Corynebacterium lactis: Antimicrobial Resistance and Impact on Invertebrate Model Systems

Abstract

Corynebacterium lactis was isolated from the skin abscess of a companion dog and from raw milk of a cow with unspecific mastitis. As information about the species was scarce and a zoonotic potential could not be excluded, we started a basic characterization of C. lactis strain RW3-42 with respect to antibiotic susceptibility and the response of invertebrate animal model systems to infection. C. lactis showed a number of antimicrobial resistances and is able to colonize Caenorhabditis elegans. In contrast, Galleria mellonella larvae were not impaired by C. lactis. Genome analyses of strain RW2-5 revealed the absence of toxin-encoding genes, and only a rather small number of other virulence factors were found, i.e., SpaA- and SpaH-type pili and the non-fimbrial adhesins DIP0733, DIP1281, DIP1621 and EmbC in addition to a homologue of Salmonella RhuM involved in killing of C. elegans. The results obtained indicate a limited pathogenic potential of the species.

1. Introduction

Corynebacteria belong to the phylum Actinobacteriota, Gram-positive bacteria with a high G+C DNA content [1,2]. Within this phylum, they form the CMNR group together with the genera Mycobacterium, Nocardia and Rhodococcus based on a complex mycolic acid-containing cell wall structure, which is shared by these bacteria [3]. To date, 168 taxonomically valid names of Corynebacterium species are known [4,5]. The genus iincludes biotechnologically important members such as Corynebacterium glutamicum [6,7] and toxin-producing pathogens such as Corynebacterium diphtheriae [8,9,10].

Several members of the genus are able to infect animals and cause zoonoses in humans. A prominent example of these zoonotic pathogens is Corynebacterium ulcerans, which was first isolated from a case of diphtheria-like illness [11]. The range of mammal hosts observed for C. ulcerans is extremely broad and includes cattle, goats, pigs, wild boars, dogs, cats, ground squirrels, otters, camels, monkeys, whales, water rats and others (reviewed in [12]). Human infections associated with C. ulcerans appear to be increasing in various countries and can most often be ascribed to zoonotic transmission (for recent examples, see [13,14,15,16]).

Corynebacterium pseudotuberculosis is another species known for zoonotic transmission. The pathogen infects small ruminants like sheep and goats, besides buffalos, camelids and others [17,18,19] and can be transmitted to humans, causing mainly purulent inflammation and chronic granulomas [20,21].

While the above-mentioned members of the Corynebacterium genus are long-known and rather well investigated, other more recently isolated species have not been studied in detail. One of these new species is Corynebacterium silvaticum, originally isolated from roe deer and wild boar and taxonomically described in 2000 [22]. C. silvaticum strain W25 was characterized with respect to its interaction with human cell lines, indicating a certain pathogenic potential [23], and in fact, two human infections with C. silvaticum were reported recently [24].

An even more neglected species is Corynebacterium lactis, mentioned in only three publications in the NCBI PubMed database when the species name is used as a query [25]. In the literature, the isolation from the milk of a cow without pathological findings, a cow with unspecific mastitis [26] and from an abscess of a companion dog [27] was reported. Characterization was mainly focusing on taxonomical investigations and markers such as biochemical tests, the presence of menaquinones and fatty acid composition and others [26,27]. Since further information about this species with respect to pathogenic and zoonotic potential was scarce, we started an approach to characterize C. lactis strain RW3-42 with respect to growth, antimicrobial susceptibility and behavior in invertebrate infection model systems, which is presented here.

2. Materials and Methods

2.1. Bacteria and Growth Conditions

Bacteria used in this study (

Table 1. Bacteria used in this study.

Strain	Reference
Corynebacterium glutamicum ATCC13032	[28]
Corynebacterium lactis RW3-42	[26]
Corynebacterium pseudotuberculosis 12CS0282	[29]
Corynebacterium silvaticum W25	[30]
Corynebacterium ulcerans 809	[31]
Escherichia coli OP50	[32]

) were streaked out from permanent cultures of our laboratory collection, with the exception of RW3-42, provided by C. Riedel (University of Ulm). Strains were routinely grown in Brain Heart Infusion (BHI; Oxoid, Wesel, Germany) at 37 °C under shaking at 125 rpm in baffled flasks or on BHI plates containing 1.5% agar-agar (Oxoid, Wesel, Germany). Medium for C. silvaticum and C. pseudotuberculosis was supplemented with 0.5% Tween 80 to avoid aggregation of bacteria d 10% Fetal Bovine Serum (FBS) to support bacterial growth.

2.2. Transformation of C. lactis

For fluorescence microscopy, electrocompetent C. lactis, prepared according to the protocol established for C. diphtheriae [33], were transformed with gfp-carrying plasmid pEPR1-p45gfp, which leads to constitutive expression of GFP controlled by the p45 corynebacteriophage promoter [34].

2.3. Antimicrobial Resistance Testing

For simple manual antimicrobial susceptibility testing of C. lactis RW3-42, Thermo Scientific™ Oxoid™ Antimicrobial Susceptibility Test Discs were used in a semi-quantitative agar diffusion test following the instructions of the supplier (Oxoid, Wesel, Germany). In short, agar plates were inoculated with C. lactis RW3-42, and Oxoid™ Antimicrobial Susceptibility Test Discs were immediately placed on the surface of the plates. After overnight incubation at 37 °C, zones of growth inhibition appearing around the discs were measured.

For a more detailed analysis, MIC Test Strips (Liofilchem Diagnostici, Roseto degli Abruzzi, Italy) were used. According to the manufacturer’s recommendation, 3 × 10⁷ bacteria were plated out on agar plates before placing the MIC Test Strip on the surface. After overnight incubation at 37 °C, a symmetrical inhibition ellipse centered along the strip was formed. The respective MIC was read directly from the scale in terms of µg mL⁻¹, at the point where the edge of the inhibition ellipse intersects with the MIC Test Strip.

As in all other experiments in this study, BHI was used, instead of Mueller–Hinton agar supplemented with 5% defibrinated horse blood and 20 mg/L β-NAD (MH-F) as recommended for corynebacteria by European Committee on Antimicrobial Susceptibility Testing (EUCAST [35]) and cells were grown under ambient atmosphere instead of 5% CO₂, to establish identical growth conditions throughout this study, e.g., for transformation of bacteria and infection experiments.

2.4. Infection of Caenorhabditis Elegans

2.4.1. Fluorescence Microscopy

C. elegans N2 were maintained on agar plates inoculated with E. coli OP50 for three to seven days until the worms became starved, indicated by clumping behavior [36]. Subsequently, the nematodes were infected with C. lactis strain RW3-42 pEPR1-p45gfp, grown in BHI medium supplemented with 25 µg mL⁻¹ kanamycin, as well as E. coli OP50 grown in LB medium. Infection of 20 L4 stage larval worms was carried out with 20 µL of each bacterial strain (from an overnight culture) on NGM plates at 21 °C for 24 h. Worms were then transferred to plates with 20 µL of unlabeled E. coli OP50 for a further 24 h, to allow the gut to clear of fluorescent cell debris and to expel non-adhering bacteria. Nematodes were paralyzed with 5 to 10 µL of 0.6% 2-phenoxy-2-propanol (Sigma, St. Louis, MI, USA) or 20 mM sodium azide, mounted onto agar pads and photographed using a Leica DMR fluorescence microscope.

2.4.2. Survival Assay

For the survival assay, adult C. elegans were transferred onto NGM plates with E. coli lawns and incubated overnight at 21 °C. Subsequently, adult worms were removed, while eggs remained on the plates. These were incubated for four to six days until the worms hatched from the eggs and became starved. Bacterial strains from overnight cultures grown as described above in Section 2.1 were inoculated in the fresh medium BHI medium and grown to an OD₆₀₀ of 0.6. Cells were harvested by centrifugation (10 min, 4500× g) and resuspended in 10 mM MgSO₄ to an OD₆₀₀ of 1.0. 50 µL aliquots of the bacterial suspension were spotted in the center of NGM plates and allowed to dry for approx. 1 h at room temperature. Subsequently, plates were incubated for 18 h at 37 °C. Ten synchronized L4 stage C. elegans (per plate) were placed around the bacterial lawn and incubated at 21 °C for two days. Dead worms were scored and removed from the plates. Worms that left the plates were counted separately and defined as censored. After 20 h, surviving worms were transferred to fresh NGM plates seeded with the respective bacterial strain to avoid incorrect scoring due to the presence of offspring of similar size.

2.5. Chemotactic Behavior

Chemotaxis of C. elegans was analyzed as described previously [37,38]. The chemotactic behavior of the nematodes towards different bacteria was evaluated based on a choice index with −1.0 representing a complete preference for C. lactis RW3-42, an index of 1.0 representing complete preference for the compared bacterial strain and an index of 0 representing an equal distribution.

$$Choice index={\displaystyle \frac{\mathrm{T}\mathrm{e}\mathrm{s}\mathrm{t} \mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}-\mathit{C.\; lactis} \mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r}}}$$

For the pairwise comparison of C. lactis with the tested bacterial strains, bacterial suspensions (30 µL, OD600 = 1.0) were spotted onto NGM plates and incubated for 24 h at 37 °C. C. elegans were grown on NGM plates with E. coli as a food source until the adult stage was reached and starvation commenced. Worms were collected in M9 buffer (3 g KH₂PO₄, 6 g Na₂HPO₄, 5 g NaCl, 1 mL 1 M MgSO₄ per l deionized H₂O), synchronized by bleaching (0.8% NaOCl, 1% NaOH, dissolved in water) for 7 min and washed three times in M9 buffer. Eggs were hatched overnight in M9 buffer at room temperature. Synchronized worms (5 µL aliquots) were transferred onto the NGM plates containing the bacterial suspension. Approximately 20–40 nematodes were used per plate. Plates were incubated at 21 °C, and assays were performed in triplicate.

2.6. Infection of Greater Wax Moth (Galleria mellonella) Larvae

For infections of wax moth larvae, bacteria from an overnight culture were inoculated in fresh medium and grown until an OD₆₀₀ of 0.6 was reached. Bacteria were harvested by centrifugation (10 min, 4500× g) and resuspended in 10 mM MgSO₄ to an OD₆₀₀ of 1 (approximately 3 × 10⁸ CFU mL⁻¹). For infection, a 50 µL Hamilton syringe was used to inject 5 µL aliquots into the hindmost left proleg of the larvae. To avoid light-induced melanization, G. mellonella larvae were incubated in the dark and monitored daily for activity, melanization, cocoon formation and survival. For each bacterial strain, five wax moth larvae were infected and incubated at room temperature. Each infected individual larva was monitored on the day of infection and daily for seven days using the Health Index Scoring System [39]. The scores of the four categories were summed to obtain a total score, which was divided by the number of larvae to receive the average health score. A score between 9 and 10 suggests healthy, uninfected larvae, whereas a lower score indicates impaired health.

2.7. Genome Screening

In order to evaluate a potential pathogenic potential of C. lactis, the only available complete, ungapped 2.8 Mb genome sequence of type strain RW2-5 (BioSample ID SAMN04012704, GenBank CP006841, genome coverage 128.56x) was screened for genes encoding virulence factors using known corynebacterial proteins involved in pathogenicity as query in blastp searches in addition to use of the VFAnalyzer and the information collected at the Virulence Factor DataBase (VFDB) [40].

3. Results

3.1. Growth of C. lactis

C. lactis RW3-42 is a fast-growing Corynebacterium species (Figure 1) reaching a doubling time of 62 ± 5 min and a final OD₆₀₀ of 12.8 ± 1.8 after 24 h in BHI medium. Growth in other complex media, e. g. Luria broth, is only slightly slower and also growth in minimal medium was established [41].

3.2. Antibiotic Susceptibility and Resistance of C. lactis

To generate a first set of data as an overview, inhibition zone diameters were determined for various antibiotics using Oxoid (Wesel, Germany) antimicrobial susceptibility filter discs (

Table 2. Antimicrobial sensitivity and resistance of C. lactis RW3-42. Two independent biological replicates were carried out for each antibiotic using Oxoid (Wesel, Germany) antimicrobial susceptibility filter discs. The amount of antibiotics on the discs is indicated.

Antibiotics	Amount	Diameter of Inhibition Zone
Amoxicillin	10 µg	18 mm/20 mm
Ampicillin	25 µg	27 mm/28 mm
Azithromycin	15 µg	-/-
Aztreonam	30 µg	-/-
Cefalexin	30 µg	24 mm/26 mm
Cefazolin	30 µg	25 mm/25 mm
Cefepime	30 µg	-/-
Cefoperazone	30 µg	7 mm/8 mm
Cefotaxime	30 µg	-/-
Cefotetan	30 µg	-/-
Cefoxitin	30 µg	19 mm/20 mm
Ceftriaxone	30 µg	17 mm/20 mm
Cefuroxime	30 µg	22 mm/25 mm
Chloramphenicol	30 µg	26 mm/27 mm
Ciprofloxacin	5 µg	25 mm/26 mm
Clindamycin	2 µg	-/-
Cloxacillin	5 µg	9 mm/11 mm
Colistin	25 µg	-/-
Doxycycline	30 µg	9 mm/10 mm
Erythromycin	15 µg	-/-
Gentamicin	10 µg	22 mm/22 mm
Imipenem	10 µg	32 mm/34 mm
Kanamycin	30 µg	19 mm/21 mm
Lincomycin	15 µg	-/-
Meropenem	10 µg	8 mm/9 mm
Metronidazole	5 µg	-/-
Mupirocin	5 µg	-/-
Nalidixic acid	30 µg	-/-
Nitrofurantoin	200 µg	15 mm/20 mm
Oxacillin	5 µg	10 mm/11 mm
Penicillin G	10 IU	29 mm/34 mm
Streptomycin	25 µg	-/-
Sulfamethoxazole	100 µg	-/-
Tetracycline	30 µg	-/-
Ticarcillin	75 µg	30 mm/32 mm
Trimethoprim	2.5 µg	-/-
Vancomycin	30 µg	13 mm/15 mm

). C. lactis was found to be resistant to the aminoglycoside streptomycin, the cephalosporins cefepime, cefotaxime and cefotetan, the diazanaphthaline nalidixic acid, the lincosamides clindamycin and lincomycin, the macrolides azithromycin and erythromycin, the monobactam aztreonam, the nitroimidazole metronidazole, the polymyxine colistin, the sulfonamide sulfamethoxazole, tetracycline, trimethoprim and the Pseudomonas aeruginosa-derived antimicrobial metabolite mupirocin.

In contrast, C. lactis was susceptible to the aminoglycosides gentamicin and kanamycin, the β-lactams amoxicillin, ampicillin, cefazolin, cefoxitin, cefuroxime, cloxacillin, oxacillin, penicillin G and ticarcillin, the carbapenemes imipenem and meropenem, the cephalosporines cefalexin, cefoperazone and ceftriaxone, the fluoroquinolone ciprofloxacin, the glycopeptide vancomycin, the nitrofurane nitrofurantoin, the tetracycline doxycycline and the aromatic antibiotic chloramphenicol.

For antibiotics relevant for molecular biology purposes (e.g., cloning, induction of expression) and infection assays, minimal inhibitory concentrations were determined for cells grown on BHI medium using MIC Test Strips (

Table 3. Determination of minimum inhibitory concentrations (MIC) of antimicrobial agents against C. lactis RW3-42. For selected antibiotics, minimal inhibitory concentrations were determined using MIC Test Strips. After overnight incubation at 37 °C, a symmetrical inhibition ellipse centered along the strip was formed in case of antimicrobial susceptibility. The respective MIC was read directly from the scale at the point where the edge of the inhibition ellipse intersects with the MIC Test Strip. Experiments were carried out in triplicate (three independent biological replicates).

Antibiotics	MIC [µg mL−1]
Ampicillin	0.125–0.19/0.125–0.19/0.094–0.125
Chloramphenicol	2/1/0.75
Doxycycline	3–4/2–3/2–3
Gentamicin	0.19–0.25/0.064–0.094/0.125
Kanamycin	0.75–1/0.75/0.75
Tetracycline	24–32/32–48/32

). In general, β-lactam antibiotics effectively inhibited C. lactis strain RW3-42, followed by aminoglycosides and chloramphenicol. The minimal inhibitory concentration of doxycycline was rather high, while bacteria were almost insensitive to tetracycline, as indicated previously by the negative disc diffusion test.

3.3. Interaction with Invertebrate Infection Model Systems

The three reported strains of C. lactis were isolated from animal-related material, i.e., from raw cow milk and from a canine skin ulcer [26,27]. To add more information to this limited data set with respect to the interaction of C. lactis with animals, we tested the response of invertebrate model systems established for corynebacteria before [42,43,44,45,46,47], the nematode C. elegans and larvae of the greater wax moth G. mellonella, to contact with the bacteria.

3.3.1. Colonization of C. elegans

When C. elegans N2 was incubated in liquid medium in the presence of C. lactis, anatomical deformations as a result of contact were observed. Worms with deformed anal region (Dar) due to bacterial colonization and star formation, an aggregation of worms sticking together with their tails were found (Figure 2), indicating a pathogenic effect of C. lactis RW3-42 on C. elegans [38].

When GFP-labelled bacteria were fed to C. elegans N2 for 24 h, microscopy revealed a fluorescence signal distributed throughout the nematode with a strong signal in the gut (Figure 3a–c). When the worms were transferred to plates with E. coli OP50, to expel non-adhering C. lactis and undigested GFP protein from the gut, after 24 h the gut was cleared from the bacteria, while the body of the worms showed a clear fluorescence signal (Figure 3d–f), indicating that either the tissue or the surface of the nematodes was permanently colonized by C. lactis strain RW3-42 pEPR1-p45gfp.

3.3.2. C. elegans Survival Assay

To further study a putative pathogenic effect of C. lactis on C. elegans, a nematode survival assay was carried out. E. coli OP50 and C. ulcerans 809 were used as negative and positive controls, respectively. After 24 h of infection with C. lactis RW3-42, a 50% probability of death was reached. Within 54 h, this value increased to 82%, indicating a pathogenic effect of C. lactis on C. elegans (Figure 4a). Furthermore, avoidance behavior of the nematodes towards the bacteria was observed, as revealed by crawling off the plates as described before [48]. Within 24 h post-infection, 22.5% of the worms had left the plates, increasing to 32.5% after 54 h (Figure 4b).

3.3.3. Chemotactic Behavior of C. elegans

Taken together, the nematode colonization and survival assays indicated at least a small, minor detrimental effect of C. lactis. This was supported by an analysis of the chemotactic behavior of the worms. When C. elegans can choose between different feed sources, the worms are able to evaluate these and migrate to the preferred one. From the number of worms present at the different bacterial cultures, a choice index can be calculated [36]. We used this approach to compare C. lactis with different other Corynebacterium species; E. coli was tested as the control (Figure 5 and Supplementary Material Figure S1). As shown before [38], E. coli OP50, the standard feed for C. elegans, is preferred when C. glutamicum ATCC13032 is given as the choice. In contrast, C. glutamicum ATCC13032 is preferred in comparison to C. lactis RW3-42, similar to C. pseudotuberculosis strain 12CS0282, C. silvaticum strain W25 and C. ulcerans 809.

3.3.4. Infection of G. mellonella Larvae

As a second invertebrate animal model system, larvae of the greater wax moth were infected with C. lactis RW3-42 and C. ulcerans 809. Injection of MgSO₄ and untreated larvae were used as controls. As expected, untreated larvae and those injected with buffer showed no signs of any immunological reaction. Also, infection with C. lactis strains showed no detrimental effects. In contrast, injection of C. ulcerans 809 led to strong melanization of larvae (Figure 6). For a more quantitative analysis, larvae were monitored for seven days with respect to motility, melanization and cell death, and a health score index was determined. In contrast to the C. elegans model, wax moth larvae showed no negative reaction to C. lactis contact, i.e., melanization in response to injection, reduced motility or death, typical effects observed after C. ulcerans injection (Figure 7).

3.4. Putative Virulence Factors

A screening of the genome sequence of C. lactis type strain RW2-5 revealed a reduced number of virulence factors compared with important corynebacterial pathogens (

Table 4. Corynebacterial virulence factors predicted in C. lactis RW3-42. Presence or absence is indicated by + and −. Table adapted from [51,52].

Protein	Occurrence	Reference
ABC heme uptake system	+	[53]
ABC iron chelate uptake system	−	[53]
ABC iron siderophore uptake system	+	[53]
Acid phosphatase	+	[49]
Ceramidase	+	[50]
Cholesterol esterase	−	[49,50]
Cholesterol oxidase	+	[49]
CPP/CP40	−	[54]
DIP0733	+	[55]
DIP1281	+	[56]
DIP1621	+	[57]
DIP2093	−	[58]
Diphtheria toxin	−	[59]
EmbC/MptC	+	[60]
MdbA	+	[61]
Mycolic acids	−	[26]
Neuraminidase H	−	[54]
Phospholipase D/ovis toxin	−	[54]
Ribosome-binding protein/Shiga-like protein	−	[31,44]
RhuM	+	[50]
SpaA-type adhesive pili	+	[62]
SpaH-type adhesive pili	+	[62]
Superoxide dismutase C	+	[54]

, Supplementary Material Table S1). C. lactis is a non-toxigenic Corynebacterium species, lacking the diphtheria toxin-encoding tox gene. Compared with C. pseudotuberculosis, the gene for the ovis toxin, i.e., phospholipase D, is missing and compared with C. ulcerans, no Shiga-like toxin-encoding rbp gene was observed. As in C. jeikeium [49,50], enzymes like acid phosphatase, ceramidase and two cholesterol oxidases were found, indicating the putative utilization of host membrane compounds and fat as a nutrient source. Furthermore, genes encoding proteins involved in adhesion, i.e., DIP0733, DIP1281, DIP1621, EmbC and SpaA-type pili, are encoded by the C. lactis genome.

4. Discussion

C. lactis is a rarely isolated Corynebacterium species, connected to infected animals, i.e., a dog and a cow, which implies a certain danger of zoonotic transmission [26,27]. The observed antimicrobial susceptibility of C. lactis RW3-42 investigated in this study was similar to the profile reported for C. lactis strain 2447 isolated from a cutaneous abscess in a companion dog in Brazil previously [27]. However, in contrast to isolate 2447, strain RW3-42 was resistant to cefotaxime, clindamycin and erythromycin and exhibited limited susceptibility to tetracycline. This observation indicates that C. lactis can acquire different resistant determinants and may have a potential role as a putative source of shared resistance genes in microbial populations, especially when keeping in mind that the known strains were isolated from mixed infections. It has to be mentioned that, in contrast to the EUCAST standard, BHI medium was used in this work, which impairs a direct comparison of the data presented here with former studies [27]. Although beyond the aim of this study, sequencing and annotation of the RW3-42 genome, in addition to data from further isolates, would be helpful for a deeper understanding of antimicrobial resistance and the role of horizontal gene transfer in this species.

To evaluate the pathogenic potential of C. lactis without harming mammals, the invertebrate animal model systems C. elegans and G. mellonella established for corynebacteria before [42,43,44,45,46,47] were applied in this study. C. lactis strain RW3-42 was able to colonize the surface of the nematodes, most likely based on its multiple adherence proteins, the SpaA- and SpaH-type pili [62] and the non-fimbrial adhesins DIP0733 [55], DIP1281 [56], DIP1621 [57] and EmbC [58]. As a consequence, the nematodes showed morphological changes, i.e., Dar and star formation. The observed negative effect of the bacteria on the survival of C. elegans might be explained by C. lactis RhuM, as the RhuM protein of Salmonella was found to be involved in the killing of C. elegans by an unknown mechanism [63] and by the degrading enzymes cholesterol oxidase, acid phosphatase and ceramidase [49]. Interestingly, the nematodes strictly avoided C. lactis RW3-42 as a feed source when other bacteria were available. This negative chemotaxis may be explained not only by the decreased survival rate but also by the presence of two cholesterol oxidases in C. lactis. It may be hypothesized that these may allow C. lactis to degrade and metabolize the cholesterol added to the medium to support C. elegans growth, and consequently, the worms may avoid the area colonized by the bacteria due to nutrient depletion.

A negative effect on wax moth larvae was not observed when this model system was tested, maybe as a result of a limited number of virulence factors found.

5. Conclusions

The exclusive isolation of C. lactis strains from animal sources, i.e., raw cow milk and a companion dog, may have implications for public and veterinary health within the concept of global health. In this study, several antimicrobial resistances and a certain pathogenic potential towards C. elegans were observed for C. lactis RW3-42. Further investigation, especially on other C. lactis strains, are needed to evaluate and fully understand the pathogenic potential of the species.