Mutated Von Hippel-Lindau-renal cell carcinoma (RCC) promotes patients specific natural killer (NK) cytotoxicity

Background Previous evidence demonstrated that restoration of wild type VHL in human renal cancer cells decreased in vitro NK susceptibility. To investigate on the role of tumoral VHL status versus NK capability in renal cancer patients, 51 RCC patients were characterized for VHL mutational status and NK function. Methods VHL mutational status was determined by direct DNA sequencing on tumor tissue. NK cytotoxicity was measured against specific target cells K562, VHL-wild type (CAKI-1) and VHL-mutated (A498) human renal cancer cells through externalization of CD107a and IFN-γ production. Activating NK receptors, NKp30, NKp44, NKp46, NKG2D, DNAM-1, NCAM-1 and FcγRIIIa were evaluated through quantitative RT-PCR. RCC tumoral Tregs were characterized as CD4+CD25+CD127lowFoxp3+ and Treg function was evaluated as inhibition of T-effector proliferation. Results VHL mutations were detected in 26/55 (47%) RCC patients. IL-2 activated whole-blood samples (28 VHL-WT-RCC and 23 VHL-MUT-RCC) were evaluated for NK cytotoxicity toward human renal cancer cells A498, VHL-MUT and CAKI-1, VHL-WT. Efficient NK degranulation and increase in IFN-γ production was detected when IL-2 activated whole-blood from VHL-MUT-RCC patients were tested toward A498 as compared to CAKI-1 cells (CD107a+NK: 7 ± 2% vs 1 ± 0.41%, p = 0.015; IFN-γ+NK: 6.26 ± 3.4% vs 1.78 ± 0.9% respectively). In addition, IL-2 activated NKs induced higher CD107a exposure in the presence of RCC autologous tumor cells or A498 as compared to SN12C (average CD107a+NK: 4.7 and 2.7% vs 0.3% respectively at 10E:1 T ratio). VHL-MUT-RCC tumors were NKp46+ cells infiltrated and expressed high NKp30 and NKp46 receptors as compared to VHL-WT-RCC tumors. A significant lower number of Tregs was detected in the tumor microenvironment of 13 VHL-MUT-RCC as compared to 13 VHL-WT-RCC tumors (1.84 ± 0.36% vs 3.79 ± 0.74% respectively, p = 0.04). Tregs isolated from VHL-MUT-RCC patients were less suppressive of patients T effector proliferation compared to Tregs from VHL-WT-RCC patients (Teff proliferation: 6.7 ± 3.9% vs 2.8 ± 1.1%). Conclusions VHL tumoral mutations improve NKs effectiveness in RCC patients and need to be considered in the evaluation of immune response. Moreover therapeutic strategies designed to target NK cells could be beneficial in VHL-mutated-RCCs alone or in association with immune checkpoints inhibitors. Electronic supplementary material The online version of this article (10.1186/s13046-018-0952-7) contains supplementary material, which is available to authorized users.

T regulatory cells (Tregs) partecipate in the control of tumor immunity attenuating antitumor immune response [17]. Increase in peripheral and tumoral Tregs were observed in several cancers and inversely correlated to NKs number and function [18]. Tregs mainly regulate NK activity in a transforming growth factor-beta (TGF-β)-dependent manner. Soluble and membrane-bound TGF-β on regulatory T cells impaired NK cytolytic activity through down-regulation of NKG2D receptor on the NK cells [19].
To investigate on the role of tumoral VHL status versus NK capability in renal cancer patients, 51 RCC patients were characterized for VHL mutational status and NK cytolytic activity toward A498-VHL-MUT and CAKI-1-VHL-WT human renal cancer cells.

Patients and samples
Fifty five patients with primary RCC (41 clear cell; 4 papillary; 5 cromophobe; 5 unknown) and twelve healthy donors (HD) were enrolled in the study. The patients underwent surgery as part of their standard treatment at the Department of Urology, National Cancer Institute "G. Pascale" (Naples, Italy) and Genitourinary Oncology and Rare Cancer Center, Federico II University (Naples, Italy). RCC tumor and peritumoral tissues were obtained immediately after surgical resection, stabilized in RNA later (Qiagen) and stored at − 80°C. The distance of 1 cm was the minimal distance between tumor and normal-appearing renal tissue sampled. Informed consent from each patient was sought. The research protocol was approved by Human Ethical Committee of Institute (n. CEI/423/13).

CD107a degranulation assay and IFN-γ production
Lysosome-associated membrane protein LAMP-1 (CD107a) is a marker of NK-cell degranulation. CD107a expression was measured by flow cytometry as previously described [23,24]. Briefly, 1 ml whole blood was diluted with one volume RPMI-1640 containing 10% heat-inactivated fetal bovine serum and incubated with IL-2 (100 U/ml) overnight at 37°C, 5% CO 2 . The cytotoxic activity of NK cells was tested against NK-sensitive cell line K562 (HLA class I-negative) [25], human renal cancer cell lines CAKI-1-VHL-WT, A498 and 786-OVHL-MUT. 200 μl of IL-2 activated blood was co-cultured with 2 × 10 5 target cell lines in presence of PE-anti-CD107a antibody for 3-h at 37°C in 5% CO 2 . Alternatively, purified IL-2 activated NK cells from peripheral blood of RCC patients were used as effector cells. The autologous RCC tumor cells, A498-VHL-MUT and SN12C-VHL-WT cell lines were used as target cells. Assays were performed at the indicated effector-target ratio (10E:1 T; 5E:1 T). To detect spontaneous degranulation samples were incubated in the absence of target cells. Following 3-h culture, cells were stained with anti-CD56 and NK cells were identified as CD3 − CD56 + in the lymphocyte gate. CD107a was analyzed on CD56 + CD3 − NK. For intracellular staining of IFN-γ, 200 μl of IL-2 preactivated blood samples were incubated with 2 × 10 5 K562 or RCC target cells for 6 h at 37°C in a humidified 5% CO 2 incubator. The secretion inhibitor Brefeldin A (Sigma-Aldrich, Seelze, Germany) was added to the assay medium at a concentration of 10 μg/ml during the 6 h coincubation. After incubation, samples were placed on ice and stained with anti-CD56 and anti-CD3 mAbs, followed by erythrocyte lysis of blood samples with FACS Lysing Solution (BD Biosciences). Thereafter, cells were permeabilized using FACS permeabilizing solution (Perm2, BD Biosciences) and stained for intracellular IFN-γ with a PE-anti-IFN-γ antibody. Analysis was performed on a FACSAria II Cell Sorter 8-colour flow cytometer, blinded to patients' clinical characteristics.

Real-time PCR
RNA was isolated from tumoral and peritumoral RCC tissues using the RNeasy Mini kit (Qiagen, Hilden, Germany), according to manufacturer's instructions. cDNA was synthesized from 1 μg of total RNA using SuperScript III Reverse Transciptase (Invitrogen). Quantitative real-time PCR was performed using SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA) and data were collected and quantitatively analyzed on an ABI Prism 7000 System (Applied Biosystems). Primers sequences for NCAM-1, FCGR3a (CD16), DNAM-1, NKG2D, NKp46, NKp44, NKp30,KLRB1 and TGFB1 are detailed in Additional file 1: Table S1. Relative mRNA expression was normalized with GUSB gene expression.

Immunohistochemistry (IHC)
Paraffin-embedded RCC tissue blocks, after heat induced epitope retrieval with citrate buffer (pH 6), were incubated overnight at 4°C with the mouse monoclonal IgG2B, anti-human NKp46 antibody, Clone 195,314 (MAB1850 R&D Systems) diluted 1:50, followed by incubation with peroxidase-conjugated secondary antibodies according to the manufacturer's instructions (EnVision+System, Dako North America, Inc., Carpinteria, CA). The NKp46 receptor demonstrates a high degree of lineage-specificity, being expressed almost exclusively in natural killer cells. . Five randomly selected areas were identified across the entire section from the center to the invasive margin (100× magnification). NKp46 + cells were counted in 5 consecutive non-overlapping high-power fields (HPF) 400x magnification (0.237 mm 2 /field), using an Olympus BX51 microscope (Olympus, Tokyo, Japan). The results were expressed as the mean of NKp46+/mm 2 out of 5 regions of interest. Immunostaining and scoring were evaluated by 4 independent observers (C.D., G.B., E.L. and A.A.) blinded to patients' clinical characteristics.

Cytokine assay
IFN-γ was measured by ELISA on the culture supernatant collected on day 5 from suppression experiments. Cytokine concentration was assessed by Human IFNgamma Instant ELISA (Bender MedSystems). Samples were acquired by LB 940 Multimode Reader Mithras (Berthold Technologies).

Statistical analysis
Statistical analysis were performed using the MedCalc 9.3.7.0 and Excel software. Unpaired Student t test or non-parametrical Mann-Whitney test was conducted. Statistical tests of hypotheses was two-sided with p < 0.05 taken as the criterion for statistical significance.

Discussion
In the present study RCC-VHL mutational status was correlated to patients NK cytotoxicity through ex vivo evaluation toward human renal cancer cell lines A498  [31,32]. These differences might accomplish for different phenotypes affecting NK sensitivity [15]. Nowadays the effects of VHL mutations on NK activity is controversial. Previous evidence demonstrated that restoration of wild type VHL in human renal cancer cells decreased their NK susceptibility [33] while VHL mutations increased resistance to NK-mediated lysis through EPAS1/HIF-2α stabilization [16,34]. NK cells induced a contact-dependent autophagy in ccRCC cells that was mediated by the HIF-2a targeted, inositol triphosphate receptor1 (ITPR1) in tumor cells. Blocking ITPR1 expression in ccRCC cells inhibited NK cell-induced autophagy and suppressed ccRCC resistance to NK cells [35].
In our study blood derived from patients carrying RCC-VHL-MUT tumors display better NK susceptibility toward human renal cancer cells VHL-MUT suggesting an improvement in the innate immune response toward RCC.
To further correlate peripheral and tumoral NK status, NK cell infiltration was analyzed through the evaluation of NKp46 + cells and the NKs activating receptors NKp30, NKp44, NKp46, NKG2D, DNAM-1, NCAM-1 and FcγRIIIa. NK cell-activating ligands expression on A498 and CAKI-1 cells was recently reported [27][28][29]. Although a significant increase in NKG2D+ cells was not detected in our tested cells, A498 and CAKI-1 express high levels of DNAM-1 ligand PVR (94.8 and 99.1% respectively, Additional file 2: Figure S5). Previous authors stated that VHL-mutational status may not affect the expression of ligands for NKG2D and DNAM-1 activating NK receptors [33]. Higher infiltration of NKp46 + cells and increased expression of NKp30 and NKp46 [36][37][38] were detected in VHL-MUT-RCC compared to VHL-WT-RCC. Although the NKp46 receptor demonstrates a high degree of lineagespecificity, being expressed almost exclusively in natural killer cells [39] the presence of a rare NKp46 + CD3 + population was reported. These include rare subsets of T cells (αβ and δγ) in humans and mice and a subset of group 3 innate lymphoid cells (ILCs) (NCR + ILC3) [36,40]. Also DNAM1 or NKG2D may be expressed by T cells [36,40,41]. This is a limit of the present study. However the concomitant expression of NKp46 plus NKp30, NKp44 and DNAM-1 strongly suggest the hypothesis that NKs are involved.
To better define the role of NK cells in RCC tumors, Tregs and NKs are currently evaluated within the study REVOLUTION aimed at identify biomarkers of the checkpoint inhibitor nivolumab efficacy in metastatic renal cancer patients. Preliminary evidence showed that while the absolute value of CD3 − CD56 dim was not affected by nivolumab treatment NK cytotoxicity correlated with the responsiveness to treatment (Proceedings AACR 2017: Abstract #580).
Tregs suppress NK cell effector functions [42] impairing TGF-β pathway which, among other effects, induces downregulation of the activating NK cell receptor NKG2D [43]. Herein we report high infiltration of NKp46 + cells and reduced CD4 + CD25 hi Foxp3 + T cells in VHL-MUT-RCC tumors, although the reduction of TGF-β levels did not correspond to increased expression of NKG2D gene in VHL-MUT-RCC tumors. As in our results, Shen showed no significant correlation between peripheral Tregs and NKG2D expression in colorectal cancer patients NK cells [44]. Since VHL mutations induces PD-L1 expression in RCC, NK function may be a crucial element in nivolumab sensitivity [45]. NKs isolated from renal carcinoma patients expressed PD-1 on their surface and engagement of PD-1 signaling reduced their cytolytic potential. Treatment of patient-derived PD-1 + NK cells with an anti-PD-1 antibody (pidilizumab, CT-011) increased NK killing of autologous cancer cells in vitro [46]. NK cells also engage antibody-dependent cell mediated cytotoxicity (ADCC) and ex vivo treatment with the anti-CTLA4 ipilimumab enhances cetuximabmediated ADCC targeting CTLA-4 + Treg in head and neck squamous cell cancer (SCCHN) [47,48]. Nowadays the immune checkpoint targeting agents nivolumab and ipilimumab represent a clinical standard for mRCC. Nevertheless tyrosine kinase inhibitors (TKI) still play a central role in metastatic RCC treament [49] and modify the tumor immune sensitivity. TKIs affect the frequency and composition of tumor immune cell subpopulations including Tregs, myeloid-derived suppressor cells as well as T and NK cells [50][51][52]. Sorafenib significantly reduce NKs activity against the RCC cell lines A498, ACHN and CAKI-2 [53] while axitinib promotes NK cell recognition and degranulation toward A498 RCC cells in a ROS-dependent manner [28].

Conclusions
VHL tumoral mutations improve the NKs effectiveness in RCC patients and thus need to be considered in the evaluation of TKIs or immune based therapies. Moreover therapeutic strategies designed to target NK cells could be beneficial in VHL-mutated-RCCs alone or in association with immune checkpoints inhibitors.