Modulation of tissue glucocorticoid exposure by cleavage of corticosteroid binding globulin in humans

Abstract

Corticosteroid Binding Globulin (CBG) binds over 85% of plasma cortisol and modulates free cortisol levels. Observations in vitro show that CBG undergoes proteolytic cleavage by neutrophil elastase (NE) at its reactive centre loop (RCL), a mechanism proposed to reduce CBG binding capacity by approximately ten-fold and increase the availability of free cortisol to tissues at sites of inflammation. However, detection of cleaved CBG in vivo in human plasma is controversial, and any influence of NE on CBG cleavage has not been tested in vivo. The CORtisol NETwork (CORNET) consortium found that genetic variation at a locus spanning SERPINA1 (encoding alpha-1 antitrypsin, AAT, the endogenous inhibitor of NE) and SERPINA6 (CBG) contributes to morning total plasma cortisol variation.

We tested the hypotheses that: (i) CBG cleavage occurs in tissues in vivo, controlling tissue cortisol delivery; (ii) AAT deficiency increases CBG cleavage and hence free plasma cortisol; (iii) greater cleavage of CBG results in increased tissue cortisol delivery in adipose and in hypothalamic-pituitary-adrenal (HPA) axis negative feedback; and (iv) NE inhibition inhibits NE-mediated CBG cleavage and thus plasma free cortisol in vivo.

To test tissue-specific CBG cleavage we recruited 48 men and collected arterialised blood and samples from veins draining abdominal subcutaneous adipose, forearm skeletal muscle, brain and liver. Arterio-venous differences in CBG were calculated, adjusting for blood flow. Net CBG production from the liver was detected in people with obesity and type 2 diabetes, but no in vivo cleavage of CBG was observed across the tissues studied.

In recall-by-genotype studies of people who are heterozygous for inactivating mutations in SERPINA1, 16 healthy carriers of either of the two most common AAT-deficiency single nucleotide polymorphisms (rs17580 & rs28929474) and 16 age-, sex- and BMI-matched controls were recruited from the Generation Scotland Biobank. Participants underwent placebo-controlled combined receptor antagonist stimulation of the HPA axis (‘CRASH’) testing using the glucocorticoid receptor antagonist RU486 plus mineralocorticoid receptor antagonist spironolactone, in a double-blind randomised crossover design. No measurable differences in CBG were observed. However, plasma free cortisol fraction was higher in those carrying AAT mutations. Adipose cortisol concentrations were not significantly different but transcripts of glucocorticoid-responsive genes were higher in adipose from AAT-deficient subjects. Plasma cortisol was elevated during CRASH testing in both groups, with the increment versus placebo tending to be lower in AAT-deficient subjects.

Using coronary artery bypass graft (CABG) surgery as a model of acute neutrophil-mediated inflammation, we measured CBG and cortisol in a randomised double-blind parallel group clinical trial of 35 patients administered placebo or elafin, an endogenous NE inhibitor, intravenously immediately before CABG surgery. Plasma CBG concentration and binding capacity fell by >30% following surgery, with corresponding increases in total and free cortisol, with a trend towards higher, rather than lower free cortisol in the elafin-treated group.

In conclusion, these data suggest that AAT deficiency brings about changes consistent with enhanced cleavage of CBG, including enhanced delivery of glucocorticoid to adipose tissue and reduced tonic HPA axis negative feedback. This is consistent with a role for intact CBG in delivery of cortisol to the central HPA axis and a role for CBG cleavage in releasing cortisol to enhance tissue access in inflamed adipose tissue.