Contagious agalactia (CA) is an infectious syndrome with an important socioeconomic impact on the small ruminant dairy sector due to negative effects on milk production, premature culling, lessen growth rates, and the high costs of control measures. It is characterized by a triad of clinical manifestations: mastitis, arthritis, and keratoconjunctivitis, but can occasionally affect the reproductive and respiratory tract (1–3). It is a multi-etiological syndrome as four different species from the genus Mycoplasma are involved in goats: Mycoplasma agalactiae (Ma), Mycoplasma mycoides subsp. capri, Mycoplasma capricolum subsp. capricolum, and Mycoplasma putrefaciens. Ma is considered as the main etiological agent that affects goats and sheep, as the other three species of mycoplasmas have only been described sporadically as the cause of the disease in the ovine specie (4, 5).

Nowadays, the fight against CA is based on vaccination and antibiotic therapy but the absence of a satisfactory strategy causes difficulties to eradicate CA in endemic regions. In Spain, a national voluntary program based on an accurate diagnosis and the control of the disease has been put in place (6). On one hand, vaccination against CA has its limitation; while commercial vaccines can reduce symptoms and excretion (3), it does not prevent shedding in milk (7) and therefore the carrier state persists (8). Different explanations have been suggested for the lack of an efficient vaccination such as the complex etiology in goats, the high plasticity of the genome of circulating strains or their capacity to evade the immune system (3). In this sense, the development of vaccines that can prevent satisfactorily the infection in flocks or the entrance in areas free of CA does not seem to be a short- and medium-term achievement.

On the other hand, antimicrobial therapy can improve the animals’ health, but it does not eliminate the pathogen (8). It is assumed that antimicrobial agents can reduce the bacterial excretion and clinical symptoms. Nevertheless, the use of antimicrobial agents can generate antimicrobial resistances (AMR), which can compromise the effectiveness of the antimicrobial therapy (3). Indeed, several studies have reported a reduction in the antibiotic susceptibilities of the mycoplasma species associated with CA in different countries (9–16).

In this context surrounding the control and prevention of CA, the necessity to explore alternative therapies, such as the use of probiotics in recent years in people and animals, has emerged. Probiotics are live microorganisms which when administered in adequate amounts confer a health benefit on the host (17). Lactic acid bacteria (LAB) offer various advantages as potential probiotics and can be considered as alternatives to antibiotics (18). They are safe microorganisms able to produce different compounds such as bacteriocins, organic acids as lactic acid, hydrogen peroxide, diacetyl, and carbon dioxide that favor the inhibition of pathogenic microorganisms. Lactic acid bacteria are Gram-positive bacteria, they can be found in the microbiota of various anatomical locations such as the oral cavity, the skin, the gastro-intestinal tract, and the reproductive tract (19–23). Their presence in the raw milk of small ruminants is well known (24), and some strains have been tested in vitro for their potential probiotic characteristics (25, 26). As far as we know, LAB isolated in small ruminants have never been challenged against pathogens belonging to Mycoplasma spp.

In a previous study, a first dose of a commercial vaginal probiotic for women “L1” was intravaginally inoculated in ewes in order to prevent the vaginosis produced by the use of intravaginal devices. This study reported the capacity of L1 to reduce the vaginal neutrophilia produced by estrus-synchronization sponges without altering the animal health status (27). In addition, a higher dose “L2” of this commercial probiotic has been tested in vitro against Mycoplasma bovis (Mb) in bovine semen and cervical mucus and showed antimicrobial activity against the pathogen. This antimicrobial activity of Lactobacillus spp. was associated to their capacity of acidifying the medium (28, 29). In this sense, the in vitro sensitivity of Ma to acid pH has been reported in diluted semen of bucks (30). Mycoplasma bovis shares 99% of its genome with Ma (31) and both belong to the hominis group, sharing relevant similarities of intrinsic AMR and therefore control measures (32, 33).

The aim of this microbiological study was to evaluate the in vitro antimicrobial potential of lactic acid bacteria, isolated from ovine and caprine raw milk, against Mycoplasma agalactiae and compare it with the efficacity of the commercial probiotic L2 dose. To achieve this objective, the viability of Mycoplasma agalactiae and lactic acid bacteria as well as the extracellular pH oscillations were evaluated in commercial goat milk and in a Mycoplasma spp. specific culture medium.

The study design included various steps. The first one was the sampling of 72 animals from nine different farms. The second step involved the isolation of LAB from the raw goat and sheep milk obtained in the first step and the evaluation of their in vitro growth capacity in a specific Mycoplasma culture medium. The third step consisted in the molecular characterization of the selected strains. The final step was the carrying out of the in vitro experiment to assess the antibacterial activity of the different LAB against Ma. All the results from the in vitro experiment were statistically analyzed a posteriori. In addition, we also analyzed the LAB composition of L2 overtime.

The present in vitro study reports the antimicrobial effect against Ma of a selection of wild LAB isolates from the milk of healthy sheep and goats. These strains were isolated in herds located in the mainland of Spain (Table 1), an area where etiological agents associated with CA have been frequently isolated in ovine and caprine species (4, 42). Contagious agalactia control and prevention represent a challenge due to several factors: presence of asymptomatic carriers, uncontrolled movement of animals, variability in etiology and antigenicity, the limitations of commercially available vaccines and the increasing AMR of mycoplasmas associated with CA (1–3, 10). Our results suggest that the commercial probiotic used in this study, based on a combination of Lactobacillus spp., or wild LAB of ovine and caprine origin could have the potential of being used as antimicrobials for the control or prevention of mastitis caused by Ma.

Our work evinces that an important number of bacterial isolations is necessary in various flocks to obtain LAB strains capable of growing in a culture medium that allow their in vitro testing and demonstrate a possible commercial use (Table 1). All the three wild strains of LAB involved in the in vitro experiments were able to inhibit the growth of Ma in GM. Strain 248D had a bacteriostatic effect as it did not significantly decrease the number of Ma at T15, but it did prevent its ease to replicate and increase its concentration in GM at 37°C (Table 5, C5 and C1). Strains 33B and 120B were able to significantly reduce the concentration of Ma at T15 in GM (Tables 4, 6; C5) although the inhibition by 33B was significantly greater than the inhibition produced by 120B (p < 0.001).

Lowering the pH is an important feature of LAB as it can inhibit the growth of pathogenic bacteria (43). The acidification of the medium has been suggested to inhibit Ma and M. mycoides subsp. capri. in diluted semen of bucks as these species are sensitive to pH changes (30). Therefore, one of the causes of the inhibition produced by these LAB may be the drop in the pH of the GM they produced which does not occur when the GM only carries Ma and so the pathogen increases its concentration (C1). All the wild ovine and caprine strains tested in the in vitro experiments were able to acidify the GM (Tables 4–6; C4) as there was a significant difference between the GM pH of T0 and T15.

Nevertheless, the strain 33B, identified as E. mundtii, was able to inhibit Ma in PH medium (Table 4, C10) with a pH close to neutral and could therefore show better antimicrobial capacity than L2 (Tables 5, 6, C8) in environments where the pH is neutral, and the acidification of the medium is not possible. This suggests that pH acidification may not be the only antimicrobial effect of LAB against Ma and that other antimicrobial mechanisms should be sought.

Probiotics bacteria have several mechanisms of action to inhibit pathogenic bacteria in vivo: competing for nutrients, preventing the adhesion of the pathogens, producing inhibitory substances, modulating the host immune response, and reducing the bioavailability of toxins (18). It is unlikely that LAB and Ma compete for the same nutrients given that LAB use glucose to produce lactic acid (44) and Ma cannot ferment glucose unlike other species such as M. mycoides subsp. capri, M. capricolum subsp. capricolum and M. putrefasciens (30). Therefore, we propose the hypothesis that these bacteria could have a greater inhibitory effect against sugar-fermenting mycoplasma species. In the case of Ma, the production of inhibitory substances in vitro could be one of the antimicrobial mechanisms used by LAB, in addition to the harmful effect produced by acidification of the extracellular pH, given that the inhibition in PH medium (Table 4, C10) occurred without a medium acidification for the strain 33B.

One of the inhibitory substances produced by LAB are bacteriocins, and raw milk can be considered as a source of LAB strains with bacteriogenic potential (45, 46). The E. mundtii strain CRL 1656 isolated from cow’s milk has been reported as bacteriocin-producing strain and showed a bacteriocigenic activity against the pathogen Listeria monocytogenes Scott A and L. innocua 7. This strain also able to produce a good amount of hydrogen peroxide, another inhibitory substance produced by LAB. Its use as a probiotic in cows has been recommended (46). Another strain, E. mundtii EM ML2/2, isolated from raw goat milk, produced a bacteriocin substance and showed an optimal activity at pH 6.3 (47).

The two other LAB strains, both isolated from meat sheep, with a bacteriostatic (248D) and bactericidal (120B) potential were identified as Enterococcus hirae. These results evince that different antimicrobial effects against Ma can be observed for different strains of same LAB specie. Other strains of E. hirae ST57ACC and DF105Mi have shown antimicrobial activity against L. monocytogenes by producing bacteriocins capable of resisting food processing (25, 45). A strain isolated from GM was also able to modulate the gut microbiota in dogs and did not present any virulence gene (43).

Regarding the evaluation of the commercial probiotic, the addition of L2 in GM (C3) showed a significantly higher bactericidal activity (p < 0.001) against Ma than that observed with strain 33B and 120B in GM (Tables 4, 6; C2-C5). This could also be related to the significant pH decrease observed throughout all experiments. Indeed, L2 significantly reduced the pH of the GM below five, when with Ma (Tables 4–6, C3), while strains 33B, 120B, or even 248D lowered the pH to values between 5.14 and 5.43 (Tables 4–6, C5). In previous studies, a similar inoculum was evaluated in vitro against Mb in bovine diluted semen and cervical mucus of cattle, and a significant reduction in the pH was also observed (28, 29). Consistent with these studies, our results showed that L2 can also grow and acidify the extracellular medium in GM even when contaminated by Ma and could be a tool used as an antimicrobial strategy as it has been suggested (30). However, as mentioned previously, other possible influences such as competition for nutrients or the possible presence of bioactive peptides should not be ruled out as an antimicrobial mechanism of LAB (18).

These data regarding L2 could show a possible increase in the antimicrobial potential against Ma when several species of Lactobacillus spp. are used together as probiotics in GM. In this sense, the combination of various LAB strains is usually employed in commercial probiotics due to their synergy that increases their biological activity (48). The exact composition of this inoculum or one of similar composition had not been evaluated in previous studies (28, 29). In the present study, metagenomic analysis of the conditions with L2 evidenced for the first time, the real composition of this inoculum developed from a commercial probiotic for human use. The results showed that indeed, three species of Lactobacillus spp. are inoculated with our protocol (Figure 1). Our metagenomic study of the dynamics of the three species of Lactobacillus spp. of L2 showed that L. brevis increased its concentration to the detriment of L. crispatus in GM contaminated with Ma or not while L. gasseri had a steady RA over time. This provides a first approximation of the dynamics of these lactobacilli species in two different media and the possible role of L. brevis in the inhibition of Ma in GM. This specie has been isolated in raw milk of goats (49) and seems to have an antimicrobial effect against several pathogens such as Bacillus cereus (50), Escherichia coli, S. aureus, Klebsiella pneumoniae, and Pseudomonas aeruginosa (51). Nevertheless, it was reported that L. brevis was unable to acidify milk during a 20.5 h fermentation at 37°C (50) and could therefore not be responsible for the significantly lower pH observed at T15 in GM (C2-C3) and use different mechanisms to inhibit pathogens. Lactobacillus gasseri has also been isolated in caprine raw milk (52, 53) and seems to be a main component of the human vaginal flora, as well as L. crispastus (54), although the latest has never been isolated in milk to our knowledge.

Generally, this work suggests the antimicrobial potential of LAB against Ma under in vitro conditions, an important pathogen of the mammary gland of small ruminants. The necessity to explore possible applications of LAB, present in the microbiota of the mammary gland, as a control and prevention strategy against small ruminants’ mastitis was previously suggested (24). Different studies have demonstrated the positive effect of LAB and its metabolites on the welfare of farm animals. It has been shown that the use of probiotics based on LAB reduces the occurrence of pathogens in large-scale farms (55). The results of the present in vitro study would suggest the need to inoculate, in vivo in caprine and ovine models, the strains identified in this study with an antimicrobial potential against Ma. In this sense, a preliminary study developed an intravaginal inoculation method in ewes, with doses inferior to L2 of the commercial probiotic used in this study, which showed the first signs of anti-inflammatory effects and had no prejudicial effects on the animals’ health (27).

On the other hand, from an epidemiologic point of view, our results show that LAB with a negative effect against Ma can be naturally present in the mammary gland of ewes (248D, 120B) and goats (33B) from endemic regions of CA (Table 1). In all the three herds where the strains with antimicrobial potential were isolated, the use of antibiotics was anecdotic. The herds where 33B and 248D were isolated did not have any CA outbreaks, at least in the last decade, although they did manifest symptoms compatible with CA in the past. On the contrary, the ovine flock where 120B was isolated, had a clinical history of CA a year before this study took place. It is known that after a clinical outbreak of CA, the affected herds usually become chronically infected. This is normally attributed to an equilibrium created between the host and the pathogen, depending on the immune status of the herd. Moreover, it is accepted that the infection is not usually eliminated after the use of antibiotics and vaccines (2, 56, 57). Our results show the existence of LAB with antimicrobial potential against Ma in a CA chronically infected herd (strain 120B, Herd B, Table 1). Curiously, approximately one year after of this isolation, a new episode of decreased milk production was observed in this herd in animals where Ma was isolated again but no LAB was isolated. Therefore, the isolation of LAB never coincided with that of Ma and vice versa. In the herd that had a clinical outbreak of CA at the time of this study (Table 1, herd I), LAB were not isolated either. We suggest the hypothesis that this type of bacterial population (LAB) could contribute to the maintenance of the apparent asymptomatic status of a high number of animals in infected flocks. Furthermore, we need to consider that pathogenic species of Mycoplasma in ruminants such as the ones associated to CA (3, 58, 59), in asymptomatic animals, are usually found in anatomic locations such as articular liquid, lymph nodes, brain or external auditive canal, perpetuating the infection in the herds. We propose that with this strategy the pathogens not only try to avoid the immune system and the antimicrobial therapy (1, 2) but also the cohabitation with bacterial groups with antimicrobial potential such as the LAB. Indeed, these LAB populations can be found in the microbiota of the epithelium of the respiratory, mammary and reproductive tracts (23, 24, 60), which are anatomical locations that are colonized by mycoplasma associated with CA and linked to excretion route (34). In this sense, in a previous study involving Salmonella sp., the isolation of LAB was less important in dogs that were positive to this pathogen (21). Based on this hypothesis, the use of antibiotics could harm the natural barrier, that LAB with antimicrobial capacity represent, in locations such as the mammary gland of small ruminants. The results reported here could be the first indication of an undervalued interaction of LAB with other microbial agents, such as Ma, and suggests the need to carry out new studies on the bacterial ecology in CA infected animals.

In conclusion, this study marks the first description of the antimicrobial potential of LAB against Ma, hence a possible new alternative to the antibiotics used for the control of CA. To the authors’ knowledge, the assessment of the antimicrobial potential of wild LAB against mycoplasmas of the hominis group has not previously been reported. In this sense, the inoculum L2, elaborated from a human commercial probiotic based on Lactobacillus spp., evinces itself as a strategy capable of achieving the complete inhibition of Ma in vitro in GM. The presence of E. hirae and E. mundtii is also confirmed in ovine and caprine milk with an in vitro bacteriostatic or bactericidal capacity against Ma in milk. The interaction between LAB and Ma reported here suggests a possible role of LAB in the dynamics of mycoplasmosis that should be studied. Our results suggest the necessity to design further in vitro studies to characterize other aspects of these LAB strains, such as other functional properties, bio-preservation and safety, as well as try to understand the inhibitory mechanisms, in order to corroborate their probiotic potential. In addition, in vivo studies would be needed to confirm its antimicrobial potential against mycoplasmas associated with CA and its innocuity on animals’ health.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: https://www.ncbi.nlm.nih.gov/; OQ538168, OQ538169, and OQ538170.

MT, EB, and ÁG-M designed the study and wrote the manuscript with input from all authors. MT, JG, RT-P, and ÁG-M collected the samples. MT and RT-P processed and analyzed the samples. MT, EB, JG, RT-P, and ÁG-M performed the laboratory experiments. MT recollected and prepared the data. AS analyzed the data. MT, EB, AS, JC, CF, and ÁG-M interpreted the data. All authors contributed to the article and approved the submitted version.

This work was supported by Generalitat Valenciana (Spain; GVA/2020/026) and the Spanish Ministry of Science and Innovation (PID2020-119462RA-I00/AEI/10.13039/501100011033). MT is supported by a pre-doctoral contract of the CEU-UCH, RT-P by a pre-doctoral contract of the Generalitat Valenciana (CIACIF/2021/245), and ÁG-M by a “Ramón y Cajal” contract of the Spanish Ministry of Science and Innovation (RYC2021-032245-I).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.