Clinical localization of primary tumors and sites of metastasis by PET is based on the enhanced cellular uptake of 2-deoxy-2-[18F]-fluoro-D-glucose (FDG). In prostate cancer, however, PET-FDG imaging has shown limited clinical applicability, suggesting that prostate cancer cells may utilize hexoses other than glucose, such as fructose, as the preferred energy source. Our previous studies suggested that prostate cancer cells overexpress fructose transporters, but not glucose transporters, compared with benign cells. Here, we focused on validating the functional expression of fructose transporters and determining whether fructose can modulate the biology of prostate cancer cells in vitro and in vivo. Fructose transporters, Glut5 and Glut9, were significantly upregulated in clinical specimens of prostate cancer when compared with their benign counterparts. Fructose levels in the serum of patients with prostate cancer were significantly higher than healthy subjects. Functional expression of fructose transporters was confirmed in prostate cancer cell lines. A detailed kinetic characterization indicated that Glut5 represents the main functional contributor in mediating fructose transport in prostate cancer cells. Fructose stimulated proliferation and invasion of prostate cancer cells in vitro. In addition, dietary fructose increased the growth of prostate cancer cell line–derived xenograft tumors and promoted prostate cancer cell proliferation in patient-derived xenografts. Gene set enrichment analysis confirmed that fructose stimulation enriched for proliferation-related pathways in prostate cancer cells. These results demonstrate that fructose promotes prostate cancer cell growth and aggressiveness in vitro and in vivo and may represent an alternative energy source for prostate cancer cells.

Significance:

This study identifies increased expression of fructose transporters in prostate cancer and demonstrates a role for fructose as a key metabolic substrate supporting prostate cancer cells, revealing potential therapeutic targets and biomarkers.

Prostate cancer represents the second leading cause of cancer-related deaths in adult men in the United States (1). The most characterized risk factors for prostate cancer are advanced age, black race/ethnicity, and family history of the disease. Other modifiable risk factors, such as lifestyle, have been under study in recent years. Western diet has been linked to an increase in incidence of pathologies, such as obesity, metabolic syndrome, and cancer. Sweeteners have become one of the principal sources of carbohydrates in the Western diet and can be found in processed foods, beverages, and fruit juices. High fructose corn syrup (HFCS, 10%–53% glucose and 42%–90% fructose) is one of the most used sweeteners in the food industry and represents more than 40% of total sweeteners consumed in the United States (2). The idea that dietary fructose is metabolized almost completely in the liver has been challenged. It has been reported that after consumption of beverages containing HFCS, fructose can enter and reach a steady level in the blood stream and maintain it for hours (3).

Hexose uptake in human cells is primarily mediated by the glucose transporter family (Gluts), from which, 14 members (Glut1 to Glut14) have been described previously (4). Gluts exhibit both tissue specificity and multi-functional transport capacity (4). Glut1 is the most ubiquitous isoform and it transports glucose and galactose. Fructose can be transported by Glut2 [Michaelis constant (Km), 76 mmol/l), Glut5 (Km, 6 mmol/l), Glut9 (Km, 0.42 mmol/l), and Glut11 (Km, 0.06 mmol/L), with Glut5 being an obligate fructose transporter (4).

Activation of Glut gene expression, specifically Glut1, increases glucose transport/metabolism and is a molecular feature of the malignant phenotype in cancer. The high level of glucose uptake represents the biological basis for the clinical localization of primary cancers and sites of metastasis using PET, based on the use of 2-deoxy-2-[18F]-fluoro-D-glucose (FDG). Unlike most human neoplasms, PET-FDG has shown limited clinical applicability in prostate cancer (5, 6), which suggests: (i) a low metabolic activity of prostate cancer cells or (ii) prostate cancer cells use hexoses other than glucose, such as fructose, as an energy source to maintain their metabolic requirements.

Epidemiologic studies (7, 8) have established a positive correlation between the incidence/progression of prostate cancer and Western diet. A recent study (9) indicated a direct relationship between the increase in the consumption of dietary sugar, particularly HFCS, and a higher risk of developing symptomatic prostate cancer. Fructose has been involved in accelerating pancreatic and breast cancer cell proliferation and incidence of metastasis, respectively (10, 11). In addition, fructose enhanced tumor growth and grade in a murine model of intestinal cancer (12). The characteristics observed in other human neoplasms could be exacerbated in prostate cancer because of low level of Glut1 and the demonstration that PET-FDG has limited clinical applicability for imaging/detection of prostate cancer, which reflects the lack of functional glucose transporters.

Blood samples and clinical specimens

Serum samples were collected from venous blood of 65 healthy individuals and 65 patients diagnosed with prostate cancer. All healthy subjects did not have any background related to prostate cancer and had PSA levels significantly lower than patients with prostate cancer. Prostate cancer diagnosed patients were determined by needle transrectal biopsies. Human benign and malignant tissue sections were obtained commercially at Roswell Park Comprehensive Cancer Center (Buffalo, NY) through deidentified clinical specimens of patients with prostate cancer. For patient-derived xenograft (PDX) analysis, surgically resected prostate tissue was collected from patients with prostate cancer during radical prostatectomy. Tissue specimens were collected in the operating room, submerged immediately in ice-cold ViaSpan Solution (Barr Laboratories Inc.), and transported on ice for transplantation. All blood and tissue specimens were collected after obtaining a written informed consent from the patients. All studies were conducted in accordance with recognized international ethical guidelines (Declaration of Helsinki, CIOMS, and Belmont Report) and approved by the Institutional Review Board at Pontificia Universidad Católica de Chile (Santiago, Chile, PUC).

IHC

Antigen retrieval was performed using 0.01 mol/L sodium citrate buffer pH 6 for 30 minutes. Endogenous peroxidase activity was inhibited by using 3% H2O2 in methanol and nonspecific binding of antibodies was blocked with 2% BSA for 20 minutes. Tissue sections were incubated with anti-Glut1, -Glut2, -Glut5, -Glut7, -Glut9, or -Glut11 (1:100–1:1,000 dilution, Alpha Diagnostic International) and anti-Ki-67 (1:500 dilution, Abcam) antibody for 12 hours. Tissue sections were then incubated with a horseradish peroxidase (HRP)-conjugated secondary antibody (1:100, Dako) for 1 hour at room temperature. Peroxidase activity was developed using 3,3-Diaminobenzidine (Dako) and H2O2. Hematoxylin was used to counterstain nuclei. Slides were dehydrated through graded alcohol to xylene and mounted with coverslips. For human clinical specimens, quantitation of the immunostaining was performed in 10 images from each specimen, which were scored blindly by three independent observers who scored the immunostaining on a scale ranging from 0 (no staining) to 3 (strong staining) in 100 epithelial cells from each image. For tumor tissues obtained from cell line–derived and PDXs, immunostaining was quantified using ImageJ 1.52k software (NIH) and reported as integrated density or percentage of positive cells. The results were averaged from 10 image fields for each sample.

Fructose and glucose determination

Serum fructose and glucose levels from patients were measured using a commercially available kit (Abcam). Free fructose was enzymatically converted to β-glucose, and then converted to a product that reacts with OxiRed probe to emit fluorescence (excitation/emission = 535/587 nm). Concentrations of fructose and glucose were calculated by extrapolating the sample readings from a standard curve.

Cell culture

All human prostate cancer cell lines, LNCaP (androgen sensitive) and PC3 (metastatic), and the benign human prostate epithelial cell line, RWPE-1, were commercially obtained from the ATCC and used for no more than 20 passages. Cells were maintained at 37°C in a humidified atmosphere of 5% CO2 and cultured in RPMI1640 Medium (Gibco-Life Technologies) supplemented with 10% FBS (Gibco-Life Technologies) and 1% penicillin/streptomycin. All cell lines were tested every 2 weeks for the presence of Mycoplasma using the EZ-PCR Mycoplasma Test Kit (Biological Industries).

Western blotting

Cells were lysed in RIPA lysis buffer [50 mmol/L Tris (pH 8), 1% Triton x-100, 0.1% SDS, 150 mmol/L NaCl, 5 mmol/L EDTA (pH 8), and 0.5% sodium deoxycholate] supplemented with 1× protease inhibitor cocktail and separated by centrifugation. Protein (50 μg, determined by the Lowry method) was subjected to electrophoresis in 10% SDS-PAGE. The proteins were transferred onto a nitrocellulose membrane and blocked with 5% BSA in 1× TBS-Tween. The blot was incubated overnight with anti-Glut1, -Glut2, -Glut5, -Glut7, -Glut9, and -Glut11 (1:500–1:1,000 dilution, Alpha Diagnostic International), -tubulin (1:3,000 dilution, Santa Cruz Biotechnology), and -GAPDH (1:1,000, Santa Cruz Biotechnology) antibody. Membranes were then incubated with an HRP-conjugated secondary antibody (1:500, Dako) for 1 hour at room temperature. The reaction was revealed by the ECL Western Blot Analysis System Kit (Thermo Scientific Pierce).

Transport assays

Cell lines were incubated at 37°C in incubation buffer containing 0.1 mCi/mL of D-[U-14C]-fructose or 1 mCi/mL of 2-[1,2–3H]-deoxy-D-[3H] glucose (American Radiolabeled Chemicals). To establish the initial velocity of fructose transport in LNCaP and PC3 prostate cancer cell lines, time-course (2–60 seconds) analyses were performed using the radioactive tracer D-[U-14C]-fructose. Subsequently, based on the initial velocity data, dose–response (Michaelis–Menten) analyses were developed by incubation of prostate cancer cells with increasing concentrations (0.01–50 mmol/L) of radioactive fructose under initial velocity conditions (6 seconds). Transport reactions were stopped with ice-cod stopping solution (1× PBS containing 10 mol/L HgCl2). Cells were washed twice with stopping solution and radioactivity was determined by scintillation counting. The kinetic parameters (apparent Km and Vmax) associated with fructose transport were determined using the Michaelis–Menten equation and the double reciprocal (Lineweaver–Burk) plot.

Proliferation rate determination

Cells were seeded in 12-well plates (4,000 cells/well) in RPMI1640 medium supplemented with 10% FBS. After that, cells were incubated with glucose-free RPMI1640 Medium (Gibco-Life Technologies) supplemented with 5 mmol/L glucose, fructose, or galactose. Cell growth was measured every day for a total of 7–10 days by direct counting of cell number using the trypan blue exclusion assay.

Immunofluorescence

Cells were seeded on a glass coverslip (12 mm) into a 24-well plate in RPMI1640 medium supplemented with 10% FBS. After that, cells were incubated with glucose-free RPMI1640 medium supplemented with 5 mmol/L glucose or 5 mmol/L fructose for 6 days and cells were fixed with 4% paraformaldehyde for 30 minutes at room temperature. Cell permeabilization was performed using 0.1% Triton X-100 in Tris-HCl pH 7.8 for 15 minutes. Subsequently, cells were incubated overnight with anti-Ki-67 primary antibody (1:500 dilution, Abcam) at room temperature. An Alexa 488–conjugated anti-rabbit IgG (1:100, Molecular Probes) was used as a secondary antibody. Nuclei were visualized using 4′, 6′-diamidino-2-phenylindole diluted in Tris-HCl pH 7.8 (1:5,000) for 5 minutes at room temperature. Images were acquired using a Fluorescence Microscope (Zeiss, Axio Scope.A1).

Migration and invasion assays

Migration and invasion capacities were measured using Cell Biolabs Cytoselect 96-well Cell Migration and Invasion Assay (Cell Biolabs, Inc.). LNCaP and PC3 cells were cultured in 100 mm plates until 50%–60% confluence and then treated with 5 mmol/L glucose or 5 mmol/L fructose in serum-free RPMI1640 media for 48 hours. After treatment, 5 × 104 (for migration assays) or 2 × 105 (for invasion assays) cells were seeded on the top chamber of each transwell assay in serum-free RPMI1640 medium supplemented with 5 mmol/L glucose or 5 mmol/L fructose, while bottom chambers were filled with RPMI1640 medium supplemented with 5 mmol/L glucose or 5 mmol/L fructose and 10% FBS to create a chemotaxis gradient. Cells were incubated for 24 hours (for migration) or 48 hours (for invasion) at 37°C, and their migration/invasion capacities were assessed by quantifying the number of cells that migrate or invade into the bottom chamber using a fluorescence probe provided by the kit and a Fluorescence Multiplate Reader Synergy2 (BioTek; excitation/emission = 480/520 nm). Migration or invasion values were reported as the ratio between the signal of cells treated with fructose compared with glucose as mean ± SD of triplicate biological samples. Analyses using the Synergy2 reader were performed with the technical support of the Advanced Microscopy Facility (UMA) from PUC (Santiago, Chile).

siRNA analysis

The PC3 prostate cancer cell line was transiently transfected with targeting human Glut5 and nontargeting siRNAs (75 nmol/L, Santa Cruz Biotechnology) using TransIT-siQUEST Transfection Reagent (Mirus Bio LLC) according to the manufacturer's instructions. Downregulation of the expression of Glut5 in the PC3 prostate cancer cell line was confirmed using Western blot analyses 72 hours after transfection.

Xenograft models

Six-week-old male NOD-scid IL2Rgammanull(NSG) mice (The Jackson Laboratories) were separated into three experimental groups (n = 8/group): (i) a control group, which received regular diet and normal water, (ii) a glucose-treated group, which received regular diet and water supplemented with 15% w/v glucose, and (iii) a fructose-treated group, which received regular diet and water supplemented with 15% w/v fructose. For cell line–derived xenografts, 2 × 106 PC3 cells were injected into the flank of male NSG mice. Tumor volume was measured for up to 8 weeks or until tumor reached a volume of 2 cm3 using a caliper and the tumors were weighed upon excision. For PDXs, prostate cancer tissue specimens were cut into wedge-shaped pieces of 2 mm3. Four to six pieces of tissue were dipped into Matrigel and transplanted individually into the subcutaneous compartment through a 10-gauge trocar device into the flank of male NSG mice. Twenty-one days postimplantation, the pieces of tissue were surgically removed and processed for histologic analyses. All animal experimentation was conducted with the approval and under the supervision of the Institutional Animal Care and Use Committee of PUC (Santiago, Chile).

RNA sequencing for transcriptome analysis

Sequencing libraries were prepared with the RNA HyperPrep Kit with RiboErase (HMR) Kit (Roche Sequencing Solutions), from 500 ng total RNA obtained using the PureLink RNA Mini Kit-Ambion (Invitrogen) following ribosomal depletion. The libraries were quantitated using KAPA Biosystems qPCR Kit, and were pooled together in an equimolar fashion, following experimental design criteria. Each pool was denatured and diluted to 350 pmol/L with 1% PhiX control library added. The resulting pool was loaded into the appropriate NovaSeq reagent cartridge for 100-cycle paired end sequencing, performed on a NovaSeq6000 per the manufacturer's protocol (Illumina Inc.).

For the bioinformatics analysis, raw reads that passed quality filter from Illumina RTA were mapped to the hg38 human reference genomes and the corresponding GENCODE (v25) annotation databases using STAR. The mapped bam files were further QCed using RSeQC, a quality control toolbox for RNA sequencing (RNA-seq) data, to identify potential RNA-seq library preparation problems. From the mapping results, the read counts for genes were obtained by featureCounts from Subread. Data normalization and differential expression analysis were preformed using DESeq2, a variance analysis package developed to infer the statistically significant difference in RNA-seq data. Genes with a Padj value of less than 0.05 and a fold change of greater than 1.5 in expression were considered differentially expressed. Pathway analysis was done by gene set enrichment analysis (GSEA) preranked mode using ranked gene list based on test statistics from differential expression analysis against MSigDB's (v7.2) gene sets: hallmark (h), curated (c2.cp), and selected ontology (c5).

Statistical analysis

All statistical data analysis was performed using GraphPad Prism 8.0 through unpaired Student t test, one-way ANOVA, or two-way ANOVA as was appropriate. P < 0.05 was set as threshold for significant results. Linear and nonlinear regressions were calculated using GraphPad Prism 8.0 software (only correlations >0.9 were accepted).

The fructose transporters, Glut5 and Glut9, are overexpressed in human prostate cancer specimens

Immunostaining analyses of Glut1, Glut2, Glut5, Glut7, Glut9, and Glut11 were performed in 25 matched clinical specimens of nontumor and tumor human prostate tissues. Glut1 was not detected in nontumor and tumor tissues (Fig. 1A and B). Glut5 and Glut9 were significantly overexpressed in tumor tissues when compared with nontumor tissues (Fig. 1A and B). Glut2 and Glut7 were undetected or detected at very low levels in nontumor and tumor tissues (Fig. 1A and B). For Glut11, although not statistically significant, we observed an increased expression in tumor tissues compared with nontumor tissues. These results indicated that Glut5, Glut9, and probably Glut11, could have a potential role to mediate fructose uptake in benign and malignant human prostate epithelial cells. These results confirmed our previous research (13), indicating that Glut1 was not detected, or detected at very low levels in prostate cancer. Although expression of Glut1 has been studied to some extent in human clinical specimens of prostate cancer, until now the presence/functionality of this transporter remains controversial. An IHC (14) analysis of Glut1 in 195 clinical specimens of malignant prostate tissues indicated that 47% of the specimens showed positive immunostaining and 53% were negative. Among positive cases, Glut1 immunostaining showed mostly cytoplasmic localization. A similar pattern of expression of Glut1 was reported by Xiao and colleagues (15) and by Gasinska and colleagues (16). Low levels of expression, or cytoplasmic/intracellular localization of Glut1 could partially explain the limited clinical applicability of PET-FDG for imaging prostate cancer. Together, this evidence suggests that uptake of fructose through a mixture of different Gluts could be a favored mechanism to fuel metabolic demands in prostate cancer cells.

Figure 1.

Glut(s) expression in benign and malignant human prostate tissues. A, IHC analyses of Glut(s) in 25 matched clinical specimens of benign and malignant human prostate tissues. Human brain (Glut1), liver (Glut2), and kidney (Glut5, 7, 9, and 11) tissues were utilized as positive controls. Black bars, 50 μm. B, Average score of the immunostaining analysis. C, Glucose and fructose serum determination in benign patients and patients with prostate cancer. ns, not statistically significant. , P < 0.05 (t test).

Figure 1.

Glut(s) expression in benign and malignant human prostate tissues. A, IHC analyses of Glut(s) in 25 matched clinical specimens of benign and malignant human prostate tissues. Human brain (Glut1), liver (Glut2), and kidney (Glut5, 7, 9, and 11) tissues were utilized as positive controls. Black bars, 50 μm. B, Average score of the immunostaining analysis. C, Glucose and fructose serum determination in benign patients and patients with prostate cancer. ns, not statistically significant. , P < 0.05 (t test).

Close modal

We examined fasting serum fructose and glucose levels in 65 patients with nondiabetic benign and 65 patients with nondiabetic prostate cancer (Fig. 1C; Supplementary Table S1). Clinical parameters, like age and PSA levels, were analyzed in benign patients and in patients with prostate cancer, while Gleason score was analyzed only in patients with prostate cancer (Supplementary Table S1). Gleason scores were consistent with patients with prostate cancer and PSA levels were significantly higher (P = 0.0448) in patients with prostate cancer compared with benign patients. Fructose, but not glucose, levels were significantly higher (P = 0.0030) in patients with prostate cancer compared with benign patients (Fig. 1C). High levels of fructose were also reported in the serum of patients with pancreatic cancer (17). Even though the biological significance of this observation is not fully understood, it indicates that that the relationship between fructose metabolism and development/progression of cancer might be broader than expected. Although the patients selected for this analysis were not diabetic, we cannot completely rule out the possibility of alterations in liver and/or kidney function and the possible effect of diet, among other parameters that could have a significant impact on the global and/or specific metabolism of hexoses. Therefore, more studies are required to validate the presence of fructose in the blood as a potential biomarker of malignancy of the prostate or any other tissue.

Glut5 represents the main functional contributor to the fructose transport in prostate cancer cells

Levels of expression of Glut1, Glut2, Glut5, Glut7, Glut9, and Glut11 were analyzed in benign (RWPE-1) and malignant (LNCaP and PC3) human prostate cell lines (Fig. 2A). Glut5 and Glut9 were overexpressed in LNCaP and PC3 cell lines when compared with RWPE-1 cell line (Fig. 2A). Glut2 was not detected in benign and malignant human prostate cell lines, while Glut7 and Glut11 were detected at similar levels of expression (Fig. 2A). Glut1 was highly expressed in benign and malignant human prostate cell lines at similar level of expression (Fig. 2A). This observation is consistent with the cell immortalization process being intrinsically associated with reexpression of isoform Glut1 (18). Cell immortalization process together with selective culture conditions (high glucose concentration in the culture medium) could at least partially explain the high level of expression of Glut1 observed in prostate cancer cell lines when compared with cancer epithelial cells from human prostate tissue.

Figure 2.

Glut(s) expression and functionality in benign and malignant human prostate cell lines. A, Western blot analyses for the expression of Glut(s) in benign (RWPE-1) and malignant (LNCaP and PC3) human prostate cell lines. Erythrocytes membranes (Glut1), Caco-2 (Glut2), and HEK293 (Glut5, Glut7, Glut9, and Glut11) cells were utilized as positive controls. Tubulin was used as loading control. B and C, Radiolabeled glucose and fructose uptake analyses at 0 and 30 seconds in RPWE-1, LNCaP, and PC3 cell lines. D, Uptake (0–8 seconds), Michaelis–Menten, and Lineweaver–Burk analyses for the transport of fructose in LNCaP and PC3 cell lines. 1/V, 1/transport velocity; 1/S, 1/substrate concentration; Km, Michaelis–Menten constant; ns, not statistically significant. , P < 0.05 (one-way ANOVA).

Figure 2.

Glut(s) expression and functionality in benign and malignant human prostate cell lines. A, Western blot analyses for the expression of Glut(s) in benign (RWPE-1) and malignant (LNCaP and PC3) human prostate cell lines. Erythrocytes membranes (Glut1), Caco-2 (Glut2), and HEK293 (Glut5, Glut7, Glut9, and Glut11) cells were utilized as positive controls. Tubulin was used as loading control. B and C, Radiolabeled glucose and fructose uptake analyses at 0 and 30 seconds in RPWE-1, LNCaP, and PC3 cell lines. D, Uptake (0–8 seconds), Michaelis–Menten, and Lineweaver–Burk analyses for the transport of fructose in LNCaP and PC3 cell lines. 1/V, 1/transport velocity; 1/S, 1/substrate concentration; Km, Michaelis–Menten constant; ns, not statistically significant. , P < 0.05 (one-way ANOVA).

Close modal

To validate in vitro the functionality of fructose and glucose transporters in benign and malignant human prostate cell lines, we performed basic uptake analyses of D-[U-14C]-fructose or 2-[1,2–3H]-deoxy-D-[3H] glucose (Fig. 2B and C). Fructose uptake was significantly higher in LNCaP and PC3 cells when compared with RWPE-1 cells (Fig. 2C), reaching a maximum fructose uptake of approximately 40 pmol/106 cells in LNCaP and approximately 50 pmol/106 cells in PC3 cells compared with a maximum of approximately 5 pmol/106 cells in RWPE-1 cells. Therefore, prostate cancer cell lines showed a 10-fold higher fructose uptake capacity when compared with benign cells. Consistently with the high level of expression of Glut1, glucose uptake reached similar levels (∼2–3 pmol/106 cells) in benign and malignant human prostate cell lines (Fig. 2B).

To functionally identify the kinetic component(s) associated with the transport of fructose in prostate cancer cells and determine the contribution of each component to the total fructose transport capacity, we performed time–course and Michaelis–Menten analyses in LNCaP and PC3 cell lines (Fig. 2D). Time–course analyses indicated that LNCaP and PC3 cells take up fructose at a constant rate (initial velocity) of 1.11 and 1.62 pmol/106 cells/second for approximately 10 seconds, respectively (Fig. 2D, uptake 0 to 8 seconds). Michaelis–Menten analyses were performed under initial velocity conditions. For LNCaP cells, fructose transport approached saturation at approximately 240 pmol/106 cells/minute (Fig. 2D, LNCaP, Michaelis–Menten). Linear transformation of the transport data through the Lineweaver–Burk plot indicated the presence of a single kinetic component with an apparent Km of 6.8 mmol/L. For PC3 cells, fructose transport approached saturation at approximately 80 pmol/106 cells/minute (Fig. 2D, PC3, Michaelis–Menten). Linear transformation of the transport data through the Lineweaver–Burk plot indicated the presence of a single kinetic component with an apparent Km of 7.1 mmol/L. The apparent Km for the transport of fructose in LNCaP (6.8 mmol/L) and PC3 (7.1 mmol/L) cell lines was fully compatible with the isoform Glut5 (Km, ∼6 mmol/L), which represented the highest functional contributor in prostate cancer cells.

Fructose supports proliferative rate and invasion capacity in prostate cancer cells

Growth curve analyses showed that, in RWPE-1 cell line, glucose was more effective than fructose or galactose to stimulate proliferation (Fig. 3A). In prostate cancer cells, however, fructose stimulated proliferation at comparable levels with glucose (Fig. 3A). Galactose was less effective than fructose or glucose to stimulate prostate cancer cell proliferation (Fig. 3A), which indicates that prostate cancer cells may be preferentially sensitive to either fructose or glucose in vitro. This observation was confirmed using an immunofluorescence analysis of Ki-67 in prostate cancer cells (Fig. 3B). Our immunostaining data showed no significant differences in Ki-67 expression in LNCaP and PC3 cell lines in the presence of glucose or fructose (Fig. 3B). In addition, we analyzed the effect of fructose in comparison with glucose on migration and invasion capacities. Fructose or glucose was maintained in both the top and bottom chambers. FBS was utilized as a chemoattractant (Fig. 3C and D, schematic representations). While migration capacity (Fig. 3C) was not altered by fructose, invasion capacity (Fig. 3D) was significantly enhanced in the presence of fructose when compared with glucose in both prostate cancer cell lines. Previous reports indicated that invasion capacity of prostate cancer cells can be suppressed by downregulating GTPases, like Rab7 (19) or Rab23 (20), which are required for the transport of proteases-containing vesicles. An increase in GTP pools could lead to increased activity of these GTPases, and consequently, an enhanced invasion capacity of cancer cells. Fructose metabolism has been proven to stimulate nucleotide biosynthesis in pancreatic cancer (10). In addition, fructose administration in the drinking water can enhance metastatic sites in breast cancer (11). Together, our results suggest that fructose transport/metabolism can promote invasion capacity/metastatic stage of prostate cancer cells probably through an increase in GTP pools. Further studies are needed to prove this hypothesis.

Figure 3.

In vitro effects of fructose on proliferation, migration, and invasion capacities in prostate cancer cells. A, Growth curves of RWPE-1, LNCaP, and PC3 cell lines exposed to 5 mmol/L fructose, glucose, or galactose. B, Ki-67 immunofluorescence in LNCaP and PC3 cells exposed to 5 mmol/L fructose or glucose (6 days of incubation). Red (glucose) and blue (fructose) bars represent quantitation of the immunostaining analyses in each cell line. Migration (C) and invasion (D) capacities were measured using the CytoSelect migration and invasion assay. Glucose or fructose treatment (s, sugars) was applied to top and bottom chambers. C and D, Schematic representations. The number of migrating/invading cells was quantified in the bottom chamber using a fluorescence dye in a Synergy plate reader (P < 0.001, t test). ns, not statistically significant (t test). *, P < 0.05. PCa, prostate cancer.

Figure 3.

In vitro effects of fructose on proliferation, migration, and invasion capacities in prostate cancer cells. A, Growth curves of RWPE-1, LNCaP, and PC3 cell lines exposed to 5 mmol/L fructose, glucose, or galactose. B, Ki-67 immunofluorescence in LNCaP and PC3 cells exposed to 5 mmol/L fructose or glucose (6 days of incubation). Red (glucose) and blue (fructose) bars represent quantitation of the immunostaining analyses in each cell line. Migration (C) and invasion (D) capacities were measured using the CytoSelect migration and invasion assay. Glucose or fructose treatment (s, sugars) was applied to top and bottom chambers. C and D, Schematic representations. The number of migrating/invading cells was quantified in the bottom chamber using a fluorescence dye in a Synergy plate reader (P < 0.001, t test). ns, not statistically significant (t test). *, P < 0.05. PCa, prostate cancer.

Close modal

To provide a proof of concept that transport of fructose was required to promote the biological impact of this hexose on prostate cancer cells, PC3 cells were transiently transfected with a specific Glut5–targeting siRNA (Supplementary Fig. S1A and S1B). Downregulation of the expression of Glut5 in PC3 cells significantly reduced (∼50%) fructose-stimulated cell proliferation in vitro, an effect that was not observed in wild-type PC3 cells or in PC3 cells transiently transfected with a nontargeting siRNA (Supplementary Fig. S1C and S1D).

Dietary fructose boosts tumor growth/proliferation of human prostate cancer xenograft models and promotes proliferation/cell cycle and carcinogenesis/progression pathways in vitro in prostate cancer cells

The effect of fructose on in vivo tumor growth/prostate cancer cell proliferation was evaluated by injecting PC3 cells or transplanting human prostate cancer tissues subcutaneously into the flank of NSG mice. For PC3 cell line–derived xenografts, tumor volume was monitored weekly for 8 weeks (Fig. 4A). Four weeks after cell injection, differences between fructose and control condition were observed, reached significance 7 weeks postinjection and were maintained throughout the experiment. No significant differences in tumor growth were observed between glucose and control condition (Fig. 4A). Tumors dissected from fructose condition were visually larger in size (Fig. 4B) and significantly higher in weight (Fig. 4C) when compared with tumors dissected from glucose or control condition. Size (Fig. 4B) and weight (Fig. 4C) of tumors dissected from glucose condition did not differ significantly from control condition. This apparent contradictory effect of glucose on in vitro versus in vivo analyses on PC3 cell growth may be the result of a metabolic adaptation of PC3 cells grown in vivo. Consequently, we analyzed the expression of Glut1, Glut2, Glut5, Glut7, Glut9, and Glut11 in tissue specimens of PC3 cell line–derived xenograft tumors (Supplementary Fig. S2A). Our results indicated that Glut5 and Glut9 were significantly upregulated in fructose when compared with both glucose and control conditions (Supplementary Fig. S2B). Although not statistically significant, Glut1 was downregulated in fructose condition when compared with both glucose or control condition (Supplementary Fig. S2A and S2B). Lack of statistical significance in Glut1 expression could be attributable to the masking effect caused by the high level of expression of this isoform in hypoxic/peri-necrotic areas observed in all treatment conditions.

Figure 4.

Effect of fructose oral administration on preclinical models of human prostate cancer. A, Tumor growth of PC3 cell line–derived xenografts was measured every week for a total of 8 weeks. B and C, Representative images of PC3 cell line–derived xenograft tumors (B) and tumor weight of dissected tumors (C). D, Histochemical and IHC analyses of PDXs (n = 3). Representative images of hematoxylin and eosin (H&E) staining and IHC analyses for Ki-67, Glut5, and Glut9 in tissue sections from PDXs. Initial tissue obtained from human prostate cancer specimens before transplantation was used as control of prostate tissue architecture. Plots represent immunostaining quantitation of 10 images from each specimen using the ImageJ software. ns, not statistically significant; *, P < 0.05; **, P < 0.01 (two-way ANOVA).

Figure 4.

Effect of fructose oral administration on preclinical models of human prostate cancer. A, Tumor growth of PC3 cell line–derived xenografts was measured every week for a total of 8 weeks. B and C, Representative images of PC3 cell line–derived xenograft tumors (B) and tumor weight of dissected tumors (C). D, Histochemical and IHC analyses of PDXs (n = 3). Representative images of hematoxylin and eosin (H&E) staining and IHC analyses for Ki-67, Glut5, and Glut9 in tissue sections from PDXs. Initial tissue obtained from human prostate cancer specimens before transplantation was used as control of prostate tissue architecture. Plots represent immunostaining quantitation of 10 images from each specimen using the ImageJ software. ns, not statistically significant; *, P < 0.05; **, P < 0.01 (two-way ANOVA).

Close modal

In parallel, we utilized a PDX model in which tissue integrity and architecture remained constant throughout the experiment (Fig. 4D, H&E). Immunostaining analysis of Ki-67 in tissue sections from PDX showed a significant increase in the percentage of Ki-67–positive prostate cancer cells in fructose when compared with both glucose and control conditions (Fig. 4D). In addition, Glut5 and Glut9 expression was significantly increased in fructose when compared with control condition (Fig. 4D). No significant differences in Glut5 and Glut9 expression were observed between fructose and glucose or between glucose and control condition (Fig. 4D). Stimulatory effect of fructose on tumor growth also was reported in certain subtypes of breast (11) and intestinal (12) cancers. In prostate cancer, however, these findings become more relevant because prostate cancer cells express very low or intracellular levels of Glut1 (13) and prostate cancer tumors are not efficiently imaged using PET-FDG (6).

To get insights into the long-term effect of fructose on gene pathways that drive proliferation in prostate cancer cells, we performed a transcriptome analysis of PC3 cells stimulated in vitro with either 5 mmol/L fructose or 5 mmol/L glucose. Interestingly, and according to our in vitro and in vivo data, GSEA performed on the RNA-seq data showed that fructose stimulation significantly enriched (P = 0.0004) for cell proliferation–related genes (GO_CELL_PROLIFERATION, https://www.gsea-msigdb.org) when compared with glucose stimulation (Supplementary Fig. S3A). Among the top 10 fructose-upregulated genes included in this pathway, our study found overexpression of classic protumorigenic/proliferation cytokines/paracrine factors, such as the TGFβ2, ligand for the receptor-type protein-tyrosine kinase (KITLG), wingless-type MMTV integration site family-4 (WNT4), and FGF5, among others (Supplementary Fig. S3B). Further in vitro and in vivo knocked down analyses on some of these genes might provide potential target genes through which fructose or fructose metabolites promote proliferation of prostate cancer cells and probably other cancer cell types as well.

Taken together, our findings provide compelling evidence to support a key role for fructose as a potential metabolic substrate capable of increasing prostate cancer cell proliferation and aggressiveness. Further expression and functional studies of Glut5 in prostate cancer may provide clinical and therapeutic opportunities, such as the validation of Glut5 as a potential biomarker and a therapeutic target.

No disclosures were reported.

D.V. Carreño: Conceptualization, investigation, writing–original draft. N.B. Corro: Conceptualization, validation, investigation. J.F. Cerda-Infante: Conceptualization, validation, investigation, visualization, methodology. C.E. Echeverría: Validation, investigation, visualization, methodology. C.A. Asencio-Barría: Conceptualization, validation, investigation, visualization, methodology. V.A. Torres-Estay: Conceptualization, validation, investigation, visualization. G.A. Mayorga-Weber: Conceptualization, validation, investigation, visualization. P.A. Rojas: Conceptualization, validation, investigation, visualization, methodology. L.P. Véliz: Conceptualization, validation, investigation, visualization, methodology. P.A. Cisternas: Conceptualization, validation, investigation, visualization. V.P. Montecinos: Conceptualization, validation, investigation, visualization. I.F. San Francisco: Conceptualization, validation, visualization. M.A. Varas-Godoy: Validation, investigation, visualization, methodology. P.C. Sotomayor: Conceptualization, supervision, validation, investigation. M.A. Castro: Conceptualization, validation, investigation, visualization, methodology. F.J. Nualart: Conceptualization, validation, investigation, visualization. N.C. Inestrosa: Validation, investigation, visualization. A.S. Godoy: Conceptualization, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing.

This work was supported by grants from the Department of Defense (W81XWH-12-1-0341 to A.S. Godoy), HHMI Janelia Visitor Program (JVS0028700 to M.A. Castro), FONDECYT (1161115 to A.S. Godoy), FONDECYT (1160724 to N.C. Inestrosa), FONDECYT (11140255 to P.C. Sotomayor), FONDECYT (11160651 to P.A. Cisternas), FONDECYT (1191620 to M.A. Castro), FONDECYT (1150397 to V.P. Montecinos), CMA BIO BIO PIA-Conicyt (ECM-12 to F.J. Nualart), Basal Center of Excellence in Aging and Regeneration (CONICYT-PFB12/2007 to N.C. Inestrosa), and fellowships from FONDECYT-Post-Doctoral (3160717 to D.V. Carreño), CONICYT-PhD (22140138 to V.A. Torres-Estay), CONICYT-PhD (21171084 to N.B. Corro), PhD Scholarship FAI-Universidad de los Andes-Chile (to C.E. Echeverría).

1.
Siegel
RL
,
Miller
KD
,
Jemal
A
. 
Cancer statistics, 2019
.
CA Cancer J Clin
2019
;
69
:
7
34
.
2.
Tappy
L
,
Le
KA
,
Tran
C
,
Paquot
N
. 
Fructose and metabolic diseases: new findings, new questions
.
Nutrition
2010
;
26
:
1044
9
.
3.
Le
MT
,
Frye
RF
,
Rivard
CJ
,
Cheng
J
,
McFann
KK
,
Segal
MS
, et al
Effects of high-fructose corn syrup and sucrose on the pharmacokinetics of fructose and acute metabolic and hemodynamic responses in healthy subjects
.
Metabolism
2012
;
61
:
641
51
.
4.
Echeverria
C
,
Nualart
F
,
Ferrada
L
,
Smith
GJ
,
Godoy
AS
. 
Hexose transporters in cancer: from multifunctionality to diagnosis and therapy
.
Trends Endocrinol Metab
2021
;
32
:
198
211
.
5.
Bender
H
,
Schomburg
A
,
Albers
P
,
Ruhlmann
J
,
Biersack
HJ
. 
Possible role of FDG-PET in the evaluation of urologic malignancies
.
Anticancer Res
1997
;
17
:
1655
60
.
6.
Salminen
E
,
Hogg
A
,
Binns
D
,
Frydenberg
M
,
Hicks
R
. 
Investigations with FDG-PET scanning in prostate cancer show limited value for clinical practice
.
Acta Oncol
2002
;
41
:
425
9
.
7.
Fabiani
R
,
Minelli
L
,
Bertarelli
G
,
Bacci
S
. 
A western dietary pattern increases prostate cancer risk: a systematic review and meta-analysis
.
Nutrients
2016
;
8
:
626
.
8.
Lin
PH
,
Aronson
W
,
Freedland
SJ
. 
Nutrition, dietary interventions and prostate cancer: the latest evidence
.
BMC Med
2015
;
13
:
3
.
9.
Makarem
N
,
Bandera
EV
,
Lin
Y
,
Jacques
PF
,
Hayes
RB
,
Parekh
N
. 
Consumption of sugars, sugary foods, and sugary beverages in relation to adiposity-related cancer risk in the Framingham offspring cohort (1991–2013)
.
Cancer Prev Res
2018
;
11
:
347
58
.
10.
Liu
H
,
Huang
D
,
McArthur
DL
,
Boros
LG
,
Nissen
N
,
Heaney
AP
. 
Fructose induces transketolase flux to promote pancreatic cancer growth
.
Cancer Res
2010
;
70
:
6368
76
.
11.
Fan
X
,
Liu
H
,
Liu
M
,
Wang
Y
,
Qiu
L
,
Cui
Y
. 
Increased utilization of fructose has a positive effect on the development of breast cancer
.
PeerJ
2017
;
5
:
e3804
.
12.
Goncalves
MD
,
Lu
C
,
Tutnauer
J
,
Hartman
TE
,
Hwang
SK
,
Murphy
CJ
, et al
High-fructose corn syrup enhances intestinal tumor growth in mice
.
Science
2019
;
363
:
1345
9
.
13.
Reinicke
K
,
Sotomayor
P
,
Cisterna
P
,
Delgado
C
,
Nualart
F
,
Godoy
A
. 
Cellular distribution of Glut-1 and Glut-5 in benign and malignant human prostate tissue
.
J Cell Biochem
2012
;
113
:
553
62
.
14.
Carvalho
KC
,
Cunha
IW
,
Rocha
RM
,
Ayala
FR
,
Cajaiba
MM
,
Begnami
MD
, et al
GLUT1 expression in malignant tumors and its use as an immunodiagnostic marker
.
Clinics
2011
;
66
:
965
72
.
15.
Xiao
H
,
Wang
J
,
Yan
W
,
Cui
Y
,
Chen
Z
,
Gao
X
, et al
GLUT1 regulates cell glycolysis and proliferation in prostate cancer
.
Prostate
2018
;
78
:
86
94
.
16.
Gasinska
A
,
Jaszczynski
J
,
Rychlik
U
,
Luczynska
E
,
Pogodzinski
M
,
Palaczynski
M
. 
Prognostic significance of serum PSA Level and telomerase, VEGF and GLUT-1 protein expression for the biochemical recurrence in prostate cancer patients after radical prostatectomy
.
Pathol Oncol Res
2019
;
26
:
1049
56
.
17.
Hui
H
,
Huang
D
,
McArthur
D
,
Nissen
N
,
Boros
LG
,
Heaney
AP
. 
Direct spectrophotometric determination of serum fructose in pancreatic cancer patients
.
Pancreas
2009
;
38
:
706
12
.
18.
Regina
A
,
Roux
F
,
Revest
PA
. 
Glucose transport in immortalized rat brain capillary endothelial cells in vitro: transport activity and GLUT1 expression
.
Biochim Biophys Acta
1997
;
1335
:
135
43
.
19.
Steffan
JJ
,
Dykes
SS
,
Coleman
DT
,
Adams
LK
,
Rogers
D
,
Carroll
JL
, et al
Supporting a role for the GTPase Rab7 in prostate cancer progression
.
PLoS One
2014
;
9
:
e87882
.
20.
Chang
J
,
Xu
W
,
Liu
G
,
Du
X
,
Li
X
. 
Downregulation of Rab23 in prostate cancer inhibits tumor growth in vitro and in vivo
.
Oncol Res
2017
;
25
:
241
8
.