Decades ago, observations from laboratory experiments with mouse embryos confirmed what we already inferred from nature –a male and female parent are necessary to produce healthy offspring. Uniparental embryos – embryos with either two sets of maternal chromosomes or two sets of paternal chromosomes are not viable.
What is happening in biparental embryos that cannot happen in uniparental embryos? The exact mechanisms and evolutionary reasons are ongoing areas of vigorous research. We do know that uniparental embryos undergo several cycles of cell division, but never come to term, therefore the chromosomes from each parent are not equal in function.
Offspring inherit one copy of each chromosome from each parent. For most genes, offspring express both copies, or alleles, from each parent equally. For many genes, the alleles are expressed unequally, and researchers estimate that less than 1% of our genes are imprinted, i.e. programmed such that one allele is silenced, or at least suppressed, in a way that depends on which parent that allele was inherited from.
Genomic imprinting occurs actively during the earliest stages of embryonic development, from a fertilized single-cell to the 2-cell, 4-cell, and later multi-cell stages prior to implantation. Imprinting occurs through epigenetic modifications – that is, chemical changes to the DNA or to the structures around which DNA is packaged. These chemical changes include methylation of DNA and chemical modifications to the histones, the proteins that make up the nucleosomes around which DNA is “wrapped” before further folding into the structures known as chromosomes. These epigenetic “marks” do not change the sequence of the DNA, but they are inherited by subsequent generations of cells.
During the earliest stages of embryonic development, epigenetic changes to the DNA are quite dynamic. Marks from the parent genomes get removed while genomic imprinting occurs. Uniparental embryos can undergo only several cycles of cell division and never come to term, therefore it this preimplantation phase is crucial for epigenetic reprogramming of the embryo.
Resolving the role of the maternal and paternal genomes in producing viable offspring requires systematically analyzing the transcription patterns at these single- to few- cell stages of uniparental and biparental embryos. The nature of the scientific problem demands the nexus of uniparental embryo culture and single-cell genomics and transcriptomics. The exclusively maternal or paternal nature of uniparental embryos enables researchers to quantify transcripts from each genome separately and in reference to transcript levels of the corresponding genes of biparental embryos.
It may sound conceptually straightforward to perform the comparisons needed to discern parent-dependent allelic transcription, but numerous steps along the sample preparation and RNA-sequencing (RNA-seq) pipeline can yield different quantities of reads associated with each allele. In other words, researchers must be able to validate whether differential read numbers in RNA-seq data originate from a parent-of-origin effect or an artifact of the experiment.
For example, with RNA already being a fragile molecule, it is easy to lose low-copy number transcripts, and probability alone may favor those transcripts remaining in higher numbers with relatively high copy number to be overrepresented in the sequencing library. With sample extraction, read mapping, designing experiments with sufficient replicates, obtaining sufficient sample material, monitoring library complexity throughout the experiments, setting stringent thresholds and exclusion criteria during computational alignment, choosing the appropriate reference, and gene-by-gene validation - some researchers in the field to describe the approach as “more insidious than expected”.
Thus, researchers and technologists continue to polish strategies for all stages of such studies and to customize integrative approaches for the unique challenges of each investigation.
In 2019, Leng et al. contributed a comprehensive survey comparing transcription between uniparental embryos and biparental embryos from the single cell to preimplantation multi-cellular stages.
Parthenogenic (PG) embryos, embryos with solely maternal genomes, were generated by chemically activating oocytes to undergo cell division. Androgenic (AG) embryos, those with solely paternal genomes, were generated by removing the nucleus from an oocyte and injecting the nuclei from two sperm. For these studies, biparental embryos are generated by injecting one set of sperm chromosomes into an oocyte, a typical procedure for in vitro fertilization.
Observations during the first few cell cycles alone indicate that maternal and parental genomes influence the cell cycle. By the third cell division, AG embryos took measurably less time (15.8 +/- 18 hr) to divide than biparental embryos (17.6 +/- 2.0 hr), while PG embryos (19.3 +/- 2.7 hr) took longer. These observations also suggest that, at these stages, the maternal and paternal genomes have opposite control effects.
An analysis of approximately 2,000 Gb of clean RNA-seq data from 296 single cells revealed 801 maternally biased genes (MBGs) – genes with transcript numbers were higher from the maternal allele compared with the paternal allele, and 581 paternally biased genes (PBGs).
Deeper analysis of the functions of the MBGs showed that many were involved in transcriptional activation. The most enriched gene, DUX4, helps activate transcription by opening the chromatin around start sites for other transcription factors that are upregulated during embryonic development. Many of the PBGs were related to processes that shape the embryo and form the blastocoel, the fluid-filled cavity inside the early embryo prior to implantation.
To look deeper into how one type of epigenetic mark on parent genomes relates to genomic imprinting in embryos, Leng et al. performed single-embryo whole genome bisulfite sequencing to reveal the DNA methylation patterns on embryos from the 2-cell to 8-cell stages. The strategy was able to reveal some differences between the methylation patterns on AG and PG genomes, an essential first step to resolving relationships between parental DNA methylation and imprinting in embryos.
Ongoing studies are interrogating the dynamic changes in transcription factors during embryogenesis. Altogether, these strategies shed light on fundamental questions about life for mammals as well as potential applications in regenerative therapy and medicine.