Research ArticleProstate Cancer

Suppressing fatty acid uptake has therapeutic effects in preclinical models of prostate cancer

See allHide authors and affiliations

Science Translational Medicine  06 Feb 2019:
Vol. 11, Issue 478, eaau5758
DOI: 10.1126/scitranslmed.aau5758
  • Fig. 1 Fatty acid uptake and storage are increased in malignant compared with benign human prostate tissue.

    (A) Glucose uptake (n = 8), (B) glucose oxidation (n = 8), (C) de novo lipogenesis (n = 12), (D) fatty acid uptake (n = 11), (E) the percent difference in fatty acid uptake between patient-matched malignant (mal) and benign tissue, (F) fatty acid oxidation (n = 14), and (G) fatty acid incorporation into various lipids (n = 6 to 8 per lipid type). Cer, ceramide; PL, phospholipid; CE, cholesterol ester; DAG, diacylglycerol; TAG, triacylglycerol. *P < 0.05 malignant versus benign by two-tailed paired t test. For (A) to (G), data are presented as means ± SEM. (H) Kaplan-Meier curve showing estimated biochemical recurrence (BCR)–free probability after radical prostatectomy in patients with high (upper quartile) or low levels of tumor CD36 mRNA in the TCGA (provisional) cohort (27). This cohort contains genomic and transcriptomic data from 491 tumor and 40 normal human prostate samples. P = 0.022 and hazard ratio (HR) of 1.8 were determined using a log-rank test. (I) Histogram showing the alteration frequency of CD36 in seven distinct prostate cancer cohorts.

  • Fig. 2 shRNA-mediated silencing of CD36 in prostate cancer cells reduces fatty acid uptake and impairs cancer aggressiveness.

    (A) CD36 mRNA expression in PC3 parental cells, cells transfected with a shRNA with a scrambled sequence (shRNA neg), and two independent PC3 cell lines stably transfected with shCD36. n = 4 per group. *P < 0.05 versus PC3 by one-way analysis of variance (ANOVA) and Bonferroni multiple comparisons test. (B to D) Fatty acid metabolism was assessed in cells treated with 500 μM oleate (1 mCi/ml of 1-14C-oleate) for 2 hours for (B) fatty acid uptake, (C) fatty acid oxidation, and (D) fatty acid incorporation into cellular lipids. Data are presented as means ± SEM, n = 10 to 12 per group from four independent experiments. *P <0.05 versus PC3 by one-way ANOVA and Bonferroni multiple comparisons test. (E) In vitro proliferation of PC3 cells without or with CD36 knockdown using shRNA. Data from PC3 parental and shRNA neg are reported together as PC3. Cells were treated without (−FA) or with (+FA) fatty acids (oleate:palmitate, 1:1 conjugated to 1% bovine serum albumin) for 24 hours. Data represent means ± SEM, n = 3 per condition from one of three independent experiments. *P < 0.05 by two-way ANOVA and Bonferroni multiple comparisons test. (F and G) Migration efficiency calculated by recovery of the cell surface after manually scratching the cell surface. Data are presented as means ± SEM, n = 4 to 6 per group from three independent experiments. *P < 0.05 by two-way ANOVA and Bonferroni multiple comparisons test. (F) Representative images of cells at start and end point. Scale bars, 50 μm. (H) Cell motility assessed by Transwell cell migration assays calculated as the number of cells migrated through the Transwell filter relative to control after 18 hours. Data are presented as means ± SEM, n = 3 to 7 per group from three independent experiments. *P < 0.05 by two-way ANOVA and Bonferroni multiple comparisons test. (I) In vivo growth of subcutaneous tumors after injection of shCD36 PC3 cells compared and shRNA neg PC3 control cells in SCID mice (subcutaneous implantation of 1.8 × 106 cells in collagen). Tumor volume was measured at the indicated time points. Data are presented as means ± SEM; n = 6 for shRNA neg, n = 10 for shCD36 from independent clones #1 and #3. Data assessed by two-way ANOVA. Ptime×genotype = 0.07, Ptime < 0.0001, Pgenotype = 0.005.

  • Fig. 3 Genetic ablation of Cd36 alters fatty acid metabolism in the prostate tissue of Pten-deficient mice.

    (A) Immunohistochemical staining of Cd36 in lateral prostate lobes of WT, Pten−/−, and Pten−/−.Cd36−/− mice (isotype-matched negative control in bottom right on Pten−/−.Cd36−/− mice). Images are representative of n = 4 animals per genotype. Scale bars, 50 μm. (B to E) Fatty acid metabolism was assessed in the anterior prostate of mice using 500 μM oleate (1 mCi/ml of 1-14C-oleate) for 2 hours. (B) Fatty acid uptake, (C) fatty acid oxidation, (D) fatty acid storage into various lipids, and (E) the fatty acid storage to oxidation ratio. Data are presented as means ± SEM; n = 11 for WT, n = 4 for Pten−/−, n = 15 for Pten−/−.Cd36−/− mice. (F) De novo lipogenesis was assessed in the anterior prostate lobe using 5 mM glucose (3 μCi/ml of [14C(U)]-glucose). Data are presented as means ± SEM; n = 9 for WT, n = 7 for Pten−/, n = 5 for Pten−/−.Cd36−/− mice. For all figures, horizontal lines denote P < 0.05 between adjoining bars. Data are assessed by one-way ANOVA and Bonferroni multiple comparisons tests.

  • Fig. 4 Cd36 inactivation slows prostate cancer progression and reduces the primary prostate tumor burden in prostate-specific Pten−/−mice.

    (A) Prostate tissue mass of lateral, anterior, ventral, and dorsal lobes from WT, Pten−/−, and Pten−/−.Cd36−/− mice fed a standard chow diet for 8, 12, and 24 weeks. Data are presented as means ± SEM and were analyzed using two-way ANOVA and Bonferroni post hoc test. For each genotype, the n is reported for 8, 12, and 24 weeks, respectively, as follows: n = 8, 20, and 9 for WT, n = 10, 10, and 9 for Pten−/−, n = 8, 17, and 9 for Pten−/−.Cd36−/−. P < 0.05 denoted by “8” for WT versus Pten−/−, “#” for WT versus Pten−/−.Cd36−/−, and “†” for Pten−/− versus Pten−/−.Cd36−/−. Flow cytometric quantitation of (B) total prostate cellularity, (C) prostate epithelial cells, (D) epithelial cell size [mean forward scatter (FSC)], and (E) percentage of epithelial cell proliferation (ki67+) as quantitated by flow cytometric analysis of whole prostate glands from WT, Pten−/−, and Pten−/−.Cd36−/− mice aged 12 to 17 weeks. Results are presented as means ± SEM for all datasets and assessed by one-way ANOVA and Bonferroni multiple comparisons tests. Horizontal lines denote P < 0.05 between adjoining bars; n = 5 for WT, n = 6 to 8 for Pten−/−, n = 4 for Pten−/−.Cd36−/−. (F) Hematoxylin and eosin staining of lateral prostate lobes from mice aged 12 weeks. Pten−/− prostates contain CIS, which is evident by glandular ducts filled with noninvasive malignant cells (black arrow) contained within the basement membrane (white arrow), while Pten−/−.Cd36−/− prostates predominantly show HG-PIN (less CIS), containing cytological atypical cells, but retain the ductal lumen (“*”). Scale bars, 100 μm. (G) Pathology was scored as normal, HG-PIN, or CIS. The proportion of each pathology is shown for 8 weeks (n = 8 for WT, n = 10 for Pten−/−, n = 8 for Pten−/−.Cd36−/− mice), 12 weeks (n = 20 for WT, n = 10 for Pten−/−, n = 17 for Pten−/−.Cd36−/− mice), or 24 weeks (n = 9 for WT, n = 9 for Pten−/−, n = 9 for Pten−/−.Cd36−/− mice). Results are presented as means ± SEM for all datasets. Data assessed by one-way ANOVA and Bonferroni multiple comparisons tests. “*” denotes P < 0.05 between Pten−/− and Pten−/−.Cd36−/− mice.

  • Fig. 5 Cd36 inactivation suppresses oncogenic lipid production of murine prostate cancer.

    Lipidomic analysis of prostate tissue in WT, Pten−/−, and Pten−/−.Cd36−/− mice aged 12 weeks. (A) Heat map showing the changes of the prostate lipidome in Pten−/− and in Pten−/−.Cd36−/− mice (bottom). Lipid class and individual lipid species are denoted at the top of the figure. (B) Representative lipids altered in Pten−/− compared with WT mice. *P < 0.05 by unpaired t test. (C) Lipids detected in Pten−/− and Pten−/−.Cd36−/− mice. Data are assessed by one-way ANOVA and Bonferroni multiple comparisons tests, and horizontal lines denote P < 0.05 between adjoining bars. n = 7 for WT, n = 6 for Pten−/−, n = 5 for Pten−/−.Cd36−/−.

  • Fig. 6 CD36 mAb therapy reduces tumor growth in prostate cancer PDXs.

    PDXs were established in NSG male mice from localized hormone-sensitive prostate cancer tissues. Two independent PDX lines were used (PDXlo and PDXhi), which indicate their propensity for fatty acid uptake. (A) Pathology and histological assessment including hematoxylin and eosin (H&E), and expression of AR, PSA, and PSMA. Scale bars, 100 μm. Metabolic assessment of PDXlo and PDXhi showing (B) fatty acid uptake, (C) fatty acid oxidation, (D) fatty acid storage into lipids, and (E) de novo lipogenesis, which was assessed in the PDX grafts using 500 μM oleate (1 mCi/ml of [1-14C]-oleate) for 2 hours. Data are presented as means ± SEM, n = 8 for PDXlo and n = 3 for PDXhi. For therapeutic studies, pair-matched precision slices of PDX were established in IgA control or CD36 mAb–treated mice. (F) PDX graft volume from individual grafts at harvest in PDXlo and PDXhi lines. Paired samples are joined by dashed lines. n = 15 control and treatment for PDXlo and n = 7 for control and treatment for PDXhi. Data assessed by paired Student’s t test, *P < 0.05, main effect of treatment by two-way ANOVA. n = 8 to 12 per group. (G) Metabolic assessment of PDXhi grafts from vehicle and CD36 mAb–treated mice showing fatty acid uptake. Data are presented as means ± SEM, n = 3 PDX grafts per group. Data assessed by paired Student’s t test, *P < 0.05. (H) PDX graft volume from individual grafts at harvest in PDXhi lines after CD36 mAB treatment compared to docetaxel treatment for 3 weeks. Paired samples are joined by lines. Data assessed by paired Student’s t test, *P < 0.05, main effect of treatment by two-way ANOVA. n = 8 to 12 per group.

  • Fig. 7 Inhibition of de novo lipogenesis increases the efficacy of CD36 blockade.

    (A and B) Human PDX-derived organoids treated with C75 fatty acid synthase inhibitor, CD36 mAb, or both in combination for 11 days. (A) Images of day 11 control/treated organoid. Scale bars, 200 μm. (B) Organoid viability after treatment was assessed using the PrestoBlue assay. Data are presented as means ± SEM; n = 6 per group; relative to vehicle control (VC). P < 0.05 is shown by adjoining lines between groups by one-way ANOVA and Bonferroni multiple comparisons test. (C to E) Mouse organoids established from Pten−/− and Pten−/−.Cd36−/− mice treated with and without C75 fatty acid synthase inhibitor for 7 days. (C) Fatty acid uptake was assessed in cells treated with 500 μM oleate (1 mCi/ml of 1-14C-oleate) for 2 hours. (D) De novo lipogenesis was assessed using 5 mM glucose (3 μCi/ml of [14C(U)]-glucose). Data are presented as means ± SEM; n = 6 for all groups. For (C) and (D), horizontal lines between adjoining bars denote main effect for genotype, P < 0.05; *P < 0.05 versus −C75 within the same genotype. Data assessed by one-way ANOVA and Bonferroni multiple comparisons tests. (E) Organoid viability after treatment was assessed using the PrestoBlue assay. Data are presented as means ± SEM; n = 6 per group from two independent experiments. *P < 0.05 versus vehicle control; adjoining line denotes main effect for genotype by one-way ANOVA and Bonferroni multiple comparisons test.

Supplementary Materials

  • www.sciencetranslationalmedicine.org/cgi/content/full/11/478/eaau5758/DC1

    Materials and Methods

    Fig. S1. Survival curves for CD36 and other genes encoding proteins regulating fatty acid uptake.

    Fig. S2. Fatty acid uptake in prostate cancer cell lines.

    Fig. S3. shRNA knockdown of CD36 in PC3 and LNCaP cells.

    Fig. S4. Mouse phenotyping by flow cytometric analysis.

    Fig. S5. Phenotype of WT.Cd36−/− mice.

    Fig. S6. Assessment of metabolism and cancer pathology in Pten−/−.Cd36−/− mice fed a high-fat diet for 6 weeks.

    Fig. S7. Schematic depicting the changes in Pten−/−-induced lipid metabolism that link the production of oncogenic lipids to prostate cancer progression.

    Fig. S8. Linking cPLA2α inhibition to the antitumorigenic effects of Cd36 deletion.

    Fig. S9. Systemic effects of CD36 mAb treatment in mice.

    Table S1. Patient characteristics of specimens used for metabolism studies.

    Table S2. Raw data (Excel file).

    References (5878)

  • The PDF file includes:

    • Materials and Methods
    • Fig. S1. Survival curves for CD36 and other genes encoding proteins regulating fatty acid uptake.
    • Fig. S2. Fatty acid uptake in prostate cancer cell lines.
    • Fig. S3. shRNA knockdown of CD36 in PC3 and LNCaP cells.
    • Fig. S4. Mouse phenotyping by flow cytometric analysis.
    • Fig. S5. Phenotype of WT.Cd36−/− mice.
    • Fig. S6. Assessment of metabolism and cancer pathology in Pten−/−.Cd36−/− mice fed a high-fat diet for 6 weeks.
    • Fig. S7. Schematic depicting the changes in Pten−/−-induced lipid metabolism that link the production of oncogenic lipids to prostate cancer progression.
    • Fig. S8. Linking cPLA2α inhibition to the antitumorigenic effects of Cd36 deletion.
    • Fig. S9. Systemic effects of CD36 mAb treatment in mice.
    • Table S1. Patient characteristics of specimens used for metabolism studies.
    • References (5878)

    [Download PDF]

    Other Supplementary Material for this manuscript includes the following:

Stay Connected to Science Translational Medicine

Navigate This Article