No abstract

Transfer RNAs (tRNAs) play an important role linking mitochondrial RNA and amino acids during protein biogenesis. Four types of tRNA genes have been identified in living organisms. However, the evolutionary origin of tRNAs remains largely unknown. In this article, we conduct a deep sequence analysis of diverse genomes that cover all three domains of life to unveil the evolutionary history of tRNA genes from tRNA halves. tRNA half homologs were detected in diverse organisms, and some of them were expressed in mouse tissues. Continuous tRNA genes have a conserved pattern similar to indels, which is, more closely flanking regions have higher single nucleotide substitution rates, whereas tRNA half homologs do not have this pattern. In addition, tRNAs tend to break into tRNA halves when tissues are incubated in vitro, the tendency of tRNA to break into tRNA halves may be a "side-effect" of tRNA genes evolving from tRNA halves. These results suggest that modern tRNAs originated from tRNA halves through a repeat element-mediated mechanism. These findings provide insight into the evolutionary origin of tRNA genes.

Age-related macular degeneration (AMD) is a progressive neurodegenerative disease of the central retina, with no suitable biomarkers for early diagnosis and treatment. This study aimed to find potential diagnostic biomarker candidates for AMD and investigate their immune-related roles in this pathology. Weight gene correlation analysis was first performed based on data from the Gene Expression Omnibus database and 20 hub genes were identified. The functional enrichment analyses showed that the innate immune response, inflammatory response, and complement activation were key pathways associated with AMD. Complement C1s (C1S), adrenomedullin (ADM), and immediate early response 5 like (IER5L) were identified as the crucial genes with favorable diagnostic values for AMD by using LASSO analysis and multiple logistic regression. Furthermore, a 3-gene model was constructed and proved to be of good diagnostic and predictive performance for AMD (AUC = 0.785, 0.840, and 0.810 in training, test, and validation set, respectively). Finally, CIBERSORT was used to evaluate the infiltration of immune cells in AMD tissues. The results showed that the NK cells, CD4 memory T cell activation, and macrophage polarization may be involved in the AMD process. C1S, ADM, and IER5L were correlated with the infiltration of the above immune cells. In conclusion, our study suggests that C1S, ADM, and IER5L are promising diagnostic biomarker candidates for AMD and may regulate the infiltration of immune cells in the occurrence and progression of AMD.