A survey of small RNAs in human sperm

Abstract

Background: There has been substantial interest in assessing whether RNAs (mRNAs and sncRNAs, i.e. small non-coding) delivered from mammalian spermatozoa play a functional role in early embryo development. While the cadre of spermatozoal mRNAs has been characterized, comparatively little is known about the distribution or function of the estimated 24,000 sncRNAs within each normal human spermatozoon.

Methods: RNAs of <200 bases in length were isolated from the ejaculates from three donors of proved fertility. RNAs of 18-30 nucleotides in length were then used to construct small RNA Digital Gene Expression libraries for Next Generation Sequencing. Known sncRNAs that uniquely mapped to a single location in the human genome were identified.

Results: Bioinformatic analysis revealed the presence of multiple classes of small RNAs in human spermatozoa. The primary classes resolved included microRNA (miRNAs) (≈ 7%), Piwi-interacting piRNAs (≈ 17%), repeat-associated small RNAs (≈ 65%). A minor subset of short RNAs within the transcription start site/promoter fraction (≈ 11%) frames the histone promoter-associated regions enriched in genes of early embryonic development. These have been termed quiescent RNAs.

Conclusions: A complex population of male derived sncRNAs that are available for delivery upon fertilization was revealed. Sperm miRNA-targeted enrichment in the human oocyte is consistent with their role as modifiers of early post-fertilization. The relative abundance of piRNAs and repeat-associated RNAs suggests that they may assume a role in confrontation and consolidation. This may ensure the compatibility of the genomes at fertilization.

Figure 1

Sequences shared among the three human sperm sncRNA libraries. The proportion of the total number of sequences that are unique to a library, shared with at least one other library or present in all three libraries (in common to all three libraries) are presented for each of the libraries. For each library, contigs were constructed according to information on chromosome, strand, start and end position, allowing new reads to join to a contig if there is overlap or if the read is less than or equal to four nucleotides away on the same chromosome/strand. A contig is synonymous with a unique locus. (A)–(C) Represent counts of reads in common between AS062, AS064 and AS066, based on these contigs. (A). For each library, reads were individually compared with the sets of contigs of the other libraries, resulting in numbers of reads that were unique to the library, or in common to one or both of the other libraries. The results were normalized to a million counts and plotted with a color code to show the percentage for each commonality subset, described in the legend, with purple representing reads common to all libraries and blue for reads that are donor-specific. (B). Identical to 1A, except showing only the reads that are only in one library, and the corresponding percentages. (C). This figure compares the relative percentages of shared reads—the percentages were calculated after the donor-specific reads were subtracted out.

Figure 2

Classification of all the sequences found in the three human sperm sncRNA libraries. (A) Sequences mapping to known genomic elements, TSS and promoters. The distribution of sequences mapping to miRNA, piRNA, snoRNA, snRNA and repeats are shown here. In addition, those sequences that did not map to known genomic elements were analyzed for TSS and promoter association. (B) Sequences mapping to miRNA or piRNA as well as repeats, CpG islands, histones and TSS or promoters. All sequences associated with miRNA (blue) or piRNA (red) were further analyzed to determine if they also map to repeats, CpG islands, histones or TSS/promoters. (C) Sequences mapping to known repeat classes. This figure shows the majority of sequences associated with an undefined repeat category. Of the remaining categories, LINE, LTR and SINE are highly represented.

Figure 3

piRNA-targeting repetitive elements. Distribution of repetitive elements targeted by piRNAs. For each repeat consensus the relative affinity of the piRNA in each category was determined. The value for each piRNA in each repeat class is standardized to account for the number of positions within the repeat a piRNA targets, the length of the repeat and its frequency in the genome. Only piRNA with a maximum relative affinity in the top 5% are illustrated.