Letter to the Editor: Comments on “Mammary microbial dysbiosis leads to the zoonosis of bovine mastitis: a One-Health perspective” by Maity and Ambatipudi

Dr Horn,

Maity and Ambatipudi (2021) provide a narrative minireview ‘Mammary microbial dysbiosis leads to the zoonosis of bovine mastitis: a One-Health perspective.’ I find many statements in this article are imprecise, incorrect, unsupported by the references cited, or over-interpret the data presented in the citations.

Beginning with the title; ‘the zoonosis of mastitis’ is an imprecise characterization, which might lead to confusion. Mastitis is defined as inflammation of the mammary gland (Watts 1988). A zoonosis (singular) is an infectious disease transmitted or shared between humans and other vertebrate animals (Schwabe 1984). Mastitis, per se, is not transmitted between cows and humans. Some pathogens that cause bovine mastitis are zoonotic. I believe semantics are important here. The title characterizes bovine mastitis as a single zoonotic disease. Mastitis is a complex, multifactorial disease with many etiologies; Watts (1988) described 137 microbial species, subspecies and serovars isolated from the bovine mammary gland. Many microbial species and strains associated with mastitis are opportunistic pathogens, some are zoonotic, many are not (Zadoks et al. 2011).

Continuing with the first sentence of the abstract, Maity and Ambatipudi characterize mastitis as a ‘prototypic emerging and reemerging bacterial disease.’ I believe this is inaccurate. Emerging infectious diseases (EIDs) are ‘infections that have newly appeared in a population or have existed previously but are rapidly increasing in incidence or geographic range’ (Morse 1995). Trends in mastitis epidemiology over the past six decades are complicated. The incidence and prevalence of some mastitis pathogens has declined, while other pathogens have increased in prevalence; some previously unrecognized species or strains have been identified, while the clinical significance of other pathogens is being reconsidered; and, there are geographical differences in these trends (Hillerton and Berry 2005; Pyörälä and Taponen 2009; Zadoks and Fitzpatrick 2009). Some bacterial strains emphasized in the minireview, e.g. Methicillin-resistant Staphylococcus aureus sequence type 398 (ST398) and group B Streptococcus strains, have recently emerged or are identified as zoonotic strains isolated from milk of dairy cattle or cases of bovine mastitis. In my opinion, these examples do not support broadly defining mastitis as an emerging or re-emerging disease. An approach to defining EIDs has been proposed (Funk et al. 2013).

Maity and Ambatipudi omit many mammary or milk association zoonotic pathogens (e.g. Brucella abortus, Coxiella burnetii, Listeria monocytogenes, Mycobacterium spp., Salmonella spp. and Campylobacter jejuni). Perhaps, these zoonotic pathogens were omitted because it is unclear how they might relate to putative bovine mammary dysbiosis, although genomic signals for these or closely related taxa are identified in mammary microbiome studies targeting 16S RNA gene amplification or using a shotgun metagenomic approach (Quigley et al. 2013; Andrews et al. 2019; Hoque et al. 2020). This brings me back to another problem with the provocative title, which is phrased as a conclusion that mammary dysbiosis causes (‘leads to’) mastitis associated zoonoses. Where is the evidence supporting this association or casual path? I find no compelling evidence that changes in microbial diversity, loss of beneficial commensals, or expansion of putative intramammary pathobionts lead to mastitis and increase the risk of mammary shedding of zoonotic pathogens. The temporal or causal relationship between putative intramammary dysbiosis and mastitis pathogen colonization or infection remains unresolved. It is equally likely that intramammary infection leads to changes in the results obtained from culture independent targeted amplicon or shotgun metagenomic sequencing.

In the manuscript by Maity and Ambatipudi (2021), the title of Table 1 ‘Records of the transmission to humans’ is misleading. It is not possible to demonstrate the direction of zoonotic transmission for any of the 16 descriptive studies or the two observational cohort studies in the table, and few of these publications are records of zoonotic transmission; rather they mostly demonstrate potential human exposure risk obtained from surveys of animals or animal products.

Maity and Ambatipudi state, ‘The emergence of a particular group of S. aureus strains, for instance, Methicillin-Resistant S. aureus (MRSA), has been primarily responsible for zoonotic transmission since its first outbreak in humans in 1961. Here they cite van Cleef et al. (2011), where I find no data supporting this statement. The concept that MRSA has been primarily responsible for zoonotic transmission since 1961 seems speculative and evidence that MRSA zoonotic transmission is historically more frequent than methicillin sensitive S. aureus (MSSA) zoonotic transmission needs documentation. Of particular importance to this issue is sampling bias due to use of selective microbiological methods. Most studies have focused on MRSA identification and MSSA colonization is rarely identified or quantified (Kock et al. 2014). Maity and Ambatipudi incorrectly cite Goerge et al. (2017) suggesting that LA-MRSA was ‘detected in 32–37% of cattle in 40–45% of farms across Europe.’ Goerge et al. (2017) provide no primary data on the prevalence of MRSA in cattle on cattle farms, rather they report data on human occupational risk for nasal colonization with MRSA, 31–37% of cattle-farmers.

Maity and Ambatipudi state ‘All the zoonotic outbreaks, dead-end infections without adaptation to humans (e.g. Ebola) or stably adaptation to humans (e.g. SARS), has led to sustained person-to-person transmission.’ The main independent clause ‘All the zoonotic outbreaks has led to sustained person-to-person transmission’ is incorrect. Many zoonotic outbreaks do not lead to sustained person-to-person transmission.

The authors state, ‘host immune system and the mammary microbiota synergistically set up a robust mutual control’ citing Thompson-Crispi et al. (2014). I find no reference to synergism between the host immune system and the mammary microbiome in the review article by Thompson-Crispi et al.

Maity and Ambatipudi suggest ‘commensal microbiota suppresses the proliferation of the existing indigenous pathobionts that could be triggered by external (e.g. diet, pharmaceuticals and hygienic environment) and internal factors’ citing Butto, Schaubeck and Haller (2015). In fact, Buttó, Schaubeck and Haller cite no primary data supporting functional suppression of pathobionts by commensal bacteria. Maity and Ambatipudi present no evidence that commensal microbiota suppress indigenous pathobionts of the mammary gland. Further, the concept of defining ‘pathobionts’ is not without controversy. Jochum and Stecher (2020) suggest, ‘We refrain from using the term.’ Rainard (2017) discussed the challenge of using research on the gut microbiota and mucosal immunity to make inferences to the interactions between a putative functional intramammary microbiota and mammary immune surveillance. There may be some recent publications suggesting interactions between the mammary microbiome and the immune system in humans and cattle (e.g. Sakwinska and Bosco 2019), but none of these are cited in this minireview.

Maity and Ambatipudi suggest in subclinical mastitis, ‘what seems like a peaceful coexistence of commensals and pathobionts; however, represents a constant struggle between the two groups with the immune system driving autoimmunity, resulting in the development of the clinical mastitis.’ Neither of the references cited provide primary data demonstrating a struggle or peaceful coexistence between commensals and pathobionts, nor does either study provide primary data on the host immune response, let alone a shift to autoimmunity in the mammary gland.

Maity and Ambatipudi state ‘Several studies have revealed the threats posed by discontinued use of antimicrobials on organic dairy farms and the subsequent consequences of the elevated burden of increased antimicrobial residues and AMR pathogens into the food chain.’ Neither of the references cited appear to describe any study where the discontinued use of antibiotics on organic farms results in the subsequent consequence of an elevated burden of increased antimicrobial residues and AMR pathogens into the food chain. It is not clear what the authors mean by ‘threats posed by discontinued use of antimicrobials on organic dairy farms.’ How does discontinuing antibiotic use lead to increased antimicrobial residues? I hope the authors will clarify.

This is only a partial list of issues I find in this minireview. In addition to the above issues, with careful review of only the first two pages of this paper, I find eleven other inaccuracies, imprecisions or statements unsupported by the citations provided. My intent is not to minimize the potential public health importance of zoonoses associated with dairy cattle or dairy products, especially for people in close contact with livestock or consuming raw products such as unpasteurized milk. I agree with the authors that bovine mastitis and dairy production should be considered from ecological and One Health perspectives. Others also recently addressed mastitis from these perspectives (e.g. Vanderhaeghen et al. 2015; Garcia, Osburn and Cullor 2019). My concerns are with the frequent inaccurate representations of prior research and the misleading conjecture, without critical appraisal, describing a causal relationship between dysbiosis, putative intramammary pathobiont expansion and zoonotic mammary pathogen emergence and transmission.

Thank you for the opportunity to present my concerns.

Conflicts of interest

None declared.