Placental expression of GKN1 and diminished pancreatic beta cell function during pregnancy.

AIMS/HYPOTHESIS: Pregnancy is characterised by augmented pancreatic beta cell function, which is impaired in gestational diabetes mellitus (GDM). Using placental transcriptomic analysis, we aimed to identify secreted proteins of placental origin that may impact beta cell function in pregnancy.

METHODS

We conducted a whole-transcriptome analysis of 433 placental samples from the prospective Genetics of Glucose regulation in Gestation and Growth (Gen3G) cohort. At 24-30 weeks' gestation, participants underwent a 75 g OGTT, with glucose and insulin levels measured fasting and 1 and 2 h post oral glucose load. We used glucose and insulin levels to calculate the Pregnancy Insulin Physiology (PIP) index, a measure of beta cell function in pregnancy that is equivalent to the disposition index. Placental tissue samples were collected at delivery. Using differential gene expression analysis, we identified associations between gene transcripts and beta cell function (PIP index) after adjustment for surrogate variables capturing confounders and latent variation. We used the false discovery rate (FDR) to adjust for multiple hypothesis testing. We quantified the relationship between the gene identified in the primary analysis and other glycaemic traits using weighted linear regression models.

RESULTS

We identified GKN1, encoding gastrokine 1, as the top differentially expressed gene (FDR <0.05) in analyses relating placental gene expression to pregnancy beta cell function. Higher placental GKN1 expression was associated with lower beta cell function (log2 change in GKN1 expression per SD change in beta cell function=-0.44, FDR=7.7 × 10-6). Placentas from pregnancies with GDM had higher GKN1 expression levels than placentas from unaffected pregnancies (log2 fold change=1.13, FDR=4.2 × 10-3). Higher placenta GKN1 expression was associated with higher BMI (β=0.16, p=0.035), post-load glucose (1 h: β=0.30, p<0.001; 2 h: β=0.32, p<0.001) and post-load insulin (1 h: β=0.19, p=0.008; 2 h: β=0.22, p=0.003) and a lower insulin sensitivity-adjusted Stumvoll first-phase estimate (β=-0.53, p<0.001).

CONCLUSIONS/INTERPRETATION: Higher placental expression of GKN1 is associated with reduced beta cell function in pregnancy and GDM.

DATA AND CODE AVAILABILITY: RNA-seq data are available from the Database of Genotypes and Phenotypes (dbGAP; https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs003151.v1.p1 ). Code for these analyses is available from https://github.com/labjacquespe/diff-exp .