Convergent loss of the D-arm in mitochondrial tRNA-Cys across sampled Serpentes mitogenomes: a systematic comparative analysis

Background The dihydrouridine arm (D-arm) is a conserved structural element of canonical transfer RNAs (tRNAs) and contributes to the characteristic L-shaped tertiary structure, required for ribosomal interaction. In mitochondrial tRNAs (mt-tRNAs), however, structural constraints are relaxed, and loss of the D-arm has been observed in several metazoan phylogenies, indicating that alternative stabilisation mechanisms may compensate for its absence. In snakes (Serpentes)—a species-rich reptile order with highly derived mitochondrial genomes—it has not been systematically investigated whether this structural simplification has occurred. Building on our recent description of complete D-arm loss in the mitochondrial tRNA-Cys of Macrovipera lebetinus lebetinus , we present the first comparative analysis of this structural feature in Serpentes. Results A total of 136 serpent mitogenomes and two lizard outgroups ( Anolis carolinensis , D-arm present; Pogona vitticeps , D-arm lost) were analysed using tRNAscan-SE 2.0.12 with the vertebrate mitochondrial model. Loss of the D-arm was detected in 20 sequences (14.7%). This modification was observed in all sampled Viperinae (8/8), Xenodermidae (3/3), Acrochordidae (1/1), and Tropidophiidae (1/1), and partially prevalent in Pythonidae (2/4, 50%), Elapidae (2/13, 15.4%), and Crotalinae (2/31, 6.5%), with a single case in Colubridae (1/58, 1.7%). Maximum parsimony reconstruction required nine changes of state, corresponding to seven to nine independent loss events; stochastic character mapping gave a mean of 9.14 ( 95% CI: 8-11). Cysteine codon usage showed no significant association with D-arm loss (t-test: p = 0.755; Mann-Whitney U test: p = 0.837). D-arm loss was, however, associated with a significantly shorter TΨC-arm stem (phylANOVA: p = 0.029), unlike the compensatory T-arm elongation described for human tRNA-Ser. A sensitivity analysis of five sequences with discordant annotations indicates five or six independent losses under alternative annotations. Conclusions Predicted D-arm loss in mitochondrial tRNA-Cys is a recurring, independently evolved, structural reduction in Serpentes. No association with codon usage was detected, either before or after phylogenetic correction (phylANOVA: p = 0.883), whereas TΨC-arm stem length differed significantly, suggesting a distinct mode of structural accommodation from that reported for tRNA-Ser. These results highlight the structural plasticity of mitochondrial tRNA in reptiles and raise questions about the compensatory molecular mechanisms—including possible mitochondrial-nuclear coevolution—that enable functional tRNA activity even in the absence of a canonical structural element.