Maternal behavioral and metabolic phenotypes have a much greater effect on gene expression (and the cellular "ecosystem" of their offspring) than other epigenetic mechanisms. See:

Archer E, Lavie CJ, Dobersek U, et al. Metabolic Inheritance and the Competition for Calories between Mother and Fetus. Metabolites 2023;13(4):545.

Archer E, Lavie CJ. Obesity Subtyping: The Etiology, Prevention, and Management of Acquired versus Inherited Obese Phenotypes. Nutrients 2022;14(11):2286.

Archer E, Lavie CJ. The failure of gene-centrism. Behavioral and Brain Sciences 2023;46:e209. doi: 10.1017/S0140525X22002436 [published Online First: 2023/09/11]

Archer E. The Childhood Obesity Epidemic as a Result of Nongenetic Evolution: The Maternal Resources Hypothesis. Mayo Clinic Proceedings 2015;90(1):77-92. doi: 10.1016/j.mayocp.2014.08.006

Thanks Ed. My original question is not necessarily limited to metabolic pathways, but it fits. Your work integrates a large number of concepts involving maternal and embryonic energy budgets, and provides a good example where we could define the relative importance of epigenetics within the larger field of nongenetic evolution. Focusing on the role of glucose, a hypothesis posed in your work is that gestational hyperglycemia provides a signal to embryonic adipocytes that become either too numerous (adipogenesis) or too competitive (anabolism). Either explanation invokes phenotypes maintained by differential gene expression between offspring of hyperglycemic and normal mothers.

What is the subcellular signal preceding these changes in gene expression? If we take glucose as a regulator of gene expression, there are known and presumed loci where glucose interacts directly with DNA. Surely there are epigenetic interactions at these target sites, such as methylation or histone acetylation, which could mediate individual response to maternal hyperglycemia. Given this metabolic example, I'm specifically interested in the structural conformation of these sites on the genome. This doesn't discount glucose or some metabolite as a source signal, but seeks to characterize which class of structural modification is important for mediating the individual response to hyperglycemia.

Do we have enough information to generalize whether methylation or histone acetylation are more important contributors to epigenome heritability? I'm still not sure.

Interesting question (the original one), and the best answer can be found in this publication:https://academic.oup.com/eep/article/8/1/dvac001/6529222

DNA methylation is not the only epigenetic mechanism able to transgenerationally pass.