Osteocytes and cancer

Abstract

Purpose of Review:

While the function of osteocytes under physiologic conditions is well defined, their role and involvement in cancer disease remains relatively unexplored, especially in a context of non-bone metastatic cancer. This review will focus on describing the more advanced knowledge regarding the interactions between osteocytes and cancer.

Recent Findings:

We will discuss the involvement of osteocytes in the onset and progression of osteosarcoma, with the common bone cancers, as well as the interaction that is established between osteocytes and multiple myeloma. Mechanisms responsible for cancer dissemination to bone, as frequently occurs with advanced breast and prostate cancers will be reviewed. While a role for osteocytes in the stimulation and proliferation of cancer cells has been reported, protective effects of osteocytes against bone colonization have been described as well, thus increasing ambiguity regarding the role of osteocytes in cancer progression and dissemination. Lastly, supporting the idea that skeletal defects can occur also in the absence of direct cancer dissemination or osteolytic lesions directly adjacent to the bone, our recent findings will be presented showing that in the absence of bone metastases, the bone microenvironment and, particularly, osteocytes, can manifest a clear and dramatic response to the distant, non-metastatic tumor.

Summary:

Our observations support new studies to clarify whether treatments designed to preserve the osteocytes can be combined with traditional anticancer therapies, even when bone is not directly affected by tumor growth.

Osteocytes in physiologic conditions

The maintenance of skeletal health requires the coordinated activity of bone cells, such as the removal of old or damaged bone matrix by osteoclasts and the formation of new bone by osteoblasts. Osteocytes, as the most numerous of all bone cells [1], play essential roles in regulating both osteoblast and osteoclast activity under physiological and pathological conditions. For example, osteocytes are known to contribute to the bone pathology observed with cancer metastases [2–4]. However, the effect of non-metastatic cancer on bone pathology and the role that osteocytes in particular play in this process is less defined.

Osteocytes are derived from terminally differentiated osteoblasts, which become embedded within the secreted bone matrix in small pockets, known as lacunae [5]. As these cells become embedded, they promote the mineralization of the collagen rich matrix surrounding them via the induction of proteins such as alkaline phosphatase and dentin matrix protein 1 (DMP1) [6,7]. Although the osteocytes are subsequently entombed within the mineralized matrix, they are densely connected to their neighboring osteocytes and cells on the bone surface via long protruding cellular processes called dendrites housed in tunnels called canaliculi. Furthermore, osteocytes are able to communicate with cells in tissues beyond the bone environment, through the intimate connectivity of this lacuna-canalicular system and the bone vasculature [7–9]. Osteocytes are known to respond to circulating hormones such as parathyroid hormone (PTH), parathyroid hormone related protein (PTHrP) and 1, 25 dihydroxyvitamin D₃ (1, 25 D₃), leading to changes in the expression of key bone remodeling proteins such as receptor activator of nuclear factor kappa-B ligand (RANKL), osteoprotegerin (OPG) and sclerostin [10,11] [12]. While these proteins are important regulators of bone homeostasis, dysregulation of their expression under pathological conditions such as cancer can have devastating effects on bone [13,14]. RANKL promotes osteoclast differentiation and activation by binding to its receptor, RANK, on the osteoclast surface and OPG acts as a decoy receptor for RANKL, thereby preventing osteoclast activation [15,16]. An increase in the ratio of RANKL/OPG leads to increased osteoclast activity and bone resorption [15]. Conditional deletion of RANKL in late osteoblasts and osteocytes resulted in reduced osteoclast number and dramatically increased trabecular bone mass in skeletally mature mice [17,18] and these findings were reproduced when RANKL was specifically deleted in mature osteocytes, but not osteoblasts [19]. This therefore identifies osteocytes as the primary source of RANKL required for bone remodeling in the adult skeleton.

In addition to regulating osteoclast activity, osteocytes can also control osteoblast differentiation and bone formation via the secretion of sclerostin and Dikkopf-related protein 1 (DKK1) [20]. Both sclerostin and DKK1 are potent inhibitors of the Wnt/β-catenin signaling pathway which act through antagonism of the low-density lipoprotein receptor-related protein, Lrp 5/6 [21,22]. Deletion of β-catenin signaling in osteoblasts inhibits their differentiation and activity and impairs bone formation [23]. Mice lacking Sost, the gene encoding sclerostin, are characterized by high bone mass due to increased bone formation[24] whereas mice overexpressing the human SOST transgene have decreased bone formation leading to osteopenia [25]. DKK1 is also highly expressed by osteocytes, as well as osteoblasts and chondrocytes [26] and deletion of Dkk1 using the 10kb Dmp1-Cre can protect against pathological bone loss [27]. The beneficial effects of DKK1 inhibition may be dependent on endogenous sclerostin levels, as upregulation of sclerostin in response to loss of DKK1 was found to restrain bone formation in the Dkk1^fl/fl/Dmp1-Cre mice [28]. Both sclerostin and DKK1 have been demonstrated to play important roles in the effects of cancer on bone, which will be discussed later in this review.

The osteocyte is highly sensitive to mechanical stress and is believed to be the key mechanosensor within bone [20,29]. Many of the afore mentioned osteocyte-selective proteins are known to be regulated by mechanical strain, with sclerostin in particular being highly downregulated after mechanical loading in vivo and in vitro, which potentiates the anabolic response of bone to loading [30,31]. Furthermore, exercise also induces the release of myokines such as beta-aminoisobutyric acid (BAIBA) and irisin from the skeletal muscle, which are known to regulate bone mass via their actions on osteocytes [32–34]. Mechanical loading has been shown to attenuate bone loss in murine metastatic cancer models [35,36] and conditioned media from fluid flow-stimulated osteocytes attenuated the metastatic potential of breast cancer cells in vitro [37]. However, whether mechanical loading of osteocytes could have a similar beneficial effect on bone loss resulting from non-metastatic cancer is currently unknown. Additionally, while exercise has been shown to be beneficial to bone health in patients with skeletal metastases [38], weight bearing exercises may be problematic in cancer patients with very low bone mass and/or advanced cachexia. However, pharmacological treatments which mimic the molecular effects of exercise while limiting physical exertion may be an attractive alternative to increase bone mass in such individuals.

One of the functions of osteocytes that is receiving renewed interest is the ability to directly remodel their perilacunar matrix. This is thought to occur through an upregulation of osteoclastic genes such as tartrate-resistant acid phosphatase (Trap), cathepsin K (Ctsk), the v-ATPase subunit Atp6v0d2, matrix metalloproteinase 13 (Mmp13) and carbonic anhydrase 2 (Car2) [39–41]. This removal of their perilacunar matrix was previously referred to as ‘osteocytic osteolysis’ and thought only to occur under pathological conditions [42]. Nonetheless, it is now known that this process allows for calcium release from the bone under physiological calcium-demanding conditions such as lactation, using a similar mechanism to that of osteoclastic bone resorption [43]. Circulating hormones such as PTH [44], PTHrP [39] and transforming growth factor beta (TGFβ) [45] have been shown to promote perilacunar resorption, whereas 1, 25 D₃ [46] and calcitonin [47] signaling inhibit osteocyte lacunar enlargement.

Evidence also suggests that upon cessation of the increased calcium demand, osteocytes can replace the bone mineral that was previously removed which completes the process called perilacunar ‘remodeling’. The increased lacunar area in lactating mice was no longer observed 7 days post weaning when lacunar size returned to normal [39]. Additionally, tetracycline labeling of mineralizing bone surface was seen around the lacunae of calcium-depleted egg-laying hens after the resumption of dietary calcium [48]. While the relevance of perilacunar remodeling/osteocytic osteolysis in cancer-induced bone disease has still not been determined, we have identified a role for osteocytic osteolysis in the bones of mouse models of non-metastatic cancers [49], which will be discussed in more depth later in this review.

Osteocytes in cancer diseases

Under physiological conditions, the osteocytes, responsible for maintaining bone homeostasis by orchestrating the activity of osteoclasts and osteoblasts, are the master regulators of bone remodeling. Nevertheless, their involvement in cancer remains mostly unknown and underestimated in favor of a direct involvement of osteoclasts and osteoblasts in the formation of osteolytic and osteoblastic lesions, respectively. We will next highlight the most advanced and up-to-date knowledge of the role of osteocytes in the pathogenesis of cancer-associated bone disease. In particular, the role of osteocytes in osteosarcoma, multiple myeloma and metastatic-bone disease, as well as their involvement in promoting cancer-derived comorbidities in the absence of direct tumor dissemination to the bone will be described (Figure 1).

Figure 1.

Schematic representation of tumor-osteocyte interactions. A) Osteocytes adjacent to tumors, as in the in case of multiple myeloma or bone metastatic disease. The osteocytes can interact with the cancer cells physically or by means of soluble factors, acting as either tumor suppressors or stimulators. B) Osteocytes distant from primary tumor (e.g., lung). The cancer cells can communicate with the osteocytes by means of soluble factors.

Osteosarcoma and osteocytes

Osteosarcoma (OS), the most common malignancy directly originating from bone tissue, is a relatively rare disease with an incidence of three cases per million, affecting mostly children and adolescents [50]. Common sites of OS formation are usually found near the growth plate in long bones such as tibia, femur and humerus [51]. The 5-year relative survival rate in childhood is currently around 67% [52], and unfortunately, the survival rate has not improved significantly in the last few decades, in patients with OS-related metastatic growth [53]. The cytological origin of OS remains unclear, and whether the cancer cells have osteoblastic origins or derive from mesenchymal stem cells (MSCs) is yet to be determined [54]. OS shows osteoblastic/osteolytic mixed lesions, and different histological cell types can be found [55]. Interestingly, expression of the osteocyte marker DMP1, the osteocyte-associated gene matrix extracellular phosphoglycoprotein (MEPE) and the phosphate-regulating neutral endopeptidase homolog x-linked (PHEX) was observed in the osteoblastic subtype of human OS, thus suggesting that osteocytes can serve as progenitors for OS [56,57]. Interestingly, the immortalized murine osteocyte cell line MLO-Y4 was shown to generate solid tumors when implanted subcutaneously or intratibially in immunodeficient SCID mice, and the MLO-Y4 intratibial tumors were characterized by showing similar osteoblastic/osteolytic lesions as observed in OS patients [56]. Moreover, OS cell lines showed mixed expression of OPG, RANKL and SOST, thus explaining the mixed nature of the osteolytic/osteoblastic lesion[56]. Under physiological conditions osteocyte-derived OPG, RANKL and SOST coordinate bone remodeling, thereby supporting the hypothesis that osteocytes can act as potential precursors of OS, responsible for promoting the formation of mixed lesions [56].

Multiple myeloma and osteocytes

Multiple myeloma (MM) originates from mature plasma cells in the bone marrow and represents the second most common hematological cancer [58]. The majority of MM patients develop multiple myeloma bone disease (MMBD), a severe bone complication resulting in reduced bone formation and increased bone destruction. The presence of osteolytic lesions increases the risk of fractures, as well as morbidity and mortality [59]. Osteocyte viability is decreased in the bone of MM patients, and this event positively correlates with the severity of the osteolytic lesions [60]. The increase of osteocyte death was also described in primary human bone ingrafted in SCID-hu mice bearing MM [61]. In particular, in this study osteocyte death was found also in the contralateral limb not injected with tumor cells. This therefore suggests that soluble factors released from the tumor cells may act systemically to affect osteocyte viability. To clarify the relationship between osteocyte apoptosis and tumor growth, Trotter and collaborators used an in vivo model of diphtheria toxin-induced osteocyte death to show that osteocyte apoptosis directly supports tumor growth and homing of the MM cells, mainly by promoting changes in the microenvironment that favor the formation of cancer metastases [61]. Elevated circulating sclerostin was also found to correlate with osteocyte death and the severity of osteolytic lesions and bone loss [62]. As showed by analyzing human and murine samples, SOST was not expressed in myeloma tumor cells [63]. The levels of SOST detected in the osteocytes were reduced in the 5TGM1 tumor-bearing mice compared with the controls, although no change in sclerostin was observed by immunohistochemistry in either group. Moreover, an anti-sclerostin antibody was shown to improve multiple myeloma bone disease by preventing bone loss and increasing bone formation, as well as by increasing resistance to fracture [63]. Using the same in vivo model, others reported that the numbers of sclerostin positive osteocytes was increased in the bone of the 5TGM1 tumor-bearing mice [64]. The global deletion of Sost (Sost^−/−/Scid mice) as well as the administration of an anti-sclerostin antibody were able to decrease osteolysis and reduce bone loss without affecting tumor growth [64]. Moreover, MM cells stimulate the production from the osteocyte of IL-11 a factor able to stimulates osteoclastogenesis [60]. Interestingly, IL-11 expression was found increased in the osteocytes of MM patients with bone lesion compared with patients without bone lesion [60]. Furthermore, the relationship between osteocytes and MM cells was clearly shown by the physical interaction between the osteocyte dendrites and the MM cells [65], whose bidirectional interaction was found to be regulated by the Notch pathway [65].

Recent evidence also shows the role of osteocytes in increasing bone marrow angiogenesis in MM. Indeed, in the hypoxic environment that characterizes MM, osteocytes acquire a pro-angiogenic phenotype by producing vascular endothelial growth factor A (VEGF-A), with its secretion stimulated by the release of the osteocyte hormone Fibroblast growth factor 23 (FGF-23), a known activator of the Egr1-Vegf-a signaling pathway [66]. Accordingly, the circulating levels of FGF-23 were found elevated in MM patients compared to healthy subjects [67]. Interestingly, osteocytes were found to be the primary source of FGF-23, as elegantly shown by an in vitro study in which FGF-23 expression was enhanced in osteocytes co-cultured with human MM cells [67]. The same study showed that MM cells express the FGF- receptors (FGFRs), and its activation drives the production of pro-osteolytic factors in the tumor cells [67]. Altogether, these observations support a direct involvement of osteocytes in the pathogenesis of MM.

Osteocytes in metastatic-bone disease

Every year in the United States more than 400,000 cancer patients are affected by bone metastases, representing one of the most common sites of cancer dissemination [68]. Unfortunately, formation of bone metastases is commonly considered an incurable consequence of terminal malignancies, and each year around 350,000 patients succumb to such complication of cancer [69]. Several different types of solid cancers are known to metastasize to bone, with the highest incidence of bone dissemination found in lung cancer, followed by esophageal and prostatic malignancies [70]. However, based on the incidence of metastatic bone disease, patients affected with prostate, breast and renal cancers are usually most susceptible to developing highly debilitating bone lesions [70].

Skeletal-related events (SREs) including bone fractures, hypercalcemia, ostealgia, and spinal cord and nerve-compression syndromes are severe consequences of bone metastasis, ultimately resulting in reduced quality of life and increased mortality and morbidity rates [69]. The colonization of bone by metastatic cancer may result in abnormal bone resorption (i.e., osteolytic lesions) or bone formation (i.e., osteoblastic lesions). While osteolytic metastases are predominantly associated with advanced breast cancer, osteoblastic lesions are more frequent in prostatic cancer dissemination [71]. Bone metastases can occur from a complex interplay between cancer cells and bone cells. Bone tissue can be favorable environment for cancer cells, with bone alterations and release of bone factors induced by the tumor cells further accelerating the formation of cancer lesions, thereby representing a “vicious cycle” [72]. Osteoclast and osteoblast functions in the metastatic process have been well characterized; however, whether osteocytes play a similar role remains to be clarified.

Breast cancer and osteocytes

Osteocytes connect and communicate by means of dendrites that generate a 3D network that ultimately plays an important role in the regulation of bone homeostasis. In the presence of primary tumors or bone metastases, the osteocyte-network can change drastically. The osteocytes can acquire a spherical and undeveloped shape with fewer and shorter dendrites, and the number of connections can be either reduced or increased [73]. As a result, mechanosensation and intracellular signaling are also significantly affected due to changes in bone fluid flow within the lacuna-canalicular system that has occurred in this tightly organized structure [8,74]. Of note, a recent study showed that, differently from osteolytic and osteosclerotic lesions, normally associated with increased lacunar size and vascular canals, bone-metastatic breast cancer cells can interact with the osteocytes and induce the reorganization of the lacuna-canalicular network, consistently with augmented number and size of empty lacunae as well as reduced vascular porosity [75].

A recent study by Sano and collaborators revealed some beneficial effects of osteocyte-derived factors in the treatment of breast cancer-associated brain metastases [76]. Indeed, in addition to bone, breast cancer can disseminate to the brain, an event that severely reduces the survival rate of patients [77]. In this study, the conditioned medium derived from MLO-A5 osteocyte-like cells was shown to inhibit tumor growth in both mammary fat pad and tumor-invaded tibiae in an in vivo model, whereas the injection of osteocytes in the frontal lobe was described as a powerful approach capable of suppressing the growth of brain tumors. These positive effects were enhanced by the overexpression of Lrp5, β-catenin, and interleukin 1 receptor alpha (IL1rα) in the osteocytes. The authors also showed the osteocyte conditioned medium was particularly enriched with histone H4, which was found to limit tumor growth in a multitude of different ways, including downregulation of tumor promotors such as C-X-C motif ligand 1 (CXCL1) and CXCL5, and upregulation of tumor suppressors such as p53, tumor necrosis factor-related apoptosis-inducing ligand (TRAIL), lim domain and actin binding 1 (LIMA1), and desmoplakin (DSP). Overall, these findings underline the potential beneficial effects of a conditioned media-based therapy in the treatment of metastatic breast cancer.

Unfortunately, the data regarding the role of the osteocytes in breast cancer are controversial. Recent evidence suggests that osteocytes can have a positive role against breast cancer growth and metastasization. Contrarily, others suggest that the osteocytes can enhance proliferation, migration and invasion of breast cancer cells (Table 1).

Table 1.

Osteocyte roles in breast cancer

Osteocyte roles in breast cancer
Positive
CM derived from mechanically stimulated osteocytes reduce the metastatic potential in vitro [37] CM derived from MLO-A5 cells inhibits tumor growth in vivo [77] Osteocytes suppress brain metastasization of breast cancer [77] CM from mechanically stimulated MLO-Y4 cells reduces the migration and invasion of MDA-MB-231 breast cancer cells in vitro in a Cx43 dependent manner [80] CM from osteocytes exposed to low intensity mechanical stimulation induced mesenchymal-to-epithelial transition [86]
Negative
Mechanically-stimulated osteocytes secrete factors such as CXCL1 and CXCL2 that promote cell migration and proliferation [82] Mechanically-stimulated osteocytes release pro-tumorigenic mediators such as CCL5 and MMPs that facilitate cancer invasion in bone [83] CM from osteocytes exposed to high intensity mechanical stimulation induced epithelial-to-mesenchymal transition [86]

Prostate cancer and osteocytes

The paucity of evidence relative to the role of the osteocytes in cancer biology also results from the lack of in vitro systems that resemble the physical interaction between bone metastases and osteocytes. In order to address this point and study the crosstalk between bone and cancer, Choudhary and collaborators used engineered 3D bone tissue composed of primary osteocytes and human prostate cancer cells [78]. Their observations suggested that prostate cancer cells stimulate the release of FGF23 and DKK-1 from osteocytes. Moreover, the increased mineralization and the high level of alkaline phosphatase resembled the osteoblastic lesions that predominantly affect prostatic cancer patients [78]. In another study it was shown that conditioned media from MLO-Y4 osteocytes, cultured in a monolayer or in 3D culture, were able to stimulate human prostatic cancer cell proliferation and post-wound migration in a cancer cell dependent manner [2]. Other evidence showed that osteocytes are able to promote prostate cancer bone metastasis progression via the synthesis and release of growth-derived factor 15 (GDF15). Interestingly, GDF15 receptor, GFRAL, is normally highly expressed in several prostatic cancer cell lines, and activation of the downstream signaling leads to stimulation of the EGR1 pathway, known to drive cancer cell migration and invasion [4].

Osteocytes, bone-mechanostimulation and cancer

Several studies suggest that both mechanical stimulation and exercise can contribute to preventing cancer progression in bone. Using an in vivo model, Lynch and collaborators showed that mechanical stimulation of the tibia inhibits bone metastatic colonization and osteolysis and increases trabecular bone mass[36]. Interestingly, these effects were not the result of increased cancer cell death induced by tibia compression, but rather of the modulation of genes involved in the regulation of bone homeostasis [36]. To further support this idea, another study showed that mechanical signaling induced by low-intensity vibration (LIV) was able to reduce the spread of MM cells in the bone marrow compartment and protect bone quality against tumor cells [79]. The same group also described how LIV improved bone loss induced in a model of spontaneous granulosa cell ovarian cancer [35].

As previously mentioned, osteocytes are the principal mechanosensors in bone, primarily due to their ability to translate mechanical stimuli into molecular signals. In this regard, connexin 43 (Cx43), a transmembrane protein sensitive to mechanical stimulation and highly enriched in osteocytes, seems to play an important protective role against breast cancer cell metastasis [80]. Zhou et al. showed that the conditioned medium deriving from MLO-Y4 osteocytic cell cultures subjected to fluid flow shear stress was able to reduce the migration and invasion of MDA-MB-231 human breast cancer cells in vitro, whereas the use of a specific antibody against Cx43 hemichannels reduced their positive effect [80]. Furthermore, in the same study it was reported that osteocyte-specific Cx43 knockout mice have increased tumor growth and metastasis [80]. The positive effects associated with Cx43 were mediated by mechanical stimulation promoting the opening of Cx43 hemichannel, thus allowing the release of ATP from the osteocytes and the consequent inhibitory effect on cancer cells [80,81].

Osteocyte mechanosensing can also be modulated by mechanically-loaded cancer cells, thus stimulating dendrite formation and bone loss. More specifically, it was shown that the conditioned medium derived from breast cancer cells subjected to fluid flow sheer stress (FFSS) increased MLO-Y4 dendrite formation to a degree comparable to direct FFSS stimulation of MLO-Y4 cells [82]. Consistent with this morphological change, increased expression of E11, a gene involved in dendrite formation, was also observed. Further, the RANKL/OPG ratio was increased in the MLO-Y4 cells exposed to the conditioned medium from mechanically-loaded breast cancer cells, clearly suggesting that cancer cells stimulate osteocytes to initiate bone resorption [82]. In agreement with such findings, Dwivedi and collaborators also showed that in vitro mechanically stimulated osteocytes secrete factors that promote migration and proliferation of breast cancer cells [83]. The analysis of the secretome revealed that dynamic fluid flow stimulates the production of cytokines such as CXCL1 and CXCL2, known as important regulators of cancer cell migration. Consistently, the CXCR2 receptor was highly expressed in breast cancer cells, therefore suggesting that the stimulated osteocyte-derived cytokines may directly promote breast cancer metastasis to bone [83]. Of note, bone metastatic colonization was also found to induce mechanical pressure on the bone, thus fostering cancer growth through osteocyte activation. Indeed, as also shown in a model of prostatic cancer, the development of metastases in bone was found to respond to pressure and osteocyte mechanical stimulation. In turn, the cells produce and release pro-tumorigenic mediators, such as CC chemokine ligand 5 (CCL5) and metalloproteases (MMPs), that facilitate cancer invasion in bone [84].

Using mice bearing mesenchymal-like triple-negative Py8119 cells, Wang and collaborators showed that moderate tibia loading (4.5 N) and moderate treadmill running counteract breast cancer-induced bone destruction, whereas overloading (8 N) accelerates the bone derangements [85]. Similarly, the conditioned media derived from 0.25 Pa mechanical stimulated osteocytes was found to induce the mesenchymal to epithelial transition, whereas, on the contrary, a 1 Pa stimulation was promoting such process [86]. Another study showed that mechanical tibia loading reduces osteolysis and improves trabecular bone microarchitecture in mice intratibially injected with 4T1 breast cancer cells [87]. In particular these positive effects of mechanical stimulation were associated with reduced tumor growth, bone metastasization and osteoclast differentiation, as well as improved osteoblast number and mineralization [87]. In a similar manner, knee loading was shown to reduce the growth of EO771 breast cancers in the mammary fat pad, as well as their local progression in tumor-invaded tibiae [88]. In this case, the positive effects were primarily associated with reduced levels of WNT1-inducible signaling pathway protein 1 (WISP1), a factor found increased in patients with breast cancer and known to negatively correlate with survival [88]. Moreover, to the extent of identifying bio-fluids endowed with antitumor properties, Wu and colleagues presented evidence that a short period of loading (5 N) in mice, similar to a 10-minutes step aerobics in human, can generate changes in urine able to reduce proliferation, migration and invasion of cancer cells in both in vitro and in vivo setting [89].

A recent study using an in vitro approach showed that mechanical stimulation of osteocytes is able to activate signaling pathways involved in the control of genome integrity that could promote cancer progression [90]. In particular, in their study Santos and collaborators compared the secretome and transcriptome of murine and human osteocyte-like cell lines undergoing mechanical stimulation and found that the stretched cells displayed increased P53, cell cycle, DNA repair and nucleotide excision pathways. Moreover, pathways related to extracellular matrix remodeling, cell-matrix interaction and bone remodeling were the most highly regulated upon mechanical stimulation. These observations suggest that changes in the secretome and transcriptome in osteocytes subjected to mechanical stimulation may affect cancer progression [90]. Altogether, these findings highlight a role for osteocyte mechanosensing in the modulation of bone microenvironment and cancer biology and corroborate the idea that the intensity of exercise is a fundamental variable to be considered whenever physical exercise is suggested as therapeutic intervention in association with conventional anticancer therapies [85].

Osteocytes in non-bone metastatic tumors

Recent experimental findings by ours and other groups no longer support the concept that bone derangements occur solely as a consequence of cancer dissemination to bone. Indeed, several preclinical studies have clearly shown that mice bearing non-bone metastatic tumors develop different degrees of bone loss [91–93]. These experimental observations are further supported by clinical evidence showing that non-bone metastatic tumors, such as non-small lung cancer, non-metastatic breast cancer and invasive cervical cancer can cause osteoporosis or reduced bone mineral density [94–96]. This therefore corroborates the idea that tumor-derived soluble factors or host factors produced in response to the tumor can impair bone homeostasis by stimulating bone resorption and, consequently, cause bone loss.

Despite all of the evidence supporting the presence of bone derangements due to non-bone metastatic cancers, little is known regarding the effects on osteocyte physiology and function. In a recent study aimed at clarifying this point, we have reported severe changes in osteocyte viability and function in rodents bearing the Colon-26 adenocarcinoma (C26), the ES-2 ovarian cancer (ES-2), and Lewis lung carcinoma (LLC) [49], with massive increases in osteocyte death and empty lacunae across these three in vivo models (Figure 2). By co-culturing IDG-SW3 osteocytes and cancer cells, we showed that unknown tumor-derived factors are directly responsible for inducing osteocyte death. Interestingly, all three models presented elevated RANKL/OPG ratio and higher mRNA expression of Tnfa in the osteocytes, both in vivo and in vitro, thus suggesting that these soluble factors may be contributing to the differential effects on osteoblast death and osteoclast activation that characterize these cancer models. Furthermore, for the first time we showed that cancer cells reprogram osteocytes toward an osteoclastic phenotype, as supported by increased expression of genes normally expressed by osteoclasts such as Apc5, CtsK, Atp6v0d2, Mmp13 in the IDG-SW3 osteocytes exposed to cancer cells, as well as in the osteocyte enriched samples derived from the bone of tumor hosts. These molecular changes were in line with in vivo evidence of osteocytic osteolysis, as suggested by increased lacunar size due to osteocyte-dependent degradation of the perilacunar matrix [49].

Figure 2.

A) TUNEL-stained transverse sections in the femur midshaft of control and C26 tumor-bearing mice. B) TRAP stained transverse sections from the femoral mid-diaphysis of control and C26 tumor-bearing mice. C) Backscatter SEM (BSEM) images of longitudinal sections from the mid-diaphysis of control and C26 tumor-bearing mice. The red arrows indicate the lacunae with a larger area.

Altogether, these data show that non-bone metastatic cancers can affect bone tissue not only by impacting osteoclasts and osteoblasts, but also osteocytes, thereby causing osteonecrosis and osteocytic osteolysis. These, in turn, can have detrimental effects on the bone mechanical properties, strength and capacity to remodel, thus generating a musculoskeletal system that is incapable of recovering once the cancer is removed or in remission.

Conclusions and Future Directions

While the role of osteocytes in the regulation of bone homeostasis in physiological conditions has been thoroughly investigated, their functions in the onset and progression of cancer remain understudied, and whether osteocytes directly participate in tumor dissemination to bone is currently being debated.

Little is known about the role(s) of osteocytes in the early stages of cancer dissemination to bone, and how osteocytes influence cancer cells, osteoblasts, and osteoclasts, as well as their consequences for bone homeostasis remain to be elucidated. The field of research would also benefit from the characterization of new models for the study of metastatic cancer, especially considering that most of the available data were generated using preclinical mouse models for the study of myeloma and breast cancer only. On the other hand, a limitation that has thus far prevented the progress in the field is associated with the fact that, while most of the current research has focused on investigating the mechanisms of bone-metastatic disease, very few studies have been devoted to understanding the role of osteocytes in a setting of non-bone metastatic cancer (i.e., when the bone is not macroscopically affected by the tumors). In this regard, collaborative efforts from our group have shown that tumor-derived factors released from non-bone metastatic tumors induce osteocytic osteolysis and osteocyte death, thus clearly suggesting that even cancers that do not usually metastasize to bone can severely impact bone health by prompting osteocytic bone destruction. In particular, it is possible that such cancer-associated osteocyte defects may contribute to the generation of a musculoskeletal system unable to fully recover upon cancer remission, thus resulting in chronic or permanent musculoskeletal complications.

Altogether, these observations warrant studies to investigate whether treatments aimed at preserving the osteocytes or in general to improve bone health should be combined with traditional anticancer therapies even when bone is not visibly affected by tumor growth.