Recurrent Innovation at Genes Required for Telomere Integrity in Drosophila

Abstract

Telomeres are nucleoprotein complexes at the ends of linear chromosomes. These specialized structures ensure genome integrity and faithful chromosome inheritance. Recurrent addition of repetitive, telomere-specific DNA elements to chromosome ends combats end-attrition, while specialized telomere-associated proteins protect naked, double-stranded chromosome ends from promiscuous repair into end-to-end fusions. Although telomere length homeostasis and end-protection are ubiquitous across eukaryotes, there is sporadic but building evidence that the molecular machinery supporting these essential processes evolves rapidly. Nevertheless, no global analysis of the evolutionary forces that shape these fast-evolving proteins has been performed on any eukaryote. The abundant population and comparative genomic resources of Drosophila melanogaster and its close relatives offer us a unique opportunity to fill this gap. Here we leverage population genetics, molecular evolution, and phylogenomics to define the scope and evolutionary mechanisms driving fast evolution of genes required for telomere integrity. We uncover evidence of pervasive positive selection across multiple evolutionary timescales. We also document prolific expansion, turnover, and expression evolution in gene families founded by telomeric proteins. Motivated by the mutant phenotypes and molecular roles of these fast-evolving genes, we put forward four alternative, but not mutually exclusive, models of intra-genomic conflict that may play out at very termini of eukaryotic chromosomes. Our findings set the stage for investigating both the genetic causes and functional consequences of telomere protein evolution in Drosophila and beyond.

Keywords: Drosophila; gene turnover; positive selection; telomere; terminin.; transposable element.

Fig. 1

Two estimates of adaptive evolution across mutant classes and broadly defined molecular functions. dN = rate of nonsynonymous substitutions, dS = rate of synonymous substitutions. α refers to the proportion of adaptive sites. (MWU test of gene groups compared with all genes: *P < 0.05, **P < 0.01, ***P < 0.001).

Fig. 2

Results of phylogenomic analysis across the melanogaster group. Bayesian phylogenetic trees constructed from alignable regions of (A) HipHop-like genes, (B) caravaggio-like genes, and (C) nucleosome assembly protein 1-like genes. Paralog gene names refer to the parent gene and cytolocation in the syntenic region of D. melanogaster. (D) Syntenic locations of parent and daughter genes across the D. melanogaster reference chromosomes. (E) Birth (solid arrow) and death (dotted line arrow) events inferred by and mapped onto species tree. Posterior probabilities > 0.70 indicated on branches and only for nodes of interest.

Fig. 3

Expression evolution across parent and daughter telomere integrity genes. (A) Gene families founded by Hiphop, (B) cav, and (C) nap-1 RT-PCR products amplified with species- and gene-specific primers (left) across a panel of adult tissues (H = head, O = ovaries, C = remaining carcass, T = testis). *Refers to the smaller isoform of two detected across all tissues.

Fig. 4

Models of intra-genomic conflict proposed to drive telomere protein evolution. Under the Interference model, the selfish element (red box) sabotages the alternative allele, in this case, the sperm that did not inherit the element (yellow chromosome). These sperm either fail to mature or paternal chromosomes mis-segregate in the early embryo. Under the overproliferation model, (1) escaped telomere-restricted transposable elements (red triangles) invade the chromosome end and/or the chromosome interior or (II) the DNA repair process is hijacked in one of two ways: a selfish element (red box) hijacks Break-Induced-Replication (BIR) such that the purple chromosome telomere recurrently serves as the donor sequence for telomere capture either on the homologous chromosome or even non-homologous chromosomes. Alternatively, a telomere-specialized transposable elements opportunistically inserts at an internal double stranded break. Under the Meiotic Drive model, the selfish element migrates to the egg position rather than to the polar bodies that are not transmitted to the next generation. Meiosis I or II may be hijacked.