Purpose:

Tumor-associated macrophages (TAMs) and the hyperactivation of the PI3K/AKT pathway are involved in the pathogenesis of Hodgkin lymphoma and affect disease outcome. Because the δ and γ isoforms of PI3K are overexpressed in Hodgkin/Reed–Sternberg (HRS) cells and the tumor microenvironment (TME), we propose that the PI3Kδ/γ inhibitor RP6530 might affect both HRS cells and TME, ultimately leading to an enhanced antitumor response.

Experimental Design:

Hodgkin lymphoma cell lines (L-540, KM-H2, and L-428) and primary human macrophages were used to investigate the activity of RP6530 in vitro and in vivo in Hodgkin lymphoma cell line xenografts.

Results:

In vitro, RP6530 besides killing and inhibiting the proliferation of Hodgkin lymphoma cells, downregulated lactic acid metabolism, switching the activation of macrophages from an immunosuppressive M2-like phenotype to a more inflammatory M1-like state. By RNA sequencing, we define tumor glycolysis as a specific PI3Kδ/γ-dependent pathway implicated in the metabolic reprogramming of cancer cells. We identify the metabolic regulator pyruvate kinase M2 as the main mediator of tumor-induced immunosuppressive phenotype of macrophages. Furthermore, we show in human tumor xenografts that RP6530 repolarizes TAMs into proinflammatory macrophages and inhibits tumor vasculature, leading to tumor regression. Interestingly, patients with Hodgkin lymphoma experiencing objective responses (complete response and partial response) in a phase I trial using RP6530 showed a significant inhibition of circulating myeloid-derived suppressor cells and an average mean reduction in serum thymus and activation-regulated chemokine levels of 40% (range, 4%–76%).

Conclusions:

Our results support PI3Kδ/γ inhibition as a novel therapeutic strategy that targets both malignant cells and the TME to treat patients with Hodgkin lymphoma.

Translational Relevance

In preclinical and clinical models of Hodgkin lymphoma, the PI3Kδ/γ inhibitor RP6530 exerts a direct cytotoxic effect on Hodgkin lymphoma cells and reprogrammes the tumor-associated macrophages (TAMs) from a protumor M2 phenotype to an antitumor M1 phenotype, thus reshaping the interplay between cancer cells and their tumor microenvironment (TME). As a consequence of TME reprogramming induced by the PI3Kδ/γ inhibition, Hodgkin lymphoma tumor cells and tumor vasculature are effectively reduced. Our data establish the first evidence of the translational potential of PI3Kδ/γ inhibition in suppressing malignant cell growth and reshaping the microenvironment of Hodgkin lymphoma, suggesting that a novel unique therapeutic opportunity may be achievable for treatment of patients with Hodgkin lymphoma.

Primary refractory and early-relapsed patients with Hodgkin lymphoma experience poor responses to salvage chemotherapy and dismal long-term disease control (1–3). Despite the variety of novel therapeutic options, they represent an unmet medical need urgently requiring novel therapeutic agents to overcome the chemo-refractory phenotype (4–8).

The PI3K/Akt pathway is implicated in the pathogenesis of Hodgkin lymphoma (9). The δ isoform of PI3K is highly expressed in tissues of hematopoietic origin and is involved in the activation, proliferation, survival, homing, and retention of B-cells in lymphoid tissues (10). Idelalisib is the first PI3Kδ inhibitor to be approved for follicular lymphoma and chronic lymphocytic leukemia (11, 12), and we previously reported that PI3Kδ isoform inhibition results in direct Hodgkin lymphoma cell killing (13). The PI3Kγ isoform, although highly expressed in leukocytes, may play a more crucial role in the immune system than that in oncogenesis (10). To date, much effort has been devoted to PI3Kγ as a target in inflammatory diseases driven by leukocytes (14). Inflammation driven by tumor-associated macrophages (TAMs) is now considered a hallmark of cancer, contributing to both cancer cell expansion and angiogenesis (15). Recent data in solid tumors show that selectively targeting the γ isoform of PI3K in TAMs modulates the immunosuppressive tumor microenvironment (TME), resulting in tumor regression (16–18).

TAMs have been implicated in the pathogenesis of Hodgkin lymphoma and have been suggested to negatively impact clinical outcome (19–23). Both the δ and γ isoforms of PI3K are overexpressed in HRS cells and cell of the microenvironment, respectively, thereby representing attractive therapeutic targets (24). Here, we describe the effects of RP6530, a novel PI3Kδ/γ inhibitor currently in phase I to II clinical trials in Europe and the United States. RP6530 exhibits high antiproliferative and cytotoxic activity in Hodgkin lymphoma cell lines in vitro and potent efficacy in vivo in preclinical xenograft mouse models. We show that RP6530 downregulates lactic acid metabolism in Hodgkin lymphoma cells, reducing M2-like polarization of macrophages. Furthermore, we demonstrate that this effect is mediated by the metabolic regulator pyruvate kinase M2 (PKM2). RP6530 reshapes the TME, inhibits angiogenesis, and switches TAM activation from an immunosuppressive M2-like phenotype to a more inflammatory M1-like state. We propose that dual pharmacologic targeting of the δ and γ PI3K isoforms affects both malignant tumor cells and the TME, ultimately leading to an enhanced antitumor response. The ability of RP6530 to affect both Hodgkin lymphoma tumor cells and the Hodgkin lymphoma TME indicates that a novel unique therapeutic opportunity may be achievable for treatment of patients with Hodgkin lymphoma.

Reagents

Rhizen Pharmaceuticals, SA (La Chaux-de-Fonds) kindly provided the PI3Kδ/γ inhibitor RP6530. RP6530 had high potency against PI3Kδ (IC50 = 24.5 nmol/L) and γ (IC50 = 33.2 nmol/L) enzymes with selectivity over α (>300-fold) and β (>120-fold) isoforms with specificity being similar in isoform-specific cell-based assays as well (25). For in vitro experiments, RP6530 was reconstituted in 100% DMSO and further diluted in RPMI1640 to final concentrations of 0.05% and 0.1% DMSO (v/v). For in vivo experiments, RP6530 was dissolved in 0.5% methylcellulose (pH 2.2).

Cell death and cell proliferation assay

Hodgkin lymphoma cell lines (4 × 105 mL−1) were cultured in the absence or presence of RP6530 (ranging between 1.25 and 10 μmol/L) for 24, 48, and 72 hours. Dead and proliferating cells were detected by annexin-V/propidium iodide double staining (Immunostep) and WST assay (BioVision), respectively, according to the manufacturer's instructions. See Supplementary Information for further information.

Cell-cycle analysis

Hodgkin lymphoma cell lines were cultured in the absence or presence of RP6530 (5 or 10 μmol/L) for 48 hours, fixed in 70% ethanol, and then stained with 2.5 μg/mL propidium iodide (PI) (Calbiochem). Cell-cycle status was measured using a FACSCalibur flow cytometry system (BD Biosciences) and analyzed using FlowJo software (Treestar).

RNA preparation and sequencing

polyA-RNA-seq was performed on Hodgkin lymphoma cell lines (L-540 and KM-H2) after RP6530 100 nmol/L for 6 hours and after RP6530 10 μmol/L for 6 and 24 hours. Two biological replicates were profiled for each experimental condition. RNA was purified using an RNeasy Mini Kit (Qiagen) and treated with DNase according to the manufacturer's protocol. Total RNA quality was evaluated using an Agilent Bioanalyzer and an RNA Nano Kit. Only samples with an RIN score ≥8 were processed further. RNA-seq libraries were sequenced using a TruSeq stranded mRNA Kit (Illumina) according to the manufacturer's instructions. See Supplementary Information for further information.

RNA-Seq analysis

RNA-Seq samples were demultiplexed, and FASTQ files were created from BCL files using bcl2fastq (Illumina). Quality control and assessment were performed using FastQC v0.11 (http://www.bioinformatics.babraham.ac.uk/projects/fastqc). STAR v2 was used to align each sample's single-end reads to the Gencode Human reference genome (build GRCh38). The raw read counts were normalized with TMM using edgeR package in R/Bioconductor. Differentially expressed transcripts were calculated in RPKM using edgeR. Significant genes were selected based on an FDR ≤0.1 and logCPM ≥0.5. Hierarchical clustering of the log2 fold change (log2FC) between the vehicle and RP6530-treated samples at 6 and 24 hours was performed with cluster 3.0 using the complete linkage method and visualized using Java TreeView (NCBI Gene Express Omnibus website, GSE105439).

Pathway analysis

Pathway analysis was performed using QIAGEN's commercial Ingenuity Pathway Analysis (IPA, www.qiagen.com/ingenuity) software. GSEA was used in preranked mode. The rank metric was calculated as the sign of log2FCs multiplied by the inverse of adjusted P values. The gene sets used for this analysis were the C2:CP:KEGG and C2:CP:PID Curated Molecular Signatures Database (MSigDB) gene sets. Enrichr was used to find significantly enriched (adjusted P value ≤0.1) KEGG and PID terms for the gene set continuously modulated during the time course. PPI networks for the select MSigDB gene set pathways (KEGG_GLYCOLYSIS_GLUCONEOGENESIS and PID_HIF1_TFPATHWAY) were constructed using the STRING database (http://www.string-db.org/). See Supplementary Information for further information.

Human macrophage differentiation and culture

Human leukocytes from apheresis blood products were obtained from the Pavia Blood Bank. Cells were diluted in PBS and centrifuged over Histopaque 1077 to purify mononuclear cells. An EasySep Human Monocyte Enrichment Kit (Stemcell Technologies) was used to isolate monocytes from peripheral blood mononuclear cells by negative selection. Purified monocytes were cultured in RPMI supplemented with 10% FBS and 50 ng/mL human mCSF (PeproTech). Nonadherent cells were removed after 2 hours by washing, and adherent cells were cultured for 6 days to fully differentiate macrophages.

Macrophage polarization assay

Peripheral blood-derived macrophages were polarized toward an M1 phenotype with the addition of IFNγ (20 ng/mL; Peprotech) and LPS (100 ng/mL; Alexis) or toward an M2 phenotype with the addition of IL4 (20 ng/mL; Peprotech) for 24 hours. Polarized macrophages were incubated with RP6530 (10 μmol/L) for 24 hours, and mRNA expression and flow cytofluorimetric analyses were then performed. RNA was harvested from the cells (Qiagen RNeasy), and SYBR green-based qPCR was performed using primers for human CXCL9, IL12p40, CXCL11, CXCL10, CD80, CCL17, CCL22, IL10, CD301, and CD163. mRNA levels were normalized to actin expression (ΔCt = Ctgene of interestCtActin) and reported as relative mRNA expression (ΔΔCt = 2−(ΔCtsample−ΔCtcontrol)). Primary antibodies to cell surface markers directed against CD14 (M5E2) and CD80 (L307) were from BD Pharmingen, and primary antibodies to cell surface markers directed against CD40 (HB14), CD209 (9E9A8), and CD301 (H037G3) were from eBioscience.

siRNA-mediated gene silencing

Hodgkin lymphoma cell lines (3 × 105) were transfected using Lipofectamine RNAiMAX transfection reagent (Thermo Fisher Scientific) with 1 nmol PKM2 siRNA (#285 and #286) or Silencer Select Negative Control siRNA (#AM4621 and #AM4611; Thermo Fisher Scientific) for 48 hours according to the manufacturer's instructions.

Transwell coculture assay

Hodgkin lymphoma cells (3 × 105 cells) were plated in the lower compartment of 6.5-mm polycarbonate Transwell inserts (pore size 0.4 μm; Corning) and exposed to RP6530 (10 μmol/L) or siRNA transfection as described in the previous section. The next day, 3 × 105 M2-polarized macrophages were placed in the upper compartment of the Transwell inserts and cocultured with the Hodgkin lymphoma cells. After 24 to 48 hours of coculture, RNA was extracted from the macrophages in the upper inserts, and Hodgkin lymphoma cells were washed and then used for subsequent experiments. Control wells contained either Hodgkin lymphoma cells only in the lower compartment with RPMI in the upper compartment or RPMI in the lower compartment with only M2-polarized macrophages or M0 macrophages in the upper compartment.

Immunoblotting

Hodgkin lymphoma cell lines or polarized M1 and M2 macrophages were treated with RP6530 as described in the figure legends and proteins were detected with the indicated antibodies. See Supplementary Information for further information.

Biochemical assays

The lactate concentration in Hodgkin lymphoma cell and polarized macrophage lysates was measured after RP6530 (10 μmol/L) treatment with an L-Lactate Assay Kit (Abcam) according to the manufacturer's instructions.

ELISA

Serum samples were centrifuged at 4,000 rpm for 10 minutes and then kept at −80°C until ELISA assessment. Serum levels of thymus and activation-regulated chemokine (TARC/CCL17) were determined according to the manufacturer's instructions (R&D Systems). The TARC/CCL17 level in patient sera samples was determined by correlating each value duplicate with a standard curve based on a 2-fold serial dilution of recombinant TARC/CCL17 with known concentration.

Flow cytometry staining and analysis

Single-cell suspensions (106 cells in 100 μL total volume) were incubated with FcR-blocking reagent (BD Biosciences) at 4°C for 30 minutes. Staining of cell surface markers was performed at 4°C for 20 minutes with the following primary antibodies: anti-mouse F4/80 (BM8), anti-mouse CD45 (30-F11), anti-mouse/human CD11b (M1/70), anti-mouse CD86 (GL-1), anti-mouse CD301 (LOM-14), anti-mouse MHC-II (M5/114.15.2), anti-mouse CD206 (C068C2), anti-mouse Ki-67 (16A8), anti-human HLA-DR (L243), anti-human CD14 (M5E2), anti-human CD33 (WM53; Biolegend); and anti-mouse Ly6G (1A8) and anti-mouse Ly6C (HK1.4; eBioscence). Unconjugated rabbit anti-mouse NOS2 (Abcam) was also used followed by incubation with secondary goat anti-rabbit Alexa Fluor 647-conjugated antibody (Invitrogen, Molecular Probes). For the gating of the viable cells, a LIVE/DEAD Fixable Violet Dead Cell Stain Kit (Thermo Fisher) was used. For intracellular staining, a Cytofix/Cytoperm and Permwash Staining Kit (BD Pharmigen) was used according to the manufacturer's instructions. Multicolor FACS analysis was performed on a BD FACSCanto II flow cytometer. All data analysis was performed using the flow cytometry analysis program FlowJo (Treestar).

Tumor challenge and treatment experiments

Six- to 8-week-old NOD/SCID mice (20–25 g) were purchased from Charles River Labs and xenografted with L-540 (25 × 106 cells) and KM-H2 (20 × 106 cells) cells by subcutaneous inoculation into the right flank. The treatments started when the tumors were palpable (approximately 200 mm3). RP6530 was administered by oral gavage twice per day at 100 and 150 mg/kg for 3 weeks or at 150 mg/kg for 5 days. The control groups received a vehicle (0.5% methylcellulose, pH 2.2) without the active product. When appropriate, in vivo biotinylation of tumor vasculature (26) was performed 5 days after RP6530 treatment and 3 hours after the last drug administration. The animal experiments were performed according to EU 86/109 Directive (D.L. 116/92 and following additions) and were approved by the institutional Ethical Committee for Animal Experimentation of the Humanitas Clinical and Research Center. See Supplementary Information for further information.

Tumor-infiltrating myeloid cell analysis

Six- to 8-week-old NOD/SCID mice were injected subcutaneously with L-540 (25 × 106 cells) and KM-H2 (20 × 106 cells) cells in 400 μL of PBS in the right flank. On days 19 (for L-540) or 25 (for KM-H2) after tumor injection, the tumor-bearing mice were grouped and treated with RP6530 (150 mg/kg, twice per day, orally) or vehicle (0.5% methylcellulose, pH 2.2) for 5 days. Three hours after the final treatment, the mice were euthanized, and the tumors were snap frozen or digested in a mixture of 0.5 mg/mL collagenase IV and 150 U/mL DNase I in RPMI1640 for 30 minutes at 37°C. Tumor-infiltrating myeloid cells were analyzed by IHC and flow cytometry. Tumor macrophage enrichment was performed by plating cells in FBS-free RPMI containing 1% penicillin/streptomycin for 1 hour at 37 °C and 5% CO2. After 1 hour, nonadherent cells were removed with 3 PBS washes, and RNA was harvested from tumor macrophages (Qiagen RNeasy). SYBR green-based qPCR was performed using primers for murine IL1β, CXCL10, iNOS, TNFα, CD80, CXCL11, IL6, Arg1, CCL22, CCL2, IL10, CD163, CD206, TGFβ, PGF, IGF1, EGF, FGF2, VEGFA, HIF-1α, and PD-L1 (Sigma-Aldrich). mRNA levels were normalized to actin levels (ΔCt = Ctgene of interest − CtActin) and reported as a fold change in the RP6530-treated mice over the vehicle-treated controls.

Immunofluorescence

The in vivo biotinylated tumor nodules were incubated with incubated with Alexa Fluor 568-streptavidin (Invitrogen), pAKT (Ser473; 736E11), pERK1/2 (197G2), and caspase-3 (E-8) antibodies (Santa Cruz), PKM2 (D78A4) from Cell Signaling Technology, CD163 (EPR19518) from Novus Biologicals, MHC-II (OX-6) from Abcam and F4/80 (CI:A3-1) from Bio-Rad. See Supplementary Information for further information.

Immunohistochemistry

Sections from FFPE lymph node biopsies obtained from consenting patients at the time of diagnostic workup or Hodgkin lymphoma cell line cytospins were stained with the following antibodies: p110δ (EPR986) from Novus Biologicals; p110γ (#PA5-28070) from Thermo Fisher Scientific; and pAkt (S473) (736E11), pAkt (T308) (L32A4), pS6 (#2211), and pERK1/2 (20G11) from Cell Signaling Technology. Cryostat sections of in vivo biotinylated tumor nodules were stained with F4/80 (CI:A3-1) from Bio-Rad, HRP-conjugated streptavidin, Ki-67 (MIB-1) from Dako, and VEGF (ab46154) from Abcam. Tumor apoptosis and necrosis were detected via TUNEL staining (Roche). See Supplementary Information for further information.

Statistical analysis

Analysis for significance was performed using 1-way ANOVA with a Tukey's post hoc test for multiple pairwise testing of more than 2 groups and by parametric Student t test when only 2 groups were compared. Two-way ANOVA with Tukey's post hoc test and Dunnett post hoc test was performed when comparing more than 2 groups and 2 variables.

RP6530 inhibits the PI3K/AKT and ERK signaling pathways in Hodgkin lymphoma cells

The PI3K/Akt pathway is consistently activated in Hodgkin lymphoma (24). The δ and γ isoforms of PI3K are highly expressed both in primary HRS cells and microenvironmental cells (Supplementary Fig. S1A). The multiple roles of PI3Kδ and PI3Kγ in HRS cells and reactive cells of the microenvironment support the hypothesis that blocking δ and γ isoforms may provide a therapeutic benefit. To investigate these hypothesis, Hodgkin lymphoma cell lines (L-540, KM-H2, and L-428) displaying an expression pattern of PI3K δ and γ isoforms similar to that of primary HRS cells were used to assess the effects of the PI3Kδ/γ dual inhibitor RP6530 (Supplementary Fig. S1B). Based on pharmacokinetic studies performed in patients enrolled in phase I trial (clinicaltrials.gov identifier NCT02017613) using RP6530, drug concentrations up to 10 μmol/L are pharmacologically achievable (C. Carlo-Stella and colleagues, manuscript in preparation). Exposure of Hodgkin lymphoma cell lines to RP6530 (1.25–10 μmol/L) led to a dose-dependent inhibition of phosphorylation of ERK1/2, Akt and its downstream target proteins (Fig. 1A–C), indicating that targeting PI3Kδ/γ affects both the Akt and MAPK pathways.

RP6530 inhibits cell proliferation through induction of G0–G1 arrest and apoptosis

Incubating L-540, KM-H2, and L-428 cell lines for up to 72 hours with RP6530 (1.25–10 μmol/L) resulted in a significant dose- and time-dependent decrease in cell proliferation down to 30% (Fig. 1D). The cell-cycle data showed that treatment with RP6530 induced 2-fold accumulation of cells in the G0–G1 phase with an accompanying 4-fold decrease in cells in the S phase compared with vehicle controls (Fig. 1E). In addition, caspase-dependent cell death was increased (up to 50%) after RP6530 treatment (Fig. 1F–H). These results showed that RP6530 inhibits cell proliferation through induction of G0–G1 arrest and apoptosis in Hodgkin lymphoma cell lines.

Functional characterization of differentially expressed genes in RP6530-treated Hodgkin lymphoma cell lines

Next, we evaluated by RNA-Seq the anti-lymphoma functions and the underlying genes affected by dual PI3Kδ/γ inhibition. Concordantly downregulated genes in L-540 and KM-H2 cells at 6- or 24-hour time points were involved in cell proliferation, MAPK, JAK/STAT, IL2, IL4/STAT5, glycolysis, HIF1α, and MYC signaling, whereas concordantly upregulated genes were involved in cell death, apoptosis, and cell-cycle deregulation (Supplementary Figs. S2A–S2D, S3, and S4). We confirmed these findings with a much lower concentration of RP6530 (100 nmol/L), which reduced the extent of off-target effects (Supplementary Fig. S5A). A common gene expression signature was created consisting of genes consistently up- and downregulated across all time points in both cell lines (Supplementary Fig. S2B and Fig. 2A). Although no significantly enriched pathways were assessed in the 39 upregulated genes, the 111 genes downregulated following RP6530 treatment were highly enriched in tumor glycolysis and HIF1α signaling (Fig. 2A), a phenomenon already detectable at nmol/L concentration of RP6530 (Supplementary Fig. S5B and S5C), supporting the strong relevance of those 2 pathways in the mechanism of action of RP6530. Among these genes, we identified 28 highly connected hub genes likely to be key drivers in both tumor glycolysis and HIF1α signaling (Fig. 2B). PKM2 was selected as the most important hub gene (Fig. 2C; Supplementary Fig. S6A and S6B). PKM2 catalyzes the final and rate-limiting reaction of the aerobic glycolysis, thus regulating lactic acid production (27) and the Warburg effect in cancer cells (28). Because lactic acid secreted by tumor cells also functions as a critical signaling factor for M2-like polarization of macrophages (28), we hypothesized that by downregulating PKM2, RP6530 might reduce M2 markers expression in M2 polarized macrophages (Fig. 3A). According to the decreased expression of PKM2 after RP6530 and PKM2 siRNA treatment (Supplementary Fig. S6C), we detected a 50% reduction in lactate levels in Hodgkin lymphoma cell lines (Fig. 3B) and significant downregulation of the expression of the M2 markers CCL17 and CCL22 in the M2-like macrophage population (Fig. 3C), supporting an attenuation of the M2 phenotype. PKM2 therefore appears to be a critical determinant of the RP6530-mediated crosstalk between Hodgkin lymphoma tumor cells and macrophages (Fig. 3D).

RP6530 repolarizes macrophages to an M1-like phenotype

Many studies have implicated the PI3K/Akt pathway in macrophage activation (29, 30). RP6530 reduced Akt phosphorylation (Fig. 3E) in primary human M1 or M2 stimulated macrophages (Supplementary Fig. S7A). Differential expression of PI3Kδ and PI3Kγ isoforms was observed in M1 or M2 macrophages; with abundant p110γ in both M1 and M2 macrophages, whereas low levels of p110δ in M1 macrophages compared with M2 macrophages (Fig. 3E). Indeed, M2 macrophages were more sensitive than M1 macrophages to RP6530-induced cell death (Supplementary Fig. S7B) after 48 hours, suggesting that the expression of the δ isoform of PI3K is a prerequisite for the cytotoxic activity of RP6530.

IRF/STAT signaling is crucial in modulating macrophage polarization. The activation of IRF/STAT pathway by IFNs induces an M1-like phenotype (via STAT1), whereas IL4-induced IRF/STAT pathway activation generates an M2-like phenotype (via STAT6; ref. 31). RP6530 sustained and activated STAT1 phosphorylation in M1 and M2 macrophages, respectively, while inhibiting STAT6 phosphorylation in M2 macrophages (Fig. 3F), suggesting that RP6530 directly regulates macrophage polarization. Consistent with these findings, RP6530 markedly inhibited PKM2 mRNA and protein expression (Supplementary Fig. S8A and S8B), leading to the inhibition of lactic acid production in both M1 and M2-like macrophages (Supplementary Fig. S8C). We further tested RNA and protein expression of M1 and M2 markers in M1 or M2 macrophages after RP6530 treatment. The expression of representative M2 markers (CCL17, CCL22, CD163) was reduced, whereas M1 markers (CXCL9, CXCL10, CXCL11) were higher in RP6530-treated M2 macrophages (Fig. 3G and Supplementary Fig. S8D). Taken together, these findings demonstrated that RP6530 switches the activation of macrophages from an immunosuppressive M2-like phenotype to an inflammatory M1-like state.

RP6530 induces functional reprogramming of TAMs and decreases MDSCs in preclinical models and clinical samples from a phase I study

Based on previously published data showing that PI3Kγ is predominantly expressed in the myeloid cell compartment (32), we reasoned that RP6530, as a dual PI3Kδ/γ inhibitor, might affect the accumulation of TAMs in vivo in Hodgkin lymphoma tumors. Indeed, we showed a significant reduction of F4/80+ TAMs in L-540 or KM-H2 tumors after RP6530 treatment (P < 0.0001; Fig. 4A and B). In addition, RP6530 skewed the macrophage phenotype toward classically activated macrophages (M1) in vivo. In RP6530-treated Hodgkin lymphoma xenograft, a shift within the macrophage population towards fewer CD206+ and CD301+ macrophages (M2) and more CD86+ and MHC-II+ macrophages (M1) was detected (Fig. 4C). We further investigated whether RP6530 could influence the function of TAMs by directly modulating their activity, and indeed we observed a significant downregulation of common M2-related genes such as arginase-1 (Arg1), IL10, and CCL2, and a concomitant upregulation of M1-related genes such as inducible nitric oxide synthase (iNOS), CD80 and CXCL11 (Fig. 4D; Supplementary Fig. S9), resulting in a significantly increased M1:M2 ratio and in a less immunosuppressive TME. To strengthen this observation, we examined whether RP6530 affected the myeloid-derived suppressor cell (MDSC) compartment in vivo. Besides reducing the percentage of tumor-infiltrating and splenic MDSCs (Fig. 4E), RP6530 was found to downregulate the expression of iNOS by M-MDSCs, thereby inhibiting their suppressive function in Hodgkin lymphoma tumors (Fig. 4F). Furthermore, inhibition of circulating M-MDSCs was correlated with the clinical outcomes of patients with Hodgkin lymphoma treated with RP6530 (Fig. 4G).

Our current findings demonstrate a PI3Kδ/γ-dependent inhibition of several macrophage-attracting chemokines, such as colony-stimulating factor-1 (CSF-1), CC chemokine ligand 5 (CCL5), and thymus and activation-regulated chemokine (TARC/CCL17; Supplementary Fig. S3). TARC is highly expressed by HRS cells (33), suggesting its role as a biomarker for response evaluation (34). We therefore investigated whether serum TARC levels were correlated with the clinical outcomes of patients with Hodgkin lymphoma enrolled in a phase I trial (clinicaltrials.gov identifier NCT02017613) using RP6530 (35). Upon RP6530 treatment, serum TARC levels were evaluated in 14 patients with Hodgkin lymphoma. Patients achieving complete or partial remission (n = 4) experienced an average mean reduction in serum TARC levels of 40% (range, 4–76%) after 1 month of therapy, whereas the levels were unchanged in patients experiencing SD (n = 7) or PD (n = 3), suggesting that the ability of RP6530 to reduce TARC is likely a result of HL tumor cell death and TAMs repolarization to an M1-like phenotype (Fig. 4H and I).

RP6530 reduces tumor angiogenesis and TAM expression of proangiogenic factors

In addition to being immunosuppressive, pro-tumor TAMs contribute to abnormal tumor vasculature (15, 36). Re-polarization of the TAMs phenotype toward M1 by PI3Kδ/γ inhibition was associated with a marked reduction of pro-angiogenic factors (EGF, VEGFA, HIF1α; Fig. 5A). Given the notion that VEGF is a primary activating factor of angiogenesis and a macrophage chemotactic protein (37, 38), we showed that RP6530 treatment almost completely decreased the expression of VEGF in L-540 and KM-H2 tumors (Fig. 5B). In addition, microvessel density (average 80% inhibition of endothelial areas, P < 0.0001; Fig. 5C), and endothelial and tumor Akt and ERK1/2 phosphorylation (Fig. 5D) were reduced in RP6530-treated mice. In line with the strong inhibition of ERK1/2 and/or Akt phosphorylation on vascular cells, we detected a severe increase in tumor endothelial cell apoptosis manifested by increased expression of caspase-3 (Supplementary Fig. S10).

RP6530 suppresses tumor growth in an Hodgkin lymphoma xenograft model and exerts in vivo antiproliferative, apoptotic, and necrotic effects

The effect of RP6530 on Hodgkin lymphoma tumor growth was determined. RP6530 significantly (P < 0.0001) reduced the in vivo growth of L-540 and KM-H2 xenografts [tumor growth inhibition (TGI) = 52% and 46% at 150 mg/kg, respectively; Fig. 6A and B]. These findings were associated with a strong decrease in Ki-67 expression in tumor cells (Fig. 6C), suggesting that RP6530 inhibits tumor cell proliferation. In addition, the antiproliferative effect was associated with 12-fold increase in tumor cell apoptosis in the KM-H2 nodules, compared with that in the vehicle-treated controls (Fig. 6D), suggesting that the effect of RP6530 on tumor cell growth is both cytostatic and cytotoxic. Because apoptosis was a prominent feature of tumor or endothelial cells in RP6530-treated mice, tumors from these animals even showed large areas of nonhemorrhagic tumor necrosis by hematoxylin/eosin and TUNEL staining (Fig. 6E), suggesting that hypoxic conditions after tumor vessel inhibition might have triggered tumor destruction. RP6530 significantly increased necrotic areas in mice bearing L-540 (3% vs. 20%, P < 0.01) and KM-H2 (4% vs. 26%, P < 0.0001) xenografts compared with those in the vehicle-treated controls (Fig. 6E).

In this study, we demonstrate that the dual PI3Kδ/γ inhibitor RP6530 directly targets Hodgkin lymphoma tumor cells and acts as a critical regulator of signals involved in communication between tumor cells and macrophages. Our data show that PI3Kδ and PI3Kγ are expressed in Hodgkin lymphoma tumor and microenvironmental cells and that inhibition of these isoforms not only suppresses tumor growth and tumor vasculature but also repolarizes tumor-promoting M2-like TAMs toward tumor-suppressive M1-like TAMs by downregulating the limiting glycolytic enzyme PKM2 (39).

Over the past decade, new biological insights have revealed the key role of the TME in the pathogenesis of Hodgkin lymphoma (19). The cross-talk between HRS cells and the cells of the Hodgkin lymphoma microenvironment sustains tumor growth and survival (40). TAMs and MDSCs release immune-suppressive factors that inhibit T-cell-mediated antitumor responses (41), and therapeutic approaches that alter the Hodgkin lymphoma microenvironment, changing it from protective to cytotoxic, hold some promise as novel therapeutics for patients with Hodgkin lymphoma (42). In addition, genomic advances in Hodgkin lymphoma have provided insights into deregulation of key nodal signaling pathways, including the PI3K, NF-κB, and JAK/STAT pathways, which are amenable to small-molecule targeting (43). Among these pathways, the PI3K/Akt pathway and its downstream targets have emerged as central regulators of M2 phenotype activation in macrophages (29). In this context, we considered the high therapeutic potential of targeting the PI3K/Akt pathway to kill Hodgkin lymphoma tumor cells and circumvent the supportive Hodgkin lymphoma microenvironment.

Recent studies have revealed that PKM2-dependent lactic acid production by tumor cells has an important role in the Warburg effect (44). We found that RP6530 inhibits PKM2 in Hodgkin lymphoma cells, preventing its function in regulating the M2-like macrophage polarization through lactate production. Indeed, RP6530 indirectly downregulated the expression of M2 markers in macrophages via Hodgkin lymphoma cells, thus identifying the therapeutic value of reprogramming macrophages in Hodgkin lymphoma. In addition, using specific PKM2 siRNAs, we showed that inhibition of PKM2 is critical for lactate production in HL cells and for stabilization of the M2 phenotype in macrophages; these findings agree with previous studies demonstrating that the activation of PKM2 attenuates the LPS-induced proinflammatory M1 macrophage phenotype while promoting traits typical of an M2 macrophage (45). The effects on lactate production by tumor cells and on cytokines secretion by macrophages were early events detected at 24 hours of coculture and were not influenced by RP6530-induced apoptosis of Hodgkin lymphoma cell lines, an event appearing only after 48 hours of RP6530 exposure (Fig. 1F; Supplementary Fig. S6B). In addition, RNA-Seq identified downregulation of PKM2 as a key player in the modulation of macrophages occurring at 6 hours of incubation.

Clinical evidences have shown that an increased number of M2-like TAMs is correlated with treatment failure and poor prognosis in Hodgkin lymphoma (21, 22). Therefore, TAM-targeting immunotherapies represent a promising cancer therapeutic approach (41). In addition to preferentially inducing apoptosis in M2-like macrophages, RP6530 inhibits the expression of several macrophage-attracting chemokines, such as CSF-1, CCL5, and TARC/CCL17 in Hodgkin lymphoma cell lines. These in vitro findings were further validated by in vivo experiments in human tumor xenografts showing that RP6530 not only exerted potent antitumor effects as shown by significant TGI, but also reprogrammed TAMs to an M1-like phenotype. RP6530 directly downregulated the immune-suppressive transcriptional signature of tumor-derived macrophages, thus suppressing the expression of Arg1, TGFβ, and IL10 and stimulating the expression of IL1β and CXCL11. Moreover, RP6530 reduced the percentage of tumor-infiltrating and splenic MDSCs in Hodgkin lymphoma xenografts. Our system using Hodgkin lymphoma cell lines, which are by their very nature microenvironment-independent, represents clearly a limitation to investigate the interaction between the TME and tumor cells. However, the effects that we observed on macrophage reprogramming in vitro and the reduction of MDSCs in mice by RP6530, have been validated in a phase I trial using RP6530. In this trial, responsive patients with Hodgkin lymphoma experienced significant reduction of circulating TARC levels and inhibition of circulating MDSCs. TARC has been implicated in the recruitment of Th2 lymphocytes (46), and in the suppression of classically activated M1 macrophages (47). Reductions in TARC levels were observed in patients with Hodgkin lymphoma under standard chemotherapy (33), as well as in patients with Hodgkin lymphoma treated with Akt and multikinase inhibitors, supporting its role as a biomarker for response evaluation (34).

Notably, recent data in solid tumors show that PI3Kγ signaling regulates the switch between macrophage polarization and that selectively targeting the γ isoform of PI3K in TAMs inhibited their immunosuppressive phenotype resulting in tumor regression (16, 17). Our previous finding on the anti-lymphoma efficacy of the PI3Kδ inhibitor TGR-1202 also highlighted the potential of targeting the δ isoform of PI3K in Hodgkin lymphoma (13). Thus, our current work confirmed and expanded these observations by using a dual PI3Kδ/γ inhibitor. In addition, we demonstrated the effects of PI3Kδ/γ inhibition on tumor vasculature and identified TARC as a potential biomarker of response in patients with HL receiving PI3Kδ/γ inhibitor. These findings are particularly relevant to HL patients as the survival, proliferation, and immune escape of malignant HRS cells are highly dependent on the interactions with immune microenvironment.

Macrophage recruitment and reprogramming by tumor cells are well known to produce several angiogenic factors that mediate tumor angiogenesis (37). In addition to its critical role in tumor cell survival and cellular metabolism, the PI3K pathway is also involved in angiogenesis (32). Thus, we speculate that targeting PI3Kδ/γ as a common regulator of angiogenesis in macrophages and of tumor cell survival will likely provide a more effective strategy for HL treatment. RP6530 downregulated VEGF expression in tumor cells and TAMs, which led to tumor angiogenesis inhibition and tumor growth reduction. Thus, these findings demonstrated that RP6530 has antiangiogenic effects in Hodgkin lymphoma.

In conclusion, our findings reveal the important signaling role of PI3Kδ/γ in the induction of TAMs polarization and the subsequent promotion of tumor growth in Hodgkin lymphoma. We show that modulating the suppressive phenotype of these cells towards a more inflammatory one can be achieved by targeting PI3Kδ/γ with RP6530. As a consequence, Hodgkin lymphoma tumor burden and tumor vasculature were effectively reduced. Our data establish the first evidence of the translational potential of PI3Kδ/γ inhibition in targeting malignant cells and reshaping the TME in Hodgkin lymphoma.

A. Santoro is a consultant/advisory board member for Bristol-Myers Squibb, Eisai, Gilead, Novartis, Pfizer, Roche, Sandoz, and Servier. C. Carlo-Stella reports receiving commercial research grants from Rhizen Pharmaceuticals. No potential conflicts of interest were disclosed by the other authors.

Conception and design: S.L. Locatelli, S. Viswanadha, A. Santoro, A. Sica, C. Carlo-Stella

Development of methodology: S.L. Locatelli, G. Careddu

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S.L. Locatelli, G. Careddu, F.M. Consonni, A. Maeda, A. Sica

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): S.L. Locatelli, S. Serio, A. Maeda, P. Allavena, A. Sica, C. Carlo-Stella

Writing, review, and/or revision of the manuscript: S.L. Locatelli, S. Viswanadha, L. Castagna, A. Santoro, A. Sica, C. Carlo-Stella

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): S.L. Locatelli, S. Serio

Study supervision: S.L. Locatelli, S. Vakkalanka, C. Carlo-Stella

Other (I have performed ELISA assay for the samples from patients.): A. Maeda

Authors thank Alberto Mantovani (Humanitas University, Rozzano, Milan, Italy), and Giorgio Inghirami (Department of Pathology and Laboratory Medicine, Weill Cornell Medical College, Cornell University, NJ) for review of the manuscript and discussion. This work was supported in part by grants from the Italian Association for Cancer Research, Milan, Italy (AIRC, project No. 20575 to C. Carlo-Stella); Fondazione Regionale Ricerca Biomedica, Milan, Italy (FRRB project No. 2015-0042 to A. Sica); Worldwide Cancer Research, UK (grant #15-1346 to P. Allavena); Ministero dell'Istruzione dell'Università e della Ricerca (MIUR), Milan, Italy (PRIN, project No. C52F16000940001 to A. Sica); Irish Research Council 2017, Ireland (IRC, project No. 19885 to A. Sica). S.L. Locatelli is supported by Fondazione Regionale Ricerca Biomedica, Milan, Italy (FRRB project No. 2015-0042). A. Maeda is recipient of a Marie Skłodowska-Curie Individual European Fellowship-H2020-MSCA-IF-2015-EF-ST (No. 706557).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Engert
A
,
Diehl
V
,
Franklin
J
,
Lohri
A
,
Dorken
B
,
Ludwig
WD
, et al
Escalated-dose BEACOPP in the treatment of patients with advanced-stage Hodgkin's lymphoma: 10 years of follow-up of the GHSG HD9 study
.
J Clin Oncol
2009
;
27
:
4548
54
.
2.
Canellos
GP
,
Rosenberg
SA
,
Friedberg
JW
,
Lister
TA
,
Devita
VT
. 
Treatment of Hodgkin lymphoma: a 50-year perspective
.
J Clin Oncol
2014
;
32
:
163
8
.
3.
Moskowitz
AJ
,
Perales
MA
,
Kewalramani
T
,
Yahalom
J
,
Castro-Malaspina
H
,
Zhang
Z
, et al
Outcomes for patients who fail high dose chemoradiotherapy and autologous stem cell rescue for relapsed and primary refractory Hodgkin lymphoma
.
Br J Haematol
2009
;
146
:
158
63
.
4.
Younes
A.
Beyond chemotherapy: new agents for targeted treatment of lymphoma
.
Nat Rev Clin Oncol
2011
;
8
:
85
96
.
5.
Younes
A
,
Bartlett
NL
,
Leonard
JP
,
Kennedy
DA
,
Lynch
CM
,
Sievers
EL
, et al
Brentuximab vedotin (SGN-35) for relapsed CD30-positive lymphomas
.
N Engl J Med
2010
;
363
:
1812
21
.
6.
Younes
A
,
Gopal
AK
,
Smith
SE
,
Ansell
SM
,
Rosenblatt
JD
,
Savage
KJ
, et al
Results of a pivotal phase II study of brentuximab vedotin for patients with relapsed or refractory Hodgkin's lymphoma
.
J Clin Oncol
2012
;
30
:
2183
9
.
7.
Ansell
SM
,
Lesokhin
AM
,
Borrello
I
,
Halwani
A
,
Scott
EC
,
Gutierrez
M
, et al
PD-1 blockade with nivolumab in relapsed or refractory Hodgkin's lymphoma
.
N Engl J Med
2015
;
372
:
311
9
.
8.
Carbone
A
,
Gloghini
A
,
Castagna
L
,
Santoro
A
,
Carlo-Stella
C
. 
Primary refractory and early-relapsed Hodgkin's lymphoma: strategies for therapeutic targeting based on the tumour microenvironment
.
J Pathol
2015
;
237
:
4
13
.
9.
Kuppers
R
,
Engert
A
,
Hansmann
ML
. 
Hodgkin lymphoma
.
J Clin Invest
2012
;
122
:
3439
47
.
10.
Fruman
DA
,
Rommel
C
. 
PI3Kdelta inhibitors in cancer: rationale and serendipity merge in the clinic
.
Cancer Discov
2011
;
1
:
562
72
.
11.
Gopal
AK
,
Kahl
BS
,
de Vos
S
,
Wagner-Johnston
ND
,
Schuster
SJ
,
Jurczak
WJ
, et al
PI3Kdelta inhibition by idelalisib in patients with relapsed indolent lymphoma
.
N Engl J Med
2014
;
370
:
1008
18
.
12.
Brown
JR
,
Byrd
JC
,
Coutre
SE
,
Benson
DM
,
Flinn
IW
,
Wagner-Johnston
ND
, et al
Idelalisib, an inhibitor of phosphatidylinositol 3-kinase p110delta, for relapsed/refractory chronic lymphocytic leukemia
.
Blood
2014
;
123
:
3390
7
.
13.
Locatelli
SL
,
Careddu
G
,
Inghirami
G
,
Castagna
L
,
Sportelli
P
,
Santoro
A
, et al
The novel PI3K-delta inhibitor TGR-1202 enhances Brentuximab Vedotin-induced Hodgkin lymphoma cell death via mitotic arrest
.
Leukemia
2016
;
30
:
2402
5
.
14.
Rommel
C
,
Camps
M
,
Ji
H
. 
PI3K delta and PI3K gamma: partners in crime in inflammation in rheumatoid arthritis and beyond?
Nat Rev Immunol
2007
;
7
:
191
201
.
15.
Hanahan
D
,
Weinberg
RA
. 
Hallmarks of cancer: the next generation
.
Cell
2011
;
144
:
646
74
.
16.
Kaneda
MM
,
Messer
KS
,
Ralainirina
N
,
Li
H
,
Leem
CJ
,
Gorjestani
S
, et al
PI3Kgamma is a molecular switch that controls immune suppression
.
Nature
2016
;
539
:
437
42
.
17.
De Henau
O
,
Rausch
M
,
Winkler
D
,
Campesato
LF
,
Liu
C
,
Cymerman
DH
, et al
Overcoming resistance to checkpoint blockade therapy by targeting PI3Kgamma in myeloid cells
.
Nature
2016
;
539
:
443
7
.
18.
Kaneda
MM
,
Cappello
P
,
Nguyen
AV
,
Ralainirina
N
,
Hardamon
CR
,
Foubert
P
, et al
Macrophage PI3Kgamma drives pancreatic ductal adenocarcinoma progression
.
Cancer Discov
2016
;
6
:
870
85
.
19.
Steidl
C
,
Connors
JM
,
Gascoyne
RD
. 
Molecular pathogenesis of Hodgkin's lymphoma: increasing evidence of the importance of the microenvironment
.
J Clin Oncol
2011
;
29
:
1812
26
.
20.
Liu
WR
,
Shipp
MA
. 
Signaling pathways and immune evasion mechanisms in classical Hodgkin lymphoma
.
Blood
2017
;
130
:
2265
70
.
21.
Steidl
C
,
Lee
T
,
Shah
SP
,
Farinha
P
,
Han
G
,
Nayar
T
, et al
Tumor-associated macrophages and survival in classic Hodgkin's lymphoma
.
N Engl J Med
2010
;
362
:
875
85
.
22.
Tan
KL
,
Scott
DW
,
Hong
F
,
Kahl
BS
,
Fisher
RI
,
Bartlett
NL
, et al
Tumor-associated macrophages predict inferior outcomes in classic Hodgkin lymphoma: a correlative study from the E2496 Intergroup trial
.
Blood
2012
;
120
:
3280
7
.
23.
Koh
YW
,
Park
CS
,
Yoon
DH
,
Suh
C
,
Huh
J
. 
CD163 expression was associated with angiogenesis and shortened survival in patients with uniformly treated classical Hodgkin lymphoma
.
PLoS One
2014
;
9
:
e87066
.
24.
Locatelli
SL
,
Careddu
G
,
Stirparo
GG
,
Castagna
L
,
Santoro
A
,
Carlo-Stella
C
. 
Dual PI3K/ERK inhibition induces necroptotic cell death of Hodgkin Lymphoma cells through IER3 downregulation
.
Sci Rep
2016
;
6
:
35745
.
25.
Vakkalanka
S
,
Nyayapathy
S
,
Viswanadha
S
. 
RP6530, a dual PI3K delta/gamma inhibitor, potentiates ruxolitinib activity in the JAK2-V617F mutant erythroleukemia cell lines
.
[abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18–22; Philadelphia; Philadelphia (PA): AACR; 2015. Abstract No. 2704
.
26.
Lavazza
C
,
Carlo-Stella
C
,
Giacomini
A
,
Cleris
L
,
Righi
M
,
Sia
D
, et al
Human CD34+ cells engineered to express membrane-bound tumor necrosis factor-related apoptosis-inducing ligand target both tumor cells and tumor vasculature
.
Blood
2010
;
115
:
2231
40
.
27.
Alves-Filho
JC
,
Palsson-McDermott
EM
. 
Pyruvate kinase M2: a potential target for regulating inflammation
.
Front Immunol
2016
;
7
:
145
.
28.
Colegio
OR
,
Chu
NQ
,
Szabo
AL
,
Chu
T
,
Rhebergen
AM
,
Jairam
V
, et al
Functional polarization of tumour-associated macrophages by tumour-derived lactic acid
.
Nature
2014
;
513
:
559
63
.
29.
Vergadi
E
,
Ieronymaki
E
,
Lyroni
K
,
Vaporidi
K
,
Tsatsanis
C
. 
Akt signaling pathway in macrophage activation and M1/M2 polarization
.
J Immunol
2017
;
198
:
1006
14
.
30.
Covarrubias
AJ
,
Aksoylar
HI
,
Horng
T
. 
Control of macrophage metabolism and activation by mTOR and Akt signaling
.
Semin Immunol
2015
;
27
:
286
96
.
31.
Sica
A
,
Mantovani
A
. 
Macrophage plasticity and polarization: in vivo veritas
.
J Clin Invest
2012
;
122
:
787
95
.
32.
Okkenhaug
K
,
Graupera
M
,
Vanhaesebroeck
B
. 
Targeting PI3K in cancer: impact on tumor cells, their protective stroma, angiogenesis, and immunotherapy
.
Cancer Discov
2016
;
6
:
1090
105
.
33.
Weihrauch
MR
,
Manzke
O
,
Beyer
M
,
Haverkamp
H
,
Diehl
V
,
Bohlen
H
, et al
Elevated serum levels of CC thymus and activation-related chemokine (TARC) in primary Hodgkin's disease: potential for a prognostic factor
.
Cancer Res
2005
;
65
:
5516
9
.
34.
Guidetti
A
,
Carlo-Stella
C
,
Locatelli
SL
,
Malorni
W
,
Mortarini
R
,
Viviani
S
, et al
Phase II study of perifosine and sorafenib dual-targeted therapy in patients with relapsed or refractory lymphoproliferative diseases
.
Clin Cancer Res
2014
;
20
:
5641
51
.
35.
Carlo-Stella
C
,
Delarue
R
,
Barde
PJ
,
Scarfo
L
,
Kumar
U
,
Viswanadha
S
, et al
Clinical activity and safety of RP6530, a dual PI3Kδ/γ inhibitor, in patients with advanced hematologic malignancies: final analysis of a phase I multicenter study
.
[abstract]. In: Blood; 2016 Dec 3–6; San Diego, CA. Washington (DC): American Society of Hematology; 2016. Abstract No. 3011
.
36.
Joshi
S
,
Singh
AR
,
Zulcic
M
,
Durden
DL
. 
A macrophage-dominant PI3K isoform controls hypoxia-induced HIF1alpha and HIF2alpha stability and tumor growth, angiogenesis, and metastasis
.
Mol Cancer Res
2014
;
12
:
1520
31
.
37.
Murdoch
C
,
Muthana
M
,
Coffelt
SB
,
Lewis
CE
. 
The role of myeloid cells in the promotion of tumour angiogenesis
.
Nat Rev Cancer
2008
;
8
:
618
31
.
38.
Lewis
CE
,
Harney
AS
,
Pollard
JW
. 
The multifaceted role of perivascular macrophages in tumors
.
Cancer Cell
2016
;
30
:
18
25
.
39.
Iqbal
MA
,
Gupta
V
,
Gopinath
P
,
Mazurek
S
,
Bamezai
RN
. 
Pyruvate kinase M2 and cancer: an updated assessment
.
FEBS Lett
2014
;
588
:
2685
92
.
40.
Aldinucci
D
,
Gloghini
A
,
Pinto
A
,
De Filippi
R
,
Carbone
A
. 
The classical Hodgkin's lymphoma microenvironment and its role in promoting tumour growth and immune escape
.
J Pathol
2010
;
221
:
248
63
.
41.
Mantovani
A
,
Marchesi
F
,
Malesci
A
,
Laghi
L
,
Allavena
P
. 
Tumour-associated macrophages as treatment targets in oncology
.
Nat Rev Clin Oncol
2017
;
14
:
399
416
.
42.
Vardhana
S
,
Younes
A
. 
The immune microenvironment in Hodgkin lymphoma: T cells, B cells, and immune checkpoints
.
Haematologica
2016
;
101
:
794
802
.
43.
Kuppers
R
. 
The biology of Hodgkin's lymphoma
.
Nat Rev Cancer
2009
;
9
:
15
27
.
44.
Yang
L
,
Xie
M
,
Yang
M
,
Yu
Y
,
Zhu
S
,
Hou
W
, et al
PKM2 regulates the Warburg effect and promotes HMGB1 release in sepsis
.
Nat Commun
2014
;
5
:
4436
.
45.
Palsson-McDermott
EM
,
Curtis
AM
,
Goel
G
,
Lauterbach
MA
,
Sheedy
FJ
,
Gleeson
LE
, et al
Pyruvate kinase M2 regulates Hif-1alpha activity and IL-1beta induction and is a critical determinant of the warburg effect in LPS-activated macrophages
.
Cell Metab
2015
;
21
:
65
80
.
46.
Andrew
DP
,
Ruffing
N
,
Kim
CH
,
Miao
W
,
Heath
H
,
Li
Y
, et al
C-C chemokine receptor 4 expression defines a major subset of circulating nonintestinal memory T cells of both Th1 and Th2 potential
.
J Immunol
2001
;
166
:
103
11
.
47.
Katakura
T
,
Miyazaki
M
,
Kobayashi
M
,
Herndon
DN
,
Suzuki
F
. 
CCL17 and IL-10 as effectors that enable alternatively activated macrophages to inhibit the generation of classically activated macrophages
.
J Immunol
2004
;
172
:
1407
13
.