DISCUSSION

Despite recent progress in understanding transcriptome-wide m6A in mammalian preimplantation embryos,10,12-15 the landscape of m6A dynamics at single-nucleotide resolution during this critical period remains incomplete. Such resolution is necessary to characterize the functions of site-specific m6A marks, particularly those exhibiting pronounced dynamics during MZT. In this study, we employed SAC-seq to simultaneously characterize the m6A epitranscriptome and transcriptome from the same samples, generating the first single-base-resolution m6A landscapes across bovine oocytes and preimplantation embryos.

Through comprehensive bioinformatic analysis, several key insights emerged. First, m6A modification was widespread and highly dynamic during preimplantation development, with elevated m6A levels in GV and MII oocytes, followed by a sharp decrease immediately after fertilization, coinciding with MZT.17,35 This pattern is consistent with observations in mouse and human embryos,12,15 where m6A has been implicated in maternal mRNA clearance and zygotic transcript stabilization.4,9 Second, m6A was enriched among DEGs, particularly those involved in translation, mitochondrial metabolism, and cell cycle regulation, suggesting that m6A acts as a selective regulator of developmental transcript fate. Within DEG clusters, subcluster-based analysis further revealed distinct dynamic patterns of m6A modification associated with stage-specific biological functions. Notably, transcripts activated after ZGA were enriched in translation-related processes, which are essential for developmental competence.42 We further observed that transcripts with higher m6A site density often showed increased expression at developmental stages where their functions become prominent, suggesting that m6A density may fine-tune translational efficiency and/or RNA stability in a context-dependent manner. However, in many subclusters, changes in m6A density were not synchronized with gene expression changes, indicating partial uncoupling between epitranscriptomic and transcriptional regulation. Third, we characterized m6A in non-coding transcripts and observed dynamic, class-specific enrichment patterns in lncRNAs, snRNAs, and snoRNAs. Finally, to further explore m6A regulation of ribosome-associated gene sets, we integrated transcriptome, m6A epitranscriptome, and translatome data.35 These analyses showed that m6A contributes to the regulation of ribosome-related gene sets during early embryogenesis, potentially influencing translational dynamics in a stage- or context-specific manner, thereby supporting models of ribosome heterogeneity.28,36

Interestingly, we observed limited concordance between m6A dynamics and gene expression changes across developmental stages, differing from trends reported in mouse and human systems where m6A often correlates with transcript abundance. Several factors may explain this discrepancy. First, m6A regulates multiple layers of RNA metabolism, including translation efficiency and mRNA stability, which may not directly translate into steady-state transcript changes. Second, early embryogenesis is dominated by post-transcriptional regulation, particularly during MZT, where translational control may predominate over transcriptional output. Third, species-specific differences between bovine and other mammals may contribute to distinct m6A-expression relationships. Finally, methodological differences, including SAC-seq single-nucleotide resolution versus peak-based approaches, may also affect correlation patterns. Together, these findings suggest that m6A-mediated regulation during early embryogenesis extends beyond transcriptional control and involves complex, stage-specific post-transcriptional mechanisms.

To demonstrate gene- and site-specific functional effects, we selected RPL12, an RPG harboring a dynamically regulated m6A site essential for developmental progression. Both prime editing and dCas13Rx-ALKBH5-mediated demethylation at the A148 m6A site of RPL12 led to impaired ZGA and reduced blastocyst formation, demonstrating that loss of m6A at this single site is sufficient to compromise embryo development. Remarkably, supplementation with wild-type RPL12 mRNA failed to rescue this phenotype, suggesting that m6A regulates developmental competence not through transcript abundance, but likely through modulation of translation efficiency or ribosomal-binding protein recruitment, consistent with prior studies.7,43 Consistently, global translation assays showed that RPL12 m6A disruption or demethylation reduced protein synthesis, aligning with transcriptomic analyses demonstrating coordinated downregulation of ribosomal genes, translation factors, and metabolic regulators in edited embryos. This transcriptional dysregulation was evident as early as the 4-cell stage and persisted through the 8-cell stage, reinforcing the role of m6A as a temporal regulator of gene expression and protein synthesis during early embryogenesis. These findings support emerging models in which m6A regulates not only RNA stability and translation efficiency but also overall translational output during key developmental transitions.7,12,44,45 While HPG incorporation and transcriptomic analyses strongly suggest that m6A at RPL12 modulates global translation, we were unable to directly measure RPL12 protein levels or ribosome association due to limited material and the lack of suitable bovine-specific antibodies. Future studies using stage-specific RPL12 quantification and low-input polysome profiling will be essential to fully elucidate the mechanism by which this m6A site regulates translational output.

Importantly, this study provides functional evidence that a single m6A site can influence broader transcriptomic programs and contributes to early developmental failure. Early embryonic loss is a major cause of infertility in both humans and cattle. Although in vitro fertilization (IVF) is widely used to treat infertility, more than 50% of IVF embryos fail to progress to the blastocyst stage in both humans and cattle. Moreover, the developmental competence of IVF embryos after transfer remains lower than that of in vivo-derived embryos in cattle.46 Environmental stressors such as in vitro culture conditions can induce aberrant molecular changes in embryos, leading to developmental arrest.47 It is therefore possible that such stress also induces abnormal m6A patterns in IVF embryos during this highly plastic preimplantation period. Our findings suggest that identifying and functionally validating beneficial m6A sites may improve embryo competence and that targeted modulation of m6A in IVF embryos could represent a potential strategy to enhance fertility outcomes.

Overall, our study reveals a critical role for m6A RNA methylation in orchestrating molecular transitions during bovine preimplantation development. This work provides an unprecedented resource for exploring m6A regulation in oocyte maturation and preimplantation development.

Limitations of the study

Our experimental approach has several limitations. First, due to the limited availability of oocytes and preimplantation embryos, both RNA-seq and SAC-seq experiments were performed in duplicate biological samples. Although the datasets are of high quality and internally consistent, additional biological replicates could further strengthen robustness. Second, embryos were generated by IVF, which may introduce stress-associated epitranscriptomic changes compared with in vivo development. In addition, SAC-seq was performed on pooled embryos, which may obscure cell-to-cell and lineage-specific variability. Third, eTAM-seq validation showed differences in relative enrichment patterns for some candidate m6A sites compared with SAC-seq. However, both SAC-seq and eTAM-seq are relatively recent technologies, and neither provides 100% accuracy, bias-free quantification of m6A. Fourth, although this study provides a comprehensive m6A landscape in bovine preimplantation development, functional validation was primarily focused on RPL12. Future studies should investigate additional m6A-modified genes, including both RPL and RPS ribosomal subunit genes, and determine how perturbations of these regulatory layers contribute to early embryonic loss. Finally, the heterogeneous and dynamic patterns observed highlight that both the regulation and functional consequences of m6A are highly context-dependent and complex. Large-scale gene- and site-specific functional investigations will be required in future studies.