Mass Spectral Analysis of Sterols and Other Steroids in Different Ionization Modes: Sensitivity and Oxidation Artifacts

Abstract

In the course of synthetic work and mass spectrometry (MS) with hydroxy steroids, we observed not only the loss of H₂O but prominent 2 and 4 amu losses using atmospheric pressure chemical ionization (APCI), leading to confusion of the structural assignments. This loss of 2 amu, which we attributed mainly to oxidation of hydroxyls, varied among 44 steroids and sterols analyzed; 36 showed losses of 2n amu in APCI MS analysis (17/22 Δ⁵ steroids, 17/19 Δ⁴ steroids, and 2/3 estrogens). With the Δ⁵ steroids and sterols, the precursor MH⁺ was either observed as a minor ion or (more frequently) not detected at all, constituting the base peak in 7/22 cases. With heated electrospray ionization (HESI) MS, 2n amu losses were detected (generally weakly) in 9/44 cases but constituted the base peak in 3/9. In general, the sensitivity (base peak intensity) of steroids correlated with conjugation of the of the steroid frame. Δ⁴ steroids generally performed best on HESI⁺ (up to a maximum factor of 8-fold), while Δ⁵ steroids and sterols performed better on APCI⁺ (up to >137-fold), except for two trihydroxypregnanes. Estrogens did not show a clear trend. Sensitivity generally increased with the use of NH₄F as a mobile phase additive in ESI⁺ (up to a maximum of 7-fold). We conclude that the prominence of 2n amu losses is variable among steroids and sterols but is more commonly an artifact of APCI MS. These m/z losses can constitute dominant ions that impede detection of the precursor MH⁺ and complicate structural assignments.

Keywords: APCI; HESI; ammonium fluoride; artifacts; hydroxy steroids; mass spectrometry; oxidation; steroids.

1. Complications in Correct Product Ion Characterization in Cases of In-Source Oxidation

2. Structures and Names of Steroids Used in Analysis

2. Structures and Names of Steroids Used in Analysis

1

Mass spectra of four molecules showing strong 2n amu losses (2-electron oxidations, n= 1, 2, or 3). (A) 22­(R)-OH cholesterol (13), (B) 6β-OH testosterone (6), (C) pregnenolone (16), and (D) dehydroepiandrosterone (DHEA, 18). The exact m/z value of the MH⁺ precursor ion is indicated above each panel. The major ions detected are printed on each panel (note: the MH⁺ ion is not always detected).

2

Comparison of the APCI mass spectra d ₀-(35) and d ₂-17β-OH methyl-Δ⁵-androstene-3β-ol (36). (A) Scheme illustrating the products of the oxidation of the 3β-OH group followed by the 17β-OH group of d ₀- (upper) and d ₂-17β-OH methyl-Δ⁵-androstene-3β-ol (lower). The exact m/z values are displayed under each molecule. Mass spectra of 100 μM standards of (B) d ₀-17β-OH methyl-Δ⁵-androstene-3β-ol (35) and (C) d ₂-17β-OH methyl-Δ⁵-androstene-3β-ol (36) are shown (1 nmol injected), with the most prominent ions labeled. Ions colored red result from the loss of deuterium.

3

Comparison of HESI and APCI MS spectra of androst-5-ene-3β,17-diol (34). (A) Scheme illustrating the products of the loss of the 3β-OH group (−18 amu, upper) or two sequential oxidations of the 3β-OH and 17β-OH groups (−2 amu, lower) of androst-5-ene-3β,17-diol (34). The exact m/z values are displayed under each molecule. (B) HESI and (C) APCI MS spectra of a 100 μM standard of androst-5-ene-3β,17-diol (34), with the most prominent ions labeled.