FcγRIIIa receptor interacts with androgen receptor and PIP5K1α to promote growth and metastasis of prostate cancer

Low‐affinity immunoglobulin gamma Fc region receptor III‐A (FcγRIIIa) is a cell surface protein that belongs to a family of Fc receptors that facilitate the protective function of the immune system against pathogens. However, the role of FcγRIIIa in prostate cancer (PCa) progression remained unknown. In this study, we found that FcγRIIIa expression was present in PCa cells and its level was significantly higher in metastatic lesions than in primary tumors from the PCa cohort (P = 0.006). PCa patients with an elevated level of FcγRIIIa expression had poorer biochemical recurrence (BCR)‐free survival compared with those with lower FcγRIIIa expression, suggesting that FcγRIIIa is of clinical importance in PCa. We demonstrated that overexpression of FcγRIIIa increased the proliferative ability of PCa cell line C4‐2 cells, which was accompanied by the upregulation of androgen receptor (AR) and phosphatidylinositol‐4‐phosphate 5‐kinase alpha (PIP5Kα), which are the key players in controlling PCa progression. Conversely, targeted inhibition of FcγRIIIa via siRNA‐mediated knockdown or using its inhibitory antibody suppressed growth of xenograft PC‐3 and PC‐3M prostate tumors and reduced distant metastasis in xenograft mouse models. We further showed that elevated expression of AR enhanced FcγRIIIa expression, whereas inhibition of AR activity using enzalutamide led to a significant downregulation of FcγRIIIa protein expression. Similarly, inhibition of PIP5K1α decreased FcγRIIIa expression in PCa cells. FcγRIIIa physically interacted with PIP5K1α and AR via formation of protein–protein complexes, suggesting that FcγRIIIa is functionally associated with AR and PIP5K1α in PCa cells. Our study identified FcγRIIIa as an important factor in promoting PCa growth and invasion. Further, the elevated activation of FcγRIII and AR and PIP5K1α pathways may cooperatively promote PCa growth and invasion. Thus, FcγRIIIa may serve as a potential new target for improved treatment of metastatic and castration‐resistant PCa.


Low-affinity immunoglobulin gamma Fc region receptor III-A (FccRIIIa)
is a cell surface protein that belongs to a family of Fc receptors that facilitate the protective function of the immune system against pathogens. However, the role of FccRIIIa in prostate cancer (PCa) progression remained unknown. In this study, we found that FccRIIIa expression was present in PCa cells and its level was significantly higher in metastatic lesions than in primary tumors from the PCa cohort (P = 0.006). PCa patients with an elevated level of FccRIIIa expression had poorer biochemical recurrence (BCR)-free survival compared with those with lower FccRIIIa expression, suggesting that FccRIIIa is of clinical importance in PCa. We demonstrated that overexpression of FccRIIIa increased the proliferative ability of PCa cell line C4-2 cells, which was accompanied by the upregulation of androgen receptor (AR) and phosphatidylinositol-4-phosphate 5-kinase alpha (PIP5Ka), which are the key players in controlling PCa progression. Conversely, targeted inhibition of FccRIIIa via siRNA-mediated knockdown or using its inhibitory antibody suppressed growth of xenograft PC-3 and PC-3M prostate tumors and reduced distant metastasis in xenograft mouse models. We further showed that elevated expression of AR enhanced FccRIIIa expression, whereas inhibition of AR activity using enzalutamide led to a significant downregulation of FccRIIIa protein expression. Similarly, inhibition of PIP5K1a decreased FccRIIIa expression

Introduction
Fc receptors are a family of cell surface receptors that are commonly expressed by the cells in the immune system against pathogens [1,2]. FccRIIIa (CD16a) is an activating Fc receptor and is mainly expressed by mast cells, macrophages, neutrophils, and NK cells [3,4]. The activation of FccRIIIa is in part dependent on its binding to the Fc portion of IgG1 antibody as shown in cocrystal structure of FccRIIIa in complex with IgG [1].
The increased FccRIIIa expression in monocytes/macrophages is associated with the increased cytokine production that may trigger the inflammatory and autoimmune disease conditions [5][6][7][8][9][10][11]. The spontaneously expansion of the circulating monocytes expressing FccRIIIa was detected in patients with metastatic gastrointestinal carcinoma [12]. The expanded monocytes expressing FccRIIIa have also been found in the peripheral circulations of patients with breast cancer [13]. The FccRIIIa level was increased in blood serum from mice bearing xenograft tumors compared with mice without tumors, suggesting that FccRIIIa expression may be associated with tumorigenesis [14]. Interestingly, FccRIII expression was detected in prostate cancer (PCa) cell lines including LNCaP and PC-3 cells using flow cytometry analysis [15].
Recent advances in tumor immunology suggest that various types of tumors are able to escape from immunological attack by hijacking the key factors from immune cells. It is believed that tumor cells by expressing Fc receptor can block B-cell normal function, thereby allowing tumor cells to escape T-cellmediated cytotoxicity [16]. However, the expression and localization of FccRIII in cancer cells and its role in tumorigenesis remains obscure.
Mice with deletion of FccRIIIa allele had better survivals from the severe sepsis compared with the wildtype controls [17]. Moreover, mice lacking FccRIIIa allele had reduced phagocytosis activity and decreased pro-inflammatory cytokine production in their blood cells in response to E. coli bacteria infection [18]. The monoclonal antibody mAb 3G8 against FccRIIIa has shown promising effect on autoimmune diseases, as mAb 3G8 is able to induce clearance of the inflammatory immune complexes by selectively blocking FccRIIIa binding to IgG [19,20]. A bispecificmonoclonal antibody (2B1) was produced to against both c-erbR-2 onco-protein and FccRIII protein, and 2B1 treatment suppressed growth of SK-OV-3 human ovarian tumors in xenograft mice without obvious toxicity [21]. These findings provide evidence, supporting a role of FccRIIIa not only in autoimmune diseases but also in tumorigenesis.
In this study, we reported our novel findings on the identification of FccRIII expression in primary cancer and metastatic tissues from PCa patient cohorts and in various subtypes of PCa cell lines. We demonstrated that FccRIII is functionally associated with PIP5K1a/ AKT and AR pathways and promoted tumor growth and invasion. We further shown that targeted inhibition of FccRIIIa via siRNA-mediated knockdown or using inhibitory antibody suppressed growth of primary prostate tumors and reduced distant metastasis in xenograft mouse models. Our findings provide important information on new targets and options for combinational-targeted therapies for treatment of metastatic PCa.

Materials and methods
2.1. Tissue specimens, tissue microarrays, cDNA microarrays, and CGH arrays Tissue microarrays (TMAs) containing benign prostatic hyperplasia (BPH) (n = 48) vs. matched PCa tissues (n = 48) from a patient cohort (n = 48 patients), and primary PCa (n = 14) and metastatic PCa lesions in different organs including lymph node, liver, lung and bone/bone marrow (n = 43) from 14 PCa patients were constructed at Sk ane University Hospital, Malm€ o. The mRNA expression and copy number alteration data were extracted from the two cohorts of MSKCC datasets (n = 181 primary; n = 37 metastatic prostate cancer samples) [35][36][37][38], the SU2C/PCF metastatic patient cohort (n = 429 cases) [33], and the TCGA cohort (n = 333 cases) [34] from the Prostate Oncogenenome Project dataset in cBioPortal databases [35][36][37][38]. This study was approved by the Ethics Committee, Lund University and Ume a University. The General Data Protection Regulation (GDPR) was applied, and written informed consent was obtained when required. The Helsinki Declaration of Human Rights was strictly observed.

Immunohistochemical analysis
Immunohistochemistry on tumor tissue microarrays was performed as previously described [39]. The staining procedure was performed using a semiautomatic staining machine (Ventana ES; Ventana Inc.). The staining intensity was scored as 0 (negative), 1 (weakly positive or positive), 2 (moderate positive), or 3 (strongly or very strongly positive). The specimens were evaluated and scored by three different scientists, one of them a specialist in pathology. To evaluate the metastatic invasion of PCa cells in the bone/bone marrow of mice, femurs were fixed in 4% paraformaldehyde for 24 h before decalcification in formic acid and embedded in paraffin sections.

Mouse models of PCa and PCa distant metastasis
The animal studies were approved by the Swedish Regional Ethical Animal Welfare Committee. Three sets of mouse experiments were performed. Athymic NMRI-Foxn1 nu nude mice were purchased from the Charles River Laboratories (Sulzfeld, Germany). The male mice (n = 4-6, per experiment group) aged 4-6 weeks and weight 25-27 gram each were used in each experimental setting. (a) 1 9 10 6 PC-3M cells transfected with scrambled-siRNA control or FCGR3A-siRNA were implanted subcutaneously into left flank of each nude mouse. Growth and invasion of tumors in xenograft mice was assessed (n = 4 mice/group). (b) For in vivo antibody treatment of subcutaneous PCa tumors, 1 9 10 6 PC-3 cells were injected subcutaneously into left flank of each mouse (n = 4 per group). (c). For antibody treatment of metastatic PCa, 40 tumor spheroids derived from ALDH high PC-3M cells were injected subcutaneously into left flank of each mouse (n = 4 or 6 per group), and two independent experiments were performed. To assess the frequency of PCa cells metastasized to the distant organs, xenograft mice were injected with HLA-ABC antibody conjugated with 680-DyLight NHS-ester (Life Technologies, Stockholm, Sweden) 6 h before imaging. The in vivo imaging device (IVIS imaging system, PerkinElmer, USA) was used. For treatment of mice with purified anti-human FccRIIIa (M3G8) monoclonal antibody (M3G8 leaf antibody, BioLegend, USA) at 5 mgÁkg À1 or isotype IgG1 via intraperitoneal injection were used twice a week. For quantitative analysis of tumor size, Living IMAGE Ò software was used to measure metastatic areas and signal intensities. Bone/bone marrow samples were collected postmortem and used for immunohistochemical and immunoblot analyses. The animal experiments were under the license numbers: A-12-16, A-13-16, and A3-19 approved by the Swedish Regional Ethical Committee. The animal welfare and guidelines were strictly followed. All experimental mice were kept in the ventilation cages under highly sterile conditions with 12-h light/dark cycles. The maximum number of mice was limited to four per cage. The diet and water for feeding the animals were highly sterile.

Immunoblot and immunoprecipitation analyses
Immunoblot and immunoprecipitation analyses were performed as described previously [40]. Briefly, protein from different subcellular fractions (cytoplasmic and nuclear) was isolated by using NE-PER TM Nuclear and Cytoplasmic Extraction Reagents according to manufacturer's protocol (Thermo Fisher Scientific). Densitometric quantification of immunoblots was performed by using the IMAGEJ Image Analysis Software (NIH, Baltimore, USA). For immunoblot analysis, antibodies against FCGR3A (CD16a) were purchased from Biosite. For immunoprecipitation analysis, antibody against PIP5K1a was used to pull down the immune complexes, and antibody to IgG (Thermo Fisher Scientific, Sweden) was used as a negative control.

ALDEFLUOR assay
Aldehyde dehydrogenases (ALDH) expression in PC-3M cells is used for define the cancer stem cell enriched population ALDH high vs. noncancer stem cell population ALDH low populations. The ALDH high and ALDH low were sorted from PC-3M cells on FACS Aria (BD Biosciences) as previously described [40]. The ALDEFLUOR Kit (StemCell Technologies, Vancouver, British Columbia, Canada) was used according to manufacturers' protocol.

Immunofluorescence analysis
For staining with primary and secondary antibodies, alternatively, cell suspensions were fixed on slides in methanol in À20°C for 10 min. PCa cells were grown on glass coverslips in phenol red-free RPMI-1640 medium containing 10% FBS for 24 h and fixed with 4% paraformaldehyde in PBS. The slides were stained with primary antibodies. Primary antibodies including anti HLA-ABC conjugated with FITC and anti-FCGR3A (CD16a), was purchased from Biosite (Bioss MA, USA), and PIP5K1a (Protein Technologies, UK) was used. Secondary antibodies include anti-rabbit conjugated to Alexa Fluor 488 (Invitrogen, Stockholm, Sweden), antimouse conjugated to Alexa Fluor 546 (Invitrogen, Stockholm, Sweden), and anti-rabbit conjugated to Rhodamine (Chemicon International Inc, Temecula, CA). Cells were counterstained with DAPI (4',6diamidino-2-phenylindole, dihydrochloride) (SERVA Electrophoresis GmbH, Heidelberg, Germany). The cells were examined under an Olympus AX70 microscope using NIS Elements F 2.20 software or a Zeiss Apoptome microscope (Zeiss, Germany) and the ZEN 2.3 lite software (Zeiss, Germany).

Proliferation assay
Proliferation of the cells was determined using MTS proliferation assays (Promega Biotech Sweden, Stockholm, Sweden) according to manufacturer's protocol. Cells at 5 9 10 3 cells per well were cultured in a 96well plates for 48 h. MTS incorporation into the DNA was determined by measuring the absorbance at both 490 nm on an ELISA plate reader Infinite Ò M200 multimode microplate reader (Tecan Sunrise TM , Tecan Group, M€ annedorf, Switzerland).

Luciferase assays
PC-3 cells were transiently transfected with different vectors along with the reporter vector containing luciferase gene (Luc) or full-length cyclin A1 promoter in Luc reporter vector (cyclin A1-Luc) as indicated. The Firefly Luciferase and Renilla Luciferase activity were determined by using an Infinite Ò M200 multimode microplate reader (Tecan Sunrise TM ), equipped with dual injector. For AR receptor activity assay in LNCaP cells, AR Cignal Reporter Assay Kit (Qiagen Inc.) was used according to the manufacturer's protocol. Briefly, LNCAP cells were transiently transfected with different vectors along with the reporter vector ARE-luciferase (ARE-Luc) vector as indicated. The Firefly and Renilla Luciferase with Dual Luciferase Assay Kit (Promega, Biotech Sweden, Stockholm, Sweden) according to standard protocol in the Tecan Infinite M200 (Tecan Trading AG, Mannedorf, Switzerland) plate reader equipped with dual injector.

Migration assay
Cell migration assays were performed using Transparent PET Membrane chambers (Corning, Germany). A total of 0.5-2 9 10 5 cells in RPMI-1640 phenol redfree and serum-free medium were seeded in the upper chamber, and RPMI media supplemented with 50% serum as a chemoattractant was loaded in the lower chamber to allow the migration to proceed. The migrated cells were stained after 18 h or 24 h, and the proportion of migrated cells was calculated as described [41].

Statistical analysis
Tukey-test, t-test, Kruskal-Wallis/ANOVA, and Spearman rank correlation tests were performed. Student t-test was used for statistical analyses of the experimental data. The standard deviation (SD) is an indication of variability of all samples. The precision of the sample mean is indicated by standard error. Confidence levels are expressed using 95% confidence interval (CI). All statistical testes were two-sided, and P-values less than 0.05 were considered to be statistically significant. Data presented are representative of at least two or three independent experiments. Distribution of overall survival (OS) or disease-free survival/ biochemical recurrence-free survival (BRFS) was estimated by the method of Kaplan-Meier, with 95% confidence intervals. Differences between survival curves were calculated using the log-rank test. Statistical software, Social Sciences software (SPSS, version 21, Chicago), and GRAPHPAD Software were used.

Clinical relevance of FccRIIIa expression in patients with PCa metastasis and its correlation with PIP5K1a
To investigate the role of FccRIIIa in tumorigenesis and progression of PCa, we firstly examined FccRIIIa expression in primary tumors and metastatic lesions from PCa patients. FccRIIIa protein expression was assessed by using immunohistochemical analysis on the tissue microarrays (TMAs) consisting of benign prostate hyperplasia (BPH), adjacent primary PCa and metastatic PCa lesions in lymph nodes, bone marrows, and lungs from patients with primary and metastatic PCa. FccRIIIa expression was found in the epithelium of BPH tissues (Fig. 1A). Further, FccRIIIa was expressed by the primary cancer tissues and its expression was nearly significantly higher in the primary PCa tissues as compared with that in the BPH (P = 0.051; Fig. 1A). We found that FccRIIIa was highly expressed in the metastatic lesions in the bone marrows, lymph nodes, and lungs from patients who suffered metastatic PCa (Fig. 1A). Statistical analysis revealed that FccRIIIa protein expression was significantly higher in metastatic lesions compared with that of primary tumors (P = 0.006) (Fig. 1B). We next examined FccRIIIa mRNA expression in the MSKCC patient cohort from the Prostate Oncogenenome Project dataset in cBioPortal databases [35]. Similar to its protein expression in PCa tissues, FccRIIIa mRNA expression was found in primary and metastatic PCa from the MSKCC cohort which contained over 95% of cancer cells in the tumor tissues. Further, the expression of FccRIIIa mRNA was significantly higher in metastatic lesions (n = 19) than that of primary tumors (n = 131) (P < 0.01; Fig. 1C). We have previously reported that PIP5K1a is a key player in PCa progression and metastasis. We therefore wanted to assess whether elevated level of FccRIIIa expression might be associated with abnormal PIP5K1a expression in PCa primary and metastatic cancer tissues. Spearman's rank correlation test was performed, which revealed that there was a statistically significant correlation between FccRIIIa and PIP5K1a protein expression in primary and metastatic cancer tissues from the PCa patient cohorts (R 2 = 0.480, P < 0.001; Table 1). In addition, FccRIIIa expression correlated with cyclin A1 in cancer tissues from the same patient cohorts (P = 0.001; Table 1). To further assess the clinical importance of FccRIIIa expression, we examined the association between FccRIIIa expression and biochemical recurrence (BCR)-free survival of the patients using Kaplan-Meier survival analysis. We found that PCa patients with higher FccRIIIa mRNA expression in their tumors (n = 12) had worse biochemical recurrence (BCR)-free survival compared to those with lower FccRIIIa expression (n = 127) (P = 0.026) (Fig. 1D). Next, we wanted to assess whether alterations in FCGR3A gene encodes for FccRIIIa might be a frequent event in metastatic PCa. To this end, we examined genomic alterations in FCGR3A along with AR, PIP5K1A, and PTEN using a large SU2C/PCF metastatic PCa cohort (n = 429) [33]. We found that FCGR3A gene amplifications and mRNA upregulation account for 9% of the metastatic PCa cases. Interestingly, 61% cases of this cohort had AR gene amplifications and mRNA upregulation, and 19% had PIP5K1A gene amplification and mRNA upregulation. Conversely, PTEN gene mutation, deletion, or mRNA downregulation accounted for 37% of this metastatic PCa cohort (Fig. 1E). These data indicate that FCGR3A amplification and mRNA upregulation are likely associated with PCa metastatic status. To further assess the clinical importance of FccRIIIa mRNA expression in PCa progression, we examined FccRIIIa mRNA expression in primary tumors from the TCGA PCa cohort, which were divided into four subgroups based on the Gleason grade scores (n = 333 , as assessed by immunohistochemical analysis of TMAs using antibody against FccRIIIa. The scale bars: 300 µm and 50 µm are indicated and apply to all images in the panel. (B) Box-plot quantitative comparison of FccRIIIa protein expression between primary PCa, n = 52 and metastatic PCa, n = 19 (mean scores in primary and metastatic lesions were 1.73 and 2.17, difference = 0.74; 95% CI = 1.81-2.27, P = 0.006). **P < 0.01 is indicated. The error bars indicate SD. The student t-test was used to determine the significance. (C) Box-plot quantitative comparison of FCGR3A mRNA expression between primary (n = 131) and metastatic lesions (n = 19) (P < 0.01), **P < 0.01 is indicated. The error bars indicate SD. The student t-test was used to determine the significance. (D). Kaplan-Meier survival analysis based on biochemical recurrence-free (BCR-free) survival shows the difference between patients with low and high expression of FCGR3A. Differences in BCR-free survival between two groups and P-values were calculated using the log-rank test. P = 0.026 is indicated. (E). Alterations in genes and mRNA expression of FCGR3A, AR and PIP5K1A and PTEN in metastatic PCa cohort SU2C/PCF (n = 429) are shown in onco-prints. Different types of alterations in genes and their respective mRNA expression are indicated. (F). Dot plot graph shows the FCGR3A mRNA expression in four subgroups of PCa that were categorized using Gleason scores (TCGA PCa cohort, n = 333). Subgroups with Gleason score 3 + 3 (n = 65), Gleason score 3 + 4 (n = 102), Gleason score 4 + 3 (n = 78), and Gleason score >=8 (n = 88). The expression of FCGR3A mRNA in comparison between group of Gleason score >=8 and Gleason score 3 + 4, P = 0.003; the expression of FCGR3A in comparison between group of Gleason score 4 + 3 and Gleason score 3 + 3, P < 0.001. The ANOVA test was used to determine the significance. (G). Alterations in genes and mRNA expression of FCGR3A and AR in the PCa cohort mentioned in (F) are shown in onco-prints. Table 1. Correlations for protein expression in patient samples.

FccRIIIa and PIP5K1a
FccRIIIa and Cyclin A1 cases) [34]. FccRIIIa mRNA expression was significantly higher in the primary tumors with Gleason scores higher than 8 compared with those with lower Gleason grades (3 + 3 or 3 + 4) (P < 0.01) (Fig. 1F). These data showed that FccRIIIa mRNA expression was elevated in the advanced PCa, suggesting the clinical importance of FccRIIIa expression in PCa. Gene alterations in the FCGR3A loci were found in 6% of this PCa cohort, which was similar to what was observed for AR (Fig. 1G). In the MSKCC/DFCI patient cohort consisting primary PCa (n = 1013 PCa cases) from the Prostate Oncogenenome Project dataset in cBioPortal databases [42], FCGR3A gene alterations were found in 3% of PCa cases, which was similar to PIP5K1A gene alterations accounted for 5% in this PCa cohort (Fig. S1). No statistically significant correlations between AR and FccRIIIa mRNA expression were found in SU2C/PCF metastatic PCa cohort and TCGA cohort (Figs. S2,S3).

FccRIIIa expression in PCa cells is involved in tumor growth
To validate the presence of FccRIIIa expression in PCa cells, we used castration-resistant PCa cell line, PC-3 cells to examine the FccRIIIa mRNA expression.
Since cancer cells utilize inflammatory myeloid cells to stimulate growth signaling in cancer cell, we cocultured PC-3 cells with monocytes U-937 cells to examine the expression of FccRIIIa in PC-3 cells after being cocultured with U-937 cells. FccRIIIa mRNA expression was observed in PC-3 cells and U-937 cells cultured alone as well as in PC-3 cells from the coculture with U-937 cells as determined by the semiquantitative RT-PCR analysis using the primers specific for FCGR3A gene (Fig. S4A). The sequence of the PCR product of FCGR3A from PC-3 cells exhibited 100% match with the consensus sequence of FCGR3A from NCBI database (Fig. S4B). Next, we examined FccRIIIa protein expression in PC-3 cells along with various types of PCa cell lines by using immunoblot analysis. Interestingly, FccRIIIa protein exhibited high level in the castration-resistant cell lines including C4-2, VCaP, PC-3 and PC-3M cells, while its expression level appeared to be relatively low in the nonmalignant PNT1A cells (Fig. S5). Since FccRIIIa expression is significantly elevated in metastatic PCa tissues compared with the primary PCa tumors, we wanted to elucidate the role of FccRIIIa in PCa progression. To this end, we transfected C4-2 cells with a full-length FccRIIIa expressing plasmid or a control vector to induce FccRIIIa overexpression. Overexpression of FccRIIIa was confirmed by immunoblot analysis ( Fig. 2A). To examine the effect of FccRIIIa overexpression on tumor growth, we subjected C4-2 cells that expressed FccRIIIa or control vector to the 3-D tumor spheroid assays. We found that C4-2 cells overexpressing FccRIIIa gave rise to higher numbers of tumor spheroids than that of controls (P = 0.0057; Fig. 2B). This suggests that elevated The student ttest was used to determine the significance. (H) The formation of protein complexes among FccRIIIa, PIP5K1a, and AR was assessed by using immunoprecipitation (IP) assay. C4-2 cells were subjected to immunoprecipitation (IP) assay in which antibody against PIP5K1a was used to pull down the immuno-complexes, and antibody to IgG was used as a negative control. Antibodies against FccRIIIa and AR were used for immunoblot analysis (IB). The equal amount of total lysates was used as input control for immunoblot analysis of the immunoprecipitated protein lysates. Data are representative of at least two independent experiments (n = 2). (I) Immunofluorescence analysis was performed to assess the subcellular localization and its colocalization with PIP5K1a. Representative images of the subcellular localizations of level of FccRIIIa in PCa cells led to increased ability of tumorigenesis of PCa cells in vitro. Since steroid hormone DHT promotes growth of PCa by inducing activation of AR pathways that are associated with PCa growth and invasion, we therefore examined the relationship between DHT/AR pathways and FccRIIIa expression in PCa cells. We employed C4-2 cell line to DHT treatment and examined the effect of DHT on FccRIIIa expression. We found that DHT treatment at 5 nM resulted in increased FccRIIIa expression, which was equivalent to its effect on induction of AR expression in C4-2 cells (Fig. 2C). Thus, FccRIIIa expression is responsive to androgen stimulation in PCa cells. As mentioned above, we found a significant correlation between FccRIIIa and PIP5K1a in primary tumor and metastatic lesions from PCa patients. Given that PIP5K1a acts on upstream of AR pathways, we next investigated the relationship between FccRIIIa and AR, as well as the interaction between FcRIIIa and PIP5K1. Induced overexpression of FccRIIIa led to a slight increase in AR and PIP5K1a in C4-2 cells compared with the controls. However, the statistically significance was not achieved (Fig. 2D,E). Next, we examined the effect of FccRIIIa depletion on expression of AR and PIP5K1a in C4-2 cells. FccRIIIa was silenced via siRNA-mediated knockdown in C4-2 cells. The significant downregulation of FccRIIIa in C4-2 cells compared with that of siRNA control was confirmed using immunoblot analysis (P = 0.024; Fig. 2F,G). We found that silence of FccRIIIa led to the downregulation of PIP5K1a and AR expression as compared with that of siRNA controls in C4-2 cells (for PIP5K1a, P = 0.005; for AR, P = 0.001; Fig. 2F,G). Thus, FccRIIIa depletion has significant effect on the expression of PIP5K1a and AR, suggesting that FccRIIIa is required by PCa cells to regulate PIP5K1a and AR expression.
To further determine the relationship among FccRIIIa, AR, and PIP5K1a in PCa cells, we performed immunoprecipitation assays to examine whether FccRIIIa may form protein-protein complexes with AR and PIP5K1a in C4-2 cells. We found that both FccRIIIa and AR were present in the immuno-complexes that are associated with PIP5K1a (Fig. 2H). This suggests that FccRIIIa is able to form protein complexes with PIP5K1a and AR as well. Thus, FccRIIIa is functionally linked to AR and PIP5K1a via protein-protein interaction. Immunofluorescent analysis was performed to examine the subcellular localization of FccRIIIa in C4-2 cells. We found that FccRIIIa expression was highly enriched in the membrane/cytoplasmic compartments, and appeared to be colocalized with PIP5K1a in the membrane compartment of C4-2 cells (Fig. 2I).

The functional association between AR and FccRIIIa in PCa cells
To gain deeper understanding of the mechanisms underlying the association between FccRIIIa and AR in PCa cells, we examined the effect of AR overexpression on FccRIIIa expression at protein and mRNA levels. AR overexpression was induced in LNCaP cells by transfecting the cells with a vector carrying full-length AR or a control vector. We found that elevated expression of AR resulted in a significant increase in FccRIIIa protein expression in LNCaP cells as compared with that of control (P = 0.03; Fig. 3A). However, AR overexpression had no significant effect on FccRIIIa mRNA expression in LNCaP cells, as determined by using semiquantitative RT-PCR analysis (Fig. S6). Since enzalutamide is an inhibitor for AR and it targets ligand-binding domain of AR, resulting in inhibition of AR activity. We therefore examined whether inhibition of AR using enzalutamide might have a direct effect on FccRIIIa protein and mRNA expression. To this end, we treated LNCaP cells with enzalutamide for 6 h, 12 h, and 24 h, respectively. We then examined the effect of AR inhibition by enzalutamide on the expression of AR, FccRIIIa, and PSA, a known downstream target of AR. As expected, enzalutamide treatment resulted in significant decrease in AR expression after 6 h of treatment and throughout 12 to 24 h (for AR, enzalutamide treatment vs. control treatment for 6 h, P = 0.026; 12 h, P = 0.003; and 24 h, P = 0.021; Fig. 3B,C,D). Interestingly, enzalutamide treatment led to a significant downregulation of FccRIIIa protein expression readily after 12 h and throughout 24 h of post-treatment (enzalutamide treatment vs. control treatment for 6 h, P = 0.325, for 12 h, P = 0.03 and for 24 h, P = 0.034; Fig. 3B-D). Enzalutamide did not appear to have significant effect on FccRIIIa mRNA expression in LNCaP cells as measured by quantitative RT-PCR (Fig. S7). The effect of AR inhibition on FccRIIIa readily appeared at early time point of 12 h, and LNCaP is a slow proliferating PCa cell line with a doubling time at approximately 60 h; these findings suggest that AR and FccRIIIa may be directly under each other's control.

The role of FccRIIIa in AR-independent castration-resistant PC-3 cells
To further elucidate the role of FccRIIIa and the underlying cellular mechanisms in PCa growth and progression, we employed PC-3 cells that lack functional AR, which represent an ideal model to assess the relationship between FccRIIIa and PIP5K1a. To do this, FccRIIIa overexpression was induced in PC-3 cells by transfecting the cells with a vector carrying full-length FccRIIIa or a control vector, and FccRIIIa overexpression was verified by immunoblot analysis (Fig. 4A). We found that elevated expression of FccRIIIa resulted in a significant increase in PIP5K1a expression as compared with the control (P = 0.02; Fig. 4A). The effect of FccRIIIa overexpression on growth of PC-3 tumor was determined using tumor-spheroid assays. Similar to what was observed in C4-2 cells, FccRIIIa overexpression resulted in increased ability of PC-3 cells to form tumor spheroids as compared with the controls (P = 0.001; Fig. 4B). This data suggest that FccRIIIa overexpression is able to promote tumorigenesis in PC-3 cell model.
To elucidate the role of FccRIIIa in PCa progression, we assessed the effect of FccRIIIa overexpression on migratory ability of PC-3 cells. PC-3 cells that overexpressed FccRIIIa displayed significantly higher migratory ability compared with the control (P < 0.001; Fig. 4C).
To further assess the functional importance of FccRIIIa in tumor progression, we silenced FccRIIIa using siRNA-mediated knockdown. PC-3 cells that were transfected with siRNA to FccRIIIa or control siRNA were subjected to the tumor-spheroid formation assays. In contrast to the effect of FccRIIIa overexpression on tumor-spheroid formation, silence of FccRIIIa led to remarkable decrease in ability of PC-3 cells to form tumor spheroids relative to that of controls (P = 0.012; Fig. 4D,E). This suggests that elevated level of FccRIIIa is functional important for PCa cells to gain tumorigenic ability. Similar to what was observed in C4-2 cells, siRNA-mediated knockdown of FccRIIIa led to a significant downregulation of PIP5K1a expression (P = 0.011; Fig. 4F). Next, we examined the effect of inhibition of PIP5K1a on FccRIIIa in PC-3 cells. We applied a selective inhibitor of PIP5K1a, ISA-2011B and examined ISA-2011B on FccRIIIa in PC-3 cells. Interestingly, inhibition of PIP5K1a by ISA-2011B resulted in decreased FccRIIIa expression, which was coincident with the downregulation of pAKT induced by ISA-2011B in PC-3 cells as compared with controls (Fig. 4G). Similar to what was shown in Fig. 3, these data also show that PIP5K1a and FccRIIIa mutually affect each other, which further support our observation on that PIP5K1a and FccRIIIa form protein-protein complexes, as mentioned in Fig. 2. PIP5K1a promotes prostate cancer cell survival and invasion through regulation of expression of AR in PCa cells [28,29]. To this end, we wanted to investigate whether FccRIIIa and PIP5K1a might act as coregulators of AR to enhance transcriptional activity of AR on its target genes, and we utilized a cyclin A1 full-length promoter-luciferase reporter construct as described [41] and examined the effect of FccRIIIa and PIP5K1a on AR transcriptional activity on its target gene cyclin A1 in PC-3 cells. AR alone increased remarkably cyclin A1-luciferase activity as compared with that of controls (P = 0.003; Fig. 4H). FccRIIIa had no additive effect on AR to further enhance cyclin A1 promoter activation (Fig. 4H). Interestingly, PIP5K1a and AR in combination increased remarkably cyclin A1luciferase activity as compared to that of AR alone (P = 0.008; Fig. 4H). These data suggest that FccRIIIa may serve as a coregulator of AR via PIP5K1a. To further elucidate the functional impact of FccRIIIa on the downstream target genes in PCa cells, we employed androgen-dependent LNCaP cells and carried out dual-luciferase assays by using androgen-responsive (ARE) luciferase reporter construct. FccRIIIa overexpression induced ARE reporter luciferase activity, which led to an increase in ARE promoter activity by 100% relative to controls in LNCaP cells (P = 0.013; Fig. 4I). Thus, FccRIIIa is able to mediate the transcriptional activity of the key factors that contribute to PCa progression.

Targeted inhibition of FccRIIIa in PC-3M cells reduced tumor growth in xenograft mice
We have previously reported that PC-3M cells are able to initiate metastasis to distant organs in xenograft mice [40]. We therefore employed PC-3M xenograft tumor models in mice to elucidate the role of FccRIIIa in PCa progression. To this end, we silenced FccRIIIa in PC-3M cells by using siRNA-mediated knockdown. We then implanted subcutaneously equal amount of si-FccRIIIa PC-3M cells and si-control PC-3M cells into the nude mice. The growth of PC-3M tumors in xenograft mice was measured and monitored. At the end of experiments, the mean tumor volumes in mice that received si-FccRIIIa PC-3M cells were significantly smaller than that of controls (P = 0.020; Fig. 5A). We then assessed expression of the key marker proteins including Ki-67, phosphorylated AKT, MMP9, and VEGFR2 that control proliferation and invasiveness of PCa cells. Consistent with what was observed on the tumor volumes, si-FccRIIIa PC-3M tumors displayed a significantly reduced proliferation rate relative to controls, as determined by using Ki-67 staining (P < 0.001; Fig. 5B). Similarly, we found that expression of PIP5K1a, pSer-473AKT, MMP9, and VEGFR2 was significantly downregulated in si-FccRIIIa PC-3M tumors compared with that of sicontrol tumors, which was coincident with the reduced volumes of si-FccRIIIa PC-3 M tumors (for PIP5K1a expression, P = 0.048; for pAKT, P = 0.0045; for MMP9 expression, P = 0.028; and for VEGFR2, P < 0.001; Fig. 5C-F). We have previously reported that PC-3M cells are able to initiate metastasis to distant organs in xenograft mice [40]. We therefore examined the apparent metastasis in the lymph nodes in mice that have received si-FccRIIIa PC-3M or sicontrol PC-3M cells. We found that mice bearing sicontrol PC-3M tumors had lymph node metastasis, whereas mice bearing si-FccRIIIa tumors were free of lymph node metastasis (Fig. 5G). Further, sicontrol PC-3M tumors expressed cytokeratin 19 (CK19) and vimentin, the human epithelial cell markers. In contrast, si-FccRIIIa tumors were negative to CK19 and vimentin expression. These data suggest that inhibition of FccRIIIa greatly reduced growth and metastatic potentials of primary tumors in xenograft mouse models. It is likely that FccRIIIa promotes PCa growth and invasion via its downstream PIP5K1a/AKT and VEGFR2 signaling pathways.  82 and 1.44, difference=0.37, 95% CI in si-control = 1.8-1.84, and in si-FccRIIIa = 1.39-1.5, P = 0.028. **P < 0.01 and *P < 0.05 are indicated. Tumors from the two groups (for si-control group, n = 3; for si-FccRIIIa group, n = 2) were stained with the indicated antibodies and were evaluated. The error bars indicate SEM. The student t-test was used to determine the significance. The scale bars: 2 mm and 100 µm in the images in B, C, D, E, and F are indicated. (G) Representative images of the lymph nodes containing metastatic lesions from the xenograft mice bearing si-FccRIIIa tumors as compared with that of si-control RNA (siCtrl) are shown. Tumors were immune-stained with the antibodies against CD19 and Vimentin that are markers for cancer cells. Tumor cells positive to the markers were indicated by the arrows. Tumors from the two groups (for si-control group, n = 3; for si-FccRIIIa group, n = 2) were stained with the indicated antibodies and were evaluated. The scale bars: 1 mm and 100 µm are indicated.

Inhibitory effect of anti-FccRIIIa antibody on PCa growth in PCa cell line models and in PCa xenograft mice
To further study the role and cellular mechanisms of FccRIIIa in tumor growth and invasion, we examined the antitumor effect of anti-FccRIII antibody (M3G8) in in vitro and in vivo systems. We subjected PC-3 cells to the formation of tumor spheroids. The tumor spheroids were then subjected to the treatment with M3G8 or control antibody. We observed that M3G8 treatment led to a remarkably reduced number of tumor spheroids as compared with controls (P = 0.028; Fig. 6A). Also, there was a pronounced alteration in cell-cell contacts and a reduced phalloidin staining in tumor spheroids treated with M3G8 compared with that of controls (Fig. 6B).
Next, we wanted to investigate whether blockade of FccRIIIa using purified anti-human FccRIII monoclonal antibody termed M3G8 may suppress growth of PCa tumors in xenograft mice. To this end, we established xenograft mice bearing subcutaneously implanted PC-3 tumors, which were less invasive, but grow rapidly as compared to PC-3M tumors. PC-3 tumors were allowed to grow into approximately 300 mm 3 in size and were randomized into two groups. The two groups of mice were treated with M3G8 or control antibody. The M3G8-treated group had tumors which were fourfold smaller in size relative to the control group after treatment for 21 days (mean volume of tumors in control group and M3G8 group was 1612 mm 3 and 436 mm 3 , respectively, difference = 1176 mm 3 ; 95% CI = 384-486, P < 0.01; Fig. 6C). This was consistent with the inhibitory effect of targeted inhibition of FccRIIIa on PC-3M tumors shown above. Immunohistochemical analysis of tumor tissues further revealed that M3G8-treated tumors exhibited reduced expression of PIP5K1a and VEGFR2 as compared to that of controls (for PIP5K1a, P = 0.01; for pAKT, P = 0.02; and for VEGFR2, P = 0.01; Fig. 6D,E).

Inhibitory effect of anti-FccRIIIa antibody on PCa growth and metastasis in mice
Next, we wanted to investigate whether blockade of FccRIIIa using M3G8 may reduce/inhibit distant metastasis of PCa. We have previously reported that ALDH high stem-like subpopulations isolated from PC-3M cells initiated metastatic growth in distant organs such as bone/bone marrow in xenograft mice [40]. To this end, we sorted stem-like ALDH high subpopulations of PC-3M cells using FACS-based ALDEFLUOR assay and subjected the stem-like ALDH high subpopulations to the formation of 3-dimensional tumor spheroid (Fig. 7A). The tumor spheroids were then implanted subcutaneously into the nude mice (40 tumor spheroids/mouse) to allow formation of distant metastasis (Fig. 7A). Virtually all mice that received tumor spheroids had developed distant metastasis 60 days postimplantation, as measured and quantified by using in vivo imaging assays as described previously [40] (Fig. 7B,C). Mice bearing metastatic lesions were randomized into two groups and were treated with intraperitoneal injection of M3G8 or control antibody at 5 mgÁkg À1 dose (Fig. 7B,C). At the end of the experiments, there was a significant reduction in metastatic burdens in mice treated with M3G8 compared with that of control, as quantified using in vivo imaging analysis (P = 0.039; Fig. 7B,C). M3G8 treatment did not induce weight loss or other detectable adverse events in the mice. There was a significant higher proportion of cells positive to cytokeratin 19 (CK19), a marker of human epithelial cell origin, in the bone marrows from xenograft mice treated with control antibody compared with those treated with M3G8 (P = 0.02; Fig. 7D). These data suggest that inhibition of FccRIIIa inhibits metastatic growth of PCa cells in distant organs in xenograft mice.
To test the therapeutic potentials of combination therapies of M3G8 and ISA-2011B, we assessed the effects of M3G8, ISA-2011B alone or in combination on the invasiveness of C4-2 cells. C4-2 cells that were treated with M3G8 or ISA-2011B alone or in combination were subjected to the migration assays. Similar to ISA-2011B, M3G8 treatment alone significantly reduced migratory ability of C4-2 cells (for M3G8, P = 0.016; for ISA-2011B, P = 0.006; Fig. 7E). Combination of M3G8 and ISA-2011B had greater inhibitory effect as compared to that of M3G8 alone on the migratory ability of C4-2 cells (for combination of M3G8 and ISA-2011B vs. control, P = 0.003; for combination of M3G8 and ISA-2011B vs. M3G8, P = 0.011; Fig. 7E). These data suggest that combination treatment using M3G8 and ISA-2011B may have an additive inhibitory effect on PCa cells.

Discussion
In this study, we discovered that the expression of FccRIIIa was significantly higher in metastatic lesions than that of primary cancer tissues. Moreover, high level of FccRIIIa was significantly associated with poor prognosis in PCa patients. We for the first time showed that FccRIIIa was expressed in PCa cells from primary tumor tissues and metastatic lesions and PCa  *P < 0.05 is indicated. The student t-test was used to determine the significance. The scale bar: 50 µm is indicated. (E) The effect of M3G8 and ISA-2011B alone or in combination on the migratory ability of C4-2 cells was assessed by using the migration assays. After treatment with the agents, the equal amount of the cells from different groups was subjected to the Boyden chamber migration assay for 18 h. M3G8 treatment or ISA-2011B treatment alone reduced migratory ability of C4-2 cells as compared with that of controls (for M3G8, P = 0.016; for ISA-2011B, P = 0.006). Combination of M3G8 and ISA-2011B reduced the migratory ability of the cells as compared with that of controls (P = 0.003). Data are presented as average of two independent experiments (n = 2). **P < 0.01 and *P < 0.05 are indicated. The error bars indicate SEM. The student t-test was used to determine the significance. cell lines as well. FccRIIIa expression was significantly higher in metastatic lesion compared to that of primary tumor tissues. We showed that FCGR3A gene amplifications and mRNA upregulation accounted for 9% of the metastatic PCa cases, in which 61% cases had AR gene amplifications and mRNA upregulation, 19% had PIP5K1A gene amplification and mRNA upregulation, and 37% cases had PTEN gene mutation, deletion, or mRNA downregulation. Furthermore, PCa patients with higher FccRIIIa mRNA expression in their tumors had worse biochemical recurrence (BCR)-free survival compared to those with lower FccRIIIa expression. Our data suggest that FccRIIIa expression is highly clinical relevant and may reflect its role in PCa development and progression.
In this study, we aimed to investigate whether FccRIIIa may play an important role in growth and invasion of PCa at both AR-dependent and ARindependent fashions. We found that induced overexpression of FccRIIIa in C4-2 cells promoted cancer cell growth. Conversely, inhibition of FccRIIIa via siRNA-mediated knockdown reduced growth ability of C4-2 cells. Similarly, induced overexpression of FccRIIIa led to increased expression of AR and PIP5Ka, the key factors that promote PCa growth and invasion, while inhibition of FccRIIIa led to decreased expression of AR and PIP5K1a.
One of the striking findings in this study is the identification of the underlying mechanism by which FccRIIIa and AR interact with each other in PCa cells. We found that FccRIIIa was capable of inducing AR target gene promoter activation as determined by using the ARE reporter luciferase activity assays. Furthermore, overexpression of AR led to significant increase in FccRIIIa protein expression, while inhibition of AR by using enzalutamide decreased FccRIIIa protein expression readily after 12 h of enzalutamide treatment of LNaP cells. In addition, we showed that AR and FccRIIIa interacted with each other through formation of protein-protein complexes together with PIP5K1a. Our findings suggest that the observed effect of FccRIIIa on AR may not be the consequence of FccRIIIa-induced cell proliferation in AR-expressing PCa cells, but rather due to that FccRIIIa is functionally associated with AR and PIP5K1a associated pathways via protein-protein interactions.
Interestingly, induced overexpression of FccRIIIa in androgen-independent PC-3 cells that do not express functional AR also promoted proliferation and invasion of PC-3 cells. Conversely, inhibition of FccRIIIa using siRNA-mediated knockdown led to significant decrease in growth and invasion of PC-3 cells in vitro and PC-3 tumors in xenograft mice. Although the underlying mechanisms by which FccRIIIa promotes PCa growth and invasion at AR-independent fashion remain obscure, our findings in in vitro and in vivo model systems provide strong evidence, suggesting that FccRIIIa plays an important role in AR-independent fashion. It has been reported that activation and expression of FccRIIIa in immune cells are associated with the formation of immune complexes, and increased FccRIIIa expression can lead to the subsequent activation of PI3K/AKT pathways in immune cells [1]. It is known that FccRIIIa is activated by IgG immune complexes. Thus, the interaction between FccRIIIa and IgG immune complexes is critical for FccRIIIa internalization to enable the activation of FccRIIIa downstream signaling events related to migration and survival of leukocytes [2]. It will be of great interests to investigate whether FccRIIIa may utilize the IgG immune complexes from the PCa-associated immune cells/tumor microenvironment to promote growth and progression of castration-resistant PCa.
Our results further showed that PIP5K1a was functionally associated with FccRIIIa, as inhibition of both PIP5K1a and FccRIIIa resulted in greater inhibition of invasiveness of PCa cells as compared with inhibition of FccRIIIa alone. Further, these key molecules organize and activate several signaling pathways, leading to tumor cell survival and invasion. We plan to further investigate the role of FccRIII and the underlying mechanisms in PCa progression from androgen dependence to castration-resistant state in the near future.
In this study, we applied xenograft models Targeted inhibition of FccRIIIa via siRNA-mediated knockdown or using inhibitory antibody suppressed growth of primary prostate tumors and reduced distant metastasis in xenograft mouse models. We further established novel metastatic xenograft mouse models to examine the effect of inhibition of FccRIIIa activity on PCa metastasis. Further, our findings suggest that FccRIIIa plays an important role in PCa progression and is a potential therapeutic target for the development of the new treatment strategies for advanced and metastatic PCa. Since elevated activity of FccRIIIa can be inhibited using blockade antibody M3G8, we therefore examined the effect of M3G8 on PCa tumor growth and metastasis. Our data showed that M3G8 significantly suppressed tumor growth in vitro and in xenograft mouse models. M3G8 blocks both FccRIIIa and FccRIIIb, we found that M3G8 treatment led to an inhibition of FccRIIIa and reduced expression of PIP5K1a/AKT, as determined by our immunoblot analysis by using antibody against FccRIIIa, Further, the inhibitory effect of M3G8 treatment on PCa tumor growth is comparable to the effect of FccRIIIa knockdown on PCa tumor growth.
Several previous studies have demonstrated that FccRIIIa is a signal molecule that induces rapid and transient PIP5K1a membrane recruitment on NK cells to facilitate cytotoxic killing [43]. It is likely that PCa cells utilize FccRIIIa to mimic immune cells and to evade cytotoxic cell-mediated antitumor immunity.

Conclusions
Our results showed that treatment approach by optimizing activity to blocking antibody to FccRIIIa is likely the good strategy to improve the therapeutic outcome by using antibody-mediated destruction of malignant cells. Taken together, our findings suggest that FccRIIIa may serve as a potential new target for the improvement of treatment of metastatic and castration-resistant PCa.

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article. Fig. S1. In the MSKCC/DFCI patient cohort consisting primary PCa (n = 1013 PCa cases) from the Prostate Oncogenenome Project dataset in cBioPortal databases, FCGR3A gene alterations were found in 3% of PCa cases, which was similar to PIP5K1A gene alterations accounted for 5% in this PCa cohort. Fig. S2. Dot plots graph shows the correlations between AR and FccRIIIa mRNA expression in log 2 by using the SU2C/PCF metastatic PCa cohort (n = 429). The R 2 vaule and P value are indicated. Fig. S3. Dot plots graph shows the correlations between AR and FccRIIIa mRNA expression in log 2 by using the TCGA PCa cohort (n = 333). The R 2 vaule and P value are indicated. Fig. S4. Expression of FccRIIIa in PC-3 cells and U-937 cells. Fig. S5. FccRIIIa protein expression in PC-3 cells along with various types of PCa cell lines by using immunoblot analysis. U-937 monocytes and PCa cell lines including C4-2, VCaP, PC-3 and PC-3M cells were subjected to the immunoblotting analysis. Antibodies against FccRIIIa and GAPDH were used. Fig. S6. The effect of AR overexpression on FCGR3A mRNA expression. AR overexpression was induced in LNCaP cells by transfecting the cells with a vector carrying full-length AR or a control vector. The semiquantitative RT-PCR analysis using the primers specific for FCGR3A was performed. Fig. S7. The effect of AR inhibition on FCGR3A mRNA expression. LNCaP cells were treated with enzalutamide for 6 h, 12 h and 24 h respectively. The effect of AR inhibition by enzalutamide on the expression of AR was measured by quantitative RT-PCR.