Epigenetics meets metabolomics

posted in: Miscellaneous


Read our newest paper “Deep molecular phenotypes link complex disorders and physiological insult to CpG methylation” in Human Molecular Genetics (published online on 08 Jan 2018) … full blog page coming soon! This paper focuses on 20 CpG sites reported in our  Petersen et al., 2014 paper (see blog below) with a much broader multi-omics panel.

An epigenome-wide association study with blood serum metabolic traits

We now know over 150 loci of genetically influenced metabotypes (GIMs) and a correspondingly large number of mQTLs (see this post). Each of these loci represent the outcome of an experiment conducted by nature, and many of them provide new biomedical insights that can help to better understand complex diseases (see this post)

But how about associations between epigenetic marks and metabolomics?

Some time ago we published the first epigenome-wide association study between DNA CpG methylation and metabolomics [Petersen et al., 2014], testing the methylation state of 450,000 CpGs for association with metabolomics data from three platforms: Metabolon, Biocrates and Lipofit (now numares).

It is worth revisiting some of the associations from this study and putting them into context with more recent papers.

After elimination of associations that were confounded by a genetic variant, Petersen et al. identified about 20 purely epigenetic associations with metabolomics that stood out (see Table below).

Three families of CpG associations are particularly noteworthy:

  1. Diabetes: three CpGs, located near TXNIP, ABCG1, and CPT1A, are showing a diabetes-specific metabolic phenotype (see Table 4 in Al Muftah et al.);
  2. Smoking: a second set of CpGs, led by a CpG near AHRR, is associated with smoking and is marked by the lead association with 4-vinylphenol sulfate (4-vs) (see Zaghlool et al.);
  3. Steroid-related: a third set  of CpGs is associated with the steroid 4-androsten-3beta, 17beta-diol disulfate (A-diol) – the common factor behind these associations has not been revealed (the likely candidates age and gender were corrected for in the association).

The top metabotypes from Petersen et al. that associate with these CpGs are listed in the table below – but also check out the Supplementary Tables of Petersen et al., 2014. These tables provide a wealth of further information and provide further CpG associations that are worth investigating.

The hunt is open for further epigenome-metabolome associations. Please mail me if you find new ones!

Locus nameCpGChrPosMetabolic traitNBeta’r2P-value
DHCR24cg17901584155353706PC ae C36:517804.0010.0363.65E−18
ABCG1cg065001612143656587SM C16:01781−0.8170.0081.04E−14



Petersen AK, Zeilinger S, Kastenmüller G, Römisch-Margl W, Brugger M, Peters A, Meisinger C, Strauch K, Hengstenberg C, Pagel P, Huber F, Mohney RP, Grallert H, Illig T, Adamski J, Waldenberger M, Gieger C, Suhre K, Epigenetics meets metabolomics: an epigenome-wide association study with blood serum metabolic traits, Human Molecular Genetics, 23:534-545, 2014 [PubMed].

Al Muftah WA, Al-Shafai M, Zaghlool SB, Visconti A, Tsai PC, Kumar P, Spector T, Bell J, Falchi M, Suhre K, Epigenetic associations of type 2 diabetes and BMI in an Arab population, Clin Epigenetics, 8:13, 2016 [PubMed].

Zaghlool SB, Al-Shafai M, Al Muftah WA, Kumar P, Falchi M, Suhre K, Association of DNA methylation with age, gender, and smoking in an Arab population, Clin Epigenetics, 7:6, 2015 [PubMed].