Epigenetic Regulation of Bone Remodeling and Bone Metastasis

Abstract

Bone remodeling is a continuous and dynamic process of bone formation and resorption to maintain its integrity and homeostasis. Bone marrow is a source of various cell lineages, including osteoblasts and osteoclasts, which are involved in bone formation and resorption, respectively, to maintain bone homeostasis. Epigenetics is one of the elementary regulations governing the physiology of bone remodeling. Epigenetic modifications, mainly DNA methylation, histone modifications, and non-coding RNAs, regulate stable transcriptional programs without causing specific heritable alterations. DNA methylation in CpG-rich promoters of the gene is primarily correlated with gene silencing, and histone modifications are associated with transcriptional activation/inactivation. However, non-coding RNAs regulate the metastatic potential of cancer cells to metastasize at secondary sites. Deregulated or altered epigenetic modifications are often seen in many cancers and interwound with bone-specific tropism and cancer metastasis. Histone acetyltransferases, histone deacetylase, and DNA methyltransferases are promising targets in epigenetically altered cancer. High throughput epigenome mapping and targeting specific epi-enzymes will be helpful in the development of personalized epi-drugs for advanced and bone metastasis cancer patients. This review aims to discuss and gather more knowledge about different epigenetic modifications in bone remodeling and metastasis. Further, it provides new approaches for targeting epigenetic changes and therapy research.

1. Introduction

The skeletal system primarily consists of bones and soft tissues, which act as the central framework of the human body. Bone is a highly dynamic and metabolically active organ capable of providing physiological support to the entire body. Bone remodeling is a continuous restructuring process that involves the replacement of matured old bone tissues with new bone. This process helps to maintain normal calcium levels in the body. Homeostasis between anabolic and catabolic activities is crucial to maintain the integrity of the bone for structural and immunological support [1, 2]. Cancer is the second leading cause of death in the US, having the ability to spread to any part of the human body. When cancer cells dissociate from their primary site, they move through the bloodstream and lymph nodes and settle in any organ far from the place of origin. Some of the tumors commonly spread to the bone are the breast, kidney, prostate, lung, thyroid, and bladder [3]. Most cancer patients develop bone metastasis, drastically reducing their survival [4]. The post-metastatic condition is especially concerning since it is linked to cancer-induced bone pain, fractures, spinal cord compression, cachexia, and compromised life style [5].

Increasing evidence supports two general phenomena, i.e., genetic and epigenetic alterations in tumor initiation and metastasis. Change in gene expression without changing DNA sequence is referred to as epigenetics, which is sufficiently robust to regulate gene expression dynamics [6]. Initial studies that reported the reprogramming and loss of the malignant character of metastatic tumor cells in unique microenvironments suggest that some metastatic traits are reversible and not due to genetic mutations but due to epigenetic modifications [7, 8]. Recent genome-wide sequencing has revealed the importance of genetic and epigenetic alterations in bone remodeling and metastasis [9-11]. The most prominent and earliest recognized epigenetic alteration of cancer cells is the genome-wide deregulation of DNA methylation [12]. Several other phenomena, such as acetylation, phosphorylation, deimination, and ubiquitination (DNA and histones), are also involved in altering gene expression by epigenetic modifications. Oncogenesis from tumor initiation to metastasis is full of genome-wide epigenetic changes followed by reprogramming of various gene regulations [13]. Hypermethylation at the promotor region can inactivate the tumor suppressor genes, whereas hypomethylation induces the overexpression of the genes in tumorigenesis [14]. Furthermore, acetylation of lysine and post-translational modifications of histones modulate the several genes expression, which is critical for cancer progression and metastasis [15]. However, multiple chemokines and cytokines networks have recently been discussed in various studies elucidating their roles in cancer and bone metastasis [16-19]. Several external factors can modulate epigenetic modification patterns, including nutritional, physical inactivity, and environmental and psychological stress (Figure 1). Apart from external factors, individual characteristics such as age, gender, and genetic makeup are prominent factors for epigenetic modifications [20-22]. DNA methylation, histone modifications, and post-transcriptional gene regulation by non-coding RNA (ncRNA) are the primary epigenetic mechanisms to regulate gene functions [23]. Above mentioned modifications are crucial in the normal regulation of growth and development, but alterations contribute to neoplastic phenotypes. The major microenvironment-epigenetic modifications (as discussed above) support stable transcriptional programs without specific heritable changes in the genome or DNA and are associated with the initiation and progression of cancer [24]. These changes in cancer cells can make an autoregulatory transcriptional circuit that can control their cancer cell plasticity and metastatic potential. Identifying these epigenetic changes associated with bone metastasis in various cancer is crucial. Here, we will discuss the physiology of bone remodeling and epigenetic modifications in regulating bone homeostasis. Afterward, epigenetic alterations in cancer and the vicious cycle of bone metastasis will be discussed.

Figure 1: Epigenetic modifications and consequences in cancer:

(A) Major lifestyle factors such as physical activity, aging, and food habits are associated with epigenetic changes in normal and cancer cells. (B) DNA methylation, histone modifications, and non-coding RNAs (micro-RNA, long non-coding RNAs, and circular RNAs) are essential mechanisms for the epigenetic reprogramming of tumor suppressors and proto-oncogenes. (C) Altered epigenetic modifications in cancer cells regulate various cellular responses, including cell proliferation, metabolism, angiogenesis, cell death, invasion, and metastasis. Most commonly, hypermethylation is associated with gene silencing, such as tumor suppressor genes and hypomethylations activate the genes, such as proto-oncogenes. However, histone modifications are associated with transcriptional activation/inactivation.

2. Bone cells and bone remodeling

Under normal physiological conditions, bone resorption and bone formation are highly coupled with a process known as bone remodeling. Bone remodeling occurs in discrete locations and comprises a group of different types of cells. Bone cells consist of two lineages: namely osteoblasts and osteoclasts. Osteoblasts belong to the central axis for sequential bone formation (containing mesenchymal stem cells, pre-osteoblasts, mature osteoblasts, bone-lining cells, and osteocytes). Osteoclasts (containing hematopoietic stem cells, macrophages, osteoclasts, and multinucleated giant cells) are responsible for the resorption of bone [25, 26]. In general, osteoclasts and osteoblasts fit into a temporary structure commonly known as a basic multicellular unit (BMU) to control bone remodeling. Briefly, the BMU is an association of different lineages in the bone marrow associated with the dynamic modeling of the bone. The resorbed surface of the bone is cleaned up by the lining cells and macrophages, and osteoclast precursors differentiate to fill the space that has been resorbed. In addition to osteoblasts and osteoclasts lineages, different bone residence immune cells such as T cells, B cells, dendritic cells, natural killer T cells, regulatory T cells, neutrophils, and myeloid-derived suppressor cells participate in bone homeostasis [27].

Osteoblast lineages

Osteoblasts are the lineage of pre-osteoblast and mesenchymal stem cells (MSCs) in the stroma of bone marrow and are accountable for bone formation and matrix maturation. These are polarized cuboidal cells with a comparatively short lifespan (nearly three months) [28]. The unique function of osteoblasts is mineralization, which subsequently helps in states of new bone formation and remodeling [29]. MSCs are non-hematopoietic stem cells that possess the capacity to self-renew and are multipotent. In addition to the bone cells, MSCs can differentiate into cartilage and adipose tissue [30, 31]. In the presence of exogenous factors (dexamethasone, ascorbate, and β-glycerol phosphate), MSCs can direct their fate toward osteogenic cell differentiation and form mineralized nodules due to the increased alkaline phosphatase (ALP) expression [30]. Runt-related transcription factor 2 (RUNX2) and osterix (OSX) are the main transcription factors that are involved in the initiation and promotion of osteogenic differentiation of MSCs [32, 33]. The differentiation of bone marrow MSCs towards the osteogenic niche is tightly regulated by the nuclear transcriptional factor RUNX2 [34]. Some studies demonstrated that the mice lacking RUNX2 compromised osteoblast maturation and showed a non-mineralized skeleton [35, 36]. Transforming growth factor beta-1 (TGFβ-1) signaling is involved in the recruitment of MSCs at active sites of bone-resorption to differentiate into osteoblasts [37].

Pre-osteoblasts, a heterogeneous cell population in the proliferative phase, are essential for the osteogenic process, particularly in the regulation of mineralization with the expression of functional proteins, ALP, and osteocalcin [38, 39]. At a more advanced stage of differentiation, pre-osteoblasts are usually considered to express RUNX2 or both RUNX2 and OSX [40]. In addition to bone matrix maturation and new bone formation, the osteoblasts regulate the activity of bone-resorbing osteoclasts cells. At the sites of active bone formation, the osteoblasts secrete the type1 collagen and non-collagenous proteins containing osteonectin and osteocalcin [41]. The mineralized bone matrix contains higher hydroxyapatite (a form of calcium phosphate) which is beneficial for the skeletal system. After active bone formation and matrix maturation, matured osteoblasts either undergo apoptosis or differentiate further into osteocytes and become quiescent bone-lining cells [42]. Osteocytes are the most abundant, quiescent, terminally differentiated cells of the osteoblast lineage. These are the main mechano-sensitive skeletal cell types surrounded by an unmineralized matrix called osteoid during bone formation. Osteocytes are one of the primary sources of receptor activators of nuclear factor-κB ligand (RANKL) in bone remodeling. They are also involved in the inhibition of osteoblast differentiation and produce macrophage colony-stimulating factor (MCSF) and CCL2, which favors the recruitment of osteoclast precursors and bone resorption [43].

Another important osteoblast lineage is bone lining cells (BLCs). These cells are quiescent, perennial flat cells covering the bone surface. Initially, it was believed the primary function of BLCs was the removal of demineralized matrix on the bone surface [44]. However, several recent studies have explained the role of BLCs in bone remodeling, signifying that anabolic stimulus can induce BLCs to provide the source for bone formation [45, 46]. Parathyroid hormone (PTH) treatment can also cause the transformation of BLCs into active osteoblasts [47]. Taken together, osteoblasts, osteocytes, and BLCs originate from MSCs and are responsible for ossification-related phenomena.

Osteoclast lineages and Immune components

Osteoclasts are large multinucleated cells originating from the fusion of mononuclear progenitors of monocyte/macrophage. Together with the self-renewal capability of hematopoietic stem cells (HSCs), they produce multipotent progenitors. These multipotent progenitors differentiate into granulocyte-macrophage progenitors, and osteoclast lineages such as macrophages, osteoclasts, dendritic cells, and granulocytes originate from granulocyte-macrophage progenitor cells [48]. Additionally, multipotent progenitors bring lymphoid progenitors, which give rise to T, B, and natural killer cells [49]. Macrophages residing in the bone are also called osteal macrophages or osteomacs and participate in the regulation of bone formation, homeostasis, and remodeling [50]. The c-Fms signaling regulates the osteoclast differentiation from mononuclear precursor cells in response to MCSF, which upregulates RANK expression. Dendritic cell-specific transmembrane protein (DC-STAMP) is crucial for the fusion of the mononuclear precursors or osteoclast differentiation [51]. These multinucleated osteoclasts resorb the bone matrix by releasing several enzymes and proteins [52].

Megakaryocytes (MK) are derived from hematopoietic stem cells, which produce thrombocytes or platelets crucial for normal blood clotting. Increasing evidence suggests that MK and its secreted factors have a role in bone homeostasis. MK has been shown to express several factors (osteocalcin, osteonectin, TGF-β, and PDGF) involved in bone turnover [53-57]. Since MKs express estrogen receptors (ERs) and also synthesize TGF-β, suggesting the role of MKs in estrogen-mediated bone remodeling [58]. In vitro, MK enhances osteoblast proliferation and differentiation and expresses RANKL and Osteoprotegerin (OPG) [59]. MKs modulate the type-1 collagen synthesis suggesting a straight involvement in bone formation, whereas bone resorption is indirectly regulated by OPG and RANKL ratio [60]. Overall, MKs can modulate bone resorption and formation processes during active remodeling.

In addition to the bone cells, major immune components found in the bone marrow and lymphoid progenitor cells give rise to B cells, T cells, and natural killer (NK) cells. Nevertheless, it is well known that B cells or lymphocytes are generated from HSCs and developed in the bone marrow as humoral immunity, a component of the adaptive immune system. Similarly, T cells also originate from HSCs and play a fundamental role in the adaptive immune response [27]. Furthermore, NK cells are cytotoxic lymphocyte that actively participates in the innate immune system. NK cells differentiation and maturation occur in different organs, including lymph nodes, tonsils, thymus, spleen, and bone marrow [61]. Therefore, the bone marrow provides immune (innate and adaptive) networks and is capable of fine-tuning of immunity.

3. Physiological bone remodeling

Bone remodeling is a vital active process tightly regulated under physiological conditions. It is governed by the unique microenvironment and BMU residing in the bone [62]. BMU creates a unique environment to facilitate bone remodeling [63]. Physiological bone remodeling can be divided into five sequential phases: activation, resorption, reversal, formation, and termination.

Activation phase:

Various physiological and pathophysiological conditions or stimuli can trigger the activation phase. The precise selection of the remodeling site or BMU among the million possible sites is crucial in this process. The parathyroid hormone (PTH; a calciotropic hormone) binds with a G-protein-coupled receptor called PTH receptor or calcium-sensing receptor (CaSR) present on the pre-osteoblasts to generate signals for the maintenance of calcium homeostasis [64]. Osteocytes-mediated mechanosensory signals, the release of chemokines, cytokines, and osteocytes apoptosis, all initiate the formation of multinucleated osteoclasts to facilitate the formation of a remodeling site [65-67]. Mechanistically, TGF-β, secreted by osteocytes, regulates the activation of bone remodeling by inhibiting osteoclastogenesis. Resorption phase: The second phase is known as the resorption phase, and the active osteoclastic resorption distinguishes this phase of BMU. The degree of osteoclastogenesis is depended upon the systemic and local stimuli and several chemokines. The resorption phase is tightly controlled by the cells of osteoblastic lineages. One such chemokine is monocyte chemoattractant protein-1 (MCP-1) or CCL2, which release from osteoblasts in response to parathyroid hormone (PTH) to recruit preosteoclasts for bone resorption [68, 69]. Moreover, decreased expression of OPG by osteoblasts and increased production of colony-stimulating factor-1 (CSF-1) and RANKL facilitates osteoclasts differentiation resulting in bone resorption [68, 70]. On the resorption site, multinucleated osteoclasts adhere and form a sealing zone to the bone matrix. The osteoclasts-secreted proteolytic enzymes such as cathepsin K and acidic environment (due to the H⁺ ions) facilitate the dissolution of bone minerals [71]. Reversal phase: In the post-resorption phase, the osteoclasts undergo apoptosis, and the reversal phase occurs. The reversal cells (proposed to be monocytic phagocytes from the osteoblast lineage) remove the collagen (demineralized and undigested) from the bone surface [44]. The reversal phase is a transition between bone resorption and bone formation [72-74]. Formation phase: Bone matrix-embedded factors, including bone morphogenetic proteins (BMPs) and TGF-β, are released during the resorption phase. These factors attract the osteoblasts toward the reabsorbed area. Several systemic and growth factors such as PTH, estrogens, glucocorticoids, and vitamin D and their downstream signaling have been proposed to regulate the formation phase [75]. Termination phase: After the formation phase, osteoblasts either undergo apoptosis or are terminally differentiated in osteocytes embedded in the bone matrix. Sclerostin expression is crucial and ceases bone formation [76, 77]. Bone surfaces return to a resting state after mineralization of osteoid and bone-lining cells intercalated with osteomacs. Thus, the unique bone microenvironment and BMU govern the whole physiology of bone remodeling, where genetic and epigenetic modifications play a crucial role.

4. Epigenetic modifications and bone remodeling

Different epigenetic modifications are reported that govern normal gene regulation for bone formation and bone resorption. Epigenetics designates the following significant phenomena: DNA methylation, histone modification, the determinants of these modifications, chromatin remodeling, and the effects of ncRNA (Figure 2) [12]. The cellular machinery involved in these events can be divided into three elements: the writer, the eraser, which are required in adding and removing, respectively, and the reader (a protein involved in identifying different groups for DNA modifications) [78]. Cellular signaling pathways and extracellular stimuli also influence epigenetic modification. Investigating the cellular epigenetic machinery is cardinal to understand the conventional development process and diseases, including cancer. Recent achievements in understanding the mechanism underlying the epigenetic phenomena in different types of cancer have greatly enhanced our interest in epigenetic research.

Figure 2: Epigenetic regulations in bone metastasis and remodeling.

DNA methylation maintains the OPG and RANKL proportion by hypermethylation gene silencing. Histone modifications also regulate OPG and osteoclast differentiation by the NFATC1 gene. Whereas non-coding RNAs control the expression of RUNX2 and NFATC1 genes and modulate the osteoblast and osteoclast differentiation, respectively. Additionally, DNMTs interfere with the methylation-mediated silencing of miRNAs via miR-133a-3p/MAML1/DNMT3A positive feedback loop mechanism, which potentiates cancer cells for bone metastatic niche formation through osteoblast and osteoclast differentiation. PI3K/AKT, MAPK, WNT/β-catenin, TGF-β, JAK/STAT, and mTOR are the major signaling pathways in miRNA-mediated bone metastasis and remodeling.

DNA methylation is the principal epigenetic modification that describes cell type lineage through the regulation of genome stability and gene expression. In this chemical modification, the C5 position of the cytosine in CpG dinucleotide is methylated through a reversible covalent attachment. Though non-CpG dinucleotide cytosine can also be methylated, it is not common and restricted to mammalian cell expression. However, recent studies show that it plays an essential role in gene regulation for stem cells [79, 80]. Among several possible mechanisms of epigenetic modifications, the best characterized is DNA cytosine methylation, where the incorporation of a methyl group to form 5-methylcytosine at CpG dinucleotides [81]. DNA methylation is regulated by an enzyme family called DNA methyltransferase (DNMT), and mainly five isotypes, DNMT1, DNMT2, DNMT3a, DNMT3b, and DNMT3L, catalyze DNA methylation. DNMT transfers one methyl group from S-adenosyl-L-methionine to cytosine [82]. DNMT3A and 3B are mainly involved in de novo DNA methylation [83, 84]. DNMTs also play a decisive role in regulating gene expression, genomic imprinting, inactivating the X chromosome, silencing transposons, and genome stability [85-95].

Besides DNA methylation, epigenetic modification of the DNA sequence also includes histone modification. Histones are the positively charged protein complex termed H2A, H2B, H3A, and H4A [96]. These histones octamer is wrapped in 147 bp of DNA to form nucleosomes, which are the structural unit of chromatin. Histones are post-translationally modified in their flexible tails, and their core domain site is buried in the DNA. These modifications include acetylation, methylation, phosphorylation, ubiquitylation, sumoylation, glycosylation, and ADP-ribosylation. Histone modifications at the site of enhancer and promoter are essential in regulating gene expression and mainly invariant, except a few are dynamic. These modifications are facilitated by specific enzymes with substrate specificity for the histone and can affect locally or globally [97]. Aberrant histone modifications were observed in most cancer cases, and various histone-modifying enzymes inappropriately target the promoter sites of the gene. Other than the promoter site, an extensive range of chromatin modifications in the coding region and non-promoter sequence have been reported in cancer [98]. Histone-modifying enzymes form complexes with other proteins and reside in those multi-protein complexes. Changes in one of the proteins of that complex can cause aberrant modifications of the histone, which may also lead to disease pathogenesis.

ncRNA is a cluster of RNA that remains untranslated and is thought to be non-functional or have a mere role in regulating gene expression. Several studies have proved these ncRNA’s ability to regulate gene expression [98]. Moreover, these ncRNA can interact with chromatin to induce gene silencing [99]. There are several types of ncRNA, including micro-RNA (miRNA), small interfering RNA (siRNA), and long non-coding RNA (lncRNA), having specialized function(s). miRNAs are single-stranded RNAs of approximately 19–24 nucleotides (nt) long, and siRNA derived from the long double-strand DNA fragmented to 19-24nt long siRNA. miRNA and siRNA are the small ncRNAs usually associated with repressing the specific gene by RNA-induced post-translation modification [100]. LncRNA is 200nt long RNA, also associated with gene repression. Several mechanisms of lncRNA responsible for gene repression have been reported, including guiding the epigenetic modifier to the gene locus or competing with the endogenous RNA, acting as a scaffold for protein complexes, and decoying the transcriptional repressor [101, 102]. Several studies have reported the link of the aberrant expression of these ncRNA with the cellular proliferation and apoptotic pathway. Therefore, manipulating these ncRNA expressions and their role can impact tumorigenesis and help to regulate the aberrant physiological events in cancer. Elucidating the role of these ncRNA in tumorigenesis can open a new avenue for developing therapeutics for cancer treatment.

DNA methylation, histone modification, and ncRNA profile may affect bone cells’ osteogenic potential, thereby persuading bone remodeling under different physiological conditions [103]. The expression of well-known factors RANKL and OPG are required for osteoclastogenesis coordination with cytokines and is modulated by CpG island methylation [104]. Similarly, the role of demethylation of histone H3K27me3 has been mainly elucidated in the regulation of osteoclast transcription factor NFATc1 expression [105]. A cell model study of osteogenic differentiation found different histone modifications. Briefly, higher demethylation is associated with H3K4 and increased acetylation with H3 and H4 histone. Moreover, decreased demethylation and trimethylation of histone H3K9 and H3K27 were found, respectively [103].

Among ncRNA-mediated modifications, miRNA is well explored in normal bone remodeling. Numerous miRNAs have also been implicated in the regulation of osteoblast differentiation by targeting genes such as PPAR, BAMBI, CRIM1, DKK1, KREMEN2, SFRP2, and SPARC, SOST, DKK2, and SFRP2. While many other miRNAs target Runx2 and inhibit osteoblastogenesis. miR-21 and miR-214 stimulate mineralization and osteoblastic function, respectively. In addition, few miRNAs act in the late stage of the bone remodeling process by modulating osteoblastic lineage [106, 107]. Similarly, miRNAs, such as miR-21, participates in bone resorption by inducing osteoclast differentiation [108]. Several reports suggest that NFATc1, a known transcriptional factor that regulates osteoclastogenesis, is inhibited by different miRNAs targeting c-fos and MMP2 (miR-29b), NFIA (miR-223), TRAF6 (miR-146a and miR-125a), and RANK (miR-503) and stimulated by miR-148a [109]. In a genome-wide DNase I hypersensitive sites sequencing analysis, a novel network of transcription factors such as CREB1, Jun, ATF2, and ATF4 were identified for osteoclastogenesis [110].

5. Epigenetic regulators in cancer and bone metastasis:

Genomic instability and altered DNA methylation is the hallmark of various diseases, including cancer [111]. More than 90% of the genes that become methylated during tumorigenesis are already repressed in normal cells, suggesting that DNA methylation prevents the reactivation of the repressed gene in tumor cells [112]. In some cases, hyper and hypo DNA methylation (sequence-specific) patterns have been established for gene silencing and gene activation, respectively. These gene-specific DNA methylations are usually guided by transcription factors in normal conditions and altered in disease conditions, including cancer [113]. The second class of epigenetic regulation is covalent modifications of the histones linked to genome instability and chromosome segregation defects [114]. ncRNAs (miRNA, lncRNA, and circRNA) are the third less well-characterized mechanisms that contribute to epigenetic regulation [115].

There are several proposed mechanisms by which DNA methylation regulates gene expression. The first proposed mechanism is the methylation of cytosines at the CpG island of a specific gene promoter site to inhibit the binding of transcription factors. In addition, several transcription factors recognize the sequence containing the repeating CpG dinucleotide at the promoter site. Various evidence shows the disparity in DNA methylation patterns in cancer cells. Broadly, low CpG density regions are hypo-methylated, but at the same time, hypermethylation have also been seen at some CpG islands [12]. Moreover, H3K27me3, in association with several proteins such as embryonic ectoderm development (EED), enhancer of zeste homolog 2 (EZH2) and RING1B, methylates a large portion of the DNA, results in the repression of the gene in cancer cells [116].

Second, genome-wide DNA hypomethylation is correlated with chromosomal instability, activation of genes and retrotransposons, and aberrant gene expression [117]. Recently, partially methylated domains have been discovered with genome sequencing analysis, where DNA hypomethylation is primarily found at mega-base-scale DNA segments. The hypomethylation of cancer associated CpG island is also linked with the overexpression of oncogenic drivers [118]. Hypermethylation at intergenic regions directs the binding of transcription factors with the specific sites of DNA or genome, which regulate the expression of oncogenic driver genes such as AR, MYC, and ERG in metastatic castration-resistant prostate cancer (mCRPC) [119]. Compared to transcribed region, DNA methylation occurs at CpG islands of the regulatory region of genes resulting in gene silencing though methylation in the transcribed region is poorly understood [81].

Dysregulated methylation is generally linked with tumor initiation and metastasis. The pattern of methylation is quite similar in primary and metastatic prostate cancer (PCa) patients but differs highly in healthy controls [119]. Aberrant methylation of the p16INK4a/CDKN2A promoter has been identified in human squamous cell carcinomas and in the early phases of neoplastic transformation. This study demonstrated the potential use of epigenetic alteration as a biomarker to diagnose cancer patients [120]. Interestingly, a high difference has been observed in the primary tumor methylation index vs. metastatic breast cancer (BC) samples. In another study, analysis of BC cell carcinoma patient samples revealed that E-cadherin gene (CDH1) promoter methylation correlates with poor prognosis in BC patients. Loss of CDH1 function in BC patients is more likely caused by epigenetic silencing [121]. Several genes, including RUNX3, NEUROG1, CACNA1G, SFRP2, IGF2 DMR0, hMLH1, and CDKN2A, exhibited substantial alterations between precancerous conditions and malignancies during an examination of colorectal disease methylation patterns [122]. In the human colon cancer cell line HCT116, hMLH1 and CDKN2A are usually mutated and hypermethylated on one allele, leading to the inactivation of important tumor suppressors [123]. Additionally, altered methylations in cancer, methylation-mediated silencing of microRNAs (miRNAs) has also been shown to promote the stemness and migration ability of BC cells via miR-133a-3p/MAML1/DNMT3A positive feedback loop (Figure 2) [124]. Several other metastasis-suppressing miRNAs have been reported in multiple human carcinogenesis and metastasis. Overall, aberrant DNA methylation, directly or indirectly (via miRNA), induced the silencing of several downstream metastasis genes.

For decades, altered epigenetic modifications have been known to occur in tumor initiation and metastasis. Aberrant DNA methylation and histone modifications were described as early events in tumorigenesis. Along with DNA methylation, histone modifications are potent epigenetic mediators which regulate the transcription of several genes in normal as well as in oncogenic conditions [125]. Recently, in post-translational modifications of nucleosomal histones study, the global level of trimethylation of H4K20 (H4K20me3) and acetylation of H4K16 (H4K16ac) have been revealed in several types of cancers [126]. Whereas in primary PCa tissues, a global level of the acetylation of H3K9, H3K18, and H4K12, as well as di-methylation of H3K4 (H3K4me2) and H3R3 (H3R3me2) have been observed [127].

The involvement of histone deacetylases (HDACs) and histone acetyltransferases (HATs) has been revealed in the specific removal of acetyl groups, followed by a reorganization of chromatin and transcriptional repression resulting in tumor development [128]. In a series of HDACs, SIRT1, a nicotinamide adenine dinucleotide (NAD)-dependent histone deacetylase, induces epithelial to mesenchymal transition (EMT) through repression of CDH1 results PCa cell migration and metastasis [129]. Further, SIRT1 knockdown reverts the epithelial phenotype of cancer cells with reduced invasion and metastatic inhibition [130]. Histone methylation either turned on or off transcriptional repressors in a cancer-specific manner, combined with methyltransferases and demethylases. Notably, histone methyltransferases display a specific activity for specific histone and lysine or arginine residue. For example, polycomb protein EZH2 potentiates tumor metastasis via transcriptional repression of the Raf kinase inhibitory protein and acts as a metastasis suppressor in BC and PCa [131]. Moreover, EZH2 causes trimethylation of histone H3 and promotes metastasis by silencing the expression of CDH1 and FOXC1 genes [132]. Similarly, G9a (known as EHMT2) methylates histone 3 (H3) lysine 9 (K9) to silence Ep-Cam, increasing invasion and metastasis [133]. These methylations and acetylations influence the chromatin structure or histone modifications and activate or repress several genes and their downstream signaling cascade.

Integrated epigenome and transcriptome analysis describe the distinct hypermethylation in bone metastasis compared to localized tumors. Recently, a promoter methylation signature of PCa- bone metastasis has been revealed. This signature is associated with androgen receptor (AR) activity and prognosis after androgen deprivation therapy (ADT) in PCa patients with bone metastasis [134]. However, deep whole-genome and transcriptome analysis of 100 castration resistance prostate cancer (CRPC) metastasis elucidated the altered role of DNMT, TET2, and BRAF in hypermethylation and somatic mutations [119]. In an advanced stage of PCa, several oncogenic driver’s genes (AR, ERG, and MYC) expression are regulated by methylation status [119]. PCa cells reactivate the epigenomic programs such as DNA methylation and histone modifications during bone metastasis progression and are mediated by metastasis-specific enhancers at numerous transcription factors such as HOXB13, FOXA1, and NKX3-1 [135].

Another mechanism of epigenetic regulation occurs by ncRNAs, such as miRNA, lncRNA & circRNA. These ncRNAs accomplish numerous regulatory functions by controlling each other expressions. Further, miRNAs regulate the steps of bone metastasis, such as initiation, bone tropism, anchoring, and niche formation into bone. miRNAs act as oncogenes, known as OncomiRs, whereas other miRNAs play a role in tumor suppression. In bone metastasis, altered miRNAs modulate the EMT, acquisition of invasive properties, osteomimicry, bone colonization, and, subsequently, bone remodeling [136]. miRNAs derived from the bone marrow stroma show novel mechanism in bone development and remodeling. Additionally, these bone stromal miRNAs influence the primary carcinomas to bone metastasis by interfering with bone homeostasis together with metastatic tumor cells [137].

In a study, loss of miR-15, 16 and upregulation of miR-21 promote PCa invasion and progression of bone metastasis in a mouse model. These miRNAs activate numerous molecules and pathways, such as TGF-β and Hedgehog signaling, which promote PCa cells for bone metastasis followed by osteolysis [138]. In addition to their role as biomarkers, these miRNAs can be targeted in the prevention of bone metastasis. Several miRNAs are involved in the various signaling pathways viz., PI3K/Akt, MAPK, Wnt, TGF-β, Jak-STAT, and mTOR in different cancer. Transportation and precise delivery are the main challenges in using miRNAs as therapeutic agents for cancer treatment. miR-141 and 219 are identified to inhibit osteolysis in BC bone metastasis in vivo [139, 140]. In addition, miR-203 and 135 may be able to reduce bone metastasis of BC by Runx2 regulation [141]. In contrast, miR-34 is under-expressed in BC bone metastasis to defend against the osteolysis lesions induced by BC cells [142, 143].

Furthermore, histone modifications show influential regulatory phenomena in the metastatic transformation of tumor cells. Histone demethylases are involved in the precise removal of methyl groups in both histone and non-histone substrates in a context-specific manner. The identification of KDM4A, a lysine demethylase, which promotes the recruitment of AP1 to the promoter region of target genes such as VEGF-A, JUN, and FOSL1/2, subsequently metastasis in squamous cell carcinoma. Therefore, histone modifications play important regulation in the metastatic progression of cancer cells [144]. However, improvement has been achieved in identifying precise epigenetic changes associated with tumor progression, though how epigenetic alterations are specified during metastasis is poorly understood.

Epigenetic patterns depend highly on stimuli and local genomic markings; subsequently, it leads to convert in the global epigenetic land scape. Therefore, inappropriate regulation of these modifications can be cascaded during tumor initiation and metastasis. Analysis of metastatic PCa patient samples shows high variation in both DNA copy numbers, specifically in the hypermethylation regions [145]. But it is not yet known that hypermethylation either represents driver events or only passenger alterations.

The etiology of bone metastasis dysregulation is more likely to be an interaction between the environment and the underlying genetic code [146]. Genetic modification of the cancer cell’s DNA is usually detectable, and small-scale studies of cancer epigenetics reveal the profound distortion of the cancer epigenetics landscape. DNA in human tumors either acquire the unique pattern of hypermethylation in the promoter site or overall loss of methylation. Sometimes, the global level of histone modification in the tumor tissues in the whole genome or a specific gene is also associated with cancer [147]. Differences in these epigenetic modifications are also indicative in clinical settings where different patients are in the same cancer stages but represent different outcomes, indicating the difference in epigenetics at the cellular level for a diverse population. Enzymes that modify the histone proteins can also modify the non-histone proteins and affect gene expression. Using the molecular biomarker of dividing the cancer patients into groups is potentially a better idea for better characterization of the disease states. However, cancer is an accumulative result of genetic mutation and multiple altered epigenetic events. Notably, other numerous cellular events, including the tumor microenvironment, cohesively influence tumorigenesis and disease progression.

6. Therapy based on targeting epigenetics modifications

Bisphosphonates, denosumab, and anti-bone resorptive agents are generally used to manage bone metastasis in different types of cancers, though these therapies cannot improve morbidity and mortality in cancer patients [148]. Denosumab (anti-RANKL antibody) is used as adjuvant therapy in most bone metastatic cancers. Recently, a study on BC and colorectal cancer showed disease-free survival in postmenopausal ER-positive patients by treating them in combination with aromatase inhibitor and denosumab [149]. Thus, identifying novel targets for developing therapies against bone metastasis is an urgent need to improve the morbidity and mortality of cancer patients. Understanding the interaction between the tumor microenvironment and cancer cells with concurrent epigenetics modifications may help to develop an effective cancer therapy strategy. Notably, the epigenetic modifier may give rise to cancer chemoresistance [150].

However, this epigenetic modifier-related chemoresistance can be reversed by using epigenome-modifying drugs. Mechanistic understanding of the interaction between the tumor microenvironment and epigenetic modifiers can help to develop anti-cancer therapy [151]. These anti-cancer therapies may include those epigenome modifiers as a monotherapy or in combination with other anti-cancer drugs. Several studies have shown the role of epigenome modifier involvement in the initiation and progression of cancer by altering the oncogene expression by mutating the nucleotide sequence [152]. Epigenetic modification, including global deregulation of DNA methylation, post-translational modification of histone, and lysine acetylation, regulates the gene expression in tumor cells which epigenetic modifier drugs can target. Several mechanisms are involved in the anti-cancer activity of epigenomic drugs, including disrupting anti-apoptotic signaling, regulating the cell cycle, preventing DNA damage repair, disrupting the interaction between the tumor microenvironment and cancer cells, and altered cancer cell metabolism [153-156]. One of the drawbacks of developing epigenomic drugs in cancer is the global gene expression that can interfere with other chemotherapeutic drugs and afterward can promote or inhibit chemoresistance [157]. Mechanistic understanding of altered epigenomic modification due to drug action is important for developing effective combination therapy. Histone acetylase (HAT) and HDAC are two important targets for cancer therapeutics development, and there is some HDAC inhibitor which is already FDA-approved and in clinical trials [158]. These therapeutic targets can also be explored to develop drugs to target the distant site of the cancer metastasis, including bone metastasis.

In search for new therapeutic targets, many groups found altered expression of several genes, including proto-onco and tumor suppressor act as potential drivers of tumor initiation and metastasis [24]. Based on the current knowledge of the interaction between tumor microenvironment and epigenome modifier, several epigenomes modifying drugs have been studied, including inhibitors of DNMT, the polycomb group protein EZH2, EED, and bromodomain extra terminal [159, 160]. In myelodysplastic syndromes (abnormal blood-forming cells in the bone marrow) and several other cancers, DNMTs targeted agents such as azacytidine and decitabine have been tested. Altered DNA methylation and aberrant DNMT expression are associated with tumor initiation and progression in PCa, respectively [161]. Interestingly, DNMT inhibitor disulfiram inhibits PCa cell growth in vitro and in vivo [162]. At present, a combination of DNMT inhibitors, AR pathway inhibitors, and immunotherapy is in clinical development for PCa. Briefly, a phase I/II clinical trial of the DNMT inhibitor azacytidine, in combination with docetaxel and prednisolone, has been reported to improve the disease condition, including improving PSA response in more than 50% PCa patients [163]. HDAC inhibitor, when used with other chemotherapeutic agents, has been demonstrated to improve the patient radiographic progression-free survival. Several EZH2 inhibitors have also been developed as a monotherapy or in combination with AR inhibitors or immunotherapy to treat PCa [164]. Currently, numerous FDA-approved HDAC inhibitors, such as romidepsin, belinostat, panobinostat, and vorinostat, are being used to treat lymphoma and melanoma. These DNMT inhibitors are also tested in PCa progression and metastasis [165]. HDAC inhibitor (panobinostat) in combination with anti-androgen (bicalutamide) enhanced the survival in mCRPC patients [166].

The Polycomb group proteins, including EZH2 and its interacting protein EED, are being explored as therapeutic targets for AR-positive PCa. EED is involved in the direct regulation of AR and downstream signaling molecules along with EZH2 in the context of advanced AR-positive PCa [167]. However, an EZH2 inhibitor, tazemetostat, is granted approval by FDA for EZH2 mutation-positive follicular lymphoma patients [168]. Now, tazemetostat is also being evaluated for patients with mCRPC in combination with an AR inhibitor [169]. Furthermore, CPI-1205, a selective EZH2 inhibitor in combination with ipilimumab (monoclonal antibody targeting CTLA-4), is in a clinical trial for mCRPC patients [170]. In a clinical trial, MAK683 (an EED inhibitor) was tested for metastatic and advanced malignancies [171]. These epi-drugs have been tested in biomarker-unselected populations to understand the efficacy of AR-positive and negative CRPC patients. Several other epi-drugs have been developed to target chromatin readers family proteins, such as bromodomain-containing protein (BRD) 2, BRD3, and BRD4 [172]. Again, ZEN3694, an inhibitor of the bromodomain and extra-terminal (BET) family of proteins, was evaluated in combination with AR targeting enzalutamide in mCRPC patients [173]. Similarly, a small molecule and potent inhibitor of bromodomain of CBP/p300, CCS1477 reduces levels of the H3K27Ac histone modification specifically at enhancers and downregulates AR, AR-splice variants, and c-Myc expression in CRPC [174]. Remarkably, till now, several ongoing clinical trials exist for specific biomarkers, such as small molecules for reverse epigenetics. Understanding cancer-specific epigenetic therapies and their consequences are under crucial consideration to improve the therapeutic outcome and combination approaches in the future.

7. Concluding remarks:

Bone is a basic framework for the human body. In addition to the locomotion and protection of internal organs, bone has various metabolic functions; particularly, it maintains the body’s mineral balance. Bone tissue always maintains a dynamic equilibrium between bone resorption and bone formation. Altered homeostasis between bone formation and resorption creates pathological abnormality in the skeleton. Epigenetic dysregulation is one of the significant non-genetic factors that participate in the initiation and progression of numerous diseases, including cancer and bone metastasis. Despite improvement in survival and outcome in cancer patients, metastatic disease, especially bone metastasis, remains the leading cause of morbidity and mortality. The most reported cancers disseminated in the bone are breast, prostate, lung, thyroid, and renal cell carcinoma.

While the role of epigenetic regulation of bone remodeling is relatively well established, understanding the epigenetics dysregulation in bone metastasis is currently minimal. Additional mechanisms, including epigenetic modifications in bone metastasis, remain to be analyzed. The extensive implementation of epigenetic microarrays and next-generation whole genome sequencing will establish a broad epigenetic spectrum of bone metastatic cancer cells, which will be helpful to the identification of specific epigenetic markers for bone metastatic in a cancer-specific manner and will help to identify the clinical outcome.

Further research needs to understand epigenetic regulation, bone remodeling, and metastatic niche formation, which may reveal new targets for diagnosis, prevention, or therapeutics. In earlier studies, the heterogeneity and plasticity of metastatic tumors posed challenges for more accurate and personalized treatment. We believe that high throughput epigenome mapping will be an advanced tool for genome and epigenome mapping and may prove helpful for screening and testing for personalized drug development for advanced and bone metastasis patients. Thus, therapy for advanced cancer patients with bone metastasis can be customized for each patient with far more efficacy and fewer off-target consequences.

Abbreviations:

ADT

androgen deprivation therapy

ALP

alkaline phosphatase

AR

androgen receptor

BC

breast cancer

BLCs

Bone lining cells

BMU

basic multicellular unit

circRNA

circular RNA

CaSR

calcium-sensing receptor

CSF-1

colony-stimulating factor-1

DNMT

DNA methyltransferase

EED

embryonic ectoderm development

ERs

estrogen receptors

EZH2

enhancer of zeste homolog 2

HATs

histone acetyltransferases

HDACs

histone deacetylases

HSCs

Hematopoietic stem cells

lncRNA

long-non-coding RNA

MCSF

macrophage colony-stimulating factor

miRNA

micro-RNA

MSCs

mesenchymal stem cells

MK

megakaryocytes

MCP-1

monocyte chemoattractant protein-1

mCRPC

metastatic castration-resistant prostate cancer

NAD

nicotinamide adenine dinucleotide

ncRNA

non-coding RNA

NK cells

natural killer cells

nt

nucleotides

OPG

Osteoprotegerin

OSX

osterix

PCa

prostate cancer

PTH

parathyroid hormone

RANKL

receptor activator of nuclear factor-κB ligand

RUNX2

runt-related transcription factor 2

TGFβ-1

Transforming growth factor beta-1