Ferroptosis Inducers Are a Novel Therapeutic Approach for Advanced Prostate Cancer

Prostate cancer is the most commonly diagnosed noncutaneous malignancy in U.S. men (1). Prostate cancer accounts for 30,000 deaths annually in the United States, almost always from metastatic disease (1). The mainstay of treatment for advanced prostate cancer is androgen deprivation therapy (ADT; ref. 2). Although ADT is initially effective in nearly all men, the disease commonly recurs, referred to as castration-resistant prostate cancer (CRPC; ref. 3), which is largely responsible for prostate cancer–associated deaths. Current therapeutic approaches for CRPC include second-generation antiandrogens, such as enzalutamide, abiraterone, apalutamide, and darolutamide, taxane-based chemotherapy, immunotherapy, and targeted therapies (4–13). While adenocarcinoma positive for androgen receptor (AR) is the predominant histologic variant of CRPC (adeno-CRPC), 10%–15% of metastatic CRPCs present neuroendocrine phenotype called neuroendocrine prostate cancer (NEPC) and 20%–25% present double-negative phenotype (DNPC; refs. 14–19). NEPC and DNPC are characterized with loss of expression of AR resulting in resistance to therapies that target the AR pathway and aggressive clinical behavior (14–18). Presence of neuroendocrine markers is a characteristic of NEPC, while DNPC exhibits lack of neuroendocrine markers (14–18). Currently, there are no long-term effective or curative treatments available for adeno-CRPC, NEPC, and DNPC, and thus, exploring novel therapeutic approaches for advanced prostate cancer is critical.

Ferroptosis is an iron-dependent programmed cell death mechanism that is induced by the accumulation of lipid peroxidation (20–24). Previous studies revealed that ferroptosis is characterized by accumulation of peroxidation of phospholipids enriched with polyunsaturated fatty acids and reactive oxygen species (ROS; ref. 25). Glutathione peroxidase (GPX4) and solute carrier family 7 member 11 (SLC7A11) are two major regulators of ferroptosis (6, 25–29). GPX4 utilizes reduced glutathione to convert lipid hydroperoxides to lipid alcohols, thereby alleviating lipid peroxidation and inhibiting ferroptosis, while SLC7A11 is a transmembrane transporter that exchanges extracellular cystine for intracellular glutamate (6, 25–29). Loss or pharmacologic inhibition of GPX4 or SLC7A11 leads to ferroptosis induction (26, 27, 30). Ferroptosis is involved in pathophysiologic processes of various diseases, including cancers, and can act as a natural barrier to tumor progression (31, 32). In cancer, ferroptosis inducers (FIN) have shown a promising anticancer activity in models of multiple cancer types (24, 33, 34). Ferroptosis was recognized as a distinct mechanism of nonapoptotic programmed cell death through a small-molecule screen, and erastin and RSL3 were identified as compounds that induce selective lethality in cancer cells that express mutant HRAS (24, 35–38). RSL3 is known to inhibit GPX4, and loss or inhibition of GPX4 leads to induction of ferroptosis in cancer cells (24, 26). Erastin inhibits the cystine/glutamate transporter system Xc- composed of SLC7A11 and SLC3A2 amino acid transporters, and has been shown to induce ferroptosis across cancer types (24, 38, 39). Previous studies have shown that drug-resistant cancer cells are vulnerable to GPX4 inhibition and ferroptosis induction (33, 34). Moreover, induction of ferroptosis enhances the therapeutic efficacy of cisplatin in cancer cells (40, 41), suggesting that FINs may be even more potent in combination therapy settings. In the context of prostate cancer, it has been shown that DECR1 is an androgen-regulated survival protein that protects cells from ferroptosis and targeting DECR1 induces ferroptosis (42). In addition, treatment with enzalutamide induces lipid peroxidation and leads to sensitivity to GPX4 inhibition and ferroptosis in vitro (43). However, the therapeutic potential of FINs, erastin and RSL3, in prostate cancer has not been tested in vivo.

Herein, we performed preclinical assessment of the therapeutic potential of two FINs, erastin and RSL3, in treatment-resistant prostate cancer. We demonstrated that ferroptosis mediators, SLC7A11, SLC3A2, and GPX4, are expressed in adeno-CRPC, DNPC, and NEPC xenografts. Erastin and RSL3 increase ROS production and impair cell viability, growth, and migration of prostate cancer cells in vitro. Erastin and RSL3 also significantly delay prostate cancer tumor growth in vivo. Treatment with FINs in combination with second-generation antiandrogens, enzalutamide and abiraterone, halted prostate cancer cell growth and migration in vitro and tumor growth in vivo. Our study demonstrates that prostate cancer cells are vulnerable to ferroptosis induction, and FINs may represent a new class of therapeutic agents for advanced prostate cancer as single agents and in combination with standard-of-care therapies for CRPC.

Indicated cell line–derived xenograft tumor samples were fixed in 10% formalin overnight at 4°C, followed by immersion in 70% ethanol, and subsequently embedded in paraffin. Samples were sectioned at 4 μm and affixed to slides. Sections from cell line–derived xenograft or patient-derived xenograft (PDX) tumor samples were heated to 65°C for 1 hour, then moved to clarify to remove paraffin, followed by rehydration in sequential ethanol (100%, 95%, and 70%) to rehydrate. After 10-minute water incubation, antigen retrieval was performed in 10 mmol/L sodium citrate buffer, pH 6, at 95°C for 20 minutes. Samples were allowed to cool to room temperature and rinsed, and then incubated with 3% hydrogen peroxide for 10 minutes to block endogenous peroxidase activity. Tumor samples were blocked with 2.5% goat serum for 1 hour, followed by incubation with primary antibodies, anti-SLC7A11 (PA5-33050, 1:100) and anti-GPX4 (sc-166120, 1:100), overnight at 4°C in a humidifying chamber. After three washes with 1× PBS, samples were incubated with anti-mouse horseradish peroxidase (HRP)-conjugated secondary antibody for 1 hour at room temperature, followed by developing with DAB Reagent (Dako), and counterstaining with hematoxylin. The PDX tissue samples were subjected to manual blinded scoring for intensity of staining as low, moderate, and high, as shown in Supplementary Fig. S1A and S1B.

Human prostate cancer cell lines, DU145, PC3, 22Rv1, LNCaP, and NCI-H660, were purchased from the ATCC. ARCaP was purchased from Novicure Biotechnology. C4-2 cells were a kind gift from Dr. Owen Witte (University of California, Los Angeles, Los Angeles, CA). DU145, PC3, 22RV1, ARCaP, C4-2, and LNCaP were cultured and maintained in RPMI1640 Medium (Thermo Fisher Scientific) supplemented with 10% FBS, 100 U/mL penicillin, and 0.1% streptomycin, and 1% GlutaMAX. NCI-H660 cells were maintained in RPMI1640 medium supplemented with 5% FBS, 0.005 mg/mL insulin, 0.01 mg/mL transferrin, 30 nmol/L sodium selenite, 10 nmol/L hydrocortisone, 10 nmol/L beta-estradiol, and 1% GlutaMAX. Cell lines were authenticated at the Stanford Functional Genomics Facility for short tandem repeat profiling and routinely assayed for Mycoplasma using MycoAlert Mycoplasma Detection Kit (Lonza).

Erastin, (1S,3R)-RSL3, and ferrostatin-1 were purchased from APExBIO for in vitro experiments. The compounds were dissolved in DMSO at 10 mmol/L concentration and stored at −20°C. For in vivo experiments, erastin was obtained from BOC Sciences and (1S,3R)-RSL3 was purchased from Cayman Chemical. Enzalutamide and abiraterone acetate were purchased from TargetMol. Enzalutamide was dissolved in DMSO, and abiraterone was dissolved in either absolute ethanol or DMSO for in vitro, as described previously (44–48) and in vivo studies, respectively.

A total of 5,000 cells per well were seeded into 96-well plates. After incubation for 24 hours at 37°C, cells were treated with erastin (1.25, 2.5, 5, 10, and 20 μmol/L) or (1S,3R)-RSL3 (0.125, 0.25, 0.5, 1, 2, and 4 μmol/L), and matched volumes of DMSO were used as a vehicle control. Seventy-two hours posttreatment, the viability of cells was assayed with the CellTiter-Blue Cell Viability Assay Kit (Promega) according to the manufacturer's instructions and measured with a Tecan plate reader. Cell viability was calculated as percentage (%) compared with the control (0 or vehicle treatment) for each cell line as follows: the mean absorbance of the control (0 or vehicle treatment) was set to 100%. Percentage viability in each control technical replica was calculated: each control technical replicate/mean absorbance control × 100. For the treatment arms, the percentage cell viability was equal to the absorbance of each treatment technical replicate/mean absorbance control × 100.

A total of 5 × 10⁴ DU145, PC3, or C4-2 cells were cultured in 24-well plates overnight at 37°C. The next day, cells were treated with erastin (1.25, 2.5, and 5 μmol/L) or (1S,3R)-RSL3 (0.125, 0.25, and 0.5 μmol/L) for 72 hours. Cells were washed with PBS and harvested with trypsin/EDTA (0.25%), followed by resuspension in PBS with 0.5 μg/mL 7AAD (BioLegend, 420403) for 10 minutes in the dark before analysis using flow cytometry. To analyze live–dead cell ratios, treated cells were stained with 0.4% trypan blue solution. The percentage of live–dead cells was quantified using Countess II Automated Cell Counter (Thermo Fisher Scientific).

DU145, PC3, ARCaP, 22Rv1, C4-2 (500 cells/well), and LNCaP (5,000 cells/well) cell lines were grown in 6-well plates overnight to allow cells to attach. Cells were treated with erastin (5 μmol/L) or (1S,3R)-RSL3 (0.5 μmol/L) for single-treatment experiments. For combination treatment experiments, C4-2 cells were treated with erastin (2 μmol/L), (1S,3R)-RSL3 (50 nmol/L), enzalutamide (2 μmol/L), and abiraterone acetate (2 μmol/L) in vitro as described previously (44–48). Cells were cultured for 9 days, and media containing 10% FBS and compounds were changed every 3 days. Colonies were fixed with methanol and stained with 0.01% crystal violet solution for 1 hour at room temperature and washed with water. An equal volume of DMSO was used as a control.

Forty-eight hours prior seeding into 24-well transwell inserts, pore size 8 μm and 6.5 mm diameter (Transwell Permeable Polyester Membrane Inserts, Corning Inc), cells were treated with erastin (1.25 μmol/L), (1S,3R)-RSL3 (0.125 μmol/L), or an equal volume of DMSO as a control. A total of 5 × 10⁴ cells were then seeded in serum-free media in 24-well transwell inserts. The inserts were incubated in 10% FBS-supplemented media in 24-well plates for 20 hours. The bottom chamber was filled with 10% FBS-supplemented media containing erastin (1.25 μmol/L), (1S,3R)-RSL3 (0.125 μmol/L), or an equal volume of DMSO as a vehicle control. Cells that passed through the membrane were fixed and stained with 0.01% crystal violet solution. For migration assays with higher concentrations of compounds, cells were seeded in serum-free medium including erastin (5 μmol/L) or (1S,3R)-RSL3 (0.5 μmol/L) in 24-well transwell inserts. The inserts were then incubated in 10% FBS-supplemented medium including erastin (5 μmol/L) or (1S,3R)-RSL3 (0.5 μmol/L) in 24-well plates for 20 hours.

A total of 5 × 10⁴ DU145, PC3, or C4-2 cells were suspended in 10 μL Matrigel and pipetted as a droplet into a 24-well plate for 20 minutes to form Matrigel drop prior to adding media and compounds, as described previously (49, 50). DU145 and PC3 tumoroids were treated with erastin (1.25 and 5 μmol/L) or (1S,3R)-RSL3 (0.125 and 0.5 μmol/L) every 3 days for 6 days. C4-2 tumoroids were treated with erastin (5 μmol/L) or (1S,3R)-RSL3 (0.5 μmol/L). The radial distance the cell had migrated away from the edge of tumoroids was measured as radial migration on day 6. For combination therapy experiments, erastin or (1S,3R)-RSL3 was combined with enzalutamide or abiraterone acetate at the following doses: erastin (5 μmol/L), (1S,3R)-RSL3 (0.5 μmol/L), enzalutamide (5 μmol/L), and abiraterone acetate (5 μmol/L). Media containing 10% FBS and compounds were changed once at day 3. DMSO and ethanol were used as vehicle controls.

Briefly, 1 × 10⁵ cells were grown in 12-well plates overnight at 37°C. The next day, cells were treated with erastin (5 μmol/L) or (1S,3R)-RSL3 (1 μmol/L) for 6 hours, followed by the addition of 1 μmol/L H2DCF for 20 minutes. Cells were washed with PBS and harvested with trypsin/EDTA (0.25%), followed by washing twice with PBS. Cells were subjected to flow cytometry to measure the levels of cellular ROS.

Xenograft tissues (50 mg) were homogenized and lysed using RIPA lysis buffer containing Protease Inhibitor Cocktails (Thermo Fisher Scientific). Protein was quantified using the BCA assay, and an equal amount of protein for each lysate (50 μg) was resolved by 4%–12% gradient SDS-PAGE, followed by transfer onto nitrocellulose membranes. Membranes were blocked with 5% milk for 1 hour in TBS and probed with primary antibodies for overnight at 4°C in TBS. Washing was followed by incubation with secondary antibody for 1 hour at room temperature in TBS containing 0.1% Tween-20. Anti-GPX4 (sc-166120, 1:1,000 for Western blotting) and anti-GAPDH antibodies (sc-32233, 1:3,000) were obtained from Santa Cruz Biotechnology. Anti-SLC7A11 antibody was purchased from Invitrogen, Thermo Fisher Scientific (PA1-16775, 1:500 for Western blotting). Anti-SLC3A2 was purchased from Santa Cruz Biotechnology (sc-376815). Secondary antibodies with HRP were obtained from Thermo Fisher Scientific (anti-mouse PI31432 and anti-rabbit PI31462, 1:2,000). Pierce ECL Western Blotting Substrate (Thermo Fisher Scientific, PI32106) was used to develop chemiluminescence signals that were measured on an IVIS Lumina imager.

All animal studies and procedures were approved and performed in accordance with Stanford Administrative Panel on Laboratory Animal Care, Institutional Animal Care and Use Committee, as well as the USAMRMC Animal Care and Use Review Office. DU145, PC3, ARCaP, or C4-2 (5 × 10⁵) and NCI-H660 (1 × 10⁶) cells were suspended in 50 μL of 80% Matrigel. Tumor cells were implanted subcutaneously into the flank of 6- to 8-week-old NOD-SCID-IL2R γ (NSG; The Jackson Laboratory) male mice. Mice with established tumors with approximately 50–80 mm³ average volumes [measured by calipers and calculated as (length × width × height)/2] were randomized into treatment groups, including vehicle (DMSO), erastin (20 mg/kg in 20 μL DMSO plus 130 μL corn oil, i.p., daily) as described previously (51), and RSL3 (100 mg/kg in 20 μL DMSO plus 80 μL corn oil, i.p., biweekly; ref. 26). For combination therapy experiments, mice were randomized into different treatment groups, including vehicle, erastin (20 mg/kg), RSL3 (100 mg/kg), enzalutamide (10 mg/kg in 5% DMSO, 30% PEG 300, and 65% H₂O, oral gavage, daily), the combination of erastin with enzalutamide, or the combination of RSL3 with enzalutamide.

To examine the clinical relevance of ferroptosis in prostate cancer, we first assessed SLC7A11 and GPX4 protein levels in LuCaP PDX models derived from metastatic prostate cancer (Fig. 1A and B; Supplementary Fig. S1A and S1B; ref. 52). High levels of GPX4 were detected in adeno-CRPC (PDX, n = 36 and sample, n = 108), as well as NEPC PDX samples (PDX, n = 3 and sample, n = 9; Fig. 1B), while high levels of SLC7A11 were predominantly observed in adeno-CRPC (PDX, n = 35 and sample, n = 105; Fig. 1A). The protein levels of SLC7A11 and GPX4 were further tested in prostate cancer cell lines in vitro (Supplementary Fig. S1C and S1D). We also assessed SLC7A11 and SLC3A2 along with GPX4 protein levels in prostate cancer cell line–derived xenografts, including LNCaP (androgen sensitive), ARCaP AR low (adeno-CRPC), C4-2, 22Rv1 (adeno-CRPC), PC3 (NEPC-like characterized with lack of AR and expression of some of the neuroendocrine markers), DU145 (DNPC), H660 (NEPC), and Trop2-derived NEPC (TD-NEPC; ref. 49) by IHC (Fig. 1C–E). SLC7A11, SLC3A2, and GPX4 were expressed across all tested xenografts (Fig. 1C–E), but slightly different from what we observed by Western blotting (Supplementary Fig. S1C and S1D). Taken together, our results reveal that SLC7A11, SLC3A2, and GPX4 are expressed across androgen-sensitive, adeno-CRPC, NEPC, and DNPC xenografts.

To test whether prostate cancer is sensitive to ferroptosis induction, we treated prostate cancer cell representing different variants of prostate cancer, including AR-positive androgen-sensitive (LNCaP), AR-positive adeno-CRPC (C4-2), AR- and AR-V7 isoform–positive adeno-CRPC (22Rv1), NEPC (H660), NEPC like (PC3), AR-low adeno-CRPC (ARCaP), and DNPC (DU145) in vitro with different doses of erastin and RSL3 (Fig. 2A and B). All prostate cancer cell lines were vulnerable to ferroptosis induction mediated by erastin and RSL3 (Fig. 2A and B). Treatment with ferrostatin-1, an inhibitor of ferroptosis (21), rescued the cells from erastin-induced ferroptosis in vitro (Supplementary Fig. S2A). Treatment with erastin and RSL3 led to an increase in intracellular ROS levels, a hallmark of ferroptosis induction, in all prostate cancer cell lines, except H660, at 6 hours posttreatment initiation (Fig. 2C and D). As H660 cell lines were sensitive to erastin and RSL3 on viability assay at 72 hours posttreatment (Fig. 2A and B), it is plausible that in H660 cell line ROS production occurs later and immediately preceding the effects on viability or their sensitivity is potentially due to high levels of iron in these cells. In addition, H660 cell lines are slow growing nonadherent cells, which could be possibly associated with the time at which they release ROS. Both, erastin and RSL3 diminished colony formation of all tested prostate cancer cell lines (Fig. 2E and F). Together, these results suggest that ferroptosis induction is a promising approach to impair the growth of prostate cancer cells independent of their phenotype and AR status in vitro.

Cell motility and migration are critical for cancer progression, invasiveness, and metastasis. We investigated whether ferroptosis induction affected prostate cancer cell migration and invasion by utilizing a 3D Matrigel drop invasion assay (49, 50) and a transwell chamber migration assay. To ensure the rate of migration was not affected by cell death, we treated DU145, PC3, and C4-2 cells with erastin and RSL3 at concentrations that did not affect cell viability (Supplementary Fig. S2B and S2C). For 3D Matrigel drop assay, the radial distance the cells had migrated from edge of the Matrigel drop was measured on day 6 after plating the cells (Fig. 3A–D). Treatment of prostate cancer cells with two different doses of erastin or RSL3 significantly inhibited prostate cancer cell migration and invasion (Fig. 3A–D; Supplementary Fig. S3A and S3B). Likewise, treatment with erastin and RSL3 decreased the migration of PC3 and DU145 cells in a transwell migration assay (Fig. 3E and F; Supplementary Fig. S3C and S3D). Taken together, the results suggest that ferroptosis induction induces intracellular ROS production and reduces prostate cancer cell growth and migration.

Currently, there are no effective therapies for NEPC and DNPC. Hence, we further tested the effect of erastin and RSL3 on DNPC, adeno-CRPC with low AR, and NEPC. To test the therapeutic efficacy of erastin and RSL3 in vivo, we established subcutaneous xenograft tumor models of DU145 (DNPC), ARCaP (adeno-CRPC with low AR), PC3 (NEPC like), and H660 (NEPC) human prostate cancer cell lines in the flanks of immunocompromised (NSG) male mice. When tumor volumes reached an average size of approximately 50–80 mm³, animals were treated with erastin (20 mg/kg) or vehicle, administered intraperitoneally once daily (Fig. 4A). There were no measurable side effects observed in erastin-treated animals as assessed by animal body weight and lack of distressed behavior (Supplementary Fig. S4A). Treatment of mice with erastin led to a significant decrease in tumor growth with an increase in tumor necrosis in DU145, ARCaP, PC3, and H660 (P < 0.05) xenografts (Fig. 4B and C; Supplementary Fig. S4B). Similarly, treatment with RSL3 significantly decreased tumor growth and tumor weight at endpoint of DU145 (P < 0.0001) and PC3 (P < 0.01) xenografts with no measurable side effects as assessed by animal body weight and any signs of distress (Fig. 4D–F; Supplementary Fig. S4C and S4D). These data indicate that erastin and RSL3 decrease prostate cancer tumor growth in vivo.

We further tested the therapeutic potential of erastin and RSL3 in combination therapies with second-generation antiandrogens, enzalutamide or abiraterone. C4-2 adeno-CRPC cells that express AR were used for the studies as they are responsive to enzalutamide and abiraterone. Cells that are AR low or AR negative were not used for testing combinations, due to their resistance to agents targeting AR signaling axis. Erastin in combination with either enzalutamide or abiraterone dramatically reduced colony formation when compared with cells treated with either agent alone (Fig. 5A and B). Likewise, RSL3 in combination with enzalutamide or abiraterone decreased colony formation of prostate cancer cells when compared with treatment with RSL3, enzalutamide, or abiraterone alone (Fig. 5C and D). Treatment of C4-2 cells with erastin in combination with either enzalutamide or abiraterone significantly reduced C4-2 cell migration and invasion in vitro (Fig. 5E and F). Similarly, RSL3 in combination with enzalutamide or abiraterone inhibited C4-2 cell migration and invasion when compared with RSL3, enzalutamide, or abiraterone alone (Fig. 5G and H).

To evaluate the therapeutic efficacy of erastin and RSL3 in combination with antiandrogens in vivo, we treated established C4-2 xenografts with erastin (20 mg/kg, i.p., daily) and enzalutamide (10 mg/kg, oral gavage, daily) when tumor volumes reached approximately 50–80 mm³ on average (Fig. 6A). Consistent with the effects of erastin and RSL3 on adeno-CRPC with low AR, DNPC, and NEPC xenografts (Fig. 4), erastin and RSL3 significantly delayed the tumor growth of C4-2 xenografts (Fig. 6). Furthermore, combined treatment with erastin and enzalutamide significantly inhibited tumor growth as assessed by tumor volumes and tumor weights at endpoint when compared with treatment with vehicle, erastin, or enzalutamide alone (Fig. 6B and C). We did not observe any significant differences in body weight of animals treated with erastin and enzalutamide when compared with vehicle control and single-therapy arms (Fig. 6D). Furthermore, we tested the therapeutic potential of RSL3 in combination with enzalutamide in vivo using C4-2 xenograft model (Fig. 6E–H). Consistent with erastin, RSL3 in combination with enzalutamide halted tumor growth and was more potent than RSL3 and enzalutamide alone in vivo (Fig. 6E–G). We did not detect any measurable side effects as assessed by animal body weight and signs of distress in any of the treatments when compared with vehicle control (Fig. 6H). Therefore, the combination of erastin or RSL with second-generation antiandrogens represents a potent therapeutic strategy when compared with single-agent treatments for adeno-CRPC.

Our study provides the first demonstration of the therapeutic potential of erastin and RSL3 across prostate cancer variants, including adeno-CRPC, NEPC, and DNPC, in a preclinical setting. Our results warrant further evaluation of the efficacy of FINs in clinical settings for treatment of prostate cancer. In fact, multiple agents that have been shown to induce ferroptosis have been approved by the FDA for treatment of malignancies and other conditions. For instance, the multi-kinase inhibitor, sorafenib, used for treatment of hepatocellular carcinoma, renal cell carcinoma, and refractory differentiated thyroid carcinoma, has been shown to induce ferroptosis (39, 53). Sulfasalazine, a compound approved by the FDA for the treatment of rheumatoid arthritis and ulcerative colitis, has also been demonstrated to induce ferroptosis in cancer cells (54). Other FDA-approved agents, such as altretamine, used for treatment of ovarian cancer, inhibits GPX4 and induces ferroptosis in cancer cells (55). Testing these clinically used agents as CRPC therapies should be further tested in preclinical models and if effective, they could be rapidly translated into clinical trials.

We have also demonstrated that erastin and RSL3 are more effective in inhibiting tumor growth when combined with enzalutamide or abiraterone in a preclinical model of adeno-CRPC. The combination of erastin or RSL3 with enzalutamide or abiraterone inhibited cell growth and migration. Furthermore, combination of erastin or RSL3 with enzalutamide halted tumor growth when compared with erastin, RSL3, or enzalutamide alone in vivo. Therefore, agents effective at inducing ferroptosis alone should be tested in combination with second-generation antiandrogens in preclinical and clinical trials.

Enzalutamide acts, in part, through preventing the nuclear translocation of AR (56) and can inhibit stabilization of AR mediated by HSPs (57). Interestingly, several studies have shown that HSPs might inhibit ferroptosis induction. For example, overexpression of HSPB1 inhibits ferroptosis induction by erastin (58). HSPA5 also negatively regulates ferroptosis induction in human pancreatic ductal adenocarcinoma cells, and increased expression of HSPA5 represses ferroptosis induction by inhibition of GPX4 protein degradation (59). Therefore, enzalutamide might synergize with erastin and RSL3 by decreasing the expression of HSPs, negative regulator of ferroptosis. Another plausible mechanism of the increased efficacy of the combination of FINs with antiandrogens is that they act through separate programmed cell death pathways, namely apoptosis ferroptosis. Indeed, enzalutamide has been shown to induce apoptosis via increased expression of BAX, and decreased Bcl-2 expression (57). Further studies need to be conducted to delineate the precise mechanisms underlying the efficacy of FINs and antiandrogens combination.

In summary, our study demonstrates that erastin and RSL3 decrease the viability, growth, migration, and invasion of multiple prostate cancer cells in vitro. Furthermore, erastin and RSL3 significantly delay tumor growth of adeno-CRPC, NEPC, and DNPC xenografts in vivo with no measurable side effects. The combination of erastin or RSL3 with second-generation antiandrogens inhibited prostate cancer cell growth and migration in vitro and halted tumor growth of adeno-CRPC xenografts when compared with erastin, RSL3, or second-generation antiandrogens alone in vivo. Overall, our finding suggests that ferroptosis induction may represent a promising therapeutic strategy across prostate cancer variants either as a monotherapy or in combination with standard-of-care second-generation antiandrogens used for treatment of CRPC.

No disclosures were reported.

Opinions, interpretation, conclusions, and recommendations are those of the authors and not necessarily endorsed by the funding agencies.

A. Ghoochani: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, validation, investigation, methodology, writing–original draft. E.-C. Hsu: Conceptualization, data curation, formal analysis, validation, investigation, methodology, writing–original draft, writing–review and editing. M. Aslan: Data curation, formal analysis, validation, investigation, writing–review and editing. M.A. Rice: Validation, investigation, writing–review and editing. H.M. Nguyen: Resources, data curation, formal analysis, validation, investigation, visualization, writing–review and editing. J.D. Brooks: Resources, data curation, formal analysis, supervision, investigation, visualization, writing–original draft. E. Corey: Resources, data curation, formal analysis, supervision, investigation, methodology, writing–original draft, writing–review and editing. R. Paulmurugan: Resources, data curation, formal analysis, supervision, validation, methodology, writing–original draft, writing–review and editing. T. Stoyanova: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, validation, investigation, methodology, writing–original draft.

T. Stoyanova was supported by the Canary Foundation, the NIH/NCI R03CA230819, R37CA240822, and R01CA244281. The establishment and characterization of the LuCaP PDX models have been supported by the PNW Prostate Cancer SPORE P50CA097186 and P01CA163227. J.D. Brooks was supported by NIH CA196387. A. Ghoochani was supported by the U.S. Army Medical Research Acquisition Activity, through the CDMRP award no. W81XWH-19-1-0333. The work was also supported by NIH S10 OD023518-01A1 for the Celigo S Imaging Cytometer (200- BFFL-S).

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.