108: Molecular similarities and dissimilarities between hellp syndrome and preeclampsia (PE) using shotgun and targeted proteomics

Abstract

Although clinical management of HELLP syndrome is similar to PE, hypertension and proteinuria do not always occur with HELPP. Thus, it was suggested that HELLP and PE could be separate clinical entities. Based on the ability of the urine misfolded proteins to retain Congo red (congophilia) it was suggested PE is a protein conformational disorder similar to Alzheimer’s disease. Molecular diagnostics are becoming more often part of modern clinical medicine. Here we used state of the art shotgun proteomics and targeted multiple reaction monitoring (MRM) mass spectrometry to discover and validate molecular signatures in crude urine samples of pregnant women with PE and HELLP that potentially differentiate between these two clinical conditions. We analyzed urine samples from six groups of patients classified phenotypically as: non-hypertensive healthy controls (CRL, n=9, 30±1 wks), women with idiopathic proteinuria (IP, n=15, 28±2 wks), PE mild features (mPE, n=16, 33±1 wks), superimposed PE (spPE, n=16, 31±1 wks), PE severe features (sPE, n=16, 32±1 wks) and HELLP (n=16, 29±1 wks). In the discovery phase, urine proteins were enriched by CR precipitation and subjected to UPLC-tandem mass spectrometry. Linear discriminant analysis (LDA) was used to cluster clinical groups based on molecular signatures. We choose a non-parsimonious set of peptides that distinguished among clinical phenotypes and were suitable for targeted validation using MRM in crude urine samples. 1,747 molecular IDs were identified in CR precipitates. These merged into 690 protein groups that generated good separation among clinical phenotypes with a false discovery rate of 1.12% (Fig. A). 21 key peptides were prioritized through pairwise comparison among groups (Wilkes Lambda MANOVA=0.006, p-value of 3.6 x 10-14) (Fig. B). Based on a full MS/MS spectral library, 98 MRM transitions were generated. CLIC1 (oxidant sensor activated by misfolded proteins), ABHD14B (lipid metabolizing-enzyme) and β2-microglobulin were among the parent proteins to distinguish sPE from HELLP. Principal component analysis demonstrated that targeted proteome from raw urine represented CR precipitate well and closely correlated with clinical criteria. Based on molecular signatures, HELLP syndrome appears to be a different clinical entity from sPE.