Plasma proteomic analysis to identify potential biomarkers of histologic chorioamnionitis in women with preterm premature rupture of membranes

Abstract

Introduction: To identify potential biomarkers in the plasma that could predict histologic chorioamnionitis (HCA) in women with preterm premature rupture of membranes (PPROM), using shotgun and targeted proteomic analyses.

Methods: This retrospective cohort study included 78 singleton pregnant women with PPROM (24-34 gestational weeks) who delivered within 96 h of blood sampling. Maternal plasma samples were analyzed by label-free liquid chromatography-tandem mass spectrometry for proteome profiling in a nested case-control study design (HCA cases vs. non-HCA controls [n = 9 each]). Differential expression of 12 candidate proteins was assessed by multiple reaction monitoring-mass spectrometry (MRM-MS) analysis in individual plasma samples from cases and controls matched by gestational age at sampling (n = 40, cohort 1). A validation study was further performed in an independent study group (n = 38, cohort 2) using ELISA and turbidimetric immunoassay for three differentially expressed proteins.

Results: Shotgun proteomics analyses yielded 18 proteins that were differentially expressed (P < 0.05) between HCA cases and non-HCA controls. MRM-MS analysis of 12 differentially expressed proteins further revealed that the CRP, C4A, and SAA4 levels were significantly increased in women with HCA. A multi-marker panel comprising plasma SAA4 and C4A showed enhanced potential for differentiating HCA from non-HCA women (area under the curve = 0.899). Additional validation of these findings by ELISA assays revealed that the CRP levels were significantly higher in women with HCA than in those without HCA, whereas the plasma levels of C4A and SAA4 did not significantly differ between the two groups.

Conclusions: Plasma C4A, SAA4, and CRP were identified as potential biomarkers for detecting HCA in women with PPROM, based on targeted and shotgun proteomic analyses, showing good accuracy when used as a combined dual-biomarker panel (C4A and SAA4). Nevertheless, ELISA validation of these proteins, except for CRP, may not yield clinically useful markers for predicting HCA.

Fig 1. A brief flow chart of patients included in the proteomic and immunoassay analyses.

ELISA, enzyme-linked immunosorbent assay; HCA, histologic chorioamnionitis; PPROM, preterm premature rupture of membranes; MRM-MS, multiple reaction monitoring-mass spectrometry.

Fig 2. Schematic workflow of the discovery (label-free LC-MS/MS), verification (LC-MRM-MS), and validation (immunoassay) experiments.

C4A, complement C4-A; CRP, C-reactive protein; HCA, histologic chorioamnionitis; LC, liquid chromatography; MRM-MS, multiple reaction monitoring-mass spectrometry; MS/MS, tandem mass spectrometry; SAA4; serum amyloid A4; SIS, stable isotope-labeled standard; TGFBI, transforming growth factor-beta-induced.

Fig 3

(A) ROC curves for three plasma GNYDAAQR (SAA4) and VLSLAQEQVGGSPEK (C4A) levels to predict HCA (area under the curve [AUC] ± standard error [SE] = 0.830 ± 0.071 and 0.695 ± 0.084, respectively). (B) ROC curve for the best combined predictive model [including plasma VLSLAQEQVGGSPEK (C4A) and GNYDAAQR (SAA4)] to predict HCA, with an AUC = 0.899 (P < 0.05 for VLSLAQEQVGGSPEK vs. the combined predictive model, and P = 0.19 for GNYDAAQR vs. the combined predictive model). C4A, complement C4-A; HCA, histologic chorioamnionitis; ROC, receiver operating characteristic; SAA4; serum amyloid A4.

Fig 4. Plasma levels of C-reactive protein (CRP), complement C4-A (C4A), serum amyloid A4 (SAA4), and transforming growth factor-beta-induced (TGFBI) in histologic chorioamnionitis (HCA) cases and non-HCA controls.

All samples analysed (n = 20 and 18, respectively) were from study cohort 2. Horizontal lines indicate the respective median values.