Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Nov;7(11):1266-1283.
doi: 10.1158/2159-8290.CD-17-0741. Epub 2017 Sep 12.

A Unified Approach to Targeting the Lysosome's Degradative and Growth Signaling Roles

Affiliations
Free PMC article

A Unified Approach to Targeting the Lysosome's Degradative and Growth Signaling Roles

Vito W Rebecca et al. Cancer Discov. .
Free PMC article

Abstract

Lysosomes serve dual roles in cancer metabolism, executing catabolic programs (i.e., autophagy and macropinocytosis) while promoting mTORC1-dependent anabolism. Antimalarial compounds such as chloroquine or quinacrine have been used as lysosomal inhibitors, but fail to inhibit mTOR signaling. Further, the molecular target of these agents has not been identified. We report a screen of novel dimeric antimalarials that identifies dimeric quinacrines (DQ) as potent anticancer compounds, which concurrently inhibit mTOR and autophagy. Central nitrogen methylation of the DQ linker enhances lysosomal localization and potency. An in situ photoaffinity pulldown identified palmitoyl-protein thioesterase 1 (PPT1) as the molecular target of DQ661. PPT1 inhibition concurrently impairs mTOR and lysosomal catabolism through the rapid accumulation of palmitoylated proteins. DQ661 inhibits the in vivo tumor growth of melanoma, pancreatic cancer, and colorectal cancer mouse models and can be safely combined with chemotherapy. Thus, lysosome-directed PPT1 inhibitors represent a new approach to concurrently targeting mTORC1 and lysosomal catabolism in cancer.Significance: This study identifies chemical features of dimeric compounds that increase their lysosomal specificity, and a new molecular target for these compounds, reclassifying these compounds as targeted therapies. Targeting PPT1 blocks mTOR signaling in a manner distinct from catalytic inhibitors, while concurrently inhibiting autophagy, thereby providing a new strategy for cancer therapy. Cancer Discov; 7(11); 1266-83. ©2017 AACR.See related commentary by Towers and Thorburn, p. 1218This article is highlighted in the In This Issue feature, p. 1201.

Conflict of interest statement

Conflict of Interest Statement: RA and JW are inventors on 3 patent applications related to this work. One patent has been licensed to a biotech company to promote clinical development of Lys05 derivatives.

Figures

Figure 1
Figure 1. DQs have superior anti-cancer efficacy amongst dimeric anti-malarials
(A) Chemical structures of dimeric antimalarials. Ar: aromatic ring; CQ (chloroquine), QN (quinacrine); PQ (primaquine); MQ (mefloquine). (B) Cells were treated with compounds shown (2 weeks, 3 μM) in colony formation assays. (C) Calculated Log IC50s from 72-hour MTT assays in A375P (melanoma) cells treated with compounds shown. (D) Calculated Log IC50s of indicated compounds (72 hr, 1 nM - 30 μM) from MTT assays in A375P cells. (E) PANC1 cells were treated as indicated (24 hr, 3 μM), stained with Annexin-V and analyzed using flow cytometry. (B-D) for all graphs, mean ± standard deviation (s.d.) for N = 3 independent experiments are presented; *p<0.05.
Figure 2
Figure 2. Central nitrogen methylation status directs effects upon autophagy, induction versus inhibition
(A) Fluorescence spectroscopy emission spectra are shown (excitation 424 nm) (B) Fluorescent microscopy of PANC1 (pancreas cancer) cells treated as indicated (6 hr, 3 μM) and co-stained with LysoTracker deep red (shown green); arrows: co-localization of DQ compound (red) with LysoTracker (green). (C) LC3B immunoblot in lysates from C8161 (melanoma cells) treated as indicated (4 hr, 3 μM). (D) Bafilomycin clamp (100 nM) to measure autophagic flux was performed on A375P cells treated with DQ compounds (4 hr, 3 μM). Lysate was immunoblotted for LC3B. (E) PANC1 cells were treated with compounds shown (6 hr, 3 μM). Lysates were immunoblotted and change in p62 densitometry levels were quantified and are depicted in the graph to the right; *p<0.05. (F) A375P cells were treated with DQ660 or DQ661 (6 hr, 3 μM) and lysosomal and cytosolic fractions were subsequently isolated. Lysates were immunoblotted for LAMP2 and Tubulin (left panel). Bio-orthogonal fluorescence of DQ660 and DQ661 was measured on a fluorescent plate reader to quantify the lysosomal quantity of DQ660 and DQ661.
Figure 3
Figure 3. Central nitrogen methylation status dictates DNA damage versus lysosomal membrane permeability
(A) IF microscopy of A375P cells treated as indicated (6 hr, 3 μM) and stained for phospo-H2AX (red; arrows) and DAPI (blue); and scored using ImageJ (mean ± SD); *p<0.05. (B) IF microscopy of A375P cells treated as in (A) with the addition of the positive control LLoMe (3 hr, 2 mM). IF against galectin-3 is shown. White arrows: galectin-3 punctae, reflecting lysosomal membrane permeabilization. (C) Lysosomal sub-fractionation and immunoblotting in A375P cells treated as indicated (6 hr, 3 μM. WC: whole lysate, L: lysosomal fraction. (D) A375P cells were treated with DQ661 (24 hr, 3 μM) and lysates were subjected to reverse phase protein array (RPPA; see supplemental figure S3A for complete dataset). Bar graphs show fold change of a selected panel of proteins at 24 hrs. *p<0.05. (E) Cells treated with DQ661 (6 hr, 3 μM) and lysate was immunoblotted. (F) A375P cells were treated with DQ661 (0 – 24 hr; 3 μM) and lysate was immunoblotted. (G) A375P cells were treated with DQ661 (0 – 4 hr; 3 μM) or Llome (0 – 24 hr; 2 mM) and subsequently stained for galect-3 and imaged by IF microscopy. Below: quantitation of percentage of galectin-3 puncta positive cells. (H) A375P cells were treated with DQ661 (6 hr, 3 μM), PBS, Pepstatin A (10 μG/mL), E64 (10 μG/mL), Pepstatin A + E64, Siramesine (8 μM), PES (10 μM), PET (10 μM), Bafilomycin A1 (100 nM), or Bafilomycin A1 + DQ661. Lysate was immunoblotted against phospho-S6K T389, total S6K, phospho-4E-BP1 S65, total 4E-BP1, LC3B and Actin. (I) A375P cells were treated with a CQ, Lys05, QC, DQ660 and DQ661 (6 hr, 3 μM) and lysate was subjected to RPPA. Shown are graphs reflecting fold decrease of mTORC1 substrates. (A – B) a Cy5 secondary antibody was used for pH2AX and galectin-3 to avoid spectral overlap with DQ661. Scale bars: 80 μm in (A) and 25 μm in (B).
Figure 4
Figure 4. PPT1 is a target of DQ661
(A) Schematic demonstrating the pulldown strategy. (B) Chemical structure of the DQ661-photoprobe. (C) A375P cells were treated with the DQ661-photoprobe (24 hr, 0 – 100 μM) and lysates were immunoblotted. (D) Graph depicting the mass spectrometry analysis of lysate from pulldown with the DQ661-photoprobe. Conditions analyzed were cells that were treated with the DQ661-photoprobe ± UV and ± 10X concentration of DQ661 competition. (E) Cathepsin activity in A375P cells following treatment with DQ660 or DQ661 (3 hr, 3 μM). (F) PPT1 enzymatic activity in A375P cells following treatment with DQ660 or DQ661 (3 hr, 3 μM). (G) In vitro binding of DQ661 to PPT1. Differential Scanning Calorimetry of 1 mg/mL PPT1 (29.4 μM) in the absence (black) or presence of DQ661 (100 μM, green). Tm: melting temperature. (H) A375P cells were treated with DQ661 (0 – 240 minutes, 3 μM). CD44 palmitoylation measured using the acyl-biotinyl exchange (ABE) assay increases with DQ661 treatment compared with control. Samples not treated with hydroxylamine (-HAM) serve as a negative control. (I) A375P cells were treated with 25 nM PPT1 siRNA or non-targeting (siNT). Cells were transfected overnight in the absence of serum. Upon serum restimulation, cells were collected 3 or 6 hr thereafter and lysate was immunoblotted. (J) A375P cells were treated with DQ661 (1 hr, 3 μM) in the presence or absence of NTBHA (2 mM). Lysate was immunoblotted.
Figure 5
Figure 5. DQ661 functionally inhibits mTORC1
(A) Schematic depicting interactions between vATPase/Ragulator/Rag/mTORC1 and Rheb on the lysosome surface. (B) A375P cells were treated with DQ661 (6 hr, 3 μM) and membrane fractions were immunoblotted for vATPase machinery. Densitometry shown below corresponding blot. (C) PLA was performed on A375P cells treated with DQ661 (6 hr, 1 μM) for the p18 (Ragulator) – V1A physical interaction. (D) A375P cells were treated with DQ661 or QN (6 hr, 3 μM) and IF microscopy was performed to detect changes in p18/LAMP2 localization or (E) RagC/LAMP2 localization or (F) mTOR/LAMP2 localization. Below each panel is the respective co-localization analysis. Data are represented as mean ± SD. N = 50 cells per condition (ANOVA with Dunnett’s multiple comparison test). *p<0.0001 versus Untreated group. (G) A375P cells were treated with DQ661 (6 hr, 3 μM) and lysosome fractions were isolated and immunoblotted. (H) A375P cells were treated with DQ661 (3 hr, 3 μM) and PLA was performed. Blue represents DAPI and red fluorescence represents mTOR-RagC interactions. (I) HEK293T cells expressing FLAG-RagB were treated with DQ661 (4 hr, 3 μM) and immunoprecipitation lysates were probed for mTOR. (J) PLA was performed on A375P cells were treated with rapamycin (3 μM), torin-1 (3 μM), baf (100 nM), siramesine (8 μM), or DQ661 (1 μM) for 3 hr. Blue is DAPI and red fluorescence reflects mTOR-Rheb interaction. (K) A375P cells were treated with HDSF (6 hr, 40 μM) and membrane fractions were isolated and immunoblotted. (L) A375P cells were treated with NT or PPT1 siRNA for 24 hours. Membranes were subsequently fractionated and immunoblotted.
Figure 6
Figure 6. DQ661 has significant single-agent in vivo activity in melanoma xenograft model
(A) 1205Lu cells were injected subcutaneously in the flanks of NSG-mice (2 × 106 cells/mouse) and grown until tumors were palpable. Mice (n=8/arm) were treated with vehicle control (water, i.p.), quinacrine (8 mg/kg, i.p.), DQ660 (8 mg/kg, i.p.) or DQ661 (8 mg/kg, i.p.). Treatments were given as shown by the black arrows, (2 days on, 2 days off). Mean +/− SEM is presented; *P<0.05. A linear mixed-effect model was used to test the difference of the tumor growth trends among treatment groups. (B) Mean +/− SD tumor growth rate. (C) Tumor tissues stained for phopsho-H2AX Ser139. Scale bar: 40 μm. (D) Representative electron micrographs of tumors harvested from mice after 2 days of treatment with each agent. Arrows: autophagic vesicles. Scale bars represent 730 nm. (E) 1205Lu cells were injected subcutaneously in the flanks of NSG-mice (2 × 106 cells/mouse) and grown until tumors were palpable (1 – 2 weeks). Mice (n=8 per arm) were treated with water (i.p.) or DQ661 (4 mg/kg, i.p.). Black arrows: treatment schedule (2 days on, 2 days off) Mean +/− SEM are presented for daily tumor volumes; *p<0.05 (F) Average tumor growth rate. (G) Protein lysate from mouse tumors at the end of the experiment was immunoblotted as indicated. (H) In vivo PLA in mice treated with vehicle control or 4 mg/kg DQ661. Red fluorescence indicates mTOR-Rheb interaction, blue represents DAPI staining of nuclei. Depicted below is quantitation of PLA signal intensity. The graph reflects mean intensity, N=200 cells were quantified from 2 mice in each arm.
Figure 7
Figure 7. DQ661 improves survival in colon cancer model and potentiates activity of gemcitabine in KPC pancreatic cancer syngeneic model
(A) HT-29 cells were injected s.c. into the flanks of NSG-mice and grown until tumor were palpable. Mice (n=8/arm) were treated with water or DQ661 (4 mg/kg) i.p. DQ661 treated mice were treated 2 days on, 2 days off. Mean +/− SEM is presented. (B) Survival curve for (A) displaying the time it took for mice to reach death (defined as time when tumor volume exceeded 1000mm3). (C) A linear mixed-effect model was used to test the difference of the tumor growth trends among treatment groups (A) Mean +/− SEM tumor growth rate. (D) MTT assay of KPC cell lines 4662 and G43 treated with gemcitabine (72 hr, 3 – 30 nM). *p<0.05. (E) G43 cells were treated with gemcitabine (24 hr, 10 nM) and lysate was immunoblotted. (F) G43 cells were treated chronically for 2 weeks with gemcitabine (3 – 30 nM) in the presence or absence of DQ661 in colony formation assays. Cells were stained with crystal violet and imaged. (G) G43 cells were injected subcutaneously into the flanks of C57BL/6 mice (2 × 106 cells/mouse. Once palpable, mice (n=8 mice per treatment arm) were treated with vehicle (PBS), gemcitabine (120 mg/kg, i.p.), DQ661 (4 mg/kg, i.p., 2 days on, 3 days off) or a combination of gemcitabine and DQ661. (H) A linear mixed-effect model was used to test the difference of the tumor growth trends among treatment groups (D) Mean +/− SEM tumor growth rate.

Comment in

Similar articles

See all similar articles

Cited by 32 articles

See all "Cited by" articles

MeSH terms

Feedback