![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Published online before print
September 20, 2001, 10.1101/gr.192001
Vol. 11, Issue 10, 1625-1631, October 2001
LETTER
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
RESULTS
DISCUSSION
METHODS
REFERENCES
|
|---|
Vertebrate genomes contain multiple copies of related genes that
arose through gene duplication. In the past it has been proposed that
these duplicated genes were retained because of acquisition of novel
beneficial functions. A more recent model, the
duplication-degeneration-complementation hypothesis (DDC), posits that
the functions of a single gene may become separately allocated among
the duplicated genes, rendering both duplicates essential. Thus far,
empirical evidence for this model has been limited to the
engrailed and sox family of developmental regulators,
and it has been unclear whether it may also apply to ubiquitously
expressed genes with essential functions for cell survival. Here we
describe the cloning of three zebrafish 
subunits of the
Na(+),K(+)-ATPase and a comprehensive evolutionary analysis of this
gene family. The predicted amino acid sequences are extremely well
conserved among vertebrates. The evolutionary relationships and the map
positions of these genes and of other -like sequences indicate that
both tandem and ploidy duplications contributed to the expansion of
this gene family in the teleost lineage. The duplications are
accompanied by acquisition of clear functional specialization,
consistent with the DDC model of genome evolution.
[The sequence data described in this paper have been submitted to the GenBank data library under accession nos. AY028628, AY028629, and AY028630]
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
RESULTS
DISCUSSION
METHODS
REFERENCES
|
|---|
Vertebrate genomes have been shaped by several
episodes of large-scale gene duplications. During a very short time in
the evolution of early vertebrates, for example, many gene families expanded from one to several paralogs (Sidow 1992
; Suga et al. 1999).
Such large-scale duplications are the consequence both of individual
gene duplications and ploidy duplications, the latter known to have
occurred recently in Xenopus and salmonid fishes. There is
strong evidence for an older ploidy duplication during the evolution of
bony fishes (Postlethwait et al. 1998). For example, seven Hox
clusters are found in zebrafish resulting from a duplication of the
four Hox clusters found in most other higher vertebrates (Amores et al., 1998; Prince et al. 1998). As would be predicted from a
ploidy duplication, genes linked to the seven Hox clusters are
also duplicated with syntenic conservation to the corresponding cluster
in mammals (Amores et al. 1998; Woods et al. 2000). Other syntenic
regions between human and zebrafish chromosomes lend further support to
this model (Postlethwait et al. 2000).
In classical models of genome evolution, the fate of a duplicated gene
is dependent upon the nature of the mutations it accumulates. If a new
beneficial function is acquired, the duplicate will be retained; if no
new function is acquired, it will be lost (Ohno 1970; Sidow 1996).
Recently, Force et al. (1999) have proposed an alternate model for
certain duplications, one that takes into account the existence of
multiple expression domains typically found for products of single
genes. Accordingly, the duplication-degeneration-complementation (DDC)
model, partial loss-of-function mutations accumulate in the
cis regulatory regions of both paralogs, such that each copy of the gene is expressed in a particular spatio-temporal manner, and
both copies of the gene must be maintained to preserve the overall
function of the original gene. Thus, paradoxically, the accumulation of
degenerative mutations enhances the chances of survival of both
paralogs. The first example described for the DDC model involves the
two engrailed1 genes in zebrafish in which two sites, the
pectoral fin buds and spinal cord neurons, express only one of the
paralogs (Force et al. 1999). Members of the sox gene family
of developmental regulators show a similar partitioning of expression
(de Martino et al. 2000). The engrailed and sox duplicate pairs arose from a chromosomal duplication. However, in
principle, the nature of the duplication event giving rise to the two
paralogs should be irrelevant.
The Na(+),K(+)-ATPase or sodium pump is responsible for maintaining
proper intracellular and extracellular concentrations of sodium and
potassium ions, and is essential to membrane potential generation
(Glynn 1993; Lingrel and Kuntzweiler 1994). The protein consists of a
large multi-pass subunit and an associated glycosylated single-pass
subunit, and has been studied extensively at the biochemical level.
The highly conserved family of genes that encode the catalytic
subunits of the pump contain the binding sites for Na(+), K(+), ATP and
the digitalis glycosides (Lingrel and Kuntzweiler 1994). The mature
protein can be associated with a third, smaller 
subunit (Mercer et
al. 1993). Invertebrate genomes possess one subunit of the
Na(+),K(+)-ATPase (Emery et al. 1995). Four subunits have been
reported in mammalian systems (Kawakami et al. 1986; Shull et al. 1986;
Martin-Vasallo et al. 1989; Shamraj and Lingrel 1994; Malik et al.
1996). In mammals, the expression patterns of the different genes show
strong tissue preferences (Sweadner 1989; Blanco and Mercer 1998). The 1 gene (atp1a1) is preferentially expressed in the
kidney, gut, and heart, and ubiquitously at lower levels. The
2 (atp1a2) gene is expressed in muscle,
adipocytes, heart, and brain. The 3 (atp1a3) gene
is expressed throughout the nervous system. A less conserved fourth
isoform, 4, is expressed in the testes (Shamraj and Lingrel 1994). The 1 isoform is essential to cell survival.
Homozygous disruption of this gene in mice causes embryonic lethality
(James et al. 1999).
We evaluated these genes in the zebrafish because of evidence that it
has undergone an additional round of genome duplication over 100 million years ago (Amores et al., 1998). Through cloning and database
searches, we discovered that zebrafish have at least eight subunits. Five are in the 1 class, two in the 3
class, and one in 2 class. Sequence, mapping, and
expression analyses all suggest that even in the relatively short
period of time since gene duplication, they have acquired specialized
functions, supporting the subfunctionalization model of genome evolution.
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
RESULTS
DISCUSSION
METHODS
REFERENCES
|
|---|
Homology to Other Subunits
We cloned three subunits of the Na(+),K(+)-ATPase. Two of the
genes belong to the 1 family and the third is in the
2 group. From the predicted amino acid sequences, the two
1 genes share 91% identity between each other and 83%
identity with the 2 homolog (Fig.
1). This amino acid identity is preserved
when comparing the zebrafish subunits with their human counterparts.
atp1A1 and atp1B1 show 88% and 89% identity,
respectively, to the human 1 subunit and atp2 shares
86% identity to the human 2 protein.
|
A number of motifs essential for the sodium pump's function are noted
in Figure 1. The conserved cytoplasmic TGES/A motif, critical to the
catalytic cycle of all P-type ATPases, is present in all three of the
zebrafish subunits. The catalytically phosphorylated aspartyl
residue is also conserved and found within the consensus DKTGTLTQ
sequence. Protein kinases can modulate the catalytic activity of the
sodium pump in a complex fashion (Therien and Blostein 2000).
Phosphorylation by PKA is thought to occur at Ser 943 of the human and
rat sequences. This serine is also found in all three of the zebrafish
isoforms within the context of a PKA consensus site (XRRXSX). The
Na(+),K(+)-ATPase can also bind to ankyrin repeats (Zhang et al. 1998),
and this binding site is conserved in all three fish isoforms. The
presence of a leucine residue in the 1 genes instead of a
methionine is unusual, although it is found in the related P-type pump,
the gastric H(+),K(+)-ATPase.
Ouabain resistance in the rat 1 isoform has been mapped to
the extracellular loop between the first and second membrane-spanning segments. Mutation of two amino acid residues within this loop to those
of rat 1 can confer resistance to ouabain-sensitive isoforms (Price et al. 1990; Jewell and Lingrel 1992). The zebrafish genes have sequences predicted to be ouabain sensitive in that they
contain a glutamine or leucine rather than an arginine residue found at
the first position of the resistant isoform, and an asparigine instead
of an aspartate at the second.
Phylogenetic Analysis
To generate an evolutionary comparison, we identified additional
vertebrate genes in GenBank. Between gene cloning and
database analysis, we have identified a total of eight genes. Five are 1 paralogs, two are 3 paralogs,
and one is an 2 gene. (The 4 gene in mammals is
far less conserved and sequence information from multiple species is
lacking, so we did not include it in the analysis.) Protein maximum
likelihood analysis predicts the evolutionary tree shown in Figure
2.
|
Invertebrate genomes contain only one subunit (Emery et al. 1995),
which we used to root the tree. The first gene duplication created the
3 branch and an ancestral 12 gene, which then
duplicated to give rise to the 1 and 2
paralogs. Nodes that correspond to these duplications are shown in
black in Figure 2, and are likely due to the large-scale gene
duplications that occurred in the ancestral lineage of vertebrates. All
additional gene duplications appear only in the teleosts. The
duplication giving rise to the two 3 paralogs
(atp3A and atp3B) appears to have preceded the divergence of the Tilapia and zebrafish lineages (ProtML
BootP value = 83%) and is likely to have occurred as part
of the known large-scale gene duplication in the teleost lineage.
The 1 gene appears to have undergone additional rounds of
duplication. The first 1 gene duplication event appears to
have occurred in an early bony fish ancestor prior to the divergence of
the lineage leading to present day eels (Fig. 2, blue node). Placement
of the eel lineage before the gene duplication is rejected at a ProtML
BootP value of 0.001%, confirming that this duplication was
not due to the known large-scale teleost duplication. The two resulting
1 homologs then further duplicated, generating at least
five 1 genes in zebrafish. Duplication of the
atp1A group (gray node) occurred after divergence from the
Tilapia lineage (supported at BootP = 100%). A
subsequent duplication (yellow node) generated the other two known
atp1A homologs (atp1A2a and
atp1A2b). The one gene duplication in the subtree
containing atp1B1 (red node) occurred before the divergence
of the lineage leading to the sucker fish (Catostomus
commersoni), suggesting that this duplication may have been due to
the known large-scale teleost genome duplication.
Map Positions of the Zebrafish Catalytic Subunits
To examine the origin of the duplications, we mapped all of these
genes by somatic cell radiation hybrid (RH) mapping (Geisler et al.
1999). As shown in Figure 2, atp2 is on linkage group 2 (LG
2), whereas atp1A1 and atp1B1 both map to LG 1 in the z9394-z9382 interval. The atp3A gene maps to LG 16 close to the Hoxab cluster, and atp3B maps to LG
19 near the Hoxaa cluster. Both atp1A2a and
atp1A2b map to LG 1 in the same interval as atp1A1 and atp1B1, whereas atp1B2
maps to LG 9 between markers z9112 and z20031 (Shimoda et al. 1999).
Expression of Zebrafish 1 Isoforms
If the DDC model of subfunctionalization of genes after duplication
applies, expression patterns should have been partitioned among the
zebrafish ATPases relative to, for example, the single mammalian
1 gene. Therefore, we examined the expression of
atp1A1 and atp1B1.
The expression patterns of atp1A1 and atp1B1 in
early development are quite distinct. At the 17 somite stage,
atp1A1 is expressed in the intermediate mesoderm and, at
lower levels, in the otic placode (Fig.
3A,D), whereas the atp1B1 gene
is expressed in the optic cup, the developing nervous system, the otic
placode, and, weakly, in the intermediate mesoderm (Fig. 3B,E). The
atp2 gene is expressed in the developing somites (Fig.
3C,F). Thus, at this early stage, atp1A1 is the dominant
kidney isoform, whereas atp1B1 is expressed in a variety of
other tissues.
|
By 24 h postfertilization (hpf), atp1A1 is still expressed
in the nephric duct (Fig. 3G) and, to a lesser extent, in the otic
vesicle. We also detect weak expression of atp1B1 in the kidney, but it remains expressed in many tissues (Fig. 3H). Expression of atp2 remains exclusive to the myogenic lineage (Fig.
3I). At 48 hpf, expression of atp1A1 in the ear has
increased and is comparable with the levels seen in the nephric duct
(Fig. 3J). The dominant isoform in the nervous system is
atp1B1 (Fig. 3K). It is also transiently expressed in the
pectoral fin buds (Fig. 3K) and the heart, in which the ventricular
myocardium expresses much higher levels of atp1B1, than
does the atrium (Fig. 3L). The atp2 gene remains in the
somites and is also seen in the fin buds. (Fig. 3M).
At 96 hpf, atp1B1 is highly expressed in the brain (Fig.
3O). Low levels of atp1A1 can also be detected in this
tissue (Fig. 3N). Both genes are expressed in the pronephric tubule and
gut to a similar level (Fig. 3N,O). However, only the
atp1B1 paralog is detected in the developing liver (Fig. 3,
cf. N and O).
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
RESULTS
DISCUSSION
METHODS
REFERENCES
|
|---|
Tandem and Ploidy Gene Duplications
Invertebrate genomes contain one sodium pump catalytic subunit
(Emery et al. 1995). We show here that the three vertebrate paralogs
originated by gene duplications that preceded the last common ancestor
of jawed vertebrates (which is represented by the node in the 1
subtree from which the lineage to Torpedo californica originated; Fig. 2). We find that the zebrafish genome encodes at least
eight subunits that fall into the 1, 2, or
3 subfamilies. Our phylogenetic analysis indicates that a
gene duplication of the 1 gene took place in the bony fish
lineage, creating the 1A and 1B subfamilies.
Further duplications within each of the branches generated the
additional 1 genes found in the zebrafish. Whether this
additional expansion is unique to zebrafish awaits further genomic
analysis of other fish.
Combining data from the phylogenetic analysis and mapping, a clear
picture of the gene duplication events in teleosts emerges (Fig.
4). The first duplication, which created
the 1A and 1B subfamilies, was likely a tandem
duplication because (1) it preceded the known large-scale duplication,
which occurred after divergence of the lineage leading to eels, and (2)
1A and 1B both map to the same small interval
on LG 1 (Fig. 2).
|
The duplication giving rise to the two 3 paralogs and the
duplication in the 1B subfamily are likely due to the
teleost large-scale duplication. Both the linkage data and the
phylogenetic analysis favor this interpretation. The two 3
paralogs map to LG 16 and LG 19, linked to the duplicated Hoxa
clusters (Amores et al. 1998; Gates et al. 1999; Woods et al. 2000),
and the duplication occurred prior to divergence of the lineage leading
to Tilapia. Similarly, the 1B paralogs map to LG 1 and LG 9, which have been suggested to contain paralogous regions, as
teleost-specific members of other gene families, such as
distalless and engrailed, map to these chromosomes
(Gates et al. 1999; Woods et al. 2000). Here too, the phylogenetic
position of the duplications is consistent with the large-scale
duplication. Duplicates of the atp1A lineage are unlikely
to be due to the large scale gene duplication, as both duplications
follow the species divergence node with Tilapia. Radiation
hybrid mapping of all three genes indicates that they lie in the same
area of LG 1, suggesting that they arose by tandem duplication.
Divergence of Expression and the DDC Model
Gene duplications have been proposed to be an important mechanism
driving diversification during evolution. According to classical models, duplicated genes can accumulate mutations and acquire novel
functions. Alternatively, they may be lost, such as one of the
Hoxd clusters in zebrafish, or may degenerate to a pseudogene, as Hoxaa-10 has in zebrafish (Stellwag 1999). According to the DDC model (Force et al. 1999), some degenerative mutations may actually
favor the preservation of both paralogs through complementation of
their subfunctions.
Thus far, evidence for the DDC model has come from genes encoding
developmental regulators. The zebrafish genome contains two
engrailed1 genes expressed in two distinct locales of a few cells each, in the fin buds and spinal cord (Force et al., 1999). The
expression pattern of the duplicated sox11 genes in zebrafish is also partitioned. Whereas the single sox11 gene is
expressed throughout the somites in the mouse, two zebrafish paralogs
share this expression domain; sox11a is expressed anteriorly
and sox11b posteriorly (de Martino et al., 2000). Here we
present data that extend this model to genes essential for survival of
every cell in the body. Interestingly, data from studies of the
tissue-specific expression of isozymes of lactate dehydrogenase and
alcohol dehydrogenase (Li et al. 1983; Edenberg 2000) are also
consistent with a DDC mechanism; future genetic mapping and
evolutionary anlyses of these genes may provide further insights into
the molecular mechanisms of DDC.
The DDC model predicts that duplicate genes partition functions
normally performed by the original gene. Our expression results for
atp1A1 and atp1B1 are consistent with this
model. The 1 isoform in mice is expressed in the kidney,
heart, brain, and gut. In zebrafish, the atp1A1 gene is the
dominant kidney isoform expressed very early in nephrogenesis and at
very high levels. The atp1B1 gene is highly expressed in
the brain. The other 1 paralog (atp1A1) is also
expressed in the brain, but much later in development and at lower
levels. Only atp1B1 expression is detected in the heart and
liver. The model further predicts that the broader the expression
domains associated with a gene, the more likely duplicated paralogs
will be maintained. The 1 gene in mammals is ubiquitously
expressed, performing essential functions in all cell types of the
organism. In other words, it contains the greatest potential number of
subfunctions. Consistent with this prediction, the 1 family
in zebrafish contains the greatest number of paralogs with five.
Prior data for the engrailed1 genes indicate that partitioning
of expression domains can follow a chromosomal duplication event. Our
mapping, expression, and evolutionary data for the family members
are consistent with these prior observations and reveal, as well,
examples of subfunctionalization following a tandem duplication event.
These data lend further support to the DDC model, extending it to
ubiquitously expressed genes and to tandem duplications.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
RESULTS
DISCUSSION
METHODS
REFERENCES
|
|---|
Zebrafish Lines
Fish were raised and maintained in the Cardiovascular Research
Center fish facility at the Massachusetts General Hospital. Embryos
were kept in E3 medium at 28.5°C and staged according to somite
number or hours postfertilization (hpf) (Kimmel et al. 1995). Wild-type
fish were of the Tübingen background (obtained from Dr. Christianne
Nüsslein-Volhard, Tübingen).
cDNA Cloning and Sequence Analysis
We identified an expressed sequence tag (accession no. AA494679)
encoding a zebrafish subunit in the public databases and PCR
primers were designed to amplify a portion of this coding sequence. We
used the PCR product to screen 5 × 106 plaques at low
stringency from a 24-h zebrafish cDNA 
ZAP Express library
(Stratagene). We selected 20 plaques for further study of ~300
positives obtained. Ten strong positives and ten weaker positives were
selected and rescued. Full-length clones were completely sequenced on
both strands and characterized. We retrieved homologs from GenBank by
BLAST search, and aligned the amino acid sequences with
CLUSTALW (Thompson et al. 1994). Regions of uncertain
homology around gaps were excluded from subsequent phylogenetic
analysis by use of PROTML (Adachi and Hasegawa 1996). The
tree was rooted with invertebrate sequences that were later excluded
from the analyses of the vertebrate sequences to maximize the number of
unambiguously homologous sequence positions. The topologies of 1,
2, and 3 subtrees were determined in independent analyses, and a
global analysis allowed determination of the relative branching order
of the three paralogs. Branching order in those parts of the tree that
are important for our interpretation of the nature of the nodes was
tested explicitly by use of the user tree option in
PROTML.
In Situ Hybridizations
Embryos of the appropriate stage were fixed in 4%
paraformaldehyde, stored in methanol and processed as described
previously (Jowett 1999). The cDNA clones (in pBK-CMV) were linearized
with BamHI and T7 RNA polymerase used to transcribe antisense
DIG-labeled riboprobes.
Radiation Hybrid Mapping and Genotyping
We designed primers to amplify a portion of the 3' untranslated
region of each subunit. We also mapped GenBank entries representing other -like genes. In Table 1, we show
our proposed nomenclature consistent with the phylogenetic data
presented in this study.
|
The cycling profile was as follows: 90 sec of denaturation at 94°C, 30 cycles with 30 sec at 94°C, 30 sec at 60°C, and 1 min at 72°C. Somatic cell radiation hybrid mapping was carried out by use of the Goodfellow T51 panel (Research Genetics) and map positions calculated by use of the RH mapping service at Boston's Children's Hospital (http://genetics.med.harvard.edu/~zonlab/).
| |
ACKNOWLEDGMENTS |
|---|
We thank Alex Therien for critical comments on earlier versions of this manuscript and for helpful discussions and Sarah Childs for helpful criticism of the manuscript. This work was supported in part by National Institutes of Health grants RO1DK55383 and RO1HL63206 to M.C.F.
The publication costs of this article were defrayed in part by payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 USC section 1734 solely to indicate this fact.
| |
NOTE ADDED IN PROOF |
|---|
An analysis of zebrafish subunit genes has also been performed
by Rajarao et al. (2001).
| |
FOOTNOTES |
|---|
3 Corresponding author.
E-MAIL fishman{at}cvrc.mgh.harvard.edu; FAX (617) 726-5806.
Article published on-line before print: Genome Res., 10.1101/gr.192001.
Article and publication are at http://www.genome.org/cgi/doi/10.1101/gr.192001.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
RESULTS
DISCUSSION
METHODS
REFERENCES
|
|---|
and subunit genes expressed in the zebrafish, Danio rerio.
Genome Res.
11:
1211-1220.Received April 12, 2001; accepted in revised form June 4, 2001.
CiteULike 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  X. Shu, J. Huang, Y. Dong, J. Choi, A. Langenbacher, and J.-N. Chen Na,K-ATPase {alpha}2 and Ncx4a regulate zebrafish left-right patterning Development, May 15, 2007; 134(10): 1921 - 1930. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. S. Silva, A. I. Duarte, A. C. Rego, C. R. Oliveira, and P. P. Goncalves Effect of Chronic Exposure to Aluminium on Isoform Expression and Activity of Rat (Na+/K+)ATPase Toxicol. Sci., December 1, 2005; 88(2): 485 - 494. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Yuan and E. M. Joseph The small heart Mutation Reveals Novel Roles of Na+/K+-ATPase in Maintaining Ventricular Cardiomyocyte Morphology and Viability in Zebrafish Circ. Res., September 17, 2004; 95(6): 595 - 603. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Shu, K. Cheng, N. Patel, F. Chen, E. Joseph, H.-J. Tsai, and J.-N. Chen Na,K-ATPase is essential for embryonic heart development in the zebrafish Development, December 22, 2003; 130(25): 6165 - 6173. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Dorus, S. L. Gilbert, M. L. Forster, R. J. Barndt, and B. T. Lahn The CDY-related gene family: coordinated evolution in copy number, expression profile and protein sequence Hum. Mol. Genet., July 15, 2003; 12(14): 1643 - 1650. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Winkler, M. Schafer, J. Duschl, M. Schartl, and J.-N. Volff Functional Divergence of Two Zebrafish Midkine Growth Factors Following Fish-Specific Gene Duplication Genome Res., June 1, 2003; 13(6): 1067 - 1081. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Altschmied, J. Delfgaauw, B. Wilde, J. Duschl, L. Bouneau, J.-N. Volff, and M. Schartl Subfunctionalization of Duplicate mitf Genes Associated With Differential Degeneration of Alternative Exons in Fish Genetics, May 1, 2002; 161(1): 259 - 267. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.-K. J. Shen Sharing Duties in the Family Genome Res., October 1, 2001; 11(10): 1615 - 1615. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Subunit Gene
