Stallion sperm transcriptome comprises functionally coherent coding and regulatory RNAs as revealed by microarray analysis and RNA-seq

Abstract

Mature mammalian sperm contain a complex population of RNAs some of which might regulate spermatogenesis while others probably play a role in fertilization and early development. Due to this limited knowledge, the biological functions of sperm RNAs remain enigmatic. Here we report the first characterization of the global transcriptome of the sperm of fertile stallions. The findings improved understanding of the biological significance of sperm RNAs which in turn will allow the discovery of sperm-based biomarkers for stallion fertility. The stallion sperm transcriptome was interrogated by analyzing sperm and testes RNA on a 21,000-element equine whole-genome oligoarray and by RNA-seq. Microarray analysis revealed 6,761 transcripts in the sperm, of which 165 were sperm-enriched, and 155 were differentially expressed between the sperm and testes. Next, 70 million raw reads were generated by RNA-seq of which 50% could be aligned with the horse reference genome. A total of 19,257 sequence tags were mapped to all horse chromosomes and the mitochondrial genome. The highest density of mapped transcripts was in gene-rich ECA11, 12 and 13, and the lowest in gene-poor ECA9 and X; 7 gene transcripts originated from ECAY. Structural annotation aligned sperm transcripts with 4,504 known horse and/or human genes, rRNAs and 82 miRNAs, whereas 13,354 sequence tags remained anonymous. The data were aligned with selected equine gene models to identify additional exons and splice variants. Gene Ontology annotations showed that sperm transcripts were associated with molecular processes (chemoattractant-activated signal transduction, ion transport) and cellular components (membranes and vesicles) related to known sperm functions at fertilization, while some messenger and micro RNAs might be critical for early development. The findings suggest that the rich repertoire of coding and non-coding RNAs in stallion sperm is not a random remnant from spermatogenesis in testes but a selectively retained and functionally coherent collection of RNAs.

Figure 1. Venn diagram of transcripts detected in stallion sperm and testes by microarray analysis (SNR ≥2).

Figure 2. Heat maps of GO functional groups for sperm up-regulated (a) and sperm down-regulated (b) transcripts.

Blue boxes denote that the gene has not been associated with the corresponding GO category. Genes with symbols in red font were validated by qRT-PCR.

Figure 3. Validation of significantly (p<0.05) sperm up-regulated (a) and sperm down-regulated (b) genes by qRT-PCR (see also Table 1).

Figure 4. Summary statistics for mapped RNA sequence tags

: (a) Comparison of mapped tags (AC≥1) between the two sperm samples; (b) Proportions of tags with very high (AC≥100), high (10<AC<100), and medium (1≤AC≤10) expression.

Figure 5. Structural annotation of 19, 257 mapped RNA sequence tags (AC≥1):

(a) Distribution of the tags in structural annotation categories by ERANGE; (b) Comparison of annotated genes by GOanna (human genome) and ERANGE (horse genome).

Figure 6. Comparison of RNA-seq data with current equine gene models:

(a) PKM2 showing 9 in silico prediction sites, of which two are positioned 5′ upstream to exon 1; (b) CRISP3 with 3 in silico prediction sites, all located 5′ upstream to exon 1; (c) PRM1 and TNP2 cluster (the protamine cluster) with 12 in silico prediction sites of which only two align with PRM1 and TNP2 exons. Black boxes with numbers –exons in current gene models; blue boxes –very highly expressed tags (AC≥100); red boxes–highly expressed tags (10<AC<100); green boxes–tags with medium expression (1≤AC≤100). Exact start and end sites of all mapped tags are presented in Additional file 7.

Figure 7. ECAY transcripts in stallion sperm.

Agarose gel images showing RT-PCR amplicons of 7 ECAY genes and transcripts in stallion sperm.