Gr1−/lowCD11b−/lowMHCII+ myeloid cells boost T cell anti‐tumor efficacy

Abstract Conventional APCs that express MHC class II (MHCII) and co‐stimulatory molecules include dendritic cells (DCs) and macrophages. Beyond these conventional APCs, immune stimulatory cells have been more recently shown to extend to a class of atypical APCs, composed of mast cells, basophils, and eosinophils. Here, we describe a unique type of APC, Gr1−/lowCD11b−/low cells with a granularity and size characteristic of myeloid cells and with the ability to present Ag for crosspresentation. These cells constitutively express MHCII and the costimulatory molecules, CD80, CD86, and CD40. They do not express pan markers of myeloid DCs (CD11c), plasmacytoid DCs (Ly6C), or macrophages (F4/80), and their frequency is inversely correlated with myeloid‐derived suppressor cells (MDSCs) in tumor‐bearing mice. Among splenocytes, they are more abundant than DCs and macrophages, and they exhibit antitumor immune stimulatory function at a steady state without further activation, ex vivo. They are also found within the tumor bed where they retain their immune stimulatory function. Our findings suggest the use of these novel APCs in additional preclinical studies to further investigate their utility in APC‐based cancer immunotherapies.


INTRODUCTION
Dendritic cells (DCs) play a central role in inducing immune responses against infectious diseases and cancer. However, their efficacy as Abbreviations: AIT, adoptive immunotherapy; DC, Dendritic cell; ILC, innate lymphoid cell; MDSC, myeloid-derived suppressor cell responses to cancer and infectious diseases such as tuberculosis, 1,2 AIDS, [3][4][5] hepatitis C, 6,7 hepatitis B, 8,9 pneumonia, 10,11 and Staphylococcus aurous infection. 12 Importantly, an elevation of MDSCs is associated with a reduced efficacy of vaccines. 13,14 In addition, the generation of monocyte-derived DCs or bone marrow-derived DCs requires extensive ex vivo culturing, conceivably hampering the immunogenicity of the vaccine. Recent studies, therefore, have focused on vaccines that make use of primary DCs. 15 For instance, Sipuleucel-T is the only FDA-approved therapeutic vaccine for metastatic prostate cancer. 16 The vaccine uses readily isolated circulating DCs cultured with prostate tumor Ag and GM-CSF. However, circulating DCs are very rare and tumor-induced immune suppressive cells, such as MDSCs, limit their efficacy in inducing a sustained antitumor immune response. Therefore, there is an urgent need to identify a new class of APC that are highly efficient in orchestrating profound antitumor immunity to facilitate the development of a new class of cell-based cancer vaccines.
In recent years, there has been a rapid increase in our understanding of the biology of cells with APC characteristics, namely the ability to activate T cells. For instance, mouse neutrophils can induce Th1 and Th17 responses 17,18 and tumor-associated neutrophils have been demonstrated to stimulate T cell responses in early-stage human lung cancer. 19 A recent review discusses a number of atypical APCs including mast cells, basophils, eosinophils, and innate lymphoid cells (ILC). 20,21 However, these APCs are rare in the circulation and their maintenance of effective antitumor immune responses is likely to be inhibited due to high frequencies of MDSCs in locations of T cell priming. Very recently, it was reported that activated NKT cells decrease the frequency and immunosuppressive activity of MDSCs in tumor-bearing mice. 22 In an animal model, activated NKT cells converted MDSCs into immunogenic APCs. 23 Using peripheral blood mononuclear cells (PBMC) of patients with early stage breast cancer, we also demonstrated that conversion of MDSCs to CD33 + CD11b −/low HLA-DR + APCs, in vitro, was associated with an increased frequency of CD25+ NKT cells in reprogrammed immune cells. 24 In an effort to understand this MDSC-APC axis during the application of adoptive immunotherapy (AIT) to treat breast cancer, we identified a class of Gr1 −/low CD11b −/low MHCII+ APCs. These cells retain their immune stimulatory function during tumor progression and are inversely correlated to the frequency of splenic and tumor-infiltrating MDSCs. Importantly, we identified the presence of these cells in nonpathological conditions, whereupon we confirmed their ability to cross-present Ag to stimulate T cells. Therefore, these APCs offer a potentially novel APC-based vaccine for cancer therapy.

Mouse model
FVBN202 transgenic female mice (The Jackson Laboratory; Bar Harbor, ME) were used between 8 and 12 weeks of age throughout these experiments. These mice overexpress a nonmutated, nonactivated rat neu transgene under the regulation of the mouse mammary tumor virus promoter. 25 These mice develop premalignant mammary hyperplasia similar to ductal carcinoma in situ prior to the development of spontaneous carcinoma. 26 Premalignant events in FVBN202 mice include the accumulation of endogenous MDSCs. 26 These studies have been reviewed and approved by the Institutional Animal Care and Use Committee at Virginia Commonwealth University.

Tumor cell lines
The neu overexpressing mouse mammary carcinoma (MMC) cell line was established from a spontaneous mammary tumor harvested from FVBN202 mice. Tumor cells were maintained in RPMI 1640 supplemented with 10% FBS.

Ex vivo reprogramming and expansion of splenocytes
Reprogramming of tumor-sensitized immune cells was performed as previously described by our group. 5 Briefly, FVBN202 transgenic mice were inoculated in the mammary fat pad with 3 × 10 6 MMC cells.
After 48 h, cells were washed and cultured at 10 6 cells/ml in complete medium with 40 U/ml of IL-2. After 24 h, lymphocytes were again washed and cultured at 10 6 cells/ml in complete medium with 40 U/ml of IL-2. Lymphocytes were harvested 24 h later on the sixth day and were then either used in AIT or analyzed ex vivo. Reprogramming of splenocytes consistently yielded 5-fold expansion with greater than 40% memory T cells and 35% CD25+ NKT cells. 27

Adoptive cellular therapy
Twenty-four hours prior to AIT, FVBN202 mice were injected i.p. with CYP (100 mg/kg) to induce lymphopenia. Approximately 18 h later FVBN202 mice were challenged i.v. with MMC cells (1 × 10 5 ). Mice then received adoptive transfer of reprogrammed splenocytes i.v. at a dose of 70 × 10 6 /mouse later the same day (AIT), or remained untreated (Control). The study end-point and euthanasia occurred when the animals were considered moribund upon losing 10-20% of their initial body weight due to disease progression.

Characterization of splenocytes and tumor-infiltrating leukocytes
Spleens and metastases of tumor-bearing FVBN202 mice were harvested when the animals became moribund, and were then homogenized into a single cell suspension as described previously 28  Propidium Iodide (PI) was purchased from Sigma. (All reagents were used at the manufacturer's recommended concentration. Cellular staining was performed as previously described by our group. 24 Multicolor data acquisition was performed using a LSRFortessa X-20 (BD Biosciences) and a ImageStreamX Mark II Imaging Flow Cytometer (Millipore Sigma, Billaerica, MA). Data was analyzed using FCS Express v4.07 and v5.0 (De Novo Software; Glendale, CA).

Sorting of myeloid cells by FACS
Splenocytes were stained for surface expression of CD11b and Gr1 as described above. Isolated cells were gated on the myeloid cell population based on their inherent light scattering properties 29 thereby excluding cells of lymphoid origin. Gr1 −/low CD11b −/low myeloid cells from the Control and AIT groups were then sorted into independent populations using a FACSAria (BD Biosciences) as previously described. 30 Purity of sorted cells was consistently greater than 90%.

IFN-ELISA
Splenocytes from the Control and AIT groups were independently cultured in serum-free RPMI 1640 in order to enrich for nonadherent cells. 31  Supernatants were then collected and stored at −80 • C until assayed.
IFN-was detected in the supernatant using a Mouse IFN-ELISA kit (BD Biosciences), according to the manufacturer's protocol.

Cytotoxicity assay
Antitumor efficacy of T cells was determined in a cytotoxicity assay, in vitro, using flow cytometry as previously described by our group 32

Isolation of tumor-infiltrating leukocytes from lung metastases
Lungs were harvested from the Control and AIT groups after animals became moribund. Metastatic lesions were individually excised from the residual lung tissue, and were minced and digested in Trypsin-EDTA (0.25%; Life Technologies) overnight at 4 • C. The following day, the suspension was incubated at 37 • C for 30 min, followed by gentle tissue homogenization to create a cellular suspension. The cell suspension was then washed twice with RPMI supplemented with 10% FBS. Residual red blood cells were then lysed using ACK lysing buffer, followed by an additional wash with RPMI 10% FBS. 10 6 cells of the suspension were then stained for surface molecules as described above. All analysis was performed by gating on viable leukocytes (CD45+ Annexin V − ).

Statistical analysis
Outcomes are summarized by basic descriptive statistics such as mean and SEM; differences between groups are illustrated using graphical data presented as mean ± SEM. Statistical comparisons between groups were made using one-tailed and two-tailed Student's t-test per the specific hypothesis. Time to death in the in vivo survival studies was calculated from baseline to the date of death. Mice were euthanized when they had a weight loss of ≥10%. Kaplan-Meier curves and log-rank tests are used to illustrate time to death and to test the difference between each group. A P-value ≤ 0.05 was considered statistically significant.

Gr1 −/low CD11b −/low cells demonstrate characteristics of professional APCs
Antitumor immune responses are often corrupted in tumor bearing hosts due to pathological emergency myelopoiesis, which leads  Fig. 2).

Adoptive immunotherapy modulates Gr1 −/low CD11b −/low APCs
It has been reported that activated NKT cells can convert MDSCs into immune-stimulatory APCs. 22,23 We have reported that reprogrammed lymphocytes containing CD25+ NKT cells can induce maturation of human CD33 + CD11b + HLA-DR − MDSCs into stimulatory CD33 + CD11b −/low HLA-DR + APCs, in vitro. 24,27 Given the inverse correlation between Gr1 + CD11b + cells and Gr1 −/low CD11b −/low APCs in naïve mice (Fig. 1A), we sought to determine the impact of tumor burden as well as AIT, containing conventional tumor-specific T cells and  As the Ly6G + Ly6C − subset had a higher expression of costimulatory molecules than the Ly6G − Ly6C − subset in naïve mice ( Fig. 3), we sought to determine whether this trend was also present during tumor burden or following AIT. Subset analysis of Gr1 −/low CD11b −/low APCs showed the emergence of a Ly6G + Ly6C − cell population in tumor-bearing mice that received AIT when compared with the control group (Fig. 5A, 35%

Gr1 −/low CD11b −/low APCs exhibit immune stimulatory function
In order to determine if Gr1 −/low CD11b −/low APCs possess immune stimulatory function during tumor burden and/or following AIT, splenic lymphocytes from the AIT and control group were independently cultured with MMC tumor cells in the presence or absence of sorted autologous Gr1 −/low CD11b −/low cells. As shown in Fig. 6A,  = 0.0001). Importantly, the IFNproducing immune response to MMC was significantly boosted by autologous Gr1 −/low CD11b −/low APCs (Fig. 6A, p = 0.015). On the other hand, lymphocytes derived from the control group did not demonstrate significant IFN-release in the presence of MMC; the addition of autologous Gr1 −/low CD11b −/low APCs did not enhance this response (Fig. 6B). In order to determine if Gr1 −/low CD11b −/low myeloid cells from the control group retained their immune stimulatory function, they were co-cultured with tumor-reactive T cells from the AIT group in the presence or absence of MMC. We hypothesized that T cell specific killing of MMC cells from the AIT group could facilitate cross presentation of tumor Ags by Gr1 −/low CD11b −/low APCs, resulting in the enhancement of the immune response. As shown in Fig. 6C, the presence of Gr1 −/low CD11b −/low APCs boosted tumor-reactive IFN-production by splenic T cells derived from the AIT group (P = 0.0002). This was associated with the induction of apoptosis in MMC by reprogrammed T cells that were used for AIT compared with those of the control group (Fig. 6D, P = 0.0004).
To assess the possibility of Gr1 −/low CD11b −/low myeloid cells to potentially uptake and cross-present Ag to T cells, we first pulsed these cells with ovalbumin conjugated to a fluorophore. As shown in Although it appears these cells have a reduced efficiency to uptake this Ag compared to DCs, these data suggest that Gr1 −/low CD11b −/low F I G U R E 5 Continued (C) Gr1 −/low CD11b −/low APCs were sorted from the spleens of the AIT and control groups, and cultured for 24 h in the presence or absence of LPS (+LPS and -LPS). Data represent mean ± SEM of triplicate experiments myeloid cells may potentially function to cross-present processed Ag to T cells.
Thus, to specifically determine if Gr1 −/low CD11b −/low myeloid cells could cross-present Ag to provoke a T cell response, we sorted splenic Gr1 −/low CD11b −/low myeloid cells and pulsed them with recombinant Neu ECD protein, followed by a culturing period with tumor-sensitized T cells. In fact, as can be seen in Fig. 6F, Gr1 −/low CD11b −/low myeloid cells were able to induce IFN-production from tumor-sensitized T cells only after they were pulsed with Neu ECD, suggesting these cells possess Ag-processing and presentation functionality. Bone marrowderived DCs were used a positive control for Ag cross presentation; irradiated MMC cells were used as a specificity control for assessing Neu-reactive T cell function. We then utilized a direct cytotoxicity assay to demonstrate that sorted Gr1 We thank Julie Farnsworth and Qingzhao Zhang of Virginia Commonwealth University for their assistance with cell sorting.
We also thank Dr. Jose R. Conejo-Garcia for graciously providing key reagents.