A Combination of Xanthohumol and Ursolic Acid in the Diet Leads to Synergistic Inhibition of Prostate Cancer Progression

ABSTRACT

Prostate cancer (PCa) is the second most common cancer and second leading cause of cancer death for American men. Chemoprevention by using phytochemicals offers a promising approach to improve outcomes due to their ability to act on cancer cell metabolism and growth while maintaining low toxicity profiles. The goal of this study was to assess the combination of xanthohumol (XAN) and ursolic acid (UA) given in the diet for synergistic efficacy against PCa progression and identify potential mechanisms of action. PCa cells were treated with the combination to evaluate cell survival and colony formation. Two mouse models of PCa were used to evaluate tolerability and efficacy of dietary administration of the combination and to further understand mechanism(s) of action. The combination of XAN + UA reduced PCa cell survival and colony formation. The combination given in the diet significantly and synergistically inhibited growth of HMVP2 PCa allograft tumors and also inhibited PCa progression in HiMyc mice. Mechanistically, inhibition of polyamine synthesis and epithelial‐to‐mesenchymal transition contributed to the inhibition of HMVP2 allograft tumor growth, while the inhibition of PCa progression in HiMyc mice was associated with activation of the unfolded protein response pathway and apoptosis. Further studies in cultured PCa cells revealed additional effects of the combination on several oncogenic signaling pathways (e.g, phospho‐STAT3) and cell cycle regulatory proteins (e.g, cyclin D1, phospho‐Rb).

1. Introduction

Prostate cancer (PCa) is the second most commonly diagnosed cancer in American men, with approximately 1 in 8 receiving a diagnosis in their lifetime [1]. PCa is also the second deadliest cancer for American men, with about 1 in 44 men dying due to the disease. In 2025, the American Cancer Society estimates over 300,000 new cases and 35,770 deaths due to PCa [1]. Men who are diagnosed with localized or regional disease have over 99% survival after 5 years, though this number drops severely for men who are diagnosed with distant PCa, with only 37% survival [2]. Following a diagnosis, there are numerous treatment options for patients based on their individual circumstances, including active surveillance, surgery, hormone therapy, and chemotherapy. However, these options can come with side effects that may diminish quality of life, such as urinary incontinence, impotence, and osteoporosis [1, 3, 4, 5]. Overall, there is room for improvement in how PCa is approached to further improve survival and reduce adverse side effects for patients.

An advantageous approach to improve PCa outcomes is through chemoprevention. According to the World Health Organization, between 30% and 50% of all cases of cancer are preventable [6]. PCa represents an ideal candidate for implementing chemoprevention strategies due to the timeline of development. High‐grade prostatic intraepithelial neoplasia (PIN) is a precursor condition to developing PCa later on [7], and studies have shown that there is a long latency period between PIN development and PCa development, of at least 5 years and possibly up to 10 years [8]. This latency period offers ample opportunity to implement chemoprevention strategies that can reduce the number of PIN cases that progress to clinically evident cancer. One such strategy is chemoprevention using phytochemicals, which are biologically active compounds derived from plants.

Phytochemicals have demonstrated broad mechanisms of action and an ability to interfere with multiple steps in the carcinogenesis process while maintaining relatively low toxicity profiles [9, 10, 11]. Combinations of phytochemicals are of interest due to the ability to target multiple pathways, while reducing the risk of developing chemoresistance and improving tolerability by being able to use lower doses of each compound. Utilizing a previously established screening process to identify phytochemical combinations that may act synergistically for cancer chemoprevention [12], the combination of xanthohumol (XAN) and ursolic acid (UA) was identified and chosen for further study. XAN is a prenylated chalcone that is derived from the female Humulus lupulus, or hops plant [13]. XAN has demonstrated the ability to induce apoptosis [14, 15, 16] and to inhibit oncogenic signaling pathways [15]. UA is a pentacyclic triterpenoid that is found in various fruits and herbs, including apple, cranberry, and thyme [17]. Similar to XAN, UA has been shown to be proapoptotic and inhibit oncogenic signaling pathways [18, 19, 20, 21]. Individually, both XAN and UA have demonstrated the ability to inhibit PCa tumor growth in mouse models [18, 22].

In the current study, the ability of the combination of XAN + UA to inhibit PCa growth and progression was evaluated. The results indicated that PCa cells were sensitive to XAN + UA combination treatment in a synergistic manner, which was associated with a significant inhibition of oncogenic signaling and activation of the UPR pathway. XAN + UA given in the diet provided greater inhibition of PCa growth and progression compared to either single agent in a mouse allograft tumor model and the HiMyc transgenic mouse model. Analysis of the allograft tumors indicated a unique metabolic profile in mice that were fed the combination diet and inhibition of oncogenic signaling. In HiMyc mice, the inhibition of PCa progression by the combination was associated most closely with induction of the UPR pathway and apoptosis. Together, these various mechanisms likely converge to the overall synergistic inhibition of PCa growth and progression that is observed with XAN + UA. The current data suggest that a combination of XAN + UA may represent a novel option for the chemoprevention and/or treatment of PCa in humans.

2. Materials and Methods

2.1. Chemicals

XAN (> 98% purity) was purchased from Adooq Bioscience and Loquat leaf extract (98% UA) was purchased from Stanford Chemicals. 4‐Phenylbutyric acid (99% purity) was purchased from Sigma‐Aldrich.

2.2. Diet Formulation

Diets for the allograft tumor study were purchased from Custom Animal Diets LLC (Bangor, PA). Diets for the HiMyc tumor study were purchased from Dyets, Inc (Bethlehem, PA). For both, AIN‐93M base diets were supplemented with XAN (2 g/kg), UA (2 g/kg), or the combination (XAN 2 g/kg + UA 2 g/kg).

2.3. Animals

Mouse husbandry was according to NIH guidelines in a facility that is accredited by the American Association for the Accreditation of Laboratory Animal Care (AAALAC). Experimental protocols were approved and performed according to the guidelines of the Institutional Animal Care and Use Committees (IACUC) of The University of Texas at Austin. For the allograft tumor study, male FVB/N mice were purchased from Charles River and allowed to acclimate for 1 week prior to study start. For the HiMyc tumor study, HiMyc mice on an FVB/N background were bred in‐house. The generation and characterization of HiMyc mice have been previously described [23]. Genotype was confirmed via PCR using DNA from tail snips.

2.4. Allograft Tumor Study

Six‐ to 7‐week‐old male FVB/N mice were injected subcutaneously with 5 × 106 HMVP2 cells into both flanks. Once tumors were palpable, mice were divided into four groups and fed ad libitum with control, XAN, UA, or the combination diet. Tumor size was measured at least once a week with digital calipers. Tumor volume was calculated using the formula 0.5236 D1(D2)², where D1 and D2 are the long and short diameter, respectively. The study was concluded once tumor size in the control group reached the maximum limit as specified by the IACUC. Tumors were removed and used in metabolomics analyses.

2.5. HiMyc Tumor Study

Four‐ to 6‐week‐old male HiMyc mice were placed on control, XAN, UA, or the combination diet until they were 6 months of age. Body weight and food consumption were monitored weekly. At study conclusion, genitourinary tracts were collected for histological analysis or individual prostate lobes were isolated and snap frozen for protein analysis.

2.6. Metabolomics Sample Preparation and Acquisition

Following allograft tumor collection and snap freezing, the tissues underwent metabolomic analysis sample preparation and acquisition procedures. HPLC grade methanol, water, chloroform, ammonium bicarbonate, and formic acid were purchased from Thermo Fisher Scientific.

For metabolite extraction, tissues were weighed and homogenized in ice‐cold 1:1 methanol:water containing 10 mM ammonium bicarbonate, followed by addition of equal parts chloroform containing butylated hydroxytoluene (BHT) to preserve oxidation‐sensitive metabolites. Samples were kept on ice throughout the extraction process. The polar fractions were collected and filtered using Nanosep Omega centrifugal devices with 3 kDa molecular weight cutoff (Pall Life Biosciences). A mixture of 14 deuterated internal standards was spiked into each sample prior to analysis.

Quality control (QC) samples were prepared by pooling small aliquots from each sample. Sample injection order was randomized, with one QC sample analyzed every six samples. Analytical blanks were run before and after the entire acquisition sequence to monitor background and carryover.

Metabolomic analysis was performed using a Vanquish UHPLC system coupled to a Q Exactive Hybrid Quadrupole‐Orbitrap mass spectrometer (Thermo Fisher Scientific). Chromatographic separation was achieved using a SeQuant ZIC‐HILIC column (2.1 × 150 mm, 3.5 μm particle size, Millipore‐Sigma) operated at 25°C. The mobile phase consisted of (A) 10 mM ammonium bicarbonate in water with 0.1% formic acid and (B) 100% methanol. A gradient elution was employed with the following program: (1) 0–2 min: 80% B; (2) 2–20 min: linear gradient from 80% to 20% B; (3) 20–22 min: hold at 20% B; (4) 22–23 min: return to 80% B; (5) 23–30 min: re‐equilibration at 80% B. The flow rate was maintained at 0.25 mL/min, and the injection volume was 5 μL. The samples were then dried in vacuum concentrator (Labconco) and resuspended in an equal volume of water before chromatographic separation was completed using a Kinetex C18 column (2.1 × 150 mm, 2.6 μm particle size, Phenomenex) operated at 25°C. The mobile phase consisted of (A) 0.2% formic acid in water and (B) 100% methanol. The gradient elution program was as follows: (1) 0–4 min: 2% B; (2) 4–14 min: linear gradient from 2% to 80% B; (3) 14–15 min: linear gradient from 80% to 98% B; (4) 15–21 min: hold at 98% B; (5) 21–21.5 min: linear gradient from 98% to 2% B, (6) 21.5–35 min: Hold at 2% B. The flow rate was 0.15 mL/min and the injection volume was 5 μL.

Mass spectrometric data were acquired in both positive and negative electrospray ionization modes with the following parameters: spray voltage, 3.5 kV; capillary temperature, 300°C; sheath gas flow rate, 45 arbitrary units; auxiliary gas flow rate, 15 arbitrary units; scan range, m/z 70–1000; resolution, 70,000; automatic gain control (AGC) target, 1e6; maximum injection time, 100 ms. Data were acquired in profile MS mode using Xcalibur software (Thermo Fisher Scientific).

2.7. Metabolomics Sample Analysis

Raw data files were processed using MS‐DIAL 4.0 (http://prime.psc.riken.jp/) and an in‐house MATLAB script for peak detection, alignment, and annotation. Metabolite identification was performed by matching accurate masses and retention times to an in‐house library of authentic standards, as previously described [24]. Peaks were included in the final analysis only if they met strict quality criteria: coefficient of variation (CV) less than 25% in the QC replicates and signal‐to‐noise ratio greater than 3. The pooled QC samples were used to monitor instrument stability throughout the analysis, and analytical blanks were used for background subtraction. For statistical analysis, the data matrix was normalized to tissue weight and internal standards, and missing values were imputed using k‐nearest neighbor algorithm. Statistical analyses were performed using R (version 4.3.0) and MetaboAnalyst 5.0. and 6.0.

2.8. Histological Analyses

Male genitourinary tracts were fixed in 10% formalin, then embedded and sectioned. Prior to embedding, tissues were bisected along the bladder and oriented perpendicular to the cassette to allow for visualization of all prostate lobes upon sectioning. Tissues were embedded in paraffin and sectioned into 4 µm sections. Sections were stained with hematoxylin and eosin (H&E) for visualization. All histopathologic diagnoses were based on published criteria [25, 26].

2.9. Cell Culture

Human prostate cancer cell lines (22Rv1, C4‐2B, LNCaP, PC3, DU145) and mouse prostate cancer cell line HMVP2 [27] were cultured in RPMI 1640 medium with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (Life Technologies). Cells were periodically confirmed to be mycoplasma negative using PCR amplification (Applied Biological Materials Inc.). Cells were incubated at 37°C in 95% air and 5% CO₂.

2.10. Cell Survival Assay

Cells were plated in 96‐well plates and allowed to attach for 24 h, then treated with the indicated compounds. After treatment, viable cells were fixed with 10% formalin then stained with 0.05% crystal violet. After staining, the cells were washed once with water, and the remaining dye was extracted with 10% acetic acid. Absorbance was measured at 595 nm.

2.11. Colony Formation Assay

Cells were plated in 12‐well plates at a low density and allowed to attach for 24 h, then treated with UA, XAN, or the combination. Treatment durations varied by cell line and were concluded based on the formation of well‐defined colonies in the vehicle treated wells. After treatment, cells were fixed with 10% formalin then stained with 0.05% crystal violet. After staining, the cells were washed twice with water and colonies counted visually.

2.12. Measurement of Spermine

The abundance of spermine was measured using a competitive ELISA kit (Biomatik). HMVP2 cell lysates were collected according to manufacturer's instructions following treatment with XAN and UA for 24 h, then assay performed.

2.13. Western Blot Analyses

For in vitro samples, cells were treated with UA, XAN, or the combination for the indicated timepoints. After treatment, cells were washed with PBS and lysed with RIPA buffer supplemented with phosphatase inhibitor cocktails 2 and 3, and protease inhibitor cocktail (Sigma). For in vivo samples, tissues were homogenized using a mortar and pestle in liquid nitrogen, then transferred to a microfuge tube containing lysis buffer. Protein quantification was performed by DC protein assay (Bio‐Rad). Proteins were separated by SDS‐PAGE and transferred to nitrocellulose membranes. Membranes were blocked with 5% BSA (in TBST) for 1 h at room temperature (RT), followed by primary antibody incubation overnight at 4°C. Membranes were then washed with 0.1% TBST, incubated in secondary antibody (Cytiva) in 5% milk (in TBST) for 1 h at RT, and washed. Visualization was performed with commercial chemiluminescent detection kits (SuperSignal West Pico PLUS, Thermo Fisher, or Clarity Western, Bio‐Rad, for strong targets, or WesternBright Sirius, Advansta, for weaker targets).

The following primary antibodies were used: ODC1 (ProteinTech, 28728‐1‐AP), e‐cadherin (Cell Signaling, 3195), Twist1 (GeneTex, GTX60776), p21 (Santa Cruz, SC‐6246), actin (Sigma, A5316), pSTAT3^Y705 (Cell Signaling, 9145), STAT3 (Cell Signaling, 9139), cyclin D1 (Cell Signaling, 2922), cdc6 (Cell Signaling, 3387), p‐Rb^S807/811 (Cell Signaling, 9308), Rb (Invitrogen, MA5‐11387), c‐MYC (Cell Signaling, 5605), ATF4 (Cell Signaling, 11815), CHOP (Cell Signaling, 2895), vinculin (Cell Signaling, 13901), GAPDH (Cell Signaling, 2118), PARP (Cell Signaling, 9542), cleaved caspase 3 (Cell Signaling, 9664).

2.14. Statistical Analyses

All statistical analyses were performed using GraphPad Prism 9 software. Data are presented as mean ± SEM. Statistical tests used are indicated in Figure Legends. Briefly, one‐way ANOVA with Tukey's multiple comparisons test was used where data was normally distributed; Kruskal‐Wallis with Dunn's multiple comparisons test was used where data from one or more group(s) was not normally distributed; Welch ANOVA with Dunnett's T3 multiple comparisons test was used where data was not normally distributed and groups had unequal standard deviations. For in vivo experiments, allograft tumor growth and body weights over time were analyzed by 2way ANOVA or mixed effects analysis with Tukey's multiple comparisons test. Tumor incidence was analyzed by 2‐sided Fisher's exact tests to compare groups. Statistical significance was defined as p ≤ 0.05. Synergy was determined by Bliss index according to the formula C = A + B – AB, where A and B are the individual affected fractions and C is the theoretical affected fraction of the combination [28]. When the experimental combined affected fraction is greater than the predicted one it is considered synergistic.

3. Results

3.1. A Combination of XAN + UA Synergistically Inhibited PCa Cell Survival and Colony Formation

The combination of XAN + UA was tested for its ability to inhibit PCa cell survival using crystal violet assays. The combination significantly inhibited the survival of 6 different PCa cell lines, including mouse (HMVP2, Figure 1A) and human [22Rv1, C4‐2B (Figure 1B,C) and LNCaP, PC3, and DU145 (data not shown)]. The combination concentrations tested that produced synergistic reductions in cell survival in the cell lines is shown by the Bliss Index plot in (Figure 1D). The combination of XAN + UA also caused a significant inhibition of colony formation, demonstrated by a significant reduction in colonies formed from both HMVP2 and 22Rv1 cell lines (Figure 1E,F). Taken together, these data indicate that PCa cells are sensitive to the combination of XAN + UA in terms of cell survival and proliferation in a significant and synergistic manner compared to the individual compounds.

Figure 1.

The combination of XAN + UA inhibited PCa cell survival and proliferation. Cell survival and colony formation measured by crystal violet staining in PCa cell lines. Survival of HMVP2 (A), 22Rv1 (B), and C4‐2B (C) cells with combinations of XAN + UA. The reduction in cell survival produced by the combination treatment was assessed for synergy using the Bliss independence model (D). #, $, Φ Significantly different from control, UA and XAN alone respectively. Representative images of colonies (E) and quantitation (F). Data represents mean ± Φ SEM (n = 3) and a one‐way ANOVA with Tukey's multiple comparisons test was used to compare groups. Significance is reported at ***p < 0.001, ****p < 0.0001.

3.2. Effect of XAN + UA on Oncogenic Signaling in PCa Cells

To understand the mechanism of action of the combination of XAN + UA on cell survival and proliferation, we examined the effects on STAT3 signaling, an important signaling pathway in cell survival and in cancer development and progression. In HMVP2 cells, the combination led to a greater reduction of STAT3 phosphorylation at Tyr705 compared to either single agent (Figure 2A). To assess for reduced STAT3 activity, known downstream targets of STAT3, cyclin D1 [29] and c‐MYC [30], were investigated. Cdc6 was also selected based on a report that signaling through STAT3‐cyclin D1 can impact Cdc6 levels [31]. Aligning with the reduction of STAT3 phosphorylation by the combination of XAN + UA, there was a significant decrease in protein levels of cyclin D1, Cdc6, and c‐MYC in the combination treatment group compared to all other groups (Figure 2A,B). Similar reductions were also observed in LNCaP cells (Supplementary Figure 1A,B). Notably, these are all proteins that play a role in the cell cycle and determining G1 to S phase progression [32, 33, 34], suggesting that the combination may act in part by causing cell cycle arrest. Therefore, other proteins involved in this G1‐S phase transition were also investigated. The combination of XAN + UA significantly reduced phosphorylation of Rb compared to all other groups, and increased protein levels of p21 (Figure 2A,C). Collectively, these data indicate significant and synergistic effects on oncogenic signaling and cell cycle proteins associated with PCa survival and proliferation.

Figure 2.

The combination of XAN + UA impacted oncogenic signaling in PCa cells. Levels of oncogenic signaling proteins related to cell cycle progression in HMVP2 cells treated with vehicle (DMSO), 10 µM UA, 7.5 µM XAN, or the combination for 6 and 24 h were assessed by western blot. Representative blots are shown along with quantitation. (A) and (B) were acquired by x‐ray film and (C) was acquired with a digital imaging machine. Data represents mean ± SEM (n = 3) and a one‐way ANOVA with Tukey's multiple comparisons test was used to compare groups. Significance is reported at *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

3.3. Effect of XAN + UA on Unfolded Protein Response (UPR) Signaling in PCa Cells

To further understand the mechanisms underlying the synergistic effects of the combination of XAN + UA on PCa survival and proliferation, we examined the effects on UPR signaling. Following treatment of HMVP2 cells with the phytochemicals, UPR signaling was activated, with an increase in ATF4 protein levels followed by an increase in CHOP protein levels (Figure 3A,B). Similar increases in ATF4 and CHOP protein levels were also observed in LNCaP cells with the XAN + UA combination (Supplementary Figure 1C). Since UPR signaling activation is caused by ER stress, we assessed cell survival with and without ER stress inhibition [using 4‐phenylbutyric acid (4‐PBA), a chemical chaperone that assists with protein folding at the ER]. Pretreatment of HMVP2 cells with 4‐PBA for 2 h prior to the addition of phytochemical treatments partially rescued the survival of cells receiving the combination of XAN + UA (Figure 3C). This rescue was also observed when tested in LNCaP cells (Supplementary Figure 1D). These data suggest that ER stress with UPR signaling activation may also contribute to the synergistic effects of the combination of XAN + UA on PCa cells.

Figure 3.

The combination of XAN + UA impacted unfolded protein response signaling in PCa cells. HMVP2 cells treated with vehicle (DMSO), 10 µM UA, 7.5 µM XAN, or the combination for 2 (A) and 6 (B) hours (n = 3). Representative blots are shown along with quantitation. (A) was acquired by digital imaging machine and (B) was acquired by x‐ray film. (C) Survival of HMVP2 cells after 24 h of treatment with XAN and UA with and without pretreatment with the ER stress inhibiting compound 4‐PBA for 2 h (n = 2–3). For all, data represents mean ± SEM and a one‐way ANOVA with Tukey's multiple comparisons test was used to compare groups. Significance is reported at *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

3.4. XAN + UA in the Diet Synergistically Inhibited HMVP2 PCa Allograft Tumor Growth and Impacted Oncogenic Signaling Pathways

Since the combination of XAN + UA demonstrated a synergistic ability to reduce PCa cell survival and proliferation and modulated oncogenic pathways important in cancer development and progression in cultured PCa cells, it was tested for tolerability and efficacy in vivo. For this experiment, the phytochemicals were supplemented individually and in combination into the diets of groups of mice in an allograft tumor model of PCa based on the HMVP2 cell line [12, 27]. The mice that were fed the combination diet had significantly smaller tumor volumes compared to all other groups at the conclusion of the study (Figure 4A). The dietary administration of the phytochemicals was well tolerated at the chosen concentrations as demonstrated by no significant differences in body weight (Figure 4B). Additionally, the inhibition of tumor growth produced by the combination diet was synergistic starting at 21 days after tumor inoculation through the end of the study, as indicated by Bliss Index values of > 0.1 (Figure 4C). These data indicate that dietary administration of the combination of XAN + UA at the concentrations used was well tolerated and effective at inhibiting PCa tumor growth better than either single compound given in the diet at the same dietary dose.

Figure 4.

The combination of XAN + UA administered in the diet synergistically inhibited PCa HMVP2 allograft tumor growth without toxicity. (A) Growth of allografted HMVP2 PCa tumors over time in male FVB/N mice that were fed diets supplemented with UA (0.2%), XAN (0.2%), XAN + UA (0.2% + 0.2%), or control. Diet start (red arrow) was on day 9 after inoculation, when tumors became palpable. (B) Average body weight over time. Data represents mean ± SEM (n = 7–8 mice per group) and a two‐way repeated measures ANOVA with Bonferroni's multiple comparisons test was used to compare groups. ****, $, # Significantly different from control, UA, and XAN alone respectively. (C) ^aTumor volumes were normalized to control tumor volume. Bliss index values were calculated as the difference between the observed and predicted effect, where the observed effect was the reduction in tumor volume provided by the UA + XAN diet, and the predicted effect is the theoretical reduction in tumor volume provided based on the UA and XAN individual agent diets. A value of > 0.1 is defined as synergistic.

EMT related proteins were evaluated in pooled allograft tumors and mice that were fed the combination diet had greater changes in e‐cadherin (increased) and Twist1 (decreased) compared to the other diet groups (Supplementary Figure 2). These data provide additional support that the combination had an overall effect on progression of PCa tumors in vivo.

3.5. Effect of XAN + UA in the Diet on the Metabolic Profile of Allograft Tumors

A metabolomics analysis was performed on the HMVP2 allograft tumors to further understand changes in metabolite profiles produced by dietary administration of the compounds. Partial Least Squares Discriminant Analysis of the data indicated a separation between tumors from mice fed the control diet compared to all other diet groups (Figure 5A). While tumors from the combination diet were clearly distinguished from the UA diet tumors, there was some overlap between these tumors and tumors from the XAN only diet group, indicating that the metabolic shift observed in tumors from the combination diet group may be driven by the actions of XAN to a greater extent than UA. Pathway analysis revealed 14 metabolic pathways that were significantly impacted in tumors from the combination diet group compared to tumors from control diet group, which were not significantly impacted in the single‐agent diet tumors compared to control diet tumors (Figure 5B).

Figure 5.

The combination of XAN + UA produced a unique metabolite profile in mouse PCa allograft tumors. (A) PLS‐DA from unbiased metabolomics analyses were performed on allograft tumor tissue to probe for differences in metabolite profile between mice fed control diet or diet supplemented with UA (0.2%), XAN (0.2%), or the combination. n = 8–14 tumors per group. (B) Metabolic pathways that were significantly impacted in the tumors from mice fed the combination diet which were not significantly impacted in the tumors from mice fed either single agent. Significance is reported at FDR < 0.05.

3.6. XAN + UA Impact on the Polyamine Synthesis Pathway

As shown in Figure 5B, synthesis of the polyamines spermidine and spermine was identified as one of the 14 metabolic pathways uniquely impacted in the tumors from the combination diet group. This pathway was selected for further investigation (Figure 6A). As shown, there was a trend toward an increase in the input metabolite L‐ornithine in the combination diet tumors (Figure 6B). L‐ornithine is converted into putrescine by the rate‐limiting enzyme ODC1, though no clear trends were observed for putrescine. A trend toward decreased spermidine levels was observed in the combination diet tumors compared to the single agent diet tumors. Of note, there was a significant decrease in the levels of S‐adenosyl‐l‐methionine (SAM) and 5'‐methylthioadenosine (5'‐MTA) in the tumors from the combination diet group, which are either required for or produced, respectively, during the conversion of putrescine to spermine (Figure 6B). As shown in Supplementary Figure 3, spermine levels were lowest in HMVP2 PCa cells treated with the combination. Collectively, these data indicate that synthesis of metabolites through this pathway may be reduced by the combination of XAN + UA leading to significant reductions of 5'‐MTA.

Figure 6.

The combination of XAN + UA impacted the polyamine synthesis pathway. (A) Diagram of the polyamine synthesis pathway. (B) Peak area of metabolites related to the polyamine synthesis pathway, derived from allograft tumors. Data represents minimum, median, and maximum values. For ornithine, putrescine, and spermidine a Kruskal‐Wallis test with Dunn's multiple comparisons test was used. For SAM and 5'‐MTA a Brown‐Forsythe test with Dunnett's T3 multiple comparisons test was used. (C) Representative western blot from HMVP2 cells treated with vehicle control, UA (10 µM), XAN (7.5 µM), or the combination for 24 h, acquired by x‐ray film, with average fold change quantitation. Data represents mean ± SEM (n = 3) and a one‐way ANOVA with Tukey's multiple comparisons test was used to compare groups. Significance is reported at *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

To provide further support for the metabolomics data, protein levels of the rate‐limiting enzyme ODC1 were evaluated. Following treatment of HMVP2 cells with the combination of XAN + UA, there was a significant reduction in protein levels of ODC1 compared to all other groups (Figure 6C).

3.7. XAN + UA in the Diet Significantly Inhibited Adenocarcinoma (AC) Formation in Ventral Prostate (VP) of HiMyc Mice and Impacted UPR Signaling

To investigate whether dietary administration of XAN + UA inhibited the progression of PCa from precancerous to cancerous lesions, we utilized the HiMyc transgenic mouse model [23]. In this model, mice develop PIN lesions which progress to invasive adenocarcinomas (AC) by 6 months of age [23]. Mice were placed on AIN‐93M diet control or diet supplemented with XAN, UA, or the combination and remained on diet until 6 months of age. Histopathological analyses of the genitourinary tracts revealed that mice fed the combination diet had significantly lower incidence of in situ and locally invasive AC in the VP compared to mice fed the control and single agent diets (Figure 7A,B). Similar to the allograft tumor study, the dietary administration of the phytochemicals was well tolerated at the chosen concentrations, as demonstrated by no significant differences in body weight or food consumption during the experimental period (Figure 7C,D). Representative images of H&E‐stained sections of the VP from mice on the respective diets are shown in (Figure 7E). The overall tumor burden of the mice that were fed the combination was reduced, as indicated by a significantly lower in situ AC percent area (Figure 7F). These data indicate that the combination of XAN + UA in the diet was well tolerated and effective at inhibiting PCa progression from precancerous lesions to AC and also reduced overall tumor burden compared to mice that received the control diet or single agent diets.

Figure 7.

The combination of XAN + UA administered in the diet inhibited PCa progression in HiMyc mice without toxicity. (A, B) Percent incidence of in situ and locally invasive AC in the ventral prostates of HiMyc mice that were fed diets supplemented with UA (0.2%), XAN (0.2%), XAN + UA (0.2% + 0.2%), or control. Data represents percent of mice that presented with each type of AC in each group, n = 20–21 mice per group. Two‐sided Fisher's exact test was used for head‐to‐head comparisons. (C) Average body weight over time. Data represents mean ± SEM (n = 27–29 mice per group) and a mixed‐effects analysis with Tukey's multiple comparisons test was used to compare groups. (D) Average food consumption per day, n = 27–29. Representative images of H&E‐stained VP tissue (E) with quantitation of percent area affected by in situ AC in the VP (F). For (D) and (F) data represents mean ± SEM, and a Kruskal‐Wallis test with Dunn's multiple comparisons test was used to compare groups. Significance is reported at *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Activation of UPR signaling, which was demonstrated by the combination in cultured PCa cells, was evaluated in pooled VP tumors from the HiMyc mice. The mice that were fed the combination diet had elevated protein levels of ATF4 and had higher levels of CHOP compared to tumors from the other diet groups (Supplementary Figure 4), indicating activation of UPR signaling. Additionally, since CHOP is known as a proapoptotic protein [35], other markers of apoptosis were evaluated. Protein levels of cleaved PARP and cleaved caspase 3 were both increased in the pooled tumors from the VP of the HiMyc mice fed the combination diet compared to the other diet groups (Supplementary Figure 4). Taken together, these data suggest that UPR signaling activation and subsequent induction of apoptosis may be important to the overall synergistic inhibition of PCa growth and progression caused by the combination of XAN + UA in HiMyc mice.

4. Discussion

In the current study, the combination of XAN + UA demonstrated an ability to significantly and synergistically inhibit PCa growth and progression. PCa cells were sensitive to the combination, demonstrating a synergistic reduction in cell survival and proliferation. The combination inhibited oncogenic signaling related to cell cycle progression and induced ER stress with UPR activation in PCa cells. Administration of the combination in the diets of mice was well tolerated and led to a synergistic inhibition of growth of established tumors in an allograft study using HMVP2 PCa cells, as well as a significant inhibition of PCa progression from precancerous lesions to AC in the VP of HiMyc mice. Notably, the single agent diets, at the doses used, did not provide any inhibitory effect on tumor growth or progression in either mouse model emphasizing the synergistic capacity of XAN + UA when used in combination. Mechanistic studies conducted in both cultured PCa cells and tumors revealed that the combination induced pleiotropic effects on PCa, including inhibition of the polyamine synthesis pathway, altered cell cycle and EMT related proteins and induced ER stress leading to UPR signaling activation and subsequent apoptosis. Collectively, these data suggest that combining XAN + UA may represent a novel chemopreventive strategy for PCa and possibly other cancers.

In cultured PCa cells, the STAT3 signaling pathway was chosen for analysis due to its role in maintaining sustained proliferation of PCa [36]. In HMVP2 PCa cells, the combination significantly reduced levels of phosphorylated STAT3 while total STAT3 protein levels were unchanged, suggesting that the combination acted to inhibit STAT3 activation. To evaluate this further, protein levels of STAT3 downstream targets, including cyclin D1, cdc6, and c‐MYC were assessed and found to be significantly reduced by the combination. Notably, these proteins are all implicated in promoting cell cycle progression through the G1 to S phase transition [32, 33, 34]. Further assessment of proteins involved in this transition revealed reduced phosphorylation of Rb and increased p21 in a manner consistent with inhibition of cell cycle progression. These results with the combination of XAN + UA are similar to data obtained with the combination of curcumin (CURC) + UA published in our recent study [37] and suggest that STAT3 is a major target of phytochemicals in PCa cells most likely via inhibition of JAK2 [38].

UPR signaling is a biologically conserved mechanism which is activated in cells that have impaired folding capacity at the ER (ER stress) in order to try to restore homeostasis [39, 40]. The role of UPR signaling activation in cancer development and progression has been debated and ultimately has demonstrated both pro‐survival and pro‐apoptotic effects depending on the context and cancer type. Interestingly, a study of gene expression at different stages of PCa development in two relevant transgenic mouse models revealed that UPR regulators and their downstream target genes had lower levels of expression in PCa than in normal prostate tissue, suggesting that UPR activation may not contribute pro‐survival effects in PCa development and progression [40]. In situations of persistent activation of the UPR pathway failing to restore homeostasis, the outcome is pro‐apoptotic, and studies have reported both XAN [41, 42] and UA [43] as individual agents exert anti‐cancer effects through this mechanism. In both mouse and human PCa cells in the current study, the combination of XAN + UA produced elevated protein levels of ATF4 with elevated levels of CHOP greater than the single agent treatments, indicating an activation of UPR signaling that was enhanced by using the combination. Further, pretreatment with an ER stress inhibitor, 4‐PBA, partially rescued cell survival following XAN + UA treatment. These results are also similar to our recent observations with the combination of CURC and UA [37] and suggest that activation of UPR is also a common mechanistic pathway associated with phytochemical combinations, especially combinations with UA.

Dietary administration of the compounds produced significant and synergistic inhibition of HMVP2 allograft tumor growth and was well tolerated at the doses used over this short‐term experiment. Metabolomics analysis of the HMVP2 allograft tumors showed that a number of metabolic pathways were impacted by the combination diet not impacted by the individual compound diets. One pathway of interest was the polyamine synthesis pathway. Previous studies have demonstrated that overexpression of the rate limiting enzyme in this pathway, ODC1, was sufficient to induce tumorigenesis of normal prostate epithelial cells in a xenograft model [44]. Additionally, in humans, ODC1 gene expression is elevated in both high‐grade PIN and PCa compared to normal prostate tissue [44, 45]. A year‐long clinical trial in men which targeted ODC1 pharmacologically with the drug difluoromethylornithine (DFMO) reported that treatment led to lesser increases in prostate volume compared to placebo treatment, with a 0.94% increase in volume compared to an 11.4% increase, respectively [46]. In the allograft tumors from mice fed the combination diet, there was an apparent accumulation of ornithine, suggesting ODC1 inhibition. PCa cells treated with the combination exhibited reduced ODC1 to a greater extent than either single agent treatment supporting this hypothesis. Additionally, trends towards a decrease in spermidine and spermine were observed, as well as significant reductions of SAM and 5'‐MTA in the allograft tumors providing evidence that metabolite flow through this pathway was reduced by the combination of XAN + UA.

Notably, pharmacologic reduction of polyamine levels, especially spermine, led to a significant inhibition of PCa xenograft tumor growth associated with reductions in SAM levels [47]. The reductions in SAM and 5'‐MTA observed in allograft tumors treated with the combination of XAN + UA may have also contributed to the overall mechanism of inhibition of allograft tumor growth. The reductions in SAM may, at least in part, be the result of the reduced 5'‐MTA levels observed, as this metabolite can be recycled into methionine (MET) by a salvage pathway (MTA cycle) via the enzyme methylthioadenosine phosphorylase (MTAP) [48] and could lead to reduced synthesis of SAM due to reduced MET salvage. Furthermore, in a preliminary experiment (data not shown), we found that HMVP2 cells treated with the combination had significantly reduced expression of MAT2A, the enzyme responsible for converting MET to SAM [49], suggesting that reductions in SAM synthesis may have also contributed to reduced levels of this metabolite. Future experiments will address these interesting observations in more detail. As the universal methyl donor, reductions in SAM could lead to epigenetic changes such as hypomethylation of tumor suppressor genes thereby allowing their expression and inhibition of tumor progression. Studies of UA as a single agent have reported its ability to cause epigenetic changes in PCa, including hypomethylation, which contributed to the inhibitory effects on tumor growth [50]. Thus, the combination of XAN + UA may cause epigenetic effects in PCa cells as a result of modulation of polyamine synthesis and altered levels of 5'‐MTA and SAM.

In HiMyc mice, the combination diet synergistically reduced PCa progression. Analysis of pooled VP tumors from these mice revealed activation of UPR signaling similar to that observed in cultured PCa cells that was associated with increased levels of markers of apoptosis (Figure 3, Supplementary Figures 1 and 4), These data suggest that at the doses used the predominant mechanism in the HiMyc mouse model was induction of ER stress and activation of the UPR signaling pathway with subsequent induction of apoptosis.

In summary, the current results indicate that supplementation of the combination of XAN + UA in the diets of mice provided a synergistic inhibition of growth of established allograft tumors, as well as a significant inhibition of progression from precancerous to cancerous lesions in the VP of HiMyc mice. These combinatorial effects were synergistic at the doses used since dietary supplementation of XAN or UA alone provided no significant inhibitory effects in either mouse model. Further experiments revealed that the inhibitory effect of the combination of XAN + UA in the allograft tumor model was primarily associated with inhibition of the polyamine synthesis pathway and reduced EMT while the chemopreventive effects of the combination on PCa progression in the HiMyc mouse model was associated primarily with induction of ER stress with UPR signaling activation and apoptosis. These mechanisms, with the exception of reduced EMT, were observed in both mouse and human PCa cells in culture when treated with the combination. The differences in predominant mechanism observed between the two in vivo models is likely due to differences in biology of the models as well as differences in levels of the compounds reaching the tumor sites. Further studies are needed to explore these possibilities as well as the clinical development of the combination as a novel chemopreventive strategy for PCa.

In conclusion, taken together with our previous results [37, 51], the current study confirms that certain combinations of natural compounds can be identified through screening approaches that provide synergistic inhibition of PCa progression in relevant mouse models. An ideal clinical setting for testing these combinations is men who are placed on “Active Surveillance” protocols [4]. Understanding dosing regimens that minimize potential toxicity and developing enhanced bioavailable forms of these compounds is an important goal for future clinical application [52].

Author Contributions

Rachel Clark: conceptualization, data curation, formal analysis, investigation, methodology, visualization, writing – original draft, writing – review and editing. Achinto Saha: concenptualization, formal analysis, investigation, methodology writing – review and editing. G. Lavender Hackman: data curation, formal analysis, investigation, writing – review and editing. Chelsea A. Friedman: investigation, writing – review and editing. Ruggiero Gorgoglione: data curation, formal analysis, investigation, writing – review and editing. Stefano Tiziani: concenptualization, funding acquisition, resources, writing – review and editing. John DiGiovanni: concenptualization, formal analysis, funding acquisition, methodology, resources, writing – original draft, writing – review and editing.

Ethics Statement

All animal experiments in this study were approved and performed according to the guidelines of the Institutional Animal Care and Use Committees (IACUC) of The University of Texas at Austin (AUP‐2025‐00005).

Consent

All authors have approved the publication of this submission.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supplementary Figure 1: The combination of XAN + UA impacted oncogenic and UPR signaling in LNCaP PCa cells. Supplementary Figure 2: The combination of XAN + UA impacted EMT protein levels in PCa allograft tumors. Supplementary Figure 3: The combination of XAN + UA reduced the abundance of spermine in HMVP2 PCa cells. Supplementary Figure 4: The combination of XAN + UA impacted unfolded protein response and apoptosis in HiMyc VP tissue.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.