3.1.1 Angiogenesis is a critical component of cancer initiation and progression
In tumorigenesis, early activation of angiogenic processes is mandatory to sustain the higher energetic demands associated with the enhanced proliferation of the tumor cells and the associated tumor tissue growth. As stated above, new blood vessel formation occurs when pro-angiogenic signals surpass anti-angiogenic signals, and in tumors, this process is termed the “angiogenic switch” [
13]. The pioneering work of Douglas Hanahan and Judah Folkman using the
Rip-
Tag transgenic mice overexpressing the SV40 large T antigen in pancreatic β-cells established the role of the angiogenic switch as a rate-limiting factor sustaining tumor growth and progression [
1,
68,
87]. The angiogenic switch promotes the growth of the malignant cells and allows the escape of tumors from dormancy. Thanks to Judah Folkman’s revolutionary hypothesis that blocking angiogenesis might stop tumor growth, two decades ago, anti-angiogenic therapy became a new modality in cancer treatment in addition to surgery, radio- and chemo-therapy.
Tumor angiogenesis is typically initiated from the capillaries and includes distinct cellular processes, including sprouting angiogenesis, intussusceptive angiogenesis, vasculogenesis, and transdifferentiation of cancer stem cells (reviewed in [
149]). Mechanistically, activation of angiogenic processes involves the degradation of the vascular ECM, followed by endothelial cell proliferation and migration [
263]. ECM remodeling is essential for the migration of endothelial cells and the formation of capillary sprouts, a process that requires the activation of matrix metalloproteases (MMPs), which degrade the basal membrane and the ECM. The
Rip-
Tag mouse model was instrumental in establishing a role for VEGF in the activation of ECM-degrading enzymes, thereby supporting the angiogenic switch [
19,
112,
176] (Fig.
2).
Tumor angiogenesis results in the formation of poorly organized and malformed vascular networks, characterized by a loss of endothelial cell junctions. This increases both vessel permeabilization and interstitial pressure, which reduce tumor tissue perfusion. Hypoxia, the condition of low oxygen in tissues, is typically associated with the progression of solid tumors since with the expansion of the tumor mass, the distance between tumor cells and vessels increases, leading to local regions of poor oxygenation. Hypoxia is also a plausible cause of obesity-associated tissue inflammation (Fig.
2). Hypoxia is also a consequence of the malformed and dysfunctional vasculature and the low perfusion of the tumors. To overcome the limited oxygen and nutrient supply, hypoxic tumor cells upregulate the hypoxia-inducible factors, which then initiate pro-angiogenic programs by upregulating the transcription of downstream targets involved in angiogenesis, cell survival, and proliferation [
197].
Noteworthy, in addition to serving as nutrient, oxygen, and waste transport providers, vessels also facilitate the dissemination of tumor cells to distant sites, thereby promoting metastasis. The development of metastases is the major cause of cancer morbidity and mortality and accounts for the vast majority (up to 90%) of cancer deaths [
135].
Hypoxia-mediated HIF activation increases the expression of
VEGFA and
ANGPTL4, among others. VEGFA is a member of a family of growth factors comprising also VEGFB, VEGFC, VEGFD, and PGF (placental growth factor), and functions as the main endothelial cell survival factor. It binds to two tyrosine kinase receptors, i.e., VEGF receptor 1 (VEGFR1) and receptor 2 (VEGFR2) and prompts their homo- and heterodimerization, transphosphorylation, and the activation of downstream signaling. VEGFA is primarily secreted by tumor and stromal cells in the TME and acts chiefly on endothelial cells by interacting with VEGFR2, the principal mediator of the cellular responses to the growth factor. VEGFR2 activation promotes endothelial cell proliferation, survival, and migration via the stimulation of PLCγ-PKC, MAPK-ERK1/2, and PI3K-AKT pathways [
122,
232,
233,
244]. The function of VEGFR1 is less defined: it has been shown that a soluble version of VEGFR1 (sVEGFR1) is secreted by endothelial cells and can act as an endogenous decoy for VEGFA by sequestering it and blocking its access to VEGF receptors [
105]. VEGFA also induces vascular permeability by activating mechanisms such as the phosphorylation of vascular-endothelial (VE)-cadherin and β-catenin, which cause the destabilization of endothelial cell–cell contacts and the opening of endothelial cell junctions [
10]. Interestingly, adipocytes also produce VEGF and other pro-angiogenic factors (Fig.
2).
Given its central role in angiogenesis, VEGF signaling is upregulated in a variety of human cancers. The expression of VEGFR2 is elevated in HCC patients [
8], as well as the levels of circulating VEGFA, which correlate with tumor angiogenesis, rapid disease progression, and decreased survival [
209]. Based on these observations, therapies targeting the VEGFA pathway have been first evaluated in preclinical models of HCC, and then implemented in clinics with promising results [
66,
143]. Sorafenib, a multikinase inhibitor targeting VEGFR, PDGFR, c-Kit, and RAF, has been the standard of care for patients with advanced, unresectable HCC for over a decade given its ability to increase patients’ survival [
144]. Lenvatinib, another anti-angiogenic drug inhibiting VEGFR, fibroblast growth factor (FGF), and PDGFR with higher potency than sorafenib, is also used for unresectable HCC and showed an improvement in overall survival of the patients which was comparable to sorafenib [
131]. Additionally, bevacizumab (a humanized monoclonal antibody that sequesters VEGFA) in combination with atezolizumab (immune checkpoint inhibitor) was approved by the FDA in 2020 for the treatment of patients with unresectable or metastatic HCC. Currently, this combination therapy is the first-line systemic therapy for such patients [
67].
While HCC is a highly vascular tumor, with angiogenesis playing an important role in its growth and dissemination, PDAC exhibits poor vasculature, low blood flow, and reduced perfusion compared with normal pancreas. The resulting hypoxic environment further exacerbates pathological changes, such as the development of stromal fibrosis, a hallmark of PDAC. Tissue hypoxia promotes the release of various pro-angiogenic factors including VEGFA [
261]. Accordingly, it was initially reported that PDAC tissues show an increase in
VEGFA gene expression when compared to the normal human pancreas and that 60–65% of human PDAC samples show detectable VEGFA immunoreactivity[
100,
104,
214]. However, recent RNA-Seq data (The Cancer Genome Atlas-TCGA dataset) challenged these findings given that only 8 out of 178 (4%) human PDAC samples were found to overexpress
VEGFA, thereby suggesting that it may not be as relevant a player in PDAC tumorigenesis as firstly assumed [
243]. Several genetically engineered mouse models of PDAC that successfully recapitulated the histological and molecular evolution of the tumors have been generated and proven to be instrumental in assessing the efficacy of drugs targeting the TME and angiogenesis. Among the tested drugs is sunitinib, a multi-kinase inhibitor targeting VEGFR and PDGFR signaling [
178], which could not reduce tumor burden in PDAC mouse models. Clinical trials mirrored these results by showing a lack of efficacy of sunitinib when combined with gemcitabine for treating patients with advanced/metastatic PDAC[
20], or when used as second-line therapy after failure to respond to gemcitabine [
177]. Two additional anti-angiogenic drugs, bevacizumab and axitinib, also failed to improve the survival of PDAC patients [
137]. Altogether, targeting VEGF signaling does not seem to represent a valuable strategy for the treatment of PDAC.
The interaction of VEGFA and VEGF receptors (VEGFR) and the resulting angiogenesis have been heavily implicated in breast cancer development, progression, and metastasis. High VEGFA and VEGFR expression in breast cancer patients correlates with worse outcomes and resistance to systemic therapy. Drugs with anti-angiogenic effects have shown beneficial effects in breast cancer patients, starting from tamoxifen, the mainstay adjuvant therapy for hormone-positive breast cancer. Originally believed to be a mere competitor of estradiol, tamoxifen was later found to inhibit VEGFA and angiogenesis [
24,
72]. VEGFA is upregulated in breast cancers overexpressing the receptor tyrosine kinase human epidermal growth factor 2 (HER2). For these patients, the standard-of-care therapy is the anti-HER2 antibody trastuzumab alone or in combination with chemotherapy. Trastuzumab was shown to inhibit angiogenesis and normalize the tumor vasculature [
107].
Another important mediator of angiogenesis is ANGPTL4, a secreted factor belonging to a superfamily of proteins structurally related to angiopoietins, which are growth factors binding the receptor tyrosine kinase Tie2 on endothelial cells and regulating vasculogenesis, vessel homeostasis, and vascular remodeling [
107]. Unlike angiopoietins, however, ANGPTLs do not bind to either the Tie2 receptor, or the related protein Tie1, and are therefore considered orphan ligands. ANGPTL4 is considered a member of a new class of proteins named matricellular proteins, which are nonstructural glycoproteins secreted by cancer cells and neighboring stromal cells into the TME, where they associate with the ECM. ANGPTL4 was discovered independently by three groups in the year 2000 as a fasting-induced factor, prevalently expressed in the liver and in the adipose tissue [
119]. We now know that ANGPTL4 is a multifaceted protein involved in several metabolic and non-metabolic conditions, in both physiological and pathological situations, including angiogenesis, vascular permeability, tumorigenesis, lipid metabolism, glucose homeostasis, wound healing, and inflammation, among others.
A C-terminal circulating ANGPTL4 fragment (cANGPTL4) binds to ECM proteins and integrins, and this was originally shown to facilitate wound healing [
75,
94]. Later, it was found that the binding of cANGPTL4 to integrins β1 and β5 and their subsequent activation regulates cell migration via the focal adhesion kinase (FAK)/p21-activated kinase (PAK)–signaling cascade [
25]. Recently, it was demonstrated that cANGPTL4, via the activation of integrin α5β1, increases vascular leakiness by binding to VE-cadherin and claudin-5 and disrupting their intercellular clusters [
97]. Furthermore, cANGPTL4 can also associate with specific ECM proteins and delay their proteolytic degradation by MMPs [
97]. Thus, ANGPTL4 expression disrupts vascular endothelial tight junctions, augments vessel permeability, and alters trans-endothelial barriers [
180], ultimately facilitating tumor cell motility and the formation of metastases. Accordingly, ANGPTL4 was reported to promote venous invasion and distant spread in colorectal and renal cell cancer, as well as in gastric and breast cancer [
262]. However, not in every tumor type, ANGPTL4 promotes tumor progression. Indeed, ANGPTL4 was found to prevent lung carcinoma and melanoma metastases via the inhibition of vascular permeability, tumor cell motility, and invasiveness [
70]. These contradicting findings may be caused by the different functions of the cleaved forms of ANGPTL4 (N‐or C‐terminus), and highlight the dual role of ANGPTL4 in tumorigenesis.
In HCC and in chronic hepatitis patients, the amount of circulating ANGPTL4 at both mRNA and protein levels is significantly elevated when compared to control individuals [
63]. Similarly, in patients with alcoholic liver cirrhosis, serum levels of ANGPTL4 are increased
versus healthy controls [
192]. While these data are suggestive of an oncogenic role of ANGPTL4, other studies reported that both the levels of
ANGPT4 mRNA and the copy number of the gene are lower in HCC samples than in non-tumor tissues of the same patients. A possible mechanism for
ANGPTL4 downregulation in tumors is increased methylation at CpG sites located in the gene promoter [
173]. Lower expression levels of
ANGPTL4 mRNA are significantly associated with advanced tumor stage, poor differentiation, tumor recurrence, and decreased post-operative overall and disease-free survival of HCC patients, thereby pointing to a tumor suppressive role of ANGPTL4 [
173]. Studies in animal models helped elucidate the role of this protein in liver tumorigenesis. Indeed, the injection of ANGPTL4-overexpressing adenoviral vectors via the portal vein in mice bearing orthotopic liver cancer xenografts resulted in the suppression of both tumor growth and metastasis formation [
173]. These findings further supported the inhibitory role of ANGPTL4 in HCC development.
Mutation of the KRAS oncogene is a driver event in PDAC initiation and lineage tracing studies in mice showed that introducing a KRAS
G12D activating mutation in acinar/centroacinar cells promotes their differentiation in ductal-like cells and the formation of ADM/PanIN lesions [
80]. It was later found that Angptl4 overexpression in these mice increased the number of ADM lesions. Additionally, upregulation of Angptl4 enhanced tumor growth in a xenograft model of PDAC (Panc-1) cells [
252]. These findings point to a role for ANGPTL4 in promoting the initiation and progression of pancreatic tumorigenesis.
In patients, four molecularly-defined PDAC subtypes have been identified by integrated genomic analyses, one of them being the squamous subtype associated with mutations in TP53 and KDM6A [
14].
ANGPTL4 expression was the highest in tumors belonging to this subtype, which is associated with a particularly poor prognosis. In an independent study, transcriptome analysis showed that low expression levels of the
ANGPTL4 gene are associated with longer post-operative survival of the patients [
125]. Conversely, high
ANGPTL4 expression was observed in patients with shorter survival, as well as in PDAC cell lines resistant to gemcitabine, the standard first-line treatment for advanced or metastatic PDAC [
125]. Functionally, knockdown of
ANGPTL4 in a gemcitabine-resistant PDAC line (Panc-1 cells) led to a significant reduction in cell proliferation [
125]. Based on these results, ANGPTL4 emerges as a promising candidate target for tumors resistant to gemcitabine, and a potential marker for patients’ stratification in view of this treatment strategy.
Overexpression of
ANGPTL4 is associated with lower disease‐free survival in young breast cancer patients [
111]. Moreover, in circulating tumor cells from women with breast cancer,
ANGPTL4 copy number gain was found to be part of a signature of tumor aggressiveness and increased metastatic potential [
116].
Triple-negative breast cancer (TNBC) is an aggressive subtype having a dismal prognosis due to its propensity to metastasize to the brain or liver [
69]. ANGPTL4 is upregulated in primary tumors, in serum, and in metastases of TNBC patients [
26,
96,
97,
161,
180,
216,
256]. Ectopic ANGPTL4 overexpression in a TNBC cell line promoted the formation of 3D mammosphere cultures in vitro and led to the development of larger primary tumors, and to more liver and brain metastases in xenograft models in vivo when compared to cells with endogenous ANGPTL4 levels [
220]
.
Mechanistically, it was shown that soluble ANGPTL4 secreted by metastatic TNBC cells disrupts the integrity of endothelial cell junctions in capillaries within the lungs and brain, thereby allowing tumor cells to access and seed in the respective parenchyma [
26,
76,
180].
In conclusion, VEGF signaling promotes disease progression and aggressive behavior in HCC and breast cancer, and consequently, its inhibition has significant anti-tumor effects. In contrast, it does not seem to play a role in PDAC, and this is compatible with PDAC being intrinsically a hypovascular cancer. ANGPTL4 plays a tumor suppressive role in HCC, whereas in PDAC and breast cancer it promotes aggressiveness thanks to its effects on endothelial integrity and cellular migration. The different properties of full-length and truncated ANGPTL4 variants and their tissue distribution likely explain the different phenotypes associated with its upregulation in tumors. Targeting ANGPTL4 is considered a promising strategy to reduce tumor growth in PDAC and metastases formation in breast cancer (Fig.
2).