Introduction

Mammalian heparanase (HPA) is the only endoglycosidase that degrades heparan sulfate, a vital element of the extracellular matrix (ECM) and tumor microenvironment [1, 2]. HPA activity is strongly implicated in the metastatic potential of tumor cells, a result of heparan sulfate degradation and remodeling of ECM barriers [3, 4]. Importantly, Epithelial–mesenchymal transition (EMT) is a key biological process for HPA-derived cells to acquire the ability of migration and invasion. Previous studies have reported that HPA modulates the activity of factors such as FGF-2 and TGF-β-induced EMT [5, 6]. Besides, EMT makers were also found higher in gastric signet ring cell adenocarcinoma with higher HPA expression [7]. Similarly, HPA activity contributes to angiogenesis, inflammation, and autoimmunity [8,9,10]. Evidence accumulating over the past two decades, indicates that highly expressed HPA exists in the majority of primary tumors, including carcinomas, sarcomas, and hematological malignancies, correlates with larger tumor size, higher microvessel density, and worse prognosis [11, 12]. Apart from its roles as an enzymatic element, heparanase also fulfils enzymatic-independent biological functions, including signal transduction [13, 14] and gene transcription [15, 16]. Activation of serine/threonine kinase (AKT) is regarded as a prominent nonenzymatic signaling event exerted by HPA [17], and its inhibition attenuates cell growth [18]. HPA can also promote cell proliferation by enhancing the phosphorylation of epidermal growth factor receptor, independent of its enzymatic activity [19, 20].

Autophagy is an intracellular catabolic process that degrades cytoplasmic macromolecules and subcellular organelles, essential for cell survival and homeostasis [21]. Autophagy is a dynamic recycling system, consisting of the formation of autophagosomes, fusion of autophagosome–lysosome, and the degradation of autolysosomes [22, 55,56]. To further analyzed the expression of HPA and TFEB in GC tissues, we used a GC tissue microarray containing 15 pairs of primary GC tissues and adjacent nontumor tissues. Immunohistochemistry was applied to detect HPA and TFEB in human specimens derived from fifteen paired patients with gastric cancer. As shown in Fig. 8a, b, expression of HPA and TFEB was markedly increased in primary GC tissues compared to their adjacent normal-looking tissues. We further analyzed TFEB and HPA in gastric tumors using the TCGA database. Expression of HPA was significantly increased in gastric cancer (Fig. 8c and Supplementary Data 1). Moreover, clinicopathological analysis revealed that the patients with poorly differentiated tumors (Fig. 8d and Supplementary Data 2) and advanced tumor stage (Fig. 8e and Supplementary Data 3) exhibited significantly higher HPA expression. We next applied the Kaplan–Meier analysis and explored the relationship between TFEB expression and the GC patient outcome. The data revealed that TFEB elevation is significantly correlated with poor prognosis of in patients wih GC (OS HR = 1.61, log-rank P = 8.5e-06; FP HP = 1.44, log-rank P = 0.0011; PPS HP = 1.68, log-rank P = 4.9e-06) (Fig. 8f–h). Taken together, these results indicate that increased expression of HPA and TFEB in gastric cancer is correlated with advanced tumor stage and poor prognosis.

Fig. 8: Aberrant expression of HPA and TFEB in gastric cancer.
figure 8

Immunohistochemical analysis of HPA (a) and TFEB (b) expression in gastric tissues. c Relative expression level of HPA in normal gastric tissue and gastric primary tumors from UALCAN database. d Relative expression of HPA in gastric cancer based on tumor grade from UALCAN database. e expression of HPA in gastric cancer based on lymph nodal metastasis status from UALCAN database. Kaplan–Meier analysis from TCGA database of the correlation between relative expression of TFEB and overall survival (OS) (f), first progression (FP) (g), or post-progression survival (h). *P < 0.05, **P < 0.01 versus the adjacent or normal group. STAD stomach adenocarcinoma.

Full size image

Discussion

Mammalian cells express a single functional endoglycosidase (heparanase, HPA) which degrades heparan sulfate (HS). Cleavage of HS by HPA contributes to the disassembly of the ECM, thereby facilitating cancer metastasis and inflammation. Compelling evidence strongly implies that heparanase serve crucial roles in all aspects of the tumorigenic process, namely tumor initiation, growth, chemotherapy resistance, and metastasis [1,2,3,4,5,6,7,8,9,10,11,12,13], pointing to the potential of heparanase as an anti-cancer drug target and incentivizing the development of small molecular inhibitors of heparanase [14, 15]. A critical venue by which heparanase exerts multiple effects on its target tissues and cells is by modulating the bioavailability of HS-bound cytokines, chemokines, and growth factors, initiating the tissue microenvironment. Such architecture allows heparanase-mediated tumor-host crosstalk and promotes several basic cellular processes (i.e., exosome formation, autophagy, immune-inflammatory responses) that together coordinate tissue remodeling [16, 17]. Heparanase functions as an “activator” of HS proteoglycans and therefore is a pivotal player in creating a supportive environment for tumor cell proliferation and dissemination. In addition, it is required for various cellular processes from gene expression regulation to signal transduction and DNA damage signaling [3, 18]. In addition to its enzymatic activity, nonenzymatic functions of HPA have been reported and are well-documented [57, 58]. For example, HPA could augment cell proliferation, mobility, and angiogenesis through activation of β1 integrin, HIF-2α, Flk-1, and/or AKT signaling, independent of its enzymatic activity [17, 18, 59]. In addition, HPA has been documented to impact on the blood clotting system, independent of its catalytic function [60, 61]. Altogether, these and other results indicate that nonenzymatic activities of HPA play essential roles in a variety of pathological processes. Here, we demonstrate that, among other effects, overexpression of enzymatically inactive heparanase promotes GC cell proliferation and autophagy.

Autophagy is a major cellular degradation and recycling process, critical for maintaining cellular homeostasis [62, 63]. Autophagy is a dynamic process that sequesters misfolded and/or potentially dangerous proteins and damaged organelles in double-membranes vesicles called autophagosomes, that are ultimately degraded within lysosomes [64, 65]. The process of autophagy is schematically subdivided into three critical steps: autophagosome formation and maturation, autophagosome–lysosome fusion, and auto-lysosomal acidification [66]. Briefly, the activation of ULK1 and BECN1-VPS34 complexes leads to the formation of autophagosomes [65, 67] that are then fused with lysosomes to produce autolysosomes, followed by intraluminal acidification and subsequent activation of lysosomal hydrolases to mediate autophagic cargo degradation [68, 69]. Given its major role in cellular metabolism, autophagy is connected with numerous disease states [24, 41], yet the role of autophagy in cancer is disputable and context-dependent [24, 26, 48, 70,71,72]. It has previously been shown that HPA promotes the growth of head and neck carcinoma by enhancing autophagy [32]. In the present study, we found that both active or enzymatically inactive HPA increase autophagosome formation and the expression of related genes in gastric cancer cells. We show that nonenzymatic HPA-induced LC3-II protein expression and cell viability were attenuated by 3MA and CQ, compounds that inhibit autophagosome formation and disrupt autophagosome–lysosome fusion, respectively. We next assessed the biogenesis and formation of lysosomes to further ascertain the role of autophagy in nonenzymatic HPA-induced GC cell proliferation. We observed that both wild-type and mutant HPA increased the expression of autophagy- and lysosomal-related genes as well as the level of the lysosomal membrane protein LAMP2. Fluorescence imaging showed that the formation of autophagosomes, autolysosomes, and lysosomes was increased after transfection with native or nonenzymatic HPA, indicating that nonenzymatic HPA may also trigger cell autophagy by inducing lysosomal biogenesis and formation of autolysosomes in gastric cancer cells.

The MiT/TFE transcription factors are composed of four members: TFEB, TFEC, and TFE3 and MITF. Among them, TFEB and TFE3 are master regulators of lysosomal and autophagosome biogenesis, which bind to a conserved 10-base palindromic sequences, named Coordinated Lysosomal Expression and Regulation (CELEAR) site that promotes the transcription of several lysosomal and autophagy genes [50, 73,74,75]. A variety of stimuli are related to TFEB and TFE3 activities, among which nutritional deprivation is well-characterized [52, 76, 77]. In the presence of nutrients, TFEB and TFE3 have been shown to reside in the cell cytoplasm, but are rapidly translocated to the nucleus in the absence of nutrients [76, 78]. Subsequently, molecules involved in nutrient sensing and cellular growth have been shown to regulate TFEB and TFE3 activity [43, 79, 80]. The most studied regulatory mechanism of TFEB and TFE3 subcellular localization is involved in modulating the phosphorylation status of multiple serine residues [52]. In particular, mTOR kinase was shown to phosphorylate TFEB and TFE3 and act as a regulatory in TFEB and TFE3 subcellular localization [51, 52]. Several serine residues in TFEB and TFE3 are phosphorylated by mTOR, but S211 in TFEB and S321 in TFE3 are particularly relevant [81,82,83,84]. In the presence of adequate nutrition, mTOR phosphorylates these residues to create a cytoplasmic chaperone 14-3-3 binding site. Interaction with 14-3-3 leads to the accumulation of TFEB/TFE3 in the cytoplasm. In contrast, when nutrition is insufficient, inactivation of mTORC1 dephosphorylates S211 and S321 and prevents binding to 14-3-3 leading to rapid TFEB and TFE3 nucleus translocation occurred. The mechanism underlying autophagy induction by HPA is not clear. Shteingauz et al reported that HPA-induced autophagy is mediated by mTOR signaling because HPA overexpression is associated with decreased mTOR activity, whereas HPA inhibition leads to increased mTOR activity and substrate phosphorylation [32]. Here, we report that the mRNA and protein levels of TFEB, but not TFE3, were significantly changed after transfection with either wild-type or nonenzymatic HPA. Moreover, active and mutant HPA promoted nuclear translocation of TFEB but had no significant effect on TFE3 subcellular localization. These data suggest that TFEB- but not TFE3-mediated autophagy is predominant involved in HPA-induced autophagy in GC cells, independent of HPA enzymatic activity. Additionally, upregulation of nonenzymatic HPA decreased phosphorylation of TFEB S211, and abolished the interaction between TFEB and 14-3-3 protein, further indicating that nonenzymatic HPA induces TFEB-mediated autophagy in TFEB dephosphorylation-dependent manner.

Aberrant activation of TFEB is closely related to tumor oncogenesis and development [51, 85, 86]. Cells rely on effective lysosomal function, and increasing evidence indicates that cancer cells may utilize TFEB-dependent transcriptional activation of lysosomal degradation pathways to maintain survival. Overexpression of TFEB was reported to drive pancreatic ductal adenocarcinoma by inducing autophagy [48, 87]. TFEB gene silencing results in a significant decrease in the hyperproliferative phenotype and progression of pancreatic carcinoma to advanced stages [31, 54]. Here, we report that TFEB gene silencing significantly attenuated hyperproliferation induced by nonenzymatic HPA via decreased autophagy and lysosome biogenesis. Moreover, upregulation of HPA in GC is significantly linked to advanced tumor stage, while increased TFEB expression in GC tissues is correlated with poor prognosis. These results further indicate that HPA induces GC progression through TFEB-mediated autophagy, independent of HPA enzymatic activity (Graphical abstract), offering a new strategy for the design of nonenzymatic HPA targeting drugs.

However, our results are mainly derived from gastric cancer cell lines, and it is necessary to repeatedly verify the regulatory relationship between HPA and TFEB in animal in vivo experiments. We will make further improvements in follow-up research.