L et us assume I gave a seminar. I would tell the audience about my latest results on the population history of the pigmy shrew. My findings would be based on a stretch of DNA comprising several metabolic genes, showing no signs of genetic recombination. Armed with sequences from a large number of individuals sampled over a broad geographical area, I would make some inference on the colonization routes and times. To make life easier, I would restrict my analysis to the mutations I liked best, with nice names having been given to related sequences, rather than relying on dull mathematical quantities. As I reach one of the key conclusions of the lecture, which would go as follows: ‘It is obvious from the distribution of haplotypes Amanda, Eugenie* and Hector_2a that the Outer Hebrides were colonised about 50,000 years ago, this was followed by considerable population fluctuations, a bottleneck during the last Ice Age, a swift recovery and a dramatic recent expansion over the last 200 years andy’. Imagine that, at that climactic stage I was interrupted by someone in the audience. The impertinent would say, ‘Sir, can I just ask you whether this confidence in your conclusions may not be misplaced; your analysis is based on a single genetic marker, which comprises genes with a central role in metabolism and is thus likely to have been affected by natural selection’. An awkward silence may ensue, as I would find it difficult to dismiss this criticism easily. There are parallels between this hypothetical story and a large body of work using mitochondrial (mt)DNA polymorphisms to reconstruct the past history of innumerable species, including our own. Yet, no one in the audience seems keen to interrupt the speaker. I can only conclude that the limitations of mtDNA as a genetic marker are not fully recognized. This is arguably not because no one has previously expressed such misgivings. Excessive reliance on uniparental markers (that is, mtDNA and Y chromosome) has been criticized before (Ballard and Whitlock, 2004; Pakendorf and Stoneking, 2005; Bazin et al., 2006). Arguably, the most scathing assessment was reached by Harpending (2006) who described the entire field as a series of anecdotes ranging from the plausible and interesting to the absurd. It does not help that human phylogeography has become largely divorced from the rest of population genetics and relies on some particularly arcane jargon. As none of these earlier criticisms seem to have had much effect in stemming the tidal wave of papers using mtDNA to make fine demographic inference, I feel compelled to summarize once again what the problems are and why they cannot be swept under the carpet indefinitely. Unlinked genetic markers trickle down through species’ pedigrees independently from one another. Owing to the high stochasticity of population demography, some genetic markers will be reflective of the populations’ history and some would not. mtDNA and the Y chromosome do not recombine and therefore represent only a single realization of the many possible outcomes within a given demographic history, irrespective of the number of polymorphic sites typed (Ballard and Whitlock, 2004). As a consequence, the phylogenetic tree of uniparental markers may or may not be informative on the demography of the populations studied. Although it has been said forcefully before that gene trees should not be equated to species or population trees (Nichols, 2001), it seems that this subtle yet important distinction is rarely made. A frankly baffling trend from a population genetics perspective is the apparent increase of papers considering only a single haplogroup, as the problems with demographic stochasticity will be exacerbated even further. This demographic variance is not a major issue when one has access to a reasonable number of autosomal markers, which can be treated as replicates of the same process and averaged over loci to make inference on the population history. A crucial assumption is that the genetic markers used to make inference on the past history of populations are evolving neutrally; that a non-recombining stretch of DNA comprising 37 genes should be neutral seems a bold hypothesis. Detecting natural selection in non-recombining DNA is difficult. Despite this, there is evidence for natural selection on mtDNA in various taxa (for example, Ballard et al., 2007; Fontanillas et al., 2005; Oliveira et al., 2008), with temperature being often invoked as the likely selective force. The situation in humans is still far from clear. There have been claims based on ratios of synonymous versus nonsynonymous mutations (dN/dS ratios), and to a lesser extent, the evolutionary persistence of mutations that human mtDNA may have been affected by climate (Torroni et al., 2001; Mishmar et al., 2003; Ruiz-Pesini et al., 2004). This has been refuted by other studies, which concluded that human mtDNA sequence variation has not been significantly shaped by natural selection (Elson et al., 2004; Kivisild et al., 2006; Amo and Brand, 2007; Ingman and Gyllensten, 2007). However, all these results (both for and against selection) are questionable as dN/dS ratios are generally inadequate tests for natural selection when working over limited evolutionary time scales at a withinpopulation level (Rocha et al., 2006; Kryazhimskiy and Plotkin, 2008). Interestingly, the most comprehensive of the studies concluding that human mtDNA was evolving neutrally did actually highlight the single-nucleotide polymorphism at position 10 398 as a target for natural selection (Kivisild et al., 2006). This single-nucleotide polymorphism was also identified again in study, which showed in addition that mitochondrial diversity correlated with the temperature to which human populations were exposed (Balloux et al., 2009). Finally, the same single-nucleotide polymorphism has been shown to affect mitochondrial matrix pH and mitochondrial calcium dynamics (Kazuno et al., 2006). There is even a sensible biological explanation why temperature may affect mtDNA sequence variation (Coskun et al., 2003; Mishmar et al., 2003). As shivering frantically to keep warm is not such a pleasant prospect, we have to rely largely on the heat generated by the oxidative phosphorylation (OXPHOS) cycle, which comprises 13 genes encoded by the mitochondrial genome. The primary Heredity (2010) 104, 419–420 & 2010 Macmillan Publishers Limited All rights reserved 0018-067X/10 $32.00

中文翻译：

---

线粒体DNA树果实中的蠕虫

让我们假设我举办了一个研讨会。我会告诉观众我对侏儒种群历史的最新研究结果。我的发现将基于一段包含几个代谢基因的 DNA，没有显示出基因重组的迹象。有了在广泛地理区域采样的大量个体的序列，我将对殖民路线和时间做出一些推断。为了让生活更轻松，我会将我的分析限制在我最喜欢的突变上，并为相关序列命名，而不是依赖于枯燥的数学量。当我得出讲座的一个关键结论时，该结论如下：“从单倍型 Amanda、Eugenie* 和 Hector_2a 的分布中可以明显看出，外赫布里底群岛大约在 50,000 年前被殖民，紧随其后的是相当大的人口波动、上一个冰河时代的瓶颈、迅速恢复和最近 200 年的急剧扩张。想象一下，在那个高潮阶段，我被观众中的某个人打断了。无礼的人会说，‘先生，我能不能问问你对你的结论的这种信心是否可能是错误的？您的分析基于单个遗传标记，其中包含在新陈代谢中起核心作用的基因，因此很可能受到自然选择的影响”。可能会出现尴尬的沉默，因为我发现很难轻易驳回这种批评。这个假设的故事与使用线粒体 (mt)DNA 多态性重建无数物种的过去历史的大量工作之间存在相似之处，包括我们自己的物种。然而，听众中似乎没有人愿意打断演讲者。我只能得出结论，尚未完全认识到 mtDNA 作为遗传标记的局限性。这可以说不是因为之前没有人表达过这种疑虑。过分依赖单亲标记（即 mtDNA 和 Y 染色体）之前曾受到批评（Ballard 和 Whitlock，2004；Pakendorf 和 Stoneking，2005；Bazin 等，2006）。可以说，最严厉的评估是由 Harpending (2006) 提出的，他将整个领域描述为一系列从合理有趣到荒谬的轶事。人类系统地理学在很大程度上脱离了种群遗传学的其余部分，并依赖于一些特别神秘的行话，这无济于事。由于这些早期的批评似乎都没有对阻止使用 mtDNA 进行精细人口统计推断的论文浪潮产生太大影响，我觉得有必要再次总结问题是什么以及为什么不能无限期地将它们扫到地毯下。互不关联的遗传标记在物种的谱系中相互独立。由于人口统计学的高度随机性，一些遗传标记会反映人口的历史，而有些则不会。mtDNA 和 Y 染色体不会重组，因此仅代表给定人口统计历史中许多可能结果的单一实现，而与类型的多态位点数量无关（Ballard 和 Whitlock，2004）。作为结果，单亲标记的系统发育树可能会或可能不会提供所研究种群的人口统计信息。尽管之前曾强硬地说基因树不应等同于物种或种群树（Nichols，2001），但似乎很少有人做出这种微妙而重要的区别。从人口遗传学的角度来看，一个坦率的令人困惑的趋势是仅考虑单个单倍群的论文明显增加，因为人口随机性问题将进一步加剧。当人们可以获得合理数量的常染色体标记时，这种人口统计学差异不是主要问题，这些标记可以被视为同一过程的重复，并在基因座上取平均值以推断人口历史。一个关键的假设是，用于推断种群过去历史的遗传标记是中性进化的。包含 37 个基因的非重组 DNA 片段应该是中性的，这似乎是一个大胆的假设。在非重组 DNA 中检测自然选择是困难的。尽管如此，有证据表明在各种分类群中对 mtDNA 进行自然选择（例如，Ballard 等人，2007 年；Fontanillas 等人，2005 年；Oliveira 等人，2008 年），温度通常被认为是可能的选择力. 人类的情况还远未明朗。有人声称基于同义突变与非同义突变的比率（dN/dS 比率），以及在较小程度上，人类 mtDNA 可能受到气候影响的突变的进化持续性（Torroni 等人，2001 年；Mishmar 等人., 2003; Ruiz-Pesini 等。, 2004)。其他研究驳斥了这一点，这些研究得出的结论是，人类 mtDNA 序列变异并未受到自然选择的显着影响（Elson 等，2004；Kivsild 等，2006；Amo 和 Brand，2007；Ingman 和 Gyllensten，2007） . 然而，所有这些结果（支持和反对选择）都是有问题的，因为当在种群内水平有限的进化时间尺度上工作时，dN/dS 比率通常不足以测试自然选择（Rocha 等人，2006 年；Kryazhimskiy 和 Plotkin，2008 年） ）。有趣的是，最全面的研究得出的结论是人类 mtDNA 是中性进化的，实际上强调了 10 398 位的单核苷酸多态性作为自然选择的目标（Kivsild 等，2006）。这种单核苷酸多态性在研究中也再次被鉴定出来，此外，这表明线粒体多样性与人类暴露的温度相关（Balloux 等，2009）。最后，已显示相同的单核苷酸多态性影响线粒体基质 pH 值和线粒体钙动力学（Kazuno 等，2006）。甚至有一个合理的生物学解释为什么温度可能会影响 mtDNA 序列变异（Coskun 等人，2003 年；Mishmar 等人，2003 年）。由于疯狂地颤抖着取暖并不是一个令人愉快的前景，我们必须在很大程度上依赖氧化磷酸化 (OXPHOS) 循环产生的热量，该循环由线粒体基因组编码的 13 个基因组成。The primary Heredity (2010) 104, 419–420 & 2010 Macmillan Publishers Limited 版权所有 0018-067X/10 $32.00 最后，相同的单核苷酸多态性已被证明会影响线粒体基质 pH 值和线粒体钙动力学（Kazuno 等，2006）。甚至有一个合理的生物学解释为什么温度可能会影响 mtDNA 序列变异（Coskun 等人，2003 年；Mishmar 等人，2003 年）。由于疯狂地颤抖着取暖并不是一个令人愉快的前景，我们必须在很大程度上依赖氧化磷酸化 (OXPHOS) 循环产生的热量，该循环由线粒体基因组编码的 13 个基因组成。The primary Heredity (2010) 104, 419–420 & 2010 Macmillan Publishers Limited 版权所有 0018-067X/10 $32.00 最后，已显示相同的单核苷酸多态性影响线粒体基质 pH 值和线粒体钙动力学（Kazuno 等，2006）。甚至有一个合理的生物学解释为什么温度可能会影响 mtDNA 序列变异（Coskun 等人，2003 年；Mishmar 等人，2003 年）。由于疯狂地颤抖着取暖并不是一个令人愉快的前景，我们必须在很大程度上依赖氧化磷酸化 (OXPHOS) 循环产生的热量，该循环由线粒体基因组编码的 13 个基因组成。The primary Heredity (2010) 104, 419–420 & 2010 Macmillan Publishers Limited 版权所有 0018-067X/10 $32.00 甚至有一个合理的生物学解释为什么温度可能会影响 mtDNA 序列变异（Coskun 等人，2003 年；Mishmar 等人，2003 年）。由于疯狂地颤抖以取暖并不是一个令人愉快的前景，我们必须在很大程度上依赖氧化磷酸化 (OXPHOS) 循环产生的热量，该循环由线粒体基因组编码的 13 个基因组成。The primary Heredity (2010) 104, 419–420 & 2010 Macmillan Publishers Limited 版权所有 0018-067X/10 $32.00 甚至有一个合理的生物学解释为什么温度可能会影响 mtDNA 序列变异（Coskun 等人，2003 年；Mishmar 等人，2003 年）。由于疯狂地颤抖着取暖并不是一个令人愉快的前景，我们必须在很大程度上依赖氧化磷酸化 (OXPHOS) 循环产生的热量，该循环由线粒体基因组编码的 13 个基因组成。The primary Heredity (2010) 104, 419–420 & 2010 Macmillan Publishers Limited 版权所有 0018-067X/10 $32.00