Eur. J. Biochem.

245, 156-163 (1997)

0 FEBS 1997

Differential expression of three Arabidopsis genes

encoding the B’ regulatory subunit of protein phosphatase 2A
Keith A. LATORRE, Darby M. HARRIS and Sabine J. RUNDLE
Department of Biology, Western Carolina University, Cullowhee NC, USA

(Received 11 November 1996/16 January 1997) - EJB 96 1677/2

Numerous plant processes ranging from signal transduction to metabolism appear to be mediated, in
part, by type 2A protein serinekhreonine phosphatases (PP2A). In an effort to identify factors that control
the activity of this enzyme in plants, we have isolated and characterized DNA sequences encoding the
B‘ regulatory subunit of PP2A from Arabidopsis thaliana. Specifically, we used PCR to amplify a segment
of Arabidopsis cDNA that encodes a conserved section of the B’ polypeptide. This PCR fragment was
subsequently used as a probe to screen an Arabidopsis cDNA library and cDNA clones derived from
three distinct genes were identified. The AtB’a and AtB’P genes encode highly similar 57-kDa B’ regula-
tory subunits while the third gene, AtB‘y, encodes a more divergent 59-kDa B’ protein. A comparison of
the three Arabidopsis B‘ polypeptides to those of yeast and animals shows the core region of this protein
to be the most conserved while the amino and carboxy termini vary both in length and sequence. Genomic
Southern blots indicate that at most the Arabidopsis genome contains five genes encoding the B‘ regula-
tory subunit. The three genes identified in this study are expressed in all Arabidopsis organs, albeit at
varying levels. In addition, mRNAs derived from the three genes accumulate differentially in response to
heat shock. Our results indicate that the activity of plant PP2A might be regulated by a B’ type regulatory
subunit similar to those found in animals and yeast, and suggest possible roles for B’-containing PP2A
complexes within plant cells.
Keywords: Arabidopsis thaliana; protein phosphorylation ; protein phosphatase 2A; regulatory subunit;
heat-shock protein.

Type 1 and 2A serinekhreonine protein phosphatases (PP)
constitute two of the major types of protein phosphatases in eu-
caryotic cells. They are sensitive to nanomolar concentrations of
inhibitors, such as okadaic acid and microcystin-LR, and hence
these, as well as other agents, have been used extensively to
examine the role of PP1 and/or PP2A in various cell and devel-
opmental events (Mumby and Walter, 1993; Wera and Hem-
mings, 1995). In plants the use of PPl/PP2A inhibitors has led
to the identification of a role for reversible protein phosphoryla-
tion in processes ranging from hormone signal transduction to
metabolism (Smith and Walker, 1996). For example, both gib-
berellin-induced and light-controlled gene expression in plants
can be blocked by incubating plant cells with okadaic acid (Kuo
et al., 1996; Sheen, 1993), while pathogenesis-related gene ex-
pression can be induced in the absence of pathogens if PP1 and/
or PP2A are inhibited (MacKintosh et al., 1994; Gianfagna and
Lawton, 1995). Cell communication between the pollen grain
and cells of the stigma in certain plant species can also be
disrupted in the presence of okadaic acid or microcystin-LR
Correspondence to S. J. Rundle, Department of Biology, Western
Carolina University, Cullowhee, NC 28723, USA
F a x : + I 704 227 7647.
Abbreviations. PP, serinehhreonine protein phosphatase; P E A , type
2A serinekhreonine protein phosphatase; PP2Ac, catalytic subunit of
PP2A.
Note. The novel nucleotide sequence data published here have been
deposited with the EMBL, GenBank and DDBJ sequence data banks and
are available under accession numbers U73256-U73528.
(Rundle et al., 1993), as can the activity of various ion channels
(Li et al., 1994; Thiel and Blatt, 1994). Finally a number of
plant metabolic enzymes, including sucrose phosphate synthase,
nitrate reductase and phosphoenolpyruvate carboxylase are regu-
lated by dephosphorylation by PP2A (Huber et al., 1994).
In animals the activity of PP2A is regulated by numerous
factors, including the lipid mediator ceramide, covalent modifi-
cation, and regulatory subunits (Wera and Hemmings, 1995). Of
these regulatory mechanisms, only the latter has thus far also
been identified in plants (Rundle et al., 1995; Corum et al.,
1996). Regulation of the activity of the catalytic subunit of
PP2A (PP2Ac) by its association with specific cellular regula-
tory proteins leads to the formation of either heterodimers com-
posed of PP2Ac and a 65-kDa subunit termed A, or heterotri-
mers composed of PP2Ac, A and a variable third subunit termed
B. The activity and specificity of PP2Ac is mediated by its asso-
ciation with these subunits (Mumby and Walter, 1993). Hetero-
triineric PP2A holoenzymes containing different B subunits
have been purified to homogeneity from various mammalian
sources and have led to the identification of PP2A enzyme com-
plexes containing B regulatory subunits of 54, 55,72 or 74 kDa.
In addition, a number of differentially expressed genes encoding
various B-type regulatory subunits have been identified in mam-
mals (Wera and Hemmings, 1995). Based on these data, it is
predicted that multiple distinct PP2A complexes can exist within
mammalian cells, each presumably with unique enzyme speci-
ficity.
Homologs to the A and 55-kDa B regulatory subunit have
also been identified in plants. In Arubidopsis thaliana three dis-
 Latorre et al. (Eur: J. Biochem. 245)
tinct genes encode the A regulatory subunit and these genes ap-
pear to be expressed in all plant organs (Slabas et al., 1994;
Corum et al., 1996). Mutations in one of these genes results in
plants with altered transport of the hormone auxin, indicating
that the three A regulatory subunit proteins are not functionally
equivalent (Garbers et al., 1996). In addition, two genes encod-
ing the Arabidopsis 55-kDa B regulatory subunit have been
identified (Rundle et al., 1995; Corum et al., 1996), these two
genes also show a constitutive expression pattern. Taken to-
gether, these data indicate that plant PP2Ac are most likely regu-
lated in a similar manner to their yeast and animal counterparts
and that, as is the case in mammals, multiple distinct PP2A com-
plexes can assemble within each plant cell.
Recently a number of cDNAs encoding a B regulatory sub-
unit termed B’ have been identified in mammals (McCright and
Virshup, 1995; McCright et al., 1996; Csortos et al., 1996; Zol-
nierowicz et al., 1996; Tehrani et al., 1996). Analysis of these
cDNAs indicates that in mammals at least five genes appear to
encode this subunit and that the mRNAs from at least three of
these genes are alternatively spliced (Csortos et al., 1996; Zol-
nierowicz et al., 1996; McCright et al., 1996). In total, the mam-
malian cDNA population characterized predicts at least 11 dis-
tinct B’ isoforms ranging in size over 52-69 kDa (Zolnierowicz
et al., 1996). A number of the B’ regulatory subunit proteins are
phosphoproteins (McCright et al., 1996) and, while some are
located in the cytoplasm, others appear to function in the nucleus
(McCright et al., 1996; Tehrani et al., 1996). Hence the B’ sub-
unit is thought to add considerably to the variation in structure,
subcellular localization and means of regulation of PP2A activ-
ity. The genome of Succhuromyces cerevisiae also contains a
homolog (RTSl/SCSI) to the mammalian B’ gene (Shu and
Hallberg, 1995). This gene was identified as a suppressor of
mutations in genes involved in the stress response of yeast cells.
Specifically, overexpression of RTSI/SCSI allows rescue of
cells with temperature-sensitive mutations in the mitochondria1
heat-shock protein HSP60 or with mutations in a transcriptional
activator (ROX3) necessary for S. cerevisiue cells to survive an-
aerobic stress (Shu and Hallberg, 1995; Rosenblum-Vos et al.,
1991).
Here we present evidence regarding the plant homolog of
the B’ regulatory subunit of PP2A. We show that the Arabi-
dopsis genome contains at least three genes encoding the B’ sub-
unit and that each of these genes is ubiquitously expressed. In
addition we show that mRNAs derived from these genes accu-
mulate differentially in response to heat shock. Our results indi-
cate plant PP2A might be regulated by a B‘ type regulatory sub-
unit and provide hints regarding a possible role of specific B‘-
containing PP2A complexes in plant cells.
MATERIALS AND METHODS
Polymerase chain reaction. PCR was performed using Ara-
bidopsis cDNA as a substrate and the following degenerate
primers reflecting conserved regions of B’ from mammals and
yeast (McCright and Virshup, 1995) : GAT/CCCNGAA/GGAA/
GGATKGANGCCN encoding the amino acids DPEEDEP and
A/GAAA/GTAT/CTCA/GTTA/GTTCCAA/GTA reflecting the
complementary strand of DNA encoding the amino acids YWN-
NEYF. PCR reactions contained 60 mM Tris/HCI, 15 mM
(NHJ2SO4, 2.0 mM MgCl, pH 8.5, 0.25 mM dNTPs, 1 pM of
each primer, 0.05 pg of cDNA and 1 U Taq polymerase. Sam-
ples were heated to 94°C for 2 min followed by 30 cycles of
94 “C for 1 min, 55 “C for 2 min and 72 “C for 3 min. Upon com-
pletion of the last cycle the samples were incubated at 72°C for
157
7 min. PCR products were separated by gel electrophoresis and
DNA fragments of the expected size (71 1 bp) were excised from
the gel and submitted to another round of PCR amplification as
described above. The resulting PCR fragments were cloned into
the pCR2.1 vector of InVitrogen and analyzed by DNA sequenc-
ing.
cDNA isolation and characterization. Approximately
I50000 /z phage of an Arabidopsis cDNA library were screened
using the 711-bp PCR fragment described above. The library
was obtained from the Arabidopsis Biological Resource Center
at Ohio State University and consisted of 2-3-kb size-fraction-
ated cDNA prepared from Arubidopsis hypocotyls (Kieber et al.,
1993). Screening was performed as described in Rundle et al.
(1995). Positive clones were placed into three groups based on
restriction mapping. The complete nucleotide sequence of both
strands of DNA were determined for a representative clone from
each group using nested deletions (Henikoff, 1984) and custom-
prepared oligonucleotides.
Genomic Southern blot analysis. DNA was isolated from
Arabidopsis (Columbia) seedlings using a small-scale DNA
preparation as described by Stein et al. (1991). The DNA was
digested with the appropriate restriction enzyme and submitted
to agarose gel electrophoresis. Gels were blotted to GeneScreen
Plus nylon membranes and hybridized as described by Nasrallah
et al. (1994) except that hybridizations were performed at 60°C.
Blots were washed twice for 30min at 60°C in 2XNaCliCit
(NaCVCit = 0.15 M sodium chloride, 0.015 M sodium citrate),
0.2% (masdvol.) SDS.
Northern blot analysis. Poly(A)-rich RNA was isolated
from various Arubidopsis (Columbia) organs using the Microfast
track kit from InVitrogen. RNA was separated by formaldehyde/
agarose gel electrophoresis and blotted to Genescreen plus ny-
lon membranes. Blots were probed and washed as described
above. For heat-shock experiments, Arabidopsis seeds were ger-
minated on MS medium (Murashige and Skoog, 1962) and
grown at 23°C for 10 days with a 12-h light/dark regime. Heat
shock was performed by placing plates with seedlings at 37°C
for 2 h, control plates being maintained at 23 “C. After 2 h, seed-
lings were collected and quick-frozen in liquid nitrogen. Po-
ly(A)-rich RNA was isolated from seedlings as described
above.
RESULTS
Isolation of cDNA sequences encoding the B’ subunit of
PP2A of Arubidopsis. In an effort to determine if plant species
of PP2A exist that contain B’ type regulatory subunits, we per-
formed PCR using Arubidopsis cDNA as a substrate and primers
reflecting conserved regions of the human and yeast B’ regula-
tory subunit (see Materials and Methods). A DNA fragment of
the predicted size (711 bp) was amplified, cloned and shown,
by partial DNA sequence analysis, to encode a segment of an
Arubidopsis B’ type regulatory subunit. This PCR fragment was
subsequently used as a probe to screen an Arabidopsis cDNA
library and 13 cDNAs falling into three distinct groups were
identified. A representative cDNA from each group (AtB’a,
AtB’P and AtB’y) was submitted to complete DNA sequence
analysis and the resulting predicted amino acid sequences are
presented in Fig. 1. The AtB‘a and AtB‘P cDNAs encode pro-
teins of 57 kDa with isoelectric points of 6.1 and 6.2 respective-
ly, while the polypeptide derived from AtB’y is slightly larger
(59 kDa) and is predicted to have an isoelectric point of 8.3.
Both the AtB’a and AtB’p genes produce mRNAs that can be
polyadenylated at more than one site, as indicated by sequence
158 Latorre et al. ( E m J. Biochern. 245)

AtB'u MFKKIMKGANRKASKAEANDS ........ MYGFDPP...GRSGPGSNMIV.N..HASRGSLV.PSSPNSM.............. ...... 56

AtB'P --------GfI--P--S---EP-........S--IG L-D..N------- W-S...-----A--N .... ....................
S W 57
AtB'y .I-Q.FGKLP..P..SSH...NPNGEGGVNS.YIPNSGISSI.K-S.KSSAS.SNG.NGTVIA.....T.S .................... 69
HSB56P ...................... METKLPPASTPTS-SSPOLSPV-PPDKVDGFSRRSL-RAS-S .................... 48
CeB' .................................... MIPADAPP-TNIGRTNTYOOGWIPRRERRPSS--.................... 35
RTSl ............................................. (1-220)-..---SQ-IDI-R-SH-FERLPTPTKLNPDTDLELIKT 257

AtB'a
AtB'p
AtB'y
HSB56$
W'a3
CeB'
RTS 1

AtB'u
AtB'P
AtB 'y
HsB56P
m'u3
CeB'
RTS 1

AtB 'u
AtB'$
AtB'y
HaB56P
m'u3
CeB'
RTS 1

AtB'u
AtB'P
AtB' y
HsB56p
W'a3
CeB'
RTS 1

AtB'a
AtB'P
AtB' y
HsB56p
MnB'tr3
CeB'
RTS 1

AtB'u RWRRLDFAVEERER. .EDPMITS .. 495

AtB'P T-K--A--M--mE--H---- 499
AtB 'y T-KL-E-LAASKSVSNEAVLVPRF-SSVTLATGKTSGS 522
HSB56P L-QO-E-LRLR-~~~~QRLTPQVAASOGQS 499
MnB'a3 A-VKIE"PQVLWRVTREC 435
CeB' I - N N I E K Q ~ G ~ ~ ~ F ~ D E I I S S R Q Q N O V D E N M K T S T V L S K D E I L K N A V G V S S ~ ~ M D F G P N E I K Q S D F P P D E Q T540
PRALGE
RTS 1 N-SK-E-Y-KNLRINNDKLQYTEKNPELRNSFlTASENNTE IQ 757

CeB' YKRHDPFLKKWSTDEQ 557

Fig.1. Alignment of the amino acid sequence of the B' regulatory subunit of PP2A from various organisms. The method of Pearson and
Lipman (1988) was used to align the amino acid sequences encoded by the three B' regulatory subunit genes of Ar~hidopsis(AtB'a, AtB'P, AtB'y),
human (HsB56/3; McCright and Virshup, 1995), mouse (MmB'a3; Tehrani et al., 1996), C. elegans (CeB'; Wilson et al., 1994) and S. cerevisiae
(RTSI ). Dashes indicate amino acid identity, dots indicate gaps introduced to maximize amino acid alignment.

characterization of various cDNAs derived from these genes. In
addition, the 5' untranslated regions of all three Arabidopsis B'
inRNAs contain one or more small open reading frames, repre-
senting either possible cloning artifacts or, if truly present in the
B' gene, potential translational control elements (Hinnebusch,
1984). Finally, the PCR fragment used to screen the Arabidopsis
cDNA library is derived from nucleotides 861 - 1571 of AtB'P
and probably explains why most of the cDNAs (9 of 13) iden-
tified in our screen were derived from this gene.
An alignment of the three Arabidopsis B' proteins with those
of S. cerevisiae (Shu and Hallberg, 1995) and various animals
(human B isoform according to McCright and Virshup, 1995;
mouse a isoform according to Tehrani et al., 1996; Caenorhab-
ditis elegans according to Wilson et al., 1994) is shown in Fig. 1.
The a and P isoforms of the Arubidopsis B' regulatory subunit
are most closely related showing 84% identity and 90% simi-
larity, while the y isoform shows only 61 % identity and 72-
73% similarity to the a and p isoforms. The greatest degree of
 Latorre et al. (Eur: J. Biochem. 245)
A 1 2 8 1 2 C 1 2 D 1 2
kb
5-
3-
2-
1-
Fig. 2. Genomic organization of the Arubidopsis B’ PPZA regulatory
subunit genes. Genomic DNA was isolated from Arabidopsfs (colum-
bia) seedlings, digested with EcoRI (lane 1) or HindIII (lane 2) and
subjected to agarose gel electrophoresis. Blots were hybridized with the
following probes: (A) a mixture of the three B’ cDNAs (nucleotides
407- 1920 of AtB’a, 293- 1875 of AtB’P, and 622-2045 of AtB’y) ;
(B) a segment from the 3’ end of the AtB’a cDNA starting at the Sac1
site at nucleotide 1643 and spanning to the end of the cDNA at nucleo-
tide 1920; (C) the 711-bp PCR fragment reflecting a section of AtB’P;
(D) a HindIII-EcoRI fragment spanning nucleotides 310-576 of the
AtB’y cDNA. All blots were washed twice for 30 min in 2XNaCI/Cit,
0.2% (massivol.) SDS at 60°C; the blot in B was subjected to an addi-
tional wash of 20 min at 60°C in 0,2XNaCl/Cit, 0.2% (masshol.) SDS.
The markers consist of Life Technologies I-kb ladder.
variation among the three Arubidopsis B’ isoforms is seen in the
amino and carboxy terminal regions of the proteins. In fact the
AtB‘y protein is slightly larger than the a and p isoforms due to
the presence of an extended carboxy terminus and an insertion
in its amino terminus. The Arubidopsis B’ regulatory subunits
are 43-47% identical and 57-61% similar to their yeast and
animal counterparts. As is the case among the Arubidopsis B’
isoforms, the largest variation among the different B’ subunits
occurs in the amino and carboxy termini of the proteins. The
most conserved region of the B’ protein consists of a region
spanning amino acids 170-424 of AtB’a, this region contains a
number of amino acid sections that show sequence conservation
among all organisms examined. The largest of these sections
spans positions 269-284 of AtB’a and a comparison of this
region to all five of the human B’ isoforms indicates perfect
conservation among all the human B‘ proteins as well.
Genomic organization of the Arabidopsis B‘ genes. In an effort
to establish the genomic organization of the B’-encoding genes
in Arabidopsis, we performed a series of Southern blots. Fig. 2
shows the results from these experiments. We began by hybrid-
izing a blot of Arabidopsis genomic DNA with a mixed probe
consisting primarily of the coding region of each of the three
AtB’ cDNAs (Fig. 2A). This probe yielded five hybridizing
bands in Arubidopsis DNA digested with either EcoRI or
HindIII. We next attempted to assign these hybridizing frag-
ments to specific AtB’ genes. For this purpose we developed
gene-specific probes for each of the B‘ cDNAs. Fig. 2B shows
that the 3.1-kb EcoRI fragment and the 4.8-kb HindIII fragment
of Fig. 2A correspond to the AtB‘a gene. The probe used for
this experiment consists of a short segment from the 3’ end of
the AtB’a cDNA (spanning from the Sac1 site at position 1643
to the end of the cDNA at position 1920). Fig. 2 C shows that,
under the hybridization conditions used for genomic Southern
159
1 2 3 4 5
9-
4-
2-
1-
kb
Fig. 3. Genomic organization of the B’ PP2A regulatory subunit gene
in a number of different plant species. Genomic DNA was isolated
from seedlings of: (1) corn (Zea mays); (2) wheat (Trificurnaestivuni);
(3) pea (Pisum sativurn); (4) bean (Phaseolus vulgaris) and (5) radish
(Raphanus safivus), digested with EcoRI and subjected to agarose gel
electrophoresis. The blot was probed with a mixture of the three Arabi-
dopsis B’ cDNAs as described for Fig. 2. The markers consist of Life
Technologies 1-kb ladder.
blots, the PCR fragment described in the above section serves
as a probe specific for AtB’p. This probe hybridizes to a single
EcoRI fragment and two of the HindIII fragments noted in
Fig. 2A, consistent with the presence of a single HindIII site
detected within the PCR fragment. Lastly (Fig. 2D), AtBy was
assigned to the 2.9-kb EcoRI fragment of Fig. 2A and a 1.5-kb
HindIII fragment. This probe consisted of a small section of the
5‘ region of the AtB’y cDNA spanning from the HindIII site at
position 310 to the EcoRI site at position 576. The HindIII frag-
ment detected by this probe is not noted in Fig. 2A as the seg-
ments of AtB’y used as probes for these two blots do not over-
lap. Our result indicates that most of the hybridizing fragments
identified in Fig. 2A correspond to the three AtB’ cDNA iso-
lated during our library screen. The two hybridizing fragments
in each lane of Fig. 2A that are not assigned to AtB’a, p or y
might reflect additional B’ genes or might correspond to regions
of AtB‘ a, p or y that do not hybridize to the AtB’ gene-specific
probes used in our experiments. Taken together, our results indi-
cate that at most five genes encoding the B’ regulatory subunit
are present in the Arubidopsis genome.
In addition to examining the genomic organization of B’ in
Arabidopsis, we also tested other plant genomes for the presence
of homologs of B’. Fig. 3 shows a Southern blot of DNA iso-
lated form a variety of plants, each showing a number of frag-
ments that hybridize to the Arabidopsis B’ probe. Both represen-
tatives of the monocot (wheat and corn) and dicot (radish, pea
and bean) lineage appear to contain B’-like DNA sequences.
These results suggest that the possible regulation of plant PP2Ac
by a B’-type regulatory subunit is most likely ubiquitous.
Expression of the Arabidopsis B’ genes. In an effort to establish
the expression pattern of the Arabidopsis B’ genes, we per-
formed northern blots. For this purpose we isolated mRNA from
a variety of plant organs (Fig. 4). In order to establish the rela-
tive level of mRNA isolated from each organ, we first probed
our blot with an actin gene (Fig. 4D). Subsequently we used the
gene-specific probes described in Fig. 2 to examine the levels
of AtB‘a, p and y mRNA in flowers, stems, leaves, roots and
160 Latorre et al. ( E M J. Biochern. 245)

A 12345 B12345 c12345 D12345

kb

4.4-
2.4-
1.4-

.24-

Fig. 4. Expression of the Arabidopsis B’ regulatory subunit genes. Poly(A)-rich mRNA was isolated from flowers (l), cotyledons (2), stems (3),
leaves (4) and roots (5) of Arabidopsis (Columbia) plants. The mRNA was separated by formaldehyde gel electrophoresis and the resulting gel
blotted to a nylon membrane. Blots were probed as described in Fig. 2. (A) AtB’a-specific probe; (B) AtB;C-specific probe; (C) AtB’y-specific
probe; (D) Brassica actin probe. The markers consist of Life Technologies RNA ladder.

A 1 2 3 identified. In all Arabidopsis organs the larger mRNA appears

-+ -+ -+ to be present at lower levels than the 1.8-kb mRNA. In addition,
kb the 1.5-kb mRNA is not detectable in stems and leaves but this
might be due to the relatively low levels of RNA isolated from
these organs (Fig. 4D). The identification of different sized
mRNAs derived from the AtB‘y gene indicates that the primary
2.4- transcript produced from this gene might be alternatively
spliced, as is the case for a number of the mammalian B‘ genes
1.4-
(McCright et al., 1996; Csortos et al., 1996; Zolnierowicz et al.,
1996).
As the B’ regulatory subunit homolog of S. cerevisiae ap-
pears to be involved in the heat-stress response (Shu and Hall-
3 4 berg, 1995), we examined if heat shock could affect the expres-
B 1 2
-+ -+ -+ -+ sion of the B‘ genes of Arubidopsis. For this purpose Arabi-
dopsis seedlings were subjected to heat stress at 37°C for 2 h
kb
and B’ mRNA levels were examined using northern blots. As a
control for the heat-shock response, we probed the blots with
the HSP17.6 gene of Arubidopsis (Helm and Vierling, 1989;
Fig. 5, A1 and Bl). In addition we compared the relative level
2.4- of mRNA in each lane of our blots using an actin probe (Fig. 5 ,
1.4- A2 and B2). Finally we examined the levels of the B’ mRNAs
in control and heat-shocked seedlings. Fig. 5 (A3, B3 and B4)
show that the levels of AtB’a mRNA, AtB’P mRNA and
Fig. 5. Expression of the B‘ regulatory subunit genes of Arabidopsis 2.5-kb and 1.8-kb AtB’y mRNAs are unresponsive to heat shock.
during heat stress. Ten-day-old Arubidopsis seedlings were either main- However, the level of the 1.5-kb AtB’y mRNA appears to
tained at 23°C (-) or heat-shocked at 37OC (+) for 2 h. Poly(A)-rich increase significantly in heat-stressed seedlings. Our results indi-
mRNA was isolated from the seedlings, separated by gel electrophoresis cate that the Arabidopsis AtB’ genes are differentially expressed
and blotted to a nylon membrane. Two separate membranes were pre- in response to heat shock and suggests that a heterotrimeric
pared (A and B) and hybridized sequentially with the following probes: PP2A enzyme containing the B’y protein (derived from the
HSP17.6 (Al, BI); actin (A2, B2); AtB’n-specific probe (A3); AtB’P- 1.5-kb B’y mRNA) might possibly play a role in the heat-stress
specific probe (B3); AtB’y-specific probe (B4). The markers consist of response in plant cells.
Life technologies RNA ladder.

cotyledons. Our results indicate that each of the AtB‘ genes is
expressed in all Arubidopsis organs albeit at different relative
levels. The AtB‘a mRNA is approximately 2.1 kb (Fig. 4A), in-
dicating that the AtB’u cDNAs we identified are near full length.
AtB’a mRNA is detected in all Arabidopsis organs; however, a
comparison of the relative levels of RNA present in each lane
(Fig. 4D) indicates that the AtB’a mRNA accumulates to a
higher degree in leaves compared to roots and stems. The AtB‘P
mRNA is approximately 2.1 -2.2 kb (Fig. 4B), in good
agreement with the 2.2-kb size of the AtB’P cDNA, and tran-
scripts derived from the this gene appear to accumulate to their
highest levels in cotyledons and flowers. Finally, hybridization
of the AtB’y-specific probe with Arubidopsis mRNA detects
mRNAs of approximately 2.5, 1.8 and 1.5 kb (Fig, 4C). These
mRNAs are slightly larger or smaller than the AtB‘y cDNA we
DISCUSSION
In this paper we present evidence regarding the organization
and expression of the PP2A B’ regulatory subunit genes of Aru-
bidopsis thuliunu. Our results show that the Arabidopsis genome
contains a number of genes encoding this subunit, three of which
we have characterized in detail. The proteins encoded by the
AtB’a and AtB’P genes are highly similar (90%) while the third
B’ subunit gene (AtB’y) encodes a protein that is more divergent
but equidistantly related to the B‘a and B’P isoforms. These
comparisons indicate that the AtB’a and AtB’P genes most likely
arose by a recent gene duplication event. However, as the 5’ and
3‘ untranslated regions of the AtB’u and AtB‘P genes show little
similarity, this gene duplication event must have occurred long
enough ago to allow for extensive mutations to accumulate in
the untranslated sections of these genes.
 Latorre et al. ( E M J. Bioclzem. 245)
Alignment of the three Arabidopsis B‘ isoforms supports the
notion that the a and p isoform arose through a recent gene
duplication and identifies sections of the B’ protein that are vari-
able or conserved in plants. First, the a and p isoforms show
extensive similarity along their entire length, with the only nota-
ble divergence occurring in their very termini, again suggesting
they are closely related. Second, the y B’ isoform shows conser-
vation with a and p in its core region but diverges significantly
in both sequence and length in its amino and carboxy terminus.
These comparisons indicate that the core region of the B’ protein
is most likely necessary for the basic interaction of the B’ sub-
unit with the A and PP2Ac subunit of PP2A, while the amino
and carboxy termini of B’ might control the specific effect of
each isoform on PP2A specificity and activity.
Alignment of the Arabidopsis B‘ proteins with those of yeast
and animal supports the notion that the core region of B’ is criti-
cal for its basic interaction with the other PP2A subunits (Shu
and Hallberg, 1995; McCright and Virshup, 1995). The region
spanning amino acids 170-424 of AtB’a are highly similar
among all B‘ proteins examined to date (McCright and Virshup,
1995; Csortos et al., 1996; McCright et al., 1996; Tehrani et al.,
1996; Zolnierowicz et al., 1996; Wilson et al., 1994). On the
other hand, the amino and carboxy termini of the various B’
proteins show variation in both sequence and length. Tehrani et
al. (1996) and McCright et al. (1996) have identified the carboxy
terminus of several mammalian B’ isoforms as possibly being
responsible for the subcellular targeting of PP2A and as being
the subject for phosphorylation, hence supporting the notion that
it is the variable regions of B’ that control, in an isoform-specific
manner, the function of PP2A.
As indicated above, a number of mammalian B’ regulatory
subunits appear to function in part as nuclear targeting subunits
for PP2A (Tehrani et al., 1996; McCright et al., 1996). As plant
PP2A activity has been found associated with the nuclear com-
partment (Sheen, 1993), we examined the Arubidopsis B’ pro-
teins for possible consensus nuclear localization sequences
(Raikhel, 1992). However no such consensus sequences were
identified in the three currently characterized B’ proteins. This
analysis does not rule out the possible role of the plant B’ pro-
teins in the subcellular targeting of PP2A complexes as several
of the nuclear-localized mammalian B‘ subunits show no obvi-
ous nuclear targeting sequences (Tehrani et al., 1996; McCright
et al., 1996). Indeed, a number of known nuclear proteins lack
consensus nuclear targeting sequences but contain at least two
clusters of basic amino acids separated by a spacer of variable
length (Dingwall and Laskey, 1991). These types of sequences
are notable in the carboxy terminus of both the AtB’a and AtB’p
proteins. Thus, until B’ subcellular localization studies are per-
formed, we cannot rule out the possibility that B’ might be a
nuclear protein which, if shown to interact with other subunits
of Arabidopsis PP2A, might serve as a nuclear targetting subunit
for plant PP2A.
An examination of the expression pattern of the three Arabi-
dopsis B‘ genes shows that each gene is expressed in all Arabi-
dopsis organs albeit at different levels. Hence the B‘ proteins
appear to have a basic function in plants but certain isoforms
might have unique roles in particular cells, tissues or organ
types. For example the AtB’a mRNA appears to accumulate to
higher levels in leaves compared to roots and stems, while the
mRNA derived from AtB’/3 preferentially accumulates in flow-
ers and cotyledons. Finally the 1.5-kb B’y mRNA appears to be
present only in flowers, roots and cotyledons. A number of
studies have implicated PP2A in specific processes associated
with each of these organs (Sheen, 1993; Smith et al., 1994;
Huber et al., 1994; Rundle et al., 1993; MacKintosh et a]., 1994)
and hence our results hint at a possible role for certain PP2A
161
heterotrimeric complexes in controlling particular developmen-
tal andor metabolic plant processes.
As is the case in plants, a number of mammalian B‘ genes
are differentially expressed (Csortos et al., 1996; McCright et
al., 1996). In addition variation in the type of B‘ subunit present
in mammalian cells is often increased because of alternative
splicing of B’ transcripts. These alternatively spliced transcripts
encode B’ proteins with conserved cores but unique amino or
carboxy termini (Csortos et al., 1996; Tehrani et al., 1996;
McCright et al., 1996; Zolnierowicz et al., 1996). We thus exam-
ined all the B’ cDNAs isolated during this study for possible
evidence regarding alternative splicing of the Arubidopsis B’
mRNAs. However, none of the AtB’p cDNAs obtained in our
experiments showed evidence for alternative splicing, nor was
this the case for cDNAs derived from the AtB’a gene (we were
unable to perform this analysis for the AtB’y gene as we only
had one cDNA available from this gene). Nonetheless it is of
interest to note that northern blot data indicate that the AtB’y
gene produces a number of different sized transcripts and that
all these mRNAs are large enough to encode the predicted B‘y
proteins presented in Fig. 1 with the exception of the 1.5-kb
mRNA. Fig. 1 predicts that a B’y mRNA would need to be at
least 1565 nucleotides long, excluding the 5’ and 3’ untranslated
regions, to encode the 59-kDa B’y protein. Hence it would ap-
pear that, since the probe used to detect the 1.5-kb mRNA is
derived from the 5‘ region of AtB’y, the B’y polypeptide encoded
by the 1.5-kb mRNA might resemble B’y of Fig. 1 in its amino
terminus but differ in sequence at its carboxy terminus. These
differences might be the consequence of alternative splicing of
the primary transcript derived from the AtB’y gene; however,
future more extensive experimental analyses will be necessary
to establish if the AtB’y mRNA is indeed alternatively spliced,
and to determine if such splicing adds to the variety of B type
regulatory subunits present in Arubidopsis cells.
Finally in S. cerevisiae the B‘ regulatory subunit appears to
be involved in the response of yeast cells to heat stress (Shu and
Hallberg, 1995). While levels of mRNA derived from the yeast
B’ gene do not alter in response to heat shock, overexpression
of the this gene is able to rescue temperature-sensitive mutations
in the nuclear-encoded mitochondria1 HSP60 heat-shock protein.
This rescue appears to be mediated by increased expression of
other chaperonin-encoding genes ; indeed, S. cerevisiae cells
lacking B’ show decreased levels of mRNA for these chapero-
nins under normal growth conditions (30°C). Taken together,
these data suggest a possible role of B‘-containing PP2A com-
plexes in yeast gene expression, a suggestion supported by the
well documented role of PP2A in animal transcription (Alberts
et al., 1993; Wadzinski et al., 1993; Lerga et al., 1995). Our
results indicate that the level of most of the mRNAs derived
from the Arabidopsis B‘ genes show little response to heat
shock, as is the case in S. cerevisiae, and as such these results
shed little light on the role of B’ within plant cells. However,
the level of the 1.5-kb mRNA derived from AtB’y rises dramati-
cally in response to heat stress, suggesting that if PP2A com-
plexes containing the B‘ protein encoded by the 1.5-kb mRNA
exists within plant cells, these complexes might possibly per-
form specific functions in these cells during heat stress. Based
on the predicted role of B’ in S. cerevisiue, this might involve
specific regulation of transcription during plant growth under
non-optimal conditions. However, no data are currently available
to substantiate this idea. Thus, future experiments, including
gene expression patterns in Arabidopsis plants with reduced
levels or mutant forms of B’y, will be necessary in order to
establish what role, if any, the B’ protein has in the plant cell
stress response.
162 Latorre et al. ( E m J. Biochem. 245)
In conclusion we have identified three ubiquitously ex-
pressed Arabidopsis genes encoding the B’ regulatory subunit of
PP2A. Taken together with expression data from the other genes
encoding Arabidopsis PP2A subunits, our results predict that a
total of 75 distinct PP2A complexes might be able to assemble
within each Arabidopsis cell (Ariiio et al., 1993; Casamayor et
al., 1994; Rundle t t al., 1995; Corum et al., 1996). It is thus
possible to envision how PP2A is able to control so many dis-
tinct plant processes and to begin to establish, using these newly
identified genes as tools, which specific PP2A complexes func-
tion to control specific plant metabolic and developmental pro-
cesses.
We would like to thank Dr E. Vierling (University of Arizona) for
kindly providing the HSP17.6 probe. In addition we thank Trevor Rundle
for help in preparing figures, April Livengood for help with northern
blot experiments, and Dr Roger Lumb for critical reading of the manu-
script. This research was supported by United States Department of
Agriculture Grants (94-37304-1 109 and 96-35304-3863) awarded to
S. J. R.
REFERENCES
Alberts, A. S., Deng, T., Lin, A,, Meinkoth, J. L., Schonthal, A., Mumby,
M. C., Karin, M. & Feramisco, J. R. (1993) Protein phosphatase 2A
potentiates activity of promoters containing AP-1 -binding elements,
Mol. Cell. B i d . 13, 2104-2112.
Ariiio, J., Perez-Callejon, E., Cunillera, N., Camps, M., Posas, F. & Fer-
rer, A. (1 993) Protein phosphatases in higher plants : multiplicity of
type 2A phosphatases in Arabidopsis thaliana, Plant Mol. Biol. 21,
475 -485.
Casamayor, A,, Perez-Callejon, E., Pujol, G., Ariiio, J. & Ferrer, A.
(1994) Molecular characterization of a fourth isoform of the catalytic
subunit of protein phosphatase 2A froin Arabidopsis thaliuna, Plant
Mol. Biol. 26, 523-528.
Corom, J. W., Hartung, A. J., Stamey, R. T. & Rundle, S. J. (1996)
Characterization of DNA sequences encoding a novel isoform of the
55 kDa B regulatory subunit of the type 2A protein serinekhreonine
phosphatase of Arabidopsis thaliana, Plant Mol. Biol. 31, 419-
421.
Csortos, C., Zolnierowicz, S., Bako, E., Durbin, S. D. & DePaoli-Roach,
A. A. (1996) High complexity in the expression of the B’ subunit of
protein phosphatase 2A,,, J . Biol. Chem. 271, 2578-2588.
Dingwall, C. & Laskey, R. A. (1991) Nuclear targeting sequences - a
consensus? Trends Biochem. Sci. 16, 478-481.
Garbers, C., DeLong, A,, Derutre, J., Bemasconi, P. & So11, D. (1996)
A mutation in protein phosphatase 2A regulatory subunit A affects
auxin transport in Arahidopsis, EMBO J. 15, 2115-2124.
Gianfagna, T. J. & Lawton, M. A. (1995) Specific activation of soybean
defense genes by the phosphoprotein phosphatase inhibitor okadaic
acid, Plant Sci. 109, 165- 170.
Helm, K. W. & Vierling, E. (1989) An Arabidopsis thaliana cDNA clone
encoding a low molecular mass heat shock protein, Nucleic Acids
Res. 17, 7995.
Henikoff, S. (1984) A unidirectional digestion with exonuclease 111 cre-
ates targeted breakpoints for DNA sequencing, Gene 28, 351 -359.
Hinnebusch, A. (1984) Evidence for translational regulation of the acti-
vator of general amino acid control in yeast, Proc. Natl Acad. Sci.
USA 81, 6442-6446.
Huber, S. C., Huber, J. L. & Kaiser, W. M. (1994) Control of plant
enzyme activity by reversible protein phosphorylation, Znt. Rev. Cy-
mi. ~ 4 7 - 9 8 .
Kieber, J. J., Rothenberg, M., Roman, G., Feldmann, K. A. & Ecker, J.
R. (1993) CTRI, a negative regulator of the ethylene response path-
way in Arabidopsis, encodes a member of the Raf family of protein
kinases, Cell 72, 427-441.
Kuo, A., Cappelluti, S., Cervantes-Cervantes, M., Rodriguez, M. &
Bush, D. S. (1996) Okadaic acid, a protein phosphatase inhibitor,
blocks calcium changes, gene expression, and cell death induced by
gibberellin in wheat aleurone cells, Plant Cell 8, 259-269.
Lerga, A,, Belandia, B., Delgado, M. D., Cuadrado, M. A,, Richard, C.,
Ortiz, J. M., Martin-Perez, J. & L e h , J. (1995) Down-regulation of
c-myc and max genes is associated to inhibition of protein phospha-
tase 2A in K562 human leukemia cells, Biochem. Biophys. Res. Com-
mun. 215, 889-895.
Li, W., Luan, S., Screiber, S. L. & Assmann, S. M. (1994) Evidence for
protein pbosphatase 1 and 2A regulation of K’ channels in two types
of leaf cells, Plant Physiol. 106, 963-970.
MacKintosh, C., Lyon, G . D. & MacKintosh, R. W. (1994) Protein phos-
phatase inhibitors activate anti-fungal defence responses of soybean
cotyledons and cell cultures, Plant J. 5, 137-147.
McCright, B. & Virshup, D. M. (1995) Identification of a new family
of protein phosphatase 2A regulatory subunits, J. B i d . Chem. 270,
26 123-26 128.
McCright, B., Rivers, A. M., Audlin, S. & Virshup, D. M. (1996) The
B56 family of protein phosphatase 2A (PP2A) regulatory subunits
encodes differentiation-induced phosphoproteins that target PP2A to
both nucleus and cytoplasm, J. Biol. Chem. 271, 22081-22089.
Mumby, M. C. & Walter, G. (1993) Protein serinekhreonine phospha-
tases: structure, regulation and function in cell growth, Physiol. Rev.
73, 673-699.
Murashige, T. & Skoog, F. (1962) A revised medium for rapid growth
and bio assay with tobacco tissue cultures, Physiol. Plant. 15, 473-
497.
Nasrallah, J. B., Rundle, S. J. & Nasrallah, M. E. (1994) Genetic evi-
dence for the requirement of the Brassica S-locus receptor kinase
gene in the self-incompatibility response, Plant J. 5, 373-384.
Pearson, W. & Lipman, D. (1988) Improved tools for biological se-
quence comparison, Proc. Natl Acad. Sci. USA 85, 2444-2448.
Raikhel, N. V. (1992) Nuclear targeting in plants, Plant Physiol. ZOO,
1627-1632.
Rosenblum-Vos, L. S., Rhodes, L., Evangelista, C. C., Boayke, K. A. &
Zitomer, R. S. (1991) The ROX3 gene encodes an essential nuclear
protein involved in CYC7 gene expression in Saccharomyces cerevi-
siae, Mol. Cell. Biol. 11, 5639-5647.
Rundle, S. J., Nasrallah, M. E. & Nasrallah, J. B. (1993) Effects of
inhibitors of protein serinekhreonine phosphatases on pollination in
Brassica, Plant Physiol. 103, 1165-1171.
Rundle, S. J., Hartung, A. J., Corum, J. W. & O’Neill, M. (1995) Charac-
terization of a cDNA encoding the 55 kDa B regulatory subunit of
Arabidopsis protein phosphatase 2A, Plant Mol. Biol. 28, 257-
266.
Sheen, J. (1993) Protein phosphatase activity is required for light-induc-
ible gene expression in maize, EMBO J. 12, 3497-3505.
Shu, Y. & Hallberg, R. (1995) SCSI, a multicopy suppressor of hsp60-
ts mutant alleles, does not encode a mitochondrially targeted protein,
Mol. Cell. B i d . 15, 5618-5626.
Slabas, A. R., Fordham-Skelton, A. P., Fletcher, D., Martinez-Rivas, J.
M., Swinhoe, R., Croy, R. R. D. & Evans, I. M. (1994) Characteriza-
tion of cDNA and genomic clones encoding homologues of the
65 kDa regulatory subunit of protein phosphatase 2A in Arabidopsis
thaliana, Plant Mol. Biol. 26, 1125- 1138.
Smith, R. D., Wilson, J. E., Walker, J. C. & Baskin, T. I. (1994) Protein-
phosphatase inhibitors block root hair growth and alter cortical cell
shape in Arabidopsis roots, Planta 194, 516-524.
Smith, R. D. & Walker, J. C. (1996) Plant protein phosphatases, Annu.
Rev. Plant Physiol. Plant Mol. Biol. 47, 101-125.
Stein, J. C., Howlett, B., Boyes, D. C., Nasrallah, M. E. & Nasrallah, J.
B. (1991) Molecular cloning of a putative receptor protein kinase
gene encoded at the self-incompatibility locus of Brassica oleracea,
Proc. Natl Acud. Sci. USA 88, 8816-8820.
Tehrani, M. A,, Mumby, M. C. & Kamibayashi, C. (1996) Identification
of a novel protein phosphatase 2A regulatory subunit highly ex-
pressed in muscle, J . Bid. Chem. 271, 5146-5170.
Thiel, G. & Blatt, M. R. (1994) Phosphatase antagonist okadaic acid
inhibits steady-state K’ currents in guard cells of Vicia faba, Plant
J . 5, 727-733.
Wadzinski, B. E., Wheat, W. H., Jaspers, S., Peruski, Jr, L. F., Lickteig,
R. L., Johnson, G. L. & Klemm, D. J. (1993) Nuclear protein phos-
phatase 2A dephosphorylates protein kinase A-phosphorylated
CREB and regulates CREB transcriptional stimulation, Mol. Cell.
Biol. 13, 2822-2834.
Wera, S. & Hemmings, B. A. (1995) Serinekhreonine protein phospha-
tases, Biochem. J. 311, 17-29.
 Latorre et al. (ELKJ. Biochem. 245) 163
Wilson, R., Ainscough, R., Anderson, K., Baynes, C., Berks, M., Bon- ton, J., Thierry-Mieg, J., Thomas, K., Vaudin, M., Wilkinson-Sproat,
field. J., Burton, J., Connell, M., Copsey, T., Cooper, J., Coulson, A., J. & Wohldman, P. (1994) 2.2 Mb of contigous nucleotide sequence
Craxton, M., Dear, S., Du, Z., Durbin, R., Favello, A,, Fulton, L., from chromosome I11 of C. elegans, Nature 368, 32-38.
Gardner, A., Green, P., Hawkins, T., Hillier, L., Jier, M., Johnston, Zolnierowicz, S., Van Hoof, C., Andjelkovic, N., Cron, P., Stevens, I.,
L., Jones, M., Kershaw, J., Kirsten, J., Laister, N., Latreille, P., Light- Merlevede, W., Goris, J. & Hemmings, B. A. (1996) The variable
ning, J., Lloyd, C., McMurray, A,, Mortimore, B., O’Callaghan, M., subunit associated with protein phosphatase 2A, defines a novel
Parsons, J., Percy, C., Riflcen, L., Roopra, A., Saunders, D., Shown- multimember family of regulatory subunits, Biochem. J. 317, 187-
keen, R., Smaldon, N., Smith, A., Sonnhammer, E., Staden, R., Suls- 194.