Transcriptome sequencing of the blind subterranean mole rat, Spalax galili: utility and potential for the discovery of novel evolutionary patterns

Abstract

The blind subterranean mole rat (Spalax ehrenbergi superspecies) is a model animal for survival under extreme environments due to its ability to live in underground habitats under severe hypoxic stress and darkness. Here we report the transcriptome sequencing of Spalax galili, a chromosomal type of S. ehrenbergi. cDNA pools from muscle and brain tissues isolated from animals exposed to hypoxic and normoxic conditions were sequenced using Sanger, GS FLX, and GS FLX Titanium technologies. Assembly of the sequences yielded over 51,000 isotigs with homology to ∼12,000 mouse, rat or human genes. Based on these results, it was possible to detect large numbers of splice variants, SNPs, and novel transcribed regions. In addition, multiple differential expression patterns were detected between tissues and treatments. The results presented here will serve as a valuable resource for future studies aimed at identifying genes and gene regions evolved during the adaptive radiation associated with underground life of the blind mole rat.

Figure 1. Count and proportion of annotated isotigs as a function of HSP size.

This data refers to isotigs from the combined assembly, based on Blast results against mouse transcripts. Left: the total number isotigs with unique or multiple hits for different HSP sizes. Right: the proportion of isotigs with unique or multiple hits for different HSP sizes.

Figure 2. Shared orthology categories.

(a) Category-3: Three target genes share one orthologous group (upper left). Category-2: Two target genes share one orthologous group (below). Category-1: the target genes are mapped to separate orthologous groups (upper right). Asterisk, near species name, indicates that the target gene is mapped to separate orthologous group, or that no hit was found. (b) The number of isotigs in each category (i.e., 1, 2, and 3) as a function of the HSP coding portion, measured as the total isotig size mapped to target coding region divided by the size mapped to both the coding and untranslated regions; (c) The proportion of isotigs in each category as a function of the HSP coding portion. (d) The number of isotigs in each category as a function of the total length mapped to target transcript. (e) The proportion of isotigs in each category as a function of the total length mapped to target transcript.

Figure 3. Pairwise alignment of S. galili transcribed insert against the homologous Cavia porcellus melusin genomic region.

In order to validate Blast and Lastz indications that the S. galili novel transcribed regions are non-conserved in target genomes we locally aligned the Spalax insert as well as the conserved flanking regions to the target genome. (see Methods). An example is shown above for the putative novel transcribed region in the Spalax melusin transcript (isotig19920). The highly conserved flanking regions (gray blocks) mapped to target regions harboring Cavia exons 4 and 5. The novel Spalax transcribed region (black blocks) is only weakly aligned to the Cavia intronic region, indicating that a homologous exon is probably missing in the target genomic region. Similar alignments were constructed for different target mammals in order to decrease the probability of false positive detection of novel exons. The Cavia exons are labeled in italics. The Cavia intron is underlined.

Figure 4. A model of the S. galil sec23a/sec24 hetero-dimer, color-coded by functional domains.

The protein sec23a appears on the left in bright colors in its putative docked conformation with its sec24 partner (white and grey chains on the right). The 5 known domains are colored in yellow (trunk domain), green (zinc finger domain), violet (Gelsolin), orange (helical domain), and blue (beta-sandwich domain). Putative novel S. galili splice variant flanking amino acids (the start and end point of the new insertion) are presented in red spheres (the novel region is not modeled), and the F382L mutation in cyan spheres.