Interplay Between Extracellular Matrix Remodeling and Angiogenesis in Tumor Ecosystem

The global burden of cancer is projected to be doubled over the next two decades (1). With 19.3 million new cases and 10.0 million deaths reported in 2020 as per GLOBOCON 2020 estimates, cancer remains the leading cause of mortality worldwide. The success of the treatment depends upon surgical resection at early stage or use of adjuvant therapies to treat clinically undetectable micrometastases. Nevertheless, cancer relapses and progresses to clinically detectable metastatic disease in 50% of the patients and such patients are treated with systemic therapies. Despite the treatment advancements, significant number of patients dies of tumor metastasis. Thus, it is clearly evident that early detection and prevention or development of effective therapies to treat metastasis are important for no relapse and complete cure. Nevertheless, traditional forms of cytotoxic chemotherapy fail to reduce mortality rates of the patients due to relapse with intractable metastatic disease. The mutant cells residing in a favorable tumor ecosystem undergo neoplastic cell expansion and growth. Thus, studying the complex mechanisms of metastases and examining the interactions of neoplastic cells within the tumor ecosystem are critical to explore the effective cancer prevention modalities (2). The dynamic tumor ecosystem also defined as microenvironment is constantly evolving and describes the entirety of the tumor components. It is composed of noncellular component termed as extracellular matrix (ECM), secreted factors and the cellular components including stromal cancer-associated fibroblasts (CAF), pericytes, endothelial cells (EC), adipocytes, and immune cells. ECM is a highly dynamic entity of vital importance and it determines and controls the most fundamental behaviors and characteristics of cells such as proliferation, adhesion, migration, polarity, differentiation, apoptosis, wound healing, inflammation, and cancer. ECM can be remodeled by synthesis, contraction, and/or proteolytic degradation. Cell–matrix interactions and major alterations in ECM structure and composition control or regulate the cell behavior. On the basis of the structural and biochemical characteristics, ECM is classified into interstitial matrix (surrounds cells) and basement membrane (BM; separates epithelia from its underlying mesenchyme and lines blood vessels). Over the last decade, many studies have strengthened our understanding on the role of ECM in malignancy as well as the effect of its modification and/or remodeling on disease progression and response to therapy. Invasion of epithelial tumor cells by breaking the BM, entry into vasculature and extravasation to distant sites are necessary to establish new tumors (3). Local invasion and systemic metastatic spread depend upon the extent of degradation of ECM proteins, attachment and detachment of tumor cells from ECM proteins and their movement through the defects in ECM. The ECM provides a three-dimensional intricate network of array of multidomain macromolecules which are organized in a cell/tissue-specific manner. ECM components interact with each other and form a structurally stable composite that contributes to the mechanical properties of tissues. The “core matrisome’’ is composed of approximately 300 unique matrix macromolecules and are classified into proteins (collagens, elastin), cell adhesion proteins (fibronectin, laminin), non-proteoglycan polysaccharide (hyaluronic acid), and proteoglycans (PG) [heparan sulphate (perlecans, glypicans, and syndecans), chondroitin sulphate (versican), keratan sulphate (lumican, keratocan, mimecan, fibromodulin, prolargin, osteoadherin, and aggrecan)] (4). Heparan sulphate PGs (HSPG) including syndecans, glypicans, and perlecans are the most represented HSPG in endothelial ECM. Matricellular proteins such as thrombospondins, cartilage oligomeric matrix protein, secreted protein acidic and rich in cysteine (SPARC), osteopontin and tenascins interact with growth factors, cell surface receptors and other matrix components and contribute to tissue-specific functions (5). The ECM is also a reservoir of growth factors and bioactive molecules. These include angiogenic growth factors (AGF), proangiogenic receptors, antiangiogenic factors, and angiogenesis effectors. Tumors leverage ECM remodeling by releasing matrix-sequestered growth factors to create a microenvironment that promotes EC proliferation and migration and supports tumor angiogenesis (6). Number of AGFs including VEGFs family (VEGF-A, -B, -C, -D, and -E and placental growth factor) interact and activate integrin αvβ3, neurophilin-1 (NRP-1) and HSPGs, thus induce complete angiogenic response in ECs (7, 8). Around 22 members of the FGF family are identified. They interact with FGFR1 (widely expressed on ECs) and activate proangiogenic program. Family of FGF growth factors interact with integrin αvβ3, ganglioside GM1, NRP-1, and HSPGs and induce a full angiogenic response. Beside VEGFs and FGFs, many other canonical and noncanonical AGFs induce neovascularization (8).

The pivotal multistep process of angiogenesis is initiated with the increased EC permeability and cellular proliferation in response to angiogenic stimuli to support the elongation of new capillary sprouts. Proteolytic degradation of matricellular components of BM is required for increased invasion of ECs into the stroma of neighboring tissues. The sprout forms a multicellular structure. Migrated ECs triggers lumen formation and as a result, new capillary channel is formed. Construction of BM, adherent junctions, and ECs finally stabilize the capillary.

Tumor-driven changes are further intervened by the cooperative activity of the plasminogen activator (PA) system and various matrix metalloproteinases (MMP) and their corresponding family of specific inhibitors, also called matrixins (9, 10). They function in the extracellular environment of cells, degrade both matrix and non-matrix proteins and play central roles in morphogenesis, wound healing, tissue repair, cancer and remodeling in response to injury. Therefore, these molecules are potential targets which can be used to manipulate angiogenesis in cancer treatment strategies. Understanding the underlying mechanisms of molecular regulators and complex cascades in angiogenesis and ECM remodeling during tumor growth open new possibilities for diagnosis and explore the therapeutic potential of inhibitors of tumor development. The current review summarizes the recent discoveries on the intricate relationship between ECM remodeling and angiogenesis, their role in dictating cell behavior during cancer progression and their clinical implications during tumorigenesis.

Being involved in tissue architecture and homeostasis, ECM is a constantly evolving, highly complexed but organized structure (Fig. 1). It is composed of diverse collection of soluble and nonsoluble proteinaceous, non-proteinaceous, sugar, and secretory molecules within all tissues and organs. Collagen is the principal proteinaceous component of ECM. It is produced and secreted by many stromal cells including fibroblasts and provides necessary scaffold for the organization of cells. Higher expression of collagen-processing enzymes like lysyl hydroxylases (LH) and prolyl hydroxylases (P4H) leading to increased synthesis of collagen is observed (11). Higher transcriptional and translational activities of lysyl oxidases [LOX (involved in cross-linking of collagen)] and elastin render rigidity to the tumors, increased mechanical stress, reduced overall oxygen supply, and promote focal adhesion kinase (FAK)-mediated signaling in the tumor microenvironment (12–15). Increased expression of P4Hs, an ascorbic acid–dependent enzyme, leads to the formation of hydroxyproline and thereby stabilizes the helical collagen. The study observes the strong correlation of intratumoral deposition, chemotherapeutic resistance and reduced survival of patients with triple-negative breast cancer (16–18). Collagen thickness and its alignment (induced by collagen cross-linking) significantly reduce the overall survival of patients with pancreatic ductal adenocarcinoma (PDAC; refs. 17, 18). Another important component of ECM is PGs. It is heavily glycosylated with the chains of glycosaminoglycans (GAG) by glycosyltransferases followed by chain modification by sulfotransferases and epimerases. Varied composition of GAGs is observed in chondroitin sulfate, heparan sulfate, keratan sulfate, and dermatan sulfate (19). GAG sulfation patterns are examined to serve as specific recognition motifs for number of cytokines, chemokines, growth factors, and pathogens. Alterations in GAG composition of PGs are noted in cancer cells. Studies report higher expression of chondroitin sulfate 4-O sulfotransferases in ovarian and breast cancers. Elevated 6-O-sulfated chondroitin sulfate is identified in cancerous lung tissues compared with nonmalignant tissues (20). Glypican-3 is studied to influence many central signaling pathways (including Wnt signaling) critical for cell proliferation in hepatocellular carcinoma. Chondroitin sulfate proteoglycan 4 (CSPG4) positively regulates cancer cell proliferation and mediates activation of integrins which induce chemoresistance and survival of tumor cells. Suppressed proliferation and anchorage-independent growth of breast cancer cell lines upon de novo expression of decorin highlight its potential function as tumor suppressor (20). Hyaluronic acid (HA), another component of ECM is a GAG. It is not conjugated to proteins and lacks protein core. It is synthesized and exported outside the cell via direct extrusion through an enzymatic pore by the plasma-membrane bound family of three transmembrane glycosyltransferases namely, hyaluronan synthetase 1–3 (HAS1–3; ref. 21). Activation of number of signaling pathways involved in cell proliferation, differentiation, motility, and adhesion is known to be triggered by HA. Elevated synthesis of HA is noted in colorectal cancer, pancreatic cancer, breast cancer, prostate cancer, and brain tumors (19). Its role in epithelial-to-mesenchymal transition (EMT), increased invasion, and metastasis is examined. Higher levels of both HA and HAS1–3 are correlated with poor prognosis (22).

Laminin is the complex adhesion glycoprotein and is the main component of BM. It is ubiquitously distributed in the stromal cells of tumors. Its increased expression and aberrant distribution are correlated with poor prognosis and invasiveness (19). Another important component of ECM is a high-molecular-weight glycoprotein called fibronectin (FN) which is made of two subunits that are covalently linked by a pair of disulfide bonds at the C-termini. Single gene undergoes alternative splicing to produce 20 isoforms of human FN. The insoluble cellular FN is a part of ECM while it exists as a soluble dimer in the plasma (pFN). It interacts with many other ECM components including integrins, collagen, tenascin-C, fibrillin, glycosaminoglycans as well as with growth factors (5). Secretion and assembly of FN as parallel fibers by CAFs help to mediate directional cancer cell migration. Role of FN in cancer cell growth, survival, and invasion is extensively studied. Potential role of FN in inhibition of mitochondrial dysfunction and caspase activity, increased intracellular reactive oxygen species production and NADPH oxidase activation, and transactivation of IGF-1 receptor contributes to prosurvival effects of pancreatic cancer cells. Further functions of FN include cell proliferation via integrin α5β1 through activation of AKT/mTOR/p70S6K pathway and inhibition of AMPK and LKB1 in lung cancer; activation of AKT/mTOR/4E-BP1 signaling cascade in gall bladder cancer cell lines; activated Src and TGFβ1 signaling in renal cancer cells; and activation of PI3K/AKT/SOX2 and CDC42/F-actin/YAP-1/Nupr1/Nestin signaling pathways via integrin αvβ3 in glioma cells (5). Elevated FN expression correlates with tumor grade and/or aggressiveness in many solid tumors.

ECM modeling or stiffening or rigidity is best described as the extent to which the matrix resists itself in response to the applied mechanical forces (Fig. 2). Endogenous forces are generated by the cytoskeletal contractility within cells while the exogenous forces including shear stress, gravity, and tensile and compressive forces are applied to the cells by the surrounding microenvironment (23, 24). These forces are responsible for the remodeling or stiffening of the ECM. ECM stiffening is essential for normal tissue homeostasis where physiologic needs in different tissues and/or organs decide the degree of stiffness. However, excessive deposition of matrix and prolonged changes in ECM mechanics lead to the disease condition. The exact composition and spatial organization of ECM components dictate the local tumor microenvironment (25). Biochemical composition, integrity, physical and mechanical properties of ECM impact all the cancer hallmarks including cellular processes that lead to cancer initiation, progression, invasion, and the formation of metastatic niche. These changes are greatly influenced by CAFs. CAFs are derived from variety of precursor cells including tissue-resident fibroblasts, epithelial cells, ECs, bone marrow–derived mesenchymal stem cells, pericytes, and adipocytes. Growth factors including, platelet-derived growth factors (PDGF), EGFs, FGF, TGF superfamily (TGFβ), sonic hedgehog (SHH), secretion of chemokines, cytokines, oxidative stress and hypoxia activate quiescent fibroblasts into stellate shaped CAFs (26–29). CAFs are identified with the presence of increased levels of nonspecific markers namely, FSP-1 (fibroblast specific protein-1, also known as S100A4), FAP (fibroblast activation protein), α-SMA (α-smooth muscle actin), vimentin, PDGFR-α (platelet-derived growth factor receptor-α), PDGFR-β, and PDPN (podoplanin; ref. 30). Deposition of FN, generation of intracellular tension, translocation of Yes-associated protein (YAP) to the nucleus and activation of Rho-ROCK-Myosin II (Ras homolog-Rho-associated protein kinase-myosin II) signaling pathway and the incorporation of α-SMA into actin-myosin fibers lead to increased contractility of CAFs. Tensile forces generated by CAFs trigger ECM remodeling at different levels (30). Increase in ECM stiffness regulates the mechanics of the tumor ecosystem and is one of the hallmarks of tumor development. A total of 10-fold higher stiffening is observed in breast cancer tissues compared with normal breast tissues and it continues to increase in malignant tissues. Elastographic studies observe a halo of stiffer tissue at the margins of invasive tumors (31). Increase in stiffness due to deposition of ECM, described as desmoplasia, is being noted in tumor microenvironment.

CAFs synthesize, produce, and deposit substantial amounts of ECM components. CAFs also contract the tissue and thus alter the ECM of the tumor stroma quantitatively and qualitatively. The subsequent section discusses the role of CAFs in creating various alterations in ECM components and thus ECM stiffening in tumorigenesis. Constitutive activation of collagen concentration, its increased amounts, density, and enzymatic cross-linking of collagen are the most common alterations of ECM. These changes promote PI3K activity, invasion of an oncogene-initiated epithelium, and ECM stiffening during tumor progression. Altered ECM composition and decrease in elasticity due to higher amounts of FN fibrils and collagen type I II, III, V, and IX are observed to be associated with tumorigenesis. Observed increased type V collagen deposition is noted to modify the mechanical and structural properties of ECM (32). Increased infiltration of fibroblasts and/or myofibroblasts and significant accumulation of collagen in the ECM of many solid tumors are shown to be linked with poor prognosis and resistance to systemic therapy (33). Excessive deposition of major ECM components including FN, HA, and tenascin C into the interstitial matrix results in a fibrotic phenotype, also termed as desmoplasia, a key characteristic of various cancers including breast cancer and pancreatic ductal adenocarcinoma (34, 35). Around 15% of the collagenous matrix is made of collagen type V in the desmoplastic stroma of breast carcinomas. This type of collagen isoform is found in lower amounts in normal or fibrocystic breast tissue (36). Reduction in length and organization of collagen fibers in high collagen V/collagen I ratio is examined to be responsible for gel like ECM (37). Reduced expression of collagen IV in ovarian carcinomas compared with benign tissue inversely correlates with stage and markers of malignancy (38). Human lung tumors are identified with high collagen I/collagen III ratio and reduced collagen IV expression (39). Increased collagen I levels show strong association with invasiveness, angiogenesis, and reduced survival of melanoma patients (40). Higher expression and activity of LOX in response to increased collagen amounts enhance the cross-linking of collagen fibrils and thus promote ECM stiffening and invasiveness in many cancer types (41). Expression of LOX is shown to be associated with metastasis and reduced survival of patients. Overexpression of HSPGs is also known to contribute to the dense fibrous tissue surrounding in malignant tumors (42). A prometastatic and proangiogenic function of heparanase and its overexpression is observed in many cancers including breast cancer, gastrointestinal tumors, and esophageal carcinomas. Heparanase is an endo-β-D-glucuronidase and can catalyze the cleavage of β (1, 4)-glycosidic bond between glucuronic acid and glucosamine residue. It thus releases HS which further regulates the release of many HS-linked molecules including cytokines, growth factors, enzymes involved in inflammation, wound healing, and tumor invasion. Overexpression of heparanase mRNA correlates with reduced survival of patients with cancer (43). Triple-negative breast cancer [estrogen receptor negative, progesterone receptor negative, HER2 negative (ER⁻/PR⁻/HER2⁻)] exhibits increased deposition of collagen and enhanced invasion with CAFs (44, 45). Strong correlation between elevated ECM expression and poor prognosis is observed in luminal breast cancer (46). Another study by Mintz reports the prognostic significance of high ECM turnover for chemoresistance in pediatric osteosarcoma (47). Higher levels of collagen (Col1a1), fibronectin-1 (Fn1), and mucin 5a (Muc5a) are reported as predictive indicators of overall and disease-free survival in gastric cancer (48). Abundant, denser, and stiffer ECM in tumor tissues reduces oxygen supply, increases metabolic stress, drives activation of EMT, antiapoptotic and drug-resistant pathways and thereby contributes to metastasis (49). PGs along with collagen type II are the main components of cartilage and contribute to tissue stiffness and sturdiness. Normal and tumor tissues differ with respect to high-molecular-weight and low-molecular-weight PGs. Increased expression of aggrecan and CSPG4 but reduced decorin levels are observed in prostate cancer (50). Another study identifies higher mRNA levels of decorin, lumican, and versican in prostate cancer (51). Increase in versican and decorin levels and doubling of GAG content are noted in gastric cancer (52). Stroma of breast carcinoma is examined to be enriched with versican and decorin along with collagen type I and FN (53). Upregulated versican and decorin and downregulated aggrecan are noted in squamous cell laryngeal carcinoma (54). Shift in sulfation from 4 to 6 is examined in colorectal cancer and gastric cancer while a reverse shift pattern is observed in laryngeal carcinoma (19). Altered structural and mechanical properties resulting ECM stiffness lead to changes in cellular behavior in tumor ecosystem. Enhanced mechanotransduction and cell migration due to integrin clustering, focal adhesion activation and increased cell adhesion to ECM are associated with ECM stiffness and cancer cell invasiveness (55). ECM remodeling shifts the cells phenotype from differentiated to proliferative type. Human breast cancer cells in a stiff matrix are examined to be associated with the overexpression of proliferation-related genes including cyclin D1 (56). Growth signals in tumor ecosystem dictate the ECM in the surrounding tumor tissue to undergo changes in the amounts of deposition, composition, organization, and posttranslational modification of ECM molecules compared with normal tissues. Posttranslational modifications in ECM components lead to the morphologic changes in the ECM architecture thereby contribute to tumor progression and motility (57, 58). Glycosylation of surface proteins frequently change and foster pathologic cell behaviors in cancer (59, 60). Glycosylation of lysine and hydroxylysine residues by addition of galactose and glucose; and hydroxylation of lysine residues by LHs [encoded by procollagen-lysine 2-oxoglutarate 5- dioxygenase (PLOD) genes] modify procollagens (61). These modified procollagens form triple helices which are further processed extracellularly by proteases to form collagen fibrils. Cross linking of collagen fibrils by extracellular LOX and LOX-like isoenzymes is essential for correct collagen fiber assembly, increased tensile strength and ECM stiffness (62, 63). Tissue transglutaminase 2 (TG2) cross-links with ECM molecules like FN, HSPG, fibrinogen, and collagen IV via transamidation of glutamine residues to the amino group of a lysine residue of another protein chain. This transamidation process results in the formation of covalent N-γ-glutaminyl-ɛ-lysyl-isopeptide bonds which are resistant to proteolytic degradation (64). Overexpression of PLODs, LOXs, and TG2 is frequently identified in cancer and are responsible for increased cross-linking and linearization of ECM molecules and protumorigenic effects (64). There are several glycosylated ECM molecules including HSPGs that bind with growth factors and modify receptor tyrosine kinase (RTK) signaling. Tumor cells are identified with increased levels of bulky glycoproteins on the cell surface which can exert tension to ECM-bound integrins and induce integrin clustering. This contributes to the increased protumorigenic integrin signaling which is identified to be regulated by extracellular modification enzymes such as endosulphatases (Sulfs; ref. 59). Endosulphatases are examined to be dysregulated by modifying the sulphation pattern of HS. PGs are observed to affect the binding of growth factors in various cancers. Higher levels of periostin and galectin-1 glycoproteins are correlated with poor survival of patients with PDAC. Another ECM calcium-binding matricellular glycoprotein known as SPARC has an opposite effect on tumor growth. It inhibits proliferation and induces apoptosis, but surprisingly its expression in PDAC fibroblasts is associated with poor survival of patients. SPARC with its complex role in PDAC progression is studied as an independent prognostic marker (30). Changes in the structural composition, organization, and degradation of ECM in a tumor ecosystem regulate pathologic angiogenesis. ECM breakdown facilitating EC invasion and tube formation is identified as an essential activator of angiogenic process. Stromal CAFs are localized at the leading edge of tumor and therefore, demand expanded vessel supply. These cells release multiple angiogenic cues in the ECM and support aggressive tumor growth (65).

Angiogenesis is the process of sprouting and the formation of new blood vessels from preexisting vasculature. During tumor progression, an “angiogenic switch” is activated most of the times. This switch turns normal quiescent vasculature to continuously sprout new vessels and helps expanding the neoplastic growths (66). Three types of angiogenic processes are known which include (i) sprouting angiogenesis, (ii) splitting angiogenesis or intussusceptions, and (iii) looping angiogenesis (67, 68). Sprouting angiogenesis is the most widely studied and characterized form of angiogenesis. Tissue hypoxia plays important role in sprouting angiogenesis by triggering the budding of a new capillary sprout laterally from a preexisting vessel. Biological signals and/or proangiogenic factors secreted in the areas devoid of vasculature activate ECs present in preexisting blood vessels. Activated ECs/tip cells release proteases, degrade BM, allow ECs to escape from the original (parent) vessel walls, allow the proliferation of ECs in surrounding matrix and form solid sprouts connecting neighboring vessels, thus grow in length simultaneously. Intussusceptions and/or splitting angiogenesis is mediated by the extension of capillary walls of an existing vessel into the lumen and finally bifurcating it into two vessels. Looping angiogenesis is mediated by the involvement of biomechanical forces for the expansion of translocated vessel loops. The vessel loops are mechanically dragged into the tissue. Tissue hypoxia and/or ischemia initiates angiogenesis and this process is further controlled by the activity of various factors.

Other than hypoxia and/or ischemia, inflammation also stimulates angiogenesis and triggers the release of cytokines and the basic proangiogenic factors namely, VEGF or FGF (69, 70). Platelets regulate and control the localized angiogenesis by gathering numerous angiogenesis-related proteins and storing them in various compartments called alpha granules. During an injury, platelets are actively recruited to the wound site where they provide plethora of angiogenic factors. Thus, they facilitate the formation of new blood vessels through sequential release of growth factors by degranulation (71, 72). The initial signal is the flaccidity of the existing vessels caused by nitric oxide (NO). NO provokes the launch of proteolytic enzymes like MMPs that activate the production of the new vascular structures by the degradation of the BM. It promotes the invasion of ECs into the stroma of the neighboring tissue. The integrins are responsible for facilitating and regulating the adhesion and migration of ECs. The angiogenic stimulation causes increased permeability and cellular proliferation of ECs. Surrounding pericytes dissociate from the vessel wall and the ECs loosen their junctions. Subsequently, ECs are being transformed into motile tip cells which are located at the growing ends of sprouting vessels with long filopodia. ECs proliferation is regulated by (i) pericytes which sequester mitogens in the ECM, (ii) changes in EC shapes that reduce the sensitivity of the cells to growth factors, and (iii) endothelial integrins. While sensing the proangiogenic directional cues in the environment, tip cells migrate in a specific direction (73, 74). At the same time, specific tip cells release inhibitory signals into the surroundings for preventing the uncontrolled migration of ECs. Migrated ECs trigger lumen formation as the sprout forms a multicellular structure. Newly formed ECs determine the genesis of the new tubular structures that mature and stabilize through the construction of BM, adherent junctions, and ECs (75).

An imbalance between proangiogenic and antiangiogenic factors in response to environmental, chemical, and mechanical stimuli results in tumor angiogenesis. Tumor angiogenesis is best defined as the multistep process of blood vessel creation, penetration, and growth in tumor ecosystem. It begins in ECs and directs them to switch from quiescent state to angiogenic state in response to angiogenic factors. Eventually, enzymatic degradation of BM, increased vascular permeability, and extravasation of blood proteins into interstitial collagen matrix contribute to the formation of provisional ECM. As a result, ECs multiply, invade the ECM and participate in the formation of immature capillary structure and deposition of new complex BM. The BM of vascular ECM is thick and composed of type IV collagen and laminin and provides broader binding surface area for ECM proteins, growth factors and integrin receptors. BM acts as a reservoir of number of angiogenic growth factors, cytokines, and proteolytic enzymes. Increased activity of these molecules and their interaction with ECM proteins result in activation of multiple signaling pathways and enhanced proangiogenic and promigratory properties (Fig. 3). It provides nutrients, oxygen, helps in waste disposal, facilitates dissemination of tumor cells, and thus promotes metastasis (76). Table 1 describes the functional/mechanistic role of proangiogenic factors during tumor development (7, 77, 78–140).

Many of the ECM components act as ligands for various cell surface receptors such as syndecans, integrin, and RTKs and thus, modifications/alterations of ECM molecules have direct implications on the complex network of signaling pathways. Vascular abnormalities in mice model as well as in vitro systems are observed to be associated with the activation of multiple signaling pathways in response to angiogenic cues. It facilitates the interaction of ECs with ECM molecules (collagen, FN, laminin; refs. 76, 141). Interaction of collagen I with integrins (α1β1, α2β1, αvβ3, and αvβ5) on cell surfaces in human dermal microvascular ECs (isolated from neonatal foreskins and anchored to collagen I) and mouse model of skin angiogenesis (produced with subdermal injection of Matrigel and immortalized human cells stably transfected with VEGF) results in activation of MAPK pathway, proliferation of ECs, and suppression of apoptosis (142). Suppression of cyclic adenosine monophosphate–dependent PKA, reorganization of actin fibers and changes in cell shape upon binding of integrins (α1β1 and α2β1) to collagen I are demonstrated in above-mentioned in vitro model (142). Activation of Src (tyrosine-protein kinase) and Rho, suppression of Rac (Ras-related C3 botulinum toxin substrate 1) activity and disruption of intercellular junctions following interaction of ECs to collagen I are observed in mouse skin model (with VEGF-expressing cells and packaging cells producing retroviruses encoding Rho A GTPase mutants; ref. 143). Another important protein molecule of BM is collagen IV which participates in the proliferation of EC and cell behavior. In vitro study identifies the angiogenesis dependence on secretion and extracellular deposition of collagen type IV (144). Proteolytic cleavage of collagen type IV results in the exposure of a functionally important cryptic site which is normally hidden within its triple helical structure of collagen type IV. Mouse embryonic cell cultures lacking endogenous FN demonstrate the effect of arginine—glycine–aspartic acid (RGD) motif (present in FN) binding to the integrins α5β1 and αvβ3 in ECs on polymerization of FN, regulation of cytoskeletal organization, stabilization of cell-matrix adhesion, EC survival by suppressing activity of PKA and cell growth (145–149). Proteolytic remodeling of ECM promotes novel integrin–ligand interactions (loss of α1β1 integrin binding and gain of αvβ3) and is required for angiogenesis in vivo (150). Laminin is examined to be responsible for activation of proteinases, degradation of matrix, cessation of proliferation of ECs, recruitment of pericytes, stabilization of vessels through the activation of Notch pathway (151, 152).

ECM components and/or matrix molecules undergo degradation or partial modification due to the enzymatic activity of proteinases produced by connective tissue/ECs. As a result, small peptides also called as matrikines are produced and cryptic sites are exposed. Cryptic sites also called as matricryptic sites are biologically active sites which are absent in mature and secreted form of ECM molecules but are produced as a result of structural/conformational changes. Many of the cryptic sites are identified as important physiologic angiogenesis inhibitors. The released matrix fragments bind to specific cell receptors, activate intracellular signaling pathways and regulate the cellular activity. PA/plasmin system and MMPs possess proteinase activity and participate in the degradation of ECM components (76). Urokinase-type PA or tissue-type PA converts plasminogen (inactive plasma zymogen) to the active serine protease plasmin and is involved in fibrin degradation, matrix turnover, and cell invasion. PA/plasmin system is directly or indirectly involved in cleaving ECM molecules (von Willebrand factor, FN, laminin, thrombospondin), activating MMPs and releasing and/or sequestering growth factors and cytokines in the ECM (153, 154). MMPs (zinc endopeptidases) exist in soluble and membrane-bound forms. They are produced by many cells including epithelial, inflammatory, fibroblasts, and ECs and are shown to be primarily involved as regulators of modifying ECM composition, new blood vessel growth, cell migration, and adhesion to promote tumor growth and metastasis. An imbalanced expression and activity of MMPs and tissue inhibitors of metalloproteinases (TIMP) are observed during cancer development (155). Downregulated MMPs and upregulated TIMPs lead to incomplete matrix remodeling and irreversible fibrosis and tumor progression. Nevertheless, other studies report the role of upregulated MMP in progressive destruction of normal ECM and the formation of stiffer ECM in tumor cells. Generation of bioactive cleaved peptides, release of growth factors and chemokines in ECM due to overactive MMPs further promote cell migration, metastasis and make tumor cells resistant to apoptosis (156). Secretion of proteases from CAFs contributes to production, deposition, and degradation of ECM and thus plays dual role in ECM remodeling (157). Matrix stiffening with a concomitant rise in cytokine localization and MMP production independent of matrix density is responsible for increased vascular growth (158). Number of MMPs including MMP-1, -2, -3, -7, -9, and -14 mediate ECM degradation and their higher expression is observed in ECs in pathologic settings (88). MMPs also facilitate the release of proangiogenic factors like VEGF, TGFβ, and FGF in the stroma, create a favorable metastatic niche and promote tumor growth (159). One of the most important molecules secreted by stromal CAFs is VEGF-A in both spontaneously arising and implanted tumors of genetically engineered mice. The expression of integrins (α1β1 and α2β1) in microvascular ECs is examined to be induced by VEGF. VEGF can bind with the cell surface or ECM components and influence EC migration and proliferation (160, 161). MMP-7, for example, degrades human soluble VEGFR-1 (an endogenous VEGFR) and increases VEGF bioavailability around the ECs (162). MMP-7 is also known to release VEGF stored by fibroblasts in its latent form in the ECM (163). FGF-2 binds to HSPGs present on the cell surface and within ECM. Binding of FGF-2 to its specific tyrosine kinase receptors is modulated by perlecan which protects it from degradation by extracellular proteinases. In vitro and in vivo studies identify its role in EC survival, proliferation, migration, and differentiation (164). Three-dimensional cultures exhibit formation of calcium- and magnesium-dependent tube-like structures which mimic angiogenesis upon TGFβ1 treatment (165). MMP7 cleaves Fas ligand, removes it from cell surface and prevents it from stimulating the Fas death receptor and thus inhibits innate apoptotic pathways (166). MMP14 is also reported to exhibit antiapoptotic interactions with the surrounding microenvironment (159). Studies highlight the denser accumulation of fibrillar collagen in the stroma leading to specific covalent intermolecular connections between collagen fibers. It results in masking of active sites for MMP activity, and in turn leads to accumulation of MMP-resistant collagen fibers (159, 167). This creates an imbalance between tissue degradation by proteases such as MMPs and their inhibitors, TIMPs, thus contributes to alterations of ECM topology, and promotes ECM stiffening, metastasis, and infiltration of tumor-supporting immune cells.

Conventional anticancer therapies not only result in drug resistance but also impaired drug delivery due to genomic instability and the presence of CAFs in tumor mass. CAFs create proinflammatory, immunosuppressive, and oxygen-rich microenvironment and facilitate tumor development. Marker proteins expressed by CAFs including (α-SMA, FAK, etc.) or secreted (TGFβ, ECM proteins, etc.) are identified as potential drug targets; however, these targets are also expressed by other cells within the tumor. Therapies based on direct genetic deletion or indirect pharmacologic depletion of CAFs by targeting FAP⁺ CAFs, α-SMA⁺ CAFs, and SHH pathway lead to dramatic modification of tumor stroma and increased intratumoral vascularization thus contribute to tumor dissemination and chemoresistance (168). Many recent studies advocate the cancer therapies based on targeting angiogenesis which may potentially halt the growth, invasion, and spread of cancer.

Compressed blood supply in the regions of tumor growth results in neovascularization. Nevertheless, neovascularization leads to poor tumor's accessibility to conventional chemotherapeutic agents. Antiangiogenic molecules and/or drugs in animal models have shown to directly reduce the growth of blood vessels as well as increase the delivery of chemotherapeutic agents to a tumor and thus are considered as excellent potential therapeutic targets of angiogenesis in many types of cancers. Over the past few years, around 80 drugs are being investigated in preclinical studies and phase I–III clinical trials where the therapies are mainly centered on the inhibition of RTKs or blockade of VEGFRs or ligands by neutralizing antibodies. Bevacizumab (Avastin; a mAb) is an FDA-approved drug which binds to VEGF and prevents it from activating VEGFR. It is the first approved angiogenesis inhibitor drug used to slow down the tumor growth and extend the survival time of some patients with cancer. Bevacizumab in combination with other drugs are successfully shown to treat cancers including metastatic renal cell cancer, metastatic colorectal cancer, and non–small cell lung cancers (169). Despite promising preclinical results, these monotherapies could not avert the development of resistance. Other FDA-approved antiangiogenic drugs targeting the multiple growth factor receptors are sorafenib (Nexavar), sunitinib (Sutent), pazopanib (Votrient), and imatinib (Glivec). These drugs are used in the treatment of hepatocellular carcinoma and kidney cancer (sorafenib), kidney cancer and neuroendocrine tumors (sunitinib), kidney cancer and neuroendocrine tumors (pazopanib), and renal cell carcinoma and chronic myelogenous leukemia (imatinib; ref. 169). Antiangiogenic therapies may require a long time to treat cancer as these drugs do not kill tumor cells but slow down the growth.

Matrikines and matricryptic sites produced as a result of PA/plasmin system and MMPs activity are identified as the physiologic inhibitors of angiogenesis. One of such inhibitors is endostatin which is isolated from nonmetastatic murine hemangioendothelioma cell line and is identified as the 20-kDa C-terminal proteolytic fragment of type XVIII collagen. It blocks G₁–S phase of cell cycle, promotes apoptosis, inhibits proliferation and migration of ECs. It exhibits antiangiogenic effects upon interacting with EC surface via HSPGs (170). It is shown to inhibit VEGF-induced migration of ECs in a dose-dependent manner (171). Proteolytic activity of MMP-9 results in the generation of tumstatin, a 244 amino acid long degradative product of α3 of type IV collagen. With antiangiogenic activity restricted to amino acids 54–132, it binds to ECs via αvβ3 integrin, regulates proliferation and promotes apoptosis. Interaction of canstatin [derived from NC1 (non-collagenous 1) domain of the α2 chain of type IV collagen] with αvβ3 and αvβ5 integrins on the surface of ECs results in inhibition of EC proliferation by inhibiting Akt and FAK phosphorylation and induction of the expression of Fas ligand and caspase-dependent apoptosis (172–174). Binding of arresten to α1β1 integrins and HSPGs leads to inhibition of MAPK signaling, proliferation of mouse retinal ECs [cultured on type IV collagen and stimulated with FGF-2] and neovascularization (175). Degradation of type IV collagen, from α4, α5, and α6 chain produces antiangiogenic molecules namely, tetrastatin, pentastatin, and hexastatin, respectively (176, 177).

Endorepellin, a novel antiangiogenic product is a C-terminal angiostatic fragment of HSPG perlecan. It interacts with α2β1 integrin receptor and triggers the disruption of the endothelial actin cytoskeleton (178). It also interacts with VEGFR through its laminin G-like domains and inhibits angiogenesis (179). Given the strategic role of antiangiogenic proteins/angiostatic factors in ECM remodeling and regulation of angiogenesis, these proteins could be investigated in preclinical and clinical studies with the aim to overcome resistance and improve the clinical outcomes of patients with cancer. Different preclinical models have shown the association of angiogenesis blockade with tumor invasiveness. Treatment with inhibitors of angiogenesis brings about vascular changes that include decreased expression of adherens junction proteins, reduced BM and pericyte coverage, and increased leakiness. These changes facilitate local intravasation and extravasation of tumor cells, ECM remodeling and metastatic colonization (180, 181). Many recent studies are focusing on the efficacy of direct or indirect targeting of ECM components in sensitizing the tumor cells to chemotherapeutic drugs.

Abnormally high interstitial fluid pressure (IFP) in the ECM of tumor mass acts as a barrier for effective drug delivery. Higher amounts of laminins, collagens, FN, integrins, HA, and periostin ECM proteins lead to increased cancer cell motility and reduced response to chemotherapeutic agents. Laminin and FN induced activation of PI3K-Akt pathways and chemoresistance are reported in ovarian, lung, and breast cancers. Higher levels of collagen VI result in cisplatin resistance in vivo in ovarian tumors. HA along with CD44 and integrin αV induce resistance to alkylating agents and promote glioblastoma (30). Therapeutic application of PH20 hyaluronidase (HA degrading enzyme) in osteosarcoma and melanoma mouse models is based on promoting degradation of ECM network by improving the efficacy of coadministrated chemotherapies (docetaxel and liposomal doxorubicin) and facilitating their accumulation into primary tumor. Efficacy of the combination of PEGylated human hyaluronidase (PEGPH20) with gemcitabine and nab-paclitaxel compared with gemcitabine and nab-paclitaxel chemotherapy alone was evaluated on the patients with high tumor HA content during phase III study (182). However, the results were discouraging as no difference was observed in terms of overall survival of patients. Collagenases are currently evaluated for their therapeutic effects. It results in degradation of ECM, reduced IFP, and increased drug uptake (mAbs) by the tumor mass in human osteosarcoma xenografts. Intratumoral injection of collagenase or cathepsin C (enzyme degrading decorin) is examined to increase the diffusivity of fluorescein isothiocyanate-dextran in melanoma model (183, 184). Owing to the important role of endogenous collagenases (such as MMP-1, - 8, and -13) in tumor invasion, degradation of ECM via these strategies may have adverse effects on cancer treatment. Enhanced cancer metastasis advocates the indirect targeting of ECM as the potential anticancer strategy.

Owing to significant role of MMPs in tumor invasion, angiogenesis, and drug resistance, many preclinical and clinical studies are evaluating the therapeutic efficacy of agents that target MMPs. Clinical studies on Tanomastat, Prinomastat, Rebimastat (MMP9 inhibitors) in colorectal cancer were stopped in phase III due to severe side effects and absence of any improvement in patient survival. Batimastat-treated rats exhibit inhibition of MMPs (MMP-1, -2, -3), reduced peritoneal carcinomatosis and hepatic metastases number and significantly prolonged survival. For safe and specific targeting of MMPs, humanized mAb, an ecaliximab against MMP9 is under evaluation for its antitumoral effect in phase II along with immunotherapy and in phase III along with chemotherapy in patients with gastroesophageal junction adenocarcinoma. Preclinical cancer models are being examined with mAbs namely, DX-2400, DX-2802, and DX-2712 against MMP 14, MMP9, and MMP12, respectively (185, 186). Inhibitor of heparinase, known as SST0001 is shown to reduce the levels of HGF, VEGF, and MMP9 in multiple myeloma. Heparan sulfate mimetic known as necuparanib (rationally engineered low-molecular-weight heparin) showed promising results during phase Ib evaluation, nevertheless, it did not exhibit any improvement in overall survival of patients with metastatic PDAC when it is combined with gemcitabine/nab-paclitaxel (187, 188). Another heparanase inhibitor, Muparfostat (Pi-88) is under clinical evaluation. It is shown to block angiogenesis, metastasis, and tumor growth in preclinical animal models and became the first heparanase inhibitor to enter clinical trials. It progressed to phase III clinicals but ultimately was not approved for use (189). Inhibition of LOX is shown to reduce collagen cross-linking, FN assembly, increased drug penetration, increased sensitization of breast cancer cells to doxorubicin treatment and enhanced sensitivity of prostate cancer cells to radiotherapy. Combinations of LOX blocking antibody and gemcitabine exhibits antitumor effects in early-stage tumors rather than locally advanced cancers (with well cross-linked matrix; ref. 190). Inhibition of ECM receptors and αVβ3/αVβ5 integrins with the drug, cilengitide inhibits cancer cell proliferation in head and neck squamous cell carcinoma (191). Application of inhibitor of αVβ1 integrin along with endostatin is shown to inhibit hemangioendothelioma by reducing the levels of NFκβ-induced CXCL1 expression (192). Another drug, E7820 is shown to inhibit the collagen-α2β1 integrin binding, inactivate the ECM-induced PI3K/AKT Snail signaling pathway, and potentiate the chemotherapeutic functions of oxaliplatin and 5-fluorouracil (193). ROCK regulates maintenance of cell shape and actomyosin contractility and thus controls motility, adhesion, differentiation, and cell proliferation. Fasudil, a ROCK inhibitor is shown to reduce cell contractility and ECM stiffening, thus improves response to gemcitabine/Abraxane at both primary and secondary sites and reduces pancreatic cancer fibrosis (194). Its effects on reduced cell migration in breast cancer cells are also observed (195).

The significant participation of CAFs in tumorigenic development, immunosuppression, and therapy resistance is well understood. CAFs when interact with TME exercise phenotypic and functional heterogeneity in a context-dependent manner in a given host. Hence, identification of CAF-specific markers and targeting them may optimize the sensitivity of anticancer therapies.

Compressed blood supply in the regions of tumor growth results in neovascularization. Nevertheless, neovascularization leads to poor tumor accessibility to conventional chemotherapeutic agents. Targeting angiogenesis may potentially halt the growth, invasion and spread of cancer. Nevertheless, treatment with inhibitors of angiogenesis brings about vascular changes that include decreased expression of adherens junction proteins, reduced BM and pericyte coverage, and increased leakiness. Therefore, these changes facilitate local intravasation, extravasation of tumor cells, tumor dissemination, ECM remodeling, and chemoresistance.

Increased solid stress and interstitial fluid pressure due to denser and stiffer ECM in tumors create hypoxia and metabolic-stressed milieu. This results in the increased expression of antiapoptotic proteins and signaling pathways and ultimately leads to cancer stem cell phenotype, increased tumorigenesis and therapeutic resistance. Cell adhesion to ECM proteins through integrins leads to the induction of many cell survival pathways including PI3K/Akt, MAPK, p53, ERK, and Rho/ROCK and thus contributes to chemoresistance. Owing to significant role of ECM proteins in inducing chemoresistance, direct or indirect targeting of ECM components is being reported as major axis of anticancer treatment.

In vitro and in vivo research models should be examined at different levels in multiple cancer types for combinatorial therapies. Selection of combination of agents should be based on their potential in killing CAFs, reducing angiogenesis and modifying ECM remodeling by targeting ECM proteins in the tumor microenvironment. Designing the optimum doses of combinational agents in a context specific manner may help in overcoming the hurdles of therapy resistance by promoting the conventional therapeutic effectiveness and reducing tumor burden.

No disclosures were reported.

This work was supported by Indian Council of Medical Research (ICMR), Govt. of India (grant no. 5/3/8/24/2020-ITR).