Origin and evolution of a chimeric fusion gene in Drosophila subobscura, D. madeirensis and D. guanche

Abstract

An understanding of the mutational and evolutionary mechanisms underlying the emergence of novel genes is critical to studies of phenotypic and genomic evolution. Here we describe a new example of a recently formed chimeric fusion gene that occurs in Drosophila guanche, D. madeirensis, and D. subobscura. This new gene, which we name Adh-Twain, resulted from an Adh mRNA that retrotransposed into the Gapdh-like gene, CG9010. Adh-Twain is transcribed; its 5' promoters and transcription patterns appear similar to those of CG9010. Population genetic and phylogenetic analyses suggest that the amino acid sequence of Adh-Twain evolved rapidly via directional selection shortly after it arose. Its more recent history, however, is characterized by slower evolution consistent with increasing functional constraints. We present a model for the origin of this new gene and discuss genetic and evolutionary factors affecting the evolution of new genes and functions.

Figure 1.—

Transcription patterns of Adh-Twain and CG9010 are similar. We used RT-PCR to amplify a fragment of fh, Adh-Twain (fusion), and CG9010 from cDNA made from poly(A⁺) RNA extracted from D. subobscura adult females, adult males, and larvae. The same cDNA prep was used for all three amplifications. Adh-Twain is actively transcribed in all three samples. Qualitatively, the expression of Adh-Twain is similar to that of CG9010. The fh gene was a control to show that our cDNA did not contain genomic DNA. The fh amplicon spans an intron; thus, if the cDNA were contaminated with genomic DNA we would observe a second band in the fh lanes.

Figure 2.—

5′-UTR and promoters are similar for Adh-Twain (D. sub Fusion) and CG9010 and conserved across taxa.

Figure 3.—

Major sequence features of Adh-Twain compared to patterns of polymorphism and divergence. (A) Comparison of the nucleotide polymorphism observed at Adh-Twain in our population of D. subobscura (black line) and the nucleotide divergence between D. subobscura and D. guanche Adh-Twain (gray line). (B) Divergence between D. subobscura and D. guanche CG9010 genes (black line) and divergence between D. subobscura Adh-Twain and CG9010 (gray line). (C) Divergence between D. subobscura and D. guanche Adh genes (black line) and divergence between D. subobscura Adh-Twain and Adh (gray line). Both B and C suggest that, relative to Adh, a distinctly different set of nucleotides are under constraint in Adh-Twain.

Figure 4.—

Nucleotide substitutions in Adh and CG9010 regions of Adh-Twain and their parental genes. These results are based on our PAML analysis. The solid bars represent the time period during which we hypothesize that directional selection acted on Adh-Twain. Estimates of the number of nonsynonymous and synonymous changes are beneath the branches of the Adh-Twain lineages. In both the Adh-derived and the CG9010-derived regions of the fusion gene there has been an increase in the rates of amino acid substitution, but not synonymous substitutions, relative to the ancestral genes.

Figure 5.—

A model for the evolution of Adh-Twain. (A) Chromosomal duplication of CG9010 in the ancestor of D. subobscura, D. madeirensis, and D. guanche. (B) Retrotransposition of an Adh mRNA into one of the copies of CG9010. Note that the 5′ regulatory regions of CG9010 are conserved. (C) Subsequent formation of a contiguous ORF spanning the remainder of CG9010 and the Adh retrosequence. The 3′ end of CG9010 is lost.