Plasma proteomes associated with status of vitamins A and E in school aged Nepalese children (804.12)

Abstract

Vitamin deficiencies afflict undernourished populations, but remain latent from public health view and lack evidence‐guided interventions due to difficulty and cost of assessment. Plasma proteomics may offer a future approach to identify nutrient‐linked biomarkers, modeled to estimate status and predict prevalence of multiple vitamin deficiencies (Cole et al J Nutr 2013). We measured concentrations of retinol and α‐tocopherol (α‐T) by HPLC and relative abundance of plasma proteins by iTRAQ‐labeled, tandem mass spectrometry in plasma of 500, rural Nepalese children 6‐8 yr of age. Among ~980 proteins quantified in >10% of subjects, 77 were correlated with retinol (q <0.01, p<0.0006) and 42 with α‐T (q<0.01, p<0.0002), which included (Gene Symbols) RBP4, TTR, and many APO lipoproteins and novel correlates. Using linear mixed effects models, relative abundance of 3 proteins (RBP4, C1R and SCAF11) predicted serum retinol (R2=~82%) and 6 proteins (APOC3, APOB, PKM, RGS8, LRRC47 and FOX04) predicted α‐T status (R2=~72%); however, each nutrient had >9 other candidate protein predictors to explore further. A predictive plasma proteome may exist to assess vitamins A and E status and prevalence of deficiencies in populations if candidate proteins undergo absolute quantification.Grant Funding Source: Supported by the Bill and Melinda Gates Foundation Grants OPP5251 and GH614