Context Glucocorticoids used in pharmacological doses for the treatment of a variety of medical conditions, and endogenous glucocorticoid excess – Cushing’s syndrome, may result in several adverse effects, but currently there is no clinically useful biomarker of glucocorticoid activity. Objective To identify glucocorticoid-responsive proteins potentially measurable in human serum. Setting and Design A three-phase protein biomarker discovery strategy was used. Proteomic biomarker discovery and qualification was conducted on the secretome of ex vivo-stimulated peripheral blood mononuclear cells (PBMC) isolated from 6 volunteers, incubated ± dexamethasone 100 ng/mL for 4h and 24h. Untargeted proteomics with label-free quantification (LFQ) was conducted to discover candidate proteins which were quantified using targeted proteomics by a custom multiple reaction monitoring mass spectrometry (MRM-MS) assay. Five proteins were selected for serum measurement by immunoassay in 20 healthy volunteers, with blood drawn at baseline and 12h after 4 mg oral dexamethasone.

A scheduled multiple reaction monitoring (MRM) method measuring 179 peptides from 44 proteins (43 biomarker candidates plus internal standards, chicken ovalbumin and iRT peptides) was developed and used to qualify candidate biomarkers in the secretome samples. Samples were analysed on a standard-flow Nexera X2 UHPLC coupled to a LCMS-8050 triple quadrupole mass spectrometer (Shimadzu). Internal standard protein (225 ng of chicken ovalbumin) was spiked into each secretome sample, and then trypsin co-precipitation method was used for sample processing.Ovalbumin and iRT peptides were checked for inconsistencies in sample preparation and LC/MS analysis. A retention time predictor was generated based on iRT peptide results to increase confidence of peptide identification. Peak picking was manually inspected and adjusted accordingly by comparing retention time and normalised peak area information across runs. The relative ratio of each transition normalised to the total peak area acts as a ‘fingerprint’ for a compound and was therefore a crucial indicator of a positive identification.

PBMCs were prepared from 6 healthy volunteers for acute and chronic glucocorticoid treatments, respectively. Ten milliliters of blood collected in EDTA was gently layered under 7 mL of Ficoll-Paque PLUS (GE Healthcare, Uppsala, Sweden) and centrifuged at 500 g for 20 min without brake. The interphase containing PBMCs was collected using a transfer pipette and transferred to a fresh tube. Cells were washed twice by resuspending up to 50 mL with PBS and spinning at 500 g for 5 min. Cells were resuspended up to 25 mL with PBS and spun at 200 g for 10 min to remove platelets. Aliquots of 1x106 PBMCs in Dulbecco’s Modified Eagle Medium (DMEM) + 10% Fetal Bovine Serum (FBS) were exposed to 100 ng/mL dexamethasone for 4 or 24 hours, while the corresponding control 1x106 PBMCs were exposed to media without dexamethasone for the same duration. The treatment media were removed, and the cells were incubated with serum-free and dexamethasone-free media for 3 hours to collect secreted proteins (secretome). Secretome samples were concentrated, and protein concentration estimated by Bradford assay.