Limits...
Analysis of Canis mitochondrial DNA demonstrates high concordance between the control region and ATPase genes

Rutledge LY, Patterson BR, White BN - BMC Evol. Biol. (2010)

Bottom Line: We found high concordance across analyses between the mtDNA regions studied.Neutrality tests on both genes were indicative of the population expansion of coyotes across eastern North America, and dN/dS ratios suggest a possible role for purifying selection in the evolution of North American lineages. dN/dS ratios were higher in European evolved lineages from northern climates compared to North American evolved lineages from temperate regions, but these differences were not statistically significant.Increased sampling and development of alternative analytical tools will be necessary to disentangle demographic history from processes of natural selection.

Affiliation: Environmental and Life Sciences Graduate Program, DNA Building, Trent University, 2140 East Bank Drive, Peterborough, ON, K9J 7B8, Canada. lrutledge@nrdpfc.ca

ABSTRACT

Background: Phylogenetic studies of wild Canis species have relied heavily on the mitochondrial DNA control region (mtDNA CR) to infer species relationships and evolutionary lineages. Previous analyses of the CR provided evidence for a North American evolved eastern wolf (C. lycaon), that is more closely related to red wolves (C. rufus) and coyotes (C. latrans) than grey wolves (C. lupus). Eastern wolf origins, however, continue to be questioned. Therefore, we analyzed mtDNA from 89 wolves and coyotes across North America and Eurasia at 347 base pairs (bp) of the CR and 1067 bp that included the ATPase6 and ATPase8 genes. Phylogenies and divergence estimates were used to clarify the evolutionary history of eastern wolves, and regional comparisons of nonsynonomous to synonomous substitutions (dN/dS) at the ATPase6 and ATPase8 genes were used to elucidate the potential role of selection in shaping mtDNA geographic distribution.

Results: We found high concordance across analyses between the mtDNA regions studied. Both had a high percentage of variable sites (CR = 14.6%; ATP = 9.7%) and both phylogenies clustered eastern wolf haplotypes monophyletically within a North American evolved lineage apart from coyotes. Divergence estimates suggest the putative red wolf sequence is more closely related to coyotes (DxyCR = 0.01982 +/- 0.00494 SD; DxyATP = 0.00332 +/- 0.00097 SD) than the eastern wolf sequences (DxyCR = 0.03047 +/- 0.00664 SD; DxyATP = 0.00931 +/- 0.00205 SD). Neutrality tests on both genes were indicative of the population expansion of coyotes across eastern North America, and dN/dS ratios suggest a possible role for purifying selection in the evolution of North American lineages. dN/dS ratios were higher in European evolved lineages from northern climates compared to North American evolved lineages from temperate regions, but these differences were not statistically significant.

Conclusions: These results demonstrate high concordance between coding and non-coding regions of mtDNA, and provide further evidence that the eastern wolf possessed distinct mtDNA lineages prior to recent coyote introgression. Purifying selection may have influenced North American evolved Canis lineages, but detection of adaptive selection in response to climate is limited by the power of current statistical tests. Increased sampling and development of alternative analytical tools will be necessary to disentangle demographic history from processes of natural selection.

Cladograms of Canis mtDNA CR and ATPase sequences. Cladograms of Canis sequences from Bayesian analysis in BEAST of a) 347 bp of the mitochondrial DNA (mtDNA) control region and b) 1067 bp from the mtDNA ATPase region. fam is a Husky dog sample from Sweden; * is a wolf from Saudi Arabia; ruf is the red wolf sequence, NW represents New World evolved clades, OW represents Old World evolved clades, NA in b) identifies North American grey wolf samples. Node labels show posterior probabilities rounded to the nearest hundredth. For each genetic region the eastern wolf clade is shown in red (NWCR1ew & NWATP1ew), coyote clades I and II are shown in yellow (NWCR2coyI, NWATP2coyI) and orange (NWCR3coyII and NWATP3coyII), and Old World (OW) clades from North America (OWCR4gwNA) and Eurasia (OWCR5gwEU, OWATP4gwNAEU) are shown in blue. Insets show radial view of tree.
© Copyright Policy

Figure 2: Cladograms of Canis mtDNA CR and ATPase sequences. Cladograms of Canis sequences from Bayesian analysis in BEAST of a) 347 bp of the mitochondrial DNA (mtDNA) control region and b) 1067 bp from the mtDNA ATPase region. fam is a Husky dog sample from Sweden; * is a wolf from Saudi Arabia; ruf is the red wolf sequence, NW represents New World evolved clades, OW represents Old World evolved clades, NA in b) identifies North American grey wolf samples. Node labels show posterior probabilities rounded to the nearest hundredth. For each genetic region the eastern wolf clade is shown in red (NWCR1ew & NWATP1ew), coyote clades I and II are shown in yellow (NWCR2coyI, NWATP2coyI) and orange (NWCR3coyII and NWATP3coyII), and Old World (OW) clades from North America (OWCR4gwNA) and Eurasia (OWCR5gwEU, OWATP4gwNAEU) are shown in blue. Insets show radial view of tree.

Mentions: There was a wide geographic distribution of NW coyote-like ATPase haplotypes (Figure 1). Cladograms from each genetic region had very similar topologies, which is indicative of the haplotype association of the linked regions. Eastern wolf sequences (Ccr13, Ccr12; Catp13, Catp16) clustered monophyletically with high (> 0.9) posterior probability within the NW clade, but apart from coyotes, whereas the putative red wolf sequence clustered among coyote sequences within the NWCR2coyI/NWATP2coyI clade (Figures 2a, b). Similar clustering of the red wolf control region sequence among coyote sequences has been reported [22], but the haplotype is attributed to red wolves because it is not known to occur in non-hybridizing coyotes from western North America. A similar argument has been made for coyote-like eastern wolf haplotypes [27,32]. These geographic distinctions of coyote-like sequences in eastern and red wolves combined with evidence of coyote-like sequences in eastern wolves prior to European settlement in North America [25] provide evidence for incomplete lineage sorting within the NW lineages, although ancient (~11,000 years ago) hybridization during the Wisconsin glaciation is difficult to rule out. This, combined with extensive hybridization in eastern Canis populations [23,24,26,27,33,34], makes species designations of wild Canis difficult when based on phylogenetic inference from mtDNA alone. For example, a conspecific nature of eastern wolves and red wolves is suggested by nuclear data [19], but is not demonstrated by analysis of mtDNA. These differences are not unexpected and do not undermine the conspecific nature of eastern wolves and red wolves because gene trees based on mtDNA are not necessarily indicative of specific species relationships, and discordance is often found when comparing mtDNA and nuclear genetic signatures [35]. Given the recent divergence of NW lineages (see below) these issues are not unexpected [35]. New approaches to phylogenetic analysis that utilize multiple loci, including nuclear genes, may help reconcile inferred Canis species relationships [35,36], although extensive hybridization will likely continue to plague contemporary species designations based on nuclear markers. Regardless, the distinction of the two eastern wolf haplotypes shown in both the control region and ATPase region is indisputable, providing clear evidence for the presence of a North American evolved wolf lineage, distinct from coyotes and grey wolves.

View Similar Images In: Results  - Collection
View Article: Pubmed Central - HTML -  PubMed
Show All Figures - Show MeSH
getmorefigures.php?pmc=2927920&rFormat=json&query=null&req=5
Analysis of Canis mitochondrial DNA demonstrates high concordance between the control region and ATPase genes

Rutledge LY, Patterson BR, White BN - BMC Evol. Biol. (2010)

Cladograms of Canis mtDNA CR and ATPase sequences. Cladograms of Canis sequences from Bayesian analysis in BEAST of a) 347 bp of the mitochondrial DNA (mtDNA) control region and b) 1067 bp from the mtDNA ATPase region. fam is a Husky dog sample from Sweden; * is a wolf from Saudi Arabia; ruf is the red wolf sequence, NW represents New World evolved clades, OW represents Old World evolved clades, NA in b) identifies North American grey wolf samples. Node labels show posterior probabilities rounded to the nearest hundredth. For each genetic region the eastern wolf clade is shown in red (NWCR1ew & NWATP1ew), coyote clades I and II are shown in yellow (NWCR2coyI, NWATP2coyI) and orange (NWCR3coyII and NWATP3coyII), and Old World (OW) clades from North America (OWCR4gwNA) and Eurasia (OWCR5gwEU, OWATP4gwNAEU) are shown in blue. Insets show radial view of tree.
© Copyright Policy
Figure 2: Cladograms of Canis mtDNA CR and ATPase sequences. Cladograms of Canis sequences from Bayesian analysis in BEAST of a) 347 bp of the mitochondrial DNA (mtDNA) control region and b) 1067 bp from the mtDNA ATPase region. fam is a Husky dog sample from Sweden; * is a wolf from Saudi Arabia; ruf is the red wolf sequence, NW represents New World evolved clades, OW represents Old World evolved clades, NA in b) identifies North American grey wolf samples. Node labels show posterior probabilities rounded to the nearest hundredth. For each genetic region the eastern wolf clade is shown in red (NWCR1ew & NWATP1ew), coyote clades I and II are shown in yellow (NWCR2coyI, NWATP2coyI) and orange (NWCR3coyII and NWATP3coyII), and Old World (OW) clades from North America (OWCR4gwNA) and Eurasia (OWCR5gwEU, OWATP4gwNAEU) are shown in blue. Insets show radial view of tree.
Mentions: There was a wide geographic distribution of NW coyote-like ATPase haplotypes (Figure 1). Cladograms from each genetic region had very similar topologies, which is indicative of the haplotype association of the linked regions. Eastern wolf sequences (Ccr13, Ccr12; Catp13, Catp16) clustered monophyletically with high (> 0.9) posterior probability within the NW clade, but apart from coyotes, whereas the putative red wolf sequence clustered among coyote sequences within the NWCR2coyI/NWATP2coyI clade (Figures 2a, b). Similar clustering of the red wolf control region sequence among coyote sequences has been reported [22], but the haplotype is attributed to red wolves because it is not known to occur in non-hybridizing coyotes from western North America. A similar argument has been made for coyote-like eastern wolf haplotypes [27,32]. These geographic distinctions of coyote-like sequences in eastern and red wolves combined with evidence of coyote-like sequences in eastern wolves prior to European settlement in North America [25] provide evidence for incomplete lineage sorting within the NW lineages, although ancient (~11,000 years ago) hybridization during the Wisconsin glaciation is difficult to rule out. This, combined with extensive hybridization in eastern Canis populations [23,24,26,27,33,34], makes species designations of wild Canis difficult when based on phylogenetic inference from mtDNA alone. For example, a conspecific nature of eastern wolves and red wolves is suggested by nuclear data [19], but is not demonstrated by analysis of mtDNA. These differences are not unexpected and do not undermine the conspecific nature of eastern wolves and red wolves because gene trees based on mtDNA are not necessarily indicative of specific species relationships, and discordance is often found when comparing mtDNA and nuclear genetic signatures [35]. Given the recent divergence of NW lineages (see below) these issues are not unexpected [35]. New approaches to phylogenetic analysis that utilize multiple loci, including nuclear genes, may help reconcile inferred Canis species relationships [35,36], although extensive hybridization will likely continue to plague contemporary species designations based on nuclear markers. Regardless, the distinction of the two eastern wolf haplotypes shown in both the control region and ATPase region is indisputable, providing clear evidence for the presence of a North American evolved wolf lineage, distinct from coyotes and grey wolves.

Bottom Line: We found high concordance across analyses between the mtDNA regions studied.Neutrality tests on both genes were indicative of the population expansion of coyotes across eastern North America, and dN/dS ratios suggest a possible role for purifying selection in the evolution of North American lineages. dN/dS ratios were higher in European evolved lineages from northern climates compared to North American evolved lineages from temperate regions, but these differences were not statistically significant.Increased sampling and development of alternative analytical tools will be necessary to disentangle demographic history from processes of natural selection.

Affiliation: Environmental and Life Sciences Graduate Program, DNA Building, Trent University, 2140 East Bank Drive, Peterborough, ON, K9J 7B8, Canada. lrutledge@nrdpfc.ca

ABSTRACT

Background:

Background: Phylogenetic studies of wild Canis species have relied heavily on the mitochondrial DNA control region (mtDNA CR) to infer species relationships and evolutionary lineages. Previous analyses of the CR provided evidence for a North American evolved eastern wolf (C. lycaon), that is more closely related to red wolves (C. rufus) and coyotes (C. latrans) than grey wolves (C. lupus). Eastern wolf origins, however, continue to be questioned. Therefore, we analyzed mtDNA from 89 wolves and coyotes across North America and Eurasia at 347 base pairs (bp) of the CR and 1067 bp that included the ATPase6 and ATPase8 genes. Phylogenies and divergence estimates were used to clarify the evolutionary history of eastern wolves, and regional comparisons of nonsynonomous to synonomous substitutions (dN/dS) at the ATPase6 and ATPase8 genes were used to elucidate the potential role of selection in shaping mtDNA geographic distribution.

Results: We found high concordance across analyses between the mtDNA regions studied. Both had a high percentage of variable sites (CR = 14.6%; ATP = 9.7%) and both phylogenies clustered eastern wolf haplotypes monophyletically within a North American evolved lineage apart from coyotes. Divergence estimates suggest the putative red wolf sequence is more closely related to coyotes (DxyCR = 0.01982 +/- 0.00494 SD; DxyATP = 0.00332 +/- 0.00097 SD) than the eastern wolf sequences (DxyCR = 0.03047 +/- 0.00664 SD; DxyATP = 0.00931 +/- 0.00205 SD). Neutrality tests on both genes were indicative of the population expansion of coyotes across eastern North America, and dN/dS ratios suggest a possible role for purifying selection in the evolution of North American lineages. dN/dS ratios were higher in European evolved lineages from northern climates compared to North American evolved lineages from temperate regions, but these differences were not statistically significant.

Conclusions: These results demonstrate high concordance between coding and non-coding regions of mtDNA, and provide further evidence that the eastern wolf possessed distinct mtDNA lineages prior to recent coyote introgression. Purifying selection may have influenced North American evolved Canis lineages, but detection of adaptive selection in response to climate is limited by the power of current statistical tests. Increased sampling and development of alternative analytical tools will be necessary to disentangle demographic history from processes of natural selection.

View Similar Images In: Results  - Collection
View Article: Pubmed Central - HTML -  PubMed
Show All Figures - Show MeSH
getmorefigures.php?pmc=2927920&rFormat=json&query=null&req=5