Larue B, Cedergren R J, Sankoff D, Grosjean H
J Mol Evol. 1979 Dec;14(4):287-300. doi: 10.1007/BF01732496.
Sequence data from methionine initiator and phenylalanine transfer RNAs were used to construct phylogenetic trees by the maximum parsimony method. Although eukaryotes, prokaryotes and chloroplasts appear related to a common ancestor, no firm conclusion can be drawn at this time about mitochondrial-coded transfer RNAs. tRNA evolution is not appropriately described by random hit models, since the various regions of the molecule differ sharply in their mutational fixation rates. "Hot" mutational spots are identified in the Tpsic, the amino acceptor and the upper anticodon stems; the D arm and the loop areas on the other hand are highly conserved. Crucial tertiary interactions are thus essentially preserved while most of the double helical domain undergoes base pair interchange. Transitions are about half as costly as transversions, suggesting that base pair interchanges proceed mostly through G-U and A-C intermediates. There is a preponderance of replacements starting from G and C but this bias appears to follow the high G + C content of the easily mutated base paired regions.
利用甲硫氨酸起始转运RNA和苯丙氨酸转运RNA的序列数据，通过最大简约法构建系统发育树。尽管真核生物、原核生物和叶绿体似乎与一个共同祖先相关，但目前关于线粒体编码的转运RNA还无法得出确凿结论。随机命中模型无法恰当地描述tRNA的进化，因为分子的各个区域在突变固定率上差异很大。在TψC、氨基受体和反密码子上部茎区发现了“热点”突变位点；另一方面，D臂和环区高度保守。因此，关键的三级相互作用基本得以保留，而大部分双螺旋结构域发生碱基对互换。转换的代价约为颠换的一半，这表明碱基对互换大多通过G-U和A-C中间体进行。从G和C开始的替换占优势，但这种偏向似乎与易突变碱基对区域的高G + C含量有关。
