1. Introduction
1.1. Introduction of Pancreatic Cancer
Pancreatic carcinoma is the deadliest malignancy afflicting the exocrine digestive organ. This cancer is well known for lacking screening tools and having early metastatic spread, followed by chemoresistance, leading to limited treatment strategies and poor prognostic outcomes
[1][2]. As such, it took 466,003 lives across 185 countries in 2020 and is presently the seventh leading cause of deaths from cancers in both genders
[3]. Trends forecasted through 2040 predict that pancreatic cancer will become the second-most-leading cause of cancer-related death in the United States
[4], and approximately 355,317 new cases will occur globally
[5]. Among them, nearly 95% of pancreatic cancer incidences are pancreatic ductal adenocarcinoma (PDAC)
[6]. Approximately 80% of pancreatic cancer patients present with advanced-to-late stages of nonresectable and disseminated disease
[7]. The two most common first-line chemotherapeutic regimens include blends of 5-fluorouracil, leucovorin, irinotecan, and oxaliplatin (FOLFIRINOX) and gemcitabine (GEM) plus Nab-Paclitaxel
[8]. However, therapeutic intervention scarcely improves overall prognosis, and the 5-year survival rate remains disappointing
[9][10].
PDAC develops sporadically but is largely due to the acquisition of constitutively active mutant Kirsten rat sarcoma (KRAS) derived from the most frequent driver mutations: G12R, G12V, and G12D, which comprise approximately 90% of occurrence
[11]. Among them, G12D accounts for about 40% of incidents
[11][12]. Initial progression to pancreatic cancer embarks from the cells harboring KRAS mutations engaging in networking with proinflammatory cytokines
[13][14][15]. For instance, in response to oncogenic mutant KRAS, interleukin (IL)-6 induces the expression and activation of signal transducer and activator of transcription 3 (STAT3)
[15][16][17][18]. Accordingly, persistent STAT3 signaling was demonstrated to play a pivotal role in mutant KRAS-induced pancreatic tumorigenesis
[19], and demonstrated that Janus kinase (JAK)–STAT3 axis activation correlates with a poor outcome in PDAC patients following surgical resections
[20]. Moreover, oncogenic mutant KRAS unleashes a plethora of signaling cascades, including rapidly accelerated fibrosarcoma (RAF)/mitogen-activated protein kinase (MEK)/extracellular signal-regulated kinase (ERK), and phosphoinositol 3-kinase (PI3K)/protein kinase B (AKT) pathways in various malignant entities including pancreatic cancer
[11][21][22][23]. RAF/MEK/ERK is the first well-known Ras effector in cancers. GTP-bound KRAS interacts with and triggers RAF, which further induces the phosphorylation and activation of MEK1 and MEK2. This scenario subsequently enhances ERK1 and ERK2 serine/threonine kinases activities. Activated ERK1/2 then phosphorylates over 200 targets, many of which are transcription factors controlling cell proliferation
[24][25]. Mounting evidence demonstrates the critical role of PI3K being a regulator for embarking oncogenic KRAS-driven carcinogenesis, largely by governing cell survival and proliferation
[26][27]. Another independent study utilizing a genetically engineered mouse model containing mutant
Kras elucidates a similar finding that the PI3K pathway can augment PDAC through the activation of STAT3 and nuclear factor kappa B (NF-κB) signaling
[28].
The first histological alteration occurring in PDAC pathogenesis is the transdifferentiation of acinar cells into duct-like cells, named acinar-to-ductal metaplasia (ADM)
[29][30]. The molecular causes underlying dysregulated ADM were recently elucidated to be associated with a loss of AT-rich interactive domain containing protein 1A (ARID1A)
[31], followed by interaction between PAF1 (RNA polymerase II-associated factor 1) and YAP1 (yes activated protein-1)
[32]. For ADM, infiltrating macrophages secrete inflammatory cytokines including regulated on activation normal T cell expressed and secreted (RANTES)
[33] and tumor necrosis factor-alpha (TNF-α). Together, they lead to the activation of NF-κB signaling and expression of matrix metalloproteinases (MMPs)
[33][34]. In response to chronic inflammation, acinar pancreatic cells adopt ADM
[29] and then develop precancerous lesions, which are not only frequently observed in pancreatitis
[35], but also develop into pancreatic intraepithelial neoplasia (PanIN) following the acquisition of oncogenic mutations such as KRAS
[29]. Both ADM and PanIN constitute crucial aberrations in PDAC and persist throughout tumor development
[34][36]. During this neoplastic progression, macrophage depletion not only blocks the progression of ADM to PanIN, but also lightens PDAC burden in mice
[34][37], underscoring the imperative role played by these immune cells.
Although oncogenic
Kras mutation in mouse PDAC is critical for cancer initiation, constitutively activated mutant KRAS alone is insufficient for tumor onset; rather, it requires partner mutations such as the
P53 tumor suppressor gene, as well as cytokines produced by different cell types within the tumor mass
[38]. A genetically engineered mouse model combining both mutations,
LSL-KrasG12D;
Trp53flox/flox;
Pdx-1-Cre (KPC), has been established as a clinically relevant PDAC model that recapitulates many key features of human PDAC with a robust inflammatory response
[39] and elevated immunosuppressive features
[40].
1.2. Introduction of Tumor Microenvironment and Immune Evasion
Marked by extensive fibrosis and inflammation, PDAC’s tumor microenvironment (TME) consists of fibroblasts, immune cells, endothelial cells, and an acellular extracellular matrix (ECM) that contains various growth factors, chemokines, and cytokines
[41]. Within the TME, cancer cells interplay with nearby stroma and acellular constituents that synergistically controls malignant traits and therapeutic outcomes
[42][43]. Fibroblastic stroma can hinder drug entry by safeguarding tumor cells from therapeutic insults
[44], and then advancing tumor progression characterized by invasion, angiogenesis, metastasis, and chemoresistance
[45]. PDAC is initially featured with chronic inflammation triggered by immune aberrations
[46]. Then, oncogenic mutant KRAS augments inflammation and launches an immunosuppressive TME that subsequently plays a pivotal role in cancer progression
[47][48][49][50].
In general, immune responses are modulated by a plethora of checkpoint regulators that act as “security brakes” and establish a “do not eat me” cue when inflammation reactions shall be ended from prior infections, or autoimmunity shall be circumvented by enhancing self-tolerance. Cancers exploit various immune checkpoint modulators, attempting to evade tumoricidal responses, favor immune tolerance, and escape recognition and clearance by immune surveillance cells
[51]. Therapeutic agents abolishing such functions are recognized as immune checkpoint blockades (ICBs) that have been proven to improve clinical outcomes
[52]. Yet, PDAC remains largely embraced by an immunosuppressive TME with limited infiltration of tumoricidal immune cells, thereby resulting in a poor response to ICBs
[53][54]. The TME attracts several immunosuppressive cell types that circumvent the surveillance normally conducted by cytotoxic cluster of differentiated (CD)8
+ T lymphocytes and by dendritic cells (DC)
[53].
Within the TME of PDAC, infiltration of tumoricidal CD8
+ T lymphocytes is rare. Accordingly, a few well-known ICBs attempting to revive T lymphocytes to date have manifested disappointing efficacy
[55]. Instead, the tumor bed is infiltrated with largely protumorigenic immune-suppressive cells including myeloid-derived suppressor cells (MDSC), regulatory T cells (Treg), and tumor-associated macrophages (TAM)
[47][53]. TAMs are the earliest infiltrating cells in PanIN lesions and continue to rise throughout cancer progression
[56]. Macrophages in PDAC are derived from blends of circulating monocytes and phagocytes that reside in the pancreas. Moreover, TAMs are the most abundant immune cells in the stroma and are the key drivers shaping the immunosuppressive landscape
[57]. TAMs enhance tumor immune evasion, mainly by enhancing tumor fibrosis and excluding tumoricidal T lymphocytes
[58]. TAM infiltration not only correlates with lymph node metastasis and poor prognosis
[59], but also plays multifaceted roles in the carcinogenesis of PDAC
[60].
As a vital innate immune population for maintaining body homeostasis and warding off foreign particles or pathogens, macrophages can regularly sense their microenvironment, display high plasticity, and execute diverse functions adapted to different environmental contexts. Depending on the inflammatory cues, macrophages can develop two distinct subtypes, these being either classically activated M1 or alternatively activated M2 subpopulations
[61]. M1 macrophages are proinflammatory and tumoricidal, whereas M2 macrophages are anti-inflammatory, protumorigenic, and immunosuppressive
[61][62]. Furthermore, fully polarized macrophages can depolarize and transform reciprocally in response to environmental triggers
[63]. The M1 subtype commonly produces higher levels of IL-1, IL-6, IL-12, IL-23, TNFα, chemokine C-X-C motif ligand (CXCL)9, CXCL10, and inducible nitric oxide synthase (iNOS)
[64]. Conversely, the M2-type commonly expresses higher levels of IL-10, transforming growth factor-β1 (TGF-β1), and arginase 1 (ARG1)
[65][66][67][68][69][70]. M2 is the most abundant immunosuppressive subpopulation representing approximately 85% of TAMs
[53][57][71][72]. Infiltration and the abundance of M2 is not only a malignant hallmark but also correlates with poor prognosis
[73][74]. Yang et al. demonstrated that targeting proliferating F4/80+ macrophages by the pharmacological inhibitor, clodronate liposomes, fostered CD8
+ T cell infiltration and promoted their spatial redistribution, thereby enhancing antitumor immunity
[75]. Furthermore, closer proximity of M2 macrophages to the tumor core strongly correlates with poor disease-free survival
[69], highlighting the clinical impact of M2 macrophages on molding a cancer-promoting landscape
[61][71][76].
Macrophages exist on a spectrum of polarization states between the M1 and M2 phenotypic extremes and exhibit functional plasticity within the TME
[77]. The early stages of tumor lesions initially have a high abundance of M1 macrophages that are later polarized to the M2 population as PDAC progresses
[78]. Preclinical and clinical trials have been completed, or are still ongoing, attempting to target TAMs and treat various cancer types including pancreatic cancer (e.g., NCT03662412, NCT03184870, and NCT01921699)
[79]. Although M2 macrophages are still under substantive studies, this aims to extrapolate PDAC-fostered M2 macrophages, delineate TME-orchestrated mechanisms responsible for M2 polarization, and then discern how the M2 population synergizes cancer cells and TME factors to convey multifaceted impacts on PDAC.
2. Factors Modulate Polarization of TAM
2.1. Factors Released from Malignant Cells or Cancer-Associated Fibroblasts (CAFs)
Crosstalk between neoplastic cells and infiltrating macrophages in the tumor milieu governs PDAC carcinogenesis. TAMs are in close contact with cancer-secreted factors and thereby are polarized towards the M2 phenotype
[37][80][81]. Intriguingly, oncogenic mutant KRAS can recruit TAMs and then promote carcinogenesis
[80]. Mutant KRAS not only releases growth factors but also regulates glucose metabolism in PDAC
[82]. Accordingly, lactate and granulocyte-macrophage colony-stimulating factor (GM-CSF) are known to be profoundly released from cancer cells expressing oncogenic mutant KRAS
[82][83] (
Figure 1). This aberration is mediated through the PI3K/AKT signaling cascade that partly enhances macrophage polarization
[84]. Moreover, regenerating gene family member 4 (REG4) released from PDAC cancer cells
[85][86] can promote macrophage polarization to M2, as well as orchestrate the TME to favor cancer growth and metastasis
[87] (
Figure 1). Consequently, high numbers of M2-polarized TAMs correlate with an increased incidence of lymph node metastasis
[87]. The underlying molecular mechanism accounting for this scenario was deemed to be mediated through the epidermal growth factor receptor (EGFR)/AKT/cAMP-response element binding protein (CREB) signaling pathway
[87]. A further study elucidated that the overexpression of recombinant REG4 enhanced the expression of IL-10, CD163, and many other M2 signature genes in TAMs
[87]. Additionally, the secretion of IL-10 can be upregulated by insulin-like growth factor binding protein 2 (IGFBP2) released from cancer cells following STAT3 activation
[88] (
Figure 1). IGFBP2 favors M2 macrophages and exacerbates an immunosuppressive TME by increasing Treg infiltration and inhibiting antitumor T cell immunity in a mouse model
[88]. Hence, blocking the IGFBP2 axis constitutes a promising treatment protocol through which TAM polarization can be attenuated and a tumoricidal state of the TME can be revived
[88]. Together, multiple networks maneuver TAM polarization toward an M2 state.
Figure 1. Pancreatic ductal adenocarcinoma cells synergize with the tumor microenvironment to provoke polarization of M2 macrophages. Arrows with pointed or with blocked ends indicate activation or inhibition between regulators, respectively, while a fading effect at the start of arrows represents secretion of modulators. Plain straight lines depict interaction between ligands and receptors. Cell surface proteins are noted in rectangular boxes on cell membranes. The circular lipid bilayer depicts an extracellular vesicle. Abbreviations used include aryl hydrocarbon receptor (AhR), protein kinases B (AKT), Bcl-2-associated athanogene 3 (BAG3), cancer-associated fibroblast (CAF), CC-chemokine ligand (CCL), cluster of differentiated (CD), CXC chemokine ligand (CXCL), colony-stimulating factor (CSF) and receptor (CSF1R), dendritic cell (DC), double cortin-like kinase 1 (Dclk1), endothelial cell (Endo), epidermal growth factor receptor (EGFR), epithelial cell (Epi), ezrin (EZR), galectin (gal), granulocyte-macrophage colony-stimulating factor (GM-CSF), hypoxia-inducible factor (HIF), interleukin (IL), insulin-like growth factor binding protein 2 (IGFBP2), interferon-induced transmembrane protein 2 (IFITM-2), interferon regulatory factor 4 (Irf4), Kirsten rat sarcoma (KRAS), mammalian target of rapamycin (mTOR), microRNA (miR), monocyte (M), nucleotide-binding and leucine-rich repeat receptor containing pyrin domain 3 (NLRP3), pancreatic ductal adenocarcinoma (PDAC), phosphatidylinositol 3-kinase (PI3K), reactive oxygen species (ROS), regenerating gene family member 4 (REG4), sialic-acid-binding immunoglobulin-like lectin 15 (SIGLEC15), signal transducer and activator of transcription (STAT), spleen tyrosine kinase (SYK), suppression of tumorigenicity 2 (ST2), tumor-associated macrophage (TAM), transforming growth factor β (TGF-β), and T helper-2 (TH2).
Double cortin-like kinase 1 (Dclk1) is overexpressed in the cancer cores and PanIN lesions, based off various pancreatic cancer models
[89] (
Figure 1). By releasing various chemokines and cytokines, the elevated Dclk1-isoform 2 resulted in the polarization towards the M2 phenotype (
Figure 1). This aberration is demarcated by a high abundance of M2 macrophages and low occupancy of CD8
+ T cell infiltration with weakened tumoricidal activities
[90]. These M2 macrophages enhance cell migration, invasion, and self-renewal, along with increased expression of Snail and Slug, both of which are indicatives of cancer stem-like cells
[90][91]. Moreover, galectin-9 (gal-9), a member of the P-galactoside-binding family of lectins, was found to be highly expressed in both mouse and human PDAC. The binding of gal-9 to its receptor, Dectin-1, a crucial innate immune regulator expressed on the surface of macrophages, polarizes macrophages to the M2 phenotype (
Figure 1). Disruption of the gal-9/dectin-1 interaction reverts immunosuppression, enhances cytotoxic T lymphocytes recruitment, downregulates Tregs, impedes tumor growth, and achieves improved therapeutic efficacy
[92][93][94]. Moreover, Ezrin (EZR) expression is upregulated in PDAC and is associated with tumor progression
[95]. Chang et al. demonstrated that extracellular vesicles (EVs)-capsulated EZR is strikingly correlated with poor survival in PDAC patients
[96]. Molecular investigations further discerned that overexpressed EZR regulates STAT3 activation
[97] that further synergizes with STAT6 to augment the polarization of TAMs towards the M2 phenotypes
[98] (
Figure 1). Consistently, Su et al. reported miR-155 and miR-125b2 as being the key regulator encapsulated in the PDAC cell-line-derived EV that exploits a dose-dependent effect on macrophage plasticity
[99].
On the other hand, CAFs release colony-stimulating factor (CSF) and induce M2 polarization through binding to receptor CSF1R within the PDAC milieu, and then enhance reactive oxygen species (ROS) production in monocytes
[100] (
Figure 1). The importance of ROS activation on M2 polarization was illustrated by the evidence that ROS ablation abrogates this effect
[101]. Anti-CSF1R therapy favors the M1-like subpopulation in vivo, thereby exerting a powerful antitumor effect on glioma neoplasm
[102]. Furthermore, stromal fibroblasts are the predominant cell types for producing IL-33 that mainly targets its receptor, known as suppression of tumorigenicity 2 (ST2), on TAMs and induces the polarization of M2
[103][104]. Upon activation, IL-33-polarized TAMs subsequently release CXCL3 to further amplify CAFs. Together, this interactive axis constitutes a paracrine and positive feedback loop amplifying both CAF and TAM cell types
[105] (
Figure 1).
2.2. Factors Produced from Stromal Immune Cells
Abundantly in PDAC, oncogenic mutant Kras can activate the downstream PI3K/AKT/mammalian target of the rapamycin (mTOR) signaling pathway
[106]. Consequently, the aberrant activation of this cascade conveys tumor initiation, cancer progression, and metastatic spread, followed by emerging chemoresistance
[107]. This signaling axis can be effectively abrogated by urolithin A (Uro A)
[108]. The treatment of PDAC cells with Uro A not only inhibited the growth of tumor xenografts and improved the overall survival (OS) of
Ptf1aCre/+;LSL-KrasG12D/+;Tgfbr2flox/flox (PKT) mice, but also reprogrammed the tumor microenvironment by attenuating infiltrated immunosuppressive cells such as TAMs, MDSCs, and Tregs
[108].
Oncogenic mutant Kras
G12D elevates IL-33 expression in PDAC cells, which recruits and activates T
H2 cells. Then, T
H2 cells stimulate tumor growth by secreting protumorigenic cytokines such as IL-4 that exerts major impacts on neighboring innate immune cells (
Figure 1). Studies on animal models unveiled that IL-4-initiated signaling in macrophages can be further orchestrated by Stat6, which in turn regulates interferon regulatory factor 4 (Irf4) that acts as an important transcription factor and harnesses M2 polarization
[109] (
Figure 1). Conversely, Irf4 deficiency impeded the expression of M2-associated signature genes
[110]. In a syngeneic model of PDAC, the inhibition of Irf4 using the immunomodulatory agent pomalidomide resulted in a shift of macrophages towards the M1 population and fosters an immune surveillance antitumor environment along with an improved infiltration of cytotoxic T lymphocytes and enhanced immune responses
[111].
Proinflammatory cytokine IL-20 is a member of the IL-10 family and is expressed predominantly by epithelial cells, monocytes, dendritic cells, and endothelial cells in the TME (
Figure 1). IL-20 was demonstrated to promote M2 polarization, and elevated IL-20 levels in PDAC tumor tissue correlate with poor overall survival
[112]. Inhibiting IL-20 using an antagonistic antibody, 7E, reshapes the TME toward scenarios unfavorable for malignancies in multiple aspects including diminished M2 macrophage infiltration, lightened fibrosis, inhibited tumor growth, and reduced expression of the immunosuppressive regulator PD-L1 on tumor cells
[112].
TAMs remain the primary cell type molding the immune landscape
[75][113], partly fortified by a self-amplifying mechanism. Sialic-acid-binding immunoglobulin-like lectin 15 (SIGLEC15) is upregulated in M2 macrophages and could directly enact immunosuppressive function via binding α-2,3 sialic acid
[114]. Stimulation of the extracellular domain of SIGLEC15 promotes the tyrosine phosphorylation of DNAX-activating protein of 12 kDa (DAP12) and leads to the activation and recruitment of spleen tyrosine kinase (SYK)
[115] (
Figure 1). Joshi et al. further revealed an autocrine-positive feedback loop phenomenon by demonstrating that SYK, in conjunction with the PI3K axis, synergizes M2 polarization, which can be abolished by a dual SYK/PI3K inhibitor, SRX3207
[116]. α-2,3 sialic-acid-bound SIGLEC15 enhances the production of C-C motif chemokine ligand (CCL)2, C-X-C motif chemokine (CXCL)2, and CXCL8 in TAMs, which not only exacerbates immune suppression but also accelerates tumor progression in gastric
[117], esophageal
[118], and bladder carcinomas
[119]. Among them, CCL2 facilitates the mobilization of receptor CCR2
+ inflammatory monocytes from bone marrow to the tumor bed, where they become immunosuppressive TAMs
[120]. Together, SIGLEC15 expression, monocyte mobilization, and M2 polarization form a positive feedback circuit, enabling the recruitment and amplification of TAMs
[114]. In this regard, a clinical trial in patients with nonmetastatic PDAC using the orally dosed small-molecule CCR2 inhibitor (CCR2i) PF-04136309, in combination with FOLFIRINOX, demonstrated improved antitumor efficacy (trial number NCT01413022). However, a compensatory influx of CXCR2
+ neutrophils resulted in a relapse. Yet, this therapeutic resistance can be circumvented by combinatorial blockades targeting both types of infiltrating myeloid cells. Dual treatments not only promote a robust antitumor effect compared to either inhibitor alone, but also improve the overall response to FOLFIRINOX
[121].
On the other hand, CD40, a cell surface receptor belonging to the TNF superfamily, can regulate myeloid cell function and adaptive immunity. Similar to Toll-like receptors (TLRs), the CD40 pathway acts as a linkage between DCs and adaptive immunity in cancer. Ligands of CD40 (CD40L) connect DCs and other immune cells in response to malignancies or pathogenic insults with memory. Yet, agonistic anti-CD40 (αCD40) monoclonal antibodies mimic CD40L in vivo and have been shown to enhance the immunogenicity of cancer vaccines and trigger cancer regressions
[122][123][124], including in pancreatic cancer
[125]. Interestingly, one of the well-studied αCD40 antibodies, selicrelumab, was taken into clinical evaluation as a novel agent for immune activation and cancer immunotherapy, independent from ICB
[126]. CD40 activation by selicrelumab enhanced the polarization of TAMs towards the M1 phenotype, as well as activated the proliferation and infiltration of CD8
+ T lymphocytes and DCs
[126][127]. Together, this treatment transforms the TME from “cold” to “hot” immunity
[126][127]. Surgical samples from patients receiving selicrelumab preoperatively exhibited decreased tumor fibrosis, fewer M2 macrophages, and a greater maturation of intratumoral DCs
[127]. It is noteworthy to mention that, clinically, combinatorial treatments using αCD40 antibodies and ICB ameliorate efficacy in patients who are initially refractory to immunotherapies. Accordingly, Winograd et al. developed an effective treatment regimen with αCD40 antibodies and ICB (αPD-1/αCTLA-4) using a genetically engineered KPC mouse model
[128]. Such success exemplified that the combination of αCD40/ICB, but not either of αCD40 or ICB alone, results in a prominent decline in tumor burden and gain of immunological memory
[128].
2.3. Aberrant Metabolism, Hypoxic TME, and Dysregulated Epigenetics
Indole compounds are evolved from dietary tryptophane upon metabolizations by gut microbials such as lactobacilli. Indoles are the key activators for aryl hydrocarbon receptor (AhR), although tryptophane metabolism by human cells rendered negligible effects. By promoting the polarization of TAMs to M2, elevated AhR expression has been recognized as a central driver of TAM function in responding to multiple cues to promote an immune-suppressive state of the TME
[129] (
Figure 1). Molecular studies delineate that high expression of AhR inhibits IFNγ expression in CD8
+ T cells
[129], while it enhances the expression of immunosuppressive IL-10
[130], TGF-β, and Arg1
[131][132] (
Figure 1). The aforementioned data on animal models coincide with clinical facts in which patients with high AhR expression are strongly correlated with rapid disease progression and increased mortality, along with the immune-suppressive properties associated with TAMs, underscoring the conservation of this regulatory axis in PDAC
[129].
PDAC ubiquitously fosters a hypoxic TME. Hypoxia is a condition where the oxygen pressure is below 5–10 mm of mercury, and this phenomenon can empower cancer metastasis
[133]. The major mechanism executing cellular responses toward hypoxia is the activation and sustainment of hypoxia-inducible factors (HIFs), mainly HIF1 and HIF2, that activate a set of genes facilitating tumor growth, angiogenesis, and metastasis
[134][135]. On the other hand, the endocytosis of cancer, or immune or endothelial cells, can form and release extracellular exosomes
[136]. The tumor-derived exosomal miR-301a-3p, for example, not only is released from hypoxic PDAC, but also promotes M2 polarization and ameliorates the PTEN/PI3Kγ pathway, thereby enhancing metastasis in vitro and in vivo
[137]. Stimulated by a hypoxic TME, HIF-1α further augments the expression of glycolytic enzymes contributing to maintaining bioenergetic homeostasis during hypoxic stress
[138]. In support of this notion, inflammatory cells such as TAMs tend to maneuver metabolism toward glycolysis to meet high energetic demand
[139]. Recent studies have unveiled that hypoxia and glycolysis-related gene signatures are concurrently associated with an unfavorable TME and are used to predict a poor prognosis of PDAC patients
[140]. Hypoxia and glycolysis pathways are upregulated in the prognostically high-risk cohorts compared to the low-risk counterparts
[141][142]. Apart from glucose metabolism, the ablation of HIF2 in CAFs modestly reduces fibrosis and significantly decreases the intratumoral recruitment of M2 macrophages and Treg cells. Similarly, treatment with the clinical HIF2 inhibitor PT2399 abolishes paracrine signaling driven by HIF2, and significantly reduces M2 polarization as well as improves tumor responses to immunotherapy using ICB in PDAC mouse models
[143].
GEM treatment favors TAM infiltration into the tumor mass and shifts the stroma to a predominantly M2 phenotype that conveys notorious survival
[79], owing to the destruction of gemcitabine by M2-released pyrimidines
[144]. Furthermore, paracrine signals from the removal of chemotherapy-generated apoptotic cells can stimulate immune-suppressive controllers in the TME. The phagocytosis of apoptotic cells increases the production of TGF-β1, prostaglandin E2 (PGE2), and platelet-activating factor (PAF), all of which are known to act as anti-inflammatory and immune-suppressive modulators
[145].
Dysregulated epigenetic modulators can influence TAM polarization. An epigenomic analysis of TAMs isolated from PDAC tissues revealed the overexpression of CCCTC binding factor (CTCF), an important epigenetic regulator in TAMs. CTCF can enhance M2 polarization and favor the tumor-promoting properties of the TAMs. CTCF-transcribed long noncoding RNA (LncRNA) of prostaglandin-endoperoxide synthase 2 (PTGS2) antisense NF-
κB1 complex-mediated expression regulator RNA (PACERR) can orchestrate PTGS2 expression. A novel investigation demarcated that transcribed LncRNA PACERR binds CTCF, forming the CTCF/PACERR complex to recruit the E1A binding protein p300 (EP300), which is one of the histone acetyltransferases. Being an epigenetic regulator, this complex not only enhances chromatin accessibility, but also elevates PTGS2 transcription. Excessively expressed PTGS2 is one of the key activators for polarizing M2
[146].
Moreover, cancer progression and the chemoresistance of PDACs have been associated with elevated histone deacetylases (HDACs) and glycogen synthase kinase 3 beta (GSK3B) activity. Accordingly, treatment by the dual inhibitor, metavert, lowers the abundance of M2 macrophages by more than 50%, although the total number of macrophages are unaffected significantly
[147]. These data implicate the molecular cue leading to cancer inhibition by metavert is partially due to the reversion of M2 to the M1 phenotype
[147]. Metavert treatment further downregulates procancer cytokines like IL-6 and IL-4, induces cancer cell apoptosis, and attenuates the expression of cancer stem cell markers, as well as impedes cancer growth and metastases
[147].