Hazkani-Covo Einat, Wool David, Graur Dan
Department of Zoology, George S. Wise Faculty of Life Sciences, Tel Aviv University, Ramat Aviv 69978, Israel.
J Exp Zool B Mol Dev Evol. 2005 Mar 15;304(2):150-8. doi: 10.1002/jez.b.21033.
In 1828, Karl von Baer proposed a set of four evolutionary "laws" pertaining to embryological development. According to von Baer's third law, young embryos from different species are relatively undifferentiated and resemble one another but as development proceeds, distinguishing features of the species begin to appear and embryos of different species progressively diverge from one another. An expansion of this law, called "the hourglass model," has been proposed independently by Denis Duboule and Rudolf Raff in the 1990s. According to the hourglass model, ontogeny is characterized by a starting point at which different taxa differ markedly from one another, followed by a stage of reduced intertaxonomic variability (the phylotypic stage), and ending in a von-Baer-like progressive divergence among the taxa. A possible "translation" of the hourglass model into molecular terminology would suggest that orthologs expressed in stages described by the tapered part of the hourglass should resemble one another more than orthologs expressed in the expansive parts that precede or succeed the phylotypic stage. We tested this hypothesis using 1,585 mouse genes expressed during 26 embryonic stages, and their human orthologs. Evolutionary divergence was estimated at different embryonic stages by calculating pairwise distances between corresponding orthologous proteins from mouse and human. Two independent datasets were used. One dataset contained genes that are expressed solely in a single developmental stage; the second was made of genes expressed at different developmental stages. In the second dataset the genes were classified according to their earliest stage of expression. We fitted second order polynomials to the two datasets. The two polynomials displayed minima as expected from the hourglass model. The molecular results suggest, albeit weakly, that a phylotypic stage (or period) indeed exists. Its temporal location, sometimes between the first-somites stage and the formation of the posterior neuropore, was in approximate agreement with the morphologically defined phylotypic stage. The molecular evidence for the later parts of the hourglass model, i.e., for von Baer's third law, was stronger than that for the earlier parts.
1828年,卡尔·冯·贝尔提出了一套与胚胎发育有关的四条进化“定律”。根据冯·贝尔的第三定律,不同物种的幼胚相对未分化且彼此相似,但随着发育的进行,物种的显著特征开始显现,不同物种的胚胎逐渐彼此分化。20世纪90年代,丹尼斯·杜布勒和鲁道夫·拉夫分别独立提出了这条定律的扩展形式,即“沙漏模型”。根据沙漏模型,个体发育的特征是起点处不同分类群差异显著,随后是分类群间变异性降低的阶段(系统发育型阶段),最后以分类群间类似冯·贝尔定律的渐进性分化结束。将沙漏模型“翻译”成分子术语可能意味着,在沙漏变窄部分所描述的阶段表达的直系同源基因应该比在系统发育型阶段之前或之后的扩展部分表达的直系同源基因彼此更相似。我们使用了在26个胚胎阶段表达的1585个小鼠基因及其人类直系同源基因来检验这一假设。通过计算小鼠和人类相应直系同源蛋白之间的成对距离,估计了不同胚胎阶段的进化分歧。使用了两个独立的数据集。一个数据集包含仅在单个发育阶段表达的基因;第二个数据集由在不同发育阶段表达的基因组成。在第二个数据集中,基因根据其最早的表达阶段进行分类。我们对这两个数据集拟合了二阶多项式。这两个多项式如沙漏模型所预期的那样显示出最小值。分子结果尽管不太明显,但表明确实存在一个系统发育型阶段(或时期)。其时间位置有时在第一体节阶段和后神经孔形成之间,与形态学定义的系统发育型阶段大致相符。沙漏模型后期部分(即冯·贝尔第三定律)的分子证据比早期部分更强。