Bavarian State Collection of Zoology, Münchhausenstr. 21, 81247 Munich, Germany.
Mol Phylogenet Evol. 2013 Nov;69(2):376-92. doi: 10.1016/j.ympev.2012.11.017. Epub 2012 Dec 8.
The origin of molluscs among lophotrochozoan metazoans is unresolved and interclass relationships are contradictory between morphology-based, multi-locus, and recent phylogenomic analyses. Within the "Deep Metazoan Phylogeny" framework, all available molluscan mitochondrial genomes were compiled, covering 6 of 8 classes. Genomes were reannotated, and 13 protein coding genes (PCGs) were analyzed in various taxon settings, under multiple masking and coding regimes. Maximum Likelihood based methods were used for phylogenetic reconstructions. In all cases, molluscs result mixed up with lophotrochozoan outgroups, and most molluscan classes with more than single representatives available are non-monophyletic. We discuss systematic errors such as long branch attraction to cause aberrant, basal positions of fast evolving ingroups such as scaphopods, patellogastropods and, in particular, the gastropod subgroup Heterobranchia. Mitochondrial sequences analyzed either as amino acids or nucleotides may perform well in some (Cephalopoda) but not in other palaeozoic molluscan groups; they are not suitable to reconstruct deep (Cambrian) molluscan evolution. Supposedly "rare" mitochondrial genome level features have long been promoted as phylogenetically informative. In our newly annotated data set, features such as genome size, transcription on one or both strands, and certain coupled pairs of PCGs show a homoplastic, but obviously non-random distribution. Apparently congruent (but not unambiguous) signal for non-trivial subclades, e.g. for a clade composed of pteriomorph and heterodont bivalves, needs confirmation from a more comprehensive bivalve sampling. We found that larger clusters not only of PCGs but also of rRNAs and even tRNAs can bear local phylogenetic signal; adding trnG-trnE to the end of the ancestral cluster trnM-trnC-trnY-trnW-trnQ might be synapomorphic for Mollusca. Mitochondrial gene arrangement and other genome level features explored and reviewed herein thus failed as golden bullets, but are promising as additional characters or evidence supporting deep molluscan clades revealed by other data sets. A representative and dense sampling of molluscan subgroups may contribute to resolve contentious interclass relationships in the future, and is vital for exploring the evolution of especially diverse mitochondrial genomes in molluscs.
在有胚动物中，软体动物的起源尚未解决，形态学、多基因座和最近的系统基因组分析之间的类间关系相互矛盾。在“深层后生动物系统发育”框架内，编译了所有可用的软体动物线粒体基因组，涵盖了 8 个门中的 6 个。对基因组进行了重新注释，并在多种屏蔽和编码方案下，在各种分类群设置中分析了 13 个蛋白质编码基因（PCG）。使用最大似然法进行系统发育重建。在所有情况下，软体动物与有胚动物的外群混合在一起，而且大多数有多个代表的软体动物类群不是单系的。我们讨论了系统错误，例如长枝吸引，导致快速进化的内群（如石鳖、腹足纲和特别的腹足纲异齿亚纲）的异常、基部位置。分析的线粒体序列无论是作为氨基酸还是核苷酸，在某些（头足纲）软体动物中表现良好，但在其他古生代软体动物群中表现不佳；它们不适合重建深（寒武纪）软体动物的进化。据称，“罕见”的线粒体基因组水平特征长期以来一直被认为具有系统发育信息。在我们新注释的数据集中，基因组大小、一条或两条链上的转录以及某些 PCG 对的特定偶联对显示出同态，但显然不是随机分布。显然一致（但不是明确）的非平凡亚群信号，例如由翼足类和异齿类双壳类组成的类群，需要来自更全面的双壳类采样的确认。我们发现，不仅 PCG 而且 rRNA 甚至 tRNA 的较大簇都可以承载局部系统发育信号；将 trnG-trnE 添加到祖先簇 trnM-trnC-trnY-trnW-trnQ 的末端可能是软体动物的 synapomorphic。因此，本文探索和回顾的线粒体基因排列和其他基因组水平特征未能成为“金弹”，但作为支持其他数据集揭示的深层软体动物类群的附加特征或证据很有前途。软体动物亚群的代表性和密集采样可能有助于解决未来有争议的类间关系，对于探索软体动物中多样化的线粒体基因组的进化至关重要。
