Symposium Presentation Australian Society for Microbiology Annual Scientific Meeting 2024

Introduction

Understanding the bacterial causes of mastitis is crucial for the health and production of dairy cows.  This study aimed to identify pathogens isolated from the milk of cows with clinical mastitis in the subtropical dairy region of Australia and determine antimicrobial susceptibility. Additionally, it sought to profile and explore associations in milk microbiota between healthy, subclinical and clinical mastitis cows and identify predictors of health status.

Methods

Thirty dairy herds submitted milk samples for the first five cases of clinical mastitis each month for 12 months. Samples underwent aerobic culture, and isolates were identified by MALDI-TOF mass spectrometry. Antimicrobial susceptibility testing was performed for selected bacteria using the Sensititre microdilution system.

On one farm, additional healthy (n=28) and subclinical mastitis (n=27) milk samples were collected on February 18^th, 2022, alongside clinical mastitis samples (n=30). These underwent culture, 16S rRNA gene amplicon sequencing, and bioinformatics analysis. Twenty multinomial elastic net regression models were conducted using culture result, parity, lactation stage, and genus-level relative abundance to identify predictors (i.e., positive coefficients present in 70% of models) of healthy cows and those with subclinical and clinical mastitis.

Results

Between March 2021 and July 2022, 1,230 milk samples were collected. The most common isolate cultured was Streptococcus uberis (23.6%), followed by non-aureus staphylococci and mammaliicocci group (NASM) (15.0%). Pathogen proportion varied among individual farms. Overall, clinical mastitis pathogens displayed low levels of resistance to the antimicrobials tested.

The most abundant genera in the clinical mastitis samples were Streptococcus (16.2%) and Escherichia-Shigella (12.5%), while Staphylococcus dominated the subclinical mastitis (13.9%) and healthy samples (7.0%). Significant differences in alpha diversity measures were observed, and beta diversity showed clustering between sample types. Clinical mastitis samples displayed greater microbial dysbiosis. Elastic net identified 40 predictors of healthy cows, five for those with clinical mastitis, and 16 for subclinical mastitis.

Conclusion

Streptococcus uberis and the NASM group are common causes of clinical mastitis in the subtropical dairy region. The microbiota findings offer promising potential for machine learning to predict mastitis using cow-based and milk microbiota data, reducing disease burden and enhancing cow health and milk production.