ABCC1 Influences Negative Feedback of the Hypothalamic-Pituitary-Adrenal (HPA) Axis in Humans In Vivo

Presentation Number: OR03-1
Date of Presentation: April 2nd, 2017

Catriona Kyle*, Mark Nixon, Natalie Z Homer, Ruth Andrew, Roland H Stimson and Brian R Walker
University of Edinburgh, Edinburgh, United Kingdom


Cortisol and corticosterone both circulate in humans and may exert different effects in target tissues because of differential transmembrane export by the ATP-binding cassette (ABC) transporters. ABCB1 exports cortisol but not corticosterone while ABCC1 exports corticosterone not cortisol. ABCC1, but not ABCB1, is expressed in human adipose and intracellular corticosterone concentration increases with ABCC1 inhibition in vitro and in mice (1). We hypothesised that inhibition of ABCC1 increases occupancy of corticosteroid receptors by corticosterone but not cortisol.

A double-blind randomised crossover study was conducted in 14 healthy men (mean±SD age 28.7±13.5 years, BMI 24.1±2.6 kg/m2) comparing 5 days of placebo or probenecid (1g BID), an ABCC1 inhibitor. After each phase, blood sampling was undertaken at 3 sites (arterialised, skeletal muscle and adipose veins) before and after IV administration of the mineralocorticoid receptor (MR) antagonist potassium canrenoate followed by oral administration of the glucocorticoid receptor (GR) antagonist mifepristone (RU486). Blood flow was measured in adipose tissue and skeletal muscle using xenon washout and venous occlusion plethysmography, respectively.

During the placebo phase, systemic plasma cortisol and corticosterone concentrations increased promptly after canrenoate. In skeletal muscle, there was net uptake of cortisol throughout but net release of corticosterone after RU486. In adipose tissue, there was net uptake of both cortisol and corticosterone throughout. Probenecid had no effect on plasma concentrations of canrenoate or RU486 but increased blood flow in skeletal muscle but not adipose. In the systemic circulation, probenecid had no effect on glucocorticoid or ACTH levels at baseline or after MR antagonism. However, with the addition of GR antagonism, probenecid substantially increased cortisol concentrations (by 31% by AUC) and had a similar though non-significant effect on corticosterone and ACTH levels. In both skeletal muscle and adipose tissue, probenecid had no measurable effect on net glucocorticoid uptake or release at baseline or after MR or GR antagonism.

We conclude that ABCC1 inhibition induces a greater incremental activation of the HPA axis after GR/MR blockade but not after selective MR blockade. Establishing whether ABCC1 contributes to GR/MR occupancy in skeletal muscle or adipose will likely require more sensitive measurements using stable isotope tracers. Our findings suggest that ABCC1 contributes to export of glucocorticoids from sites involved in HPA axis negative feedback, thereby limiting activation of GR. This mechanism is likely to depend upon export of corticosterone not cortisol, operating over and above the previously reported role of ABCB1 in limiting availability of cortisol in CNS (2). This further reinforces the discrete roles for corticosterone and cortisol in humans.


Nothing to Disclose: CK, MN, NZH, RA, RHS, BRW