(ULK1) that induces cytoprotective autophagy for Parkinson's disease

Subscriber access provided by Queen Mary, University of London

A small-molecule activator of UNC-51-like kinase 1 (ULK1) that induces cytoprotective autophagy for Parkinson's disease treatment Liang Ouyang, Lan Zhang, Shouyue Zhang, Dahong Yao, Yuqian Zhao, Guan Wang, Leilei Fu, Peng Lei, and Bo Liu J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01575 • Publication Date (Web): 21 Mar 2018 Downloaded from http://pubs.acs.org on March 21, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Medicinal Chemistry

A small-molecule activator of UNC-51-like kinase 1 (ULK1) that induces cytoprotective autophagy for Parkinson's disease treatment Liang Ouyang#, Lan Zhang#, Shouyue Zhang, Dahong Yao, Yuqian Zhao, Guan Wang, Leilei Fu, Peng Lei, Bo Liu* State Key Laboratory of Biotherapy and Cancer Center, West China Hospital, Sichuan University, and Collaborative Innovation Center for Biotherapy, Chengdu 610041, China

ABSTRACT UNC-51-like kinase 1 (ULK1), the yeast Atg1 ortholog, is the sole serine-threonine kinase and initiating enzyme in autophagy, which may be regarded as a target in Parkinson's disease (PD). Herein, we discovered a small molecule 33i (BL-918) as a potent activator of ULK1 by structure-based drug design. Subsequently, some key amino acid residues (Arg18, Lys50, Asn86 and Tyr89) were found to be crucial to the binding pocket between ULK1 and 33i by site-directed mutagenesis. Moreover, we found that 33i induced autophagy via the ULK complex in SH-SY5Y cells. Intriguingly, this activator displayed a cytoprotective effect on MPP+-treated SH-SY5Y cells, as well as protected against MPTP-induced motor dysfunction and loss of dopaminergic neurons by targeting ULK1-modulated autophagy in mouse models of PD. Together, these results demonstrate the therapeutic potential to target ULK1, and 33i, the novel activator of ULK1 may serve as a candidate drug for future PD treatment.

Keywords: UNC-51-like kinase 1; Autophagy; Parkinson's disease; ULK1 activator; The ULK complex

1

ACS Paragon Plus Environment

Journal of Medicinal Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 50

INTRODUCTION Autophagy is an evolutionarily conserved and multi-step lysosomal degradation process that may degrade some long-lived proteins or damaged organelles in cells.1,2 It is well-known that autophagy can be modulated by a number of autophagy-related (Atg) genes, especially UNC-51-like kinase 1 (ULK1) and its complex.3,4 As the homolog of Atg1 in mammals, ULK1 has similar functions with Atg1 which was found as the ﬁrst autophagy-related gene in yeast.5 And, the ULK complex formed by ULK1, mAtg13, FIP200 and Atg101, is required to initiate the autophagic process.6,7 Interestingly, AMP-Activated Protein Kinase (AMPK) can activate ULK1 via directly phosphorylating ULK1 or relieving the mammalian target of rapamycin complex 1 (mTORC1) negative regulation of ULK1 activity.8,9 Subsequently, the downstream signaling pathways can be regulated by the activation of ULK1 and its complex.10,11 As autophagy is a one of key mechanisms to maintain the nutrient and energy homeostasis of cells, its dysregulation may impede the clearance of abnormal proteins and damaged organelles, which is identified as one of pathogenesis of neurodegenerative diseases, such as Parkinson’s disease (PD).12,13 PD is often characterized by the accumulation of α-synuclein that can be visible as Lewy's body inclusions and by loss of nigrostriatal dopaminergic neurons.14 Autophagy may promote the removal of such aggregated proteins for protecting neuron cells against the toxicity,15-17 which would be regarded as an attractive approach for the treatment of PD.18,19 Further, applications of autophagy enhancers have been found to alleviate dopaminergic neurodegeneration of PD models in vitro and in vivo, indicating that autophagic modulators have potential therapeutic effects on PD.20-22 More recently, ULK1 has been reported to trigger starvation-induced cytoprotective autophagy in SH-SY5Y cells, thus serving as a potential target for PD therapy.23 Therefore, we hypothesize that discovery of a new activator of ULK1 to regulate cytoprotective autophagy would be a promising avenue to treat PD. Thus, in this study, we discovered a novel ULK1 activator 33i (BL-918) that potently activated ULK1. 33i could induce cytoprotective autophagy via the ULK1 complex in SH-SY5Y cells, and also exerted its neuroprotective effects by targeting ULK1-modulated 2


Page 3 of 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


autophagy in a MPTP-induced PD mouse model. Together, these results demonstrate the therapeutic potential to target ULK1 in PD, and 33i, as the activator of ULK1 may serve as a candidate drug in PD therapy.

RESULTS Structure-based ligand design for ULK1. The ULK complex which consists of ULK1, mAtg13, FIP200 and Atg101, is able to initiate the autophagic process.11 ULK1 increases the phosphorylation of mAtg13 and interacts with FIP200 and Atg101 to form the ULK complex; thereby eventually triggering autophagy.24 Given that the ULK complex is required to trigger autophagy by an activator of ULK1, we designed a small molecule that could activate ULK1 and the rest of ULK complex (Fig. 1A). A recent study has reported a kinase activator-AMPK complex structure (PDB code: 4ZHX25). Of note, ULK1 and AMPK are the members of the serine/threonine kinase family with similar molecular structures; therefore, they could be speculated to have similar kinase binding modes. Compared ULK1 kinase domain (Gly7-Ala280) (PDB code: 4WNP26) with the kinase activator-AMPK crystal structure, we found that they share a similar structure with a low RMSD value of 0.985 in kinase domain, indicating that ULK1 may possess a putative activation site in the corresponding region (Glu9-Arg18, Ile48-Leu53 and Gln82-Tyr89) which is far away from the known inhibitor binding site (ATP-binding site) (Fig. S1 and Fig. 1B).26,27 In addition, we applied solvent accessible surface (SAS) calculation to further identify the hot spots (Gly7-Lys55, Asp80-Leu90) that were druggable contact surfaces covering the putative activator binding site (Fig. 1B).28 Based upon high-throughput screening from ZINC database, we achieved the top 50 potential small-molecule compounds according to their scores and energies. Subsequently, considering the diversity of their chemical structures, we reselected the top 20 hits (docked compounds). (Fig. 1C and Fig. S2). Combined with kinase assay of ULK1, we found 15 of the top-ranked 20 hits enhanced the ULK1 kinase activity at different levels. And, we selected compound 3 as the best leading compound (Fig. 1D). According to the interactions between ULK1 and 3, four key amino acid 3



Page 4 of 50

residues (Asn86, Met84, Lys50 and Tyr89) were identified, which may provide a clue to optimize 3 for a better ULK1 activator. (Fig. 1E).

Fig. 1 Structure-based ligand design for ULK1. (A) The schematic model of autophagy initiation process regulated by the ULK complex activation. (B) The kinase domain structure of ULK1 (PDB code: 4WNP) was colored in grey. Hot spots in activator binding sites were colored in blue and potential ULK1 activator binding sites were colored in red. (C) Docking the top-ranked 20 candidate compounds. (D) The top-ranked 20 candidate compounds (1 µM) were screened for kinase activity by ULK1 kinase assay. Each point showed luminescence normalized to basal level at 10 ng ULK1. (E) A detailed view showed the binding conformation between ULK1 and 3. Four key amino acid residues (Asn86, Met84, Lys50 and Tyr89) were surrounded an activator binding pocket with 3.

Structural optimization and discovery of the ULK1 activator Next, we drove structural optimization of the lead compound rationally (Fig. 2A). The bridging oxygen atom was replaced by ester group to form an additional hydrogen bond with Tyr89, and the imidazole ring was transformed into piperazine ring to yield compounds 24a-s (Fig. 2B and Scheme 1). Based upon ULK1 kinase activity, 24c bear the best maximum efficacy (Emax) with half maximal effective concentration (EC50) of 54.84 nM and, resulting in a 4.8-fold (EC50) and 2.1-fold (Emax) improvement compared to 3 (Table 1 and 4). According to the interactions between 24c and ULK1, an additional 4


Page 5 of 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Pi-Cation interaction and two hydrogen bonds were generated between 24c and Arg18, Lys50 and Tyr89. Subsequently, the D-(-)-2-Phenylglycine portion of the scaffold offered a straightforward entry for the generation from phenylamine to obtain compounds 29a-t (Fig. 2B and Scheme 2). Compared with 24c, 29n possessed a slight improvement of EC50 and Emax in ULK1 kinase activity, but it showed 2-fold improvement in autophagy activity measured by flow cytometry analysis of monodansylcadaverine (MDC) staining (Fig. S3 and S4, Table 2 and 4). The residues located on the edge of active pocket were not fully utilized, such as Asn86 and Tyr89. Thus, we synthesized some compounds 32a-i, 33a-i, 34a-i, 35a-i and 36a-i by introducing a urea group as the linker (Fig. 2B and Scheme 3). The most potent compound 33i had 1.7-fold (EC50) and 1.35-fold (Emax) improvement in ULK1 kinase activity, as well as a 1.5-fold improvement in autophagy activity over 29n, respectively (Fig. S3, 4 and 5, Table 3 and 4). Additionally, AMPK and eEF2K could not be activated by 33i in the kinase assays (Fig. S6), indicating 33i has a priority in activating ULK1. Moreover, 33i maintained all the critical interactions observed in compounds 24c and 29n. Three additional halogen bonds were observed by the trifluoromethyl moiety with Ala85 and Asn86, which reinforced this binding conformation, and the carbonyl of scaffold and phenolic hydroxyl of Tyr89 formed a hydrogen bond as expected (Fig. 3A). As mentioned above, the crucial residues Arg18 and Lys50 are required for the high affinity of 33i. In addition, the residues Ala85, Asn86 and Tyr89, can improve its potency and selectivity.

5



Page 6 of 50

6


Page 7 of 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


HN

Boc COOH

HN (a)

Boc H N

NH2 R1

(b)

O 26

27a,b;30a-c

H N

HN R1

(c)

X

R2 NH

O 28a,b;31a-c

H N

R1

O 32a-i;33a-i;34a-i;35a-i;36a-i

32a: X=O, R1=(S)-1-phenylethly, R2=4-methoxyphenyl 32b: X=O, R1 =(S)-1-phenylethly, R2=2-chlorophenyl 32c: X=O, R1=(S)-1-phenylethly, R2=3-chlorophenyl 32d: X=O, R1 =(S)-1-phenylethly, R2=4-methylphenyl 32e: X=O, R1=(S)-1-phenylethly, R2=2,6-dimethylphenyl 32f: X=O, R1=(S)-1-phenylethly, R2=4-trifluoromethylphenyl 32g: X=O, R1 =(S)-1-phenylethly, R2=1-naphthyl 32h: X=S, R1=(S)-1-phenylethly, R 2=phenyl 32i: X=S, R1=(S)-1-phenylethly, R2=3,5-ditrifluoromethylphenyl 33a: X=O, R1=2,4-difluorophenyl, R2=4-methoxyphenyl 33b: X=O, R1 =2,4-difluorophenyl, R2=2-chlorophenyl 33c: X=O, R1=2,4-difluorophenyl, R2=3-chlorophenyl 33d: X=O, R1 =2,4-difluorophenyl, R2=4-methylphenyl 33e: X=O, R1=2,4-difluorophenyl, R2=2,6-dimethylphenyl 33f: X=O, R1=2,4-difluorophenyl, R2=4-trifluoromethylphenyl 33g: X=O, R1 =2,4-difluorophenyl, R2=1-naphthyl 33h: X=S, R1=2,4-difluorophenyl, R2 =phenyl 33i: X=S, R1=2,4-difluorophenyl, R 2=3,5-ditrifluoromethylphenyl 34a: X=O, R1=2-methylphenyl, R2=4-methoxyphenyl 34b:X=O, R1 =2-methylphenyl, R2=2-chlorophenyl 34c: X=O, R1=2-methylphenyl, R2=3-chlorophenyl 34d: X=O, R1=2-methylphenyl, R2 =4-methylphenyl 34e: X=O, R1=2-methylphenyl, R2=2,6-dimethylphenyl

34f: X=O, R1 =2-methylphenyl, R2=4-trifluoromethylphenyl 34g: X=O, R1=2-methylphenyl, R2=1-naphthyl 34h: X=S, R1 =2-methylphenyl, R2=phenyl 34i: X=S, R1 =2-methylphenyl, R2=3,5-ditrifluoromethylphenyl 35a: X=O, R1=3-bromophenyl, R2 =4-methoxyphenyl 35b: X=O, R1=3-bromophenyl, R2=2-chlorophenyl 35c: X=O, R1=3-bromophenyl, R2 =3-chlorophenyl 35d: X=O, R1=3-bromophenyl, R2=4-methylphenyl 35e: X=O, R1=3-bromophenyl, R2 =2,6-dimethylphenyl 35f: X=O, R1 =3-bromophenyl, R2=4-trifluoromethylphenyl 35g: X=O, R1=3-bromophenyl, R2=1-naphthyl 35h: X=S, R1 =3-bromophenyl, R2=phenyl 35i: X=S, R1 =3-bromophenyl, R2 =3,5-ditrifluoromethylphenyl 36a: X=O, R1=3,4-dimethyloxyphenylethyl, R2=4-methoxyphenyl 36b: X=O, R1=3,4-dimethyloxyphenylethyl, R2 =2-chlorophenyl 36c: X=O, R1=3,4-dimethyloxyphenylethyl, R2=3-chlorophenyl 36d: X=O, R1=3,4-dimethyloxyphenylethyl, R2=4-methylphenyl 36e: X=O, R1=3,4-dimethyloxyphenylethyl, R2=2,6-dimethylphenyl 36f: X=O, R1 =3,4-dimethyloxyphenylethyl, R 2=4-trifluoromethylphenyl 36g: X=O, R1=3,4-dimethyloxyphenylethyl, R2=1-naphthyl 36h: X=S, R1=3,4-dimethyloxyphenylethyl, R2=phenyl 36i: X=S, R1=3,4-dimethyloxyphenylethyl, R2 =3,5-ditrifluoromethylphenyl

Scheme 3 Reagents and conditions: (a) N-Methylmorpholine, iso-butylchloroformate, THF, -20 ℃, R 1NH2; (b) HCl/MeOH, r.t.; (c) CH2Cl 2 , TEA, R2 NCO or R2NCS, r.t.

Fig. 2 Structural optimization and discovery of the ULK1 activator. (A) Biological evaluation guided optimization towards 33i. (B) ULK1-33i interactions included hydrogen bonds (green dash), Pi-Sulfur 7



Page 8 of 50

(yellow dash), Halogen interaction (blue dash), Pi-Cation (orange dash) and Carbon Hydrogen bond (light blue). Insets showed selected data from the structure-activity relationship.

Table 1. Kinase activities of compounds 24a-s against recombinant human ULK1 and relevant MDC positive ratio in SH-SY5Y cells

Kinase

Compound

R1

R2

activity%(100 nM)

a

MDC positive ratio%(1 µM)

b

24a

155.76±4.09

9.59±0.91

24b

148.68±4.28

6.56±1.25

24c

197.06±6.55

14.72±0.80

24d

180.73±3.31

12.49±0.59

24e

150.60±7.49

10.04±0.44

24f

141.88±3.86

7.41±0.44

24g

138.59±4.07

6.22±0.57

24h

143.06±4.56

9.99±0.39

8


Page 9 of 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


24i

158.91±4.50

11.42±0.67

24j

151.86±7.45

11.80±0.65

24k

140.89±6.17

8.56±0.92

24l

162.89±8.39

13.42±0.59

138.92±8.63

4.51±0.84

24n

162.08±11.02

11.58±1.15

24o

170.86±7.79

14.17±0.28

24p

167.93±4.95

11.05±0.57

24q

150.37±6.81

11.76±0.52

24r

154.81±6.58

11.21±0.40

24s

147.00±5.78

9.20±0.40

24m

a

CH3

Each compound was determined by three independent experiments (values are the mean ±

SEM). b

Determined by flow cytometry analysis using MDC staining (values are the mean ± SEM).

Table 2. Kinase activities of compounds 29a-t against recombinant human ULK1 and relevant MDC positive ratio in SH-SY5Y cells

9



Page 10 of 50

Kinase

Compound

R1

R2

activity%(100 nM)

a

MDC positive ratio%(1 µM)b

29a

137.05±5.74

10.96±0.81

29b

142.75±6.47

13.52±0.88

29c

146.10±5.67

15.30±0.58

29d

139.01±5.39

13.80±0.60

29e

142.11±5.59

16.61±1.01

29f

139.88±8.68

16.21±0.61

29g

147.13±8.17

18.13±0.78

10


Page 11 of 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


29h

144.09±5.26

16.75±0.40

29i

158.24±6.19

21.04±0.83

29j

163.77±8.51

24.47±0.56

29k

152.80±6.20

20.31±1.08

29l

161.03±9.18

23.44±1.24

29m

148.19±4.23

20.55±0.59

29n

201.73±9.12

30.13±1.48

29o

171.04±6.13

28.24±1.47

29p

149.18±4.44

20.25±1.23

11



Page 12 of 50

29q

150.69±5.32

22.42±1.21

29r

198.31±7.32

32.10±1.72

29s

167.90±6.56

26.18±1.60

29t

174.10±12.15

31.36±1.48

a


SEM). b


Table 3. Kinase activities of compounds 32a-i, 33a-i, 34a-i, 35a-i, 36a-i against recombinant human ULK1 and relevant MDC positive ratio in SH-SY5Y cells

R1

Kinase

MDC positive

Compound

X

32a

O

145.21±9.96

19.19±0.95

32b

O

158.14±8.91

19.29±0.95

R2

activity%(100 nM)

a

ratio%(1 µM)b

12


Page 13 of 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


169.07±9.96

22.70±1.10

O

148.05±11.74

16.56±1.17

32e

O

144.21±11.2

16.24±0.82

32f

O

209.33±10.17

38.34±1.61

32g

O

149.21±8.35

19.61±0.77

32h

S

169.03±4.54

32.25±2.37

32i

S

177.85±7.24

32.92±2.07

33a

O

152.79±6.18

30.95±0.39

33b

O

143.91±2.75

19.65±0.60

33c

O

159.01±2.61

24.72±0.93

33d

O

152.79±2.84

24.02±0.86

32c

O

32d

CH3

CH3

13



Page 14 of 50

33e

O

147.09±1.05

24.89±1.67

33f

O

226.95±2.17

43.61±1.11

33g

O

151.17±5.10

22.26±0.84

33h

S

196.87±4.77

39.44±1.35

33i

S

243.21±4.45

43.90±1.73

34a

O

157.78±3.62

30.35±0.93

34b

O

30.35±0.93

36.75±1.03

34c

O

156.17±5.41

27.35±0.93

34d

O

174.42±5.64

34.88±0.86

34e

O

149.37±4.52

24.25±0.80

34f

O

182.68±3.04

38.38±0.59

34g

O

145.81±5.19

24.40±1.13

CH3

14


Page 15 of 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


34h

S

179.30±4.30

36.21±0.94

34i

S

204.95±5.78

43.09±1.17

35a

O

157.00±5.34

28.61±1.01

35b

O

160.81±6.31

26.34±0.86

35c

O

151.86±5.58

24.11±1.10

35d

O

147.06±4.86

25.43±0.67

35e

O

154.80±3.50

27.12±0.63

35f

O

179.01±2.35

33.57±2.74

35g

O

150.33±3.40

24.63±0.54

35h

S

175.71±3.59

32.93±1.39

35i

S

217.27±5.26

39.63±1.03

CH3

15



Page 16 of 50

36a

O

140.82±5.51

21.05±0.30

36b

O

152.12±4.46

27.89±0.67

36c

O

167.93±7.71

31.09±0.96

36d

O

148.01±6.56

24.31±1.28

36e

O

143.99±4.30

23.21±1.33

36f

O

162.67±4.14

33.48±1.20

36g

O

139.48±4.13

20.95±0.75

36h

S

170.26±1.82

36.66±1.11

36i

S

184.33±7.33

38.95±0.81

CH3

a


SEM). b


Table 4. Emax values and EC50 values for representative compounds Compound

Emaxa

EC50 (nM)

3

0.324 ± 0.024

263.1 16


Page 17 of 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


a

24c

0.692 ± 0.032

54.84

29n

0.742 ± 0.028

40.80

33i

1.000 ± 0.037

24.14

All Emax values were normalized to the maximal response of 33i.

Identification of the key amino acids between 33i and ULK1 To determine the key amino acids in activator binding site between 33i and ULK1, we constructed several mutants of ULK1 using site-directed mutagenesis. The amino acid residues of Lys50, Arg18, Asn86 and Tyr89 were mutated to alanine (ULK1K50A, ULK1R18A, ULK1N86A ULK1Y89A). ULK1 kinase activity assay indicated that ULK1K50A, ULK1N86A and ULK1Y89A mutants induced substantial decrease of kinase activity compared to ULK1WT after treatment with 33i, while ULK1R18A mutant only induced relatively little reduction of kinase activity (Fig. 3B and Fig. S7). Subsequently, we applied in vitro kinase assay to demonstrate that 33i could enhance the phosphorylation level of mAtg13 in HEK-293T cells transfected with ULK1WT, indicating 33i activates ULK1 in vitro.29,30 And, we found that ULK1R18A mutant did not result in apparent reduction on phosphorylation of mAtg13 in the presence of 33i, but the other three mutants induced a significant decrease of mAtg13 phosphorylation, especially for ULK1N86A and ULK1Y89A mutants (Fig. 3C). Intriguingly, we directly co-transfected Flag-tagged ULK1 (WT or mutant ULK1) and GST-tagged mAtg13 into HEK-293T cells, and found the similar results with the in vitro kinase assay indicated by the phosphorylation levels of p-mAtg13 (Fig. 3D). Moreover, in an analysis of Surface Plasmon Resonance (SPR), we found that 33i bound to ULK1 with a high binding affinity (KD = 0.719 µM), but the ULK1K50A, ULK1N86A and ULK1Y89A mutants lead to obvious decrease of binding affinity than ULK1WT at different levels (Fig. S8). Notably, these biochemical experiments suggested that Lys50, Asn86 and Tyr89 may be more crucial to the activator binding site of ULK1. More importantly,they were in consistent with the aforementioned key amino acid residues to the activator binding pocket with 3. Together, these results suggest that Lys50, Asn86 and Tyr89 are the key amino acids between ULK1 and its activator. 17



Page 18 of 50

Fig. 3 Identification of the key amino acids between ULK1 and 33i. (A) A docking pose based upon the kinase domain structure of ULK1 (PDB code: 4WNP) showed the interaction between 33i and ULK1. The key amino acid residues are marked red. (B) Lys50, Arg18, Asn86 and Tyr89 in the kinase domain of ULK1 were mutated to alanine by site-directed mutagenesis. And, the recombinant proteins were purified from insect cells expressing ULK1WT and ULK1 mutants. Then, ULK1 kinase activity was measured by using kinase assay and purified ULK1 mutant proteins in the presence of different concentrations of 33i. The relative kinase activity of ULK1 mutants were normalized to the maximal response of ULK1

WT

. (C)

Flag-tagged ULK1WT and ULK1 mutants were expressed in HEK-293T cells and immunoprecipitated by anti-Flag antibody, then incubated with GST-tagged mAtg13 in a kinase reaction buffer in the presence or absence of 33i. The reaction was stopped and analyzed by western blot with p-mAtg13 antibody. (D) HEK-293T cells was co-transfected with Flag-tagged ULK1 (WT or mutant ULK1) and GST-tagged mAtg13, then treated with or without 33i. The phosphorylation of mAtg13 in total cell lysates were 18


Page 19 of 50 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


detected with p-mAtg13 antibody.

33i induces autophagy in neuron-like cells To explore the mechanisms of 33i-induced autophagy, we observed the cellular ultrastructure by the electron microscopy. 33i treatment induced some vacuolar elements that were most likely to be of autophagic origin in SH-SY5Y cells (Fig. 4A and 4B). And, we found that 33i time-dependently elevated the expression levels of LC3-II (a key marker of autophagy), Beclin-1 and its phosphorylation status, whereas the level of the selective autophagy substrate SQSTM1/p62 was reduced after treatment with 33i (Fig. 4C). Moreover, LC3 and SQSTM1/p62 were significantly accumulated after co-treatment with 33i and Bafilomycin A1, indicating that 33i treatment enhances the autophagic flux (Fig. 4D). To further demonstrate that whether 33i could induce autophagic flux in other neuron-like cells, we treated highly differentiated PC-12 cell with 33i and Bafilomycin A1. Interestingly, 33i treatment induced autophagosome accumulation in PC-12 cells, which was indicated by increased expression levels of LC3-II and SQSTM1/p62, as well as the aggregated LC3 puncta in PC-12 cells (Fig. 4E and 4F). These results show that 33i can induce autophagy in both undifferentiated and differentiated neuron-like cells. Next, we examined whether 3-methyladenine (3-MA), a class III PI3K autophagy inhibitor, could block 33i-induced autophagy. MDC staining analysis were conducted by 33i-treated SH-SY5Y cells within or without 3-MA. We observed an increase in green dots with fluorescence in 33i-treated cells; these green dots were significantly reduced by the 3-MA treatment (Fig. 4G). Meanwhile, 33i treatment led to the increase of the GFP-LC3 puncta in SH-SY5Y cells, which was markedly decreased under the treatment of 3-MA (Fig. 4H). In addition, 3-MA was found to restrain the transformation of LC3-I to LC3-II, as well as the degradation of SQSTM1/p62, indicating that 3-MA can block 33i-induced autophagy (Fig. 4I).

19



Page 20 of 50

Fig. 4 33i induces autophagy in neuron-like cells. (A) SH-SY5Y cells were treated with or without 5 µM 33i for 24 h. The ultrastructure was examined via transmission electron microscopy (TEM). Arrows indicate autophagic bodies. (B) The numbers of cells with autophagosome and numbers of autophagic vacuoles per cell were analyzed from at least 30 randomly chosen fields in the TEM analysis,

***

p

(ULK1) that induces cytoprotective autophagy for Parkinson's disease

Recommend Documents