Integrating multiple data layers could offer a better understanding of the pathogenetic mechanisms behind the development of mastitis and improve the identification of robust biomarkers for its detection. Hence, in this work we integrated serum metabolomic data from 1H nuclear magnetic resonance (1H-NMR) with milk leucocytes subpopulations measured with flow cytometry and milk udder health traits in healthy cows (n = 15, NEG) and cows with spontaneous subclinical intramammary infection (sIMI) (n = 19). The Data Integration Analysis for Biomarker discovery using Latent Components (DIABLO) algorithm was used to combine these datasets and identify key features for sIMI detection. The predictive ability of the selected hub variables in discriminating between NEG/sIMI animals was then assessed using receiver operating characteristic (ROC) analysis. This approach revealed a strong correlation (r = .73) between serum metabolomic and milk leukocytes categories. Among the most informative features selected by DIABLO we observed an increased concentration of lactose, due to the altered permeability of the mammary gland and histidine, potentially modulating immune responses. Additionally, the decreased concentrations of ruminal fermentation-associated metabolites (e.g. acetone, methanol, ethanol and 3-Hydroxybutyrate) suggest, with the onset of inflammation, a shift in energy allocation from physiological processes towards immune response. The predictive performance of the selected metabolites ranged from moderate to good. Lactose emerged as the most promising biomarker, while allantoin demonstrated high sensitivity (0.93). These findings demonstrate the potential of combining immunological and metabolomic profiling across different biofluids for mastitis detection, though larger cohort is required for testing its application. HIGHLIGHTS Strong relationship highlighted between milk immune cells and serum metabolomic variables Moderate correlations between milk leucocytes and serum metabolomic variables, but not with specific immune cells Infected animals show higher lactose and histidine levels Infected animals have lower ruminal fermentation-associated metabolites
Integrated serum metabolomics and milk immune cell profiling reveals systemic and local responses in Holstein cows with subclinical intramammary infection
Bisutti V.;Lisuzzo A.;Giannuzzi D.;Gelain M. E.;Fiore E.;Gianesella M.;Cecchinato A.;Pegolo S.
2026
Abstract
Integrating multiple data layers could offer a better understanding of the pathogenetic mechanisms behind the development of mastitis and improve the identification of robust biomarkers for its detection. Hence, in this work we integrated serum metabolomic data from 1H nuclear magnetic resonance (1H-NMR) with milk leucocytes subpopulations measured with flow cytometry and milk udder health traits in healthy cows (n = 15, NEG) and cows with spontaneous subclinical intramammary infection (sIMI) (n = 19). The Data Integration Analysis for Biomarker discovery using Latent Components (DIABLO) algorithm was used to combine these datasets and identify key features for sIMI detection. The predictive ability of the selected hub variables in discriminating between NEG/sIMI animals was then assessed using receiver operating characteristic (ROC) analysis. This approach revealed a strong correlation (r = .73) between serum metabolomic and milk leukocytes categories. Among the most informative features selected by DIABLO we observed an increased concentration of lactose, due to the altered permeability of the mammary gland and histidine, potentially modulating immune responses. Additionally, the decreased concentrations of ruminal fermentation-associated metabolites (e.g. acetone, methanol, ethanol and 3-Hydroxybutyrate) suggest, with the onset of inflammation, a shift in energy allocation from physiological processes towards immune response. The predictive performance of the selected metabolites ranged from moderate to good. Lactose emerged as the most promising biomarker, while allantoin demonstrated high sensitivity (0.93). These findings demonstrate the potential of combining immunological and metabolomic profiling across different biofluids for mastitis detection, though larger cohort is required for testing its application. HIGHLIGHTS Strong relationship highlighted between milk immune cells and serum metabolomic variables Moderate correlations between milk leucocytes and serum metabolomic variables, but not with specific immune cells Infected animals show higher lactose and histidine levels Infected animals have lower ruminal fermentation-associated metabolites
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
