Deciphering and Targeting Epigenetics in Cancer Metastasis

Abstract

Once cancer metastasizes to distant organs like the bone, liver, lung, and brain, it is in an advanced stage. Metastasis is a major contributor to cancer-associated deaths. Countless molecules and complex pathways are involved in the dissemination and colonization of cancer cells from a primary tumor at metastatic sites. Establishing the biological mechanisms of the metastatic process is crucial in finding open therapeutic windows for successful interventions. Emerging evidence suggested a variety of epigenetic regulations were identified to regulate cancer metastasis. Here we summarize the procedures and routes of cancer metastasis as well as the roles of epigenetics including ncRNA, DNA methylation, and histone modifications in common metastases. Then we further discuss the potentials and limitations of epigenetics-related target molecules in diagnosis, therapy, and prognosis.

1. Introduction

Cancer is the leading cause of death all over the world, accounting for nearly 10 million deaths in 2020 according to the World Health Organization (WHO). Metastatic cancer, the main contributor to high mortality, results in more than 90% of cancer death. This is because when cancer metastasizes to distant organs, especially the bone, liver, lung, and brain, this secondary tumor is formed. And this kind of tumor is difficult to remove despite the various systemic treatments including chemotherapy, screening, and immunotherapy. Efforts from doctors, researchers, and other aspects to promote cancer killing over the past years have paid off in some countries and in some cancers. Since 1991, the cancer death rate has fallen continuously in the United States. Up to 2018, the total mortality rate fell by 31%. Yet the mortality rate has been increasing in other places, such as China [1]. In most cases, many patients with metastatic cancer will face death within 5 years after their diagnosis, which is a horrible thing. Therefore, knowing the mechanism of cancer metastasis to treat metastasis is meaningful for patients, and is a challenging project for oncologists and clinical investigators.

Exploring the physiological mechanisms of the metastatic process is the foundation to find successful interventions. In the beginning, body fluids were thought to be responsible for tumor metastasis. In 1929, James Ewing proposed a theory that believed the anatomical structure of the vascular system contributes to metastasis and dissemination of cancer cells [2]. This view prevailed for decades. Nonetheless, the most classic and now popular is the “seed and soil” hypothesis proposed by Stephen Paget in 1889 [3]. Over the next few decades, cancer scientists gradually enhanced our knowledge of this mechanism based on molecular and cellular aspects. During this process of metastasis, countless molecules, and complex pathways, including epigenetic regulations, are involved in the dissemination and colonization of cancer cells from a primary tumor at metastatic sites. Epigenetics, a reversible process, refers to the study of heritable changes in gene expression without DNA sequence changes. Increasing studies of epigenetic regulation suggest that such regulations without altering the DNA sequence are critical for the normal physiological activities and the maintenance and development of tissue-specific gene expression in mammals [4]. The location of modified residue and the degree of methylation determines whether the transcriptional activation or repression. For example, the trimethylation of lysine 4 on histone H3 (H3K4me3) can be observed at the promoters of activated genes transcriptionally, yet trimethylation of H3K9 (H3K9me3) and H3K27 (H3K27me3) is enriched at repressed gene promoters transcriptionally [5].

Moreover, the importance of epigenetic changes in early tumorigenesis and cancer metastasis also has been shown, including non-coding RNA (ncRNA), DNA methylation, and histone modifications. Some such examples are increased N6-methyladenosine (m6A) modification of c-Myc mRNA enhances tumor cell growth, invasion, and tumorigenesis in animal models [6]. Upregulated Lysine Demethylase 6B (KDM6B) facilitates lung metastasis in osteosarcoma by modulating the H3K27me3 demethylation level of lactate dehydrogenase (LDHA) [7]. In addition, the enhancer of zeste homolog 2 (EZH2), the histone methyltransferase (HMT) of H3K27, is increased in cancers and promotes tumor metastasis [8, 9]. overexpressed long non-coding RNA (lncRNA) H19 enhances the migration of malignant cells and promotes the occurrence of epithelial to mesenchymal transition (EMT) in endometrial cancer [10].

Given the distinguished functions of epigenetics in cancer progression, and numerous crucial pathways and key biomarkers discovered by researchers, various potent and specific inhibitors targeting biomarkers have been studied and applied in clinics for treating cancer, since azacytidine, the first epigenetic drug approved by Food and Drug Administration (FDA) in 2004. In addition, as inhibitors of DNA methyltransferase (DNMT) enzymes (also termed hypomethylating agents), decitabine (5-aza-2′-deoxycytidine) and guadecitabin are the most extensively applied epigenetic therapies to kill various cancer cells, such as mutated monocyte in acute myeloid leukemia (AML) [11]. Several histone deacetylase (HDAC) inhibitors also have been extensively applied to anticancer (i.e., vorinostat, romidepsin, panobinostat, and belinostat), gaining the approval of FAD for hematological malignancies based on the activity of the single drug [12]. What’s more, histone methyltransferase (HMTs), like EZH2, protein arginine methyltransferase 6 (PRMT6), SET domain bifurcated 1 (SETDB1), SUV39H1, and disruptor of telomeric silencing 1-like (DOT1L), also are the targets of cancer treatment. Note that, several small-molecule inhibitors of EZH2 (i.e.,tazemetostat, SHR2554, MAK683) and DOT1L (i.e., EPZ-5676) have entered into clinic phases [13].

Based on the increasing knowledge about the mechanism of metastasis and drug development, the prognosis and survival in patients with cancer will gain an effective improvement in clinical outcomes. That is because using the vulnerabilities of metastatic cancer cells and the properties of metastatic tumor microenvironments are a great entry point to prevent cancer metastasis. As a result, several related drugs involved in cancer metastasis to treat cancer came into being. For instance, gefitinib and erlotinib are tyrosine kinase inhibitors (TKIs) that target activating epidermal growth factor receptor (EGFR) mutations and can improve overall survival by inhibiting metastasis in non-small cell lung cancer (NSCLC) [14, 15]. In general, cancer metastasizes to bone, liver, lung, and brain at an advanced stage, which is difficult for clinicians to destroy the secondary tumor. This is an urgent task and of great significance to patients. Thus, cancer biologists are working to deepen their understanding of epigenetic mechanisms in cancer metastasis to develop better therapy.

Although these drugs mentioned above contribute to clinical improvements in cancer patients, there exist some challenges. The first obstacle is that tumor heterogeneity, one of the characteristics of malignant tumors, which leads to the differences in immune characteristics, growth rate, aggressive ability, sensitivity to drugs, prognosis, and other phenotypic aspects after taking the same drugs. That means precision medicine and personalized medicine are the points of future medical development. Besides, the vast majority of genetic changes of epigenetics are inactivating mutations that are inherently difficult to treat, even though cancer biologists are designing drugs to interfere with adaptive mechanisms. Epigenetics provides a novel insight for researchers to improve the prognosis and survival of patients.

In this review, we summarize the procedures and routes of cancer metastasis as well as the roles of epigenetics including lncRNA, DNA methylation, and histone modifications in common metastases including bone, liver, lung, and brain, followed by discussing about potentials and limitations of epigenetics-related molecules in diagnosis, therapy, and prognosis.

2. Cancer metastasis

At present, metastasis is known as the result of a complex multistep cell-biological process collectively known as the invasion-metastasis cascade. It involves the changes of cancer cells in the physical position from the primary tumor to distant or adjacent sites of dissemination, and the colonization of the “seed cells” by adapting to the alien tissue microenvironments. More specifically, during metastatic progression, the first change is the cellular adhesion and morphology of cancer cells are reduced by EMT. Then, cancer cells improved the capability of invading the normal tissue surrounding (local invasion). Next, cancer cells make a way into (intravasation) and out of (extravasation) systemic circulation such as the lymphatic or circulatory system to land at distant sites. In this step, surviving cancer cells are termed circulating tumor cells. Lastly, the cancer cells proliferate and colonize in an unknown tissue microenvironment of different distant organs (Figure 1).

Figure 1.

Overview of metastatic Cascade. During metastatic progression, the first change is that the cellular adhesion and morphology of cancer cells are reduced by epithelial-mesenchymal transition (EMT). Then, the ability of cancer cells to invade the surrounding normal tissue (local invasion) is increased. Next, cancer cells make a way into (intravasation) and out of (extravasation) systemic circulation such as the lymphatic or circulatory system to land at a distant site. In this step, surviving cancer cells are termed circulating tumor cells. Lastly, the cancer cells proliferate and colonize an unknown tissue microenvironment of different distant organs. This figure was created with BioRender.com.

2.1 The invasion-metastasis cascade

2.1.1 Local invasion

The local invasion of cancer cells is the foundation for metastatic cancer process. Local invasion refers to the entry of cancer cells into the surrounding tumor-associated stroma, subsequently entering the adjacent normal parenchymal tissue. To invade the stroma, cancer cells must first break the basement membrane (BM) located at the interface of the epithelial tissue and connective tissue. The specialized extracellular matrix (ECM) can exchange material and regulate tissue growth, differentiation, and regeneration [16]. When these cancer cells invade the stroma, they have to be confronted with diverse cancer-associated stromal cells, including myofibroblasts, fibroblasts, adipocytes, endothelial cells, and plenty of bone marrow-derived cells (BMDCs) (i.e., macrophages, mesenchymal stem cells) [17]. Then, these stromal cells are able to enhance the aggressiveness of cancer cells through various cytokines. Secretion of interleukin-6 (IL-6) by cancer-associated fibroblasts (CAFs), stimulates the migration and invasiveness of colorectal cancer cells by the STAT3-LRG1 axis [18]. Increased IL-4 in endothelial cells can lead to enhanced invasiveness of liver cancer cells via the ERK-AKT signaling axis [19]. Besides, colorectal cancer cells secret IL-4 to promote M2-like tumor-associated macrophage (TAM) polarization [20]. These findings suggest there exists a positive-feedback loop in the tumor microenvironment, which is that cancer cells maintain high inflamed surroundings, and these stromal cells further enhance the malignant characteristics of cancer cells.

Researchers have observed various patterns of invasion when cancer cells infiltrate the substrates of adjacent tissues. Due to the dissemination of cancer cells, as individuals and collectives, these researchers divide migrations into individual cell migration and collective cell migration. Both types of migration are simultaneously present in many cancers [21]. In cancer progression, the plastic changes of numerous cancer cells are shown by morphological and phenotypical conversions, such as EMT and its reverse process the mesenchymal-epithelial transition (MET) [22], the collective-amoeboid transition (CAT) [23], the mesenchymal-amoeboid transition (MAT) [24]. Among these conversions, EMT has been increasingly considered a crucial and indispensable stage in the cancer metastatic process over the last decade [22], despite the studies of Seyfried et al. in the VM mouse model of systemic metastasis suggesting that EMT is unnecessary for the initial cancer metastasis [25, 26].

EMT is a cellular process activated by master transcription regulators, includingZEB1, ZEB2, Twist, Slug, and Snail, which enhance cell motility and migration ability to invade stroma. Besides, transforming growth factor (TGF)-β has proved to be a strong inducer of EMT by collaborating with other signaling pathways, especially the RAS-MAPK cascade [27]. Moreover, increasing emerging evidence shows the potent roles in invasion and EMT of many long noncoding RNAs (lncRNAs), such as lncRNA MEG3, lncRNA PNUTS, and lncRNA MIR100HG [28, 29, 30]. During this process, cells lost epithelial characteristics and markers like E-cadherin and cytokeratin, instead of gaining mesenchymal characteristics and markers like vimentin, N-cadherin, and fibronectin [22]. In addition to cancer metastasis, EMT has been involved in different cancer stages, including cancer initiation, malignant progression, cancer stemness, and drug resistance [31].

2.1.2 Intravasation

It is a vital and indispensable step for cancer cells to disseminate to distant organs during which cancer cells infiltrate into the vascular or lymphatic wall and then enter circulation, becoming circulating tumor cells (CTCs) and potential metastatic seeds. The formation of new blood vessels around cancer cells has a great influence on cancer cells entering the circulatory system, thus understanding the various mechanisms of neoangiogenesis stimulated by cancer cells in local microenvironment will help us comprehend intravasation. Vascular endothelial growth factor (VEGF), a highly bioactive functional glycoprotein, promotes blood vessel growth and lymphatic vessels, which plays an irreplaceable role in angiogenesis. However, the neo-vasculatures generated by cancer cells increase capillary permeability compared with the blood vessels produced by normal cells and tissues [32]. During the lung metastasis of breast cancer, VEGF/VEGF receptor 2 (VEGFR2) and its target proteins such as ERK1/2, Src, and FAK regulate neo-angiogenesis and blood vessel permeability to enhance metastasis [33].

On the other hand, a bunch of studies reveal intravasation can be improved by boosting the penetrability of cancer cells to pass the barrier of endothelial cells. For example, secretion of epidermal growth factor (EGF) by TAMs enhances the intravasation of breast cancer cells [34]. Additionally, the TGF-β enhances mammary cancer intravasation by increasing carcinoma cell penetration of micro-vessel walls or more generally strengthening invasiveness [35]. What’s more, in melanoma, the migration of cancer cells to endothelial cells and intravasation are promoted via endothelial-derived SLIT2 protein and its receptor ROBO1 [36]; activated Notch1 receptors (N1ICD) can promote neutrophil infiltration into the tumor, the intravasation of cancer cells and postsurgical metastasis [37]. In the study of Wei et al., increased IL-6 from TAMs is observed and can promote the invasiveness of cancer cells through the STAT3/miR-506-3p/FoxQ1 axis, then increases CCL2 level to boost the recruitment of macrophages. Besides, the authors suggested that there exists a feedback loop between TAMs and cancer cells, which was essential for the EMT and intravasation into the blood vessels [38].

2.1.3 Circulation

Once cancer cells have successfully entered lymph and blood, these malignant cells have the chance to disseminate throughout the body. In blood and lymphatic vessels, these cancer cells must escape the killing of immune cells and physical damage from hemodynamic shear forces to survive. In general, CTCs are in a dormant state that can cause relapse and poor prognosis for patients. This is because conventional surgery, radiotherapy, and chemotherapy are powerless against these CTCs in the blood, lymph, and body fluids, as well as dormant cancer cells, further leading to a decrease in immunity and the rapid growth and metastasis of hidden CTCs.

Many studies have verified the prognostic role and value of CTCs in the early and metastatic stages of cancer by measuring biomarkers [39]. An informative meta-analysis including 1847 patients with colorectal cancer under chemotherapy studied by Huang et al., demonstrated the high expression of CTCs in the bloodstream has a positive correlation with decreased progression-free survival (PFS) (hazard ratios = 2.500, 95% CI [1.746–3.580], P < 0.001) [40]. Moreover, CTCs in blood samples of 100 patients with head and neck squamous cell carcinoma were enriched and isolated and the PFS and overall survival of these patients were observed and recorded. The result showed a worse prognosis like decreased PFS and overall survival in CTCs-high patients [41]. Rink et al. also observed patients with ≥1CTCs C per 7.5 ml of blood in distant metastatic bladder cancer shortened the time of disease recurrence and cancer-specific death, resulting in worse clinical outcomes [42]. With the improvement of technologies and the depth of research, plenty of CTCs-related biomarkers are uncovered. At present, a set of biomarkers has been applied to detect CTCs in various cancers. Lin D et al. summarized the CTC-related biomarkers in different cancers [43]. EpCAM as the most common marker can be found in most cancer (i.e., breast cancer, liver cancer, prostate cancer, kidney cancer, melanoma, bladder cancer), which is because most cancers originate from the epithelium [44]. Just like EpCAM, human epidermal growth factor receptor-2 (HER-2), estrogen receptor (ER), prostate-specific membrane antigen (PSMA), and folate receptor (FR) also have been applied to detect CTCs in some cancers, with outstanding clinical significance [45, 46, 47, 48, 49].

In addition to CTCs-related biomarkers, the mechanism by which CTCs escape the detrimental shear stress and anoikis in the circulatory system is becoming clearer. There is evidence that CTCs in the blood can stay away from immune cells’ killing to increase survivability by bounding tightly to blood constituents like neutrophils, myeloid-derived suppressor cells (MDSCs), CAFs, or platelet [50, 51]. A few years ago, Szczerba et al. found the concentration of CTCs and neutrophils have a significant correlation in animal models and patients with breast cancer, which displays greater metastatic potential and higher gene expression involving cell proliferation. They thought the binding of CTC and neutrophil is possibly mediated by vascular cell adhesion molecule [52]. Besides, Spicer et al. suggested neutrophils could directly adhere to CTCs by the neutrophil Mac-1/ICAM-1, which becomes a bridge between cancer cells and the liver to accelerate CTCs extravasation and colonization [53]. Neutrophils can also enhance metastasis in an indirect manner by trapping CTCs in the circulation through neutrophil extracellular traps (NETs) [54]. In several in vivo experiments, liver or lung NETs were found to collect cancer cells to promote distant metastases by a transmembrane protein named coiled-coil domain containing 25 (CCDC25) to activate the ILK-β-parvin pathway, leading to enhance cell motility [55].

TAMs play crucial roles in the mechanical adhesiveness and endurance of CTCs, which contribute to the formation of protective cell clusters and the resistance to shear stress [56]. Liu et al. proposed that CTCs interacting with adhesive immune cells like MDSCs could create a defensive shield to allow evasion of immune surveillance, facilitating distant metastatic lesions [57]. Sprouse et al. found reactive oxygen species (ROS) from MDSCs could activate the Notch pathway in CTCs, promoting CTCs proliferation [58]. In addition, CAFs could protect CTCs from the fluid shear forces in the peripheral blood via intercellular contact and soluble derived factors in prostate cancer [59]. As an important component in blood, platelet also supports the survival and metastasis of CTCs in a CTCs-platelet cluster manner. Platelets have been shown to help CTCs evade attack by NK cells by creating a surface shield and normal MHC-I [60], or by downregulating natural killer group 2 member D (NKG2D) and its ligands, further stimulating glucocorticoid-induced tumor necrosis factor receptor-related protein (GITR) to exert functions in NK cells [61, 62, 63]. Furthermore, platelets involve the adhesion process of endothelial cells. The attachment between platelets and CTCs is enhanced by integrin αIIbβ3 and P-selectin (platelet adhesion receptors), in which supports the strong adherence of CTCs to the endothelial wall [64, 65, 66].

2.1.4 Extravasation

Cancer cells extravasate from a vascular lumen into tissues such as the lung, liver, and brain by passing through the endothelial cell and pericyte layers. Extravasation is comparable to intravasation in that it is morphologically similar to invadopodia but mechanistically different. Certain cell types in the primary tumor microenvironment, such as TAMs [34], can initiate intravasation, but these same cells do not have the same promoting function in the extravasation process of disseminated CTCs. Indeed, macrophage phenotype and function differ between primary and metastatic tumor locations. For example, macrophage seeding at distant places is VEGFR+, CCR2+, CXCR4-, Tie2-, and the subpopulation of macrophage at perivascular macrophage is phagocytic [67].

CTCs must overcome the physical barriers of the microvascular wall to extravasate. According to some studies, primary tumors have been shown to release substances that interfere with these distant microenvironments and cause vascular hyperpermeability. Secreted protein angiopoietin-like-4 (Angptl4), as well as the pleiotropically active proteins like NOX4, MMP-1, and MMP-9, disrupt pulmonary vascular endothelial cell-cell junctions, allowing colorectal cancer cells to extravasate into the lungs [68]. Angiopoietin2 (Angpt2), MMP-3, MMP-10, placental growth factor, and VEGF, all of which are secreted by many types of primary tumors, can induce pulmonary hyperpermeability prior to the arrival of cancer cells in the lungs, allowing CTCs to extravasate more easily [69, 70]. Finally, by secreting VEGF, inflammatory monocytes recruited to pulmonary metastases via CCL2-dependent processes increase breast cancer cell extravasation in the lung [71].

Interestingly, whereas Anglptl4 improved the extravasation of breast carcinoma cells in the lung, it did not increase the extravasation or intravasation efficiency of these same breast cancer cells in the bone [72]. As a result, Anglptl4 selectively and only increases extravasation within the lung tissue environment.

2.1.5 Colonization and metastatic growth

Cell-nonautonomous mechanisms required to transform a foreign microenvironment into a more friendly niche may be required for disseminated tumor cells to emerge from hibernation and begin active proliferation. For example, the growth of other inactivated disseminated tumor cells might need to stimulate BMDCs to enter the circulation system, as well as the followed recruitment of these cells to the metastatic location; in some situations, the process may be activated through systemic signals such as osteopontin (OPN) or SDF-1 produced by cancer cells, [73, 74].

Alternatively, because the body is in a constant state of homeostasis, dormant cancer cells could continue to proliferate without a net increase. The reasons driving such high rates of attrition are unknown, however, a lack of disseminated tumor cells to initiate neoangiogenesis has been hypothesized as one possible explanation. Prostate tumor cell-secreted prosaposin (Psap) may limit metastatic colonization by increasing the expression of the anti-angiogenic factor thrombo-spondin-1 in stromal cells, which is consistent with this theory [75]. Angpt2, on the other hand, promotes the metastatic colonization of breast and pancreatic cancer by improving the infiltrating capability of myeloid cells to support the vascularization of metastatic nodules [76].

Numerous genes promote the metastatic colonization of cells in breast cancer to bone, lung, brain, or liver, which have recently been discovered. These genes are able to adapt and overcome incompatibilities between the special development procedure of disseminated cancer cells and the demands from foreign tissue milieu, in parallel, researchers come up with the idea that these genes could control organ-specific metastatic tropism. The osteoclastic cytokine IL-11 is an excellent example of this, IL-11 works through a receptor activator for nuclear factor kB (RANK), which disrupts the normal crosstalk between osteoclasts and osteoblasts [77]. Moreover, it strengthens metastatic tumor growth in breast cancer and osteolysis by JAK1/STAT3/c-myc signaling pathway rather than in a RANKL-dependent manner [78].

Similarly, the Notch ligand Jagged1 enhances the osteolytic bone metastases in breast cancer cells by boosting osteoclast activity through IL-6 released by osteoblasts [79]. By encouraging osteoclast action, IL-11 and Jagged1 are able to cause osteolysis and release the rich deposits of growth factors from the bone matrix. The fact that genes identified as candidate mediators of breast cancer cell metastatic colonization in bone, lung, brain, or liver show very little overlap, illustrates the idea that different tissue microenvironments are needed to be organ-specific for metastatic colonization.

2.2 Routes of cancer metastasis

2.2.1 The circulatory system

Despite the fact that lymphatic diffusion of cancer cells is a key prognostic marker for cancer progression, spreading through the blood circulation seems to be the main mechanism of dispersal of metastatic carcinoma cells. Based on intravital imaging studies, tumor cells can travel toward blood arteries. Li et al. injected metastasized breast cancer cells in mice and found that these cells move toward arteries, illustrating that metastatic cells have the ability of directional migration toward blood streams [80]. Morphologically, compared with non-metastatic cells, metastatic cells are more round, and this kind of morphology boost both their ability to spread, enter, and colonic tumor vasculature [81, 82]. These findings show that tumor cells with an elongated morphology may need to change their shape to become more rounded in order to successfully intravasate and endure shear forces within blood arteries. It’s tempting to think that higher cortical acto–myosin contraction, which promotes the rounded morphology, also allows the cortical cytoskeleton to withstand more mechanical stress. Tumor cells have been found to form part of blood vessel walls in other imaging studies. GFP-labeled tumor cells have been shown to constitute part of the lumen of blood arteries using in situ imaging [80]. This behavior is likely to be linked to tumor cells expressing genes that are normally restricted to endothelial cells. In many circumstances, entering the bloodstream could be as simple as separating tumor cells from the walls of blood vessels.

2.2.2 The lymphatic system

Tumor cell entry into the lymphatic system is another major mechanism for tumor cell dissemination. Even in the early phases of tumor formation, changes in lymphatic artery architecture have been seen, under the aid of VEGFC lymphangiogensis can be established quickly using these cells [83]. The lymphatic vasculature can be examined and the interaction of tumor cells with lymphatics can be seen in live tissue by injecting dyes or fluorescent tracers [84]. Cancer cells may extend protrusions via holes in lymph vessel walls before entering the vessels, according to electron microscope photos [85, 86]. To learn more about the cytoskeletal architecture of intravasation cells and their interactions with the lymphatic endothelium, time-lapse imaging should be possible. Interstitial pressure within the tumor affects lymphatic outflow [87]. Interstitial flow can produce autocrine gradients that signal through CC chemokine receptor 7 (CCR7) to induce cell migration in the same direction as the interstitial flow, according to in vitro research. In vivo, however, it is uncertain whether lymphatic channel movement is influenced by interstitial pressure. Trapped breast cancer cells can be discovered in the subcapsular region of lymph nodes after entering lymphatic channels [88]. However, clinically, lymph node metastases might stay in this place or spread to other parts of the node. The admission of tumor cells into the central sections of lymph nodes, as well as their interactions with immune cells, has yet to be captured in high resolution, but when it is, it will likely provide fascinating discoveries.

3. Epigenetics in cancer metastasis

Due to the devilishness of cancer metastasis, understanding how cancer cells acquire and maintain metastatic characteristics is critical. However, metastasis-specific genetic alterations cannot be discovered in most exome or genome sequencing investigations. Reversible epigenetic pathways control important phases in metastasis, which can be targeted to prevent and treat metastatic illness in increasingly emerging data. ncRNA, DNA methylation, and histone changes are only a few of the epigenetic processes that have been discovered to modulate the cancer metastasis process. Large-scale chromatin structural changes, such as enhancer reprogramming and chromatin accessibility to transcription factors, have been revealed to be a possible driving force of cancer metastasis in diverse malignancies in recent years. Given that numerous researchers have reviewed the function of epigenetic markers in different stages of metastasis [89, 90, 91], we will concentrate on well-defined particular metastatic locations such as bone, liver, lung, and brain in various malignancies over the last 5 years.

3.1 Epigenetics in bone metastasis

Bone is one of the most common sites of metastasis for a variety of solid tumors, including lung, liver, breast, and prostate, with bone metastases being seen in 70% of metastatic prostate and breast cancer patients [92]. Unfortunately, once cancer has progressed to the bone, it is seldom treated and is associated with countless complications such as discomfort, increased fracture risk, and hypercalcemia. This finding has prompted scientists interested in bone and cancer biology to investigate the bone, revealing a number of mechanisms, including epigenetically related elements that promote cancer spread to the bone (Table 1).

Cancer types	Epigenetic types	Biomarkers	Pathways	References
Lung cancer	ncRNA	lncRNA-SOX2OT	miRNA-194-5p/RAC1 signaling axis	[93]
		miRNA-660-5p	miR-660-5p/SMARCA5/RANKL axis	[94]
		miR-574-5p, miR-328-3p, and miR-423-3p	Wnt/β-catenin pathway	[95]
		miR-106a	miR-106a/TP53INP1	[96]
		miR-139-5p	miR-139-5p/Notch 1	[97]
		miR-17-5p	PTEN/PI3K/Akt	[98]
		miR-365	NKX2-1/EGFR/PI3K	[99]
		miR-192-5p	miR-192-5p/ TRIM-44	[100]
		miR-886-3p	miR-886-3p-PLK1/TGF-β1 pathway	[101]
	DNA methylation	DNMTs	miR-886-3p-PLK1/TGF-β1 pathway	[101]
		DNMTs	WIF-1	[102]
Prostate cancer	ncRNA	lncRNA NEAT1	CYCLINL1/CDK19/NEAT1-1	[103]
		miR-940	miR-940/ARHGAP1 and FAM134A	[104]
		miR-181b-5p	miR-181b-5p/ Oncostatin M axis	[105]
		lncRNA HOXA11-AS	HOXB13/lncRNA HOXA11-AS/ IBSP	[106]
		lncRNA PCAT6	PCAT6/IGF2BP2/IGF1R axis	[107]
		lncRNA NEAT1	miR-205-5p/RUNX2/ SFPQ/PTBP2 axis	[108]
		miR-378a-3p	miR-378a-3p/ Dyrk1a/Nfatc1/Angptl2	[109]
		lncRNA MAYA	ROR1-HER3-lnc RNA MAYA/Hippo-YAP pathway	[110]
		miR-214	miR-214/TRAF3	[111]
		miR-124	miR-124/IL-11 axis	[112]
		miR-218	miR-218/COL1A1	[113]
		miR-125b	miR-125b/ HIF1A/PTGS2	[114]
		miR-429	miR-429/CrkL/MMP-9	[115]
		miR-21	miR-21/PDCD4	[116]
		miR-19a	miR-19a/IBSP	[117]
		lncRNA SNHG3	miR-1273 g-3p/BMP3 axis	[118]
	DNA methylation	DNMTs	HGF/Met receptor signaling; E-cadherin, Twist transactivation	[119]
Liver cancer	ncRNA	lncRNA 34a	PHB2/DNMT3a/ miR-34a; TGF-β/Smad4 pathway	[120]
		lncRNA H19	H19/p38MPAK/OPG; H19/miR200b-3p/ZEB1	[121]

Table 1.

Epigenetic biomarkers in bone metastases of various cancers.

Notes: ncRNA: non-coding RNA; TP53INP1: tumor protein 53-induced nuclear protein 1; NKX2-1: NKX homeobox-1, EGFR: epidermal growth factor receptor, PI3K: phosphoinositide-3-kinase; TRIM-44: tripartite motif-44; PHB2: Prohibitin 2; TGF-β: transforming growth factor β; MPAK: mitogen-activated protein kinase; OPG: osteoprotegerin; ROR1: receptor tyrosine kinase (RTKs)-like orphan receptor-1; YAP: yes-associated protein; TRAF-3: TNF receptor-associated factor-3; COL1A1: collagen type I alpha 1 chain; PTGS2: prostaglandin-endoperoxide synthase 2; HIF1A: hypoxia-inducible factor 1 alpha subunit; CrkL: v-crk avian sarcoma virus CT10 oncogene homolog-like; PDCD4: programmed cell death 4; IBSP: integrin-binding sialoprotein; RUNX2: runt-related transcription factor 2; SFPQ: splicing factor proline- and glutamine-rich; and PTBP2: polypyrimidine tract-binding protein 2.

3.1.1 Lung cancer

In the last year, Ni et al. extracted exosomes from the plasma of non-small cell lung cancer (NSCLC) patients with or without bone metastasis. They found exosomal lncRNA-SOX2OT enhanced bone metastasis of NSCLC by targeting the miRNA-194-5p/RAC1 signaling axis in osteoclasts [93]. In organ-specific metastatic lung cancer cells, Ai et al. observed miR-660-5p involved tumor progression and bone-specific metastasis by nm23-H1/miR-660-5p/SMARCA5/RANKL axis [94]. Yang et al. identified an exosomal microRNA cluster that has an association with bone metastasis. Specifically, in this cluster miR-574-5p was down-regulated, miR-328-3p and miR-423-3p were up-regulated in patients with bone metastasis, which suppressed or activated the Wnt/β-catenin pathway [95]. Han et al. observed that upregulated miR-106a promoted bone metastasis by targeting tumor protein 53-induced nuclear protein 1 (TP53INP1), including cell migration, death, and EMT [96]. Xu et al. found miR-139-5p was downregulated in serum to facilitate lytic bone metastasis by targeting Notch1 [97]. In addition, miR-17-5p promotes osteoclastogenesis through the PI3K/Akt pathway via targeting PTEN [98]. Liu et al. revealed that miR-365 was reduced in patients with bone metastasis of NSCLC, and miR-365 could suppress lung metastasis via NKX2-1/EGFR/PI3K axis [99]. Zou et al. demonstrated that increased miR-192-5p in patient serum inhibited lung cancer metastasis, possibly by reducing TRIM44 [100]. Loss of miR-886-3p expression was mediated by DNA hypermethylation of its promoter in both cultured small cell lung cancer (SCLC) cells and tumor samples. What’s more, upregulated miR-886-3p greatly inhibited bone metastasis [101]. The downregulation of Wnt inhibitory factor 1 (WIF-1) expression was linked to hypermethylation of its promoter, which increased lung metastasis [102].

3.1.2 Liver cancer

In liver cancer bone metastasis, Zhang et al. revealed the molecular function of lncRNA 34a regulated bone metastasis. Mechanistically, lncRNA 34a epigenetically suppressed miR-34a level via the recruitment of DNMT3a by Prohibitin 2 (PHB2) to methylate miR-34a promoter and histone deacetylase (HDAC) 1 to promote histones deacetylation. On the other hand, miR-34a regulated Smad4 through the transforming growth factor-β (TGF-β) pathway, impacting the downstream genes (i.e., connective tissue growth factor (CTGF) and IL-11) associated with bone metastasis [120]. Huang et al. identified lnRNA H19/p38 mitogen-activated protein kinase (MPAK)/ osteoprotegerin (OPG) and lncRNA H19/miR200b-3p/ZEB1 axes contributed to hepatocellular carcinoma bone metastasis [121].

3.1.3 Breast cancer

In breast cancer, the Hippo-YAP pathway was controlled by a ROR1-HER3-LncRNA signaling axis to govern bone metastases [110]. Both osteoclastic miR-214/ TNF receptor-associated factor-3 (TRAF-3) pathway and dysregulated miR-124/IL-11 axis were devoted to the understanding of breast cancer metastases to the bone [111, 112]. A study pointed to a concept in which cancer-derived miR-218 impairs osteoblast function by directly targeting collagen type I alpha 1 chain (COL1A1) and regulating inhibin βA expression [113]. miR-125b may reduce the effect of hypoxia-inducible factor 1 alpha subunit (HIF1A), which is known to enhance metastatic spread by upregulating prostaglandin-endoperoxide synthase 2 (PTGS2) [114]. To investigate the influence of miR-429 on the metastatic bone environment in vivo, Zhang et al. created an orthotopic bone degradation model and a left ventricle implantation paradigm. The levels of V-crk sarcoma virus CT10 oncogene homolog-like (CrkL) and MMP-9 were negatively influenced by miR-429 [115]. Exosomal miR-21 generates from breast cancer cells promotes osteoclastogenesis by modulating the levels of the protein programmed cell death 4 (PDCD4). Furthermore, the amount of miR-21 in breast cancer patients with bone metastases is considerably greater in serum exosomes [116]. Exosomal miR-19a and integrin-binding sialoprotein (IBSP) are highly increased and secreted from bone-tropic estrogen receptor-positive (ER+) breast cancer cells, resulting in a milieu conducive to colonization in the bone [117]. Teng et al. identified many key lncRNAs such as lncRNA RP11-317-J19.1 related to bone metastasis in breast cancer [122]. By influencing the miR-1273 g-3p/BMP3 axis, LncRNA SNHG3 regulates BMSC osteogenic development in breast cancer bone metastases [118].

In breast cancer, the level of DNA methylation is increased to further regulate Wwox, following to stimulate HGF/Met receptor signaling and E-cadherin, downregulating Twist transactivation, leading to bone metastasis [119].

3.1.4 Prostate cancer

Wen et al. analyzed the m6A status using patient samples and bone metastatic patient-derived xenografts (PDXs) with prostate cancer through m6A high-throughput sequencing, and they found 4 credible m6A sites on lncRNA NEAT1-1. Besides, NEAT1-1 acted as a bridge to strengthen the combination between CYCLINL1 and CDK19 and promoted the Pol II ser2 phosphorylation in the promoter of RUNX2, leading to the development of bone metastatic prostate cancer [103]. By targeting ARHGAP1 and FAM134A, miR-940 boosted osteogenic differentiation of human mesenchymal stem cells [104]. miR-181b/Oncostatin m axis also contributes to prostate cancer bone metastasis by altering osteoclast differentiation [105]. To promote the bone-specific metastasis of prostate cancer, HOXA11-AS controlled the expression of chemokines, integrins, and associated genes like IBSP in collaboration with HOXB13 [106]. lncRNA PCAT6 enhances prostate cancer bone metastasis and tumor growth by upregulating IGF1R expression via increasing IGF1R mRNA stability through the PCAT6/IGF2BP2/IGF1R pathway [107]. lncRNA NEAT1/miR-205-5p/RUNX2/SFPQ/PTBP2 axis and miR-378a-3p/ Dyrk1a/Nfatc1/Angptl2 axis are also devoted to bone metastasis [108, 109].

3.2 Epigenetics in liver metastasis

3.2.1 Esophageal squamous cell carcinoma

Tang et al. explored the function of lncRNA LOC146880 in esophageal squamous cell carcinoma (ESCC) progression. The result of in vivo and in vitro experiments showed LOC146880 sponged miR-328-5p to regulate fascin actin-bundling protein 1 (FSCN1) activating MAPK signaling pathway, resulting in liver metastasis [123] (Table 2).

Cancer types	Epigenetic types	Biomarkers	Pathways	References
Liver metastases
ESCC	ncRNA	lncRNA LOC146880	LOC146880/miR-328-5p/FSCN1/MAPK axis	[123]
Breast cancer	ncRNA	miR-190	ZEB1-miR-190-SMAD2 axis	[124]
		miR-1204	miR-1204/VDR	[125]
		cirRNA ciRS-7	ciRS-7/miR-1299	[126]
		circROBO1	circROBO1/ KLF5/FUS	[127]
Colorectal cancer	ncRNA	miR-221; miR-222	—	[128]
Brain metastases
Breast cancer	ncRNA	lncRNA XIST	EMT and MSN/c-Met	[129]
		miR-10b	—	[130]
		miR-576-3p	—	[131]
		lncRNA BCBM	lncRNA BCBM /JAK2/STAT3	[132]
		circBCBM1	circBCBM1/miR-125a/BRD4 axis	[133]
		lncRNA-CCRR	lncRNA-CCRR/connexin 43	[134]
		miRNA let-7d	PDGF/PDGFR axis	[135]
		miR-802-5p; miR-194-5p	—	[136]
		miR-132-3p; miR-199a-5p; miR-150-5p; miR-155-5p	—	[137]
		miR-211	SOX11/NGN2 axis	[138]

Table 2.

Epigenetic biomarkers in liver and brain metastases of various cancers.

Notes: ESCC: esophageal squamous cell carcinoma; FSCN1: fascin actin-bundling protein 1; VDR: vitamin D receptor; KLF5: Kruppel like factor 5; bHLH transcription factor 2; CDK6: Cyclin Dependent Kinase-6; STAT3: signal transducer and activator of transcription-3; HDAC: histone deacetylase 1; ZEB1: Zinc Finger E-Box Binding Homeobox 1; ROCK1: Rho associated coiled-coil containing protein kinase 1; CRYAB: αB-crystallin; DNMTs: DNA methyltransferase; ERα: estrogen receptor alpha; TET2: ten-eleven translocation 2; IRX1: iroquois homeobox 1; PDK1: phosphoinositide-dependent kinase-1; ST7L: suppression of tumorigenicity 7 like; and BRD4: bromodomain containing 4.

3.2.2 Breast cancer

In breast cancer, the TGF network in liver metastasis can be explained by the ZEB1-miR-190-SMAD2 axis [124]. miR-1204 inhibits vitamin D receptors (VDR), which promotes epithelial-mesenchymal transition and metastasis [125]. As a ceRNA of miR-1299, circular RNA ciRS-7 promotes lung and liver metastases by targeting MMPs [126]. Wang et al. found circROBO1 was upregulated to boost tumor development and liver metastasis in vivo. Further research revealed the mechanism that circROBO1 upregulated KLF5 by sponging miR-217-5p, allowing KLF5 to activate FUS transcription, hence promoting circROBO1 back splicing [127].

3.3 Epigenetics in lung metastasis

3.3.1 Osteosarcoma

LncRNA-CASC15 promotes lung metastasis in osteosarcoma by regulating EMT via the Wnt/β-catenin signaling pathway [139] (Table 3). MIR205HG also can drive lung metastatic osteosarcoma via regulating the axis of miR-2114-3p/twist family bHLH transcription factor 2 (TWIST2) [140]. miR-485-3p regulated by lncRNA MALAT1 inhibites osteosarcoma glycolysis and lung metastasis by directly suppressing c-MET and AKT3/mTOR signaling, meanwhile, MALAT1 also facilitated lung metastasis of osteosarcomas through miR-202 sponging [141, 142]. Besides, Chen et al. also observed that the LOC100129620/miR-335-3p/CDK6 signaling promoted the lung metastasis of osteosarcoma by mediating the osteosarcoma cells proliferation, macrophage polarization, and angiogenesis [144]. The lncRNA NEAT1/miR-483/STAT3 axis also exerts a crucial role in regulating the lung metastasis process in osteosarcoma, especially in EMT [145]. miR-326 inhibited by SP1/HDAC1 has a great impact on proliferation and metastasis of osteosarcoma through stimulating SMO/Hedgehog pathway [146]. The miR-19a/RhoB/AKT1 network and miR-491/αB-crystallin (CRYAB) axis also may help us to better know the lung metastatic mechanism of osteosarcoma [147, 149]. In Ewing sarcoma, miR-130b directly targets ARHGAP1 to activate a lung metastatic CDC42-PAK1-AP1 positive feedback loop [157].

Cancer types	Epigenetic types	Biomarkers	Pathways	References
Osteosarcoma	ncRNA	lncRNA-CASC15	Wnt/β-catenin signaling	[139]
		MIR205HG	MIR205HG/miR-2114-3p/TWIST2 axis	[140]
		lncRNA MALAT1	miR-485-3p/c-MET; miR-485-3p /AKT3/mTOR signaling; lncRNA MALAT1/miR-202; miR-129-5p/RET/ PI3K-Akt axis	[141, 142, 143]
		lncRNA LOC100129620	LOC100129620/miR-335-3p/CDK6 signaling	[144]
		lncRNA NEAT1	lncRNA NEAT1/miR-483/STAT3 axis	[145]
		miR-326	Sp1/HDAC1/miR-326/SMO/Hedgehog axis	[146]
		miR-19a	miR-19a/RhoB/AKT1	[147]
		lncRNA DANCR	miR-335-5p and miR-1972/ROCK1	[148]
		miR-491	miR-491/CRYAB	[149]
	DNA methylation	DNMTs	ERα	[150]
		DNMTs	SPARCL1/ WNT/β-catenin signaling	[151]
	DNA demethylation	—	TET2/IL-6	[152]
		—	IRX1/ CXCL14/NF-κB signaling	[153]
Breast cancer	ncRNA	linc-ZNF469-3	miR-574-5p-ZEB1 axis	[154]
		cirRNA ciRS-7	ciRS-7/ miR-1299	[126]
		lncRNA MIR31HG	miR-575/ ST7L	[155]
Colorectal cancer	Histone acetylation	CBP	CBP-DOT1L/ RNF8/H3K79	[156]
Ewing Sarcoma	ncRNA	miR-130b	miR-130b-AP-1/ CDC42-PAK1-AP1 axis	[157]
Gastric cancer	ncRNA	lncRNA GMAN	ephrin A1	[158]
		Lnc RNA MIR17HG	Wnt/β-catenin signaling	[159]
		lncRNAs AC093818.1	PDK1	[160]

Table 3.

Epigenetic biomarkers in lung metastases of various cancers.

Notes: TWIST2: twist family bHLH transcription factor 2; CDK6: Cyclin Dependent Kinase-6; STAT3: signal transducer and activator of transcription-3; HDAC: histone deacetylase 1; ZEB1: Zinc Finger E-Box Binding Homeobox 1; ROCK1: Rho associated coiled-coil containing protein kinase 1; CRYAB: αB-crystallin; DNMTs: DNA methyltransferase; ERα: estrogen receptor alpha; TET2: ten-eleven translocation 2; IRX1: iroquois homeobox 1; PDK1: phosphoinositide-dependent kinase-1; and ST7L: suppression of tumorigenicity 7 like.

Additionally, Lillo et al. found estrogen receptor alpha (ERα) was not expressed in osteosarcoma due to promoter DNA methylation. They took Decitabine, a DNA methyltransferase (DNMTs) inhibitor to activate ERα, further inhibiting osteosarcoma growth and lung metastasis [150]. In primary osteosarcoma cells, increased IL-6 expression regulated by DNA demethylation of the promoter of ten-eleven translocation 2 (TET2) promotes lung metastasis in osteosarcoma [152]. Secreted protein acidic and rich in cysteine (SPARCL1) downregulated by epigenetic promoter DNA methylation in osteosarcoma promotes lung metastasis via canonical WNT/β-catenin signaling activated through stabilization of the WNT–receptor complex [151]. Hypomethylation of iroquois homeobox 1 (IRX1) in osteosarcoma cell lines substantially affected metastatic behavior in vitro, including migration, invasion, and resistance to anoikis and influenced lung metastasis in animal models by upregulating CXCL14/NF-B signaling, according to another study [153].

3.3.2 Breast cancer

In the study by Wang et al. they showed that linc-ZNF469-3 accelerated lung metastasis of triple-negative breast cancer (TNBC) via miR-574-5p-ZEB1, which may be acted as a potential and promising prognostic marker for TNBC patients [154]. MIR31HG, a long noncoding RNA that sponges miRNA-575 to control ST7L expression, suppresses hepatocellular carcinoma proliferation and metastasis [155].

3.3.3 Colorectal cancer

Colostrum basic protein (CBP), a histone acetyltransferase (HAT), mediates DOT1L K358 acetylation and has a positive correlation with colorectal cancer stages. DOT1L acetylation confers DOT1L stability by blocking RNF8 binding to the protein and subsequent proteasomal degradation, but it has no effect on the enzyme’s activity. DOT1L can catalyze the H3K79 methylation of genes involved in epithelial-mesenchymal transition, such as SNAIL and ZEB1, once stabilized [156].

3.3.4 Gastric cancer

GMAN, a long non-coding RNA, is upregulated in stomach cancer patients and is linked to overall survival and metastasis process. It inhibits the translation of ephrin-A1 mRNA by binding to GMAN-AS in a competitive manner [158]. Interferon regulatory factor-1 (IRF-1) suppresses gastric cancer spread by suppressing Wnt/−catenin signaling and downregulating the MIR17HG-miR-18a/miR-19a axis to inhibit gastric cancer lung metastasis [159].

3.4 Epigenetics in brain metastasis

Through activation of EMT- and MSN-mediated up-regulation of c-Met, the loss of lncRNA XIST increases breast cancer brain metastasis by boosting both stemness and aggressiveness of tumor cells [129] (Table 2). Yoo et al. proved the therapeutics function of miRNA-10b by targeting the brain metastases process in breast cancer [130]. JAK2-binding lncRNA BCBM promotes breast cancer brain metastasis by regulating STAT3 [132]. circBCBM1 is involved in breast cancer brain metastasis via circBCBM1/miR-125a/BRD4 axis [133]. By modulating connexin 43 expression, dysregulation of lncRNA-CCRR contributes to breast cancer brain metastases through intercellular coupling [134]. Loss of miRNA let-7d and active hypoxia- inducible factor-1 (HIF1) signaling enhances breast cancer brain metastasis via platelet-derived growth factor (PDGF), while pharmacologic inhibition of PDGF receptor (PDGFR) inhibits brain metastasis, implying new therapeutic possibilities [135]. miR-132-3p, miR-199a-5p, miR-150-5p, miR-155-5p, miR-802-5p and miR-194-5p from breast cancer cells also were identified the important role in brain metastasis [136, 137]. In triple-negative breast cancer, miR-211 regulates brain metastatic selectivity via the SOX11/NGN2 axis [138].

4. Therapeutic potentials and limitations

Epigenetic drugs are chemicals that alter DNA and chromatin structure, promoting the disruption of transcriptional and post-transcriptional modifications, primarily by regulating the enzymes required for their establishment and maintenance, and reactivating epigenetically silenced tumor-suppressor and DNA repair genes [161]. The development of treatment techniques incorporating epigenetic medicines, which focus on the cancer epigenome to generate pharmacological molecules that could restore a “normal” epigenetic landscape, is a developing field of drug discovery [161]. Epigenetic medicines target the enzymes that are required for the maintenance and establishment of epigenetic alterations, with the inhibition of DNMTs and HDACs being the most common technique [161]. The epigenetic alterations caused by these medications can regulate the temporal and spatial expression of genes [162], and they have ramifications for the regulation and dysregulation of physiological and pathological processes. Because epigenetic markings are tightly linked to the type of tumor and stage of disease, as well as individual genetic variation, such as in personalized medicine [163, 164], they have a lot of promise to give molecular biomarkers for diagnosis and treatment alternatives for cancer therapy [165].

The FDA has approved six new epigenetic medicines and multi-drug regimens for use in clinical cancer treatment. Some side effects will happen, so novel epigenetic therapeutic compounds are continually being tested in preclinical research, as well as clinical trials for the development and release of new medicines, for cytotoxicity, and pharmacological characteristics, and to better understand their mechanism of action. The majority of epigenetic medication studies are focused on cancer detection, therapy, and prognosis.

NcRNAs have shown new promise and insight as therapeutic targets for cancer treatment and preventing cancer metastasis in vivo preclinical models of metastatic illness. Research has shown that lncMAYA, MALAT1, and lncARSR have all been targeted for in vivo suppression using ASOs in mice models to alleviate the burden of metastatic disease [110, 166, 167, 168]. When targeting lncRNAs with ASO therapies, however, it will be vital to ensure minimal off-target effects [169, 170], which could offer additional challenges given the decreased quantity of lncRNA transcripts in vivo. ASOs disrupt target RNAs by premature transcriptional termination [171, 172], in addition to RNase H-mediated destruction of mature RNA, according to new findings, which should be taken into account when estimating the efficacy of ASO therapies. The biggest problem is the species conservatism in ncRNA, especially lncRNAs. Animal testing is also required before conducting clinical trials.

5. Conclusions

Cancer metastasis is a common cause of death. The role of epigenetics in the etiology of metastases cannot be ignored. To overcome cancer and its metastasis, many methods and technologies are applied to developing new drugs. In parallel with the development of specific and potent small-molecule inhibitors, some novel and cutting-edge technologies like proteolysis-targeting chimeras (PROTACs), RNA interference (RNAi), clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9-based genome editing, and artificial intelligence (AI)-based drug design, in chemical biology show huge potentials for cancer treatment, which allows to screening therapeutics targeting almost all kinds of molecules, like proteins and epigenetic regulators [173].

Acknowledgments

This section is supported by the Natural Science Foundation Council of China (81700780 and 81922081), the Department of Education of Guangdong Province (2021KTSCX104), and the Guangdong Basic and Applied Basic Research Foundation (2022A1515012164), and the Science, Technology, and Innovation Commission of Shenzhen (JCYJ20210324104201005).

We thank the administrative assistant (Ms. Yufang Zuo) for providing help and support.

Conflict of interest

The authors declare no conflict of interest.