Body fat distribution is a predictor of metabolic and cardiovascular disease independent of body mass index (BMI) and is heritable. To understand the genetic determinants of the differences observed in fat distribution in the general population, genome-wide association studies (GWAS) have emerged as an extremely useful approach and have convincingly associated many single nucleotide polymorphisms (SNPs) with body fat distribution. Unfortunately, it has proven very difficult to link specific genes or genetic variants with these findings, reducing their translational impact. However, massive increases in the size of human genetic studies have increased the power of studies to confidently link rare missense (coding) variants with a range of phenotypes. In collaboration with colleagues in the MRC Epidemiology Unit in Cambridge, we recently identified rare missense variants in ALK7 (p.I195T and p.N150H), CALCRL (p.L87P), PLIN1 (p.L90P) and PDE3B (p.R783X) which were associated with a favourable body fat distribution and with protection from cardiometabolic disease, while a variant in gene PNPLA2 (p.N252K) was associated with an adverse phenotype (1).

To understand the roles of these genes, initial loss-of-function studies were carried out using siRNA-mediated knockdown in 3T3-L1 (pre)adipocytes, as all the genes are expressed in adipose tissue and have been at least loosely implicated in the regulation of intracellular lipolysis. In this cell model, Plin1 depletion resulted in an increase in basal lipolysis while isoproterenol stimulated lipolysis was suppressed. Pnpla2 depletion, on the other hand, impaired lipolysis under both basal and stimulated conditions. Alk7 or Calcrl knockdown resulted in a consistent decrease in lipolysis while the impact on adipogenesis was variable. Depletion of Pde3b did not have any impact on adipogenesis or lipolysis.

Subsequent studies focussed on functional analysis of a subset of the specific missense variants. mRNA expression of the PNPLA2 p.N252K variant was lower in the heterozygous carriers of the mutant allele than in homozygous wild-type allele carriers, although notably the variant retained its ability to target lipid droplets and also its catalytic activity. Functional analysis of the ALK7 variants indicated that the p.I195T variant failed to initiate signal transduction upon stimulation with known ligand Nodal, while the impact of the p.N150H mutant was far more subtle. Meanwhile, the PLIN1 p.L90P variant seemed to manifest a stronger interaction with HSL compared to wild type in the basal state. However, further investigation suggested that this variant suppressed stimulated lipolysis while the ability to promote lipid accumulation was unaffected.

Collectively, these findings provide some early insights into the potential impact of variants convincingly linked with altered body fat distribution and metabolic disease risk, but further work is required to more clearly understand exactly how they affect adipogenesis and/or lipolysis and thereby modify fat distribution and disease risk in vivo.