Variation in the relative isomer abundance of synthetic and biologically derived phosphatidylethanols and its consequences for reliable quantification.

Phosphatidylethanol (PEth) in human blood samples is a marker for alcohol usage. Typically, PEth is detected by reversed-phase liquid chromatography coupled with negative ion tandem mass spectrometry, investigating the fatty acyl anions released from the precursor ion upon collision-induced dissociation (CID). It has been established that in other classes of asymmetric glycerophospholipids the unimolecular fragmentation upon CID is biased depending on the relative position (known as sn-position) of each fatty acyl chain on the glycerol backbone. As such, the use of product ions in selected-reaction-monitoring (SRM) transitions could be prone to variability if more than one regioisomer is present in either the reference materials or the sample. Here, we have investigated the regioisomeric purity of three reference materials supplied by different vendors, labelled as PEth 16:0/18:1. Using CID coupled with ozone-induced dissociation, the regioisomeric purity (% 16:0 at sn-1) was determined to be 76%, 80% and 99%. The parallel investigation of the negative ion CID mass spectra of standards revealed differences in product ion ratios for both fatty acyl chain product ions and ketene neutral loss product ions. Furthermore, investigation of the product ion abundances in CID spectra of PEth within authentic blood samples appears to indicate a limited natural variation in isomer populations between samples, with the cannonical, PEth 16:0/18:1 (16:0 at sn-1) predominant in all cases. Different reference material isomer distributions led to variation in fully automated quantification of PEth in 56 authentic dried blood spot (DBS) samples when a single quantifier ion was used. Our results suggest caution in ensuring the regioisomeric composition of reference materials are well-matched with the authentic blood samples.