Effect of Enterococcus faecalis OG1RF on human calvarial osteoblast apoptosis

Persistent apical periodontitis (PAP) is an endodontic inflammatory condition characterised by bone destruction in the periapical region [1]. Inflammatory bone destruction is generally associated with an imbalance between bone degradation and formation [2]. In addition to enhanced osteoclastogenesis and subsequential bone resorption, increased osteoblast apoptosis decreases the synthesis of bone matrix and deposition of new bone, resulting in bone loss in inflammatory bone diseases [3]. Bacteria and their by-products are considered responsible for osteoblast apoptosis in osteomyelitis, periodontitis, and PAP [4‐6]. Evidence has highlighted that Enterococcus faecalis is one of the causative microorganisms of PAP [7, 8]. The detection of E. faecalis in infected root canals has been associated with the occurrence of periapical lesions larger than 3 mm, indicating the role of E. faecalis in PAP and endodontic treatment failure [9]. Recent studies have found that E. faecalis induces apoptosis in mouse osteoblast-like MC3T3 cells and human osteosarcoma MG63 cells [10‐12]. However, there may be a difference between the effect of E. faecalis infection in primary osteoblasts and osteoblast-like cells from different origins [10, 13]. The apoptosis-inducing effect of E. faecalis and its mechanisms in human primary osteoblasts remain elusive.

Bacteria may induce apoptosis in osteoblasts via intrinsic (caspase-9 dependent) and/or extrinsic (caspase-8 dependent) pathway(s) through the activation of executioner caspase-3 [5, 11]. Apoptosis-related genes, such as tumour necrosis factor receptor superfamily (TNFRSF) and caspase activity inhibitors (e.g., BIRC6, NAIP, and XIAP), may also play important roles in infection-induced apoptosis [14]. Notably, the B-cell lymphoma-2 (BCL-2) family plays a key role in the intrinsic apoptotic pathway [15]. The sophisticated modulation of the balance between anti-apoptotic and pro-apoptotic members of the BCL-2 family can determine cell fate decisions of life or death, giving rise to numerous diseases, such as cancer and autoimmune diseases [16, 17]. Studies have reported that anti-apoptotic BCL-2 and BCL2L1 (also known as BCLXL), and pro-apoptotic BID, BIM, and BAX may play roles in the pathogenesis of periodontitis [5, 18, 19]. BCL-2, BCL2L1, BAD, and BAX may also be involved in the molecular mechanisms linking periodontal disease with cancer [20]. However, the putative role of the BCL-2 family and other apoptosis-related genes in E. faecalis-infected primary osteoblasts remains inconclusive.

Here, we investigated the apoptotic effect of E. faecalis OG1RF on primary human calvarial osteoblasts. The strain OG1RF is a rifampicin- and fusidic acid-resistant derivative of a caries-related strain OG1 isolated in 1975 [21, 22]. Because of its oral rather than urinary origin (e.g., ATCC 29,212), it is commonly used to investigate the pathogenicity of E. faecalis in apical periodontitis [23, 24]. We examined the mRNA expression of 84 apoptosis-related genes, and explored the role of anti-apoptotic and pro-apoptotic members of the BCL-2 family in E. faecalis-induced osteoblast apoptosis.

PAP, a common endodontic disease induced by bacterial infection and consequent inflammatory responses, is characterised by persistent bone destruction in the periapical region [1]. Recent studies have shown that PAP lesion is associated with osteoblast apoptosis, suggesting a role of osteoblast apoptosis in the pathogenesis of PAP [27, 28]. E. faecalis, as one of the primary pathogens in PAP, has been reported to be highly associated with endodontic infection, promoting the disruption of bone homeostasis [9]. Hence, it is crucial to examine the role of E. faecalis on osteoblast apoptosis in PAP. Our previous study found that E. faecalis strains from the root canals of teeth with PAP trigger apoptosis in mouse MC3TE-E1 and human MG63 cell lines [10, 12]. However, the effect of E. faecalis on apoptosis in human primary osteoblasts and its mechanisms remain unclear.

In the present study, E. faecalis infection increased the number of TUNEL-positive cells and apoptotic cells in the early and late phases, which was also observed in our previous study on MC3T3-E1 and MG63 cells infected with E. faecalis [10, 12]. We also noticed that necrotic cells of which the cell membrane integrity is damaged may also be stained as Annexin V-FITC/PI double-positive, as previous studies reported [29, 30]. In this study, we observed that E. faecalis infection with a higher MOI can trigger more late apoptotic cells (and necrotic cells, if any), whereas a lower MOI infection induced more early apoptotic cells, suggesting that E. faecalis infection causes osteoblast apoptosis (and necrosis, if any) in an MOI-dependent way. Additionally, the mRNA expression level and activity of caspase-3/-8/-9 were elevated in the infected osteoblast group, indicating that both intrinsic and extrinsic apoptosis were activated in the osteoblasts infected with E. faecalis OG1RF. More precisely, the significantly decreased ΔΨm and increased expression of apoptotic peptidase activating factor 1 (APAF1) indicated the activation of intrinsic apoptosis, whereas the enhanced expression of caspase-10, which is a homologue of caspase-8, indicated the activation of the extrinsic pathway [31]. Moreover, the upregulation of TNFRSF1B, TNFRSF8, TNFRSF9, and TNFRSF10 expression, as well as elevated levels of TNF, suggested apoptotic signal transition through the TNF family membrane receptors, which was also detected in tumour cells treated with chemotherapeutic drugs [32, 33]. Furthermore, the expression of caspase inhibitors NAIP, BIRC6, and XIAP was also downregulated, indicating the promotion of apoptosis. Taken all together, these data demonstrated that both intrinsic and extrinsic apoptosis were involved in human primary osteoblasts infected with E. faecalis.

To further explore the activation of the intrinsic apoptotic pathway, we analysed the expression profile of the BCL-2 family in E. faecalis-infected osteoblasts. The BCL-2 family has two subfamilies: anti-apoptotic and pro-apoptotic [34]. The anti-apoptotic members BCL-2, BCL2L1, BCL2L2, BCL2L10, BCL2A1, BCL2L12, and MCL1 exert their function via the sequestration of the pro-apoptotic members. The pro-apoptotic subfamily is further categorised into multi-domain executioners (BAK, BAX, and BOK) and the BH3-only proteins that possess only the BCL-2 homology (BH) 3 domain. The BH3-only activators, BCL2L11, BID, PUMA, and MULE, interact with both pro-apoptotic executioners and anti-apoptotic members to trigger apoptosis, whereas BH3-only sensitisers (BAD, BMF, HRK, NOXA, BIK, and Beclin-1) displace the BH3-only activators and executioners from the anti-apoptotic protein heteromeric complex to promote apoptosis [35]. The interplay between the BCL-2 family members triggers the multi-domain executioner oligomerisation on the mitochondrial membrane, resulting in mitochondrial outer membrane permeabilization (MOMP) and subsequent activation of the caspase cascade [36]. Here, we found that the mRNA expression of anti-apoptotic BCL2 was downregulated, whereas that of pro-apoptotic BCL2L11, HRK, BIK, BMF, NOXA, and BECN1 was upregulated in the infected cells. Downregulated BCL2 expression was also detected in MC3T3-E1 cells infected with E. faecalis [10]. Moreover, it has also been reported that the activator BCL2L11 and sensitisers Beclin-1 and BIK participate in osteoblast-like cell apoptosis induced by glucocorticoids, sodium fluoride, oxidative stress, and chemotherapeutic drugs [37‐40]. Furthermore, decreased expression of sensitisers BMF and NOXA may increase osteoclast survival to promote bone loss [41, 42]. However, the role of harakiri (HRK), a novel regulator of cell death, in bone homeostasis remains unclear. Evidence has shown that the inactivation of HRK in prostate cancer constitutes a critical component in decreased apoptosis of tumour cells, whereas HRK-mediated mitochondrial dysfunction contributes to enhanced apoptosis in human malignancies [43]. However, in lipopolysaccharide-stimulated osteoclasts, HRK expression is not altered [41]. In this study, we reported an increase in HRK expression in the infected osteoblasts, suggesting its role in bone remodelling in PAP. In addition, interestingly, we observed that the expression of anti-apoptotic BCL2A1 was significantly higher in the infected osteoblasts than in the control sample. The pro-survival potential of BCL2A1 is reported to be associated with its interaction with BCL2L11, BIK, HRK, NOXA, BID, and PUMA [44]. The level of BCL2A1 is enhanced in different types of cancer cells, resulting in tumour progression and chemotherapy resistance [45]. Porphyromonas gingivalis infection also induces the upregulation of BCL2A1 in epithelial cells [46]. In this study, as the mRNA levels of BCL2L11, BIK, HRK, and NOXA were increased, we inferred that the elevated BCL2A1 expression may act as a negative feedback factor that interacts with these four aforementioned BH3-only proteins to protect the infected osteoblasts from apoptosis. Evidently, our data suggested that these BCL-2 family members play important roles in E. faecalis-induced osteoblast apoptosis. Further studies are needed to elucidate the exact role and mechanism of the interaction of BCL-2 family members in the pathogenesis of PAP and their possible role(s) in the relationship of PAP with system diseases.

In conclusion, our findings demonstrated that E. faecalis OG1RF induced apoptosis in human primary osteoblasts, and both the extrinsic and intrinsic pathways were involved. Anti-apoptotic BCL-2 and BCL2A1, and pro-apoptotic BCL2L11, HRK, BIK, BMF, NOXA, and Beclin-1 were also involved in osteoblast apoptosis induced by E. faecalis (Fig. 7). These results suggest that BCL-2 family members act as novel regulators of osteoblast apoptosis induced by E. faecalis and are, thus, potential therapeutic targets for PAP treatment.