Small Molecule Inhibition of the Autophagy Kinase ULK1 and Identification of ULK1 Substrates - PubMed (original) (raw)

. 2015 Jul 16;59(2):285-97.

doi: 10.1016/j.molcel.2015.05.031. Epub 2015 Jun 25.

Matthew G H Chun 1, Mitchell Vamos 2, Haixia Zou 2, Juan Rong 2, Chad J Miller 3, Hua Jane Lou 3, Dhanya Raveendra-Panickar 2, Chih-Cheng Yang 4, Douglas J Sheffler 2, Peter Teriete 2, John M Asara 5, Benjamin E Turk 3, Nicholas D P Cosford 6, Reuben J Shaw 7

Affiliations

Small Molecule Inhibition of the Autophagy Kinase ULK1 and Identification of ULK1 Substrates

Daniel F Egan et al. Mol Cell. 2015.

Abstract

Many tumors become addicted to autophagy for survival, suggesting inhibition of autophagy as a potential broadly applicable cancer therapy. ULK1/Atg1 is the only serine/threonine kinase in the core autophagy pathway and thus represents an excellent drug target. Despite recent advances in the understanding of ULK1 activation by nutrient deprivation, how ULK1 promotes autophagy remains poorly understood. Here, we screened degenerate peptide libraries to deduce the optimal ULK1 substrate motif and discovered 15 phosphorylation sites in core autophagy proteins that were verified as in vivo ULK1 targets. We utilized these ULK1 substrates to perform a cell-based screen to identify and characterize a potent ULK1 small molecule inhibitor. The compound SBI-0206965 is a highly selective ULK1 kinase inhibitor in vitro and suppressed ULK1-mediated phosphorylation events in cells, regulating autophagy and cell survival. SBI-0206965 greatly synergized with mechanistic target of rapamycin (mTOR) inhibitors to kill tumor cells, providing a strong rationale for their combined use in the clinic.

Copyright © 2015 Elsevier Inc. All rights reserved.

PubMed Disclaimer

Figures

Figure 1

Figure 1. Determination of the optimal ULK1 consensus phosphorylation motif

(A) A positional scanning peptide library (PSPL) approach was used to define the optimal ULK1 consensus phosphorylation motif. A spatially arrayed PSPL was subjected to in vitro phosphorylation using active ULK1 and radiolabeled ATP. Each peptide contained a fixed residue at one of nine positions relative to the centrally fixed phosphoacceptor (an equal mix of Ser and Thr), with the remaining positions being equimolar mixtures of the 17 amino acids (excluding Ser, Thr, and Cys). Aliquots of each reaction were spotted onto an avidin membrane, which was washed, dried, and exposed to a phosphor-storage screen to quantify radiolabel incorporation into each peptide. The heat map was generated in Microsoft Excel using normalized and log transformed signals corresponding to peptides with the indicated amino acid at the indicated position. (B) Scaled sequence logo showing ULK1 sequence specificity obtained from the PSPL data. ULK1 prefers hydrophobic residues at the –3, +1, and +2 positions. The logo was generated using EnoLogos, with the height of the letter being proportional to the signal for the corresponding amino acid at the indicated position. (C) Rates of ULK1 phosphorylation for ULKtide variants with individual amino acid substitutions are shown (normalized to unmodified ULKtide peptide). ULKtide was estimated to have a 15 μM Km. Peptide phosphorylation was assayed at a 5 μM concentration using a radiolabeled kinase assay. Incorporation of γ-32P-ATP into these peptides was determined by a phosphocellulose filter-binding assay.

Figure 2

Figure 2. Identification of novel ULK1-dependent phosphorylation sites in vivo

(A) Myc-tagged wild-type ULK1 (WT ULK1; top) or Myc-tagged kinase-inactive ULK1 (KI ULK1; bottom) was transfected into HEK-293T cells along with wild-type Flag-tagged Atg101 (Flag-Atg101) and immunoprecipitated with M2 agarose. The immunoprecipitate was run out on an SDS-PAGE gel, stained with coomassie, and the band corresponding to Flag-Atg101 was cut out, isolated, and subjected to tryptic digest and LC/MS/MS analysis. The phosphorylated sites that conform to the optimal ULK1 phosphorylation motif that were identified by this analysis are boxed. Green bars indicate peptide coverage, and purple highlights indicate phosphorylation events. (B) WT ULK1 or KI ULK1 was transfected into HEK-293T cells along with Flag-Atg101 or Flag-Atg101 serine-to-alanine point-mutants. The specific mutants used in this analysis are indicated by the position(s) of the substituted amino acid (top). Cellular lysates were isolated 24 hours post-transfection, run out on an SDS-PAGE gel containing the Phos-Tag reagent (middle) or a standard SDS-PAGE gel lacking the Phos-Tag reagent (bottom), and transferred to PVDF membranes, which were subsequently immunoblotted with the indicated antibodies. (C) Same as 2A except using wild-type Flag-tagged Beclin1 as a substrate. (D) Same as 2A except using wild-type Flag-tagged Ambra1 as a substrate. (E) Same as 2B except using wild-type Flag-tagged Syntenin-1 (Flag-Syntenin-1) or Flag-tagged Syntenin-1 serine-to-alanine point mutants. The specific mutants used in this analysis are indicated by the position(s) of the substituted amino acid (top). Cellular lysates were isolated 24 hours post-transfection, run out on an SDS-PAGE gel containing the Phos-Tag reagent (middle) or a standard SDS-PAGE gel lacking the Phos-Tag reagent (bottom), and transferred to PVDF membranes, which were subsequently immunoblotted with the indicated antibodies. (F) Alignment of all novel ULK1 phosphorylation sites identified in this analysis, alongside the STING phosphorylation site which was previously reported as a ULK1 site (Konno et al., 2013). Phosphorylation sites that contain residues conforming to the optimal ULK1 phosphorylation motif at the –3 (green), +1 (green), and +2 (yellow) positions are highlighted.

Figure 3

Figure 3. Vps34 Ser249 is a novel ULK1 phosphorylation site in vivo

(A) Either Myc-tagged wild-type ULK1 (WT ULK1; bottom) or Myc-tagged kinase-inactive ULK1 (KI ULK1; top) was transfected into HEK-293T cells along with wild-type Flag-tagged Vps34 (WT Vps34) and immunoprecipitated with M2 agarose. The immunoprecipitate was run out on an SDS-PAGE gel, stained with coomassie, and the band corresponding to WT Vps34 was cut out, isolated, and subjected to tryptic digest and LC/MS/MS analysis. Green bars indicate peptide coverage, and purple highlights indicate phosphorylation events. Arrow indicates serine 249. (B) Clustal alignment of Vps34 serine 249 across species shows that it is conserved throughout evolution and conforms to the optimal ULK1 phosphorylation motif. (C) An in vitro kinase assay was performed using Flag-tagged WT Vps34 (Vps34 WT) or a Flag-tagged serine-to-alanine point mutant Vps34 (Vps34 S249A) as substrates for either WT ULK1 or KI ULK1. The in vitro kinase assay was performed in the presence of radiolabeled γ-32P-ATP (top). Vps34 WT, Vps34 S249A, WT ULK1, and KI ULK1 were produced in HEK-293T cells (bottom). (D). Vps34 WT or Vps34 S249A and WT ULK1 or KI ULK1 were transfected into HEK-293T cells. Cellular lysates were isolated 24 hours post-transfection, run out on an SDS-PAGE gel, and transferred to PVDF membranes, which were subsequently probed with the indicated antibodies. Arrow indicates a mobility shift representative of phosphorylation that only occurs with the Vps34 WT and WT ULK1 combination. (E). HEK-293T cells were transfected with Vps34 WT or Vps34 S249A and WT ULK1, wild-type Myctagged ULK2 (ULK2), or wild-type Myc-tagged ULK3 (ULK3). Cellular lysates were isolated 24 hours post-transfection and immunoblotted with the indicated antibodies.

Figure 4

Figure 4. In cellulo screen identifies SBI-0206965 as a potent ULK1 kinase inhibitor against its downstream substrate phosphorylation sites

(A) Chemical structure of SBI-0206965 (“6965”), the lead ULK1 competitive kinase inhibitor. (B) The IC50 value for 6965 against wild-type ULK1 and ULK2 was determined using an in vitro kinase assay. Wild-type ULK1 (left) and wild-type ULK2 (right) were assayed using 10 μM MBP in the presence of 30 μM radiolabeled γ-32P-ATP. 6965 was tested in triplicate in a 10-dose IC50 mode with 3-fold serial dilution and a starting dose of 1 μM. (C). Human embryonic kidney cells (HEK-293T) were transfected with wild-type (WT) or kinase inactive (KI) Myc-tagged ULK1 and wild-type Flag-tagged Vps34 (WT Vps34). 24 hours post-transfection, cells were treated with a panel of putative ULK1 competitive inhibitors in a dose response manner (1, 10, 50 μM). Cellular lysates were isolated after 1 hour of treatment and immunoblotted with the indicated antibodies. Representative results for SBI-0206965 are shown. (D) HEK-293T cells were transfected with WT or KI ULK1 and WT Vps34 (left) or wild-type Flag-tagged Beclin1 (WT Beclin1; right). 24 hours post-transfection, cells were treated with 6965 (10 μM) or DMSO. Cellular lysates were isolated after 1 hour of treatment and immunoblotted with the indicated antibodies. (E) Wild-type or Ulk1/Ulk2 double knockout mouse embryonic fibroblasts (MEFs) (Cheong et al., 2011) were treated with fresh media (Dulbecco’s modified Eagle medium [DMEM] containing 10% fetal bovine serum [FBS]) containing 1 μM INK128, 1 μM AZD8055, or DMSO or with starvation media (Earle’s balanced salt solution [EBSS]) in the presence or absence of 10 μM 6965. Cellular lysates were isolated after 1 hour of treatment and immunoblotted with the indicated antibodies.

Figure 5

Figure 5. SBI-0206965 is a highly selective kinase inhibitor

(A) The kinase selectivity profile for SBI-0206965 was determined using the DiscoveRx KINOMEscan profiling service (

http://www.discoverx.com/

). Briefly, 6965 was screened at a 1 μM dose for its ability to impair binding of a panel of 456 kinases to substrate in an in vitro binding assay. Scores for the primary screen hits are reported as a percent of the DMSO control (% Control). Lower scores reflect stronger inhibitory effects of 6965 on the target kinase. (B) A TREEspot interaction map (

http://www.discoverx.com/tools-resources/interaction-maps

) was generated to visually represent the selectivity profile for 6965 against the panel of kinases tested in 5A. Kinases whose binding was inhibited by 6965 are marked with red circles, with larger circles indicating stronger inhibitory effects. Kinases tested in this analysis are arrayed according to their phylogenetic groupings in the human kinome. (C) The list of the kinases whose binding was found to be inhibited by 6965 to less than 4% of the DMSO negative control in the screen described in 5A. Kinases whose binding was inhibited to less than 1% of the DMSO negative control are highlighted in red. (D) In vitro kinase assays were performed for the kinases highlighted in 5C. These assays were performed in the presence of 6965 in a dose response manner to identify the IC50 value for 6965 for each of these individual kinases. Kinases whose IC50 value was less than 1-fold difference than ULK1 are highlighted in yellow. Of the remaining kinases, those kinases whose IC50 value was less than what was identified for ULK2 are highlighted in brown. In vitro kinase assays performed by Reaction Biology (

http://www.reactionbiology.com/

). (E) Immunoflourescence imaging of A549 cells treated in the presence or absence of 5 μM 6965 for 2 hours followed by 4 μM of the mTOR catalytic inhibitor AZD8055 for 24 hours. Autophagic vacuoles were detected using the Cyto-ID autophagy detection kit (Chan et al., 2012) and are visualized in green while cell nuclei were counterstained by DAPI and are visualized in blue. (F) PC3 human prostate cells that stably express a construct encoding LC3 fused to green fluorescent protein (GFP-LC3) were transfected with siRNAs against the top 18 kinases whose binding was shown to be inhibited by 6965 (Figure 5A). 48 hours after RNAi transfection, the cells were treated with 1 μM of the catalytic mTOR inhibitors AZD8055 or INK128 for 4 hours and assessed for the presence of GFP-LC3 puncta (see Experimental Procedures). The average number of GFP-LC3 puncta and standard deviation for each siRNA and drug treatment are shown. Control immunoblots demonstrating siRNA efficiency in Figure S5E. [“SRC” = c-Src; “c-Src kinase” = CSK] (G) Representative immunofluorescence images for the data shown in 5F. GFP-LC3 puncta are visualized in green and cell nuclei, which were counterstained with DAPI, are visualized in blue.

Figure 6

Figure 6. SBI-0206965 synergizes with starvation and mTOR inhibition to induce an enhanced apoptotic response

(A) Wild-type mouse embryonic fibroblasts (MEFs) were treated with fresh media (Dulbecco’s modified Eagle medium [DMEM] containing 10% fetal bovine serum [FBS]) or starvation media (Earle’s balanced salt solution [EBSS]) and 10 μM 6965 or DMSO for 18 hours. Cells were collected, stained with 7-AAD and PE-AnnexinV, and quantified for apoptosis by fluorescence-activated cell sorting (FACS) analysis. Red numbers indicate the percentage of 7-AAD/PE-AnnexinV doublepositive cells, representative of end-stage apoptosis and death. (B) Western analysis of wild-type MEFs treated as in 6A for 1, 6, and 18 hours. Cellular lysates were immunoblotted with the indicated antibodies. Chloroquine (CQ) at 20 μM serves as a positive control for autophagy inhibition. (C) Same as 6A except using a lung cancer cell line derived from a genetically engineered mouse model of lung cancer driven by mutant Kras and deletion of p53. Red numbers indicate the percentage of AnnexinV-positive cells. (D) Same as 6A except using the human glioblastoma cell line U87MG. Red numbers indicate the percentage of AnnexinV-positive cells.

Figure 7

Figure 7. SBI-0206965 converts the cytostatic effects of three different mTOR inhibitors into a cytotoxic apoptotic response in lung cancer cells

(A) A549 human lung cancer cells were treated with the catalytic mTOR inhibitor AZD8055 (1 μM) or DMSO and increasing doses of 6965. Cells were treated for 72 hours and then collected, stained wit PE-AnnexinV, and quantified by FACS analysis. Red numbers indicate the percentage of AnnexinV-positive cells, representing cells actively undergoing apoptosis. (B) A549 cells were treated with DMSO, 1 μM INK128, 500 nM rapamycin, or 1 μM AZD8055 in the presence or absence of 10 μM 6965 or 10 μM chloroquine (CQ). Cells were treated for 48 hours and then collected, stained with PE-AnnexinV, and quantified by FACS analysis. (C) Western analysis of A549 cells treated as in 7B. Cellular lysates were immunoblotted with the indicated antibodies. (D) Model: mTOR inhibitors trigger autophagy by preventing mTOR-dependent inhibitory phosphorylation events on autophagy regulators, including the serine/threonine kinase ULK1. Thus mTOR inhibition triggers autophagy, which promotes cell survival. Co-administration of an mTOR inhibitor with a ULK1 catalytic inhibitor results in a loss of autophagy induction and prevents the cell survival effect resulting from mTOR inhibition, converting a cytostatic response into a cytotoxic response.

Comment in

References

    1. Benjamin D, Colombi M, Moroni C, Hall MN. Rapamycin passes the torch: a new generation of mTOR inhibitors. Nat Rev Drug Discov. 2011;10:868–880. - PubMed
    1. Bissler JJ, McCormack FX, Young LR, Elwing JM, Chuck G, Leonard JM, Schmithorst VJ, Laor T, Brody AS, Bean J, et al. Sirolimus for angiomyolipoma in tuberous sclerosis complex or lymphangioleiomyomatosis. N Engl J Med. 2008;358:140–151. - PMC - PubMed
    1. Chan EY. mTORC1 phosphorylates the ULK1-mAtg13-FIP200 autophagy regulatory complex. Sci Signal. 2009;2:e51. - PubMed
    1. Chan LL, Shen D, Wilkinson AR, Patton W, Lai N, Chan E, Kuksin D, Lin B, Qiu J. A novel image-based cytometry method for autophagy detection in living cells. Autophagy. 2012;8:1371–1382. - PMC - PubMed
    1. Cheong H, Lindsten T, Wu J, Lu C, Thompson CB. Ammonia-induced autophagy is independent of ULK1/ULK2 kinases. Proc Natl Acad Sci U S A. 2011;108:11121–11126. - PMC - PubMed

Publication types

MeSH terms

Substances

Grants and funding

LinkOut - more resources