Page 83 - I. Chemical biology and drug discovery

Discovery of the first potent proteolysis targeting chimera (PROTAC) degrader of

AIMP2-DX2 as a lung cancer drug

BoRa Lee , Dae Gyu Kim , Aram Lee , Young Mi Kim , Lianji Cui , Sunghoon Kim *, Inhee Choi *
2,
1,
1
1
1
2,§
1,§
1. Medicinal Chemistry, Institut Pasteur Korea, 16, Daewangpangyo-ro 712 beon-gil, Bundang-gu, Seongnam-si, Gyeonggi-do, 13488, Korea

2. Medicinal Bioconvergence Research center, College of Pharmacy and College of Medicine, Gangnam Severance hospital, Yonsei University, Incheon, 21983, Korea

BACKGROUND AIM

Proteolysis targeting chimera (PROTAC) technology have attracted To target DX2, specific compounds are required to interrupt the

great attention in drug discovery and development areas. A splicing interaction between DX2 and HSP70. Here we designed and

variant of AIMP2, AIMP2-DX2 (DX2), is highly expressed in lung synthesized DX2 PROTACs using the most potent hit compound

cancer cells and lung cancer patient tissues. To date, many attempts BC-DXI-495 as the targeting protein ligase ligand and CRBN

have been made to develop DX2 inhibitors but due to the lack of (pomalidomide and thalidomide) as the Siah1 E3 ligase ligand

satisfactory activity against DX2, new therapeutic strategies are based on the binding assay. Linker composition and length

needed to develop a novel drug for DX2. It is already reported that activity relationship of PROTACs were studied and measured by

DX2 selectively binds to 70 kDa heat shock proteins (HSP70s) but the luciferase assay.

HSP70s inhibit ubiquitination mediated degradation of DX2 and

Siah1, a specific E3 ligase of DX2.

METHODS RESULTS

SAR map of DX2-PROTAC SAR map of DX2-PROTAC

Mechanism of PROTAC
Position 1 Position 2

H O
PROTAC NH O O N

O N O
AIMP2-DX2 Linker NH N Linker N

E3 ligase ligand E3 ligase ligand N O N O
H

O S NH “Variable length
2

“Variable lengths & O H O NH N &
H
N
functional groups” O O S functional groups
2

Linker N ”

O

Number of Number of
% inhibition of DX2 % inhibition of DX2

MC code non-H MC code non-H
luciferase (at 4 µM) luciferase (at 4 µM)
linker atoms linker atoms

BC-DXI-495 37.67 - MC279 25.05 3

MC285 19.43 2 MC249 51.10 4

MC316 41.14 3 MC314 35.33 4

MC233 37.50 4 MC347 44.44 5

MC337 44.74 5 MC250 51.99 6
General chemical structure of MC283 48.30 5 MC348 42.53 6

9
putative PROTAC for DX2 MC310 35.88 6 MC251 45.79 12

39.74
MC252
MC235 40.03 6 MC277 26.11 15

NH O MC232 40.76 6

N Position 2
O H O MC313 30.88  The same linker + E3 ligase ligand
O N 7
N O MC287 34.11 7 were used for MC233 and MC249,

NH H
O S N MC323 46.16 8 MC235 and MC250, MC236 and
2

O
MC311 24.65 9 MC251, MC237 and MC252. These

Position 1 MC345 32.40 9 linkers and E3 ligase ligand

CRBN E3 ligase ligand

BC-DXI-495 K = 25µM MC236 44.61 9 combination increased inhibitory

K = 14 µM D MC267 36.63 9 activities regardless of positions.
D

MC275 28.79 9

 BC-DXI-495 specifically inhibits the MC234 31.73 10  Linker composition has effect on the

tumor-promoting interaction between MC290 38.13 10 ability of DX2 PROTAC degraders to

DX2- and HSP70. MC312 24.37 12 induce degradation of the target

MC324 46.39 12 protein.
 BC-DXI-495 binds to DX2 with a K of MC344 34.00 12

D
approximately 14 µM. MC237 37.80 12  Linker length dependence was

 CRBN E3 ligase ligand binds to Siah1 MC268 47.09 12 observed in DX2 degraders.

with a K D of approximately 25 µM MC280 33.79 13 PROTACs with longer linker

whereas other E3 ligase ligand such as MC269 41.48 15 possessed lower degradation

24.03
MC284
VHL doesn’t bind to Siah1. MC346 14.84 16 capacity, which implies linkers

18
should be around 4-9 in lengths.

 BC-DXI-495 and CRBN are connected MC270 32.97 18

via various linkers through pathway 1 MC288 29.10 19  Inhibition is more effective when

and pathway 2. MC276 31.87 21 linker + E3 ligase ligands are

MC281 23.61 22 connected to Position 2 in general.

CONCLUSION ACKNOWLEDGEMENTS

 We believe that MC250 is the best DX2-PROTAC better than BC- This was supported by the National Research foundation of

DXI-495 based on our structure-based discovery. Korea (NRF) grant funded by the Korea government (MSIT) (No.

 DX2-PROTAC is more effective when linkers were optimized in 4- 2017M3A9G6068257), Gyeonggi-do.

9 lengths and are connected at position 2.

 Our data confirmed the applicability of PROTAC technology to Contact information

target DX2 and proved DX2 PROTAC compounds could potentially

lead to therapeutic opportunities to treatment of lung cancer. Dr. Inhee Choi (inhee.choi@ip-korea.org)