Abstract
Cancer-associated fibroblasts conduct an aberrant wound-healing response, including mechanisms that restrain and others that support tumor progression. In this issue of Cancer Discovery, Francescone and colleagues demonstrate expression of presynaptic protein NetG1 on fibroblasts in pancreatic ductal adenocarcinoma and characterize tumor-supportive functions of NetG1 in this context, including metabolic and immune-modulatory mechanisms.
See related article by Francescone et al., p. 446.
Despite substantial advances in precision medicine (1, 2), pancreatic ductal adenocarcinoma (PDAC) remains near-uniformly lethal. Generating an immunologic milieu permissive to antitumor immunity is perhaps the ultimate goal in the broad pursuit of effective treatments for this highly aggressive cancer. PDAC exhibits substantial heterogeneity with respect to composition of the immune microenvironment, yet although some tumors harbor high levels of infiltrating cytotoxic leukocytes, multiple mechanisms enable immune suppression. Cancer-associated fibroblasts (CAF) are abundant in the PDAC microenvironment and have been shown to suppress antitumor immunity (3). Understanding immune-modulatory pathways stemming from CAFs may aid the identification of new combination therapeutic strategies that foster immune-mediated PDAC cell killing. In light of increasing evidence that the metabolic state of the tumor microenvironment profoundly influences immune cell function and the potential for antitumor immune responses, the capacity for CAFs to regulate metabolite levels in the tumor microenvironment and the metabolic state of tumor cells (4) may also point to mechanisms by which CAFs modulate intratumoral immune cell functions, though these metabolic and immune-modulatory roles for CAFs have not been linked previously to a common pathway or mechanism. Francescone and colleagues set out to identify features and functions unique to PDAC CAFs (5), and identified one such mechanism in the process.
As the cell, or cells, of origin for pancreatic CAFs remain unclear, the pathways and transcriptional programs that distinguish CAFs from their non–tumor-associated counterparts are largely unknown. To gain insights into fibroblast functions specific to or enhanced among CAFs, the authors compared transcriptional programs in PDAC patient-derived CAFs compared with fibroblasts from tumor-adjacent regions in a matched tissue specimen. For this, fibroblasts were isolated from freshly resected surgical samples and grown in a three-dimensional culture system featuring cell-derived extracellular matrix components. The authors have previously demonstrated that this culture system enables maintenance of fibroblast features observed in vivo which are altered or lost in conventional, two-dimensional culture on an artificially stiff growth substrate (6). Transcriptional profiling of these patient-derived materials showed that the second most upregulated gene in CAFs is NTNG1, encoding the presynaptic, lipid-anchored protein Netrin G1 (NetG1). PDAC tissue staining confirmed elevated expression of NetG1 protein in the tumor stroma compared with adjacent nontumor regions, whereas the only known NetG1 ligand, NGL1, was detected in the epithelial compartment. Importantly, stromal fibroblast NetG1 expression correlated inversely with overall patient survival. This sets up an interesting contrast with a recently published study documenting a worse prognosis for patients with PDAC with lower stromal density (7), taking cellular and acellular stromal features into account. These findings suggest that, although bulk stromal content may be protective, specific stromal factors may promote tumor growth, such that stroma-targeted therapeutic approaches should be aimed at specific, protumorigenic mechanisms. Although NetG1 has documented roles in neurons regulating axon guidance and stabilization of excitatory glutamatergic synapses, it has been subject to very little study in cancer, with prior studies limited to biomarker analyses. As such, its functional role in PDAC or on CAFs was entirely unknown.
To test the significance of NetG1 in PDAC progression, the authors generated several loss-of-function systems to disrupt the NetG1/NGL1 axis in vivo using orthotopic transplant models. This included use of a neutralizing monoclonal antibody against NetG1, as well as Ngl1 knockout PDAC cells and NetG1 knockout CAFs to be used in cotransplantation assays. Antibody-mediated or genetic disruption of this axis significantly reduced PDAC progression across models, motivating efforts to understand the tumor-promoting mechanisms relevant to NetG1 function. To begin to understand the significance of NetG1 expression for tumor–stroma cross-talk, the authors performed coculture experiments and found that loss of NetG1 expression in CAFs profoundly reduced heterotypic cell–cell interaction between CAFs and PDAC cells, and at the same time increased PDAC cell motility. Interestingly, loss of NGL1 in PDAC cells did not phenocopy these cell adhesion results, raising the possibility that NetG1 associates with an additional and yet unknown ligand on PDAC cells. Probing further, the authors found that efficient transfer of intracellular content directly from CAFs to PDAC cells, using GFP as a marker, required NetG1 and NGL1. When intracellular content transfer was tested using conditioned media independent of cell–cell contact, NetG1 was dispensable but NGL1 was required, and PDAC cell NGL1 was implicated in regulation of macropinocytosis.
The significance of NetG1/NGL1 for direct and indirect CAF–PDAC cell interaction led the authors to address a role for this axis in paracrine metabolite transfer. CAFs provide nutritional support to PDAC cells in multiple forms and via multiple mechanisms (8), which is thought to promote PDAC cell survival within a nutrient-poor tumor microenvironment. In assays featuring cell–cell contact, secretome transfer via conditioned media, or CAF-derived extracellular matrix, the NetG1/NGL1 axis promoted PDAC cell survival under culture conditions of nutrient challenge. Though these results suggest that NetG1 regulates several CAF functions relevant to paracrine metabolic support, the authors specifically addressed secretion of amino acids as a stromal mechanism to promote PDAC cell viability under nutrient limitation. They found that NetG1 profoundly affects glutamate and glutamine secretion by CAFs. Although these amino acids were shown to support PDAC cell viability in this study, they may also link stromal NetG1 to regulation of additional cell types in the tumor microenvironment sensitive to glutamate and glutamine levels.
In further characterizing stromal NetG1 function, the authors took note of prior publications documenting cytokine production by CAFs and their potential for immune suppression (3, 9), and tested whether NetG1 regulates CAF immune-modulatory potential with a focus on cytokine and growth factor secretion. NetG1 loss results in a striking reduction in the production of GM-CSF and CCL20, with more modest but significant reductions in additional cytokines. IL15, a key activator of NK cells, was independent of NetG1; maintained IL15 expression in the context of secretome alterations with NetG1 loss unleashed potent NK cell activity and PDAC cell killing in vitro in the presence of NetG1 knockout CAFs compared with control CAFs. Though NK cells were tested directly, the broad regulation of stromal cytokine production suggests that NetG1 may promote immune suppression via regulation of additional immune cell types.
Mechanistically, the authors found that p38 signaling downstream of NetG1 regulates the immunosuppressive phenotypes observed, whereas AKT signaling regulates the metabolic functions reported, though additional pathways surely contribute to these complex phenotypes. Additional mechanistic studies implicated FRA1 and 4E-BP1 as functionally relevant effectors of p38 and AKT, respectively. In addition, as vesicular glutamate transporter 1 (VGlut1) and glutamine synthetase were downregulated in the NetG1 knockout setting, the significance of these factors in CAFs was tested. VGlut1 has a well-appreciated role in loading presynaptic vesicles with glutamate to assure appropriate synaptic glutamate secretion in the context of neuronal communication. Thus, roles for VGlut1 and glutamine synthetase, an enzyme with a critical role in the metabolism of nitrogen and production of glutamine, were expected with respect to metabolic functions downstream of NetG1 in CAFs. However, both VGlut1 and glutamine synthetase coregulated aspects of metabolic as well as immune-modulatory CAF functions regulated by NetG1, suggesting a novel signaling circuit in the PDAC stroma with diverse functional roles including those described here and others yet to be identified (Fig. 1).
Model of the diverse functions of NetG1 on pancreatic CAFs. Stromal NetG1 supports pancreatic tumor progression in part by providing metabolic support to enable pancreatic cancer cell survival under relevant conditions of nutrient challenge. This metabolic support includes secretion of amino acids including glutamine and glutamate, which can be taken up by pancreatic cancer cells to support their viability when nutrient levels are low. Mechanistically, NetG1 supports PDAC cell metabolism at least in part by signaling through AKT and 4EBP1, and by regulating expression of glutamine synthetase (GS) and perhaps other metabolic enzymes. In addition, the NetG1 ligand NGL1 on PDAC cells positively regulates micropinocytosis, which is known to promote PDAC cell metabolism and tumor growth. NetG1 also promotes immune suppression by regulating production of numerous immune-suppressive cytokines. In addition to these immune-suppressive factors, PDAC CAFs secrete the NK cell–activating cytokine IL15, which is independent of NetG1, such that NK cell capacity to kill PDAC cells is unleashed in the presence of CAFs lacking NetG1.
Model of the diverse functions of NetG1 on pancreatic CAFs. Stromal NetG1 supports pancreatic tumor progression in part by providing metabolic support to enable pancreatic cancer cell survival under relevant conditions of nutrient challenge. This metabolic support includes secretion of amino acids including glutamine and glutamate, which can be taken up by pancreatic cancer cells to support their viability when nutrient levels are low. Mechanistically, NetG1 supports PDAC cell metabolism at least in part by signaling through AKT and 4EBP1, and by regulating expression of glutamine synthetase (GS) and perhaps other metabolic enzymes. In addition, the NetG1 ligand NGL1 on PDAC cells positively regulates micropinocytosis, which is known to promote PDAC cell metabolism and tumor growth. NetG1 also promotes immune suppression by regulating production of numerous immune-suppressive cytokines. In addition to these immune-suppressive factors, PDAC CAFs secrete the NK cell–activating cytokine IL15, which is independent of NetG1, such that NK cell capacity to kill PDAC cells is unleashed in the presence of CAFs lacking NetG1.
This study welcomes numerous new lines of investigation in the area of pancreatic tumor–stroma cross-talk. That NGL1 was dispensable for the heterotypic interaction between CAFs and PDAC cells whereas NetG1 was required for extensive interaction suggests that PDAC cells may express additional NetG1 ligands, perhaps specific to pancreas tissue or to the cancer context. Further, the implication of NGL1 in regulation of macropinocytosis warrants additional mechanistic investigation in light of the significance of this mechanism in metabolic support of RAS-driven cancers and the previously unappreciated role of NGL1 in this process. The relevance of stromal NetG1 in suppression of NK cell activity gains significance in light of recent studies documenting a substantial NK cell population in human PDAC (9, 10). Although mechanisms regulating NK cell recruitment into the PDAC microenvironment remain unknown, therapeutic strategies that increase NK cell abundance may show efficacy together with inhibition of NetG1 to promote NK cell antitumor activity as a novel immune-modulatory treatment approach in this functionally immune-suppressed cancer. In addition, the recent identification of PDAC CAF subtypes raises the possibility that CAFs exhibit heterogeneity with respect to NetG1 expression, perhaps with preferential NetG1 expression on a specific CAF subtype. Though intratumoral CAF heterogeneity has been documented (4), the extent of intertumoral heterogeneity remains unclear. The dominant stromal immune suppression mechanism may indeed vary from tumor to tumor, such that PDAC with relatively low levels of CAF NetG1 expression may be subject to suppression of NK cell function by other mechanisms yet to be discovered. Finally, this work by Francescone and colleagues inspires investigation of stromal NetG1 in other cancer types to probe both functional significance and therapeutic opportunity.
Author's Disclosures
No disclosures were reported.