The Journal of Heredity 2002:93(2)
© 2002 The American Genetic Association 93:81-85
Population Genetics of a Polyploid: Is There Hybridization Between Lineages of Hyla versicolor?
From the Department of Biological Sciences, Life Sciences Bldg. 206, Louisiana State University, Baton Rouge, Louisiana 70803. N. R. Espinoza is currently at the Department of Medicine/Division of Oncology, Stanford University Medical Center, Stanford, CA 94305.
Address correspondence to Mohamed Noor at the address above or e-mail: mnoor{at}lsu.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Several studies have demonstrated that polyploid species can form recurrently from their progenitors, but few studies have evaluated gene flow between the resultant polyploid lineages. Here we examine the possibility of hybridization between lineages of the tetraploid common gray treefrog (Hyla versicolor). We utilize a polymerase chain reaction (PCR) cloning approach to estimate the genotypes of tetraploid individuals and measure genetic differentiation between (1) sympatric populations of two lineages and (2) allopatric populations of a single lineage. We find that allele frequencies in sympatric populations of two lineages do not differ, suggesting that frogs of these two lineages hybridize in areas where they co-occur.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
One of the most exciting recent discoveries in speciation is that many polyploid species formed recurrently from natural populations of their progenitors [for reviews see Soltis and Soltis (1999, 2000)]. Hence, rather than representing a single, discrete speciation event, polyploidy events creating a particular new species can occur many times, and more variation can be introduced into the polyploid species from the diploid species than was previously thought. Several recognized polyploid species may thus be polyphyletic.
One question that has not been fully addressed, however, is how often the resultant species of separate polyploidy events are reproductively compatible. Although these resultant species will have the same number of chromosomes, they may have sufficient sequence divergence to produce some (or possibly complete) reproductive isolation. In this study we examine the possibility of hybridization between distinct lineages of the common gray treefrog (Hyla versicolor).
Using mitochondrial cytochrome b sequences, Ptacek et al. (1994) concluded that the North American tetraploid H. versicolor originated from the diploid Hyla chrysoscelis at least three times (see also Wiley and Little 2000). Given that polyhaploidy is extremely uncommon (Jones 1970; Stebbins 1970), and given that all other closely related Hyla species are diploid, this conclusion is more parsimonious than the multiple origins of the diploid from the tetraploid species in a few million years. The three cytochrome b lineages of H. versicolor were somewhat separated geographically, comprising a "southwestern" (SW) race in Tennessee, Oklahoma, Texas, and Louisiana; an "eastern" (E) race found in areas including Virginia, West Virginia, and Maine; and a "northwestern" (NW) race in Missouri, Minnesota, and Canada. The NW and SW races are sympatric in Missouri, as evidenced by their mitochondrial haplotypes being isolated from the same ponds (Porter D and Ptacek M, unpublished manuscript). Sequences of individuals within lineages differed by 0.18–0.53%, while sequences of individuals between lineages differed by 1.9–3.4%. The NW lineage was the outgroup to the other two lineages, and a cytochrome b sequence divergence of 3.4% could suggest a long enough separation to have allowed for the evolution of reproductive isolation (Johns and Avise 1998). The strong association of these lineages with particular geographic areas suggests that they may not be freely exchanging migrants, or at least female migrants.
Nuclear genetic data can help to address whether gene flow between lineages is occurring, particularly when highly variable markers are used. However, evaluating gene flow between tetraploid (or other types of polyploid) populations is more challenging than between diploid populations. When using molecular genetic markers such as allozymes, restrictions fragment length polymorphisms (RFLPs), or microsatellites, one cannot easily identify the genotype of a particular individual if it bears two distinct alleles. For example, if a tetraploid individual is heterozygous for two size alleles denoted A and a, the following three genotypes would appear identical on a standard gel: one copy of the A allele and three copies of a, two copies of A and two copies of a, or three copies of A and one copy of a (see the top of Figure 1). Investigators have circumvented this problem by either estimating genotype based on the apparent density of particular bands in heterozygotes (e.g., Romano et al. 1987) or by arbitrarily assigning heterozygotes a genotype of two copies of A and two copies of a (e.g., Allard et al. 1993). Neither of these approaches is fully satisfying.
|
Here we used a PCR-product cloning approach for estimating the genotype of polyploid individuals that are heterozygous for two or more alleles. We then apply this approach in evaluating whether distinct lineages of H. versicolor hybridize in sympatry. To address this question we compare allele frequencies between (1) geographically proximate populations represented by distinct cytochrome b lineages (NW and SW) and (2) two geographically distant populations of a single cytochrome b lineage (NW). We predict that if these lineages hybridize in nature, geographically close populations will harbor similar allele frequencies irrespective of lineage. Conversely, in the absence of gene flow between lineages, geographically distant populations within a particular cytochrome b lineage will bear more similar allele frequencies than geographically close populations of different cytochrome b lineages. We find that allele frequencies vary more with geographic location than with lineage, and we conclude that the NW and SW lineages of H. versicolor do appear to hybridize in sympatry.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Sampling Localities
Ethanol-preserved frog muscle samples were generously provided by Margaret Ptacek, Idaho State University. We obtained samples of H. versicolor from Saline, MO (2); Boone, MO (3); Ottawa, OK (4); Guelph, Ontario (5); Howell, MO (3); Hoard, TX(4); and Mecklenberg, VA (10). In one case (see below), we used a sample of the diploid H. chrysoscelis from Stillwater, OK.
Determination of Lineage
The lineages of H. versicolor were identified by amplifying and sequencing cytochrome b in each of our samples. Frog DNA was prepared using the Puregene DNA isolation kit (Gentra Systems). PCR was performed in a 50 µl reaction volume with 0.5 µM of each primer, 200 µM dNTP, 5 µl 10x buffer (100 mM Tris, pH 8.3, 500 mM KCl, 15 mM MgCl2), 1 U Taq polymerase, and 1 µl from the DNA preparation. PCR was executed with 30 cycles of 1 min at 94°C, 30 s at 58°C, 2 min at 72°C, and 5 min at 95°C. PCR products were cleaned using a Qiagen PCR purification kit and resuspended in 30 µl of water. We then used the ABI BigDye sequencing kit and protocols for the sequencing reaction, and the products were run on the ABI 377 automated sequencer at the Museum of Natural Science, Louisiana State University.
We performed a parsimony analysis on 526 bases of sequence with equal weighting of transitions and transversions using PAUP* (Swofford 1998). We used published data from a sample of Hyla arenicolor as an outgroup (Ptacek et al. 1994). The parsimony analysis was subjected to a nonparametric bootstrap with 1000 pseudoreplicates to determine the support of the identified branches (Felsenstein 1985). The lineages we identified were then compared to previously documented lineages by sequence similarity (Ptacek et al. 1994).
Genetic Markers
Three of the four genetic markers used in this study were microsatellites documented previously (Krenz et al. 1999): 6M8C, 19F9G, and 26J1A. We identified a fourth marker by sequencing an approximately 800 bp random clone from a genomic DNA library of H. versicolor. When primers of this clone were designed (F: 5'-TGGCTAGGACGGATGCGTAC-3'; R: 5'-ACAATCGCCTGCTCTTGAATATGC-3') and further sequences performed, we found that the restriction enzyme, HphI, differentiated two alleles by cutting one in a single location and the other at two locations. The sequences of these alleles have been submitted to GenBank (accession number AZ694804).
Genotyping Procedure
Using the frog DNA isolated above, PCR was performed for all markers in a 20 µl reaction volume using the ingredients described above, with PCR following a standard touchdown protocol (Palumbi 1996). PCRs of the random clone were digested for 1 h with one unit of HphI at 37°C. Products were visualized on a 2.5% MetaPhor agarose gel. Alleles were confirmed on acrylamide gels using a LiCor DNA sequencer/analyzer. If frogs were homozygous for a particular marker, they were scored as possessing either two copies of the particular allele (if diploid, H. chrysoscelis) or four copies of the particular allele (if tetraploid, H. versicolor).
If the frogs were heterozygous for alleles at a locus, PCR was repeated in a 50 µl reaction volume. PCR products were cleaned using the Qiagen gel extraction kit, and cloned into plasmids in Escherichia coli using the Invitrogen TOPO-TA cloning kit, following the procedures of both kits. E. coli was plated onto ampicillin-resistant LB plates, and 20–55 clones (generally more than 30) were analyzed by PCR a third time in a 50 µl reaction volume. The PCR products from the clones were run alongside the remaining original PCR products on an agarose gel. Using the frequencies of particular alleles from among the clones, we estimated the genotype of the particular frogs. For example, if 40 clones were assayed, and 30 possessed one allele while 10 possessed another, we estimated that the frog had three copies of the first allele and one copy of the second (see Figure 1). With a sample size of 30, we have a greater than 75% probability of correctly identifying the genotype of an individual with this method, assuming equal amplification of products (see below). No products were observed in clones that were not observed in the genomic DNA PCR, suggesting little or no in vitro mutations relevant to our study.
Our methodology was tested using the random clone RFLP in the diploid H. chrysoscelis. The individual tested was a heterozygote, necessarily possessing a copy of each allele. Our method correctly identified the frequency of alleles within this individual to be nearly 50% (43% of clones possessed one allele).
Analysis
We hypothesized that if respective lineages of H. versicolor are reproductively isolated, allopatric populations of the NW lineage should be more similar to each other in allele frequencies than are sympatric populations of the NW and SW lineages. The NW lineage diverged from the E and SW lineages several million years ago according to the cytochrome b sequence data (Ptacek et al. 1994), suggesting that it should have diverged in allele frequencies substantially, particularly as the origin of each lineage should have been accompanied by a strong founder effect. Conversely, if genetic exchange occurs between the two lineages in sympatry, interlineage populations should exhibit much less differentiation than should geographically distant NW populations. This prediction may also apply to the two SW lineage populations; however, the geographic distance between SW populations is much smaller (less than 1000 km), which may result in a smaller genetic difference.
Allele frequency data were entered into FSTAT (Goudet 1995), and FST statistics were calculated. The statistical significance of differentiation between the populations was evaluated by randomizing genotypes across samples 1000 times, without assuming random mating within samples (see Goudet et al. 1996).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
From our sequencing efforts, we identified three strongly supported mitochondrial lineages among our H. versicolor individuals (Figure 2). The H. arenicolor individual differed from all the ingroup samples at 64 base positions, and an additional 27 base positions were phylogenetically informative among the remaining individuals. One lineage corresponds with the NW lineage documented by Ptacek et al. (1994), and consists of the individuals from Boone, MO; Saline, MO; Ottawa, OK; and Guelph, Ontario. A second lineage, corresponding with the published SW lineage, consisted of the individuals from Howell, MO, and Hoard, TX. The 10 individuals from Mecklenberg, VA, formed a third lineage (eastern). These three H. versicolor lineages correspond with the three lineages identified previously in a separate study (Ptacek et al. 1994). Although our samples did not include a sampling location with frogs of two lineages, Boone and Howell are only about 100 km apart, suggesting that frogs of the NW and SW lineages probably encounter one another, as proven by Porter and Ptacek (unpublished manuscript).
|
The NW lineage H. versicolor samples were pooled as follows. First, all Missouri and Oklahoma samples were considered to form a single population that is sympatric with the SW lineage (Ottawa, OK, is on the Missouri border) designated as "NW Missouri." These populations were all within 400 km of each other. Previously Romano et al. (1987) evaluated allele frequencies at 11 polymorphic allozyme loci from two H. versicolor populations in this area approximately 400 km apart (Miami, MO, and Manito, IL), and virtually no difference in allele frequencies were detected, so this pooling is reasonable. Second, the Guelph, Ontario, samples were considered to form a second, allopatric population. The closest sampling points between these populations are approximately 1200 km apart. All Missouri samples of the SW lineage were considered to be sympatric with the NW lineage, while the Texas samples of the SW lineage were considered to be allopatric.
All four nuclear markers exhibited variability within the NW and SW lineages of H. versicolor, though two markers were fixed in the sampled E lineage population (Table 1). The three microsatellite markers each had three to four size alleles that could be reliably diagnosed on high-density agarose gels, and two alleles were noted in our RFLP analysis of the random genomic clone. Allele frequencies are different between populations within the lineages and between the lineages as a whole. The E lineage population was the most genetically distinct.
|
FST between the NW Ontario and NW Missouri populations was 0.118 and statistically significant (P < .05). In contrast, FST between the NW Missouri and SW Missouri population was 0.037 and not statistically significant (P > .10). The two Missouri populations are thus more genetically similar to each other than are the two NW lineage populations. FST between the SW Missouri and Texas populations was 0.072, though also not statistically significant (P > .10). The smaller genetic distance between the two SW lineage populations than between the two NW lineage populations may reflect the closer geographic proximity of the former (see Materials and Methods).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Using a PCR-product cloning approach, we estimated the genotypes of tetraploid individuals and used these data to determine whether gene flow has occurred between the independent lineages of H. versicolor. We found that geographically distant populations of the NW lineage were significantly differentiated (FST = 0.12, P < .05), but geographically close populations of different lineages were not (FST = 0.04, NS). These genetic data suggest that the different lineages hybridize in areas where they co-occur.
If these lineages were not hybridizing, we would have expected substantial differentiation between the NW and SW lineages, in part because polyploidy events are necessarily associated with severe bottlenecks (down to four chromosomes). The cytochrome b sequence divergence between these two lineages is approximately 3%. Assuming a divergence rate of 0.75%/million years (e.g., Martin and Palumbi 1993), the NW lineage was formed by a polyploidy event (and associated genetic bottleneck) that occurred approximately 4 million years ago. Similarly the bottleneck resulting in the formation of the SW lineage would have occurred more than 2 million years ago. Thus it seems highly improbable that these two lineages would have allele frequencies that do not differ significantly in the absence of hybridization.
Other data also support the conclusion of Porter and Ptacek (unpublished manuscript) that hybridization occurs where the NW and SW lineages are sympatric. Carl Gerhardt's group (University of Missouri, personal communication) has also identified call differences between the NW and SW lineages of H. versicolor, and they noted frogs with intermediate calls in the Missouri area, suggesting possible hybridization. Continued sampling of frogs from the Missouri area may produce a hybrid between the two lineages that can be confirmed by genetic analysis.
Whether independent polyploidy events arising from a single species necessarily produce genetically compatible polyploid lineages has not been addressed for many polyploid taxa. This deficiency may result from the technical difficulties associated with genetic studies of such species. If the time between polyploidy events is significant, these events would likely lead to the formation of evolutionarily distinct polyploid species. However, this does not appear to be the case for the three lineages of H. versicolor, given our evidence for hybridization between an early (NW) and a later-emerging lineage (SW). We did not have sympatric population samples that would have been required to test for similar hybridization between eastern H. versicolor and either of the two other lineages. Nonetheless, given that the NW and SW lineages hybridize, it seems possible that the more closely related SW and E lineages would hybridize if and where they co-occur.
It has not been conclusively determined whether H. versicolor is an autotetraploid or an allotetraploid from a hybridization of two lineages of H. chrysoscelis (Ptacek et al. 1994). If the NW and SW lineages of H. versicolor were formed by recent allopolyploidy events of H. chrysoscelis lineages, then our estimates of divergence time could be inflated, and similarities between NW and SW lineages in nuclear allele frequencies could reflect recent (or repeated) origins rather than gene flow. However, the NW lineage cytochrome b sequence is an unambiguous outgroup to all lineages of both H. versicolor and H. chrysoscelis [this study and the more extensive survey of Ptacek et al. (1994)], suggesting that the NW lineage does not have a recent origin from H. chrysoscelis and that similarities in nuclear allele frequencies between co-occurring NW and SW lineages of H. versicolor likely result from gene flow.
The method we used to estimate the genotypes of tetraploid frogs can be applied readily to other polyploid species. With sufficient sampling of clones and appropriate diploid (or known genotype) controls to test for differential amplification of alleles, allelic assignments are highly reliable. However, the approach is costly and may force investigators to accept small sample sizes of individuals and genetic markers. Nonetheless, it may be useful to genetic and population genetic studies of polyploid species (e.g., Brochmann and Elven 1992; Marsden et al. 1987; Wendel 2000).
|
Acknowledgments |
|---|
We thank J. McGuire and J. Reiland for discussions and technical assistance. This work was supported by National Science Foundation grant DEB-9980797 and National Institutes of Health grant GM58060 subcontracted through J. Hey at Rutgers University.
|
Footnotes |
|---|
Corresponding Editor: Ross MacIntyre
Received June 11, 2001
Accepted December 31, 2001
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
-
Allard RW, Garcia P, Saenz-de-Miera LE, and de la Vega MP, 1993. Evolution of multilocus genetic structure in Avena hirtula and Avena barbata. Genetics 135:1125–1139.[Abstract]
Brochmann C and Elven R, 1992. Ecological and genetic consequences of polyploidy in arctic Draba (Brassicaceae). Evol Trends Plants 6:111–124.
Felsenstein J, 1985. Confidence limits on phylogenies: an analysis using the bootstrap. Evolution 39:783–791.[CrossRef][Web of Science]
Goudet J, 1995. FSTAT, a program for IBM PC compatibles to calculate Weir and Cockerham's estimators of F-statistics.
Goudet J, Raymond M, Demeeus T, and Rousset F, 1996. Testing differentiation in diploid populations. Genetics 144:1933–1940.[Abstract]
Johns GC and Avise JC, 1998. A comparative study of genetic distances in vertebrates from the mitochondrial cytochrome b gene. Mol Biol Evol 15:1481–1490.
Jones K, 1970. Chromosome changes in plant evolution. Taxon 19:172–179.[CrossRef]
Krenz JD, Semlitsch RD, Gerhardt HC, and Mahoney PA, 1999. Isolation and characterization of simple sequence repeat loci in the gray treefrog, Hyla chrysoscelis. Genome 42:676–680.[Medline]
Marsden JE, Schwager SJ, and May B, 1987. Single-locus inheritance in the tetraploid treefrog Hyla versicolor with an analysis of expected progeny ratios in tetraploid organisms. Genetics 116:299–311.
Martin AP and Palumbi SR, 1993. Body size, metabolic rate, generation time, and the molecular clock. Proc Natl Acad Sci USA 90:4087–4091.
Palumbi S, 1996. Nucleic acids II: the polymerase chain reaction. In: Molecular systematics (Hillis DM, Moritz C, and Mable BK, eds). Sunderland, MA: Sinauer; 205–247.
Ptacek MB, Gerhardt HC, and Sage RD, 1994. Speciation by polyploidy in treefrogs: multiple origins of the tetraploid, Hyla versicolor. Evolution 48:898–908.[CrossRef]
Romano MA, Ralin DB, Guttman SI, and Skillings JH, 1987. Parallel electromorph variation in the diploid-tetraploid gray treefrog complex. Am Nat 130:864–878.[CrossRef]
Soltis DE and Soltis PS, 1999. Polyploidy: recurrent formation and genome evolution. Trends Ecol Evol 14:348–352.[CrossRef][Medline]
Soltis PS and Soltis DE, 2000. The role of genetic and genomic attributes in the success of polyploids. Proc Natl Acad Sci USA 97:7051–7057.
Stebbins GL, 1970. Variation and evolution in plants: progress during the last twenty years. In: Essays in evolution and genetics (Hecht MK and Steere WC, eds). New York: Appleton-Century-Crofts.
Swofford DL, 1998. PAUP*: phylogenetic analysis using parsimony. Sunderland, MA: Sinauer.
Wendel JF, 2000. Genome evolution in polyploids. Plant Mol Biol 42:225–249.[CrossRef][Web of Science][Medline]
Wiley JE and Little ML, 2000. Replication banding patterns of the diploid-tetraploid treefrogs Hyla chrysoscelis and H. versicolor. Cytogenet Cell Genet 88:11–14.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  H. C. Gerhardt, C. C. Martinez-Rivera, J. J. Schwartz, V. T. Marshall, and C. G. Murphy Preferences based on spectral differences in acoustic signals in four species of treefrogs (Anura: Hylidae) J. Exp. Biol., September 1, 2007; 210(17): 2990 - 2998. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Church and D. R. Taylor Speciation and hybridization among Houstonia (Rubiaceae) species: the influence of polyploidy on reticulate evolution Am. J. Botany, August 1, 2005; 92(8): 1372 - 1380. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||