Bovine mastitis is the inflammation of udder due to physical injury or microbial infections. The milk from different mastitis statuses present a different microbial profile that can impact the mechanisms and pathophysiology of mastitis. Various studies were done in Malaysia using culturable methods to reveal the causative microorganisms in order to get an insight into the bacteria involved in bovine mastitis (Ariffin et al., 2019); (Marimuthu et al., 2014). However, studies examining the intramammary bacteria population of dairy cattle through conventional culture methods have been restricted to limited selective media, which isolate a limited spectrum of bacteria. Such studies are not conclusive in representing the total bacterial diversity existing in different clinical statuses of mastitis.

Numerous studies reported that 10 % to 40 % of clinical mastitis cases with “no growth”, which is defined as no bacterial growth was observed, when subjected to conventional culture methods (Kuehn et al., 2013), and evidence suggests that such cases are increasing (Kuehn et al., 2013) There is no scientific evidence on the real reason behind this phenomenon.

However, the failure to culture bacteria might be due to their concentrations being below detectable levels, or the mastitis infection may be caused by non-bacterial agents. (Kuehn et al., 2013).

High-throughput next-generation sequencing (NGS) technology involving 16S amplicon analysis and bioinformatic tools is a way to overcome the limitations of culture-based approaches (Kennedy et al., 2016). This method has proven it to be possible to identify more than 90% of the majority of bacterial or pathogen communities at the genus level (Hoque et al., 2019).

Mastitis is thought to be linked to alterations of microbiota composition in the udder, which can trigger an inflammatory response (Derakhshani et al., 2018). Mastitis may not only be caused by pathogenic microorganisms but also by microbial imbalance in the milk, a condition referred to as dysbiosis (Derakhshani et al., 2018); (Kuehn et al., 2013). To date, the microbial profiles of healthy versus mastitic mammary glands remain insufficiently characterized (Derakhshani et al., 2018). In Malaysia, studies on mastitis milk microbiota are limited. One study conducted in Malaysia has proven that mastitis related to dysbiosis and the balance of microbiota is a condition that provides a protective role against mastitis (Tan et al., 2023). However, there is a lack of reported evidence demonstrating the shift of the microbiota profile over time. Therefore, our study aimed to investigate the microbiome in healthy bovine and clinical mastitis milk, and the shifts in mastitis milk microbiota over a three-week period, which allows a comprehensive and practical time frame, within a Jersey Friesian mastitis herd in a local farm. This duration is ideal for studying the invasion pattern and the host-pathogen population interactions.

A total of 20 samples were grouped (n=5) in four groups, which consist of HT (HT[a], HT[b], HT[c], HT[d], HT[e]), W1 (W1[a], W1[b], W1[c], W1[d], W1[e]), W2 (W2[f], W2[g], W2[h], W2[i], W2[j]), and W3 (W3[k], W3[l], W3[m], W3[n], W3[o]). The clinical status is categorized into healthy non-mastitis (HT), and clinical mastitis throughout three weeks (W1, W2, W3), with 5 animals in each category representing different stages of severity of the condition to investigate the microbiota dynamic shift across three weeks in the same herd. Microbial DNA was isolated from each milk sample. The V1-V2 hypervariable regions were amplified (Table 1). Sequencing of 20 milk samples produced 8,000,852 sequences with sizes ranging from 35 to 301 bp. A total of 2,481,342 sequences were eventually used for analysis after trimming and chimeric sequence exclusion.

The median and distribution of the Chao and Shannon index demonstrated that Clinical groups across the three weeks decreased (Figure 1). The diversity of W2 and W3 is significantly lower compared to HT and W1. The microbial diversity, as well as species richness and evenness, were shifted in the period of 3 weeks of study. Mostly similar genera of pathogens were domineering the herd across the time frame, as shown in the diversity indexes. The PCA analysis illustrated varying bacterial sequence compositions in the HT group in contrast to the W1, W2, and W3 groups. The HT group exhibited a unique distribution pattern, whereas the W1, W2, and W3 groups showed overlapping patterns. Despite this overlap, distinct differences in bacterial distributions were observed, indicating clear disparities among these groups (Figure 2).

The HT group displays a more balanced distribution of phyla compared to W1, W2, and W3 (Figure 3). Over the specified time frame, there was a noticeable increase in Firmicutes within the population (HT=35.40%; W1=63.10%; W2=89.32%; W3=90.86%). Conversely, most other phyla, including Actinobacteriota (HT=34.08%; W1=7.87%; W2=1.01%; W3=6.95%), Proteobacteria (HT=11.17%; W1=18.69%; W2=7.50%; W3=1.14%), and Bacteroidota (HT=14.77%; W1=1.86%; W2=1.01%; W3=0.88%) showed a decreasing trend from W1 to W3 (Figure 4). This study supports the idea of dysbiosis in mastitis samples. The prevalence of the dominant bacteria in clinical mastitis samples indicates their potential as causative mastitis pathogens on the respective farm.

The data indicated that Mycoplasma spp. dominated the clinical mastitis cases during the study period (Figure 5). The W1, W2, and W3 exhibited a rapid increase in the percentage of Mycoplasma spp., reaching as high as 86.9% in W3, whereas this percentage was only 0.35% in the HT group. Conversely, the HT group showed a high prevalence of Rhodococcus spp. (27.81%) and Streptococcus sp. (11.76%), while other genera accounted for approximately 5% or less (Table 2).

The data from 16S amplicon sequencing was collected over three weeks of milk sampling from the same cow herd. The microbial population went through a dynamic change in its composition (Figure 5). From the HT milk microbial population, these changes in the microbial population dynamics range from the dominance of mutual symbionts to opportunistic pathogens while entering clinical mastitis and eventually towards the dominance of a pathogenic species during extreme conditions, as represented by microbial dysbiosis (Porcella et al., 2020).

In the overall analysis, Firmicutes clearly dominated the farm during the sampling period, leaving no doubt that this bacterial group was the phyla of the causative agent of mastitis at the sampled point (Figure 3). The data provides a fairly unbiased snapshot of the phyla and genera present in the milk microbial community. The data might be a good representation of the insight into cow mastitis intramammary of untreated cows across the timeline of infection, as a mastitis treatment will take approximately 5-21 days. The escalating population of Firmicutes coincided with a decrease in other phyla, such as Proteobacteria, Actinobacteria, Desulfobacterota, Bacteroidota, Patescribacteria, Campylo-bacterota, Chloroflexi, and Cyanobacteria, indicating that Firmicutes dominated other phyla (Figure 4).

At the genus level, Mycoplasma, known as one of the main pathogens in bovine mastitis, exhibited a noticeable increase, reaching 89.66% in W3, becoming the predominant member within the Firmicutes phylum. Mycoplasma species are significant contributors to bovine mastitis, causing both clinical and subclinical infections in dairy cows (Fox, 2012). Mycoplasma tends to spread rapidly within a herd, as evidenced by our investigation of clinical mastitis infection over a three-week sampling period (Figure 5). Mycoplasma infection led to alterations in milk SCC, as well as changes in milk composition and quality (Al-Farha et al., 2017). Our findings suggest that Mycoplasma infections in cows’ mastitis can be rapidly transmitted, leading to a significant prevalence of the bacteria in affected animals.

Comparatively, Mycoplasma was found at 0.35% in the HT group, indicating its role as one of the balanced microbes within a diverse microbial population, similar to other mastitis pathogens such as Streptococcus (11.76 %), Staphylococcus (0.24 %), Clostridium (1.64 %), Pseudomonas (1.41 %), and Enterococcus (0.64 %) (Table 2). Mycoplasma mastitis is usually excluded from standard mastitis screening tests due to its unique growth needs and the time delay involved (Okella et al., 2023). Traditional culturing of Mycoplasma from milk samples used to be a slow process, often taking up to two weeks and frequently resulting in non-growth due to these bacteria’s specific culture requirements (Parker et al., 2018). Recent research has shown that some Mycoplasma-infected samples yield negative cultures but positive PCR results (Al-Farha et al., 2017). However, metagenomic technology provides comprehensive information about the causative agents in respective farms.

Rhodococcus in the phyla of Actinobacteria is a genus of aerobic, gram-positive bacteria. It has a relatively fast growth rate and a simple development cycle, and exists in a high percentage in the HT group (Figure 3). Rhodococcus can be pathogenic, but most of the species are benign and largely found in environments such as water and soil (Patek et al., 2021). Livestock, according to research by (Vechi et al., 2018) and (Żychska et al., 2021), can serve as hosts for Rhodococcus. However, in our study, Rhodococcus was identified as one of the mutual microbes in the HT milk samples, as it does not affect the milk quality despite the high population in the HT milk (27.81 %).

Besides, Streptococcus did not emerge as a pathogenic bacteria during this specific sampling period, presenting a contrast to previous findings. In earlier samplings, Streptococcus exhibited a significantly high prevalence in the CM milk group and is known as the causative agent of mastitis (Table 2). Conversely, Streptococcus was identified as one of the common microbes in the HT group during other sampling instead (Table 2). It highlights that different pathogens might cause distinct clinical mastitis cases, contingent on the prevailing causative agents in each situation. Intriguingly, pathogenic bacteria were also detected in the Healthy group’s milk, indicating their existence as mutual symbionts within a balanced microbial population. This further supports the concept of dysbiosis in mastitis disease within our local farm.

Ralstonia was associated with the contamination of water, as well as water purifying systems, potentially as a source of contamination in milking that relies heavily on water (Ryan et al., 2011). Therefore, the genera may exist in higher amounts in the farm, as the local farm is highly reliant on underground water. As for other genus, decreasing trends were observed except for Mycoplasma. This finding provides valuable information for developing targeted treatment strategies for dairy herds. Therefore, vigilant monitoring and management strategies are crucial to prevent the spread of Mycoplasma-induced mastitis and maintain the overall health of the dairy herd on the farm.

Analysis of microbiota dynamics over the three weeks reinforced the association between the incidence of microbiota dysbiosis and mastitis. The decreasing Chao and Shannon diversity index during this time (Figure 1) suggest a shift in microbial community composition. This imbalance, characterized by the overrepresentation of certain pathogens, likely contributes to mastitis development. While the precise mechanisms by which specific microbial taxa influence this process remain unclear, our findings further support the link between microbiota dysbiosis and clinical mastitis. The balance of microbiota provides a certain degree of protection against infectious diseases, like mastitis. The investigation reported in Malaysia to characterize shifts in milk microbiota over time using 16S rRNA amplicon sequencing, offering new insights into microbial dynamics associated with mastitis, and the importance of incorporating molecular tools for detection of mastitis-related pathogens and targeted treatment strategies for the dairy herds.

The dominance of Firmicutes persisted throughout the weeks, particularly Mycoplasma in the clinical mastitis (CM) group, while other populations decreased over the specified time frame. The healthy (HT) group maintained a more diverse distribution of phyla. This study supports the idea of dysbiosis in mastitis samples. The prevalence of specific bacteria in clinical mastitis samples indicates their potential as causative mastitis pathogens on the respective farm. This finding provides valuable information in developing targeted treatment strategies for the dairy herds.

We certify that there is no conflict of interest with any financial, personal, or other relationships with other people or organizations related to the material discussed in the manuscript.

Special thanks to the research team of Apical Scientific Sdn. Bhd assisting in data processing, analysis, and troubleshooting.