Profiling Mannheimia haemolytica infection in dairy calves using near infrared spectroscopy (NIRS) and multivariate analysis (MVA)

Authors: SANTOS-RIVERA, MARIANNA - Mississippi State University; WOOLUMS, AMELIA - Mississippi State University; THORESEN, MERRILEE - Mississippi State University; BLAIR, ELLIANNA - Mississippi State University; JEFFERSON, VICTORIA - Mississippi State University; MEYER, FLORENICA - Mississippi State University; VANCE, CARRIE - Mississippi State University

Submitted to: Nature Scientific Reports
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/23/2020
Publication Date: 1/14/2021
Citation: Santos-Rivera, M., Woolums, A., Thoresen, M., Blair, E., Jefferson, V., Meyer, F., Vance, C.K. 2021. Profiling Mannheimia haemolytica infection in dairy calves using near infrared spectroscopy (NIRS) and multivariate analysis (MVA). Nature Scientific Reports. https://doi.org/10.1038/s41598-021-81032-x.
DOI: https://doi.org/10.1038/s41598-021-81032-x

Interpretive Summary: Bovine Respiratory Disease (BRD) is a multi-factorial, multi-pathogen condition causing billion-dollar losses to the worldwide cattle industry. BRD associated with Mannheimia haemolytica, a gram-negative coccobacillus that is a constituent of the normal flora of the upper respiratory system of ruminants, is the principal cause of pneumonia in cattle. Visual-clinical diagnosis of cattle appearance and behavior, after signs of infection are presented, is currently the standard approach to detecting BRD but has low specificity and sensitivity, meaning asymptomatic or mildly symptomatic cattle in the early stages of the disease are not identified or treated, leading to the spread of the infection throughout the herd. We propose that using the biophotonics technique of near infrared spectroscopy(NIRS) to assess easily collectible biofluids could be a more rapid and accurate diagnostic approach to detecting BRD in cattle. Thus, this study aims to test and create prediction models of respiratory disease infection by identifying and discriminating the key clinical signs, blood parameters, and NIRS spectra of blood plasma in order to profile responses of dairy calves to infectious agents. Our long-term goal is to create a diagnostic strategy for the detection and treatment of animals in the early stages of BRD, thereby contributing to the sustainability of the food supply chain.

Technical Abstract: Bovine respiratory disease (BRD) linked with Mannheimia haemolytica is the principal cause of pneumonia in cattle. Diagnosis of BRD traditionally relies on visual assessment, which can be untimely, insensitive, and nonspecific leading to inadequate treatment and further spread of disease. Near Infrared Spectroscopy (NIRS) is a rapid acquisition vibrational spectroscopy that can profile changes in biofluids, and when used in combination with multivariate analysis, has potential for disease diagnosis. This study characterizes the NIR spectral profile of blood plasma from dairy calves infected with M. haemolytica and validates the spectral biochemistry using standardized clinical and hematological reference parameters. Blood samples were collected for four days prior to (baseline), and 23 days after, a controlled intrabronchial challenge. NIR spectral profiles of blood plasma discriminated and predicted Baseline and Infected states of animal disease progression with accuracy, sensitivity, and specificity = 90% using PCA–LDA models. These results show that physiological and biochemical changes occurring in the bloodstream of dairy calves during M. haemolytica infection are reflected in the NIR spectral profiles, demonstrating the potential of NIRS as a diagnostic and monitoring tool of BRD over time.