Emerging evidence that the mammalian sperm epigenome serves as a template for embryo development

Abstract

Although more studies are demonstrating that a father's environment can influence child health and disease, the molecular mechanisms underlying non-genetic inheritance remain unclear. It was previously thought that sperm exclusively contributed its genome to the egg. More recently, association studies have shown that various environmental exposures including poor diet, toxicants, and stress, perturbed epigenetic marks in sperm at important reproductive and developmental loci that were associated with offspring phenotypes. The molecular and cellular routes that underlie how epigenetic marks are transmitted at fertilization, to resist epigenetic reprogramming in the embryo, and drive phenotypic changes are only now beginning to be unraveled. Here, we provide an overview of the state of the field of intergenerational paternal epigenetic inheritance in mammals and present new insights into the relationship between embryo development and the three pillars of epigenetic inheritance: chromatin, DNA methylation, and non-coding RNAs. We evaluate compelling evidence of sperm-mediated transmission and retention of paternal epigenetic marks in the embryo. Using landmark examples, we discuss how sperm-inherited regions may escape reprogramming to impact development via mechanisms that implicate transcription factors, chromatin organization, and transposable elements. Finally, we link paternally transmitted epigenetic marks to functional changes in the pre- and post-implantation embryo. Understanding how sperm-inherited epigenetic factors influence embryo development will permit a greater understanding related to the developmental origins of health and disease.

Fig. 1. The final composition of the sperm epigenome prepares the embryo for developmental competency.

A A schematic showing the global epigenetic composition of sperm relative to functional loci of interest from the perspective of spermatogenesis and embryo development, based on whole-genome sequencing experiments^,,,–,,,,,. For details on the dynamic nucleoprotein composition through spermatogenesis refer to Box 1. Active and narrow marks in sperm including H3K4me1, H3K4me2, H3K4me3 and H3K27ac, are enriched at many promoters implicated in spermatogenesis, basic cellular processes, and development. Interestingly, these active marks are also enriched in intergenic space and putative enhancers. The enrichment of these marks relative to each other may influence function. Other marks such as the active H3K36me3, and the repressive H3K9me3 and H2AK119ub1, broadly cover certain gene bodies, promoters, and intergenic space, including transposable elements. Finally, although over 70% of the sperm epigenome is in a methylated state, DNA methylation remains mostly devoid at CpG dense promoters bearing nucleosomes. Panel created with BioRender.com. B Non-coding RNAs coincide histone post translational modifications. Z-scores between the relationship of non-coding RNAs and different histone post-translational modifications in sperm. Raw FASTQ files from the sperm PANDORA-seq small non-coding RNA dataset were downloaded (accession number GSE144666). PANDORA-seq small non-coding RNA datasets were aligned to the mm10 genome and annotated based on the SPORTS1.1 pipeline with one mismatch tolerance (github pipeline: https://github.com/junchaoshi/sports1.1). Classes of small non-coding RNAs are curated by SPORTS1.1. As an output, a FASTA file is obtained, and this file subsequently provides the specific small non-coding RNA sequences captured by the PANDORA-seq in sperm. The DNA sequence that coincided with different histone modification peaks in sperm was extracted, and whether the DNA sequence from the PANDORA-seq FASTA files overlapped to the DNA sequence from the histone modifications was determined.

Fig. 2. Chromatin marks in sperm and their relationship to the oocyte and 2-cell embryo.

A Heatmaps indicating H3K4me3, H3K27me3, H3K27ac, H3K36me3, or H3K9me3 regions in sperm and these marks’ enrichment in the oocyte and in 2-cell embryos. Color of the signal corresponds to relative RPKM counts. Sperm H3K4me3 and H3K27me3 datasets were generated from Lismer et al.; oocyte H3K4me3 dataset was retrieved from Liu et al.; 2-cell H3K4me3 dataset was retrieved from Liu et al.. Sperm H3K27ac dataset was generated in house; oocyte and 2-cell H3K27ac datasets were retrieved from Dahl et al.. Sperm, oocyte, and 2-cell H3K36me3 datasets were retrieved from Xu et al., 2018. Sperm, oocyte, and 2-cell H3K9me3 datasets were retrieved from Wang et al.. Raw FASTQ files were aligned as indicated in Lismer et al.. Regions with histone post-translational histone enrichments in sperm were identified using the csaw Bioconductor package. Heatmaps were generated using the DeepTools software. B DNA methylation (DNAme) is altered in sperm from DDT-exposed Greenlandic Inuit and South African Vhavenda men at regions that predicted to persist in the developing embryo. Ridge plots corresponding to the density of low, dynamic, or high CpG DNAme levels (x-axis) for different stages of pre-implantation embryo development (y-axis) at CpGs overlapping MCC-seq background (gray), Greenlandic sperm DNAme gain DMCs (dark blue), Greenlandic sperm DNAme loss DMCs (light blue), South African sperm DNAme gain DMCs (red), and South African sperm DNAme loss DMCs (yellow). Gray shaded boxes correspond to DMCs that retain dynamic CpG levels throughout pre-implantation embryogenesis. See Lismer et al., 2022 for full experimental details.

Fig. 3. Transcription factors and transposable elements may promote paternal epigenetic inheritance.

A Table summarizing what is known regarding the enrichment or disruption of epigenetic marks at specific classes of transposable elements in sperm or embryos from mice or men. B Diagram illustrating the divergence of transposable element copies after integration in the genome and relationships to transposable element age. For each transposable element integrated in the genome, a percent divergence score can be calculated from its consensus sequence. A low percent divergence score reflects a “young” transposable element that has more recently been integrated in the genome as it has not yet accumulated many sequence substitutions (stars), deletions, insertions, truncations, or undergone recombination. Conversely, a high percent divergence corresponds to an “old” transposable element. These old transposable elements may have diverged too far from the original transcription factor binding site, preventing the recognition by transcription factors and downstream activity. Because young transposable elements retain their recognizable transcription factor binding site, they have higher activity and may be epigenetically silenced if not beneficial to the host. Panel created with BioRender.com. C Narrow sperm H3K4me3 peaks that overlap the insulator transcription factor CTCF, important for the 3D chromatin organization in cells, retain H3K4me3 in the pre-implantation embryo. Broad H3K4me3 peaks are not enriched for CTCF in sperm and lose H3K4me3 in the pre-implantation embryo, conferring that CTCF may play a role in the transmission of sperm H3K4me3 post-fertilization. The insulator protein CTCF acts alongside cohesin to regulate topological associated domains in cells by folding domains into loop structures. In line with this, Hi-C experiments have shown a high conservation between the 3D chromatin organization in sperm and in pre-implantation embryos post zygotic genome activation, highlighting the interplay between architectural proteins, and inherited histone modifications in sperm. Panel created with BioRender.com.

Fig. 4. Moving from association to function in the field of paternal epigenetic inheritance.

Overview of remaining questions that will permit the field of paternal epigenetic inheritance to understand how transmitted epigenetic marks elicit changes in embryo development intergenerationally and transgenerationally. While the molecular mechanisms underlying epigenetic inheritance are partially resolved there remain unanswered questions. Current studies suggest that exposures can alter epigenetic marks including DNA and histone methylation (H3K4me3), some of which may alter transcription factor binding and consequently chromatin organization. These epigenetic changes are partially retained on the paternal chromatin where they alter transcription in the early embryo leading to downstream phenotypes. It is unclear what marks may persist through lineage segregation and cell differentiation. Boxed are examples of genes that have altered H3K4me3 and / or DNA methylation in folate deficient mouse sperm or DDT-exposed South African human sperm. As highlighted, genes with altered epigenetic marks are involved in both pre- and post-implantation development. Whether genes that are expressed at later stages of development carry sperm-inherited epigenomic alterations throughout embryogenesis until their expression, has never been explored. How sperm RNA content is altered in relation to chromatin and DNA methylation is unknown as is how sperm-transmitted RNA leads to altered embryonic gene expression. Figure created with BioRender.com.