Total synthesis of AMF-26, an antitumor agent for inhibition of the golgi system, targeting adp-ribosylation factor 1

Article abstract of DOI:10.1021/jm301695c

An effective method for the total synthesis of 1 (AMF-26), a potentially promising new anticancer drug that disrupts the Golgi system by inhibiting the ADP-ribosylation factor 1 (Arf1) activation, has been developed for the first time. The construction of

Full text of DOI:10.1021/jm301695c

Subscriber access provided by Brown University Library
Article
Total Synthesis and Anticancer Activity of AMF-26
Targeting ADP-ribosylation Factor 1 (Arf1)
Isamu Shiina, Yuma Umezaki, Yoshimi Ohashi, Yuta Yamazaki, Shingo Dan, and Takao Yamori
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm301695c • Publication Date (Web): 06 Dec 2012
Downloaded from http://pubs.acs.org on December 11, 2012
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 39
Journal of Medicinal Chemistry
1
2
1
3
4
5
6
7
8
9
Total Synthesis of AMF-26, an Anti-tumor Agent for Inhibition of
the Golgi System, Targeting ADP-ribosylation Factor 1 (Arf1)
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
Isamu Shiina,*,† Yuma Umezaki,† Yoshimi Ohashi,‡ Yuta Yamazaki,† Shingo Dan,‡
and Takao Yamori‡
†Department of Applied Chemistry, Faculty of Science, Tokyo University of Science,
1-3 Kagurazaka, Shinjuku-ku, Tokyo 162-8601, Japan
‡Division of Molecular Pharmacology, Cancer Chemotherapy Center, Japanese
Foundation for Cancer Research,
3-8-31 Ariake, Koto-ku, Tokyo 135-8550, Japan
*To whom correspondence should be addressed. For I.S.: phone,
+81-3-3260-5609; fax, +81-3-3260-5609; e-mail, shiina@rs.kagu.tus.ac.jp
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 39
1
2
2
3
4
5
6
7
ABSTRACT:
8
9
An effective method for the total synthesis of 1 (AMF-26), a potentially promising new
anti-cancer drug that disrupts the Golgi system by inhibiting the Arf1 activation, has
been developed for the first time. The construction of the chiral linear precursor (a key
to the synthesis) was achieved by the asymmetric aldol reaction followed by the
computer-assisted predictive stereoselective intramolecular Diels–Alder (IMDA)
reaction. The global anti-tumor activity of the totally synthetic 1 against a variety of
human cancer cells was assessed using a panel of 39 human cancer cell lines (JFCR39),
and it was shown that the synthetic 1 strongly inhibited the growth of several cancer cell
lines at concentrations of less than 0.04 µM. Biological assays of novel derivatives, 26
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
1,2
and 31, which have different side-chains at the C-4 positions in the Δ -octalin
backbone, disclosed the importance of the suitable structure of the side-chain containing
conjugated multi-double bonds.
———————————————————————————————————
Figure 1. Structures of AMF-26 (1) and AMF-14 (2).
ACS Paragon Plus Environment

Page 3 of 39
Journal of Medicinal Chemistry
1
2
3
3
4
5
6
7
■ INTRODUCTION
8
9
(2E,4E)-5-((3S,4S,4aS,5S,6S,7R,8aR)-3,4,4a,5,6,7,8,8a-Octahydro-6-hydroxy-3,5,7-trim
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
ethylnaphthalen-4-yl)-2-methyl-N-((pyridin-3-yl)methyl)penta-2,4-dienamide
1
(AMF-26)
(1)
has
been
semi-synthesized
from
(2E,4E)-5-((3S,4S,4aS,5S,6S,7R,8aR)-3,4,4a,5,6,7,8,8a-octahydro-6-hydroxy-3,5,7-trim
2
ethylnaphthalen-4-yl)-2-methylpenta-2,4-dienoic acid (AMF-14) (2), a natural product
isolated from the genus Trichoderma NFS-932 by Nippon Shinyaku Co., Ltd., in 2005
(Figure 1). The compound 1 is a potentially promising new anti-cancer drug lead
because its oral administration to female nude mice induced complete regression of
2
human breast cancer BSY-1 xenografts in vivo. It was revealed that 1 induced Golgi
disruption, apoptosis, and cancer-cell growth inhibition through a mechanism for
preventing the ADP-ribosylation factor 1 (Arf1) activation introduced by Ohashi and
2
Yamori et al.
Arf1 has an important role in mediating vesicular transport. Recently,
protein–protein interaction, such as the activation of Arf1 by its guanine nucleotide
exchange factor (ArfGEF), has become a very attractive new target for providing
3
effective therapeutic agents, including novel anti-cancer drugs.
From computer
modeling, it is postulated that 1 binds to the contact surface of the Arf1-Sec7 domain
where brefeldin A bound. Brefeldin A was first introduced as the inhibitor of the
4–6
Arf1-ArfGEF interaction;
however, brefeldin A has not progressed beyond the
6a,7
pre-clinical stage of drug development because of its poor bioavailability.
Therefore, 1, a new Golgi inhibitor targeting Arf1 activation, can be one of the first
candidates for a new type of medicine for cancer treatment in clinical use.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 39
1
2
4
3
4
5
6
7
8
9
Furthermore, another biological activity of 1, angiogenesis inhibition of the VEGF
receptor phosphorylation and NF-κB signaling, has been investigated by Watari and
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
8
Nakamura et al.
In this study, we report the first total synthesis of 1 using the asymmetric aldol
reaction for constructing the chiral linear precursor and computer-assisted predictive
stereoselective intramolecular Diels–Alder (IMDA) reaction to provide the core
structure of the multi-substituted octahydronaphthalene.
■ TOTAL SYNTHESIS OF AMF-26
Retrosynthetic Analysis of 1 via IMDA Reaction. Scheme 1 outlines our
designed asymmetric synthesis of 1, including the three following notable
9
methodologies: i) the diastereoselective Evans’ alkylation of the chiral oxazolidinone
16 with (E,E)-1-bromohexa-2,4-diene (17) to form a chiral diene 15, ii) the
10
enantioselective Mukaiyama aldol reaction of an aldehyde 12 with ketene silyl acetal
(KSA) 13 to form an α,β,γ-trisubstituted chiral thioester 11, and iii) the
11
diastereoselective IMDA reaction
of the chiral linear ester 8 to form
octahydronaphthalene 7a. Other transformations, such as the Wittig reaction of an
aldehyde with an ylide 10 to provide an ester 8, the HWE
9
(Horner–Wadsworth–Emmons) reaction of an aldehyde 5 with phosphate 6 to provide a
12
carboxylic acid 3, and the MNBA coupling reaction of 3 with 3-pyridylmethylamine
(4) to afford the corresponding amide, might be successfully carried out for the
preparation of the basic skeleton of the target molecule 1 in this plan.
ACS Paragon Plus Environment

Page 5 of 39
Journal of Medicinal Chemistry
1
2
5
3
4
5
6
7
Scheme 1. Retrosynthetic Analysis of 1
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

Journal of Medicinal Chemistry
Page 6 of 39
1
2
6
3
4
5
6
7
Asymmetric Synthesis of Chiral Linear Ester 8 and Prediction of the
8
9
1,2
Stereoselectivity in the IMDA Reaction of 8M to Form Δ -Octalins 7Ma–d.
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
9
First, Evans’ chiral oxazolidinone 16 was converted into the sodium enolate by the
13
reaction with NaHMDS, followed by the treatment with 17 in the presence of TBAI
to diastereoselectively form a diene 15 (Scheme 2). Removal of the chiral auxiliary in
15 by reduction with lithium aluminum hydride, and the successive oxidation of the
primary alcohol 18 using TPAP afforded the (E,E)-dienealdehyde 12 in good yield.
10
The enantioselective Mukaiyama aldol reaction of 12 with KSA 13 in the presence of
the chiral Sn(II)-complex 14 smoothly took place to exclusively provide the desired
α,β,γ-trisubstituted thioester 11. Protection of the free hydroxyl group in 11 and
reduction of the resulting TBS ether 19 with DIBAL were carried out to afford the
chiral aldehyde 9, and then the sequential treatment of 9 with Wittig reagent 10 afforded
the α,β-unsaturated ester 8 as the key synthetic intermediate.
Before attempting the formation of the main skeleton of 1 from 8 by the IMDA reaction,
we investigated the computer-aided prediction of the preferable stereoselectivity
14,15
through theoretical calculations of the stability of the transition structures
using a
simplified model compound 8M, in which the TBSO group of 8 was replaced with a
MeO group (Scheme 3).
DFT calculations were performed to determine the transition structures that form the
fused bicyclic backbone of the multi-substituted 3,4,4a,5,6,7,8,8a-octahydronaphthalene
ACS Paragon Plus Environment

Page 7 of 39
Journal of Medicinal Chemistry
1
2
7
3
4
5
6
7
8
9
(Δ1,2-octalin). The geometries of all the stationary points were fully optimized at the
B3LYP/6-31G* level of theory, and the properties of the molecules were also calculated
at the same or advanced level using 6-31G**, 6-311G*, and 6-311G** basis functions.
All points were characterized as minima or saddle points by calculation of the harmonic
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
16
vibrational frequencies using analytical second derivatives.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 39
1
2
8
3
4
5
6
7
a
Scheme 2. Preparation of the Key Chiral Linear Ester 8 Intermediate
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Reagents and conditions: (a) NaHMDS, THF, –78 °C, 15 min, and then bromide 17, TBAI, rt,
1 h, 87%; (b) LiAlH , THF, 0 °C to rt, 12 h, 81%; (c) TPAP, NMO, 4 Å molecular sieves,
4
n
CH Cl , 0 °C, 30 min; (d) KSA 13, (R)-Sn(II)-complex 14, Bu Sn(OAc) , CH Cl , –78 °C, 3
2
2
2
2
2
2
h, 62% from 18; (e) TBSOTf, 2,6-lutidine, CH Cl , 0 °C, 1 h, 98%; (f) DIBAL, CH Cl ,
2
2
2
2
–78 °C, 5 min; (g) Ph P=CHCO Et (10), toluene, 110 °C, 3 h, 83% from 19.
3
2
ACS Paragon Plus Environment

Page 9 of 39
Journal of Medicinal Chemistry
1
2
9
3
4
5
6
7
8
9
Scheme 3. Predicted Preference of Diastereoselectivity in the IMDA Reaction of
the Model Compound 8M Based on DFT Calculations
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
Four transition states provide the [4 + 2] cycloadducts 7Ma–d, and these transition
structures ([ts-a1]–[ts-d1]) are categorized as endo- and exo-modes (endo, [ts-a1] and
[ts-d1]; exo, [ts-b1] and [ts-c1]) as depicted in Scheme 4 and Figure 2. The desired
(9R,10R)-Δ1,2-octalin derivative 7Ma could be generated via the transition state [ts-a1]
by the IMDA reaction of the model triene 8M. Another endo-mode transition state
[ts-d1] to form the diastereoisomeric (9S,10S)-Δ1,2-octalin derivative 7Md is an
unfavorable structure for the preparation of 1. Similarly, the exo-mode transition
states, [ts-b1] and [ts-c1], could be converted into the undesired (9S,10R)- and
(9R,10S)-Δ1,2-octalin derivatives 7Mb and 7Mc through the imaginary IMDA reactions,
respectively. The calculated relative Gibbs free energy (ΔGrel) is shown under each
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 39
1
2
10
3
4
5
6
7
8
9
transition structure in Scheme 4 and Table 1. It was revealed that [ts-c1] and [ts-d1]
have significantly high ΔGrel values compared with those of [ts-a1] and [ts-b1] because
of the steric repulsion between the methyl group at the C-4 position in 8M and the
surroundings in these transition structures. Therefore, the corresponding (9R,10S)- and
(9S,10S)-Δ1,2-octalin derivatives 7Mc and 7Md will hardly form via these two highly
strained transition states (ΔGrel = 5.21–5.08 kcal/mol for [ts-c1]; ΔGrel = 4.65–4.28
kcal/mol for [ts-d1]). Within the stable [ts-a1] and [ts-b1] structures having lower
values of ΔGrel, the formation of the latter transition structure is preferable. The
calculated free energy difference (ΔGrel = 0.49–0.19 kcal/mol at 298.15 K (25 °C);
0.40–0.14 kcal/mol at 383.15 K (110 °C)) corresponds to the medium diastereomeric
ratio of the formed [4 + 2] cycloadducts according to the Boltzmann distribution
(estimated selectivity of 7Ma:7Mb = ~30:70–42:58 at 298.15 K (25 °C); ~37:63–45:55
at 383.15 K (110 °C)).
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 11 of 39
Journal of Medicinal Chemistry
1
2
11
3
4
5
6
7
8
9
Scheme 4. Transition Structures of the Thermal IMDA Reaction of 8M
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 12 of 39
1
2
12
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
Figure 2. Three-dimensional transition structures, [ts-a1] (endo), [ts-b1] (exo),
[ts-c1] (exo), and [ts-d1] (endo), in the thermal IMDA reaction of 8M.
Table 1. Calculated Relative Gibbs Free Energies (ΔG ) (in kcal/mol) of the
rel
Transition Structures in the Thermal IMDA Reaction of 8M
B3LYP/6-31G* B3LYP/6-31G** B3LYP/6-311G* B3LYP/6-311G** Predict (110 °C)
[ts-a1]
[ts-b1]
[ts-c1]
[ts-d1]
0.49 (0.40)a
0.00 (0.00)a
5.21 (5.12)a
4.65 (4.64)a
0.50
0.00
5.19
4.62
0.19 (0.14)a
0.00 (0.00)a
5.26 (5.23)a
4.39 (4.48)a
0.33
0.00
5.08
4.28
37
··
63
·· ! ··
0
··
0
45
··
55
0
··
0
a The values in parentheses are !Grel!s at 383.15 K (110 °C).
ACS Paragon Plus Environment

Page 13 of 39
Journal of Medicinal Chemistry
1
2
13
3
4
5
6
7
1,2
Thermal IMDA Reaction of 8 to Form Δ -Octalins 20a and 20b, and
8
Et AlCl-Mediated Diastereoselective IMDA Reaction of 8 to Form
2
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
1,2
Δ
-Octalin 20a. After determining the transition structures and understanding the
nature of the predictable stereoselectivity of the IMDA reaction of 8M, we eventually
examined the thermal transformation of 8, which has a TBSO group at the C-5 position
instead of a MeO group at the same position in 8M. By heating 8 in toluene at reflux
temperature for three days (Scheme 5, upper equation), the desired thermal [4 + 2]
cycloaddition reaction successfully proceeded, and we obtained a mixture of two
stereoisomeric products, 7a and 7b, corresponding to the model Δ1,2-octalins 7Ma and
7Mb, respectively. A mixture of the cycloadducts, 7a and 7b, was then treated with
DIBAL to provide the desired primary alcohols, 20a and 20b, with a low
diastereoselectivity (20a:20b = ~44:56) in this two-step transformation. These
experimental results are in good agreement with the assumed stereoselectivity as
follows: (1) Two products, 7a and 7b, possessing (R) configurations at the C-10
position were predominantly produced, and two other Δ1,2-octalins possessing (S)
configurations at the same position were not obtained at all. (2) Both the major
stereoisomer, (9S,10R)-7b, generated from the exo-mode transition state and the minor
one, (9R,10R)-7a, generated from the endo-mode transition state were simultaneously
produced because the predicted preference of the exo-mode transition state over the
endo-mode transition state is not very high.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 39
1
2
14
3
4
5
6
7
Scheme 5. Thermal or Et AlCl-Mediated IMDA Reaction of 8 to Form endo-Type
2
8
9
1,2
1,2
a
(9R,10R)-Δ -Octalin 20a via 7a and exo-Type (9S,10R)-Δ -Octalin 20b via 7b
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Reagents and conditions: (a) toluene, reflux, 3 d; (b) DIBAL, CH Cl , 0 °C, 30 min, 38% of
2
2
20a and 48% of 20b from 8; (c) Et AlCl, CH Cl , –78 °C to rt, 16 h; (d) DIBAL, CH Cl , 0 °C,
2
2
2
2
2
20 min, 62% of 20a from 8.
To improve the yield of the endo-type (9R,10R)-Δ1,2-octalin derivative 7a, a favorable
intermediate for the synthesis of 1, it is necessary to develop an alternative activation
method for the IMDA reaction that proceeds via the endo-mode transition state [ts-a1]
but not via the exo-mode transition state [ts-b1]. There are many successful examples
of obtaining endo-type octalin derivatives via the diastereoselective Diels–Alder
17
reaction using a Lewis acid in steroid syntheses, therefore, we decided to use the
Et2AlCl-mediated Diels–Alder reaction to control the diastereoselectivity of the IMDA
reaction of 8.
ACS Paragon Plus Environment

Page 15 of 39
Journal of Medicinal Chemistry
1
2
15
3
4
5
6
7
8
9
Determination of the structures of the endo- and exo-mode transition states with
Me2AlCl and the relative Gibbs free energies was next carried out by theoretical
calculations according to the above mentioned method for predicting the
diastereoselectivity that uses the model substrate 8M. The optimized endo-mode
transition structures [ts-a2] and [ts-d2], and exo-mode transition structures [ts-b2] and
[ts-c2] are described in Scheme 6 and Figure 3, in which all ester-carbonyl groups are
activated by Me2AlCl. Each chloride atom in [ts-a2]–[ts-d2] locates near an α-proton
of the ester group because a hydrogen-bonding interaction occurs between the
electronically negative chloride atom and electronically positive hydrogen atom.
Therefore, stable six-membered rings, including (i) a strong coordination of the oxygen
atoms in ester groups to aluminum atoms, and (ii) an additional hydrogen-bonding
interaction between the chloride atoms in Me2AlCl and C-2 protons of the ester groups,
were efficiently created in these transition states. However, the calculated relative
Gibbs free energies in Table 2 indicate that (i) [ts-c2] and [ts-d2] have higher ΔGrel
values (5.33–5.01 kcal/mol for [ts-c2]; 3.51–3.23 kcal/mol for [ts-d2]) compared with
those of [ts-a2] and [ts-b2] similar to the thermal IMDA reaction of 8M, (ii) the
endo-mode transition state [ts-a2] is more stable than the exo-mode transition state
[ts-b2] because the ΔGrel of [ts-a2] is much lower than that of [ts-b2] (ΔGrel = 1.50–1.22
kcal/mol at 298.15 K (25 °C)). It is assumed that the steric repulsion between one of
the methyl groups in Me2AlCl and the substrate is quite high during the conversion of
the triene 8M–Me2AlCl complex into the transition structure [ts-b2] as depicted in
Scheme 6.
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 16 of 39
1
2
16
3
4
5
6
7
Scheme 6. Transition Structures of the Me AlCl-Mediated IMDA Reaction of
2
8
9
8M
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 17 of 39
Journal of Medicinal Chemistry
1
2
17
3
4
5
6
7
8
9
2.724 Å
2.786 Å
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
2.726 Å
1.929Å
1.920Å
2.015 Å
1.992 Å
2.776 Å
[ts-a2] (ΔGref = 0.00 kcal/mol)
[ts-b2] (ΔGref = 1.22 kcal/mol)
(B3LYP/6-311G**//B3LYP/6-31G*)
(B3LYP/6-311G**//B3LYP/6-31G*)
1.916Å
1.929 Å
1.997 Å
2.002Å
2.711Å
2.731 Å
2.700 Å
2.700Å
[ts-c2] (ΔGref = 5.01 kcal/mol)
[ts-d2] (ΔGref = 3.23 kcal/mol)
(B3LYP/6-311G**//B3LYP/6-31G*)
(B3LYP/6-311G**//B3LYP/6-31G*)
Figure 3. Three-dimensional transition structures, [ts-a2] (endo), [ts-b2] (exo),
[ts-c2] (exo), and [ts-d2] (endo) in the Me AlCl-mediated IMDA reaction of 8M.
2
Table 2. Calculated Relative Gibbs Free Energies (ΔG ) (in kcal/mol) of the
rel
Transition Structures in the Me AlCl-Mediated IMDA Reaction of 8M
2
B3LYP/6-31G* B3LYP/6-31G** B3LYP/6-311G* B3LYP/6-311G** Predict (25 °C)
[ts-a2]
[ts-b2]
[ts-c2]
[ts-d2]
0.00
1.50
5.33
3.48
0.00
1.41
5.27
3.34
0.00
1.32
5.23
3.51
0.00
1.22
5.01
3.23
93
··
7
·· ! ··
0
··
0
89
··
11
0
··
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 39
1
2
18
3
4
5
6
7
8
9
Based on the present preferable theoretical prediction, we actually examined the IMDA
reaction of 8 using Et2AlCl instead of the ideal reaction of 8M with Me2AlCl, and the
desired 20a was solely obtained after reducing the temporarily formed 7a; namely, the
production of the endo-type (9R,10R)-Δ1,2-octalin derivative 7a exclusively occurs
without formation of not only the exo-type (9S,10R)-Δ1,2-octalin derivative 7b, but also
other stereoisomers, such as the (9R,10S)- and (9S,10S)-Δ1,2-octalin derivatives 7c and
7d (Scheme 5, bottom equation).
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
Completion of the First Total Synthesis.
Next, installation of the
α,β,γ,δ-bisunsaturated ester moiety in the (9R,10R)-Δ1,2-octalin derivative 20a was
attempted, as shown in Scheme 7. Oxidation of 20a using TPAP with NMO was
carried out to provide the corresponding aldehyde 5, and the following HWE olefination
18
of the resulting aldehyde 5 with phosphate 6
smoothly proceeded to afford the
desired triene 21 in good yield. After hydrolysis of the ester moiety in 21 with lithium
hydroxide, deprotection of the silyl ether group at the C-6 position of the resulting
carboxylic acid 3 was eventually examined using 12 M hydrochloric acid, and the
19
desired product 2 was readily prepared in excellent yield by this facile procedure.
All the spectral data, including the optical rotation of the totally synthetic 2,
corresponded to those reported for naturally occurring (+)-2 (our synthetic sample;
24
20
1,2
[α]D +56.4 (c 0.593, MeOH); lit
[α]D +58.71 (c 0.998, MeOH)), and all the
absolute configurations of the stereogenic centers at the C-3, C-4, C-5, C-6, C-7, C-9,
and C-10 positions were unequivocally determined by this identification. On the other
ACS Paragon Plus Environment

Page 19 of 39
Journal of Medicinal Chemistry
1
2
19
3
4
5
6
7
8
9
12
hand, it was revealed that MNBA-mediated amidation using an excess amount of
DMAP (3.0 equiv.) in the absence of triethylamine produced the 3-picolylamide 22
from the carboxylic acid 3 in a satisfactory yield. Finally, the transformation of 22 by
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
19
deprotection of the TBS group furnished the targeted molecule 1 in very high yield.
Thus, an efficient method for the synthesis of 1 was established via Et AlCl-mediated
2
diastereoselective IMDA reaction of the chiral linear precursor 8. The synthesis of 1
1
proceeds in 14 steps and 19% overall yield from the chiral oxazolidinone 16. The H
13
and C NMR spectra of the obtained 1 showed that the synthetic sample has the same
relative stereochemistry as formerly prepared 1 derived from naturally occurring 2.
Other properties of the synthetic 1 including its optical rotation were identical to those
26
of formerly prepared (+)-1 derived from (+)-2 (our synthetic sample; [α]D +73.6 (c
1,2
0.480, MeOH); lit
[α]D20 +76.69 (c 1.103, MeOH)).
a
Scheme 7. Completion of Total Syntheses of 1 and 2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 39
1
2
20
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Reagents and conditions: (a) TPAP, NMO, CH Cl , 0 °C, 30 min; (b) phosphate 6, LHMDS,
2
2
THF, –78 °C to rt, 14 h, 92% from 20a; (c) 4 M LiOH, MeOH, THF, rt, 1 d; (d) 12 M HCl,
MeOH, THF, rt, 16 h, 94% from 21; (e) amine 4, MNBA, DMAP, CH Cl , rt, 14 h, 99% from
2
2
21; (f) 12 M HCl, MeOH, THF, rt, 14 h, 95%.
ACS Paragon Plus Environment

Page 21 of 39
Journal of Medicinal Chemistry
1
2
21
3
4
5
6
7
8
9
■ BIOLOGICAL ASSAYS OF THE TOTALLY SYNTHETIC AMF-26
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
Anti-Tumor Activity of the Totally Synthetic 1 against a Variety of
Human Cancer Cell Lines. To obtain a preliminary evaluation of the totally
synthetic 1, we determined the biological activity of the synthetic 1 for cell growth
inhibition and disrupting the Golgi system. As shown in Figure 4, the global
anti-tumor activity of the totally synthetic 1 against a variety of human cancer cells was
2,20
assessed on a panel of 39 human cancer cell lines (JFCR39).
Among them, the
synthetic 1 strongly inhibited the growth of several cancer cell lines at concentrations of
less than 0.04 M (GI50 values at 18.2 nM for BSY-1, 22.9 nM for MDA-MB-231
[breast], 30.2 nM for SNB-75 [central nervous system], 35.5 nM for KM-12, 33.9 nM
for HT-29 [colon], 30.9 nM for NCI-H226 [lung], 33.1 nM for LOX-IMVI [melanoma],
34.7 nM for OVCAR-4, 37.2 nM for OVCAR-8 [ovary], and 36.3 nM for MKN1, 27.5
nM for MKN45 [stomach]).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 39
1
2
22
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
Figure 4. Growth inhibition against a panel of human cancer cell lines. Dose-dependent
effects of the totally synthetic 1 on each cancer cell were measured as described in the
experimental procedures. The x-axis represents the logarithm of the mean of the GI50 values
(log[concentration (M)]) for each cell line. The y-axis represents the growth rate of each
human cancer cell line (% of control). Quantification of the GI50 value was represented as the
mean of four different experiments. Br, breast; CNS, central nervous system; Co, colon; Lu,
lung; Me, melanoma; Ov, ovary; Re, renal; St, stomach; xPg, prostate.
The informatic anti-cancer drug discovery system, JFCR39, is a powerful tool for
20
extracting compounds with similar modes of action from a database.
The JFCR39
panel was coupled with our own drug sensitivity database, which is comparable to that
21
of the NCI60 panel from the National Cancer Institute.
By using the COMPARE
computer algorithm (details described in the experimental procedures), it is possible to
ACS Paragon Plus Environment

Page 23 of 39
Journal of Medicinal Chemistry
1
2
23
3
4
5
6
7
8
9
correlate the growth inhibitory patterns of JFCR39 (termed “fingerprints”) of a test
compound with those of a known compound in the database. We compared the
fingerprint of the totally synthetic 1 with that of the formerly prepared 1 derived from a
natural molecule 2 (Figure 5), and demonstrated a significant correlation between the
two (correlation coefficient; r = 0.843).
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
Figure 5. Fingerprints of the formerly prepared 1 and totally synthetic 1. Growth inhibition
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 39
1
2
24
3
4
5
6
7
8
9
was assessed against a panel of 39 human cancer cell lines. Fingerprints were produced by
computer processing of the 50% growth inhibition (GI50) values. The logarithm of the GI50
value for each cell line is indicated in the figure. In the plot, columns to the right of zero
indicate the sensitivity of the cell line to the compound, and columns to the left indicate
resistance to the compound. The x-axis represents the logarithm of the difference between the
mean of the GI50 values for 39 cell lines and the GI50 value for each cell line in the JFCR39
panel. The fingerprint of the totally synthetic 1 (right column) significantly correlated with
that of the formerly prepared 1 derived from a natural molecule 2 (left column) (Pearson
correlation coefficients; r = 0.843). MG-MID, the mean of the log GI50 values for the 39 cell
lines; delta, the logarithm of the difference between the MG-MID and log GI50 of the most
sensitive cell line; range, the logarithm of the difference between the log GI50 of the most
resistant cell line and log GI50 of the most sensitive cell line. One scale represents one
logarithm difference.
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
Golgi Disruption and Proliferation Tests of BSY-1 and HBC-4 Cells by
the Totally Synthetic AMF-26. We previously reported that 1 disrupted the Golgi
2
system by inhibiting the Arf1 activation.
We now examined whether the totally
synthetic 1 induced Golgi disassembly by use of immunofluorescence staining with a
monoclonal antibody to GBF1 (Golgi brefeldin A resistant guanine nucleotide
exchange factor 1), a marker of Golgi. In the control cells, GBF1 was observed in the
perinuclear region. The addition of the totally synthetic 1 caused a rapid release of
GBF1 into the cytoplasm as shown in Figure 6 (B,D), similar to the addition of the
ACS Paragon Plus Environment

Page 25 of 39
Journal of Medicinal Chemistry
1
2
25
3
4
5
6
7
formerly prepared 1 derived from a natural molecule 2, as shown in Figure 6 (A,C).
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
Figure 6. Disrupting Golgi apparatus by the formerly prepared 1 and totally synthetic 1
[dispersion of GBF1 in BSY-1 cells (A,B) or HBC-4 cells (C,D)]. Cells were treated with
either the formerly prepared 1 (A,C) or the totally synthetic 1 (B,D) at different concentrations
for 1 h. The cells were then fixed and labeled with anti GBF1 (green) and DAPI (blue). In
the BSY-1 and HBC-4 cells treated with the formerly prepared 1 or the totally synthetic 1,
GBF1 was dispersed from the perinuclear region, which gave a ribbon-like appearance (white
arrowhead), into the cytoplasm. Scale bar: 20 µm.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 39
1
2
26
3
4
5
6
7
8
9
The proliferation of BSY-1 or HBC-4 cells treated with drugs was monitored as shown
in Figure 7 (A). Changes in the amounts of the total cellular protein after 48 h of drug
treatment were analyzed using the SRB assay. Furthermore, the BSY-1 or HBC-4
cells were treated with the formerly prepared 1 and totally synthetic 1 at different
concentrations for 1 h, and then stained with antibodies against GBF1 (Figure 7 (B)).
Over 100 features were measured under each condition, and the numbers of cells whose
Golgi was dispersed were then counted. The biological efficacy of the totally
synthetic 1 was consistent with that of the semi-synthesized 1 based on these data;
therefore, we concluded that the structure of the totally synthetic 1 was unambiguously
identified with that of the formerly prepared 1, which was derived from a natural
compound 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
Figure 7. Cell growth inhibition and Golgi disruption by the formerly prepared 1 and totally
synthetic 1. Symbols represent the following: open circle, BSY-1 cells treated with the
formerly prepared 1; open square, BSY-1 cells treated with the totally synthetic 1; closed circle,
ACS Paragon Plus Environment

Page 27 of 39
Journal of Medicinal Chemistry
1
2
27
3
4
5
6
7
8
9
HBC-4 cells treated with the formerly prepared 1; closed square, HBC-4 cells treated with the
totally synthetic 1.
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
■ SYNTHESIS AND SAR STUDIES OF AMF-26 DERIVATIVES
By using intermediate 20a, we further prepared compounds 26 and 31 as derivatives of
1 as depicted in Schemes 8 and 9, respectively. The two-carbon elongation of the
aldehyde derived from 20a was carried out to afford the corresponding unsaturated ester
23 (Scheme 8). The saponification of 23 and the successive MNBA-mediated
amidation of 24 provided 25 in a very high yield. Finally, compound 26, which
includes a shortened side-chain compared to 1, was successfully obtained by the
treatment of 25 with hydrochloric acid. We also prepared another derivative 31
according to the procedure for the synthesis of 1. As depicted in Scheme 9, compound
31 has a longer alkyl group (ethyl group) at the α-position of the amide group instead of
a methyl group at the same position in 1. We then tried to evaluate the efficacy of
these novel derivatives, 26 and 31, based on SAR studies of the side-chain in 1. We
first determined the biological activity of these derivatives for cell growth inhibition
using JFCR39. Supplementary Figure 1 in the supporting information represents
fingerprints of the formerly prepared 1, the totally synthetic 1, compound 26, and
compound 31. The experimental conditions and abbreviations are described in Figure
5. The logarithm of the mean GI50 values of compound 31 was –6.11, meaning that
the growth inhibition efficacy of compound 31 was about 10-fold weaker than that of 1
and its efficacy of the growth inhibition was moderate. The fingerprint of compound
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 39
1
2
28
3
4
5
6
7
8
9
31 significantly correlated that of the formerly prepared 1 (correlation coefficient; r =
0.816). On the other hand, the logarithm of the mean GI50 values of compound 26
was –4.02, suggesting that this compound showed a very weak efficacy. There was no
correlation between the fingerprint of compound 26 and that of the formerly prepared
(r = 0.117).
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, we examined the Golgi disruption efficacy of the two derivatives. The addition
of compound 31 caused a release of GBF1 into the cytoplasm as depicted in
Supplementary Figure 2 (B,D) in the supporting information. However, no Golgi
disruption was observed at the concentration of 1 µM in the compound-26-treated
BSY-1 cells as depicted in Supplementary Figure 2 (A), or 10 µM in the
compound-26-treated HBC-4 cells as depicted in Supplementary Figure 2 (C).
The proliferation of BSY-1 or HBC-4 cells treated with drugs (formerly prepared 1, the
totally synthetic 1, compound 26, and compound 31) was monitored as shown in
Supplementary Figure 3 (A,B) in the supporting information. Furthermore, the
observed Golgi disrupting efficacy of all drugs was shown in Supplementary Figure 3
(C,D). Modification of the side-chain induced a wide range of biological efficacies.
Compound 31 moderately inhibited cell growth (GI50 value at 0.55 µM for HBC-4, and
0.19 µM for BSY-1), and disrupted the Golgi apparatus (EC50 value at 0.85 µM for
HBC-4, and 0.34 µM for BSY-1), however, compound 26 showed very weak biological
22
activities (both the GI50 value and EC50 value were above 10 µM).
ACS Paragon Plus Environment

Page 29 of 39
Journal of Medicinal Chemistry
1
2
29
3
4
5
6
7
a
Scheme 8. Preparation of a Novel Derivative 26
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Reagents and conditions: (a) TPAP, NMO, MS 4Å, CH Cl , 0 °C, 30 min; (b)
2
2
(EtO) P(O)CH CO Et, LHMDS, THF, –78 °C to rt, 14 h, 64% from 20a; (c) 4 M LiOH,
2
2
2
MeOH, THF, rt, 14 h, 99%; (d) amine 4, MNBA, DMAP, CH Cl , rt, 13 h, 99%; (e) 12 M HCl,
2
2
MeOH, THF, rt, 13 h, 94%.
ACS Paragon Plus Environment