Activation of IGF signaling is a major oncogenic event in diverse cancers, including hepatocellular carcinoma (HCC). In this setting, the insulin-like growth factor binding protein IGFBP7 inhibits IGF signaling by binding the IGF1 receptor (IGF1R), functioning as a candidate tumor suppressor. IGFBP7 abrogates tumors by inhibiting angiogenesis and inducing cancer-specific senescence and apoptosis. Here, we report that Igfbp7-deficient mice exhibit constitutively active IGF signaling, presenting with proinflammatory and immunosuppressive microenvironments and spontaneous liver and lung tumors occurring with increased incidence in carcinogen-treated subjects. Igfbp7 deletion increased proliferation and decreased senescence of hepatocytes and mouse embryonic fibroblasts, effects that were blocked by treatment with IGF1 receptor inhibitor. Significant inhibition of genes regulating immune surveillance was observed in Igfbp7−/− murine livers, which was associated with a marked inhibition in antigen cross-presentation by Igfbp7−/− dendritic cells. Conversely, IGFBP7 overexpression inhibited growth of HCC cells in syngeneic immunocompetent mice. Depletion of CD4+ or CD8+ T lymphocytes abolished this growth inhibition, identifying it as an immune-mediated response. Our findings define an immune component of the pleiotropic mechanisms through which IGFBP7 suppresses HCC. Furthermore, they offer a genetically based preclinical proof of concept for IGFBP7 as a therapeutic target for immune management of HCC. Cancer Res; 77(15); 4014–25. ©2017 AACR.

Evasion from immune surveillance is an important mechanism facilitating tumor development and progression (1). Three equally important events are necessary for immune evasion, the masking of neoantigens, downregulation of antigen presentation machinery, and irresponsiveness to IFNγ-mediated killing by immune cells (2). Tumor antigenic peptides are generally produced in the cytosol via processing by proteasomes that contain interferon-γ-inducible subunits LMP-2, LMP-7, and LMP-10 (3). The 8 to 9 amino acids antigen peptides thus produced are translocated by transporters associated with antigen processing (TAP-1 and TAP-2) to the endoplasmic reticulum (ER) where the peptides are assembled with MHC class I heavy chain and β2-microglobulin light chain and are transported to the cell surface to be presented to CD8+ cytotoxic T lymphocytes (CTL). The components of the antigen presentation pathway, such as TAP1/2 and LMP2/7, are downregulated in cancers, including HCC, resulting in a loss of immune-surveillance and initiation and progression of the disease (4).

Insulin-like growth factor (IGF) signaling plays an important oncogenic role in hepatocellular carcinoma (HCC; ref. 5). Insulin-like growth factor binding protein-7 (IGFBP7) is a secreted protein that binds to IGF1 receptor (IGF1R) and blocks activation by IGFs (6). IGFBP7 functions as a tumor suppressor in a variety of cancers, including HCC, where its expression is markedly downregulated (7–9). Genomic deletion and promoter hypermethylation cause downregulation of IGFBP7 in HCC (7, 10). Recombinant IGFBP7 (rIGFBP7) protein induces senescence and/or apoptosis and inhibits angiogenesis in diverse cancers either in vitro or in nude mice xenograft models (8, 9, 11). Stable overexpression of IGFBP7 in aggressive human HCC cells led to inhibition in IGF signaling, induced senescence, inhibited proliferation, and resulted in profound inhibition in xenograft growth in nude mice, which was accompanied by marked inhibition in angiogenesis (7). Intratumoral injection of an adenovirus expressing IGFBP7 (Ad.IGFBP7) eradicated both injected tumors and noninjected tumors established in the other flank of nude mice, indicating that IGFBP7 not only has direct effect on primary cancer but also exerts a “by-stander” antitumor effect (12). However, the mechanism by which IGFBP7 exerts this “by-stander” effect remains to be determined.

In this article, we describe the generation and characterization of an Igfbp7 knockout (Igfbp7−/−) mouse that establishes the tumor suppressor functions of IGFBP7 and unravels a novel role of IGFBP7 in regulating an antitumor immune response. These studies indicate that IGFBP7 inhibits cancer by pleiotropic mechanisms and might be an effective therapeutic for HCC and other cancers.

Generation of Igfbp7−/− mouse

Igfbp-7−/− mouse was created in a pure C57BL/6 background using a Cre-loxP strategy (Supplementary Fig. S1). All animal studies were approved by the Institutional Animal Care and Use Committee at Virginia Commonwealth University and were performed in accordance with the Animal Welfare Act, the PHS Policy on Humane Care and Use of Laboratory Animals, and the U.S. Government Principles for the Utilization and Care of Vertebrate Animals Used in Testing, Research, and Training.

Primary cells isolation and culture conditions

All primary cells were used immediately after isolation and were mycoplasma free. Primary mouse hepatocytes were isolated and cultured in Williams E Medium containing NaHCO3, l-glutamine, insulin (1.5 mmol/L), and dexamethasone (0.1 mmol/L) as described (13). Kupffer cells were isolated from liver homogenates by centrifuging at 500 RPM for 10 minutes (14). The supernatant containing immune cells was sorted for CD11b+F4/80+ cells using FACSAria II (BD Biosciences). Mouse embryonic fibroblasts (MEF) were isolated as described from (E13.5 embryos) and were cultured in DMEM containing 10% fetal bovine serum (FBS; ref. 15). Bone marrow–derived macrophages and peritoneal macrophages were isolated according to standard protocols (16). Bone marrow cells were isolated from femurs of C57BL/6 mice and were differentiated into macrophages using RMPI-1640 medium supplemented with 10% heat-inactivated FBS and 20% L929 conditioned media for 7days. At day 7, the media was changed to complete RPMI-1640 containing 10% heat-inactivated FBS. Macrophages were cultured in complete media for at least 12 hours prior to using for experiments. Stellate cells were isolated according to standard protocol (17). All primary cells were isolated from male mice of 6 to 12 weeks of age were cultured at 37C° and in 5% CO2 with 100% humidity and were used for experiments at 60% to 80% confluence. Cells from N-Nitrosodiethylamine (DEN)-treated animals were obtained at 20 to 32 weeks of age.

DEN-induced HCC

A single dose of 10 μg/g body weight of DEN was administered intraperitoneally (i.p.) to 14-day-old male pups. The animals were monitored and euthanized at 32 weeks of age. Serum liver enzymes were analyzed in the Molecular Diagnostic Laboratory, Department of Pathology, VCU, using standard procedures.

RNA extraction, cDNA synthesis, and quantitative real-time PCR

Total RNA was extracted using the QIAGEN RNeasy Mini Kit (Qiagen; Cat# 74104). Two micrograms of RNA was used for cDNA synthesis using ABI cDNA Synthesis Kit (Applied Biosystems). qRT-PCR was performed using an ABI ViiA7 fast real-time PCR system, and Taqman gene expression assays using predesigned best coverage Taqman probes for standard gene expression (Applied Biosystems) according to the manufacturer's protocol. All mRNA levels were normalized by GAPDH mRNA levels.

RNA sequencing

RNA, extracted from livers of 3 adult mice per group, was used. RNA sequencing (RNA-Seq) library was prepared using the Illumina TruSeq RNA Sample Preparation Kit and sequenced on an Illumina HiSeq2000 platform. RNA-Seq libraries were pooled together to aim about 25 to 40 million read passed filtered reads per sample. All sequencing reads were aligned with their reference genome (UCSC mouse genome build mm9) using TopHat2. Bam files from alignment were processed using HTSeq-count to obtain the counts per gene in all samples. The counts for all samples were read into R software using DESeq package. For each condition a pairwise test was performed using the functions in DESeq and plot distributions were analyzed using reads per kilobase million (RPKM) values. Data were filtered on the basis of low count or low RPKM value (<40 percentile). Genes showing log2 fold change of >1.5 or <−1.5, FDR of <0.1 and P value of <0.05 were selected. A complete list of differentially regulated genes is available in GSE85427 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?token=ctoloyiiznwbtgt&acc=GSE85427)

In vitro antigen cross-presentation of bone marrow–derived dendritic cells and T-cell priming

Bone marrow-derived dendritic cells (BMDC) were isolated as described (18). Bone marrow–derived cells were differentiated into dendritic cells (DC) by culturing in RPMI-1640 medium containing granulocyte macrophage colony-stimulating factor (GM-CSF; 20 ng/mL) for 8 days. Day 8 Igfbp7+/+ and Igfbp7−/− BMDCs were loaded with 1 μg/mL gp100 (25-33a.a.) peptide for 3 hours, followed by 500 ng/mL lipopolysaccharide (LPS) for 2 hours (priming). LPS was washed before coculturing the BMDCs with pmel-17 T lymphocytes at a molar ratio of 1:10, 1:20, and 1:40 (DC:TC) for 60 hours. In a second experiment, gp100-loaded BMDCs were treated or not with rIGFBP7 protein (R&D) for 3 hours followed by OSI-906 (4 μmol/L) for 2 hours and LPS (500 ng/mL) for 2 hours. The cells were washed and cocultured with pmel-17 T cells for 60 hours. Supernatants were collected at 48 hours for ELISA. Cells were pulsed with 0.5 μCi/well [3H]-thymidine for the last 16 hours of incubation. Proliferation was measured by [3H]-thymidine incorporation in triplicate wells.

Establishment of stable Igfbp7-overexpressing clone in Hepa1-6 cells

The Hepa-1-6 cell line (CRL-1830) was obtained from ATCC and was cultured as instructed. Mouse Igfbp7 (NM_001159518) expression plasmid was obtained from Origene Technologies, Inc. (MR222256). Hepa1-6 cells were transfected with 8 μg of mIgfbp7 expression plasmid DNA and corresponding empty vector using FuGENE HD transfection reagent at a ratio of 3.5:1 (reagent:DNA). Transfected cells were selected in 450 μg/mL G418 for 4 weeks to establish BP7-OE pooled clone.

Xenograft studies in syngeneic mice

C57L/J mice were purchased from The Jackson Laboratory. Control and BP7-OE pooled clones of Hepa1-6 cells (4 × 106) were injected subcutaneously into the right dorsal flank of C57L/J mice. Five mice per group were used. Tumor volume was measured by the following formula: π/6 × (small diameter)2 × (large diameter). Mice were monitored twice a week and euthanized after 4 weeks, at which point the tumors were harvested. Immune cells infiltration was determined by FACS as described (19).

CD8+ and CD4+ depletion assays

Subcutaneous xenografts were established using BP7-OE clones in C57L mice. Depletion of CD8+ and CD4+ cells was performed by injecting 200 μg of neutralizing antibodies i.p. on days 1, 5, 12, and 19 as described (19). Tumors were collected 4 weeks after tumor implantation. Neutralizing antibodies were obtained from Bioxcell, CD8+ Ab (Clone: 2.43 catalog #BP0061), CD4+ Ab (Clone: GK1.5 catalog #BP0003-1).

IHC and immunofluorescence assays

IHC was performed on formalin-fixed paraffin-embedded (FFPE) sections as described (20) using the following antibodies: AFP (Santa Cruz Biotechnology: sc-15375), PCNA (Cell Signaling Technology; #13110), c-Myc (Cell Signaling Technology; #13987), CD31 (Dako; #JC70A), F4/80 (AbD Serotec; #MCA497RT), glutamine synthetase (Sigma; #G2781), and IGFBP7 (R&D Systems; #MAB21201). Immunofluorescence was performed on cells cultured in 4-chamber slides for hepatocytes (collagen-1 coated) and MEFs using antibodies against p65 (Cell Signaling Technology; #8242) and γ-H2AX (Cell Signaling Technology; #5438). The images were taken by a confocal laser scanning microscope. Antibody dilutions were used as recommended by the manufacturer.

Senescence-associated β-galactosidase assay

Hepatocytes were cultured for 8 days and senescence-associated β-galactosidase (SA-β-Gal) activity was measured as described (7).

Cell proliferation, BrdUrd incorporation, and colony formation assays

Cells (1 × 103) were plated in each well of a 96-well plate for measuring proliferation by a standard MTT assay (7). BrdUrd incorporation was measured using the BrdUrd Cell Proliferation Assay Kit (Cell Signaling Technology; #6813) according to the manufacturer's protocol. Colony formation assay was performed by plating 1 × 104 MEFs on 6-cm dishes and culturing for 4 weeks.

Western blotting analysis

Cell lysates and tissue extracts were prepared and Western blotting was performed as described (20). The primary antibodies used were (all from Cell Signaling Technology): p-AKT (#4060), AKT (#4685), pERK (#9101), and ERK (#4695). p-IGF1Rβ (Cell Signaling Technology; #6113), IGF1Rβ (Cell Signaling Technology; #9750), GAPDH (Santa Cruz Biotechnology; #sc-166545), EF1α (Millipore; #05-235), p-GSK3β (Cell Signaling Technology; #5558), GSK3β (Cell Signaling Technology; #12456), p65 NF-κB (Cell Signaling Technology; #5438), p-p65 (Cell Signaling Technology; #8242), cyclin D1 (Cell Signaling Technology; #2978), cyclin E1 (Cell Signaling Technology; #4129), arginase-1 (Santa Cruz Biotechnology; #sc-18355), Myc-tag (Cell Signaling Technology; #2276), p-STAT1 (Cell Signaling Technology; #9167), STAT1 (Cell Signaling Technology; #14994), Igfbp7 (R&D Systems Technology; #MAB21201) and β-actin (Sigma Aldrich; #A5316). Antibodies were diluted as recommended by the manufacturer. Densitometric analysis was performed by ImageJ software.

In vitro treatment of cells

Recombinant mouse IGF1 (R&D Systems; #Q8CAR0) was used at 20 ng/mL for signaling experiments and 100 ng/mL for cell proliferation experiments. LPS (Sigma Aldrich; #L3024) was used at 200 ng/mL. OSI-906 (Selleckchem #S1091) was used at 4 μmol/L. In vitro DEN was used at 10 ng/mL.

Cell-cycle analysis

MEFs (1 × 105) were synchronized in serum-free media for 18 hours and released into complete growth media. Cells were washed with PBS and fixed in 70% ethanol for 30 minutes at −20°C. The fixed cells were washed three times with PBS and were incubated with PBS containing 10 μg/mL RNase A for 30 minutes at 37°C, following which, cells were incubated with 30 μg/mL propidium iodide (PI) for 30 minutes in the dark. The samples were acquired in a FACSCanto II system. For each measurement 10,000 cells were acquired and the data were analyzed using the FACSDiva software.

Fluorescence-activated cell sorting analysis

To analyze immune cell population, livers from 5-month-old DEN-injected littermates were perfused and resuspended into a single-cell suspension as described (13). The cell suspension was subjected to fluorescence-activated cell sorting (FACS) analysis as described (19). The s.c. xenograft tumor tissues were digested with collagenase D (10 mg/mL) and DNase I (100 mg/mL), and cell suspensions were filtered through a 70-mm cell strainer prior to FACS analysis (19). Antibodies were obtained from Biolegend: CD4 (GK1.5), CD8 (2.43), INF-γ (XMG1.2), Ly6G (1A8), Ly6C (HK1.4), CD80 (16-10A1), CD86 (GL-1), CD11b (M1/70), CD11c (N418), F4/80 (BM8), CD3 (17A2), NK1.1 (PK136), B220 (RA3-6B2), and Gr1 (RB6-8C5).

NF-κB luciferase reporter assay

NF-κB luciferase reporter assay was performed in Igfbp7+/+ and Igfbp7−/− hepatocytes exactly as described (20). Experiments were performed in triplicates with two independent experiments.

Coinjection of HCC cells and bone marrow–derived macrophages in NOD scid gamma mice

Dihxy cells (1 × 106), developed from DEN-injected C57BL/6 mice in Dr. Karin's laboratory (21), were co-injected with bone marrow–derived macrophages (10 × 104 or 5 × 104) from Igfbp7+/+ and Igfbp7−/− mice subcutaneously in 7-week-old male NOD scid gamma (NSG) mice and tumor development were monitored for 7 weeks. Tumor volume was measured with calipers using the formula: (width)2 × length/2. For each group, 5 mice were used.

Statistical analysis

Data were presented as the mean ± SD and analyzed for statistical significance using two-tailed Student t test. For canonical pathway analysis, the P value was calculated using the right-tailed Fisher exact test.

Generation of Igfbp7−/− mouse

We generated Igfbp7−/− mice on C57BL/6 background by targeting promoter region and exon 1 of mIgfbp7 gene by a Cre-loxP strategy (Supplementary Fig. S1A–B and Fig. 1A–C). IHC staining of the Igfbp7+/+ liver showed more IGFBP7 staining in interstitial cells when compared with hepatocytes (Fig. 1C). qRT-PCR analysis demonstrated significantly more Igfbp7 mRNA expression in stellate cells and macrophages compared with hepatocytes (Fig. 1D). In liver and spleen, macrophage marker F4/80-positive cells showed positive staining for IGFBP7, further indicating that nonparenchymal cells are the major source of IGFBP7 (Supplementary Fig. S1C). The levels of other IGFBPs and IGF1 and IGF-2 mRNAs were similar in Igfbp7+/+ and Igfbp7−/− mice (Supplementary Fig. S1D).

Loss of Igfbp7 increases spontaneous tumorigenesis without affecting normal development

Igfbp7−/− mice were viable and fertile having normal litter size. Histological analysis of internal organs from 8 weeks old mice did not show any discernable difference between Igfbp7+/+ and Igfbp7−/− mice, indicating that Igfbp7 may not play a role in normal development (Supplementary Fig. S2A). However, when monitored for 24 months, Igfbp7−/− mice (3 of 10) developed spontaneous tumors in liver and lung, while no tumors were detected in Igfbp7+/+ mice (n = 7; Fig. 1E). The tumors in the liver showed complete loss of hepatic architecture, indicating HCC (Fig. 1E). At this age, splenic architecture was distorted in Igfbp7−/− mice compared with +/+. Additionally, increased infiltration of immune cells in the tissue microenvironment of liver, lung, and kidney was observed in Igfbp7−/− mice, suggesting increased inflammation (Fig. 1E). In 24-month-old liver and lung, increase in Il6 mRNA levels (marker of inflammation) was detected in both liver and lung tumors, but increase in α-feto protein (Afp) mRNA levels (marker of HCC) was observed only in liver tumors in Igfbp7−/− mice, indicating that the lung tumors are primary tumors and not metastatic lesions from the liver (Fig. 1F). Increased staining for glutamine synthetase (a marker for HCC) and increased infiltration of macrophages (determined by F4/80 staining) was observed in 24-month-old Igfbp7−/− liver sections compared with Igfbp7+/+ liver sections (Supplementary Fig. S2B–C). Staining of lung tumors for F4/80 identified macrophages in between large tumor cells and the tumors showed strong positive staining for the proliferation marker PCNA and increased expression of angiogenesis marker Vegf in Igfbp7−/− but not in Igfbp7+/+ (Supplementary Fig. S2D-F).

Loss of Igfbp7 increases proliferation and prevents senescence

We next checked the effect of genetic deletion of Igfbp7 on cell proliferation and senescence. Igfbp7−/− MEFs showed increased proliferation as measured by MTT and colony formation assays when compared with Igfbp7+/+ (Fig. 2A). Similarly, Igfbp7−/− hepatocytes showed increased proliferation by MTT assay and increased BrdUrd incorporation upon IGF1 treatment versus Igfbp7+/+ (Fig. 2B). Mouse hepatocytes do not proliferate in vitro and start undergoing senescence by 96 hours. At day 8 of culture, ∼80% of Igfbp7+/+ hepatocytes showed senescence versus only ∼30% Igfbp7−/− hepatocytes (Fig. 2C and Supplementary Fig. S3A). As a corollary, γ-H2AX–positive nuclei were significantly reduced in Igfbp7−/− MEFs after 12 hours of serum starvation (Fig. 2D and Supplementary Fig. S3B). To check the effect of Igfbp7 deletion on cell-cycle progression, MEFs were synchronized in serum-free media for 24 hours, following which, they were released in complete growth media and subjected to cell-cycle analysis. At 18 and 36 hours after release, there was a significant decrease in cells in G1 phase and a corresponding increase in cells in G2–M phase in Igfbp7−/− MEFs compared with Igfbp7+/+ MEFs, indicating that Igfbp7−/− MEFs cycle faster than their wild-type counterparts (Fig. 2E).

Loss of Igfbp7 activates the IGF1 signaling pathway

Igfbp7+/+and Igfbp7−/− hepatocytes were treated with recombinant mouse IGF1 (20 ng/mL). Constitutive activation of IGF1R and downstream Akt and GSK3β was observed in Igfbp7−/− hepatocytes, and there was a further increase in IGF1R, Akt, and GSK3β activation in Igfbp7−/− hepatocytes, compared with Igfbp7+/+, at 5 minutes of exposure to IGF1, suggesting that Igfbp7-null hepatocytes have increased propensity for activation of the IGF1 pathway (Fig. 2F). In both cell types, activation persisted until 60 minutes. Analysis of livers of adult mice also showed increased activation of Akt and its downstream target GSK3β, further confirming constitutive activation of IGF1 signaling in Igfbp7−/− mice (Fig. 2G and Supplementary Fig. S4A). Constitutive activation of Akt, GSK3β, and ERK was also observed in Igfbp7−/− hepatocytes, macrophages, and MEFs compared with Igfbp7+/+ (Fig. 2H and Supplementary Fig. S4B). Because Igfbp7−/− MEFs cycle faster than Igfbp7+/+ MEFs, we checked the levels of cyclin D1 (CCND1) and cyclin E1 (CCNE1), which facilitate G1 to S cell-cycle progression. Increased levels of CCND1 and CCNE1 were observed in Igfbp7−/− MEFs versus Igfbp7+/+, thus explaining the accelerated cell-cycle progression in the former (Fig. 2H and Supplementary Fig. S4B). We next checked how constitutive activation of IGF1 signaling pathway modulates response of Igfbp7−/− MEFS to OSI-906, a dual kinase inhibitor of IGF1R and insulin receptor (INSR). Igfbp7−/− MEFS were more sensitive to growth inhibition by OSI-906 compared with Igfbp7+/+, suggesting that Igfbp7−/− MEFs might have preferential addiction to IGF signaling (Fig. 2I). Overall, these results indicate that activation of the IGF1 signaling pathway contributes to increased proliferation upon Igfbp7 deletion.

Loss of Igfbp7 results in markedly accelerated DEN-induced HCC

Because aged Igfbp7−/− mice develop spontaneous HCC, we further analyzed HCC development upon exposure to the hepatocarcinogen DEN, a well-established HCC model (22, 23). HCC was induced in Igfbp7+/+, +/− and −/− littermates by a single i.p. injection of DEN (10 μg/g) when the mice were two weeks of age. At 32 weeks, Igfbp7−/− mice demonstrated a marked increase in both number and size of hepatic nodules when compared with the other genotypes (Fig. 3A–B). This increase in hepatic nodules contributed to increased liver-to-body-weight ratio in Igfbp7−/− mice (Fig. 3C). Liver enzymes, aspartate aminotransferase (AST) and alanine aminotransferase (ALT), and serum total protein were significantly increased in Igfbp7−/− mice versus +/+ and +/− mice (Fig. 3D–E). Histological analysis of the liver showed extensive abnormality in liver architecture showing pleomorphic, hyperchromatic nuclei with increased infiltration of immune cells, and abnormal blood vessel formation (Fig. 3F and Supplementary Fig. S5A). Increased staining for AFP, proliferating cell nuclear antigen (PCNA; proliferation marker), c-MYC, CD31 (angiogenesis marker), and F4/80 (macrophage marker) was observed in Igfbp7−/− liver sections when compared with Igfbp7+/+, indicating increased proliferation, angiogenesis, inflammation, and development of frank HCC (Fig. 3F). Increased staining for glutamine synthetase was also observed in DEN-treated Igfbp7−/− livers compared with Igfbp7+/+ (Supplementary Fig. S5B). To exclude the possibility that the increased HCC development in Igfbp7−/− mice is because of increased sensitivity to DEN, we analyzed liver functions 48 hours after injection of DEN. Both Igfbp7+/+ and Igfbp7−/− littermates showed similar levels of changes in AST, ALT, and alkaline phosphatase (ALP), a marker of acute liver damage, suggesting that littermates of both genotypes respond similarly to DEN-induced liver injury (Supplementary Fig. S5C). DEN needs to be metabolically activated by CYP2E1 in zone three hepatocytes. Cyp2e1 mRNA levels were similar in Igfbp7+/+ and Igfbp7−/− livers (data not shown), suggesting that altered metabolism of DEN does not underlie differential tumorigenic response.

Igfbp7 deletion creates an inflammatory and immunosuppressive tumor microenvironment

DEN-induced HCC is an inflammatory-type of HCC in which damaged hepatocytes release cytokines, mainly IL1β, which stimulate Kupffer cells to release IL6 and TNFα by activating NF-κB signaling (24–26). In this context, we analyzed the immune cell profile in Igfbp7+/+ and Igfbp7−/− mice by FACS. Examination of 8-week-old naïve liver and spleen showed no significant difference in CD8+ and CD4+ T cells, natural killer (NK) cells, B cells, DCs, splenic and liver macrophages, and myeloid-derived suppressor cells (MDSC), indicating that IGFBP7 may not be required for immune cell development (Supplementary Fig. S6A–B). However, upon examining the livers of DEN-treated mice at 20 weeks of age, a time point before the development of overt hepatic nodules but having microscopic tumors as evidenced by increased AFP and c-Myc expression (Supplementary Fig. S6C), we found a significant increase in Kupffer cells (CD11bhigh/F4/80low), infiltrating macrophages (CD11blow/F4/80high) and CD11b+ monocytes (Ly6G/Ly6C+) in Igfbp7−/− livers compared with +/+ livers (Fig. 4A). These results suggest that Igfbp7−/− mice have an inflammatory and immunosuppressive tumor microenvironment. M-MDSCs mediate immune suppressive function through l-arginine depletion by increasing expression of arginase-1 (ARG1) that abrogate antitumor responses by T cells (27). Indeed, DEN-treated Igfbp7−/− livers showed substantially increased levels of ARG1 compared with Igfbp7+/+ (Supplementary Fig. S6D). This increased immune suppressive environment might contribute to a decrease in liver-infiltrating NK cells as well as CD4+ and CD8+ T cells (Fig. 4A).

The inflammatory environment in Igfbp7−/− cells was further checked at gene expression levels. Il6, Il1b, Tnfa, and c-Myc (marker of proliferation) mRNAs showed significantly increased levels in the livers of DEN-treated Igfbp7−/− livers when compared with Igfbp7+/+ livers at 32 weeks (Fig. 4B). Igfbp7−/− hepatocytes, isolated at 24 weeks after DEN treatment, and Igfbp7−/− Kupffer cells, isolated at 5 weeks after DEN treatment, also showed increased levels of Il6, Il1b, and Tnfa versus Igfbp7+/+ (Fig. 4B). Primary hepatocytes treated in vitro with DEN for 24 hours showed significantly increased levels of Il1b and Tnfa, but not Il6, in Igfbp7−/− compared with Igfbp7+/+, indicating that Igfbp7−/− hepatocytes respond more to DEN treatment in terms of generating cytokines, leading to a robust activation of Kupffer cells generating Il6 (Fig. 4C, top). We treated macrophages with IL1β (10 ng/mL) and measured Il6 and TNFa mRNA levels. Basal Il6 and TNFa levels were significantly higher in Igfbp7−/− macrophages versus +/+ (Fig. 4C, bottom). IL1β treatment induced Il6 and TNFa mRNA expression in both Igfbp7+/+ and Igfbp7−/− macrophages. However, the magnitude of induction was significantly more in Igfbp7−/− macrophages when compared with Igfbp7+/+. While inflammatory cytokine levels were similar in 6-week-old Igfbp7+/+ and Igfbp7−/− livers, they were higher in 24-week-old Igfbp7−/− livers compared with Igfbp7+/+, indicating that Igfbp7 deficiency with age creates a precancerous milieu for development of spontaneous HCC (Supplementary Fig. S7A).

It is well-established that NF-κB is the pivotal regulator of inflammation-driven HCC and other cancers (26, 28). Under basal condition, p50/p65 NF-κB resides in the cytoplasm and upon stimulation translocates to the nucleus. In Igfbp7+/+ hepatocytes, p65 NF-κB was located exclusively in the cytoplasm, while some nuclear staining of p65 was detected in Igfbp7−/− hepatocytes (Fig. 4D). Upon stimulation with LPS, p65 translocated to the nucleus in both cells. However, the magnitude of translocation was more robust in Igfbp7−/− hepatocytes compared with Igfbp7+/+. To quantify these changes, we measured NF-κB luciferase reporter activity in Igfbp7+/+ and Igfbp7−/− primary hepatocytes. Both basal and LPS-induced luciferase activity was significantly increased in Igfbp7−/− hepatocytes when compared with Igfbp7+/+ (Fig. 4E). Igfbp7−/− hepatocytes and bone marrow–derived macrophages showed increased levels of phosphorylated p65 upon LPS treatment, further confirming that Igfbp7 deletion results in activation of NF-κB (Fig. 4F and Supplementary Fig. S7B).

We checked whether Igfbp7 status in the macrophages affects tumorigenesis by HCC cells. For this purpose, we injected mouse HCC cells Dihxy (21) either alone or with Igfbp7+/+ or Igfbp7−/− bone marrow–derived macrophages at a tumor:macrophage ratio of 10:1 or 20:1 s.c. in NSG mice and monitored tumor development. Coinjection of Igfbp7−/− macrophages significantly stimulated tumorigenesis by Dihxy cells compared with Igfbp7+/+ cells, indicating that Igfbp7 loss intrinsically changes macrophage properties, rendering them more tumor promoting (Supplementary Fig. S7C). We cultured bone marrow–derived macrophages either in serum-free medium (control) or in conditioned media (CM) from Dihxy cells and analyzed expression of Il6 (marker for M1 macrophage) and Arg1 (marker for M2 macrophage; Supplementary Fig. S7D). Dihxy CM induced Il6 and Arg1 expression in both Igfbp7+/+ and Igfbp7−/− macrophages. However, this induction was more robust in Igfbp7−/− macrophages, especially for Arg1, indicating that Igfbp7−/− macrophages have a propensity for polarization to M2 subtype, which promotes angiogenesis and tissue remodeling as well as immunosuppression (29–31).

RNA-Seq identifies inhibition of immune surveillance in Igfbp7−/− mice

To obtain insights into the gene expression changes facilitating HCC in Igfb7−/− mice, we performed RNA-Seq using total RNA from naïve liver samples from 8-week-old Igfbp7+/+ and −/− mice (n = 3). Genes showing absolute fold change of >1.5, FDR of <0.1 and P value of <0.05 were selected. Genes (473) showed differential changes, out of which 209 were upregulated and 264 were downregulated in Igfbp7−/− livers when compared with Igfbp7+/+ (Supplementary Table S1). We performed biological processes and pathway analysis using the softwares DAVID and Ingenuity. Surprisingly, the topmost biological pathways identified to be inhibited in Igfbp7−/− livers are all associated with immune surveillance, such as antigen presentation or DCs maturation (Fig. 5A and Supplementary Table S2). The genes regulating antigen presentation that are downregulated in Igfbp7−/− liver include MHC class I and II genes, proteasome components LMP2 and LMP7 that process the antigenic peptides, Transporter associated with Antigen Processing 1 (Tap1) that transports the processed peptide to the ER for assembly with MHC molecules, and costimulators CD80 and CD86. Downregulation of Tap1 mRNA in Igfbp7−/− liver was confirmed by Taqman qRT-PCR (Fig. 5B). Mature DCs present high cell surface expression of CD80 and CD86 (32). Downregulation of CD86 and CD80 in Igfbp7−/− CD11b+CD11c+ BMDCs were confirmed by flow cytometry (Fig. 5C). Mean fluorescence intensity (MFI) for CD86 was 138 and 155, while MFI for CD80 was 240 and 165, respectively for Igfbp7+/+ and Igfbp7−/− DCs. Because there was a discrepancy between percentage of cell surface expression versus MFI for CD86, we confirmed these observations by biological assays. We measured antigen presentation capacity of BMDCs from 6-week-old Igfbp7+/+ and −/− littermates. DCs, loaded with a tumor antigen gp100 peptide, was used to stimulate T cells from pmel-17 mice, a transgenic mouse expressing rearranged T cell receptor (TCR) recognizing gp100. T-cell activation by Igfbp7−/− DCs was significantly blunted when compared with that by Igfbp7+/+ DCs as evidenced by decreased proliferation of T cells by [3H]-thymidine incorporation assay and decreased production of IL12 and IL2 in the conditioned media (Fig. 5D). The inhibition in antigen presentation by Igfbp7−/− DCs could be rescued by the addition of recombinant IGFBP7 protein (rIGFBP7) or treatment with OSI-906 (Fig. 5E). In Igfbp7−/− DCs, cotreatment of OSI-906 and rIGFBP7 did not further augment T-cell proliferation or IFNγ production when compared with rIGFBP7 treatment alone, although a significant increase in IL2 production was observed. These findings suggest a possible IGF1R-independent function of IGFBP7. Treatment of Igfbp7+/+ DCs with rIGFBP7 did not modulate T-cell proliferation or IFNγ or IL2 production in antigen presentation assay, suggesting a lack of activation of IGF signaling in these cells, which may not be affected by rIGFBP7 or OSI-906 (Supplementary Fig. S8A). Antigen presentation is positively regulated by IFNγ that functions by stimulating the JAK–STAT1 pathway. Upon LPS treatment a substantial inhibition in STAT1 activation was observed in Igfbp7−/− DCs versus Igfbp7+/+ DCs, indicating that the IFNγ signaling pathway is interfered in Igfbp7−/− DCs (Fig. 5F and S8B). To check potential role of Igfbp7 in regulating NK cell function splenocytes from Igfbp7+/+ and Igfbp7−/− mice were cultured in the presence of NK cell activating cytokines (IL2 and IL15) for 48 hours and granzyme B and IFNγ production from NK cells (NK1.1+CD3) was assayed by flow cytometry. No significant difference in NK cell activation was observed between Igfbp7+/+ and Igfbp7−/−, indicating that Igfbp7 does not affect intrinsic functions of NK cells (Supplementary Fig. S8C).

Overexpression of Igfbp7 decreases tumor growth by activating an antitumor immune response

As yet, all studies analyzing the effect of IGFBP7, either recombinant protein, stable overexpression, or via an adenovirus, have been performed in athymic nude mice, thereby not detecting immunomodulatory properties of IGFBP7 (7). To examine the impact of IGFBP7 on tumor immune environment, we established stable pooled clone of mouse HCC Hepa1-6 cells overexpressing mIGFBP7 (BP7-OE). Overexpression of secreted mIGFBP7 in BP7-OE clones was confirmed by Western blot analysis in the conditioned media and by qRT-PCR in the cells (Fig. 6A). Subcutaneous xenografts from BP7-OE cells grew significantly slower in syngeneic C57L mice, from which Hepa1-6 cells were established, compared with the control tumors (Fig. 6B). BP7-OE tumors showed decreased activation of Akt and ERK and marked decrease in CD31 staining, confirming inhibition of IGF1 signaling and angiogenesis (Fig. 6C and Fig. S9A-B). Interestingly, BP7-OE tumors presented with marked tumor infiltration of IFNγ–producing CD8+ or CD4+ T cells (Fig. 6D), marked elevation of cytokine IL12 (Fig. 6E), which is crucial for antitumor Th1 immunity, moderate elevation of IFNγ (Fig. 6E) and decrease in MDSCs in the tumors (Fig. 6D). These tumors also showed significant upregulation of Tap1 (Fig. 6E). Collectively these findings in BP7-OE tumors reflect reverse findings of what is observed in Igfbp7−/− mice. To further confirm the role of induction of antitumor immunity, we established BP7-OE tumors in C57L mice and depleted CD8+ and CD4+ cells with neutralizing antibody. The depletion of CD8+ or CD4+ cells was confirmed by FACS analysis of the tumor (Supplementary Fig. S9C). Depletion of CD8+ and CD4+ cells significantly rescued the growth of BP7-OE tumors, supporting a major role of immune activation in mediating tumor suppressor function of IGFBP7 (Fig. 6F).

Previous studies in multiple tumor models, including HCC, have identified IGFBP7 as a potential tumor suppressor (7–9). However, an in vivo model is necessary to conclusively establish the tumor suppressor function of IGFBP7 and the present studies fulfill this need. We document that Igfbp7−/− mice develop spontaneous tumors in multiple organs, although not at 100% penetrance, and present with a proinflammatory milieu that might provide a fertile ground for the sustenance and progression of cancers once cells become transformed following a mutagenic event. Indeed, hepatocarcinogenesis following DEN treatment is markedly accelerated in Igfbp7−/− mice compared with Igfbp7+/+ and Igfbp7+/− mice, thereby establishing IGFBP7 as a bona fide tumor suppressor according to the two-hit hypothesis in which both alleles need to be mutated to manifest the tumorigenic phenotype.

The loss of Igfbp7 resulted in constitutive activation of the IGF pathway in parenchymal cells, such as hepatocytes and MEFs, which translated into increased proliferation, accelerated cell-cycle progression, and inhibition of senescence. Activation of the IGF1 pathway protects from stress-induced premature senescence, IGFBP7 is induced in senescent cells and loss of IGFBP7 mediates escape from oncogene-induced senescence (8, 33, 34). In DEN-induced HCC model, DEN causes DNA damage, induces reactive oxygen species (ROS), and ultimately apoptosis in the hepatocytes (25). However, there is also compensatory proliferation to mitigate this effect. Upon activation of survival pathways, such as IGF1, the damaged (and mutated) hepatocytes do not undergo apoptosis and continue to survive and proliferate, resulting in expansion of mutated, transformed cells. Induction of senescence in premalignant cells and clearing of these senescent cells by immune system serve as a mechanism of inhibition of HCC (35). Thus, deleting Igfbp7 in hepatocytes confers both proliferative and survival advantage and protection from senescence (Fig. 6G). On the other hand, we observed that the expression of IGFBP7 in macrophages is significantly more than that in hepatocytes, suggesting an important role of IGFBP7 in modulating liver microenvironment. Igfbp7−/− macrophages show activation of NF-κB and increased expression of proinflammatory cytokines, thereby establishing a chronic inflammatory environment in which senescence-resistant transformed hepatocytes might thrive. Activated Akt, following activation of IGF1R, phosphorylates IKK, leading to activation of the NF-κB pathway (36). Thus, activation of IGF signaling might contribute to both proliferative and proinflammatory phenotypes (Fig. 6G).

Cross-talk between hepatocytes and macrophages is fundamental in HCC development. Macrophages require secreted factors, such as IL1β, released from damaged hepatocytes, for activation, while damaged hepatocytes require macrophage-released factors, such as IL6, for survival and proliferation and NF-κB play an important role in regulating this cross-talk (26, 37, 38). The tumorigenic effects of NF-κB in HCC are highly dependent on cell type. For instance, hepatocyte-specific knockout of IKKβ promoted DEN-induced HCC via increased ROS and hepatocyte compensatory proliferation, while double knockout of IKKβ in the macrophage and hepatocyte reduced DEN-induced HCC (25). In Igfbp7−/− mice, NF-κB is activated in both hepatocytes and macrophages, thereby creating a situation that is mirror image of IKKβ double knockout mice. The simultaneous activation of NF-κB in our model explains the concomitant increase in inflammatory cytokines, such as IL6, IL1β, and TNFα, and the subsequent increase in the Kupffer cell population in DEN-induced HCC.

One surprising finding deduced from our model is the ability of IGFBP7 to modulate the antigen presentation machinery and thereby an antitumor immune response. We document that deletion of Igfbp7 resulted in downregulation of an IFNγ-regulated cluster of genes regulating antigen presentation. Indeed, we observe decreased activation of IFNγ signaling in Igfbp7−/− BMDCs, which functionally translated into decreased antigen presentation and decreased infiltration of CD8+ and CD4+ T cells and NK cells in DEN-induced tumor. Treatment of Igfbp7−/− BMDCs with rIGFBP7 protein restored antigen presentation capacity in BMDCs documenting a key role of IGFBP7 in proper functioning of antigen-presenting cells. As a corollary, overexpression of Igfbp7 in mouse HCC cells inhibited tumorigenesis in syngeneic mice with robust infiltration of CD8+ and CD4+ T cells and depletion of CD8+ and CD4+ T cells facilitated tumor growth. While the immunomodulatory effects of IGF1 are less understood, several lines of evidence supports the potential involvement of IGF signaling in regulating tumor immunogenicity and immune cell phenotype or function. Silencing IGF1 expression in Hepa1-6 cells increased its immunogenicity and decreased tumorigenicity by upregulation of MHC class I molecules and mobilization of T cells (39). In T cells, IGF1 treatment results in decreased surface expression of IFNγR2 and inhibits IFNγ–STAT1 signaling (40). Additionally, DCs, upon treatment with IGF1, showed a defect in maturation following LPS stimulation and produced immune suppressive cytokines such as IL10 (41). Indeed, we observe an increase in arginase-1 and MDSCs in DEN-induced HCC in Igfbp7−/− mice and a decrease in MDSCs in BP7-OE tumors. Thus, Igfbp7 deletion creates an immunosuppressive environment preventing clearance of transformed hepatocytes by the immune system (Fig. 6G).

In summary, we document that IGFBP7 can not only inhibit cancer cells but also modulate tumor microenvironment and this dual effect might have a lasting effect in inhibiting both primary tumors and distant metastasis. Even though HCC has an immunosuppressive milieu, immune-targeted therapies are beginning to demonstrate significant objective responses in clinical trials. Targeted delivery of rIGFBP7 protein might be an effective therapeutic for HCC and other cancers.

No potential conflicts of interest were disclosed.

Conception and design: M. Akiel, D. Rajasekaran, M.A. Subler, J. Windle, X.-Y. Wang, D. Sarkar

Development of methodology: M. Akiel, D. Sarkar

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M. Akiel, C. Guo, X. Li, D. Rajasekaran, R.G. Mendoza, C.L. Robertson, N. Jariwala, F. Yuan, M.A. Subler, J. Windle, D.K. Garcia, Z. Lai, S. Giashuddin, D. Sarkar

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M. Akiel, X. Li, D. Rajasekaran, C.L. Robertson, H.-I.H. Chen, Y. Chen, P.B. Fisher, X.-Y. Wang, D. Sarkar

Writing, review, and/or revision of the manuscript: M. Akiel, D. Rajasekaran, M.A. Subler, J. Windle, P.B. Fisher, X.-Y. Wang, D. Sarkar

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M. Akiel, R.G. Mendoza, Y. Chen, S. Giashuddin, D. Sarkar

Study supervision: M. Akiel, X.-Y. Wang, D. Sarkar

P.B. Fisher holds the Thelma Newmeyer Corman Chair in Cancer Research and is a Samuel Waxman Cancer Research Foundation (SWCRF) Investigator. X.-Y. Wang is the Mary Anderson Harrison Distinguished Professor in Cancer Research in VCU MCC. D. Sarkar is the Harrison Foundation Distinguished Professor in Cancer Research in VCU MCC.

The present study was supported in part by NCI grant R21 CA183954, National Institute of Diabetes and Digestive and Kidney Diseases Grant R01 DK107451 and VCU Massey Cancer Center (MCC) Pilot Project Grant (D. Sarkar), and R01 CA175033 and R01 CA154708 (X-Y. Wang). C.L. Robertson is supported by a National Institute of Diabetes And Digestive and Kidney Diseases Grant T32DK007150. Services in support of this project were provided by the VCU Massey Cancer Center Transgenic/Knock-out Mouse Facility and Flow Cytometry Facility, supported in part with funding from NIH-NCI Cancer Center Support Grant P30 CA016059.

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.
Swann
JB
,
Smyth
MJ
. 
Immune surveillance of tumors
.
J Clin Invest
2007
;
117
:
1137
46
.
2.
Mittal
D
,
Gubin
MM
,
Schreiber
RD
,
Smyth
MJ
. 
New insights into cancer immunoediting and its three component phases–elimination, equilibrium and escape
.
Curr Opin immunol
2014
;
27
:
16
25
.
3.
Neefjes
J
,
Jongsma
ML
,
Paul
P
,
Bakke
O
. 
Towards a systems understanding of MHC class I and MHC class II antigen presentation
.
Nat Rev Immunol
2011
;
11
:
823
36
.
4.
Makarova-Rusher
OV
,
Medina-Echeverz
J
,
Duffy
AG
,
Greten
TF
. 
The yin and yang of evasion and immune activation in HCC
.
J Hepatol
2015
;
62
:
1420
9
.
5.
Scharf
JG
,
Braulke
T
. 
The role of the IGF axis in hepatocarcinogenesis
.
Hormone Metab Res
2003
;
35
:
685
93
.
6.
Evdokimova
V
,
Tognon
CE
,
Benatar
T
,
Yang
W
,
Krutikov
K
,
Pollak
M
, et al
IGFBP7 binds to the IGF-1 receptor and blocks its activation by insulin-like growth factors
.
Sci Signal
2012
;
5:ra92
.
7.
Chen
D
,
Yoo
BK
,
Santhekadur
PK
,
Gredler
R
,
Bhutia
SK
,
Das
SK
, et al
Insulin-like growth factor-binding protein-7 functions as a potential tumor suppressor in hepatocellular carcinoma
.
Clin Cancer Res
2011
;
17
:
6693
701
.
8.
Wajapeyee
N
,
Serra
RW
,
Zhu
X
,
Mahalingam
M
,
Green
MR
. 
Oncogenic BRAF induces senescence and apoptosis through pathways mediated by the secreted protein IGFBP7
.
Cell
2008
;
132
:
363
74
.
9.
Sprenger
CC
,
Damon
SE
,
Hwa
V
,
Rosenfeld
RG
,
Plymate
SR
. 
Insulin-like growth factor binding protein-related protein 1 (IGFBP-rP1) is a potential tumor suppressor protein for prostate cancer
.
Cancer Res
1999
;
59
:
2370
5
.
10.
Li
F
,
Fan
YC
,
Gao
S
,
Sun
FK
,
Yang
Y
,
Wang
K
. 
Methylation of serum insulin-like growth factor-binding protein 7 promoter in hepatitis B virus-associated hepatocellular carcinoma
.
Genes Chromosomes Cancer
2014
;
53
:
90
7
.
11.
Wajapeyee
N
,
Kapoor
V
,
Mahalingam
M
,
Green
MR
. 
Efficacy of IGFBP7 for treatment of metastatic melanoma and other cancers in mouse models and human cell lines
.
Mol Cancer Ther
2009
;
8
:
3009
14
.
12.
Chen
D
,
Siddiq
A
,
Emdad
L
,
Rajasekaran
D
,
Gredler
R
,
Shen
XN
, et al
Insulin-like growth factor-binding protein-7 (IGFBP7): a promising gene therapeutic for hepatocellular carcinoma (HCC)
.
Mol Ther
2013
;
21
:
758
66
.
13.
Bissell
DM
,
Guzelian
PS
. 
Degradation of endogenous hepatic heme by pathways not yielding carbon monoxide. Studies in normal rat liver and in primary hepatocyte culture
.
J Clin Invest
1980
;
65
:
1135
40
.
14.
Bissell
DM
,
Guzelian
PS
. 
Degradation of endogenous hepatic heme by pathways not yielding carbon monoxide. Studies in normal rat liver and in primary hepatocyte culture
.
J Clin Invest
1980
;
65
:
1135
40
.
15.
Xu
J
. 
Preparation, culture, and immortalization of mouse embryonic fibroblasts
.
Current Protoc Mol Biol
2005
;
Chapter 28:Unit 28 1
.
16.
Zhang
X
,
Goncalves
R
,
Mosser
DM
. 
The isolation and characterization of murine macrophages
.
Current Protoc Immunol
2008
;
Chapter 14:Unit 14 1
.
17.
Maschmeyer
P
,
Flach
M
,
Winau
F
. 
Seven steps to stellate cells
.
J Vis Exp
2011
;
51
:
2710
.
18.
Guo
C
,
Yi
H
,
Yu
X
,
Hu
F
,
Zuo
D
,
Subjeck
JR
, et al
Absence of scavenger receptor A promotes dendritic cell-mediated cross-presentation of cell-associated antigen and antitumor immune response
.
Immunol Cell Biol
2012
;
90
:
101
8
.
19.
Yu
X
,
Guo
C
,
Yi
H
,
Qian
J
,
Fisher
PB
,
Subjeck
JR
, et al
A multifunctional chimeric chaperone serves as a novel immune modulator inducing therapeutic antitumor immunity
.
Cancer Res
2013
;
73
:
2093
103
.
20.
Robertson
CL
,
Srivastava
J
,
Siddiq
A
,
Gredler
R
,
Emdad
L
,
Rajasekaran
D
, et al
Genetic deletion of AEG-1 prevents hepatocarcinogenesis
.
Cancer Res
2014
;
74
:
6184
93
.
21.
He
G
,
Dhar
D
,
Nakagawa
H
,
Font-Burgada
J
,
Ogata
H
,
Jiang
Y
, et al
Identification of liver cancer progenitors whose malignant progression depends on autocrine IL-6 signaling
.
Cell
2013
;
155
:
384
96
.
22.
Lee
JS
,
Chu
IS
,
Mikaelyan
A
,
Calvisi
DF
,
Heo
J
,
Reddy
JK
, et al
Application of comparative functional genomics to identify best-fit mouse models to study human cancer
.
Nat Genet
2004
;
36
:
1306
11
.
23.
Lim
IK
. 
Spectrum of molecular changes during hepatocarcinogenesis induced by DEN and other chemicals in Fischer 344 male rats
.
Mech Ageing Dev
2002
;
123
:
1665
80
.
24.
Sakurai
T
,
He
G
,
Matsuzawa
A
,
Yu
GY
,
Maeda
S
,
Hardiman
G
, et al
Hepatocyte necrosis induced by oxidative stress and IL-1 alpha release mediate carcinogen-induced compensatory proliferation and liver tumorigenesis
.
Cancer Cell
2008
;
14
:
156
65
.
25.
Maeda
S
,
Kamata
H
,
Luo
JL
,
Leffert
H
,
Karin
M
. 
IKKbeta couples hepatocyte death to cytokine-driven compensatory proliferation that promotes chemical hepatocarcinogenesis
.
Cell
2005
;
121
:
977
90
.
26.
Pikarsky
E
,
Porat
RM
,
Stein
I
,
Abramovitch
R
,
Amit
S
,
Kasem
S
, et al
NF-kappaB functions as a tumour promoter in inflammation-associated cancer
.
Nature
2004
;
431
:
461
6
.
27.
Fletcher
M
,
Ramirez
ME
,
Sierra
RA
,
Raber
P
,
Thevenot
P
,
Al-Khami
AA
, et al
l-Arginine depletion blunts antitumor T-cell responses by inducing myeloid-derived suppressor cells
.
Cancer Res
2015
;
75
:
275
83
.
28.
Ben-Neriah
Y
,
Karin
M
. 
Inflammation meets cancer, with NF-kappaB as the matchmaker
.
Nat Immunol
2011
;
12
:
715
23
.
29.
Qian
BZ
,
Pollard
JW
. 
Macrophage diversity enhances tumor progression and metastasis
.
Cell
2010
;
141
:
39
51
.
30.
Mantovani
A
,
Allavena
P
,
Sica
A
. 
Tumour-associated macrophages as a prototypic type II polarised phagocyte population: role in tumour progression
.
Eur J Cancer
2004
;
40
:
1660
7
.
31.
Lewis
CE
,
Pollard
JW
. 
Distinct role of macrophages in different tumor microenvironments
.
Cancer Res
2006
;
66
:
605
12
.
32.
Banchereau
J
,
Briere
F
,
Caux
C
,
Davoust
J
,
Lebecque
S
,
Liu
YJ
, et al
Immunobiology of dendritic cells
.
Ann Rev Immunol
2000
;
18
:
767
811
.
33.
Gruber
HE
,
Hoelscher
GL
,
Ingram
JA
,
Bethea
S
,
Hanley
EN
. 
IGF-1 rescues human intervertebral annulus cells from in vitro stress-induced premature senescence
.
Growth Factors
2008
;
26
:
220
5
.
34.
Swisshelm
K
,
Ryan
K
,
Tsuchiya
K
,
Sager
R
. 
Enhanced expression of an insulin growth factor-like binding protein (mac25) in senescent human mammary epithelial cells and induced expression with retinoic acid
.
Proc Natl Acad Sci USA
1995
;
92
:
4472
6
.
35.
Kang
TW
,
Yevsa
T
,
Woller
N
,
Hoenicke
L
,
Wuestefeld
T
,
Dauch
D
, et al
Senescence surveillance of pre-malignant hepatocytes limits liver cancer development
.
Nature
2011
;
479
:
547
51
.
36.
Mitsiades
CS
,
Mitsiades
N
,
Poulaki
V
,
Schlossman
R
,
Akiyama
M
,
Chauhan
D
, et al
Activation of NF-kappaB and upregulation of intracellular anti-apoptotic proteins via the IGF-1/Akt signaling in human multiple myeloma cells: therapeutic implications
.
Oncogene
2002
;
21
:
5673
83
.
37.
Karin
M
,
Cao
Y
,
Greten
FR
,
Li
ZW
. 
NF-kappaB in cancer: from innocent bystander to major culprit
.
Nat Rev Cancer
2002
;
2
:
301
10
.
38.
Finkin
S
,
Yuan
D
,
Stein
I
,
Taniguchi
K
,
Weber
A
,
Unger
K
, et al
Ectopic lymphoid structures function as microniches for tumor progenitor cells in hepatocellular carcinoma
.
Nat Immunol
2015
;
16
:
1235
44
.
39.
Liu
Y
,
Wang
H
,
Zhao
J
,
Ma
J
,
Wei
L
,
Wu
S
, et al
Enhancement of immunogenicity of tumor cells by cotransfection with genes encoding antisense insulin-like growth factor-1 and B7.1 molecules
.
Cancer Gene Ther
2000
;
7
:
456
65
.
40.
Bernabei
P
,
Bosticardo
M
,
Losana
G
,
Regis
G
,
Di Paola
F
,
De Angelis
S
, et al
IGF-1 down-regulates IFN-gamma R2 chain surface expression and desensitizes IFN-gamma/STAT-1 signaling in human T lymphocytes
.
Blood
2003
;
102
:
2933
9
.
41.
Huang
CT
,
Chang
MC
,
Chen
YL
,
Chen
TC
,
Chen
CA
,
Cheng
WF
. 
Insulin-like growth factors inhibit dendritic cell-mediated anti-tumor immunity through regulating ERK1/2 phosphorylation and p38 dephosphorylation
.
Cancer Lett
2015
;
359
:
117
26
.