-
6326.
54. Yamamoto, A., M. Iwahasahi, M. F. Yanofsky, E. W. Nester, I.
Takebe, and Y. Machida. 1987. The promoter proximal region in
the virD locus of Agrobacterium tumefaciens is necessary for
the plant-inducible circularization of T-DNA. Mol. Gen. Genet.
206:174-177.
55. Yanofsky, M. F., and E. W. Nester. 1986. Molecular character-
ization of a host-range determining locus from Agrobacterium
tumefaciens. J. Bacteriol. 168:244-250.
56. Yanofsky, M. F., S. G. Porter, C. Young, L. Albright, M. P.
Gordon, and E. W. Nester. 1986. The virD operon of Agrobac-
terium tumefaciens encodes a site-specific endonuclease. Cell
47:471-477.