Introduction

Mastitis, inflammation of the mammary gland, is a common disease in cattle. It is related to the presence of pathogenic microorganisms, the most common of them being E. coli and various Streptococcus and Staphylococcus species1. Non-aureus staphylococci are pathogenic microorganisms frequently isolated in mammary gland infections2,3,4. Another group of microorganisms, which are less frequently isolated from milk but represent a considerable interest, are pathogenic bacilli5. The fact that pathogenic bacilli are capable of sporulation, which allows them to survive thermal treatment of milk and dairy products, further emphasizes their significance. The development of associations composed of different microbial species and their adaptation to their environments are currently increasingly attracting researchers' interest. In the present work, we have for the first time described consortia of Staphylococcus haemolyticus and Bacillus paranthracis isolated from the milk of cows diagnosed with mastitis in three geographically remote regions of Russia.

Results

General genomic characteristics

Each of the three studied bacterial consortia (4M, 1702, and 1710) was composed of two pathogenic strains. Genome quality estimation with CheckM showed that all genomes were of high quality (> 98% completeness and < 0.5% contamination). The Genome Taxonomy Database tool kit (GTDB-tk) was used to classify the bacterial genomes. GTDB-tk analysis classified one of strains in each consortium as a member of the Bacillus_A group, namely Bacillus paranthracis. The other strain in each consortium was Staphylococcus haemolyticus based on the taxonomic classification defined by topology and ANI. B. paranthracis strains exhibited a high level of similarity among different consortia, and so did the S. haemolyticus strains (Fig. 1). Each of the B. paranthracis strains possessed ten replicons, including one chromosome and nine plasmids ranging in size from 3124 bp to more than 300 kb. Five larger plasmids apparently had the theta-type replication, whereas the smaller plasmids were of the RCR type. The S. haemolyticus strains contained four replicons: one chromosome and three small plasmids of 6539, 3048, and 2362 bp (Table 1). Genetic features of plasmids found in B. paranthracis and S. haemolyticus strains are listed in Supplementary Data.

Figure 1
figure 1

Mauve whole-genome alignments of B. paranthracis (a) and S. haemolyticus (b) strains.

Full size image
Table 1 Genome size of B. paranthracis and S. haemolyticus strains.
Full size table

To find out in how much the strains composing the consortia differed genetically, pairwise comparison of the B. paranthracis genomes, as well as of the S. haemolyticus genomes, was performed using the ANI calculator. The results of this comparison are shown in Table 2. It can be seen that none of the strains studied was identical to some other strain. Most likely, these consortia originated from a common ancestor consortium and have been evolving independently. The characteristics of the three B. paranthracis genomes were absolutely identical, and so were the characteristics of the three S. haemolyticus genomes; they are listed in Table 3.

Table 2 Pairwise comparison of genomes by ANI calculator.
Full size table
Table 3 Genomic features of B. paranthracis and S. haemolyticus strains.
Full size table

Putative virulence genes

The isolated B. paranthracis strains were found to possess several putative virulence genes. The genes nheA, nheB, and nheC encode a pore-forming toxin composed of a cytolytic protein NheA and two associated protein components NheB and NheC, which enhance the biological activity of the cytolytic protein. There were also three homologous reading frames encoding a secreted metalloprotease InhA, which can cleave multiple proteins of the host organism cells. The multiple alignment and the evolutionary tree of the three inaA orthologs are presented in Supplementary Materials, Fig. S1. All B. paranthracis strains possessed two copies of the gene encoding thiol-activated cytolysin ALO, which belongs to the family of cholesterol-dependent cytolysins. ALO is a pore-forming toxin that requires the presence of cholesterol in the membrane for pore formation; to date, the mechanism of this process is not fully understood. In addition, the B. paranthracis genomes included a gene cluster composed of the genes hblA1, hblA2, and hblB, which encodes a component of pore-forming hemolysin BL. However,the most interesting finding in B. paranthracis strains concerned the gene of the pore-forming toxin cytK-2. In all three strains, the sequence of this gene carried a point mutation, an T → A transition at position 357 that produced a TAA stop codon interrupting the reading frame. As a result, instead of the full-size CytK-2 (336 amino acids), these cells probably synthesize its truncated variant of 103 amino acids, which rather resembles leukotoxin LukDv (Fig. 2). The multiple alignment of cytK2 nucleotide sequence of B. paranthracis 4M, B. paranthracis 1710, B. paranthracis 1702, B. cereus E33L, and B. thuringiensis BGSC 4Y1 is presented in Supplementary Materials, Fig. S2. In addition, this truncated protein exhibits 99% homology to subunit C of hemolysin gamma from Streptococcus pneumoniae (GenBank Sequence IDs CKI41873, COH50677, and CJA68379) (Fig. 3).

Figure 2
figure 2

Comparison between the neighborhoods of the cytK gene in the B. paranthracis, B. cereus and B. thuringiensis genomes.

Full size image
Figure 3
figure 3

Multiple alignments of truncated CytK-2 and subunit C of hemolysin gamma amino acid sequences.

Full size image

The S. haemolyticus strains lacked genes of pore-forming toxins but possessed several genes responsible for prevention of phagocytosis. These were type 8 capsule genes: capA, capB, capC, capD, and capP. These genes are also involved in formation of biofilms.

Biofilm formation

In addition to the genes capA, capB, capC, capD, and capP mentioned above, the isolated S. haemolyticus strains were found to possess the icaC gene, which is usually a part of the ica operon of intercellular adhesion required for biofilm formation. However, the other genes of this operon, icaADB, were not detected in the S. haemolyticus strains studied. In addition, the S. haemolyticus strains possessed the genes of the adapter proteins MecA and MecB, which play a central role in the process of biofilm formation.

The B. paranthracis genomes included the eps1 cluster comprising 17 open reading frames (Supplementary Materials, Fig. S3); these strains lacked the eps2 cluster. The genomes of B. paranthracis also included the genes of the multifunctional regulators papR, plcR, and sinR, which are of significant importance both for the expression of virulence factors and for biofilm formation. In addition, the genomes of B. paranthracis had other genes involved in biofilm formation: sipW and tasA; furthermore, tasA was represented by three orthologs. The first of them was located remote from sipW, upstream from a gene encoding a serine peptidase of the family S8. Two other tasA orthologs were located immediately downstream from sipW and directly upstream from sinR and inaA_2.

Antibiotic resistance genes

The B. paranthracis strains possessed two chromosomal antibiotic resistance genes. One of them, vmlR, encodes an ATPase that binds to the ribosome and provides resistance to virginiamycin and lincomycin by the target protection mechanism. The other gene, fosB, encodes a magnesium-dependent thiol transferase that degrades fosfomycin.

The chromosome of the S. haemolyticus strains featured two resistance genes: fusC is responsible for inactivation of fusidic acid; mgrA ensures active efflux of a number of antibiotics, such as ciprofloxacin, methicillin, and tetracycline. Furthermore, the 6539-bp plasmid of S. haemolyticus carried two further resistance genes. One of them, aacA–aphD, encodes an aminoglycoside acetyltransferase and provides resistance to aminoglycoside antibiotics (e.g., neomycin, gentamicin, and kanamycin). The other gene was initially described as tetB in Geobacillus stearothermophilus; it is responsible for tetracycline efflux.

Secondary metabolites

The B. paranthracis strains contained two non-ribosomal peptide synthetases responsible for the synthesis of bacillibactin and fengycin, respectively. The S. haemolyticus strains were capable of synthesizing staphyloferrin.

Summary of genes for the virulence, biofilm formation, and antibiotic resistance in B. paranthracis and S. haemolyticus strains is presented in Table 4.

Table 4 Summary of genes for the virulence, biofilm formation, and antibiotic resistance presented in each strain.
Full size table

Discussion

Mastitis is a common and economically important disease of cattle. In this work, we used the megasequencing technology to analyze the bacterial communities of the infected milk samples and identify the presence of genes that can participate both in the pathogenic process and in overcoming the host's immune response. It was found that milk samples collected from cows with mastitis consistently contained consortia of two gram-positive bacteria, S. haemolyticus and B. paranthracis. The best-known pathogenic agent associated with cattle mastitis is Staphylococcus aureus, a gram-positive coccus that also causes other clinical infections. It is also recognized as a pathogen provoking food poisoning outbreaks. The pathogenicity of S. aureus is due to a broad range of virulence factors, including hemolysins, leukocidins, enterotoxins, and secretion systems of numerous toxins6,7.

While S. aureus has been described as a major pathogen causing cattle mastitis, coagulase-negative staphylococci (CNS) are also increasingly recognized worldwide as etiologic agents associated with intramammary infections (IMI). However, clinical and pathogenetic significance of their detection in milk sample cultures is still debated. Some researchers believe that CNS are true mastitis-causing agents with major virulence factors, a high level of antimicrobial resistance, and the ability to cause chronic infections. At the same time, others consider them secondary pathogens of cattle.

Bacterial pathogens have to survive, proliferate, and spread within the host; they must be able to adapt to an extremely hostile environment by responding to the existing immune barriers. The interaction between a host and a pathogen should be viewed not as a static phenomenon, but as an arms race in which each competitor tries to act as effectively as possible8. As a consequence, pathogens have developed extremely sophisticated strategies to overcome the host's immune response. In particular, depending on the lifestyle of various microbial pathogens, phagocytosis as such can represent not only an obstacle but also an opportunity for their propagation: some of them employ numerous intricate strategies to enter phagocyte cells and survive in them, whereas other bacteria have developed mechanisms to prevent or evade phagocytosis. For successful infection, pathogens must overcome or avoid the activity of the host's immune system cells. Infection with virulent bacilli is characterized by bacterial proliferation in spite of inflammation at the infection site9. This implies that these bacteria have developed means to counteract inflammatory cells and thus the host immune system.

We performed a comparative analysis of the pathogenicity factors of the other consortium member B. paranthracis, which can attack epithelial cells, as well as overcome and/or suppress the host's immune response. Taxonomically, B. paranthracis is very close to Bacillus cereus, a gram-positive spore-forming bacterium that provokes food poisoning and severe opportunistic infections10,11. B. cereus is present in soil and in food products; it contaminates human skin and nearly all surfaces in hospital environments. It is the second most common cause of mass food-related infection outbreaks after S. aureus12. Therefore, B. cereus sensu lato is of particular interest in terms of food safety and public health because of its ability to spoil food and cause disease due to synthesis of various toxins13. In this study, we determined the presence of pathogenicity factors in the identified consortia. All B. paranthracis strains present in the consortia possessed a cytK-2 gene14 with a mutation disrupting the reading frame. The protein encoded by this new reading frame has a high level of homology to LukDv, a pore-forming toxin protein described in S. pneumoniae. This protein can form pores in the membranes of eukaryotic cells, unbalancing the partial pressure inside the host cells, which results in cell lysis, release of the nutrients into the environment, and ultimately in the death of these cells15. The novel protein is encoded by a gene with a high homology to cytK, which suggests a possibility of horizontal transfer of this gene among various gram-positive microorganisms that are in a physical contact.

The second best-studied genetic determinant of a pore-forming toxin in B. cereus sensu lato is the hbl gene cluster present in all B. paranthracis strains described in this work. Hbl-dependent pore formation requires all Hbl components, which can individually bind to erythrocytes and form a membrane attack complex that ultimately provokes cell lysis16. The presence of the genes encoding PlcR and its inducer PapR in all strains described in our work also serves for positive control of the expression of Hbl-encoding genes. Furthermore, the expression of the gene that encodes the protein highly homologous to LukDv apparently remained under the characteristic positive transcriptional control of the PlcR system17 as described for the wild-type cytK-2 gene.

B. cereus cells can survive in the host organism and cause infection in spite of attracting phagocytes. The genome of B. cereus includes over 350 genes that encode exoproteins; among them, there are at least 50 genes of proteases with several putative pathogenetic functions18. At all stages of infection with pathogenic microorganisms, proteases are key virulence factors. They contribute to colonization of the host by degrading the extracellular matrix of host tissues, including collage and elastin. Together with the cytolytic activity of pore-forming toxins, this protease-mediated degradation promotes bacterial proliferation by providing nutrients, interfering with the host's immune response, damaging the protective endothelium, and disrupting the epithelial barriers19. Among exoproteases of B. cereus, there are two zinc-dependent proteases, InhA1 and NprA, which were quantitatively determined in a study of several exoproteomes20,21,22. InhA1 plays a central role in the virulence of B. cereus sensu lato due to its interactions with both bacterial and host proteins in the course of infection. InhA1 and NprA both feature zinc-binding motifs and include amino acid residues of the active catalytic center common for metalloproteases (HEXXH). InhA1 is lethal when injected into the hemolymph of insects and is capable of cleaving antibacterial peptides, such as cecropin and attacin23. InhA1 also determines the ability of B. cereus cells to evade host macrophages24. The B. paranthracis strains described herein possess genes of three homologous proteins: InhA1, InhA2, and InhA3. NprA enables bacteria to escape from macrophages and therefore represents a key element required to combat the host's immune system. Mutants with inactivated nprA or inhA1 genes cannot escape undamaged if captured by macrophages. NprA degrades host cell components, which can explain the role this proteins plays in the consortium during the escape from macrophages. InhA1-mediated cleavage of NprA is a key step in enhancing NprA activity. Indeed, an active C-terminal domain of NprA suffices to stimulate the release of bacteria from host macrophages after phagocytosis25.

Bacteria of the consortium are constantly exposed to a flow of milk and therefore require the ability to form biofilms. In the biofilm state, they adhere to the udder wall and acquire resistance to stress factors and most antibiotics, while continuing to secrete various pathogenicity factors26. Protection provided by a biofilm enables the bacteria to resist not only the host's defense mechanisms but also the standard antibiotic therapy. In staphylococci, biofilm formation involves the genes of the intercellular adhesion operon (ica), icaADBC, responsible for the synthesis of polysaccharide intercellular adhesin (PIA), the principal component of the exopolysaccharide matrix surrounding bacterial cells within biofilms. Most biofilms found in natural environments are composed of several bacterial species. B. cereus sensu lato is no exception. It is frequently observed in associations with other microorganisms, when biofilms can be described as cooperative consortia where each partner contributes to stability and development of the community27.

Similarly to cytK28, individual genes involved in biofilm formation in B. cereus are regulated by the global transcriptional regulator PlcR together with the oligopeptide encoded by papR, since quorum sensing is critically important not only for toxin expression levels but also for biofilm formation in gram-positive microorganisms. In addition, bacterial adhesion to surfaces, intercellular interactions, cell aggregation, and biofilm formation in B. cereus involve genes of the eps2 cluster, whereas the eps1 gene cluster apparently participates in some sort of social motility of these bacteria. That is, the eps2 cluster is more important for adhesion to epithelial cells29.

In addition, the expression of enterotoxin genes is affected by the phase transition regulator SinR. SinR and PlcR regulate biofilm formation in an interactive manner. SinR was shown to control biofilm formation and swimming motility in B. thuringiensis30. Interestingly, only a small subpopulation of cells in the biofilms expressed hbl, which depended on the expression of sinI30. Two SinR recognition sites were detected upstream from the hbl operon, and one more SinR recognition site was identified upstream from the nhe operon31. It should be mentioned that the spo0A–sinI–sinR regulatory chain in B. cereus AR156 was also shown to be involved in biofilm formation, in cell differentiation, and in host–pathogen interaction32.

Methods

Strains, culturing conditions, and genomic DNA extraction

Microbial consortium 4M was isolated from milk of black-motley cows in the Moscow region. Microbial consortium 1702 was isolated from milk of Simmental cows in the Saratov region. Microbial consortium 1710 was isolated from milk of black-motley Holstein hybrid cows in the Republic of Mordovia. On the map, the sites of sample collection form a triangle with sides of approximately 600, 400, and 300 km (Fig. 4). Based on the results of veterinary and laboratory control, all animals included in the study were diagnosed with the clinical form of mastitis. To obtain pure cultures, milk samples were first inoculated as enrichment cultures in Salt Meat Broth, Azide Dextrose Broth, and Tryptone Soya Broth (HiMedia Laboratories Pvt. Ltd., India) and subsequently transferred onto differential diagnostic media: Baird Parker Agar (HiMedia Laboratories) and Azide Blood Agar Pronadisa (Conda, Spain). For all cultures studied, preliminary microscopy of individual colonies from Baird Parker Agar showed cells of different size and shape (cocci and bacilli) and of the same positive Gram staining. To isolate individual strains and obtain uniform colonies, a series of five passages on the same differential diagnosis medium was performed. After each passage, microscopic analysis revealed the same two types of cells in the field of view.

Figure 4
figure 4

Geographical location of sample collection sites on the map of Western Russia.

Full size image

Cells from a pure, fresh 50 mL liquid culture from each consortium were harvested by centrifugation at 15,000×g for 3 min and washed three times with sterile Milli Q water under sterile conditions. Total genomic DNA was extracted using the GeneJET Genomic DNA Purification Kit (Thermo Fisher Scientific, Bremen, Germany) following the manufacturer’s instructions. Quality and quantity of the extracted DNA was determined using QubitTM fluorometric quantitation and NanoDrop 2000 (both Thermo Fisher Scientific, Bremen, Germany).

Genome sequencing, assembly and annotation

The genome of both strains in each consortium was sequenced on an Illumina MiSeq platform with a read length of 300 bp (paired end) and Oxford Nanopore MinION. The genome from each member of consortia was assembled using Unicycler from both short and long sequencing reads33. Genome quality estimation was determined with CheckM. Genome circularization was checked manually using BLAST. Genome annotation was performed with Prokka34. Protein coding genes were classified based on the annotation into Cluster of Orthologous Groups (COG) functional categories with the automatic classification COG tool at MicroScope platform35. The data for this study have been deposited in the GenBank under project number PRJNA865942 with accession numbers CP102659-CP102662, and CP113365-CP113402.

Genomes comparison

Classification of the genomes was determined according to the Genome Taxonomy Database (GTDB) using the GTDB-tool kit (GTDB-tk) v.1.1.0 integrated in the MicroScope web-based service35. GTDB-tk provides a taxonomic classification of bacterial and archaeal genomes based on the combination of the GTDB reference tree, the relative evolutionary divergence and the ANI value against reference genomes36. CJ Bioscience's online Average Nucleotide Identity (ANI) calculator was used to compare two prokaryotic genome sequences37. Whole-genome alignments were performed by Mauve v. 2015022638.

Ethics statement

The animal study was reviewed and approved by L.K. Ernst Federal Science Center for Animal Husbandry Bioethics Committee, protocol number 2021-3/4. All methods were carried out in accordance with the approved guidelines and relevant regulations. All methods are reported in accordance with ARRIVE guidelines.