Anti-Staphylococcus aureus Single-Chain Fragment Variables Play a Protective Anti-Inflammatory Role In Vitro and In Vivo

Staphylococcus aureus is a causative agent of bovine mastitis, capable of causing significant economic losses to the dairy industry worldwide. This study focuses on obtaining single-chain fragment variables (scFvs) against the virulence factors of S. aureus and evaluates the protective effect of scFvs on bovine mammary epithelial (MAC-T) cells and mice mammary gland tissues infected by S. aureus. After five rounds of bio-panning, four scFvs targeting four virulence factors of S. aureus were obtained. The complementarity-determining regions (CDRs) of these scFvs exhibited significant diversities, especially CDR3 of the VH domain. In vitro, each of scFvs was capable of inhibiting S. aureus growth and reducing the damage of MAC-T cells infected by S. aureus. Preincubation of MAC-T cells with scFvs could significantly attenuate the effect of apoptosis and necrosis compared with the negative control group. In vivo, the qPCR and ELISA results demonstrated that scFvs reduced the transcription and expression of Tumor Necrosis Factor alpha (TNF-α), interleukin-1β (IL-1β), IL-6, IL-8, and IL-18. Histopathology and myeloperoxidase (MPO) results showed that scFvs ameliorated the histopathological damages and reduced the inflammatory cells infiltration. The overall results demonstrated the positive anti-inflammatory effect of scFvs, revealing the potential role of scFvs in the prevention and treatment of S. aureus infections.


Introduction
Staphylococcus aureus refers to a natural inhabitant of mammalian skin and mucous epithelia bacterium, acting as the leading causative agent of various diseases in man and domestic animals [1]. The organism accounts for countless diseases, from the skin and soft tissue infections to more invasive diseases (e.g., necrotizing pneumonia and sepsis, especially bovine mastitis) [2]. S. aureus is largely involved in intramammary infections (IMI) of lactating cow, which causes significant economic losses to the dairy industry worldwide [3]. After invading the udder via the nipple duct, S. aureus can recognize mammary epithelial and inflammatory cells by secreting a series of related toxins and adhesion factors, destroying the integrity of the cell membrane through the joint action of other virulence factors; then, the microbes invade into the cells, causing damage of mammary gland tissues and inflammation [4,5]. The following four proteins (fibronectin binding protein A (FnBPA), coagulase (Coa), β-hemolysin (Hlb), glyceraldehyde-3-phosphate dehydrogenase (GapC)), as important virulence factors, play a vital role in the pathogenicity epizootic diarrhea. Wang et al. reported that scFvs targeting the whole-cell antigen could effectively inhibit the growth of S. aureus in vitro and weaken the damage of mammary gland tissues infected by S. aureus in vivo [19]. Zhang et al. also reported that scFvs targeting PEDV could neutralize PEDV infection and provide protective efficiency in piglets against PEDV challenge [29].
Proinflammatory cytokine is one of the important indicators of inflammatory response and plays a key role in host defense against invasive pathogenic microorganisms [30]. TNF-α is an "early" multifunctional proinflammatory cytokine that can induce the cascade release of other inflammatory cytokines, such as IL-6 [31]. As a multifunctional cytokine, IL-6 plays a role in regulating inflammation and immune response. As indicated by the data, bovine mastitis has a higher level of IL-6 in milk [32]. IL-1β can regulate S. aureusinduced host immune response and recruit inflammatory cells, such as neutrophils, into the lesion area, thus causing a series of inflammatory response [33]. IL-8 and IL-18 are the proinflammatory molecules mediated by NFκB signal pathway in an inflammatory microenvironment. Previous studies showed that IL-8 and IL-18 contributed to neutrophil chemotaxis, which plays an important role in antibacterial immune response [34][35][36]. Inhibiting the production of TNF-α, IL-18, IL-6, IL-1β, and IL-8 is helpful to prevent inflammation. Appropriate levels of proinflammatory cytokines are important for the immune response against pathogens, but excessive production of cytokines may lead to severe tissue damage [37].
In the present study, the binding characteristics of scFvs against virulence factors related to inflammation were evaluated, and the bacterial growth inhibition assay was performed with scFvs. The protective properties of scFvs on MAC-T cells and mice mammary gland tissues infected by S. aureus were further confirmed. These results provide basis for the development of scFv-based therapeutic agents, and the mentioned novel bovine scFvs may be suitable candidates to prevent and treat S. aureus-induced bovine mastitis.

Bacterial Strains and Cell Culture
Five S. aureus strains applied here were termed as USA300 (human-associated S. aureus strain kindly provided by professor Jianhe Sun, Shanghai Jiao Tong University, China), ATCC25923 and ATCC29213 (standard S. aureus strains preserved in our lab), XD69 and XD79 (cow-associated S. aureus strains isolated from Shanghai Xi Di dairy farm, China). S. aureus strains were cultured aerobically in triplicate for different hours in Tryptic Soy Broth (TSB) with shaking at 37 • C depending on experiment needs. The supernatants of S. aureus were produced through centrifugation, sterilized with 0.22 µm filters, and then stored at −20 • C until use. MAC-T cells (kindly provided by professor Jianhe Sun, Shanghai Jiao Tong University, China) were maintained at 37 • C with 5% CO 2 in RPMI 1640 supplemented with 10% fetal bovine serum (FBS), no penicillin and streptomycin were added during incubation with S. aureus unless stated otherwise.

Preparation of S. aureus Virulence Factors
The genomic DNA of S. aureus ATCC25923 strain were extracted with boiling method and served as template to amplify four genes (coa, hlb, gapC and fnbpA). The primers used to amplify the gene sequence of four virulence factors was designed by our lab ( Table 1). The underlined nucleic acid sequence is the restriction site that binds to the specific restriction enzyme. The PCR products of the four genes were subcloned into the pET-32a (+) plasmid via double enzyme digestion. Subsequently, the recombinant plasmid was transformed into BL21 (DE3) competent Escherichia coli cells (Invitrogen, Carlsbad, CA, USA) [38]. Four hundred milliliters of LB cultures were incubated at 37 • C, until the cultures reached an OD 600 of 0.6. Next, the cultures were induced with 0.6 mM IPTG (Sigma-Aldrich, St. Louis, MO, USA) at 16 • C, with shaking for 16 h. After centrifugation at 6000× g for 20 min, the bacterial pellets were resuspended with phosphate buffered saline (PBS) and then sonicated on ice for 40 min. Afterwards, cell lysates were clarified Vaccines 2021, 9, 1300 4 of 21 through centrifugation at 12,000 rpm for 25 min [39]. The soluble proteins with N-terminal 6× His tag were purified with His-Bind Resin (Merk-Novagen, Darmstadt, Germany). The proteins were eluted with 50, 100, 150, 200, 250, 300 mM imidazole contained successively. The protein concentrations were measured by the BCA Protein Assay Kit (Sangon Biotech, Shanghai, China) [19]. The proteins were stored at -80 • C until use.

Optimization and Bio-Panning of the ScFv Phage Display Library
The scFv phage display library was already constructed by Wang et al. in our lab [19]. To obtain a phage library with higher positive rate, we optimized the scFv phage library. Firstly, four virulence factor proteins of S. aureus were mixed and coated on a 96-well plate. After five rounds of enrichment and screening of the phage library, a secondary library with higher positive rate was generated. Subsequently, we used the single virulence factor as the coated antigen. The process of the enrichment and the screening can be referenced from previous research conducted in our lab [19,29].

Affinity Selection against Virulence Factors
Phage ELISA was adopted to assess the binding affinity of phages with virulence factor antigens. The bacterial clones were randomly selected for the virulence factor antigens binding analysis after five rounds of panning. Each clone infected by the M13KO7 helper phage was precipitated as described in the previous section. During the screening process of phage ELISA, M13KO7 helper phage was used as the negative control. The description of the procedures performed next can be referenced from previous studies [19,43].

Expression and Purification of ScFvs
To verify the interaction of antigens and scFv in vivo and in vitro experiment, the gene fragment of four scFvs, namely ZL4, ZL15, ZL26, and ZL125 (targeting Hlb, Coa, GapC and FnBPA-A, respectively), were subcloned into pGEX-4T-1 vector carrying N-GST tag. The primers and restriction sites of the four scFvs were identical as below: forward primer named VL-EcoRI-F (5 -CCGGAATTCATGGCCCAGGCTGTGCTGACTCAG) and reverse primer named VH-XhoI-R (5 -CCGCTCGAGACTAGTGGAGGAGACGGTGAC). The template of the constructed scFv was amplified with the phagemid DNA of positive clones. The subsequent expression and purification of four scFv proteins can be referenced from previous studies [44]. Four soluble scFv proteins with N-GST tag were purified with Vaccines 2021, 9, 1300 5 of 21 the GST-Sefinose ™ Kit (Sangon Biotech, Shanghai, China). Afterwards, the proteins were stored at −80 • C until use.

Adhesion and Invasion of S. aureus on MAC-T Cells
RPMI-1640 medium supplied with 10% FBS was used to culture MAC-T cells at 37 • C in 5% CO 2 . Specific to the adhesion assay, 1 × 10 5 MAC-T cells per well were incubated with S. aureus or plus with scFvs (final scFv or mixture concentration was 40 µg/mL in the medium) at a multiplicity of infection (MOI) of 10. After incubation for 3 h, the plates were washed with PBS three times to eliminate bacterial not adhered to MAC-T cells. Next, the cells in every well were lysed with PBS containing 0.1% Triton X-100 for 5 min. The adhered S. aureus through the serial dilutions were incubated on TSA plates at 37 • C overnight. Specific to the invasion assay, 1 × 10 5 MAC-T cells per well were incubated with S. aureus or plus with scFvs (final scFv or mixture concentration was 40 µg/mL in the medium) at a MOI of 10. After the cell incubation for 3 h and washing with PBS three times, the medium was substituted with RPMI 1640 medium supplied with 1% FBS and 50 µg/mL of gentamicin (aiming to kill the extracellular S. aureus), and then incubated at 37 • C for 1 h. Afterwards, the intracellular invading S. aureus were released from the MAC-T cells lysed with PBS containing 0.1% Triton X-100 at 37 • C for 5 min. The adhesion/invasion rate was calculated as (number of CFU of the adhered or invading bacteria/ number of CFU of the initial number of bacteria) × 100% [45].

Bacterial Growth Inhibition Assay
To demonstrate the effect of scFvs aiming at various virulence factors, scFv ZW88 obtained in the previous study was used as the positive control in the bacterial growth inhibition assay. The specific steps of the inhibition assay are as follows. 50 µL S. aureus diluted to 10 6 cfu/mL with TSB were added to a 96-well plate, combined with 50 µL purified scFv (or mixture of four scFvs diluted with TSB), and then added with 100 µL of TSB (final scFv or mixture concentration was 40 µg/mL) with shaking at 37 • C. The mixture of 150 µL of TSB and 50 µL 10 6 cfu/mL S. aureus was set as the negative control. Moreover, the mixture of 150 µL of TSB and 50 µL 10 6 cfu/mL S. aureus containing 40 µg/mL penicillin were set as the positive control. Next, an ELISA plate reader was used to measure the OD 600 at 6-h intervals [19].

Assessment of Apoptosis by Immunofluorescence Assay (IFA)
For the apoptosis assay, YO-PRO-1 reagent can be used as a sensitive indicator of the apoptosis [46]. Continuously increasing green fluorescence alone indicates apoptotic cells, while the simultaneous red and green fluorescence indicates dead cells. Here, we use YO-PRO TM -1/ Propidium Iodide (PI) (Vybrant Apoptosis Assay kit 4; Thermo Fisher, Waltham, MA, USA) to assess the apoptosis of S. aureus-induced MAC-T cells. In brief, 1 × 10 5 MAC-T cells per well was seeded in 96-well culture plates, then washed three times and infected by 1 × 10 6 cfu/mL of S. aureus (MOI = 10) or plus with scFvs (final scFv or mixture concentration was 40 µg/mL) at a range of time points. Next, 100 µL of standard staining working solution was added to the respective well in darkness at 37 • C for 20 min. After being dyed, the stained MAC-T cells were visualized under the epi-fluorescence microscope (EVOS FL-2, Life technologies, New York, NY, USA) [47].

LDH Cytotoxicity Assay
Since the LDH concentration in cell culture media is an indicator of cellular cytotoxicity, the assay can be adopted to monitor cytotoxicity from the identical sample. The level of LDH released in the cell culture supernatant was determined with the LDH Assay Kit (Abcam, Cambridge, UK) in line with the manufacturer's instructions. To perform the assay, 1 × 10 5 cells per well were seeded in 96-well culture plates. After being washed with PBS three times, the cells were then infected by 1 × 10 6 cfu/mL of S. aureus (diluted with 1% FBS-RPMI 1640 medium) or plus with scFvs (final scFv or mixture concentration was 40 µg/mL) Vaccines 2021, 9, 1300 6 of 21 over time. The negative controls consisted of 200 µL of 1% FBS-RPMI 1640 medium. After the predetermined time points, the 96-well culture plate was centrifuged at 400× g for 5-10 min. Subsequently, 120 µL aliquot of the cell culture media was transferred to a novel plate, then, 60 µL detection kit reaction mixture was added to the plate. After a 30-min incubation at the ambient temperature without light, the absorbance was measured with a microplate reader at 490 nm. The cytotoxicity rate was calculated in accordance with the specification [48].

Cell Viability Assay
Cell viability assays were performed by using the CellTiter-Glo ® Luminescent Cell Viability Assay (Promega, Madison, WI, USA) according to the manufacturer's instruction. 1 × 10 5 MAC-T cells per well was seeded in 96-well culture plates. Subsequently, the supernatant of different S. aureus strains or plus with scFvs (final scFv or mixture concentration was 40 µg/mL) were added to the plates at various time points when MAC-T cells grew to about 90%. RPMI-1640 medium was added to the plate as the negative control. After the MAC-T cells were lysed by CellTiter Glo reagent in 10-30 min incubation at ambient temperature, the relative cell viability was determined by luminescence assay based on adenosine triphosphate quantification (CellTiter-Glo). Details of the specific operation steps and calculation methods are in accordance with the specification [49].

Animal Experiment
In the present study, 132 adult female BALB/c mice (6-8 weeks old) that had delivered their first litter of pups 10-15 days in advance were randomly allocated into the control group (C), negative control group (NC), positive control group (PC) and scFv treatment groups (including ZL4, ZL15, ZL26, ZL125, and scFvs mixture group) for intramammary bacterial challenge. Each group was infected by three S. aureus trains (USA300, ATCC25923 and XD69). The mastitis model was established as described previously [19]. First, 100 µL of physiological saline were injected subcutaneously into the teats of the R4 and L4 of the mammary gland tissues in C and NC groups. The PC group had 100 µL of 100 mg/mL penicillin injected, while the scFv group injected single scFv or the mixture of four scFvs (20 mg/Kg). Six hours later, 100 µL of physiological saline was injected into the teats of the C group. Meanwhile, the NC, PC and scFv treatment groups had 100 µL of 10 8 cfu/mL S. aureus (USA300, ATCC25923, and XD69) injected. After sodium pentobarbital anesthetized (40 mg/kg), all mice were euthanized with CO 2 inhalation at 48 h post-inoculation. Subsequently, the blood samples and mammary gland tissues were collected for analysis of the following experimental indicators.

Bacteria Count of Mammary Gland Tissues
Bacterial burden assay was performed to evaluate the inflammatory index. One hundred mg aliquot of mammary gland tissues were homogenized in PBS and then plated onto tryptic soy agar (TSA) plates to enumerate cfu through serial dilutions. Bacterial burden assay was performed with groups of six mice [50].

Histopathological Evaluation of Mammary Gland Tissues
Severe infiltration of neutrophils and macrophages occurs in most fields and areas of mammary gland tissues caused by S. aureus. Histopathological evaluation of mammary gland tissues reflects the treatment effect of the scFvs. The mammary gland tissues were excised and fixed with 4% paraformaldehyde, dehydrated, and embedded with paraffin wax. Afterwards, 4 µm thick slices were stained with hematoxylin and eosin (H & E) and observed under an optical microscope (Olympus, Tokyo, Japan) [51,52]. The mammary gland tissues alternation index was determined as described previously [53,54], such as the damage level of mammary acini and mammary epithelial cells, the integrity of lobuli mammae, the inflammatory cells infiltration and the thickening of alveolus walls. Each histological characteristic was graded 0 to 5.

MPO Activity Assay
Myeloperoxidase (MPO) exists in polymorphonuclear neutrophils (PMNs, the main components of inflammation and immune response in bovine mastitis). The inflammatory response severity of the parenchymal infiltration of neutrophils and macrophages in mammary gland tissues was assessed by MPO activity. One hundred mg aliquot of mammary gland tissues were homogenized with reaction buffer (volume ratio 1:9) by homogenizer (GentleMACS Dissociator, Miltenyi Biotec GmbH, Bergisch-Gladbach, Germany). The operating steps were in accordance with the manufacturer's instructions (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). The value of OD 460 measured with a 96-well microplate reader was used to evaluate the MPO activity [41,42].

ELISA Assay of Cytokine Expression Levels
In this study, the secretion level of five inflammatory cytokines (TNF-α, IL-18, IL-6, IL-1β, IL-8) was used to evaluate the inflammatory response of mammary gland tissues. The different groups of mammary gland tissues were collected and homogenized with PBS on ice, and then the supernatants were collected by centrifugation. The levels of the five inflammatory cytokines were determined by ELISA kits (Multisciences, Hangzhou, China) according to the manufacturer's instructions [53,55].

qPCR Assay of Cytokine Transcription Levels
The mammary gland tissues were homogenized in PBS, frozen and thawed three times; the supernatants were obtained after centrifugation at 9000× g for 20 min. The total RNA was then extracted from the supernatants of the mammary gland tissues by the TRIzol reagent, and the cDNA was generated by using a reverse transcription kit (Invitrogen, CA, USA) according to the manufacturer's instructions. The specific primers of five cytokines were listed in Table 1 [35,[40][41][42]. Quantitative real-time PCR was performed on an ABI 7300 Real-Time PCR Detection System (Applied Biosystems, Foster, CA, USA) in a 20-µL reaction volume. Each sample was assessed in triplicate. The reaction volume and the reaction condition can be referenced from our lab's previous studies [29]. The results (fold changes) were normalized by GAPDH using the 2 −∆∆Ct comparative method [41].

Statistical Analysis
All tests of the statistical analysis were performed on GraphPad Prism 6 software (San Diego, CA, USA) and SPSS 18.0 software (SPSS, Inc., Chicago, IL, USA). The data were expressed as mean ± standard error. Moreover, the intergroup differences were assessed by multiple comparisons of variance (one-way analysis of variance). The differences were considered statistically significant when p < 0.05.

Expression and Purification of Virulence Factors
Four genes ( Figure 1A-D) were subcloned into pET-32a (+) plasmid and then expressed on BL21 (DE3) competent E. coli cells. The length of four DNA fragments were about 1000 bp (hlb and gapC), 1600 bp (fnbpA-A), and 2000 bp (coa). Four protein samples were separated by 10% SDS-PAGE and then transferred onto nitrocellulose. The western blotting assay displayed a distinct single band after dyeing, suggesting that the purity of the four antigen proteins was high enough to apply in the following experiment ( Figure 1E-H). Four genes ( Figure 1A-D) were subcloned into pET-32a (+) plasmid and then expressed on BL21 (DE3) competent E. coli cells. The length of four DNA fragments were about 1000 bp (hlb and gapC), 1600 bp (fnbpA-A), and 2000 bp (coa). Four protein samples were separated by 10% SDS-PAGE and then transferred onto nitrocellulose. The western blotting assay displayed a distinct single band after dyeing, suggesting that the purity of the four antigen proteins was high enough to apply in the following experiment ( Figure  1E-H).

The Characteristics and the Screening of ScFv Phage Library
After PCR and overlap PCR, DNA sequences encoding the VL and VH regions were confirmed to be approximately 350 and 380 bp in length, respectively, and the full length of scFv product was around 760-790 bp ( Figure 1I,K), including the VL, VH, and (Gly4Ser)3 linker. The phage display scFv library size was determined as 3.1 × 10 6 cfu/mL. The positive rate of the library reached over 90% (data not shown). These results revealed the scFv phage library is promising for scFv selection.
After five rounds of bio-panning, the final output titer of the scFv phage display library reached 1.4 × 10 8 cfu/mL, which was 150-fold higher than the 1st round ( Table 2). The positive clone was defined as 2.5-times as much as the OD450 of the negative control. Four unique clones targeting the four virulence factors showed a comparatively higher affinity than other positive clones ( Figure 1J). These scFvs were designated as ZL4 (accession no. MZ392847), ZL15 (accession no. MZ392848), ZL26(accession no. MZ392849) and ZL125 (accession no. MZ392850) (aiming at Hlb, Coa, GapC and FnBPA-A, respectively).

The Characteristics and the Screening of ScFv Phage Library
After PCR and overlap PCR, DNA sequences encoding the VL and VH regions were confirmed to be approximately 350 and 380 bp in length, respectively, and the full length of scFv product was around 760-790 bp ( Figure 1I,K), including the VL, VH, and (Gly 4 Ser) 3 linker. The phage display scFv library size was determined as 3.1 × 10 6 cfu/mL. The positive rate of the library reached over 90% (data not shown). These results revealed the scFv phage library is promising for scFv selection.
After five rounds of bio-panning, the final output titer of the scFv phage display library reached 1.4 × 10 8 cfu/mL, which was 150-fold higher than the 1st round ( Table 2). The positive clone was defined as 2.5-times as much as the OD 450 of the negative control. Four unique clones targeting the four virulence factors showed a comparatively higher affinity than other positive clones ( Figure 1J). These scFvs were designated as ZL4 (accession no. MZ392847), ZL15 (accession no. MZ392848), ZL26(accession no. MZ392849) and ZL125 (accession no. MZ392850) (aiming at Hlb, Coa, GapC and FnBPA-A, respectively).

Identification and Expression of ScFv
Four scFvs carrying N-GST tag were constructed to distinguish the relevant antigens carrying N-6 × His tag. The lengths of the four scFv genes were around 760-790 bp ( Figure 1K). All scFvs protein samples were run on 10% SDS-PAGE gels and transferred to nitrocellulose at 150 mA for 1.5 h. As revealed from the SDS-PAGE assay and the Western blots assay, four scFv preteins were approximately 54 kDa (contained 26 kDa of GST tag) in size and displayed the single band, revealing that the purity of the four scFv proteins is satisfactory for the following experiments ( Figure 1L).

Characteristics of Purified ScFvs
The amino acid sequences of the four scFvs are presented in Figure 2. The VL and VH amino acid sequences of the four scFvs were successively joined with a (Gly 4 -Ser) 3 peptide linker. Four frame regions (FR) and three complementary determination regions (CDR) constitute VH domain and VL domain, respectively. The FRs of scFv were highly conserved according to amino acid sequencing results. The CDRs of scFv demonstrated a significant diversity, especially CDR3 of VH.

ScFvs Can Effectively Inhibit the Bacteria Proliferation In Vitro and In Vivo
The growth of S. aureus USA300, ATCC25923, ATCC29213, XD69, and XD79 in TSB were inhibited to some extent by the single scFv (ZL4, ZL15, ZL26, ZL125 and ZW88) or the mixture of four scFvs compared with NC in 24 h (Figure 4A-E). Furthermore, the inhibiting effect of scFvs mixture group was better than that of scFv ZW88 group in five

ScFvs Can Affect the Adhesion and Invasion of S. aureus on MAC-T Cells
Three S. aureus strains USA300, ATCC25923, and XD69 were performed in the adhesion and invasion assay ( Figure 3A,B). Specific to the adhesion assay, the adhesion rate of MAC-T cells infected by USA300 and XD69 strains was higher (60%) than that by ATCC25923 strain (44%). After the addition of the mixture of four scFvs, the adhesion rate exhibited a sharp decrease (p < 0.001). In the invasion experiment, the invasion rate of USA300, ATCC25923, and XD69 were nearly 39%, 26%, and 42%, respectively. After adding the mixture of four scFvs, the invasion rates were decreased to 21%, 14%, and 22% individually. As suggested from the data, the adhesion rate and invasion rate of MAC-T cells infected by three S. aureus strains decreased significantly after adding ZL15, ZL26, ZL125 or the mixture of four scFvs compared with the NC group (S. aureus group) (p < 0.05), while ZL4 exhibited no effect in adhesion and invasion assay (p > 0.05). In addition, no significant difference was found in adhesion and invasion assay between the NC group and the blank control group (S. aureus + GST group) (p > 0.05).

ScFvs Can Effectively Inhibit the Bacteria Proliferation In Vitro and In Vivo
The growth of S. aureus USA300, ATCC25923, ATCC29213, XD69, and XD79 in TSB were inhibited to some extent by the single scFv (ZL4, ZL15, ZL26, ZL125 and ZW88) or the mixture of four scFvs compared with NC in 24 h ( Figure 4A-E). Furthermore, the inhibiting effect of scFvs mixture group was better than that of scFv ZW88 group in five strains. Nevertheless, the supernatant of five S. aureus strains incubated with the mixture of four scFvs and scFv ZW88 became turbid in TSB in 18 h, the OD 600 (about 0.25) of which were less than the NC group (about 1.12). To determine the specific contribution of scFvs to the inflammatory response, we monitored colonization and bacterial replication in mammary gland tissues systemically infected with S. aureus. The results showed that the bacterial burden of USA300 and XD69 were higher than that of ATCC25923 in mammary gland tissues. In addition, the effect of the mixture of four scFvs exhibited a 100-fold decrease in bacterial burden compared with the NC group (p < 0.05). Furthermore, the scFvs mixture group significantly inhibited the bacteria proliferation compared with the single scFv group at 48 h (p < 0.05), which exhibited a 10-100-fold decrease in bacterial burden ( Figure 4F). The results showed that scFvs could effectively inhibit the proliferation of S. aureus in vitro and in vivo.

ScFvs Can Attenuate the Apoptosis and Necrosis of S. aureus on MAC-T Cells
The apoptosis assay was performed for MAC-T cells infected by USA300, XD69, and ATCC25923, respectively ( Figure 5A-C). As revealed from the results, S. aureus could cause apoptosis and necrosis of MAC-T cells. After adding the mixture of four scFvs, the number of apoptotic and necrotic MAC-T cells infected by three S. aureus strains decreased significantly from 1300 to 750 (USA300 group) (p < 0.01), from 1240 to 640 (ATCC25923) (p < 0.01), and from 1590 to 860 (XD69 group) (p < 0.01) ( Figure 5D). S. aureus strains were cultured in TSB containing physiological saline, scFvs or penicillin (final concentration of both scFv/mixture and penicillin were 40 μg/mL). Data represent mean results ± SD (n = 3 or 6). # p < 0.05 vs. scFvs mixture group. * p < 0.05 vs. negative control group.

ScFvs Can Weaken the Cytotoxicity of S. aureus on MAC-T Cells
The LDH cytotoxicity assay detection kit was adopted to detect the level of LDH released in the cell culture supernatant. Three S. aureus strains (i.e., USA300, ATCC25923, and XD69) were selected to perform the assay. The results revealed that the level of LDH concentrations in the supernatant were continually elevated over time. After the infections for 6 h, the LDH release of MAC-T cells infected by strain USA300 (from 50% to 34%) (p < 0.01), ATCC25923 (from 40% to 25%) (p < 0.01), and XD69 (from 55% to 37%) (p < 0.001) decreased significantly after incubation with the mixture of four scFvs ( Figure 6A-C). It is worth noting that the mixed scFvs had a dose-dependent protective effect on MAC-T cells ( Figure 6D). Meanwhile, the relative cell viability assay was conducted by detecting the level of ATP release of MAC-T cells infected by the supernatant of USA300, ATCC25923, ATCC29213, XD69, and XD79, showing that the percentage of the relative cell viability of MAC-T cells infected by the supernatant of XD69 and USA300 strains was lower than that of the cells infected by the supernatant of other three strains. The differences between the supernatant of five S. aureus and plus with the mixture of scFvs for infections to MAC-T cells showed the statistical significance (p < 0.05) ( Figure 6E,F). The downtrend of the relative cell viability of MAC-T cells incubated with the mixture of scFvs and the culture supernatants were distinct to varying degrees by the change over time.

Effect of ScFvs on S. aureus-Induced Histopathologic Changes and MPO Activity of Mammary Gland Tissues
The pathological section of mammary gland tissues was detected by H & E staining and microscopic observation. The results showed no histopathological changes in the C group ( Figure 7A). The mammary gland tissue structures of the NC group treated with three S. aureus strains were destroyed, showing severe histopathologic changes such as damaged mammary acini and mammary epithelial cells, incomplete lobuli mammae, and inflammatory cells infiltration ( Figure 7B). In contrast, these histological changes of the scFv treatment groups were significantly ameliorated (p < 0.05), particularly the scFvs mixture group, the effect of which was close to the PC group exhibiting very slight inflammatory injury and mild inflammatory cells interstitial infiltration ( Figure 7C-H). The mammary gland tissues alternation index was used to evaluate the histopathological changes in the mammary gland tissues, such as the integrity of mammary tissues and the numbers of infiltrated inflammatory cells ( Figure 7I). Besides, the MPO activity of five scFv treatment groups, including single scFv group and scFvs mixture group, was significantly reduced compared with that of the NC group (p < 0.05) ( Figure 7J).

Effect of ScFvs on Inflammatory Cytokine Production in Mammary Gland Tissues
To investigate the effects of scFvs on the S. aureus-induced inflammatory cytokines production, the levels of TNF-α, IL-18, IL-6, IL-1β, and IL-8 were measured by qPCR and ELISA assays. The results indicated that every single scFv could dramatically reduce the expression levels of five inflammatory cytokines (p < 0.05). Compared with that of the NC group, the expression levels of TNF-α (from 480 pg/mL to 190 pg/mL), IL-18 (from 880 pg/mL to 350 pg/mL), IL-6 (from 520 pg/mL to 210 pg/mL), IL-1β (from 770 pg/mL to 300 pg/mL), and IL-8 (from 460 pg/mL to 190 pg/mL) in the scFvs mixture treatment group approximately exhibited a 2.5-fold decrease (p < 0.05) ( Figure 8B). Besides, the transcription levels of the five inflammatory cytokines in mammary gland tissues infected by three S. aureus strains were significantly suppressed by the scFvs mixture group compared with the NC group (p < 0.05), exhibiting a 2.5-fold decrease (the fold changes decrease approximately from 6.74 to 2.92 in TNF-α group, from 12.02 to 2.48 in IL-18 group, from 7.22 to 3.16 in IL-6 group, from 8.52 to 2.85 in IL-1β group, and from 8.03 to 2.87 in IL-8 group). Moreover, Almost all of the scFvs mixture groups (except for the expression of TNF-α in ATCC25923 group and IL-8 in XD69 group) showed remarkable inhibitory effects compared with the single scFv group (p < 0.05) ( Figure 8A,B).

ScFvs Can Weaken the Cytotoxicity of S. aureus on MAC-T Cells
The LDH cytotoxicity assay detection kit was adopted to detect the level of LDH released in the cell culture supernatant. Three S. aureus strains (i.e., USA300, ATCC25923, and XD69) were selected to perform the assay. The results revealed that the level of LDH concentrations in the supernatant were continually elevated over time. After the infections for 6 h, the LDH release of MAC-T cells infected by strain USA300 (from 50% to 34%) (p < 0.01), ATCC25923 (from 40% to 25%) (p < 0.01), and XD69 (from 55% to 37%) (p < 0.001) decreased significantly after incubation with the mixture of four scFvs ( Figure 6A-C). It is worth noting that the mixed scFvs had a dose-dependent protective effect on MAC-T cells ( Figure 6D). Meanwhile, the relative cell viability assay was conducted by detecting the level of ATP release of MAC-T cells infected by the supernatant of USA300, ATCC25923, ATCC29213, XD69, and XD79, showing that the percentage of the relative cell viability of MAC-T cells infected by the supernatant of XD69 and USA300 strains was lower than that of the cells infected by the supernatant of other three strains. The differences between the supernatant of five S. aureus and plus with the mixture of scFvs for infections to MAC-T cells showed the statistical significance (p < 0.05) ( Figure 6E,F). The downtrend of the relative cell viability of MAC-T cells incubated with the mixture of scFvs and the culture supernatants were distinct to varying degrees by the change over time.

Discussion
Bovine mastitis is generally known as one of the most prevalent diseases affecting the dairy industry and causing considerable economic losses. S. aureus is one of the extremely important causative organisms that can cause bovine mastitis [56]. At present, the

Discussion
Bovine mastitis is generally known as one of the most prevalent diseases affecting the dairy industry and causing considerable economic losses. S. aureus is one of the extremely important causative organisms that can cause bovine mastitis [56]. At present, the enhancement of drug resistance and antibiotic residues have increasingly become important factors endangering public health, despite certain therapeutic effects in the treatment of bovine mastitis. The existing studies demonstrated that the killing effect of S. aureus on cells and mice varies significantly with the source of the strains [57]. Previous studies revealed that specific S. aureus genotypes (i.e., USA300) showed an increased cow-tohuman cross-infection [58]. Therefore, S. aureus (five strains) or their culture supernatants were studied to verify the pathogenicity of S. aureus from different sources. In brief, four scFvs exhibiting a strong interaction with four virulence factors were obtained by using the phage display technique. The overall results showed that USA300 and XD69 had stronger pathogenicity compared with ATCC25923, ATCC25923, and XD79, and the protective effects of the four scFvs on MAC-T cells and mice mammary gland tissues infected by S. aureus was confirmed.
The positive rate of the phage library was improved by optimizing the screening condition, such as reducing the concentration of coated antigen from 10 µg/mL in the first round to 5 µg/mL in subsequent rounds, and increasing the washing times from 10 to 20 times in the subsequent rounds with PBST and PBS. The results revealed that the output/input of the scFv phage display library in the 5th round (1.33 × 10 −5 ) is 150 folds higher than that in the 1st round (8.93 × 10 −8 ). Previously, Wang et al. reported that the final output titer of the scFv phage display library reached 2.32 × 10 7 cfu/mL, which is 6 folds lower than in this study (1.4 × 10 8 cfu/mL) [19,29].
As indicated from the comparison of amino acid sequence, Framework regions (FRs) are highly conserved, while the CDRs of antibody binding sites and the specific determination are highly variable, suggesting that scFvs obtained in this study may well have various epitope specificities. For the composition and the numbers of amino acids, the largest variation was in the CDR3 of VH, suggesting that in the process of binding to antigen, the CDR3 may be the most important antigen recognition site. Future studies should focus on the spatial structure and binding mode of scFv to explore the antigen-antibody interaction mechanism. Also, it will be interesting to study whether the region interacting with the corresponding virulence factors of S. aureus belongs to the CDR regions of scFv, especially the CDR3 region of VH.
Compared with that of the NC group, the adhesion rate and invasion rate of USA300, ATCC25923 and XD69 strains decreased in different degrees after incubation with ZL26, ZL125, and the mixture of four scFvs, respectively. The adhesion rate and invasion rate of ZL26 and ZL125 treatment groups decreased significantly compared with that of the NC group (p < 0.01), illustrating that the interaction between scFvs (ZL125 and ZL26) and the adhesins (i.e., FnBPA and GapC) secreted by S. aureus may be responsible for the weakening of the adhesion of S. aureus to MAC-T cells, thus effecting the invasion. Previous studies demonstrated that FnBPA and GapC contributed to the adhesion and invasion of S. aureus to the host cells, complying with the results in this study [6,11]. The adhesion rate of USA300 (from 56% to 51% in ZL15 group (p < 0.05), and from 56% to 42% in ZL125 group (p < 0.01)), ATCC25923 (from 44% to 39% in ZL15 group (p < 0.05), and from 44% to 30% in ZL125 group (p < 0.001)) and XD69 (from 58% to 54% in ZL15 group (p < 0.05) and from 58% to 45% in ZL125 group (p < 0.001)) decreased in varying degrees after the interaction with ZL15 (against Coa) and ZL125 (against FnBPA). As reported, Coa could combine with FnBPA to form fibrin agglutination on the staphylococci surface and promote bacterial adhesion [8]. Moreover, scFvs mixture group exhibited a stronger inhibiting effect in adhesion and invasion compared with that of single scFv treatment group. The overall results suggested that scFvs may interact with Coa, FnBPA and GapC, thus blocking the formation of fibrin agglutination and decreasing the adhesive attraction.
As suggested from the apoptosis assay, S. aureus could cause apoptosis and necrosis of MAC-T cells, complying with the results of the existing studies [59]. The number of apoptotic cells and necrotic cells in scFv treatment groups decreased approximately by half (from 1300 to 750 in USA300 group (p < 0.01), from 1240 to 640 in ATCC25923 group (p < 0.01) and from 1590 to 860 in XD69 group (p < 0.01), respectively) compared with that of the NC group. Previous studies demonstrated that Panton-Valentine leucocidin and enterotoxin H of S. aureus could induce apoptosis of bovine mammary epithelial cells [60,61].
Additionally, Coa, FnBPA, and GapC can affect the adhesion and invasion of S. aureus to the host cells [6,8,11], and Hlb can destroy the integrity of cell membrane [9,10]. Therefore, it is speculated that the four virulence factors possibly worked together to destroy the cell membrane, thus leading to apoptosis and necrosis of MAC-T cells. More mechanisms of apoptosis resulted by S. aureus need to be investigated in the following research.
LDH acts as a vital index of cell membrane integrity. The concentration of LDH in the culture supernatant of MAC-T cells infected by USA300 (p < 0.01), ATCC25923 (p < 0.01) and XD69 (p < 0.001) in scFv treatment groups decreased significantly compared with the NC group (p < 0.01). Wang et al. reported that S. aureus could lead to nearly 80% of LDH release on A539 cells, the results of which were similar with the present studies [62]. Besides, the mixed scFvs had a dose-dependent protective effect on MAC-T cells. Furthermore, the relative cell viability assay showed that USA300 and XD69 exhibited stronger pathogenicity to MAC-T cells compared with that of other S. aureus strains in vitro. The relative cell viability assay also presented similar results with cytotoxicity assay [63]. These studies indicated that the four virulence factors of S. aureus can destroy the cell membrane [6,8,9,11]. Moreover, relevant scFvs can inhibit the cytotoxic effects of S. aureus on MAC-T cells, thus possibly providing a potential protective effect on dairy cow mastitis caused by S. aureus.
A mouse model of S. aureus-infected mastitis was established to evaluate the effects of scFvs on the inflammatory response in mammary gland tissues. The results of histopathological changes in the NC group revealed that the mammary acini and mammary epithelial cells were damaged, the lobuli mammae was incomplete, and the inflammatory cells were infiltrated in mammary gland tissues. These results were accordant to the previous studies [53,54]. Comparatively, these histological changes were significantly ameliorated and the MPO activity was significantly reduced in the scFv treatment groups (p < 0.05). Moreover, the effect of scFvs mixture group on the decrease of MPO activity was better than that of the single scFv group, indicating that the scFvs probably produce potential anti-inflammatory effects, especially the scFvs mixture.
We detected the production of five proinflammatory cytokines to verify the key role in host defense against invasive pathogenic micro-organisms. The results showed that compared with the C group, the transcription level and expression level of TNF-α, IL-18, IL-6, IL-1β and IL-8 increased significantly in mammary gland tissues infected by S. aureus, especially IL-1β and IL-8 (the expression levels reached over 700 pg/mL) (p < 0.05). Additionally, MPO activity (from about 1.21 to 0.78) assay also manifested that the neutrophils could be recruited to the inflammation area in the NC group. Similarly, previous studies demonstrated that TNF-α could induce the cascade release of other inflammatory cytokines, such as IL-6 [31]. IL-6 played a role in regulating inflammation and immune response [32]. IL-1β can regulate S. aureus-induced host immune response and recruit inflammatory cells, such as neutrophils, into the lesion area, thus causing a series of inflammatory response [33]. In the present study, TNF-α, IL-6, and IL-1β exhibited a sharp decrease in scFvs mixture group compared with the NC group (p < 0.05), indicating that scFvs have anti-inflammatory effects on mammary gland tissues. IL-8 and IL-18 contributed to neutrophil chemotaxis, which plays an important role in anti-bacterial immune response [34][35][36]. These studies indicated that suitable levels of five proinflammatory cytokines are vital for host defenses, whereas overmuch production can cause systemic inflammation and damage rather than protective effects. Here in this study, the production of five proinflammatory cytokines in three groups (USA300, ATCC25923 and XD69) exhibited no obvious difference. The results revealed that the transcription level and expression level of five proinflammatory cytokines in scFv mixture group reduced significantly compared with the NC group (p < 0.05). In addition, the anti-inflammatory effect in scFv mixture treatment group (except for the expression of TNF-α in ATCC25923 group and IL-8 in XD69 group) was more significant than that of single scFv group (p < 0.05). Generally, the overall results indicated that scFvs attenuated the damage of mammary gland tissues in vivo by regulating the transcription level and expression level of five proinflammatory cytokines, which is consistent with the study of Wang et al. [19].

Conclusions
In summary, four scFvs against different virulence factors were screened by using phage display techniques in this study. These scFvs could significantly inhibit the adhesion and invasion of S. aureus to MAC-T cells and exhibited bacteriostatic effects in vitro and in vivo. The scFvs could also reduce the apoptosis and cytotoxicity of S. aureus on MAC-T cells. Besides, as proven from a variety of experiments, scFvs have certain protection effects on mammary gland tissues of mice and MAC-T cells infected by S. aureus. In the future, in-depth investigations on the anti-inflammatory function of scFvs and the possible molecular mechanisms need to be conducted. The data presented in this study could be useful for further studies on the pathogenic mechanism of virulence factors and ideal drug targets screening, which will assist in the prevention and treatment of dairy cow mastitis attributed to S. aureus.

Informed Consent Statement: Not applicable.
Data Availability Statement: The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.