Hepatic transcriptomic signatures of statin treatment are associated with impaired glucose homeostasis in severely obese patients

Daniel Margerie†,
Philippe Lefebvre†,
Violeta Raverdy,
Uwe Schwahn,
Hartmut Ruetten,
Philip Larsen,
Alain Duhamel,
Julien Labreuche,
Dorothée Thuillier,
Bruno Derudas,
Céline Gheeraert,
Hélène Dehondt,
Quentin Dhalluin,
Jérémy Alexandre,
Robert Caiazzo,
Pamela Nesslany,
Helene Verkindt,
François Pattou† and
Bart Staels†Email author View ORCID ID profile

†Contributed equally

BMC Medical Genomics201912:80

https://doi.org/10.1186/s12920-019-0536-1

Received: 13 December 2018
Accepted: 21 May 2019
Published: 3 June 2019

Open Peer Review reports

Abstract

Background

Clinical data identified an association between the use of HMG-CoA reductase inhibitors (statins) and incident diabetes in patients with underlying diabetes risk factors such as obesity, hypertension and dyslipidemia. The molecular mechanisms however are unknown.

Methods

An observational cross-sectional study included 910 severely obese patients, mean (SD) body mass index (BMI) 46.7 (8.7), treated with or without statins (ABOS cohort: a biological atlas of severe obesity). Data and sample collection took place in France between 2006 and 2016. Transcriptomic signatures of statin treatment in human liver obtained from genome-wide transcriptomic profiling of five different statin drugs using microarrays were correlated to clinico-biological phenotypes and also assigned to biological pathways and mechanisms. Patients from the non-statin-users group were matched to patients in the statin users group by propensity score analysis to minimize confounding effects from age, gender, parental familial history of diabetes, BMI, waist circumference, systolic and diastolic blood pressure and use of anti-hypertensive drugs as pre-specified covariates.

Results

We determined the hepatic, statin-related gene signature from genome-wide transcriptomic profiling in severely obese patients with varying degrees of glucose tolerance and cardio-metabolic comorbidities. One hundred and fifty seven patients on statin treatment in the matched cohort showed higher diabetes prevalence (OR = 2.67; 95%CI, 1.60–4.45; P = 0.0002) and impairment of glucose homeostasis. This phenotype was associated with molecular signatures of increased hepatic de novo lipogenesis (DNL) via activation of sterol regulatory element-binding protein 1 (SREBP1) and concomitant upregulation of the expression of key genes in both fatty acid and triglyceride metabolism.

Conclusions

A DNL gene activation profile in response to statins is associated with insulin resistance and the diabetic status of the patients. Identified molecular signatures thus suggest that statin treatment increases the risk for diabetes in humans at least in part via induction of DNL.