Pyrazolylamine Derivatives Reveal the Conformational Switching between Type i and Type II Binding Modes of Anaplastic Lymphoma Kinase (ALK)

Article abstract of DOI:10.1021/acs.jmedchem.6b00106

Most anaplastic lymphoma kinase (ALK) inhibitors adopt a type I binding mode, but only limited type II ALK structural studies are available. Herein, we present the structure of ALK in complex with N1-(3-4-[([5-(tert-butyl)-3-isoxazolyl]aminocarbonyl)amino

Full text of DOI:10.1021/acs.jmedchem.6b00106

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES
Article
Pyrazolylamine derivatives reveal the conformational switching between
type-I and type-II binding modes of anaplastic lymphoma kinase (ALK)
Chih-Hsiang Tu, Wen-Hsing Lin, Yi-Hui Peng, Tsu Hsu, Jian-Sung Wu, Chun-Yu
Chang, Cheng-Tai Lu, Ping-Chiang Lyu, Chuan Shih, Weir-Torn Jiaang, and Suying Wu
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00106 • Publication Date (Web): 31 Mar 2016
Downloaded from http://pubs.acs.org on March 31, 2016
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 free 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 accessible to all
readers and 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.
Journal of Medicinal Chemistry 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 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Pyrazolylamine derivatives reveal the conformational
switching between typeꢀI and typeꢀII binding modes
of anaplastic lymphoma kinase (ALK)
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
Chih-Hsiang Tu, † Wen-Hsing Lin, †§ Yi-Hui Peng, †§ Tsu Hsu, † Jian-Sung Wu,† Chun-Yu
∥
Chang,† Cheng-Tai Lu,† Ping-Chiang Lyu, ∥ Chuan Shih,† Weir-Torn Jiaang, †* Su-Ying Wu†*
†Institute of Biotechnology and Pharmaceutical Research, National Health Research Institutes,
35 Keyan Road, Zhunan Town, Miaoli County 350, Taiwan, ROC
∥ Institute of Bioinformatics and Structural Biology, National Tsing Hua University, 101, Sect.2,
Guangfu Road, Hsinchu 300, Taiwan, ROC
§ W.ꢀH. L. and Y.ꢀH. P. contributed equally
KEYWORDS: anaplastic lymphoma kinase (ALK), typeꢀII kinase inhibitors, DFG (AspꢀPheꢀ
Gly) motif, juxtamembrane (JM) domain.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 48
1
2
3
4
ABSTRCT
5
6
7
8
Most Anaplastic lymphoma kinase (ALK) inhibitors adopt a typeꢀI binding mode, but only
limited typeꢀII ALK structural studies are available. Herein, we present the structure of ALK in
complex with N1ꢀ(3ꢀ4ꢀ[([5ꢀ(tertꢀbutyl)ꢀ3ꢀisoxazolyl]aminocarbonyl)amino]ꢀ3ꢀmethylphenylꢀ1Hꢀ
5ꢀpyrazolyl)ꢀ4ꢀ[(4ꢀmethylpiperazino)methyl]benzamide (5a), a novel ALK inhibitor adopting a
typeꢀII binding mode. It revealed binding of 5a resulted in the conformational change and
reposition of the activation loop, αCꢀhelix, and juxtamembrane domain, which are all important
domains for the autoinhibition mechanism and downstream signal pathway regulation of ALK. A
structureꢀactivity relationship study revealed that modifications to the structure of 5a led to
significant differences in the ALK potency and altered the protein structure of ALK. To the best
of our knowledge, this is the first structural biology study to directly observe how changes in the
structure of a small molecule can regulate the switch between the type I and type II binding
modes, and induce dramatic conformational changes.
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
Introduction
Anaplastic lymphoma kinase (ALK) was first discovered as the Nucleophosmin (NPM)ꢀALK
fusion protein in the anaplastic large cell lymphoma (ALCL) of nonꢀHodgkin’s lymphoma
(NHL) patients in 1994.1 It is classified as a receptor tyrosine kinase (RTK), and consists of an
extracellular ligand binding domain, a transmembrane domain, a juxtamembrane domain, an
intracellular kinase domain, and a Cꢀterminal tail.2, 3 Midkine (MK) and pleiotrophin (PTN), two
neurite growthꢀpromoting factors (NEGF), have been reported as the activating ligands for
mammalian ALK.2, 3 Since the first discovery of NPMꢀALK in NHL patients, various ALKꢀ
fused partners have been identified, including cysteinylꢀtRNA synthetase (CARS)ꢀALK in
ACS Paragon Plus Environment

Page 3 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
inflammatory myofibroblastic tumors (IMT); clathrin (CLTC)ꢀALK in diffuse large Bꢀcell
lymphoma (DLBCL); and echinoderm microtubule associated protein like 4 (EML4)ꢀALK in
nonꢀsmallꢀcell lung carcinoma (NSCLC).2 Among these, EML4ꢀALK, identified in about 3 to
7% of NSCLS patients,4 is one of the most extensively studied ALK fusions. The Nꢀterminal
portion of EML4 facilitates ALK oligomerization, kinase activation, and subsequently initiates
the signal transduction network. Studies have shown that mutations, gene amplification and
chromosomal translocation of ALK are associated with various cancers.2 Therefore, ALK has
been proposed as a drug target in the treatment of tumours; and the inhibition of ALK by small
molecules constitutes a promising approach due to the low ALK mRNA and protein expression
levels in normal human tissue.2, 5 Currently, several ALK inhibitors, including crizotinib,6
ceritinib,7 and alectinib5 are FDAꢀapproved for the treatment of ALK positive NSCLC patients;
and several secondꢀgeneration inhibitors such as N2ꢀ[2ꢀMethoxyꢀ4ꢀ[4ꢀ(4ꢀmethylꢀ1ꢀpiperazinyl)ꢀ
1ꢀpiperidinyl]phenyl]ꢀN4ꢀ[2ꢀ[(1ꢀmethylethyl)sulfonyl]phenyl]ꢀ1,3,5ꢀtriazineꢀ2,4ꢀdiamine
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
(ASP3016),8
5ꢀChloroꢀN2ꢀ{4ꢀ[4ꢀ(dimethylamino)ꢀ1ꢀpiperidinyl]ꢀ2ꢀmethoxyphenyl}ꢀN4ꢀ[2ꢀ
(dimethylphosphoryl)phenyl]ꢀ2,4ꢀpyrimidinediamine (AP26113),9 and (10R)ꢀ7ꢀAminoꢀ12ꢀ
fluoroꢀ2,10,16ꢀtrimethylꢀ15ꢀoxoꢀ10,15,16,17ꢀtetrahydroꢀ2Hꢀ8,4ꢀ(metheno)pyrazolo[4,3ꢀh]ꢀ
[2,5,11]benzoxadiazacyclotetradecineꢀ3ꢀcarbonitrile (PFꢀ06463922)10 are in clinical trials.
Unfortunately, some of these are associated with adverse effects. For example, nausea, vomiting,
diarrhea, visual disorders and peripheral edema were observed in patients after receiving
crizotinib treatment.11, 12 Ceritinib produced adverse effects; nausea, diarrhea, vomiting and
fatigue were observed in clinical trials.13 The reported side effects of alectinib were dysgeusia,
anemia, elevated aminotransferase levels, rash and constipation.14 Hence, further development of
ALK inhibitors for the treatment of cancer is needed.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 48
1
2
3
4
5
6
7
8
9
TypeꢀI kinase inhibitors are ATP competitors and target the DFGꢀin kinase conformation
whereas typeꢀII kinase inhibitors recognize the DFGꢀout conformation.15 In addition, typeꢀII
kinase inhibitors not only occupy the ATP binding site but also exploit the less conserved
allosteric pocket by flipping the DFG motif, which might improve kinase selectivity and reduce
the potential for resistance.16, 17 Moreover, typeꢀII kinase inhibitors generally display a lower
dissociation rate constant (koff) with prolonged residence time, which would extend the efficacy
both in vitro and in vivo.18, 19 However, most typeꢀII inhibitors tend to have high molecular
weight and lipophilicity in order to exploit the additional hydrophobic pocket. In consequence,
the majority of typeꢀII inhibitors have the potential problems of limited solubility and worse
pharmacokinetic (PK) properties.15 These require further efforts in the late stage of drug
development to improve the druggability of typeꢀII inhibitors.
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
Nevertheless, the use of type II kinase inhibitors, such as imatinib,20 has been viewed as
beneficial approach for the treatment of cancers. Unfortunately, due to the limitations of the
traditional enzymatic assay, few typeꢀII ALK kinase inhibitors have been identified.21, 22 In
addition, limited DFGꢀout ALK/inhibitor structures available for structureꢀguided in drug design
may impede the effort on the design and optimization of typeꢀII ALK inhibitors.
In this study, a screening of inꢀhouse kinases led to the identification of 5a, a 3ꢀphenylꢀ1Hꢀ5ꢀ
pyrazolylamineꢀbased compound bearing a urea substituent, as an ALK inhibitor with an IC50
value of 177 nM. Herein, we describe structural biology studies of 5a with ALK, revealing it to
adopt a typeꢀII binding mode and concurrently rearrange the conformations of the activation loop,
αCꢀhelix, and juxtamembrane domain. Upon binding to ALK, compound 5a induces movement
of the αCꢀhelix which allosterically regulates the autoinhibition state of the juxtamembrane
domain from antiꢀparallel βꢀsheets to a less compact, short helix conformation. A subsequent
ACS Paragon Plus Environment

Page 5 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
structureꢀactivity relationship (SAR) study revealed that modifications to the structure of 5a led
to a significant difference in the ALK inhibitor potency. Moreover, comparison of the coꢀcrystal
structure of ALK in complex with 5a with that of its analogue, 4, demonstrated that modification
of inhibitor structure could switch it between a type I and type II binding mode, and induce
dramatic conformational changes.
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
Results
Identification of pyrazolylamine derivatives as ALK inhibitors. A novel class of 3ꢀphenylꢀ
1Hꢀ5ꢀpyrazolylamine derivatives23 has been utilized as a versatile template for the development
of kinase inhibitors. Further screening of this series of compounds against inꢀhouse kinases led to
the identification of ureaꢀsubstituted series of compounds as ALK inhibitors. As shown in Table
1, 5a was the most potent ALK inhibitor, with an IC50 value of 177 nM. The general synthetic
route to 3ꢀphenylꢀ1Hꢀ5ꢀpyrazolylamine benzamides, 5a and its analogs (Table 1), is shown in
Scheme 1. Structureꢀactivity relationship (SAR) studies showed that modifications to the ureaꢀ
substituted tail moiety (R1 group in Table 1) could significantly affect ALK potency. Removing
the ureaꢀsubstituted moiety of 5a to give 4 resulted in the complete loss of activity (IC50 > 10
uM), suggesting the ureaꢀsubstituted moiety to be very important for the inhibition of ALK.
Replacement of the isoxazole ring of 5a with a phenyl ring to give 5b decreased the activity to
IC50 = 6.2 uM. Adding a methyl group next to the urea moiety to give 5c also reduced the
activity (IC50 = 4.5 uM).
The effect of alkyl substituents on the terminal isoxazole ring on ALK inhibition was also
evaluated. In comparison with the bulky tꢀbutyl group of 5a, compounds bearing the less
sterically hindered isopropyl group (5d), ethyl group (5e) and methyl group (5f) were
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 48
1
2
3
4
5
6
7
8
9
synthesized and tested, and the substitution found to result in a 2ꢀfold, 4ꢀfold and 32ꢀfold
decrease in potency, respectively. Removal of the methyl substituent (R2 group in Table 1) on
the phenyl ring connected to the pyrazolylamine group to give 5g and 5h resulted in decreased
ALK inhibition compared with their methyl analogues, 5a and 5d.
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
The effect of adding a waterꢀsolubilizing group on ALK inhibition was investigated. Previous
studies have demonstrated the importance of a waterꢀsolubilizing substituent on the benzamide
moiety for optimal activity toward the FLT3 kinase.23ꢀ25 In this study, addition of various waterꢀ
solubilizing benzamide substituents to the urea substituted moiety had dramatic effects on
potency. With conservation of the 5ꢀtertꢀbutylꢀisoxazole urea moiety, inhibitors bearing a
pyrrolidine group tethered by either a twoꢀcarbon (5i) or threeꢀcarbon ether (5j) were found to be
3ꢀfold and 7ꢀfold less potent than a Nꢀmethylpiperazine group linked by a oneꢀcarbon tether (5a).
A sharp drop in potency was observed when the pyrrolidine group (5i) was replaced by a
morpholine group (5k), or an Nꢀmethylpiperazine group was directly attached to benzamide (5l).
Both compounds 5k and 5l showed no inhibition (IC50 > 10 ꢁM).
Crystal structure of 5a. To further elucidate the binding mode and get structural insights into
the detailed interactions between compound 5a and ALK, a structural biology study of 5a in
complex with the ALK kinase domain (amino acids 1084ꢀ1410) was performed. Compound 5a
was soaked into the ALK native crystal and the structure of 5a bound to ALK determined to a
resolution of 1.92 Å. Data collection and refinement statistics are summarized in Table 2. All the
protein residues were well defined in the electron density map except for some flexible loops,
including residues 1084ꢀ1092, 1124ꢀ1127, 1137ꢀ1143, 1216ꢀ1219, 1275ꢀ1287 and 1405ꢀ1410.
ACS Paragon Plus Environment

Page 7 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The structural moieties of ALK in complex with 5a were: the small Nꢀterminal lobe (Nꢀlobe);
the large Cꢀterminal lobe (Cꢀlobe); and the interconnection hinge region. The Nꢀlobe is
comprised of five antiparallel βꢀsheets and one dominant helix (αCꢀhelix). The Cꢀlobe is
comprised predominantly of eight αꢀhelices and an activationꢀloop (Aꢀloop). The electron
density map clearly shows that 5a is located in the ATP binding site between the Nꢀlobe and Cꢀ
lobe. Unlike most known ALK inhibitors (which present as typeꢀI inhibitors), 5a adopts the typeꢀ
II inhibitor binding mode and exploits the additional hydrophobic pocket, creating by the
rearrangement of the activation loop of ALK from the DFGꢀin to the DFGꢀout conformation.
Indeed, 5a bears several structural features that bind to ALK: the hinge scaffold
(pyrazolylamine), the waterꢀsolubilizing head group (Nꢀmethylpiperazine), the middle linker
(tolyl), hydrogenꢀbonding moiety (urea) and hydrophobic tail group (tꢀbutylꢀisoxazole) (Figure
1A). The pyrazolylamine scaffold of 5a forms hydrogen bonds with Glu1197 and Met1199 in the
hinge region (Figure 1B), stabilizing the binding orientation and coordinating 5ꢀtꢀbutylꢀisoxazole
group deeply into the hydrophobic pocket. The tolyl group at the 3ꢀposition next to the
pyrazolylamine scaffold makes hydrophobic interactions with the surrounding residues,
including Val1130, Leu1196, Leu1256, and Phe1271. Moreover, the rearrangement of the
sidechain of Phe1271 of DFG motif provides the edgeꢀtoꢀface hydrophobic interactions with the
tolyl group. The methyl substituent in the tolyl group has interactions with Phe1271 and also
contributes to the inhibitory activity to ALK. Removal of this methyl substitute results in 5ꢀfold
decreased activity of 5g as compared with 5a. The Nꢀmethylpiperazine head group is located at
the acidic zone, formed by the side chains of Asp1203, Ser1206 and Glu1210 and the backbone
carbonyl oxygens of Ala1200 and Gly1201 (Figure 1B). The methylpiperazine group is a waterꢀ
solubilizing group, often used in kinase inhibitor design to improve the physicochemical
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 48
1
2
properties of compounds.26 However, the methylpiperazine group not only improved the
physicochemical properties of 5a, but also significantly influences potency against ALK, as
revealed in the aforementioned SAR study. The two tertiary amines of the Nꢀmethylpiperazine of
5a serve as bases to form the longꢀrange, chargeꢀcharge interactions with the carboxylic acid
group of Glu1210 and the backbone carbonyl oxygens of Ala1200 and Gly1201. With an
increase (5k) or decrease (5l) in chain length, the inhibitory activities decrease. One reason for
the dramatic loss of activity of 5k could be that the waterꢀsolubilizing group is away from the
acidic region, which leads to the loss of chargeꢀcharge interactions with the surrounding
residues. Another possibility is that the basicity of the waterꢀsolubilizing group might also
contribute to the binding with the protein. The pKa of the nitrogen atom in the morpholineꢀ4ꢀylꢀ
methoxy group of 5k is 4.09, while pKa of two nitrogen atoms in Nꢀmethylpiperazine of 5a are
7.84 and 8.07 as calculated by MarvinSketch program implemented in ChemAxon.27 The
nitrogen atom in the morpholineꢀ4ꢀylꢀmethoxy group with the lower pKa value and less basicity
might have weaker charge interactions with the acidic residues of ALK, resulting in a loss in
activities. Moreover, the reason for the decreased activity of 5l could be that the shorter chain
length restricted the orientation of Nꢀmethylpiperazine to form optimized interactions with the
surrounding acidic residues.
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
The urea moiety and the tꢀbutyl isoxazole group of 5a are located at the additional
hydrophobic pocket, creating by the shift of the DFG motif from the DFGꢀin to the DFGꢀout
conformation. The urea moiety forms one hydrogen bond with the backbone of Asp1270 in the
DFG motif, and two hydrogen bonds with the sidechain of Glu1167 in αCꢀhelix. The tꢀbutyl
isoxazole group extends into the allosteric site surrounded by the HRD motif and the αCꢀhelix,
where it makes extensive hydrophobic interactions with Ile1171, Phe1174, Ile1179, Phe1245,
ACS Paragon Plus Environment

Page 9 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
His1247 and Ile1268. That these interactions between the tꢀbutyl moiety and these hydrophobic
residues contribute to the activity of 5a is evidenced by compounds 5d, 5e and 5f. When the tꢀ
butyl moiety is replaced with the smaller alkyl substituents, such as isoꢀpropyl (5d), ethyl (5e), or
methyl (5f), the activity decreases accordingly, suggesting that the hydrophobicity of the alkyl
substituents correlates with the activity of compounds. Indeed, superimposition of the crystal
structure of ALK/5a and of ALK/5d shows that the isoꢀpropyl group of 5d loss of the
hydrophobic interactions (Phe1174, Ile1179 and Phe1245) compared with the tꢀbutyl group of 5a
(Figure 2).
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
5a induces a dramatic conformational change of ALK. Superimposition of the ALK/5a
complex structure with the apo ALK structure reveals that ALK undergoes significant
conformational changes upon 5a binding. The rootꢀmeanꢀsquare deviation (RMSD) of these two
aligned structures is 2.11 Å for 282 Cα atoms. Binding of 5a, a typeꢀII kinase inhibitor, not only
induces the conformational flipping of DFG motif but also results in the reconstitution of the αCꢀ
helix, activation loop (Aꢀloop) and juxtamembrane (JM) domain.
(i) Rearrangement of Activationꢀloop (Aꢀloop) and αCꢀhelix. To elucidate the conformational
change induced by 5a, the apo ALK crystal structure was solved to a resolution of 1.50 Å and
compared with ALK/5a complex structure (Figure 3). There are various differences between
these two structures. Firstly, in the apo ALK structure, the nonꢀconserved region of the Aꢀloop
following the DFG motif adopted a short helix (αꢀAL, residue 1272ꢀ1280) and positioned
orthogonally underneath the αCꢀhelix to form the salt bridge and hydrogen bond interactions
with αCꢀhelix (i.e. Gln1159 in αCꢀhelix/Arg1279 in αꢀAL; Asp1163 in αCꢀhelix/Arg1275 and
Arg1279
in
αꢀAL;
Glu1167
in
αCꢀhelix/
Arg1275
in
αꢀAL).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 48
1
2
3
4
5
6
7
8
9
Tyr1278, located in the terminal of αꢀAL, is an autophosphorylation site in the Aꢀloop and plays
an important role in the regulation of ALK activation.28 Through edgeꢀtoꢀface, πꢀπ interactions
with Tyr1096 and the hydrogen bonding with backbone nitrogen atom of Ser1097 in JM domain,
Tyr1278 serves as the latch which clamps the orientation of αCꢀhelix and stabilizes the JM
domain in an antiꢀparallel βꢀturn conformation (Figure 3A). In comparison, binding of 5a
unfolded the αꢀAL to a nonꢀstructural loop and moved αꢀAL away from the αCꢀhelix (Figure
3B), resulting in the loss of the salt bridge and hydrogen bond interactions with αCꢀhelix. The
loss of the interactions between αꢀAL and αCꢀhelix moves the αCꢀhelix 2 Å away from Cꢀlobe to
an open state (Figure 3C). Consequently, the interactions of Tyr1278 in αꢀAL with Tyr1096 and
S1097 in the JM domain were concurrently lost.
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
Secondly, 5a flipped the DFG motif from an in to an out conformation. This conformational
change altered the internal architecture of the ALK kinase domain. In the apo ALK structure, the
neighbouring residues (His1247 in HRD motif, Phe1271 in DFG motif, Ile1171 in the αCꢀhelix
and Cys1182 in the β4ꢀstrand) constitute the functional hydrophobic network, which is
designated as regulatory spine (Rꢀspine) (Figure 4A). The backbone nitrogen of His1247 in Rꢀ
spine hydrogen bonded to the sidechain of Asp1311 in the helixꢀF, served as the base for the Rꢀ
spine. Studies have shown that assembling and disassembling of the Rꢀspine regulates the
activity of protein kinase.29ꢀ31 Upon the binding of 5a to ALK, the Rꢀspine is disrupted by the
movement of Phe1271 in the DFGꢀmotif and reꢀaccommodation by the tꢀbutylꢀisoxazole group
of 5a. In addition, the catalytic spine (Cꢀspine) also showed differences between the apo ALK
and ALK/5a complex structures. It also regulates activity of protein kinase; however, assembling
or disassembling of the Cꢀspine is dependent on the incorporation of ATP or substrate. In the apo
ALK structure, without the ATP or substrate, the Cꢀspine is incomplete; while in the ALK/5a
ACS Paragon Plus Environment

Page 11 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
complex structure, the pyrazole group of 5a interconnects the Nꢀlobe and Cꢀlobe through
neighbouring hydrophobic residues and completes the Cꢀspine (Figure 4B).
7
8
9
(ii) Conformational change of juxtamembrane (JM) domain. Most of the published ALK
structures revealed that the juxtamembrane (JM) domain (residues 1087ꢀ1115), flanked by the
αCꢀhelix, contain an Nꢀterminal helix segment (residues 1086ꢀ1092) and an antiparallel βꢀturn
motif (residues 1096ꢀ1103) (Figure 3A and Figure 3C). The βꢀturn motif in the JM domain is a
unique feature for ALK, compared with other kinase structures. The hydrophobic residues from
the JM domain, including Ile1088, Met1089, Pro1094, Tyr1096, Phe1098, Ile1105 and Leu1108,
facilitate the close packing of the JM domain on the interface of αCꢀhelix. In addition, the
interactions between the JM domain (Tyr1096 and Ser1097) and the αꢀAL (Tyr1278) also assist
the stabilization of the ALK kinase domain in the native conformation (Figure 3A).
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
Interestingly, upon the binding of 5a, the JM domain transformed from the antiꢀparallel βꢀturn
conformation to the helix conformation (residue 1096ꢀ1102), with concomitant disorder of the
Nꢀterminal helix segment of the JM domain (Figure 3B and Figure 3C). This conformational
change in the JM domain results in significant loss of the contacted interface area with αCꢀhelix,
and disrupts the interactions with αꢀAL. Superimposition of the apo ALK structure with the
ALK/5a complex structure revealed that the opened state αCꢀhelix induced by the binding of 5a
would clash with the JM domain (Figure 3D). Ile1088 and Met1089 in the Nꢀterminal JM
domain experience a steric clash with the Glu1158 and Glu1161 in the Nꢀterminal αCꢀhelix;
while Tyr1096 and Phe1098 in the βꢀturn motif of the JM domain clash with Ile1170 and
Phe1174 in the Cꢀterminal part of αCꢀhelix, respectively (Figure 3D). To avoid this, the JM
domain undergoes a conformational change, from βꢀturn motif to the helix conformation and also
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 48
1
2
3
4
5
6
disorders the Nꢀterminal helix segment to a more flexible conformation to accommodate the
opened state αCꢀhelix.
7
8
9
As a result of the conformational change of the JM domain in the ALK/5a structure, two
potential autophosphorylation tyrosine residues, Tyr1092 and Tyr109632, 33 were repositioned. In
the apoꢀALK structure, Tyr1092 in the Nꢀterminal helix segment was exposed to solvent region
while Tyr1096 was deeply immersed in a hydrophobic pocket composed of Phe1098, Leu1169,
Ile1170, Lys1173, Phe1174, Phe1245 and Tyr1278 (Figure 3A). However, in the ALK/5a
structure, Tyr1092 was structurally disordered and Tyr1096 was exposed to solvent, where it
forms hydrogen bonds with Asp1163 in the αCꢀhelix and Asn1093 in the JM domain (Figure 3B).
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
In the previous studies, Tyr1096 has been identified as the phosphoprotein binding domain (PBD)
for insulin receptor substrate1 (IRS1) and Suc1ꢀassociated neurotrophic factorꢀinduced tyrosineꢀ
phosphorylated target (SNT).32 IRS1 binds to pTyr1096 and activates the downstream of cell
division of the Ras/Erk1/2 signal pathway. It is suggested that binding of 5a to ALK results in
reconstitution of the JMꢀdomain and alters the interaction network of Tyr1096 with its
surroundings, which may affect the autoinhibition state of ALK and the downstream signalling
pathway inhibition. Tyr1096 is not only a key autophosphorylation site, but might also serve as a
structural switcher to regulate the JM domain autoinhibition and kinase catalytic activity.
The conformational change of the JM domain induced by the typeꢀII kinase inhibitors kinase
has also been observed for other tyrosine kinases, including FLT3/ 1ꢀ(5ꢀtertꢀbutylꢀ1,2ꢀoxazolꢀ3ꢀ
yl)ꢀ3ꢀ[4ꢀ[6ꢀ(2ꢀmorpholinꢀ4ꢀylethoxy)imidazo[2,1ꢀb][1,3]benzothiazolꢀ2ꢀyl]phenyl]urea (AC220)
(PDB code: 4XUF),34 cꢀKIT/imatinib (PDB code: 1T46)35 and VEGFR/sorafenib (PDB code:
4ASD).36 In addition, reposition of those autoꢀphosphorylation tyrosine residues has always
accompanied rearrangement of the JM domain. For example, in the cꢀKIT structure, Tyr568 and
ACS Paragon Plus Environment

Page 13 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Tyr570 are key autoꢀphosphorylation tyrosine residues that regulate the autoinhibition state of cꢀ
KIT. The binding of imatinib to cꢀKIT led to the rearrangement of the JM domain and the
repositioning of Tyr568 and Tyr570. However, most of those typeꢀII inhibitors would directly
clash with the JM domain; hence, the conformational change of JM domain is required to avoid
this. In contrast, 5a is far away and has no directly steric clash with the JM domain as revealed in
the ALK/5a complex structure. 5a induced an allosteric effect through the conformational
change of αCꢀhelix moving toward the JM domain, and consequently led to the rearrangement of
JM domain to be compatible with the opened state of αCꢀhelix.
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
In conclusion, 5a induced conformation changes in the αꢀAL, αCꢀhelix and JM domain, and
thus disrupted the extensive interaction networks in the ALK structure. Consequently, the
interactions between Tyr1278 in αꢀAL and Tyr1096 and S1097 in the JM domain were lost.
These results suggest that 5a altered the interaction network formed by αCꢀhelix, αꢀAL and JM
domain, which regulates the autoinhibition and activation of ALK.
The impact of substitutedꢀurea tail moiety on 5a binding. As revealed in the SAR study, 4 did
not impart any ALK inhibitory activity (IC50 >10 ꢁM), suggesting that the ureaꢀsubstituted tail
moiety was the dominant feature of 5a and contributed to the kinase inhibition activities. In order
to evaluate the effect of the ureaꢀsubstituted tail group on the binding of ALK at the molecular
level, 4, a nonꢀurea aryl analogue of 5a, was soaked into ALK crystal and the structure of the
complex resolved to a resolution of 1.87 Å. The electron density map clearly shows that 4
presented as a typeꢀI inhibitor, bound in the DFGꢀin conformation.
The structure of ALK/4 is significantly different from the ALK/5a structure. Superimposition
of these two structures gave a RMSD of 2.147 Å for 280 Cα atoms. Firstly, in the ALK/5a
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 48
1
2
3
4
5
6
7
8
9
structure, the DFG motif adopts the out conformation to allocate the ureaꢀsubstituted tail and
makes the faceꢀtoꢀedge πꢀπ interactions with the pyrazolylamine scaffold of 5a; while in ALK/4
structure, in the absence of the ureaꢀsubstituted tail, the DFG motif adopts the in conformation,
and has no direct interactions with the pyrazolylamine scaffold of 4 (Figure 5A).
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
Secondly, in the ALK/5a structure, the αCꢀhelix exhibits the open state conformation and the
Glu1167 in the αCꢀhelix forms hydrogen bonding interactions with the urea moiety (Figure 1).
The JM domain was fully unfolded due to the allosteric effect derived from movement of αCꢀ
helix. However, in the ALK/4 structure, the αCꢀhelix adopts the close state conformation and the
Glu1167 interacts with the aminoꢀsubstituent on the 4ꢀposition of the tolyl group, through a
waterꢀmediated hydrogen bond (Figure 5B). Without the ligandꢀinduced allosteric effect, the αCꢀ
helix and JM domain of ALK/4 structure would adopt a similar conformation to the apo ALK
structure (Figure 3A) and other published ALK complex structures with type I inhibitors.
As revealed in the structural biology studies of 5a and 4, 5a is a typeꢀII inhibitor bound to the
DFGꢀout conformation, while 4 is a typeꢀI inhibitor bound to the DFGꢀin conformation. Many
reports have shown that the binding kinetics of typeꢀII inhibitors and typeꢀI inhibitors are
different.18 Therefore, the binding kinetics of 5a and 4 were further evaluated by surface plasmon
resonance (SPR). Binding of 4 showed a fragmentꢀlike squaredꢀshaped sensorgram, indicating
very fast on and fast off rates (Supporting information, Figure S1). In contrast, the association
rate (kon) and dissociation rate (koff) for 5a were 5.17x104 1/Ms and 2.4x10ꢀ3 1/s, respectively,
suggesting that 5a bound to ALK with slow association and dissociation rates. It has been
reported that these slow binding kinetics correlate with protein conformation change induced by
the ligand binding.37, 38 The ureaꢀsubstituted tail moiety endows 5a with a characteristic slow offꢀ
rate, consistent with the requirement of DFG motif conformational change.
ACS Paragon Plus Environment

Page 15 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In addition to the ALK/5a structure, the structure of ALK/4 also compared with apo ALK
structure. The most significant difference between ALK/4 and apo ALK structures is the
conformation change of the phosphateꢀbinding loop (Pꢀloop) upon the binding of 4. In the
ALK/4 structure, the Pꢀloop adopts a bent conformation and orients the conserved aromatic
residue Phe1127 into the catalytic site, directly stacked onto the tolyl group of 4 by edgeꢀtoꢀface
πꢀπ interaction, whereas the Pꢀloop adopted an extended conformation in most of the ALK
structures (Figure 5C). In many kinase/inhibitor complex structures, bending of the Pꢀloop
always creates the faceꢀtoꢀface or edgeꢀtoꢀface aromatic interaction between the conserved
aromatic residue and inhibitors such as AuroraꢀA/Nꢀ[4ꢀ[4ꢀ(4ꢀmethylpiperazinꢀ1ꢀyl)ꢀ6ꢀ[(5ꢀ
methylꢀ1Hꢀpyrazolꢀ3ꢀyl)amino]pyrimidinꢀ2ꢀyl]sulfanylphenyl]cyclopropanecarboxamide (VXꢀ
680) (PDB code: 3E5A),39 cꢀAbl/imatinib (PDB code: 1IEP),40 and p38α/1ꢀ[5ꢀtertꢀbutylꢀ2ꢀ(4ꢀ
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
methylphenyl)pyrazolꢀ3ꢀyl]ꢀ3ꢀ[4ꢀ(2ꢀmorpholinꢀ4ꢀylethoxy)naphthalenꢀ1ꢀyl]urea
(PDB code: 1KV2)18 complex structures.
(BIRB796)
Comparing 5a with others ALK kinase inhibitors. To elucidate the similarities and
differences of structure conformation and proteinꢀligand interactions, crystal structures of 5a and
different ALK kinase inhibitor (Figure 6A) complex structures were superimposed.
Superimposition of 5a and crizotinib (PDB code: 2XP2)6 revealed that the structural
conformations are quite different (Figure 6 B and C). The structure of ALK/crizotinib resembles
apo ALK and the binding of crizotinib to ALK does not induce dramatic conformational changes.
The conformation of ALK/crizotinib is the DFGꢀin conformation and the αCꢀhelix adopts the
close state accompanying with the JM domain in the antiꢀparallel βꢀsheets conformation. In
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 48
1
2
3
4
5
6
contrast, ALK/5a was found to present the DFGꢀout conformation, where the αCꢀhelix adopts
the open state accompanying with the JM domain in the helix conformation.
7
8
9
Both the pyrazolylꢀ5ꢀamine of 5a and pyridineꢀ2ꢀamine of crizotinib (Figure 6A, highlighted
in cyan) formed hydrogen bonds to the hinge residues Glu1197 and Met1199. However, the most
significant differences between 5a and crizotinib are that the tꢀbutylꢀisoxazole urea moiety
(Figure 6A, highlighted in purple) of 5a extended to the allosteric pocket, whereas the 2,6ꢀ
dichloroꢀ3ꢀfluoroꢀphenyl group (Figure 6A, highlighted in red) of crizotinib was oriented to
small hydrophobic cavity underneath the Pꢀloop instead. Conformational changes in ALK
induced by 5a, accompanied by additional interactions with the allosteric pocket, led to the slowꢀ
binding kinetics of 5a as compared with crizotinib. The binding association and dissociation rate
constants of 5a (kon: 5.17x104 1/Ms; koff: 2.4 x10ꢀ3 1/s) were about 226ꢀfold and 15ꢀfold slower
than crizotinib (kon: 1.17x107 1/Ms; koff: 3.55x10ꢀ2 1/s). The slower association rate of 5a may
due to the higher energy barrier for inducing ALK conformational change. Slow dissociationꢀrate
has shown to be correlated with the potency and prolonged drug effects both in vitro and in
vivo.18, 37, 38, 41
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
Next, the structure of 5a was further compared with 6, a typeꢀII ALK inhibitor (PDB code:
4FNY).22 Superimposing the crystal structure of 5a with 6 show differences between these two
structures. Firstly, the JM domain is visible in ALK/5a structure but absent in ALK/6 structure
(Figure 6D). Binding of 5a reorients the αCꢀhelix to an open state and allosteric effects regulate
the refolding of the JM domain to a short helix segment. However, in the ALK/6 structure, the
refolding of the JM domain is not observed due to the disordered in this region. Secondly, in the
ALK/5a structure, αꢀAL was unfolded to a nonꢀstructural loop and moved away from the αCꢀ
helix while αꢀAL was completely invisible in the ALK/6 structure.
ACS Paragon Plus Environment

Page 17 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Moreover, by superimposing the structure complex of 6 onto 5a, it revealed that pyrazolylꢀ5ꢀ
amine (Figure 6A, highlighted in cyan) of 5a formed three hydrogen bonds with hinge residues
Glu1197 and Met1199, while the nitrogen atom of the dimethoxyquinoline group (highlighted in
cyan) in 6 made one hydrogen bond with Met1199. Moreover, the phenyl group together with
the Nꢀmethylpiperidine group of 5a extend to the solvent exposed region of ALK to form
hydrophobic interactions with the surrounding residues and made chargeꢀcharge interactions
with the acidic zone. In contrast, in ALK/6 complex structure, the solvent exposed region is
unoccupied and two methoxy substitutions at quinoline group of 6 were much less exposed to the
bulk solvent and had no interactions with the acidic zone.
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
Discussion and Conclusions. In this study, 5a, a 3ꢀphenylꢀ1Hꢀ5ꢀpyrazolylamineꢀbased
compound bearing a urea substituent, was identified by inꢀhouse kinase screening as a novel
ALK inhibitor. Further structural biology studies of 5a with ALK revealed that 5a adopts a typeꢀ
II binding mode. The ureaꢀsubstituted moiety of 5a flips the DFGꢀmotif from an inꢀtoꢀout
inactive conformation and coordinates the orientation of 5a through the hydrogen bonds with
Glu1167 in the αCꢀhelix and Asp1270 in DFG motif. In addition, the tꢀbutyl group of 5a allows
for a hydrophobic interaction with the allosteric pocket created by the conformational change of
DFG motif. Interestingly, binding of 5a induces a dramatic conformational change of ALK by
spatial rearrangement of Aꢀloop, αCꢀhelix, and JM domain. 5a unfolded the Aꢀloop to a nonꢀ
structural loop and moved the αꢀAL away from the αCꢀhelix, resulting in the loss of the salt
bridge and hydrogen bond interactions with αCꢀhelix. The loss of the interactions between αꢀAL
and the αCꢀhelix reorients the αCꢀhelix 2 Å away from the Cꢀlobe to an opened state.
Furthermore, the binding of 5a results in the transformation of the JM domain from the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 48
1
2
3
4
5
6
7
8
9
antiparallel βꢀturn conformation to the helix conformation and the disorder of Nꢀterminal helix
segment of JM domain. The conformational change of JM domain significantly reduces the
contacted interface area with αCꢀhelix and disrupts the interactions with αꢀAL. As a consequence
of this conformational change and spatial rearrangement, the binding of 5a to the ALK kinase
domain alters the interaction network of Tyr1096, an important autophosphorylation residue
located in juxtamembrane domain, and may affect the autoinhibition mechanism of ALK and
downstream signal pathway regulation.
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
Further comparison of the coꢀcrystal structures of ALK/5a with ALK/4, revealed that the
substitutedꢀurea tail moiety controls the conversion between the DFGꢀin and DFGꢀout
conformation and reꢀallocate the Aꢀloop, αCꢀhelix, and JM domain. Further studies of 5a and 4
showed these two compounds have different molecular properties, such as the polarizability,
lipophilicity and molecular surface area. The polarizability of compound 5a and 4 are 62.85 Å3
and 46.99 Å3, respectively, indicating they have different electronic properties. The logP values
of compound 5a and 4 are 4.54 and 2.61, respectively, suggesting 5a exhibited higher
lipophilicity. Moreover, the molecular surface area of compound 5a is 863.6 Å2 and that of
compound 4 is 615.3 Å2, suggesting they have different impacts on steric effects with the
protein. The differences in physicochemical and structural properties of inhibitors would also
contribute to the conversion between the typeꢀI and typeꢀII binding modes and the
conformational change of ALK.
Furthermore, the ligand efficiency (LE) of 5a and analogues were calculated (Table 1). Based
on the structural observation from ALK/5a, it is suggested that replacing the Nꢀmethylpiperazine
group of 5a with a guanidine group may potentially provide electrostatic interactions and
hydrogen bond interactions with the acid region in ALK to improve the potency (Figure S3 in the
ACS Paragon Plus Environment

Page 19 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
supporting information). In addition, the replacement of Nꢀmethylpiperazine group of 5a with a
guanidine group would reduce the molecular weight (Figure S3 in the supporting information),
which would potentially improve the ligand efficiency of 5a.
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
To the best of our knowledge, this is the first structural biology study to directly observe how
changes in a small molecule can regulate the switch between the type I and type II binding
modes, and induce dramatic conformational changes in protein structures. Our results are
expected to be useful for activation/inhibition mechanism studies of ALK, and should give
insights into the development of the novel typeꢀII inhibitors for ALK or the other kinases in the
future.
Experimental Section
ALK Expression, Purification and Crystallization. Human ALK kinase domain (residues
1084ꢀ1410) was expressed using baculovirus system (Sf21 cells). The residue Cys1097 was
mutated to serine to facilitate protein expression and crystallization as reported previously.22 Cell
suspension was centrifuged at 4500 r.p.m. for 15 minutes and the cell pellet was suspended in
Tris buffer (50 mM TrisꢀHCl pH 7.5, 150 mM NaCl, 10 mM imidazole and 1mM TCEP) and
lysed by sonication. The supernatant was filtered with 0.22 filters and loaded onto HiTrap HP
Niꢀaffinity column (GE Life Sciences). ALK was eluted with 75 mM imidazole in wash buffer
(50 mM TrisꢀHCl, 150 mM NaCl, pH 7.5) and buffer exchanged to Factor Xa cleavage buffer
(50 mM TrisꢀHCl, 100 mM NaCl. 5 mM CaCl2, pH 8.0) by Hiprep 26/10 desalting column (GE
Life Sciences). His tag was then removed by Factor Xa (Novagen) cleavage at 4°C. ALK was
finally exchanged to the protein buffer (25 mM TrisꢀHCl, 125 mM NaCl, 14 mM βꢀME, pH 8.5)
and concentrated to 12ꢀ15 mg/ml.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 48
1
2
3
4
5
6
7
8
9
ALK was crystallized by sitting drop method. Equal volumes of reservoir solution (18%
PEG2000 MME, 0.1 M TrisꢀHCl pH 8.6) and ALK were mixed and grown at 4°C. ALK crystals
were soaked with 0.5 mM tested compounds for more than 6 hours. The ALK crystal was then
quickly immersed in the reservoir solution with additional 20% ethylene glycol and then cooled
in the liquid nitrogen. Diffraction data were collected at beamline BL15A1 (NSRRC, Taiwan),
BL12B2 and 44XU (Springꢀ8, Japan). Data were processed by HKL2000 program.42 The phase
of initial ALK structure was resolved by molecular replacement using PHENIX.43 Structure
refinement and model building were performed using PHENIX and COOTꢀ0.8.1.44 Figures were
prepared with PyMOL v1.7.2 (Schrodinger, LLC).
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
Kinase Inhibition Assay. In vitro kinase assays were carried out in 96ꢀwell plate with tested
compounds in the reaction buffer (25 mM HEPES pH 7.4, 10 mM MgCl2, 4 mM MnCl2, 0.5 mM
Na3VO4, 2 mM DTT, 0.02% TritonXꢀ100, and 0.01% BSA). Total of 50 ꢁl of reaction mixture
contained the testing compounds, l ꢁM ATP (Sigma), 2 ꢁM polyGlu4:Tyr peptide (Sigma), and
100 ng ALK protein. The reaction was performed at 30°C for an hour. The kinase assay was run
with phosphorylated ALK. Equal volume of kinaseꢀGlo plus reagent (Promega) was added after
one hour of incubation. The reaction mixture was transferred to a black plate and recorded the
luminescence by Wallac Vector 1420 multilabel counter (PerkinElmer,Shelton, CT, USA).
Surface Plasmon Resonance (SPR). The ligand was immobilized on the CM5 chip (GE Life
Sciences) using amineꢀcoupling method. The immobilization procedure was based on the default
setting on the BiacoreꢀT200 control software. PBS buffer (pH 7.5, contained 5% fresh DMSO)
was used as the running buffer. 5a (50 mM) in 100% DMSO was dissolved in 1x running buffer
to give the analyte solution concentration 1.25 to 0.046 ꢁM. Crizotinib (1mM) was used to
ACS Paragon Plus Environment

Page 21 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
prepare analyte solution 40 to 2.5 nM. Three cycles of startꢀup and one cycle of solvent
correction were performed before sampleꢀcycle run. Solvent correction (8 concentrations, ranged
from 4.5% to 5.5% DMSO in 1xPBS pH 8.5) was used to adjust response error due to the
content mismatch between running buffer and sample. The LMW singleꢀcycle kinetics analysis
consisted of 120 seconds of association (30 ꢁl/min), 300 seconds of dissociation (30 ꢁl/min), 30
seconds of carryꢀover control (40 ꢁl/min) and finally 60 seconds of surface stabilization (30
ꢁl/min). Except five concentrations of sample injection, two double reference cycleꢀrun were
repeated. BiacoreꢀT200 evaluation software version2.0 was used for the affinity and kinetic
analysis.
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
Molecular properties calculation. The polarizability and molecular surface area of
compound 5a and 4 were calculated by MarvinSketch program implemented in ChemAxon.27
The logP value of compound 5a and 4 were calculated by Chemdraw Ultra 7.0.
Chemistry methods. All commercial chemicals and solvents are reagent grade and were used
1
without further treatment unless otherwise noted. H NMR spectra were obtained with a Varian
Mercuryꢀ300 or a Varian Mercuryꢀ400 spectrometer. Chemical shifts were recorded in parts per
million (ppm, δ) and were reported relative to the solvent peak or TMS. LC/MS data were
measured on an Agilent MSDꢀ1100 ESIꢀMS/MS System. Highꢀresolution mass spectra (HRMS)
were measured with a Thermo Finnigan (TSQ Quantum) electrospray ionization (ESI) mass
spectrometer. Flash column chromatography was done using silica gel (Merck Kieselgel 60, No.
9385, 230ꢀ400 mesh ASTM). Reactions were monitored by TLC using Merck 60 F254 silica gel
glass backed plates (5 × 10 cm); zones were detected visually under ultraviolet irradiation (254
nm) or by spraying with phosphomolybdic acid reagent (Aldrich) followed by heating at 80 °C.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 48
1
2
3
4
5
6
7
8
9
All starting materials and amines were commercially available unless otherwise indicated. The
purity of compounds was determined by a Hitachi 2000 series HPLC system. Compound purities
were determined by reverse phase C18 column (Agilent ZORBAX Eclipse XDBꢀC18 5 µm, 4.6
mm × 150 mm) under the following gradient elution condition: Mobile phase Aꢀacetonitrile
(10% to 90%, 0 to 45 min) and mobile phase Bꢀ10 mM NH4OAc aqueous solution containing
0.1% formic acid (90% to 10%, 0 to 45 min). The flowꢀrate was 0.5 ml/min and the injection
volume was 20 µl. The system operated at 25 °C. Peaks were detected at 254 nm.
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
Nꢀ[3ꢀ(4ꢀAminoꢀ3ꢀmethylphenyl)ꢀ1Hꢀpyrazolꢀ5ꢀyl]ꢀ4ꢀ[(4ꢀmethylpiperazinꢀ1ꢀ
yl)methyl]benzamide (4). Mp 229–230 °C. 1H NMR (300 MHz, DMSOꢀd6): 11.12 (s, 1H), 8.11
(d, J = 8.1 Hz, 2H), 8.00 (d, J = 8.4 Hz, 1H), 7.84 (d, J = 8.1 Hz, 2H), 7.77 (s, 1H), 7.70 (d, J =
8.1 Hz, 1H), 7.54 (d, J = 8.4 Hz, 1H), 7.03 (s, 1H), 4.51 (bs, 2H), 3.78ꢀ3.30 (m, 6H), 3.15 (s,
3H), 2.90ꢀ2.70 (m, 4H), 2.43 (s, 3H); MS (ES+) m/z calcd. for C23H28N6O: 404.23; found: 405.2
(M+H+). HPLC: tR = 14.09 min, 95.0%.
N1ꢀ(3ꢀ4ꢀ[([5ꢀ(Tertꢀbutyl)ꢀ3ꢀisoxazolyl]aminocarbonyl)amino]ꢀ3ꢀmethylphenylꢀ1Hꢀ5ꢀ
1
pyrazolyl)ꢀ4ꢀ[(4ꢀmethylpiperazino)methyl]benzamide (5a). Mp 263–264 °C. H NMR (300
MHz, DMSOꢀd6): 10.78 (s, 1H), 9.95 (s, 1H), 8.35 (s, 1H), 7.99ꢀ7.97 (m, 3H), 7.73 (s, 1H), 7.57
(d, J = 8.4 Hz, 1H), 7.41 (d, J = 8.1 Hz, 2H), 7.11 (s, 1H), 6.48 (s, 1H), 3.52 (s, 2H), 3.35 (s, 4H),
2.37 (s, 4H), 2.30 (s, 3H), 2.15 (s, 3H), 1.30 (s, 9H); MS (ES+) m/z calcd. for C31H38N8O3:
570.31; found: 571.4 (M+H+); HRMS (ESI) calcd. for C31H39N8O3: 571.3145; found: 571.3136
(M+H+). HPLC: tR = 22.66 min, 99.1%.
Nꢀ[3ꢀ(4ꢀ{[(3ꢀFluorophenyl)carbamoyl]amino}phenyl)ꢀ1Hꢀpyrazolꢀ5ꢀyl]ꢀ4ꢀ[(4ꢀ
1
methylpiperazinꢀ1ꢀyl)methyl]benzamide (5b). Mp 225–227 °C. H NMR (400 MHz, DMSOꢀ
ACS Paragon Plus Environment

Page 23 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
d6): 12.80 (s, 1H), 10.78 (s, 1H), 9.07 (s, 1H), 8.99 (s, 1H), 7.99 (d, J = 8.0 Hz, 2H), 7.69 (d, J =
8.4 Hz, 2H), 7.54 (d, J = 8.8 Hz, 2H), 7.49 (s, 1H), 7.41 (d, J = 7.6 Hz, 2H), 7.31 (q, J = 7.8 Hz,
1H), 7.13 (d, J = 8.0 Hz, 1H), 6.97 (s, 1H), 6.79 (td, J = 8.4, 2.4 Hz , 1H), 3.67 (s, 2H), 2.62ꢀ2.00
(m, 8H), 2.16 (s, 3H); MS (ES+) m/z calcd. for C29H30FN7O2: 527.24; found: 528.2 (M+H+).
HPLC: tR = 21.21 min, 98.7%.
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
N1ꢀ[3ꢀ(4ꢀ[([5ꢀ(Tertꢀbutyl)ꢀ3ꢀisoxazolyl]aminocarbonyl)amino]methylphenyl)ꢀ1Hꢀ5ꢀ
o
1
pyrazolyl]ꢀ4ꢀ[(4ꢀmethylpiperazino)methyl]benzamide (5c). Mp 226–227 C. H NMR (300
MHz, DMSOꢀd6): 10.82 (s, 1H), 10.71 (s, 1H), 9.50 (s, 1H), 7.98 (d, J = 8.1 Hz, 2H), 7.72 (d, J
= 8.1 Hz, 2H), 7.41 (d, J = 8.1 Hz, 2H), 7.36 (t, J = 8.4 Hz, 2H), 7.02 (t, J = 5.7 Hz, 2H), 6.38 (s,
1H), 4.35 (d, J = 5.7 Hz, 2H), 3.52 (s, 2H), 3.25 (s, 4H), 2.37 (s, 4H), 2.15 (s, 3H), 1.27 (s, 9H);
MS (ES+) m/z calcd. for C31H38N8O3: 570.31; found: 571.4 (M+H+); HRMS (ESI) calcd. for
C31H39N8O3: 571.3140; found: 571.3147 (M+H+). HPLC: tR = 20.98 min, 98.1%.
4ꢀ[(4ꢀMethylpiperazinꢀ1ꢀyl)methyl]ꢀNꢀ{3ꢀ[3ꢀmethylꢀ4ꢀ({[5ꢀ(propanꢀ2ꢀyl)ꢀ1,2ꢀoxazolꢀ3ꢀ
1
yl]carbamoyl}amino)phenyl]ꢀ1Hꢀpyrazolꢀ5ꢀyl}benzamide (5d). Mp 235–236 °C. H NMR
(400 MHz, DMSOꢀd6): 12.80 (s, 1H), 10.78 (s, 1H), 9.93 (s, 1H), 8.35 (s, 1H), 8.05ꢀ7.95 (m,
3H), 7.62 (s, 1H), 7.57 (d, J = 8.8 Hz, 1H), 7.41 (d, J = 7.6 Hz, 2H), 7.00 (s, 1H), 6.49 (s, 1H),
3.52 (s, 2H), 3.02 (sep, J = 7.2 Hz, 1H), 2.60ꢀ2.00 (m, 8H), 2.37 (s, 3H), 1.25 (d, J = 7.2 Hz,
6H); MS (ES+) m/z calcd. for C30H36N8O3: 556.29; found: 557.3 (M+H+). HRMS (ESI) calcd.
for C30H37N8O3: 557.2989; found: 557.2994 (M+H+). HPLC: tR = 21.79 min, 99.9%.
N1ꢀ3ꢀ[4ꢀ([(5ꢀEthylꢀ3ꢀisoxazolyl)amino]carbonylamino)ꢀ3ꢀmethylphenyl]ꢀ1Hꢀ5ꢀpyrazolylꢀ
1
4ꢀ[(4ꢀmethylpiperazino)methyl]benzamide (5e). Mp 225–227 °C. H NMR (400 MHz,
DMSOꢀd6): 12.80 (s, 1H), 10.77 (s, 1H), 9.91 (s, 1H), 8.36 (s, 1H), 7.99ꢀ7.97 (m, 3H), 7.62ꢀ7.56
(m, 2H), 7.41 (d, J = 8.0 Hz, 2H), 7.00 (s, 1H), 6.52 (s, 1H), 3.50 (s, 2H), 3.17 (d, J = 4.8 Hz,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 48
1
2
3H), 2.67 (q, J = 7.6 Hz, 2H), 2.37 (br s, 8H), 2.15 (s, 3H), 1.22 (t, J = 7.6 Hz, 3H); MS (ES+)
m/z calcd. for C29H34N8O3: 542.28; found: 543.3 (M+H+). HRMS (ESI) calcd. for C29H35N8O3:
543.2832; found: 543.2839 (M+H+). HPLC: tR = 20.16 min, 95.1%.
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
N1ꢀ3ꢀ[3ꢀMethylꢀ4ꢀ([(5ꢀmethylꢀ3ꢀisoxazolyl)amino]carbonylamino)phenyl]ꢀ1Hꢀ5ꢀ
pyrazolylꢀ4ꢀ[(4ꢀmethylpiperazino)methyl]benzamide (5f). Mp 155–157 °C.1H NMR (300
MHz, DMSOꢀd6): 10.78 (s, 1H), 9.89 (s, 1H), 8.36 (s, 1H), 7.99ꢀ7.97 (m, 3H), 7.62 (s, 1H), 7.57
(d, J = 8.7 Hz, 1H), 7.41 (d, J = 8.1 Hz, 2H), 7.00 (s, 1H), 6.52 (s, 1H), 3.52 (s, 2H), 3.36 (s, 4H),
2.37 (s, 7H), 2.30 (s, 3H), 2.15 (s, 3H); MS (ES+) m/z calcd. for C28H32N8O3: 528.26; found:
529.3 (M+H+). HRMS (ESI) calcd. for C28H32N8O3: 528.2597; found: 528.2598 (M+). HPLC: tR
= 18.55 min, 99.5%.
N1ꢀ(3ꢀ4ꢀ[([5ꢀ(Tertꢀbutyl)ꢀ3ꢀisoxazolyl]aminocarbonyl)amino]phenylꢀ1Hꢀ5ꢀpyrazolyl)ꢀ4ꢀ
1
[(4ꢀmethylpiperazino)methyl]benzamide (5g). Mp 247–248 °C. H NMR (300 MHz, DMSOꢀ
d6): 10.79 (s, 1H), 9.73 (s, 1H), 9.23 (s, 1H), 7.98 (d, J = 8.1 Hz, 2H), 7.69 (d, J = 8.4 Hz, 2H),
7.54 (d, J = 8.4 Hz, 2H), 7.41 (d, J = 8.1 Hz, 2H), 6.97 (s, 1H), 6.52 (s, 1H), 3.51 (d, J = 5.1 Hz,
2H), 3.28 (t, J = 4.8 Hz, 4H), 2.37 (t, J = 4.8 Hz, 4H), 2.16 (s, 3H), 1.30 (s, 9H); MS (ES+) m/z
calcd. for C30H36N8O3: 556.29; found: 557.4 (M+H+); HRMS (ESI) calcd. for C30H37N8O3:
557.2993; found: 557.2988 (M+H+). HPLC: tR = 22.29 min, 93.3%.
4ꢀ[(4ꢀMethylpiperazinꢀ1ꢀyl)methyl]ꢀNꢀ{3ꢀ[4ꢀ({[5ꢀ(propanꢀ2ꢀyl)ꢀ1,2ꢀoxazolꢀ3ꢀ
1
yl]carbamoyl}amino)phenyl]ꢀ1Hꢀpyrazolꢀ5ꢀyl}benzamide (5h). Mp 229–230 °C. H NMR
(400 MHz, DMSOꢀd6): 12.82 (s, 1H), 10.78 (s, 1H), 9.60 (s, 1H), 9.09 (s, 1H), 7.98 (d, J = 8.0
Hz, 2H), 7.69 (d, J = 8.8 Hz, 2H), 7.54 (d, J = 8.4 Hz, 2H), 7.41 (d, J = 7.6 Hz, 2H), 6.98 (s, 1H),
6.54 (s, 1H), 3.52 (s, 2H), 3.03 (sep, J = 6.8 Hz, 1H), 2.60ꢀ2.02 (m, 8H), 2.16 (s, 3H), 1.25 (d, J
ACS Paragon Plus Environment

Page 25 of 48
Journal of Medicinal Chemistry
1
2
= 6.8 Hz, 6H); MS (ES+) m/z calcd. for C29H34N8O3: 542.28; found: 543.3 (M+H+). HPLC: tR =
21.28 min, 97.2%.
3
4
5
6
7
8
9
N1ꢀ(3ꢀ4ꢀ[([5ꢀ(Tertꢀbutyl)ꢀ3ꢀisoxazolyl]aminocarbonyl)amino]phenylꢀ1Hꢀ5ꢀpyrazolyl)ꢀ4ꢀ(2ꢀ
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
o
1
tetrahydroꢀ1Hꢀ1ꢀpyrrolylethoxy)benzamide (5i). Mp 256–257 C. H NMR (400 MHz,
DMSOꢀd6): 12.79 (s, 1H), 10.68 (s, 1H), 9.59 (s, 1H), 8.99 (s, 1H), 8.02 (d, J = 8.8 Hz, 2H),
7.69 (d, J = 8.8 Hz, 2H), 7.54 (d, J = 8.4 Hz, 2H), 7.04 (d, J = 8.8 Hz, 2H), 6.96 (bs, 1H), 6.52 (s,
1H), 4.16 (t, J = 5.8 Hz, 2H), 2.83 (t, J = 5.8 Hz, 2H), 2.62ꢀ2.50 (m, 4H), 1.80ꢀ1.64 (m, 4H), 1.32
(s, 9H); MS (ES+) m/z calcd. for C30H35N7O4: 557.64; found: 558.6(M+H+); HRMS (ESI) calcd.
for C30H36N7O4: 558.2823; found: 558.2841 (M+H+). HPLC: tR = 42.79 min, 95.6%.
Nꢀ[3ꢀ4ꢀ{[(5ꢀ(Tertꢀbutyl)ꢀ1,2ꢀoxazolꢀ3ꢀyl)carbamoyl]amino}phenyl]ꢀ1Hꢀpyrazolꢀ5ꢀyl]ꢀ4ꢀ(3ꢀ
1
(pyrrolidinꢀ1ꢀyl)propoxy]benzamide (5j). Mp 172–174 °C. H NMR (300 MHz, DMSOꢀd6):
10.65 (s, 1H), 9.77 (s, 1H), 9.61 (s, 1H), 7.98 (d, J = 9.3 Hz, 2H), 7.64 (d, J = 8.4 Hz, 2H), 7.59
(d, J = 8.4 Hz, 2H), 6.99 (d, J = 9.0 Hz, 2H), 6.86 (bs, 1H), 6.48 (s, 1H), 4.08 (t, J = 6.2 Hz, 2H),
3.10ꢀ2.60 (m, 6H), 2.03 (t, J = 7.1 Hz, 2H), 1.85ꢀ1.60 (m, 4H), 1.25 (s, 9H); MS (ES+) m/z calcd.
for C31H37N7O4: 571.29; found: 572.1(M+H+). HPLC: tR = 24.24 min, 95.6%.
N1ꢀ(3ꢀ4ꢀ[([5ꢀ(Tertꢀbutyl)ꢀ3ꢀisoxazolyl]aminocarbonyl)amino]ꢀ3ꢀmethylphenylꢀ1Hꢀ5ꢀ
o
1
pyrazolyl)ꢀ4ꢀ(2ꢀmorpholinoethoxy)benzamide (5k). Mp 258–260 C. H NMR (300 MHz,
DMSOꢀd6): 12.77 (s, 1H), 10.69 (s, 1H), 9.98 (s, 1H), 8.38 (s, 1H), 7.99 (t, J = 9.6 Hz, 3H),
7.61ꢀ7.55 (m, 2H), 7.04 (d, J = 9.0 Hz, 2H), 6.91 (s, 1H), 6.48 (s, 1H), 4.16 (t, J = 5.7 Hz, 2H),
3.58 (t, J = 4.5 Hz, 4H), 2.71 (t, J = 5.7 Hz, 2H), 2.50ꢀ2.46 (m, 4H), 2.30 (s 3H), 1.30 (s, 9H);
MS (ES+) m/z calcd. for C31H37N7O5: 587.29; found: 588.4 (M+H+); HRMS (ESI) calcd. for
C31H38N7O5: 588.2934; found: 588.2926 (M+H+). HPLC: tR = 23.60 min, 98.0%.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 48
1
2
3
4
5
6
7
8
9
N1ꢀ(3ꢀ4ꢀ[([5ꢀ(Tertꢀbutyl)ꢀ3ꢀisoxazolyl]aminocarbonyl)amino]phenylꢀ1Hꢀ5ꢀpyrazolyl)ꢀ4ꢀ(4ꢀ
methylpiperazino)benzamide (5l). Mp 332–334 oC. 1H NMR (400 MHz, DMSOꢀd6): 12.74 (s,
1H), 10.51 (s, 1H), 9.58 (s, 1H), 8.99 (s, 1H), 7.92 (d, J = 9.2 Hz, 2H), 7.69 (d, J = 8.4 Hz, 2H),
7.53 (d, J = 8.8 Hz, 2H), 6.99ꢀ6.92 (m, 3H), 6.53 (s, 1H), 3.29ꢀ3.27 (m, 4H), 2.47ꢀ2.43 (m, 4H),
2.23 (s, 3H), 1.27 (s, 9H); MS(ES+) m/z calcd. for C29H34N8O3: 542.28; found: 543.3 (M+H+);
HRMS (ESI) calcd. for C29H34N8O3: 543.2827; found: 543.2832 (M+H+). HPLC: tR = 22.70 min,
99.9%.
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
ACS Paragon Plus Environment

Page 27 of 48
Journal of Medicinal Chemistry
1
2
3
4
ASSOCIATED CONTENT
5
6
Supporting Information
7
8
9
Binding kinetics analysis of compound 5a, 4 and crizotinib. The density maps of 5a, 4 and 5d in
the active site of ALK. The docking position of the designed ALK inhibitor in the binding site of
ALK. This material is available free of charge via the Internet at http://pubs.acs.org.
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
PDB ID Codes: 5IUG (ALK/5a), 5IUH(ALK/5d), 5IUI(ALK/4)
Authors will release the atomic coordinates with experimental data upon article publication.
AUTHOR INFORMATION
Corresponding Author
* Phone: +886ꢀ37ꢀ246ꢀ166 extension 35713. Fax: + 886ꢀ37ꢀ586ꢀ456. Email: suying@nhri.org.tw
for SYW. Phone: +886ꢀ37ꢀ246ꢀ166 extension 35712. Fax: + 886ꢀ37ꢀ586ꢀ456. Email:
wtjiaang@nhri.org.tw for WTJ
Funding Sources
This work was supported by National Health Research Institute, Taiwan, ROC and National
Science Council (NSC 104ꢀ2325ꢀBꢀ400 ꢀ003).
ABBREVIATIONS
ALK, anaplastic lymphoma kinase; NHL, nonꢀHodgkin’s lymphoma; CARS, cysteinylꢀtRNA
synthetase ; IMT, inflammatory myofibroblastic tumors; DLBCL, diffuse large Bꢀcell
lymphoma; EML4, echinoderm microtubule associated protein like 4; NSCLC, nonꢀsmallꢀcell
lung carcinoma; DFG, AspꢀPheꢀGly; HRD, HisꢀArgꢀAsp; JM domain, juxtamembrane domain;
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 48
1
2
3
4
5
6
Aꢀloop, activation loop; Pꢀloop, phosphateꢀbinding loop; Rꢀspine, regulatory spine; Cꢀspine,
catalytic spine.
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
ACS Paragon Plus Environment

Page 29 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
References
(1) Morris, S. W.; Kirstein, M. N.; Valentine, M. B.; Dittmer, K. G.; Shapiro, D. N.; Saltman,
7
8
9
D. L.; Look, A. T. Fusion of a kinase gene, ALK, to a nucleolar protein gene, NPM, in nonꢀ
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
Hodgkin's lymphoma. Science 1994, 263, 1281ꢀ1284.
(2)
human cancer biology. Nat. Rev. Cancer 2013, 13, 685ꢀ700.
(3) Roskoski, R., Jr. Anaplastic lymphoma kinase (ALK): structure, oncogenic activation,
and pharmacological inhibition. Pharmacol. Res. 2013, 68, 68ꢀ94.
(4) Sasaki, T.; Janne, P. A. New strategies for treatment of ALKꢀrearranged nonꢀsmall cell
lung cancers. Clin. Cancer Res. 2011, 17, 7213ꢀ7218.
(5) Sakamoto, H.; Tsukaguchi, T.; Hiroshima, S.; Kodama, T.; Kobayashi, T.; Fukami, T. A.;
Hallberg, B.; Palmer, R. H. Mechanistic insight into ALK receptor tyrosine kinase in
Oikawa, N.; Tsukuda, T.; Ishii, N.; Aoki, Y. CH5424802, a selective ALK inhibitor capable of
blocking the resistant gatekeeper mutant. Cancer Cell 2011, 19, 679ꢀ690.
(6)
Cui, J. J.; TranꢀDube, M.; Shen, H.; Nambu, M.; Kung, P. P.; Pairish, M.; Jia, L.; Meng,
J.; Funk, L.; Botrous, I.; McTigue, M.; Grodsky, N.; Ryan, K.; Padrique, E.; Alton, G.;
Timofeevski, S.; Yamazaki, S.; Li, Q.; Zou, H.; Christensen, J.; Mroczkowski, B.; Bender, S.;
Kania, R. S.; Edwards, M. P. Structure based drug design of crizotinib (PFꢀ02341066), a potent
and selective dual inhibitor of mesenchymalꢀepithelial transition factor (cꢀMET) kinase and
anaplastic lymphoma kinase (ALK). J. Med. Chem. 2011, 54, 6342ꢀ6363.
(7)
Marsilje, T. H.; Pei, W.; Chen, B.; Lu, W.; Uno, T.; Jin, Y.; Jiang, T.; Kim, S.; Li, N.;
Warmuth, M.; Sarkisova, Y.; Sun, F.; Steffy, A.; Pferdekamper, A. C.; Li, A. G.; Joseph, S. B.;
Kim, Y.; Liu, B.; Tuntland, T.; Cui, X.; Gray, N. S.; Steensma, R.; Wan, Y.; Jiang, J.; Chopiuk,
G.; Li, J.; Gordon, W. P.; Richmond, W.; Johnson, K.; Chang, J.; Groessl, T.; He, Y. Q.;
ACS Paragon Plus Environment