Versican and versikine: The dynamism of the extracellular matrix

Abstract

Versican is a large chondroitin sulfate/dermatan sulfate proteoglycan in the extracellular matrix and one of the aggrecan/lectican family. Whereas versican is constitutively expressed and serves as a structural macromolecule in some tissues, it is transiently expressed at high levels when the extracellular matrix dynamically changes. There, versican plays an important role in forming the provisional matrix, which is replaced with the “authentic” extracellular matrix, that is, the matrix as it should be. ADAMTS‐1, 4, 5, 9, 15, and 20 cleave versican core protein and are therefore named versicanases. These proteinases have been believed to play a critical role in versican turnover. A cleaved N‐terminal fragment harbors biological functions, and it is termed “versikine.” This review discusses recent advances in the research on the in vivo function of versican and versikine generated by versicanases.

Abbreviations

• Acan
• aggrecan
• ADAMTS
• a disintegrin and metalloproteinase with thrombospondin motifs
• BAV
• bicuspid aortic valve
• BMP
• bone morphogenetic protein
• BPV
• bicuspid pulmonary valve
• bt
• belted
• CAF
• cancer-associated fibroblast
• CCM
• cerebral cavernous malformation
• cDC
• conventional dendritic cell
• CS
• chondroitin sulfate
• ECM
• extracellular matrix
• Egr
• early growth response
• EMT
• epithelial–mesenchymal transition
• FGF
• fibroblast growth factor
• HA
• hyaluronan
• HAPLN
• hyaluronan and proteoglycan link protein
• HC
• heavy chain
• hdf
• heart defect
• LRP-1
• low-density lipoprotein receptor protein-1
• MMP
• matrix metalloproteinase
• PDGF
• platelet-derived growth factor
• SHAP
• serum-derived hyaluronan-associated protein
• SHF
• second heart field
• TAM
• tumor-associated macrophage
• TGF-β
• transforming growth factor
• Vcan
• versican
• VSD
• ventricular septal defect

INTRODUCTION

The tissue most abundant in proteoglycans is cartilage. There, the major proteoglycan is aggrecan (Acan),^1,2 a large chondroitin sulfate (CS) proteoglycan. Although the presence of CS in cartilage was already known, it was not until 1968 that CS was discovered to exist in the form of proteoglycans, the first proteoglycans. Subsequently, various biochemical analyses, especially isopycnic ultracentrifugation, showed that the major proteoglycan of cartilage forms aggregates with hyaluronic acid (HA) and link protein. Molecular cloning was performed, and this proteoglycan was named Acan (gene symbol, Agc, Acan). Acan has a gene family. The family includes versican (Vcan), neurocan, brevican, and dermacan.³ Whereas neurocan and brevican are restricted in nervous tissues, Vcan is expressed ubiquitously. Therefore, Vcan is assumed to be a prototype of the family.

Vcan was initially extracted and isolated from the bovine aorta.⁴ In parallel, an ortholog was extracted and isolated from the chick limb bud. By isopycnic centrifugation, this proteoglycan moved to the middle fraction, whereas Acan was in the bottom fraction, and therefore, the new proteoglycan was named PG-M.⁵

Vcan exhibits characteristic expression patterns. It is constitutively expressed in the aorta⁶ and brain,⁷ serving as a structural macromolecule of the extracellular matrix (ECM). In adults, Vcan is constitutively expressed in the cardiovascular and central nervous system and serves as a structural macromolecule. During embryonic development and when the ECM structure dramatically changes in adults, such as inflammation and repair, Vcan is transiently expressed at high levels, serving as a central molecule of the provisional matrix. Cells start to express molecules of “authentic” ECM, that is, the matrix as it should be, which replace provisional matrix molecules, including Vcan, completing the ECM formation. Researchers explored the mechanism of how transiently expressed Vcan disappears. Since ADAMTS-4 was identified as a critical proteinase to cleave Acan, thereby termed “aggrecanase,⁸” following studies revealed that ADAMTS-1, 5, 9, 15, and 20 cleave Vcan core protein, thereby named versicanases.⁹

Whereas an ECM molecule usually serves as a structural component, it may be degraded by proteinases. In some cases, a cleaved fragment is released from the ECM and gains function. The biologically active molecule generated by cleavage is termed “matrikines” and “matricryptins.”¹⁰ Examples are endostatin from type XVIII collagen,¹¹ canstatin,¹² tetrastatin,¹³ and tumstatin¹⁴ from type IV collagen, and endorepellin from perlecan.¹⁵ The cleavage of the Vcan core protein generates a biologically active molecule termed “versikine.”^16,17 A review has been published on versikine recently.¹⁸ Including recent articles, this review discusses the roles of Vcan and versikine in embryonic development and diseases and the regulation of Vcan degradation and versikine production by versicanases.

VCAN FUNCTIONAL DOMAINS AND MOTIFS

Acan family members share two globular domains, G1 and G3, at N- and C-termini, respectively. One or two CS-attachment domains are located in the middle. Acan has another globular domain, G2, and a keratan sulfate-attachment domain. Compared with Acan, the Vcan core protein contains two EGF-like repeats in the G3 domain.¹⁹ Because of this “versatility,” this PG was named “versican.” Based on these structural features, Vcan is thought to have four functional domains/motifs: N-terminal G1 domain, C-terminal G3 domain, EGF-like motifs in the G3 domain, and CS chains attached to CS domains.

Vcan was known to have four splice variants, V0 (G1-CSα-CSβ-G3), V1 (G1-CSβ-G3), V2 (G1-CSα-G3), V3 (G1-G3).²⁰ While these four variants are commonly referred to as classical, a new variant, V4 (G1-part of CSβ-G3), was discovered several years ago in breast cancer tissue.²¹ Furthermore, another Vcan variant was found in the dorsal root ganglia of a rat spared nerve injury model and named V5 variant.²² The V5 variant has another exon, named exon 8β, in intron 8. The exon 8 contains two termination codons, and western blot analysis revealed the presence of a core protein like G1-CSβ in the dorsal root ganglia of the injury model but not in sham control (Figure 1A).

[IMAGE OMITTED. SEE PDF]

The G1 domain comprises A, B, and B′ subdomains. The A subdomain contains Ig-motif. The B and B′ subdomains are homologous, and they are called “proteoglycan tandem repeat.” Either subdomain contains a “link module,” essential for specific hyaluronan (HA)-binding.²³ Interestingly, whereas CD44 and TGS-6 contain only one link module and binds HA, both Acan and Vcan G1 domain require two link modules tandemly located.^24,25 The G1 domain binds another molecule termed hyaluronan and proteoglycan link protein (HAPLN), and at present, four HAPLNs have been reported.²⁶ Acan A subdomain of the G1 binds HAPLN1 A domain. Acan B–B′ region of the G1 binds HA. HAPLN1 B-B′ region binds HA. Through these interactions, the three molecules form a complex named “proteoglycan aggregate (Acan aggregate).” Although HA and Acan appeared to be adequate to form an aggregate, biochemical analysis²⁴ and analysis of HAPLN1-null mice demonstrated that HAPLN1²⁷ substantially stabilizes the aggregates. Similarly to Acan, Vcan also interacts with both HA and HAPLN1, forming the proteoglycan aggregate (Vcan aggregate). In this context, stable incorporation of Vcan in the ECM is likely dependent on the presence of HA and HAPLN1. Vcan G1 and HAPLN1 bind at their respective B–B′ stretches,²⁵ distinct from the interaction between Acan G1 and HAPLN1 (Figure 1B).²⁴ The mechanism of their interaction remains unclear.

The G3 domain interacts with various ECM molecules, including fibrillin-1, 2, fibulin-1, 2, and tenascin C and R.^28,29 The interaction of Vcan at the G3 domain with various ECM molecules may enable the formation of the supramolecular structure of the ECM, generating a concept of versican interactome.³⁰ Analyses of Vcan gene knockout mice have demonstrated that Vcan regulates transforming growth factor-β (TGF-β) and bone morphogenetic protein (BMP) signaling. As fibrillins sequester TGF-β and BMPs,³¹ Vcan may regulate TGF-β and BMP via fibrillins.

The number and the location of CS chains attached to the CS domains have not been determined. CS chains have been shown to affect ADAMTS5 activity. Mutations at Ser (507) and Ser (525) attenuate ADAMT5 cleavage at the initial cleavage site.³²

VERSICANASES THAT CLEAVE VCAN CORE PROTEIN

In the 1980s, research on osteoarthritis had suggested that Acan is digested by proteinases distinct from matrix metalloproteinases (MMPs), and in 1999, a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS)-4 was identified as a proteinase critical for Acan degradation.³³ The ADAMTS proteinases are secreted zinc endopeptidases comprising 19 family members (ADAMTS-11 was demonstrated to be identical to ADAMTS5). To date, ADAMTS-1, 4, 5, 8, 9, 15, and 20 are known to cleave large proteoglycans of Acan family, thereby termed “proteoglycanases,” and when focusing on Vcan degradation, termed versicanases.”

The initial cleavage site of Vcan core protein by versicanases is E(441)-A(442) (human V1 sequence) located close to the N-terminus of the CSβ domain.⁹ This cleavage generates a sequence DPEAAE, and a specific antibody recognizing a neoepitope DPEAAE is often used for the detection of the N-terminal processed fragment, later termed versikine. A mutation of this initial cleavage site has been shown to generate another cleavage site, E(438)-A(439) (human V1 sequence).^32,34

The enzymatic activity of versicanases on Vcan core protein is different among them. ADAMTS-5 exhibits more potent activity (~18-fold) than ADAMTS-4. ADAMTS-1 shows substantially low activity compared with ADAMTS-4.^35,36 With six members, versicanases may have functional redundancies depending on their expression levels and patterns in tissues and their enzymatic activity levels, raising a concern that knockout mice of one gene would not generate abnormalities. Contrary to this, Adamts9-null mice display embryonic lethality before somite formation; Adamts20-null by spontaneous and ENU-induced mutations, abnormal hair/coat pigmentation with a white belt pattern³⁷; Adamts1-null, growth retardation with adipose tissue malformation, impaired female fertility, enlarged renal calyces, and abnormal adrenal medullary architecture; Adamts4-null, impaired coordination and increased susceptibility to pharmacologically induced seizures. Adamts5-null mice exhibit a significant reduction in cartilage degradation after induction of osteoarthritis, but they show no morphological abnormalities in physiological conditions (information available at ). Besides putative functional redundancies, these versicanases digest other molecules as substrates, which may be associated with the phenotype. As the application of anti-versicanase antibodies for immunostaining is limited, the contribution of individual versicanases is often evaluated by quantitative real-time polymerase chain reaction. Several posttranslational regulations of ADAMTS proteases are present.³⁸ For example, ADAMTS-5 half-life is regulated by endocytosis through the low-density lipoprotein receptor protein-1 (LRP-1) receptor.³⁹ These points above make it difficult to understand the biological functions of vericanases.

VCAN AND ITS PROCESSED FORM VERSIKINE

Genetically engineered Vcan mutant mice (Figure 2) and their combinations with Adamts-null mice have displayed a series of abnormalities in embryonic development and in pathological models. However, it is still difficult to determine whether these abnormalities are caused by deletion/accumulation of Vcan or versikine generated through cleavage of Vcan core protein. Overexpression of versikine provides additive or competitively adverse effects on the original function of Vcan, and it is difficult to discern whether its effects are direct or by interfering with endogenous Vcan function.

[IMAGE OMITTED. SEE PDF]

To understand the impact of Vcan cleavage, an effective way is to replace Vcan with a versicanase-resistant one by mutating the cleavage site(s) and applying it to mouse analysis. Currently, there are two mouse strains whose Vcan is resistant to ADAMTS cleavage, i.e., V1R (R/R for homozygotes),⁴⁰ and Vcan^AA.³⁴ In V1R Vcan, the initial cleavage site 441EARRG445 is replaced with ENVYG. In Vcan^AA Vcan, a stretch encompassing two cleavage sites, DPEAAEARRG, is replaced with DPAAAAARRG. A series of studies using various mutant mice and their abnormalities are discussed below.

Embryonic development

Heart, blood vessels, and smooth muscle

During cardiac development, Vcan expression initiates as early as E8.5, and the regions of Vcan expression shift dramatically,⁴¹ suggesting its involvement in different stages of cardiac development. The Vcan-mutant mice reported for the first time is heart defect (hdf). The homozygote hdf/hdf cardiac defects include the absence of endocardial cushions and a highly dilated primitive atrium and ventricle.⁴² Another strain of Vcan-mutant mice termed Vcan^Δ3/Δ3 express a decreased level of Vcan without the A subdomain and display various heart defects at different developmental stages.⁴³ These mice with C57Bl/6 background die at E10.5 with heart defects similar to hdf/hdf, including dilated ventricles and lack of cardiac jelly. In contrast, Vcan^Δ3/Δ3 mice with a mixed background exhibit ventricular septal defects (VSD) and survive longer up to P0. Their atrioventricular canal (AVC) cushion is much smaller, and the cells there exhibit both endocardial and mesenchymal characteristics with higher levels of proliferation. To investigate the role of Vcan V2 variant, mainly expressed in the brain after birth, Vcan^tm1.1Dzim mice were generated.⁴⁴ Their Vcan genome lacks CSα-encoding exon7, and they do not express V0 and V2 variants. Besides impaired assembling of Ranvier node ECM, homozygote Vcan^tm1.1Dzim mice exhibit cardiac abnormalities, including VSD, the inadequate volume of cushion/leaflet region, and altered pulmonary and aortic outflow tracts.⁴⁵ As developing hearts express V0 and V1, these cardiac abnormalities indicate that V0 contributes to cardiac development. Both hdf heterozygotes and HAPLN1-null mice display VSD, indicating that the proteoglycan aggregate of Vcan, HA, and HAPLN1 is required to complete ventricular septa.⁴⁶ These observations suggest that Vcan is necessary for the generation of space where cells migrate, differentiate, and transform. In the process of cardiac jelly formation, Vcan and HA may generate a temporary hydrated area for endocardial epithelial–mesenchymal transition (EMT). During the formation of septa and valves, Vcan aggregates appear to generate a path for migration of endocardial cells, followed by transformation to the myocardium.

Several studies have shown the physiological relevance of Vcan degradation in cardiovascular development. There are two cardiac progenitor cell lineages: the second heart field (SHF)⁴⁷ and cardiac neural crest cells, and both contribute to the development of the cardiac outflow tract. The expression of Adamts9 continues in the derivatives of SHF, which include the aortic wall, the valvulosinus, and the aortic leaflets. Adamts9^+/LacZ heterozygous mice are viable, and their adult mice show anomalies in the regions where Adamts9 is expressed. They exhibit thickened aortic leaflets with a CS-rich matrix, myxomatous mitral valves with Vcan accumulation, and cartilaginous nodules in the valvular sinus region. These mice also exhibit sponge-like myocardium with protrusions in the left ventricular chamber, presumably due to impaired compaction of myocardium.⁴⁸ As a deletion of fibulin-1, a cofactor for cleavage of Vcan by ADAMTS-1,⁴⁹ displays a similar phenotype,⁵⁰ the sponge-like myocardium of Adamts9^+/LacZ heterozygous mice is likely due to impaired Vcan digestion. Whereas fibrin-1 was reported as a cofactor for Vcan degradation by ADAMTS-1,⁴⁹ a recent study demonstrated that the complex of fibulin-1 and Vcan form the expansive subendothelial ECM and facilitate the migration of smooth muscle cells, leading to the closure of ductus arteriosus.⁵¹ Collectively, the effects of fibulin-1 on Vcan accumulation and degradation may be context-dependent. Adamts5-null mice display bicuspid aortic valve (BAV) and bicuspid pulmonary valve (BPV), concomitant with Vcan accumulation. A reduction of Smad2-signaling increases the penetrance, implicating regulation of TGFβ-signaling by Vcan degradation.⁵² These valvular abnormalities in Adamts5-null mice are transiently rescued in P0~P7 and could be ascribed to compensatory Vcan degradation by ADAMTS-1, MMP2, and MMP9.⁵³

Cerebral cavernous malformations (CCMs) are common vascular abnormalities that show small blood vessels with thin walls and lead to blood leaking and hemorrhage. Their molecular basis is the loss of function of the CCM complex, which negatively regulates MEKK3 signaling. Endothelial loss of CCM function enhances MEKK3 signaling and causes CCM. Endothelial loss of ADAMTS-5 reduces CCM formation in the neonatal mouse model, and the gain of ADAMTS-5 generates the early lesion. Lowering Vcan expression reduces the CCM burden, indicating that the lesion formation requires Vcan degradation by ADAMTS-5 but not the loss of Vcan.⁵⁴

Recently, analysis of uterine-specific Vcan deletion mice (uKO mice) revealed the role of Vcan in pregnancy.⁵⁵ These mice exhibit fetal growth restriction and maternal hypertension due to insufficient spiral artery dilation. This insufficiency is likely caused by a decrease in the proliferation of tissue-resident natural killer cells required for spiral artery dilation.

Cartilage and joint development and interdigital web regression

Vcan is transiently expressed at high levels in mesenchymal condensation areas of cartilage primordium, and it rapidly disappears during chondrocyte differentiation.⁵ Micromass culture of hdf/hdf limb bud cells revealed the absence of type II collagen expression, indicating that Vcan is required for chondrocyte differentiation of limb bud cells.⁵⁶ In contrast, Prx1-cre: Vcan^flox/flox (Prx1-Vcan) mice lacking Vcan expression in mesenchymal condensation areas demonstrate that Vcan is dispensable for cartilage development. These mice exhibit endochondral ossification but display tilting of the joint and a slight delay of chondrocyte differentiation due to altered localization of extracellular TGF-β and its signaling.⁵⁷ Vcan expression remains on the surface of articular cartilage⁵⁸ and synovia.⁵⁹ Prx1-Vcan mice show the impaired formation of cruciate ligaments.⁶⁰ While Vcan is not essential for cartilage development, it is required for joint and ligament formation. Analysis of Prx1-Vcan mice alone does not clarify the impact of Vcan processing, which needs to be investigated.

Adamts9^+/−; bt/bt mice exhibit fully penetrant cleft palate, whereas both Adamts9^+/−; bt/+ and bt/bt mice delayed cleft closure, indicating the co-operation of these proteinases. Because of different expression patterns of Adamts5 and Adamts20 in the cell lineages in the palate, together with different frequencies of the cleft palate between Adamts9^+/−; bt/+ and bt/bt mice, the contribution of these enzymes appear independent. The palatal mesenchyme of Adamts9^+/−; bt/bt mice exhibits decreased cell proliferation, a lower cell density, and decreased processing of Vcan. A combination with haploinsufficient Vcan^hdf/+ mice increases the penetrance of bt/bt cleft palate, suggesting versikine is required for cell proliferation.⁶¹

The action of versikine was demonstrated in a study on soft tissue syndactyly in combined Adamts mutant mice.⁶² Different versikine activities by a combination of Adamts5−, Adamts9−, and Adamts20− knockout alleles exhibit a different ratio of the penetrance of interdigital web regression by reduced apoptosis in the autopods. Adamts5−/−; Adamts20−/− (bt/bt), Adamts5−/−; Adamts9 +/−; Adamts20+/− (bt/bt), and Adamts9+/−; Adamts20−/− (bt/bt), show soft tissue syndactyly with full penetrance. These versicanases, Vcan, and fibulin-1 are coexpressed in the interdigital tissue, and haploinsufficiency of either Vcan or Fbln-1 leads to highly penetrant syndactyly in bt/bt mice, suggesting that versikine facilitates the apoptosis. Importantly, the application of versikine induces apoptosis in Adamts5−/−; Adamts20−/− (bt/bt) web.

Homozygote mice of both homozygote V1R (R/R) and Vcan^AA/AA mice exhibit syndactyly with a ~20% penetrance, supporting the previous report showing that digestion of Vcan by ADAMTSs is required for interdigital web regression.⁶² Whereas Vcan^AA/AA mice show ~20% penetrance of syndactyly, Adamts20-null bt/bt; Vcan^AA/AA mice achieve its full penetrance, suggesting that ADAMTS20 cleaves Vcan core protein at other sites, or that digestion of other substrates by ADAMTS20 is required for interdigital web regression.³⁴

Skin and hair follicles

Vcan^Δ3/Δ3 dermis at P0 exhibits decreased collagen content and cell density. In cultured dermal fibroblasts, while the level of collagen deposition is similar, collagen biosynthesis significantly decreases in Vcan^Δ3/Δ3 fibroblasts. Vcan^Δ3/Δ3 fibroblasts in culture show a reduced TGF-β storage in the ECM and downregulation of TGF-β-Smad2/3 signal transduction, concomitant with downregulation of expression of transcription factors, early growth response (Egr) 2 and 4, which act downstream of TGF-β signaling.⁶³

The mouse has a clear hair cycle consisting of anagen (growth phase), catagen (transition phase), and telogen (resting phase). Vcan is expressed at high levels in dermal papillae of anagen, followed by catagen. In contrast, it is absent in vellus hair follicles of bald male patterns even in anagen.^64,65 Dermal papillae cells isolated from Vcan promoter driven-GFP mice were applied to a spheroid culture, and a half-diluted combination of adipogenic and osteogenic medium together with FGF2 and PDGF-AA promoted expression of dermal papillae-specific gene expression.⁶⁶ Although several researchers used Vcan expression as a dermal papillae cell marker,⁶⁷ only a limited number of reports have implied its role in dermal papillae formation. The Vcan^Δ3/Δ3 dermal papillae show decreased levels of Vcan deposition as well as the number of hair follicles.⁶³ RNAi targeted to Vcan efficiently suppresses the aggregative growth of dermal papilla cells in vitro.⁶⁸ HA is unlikely to be involved, as it is not colocalized in dermal papillae where Vcan is present.⁶³ Whereas Vcan is used as a marker for dermal papilla cells, the role of Vcan in the growth and maintenance of hair follicles remains to be understood.

Pathological conditions

Vcan has been shown to play important roles in cardiovascular and neurological diseases. There are excellent reviews on Vcan in cardiovascular diseases^69–71 and neurological disorders.⁷² This section focuses on the role of Vcan in inflammation and tumor progression.

Inflammation and fibrosis

Vcan secreted by monocyte/macrophages and stromal fibroblasts may regulate inflammation via HA-matrix.⁷³ In a mouse model of acute pulmonary inflammation, exposure with lipopolysaccharide (LPS), TLR3 agonist Pseudomonas aeruginosa, poly (I: C), and cockroach allergen, increases Vcan expression.⁷⁴ Local Vcan deficiency in the lung decreases the HA-Vcan matrix and inhibits leukocyte infiltration, ameliorating poly(I: C)-treated acute inflammation.⁷⁵ Analysis of Vcan expression in bone marrow-derived macrophages and macrophage-specific conditional knockout mice of Vcan (LyzM-Vcan) revealed that type I IFN signaling induces Vcan expression, leading to the production of IFN-β and IL-10, two important anti-inflammatory cytokines.⁷⁴

Vcan has been reported to serve as a marker for hepatic⁷⁶ and renal fibrosis.⁷⁷ Its expression is upregulated in circulating monocytes of systemic sclerosis patients.⁷⁸ These findings suggest the involvement of Vcan in organ and systemic fibrosis.

In vitro studies using Adamts5-null and Adamts5-null; Vcan^hdf/+ dermal fibroblasts suggested that accumulation of Vcan by deficient ADAMTS activity increases the pericellular HA-Vcan matrix and facilitates the transition of fibroblasts to myofibroblasts.⁷⁹ In wound healing experiments, V1R homozygote (R/R) mice show faster wound healing than the wild type, concomitant with more accumulation of Vcan, activated TGF-β-signaling, and an increasing number of myofibroblasts. Cultured R/R dermal fibroblasts exhibit increased deposition of Vcan, collagens, and HA, with upregulation of Smad2/3 signaling. Therefore, the initial ADAMTS-cleavage site is the major cleavage site for Vcan turnover, although ADAMTSs digest Vcan core protein at various sites,⁸⁰ and accumulation in the ECM of R/R affects the distribution of other ECM molecules and TGF-β-signaling.⁴⁰

Generation and analysis of the dextran sodium sulfate (DSS)-colitis model in R/R revealed several functional aspects of Vcan and versikine.⁸¹ R/R colitis is less severe than wild-type. Even with higher levels of macrophage infiltration, R/R inflamed colon retains the structure of collagen fibers and higher levels of myofibroblasts. The phenotype of R/R mice in inflammation does not determine whether it is due to the accumulation of versicanase-resistant Vcan or decreased generation of versikine. In vitro analysis for the coculture of macrophages and stromal fibroblasts and versikine overexpression revealed that versikine directly inhibits the differentiation of fibroblasts toward myofibroblasts. In both wound healing and DSS-colitis models, deposition of the accumulated Vcan gradually disappeared, indicating that Vcan core protein is cleaved at other sites in vivo. In these studies, the cleavage and the presence of versikine were evaluated using a neoepitope antibody. Even though the cleavage at the site determines the level of tissue damage and the rate of the repair process, several proteinases are likely involved in them.

Related to inflammation, recently, Vcan was reported to be associated with abnormal hematopoiesis after cardiovascular diseases. In patients with arterial hypertension alone, both hypertension and atherosclerosis or acute myocardial infarction show increased proliferation of hematopoietic stem cells. A mouse model of myocardial infarction exhibited changes in bone marrow vasculature with prominent angiogenesis. Further analysis using Cdh5^CreERT2; Vcan^flox/flox mice, lacking bone marrow endothelial Vcan expression, revealed that Vcan participates in abnormal hematopoiesis after myocardial infarction.⁸²

Cancer progression

Vcan is expressed in cancer cells, stromal cells, and macrophages in the tumor microenvironment. Elevated expression levels of Vcan in cancer have been reported in most cancer types, such as glioblastoma, melanoma, osteosarcoma, lymphoma, and cancers of the breast, prostate, and ovary.⁸³ Some studies have demonstrated the effects of Vcan on cancer cell behavior. Silencing of Vcan V0/V1 and CD44 in the SK-mel-131 human melanoma cells decreases their cell proliferation and migration, indicating Vcan facilitates tumor progression via CD44-mediated signaling.⁸⁴ Vcan expression in Lewis lung carcinoma cells activates the production of TNF-α and IL-6 through TLR2–TLR6–CD14 signaling in macrophages and myeloid cells, which enhances metastatic tumor growth.⁸⁵

Vcan expressed by stromal cells affects cancer behavior. In coculture experiments of ovarian cancer cells and cancer-associated fibroblasts (CAFs), TGF-β upregulates Vcan expression in CAFs, enhancing the aggressiveness of cancer cells.⁸⁶ Vcan expression in the stroma correlates with the accumulation of tumor-associated macrophages (TAM) at advanced stages of tumor progression.⁸⁷ In QRsP11 fibrosarcoma implant experiments, local deletion of stromal Vcan expression increases the tumor growth, and V3 overexpression inhibits tumor cell growth. The increased tumor mass is associated with a decrease in fibroblasts and collagen fibers, suggesting that Vcan is required for maintaining fibroblasts and collagen fiber structure in the tumor microenvironment.⁸⁸ In B16F10 implant experiments, hdf/+ mice show reduced tumor mass with a lower capillary density and accumulation of capillaries at the tumor periphery, suggesting a role of stromal Vcan in tumor angiogenesis.⁸⁹

In MMTV-PyMT transgenic mice that exhibit spontaneous breast cancer, monocytic cells expressing Vcan in bone marrow migrate to the lungs, recruit metastatic breast cancer cells⁹⁰ and induce mesenchymal to epithelial transition.⁹¹

There are increasing reports addressing the Vcan and versikine on cancer behavior. Vcan-cleavage was shown to be involved in the progression of multiple myeloma.¹⁷ Vcan is abundantly secreted by myeloma-associated macrophages, and Vcan itself prompts tolerogenic polarization of antigen-presenting cells. Stroma-derived ADAMTS-1 cleaves Vcan, and versikine induces early expression of interleukin-1β (IL-1β) and IL-6 by myeloma-associated macrophages via both Tpl2-dependent and independent signaling pathways, suggesting versikine as a damage-associated molecular pattern, which facilitates immune-sensing. The role of versikine was further investigated for colorectal cancer (CRC). Tissue microarray analysis revealed a good correlation of T-cell infiltration with the levels of versikine.⁹² Recently, the involvement of versikine in cancer stromal reaction with T-cell-induced inflammation has been reported.⁹³ Although localized in the tumor marginal region, stimulatory type 1 conventional dendritic cells (cDC1s) regulate effector cell influx into the tumor microenvironment. Versikine promotes their accumulation in vivo, and tumor stroma with abundant cDC1 elicits T cell-inflamed tumor microenvironment. ADAMTS-15, Vcan, and versikine are colocalized in human prostate cancer biopsies. Overexpression of ADAMTS15 inhibits the proliferation and migration of PC3 prostate cancer cells, implicating the tumor-suppressing function of ADAMTS-15, partly via Vcan processing.⁹⁴

Several factors are involved in building up the cancer microenvironment. Cancer cells, stromal fibroblasts, and inflammatory cells synthesize Vcan, versicanases, and HA. The effects of Vcan and versikine are likely context-dependent.

VERSIKINE: MECHANISM OF ACTION

Since versikine is generated from Vcan, versikine is not present in areas where Vcan is or was not expressed, and the expression levels of Vcan as substrate affect the levels of versikine, as well as versicanase activity.

The structure of versikine is the G1 domain with a short stretch of a peptide sequence when originated from the V1 variant. Therefore, its function is likely attributed to the HA-binding property. Vcan^Δ3/Δ3 embryonic fibroblasts, expressing a decreased level of Vcan with weaker HA-binding affinity, show premature senescence via constitutively active CD44-ERK1/2 mediated signaling,⁹⁵ suggesting that Vcan binding with HA influences CD44-mediated signal transduction.

As the core protein of V0/V1 weighs more than 200 kDa and intact Vcan could even exceed 1000 kDa if considering the binding of the CS chain, the G1 domain of intact Vcan may bind to HA at regular intervals. In contrast, versikine from V1 weighing ~37 kDa may bind in large numbers, occupying the empty spaces in the HA chain and may alter the structure of HA, which could potentially impact the signal transduction mediated by HA and CD44.

Treatment of cultured dermal fibroblasts with recombinant Vcan G1 domain has been shown to alter the structure of HA into an aggregated cable-like structure, which bridges fibroblasts,⁹⁶ whereas endogenous intact Vcan V0/V1 does not prevent the cable formation. Interestingly, the G1 remains on the HA cables for up to 4 weeks, presumably maintaining the HA-cable structure. Vcan G1 has been shown to be present not only in the tissue repair process after acute inflammation but also in the dermis under chronic inflammation. In the pressure ulcers generated through chronic refractory wounds, the cleaved G1 domain colocalizes with a heavy chain (HC)-HA complex (also termed serum-derived HA-associated protein (SHAP)-HA complex). Vcan G1 directly binds to two HCs and forms a G1-HC (SHAP)-HA complex, where Vcan G1 is in the form of aggregates, facilitating the expansion of the pericellular HA matrix.⁹⁷ With these observations, versikine may assemble individual HA chains to form a cable structure and alter the pericellular microenvironment, by forming HC: HA complex. So far, changes in the microstructure of HA chains cannot be observed in cell biological experiments. Once an assay system capable of analyzing the fine structure of HA is established, the mechanism of versikine action in detail could be elucidated.

Some studies have provided evidence against HA-mediated signaling by versikine. Vcan G1 enhances adenoviral vector transgene expression, whose mechanism is independent of HA-CD44 signaling.⁹⁸ Studies using versikine-overexpressing cells have shown cell-autonomous signaling, unaffected by CD44 expression.⁹³ HAPLN1 secreted from bone marrow stromal cells enhances resistance to bortezomib, a proteasome inhibitor, via atypical nuclear factor-kB (NF-kB) activation in multiple myeloma cells.⁹⁹ Domain functional analysis demonstrated that a link module of either B or B′ fragment has adequate activity, and their NF-kB activation does not involve HA-binding. It remains to be understood how versikine signals to the target cells.

CONCLUSION

The Vcan-versikine axis is an elaborate system that supports the dynamism of the ECM. Transiently expressed Vcan serves as a key player in the provisional matrix. Versicanases are expressed at various times and locations and participate in the breakdown of Vcan and the generation of bioactive molecule versikine. Other ECM molecules, such as fibulin-1, may affect the Vcan-versikine axis. Furthermore, ADAMTS-4 has recently been shown to degrade fibronectin and latent TGF-β-binding protein, regulating TGF-β-signaling.¹⁰⁰ Further studies are necessary to clarify the dynamics of ECM, which includes not only the Vcan-ADAMTS axis but also the degradation of various ECM molecules.

AUTHOR CONTRIBUTION

Hideto Watanabe: Writing—review and editing.

ACKNOWLEDGMENTS

I thank Dr. Anthony Day for the valuable discussion.

CONFLICT OF INTEREST STATEMENT

The author declares no conflict of interest.

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

ETHICS STATEMENT

The author has nothing to report.

References

H. Watanabe, Y. Yamada, K. Kimata, J. Biochem. 1998, 124(4), 687.

A. J. Hayes, J. Melrose, Biomolecules 2020, 10(9), 1244.

J. S. Kang, T. Oohashi, Y. Kawakami, Y. Bekku, J. C. Izpisúa Belmonte, Y. Ninomiya, Mech. Dev. 2004, 121(3), 301.

T. R. Oegema, Jr., V. C. Hascall, R. Eisenstein, J. Biol. Chem. 1979, 254(4), 1312.

K. Kimata, Y. Oike, K. Tani, T. Shinomura, M. Yamagata, M. Uritani, S. Suzuki, J. Biol. Chem. 1986, 261(29), [eLocator: 13517].

K. L. Olin, S. Potter‐Perigo, P. H. R. Barrett, T. N. Wight, A. Chait, J. Biol. Chem. 1999, 274(49), [eLocator: 34629].

Y. Yamaguchi, Cell. Mol. Life Sci. 2000, 57(2), 276.

E. Arner, Curr. Opin. Pharmacol. 2002, 2(3), 322.

J. D. Sandy, J. Westling, R. D. Kenagy, M. L. Iruela‐Arispe, C. Verscharen, J. C. Rodriguez‐Mazaneque, D. R. Zimmermann, J. M. Lemire, J. W. Fischer, T. N. Wight, A. W. Clowes, J. Biol. Chem. 2001, 276(16), [eLocator: 13372].

S. Ricard‐Blum, R. Salza, Exp. Dermatol. 2014, 23(7), 457.

M. S. O'Reilly, T. Boehm, Y. Shing, N. Fukai, G. Vasios, W. S. Lane, E. Flynn, J. R. Birkhead, B. R. Olsen, J. Folkman, Cell 1997, 88(2), 277.

G. D. Kamphaus, P. C. Colorado, D. J. Panka, H. Hopfer, R. Ramchandran, A. Torre, Y. Maeshima, J. W. Mier, V. P. Sukhatme, R. Kalluri, J. Biol. Chem. 2000, 275(2), 1209.

S. Brassart‐Pasco, K. Sénéchal, J. Thevenard, L. Ramont, J. Devy, L. Di Stefano, A. Dupont‐Deshorgue, S. Brézillon, J. Feru, J. F. Jazeron, M. D. Diebold, S. Ricard‐Blum, F. X. Maquart, J. C. Monboisse, PLoS One 2012, 7(4), [eLocator: e29587].

Y. Maeshima, A. Sudhakar, J. C. Lively, K. Ueki, S. Kharbanda, C. R. Kahn, N. Sonenberg, R. O. Hynes, R. Kalluri, Science 2002, 295(5552), 140.

M. Mongiat, S. M. Sweeney, J. D. San Antonio, J. Fu, R. V. Iozzo, J. Biol. Chem. 2003, 278(6), 4238.

S. Nandadasa, S. Foulcer, S. S. Apte, Matrix Biol. 2014, 35, 34.

C. Hope, S. Foulcer, J. Jagodinsky, S. X. Chen, J. L. Jensen, S. Patel, C. Leith, I. Maroulakou, N. Callander, S. Miyamoto, P. Hematti, S. S. Apte, F. Asimakopoulos, Blood 2016, 128(5), 680.

K. P. Timms, S. B. Maurice, Glycobiology 2020, 30(6), 365.

D. R. Zimmermann, E. Ruoslahti, EMBO J. 1989, 8(10), 2975.

T. Shinomura, M. Zako, K. Ito, M. Ujita, K. Kimata, J. Biol. Chem. 1995, 270(17), [eLocator: 10328].

P. Kischel, D. Waltregny, B. Dumont, A. Turtoi, Y. Greffe, S. Kirsch, E. De Pauw, V. Castronovo, Int. J. Cancer 2010, 126(3), 640.

O. Bogen, O. Bender, P. Alvarez, M. Kern, S. Tomiuk, F. Hucho, J. D. Levine, Mol. Pain. 2019, 15, [eLocator: 174480691987455].

A. J. Day, G. D. Prestwich, J. Biol. Chem. 2002, 277(7), 4585.

H. Watanabe, S. C. Cheung, N. Itano, K. Kimata, Y. Yamada, J. Biol. Chem. 1997, 272(44), [eLocator: 28057].

K. Matsumoto, M. Shionyu, M. Go, K. Shimizu, T. Shinomura, K. Kimata, H. Watanabe, J. Biol. Chem. 2003, 278(42), [eLocator: 41205].

A. P. Spicer, A. Joo, R. A. Bowling, Jr., J. Biol. Chem. 2003, 278(23), [eLocator: 21083].

H. Watanabe, Y. Yamada, Nat. Genet. 1999, 21(2), 225.

Y. J. Wu, D. P. L. Pierre, J. Wu, A. J. Yee, B. B. Yang, Cell Res. 2005, 15(7), 483.

S. Islam, H. Watanabe, J. Histochem. Cytochem. 2020, 68(11), 763.

F. Tang, J. E. Brune, M. Y. Chang, S. R. Reeves, W. A. Altemeier, C. W. Frevert, Am. J. Physiol. Cell Physiol. 2022, 323(2), C249.

G. Sengle, L. Y. Sakai, Matrix Biol. 2015, 47, 3.

S. J. Foulcer, C. M. Nelson, M. V. Quintero, B. Kuberan, J. Larkin, M. T. Dours‐Zimmermann, D. R. Zimmermann, S. S. Apte, J. Biol. Chem. 2014, 289(40), [eLocator: 27859].

M. D. Tortorella, T. C. Burn, M. A. Pratta, I. Abbaszade, J. M. Hollis, R. Liu, S. A. Rosenfeld, R. A. Copeland, C. P. Decicco, R. Wynn, A. Rockwell, F. Yang, J. L. Duke, K. Solomon, H. George, R. Bruckner, H. Nagase, Y. Itoh, D. M. Ellis, H. Ross, B. H. Wiswall, K. Murphy, M. C. Hillman, G. F. Hollis, R. C. Newton, R. L. Magolda, J. M. Trzaskos, E. C. Arner, Science 1999, 284(5420), 1664.

S. Nandadasa, C. Burin des Roziers, C. Koch, K. Tran‐Lundmark, M. T. Dours‐Zimmermann, D. R. Zimmermann, S. Valleix, S. S. Apte, Matrix Biol. Plus 2021, 10, [eLocator: 100064].

S. Santamaria, K. Yamamoto, A. Teraz‐Orosz, C. Koch, S. S. Apte, R. de Groot, D. A. Lane, J. Ahnström, Sci. Rep. 2019, 9(1), [eLocator: 10914].

A. F. Minns, Y. Qi, K. Yamamoto, K. Lee, J. Ahnström, S. Santamaria, J. Biol. Chem. 2023, 299(4), [eLocator: 103048].

R. Kelwick, I. Desanlis, G. N. Wheeler, D. R. Edwards, Genome Biol. 2015, 16(1), 113.

K. W. J. Rose, N. Taye, S. Z. Karoulias, D. Hubmacher, Front. Mol. Biosci. 2021, 8, [eLocator: 701959].

K. Yamamoto, L. Troeberg, S. D. Scilabra, M. Pelosi, C. L. Murphy, D. K. Strickland, H. Nagase, FASEB J. 2013, 27(2), 511.

S. Islam, K. Chuensirikulchai, S. Khummuang, T. Keratibumrungpong, P. Kongtawelert, W. Kasinrerk, S. Hatano, A. Nagamachi, H. Honda, H. Watanabe, Matrix Biol. 2020, 87, 77.

D. J. Henderson, A. J. Copp, Circ. Res. 1998, 83(5), 523.

C. H. Mjaatvedt, H. Yamamura, A. A. Capehart, D. Turner, R. R. Markwald, Dev. Biol. 1998, 202(1), 56.

S. Hatano, K. Kimata, N. Hiraiwa, M. Kusakabe, Z. Isogai, E. Adachi, T. Shinomura, H. Watanabe, Glycobiology 2012, 22(9), 1268.

M. T. Dours‐Zimmermann, K. Maurer, U. Rauch, W. Stoffel, R. Fassler, D. R. Zimmermann, J. Neurosci. 2009, 29(24), 7731.

T. A. Burns, M. T. Dours‐Zimmermann, D. R. Zimmermann, E. L. Krug, S. Comte‐Walters, L. Reyes, M. A. Davis, K. L. Schey, J. H. Schwacke, C. B. Kern, C. H. Mjaatvedt, PLoS One 2014, 9(2), [eLocator: e89133].

E. E. Wirrig, B. S. Snarr, M. R. Chintalapudi, J. L. O'Neal, A. L. Phelps, J. L. Barth, V. M. Fresco, C. B. Kern, C. H. Mjaatvedt, B. P. Toole, S. Hoffman, T. C. Trusk, W. S. Argraves, A. Wessels, Dev. Biol. 2007, 310(2), 291.

R. G. Kelly, Curr. Top. Dev. Biol. 2012, 100, 33.

C. B. Kern, A. Wessels, J. McGarity, L. J. Dixon, E. Alston, W. S. Argraves, D. Geeting, C. M. Nelson, D. R. Menick, S. S. Apte, Matrix Biol. 2010, 29(4), 304.

N. V. Lee, J. C. Rodriguez‐Manzaneque, S. N. M. Thai, W. O. Twal, A. Luque, K. M. Lyons, W. S. Argraves, M. L. Iruela‐Arispe, J. Biol. Chem. 2005, 280(41), [eLocator: 34796].

M. A. Cooley, V. M. Fresco, M. E. Dorlon, W. O. Twal, N. V. Lee, J. L. Barth, C. B. Kern, M. L. Iruela‐Arispe, W. S. Argraves, Dev. Dyn. 2012, 241(2), 303.

S. Ito, U. Yokoyama, T. Nakakoji, M. A. Cooley, T. Sasaki, S. Hatano, Y. Kato, J. Saito, N. Nicho, S. Iwasaki, M. Umemura, T. Fujita, M. Masuda, T. Asou, Y. Ishikawa, Arterioscler., Thromb., Vasc. Biol. 2020, 40(9), 2212.

L. E. Dupuis, H. Osinska, M. B. Weinstein, R. B. Hinton, C. B. Kern, J. Mol. Cell. Cardiol. 2013, 60, 50.

L. E. Dupuis, S. E. Evins, M. C. Abell, M. E. Blakley, S. L. Horkey, J. L. Barth, C. B. Kern, J. Cardiovasc. Dev. Dis. 2023, 10(1), 27.

C. C. Hong, A. T. Tang, M. R. Detter, J. P. Choi, R. Wang, X. Yang, A. A. Guerrero, C. F. Wittig, N. Hobson, R. Girard, R. Lightle, T. Moore, R. Shenkar, S. P. Polster, L. M. Goddard, A. A. Ren, N. A. Leu, S. Sterling, J. Yang, L. Li, M. Chen, P. Mericko‐Ishizuka, L. E. Dow, H. Watanabe, M. Schwaninger, W. Min, D. A. Marchuk, X. Zheng, I. A. Awad, M. L. Kahn, J. Exp. Med. 2020, 217(10), 1.

Y. Sagae, A. Horie, A. Yanai, T. Ohara, B. Nakakita, Y. Kitawaki, A. Okunomiya, H. Tani, K. Yamaguchi, J. Hamanishi, J. P. Lydon, T. Daikoku, H. Watanabe, M. Mandai, Matrix Biol. 2023, 115, 16.

D. R. Williams, Jr., A. R. Presar, A. T. Richmond, C. H. Mjaatvedt, S. Hoffman, A. A. Capehart, Biochem. Biophys. Res. Commun. 2005, 334(3), 960.

K. Choocheep, S. Hatano, H. Takagi, H. Watanabe, K. Kimata, P. Kongtawelert, H. Watanabe, J. Biol. Chem. 2010, 285(27), [eLocator: 21114].

K. Matsumoto, N. Kamiya, K. Suwan, F. Atsumi, K. Shimizu, T. Shinomura, Y. Yamada, K. Kimata, H. Watanabe, J. Biol. Chem. 2006, 281(26), [eLocator: 18257].

J. B. Shepard, H. A. Krug, B. A. LaFoon, S. Hoffman, A. A. Capehart, Int. J. Biol. Sci. 2007, 3(6), 380.

T. Higuchi, D. Suzuki, T. Watanabe, K. Fanhchaksai, K. Ota, K. Yokoo, H. Furukawa, H. Watanabe, PLoS One 2021, 16(4), [eLocator: e0250366].

H. Enomoto, C. M. Nelson, R. P. T. Somerville, K. Mielke, L. J. Dixon, K. Powell, S. S. Apte, Development 2010, 137(23), 4029.

D. R. McCulloch, C. M. Nelson, L. J. Dixon, D. L. Silver, J. D. Wylie, V. Lindner, T. Sasaki, M. A. Cooley, W. S. Argraves, S. S. Apte, Dev. Cell 2009, 17(5), 687.

S. Hatano, N. Nagai, N. Sugiura, J. Tsuchimoto, Z. Isogai, K. Kimata, A. Ota, S. Karnan, Y. Hosokawa, H. Watanabe, Connect. Tissue Res. 2018, 59(2), 178.

T. Soma, M. Tajima, J. Kishimoto, J. Dermatol. Sci. 2005, 39(3), 147.

D. L. du Cros, R. G. LeBaron, J. R. Couchman, J. Invest. Dermatol. 1995, 105(3), 426.

T. Kazi, I. Niibe, A. Nishikawa, T. Matsuzaki, FEBS Open Bio 2019, 10(2), 197.

C. F. Huang, Y. J. Chang, Y. Y. Hsueh, C. W. Huang, D. H. Wang, T. C. Huang, Y. T. Wu, F. C. Su, M. Hughes, C. M. Chuong, C. C. Wu, Sci. Rep. 2016, 6, [eLocator: 26436].

M. Feng, G. Yang, J. Wu, Clin. Exp. Dermatol. 2011, 36(1), 77.

T. N. Wight, Matrix Biol. 2018, 71‐72, 396.

T. N. Wight, J. Am. Coll. Cardiol. 2018, 72(13), 1504.

T. N. Wight, I. Kang, S. P. Evanko, I. A. Harten, M. Y. Chang, O. M. T. Pearce, C. E. Allen, C. W. Frevert, Front. Immunol. 2020, 11, 512.

M. M. Anwar, E. Özkan, Y. Gürsoy‐Özdemir, Eur. J. Neurosci. 2022, 56(9), 5453.

A. C. Petrey, C. A. de la Motte, Front. Immunol. 2014, 5, 101.

M. Y. Chang, I. Kang, M. Gale, A. M. Manicone, M. G. Kinsella, K. R. Braun, T. Wigmosta, W. C. Parks, W. A. Altemeier, T. N. Wight, C. W. Frevert, Am. J. Physiol. Lung Cell. Mol. Physiol. 2017, 313(6), L1069.

I. Kang, I. A. Harten, M. Y. Chang, K. R. Braun, A. Sheih, M. P. Nivison, P. Y. Johnson, G. Workman, G. Kaber, S. P. Evanko, C. K. Chan, M. J. Merrilees, S. F. Ziegler, M. G. Kinsella, C. W. Frevert, T. N. Wight, J. Biol. Chem. 2017, 292(1), 51.

T. N. Bukong, S. B. Maurice, B. Chahal, D. F. Schaeffer, P. J. Winwood, Lab. Invest. 2016, 96(3), 361.

M. Rudnicki, P. Perco, H. Neuwirt, S. J. Noppert, J. Leierer, J. Sunzenauer, S. Eder, C. Zoja, K. Eller, A. R. Rosenkranz, G. A. Müller, B. Mayer, G. Mayer, PLoS One 2012, 7(9), [eLocator: e44891].

A. Masuda, H. Yasuoka, T. Satoh, Y. Okazaki, Y. Yamaguchi, M. Kuwana, Arthritis Res. Ther. 2013, 15(4), R74.

N. Hattori, D. A. Carrino, M. E. Lauer, A. Vasanji, J. D. Wylie, C. M. Nelson, S. S. Apte, J. Biol. Chem. 2011, 286(39), [eLocator: 34298].

D. R. Martin, S. Santamaria, C. D. Koch, J. Ahnström, S. S. Apte, J. Proteomics 2021, 249, [eLocator: 104358].

S. Islam, N. Jahan, A. Shahida, S. Karnan, H. Watanabe, Matrix Biol. 2022, 107, 59.

D. Rohde, K. Vandoorne, I. H. Lee, J. Grune, S. Zhang, C. S. McAlpine, M. J. Schloss, R. Nayar, G. Courties, V. Frodermann, G. Wojtkiewicz, L. Honold, Q. Chen, S. Schmidt, Y. Iwamoto, Y. Sun, S. Cremer, F. F. Hoyer, O. Iborra‐Egea, C. Muñoz‐Guijosa, F. Ji, B. Zhou, R. H. Adams, J. D. Wythe, J. Hidalgo, H. Watanabe, Y. Jung, A. M. van der Laan, J. J. Piek, Y. Kfoury, P. A. Désogère, C. Vinegoni, P. Dutta, R. I. Sadreyev, P. Caravan, A. Bayes‐Genis, P. Libby, D. T. Scadden, C. P. Lin, K. Naxerova, F. K. Swirski, M. Nahrendorf, Nat. Cardiovasc. Res. 2022, 1(1), 28.

C. Ricciardelli, A. J. Sakko, M. P. Ween, D. L. Russell, D. J. Horsfall, Cancer Metastasis Rev. 2009, 28(1–2), 233.

D. Hernández, L. Miquel‐Serra, M. J. Docampo, A. Marco‐Ramell, A. Bassols, Int. J. Mol. Med. 2011, 27(2), 269.

S. Kim, H. Takahashi, W. W. Lin, P. Descargues, S. Grivennikov, Y. Kim, J. L. Luo, M. Karin, Nature 2009, 457(7225), 102.

T. L. Yeung, C. S. Leung, K. K. Wong, G. Samimi, M. S. Thompson, J. Liu, T. M. Zaid, S. Ghosh, M. J. Birrer, S. C. Mok, Cancer Res. 2013, 73(16), 5016.

D. C. Dos Reis, K. A. Damasceno, C. B. de Campos, E. S. Veloso, G. R. A. Pêgas, L. R. Kraemer, M. A. Rodrigues, M. S. Mattos, D. A. Gomes, P. P. Campos, E. Ferreira, R. C. Russo, G. D. Cassali, Front. Oncol. 2019, 9, 577.

K. Fanhchaksai, F. Okada, N. Nagai, P. Pothacharoen, P. Kongtawelert, S. Hatano, T. Nakamura, H. Watanabe, Int. J. Cancer 2016, 138, 630–641.

K. Asano, C. M. Nelson, S. Nandadasa, N. Aramaki‐Hattori, D. J. Lindner, T. Alban, J. Inagaki, T. Ohtsuki, T. Oohashi, S. S. Apte, S. Hirohata, Sci. Rep. 2017, 7(1), [eLocator: 17225].

D. Gao, N. Joshi, H. Choi, S. Ryu, M. Hahn, R. Catena, H. Sadik, P. Argani, P. Wagner, L. T. Vahdat, J. L. Port, B. Stiles, S. Sukumar, N. K. Altorki, S. Rafii, V. Mittal, Cancer Res. 2012, 72(6), 1384.

D. Gao, L. T. Vahdat, S. Wong, J. C. Chang, V. Mittal, Cancer Res. 2012, 72(19), 4883.

C. Hope, P. B. Emmerich, A. Papadas, A. Pagenkopf, K. A. Matkowskyj, D. R. Van De Hey, S. N. Payne, L. Clipson, N. S. Callander, P. Hematti, S. Miyamoto, M. G. Johnson, D. A. Deming, F. Asimakopoulos, J. Immunol. 2017, 199(5), 1933.

A. Papadas, G. Deb, A. Cicala, A. Officer, C. Hope, A. Pagenkopf, E. Flietner, Z. T. Morrow, P. Emmerich, J. Wiesner, G. Arauz, V. Bansal, K. Esbona, C. M. Capitini, K. A. Matkowskyj, D. A. Deming, K. Politi, S. I. Abrams, O. Harismendy, F. Asimakopoulos, Cell Rep. 2022, 40(7), [eLocator: 111201].

M. J. Binder, S. McCoombe, E. D. Williams, D. R. McCulloch, A. C. Ward, Biomolecules 2020, 10(5), 682.

K. Suwan, K. Choocheep, S. Hatano, P. Kongtawelert, K. Kimata, H. Watanabe, J. Biol. Chem. 2009, 284(13), 8596.

M. J. Merrilees, N. Zuo, S. P. Evanko, A. J. Day, T. N. Wight, J. Histochem. Cytochem. 2016, 64, 353.

Y. Murasawa, H. Nakamura, K. Watanabe, H. Kanoh, E. Koyama, S. Fujii, K. Kimata, M. Zako, M. Yoneda, Z. Isogai, Am. J. Pathol. 2018, 188(2), 432.

P. Y. Akinfenwa, W. S. Bond, C. J. Ildefonso, M. Y. Hurwitz, R. L. Hurwitz, J. Biol. Chem. 2017, 292(35), [eLocator: 14381].

M. Huynh, C. Pak, S. Markovina, N. S. Callander, K. S. Chng, S. M. Wuerzberger‐Davis, D. D. Bakshi, J. A. Kink, P. Hematti, C. Hope, F. Asimakopoulos, L. Rui, S. Miyamoto, J. Biol. Chem. 2018, 293(7), 2452.

M. Vistnes, P. M. Erusappan, A. Sasi, E. S. Nordén, K. K. Bergo, A. Romaine, I. G. Lunde, L. Zhang, M. B. Olsen, J. Øgaard, C. R. Carlson, C. H. Wang, J. Riise, Christen Peder Dahl, A. E. Fiane, I. M. Hauge‐Iversen, E. Espe, A. O. Melleby, T. Tønnessen, J. M. Aronsen, I. Sjaastad, G. Christensen, Cardiovasc. Res. 2023, 119(10), 1915.