Prevalence, Virulence, and Antibiotics Gene Profiles in Lactococcus garvieae Isolated from Cows with Clinical Mastitis in China

Lactococcus garvieae (L. garvieae) is a pathogenic gram-positive, catalase-negative (GPCN) bacterium that causes bovine mastitis. A total of 49 L. garvieae isolates were identified from 1441 clinical mastitis (CM) samples. The pathogenic effects of L. garvieae were studied with two infection models: bovine mammary epithelial cells cultured in vitro and murine mammary infections in vivo. The overall farm prevalence was 15.5% (13/84 farms in 9/19 provinces) and sample prevalence was 3.40% (49/1441). Post-treatment somatic cell count (SCC) post L. garvieae infection was significantly higher than the other GPCN pathogens isolated, and the bacteriological cure fraction was 41.94% (13/31) after intramammary antibiotic treatment. All L. garvieae isolates were resistant to rifaximin, 12.24% of isolates were resistant to cephalexin, and 10.20% (5/49) were multidrug-resistant (MDR). The most prevalent virulence genes were Hemolysin 1 (hly1)(100%), Hemolysin 2 (hly2) (97.96%), NADH oxidase (NADHO) (100%), Superoxide dismutase (SOD) (100%), Adhesin Pav (Pav) (100%), Adhesin PsaA (PsaA) (100%), Enolase (eno) (100%), Adhesin cluster 1(AC1) (100%), Adhesin cluster 2 (AC2) (100%), and several exopolysaccharides. L. garvieae rapidly adhered to bovine mammary epithelial cells, resulting in an elevated lactate dehydrogenase release. Edema and congestion were observed in challenged murine mammary glands and bacteria were consistently isolated at 12, 24, 48, 72, and 120 h after infection. We concluded that L. garvieae had good adaptive ability in the bovine and murine mammary cells and tissue. Given the resistance profile, penicillin and ampicillin are potential treatments for CM cases caused by L. garvieae.


Introduction
Mastitis, an inflammatory process of the mammary gland, is the most common bacterial disease [1] and one of the most costly diseases of dairy cattle [2]. Lactococcus species have been known to be associated with mastitis since as early as 1932 [3]. However, GPCN streptococci or streptococci-like bacteria including Streptococcus, Enterococcus, Aerococcus, and Lactococcus are phenotypically and biochemically alike. This means the identification of Lactococcus species has been inaccurate and unreliable in many studies and diagnostic laboratories [4]. The incidence of Lactococcus species identified on-farm may have been historically underreported or was phenotypically identified as Streptococcus uberis (S. uberis) or Streptococcus spp. [5]. Thus, little is known about the clinical importance of this genus as a mastitis pathogen, and awareness and focus have only increased in recent years.
Based on phenotypic similarities, Lactococcus species were initially assigned to the genus Streptococcus, and a new genus was assigned to Lactococcus in 1985 [6]. L. garvieae was known as an emergent disease affecting many fish species and it is considered a potential zoonotic microorganism [7]. This is because it is known to cause several opportunistic human infections, such as endocarditis [8], diverticulitis [9], peritonitis [9], infective Microorganisms 2023, 11, 379 3 of 20 identified by farm personnel as CM samples. CM samples were aseptically collected from individual quarters by the authors of this study or trained on-farm personnel. The CM samples for each farm were collected within a 7 day time span. The samples were quickly frozen (−20 • C) overnight and then shipped to Shenyang Agricultural University in Shenyang, China for further identification. The CM sample collection period was from March 2020 to July 2021.

Bacterial Culture of L. garvieae
Sheep blood agar and brain heart infusion (BHI) agar were used to isolate and purify the colony of the GPCN cocci growing on sheep blood agar from 1441 CM milk samples. Then, BHI broth was used to proliferate bacteria from the colonies that grew on the agar. Gram staining and scanning electron microscopy (SEM) were carried out in order to observe the morphology of L. garvieae.

16 S rDNA Sequencing Identification and Biochemical Testing
Bacterial DNA was extracted from 248 isolates using a bacterial DNA extraction kit (Sangon Biotech (Shanghai) Co., Ltd., Shanghai, China) according to the manufacturer's instructions. The extracted DNA was used as a PCR template for amplification; GPCN streptococci-like isolates were determined by 16S rDNA sequencing [41], where primer p27f (5 -AGAGTTTGATCCTGGCTCAG-3 ) and primer 1492r (5 -TACGGCTACCTTGTTACGA CTT-3 ) were used to amplify a 1460-bp product of the 16S rDNA gene. The PCR cycling conditions included an initial denaturation step at 95 • C for 3 min, followed by 35 cycles at 95 • C for 15 s, 55 • C for 15 s, and 72 • C for 1 min, with a final extension step at 72 • C for 5 min. The PCR products were subjected to sequencing (Sanger sequencing by Sangon Biotech (Shanghai) Co., Ltd., Shanghai, China) after verification on 1.2% agarose gel. The 16S rDNA sequences were compared with sequences deposited in the nucleotide database of the National Center for Biotechnology Information. Identification was deemed reliable if the values for sequence similarities were ≥99%. The biochemical reacting kit (Qingdao Hi-Tech Industrial Park Haibo Biotechnology Co., Ltd., Qingdao, China) was used for biochemical testing. A total of 11 reagents (ribose, sucrose, lactose, liquid gelatin, sorbitol, maltose, esculin, galactose, VP, trehalose, and glucose) were fermented with isolates, following the manufacturer's instructions. Briefly, isolates were seeded in the tubes that were subsequently cultured in an incubator at 37 • C for the required time; some of the reagents needed further operations and the colors were compared with negative tubes.

Post-Treatment Milk Sample Collection for SCC and Bacteriological Cure Evaluation
The post-intramammary-antibiotic-treatment milk samples were collected from the farm with the highest prevalence of L. garvieae. The dairy farm, located in northwest China, had an average of 6400 milking cows during the study period. Cows were fed a TMR and housed in freestall barns with sand bedding. Lactating cows were milked thrice daily in two rotary parlors. Milk samples were aseptically collected from the same individual quarter first identified as CM 17 ± 3 d after an extended therapy of 5 d of antimicrobial intramammary treatment (Ubrolexin, Boehringer Ingelheim, Ingelheim am Rhein, Germany) instead of the standard 2 d regimen, and with anti-inflammatory treatment (meloxicam, Boehringer Ingelheim, Germany) for bacteriological cure evaluation. A bacteriological cure was defined as being L. garvieae culture-positive in the clinical mastitis sample, and culture-negative in the post-treatment milk sample. After collection, the quarter milk sample was shaken several times to ensure good mixing and then tested for SCC with the DeLaval DCC instrument (Delaval (Tianjian) Co., Ltd., Tianjin, China).

Growth Curve of Lactococcus garvieae
The growth curves of one isolate of L. garvieae (LG41), one isolate of L. lactis, one strain of Staphylococcus aureus (S. aureus), and one isolate of Enterococcus faecalis (E. faecalis) were assessed simultaneously. The S. aureus was ATCC 29213, and the remaining isolates from CM milk samples from the same dairy farm had the highest L. garvieae prevalence. The mediums were prepared according to the manufacturer's instructions. For each isolate, 30 µL of the bacterial solution was added to 3 mL BHI in 5 mL sterile tubes for each different isolate, and they were placed on a constant temperature shaker (37 • C, 220 rpm). At 0, 0.5, 1, 1.5, 2, 2.5, 3, 3.5, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, and 24 h, the optical density (OD) of each bacterial suspension was determined using 3 tubes per isolate at 600 nm in a UV spectrophotometer (Shimadzu Corporation, Kyoto, Japan). Each experiment was performed in triplicate.

Cytotoxic Lactate Dehydrogenase (LDH) Release Assay
LDH release was used to identify the most and least cytotoxic isolates among the 49 isolates for further pathogenicity studies, as well as to study the cytotoxic effects of L. garvieae on bovine mammary alveolar cell T (MAC-T) (Shanghai Baiye Biotechnology Center, Shanghai, China). This was assessed using an LDH assay kit (Beyotime Biotechnology, Beijing, China). Cells were cultured at 37 • C with 5% CO 2 in 96-well plates (Corning Inc., Corning, NY, USA) and confluent growth (approximately 80% full) was achieved, then challenged with different L. garvieae isolates (n = 49) at a multiplicity of infection (MOI, ratio of L. garvieae to cells) of 5:1 for 12 h. The most cytotoxic isolate should have the highest LDH release, and the least cytotoxic should have the lowest LDH release; thus, the most and least cytotoxic isolates were selected for mouse mastitis model experiments. Then, MAC-T cells were cultured again at 37 • C with 5% CO 2 in 96-well plates and challenged with the most and least cytotoxic isolates with an MOI of 5:1 at 1, 3, 6, 12, and 24 h. Uninfected cells were cultured as the control group. After incubation, 200 µL of the supernatant was collected from each well and transferred to a centrifuge tube and centrifuged (8000× g, 5 min at 4 • C). Then, 120 µL of the supernatant was transferred to a new 96-well polystyrene plate and 60 µL of reaction mixture was added to each well. The reaction mixture was then incubated in the dark on a rotating shaker (150 rpm) at room temperature for 30 min. The absorbance was read at 490 nm (QuantStudio3, Thermo Fisher, Waltham, MA, USA). Each experiment was performed in triplicate.

Adhesion Assay
To assess the adhesion capacity of L. garvieae to MAC-T, the bacterial adhesion of two L. garvieae isolates (LG41 and LG47) was slightly modified as described [2]. The MAC-T were cultured in 6-well plates (Corning Inc.) and confluent growth (approximately 80% full) was achieved in an antibiotic-free medium, followed by infection with L. garvieae at an MOI of 50:1 for 30 min, 1, 2, and 3 h, and cultured at 37 • C and 5% CO 2 . After incubation, cells were washed twice with sterile PBS (pH 7.4) to remove unbound bacteria. Adhered bacteria were released by adding 1 mL of PBS and 1 mL of 1% triton X-100 (0.5% vol/vol) to lyse cells. Both the bacterial suspension (1 mL) and cells in the control group were treated with 1 mL triton X-100. The cell lysates and treated bacterial supernatant of the infected group and bacterial supernatant of the control group were diluted using a 10-fold serial method, cultured on SBA, incubated at 37 • C for 24 h, and CFU counts were determined. The adhesion fraction was determined as follows: Bacterial adhesion fraction = bacterial CFU count of Cell lysates of infected group(CFU/mL) bacterial CFU count of control group (CFU/mL) × 100% The adhesion assays were repeated 3 times, in triplicate for each test.

Morphology of Lactococcus garvieae on MAC-T
Gram staining and SEM of cell slides were carried out in order to observe the morphology of the co-culture of MAC-T and the two L. garvieae isolates (LG41 and LG47). The MAC-T were cultured in 6-well plates (Corning Inc.) for 2 days and grown to confluence in an antibiotic-free medium, followed by infection with L. garvieae (MOI 50:1) for 24 h. After incubation, cells were washed twice with sterile PBS (pH 7.4) to remove unbound bacteria. Gram staining was performed with stained with hematoxylin-eosin. After co-culturing for 24 h, an electron microscope fixing solution was added rapidly to fix the sample at room temperature for 2 h. Then, the samples were rinsed 3 times with 0.1 M phosphate buffer Pb (pH 7.4) for 15 min each time, and the samples were fixed with 1% osmic acid with 0.1 M phosphate buffer Pb (pH 7.4) at room temperature in the dark for 1-2 h. Afterwards, they were rinsed 3 times with 0.1 M phosphate buffer Pb (pH 7.4) for 15 min each time. After that, 30%-50%-70%-80%-90%-95%-100%-100% ethanol was injected into the tissue for 15 min each time. The final process of dehydration involved adding isoamyl acetate for 15 min; then, the sample was put into the critical point dryer for drying. The dried sample was placed on the sample table of the ion sputtering instrument and sprayed with gold for approximately 30 s. Finally, a scanning electron microscope was used to observe and collect pictures.

Lactococcus garvieae Experimental Infection in a Mouse Mammary Gland
The pathogenic effect of two L. garvieae isolates (LG41 and LG47) during intramammary infection was determined using 6-8 week old female specific-pathogen-free BALB/c mice (Liaoning Changsheng Biotechnology Co., Ltd., Benxi, China) [43]. Pregnant (19 d Microorganisms 2023, 11, 379 7 of 20 of gestation) mice were kept in germ-free isolators and fed ad libitum in a controlled environment with light and dark cycles (12 h light and 12 h darkness). On the third day after parturition, mice were anesthetized by intramuscular injection of 50 mg/kg Zoletil 50 (Virbac, Carros, France). The fourth pair of mammary glands ducts were exposed by cutting the teat tip and 50 µL of bacterial suspension (5 × 107 CFU) was slowly injected using a small-gauge blunt-tipped needle (Guangdong Xiapute Technology Co., Ltd., Yangjiang, China). Three groups (n = 25 per group) of mice were allocated as 2 challenge groups (LG41 and LG47, respectively) and 1 negative control group (sterile PBS). The pups were removed 1 h before intramammary inoculation. The sedated mice were euthanized with cervical dislocation. The skin was fixed using pins before photographing the mammary glands. The bacterial load in the mammary glands at 12, 24, 48, 72, and 120 h after challenge (5 mice per time point) was measured as described [43]. Briefly, mammary gland tissue (0.1 g) was separated into a sterile Petri dish under a germ-free environment. After homogenization, 50 µL of supernatant was spread on sheep blood plates (multiple dilutions). The numbers of viable colonies were expressed as CFU/g. Mammary gland tissue was fixed with 5% paraformaldehyde, and embedded, sectioned, and stained with hematoxylin-eosin. Histological evaluation was performed to assess tissue necrosis, polymorphonuclear neutrophilic granulocyte inflammation (i.e., neutrophilic inflammation), and lymphocytic inflammation, as described [43]. The flow diagram showing the sample collection and identification for enrollment in the in vivo and in vitro study are shown in Figure 1.

Statistical Analyses
SPSS 22.0 (SPSS Corporation, Chicago, Illinois) was used to analyze the data. way analysis of variance (ANOVA) was used to compare post-treatment SCC betwe garvieae and L. lactis, or other pathogen infected CM milk samples, LDH release, inva and adhesion fractions between the treatment groups, and the Duncan test was us determine the difference. If p < 0.05, the difference was considered to be statistically nificant.

Morphological Characteristics of Lactococcus garvieae
L. garvieae formed round, medium-sized (approximately 1-2 mm in diam colonies on sheep blood TSA plates, with smooth edges, moist surfaces, and α light g 2.14. Statistical Analyses SPSS 22.0 (SPSS Corporation, Chicago, IL, USA) was used to analyze the data. Oneway analysis of variance (ANOVA) was used to compare post-treatment SCC between L. garvieae and L. lactis, or other pathogen infected CM milk samples, LDH release, invasion and adhesion fractions between the treatment groups, and the Duncan test was used to determine the difference. If p < 0.05, the difference was considered to be statistically significant.

Morphological Characteristics of Lactococcus garvieae
L. garvieae formed round, medium-sized (approximately 1-2 mm in diameter) colonies on sheep blood TSA plates, with smooth edges, moist surfaces, and α light green hemolysis around the colony. The colony morphology and hemolysis was very similar to streptococcus after incubation at 37 • C for 24 h (Figure 2A). The bacteria stained gram-positive, and, microscopically, the bacterial body appeared spherical or ellipsoidal. The arrangement shape was either two to three bacteria lined up in short chains, or a single bacterium was present ( Figure 2B). The results of the scanning electron microscopy showed that the bacteria were ellipsoidal in shape and approximately 1.5-2 µm in diameter. (Figure 2C,D).

Identification of Suspected Isolates by 16S rDNA Sequence
After the initial isolation and identification of the 1441 milk samples, GPCN isolates were subjected to 16S rDNA gene sequencing. A total of 248 GPCN cocci iso (all suspected) were sequenced with 16S rDNA gene fragment amplicons, and the t bacteria with the highest percentage were L. garvieae (19.76%), L. lactis (16.53%), and S tococcus agalactiae (13.71%). The details are shown in Table 2 Table S1.

Identification of Suspected Isolates by 16S rDNA Sequence
After the initial isolation and identification of the 1441 milk samples, GPCN cocci isolates were subjected to 16S rDNA gene sequencing. A total of 248 GPCN cocci isolates (all suspected) were sequenced with 16S rDNA gene fragment amplicons, and the three bacteria with the highest percentage were L. garvieae (19.76%), L. lactis (16.53%), and Streptococcus agalactiae (13.71%). The details are shown in Table 2. Positive samples of L. garvieae were isolated from nine provinces. A total of 49 strains of L. garvieae were isolated and identified on 13 farms, with a sample detection frequency of 3.40% (49/1441). Between farms, there was a positive detection frequency of 16.25% (13/84). A farm in Ningxia had the highest detection frequency, with 31 strains of bacteria isolated from 149 clinical mastitis milk samples, and a detection frequency of 20.81%. The distribution of L. garvieae isolated from different Chinese commercial dairy farms are listed in Supplementary Materials in Table S1.

Post-Treatment SCC and Bacteriological Cure
Post-treatment SCC of L. garvieae (31 isolates), L. lactis (7 isolates), and other bacteria (7 L. lactis isolates, 2 Aerococcus viridans isolates, 1 Enterococcus faecium isolate, and 1 S. uberis isolate) are shown in Figure 3. Post-treatment SCC of L. garvieae infections was not significantly different to L. lactis but was significantly different from other bacterial isolates. The bacteriological cure fraction was 41.94% (13/31) for L. garvieae, 71.43% (5/7) for L. lactis, and 54.55% (6/11) for other bacteria (detailed bacteriological cure data not shown). There was no significant difference in the bacterial cure percentage, which is not surprising given the low numbers in the L. lactis and other groups. All 31 cows identified as being infected with mastitis caused by L. garvieae were classified as mild CM cases (abnormal milk only); however, eight had a recurrence within 30 days after initial diagnosis. The recurrence rate was 25.81% (8/31), while the recurrence rate of other pathogens was 9.10% (1/11).

Post-Treatment SCC and Bacteriological Cure
Post-treatment SCC of L. garvieae (31 isolates), L. lactis (7 isolates), and other bacteria (7 L. lactis isolates, 2 Aerococcus viridans isolates, 1 Enterococcus faecium isolate, and 1 S. uberis isolate) are shown in Figure 3. Post-treatment SCC of L. garvieae infections was not significantly different to L. lactis but was significantly different from other bacterial isolates. The bacteriological cure fraction was 41.94% (13/31) for L. garvieae, 71.43% (5/7) for L. lactis, and 54.55% (6/11) for other bacteria (detailed bacteriological cure data not shown). There was no significant difference in the bacterial cure percentage, which is not surprising given the low numbers in the L. lactis and other groups. All 31 cows identified as being infected with mastitis caused by L. garvieae were classified as mild CM cases (abnormal milk only); however, eight had a recurrence within 30 days after initial diagnosis. The recurrence rate was 25.81% (8/31), while the recurrence rate of other pathogens was 9.10% (1/11).

Growth Ability of Lactococcus garvieae
Growth curves of L. garvieae, L. lactis, S. aureus, and E. faecalis isolates cultured in BHI broth are shown in Figure 4. For L. garvieae, the bacterial growth curve consisted of a lag phase (~2 h), a log phase (~4 h), and ultimately, a stationary phase. Based on subjective

Growth Ability of Lactococcus garvieae
Growth curves of L. garvieae, L. lactis, S. aureus, and E. faecalis isolates cultured in BHI broth are shown in Figure 4. For L. garvieae, the bacterial growth curve consisted of a lag phase (~2 h), a log phase (~4 h), and ultimately, a stationary phase. Based on subjective observations, the growth curve of L. garvieae had a similar lag phase to other isolates. Additionally, the OD600nm value of L. garvieae isolates reached as high as~1.0, whereas it was almost the same for L. lactis and up to~1.3 for S. aureus and E. faecalis isolates.

Antimicrobial Resistance Profiles of Lactococcus garvieae
All 49 L. garvieae isolates were susceptible to penicillin, ampicillin, ceftiofur, and cefquinome among the β-lactam antibiotics, but 12.24% were resistant to cephalexin. All L. garvieae isolates were resistant to lincomycin and rifaximin, and 73.47% of isolates were resistant to oxytetracycline. All L. garvieae isolates were sensitive to marbofloxacin and vancomycin, see Table 4. As L. garvieae is intrinsically resistant to clindamycin [44], it was excluded from the drug-resistant (DR) or multidrug-resistant (MDR) statistics in this paper. In summary, 89.20% (44/49) of L. garvieae were DR and 10.20% (5/49) were MDR.

Antimicrobial Resistance Profiles of Lactococcus garvieae
All 49 L. garvieae isolates were susceptible to penicillin, ampicillin, ceftiofur, and cefquinome among the β-lactam antibiotics, but 12.24% were resistant to cephalexin. All L. garvieae isolates were resistant to lincomycin and rifaximin, and 73.47% of isolates were resistant to oxytetracycline. All L. garvieae isolates were sensitive to marbofloxacin and vancomycin, see Table 4. As L. garvieae is intrinsically resistant to clindamycin [44], it was excluded from the drug-resistant (DR) or multidrug-resistant (MDR) statistics in this paper. In summary, 89.20% (44/49) of L. garvieae were DR and 10.20% (5/49) were MDR.

LDH release 12 h after infection among different isolates indicated that
LG41 was the most cytotoxic, while LG47 was the least cytotoxic. Therefore, these two isolates were selected. At 1, 3, 6, 12, and 24 h after infection, LDH release of LG41 was higher than in the control group (p < 0.01). At 1, 3, and 24 h after infection, LDH release of LG47 was higher than in the control group (p < 0.05). At 3, 6, 12, and 24 h after infection, LDH release of LG41 was higher than LG47 (p < 0.01, Figure 6A). The LG41 isolate adhered to MAC-T at a higher frequency compared with LG47 at the different time points (p < 0.01, Figure 6B).
Microorganisms 2023, 11, x FOR PEER REVIEW 13 of

Pathogenic Effects of Lactococcus garvieae on MAC-T
LDH release 12 h after infection among different isolates indicated that LG41 was t most cytotoxic, while LG47 was the least cytotoxic. Therefore, these two isolates were lected. At 1, 3, 6, 12, and 24 h after infection, LDH release of LG41 was higher than in t control group (p < 0.01). At 1, 3, and 24 h after infection, LDH release of LG47 was high than in the control group (p < 0.05). At 3, 6, 12, and 24 h after infection, LDH release LG41 was higher than LG47 (p < 0.01, Figure 6A). The LG41 isolate adhered to MAC-T a higher frequency compared with LG47 at the different time points (p < 0.01, Figure 6

Morphology of Lactococcus garvieae on MAC-T
The cells in the control group were closely attached to the round coverslip, the c morphology was paving stone-like, the surface of the cell membrane was covered w rich microvilli, the cells were arranged in an orderly manner, and there were elongat and rich pseudopods scattered in the cells, which was conducive to cell sticking ( Figu  7A,B). After 24 h of LG41 challenge, the microvilli on the cell surface were broken, a lar number of bacteria (red arrow) adhered to the cell surface, and there was damage to t cell surface where bacteria were attached. This phenomenon may be a manifestation endocytosis (Figure 7C,D). After 24 h of LG47 challenge, there was a much smaller nu ber of bacteria (red arrow) adhered to the cell surface than the LG41 group ( Figure 7E, Data are mean ± SD of three independent experiments. * indicates a significant difference between different treatment groups (p < 0.05 by ANOVA test), and ** indicates a very significant difference between different treatment groups (p < 0.01 by ANOVA test). (B) Adhesion of L. garvieae (up to 3 h after infection) into MAC-T. ** indicates a very significant difference between different treatment groups (p < 0.01 by ANOVA test). The bar charts were calculated and plotted using GraphPad prism 8.

Morphology of Lactococcus garvieae on MAC-T
The cells in the control group were closely attached to the round coverslip, the cell morphology was paving stone-like, the surface of the cell membrane was covered with rich microvilli, the cells were arranged in an orderly manner, and there were elongated and rich pseudopods scattered in the cells, which was conducive to cell sticking ( Figure 7A,B). After 24 h of LG41 challenge, the microvilli on the cell surface were broken, a large number of bacteria (red arrow) adhered to the cell surface, and there was damage to the cell surface where bacteria were attached. This phenomenon may be a manifestation of endocytosis ( Figure 7C,D). After 24 h of LG47 challenge, there was a much smaller number of bacteria (red arrow) adhered to the cell surface than the LG41 group ( Figure 7E,F).

Inflammation of Murine Mammary Gland Infected by Lactococcus garvieae
Edema and hyperemia were evident in murine mammary glands at 12 h after infec tion with L. garvieae, with more profound pathological changes seen over time (24,48, and 72 h after infection; Figure 8A). Histological characteristics of the mammary glands in fected with L. garvieae were observed ( Figure 8B). The structure of the gland alveoli wa destroyed, the walls of the gland alveoli was thicker, and infiltrating inflammatory cell (mainly neutrophils) were observed in the gland alveoli and interstitium of the infected mammary gland at 24, 48, 72, and 120 h post infection. At 48 h post infection, the LG41 group showed interstitial tissue hyperplasia. Acute inflammation resolved at 120 h afte infection, leaving connective tissue to fill the gap after epithelial death. No evidence o inflammation in mice from the control group (uninfected) was observed.

Inflammation of Murine Mammary Gland Infected by Lactococcus garvieae
Edema and hyperemia were evident in murine mammary glands at 12 h after infection with L. garvieae, with more profound pathological changes seen over time (24,48, and 72 h after infection; Figure 8A). Histological characteristics of the mammary glands infected with L. garvieae were observed ( Figure 8B). The structure of the gland alveoli was destroyed, the walls of the gland alveoli was thicker, and infiltrating inflammatory cells (mainly neutrophils) were observed in the gland alveoli and interstitium of the infected mammary gland at 24, 48, 72, and 120 h post infection. At 48 h post infection, the LG41 group showed interstitial tissue hyperplasia. Acute inflammation resolved at 120 h after infection, leaving connective tissue to fill the gap after epithelial death. No evidence of inflammation in mice from the control group (uninfected) was observed. Bacteria were isolated from the mammary glands of mice challenged with L. garvieae, whereas no bacteria were isolated from the non-infected control group. The bacterial load (mean value) was 3.80 × 10 8 CFU/g at 12 h after infection, but rapidly increased to 2.10 × 10 9 CFU/g at 24 h after infection and started to drop to 7.55 × 10 7 CFU/g at 48 h, 4.31 × 10 5 CFU/g at 72 h, and 7.49 × 10 4 CFU/g at 120 h ( Figure 8C).

Discussion
This is the first report of L. garvieae associated with bovine CM cases from multiple farms. L. garvieae is an emerging pathogen that has been confirmed with molecular testing methods such as PCR [5], RAPD, REP-PCR, MLRT, and MALDI-TOF [8]. The strain (98/4289) isolated from water was genetically more closely related to that from bovines than fish-oriented strains. This raises the possibility some environmental L. garvieae strains having evolved from mammals, and this may be involved in the epidemiology of fish lactococcosis [20]. This highlights the importance of implementing screening for L. garvieae as an emerging zoonotic bacterium. Humans, particularly those who have an anatomically or physiologically altered gastrointestinal tract or coexisting local predisposing health problems, are considered to be at-risk individuals. Some human cases have been associated with consuming raw seafood [8]. Consuming raw milk, therefore, has the potential for serious morbidity and mortality [10].
In the present study, the farms suffering from L. garvieae infection were using sand bedding. Sand bedding can be a reservoir of L. garvieae strains and be a potential vehicle for their dissemination in dairy farms. Contaminated sand bedding could also transfer infection between cows [21]. When L. garvieae was initially identified as causing bovine mastitis outbreaks on farms, it is possible the Lactococcus strain already existed and changes in the environment selectively favored the strain responsible for the outbreak. Alternatively, a new Lactococcus strain could have been introduced. Lactococcus has been found in samples from mastitic and normal milk, the bulk tank, and sand bedding. The Bacteria were isolated from the mammary glands of mice challenged with L. garvieae, whereas no bacteria were isolated from the non-infected control group. The bacterial load (mean value) was 3.80 × 10 8 CFU/g at 12 h after infection, but rapidly increased to 2.10 × 10 9 CFU/g at 24 h after infection and started to drop to 7.55 × 10 7 CFU/g at 48 h, 4.31 × 10 5 CFU/g at 72 h, and 7.49 × 10 4 CFU/g at 120 h ( Figure 8C).

Discussion
This is the first report of L. garvieae associated with bovine CM cases from multiple farms. L. garvieae is an emerging pathogen that has been confirmed with molecular testing methods such as PCR [5], RAPD, REP-PCR, MLRT, and MALDI-TOF [8]. The strain (98/4289) isolated from water was genetically more closely related to that from bovines than fish-oriented strains. This raises the possibility some environmental L. garvieae strains having evolved from mammals, and this may be involved in the epidemiology of fish lactococcosis [20]. This highlights the importance of implementing screening for L. garvieae as an emerging zoonotic bacterium. Humans, particularly those who have an anatomically or physiologically altered gastrointestinal tract or coexisting local predisposing health problems, are considered to be at-risk individuals. Some human cases have been associated with consuming raw seafood [8]. Consuming raw milk, therefore, has the potential for serious morbidity and mortality [10].
In the present study, the farms suffering from L. garvieae infection were using sand bedding. Sand bedding can be a reservoir of L. garvieae strains and be a potential vehicle for their dissemination in dairy farms. Contaminated sand bedding could also transfer infection between cows [21]. When L. garvieae was initially identified as causing bovine mastitis outbreaks on farms, it is possible the Lactococcus strain already existed and changes in the environment selectively favored the strain responsible for the outbreak. Alternatively, a new Lactococcus strain could have been introduced. Lactococcus has been found in samples from mastitic and normal milk, the bulk tank, and sand bedding. The relative abundance of the Lactococcus genus would be higher in the microbiome of mastitic samples, compared with milk samples from healthy animals [33].
The results of the MIC tests performed in this study agree with other studies that found L. garvieae to be resistant to clindamycin [8,47]. All isolates were sensitive to penicillin, ampicillin, ceftiofur, and cefquinome. However, 12.24% of isolates were resistant to cephalexin; this may be slightly biased in that the majority of isolates were from one farm that also had the most cytotoxic isolate. The IMM antibiotic tube used by the farm at the time of the study was a combination of cefalexin and kanamycin (Ubrolexin). Using a breakpoint of 16/1.6, as used by Sorge et al. (2021), 14.9% of the L. garvieae isolates were resistant to the cefalexin-kanamycin combination. In contrast, L. garvieae had a 5.3% resistance rate to marbofloxacin [48], but in this study, all isolates were sensitive; furthermore, all isolates were resistant to rifaximin. Rifaximin has frequently been used as an antibiotic in dry cow intramammary tubes on some Chinese dairy farms. It concerns us that, if a Lactococcus IMI outbreak occurred, rifaximin might not cure the subclinical infection in dry cows. Therefore, during the next lactation, these infected cows might be more likely to have a flare-up of a clinical case of Lactococcus mastitis. These cows would also be more likely to have a high individual SCC. The results of the MIC tests performed in this study also agree with other studies that L. garvieae from fish-derived [49] and from bovine-milkderived [50] sources is resistance to tetracycline, and tetS genes were detected from all isolates. Of the 31 milk-derived L. garvieae isolates studied by Walther in 2008, 45.2% were resistant to tetracycline [50]. In comparison, the resistant fraction was 73.47% in this study. The MDR fraction of L. garvieae in this study was 10.20%, which was lower than other mastitic-milk-derived pathogenic bacteria reported in China, including 33% for S. aureus, 56% for non-aureus staphylococci, and 21% for Streptococcus species [51].
Previous reports have agreed that SCC will decline with time when a bacteriological cure is achieved, and this measure is a practical and reliable indicator of treatment success [1]. Lactococcus genus showed a lower bacteriological cure fraction and slower individual SCC resolution than Streptococcus dysaglactiae or S. uberis [52]. The bacteriological cure fraction of L. garvieae was comparable to that of S. aureus, ranging from 38.8% to 52% [53]. The bacteriological cure fraction of other GPCN bacteria isolated in this study (L. lactis, Aerococcus viridans, Enterococcus faecium, and S. uberis) were higher, suggesting that L. garvieae may be comparable to the major pathogenic bacteria in terms of the bacteriological cure fraction. On the other hand, the bacteriological cure fraction of L. lactis was significantly higher, suggesting that L. lactis might be less pathogenic than other pathogens.
Extended therapy with ceftiofur was reported to provide a greater probability of bacteriological cure for gram-positive pathogens, both in clinical and subclinical cases [54,55]. Based on these results from previous studies and the MIC result of our study, we inferred that a low bacteriological cure might be improved by extended therapy. Control measures could also include vaccines, autovaccines, bacteriophages, and antiserum [7]. In bovines, an effective dry cow antibiotic tube could also be a good control method [56].
Furthermore, hly-1 was detected in all isolates, and hly-2 was detected in 97.96% of isolates. This was demonstrated by α-hemolysis of the bacterial colonies, as shown in Figure 2. The three adhesion genes (PavA, eno, and PsaA) were detected in all isolates [34]. Pgm is a metabolic enzyme conferring resistance to peptide antimicrobials and was detected in the most cytotoxic isolate; however, it was not detected in the least cytotoxic isolate. In this study, we investigated the presence of four LPxTG genes, and the LP3 gene was detected in 22.45% of the isolates; however, the LP1, LP2, and LP4 genes were not detected in any isolates. CGC was detected from fish-derived L. garvieae, and four different primers were used to detect CGC [37]; however, none of the isolates from CM cases had all four target stripes, which suggests the similarity between fish and bovines is not high. Seven genes (epsRXABCD) were conserved in the exopolysaccharide (EPS) biosynthetic gene cluster of 49 L. garvieae isolates. Similarly, these data suggest that capsule gene clusters may have spread wildly in Lactococcus spp. as genomic islands. The seven genes also appeared to encode enzymes involved in the polysaccharide structure of the capsule. It has been known that having only a few virulence genes is not enough to cause a pathogenic state, and an appropriate combination of virulence genes must be obtained to cause disease in a particular host species. The most cytotoxic isolate had more virulent genes identified than the least cytotoxic isolate. We inferred the difference in phenotypic virulence might be related to genotypic virulence, but further studies need to be carried out to confirm this.
In vivo challenge models in mice with bovine mastitis pathogens have been successfully used to assess bacterial infection and tissue damage [43]. In the current study, L. garvieae infections stimulated the inflammatory response of the murine mammary gland, which was manifested by the general appearance of concentrated inflammatory cell infiltration, progressive mammary alveolar damage, and the concentration of bacteria in the tissue. L. garvieae multiplied rapidly, leading to the migration of inflammatory cells to the mammary tissue, resulting in edema and congestion [2]. In addition to rapid growth in vitro and a high bacterial count, L. garvieae then declined but was not cleared. Infection with L. garvieae in the murine model indicated that the organism is well adapted to proliferation in the mammary gland and to cause tissue damage. The in vitro study (MAC-T) also supported the bacteria having good adaptive ability in bovine mammary cells.

Conclusions
This was the first time the zoonotic pathogen L. garvieae was isolated in CM milk samples from large dairy farms in China (prevalence of 3.40%). All L. garvieae isolates were susceptible to penicillin, ampicillin, cephalexin, cefquinome, ceftiofur, marbofloxacin, and vancomycin. L. garvieae had high resistance to lincomycin, oxytetracycline, and rifaximin, and 12.24% of isolates were resistant to cephalexin, with 10.20% (5/49) being multidrugresistant (MDR). This suggests bacterial clearance may be decreased during the dry period after the application of dry cow antibiotic preparations and that extended therapy may result in better bacteriological cures in CM cases. The study demonstrated the adhesive ability of L. garvieae in MAC-T and how it can cause cell damage both in vitro and in vivo in the murine model of intramammary infection. The findings of this study help to explain the high prevalence, tissue-damaging nature, and antimicrobial resistance of L. garvieae as an emerging mastitis pathogen.