(Received for publication, October 30, 1996, and in revised form, December 12, 1996)
From the Department of Molecular Pharmacology, Atran Laboratories, Albert Einstein College of Medicine, Bronx, New York 10461
The molecular and cellular basis for concerted
Ca2+/lipid signaling in Caenorhabditis elegans
was investigated. A unique gene (pkc-2) and cognate
cDNAs that encode six Ca2+/diacylglycerol-stimulated
PKC2 isoenzymes were characterized. PKC2 polypeptides (680-717 amino
acid residues) share identical catalytic, Ca2+-binding,
diacylglycerol-activation and pseudosubstrate domains. However,
sequences of the N- and C-terminal regions of the kinases diverge. PKC2
diversity is partly due to differential activation of transcription by
distinct promoters. Each promoter precedes an adjacent exon that
encodes 5-untranslated RNA, an initiator AUG codon and a unique open
reading frame. PKC2 mRNAs also incorporate one of two 3
-terminal
exons via alternative splicing. Cells that are capable of receiving and
propagating signals carried by Ca2+/diacylglycerol were
identified by assessing activities of pkc-2 gene promoters
in transgenic C. elegans and visualizing the distribution of PKC2 polypeptides via immunofluorescence. Highly-selective expression of certain PKC2 isoforms was observed in distinct subsets of
neurons, intestinal and muscle cells. A low level of PKC2 isoforms is
observed in embryos. When L1 larvae hatch and interact with the
external environment PKC2 content increases 10-fold. Although 77- and
78-kDa PKC2 isoforms are evident throughout post-embryonic development,
an 81-kDa isoform appears to be adapted for function in L1 and L2
larvae.
Numerous hormones, growth factors, and neurotransmitters elicit
receptor-mediated activation of phospholipases (C,
C
, D, and A2) and the concomitant production
of phospholipid-derived second messenger molecules (1-3). Signals
carried by diacylglycerol and other phospholipid metabolites are
received, amplified, and distributed to multiple intracellular
compartments by various protein kinase C
(PKC)1 isoforms (1-8). Activated PKCs
catalyze the phosphorylation of substrate/effector proteins at specific
Ser and/or Thr residues, thereby modulating their functions.
PKC-mediated protein phosphorylation is involved in the regulation of
many important processes, including secretion, ion channel activity,
mitogenesis, gene transcription, and cell differentiation (1-5).
Functions controlled by lipid second messengers vary with cell type
(1-5). This reflects differences in expression of hormone/growth factor receptors, phospholipases, and PKC substrates. Distinctive properties of PKC isoenzymes also contribute to the diversity of
responses elicited by phospholipid metabolites. In mammals nine genes
encode 10 PKC isoforms (1-7). These kinases differ in their substrate
specificities, susceptibilities to activation by calcium and lipid
second messengers, intracellular destinations after activation, and
ability to undergo down-regulation. Reconstitution of PKC-mediated
signaling in permeabilized cells and manipulation of PKC signaling
pathways via transfection and microinjection indicate that individual
PKCs control discrete physiological functions. Examples include the
regulation of secretion, phospholipase C activity, and
mitogenesis by PKCs
,
, and
, respectively (9-11). An
important caveat is that these studies were performed with immortalized
or tumor-derived cultured cells. The relevance of the observations to
physiological roles for individual PKCs in specific cells in intact
organisms remains to be established.
The spectrum of cellular responses to lipid second messengers also reflects differential expression of genes encoding PKC isoforms. Qualitative and quantitative differences in PKC isoform content are evident in many mammalian tissues (12-15). Moreover, types and levels of PKC isoenzymes expressed in a given cell can be altered by differentiation, hormones, and phorbol esters (1-5, 16-20). In model cell systems, an increase in PKC isoform content has been correlated with elevations in the rate of transcription of the cognate gene and the level of mRNA encoding the isoform (17, 18). Although direct transcriptional activation probably plays a central role in generating PKC diversity (20-22), underlying control mechanisms have not been elucidated. Little is known about cis-regulatory elements and trans-acting proteins that govern activation or inhibition of PKC gene transcription.
The non-parasitic nematode Caenorhabditis elegans provides a powerful system for investigations on functions and regulated expression of PKC isoforms. Adult C. elegans are composed of 959 somatic cells, which are organized into tissues that constitute digestive, reproductive, muscular, hypodermal, and nervous systems (23-25). The cellular and developmental biology of C. elegans have been characterized in exceptional detail and the lineage of each cell in the animal has been determined (23-25). Numerous aspects of C. elegans development and homeostasis are controlled by signal transduction systems that are analogous to, or identical with, those operative in mammals (26, 27). Methods for production of mutant, transgenic and "knock-out" strains of C. elegans enable the analysis of gene promoter activities and gene functions in individual cells in situ (28-32). Confocal immunofluorescence microscopy (33) permits detection of specific proteins in individual cells of intact C. elegans at all developmental stages.
We demonstrated the utility of this system by characterizing the
C. elegans PKC1 gene (34). This gene encodes a novel
calcium-independent, diacylglycerol-activated PKC (nPKC) that is
expressed exclusively in ~75 sensory neurons and related
interneurons. Another C. elegans nPKC (the product of the
tpa-1 gene) has been characterized by Miwa and colleagues
(35, 36). It is thought that nPKCs mediate sustained, long-term,
lipid-controlled signaling. In contrast, rapid (but transient)
responses to many hormones/growth factors often require integration of
signals carried by both lipids and calcium ions (1-5). Only classical
PKCs (cPKCs, which correspond to the ,
I,
II, and
isoforms
in mammals) contain distinct binding sites for diacylglycerol and
calcium. Concerted actions of the two activators promote translocation
of cytoplasmic cPKCs to membranes and cytoskeleton and generate maximal
levels of kinase activity (1-8). Previous studies documented the
occurrence of a cPKC-like enzyme in C. elegans (37),
suggesting that concerted lipid/calcium signaling is operative in
C. elegans. However, C. elegans cPKCs have not
been characterized at the molecular level. Thus, to initiate
investigations on functions and regulated expression of C. elegans cPKCs it is essential to clone and characterize relevant
cDNAs and genes; discover mechanisms that govern the generation of
cPKC diversity; identify cells in which the cPKC promoter(s) is/are
active and cPKC polypeptides accumulate (to provide clues regarding
isoform function in vivo); and characterize alterations in
cPKC expression and intracellular localization during development.
The Bristol N2 strain of C. elegans was grown at 20 °C as described previously (38). To synchronize C. elegans for developmental studies nematodes were hatched in the absence of nutrients and then transferred to plates containing Escherichia coli as a food source. Under these conditions, the worms develop synchronously into reproductive adults (39). L1 larvae were harvested 6 h after feeding, L2 larvae at 20 h, L3 larvae at 29 h, L4 larvae at 40 h, young adult worms at 53 h, and egg-laying adult nematodes at 75 h. A purified population of embryos was obtained by alkaline hypochlorite treatment of gravid C. elegans, as described by Sulston and Hodgkin (40).
Isolation of cDNAs Encoding PKC2 IsoformsA cDNA
that encodes a segment (residues 287-665, Fig. 1, A and
B) of a novel C. elegans protein kinase C (named
PKC2) was isolated from a complementary DNA library in bacteriophage
gt10, as described previously (34). A fragment (394 bp) was excised from the 5
-end of the cDNA (by digestion with EcoRI and
NcoI) and used as a template to generate a random-primed,
32P-labeled probe. This probe was used to screen a C. elegans cDNA library in bacteriophage
ZAP II (Clontech) as
indicated in previous papers (34, 38). Seven positive recombinant phage
clones were plaque purified and the cDNAs (0.4-2.4 kbp) were
subcloned in the plasmids pGEM7Z (Promega) and pBluescript (Stratagene)
and sequenced.
Computer Analysis
Analyses of sequence data, sequence comparisons, and data base searches were performed using PCGENE-IntelliGenetics software (IntelliGenetics, Mountainview, CA) and the BLAST and FASTA programs (41, 42) provided by the NCBI server and the National Library of Medicine/National Institutes of Health.
Southern Gel AnalysisFragments of C. elegans genomic DNA were generated by digestion with restriction endonucleases, fractionated in a 0.6% agarose gel, and transferred to a Nytran membrane as described previously (38). The Southern blot was probed with the 32P-labeled PKC2 cDNA (2 × 106 cpm/ml) described below. Conditions for hybridization, as well as high and low stringency washing of the membrane, are given in Hu and Rubin (38).
Preparation of RNA and Northern Gel AnalysisTotal C. elegans RNA was prepared as indicated in a previous paper (38). Poly(A+) RNA was purified according to Sambrook et al. (43). Northern blot analysis was performed as described previously (34). A 32P-labeled, EcoRI fragment of PKC2 cDNA (nucleotides 885-2019, Fig. 1, A and B), which encodes the catalytic domain of the kinase, was used as a probe.
DNA Sequence AnalysisPKC2 cDNAs and genomic DNA fragments containing the pkc-2 gene were subcloned into the plasmid pGEM7Z. DNA inserts were sequenced by the dideoxynucleotide chain termination procedure of Sanger et al. (44) using T7, SP6, and custom oligonucleotide primers as described previously (38).
Characterization of the Extreme 5 Complementary DNAs corresponding to the 5-terminal
regions of PKC2 mRNAs were synthesized, amplified, cloned, and
sequenced as detailed in Land et al. (34, 45). Three rounds
of amplification, via the polymerase chain reaction, were used to
obtain PKC2 cDNAs. The 5
primers were described previously (34).
The initial 3
primer (5
-GCTCTAGATTATGCTGCTGGCGAGGATC-3
) contains the
inverse complement of nucleotides 321-340 in the cDNA encoding
PKC2 isoforms (Fig. 1A); the second 3
primer
(5
-GCTCTAGAGGTATCTACTCCTTTATCTGC-3
) contains the inverse
complement of nucleotides 297-317 in PKC2 cDNA; the final 3
primer (5
-GCTCTAGATGCGATTTGATTTCGTGAAT-3
) contains the inverse
complement of nucleotides 123-142 in PKC2 cDNA. The first 8 nucleotides of each 3
primer correspond to two irrelevant nucleotides
and an XbaI recognition sequence (TCTAGA). Before the final
round of cDNA amplification, primers and template were incubated at
100 °C for 2 min and then annealed for 15 s at
70 °C in 50 mM Tris-HCl, pH 8.3, containing 70 mM KCl.
Amplified cDNAs were cloned into the XbaI site of
plasmid pGEM7Z and sequenced.
Complementary DNAs encoded by
three alternative 5-terminal exons in the C. elegans
pkc-2 gene were synthesized and cloned in pGEM7Z as
described above (see Fig. 2 for sequences and nomenclature). Recombinant plasmids were linearized by digestion with PvuII
and 32P-labeled antisense RNAs were synthesized by
bacteriophage T7 RNA polymerase as described previously (45). Antisense
RNAs for exons 1A, 1B, and 1C contain unique sequences of 122, 70, and
94 nucleotides, respectively, that are complementary to corresponding 5
termini in subsets of PKC2 mRNAs. RNase protection analysis was
performed with 20 µg of total RNA isolated from C. elegans at seven stages of development as described previously (34, 45).
Expression and Purification of Recombinant PKC2 Fusion Protein
A 729-bp HpaI-NcoI restriction
fragment of PKC2 cDNA was subcloned into the pRSET-A expression
plasmid (Invitrogen). This places cDNA encoding the calcium-binding
region and part of the catalytic domain of PKC2 isoforms (residues
176-417, Fig. 1A) downstream from the T7 RNA polymerase
promoter and 42 codons that direct the synthesis of an N-terminal
fusion peptide. The peptide contains a stretch of six consecutive His
residues, which form a nickel binding domain. E. coli BL21
(DE3) was transformed with the expression plasmid and induced with 0.4 mM isopropyl-1-thio--D-galactopyranoside for 14 h at 22 °C. The host bacterium contains a chromosomal
copy of the phage T7 RNA polymerase gene under control of the
lac promoter. Bacteria were harvested, disrupted, and
separated into soluble and particulate fractions as described for
previous studies (34). The PKC2 fusion protein was recovered in the
pellet fraction. Recombinant PKC2 fusion protein was dissolved in 20 mM Tris-HCl, pH 8.0, 0.5 M NaCl supplemented
with 6 M urea and purified to near-homogeneity by
nickel-chelate chromatography (in the presence of 6 M urea)
as described previously (46). When urea was eliminated by extensive
dialysis against 50 mM sodium acetate, pH 5.0, the fusion
protein remained soluble. Approximately 3 mg of highly-purified PKC2
fusion protein was obtained from a 500-ml culture of E. coli.
Samples of the PKC2 fusion protein were injected into rabbits (0.35-mg initial injection; 0.2 mg for each of three booster injections) at Hazelton Corning Laboratories (Vienna, VA) for the generation of antisera. Serum was collected at 3-week intervals.
Affinity Purification of Anti-PKC2 ImmunoglobulinsPurified
PKC2 fusion protein (0.7 mg) was coupled to 1 ml of Affi-Gel 10 resin
(Bio-Rad), in 2 ml of 0.1 M sodium acetate, pH 5.0, at
4 °C for 4 h. Antiserum (2 ml) was adjusted to a final concentration of 20 mM MES buffer, pH 6.0, and mixed with
the affinity resin for 2 h at room temperature. Next, the resin
was packed into a column and washed with 20 mM MES, pH 6.0, containing 0.5 M NaCl until the flow-through reached an
absorbance of zero at 280 nm. Bound IgGs were eluted with 3 ml of 0.5%
acetic acid containing 0.15 M NaCl. Fractions (0.5 ml) were
collected into tubes containing sufficient 1 M
Na2HPO4 to neutralize the acid and adjust the
pH to 7.5. The IgG concentration was estimated from the absorbance at
280 nm. Fractions containing IgGs were pooled and supplemented with 5 mg of albumin/mg of IgG. Subsequently, affinity purified IgGs were
dialyzed against 10 mM sodium phosphate buffer, pH 7.4, containing 0.15 M NaCl and 50% (v/v) glycerol and stored
at 20 °C.
Samples of proteins were denatured in gel loading buffer and subjected to electrophoresis in a 9% polyacrylamide gel containing 0.1% SDS as described previously (47). Phosphorylase (Mr = 97,000), transferrin (77,000), albumin (67,000), ovalbumin (43,000), and carbonic anhydrase (29,000) were used as standards for the estimation of Mr values. Cytosolic and particulate fractions of C. elegans and Sf9 cell homogenates were prepared as described previously (34). Western blots of C. elegans proteins and polypeptides from Sf9 cells were blocked, incubated with antiserum (1:2000), and washed as described previously (48). PKC2 isoforms were visualized by an indirect chemiluminescence procedure as previously reported (48).
Expression of PKC2 in Insect CellsComplementary DNAs containing the complete coding sequences for PKC2A and PKC2B (see Fig. 1) were excised from recombinant pBluescript plasmids by digestion with BamHI and SpeI. These inserts were subcloned into the baculovirus transfer vector pVL1392, which was cleaved with BamHI and XbaI. Recombinant baculoviruses were produced and used to infect Sf9 cells as indicated in previous papers (34, 45). Aliquots of infected cells were harvested every 24 h over a period of 5 days. Cytosolic and particulate fractions of infected Sf9 cells were prepared as described previously (48). PKC activity was determined in the presence and absence of calcium, using the synthetic peptide RFARKGSLRQKNV as a substrate (34).
Preparation of Transgenic C. elegansThe cosmid E01H11,
which contains the gene encoding PKC2 isoforms and 5-flanking DNA (see
"Results"), was obtained from Dr. Alan Coulson, Medical Research
Council Laboratory of Molecular Biology, Cambridge, United Kingdom.
Fragments of genomic DNA that flank the 5
-ends of exons 1A, 1B, and 1C
were excised from cosmid E01H11 by digestion with restriction enzymes
and identified by hybridization (on Southern blots) with
32P-labeled cDNA probes corresponding to the unique
5
-terminal cDNAs described above and in Fig. 2. Lines of
transgenic nematodes were created in order to investigate the
cell-specific in vivo promoter activity of the flanking DNA
segments. The basic strategy involves the insertion of 5
-flanking DNA
into the multiple cloning site of a C. elegans expression
vector (plasmid pPD16.51) devised by Fire et al. (29).
Inserted promoter sequences will drive the expression of a
-galactosidase reporter gene (lacZ) that is immediately
preceded by 27 nucleotides encoding an initiator ATG and the nuclear
targeting sequence of SV40 large T antigen (29). The lacZ
coding region is followed by translation termination and poly(A)
addition signals.
A 1.5-kbp segment of cosmid DNA that terminates 13 bp upstream from the
initiator ATG in exon 1A was obtained by digestion with
BstXI, creation of blunt ends with T4 DNA polymerase, and subsequent cleavage with SphI. The DNA fragment was ligated
into plasmid pPD16.51 that was cut with SmaI and
SphI. A 3-kbp DNA fragment that terminates 9 bp upstream
from the initiator ATG in exon 1B was produced by digestion with
BamHI and XhoI. This DNA insert was ligated into
plasmid pPD16.51 that was cleaved with BamHI and
SalI. Finally, a 1.4-kbp fragment that terminates at codon
25 (in exon 2) was obtained by digesting cosmid DNA with ApaLI, filling in with Klenow DNA polymerase, and cleaving
with SphI. The fragment was cloned into the vector pPD16.51,
which was cleaved with SmaI and SphI. In the
resulting construct, the DNA segment that encodes -galactosidase
preceded by the nuclear targeting sequence is positioned, in-frame,
downstream from the exon 1C initiator ATG. In recombinant pPD16.51
plasmids containing DNAs that flank exons 1A and 1B, translation is
initiated at the ATG adjacent to the SV40 nuclear localization
signal.
C. elegans were transformed by microinjecting both
recombinant reporter plasmid DNA containing the putative 1A, 1B, or 1C promoter and a plasmid containing the dominant selectable marker gene
rol-6, as described previously (30, 49). Transgenic C. elegans were selected and maintained as indicated in a previous paper (49). Transgenic C. elegans were fixed and stained for -galactosidase activity as reported in Freedman et al.
(49).
C. elegans were fixed, washed, and incubated sequentially with affinity purified anti-PKC2 IgGs and fluorescein isothiocyanate-tagged goat IgGs directed against rabbit immunoglobulins, as described by Land et al. (34). Fluorescence signals corresponding to PKC2-IgG complexes were obtained with a Bio-Rad MRC 600 laser scanning confocal microscope (Image Analysis Facility, Albert Einstein College of Medicine).
Complementary DNAs that encode PKC2 were retrieved from
bacteriophage libraries. Two near-full length (~2.4 kbp) and six
partial (0.4 to 2.0 kbp) cDNAs were sequenced from both DNA
strands. All partial sequences were identical with segments of the
larger cDNAs. The 2.4-kbp cDNAs encoded two related, but
distinct PKC2 isoforms (designated PKC2A and PKC2B) (Fig.
1). PKC2A and 2B cDNAs are identical between
nucleotides 1 and 1918 (Fig. 1A). The shared cDNA
segment contains 5-untranslated nucleotides, a Met codon (nucleotides
27-29) in a C. elegans consensus context for translation initiation ((A/G)NNATGT) and a contiguous 3
-open reading frame that encodes 630 amino acids (Fig. 1A). The sequences
of PKC2A and 2B cDNAs diverge after codon 631. The novel 3
portion of PKC2A cDNA contains 150 bp of coding sequence followed by a translation termination codon (nucleotides 2067-2069) and 309 untranslated nucleotides (Fig. 1B). A consensus poly(A)
addition signal (AATAAA, nucleotides 2363-2368) precedes the poly(A)
tail by 11 nucleotides. The unique 3
region in PKC2B cDNA includes 52 codons, a translation termination signal (nucleotide 2073-2075) and
287 untranslated nucleotides (Fig. 1C). An atypical poly(A) addition signal (ACTAAA, nucleotides 2343-2348) is evident 14 nucleotides upstream from the poly(A) tail. The data suggest that the
divergent sequences in PKC2A and 2B mRNAs correspond to
alternatively-spliced, 3
-terminal exons comprising 462 and 446 nucleotides, respectively. Subsequent characterization of genomic DNA
demonstrated directly that sequences presented in Fig. 1, B
and C, constitute the final and penultimate exons,
respectively, of the kin-112
(pkc-2) gene (see below).
The predicted amino acid sequences of the C-terminal regions of PKC2A and PKC2B are divergent (only 48% identical). Dissimilarities in C-terminal domain sequences and higher order structure may differentially affect kinetic properties, stability, and/or intracellular targeting of PKC2 isoforms (e.g. see Ref. 50). Overall, the PKC2A isoform is composed of 680 amino acids and has a Mr of 77,800; PKC2B contains 682 residues and has a calculated Mr of 78,100.
PKC2 Isoform Diversity Is Amplified by the Utilization of Three Types of 5An anchored polymerase chain reaction
procedure, known as RACE (for rapid amplification of cDNA ends),
was used to establish the exact length and complete sequence for the
5-ends of PKC2 mRNAs. Three distinct 5
-terminal cDNA
sequences were cloned and were assigned the names 1A, 1B, and 1C (Fig.
2). 1A-1C were incorporated at the 5
-ends of PKC2
cDNAs in a mutually exclusive manner. Each of the novel 5
cDNA
sequences contains untranslated nucleotides upstream from a potential
initiator Met codon and a contiguous coding region, that in turn, fuses
in-frame with the second nucleotide in codon 14 (Fig. 1A).
The 1B sequence was previously identified by sequencing the PKC2A and
2B cDNAs, as shown in Fig. 1. 1C comprises 86 nucleotides and
encodes an N-terminal extension of 15 residues. 1AA
, which is derived
from two novel exons (see below), encompasses 159 nucleotides and
contains an open reading frame for an N-terminal segment of 50 amino
acid residues. Polypeptide sequences encoded in the three 5
-terminal
cDNAs are unrelated to each other and to protein sequences in the
standard data bases. The results suggest that the C. elegans
pkc-2 gene can direct the synthesis of six calcium/diacylglycerol-dependent PKC isoenzymes. Diversity
is generated by differential incorporation of alternative 5
- and 3
-exons into PKC2 mRNAs that contain an invariant core of 1848 nucleotides (codons 15-630, Fig. 1A).
A Northern blot that contains
size-fractionated C. elegans poly(A)+ RNA was
probed with 32P-labeled PKC2 cDNA. A hybridization
signal was obtained for mRNA that is composed of ~2500
nucleotides (Fig. 3A). Detection of 2.5-kilobase mRNA is consistent with sizes of cDNA sequences
that encode various PKC2 isoforms. Although six distinct mRNAs can be derived from the pkc-2 gene, their predicted sizes
cluster in a narrow range: 2360-2480 nucleotides plus poly(A) tails.
Given the limited resolving power of a 0.8% agarose gel (Fig.
3A), it is probable that the apparent 2.5-kilobase
transcript corresponds to a mixture of several or all six PKC2
mRNAs.
C. elegans genomic DNA was cleaved with restriction enzymes and analyzed by Southern gel analysis (Fig. 3B). The hybridization pattern observed when the blot was probed with 32P-labeled PKC2 cDNA indicated that multiple PKC2 isoenzymes are encoded by a single copy gene.
The chromosomal location of the pkc-2 gene was elucidated by hybridizing a panel of yeast artificial chromosomes, which contain >90% of the C. elegans genome in overlapping segments (51), with a radiolabeled PKC2 cDNA probe. Four overlapping genomic DNA fragments hybridized. Comparison of the positions of the genomic DNAs on the C. elegans physical map of chromosomes (Harvard Medical School electronic map) places pkc-2 on the left arm of the X chromosome in the vicinity of genes named sup-28 and unc-115.
Structure/Function Relationships in C. elegans PKC2 IsoformsFunctional roles for amino acids that are conserved in
S/T protein kinases have been established by a combination of
biochemical analysis, comparisons of sequences of hundreds of
phosphotransferases and determination of three-dimensional structures
for several S/T protein kinases (52-55). Application of this knowledge
to the derived amino acid sequences in Fig. 1 enables tentative
identification of functional domains in C. elegans PKC2
isoforms. The catalytic domain of all PKC2 isoenzymes consists of
residues 347-620 (Fig. 1A). A
GXGXXGX16K motif (residues
355-377) probably contributes hydrogen bonds and charged side chains
that anchor the and
phosphates of the substrate ATP.
Lys377 is essential for expression of catalytic activity.
Asp490, which appears in the conserved DFG sequence
(residues 490-492), is also directly involved in binding Mg-ATP.
Glu517, which is part of a conserved APE triad (residues
515-517), as well as Asp529 and Arg589 have
been implicated in stabilization of the catalytic core region (53, 54).
The RDLKLDN segment (residues 471-477) corresponds to the
"signature" sequence for a S/T protein kinase (52) and is highly
homologous with a critical portion of the catalytic loop in protein
kinase A (53, 54).
C. elegans PKC2 isoforms contain two copies of a Cys-rich, zinc binding motif (residues 52-88 and 117-153). These domains mediate the binding of phosphatidylserine, diacylglycerol, and phorbol esters in mammalian cPKCs and nPKCs (1-7). The binding of zinc (2 atoms per Cys-rich repeat) is required for proper higher-order folding of these regulatory domains (56). The N-terminal Cys-rich region is preceded by a pseudosubstrate sequence (residues 22-35, Fig. 1A). Ala27 and flanking basic residues generate a PKC2 substrate site that lacks Ser and Thr (57). When intracellular levels of diacylglycerol and free Ca2+ are low (unstimulated cells) the pseudosubstrate site occupies the catalytic cleft and inhibits PKC activity.
Residues 184-259 (Fig. 1A) are homologous (Table I) with the calcium-binding region (C2 domain) in mammalian cPKCs (58). Six Asp residues (Asp189, Asp195, Asp206, Asp249, Asp251, and Asp257) that ligate Ca2+ ions and hydrophobic amino acids (Trp248, Trp250, and Phe258) that orient PKC interactions with membranes are conserved in C. elegans PKC2. Binding of diacylglycerol and calcium to cPKCs results in expulsion of the pseudosubstrate domain from the catalytic site and expression of phosphotransferase activity (57).
|
Activation and intracellular translocation of cPKCs are governed (in part) by the sequential phosphorylation of three residues in the C-terminal portion of the enzymes (7, 59). By analogy with mammalian PKCs, these residues are identified as Thr510, Thr651, and Ser670 in nematode PKC2A (Fig. 1, A and B). Thr510 is transphosphorylated by an unidentified protein kinase (59); subsequent incorporation of phosphate at the latter two residues is due to autophosphorylation.
The utilization of alternative 3-exons to encode the C termini of PKC2
isoforms (Fig. 1) documents the conservation (from nematode to man) of
a splicing mechanism that generates cPKC diversity (20, 60). Mammalian
PKC
I and
II isoforms have divergent C termini encoded by
alternative exons. In nematodes and mammals the two C-terminal
sequences include 50 or 52 amino acids, but share only ~48% sequence
identity. Human PKC
II binds F-actin via a unique site located near
the C terminus of the kinase; PKC
I does not interact with F-actin
(50). The sequence that sequesters actin in man (50) is not conserved
in C. elegans PKC2 isoforms. However, a substantial
proportion of C. elegans PKC2 is associated with the
particulate fraction of homogenates (see below).
Mammalian cPKCs bind 35-kDa proteins known as RACKs (receptors for
activated protein kinase Cs) (64). RACKS facilitate translocation and
anchoring of activated cPKCs to membranes. Mammalian PKC isoforms
contain an autoregulatory region (Ser-Val-Glu-Ile-Trp-Asp) in the C2
(calcium-binding) domain that occludes the RACK-binding domain in the
absence of lipid second messengers, thereby blocking PKC
activation/translocation in unstimulated cells (65). The pseudo-RACK,
autoregulatory region is conserved
(Ser244-Ile-Glu-Val-Trp-Asp249) in C. elegans PKC2 isoforms (Fig. 1A), suggesting that
RACK-mediated PKC routing operates in C. elegans.
C. elegans PKC2 isoforms exhibit a high degree of overall
homology (67% identity) with both the and
isoforms of
mammalian cPKCs (Fig. 4). Maximal levels of sequence
identity are evident in the pseudosubstrate, catalytic, and Cys-rich
regulatory regions (Table I). However, the differential utilization of
alternative exons to produce distinct C termini is observed only for
the mammalian PKC
and the C. elegans pkc-2 genes. This
suggests that these two genes are most closely related and derived from
a common ancestor.
Organization of the C. elegans pkc-2 Gene
A 12-kbp fragment
of DNA that contains a portion of the pkc-2 gene was
obtained from a C. elegans genomic DNA library in
bacteriophage EMBL4. Sequence analysis of a 2-kbp segment of the
genomic DNA elucidated sequences for 5 introns and 6 exons that encode
residues 14-312 in the common core of all PKC2 isoforms (Fig.
1A). During the course of our studies the C. elegans genome project (67) deposited the DNA sequence for large
portions of the X chromosome in the GenBank data base. Searches of the
data base with sequences of the 5- and 3
-ends of PKC2 cDNAs
(Figs. 1, B and C, and 2) revealed that cosmid
E01H11 (accession number U29376[GenBank]) contained the entire pkc-2
gene. Comparison of cDNA sequences determined for all PKC2 isoforms
(Figs. 1 and 2) with the cosmid DNA sequence enabled the elucidation of
the intron/exon organization of the pkc-2 structural gene
(Table II). The GENEFINDER program (67) used in the
sequencing project predicted an incorrect amino acid sequence for this
region of the DNA. No portion of the gene or its transcripts was
previously characterized experimentally.
|
The pkc-2 gene contains 17 exons that are dispersed over 25 kbp of DNA (Table II). Alternative 5- and 3
-exons are separated by
large introns that account for 80% of the total length of the gene. In
contrast, invariant exons 2-12 are embedded in a relatively compact
DNA segment (3 kbp). All introns in this region are small, ranging in
size from 46 to 281 bp. The 1A-1C sequences that appear at the 5
termini of discrete PKC2 mRNAs are encoded by 4 exons. Exons 1A and
1A
(Table II) are spliced together to generate a novel 5
-terminal
cDNA sequence (Fig. 2). Alternative exons 1B and 1C encode the 1B
and 1C cDNA sequences, respectively. Alternative exons (13A and
13B, Table II) at the 3
-end of the gene are separated by a 1.6-kbp
intron and contain open reading frames for 50 or 52 residues,
translation termination codons, 3
-untranslated sequences, and poly(A)
addition signals linked in a contiguous fashion. Since only one 3
-exon
and one or two 5
-exon(s) are incorporated in mature mRNAs, PKC2
isoforms are encoded by 13 or 14 exons.
Systematic characterization of PKC2 cDNAs and RACE cDNA
products revealed the sequence GGTTATACCCAGTTAACCAAG at the extreme 5-end of 1B cDNA sequences. This sequence is donated from the 5
-end of a spliced leader RNA (encoded by ~100 tandemly-repeated SL1
RNA genes) in a trans-splicing reaction (68). C. elegans mRNAs undergo trans-splicing only when
transcripts contain an unpaired splice acceptor signal (TTTCAG) in
their 5
-untranslated regions (69). Since the pkc-2 gene
contains upstream splice donor sequences at the 3
boundaries of exons
1A and 1A
, the results indicate that a promoter (designated P1B) is
positioned upstream from exon 1B and downstream from exon 1A
. A
distinct promoter (P1A) must drive transcription of exons 1A and 1A
. A third promoter (P1C) may be associated with exon 1C because the 1AA
or
1B exons cannot be excised from transcripts initiated by promoters P1A
and P1B to yield the 1C 5
cDNA sequence. In contrast, when
transcription begins at promter P1A or P1B, exons 1B and/or 1C readily
become introns.
Sf9 insect cells were infected with
recombinant baculovirus that contained full-length PKC2A cDNA (Fig.
1) downstream from a powerful polyhedron promoter. Proteins in Sf9 cell
cytosol were size-fractionated in a 0.1% SDS-9% polyacrylamide gel,
transferred to a polyvinylidene difluoride membrane, and probed with
affinity-purified IgGs directed against a segment (residues 176-417)
of the PKC2 isoforms that corresponds to the C2 region and part of the
catalytic domain. The antibodies bound a doublet of 77/78-kDa
polypeptides in cytosol from virally-infected cells (Fig.
5, lanes 3 and 4). The sizes of
the immunoreactive proteins are in agreement with the calculated
Mr values for PKC2A and PKC2B. By analogy with mammalian PKC isoforms (59), the doublet may be due to different levels of autophosphorylation at residues Thr651 and
Ser670. The IgGs did not bind with endogenous proteins in
Sf9 cells (Fig. 5, lane 1) and detection of C. elegans PKC2 was inhibited by adding an excess of polypeptide
antigen (Fig. 5, lane 2).
Cytosol from control Sf9 cells has a low level of PKC activity, which is only weakly stimulated by Ca2+ (Table III). In contrast, PKC activity is elevated 6-8-fold by Ca2+ in cytosol derived from cells expressing 77/78-kDa PKC2A. Furthermore, immunoprecipitation with anti-PKC2 IgGs reduced Ca2+-stimulated phosphotransferase activity >90% in cytosol prepared from Sf9 cells that express PKC2A (Table III).
|
Incorporation of exon 1B into PKC2
mRNAs was documented by both cDNA sequencing and RACE analysis
(Figs. 1 and 2), whereas sequences corresponding to exons 1A, 1A, and
1C were discovered in short PCR-derived cDNAs (Fig. 2). To confirm
the utilization of exons 1A, 1A
, and 1C in vivo and monitor
expression of PKC2 mRNAs containing the various 5
termini during
C. elegans development, we performed RNase protection
analysis. 32P-Labeled antisense RNA, which complements a
unique 122-nucleotide mRNA sequence encoded by exons 1A and 1A
,
was used as a probe. A low level of 32P-labeled cRNA
hybridized with mRNA from embryos (Fig. 6A,
lane 1). A 6-fold increase in abundance of transcripts containing
exons 1A and 1A
was observed as eggs hatched and L1 larvae initiated contact with the environment (Fig. 6A, lane 2). The content
of mRNAs that contain 1A and 1A
exons declined by 80% at the L3 stage and remained low thereafter (Fig. 6A, lanes 3-7).
Similar experiments were performed with antisense probes for mRNAs
containing the 1B and 1C exons and data were quantified via
PhosphorImager analysis (Fig. 6B). In contrast to the
developmentally regulated utilization of 1A and 1A
exons, the
frequencies with which 1B and 1C exons are incorporated into PKC2
mRNAs were relatively constant throughout development.
Developmental Regulation of Expression of PKC2 Polypeptides
Expression of PKC2 isoforms was monitored via
Western immunoblot analysis. Although embryos contain PKC2 mRNAs
(Fig. 6), only small amounts of the encoded polypeptides are detected
during early development (Fig. 7, lanes 1 and
2). The concentration of PKC2 isoenzymes increases
~10-fold as L1 larvae hatch and begin to interact with the external
environment (Fig. 7, lanes 3 and 4). Substantial
amounts of the Ca2+-activated protein kinases are also
evident in C. elegans during the subsequent course of larval
(L1 to L4) and adult development (Fig. 7, lanes 5-16). A
high proportion of PKC2 polypeptides is associated with the particulate
fraction in newly-hatched (early L1) C. elegans and L3 and
L4 larvae (Fig. 7, lanes, 3, 4, and 9-12). In
mid-L1 and L2 larvae, as well as young adult nematodes, PKC2 isoforms
are nearly uniformly dispersed between cytosol and organelles and/or
cytoskeleton (Fig. 7, lanes 5-8, 13, and 14). In
contrast, ~70% of PKC2 isoforms is isolated in cytosol from egg-laying adults (Fig. 7, lanes 15 and 16).
The principal cytosolic PKC2 has an apparent Mr
of 78,000 at all developmental stages. However, a closely spaced
77/78-kDa PKC2 doublet usually appears in particulate fractions of
C. elegans extracts (Fig. 7, lanes 5, 7, 9, 11, 13, and 15). Thus, distinct PKC2 isoforms may
differentially associate with membranes and/or cytoskeleton. Both the
77- and 78-kDa kinases are potentially heterogeneous because four PKC2
isoforms have predicted Mr values of
77,000-78,000: PKC21B, PKC2
1C,
PKC2
1B, and PKC2
1C (where
and
correspond to C-terminal regions encoded by alternative 3
-exons; 1B and 1C denote N-terminal domains encoded by the alternate 1B and 1C 5
-exons, respectively (Table II)). In addition, each isoform
may exhibit an altered electrophoretic mobility because of differential
autophosphorylation (59).
Larger PKC2 isoforms (Mr ~80,000-82,000)
accumulate immediately after C. elegans embryogenesis
terminates (Fig. 7, lane 3). An 8-fold increase in an 81-kDa
PKC2 polypeptide in membranes (Fig. 7, lanes 1 and
3) is coordinated with a 6-fold elevation in content of
mRNA that contains exons 1A and 1A (Fig. 6). Expression of 80-82
kDa PKC2 proteins persists in L1 and L2 larvae (Fig. 7, lanes
5-8). The level of 80-82-kDa PKCs declines sharply in L3 animals
and these proteins are not detected in L4 and adult C. elegans (Fig. 7, lanes 9-16). A decrease in mRNA
containing exons 1A and 1A
parallels the loss of the larger PKC2
enzymes (Fig. 6). However, a low level of 81-kDa PKC2 accumulates in
embryos (Fig. 7, lane 1), which contain the same or a lesser
amount of the cognate mRNA than L4 and adult nematodes (Fig. 6).
Incorporation of 50 amino acids encoded by the 1A and 1A
exons (Table
II) at the N terminus of PKC2 generates two 81-kDa isoforms,
PKC2
1A and PKC2
1A. Different stoichiometries
of autophosphorylation could contribute another level of heterogeneity
(59).
Together, RNase protection
analysis, cDNA sequencing, and elucidation of the organization of
the pkc-2 gene suggest that three promoters govern the
synthesis of transcripts encoding PKC2 isoforms. By creating transgenic
C. elegans carrying chimeric reporter genes it is possible
to assess promoter activities in individual cells of intact animals.
Thus, 1.5-, 3.0-, and 1.4-kbp segments of DNA that flank the 5-ends
and extend into 5
-untranslated regions of exons 1A, 1B, and 1C,
respectively, were inserted upstream from a
-galactosidase reporter
gene (lacZ) in a C. elegans expression plasmid.
An octapeptide targeting domain from SV40 large T antigen was
engineered into the N terminus of
-galactosidase to direct accumulation of the reporter enzyme in cell nuclei (29). Promoter activity was revealed by histochemical staining for
-galactosidase; cells with active promoters were identified by microscopy and reference
to a well established anatomical data base. Fusion genes containing
potential promoter regions (P1A, P1B, and P1C) were designated
pkc2P1A:lacZ, pkc2P1B:lacZ,
and pkc2P1C:lacZ, respectively.
Multiple lines of transgenic C. elegans that contain the
various chimeric genes were created and assayed. Typical results are
presented in Fig. 8. The P1C promoter was active in only
~9 cells in L1-L4 larvae (Fig. 8A). This promoter directs
production of PKC2 mRNAs in pharyngeal neurons and certain sensory
neurons in the anal ganglion. Expression of the
pkc2P1A:lacZ fusion gene is observed in ~25
cells (Fig. 8B). More than half of the nuclei that contain
-galactosidase are located in neurons. In the head region, the
neurons are constituents of sensory ganglia (Fig. 8B). Their
cell bodies are positioned immediately anterior and posterior to the
nerve ring and their processes contribute to the nerve ring and ventral
nerve cord. P1A promoter activity is also observed in sensory neurons
in the tail ganglion. The cellular patterns of P1A and P1C promoter
activities do not overlap. The P1B promoter activates reporter gene
transcription in ~35 cells in L1-L4 larvae. The pattern of
-galactosidase accumulation in transgenic C. elegans
containing the pkc2P1B:lacZ chimera (Fig. 8C) is partially congruent with that observed for the
pck2P1A:lacZ construct. Nuclei in body wall
muscle, several sensory neurons posterior to the nerve ring, and in
tail ganglia and a few intestinal cells evidently employ the P1A and
P1B promoters to produce mRNAs encoding multiple PKC2 isoforms
(Fig. 8, B and C). However, the P1B promoter also
stimulates lacZ transcription in neuronal, intestinal, and
muscle nuclei that lack
-galactosidase in C. elegans
carrying pkc2P1A:lacZ and
pkc2P1C:lacZ (Fig. 8C). Intense 1B
promoter activity is evident in four cells that comprise the top of the
intestine, whereas weaker activity is observed in other intestinal
nuclei and in body wall muscle nuclei that lie near the tip of the
head. Promoter P1B also directs
-galactosidase expression in nuclei of somatic cells in distal regions of the symmetrical adult gonad (Fig.
8D).
No positively staining cells were observed when transgenic nematodes
carrying a promoterless reporter gene or lacZ downstream from a metal-inducible promoter (49) were assayed under similar conditions. The specificity of utilization of the P1A-P1C promoters was
documented further by the lack of -galactosidase staining in ~900
somatic nuclei in the transgenic lines of C. elegans.
Accumulation and localization of PKC2 proteins were analyzed
by confocal immunofluorescence microscopy. PKC2 polypeptides accumulate
in the cell bodies of multiple neurons that are components of sensory
ganglia. These neurons are positioned anterior and posterior to the
nerve ring in the head of C. elegans (Figs. 9, A-C). Representative micrographs reveal that
the Ca2+-activated kinases are maximally enriched in
neuronal processes that are incorporated into the nerve ring (the
principal site of integration of neuronal signaling) and cells that
constitute the anterior end of the intestine. In the head region PKC2
isoforms are also detected in cell bodies of pharyngeal neurons and in neuronal processes that contribute to the amphid and labial nerve fibers and the ventral nerve cord (Fig. 9, A and
B). PKC2 expression is also evident in cell bodies of
neurons that are included in the rectal and tail ganglia (Fig.
9C). Finally, a high level of PKC2 is observed in somatic
cells of the distal portions of the gonad, including the spermatheca
(Fig. 9D). Patterns of PKC2 protein expression and P1A-P1C
promoter activities are similar.
A unique C. elegans gene (the pkc-2 gene) encodes a family of calcium-stimulated PKC isoenzymes. PKC2 polypeptides contain 680-717 amino acid residues and have molecular weights of 77,000-82,000. The catalytic domain, pseudosubstrate site, calcium-binding segment, and Cys-rich regulatory regions are identical in each PKC2 isoform. However, the proteins diverge at their N and C termini. Complementary DNAs encoding PKC2 isoforms directed the expression of calcium-activated, lipid-dependent protein kinases in Sf9 cells infected with recombinant baculovirus. Anti-C. elegans PKC2 IgGs bound transgene products on Western blots and precipitated calcium-stimulated phosphotransferase activity, thereby confirming that pkc-2 gene products are members of the cPKC superfamily.
In mammals, three genes encode four cPKCs: the ,
I,
II, and
isoforms (1-7). Although these kinases have similar sizes, sequences, and kinetic properties, they subserve distinct physiological roles in certain cultured cells and the immune system (1-7, 70). They
also differ (somewhat) in intracellular distribution, sensitivity to
lipid activators and inhibitors, and stability. Levels of
,
I,
II, and
isoforms vary markedly and independently with
cell/tissue type, development, and the ambient concentrations of
hormones and growth factors. Thus, distinctive properties of cPKC genes and isoenzymes enable an organism to generate a broad spectrum of
integrated physiological responses to hormones/growth factors that
initiate concerted Ca2+/diacylglycerol-mediated signaling.
In contrast, extensive screening of cDNA libraries, reverse
transcriptase-PCR analysis, and the impressive progress of the C. elegans Genome Project (~50% complete in Sept. 1996) have
revealed only one nematode gene that codes for a
Ca2+-activated PKC, pkc-2. This raises the
question of whether multiple transcriptional and/or
post-transcriptional control mechanisms can diversify the
pkc-2 gene products sufficiently to mediate, target, and
fully integrate a variety of input signals in C. elegans.
Systematic cDNA sequencing and application of an anchored PCR
procedure (RACE) to characterize extreme 5-ends of cDNAs revealed that six distinct mRNAs can be derived from the pkc-2
gene. PKC2 mRNAs incorporate one of two 3
-terminal exons via
alternative splicing. Further diversity in PKC2 mRNAs is
contributed by the complex organization of multiple promoters that
govern pkc-2 gene transcription. The 25-kbp gene has 17 exons and is located on the X chromosome. Exons 2-12 encode shared
domains that are present in all PKC2 isoforms. However, transcription
can be initiated from three distinct promoters. Each promoter precedes
an adjacent exon that encodes 5
-untranslated RNA, an initiator ATG
codon, and a unique open reading frame. Differential promoter
utilization and splicing results in the incorporation of the
promoter-proximal exon as exon 1 in the processed PKC2 mRNA. Exons
adjacent to the alternative, non-utilized promoters are processed as
"introns" and are excluded from the mRNA. Thus, three distinct
N termini (1A-1C) contribute to PKC diversity. Derived amino acid
sequences encoded by alternative 5
-exons are not homologous with each
other or protein sequences in standard data bases. The divergent nature of the N- and C-terminal segments of the isoforms suggests that these
structural cassettes may mediate specialized functions. Possible roles
for the different N and C termini include: the modulation of substrate
specificity; selective intracellular targeting and anchoring of PKC2
isoforms to various organelles and cytoskeleton; modulation of the
susceptibility of PKC2 to regulation by a variety of lipid-derived
activators and inhibitors; and control of the stability of the
kinases.
Differential utilization of pkc-2 promoters and 3-terminal
exons provides mechanisms for generating novel, cell- and
developmental-stage specific patterns of PKC2 isoform accumulation and
intracellular distribution. In situ promoter analysis and
RNase protection studies demonstrated that some neurons and
non-neuronal cells restrict PKC2 diversity by expressing mRNAs that
contain only one type of 5
-terminal exon. This implies that only one
or two PKC2 isoforms is/are sufficient to mediate
Ca2+/diacylglycerol signaling in those cells. Regulation of
5
-exon selection suggests the speculation that properties conferred by distinct 1A-1C termini adapt PKC2 isoforms for distinct functions in
different cells. For example, the 15-residue sequence encoded by exon
1C will appear exclusively at the N terminus of PKC2 isoforms expressed
in nine neurons. Various clusters of neurons, muscle, and intestinal
cells, as well as somatic cells of the gonad, accumulate mixtures of
PKC2 mRNAs that begin with either exon 1A or exon 1B. Thus, some
cells of intact C. elegans will contain four PKC2 isoforms
if alternative 3
splicing is also operative. An increase in PKC2
diversity could alter the magnitude and duration of cellular responses
to external stimuli and facilitate recruitment of an enlarged group of
effector proteins from multiple cell compartments. Phosphorylation of
different types and increased numbers of downstream substrate/effector
molecules by several PKC2 isoforms would enable activation and
integration of multiple responding pathways to a concerted
Ca2+/diacylglycerol signal.
PKC2 isoforms are minimally expressed in C. elegans embryos. A 10-fold increase in total PKC2 content was observed in newly-hatched animals. Substantial levels of 77/78-kDa PKC2 were evident at all stages of post-embryonic development (larval stages L1-L4 and reproductive adults). The abundance of mRNAs encoding these isoforms is nearly invariant throughout development. Thus, accumulation of the 77/78-kDa PKC2 isoforms is negatively regulated at the level of translation during embryogenesis. The concentrations of 80-82-kDa PKC2 isoforms and cognate mRNAs are coordinately increased 8- and 6-fold, respectively, in newly-hatched larvae. These larger isoenzymes persist only in a relatively brief developmental period (~18 h) that terminates with the molt that demarcates the transition from L2 to L3 larvae. Thus, 80-82-kDa PKC2 isoforms apparently subserve physiological/regulatory functions associated with early stages of post-embryonic development. Expression of 80-82-kDa PKC2 isoforms seems to be regulated transcriptionally since 1A promoter activity was observed principally in L1 and L2 animals.
Compartmentalizaton of PKC2 varied markedly during development. For
example, approximately 80% of PKC2 was in the particulate fraction of
homogenates of L1 larvae, whereas 70% of PKC2 in adult nematodes
partitioned with cytosol. This could be due to developmental regulation
of alternative 3-terminal exon splicing and/or promoter selection. One
or more of the unique amino acid sequences encoded by the variable 5
-
and 3
-exons may include compartment-specific targeting/anchoring
domains. The observation that the 77-kDa isoform is restricted to the
particulate fraction of C. elegans homogenates is consistent
with this idea. In addition, immunofluorescence microscopy revealed
differential intracellular targeting of PKC2 isoforms in neurons. The
kinases are moderately abundant in cell bodies, highly enriched in
processes that comprise the nerve ring (the key locus of interneuronal
communication and integration of signaling), and excluded from
nuclei.
The low level of PKC2 expression in embryos indicates that this family of protein kinases does not play an essential role in early development. The abrupt increase in expression of PKC2 polypeptides in newly-hatched animals and the persistence of PKC2 and PKC2 gene promoter activity in sensory neurons throughout post-embryonic development suggests that, in part, these kinases mediate the reception and integration of environmental signals and the animal's responses to such signals. The occurrence of a high level of PKC2 in the somatic tissue of the hermaphrodite gonad raises the possibility that this family of lipid activated kinases may play a prominent role in supporting the development of oocytes and spermatocytes. Finally, the shift from a predominantly particulate localization in early larvae to a cytoplasmic distribution in adult C. elegans indicates that PKC2 isoenzymes may perform different functions in different cell compartments at various stages of development.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) U82935[GenBank] and U82936[GenBank].
We thank Ann Marie Alba for expert secretarial services.