Bench-to-Bedside Evaluation of Sulforaphane/BroccoMax^® on Fatty Acid Synthesis in Prostate Cancer

Abstract

Fatty acid synthesis pathway is a valid target for prevention of prostate cancer. However, a clinical grade inhibitor of fatty acid synthesis is still lacking. This bench-to-bedside study was undertaken to determine the feasibility of fatty acid synthesis inhibition using broccoli constituent sulforaphane (SFN) and its clinical grade formulation BroccoMax^® (BMAX). Oral administration of SFN to Hi-Myc mice resulted in inhibition of prostate adenocarcinoma burden by about 61% that was accompanied by a significant decrease in prostate tumor levels of c-Myc and PCNA proteins and increased apoptosis. Expression of acetyl-CoA carboxylase 1 (ACC1) and fatty acid synthase (FASN) were lower by about 46% and 31%, respectively, in the prostate tumor of SFN-treated mice when compared to that of control mice (P < 0.001). Plasma levels of total free fatty acids (TFFA), cholesterol, and total phospholipids were decreased significantly following SFN treatment. In a double-blind clinical trial, patients with histologically confirmed prostate cancer were randomized to the BMAX group (n= 19) or placebo group (n= 22). Patients were treated with 4 capsules BMAX or 4 capsules of matching placebo orally two times daily after breakfast and dinner for 4 weeks. Prostate tumor expression of c-Myc, ACC1, FASN, and Ki-67 proteins were significantly lower in the BMAX arm when compared to PBO group. However, serum or prostate tumor level of acetyl-CoA or TFFA was not decreased by BMAX treatment. A longer duration treatment with BMAX in early-stage prostate cancer patients may be necessary to lower circulating or prostate tumor level of TFFA.

Introduction

Prostate cancer is a leading cause of cancer-related deaths in American men (1). According to the American Cancer Society, diagnosis of 313,780 new prostate cancer cases was expected in the year 2025 alone. The American Cancer Society also estimates 35,770 deaths from prostate cancer in 2025. The biology underlying prostate cancer progression is complex and associated with amplification of transcription factors [e.g., androgen receptor (AR) and c-Myc], somatic or germline mutations (e.g., TP53 and PI3K), gene deletion (e.g., PTEN deletion), alterations in DNA repair pathway (BRCA2, BRCA1, and ATM), and defective cell signaling pathways (2,3).

Metabolic reprogramming is another hallmark of many solid tumors including prostate cancer (4). Fatty acid synthesis is uniquely elevated in human prostate cancer and contributes to disease progression (5,6). Overexpression of key enzymes of the fatty acid synthesis pathway including acetyl-CoA carboxylase 1 (ACC1) and fatty acid synthase (FASN) has been reported in human prostate cancer when compared to normal prostate tissue (7,8). For example, immunohistochemistry revealed a 5.7- and 8.5-fold increase in FASN protein expression in human prostatic intraepithelial neoplasia (PIN) and prostate adenocarcinoma (PAC), respectively, when compared to normal prostate tissue (7). Studies from our laboratory have revealed a statistically significant increase in expression of ACC1 protein in PAC when compared to normal prostate tissue (8). Inhibition of ATP citrate lyase (ACLY), which initiates fatty acid synthesis by converting citrate to acetyl-CoA, by a bioactive compound from cucumber (Cucurbitacin B) significantly inhibited growth of PC-3 and LNCaP human prostate cancer cells by promoting apoptotic cell death (9). Overexpression of ACC1 protein, which is the rate limiting enzyme in the fatty acid synthesis pathway and catalyzes conversion of acetyl-CoA to malonyl-CoA, in human prostate cancer patients was associated with shorter disease-free survival (10). RNA interference of FASN, which catalyzes conversion of acetyl-CoA and malonyl-CoA to saturated fatty acids (e.g., palmitic acid), in LNCaP cells resulted in a decrease in synthesis of by-products of the fatty acid synthesis pathway (e.g., triglycerides and phospholipids) that was accompanied by a reduction in cell volume, loss of cell-cell contact, and formation of spider-like protrusions (11). Moreover, FASN silencing inhibited LNCaP cell growth due to apoptosis induction (11). Increased expression of FASN protein in human prostate cancer patients was associated with a more aggressive phenotype (12). Overexpression of FASN in immortalized prostate epithelial cell lines and LNCaP cells increased cell proliferation (13). Transgenic overexpression of FASN in mouse prostate resulted in PIN lesions (13). Blood levels of saturated fatty acids were shown to be good markers of lipogenesis as well as the risk of prostate cancer (14). Collectively, these results indicate that enzymes of the fatty acid synthesis pathway are valid targets for prevention and treatment of human prostate cancer.

The pharmaceutical industry is actively pursuing drug discovery for ACLY, ACC1, and FASN inhibitors as a potential therapy for cancer (15). An FDA approved drug that can lower levels of fatty acids is Orlistat (Xenical^®) (16). However, Orlistat has side effects and is currently not approved for cancer treatment (17). Anti-microbial agent Triclosan (C75) has also been tested for cancer treatment (18). In a first-in-human clinical trial in non-small cell lung cancer, ovarian cancer, and breast cancer, the disease control rate with fatty acid synthase inhibitor TVB-2640 monotherapy was 42% (19). However, no patient treated with monotherapy had a complete or partial response (19). In phase II study of first relapse high-grade astrocytoma, TVB-2640 was found to be a well-tolerated, and it could be safely combined with bevacizumab (20). The ACC inhibitor MK-4074 reduced hepatic steatosis but elevated plasma triglycerides (21). Certain phytochemicals were also shown to decrease expression of FASN and/or lower fatty acid levels in prostate cancer cells in vitro (e.g., epigallocatechin-3-gallate from green tea) (22–24). However, a randomized clinical trial with green tea failed to show downregulation of FASN in prostate biopsies (25).

We have shown previously that broccoli constituent sulforaphane (SFN) downregulates expression of ACC1 and FASN in LNCaP and 22Rv1 human prostate cancer cells in vitro (26). SFN is stored in plant cells as a glucosinolate (glucoraphanin). Glucoraphanin is converted to SFN by catalytic mediation of myrosinase, which is released upon cutting or chewing of broccoli. The gut microbiome can also mediate this conversion (27). The primary objective of the present study was to determine whether daily oral administration of SFN-rich dietary supplement BroccoMax^® (BMAX) for 4 weeks decreases expression of ACC1 and/or FASN proteins and suppresses levels of total free fatty acids (TFFA) in the serum and/or prostate tumor of prostatectomy patients. The clinical trial was complemented with a pre-clinical study using Hi-Myc transgenic mice.

Materials and methods

Ethics statement

Use of Hi-Myc mice was approved by the Institutional Animal Care and Use Committee of the University of Pittsburgh (protocol # 17020271). The clinical trial was approved by the Institutional Review Board of the University of Pittsburgh (protocol # 19050340) and registered at the ClinicalTrials.gov (NCT03665922). We obtained written informed consent from each patient prior to distribution of PBO or BMAX. This study was conducted in accordance with the U. S. Common Rule.

Reagents

SFN (catalog #S8047) was purchased from the LKT laboratories (St. Paul, MN). Anti-c-Myc (cat.# ab39688; RRID:AB_731661) antibody was from Abcam (Waltham, MA). Anti-proliferating cell nuclear antigen (PCNA) antibody (cat. # sc-25280; RRID: AB_628109) was purchased from Santa Cruz Biotechnology (Dallas, TX). CoraLite^® Plus 647-conjugated ACC1 polyclonal antibody (cat.# CL647–21923; RRID: AB_2934949) and CoraLite^® Plus 488-conjugated FASN antibody (cat.# CL488–81079; RRID: AB_3084499) were from Proteintech Group (Rosemont, IL). Anti-ACC1 (cat.# 21923–1-AP; RRID: AB_11042445) antibody used for immunohistochemistry of human specimens were from Proteintech Group. Anti-c-Myc (cat.# ab39688; RRID:AB_731661) and anti-Ki-67 (cat.# 9449S; RRID: AB_2797703) antibodies for immunohistochemistry for human specimens were from Abcam and Cell Signaling Technology (Danvers, MA), respectively. Kits for determination of lactate (cat. # K607–100) and cholesterol (cat. # K603–100) levels were purchased from BioVision (Milpitas, CA). Kits for determination of TFFA (cat. # MAK044) and total phospholipids (cat. # MAK122) were purchased from Sigma-Aldrich (St. Louis, MO). Kits for mouse specific malonyl-CoA ELISA (cat. # MBS705127) and human specific malonyl-CoA ELISA (cat. # MBS2606598) were purchased MyBiosource (San Diego, CA). Acetyl-CoA ELISA kit (cat. # orb1146757) was from Biorbyt (San Francisco, CA). EnzyChrom citrate assay kit (cat. # ECIT-100) was from BioAssay Systems (Hayward, CA).

Chemoprevention study in Hi-Myc mice

Breeding pairs of Hi-Myc mice were purchased from the National Institutes of Health Mouse Repository. After transgene verification, 10-week-old male Hi-Myc mice were randomly divided into two groups. Control mice (n=21) were orally treated with 100 μL corn oil five times/week for 16 weeks, while second group of mice (n=19) were orally treated with 1 mg SFN/mouse in 100 μL corn oil five times/week for 16 weeks. At the end of study, mice were euthanized by carbon dioxide inhalation. Blood samples were collected and plasma specimens were obtained by centrifugation at 3,000 rpm for 5 minutes. Prostate tissue was collected at the time of sacrifice. Tissues were fixed in 10% neutral-buffered formalin for hematoxylin and eosin (H&E) staining and immunohistochemistry. Plasma samples were aliquoted and stored at −80°C.

Sample size justification and statistical tests for the Hi-Myc mice study

We assumed a 50–60% decrease in prostate cancer burden in SFN group compared to control group. A sample size of n=25 mice/group provided a power of at least 0.80 with P value of 0.05. Statistical significance of difference was determined by two-sided t-test using GraphPad Prism (version 8.0.0).

Immunohistochemistry for Hi-Myc mice study

Immunohistochemistry was performed as described by us previously (28). Formalin-fixed paraffin-embedded mouse prostate tumor tissue sections were deparaffinized, rehydrated, and washed with phosphate-buffered saline (PBS) and then antigen retrieval was performed using a citrate retrieval buffer solution (pH 6.0) for 20–30 minutes at 100°C. Sections were treated with blocking buffer for 1 hour followed by incubation with primary antibodies [anti-c-Myc antibody (1:100 dilution); anti-proliferating cell nuclear antigen (PCNA) antibody (1:200 dilution)] overnight in humidified chambers at 4°C. Slides were then washed with PBS and incubated with horseradish peroxidase-conjugated secondary antibody for 2 hours at room temperature. Sections were washed with PBS. Color was developed by incubation with 3,3’-diaminobenzidine tetrahydrochloride and counterstained with hematoxylin. Stained sections were examined under Nikon confocal microscope or Leica microscope equipped with DFC 450C digital camera or Olympus Fluorview 3000 confocal microscope.

TUNEL assay for apoptosis in Hi-Myc mice prostate tissue

TUNEL staining was performed using ApopTag^® Fluorescein In Situ Apoptosis Detection kit as suggested by the supplier. Paraffin-embedded prostate tissue sections from control and SFN-treatment groups were deparaffinized by xylene and serially washed with absolute ethanol, 95% ethanol, 70% ethanol, 50% ethanol, and PBS. Freshly prepared proteinase K (20 μg/mL) solution was then applied directly to the sections for 15 minutes at room temperature. Tissue sections were then washed with PBS and incubated in equilibrium buffer for 10 seconds and then with the TdT enzyme. The tissue sections were then incubated in a humidified chamber for 60 minutes at 37°C. Next, the tissue sections were treated with stop/wash buffer supplied with the kit for 15 seconds and washed with PBS followed by treatment with anti-digoxigenin conjugate at room temperature for 30 minutes. Sections were washed with PBS, mounted on a glass coverslip in DAPI containing mounting medium. Stained sections were examined under a Nikon confocal microscope at 20× objective. The number of TUNEL-positive cells per high-power field from five to six non-overlapping fields of each section was calculated for each group using NIS element software.

Measurement of metabolites in the plasma of Hi-Myc mice

Levels of TFFA, total phospholipids, cholesterol, and lactate in the plasma of control and SFN-treated Hi-Myc mice were determined using commercially available kits and by following the manufacturer’s protocol as described by us previously (23,26).

BMAX

The BMAX and matching placebo (PBO) capsules were generously supplied by the Jarrow Formulas (Los Angeles, CA). Each BMAX capsule contained 30 mg of SFN precursor glucoraphanin derived from broccoli seeds and myrosinase derived from broccoli sprouts. In a previous clinical trial with BMAX^® intervention in breast cancer patients, Grade 3 or 4 adverse events were not reported in any subject (29). The patients were randomized in block to take 4 capsules of BMAX or 4 capsules of matching PBO orally two times daily after breakfast and dinner for 4 weeks. Last dose was taken in the evening prior to prostatectomy.

Inclusion criteria

Inclusion criteria included: (a) men ≥ 18 years of age requiring prostatectomy after histopathological diagnosis of PAC; (b) subjects were instructed to orally take PBO or BMAX capsules daily for 4 weeks prior to surgical removal of the prostate tumor; (c) subjects are able to swallow BMAX or PBO pills; (d) subjects in good health with normal liver enzyme levels and blood count (e) subjects agreeing to refrain from dietary intake of broccoli or other cruciferous vegetables; and (f) subjects must sign a written informed consent.

Exclusion criteria

The exclusion criteria included: (a) ineligible to undergo prostatectomy due to co-morbidities; (b) subjects with diagnosis of a second malignancy at least 3 years prior to enrollment in this trial except for adequately treated basal cell or squamous cell skin cancer; (c) subjects suffering with malabsorption issues or gastrointestinal ailments; (d) subjects with prior or concurrent androgen deprivation therapy with luteinizing hormone-releasing hormone agonist or antagonists; (e) subjects taking any other investigational agent or dietary supplement; (f) subjects with active infection, uncontrolled angina, New York Heart Association class III or IV heart failure, uncontrolled or uncontrollable hypertension, severe diabetes, chronic liver disease; and (g) subjects with prior history of allergic reactions to cruciferous vegetables or specific fillers used in the PBO capsules.

Sample size justification and statistical tests for the clinical trial

The power calculation was based on our own unpublished data on TFFA levels in baseline (pre-treatment) plasma samples collected from our clinical trial in patients with atypical nevi. The mean of the log-transformed TFFA was 2.4 and the standard deviation was 0.4. A sample size of 20/arm provided at least 80% power (at α=0.05) for mean changes in log(TFFA) of 12% or greater. Statistical significance of difference in biomarkers was determined by two-sided t-test.

Baseline evaluations

Baseline evaluations included data collection on patient’s medical history, age, cancer history, review of medications, food intolerances and food habits, and concomitant medications, physical examination including height and weight measurements, blood collection (approximately 15 mL) in serum separator tiger tubes for biomarker analyses, and dispensation of BMAX or PBO capsules.

Day of prostatectomy evaluations

Height and weight measurements, blood (approximately 15 mL) collection in tiger tubes for biomarker analyses, adverse event assessment, and prostatectomy were performed on the day of surgery. Serum was separated and stored at −80°C at the Biospecimen core of the University of Pittsburgh. A portion of the surgically removed prostate tumor tissue was fresh frozen at −80°C. The remaining piece of the prostate tumor tissue was fixed in 10% neutral buffered-formalin and embedded in paraffin. These specimens were also stored at the Biospecimen core of the University of Pittsburgh. Collection of prostate tumor sample was not possible from every patient for a variety of reasons. The University of Pittsburgh mandates processing and storage of specimens from clinical trials at the Biospecimen core. The surgery for some patients was delayed and did not occur during working hours of the Biospecimen core. However, the blood samples from these patients were stored in the clinic of Dr. Jacobs before they were transferred to the Biospecimen core next day for processing and storage.

End points

Reduction in serum TFFA in subjects treated with BMAX^® compared to PBO was the primary end point. The secondary end points included: (1) prostate tumor levels of TFFA; (2) proliferative marker Ki-67 and apoptotic index (TUNEL-positive cells) in tumor sections; (3) tumor expression of ACC1, FASN, Ki-67, and c-Myc by immunohistochemistry; and (4) serum and prostate tumor levels of fatty acid metabolism intermediates (citrate, acetyl-CoA, malonyl-CoA, and TFFA).

Measurement of metabolites in the serum and prostate tumor of patients

Serum and tumor lysates from PBO and BMAX^® groups were used for measurements of citrate, acetyl-CoA, malonyl-CoA, and TFFA. For the preparation of tumor tissue lysates, tissue pieces (~10 mg) were grounded using a glass homogenizer in ice-cold PBS supplemented with protease and phosphatase inhibitors on ice and centrifuged at 14,000 rpm for 10 minutes at 4°C. Deproteinized serum and tumor lysates obtained through a 10 kDa spin filter were used for the citrate assay following supplier’s protocols.

Preparation of tissue microarray (TMA) and immunofluorescent staining for clinical specimens

Antigen retrieval was done by placing the TMA sections on a glass slide within a chamber, exposing it to light-emitting diode light at room temperature for 15 minutes. After that, 1 mL of bleaching solution containing H₂O₂, (30%, (wt/vol), 0.02 M NaOH was added into the chamber. After rinsing with PBS to remove the bleaching solution and autofluorescence, the sections were permeabilizaed and blocked with 0.2% Triton X-100 in PBS, followed by blocking with blocking solution consisting of 1% bovine serum albumin, 100 mM NH₄Cl, and 150 mM maleimide for 30–60 minutes. The sections were then treated with the desired primary antibody (anti-c-Myc- 1:100 dilution; anti-ACC1 antibody- 1:100 dilution; anti-FASN antibody- 1:100 dilution, and anti-Ki-67 antibody- 1:200 dilution) either for 2 hours at room temperature or overnight at 4°C. The sections were washed with PBS and treated with appropriate secondary antibody for 1 hour at room temperature. For fluorescence imaging, a buffer consisting of700 mM N-acetylcysteine in double distilled water adjusted to pH 7.4 was added to the TMA to prevent photo-crosslinking. Antibody elution followed each cycle of fluorescence imaging, involving washing with PBS and incubating with elution buffer containing 0.625 M L-glycine, 3.75 M Urea, 3.75 M guanidium chloride in double distilled water, adjusted to pH 2.5 and tris-(2-carboxyethyl)phosphine hydrochloride (TCEP) solution (350 mM TCEP in double distilled water for 15 minutes and this step was repeated three times.

Untargeted high-resolution lipidomic analysis using human serum

Deuterated LipidSplash internal standard mix (10 μL) was added to 50 μL of thawed plasma. A modified Folch extraction was performed using chloroform, methyl alcohol, and water. Samples were placed on ice for 10 minutes and then centrifuged at 2,500 x g for 15 minutes. The organic phase (700μL) was transferred to a clean glass tube and dried under nitrogen gas. Samples were then re-suspended in 100 μL of 1:1 acetonitrile:isopropanol. Prior to LC-HRMS analysis. Briefly, 3 μL was injected into a Thermo Vanquish UHPLC and separated over a reversed phase Thermo Accucore C-18 column (2.1×100mm, 5μm particle size) that was maintained at 55°C. The mobile phase consisted of t solvent A (50:50 H₂O:ACN with 10mM ammonium acetate/0.1% acetic acid) and solvent B (90:10 IPA:ACN with 10 mM ammonium acetate/0.1% acetic acid). The chromatographic conditions were: 30% of solvent B at a flow rate of 0.2 mL/min. The gradient was the following: an increase to 43% solvent B over 2 minutes, followed by an increase to 55% of solvent B over 0.1 minutes. Subsequently, the organic phase increased to 65% of solvent B over 10 minutes, and to 85% of solvent B over 6 minutes. The final increase of solvent B to100% was achieved in2 minutes that was held for 5 minutes followed by equilibration at 30% of solvent B for 5 minutes. The Thermo ID-X tribrid mass spectrometer was operated in negative ESI mode. A data-dependent MS² method scanning in Full MS mode from 200 to 1500 m/z at 120,000 resolution with an AGC target of 5e4 for triggering MS² fragmentation using stepped HCD collision energies at 20,40, and 60% in the orbitrap at 15,000 resolution was set. Spray voltage was set at 2.4kV and source gas parameters were 35 sheath gas, 5 auxiliary gas at 300°C, and 1 sweep gas. Individual free fatty acid peak areas were then determined using Quan Browser (Thermo Fisher Xcalibur ver. 2.7) and normalized to internal standard peak area

Data availability

All raw data for preclinical studies and de-identified patient data may be shared upon reasonable request to the corresponding author.

Results

Oral administration of SFN inhibited prostate cancer progression in Hi-Myc mice

SFN administration did not affect body weight of the mice (Figure 1A). The burden of invasive PAC was 61% lower in the SFN treatment group when compared to control mice (Figure 1B). Expression of c-Myc and PCNA were decreased by about 32% and 34%, respectively, after oral administration of SFN (Figure 1C). The number of TUNEL-positive apoptotic cells was higher by about 1.3-fold in the prostate tumor of SFN-treated mice when compared to control mice (Figure 1C).

Figure 1:

Effect of oral sulforaphane (SFN) administration on prostate tumor development in Hi-Myc mice. (A) Body weight of the mice over time of the control group and SFN treatment groups. (B) Burden of invasive prostate adenocarcinoma (PAC) in control and SFN treated mice. (C) Representative immunohistochemical images (20× objective magnification, scale bar = 100 μm) of c-Myc and PCNA expression, and TUNEL-positive apoptotic cells in the prostate tumor tissues of control and SFN-treated mice. Quantitative results shown are mean ± 95% CI (n= 21 for control group and n= 19 for SFN group). Statistical significance of difference was determined by two-sided t-test.

SFN administration decreased ACC1 and FASN protein expression in the prostate tumor of Hi-Myc mice

Figure 2 shows immunohistochemistry for ACC1 and FASN proteins. Expressions of ACC1 and FASN proteins were lower by 46% and 31%, respectively, in the prostate tumor of SFN-treated mice in comparison with control mice (Figure 2).

Figure 2:

Effect of oral sulforaphane (SFN) administration on expression of fatty acid synthesis enzyme proteins in the prostate of Hi-Myc mice. Representative immunohistochemical images (60× oil objective magnification, scale bar = 10 μm) for expression of acetyl-CoA carboxylase 1 (ACC1) and fatty acid synthase (FASN) proteins in prostate adenocarcinoma (PAC) of control and SFN-treated Hi-Myc mice. The right panels show quantitation of corresponding proteins analyzed using ImageJ software. Results shown are mean ± 95% CI (n= 20 for control and n= 19 for SFN group). Statistical significance of difference was determined by two-sided t-test.

SFN administration decreased TFFA levels in the plasma of Hi-Myc mice

SFN administration resulted in about 30% decrease in plasma TFFA levels when compared to control mice (Figure 3). The plasma levels of phospholipids and cholesterol were also decreased significantly after treatment with SFN (Figure 3). Cholesterol is a by-product of fatty acid synthesis pathway. Acetyl-CoA from fatty acid synthesis pathway is utilized to synthesize cholesterol. HMG-CoA reductase is the rate limiting enzyme in cholesterol biosynthesis and SFN was shown to downregulate HMG-CoA reductase (30). These results may explain suppression of cholesterol level by SFN treatment. A plasma level of lactate was not significantly different from that of the control mice (Figure 3).

Figure 3:

Plasma levels of total free fatty acids (TFFA), phospholipids, cholesterol, and lactate in Hi-Myc mice of the control group and SFN treatment group. Results shown are mean ± 95% CI (n= 21 for control and n= 19 for SFN groups). Statistical significance of difference was determined by two-sided t-test.

Patient characteristics

We obtained written informed consent from each patient and the studies were conducted in accordance with the United States Common Rule. In this cohort, all patients underwent a prostate biopsy to diagnose prostate cancer prior to surgery. Many of these patients had a prostate magnetic resonance imaging prior to their diagnostic prostate biopsy. Patients typically did not undergo surveillance imaging for metastatic disease unless they had high-risk disease on biopsy, this explaining why most patients were clinically metastatic disease (Table 1). If metastasis was related to pathologic stage, this is because metastatic sites are not evaluated during prostatectomy (only localized disease and lymph nodes). All patients underwent a lymph node dissection along with their prostatectomy to evaluate nodal status. A total of 49 prostate cancer patients scheduled for prostatectomy were consented. Two patients were not available because the surgery was cancelled. Three patients were removed from the study due to non-compliance. The remaining patients withdrew from the study. A total of 41 patients [n=22 in the PBO group and n=19 in the BMAX arm] completed the study. The patient characteristics are summarized in Table 1. Only one patient was African American and rest of the patients were Caucasian. Age of patients, baseline serum prostate specific antigen (PSA) level, Gleason score, and tumor stage were comparable between PBO and BMAX groups (Table 1). Representativeness of study participants is summarized in Supplementary Table S1.

Table 1.

Patient characteristics

		PBO	BMAX	P3
n		22	19
Age (years) y	Mean ± SD	65.9 ± 4.3	63.5 ± 6.5	0.170
PSA, ng/mL 1	Mean ± SD	6.6 ± 3.6	8.0 ± 3.9	0.353
	Median	5.8	7.7
	Range	2.1–17.0	2.6–15.1
Gleason score (tumor grade)	3+3 (group 1)	3 (13.6%)	1 (5.3%)	0.415
	3+4 (group 2)	14 (63.6%)	14 (73.7%)
	4+3 (group 3)	3 (13.6%)	2 (10.5%)
	4+4 (group 4)	0 (0%)	1 (5.3%)
	4+5 (group 5)	0 (0%)	1 (5.3.%)
	5+4 (group 5)	2 (9.1%)	0 (0%)
Tumor stage (pTNM) 2	T2	11 (50%)	7 (36.8%)	0.573
	T3a	7 (31.8%)	9 (47.4%)
	T3b	3 (13.6%)	2 (10.5%)
	N0	20 (90.9%)	18 (94.7%)	0.355
	N1	1 (4.5%)	0 (0%)
	Mx	21 (95.5%)	18 (94.7%)	0.631
Comorbidity	Metabolic, n (%)
	Hyperlipidemia/Hypertriglyceridemia	11 (50.0%)	2 (10.5%)	0.009
	Diabetes mellitus	4 (18.2%)	1 (5.3%)	0.350
	Fatty liver disease	1 (4.5%)	0 (0%)	1.000
	Chronic liver disease	0 (0%)	4 (21.1%)	0.038
	Thyroid disorder	4 (18.2%)	3 (15.8%)	1.000
	Cardiovascular4 n (%)	14 (63.6%)	13 (68.4%)	1.000
	Other n (%)
	BPH	8 (36.4%)	4 (21.1%)	0.325
	Prior malignancy	1 (4.5%)	4 (21.1%)	0.164

¹Prostate-specific antigen (PSA) value was not available in eight patients in the placebo (PBO) group and six patients in the BMAX arm.

²Tumor stage data was not available in one patient in both the PBO group and the BMAX treatment group.

³Statistical significance was determined by t-test or Fisher’s exact test as appropriate.

⁴Includes hypertension, coronary artery disease, atrial fibrillation, aortic valve disease

T2: cancer is completely inside the prostate gland; T3a: cancer has broken through the capsule of the prostate gland; T3b: cancer has spread into the seminal vesicles; N0: cancer has not spread to any lymph nodes; N1: cancer has spread to nearby lymph nodes in the pelvis; and Mx: metastasis; BPH, benign prostate hyperplasia.

We performed an expanded chart review to capture baseline metabolic, hepatic, cardiovascular, and prostate-relevant comorbidities that may influence lipid metabolism and fatty acid synthesis. Baseline imbalances were observed for hyperlipidemia, hypertriglyceridemia, and chronic liver disease (Table 1), which are relevant to lipid metabolism and were therefore considered in the interpretation of serum TFFA and malonyl-CoA findings. Additional metabolic covariates such as body mass index, medication use (e.g., statins and antidiabetic agents), and fasting glucose would have further strengthened interpretation. These variables were not consistently available across all participants and thus could not be reliably incorporated. We did not collect data on ethnicity (e.g., Latino or Asian). We relied on self-reporting by the patient about any side effects. Other than bloating and resolvable minor gastric issues, Grade 3 or 4 adverse events were not experienced by any patient, which is consistent with prior clinical trials of SFN in prostate cancer patients (31,32).

BMAX intake downregulated c-Myc, ACC1, FASN, and Ki-67 expression in the prostate tumor of men

Figure 4A shows immunohistochemistry images of these proteins in normal tissue adjacent to tumor (NAT). Representative H&E staining and immunohistochemistry images for c-Myc, ACC1, FASN, and Ki-67 in NAT and prostate tumor in three patients of the PBO group and three patients of the BMAX group are shown in Supplementary Figure S1, Supplementary Figure S2, Supplementary Figure S3, and Supplementary Figure S4, respectively. Expression of c-Myc was lower by 38% in the NAT of the BMAX group compared to PBO group, but the difference did not reach statistical significance (Figure 4B). Expression of ACC1, FASN, and Ki-67 did not differ between PBO group and the BMAX group in the NAT (Figure 4B). Figure 4C shows immunohistochemistry images of these proteins in the prostate tumor tissue. Expression of c-Myc, ACC1, FASN, and Ki-67 proteins were lower by about 41%, 13%, 13%, and 32%, respectively, in the BMAX group when compared to PBO group (Figure 4D).

Figure 4:

Effect of oral BroccoMax^® (BMAX) administration on protein expression of fatty acid synthesis pathway and Ki-67 in prostate cancer patients. (A) Representative immunohistochemical images for c-Myc, acetyl-CoA carboxylase 1 (ACC1), fatty acid synthase (FASN), and Ki-67 expression in normal tissue adjacent to tumor (NAT) of prostate cancer patients who received placebo (PBO) or BMAX treatment. (B) Quantitation of fluorescence intensities of proteins shown in panel A using ImageJ software. Results shown are mean ± 95% CI (n= 12 for PBO and n= 11 for BMAX groups). (C) Representative immunohistochemical images for c-Myc, ACC1, FASN, and Ki-67 expression in prostate tumor of patients who received PBO or BMAX treatment. (D) Quantitation of fluorescence intensities of proteins shown in panel C using ImageJ software. Results shown are mean ± 95% CI (n= 12 for PBO and n= 11 for BMAX groups). Statistical significance of difference was determined by two-sided t-test.

The primary end point of a decrease in serum TFFA level was not achieved

Serum levels of citrate, acetyl-CoA, malonyl-CoA, and TFFA were comparable in the PBO group and the BMAX groups (Figure 5A). Prostate tumor levels of citrate, acetyl-CoA, and TFFA did not differ significantly between PBO and BMAX groups (Figure 5B). We also performed analysis of individual fatty acids using serum. The levels of individual fatty acids did not differ significantly between PBO and BMAX groups (Supplementary Figure S5). However, the prostate tumor level of malonyl-CoA was decreased by about 41% (P= 0.049) after BMAX^® administration (Figure 5B). A possibility of accelerated fatty acid breakdown may explain these results. It is important to point out that TFFA levels can fluctuate in response to diet, fasting status, insulin dynamics, circadian rhythm, and physical activity. These details were not collected from the patients. The patients were not asked to maintain a food intake diary.

Figure 5:

Effect of oral BroccoMax^® (BMAX) administration on fatty acid synthesis intermediates and total free fatty acids (TFFA) in prostate cancer patients. Serum (A) and prostate tumor (B) levels of citrate, acetyl-CoA, malonyl-CoA, and TFFA. Results shown are mean ± 95% CI (serum- n= 22 for PBO and n= 19 for BMAX groups; prostate tumor- n= 11 for each group). Statistical significance of difference was determined by two-sided t-test.

Discussion

We have shown previously that SFN administration inhibits prostate cancer progression in Transgenic Adenocarcinoma of the Mouse Prostate (TRAMP) mice (33). Prostate cancer in TRAMP mice is driven by SV40 T-antigen mediated inactivation of p53 and Rb genes, which are frequently inactivated in human prostate cancers (34,35). We raised the question of whether suppression of fatty acid synthesis by SFN treatment is unique to the TRAMP model. The present study reveals that inhibition of prostate cancer in Hi-Myc mice, in which prostate cancer is driven by prostate-specific expression of Myc (36), by oral administration of SFN is also accompanied by downregulation of fatty acid synthesis enzyme proteins ACC1 and FASN leading to a decrease in plasma levels of TFFA. One can question clinical significance of a modest decrease in plasma TFFA level after SFN treatment. However, even a modest decrease in circulating level of a metabolite can be clinically significant. For example, serum cholesterol level of 240–300 mg/dL is considered high and a risk factor for cardiovascular disease, but a serum cholesterol level of 199 mg/dL is considered normal (about 17% and 34% lower from 240–300 mg/dL level).

We have shown previously that c-Myc is a direct regulator of ACC1 and FASN expression in prostate cancer cells in vitro and in vivo (8) and SFN downregulates c-Myc expression in prostate cancer cells (37). These results may partly explain the mechanism underlying suppression of fatty acid synthesis by SFN treatment. However, there are other mechanistic possibilities deserving attention in future studies. Adenosine monophosphate-activated protein kinase is a master regulator and sensor of cellular energy and a regulator of synthesis of fatty acids (38). Activation of nuclear factor erythroid 2-related factor 2 by sulforaphane was shown to inhibit high glucose-induced development of pancreatic cancer via adenosine monophosphate-activated protein kinase-dependent signaling (39). Sterol regulatory element-binding protein 1 is also involved in regulation of lipid synthesis (40). SFN was shown to suppress sterol regulatory element-binding protein 1 in human prostate cancer cells (26).

Clinical trials have been conducted using SFN-rich broccoli sprout extract (SFN-BSE), BMAX^® or free stabilized SFN (Prostaphane^®). Preclinical studies showed downregulation of AR expression after SFN treatment in human prostate cancer cells (41,42). These studies led to two clinical trials to determine if SFN-BSE or Prostaphane^® suppresses PSA, a downstream target of regulation by AR. In one study, prostate cancer patients (n=20) were treated with 200 μmoles of SFN-BSE for 20 weeks (32). The primary end point of a ≥50% PSA decline was observed only in one patient (32). However, there was a significant increase in PSA doubling time after SFN-BSE administration (32). Daily oral administration of 60 mg free stabilized SFN (Prostaphane^®) for 6 months to prostate cancer patients with biochemical recurrence (n=78) resulted in a significant lengthening of the PSA doubling time compared to the PBO group (31). However, the primary end point of a 0.012 log (ng/mL)/month decrease in the log PSA slope was not met (31). A modest decrease in plasma fatty acids was observed within 48 hours of treatment with 200 μmol of SFN-BSE in 10 normal healthy subjects (43). In our clinical trial, we found a statistically significant decrease in expression of c-Myc, ACC1, FASN, and Ki-67 expression in the prostate of BMAX group compared to the PBO group. Despite these encouraging changes, the serum or prostate tumor level of TFFA or serum level of individual fatty acids was not decreased.

The mouse data shows strong metabolic and oncogenic suppression, whereas the clinical trial shows modest protein-expression changes without systemic metabolic effects. Several factors could contribute to these differences, including species-specific pharmacokinetics, tumor heterogeneity, differences in exposure, and limitations of nutraceutical-grade SFN. A longer duration treatment might be required to observe suppression of TFFA. A treatment duration of >4 weeks is not possible for a prostatectomy trial because the patients may not agree to delaying surgery >4 weeks; and (b) patients with early-stage prostate cancer T1/T2 may be better suited for this study because majority of the patients in our trial were T3 or higher. Alternatively, higher doses of BMAX^® might be required to suppress fatty acid synthesis in prostate cancer patients.

A limitation of our study is that we did not measure prostate tumor levels of SFN metabolites. However, pharmacokinetics and prostate tissue distribution of SFN metabolites have been studied in mice (44). SFN metabolites were detectable in mouse prostate after 2 hours and 6 hours of oral administration of 5 μmole and 20 μmole of SFN (44). Residual amounts of SFN metabolites persisted for 24 hours in the mouse prostate after oral SFN treatment (44). Prostate tumor tissue bioavailability of SFN metabolites has also been demonstrated in patients (45). It is important to point out that preclinical SFN concentrations often exceed what is achievable in humans. There were other limitations of our study. For example, we could not enroll >1 African American patients. Latino patients were not available either for enrollment. Unmeasured dietary and medication confounders also limit interpretation of lipid end points. These confounders must be considered in future clinical trials of SFN.

Prevention relevance:

Fatty acid synthesis is a valid target for prevention of human prostate cancer. In this study, we determined the feasibility of fatty acid synthesis inhibition using sulforaphane in a mouse model and BroccoMax^®, its clinical grade formulation, in prostatectomy patients

Abbreviations

ACC1

acetyl-CoA carboxylase 1

AR

androgen receptor

BMAX

BroccoMax^®

FASN

fatty acid synthase

NAT

normal tissue adjacent to tumor

PBO

placebo group

PBS

phosphate-buffered saline

PAC

prostate adenocarcinoma

SFN

sulforaphane

SFN-BSE

sulforaphane-rich broccoli sprout extract

TFFA

total free fatty acids

TMA

tissue microarray

TRAMP

Transgenic Adenocarcinoma of the Mouse Prostate