Host defense peptides clavanins A and MO reduce in vitro osteoclastogenesis

1 Postgraduate Program in Health Sciences, University of Brasília, Brasília, Distrito Federal, Brazil. 2 Proteomic and biochemical analysis center, Postgraduate Program in Genomic Sciences and Biotechnology, Catholic University of Brasilia, Brasília, Distrito Federal, Brazil. 3 Dentistry Course, Catholic University of Brasília, Brasília, Federal District, Brazil. 4 Postgraduate Program in Biotechnology and Biodivesity, Brasilia University, Brasilia, Distrito Federal, Brazil. 5 Graduate Program in Health and Development in the Midwest Region. Faculty of Medicine. Federal University of Mato Grosso do Sul. Campo Grande. Brazil. 6 S-Inova Biotech, Graduate Program in Biotechnology, Dom Bosco Catholic University, Campo Grande, Mato Grosso do Sul, Brazil.


Introduction
Bone remodeling is a process balanced between osteoblast-mediated bone deposition and osteoclast-developed bone resorption. Many oral diseases are mediated by an inflammatory process, increasing the recruitment of osteoclasts and enhancing bone erosion 1 . Periodontal disease and apical periodontitis present bone resorption with high osteoclast formation or hyperactivation, overcoming bone formation, and decreasing osteoblast activity 2 .
Inflammatory conditions, such as local osteolysis, can be associated with inducible nitric oxide synthase (iNOS) activation 3 . NO can also promote cytokine production and bone turnover besides indirect induction of bone resorption 3 . In this regard, periodontitis results in higher production of NO compared to healthy gingiva 4 . Periodontal treatment may involve the use of several systemic antibiotics such as tetracycline (minocycline and doxycycline) as adjuvants due to its local distribution. It was demonstrated that doxycycline hyclate gel (local therapy) could aid in scaling and root planning in patients with moderate to severe chronic periodontitis, but the benefit is still uncertain 5 . For the endodontic treatment, the use of calcium hydroxide (Ca(OH) 2 ) as a local antimicrobial is accepted worldwide as intracanal dressing 6 .
Despite the high success rate in existing periodontal and endodontic therapies, there are some limitations, mainly in tissue repair activity 6 . To improve bone repair in these diseases, it is essential to develop new substances. New therapies involving a direct effect on the bone can prolong the maintenance of the tooth in the oral cavity due to tissue support health. Host defense peptides (HDPs) are biomolecules from many organisms released in early defense response to infection and invasion by bacteria and other microorganisms 7 . HDPs may possess antimicrobial and immunomodulatory properties besides tissue repair induction 8 . In this context, human cells can be potential sources of HDPs 7 . Clavanin A is a promising HDP due to its known antibacterial, immunomodulatory, antitumor, and antiviral activities 9 . Besides, clavanin A was used as a model to create clavanin MO. Five hydrophobic amino acid residues (FLPII) were added to the N-terminus, being selected based on a computational search of the conserved region of other peptides with higher immunomodulatory activities 10 . It has been shown that different HDPs could improve therapies in the dental field.
Some peptides have been reported as having the potential to inhibit osteoclastogenesis such as LL-37 11 , human beta-defensin-3 with C-terminal end contains a 15-amino acid polypeptide (HBD3 -C15) 12 , synoeca-MP, 13 and HHC-10 13 . A previous study demonstrated that LL-37 and clavanins A and MO can modulate the inflammatory response of active cytokines presented in the osteoclastogenesis process, such as TNF-α, while Ca(OH) 2 up-regulated the IL-6 and IL-1α production 14 . This fact leads us to believe that clavanins A and MO may have the potential to inhibit osteoclastogenesis, a fact that has not yet been evaluated. Thus, this study aims to evaluate the biotechnological potential of HDPs clavanin A and MO in the oral osteoimmunological context and their capability to reduce in vitro osteoclastogenesis, compared to LL-37 (HDP control), Ca(OH) 2 (used in the endodontic treatment) and doxycycline (used in the periodontal treatment).

Doxycycline and calcium hydroxide preparation
Ca(OH) 2 (Iodontosul, Porto Alegre, Brazil) was weighed and diluted in ultrapure water before each experiment. Doxycycline (Pharmac, Brasilia, Brazil) was handled in capsules (100 mg in each unit). The capsules were opened, and doxycycline was weighed and diluted in ultrapure water before each experiment.

Cytotoxicity analyses
Peptides, Ca(OH) 2 , and doxycycline cytotoxicity were analyzed by MTT colorimetric assay (Sigma-Aldrich, St. Louis, USA), read in a microplate reader (Bio-Tek Power Wave HT, USA) at 570 nm 17 . Cell viability was determined after 3 and 7 days of cul-ture. All samples were compared to a positive control group (RAW culture), considered as 100% cell viability.

Nitric oxide production analysis
Nitrite production was evaluated in supernatants of cell cultures by Griess reaction, with adaptations 18 . Briefly, 100 μL of cell culture supernatant was transferred to a new 96-well plate (Kasvi, China). Then, 100 μL of 1% sulfanilamide phosphoric acid solution and 2.5% of 1% naphthyl ethylenediamine phosphoric acid (1:1) was added. After 10 min, reading was performed in a microplate reader (Bio-Tek PowerWave HT, USA) at 490 nm. The amount of nitrite was calculated based on a standard curve of sodium nitrite (1.5625 µM to 200 µM) 18 .

Tartrate-resistant acid phosphatase (TRAP) staining
TRAP staining was performed after 7 days of incubation for the quantification of differentiated osteoclast-like cells. The tartrate-resistant acid phosphatase (TRAP) kit (Sigma-Aldrich, St. Louis, USA) was used according to the manufacturer's specifications. Osteoclast-like cells were considered as TRAP-positive cells (with red/orange TRAP staining) with more than three nuclei.

Statistical analysis
Data obtained was analyzed by the standard error of the mean for each experiment. The normality was evaluated (Kolmogorov-Smirnov test), and subsequent parametric statistical analysis was carried out by two-way analysis of variance (two-way ANOVA) for the data from MTT and NO production and one-way ANOVA for TRAP analyses. Tukey's posthoc test was applied to identify statistical differences. Analyses were considered at the 95% significance level, and statistical differences were considered when p<0.05. Statistical analysis was performed using GraphPad Prism 6.0 software (Instat California, USA).

HDP cytotoxicity
The cytotoxicity of substances was determined by cell viability assays after 3 and 7 days of cell culture in the presence of HDPs clavanin A, clavanin MO, LL-37, Ca(OH) 2 and doxycycline. HDPs and Ca(OH) 2 were not cytotoxic to pre-osteoclasts (data not shown). However, doxycycline, at the high concentration (128 μg.mL -1 ) reduced cell viability by 48% (p<0.05), after 3 days of incubation, compared to the control group (data not shown). Similar viability results were observed in osteoclast-like cells (RAW cells with rRANKL), and substances after 3 and 7 days incubation. HDPs, Ca(OH) 2 and doxycycline were not cytotoxic to rRANKL-stimulated cells (Figure 1). However, 128 μg.mL -1 of doxycycline reduced cell viability by 42% after 3 days, compared to the control group (p<0.05). Indeed, HDPs were not cytotoxic and only doxycycline at 128 μg.mL -1 demonstrated a cytotoxic effect on osteoclast-like cells (rRANKL-stimulated and RAW cells).

Nitric oxide production
Cell cultures with HDPs, Ca(OH) 2 , and doxycycline produced basal levels of NO, compared to the control group (data not shown). The rRANKL increased NO production in  2 and doxycycline cytotoxicity at 2, 8, 32 and 128 μg.mL -1 on 2.5x10 3 RAW cells, after 3 and 7 days, by MTT assay. Cultures were stimulated with 100 ng.mL -1 of rRANKL. Cell viability was represented by percentage. Control group was represented by 2.5x10 3 RAW cells stimulated with 100ng.mL -1 of rRANKL and considered 100% of cell viability. All experiments were done in technical and biological triplicates. Statistical differences by two-way ANOVA test and Tukey's post hoc were represented by *p<0.05, **p<0.005, ***p<0.0005 and ****p<0.0001 compared to each concentration and time tested conditions; Dark green bars represent statistical differences observed on day 3; Light green bars represent statistical differences observed on day 7. MO at 128 μg.mL -1 , after 7 days (p <0.05). However, the presence of 128 μg.mL -1 of LL-37 after 7 days, increased the NO levels compared to the control group (p<0.0001), while the others concentrations at 3 and 7 days were similar to baseline levels. Ca(OH) 2 was able to upregulate NO production at 8 (p<0.005), 32 (p<0.05) and 128 μg.mL -1 (p<0.005) after 7 days of cell incubation. NO levels in the presence of Ca(OH) 2 at other different concentrations were similar to the control group (p<0.05), after 3 and 7 days. The lower concentration of doxycycline up-regulated NO production after 7 days compared to all concentrations, including the control group (p<0.0001). Overall, reduced levels of NO were observed in some concentrations of all substances, except for doxycycline and Ca(OH) 2 7 days after the test. Based on the number of osteoclast-like cells differentiated by rRANKL-stimulated RAW cell culture, the concentration of 8 μg.mL -1 was the lowest common concentration for HDPs, Ca(OH) 2 , and doxycycline, capable of reducing differentiation in osteoclast-like cells. Therefore, the best results were exhibited by LL-37 and Ca(OH) 2 . LL-37 showed approximately 67% fewer osteoclasts than clavanin A, clavanin MO and doxycycline, while Ca(OH) 2 showed 59% fewer osteoclasts compared to clavanins and doxycycline. Ca(OH) 2 and LL-37 demonstrated better osteoclastogenesis downregulation, compared to the same concentration of clavanin A, clavanin MO and doxycycline (p<0.0001). Therefore, among the tested HDPs, LL-37 presented the best ability to reduce the number of osteoclasts in vitro (p<0.05).

Discussion
Although periodontal and endodontic therapies are highly effective, new substances can improve outcome expectations. Antimicrobial, immunomodulatory, and reparative activity could be better achieved by new therapies and biologic substances 19 . Indeed, antimicrobial resistance is also a current limitation for both therapies 6 . The present study demonstrated the in vitro potential of HDPs clavanins A and MO in an osteoclastogenesis model. Results of cellular NO production, cytotoxicity, and the effects of HDPs on rRANKL-mediated osteoclastogenesis were compared to Ca(OH) 2 and doxycycline, widely used medications in endodontic 6 and periodontal 5 areas, respectively.
Study related to clavanin A has demonstrated different activities regarding this peptide, including important points for dentistry, such as antibiofilm and antimicrobial activity. HDP clavanin A showed antibiofilm activity against fungal biofilms when used to coat an amniotic membrane, which is frequently used in ophthalmologic surgery for rapid ocular surface reconstruction 20 . HDPs clavanin A, clavanin MO and LL-37 did not show any degree of cytotoxicity to RAW cells. A previous study using clavanin A also showed no cytotoxicity against mammalian cells (L929) with low concentrations 9 . Moreover, another study showed that 128 µg.mL -1 of clavanin MO did not demonstrate cytotoxicity compared to the other antimicrobial agents, with or without additional stimulation. LL-37 increased cell viability on RAW cells, and Ca(OH) 2 did not interfere with cell viability at the same concentration. Besides, after 6 h of incubation, clavanins alone reduced cell viability 14 .
NO regulates bone resorption through the regulation of the synthesis of OPG/ RANKL in bone marrow cells 21 , although other factors, including cytokines, are also involved. Our results demonstrated that HDPs downregulated NO production with or without the stimulation of rRANKL. Accordingly, clavanin A demonstrated a significant reduction in the number of osteoclast-like cells in a dose-dependent manner. Clavanin MO also reduced the number of differentiated osteoclast-like cells. Similarly, LL-37 at 8 and 128 μg.mL -1 demonstrated the best inhibition activity. According to previous results, HDPs can also modulate inflammatory mediators that contribute to the bone resorption activation process, such as TNF-α, IL-6 and IL-1α, and NO production 14 .
Different substances already used in clinical practice were also evaluated in this study. Ca(OH) 2 did not show cytotoxicity, and doxycycline demonstrated a toxic effect on cells at high concentration and increased cell viability at low concentration. Accordingly, a study evaluated the effects of a sub-antimicrobial dose of doxycycline (SDD) on ligature-induced periodontitis in spontaneously hypertensive rats. It concluded that SDD therapy exerted a systemic modulating effect on inflammation, with reduced periodontal tissue destruction in hypertensive rats 22 .
Ca(OH) 2 presented similar NO results compared to HDPs, and doxycycline presented an increase in NO levels, especially at 2 µg.mL -1 . In the osteoclastogenesis process, both tested drugs decreased the number of osteoclast-like cells in the presence of all concentrations tested. These facts suggest that although NO is strongly associated with osteoclast differentiation, this is not the only factor involved in the osteoclasto-genesis process 23 . Also, doxycycline might have another mechanism for downregulating the osteoclastogenesis pathway 23 .
However, these results are in agreement with previous results that suggest the inhibition of osteoclastogenesis in RAW cells in the presence of Ca(OH) 2 23 . Other studies suggest that the alkaline pH of Ca(OH) 2 can neutralize the lactic acid secreted by osteoclasts and may help prevent the destruction of mineralized tissue 24 .
In summary, this study aims to initiate the assessment of the biotechnological potential of HDPs clavanin A and MO in the oral osteoimmunological context and their capability to reduce in vitro osteoclastogenesis, compared to LL-37 (HDP control), Ca(OH) 2 (used in the endodontic treatment) and doxycycline (used in the periodontal treatment). We highlighted the results observed in the presence of 8 μg.mL -1 of LL-37 and Ca(OH) 2 , thus considering the use of these peptides as a possible product for endodontic and periodontal applications, in order to reduce the osteoclastogenesis process. On the other hand, Ca(OH) 2 shows low production costs when compared to LL-37. This HDP presents a relatively long sequence of amino acids, which raises its cost for synthesis, and it would probably only be indicated for restricted cases. Indeed, because of its immunomodulatory benefits and its biocompatibility, by being a peptide present in the oral cavity, LL-37 presents itself as a good candidate for dentistry use. When LL-37 results were compared to the doxycycline, the HDP demonstrated better efficiency in osteoclastogenesis downregulation at low concentrations, thus showing an even greater potential in the context of periodontal bone loss.
Despite the benefits highlighted in these data, in vitro results should be interpreted with caution and other in vivo studies are necessary to evaluate the potential of this biomolecule for clinical use. Other important points for future investigations should be focused on the large-scale expression of this peptide (lowering its cost), and the analysis of its integrity through various oral conditions, such as temperature changes, pH, and presence of lytic enzymes. In addition, other parameters should be evaluated, such as the peptides' mechanism of action in the osteoclastogenesis process, in order to enhance the knowledge on these potential products indicated for bone resorption processes, present in the periradicular area and periodontitis.