PLoS ONE (Jan 2011)

Proteomics mapping of cord blood identifies haptoglobin "switch-on" pattern as biomarker of early-onset neonatal sepsis in preterm newborns.

  • Catalin S Buhimschi,
  • Vineet Bhandari,
  • Antonette T Dulay,
  • Unzila A Nayeri,
  • Sonya S Abdel-Razeq,
  • Christian M Pettker,
  • Stephen Thung,
  • Guomao Zhao,
  • Yiping W Han,
  • Matthew Bizzarro,
  • Irina A Buhimschi

DOI
https://doi.org/10.1371/journal.pone.0026111
Journal volume & issue
Vol. 6, no. 10
p. e26111

Abstract

Read online

Intra-amniotic infection and/or inflammation (IAI) are important causes of preterm birth and early-onset neonatal sepsis (EONS). A prompt and accurate diagnosis of EONS is critical for improved neonatal outcomes. We sought to explore the cord blood proteome and identify biomarkers and functional protein networks characterizing EONS in preterm newborns.We studied a prospective cohort of 180 premature newborns delivered May 2004-September 2009. A proteomics discovery phase employing two-dimensional differential gel electrophoresis (2D-DIGE) and mass spectrometry identified 19 differentially-expressed proteins in cord blood of newborns with culture-confirmed EONS (n = 3) versus GA-matched controls (n = 3). Ontological classifications of the proteins included transfer/carrier, immunity/defense, protease/extracellular matrix. The 1(st)-level external validation conducted in the remaining 174 samples confirmed elevated haptoglobin and haptoglobin-related protein immunoreactivity (Hp&HpRP) in newborns with EONS (presumed and culture-confirmed) independent of GA at birth and birthweight (P<0.001). Western blot concurred in determining that EONS babies had conspicuous Hp&HpRP bands in cord blood ("switch-on pattern") as opposed to non-EONS newborns who had near-absent "switch-off pattern" (P<0.001). Fetal Hp phenotype independently impacted Hp&HpRP. A bayesian latent-class analysis (LCA) was further used for unbiased classification of all 180 cases based on probability of "antenatal IAI exposure" as latent variable. This was then subjected to 2(nd)-level validation against indicators of adverse short-term neonatal outcome. The optimal LCA algorithm combined Hp&HpRP switch pattern (most input), interleukin-6 and neonatal hematological indices yielding two non-overlapping newborn clusters with low (≤20%) versus high (≥70%) probability of IAI exposure. This approach reclassified ∼30% of clinical EONS diagnoses lowering the number needed to harm and increasing the odds ratios for several adverse outcomes including intra-ventricular hemorrhage.Antenatal exposure to IAI results in precocious switch-on of Hp&HpRP expression. As EONS biomarker, cord blood Hp&HpRP has potential to improve the selection of newborns for prompt and targeted treatment at birth.