Purine-Type Compounds Induce Microtubule Fragmentation and Lung Cancer Cell Death through Interaction with Katanin

Article abstract of DOI:10.1021/acs.jmedchem.6b00797

Microtubule targeting agents (MTAs) constitute a class of drugs for cancer treatment. Despite many MTAs have been proven to significantly improve the treatment outcomes of various malignancies, resistance has usually occurred. By selection from a two million entry chemical library based on the efficacy and safety, we identified purine-type compounds that were active against lung small cell lung cancer (NSCLC). The purine compound 5a (GRC0321) was an MTA with good effects against NSCLC. Lung cancer cells H1975 treated with 5a could induce microtubule fragmentation, leading to G2/M cell cycle arrest and intrinsic apoptosis. Compound 5a directly targeted katanin and regulated the severing activity of katanin, which cut the cellular microtubules into short pieces and activated c-Jun N-terminal kinases (JNK). The microtubule fragmenting effect of 5a is a unique mechanism in MTAs. It might overcome the resistance problems that most of the MTAs have faced.

Full text of DOI:10.1021/acs.jmedchem.6b00797

Subscriber access provided by Northern Illinois University
Article
Purine-type Compounds Induce Microtubule Fragmentation
and Lung Cancer Cell Death through Interaction with Katanin
Ting-Chun Kuo, Ling-Wei Li, Szu-Hua Pan, Jim-Min Fang, Jyung-Hurng Liu, Ting-Jen R. Cheng, Chia-Jen
Wang, Pei-Fang Hung, Hsuan-Yu Chen, Tse-Ming Hong, Yuan-Ling Hsu, Chi-Huey Wong, and Pan-Chyr Yang
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00797 • Publication Date (Web): 18 Aug 2016
Downloaded from http://pubs.acs.org on August 19, 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 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Purine-type Compounds Induce Microtubule Fragmentation and Lung Cancer Cell Death
through Interaction with Katanin
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
¶
¥
, ¶, #, £
, ¥
¢, §, ǀ, ǂ
TingꢀChun Kuo , LingꢀWei Li , SzuꢀHua Pan*
, JimꢀMin Fang* , JyungꢀHurng Liu
,
ƞ
&
,
ζ
&
ȝ
ø
TingꢀJen Cheng , ChiaꢀJen Wang
, PeiꢀFang Hung , HsuanꢀYu Chen , TseꢀMing Hong ,
#
ƞ, δ
ǁ, ȹ, ʘ, δ
YuanꢀLing Hsu , ChiꢀHuey Wong , and PanꢀChyr Yang
.
¶
Ph.D. Program in Translational Medicine, National Taiwan University and Academia Sinica,
Taipei, Taiwan
¥
#
Department of Chemistry, National Taiwan University, Taipei 106, Taiwan.
Graduate Institute of Medical Genomics and Proteomics, College of Medicine, National
Taiwan University, Taipei 100, Taiwan
£
Genome and Systems Biology Degree Program, National Taiwan University and Academia
Sinica, Taipei 106, Taiwan
¢
Department of Life Sciences, National Chung Hsing University, Taichung 402, Taiwan
Institute of Genomics and Bioinformatics, National Chung Hsing University, Taichung 402,
§
Taiwan
ǀ
Agricultural Biotechnology Center, National Chung Hsing University, Taichung 402,
Taiwan
ǂ
Rong Hsing Research Center for Translational Medicine, National Chung Hsing University
4
02, Taichung, Taiwan.
ƞ
Genomics Research Center, Academia Sinica, Taipei 115, Taiwan.
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
&
Department of Internal Medicine, National Taiwan University College of Medicine, Taipei
100, Taiwan
ζ
ȝ
Institute of Stem Cell and Translational Cancer Research, Chang Gung Memorial Hospital at
Institute of Statistical Science, Academia Sinica, Taipei 115, Taiwan
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Ø Institute of Clinical Medicine, National Cheng Kung University College of Medicine,
Tainan 701, Taiwan.
ȹ Institute of Biomedical Sciences, Academia Sinica, Taipei 115, Taiwan
ʘ NTU Center for Genomic Medicine, National Taiwan University, Taipei 100, Taiwan.
Δ Senior authors.
*
Corresponding authors.
KEYWORDS
Nonꢀsmallꢀcell lung cancer; Microtubule targeting agent; Microtubule fragmentation; Katanin.
ACS Paragon Plus Environment
2

Page 3 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ABSTRACT
Microtubule targeting agents (MTAs) constitute a class of drugs for cancer treatment.
Despite many MTAs have been proven to significantly improve the treatment outcomes of
various malignancies, resistance has usually occurred. By selection from a 2ꢀmillion entry
chemical library based on the efficacy and safety, we identified purineꢀtype compounds that were
active against lung small cell lung cancer (NSCLC). The purine compound 5a (GRC0321) was
an MTA with good effects against NSCLC. Lung cancer cells H1975 treated with 5a could
induce microtubule fragmentation, leading to G2/M cell cycle arrest and intrinsic apoptosis.
Compound 5a directly targeted katanin and regulated the severing activity of katanin, which cut
the cellular microtubules into short pieces and activated cꢀJun Nꢀterminal kinases (JNK). The
microtubule fragmenting effect of 5a is a unique mechanism in MTAs. It might overcome the
resistance problems that most of the MTAs have faced.
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
1
Lung cancer remains the leading cause of cancerꢀrelated mortality worldwide . Although
treatment advances have been made over the past 20 years, the response rate and prognosis of
2
, 3
lung cancer patients remain poor , and drug resistance has become one of the main causes of
4
treatment failure . Therefore, the development of antiꢀcancer drugs with novel mechanisms for
lung cancer treatment is urgently needed.
Microtubules, long, hollow cylinders composed of polymerized αꢀ and βꢀtubulin dimers, are
an important component of the cytoskeleton. They play an irreplaceable role in a variety of
5
cellular functions, including the maintenance of cell shape, polarity, vesicles and mitosis .
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
During mitosis, microtubules form the mitotic spindle, which separates the daughter
chromosomes to the opposite ends of the poles. Because of this characteristic, microtubules
5
ꢀ7
become a suitable target for antiꢀcancer agents to kill rapidly dividing cancer cells . Usually,
the tubular polymers of tubulin can grow as long as 50 ꢁm, and the processes of microtubule
assembly and disassembly are highly dynamic and are controlled by GTP binding and hydrolysis
events. Regulating this unstable characteristic of microtubules is crucial for cells to respond to
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
8
, 9
different circumstances . Until now, many microtubuleꢀtargeting agents (MTAs) have been
found to target this property of the microtubules, and they can be further classified into two
groups: stabilizing and destabilizing agents. The microtubuleꢀstabilizing agents, including
taxanes and epothilones, can stabilize the polymer formation of microtubules and increase the
polymer mass of the microtubules in the cells. In contrast, microtubuleꢀdestabilizing agents, such
as the vinca alkaloids, depolymerize microtubules and decrease the polymer mass. To date,
taxanes and a series of vinca alkaloids are two major MTAs that are used in current
5
, 10
chemotherapeutic regimes with excellent effects on different human malignancies . However,
after a prolonged treatment, the clinical activity of the MTAs will usually be reduced due to drug
1
1, 12
resistance
. As a result, there is an ongoing effort to develop novel agents within each class,
to improve their efficacy and overcome drug resistance.
In addition of developing novel MTAs, targeting microtubuleꢀassociated proteins (MAPs) is
1
3
identified as an alternative mechanism to overcome the drug resistance problem of MTAs . The
MAPs are a class of proteins that associate with microtubules, usually at the ends of the
filaments. They regulate the dynamic behaviors of the microtubules by enhancing the rate of
1
3
microtubule growth or increasing the frequency of catastrophes of microtubules . Among MAPs,
stathmin and tau are two important target proteins for cancer therapy. Inhibiting the expression
ACS Paragon Plus Environment
4

Page 5 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of stathmin by antiꢀstathmin ribozyme, either as monotherapy or in combination with
chemotherapeutic agents, has been demonstrated to reduce cell proliferation and increase
1
4
apoptosis in breast cancer cells . The expression level of tau protein is associated with the
sensitivity of paclitaxel in breast cancer cells, thus decreasing tau expression would make breast
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
1
5
cancer cells hypersensitive to paclitaxel . In addition, a class of proteins known as microtubule
severing proteins can also bind to microtubules and regulate their dynamics by cutting
1
6
microtubules into short fragments . For example, katanin, one of the microtubule severing
proteins, is a heterodimer hydrolase that is composed of two subunits called p60 and p80,
according to their molecular weights. The p60 subunit is the catalytic subunit that contains the
AAA (ATPase associated with diverse cellular activities) region, which functions in microtubule
1
7, 18
severing activity. The other subunit, p80, serves a regulatory function
. Although katanin has
received little attention in the context of cancer in the past, some of the literature has recently
revealed its potential roles in cancer formation and progression. Ye and colleagues found that the
1
9
expression of katanin p60 was upꢀregulated in prostatic adenocarcinomas . Sudo and Maru also
found that LAPSER1, a tumor suppressor gene in prostate cancer, participated in cytokinesis by
interacting with katanin p80, and the disruption of this interaction might cause genetic instability
2
0
and cancer .
Recently, our laboratory has been devoted to screening novel effective inhibitors to
overcome drug resistance in nonꢀsmallꢀcell lung cancer (NSCLC). In this study, GRC0321
compound 5a) was one of the compounds that was selected from a 2ꢀmillion compound library
(
through a twoꢀstep highꢀthroughput screening of several lung cancer cell lines with different
epidermal growth factor receptor (EGFR) status (including EGFR wild type and EGFR
21
mutants) . The purineꢀtype compound 5a was selected due to its excellent in vitro and in vivo
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
activities against NSCLC. We demonstrated that 5a could disrupt the microtubule into small
fragments in NSCLC cells by interacting with microtubule severing proteins katanin p60 and
p80. The damage further activated cꢀJun Nꢀterminal kinases (JNK)ꢀrelated signaling pathways,
which caused G2/M cell cycle arrest and triggered mitochondriaꢀmediated intrinsic apoptosis.
Our results demonstrated that microtubule severing proteins could serve as a target for antiꢀ
cancer therapy.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
RESULTS
Identification of purine-type compounds that possess NSCLC inhibitory activities.
Through highꢀthroughput screening of efficient drugs for the treatment of NSCLC cells, we
found a purineꢀtype compound 5a, which was active to several lung cancer cell lines with IC50
values of 0.23–0.33 ꢁM (Table 1). The cytotoxic effect of 5a in NSCLC cells is less related to
EGFR status in these cells.
Table 1. Inhibitory activities of 5a against different lung cell lines.
a
a
Cell line
H1975
PC9
IC (ꢁM)
Cell line
CL1ꢀ0
CL1ꢀ5
A549
IC (ꢁM)
5
0
50
0.24 ± 0.03
0.33 ± 0.01
0.25 ± 0.01
21.7 ± 1.6
0.24 ± 0.01
0.28 ± 0.01
0.23 ± 0.02
PC9/gef
NBE
a
CL1ꢀ0, CL1ꢀ5 and A549 are human lung adenocarcinoma cell lines with wideꢀtype EGFR. PC9 is a human
lung adenocarcinoma cell line with EGFR exonꢀ19 deletion. H1975 and PC9/gef are two human lung
adenocarcinoma cell lines with different resistant mechanisms to EGFR TKIs. H1975 harbors EGFR T790M
ACS Paragon Plus Environment
6

Page 7 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
mutation and PC9/gef acquires EGFR TKIs resistance by longꢀterm and lowꢀdose treatment of EGFR TKIs in vitro.
NBE is a human normal bronchial epithelial cell.
To further investigate the structure–activity relationship, we synthesized a series of derivatives
with varied aryl moieties at the C2 and N9 positions. Scheme 1 shows the synthetic route to the
purine compounds 5a–g by consecutive coupling reactions of diaminomaleonitrile (1) with
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
2
2
isocyanates 2a–c and aldehydes 4a–e. The phenyl isocyanates 2a–c (1.05 equiv) that bear 2ꢀ
methoxy, 3ꢀCO Me or 4ꢀCO Me substituents were added to diaminomaleonitrile to afford the
2
2
corresponding urea compounds 3a–c, which then underwent condensation reactions with
substituted benzaldehydes 4a–e (1.05–2.0 equiv) to yield the desired 8ꢀoxopurine products 5a–g,
which have an amido group at C6 and two aryl substituents at the N9 and C2 positions.
Scheme 1. Synthesis of purineꢀtype compounds 5a–5g.
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The growth inhibitory activity of 5a–5g against H1975 was examined, and the IC values
5
0
1
were determined by MTT assays (Table 2). The effect of the R substituent in the N9ꢀaryl group
was variable. Compound 5e, which has a CO Me substituent at the metaꢀposition of the N9ꢀaryl
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
group, showed superior inhibitory activity (IC = 0.19 ꢁM) compared with the hit 5a (IC =
5
0
50
0.30 ꢁM), whereas the analog 5f has a paraꢀCO Me substituent that exhibited lower activity
2
2
(
IC = 4.73 ꢁM). In contrast, the effect of the R substituent in the C2ꢀaryl group was dramatic.
5
0
2
Alternation of R from ethoxy in 5a to hydroxyl (5b), butoxy (5c), hexoxy (5d) or phenyl (5g)
caused a large decrease in the inhibition.
Table 2. Inhibitory activities of purineꢀtype compounds against the growth of H1975 lung cancer
cells.
Compound
IC (ꢁM)
Compound
IC (ꢁM)
5
0
50
5a
0.30 ± 0.03
5.80 ± 0.13
4.66 ± 0.05
> 5.0
5f
5g
6e
8
4.73 ± 0.16
> 20.0
5b
5c
7.91 ± 0.18
> 10.0
5d
5e
0.19 ± 0.02
9
26.3 ± 0.3
Compound 5a inhibits lung cancer cell growth in vitro and in vivo. The purineꢀtype
compound 5a was selected for further efficacy testing and to explore its action mechanism in
vitro and in vivo. We first examined the cytotoxic activity of 5a in different human NSCLC cells
by MTT assay. As the results show in Figure 1A and Table 1, 5a caused NSCLC cell death in a
concentrationꢀdependent manner, with the IC50 values in the range of 0.26 to 0.33 ꢁM.
ACS Paragon Plus Environment
8

Page 9 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Compound 5a was relatively nonꢀtoxic to the normal bronchial epithelial (NBE) cells, with an
IC value of 21.66 ꢁM. To further investigate the longꢀterm effect, the colony formation assay
50
was conducted to observe whether 5a would interfere with the ability of single cells to form a
growing colony. The results indicated that a 14ꢀday treatment of 5a inhibited the colony
formation of the H1975 cells with an IC of 0.13 ꢁM (Figure 1B). These data suggested that 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
5
0
can inhibit NSCLC cell growth in vitro.
To explore the in vivo antitumor efficacy of 5a, athymic nude mice that bear established
subcutaneous H1975 tumors were treated intraperitoneal with 5a twice a week at 2 mg/kg versus
a vehicle control (n = 8 for each group). Figure 1C showed that 5a at this dosage markedly
3
reduced the H1975 tumor size to 930.4 ± 324.8 mm (P < 0.01) compared with 1699.3 ± 413.3
3
mm of the vehicleꢀtreated group on day 28. The body weights of the test mice and the
biochemical markers of the liver and kidney functions, including GOT, GPT, BUN and Cre,
were not significantly affected after a 28ꢀday treatment, which indicated that 5a was well
tolerated at the dosage of 2 mg/kg (Figure s1 in Supporting Information (SI)).
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. Compound 5a inhibits tumor growth in vitro and in vivo. (A) Inhibitory activities
of 5a against different human nonꢀsmallꢀcell lung cancer cell lines. The cells were treated with
5a for 72 h, and cell proliferation was examined by MTT assay. Data are expressed as the mean
of four determinations. (B) H1975 cells were treated with 5a (300 nM) for 14 days. The colonies
were fixed and stained with crystal violet and counted. Data are expressed as the mean ±SEM of
three determinations. (C) H1975 cells were injected subcutaneously into the right flank region of
nude mice. Three days after the injection, the mice were treated with vehicle (DMSO) or 5a (2
mg/kg) twice a week intraperitoneally for 28 days. Representative tumor images and H&Eꢀ
stained sections of tumors taken from mice treated for 28 days with either vehicle or 5a are
shown. The tumor size was measured with a caliper rule every time before compound delivery.
Data are presented as the mean tumor volumes of mice in both the treatment and vehicle groups
on the days of preꢀtreatment.
ACS Paragon Plus Environment
10

Page 11 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Compound 5a induces cell cycle arrest and apoptosis in H1975 lung cancer cells. To
confirm that the cytotoxic effect of 5a was caused by an apoptotic cell death, the 5aꢀtreated
H1975 lung cancer cells were coꢀstained with annexin V and propidium iodide (PI), and the
positive staining population was analyzed by flow cytometry. The results indicated that positive
annexin Vꢀ and PIꢀstained H1975 cells increased significantly after the treatment with 5a (Figure
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
2A). This experiment suggested that the 5aꢀtreatment could cause apoptosis in H1975 cells.
When further dissecting the potential apoptotic pathways that are activated by 5a, the expression
levels of several specific apoptotic markers were detected by immunoblotting. As expected, the
cleaved form of two important regulatory proteins in the apoptotic pathway, caspaseꢀ3 and
PARPꢀ1, appeared after compound treatment. The appearance of cleaved caspaseꢀ9 implied that
a mitochondriaꢀmediated apoptotic pathway was being activated (Figure 2B). Therefore, the
protein level of the Bclꢀ2 family proteins and the location of cytochrome C were also detected.
The protein level of Bak and Bax, two proꢀapoptotic proteins in the Bclꢀ2 family, did not change
significantly after the treatment with 5a. However, in the blotting results, we could see a band
that had a molecular weight that was higher than that of the total form of Bclꢀ2 protein, which
was detected after the compound treatment (star mark in Figure 2C). This band might be
correlated to the phosphorylation form of the Bclꢀ2 protein, which increased after compound
treatment. As is known, cytochrome C release from mitochondria to the cytosol is another
marker that is frequently found after apoptotic induction. Here, we also attempted to clarify
whether the compound treatment could cause cytochrome C release in the H1975 cells. The data
showed that a 6ꢀhour treatment of 5a induced an increase in the cytosolic level of cytochrome C
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
Figure 2D). All of these results indicated that 5a could induce cell apoptosis in H1975 lung
cancer cell lines.
In addition to the apoptotic inducing effect, we also explored whether the treatment of 5a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
could have any effect on the H1975 cell cycle progression. We found that compound treatment
also caused cell cycle arrest in H1975 lung cancer cells. The flow cytometric analysis of the cell
cycle by PI staining showed that 5a induced an increase in the cell population at the G2/ M phase
(Figure 2E).
Figure 2. Compound 5a causes cell cycle arrest and induces cell apoptosis in H1975 lung
cancer cells. H1975 cells were treated with 300ꢀnM 5a for the indicated time periods. Then, the
cells were harvested for (A) annexin VꢀPI double staining of apoptosis FACScan flow cytometer.
Data are expressed as the mean ±SEM of three determinations. (B and C) The protein
ACS Paragon Plus Environment
12

Page 13 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
expressions were detected by immunoblotting. The data are representative of three independent
experiments. (D) The cytosolic and mitochondrial fractions were obtained for the detection of the
cytochrome C protein level by immunoblotting. (E) The cells were fixed and stained with PI to
analyze the DNA content with a FACScan flow cytometer. The data are representative of three
independent experiments.
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
Compound 5a activates the JNK-related signaling pathway and induces cell apoptosis.
To characterize the molecular changes that were induced by 5a, the human phosphoꢀkinase
antibody array was used. H1975 cells were harvested after a 9ꢀhour treatment with 300 nM of 5a,
and the assay was then applied to the cell lysate. Cells treated with DMSO alone were used as
controls. The results showed that treatment with 5a increased the signal intensity of
phosphorylated JNK (JNK pT183/Y185 and pT221/Y223) and phosphorylated cꢀJun (cꢀJun
pS63) when compared with the results in the control group (Figure 3A). To further verify the
results, cells treated with 5a (300 nM) for 6 and 12hours were harvested, and the
phosphorylation levels of these two proteins were detected by immunoblotting. The results
showed that a dramatic upꢀregulation of phosphoꢀJNK and phosphoꢀcꢀJun was observed after the
compound treatment (Figure 3B).
To further determine the role of JNKꢀrelated signaling pathways in the mechanism of 5aꢀ
induced cell apoptosis, a JNK inhibitor, 1,9ꢀpyrazoloanthrone (SP600125), was adopted. As the
results demonstrated, SP600125 could inhibit the activation of the JNKꢀrelated signaling
pathways and reverse the apoptotic effects induced by 5a. The cleaved form of caspaseꢀ3 and
PARPꢀ1 decreased after combining SP600125 with the compound (Figure 3C). Additionally,
SP600125 also reduced the percentage of cells that had positive annexin V and PI staining after
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
treatment with 5a (Figure 3D). Overall, these data suggested that JNK kinase and its related
signaling pathways played important roles in the apoptotic cascade of 5a.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. JNK-related signaling pathway involved in 5a-induced cancer cell apoptosis. (A)
H1975 cells were incubated in the absence or presence of 5a (300 nM) for 9 h. The cells were
then harvested and lysed to detect the phosphorylation level of the proteins included in the
human phosphorꢀkinase antibody array. The phosphorylation level of particulate proteins was
quantified using the computerized image analysis system ImageJ. (B) H1975 cells treated with
5a (300 nM) were harvested and lysed to detect the existence of phosphoꢀJNK and phosphoꢀcꢀ
Jun. The data are representative of three independent experiments. (C) H1975 cells were
incubated in the indicated agent (5a, 300 nM; SP600125, 10 ꢁM) for 48 h. Then, the cells were
harvested and lysed to examine the expression levels of cleavage caspaseꢀ3, PARPꢀ1 and
phosphoꢀcꢀJun. The data are representative of three independent experiments. (D) H1975 cells
ACS Paragon Plus Environment
14

Page 15 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
were treated with the indicated agent (5a, 300 nM; SP600125, 10 ꢁM) for 48 h. Then, the cells
were harvested for the annexin VꢀPI double staining of the apoptosis FACScan flow cytometer.
The data are expressed as the mean ±SEM of three determinations.
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
Compound 5a induces fragmentation of microtubules and interrupts its organization in
H1975 cells. The results showed that 5aꢀtreatment would cause cell cycle arrest at the G2/M
phase and apoptosis in H1975 lung cancer cells, which led us to hypothesize that 5aꢀtreatment
could also interfere with microtubule rearrangement due to its critical role in G2/M regulation.
To test this idea, the organization of intracellular microtubules was first detected by
immunofluorescence in H1975 cells. The in situ labeling of αꢀtubulin by antibodies demonstrated
that 5a would induce microtubule fragmentation (Figure 4A). An in vitro microtubule assembly
assay was also performed to separate the assembled and disassembled microtubules into
particulate and soluble fractions by centrifugation. The data showed that a 1ꢀhour exposure of 5a
significantly reduced the amount of αꢀtubulin in the particulate fraction (the relative expression
percentage of the soluble form vs. polymer form was 68% vs. 32% at 300 nM treatment and 75%
vs. 25% at 600 nM; Figure 4B). The results from the in vitro microtubule assembly assay
confirmed the microtubuleꢀdamaging effects of 5a that were observed by the
immunofluorescence staining. To further clarify the change in the microtubule structure after
treatment with 5a, a gel filtration chromatography method was applied to roughly determine the
length of the microtubule in the H1975 cancer cells. The results showed that αꢀtubulin protein
was detected in the latter fraction in the 5aꢀtreatment group compared to the DMSO control
group. Moreover, there were fewer fractions that contained αꢀtubulin protein in the treatment
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
group compared to the control group (Figure 4C). These results indicated that 5a reduced the
length of the microtubule in H1975 lung cancer cells.
All of the above data indicated that 5a interfered with the microtubule structure of H1975
lung cancer cells. Next, the turbidity assay for tubulin polymerization and depolymerization was
adopted to examine whether 5a displayed a direct regulatory function in microtubule assembly.
As demonstrated in Figure 4D and 4E, in the presence of GTP at 37 °C, the tubulins were
triggered to polymerize in a timeꢀdependent manner. Nocodazole, vincristine and paclitaxel were
used as reference agents to compare with 5a. Surprisingly, according to the results, we found that
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5a did not influence the dynamics of microtubule polymerization and depolymerization (Figure
4D and 4E).
ACS Paragon Plus Environment
16

Page 17 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 4. Treatment of 5a interrupts microtubule organization in H1975 lung cancer cells.
(A) H1975 cells were incubated with vehicle or 5a (300 nM) for the indicated times. Then, the
cells were fixed and stained with antiꢀαꢀtubulin primary antibody. FITCꢀlabelled secondary
antibodies were used (green fluorescence), and the protein was detected by a confocal laser
microscopic system. The nuclei were detected by DAPI staining (blue fluorescence). (B) H1975
cells were incubated with vehicle, 5a, paclitaxel or nocodazole for 1 h. Then, the cells were
harvested and separated into soluble (S, tubulin monomer) and particulate form (P, tubulin
polymers), and the level of αꢀtubulin protein in each fraction was detected by immunoblotting.
The protein level of αꢀtubulin was quantified using the computerized image analysis system
ImageJ. (C) H1975 cells were incubated in the absence or presence of 5a (300 nM) for 12 h.
Then, the cells were harvested and lysed for gel filtration chromatography. The dropꢀbyꢀdrop
flowthrough was collected in ~0.5ꢀmL fractions. The protein level of αꢀtubulin in each fraction
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
was detected by immunoblotting. (D) Purified tubulins were incubated at 37 C with GTP in the
vehicle (control) or in the presence of 5a (10 ꢁM or 300 nM), paclitaxel (10 ꢁM), vincristine (10
ꢁ
M) or nocodazole (10 ꢁM). Tubulin polymerization was examined turbidimetrically. Data are
expressed as the mean of three determinations. (E) Purified tubulins were incubated with GTP
o
and paclitaxel at 37 C for 30 min and then were held at room temperature for at least 1 h to
o
obtain microtubule stock. The microtubule stock was further incubated at 37 C with GTP in the
vehicle (control) or in the presence of 5a (10 ꢁM, 3 ꢁM or 300 nM), paclitaxel (30 ꢁM) or
vincristine (30 ꢁM). Tubulin depolymerization was examined turbidimetrically. The data are
expressed as the mean of three determinations.
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Using a biotin-annexed probe to identify katanin as a direct target protein. To identify
the target proteins of 5a, its analogs carrying azido or diazirinyl group in the C2ꢀaryl moiety
were first designed as photoaffinity probes. Two analogs were synthesized by the coupling
reactions of 3a with 4ꢀazidobenzaldehyde or 3ꢀ(4ꢀ(formylphenyl)ꢀ3ꢀ(trifluoromethyl)ꢀ3Hꢀ
diazirine. However, the utilization of these two compounds as photoaffinity probes was
hampered due to their instability, poor solubility and low inhibitory activity. Alternatively, we
consider transforming the active compound 5e (IC = 0.19 ꢁM) into the desired probe using
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
50
biotin as the reporter unit. As shown in Scheme 2, ester 5e was subjected to saponification to
give acid 6e, which underwent the amideꢀforming reaction with amine 7 in the presence of
N,N,N′,N′ꢀtetramethylꢀOꢀ(1Hꢀbenzotriazolꢀ1ꢀyl)uronium hexafluorophosphate (HBTU, 1.0–1.2
equiv) and N,Nꢀdiisopropylethylamine (DIEA, 1.0–1.2 equiv), to afford the corresponding amide
product 8. The Boc protecting group in 8 was readily removed by treatment with trifluoroacetic
acid (TFA) at room temperature, and the exposed amino group was successfully linked to Dꢀ(+)ꢀ
biotin (1.0 equiv) via the amide bondꢀforming reactions, producing 9.
The biotinꢀannexed 9 was used as a probe to catch the target proteins. After incubating with
H1975 cell lysate, the proteins that were bound to 9 were pulled down by streptavidinꢀ
conjugated beads. Katanin p60 and p80 were visualized by immunoblotting (Figure 5A). As a
control, the biotin linker that lacked the purine moiety could not trap katanin p60 or p80.
ACS Paragon Plus Environment
18

Page 19 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 2. Synthesis of amides 8 and the biotinꢀannexed probes 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
Molecular computation of pharmacophore models and p60–5a complex models. To
correlate the chemical structures of the purineꢀtype compounds to their NSCLC inhibition
activities in a quantitative manner, we selected the compounds with certain IC values (Table 2)
5
0
to generate the 3DꢀQSAR pharmacophore models. The results for the top ten pharmacophore
hypotheses and their statistical parameters are shown in Table 3. In this study, the first
pharmacophore hypothesis (Hypoꢀ1) is the best because this model has the lowest total cost
value (45.488), the largest cost difference (54.763), the lowest rootꢀmeanꢀsquare deviation
(RMSD) value (0.699), and the highest correlation coefficient (0.984). Hypoꢀ1 includes one
hydrogenꢀbond acceptor (HBA), one hydrogenꢀbond donor (HBD), one hydrophobic feature
(HY), and one ring aromatic feature (RA). The 3D spatial arrangement and distance constraints
of these features are represented in Figure 5B. The features of Hypoꢀ1 mapped onto 5a are
shown in Figure 5C. HBA is mapped onto the 8ꢀoxo group, HBD is mapped onto the amino
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
moiety of the 6ꢀcarboxamide group, RA is mapped onto the C2ꢀaryl group, and HY is mapped
onto the tip of the ethoxy substituent in the C2ꢀaryl group.
To demonstrate the predictive power of Hypoꢀ1, regression analysis was used to estimate the
activity of each trainingꢀset compound. As shown in Table 4, all of the trainingꢀset compounds
showed an error value of less than 2. An error value below 10 indicated that only one order of
magnitude of difference was observed in the discrepancy between the estimated and actual
activity. The most active 5e (IC = 0.19 µM) showed a fitness score value of 7.46 when mapped
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
to Hypoꢀ1, whereas the least active 6e (IC = 7.91 µM) showed a fitness value of 6.04. This
5
0
result indicates that Hypoꢀ1 is a reliable model that consists of reasonable chemical features of
the trainingꢀset compounds and accurately estimates the experimental activity of the trainingꢀset
compounds. The calculated IC values through this 3DꢀQSAR pharmacophore model supports
5
0
that 5a and its analogue 5e have the most potent inhibitory activity. Therefore, small molecules
that completely satisfy the chemical feature arrangement of Hypoꢀ1 could be good candidates for
NSCLC inhibition. However, the high structural homology among the compounds used in this
study combined with their small activity range made this 3DꢀQSAR pharmacophore model
slightly imperfect. For example, the smaller difference (54.763) between the total and null cost
values suggests that Hypoꢀ1 might not have more than 90% statistical significance as a model.
This difference value should be greater than or equal to 60, according to the Discovery Studio
(DS) manual.
Although we found that katanin p60 and p80 can be pulled down by 9, an analogue of 5a
(Figure 5A), we do not know whether 5a binds to p60 or p80 or the interface between p60 and
p80. Because 5a shares a similar purine scaffold with ATP, we hypothesized that 5a could bind
to the catalytic AAA domain of p60. To assess this possibility, we modeled the homoꢀhexameric
ACS Paragon Plus Environment
20

Page 21 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
p60 structure using the crystal structure of transitional endoplasmic reticulum ATPase p97 from
2
3
mouse (PDB ID: 3CF1) as the template. The sequence alignment indicated that human p60
2
4
shares 30% sequence identity with mouse p97, but the protein structure prediction server
Phyre2 suggested that the confidence score for this model is 100%. The docking software
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
2
5
AutoDock Vina successfully docked 5a into the ATPꢀbinding pocket of p60, and the
theoretically calculated binding affinity is quite high (ꢂG = –10.7 kcal/mol). Two plausible
binding modes of 5a in p60 hexamer are shown in Figures 5D and 5E. The predicted binding
modes suggested that the backbone nitrogen of Thr256 could bind to the HBA feature of Hypoꢀ1.
The interaction to the HBD feature could require the backbone oxygen of Ala358 and the
sidechain of Asp308 or Asn360. We also noticed that residues from two subunits of a p60
hexamer contribute to the hydrophobic contacts with 5a. More specifically, the proposed
arginine finger of p60, Arg372, is also involved in the binding to 5a. Early study demonstrated
that ATP induces p60 oligomerization and that the ATPꢀbound p60 hexamer has a high
2
6
microtubule severing activity . Based on the docking model presented in this study, we
hypothesize that 5a could stabilize the p60 hexamer structure and enhance the severing activity
of katanin. However, this hypothesis requires further experimental investigation.
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Compound 5a direct interacts with katanin protein as a p60–5a complex. (A)
H1975 cell lysate was incubated with probe 9 or a linker without the purine component (as
control) for 24 h. Then, the streptavidinꢀconjugated beads were added into the lysate and
incubated for 2 h. The lysate was washed with IP lysis buffer 6 times. The binding proteins were
visualized by immunoblotting. The data are representative of three independent experiments.
H1975 cells were transfected with scrambled control siRNA or katanin p60 siRNA. (B to E)
Pharmacophore model and predicted binding modes between katanin p60 hexamer and 5a. (B)
Best pharmacophore model Hypoꢀ1 represented with distance constraints. (C) Hypoꢀ1 mapping
with 5a. Pharmacophore features are colorꢀcoded as follows: green, hydrogen bond acceptor
(
HBA); magenta, hydrogen bond donor (HBD); cyan, hydrophobic (HY); orange, ring aromatic
RA). (D and E) show two binding modes with equal predicted binding affinities (ꢂG = –10.7
(
ACS Paragon Plus Environment
22

Page 23 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
kcal/mol). The green background highlights the amino acids from subunit A of the p60 hexamer,
and the pink background highlights those that belong to subunit F.
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
Table 3. Statistical significance, predictive power and features of the top ten hypotheses derived
from 3DꢀQSAR pharmacophore generation.
d
Features
a
b
c
Max.
Fit
Hypothesis
Correlation
RMSD
Total cost
ꢂCost
HBA HBD
HY
1
RA
1
Hypoꢀ1
Hypoꢀ2
Hypoꢀ3
Hypoꢀ4
Hypoꢀ5
Hypoꢀ6
Hypoꢀ7
Hypoꢀ8
Hypoꢀ9
Hypoꢀ10
0.984
0.970
0.958
0.944
0.930
0.916
0.906
0.890
0.822
0.822
0.699
0.958
1.121
1.290
1.437
1.567
1.657
1.784
2.232
2.228
45.488
47.619
49.706
52.029
54.039
55.426
57.579
59.697
68.011
68.346
54.763
52.632
50.545
48.222
46.212
44.825
42.672
40.554
32.240
31.905
1
1
1
1
1
2
2
2
2
0
1
1
1
1
1
0
1
1
0
0
8.465
7.572
6.849
6.534
6.530
7.272
6.402
6.461
8.615
8.402
1
1
2
0
1
1
2
0
1
1
1
0
1
0
1
1
3
2
The null cost, fixed cost and configuration cost are 100.251, 42.969 and 18.175, respectively.
a
Correlation coefficient is calculated based on linear regression derived from the geometric fit index.
b
Root mean square difference of log(estimated activity) from log(measured activity) normalized to
log(uncertainty).
c
ꢂCost= null cost – total cost of the hypothesis.
d
HBA, hydrogenꢀbond acceptor; HBD, hydrogenꢀbond donor; HY, hydrophobic; RA, aromatic ring.
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 4. Actual and estimated IC values of 10 purineꢀtype compounds calculated based on the
pharmacophore Hypoꢀ1.
5
0
b
a
Mapping
c
d
e
Name
Fit
Act. (µM)
Est. (µM)
Error
HBA HBD HY RA
[13 11 28 15]
[13 11 28 15]
[13 11 30 15]
[30 * 28 15]
[13 11 * 15]
[13 11 * 15]
5e
5a
5c
7.46
7.37
6.29
5.93
5.92
6.04
0.19
0.30
4.66
4.73
5.80
7.91
0.24
0.29
3.6
+1.3
ꢀ1.0
ꢀ1.3
+1.7
+1.4
ꢀ1.2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5f
8.2
5
b
8.4
6
e
6.4
a
Fit value is calculated by fitting the chemical features between the hypothesis and the compound; the higher the
value is, the better the fit.
b
Matching degree of the pharmacophore element with the specific conformation of the compound.
c
Actual IC value.
5
0
d
e
Estimated IC value.
50
Error factor is calculated as the ratio of the actual IC50 to the estimated IC . A positive value indicates that the
50
estimated IC50 is higher than the actual IC ; a negative value indicates that the estimated IC is lower than the
50
50
actual IC .
5
0
Compound 5a induces cell apoptosis through interacting with the katanin protein
complex. Taking together the results of the probe study and molecular modeling, one can deduce
that katanin could be the direct target of 5a. The microtubule fragmentation phenomenon
induced by 5a was similar to the phenotype observed in the cells that overexpress the katanin
2
7
protein . To further establish the role of katanin p60 in the anticancer mechanism of 5a, a
katanin p60 knockꢀdown H1975 lung cancer cell line was created by siRNA transfection
technology. When treating the katanin p60 silencing cells with 5a, it would not reduce the
amount of αꢀtubulin in the particulate fraction. The relative expression percentage of the soluble
form vs. polymer form was 68% vs. 32% in the siControl group and 34% vs. 66% in the
siKatanin group (Figure 6A). Moreover, the microtubule structure was also maintained in a
ACS Paragon Plus Environment
24

Page 25 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
relatively undamaged status after compound treatment in the siKatanin group compared to the
siControl group (Figure 6B). These data all indicated that katanin proteins participated in the
mechanism of 5aꢀinduced microtubule damage in H1975 lung cancer cells.
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 addition to the effect of microtubule damage, the apoptotic effects of 5a could also be
reversed by the knockꢀdown katanin p60 protein. First, the phosphorylation levels of JNK and cꢀ
Jun were examined by immunoblotting. According to our expectation, the expressions of
phosphoꢀJNK and phosphoꢀcꢀJun were increased by treatment with 5a and would reduce
significantly in katanin p60 knockꢀdown cells (Figure 6C). Then, the amount of cleaved forms of
caspaseꢀ3 and PARPꢀ1 induced by treatment with 5a was also explored in katanin silencing cells.
Similar to the findings of phosphoꢀJNK and phosphoꢀcꢀJun, increasing amounts of cleaved
caspaseꢀ3 and PARPꢀ1 after treatment with 5a were decreased in the knockꢀdown cells (Figure
6D). Finally, the overall cytotoxic effect of 5a was measured by MTT assay in both the siControl
and siKatanin cell groups. The IC value of this compound increased from 0.25 ꢁM to 0.95 ꢁM
50
(Figure 6E). Taken together, these data indicated that katanin p60 protein is an important protein
for regulating 5aꢀmediated microtubule damage and apoptosis in H1975 lung cancer cells.
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Compound 5a induces cell apoptosis through interacting with the katanin protein
complex. (A) The cells were treated with 5a for 1 h. Then, the cells were harvested and separated
into soluble (S, tubulin monomer) and particulate (P, tubulin polymers) forms, and the level of αꢀ
tubulin protein in each fraction was detected by immunoblotting. The protein level of αꢀtubulin
was quantified using the computerized image analysis system Image J. (B) H1975 cells were
incubated with vehicle or 5a for 12 h. Then, the cells were fixed and stained with primary
antibody to αꢀtubulin. FITCꢀlabelled secondary antibodies were used (green fluorescence), and
the protein was detected by a confocal laser microscopic system. The nuclei were detected by
DAPI staining (blue fluorescence). (C and D) H1975 cells were incubated with vehicle or 5a for
24 h. Then, the cells were harvested for the identification of indicated proteins by
immunoblotting. (E) The cell proliferation activity was examined by MTT assay.
ACS Paragon Plus Environment
26

Page 27 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
DISCUSSION
Microtubules play a crucial role in several cellular processes, especially in the formation of
the mitotic spindle during cell mitosis. The disruption of the microtubule dynamics will lead to
cell cycle arrest in the G2/M phase; therefore, interference with the microtubule structure has
become an important mechanism in antiꢀcancer therapy. Recently, many MTAs have been
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
2
8
developed, and some of them have been approved to enter into clinical trials . In the present
study, we demonstrated that the purineꢀtype compound 5a was a novel MTA that caused
microtubule fragmentation by directly targeting the microtubule severing protein, katanin, to
induce apoptosis of the cancer cells. Compound 5a had good in vitro and in vivo activities
against NSCLC, including the gefitinibꢀresistant types. Compound 5a also shows a relatively
wide therapeutic window between proliferating and quiescent cells (over 10ꢀfold), which
indicates that the cellꢀgrowth inhibiting effects resulted from inhibiting proliferationꢀdependent
processes rather than indiscriminant cytotoxicity.
By interfering with the microtubule structure, most of the MTAs will cause cancer cell
mitotic catastrophe and induce a G2/M cell cycle arrest. The cell cycle arrest allows cells to
repair the damage caused by MTAs. However, if the damage cannot be fixed, then the prolonged
cell cycle arrest will, in the end, become a stress and trigger a mitochondriaꢀmediated intrinsic
10, 29, 30
apoptosis
. Here, we found that 5a, similar to other MTAs, can induce a sustained G2/M
cell cycle arrest and activate intrinsic apoptosis. The release of cytochrome C from mitochondria
to the cytosol is a marker for mitochondriaꢀmediated apoptosis, which is controlled by the
balance between two subfamilies of Bclꢀ2 family proteins, the antiꢀapoptotic and proꢀapoptotic
3
1
subfamilies . Among them, the Bclꢀ2 protein, as an antiꢀapoptotic protein, plays an important
role in preventing cell apoptosis by binding and neutralizing the activity of proꢀapoptotic
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
subfamily members, such as BAX and BAK. This protecting function of Bclꢀ2 protein depends
not only on their amount in the cytosol but also on their postꢀtranscriptional modification status.
Several studies have reported that Cdk1, cꢀRaf, PKA, JNK1, and p38MAPK kinase can cause the
phosphorylation of the Bclꢀ2 protein; this phosphorylation will then cause a conformational
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
2, 33
change in the Bclꢀ2 protein to inactivate the antiꢀapoptotic function
. In our study, after 6ꢀ
hour of 5aꢀtreatment, a band with higher molecular weight than unmodified Bclꢀ2 protein was
detected. This band might represent the level of phosphorylated Bclꢀ2, which implies an
inhibitory effect of its function. Because 5aꢀtreatment would induce a G2/M cell cycle arrest and
activation of JNK kinase, we suspected that the phosphorylation of Bclꢀ2 protein might be
caused by the G2/M phaseꢀspecific Cdk1 kinase or the JNK kinase.
JNK kinase can phosphorylate different substrates, which include cꢀJun, ATF2, ELK1,
SMAD4, p53 and HSF1, and it can be involved in different cellular processes, such as apoptosis,
3
4
neurodegeneration, cell differentiation and proliferation . In the apoptotic pathways, JNK kinase
plays opposing roles depending on the upstream stimuli and downstream effectors in the
35
different cells . For example, Ma and his colleagues found that the JNK/cꢀJun pathway
stimulates cell proliferation and protects cells from apoptosis in a pulmonary artery endothelial
3
6
cell model . However, in other cell types, such as human lung cancer cells or breast cancer cells,
37, 38
the activation of the JNK/cꢀJun pathway will induce caspaseꢀmediated apoptosis
. Our studies
showed that the JNK/cꢀJun pathway could induce cell apoptosis in H1975 human lung cancer
cells. The apoptotic effect of 5a could be alleviated significantly by combining the JNK inhibitor,
SP600125. In fact, several studies also reported that different MTAs could activate the JNK/cꢀ
39, 40
Jun pathway to induce an intrinsic apoptosis
.
ACS Paragon Plus Environment
28

Page 29 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Although the effects of 5a to induce G2/M cell cycle arrest and cell death are similar to
those caused by other MTAs, such as taxanes and vinca alkaloids, the target protein and detailed
mechanisms of action are different between them. Most of the MTAs directly bind to the tubulin
and disrupt the dynamics of microtubule polymerization or depolymerization. The structure of
the microtubule will be sustained as a filament when treating the cell with microtubuleꢀ
stabilizing agents and will disrupt into vague dots when treating with microtubuleꢀdestabilizing
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
4
1, 42
agents
. However, after 5aꢀtreatment, the microtubules of the cells were cut into fragments
instead of dots under microscopic observation. When we further analyzed the dynamics of the
microtubules, we found that the rates of microtubule polymerization and depolymerization did
not change. To our knowledge, this fragmentation phenotype of microtubule is seldom induced
by other MTAs and could imply that a unique microtubule targeting mechanism is used by the
purineꢀtype compound 5a.
Proteins associated with microtubules are responsible for the regulation of microtubule
1
3
structure . Microtubule severing proteins are one class of proteins that interact with
microtubules and regulate microtubule structure by a severing process. Through the severing
process, the microtubule will be cut into short fragments and transported to different regions.
This arrangement provides an efficient way of promoting microtubule disassembly and
1
6, 43
movement
. Katanin, one of the microtubule severing proteins, has been proven to be an
1
7, 18
important microtubule structure modulator in many cellular processes
. The overexpression
44, 45
of katanin will cause microtubule fragmentation
. Because the microtubule structure will
undergo intense reorganization in the G2/M phase to help cell division, katanin is also involved
4
6
in cell dividing processes . McNally and his colleagues proved that katanin p60 will concentrate
47
at the ends of the microtubule and control their lengths in mitotic cells . Maddika and his
ACS Paragon Plus Environment
29