Two mRNA species encoding calcium-dependent protein kinases are differentially expressed in sexual organs of Marchantia polymorpha through alternative splicing.

Abstract

In plants, calcium-dependent calmodulin-independent protein kinases (CDPKs) are the predominant calcium-regulated protein kinases and their genes are encoded by a multigene family. A CDPK gene was cloned from a liverwort, Marchantia polymorpha, which showed a high level of sequence similarities to other higher plant CDPK genes. The liverwort CDPK gene consisted of 9 exons and 8 introns. The 6th and 7th exons (Exon 6A and Exon 6B) were almost identical except for 4-amino acid substitutions, both of which coded for EF-hands in the calcium-binding domain. RT-PCR analysis revealed that two species of mature mRNA containing either Exon 6A or Exon 6B were generated from a single CDPK gene by mutually exclusive alternative splicing. Both histidine-tagged fusion proteins derived from cDNAs containing either Exon 6A or Exon 6B exhibited calcium-dependent protein kinase activity in vitro. Preferential accumulation of the mature mRNA with Exon 6A detected in male sexual organ implies possible sexual control of the ratio between the two CDPK isozymes through alternative splicing. Functions and evolution of CDPKs are discussed based on the structure and expression of the liverwort CDPK gene.

Full text

Page 1

Plant Cell Physiol. 40(2): 205-212 (1999)
JSPP © 1999
Two mRNA Species Encoding Calcium-Dependent Protein Kinases Are
Differentially Expressed in Sexual Organs of Marchantia polymorpha
through Alternative Splicing
Rie Nishiyama, Hiroshi Mizuno, Sachiko Okada, Tomoya Yamaguchi, Mizuki Takenaka,
Hideya Fukuzawa and Kanji Ohyama'
Laboratory of Plant Molecular Biology, Division of Applied Life Sciences, Graduate School of Agriculture, Kyoto University, Kyoto,
606-8502 Japan
In plants, calcium-dependent calmodulin-independ-
ent protein kinases (CDPKs) are the predominant calci-
um-regulated protein kinases and their genes are encoded
by a multigene family. A CDPK gene was cloned from a
liverwort, Marchantia polymorpha, which showed a high
level of sequence similarities to other higher plant CDPK
genes. The liverwort CDPK gene consisted of 9 exons and
8 introns. The 6th and 7th exons (Exon 6A and Exon 6B)
were almost identical except for 4-amino acid substitu-
tions, both of which coded for EF-hands in the calcium-
binding domain. RT-PCR analysis revealed that two spe-
cies of mature mRNA containing either Exon 6A or Exon
6B were generated from a single CDPK gene by mutually
exclusive alternative splicing. Both histidine-tagged fusion
proteins derived from cDNAs containing either Exon 6A or
Exon 6B exhibited calcium-dependent protein kinase ac-
tivity in vitro. Preferential accumulation of the mature
mRNA with Exon 6A detected in male sexual organ implies
possible sexual control of the ratio between the two CDPK
isozymes through alternative splicing. Functions and evo-
lution of CDPKs are discussed based on the structure and
expression of the liverwort CDPK gene.
Key words: Alternative splicing — Calcium-dependent
protein kinase — Liverwort — Plant — Sexual organs.
Change of intracellular calcium ion (Ca2+) level has
been observed in plants under many conditions (Trewavas
and Knight 1994). In animals, Ca2+ regulates the signal
transduction pathway by activating calcium/calmodulin-
Abbreviations: CaMK, calcium/calmodulin-dependent pro-
tein kinase; CDPK, calcium-dependent calmodulin-independent
protein kinase; EGTA, ethylene glycol bisOS-aminoethylether)-
N,MAr,A^-tetraacetic acid; RACE, rapid amplification of cDNA
ends; RT-PCR, PCR in conjunction with reverse transcription.
The nucleotide sequences reported in this paper have been
submitted to the NCBI, EMBL and DDBJ under accession num-
bers, AB017515 for CDPK gene, AB017516 for CDPK-A mRNA,
AB017517 for CDPK-B mRNA.
1 To whom correspondence should be addressed. Phone, +81-
75-753-6389; FAX, +81-75-753-6127; E-mail, kohyama@kais.
kyoto-u.ac.jp
dependent protein kinases (CaMKs) (Soderling 1996).
However, in plants, Ca2+ signal transduction is regulated
by CaMKs and CDPKs that have been identified as calci-
um-dependent calmodulin-independent serine/threonine
kinases (Stone and Walker 1995, Poovaiah et al. 1997).
CDPKs have been identified in many plants and protists,
and are encoded by multigene families. They show wide
diversity in a phylogenetic tree based on comparisons of
the amino acid sequences of CDPKs from Arabidopsis,
other plants, and protists (Hrabak et al. 1996). Plant
CDPKs are also divergent in responsiveness to Ca2+ levels,
expression patterns, subcellular localizations and physical
functions (Stone and Walker 1995, Furumoto et al. 1996,
Hong et al. 1996). Environmental stresses such as drought,
cold and salinity can induce expressions of specific CDPK
genes (Sheen 1996, Urao et al. 1994). Thus, it is suggested
that various plant CDPKs work coordinately under their
circumstances, but an understanding of their physiologi-
cal roles remain elusive.
Splicing of nuclear precursor mRNA to mature
mRNA is a central feature of eukaryotic gene expression
and takes place in the spliceosome (Simpson and Filipowicz
1996). The process requires precise recognition of 5' donor
and 3' acceptor splicing sites and ligation of contiguous
exons. One of the best-understood cases of alternative
splicing is the Sex-lethal gene in Drosophila that partici-
pates in controlling sex determination by alternative splic-
ing of the mRNA (Kelly and Kuroda 1995). Development
and differentiation of organs or tissues are also regulated
by alternative splicing in many other systems (Chabot 1996,
Edwalds-Gilbert et al. 1997). In these cases, external signals
determine the direction of cell differentiation by alternative
splicing of specific mRNAs.
At present, alternative splicing has not been reported
for plant CDPK genes. In this paper, we describe the
cloning of a novel CDPK gene from a liverwort, Mar-
chantia polymorpha, that produces two species of mature
CDPK mRNAs by alternative splicing.
Materials and Methods
Plant materials—The liverwort suspension-cultured cells
were originally derived from a female thallus and maintained on
205
by guest on October 14, 2011
pcp.oxfordjournals.org
Downloaded from

Page 2

206Alternative splicing of CDPK gene in M.polymorpha
1-M51C medium as previously described (Ono et al. 1979). The
male and female thalli were derived from a respective gemma
grown in the field of Kyoto City. They were sterilized and main-
tained on M51C solid agar medium (Ono et al. 1979). The male
and female sexual organs were collected from wild populations.
Preparation of genomic DNA and PCR—Genomic DNA was
prepared from male and female thalli using cetyltrimethylam-
monium bromide (Ausubel et al. 1987). The PCR reaction mix-
ture (20^1) contained 20 ng of genomic DNA, 10 mM Tris-HCl
(pH 8.3), 50 mM KC1, 0.1 figfiV^ gelatin, 1.5 mM MgCl2, 20^M
dNTPs, 0.1 fid of [a32P]dCTP, and 0.0125 units fiV' of Taq DNA
polymerase (PE Applied Biosystems), and 12.5 fM of each ar-
bitrary 15-mer oligonucleotide primer (primer 1: 5-NNNNNGGA-
TGCGAGT-3', primer 2: 5'-NNNNNCGAGTGGATG-3', primer
3: 5'-NNNNNGATTGGACCG-3', primer 4: 5'-NNNNNCCGCT-
CACTA-3', primer 5: 5'-NNNNNGGCGAGTCAC-3'; N is a mix-
ture of A, C, G and T). Thirty cycles of PCR (94°C for 1 min,
45°C for 2 min, 72°C for 1 min) were carried out. The PCR
products were separated by electrophoresis on 5% polyacryl-
amide gels and visualized by autoradiography.
Rapid amplification of cDNA ends (RACE)-PCR was per-
formed by using 1 fig poly(A)+ RNA and oligonucleotide primers,
5' cap primer-1 (5-CTGACCACGACCCAGTTCCCGTCC-3),
or 5' cap primer-2 (5-GAGGGTGTAAATGGATCGCACATC-
3').
Cloning and sequence analysis of PCR fragments—The 10
amplified PCR fragments which were unique to either male or
female genomes were recovered from polyacrylamide gels and
cloned into pBluescript II KS +. DNA sequencing was performed
by the dideoxy chain-termination method using ABI PRISM Dye
primer Cycle Sequencing kit (PE Applied Biosystems) and a DNA
sequencer (PE Applied Biosystems; Model 377). Sequence ho-
mology searches were performed with the Genbank, EMBL, Swiss
Prot, and PIR Databases.
DNA and RNA blot hybridizations—DNA fragments were
labelled with [a32P]dCTP using the Megaprime DNA Labelling
System (Amersham) and used as a probe. Two fig each of genomic
DNA isolated from female suspension-cultured cells was digested
with either Hindlll or Smal. The DNA fragments were elec-
trophoresed in a 0.7% agarose gel. After alkaline treatment and
blotting onto nylon membrane, hybridization was performed in a
solution containing 5 x Denhardt's reagent, 6 x SSPE, 0.5% SDS,
100 fig ml"1 denatured salmon sperm DNA, and 50% formamide
at 42°C (Ausubel et al. 1987). The filter was washed twice with a
solution containing 1 x SSPE and 0.1% SDS at room tempera-
ture followed by washing with that of 0.1 x SSPE and 0.1% SDS
at 42CC for 10 min and then autoradiographed.
Poly(A)+ RNA was isolated from the female suspension-
cultured cells using the PolyATtract System 1000 (Promega) ac-
cording to the manufacturer's instruction. One fig of poly(A)+
RNA was electrophoresed in a 0.8% denaturing agarose gel con-
taining formaldehyde (Ausubel et al. 1987) and transferred onto
nylon membrane. Hybridization was carried out according to the
method described above for the genomic Southern hybridiza-
tion.
RT-PCR and analysis of alternatively spliced products—
Using 0.5 fig of Poly(A)+ RNA from female suspension-cultured
cells, the reverse transcription was carried out using the Ready
To Go T-Primed First-Strand Kit (Pharmacia Biotech) according
to the manufacturer's instruction. The first strand of the cDNA
mixture was then amplified in a 100//I solution containing 0.5 fiM
oligonucleotide primers, EX5-F (5-AAAGCTTGCAGAGTCA-
GAAGTGC-3') and EX7-R (5'-GAGCTCCCATCCCGTGCTTA-
A-3'), 10 mM Tris-HCl (pH 8.3), 50 mM KC1, 1.5 mM MgCl2, 0.2
mM dNTP mixture, and 0.025 units /il"1 of Taq DNA poly-
merase (TAKARA). Twenty five cycles of PCR (94°C for 1 min,
57°C for 1 min, 72°C for 1 min) were carried out. The PCR
products were subjected to either a single or double digestions
with EcoJlll and Fokl, separated by electrophoresis on 5%
polyacrylamide gels, and visualized by staining with ethidium
bromide. For analysis of molar ratio of mRNA species, respec-
tive poly(A)+ RNA samples prepared from male, female sexual
organs, and male, female thalli were subjected to RT-PCR (PCR
in conjunction with reverse transcription) in the presence of 5 fid
of [a32P]dCTP. The PCR products were then washed five times
with TE-buffer using membrane filters SUPREC 02 (TAKARA).
Approximately 1,000 cpm of PCR products were digested with
EcoT221 and Fokl, separated by electrophoresis using 5% poly-
acrylamide gel, and visualized using a image analyzer (BAS 2000,
FUJIFILM).
In vitro protein kinase assay—Two cDNA fragments coding
for CDPK-A (containing Exon 6A) and CDPK-B (containing
Exon 6B) were amplified by PCR using the 5'-primer (5-
GGAGTACATATGGGCAACTGCGTG-3) and 3'-primer (5-
CAGGATCCAACCAGGTTGACATA-3)
with translation initiation codon and BamHI site used for sub-
cloning are indicated by underlines). The PCR products were
digested with Ndel and BomHI, and the resulting fragments were
inserted into the corresponding sites of the pET-15b vector
(Novagene). Both of these cloned fragments were verified by DNA
sequencing.
The fusion proteins with histidine-tags were expressed in
E. coli BL21 (DE3) pLysS and purified according to the manu-
facturer's instruction. The protein kinase assays were performed
at 25°C for 15 min in a 20 fi\ solution containing 50 mM Tris-
HCl (pH 7.6), 5 mM MgCl2, 1 mM CaCl2> 4^Ci of [y32P]ATP, 10
fig of a-casein (Sigma), and 500 ng of fusion protein. To meas-
ure the kinase activities in the absence of Ca2+ ions, 5 mM EGTA
(ethylene glycol bis(/J-aminoethylether)-iV,./V,iV,./V'-tetraacetic acid)
was added as a chelating agent. Reaction mixtures were subject-
ed to 12% SDS-polyacrylamide gel electrophoresis and to auto-
radiography.
(Ndel overlapping
Results
Isolation and characterization of the CDPK gene from
a liverwort—Genomic DNAs of male and female thalli
were amplified using five arbitrary 15-mer oligonucleotide
primers for the purpose of detection of differences be-
tween male and female genomes. Ten PCR products which
were detected to either male or female genomes were cloned
and their sequences were determined. By chance, a deduced
amino acid sequence from one of the clones, F3C con-
taining 280-bp PCR fragment amplified using primer-2,
showed significant similarities with a consensus sequence
(called EF-hand) of CDPK genes. In order to obtain the
full-length of the putative CDPK gene, a genomic library
for the female suspension-cultured cells of liverwort
(Akashi et al. 1996) was screened using the PCR frag-
ment in the F3C as a probe. One lambda clone containing
a 5.4-kb Nael-Sacl fragment was selected and the nucleo-
tide sequence of the 5.4-kb fragment was determined. By
comparing the nucleotide sequence with the databases,
by guest on October 14, 2011
pcp.oxfordjournals.org
Downloaded from

Page 3

Alternative splicing of CDPK gene in M. polymorpha
207
1 2
3.9 kb
B
Ex0n6A A3VDGDGTIDYLE
IM IMIIII
Exon 6B A 3VDGNGTIDYLE
T T
'ITATMHLNKIDKEDHLYAAFQHF3GDHSG
I I I I M I i l I M I M M I I I I
'ITATMHLNKIEKEDHLYAAFQHFDEDSSG
I I II
Fig. 1 Structure of the liverwort CDPK gene (A) and
comparison of partial amino acid sequences deduced from Exon
6A and Exon 6B (B). (A) Solid boxes represent exons and two
nearly identical exons, Exon 6A and Exon 6B, are shown as
shaded boxes with 6A and 6B, respectively. Arrow heads with F
and R indicate the positions of RT-PCR primers, EX5-F and
EX7-R, respectively. Restriction sites: B, Bglll; E, EcoKl; H,
Hindlll; N, Nael; Sc, Sad; SI, Sail; Sm, Smal; and X, Xbal. The
3.9-kb DNA fragment used as probe for genomic Southern hy-
bridization and RNA blot analyses is shown. (B) Amino acid
residues are shown in single letter symbols. Identical amino acid
residues are shown by horizontal bars. Amino acid substitutions
are indicated by arrow heads. Calcium-binding loops of the EF-
hands are boxed.
coding regions for CDPK gene were predicted as shown in
Fig. 1A. The boundaries between exons and introns were
determined by sequencing cDNA clones derived from fe-
male liverwort suspension-cultured cells (see below). The 5'
end of the first exon was confirmed by the RACE-PCR
(data not shown). The CDPK gene consisted of 9 exons and
8 introns. The lengths of the 6th and 7th exons (Exon 6A
and Exon 6B) were 128 bp long each and identical to each
other except for 20 nucleotide substitutions which corre-
spond to four amino acid substitutions in the region of the
deduced 43 amino acids (Fig. IB).
Copy number of CDPK gene and its expression in the
liverwort cells—To know whether the isolated CDPK
gene is a single copy or belongs to a multigene family,
genomic DNA of female suspension-cultured cells was
probed with the 3.9-kb Sall-Sacl fragment containing the
entire coding region of the genomic CDPK gene (Fig. 1 A).
When the genomic DNA was digested with Hindlll, two
bands of 1.9 kb and 4.9 kb were detected (Fig. 2, lane 1).
This result coincides with the fact that a single Hindlll site
is located in the CDPK coding region (see Fig. 1A). When
the genomic DNA was digested with Smal, the probe hy-
bridized to a single 15 kb band (Fig. 2, lane 2). Further-
more, RNA blot analysis using poly(A)+ RNA clearly
showed a single transcript of 2.6 kb (Fig. 2, lane 3). These
hybridization data clearly indicate that the liverwort ge-
nome contains a single copy of this type of CDPK gene.
But the organization and sequence similarity of Exon 6A
and Exon 6B suggested that this gene possibly encodes two
i
4.9 kb
«* 1.9 kb
Fig. 2 Genomic Southern and RNA blot analyses of CDPK
gene. Ten fig of genomic DNA digested by Hindlll (lane 1) or
Smal (lane 2), and one ng of poly(A)+ RNA (lane 3) were elec-
trophoresed and probed by the 3.9-kb Sall-Sacl fragment (see
Fig. 1A). Molecular sizes are indicated.
isoforms which are translated from alternatively spliced
mRNA molecules.
Accumulation of alternatively spliced mRNA mole-
cules—To identify actual splicing products containing
Exon 6A and/or Exon 6B, the region of the mRNA se-
quences ranging from Exon 5 to Exon 7 was amplified by
RT-PCR using a set of primers, EX5-F and EX7-R (as
shown in Fig. 1A). A PCR product of 224 bp was detect-
ed as a single band (Fig. 3A, lane 1), which indicates that
the 224-bp fragment derived from either Exon 6A or Exon
6B (Fig. 3B), and a mature mRNA containing both Exon
6A and Exon 6B was not detected. Based on their nucleo-
tide sequences, Exon 6A and Exon 6B have a unique site
for £coT22I and Fokl, respectively (Fig. 3B). In fact, the
PCR product digested with £coT22I produced three frag-
ments; 224-bp fragment derived from Exon 6B, and 127-
bp and 97-bp fragments from Exon 6A (Fig. 3A, lane 2).
When digested with Fokl, it gave three fragments; 224-
bp fragment derived from Exon 6A, and 160-bp and 64-
bp fragments from Exon 6B (Fig. 3A, lane 3). In addition,
by double digestions with £coT22I and Fokl, it gave four
fragments; 127-bp and 97-bp fragments from Exon 6A,
and 160-bp and 64-bp fragments from Exon 6B (Fig. 3A,
lane 4). Thus, these results indicate that two very similar
mRNAs having either Exon 6A or Exon 6B are processed
from a single CDPK gene by alternative splicing, and the
post-transcriptional product derived from this gene is a
mixture of two kinds of mRNA molecules having either
Exon 6A or Exon 6B.
Differential expression of alternatively spliced prod-
ucts—We further surveyed the ratio of the accumulation of
mRNAs having either Exon 6A or Exon 6B in different
organs of the liverwort. Due to differences in the restriction
sites of Exon 6A and Exon 6B, the ratio of alternatively
spliced mRNAs can be estimated by comparing the levels
by guest on October 14, 2011
pcp.oxfordjournals.org
Downloaded from

Page 4

Alternative splicing of CDPK gene in M. polymorpha
2 3 4 (bp)
1224
160
127
97
64
B
97
ECOT22I
127
Exon 5 Exon 6A
. 64 «*.'
Exon 7
160
Exon 5 Exon6BExon 7
Fig. 3 Restriction analysis of RT-PCR products. (A) First
strands of cDNA mixture were generated from poly(A)+ RNA
isolated from female liverwort suspension-cultured cells and am-
plified with EX5-F and EX7-R primers (lane 1). The RT-PCR
products were digested with EcoT22l (lane 2), Fokl (lane 3) and
both enzymes (lane 4). (B) Schematic illustration of RT-PCR
products shown in (A). Restriction endonuclease EcoTlll cleaved
RT-PCR product derived from Exon 6A into two fragments with
sizes of 97 bp and 127 bp, and Fokl digested that containing Exon
6B into two fragments with the sizes of 64 bp and 160 bp.
of RT-PCR products followed by restriction digestions, if
the PCR products are generated under non-saturating con-
ditions. Since the two kinds of PCR products with Exon
6A or Exon 6B have the same length of 224 bp and they
have almost identical nucleotide sequences which differ by
20 nucleotides in the exon regions, the molar ratio of the
two PCR products should reflect that of the two kinds of
mRNA molecules produced in respective male, female
sexual organs, and male, female vegetative organs called
thalli. Therefore, poly(A)+ RNA samples were isolated
from each organs and subjected to RT-PCR for 23 cycles
under non-saturating condition (data not shown) in the
presence of [a32P]dCTP using a set of primers, EX5-F and
EX7-R. The generated 224-bp PCR products were com-
pletely digested with EcoT221 and Fokl, electrophoresed in
5% polyacrylamide gel, and then followed by autoradiog-
raphy (Fig. 4A). Relative intensities of radioactivity in each
band in the autoradiogram was estimated by image ana-
lyzer and shown in Fig.4B. From the autoradiogram, the
percentages of the accumulated PCR fragments with Exon
6A in the total PCR products prepared from female sex
organ and thalli were from 38% to 44% (lanes 2-4). In
contrast, that of the accumulated PCR fragments with
1
s
(bp)
160
127
97
64
]6A
6B
Fig. 4 Alternatively spliced products in different organs of the
liverwort. (A) Poly (A)+ RNA was isolated from male sexual
organs (lane 1), female sexual organs (lane 2), male thalli (lane 3)
and female thalli (lane 4), and subjected to RT-PCR. The PCR
products labelled with 32P were completely digested with £coT22l
and Fokl. (B) Schematic illustration of the relative intensities of
radioactivity which were obtained by scanning of the autoradio-
gram as shown in panel (A). Total radioactivities of the four
bands in each lane are shown as 100%.
Exon 6A prepared from male sex organ was approximately
60% (lane 1). These data suggest that a higher level of the
mRNA transcripts with Exon 6A are accumulated in the
male sexual organ compared with those of the mRNA with
Exon 6B. This analysis was repeated ten times with two
different preparations of poly(A)+ RNA samples and two
different preparations of cDNA samples and the almost
same accumulation pattern of RT-PCR products was ob-
served in each autoradiograms (data not shown).
Purification of two fusion proteins of CDPK and
protein kinase assay—The cDNA fragments coding for
CDPK isoforms, CDPK-A derived from Exon 6A and
CDPK-B from Exon 6B, were obtained by RT-PCR using
a set of primers, 5'-primer and 3'-primer, and then inserted
into the pET-15b expression vector. The two fusion pro-
teins which were designed to have six histidine residues to
the amino terminus (His-CDPK-A and His-CDPK-B) were
expressed in E. coli and purified by Ni+-affinity column
chromatography (Fig. 5A). Each partially purified fraction
contained 63-kDa protein corresponding to the expected
molecular size of 62.6 kDa as a major component, al-
though degraded bands were detected at 40 kDa and 30
kDa in SDS-PAGE (Fig. 5A, lanes 2 and 4). Both two fu-
sion proteins, His-CDPK-A and His-CDPK-B, were used
for an in vitro kinase assay using casein as a substrate
(Fig. 5B). In the presence of Ca2+, phosphorylation of
casein was observed in the presence of either fusion pro-
by guest on October 14, 2011
pcp.oxfordjournals.org
Downloaded from

Page 5

Alternative splicing of CDPK gene in M. polymorpha
209
A
97—a
66—2
42— j§
1
3 0 — -
20 —
(kDa)
2 3 4
a -,.-;
. . • ~" *"^
B
97 —
a 66—
42 —
30 —
20 —
(kDa)
1 2
3 4 5 6
-ft
t
Fig. 5 Purification and protein kinase assay of fusion pro-
teins, His-CDPK-A and His-CDPK-B. (A) Soluble extracts of
E. coli expressing His-CDPK-A (lane 1) and His-CDPK-B (lane
3), and purified proteins (lanes 2 and 4, respectively) were re-
solved by 12% SDS-PAGE and stained with Coomassie brilliant
blue. Positions of two fusion proteins are indicated by an open
triangle. (B) 500 ng of purified proteins, His-CDPK-A (lanes 1-3)
and His-CDPK-B (lanes 4-6) were incubated with 1 mM Ca2+
(lanes 1, 2, 4 and 5) or without Ca2+ (lanes 3 and 6) in the pres-
ence of [y32P]ATP, and subjected to SDS-PAGE and to autora-
diography. Alpha-casein (0.5 mg ml"1) was added to the reaction
mixture (lanes 2, 3, 5 and 6). A closed triangle indicates the
position of a-casein.
tein (indicated by an closed triangle in Fig. 5B, lanes 2 and
5). In addition, both types of fusion proteins, His-
CDPK-A and His-CDPK-B, were autophosphorylated in
the presence of Ca2+ (indicated by the open triangle in
Fig. 5B, lanes 1, 2, 4 and 5), although a higher level of
autophosphorylation was observed in the His-CDPK-B
(lanes 4 and 5). However, by chelating Ca2+ in the pres-
ence of 5 mM EGTA, neither of the two fusion proteins
nor casein was phosphorylated (Fig. 5B, lanes 3 and 6).
These results indicate that both isozymes, CDPK-A and
CDPK-B, have Ca2+-dependent protein kinase activities in
vitro.
Discussion
Amino acid sequence conservation between liverwort
and higher plant CDPKs—The alternatively spliced liver-
wort CDPKs (CDPK-A from Exon 6A and CDPK-B from
Exon 6B) each have 548 amino acid residues with an esti-
mated molecular weights of 60418 and 60477, respective-
ly. CDPK-A and CDPK-B shared significant amino acid
sequence similarities (approximately 60% identity through-
out their length) with previously characterized calcium-
dependent and calmodulin-independent serine/threonine
protein kinases of higher plants, Arabidopsis (64%)
(Hong et al. 1996), rice (62%) (Breviario et al. 1995), ma-
ize (61%) (Estruch et al. 1994), 44% identity with Chla-
mydomonas CDPK (Siderius et al. 1997), and 33% identity
with Drosophila CaMK (Ohsako et al. 1993) (Fig. 6). Al-
though the N- and C-terminal regions of the liverwort
CDPKs did not show significant similarity to other plant
CDPKs, liverwort CDPKs contain the following three
conserved domains characteristic to CDPKs; First, the ki-
nase domain is subdivided into twelve kinase subdomains
(I-VIA, VIB-XI) as reported previously (Hanks and Hun-
ter 1995) and shows a high level of local sequence similar-
ity (193 amino acids identical in 259 amino acids) with the
kinase domain of the maize CDPK which expresses spe-
cifically in pollen (Estruch et al. 1994). Twenty-three con-
sensus amino acids predicted in the kinase domain (Stone
and Walker 1995) were all conserved in the two liverwort
CDPK isoforms (depicted as asterisks in Fig. 6). Second,
the autoinhibitory domain, which functions as a pseudo-
substrate for the kinase and inhibits kinase activity in the
absence of Ca2+ (Harmon et al. 1994), is highly con-
served in the liverwort CDPK isoforms (30 amino acids
identical in 34 amino acids with Arabidopsis CDPK6).
Third, the calcium-binding domains forming a three-di-
mensional structure that consists of four helix-loop-helix
domains (so called EF-hand motifs) are conserved in the
liverwort CDPKs. Five amino acid residues, DxDx(D/
S/N)xxx(D/T/N)xxE (x, unconserved amino acid resi-
dues), in the loops of liverwort CDPK isoforms corre-
spond to those which bind to Ca2+ in other calcium-bind-
ing proteins as depicted previously (da Silva and Reinach
1991, Marsden et al. 1990). In addition, 6th Gly residue
and 8th He residue in the EF-hands are conserved in the
liverwort CDPK isoforms as shown previously (da Silva
and Reinach 1991, Marsden et al. 1990). It has been
reported that a single amino acid substitution in the EF-
hands causes a dramatic reduction of calcium-binding
ability (Drake et al. 1996, Steinmetz et al. 1998), or drastic
decrease in enzyme activity (Drayer et al. 1995). In the two
liverwort CDPKs, there are three substitutions in the loops
and one in a helix between the loops (shown in Fig. IB and
Fig. 6). In particular, the second calcium-binding loop
contains an amino acid substitution (Asp in CDPK-A, and
Asn in CDPK-B) involved in calcium-binding. A substitu-
tion between Asp and Glu, which is located in the helix
(Fig. IB) could reflect the calcium-binding affinity as
reported that Glu residue in the helices modulate the cal-
cium-binding affinity through the stability of three dimen-
sional structure of the enzyme (Fujimori et al. 1990). These
substitutions might reflect a different calcium-binding
affinity or phosphorylation activity between the two liver-
wort isoforms, although both isozymes showed calcium-
dependent protein kinase activities in vitro (Fig.5B). Al-
though higher levels of phosphorylation of casein and
autophosphorylation were observed in one of the iso-
zyme CDPK-B (Fig. 5B), it can not be concluded that
CDPK-B has a higher specific activity than CDPK-A be-
cause of enzyme degradation or impurities in the two
different preparations. Thus, further quantitative analy-
by guest on October 14, 2011
pcp.oxfordjournals.org
Downloaded from