Screening of Functional Properties of Lactic Acid Bacteria Isolated from Animal Rennets and Their Associated Cheeses and Whey

Abstract

This study investigated the diversity and functional potential of lactic acid bacteria isolated from traditional lamb rennet, cheese, and whey collected from seven artisanal sheep farms in southern Romania. A total of 31 samples were analyzed, yielding 118 Gram-positive, catalase-negative isolates. Following dereplication by rep-PCR and 16S rRNA gene sequencing, 63 unique strains were identified across nine genera, with Lactiplantibacillus, Lactococcus, and Leuconostoc being the most prevalent. Strain distribution varied by sample type and manufacturer, with rennet and whey showing greater species diversity than cheese. Technological characterization showed strain-dependent differences in acidification and growth in cow’s and goat’s milk. Genetic screening revealed a high prevalence of functional genes such as ribA, gad, and α-amy, while genes associated with bacteriocin (nisA, pln) and folate (folK) production were less common. Most strains carried multiple functional genes, indicating a genetic potential for diverse functional traits. Antibacterial activity assays demonstrated that nearly all strains inhibited at least three indicator pathogens, with ten strains, particularly Lactiplantibacillus plantarum and Lactococcus lactis strains, exhibiting strong inhibitory effects. Bacteriocin production was confirmed for three Lact. lactis strains. Exopolysaccharide (EPS) production was confirmed in two strains, with yields varying by growth medium and sucrose supplementation. Overall, the results underscore the rich microbial diversity and promising biotechnological potential of LAB from traditional Romanian dairy ecosystems, supporting their use in food fermentation and functional product development.

1. Introduction

Lactic acid bacteria (LAB) have been utilized for centuries in the production of a wide range of fermented foods, such as dairy products, fermented vegetables, sourdough bread, certain types of sausages, as well as numerous traditional foods specific to various countries. The utmost importance of these bacteria in food fermentation is linked to their substantial technological and functional potential [1]. In particular, LAB are used due to their fast acidification capacity, as a result of organic acid production, which prevents the proliferation of harmful microorganisms during the fermentation process and extends the foods’ shelf life [2]. The antimicrobial activity of LAB may also result from the production of various metabolites, such as bacteriocins, ethanol, hydrogen peroxide, and antibiotic-like peptides. In addition to its role in the biological control of foodborne pathogens, this antimicrobial activity imparts probiotic properties to the LAB that produce them [3].

Moreover, LAB produce various beneficial metabolites during fermentation, contributing to the health-promoting properties of fermented foods. These metabolites, including organic acids, bacteriocins, exopolysaccharides, and vitamins, can improve gut health, enhance nutrient bioavailability, and exhibit anti-inflammatory and immunomodulatory effects [4,5,6]. In particular, the exopolysaccharides (EPSs) produced by LAB increase the viscosity and improve the rheological properties of food products [7], while also offering several health benefits to consumers, such as glycemic control, antitumor effects, immunomodulatory properties, and antioxidant activities [8]. The production of vitamins, short-chain fatty acids, γ-ammino-butyric acid (GABA), and enzymes, enhances the nutritional value of foods [9,10]. Additionally, GABA plays a key role as an inhibitory neurotransmitter in the central nervous system and has been shown to have antihypertensive, diuretic, and antidepressant effects [11].

Considering their technological and functional potential that is usually genetically determined by the plasmids in lactococci [12], it is clear why there is a continuous demand for new LAB strains for use in food fermentation. As a result, novel sources are being explored to isolate such functional bacterial strains. In the present study, artisan animal rennet was examined to explore its bacterial diversity and isolate LAB strains with functional properties. Rennet is an enzymatic preparation extracted from the fourth stomach (abomasum) of ruminant animals, principally calves, lambs or kids [13], traditionally used as coagulant agent in the production of many cheese types [14,15]. Studies on the microbial composition of animal rennets have revealed the presence of LAB, which are believed to be the source of non-starter LAB and are associated with the distinctive flavors of the cheeses produced [14,16].

In this study, artisan animal rennet was examined to explore its bacterial diversity and to isolate LAB strains with functional properties. In this study, LAB diversity was analyzed in thirty-one samples of lamb rennet, along with the associated cheese and whey, using culture-dependent methods. The newly isolated strains were identified on a species level and assessed for their technological and functional properties, including antimicrobial activity, EPS production, and the presence of functional genes responsible for the biosynthesis of vitamins, GABA, and certain enzymes.

2. Materials and Methods

2.1. Sample Characteristics

Thirty-one samples collected from seven anonymous sheep farms (artisanal cheese manufacturers), located in the south part of Romania (Buzău county) were analyzed in this study. Among these samples, 17 were liquid lamb rennets (artisanal products), coming from different production batches, while the others were cheese (8 samples) and whey (6 samples) associated with the collected rennets (

Table 1. Sample codification (where Pn encodes samples collected from manufacturer n) and pH measurement results.

Sample ID	Sample Type	pH
P1R1	Rennet	4.5
P1R2	Rennet	4.0
P1R3	Rennet	3.9
P1R4	Rennet	3.9
P1R5	Rennet	4.6
P1C1	Cheese	4.3
P1C2	Cheese	4.5
P2R1	Rennet	4.8
P2R2	Rennet	4.7
P2R3	Rennet	5.2
P2R4	Rennet	5.0
P2C1	Cheese	4.5
P2C2	Cheese	5.5
P2W1	Whey	4.3
P3R1	Rennet	4.5
P3R2	Rennet	4.0
P3R3	Rennet	4.5
P3C1	Cheese	6.5
P3W1	Whey	4.0
P4R1	Rennet	3.8
P4C1	Cheese	6.5
P4W1	Whey	4.3
P5R1	Rennet	4.3
P5C1	Cheese	6.0
P5W1	Whey	4.2
P6R1	Rennet	4.3
P6R2	Rennet	3.5
P6C1	Cheese	6.0
P6W1	Whey	4.3
P7R1	Rennet	4.3
P7W1	Whey	4.5

). The pH of the samples was measured with a pH meter (inoLAB, WTW, Weilheim, Germany). In the case of cheese, 1 g of each sample was diluted with 9 mL of physiological saline and homogenized in a MiniMix stomacher (Interscience, Paris, France) for 5 min at the highest speed; the pH was measured immediately after homogenization.

2.2. Isolation and Purification of LAB

Serial decimal dilutions of all samples were prepared in physiological saline and plated on de Man–Rogosa–Sharpe (MRS) agar media [17]. After 48–72 h of incubation at 30 °C, colonies were randomly picked up (about 10% of the colonies with the same morphology) and purified by successive cultivations on liquid and solid medium. They were further tested by Gram staining and for catalase activity. The presumptive LAB (Gram-positive and catalase-negative isolates) were deposited at −80 °C in an MRS medium supplemented with 25% (v/v) glycerol as a cryoprotectant.

2.3. Dereplication and Identification of Unique Isolates

Genomic DNA was extracted from overnight cultures of 118 isolates using a Pure Link Genomic DNA kit (Invitrogen, Carlsbad, CA, USA) as described in the manufacturer’s instructions. In order to remove duplicate or redundant bacterial isolates and to avoid further repetitive analysis, dereplication of the isolates [18] was performed by rep-PCR fingerprinting [19], in a DLAB TC1000-G PCR thermocycler (SciQuip, Rotherham, UK), using (GTG)₅ primers, as described previously [20]. The rep-PCR profiles were compared among isolates from each sample and clustered based on their similarities. One representative strain of each cluster was further submitted for 16S rRNA gene sequencing. For that, PCR amplifications of template DNA were performed using the universal primers 27f and 1492r [21,22]. The amplified 16S rRNA gene products were sequenced by the Sanger method (Macrogen Europe, Amsterdam, The Netherlands) using both forward and reverse primers. The resulting chromatograms were manually inspected and assembled into consensus sequences using MEGA version 11 to ensure sequence accuracy.

The sequences were compared to those available in the databases of the National Center for Biological Information using the on line BLAST search program (www.ncbi.nlm.nih.gov/BLAST/) (accessed on 21 October 2024).

2.4. Technological Screening

The 63 different LAB strains isolated from the rennet and associated cheese and whey samples were characterized for their growth and acidifying abilities in an MRS medium and in two types of commercial, pasteurized milk, namely: cow’s milk and goat’s milk. To analyze the bacterial growth in an MRS medium, strains were inoculated in triplicate (2% inoculum) in a 96-well plate filled with MRS (200 µL in each well). The plate was maintained at 30 °C and OD_600nm was measured during the following 24 h, at 1 h intervals. In parallel, strains were inoculated (2% inoculum) in 20 mL of MRS and milk, respectively, and the pH values were measured at 3, 6, 24, and 48 h after inoculation.

2.5. Screening for Functional Properties

2.5.1. Genetic Screening for Functional Genes

Genetic screening was performed by PCR amplification of specific functional genes using the extracted genomic DNA from the bacterial strains as a template. Specific primers were employed to target genes involved in the synthesis of various compounds, such as: bacteriocins, exopolysaccharides, S-layer proteins, vitamins, or gamma-aminobutyric acid. Additionally, genes encoding for enzymes, such as α-amylase and α-glucosidase were also selected for the screening, as summarized in

Table 2. Description of primers, targeted genes and encoded products.

Targeted Gene	Encoded Product	Primer Sequence	Annealing Temperature	Reference
ftf	Frunctansucrase	F: 5′-GAYRTYTGGGAYWSNTGGC-3′ R: 5′-GCWGANCCNGACCATTSTTG-3′	55 °C	[23]
dexreu	Dextransucrase	F: 5′-GTGAAGGTAACTATGTTG-3′ R: 5′-ATCCGCATTAAAGAATGG-3′	55 °C	[24]
gtf	Glucansucrase/Glycosyltransferase	F: 5′-GAYAAYWSIAAYCCIRYIGTIC-3′ R: 5′-ADRTCICCRTARTAIAVIYKIG-3′	62 °C	[25]
agl	α-Glucosidase	F: 5′-GCSAAAATGCTAGCGACYMT-3′ R: 5′-CCACTGCATYGGYGTACGY-3′	61 °C	[26]
epsD/E	Priming Glycosyltransferase	F: 5′-TCATTTTATTCGTAAAACCTCAA TTGAYGARYTNCC-3′ R: 5′-AATATTATTACGACCTSWNAYYTG CCA-3′	58 °C	[27]
malP	Maltose-phosphorylase	F: 5′-TGCCAYAAYGARTGGGARAT-3′ R: 5′-ACSCKATCWGCCCARAAAC-3′	60 °C	[26]
α-amy	Amylase	F: 5′-AGATCAGGCGCAAGTTCAGT-3′ R: 5′-TTTTATGGGCACACCACTCA-3′	62 °C	[26]
slpA	S-layer proteins	F: 5′-ATGAAGAAAAATTTAAGAATTG-3′ R: 5′-AAAGTTTGCAACCTTTACGTAAG-3′	42 °C	[28]
ribA	Riboflavin	F: 5′-TTTACGGGCGATGTTTTAGG-3′ R: 5′-CGACCCTCTTGCCGTAAATA-3′	62 °C	[26]
folK	Folate	F: 5′-CCATTTCCAGGTGGGGAATC-3′ R: 5′-GGGGTGGTCCAAGCAAACTT-3′	61 °C	[26]
gad	Glutamate decarboxylase	F: 5′-CCTCGAGAAGCCGATCGCTTAGTT CG-3′ R: 5′-TCATATTGACCGGTATAAGTGATGC CC-3′	58 °C	[29]
nisA	Nisin A	F: 5′-GGATAGTATCCATGTCTG-3′ R: 5′-CAATGATTTCGTTCGAAG-3′	55 °C	[30]
lcn972	Lactococin A	F: 5′-TTGTAGCTCCTCAGAAGGAACATGG-3′ R: 5′-GCCTTAGCTTTGAATTCTTACCAAAAG-3′	58 °C	[31]
pln	Plantaricin	F: 5′-GTACAGTACTAATGGGAG-3′ R: 5′-CTTACGCCAATCTATACG-3′	53 °C	[32]

. The amplifications were performed using the following steps: initial denaturation at 94 °C for 5 min, 35 cycles of amplification (94 °C for 30 s, t °C for 30 s, 72 °C for 1 min) followed by a final elongation at 72 °C for 10 min, where t °C represents the specific annealing temperature of each set of primers according to Table 2.

The PCR products were separated by electrophoresis on 1.5% or 2% (wt/vol) agarose gel, depending on the expected size of the amplicon, and visualized by UV transillumination after staining with the SYBR safe DNA gel stain (Invitrogen, Carlsbad, CA, USA). As molecular weight markers, a 1 kb DNA Ladder (ThermoFisher Scientific, Waltham, MA, USA), and a 50 bp ladder (Promega, Madison, WI, USA) were used.

2.5.2. Antibacterial Activity

The antibacterial activity of the 63 unique bacterial strains was tested against the following indicators: Staphylococcus aureus ATCC 25923, Escherichia coli ATCC 25922, Listeria monocytogenes ATCC 1911-1, Salmonella enterica ATCC 14028, Bacillus cereus CBAB, Bacillus subtilis ATCC 6633, and Lactobacillus delbrueckii subsp. bulgaricus LMG 6901^T. The indicator strains were grown in a BHI medium (Brain Heart Infusion, Merck, Darmstadt, Germany), except for L. delbrueckii subsp. bulgaricus, which was grown in an MRS medium. The two Bacillus strains were incubated at 28 °C with shaking at 150–200 rpm, while all the other strains were incubated at 37 °C under static conditions.

To evaluate the antibacterial activity of the isolated strains, an agar overlay assay was used [33]. Briefly, 10 µL of an overnight culture were spotted on MRS agar medium and incubated at 30 °C to allow the bacterial growth in the spots. All tested cultures were grown under identical conditions; however, the inoculum was not standardized, which could influence the size of the inhibition zones. A top-layer medium containing 100 µL of the indicator strain was then poured on the top and further incubated at the optimum growth temperature of the indicator. This top-layer represented a soft agar (0.7%) medium, either MRS or BHI, depending on the indicator strain. The inhibitory effect was evaluated by measuring the clear zone surrounding the spots.

2.5.3. Bacteriocin Production

For selected strains, further tests were done to elucidate the nature of inhibition. Supernatants of overnight cultures were neutralized to about 6.5–7.0, using 1 M NaOH. The inhibitory activity was then tested against L. delbrueckii subsp. bulgaricus, using the agar spot method [34].

2.5.4. Exopolysaccharide Production

To evaluate the capacity of the isolated strains to produce EPSs, they were grown for 24–48 h on a modified MRS agar medium, containing 50 g/L sucrose instead of glucose [35]. The mucoid or ropy phenotype of the developed colonies was recorded. The ropy phenotype was detected by gently touching the colony surface with a sterile toothpick and observing the formation of a viscous, thread-like appearance when the colony is lifted [36]. Selected strains were further inoculated in 50 mL of modified MRS (50 g/L sucrose) liquid medium, but also in cow’s and goat’s milk, both supplemented with 50 g/L sucrose. Cultures were incubated overnight at 30 °C, and EPSs were extracted by a two-step precipitation method described by De Vuyst et al. [37]. The acetone-precipitated EPSs were dried and weighed, for a quantitative determination.

3. Results

3.1. Sample Characteristics and Microbiological Analysis

The pH of the rennet samples varied between 3.8 and 5.2, with most of the values in the range of 4.0–4.9 (Table 1). The highest pH values, of 5.0 and 5.2, were measured for two rennet samples collected from producer 2 (P2R3 and P2R4). The cheese samples were less acidic, most of them having a pH between 5.5 and 6.5. Lower values were measured for the samples collected from producer 1 and one sample from producer 2 (P2C1). All whey samples had a pH between 4.0 and 4.5.

From 31 samples of rennet, cheese and whey, a total of 145 colonies were obtained on MRS agar plates. Among these, 118 were Gram-positive and catalase-negative, and were deposited as presumptive LAB. After dereplication and clustering of the isolates from each sample, 82 representative strains were submitted to 16S rRNA gene sequencing.

3.2. LAB Diversity

The identification results are summarized in Figure 1. As observed, the predominant genus was Lactiplantibacillus (43 strains, representing about 36% of the total isolated strains), closely followed by Lactococcus (35 strains, corresponding to about 30% of the total). Leuconostoc, Enterococcus and Lacticaseibacillus genera were represented by 10–13 strains (about 8–11% of the 118 isolates). Finally, there were four less-represented genera, namely Pediococcus, Levilactobacillus, Weissella, and Companilactobacillus, with only 1–2 strains (about 1–2%) (Figure 1).

At the species level, as seen in Figure 1, the relative abundance showed differences related to the product type. In rennet, for instance, the predominant species was L. plantarum (over 50% of the isolates). E. faecium/durans and L. paracasei represented 14% and 9%, respectively, while 8 other species were sporadic (less than 5% of the isolates). On the other hand, in cheese, Lact. lactis had the highest abundance (over 60% of the isolates), followed by L. plantarum (about 25%). Lact. garviae, L. paracasei, and E. faecium/durans were occasionally found in some cheese samples. About 60% of the bacteria isolated from whey belonged to Leuc. citreum and Lact. lactis (with similar relative abundances), and 14% belonged to L. plantarum. Other species were present in whey, but in low numbers.

The microbial distribution per sample (Figure 2) varied among manufacturers and among the product type. Rennet samples collected from manufacturers 1, 5, and 6, had the lowest LAB diversity (one or maximum two different species).

All five rennet samples from producer 1 contained L. paracasei as the unique species, or associated with E. faecium/durans. On the contrary, rennet samples from manufacturers 5 and 6 contained L. plantarum, alone or associated with E. faecium/durans. Rennets from the other four producers had a higher diversity, with up to four different LAB species. Among these species, L. plantarum was always present, in association with Leuc. citreum or E. faecium/durans. Lact. lactis was only found in two rennets, namely: P4R1 and P7R1. A high LAB diversity was also observed for most of the whey samples (3–4 species), except the samples from producer 2 and 3, from which only one or two species were isolated. In five of the six samples, Leuc. citreum was present alone or associated with other species. L. plantarum was also found in 3 whey samples, associated with Lact. lactis and other species. The highest diversity was observed for P4W1 and P6W2, with four different species, among which Lact. lactis and Leuc. citreum were common. Regarding the cheese samples, Lact. lactis was found in all eight samples, usually associated with L. plantarum. In the two cheeses from producer 1, Lact. lactis was only associated with L. paracasei.

Sixty-three unique strains (with a unique strain defined as one representative strain of one bacterial species found in a particular sample) were further used for the screening for technological and functional properties.

3.3. Growth and Acidification in Milk and Synthetic Media

Most of the strains grew well in the MRS broth, reaching an OD_600nm of over 1 after 24 h (Figure 3). There were, however, some very fast-growing strains, such as L. plantarum R19 and R34.1, originating from rennet and cheese, respectively, which reached an OD_600nm equal to or over 1 after 6 h of incubation, and over 2 after 12 h of incubation. On the other hand, Lact. lactis R47, and E. faecium R2 were among the slowest-growing strains, with an OD_600nm of less than 0.5 after 24 h.

Regarding the acidification rate, results depended on the medium used for the growth. As seen in Figure 4, a pH variation (ΔpH) of over 1 unit was measured in the MRS medium for most of the strains after 24 h of growth. A small increase in the median ΔpH value was observed after another 24 h of incubation. Two L. plantarum strains (R19 and R128, both isolated from rennet) stand out as fast-acidifying in MRS, with the highest ΔpH, both at 6 h and at 24/48 h of growth.

In both types of milk, the median ΔpH values were higher than in the MRS, especially after 24 and 48 h of growth. Few Lact. lactis strains were able to decrease the pH by more than 1 unit after only 6 h of growth in cow’s milk (R107, R47, and R152, all three isolated from cheese), and goat’s milk, respectively (R38, isolated from rennet, and R151 from cheese). Lact. lactis R120 and R134 were able to decrease the pH by over 2 units after 24 h, while Lact. lactis R155 and L. casei R157 decreased the pH by 2–2.5 units after 48 h of in both types of milk.

3.4. Screening for Functional Properties

3.4.1. Genetic Screening for Functional Genes

The most frequently found genes among the tested strains were ribA, α-amy, and gad, responsible for riboflavin, α-amylase, and glutamate decarboxylase synthesis, respectively (Figure 5). They were found in over 65% of the strains, with the highest incidence for ribA, of about 75%. Sixteen (about 25%) of the analyzed strains were shown to harbor the gene for α-glucosidase (agl), while the folK gene (for folate production) was present only in five strains.

Genes associated with specific bacteriocins were detected in several strains, including the nisA gene (nisin synthesis) in seven Lactococcus lactis strains, and the pln gene (plantaricin synthesis) in two L. plantarum strains. However, the lcn gene (lactococcin synthesis), was not found in any strain. No amplification was detected when using specific primers for genes involved in EPS synthesis. When looking at the co-occurrence of the analyzed genes, we observed that in most of the strains (50 out of 63) there were two or more genes present. In 13 of the strains (of which twelve are L. plantarum) there were four to five genes present, especially ribA, gad, α-amy, and agl. In another 20 strains, in which we detected three genes, there was a constant co-occurrence of gad and ribA, while the third gene was variable. When two genes were detected (in 17 strains), no rules of co-occurrence could be observed, but when only one gene was present (in 11 strains), this was, with 2 exceptions, α-amy gene.

3.4.2. Antibacterial Activity

Except for five strains (with no inhibitory effect), all tested cultures showed antibacterial activity against at least three indicator strains, with varying intensities, as indicated by the diameters of the inhibition zone ranging from 15 to 65 mm. Noticeable activities were detected for ten strains (

Table 3. Antibacterial activity of selected strains, measured by the diameter of the clear inhibition zone. The isolation source is mentioned for each strain (R—rennet, C—cheese, W—whey).

Bacterial Strain (Isolation Source)	Inhibition Zone (mm in Diameter)
	L. delbrueckii LMG 6901T	E. coli ATCC 25922	Salm. enterica ATCC 14028	List. monocytogenes ATCC 1911-1	Staph. aureus ATCC 25923	B. cereus CBAB	B. subtilis ATCC 6633
L. plantarum R137 (W)	48	0	30	65	60	60	0
Lact. lactis R38 (R)	45	22	26	30	30	0	60
Lact. lactis R151 (C)	41	40	22	32	40	46	33
L. plantarum R123 (C)	38	50	25	50	60	60	50
L. plantarum R112 (R)	38	40	24	48	60	40	47
Lact. lactis R152 (C)	35	38	30	31	0	46	16
L. plantarum R41.1 (C)	31	60	25	40	54	40	38
L. plantarum R19 (R)	30	54	28	54	54	39	48
Weissella confusa/cibaria R21.2 (R)	30	45	16	42	54	46	45
L. plantarum R39 (R)	30	46	24	45	54	40	44

), isolated mostly from rennet and cheese. These included three Lact. lactis strains (R38, R151 and R152), six L. plantarum (R137, R112, R123, R41.1, R19, and R39), and one Weissella confusa/cibaria R21.2 strain. Most of these strains were able to inhibit both L. delbrueckii subsp. bulgaricus LMG6901^T, and all the pathogenic or potential pathogenic indicator strains, displaying large inhibition zones with diameters reaching up to 65 mm. It should be noted that variations in the density of overnight cultures could affect inhibition zone sizes, and therefore comparisons of antibacterial activity among strains should be interpreted with caution.

Among the ten strains with strong antibacterial activity, five were harboring a bacteriocin-encoding gene: strains R38, R151, and R152—nis gene, and strains R41.1 and R19—pln gene (Figure 5). Also, most of the ten strains were shown to be fast acidifiers (Figure 4).

3.4.3. Bacteriocin Production

When the neutralized cell-free supernatants were tested by agar spot method, a clear inhibition zone against L. delbrueckii subsp. bulgaricus LMG6901^T was only detected in case of three Lact. lactis strains, namely R38, R151, and R152. While testing the bacterial cultures did not imply a pH control, so any inhibition observed could be due to acidity from organic acid production, the use of neutralized supernatants allowed assessment of bacteriocin or other antimicrobial activity independent of acidic effects.

3.4.4. Exopolysaccharide Production

Six strains (four Leuc. citreum, one E. durans, and one W. confusa/cibaria) exhibited a mucoid phenotype when grown on modified MRS agar, containing 50 g/L sucrose. To confirm EPS production, an extraction by acetone precipitation was performed for the culture supernatants of these strains obtained in modified MRS broth. A precipitate was only obtained for strains W. confusa/cibaria R21.2 and Leuc. citreum R139, with an EPS yield of about 20, and 10 g/L, respectively (

Table 4. Exopolysaccharide production in various media by W. confusa/cibaria R21.2, isolated from rennet, and Leuc. citreum R139 from whey.

Growth Medium	EPS Yield (g/L)
	W. confusa/cibaria R21.2	Leuc. citreum R139
MRS + 50 g/L sucrose	20.1 ± 0.6	10.2 ± 0.4
Cow’s milk	n.g.	0.6 ± 0.2
Cow’s milk + 50 g/L sucrose	16.7 ± 1.9	16.7 ± 0.7
Goat’s milk	n.g.	0.7 ± 0.1
Goat’s milk + 50 g/L sucrose	16.1 ± 0.6	13.2 ± 1.7

n.g. = no growth.

). When cultivated in cow’s or goat’s milk, strain R21.2 showed no growth after 48 h, whereas strain R139 grew well within 24 h, but produced a very low amount of EPSs (below 1 g/L). However, supplementing the milk with 50 g/L of sucrose significantly increased EPS yields in both strains. For R21.2, the yield remained lower than that achieved in modified MRS, while for R139, EPS production in milk—particularly in cow’s milk—exceeded that in MRS.

4. Discussions

Our study aimed to evaluate the LAB communities of artisanal animal rennets, and that of their associated cheese and whey from different local cheese manufacturers from Romania, as novel sources for strains with applicative potential in food industry and health. This is the first attempt of describing the LAB diversity in animal rennets collected in Romania, while very limited information exists about artisanal rennets originating from other countries, with most of the existing studies being based on culture-independent analysis [14,38].

In this work, the LAB diversity of 31 samples (of which 17 were lamb rennets) available from seven different artisanal cheese manufacturers was assessed by a culture-dependent method. After dereplication by rep-PCR, 63 unique strains, identified at species level, were kept and further analyzed. Thirty-eight unique strains, belonging to eleven LAB species, were isolated from the rennets, which are also the most diverse samples under study. Most of these strains were lactobacilli, which is in accordance with other results from the literature [14,38,39]. However, differences occur among studies when looking at species level. For instance, in our study, the most abundant species was L. plantarum (found in 12 of the 17 samples), which is also the most commonly reported lactobacilli (together with L. brevis) in the rennets used for the production of Cabrales cheese [38]. On the other hand, L. reuteri and L. fermentum were found as dominant in the rennet pastes used to manufacture Pecorino Romano cheese [39], while Cruciata et al. [14] found L. crispatus and L. reuteri as the most frequent in artisanal rennets. Such differences may be due to the fact that artisanal rennets are made by various techniques that might select distinct species, but also to the environmental differences. This can also explain the presence of L. paracasei in all five rennet samples from one producer, but not in any other.

Frequently found species in the analyzed rennets were also E. faecium/durans and L. paracasei, while other species were more sporadic, including Lact. lactis, found in only two samples. However, this was the dominant species in the associated cheese and whey, in fact it was present in all cheese samples and in four of the six whey samples. The dominance of Lact. lactis has been shown in many types of cheese during their manufacture and ripening, but it was suggested that the main source of this species is milk, and not the rennets [40,41]. Lact. lactis, a lactic acid bacterium, originates from both plant and dairy environments. While it is commonly associated with dairy products, research indicates that it likely evolved from plant-associated strains. These strains are found on fresh vegetables, fruits, and even roots or cereals, and can also be found on animal skin [41]. Nevertheless, rennet may be an important source of non-starter LAB that are involved in flavor and aroma development [42], and many other technological and probiotic properties [43], and this non-starter population is usually dominated by lactobacilli [40]. In this study, L. plantarum was detected in cheeses produced using rennet containing this species, suggesting that the rennet was the source of L. plantarum. Moreover, L. paracasei was identified in cheeses obtained from the same manufacturer, and was also isolated from the corresponding rennet, further supporting its rennet origin. Finally, enterococci were mostly present in rennet (in about 40% of the samples), but only sporadic in cheese and whey (one sample of each). This was in contradiction with other studies, reporting a higher incidence of this LAB group in dairy products made of raw milk [20]. This discrepancy may reflect differences in ecological selection during cheesemaking, but also differences in processing practices, farm hygiene, and geographic factors [44].

Most of the newly isolated strains were able to grow and acidify both cow’s and goat’s milk, with a pH decrease of about 1.5–2 units within 24 h. Some strains (most of them Lact. lactis), isolated from both rennet (three strains) and cheese (four strains) led to a fast acidification, with a pH decrease of over 1 unit within the first 6 h of growth and of about 2–2.5 units after 24 h of growth, making them suitable for use in dairy industry, thus contributing to a fast coagulation, but also to an improvement of the food’s shelf life. It is known that a low pH, caused by the organic acids generated by LAB, inhibits the growth of many spoilage microorganisms, and is one of the most important mechanisms of food bio-preservation [1]. Natural isolates are a source of new producers of bacteriocins and other antimicrobial molecules [45] so all of the different isolates were tested for antimicrobial activity against several pathogens. Most of the strains isolated in this study exhibited antibacterial activity against the indicator strains, including several (potential) foodborne pathogens and spoilage bacteria. With a few exceptions, this activity was lost in neutralized cell-free supernatants, indicating that the observed inhibition was primarily due to the production of organic acids. Moreover, strains displaying strong antibacterial activity were also identified as fast acidifiers when grown in an MRS medium or milk, likely reflecting a high level of organic acid production. This characteristic may further account for the non-specific inhibitory effects observed, which are typically associated with organic acids [46].

On the other hand, the detection of nisA and pln genes in several Lactococcus lactis and L. plantarum strains indicates a potential for the production of nisin and plantaricin, respectively, suggesting that these compounds may contribute to the antimicrobial activity observed among the tested LAB. Although seven Lactococcus lactis strains were positive for the nisA gene, only three of them retained antibacterial activity against L. delbrueckii subsp. bulgaricus 6901ᵀ after pH neutralization.

These results, correlated with the limited inhibitory spectrum common for bacteriocins [47], led us to the assumption that, for the three Lact. lactis strains, the antibacterial activity is due to the production of nisin. Although the genes involved in nisin, plantaricin, or lactococcin synthesis were detected in the genomes of several other strains, no inhibitory activity was observed in the neutralized supernatants of the harboring strains. This discrepancy may be due to the presence of incomplete nisin gene clusters, as only the structural gene (nisA) was analyzed in this study. The absence of other genes required for nisin biosynthesis, modification, regulation, or secretion (e.g., nisB, nisC, nisT, nisRK) could hinder bacteriocin production [48]. Moreover, nisin synthesis is known to be tightly regulated and influenced by environmental and physiological conditions, which might also contribute to the observed variability in activity among the nisA-positive strains [49].

Although genetic screening did not detect genes associated with EPS biosynthesis, two strains were found to produce high levels of EPS when grown in an MRS medium as well as in cow’s and goat’s milk supplemented with sucrose. This discrepancy may be explained by the presence of uncharacterized or strain-specific EPS gene clusters that are not targeted by the primers used, highlighting the potential for novel EPS biosynthesis pathways in these strains [50]. Further genomic analyses would be necessary to fully identify and characterize these genes. The observed increase in EPS production by strains R21.2 and R139 in sucrose-supplemented milk may be linked to their strain-specific metabolic capabilities. Sucrose can serve as a preferred carbon source for EPS biosynthesis, and differences in carbohydrate metabolism among strains could explain the enhanced yields. Although the current study did not investigate the underlying metabolic pathways, this phenomenon suggests that sucrose availability may regulate EPS synthesis [51].

Several other strains displayed a mucoid phenotype on solid media, but they did not produce extractable EPSs in liquid culture. This result may be due to medium-specific induction of EPS synthesis. Growth on solid surfaces can stimulate EPS production through surface attachment or stress-related signals that are absent in liquid media [52]. Therefore, EPS expression appears to be influenced by environmental conditions, and the mucoid appearance on agar does not necessarily predict the quantity of EPSs produced in liquid culture. Moreover, mucoid phenotype may be due to the production of capsular polysaccharides, that are tightly associated with the cell surface, and may not be readily released into the surrounding medium during liquid culture extraction [8].

These strains, capable of synthesizing EPSs during fermentation, could be beneficial for the food industry by reducing syneresis and enhancing the sensory qualities of the final product, all without incurring additional costs.

Genomic studies have shown the potential of the screened strains for production of several vitamins, enzymes and GABA. However, further studies are needed to confirm the biosynthesis of these metabolites. Forty-six of the tested strains have been found to carry the gene responsible for riboflavin production (ribA), while five strains possessed the gene for folate production (folK), with most of them also carrying ribA. Previous studies have demonstrated that the ability of LAB to produce vitamins is strain-dependent [53,54], and such vitamin-producing strains could potentially be utilized to boost the levels of B-group vitamins in fermented foods [55]. Therefore, our findings open opportunities to use strains isolated from rennet or dairy products for this purpose.

The gene encoding glutamate decarboxylase (gad), a key enzyme in GABA production, was also identified in a significant number of genomes (40 strains). This finding was anticipated, as it is well-established that LAB, particularly lactobacilli, as well as Lact. lactis and Streptococcus thermophilus, are important producers of GABA [56,57,58,59].

Finally, genes encoding enzymes involved in carbohydrate digestion were identified in the analyzed genomes. Specifically, the α-amy gene was found in 42 strains, while the agl gene was present in 17 strains. The production of α-amylase and α-glucosidase by strains used in fermented dairy foods could, on one hand influence fermentation efficiency, and on the other hand have beneficial effects on the digestibility, organoleptic properties, and nutritional value of the foods [60].

5. Conclusions

The LAB diversity of artisanal animal rennets collected in Romania was dominated by L. plantarum, followed by E. faecium/durans and L. paracasei, while Lact. lactis was the most prevalent in cheese. However, in half of the cheese samples, L. plantarum was also found, likely originating from the rennet, and may contribute to the development of cheese flavor and aroma. The isolated LAB strains were shown to possess functional properties, including antibacterial activity against various pathogenic or potentially pathogenic bacteria, EPS production (with one strain producing about 20 g/L EPS), and nisin production. However, no distinct distribution pattern of specific functional characteristics was observed among LAB strains isolated from different dairy ecosystems. In addition, this study identified the potential for the production of vitamins, GABA, and certain enzymes based on the presence of specific genes in the genomes of the strains, but these traits require further experimental confirmation.