Mastitis is the most common and multifactorial disease in dairy animals, especially in cattle. Subclinical form of mastitis is the major importance in comparison to clinical mastitis due to its hidden nature and incurring severe economic loss to dairy farmers. In most of the studies conducted on mastitis, more than 140 microorganisms have been found as the causative agents for mastitis. Among them, the major infectious agent is S. aureus involving more than 40% cases under contagious pathogens followed by E. coli under environmental pathogens category (Athar et al., 2007).
       
In subclinical form of mastitis, the use of antibiotics cannot resolve the disease and the condition becomes chronic. Continued treatment in chronic cases increases the risk of antibiotic resistance, which is a major threat to human and animal health (Ramasamy et al., 2021 and Regitze et al., 2021). The reasons for the failure of treatment in mastitis is bacteria producing beta-lactamase and further complicated by the bacteria embedded in a self-produced matrix of extracellular polymeric substances called biofilm in the udder (Saglam et al., 2017; Schonborn et al., 2017). Biofilm is an ancient prokaryotic adaptation by altered gene expressions, antibiotic resistance, nutrition utilization and virulence factors. Biofilms not only helps in attach to abiotic surfaces but also attach to the surrounding mucus layer or host fluids through adhesins like encoding clumping factors (clfA and clfB) and fibronectin-binding proteins (fnbA and fnbB) (Zhang and Mah, 2008; Ramasamy et al., 2021; Ali et al., 2024).
       
In biofilm forming bacteria, there is an up regulation of various genes involved in anaerobic respiration and fermentation. Among the anaerobic genes, Pyruvate Formate Lyase (pfl) gene is fourfold highly expressed (Melchiorsen et al., 2002). Under glucose-limited anaerobic conditions, Pyruvate Formate Lyase (PFL) (E.C. 2.3.1.54) protein plays a central role in the shifting homolactic to mixed-acid metabolism in which formate, acetate and ethanol are produced. In our laboratory, the association of PFL protein with biofilm forming mastitis was demonstrated by Anitha et al. (2014) and also the level of pfl gene expression in biofilm forming S. aureus isolated from mastitis cases was demonstrated by Manasa et al. (2020). Association of expression of PFL protein and subclinical mastitis showing promising results to make it as a candidature target towards the diagnosis and/or treatment. In this process, structural elucidation of PFL protein across the mastitis causing organisms becomes important to establish the PFL proteins either as a novel drug target or as a diagnostic marker.

Collection of milk samples from subclinical mastitis cases
 
Milk samples were collected from cattle presented at the Veterinary Clinical Complex (VCC), Rajiv Gandhi Institute of Veterinary Education and Research (RIVER), Puducherry and screened for subclinical mastitis using the California Mastitis Test (CMT). A total of 45 CMT-positive milk samples were obtained from 35 animals during the study period (December 2020 to May 2021). The samples were maintained at 4^oC and processed at the Department of Veterinary Biochemistry, RIVER.
 
Isolation and identification of S. aureus and E. coli from subclinical mastitis milk
 
Collected milk samples were subjected to bacterial enrichment by inoculating loopful of milk into Luria Bertani (LB) broth (HIMEDIA) and incubated at 37^oC for 18 h (ORBITEK). Enriched cultures were streaked and incubated at 37^oC for 24 h on mannitol salt agar (HIMEDIA) and MacConkey agar (HIMEDIA) for mannitol fermenters and lactose fermenters, respectively. Then, the colonies were subjected to Gram’s staining procedure. Further, the isolates were subjected to molecular identification by Polymerase Chain Reaction (PCR).
 
Molecular identification of S. aureus and E. coli
 
A single colony from the cultures streaked on Mueller Hinton agar (HIMEDIA) was inoculated into 2 ml LB broth and incubated overnight at 37^oC. After incubation, DNA was extracted from the bacterial growth by the method of Christensen et al. (1993). S. aureus and E. coli were identified by detecting the presence of nuc gene and alr gene, respectively (Brakstad et al., 1992; Yokoigawa et al., 1999) by PCR.
 
Screening the isolated organisms for biofilm formation
 
Presence of biofilm associated genes, bap gene of S. aureus and ndvB gene of E. coli were identified by PCR using the primers as described by Cucarella et al. (2004) and Mohammed et al. (2014), respectively. The isolates which possess the biofilm associated genes were further subjected to in vitro biofilm assay by TCP method which is a quantitative and gold standard method for the biofilm detection (Neopane et al., 2018 and Mishra et al., 2026). Loopful of culture was inoculated into 2 ml of LB broth (HIMEDIA) with 1% glucose and incubated overnight at 37^oC. Then, 200 μl of 1:100 diluted culture was added to sterile 96 well flat-bottom polystyrene plate (TARSONS) in triplicates. The control well was processed in a similar manner using sterile distilled water and medium. The plate was then incubated at 37^oC for 24 h. After incubation, the contents of each well were removed by gentle tapping. The wells were washed with 200 μl of phosphate buffer saline, pH 7.3 to remove free-floating bacteria. Biofilms formed by bacteria adherent to the wells were fixed by 99% methanol and stained with 0.1% crystal violet for 15 minutes at room temperature. Excess stain was washed gently and the plate was air dried. The optical density (O.D) of the stained adherent biofilm was measured using a micro-ELISA auto-reader (TECAN) at a wavelength of 620 nm. The isolates were classified based on biofilm forming capacity as described by Halim et al. (2018).
 
Amplification and sequencing of pfl gene S. aureus and E. coli
 
Among the biofilm-producing isolates, three isolates each of S. aureus and E. coli demonstrating the highest biofilm formation, based on optical density measurements obtained from the microtiter plate assay, were selected for full-length amplification of the pfl gene using primer walking with in-house designed primers. The pfl gene sequence of S. aureus (Acc. No: NC_007795) and E. coli (Acc. No: NC_000913) were retrieved from GenBank. Three sets of primers were designed for each species (Ye et al., 2012) using primer BLAST server (https://www.ncbi.nlm.nih.gov/tools/primer-blast/) to amplify the pfl gene with overlapping regions to cover the entire coding sequence. The primers (Table 1) were custom synthesized from Eurofins Genomics India Pvt. Ltd., Bengaluru, India.


       
The pfl gene of both S. aureus and E. coli was amplified by PCR. Twenty microliter reaction mixture was set up using 200 ng of template DNA and 10 pmol of forward and reverse primer. The PCR conditions for amplification of pfl gene of S.aureus consist of primary denaturation at 95^oC for 5 min followed by 30 cycles of denaturation at 95^oC for 20 sec, annealing at 58^oC for 40sec (set I and II primers), 59^oC for 40 sec (set III primers), elongation at 72^oC for 60 sec and final elongation of 72^oC for 10 min. The PCR conditions for amplification of pfl gene of E. coli consist of primary denaturation at 95^oC for 5 min followed by 30 cycles of denaturation at 95^oC for 20 sec, annealing at 58^oC for 40 sec (set I primers), 56°C for 40 sec with (set II primers), 59^oC for 40 sec (set III primers), elongation at 72^oC for 60 sec and final elongation of 72^oC for 10 min. Amplified PCR products were analysed in 2% agarose gel electrophoresis.
 
Multiple sequence alignment studies of pfl gene
 
Amplified pfl gene products of the three isolates of each species were custom sequenced at Eurofins Genomics India Pvt. Ltd., Bengaluru. The obtained sequences were subjected to blast analysis by NCBI – Nucleotide BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi?PROGRAM=blastn andPAGE_TYPE=BlastSearchandLINK_LOC=blasthome) to confirm the specificity of amplicon (Altschul et al., 1990).
       
Individual amplicon sequences were constructed from the sequence generated using forward and reverse primers. The full-length coding region of pfl gene sequence was reconstructed based on overlapping regions of each pfl set using BioEdit sequence alignment Editor (version 7.2.5) tool.
       
Full length, coding region of pfl gene nucleotide sequences were subjected for multiple sequence alignment (Clustal Omega) with their respective reference sequences (S. aureus - Acc. No: NC_007795 and E. coli - Acc. No: NC_ 000913) and the nucleotide variations within the species were analysed (https://www.ebi.ac.uk/Tools/msa/clustalo/).
 
Multiple sequence alignment of PFL protein sequence
 
pfl gene of both S. aureus and E. coli isolates of Puducherry region was in silico translated to get the Primary protein structures using ExPASy server (https://web.expasy.org/translate/).
       
Primary protein structures were subjected to multiple sequence alignment (Clustal Omega) with their respective reference sequences (S. aureus - Acc. No: NC_007795 and E. coli - Acc. No: NC_000913) and change in the amino acid sequences were analysed (https://www.ebi.ac.uk/Tools/msa/clustalo/).
       
Consensus sequences of PFL proteins were created for both S. aureus and E. coli isolates of Puducherry region using BioEdit sequence alignment editor (version 7.2.5) tool.
 
PFL protein structure prediction
 
The consensus protein sequences of both S. aureus and E. coli isolates were subjected to secondary protein structures prediction and the modeled structures were validated using Swiss-Model server. (https://swissmodel. expasy.org/interactive). Ramachandran plot analysis was carried out using MOLPROBITY server (http://molprobity. biochem.duke.edu/) and the structural stability was analysed by QMEAN scoring.
 
Superimposition analysis of PFL protein
 
The homology modeled PFL protein of S. aureus and E. coli of Puducherry isolates were validated by superimposing with their respective reference structures (S. aureus - Acc. No: NC_007795 and E. coli - Acc. No: NC_000913) for the identification of structural alignment and the difference in folding pattern using UCSF Chimera server (https://www.cgl.ucsf.edu/chimera/).
 
Superimposition of PFL protein structure between the species
 
The consensus primary structure of PFL protein of S. aureus and E. coli of Puducherry isolates were subjected to multiple sequence alignment (Clustal Omega) (https://www.ebi.ac.uk/Tools/msa/clustalo/). The homology models of PFL proteins of S. aureus and E. coli of Puducherry isolates were superimposed with each other for the identification of structural alignment and the difference in folding pattern between the two species using UCSF Chimera server (https://www.cgl.ucsf.edu/chimera/).

Isolation and identification of S. aureus and E. coli from subclinical mastitis milk samples
 
Milk samples collected from subclinical mastitis cases were used for the isolation of S. aureus and E. coli. Out of 45 collected samples, 25 samples had shown yellow colour colonies (mannitol fermenters) on mannitol salt agar and 8 samples have shown pink colour colonies (lactose fermenters) on MacConkey agar. On Gram’s staining, the colonies of mannitol fermenters appeared as Gram positive cocci arranged in clusters (Staphylococcus sp.). The colonies of lactose fermenters appeared as Gram negative rods (Coliform sp.).
 
Molecular identification of S. aureus and E. coli
 
Gram positive mannitol fermenters and Gram-negative lactose fermenters were subjected to molecular confirmation by PCR using species specific primers of nuc gene for S. aureus and alr gene for E. coli, respectively. Out of 25 Staphylococcus sp. and 8 coliform sp. subjected for PCR, 13 were found to be S. aureus and 7 were found to be E. coli (Fig 1 and 2).




 
Screening the isolates for biofilm formation
 
S. aureus and E. coli isolated from subclinical mastitis cases were analyzed for biofilm formation by screening for the presence of biofilm associated genes and also phenotypic expression of biofilm formation on tissue culture plates.
       
PCR positive isolates of S. aureus and E. coli were further screened for the presence of biofilm associated genes by PCR using the primers targeted to bap gene of S. aureus and ndvB gene of E. coli. Out of 13 S. aureus and 7 E. coli isolates, 5 were positive for bap gene and 4 were positive for ndvB gene, respectively (Fig 3 and 4).




       
The isolates found positive for the presence of biofilm associated genes by PCR were further subjected to in vitro biofilm assay using TCP method. Four out of five isolates of S. aureus were found as strong biofilm producers while one was found to be moderate biofilm producer (Table 2). All the four isolates of E. coli were strong biofilm producers (Table 3). From both S. aureus and E. coli isolates, top three biofilm producers were selected for full length amplification of pfl gene by primer walking process using in-house designed primers.




 
Amplification of pfl gene of S. aureus and E. coli
 
Standardized PCR protocol was used for the amplification of pfl gene of S. aureus and E. coli using three sets of respective overlapping primers. Amplified PCR products were analysed in 2% agarose gel electrophoresis and it had shown the expected product sizes of 830 bp, 851 bp and 783bp for S. aureus and 849 bp, 810 bp and 899 bp for E. coli as shown in Fig 5 (A, B, C) and Fig 6 (A, B, C) respectively.




       
The specificity of the PCR products was confirmed by custom sequence using forward and reverse primers separately. The sequenced results were subjected for BLAST analysis revealed 100% identity with the GenBank entries of pfl gene sequences.
 
Reconstruction of full length pfl gene sequence of S. aureus and E. coli
 
The sequence of each amplicon was generated from both the directions based upon the sequence obtained from both forward and reverse primers. The entire length of coding region of pfl gene sequence was reconstructed based on the overlapping sequence regions of individual pfl amplicons using BioEdit sequence alignment editor (version 7.2.5) tool. The full length pfl gene sequences were submitted in GenBank and accession numbers listed in Table 4.


 
Multiple sequence alignment of nucleotide sequence of pfl gene of S. aureus and E. coli
 
The coding region of pfl sequence of each isolate of S. aureus and E. coli were subjected to multiple sequence alignment with reference to their respective reference sequence. In S. aureus, a total of 16 mutations were observed, which corresponds to 11 nucleotide positions (Table 5). Out of 16 mutations in S. aureus, 9 transition mutations and 7 transversion mutations were noticed. Similarly, in E. coli, a total of 13 mutations were observed, which corresponds to 13 nucleotide positions (Table 6). Out of 13 mutations in E. coli, 7 transition mutations and 6 transversion mutations were noticed.




 
Multiple sequence alignment of amino acid sequences
 
The entire coding region of the individual nucleotide sequence were subjected to in silico translation using ExPASy Translate tool. The amino acid sequence of the PFL protein of the Puducherry isolates were subjected for multiple sequence alignment with their respective reference sequences using Clustal Omega server and the amino acid variations within the species were analysed. The identity score was predicted by multiple sequence alignment analysis using residue matches and each of the sequences were more than 99% of identity with the reference sequence. A total of 3 variations in S. aureus isolates and 7 variations in E. coli isolates were observed.
       
The codons corresponding to mutations and the details of amino acid variations for both S. aureus and E. coli were listed in Table 7 and Table 8. Out of 16 mutations of S. aureus, 12 were synonymous mutations and 4 were nonsynonymous mutations whereas in E. coli, out of 13 mutations, 6 were synonymous mutations and 7 were nonsynonymous mutations.




 
Protein structure prediction
 
The homology model prediction analysis was carried out using Swiss-Model server for the consensus sequence of PFL protein obtained by BioEdit sequence alignment editor (version 7.2.5) tool. In this study, PDB: IMZO.1. A. model was used as a parental template to model the PFL proteins of S. aureus (Fig 7) and E. coli (Fig 8).




       
The secondary structures were validated using Ramachandran plot analysis from MOLPROBITY server. Ramachandran plot analysis revealed that 94.3% and 98.6% of the residues were in the favored and allowed region for PFL protein of S. aureus, respectively (Fig 9). Similarly, PFL protein of E. coli had 95.3% and 99.8% of the residues in the favored and allowed region, respectively (Fig 10). Structural stability of PFL protein of both S. aureus and E. coli was analysed by Local Quality Estimate (Fig 11A and 11B) and QMEAN scoring (Fig 12A and 12B) revealed that all the amino acid residues are falling in narrow range (between 0 to 1.5) indicating predicted protein structures are highly stable.








       
The predicted models were visualized by Swiss-Model server for the 3D protein structures (Fig 13). Further 3D models were superimposed with their respective reference structures for the identification of structural alignment and the difference in folding pattern by UCSF chimera server. PFL protein of S. aureus and E. coli homology model structures were 99% superimposed with respect to their reference structures (Fig 14 and 15). Therefore, the homology models were highly stable and sharing similar domains with identical active site.






       
Primary PFL protein structure of S. aureus and E. coli revealed 66% identical with each other on Multiple sequence alignment (Clustal Omega) (Fig 16). PFL protein of S. aureus and E. coli were superimposed to study the structural similarities across the species by UCSF chimera tool. In spite of 66% identity in primary structure, pfl proteins shared 85% of their tertiary structure on superimposing the homology models of (Fig 17A). The active sites are conserved in both the species with the two consécutive cysteine amino acids at 418, 419 positions in E. coli and 413, 414 positions in S. aureus with glycine residues at 734, 724 positions, in E. coli and S. aureus, respectively (Fig 17D). Two additional regions in E. coli, “ENGVNL” and “FHHEA” corresponding to the positions 209-214 and 694-698 as showed in Fig 17B and Fig 17C forms a non-superimposed loop in the tertiary structure of superimposed structures of S. aureus and E. coli. The similar motif site residue “VASTITSHDAGY” which was present in both the sequences of S. aureus and E. coli were shown in Fig 17E. Likewise, several sites were matched highly between the PFL proteins among the two species.




       
Subclinical mastitis caused by bacterial pathogens such as S. aureus and E. coli remains a significant challenge in dairy production due to its persistent nature and impact on milk quality and animal health. The ability of these pathogens to form biofilms contributes substantially to their persistence within the host. Biofilm formation facilitates bacterial adhesion, protection from host immune responses and increased tolerance to antimicrobial agents, thereby promoting chronic infection.
       
Within the biofilm microenvironment, bacteria often encounter oxygen-limited conditions that necessitate metabolic adaptation for survival. Under such conditions, bacterial cells shift from aerobic metabolism to alternative anaerobic metabolic pathways. The pfl gene encodes pyruvate formate lyase (PFL), a key enzyme involved in anaerobic metabolism that catalyzes the conversion of pyruvate to acetyl-CoA and formate. Previous studies have shown that pfl expression is significantly increased under anaerobic conditions, indicating its role in supporting bacterial survival within biofilms (Melchiorsen et al., 2002; Fuchs et al., 2007).
       
Sequence analysis of the pfl gene among the selected isolates revealed the presence of nucleotide variations, although not all mutations resulted in amino acid substitutions. Synonymous mutations, while not altering the amino acid sequence, may still influence gene expression by affecting mRNA stability or translation efficiency (Chamary and Hurst, 2005). Such variations may therefore contribute to regulatory differences in the expression or activity of the PFL enzyme.
       
Structural modelling of the PFL protein demonstrated that the predicted protein structures of both S. aureus and E. coli isolates were highly conserved when compared with their respective reference structures. Although differences were observed in the primary amino acid sequences between the two species, the tertiary structural organization remained highly similar. Importantly, the catalytic residues responsible for enzyme activity were conserved in both organisms, suggesting that the functional integrity of the enzyme is maintained despite sequence variability.
       
The conservation of the active site residues and overall structural stability of the PFL protein indicate that this enzyme plays a crucial role in bacterial metabolic adaptation during biofilm formation. Therefore, PFL may serve as a potential target for therapeutic strategies aimed at disrupting anaerobic metabolic pathways essential for bacterial survival within biofilms. Inhibition of PFL activity could impair biofilm-associated persistence of S. aureus and E. coli in subclinical mastitis, thereby improving treatment effectiveness and reducing the likelihood of recurrent infections.

The present study demonstrates that PFL protein expression is associated with biofilm-forming mastitis pathogens, namely Staphylococcus aureus and Escherichia coli. This enzyme plays a key role in anaerobic metabolic pathways that enable bacterial survival under limited oxygen conditions within biofilms. The observed structural similarity and conserved active sites of PFL proteins in both species suggest that PFL could serve as a promising target for therapeutic intervention. Targeting this enzyme may disrupt biofilm-associated metabolic adaptation and thereby reduce the survival and persistence of these pathogens in mastitis infections.

The authors thank the Dean, Rajiv Gandhi Institute of Veterinary Education and Research (RIVER), Puducherry for providing necessary facilites to carry out research work.

 The authors declare that there is no conflict of interest.