If you’ve been following nutrition studies hoping for diet advice, you may have noticed it’s complicated. First fat is bad, then it’s just saturated fat, then just trans fat, then actually it’s sugar. This may cause some to throw their hands up in frustration (grabbing a Twinkie from the top shelf while they’re up there), but as an Epigenie reader, you’re probably hoping gene variants and epigenetics may help cut through the fat (saturation status notwithstanding).
If so, we have both good news and bad news courtesy of a new epigenetics study led by Michelle Holland, Robert Lowe, and Vardhman Rakyan in London. The bad news is confirmation of our intro: it’s complicated. The good news: recognizing that different gene variants may not be equally susceptible to epigenetic effects “could [explain] why we’ve only so far been able to explain a small fraction of the heritability of many health conditions, which makes a lot of sense in the context of metabolic diseases, such as type 2 diabetes.”
The How: Whole-Genome Methylation Analysis
The team was interested in how a mother’s diet influences her children’s epigenetics (developmental programming), so they fed mother mice one of two diets – either protein restricted (8% protein) or control (20% protein) – from conception through nursing. They only studied male pups, who all dined upon the normal control diet from the time they could down solid food.
After 16-20 weeks, the male offspring were sacrificed, and the team investigated their genome methylation with two techniques. First, they did whole-genome methylation analysis with reduced representation bisulfite sequencing (RRBS), then they analyzed target regions in-depth by amplifying them with bisulfite PCR and reading the products with high-throughput sequencing (BisPCR-seq).
The What: Diet Induces Correlations Between Genome and Epigenome
Protein restriction during early development had two clear effects:
- DNA hypermethylation at rDNA
- Lower body weight at weaning
Weirdly, protein restriction also induced interesting correlations, which were not seen in pups of moms on the control diet. To understand these, some quick background is in order. First, mice (like humans) have hundreds of copies of rDNA, and these have an important SNP (either A or C) 104 bp upstream of the transcriptional start site. Second, there is a CpG methylation site 133 bp upstream, and methylation there (which was higher for the A variants), silences the rRNA gene. Now for the weird results:
- CpG-133 methylation was negatively correlated with body weight, but only methylation at -104A variants, and only for the protein-restricted pups.
- CpG-133 methylation of A variants (but not C variants) was positively correlated with the overall percentage of A variants, and only for the protein-restricted pups.
- The percentage of variants that were both -104A and unmethylated was positively correlated with rRNA expression.
Lest we suspect this result was a one-hit wonder, the team re-analyzed data from two previous maternal diet studies, showing that mice on either high-fat or obesogenic diets had pups with hypermethylation of CpG-133, but only at -104A variants.
The Huh? Hinting At A Role For Regulatory RNA
Ok, so maternal diet affects a pup’s rDNA methylation, gene expression, and body weight, but only for -104A copies of the rDNA, and the higher the percentage of -104A alleles, the larger the effect. This is surprising, because it suggests that diet, at least in this case, only affects epigenetics if the pup is “primed” by lots of A allele copies. How would this work? One clue may lie in a brief aside.
A regulatory noncoding RNA is expressed from unsilenced rDNA copies, and it recruits DNA methyltransferases to silenced copies during S phase, which helps their methylation be inherited through cell division. Could it be that this regulatory RNA is only expressed from – or effective at – A alleles, making this RNA-based, epigenetic inheritance? Did the authors test that? Is that next paper already in review? We don’t know, but depending on when you’re reading this, the Related Posts below may or may not be correlated with an answer.
Don’t let your own diet be Science-restricted! Read your fill over at Science, 2016.