390 - Early-Life Plasma Metabolomic Signatures of Infant Vaccine Responsiveness in the Rochester Infant Cohort

Publication Number: 2379.390

Joann Diray-Arce, Boston Children's Hospital, Boston, MA, United States; Jing Chen, Boston Children's Hospital, Boston, MA, United States; Ravinder Kaur, Rochester General Hospital, Rochester, NY, United States; Caitlin Syphurs, Boston Children's Hospital, Boston, MA, United States; Nicole Prince, Brigham and Women's Hospital, Boston, MA, United States; Annmarie Hoch, Boston Children's Hospital, Cambridge, MA, United States; Kerry McEnaney, Boston Children's Hospital, Boston, MA, United States; Asimenia Angelidou, Harvard Medical School, Boston, MA, United States; Ofer Levy, Precision Vaccines Program, Boston Children's Hospital, Boston, MA, United States; Jessica A. Lasky-Su, Brigham and Women's Hospital, Bston, MA, United States; Michael Pichichero, Rochester General Hospital Research Institute, Rochester, NY, United States

Background: Infant responses to vaccines vary widely, yet the metabolic pathways can play key roles in immunity yet the metabolic factors that may associate with distinct vaccine responses remain unclear. Early-life metabolic states may shape immune maturation and the persistence of vaccine immunogenicity.

Objective: To define baseline and longitudinal plasma metabolomic profiles associated with vaccine-induced antibody responses (Ab) during infancy.

Design/Methods: Plasma samples from 109 infants (n=258 samples) in the Rochester Infant Cohort were analyzed between 5-24 months using untargeted metabolomics (Metabolon HD4 platform, 897 metabolites). IgG antibody titers against diphtheria (DT), tetanus (TT), Haemophilus influenzae type b (PRP), and pertussis antigens (PT, FHA, and PRN) were analyzed as continuous and ordinal outcomes based on published Rochester infant vaccine categories: very low (vLVR), low (LVR), normal (NVR), or high (HVR) vaccine responders. Longitudinal associations between metabolite trajectories and Ab responses were analyzed using generalized additive mixed-effects models with covariate adjustment. Weighted gene co-expression network analysis (WGCNA) defined metabolite modules and pathway-level association to identify early predictors of vaccine response.

Results: Plasma metabolites xylose (coef=0.14), lyxonate (coef=0.12), glycerol (coef=0.16), and iminodiacetate (coef=-0.14) showed significant longitudinal associations (p<=0.05 for both intercepts and slopes) over time with vaccine-induced Ab responses, implicating pentose and glycerolipid metabolism. WGCNA identified 17 metabolite modules enriched for lipid, amino acid, and cofactor/vitamin metabolism showing significant relationships with Ab response (coef=0.01). Pathway enrichment revealed methionine, cysteine, S-adenosylmethionine (SAM), and taurine metabolism, which regulate redox balance and methyl-donor availability critical for B-cell activation and antibody synthesis. Vaccine responders (NVR and HVR) demonstrated early increased in lipid metabolites involved in membrane synthesis and cellular energy pathways, suggesting that lipid remodeling supports the heightened metabolic demands of antibody-producing cells.

Conclusion(s): Distinct early-life metabolic profiles are associated with variability in vaccine-induced antibody responses. Pathways involving amino acid utilization, methylation, and lipid metabolism suggest immune maturation and antibody development during infancy. Longitudinal metabolomic profiling may identify early metabolic predictors of vaccine responsiveness and inform precision strategies for early-life immunization.

Longitudinal Changes in Key Plasma Metabolites Distinguish High and Low Infant Vaccine Responders
IDEAL_VR.pdfMetabolites xylose, lyxonate, glycerol, and iminodiacetate levels are shown over time (5–24 months) for infants categorized as high (NVR+HVR) or low (vLVR+LVR) vaccine responders. Curves represent generalized additive model (GAM) fits with shaded 95% confidence intervals. Shape and average parameters reflect the degree of curvature (slope) and average effect size (intercept), respectively, for longitudinal associations with antibody responses.