Isozyme evidence for independently derived, duplicate G3PDH loci among squamate reptiles

Transcription

1 Isozyme evidence for independently derived, duplicate G3PDH loci among squamate reptiles JACK W. SITES, JR. Department of Zoology, Brigham Young University, Provo, UT, USA AND ROBERT W. MURPHY Department of Ichthyology and Herpetology, Royal Ontario Museum, 100 Queen's Park, Toronto, Ont.. Canada M5S 2C6 Received September 1 1, 1990 SITES, J. W., JR., and MURPHY, R. W Isozyme evidence for independently derived, duplicate G3PDH loci among squamate reptiles. Can. J. Zool. 69: We report evidence for several independent gene duplications for the locus encoding the enzyme glycerol-3-phosphate dehydrogenase (G3PDH) in squamate reptiles. Evidence for the duplication comes frompopulation genetic studies demonstrating "fixed" heterozygosity in all members of some lizard species, the documentation of independent allelic heterozygosity at each of the two G3PDH loci in these same species, and tissue-specific gene expression surveys in a taxonomically diverse array of groups. The duplicated condition is present at both low and high taxonomic levels (selected populations of the phrynosomatid lizard Sceloporus grammicus, and almost all snakes, respectively), and appears to represent the derived condition in most of these groups. One notable exception is the colubrid snake genus Masticophis, which appears to be characterized by an apomorphic secondary silencing event. Evolutionary implications of the duplication and silencing events within squamates are discussed, and we suggest that the overall phylogenetic utility of this marker is low in this radiation as a result of extensive homoplasy. SITES, J. W., JR., et MURPHY, R. W Isozyme evidence for independently derived, duplicate G3PDH loci among squamate reptiles. Can. J. Zool. 69 : Nous avons constat6 l'existence de plusieurs duplications indkpendantes de gbnes au locus codant l'enzyme glycerol-3- phosphate dcshydrogcnase (G3PDH) chez des reptiles squamates. L'existence de la duplication a CtC mise en lumibre au cours d'ctudes gcnctiques de populations qui ont dcmontrc une hctcrozygotie (( fixce >> chez tous les individus de certaines espkces de lizards; l'existence de la duplication est Cgalement dtmontrce par I'hCtCrozygotie allclique indkpendante B chacun des deux locus G3PDH chez ces m$mes espbces; des etudes de l'expression de gbnes dans certains tissus specifiques au sein d'un groupe de taxons divers ont Cgalement revel6 l'existence de la duplication. La condition de duplication se retrouve 5 la fois aux echelons bas et hauts d'cchelles taxonomiques (chez des populations choisies du lczard phrynosomatidc Sceloporus grammicus d'une part, chez presque tous les serpents d'autre part) et semble une caracteristique dcrivce chez la plupart de ces groupes. Une exception cependant, celle du genre de colubridc Masticophis qui semble caractcrisc par un CvCnement secondaire masquant apomorphe. La signification Cvolutive de la duplication et des CvCnements masquants chez les squamates fait l'objet d'une discussion; il semble que I'utilitC phylogenktique gcntrale de ce marqueur soit faible au sein de cette branche B cause d'une importante homoplasie. [Traduit par la rcdaction] Introduction a correlation between an increase in locus number and tissue Understanding the nature and frequency of gene duplication specificity with phyletic advancement. Similar patterns might be events is important for several reasons. First, duplication of some expected for some of the same enzyme systems in various loci provides a mechanism for the production of gene products tetrapod groups, but tetrapods have generally not been screened : needed in large quantities in certain environments or during as intensively as fishes (for a very limited survey see Fisheret al. select stages of development. Second, duplications initially 1980). produce functionally redundant loci which are thought to serve The monophyletic group Reptilia, as defined by Gauthier et al. " as the raw material for novel adaptations (Ohta 1987). Many duplicate enzymes, for example, are functionally redundant because they catalyze the same reactions (albeit often with different efficiencies), and thus provide greater regulatory "flexibility" in that one locus is free to accumulate new mutations randomly and perhaps acquire a novel function (Ohno 1970; Whitt 1983, 1987). Gene redundancy also accelerates the spead of "compensatory" mutations (those correcting the effect of an earlier deleterious mutation), owing to relaxation of selective constraints (Ohta 1989). Third, duplicate enzymecoding loci often produce electrophoretically distinct isozyme patterns, and these may have systematic utility (Buth 1984). The fact that strong evidence for duplication events can be collected electrophoretically (MacIntyre 1976; Li 1982) has allowed for widespread and relatively inexpensive screening of isozyme patterns, especially in plants (Gottlieb 1982) and fishes (Buth 1983; Whitt 1983,1987). Frequently, these enzymes show Printed in Canada 1 Imprime au Canada (1988), occupies a key phyle& position among tetrapods, and discoveries concerning the frequencies and distribution of gene duplications in this group are of interest because they have implications for patterns of genome evolution in most other living tetrapods. We have electrophoretically surveyed several different groups of squamate reptiles during the course of ongoing population, genetic, and systematic studies in our laboratories, and have discovered isozyme evidence for independent duplications of an originally single locus coding for the enzyme glycerol-3-phosphate dehydrogenase (G3PDH, EC ; also known as a-glycerophosphate dehydrogenase, a-gpd). Previously, Fisher et al. (1980) reported only one G3PDH locus in reptiles, on the basis of screening liver and muscle extracts from a single specimen of the lizard Anolis carolinensis, while Hall and Selander (1973) and Densmore and Dessauer (1983) (see also Gartside et al. 1977) reported a twolocus system in some populations of another lizard, Sceloporus

2 2382 CAN. J. ZOOL. VOL. 69, 1991 grammicus, and in Alligator mississippiensis, respectively. In S. grammicus, both loci were expressed in the same tissue extracts, and appeared as three-banded "fixed" heterozygotes (see also Sites and Greenbaum 1983; Sites et al. 1988). A threebanded pattern was also reported by Thompson and Sites (1986) in the sagebrush lizard, Sceloporus graciosus. In Alligator, the two loci are predominantly expressed in different tissues, one in liver and a second in skeletal muscle. Finally, Murphy and Crabtree (1985) reported two G3PDH loci with the same mutually exclusive patterns of expression in the prairie rattlesnake, Crotalus viridis viridis. G3PDH appears to evolve slowly in vertebrates with respect to the rate of amino acid substitution (Ballas et al. 1984), but the emerging pattern from our work and other published reports on reptiles suggests that (i) some lineages show evidence of relatively old duplications, as inferred from restricted patterns of tissue-specific expression (A. mississippiensis and C. viridis); (ii) other lineages appear to have undergone more recent duplications, in which both loci are expressed in the same tissue, and are structurally similar enough to form interlocus heterodimers (S. graciosus and some populations of S. grammicus); and (iii) other lineages either have never fixed duplications at this locus, or these may have occurred once and then were followed by a silencing of one locus, so that only a single locus is now expressed (Anolis carolinensis (?), Fisher et al. 1980). Other reports of single-locus systems include those of Dessauer and Cole (1984) in the teiid lizard Cnemidophorus tigris, and Murphy and Matson (1986) in the tuatara, Sphenodonpunctatus. In this paper, we summarize patterns of isozyme expression in lizards and snakes, including literature reports and unpublished work from our laboratories, and document that (i) the threebanded G3PDH isozyme phenotypes reported as present in most or all individuals of some samples are most parsimoniously interpreted as the expression of two loci, rather than two electromorphs segregating at a single locus or epigenetic modification of a single locus product; and that (ii) this presumed duplication has probably occurred independently in several groups of squamates. Materials and methods Our electrophoretic and histochemical staining methods follow Murphy et al. (1990) unless specified otherwise. The isozyme nomenclature follows Murphy and Crabtree (1985) in abbreviating the entire enzyme system with capital letters (G3PDH), identifying a specific locus by a capital letter postscript following the enzyme abbreviation with only the first letter capitalized (G3pdh-A), and identifying specific electromorphs (i.e., presumed alleles, see Allendorf 1977) by lower case letters in following the locus designation (~3pdh-~(aa)). Ideally. demonstration of two distinct structural loci encoding the same enzyme should be based on breeding studies showing indepenhent inheritance of the separate gene products (Soltis et al. 1987). In a broad survey such as this, it is impossible to identify intraspecific allozyme polymorphisms at each locus and then undertake a breeding program for each species. We therefore indirectly inferred the number of G3PDH loci (see below). The data set consists of previously published surveys of squamates for which the number of G3PDH loci was reported, and unpublished data from our own laboratories. All of these are included in the Appendix, which summarizes for each species the number of individuals and types of tissues screened and the observed number of G3PDH loci. For some taxa, two loci were inferred by the within-sample presence of a three-banded isozyme phenotype in frequencies much greater than would be expected from a simple diallelic segregation at a single locus. For two species, we have demonstrated apparent independent allelic segregation at each of the two loci, and have illustrated the resulting multibanded isozyme phenotypes. In the majority of taxa, tissuespecific mobility and activity differences were used to infer the presence of two loci, even though these species were usually represented by few individuals and localities. However, we are reasonably confident about the reported number of loci encoding G3PDH in these species because most electrophoretic surveys included liver and muscle samples. G3PDH is either most strongly expressed in liver and skeletal muscle, or entirely restricted to one or both of these tissues in species ' for which extensive tissue surveys have been carried out, including the alligator (Dessauer and Densmore 1983), several species of the lizard genera Cnemidophorus (Dessauer and Cole 1984) and Sceloporus (this paper), C. v. viridis (Murphy and Crabtree 1985), and the tuatara, (Murphy and Matson 1986). To infer patterns of G3PDH evolution within squamates, the distribution of duplications was mapped onto two different cladograms (conceptual issues are discussed by Felsenstein 1985 and Donoghue 1989). The first was derived by Hall (1973, 1980) for the lizard genus Sceloporus on the basis of a chromosome data set in which speciesgroups were defined by one or more synapomorphic autosomal or sexchromosome rearrangements relative to the symplesiomorphic 2n = 34 (XY male) karyotype (see also Paul1 et al. 1976). A second cladogram was modified from those produced for all Squamata by Estes'et al. (1988) on the basis of extensive morphological and osteological data sets. Estes et al. produced several alternative cladograms, but since no single hypothesis was heavily supported relative to the alternatives, we adopted their conservative consensus hypothesis for our comparisons (Estes et al. 1988, Fig. 6, p. 140). Results Evidence for a G3PDH gene duplication The best-documented examples of the presumed G3PDH duplication are found in two species of the North American lizard genus Sceloporus. The S. grammicus complex includes a minimum of seven chromosomally differentiated populations (cytotypes) ranging across most of mainland Mexico and into southern Texas. A number of intensive studies have examined the interplay between population structure and the fixation of chromosomal rearrangements in this group (reviewed by Sites et, al. 1987; Arevalo et al. 1991), and four electrophoretic studies have been undertaken on various portions of this complex (Hall and Selander 1973; Sites and Greenbaum 1983; Sites et al. 1988; ' Sites and Davis 1989). Results of thse studies have revealed between-populations variability in the number of G3PDH isozymes expressed in homologous tissues. Figure 1 shows the geographic locations from which these populations have been sampled, and Table 1 summarizes relevant data concerning sample sizes and number of expressed G3PDH loci. Table 1 shows that nine populations of the 2n = 32 cytotype and one of the 2n = 34 race consistently display two loci, while all others show only a single locus. The pattern within the 2n = 32 race is FIG. 1. Approximate ranges of several chromosome races of Sceloporus grammicus, and the localities of origin of the samples that were electrophoretically screened for G3PDH. Sample sizes and G3PDH character states (one or two loci) are given for each locality in Table 1; "FM" refers to three different "multiple fission" races with diploid numbers ranging from 38 to 46 (see details in Arevalo et al ). The heavy broken line indicates the Texas-Mexico border, and the light broken lines the borders of Mexican states, which are indicated as follows: Coah, Coahuila; Gto, Guanajuato; Hgo, Hidalgo; Mich, Michoacan; Mx, MCxico; NL, Nuevo Leon; Qto, Queretaro; SLP, San Luis Potosi; Tamp, Tamaulipas; Ver, Veracruz; and Zac, Zacatecas. The shaded areas at the bottom identify areas of mountain ranges above 3000 m that surround the Valley of Mexico.

3 SITES AND MURPHY 2383

4

5

6

7

8

9

10

11

12

13

14

15

16