Anti-S100A4 Antibody Therapy Is Efficient in Treating Aggressive Prostate Cancer and Reversing Immunosuppression: Serum and Biopsy S100A4 as a Clinical Predictor

Recent advances in immunotherapies have revolutionized the management of cancer in patients; however, the impact has been found less pronounced in prostate cancer (CaP; ref. 1). Nevertheless, the modest outcome in vaccine therapy and negative outcome of checkpoint blockade therapy has propelled investigators in investigating the causes of failure of immunotherapy in prostate cancer. It is now widely accepted that the success of immunotherapies for prostate cancer disease would depend on the genetic status of tumors such as mutation burden, for example, BRCA mutation (2). The emerging preclinical data suggest that targeting immune system components other than conventional PDL1/PDL or CTL4 pathways could be an alternative option for successful immunotherapies for prostate cancer disease in men. The prostate tumor in general is recognized by (i) the absence of natural killer (NK)-immune cells, which are needed to kill the neoplastic cells; (ii) the presence of immunosuppressive cells [myeloid-derived suppressor cells (MDSC), Tregs]; and (iii) extracellular proteins, which sustain the immunosuppressive microenvironment (3). The new approach for making immunotherapy successful in prostate cancer disease would be the conversion of an immune-cold tumor into an immune-HOT tissue.

S100A4 is a calcium-binding protein reported to associated with rheumatoid arthritis, fibrosis, and inflammatory disease conditions (4). Mesenchymal and epithelial–mesenchymal transition (EMT) cells have an inherent tropism for sites of inflammation, which are frequently present in sites of cancer including prostatic lesions (5). Notably, S100A4 is highly expressed in EMT cells and immune-system cells such as monocytes, macrophages, and polymorphonuclear granulocytes (6). We previously have comprehensively reviewed the role of S100A4 in human diseases (4). Tumor-associated macrophages (TAM) and tumor-associated fibroblasts (TAF) are reported to enhance metastases and level of cytokines, which inhibit an adaptive tumor-specific immune response (7). It is noteworthy that both protumorigenic TAMs and TAFs are reported to express high levels of S100A4 protein (8). Weatherly and colleagues reported an association between reduced T-cell levels to S100A4 deficiency (9).

We previously reported that S100A4 expression increases in prostatic tissues with the increasing grade of prostate cancer disease in TRAMP mice, an autochthonous GEM model of prostate cancer (10). Gupta and colleagues showed a positive association between the S100A4 protein and increasing Gleason grade in patients with prostate cancer (11). We showed that S100A4 oncoprotein plays a critical role in the prostate tumorigenesis and metastasis in TRAMP mouse model (12). We provided evidence that S100A4 exists in both intracellular as well as secretory form in prostate cancer disease and showed the growth-promoting role of extracellular-S100A4 in prostate cancer disease (12). Recent studies have shown an association of S100A4 to the poor survival of patients (13). However, the relevance of extracellular-S100A4 in prostate cancer disease has remained an understudied area.

We hypothesized that extracellular-S100A4 protein regulates metastasis and generation of aggressive phenotype by sustaining a chronic inflammatory and immunosuppressive microenvironment in the prostatic tissue. We speculate that extracellular-S100A4 protein could be developed as a fluid-biopsy biomarker and therapeutic target for advanced prostate cancer in humans. This entails a detailed investigation of secretory-S100A4 levels in patients and therapies targeting extracellular-S100A4. In this study, we investigated the (i) relevance of S100A4 as a predictive biomarker of outcome in prostate cancer disease using RNA sequencing (RNA-seq)–based Decipher-genomic test, multicenter clinical data; and (ii) therapeutic efficacy of S100A4-targeted mab6B12 antibody using relevant drug-testing techniques and models.

Human and murine cell models representing either primary or metastatic prostate cancer disease were selected for the study. TRAMP-C2 (mouse primary prostate cancer; RRID:CVCL_3615), Myc-CaP (mouse primary prostate cancer; RRID:CVCL_J703), PC-3 (human bone-derived metastatic prostate cancer; ATCC#CRL-7934, RRID:CVCL_0035), DU145 (human brain-derived metastatic prostate cancer; (ATCC#HTB-81, RRID:CVCL_0105), 22Rν1 (CRPC cell model), and mesenchymal stem cells (MSCs) were used in this study. The PC-3, DU145 TRAMP-C2 and mouse MSC cell models were procured from ATCC, whereas human MSCs (hMSC) were obtained from PromoCell. MYC-CaP was procured from Dr. Charles Sawyer's laboratory (Memorial Sloan Kettering Cancer Center, New York, NY). DU145, TRAMP-C2, and Myc-CaP were grown in DMEM, whereas PC-3 cells were grown in RPMI + 10% FBS. MSCs were grown in MesenCult MSC Basal Media and MesenCult Osteogenic Stimulatory Supplement. All the human cells were tested for contamination and genetic accuracy using STR analysis (Genetica Cell Line Testing).

We collected human umbilical cords from consenting deidentified patients through IRB-approved tissue procurement facility (BioNet) at the University of Minnesota (Minneapolis, MN). Umbilical cords were rinsed twice with Hank buffered salt solution, and one end of the cord was sealed and the umbilical veins were filled with 10 mL of collagenase type II (1 U/mL, STEMCELL Technologies). The digestion was done at 37°C for 15 minutes in the incubator. Detached cells were then released by flushing the veins with HBSS. The mixture was centrifuged at 1,500 rpm for 10 minutes and the cell pellets were resuspended in medium Endothelial Cell Growth Medium MV2 (PromoCell). The medium was replaced every 48 hours. Cells were subcultured after treatment with 0.01% trypsin/EDTA.

We have previously established the expression plasmid of S100A4 (pcDNA-S100A4) plasmid (12). The anti-S100A4 neutralizing antibody (mab6B12) was provided by J. Klingelhöfer and N. Ambartsumian under a user agreement between the corresponding author and Cancer Research Technology (London, UK).

The archived and deidentified patient sera were procured from the Department of Hematology and Oncology, Bio-Net (IRB-approved bio-specimen facility) of the University of Minnesota and commercial serum bank (BioServe). In addition, we used the archived sera to measure serum-S100A4 level, which we had used for our previous study to measure circulatory BMI1 in patients with prostate cancer (14).

This study profiled whole transcriptome (RNA-seq) from primary prostate tumor specimens from patients with prostate cancer. The cohort consisted of patients who underwent radical prostatectomy (RP) at the Masonic Cancer Center, University of Minnesota. Decipher genomic classifier (GC) testing was offered to patients with adverse pathologic features on the final pathology such as positive surgical margin, presence of extraprostatic extension, seminal vesicle invasion, or PSA persistence. After exclusion for tissue unavailability and quality control, the study consisted of 228 patients. The tumor material was shared with Decipher Biosciences. The RNA extraction was all done in a Clinical Laboratory Improvement Amendments–certified laboratory facility (Decipher Biosciences). The purified total RNA was labeled and hybridized to a Human Exon 1.0 ST GeneChips (Thermo Fisher Scientific). Quality control was performed using Affymetrix Power Tools, and normalization was performed using the Single Channel Array Normalization algorithm (15). A whole-transcriptome RNA-seq assay, which measures the expression of over 46,000 genes and noncoding RNA was performed as described previously (16). The final normalized dataset was provided back to the University of Minnesota and the bioinformatics was performed by Dr. Jinhua Wang (Institute for Health Informatics). We tested the hypothesis that S100A4 association with tumor aggression and metastasis can be highly reflective of clinical outcomes in patients with prostate cancer. To examine the relevance of biopsy-S100A4 gene expression as a predictive biomarker for therapy outcome and prostate cancer–specific mortality, we grouped patients based on the assignment with the DECIPHER-TEST's ars_1 and Penny-2011 classification algorithms (17, 18). Both prediction models use RNA transcriptome profiling data as input, extract signature genes, and build clinical outcomes using SVM algorithms. Ars_1 attempt to classify patient based on response to androgen deprivation therapy (ADT). Penny-2011 algorithm classifies patients based on prostate-specific mortality and Gleason score as the primary metric (18).

To clone human S100A4 gene in pET28a, the gene was amplified by PCR by using pcDNA-S100A4 as a template. The forward primer (CGACATATGATGGCGTGCCCTCTGGAGAAG) contained the Ndel restriction site, whereas the reverse primer (CGACTCGAGTCATTTCTTCCTGGGCTGCTT) contained the Xhol restriction site. Purified PCR product as well as expression vector pET-28a vector were digested with XhoI and NdeI. Recombinant plasmids pET28a-S100A4 were transferred into E. coli BL21 (DE3) and induced to express the S100A4 protein by 1.0 mmol/L IPTG to isopropyl-β-d-thiogalactopyranoside (IPTG) at 30°C for 6 hours. Following this, the cell lysate was sonicated and centrifuged. Protein samples from the supernatant, precipitation, and cell lysate were analyzed by SDS-PAGE and stained with Coomassie brilliant blue to confirm the pET28a-S100A4 expression. The supernatant sample was purified by Ni-NTA Column Chromatography. The recombinant S100A4 protein was concentrated and dialyzed overnight against 50 mmol/L Hepes (pH 7.5), 150 mmol/L NaCl, 10 mmol/L CaCl₂ and followed by centrifugation (3,000 × g for 3 hours at 4°C).

Physical Interaction between S100A4 protein and anti-S100A4 neutralizing (mab6B12) was performed by employing surface plasmon resonance (SPR) using a BIAcore 2000 instrument as per vendor's protocol (BIAcore, GE Life Sciences) at the Hormel Institute, University of Minnesota. Briefly, 1,800 resonance units (RU) of a recombinant human S100A4 protein were immobilized covalently on a CM4 sensor chip using an amino coupling procedure according to the manufacturer's instructions.

Native S100A4 dimer protein (200 μmol/L solution) was loaded into the calorimetric syringe and titrated with 30 μmol/L of mab6B12 antibody in the sample cell. An initial injection of 2 μL solution was followed by 28 injections (10 μL/injection) with enough time period between injections (to allow heat recovery back to baseline). The heat of dilution was measured in controlled experiments and subtracted from experimental heats before fitting. The data were fitted with a single-site binding model using an appropriate software (Origin 7, MicroCal software) provided with the instrument.

Cells were allowed to grow up to 80% confluence in complete media. At 80% confluent level, media were discarded, and cells were washed with PBS twice. After washing, cells were added with serum-free media. Cells were cultured in serum-free media for 24 hours. After 24 hours, media were collected and analyzed for S100A4 protein by using Immuno-Slot-blot assay as per the manufacturers' protocol (Whatman). Briefly, Slot-blot apparatus was assembled using Whatman filter paper and a prewetted nitrocellulose membrane and connected to a vacuum pump. Slots were filled with samples (media/serum) and unused slots were filled with PBS. The membranes were processed for S100A4 detection using a standard immunoblot/Western blot assay protocol.

This was performed by using ELISA technique. The S100A4 ELISA Kit (Circulex cat#CY-8059) and RANTES ELISA Kits (mouse specific cat# RAB0077-1kt, Sigma; human-specific RANTES, Abcam, cat# ab174446) were used as per the vendor's protocols.

Cell invasion ability was assessed using a modified Boyden chamber, 8.0-μm pore polycarbonate filter inserts (Costar-Corning). Human umbilical vein endothelial cells (HUVEC) (1.0 × 10⁵) were seeded and growth overnight at 37°C. Next, TRAMP-C2, DU145, and PC-3 cells (1.0 × 10⁴) were seeded into the upper chamber either with/or without the recombinant-S100A4 protein or mab6B12 antibody. The bottom chamber was filled with growth medium. After incubation for 48 hours, invaded cells were fixed in 3.7% PFA, stained with 0.1% crystal violet, and photographed under a microscope.

Mouse or human mesenchymal stem cells (5 × 10³) were seeded into 6-well plate in a complete MesenCult expansion medium with MesenPure for proliferation. When cells reached appropriate confluence, expansion culture medium was replaced with complete MesenCult Osteogenic medium. The conditioned medium from TRAMP-C2 and PC-3 cell types were mixed with complete MesenCult Osteogenic Medium in a 1:1 ratio. The mixed osteogenic medium and conditioned medium was then supplemented with 10 mmol/L β-glycerophosphate and 50 mg/mL ascorbic acid. The culture medium was changed at every 48 hours. At day 21, cells were stained for calcium deposition by using alizarin red S staining.

These studies were conducted in accordance with the guidelines of University of Minnesota Institutional Animal Care and Use Committee protocol (IACUC protocol#1710-35262A). For the rigor and reproducibility, we followed the NIH guidelines for considering the sex as a biological variable and selected only male mice for tumor studies because prostate cancer is a male-specific disease. One million cells of TRAMP-C2 cells mixed with Matrigel (1:1) were injected subcutaneously in the right flank male B6/J (C57BL/6J; RRID:IMSR_JAX:000664) mice. When the tumors where visible to naked eye and of approximately 100 mm³ volume, the mice were randomly distributed in a blinded manner in two experimental groups. On the basis of our experience, the tumor intake for TRAMPC2 cells is recorded 100% in male C57BL/6J mice. Therefore, for the statistical and power analysis consideration, a 10 mice/group provides a power of >80% at 0.05 significance. The 1st group served as control and received IgG control antibody (7.5 mg/kg; intraperitoneal administration) at a frequency of 3 times/week. The 2nd group received anti-S100A4 neutralizing antibody (mab6B12) therapy (7.5 mg/kg; 3 times/week). Mice received treatments for a total of 6 weeks. All procedures conducted were in accordance with IACUC guidelines.

The mouse cytokine antibody array was used as per the vendor's protocol (cat# ab197465; Abcam). Serum was incubated in cytokine-antibody array plate and the plate images were captured using GE-immunoblot imager system. The images for individual cytokine was quantified using an ImageJ software (RRID:SCR_003070).

Activated CD4⁺ cells were isolated by CD4⁺ Isolation Kit mouse (Miltenyi Biotec) and were cocultured with TRAMP-C2 cells in the presence or absence of either recombinant S100A4 protein or mab6B12 antibody (6 μg/mL). At 48 hours posttreatment, the CD4⁺ T cells were harvested and total RNA was extracted using the RNeasy Mini Kit (Qiagen). The high-quality RNA (1 μg) was converted to cDNA using the RT² First Strand Kit. Samples were loaded onto a 96-well mouse Th1 and Th2 Profiler PCR array as per the manufacturer's instructions (cat#PAMM-034ZA, Qiagen) and read using a Quant Studio 3 thermocycler system. The thermocycling conditions were 1 cycle of 95°C for 10 minutes (hot start) followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute. Melting curves were performed each time from 55°C to 95°C with increments of 0.5°C for 5 seconds to ensure that only one amplification product was formed.

Statistical comparisons were analyzed by Student t test (two groups only) and ANOVA (greater than two groups) with Prism8.4 from GraphPad Software (RRID:SCR_002798). Data are expressed either as mean ± SD or mean ± SEM. Statistical significances were accepted at P < 0.05.

Previously, we and others reported that S100A4 protein (based on IHC) is elevated during the progression of prostate cancer in transgenic TRAMP mouse model (10, 12). The study warranted thorough investigation about the relevance of S100A4 in clinical outcome in patients with prostate cancer. In the current study, we determined the (i) nature of S100A4 gene alteration (amplification or mutation) that occurs in prostate cancer patients; and (ii) association of S100A4 gene alteration to survival, mortality, therapy outcome, and biochemical recurrence by analyzing the tumor genome data of patients from multiple clinical studies. These studies included SU2C-2019, Prostate (MICH), Metastatic Prostate (SU2C), adenocarcinoma-FHCRC 2020, TCGA-PRAD, and DECIPHER-Minnesota cohort (n = 228).

We performed the data analysis of three clinical studies (n = 655 patients) available at cBioportal platform (19). The genomic analysis of tumors (n = 655) showed 4% to 12% alteration frequency in S100A4 gene and most of the alterations constituted the amplification of gene (Fig. 1Ai). We next conducted a comparative analysis of FHCRC and SU2C clinical studies (RNA-seq data of primary adenocarcinoma and metastatic prostate cancer) and observed metastatic prostate cancer exhibiting 3- to 4-fold higher S100A4 expression than primary adenocarcinoma (Fig. 1Aii). These data suggest that S100A4 expression has a strong association to the progression of disease in patients with prostate cancer.

The Decipher-classifier assay is a genomic test that serves as a marker of outcomes in patients who have undergone primary cancer treatments (15–18). The DECIPHER-test allows risk stratification of RP-treated patients to predict probability or risk of recurrence, metastases, prostate cancer–specific mortality, need for adjuvant therapy and, in patients with disease recurrence guide physicians to take appropriate treatment decision (15–18). We asked if biopsy-S100A4 expression independently predicts the outcome of secondary treatments, for example, ADT in patients who are designated as high risk of disease recurrence (after receiving primary treatment such as RP). For this purpose, we used whole transcriptome data of patients who were classified as high-responsive, low-responsive, and average-responsive patients on the basis of DECIPHER-genetic classifier test. By employing a stringent significance cut-off value at P = 0.05, we evaluated the significance of S100A4 as the predictor of treatment response in patients using ADT-response algorithm. The probability of ADT-response in patients was classified into low-ADT and high-ADT. We observed that probability of low ADT-response was increased in patients with high-S100A4 expression, thus suggesting an association of S100A4 transcriptional level with therapy failure (Fig. 1Aiii). The data generated from the algorithm suggests a significant (P = 0.05) difference between low-ADT response and high-ADT response. We next evaluated risk of prostate cancer–specific mortality in RP-treated patients based on S100A4 expression. We found that the risk of mortality was significantly (P = 0.44) reduced in patients exhibiting low levels of S100A4 expression in prostate tissues (Fig. 1Aiv). These data suggest that patients with high prostatic S100A4 expression are likely to be nontreatable by conventional therapies (such as ADT) and have increased risk of mortality.

Next, we investigated a correlation of copy number of S100A4 with the disease recurrence and survival of patients with prostate cancer using TCGA PRAD clinical data. Employing USCS xena data analysis platform (20), we found a positive correlation (P = 0.001613) between high S100A4 copy number patients to the biochemical recurrence and poor survival probability of patients with prostate cancer (Fig. 1Av). In contrast, patients exhibiting low/medium S100A4 copy number (n = 464) exhibited no biochemical recurrence and showed better survival probability (Fig. 1Av). Prostate cancer is a major health burden with significant overtreatment because of difficulty segregating high- and low-risk disease. Discovery of biomarkers that stratify risk could have a broad public health impact but requires cohorts with comprehensive molecular and clinical follow-up. By analyzing genomic and clinical data from the TCGA data (n = 380 samples) from oncomine database platform (21), patients with high Gleason score exhibited an increased S100A4 copy number (Supplementary Fig. S1A). These data demonstrate that S100A4 copy number is associated with disease progression, poor survival, and high-grade morphologic features in patients with prostate cancer.

We previously reported that S100A4 protein is secreted by PC3 and DU145 metastatic cells (12). In addition, we showed elevated levels of S100A4 protein in stromal regions of prostate tumor in TRAMP mice, indicating the extracellular nature of this protein (12). It becomes imperative to investigate if S100A4 protein is secreted by tumor cells, which grow as aggressive primary prostate cancer types. For this purpose, we measured S100A4 expression in MYC-CaP cell and TRAMP-C2 cell (derived from prostate tissue of TRAMP mouse). The immunoblot analysis validated that TRAMP-C2 and MYC-CaP murine cells express high level of S100A4 protein (Fig. 1Bi). We next performed immune slot-blot assay to analyze the free S100A4 protein in FBS-free conditioned culture medium harvested from cell cultures as per the method previously described by us (14). TRAMP-C2 and MYC-CaP cell cultures exhibited high levels of secretory S100A4 protein in respective culture media (Fig. 1Bii). Because LNCaP cells do not express S100A4, this model served as a negative control.

We next asked whether S100A4 protein could be detected in the serum of patients with prostate cancer. For this purpose, we used the archived sera collected from human patients with prostate cancer (primary and metastatic) to measure S100A4. We analyzed the archived albumin-cleared sera of patients with prostate cancer for S100A4 protein by employing Slot-blot analysis as described previously (14). We now show that secretory S100A4 protein are detectable in the serum of patients with prostate cancer and elevated in advanced stage prostate cancer (Supplementary Fig. S1Aii).

Next, we quantified the serum-S100A4 levels in prostate cancer patient cohort of patients: normal (n  =  10), stage II (n  =  23), stage III (n  =  16), stage IV (n  =  16), and metastatic prostate cancer (22). The values were calculated on a linear scale and significance was calculated employing ordinary one-way ANOVA and t test by using PRISM 8.4 statistical software (GraphPad). The mean ± SD value of serum-S100A4 protein level was recorded in subjects as: normal 3.3 ± 1.2 ng/mL, stage II prostate cancer 7.9 ± 4.4 ng/mL, stage III prostate cancer 12.5 ± 4.5 ng/mL, stage IV prostate cancer 18.3 ± 9.7 ng/mL, and metastatic prostate cancer 11.2 ± 6.6 ng/mL (Fig. 1C). The ANOVA analysis of data shows that serum-S100A4 protein levels progressively increase with the increasing stage of prostate cancer in humans (r = 0.35, P < 0.0001, Fig. 1C). These data suggest the significance of serum-S100A4 as a noninvasive biomarker for prostate cancer stage stratification; however, this warrants an investigation in a large cohort of patients.

Recent preclinical and clinical studies have shown that soluble growth factors are amenable to neutralization antibody therapy (22). Several neutralizing antibodies such as rilotumumab (anti-HGF), bevacizumab (anti-VEGFA), and fresolimumab (TGFβ) have moved to clinical trials (22, 23). The neutralizing of biological activity of growth factors could be an important node in blocking the oncogenic signaling and preventing drug resistance in various cancers. Previously, we showed that extracellular-S100A4 has a growth promoting function in vitro (10, 12). We showed that extracellular-S100A4 confers metastatic ability to less aggressive prostate cancer cells (12). Semov and colleagues showed that extracellular-S100A4 act as a proangiogenic factor in cancer (24). On the basis of these data, we hypothesized that extracellular-S100A4 protein could be an ideal target for a neutralizing antibody, which ultimately could be used a therapy to treat aggressive prostate cancer in humans. For this purpose, we collaborated with J. Klingelhöfer and N. Ambartsumian who have developed an anti-S100A4 neutralizing antibody (mab6B12; ref. 25).

By employing relevant drug–target interaction techniques such as SPR and isothermal titration calorimetry (ITC) assays, we determined whether ma6B12 antibody interacts and neutralizes the soluble S100A4 protein. The SPR is a label-free technique and enables measurement of real-time quantification of ligand-binding affinities using relatively small amounts of protein in native environment, and has potential to be medium-throughput (26). This technique is considered a gold standard for analyzing drug–target interaction. We conducted SPR assay to determine the interaction of ma6B12 antibody with S100A4 protein. For this purpose, we generated a recombinant S100A4 protein by cloning S100A4 in pET-28a vector (Supplementary Fig. S1Bi). The purified protein was of 98% purity on SDS page (Supplementary Fig. S1Bii). S100A4 protein is reported to homodimerize in the presence of calcium ion. The homodimerization of S100A4 protein has been shown to induce its activity as well as interaction to its target proteins (27). Next, the mab6B12 was immobilized on the sensor chip and served as surface-bound ligand. The experiment was performed on a BIAcore 2000 instrument (BIAcore) as described in the Materials and Methods section with 5.0 mmol/L Ca²⁺ added to the running buffer. The sensorgram was generated with recombinant-S100A4 protein as an analyte injected at different concentrations of 0.625, 1.25, 2.5, and 5 μmol/L (Fig. 1D). The sensorgram shows a significant dose-dependent response with characteristic association and dissociation slopes, indicating a true interaction between both molecules in presence of Ca²⁺ (Fig. 1D)

ITC follows the heat change when a test compound binds to a target protein. This technique is label-free, allows precise measurement of affinity, and does not introduce artifacts. The method is direct, making interpretation facile, because there is no requirement for competing molecules (28). The ability of ITC technique to measure high-affinity interactions is worth noting. ITC technique can measure binding affinities over 5 log units from 100 μmol/L to 1 nmol/L. This technique is frequently used in drug development process. We next conducted an ITC assay to determine the binding affinities of mab6B12 antibody to soluble (recombinant) S100A4 protein. Titrations of S100A4 protein with mab6B12 antibody revealed an exothermic association with a K_d of 7.37 μmol/L as shown in Fig. 1E (inset table). The one-fitting model of midpoint ITC titration calculations suggest a binding stoichiometry of S100A4 protein and mab6B12 in a ratio of 1:1 (Fig. 1E).

We previously reported that extracellular-S100A4 protein induces migration and invasiveness of prostate cancer cells (12). We asked whether mab6B12 could neutralize the S100A4 protein–induced proinvasive and promigratory potential of prostate cancer cells. By employing a wound-healing assay, we determined the migratory potential of TRAMP-C2, DU145 and PC3 cells incubated in the presence or absence of recombinant S100A4 protein (1 μg/mL) or mab6B12 antibody (6 μg/mL) or both. Cells were monitored for longitude migration using phase-contrast microscopy and images of live cells were captured by a camera fitted to microscope. The recombinant S100A4 protein significantly induced the migration of prostate cancer cells. Notably, addition of mab6B12 to culture condition was observed to rescue cells from the effects of recombinant-S100A4 protein (Fig. 2A). Prostate cancer cells (PC3, TRAMP-C2, DU145) secrete S100A4 protein in culture medium. Therefore, we tested the efficacy of mab6B12 on the migratory potential of prostate cancer cells (without adding recombinant S100A4). We observed a significant inhibition of migration of prostate cancer cells by mab6B12 therapy (Fig. 2A). These data suggest that mab6B12 has the potential to block the proinvasive/promigratory activities of secretory-S100A4 protein.

A crucial event in the metastatic cascade is the extravasation of circulating cancer cells from blood capillaries to the surrounding tissues (29). Transendothelial migration is the extravasation of tumor cells across the endothelium barrier of blood vessels and is a critical step for metastasis to distant regions of the body. Inhibiting cancer cell extravasation represents a promising strategy to break the metastatic cascade and inhibit tumor angiogenesis (29). Transwell endothelial migration assay utilizes endothelial cell lining cultured on a permeable filter that allows growth factor exchange and cells to migrate across. First, we allowed HUVECs to settle down in transwell for 2 days to ensure HUVEC stabilization as a monolayer on the membrane. Second, we cultured TRAMP-C2, DU145, and PC3 cells on a monolayer of HUVECs (described in the Materials and Methods section) and followed their invasion across the membrane after 12 hours of incubation with recombinant S100A4 ± mab6B12 antibody. The recombinant-S100A4 protein significantly induced prostate cancer cells extravasation through the endothelial barrier as indicated by the number of cells crossing the membrane (Fig. 2Bi–iii). However, mab6B12 therapy significantly (P < 0.05) reduced the invasion-inducing effects of recombinant-S100A4 protein as evident from the number of prostate cancer cells extravasating the endothelial cell barrier (Fig. 2Bi–iii). Of note, we found that mab6B12 therapy alone also decreased the transendothelial migration of TRAMP-C2, DU145, and PC-3 cells (Fig. 2Bi–iii).

The most common sites where prostate tumor cells metastasize are lymph nodes and bone (70%–80% autopsies) in patients (30). The biology of normal bone development involves mineralization, laying down of calcium/or phosphate, collagenous proteins, and noncollagenous protein layers to form the framework of mature skeletal bone. Koenemen and colleagues showed that the presence of bone matrix noncollagenous proteins enable prostate cancer cells to acquire the ability to “home” to bones, and become “bone-like” or express osteomimetic properties (31). According to Magnusson and colleagues, prostate cancer metastases in bone are primarily sclerotic (osteoblastic), that is, they build bone (32). Notably, both bone deformation and bone loss are prominent features in metastatic prostate cancer (33). Koenemen and colleagues showed that both osteoblast (bone formation) and osteoclast (bone degradation) events are observed during progressive phases of prostate cancer metastasis (31). Studies have shown an abundant expression of S100A4 in rheumatoid arthritis and bone damage (34). Erlandsson and colleagues showed that S100A4 is essential for bone resorption and regulates osteoclast function (35). These data prompted us to investigate the effect of extracellular-S100A4 on matrix bone mineralization. For this purpose, we used hMSCs (PromoCell), which are derived from bone marrow matrix and can differentiate under in vitro conditions into chondrocytes and osteoblasts. We generated bone microenvironment by culturing hMSCs in osteogenic culture medium in the presence or absence of recombinant S100A4 protein (1 μg/mL) for 22 days. At this point, we determined the level of mineralization on hMSCs by employing Alizarin-red staining. We observed that the addition of recombinant S100A4 protein does not exert any effect on the proliferation, although significantly delayed the matrix mineralization of hMSCs (Fig. 2Ci). Similarly, hMSCs exposed to conditioned culture media of PC3 and TRAMP-C2 cells exhibited a reduced matrix mineralization (Fig. 2Cii–iii). The mab6B12 antibody significantly neutralized the S100A4 protein-induced delay of matrix mineralization of hMSCs (Fig. 2Ci–ii).

We next validated the effects of S100A4 protein induced matrix demineralization of hMSC by performing immunoblot analysis of hMSC cell extracts. Osterix (or SP7) transcription factor is expressed in osteoblasts of all endochondral and membranous bones (36). Osterix is a critical component of the bone matrix and is necessary for mineralization in hMSCs (36).The recombinant S100A4 protein was found to reduce the expression of Osterix in hMSCs, indicating reduced mineralization or bone loss (Fig. 2Di–ii and Ei–ii). These data are concomitant to the Alizarin staining data of MSCs as presented in Fig. 2Ci–Cii. Notably, mab6B12 treatment significantly neutralized the effects of recombinant S100A4 protein on the bone mineralization as evident from the level of Osterix protein in hMSCs and mMSCs (Fig. 2Di–ii and Ei–ii). These data prompted us to investigate relevance of Osterix in prostate cancer disease in human patients. Using an UALCAN web platform (37), we analyzed the tumor genome data of TCGA-PRAD clinical study comprised of normal subjects (n = 52) and patients with prostate cancer (n = 497). The data show that Osterix expression is reduced in prostate cancer when compared with normal (Fig. 1F). We next determined the status of Osterix gene in metastatic bones of patients with prostate cancer. We analyzed tumor genome data of metastatic prostate cancer Provisional Project 2019 for Osterix using cBioportal platform (19). We observed that Osterix gene copy number alterations (deep deletions, diploid; Supplementary Fig. S1D). An insight into the mechanism that drives Osterix expression by extracellular-S100A4 warrants a detailed investigation, which is beyond the scope of current study.

We reported the potential of extracellular S100A4 protein as a growth factor in prostate cancer (12). Because mab6B12 showed the potential of neutralizing S100A4 protein in situ, we asked whether mab6B12-neutralizing antibody therapy can rescue prostate tumor cells from the growth-inducing effects of extracellular-S100A4. We tested the efficacy of mab6B12 against the proliferation of prostate cancer cells representative of bone metastasis (PC3, which secretes S100A4 in abundance) and primary CRPC (22Rν1, which secrets low amount of S100A4) by employing a ³[H]thymidine uptake assay (12). Prostate cancer cells incubated with recombinant S100A4 protein (2.0 μg/mL) exhibited significantly increased rate of proliferation (Fig. 3Ai; Supplementary Fig. S1C). The mab6b12 therapy was found to reduce the proliferation of prostate cancer cells. Notably, mab6B12 (1 μg/mL) significantly blocked the growth-inducing effect of extracellular-S100A4 protein on PC3 (Fig. 3Ai) and 22Rν1 cells in vitro (Supplementary Fig. S1C). These data strengthened the hypothesis that S100A4-targeted mab6B12 therapy has the efficacy to block the proliferation and invasiveness of prostate cancer cells, which grow in a microenvironment rich with extracellular-S100A4 protein.

To test the efficacy of mab6B12 against prostate cancer disease, we employed a TRAMP-C2 cell-derived tumor growth syngeneic mouse model. TRAMPC2 cells are derived from prostate tumor of TRAMP mice, an autochthonous transgenic model that mimics human prostate cancer development. These cells also express molecules/receptors associated with immune system function. The advantage of using TRAMP-C2 cells is that they express S100A4 protein and secrete S100A4 protein in abundance (Fig. 1Bi–ii). To achieve our objective, TRAMP-C2 cells (1 × 10⁶) were transplanted in subcutaneous region in 6-week-old male immunocompetent C57BL/6J mice (The Jackson Laboratory). Tumor in these mice could grow till an approximate volume of 100 mm³ and animals at this point were randomly distributed in two groups. Group I (n = 10) served as control and received an IgG isotype control (7.7 mg/kg; 3 times/week). Group II (n = 10) received mab6B12 (7.5 mg/kg, 3 times/week; intraperitoneal route). Body weight and tumor volume were assessed on a weekly basis. A tumor volume of 1,000 mm³ was a preset endpoint for animal withdrawal from study. During the study, one mouse from control and treated group had to be withdrawn and censored from the study because of humane reasons.

The tumor growth was noted as progressive in control group when measured in terms of tumor volume as a function of time. We found that the majority (>80%) of control mice reached a preset endpoint tumor volume by 6 weeks of postimplantation (Fig. 3Aii). Comparing at matching time point (6 weeks), the tumor growth in mab6B12-treated group was recorded significantly (P < 0.05) lower than control group (Fig. 3Aii). The growth-inhibiting effect of mab6B12 antibody was recorded significant in 70% (P > 0.05) of mice. Control tumors were found to grow at an average of 136 mm³/week, whereas mab6B12 recipients exhibited a tumor volume turnover of 64 mm³/week. As shown in the Kaplan–Meier survival curve, the observed differences in tumor development between treated and control groups were statistically significant (P < 0.04; Fig. 3Aiii).

At 6 weeks, we sacrificed mice from all groups. We compared the weight of tumors harvested from control and treated mice. As compared with an average weight of 1,750 mg/tumor in control animals, the mab6B12-treated mice exhibited a 50% reduction (750 mg/tumor; P < 0.05; Fig. 3Aiv). The mab6B12 therapy significantly reduced the tumor weight and size, thus suggesting a strong impact on the proliferation of tumor cells similar to findings observed in Fig. 3Ai. These data establish the efficacy of mab6B12 is an effective immunotherapeutic-neutralizing antibody with a potential of slowing the growth of prostatic tumors. It is to be noted that the mab6B12 therapy did not cause any systemic toxicity in mice and the average body weight increased equally in control and treated mice.

We measured the serum-S100A4 protein levels in control and therapy-recipient mice at the end of therapy. Compared with the control, the mab6B12-treated group displayed a decreased circulatory serum-S100A4 protein level (Fig. 3Av). Control tumor-bearing mice exhibited an average of 150 ng/mL of serum-S100A4, whereas it was recorded 75 ng/mL in mab6B12-treated group (Fig. 3Av). We speculate that the lowered serum-S100A4 levels in mab6B12-treated mice could be attributed to the protein neutralization effect by mab6B12 antibody.

We previously reported that extracellular-S100A4 protein exerts its growth and metastasis promoting effects by activating NFκB pathway and MMP9 in prostate cancer cells and TRAMP mice (12). S100A4 is reported to act as an angiogenesis factor (36). NFκB-regulated genes (MMP9 and VEGF) play an important role in metastasis and angiogenesis (4, 13). It is noticeable that S100A4 regulates NFκB and MMP9 activities in prostate cancer (12, 36). Therefore, we determined the effect of mab6B12 therapy on the expression of downstream targets of S100A4 (VEGF and MMP9) by performing the IHC analysis of harvested tumors. The IHC analysis showed an increased level of MMP9 and VEGF proteins in tumors of control mice (Fig. 3B). On the contrary, the mab6B12-treated animals exhibited a decreased level of MMP9 and VEGF expression in tumors (Fig. 3Bi–ii). We measured the differences between control and treated mice by quantifying the immunostaining of tissue sections who underwent IHC. On the basis of intensity score (0 = none, 1–2 = weak/modest, 3–4 = high/very high), we observed a significant reduction in VEGF and MMP9 protein levels concomitant to the reduced volume of tumors in mab6B12 therapy–respondent animals (Fig. 3Bvi–Bv). Because mab6B12 antibody treatment was observed to reduce the rate of proliferation in prostate cancer cells in vitro, we asked whether similar could be observed under in vivo conditions. This was validated by counting Ki-67 immuno-positive (proliferation marker) cells in tumors. The IHC of tumors showed that therapy-respondent animals exhibit reduced number of Ki-67 immune-positive cell, thus suggesting a positive impact of mab6B12 therapy on prostate cancer cell growth in vivo (Fig. 3Biii and 3Bvi). These data suggest that mab6B12 therapy hits the target perfectly and performs well under in vivo conditions.

S100A4 protein is secreted by tumor cells, stroma-associated fibroblasts, and immune cells to generate an inflammation-rich microenvironment (13). S100A4 is a major chemoattractant for inflammatory cells, including macrophages, neutrophils, lymphocytes, dendritic cells (DC), and MDSCs (38, 39). It is reported that the presence of Th2 cells and M2-polarized macrophages at the tumor site help in tumor progression (40). We next asked whether neutralizing of the extracellular S100A4 protein by mab6B12 therapy modulates the tumor immune profile. For this purpose, the tumor tissues harvested from control and therapy-responsive groups were analyzed for infiltration of T cells, macrophages, and neutrophils. The IHC analysis of tumors for CD3, CD4, and CD8 markers showed that mab6B12 therapy causes a reduction in the level of T cells (Fig. 4Ai). The number of GATA-3, CD45, and CD68-positive cells were significantly reduced, whereas the number of T-Bet and STAT-4–positive cells showed significant increase in mab6B12-treated mice (Fig. 4B and C). CD209 is a marker for DCs, which play a pivotal role in the induction of antitumor immune responses (41). It is noticeable that mab6B12 therapy was observed to cause an increase in the infiltration of CD209⁺ DCs (Fig. 4B).

Previous studies have demonstrated that Th1/Th2 ratio balance is upset during tumorigenesis (39). Recent reports show that S100A4 stimulates production of cytokines from T lymphocytes and shifts the Th1/Th2 balance toward Th2 that is favorable for cancer progression (39). The IHC analysis showed a reduced T-cell Infiltration in therapy-sensitive tumors by mab6B12 therapy. We next determined the effect of mab6B12 therapy on cytokine levels in the blood harvested from control and therapy-receiving mice. Serum isolated from the blood was evaluated for cytokine profile using a microarray composed of a battery of 22 cytokines (GCSF, GM-CSF, IL2, IL3, IL4, IL5, IL6, IL9, IL10, IL12 p40/p70, IL12p70, IL13, IL17, IFNγ, MCP-1, MCP-5, RANTES (CCL5), SCF, sTNFRI, TNFα, thrombopoietin, and VEGF). The cytokines of interest showing the major differences between control and mab6B12 therapy–sensitive mice are presented in Fig. 5A. We show that mAb6B12 therapy caused a reduction in the circulatory levels of protumorigenic cytokines IL4, IL5, IL6, IL12, IL13, and RANTES levels in mice (Fig. 5Bi–vi). The data are concomitant to IHC-based findings in therapy-responsive tumors who displayed a decrease in Th cells (Th2) that are known to secrete IL4, IL6, and IL13 (42). IL4, IL6, IL12, IL13, and RANTES levels are known to be elevated in several human cancers (42).

RANTES is reported as a critical regulator of tumor cell metastasis (42, 43). The data about RANTES being the responder to S100A4-targeted antibody therapy is of utmost importance given the fact that it has previously been reported to stimulate the secretion of S100A4 from the fibroblasts (44). To establish RANTES as a bioresponder for S100A4-targeted therapies, it becomes imperative to investigate the correlation between S100A4 and RANTES under clinical settings. Because S100A4 gene alteration existed in almost 12% patients with prostate cancer in SU2C/PCF clinical study (Fig. 1A), we analyzed tumor genome data of this patient cohort (n = 429) to determine RANTES status. It is noticeable that S100A4 expression was found to significantly correlated to RANTES expression in patients with metastatic prostate cancer disease (Fig. 5Ci).

The importance of RANTES as a downstream target of S100A4 is high because it is a secretory protein, which is measurable by ELISA. We suggest that the outcome of S100A4-targeted therapies can be measured by secretory-RANTES levels in prostate cancer. For this purpose, we first determined levels of serum-RANTES in primary prostate cancer (n = 15) and metastatic prostate cancer (n = 34) patients by using a human-specific ELISA. The statistical analysis (t test) shows a significant difference between serum-RANTES levels of primary prostate cancer (8.06 ± 3.8 ng/mL; mean ± SD) and metastatic prostate cancer patients (26.90.5 ± 7.8 ng/mL; mean ± SD; P < 0.001; Fig. 5Cii).

Next, we measured serum-RANTES level in prostate cancer syngeneic mouse model, which received S100A4-targeted mab6B12 antibody therapy by employing mouse-specific ELISA. We observed that mab6B12 treatment significantly reduced the serum-RANTES levels in TRAMP-C2 tumor-bearing syngeneic mice, suggesting the impact of S100A4-targeted antibody therapy (P < 0.05; Fig. 5Ciii).

CD4⁺ T cells play a central role in immune protection. The immunosuppression is caused by an imbalance between Th1 and Th2 subpopulations of CD4⁺ T cells (40). We next determined the association of extracellular S100A4-induced changes and mab6B12 therapy on the expression of genes associated to CD4⁺ Th cells by using a qPCR-based microarray containing 96 well-characterized genes associated to Th1-Th2 genome (Th1-Th2 RT Profiler, Qiagen). For this purpose, we isolated T cells from the spleen of wild-type mice. The isolated T cells were cocultured with TRAMP-C2 cells in the presence of either control IgG or rhS100A4 protein or mab6B12 antibody. From the coculture, we harvested CD4⁺ cells using CD4⁺ Isolation Kit (Miltenyi Biotec). The RNA isolated from CD4⁺ T cells was subjected to microarray analysis (Th1-Th2 RT Profiler, Qiagen). The analysis of data show that the expression of Th2 cytokines (IL4, IL4R, IL5, STAT4) is decreased by mab6B12 therapy in CD4⁺ T cells (Fig. 5D). The data showing the effect of mab6B12 therapy on total Th1-Th2 associated genome in CD4⁺ cells is provided in Supplementary Fig. S1E. To summarize, our data suggest that the efficacy of mab6B12 could be associated to its property to quench the extracellular S100A4 protein and shifts the immune balance favoring inhibition of tumor growth in the tumor microenvironment.

Prostate cancer has seen an increase in intermediate and early-stage disease at diagnosis, with widespread use of PSA to screen for prostate cancer and an increase in the number of needle cores taken during biopsy. Although Gleason score and other clinicopathologic variables are used to guide treatment decisions. Currently, the established prognostic factors (histologic grade, stage, and PSA) at diagnosis are insufficient to separate patients with prostate cancer who are at high risk for cancer progression, and to date, molecular markers for metastatic propensity remain elusive. The discovery of molecular biomarkers for prostate cancer has been hindered by the paucity of molecular subtypes with distinct outcomes.

DECIPHER-genomic testing has recently gained significant traction in terms of disease outcome in the biopsy setting and postinitial therapy (such as RP and RT) in patients with prostate cancer (16, 17). DECIPHER-testing depends on a set of gene signature-based algorithms and each algorithm predict an outcome parameter in patients (postinitial biopsy or postprimary treatment). In this context, the performance of biopsy-S100A4 expression as a standalone marker predicting ADT outcome and risk of mortality (which is generally associated to metastasis) in RP-treated patients is significant. Furthermore, our data strongly support the notion that S100A4 expression and copy number could be used to separate the patients who are at high risk and propensity to have biochemical recurrence after therapy. Therefore, S100A4 copy number burden in biopsies could be used a global measure of the risk of recurrence and adds additional information to currently available clinicopathologic variables. We suggest that S100A4's association with poor ADT response should lead to new treatment strategies for treating ADT-resistant CRPC disease in patients. This might include potential packaging a S100A4-targeted therapy along with ADT or hormone therapy in patients

Cancer progression is an intricate process that depends on the mutual communications between tumor cells and the microenvironment. The imbalance between protumor or antitumor immune cells causes the tumor progression. We speculate that extracellular S100A4 protein, proinflammatory in nature, is a critical player that shifts the balance of immune cells in primary tumor toward disease progression. We previously reported that S100A4 drives prostate cancer metastasis in human cells and TRAMP mice (12). We had reported that prostate tumor and stromal cells secrete S100A4 protein that serves as a growth factor for prostate cancer disease (12, 45). The central theme of the current study was to capitalize on our previous findings in vitro and investigate its clinical relevance by testing the blood of patients with prostate cancer for S100A4 protein level. The significant outcome of this study is that it lays a foundation for exploiting the utility of serum-S100A4 levels as a noninvasive biomarker for prostate cancer progression and prognosis in clinical settings; however, this warrants an investigation in large cohort of patients. Al-Ebad and colleagues reported that high-serum S100A4 mRNA level is associated to poor prognosis in patients with breast cancer (46). However, our study has an advantage over the published report because it establishes a protein-based biomarker that is feasible, accurate, and economical than the RNA-based assay (that is dependent on secretory tumor cells or cell-free exosomes, which is laborious and expensive).

Targeting an oncoprotein has always remained a prime focus for investigators to treat the cancers (14). However, the conundrum is that it becomes a tiring task to tame the activity of an oncoprotein that functions both at intracellular and extracellular level in the tumors. The S100A4 oncoprotein is one of the molecules that elicits its effect at both intracellular and extracellular level (12). We reported that acquiring a metastatic phenotype and aggressiveness of prostatic tumor cells is dependent on extracellular-S100A4 protein in vitro (12). In this study, we focused our attention to tame the activity of extracellular or secretory S100A4 protein for treating prostate cancer by using a novel anti-S100A4 neutralizing antibody (25). The outcome of in situ, in vitro, and in vivo studies shows the performance of mab6B12 as a strong neutralizer of extracellular-S100A4 protein, particularly under physiologic conditions. This study also provides a peak into the changes in the immune system of animals following the mab6B12 antibody therapy.

T cells are broadly classified into Th cells (CD4⁺) or cytotoxic T cells (Tc cells, CD8⁺; ref. 47). Th1 cells primarily produce IFNγ and IL2, whereas Th2 cells produce IL4, IL5, IL6, IL10, and IL13 (47). Under physiologic condition, it is important to maintain a proportionate balance between Th1 and Th2 cytokines. In many disease conditions including cancer, the fine balance between Th1 and Th2 is disturbed and the equilibrium polarizes from Th1 to Th2, causing immunosuppression. Recently, S100A4 protein has been reported to activate T cells by JAK/STAT pathway and recruit immune cells (T, macrophages, NK cells) to the tumor site (25, 38). Grum-Schwensen and colleagues reported that S100A4 modulates the tumor microenvironment by attracting T cells and macrophages, which helps in tumor progression (39). Klingelhöfer and colleagues reported that in premetastatic lungs, cytokines secreted by T cells in response to S100A4 protein exhibit a shift in Th1/Th2 balance toward Th2 characteristic (25). However, our observation in prostatic tumor model slightly differed from Klingelhöfer and colleagues' report (25). We found that mab6B12 therapy caused no major change in T-cell infiltration in prostatic tumor status, although significantly caused a polarization of T cells (Th1/Th2) in the tumor microenvironment favoring growth inhibition. Nevertheless, our data are in agreement with the report by Weatherly and colleagues who showed decrease in T cells in primary tumors under S100A4-knockdown condition in S100A4(^−/−) PyMT in mice (9).

Bruhn and colleagues have reported an S100A4-mediated increase of IL13 cytokine production in T cells (48). IFNγ and IL4 are critical cytokines that are produced by Th1 and Th2 cells. IL12 and IFNγ signals are important for Th1 cell differentiation. The transcription factors Stat4 and T-bet are required for the development of Th1 cells (49, 50). IL4 is reported to induce the STAT-6 activation, which promotes the expression of GATA-3, an essential transcription factor for both IL4 production and the development of Th2 cells (50). GATA-3 is reported to inhibit the Th1 cell–specific transcription factors (50). In this context, our data are important because we provide evidence that mab6B12 therapy causes an upregulation of T-bet and Stat4 and downregulation of GATA-3, which are strong indicators of a shift in Th1/Th2 balance toward Th1 cells. Among the cytokines that were decreased by mab6B12 therapy, IL4, IL6, and IL13 belong to Th2 type cytokine.

RANTES, reported to be a target gene of NF-κB activity, is expressed by T lymphocytes, macrophages, platelets, tubular epithelium, synovial fibroblasts, and tumor cells (42). RANTES is reported to stimulate secretory S100A4 in mast cells and certain classes of T cells (43). RANTES plays an active role in recruiting leukocytes into inflammatory sites and functions in collaboration with IL2 and IFNγ (42). The exact function of RANTES in tumor biology is not yet completely known; however, studies show that RANTES/CCR5 (target receptor) interactions stimulate angiogenesis, induce the recruitment of stromal and inflammatory cells, and promote the immune evasion by tumor cells (42). Given the fact that RANTES is known to be regulated by NF-κB, one possible mechanism underlying the RANTES downregulation by S100A4 targeting mab6B12 therapy could be attributed to the loss of S100A4-induced RAGE/NFκB pathway in tumors growing in syngeneic mouse model.

To summarize, we show that biopsy-S100A4 combined with serum-S100A4 level could be developed as a predictive biomarker for outcome of disease in patients with prostate cancer after initial diagnosis or initial treatment (such as RP). Of note, serum-S100A4 as a potential noninvasive biomarker would add value to the existing standard clinical biomarkers in detecting recurrence of disease more accurately. In addition, we suggest that S100A4 neutralizing antibody-based therapies have a potential for their use under clinical settings for treating prostate cancer including NECaP (which so far has remained nontreatable disease in patients). These studies also open the window of opportunity of exploiting the applicability of anti-S100A4 antibodies as imaging tools for the detection of micrometastatic tumors. Our laboratory is currently developing anti-S100A4 antibodies, which could be exploited for their theranostic use in clinical settings for patients with prostate cancer.

J. Klingelhöfer reports a patent for WO2014068300A1 issued. N. Ambartsumian reports a patent for WO2014068300A1 issued. C.A. Warlick reports other from Genomic Health (research relationship) and other from Francis Medical (research relationship) outside the submitted work. No potential conflicts of interest were disclosed by the other authors.

A.A. Ganaie: Formal analysis, validation, investigation, methodology, writing-original draft, writing-review and editing. A.P. Mansini: Data curation, formal analysis, investigation, methodology. T. Hussain: Investigation. A. Rao: Resources. H.R. Siddique: Formal analysis, investigation, methodology. A. Shabaneh: Data curation, software, investigation, methodology, writing-original draft. M.G. Ferrari: Investigation, methodology. P. Murugan: Resources, formal analysis. J. Klingelhöfer: Resources. J. Wang: Data curation, software, formal analysis, investigation. N. Ambartsumian: Resources, formal analysis. C.A. Warlick: Resources, visualization, methodology. B.R. Konety: Resources, methodology. M. Saleem: Conceptualization, resources, formal analysis, supervision, funding acquisition, investigation, visualization, writing-original draft, project administration, writing-review and editing.

This study was supported by an R01 grant from US PHS/NIH (CA193739) to M. Saleem. The research activity of M. Saleem was also supported by US PHS grants (CA184685 and CA184685-02S1), University of Minnesota Medical School, and Masonic Cancer Center research support funding. B.R. Konety is supported by a US PHS grant (U54MD008620) and DOD grant (W81XWH‐17‐1‐0462). We are thankful to Dr. Elai Davicioni (Chief Scientific Officer, Decipher Biosciences, Inc.) for DECIPHER-genome testing of Minnesota cohort (228 RP-treated patient specimens). We thank Dr. Eric Roush (Principal Scientist, Biacore, Life Sciences) for helping the team in performing SPR studies at the shared resources facility of the Hormel Institute, University of Minnesota. We thank J. Klingelhöfer, N. Ambartsumian, and Cancer Research Technology (London, United Kingdom) for providing mab6B12 antibody. We thank Dr. Thomas Griffith (Dept. of Urology, University of Minnesota) for evaluating immune-cell data. We thank Colleen Forster for helping in histopathology work. We thank Neelofar Jan for providing help in animal studies.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.