Dendrogenin A drives LXR to trigger lethal autophagy in cancers

Abstract

Dendrogenin A (DDA) is a newly discovered cholesterol metabolite with tumor suppressor properties. Here, we explored its efficacy and mechanism of cell death in melanoma and acute myeloid leukemia (AML). We found that DDA induced lethal autophagy in vitro and in vivo, including primary AML patient samples, independently of melanoma Braf status or AML molecular and cytogenetic classifications. DDA is a partial agonist on liver-X-receptor (LXR) increasing Nur77, Nor1, and LC3 expression leading to autolysosome formation. Moreover, DDA inhibited the cholesterol biosynthesizing enzyme 3β-hydroxysterol-Δ^8,7-isomerase (D8D7I) leading to sterol accumulation and cooperating in autophagy induction. This mechanism of death was not observed with other LXR ligands or D8D7I inhibitors establishing DDA selectivity. The potent anti-tumor activity of DDA, its original mechanism of action and its low toxicity support its clinical evaluation. More generally, this study reveals that DDA can direct control a nuclear receptor to trigger lethal autophagy in cancers.

Fig. 1

Nur77 and NOR1 are required for the induction of cell death by DDA in melanoma. a Chemical structure of DDA and C17. b B16F10 and SKMEL-28 cells were treated or not for 24, 48, or 72 h with DDA or C17. DDA- or C17-induced cell death was determined by a trypan blue assay and expressed as the percentage cell death relative to control (vehicle). c Cells were treated with 5 µM DDA for 48 h or 72 h with or without 500 µM vitamin E (Vit E), 50 µM z-VAD-fmk, 50 µM z-DEVD-fmk, 1 µg/ml actinomycin D (Act D) or 2.5 µg/ml cycloheximide (CHX), and cell death was measured as in b. d DDA induced the accumulation of free sterols in cells. Cells were treated with solvent vehicle or 2.5 µM DDA for 48 h, then fixed and stained with filipin and analyzed by fluorescence microscopy. e Representative EM images of B16F10 (panels 1–3) and SKMEL-28 (panels 2–4) cells treated for 24 h with 2.5 µM DDA. Ly: multilamellar body-derived lysosomes, AP: autophagosomes, AL: autolysosomes, N: nucleus, C: cytoplasm. Bars: 250 nm for panel 1 and 100 nm for inserts; 500 nm for panel 2 and 100 nm for inserts; 500 nm for panel 3; and 250 nm for panel 4. f Heat map depicting transcription of genes encoding nuclear receptors (NR) and their co-regulators, using PCR arrays in B16F10 and SKMEL-28 cells treated or not with 2.5 µM DDA for 5 h, using PCR arrays. g Immunoblot for Nur77 and NOR1 protein expression in melanoma cells treated or not with DDA for 24 h. h Analysis of DDA cytotoxicity in B16F10 and SKMEL-28 cells transfected with control scramble siRNA (siSC), siNur77 and/or siNOR1. 72 h after transfection, cells were treated or not for 24 h with 2.5 µM DDA. Cell death is expressed as a percentage relative to the level of cell death induced by DDA in cells transfected with a scramble control siRNA (siSC). Data from b–d and h are the means ± S.E.M. of three independent experiments performed in triplicate, **P < 0.01, ***P < 0.001, t-test. All images and densitometry values are representative of three independent experiments

Fig. 2

DDA induces Nur77- and NOR1-dependent lethal autophagy in melanoma cells. a DDA triggers the accumulation of autophagic vesicles. Cells were treated for 24 h with or without 2.5 µM DDA then stained with monodansylcadaverine (MDC) and observed by fluorescence microscopy. MDC-specific activity was measured by fluorescence photometry. b Cells were treated for 24 h with solvent vehicle or increasing concentrations of DDA then analyzed for autophagic protein expression by immunoblotting. Blots are representative of three independent experiments. c Long-life protein degradation was determined in cells treated with solvent vehicle (control) or 1 µM DDA for 18 h in the presence or absence of the autolysosomal inhibitors bafilomycin A1 (Baf A1) and hydroxychloroquine (HCQ). Autophagic activity was measured as the level of degradation of long-lived proteins. Starvation for 18 h in Hank’s balanced salt solution (HBSS) was used as a positive control. d Immunoblots of LC3 proteins from cells treated for 24 h with or without 2.5 µM DDA and with or without E64 + pepstatin A (10 µg/ml). Images are representative of three independent experiments. e DDA induced the formation of punctate LC3 cells. Cells were transfected with a plasmid-expressing GFP-LC3 and then treated for 24 h with the solvent vehicle or 2.5 µM DDA, with or without E64 + pepstatin A (10 µg/ml) and observed by fluorescent microscopy. The percentage of GFP-LC3-positive cells with GFP-LC3 puncta was calculated. f Analysis of DDA cytotoxicity in cells transfected with scramble siRNA (siSC), siATG7, siVPS34, or siBECN1. Seventy-two hours after transfection, cells were treated or not for 24 h with 2.5 µM DDA. g Analysis of DDA cytotoxicity in SKMEL-28 cells permanently transfected with control shRNA (shCTRL) or shRNA against VPS34 (shVPS34). Cells were treated for 72 h with solvent vehicle (CTRL) or 2.5 µM DDA. h Cells were treated with 2.5 µM DDA for 24, 48 and 72 h in the presence or absence of the autolysosome inhibitors Baf A1 or HCQ. Cell death is expressed as in Fig. 1a. Data from a, c, e, f, g are the means ± S.E.M. of three experiments performed in triplicate, *P < 0.05, **P < 0.01, ***P < 0.001, t-test

Fig. 3

LXR are targets of DDA and LXRβ is required for its cytotoxicity in melanoma cells. a LXR transcriptional activity was analyzed using transient transfection reporter assays. Tranfected cells were treated with 10 µM 22(R)HC with or without DDA. b Competition binding assays on LBD-LXRα or LBD-LXRβ. c SPR sensorgrams showing the binding of DDA to the LBD-LXRβ. d Molecular docking of DDA with the LBD-LXRβ. Amino acid side chains that interact with DDA are represented (in black). The names of the amino acids known to interact with known LXR ligands are colored in blue. Gray: carbon atoms, white: hydrogen atoms, red: oxygen atoms, blue: nitrogen atoms, yellow: sulfur atoms. e Structures of DDA analogs assayed in the LXR reporter assay. f Analysis of LXRβ-dependent agonistic or antagonistic activities by DDA and analogs. g the stimulation of LC3, Nur77, and NOR1 protein expression by DDA is LXRβ-dependent. h ChIP-qPCR of LXRβ on the SCD1, LC3A, LC3B, SREBP1, NR4A1, NR4A3, ABCA1, and LDLR enhancers on SKMEL-28 cells treated or not with 10 µM 22(R)HC or 2.5 µM DDA. i ChIP-qPCR of LXRβ on the TFEB enhancer on SKMEL-28 cells treated or not with 10 µM 22(R)HC or 2.5 µM DDA. j Real-time PCR of TFEB expression in SKMEL-28 cells treated or not with 5 or 10 µM 22(R)HC, and 2 or 5 µM DDA. k Luciferase reporter gene assays with the TFEB promoter-luciferase construct in HEK293T. Cells were treated with increasing DDA concentrations. l Analysis of DDA cytotoxicity in cells transfected with control siRNA (shSC) or siLXRβ. Cells were treated with 2.5 µM DDA. m Analysis of the cytotoxicity of cells treated with or without 2 µM DDA, 5 µM 22(R)HC, 0.5 µM TO, 1 µM GW or 2 µM DDA + 10 µM 22(R)HC, 0.5 µM TO, or 1 µM GW. n TO reversed DDA induction of autophagic vesicles. Cells were treated for 24 h with or without 2 µM DDA, 0.5 µM TO, or 0.5 µM TO + 2 µM DDA. Cells were stained with MDC and observed by fluorescence microscopy. Data from a, b, f, j, k, l, m are the means ± S.E.M. of three independent experiments performed in triplicate (*P < 0.05, **P < 0.01, ***P < 0.001, t-test)

Fig. 4

DDA induces autophagy in melanoma in vivo and in vitro in an LXRβ-dependent manner. a, b Mice engrafted with human SKMEL-28 or mouse B16F10 cells (10 per group) were treated with DDA (i.p. 20 mg/kg/day) or vehicle. Mean tumor volumes ± S.E.M. are shown, **P < 0.01, analysis of variance (ANOVA). Data are representative of three independent experiments. At the end of treatments, tumors were analyzed for Nur77, NOR1 and LC3 protein expression by a, b immunoblotting or c immunohistochemistry (brown staining). c Representative immunohistochemical analysis. Magnification ×40. Insert is 4× digital amplification. d, e TEM images of cells transfected with shCTRL, sh3LXRβ, or sh4LXRβ and treated with control vehicle or 2.5 µM DDA for 24 h. e TEM images of cells transfected with shCTRL, sh3LXRβ, or sh4LXRβ and treated with 2.5 µM DDA for 24 h. d, e N, nucleus; C, cytoplasm; AM, amphisome

Fig. 5

The anti-melanoma action of DDA in vivo is LXRβ-dependent. a Mice engrafted with SKMEL-28 cells transfected with shCTRL or sh4LXRβ (10 per group) were treated with DDA (i.p. 20 mg/kg/day) or vehicle. Mean tumor volumes ± S.E.M. are shown, **P < 0.01, analysis of variance (ANOVA). Data are representative of three independent experiments. b At the end of treatments, tumors were analyzed for Nur77, NOR1, and LC3 protein expression by immunoblotting. All images and blots are representative of three independent experiments. c, d Mice engrafted with mouse B16F10 cells or human SKMEL-28 (10 per group) were treated with vehicle, DDA (i.p. 20 mg/kg/day), TO (i.p. 20 mg/kg/day), and DDA + TO (i.p. 20 mg/kg/day each). Mean tumor volumes ± S.E.M. are shown, **P < 0.01, analysis of variance (ANOVA). Quantification of Δ8-sterols and 5,6α-EC and 5,6β-EC in tumors were quantified by GC/MS. The results are reported as µg Δ8-sterols or ng 5,6-EC/g tumors. e Box plot of TCGA RNA-seq data from patients with melanoma showing that LXRβ is the predominant LXR isoform expressed. ***P < 0.001

Fig. 6

DDA induces lethal autophagy in AML cells via LXRβ. a DDA-induced cell death in KG1 and HL60 cells was determined over time as in Fig. 1b. b Quantification of the Δ8-sterols that had accumulated in KG1 and HL60 cells treated without or with 2.5 µM DDA. c May–Grünwald–Giemsa staining of KG1 and HL60 cells treated or not with DDA. d Immunoblots of LC3 proteins in cells treated or not with 2.5 µM DDA, and E64 + pepstatin A (Pep). e Effect of the pharmacological inhibitor of autophagy Baf A1 on DDA cytotoxicity at 48 and 72 h. f DDA cytotoxicity in KG1 or HL60 cells permanently transfected with control shRNA (shC) or shRNA against VPS34 (shVPS34) after 72 h treatment. g Immunoblots of ATG3 and LC3 proteins in KG1 cells transfected with control scramble siRNA (siSC) or siRNA against ATG3 (siATG3). Seventy-two hours after transfection, cells were treated for 24 h with 5 µM DDA. h Analysis of DDA cytotoxicity in KG1 cells transfected with control scramble siRNA (siSC), siATG3, siATG7, or siBECN1. Seventy-two hours after transfection, cells were treated for 24 h with 5 µM DDA. i Analysis of DDA cytotoxicity in KG1 cells transfected with shCTRL, sh3LXRβ, or sh4LXRβ. Cells were treated for 24 h with 5 µM DDA or vehicle. j Immunoblots of LC3 protein expression in cells transfected with shCTRL, sh3LXRβ, or sh4LXRβ and treated with 5 µM DDA or vehicle for 24 h. Analysis of the cytotoxicity k and acridine orange-positive vesicles l in KG1 cells treated or not with 5 µM DDA, 2 µM TO, 2 µM GW, or 10 µM 22(R)HC. The presence of Nur77 and NOR1 is required in DDA cytotoxicity (m) and autophagy (n). Data from a, b, e, f, h, i, k, l, m, n are the means ± S.E.M. of three experiments performed in triplicate, *P < 0.05, **P < 0.01, ***P < 0.001 t-test. All images and densitometry values are representative of three independent experiments

Fig. 7

DDA induces an LXRβ-, Nur77-, and NOR1-dependent lethal autophagy in AML. a Tumor volume curves of xenografts of cells transfected with shCTRL, sh3LXRβ, and sh4LXRβ and implanted into NOD/SCID mice (20 per group) who were then treated daily with DDA (20 mg/kg/day, i.p.) or solvent vehicle. b HL60 cells were injected i.v. into irradiated NSG mice (n = 10 per group) who were then treated daily with DDA (20 mg/kg/day, i.p. or 40 mg/kg/day, p.o.) for 16 days. Analyses of HL60 cell contents and viability in bone marrow (BM) and brain (BR) of NSG mice. Cells were quantified by flow cytometry using human anti-CD45 and human anti-CD33 antibodies (left panel) and viability was determined by Annexin-V staining (right panel). c Overall survival was determined for NSG mice (n = 10 per group) engrafted with HL60 cells and treated, after disease establishment, with control (vehicle) or DDA (40 mg/kg/day, p.o.), *P < 0.05, log-rank test. Samples from AML patients (n = 61, Supplementary Data  1) were exposed to increasing concentrations of DDA (d) or cytarabine (e) for 48 h. Cell death was assessed both in the AML bulk and in the progenitor/LSC cells (CD34+CD38−CD123+) using Annexin-V/7AAD staining. Data are represented as the percentage of survival. Scatter plots comparing DDA efficacy in primary AML patients according to f their prognostic risk category (LR: low risk, IR: intermediate risk, and HR: high risk), g CFU-L formation, h total white blood count, and i their Flt3-ITD and NMP1 status. Samples from AML patients were exposed to 10 nM daunorubicin (DNR) j or 100 µM cytarabine k with or without 5 µM DDA for 72 h. “daunorubicin insensitive” or “cytarabine insensitive” when cell death was lower than 20%, and “sensitive” when cell death was over 70%. Bars represent S.E.M. l Images of primary AML cells stained with MGG after treatment with vehicle or DDA for 24 h. Images are representative of three independent experiments. Data from a-e are means ± S.E.M., and is representative of 3–5 independent experiments, **P < 0.01, t-test. d–k Bars represent S.E.M., *P < 0.05, t-test

Fig. 8

DDA exerts anti-leukemic activity in vivo in patient AML samples. Primary cells from AML patients were injected i.p. into irradiated NSG mice (three AML patients were tested separately). After validation of tumor engraftment, mice were treated with DDA (20 mg/kg/d, i.p.) or control (vehicle) for 19 days. a Representative flow cytometry analysis showing the selection of the human AML population (hCD45+hCD33+) and analysis of cell viability by Annexin-V/7AAD staining. b Human leukemic cell content in the hind limb bone marrow (BM) and spleen (SP) was measured by flow cytometry using human anti-CD45 and anti-CD33 antibodies. The viability of human CD45+CD33+ cells was determined by Annexin-V staining and flow cytometry analysis, *P < 0.05, **P < 0.01, ***P < 0.001, t-test. c Histological analysis of femur and sternum sections from mice injected with primary AML cells (AML#1), stained with Goldner (femur) and HE (sternum). d Histological analysis of bone marrow staining for Nur77, NOR1, and P62. Normal human CD34+ cells from a healthy donor were implanted intravenously into NSG mice and treated with vehicle or DDA for 3 weeks (20 mg/kg/day, i.p.). e Engraftment was quantified by assessing the percentage of hCD45+ cells in the BM. f The percentage of human myeloid (CD45+/CD33+) and lymphoid (CD45+/CD19+) cells was determined by flow cytometry. e, f Bars represent S.E.M

Fig. 9

Molecular mechanisms through which DDA induced lethal autophagy in cancer cells. D8D7I, 3β-hydroxysterol-Δ8,7-isomerase; ChEH, cholesterol-5,6-epoxide hydrolase; Δ8-sterols, zymostenol and 8-DHC; Ly, lysosome; AM, amphisome; AL, autolysosome; AP, autophagosome