This study was motivated by the overall question of how damaged necrotic cells affect the paracrine environment using osteoclastogenesis and chemokine expression as a bioassay. Bone resorption occurs early upon implant insertion to remove the necrotic bone where it is mainly the dying osteocytes driving this catabolic event (2, 4). Bone resorption linked to periodontitis and peri-implantitis is a consequence of chronic inflammation (31); but may also be affected by clinical interventions that harm the periodontal cell – the gingival fibroblasts, the oral epithelial cells, and the macrophages accumulating in the inflamed periodontal soft tissue (12). Considering this clinical scenario, we have recently reported that lysates prepared from murine bone marrow stromal cell line ST2 attenuate in vitro osteoclastogenesis (19). We now extend this research by implementing cells representing the oral soft tissue and an osteocytic cell line. The major finding of this research was that (i) lysates prepared from gingival fibroblasts, oral epithelial cells, and IDG-SW3 cells consistently reduced osteoclastogenesis in vitro. Lysates prepared by RAW264.7 macrophages are somewhat less potent but support the overall conclusion namely that damaged necrotic oral cells possess a paracrine activity capable of blocking osteoclastogenesis in vitro. Another observation was that (ii) exclusively lysates from HSC2 cells and TR146 were capable of provoking a robust increase of inflammatory cues being expressed in murine ST2 cells. Thus, there are two possible modes of action – on is the direct effect by blocking of osteoclastogenesis in bone marrow cultures and the other is the indirect by changing the paracrine environment of stromal cells.
If we relate our first main observation to other studies we find support for our observations coming from lysates prepared from HEK-293 cells releasing clusterin (20), and clusterin can reduce osteoclastogenesis in vitro (21). In contrast to our findings, however, supernatants prepared from serum-starved necrotic IDG-SW3 cells increased osteoclastogenesis via a Mincle-dependent mechanism (9). There is thus a discrepancy between supernatants harvested from serum-starved necrotic cells and the total necrotic cell lysates, the first stimulating while the latter inhibiting in vitro osteoclastogenesis, respectively. Surprisingly, and even though the environmental changes are relevant in osteoclastogenesis, for instance the bidirectional ephrinB2-EphB4 signaling (32), studies on how cell lysates regulate osteoclastogenesis are rare. We can only speculate about the molecular mechanism responsible for the decrease of osteoclastogenesis in murine bone marrow cultures. One possible pathway is that cell lysates contain IFN-β produced by osteocytes and known to inhibit osteoclastogenesis (33). Most of the inhibitor cytokines, however, are produced by lymphocytes, macrophages and dendritic cells (34). Alternatively, our necrotic lysates may contain proteases cleaving the osteoclast-induction cocktail of RANKL, M-CSF and TGF-β1; however, usually proteases release membrane bound RANKL (35) and activate TGF-β1 (36). Another option would be the lysates contain molecules adsorbing and thereby neutralizing the osteoclast-induction cocktail, for instance collagen can adsorb the TGF-β1 (37) but overall the mechanism remains unclear.
Damage-associated molecular patterns (DAMPs) provide another interesting line of research. DAMPs are molecules released from damaged or dying cells causing an innate immune response (38). DAMPs include but are not limited to extracellular matrix components (39), S100 and heat shock proteins – all of which bind and active TLRs and other pattern recognition receptors (40) – and importantly, LPS as a major TLR4 agonist greatly inhibits osteoclastogenesis in murine in vitro cultures (41, 42). There is thus a rational to make DAMPs activating TLR4 signaling in bone marrow cells responsible for the blocking of osteoclastogenesis. However, this is unlikely as the respectively lysates reduce LPS-induced inflammatory response of RAW264.7 macrophages in vitro (43). Thus, the molecular mechanisms of how the lysates reduce in vitro osteoclastogenesis remains puzzling. Moreover mastication-induced gingival damage causes oral epithelial cells to produce IL6 expression (44), a major driver of osteoclastogenesis (29).
Our second main observation was that lysates from HSC2 and TR146 greatly induced the expression of CCL2, CCL5, CXCL1, IL1 and IL6 in murine ST2 cells – neither lysates from gingival fibroblasts nor those from IDG-SW3 osteocytic cells were capable of provoking such an increase. Most of the inflammatory cues such as CCL2 (24), CXCL1 (30), IL1 (45) and IL6 (29) are potent induces of osteoclastogenesis; only CCL5 is more of a suppressor for osteoclastogenesis (25). There observations are consistent with our unpublished observations showing that lysates of HSC2 and TR146 cell are stimulating the expression of cytokines and CXCL8 in human gingival fibroblasts (Sordi and Panahipour et al., in print). The question on the underlying molecular mechanism remains unclear and might involve TLR activation, as for instance, the TLR2 agonists Pam3CSK4 can increase cytokine expression on ST2 cells (46). Considering that TLR2 is sensitive to extracellular matrix components, histones, high mobility group box 1 (HMGB1) and other DAMPs (39), it can be speculated that DAMPs released by HSC2 and TR146 maybe have caused the increase in chemokine and cytokine expression. The question remains why not DAMPs from gingival fibroblasts or those from IDG-SW3 have this activity? Thus, our data are preliminary and might help in establishing a fibroblast-based bioassay to screen for DAMP being released from necrotic oral epithelial cells.
The observation that ST2 cell increasingly expressed CCL2 (24), CXCL1 (30), IL1 (47), and IL6 (29), all agonist of osteoclastogenesis and to some extend also CCL5, an antagonist of osteoclastogenesis (25) is maybe less surprising since CCL2, CCL5 and IL6 are expressed in response to TNFα in ST2 cells (46). The bioassay therefore has to be interpreted not strictly related to osteoclastogenesis but more in the direction that lysates can change the paracrine environment of stromal cells; an environment that affects mobilization and differentiation of hematopoietic cells, including neutrophils and macrophages, as well as lymphocytes. For instance, CCL2 and CCL5 are elevated in gingival crevicular fluid in patients with generalized aggressive periodontitis (27) and CCL2 is an increasingly expressed in fibroblasts from periodontitis patients (12). Likewise, CCL2 and CCL5 is elevated in other diseases such as pulmonary sarcoidosis (48) or in the synovial fluid of patients with juvenile rheumatoid arthritis (49). Thus, our findings that HSC2 and TR146 lysates increase CCL2 and CCL5 in ST2 might be linked to the recruitment and activation of monocytes and macrophages. In support of this hypothesis, the enhanced CXCL1 expression point towards a recruitment of neutrophils (50). IL1 and IL6 have pleiotropic pro-inflammatory functions (51, 52), similarly changing the paracrine environment of stromal cells.
This discussion clearly shows the study limitation namely that the findings on osteoclastogenesis but also those related to the chemokine and cytokine expression in ST2 cells are descriptive and we are left with many questions of how to explain these observations on a molecular level – and actually, if the in vitro observation have a relevance in vivo – and as we are interest in dentistry, the clinical impact of the findings in the oral cavity. We are left with the open question how invasive dental treatments affects cell necrosis and if this translates into reducing osteoclastogenesis in vivo, and if the necrotic epithelial cells – here oral squamous carcinoma cells - change the local stromal environment towards a CCL2 and CCL5-mediated migration of monocytes, and CXCL1-mediated migration of neutrophils, also involving the leading inflammatory cytokines IL1 and IL6. Our future research also relates to a possible hierarchy of the expressed genes, for instance the increasingly expressed inflammatory cytokines IL1 and IL6 might be the driver of the CCL2, CCL5 and CXCL1 expression. It can thus not be ruled out that it is IL1, IL6 and possibly other mediators that are increased in ST2 cells in turn drive the chemokine expression. Thus, HSC2 and TR146 lysates might have indirect effects by changing the autocrine environment of ST2 cells causing changes of chemokine expression. Independent if the expressing of chemokines is a consequence of the cytokines or if both are directly triggered by HSC2 and TR146 lysates, our preliminary observation basically supports the fundamental research on how fibroblasts, in our case ST2 cells, in a catabolic environment become inflammatory cells expressing chemokines and cytokines thereby attracting and activating neutrophils, lymphocytes and antigen presenting cells (12), a process that is presumably happening during transient inflammation, before its resolution, in the context of tissue regeneration (53).
Care, however, shout be taken when interpreting findings as the necrotic cell lysate is actually living cells that were sonicated and thus the process of necrosis in these cells is not clearly present. Moreover, the use of HSC2 and TR146 cancer cells is not necessarily reflecting the normal oral epithelial cells; thus, future research should test the activity of the respective lysates in the bioassays we have performed. It would be interesting to understand what cellular components the HSC2 and TR146 cells release, in contrast to the gingival fibroblasts and the other cell lines, that provoke the inflammatory response in the ST2 mouse stromal cells – while all lysates reduce in vitro osteoclastogenesis in mouse bone marrow cultures. Considering these limitations and the pilot nature of the research, our study incites the curiosity to uncover the mechanisms of how necrotic cell lysates affect osteoclastogenesis and the local response of the fibroblastic stromal environment – a research line that is of potential clinical relevance considering the invasive procedures in dental treatments and possibly in other fields of regenerative medicine.