Targeting CXCR4 reverts the suppressive activity of T-regulatory cells in renal cancer

With the intent to identify biomarkers in renal cell carcinoma (RCC) the functional status of T-regulatory cells (Tregs) was investigated in primary RCC. Tregs were isolated from tumoral-(TT), peritumoral tissue-(PT) and peripheral blood-(PB) of 42 primary RCC patients and function evaluated through effector T cells (Teff) proliferation, cytokines release and demethylation of Treg Specific Region (TSDR). The highest value of Tregs was detected in TT with the uppermost amount of effector-Tregs-(CD4+CD25hiFOXP3hiCD45RA-). PB-RCC Tregs efficiently suppress Teff proliferation compared to healthy donor (HD)-Tregs and, at the intrapatient evaluation, TT-derived Tregs were the most suppressive. Higher demethylation TSDR was detected in TT- and PB-RCC Tregs vs HD-Tregs (P <0,001). CXCR4 is highly expressed on Tregs, thus we wished to modulate Tregs function through CXCR4 inhibition. CXCR4 antagonism, elicited by a new peptidic antagonist, Peptide-R29, efficiently reversed Tregs suppression of Teff proliferation. Thus Tregs functional evaluation precisely reflects Tregs status and may be a reliable biomarker of tumoral immune response. In addition, treatment with CXCR4 antagonist, impairing Tregs function, could improve the anticancer immune response, in combination with conventional therapy and/or immunotherapy such as checkpoints inhibitors.


INTRODUCTION
Renal cell carcinomas (RCCs) is a disease with an estimated 338,000 new cases diagnosed worldwide [1] with approximately 30% of patients presenting with metastatic disease. High-dose IL-2 and Interferon-α were the principal therapies for metastatic RCC; due to the harsh-tolerance profiles and limited response rates, immunotherapies were replaced with targeted therapies but only everolimus has an impact on overall survival (NCCN guidelines Version 2.2016). The human IgG4 anti-PD-1 monoclonal antibody, Nivolumab, was recently approved to treat advanced (metastatic) clear cell renal cell carcinoma (ccRCC) patients who have received a prior anti-angiogenic therapy [2] resulting in 30% of www.impactjournals.com/oncotarget/ Oncotarget, 2017, Vol. 8, (No. 44), pp: 77110-77120 Research Paper www.impactjournals.com/oncotarget clinical response. Immunotherapy efficacy is controlled by multiple mechanisms such as the recruitment of immunosuppressive cells to the tumor microenvironment (Tregs, Myeloid derived suppressor cells-MDSCs) [3]. T regulatory cells (Tregs) suppress a whole range of immune cells including B cells, NK cells, NKT cells, CD4 + or CD8 + T cells, monocytes and dendritic cells [4]. Nevertheless, Tregs role in cancer microenvironment is controversial: Tregs are associated with poor prognosis in renal [5], ovarian [6], pancreatic [7] and liver cancer [8], while others showed no correlation between tumorinfiltrating Tregs and patient outcome [9]. Tregs are identified through the surface expression of CD3, CD4, CD25, FOXP3, CD127 and CD45RA. In addition ICOS, a member of CD28/CTLA-4 family and CD39/ENTPD1 defines a highly suppressive Treg population [10]. CTLA-4 and PD-1 have also been proposed as key molecules in generation and/or suppressive function of Tregs [11]. Cancer patients derived Tregs usually express a distinct profile of chemokine receptors, such as CCR4, CXCR4 and CCR5, which facilitates their migration into tumors in response to the corresponding chemokine ligands derived from tumor microenvironment [12]. With the intent to characterize the local antitumor immune response Tregs were functionally evaluated in 42 RCC patients. Moreover, CXCR4 antagonism was investigated as a strategy to impair Treg efficiency.

Tregs isolated from primary tumor (TT), peritumoral tissue (PT) and peripheral blood (PB) of renal cancer (RCC) patients are highly activated
Tregs, identified as percentage of CD4 + CD25 hi FOXP3 + cells, were evaluated in 42 consecutive renal cancer patients (Table 1). A significantly higher number of Tregs was detected in 42 RCC-PB as compared to 15 healthy donors (HD) (Figure 1Ai) Intrapatients analysis demonstrated higher Tregs in TT versus PT and PB. In Figure 1B a representative plot of HD-Tregs and RCC-Tregs was shown. As shown in Figure 1C-1D, TT-Tregs displayed the highest level of CTLA-4, ICOS, ENTPD1, CD45RA, PD-1 and CXCR4. According to the Tregs functional classification that distinguish CD25 hi FOXP3 hi CD45RAas effector and CD25 hi FOXP3 low CD45RAas not suppressive Tregs [13], the highest number of effector Tregs was detected in TT (PB and PT vs TT, P <0,001) ( Figure 2A); in Figure 2B a representative analysis of Tregs subpopulations is shown (PB/PT/TT). PB-Tregs from RCC patients cocultured with autologous Teff cells more efficiently suppress Teff proliferation compared to HD-Tregs ( Figure 3A-3B). In Figure 3C a representative analysis of CFSE-labeled Teff proliferation-Treg dependent was shown. As control, anti-CD3/CD28-stimulated CD4 + CD25 + T cells were anergic while Teff intensively proliferated (Supplementary Figure 1).

TT-Tregs are more suppressive than PB-and PT-Tregs in RCC patients
In Figure 4A, TT-, PT-and PB-isolated Tregs significantly suppressed autologous Teff cells proliferation (P <0,001). In particular, TT-Tregs more efficiently suppressed T-effector proliferation compared to PT-and PB-Tregs ( Figure 4A). In Figure 4B a representative suppression assay was shown. Treg function is also regulated through the status of methylation of Tregspecific demethylated region (TSDR). Thus FOXP3-TSDR demethylation status was investigated on TT-and PB-Tregs from RCC patients and compared to HD. High demethylation rate was detected in TT-RCC and PB-RCC vs HD derived Treg cells (P <0,001) ( Figure 5A). As reference sequence the methylation of CpG sequences of IFN transcription regulatory factors 8 (IRF8) was considered [14,15]. Moreover to evaluate Tregs function culture supernatants were evaluated for IFN-γ and TGF-β1 on day 5 of cocolture. As shown in Figure 5B, a significant decrease of IFN-γ was observed when TT-Tregs were added to autologous Teff cells. Of interest, very low   IFN-γ production was observed in cultures with PB-Tregs. Surprisingly, a dramatic increase in IFN-γ was observed in coculture of PT-Tregs. This increase could be ascribed to IFN-γ production from tumor infiltrating lymphocytes (TIL) that induce inhibitory T cell ligands such as PD-L1 [16]. As expected, a significant increase of TGF-β1 was observed when TT-Tregs were added to autologous Teff cells ( Figure 5C). TGF-β1 and IL-10 mRNA expression was also evaluated in TT and PT RCC tissues. Consistent with an increased Treg function, TGF-β1 expression, though not significant, increased in tumoral tissues (PT vs TT: 0.06±0.08 vs 0.20±0.30) while IL-10 expression dramatically increased in TT as compared to PT samples (tumoral vs peritumoral: 0.36±0.37 vs 0.08±0.14, p<0,01) (Supplementary Figure 2). Altogether these data suggest that, although TT-Tregs share some phenotypic similarities with both PT-and PB-Tregs, they are functionally more suppressive.

DISCUSSION
Tumors actively escape the immune system through complex mechanisms including inhibitory receptors, secreted soluble inhibitors (IL-10 and TGF-β) and Tregs [19]. In the present study Tregs were evaluated in 42 RCC patients: higher number of Tregs was detected in tumoral tissue as compared to peritumoral and peripheral Tregs as previously described in patients with head and neck squamous cell carcinoma, breast, pancreatic, stomach and liver cancer [20]. In here, Tregs expressing CTLA-4, ICOS, ENTPD1, CD45RA and PD-1 were the majority among tumoral Tregs and the so called effector Tregs, CD25 hi FOXP3 hi CD45RAhighly suppressive cells, were the most represented among tumor Tregs tissue derived [13]. Higher number of intratumoral Tregs with activated phenotype was previously reported in non-small-cell lung cancer patients and rectal cancer [21]. The elevated number of Tregs in RCC environment raises the question of their immunosuppressive capability. We found that tumoral Tregs display higher immunosuppressive activity compared to peritumoral and peripheral Tregs as demonstrated by reduced Teff proliferation and IFN-γ production. The high IFN-γ production in the cocolture of peritumoral Treg and T effectors can probably be ascribed to IFN-γ production from TIL cells that induce inhibitory T cell ligands such as PD-L1 [16]. Moreover, high Treg activity was supported by TGF-β production in culture supernatants [10]. These data suggest that functional Tregs accumulated in the RCC tumors control tumor immunity. Tregs function is regulated through methylation of Tregspecific demethylated region (TSDR). TSDR is a CpG dinucleotide rich domain and highly conserved region within the conserved non coding sequences 2 (CNS2), located in the first intron of FOXP3 gene [22]. TSDRhypermetilated Tregs impaired suppressive function [23]; recent evidence demonstrated TSDR demethylation in liver biopsy and peripheral blood samples from patients with advanced grades of HCC compared to control subjects with nonmalignant disease [24]. Herein we show that TSDR demethylation increases in RCC isolated Tregs suggesting a higher suppressive function compared to healthy donors isolated Tregs. These results were compared to the demethylation induced by DNMTi azacytidine (AZA) and IFN-γ on unrelated sequence in the promoter of the Interferon Regulated Factor IRF-8 [14,15]. Tregs from RCC patients express higher CXCR4 compared with Tregs from healthy donors. CXCR4 receptor is activated by the ligand CXCL12 that regulates T cell access and recruits immunosuppressive population to tumoral microenvironment including interleukin-10producing plasmacytoid dendritic cells, Tregs and MDSCs. In a mouse model of ovarian cancer the CXCR4 antagonist, AMD3100, determined several anti-tumor effects including increased tumor cell death, reduced dissemination and better survival of the treated animals. Significantly, it was observed a selective reduction in the recruitment of Foxp3 + T cells in comparison with CD8 + T cells [25]. Blockade of both CXCR4 and PD-1 prevents suppression of immune cell function in HCC tumors, enhances immune cell tumor penetration and activation, and ultimately delays HCC progression [26]. In our experience CXCR4 antagonism with AMD3100 or with a new class of CXCR4 antagonists (Peptide R29) reverts Treg suppressive function in RCC patients. CXCR4 transduction may regulate some genes critical in Treg function since CXCL12/CXCR4 signaling plays a key role in human thymocyte development, promoting survival and expansion of human early T-cell progenitors [27].
The mechanism through which CXCR4 antagonists impact on Treg function is under evaluation. A possibility is that CXCR4 transduction impairs the methylation status of FOXP3. In metastatic renal carcinoma IL-2 based immunotherapy leads to substantial expansion of T-cells with a demethylated TSDR. Preliminary evidence displayed a reduced demethylation in TSDR when RCC derived Tregs were treated with CXCR4 antagonists (Santagata et al, manuscript in preparation). Post translational modification such as acetylation can regulate Treg-FOXP3 activity [28]. Because CXCR4 signaling transduces also on FOXP3 and HDACis upregulated CXCR4 mRNA expression [29], it is possible that CXCR4 modulation affects the acetylation status of FOXP3 gene promoter.
Although the anti PD-1 Nivolumab was recently approved to treat advanced (metastatic) clear cell renal cell carcinoma (ccRCC), only a minority of patients display a durable anti PD-1 response. This effect might be due to the tumor microenvironment lacking the efficient T-effector cells access to tumors and to a recruitment of immunosuppressive cells. Targeting the CXCR4-CXCL12 axis could revert the tolerogenic polarization of the microenvironment rich of immunosuppressive cells such as Tregs cells M2 and N2 neutrophils [25,30,26], improving immunotherapeutic intervention in patient with RCC. Preliminary evidence from an ongoing observational trial demonstrates that Tregs functional status in peripheral blood of nivolumab treated metastatic renal cancer patients is predictive of drug sensitivity [31]. Although the number of Tregs isolated from peripheral blood was adequate to functionally evaluate the role of Tregs in RCC patients, the Tregs isolated from primary tumors was a limiting factor. Moreover the interpatients variability required an adequate number of patients to draw conclusions. In conclusion RCC patients at peripheral, tumor and peritumoral sites, contain high numbers of functional Tregs and intratumoral Tregs actively limits the tumoral immune response. Thus monitoring Treg function also in peripheral blood could be informative on the antitumor immune response and possibly predictive of immunotherapy response. Moreover, CXCR4 antagonism is able to reverse in vitro Tregs suppressive capability on effector T cells proliferation suggesting that targeting CXCR4 may improve immunotherapy in RCC.

Patients and specimens
RCC samples were collected from 42 RCC patients undergone partial/radical nephrectomy at the Urology Unit of Istituto Nazionale per lo Studio e la Cura dei Tumori, Fondazione "G. Pascale" and Genitourinary Oncology and Rare Cancer Center, Federico II University in Naples. Renal tumor (TT) and peritumoral tissue (PT) specimens were collected at the time of surgery.
The distance 1cm was the minimal distance between tumor and normal-appearing renal tissue sampled (PT). In addition, 8-mL of heparinized peripheral blood (PB) was collected from each patient before surgery. Heparinized blood was also collected from 15 healthy donors (HD). Patients features are shown in Table 1. The research protocol was approved by Human Ethical Committee of Institute (n. CEI/423/13).

Cells preparation
Peripheral blood samples from cancer patients and healthy donors were drawn into heparinized tubes (BD Biosciences) and centrifuged on Ficoll-Hypaque gradients (GE Healthcare Bioscience). For tissue-infiltrating lymphocytes isolation, freshly isolated tumors sample and corresponding peritumoral renal tissue biopsies were minced into small pieces, digested with 1 mg/mL of collagenase (Sigma-Aldrich) for 30 minutes at 37°C, then transferred to a cell strainer (70μm Nylon) (BD Biosciences) and gently separated by using a syringe plunge. The collected cells underwent to Ficoll-Hypaque gradient centrifugation, washed and immediately used for experiments.

Purification of T cell subsets
Peripheral, tumor and peritumoral CD4 + CD25 + Tregs and peripheral CD4 + CD25 − T effector (Teff) cells were isolated using the Dynabeads Regulatory CD4 + CD25 + T cell kit. Briefly, CD4 + cells were separated by negative selection, using the antibody mix human CD4. In the second step, a depletion beads solution was added to remove the non-CD4 cells. Then, CD25-beads was added to CD4 + T cells to capture the CD4 + CD25 + Tregs and the remaining fraction was used as CD4 + CD25 − Teff cells. Finally, Dynabeads CD25 was removed from the cells. All purification steps were performed according to the manufacturer's instructions (Invitrogen by Life Technologies) and collected cells were found to be > 95% pure by flow cytometry.

Methylation studies: genomic DNA isolation, bisulfite conversion, and qPCR
Genomic DNA (gDNA) of peripheral Tregs isolated from 4 HD and 5 RCC patients was obtained by using a traditional phenol/chloroform extraction method with minor modifications. Bisulfite treatment of 500 ng genomic DNA was performed by using the EZ DNA Methylation™ Kit (ZYMO Research) according to the manufacturer's instructions. qPCR was prepared by using SensiMix SYBR Kit (Bioline, London) and performed by LightCycler ® 480 System (Roche Diagnostics). Primers for methylation and demethylation-specific FOXP3-TSDR and computing of the demethylation rate (DMR) of FOXP3-TSDR were previously described [12]. Briefly, we used the following formula: 100/[1 + 2(CtTG-CtCG)] × 100%, where CtTG represents the cycle threshold (Ct) achieved with TG (demethylated) primers and CtCG represents the Ct achieved with CG (methylated) primers. For female patients, this rate was corrected by a factor of 2 because one of the two TSDR alleles is methylated as a result of X inactivation.

Cytokine assay
IFN-γ and TGF-β1 were measured by ELISA assay on the culture supernatant collected on day 5 from suppression experiments. In particular, cytokine concentration was assessed by Human IFN-gamma Instant ELISA (Bender MedSystems) and Human TGF-β1 ELISA kit (Boster Biological Technology Co). Samples were acquired by LB 940 Multimode Reader Mithras (Berthold Technologies).
Samples were run in triplicate, and their relative expression was calculated in the following formula using GUSB as endogenous control: 2 -ΔΔCt .

Author contributions
SS and MN designed, performed, and analyzed data, and wrote the manuscript. CD contributed to experimental design and data interpretation. SD and FP organized the recruitment, consent and provided information on patients. SDM, EN and LM provided the peptide and editing the manuscript. AF, MB performed experiments and analyzed methylation data. LG contributed to data interpretation and editing the manuscript. NL, S. Pignataand S. Perdonà provided patients sample, patients information and intellectual input and critically reviewed the manuscript. S. Scala conceived of, designed, supervised the project and experimental plan, interpreted experiments, and wrote the manuscript.