Titanium Particles Modulate Lymphocyte and Macrophage Polarization in Peri-Implant Gingival Tissues

Titanium dental implants are one of the modalities to replace missing teeth. The release of titanium particles from the implant’s surface may modulate the immune cells, resulting in implant failure. However, little is known about the immune microenvironment that plays a role in peri-implant inflammation as a consequence of titanium particles. In this study, the peri-implant gingival tissues were collected from patients with failed implants, successful implants and no implants, and then a whole transcriptome analysis was performed. The gene set enrichment analysis confirmed that macrophage M1/M2 polarization and lymphocyte proliferation were differentially expressed between the study groups. The functional clustering and pathway analysis of the differentially expressed genes between the failed implants and successful implants versus no implants revealed that the immune response pathways were the most common in both comparisons, implying the critical role of infiltrating immune cells in the peri-implant tissues. The H&E and IHC staining confirmed the presence of titanium particles and immune cells in the tissue samples, with an increase in the infiltration of lymphocytes and macrophages in the failed implant samples. The in vitro validation showed a significant increase in the level of IL-1β, IL-8 and IL-18 expression by macrophages. Our findings showed evidence that titanium particles modulate lymphocyte and macrophage polarization in peri-implant gingival tissues, which can help in the understanding of the imbalance in osteoblast–osteoclast activity and failure of dental implant osseointegration.


Introduction
The use of dental implants to replace missing teeth is beneficial and increasing due to the reliable and durable performance of implant-supported prostheses. This modality to replace missing teeth is expected to continue in the future as a result of the global population aging and advancements in dental treatments. Though dental implants display high success rates, it is important to note that the challenges in this approach of treatment are the failure of approximately 5% to 11% of the implants within 10 to 15 years and the necessity of removal of the implants [1]. The primary causes of this failure are attributed to biological or biomechanical factors; these factors are complex and interrelated. However, peri-implantitis is an inflammatory process that affects peri-implant tissues, leading to the cells, consequently increasing the release of TNF-α [19,20]. Furthermore, cytokines such as TNF-α and IL-1β were known to induce the expression and activate the promoter of CCL3, which was reported to play a critical role in the potential migration of macrophages [21]. During inflammation, leukocyte recruitment is regulated by various cytokines, including IL-8/CXCL8, CCL2, IL-1β, and TNFα, which act as prominent chemoattractants [22].
However, there is a gap in the knowledge about the role of titanium particles in the development of peri-implantitis and the failure of dental implant osseointegration. Exploring this area of research can shed light on the relationship between titanium particle infiltration into peri-implant tissues and the modulation of the immune cells, thereby enhancing our understanding of implant osseointegration failure. This will contribute to the understanding of the imbalance in osteoblast-osteoclast activity and the failure of dental implant osseointegration. Therefore, the primary objective of this study is to investigate the infiltration of titanium particles and immune cells into the peri-implant gingival tissues by conducting a comparative gene expression analysis using whole transcriptome analysis on gingival tissues surrounding both successful and failed dental implants, in comparison to a reference group without implants. The subsequent step was functional validation using THP-1 monocyte-derived macrophage cell lines to confirm findings and provide additional insights.

Transcriptional Profiling Distinguishes between FI, SI and NI Gingival Tissue Samples
The differential gene expression analysis showed distinct gene expression profiles among FI and SI in comparison to NI gingival tissue samples. After standard normalization and filtering, a total of 2185 and 5375 DEGs resulted in FI compared to NI and SI compared to NI, respectively. The differentially expressed genes were selected with FC > 2 at a significance level of the adjusted p-value (FDR) < 0.05 and used for unsupervised hierarchical clustering analysis. The patients with FI and NI are clustered as a single branch, clearly showing subgroups of FI and NI clusters irrespective of their origin from different patients with FI and NI ( Figure 1A). In addition, the patients with SI and NI are clustered as a single branch, clearly showing subgroups of SI and NI clusters irrespective of their origin from different patients with SI and NI ( Figure 1B). The gene tree in each unsupervised hierarchical clustering was ordered according to the enriched biological pathways by the Metascape annotation tool, and the gene clusters enriched for the top enriched pathways are indicated. Hence, the transcriptomic profiling found in the heatmaps shows a unique signature between FI vs. NI and SI vs. NI, which confirms the immune fraction and their related pathways in each subgroup.

Dysregulation of Cellular Response to Chemical Stress and Immune Response in Gingival Tissues from SI and FI
To investigate the functional relevance of the significant DEGs in the gingival tissues from FI and SI compared to tissues from NI, functional and pathway analyses were performed ( Table 1). The analyses revealed key cellular pathways such as the cellular response to cadmium ion (p < 0.01) and lymphocytes B-Cells and T-Cells activation (p < 0.0001 and 0.01, respectively) and proliferation processes (p < 0.001) that were over-enriched in FI compared to NI. In contrast, the cellular response to chemical stress (p < 0.0001) and response to alcohol were significantly downregulated in FI compared to NI. Interestingly, different pathways related to the immune response and process and the activation of immune cells were upregulated in SI, such as leukocyte migration (p < 0.0001), myeloid leucocyte activation (p < 0.00001), neutrophil and leukocyte degranulation (p < 0.0001, 0.001, respectively), in addition to the regulation of cytokine production (p < 0.001) and mainly the positive regulation of interleukin-10 production (p < 0.001). Moreover, the cellular defense response (p < 0.001) and apoptotic cell clearance (p < 0.001) pathways were highly enriched in SI compared to NI. However, the functional clustering significantly revealed that cellular

Dysregulation of Cellular Response to Chemical Stress and Immune Response in Gingival Tissues from SI and FI
To investigate the functional relevance of the significant DEGs in the gingival tissues from FI and SI compared to tissues from NI, functional and pathway analyses were performed ( Table 1). The analyses revealed key cellular pathways such as the cellular response to cadmium ion (p < 0.01) and lymphocytes B-Cells and T-Cells activation (p < 0.0001 and 0.01, respectively) and proliferation processes (p < 0.001) that were over-enriched in FI compared to NI. In contrast, the cellular response to chemical stress (p < 0.0001) and response to alcohol were significantly downregulated in FI compared to NI. Interestingly, different pathways related to the immune response and process and the activation of immune cells were upregulated in SI, such as leukocyte migration (p < 0.0001), myeloid leucocyte activation (p < 0.00001), neutrophil and leukocyte degranulation (p < 0.0001, 0.001, respectively), in addition to the regulation of cytokine production (p < 0.001) and mainly the positive regulation of interleukin-10 production (p < 0.001). Moreover, the cellular defense response (p < 0.001) and apoptotic cell clearance (p < 0.001) pathways were highly enriched in SI compared to NI. However, the functional clustering significantly revealed that cellular responses to stimuli (p < 1 × 10 −53 ) and Programmed Cell Death (p < 1 × 10 −14 ) were downregulated in SI compared to NI.

GSEA Identifies the Activated Cellular Pathways in Gingival Tissues from SI and FI
To identify more detailed pathways and the related DEGs in each comparison of FI vs. NI and SI vs. NI, an exhaustive GSEA was performed. The differentially activated significant pathways were identified for each comparison based on the cut-off nominal p < 0.05. The details of the significantly enriched gene sets are listed in Table 2. Further, the leading-edge analyses revealed important differentially activated genes that are involved in multiple inflammations and immune responses related to biological processes or molecular pathways, including IL1B, CDK3, IL27 and CD86 in FI vs. NI comparison (Figure 2A

Polarization of Macrophage in Peri-Implant Gingival Tissues
A focused analysis of the deconvolution of immune cells for the DEGs in SI and FI compared to NI was assessed using the CIBERSORT computational method ( Figure 3). The digital cytometry analysis showed that the percentage of M1 macrophages was 9.9% in SI, slightly higher than that in FI (9.4%). However, the percentage of M2 macrophage was 22.3% in FI, which was more than that in SI (18.7%). Interestingly, there was a clear variation in the percentages of T cell subtypes between FI and SI samples; for instance, the percentages of regulatory T cells were 16.4% in FI and 3.15% in SI (Table 3).

Polarization of Macrophage in Peri-Implant Gingival Tissues
A focused analysis of the deconvolution of immune cells for the DEGs in SI and FI compared to NI was assessed using the CIBERSORT computational method ( Figure 3). The digital cytometry analysis showed that the percentage of M1 macrophages was 9.9% in SI, slightly higher than that in FI (9.4%). However, the percentage of M2 macrophage was 22.3% in FI, which was more than that in SI (18.7%). Interestingly, there was a clear variation in the percentages of T cell subtypes between FI and SI samples; for instance, the percentages of regulatory T cells were 16.4% in FI and 3.15% in SI (Table 3).

Titanium Particles Result in Triggering the Immune System in Both FI and SI
The count of the DEGs across the two types of peri-implant gingival tissues (FI and SI) was investigated ( Figure 4A). Functional clustering and pathway analysis of the common DEGs in FI and SI peri-implant gingival tissues showed significant enrichment of transcripts involved in response to stimulus, immune system defense, detoxification and signaling ( Figure 4B).

Titanium Particles Result in Triggering the Immune System in Both FI and SI
The count of the DEGs across the two types of peri-implant gingival tissues (FI and SI) was investigated ( Figure 4A). Functional clustering and pathway analysis of the common DEGs in FI and SI peri-implant gingival tissues showed significant enrichment of transcripts involved in response to stimulus, immune system defense, detoxification and signaling ( Figure 4B).

Titanium Particles Result in Triggering the Immune System in Both FI and SI
The count of the DEGs across the two types of peri-implant gingival tissues (FI and SI) was investigated ( Figure 4A). Functional clustering and pathway analysis of the common DEGs in FI and SI peri-implant gingival tissues showed significant enrichment of transcripts involved in response to stimulus, immune system defense, detoxification and signaling ( Figure 4B).

Higher Infiltration of Immune Cells in Peri-Implant Gingival Tissues of FI Compared to SI or NI
To confirm immune cell infiltration in peri-implant gingival tissues of FI compared to SI and NI, H&E staining was performed. Microscopic examination showed infiltration of more inflammatory cells into the gingival tissues which were collected from patients with FI, compared to gingival tissues from SI. The predominance of macrophages and lymphocytes is an indication of chronic inflammation. Indeed, the specification of the immune cells was based on the morphology of each cell type based on the nucleus and cytoplasm. Furthermore, titanium particles were evident in the H&E images in the periimplant gingival tissues collected from patients with FI and SI. However, the gingival tissues collected from patients without dental implants revealed none to minimal infiltration of inflammatory cells ( Figure 5). The identification of titanium particles was based on the morphology as well; they are completely different from the inflammatory cells as they look like black spots. of more inflammatory cells into the gingival tissues which were collected from patients with FI, compared to gingival tissues from SI. The predominance of macrophages and lymphocytes is an indication of chronic inflammation. Indeed, the specification of the immune cells was based on the morphology of each cell type based on the nucleus and cytoplasm. Furthermore, titanium particles were evident in the H&E images in the peri-implant gingival tissues collected from patients with FI and SI. However, the gingival tissues collected from patients without dental implants revealed none to minimal infiltration of inflammatory cells ( Figure 5). The identification of titanium particles was based on the morphology as well; they are completely different from the inflammatory cells as they look like black spots.

Immunohistochemistry Analysis Validates the Infiltration of Macrophages and Lymphocytes into Peri-Implant Gingival Tissues
In order to validate the infiltration of immune cells represented by mainly macrophages and lymphocytes in the peri-implant gingival tissues, an immunoreactive score (IRS) using IHC analysis findings was performed to evaluate the expression of CD68 and CD3 markers representative of macrophages and lymphocytes, respectively. Analysis of IHC images showed positive expression of CD68 and CD3 in cases of FI, which was more pronounced than that in SI. In addition, the gingival tissues collected from patients without dental implants showed negative expression of the same markers ( Figure 6). Statistical analysis of the immune markers between the various implant groups showed that both CD68 and CD3 were significantly higher in the FI group compared to the SI group (p < 0.01), and both were slightly less significant when comparing the NI group to the FI group (p < 0.05), but they were not significant when comparing NI to SI groups.

Immunohistochemistry Analysis Validates the Infiltration of Macrophages and Lymphocytes into Peri-Implant Gingival Tissues
In order to validate the infiltration of immune cells represented by mainly macrophages and lymphocytes in the peri-implant gingival tissues, an immunoreactive score (IRS) using IHC analysis findings was performed to evaluate the expression of CD68 and CD3 markers representative of macrophages and lymphocytes, respectively. Analysis of IHC images showed positive expression of CD68 and CD3 in cases of FI, which was more pronounced than that in SI. In addition, the gingival tissues collected from patients without dental implants showed negative expression of the same markers ( Figure 6). Statistical analysis of the immune markers between the various implant groups showed that both CD68 and CD3 were significantly higher in the FI group compared to the SI group (p < 0.01), and both were slightly less significant when comparing the NI group to the FI group (p < 0.05), but they were not significant when comparing NI to SI groups.

Protein Expression of Various Cytokines by M0 Macrophages Treated with TiO2
Various macrophage inflammatory-associated cytokines were investigated upon the treatment of M0 macrophages for 24 h with 20 and 100 µg/mL of TiO2 NPs and MPs. Interestingly, there was an increase in the release of IL-1β, IL-8 and IL-18 by M0 macro-

Protein Expression of Various Cytokines by M0 Macrophages Treated with TiO 2
Various macrophage inflammatory-associated cytokines were investigated upon the treatment of M0 macrophages for 24 h with 20 and 100 µg/mL of TiO 2 NPs and MPs. Interestingly, there was an increase in the release of IL-1β, IL-8 and IL-18 by M0 macrophages after being treated for 24 h with 100 µg/mL of TiO 2 NPs (p < 0.05, p < 0.01 and p < 0.0001, respectively) and MPs (p < 0.05, p < 0.01 and p < 0.001, respectively) compared to untreated cells. Furthermore, there was a significant increase in the levels of IL-1β, CCL3, and IL-8 between the 20 and 100 µg/mL treatments of TiO 2 NPs and MPs. However, TNF-α did not show any significant change upon different treatments of TiO 2 (Figure 7). Data are represented as mean ± SEM from two separate experiments. * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001.

Protein Interaction Network of the Significantly Altered Cytokines in the Peri-Implant Gingival Tissues with the Presence of TiO2
We constructed protein interaction networks of the significantly altered cytokines in the peri-implant gingival tissues (CCL3, IL-1β, IL-8, and IL-18) which were validated in the M0 macrophages treated with TiO2. The interaction network analysis showed that all cytokines are functionally interconnected (Figure 8). Analysis of the protein interaction network showed the presence of several DEGs in FI compared to NI, including AKT1, GNA15, GNAI1, GNB2L1, GSDMD, NLRC4 and SMO that showed potential interactions and commonly affect each other's expression and function. Thus, it can be supposed that We constructed protein interaction networks of the significantly altered cytokines in the peri-implant gingival tissues (CCL3, IL-1β, IL-8, and IL-18) which were validated in the M0 macrophages treated with TiO 2 . The interaction network analysis showed that all cytokines are functionally interconnected (Figure 8). Analysis of the protein interaction network showed the presence of several DEGs in FI compared to NI, including AKT1, GNA15, GNAI1, GNB2L1, GSDMD, NLRC4 and SMO that showed potential interactions and commonly affect each other's expression and function. Thus, it can be supposed that the disruption of the cytokines CCL3, IL-1β, IL-8, and IL-18 plays an essential role in dental implant osseointegration.

Discussion
To the best of our knowledge, this is the first study using the transcriptome analys to compare peri-implant gingival tissues from FI, SI and NI patient groups. In the prese study, the distinct gene expression profiles in transcriptome analysis between tissu

Discussion
To the best of our knowledge, this is the first study using the transcriptome analysis to compare peri-implant gingival tissues from FI, SI and NI patient groups. In the present study, the distinct gene expression profiles in transcriptome analysis between tissue samples from FI, SI and NI could be considered an indication of the impact of titanium particles in the gene expression changes. Moreover, this may be correlated to the infiltration of inflammatory cells, which confirms the unique immune fraction in each subgroup resulting in the hierarchical clustering and heatmap analyses. Furthermore, the data extracted from the in silico flow cytometry analysis confirmed the profusion of macrophages in the periimplant tissues with variations in the M1 with a 9.4% and 9.9% ratio and the M2 with a 22.3% and 18.7% ratio, respectively, in FI and SI tissues compared to NI samples. The percentage of M2 macrophages and T cells in the tissue from cases with FI was more than that in SI, but the percentage of M1 macrophages was more in SI than that in FI. This finding is in accordance with orthopedic studies that have reported an abundance of CD68 macrophages in periprosthetic orthopedic tissues containing wear metal debris [18] and an increased number of CD68 macrophages in an aseptically loosened orthopedic prosthesis. Yang et al., analyzed mice implanted with human periprosthetic tissues and reported CD68 cells in much higher quantities when titanium particles were present compared to those free of titanium particles [23]. This increase in CD68 cells was thought to be associated with an increased number of particles-engulfed monocytes at sites containing titanium particles and wear debris. Our findings showed that M1 and M2 macrophages were detected in peri-implant gingival tissues from patients with FI and SI, which followed other studies where the plasticity and polarization of macrophages have been reported in the context of peri-implantitis [24]. Further future studies, which consider the use of CD163 and CD86, will be undertaken in order to investigate the polarization of M2 and M1, respectively.
The functional clustering and pathway enrichment of the differentially expressed genes in FI vs. NI, or SI vs. NI, identified various molecular pathways and biological processes. However, the immune response pathways were the most frequent in both comparisons, suggesting the important roles of the infiltrating immune cells in the periimplant tissues. For the DEGs in the FI group, we showed that B cells and T cells activation and proliferation processes were upregulated, although the cellular responses to chemical stress and response to alcohol were under-regulated in FI compared to NI. Indeed, for the DEGs in the SI group, different pathways related to the immune response, process and activation of immune cells were upregulated, including leukocyte migration, myeloid leukocyte activation, neutrophil and leukocyte degranulation, in addition to the regulation of cytokine production. These findings are consistent with previous results showing that in peri-implant tissues, the gene expression changes are mainly related to immune responses and defense responses [25]. Additionally, the GSEA further showed that gene sets related to M1/M2 macrophage polarization, immune responses, immune cell activation and proliferation and cytokine signaling were differentially over-represented between the different subgroups. Importantly, the leading-edge analyses revealed important differentially activated genes that are involved in multiple inflammations and immune responses related to biological processes or molecular pathways, including IL-1β, CDK3, IL-27 and CD86 in the FI vs. NI comparison, in addition to CXCL6, CXCL1, CCL7, CCL13 and IL18 in the SI vs. NI comparison. Previous transcriptomics studies profiling the differentially expressed genes in peri-implantitis have pointed out that the significantly altered pathways were related to extracellular matrix molecules [26], matrix-degrading enzymes [27] and inflammatory pathways including the RANKL/OPG pathway and the cyclooxygenase2 pathway [28].
On the other hand, the functional enrichment analysis of the DEGs common to FI and SI peri-implant gingival tissues showed that significant enrichment of transcripts is involved in response to stimulus, immune system defense, detoxification and signaling, indicating that in both cases of failed and successful implants, there is a release of the titanium which is considered as stress for the cells in the peri-implant area and therefore, it results in the triggering of the immune system, activating mainly M1 macrophage po-larization, and maybe later, M2 polarization. In addition, it can thus be assumed that an imbalance of macrophage M1-M2 polarization is often associated with various inflammatory conditions, which leads to implant failure in some cases and success in others. This finding is in agreement with the Tilton et al. study, and further supports the idea that the exposure of THP-1 monocytes to TiO 2 particles may induce the biological processes and differential regulation of inflammation involved in neutrophil activation and immune responses such as phagocytosis [29]. These results are in good agreement with our earlier study which showed that the phagocytosis of TiO 2 particles by macrophages is size-and dose-dependent [30].
In the present study, the images from H&E slides confirmed the presence of titanium particles and infiltration of inflammatory cells in the peri-implant gingival tissues collected from FI or SI patients. The visual examination of the images showed that titanium particles were present in approximately 85% of the tissue samples, signifying that titanium particles often dislodge from the implant surface. However, the expression of inflammatory cells in the FI samples was higher than that in the SI samples. This finding refers to the role of titanium particles to increase inflammation in the peri-implant gingival tissues. As proposed forward by Tawse-Smith et al. [7], the titanium particles were found highly abundant in all peri-implantitis tissue when compared with other test and control areas using SEM-EDS analysis. We showed that the inflammatory cells infiltrating the peri-implantitis tissues were primarily chronic, which is a common finding for peri-implantitis tissues. This is in agreement with a study by Dapunt et al. who found that the release of metal ions from the prosthesis can induce an acute phase of the chronic inflammatory disease with infiltrates of monocytes, T cells, and osteoclasts [31]. Nevertheless, the influence of microbial factors will never be neglected in the development of peri-implantitis, though, the main focus of the current study was investigating the infiltration of titanium particles and inflammatory cells into peri-implant gingival tissues and the consequences of this fact. Hence, additional assessments in the gingival tissue to identify the content of titanium particles will be warranted for future research, which will pave the way to a better understanding of the association between the content of titanium particles in the gingival tissues and chronic inflammation, peri-implantitis and failure of osseointegration.
The images from the IHC analysis showed positive staining of CD68 and CD3 in the presence of titanium particles, and the expression of the positive staining in FI cases was more than that in SI cases. This could be an indication of the increase in the infiltration of macrophages and lymphocytes into the gingival tissues around FI compared to the SI tissues, which may refer to the role of titanium particles in macrophage polarization. This finding is in agreement with the clinical study by Rao et al. who stated that the failure of orthopedic implants was related to the presence of excessive M1-polarized macrophages around titanium joints [32]. Additionally, Galarraga Vinueza et al. showed in the IHC study that there was an increase in the M1/M2 macrophages ratio in the peri-implant gingival tissues from patients with peri-implantitis [33]. Consequently, this may be considered an important factor related to the failure of the implant and could lead to absorption of the surrounding bone and loosening of the implants.
In the clinical scenario, the whole cells in the peri-implant gingival tissues are usually exposed to titanium wear particles. However, in the in vitro experiments used in the validation study, we targeted only macrophages as the main immune cell included in the response to foreign body reactions. Worthy of consideration is that IHC analysis confirmed the infiltration of macrophages into the peri-implant gingival tissues collected from FI and SI with the presence of titanium particles and the identified related genes by GSEA in both comparisons. Thus, the functional validation was crucial to confirm the impact of titanium particles in the polarization of M1 macrophages which are responsible for the expression of inflammatory cytokines such as Il-18, Il-1β, CCL3, TNF-α, and IL-8.
In the present study, the size of the titanium particles was not possible to recognize in the IHC and H&E images due to the potential modification caused by the microtome cutting technique, which limits the thickness of the samples. Therefore, to replicate the clinical scenarios, we used different concentrations of nano-and micro-sized TiO 2 particles to treat THP-1 monocyte-derived M0 macrophages, thereby validating the findings obtained from the IHC analysis. The results at the protein level confirmed that TiO 2 particles induce the secretion of IL-1β by macrophages. This finding was in agreement with previous evidence demonstrating that IL-1 serves as a pro-inflammatory mediator expressed by macrophages [34]. Additionally, our findings showed an upregulation in the expression of macrophage markers in the tissue samples from the FI compared to SI. It is worth noting that IL-1β was associated with the direct differentiation of pre-osteoclasts into osteoclasts, which potentially influenced the failure of dental implant osseointegration [35]. Furthermore, the activation of NF-κB involves numerous molecules, including IL-1β and TNF-α. Excitingly, the IL-1β promoter contains a functional NF-κB binding site, and this indicates that IL-1 positively autoregulates its synthesis as it acts as a potential inducer of NF-κB binding activity. Accordingly, the IL-1β gene may be considered an important additional member of the cytokine gene family that is regulated in part by the NF-NF-κB/rel family of transcription factors [36].
The immune response to titanium particles is a complex process influenced by the interaction between numerous immune cells. These cells are exposed to the particles and produce different cytokines, which in turn induce the activation of pro-inflammatory macrophages. In our present study, there was upregulation of these pro-inflammatory cytokines by macrophages, indicating the potential link between the release of titanium particles and the failure of dental implants. This finding is in line with previous research demonstrating that IL-1β and inflammatory cytokines impact the activation of osteoclast and are associated with the downregulation of type 1 collagen. Consequently, the IL-1β cytokine is implicated in the resorption of bone and peri-implantitis [37]. TNF-α plays a crucial role in organizing the inflammatory response and the infiltration of immune cells. While no significant alterations were observed in the protein level of TNF-α following various treatments of TiO 2 , it is broadly recognized that TNF-α, in conjunction with IL-1β provokes downstream signaling pathways, such as mitogen-activated protein kinases (MAPKs) and the transcription factor NF-κB. These pathways eventually lead to the release of additional pro-inflammatory chemokines and cytokines [38].
CCL3 plays a crucial role in the recruitment and activation of macrophages, especially at sites of infection or inflammation. Its release stimulates other innate immune cells such as neutrophils, lymphocytes and dendritic cells. Moreover, CCL3 has been identified as a regulator of acute and chronic inflammatory responses, and it acts as a potent activator of osteoclasts, leading to bone resorption [39]. As previously mentioned, TNF-α and IL-1β were reported to control the transcription and expression of CCL3 through the MAPK and NF-κB pathways [21]. Our study observed an upregulation of IL-8 expression in response to the treatment of M0 macrophages with TiO 2 , which aligns with previous findings demonstrating upregulation of IL-8, IL-1β and TNF-α production macrophages [34]. These findings are in agreement with previous studies indicating that excessive IL-8 production can affect the inflammatory response, leading to delayed tissue healing and impairment in endothelial tube growth [40]. The expression of IL-18 at the protein level was found to be increased following the treatment of M0 macrophages with TiO 2 particles. Importantly, IL-18 is involved in the priming of other immune cells such as NKT cells, leading to the expression of TNF-α, which is essential for the differentiation of osteoclasts [41].
The collection of gingival tissue samples from SI is not feasible since the explantation of well-osseointegrated implants in humans to get gingival tissue samples is a rarity to happen and is an exception [42]. Hence, the gingival tissue that is covering the SI was resected during the routine uncovering procedure stage to place the healing abutment (3-5 months after implant placement) and used as SI gingival tissue samples. Consequently, the most common reason for titanium particles' release into gingival tissue around SI could be the friction contact between the implant and alveolar bone during implant placement [43]. Those implants were considered as SI as there was successful osseointegration and no signs of peri-implantitis. The peri-implant gingival tissue samples from the FI were collected 1-2 years after the connection of the crown to the implant body; thus, the micro-movement/friction between the implant and its crown might be the reason for the development of a micro-gap at an implant-crown interface that could be considered as the source for titanium particles [44,45]. Worth noting, all implants included in the study were of the same model surface characteristics. In the present study, iatrogenic contamination of metal elements into the peri-implant gingival tissues was not probable as meticulous care was taken during the harvesting of the tissue samples to prevent the accidental release of titanium. This was in accordance with previous studies where the tissue samples were collected prior to the removal of the implants to prevent contamination with metal from corrosion happening at the time of implant removal [46].
The findings of the present study will facilitate our understanding of immune cell population-specific molecular events deriving the peri-implant inflammation in response to titanium particles. Furthermore, the activated inflammatory mediators may be the key to the imbalance in osteoblast-osteoclast activity, failure of osseointegration and dental implant treatment which paves the way for discovering potential molecular mechanisms underlying both diseases, diagnostic biomarkers and novel therapeutic perspectives (Figure 9). Despite that, this study has limitations. For the in vitro validation, THP-1 monocyte-derived macrophage cell lines were used instead of those derived from the patients due to the limited source of patient samples. Moreover, the small sample size was a limitation in this study, but we were able to overcome this limitation by using different bioinformatics tools to identify significant differentially expressed genes and confirm the differentially activated molecular pathways and biological processes in each comparison of FI, SI and NI groups. collected 1-2 years after the connection of the crown to the implant body; thus, the micromovement/friction between the implant and its crown might be the reason for the development of a micro-gap at an implant-crown interface that could be considered as the source for titanium particles [44,45]. Worth noting, all implants included in the study were of the same model surface characteristics. In the present study, iatrogenic contamination of metal elements into the peri-implant gingival tissues was not probable as meticulous care was taken during the harvesting of the tissue samples to prevent the accidental release of titanium. This was in accordance with previous studies where the tissue samples were collected prior to the removal of the implants to prevent contamination with metal from corrosion happening at the time of implant removal [46]. The findings of the present study will facilitate our understanding of immune cell population-specific molecular events deriving the peri-implant inflammation in response to titanium particles. Furthermore, the activated inflammatory mediators may be the key to the imbalance in osteoblast-osteoclast activity, failure of osseointegration and dental implant treatment which paves the way for discovering potential molecular mechanisms underlying both diseases, diagnostic biomarkers and novel therapeutic perspectives (Figure 9). Despite that, this study has limitations. For the in vitro validation, THP-1 monocyte-derived macrophage cell lines were used instead of those derived from the patients due to the limited source of patient samples. Moreover, the small sample size was a limitation in this study, but we were able to overcome this limitation by using different bioinformatics tools to identify significant differentially expressed genes and confirm the differentially activated molecular pathways and biological processes in each comparison of FI, SI and NI groups.

Materials and Methods
The study was approved by the Research Ethics Committee at the University of Sharjah (REC-19-10-30-01). All patients who volunteered as subjects in this study were informed about the study conditions, asked for their participation, and given written consent for the surgical procedure. Subjects were divided into three groups: Failed Implants (FI) group, Successful Implants (SI) group and No Implants (NI) group. The study methodology is detailed in (Figure 10).

Materials and Methods
The study was approved by the Research Ethics Committee at the University of Sharjah (REC-19-10-30-01). All patients who volunteered as subjects in this study were informed about the study conditions, asked for their participation, and given written consent for the surgical procedure. Subjects were divided into three groups: Failed Implants (FI) group, Successful Implants (SI) group and No Implants (NI) group. The study methodology is detailed in (Figure 10).

Patients' Selection Criteria
The inclusion criteria were (i) patients with an implant-supported fixed or removable prosthesis suffering from peri-implantitis and indicated for explantation (removal of the implant) [2]; these were considered in the FI group. Peri-implantitis was defined as the existence of signs of peri-implantitis represented by clinical and radiographical findings, such as bleeding on probing, probing depth beyond 5 mm and radiographic bone loss of more than 3 mm around the dental implant with or without cortical bone perforations, that is related to the mobility and non-restorability of the implant [2]. The inclusion criteria also consisted of (ii) patients with dental implants indicated for second-stage surgery and healing-abutment fixation, without any signs of peri-implantitis; these were considered in the SI group. Lastly, the inclusion criteria also included (iii) patients with partially impacted wisdom teeth without pericoronitis, who were considered in the NI group. The exclusion criteria were patients younger than 18 years and above 60 years, patients with systemic disease, smoking and alcohol consumption, and patients who had previous periodontal therapy of the dental implant (to avoid iatrogenic causes of titanium release into the peri-implant tissues). Moreover, patients with periodontal disease of the teeth adjacent to the implants were excluded from the study.

Collection of Gingival Tissue Samples
The peri-implant gingival tissue samples were collected from patients who attended the University Dental Hospital Sharjah for dental implant treatment or extraction of wisdom teeth. All patients went through clinical examination to be sure that the teeth adjacent to the implants are without periodontitis, the gingival index is 0, the plaque index is 0, and the probing depth is about 1-2 mm; besides, the dental history showed no history of periodontitis. A total of 20 patients were selected based on the inclusion and exclusion criteria of the study. The age and gender distribution of the patients included in the study are summarized in the Supplementary Table S1. Peri-implant gingival tissue samples were collected from areas adjacent to the implants and contained almost a similar content of both epithelium and connective tissue. The peri-coronal gingival operculum tissues from impacted wisdom teeth sites were collected as well.
The gingival tissue samples of the FI group were taken from eleven patients with peri-implantitis. The implant was indicated for explantation (1-2 years after implant loading). Under local anesthesia, a circular incision with releasing incisions mesial and distal

Patients' Selection Criteria
The inclusion criteria were (i) patients with an implant-supported fixed or removable prosthesis suffering from peri-implantitis and indicated for explantation (removal of the implant) [2]; these were considered in the FI group. Peri-implantitis was defined as the existence of signs of peri-implantitis represented by clinical and radiographical findings, such as bleeding on probing, probing depth beyond 5 mm and radiographic bone loss of more than 3 mm around the dental implant with or without cortical bone perforations, that is related to the mobility and non-restorability of the implant [2]. The inclusion criteria also consisted of (ii) patients with dental implants indicated for second-stage surgery and healing-abutment fixation, without any signs of peri-implantitis; these were considered in the SI group. Lastly, the inclusion criteria also included (iii) patients with partially impacted wisdom teeth without pericoronitis, who were considered in the NI group. The exclusion criteria were patients younger than 18 years and above 60 years, patients with systemic disease, smoking and alcohol consumption, and patients who had previous periodontal therapy of the dental implant (to avoid iatrogenic causes of titanium release into the periimplant tissues). Moreover, patients with periodontal disease of the teeth adjacent to the implants were excluded from the study.

Collection of Gingival Tissue Samples
The peri-implant gingival tissue samples were collected from patients who attended the University Dental Hospital Sharjah for dental implant treatment or extraction of wisdom teeth. All patients went through clinical examination to be sure that the teeth adjacent to the implants are without periodontitis, the gingival index is 0, the plaque index is 0, and the probing depth is about 1-2 mm; besides, the dental history showed no history of periodontitis. A total of 20 patients were selected based on the inclusion and exclusion criteria of the study. The age and gender distribution of the patients included in the study are summarized in the Supplementary Table S1. Peri-implant gingival tissue samples were collected from areas adjacent to the implants and contained almost a similar content of both epithelium and connective tissue. The peri-coronal gingival operculum tissues from impacted wisdom teeth sites were collected as well.
The gingival tissue samples of the FI group were taken from eleven patients with periimplantitis. The implant was indicated for explantation (1-2 years after implant loading). Under local anesthesia, a circular incision with releasing incisions mesial and distal of the implant was performed in the gingival tissues using a 15C scalpel (Sigma Aldrich, St. Louis, MO, USA). Then, 4 mm from the crest of the gingival tissue around the FI was excised, followed by the surgical removal of the dental implant. The gingival tissue samples of the SI group were taken from six patients who attended the clinic for second-stage surgery to uncover the implants and placement of healing abutments (3-5 months after the implants' placement). Under local anesthesia, the regular protocol for uncovering dental implants was followed using a hemostatic clamp and 15C scalpel, and 4 mm of the gingival tissue that was excised to uncover the implant was used as an SI tissue sample. The gingival tissue samples of the NI group were taken from three patients who attended the implant clinic for the surgical extraction of wisdom teeth. The regular surgical protocol for the surgical extraction of a wisdom tooth was performed, followed by the excision of 4 mm of healthy peri-coronal gingival tissue at the gingival crest of the extracted tooth.

Processing of Gingival Tissue Samples and Paraffin-Embedded Blocks' Preparation
Gingival tissue samples from the three groups were fixed in 10% neutral buffered formalin (Thermo Scientific, Waltham, MA, USA). After 2 days of fixation, tissues were processed by an Excelsior AS tissue processor (Thermo Scientific, Waltham, MA, USA) using a routine overnight protocol and embedded as paraffin blocks (TEC 2900, Pantigliate MI, Italy). Plain tissue slides (Medix, Newbury Park, CA, USA) were prepared in 3-5 µm sections with the HM 355S Automatic Microtome (Thermo Scientific, Waltham, MA, USA), deparaffinized and then rehydrated with graded ethanol (Sigma-Aldrich, St. Louis, MO, USA).

RNA Sequencing
The paraffin blocks of the FI, SI and NI samples were sectioned into slides at 5µm thickness using the HM 355S Automatic Microtome (Thermo Scientific). Then, the tissues were macrodissected from the slides and collected in Eppendorf tubes for RNA extraction as previously described [47]. The RNA was extracted using the RNA RecoverAll kit, followed by the addition of Turbo DNAse and further cleaned and concentrated using the Zymo RNA mini-Kit according to the manufacturer's instructions (Zymo Research, Irvine, CA, USA). Purified RNA was used for whole transcriptome profiling analysis using targeted RNA-Seq with the Ion AmpliSeq Whole Transcriptome human gene expression kit (Thermo Fisher Scientific), which is designed to target over 21,000 specific human RNA transcripts using a high-throughput multiplexed method as previously described [48]. RNA sequencing was performed using an Ion 540 Chip on an Ion S5 XL Semiconductor sequencer (Life Technologies, Carlsbad, CA, USA).

RNA-Seq Data Analysis
The RNA-seq data were analyzed using the Ion Torrent Software Suite version 5.4. The Torrent Mapping Alignment Program (TMAP) was used to align the raw sequencing reads to the GRCh37/hg19 Human reference genome. The raw read counts of the targeted genes were normalized using the Fragments Per Kilobase Million (FPKM) normalization method. The number of expressed transcripts was evaluated in the subsequent quality control step. A minimum of 10 million total reads per sample was recommended for further analysis. Next, the DESeq2 R/Bioconductor package (https://bioconductor.org/packages/ release/bioc/html/DESeq2.html) (accessed on 21 September 2021) was used to identify the differentially expressed genes (DEGs) in each comparison of patients with FI and SI against patients with NI. DESeq2, the negative binomial generalized linear model fitting was applied and the Wald test was performed to analyze the abundance data. Read count normalization was carried out using the regularized logarithm (rlog) method provided in DESeq2. Genes with low read counts were filtered out; only genes with more than ten normalized read counts were considered for further analysis. DEGs were selected with Fold change (FC) > 2 at a significance level of the adjusted p-value (FDR) < 0.05.

Clustering and Functional Enrichment Analysis of the Differentially Expressed Genes
In order to confirm the classification of the studied patients, unsupervised clustering into distinct subclusters was performed based on the expression patterns of the DEGs lists between patients with FI and SI against patients with NI independent of their clinicpathological features or intrinsic molecular subtypes. This was carried out using Euclidean distance measure and Ward linkage [49]. Moreover, the obtained upregulated and downregulated gene lists between patients with FI and SI against patients with NI were annotated and analyzed using the Metascape tool [50]. The latter is relying largely on well-annotated pathways derived from Gene Ontology (GO), Reactome and KEGG databases. Indeed, the DEGs of the two comparisons of FI and SI against NI were intersected to identify the commonly differentially regulated genes between the failed and successful implants using InteractiVenn [51]. The common genes were further analyzed for pathways annotation.

Gene Set Enrichment Analysis
The significant DEGs lists obtained were further analyzed to identify the enriched pathways and the related genes in the peri-implant inflammation in response to titanium particles in comparison to normal tissues using Gene Set Enrichment Analysis (GSEA). A GSEA search was performed in large sets of cellular pathways and biological processes obtained across Immunome, Immunologic signature gene sets C7 and ontology gene sets C5 from the GSEA database (https://www.gsea-msigdb.org) (accessed on 30 November 2021). The significantly enriched pathways with p-value (p) < 0.05 were further explored to identify the top relevant genes in the respective sets.

In Silico Prediction of the Immune Cells' Subsets in the Peri-Implant Gingival Tissues
In silico CIBERSORT, a digital cytometry that can identify cell type composition and expression from bulk tissues, was carried out on the DEGs gene lists to identify the composition of immune cells in the peri-implant gingival tissues of each subgroup of FI and SI patients compared to the NI group. As references, we used gene lists for the immune cells including naïve CD8 T cells, naïve CD4 T cells, alternatively activated macrophages, classically activated macrophages, regulatory T cells, T helper cells, natural killer cells, and dendritic cells [52].

Histological and Cytological Staining
To validate the transcriptome findings concerning the infiltration of titanium particles and immune cells represented by macrophages and lymphocytes, immunohistochemistry (IHC) and Hematoxylin-Eosin (H&E) analyses were performed. The peri-implant gingival tissue samples from the same selected FI, SI and NI patients were used in the IHC and H&E assessment. The slides were subjected to antigen retrieval of CD3 and CD68 markers ( Table 4) by steaming them in an appropriate retrieval buffer for 20 min, followed by blocking with 5% bovine serum albumin (BSA) for 30 min. Primary antibody incubation was performed by overnight incubation at 4 • C with desired antibody dilution in 1% BSA solution and followed by 1 h incubation with the horseradish peroxidase (HPR)labeled suitable secondary antibody dilution in 1% BSA solution. The color was developed with 3,3 -Diaminobenzidine (DAB) (Abcam, Cambridge, UK) and counterstained with Shandon Harris Hematoxylin (Thermo Scientific). For the two antibodies, the EDTA buffer with a pH of 8.0 for heat-induced antigen retrieval during immunohistochemistry (IHC) and the Ventana autostainer (Ventana-Benchmark XT, Roche Inc., Branchburg, NJ, USA) were used. Counterstaining was performed using Gill's hematoxylin (GHS1128, Sigma-Aldrich, USA) and slides were rinsed in a container with cold running water. To confirm the specificity of the antibody sections with human samples, tonsil tissues were used as appropriate positive controls. The images were captured with an Olympus DP74 microscope digital camera attached to a BX43 microscope (Olympus Life Sciences, Tokyo, Japan), at various magnification powers. H&E and IHC staining was validated by two experienced pathologists (IT and RAH). Given the results of previous steps, certain cytokines of interest were explored in THP-1 monocyte-derived M0 macrophage cell lines treated with TiO 2 particles.

Protein-Protein Interaction (PPI) Network Analysis
The protein-protein interaction network analysis permits the visualization of the functional relationships between proteins [53]. To identify potential interactions between the significantly altered cytokines in the peri-implant gingival tissues, the STRING tool was employed to construct the PPI network. Several interaction sources including experiments, databases, physical interactions and co-expression at a minimal confidence interaction score of 0.4 were applied to construct the PPI network between the input proteins and their neighbors at ten more layers.

Statistical Analysis
The Kruskal-Wallis test was used for IHC analysis between the three groups of FI, SI and NI. The post hoc Tukey test and one-way ANOVA were used for analyzing ELISA. For the Luminex assay, the unpaired Student's t-test was used to compare each of the two groups. p < 0.05 was statistically significant.

Conclusions
This study aimed to investigate the immune microenvironment that is involved in the peri-implant inflammation induced by titanium particles. Accordingly, the peri-implant gingival tissue samples were collected from patients with failed implants, successful implants, and no implants. The analysis of whole transcriptome data found a significant enrichment of transcripts related to immune response signaling pathways, particularly T cell-mediated immunity and M1/M2 macrophage polarization. Moreover, histological staining (H&E and IHC) confirmed the presence of lymphocytes, macrophages, and titanium particles in the peri-implant gingival tissues of both failed and successful implants. Furthermore, the functional validation showed a remarkable increase in the expression levels of IL-18, IL-1β, IL-8 and CCL3 by THP-1 monocyte-derived macrophages when exposed to different concentrations of titanium dioxide particles. These findings will contribute to highlighting the role of titanium particles in the development of peri-implant inflammation and failure of dental implant treatment. Thus, further research is warranted to investigate the early detection of titanium particle infiltration into peri-implant gingival tissues and understand the changes in immune cell composition in the affected peri-implant environment.