Application of Autologous Platelet-Rich Plasma on Tooth Extraction Site Prevents Occurence of Medication-Related Osteonecrosis of the Jaws in Rats

This study evaluated the effects of local application of autologous platelet-rich plasma (PRP) on the tooth extraction site of rats presenting the main risk factors for medication-related osteonecrosis of the jaw (MRONJ). For seven weeks, senile rats were submitted to systemic treatment with vehicle (VEH and VEH-PRP) or 100 μg/Kg of zoledronate (ZOL and ZOL-PRP) every three days. After three weeks, the first lower molar was extracted. VEH-PRP and ZOL-PRP received PRP at the tooth extraction site. Euthanasia was performed at 28 days postoperatively. Clinical, histopathological, histometric and immunohistochemical analyses were carried out in histological sections from the tooth extraction site. ZOL showed lower percentage of newly formed bone tissue (NFBT), higher percentage of non-vital bone tissue (NVBT), as well as higher immunolabeling for TNFα and IL-1β. In addition, ZOL presented lower immunolabeling for PCNA, VEGF, BMP2/4, OCN and TRAP. VEH and ZOL-PRP showed improvement in the tooth extraction site wound healing and comparable percentage of NFBT, VEGF, BMP2/4 and OCN. Local application of autologous PRP proved a viable preventive therapy, which is safe and effective to restore tissue repair capacity of the tooth extraction site and prevent the occurrence of MRONJ following tooth extraction.

SciENtific REPORTS | (2019) 9:22 | DOI: 10.1038/s41598-018-37063-y in oncologic doses 7 . Most cases of MRONJ are associated with intravenously administered nitrogen-containing BPs, among which zoledronate stands out as the most potent one. Patients most affected by this condition are women at an advanced age undergoing adjuvant therapy with BPs for cancer treatment, predominantly of multiple myeloma or breast cancer. The jaw is the most affected bone and the two main local risk factors are tooth extraction and periodontal disease 8,9 . MRONJ was first described by Marx 10 and its pathogenesis has not yet been fully elucidated. Several possible etiopathogenic factors of MRONJ have been raised concerning the use of BPs: i) induction of severe inhibition of osteoclast activity, preventing bone remodeling and resulting in non-vital bone accumulation; ii) toxic action on oral mucosa cells, reducing soft tissue repair ability, consequently impairing the underlying bone tissue; iii) antiangiogenic effect, resulting in both impairment of tissue repair and avascular necrosis of bone tissue; iv) increased infection susceptibility, since the drug facilitates adherence and bacteria colonization in the exposed bone in the oral cavity; v) dysfunction of the local immune response, resulting in impairment of the mechanisms of both defense and tissue repair [11][12][13][14] .
The limited understanding of the pathogenesis of MRONJ greatly restricts its prevention and treatment, and a standard protocol has not been established so far. Antibiotic therapy has been predominantly used as prevention 15 ; however, it may not always work efficiently 16 . MRONJ treatment has been carried out by different clinical approaches, based on the clinical staging of the disease. Therapy has been drug-based and/or surgical. Drug therapy consists mostly of the extended use of antimicrobial agents [17][18][19] . Surgical therapy ranges from conservative to aggressive, by means of curettage and/or sequestrectomy to resection of the jaws [17][18][19] . MRONJ treatment is time-consuming and generally results in failure or severe side effects [17][18][19] .
The biological properties of autologous platelet-rich plasma (PRP) make it a potential preventive therapy for MRONJ. PRP consists of a high concentration of platelets in a small volume of blood plasma 20,21 . Activated platelets, in a fibrin matrix, are a source of bioactive molecules that promote activation, proliferation and differentiation of a variety of cell types 22 . Studies show that PRP is able to accelerate the repair process of both bone tissue [23][24][25][26] and soft tissues, including epithelial and connective tissues [27][28][29][30] . Moreover, it has been reported that the use of PRP has an anti-inflammatory action 31 and antimicrobial effect 32 . Although the use of surgical debridement in combination with autologous PRP application for treating severe cases of MRONJ has been successful 33 , its effectiveness as a preventive therapy for this condition has not yet been properly evaluated. Therefore, the aim of this study was to evaluate the effects of local application of autologous PRP in the tooth extraction site in rats exhibiting the main risk factors for MRONJ.

Results
General health status, intra-oral and tooth extraction site clinical features. The general health conditions of animals used in this study remained constant and there was no loss of animals throughout the experimental period. The animals tolerated well the drug treatment, blood collection via cardiac puncture and first molar extraction. At the end of the experimental period, there was no statistically significant difference in the mean body weight of animals.
The intra-oral examination showed no macroscopic lesions in the oral cavity and tooth extraction site in groups VEH, VEH-PRP and ZOL-PRP. In contrast, great area of bone exposure (more than half of extraction socket) and impairment of mucous membrane repair were observed in the extraction site in four rats from ZOL group. The remaining rats exhibited similar clinical features; however, with small area of exposed bone (less than half of extraction socket). The scores, distribution of specimens, medians and interquartile ranges of attributed scores according to the clinical aspect of the tooth extraction site in VEH, VEH-PRP, ZOL and ZOL-PRP groups at 28 days postoperatively are presented in Table 1.
Qualitative and quantitative aspects of PRP samples. PRP smear samples showed structural integrity of platelets. The PRP samples from VEH-PRP and ZOL-PRP groups had at least 3.5 times more platelets than the blood samples that originated them.
Histopathological aspect of tooth extraction site and adjacent tissues. Parameters, scores, distribution of specimens, medians and interquartile ranges of attributed scores according to histopathological analysis at 28 days after tooth extraction in VEH, VEH-PRP, ZOL and ZOL-PRP are shown in Table 2.
In group ZOL, absence of epithelial tissue regeneration in the mucosa overlying the tooth extraction site and persistent inflammation in the lamina propria was observed. The interior of the extraction socket was empty or filled with small amount of trabecular bone and moderate amount of connective tissue. When present, bone tissue in the extraction socket was comprised of thin trabecular bone and large medullary spaces. The connective tissue at this site was composed of a large amount of inflammatory cells, a moderate amount of both fibroblasts and blood vessels, as well as a small amount of collagen fibers. Numerous and extensive areas of non-vital bone tissue interspersed with vital bone tissue were found surrounding the extraction socket. In this group, all specimens had non-vital bone areas involved with significant inflammatory cells and circumscribed by microbial colonies (Fig. 1).
VEH, VEH-PRP and ZOL-PRP groups showed similar histopathological characteristics. The extraction sockets were covered by epithelial tissue associated with connective tissue rich in fibroblasts, collagen fibers, blood vessels and some inflammatory cells. The interior of the extraction sockets were, in most specimens, predominantly filled by bone tissue, composed of moderately thick trabecular bone involving small medullary spaces filled with loose connective tissue and/or bone marrow. Large areas of non-vital bone tissue were absent in VEH and VEH-PRP. In group ZOL-PRP, larger amounts of empty osteocyte lacunae and larger amounts of small areas of non-vital bone tissue interspersed with vital bone tissue were clearly observed. In these groups, bacterial colonies were not observed in the bone tissue (Fig. 1). NFBT and NVBT in tooth extraction site and adjacent tissues. NFBT in the tooth extraction site was significantly lower in ZOL group, than in the other groups. There was no statistically significant difference between VEH and ZOL-PRP. NFBT was significantly higher in VEH-PRP than in VEH and ZOL-PRP (Fig. 1). The NVBT in ZOL and ZOL-PRP was significantly higher than in VEH and VEH-PRP groups. NVBT was significantly lower in ZOL-PRP than in ZOL group. No statistically significant difference was observed between VEH and VEH-PRP groups (Fig. 1).
TNFα and IL-1β immunolabeling in mucosa overlying tooth extraction site. In group ZOL, TNFα and IL-1β immunolabeling density in the connective tissue of the mucosa overlying the tooth extraction site was significantly larger than in the other groups. In ZOL-PRP, the immunolabeling for such pro-inflammatory cytokines was significantly lower than in ZOL group and was significantly higher than in VEH and VEH-PRP groups. There was no statistically significant difference between VEH and VEH-PRP groups (Fig. 2).
PCNA and VEGF immunolabeling in mucosa overlying tooth extraction site. The number of PCNA-positive and VEGF-positive cells in the connective tissue of the mucosa overlying the tooth extraction site was significantly lower in ZOL than in VEH, VEH-PRP and ZOL-PRP groups. There was no statistically significant difference between VEH and ZOL-PRP groups. PCNA and VEGF immunolabeling was significantly higher in VEH-PRP than in ZOL-PRP group and did not differ statistically from VEH (Fig. 3). TRAP immunolabeling in tooth extraction site. The amount of TRAP-positive cells and TRAP-positive cells coupled to the bone matrix did not differ in ZOL and ZOL-PRP groups. However, these parameters were lower in ZOL and ZOL-PRP when compared to VEH and VEH-PRP groups (Fig. 5).

Discussion
The use of antiresorptive drugs has increased significantly due to the aging of the population worldwide, as they constitute an effective therapeutic alternative for treatment of common diseases in the elderly [3][4][5] . Thus, patients making chronic use of antiresorptive drugs have also increased progressively in the dental office, frequently requiring dental interventions potentially capable of triggering MRONJ 34 . Hence, it is necessary to establish effective and safe preventive protocols during dental treatment. In this context, the purpose of this study was to evaluate the effectiveness of local application of autologous PRP in the tooth extraction site of rats exhibiting the major risk factors for MRONJ. The most important finding of this study was that local use of autologous PRP minimized the negative consequences of zoledronate on the wound healing process by stimulating tissue repair. Thus, PRP use is a favorable therapy for preventing the occurrence of MRONJ after tooth extraction.
Experimental models have contributed to the proposition and/or evaluation of preventive and/or curative therapies for MRONJ, constituting an important guideline for clinical research. In the present study, our research group designed an experimental model based on epidemiological studies. This experimental model displayed the main risk factors for MRONJ. Concerning the individual risk factor, female senile rats (20 months) were used, as the disease most frequently

Clinical Aspect of Tooth Extraction Site and Adjacent Tissues
(1) absence of exposed bone and totally repaired mucous membrane 7 7 -3 (2) absence of exposed bone and partially repaired mucous membrane ---4 (3) large extraction site with small area of exposed bone (less than half of extraction socket) and impairment of mucous membrane repair --3 -(4) large extraction site with great area of exposed bone (more than half of extraction socket) and impairment of mucous membrane repair --4 medians and interquatile ranges affects women at an advanced age 8,9 . Regarding the drug related risk factor, an oncologic dose of zoledronate was used, since this drug is the most potent BP linked to most cases of MRONJ 8,9 . As to local risk factors, experimental periodontitis was induced in the lower first molar, followed by extraction, considering the fact that tooth extraction with

Cellular Pattern and Epithelial Tissue Structure
(1) moderate layer of epithelial tissue completely recovering extraction site 4 5 -3 (2) thin layer of epithelial tissue completely recovering extraction site

Cellular Pattern and Bone Tissue Structure
(1) absence of non-vital bone in tissues adjacent to the extraction site and trabecular bone filling more than half of extraction socket 6 7 -4 (2) absence of non-vital bone in tissues adjacent to the extraction site and trabecular bone filling less than half of extraction socket 1 --2 (3) presence of few areas with non-vital bone in tissues adjacent to the extraction site and trabecular bone filling less than a third of extraction socket --2 1 (4) presence of many areas with non-vital bone in tissues adjacent to the extraction site and trabecular bone filling less than a third of extraction socket --5 medians and interquatile ranges 1  periodontal and/or periapical involvement is identified as an important precipitating factor 34,35 . In this experimental model, ZOL group showed a MRONJ condition, supported by clinical and histopathological analysis. A recent systematic review and meta-analysis of clinical studies reported that autologous PRP application in the tooth extraction site improves repair of both soft and hard tissues 36 , corroborating the findings of the present study. Keratinocytes, fibroblasts and osteoblasts are the main cells that in an orchestrated way participate in the wound healing process and are among the main targets of the PRP biostimulatory action [23][24][25][26][27][28] . In contrast, studies have demonstrated that zoledronate treatment exerts negative effects on keratinocytes 37,38 , fibroblasts 39,40 and osteoblasts 41,42 . The zoledronate action on such cells has great importance in both impaired tissue repair of the tooth extraction site and occurrence of MRONJ [11][12][13][14] , which is consistent with the results obtained in this study. Our data showed that autologous PRP application in the tooth extraction site was able to prevent the negative effects of zoledronate on such cells and tissues. In addition, epithelial tissue recovery, connective tissue repair, local inflammatory response modulation, alveolar bone neoformation and lower impairment of pre-existing bone tissue were observed in animals treated with zoledronate and autologous PRP.

Contamination Pattern of Tooth Extraction Site
Another undesirable aspect during zoledronate therapy is its action on the vasculature and main local angiogenesis modulators, which can be responsible for compromising the repair process of soft and hard tissues of the tooth extraction site. Thus, triggering avascular necrosis of alveolar bone tissue. Studies have shown that zoledronate adversely affects viability 43,44 , proliferation 45 , migration 44,45 , differentiation 46 and endothelial cell tube formation 45 , besides increasing apoptosis rate in such cells 43,44 . Zoledronate also reduces migration ability and differentiation of endothelial progenitor cells 44,46 . Santini et al. 47 showed a decrease in the amount of endothelial cells and endothelial precursor cells in peripheral blood of patients treated with zoledronate in oncologic doses. Other studies showed a severe decrease in angiogenesis of the oral mucosa of patients with MRONJ 48 and reduction in the expression of its major modulation factors, such as VEGF 49 , similar to the data from the present study. One of the indications that tissue repair was also stimulated by a favorable performance of PRP on vascular elements was an increased VEGF immunolabeling noticed in group ZOL-PRP.
Nitrogenous BPs, such as zoledronate, have osteoclasts as their main target cell 1,2 . This drug acts by blocking the action of the farnesyl diphosphate synthase enzyme, which belongs to the mevalonate pathway 1,2 . Blockage of this pathway inhibits osteoclastogenesis, prevents the activation of mature osteoclasts and induces premature apoptosis inactive osteoclasts 1,2 . A significant reduction in osteoclasts and osteoclasts coupled to the bone matrix in the tooth extraction site and surrounding areas was observed in ZOL and ZOL-PRP groups, confirming the effectiveness of the zoledronate treatment. Although some in vitro studies show that PRP promotes reduction in osteoclastogenesis 50,51 , the present study did not observe changes in the number of osteoclasts caused by the use of PRP, presumably due to the potent effect of zoledronate on osteoclasts and their precursors.
Another action of PRP, which probably contributed to the results observed in this study, was its anti-inflammatory action 31 . ZOL showed greater local inflammatory response and higher levels of key proinflammatory cytokines (TNFα and IL-1β) in the tooth extraction site. These data are in agreement with Morita et al. 52 , who suggested a close relationship between high levels of the proinflammatory cytokines TNFα, IL-1 and IL-6 in MRONJ pathogenesis, since mice deficient for these cytokines were resistant to developing osteonecrosis. In contrast, it was found that PRP use stimulated modulation of local inflammatory response in animals treated with zoledronate in our study. This was supposedly one of the responsible actions for promoting tissue repair process and preventing the occurrence of MRONJ after tooth extraction.
An additional systematic review including both in vitro and in vivo animal studies reported that platelet concentrates exhibit local antimicrobial action 32 . The present study was not aimed to perform an accurate microbiological evaluation. However, the histopathological analysis of the tooth extraction site from animals treated with zoledronate showed large areas of necrotic bone surrounded by bacterial colonies, corroborating previous studies [53][54][55] . Some studies demonstrated a close relationship between bacteria, especially of the Actinomyces genus, and MRONJ 56,57 ; nevertheless, it is unclear whether it is a primary or secondary event. In group ZOL-PRP, the histopathological analysis of the tooth extraction site revealed no bacterial colonies attached to the bone tissue similar to those detected in group ZOL. Although the antimicrobial PRP action mechanism is not yet clear, some studies have shown an extensive bacteriostatic and bactericidal spectrum of action, including against many common microorganisms in the oral microbiota 58 . It should be taken into account that an antimicrobial action is among the benefits of PRP use. However, further microbiological studies are needed to elucidate this assumption.
Barba-Recreo et al. 59 reported that rats treated with zoledronate and submitted to tooth extraction showed no improvement in the alveolar repair process after local allogeneic PRP application in the extraction site associated with mucoperiosteal flap. The use of allogeneic PRP may have compromised the benefits of this type of therapy. Such data and findings of this study highlight the importance of the origin of the PRP. In addition, the discrepancies between the scientific findings related to the use of PPR may be associated with the different preparation protocols 60 .
The present study evaluated the local use of autologous PRP as a preventative therapy for MRONJ due to the scarcity of studies on this topic. However, the combination of surgical lesion debridement and use of autologous PRP as a curative therapy for MRONJ have been investigated in experimental studies in animals and clinical studies. Sarkarat et al. 61 , performed curettage and local treatment with autologous PRP in a study using an experimental rat model after tooth extraction and clinical confirmation of the occurrence of MRONJ. These authors found no differences from the control group concerning epithelial regeneration, neoangiogenesis and bone sequestra formation; however, the amount of vital bone tissue at the surgical site was higher in rats treated with autologous PRP. Longo et al. 62 and Coviello et al. 63 comparatively evaluated the efficacy of isolated surgical therapy and surgical therapy associated with the local use of autologous PRP in clinical studies. Both studies found that the latter type therapy proved to be far more effective in treating MRONJ, which corroborates our findings and suggests that local use of autologous PRP can be a preventive and curative therapeutic option. The occurrence of MRONJ compromises the quality of life of patients 64 . Treatment of this condition can be extensive, ineffective, may impair the treatment of the underlying disease and result in serious sequelae [17][18][19] . The use of preventive therapeutic strategies is ideal in the case of MRONJ. Considering the beneficial effects of PRP, such as biomodulatory action, local antimicrobial activity and absence of adverse effects, local employment of autologous PRP may constitute a promising preventive therapy for MRONJ. In addition, despite using an experimental animal model, the findings of this study may be important in guiding future research. Thus, human clinical studies are necessary for the establishment of preventive protocols to be used in patients making use of antiresorptive drugs and requiring invasive dental interventions.

Conclusion
Within the limits of this study, it was concluded that local application of autologous PRP showed to be a viable, safe and effective preventative therapy to restore tissue repair capacity of the tooth extraction site that has been severely compromised by zoledronate treatment. The use of PRP showed a promising therapy to prevent the occurrence of MRONJ after tooth extraction.

Methods
Animals. The present study used twenty-eight senile (20 months) female rats (Wistar -Rattus norvegicus) with body weight ranging between 320-370 g. The animals were supplied by the School of Dentistry of Aracatuba (São Paulo State University -UNESP) and maintained throughout the experimental period under the following conditions: 12/12 light dark cycle at 22 ± 2 °C, relative humidity of 55 ± 5% and ventilation/exhaustion of 20 cycles  installed around the left first molar one day prior to the beginning of drug treatment (Fig. 6A,B). The ligature was maintained for three weeks in order to induce experimental periodontitis (EP) (Fig. 6C) 67,68 .
Drug treatment plan. The drug treatment was initiated one day after ligature installation and lasted seven weeks (three weeks prior to exodontia and four weeks after exodontia) (Fig. 6A)  Experimental groups. Three weeks after beginning the drug treatment, the rats were randomly distributed in four groups: VEH, VEH-PRP, ZOL and ZOL-PRP. The following drug and local procedures were performed: VEH (n = 7), treated with vehicle and submitted to tooth extraction and no local treatment; VEH-PRP (n = 7), treated with vehicle and submitted to tooth extraction and application of autologous PRP on extraction site; ZOL (n = 7), treated with zoledronate and submitted to tooth extraction and no local treatment; ZOL-PRP (n = 7), treated with zoledronate and submitted to tooth extraction and application of autologous PRP on extraction site (Fig. 6A).
PRP preparation protocol. Soon after the allocation of the animals in the different experimental groups, 1.5 ml of blood was collected via cardiac puncture in a syringe containing 0.15 mL of 3.2% sodium citrate. In groups VEH and ZOL the collected blood was discarded. In groups VEH-PRP and ZOL-PRP the collected blood was used to prepare the PRP. A sample of 25 µl of this blood from each animal in groups VEH-PRP and ZOL-PRP was also used in a blood smear for platelet count. The PRP preparation followed an adaptation of the protocol of Sonnleitner et al. 70 . Briefly, the blood samples were centrifuged at 160 G during 20 minutes for separation of blood plasma containing platelets, which was removed and centrifuged again at 400 G for 15 minutes. The platelet-poor plasma was discarded and 125 μl of PRP was separated for further use (100 µl was used to prepare the PRP gel and 25 µl reserved to be used in a PRP smear for qualitative and quantitative platelet evaluation. All centrifugation steps were performed in a refrigerated centrifuge ( Tooth extraction. Tooth extraction was performed soon after the blood collection for PRP preparation. Briefly, the ligature was removed, then antisepsis of the oral cavity, sindesmotomy, dislocation (Fig. 6D) and extraction of the first lower left molar (Fig. 6E) was done using adapted dental instruments 67,68 . No local treatment was performed in groups VEH and ZOL, that is, immediately after extraction, the edges of the surgical wounds were sutured with 4-0 silk thread (Johnson and Johnson), in order to assist the wound healing process.
Autologous PRP application in the tooth extraction site. Autologous PRP was used in the tooth extraction site in VEH-PRP and ZOL-PRP groups. The volume of PRP was submitted to activation with 10% calcium chloride (ScienceLab, TX, USA) (Fig. 6F) and placed in the center of the extraction socket (Fig. 6G), with the aid of a micropipette with an adapted tip. Next, the edges of the surgical wounds were sutured, as previously described, in order to assist the wound healing process and to guarantee the containment of the PRP gel activated inside the socket.
Euthanasia and sample collection. Euthanasia was carried out seven weeks after the beginning of the drug treatment, that is, 28 days after tooth extraction. The animals were deeply anesthetized and transcardially perfused with 0.9% sodium chloride added with 0.1% heparin (100 ml), followed by fixative solution (800 ml) of 4% formaldehyde ( The histological sections were submitted to hematoxylin-eosin staining (HE) for histopathological analysis of the tooth extraction site and adjacent tissues, and for histometric analysis of newly formed bone tissue (NFBT) and non-vital bone tissue (NVTB). For immunohistochemical analysis, the histological sections were divided into seven batches and submitted to indirect immunoperoxidase technique. The immunohistochemical processing followed the protocol described previously by Ervolino et al. 68

Analysis of Results
Analysis of the general health condition and intra-oral clinical examination. The general health condition of the animals was observed throughout the experimental period and the body weight was monitored weekly. Intra-oral clinical examination was performed, consisting of a detailed visual inspection of the oral cavity, in particular, of the tooth extraction site. The evaluated clinical parameters and their respective scores were based on Statkievicz et al. 67 and Ervolino et al. 68 and are shown in Table 2. Such data were expressed as medians and interquartile ranges of attributed scores in each evaluated parameter.
Microscopic analysis. Microscopic analyzes were performed by a certified histologist (EE) who was blinded to treatments. For histopathological and histometric analyzes, three histological sections from the buccal, middle and lingual portion of the extraction socket were used. A histological section of the central portion of the extraction socket was used for immunohistochemical analysis.

Region of interest (ROI).
In the present study, ROI (I), ROI (II) and ROI (III) were considered according to the performed microscopic analysis. ROI (I) comprised a panoramic view of the tooth extraction site and adjacent tissues, consisting of a 4 mm × 4 mm area of the portion of the extraction socket previously occupied by the mesial and distal roots of the lower left first molar and adjacent tissues. The distal limit consisted of a line situated parallel to the surface of the coronary and radicular dentin of the lower left second molar, extending 4 mm mesially. The coronary limit consisted of a line located parallel to the gingival-cement boundary of the lower left second molar, extending 4 mm apically 67,68 .
ROI (II) comprised samples of the tissue overlying the tooth extraction site, consisting of two 250 μm × 250 μm areas located in the tissue overlying the tooth extraction site. The delimitation of this area followed a line located at the center of the connective tissue, perpendicular to the long axis of the teeth and dividing such tissue at its coronal apical direction. Two other lines were used, one parallel to the central portion of the cavity formerly occupied by the mesial root and another one parallel to the central portion of the cavity formerly occupied by the distal root of the first molar. The intersection of these lines determined the center of the two analyzed areas 67,68 .
ROI (III) concomitantly encompassed samples of pre-existing bone tissue in the alveolar wall and newly formed bone tissue inside the extraction socket, both vital tissues. An area of 500 μm × 500 μm located in the apical portion of the extraction socket within the space previously occupied by the mesial root or the distal root was analyzed. The center of these areas was positioned at the boundary between the pre-existing bone tissue and  67 and Ervolino et al. 68 and are shown in Table 2. Such data were expressed as medians and interquartile ranges of attributed scores in each evaluated parameter.
Histometrical Analysis of NFBT and NVBT. In ROI (I), images were captured by using a digital camera (AxioCam ® Carl Zeiss, Gottingen, Germany) coupled to an optical microscope (AxioLab ® ) and connected to a microcomputer. With the aid of image analysis software (Axiovision 4.8.2 ® Carl Zeiss), the total amount of bone tissue was measured; then, the percentage of newly formed bone tissue (NFBT) and percentage of non-vital bone tissue (NVBT) were calculated 68 . The NVBT consisted of regions where more than ten adjacent osteocyte lacunae were empty or containing osteocyte necrotic debris. The percentage of NFBT and NVBT was expressed as mean ± standard deviation.
Immunohistochemical analysis of TRAP in tooth extraction site. In ROI (I), images of histological sections immunolabeled with TRAP were obtained as previously described. With the aid of image analysis software (Axiovision 4.8.2 ® Carl Zeiss), TRAP-positive cells and TRAP-positive cells coupled to the bone matrix were quantified 68 . The amount of TRAP-positive cells per mm² of bone was expressed as mean ± standard deviation.
Immunohistochemical analysis of TNFα and IL-1β in the mucosa overlying the tooth extraction site.
In ROI (II), images of histological sections immunolabeled with TNFα and IL-1β were obtained as previously described. The area corresponding to the immunolabeling was obtained through immunolabeling density using a color threshold tool of image analysis software (Axiovision 4.8.2 ® Carl Zeiss) 67,68 . The immunolabeling density per area was expressed in percentage as mean ± standard deviation.
Immunohistochemical analysis of PCNA and VEGF in the mucosa overlying the tooth extraction site.
In ROI (II), images of histological sections immunolabeled with PCNA and VEGF were obtained as previously described. PCNA-positive cells and VEGF-positive cells were quantified with the aid of an image analysis software (Axiovision 4.8.2 ® Carl Zeiss). The amount of immunolabeled cells per mm² was expressed as mean ± standard deviation.
Immunohistochemical analysis of BMP2/4 and OCN in bone tissue of tooth extraction site. In ROI (III), images and analysis of histological sections immunolabeled with BMP2/4 and OCN were obtained as described above. The amount of immunolabeled cells per mm² was expressed as mean ± standard deviation.

Data Availability
Data from this manuscript are available from Scientific Reports.