KM mutant highlights enhancers in minor ZGA

Before the first cell division, a fertilized embryo, also known as a zygote, contains 2 distinct nuclei called the paternal and maternal pronucleus (PN). During the zygotic period, the paternal PN undergoes massive epigenetic reprogramming characterized by removing protamines and establishing new epigenetic marks to acquire totipotency, whereas the epigenetic alteration of maternal PN is less dynamic. Reflecting this difference, the timing and levels of global gene transcription in the zygotic stage, called “minor zygotic gene activation (ZGA),” differ between paternal and maternal PNs,¹ suggesting that the mechanism of transcriptional regulation is asymmetric. Consistent with this idea, several epigenetic events occur uniquely in each PN. For instance, H3.3, an H3 variant that is incorporated into chromatin independently of DNA replication, is the best characterized in zygotes due to its preferential incorporation into paternal chromatin soon after fertilization. A few studies have demonstrated the critical role of H3.3 methylation in early embryonic development by overexpressing the lysine-arginine (K-R) substitutions: e.g., H3.3-K27R and H3.3-K36R, in which the 27^th and 36^th lysine residues (respectively) are not methylated.^2,3 However, unlike in yeast, genetic disruption of specific lysine residues in mammalian histones is impractical, since they are multiply encoded in the genome.

A recent discovery revealed that lysine-methionine (K-M) substitution of H3.3 not only abrogates methylation of substituted lysines transcribed from the mutated genome, but also inhibits global endogenous methylation of substituted lysines in non-mutated H3, including H3.1 and H3.2, by preventing the catalytic activity of SET-domain-containing methyltransferases.⁴ To take advantage of this “reverse-genetic” system, we employed 4 K-M mutants (H3.3-K4M, -K9M, -K27M, and -K36M) to re-evaluate the roles of histone methylation in early preimplantation development in mice, and uncovered a critical role of H3K4 methylation, which was missed when K-R mutants were used.⁵

Overexpression of H3.3-K4M in the early zygotic stage resulted in developmental arrest starting at the 8-cell stage. However, no such phenotype was observed when it was overexpressed in the late zygotic stage, which implies the crucial role of H3K4 methylation prior to the late zygotic stage. As expected, H3K4 methylations were significantly altered in paternal PN rather than in maternal PN, due to the paternal PN-specific incorporation of H3.3. Interestingly in H3.3-K4M expressing embryos, a 5-ethynyl uridine incorporation assay demonstrated reduced transcription levels specifically in the paternal PN, whereas a dominant transcriptional wave in 2-cell-stage embryos was maintained, suggesting that the growth interference in the H3.3-K4M mutant was caused by the alteration of minor, rather than major, ZGA in the paternal PN.

Although the H3.3-K4M mutant altered all states of H3K4 methylation in paternal PN, H3K4 monomethylation (H3K4me1) is established around the onset of minor ZGA. H3K4 trimethylation, however, is not observed until the late pronuclear stages, although it strongly correlates with transcriptional activation. This observation implies that H3K4me1 at enhancers, rather than H3K4me3 at promoters, plays a crucial role during minor ZGA, and makes Mll3/4 the most conceivable candidate for exhibiting H3K4 methyltransferase activity. Indeed, knockdown of Mll3/4 during minor ZGA by the siRNAs phenocopied H3.3-K4M in terms of developmental arrest appearing from the 4- to 8-cell stages, and attenuated H3K4me1 as well as transcription in the paternal PN. Importantly, siMll3/4 also impaired H3K27 acetylation in paternal PN. The depletion of p300, a co-factor of Mll3/4 for enhancer activation, in zygotes also caused developmental arrest starting from the 4- to 8-cell stages. Taken together, these findings strongly suggest the importance of enhancer activation in the onset of paternal minor ZGA through H3K4me1 by the Mll3/4 complex (Fig. 1).

During the zygotic period, some histone modifications in the paternal genome are quickly established to levels seen in the maternal genome, whereas others remain unmodified. However, it has been unclear whether and how this asymmetric epigenetic status influences transcriptional properties in minor ZGA. Our work using the H3.3-K4M mutant strongly suggests the involvement of enhancer activation in paternal minor ZGA. Interestingly, the injection of reporter plasmids into zygotes demonstrated distinct differences in transcription susceptibility between the paternal and maternal PN, with the former exhibiting higher transcription activity and fewer requirements of exogenous enhancers independent of chromatin structure.⁶ Thus, one could interpret that it is not the recruitment of transcription factors to promoters by opening the chromatin structure that acts as a trigger of paternal minor ZGA, but rather endogenous enhancer activation by newly incorporated H3.3 and the subsequent K4 methylation by Mll3/4. Furthermore, unlike the impairment of major ZGA, by which embryonic development was arrested in the 2- to 4-cell stages, both overexpression of H3.3-K4M and depletion of Mll3/4 independently caused the delayed phenotype beyond the 4- to 8-cell stages, even though the transcriptional alteration specifically occurred in the 1-cell stage. This phenomenon may have been observed because the effect is restricted only to the paternal PN. These observations may support the idea of a “stepwise” gene activation model, in which one transcriptional wave triggers following transcriptional waves during preimplantation development.⁷

Figure 1.

After fertilization, a paternal PN gradually establishes new epigenetic marks. In the middle pronuclear stage, newly synthesized Mll3/4 introduces H3K4me1 predominantly in paternal PNs, and it triggers transcriptional activation corresponding to minor ZGA from paternal genome. Since Mll3/4 preferentially mono-methylates H3K4 at enhancers rather than promoters, it suggests the importance of enhancer activation in the onset of paternal minor ZGA.M