Applications of gas chromatography—mass spectrometry in the study of androgen and odorous 16-androstene metabolism by human axillary bacteria

Abstract

The known involvement of axillary microflora with under-arm odour (UAO) production led us to determine whether the odorous 16-androstene steroids are formed in the axilla by bacterial metabolism of an odourless precursor such as testosterone. Axillary bacteria from 34 men were selectively cultured for aerobic coryneform bacteria (ACB), Micrococcaceae and propionibacteria. Overnight suspensions of bacteria were incubated separately at 37°C for two weeks with radiolabelled testosterone plus unlabelled testosterone (0.5 mg) and 0.5-mg quantities of 4,16-androstadien-3-one (androstadienone) and 5,16-androstadien-3β-ol (androstadienol). After extraction and purification by Sep-Pak cartridges and thin-layer chromatography, the eluted steroids were derivatised as the pentafluorobenzyl oximes (PFBO) and tert.-butyl dimethylsilyl (TBDMS) ethers. Saturated analogues were used as internal standards. Selected-ion monitoring electron-impact mass spectrometry was performed at the m/z corresponding to the M + ion for the PFBO derivatives and the [M − 57] + ion for the TBDMS ethers. Only ACB produced classical musk-like UAO (UAO +ve) in an in vitro odour-producing system with 29% being UAO −ve. ACB (UAO +ve) metabolised far more ( p = 0.001) testosterone than ACB (UAO −ve), the principal metabolites being 5α(β)-dihydrotestosterone, 5α(β)-androstane-3,17-dione and 4-androstene-3,17-dione (4-androstenedione). No non-polar 16-androstenes were formed. Micrococcus luteus (ten strains) metabolised testosterone to 4-androstenedione only; propionibacterium spp. did not metabolise testosterone at all. However, incubation of 16-androstenes with ACB gave evidence for 4-ene-5α(β)-reduction, 3α(β)-oxido-reduction and epimerisation. In general the direction of transformations favoured formation of the more odorous 5α-androst-16-en-3-one (5α-androstenone) and 5α-androst-16-en-3α-ol (3α-androstenol) from less odorous steroids. Such transformations, in vivo, would not require de novo synthesis of 5α-androstenone or 3α-androstenol and would be consistent with utilisation by ACB of 16-androstenes already present in small quantities in fresh apocrine secretions, which are odourless, to produce a more powerfully smelling mixture on the axillary skin surface.