Background
High doses of irradiation therapy are routinely administered to patients for a large number of cancers affecting many organs [
1,
2]. Furthermore, the use of bone allografts necessitates high dose irradiation for sample sterilization [
1,
2]. Under normal circumstances, irradiation passes through a number of tissues where the minimization of doses is vital to the future survival of tissues. One tissue that commonly receives high doses of irradiation therapy in maxillofacial surgery is that of bone. Due to the high volume of bones found in the maxilla, it is common in head and neck procedures to pass high doses of irradiation through bony tissues [
1,
2]. Common doses can range from 60 Gray to 70 Grays and such procedures have been highly successful for the treatment of many head and neck tumors [
3,
4]. Despite successfully treating and managing these tumors, irradiation of bone has reported some drawbacks with a certain percentage of bone losing complete vitality and becoming necrotic [
3,
4]. A group of experts have since recommended guidelines established for preventing the possibility of developing osteoradionecrosis of the jaw at doses exceeding 60 Grays [
3,
4].
Furthermore, irradiation of bone is also commonly found in bone allograft sterilization procedures for bone tissue banking [
5]. Although the full set of doses are commonly kept proprietary information [
5], the release of subsequent growth factors from its content is a trademark commonly observed in demineralized bone allografts which may demonstrate signs of osteoinduction by showing signs of ectopic bone formation in various animal models. Irradiation of bone is also used in extracorporeal irradiation and has been implicates in not only oral and maxillofacial surgery, but also other disciplines such as otolaryngology [
6] and orthopedics [
7].
Marx et al. have been key oral maxillo-facial surgeons responsible for treating and developing treatment guidelines for patients presenting necrotic bone following irradiation [
8‐
10]. While a number of attempts have been investigated to maintain optimal bone viability [
11‐
13], a limited understanding of the cellular events that take place within bone following irradiation would benefit from further investigation.
Due to the widespread use of irradiation for various bone procedures including irradiation for tumors, sterilization of autologous bone transplants and bone tissue banking, it became of interest to our group to further investigate in a bench model the effects of dose-dependent irradiation on bone cell viability and release of growth factors. While it is known that irradiation is unfavorable for bone remodeling [
14] and that an accumulation of evidence has been accumulating demonstrating that irradiation consequently affects microvasculature of bone-tissues [
15], a detailed investigation of the in vitro mechanisms was studied to further increase our understanding of bone changes following irradiation. Therefore, the purpose of the present investigation was to determine the effects of irradiation on bone cell viability following irradiation at various single-doses including 0, 7.5, 15, 30, 60 and 120 Grays. Thereafter, bone graft morphology and surface proteins were analyzed via scanning electron microscopy and release of growth factors from the bone samples was quantified for TGFβ1, BMP2, VEGF, IL1β and RANKL at 15 min and 4 h following irradiation.
Discussion
The purpose of the present manuscript was to determine in detail the cellular events that occur following irradiation of bone samples at increasing doses. Although clinically doses can range in intensity and duration, it has been suggested in the literature that doses exceeding 60 Gy are more commonly associated with osteoradionecrosis of the jaw [
3,
4]. Furthermore, a number radiation therapy complications have been reported in the literature to date [
21]. Therefore, the aim of this bench top study was to perform an in vitro investigation to further understand the cellular events taking place within bone samples following irradiation.
Interestingly, in the present study very little change in surface proteins was found between all bone samples following irradiation (Figs.
1,
2). Furthermore, it was observed that in all samples treated with irradiation, over an 85 % cell mortality was seen even following as little as 7.5 Gy (Fig.
3). This result was extremely surprising as it was initially thought that such small doses would have very little effect on cellular viability. It must be noted that this reported finding applies solely to an in vitro model and its extrapolation to the clinical reality is limited given that immune cells and regenerative cells would minimize free radicals and cells death when compared to the present in vitro model. Nevertheless, it remains striking that cell death occurred in such high numbers following such low levels of irradiation doses.
Thereafter, samples were investigated for growth factor and cytokine release following irradiation at both 15 min and 4 h post irradiation. The reason for selecting these time points was specifically to investigate the changes in cytokine release after a short time interval following initial irradiation (15 min) and also to determine how release of cytokines was affected after a later time point (4 h) following cell death from irradiation. The bone samples were first assessed for VEGF protein release. There was a marked and significant decrease in release of VEGF as early as 15 min post-irradiation (Fig.
4a). VEGF is one of the key growth factors responsible for angiogenesis and the effects of irradiation demonstrate the harsh effects on this potent growth factor. The results from this study further demonstrate and support the groups of clinical experts working with oxygen delivered in hyperbaric pressures for improved angiogenesis before and/or after irradiation therapy [
11‐
13]. Although the clinical efficacy of using such treatment has come under speculation in recent years [
22], the rational behind improving the angiogenic properties of bone for irradiated patients is logical and can further be explained by the present study as VEGF was the growth factor most notably down-regulated following irradiation (Fig.
4a). Future research aimed at addressing the in vivo release of VEGF from bone following irradiation may further add valuable data supporting the findings from the present study that irradiation has a significantly pronounced effect on VEGF protein release following irradiation at various doses.
Interestingly, it has been shown in a recent report that osteocytes are major contributors to the release of VEGF in vivo [
23]. Furthermore, a subsequent in vitro report has demonstrated that proton irradiation with as little as 2 Gy is enough to suppress angiogenic genes in certain cell types [
24]. Taken together with the results from the present study, it becomes extremely clinically relevant to further design strategies to limit the down-regulation of pro-angiogenic genes.
A second growth factor that has been extensively investigated by our group with respect to growth factor released from bone samples is TGFβ1 [
17,
25‐
27]. In several studies analyzing the released protein content from bone (termed bone conditioned medium (BCM)), it was found that one of the likely paracrine factors displayed in bone remodeling is that of TGFβ1 [
25]. It was found that by inhibiting TGFβ1 pathway, a 5-fold decrease in oral fibroblast activity was observed, thus confirming that much of the preliminary remodeling process caused by bone is likely governed by TGFβ1 signaling [
27]. Thus, in the present study, a 2 fold significant decrease in TGFβ1 protein expression released from the bone samples following irradiation is likely to have a significant effect on future bone remodeling following irradiation. Furthermore, the combinatory reduction of both TGFβ1 and VEGF is hypothesized to pose major bone remodeling challenges further illustrating the necessary regenerative procedure to counteract these major drawbacks.
It was surprisingly observed in the present study that irradiation had virtually no effect on BMP2 protein expression (Fig.
4c). It may therefore be concluded that dying cells that are found within the bone matrix are not target cells for release of osteoinductive growth factors such as BMP2. Interestingly, it is commonly reported in the literature that certain forms of demineralized freeze-dried bone allografts (DFDBA) are osteoinductive whereas most if not all non-demineralized samples are non-osteoinductive [
28]. Therefore, it may be concluded that within the present investigation, BMP2 is not a key player in bone remodeling of irradiated bone and likely BMP2 expression is only upregulated once the bone samples are resorbed by osteoclasts and BMP2 is thereafter released from content coming from within the bone matrix.
The results obtained with RANKL protein quantification also generated statistically significant differences at 4 h between control and irradiated bone. RANKL protein expression was up to 2 fold lower than certain irradiated bone samples. It must once again be highlighted that the largest percentage of cells found within bone samples are osteocytes, which account for approximately 90 % of all bone cells. It has previously been demonstrated that damaged or lack of osteocytes is routinely associated with reduced remodeling [
29,
30] and dying osteocytes are able to signal for bone resorption by attracting osteoclasts through the release of RANKL [
31,
32]. The in vitro findings in the present experiment further demonstrate and confirm the ability for bone samples undergoing high and fast rates of cell death are able to release osteoclast differentiation marker RANKL to the surrounding environment following cell death.
It must also be considered that one of the study limitation of the present study were that all bone samples received irradiation directly to bone in an in vitro model which might not simulate a clinical situation. An in vivo model would necessitate that all irradiation to bone would ultimately pass through overlying tissue which include epithelial, connective tissues, glands, muscles and a combination of them therefore absorbing some of the irradiation prior to bone. Furthermore, immune cells would counteract some of the free radicals produced by irradiation likely contributing to apoptosis of various cell types. Future investigation characterizing this interplay between these various cell types would further contribute to our understanding of bone remodeling following irradiation. Furthermore, in the present model, the bone periosteum was removed which might have a significant influence of the final outcome. Many of the progenitor cells found within bone are located within the periosteum and this complex interaction between periosteum and bone requires better understanding to better implement future regenerative procedures.
In context with some of the known literature, it has been debated for some years the influence of osteoradionecrosis on bone cell interactions. The original proposed and well-accepted ‘three-H concept’ of hypoxia, hypocellularity and hypovascularity as defined by Marx brings into question all the key elements of bone viability [
33]. In light of the present findings, it becomes apparent that one of the key components downregulated after irradiation is that of VEGF thus giving evidence for a hypovascular and hypoxic environment. We also demonstrate the drastic changes in cell viability following only 7.5 Gy of irradiated bone. Although these doses would be significantly different in a human model and that the present in vitro model can only be vaguely extrapolated to a clinical situation, it remains highly pertinent information that 2 of the most affected genes, VEGF and TGFβ1, are prominent growth factors for bone regeneration.
Furthermore, most of the accumulated evidence from this manuscript seems to suggest that it is the osteocytes that are playing a key role in this process following irradiation. As most of the cells are apoptotic following irradiation, it becomes evident that they are major key players in maintaining tissue vascularity as they are key players in VEGF production. In previous histologic studies, it was found in human specimen samples from osteoradionecrotic bone after 36 Gy, a loss of osteocytes could be observed [
34,
35]. Thus, it remains essential to further study the relationship between irradiated bone and most specifically osteocytes. Future research aimed at investigating protein release of growth factors such as TGFβ1 and VEGF using an animal model would be extremely advantageous. A further understanding of this relationship could provide more pertinent information to clinicians to better gear regenerative procedures for the treatment of osteoradionecrosis of the jaw and possibly provide better preventative measures for these patients prior to complications.
Abbreviations
TGFβ1, transforming growth factor Beta 1; BMP2, bone morphogenetic protein 2; VEGF, vascular endothelial growth factor; IL1β, interleukin 1 beta; RANKL, receptor activator of nuclear factor kappa-B ligand; PBS, phosphate buffered solution; ELISA, enzyme-Linked Immunosorbent assay; SEM, scanning electron microscopy; BCM, bone conditioned medium
Acknowledgements
This study was funded entirely by the department of Oral and Maxillofacial Surgery, University of Bern, Switzerland. We also thank Catherine Solioz for her skillful technical assistance. All authors declare no conflict of interest and have viewed and agreed to submission.