Evolution of hydra, a recently evolved testis-expressed gene with nine alternative first exons in Drosophila melanogaster

Abstract

We describe here the Drosophila gene hydra that appears to have originated de novo in the melanogaster subgroup and subsequently evolved in both structure and expression level in Drosophila melanogaster and its sibling species. D. melanogaster hydra encodes a predicted protein of approximately 300 amino acids with no apparent similarity to any previously known proteins. The syntenic region flanking hydra on both sides is found in both D. ananassae and D. pseudoobscura, but hydra is found only in melanogaster subgroup species, suggesting that it originated less than approximately 13 million y ago. Exon 1 of hydra has undergone recurrent duplications, leading to the formation of nine tandem alternative exon 1s in D. melanogaster. Seven of these alternative exons are flanked on their 3' side by the transposon DINE-1 (Drosophila interspersed element-1). We demonstrate that at least four of the nine duplicated exon 1s can function as alternative transcription start sites. The entire hydra locus has also duplicated in D. simulans and D. sechellia. D. melanogaster hydra is expressed most intensely in the proximal testis, suggesting a role in late-stage spermatogenesis. The coding region of hydra has a relatively high Ka/Ks ratio between species, but the ratio is less than 1 in all comparisons, suggesting that hydra is subject to functional constraint. Analysis of sequence polymorphism and divergence of hydra shows that it has evolved under positive selection in the lineage leading to D. melanogaster. The dramatic structural changes surrounding the first exons do not affect the tissue specificity of gene expression: hydra is expressed predominantly in the testes in D. melanogaster, D. simulans, and D. yakuba. However, we have found that expression level changed dramatically (approximately >20-fold) between D. melanogaster and D. simulans. While hydra initially evolved in the absence of nearby transposable element insertions, we suggest that the subsequent accumulation of repetitive sequences in the hydra region may have contributed to structural and expression-level evolution by inducing rearrangements and causing local heterochromatinization. Our analysis further shows that recurrent evolution of both gene structure and expression level may be characteristics of newly evolved genes. We also suggest that late-stage spermatogenesis is the functional target for newly evolved and rapidly evolving male-specific genes.

Figure 1. Evolution of hydra in Drosophila Species

(A) The hydra region of the X-chromosome of D. melanogaster, based on FlyBase genome browser release 4.3. The hydra gene was previously annotated as producing two alternative transcripts, RA and RB, derived from alternative exon 1s. The proposed annotation of seven additional exon 1s is based on evidence presented here. (B) Evolution of hydra region and hydra gene structure in seven Drosophila species. hydra and the flanking gene CG1835 are located in a recently expanded region between run and cyp6v1. hydra originated in the common ancestor of the melanogaster subgroup (arrow A). In D. melanogaster, this region between run and cyp6v1 is ~32 kb (10 kb from the 3′ end of hydra to cyp6v1 and 17 kb from the 5′ end of hydra to run), but is only ~26 kb apart in D. ananassae and D. pseudoobscura, where both hydra and CG1835 are missing. hydra has gone through multiple cycles of duplication and rearrangement in D. melanogaster and its sibling species, and accumulated insertions of the transposon DINE-1 and other repetitive sequences (arrow B). CG1835 is on the opposite strand from all other genes, as indicated by its leftward-pointing arrow. Three copies of hydra are found on two unlinked scaffolds in D. sechellia. Note that the distances are not to scale.

Figure 2. Length Variation of the Exon 1 Region of hydra in D. melanogaster

(A) Location of the probes (thick black lines) and restriction enzyme SacII cutting sites in hydra of D. melanogaster (top) and D. simulans (bottom). Note that the region homologous to the probe is duplicated in D. simulans. (B–C) Southern hybridization of genomic DNA extracted from multiple lines of D. melanogaster (B) and D. simulans (C). D. melanogaster lines are all from Zimbabwe, Africa. D. simulans lines 1–4 are from California, United States; lines 5–7 are from Zimbabwe, Africa; and lines 8–14 are from Madagascar, Africa.

Figure 3. Phylogenetic Trees of hydra

Trees are based on (A) exon 2–4 sequences and (B) exon 1, including putative 5′ UTR sequences. The trees were built using the neighbor-joining method according to the Jukes-Cantor substitution model. Branch lengths are proportional to the number of nucleotide substitutions. Percentage bootstrap values are shown at each node (500 replicates). In (B), sequences mel 1–9 correspond to the alternative exon 1s of D. melanogaster shown in Figure 1B. Sequences Ra and Rb are from the annotated transcripts of D. melanogaster hydra in FlyBase. D. simulans and D. sechellia exon 1s are labeled as shown in Figure 1B. Note that mel 6 is part of group B but clusters with group A, presumably because of having a divergent 5′ UTR.

Figure 4. Expression of hydra detected by RT-PCR

(A) D. melanogaster, D. simulans, and D. yakuba. (B) Canton S line of D. melanogaster. M, DNA kb size marker; G, genomic DNA; T, adult testis RNA; O, RNA from whole body of males with testis removed; +, positive control for cDNA O using primers for the gene GAPDH; −, negative PCR control. RNA was extracted from strain ORC of D. melanogaster (A), California strain w54 of D. simulans, and Tai18E2 of D. yakuba. Note that expression of hydra in D. melanogaster is specific to testis in the ORC strain (A), but slight expression was detected in other tissues in the Canton S strain (B).

Figure 5. In Situ Hybridization of DIG-Labeled Antisense RNA (A) and Sense Control RNA (B) of hydra to Testes of D. melanogaster

Expression is strongest in the proximal region of testes. ag, accessory glands; ed, ejaculatory duct.

Figure 6. Relative Abundance of hydra mRNA Quantified by Real-Time RT-PCR in Various Lines of D. melanogaster and D. simulans

For each line, the abundance of hydra relative to GAPDH was determined. This value was then compared for each line relative to D. melanogaster Canton S (mel_CS), which is normalized here to a value of 1.