The new study, published in Frontiers in Animal Science, examined Duolang sheep, an indigenous breed from the Xinjiang region prized for its early sexual maturity, year-round estrous cycles, and relatively high reproductive output. First estrus in this breed typically arrives at around five months of age, making the animals an attractive model for dissecting the molecular events that accompany the onset of fertility. The researchers focused on ESR2, the gene encoding estrogen receptor beta, a protein abundant in ovarian granulosa cells that mediates estrogen’s local effects on follicular development, granulosa cell differentiation, and ovulation. When ESR2 function is lost in experimental animals, granulosa cells respond poorly to gonadotropins, follicular maturation falters, and ovulation is impaired.

The logic behind the study rests on a well-established principle of epigenetics. DNA methylation, the addition of methyl groups to cytosine bases within CpG dinucleotides, tends to cluster in promoter regions called CpG islands. When these islands are heavily methylated, transcription factors may struggle to bind and chromatin may become less accessible, generally suppressing gene transcription. Because epigenetic regulation has already been implicated in the developmental activation of the female reproductive axis, the team asked a straightforward question: does methylation of the ESR2 promoter change as Duolang ewes pass through puberty, and do those changes track the gene’s activity?

To find out, the researchers collected ovarian tissue from fifteen healthy ewes divided into three groups of five: prepubertal animals at ninety days of age, pubertal animals sampled within twenty-four hours of their first observed estrus, and early postpubertal animals sampled within seven days after estrous behavior subsided. Puberty was defined behaviorally, by voluntary acceptance of mounting, restlessness, increased vocalization, vulvar swelling, and mucous discharge during daily ram-exposure checks. The team first cloned and sequenced a 2,871-base-pair stretch of the ESR2 promoter, identifying four candidate transcription start sites, a single TATA box, sixty-seven predicted transcription factor binding sites, and two CpG islands. They then used bisulfite sequencing, a technique that converts unmethylated cytosines to uracil while leaving methylated cytosines intact, to measure methylation at nine sites in the first CpG island and twenty-two sites in the second.

The results revealed a striking asymmetry between the two regulatory regions. Methylation of CpG island 1 declined steadily across the three stages, falling from 82.22 percent in prepubertal ovaries to 78.71 percent at puberty and 71.11 percent in the early postpubertal period, an absolute drop of more than eleven percentage points. CpG island 2 behaved differently, dipping from 79.09 percent to 74.45 percent at puberty before rebounding to 79.82 percent afterward, a level significantly higher than the pubertal value. Meanwhile, ESR2 messenger RNA expression climbed progressively: significantly higher at puberty than before it, and significantly higher still after puberty than at either earlier stage.

The correlation analysis sharpened the picture. In the prepubertal group, methylation of CpG island 1 showed only a weak, statistically insignificant relationship with ESR2 expression. But in the pubertal and postpubertal groups, the correlation turned strongly negative, with coefficients of -0.956 and -0.969 respectively, meaning that lower methylation was consistently paired with higher gene expression. At the level of individual CpG sites, the story grew more nuanced: some positions within each island showed nominal associations with expression in one direction, others in the opposite direction, and the average methylation of an island did not always capture what was happening at specific sites. The authors are careful to note that these site-level findings rest on unadjusted P values and should be treated as candidate associations rather than confirmed regulatory elements.

To probe whether demethylation could actually drive ESR2 expression, the team turned to cultured ovarian granulosa cells, verified as more than 95 percent pure by immunofluorescent staining for the follicle-stimulating hormone receptor. Treating the cells with 0.5 micromolar 5-aza-2′-deoxycytidine, a drug that inhibits DNA methyltransferases and strips methyl groups genome-wide, significantly increased ESR2 mRNA expression after thirty-six hours and boosted cell viability at later time points. On its face, the experiment seems to close the loop: remove methylation, and the gene wakes up.

But the researchers are refreshingly candid about the limits of that inference. Because 5-Aza-dC acts globally rather than on the ESR2 promoter specifically, and because promoter methylation was not remeasured after treatment, the rise in expression cannot be attributed directly to demethylation of either CpG island. Nor was the ERβ protein itself quantified, leaving open whether the transcriptional change translated into more receptor. The authors also acknowledge a series of constraints on the tissue study: only five animals per stage, behavioral rather than endocrine confirmation of pubertal status, just five bisulfite-sequenced clones per animal per island, and whole-ovary samples in which shifts in the proportions of granulosa, theca, stromal, and luteal cells could masquerade as regulatory changes. The expression stage-pattern, they note, may equally reflect changes in follicular composition and the ovarian endocrine environment rather than a causal role for ESR2 in triggering puberty.

Context from other species both supports and complicates the interpretation. In Holstein cows, prenatal heat exposure has been linked to site-specific methylation of the ESR2 untranslated region and altered expression of ovarian steroid receptors in adult dominant follicles, hinting that this gene’s epigenetic state is sensitive to developmental conditions. An integrated methylome and transcriptome analysis of Hu sheep ovaries found a negative relationship between methylation near transcription start sites and gene expression, and large-scale demethylation has been observed in proliferating mouse granulosa cells during follicular development. Genome-wide studies of gilt ovaries and goat hypothalami, however, show that methylation changes during reproductive maturation are tissue-specific, region-specific, and stage-dependent, meaning the declining pattern seen in Duolang CpG island 1 should not be assumed to be a universal signature of puberty.

What the study ultimately delivers is a well-mapped starting point rather than a mechanism. It establishes that the ESR2 promoter carries dynamic, stage-specific methylation patterns in a breed defined by early fertility, that one of those patterns inversely tracks gene expression precisely when the ovary is transitioning to reproductive activity, and that pharmacological demethylation is associated with increased ESR2 transcription in granulosa cells. The candidate CpG sites identified here, particularly those within island 1 that correlated negatively with expression, now form a shortlist for promoter-reporter assays, targeted methylation editing, and transcription-factor binding experiments. If those follow-up studies hold up, epigenetic markers of reproductive maturation could eventually inform selection programs aimed at breeding sheep that reach puberty sooner and reproduce more prolifically, turning a chemical footnote on DNA into a practical tool for livestock genetics.

Subject of Research: ESR2 promoter DNA methylation and mRNA expression in sheep ovaries during puberty onset

Article Title: ESR2 promoter methylation and mRNA expression patterns in Duolang sheep ovaries across pubertal stages

Article References: Gu, R., Liu, L., Wu, L., Wang, M., & Xing, F. (2026). ESR2 promoter methylation and mRNA expression patterns in Duolang sheep ovaries across pubertal stages. Frontiers in Animal Science, 7, Article 1932991. https://doi.org/10.3389/fanim.2026.1932991

Image Credits: AI Generated

DOI: 10.3389/fanim.2026.1932991

Keywords: DNA methylation, ESR2, estrogen receptor beta, puberty, Duolang sheep, ovary, granulosa cells, epigenetics, CpG islands, bisulfite sequencing, reproductive maturation, 5-Aza-dC

The debate began when Chiapperino and Paneni highlighted a fundamental technical obstacle: epigenetic studies are largely unable to incorporate fine-grained mapping data of an individual’s surrounding social milieu. The same limitation, they noted, afflicts genomics. Zaina accepts this diagnosis and praises the authors for elevating questions of scientific epistemology into public debate. His response, however, goes a step further. He argues that because epigenetic marks are tightly coupled to DNA sequence, epigenetics is structurally doomed to reproduce the very shortcomings of genomics that Chiapperino and Paneni identified. If the sequence sets the stage for most methylation patterns, then no amount of interdisciplinary ingenuity will allow epigenetics to serve as a clean molecular record of social experience.

To understand why, Zaina insists, one must revisit the history of the concept. The term epigenetics traces back to Conrad Waddington’s definition of epigenesis in the 1940s, a description of molecular development that, in Zaina’s view, sits awkwardly with the modern scientific community’s intuition of what the field does. Epigenetics remained poorly understood until genomic imprinting revived it, because DNA methylation offered a mechanism to explain that clearly non-genetic phenomenon. The next breakthrough came with the demonstration that epigenetic marks can be modified by exogenous factors, notably through studies showing that early nutrition could alter methylation of transposable elements. From these milestones emerged the definition now taught in textbooks: a DNA sequence-independent, environmentally sensitive mode of gene regulation, widely adopted as a tool to hunt for markers of non-communicable diseases, particularly in light of genetics’ notorious missing heritability problem.

That definition, Zaina argues, can be dismantled with the field’s own data. The decisive evidence comes from methylation quantitative trait loci, or methylation QTL, first documented by Tycko’s group in 2008. These are genetic variants that predict methylation states at nearby or distant sites, and subsequent surveys show they account for the vast majority of DNA methylation variation in humans. A comprehensive analysis of human tissues found roughly twice as many methylation QTL as expression QTL colocalizing with genetic variants identified by genome-wide association studies. Even a textbook example of sequence dependence has long been available but rarely stated: mammalian CpG islands, the GC-rich regions surrounding many gene promoters, remain generally unmethylated precisely because of their peculiar sequence composition. The genetic grip on epigenetics is further tightened by methodology. Mendelian randomization, which exploits genetic variants as instruments, is currently the best available tool for assessing whether any given DNA methylation profile causally influences a phenotype, a dependence on genetics that Zaina finds telling.

There is also a quantitative prediction that follows from this tight coupling. If methylation is largely sequence-driven, then individual differentially methylated CpG sites should exert comparatively small phenotypic effects, echoing the missing heritability problem that has frustrated geneticists for decades. The available evidence, Zaina notes, suggests that this prediction is correct. He also flags a conceptual confusion at the heart of the field’s self-description: independence from changes in DNA sequence is often cited as the main difference between epigenetics and genetics, yet genetics deals with sequence variation in populations rather than sequence changes, mutations being a relatively minor topic. The distinction, in other words, is less clean than the standard rhetoric implies.

The second pillar of the popular narrative, the idea that the epigenome is an open book on which the environment can freely write, erase, or rewrite information, fares no better in Zaina’s assessment. He points to the controversy surrounding a famous New Yorker article that proposed epigenetics as the mechanism underlying phenotypic divergence between identical twins, an interpretation that drew sharp criticism from prominent epigeneticists. Even more contentious is the hypothesis of epigenetic transgenerational inheritance in humans, the notion that methylation patterns altered by a grandparent’s environment could be transmitted across generations. Reviews of the evidence describe it as highly controversial, with mechanisms in mammals poorly established and human data difficult to disentangle from genetic and cultural confounding.

What remains, once these assumptions are stripped away? In Zaina’s formulation, epigenetics describes a highly sequence-dependent and comparatively weakly environmentally sensitive mode of transcriptional regulation, functionally closer to the activity of transcription factors or DNA-binding long non-coding RNAs than to a free-floating environmental ledger. He even suggests the field might be more accurately renamed paragenetics. Adding micro-RNAs and non-coding RNAs to the definition, he argues, has only deepened the confusion. The counterfactual he poses is striking: had methylation QTL been documented forty years ago rather than in 2008, the current enthusiasm for epigenetics might never have taken hold. Had the higher affinity of DNA methyltransferases for RNA relative to DNA been appreciated earlier, textbooks might describe those enzymes as RNA-binding proteins that modify DNA chemistry.

The stakes of this conceptual housekeeping extend well beyond taxonomy. Zaina’s central concern is that biosocial epigenetics, the quest to find molecular signatures of adverse social environments, whether politically or economically adverse, is likely to uncover genetics-related phenomena rather than pure environmental imprints. Researchers searching for DNA methylation profiles written exclusively by an adverse milieu to explain how social circumstances create and maintain disadvantaged groups will, on his reading, inevitably bump into sequence-dependent variation. Such findings would be highly controversial if they appeared to clash with dominant views in identity politics or social constructivism. The devils of genetics when applied to social sciences and economics are obvious, he writes, but hidden ones may lurk in epigenetics, precisely because the field’s environmental framing can mask its genetic dependencies.

Zaina is careful not to dismiss the enterprise entirely. Reducing the daylight between epigenetics and fine-grained mapping of the social milieu remains, in his words, a noble and useful pursuit, and he credits Chiapperino and Paneni with correctly identifying the problem and acknowledging its difficulty even for the best interdisciplinary research and technological advances. But he insists that the hurdles will only be overcome after a correct understanding of the essence of epigenetics, however dynamic that essence proves to be. His closing question is deliberately uncomfortable: this territory feels remote from pipettes and tubes, and scientists must decide whether they should walk that far at all. The reply, published open access in Epigenetics Communications, ensures that the debate over whether epigenetics can ever truly bridge biology and society will continue, and that any future biosocial science will have to reckon with the sequence it inherits along with the environment it measures.

Subject of Research: The debate over whether epigenetics can function as a biosocial science given its dependence on DNA sequence

Article Title: Will epigenetics ever be a biosocial science? A reply to Chiapperino and Paneni

Article References: Zaina, S. (2023). Will epigenetics ever be a biosocial science? A reply to Chiapperino and Paneni. Epigenetics Communications, 3(1), Article 2. https://doi.org/10.1186/s43682-023-00018-0

Image Credits: AI Generated

DOI: 10.1186/s43682-023-00018-0

Keywords: epigenetics, DNA methylation, methylation QTL, biosocial science, genomics, social epigenomics, transgenerational inheritance, Mendelian randomization, gene regulation, missing heritability, Waddington, CpG islands