Ssions [35]. The required period of vernalization (for Arabidopsis accessions requiring this cue to trigger flowering) is typically much IQ1 longer that the moist chilling period required for Arabidopsis seed dormancy breakage. Yet the CASIN cost similarities between the two processes are striking. It is thus interesting to speculate that a similar polymorphism-based mechanism is the underlying basis for some of the variations in cold-requirements for dormancy breakage, and for the huge variation in the depth of seeddormancy. The mechanisms observed in the context of vernalization effects at the chromatin level certainly seem comparable to the gradual accumulation of H3K27me3 and simultaneous reduction of H3K4me3 that we observed on seed dormancy regulators during moist chilling of seeds. It is conceivable that germination competence is reached at a certain threshold ratio of active to repressive marks on dormancy regulators, and that germination proceeds only when genes specifying positive germination regulators are activated. This would provide a quantitative means to measure the extent of cold exposure as a result of the output of the ratio of activating and repressive marks. Interestingly, we detected both histone marks at the corresponding loci for maturation/dormancy-related genes in our sampling population in non-dormant seeds before their transfer to germination conditions (Figs. 4 and S2). H3K27me3 thus gradually replaces H3K4me3 on the major dormancy regulators, until no or very little H3K4me3 remains detectable in seedlings. It is however unlikely that these would represent so-called bivalent marks, which are found in mammalian embryonic stem cells [38], as evidence 18297096 for this is scarce in plants [39]. Very little is known about the mechanistic aspects of the interactions between different histone modifications in plants. By combining our findings with H3K4me3 profiling in prc2 mutants [13], we found that this mark needs to be replaced by H3K27me3 in order to be removed as many of the dormancy regulators that we investigated such as ABI3, DOG1, and FLC, still show H3K4me3 in seedlings upon loss of H3K27me3 [13] (Fig. S3). This is remarkable as only a small portion of PRC2-target genes in the genome show this gain in H3K4me3 upon loss of PRC2, suggesting that its activity is necessary to replace the activating mark and effect the termination of gene expression. PRC2-mediated dormancy control appears to take place at the level of the embryo, as seeds with homozygous PRC2-defective endosperm but heterozygous embryos exhibit germination behavior that is indistinguishable from that of wild-type seeds [13]. Therefore, we expanded our nChIP analyses to isolated embryos from seeds that had been exposed to the dormancy-terminating treatment (14 d of moist chilling). The histone profiles of the embryos strongly resembled those of whole seeds (Fig. S4). Therefore the dynamic change from the activating to the repressive state very likely takes place in the embryo during dormancy breakage.Changes in Histone Methylation of the ABI3 Gene, a Major Regulator of Life Cycle Transitions, are Evolutionarily ConservedHaving found that major dormancy regulators such as ABI3 are transcriptionally regulated at the chromatin level in Arabidopsis, we asked whether the same is true for an evolutionarily distant species, the gymnosperm yellow-cedar (Callitropsis nootkatensis). The seeds of this conifer species are deeply dormant at maturity and upon dispersal they typica.Ssions [35]. The required period of vernalization (for Arabidopsis accessions requiring this cue to trigger flowering) is typically much longer that the moist chilling period required for Arabidopsis seed dormancy breakage. Yet the similarities between the two processes are striking. It is thus interesting to speculate that a similar polymorphism-based mechanism is the underlying basis for some of the variations in cold-requirements for dormancy breakage, and for the huge variation in the depth of seeddormancy. The mechanisms observed in the context of vernalization effects at the chromatin level certainly seem comparable to the gradual accumulation of H3K27me3 and simultaneous reduction of H3K4me3 that we observed on seed dormancy regulators during moist chilling of seeds. It is conceivable that germination competence is reached at a certain threshold ratio of active to repressive marks on dormancy regulators, and that germination proceeds only when genes specifying positive germination regulators are activated. This would provide a quantitative means to measure the extent of cold exposure as a result of the output of the ratio of activating and repressive marks. Interestingly, we detected both histone marks at the corresponding loci for maturation/dormancy-related genes in our sampling population in non-dormant seeds before their transfer to germination conditions (Figs. 4 and S2). H3K27me3 thus gradually replaces H3K4me3 on the major dormancy regulators, until no or very little H3K4me3 remains detectable in seedlings. It is however unlikely that these would represent so-called bivalent marks, which are found in mammalian embryonic stem cells [38], as evidence 18297096 for this is scarce in plants [39]. Very little is known about the mechanistic aspects of the interactions between different histone modifications in plants. By combining our findings with H3K4me3 profiling in prc2 mutants [13], we found that this mark needs to be replaced by H3K27me3 in order to be removed as many of the dormancy regulators that we investigated such as ABI3, DOG1, and FLC, still show H3K4me3 in seedlings upon loss of H3K27me3 [13] (Fig. S3). This is remarkable as only a small portion of PRC2-target genes in the genome show this gain in H3K4me3 upon loss of PRC2, suggesting that its activity is necessary to replace the activating mark and effect the termination of gene expression. PRC2-mediated dormancy control appears to take place at the level of the embryo, as seeds with homozygous PRC2-defective endosperm but heterozygous embryos exhibit germination behavior that is indistinguishable from that of wild-type seeds [13]. Therefore, we expanded our nChIP analyses to isolated embryos from seeds that had been exposed to the dormancy-terminating treatment (14 d of moist chilling). The histone profiles of the embryos strongly resembled those of whole seeds (Fig. S4). Therefore the dynamic change from the activating to the repressive state very likely takes place in the embryo during dormancy breakage.Changes in Histone Methylation of the ABI3 Gene, a Major Regulator of Life Cycle Transitions, are Evolutionarily ConservedHaving found that major dormancy regulators such as ABI3 are transcriptionally regulated at the chromatin level in Arabidopsis, we asked whether the same is true for an evolutionarily distant species, the gymnosperm yellow-cedar (Callitropsis nootkatensis). The seeds of this conifer species are deeply dormant at maturity and upon dispersal they typica.