Epigenetic regulation and targeting of ECM for cancer therapy

Abstract

The tumor microenvironment (TME) composed of different types of cells embedded in extracellular matrix (ECM) has crucial effects on cancer growth and metastasis. ECM is made of a variety of proteins that provide structural support to the cells and regulate biological functions by modulating the cross talk among cells, thus effecting tumor growth and progression. In this mini-review, the author discusses epigenetic modifications that regulate the expression of fibrous ECM proteins and glycoproteins and the prospects of targeting them for cancer therapy.

INTRODUCTION

The biological milieu surrounding cancer cells in living organisms, referred to as the tumor microenvironment (TME), consists of multiple types of cells embedded in extracellular matrix (ECM) surrounded by blood vessels. The TME is capable of undergoing various changes that have profound effects on the ability of cancer cells to grow, metastasize, and respond to therapy. These changes are modulated by signaling molecules including growth factors, cytokines, chemokines, galectins, and other factors. The role of ECM in the TME is not only to provide specific architecture and three-dimensional (3-D) organization, but also to regulate many biological functions by modulating crosstalk between individual cells and between cells and the environment. Changes in the expression of fibrous proteins that constitute the main structural component of the ECM as well the glycoprotein that facilitate cell-cell recognition and signaling have been linked to tumor progression and metastasis. In this mini-review, the author describes epigenetic alterations that regulate the expression of fibrous ECM proteins and glycoproteins and the prospect of targeting them for cancer therapy.

EPIGENETIC ALTERATIONS AND THEIR ROLE IN CANCER

Cancer cells are well known to acquire multiple genetic mutations that contribute to tumor growth and development; however, recent studies demonstrated that they can also acquire numerous epigenetic alterations with cancer-promoting effects (1, 2). Epigenetic changes are not the changes in the DNA but the modifications that alter DNA accessibility and chromatin structure, which effects gene expression (3). These epigenetic alterations include changes in DNA, such as methylation, and in histone proteins, for example, methylation, acetylation, phosphorylation, ubiquitination, etc. (4) and are important for cancer initiation and progression (5), as they influence multiple aspects of tumorigenesis including regulation of tumor-suppressor genes and oncogenes and modulation of signaling pathways that promote cancer cell growth, invasion, and metastasis (6). Epigenetic alterations in cancer cells have also been associated with drug resistance and response to treatment (7). Recent efforts to target epigenetic regulators for cancer therapy have led to new and effective treatments for many types of cancer (1, 6, 8). Furthermore, there is emerging evidence that epigenetic modification plays important roles in regulating the expression of fibrous ECM proteins and glycoproteins (9). Previous studies have provided us the information on the different mechanisms that regulate ECM proteins expression including microRNA (10, 11). In this mini-review, the author describes in details the epigenetic modifications of DNA and histone that regulate the expression of fibrous ECM protein and glycoproteins (Fig. 1).

Figure 1.

Epigenetic regulation and targeting of ECM for cancer therapy. Illustration of the epigenetic modifications that regulate the expression of various fibrous ECM proteins and glycoproteins and different ways by which these candidates are targeted for cancer therapy. This Illustration was prepared using BioRender.com. ECM, extracellular matrix; LOX-1, lysyl oxidase 1.

EPIGENETIC REGUALTION OF THE GENES THAT ENCODE FIBROUS ECM PROTEINS AND GLYCOPROTEINS

ECM is a complex 3-D structural milieu that surrounds and supports cells within mammalian tissues (12). ECM is composed of proteoglycans attached to specific GAGs (glycosaminoglycans), glycoproteins, and fibrous proteins. Proteoglycans are highly branched moiety composed of repeating unit of proteins conjugated to polysaccharides and are classified based on the specific GAGs. They are also the most abundant core component of the ECM and are mainly present in the connective tissue. Proteoglycans act as tissue organizer and play a role in regulating signaling pathways by binding to secreted molecules such as growth factors (13–15). Glycoproteins (fibronectin and laminin), conversely, are composed of nonrepeating unit of proteins conjugated to saccharides and are mainly localized on the cellular membrane and help in facilitating cell-cell recognition and promote cross talk, whereas fibrous proteins, which include collagens and elastin, provide mechanical strength and constitute the structural component of ECM (13, 15). ECM proteins, including fibrous proteins, bind to cells through transmembrane receptors called integrins, which serve as a link to transmit extracellular signals to the cytoskeleton and trigger signaling pathways. The complex multidynamic interrelationships among ECM proteins and integrins determine the mechanical characteristics of the ECM and can profoundly affect tumor progression and metastasis (16).

Collagens are the most abundant fibrous proteins in both, the ECM and the human body (17). Previous studies reported that several collagens are overexpressed in cancer cells and regulate key steps of tumor development (13, 18). Epigenetic regulation of collagens has also been observed in different types of cancer (19, 20). Previous studies performed on human endometrial stromal cells (T-HESCs) showed that Collagen Type I Alpha 1 Chain (COL1A1) expression was regulated by promoter histone deacetylation and histone deacetylase (HDAC) inhibitor treatment led to COL1A1 promoter hyperacetylation and increased gene expression (19, 20). Moreover, DNA methylation has also been shown to play a role in regulating collagen expression. For example, a previous study showed that collagen genes possess abundant “CpG” islands near their transcriptional start sites. An analysis performed using Methylated DNA Immunopreciptation (MeDIP) demonstrated that methylation patterns of the COL1A1 and COL4A1/A2 genes differ between cancer cells and normal cells (21, 22). Similarly, DNA hypermethylation results in repression of collagen type VII in mammary and prostate cancer cells (21). In another study, histone modifications such as H3K4me2 and H3 hypermethylation (pan-acetylated histone H3) in glioblastoma were associated with increased levels of several collagens, which in turn promoted tumor metastasis (23). These studies highlight the importance of epigenetic modifications such as DNA methylation and methylation and acetylation of histone in regulating the expression of collagens.

Epigenetic regulation of elastin and fibronectin has been linked to several types of cancer. Elastin expression in solid cancers contributes to matrix stiffness, which is associated with malignancy and the metastatic phenotype (24). In colorectal cancer, hypomethylation in the coding region of the elastin gene was found to result in overexpression of elastin (25). Fibronectins are large multidomain glycoproteins that are overexpressed in several human cancers, playing important roles in tumor growth and metastasis, immune regulation, and modulation of therapeutic responses (26). Epigenetic regulation of fibronectin using the DNA methyltransferase (DNMT) inhibitor 5-aza-CdR led to a decrease in fibronectin expression in human uterine leiomyoma primary (HULP) cells (27).

Laminins are the most abundant glycoproteins of the basement membrane of ECM and are universally present. Laminins protect the cancer cells from undergoing apoptosis and are involved in multiple processes important for cancer cell proliferation and migration (28). Laminin-332 (previously known as laminin-5) regulates cell adhesion and migration in gastric carcinoma (29). Laminin-332 is a heterotrimeric protein that consists of three polypeptide chains, Α3, Β3, and C2, which are the products of the LAMA3, LAMB3, and LAMC2 genes, respectively (30). Methylation of “CpG” islands in the promoter regions of LAMB3, LAMC2, and LAMA3 caused transcriptional silencing of the polypeptide chains in gastric cancer tissues. The methylation frequency of LAMA3 was higher than LAMB2 and LAMC2 genes, which was reversed by the treatment with demethylating agent 5-aza-2'-deoxycytidine (5-aza-dC) (29). Similar results were reported in nonsmall cell lung cancer (NSCLC) and small cell lung cancer (SCLC): methylation of LAMA3, LAMB3, and LAMC2 caused silencing of the corresponding Laminin-332 polypeptides in 60%–80% of cases in SCLC cell lines and in 15%–60% of cases in NSCLC cell lines and impacted lung cancer growth (31). These studies show that frequent epigenetic inactivation via methylation is the major cause of laminin inactivation in several cancer types.

Integrins are transmembrane receptors that constitute part of the ECM and regulate cancer growth and metastasis by facilitating cell-ECM and cell-cell interactions (32). The expression and function of integrins and integrin-linked kinase are regulated by epigenetic factors. Furthermore, epigenetic regulation of the expression of integrins β1, β2, β4, β6, β8, α4, α6, and α10 has been shown to have an important effect on cancer cells (33). Inhibition of histone deacetylase (HDAC) by trichostatin A (TSA) treatment upregulated the expression of several integrins such as α4, β2, and β6, suggesting that hyperacetylation of histone proteins leads to transcriptional activation of integrin genes (34). Other studies showed that decreased H3K9Ac and H3K4me3 and increased H3K27me3 and DNA methylation in the promoter region of integrin α4 lead to the repression of α4 integrin expression (35, 36). In addition, hypermethylation in the α4 integrin promoter was found to cause repression of α4 integrin expression in gastric cancer and cholangiocarcinoma (37). These findings showcase the role of epigenetic modifications such as acetylation and methylation in regulating the expression of integrins, which affects malignancy and the metastatic phenotype.

TARGETING ECM PROTEINS FOR CANCER THERAPY

ECM proteins play a key role in regulating ECM-mediated signaling that promotes cancer growth and metastasis. Several therapeutic strategies have been developed to target both fibrous ECM proteins and glycoproteins for effective cancer therapy (Fig. 1). Collagen inhibition combined with chemotherapy and radiotherapy has shown promising results in clinical trials. In one study, a collagen-binding epidermal growth factor receptor single-chain Fv antibody fragment was fused with a single-chain fragment of cetuximab antibody (CBD-scFv) and found to have enhanced tumor-inhibitory activity against collagen-rich cancers in vivo (38).

Another way to inhibit collagens is by using enzyme collagenase, which degrades collagens (24). There are several ways by which collagenase can be directly administered to cancer cells for therapeutic purposes. Nanoparticles (poly-lactic-co-glycolic acid nanoparticle and lipid-bilayer mesoporous silica nanoformulations) coated with collagenase that precisely targets the tumor cells have been generated and effectively used for cancer therapy (39). Other methods for collagenase delivery to cancer cells include the use of oncolytic herpes simplex virus vector (40) and the collagozome (a liposomal formulation of collagenase type I) (41). Once collagenase is delivered to sites of tumor growth, it elicits potent anticancer effects in two ways—first, by degrading the ECM collagen that promotes tumor metastasis, and second, by disassembling the dense collagen stroma and thus allowing the penetration of other anticancer drugs into the tumor such as in pancreatic ductal adenocarcinoma (PDAC) (41). The use of collagenase-based anticancer therapy is limited by the cancer stage at which it is administered. For collagenase treatment to be effective, it has to be administered at an early stage when the cancer has significant matrix stiffness but has not been yet invaded or metastasized, because collagen degradation at advanced clinical stages promotes metastasis instead of having anticancer effects (24).

Another way to inhibit collagens is the use of small-molecule inhibitors that directly target collagens. For example, one study showed that SR-T100, a glycoalkaloid extracted from Solanum, inhibits the promoter activity and expression of COL11A1 in drug-resistant ovarian cancer cell lines (42). Another study showed that LY2157299, a small-molecule inhibitor of TGFβ receptor I (TβRI), inhibits the expression of COL11A1 in tumor stroma (24, 43). Furthermore, the AKT inhibitor SC66 was recently found to inhibit the expression of COL11A1 in cancer cells (44), and the AZD compound AZD5653, which is also an AKT inhibitor was found to inhibit the transcription and promoter activity of COL11A1 (45). The antihypertension drug losartan was shown to inhibit collagen synthesis in animal models (46) and in a clinical trial (clinicaltrials.gov identifier: NCT01821729). Furthermore, clinical trial studies concluded that neoadjuvant therapy with the combination of losartan, FOLFIRINOX, and chemoradiotherapy provided downstaging of locally advanced PDAC and an R0 resection rate of 61% (47). These findings advance our knowledge of the potential of targeting collagens with antibodies, enzymes, and inhibitors for cancer treatment.

Elastin is also an important fibrous ECM protein that has been targeted for cancer therapy. A few studies have suggested that elastin peptides are potent inhibitors of tumor metastasis and can be exploited for anticancer therapeutics (48); however, a number of other studies found that peptides targeting elastin have pro-oncogenic effects that increase tumor growth and invasion (49). In one study with promising results, inhibitors of lysyl oxidase 1 (LOX-1), an enzyme involved in elastin and collagen cross linking (50) that promotes cancer growth and metastasis (51), were tested for anticancer effects in different kinds of cancer, including triple negative breast cancer (TNBC) and colorectal cancer (52–54). Another study showed that a LOX-blocking antibody in combination with gemcitabine suppressed PDAC tumor growth and metastasis (55). These studies show that targeting of lysyl oxidases (LOXs) might be an effective way to treat cancer.

Fibronectin is a predominant ECM protein that is overexpressed in several cancers and affects cancer proliferation, migration, differentiation, and survival (56). Fibronectin is present in two different isoforms: extra domain A (EDA) and extra domain B (EDB) (56). Several new therapies targeting EDB are emerging in clinics for cancer treatment. A murine monoclonal antibody called BC-1 was generated against the cryptic domain adjacent to human fibronectin EDB and fused with murine IL12 (huBC-1-mIL-12). The fusion antibody had strong growth-inhibitory effects on various kinds of cancer xenografts in immunodeficient mice, including colon cancer and prostate cancer (57, 58). In a subsequent clinical trial (clinicaltrials.gov identifier: NCT00625768), 46% of patients with metastatic renal cell carcinoma or metastatic malignant melanoma who were treated with huBC-1-mIL-12 (AS1409) were in stable condition after six or more cycles of treatment (59). Another antibody that targets human fibronectin EDB, L19, was fused with IL-2 (L19-IL-2) and found to significantly improve the tumor-inhibitory efficiency of IL-2 in tumor-bearing mice (60). Phase I/II clinical trials (clinicaltrials.gov identifier: NCT01058538) showed that L19-IL2 provided clinical benefit leading to stable disease in patients with advanced renal cell carcinoma and can be safely and repeatedly administered at the recommended phase II dose of 22.5 Mio IU IL2 equivalent in patients with advanced solid tumors (61). In another phase II clinical trial, a combination of L19-IL-2 and the chemotherapy agent dacarbazine showed encouraging clinical results in patients with metastatic melanoma (clinicaltrials.gov identifier: NCT01055522) (62). These results provide a basis for combining fibronectin inhibitors with approved anticancer drugs to treat advanced solid cancers such as renal cell carcinoma (RCC), metastatic melanoma, and others.

The high-affinity laminin 37/67 kDa receptor, also known as laminin receptor precursor/laminin receptor (LRP/LR), plays a vital role in tumor growth and metastasis. LRP/LR is overexpressed in several types of cancer including leukemia and breast, colon, cervical, lung, ovarian, gastric, thyroid, prostate, and uterine cancers (63). This overexpression promotes enhanced LRP/LR-laminin-1 interaction, resulting in increased tumor aggressiveness (64). Based on this observation, several methodologies have been developed to interrupt the laminin-LR interaction and thus inhibit tumor growth and progression. For example, the small molecule BC-K-YH16899 was discovered to suppress tumor metastasis by preventing the laminin-LR interaction (65). Epigallocatechin-3-gallate (EGCG), a small molecule derived from green tea, was also found to inhibit cancer growth by interrupting the laminin-LR interaction via competitive-binding to receptor LAMR⁶⁷ (66). EGCG is currently being tested further in a phase I study in patients with colorectal cancer (clinicaltrials.gov identifier: NCT02891538). In another study, an IgG-type polyclonal-blocking antibody called P4G was generated against the laminin-binding protein of 37 kDa (37LBP) and shown to effectively inhibit the colonization and growth of human fibrosarcoma HT1080 cells in the lungs of mice, suggesting that targeting the r37LBP may serve as an attractive strategy to prevent cancer metastasis (67). Similarly, a monoclonal IgG1-iS18 antibody that interferes with the LRP/LR-laminin-1 interaction was found to inhibit the invasion of breast and esophageal cancer cells (68). Furthermore, an immunogenic LAMR tumor-associated antigen called oncofetal antigen immature laminin receptor protein (OFA-iLRP) was successfully used as a tumor antigen for vaccine-based therapies in preclinical studies (65). In a phase I/II clinical study (clinicaltrials.gov identifier: NCT00879489), autologous dendritic cells loaded with OFA-iLRP showed promising results against metastatic breast cancers. These findings suggest that targeting laminins has the potential to be an effective treatment strategy in a variety of cancers.

Integrins are being targeted in several ways to treat cancer, including with antibodies, inhibitors, and peptides. In this context, several antibodies targeting different integrins have been generated. Etaracizumab is a humanized monoclonal antibody targeting integrin αvβ3 that is currently in phase II clinical trials for colorectal, melanoma, prostrate, and thyroid cancers (32). Intetumumab, a humanized monoclonal antibody targeting integrins αvβ1 and αvβ3 has been used to treat melanoma and prostate cancers (32) and is in a phase I/II clinical trial as combination therapy with dacarbazine for treatment of stage 4 melanoma (clinicaltrials.gov identifier: NCT00246012). Abciximab, a humanized monoclonal antibody targeting integrins αIIbβ3 and αvβ3 in melanoma and breast cancer is in preclinical trials (32). Similarly, volociximab, an antibody that binds specifically to integrin α5β1, has been used to treat ovarian cancer, peritoneal cancer, renal cancer, and metastatic melanoma (32, 69, 70). Volociximab is currently in a phase II clinical trial in combination with carboplatin and paclitaxel as a first-line treatment for advanced NSCLC (clinicaltrials.gov identifier: NCT00654758). A previous phase I clinical trial concluded that volociximab combined with carboplatin and paclitaxel was generally well tolerated and potentially effective in patients with advanced NSCLC (71). Vitaxin, a humanized monoclonal antibody targeting integrin αvβ3 has shown promising therapeutic efficacy against breast, lung, and colon cancers by preventing angiogenesis in clinical trials (72).

Apart from antibody-based therapy, the use of specific peptide antagonists to target integrins is emerging as a promising approach in cancer therapeutics. Cilengitide is a small peptide that functions by targeting integrin αvβ3 and cyclic arginine-glycine-aspartate (RGD) binding. The efficacy of cilengitide has been evaluated in a series of phases I and II studies in patients with both recurrent and newly diagnosed glioblastoma. In one study, cilengitide was used to treat patients with head and neck cancers and found to be safe and associated with good recovery in these patients (73). In a phase III clinical study, cilengitide was used in combination with radiation therapy and temozolomide chemotherapy in patients with newly diagnosed glioblastoma (clinicaltrials.gov identifier: NCT00689221; NCT00813943). Although cilengitide is primarily used for glioblastoma therapy, additional studies using cilengitide in head and neck cancers and lung cancer have been performed. For example, in one phase II study, cilengitide was used in combination with cisplatin and either vinorelbine or gemcitabine as a first-line therapy for patients with advanced NSCLC (clinicaltrials.gov identifier: NCT00842712) (74). In another trial, cilengitide in combination with cetuximab/chemotherapy showed potential clinical activity, with a trend for progression-free survival in an independent-read analysis. Another randomized phase I/II study was conducted to evaluate the safety and efficacy of cilengitide in combination with cisplatin, 5-fluorouracil, and cetuximab in patients with recurrent/metastatic small cell head and neck cancer (clinicaltrials.gov identifier: NCT00705016) (74).

Other peptides that target integrins for cancer therapy are also being investigated. ATN-161, a peptide targeting integrin α5β1, is in phase I clinical trials for patients with glioblastoma (32, 75). HM-3, an 18 amino-acid peptide targeting integrin αvβ3, was tested preclinically in xenograft models of NSCLC and gastric, breast, and colon cancers and found to have efficacy with minimal cytotoxicity (76). HM-3 is now in phase I clinical trials (32, 76). AP25, a 25 amino-acid peptide targeting integrins αvβ3 and α5β1 has been used in preclinical trials for treatment of melanoma and gastric, hepatic, and breast cancers (32).

These studies highlight the importance of using and combining integrin-targeting agents with approved anticancer drugs for more effective clinical outcomes. Nonetheless, there are other promising therapeutic targets within ECM such as hyaluronic acid, which can be used for effective treatment (77, 78). Methodical and precise determination of the specific ECM components that promote growth and metastasis in different cancers will help to reveal more new therapies targeting ECM components and provide preclinical data that can be tested in clinical trials.

CONCLUSIONS

The role of ECM proteins in modulating the TME and its effect on cancer progression is well defined. However, the role of epigenetic modifications that regulate the expression of ECM proteins has not been well studied. More studies need to be conducted across all cancer types to provide a greater understanding of how epigenetic modifications modulate the expression of ECM proteins, thus affecting cancer signaling. Targeting of ECM proteins is emerging as a new approach to treat various cancers. Therefore, discovering new and more effective ways to target ECM proteins will improve future clinical care and quality of life for patients with cancer.

GRANTS

The author was supported by National Institutes of Health Grants R03CA230815, R03CA248913, and R01CA233481-01A1.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

This article is part of the special collection “Tumor Host Interactions in Metastasis.” Mythreye Karthikeyan, PhD, and Nadine Hempel, PhD, served as Guest Editors of this collection.

AUTHOR CONTRIBUTIONS

R.G. prepared figures; R.G. drafted manuscript; R.G. edited and revised manuscript; R.G. approved final version of manuscript.