Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Oct 1.
Published in final edited form as: Endocr Relat Cancer. 2021 Feb;28(2):135–149. doi: 10.1530/ERC-20-0337

B7 immune-checkpoints as targets for the treatment of neuroendocrine tumors

Ziqiang Yuan 1,2, Juliet C Gardiner 2, Elaine C Maggi 2,3, Shuyu Huang 4,5, Asha Adem 2, Guiying Li 6, Sylvia Lee 2, Daniel Slegowski 2, Alyssa Exarchakis 2, James R Howe 6, Edmund C Lattime 2, Xingxing Zang 4,5, Steven K Libutti 2
PMCID: PMC8486311  NIHMSID: NIHMS1740848  PMID: 33410766

Abstract

The B7 family, and their receptors, the CD28 family, are major immune checkpoints that regulate T-cell activation and function. In the present study, we explore the role of two B7 immune-checkpoints: HHLA2 and B7x, in the progression of gastrointestinal and pancreatic neuroendocrine tumors (GINETs and PNETs). We demonstrated that both HHLA2 and B7x were expressed to a high degree in human GINETs and PNETs. We determined that the expression of B7x and HHLA2 correlates with higher grade and higher incidence of nodal and distant spread. Furthermore, we confirmed that HIF-1α overexpression is associated with the upregulation of B7x both in our in vivo and in vitro models. When grown in vitro, islet tumor β-cells lack B7x expression, unless cultured under hypoxic conditions, which results in both HIF-1α and B7x upregulation. In addition, we demonstrated that Men1/B7x double knockout (KO) mice (with loss of B7x expression) exhibited decreased islet β-cell proliferation and tumor transformation accompanied by increased T-cell infiltration compared with Men1 single knockout mice. We have also shown that systemic administration of a B7x mAb to our Men1 KO mice with PNETs promotes an antitumor response mediated by increased T-cell infiltration. These findings suggest that B7x may be a critical mediator of tumor immunity in the tumor microenvironment of NETs. Therefore, targeting B7x offers an attractive strategy for the immunotherapy of patients suffering from NETs.

Keywords: B7 immune-checkpoints, neuroendocrine tumors, tumor microenvironment, HIF-1α, immunotherapy

INTRODUCTION

Gastrointestinal (GI) and pancreatic neuroendocrine tumors (GINETs and PNETs) are increasing in incidence and yet remain an understudied tumor type with few effective treatments (Yao et al. 2008). Therefore, further understanding of their pathogenesis would be of high significance in developing new therapeutic approaches. The immune system and tumor cells have dynamic interactions, (Dunn et al. 2004, Grivennikov et al. 2010, Schreiber et al. 2011) which play crucial roles in tumor development and clinical outcome. New clinical data show that tumor-infiltrating lymphocytes, particularly CD4 and CD8 T-cells, are prognostic and predictive in human GINETs and PNETs, suggesting that T-cells play a major anti-tumor role during human NET progression (Epardaud et al. 2008, Katz et al. 2010, Sato et al. 2014). In humans, PD-1, PD-L1, and CTLA-4 are not expressed on resting T-cells but are induced after T-cell activation and expressed on regulatory T-cells (Scandiuzzi et al. 2011). B7x and HHLA2 have been identified as being absent in most normal tissues and only HHLA2 is expressed on human monocytes without any stimulation (Zhao et al. 2013). However, B7x and HHLA2 are overexpressed in human tumors of the brain, thyroid, esophagus, lung, stomach, pancreas, kidney, gut, skin, ovary, uterus, prostate, and breast, often correlating with negative clinical outcomes (Hofmeyer et al. 2012, Janakiram et al. 2017, Lee et al. 2012, Wei et al. 2011, Zang et al. 2007, Zang et al. 2010). Both B7x and HHLA2 are involved in tumor development and progression by inhibiting T-cell function (Vigdorovich et al. 2013, Zang et al. 2003, Zhao et al. 2013). In the present study, we investigated the role of two B7 immune checkpoints, HHLA2 (HHLA2 is not present in mice) and B7x, in the tumor microenvironment of NETs. We then assessed the efficacy of a high-affinity antibody targeting the immune checkpoint B7x in reducing the growth of PNETs, using our previously described multiple endocrine neoplasia type 1 (MEN1) knockout (KO) mouse model (Shen et al. 2009).

During tumor development and progression, cancer and stromal cells often have restricted access to nutrients and oxygen. Hypoxia is a common feature of solid tumors, including NETS, and results in the aberrant vascularization observed in the tumor microenvironment (Kobayashi et al. 2016, Pinato et al. 2014, Pouyssegur et al.2006, Sethumadhaven et al. 2017, Westendorf et al. 2017). The hypoxic response is mainly ascribed to hypoxia-inducible factors (including HIF-1, HIF-2, HIF-3). The activation of the HIF-1α oxygen sensitive subunit is a critical effector in the hypoxic tumor microenvironment of common advanced tumors including NETs (Laitala et al. 2018, Maione et al. 2012, Soni et al. 2017,). Studies have shown an upregulation of HIF-1α in neuroendocrine tumors (Pinato et al. 2014). Hypoxia has also been shown to suppress immune function within the tumor microenvironment (Palazon et al. 2014, Palazon et al. 2017). Recent studies reported that hypoxia causes a rapid, dramatic, and selective upregulation of PD-L1, and it is important to note that HIF-1α directly binds to the promoter of PD-L1 (Noman et al. 2014). In this study, we also explored the mechanistic relationship between the expression of HIF-1α and the upregulation of B7x in the tumor microenvironment of PNET tumors.

MATERIALS AND METHODS

Human tissue specimens

Thirty-seven human specimens from gastrointestinal (GINET, N=13) and pancreatic (PNET, N=24) tissues and matched adjacent normal tissue were obtained. All tissues were collected from individuals operated on at the Albert Einstein College of Medicine (Einstein) or Roy J and Lucille A. Carver University of Iowa College of Medicine (University of Iowa). All tissues were stored at −80°C and carefully evaluated by an experienced pathologist. Patient samples were obtained after informed consent regarding the use of specimens for research purposes and the nature of all procedures used. Unidentified specimens were used in these studies. This study was approved by the Institutional Review Board at both Einstein and the University of Iowa.

In this study cohort (n=37 patients), the mean age of the patient population was fifty-seven years. The known American Joint Committee on Cancer (AJCC) staging information (n=37), stage I, II, III, and IV, were 48.6% (18/37), 32.4% (12/37), 10.8% (4/37), 8.1% (3/37), respectively (Supplemental Table 1).

Men1 knockout (KO) mice and Non-Men1 KO (RIP-Tag) mice

We utilized the Cre-lox system, driven by the tissue-restricted promoter Pdx1, to inactivate Men1 in the pancreas of mice (C57BL/6 background). Mice with homozygous loss of Men1 develop islet hyperplasia at 6 months and insulinomas at 12-months with increased vasculature in Pdx1-Cre; Men1 floxed/floxed (Men1 KO) mice (Shen et al. 2009). These Men1 KO mice had elevated serum insulin levels starting at 4 months of age that persists throughout their lifetime (Shen et al. 2009). Since these mice are patterned after the human MEN1 familial cancer syndrome, they provide an in vivo tool for the study of neuroendocrine tumors of the pancreas (Shen et al. 2009, Smith et al. 2016). The tumor tissues from these Men1 KO mice were used to characterize B7x expression and immune response, and they were used to assess the efficacy of the anti-B7x mAb in the treatment of PNETs. To explore whether the NET tumors with wild type menin also have high expression of B7x, we investigated the B7x expression in the PNETs from rat insulin promoter-T antigen (RIP-Tag) mice. This RIP-Tag transgenic mouse model expresses the SV40 T-antigen under control of the rat insulin promoter, resulting in pancreatic β-cell tumors without MEN1 mutation and loss of menin (Contractor et al. 2016). All animal experiments were conducted under the approved animal protocols at Rutgers University.

B7x KO mice

The mice lacking B7x (B7x KO) on a C57BL/6 background are as previously described (Wei et al. 2011). In our study, we did not find that the phenotype nor B7x expression have any differences between age and gender in the B7x KO mice. For our experiments, we used 50% female mice and 50% male mice. The B7x KO mice have been crossed to Men1 KO mice with PNETs on the same C57BL/6 background. We tested if loss of B7x allowed for T-cell activation and prevents the formation of PNETs. 24 mice were assigned to 4 groups for each genotype (Men1 KO/B7x KO, Men1 KO/B7x wild type (WT), Men1 WT/B7x KO, Men1 WT/B7x WT). We monitored tumor growth by measuring serum insulin, a tumor growth biomarker of Men1 KO mice, from 3 to 12 months of ages. All mice for each genotype were sacrificed at 12 months of age. Whole pancreatic tissues from the mice were formalin-fixed for further analysis. Genetic controls were age- and gender-matched littermates. All animal experiments were conducted under the approved animal protocols at Rutgers University.

Islet β-cell tumor cell line

Since our Men1 KO mouse pancreatic β-cells fail to grow in vitro, we have adopted a transplantable line from the N134 cell model for in vitro studies. N134 is a pancreatic β-cell tumor cell line derived from a tumor from RIP-Tag; RIP-Tva transgenic animal (obtained from Dr. Nancy Du, Weill Cornell Medical College, New York) at 16 week of age, and maintained in dulbecco’s modified eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS) (Du et al. 2007). The N134 β tumor cells were cultured under hypoxic (1% O2) and normoxic conditions at different time points.

Targeting B7x with anti-B7x mAb in Men1 KO mice with PNETs

  1. The first protocol evaluated the effects of therapy on tumor growth by measuring serum insulin and tracking animal survival: Twelve Men1 KO mice (12 months of age) with insulinomas (based on measurement of serum insulin levels, as outlined above) were assigned to two groups treated with either 1H3 anti-B7x mAb (Jeon et al. 2014) (n=6) or normal mouse IgG (n=6) (control group). 200μg of anti-B7x mAb or normal mouse IgG were administrated by intraperitoneal injection (IP) on days 1, 3, 7, 11, and 14. Each mouse received a total of 1mg of antibodies during the course of the experiments. All mice were treated at 12 months of age and we followed the survival time in each group during and after treatment.

  2. The second protocol was designed to examine immune response and tumor cell growth: Eight Men1 KO mice (12 months of age) with insulinomas (based on measurement of serum insulin levels, as outlined above) were assigned to two groups and treated with either 1H3 anti-B7x mAb (n=4) or normal mouse IgG (n=4) (control group). 200μg of anti-B7x mAb or normal mouse IgG were administrated by IP injection on days 1, 3, 7, 11, and 14. Each mouse received a total of 1mg of antibodies during the B7x experiments. All mice were sacrificed at 1 month post-treatment. Whole pancreatic tissues from the mice were flash-frozen at −80°C then later formalin-fixed for further analysis. The tumor growth and immune response in the tumor immune microenvironment were assessed.

Laser capture microdissection (LCM)

In order to detect the mRNA expression level of target genes in the islets of the pancreas, we used a LCM technique to isolate the islets of the pancreas from mice in our study. The pancreatic tissues from Men1 KO and WT mice were harvested and embedded in OCT. The OCT-embedded tissue sections were stained by an Arcturus Histogene Frozen Section Staining Kit (AB Applied Biosystems, CA, USA) and the islet tissues dissected using an Arcturus XT LCM system.

Real-time RT-PCR analysis

In human samples with PNETs, the endocrine tissues were isolated from primary tumors, metastases, and islets of adjacent normal tissues by LCM. For human samples from GINET patients, because we cannot isolate the endocrine cells from GI tissues, we used matched adjacent normal tissues as normal controls. Moreover, HHLA2 and B7x were reported not to be expressed in all cell types in normal GI organs. In addition, we also used LCM to isolate islet tissues of the pancreas from Men1 KO and Men1 WT mice. Furthermore, total RNA was extracted from isolated tissues using the RNeasy total RNA kit (Qiagen, Valencia, CA, USA). mRNA expression was detected by quantitative real-time RT-PCR. Measurements of fluorescent signal intensity after standardization with an 18S ribosomal RNA (rRNA) internal control allowed quantification. The primers for HHLA2 and B7x gene were purchased from Qiagen (NM_007072 and NM_001253849). The primers for Perforin, Granzyme B, and FasL genes are listed in Supplementary Table 2.

Hematoxylin and eosin (H&E) staining

Mice were sacrificed and whole pancreas and other normal tissues (brain, lung, heart, liver, spleen, kidney, and muscle) were dissected, formalin-fixed, paraffin embedding, and sectioned (5 μm) using a rotating microtome (Microm, Walldorf, Germany). Hematoxylin and eosin staining was performed according to standard procedures. For morphometric analysis, three 100μm sections were digitalized and pancreas cross-sectional area in the section with maximal tumor diameter was determined by computerized pixel counting.

Immunohistochemistry (IHC) staining

IHC staining was used to determine expression levels of HHLA2, B7x, CD4, CD8, CD11b, Ki67, PCNA, and Caspase 3 in pancreatic tissues from NET patients and Men1 KO and WT mice. Purified mouse anti-HHLA2 (Zhao et al. 2013) and anti-B7x antibodies (obtained from Dr. Zang’s laboratory, Einstein; 1:50 dilution)(Zang et al., 2007), rabbit anti-CD4 (1:1000 dilution; Abcam, MA, USA), rabbit anti-CD8 (1:200 dilution; Abcam, MA, USA), mouse-anti-HIF-1α (1:200 dilution; Abcam, MA, USA), rabbit anti-Ki67 (1:100 dilution; Abcam, MA, USA), rabbit anti-CD11B (1:400 dilution; Novus Biologicals, CO, USA), rabbit anti-proliferating cell nuclear antigen (PCNA) (1:200 dilution; Santa Cruz, CA, USA), and rabbit anti-Caspase 3 (1:200 dilution; BD Pharmingen, CA, USA) antibodies were used and incubated overnight at 4°C. A HEP/DAB (ABC) detection IHC kit (Abcam, MA, USA) was used for IHC analysis. An IHC staining without primary antibody was used as a negative control and a recommended tissue for positive expression of the antibody was used as a positive control.

Immunofluorescence (IF) staining

IF staining was used for the evaluation of the expression levels of interferon-γ (IFN-γ) and tumor necrosis factor-α (TNF-α) in pancreatic tissues from Men1 KO and WT mice. The sections were incubated overnight at 4°C with primary antibodies: pig anti-insulin antibody (1:500 dilution; DAKO, CA, USA) and rabbit anti-IFN-γ (1:100 dilution; Abcam, MA, USA), rabbit anti-TNF-α (1:500 dilution; Abcam, MA, USA) antibodies. After washing, the slides were incubated with goat anti-rabbit Alexa Fluor 488 and goat anti-guinea pig Alexa Fluor 647 (both 1:200 dilutions; Thermo Fisher Scientific, NY, USA) for 45 min in the dark. The slides were assembled in Vectashield mounting medium with DAPI (Vector Laboratories, CA, USA). Images were obtained with an Epifluorescence microscope with an attached camera. The quantification of IF staining was performed by counting cells that strongly expressed the target proteins in 4 randomly selected fields per slide under 10 and 20x magnification.

Western Blot analysis

Total protein was extracted from pancreatic tissues of Men1 KO and WT mice using Radioimmunoprecipitation assay (RIPA) lysis buffer (Pierce, Rockford, IL). The extracted proteins were quantified and equal amounts of protein were separated by SDS-PAGE and blotted. The blots were incubated overnight at 4°C with primary antibodies: rabbit anti-GAPDH (1:1000 dilution; Thermo Fisher Scientific, NY, USA), mouse-anti-HIF-1α (1:400 dilution; Abcam, MA, USA), rabbit-anti-Perforin (1:1000 dilution; Abcam, MA, USA), rabbit-anti-Granzyme B (1:1000 dilution; Abcam, MA, USA), mouse-anti-Fas L (1:1000 dilution; Santa Cruz Biotech, CA, USA) antibodies, respectively. The secondary antibodies used were: 1:5000 dilution of anti-mouse IgG-HRP, 1: 2500 dilution of anti-rabbit IgG-HRP, and 1:2500 dilution of anti-goat IgG-HRP (Santa Cruz, CA, USA), respectively. The small differences in loading were corrected by comparison to the loading control (GAPDH).

Chromatin Immunoprecipitation (ChIP)-PCR assay

We used ChIP-PCR to validate the HIF-1α-B7x promoter interaction in N134 tumor β-cells in vitro and MEN1 tumor tissues from Men1 KO mice (Magna ChIP kit,Millipore, CA) according to the manufacturer’s protocol. Briefly, cell lines and tumor tissues were collected and cross-linked with 1% formaldehyde for 10 minutes at room temperature. Cross-linking was stopped by adding glycine for 5 minutes then cells were washed three times with ice-cold PBS/proteinase inhibitor. The nuclear extract was prepared, and chromatin was sonicated to a size of 500–1000 bp. The HIF-1α ChIP assay was performed by incubating with 10 μl (1 μg/μl) of HIF-1α antibody (Abcam, MA, USA) overnight at 4°C. Rabbit IgG was used as a negative control. Protein G magnetic beads were used to collect the immunoprecipitates (IP). The HIF-1α -ChIP DNA was assayed by PCR using primers corresponding to the promoter regions of the B7x gene, and a known target of HIF-1α, PFKFB4, also was used as a positive control (Xia et al., 2009) (Supplementary Table 3).

Enzyme-linked immunosorbent assay (ELISA)

Blood was collected via a retro-orbital bleeding technique from anti-B7x treated mice and IgG-treated control mice following an 18-h fast. Serum levels of insulin were measured by ELISA with the Ultrasensitive Mouse Insulin ELISA Kit (Mercodia, NC, USA) as described previously (Shen et al., 2009). The expression levels of IFN-γ and TNF-α in tumor tissues from B7x treated mice and IgG control mice were measured by a mouse IFN-γ ELISA kit (Abcam, MA, USA) and a mouse TNF-α ELISA kit (Life Technologies, MD, USA) following the manufacturer protocols.

Terminal deoxyuridine triphosphate nick end labeling (TUNEL) assay

Apoptotic status in tumor tissues was measured in anti-B7x treated and IgG control mice by TUNEL assay. For quantification of apoptosis, the TUNEL assay was performed according to the manufacturer instructions with an In Situ Cell Death Detection Kit (Roche Diagnostic, IN, USA). The sections were labeled and stained with the TUNEL reaction mixture (label plus enzyme solutions) for 60 min at 37°C and washed twice with PBS in the dark. The slides were mounted in Vectashield mounting medium with DAPI (Vector Laboratories, CA, USA). The apoptotic fluorescent cells were counted under a fluorescent microscope and the numbers were expressed as the percentage of total cells ± standard deviation (SD). A negative control without enzyme treatment and a positive control with DNase I treatment were also performed.

Survival Analysis

We followed the survival in each group during and after treatment from our animals under the first protocol. Then we calculated the average survival rate for each group.

Statistical analysis

The two-tailed student’s t-test and ANOVA were performed to assess the significance between experimental groups. The Pearson’s correlation or linear regression analysis was used for correlations between parameters. Cumulative survival time was calculated by the Kaplan-Meier method and comparison between the groups was tested by the log-rank test. GraphPad software (GraphPad Software, Inc, version 5.02) was used for all statistical analyses. Statistically significant P values are indicated in the figures with asterisks: ***, P <0.001; **, P < 0.01; *, P < 0.05 for tested samples vs controls.

RESULTS

Increased HHLA2 and B7x with decreased tumor infiltrating lymphocytes (CD4 and CD8) in human GINETs and PNETs

To determine if HHLA2 and B7x were overexpressed in human NETs, we performed IHC assays and quantitative real-time PCR on GINETs (N=13) and PNETs (N=24) and matched normal tissues. We demonstrated that both HHLA2 and B7x proteins were expressed to a high degree in GINET and PNET tissues compared with their matched normal tissues as measured by IHC (Figure 1). The pathological analysis was reviewed by a pathologist with double-blind reviewing. The positive cell rate was scored by 0: < 1%, 1: 0–15%, 2: 15–30%, 3: 30–60%, 4: >60%, and the score in human NETs for IHC staining is also presented in the Supplemental Table 1 and 4. Tumor infiltrating T cells (CD4 and CD8) were significantly decreased in the tumor microenvironment with high expression of HHLA2 and B7x compared with the tumor microenvironment with low expression of HHLA2 and B7x (Figure 1). These results suggest an association between HHLA2 and B7x and tumor infiltrating lymphocytes. Furthermore, we demonstrated that the expressed positive cell rates of HHLA2 and B7x were statistically higher in non-functional (do not produce hormones) (expressed positive cell rate: 45.2± 7.9% and 42.6±12.8%) and functional (hormone-producing) GINETs and PNETs (expressed positive cell rate: 15.9±5.7% and 22.8±4.8%) compared to nonfunctional case matched normal tissues, 4.9±2.3% and 3.5±1.9%, respectively (P<0.001), and functional case matched normal tissues, 4.2% and 1.8%, respectively (P<0.001) as measured by IHC staining. The mRNA expression levels, as measured by qRT-PCR, of both HHLA2 and B7x, are also statistically higher in non-functional and functional GINETs and PNETs compared to case matched normal tissues (Supplemental Figure 1A). We also found a strong correlation between mRNA and protein expression of HHLA2 and B7x, correlation coefficient: 0.8998 (P<0.001) and 0.8702 (P<0.001), respectively. In addition, we determined that the mRNA and protein expression levels of HHLA2 and B7x are significantly higher in primary tumor tissues that resulted in metastasis (90.3±8.2% and 80.5±5.4%) compared to primary tumor tissues that did not result in metastasis (P<0.001) (Supplemental Figure 1A) (Supplemental Table 1). These results revealed that both the mRNA and protein levels of HHLA2 and B7x were highly expressed in human GINETs and PNETs and that primary tumors with metastasis tended to have higher levels than primary tumors without metastasis, suggesting these two new immune checkpoints may play important roles in tumor progression and metastasis.

Figure 1. Expression of HHLA2 and B7x in human pancreatic neuroendocrine tumors (PNETs).

Figure 1.

Open in a new tab

Immunohistochemistry (IHC) staining of HHLA2 (A) and B7x (B) are shown in human PNET tissue; Negative staining of HHLA2 (E) and B7x (F) is also shown in the matched normal tissues by IHC staining. Furthermore, negative staining of CD4 (C) and CD8 (D) are shown in the human PNET tissue with high expression of HHLA2 and B7x; Positive staining of CD4 (G) and CD8 (H) is also shown in the human PNET tumor with low expression of HHLA2 and B7x by IHC staining. The images were taken with the 40X objective.

Expression levels of immune checkpoint proteins HHLA2 and B7x correlate with clinical and pathological parameters.

To evaluate the clinical utility of HHLA2 and B7x as informative markers, we performed correlation analysis between the expression levels of HHLA2 and B7x and clinical-pathological parameters using Pearson’s Correlation. We found that there is a correlation between the expressed positive cell rate of HHLA2 with tumor size, correlation coefficient: 0.6488 (P=0.0002) for nonfunctional NETs (Supplemental Figure 1B1) and 0.5613 (P=0.0421) for functional NETs (Supplemental Figure 1B5), respectively, and tumor cell proliferation as measured by IHC using a Ki67 marker, correlation coefficient: 0.4222 (P=0.0064) for nonfunctional NETs (Supplemental Figure 1B3) and 0.5854 (P=0.0370) for functional NETs (Supplemental Figure 1B7), respectively. In addition, we demonstrated that a correlation between the expressed positive cell rate of B7x with tumor size, correlation coefficient: 0.6681 (P<0.0001) for nonfunctional NETs (Supplemental Figure 1B2) and 0.7818 (P=0.0082) for functional NETs (Supplemental Figure 1B6), respectively, and tumor cell proliferation as measured by IHC using a Ki67 marker, correlation coefficient: 0.5786 (P=0.0010) for nonfunctional NETs (Supplemental Figure 1B4) and 0.8621 (P=0.0025) for functional NETs (Supplemental Figure 1B8), respectively. Our results indicate that the expression of HHLA2 and B7x correlate with higher grade and higher incidence of nodal and distant spread of GINETs and PNETs (Supplemental Figure 1B)(Supplemental Table 1 and 4).

Upregulation of B7x in the PNETs of Men1 KO mice.

Previously, we developed a genetically-engineered mouse model of PNETs (Pdx1-Cre; Men1floxed/floxed) that provides an in vivo tool for the study of PNETs that closely mimics human disease (Shen et al. 2009). This model relies on the tissue-specific deletion of the Men1 gene in the pancreas, resulting in development of PNETs that secrete insulin. Using this model, we demonstrated that the expression level of B7x is upregulated in the pancreatic endocrine tumor tissue from Men1 KO mice compared to the normal pancreatic endocrine tissue from Men1 WT mice (Figure 2). Additionally, B7x is not expressed in control tissues (brain, heart, liver, spleen, kidney, and muscle) from Men1 KO mice or in any tissues of Men1 WT mice (Supplemental Figure 2).

Figure 2. B7x expression in the tumor microenvironment of Men1 knockout (KO) mice.

Figure 2.

Open in a new tab

Tumors are observed in pancreatic endocrine tissues of Men1 KO mice (B1) but not in Men1 wild type (WT) mice (A1) as shown by H&E staining. B7x expression is upregulated in tumor tissues from Men1 KO mice (B2) compared to pancreatic tissue from Men1 WT mice (A2) by immunohistochemistry (IHC) staining. The A1 and B1 images were taken with the 2.5X objective and A2 and B2 images were taken with the 20X objective.

Further, to determine whether the increased B7x expression is caused by Men1 KO or not, we demonstrated that PNET tissues with WT menin from RIP-Tag mice also have high expression of B7x (Supplemental Figure 3).

HIF-1α overexpression coincides with the upregulation of B7x in the tumor microenvironment of Men1 KO mice and in the Rip-Tag β-cell line.

By analyzing our Men1 KO model, we have demonstrated that the proliferating β-cells result in the overexpression of HIF-1α in the tumor microenvironment at 6 months (hyperplasia) and at 12 months (insulinoma) but not at earlier time points (4 months) (Figure 3). Furthermore, we have demonstrated that HIF-1α overexpression coincides with the upregulation of B7x in the tumor microenvironment during β-cell proliferation at 6 months (hyperplasia) and at 12 months (insulinoma) but not at earlier time points (4 months) without β-cell proliferation and not in islets of WT mice (Figure 3).

Figure 3. HIF-1α overexpression is associated with the upregulation of B7x in tumor microenvironment of Men1 knockout (KO) mice.

Figure 3.

Open in a new tab

The proliferated β-cells resulting in the overexpression of HIF-1α in the tumor microenvironment of Men1 KO mice at 12 months (insulinoma) (A1 and C1) and at 6 months (hyperplasia) (A2 and C2) are associated with the upregulation of B7x (E1 and E2) but not in islets of Men1 KO mice at 4 months (A3, C3, and E3) and islets of Men1 wild type (WT) mice (B1, D1, F1, B2, D2, F2, B3, D3, and F3) by immunohistochemistry (IHC) staining. The images were taken with the 20X objective.

Given that explants from our Men1 KO mice fail to grow in vitro, we have adopted a transplantable line from the N134 model for in vitro studies. When grown in vitro, RIP-Tag islet tumor β-cells (N134) lack B7x expression, presumably due to the lack of influence of the tumor microenvironment. When cultured under hypoxic conditions (1% O2), N134 cells expressed B7x and HIF-1α (Figure 4A1). However, HIF-1α and B7x were not expressed in N134 cells transduced with HIF-1α siRNA (200 nM, 48 hrs, Santa Cruz Biotechnology) under hypoxic conditions (1% O2) (Figure 4A2), providing a direct association between HIF-1α and B7x expression.

Figure 4. HIF-1α regulates the B7x expression and HIF-1α binds to B7x promoter.

Figure 4.

Open in a new tab

In vitro, when cultured under hypoxic conditions (1% O2) after 48h and 72h, N134 isle tumor β cells that lack B7x expression were found to induce the upregulation of HIF-1α and B7x (A1); however, the expression levels of HIF-1α and B7x were not induced in the islet tumor β cells following HIF-1α siRNA treatment after 48 hours by western blot (A2).

HIF-1α can bind to the B7x promoter in N134 isle tumor β cells under in vitro hypoxic conditions (B1) and tumor tissues of Men1 KO mice (B2) using a chromatin immunoprecipitation (ChIP)-PCR analysis. The HIF-1α immunoprecipitation (IP) products were amplified using quantitative real-time RT-PCR with primers corresponding to the promoter regions of the B7x gene and the relative amount of DNA was calculated using fold enrichment compared to the IgG Ab control. Statistically significant P values are indicated in the figures with asterisks: ***, P <0.001.

HIF-1α interacts with the B7x promoter

To expand our mechanistic characterization, we investigated whether HIF-1α interacts with the promoter region of B7x in both N134 tumor β-cell line in vitro and in fresh Men1 tumor tissues from Men1 KO mice. We identified that HIF-1α can bind to the B7x promoter in N134 β-cells under hypoxic conditions (1% O2) and in tumor tissues from Men1 KO mice by using a ChIP-PCR analysis (Figure 4B).

Loss of the immune checkpoint protein B7x results in T-cell infiltration associated with reduced tumor burden.

We have crossed our Men1 KO mice with mice lacking B7x (B7x KO). We observed a significant decrease in serum insulin, a tumor biomarker, in the Men1 KO/B7x KO mice with loss of B7x expression from 4 to 12 months when compared to Men1 KO/B7x WT mice (P<0.001) (Figure 5). The Men1 KO/B7x KO mice had a significant decrease in islet size compared with the Men1 KO/B7x WT mice (1477 μm2 ± 441 vs 27382 μm2 ±827; P<0.001), but the islet size in Men1 KO/B7x KO mice had a slight increase compared with the Men1 WT/B7x KO and Men1 WT/B7x WT control mice (1477 μm2 ± 441 vs 1037 μm2 ± 211 and 988 μm2 ± 177; P<0.05) at 12 months of age (Figure 6A). Furthermore, we demonstrated that the Men1 KO/B7x KO mice had significantly increased T-cell infiltration compared with Men1 KO/B7x WT mice, but both negative control groups (Men1 WT/B7x KO and Men1 WT/B7x WT) did not exhibit any T-cell infiltration in the islets of the pancreas as measured by IHC (Figure 6C and 6D). These results indicate that loss of B7x allows T-cells activation and prevents the formation of PNETs.

Figure 5. Loss of B7x decreases serum insulin level.

Figure 5.

Open in a new tab

There was a significant decrease in serum insulin in the Men1 knockout (KO)/B7x KO mice compared to Men1 KO/B7x wild type (WT) mice (P<0.01) by enzyme-linked immunosorbent assay (ELISA). Both normal negative control groups: Men1 WT/B7x KO and Men1 WT/B7x WT mice didn’t obverse any changes. Statistically significant P values are indicated in the figures with asterisks: ***, P <0.001; **, P < 0.01.

Figure 6. Loss of B7x promotes T-cell infiltration in islets infiltration and associated with reduced tumor burden in Men1 knockout (KO)/B7x KO mice.

Figure 6.

Open in a new tab

H&E and immunohistochemistry (IHC) staining showed that the Men1 KO/B7x KO mice have loss of B7x expression in islets of pancreas (B1) inhibited tumor growth in islets of pancreas (A1) and demonstrated increased T-cell infiltration in the islets (C1 and D1) compared with Men1 KO/B7x wild type (WT) mice (A2, B2, C2 and D2). Both normal negative control groups: Men1 WT/B7x KO and Men1 WT/B7x WT mice didn’t obverse any changes (A3, B3, C3, D3, A4, B4, C4, and D4).

Blockade of B7x with anti-B7x mAb decreases serum insulin levels and improves survival in Men1 KO mice

Based on the therapeutic potential of inhibiting B7x overexpression, we treated PNET-exhibiting Men1 KO mice using a high-affinity anti-B7x antibody (B7x mAb 1H3) (Jeon et al. 2014). Since Men1 KO mice develop insulinomas, which are clinically defined by a measurable increase in serum insulin levels, serum insulin can be used as a surrogate for tumor presence and size. In two groups of Men1 KO mice, after two weeks of treatment with either anti-B7x mAb (N=6) or normal mouse IgG (N=6), the group treated with anti-B7x mAb exhibited a strong and sustained decrease in serum insulin as measured by ELISA (P<0.001) (Figure 7A) and an enhanced survival rate of 83.3% (5/6) compared to the group treated with control IgG 0% (0/6)(P<0.001)(Figure 7B). These results support the conclusion that blockade of B7x can decrease serum insulin and improve survival in Men1 KO mice.

Figure 7. B7x mAb treatment induces an antitumor response and improved survival in Men1 knockout (KO) mice.

Figure 7.

Open in a new tab

The mice were treated with B7x Ab and IgG control at 12 months old. There was a significant decrease in serum insulin in the B7x treatment group but no change in the IgG control group in Men1 KO mice (P<0.001) (A). Increased survival was also observed in the B7x antibody treatment cohort compared with mice treated with IgG control (P<0.001) (B). H&E staining shows that the B7x mAb treated mice had a significant decrease in tumor size compared with IgG control mice (C). The images were scanned with the 0.5X (C1 and C3) and 3X objective (C2 and C4). Tumor cell proliferation is significantly decreased by PCNA and Ki67 analysis (D1 and D2) and apoptosis is significantly increased by Caspase 3 and Terminal deoxyuridine triphosphate nick end labeling (TUNEL) analysis (D3 and D4) in the B7x mAb treatment group compared with the IgG control group. The average across all samples ± SD were used for statistics in the histograms. Statistically significant P values are indicated in the figures with asterisks: ***, P <0.001; **, P < 0.01.

Blockade of B7x with a B7x mAb inhibits tumor cell proliferation and induces tumor cell apoptosis in Men1 KO mice

To identify the effects of targeting B7x with anti- B7x mAb treatment on tumor growth in Men1 KO mice, we compared the tumor size of the treated group and the IgG control group. We demonstrated that the tumor size was significantly reduced in the anti-B7x mAb treated group when compared to the IgG control group (1928 μm2 ± 661 vs 13196 μm2 ± 1525; P<0.0001) (Figure 7C). Furthermore, we demonstrated that tumor cell proliferation was significantly inhibited as measured by PCNA (Figure 7D1) and Ki67 staining (Figure 7D2). In addition to decreased proliferation, apoptosis was significantly increased as measured by Caspase 3 staining (Figure 7D3) and a TUNEL assay (Figure 7D4) in B7x mAb treated mice when compared with IgG control mice. These results indicate that blockage of B7x can inhibit tumor growth and induce apoptosis in Men1 KO mice.

Blockade of B7x enhances T-cell response, decreases infiltrating immune suppressor cells MDSCs, and induces the production of IFN-γ and TNF-α in the tumor microenvironment

We demonstrated that anti-B7x mAb treatment robustly increases tumor infiltrates of both CD4 (Figure 8A2 and 8B) and CD8 (Figure 8A4 and 8B) T-cells, in the PNET tumor microenvironment when compared to the control group (Figure 8A1, 8A3, and 8B). Furthermore, we investigated the effect of B7x antibody treatment on immunosuppressive myeloid-derived suppressor cell (MDSC) infiltrates and demonstrated that they are significantly decreased in the treated group (Figure 8A6 and 8B) compared to control (Figure 8A5 and 8B). We also observed that INF-γ and TNF-α are significantly increased following anti-B7x treatment, suggesting that cytotoxic T-lymphocytes (CTL) may be activated and act to enhance the anti-tumor effect in B7x mAb treated mice. This was determined by IF staining (Supplemental Figure 4) and ELISA (Figure 8C). These results indicate that blockade of B7x can enhance anti-tumor immune responses by increasing CTL and decreasing immune suppression.

Figure 8. Anti-B7x mAb induces cytotoxic T lymphocyte response and increased expression of cytokines in the tumor microenvironment with anti-B7x mAb treatment.

Figure 8.

Open in a new tab

Immunohistochemistry (IHC) staining shows that the CD4+ (A2) and CD8+ (A4) T cells are increased and CD11b+ (MDSCs) (A6) are decreased in the B7x mAb treatment group compared with the IgG control group (A1, A3, and A5); quantification was achieved by counting CD4+, CD8+, and CD11b+ (B). The levels of IFN-γ and TNF-α are also significantly increased in the tumor microenvironment in the B7x treatment group compared with the IgG control as shown by enzyme-linked immunosorbent assay (ELISA) (C). The average across all samples ± SD was used for statistics in the histograms. Statistically significant P values are indicated in the figures with asterisks: ***, P <0.001; **, P < 0.01.

Blockade of B7x facilitates anticancer immunity through increased expression of Perforin, Granzyme B, and FasL in tumors of Men1 KO mice

To further characterize the signaling pathways involved in the immune response with blockade of B7x, we determined the expression levels of perforin, granzyme B, and FasL mRNA and proteins in tumors using LCM-quantitative real-time PCR and Western blot. We demonstrated a significant upregulation of perforin, granzyme B, and FasL in the tumors of Men1 KO mice with anti-B7x antibody treatment compared to IgG controls (Figure 9). These results indicate that blockage of B7x induces apoptosis in tumor cells by upregulation of perforin/granzyme B and FasL pathways.

Figure 9. Blockade of B7x increased expression of Perforin, Granzyme B, and FasL in the tumor microenvironment with anti-B7x mAb treatment.

Figure 9.

Open in a new tab

The mRNA expression of perforin, granzyme B, and FasL were quantified by real-time PCR and we observed a significant upregulation of perforin, granzyme B, and FasL in the tumors of Men1 knockout (KO) mice with B7x antibody treatment compared to IgG controls (A). Furthermore, we demonstrated that the protein levels of perforin, granzyme B, and FasL were significantly increased in the tumors of Men1 KO mice with B7x antibody treatment compared to IgG controls (B) by Western blot. The average across all samples ± SD was used for statistics in the histograms. Statistically significant P values are indicated in the figures with asterisks: ***, P <0.001; **, P < 0.01.

DISCUSSION

The immune system is capable of playing an important role in controlling and eradicating cancer. However, immunosuppressive processes that occur in the tumor microenvironment limit the normal function of the immune system in multiple malignancies. It is thought that the main function of tumor cell-associated immune checkpoint molecules is the modulation of antitumor immune responses. To further explore this paradigm, we characterized the immunological role of two unique immune checkpoints, HHLA2 and B7x, in the development of NETs in human NET samples and our Men1 KO PNET mouse model. We observed that increased expression of HHLA2 and B7x is associated with human GINETs and PNETs. The expression of the HHLA2 and B7x proteins is also higher in non-functional tumors, which are generally identified later in disease progression compared to functional tumors. Our studies further demonstrated that the expression of HHLA2 and B7x is higher in primary tumors with associated metastases compared to the those without metastases. This correlation between the expression of HHLA2 and B7x with detrimental clinical features also extends to tumor size and tumor cell proliferation (Ki67 staining). In addition to being present in human samples, upregulated B7x was also confirmed in our Men1 KO mouse model, which serves as an excellent surrogate for human disease progression. Our results indicate that primary tumor cells may up-regulate the expression of HHLA2 and B7x, thereby equipping these cells with an increased capability to grow and metastasize, eventually resulting in the advanced progression of NETs. The over-expression of these B7 immune checkpoints by human NETs may provide the essential mechanism by which NET cells avoid destruction mediated by tumor–reactive T-cells and other immune cells. The high level of B7x expression may facilitate tumor progression, therefore providing a valuable prognostic marker to predict outcome based on a suitable drug target.

In addition, to explore whether the expression of B7x was only upregulated in the PNETs of Men1 KO mice, we investigated B7x expression in the PNETs from RIP-Tag mice. This RIP-Tag transgenic mouse model expresses the SV40 T-antigen under control of the rat insulin promoter, resulting in pancreatic β-cell tumors without MEN1 mutation nor the loss of menin. We demonstrated that the B7x upregulation was also observed in the PNETs from RIP-Tag mice (Supplemental Figure 3). Our results showed that increased B7x expression is not caused by Men1 KO.

Previous studies have reported upregulated expression of immune checkpoints that inhibit T-cell function and increase immunosuppressive MDSCs within the tumor microenvironment (Assal et al. 2015), and these reports support what we have shown in our current study. Tumor-infiltrating lymphocytes, particularly CD8 and CD4 T-cells, are prognostic and predictive in NETs, suggesting that T-cells are a significant anti-tumor population during NET progression (Epardaud et al. 2008, Katz et al. 2010, Sato et al. 2014). However, tumor cells may over-express T-cell inhibitory B7 molecules as they “evolve” to evade the immune system. Our experimental findings demonstrate that the overexpression of B7x in the tumor cells from our Men1 KO mouse model was associated with a decrease in T-cells and an increase in MDSCs within the tumor microenvironment, an occurrence we reasonably believe to be occurring in human tumors as well. These findings suggest that the B7x molecule is mediating immune inhibition in NETs, by both inhibiting T-cell activation and increasing the MDSC immunosuppressive function.

The molecular mechanism(s) behind B7x upregulation have not been entirely elucidated. We found that the Rip-Tag tumor cells lacked B7x expression, presumably due to the lack of influence of the tumor microenvironment. Hypoxia in the tumor microenvironment is a common feature of solid tumors as well as NETs. The hypoxic tumor microenvironment is caused by increased oxygen consumption due to hyperplasia and decreased oxygen delivery due to increased diffusion and is primarily driven by the hypoxia inducible transcription factors. Hypoxia causes the upregulation of PD-L1 and there are reports of HIF-1α binding directly to the promoter of PD-L1 (Noman et al. 2014). In order to further our mechanistic understanding of B7x upregulation, we evaluated whether the rapid proliferation of tumor cells drives HIF-1α overexpression resulting in upregulation of B7x in the tumor microenvironment of PNETs. We have demonstrated that HIF-1α overexpression is associated with the upregulation of B7x both in our in vivo and in vitro models. Furthermore, we identified that that HIF-1α can bind the B7x promoter in tumor β-cells cultured in vitro under hypoxic conditions and MEN1 tumor tissues from Men1 KO mice. This suggests that a potential molecular mechanism of B7x upregulation is due to the expression of HIF-1α following the relative hypoxia resulting from the rapid growth of tumor cells.

We have demonstrated that B7x is expressed to a high degree in patient-derived GINETs and PNETs. To further investigate the role that B7x plays in the maintenance of the tumor phenotype seen in Men1 knockout mice, we have crossed our Men1 KO mice with mice lacking B7x (B7x KO). We observed a significant decrease in serum insulin, a tumor biomarker, in the Men1 KO/B7x KO mice from 4 to 12 months when compared to Men1 KO/B7x WT mice (P<0.001). Furthermore, we demonstrated that the Men1 KO/B7x KO mice with loss of B7x expression inhibited tumor growth in the islets of the pancreas and increased T-cell infiltration compared with Men1 KO/B7x WT mice.

Immune checkpoint inhibitors have demonstrated considerable promise with their recent approval for the treatment of melanoma, non-small cell lung cancer, and other cancers (Shih et al. 2014). Studies have shown that targeting the PD-1 and PD-L1 molecules yields a robust antitumor response in the treatment of melanoma, non-small cell lung cancer, head and neck squamous cell cancer, urothelial cancer, Hodgkin lymphoma, and renal cell cancer (Bellmunt et al. 2017, Hodi et al. 2014, Reck et al. 2016,). In this study, our cohorts of mice administered with the anti-B7x antibody showed a strong and sustained decrease in serum insulin (a surrogate for tumor response) over the experimental timeframe compared with animals that were administered the IgG control. Importantly, mice treated with anti-B7x antibody also exhibited a significant increase in overall survival compared with mice treated with the IgG control. Furthermore, we also demonstrated that the beneficial outcome of the blockade of B7x was not limited to preventing insulin secretion but also included antiproliferative effects and apoptotic induction. This is of particular clinical relevance because the B7x antibody may not only alleviate symptoms in patients with insulinomas or other functional NETs but might actually lead to tumor regression and prolonged survival.

Blocking the interaction between an immune checkpoint molecule and its ligand(s) has been shown to result in impressive antitumor responses by altering the tumor immune microenvironment, including increasing T-cell activation and inhibiting infiltration by MDSC. We analyzed immune cells and cytokines to better understand how targeting B7x may transform immune suppression into immune stimulation in the tumor microenvironment. As evidence of the therapeutic mechanism of the anti-B7x mAb treatment, we observed that mice treated with the anti-B7x mAb had a higher percentage of CD4 and CD8 T-cell infiltration and a lower percentage of MDSC immune suppressing cells than control IgG treated mice. The production of IFN-γ and TNF-α were also significantly increased, suggesting that cytotoxic T-lymphocytes were activated. Furthermore, we have shown that targeting B7x activates the Perforin/Granzyme B/FasL pathways in the tumor microenvironment of Men1 KO mice. This marks a significant finding because Perforins and Granzyme B are critical players in tumor cell destruction by CTL. CTL kills tumor cells through Perforins and Granzymes upon contact between the T-cell receptor and presented self- major histocompatibility complex molecules on the tumor cell or antigen presenting cells (Chowdhury et al. 2008). Granzyme B activates intrinsic cell death proteases and caspases in tumor cells leading to caspase mediated degradation of hundreds of cellular protein substrates, which promotes fast and efficient apoptosis. By functioning through both reduction of proliferation and increasing apoptosis, B7x blockade exhibits significant anti-tumor effects.

In the present study we demonstrated that B7x is a critical mediator of tumor immunity in the tumor microenvironment of NETs. Furthermore, we found that B7x upregulation is a result of the expression of HIF-1α following the relative hypoxia resulting from rapid growth of tumor cells. Blockade of B7x with an anti-B7x antibody can inhibit tumor cell proliferation and induce apoptosis, both of which improve the tumor immune microenvironment and result in increased survival in the preclinical setting. This work reveals important insights that can be rapidly translated to the clinic by the targeting of B7x for the treatment of patients with NETs. Additionally, given the strong likelihood of the presence of this mechanism in other cancer types, our anti-B7x mAb may have additional potential for the treatment of other types of cancers.

Supplementary Material

Supp Table 1

Supplemental Table 1: Clinical characteristics and the expression of B7x and HHLA2 in the patients with neuroendocrine tumors. The expression of B7x and HHLA2 were analyzed in the patients with neuroendocrine tumors by immunohistochemistry (IHC) staining. The expression of B7x and HHLA2 were significantly increased in the later stage cancer to compared to early stage cancer (P=0.0014), but there is a significantly correlation between the expression of B7x and HHLA2 with age (P=0.5389) or gender (P=0.8523).

Supp Table 3

Supplemental Table 3: Primer sequences for chromatin immunoprecipitation (ChIP)-PCR.

Supp Table 2

Supplemental Table 2: Primer sequences for quantitative real-time RT-PCR.

Supp Table 4

Supplemental Table 4: Clinical characteristics and expression of B7x and HHLA2 in the individual patients with neuroendocrine tumors.

Supp Table 5
Supp Figure 1B
Supp Figure 1 A

Supplemental Figure 1. Expression levels of immune check-point proteins HHLA2 and B7x are correlated with clinical and pathological parameters.The expressed positive cell rates of HHLA2 and B7x are statistically higher in nonfunctional (42.6±12.8% and 45.2± 7.9%) and functional neuroendocrine tumors (NETs) (15.9±5.7% and 22.8±4.8%) compared to nonfunctional case matched normal tissues, 3.5±1.9% and 4.9±2.3%, respectively (P<0.001), and functional case matched normal tissues, 1.8±0.3% and 4.2±1.1%, respectively (P<0.001) by immunohistochemistry (IHC) staining (A1 and A2); the expressed positive cell rates of HHLA2 and B7x are significantly higher in primary tumor tissues with metastasis (80.5±5.4% and 90.3±8.2%) compared to tumor tissues without metastasis (A1 and A2); similar results are observed in mRNA expression by real-time RT-PCR analysis (A3 and A4). (N: Normal controls; T: Tumors). The average across all samples ± SD was used for statistics in the histograms. Statistically significant P values are indicated in the figures with asterisks: ***, P <0.001.

There is a correlation between the expressed positive cell rate of HHLA2 and B7x with tumor size in nonfunctional neuroendocrine tumors (NETs) (B1 and B2) and functional NETs (B5 and B6). There is also a correlation between the expressed positive cell rate of HHLA2 and B7x with tumor cell proliferation as measured by IHC with the Ki67 marker in nonfunctional NETS (B3 and B4) and functional NETs (B7 and B8).

Supp Figure 2

Supplemental Figure 2. Control normal tissues from Men1 wild type (WT) and knockout (KO) mice were analyzed for B7x expression byimmunohistochemistry(IHC). H&E staining of the control normal tissues (brain, heart, liver, spleen, kidney, and muscle) from Men1 WT (A1-A6) and KO mice (B1-B6). Negative IHC staining of B7x in the control normal tissues (brain, heart, lung, liver, spleen, kidney, and muscle) from Men1 WT (C1-C6) and KO mice (D1-D6).

Supp Figure 3

Supplemental Figure 3. B7x expression in the pancreatic neuroendocrine tumors (PNETs) with wild type (WT) menin from rat insulin promoter-T antigen (RIP-Tag) mice. H&E staining of the normal pancreatic tissue from WT control mice (A1) and PNET from RIP-Tag mice (A2). Negative immunohistochemistry (IHC) staining of B7x is shown in the normal pancreatic tissue from WT control mice (B1) and positive staining of B7x is shown in the PNET tissue from RIP-Tag mice (B2).

Supp Figure 5

Supplemental Figure 5. The expression levels of HIF-1α and B7x in N134 cells. The expression levels of HIF-1α and B7x were not induced in the islet tumor β cells following HIF-1α siRNA treatment after 48 hours by western blot. We provided the entire films for the three antibodies in theSupplemental Figure 5.

Supp Figure 4

Supplemental Figure 4. Anti-B7x mAb increased expression of cytokines in the tumor microenvironment. The levels of interferon-γ (IFN-γ) (B) and tumor necrosis factor-α (TNF-α) (D) are significantly increased in the tumor microenvironment in the B7x treatment group compared with the IgG control (A and C) as shown by immunofluorescence (IF) staining, immunostaining of insulin (pink), IFN-γ/TNF-α (green), and nuclei counterstained with DAPI (blue). The images were taken with the 10X objective.

Acknowledgements:

This work was supported in part by the NCI P50CA174521 (Iowa NET SPORE award) (Libutti); the Rutgers Cancer Institute of New Jersey CCSG P30-CA72720-16 (Libutti).

NIH R01CA175495 and Department of Defense PC131008 (Zang); and the CINJ Biospecimen Repository Shared Resource.

Footnotes

Disclosure of Potential Conflicts of Interest: Dr. Zang is an inventor on patent number US 9447186 B2 which covers the topic of anti-B7x cancer immunotherapy and an inventor on two pending patents which cover the topic of HHLA2-directed immunotherapy.

REFERENCES:

  1. Bellmunt J, de Wit R, Vaughn DJ, Fradet Y, Lee JL, Fong L, Vogelzang NJ, Climent MA, Petrylak DP, Choueiri TK, et al. 2017. KEYNOTE-045 Investigators. Pembrolizumab as Second-Line Therapy for Advanced Urothelial Carcinoma. N Engl J Med. 376 (11):1015–1026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Chowdhury D & Lieberman J 2008. Death by a thousand cuts: granzyme pathways of programmed cell death. Annual review of immunology. 26, 389–420. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Contractor T, Kobayashi S, da Silva E, Clausen R, Chan C, Vosburgh E, Tang LH, Levine AJ, Harris CR. 2016. Sexual dimorphism of liver metastasis by murine pancreatic neuroendocrine tumors is affected by expression of complement C5. Oncotarget. 7(21):30585–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Du YC, Lewis BC, Hanahan D, Varmus H. 2007. Assessing tumor progression factors by somatic gene transfer into a mouse model: Bcl-xL promotes islet tumor cell invasion. PLoS Biol. 16;5(10):e276. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Dunn GP, Old LJ, Schreiber RD. 2004. The three Es of cancer immunoediting. Annual review of immunology, 22:329–60. [DOI] [PubMed] [Google Scholar]
  6. Epardaud M, Elpek KG, Rubinstein MP, Yonekura AR, Bellemare-Pelletier A, Bronson R, Hamerman JA, Goldrath AW, Turley SJ. 2008. Interleukin-15/interleukin-15R alpha complexes promote destruction of established tumors by reviving tumor-resident CD8+ T cells. Cancer Res, 68(8):2972–83. [DOI] [PubMed] [Google Scholar]
  7. Grivennikov SI, Greten FR, Karin M. 2010. Immunity, inflammation, and cancer. Cell, 140(6):883–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Hodi FS, Lee S, McDermott DF, Rao UN, Butterfield LH, Tarhini AA, Leming P, Puzanov I, Shin D, Kirkwood JM. 2014. Ipilimumab plus sargramostim vs ipilimumab alone for treatment of metastatic melanoma: a randomized clinical trial. JAMA. 312(17):1744–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Janakiram M, Chinai JM, Fineberg S, Fiser A, Montagna C, Medaverepu R, Castano E, Jeon H, Ohaegbulam KC, Zhao R, et al. 2015. Expression, clinical significance, and receptor identification of the newest B7 family member HHLA2 protein Clinical cancer research : Clin Cancer Res. 23(3):825–832. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Jeon H, Vigdorovich V, Garrett-Thomson SC, Janakiram M, Ramagopal UA, Abadi YM, Lee JS, Scandiuzzi L, Ohaegbulam KC, Chinai JM, et al. 2014. Structure and cancer immunotherapy of the B7 family member B7x. Cell Rep. 9(3):1089–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Katz SC, Donkor C, Glasgow K, Pillarisetty VG, Gönen M, Espat NJ, Klimstra DS, D’Angelica MI, Allen PJ, Jarnagin W, et al. 2010. T cell infiltrate and outcome following resection of intermediate-grade primary neuroendocrine tumours and liver metastases. HPB (Oxford), 12(10):674–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Kobayashi A, Takahashi M, Imai H, Akiyama S, Sugiyama S, Komine K, Saijo K, Takahashi M, Takahashi S, Shirota H, et al. 2016. Attainment of a Long-term Favorable Outcome by Sunitinib Treatment for Pancreatic Neuroendocrine Tumor and Renal Cell Carcinoma Associated with von Hippel-Lindau Disease. Intern Med. 55(6):629–34. [DOI] [PubMed] [Google Scholar]
  13. Laitala A, Erler JT. 2018. Hypoxic Signalling in Tumour Stroma. Front Oncol. 8:189. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Lee JS, Scandiuzzi L, Ray A, Wei J, Hofmeyer KA, Abadi YM, Loke P, Lin J, Yuan J, Serreze DV, et al. 2012. B7x in the periphery abrogates pancreas-specific damage mediated by self-reactive CD8 T cells. Journal of immunology, 189(8):4165–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Maione F, Capano S, Regano D, Zentilin L, Giacca M, Casanovas O, Bussolino F, Serini G, Giraudo E. 2012. Semaphorin 3A overcomes cancer hypoxia and metastatic dissemination induced by antiangiogenic treatment in mice. J Clin Invest. 122(5):1832–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Noman MZ, Desantis G, Janji B, Hasmim M, Karray S, Dessen P, Bronte V, Chouaib S. 2014. PD-L1 is a novel direct target of HIF-1α, and its blockade under hypoxia enhanced MDSC-mediated T cell activation. J Exp Med. 211(5):781–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Palazon A, Goldrath AW, Nizet V, Johnson RS. 2014. HIF transcription factors, inflammation, and immunity. Immunity. 41(4):518–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Palazon A, Tyrakis PA, Macias D, Veliça P, Rundqvist H, Fitzpatrick S, Vojnovic N, Phan AT, Loman N, Hedenfalk I et al. 2017. An HIF-1α/VEGF-A Axis in Cytotoxic T Cells Regulates Tumor Progression. Cancer Cell. 32(5):669–683. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Pinato DJ, Tan TM, Toussi ST, Ramachandran R, Martin N, Meeran K, Ngo N, Dina R, Sharma R. 2014. An expression signature of the angiogenic response in gastrointestinal neuroendocrine tumours: correlation with tumour phenotype and survival outcomes. Br J Cancer. 110(1):115–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Pouyssegur J, Dayan F, Mazure NM. 2006. Hypoxia signalling in cancer and approaches to enforce tumour regression. Nature. 441:437–443 [DOI] [PubMed] [Google Scholar]
  21. Sato S, Tsuchikawa T, Nakamura T, Sato N, Tamoto E, Okamura K, Shichinohe T, Hirano S. 2014. Impact of the tumor microenvironment in predicting postoperative hepatic recurrence of pancreatic neuroendocrine tumors. Oncol Rep, 32(6):2753–9. [DOI] [PubMed] [Google Scholar]
  22. Scandiuzzi L, Ghosh K, Zang X. 2011. T cell costimulation and coinhibition: genetics and disease. Discovery medicine, 12(63):119–28. [PMC free article] [PubMed] [Google Scholar]
  23. Schreiber RD, Old LJ, Smyth MJ. 2011. Cancer immunoediting: integrating immunity’s roles in cancer suppression and promotion. Science. 331(6024):1565–70. [DOI] [PubMed] [Google Scholar]
  24. Sethumadhavan S, Silva M, Philbrook P, Nguyen T, Hatfield SM, Ohta A, Sitkovsky MV. 2017. Hypoxia and hypoxia-inducible factor (HIF) downregulate antigen-presenting MHC class I molecules limiting tumor cell recognition by T cells. PLoS One. 20; 12(11):e0187314. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Shen HC, He M, Powell A, Adem A, Lorang D, Heller C, Grover AC, Ylaya K, Hewitt SM, Marx SJ, et al. 2009. Recapitulation of pancreatic neuroendocrine tumors in human multiple endocrine neoplasia type I syndrome via Pdx1-directed inactivation of Men1. Cancer Res, 69(5):1858–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Shih K, Arkenau HT, Infante JR. 2014. Clinical impact of checkpoint inhibitors as novel cancer therapies. Drugs. 74:1993–2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Smith TL, Yuan Z, Cardó-Vila M, Sanchez Claros C, Adem A, Cui MH, Branch CA, Gelovani JG, Libutti SK, Sidman RL, et al. 2016. AAVP displaying octreotide for ligand-directed therapeutic transgene delivery in neuroendocrine tumors of the pancreas. Proc Natl Acad Sci U S A. 113(9):2466–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Soni S, Padwad YS. 2017. HIF-1 in cancer therapy: two decade long story of a transcription factor. Acta Oncol. 56(4):503–515. [DOI] [PubMed] [Google Scholar]
  29. Reck M, Rodríguez-Abreu D, Robinson AG, Hui R, Csőszi T, Fülöp A, Gottfried M, Peled N, Tafreshi A, Cuffe S, et al. 2016. KEYNOTE-024 Investigators. Pembrolizumab versus Chemotherapy for PD-L1-Positive Non-Small-Cell Lung Cancer. N Engl J Med. 375(19):1823–1833. [DOI] [PubMed] [Google Scholar]
  30. Vigdorovich V, Ramagopal UA, Lazar-Molnar E, Sylvestre E, Lee JS, Hofmeyer KA, Zang X, Nathenson SG, Almo SC. 2013. Structure and T cell inhibition properties of B7 family member, B7-H3. Structure. 21(5):707–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Wei J, Loke P, Zang X, Allison JP. 2011. Tissue-specific expression of B7x protects from CD4 T cell-mediated autoimmunity. The Journal of experimental medicine. 208(8):1683–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Westendorf AM, Skibbe K, Adamczyk A, Buer J, Geffers R, Hansen W, Pastille E, Jendrossek V. 2017. Hypoxia Enhances Immunosuppression by Inhibiting CD4+ Effector T Cell Function and Promoting Treg Activity. Cell Physiol Biochem. 41(4):1271–1284. [DOI] [PubMed] [Google Scholar]
  33. Xia X, Lemieux ME, Li W, Carroll JS, Brown M, Liu XS, Kung AL. 2009. Integrative analysis of HIF binding and transactivation reveals its role in maintaining histone methylation homeostasis. Proc Natl Acad Sci U S A. 106(11):4260–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Yao JC, Hassan M, Phan A, Dagohoy C, Leary C, Mares JE, Abdalla EK, Fleming JB, Vauthey JN, Rashid A, et al. 2008. One hundred years after “carcinoid”: Epidemiology of and prognostic factors for neuroendocrine tumors in 35,825 cases in the United States. Journal of Clinical Oncology, 26:3063–3072. [DOI] [PubMed] [Google Scholar]
  35. Zang X, Allison JP. 2007. The B7 family and cancer therapy: costimulation and coinhibition. Clinical cancer research, 13(18 Pt 1):5271–9. [DOI] [PubMed] [Google Scholar]
  36. Zang X, Loke P, Kim J, Murphy K, Waitz R, Allison JP. 2003. B7x: a widely expressed B7 family member that inhibits T cell activation. Proc Natl Acad Sci U S A, 100(18):10388–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Zang X, Sullivan PS, Soslow RA, Waitz R, Reuter VE, Wilton A, Thaler HT, Arul M, Slovin SF, Wei J, et al. 2010. Tumor associated endothelial expression of B7-H3 predicts survival in ovarian carcinomas. Modern pathology: 23(8):1104–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Zang X, Thompson RH, Al-Ahmadie HA, Serio AM, Reuter VE, Eastham JA, Scardino PT, Sharma P, Allison JP. 2007. B7-H3 and B7x are highly expressed in human prostate cancer and associated with disease spread and poor outcome. Proc Natl Acad Sci U S A. 104(49):19458–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Zhao R, Chinai JM, Buhl S, Scandiuzzi L, Ray A, Jeon H, Ohaegbulam KC, Ghosh K, Zhao A, Scharff MD, et al. 2013. HHLA2 is a member of the B7 family and inhibits human CD4 and CD8 T-cell function. Proceedings of the National Academy of Sciences of the United States of America, 110(24):9879–84. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp Table 1

Supplemental Table 1: Clinical characteristics and the expression of B7x and HHLA2 in the patients with neuroendocrine tumors. The expression of B7x and HHLA2 were analyzed in the patients with neuroendocrine tumors by immunohistochemistry (IHC) staining. The expression of B7x and HHLA2 were significantly increased in the later stage cancer to compared to early stage cancer (P=0.0014), but there is a significantly correlation between the expression of B7x and HHLA2 with age (P=0.5389) or gender (P=0.8523).

NIHMS1740848-supplement-Supp_Table_1.docx (13.5KB, docx)
Supp Table 3

Supplemental Table 3: Primer sequences for chromatin immunoprecipitation (ChIP)-PCR.

NIHMS1740848-supplement-Supp_Table_3.docx (12.2KB, docx)
Supp Table 2

Supplemental Table 2: Primer sequences for quantitative real-time RT-PCR.

NIHMS1740848-supplement-Supp_Table_2.docx (12.4KB, docx)
Supp Table 4

Supplemental Table 4: Clinical characteristics and expression of B7x and HHLA2 in the individual patients with neuroendocrine tumors.

NIHMS1740848-supplement-Supp_Table_4.docx (12.2KB, docx)
Supp Table 5
NIHMS1740848-supplement-Supp_Table_5.docx (14.1KB, docx)
Supp Figure 1B
NIHMS1740848-supplement-Supp_Figure_1B.tif (126KB, tif)
Supp Figure 1 A

Supplemental Figure 1. Expression levels of immune check-point proteins HHLA2 and B7x are correlated with clinical and pathological parameters.The expressed positive cell rates of HHLA2 and B7x are statistically higher in nonfunctional (42.6±12.8% and 45.2± 7.9%) and functional neuroendocrine tumors (NETs) (15.9±5.7% and 22.8±4.8%) compared to nonfunctional case matched normal tissues, 3.5±1.9% and 4.9±2.3%, respectively (P<0.001), and functional case matched normal tissues, 1.8±0.3% and 4.2±1.1%, respectively (P<0.001) by immunohistochemistry (IHC) staining (A1 and A2); the expressed positive cell rates of HHLA2 and B7x are significantly higher in primary tumor tissues with metastasis (80.5±5.4% and 90.3±8.2%) compared to tumor tissues without metastasis (A1 and A2); similar results are observed in mRNA expression by real-time RT-PCR analysis (A3 and A4). (N: Normal controls; T: Tumors). The average across all samples ± SD was used for statistics in the histograms. Statistically significant P values are indicated in the figures with asterisks: ***, P <0.001.

There is a correlation between the expressed positive cell rate of HHLA2 and B7x with tumor size in nonfunctional neuroendocrine tumors (NETs) (B1 and B2) and functional NETs (B5 and B6). There is also a correlation between the expressed positive cell rate of HHLA2 and B7x with tumor cell proliferation as measured by IHC with the Ki67 marker in nonfunctional NETS (B3 and B4) and functional NETs (B7 and B8).

NIHMS1740848-supplement-Supp_Figure_1_A.tif (96.8KB, tif)
Supp Figure 2

Supplemental Figure 2. Control normal tissues from Men1 wild type (WT) and knockout (KO) mice were analyzed for B7x expression byimmunohistochemistry(IHC). H&E staining of the control normal tissues (brain, heart, liver, spleen, kidney, and muscle) from Men1 WT (A1-A6) and KO mice (B1-B6). Negative IHC staining of B7x in the control normal tissues (brain, heart, lung, liver, spleen, kidney, and muscle) from Men1 WT (C1-C6) and KO mice (D1-D6).

NIHMS1740848-supplement-Supp_Figure_2.tif (1.4MB, tif)
Supp Figure 3

Supplemental Figure 3. B7x expression in the pancreatic neuroendocrine tumors (PNETs) with wild type (WT) menin from rat insulin promoter-T antigen (RIP-Tag) mice. H&E staining of the normal pancreatic tissue from WT control mice (A1) and PNET from RIP-Tag mice (A2). Negative immunohistochemistry (IHC) staining of B7x is shown in the normal pancreatic tissue from WT control mice (B1) and positive staining of B7x is shown in the PNET tissue from RIP-Tag mice (B2).

NIHMS1740848-supplement-Supp_Figure_3.tif (597KB, tif)
Supp Figure 5

Supplemental Figure 5. The expression levels of HIF-1α and B7x in N134 cells. The expression levels of HIF-1α and B7x were not induced in the islet tumor β cells following HIF-1α siRNA treatment after 48 hours by western blot. We provided the entire films for the three antibodies in theSupplemental Figure 5.

NIHMS1740848-supplement-Supp_Figure_5.tif (2.3MB, tif)
Supp Figure 4

Supplemental Figure 4. Anti-B7x mAb increased expression of cytokines in the tumor microenvironment. The levels of interferon-γ (IFN-γ) (B) and tumor necrosis factor-α (TNF-α) (D) are significantly increased in the tumor microenvironment in the B7x treatment group compared with the IgG control (A and C) as shown by immunofluorescence (IF) staining, immunostaining of insulin (pink), IFN-γ/TNF-α (green), and nuclei counterstained with DAPI (blue). The images were taken with the 10X objective.

NIHMS1740848-supplement-Supp_Figure_4.tif (4.1MB, tif)

RESOURCES