A novel peptidic inhibitor derived from Streptococcus cristatus ArcA attenuates virulence potential of Porphyromonas gingivalis

Periodontitis is a global health problem and the 6th most common infectious disease worldwide. Porphyromonas gingivalis is considered a keystone pathogen in the disease and is capable of elevating the virulence potential of the periodontal microbial community. Strategies that interfere with P. gingivalis colonization and expression of virulence factor are therefore attractive approaches for preventing and treating periodontitis. We have previously reported that an 11-mer peptide (SAPP) derived from Streptococcus cristatus arginine deiminase (ArcA) was able to repress the expression and production of several well-known P. gingivalis virulence factors including fimbrial proteins and gingipains. Herein we expand and develop these studies to ascertain the impact of this peptide on phenotypic properties of P. gingivalis related to virulence potential. We found that growth rate was not altered by exposure of P. gingivalis to SAPP, while monospecies and heterotypic biofilm formation, and invasion of oral epithelial cells were inhibited. Additionally, SAPP was able to impinge the ability of P. gingivalis to dysregulate innate immunity by repressing gingipain-associated degradation of interleukin-8 (IL8). Hence, SAPP has characteristics that could be exploited for the manipulation of P. gingivalis levels in oral communities and preventing realization of virulence potential.

loss 18 . Several reports have also documented that the combinations of P. gingivalis and other oral bacteria such as Treponema denticola, Fusobacterium nucleatum or S. gordonii leads to synergistic pathogenicity in animal models [19][20][21][22] . Therefore, inhibiting the colonization and accumulation of P. gingivalis in a polymicrobial community is an attractive strategy for disrupting the transition of a periodontally healthy community into a destructive one.
We have previously reported that arginine deiminase (ArcA) of S. cristatus can repress the expression of the FimA major fimbrial subunit protein in several P. gingivalis strains, including strains expressing fimA types I, II, and III 23,24 . In an in vitro study, we demonstrated that S. cristatus ArcA significantly inhibited biofilm formation by P. gingivalis 25 and using a mouse model, we found that S. cristatus ArcA can interfere with the colonization and pathogenesis of P. gingivalis 26 . More recently, we showed that an 11-mer peptide with the native sequence of ArcA (NIFKKNVGFKK, molecular-weight 1322.62), and theoretical pI 10.48, which we designated as SAPP (Streptococcal-derived Anti-P. gingivalis Peptide), represses the expression of several well-established P. gingivalis virulence-associated genes including fimA, mfa1, rgpA/B, and kgp 27 .
In this study, we further tested whether SAPP interferes with virulence-associated phenotypic properties of P. gingivalis. A series of functional assays was performed to assess the impact of SAPP on bacterial growth, biofilm formation, invasion, and gingipain activity, as well as ability of P. gingivalis to manipulate host immune responses. Our data show that SAPP does not affect the growth rate of P. gingivalis strains, but inhibits its ability to form monotypic and heterotypic biofilms, and to invade oral epithelial cells, which are key events of establishing P. gingivalis infection. In addition, arginine-and lysine-specific activities of gingipain were reduced in P. gingivalis cells and its growth media in the presence of SAPP. Our findings demonstrate the potential of SAPP as a lead compound for the development of therapeutic agents designed to inhibit P. gingivalis colonization and pathogenicity.

Results
Effect of SAPP on biofilm formation of P. gingivalis. P. gingivalis strains 33277 and W83 obtained from ATCC were selected as representatives of fimbriated, non-encapsulated and non-fimbriated, encapsulated lineages, respectively, to examine alteration of phenotypic properties of P. gingivalis in response to SAPP. As shown in Fig. 1a,b, growth rates of P. gingivalis 33277 or W83 were not significantly changed in the presence of SAPP (24 µM) compared to growth without SAPP. This observation suggests a killing-independent mechanism of SAPP action, which is in agreement with our earlier finding that SAPP represses expression and production of fimbrial proteins and gingipains in P. gingivalis 28 . While the predominant niche of P. gingivalis is the subgingival area, the organism also colonizes supragingival plaque and oral mucosal surfaces 29 . Indeed these sites, which are exposed to the salivary fluid phase, may represent early colonization events. Hence, P. gingivalis strains grown with or without SAPP were then tested for their ability to attach to saliva-covered surfaces. After a 24 h incubation, an approximate 25% decrease of attachment was detected with P. gingivalis 33277 grown with SAPP (24 µM) compared to the control without SAPP (Fig. 2a), while the decrease in attachment reached 70% after 48 h. A control peptide (peptide26 with 11 amino acids located immediately down stream of SAPP) was also tested for its role in biofilm formation of P. gingivalis, and it did not significantly affect the biofilm formation. The effect of W83 on monotypic biofilm formation was also tested with the bacterial cells grown with 24 or 48 µM SAPP. The ability of W83 to form biofilms was lower than that of 33277, and an impact of SAPP on biofilm formation by W83 was not observed after 24 h. However, after 48 h a 33% and 54% reduction in biofilm formation was observed for bacteria grown with 24 or 48 µM SAPP, respectively (Fig. 2b). These results indicate that SAPP can suppress biofilm formation by both 33277 and W83, with a more efficient inhibition occurring with 33277, likely due to the involvement of fimbrial adhesins in biofilm formation by 33277.
P. gingivalis and S. gordonii dual-species communities are one of the best-documented examples of synergistic oral bacterial interactions. The surface molecules involved in co-adhesion are well characterized and include FimA and Mfa1 of P. gingivalis and streptococcal glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and SspA/B 30 . We postulated that SAPP could prevent the formation of heterotypic P. gingivalis-S. gordonii biofilms by repressing the expression of fimA and mfa1. To test this, S. gordonii DL1 substrata were first established on saliva-coated wells and reacted with P. gingivalis grown with or without SAPP (24 µM). The amount of P. gingivalis 33277 bound to the S. gordonii DL1 biofilms was determined using qPCR. The number of P. gingivalis grown without SAPP was 2.5-fold higher than that of the bacteria grown with SAPP (Fig. 3). Not surprisingly, binding of P. gingivalis W83, an afrimbriated strain, to S. gordonii DL1 biofilms was not observed (data not shown), which further confirms the role of FimA and Mfa1 fimbriae in the interaction of P. gingivalis and S. gordonii.
We then tested whether SAPP could reduce the numbers of P. gingivalis cells in established heterotypic biofilms. To do this, we first generated heterotypic biofilms of S. gordonii-P. gingivalis that were grown without SAPP. After the removal of unbound bacteria, SAPP (24 µM) was added. Planktonic P. gingivalis was collected three times in a 24 h interval, and the biofilm was collected after 72 h. The numbers of planktonic or sessile P. gingivalis were determined using qPCR. Levels of planktonic P. gingivalis cells were significantly increased following SAPP treatment (Fig. 4a). Consistent with this, the number of P. gingivalis in the heterotypic biofilms was negatively correlated with that in the planktonic phase, and introduction of SAPP to P. gingivalis-S. gordonii heterotypic biofilms reduced the number of P. gingivalis by approximately 50% after 72 h (Fig. 4b). There was no significant difference in the total P. gingivalis cell numbers with or without SAPP (Fig. 4c), and numbers of S. gordonii in the growth media did not alter in the presence or absence of SAPP (Fig. 4d). These data suggest that SAPP does not affect bacterial viability; rather, it promotes detachment of P. gingivalis cells in the heterotypic biofilms and inhibits their re-entry into the biofilm. Furthermore, SAPP does not impact colonization of the health-related microbiota in this model system.  Impact of SAPP on intracellular invasion of P. gingivalis. The adherence of P. gingivalis to epithelial cells is mediated primarily by FimA and adhesive domains of gingipains [31][32][33][34] . Adherence subsequently initiates bacterial internalization of host cells 35 . Since SAPP represses FimA and gingipain production, we postulated that it may also inhibit P. gingivalis invasion of human oral keratinocytes (HOKs). We conducted antibiotic protection assays to compare the invasive ability of P. gingivalis 33277 and W83 grown with or without SAPP. Consistent with previous reports 36,37 , P. gingivalis W83 was much less efficient than strain 33277 in invading HOKs (Fig. 5a,b). Moreover, the invasion efficiency of 33277 grown with SAPP (24 µM) was reduced by approximate 4.8 fold compared to the control without SAPP, while a 5.2-fold difference in invasion efficiency was observed with W83 grown in the presence or absence of SAPP (48 µM). The differential potencies of SAPP on invasion inhibition of these two strains may due to different adhesins involved in their invasion process. P. gingivalis 33277 may mainly use its long and/or short fimbriae in the process, while W83 does not express these fimbriae and likely depends on other cell surface adhesins such as gingipains 38,39 . Effect of SAPP on gingipain enzymatic activity. P. gingivalis gingipains are multi-functional proteins with adhesive and catalytic domains. We have previously shown that the expression of genes encoding the arginine-and lysine-specific gingipains was repressed in P. gingivalis grown with SAPP 28 . Thus, we further tested and compared arginine-specific or lysine-specific activities in P. gingivalis strains grown with or without SAPP. An rgpA − , rgpB − , and kgp − triple mutant (KDP128) was used as a negative control, and as expected, no arginine-or lysine-specific enzyme activities were detected in the mutant (Fig. 6). Cell-associated arginine-specific protease activity of P. gingivalis 33277 grown with SAPP (48 µM) was decreased by 50%, while a 46% reduction was found in W83 under the same experimental conditions, compared to the control without SAPP (Fig. 6a). SAPP inhibited cell-associated lysine-specific protease activity of P. gingivalis 33277 and W83 to a similar degree (Fig. 6b). Rgp and Kgp activities in the growth media of 33277 and W83 were also monitored to determine the levels of secreted gingipains. After exposure to SAPP (48 µM), P. gingivalis 33277 exhibited a greater than 2.5-fold decrease in Rgp activity, and 2-fold less in Kgp activity (Fig. 6c,d). Rgp and Kgp activity in culture supernatants was also 50% lower after W83 was exposed to SAPP (48 µM). Significantly lower protease activities of Rgp and Kgp induced by SAPP may be predicted to reduce the virulence potential of P. gingivalis. Modulation of IL-8 levels by SAPP. P. gingivalis is known to selectively impair the production of certain cytokines and chemokines, such as IL-8. Thus, we compared the accumulation of IL-8 in the growth media of HOKs exposed to P. gingivalis grown with or without SAPP (48 µM) using an ELISA. The results showed that the ability of P. gingivalis 33277 or W83 to reduce the accumulation of Il-8 was significantly diminished by SAPP (Fig. 7a,b). These data indicate that SAPP has the potential to modulate manipulation of the innate immune response by P. gingivalis.

Discussion
Treatment of chronic and other forms of periodontitis usually involves physical removal of the plaque biofilm sometimes supplemented by systemic or local administration of antibiotics, especially in the cases of severe and refractory manifestations of the disease [40][41][42] . While the adjunctive use of antibiotics may improve the results of the mechanical therapy, meta-analyses suggest that the benefits of antibiotic treatment must be balanced against their side effects. Concerns regarding the use of antibiotics include breaking the symbiotic or mutualistic relationships between host and commensal microbiota 43 , and emerging oral bacterial resistance to antibiotics 44 . Therefore, targeting specific pathogenic bacteria, rather than non-selectively inhibiting both pathogens and commensal bacteria, has emerged as a strategic goal for maintaining a healthy oral microbiota.
Fimbrial proteins, FimA and Mfa1, and gingipains, RgpA, RgpB, and Kgp, are well-established virulence factors of P. gingivalis. Genes for mfa1 and fimA are present in the genomes of 20 of 21P. gingivalis strains isolated worldwide over a 25 year time period 45 , and thus far, all tested P. gingivalis strains produce gingipains that are both membrane associated and secreted in soluble form 46 . The primary role of FimA and Mfa1 is to promote the bacterial attachment to various oral surfaces, which leads to biofilm formation and internalization of epithelial cells 47,48 . The gingipains include two arginine and one lysine specific cysteine proteinases (RgpA, RgpB, and Kgp) 49 . Gingipains are multi-functional proteins that have important roles in nutrient acquisition and protein processing, and can also degrade host matrix proteins and immune effector molecules 50 . Gingipains play an important role in biofilm formation, especially for the strains lacking FimA expression, through the C-terminal adhesive regions of RgpA and Kgp, or through processing profimbrillin 51,52 . Animal studies with a murine lesion model using gingipain mutants found that gingipains, Kgp in particular, are primary contributors to the formation of skin abscesses 53 . Moreover, using the same animal model, pretreatment of P. gingivalis with a Kgp inhibitor also decreased bacterial virulence 54 , which demonstrates an essential role of gingipains in P. gingivalis infection. A previous study by Wilensky et al. also showed that alveolar bone loss was only observed in mice orally infected with RgpA-expressing P. gingivalis strains 55 . Therefore, inhibition of these P. gingivalis virulence factors has become the focus of therapeutic strategies for prevention and treatment of periodontitis, and a long list of gingipain inhibitors has been discovered, including synthetic compounds, proteins and peptides, and extracts from plants 56 . Nonetheless, the therapeutic potential of disruption of other pathogenic processes of P. gingivalis is also under investigation. Several molecules derived from marine natural products were identified as inhibitors of P. gingivalis-S. gordonii heterotypic community formation through repression of expression of mfa1 and fimA 57 . Other promising biofilm inhibitors are small peptides representing the binding domain (BAR) of S. gordonii SspB, which has been shown to disrupt P. gingivalis-S. gordonii communities and prevent bone loss in a mouse model 21,58 . One limitation of these inhibitors is that each of them only targets a single virulence factor of P. gingivalis, either a fimbrial protein or gingipains. In this report, we show that SAPP derived from S. cristatus ArcA has a much broader spectrum and is able to impede several virulence factors simultaneously. Such multiplicity of action greatly increases the value of this peptide as a potent inhibitor capable of eliminating P. gingivalis strains expressing virulence factors differentially.
In this study, we have demonstrated that SAPP does not impact growth of P. gingivalis but significantly reduces the numbers of P. gingivalis cells in both monotypic biofilms and P. gingivalis-S. gordonii heterotypic biofilms. A reduction of P. gingivalis cell number may be predicted to lead to a loss of competitive edge and a disruption of polymicrobial synergy. Another significant finding is that SAPP not only inhibits P. gingivalis biofilm formation but also disrupts established biofilms. Although the mechanism is unclear, we speculate that SAPP disrupts attachment and inhibits re-entry of the detached bacteria to the biofilm. The dispersed P. gingivalis cells may be eliminated from the oral cavity due to reduced ability to attach to oral surfaces and to invade host cells. Directly   Figure 7. Determination of IL-8 level in the grown media of HOKs. HOKs were exposed to P. gingivalis 33277 or W83, at a MOI of 10 for 2 and 18 h, and the culture supernatants and HOKs were collected, respectively. IL-8 levels in the culture media (a,b) were measured using an ELISArray Kit. Each bar represents means of IL-8 concentration, and standard deviations were calculated from three biological replicates. Asterisks indicate statistical difference of IL-8 concentration in the culture supernatants of HOKs treated with P. gingivalis grown with or without SAPP (p < 0.05; t-test).
targeting P. gingivalis attachment by SAPP should therefore stabilize a healthy microbiota and maintain homeostasis between the oral microbiota and host immune systems. Hence, it may be sufficient to prevent and treat P. gingivalis associated-periodontitis.
It has been reported that the expression of IL1-α, IL1-β, IL6, IL-8, IL-10, and NF-ƙB is modulated in oral epithelial cells infected with P. gingivalis [59][60][61][62][63][64] . In contrast to many periodontal colonizers that induce expression of IL-8 in gingival epithelial cells, P. gingivalis can inhibit IL-8 production, suggesting a unique role for P. gingivalis in regulating epithelial cell chemokine responses 65,66 . The ability of P. gingivalis to modulate IL-8 levels could play a pivotal role in initiating periodontitis, as this chemokine is involved in sustaining a healthy periodontium by maintaining a gradient for neutrophil recruitment into the gingival crevice 67 . The SAPP peptide attenuate the ability of P. gingivalis to reduce the accumulation of IL-8 in epithelial cell culture supernatants, suggesting that SAPP is able to partially restore the impairment of host immunity by P. gingivalis, which may facilitate maintenance of periodontal tissue homeostasis.
In conclusion, we have assessed the inhibitory role of SAPP in biofilm formation, invasion, and gingipain activity of P. gingivalis. We demonstrate that a unique aspect of this peptide is that it efficiently reduced all of these virulence-associated properties of P. gingivalis. It should be pointed out that effective inhibitory activity of SAPP reported here is found in standard culture media and buffers, and it is not known how SAPP works in vivo where SAPP is dissolved in saliva and/or gingival crevicular fluid with wide range of pH 68 . In addition, more than 200 distinct bacterial species inhabit the human oral cavity of any given individual 69,70 . Future studies will be designed to determine the efficacy of SAPP under much more complex conditions. Nevertheless, the ability of SAPP to selectively disperse P. gingivalis from dental plaque and attenuate its virulence potential makes it an attractive agent for controlling the composition of microbial communities and maintaining a healthy microbiota.

Methods
Bacterial strains and growth conditions. P. gingivalis 33277 and W83 were obtained from the American Type Culture Collection (ATCC, Manassas, VA), and cultured from frozen stocks in either Trypticase soy broth (TSB) or on TSB blood agar plates supplemented with yeast extract (1 mg/ml), hemin (5 μg/ml), and menadione (1 μg/ml), and incubated at 37 °C in an anaerobic chamber (85% N 2 , 10% H 2 , and 5% CO 2 ). S. gordonii DL1 was grown in Trypticase peptone broth (TPB) supplemented with 0.5% glucose at 37 °C under aerobic conditions. Peptide synthesis and activity. SAPP and a control peptide (peptide26 with 11 amino acids located immediately down stream of SAPP) were synthesized by Biomatik (Wilmington, DE) and purified with high performance liquid chromatography (HPLC) to achieve ≥ 95% purity. The purified peptide was resuspended in nuclease/proteinase-free PBS, aliquoted, and stored at −20 °C.
Monotypic biofilm assay. Attachment of P. gingivalis to saliva-coated surfaces was evaluated as described previously 71 . Briefly, P. gingivalis strains were grown to mid-log phase (OD 600 = 0.8) and collected by centrifugation. Bacterial cells (10 8 ) were resuspended in TSB, transferred to the wells of a 96-well polystyrene plate (Corning Inc., Corning, NY) that were precoated with human whole saliva that was diluted 2 times with PBS, and incubated at 37 °C. After washing, the biofilms were stained with 1% crystal violet and destained with 95% ethanol. The absorbance of the ethanol de-staining solution at 540 nm was then determined with the Ultrospec 2100 Pro spectrophotometer (Amersham Pharmacia Biotech, Piscataway, NJ).
Heterotypic biofilm assay. Heterotypic biofilms of P. gingivalis and S. gordonii were generated on a polystyrene six-well plate. S. gordonii DL1 cells (2 × 10 9 ) were first incubated aerobically in saliva-coated wells at 37 °C for 3 h, and the unbound cells were removed by washing with PBS three times. P. gingivalis cells (2 × 10 9 ) grown in TSB with or without SAPP were collected, re-suspended in ¼ TSB (1:3 TSB and PBS), added to the wells containing streptococcal biofilms, and incubated anaerobically at 37 °C for 4 h. The number of sessile P. gingivalis in S. gordonii biofilms was determined using qPCR. The bacterial cells were lysed with lysis solution (solution A; Invitrogen, Waltham, MA), and DNA was extracted using an Easy-DNA kit (Invitrogen). P. gingivalis cells in the biofilms were enumerated using a QuantiTect SYBR green PCR kit with 16S rRNA gene primers of P. gingivalis 25 . Standards used to determine numbers of P. gingivalis in the heterotypic biofilms were prepared using genomic DNA from P. gingivalis 33277. A fresh culture of 33277 was serially diluted in PBS and plated on TSB plates to obtain the colony forming units (CFUs) per milliliter for each dilution.
To determine whether SAPP promotes the release of P. gingivalis cells from the heterotypic biofilms and/ or inhibits their re-entry, a modified biofilm assay was conducted. P. gingivalis grown without SAPP was first added to wells of six-well polystyrene plates containing S. gordonii DL1 and incubated for 4 h. After removing the unbound P. gingivalis, fresh ½ TSB (1:1 TSB and PBS) containing SAPP and gentamicin (50 μg/ml) was then added and the plates were incubated with gentle shaking under anaerobic conditions. Planktonic P. gingivalis were collected at three 24 h intervals, while the sessile bacterial cells were collected at the 72-h time point. Bacterial DNA was purified using an Easy-DNA kit (Invitrogen), and the P. gingivalis cells were quantitated using qPCR.
ScIentIfIc REPORtS | 7: 16217 | DOI:10.1038/s41598-017-16522-y Bacterial invasion assay. The invasive ability of P. gingivalis was determined using an antibiotic protection assay 73 . P. gingivalis strains were grown in TSB for 16 h with or without SAPP to reach mid-log phase. The bacterial cells were collected by centrifugation. Human oral keratinocytes (HOKs, 5 × 10 5 ) (ScienCell Research Laboratories, Carlsbad, CA) were seeded in a six-well plate. After reaction with P. gingivalis 33277 or W83 for 1 h, the HOKs were washed with PBS to remove unbound bacteria and then continually cultured for another 4 h in the presence of antibiotics gentamicin (300 µg/ml) and metronidazole (200 µg/ml) to eliminate extracellular bacteria. The HOKs were then washed three times with PBS and lysed with sterile distilled dH 2 O. The internalized bacteria were plated on TSB blood agar plates. The plates were incubated anaerobically at 37 °C for 7 days, and CFUs of P. gingivalis were enumerated.
Enzyme-linked immunosorbent assay (ELISA). ELISA was performed using a human IL-8 Single Analyte ELISArray Kit (Qiagen, Redwood City, CA), according to the manufacturer's instructions. HOKs (1 × 10 5 ) were exposed to P. gingivalis (1 × 10 6 ) grown with or without SAPP for 2 and 18 h. The culture media of HOKs were collected and analyzed by ELISA. IL-8 was quantitated using a standard curve of the cytokine.

Statistical analyses.
A Student's t-test was used to determine the statistical significance of differences in functions and growth rates of P. gingivalis strains grown in the presence or absence of SAPP. A p < 0.05 was considered significant. All assays were performed with three biological replicates.