Type I Interferon Promotes Antitumor T Cell Response in CRPC by Regulating MDSC

Simple Summary Despite initial tumor regression following androgen blockade treatment, relapse of castration-resistant prostate cancer (CRPC) eventually occurs in most patients. Immunotherapy aims to activate the host immune system to fight against cancer and has achieved significant therapeutic effects in various solid tumors. The purpose of our research was to investigate the mechanisms underlying the immune response during CRPC development and to screen effective immunotherapies against CRPC. We found that interferon-α (IFNα) directly inhibited the progression of CRPC, reduced the accumulation of the immune suppressive granulocytic myeloid-derived suppressor cells (G-MDSCs) in the tumor microenvironment (TME), and impaired the inhibitory function of G-MDSCs on T cell activation. This research provides a potential strategy for the clinical treatment of CRPC. Abstract Background: Metastatic castration-resistant prostate cancer (CRPC) is the leading cause of death among prostate cancer patients. Here, our aim was to ascertain the immune regulatory mechanisms involved in CRPC development and identify potential immunotherapies against CRPC. Methods: A CRPC model was established using Myc-CaP cells in immune-competent FVB mice following castration. The immune cell profile of the tumor microenvironment (TME) was analyzed during CRPC development. Different immunotherapies were screened in the CRPC tumor model, and their efficacies and underlying mechanisms were investigated in vitro and in vivo. Results: During CRPC development, the proportion of granulocytic myeloid-derived suppressor cells (G-MDSCs) in the TME increased. Among the immunotherapies tested, IFNα was more effective than anti-PD-L1, anti-CTLA-4, anti-4-1BB, IL-2, and IL-9 in reducing Myc-CaP CRPC tumor growth. IFNα reduced the number of G-MDSCs both in vitro during differentiation and in vivo in CRPC mice. Furthermore, IFNα reduced the suppressive function of G-MDSCs on T cell proliferation and activation. Conclusion: G-MDSCs are crucial to effective immunotherapy against CRPC. Treatment with IFNα presents a promising therapeutic strategy against CRPC. Besides the direct inhibition of tumor growth and the promotion of T cell priming, IFNα reduces the number and the suppressive function of G-MDSCs and restores T cell activation.


Introduction
Prostate cancer is the second most commonly diagnosed cancer among men worldwide [1]. The incidence of prostate cancer is related to factors such as age, genetics, and ethnicity [2,3]. As the tumor grows, it spreads to tissues such as bone and lymph nodes. FVB mice were purchased from Shanghai Lingchang Biotechnology Co., Ltd. If-nar1 −/− mice were kindly provided by Dr. Anita Chong from the University of Chicago. Ifnar1 −/− FVB mice were generated by crossing the Ifnar1 −/− C57BL/6 mice and the FVB mice for 6 generations.

Myc-CaP Treatment In Vitro
A total of 3 × 10 4 Myc-CaP cells were seeded in 24 well plates with DMEM complete culture medium supplemented with PBS or IFNα4 (50 or 200 ng/mL). After 48 h, the cells were digested with trypsin, and the number of cells was counted using a hemocytometer.

MDSCs Differentiation from Bone Morrow Cells
Femurs and tibias were obtained from male FVB mice, and the bone marrow cavities were flushed with PBS using an insulin syringe. ACK (Ammonium-Chloride-Potassium) lysis buffer (BD Biosciences, San Jose, CA, USA) was used to lyse the red blood cells in all samples. Bone marrow cells were cultured in Petri dishes containing RPMI-1640 complete culture medium supplemented with 20 ng/mL GM-CSF (Sinobiological Catalog No: 51048-MNAH, Beijing, China) and induced for 4 days by treating with PBS or IFNα4 (20 ng/mL or 100 ng/mL), to generate bone marrow-derived MDSCs (BM-MDSCs). On day 4, the proportion of G-MDSCs was detected by flow cytometry, and the cell number was obtained using a hemocytometer.

Isolation of G-MDSCs
Bone marrow cells were induced for 4 days. The single-cell suspension was incubated with 2.4G2 (antibodies recognizing CD16 and CD32) for 10 min to block non-specific Fcmediated binding. The BM-MDSCs were magnetically labeled with anti-Ly-6G-Biotin (1A8) (Biolegend, San Diego, CA, USA) and Streptavidin Nanobeads (Biolegend). Subsequently, they were separated on a magnetic rack to obtain G-MDSCs.

T Cell Inhibition by G-MDSCs
Mouse spleens were ground and passed through a 70 µm cell strainer. Red blood cells were lysed using ACK lysis buffer and isolated using the MojoSort™ Mouse CD3 T Cell Isolation Kit (BioLegend), according to the manufacturer's instructions. The purified T cell suspension was incubated in the dark with 5 µM carboxyfluorescein succinimidyl ester (CSFE) (Selleck, Boston, MA, USA) for 7 min and washed twice with RPMI-1640 complete culture medium to remove the unbound CFSE. Subsequently, they were cocultured with Ly6G + BM-MDSCs in the ratio of 1:1 and 3:1 in 96-well plates with RPMI-1640 complete culture medium (10% heat-inactivated FBS, 100 units/mL penicillin, 100 µg/mL streptomycin, 2 mM L-glutamine, and 55 µM β-mercaptoethanol). Plate-bound anti-CD3 (0.5 µg/mL) and anti-CD28 (1 µg/mL) antibodies were used to stimulate the T cells in the culture. Forty-eight hours after activation, the cells and supernatants were collected for flow cytometry and cytometric bead array (CBA) analyses.

Isolation of Single Cells from Tumors
Mouse tumor tissues were sliced to pieces using surgical scissors and digested in tumor dissociation buffer (RPMI-1640 with 50 µg/mL Liberase TL (Roche) and 200 µg/mL DNase I (Sigma, St Louis, MO, USA)). The tumor tissues were ground and passed through a 70 µm cell strainer. The single cells obtained were re-suspended in staining buffer (1 × PBS with 1% FBS).

Flow Cytometry Analysis
Single-cell suspensions of cells were incubated with 2.4G2 for 10 min. Blocked samples were subsequently stained with fluorescently labeled monoclonal antibodies and a fluorescent intercalator. The anti-mouse CD45-APC/Cyanine7 (30-

Cytokine Production Analysis
Production of the cytokine, IFN-γ, in co-culture supernatants from the T cells and Ly6G + BM-MDSCs was tested using the Cytometric Bead Array Kit (BD biosciences), according to the manufacturer's instructions.

RT-qPCR
Total RNA was extracted using the E.Z.N.A. ® Total RNA Kit I (Omega Bio-Tek, Norcross, GA, USA) and reverse transcribed using the GoScript Reverse Transcription system (Promega, Madison, WI, USA). Specific gene was amplified using 2 × ChamQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, Jiangsu, China) and quantified by real-time PCR, according to manufacturer's instructions. The qPCR primers are enlisted in Supplementary Materials Table S1.

Statistical Analysis
The data were analyzed using the GraphPad Prism 8 software (GraphPad Software Inc, La Jolla, CA, USA). The significance of assays was determined using the unpaired Student's two-tailed t-test. Where indicated, * p < 0.05, ** p < 0.01, *** p < 0.001 were considered as statistically significant results.

G-MDSCs Are Increased in Prostate TME Following Castration
Although androgen deprivation therapies are initially effective against prostate cancer, resistance and relapse occur eventually. During relapse, the prostate tumor cells alter their growth pattern to an androgen-independent manner. The changes in the immune cells found in the TME during the transition of prostate cancer from an androgen-dependent (AD) to a castration-resistant (CR) form remain unclear. We established a Myc-CaP prostate tumor model in immune-competent FVB mice. Following castration, we found that the Myc-CaP prostate tumor underwent remission initially and relapsed later, which resembled the clinical development of castration-resistant prostate cancer [31]. Following the castration of the tumor-bearing mice, we analyzed the immune cell composition in the TME at different time points in both the remission and relapse periods. We found that the infiltration of CD8 + T cells, CD4 + T cells, and NK cells significantly decreased during the remission and relapse of prostate cancer ( Figure 1A-C). These results suggest the inability or weak ability of these cells to induce antitumor immunity after castration. Interestingly, a subtype of MDSCs, G-MDSCs, were significantly enriched in the TME in the relapse periods after castration ( Figure 1E), but M-MDSCs decreased in the TME ( Figure 1D), suggesting that G-MDSCs may be one of the primary reasons for the suppression of antitumor immunity and promotion of CRPC development.

IFNα Was Effective in Controlling CRPC
T cells are critical for conferring antitumor immunity. Immune checkpoint blockade antibodies targeting T cells have shown promising antitumor effects in both preclinical models as well as patients with cancer [32,33]. Since T cell infiltration was decreased in the TME after castration, we examined whether immune checkpoint antibodies could re-activate T cells to reduce the tumor burden during CRPC. For this, we combined anti-PD-L1 or anti-CTLA-4 treatment with castration. However, these two immune checkpoint antibodies showed no antitumor effects during CRPC (Figure 2A-C). Insufficient co-stimulation signal and T cell proliferation cytokines are possible mechanisms for weak T cell activation in the TME [34]. Subsequently, we combined castration with anti-4-1BB co-stimulation agonist antibody and cytokine IL-2 or IL-9. Anti-4-1BB antibodies have shown strong antitumor activity in various tumor models, including immune checkpoint blockade antibody-resistant tumor models [35][36][37][38]. The cytokine IL-2 is critical for T cell proliferation and has shown antitumor activity in both preclinical models and in patients with cancer [39]. IL-9 not only induces innate and adaptive immune responses but also directly promotes tumor apoptosis [40]. However, anti-4-1BB antibody, IL-2, and IL-9 failed to reduce tumor burden in the Myc-CaP CRPC tumor model ( Figure 2D-F). Recent studies have shown that type I interferons are critical for the generation of antitumor T cell immunity [41]. Interestingly, IFNα4 treatment showed potent antitumor activity when combined with castration ( Figure 2G). Since all five types (anti-PD-L1, anti-CTLA-4, anti-4-1BB, IL-9, and IL-2) of T cell-targeting treatments were not effective against CRPC, it suggests that IFNα treatment may generate antitumor immunity through non-T cells, such as antigen-presenting cells and immune suppressive cells, which are critical for generating antitumor immunity. Cancers 2021, 13, 6 of 18

IFNα Was Effective in Controlling CRPC
T cells are critical for conferring antitumor immunity. Immune checkpoint blockade antibodies targeting T cells have shown promising antitumor effects in both preclinical models as well as patients with cancer [32,33]. Since T cell infiltration was decreased in the TME after castration, we examined whether immune checkpoint antibodies could reactivate T cells to reduce the tumor burden during CRPC. For this, we combined anti-PD-L1 or anti-CTLA-4 treatment with castration. However, these two immune checkpoint an- munity [41]. Interestingly, IFNα4 treatment showed potent antitumor activity when combined with castration ( Figure 2G). Since all five types (anti-PD-L1, anti-CTLA-4, anti-4-1BB, IL-9, and IL-2) of T cell-targeting treatments were not effective against CRPC, it suggests that IFNα treatment may generate antitumor immunity through non-T cells, such as antigen-presenting cells and immune suppressive cells, which are critical for generating antitumor immunity.  Fourteen days after castration, the tumor relapsed. The mice were administered an intratumoral injection of immune checkpoint inhibitors or cytokines (20 µg/mouse) on days 14, 17, and 21. The tumor volumes were measured using a vernier caliper twice a week and calculated using the formula: (length × width × height)/2. (B-G) Fourteen days after castration, the mice were treated with control (hIgG), anti-mouse-PD-L1 (B), anti-mouse-CTLA-4 (C), anti-mouse-4-1BB (D), mouse IL-2 (E), mouse IL-9 (F), or mouse IFNα4 (G). Statistical significance was determined using the unpaired t-test and is represented by ** p < 0.01. Representative results from one of one or two replicates are shown (B-G) (mean ± SEM), n = 4-5 per group.

IFNα4 Reduced Immunosuppression in the TME
Since our data indicated that the T cell targeting treatment through immune checkpoint blockade, co-stimulation enhancement, and T cell growth stimulation was not sufficient to activate efficient antitumor T cell immune response, we hypothesized that non-T immune cells may be critical for IFNα-mediated tumor suppression. To verify this, we analyzed the immune cell profile in the TME of castrated Myc-CaP bearing mice on day 14 post IFNα4 treatment. We found that the infiltration of CD45 + leukocytes in the tumor tissue was significantly increased on day 14 (8.9% vs. 23%) post IFNα4 treatment ( Figure 3A). Among the CD45 + leukocytes, IFNα4 treatment had little effect on the infiltration of CD19 + B cells ( Figure 3G). However, it significantly increased the infiltration of CD4 + T cells (1.9% vs. 7.4%), CD8 + T cells (0.6% vs. 9.1%), and NK cells (1% vs. 2.5%) on day 14 post treatment ( Figure 3B-D). These findings are consistent with the reduced tumor burden following IFNα4 treatment. In addition, we also observed a significant decrease in the proportion of immune suppressive G-MDSCs on day 14 after IFNα4 treatment (6.8% vs. 1.5%) ( Figure 3F), but not M-MDSCs in the TME (2.2% vs. 6.6%) ( Figure 3E). These data suggest that IFNα4 acts on multiple immune cell types to simultaneously promote cells with tumor-killing effect and reduce tumor-promoting immune suppressive cells.

Cytotoxic T Cells Are Critical for IFNα-Mediated Therapeutic Effect on CRPC
IFNAR is widely expressed on almost all cell types, including tumor and non-tumor cells, which are potential targets of IFNα treatment. Besides their direct inhibition of tumor growth, recent studies have highlighted the importance of IFNα in activating various immune cells, including T cells and NK cells [42]. In this study, first, we tested the direct effect of IFNα on Myc-CaP cells. Consistent with previous findings [14], IFNα4 significantly reduced Myc-CaP cell growth to about 70% in vitro ( Figure 4A). Since CD4 + T cells, CD8 + T cells, and NK cells are crucial antitumor components and potential targets of IFNα, we tested if these cells were required for the IFNα-mediated tumor suppression of CRPC. We administered CD8 + T cell-, CD4 + T cell-, or NK cell-depleting Ab during the IFNα4 treatment of castrated Myc-CaP bearing FVB mice and measured tumor growth. CD8 + T

Cytotoxic T Cells Are Critical for IFNα-Mediated Therapeutic Effect on CRPC
IFNAR is widely expressed on almost all cell types, including tumor and non-tumor cells, which are potential targets of IFNα treatment. Besides their direct inhibition of tumor growth, recent studies have highlighted the importance of IFNα in activating various immune cells, including T cells and NK cells [42]. In this study, first, we tested the direct effect of IFNα on Myc-CaP cells. Consistent with previous findings [14], IFNα4 significantly reduced Myc-CaP cell growth to about 70% in vitro ( Figure 4A). Since CD4 + T cells, CD8 + T cells, and NK cells are crucial antitumor components and potential targets of IFNα, we tested if these cells were required for the IFNα-mediated tumor suppression of CRPC. We administered CD8 + T cell-, CD4 + T cell-, or NK cell-depleting Ab during the IFNα4 treatment of castrated Myc-CaP bearing FVB mice and measured tumor growth. CD8 + T cell depletion, and not CD4 + T cell or NK cell depletion, abolished the therapeutic effect of IFNα4 ( Figure 4B-D). These data suggest that the antitumor activity of IFNα4 is mediated primarily through the activation of CD8 + T cell response.
Cancers 2021, 13, 10 of 18 cell depletion, and not CD4 + T cell or NK cell depletion, abolished the therapeutic effect of IFNα4 ( Figure 4B-D). These data suggest that the antitumor activity of IFNα4 is mediated primarily through the activation of CD8 + T cell response.

IFNα4 Inhibited the Differentiation or Proliferation of G-MDSCs
Our results showed that IFNα4 reduced the accumulation of G-MDSCs and enhanced the antitumor T cell response. However, the mechanisms underlying the IFNαmediated reduction in the accumulation of G-MDSCs, and the contribution of this reduction to enhanced T cell response remained unclear. To verify this, first, we tested if IFNα4 affected the proliferation of G-MDSCs. We performed in vitro differentiation of G-MDSCs from bone marrow precursor cells and found that the yield of G-MDSCs was significantly reduced (5.85 × 10 4 vs. 3.0 × 10 4 ) when IFNα4 was present, suggesting that IFNα4 directly affects G-MDSCs differentiation and proliferation ( Figure 5A). To further confirm this, we obtained G-MDSCs from Ifnar1 −/− bone marrow cells, which lack the ability to transduce the interferon signaling pathway. We found that IFNα4 had no effect on Ifnar1 −/− G-

IFNα4 Inhibited the Differentiation or Proliferation of G-MDSCs
Our results showed that IFNα4 reduced the accumulation of G-MDSCs and enhanced the antitumor T cell response. However, the mechanisms underlying the IFNα-mediated reduction in the accumulation of G-MDSCs, and the contribution of this reduction to enhanced T cell response remained unclear. To verify this, first, we tested if IFNα4 affected the proliferation of G-MDSCs. We performed in vitro differentiation of G-MDSCs from bone marrow precursor cells and found that the yield of G-MDSCs was significantly reduced (5.85 × 10 4 vs. 3.0 × 10 4 ) when IFNα4 was present, suggesting that IFNα4 directly affects G-MDSCs differentiation and proliferation ( Figure 5A). To further confirm this, we obtained G-MDSCs from Ifnar1 −/− bone marrow cells, which lack the ability to transduce the interferon signaling pathway. We found that IFNα4 had no effect on Ifnar1 −/− G-MDSCs differentiation and proliferation ( Figure 5B). To further confirm that IFNα4-mediated downstream signal activation was critical for G-MDSCs differentiation and proliferation, we investigated whether IFNα4 possessed the same function in vivo. Three days post the second treatment of CRPC-bearing mice with IFNα4. We observed a decrease in the number of G-MDSCs in the bone marrow ( Figure 5C). Collectively, these results demonstrate that IFNα4 negatively regulates G-MDSCs differentiation and proliferation both in vitro and in vivo.
Cancers 2021, 13, 11 of 18 MDSCs differentiation and proliferation ( Figure 5B). To further confirm that IFNα4-mediated downstream signal activation was critical for G-MDSCs differentiation and proliferation, we investigated whether IFNα4 possessed the same function in vivo. Three days post the second treatment of CRPC-bearing mice with IFNα4. We observed a decrease in the number of G-MDSCs in the bone marrow ( Figure 5C). Collectively, these results demonstrate that IFNα4 negatively regulates G-MDSCs differentiation and proliferation both in vitro and in vivo.  were differentiated with 20 ng/mL GM-CSF, combined with PBS or IFNα4 (20 ng/mL or 100 ng/mL) for 4 days, in 96-well plates. The G-MDSCs (CD45 + CD11b + Ly6G + ) cell number was compared. Statistical significance was determined using unpaired t-test and is represented by ** p < 0.01, *** p < 0.001. Representative results from one of two replicates are shown (A,B) (mean ± SEM), with triplicate wells per group. (C) Similar to Figure 2, Myc-CaP CRPC-bearing mice were treated with PBS or IFNα4 on days 14 and 17 after castration. On day 20, bone marrow cells were harvested, and the G-MDSCs (CD45 + CD11b + Ly6G + ) cell number was compared. Each point represents one mouse. Statistical significance was determined using unpaired t-test and is represented by ** p < 0.01. Representative results from one of two replicates are shown (C) (mean ± SEM), n = 5 per group.

IFNα4 Inhibited the Immune Suppressive Function of G-MDSCs
To test whether IFNα4 affected the immunosuppressive function of G-MDSCs, we performed magnetic bead sorting on in vitro differentiated G-MDSCs. An equal number of purified G-MDSCs were co-cultured with purified CFSE-labeled CD3 + T cells to evaluate the suppressive function of G-MDSCs on T cell activation and proliferation. We found that the inhibitory effect of G-MDSCs on both CD4 + T cell and CD8 + T cell proliferation was significantly reduced post exposure to IFNα4 (Figure 6A-D). Furthermore, to test whether IFNα4-treated G-MDSCs influenced the effector function of T cells, we analyzed the secretion of the T cell effector molecule, IFN-γ, in the co-culture supernatant; this revealed the activation status of T cells. Consistent with the reduced inhibitory effect of G-MDSCs on T cell proliferation, IFNα4-treated G-MDSCs showed a weaker inhibitory effect on IFN-γ secretion from T cells following their activation ( Figure 6E). In addition, we also tested the immune suppressive function of IFNα4-treated Ifnar1 −/− G-MDSCs. We found that the IFNα4-mediated reduction in the immune suppressive function of G-MDSCs was abolished in Ifnar1 −/− G-MDSCs, indicating that the activation of downstream signaling pathways is crucial for the function of IFNα ( Figure 6F-J). To elucidate the detailed regulatory mechanism of the effect of IFNα4 on G-MDSCs, we analyzed the mRNA expression profile of molecules associated with T cell activation by G-MDSCs cells. We observed a significant increase in the expression levels of co-stimulatory molecules such as ICOSL, TNFSF14, and CD40L, and T cell growth factors such as IL-7 and IL-15 ( Figure 6K). These data suggest that IFNα4 not only inhibits the proliferation of G-MDSCs but also affects its immune suppressive function. The G-MDSCs obtained were purified using anti-Ly6G magnetic beads and subjected to RNA isolation. The expression profile of T cell activation-related genes was analyzed by RT-qPCR. GAPDH was used as a housekeeping gene to normalize gene expression. Statistical significance was determined using unpaired t-test and is represented by * p < 0.05, ** p < 0.01, *** p < 0.001. Representative results from one of two replicates are shown (K) (mean ± SEM).

Discussion
Prostate cancer usually depicts slow growth. Detection and treatment before symptoms appear often results in limited improvement in the health and survival of patients. Clinically, surgical resection, hormonal therapy, and radiation therapy are used to treat prostate cancer. Androgen deprivation drugs, such as abiraterone or enzalutamide, cause anemia, lower bone density, and CRPC. Once patients develop CRPC, they become resistant to most therapeutic drugs, thereby limiting the treatment option to a few drugs [43]. In recent decades, immunotherapy has been increasingly used for the clinical treatment of prostate cancers. It prolongs the survival of patients and has fewer side effects. The tumor vaccine, sipuleucel-T, which targets prostatic acid phosphatase (PAP), has received U.S. FDA approval to be used in the treatment of metastatic CRPC (asymptomatic/minimally symptomatic) [44]. Moreover, several PSMA-directed CAR-T cells have undergone clinical trials for the treatment of metastatic CRPC, and the drugs have been identified as safe and feasible when used at the appropriate dosage [45,46]. Previous studies have shown that immune checkpoint inhibitors such as anti-PD-1 and anti-CTLA-4 slightly slow down the growth of CRPC. However, when they are combined with MDSCs depleting anti-Gr1 antibody or BEZ235 (dual PI3K and mTOR inhibitor), they significantly inhibit the growth of CRPC [47]. In addition, preclinical studies have shown that anti-IL-23 antibody inhibits the growth of CRPC and increases the efficacy of enzalutamide in the treatment of CRPC [48].
In this study, we focused on understanding the changes in the immune cell profile during CRPC development and designing potential immunotherapies for the treatment of CRPC. Various mechanisms trigger the development of CRPC: the increased sensitivity of the androgen receptor (AR) pathway or AR mutations lead to androgen-independent AR activation [49]. Besides these intrinsic changes in the tumor, our results showed that the changes in the immune cell profile of the TME might also contribute to CRPC development. We found a reduction in the infiltration of CD8 + T cells and NK cells and an increase in the proportion of immunosuppressive cells such as G-MDSCs, which resulted in increased immunosuppressive ability. To treat CRPC, we evaluated several immunotherapies, including immune checkpoint inhibitors such as anti-PD-L1 Ab and anti-CTLA-4 Ab, agonistic antibodies of co-stimulatory molecules, such as anti-4-1BB Ab, and cytokines, such as IL-2, IL-9, and IFNα4. These treatments have shown potent antitumor activity in some tumor models [38,[50][51][52][53][54]. However, we found that only IFNα4 reduced the tumor burden in the Myc-CaP CRPC tumor model. IFNα directly acts on tumor cells, blocks their cell cycle progression, and induces cell apoptosis [55,56]. It also indirectly activates immune cells, promotes their effector functions, or blocks their suppression in order to kill tumor cells. Previous studies have shown that IFNα increases the production and survival of CD8 + effector T cells [57], promotes NK cell activation and effector factor release [58], promotes B cell maturation and immunoglobulin secretion [59] and increases the antigen-presenting ability of DCs [60]. In this study, our data indicated that IFNα4 may act on other non-T cells to inhibit the growth of CRPC. We found that the proportion of G-MDSCs increased during the development of CRPC and decreased significantly following IFNα4 treatment. Furthermore, IFNα reduced the proliferation of G-MDSCs both in vivo and in vitro. It also decreased the G-MDSC-mediated inhibition of T cells. It is reported that MDSC-derived IL-23 contributed to the development of castration-resistant prostate cancer [23]. It will be interesting to investigate whether IFNα could affect IL-23 production from MDSC. Immune checkpoint antibodies have shown weak to moderate efficacy in prostate cancer [61]. It is worth testing whether IFNα could be combined with immune checkpoint antibodies to enhance the antitumor efficacy.
Although our findings are promising, our study has certain limitations. First, the TME contains many types of immune cells, and IFNα, which has a wide range of effects, may affect other immune cells as well, which we did not look at. In addition to G-MDSCs, it will be interesting to elucidate the role of IFNα on other immune cells in the prostate cancer TME, such as NK cells, macrophages, and B cells. Second, the systemic delivery of IFNα has several side effects in clinical settings. Therefore, it is critical to investigate if the targeted delivery of IFNα against a specific prostate cancer antigen or an IFNα pro-drug is more effective in reducing the side effects on non-tumor tissues.
In summary, G-MDSCs are correlated with the development of CRPC. IFNα effectively inhibits the growth of CRPC, reduces the number of G-MDSCs in tumor-bearing mice, and decreases the inhibitory effect of G-MDSCs on T cells in vitro. Our work revealed that G-MDSCs may be a potential therapeutic target, thereby presenting a new strategy for the treatment of CRPC.

Conclusions
G-MDSCs infiltration is crucial for designing immunotherapies against CRPC. IFNα promotes antitumor T cell response against CRPC by regulating G-MDSCs, thereby presenting a potential approach for the treatment of CRPC in clinical settings.
Author Contributions: X.Y. designed the overall project. L.F., G.X., J.C., M.L., H.Z., F.L., X.Q., X.Z., Z.L., P.H. and X.Y. performed the experiments. L.F. and X.Y. analyzed the results and wrote the manuscript. All authors have read and agreed to the published version of the manuscript.

Informed Consent Statement: Not applicable.
Data Availability Statement: Data sharing is not applicable to this article.