E-Letter responses to:

reports: Eric Bazin, Sylvain Glémin, and Nicolas Galtier Population Size Does Not Influence Mitochondrial Genetic Diversity in Animals Science 2006; 312: 570-572 [Abstract] [Full text] [PDF]

E-Letters: Submit a response to this article

Published E-Letter responses:

Response to E-Letter by Foltz and Rand

E. Bazin, S. Glémin, N. Galtier   (29 December 2006)

Mitochondrial DNA Diversity and Population Size

David W. Foltz, David Rand   (29 December 2006)

Response to E-Letter by Hickey et al.

E. Bazin, S. Glémin, N. Galtier   (29 December 2006)

Mitochondrial Nucleotide Diversity and Population Size

Donal A. Hickey, Mehrdad Hajibabaei, Gregory A. C. Singer   (29 December 2006)

Response to E-Letter by Foltz and Rand 29 December 2006
E. Bazin, galtier@univ-montp2.fr CNRS UMR 5171–Génome, Populations, Interactions, Adaptation–Université Montpellier 2, S. Glémin, N. Galtier Respond to this E-Letter: Re: Response to E-Letter by Foltz and Rand	Response to Foltz and Rand Foltz and Rand argue that the lack of relationship between mtDNA diversity and population size we report is due to the irrelevance of the population size indicators we use and to the fact that distantly related species are compared. They propose to focus on closely related species differing by only one ecological variable and list a number of such studies in which mtDNA diversity behaved as expected. We disagree with this argument for the reasons given below: 1) If the population size indicators or the time scale we use were inappropriate and/or uncorrelated to effective size, why would nuclear DNA behave as intuitively expected? 2) Every species has an effective population size. These are numbers that can be compared, whether species are closely or distantly related. The average populations sizes of vertebrate and invertebrate species are probably very different, as suggested by the intuition and confirmed by nuclear data. This should be reflected by mtDNA diversity if mtDNA was a neutral marker, irrespective of which traits control population size, and of which traits are comparable between species. 3) We do not mean to argue that mtDNA is misleading for every animal species/taxon. In species that did not undergo a recent selective sweep, population size probably matters, as illustrated by the references provided by Foltz and Rand. Our work suggests that such cases are more exceptions than rule, however, especially given the strong publication bias. Foltz and Rand list less than 10 studies in which mtDNA diversity behaves as expected, when several research groups have been working on this topic, and thousands of data sets are available. How many case studies in which mtDNA does not fit the expectations have remained unpublished? E. Bazin, S. Glémin, N. Galtier CNRS UMR 5171–Génome, Populations, Interactions, Adaptation–Université Montpellier 2, 34095 Montpellier Cedex 5, France.
Mitochondrial DNA Diversity and Population Size 29 December 2006
David W. Foltz Department of Biological Sciences, Louisiana State University, David Rand Respond to this E-Letter: Re: Mitochondrial DNA Diversity and Population Size	The Report by E. Bazin et al. that mitochondrial DNA (mtDNA) diversity is uncorrelated with either nuclear DNA diversity or allozymic heterozygosity (“Population size does not influence mitochondrial genetic diversity in animals,” 28 Apr., p. 570) raises questions about the suitability of mtDNA sequences as markers for conservation and population biology studies. We suggest that their ecological comparisons (such as freshwater versus marine fish, invertebrates versus vertebrates, or terrestrial versus marine mollusks) are probably too crude to yield useful insights into factors influencing diversity of the mitochondrial and nuclear genomes. Average life span, fecundity, and other life history traits influencing sequence diversity are orders-of-magnitude more variable among invertebrate than vertebrate species. As a group, invertebrates also have a greater diversity in modes of reproduction, including self-fertilization, parthenogenesis, and other forms of clonal reproduction. In contrast to Bazin et al.’s phylum-level analyses, prior studies comparing closely related animals species that differ in one or more life history attributes have found differences in diversity for both mtDNA and nuclear genes that appear to be correlated with effective population size. In particular, differences in levels of nonsynonymous substitutions for mitochondrial protein-coding genes between island versus mainland populations (1, 2), endosymbiotic versus free-living bacteria and fungi (3, 4), geographically subdivided versus nonsubdivided rodents (5), and invertebrate species with nonpelagic larvae versus those with pelagic larvae (6) are all consistent with the suggestion that the former lineages have lower effective population sizes than the latter. Other possible explanations for the findings, such as differences in mutation rates or selective constraints (7), are less plausible, although some ecological differences such as endosymbiosis and island colonization could simultaneously cause low effective population sizes and relaxed selective constraints, due to ecological release. Similar patterns are seen for nuclear genes as well (2, 8). These results suggest that, although mtDNA diversity may be “essentially unpredictable” at higher taxonomic levels due to “genetic draft” or other selective forces, as proposed by Bazin et al., these forces do not seriously impede the use of mtDNA sequences for population-level studies and other comparisons at lower taxonomic levels or appropriately matched sister taxon pairs. To support the claim of repeated selective sweeps in mtDNA, there are several additional analyses that need to be considered. Specifically, differences between nuclear and mtDNA data sets in vertebrates and invertebrates most likely vary in (i) sample sizes, (ii) mtDNA gene used for analyses, (iii) nucleotide site frequencies, and (iv) evidence for low polymorphism in mtDNAs. These issues need to be addressed as follows: (i) larger sample sizes lead to reductions in neutrality index (NI) (9); (ii) different mitochondrial genes tend to be used in vertebrates and invertebrates (Cytb versus CO1), which alters ratios of fixed replacement and silent sites; (iii) selective sweeps should leave a signature of negative Tajima’s D that differs for mtDNA and nuclear genes, and for silent and replacement sites (10); and (iv) why is there so little evidence for lack of variation in mtDNA if sweeps are so common? Although we applaud the authors for pursuing a critical test of the neutral assumptions of mtDNA, something we have been advocating for more than a decade (11), the lack of information on these important factors leaves us wondering whether more critical analyses of additional neutrality tests would reveal the complete picture of selection on mtDNA. David W. Foltz, Professor Department of Biological Sciences Louisiana State University Baton Rouge, LA 70803-1715, USA David Rand Professor of Biology Department of Ecology and Evolutionary Biology Box G-W, 69 Brown Street Brown University Providence, RI 02912, USA References and Notes 1. K. P. Johnson, J. Seger, Mol. Biol. Evol. 18, 874 (2001). 2. M. Woolfit, L. Bromham, Proc. Roy. Soc. London B 272, 2277 (2005). 3. D. J. Funk, J. J. Wernegreen, N. A. Moran, Genetics 157, 477 (2001). 4. M. Woolfit, L. Bromham, Mol. Biol. Evol. 20, 1545 (2003). 5. T. A. Spradling, M. S. Hafner, J. W. Demastes, J. Mammal. 82, 65 (2001). 6. D. W. Foltz, J. Mol. Evol. 57, 607 (2003). 7. J. Leebens-Mack, C. dePamphilis, Mol. Biol. Evol. 19, 1292 (2002) 8. A. Eyre-Walker, P. D. Keightley, N. G. C. Smith, D. Gaffney, Mol. Biol. Evol. 19, 2142 (2002). 9. D. M. Weinreich, D. M. Rand, Genetics 156, 385 (2000). 10. D. M. Rand, L. M. Kann, Mol. Biol. Evol. 13, 735 (1996). 11. J. W. O. Ballard, D. M. Rand, Annu. Rev. Ecol. Evol. Syst. 36, 621 (2005).
Response to E-Letter by Hickey et al. 29 December 2006
E. Bazin CNRS UMR 5171–Génome, Populations, Interactions, Adaptation–Université Montpellier 2, S. Glémin, N. Galtier Respond to this E-Letter: Re: Response to E-Letter by Hickey et al.	Response to Hickey et al. We interpreted the lack of relationship between species abundance and within-species mitochondrial DNA (mtDNA) diversity as a consequence of recurrent adaptive evolution in mitochondrial genomes. Hickey et al. criticize this explanation on two grounds. They argue that the higher level of nucleotide diversity in mtDNA, as compared with nuclear DNA, is not consistent with the hypothesis of frequent mitochondrial selective sweeps. This argument neglects the role of mutation rate, another important determinant of genetic diversity. The mtDNA per-base mutation rate is known to be orders of magnitude higher than the nuclear one, so a rapid recovery of mtDNA diversity is expected posterior to bottlenecks or selective sweeps. In the absence of selection, we would expect a much higher ratio of average diversity between mtDNA and nuclear DNA than observed, knowing the difference in mutation rate. Hickey et al. then invoke heteroplasmy (the existence of distinct mitochondrial haplotypes within a single individual) and the high number of mtDNA molecules per cell as a major factor controlling mtDNA diversity in animals. There are two reasons why we think that this objection does not apply. First, site heteroplasmy in animal mtDNA is scarce (1). Within-individual variability, when detected, is typically much lower than between-individual diversity, probably because of the germ-line bottleneck experienced at every generation. Heteroplasmy, furthermore, has been mostly detected in the hypervariable mitochondrial control region, whereas our study focuses on coding regions. Second, assuming that heteroplasmy is high, and even assuming some level of paternal mtDNA transmission, the mitochondrial population structure could be considered as a metapopulation in which female individuals are demes, as suggested by Hickey et al. But metapopulation models predict that species diversity increases with the number of demes (2). Even under this extreme and implausible hypothesis, therefore, we would expect a relationship between female population size and mtDNA diversity. E. Bazin, S. Glémin, N. Galtier CNRS UMR 5171–Génome, Populations, Interactions, Adaptation–Université Montpellier 2, 34095 Montpellier Cedex 5, France. References 1. C. M. Barr, M. Neiman, D. M. Taylor, New Phytol. 168, 39 (2005). 2. N. H. Barton, M. C. Whitlock, in Metapopulation Biology. Ecology, Genetics and Evolution, I. A. Hanski, M. E. Gilpin, Eds. (Academic Press, San Diego, CA, 1997), pp. 183–210.
Mitochondrial Nucleotide Diversity and Population Size 29 December 2006
Donal A. Hickey Department of Biology, Concordia University, Mehrdad Hajibabaei, Gregory A. C. Singer Respond to this E-Letter: Re: Mitochondrial Nucleotide Diversity and Population Size	In their Report “Population size does not influence mitochondrial genetic diversity in animals” (28 Apr., p. 570), E. Bazin et al. show that the nucleotide diversity of animal mitochondrial genomes is independent of population size. The authors propose that these sequences are subject to repeated selective sweeps that propagate adaptive mutations within the population while, at the same time, causing fixation of neutral variants at other sites. Although the data are clear cut, the adaptive explanation appears to be at odds with the data. Given that selective sweeps tend to “clean out” the standing genetic variation from the population, we would expect to see not only the observed lack of correlation between nucleotide diversity and estimated population size, but also a relatively low average level of diversity among the sequences. This second prediction is not met. This is especially notable for the mammals, where the levels of mitochondrial diversity are more than tenfold higher than among the nuclear sequences. The high levels of standing genetic variation among mitochondrial genomes argue against frequent selective sweeps. The observed lack of correlation between mitochondrial nucleotide diversity and population size may be due to the fact that there are multiple copies of the mitochondrial genome within each eukaryotic cell. For instance, it has been shown that the human oocyte contains approximately 100,000 mitochondrial DNA copies (1) and that the frequency of sequence variants within heteroplasmic individuals can fluctuate from generation to generation (2). Therefore, a population of eukaryotic individuals could be considered as a large collection of separate mitochondrial “populations.” Looked at in this way, the critical factor may the number of mitochondrial genomes per organelle and per cell rather than the number of individuals in the population. Donal A. Hickey,1 Mehrdad Hajibabaei,2 Gregory A. C. Singer3 1Department of Biology, Concordia University, Montreal, QC H4B 1R6, Canada. 2Department of Integrative Biology, University of Guelph, Geulph, ON N1G 2W1, Canada. 3Ohio State University, Columbus, OH 43210, USA. References 1. X. Chen et al., Am. J. Hum. Genet. 57, 239 (1995). 2. S. Lutz et al., Int. J. Legal Med. 113, 155 (2000).