Obesity before pregnancy may leave molecular marks in egg cells that influence metabolism in later generations, according to new research that combines genome-wide profiling, multigenerational mouse experiments, targeted epigenetic editing and analysis of human oocytes.
The study, published in Nature Metabolism on 2 October 2026, goes beyond showing an association between maternal obesity and offspring health. By deliberately adding methylation at specific sites in mouse oocytes, the researchers found that individual epigenetic changes could reproduce metabolic abnormalities in descendants. The effects were particularly evident in female offspring.
The findings add mechanistic detail to a long-running question in developmental biology: how can a mother’s metabolic state before conception influence offspring after the embryo undergoes extensive epigenetic reprogramming?
Searching the egg cell for a molecular memory
Researchers led by teams at Nanjing Medical University and collaborating institutions studied mice exposed to a high-fat diet and compared their mature oocytes with those from control animals. Genome-wide analysis identified 940 differentially methylated regions between oocytes from the normal-diet and high-fat-diet groups.
DNA methylation is a chemical modification of DNA that can help regulate whether genes are more or less active without changing the underlying genetic sequence. Because mammalian embryos erase and rebuild much of their DNA methylation after fertilisation, persistent effects from an egg cell present a biological puzzle.
The researchers focused on metabolic genes whose methylation and expression patterns remained abnormal across generations and tissues. Three loci became especially important: Hnf1a, Thra and Pdk4. These genes are involved in metabolic regulation, including glucose production, thyroid-hormone signalling and the control of glucose oxidation.
The abnormalities were not simply copied unchanged through every developmental stage. Methylation at the affected regions was largely erased during early embryonic development, as expected, but later reappeared in relevant tissues and in the female germline. This led the authors to investigate whether another chromatin signal might preserve information through the period when DNA methylation itself disappears.
Methylation disappeared, but a chromatin signal remained
At the affected loci, the team detected enrichment of the histone modification H3K36me2 even while the abnormal DNA methylation had been erased. Histones are proteins around which DNA is packaged, and chemical modifications to them can influence how genomic regions are organised and regulated.
The persistence of H3K36me2 provides a possible explanation for how methylation could later be re-established at the same genomic locations. The study therefore supports a model in which intergenerational information is not necessarily carried by an unbroken methylation mark. Instead, another feature of chromatin may help guide the later reconstruction of that mark.
That distinction matters because it avoids a simplistic interpretation of epigenetic inheritance. The results do not suggest that all methylation changes in an obese mother’s egg cells escape embryonic reprogramming. Rather, the evidence points to selected loci that can be re-established in a tissue-specific and sex-specific pattern.
Editing egg cells tested whether the marks could cause metabolic changes
Observational epigenetic differences cannot by themselves establish cause and effect. To address this, the researchers used a targeted dCas9-Dnmt3a system to increase methylation at selected regions in mouse oocytes. This allowed them to ask whether recreating a specific obesity-associated methylation pattern, without reproducing the entire maternal metabolic environment, could alter offspring physiology.
Targeted hypermethylation of the Hnf1a region led to increased hepatic gluconeogenesis in female offspring. Gluconeogenesis is the process through which the liver produces glucose, and excessive activity can contribute to poor metabolic control. Hyper-methylating the Pdk4 region impaired glucose tolerance in female offspring.
Importantly, these changes were not simply reflections of heavier animals. Female mice with targeted Pdk4 methylation showed no significant difference in body weight or insulin sensitivity at 20 weeks compared with controls, while female mice with targeted Hnf1a methylation showed no significant difference in body weight or fasting glucose. The metabolic phenotypes were therefore more specific than a general increase in obesity.
The sex difference was also notable. In male offspring, targeted methylation at the same loci did not produce equivalent metabolic abnormalities. Separate analyses found no significant methylation differences at the key Hnf1a, Thra and Pdk4 regions in relevant tissues from male offspring of the maternal high-fat-diet model.
Some effects extended into later generations
The team followed selected methylation patterns beyond the first generation. For Hnf1a, altered methylation persisted through the female germline and was detected in liver tissue of third-generation female descendants. The researchers reported strongly altered Hnf1a expression in these animals, with a reported P value below 0.0001.
In a separate experiment, females produced from oocytes with targeted Pdk4 hypermethylation were bred with control males. Their female offspring retained site-specific methylation differences in skeletal muscle and showed altered responses during glucose-tolerance testing. The F2 glucose experiment included 15 mice per group.
These multigenerational results strengthen the argument that the altered oocyte state can influence descendants rather than merely accompanying a single pregnancy. They do not, however, mean that every consequence of maternal obesity is transmitted epigenetically or that such effects are inevitable.
Human oocytes showed related methylation differences
The researchers also analysed oocytes from women with obesity. They detected increased methylation at regions corresponding to HNF1A and THRA, providing evidence that some of the molecular patterns identified in mice are conserved in humans.
This human component is important but should be interpreted cautiously. The causal experiments were performed in mice, not people. Finding similar methylation patterns in human oocytes does not demonstrate that those marks cause diabetes, obesity or other metabolic disease in children. Human metabolic health is shaped by genetics, nutrition, socioeconomic conditions, pregnancy biology, childhood environments and many other influences that cannot be reproduced by a controlled mouse experiment.
Why the preconception period may matter
The study shifts part of the developmental-health discussion to a period before pregnancy begins. If metabolic conditions can modify oocytes in biologically consequential ways, preconception health may be relevant not only to fertility and pregnancy outcomes but also to some aspects of offspring metabolic programming.
That implication should not be turned into individual blame. The work is mechanistic research, predominantly in mice, and it does not establish that changing body weight before conception will erase specific methylation marks or prevent disease in human children. It instead identifies molecular pathways that could eventually be investigated as intervention targets.
A particularly important contribution is the targeted editing experiment. Many studies report epigenetic differences associated with parental exposures, but association can arise because methylation is a consequence rather than a cause of altered physiology. By recreating selected methylation changes in oocytes and observing specific metabolic outcomes, this study provides unusually direct evidence that defined germline methylation states can participate in programming later metabolism.
Important limitations remain
The strongest causal evidence comes from laboratory mice, so translation to humans remains uncertain. Mouse high-fat-diet models simplify a condition that is far more heterogeneous in people, and targeted epigenetic editing is an experimental intervention rather than a naturally occurring exposure.
The effects were also locus-specific, tissue-specific and strongly sex-biased. This makes broad claims about a universal epigenetic mechanism inappropriate. The persistence of H3K36me2 is consistent with a molecular memory mechanism, but further work is needed to establish exactly how it guides methylation re-establishment after embryonic reprogramming.
Finally, the human analysis demonstrates conservation of selected methylation differences rather than intergenerational causality. Longitudinal human studies would be needed to connect preconception oocyte methylation with later metabolic outcomes in offspring while accounting for the many genetic and environmental factors that operate after conception.
Source Information
Study: Han, L., Wu, Y., Li, J. et al. “Maternal obesity imprints methylation marks in oocytes to drive intergenerational metabolic dysfunction.” Nature Metabolism (2026).
Published: 2 October 2026
DOI: 10.1038/s42255-026-01617-6
Research approach: Genome-wide and locus-specific DNA methylation profiling, multigenerational mouse experiments, gene-expression analysis, chromatin profiling, targeted oocyte methylation editing, metabolic phenotyping and comparative analysis of human oocytes.
