Synergistic Inhibition of Prostate Cancer Progression in Mice With a Combination of Curcumin and Ursolic Acid in the Diet

ABSTRACT

Prostate cancer (PCa) is the second leading cause of cancer‐related death among American men, and its long latency offers a window for chemopreventive strategies. Phytochemicals, with their diverse impacts on cancer cell growth and metabolism, represent promising candidates for such strategies. Combining compounds like curcumin (Curc) and ursolic acid (UA), which target multiple pathways, can be advantageous in slowing tumor progression. Previous studies revealed the synergistic effects of Curc + UA in reducing tumor growth in a PCa allograft model. In this study, diet‐based interventions were evaluated using two transgenic mouse models of PCa. Mice fed a Curc + UA‐enriched diet exhibited significant inhibition of prostate tumor progression compared to single‐agent diets in both HiMyc and PTEN knockout mouse models. Protein analyses of ventral prostate tissues from HiMyc mice indicated that the combination suppressed oncogenic signaling pathways, including STAT3, AKT, and mTORC1, while modulating cell regulatory proteins to inhibit tumor cell proliferation. Furthert mechanistic studies in mouse and human PCa cell lines confirmed that Curc + UA exerted pleiotropic effects by influencing oncogenic signaling, cell cycle regulation, mitochondrial function, unfolded protein response (UPR), and apoptosis, collectively contributing to its synergistic efficacy. These findings highlight the potential of Curc + UA to inhibit PCa progression through multitargeted mechanisms. The combination's superior efficacy over single agents underscores its promise as a chemopreventive or therapeutic strategy. This study provides a strong rationale for further mechanistic investigations and clinical development of Curc + UA for PCa prevention and treatment.

Abbreviations

CR

calorie restriction

Curc

curcumin

DIO

diet‐induced obesity

EMT

epithelial to mesenchymal transition

H&E

hematoxylin and eosin

IF

immunofluorescence

KO

knockout

LIA

locally invasive adenocarcinomas

PCa

prostate cancer

PERKi

PERK inhibitor

PIN

prostatic intraepithelial neoplasia

RB

retinoblastoma protein

Tg

thapsigargin

UA

ursolic acid

UPR

unfolded protein response

VP

ventral prostate

1. Introduction

PCa is the most common non‐cutaneous cancer type among American men, with approximately 250,000 new diagnoses each year [1]. Approximately 1 in 8 men will receive a PCa diagnosis in their lifetime, and PCa is the second leading cause of cancer‐related deaths in American men [1]. For men diagnosed with localized or regional PCa, the 5‐year survival rate is approximately 99%, however, the survival rate is much lower for men diagnosed when the cancer has metastasized to the lungs, liver or bones (~34%) [2]. Racial disparities also exist, for example, African men have a higher incidence PCa (1 in 6), and the disease is often more aggressive in African American men than Caucasian men [3, 4]. PCa develops from the precursor condition high‐grade prostatic intraepithelial neoplasia (referred to as PIN), through in situ PCa, invasive PCa and metastatic cancer [5]. It is estimated that approximately half of all men have developed PIN by the age of 50, however, there can be a long latency period between PIN and development of clinically evident cancer [5].

The goal of chemoprevention is to prevent, delay, or suppress the development of cancer. The average age at the time of diagnosis is 66, and because PCa can be relatively slow growing, it presents an opportunity to implement preventive strategies to better manage PCa, especially in high risk individuals. Phytochemicals have pleiotropic effects, and can favorably modulate several signaling pathways involved in cancer development and progression with smaller side effect profiles and relatively lower toxicity compared to most treatments to prevent disease. Using combinations of compounds that target multiple receptors and pathways to slow tumor growth and progression represents an advantageous approach to chemoprevention.

The combination of Curc + UA was originally identified through a two‐tiered screening method involving depletion of cellular ATP levels followed by inhibition of glutamine uptake [6]. Previously published studies from our laboratory showed that Curc + UA synergistically decreased cell survival in PCa cell lines, including HMVP2 (mouse) and several human PCa cell lines [7]. Furthermore, this combination synergistically induced apoptosis in HMVP2 cells [7]. Preliminary mechanistic studies performed in HMVP2 PCa cells showed that the combination reduced glutamine uptake and activated AMPK while reducing activation of Src and Stat3 (via reduced phosphorylation). In this previously published study, the dietary administration of a combination of Curc + UA synergistically inhibited growth of HMVP2 allograft tumors compared to either agent alone at the same doses.

The overall goal of the current study was to evaluate the effects of this combination on development and progression of PCa in autochthonous mouse models and provide insights into the mechanism of synergism and potential calorie restriction (CR) mimetic activity. Two relevant transgenic models of PCa were used to evaluate the efficacy of the combination in preventing PCa development and progression. Mouse and human PCa cell lines were also used to investigate the mechanism behind the observed synergy. The current results demonstrate that Curc + UA given in the diet had greater inhibitory effects on PCa progression in both HiMyc and PTEN knockout (KO) mice compared to the single agents given alone in the diet. Significant effects of the combination on critical signaling pathways, cell cycle progression, UPR activation and apoptosis all likely contributed to the synergistic effects on prostate tumor progression. The current data suggest that combining Curc + UA may be a valuable therapeutic option for prevention and/or treatment of PCa.

2. Materials and Methods

2.1. Transgenic Animals

All studies were approved and performed according to the guidelines of the Institutional Animal Care and Use Committees of UT Austin. C57BL.6 J (Strain 000664), Tg(Pbsn‐cre)4Prb/J (Strain 026662), and PTEN^Flox/Flox (Strain 006440) were purchased from Jackson Laboratories to be used as breeders. Generation and characterization of PTEN knockout mice have been described previously [8]. HiMyc mice were obtained from the NIH MMRRC on a FVB/N genetic background. The generation and characterization of HiMyc mice have been described previously [9]. All study mice were bred in house and genotypes were confirmed by PCR using tail DNA.

2.2. Diet Formulation

Diets for the HiMyc study were purchased from Research Diets Inc (New Brunswick, NJ). Diets for the PTEN knockout studies were purchased from Dyets Inc. (Bethlehem, PA). AIN‐93M diets were supplemented with Curcumin C3 complex (10 g/kg), ursolic acid (2 g/kg), or the combination.

2.3. Chemicals

For diet formulation, Curcumin C3 Complex was purchased from Sabinsa Corporation (East Windsor, NJ) and Loquat Leaf extract (CAS #77‐52‐1, 98% ursolic acid) was purchased from Stanford Chemicals (Lake Forest, CA). For in vitro experiments, curcumin (Cat. #C7727, > 95% purity) was purchased from Sigma‐Aldrich (St. Louis, MO) and Loquat Leaf extract (CAS #77‐52‐1, 98% ursolic acid) was purchased from Stanford Chemicals (Lake Forest, CA). A free sample of GSK2606414 (CAS #1337531‐36‐8), a cell permeable PERK inhibitor, was received from MedChemExpress (Princeton, NJ). Thapsigargin (CAS #67526‐95‐8), an inhibitor of the sarco‐endoplasmic reticulum Ca²⁺‐ATPase, was purchased from Sigma (Burlington, MA).

2.4. HiMyc Diet Study

Cohorts of HiMyc mice (5‐6 weeks old) were placed on diet and fed ad libitum for 6 months. Food consumption and body weight were monitored weekly. Mice were killed and genitourinary tract tissues were isolated from 20 to 24 mice per group and weighed before fixing in formalin. Hematoxylin and Eosin (H&E) stained tissue was used for histologic diagnoses. Individual lobes were isolated from an additional 5 mice in each group, which were snap frozen for extraction of protein and RNA. Ventral prostate tissue was used for western blot.

2.5. PTEN Knockout Diet Study

Cohorts of 5–7 week old PTEN knockout mice were placed on diet and fed ad libitum for 12 weeks. Food consumption and body weight were monitored weekly. Mice were killed and genitourinary tract tissues were isolated from 28 to 29 mice per group and weight before fixing in formalin. H&E stained tissue was used for histologic diagnoses.

2.6. Histological and Immunofluorescent Analyses

Male genitourinary tracts were removed from mice at the conclusion of the studies. Tissues were fixed in 10% formalin, embedded in paraffin and transversely sectioned. 4 µm sections were stained with H&E for histopathologic diagnoses. All histopathologic diagnoses of prostate lesions were based on published criteria [10, 11, 12]. Immunofluorescent staining of genitourinary tract tissue isolated from HiMyc mice on diet were also performed to evaluate expression of Ki67, PCNA, p‐Stat3^Y705, p‐S6R^S240/244, p‐RB^S807/811, CDK2, p16, and E‐cadherin.

2.7. Western Blot Analyses

Protein was collected from various experiments and isolated using RIPA buffer, supplemented with protease inhibitor phosphatase inhibitor (Cat. #25765800 & Cat. #P5726, Sigma‐Aldrich, St. Louis, MO) for 30 min. Protein samples from each diet group in the HiMyc study were pooled for analysis (n = 5). Proteins (40 µg) were separated by SDS‐PAGE and transferred to nitrocellulose membranes, blocked with 5% BSA for 1 h at room temperature, then incubated with primary antibody overnight at 4°C. The following primary antibodies were used: p‐AKT^T308, AKT, p‐mTOR²⁴⁸¹, mTOR, p‐S6R^S233/234, p‐S6R^S240/244, S6R, PTEN, p‐Stat3^Y705, Stat3, p27, CDK2, Cyclin E, p‐RB^S807/811, RB, p‐cdc2^T161, cdc2, Cyclin D1, p‐Src^Y416, Src, ATF4, p‐eif2a^S51, E‐Cadherin, cleaved caspase 7, Survivin, CHOP, BCL2, and MCL‐1 (Cell Signaling Technologies, Danvers, MA), Bim, Cyclin A and SOX4 (Invitrogen, Carlsbad, CA), BCL2 (Proteintech, Rosemont, IL), TWIST1 (Genetex, Irvine, CA), CDK4 (Santa Cruz Biotechnology, Dallas, TX), and MTA1 (Abcam, Cambridge, UK). Vinculin, β‐actin and GAPDH (Cell Signaling Technologies, Danvers, MA) were used as endogenous controls.

2.8. Cell Culture

HMVP2 [13] and LNCaP cells were cultured in RPMI1640 (Life Technologies) medium supplemented with 10% FBS (Life Technologies), maintained in 95% air and 5% CO₂ at 37°C. Cell lines were regularly tested for mycoplasma contamination by PCR Mycoplasma Detection Kit (Applied Biological Materials Inc.).

2.9. Apoptosis Assay

The percentage of apoptotic cells was determined using a FITC Annexin V Apoptosis Detection Kit I (Cat. #556547; BD Biosciences, Franklin Lakes, NJ). HMVP2 and LNCaP cells were treated for 24 h with the indicated treatments before collection. Annexin V‐positive cells were measured by flow cytometry according to the manufacturer's instructions. Data was acquired using a BD Accuri C6.

2.10. Statistical Analysis

Fisher's exact test was used to analyze the tumor incidence in the HiMyc and PTEN knockout diet studies. One‐way ANOVA with Tukey's multiple comparisons test was performed for statistical analyses of western blots. A paired t test was performed for analysis of PERK inhibitor experiments. Significance in all cases was set at p ≤ 0.05.

3. Results

3.1. Effect of a Combination of Curc + UA in the Diet on Formation of Adenocarcinomas in HiMyc Mice

Cohorts of 5–6 week old HiMyc mice were placed on diets containing the phytochemicals (control, 1.0% Curc, 0.2% UA or the combination) and were maintained on the diets. At 6 months, mice were killed and genitourinary tracts were collected for histopathologic diagnosis of the ventral prostate (VP). Notably, mice in the combination diet group showed a statistically significant reduction of in situ adenocarcinoma and locally invasive adenocarcinomas (LIA) compared to mice fed diets containing the single agents alone as well as the control diet (Figure 1A). Additionally, the genitourinary tract weights of mice on the combination diet were significantly reduced compared to all other groups at the end of the study (Figure 1B). There were no statistically significant differences in food consumption or body weight between different groups over the course of the experiment (Figure 1C,D, respectively), indicating that the compounds supplemented in the diet were well tolerated. Representative H&E stained sections of the VP tissue isolated from HiMyc mice on the respective diets are shown in Figure 1E. The percent area of VP that was impacted by adenocarcinoma and LIA was also determined (Figure 1F), and mice on the combination diet had the lowest overall tumor burden. Taken together, these data demonstrate that a diet containing a combination of Curc + UA significantly inhibited PCa progression in HiMyc mice compared to mice consuming diets containing the individual compounds as well as the control diet group.

Figure 1.

Effect of phytochemical treatment on prostate cancer (PCa) progression in HiMyc mice.5–6 week old HiMyc mice were placed in respective diet groups and were fed ad libitum for 6 months. Genitourinary tracts were collected and weighed before formalin fixation, and were used to perform histopathologic diagnoses. N = 20 for control and Curc groups, n = 24 for UA group, n = 23 for combination group. (A) Incidence of each stage of PCa progression from histopathologic diagnoses. (B) Genitourinary tract weights. (C) Average food consumption per day. (D) Average body weight per mouse. (E) Representative H&E stained ventral prostate tissue of HiMyc mice on diet. (F) Percent area of in situ adenocarcinoma and locally invasive adenocarcinoma occurrence of ventral prostate tissue of HiMyc mice. Statistical analyses for tumor incidence was performed with Fisher's Exact Test and statistical analyses for genitourinary tract weight and percent area of tumor were performed with one‐way ANOVA; *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.005.

3.2. Impact of Dietary Curc + UA on Established Oncogenic Signaling Pathways in HiMyc Mice

To explore potential mechanism(s) behind the observed synergistic reduction in tumor progression with the combination diet, levels and phosphorylation status of various proteins known to be involved in oncogenic signaling in PCa were evaluated in pooled samples of VP tumors from HiMyc mice on the various diets. As shown in Figure 2A, the combination diet reduced phosphorylation of AKT at the Tyr308 site. Additionally, there was a reduction in activation of mTORC1, shown both by a modest reduction in phosphorylation of mTOR at Ser2481 and greatly reduced phosphorylation of the downstream target, S6R, at both phosphorylation sites (Ser233/234 and Ser240/244). PTEN protein levels were increased across all phytochemical diet groups, but especially the combination diet group. Phosphorylation of Stat3 at Tyr705, was also dramatically decreased in the combination diet group compared to all other groups (Figure 2A).

Figure 2.

Effect of phytochemical treatment on oncogenic signaling pathways, cell cycle regulatory proteins and markers of EMT in HiMyc mice. Groups of 5–6 week old HiMyc mice were placed in respective diet groups and were fed ad libitum for 6 months. Ventral prostate lobes of five mice in each group were isolated for protein extraction. Protein was pooled for Western blot analyses and all quantitation was normalized to a loading control. (A) Representative Western blots of oncogenic signaling pathways with quantitation; (B) Representative Western blots of cell cycle regulatory proteins with quantitation. (C) Representative Western blots of markers of EMT with quantitation. (D) Representative Western blots of stress‐related proteins with quantitation.

Supplemental Figure 1A shows representative IF stained sections of the VP for cell proliferation makers (PCNA and Ki67) and for both p‐Stat3 (Tyr705) and p‐S6R (Ser240/244). As shown, the combination diet group had a significant reduction in both Ki67 and PCNA staining. In addition, decreased levels of p‐Stat3 Tyr705 staining as well as p‐S6R Ser240/244 staining were observed in VP of the combination group compared to the single agent treatment groups. These latter data corroborate the western blot results shown in Figure 2A and indicate that the synergistic effects of the combination of Curc + UA on tumor progression in vivo involved reductions of AKT, Stat3, and mTORC1 signaling as well as reduced tumor cell proliferation. The data in Supplemental Figure 1A suggested that cellular proliferation in the VP was reduced by the combination diet to a greater extent than the other diet groups. As shown in Figure 2B, we observed a greater reduction in cyclins E and A and a greater increase in p27 in the combination diet group compared to all of the other diet groups. Protein levels of CDK4 and CDK2 were also reduce to a greater extent in the combination diet group compared to all other diet groups. There was also a greater reduction in cdc2 phosphorylation at the Thr161 site and RB phosphorylation at Ser807/811 in the combination diet group compared to the other diet groups (again see Figure 2B). Supplemental Figure 1B presents further IF stained sections of VP showing a decrease in p‐RB (Ser807/811) and CDK2 protein staining as well as increased p16 staining consistent with the changes observed in the Western blot data. Taken together, these data provide further evidence that the combination diet led to greater inhibition of tumor cell proliferation compared to the other diet groups.

3.3. Effect of Combining Curc + UA on Markers of EMT in HiMyc Mice

As shown in Figure 1A, the primary effect of the combination diet was on the progression of PIN lesions to adenocarcinoma. To further explore how the combination diet was able to synergistically reduce tumor progression in HiMyc mice, we evaluated levels of EMT markers in protein lysates from VP tumors. Increased E‐cadherin protein levels were observed in VP tumors with the greatest increase seen in the combination diet group (Figure 2C). Furthermore, the combination diet dramatically reduced the levels of TWIST1 protein compared to all other diet groups. Additional EMT related protein changes were also observed to a greater extent in VP tumors from mice on the combination diet, including reduced levels of SOX4 and reduced levels of metastasis‐associated protein (MTA1). Collectively, these data suggest that the combination of Curc + UA was able to reduce PCa progression in vivo, at least in part, by also altering levels of EMT related changes.

3.4. Dietary Administration of Curc + UA Increased ATF4 Protein Levels and Modulated Markers of Mitochondrial Function in VP Tumors of HiMyc Mice

As shown in Figure 2D, p‐eif2a^S51 was modestly increased in VP tissue to a similar extent in all phytochemical diet groups compared to control diet. In contrast, ATF4 protein level was increased to a greater extent in the combination diet group compared to all other diet groups. The proapoptotic protein Bim, was also increased in VP tumors to a greater extent in the combination diet group compared to all other groups (again see Figure 2D), whereas, the antiapoptotic protein Survivin was decreased to a greater extent. Taken together, these changes may have also contributed to the synergistic effects of the combination diet on tumor progression.

3.5. Effect of Curc + UA on Oncogenic Signaling and Cell Cycle Pathways in Vitro

To further confirm the observed changes and mechanism(s) behind the synergistic effects of the combination diet on PCa progression, experiments were performed using cultured PCa cells. HMVP2 (mouse) PCa cells were treated with the phytochemicals for 6 h and protein was collected for Western blot analyses (Figure 3). Quantitation and statistical analyses of triplicate experiments are provided in Supplemental Figure 2. As shown in Figure 3A, there was a significant reduction in phosphorylation of S6R at the Ser240/244 site following treatment with the combination compared to all other groups. In addition, there was a significant reduction in phosphorylation of Src (Y416) and a dramatic reduction in phosphorylation of Stat3 at the Tyr705 site (Figure 3A), although the latter decrease was not statistically different than that observed in the Curc treated goup. There was a statistically significant reduction in total Stat3 protein levels following 6 h treatment with the combination compared to the other treatment groups. MTA1 was also significantly reduced by the combination compared to the other treatment groups.

Figure 3.

Effect of phytochemical treatment on oncogenic signaling pathways, cell cycle regulatory proteins and markers of mitochondrial function in HMVP2 cells. HMVP2 cells were treated with phytochemicals for 6 h before being harvested for protein. Experiments were performed in triplicate and representative western blots are shown. All quantitation was normalized to a loading control. Phosphoproteins were also normalized to their respective total protein levels. Statistical analysis can be found in Supplemental Figure 2. (A) Representative Western blots of oncogenic signaling pathways with quantitation. (B) Representative Western blots of cell cycle regulatory proteins with quantitation. (C and D) Representative Western blots of stress‐related proteins with quantitation.

Figure 3B shows representative Western blots for cell cycle proteins from HMVP2 cells treated for 6 h. As shown, the combination treatment reduced the levels of p‐RB, Cyclin D1, CDK2 and CDK4 to a greater extent than any other treatment group. In addition, as shown in Figure 3C,D the antiapoptotic proteins MCL‐1 and BCL2 were reduced and stress related proteins CHOP and ATF4 were elevated to a greater extent by the combination. Several other proteins were also analyzed in HMVP2 cells treated with the compounds where the combination reduced their levels to a greater extent than the individual compounds, including CDK6, Cdc6, p‐Cdc2^T161, and BCL2 but the observed difference was not statistically different than one or more of the other treatment groups (data not shown). Collectively, the observed changes in HMVP2 cells treated with the combination were similar to those observed in the VP tumors from HiMyc mice given the combination diet (Figure 2 and Supplemental Figure 1) further supporting the hypothesis that the combination diet works, at least in part, by altering to a greater extent signaling pathways involved in cell cycle control, survival/stress related pathways and EMT.

Additional Western blot experiments were performed on protein lysates from LNCaP cells prepared following 6 h treatment with the phytochemicals (Supplemental Figures 3 and 4). Overall similar effects on oncogenic signaling, cell cycle, survival/stress and EMT proteins were obtained with the combination in this human PCa cell line.

Figure 4.

Effect of PERK inhibition on ATF4 levels in HMVP2 cells. HMVP2 cells were pretreated with PERKi for 1 h before treatment. Phytochemical treatments were applied for 2 h before protein harvest. Experiments were performed in triplicate. (A) Representative Western blots. (B) Quantitation of three independent experiments. Statistical analyses were performed using a paired t test. *p ≤ 0.05, **p ≤ 0.01. (C) LNCaP cells were treated for 24 h and representative blots are shown. (D) Average quantitation of three experiments. (E) Annexin V results for LNCaP cells following 24 h of treatment. (F) Annexin V results for HMVP2 cells following 24 h of treatment. Statistical analyses were performed using a one‐way ANOVA followed by Tukey's multiple comparisons test. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.005, ****p ≤ 0.0001.

3.6. Effect of Curc + UA on Unfolded Protein Response (UPR) Activation, Markers of Mitochondrial Function and Apoptosis In Vitro

To further explore the effect of the combination of Curc + UA on ATF4 levels that were observed in VP tumors from HiMyc mice (Figure 2D) and in HMVP2 cells in culture (Figure 3D), additional experiments were performed. As shown in Figure 3C,D, the increased level of ATF4 in HMVP2 cells was associated with increased levels of CHOP at 6 h, suggesting possible ER stress‐induced apoptosis. Following treatment with the combination, we also observed reduced protein levels of apoptosis proteins in both HMVP2 and LNCaP cell lines, including MCL‐1, which was significantly reduced compared to all other groups (Figure 3C and Supplemental Figure 3C). Additionally, the level of BCL2 was reduced significantly in HMVP2 cells (Figure 3C). Other proteins including BCL‐xL and PUMA showed changes (reduced and increased, respectively) by the combination treatment but the differences were not significant compared to one or more of the single agent treatment groups (data not shown).

To investigate whether the increase in expression of ATF4 caused by the combination treatment was through activation of the PERK arm of the UPR, we treated HMVP2 cells with the PERK inhibitor GSK2616414 (referred to as PERKi). Cells were pretreated for 1 h with 100 nM PERKi or vehicle (DMSO), and media was then replaced with media containing PERKi and the respective phytochemical treatments. Thapsigargin (Tg) a potent UPR inducer [14, 15] was used as a positive control for these experiments. As shown, PERK inhibition led to a reduction in ATF4 protein level caused by Tg and the phytochemical treatments (Figure 4A). The PERK inhibitor significantly reduced both ATF4 by both Curc alone and the combination and significantly reduced p‐eif2a protein levels by the combination (Figure 4B). These experiments support the hypothesis that the combination induced ATF4 protein levels, at least in part, through activation of the PERK arm of the UPR.

ATF4 expression followed by CHOP activation and reduced expression of antiapoptotic proteins (Figure 3C,D) also suggested that the combination treatment induced apoptosis in the PCa cells. Further analyses revealed that following 24 h treatment with the combination, LNCaP cells exhibited a marked increase in caspase 7 cleavage that was statistically significant compared to all other treatment groups (Figure 4C,D). Similarly, single agent treatments caused moderate reductions in levels of Survivin, while the level of this protein was significantly reduced by treatment with the combination. Reductions in antiapoptotic proteins and subsequent increases in cleavage of caspase 7 strongly suggested that the combination of Curc + UA induced apoptosis in both mouse and human PCa cells. To directly assess the populations of apoptotic cells in phytochemical treated cells, we treated cells for 24 h and measured the percentage of apoptotic cells by Annexin V staining. As shown, the percentage of apoptotic cells was highest in cells treated with the combination (Figure 4E,F). These data demonstrate that the combination of Curc + UA induced apoptosis to a greater extent than single agent treatments in vitro, consistent with diet effects (Figure 2D) observed in the tumors from HiMyc mice.

3.7. Dietary Administration of Curc + UA Reduced Pca Progression in Prostate Specific PTEN KO Mice

The effect of combining Curc + UA in the diet was also assessed in a second PCa mouse model harboring a different mutation, in this case PTEN deletion. Groups of 5–7 week old male, prostate specific PTEN KO mice (Pb. Cre x PTEN^flox/flox mice) were placed on diets (control, 1.0% Curc, 0.2% UA or the combination). After 12 weeks on diet, mice were killed and genitourinary tract tissues were collected. Genitourinary tracts isolated from prostate specific PTEN KO mice on the combination diet had significantly reduced weight compared to all other diet groups (Figure 5A). Food consumption and body weight were monitored weekly, and there were no significant differences between diet groups (Figure 5B,C). To determine whether the combination reduced progression of PCa in these mice, histopathologic diagnoses were performed on H&E stained tissue of both VP and DLP. As shown in Figure 5D, the combination diet caused significant reductions in LIA in both the ventral and dorsolateral prostate lobes compared to all other diet groups. As expected, 100% of control mice had evidence of LIA in both the ventral and dorsolateral prostate lobes. Approximately 88% and 82% of Curc mice showed evidence of LIA in ventral prostate and dorsolateral prostate, respectively. Mice receiving UA only in the diet had similar incidence, with 88% of mice showing LIA in both the ventral prostate and dorsolateral lobes. By contrast, only 24% of combination diet mice had evidence of LIA in the ventral prostate, and 18% had LIA in the dorsolateral prostate. Representative H&E stained sections of the ventral prostate tissue isolated from these prostate specific PTEN KO mice on the respective diets are shown in Figure 5E.

Figure 5.

Effect of phytochemicals supplemented in the diet on prostate specific PTEN KO mice. Groups of 5–7 week old prostate specific PTEN KO mice were placed in respective diet groups (1% Curc, 0.2% UA or 1% Curc + 0.2% UA) and were fed ad libitum for 12 weeks. (A) Genitourinary tracts were collected and weighed before formalin fixation, and were used to perform histopathologic diagnoses. N = 28 for control, Curc and UA groups; n = 29 for the combination group. Statistical analyses for genitourinary tract weight were performed with one‐way ANOVA; *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.005. (B) Average food consumption per day. (C) Average body weight per mouse. (D) Percent incidence of locally invasive adenocarcinoma (LIA) from histopathologic diagnoses. Statistical analyses for tumor incidence were performed using Fisher's Exact Test; **p ≤ 0.01, ***p ≤ 0.005, ****p ≤ 0.0001. (E) Representative H&E stained sections of the ventral prostate tissue isolated from PTEN KO mice on diet.

4. Discussion

Overall, the combination of Curc + UA significantly inhibited PCa tumor progression in both HiMyc mice and prostate specific PTEN KO mice to a greater extent than either compound alone. Curc + UA administered in the diet inhibited progression from hgPIN to in situ adenocarcinoma and LIA in HiMyc mice and progression to LIA in prostate specific PTEN KO mice to a greater extent than the individual compounds. Mechanistic studies in tumor tissue from HiMyc mice and in PCa cell lines revealed that the combination produced synergistic effects on PCa progression via alterations in oncogenic signaling, cell cycle proteins and cell survival pathways. In addition, the combination significantly reduced markers associated with EMT and increased E‐cadherin levels in tumors consistent with reduced progression of PCa. Collectively, the current results demonstrate that the previously reported screening method [6] can identify phytochemical combinations such as Curc + UA used in the current study to effectively prevent PCa progression in two different autochthonous mouse models of PCa.

Detailed mechanistic studies of both tumors and PCa cell lines revealed that the combination inhibited several oncogenic signaling pathways involved in PCa progression and significantly reduced (or increased in the case of p27) levels of key cell cycle regulatory proteins. Tumors from HiMyc mice fed the combination diet showed a reduction in Stat3 phosphorylation at Tyr705, indicating reduced Stat3 activation. Stat3 activation regulates a number of important pathways involved in tumor progression including cell cycle proteins involved in proliferation of PCa cells and antiapoptotic proteins through transcriptional regulation of these genes [16, 17, 18]. Similar results were observed in cultured PCa cells (both mouse and human) treated with the combination. In addition, Stat3 also regulates genes involved in EMT [18, 19, 20]. For example, Twist1 is a known transcriptional target of Stat3 [21], and as shown in Figure 2C, the level of this EMT‐related transcription factor was markedly lower in tumor tissue from mice treated with the combination.

There is ample evidence in the literature demonstrating that inhibition of Stat3 disrupts tumor cell proliferation and promotes apoptosis [22, 23]. Stat3 activation promotes the transcription of Cyclin D1 [24, 25, 26], which drives progression of the cell cycle through the G1‐S‐phase. In vitro, we observed synergistic decreases in Cyclin D1 protein levels following 6 h treatment with the combination in both mouse and human PCa cell lines. Other cell cycle regulatory proteins were also dramatically decreased by the combination treatment, including CDK2, CDK4, and p‐RB^S807/811. Reduced levels of multiple of proteins that regulate transition of cells into G1‐S phase suggests that the combination halted the cell cycle and prevented excessive proliferation of tumor cells. These results were confirmed by IF staining of VP tumors from HiMyc mice.

Frequent activation of the PI3K/AKT/mTOR pathway can promote tumor growth and progression through signaling cascades that control cell survival, proliferation, metabolism and differentiation [27, 28, 29]. It is estimated that approximately 40% of primary prostate tumors and 70% of metastatic tumors exhibit alterations in PI3K signaling [30, 31]. Under normal conditions, mTOR regulates cell growth and division whereas abnormal activation of mTOR helps to maintain the growth, survival and proliferation of tumor cells [32, 33]. Previous studies have reported effects of the individual compounds on mTORC1 signaling, and here we show that the combination of Curc + UA inhibited AKT phosphorylation and mTORC1 downstream signaling to a greater extent than the individual compounds. Simultaneous reductions in AKT, mTORC1 and Stat3 signaling as a result of Curc + UA treatment suggests that this combination is able to combat increased cell growth and resistance to cell death mechanisms induced by aberrant activation of these pathways, and may be a useful against a range of genetic alterations that are often found in PCa and other cancers.

Further mechanistic studies performed in the mouse and human PCa cell lines suggested that the combination of Curc + UA also significantly modulated mitochondrial function, UPR activation, and apoptosis pathways that likely contributed to its synergistic effects on PCa progression in vivo. The UPR is a double‐edged sword because it can both promote cell survival and induce cell death depending on the context [34, 35]. Under normal conditions, UPR activation is a mechanism for cells to overcome ER stress and maintain homeostasis, but when the ER stress cannot be resolved, the UPR signals to induce cell death. We found that treatment with the combination induced a significant increase in ATF4 protein levels in VP tumors and in both mouse and human PCa cells compared to treatment with the single agents. PERK inhibition partially attenuated the increase in ATF4 protein levels caused by treatment of cells with the combination, suggesting that the combination induced ATF4 protein levels, at least in part, through activation of the PERK arm of the UPR. The observed increase in ATF4 caused by treatment with the combination was followed by a marked increase in the levels of CHOP. CHOP has been shown to suppress BCL2 [36], and CHOP protein expression is associated with cell death [37, 38, 39].

Upon further analyses the combination of Curc + UA induced apoptosis to a greater extent than treatment with the single agents suggesting that this was due in part to activation of the UPR. In cultured PCa cells, we found that the combination reduced levels of several BCL‐2 family proteins, including BCL2, MCL‐1, and Survivin by 6 h after treatment of PCa cells (Figures 3C and 4C and Supplemental Figures 2, 3, 4). These changes led to increased apoptosis by 24 h with the combination to a greater extent than the individual compounds as evidenced by increased cleaved caspase 7. Increased apoptosis by the combination compared to the single agents was confirmed by Annexin V staining at the 24 h time point. Increased levels of Bim and reduced levels of Survivin (Figure 2D) in VP tissue from HiMyc mice on the combination diet suggested that the combination diet also reduced PCa cell survival in vivo.

An increase in PTEN protein level was observed in tumors from HiMyc mice given the combination compared to mice given the single agents (Figure 1E). While the mechanism for this increase is not known at present, this could explain the effect of the combination on Akt phosphorylation (reduced) suggesting that this effect may be another potential avenue through which the combination diet is able to synergistically inhibit prostate tumor progression at least in HiMyc mice. However, because we observed a significant, synergistic reduction in PCa progression in prostate specific PTEN KO mice with the combination of Curc + UA in the diet, this suggests that the increase in PTEN observed in HiMyc mice may not be essential for the overall chemopreventive efficacy of the combination.

One of the goals of the current research was to identify combinations of phytochemicals with potential CR mimetic activity. CR has been shown in numerous studies to inhibit tumor formation in animal models of cancer, including PCa [40, 41, 42]. It is interesting to compare the effects of the combination of Curc + UA on PCa progression in the current study with that of an earlier study from our laboratory that compared the impact of diet‐induced obesity (DIO) and CR on PCa development and progression in the HiMyc mouse model. Results from that study showed that relative to control diet, a DIO regimen (60 kcal% fat) enhanced progression of PCa while a 30% CR significantly delayed PCa progression similar to that observed in our current study with the combination of Curc + UA. Compared to the control diet group, DIO led to an increase in AKT, mTOR and Stat3 signaling in tumor tissue, while CR reduced signaling through these pathways, as evidenced by both IHC and Western blot analyses [40]. Furthermore, CR also reduced proliferation and cell cycle proteins (e.g., cyclin D1) similar to effects seen with the combination diet in the current study. Collectively, these observations suggest that dietary supplementation of Curc + UA modulated similar molecular signals that are associated with the anticancer effects of CR and suggest that the combination of Curc + UA fits the criteria of a CR mimetic [43].

In summary, the current data demonstrate that the combination of Curc + UA supplemented in the diet provided greater inhibition of PCa progression in two different mouse models of PCa when compared to either compound administered alone. Mechanistically, this effect was associated with reduced activation of oncogenic signaling pathways, reduced cell cycle progression and attenuated proliferation in addition to reversing expression of progression‐related markers of EMT. These results demonstrate the potential of the combination of Curc + UA as a therapeutic option for prevention and/or treatment of PCa. The fact that many of the pathways modulated by the combination of Curc + UA are altered in other cancers also suggests that this combination may display chemopreventive efficacy beyond just PCa.

Author Contributions

C.F., A.S., and J.D. designed the experiments. C.F., A.S., R.C., C.W., J.W. performed experiments and analyzed the data. C.F., A.S., and J.D. interpreted the data. C.F. and J.D. wrote the article. All authors reviewed the article and approved its content.

Ethics Statement

The in vivo experiments described in this study were conducted with the approval of the Institutional Animal Care and Use Committee (IACUC) at the University of Texas at Austin and in compliance with their established guidelines.

Consent

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflict of interest related to this study.

Supporting information

Supmat.

HiMyc Paper Figures_v121111.

Data Availability Statement

All data supporting the findings of this study are available within the article.