Synthesis and evaluation of analogues of N-phthaloyl-l-tryptophan (RG108) as inhibitors of DNA methyltransferase 1

Article abstract of DOI:10.1021/jm401419p

DNA methyltransferases (DNMT) are promising drug targets in cancer provided that new, more specific, and chemically stable inhibitors are discovered. Among the non-nucleoside DNMT inhibitors, N-phthaloyl-l-tryptophan 1 (RG108) was first identified as inhibitor of DNMT1. Here, 1 analogues were synthesized to understand its interaction with DNMT. The indole, carboxylate, and phthalimide moieties were modified. Homologated and conformationally constrained analogues were prepared. The latter were synthesized from prolinohomotryptophan derivatives through a methodology based amino-zinc-ene-enolate cyclization. All compounds were tested for their ability to inhibit DNMT1 in vitro. Among them, constrained compounds 16-18 and NPys derivatives 10-11 were found to be at least 10-fold more potent than the reference compound. The cytotoxicity on the tumor DU145 cell line of the most potent inhibitors was correlated to their inhibitory potency. Finally, docking studies were conducted in order to understand their binding mode. This study provides insights for the design of the next-generation of DNMT inhibitors.

Full text of DOI:10.1021/jm401419p

Subscriber access provided by DUESSELDORF LIBRARIES
Article
Synthesis and Evaluation of Analogues of N-Phthaloyl-L-
tryptophan (RG108) as Inhibitors of DNA Methyltransferase 1
Saâdia Asgatay, Christine Champion, Gaël Marloie, Thierry Drujon, Catherine Senamaud-
Beaufort, Alexandre Ceccaldi, Alexandre Erdmann, Arumugam Rajavelu, Philippe Schambel, Albert
Jeltsch, Olivier LEQUIN, Philippe Karoyan, Paola Barbara Arimondo, and Dominique Guianvarc'h
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 13 Dec 2013
Downloaded from http://pubs.acs.org on January 1, 2014
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 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis and Evaluation of Analogues of Nꢀ
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
Phthaloylꢀ
L
ꢀtryptophan (RG108) as Inhibitors of
DNA Methyltransferase 1
Saâdia Asgatay,# Christine Champion,$† Gaël Marloie,# Thierry Drujon,# Catherine
SenamaudꢀBeaufort,$ Alexandre Ceccaldi,$† Alexandre Erdmann,‡ Arumugam Rajavelu,£
Philippe Schambel,∫ Albert Jeltsch,£ Olivier Lequin,# Philippe Karoyan,# Paola B. Arimondo,
$‡*, Dominique Guianvarc’h#*
#Laboratoire des BioMolécules, UMR 7203, Université Pierre et Marie CurieꢀParis 6 ꢀ ENS ꢀ
CNRS, 4, place Jussieu, 75252 Paris Cedex 05, France. $MNHN CNRS UMR 7196, INSERM
U565, 43 rue Cuvier 75005 Paris France. †UPMC Université Paris 6, 75005 Paris, France. £
Institute of Biochemistry, Faculty of Chemistry, University Stuttgart, Pfaffenwaldring 55, 70569
Stuttgart, Germany. ∫Institut de Recherche Pierre Fabre, Centre de Recherche Pierre Fabre, 17
rue Jean Moulin, 81 106 Castres cedex, France. ‡USR ETaC CNRSꢀPierre Fabre n°3388 –
CRDPF BP 13562, 3 avenue Hubert Curien 31 100 Toulouse France
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 55
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
KEYWORDS. DNA methyltransferase, inhibitors, conformationally constrained analogs,
RG108, epigenetics.
ABSTRACT.
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
DNA methyltransferases (DNMT) are promising drug targets in cancer provided that new, more
specific, and chemically stable inhibitors are discovered. Among the nonꢀnucleoside DNMT
inhibitors, NꢀPhthaloylꢀLꢀtryptophan 1 (RG108) was first identified as inhibitor of DNMT1.
Here, 1 analogs were synthesized to understand its interaction with DNMT. The indole,
carboxylate and phthalimide moieties were modified. Homologated and conformationally
constrained analogs were prepared. The latter were synthesized from prolinoꢀhomotryptophan
derivatives through a methodology based aminoꢀzincꢀeneꢀenolate cyclization. All compounds
were tested for their ability to inhibit DNMT1 in vitro. Among them, constrained compounds 16ꢀ
18 and NPys derivatives 10ꢀ11 were found at least 10ꢀfold more potent than the reference
compound. The cytotoxicity on the tumor DU145 cell line of the most potent inhibitors was
correlated to their inhibitory potency. Finally, docking studies were conducted in order to
understand their binding mode. This study provides insights for the design of the nextꢀgeneration
of DNMT inhibitors.
INTRODUCTION
DNA methylation is an epigenetically inherited chemical modification that is implicated in gene
1
expression regulation. In humans, DNA methylation occurs at position 5 of the cytidine mainly
in a CpG context. CpG sequences are not distributed randomly in the genome but are grouped in
2
,3
CpG islands overrepresented in gene promoter regions, in transposons and repeated elements.
ACS Paragon Plus Environment
2

Page 3 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Methylation of CpGꢀrich promoters is often associated with transcription inhibition and
methylation of repeated regions ensures the stability of the genome. Epigenetic regulation plays
an important role not only in the normal functioning of cells (as in embryonic development,
cellular differentiation, imprinting, chromosome X inactivation), but also in pathologies 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
4
,5,6,7
particularly in cancers.
Indeed, in tumorigenesis, the epigenetic regulatory system is
disrupted leading to global hypomethylation and inappropriate hypermethylation of CpG islands,
and thus to aberrant expression of a battery of genes involved in critical cellular processes, from
DNA damage repair to apoptosis, cell cycle control, hormone response, cell adhesion/metastasis,
and carcinogen detoxification. The first example of CpG methylation of a tumor suppressor gene
8
in a human cancer was described for Rb (retinoblastoma) in 1989, but it is not until very
recently that this process has been confirmed for other genes and its importance as drug target
9
has been understood. Interestingly, DNA methylation is reversible, thus potentially allowing
malignant cells to revert to a more normal state leading to their proliferation arrest and ultimately
1
0
to their death. The methylation process is carried out by the C5ꢀDNA methyltransferases
11
(
DNMTs) and it has been shown that upon use of DNMT inhibitors it is possible to reactivate
1
2,13,14
genes silenced by promoter methylation in cancers.
So, DNMTs are valuable candidate
targets for anticancer therapy and several efforts are directed toward developing small molecules
that target these enzymes. Most active compounds are nucleosides analogs such as 5ꢀazaꢀcytidine
(5ꢀazaꢀC) and 5ꢀazaꢀdeoxycytidine (5ꢀazaꢀdC), which, once incorporated in DNA in place of
cytosine, covalently block the DNMT on position 6 of the nucleotide (Figure 1). 5ꢀazaꢀC and 5ꢀ
azaꢀdC have been both approved by FDA for myelodysplastic syndromes and acute myeloid
1
5,16
leukemia, and are currently used in clinical trials (Phases II, III) for other cancers.
Zebularine
is also a nucleoside analog that stabilizes the DNMTꢀDNA complex and shows a lower
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 55
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
cytotoxicity. However, in comparison to the irreversible analogues, zebularine has to be used at
higher concentrations to obtain the same demethylation effect in cells, and thus never entered
clinical trials. The azanucleosides are very efficient but exhibit little selectivity due to their
incorporation everywhere in the DNA and are chemically unstable although some efforts have
been made in this area, as evidenced by the identification of the dinucleotide 5ꢀazaꢀdC followed
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
1
7
by a deoxyguanosine (SGIꢀ110) (Figure 1) that shows a better pharmacokinetic profile. This
compound is also in clinical trials for the treatment of myelodysplastic syndromes, acute myeloid
leukemia and solid tumors. Nonꢀnucleoside analogs have also been found to inhibit DNA
18
methylation, even if their mechanism of action has not been systematically elucidated yet. This
inhibitors’ family is extremely heterogeneous due to the various chemical structures of the
compounds but also to their inhibitory mechanisms, which are not always clearly identified. It
19
20
includes natural compounds like tea polyphenol , (ꢀ)ꢀepigallocatechinꢀ3ꢀgallate (EGCG) ,
2
1
22
23
24
genistein , nanaomycin A , psammaplin A or most recently laccaic acid A, and also
synthetic drugs already used in the treatment of other pathologies like hypertension
25
26,27
(
hydralazine ), or as local anesthetic (procainamide and procaine
) (Figure 1). Over the past
few years, various DNMT inhibitor compounds were discovered from virtual screenings by the
2
8,29
Lyko’s group
, among which NꢀPhthaloylꢀ
Lꢀtryptophan 1 that will be further discussed
3
0,31
(Figure 1).
More recently, a quinoline derivative (SGIꢀ1027), which was initially synthesized
for its antitumor activities, was reported as an efficient DNMT inhibitor both in vitro and in cells
32
(
Figure 1). Sꢀadenosylꢀ
Lꢀhomocysteine (AdoHcy) analogs were also described as inhibitors of
human DNMTs but these compounds suffer from low specificity toward DNMTs, because they
3
3,34
target other Sꢀadenosylꢀ
Lꢀmethionine (AdoMet)ꢀdependent methyltransferases as well.
In
summary, today there are several DNMT inhibitors known, but the high toxicity, lack of
ACS Paragon Plus Environment
4

Page 5 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
35
specificity, weak activity or chemical instability of these molecules hinder their use. Therefore,
there is a great need to design new drugs that specifically target DNMTs. Currently, the nonꢀ
nucleosideꢀbased analogs are more interesting to develop and, in an effort to discover new
compounds, we recently developed several strategies. First, we have set up a HTS based on
fluorescence detection to find new inhibitors of DNMT from chemical libraries, allowing the
characterization of promising flavones and flavanones derivatives (Figure 1) with IC in the
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
3
6
nanomolar range. In a second approach, we increased the inhibitors activity by conjugating
them to a DNA ligand. We chose procainamide as a DNA binder to guide an inhibitor of DNMT
37
to CpGꢀrich regions, in order to increase its local concentration at CpG sites (Figure 1). The
pharmacomodulation of nonꢀnucleosidic validated inhibitors specific for DNMTs constitutes a
third strategy to increase the potency of these inhibitors. Among them, 1, a nonꢀnucleoside
30,31
inhibitor discovered from an in silico screen,
presents some interesting features (Figure 1).
Indeed, it has been shown to inhibit DNMT in a nonꢀcovalent manner and to cause
3
0
demethylation of genomic DNA in cells without any detectable toxicity. More recently,
maleimide derivatives of 1 were designed to trap DNMT1 through addition of the catalytic thiol
3
8
of the enzyme to the maleimide moiety (Figure 1). Herein, we conducted a structureꢀactivity
relationship (SAR) study to understand the important features of 1 and its interaction in the
DNMT active site, in order to improve its activity. We synthesized a small library of analogs of 1
shown in Figure 2. We chose to modify the indole, carboxylate and phthalimide moieties. We
designed also constrained analogs to partially reduce the flexibility of the inhibitor. This strategy
was recently used to improve other DNMT inhibitors such as: procaine and AdoHCy (Figure
3
9,40
1).
Finally, we also synthesized β2ꢀ and β3ꢀphthaloylated homotryptophane as homologated
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 55
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
compounds. We report here the synthesis, the biological activity and a structural study of these
derivatives of 1.
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
Chemistry
Synthesis of compounds 2ꢀ11 (Figure 2). The compounds 1(S)
,
1(R), 2ꢀ6 and 8 were obtained
ꢀtryptophan, ꢀtryptophan, 5ꢀ
ꢀalanine, 3ꢀ(2ꢀoxoꢀ1,2ꢀ
ꢀtryptophanamide, respectively, according to Casimir et
from the commercially available corresponding amino acids
L
D
hydroxyꢀ
Lꢀtryptophan,
Lꢀhistidine,
Lꢀphenylalanine, 3ꢀbenzothienylꢀL
dihydroꢀ4ꢀquinolinyl)ꢀ
L
ꢀalanine and
L
41
al.’s procedure (Scheme 1). The methyl 2ꢀ((succinimidooxy)carbonyl)benzoate (MSB) was
reacted with the sodium salt of the corresponding amino acid in water/acetonitrile affording the
corresponding phthaloylated compounds 2ꢀ6, 8 in good yields. The tryptophan quinazolinedione
derivative 9 was obtained by reacting 2ꢀcarbomethoxyphenyl isocyanate 20 with
Lꢀtryptophan
(LꢀTrp) in a mixture of dioxane and water containing NaOH 1M at 50°C following the method
4
2
reported by Canonne et al. (Scheme 2). Synthesis of 10 and 11 were achieved in good yields by
treatment of ꢀtryptophan with 2ꢀnitropyridinylsulfenyl chloride and 2,4ꢀdinitrophenylsulfenyl
L
chloride, respectively, in aqueous alkaline solution (Scheme 2).
Synthesis of homologated analogues 12ꢀ14 (Figure 2). β2ꢀphthaloylꢀhomotryptophan 12(R) and
12(S) were obtained from the corresponding β2ꢀhomotryptophan 26 by using the same
phthaloylation procedure as described in Scheme 1. For the synthesis of β2ꢀhomotryptophan, we
used the diastereoselective methodology based on aminomethylation of chiral silyl enol ether
4
3
recently described by Moumné et al. (Scheme 3). This method allows the preparation of
enantiomerically pure functionalized β2ꢀamino acids. Briefly, 3ꢀindolylpropanoic acid 21 was Nꢀ
benzylated leading to compound 22 that was activated to the corresponding mixed anhydride
ACS Paragon Plus Environment
6

Page 7 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
with isobutyl chloroformate and coupled to (+) or (ꢀ) Oppolzer’s Sultam as chiralꢀauxiliary (in
accordance with the desired enantiomer) to give 23 in 55% overall yield. In the key step of the
synthesis, 23 was converted to the silyl enol ether intermediate that was aminomethylated in the
presence of highly reactive trifluoroacetate dibenzylidene imminium salt to give 24 in 88% yield.
Finally, β2ꢀhomotryptophan 26 was obtained after successive protection/deprotection steps, i.e.
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
(i) deprotection of the dibenzylamine moiety by hydrogenolysis, (ii) removal of the Sultam
moiety by saponification, (iii) Boc amine protection, (iv) deprotection of the indole under Birch
conditions and (v) amine deprotection in TFA.
β3ꢀphthaloylꢀhomotryptophan 13 was obtained from the commercially available corresponding
β3ꢀhomotryptophan upon use of the same phthaloylation procedure as described in Scheme 1.
The dipeptide 14 was obtained by coupling
that was activated to the corresponding mixed anhydride with isobutyl chloroformate in THF at
15°C affording the dipeptide 27. Next, 27 was debenzylated by catalytic hydrogenation over
Lꢀtryptophan benzyl ester and the phthaloyl glycine 7
ꢀ
palladium charcoal to afford 14 in good yield (Scheme 4).
Synthesis of constrained analogues 15ꢀ18 (Figure 2). The constrained analogues 15ꢀ18 were
obtained from the Boc protected prolinoꢀhomotryptophan derivative 30 (Scheme 5). This
compound was synthesized through a methodology developed by Mothes et al. based on aminoꢀ
44
zincꢀeneꢀenolate cyclization. Carbocyclization of olefin 28 was achieved after deprotonation
with LDA and transmetallation in the presence of ZnBr . Compound 29 was obtained through
2
Negishi crossꢀcoupling of the NꢀBocꢀ3ꢀiodoꢀindole with the cyclic zinc intermediate by using
Pd(OAc) in the presence of (tꢀBu PH)ꢀBF . 29 was debenzylated by catalytic hydrogenation
2
3
4
over palladium charcoal affording 30. Finally, 15, 16 and 17 were obtained after reacting 30 in
the presence of Nꢀphthaloylglycine 7, naphthoyl or benzoyl chloride, respectively, and
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 55
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
deprotection in TFA. 30 was also treated with 2,4ꢀdinitrophenylsulfenyl chloride and 18 was
obtained after deprotection in TFA.
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 AND DISCUSSION
1 is a rationally designed small molecule inhibitor of DNMT that was found by computational
screening of a database of compounds with potential high affinity to a threeꢀdimensional model
4
5
of the human DNMT1 catalytic pocket. Since at that time the Xꢀray structure of DNMT1 was
not available, this model was established by using a homology modeling approach with known
46
structures of bacterial C5ꢀDNMT (M. Hha I and M. Hae III) and human DNMT2.
Actually, the precise molecular mode of action of this inhibitor is not known but it presumably
acts by blocking the active site of the enzyme, as predicted by the molecular modeling study.
45,46
According to Siedlecki et al.
, the phthalimide moiety seems to occupy the same part of the
binding pocket as the native cytosine. In addition, the docking study suggested that (i) the
anionic carboxylate group of 1 plays a critical role and makes hydrogen bonds with Arg 1310,
and (ii) the phthalimide group lies next to the catalytic cysteine in the active site of DNMT1,
4
5
making a hydrogen bond through its carbonyl to Arg 1308. By saturation transfer difference
(STD) NMR spectroscopy experiments, we observed that 1 interacts with human DNMT1, the
3
7
catalytic domain of murine Dnmt3A and the catalytic complex of murine Dnmt3A/3L. Suzuki
3
8
et al. docked the maleimide derivatives of 1 in the crystal structure of M.Hha I. They found that
the indole ring is located in a hydrophobic pocket formed by Tyr254, Gly88 and Gly255, while
the carboxylate forms two hydrogen bonds with Arg163. Thus, the features implicated in the
binding interaction with the enzyme are still not fully elucidated. In this context, we have
undertaken a systematic in vitro study to explore the contributions of each functional group of
ACS Paragon Plus Environment
8

Page 9 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the molecule to the binding. All synthesized compounds were tested in an in vitro DNA
3
methylation assay, measuring the incorporation of tritiated methyl from Hꢀlabeled AdoMet by
human recombinant DNMT1 into a hemimethylated DNA duplex as substrate. All compounds
were first tested at 500 ꢁM in the DNMT1 inhibition activity assay (Figure 3) and next, the
inhibitory concentration IC was evaluated on DNMT1 for our most efficient compounds to
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
further explore their inhibitory efficiency (Table 1).
3
0,45
Since the activity of a racemic mixture of 1 was initially reported,
we studied whether the
configuration of the asymmetric carbon plays an important role in the inhibitory activity and
established the stereospecificity of the activity. In our enzyme assay, 1(S) showed only a modest
IC value of 390 ꢁM in comparison with the 115 nM value initially reported by the authors
50
3
0
describing 1 activity. In fact, their value was obtained on the bacterial DNMT M.Sss I and
using an unmethylated DNA substrate. In addition, the two biological tests are different: here, we
measured the direct incorporation of a radiolabeled methyl group on a short hemimethylated
DNA substrate containing 8 CpGs and not the indirect inhibition of a restriction enzyme by
methylation of a 798 pb DNA substrate containing several CpGs. Actually, notorious variations
in compounds activity in DNMT inhibition enzymatic tests have already been observed in other
47
studies and depend on the assay and the chosen DNA substrate. However, our data are in
agreement with the recent findings of Suzuki et al. reporting a 34% inhibition of recombinant
38
human DNMT1 activity at 1 concentration of 1 mM. 1(R) showed an activity comparable to
(S) and was shown to bind to DNMT by NMR using STD experiments. This result led us to
1
reassess the interaction model proposed above. In order to further characterize the binding mode
of 1, we investigated whether 1 could compete with known ligands of DNMT by 1D NMR
spectroscopy. Indeed, the binding of 1 and AdoHcy ligands was followed using WaterLOGSY
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 55
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
48,49,50
(
WaterꢀLigand Observed via Gradient SpectroscopY).
This experiment relies on the
observation of intermolecular NOE magnetization transfer via bulk water. Small molecules in
solution are characterized by positive NOEs with bulk water while ligands interacting in fast
exchange with macromolecules typically have negative NOEs. This NMR study was carried out
on the murine catalytic domain of Dnmt3A, a DNMT model available in large amounts. As
shown in Figure 4, the adenine protons of AdoHcy display positive NOEs in the absence of
Dnmt3A (Figure 4b) whereas in the presence of Dnmt3A the NOE signals are cancelled or
become negative (Figure 4c), indicating that AdoHcy interacts with the enzyme. The addition of
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
1
(Figure 4d) leads to a 50% reduction in the intensity of the negative NOE and the appearance
of a positive NOE for the AdoHcy protons resonating at 8.26 and 8.36 ppm, respectively,
indicating decreased binding of AdoHcy. In addition, 1 protons exhibit strong negative NOEs
due to the binding to Dnmt3A. These NMR observations indicate that 1 and AdoHcy compete
for binding to Dnmt3A, leading to a new interaction model for 1.
In quest of determinants of 1 binding to DNMT, we chose to modify each functional group of the
compound: the indole, the carboxylate and the phthalimide moiety. The derivatives lacking the
indole or phthalimide moiety, i.e. phthaloyl glycine 7 and
Lꢀtryptophan, respectively, were
inactive at 500 ꢁM, emphasizing the need for these pharmacophores. Several indole substituents
were tested. The replacement of indole by a most acid heterocycle such as imidazole (compound
3), by a nonꢀheterocyclic phenyl group (compound 4) or by a quinolone moiety (compound 6)
was unfavorable as these modifications resulted in the absence of inhibitory potency at 500 ꢁM.
A 5ꢀhydroxy derivative 2 was also tested, because it was the starting material for the synthesis of
4
5
conjugated compounds. Indeed, according to the modeling of Lyko’s group , the 5ꢀposition of
the tryptophan sticks out of the DNMT active site. Accordingly, a conjugated biotinylated
ACS Paragon Plus Environment
10

Page 11 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
51
derivative of 1 was already synthesized. No significant difference was observed between
compounds 1 and 2 validating the 5ꢀposition of the tryptophan as an anchor point for potential
linkers. Interestingly, bioisosteric replacement of the indole moiety by the bicyclic
benzothiophene (i.e., variation of the heteroatom from nitrogen to sulfur) was favorable since the
analog 5 is 2ꢀfold better than 1 with an IC of 230 ꢁM. Thus, the replacement of the good Hꢀ
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
bond donor NH group by a poor Hꢀbond acceptor sulfur atom led to a better inhibitory efficacy,
indicating that the DNMT binding pocket is sensitive to changes in the electronic character of the
aromatic system. With regard to the phthalimide group, we replaced it by a quinazolinedione
substituent (compound 9) by assuming that it could mimic the pyrimidine ring of cytidine as the
4
5,46
phthalimide moiety was supposed to occupy the same part of the binding pocket as cytosine.
However, this modification did not dramatically affect the activity. Most interestingly, the
phthalimide moiety was favorably replaced by Npys group or 2,4ꢀdinitrobenzenesulfenyl:
compounds 10 and 11 showed an important increase in activity with IC values of 20 and 40
50
ꢁ
M, respectively, a 20ꢀ and 10ꢀfold gain in comparison with 1. We already observed the
importance of the presence of an electron withdrawing group such as nitro through our previous
3
6
study on another DNMT inhibitors family, i.e. flavanones. To determine the contribution of the
negative charge of the carboxylate group in 1/DNMT interaction, we replaced the carboxylate
group by a neutral isosteric amide function in tryptophanamide 8. Indeed, an analogue of 1,
named RG119 that lacks the carboxylate group was previously reported as inactive in
30
demethylating genomic DNA. 8 showed a low activity at 500 ꢁM similar to that of 1, so the
determinant role of the negative charge in 1 activity seems to be inconclusive.
Next, we examined whether an extension of compound 1 could better occupy the space of the
binding pocket and possibly improve the binding affinity for DNMT1. For this purpose,
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 55
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
homologated analogs of 1 were synthesized in order to explore the space of the ligand binding
site. Two isomers of phthaloylated βꢀtryptophan, which have their amino group bound to carbon
β instead of α, were assayed: β2, in which the side chain is also bound to carbon β (compounds
12(R) and 12(S)), and β3ꢀtryptophan, in which the side chain is bound to carbon α (compound
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
13). Furthermore, a dipeptide NꢀPhtꢀGlyTrp 14 was assayed. Unfortunately, no significant
improvement in activity was observed with this family of compounds. This could be explained
by the fact that the important moieties of the compound are located too far away to each other.
Finally, we attempted to restrict 1 to a bioactive conformer by reducing its flexibility. The two
enantiomers of a prolino analogue of tryptophan were synthesized and different aromatic
substituents were conjugated to the amine moiety of the proline ring in order to mimic the
phthaloyl part. When phthaloylglycine was conjugated through a peptide link in order to mimic
the phthaloyl moiety of 1 (see compound 15), no improvement in activity was observed in
comparison with 1. Interestingly, the linkage of a benzoyl or naphthoyl moiety was favorable and
compounds 16 and 17 showed increased activity with 50 to 100% inhibition at 500 ꢁM. IC50
determinations for these compounds confirmed their potent inhibition activity: IC50 values
ranging from 128 to 50 ꢀM were obtained as compared to 390 ꢀM for reference compound 1.
Noteworthy, in each series of prolino derivatives, the enantiomer (2R,3S) was always the most
active. The compound bearing the naphthoyl moiety (16) showed a slightly better activity than
the compound with the benzoyl one (17) (IC of 50 ꢀM and 73 ꢀM respectively, for the most
5
0
active enantiomer 16 and 17). Therefore, the prolino compounds constitute a promising new
family of 1 derivatives since our better constrained analog 16 (2R,3S) is 8ꢀfold better than 1. The
important features of each best compound found in this study were associated in compound 18
(2R,3S): a prolino analog of tryptophan bearing a Npys substituent. Unfortunately, we did not
ACS Paragon Plus Environment
12

Page 13 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
obtain the expected synergic activity since an IC50 value of 80 ꢀM was obtained, which is
comparable to that of compound 17 (2R,3S).
Finally, we determined the cytotoxicity of the most potent inhibitors on prostate cancer cells
DU145 (Table 1). We chose this cell line for our study as methylationꢀmediated repression of
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
suppressor tumor genes was clearly observed in prostate cancer and their reactivation constitutes
a potential approach for the treatment of this cancer. In the presence of 1, we observed a modest
660 ꢀM cytotoxicity value. Under the same conditions, the prolino compounds, the 2ꢀ
nitropyridinylsulfenyl and 2,4ꢀdinitrobenzenesulfenyl derivatives showed lower TC values (2
5
0
to 6 fold). Interestingly, a link between IC50 and TC50 was observed for the enantiomers of
compound 16: the 16 (2R, 3S) enantiomer is the most active on DNMT1 as well as in terms of
cytotoxicity. This correlation is less clear in the case of compound 17, as the difference in
inhibition activity is smaller. However, in general the best DNMT inhibitors showed the highest
cytotoxicities. In conclusion, the cytotoxicity values obtained with the compounds are improved
in comparison to that observed with 1, but these values remain modest, the best compound
showing a 100 ꢁM cytotoxicity.
In order to understand the gain in activity for the most potent compounds, molecular docking
studies were carried out on a threeꢀdimensional model of murine catalytic Dnmt1 (residues 732ꢀ
5
2
1
600) built from the Xꢀray structure of the catalytic complex (PDB 4DA4). The choice of this
structure has been guided by the fact that we wanted to start from a model in which both the
cofactor, here the product AdoHcy, and the DNA substrate, cytidine, are wellꢀresolved, together
with the BAH1, BAH2 and the entire MTase domain, even though the CXXC domain (residues
6
46ꢀ692) and the autoinhibitory linker domain (residues 699ꢀ733) are lacking. Inducedꢀfit
docking (IFD) studies were carried out with the constrained analogs 16 and 17. The IFD
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 55
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
procedure was validated upon docking of the AdoHcy exactly as the ligand of the crystal
structure (Supplementary Figure 1). 1 was found in the AdoMet pocket (Figure 5A); its carboxyl
group interacts with Lys1247 and Arg 1576 and the carbonyl of the phthalimide interacts also
with Lys1247 (Figure 5B). This result is consistent with our NMR study showing that 1 and
AdoHcy compete in binding to DNMT. Figure 5C compares 1 and the most potent maleimide
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
8
derivative of Suzuki et al. (compound 5 of their article). Both compounds are perfectly
superimposed and the same interactions with Arg1576 and Lys1247 were found. The SYBYLꢀX
1.3 software revealed a partially different positioning in the MTase catalytic site for 1 and the
45
38
maleimide derivative compared to previous findings on DNMT1 and M. Hha I, respectively.
Inhibitors 16 (Figure 5D) and 17 (Supplementary Figure S2) fit the AdoMet pocket and interact
through the carboxyl with Lys1247 as 1 (Figure 5E and Supplementary Figure S2). In addition,
the proline group fits in the ribose pocket of the cofactor and, unexpectedly, the naphthoyl and
the benzoyl fit towards the cytidine pocket of the crystal structure, forming an interaction with
Arg1576 through the carbonyl of the substituent. These compounds thus fill the entire accessible
catalytic pocket, thereby explaining their higher inhibition activity (up to 8ꢀfold). Similarly, the
IFD analysis suggested that compound 11, 10ꢀfold more potent than 1, also occupies the same
pockets (data not shown). Concerning compound 5, its superior inhibitory efficacy could be
explained by the hydrophobic nature of the pocket, in which the indole moiety binds.
The main conclusion regarding the interaction of the constrained analogues with both the DNA
and AdoMet pockets was further confirmed by a selectivity assay performed on compound 17
(2R, 3S). Indeed, we investigated whether 17 could affect the activity of other methyltransferases
in order to get insights on the mechanism of inhibition. 17 was tested against the histone
53
methyltransferase (HMTK) G9a, which methylates histone tails on their lysine residues, and
ACS Paragon Plus Environment
14

Page 15 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the bacterial DNA (adenine Nꢀ6)ꢀmethyltransferase, EcoDam, which methylates DNA at the
5
4,55,56
3
adenine within GATC sites.
A biotin–avidin microplate kinetic assay based on HꢀSAM
5
7
incorporation was used for the quantitative analysis of methyltransferase activities in vitro.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Interestingly, 17 showed a very good specificity for the DNMT1, as compared to the other two
methyltransferases, since it did not affect the other methyltransferase activities (up to a
concentration of 750 ꢀM for EcoDam and 1 mM for G9a), while the IC for DNMT1 was 73
50
ꢀM (Table 1).
CONCLUSION
The nonꢀnucleosidic DNMT inhibitors constitute a heterogeneous family of compounds with a
high potential therapeutic interest but more data are needed to understand their mode of action.
In this context, we examined NꢀPhthaloylꢀLꢀtryptophan 1, because it was described as the first
specific inhibitor of DNMT. Our systematic study showed that both the indole and phthaloyl
moieties are required to observe an inhibitory effect. However, it is possible to improve 1 by
replacing these groups by a benzothiophene and Npys group or 2,4ꢀdinitrobenzenesulfenyl,
respectively. Furthermore, constrained analogs also showed an increased efficiency in DNMT1
inhibition. A molecular modeling study led us to propose the following hypothesis explaining the
higher inhibition activity of the prolino compounds: unlike 1, which fits in the AdoMet pocket in
our model, the prolino compounds fit both the AdoMet and the cytidine pockets, therefore filling
the entire accessible catalytic pocket. In conclusion, our study provides a structureꢀactivityꢀ
relationship framework to be integrated for the conception of the next generation of DNMT
inhibitors. Prolino compounds seem most promising and it would be useful to design derivatives
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 55
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
exhibiting the main features from this study: prolino analogs containing the benzothiophene
heterocycle may be envisaged. Moreover, the introduction of substituted naphthoyl moieties, for
example with nitro group, should be considered.
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
EXPERIMENTAL SECTION
Chemistry. Solvents and most of the starting materials were purchased from Sigma Aldrich.
4
1
1
13
1
(S) and 1(R) were prepared according to from Lꢀ and Dꢀtryptophan. H and C NMR spectra
were recorded on Bruker ARX 250 and 400 spectrometers. Chemical shifts are given in parts per
1
13
million (ppm) using the CDCl signal as reference (δ H = 7.26 ppm, δ C = 77.16 ppm) and
3
coupling constants (J) are given in Hertz (Hz). All yields reported are unoptimized. Flash
chromatography was performed using silica gel Merck 60 with 0.040–0.063 ꢁm. Some
compounds were separated and purified by preparative reverse phase HPLC using a Symmetry,
C8 (250 x 10 mm; 5ꢁm) column with a multiple gradient of acetonitrile in water at a flow rate of
3
mL/min with UV detection at 220 and 280 nm. Analytical HPLC conditions were as follows:
column Waters X Terra MS C18 (250 x 4.6 mm, 5 mm) with a gradient of acetonitrile in water
(0 to 100% during 30 min) at a flow rate of 1 mL/min, detection at 220 and 280 nm. Purity of
tested compounds was established by analytical reverse phase HPLC confirming a purity of at
least 95%. High resolution mass spectra (HRMS) were recorded at the Institut de Chimie
Moléculaire (FR 2769) of UPMC (electrospray source).
General Synthetic Method for compounds 2ꢀ6, 8, 13.
To a stirred solution of the appropriate amino acid (1 equiv) and sodium carbonate (5 equiv)
dissolved in water/acetonitrile, was added methyl 2ꢀ((succinimidooxyl)carbonyl)benzoate (2
ACS Paragon Plus Environment
16

Page 17 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
41
equiv) freshly prepared. The mixture was vigorously stirred overnight at room temperature. The
solution was acidified to pH = 1ꢀ2 with 2N HCl, diluted with EtOAc and washed with 1N HCl
and water. The organic phase was dried with Na SO , filtered, and concentrated in vacuo. The
2
4
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
product is isolated by preparative reverse phase HPLC, on a Symmetry, C8 (250 x 10 mm; 5ꢁm)
column with a multiple gradient of acetonitrile in water (20 to 40% during 5 min, 40 to 45 %
during 20 min) at a flow rate of 3 mL/min.
2(S)ꢀ(1,3ꢀDihydroꢀ1,3ꢀdioxoꢀ2Hꢀisoindolꢀ2ꢀyl)ꢀ3ꢀ(5ꢀhydroxyꢀ1Hꢀindoleꢀ3ꢀyl)ꢀpropanoic acid
1
(
2): light yellow solid, 11 % yield; HPLC Rt = 10.3 min; HꢀNMR (MeOD): δ (ppm) 7.75 (m,
4H); 7.07 (dd, J = 0.4 Hz, J = 8.5 Hz, 1H); 6.90 (dd, J = 0.4 Hz, J = 2.2 Hz, 1H); 6.88 (s, 1H);
6.60 (dd, J = 2.2 Hz, J = 8.7 Hz, 1H); 5.20 (dd, J = 11.6 Hz, J = 4.8 Hz, 1H); 3.70 (dd, J = 15.1
13
Hz, J = 11.4 Hz, 1H); 3.60 (dd, J = 14.9 Hz, J = 4.8 Hz, 1H). CꢀNMR (MeOD): δ (ppm)
172.59; 169.11; 151.16; 135.43; 132.90; 132.85; 129.13; 124.90; 124.17; 112.65; 112.44;
+
110.50; 103.22; 54.17; 25.60. HRMS calcd for C19
H
15
N
2
O
5
[MH ]: 351.0975. Found: 351.0996
2(S)ꢀ(1,3ꢀDihydroꢀ1,3ꢀdioxoꢀ2Hꢀisoindolꢀ2ꢀyl)ꢀ3ꢀ(3Hꢀimidazolꢀ4ꢀyl)ꢀpropanoic acid (3):
1
white solid, 11 % yield; HPLC Rt = 6.7 min; HꢀNMR (DMSOꢀd
6
): δ (ppm) 8.92 (s, 1H); 7.90
(br s, 4H); 7.44 (s, 1H); 5.11 (dd, J = 10.5 Hz, J = 4.4 Hz, 1H); 3.54 (dd, J = 15.8 Hz, J = 4.3 Hz,
1
3
1
6
H); 3.38 (dd, J = 15.5 Hz, J = 10.5 Hz, 1H). CꢀNMR (DMSOꢀd ): δ (ppm) 169.10; 167.02;
12 3 4
134.97; 134.09; 130.93; 129.31; 123.51; 117.16; 50.99; 24.16. HRMS calcd for C14H N O
+
[MH ]: 286.0822. Found: 286.0802
2(S)ꢀ(1,3ꢀDihydroꢀ1,3ꢀdioxoꢀ2Hꢀisoindolꢀ2ꢀyl)ꢀ3ꢀphenylpropanoic acid (4): white solid, 32 %
1
yield; HPLC Rt = 17 min; HꢀNMR (CDCl ): δ (ppm) 7.77 (m, 2H); 7.67 (m, 2H); 7.17 (m, 5H);
3
13
5
.19 (dd, J = 9.0 Hz, J = 6.6 Hz, 1H); 3.55 (m, 2H). CꢀNMR (CDCl ): δ (ppm) 173.56; 167.36;
3
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 55
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
1
36.41; 134.14; 131.49; 128.80; 128.60; 126.93; 123.53; 52.99; 34.41. HRMS calcd for C H N
1
7
14
+
NaO [MNa ]: 318.0737. Found: 318.0736
4
3
ꢀ(Benzothienꢀ3ꢀylmethyl)ꢀ2(S)ꢀ(1,3ꢀdihydroꢀ1,3ꢀdioxoꢀ2Hꢀisoindolꢀ2ꢀyl)ꢀpropanoic
acid
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
1
(5): white solid, 30 % yield; HPLC Rt = 22.1 min; HꢀNMR (MeOD): δ (ppm) 7.82 (m, 2H);
7.76 (br s, 4H); 7.31 (m, 2H); 7.23 (s, 1H); 5.31 (dd, J = 10.9 Hz, J = 5.0 Hz, 1H); 3.86 (dd, J =
13
1
4.9 Hz, J = 10.9 Hz, 1H); 3.79 (dd, J = 5.0 Hz, J = 15.1 Hz, 1H). CꢀNMR (MeOD): δ 171.97;
168.93; 141.71; 139.77; 135.61; 133.00; 132.78; 125.36; 125.11; 124.85; 124.28; 123.78;
+
122.32; 99.13; 53.05; 28.56. HRMS calcd for C H NO S [MH ]: 352.0638. Found: 352.0628
1
9
14
4
2(S)ꢀ(1,3ꢀDihydroꢀ1,3ꢀdioxoꢀ2Hꢀisoindolꢀ2ꢀyl)ꢀ4ꢀquinolinepropanoic acid (6): white solid, 19
1
%
yield; HPLC Rt = 17.9 min; HꢀNMR (DMSOꢀd6): δ (ppm) 7.86 (br s, 4H); 7.81 (d, J = 8.1
Hz, 1H); 7.48 (dd, J = 8.1 Hz, J = 7.0 Hz, 1H); 7.28 (d, J = 8.3 Hz, 1H); 7.18 (dd, J = 8.1 Hz, J =
7.0 Hz, 1H); 6.26 (s, 1H); 5.19 (dd, J = 10.9 Hz, J = 4.4 Hz, 1H); 3.75 (dd, J = 14.7 Hz, J = 4.4
13
Hz, 1H); 3.52 (dd, J = 14.4 Hz, J = 10.9 Hz, 1H). CꢀNMR (DMSOꢀd6): δ (ppm) 169.54;
167.03; 161.03; 147.14; 138.82; 135.01; 130.67; 130.37; 124.00; 123.49; 121.84; 121.65;
+
1
18.19; 115.73; 50.91; 30.64. HRMS calcd for C H N O [MH ]: 363.0975. Found: 363.0972
2
0
15
2
5
2(S)ꢀ(1,3ꢀDihydroꢀ1,3ꢀdioxoꢀ2Hꢀisoindolꢀ2ꢀyl)ꢀ3ꢀ(1Hꢀindoleꢀ3ꢀyl)ꢀpropanamide (8): light
1
yellow solid, 35 % yield; HPLC Rt = 13 min; HꢀNMR (MeOD): δ (ppm) 7.73 (br s, 4H); 7.55
(d, J = 7.9 Hz, 1H); 7.22 (d, J = 8.1 Hz, 1H); 7.01 (td, J = 1.0 Hz, J = 8.12 Hz, J = 7.00 Hz, 1H);
6
.92 (s, 1H); 6.92 (td, J = 1.0 Hz, J = 8.0 Hz, J = 7.7 Hz, 1H); 5.19 (dd, J = 11.4 Hz, J = 5.2 Hz,
13
1
H); 3.75 (dd, J = 14.7 Hz, J = 11.4 Hz, 1H); 3.66 (dd, J = 14.7 Hz, J = 5.3 Hz, 1H). CꢀNMR
(MeOD): δ (ppm) 172.58; 167.88; 136.43; 133.76; 131.62; 126.86; 122.78; 122.54; 120.85;
+
118.20; 117.62; 110.67; 109.77; 53.99; 24.02. HRMS calcd for C H N O [MH ]: 334.1186.
1
9
16
3
3
Found: 334.1221
ACS Paragon Plus Environment
18

Page 19 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
(S)ꢀ(1,3ꢀdihydroꢀ1,3ꢀdioxoꢀ2Hꢀisoindolꢀ2ꢀyl)ꢀ3ꢀ(1Hꢀindoleꢀ3ꢀyl)ꢀ butanoic acid (13): light
1
yellow solid, 26 % yield; HPLC Rt = 15.7 min; HꢀNMR (MeOD): δ (ppm) 10.14 (s, 1H); 7.75
(br s, 4H); 7.58 (d, J = 7.9 Hz, 1H); 7.26 (d, J = 8.1 Hz, 1H); 6.95 (m, 3H); 4.99 (m, 1H); 3.41
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
3
(
dd, J = 14.2 Hz, J = 9.4 Hz, 1H); 3.28 (m, 3H); 2.92 (dd, J = 16.0 Hz, J = 5.7 Hz, 1H). Cꢀ
NMR (MeOD): δ (ppm) 174.88; 169.75; 137.99; 135.23; 133.03; 128.62; 124.15; 122.33;
+
1
3
3
19.69; 119.25; 112.18; 111.92; 54.7; 37.5; 29.3. HRMS calcd for C H N NaO [MNa ]:
2
0
16
2
4
71.1002. Found: 371.1003
(2H)ꢀQuinazolineacetic acid, 1,4ꢀdihydroꢀαꢀ(1Hꢀindolꢀ3ꢀylmethyl)ꢀ2,4ꢀdioxoꢀ, (S)ꢀ (9): The
ꢀtryptophan (1.5
commercially available isocyanate 20 (0.3 mmol) was added to a solution of
L
equiv) in a mixture of 20 mL of water/dioxane (1/1) and sodium hydroxide (1M, 10 mL). The
mixture was stirred at 50°C for 4h. The reaction mixture was diluted with H O, acidified with
2
hydrochloric acid (10%) to pH = 3 and extracted with EtOAc. The combined organic layers were
concentrated in vacuo and the crude residue was purified using reverse phase HPLC, on a
Symmetry, C8 (250 x 10 mm; 5ꢁm) column with a multiple gradient of acetonitrile in water (20
to 40% during 5 min, 40 to 45 % during 20 min) at a flow rate of 3mL/min leading to 9 as a
1
white solid, 10 % yield. HPLC Rt = 13.4 min; HꢀNMR (DMSOꢀd6): δ (ppm) 8.30 (d, J = 8.5
Hz, 1H); 7.89 (dd, J = 1.5 Hz, J = 4.6 Hz, 1H); 7.58 (d, J = 7.6 Hz, 1H); 7.44 (m, 1H); 7.34 (d, J
=
7.92 Hz, 1H); 7.18 (d, J = 2.4 Hz, 1H); 7.08 (m, 1H); 7.01 (d, J = 7.9 Hz, 1H); 6.95 (m, 1H);
1
3
4
.41 (m, 1H); 3.19 (dd, J = 14.2 Hz, J = 4.6 Hz, 1H); 3.04 (dd, J = 14.4 Hz, J = 9.6 Hz, 1H). Cꢀ
NMR (MeOD): δ (ppm) 174.1; 154.5; 142.6; 136.1; 133.5; 127.1; 123.5; 120.9; 120.2; 119.3;
+
1
18.3; 118.15; 111.3; 110.2; 53.7; 27.2. HRMS calcd for C H N O [MH ]: 350.1135. Found:
1
9
16
3
4
3
50.1139
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 55
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
N2ꢀ[(3ꢀnitroꢀ2ꢀpyridinyl)thio]ꢀ
LꢀTryptophan (10): To a solution of Lꢀtryptophan (40 mg) in
2M NaOH (1 mL) and dioxane (5 mL), nitropyridinylsulfenyl chloride (1.1 equiv) was added
dropwise over 15 min, whilst 2M NaOH (2 mL) was added dropwise with vigorous stirring. The
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
mixture was diluted with H O (50 mL), filtered from precipitated disulfide, overlayered with
2
EtOAc and acidified with 0.5 M H SO . Upon extraction with EtOAc, the combined extracts
2
4
were washed twice with H O and then dried over Na SO . The solution was concentrated and
2
2
4
purified by HPLC in the same conditions as described above leading to 10 as a yellow solid, 50
1
% yield. HPLC R = 21.6 min; HꢀNMR (MeOD): δ (ppm) 8.50 (dd, J = 8.1 Hz, J = 1.5 Hz, 1H);
t
7.98 (dd, J = 4.6 Hz, J = 1.3 Hz, 1H); 7.59 (d, J = 8.1 Hz, 1H); 7.41 (d, J = 8.1 Hz, 1H); 7.24 (m,
2H); 7.14 (t, J = 7.9 Hz, 1H); 7.02 (t, J = 7.9 Hz, 1H); 3.97 (dd, J = 8.8 Hz, J = 5 Hz, 1H); 3.43
13
(dd, J = 14.5 Hz, J = 4.8 Hz, 1H); 3.17 (dd, J = 14.5 Hz, J = 8.7 Hz, 1H). CꢀNMR (MeOD): δ
(ppm) 175.8; 163.8; 154.3; 140.7; 135.0; 129.1; 124.9; 122.4; 120.8; 119.7; 119.5; 118.9; 112.3;
+
107.0; 65.0; 30.5. HRMS calcd for C H N O S [MH ]: 359.0808. Found: 359.0797
1
6
15
4
4
N2ꢀ[(2,4ꢀdinitrophenyl)thio]ꢀLꢀTryptophan (11): 11 was obtained following the same
procedure as for compound 10 using 2,4ꢀdinitrophenylsulfenyl chloride instead of
nitropyridinylsulfenyl chloride. 11 was obtained as a yellow solid, 70 % yield; HPLC R = 24.7
t
1
min; HꢀNMR (MeOD): δ (ppm) 8.74 (br s, 1H); 7.55 (m, 1H); 7.43 (m, 2H) ; 7.31 (m, 1H) ;
1
3
7
.22 (s, 1H); 7.13 (m, 1H); 6.98 (m, 1H); 3.77 (m, 1H); 3.44 (m, 1H); 3.10 (m, 1H). CꢀNMR
MeOD): δ (ppm) 175.8; 160.3; 154.0; 143.9; 141.3; 136.7; 127.3; 125.8; 125.6; 123.8; 121.3;
20.1; 118.9; 118.7; 118.1; 117.6; 111.0; 110.5; 64.7; 28.7. HRMS calcd for C H N O6S
(
1
1
7
15
4
+
[
MH ]: 403.0706. Found: 403.0715
(Nꢀbenzylindoyl) propanoic acid (22): To a stirred solution of 3ꢀindolylpropionic acid 21 (10 g)
in DMF (55 mL) cooled to 0°C was added sodium hydride (2.2 equiv) slowly. After 30 min,
ACS Paragon Plus Environment
20

Page 21 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
benzyle bromide (1 equiv) was added. The mixture was allowed to react over 1 h, then LiOH
(1.1 equiv) was added. After 1 h, the mixture was quenched with water, acidified until pH 2 and
extracted with ether. The combined organic layers were washed with saturated aqueous NH Cl
4
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
solution, dried over MgSO and evaporated under reduced pressure. Silica gel column
4
chromatography purification (cyclohexane/EtOAc/AcOH, 7/3/0.1) afforded 22 as a white solid,
1
7
5% yield. R (cyclohexane/EtOAc: 4/1) 0.25; mp 117 °C; H NMR (CDCl ): δ 7.60 (d, J = 7.6
f
3
Hz, 1H); 7.22ꢀ7.29 (m, 5H); 7.17 (t, J = 7.1 Hz, 1H); 7.11 (t, J = 7.6 Hz, 1H); 7.07 (d, J = 7.1
1
3
Hz, 1H); 6.93 (s, 1H); 5.24 (s, 2H); 3.71 (t, J = 7.6 Hz, 2H); 2.76 (t, J = 7.6 Hz, 2H). C NMR
(
CDCl ): δ 179.5; 137.6; 136.7; 128.7; 127.7; 127.5; 126.7; 125.6; 121.9; 119.1; 118.8; 113.8;
3
109.7; 49.9; 34.8; 20.4. Anal. Calcd for C H NO : C, 77.40; H, 6.13; N, 5.01. Found: C, 77.12;
18
17
2
H, 6.37; N, 4.89.
1R)ꢀNꢀ3ꢀ(Nꢀbenzylindoylpropanoyl) camphorsultam (23): A solution of (+) 10,2ꢀ
(
ꢀ1
camphorsultam (1 equiv) in dimethoxyethane (3 mL.mmol ) was slowly added at room
temperature to a suspension of sodium hydride (1.1 equiv) in the same solvent. The reaction
mixture was stirred for 1 h. At the same time, isobutyl chloroformate and Nꢀmethylmorpholine
(1.1 eq) were added to a solution of (Nꢀbenzylindoyl) propanoic acid 22 (1 equiv) in
ꢀ1
dimethoxyethane (3 mL.mmol ) cooled at ꢀ15°C. This solution was stirred for 15 min at ꢀ15°C,
then filtered through a Celite pad under argon atmospher over the sultam sodium salt solution.
The reaction mixture was stirred for 1 h at ꢀ15°C and 1 h at room temperature and then quenched
with water, and concentrated. AcOEt was added and the organic layer was washed with aqueous
NH Cl solution, dried over MgSO and evaporated under reduced pressure. The compound was
4
4
obtained by precipitation in an etherꢀpentane mixture. Silica gel column chromatography
purification (cyclohexane/EtOAc/NEt , 8/2/0.1) afforded 23 as a white solid, 70% yield. R
3
f
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 55
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
20
1
(
cyclohexane/EtOAc/NEt : 8/2/0.1) 0.42; mp 123 °C; [α]D 57 (c 1. CHCl ); H NMR (CDCl ):
3
3
3
δ 7.63ꢀ7.65 (m, 1H); 7.20ꢀ7.28 (m, 4H); 7.07ꢀ7.16 (m, 4H); 6.97 (s, 1H); 5.23 (s, 2H); 3.85 (t, J
6.3 Hz, 1H); 3.45 (d, J = 13.9 Hz, 1H); 3.42 (d, J = 13.9 Hz, 1H); 3.06ꢀ3.19 (m, 4H); 2.03ꢀ2.05
=
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
1
3
(m, 2H); 1.81ꢀ1.87 (m, 3H); 1.29ꢀ1.39 (m, 2H); 1.05 (s, 3H); 0.92 (s, 3H); C NMR (CDCl ): δ
3
1
71.5; 137.7; 136.6; 128.7; 127.9; 127.4; 126.7; 125.8; 121.7; 119.1; 119.0; 113.8; 109.5; 65.2;
2.9; 49.8; 48.4; 47.7; 44.6; 38.5; 36.1; 32.8; 26.4; 20.7; 20.2; 19.8. Anal. Calcd for
5
C H N O S: C, 70.56; H, 6.77; N, 5.88. Found: C, 70.31; H, 6.98; N, 5.67.
28
32
2
3
(
1R)ꢀNꢀ(2ꢀDibenzylaminomethylꢀ3ꢀNꢀbenzylindoyl propanoyl) camphorsultam (24):
Trimethylsilylsulfate (1.5 equiv) and then triethylamine (1.1 equiv) were added to a
ꢀ1
dichloromethane solution of the sulfonamide 23 (1 equiv, 1 mL.mmol ). The reaction mixture
was stirred for 6 h at room temperature. A solution of a N,Nꢀdibenzyliminium trifluoroacetate
salt (1.1 equiv) in dichloromethane, freshly prepared from N,N,N’,N’,tetrabenzylꢀ
methanediamine was added and the reaction mixture was stirred for 15 min. A solution of TBAF
ꢀ1
in THF (1 mol.L , 1 equiv) was added, and then water was added. Dichloromethane was
evaporated under reduced pressure and the aqueous solution was extracted with ether. The
organic layer was washed with saturated aqueous NH Cl solution, dried over MgSO and
4
4
evaporated under reduced pressure. Silica gel column chromatography purification
(
cyclohexane/EtOAc, 9/1) afforded 24, 88% yield. R (Cyclohexane/EtOAc, 4/1) 0.35; mp 80–81
f
20
1
°
C; [α]D 25 (c 1. CHCl ); H NMR (CDCl ): δ 7.78ꢀ7.80 (m, 1H); 7.09ꢀ7.34 (m, 16H); 6.99ꢀ
3
3
7.01 (m, 2H); 6.85 (s, 1H); 5.20 (s, 2H); 3.88 (m, 2H); 3.42ꢀ3.52 (m, 6H); 3.21 (dd, J = 14.4 Hz,
.8 Hz, 1H); 2.94 (dd, J = 12.4 Hz, 10.3 Hz, 1H); 2.86 (dd, J = 14.4 Hz, 9.1 Hz, 1H); 2.56 (dd, J
12.6 Hz, 4.5 Hz, 1H); 2.06ꢀ2.15 (m, 2H); 1.85ꢀ1.87 (m, 3H); 1.30ꢀ1.37 (m, 2H); 1.23 (s, 3H);
3
=
1
3
0
.96 (s, 3H). C NMR (CDCl ): δ 174.7; 138.7; 137.9; 136.6; 129.1; 128.6; 128.4; 127.9; 127.0;
3
ACS Paragon Plus Environment
22

Page 23 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
3
7
26.7; 126.6; 121.6; 119.6; 112.4; 109.4; 65.4; 57.5; 55.8; 53.2; 49.8; 48.3; 47.7; 44.7; 38.6;
2.8; 26.5; 25.8; 21.0; 19.9. Anal. Calcd for C H N O S: C, 75.29; H, 6.91; N, 6.13. Found: C,
43 47
3
3
5.45; H, 7.20; N, 5.81.
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
(
R)ꢀ(Nꢀtertꢀbutyloxycarbonylaminomethyl)ꢀ3ꢀNꢀbenzylindolylpropanoic
acid
(25):
ꢀ1
ꢀ1
Pd(OH) /C 20% (200 mg.mmol ) was added to a THFꢀMeOH (1/1, 5 mL.mmol ) solution of
2
camphorsultam. The reaction mixture was stirred at room temperature under 5 bars hydrogen for
6
7 h. The solution was filtered through a Celite pad and then concentrated. The crude residue
was dissolved in acetonitrileꢀH O (1/1, 5 mL.mmolꢀ1). LiOH (2 equiv) was added and the
2
reaction mixture was stirred overnight at room temperature. Acetonitrile was removed under
ꢀ1
reduced pressure and the aqueous layer was extracted once with ether. Dioxane (2.5 mL.mmol ),
Boc O (1.2 equiv) and LiOH (pH = 9) were then added. The reaction mixture was stirred for 24
2
h, then acidified at pH = 2 at 0°C with HCl 1M and extracted with AcOEt. The organic layer was
washed with saturated brine, dried over MgSO and evaporated under reduced pressure. Silica
4
gel column chromatography purification (cyclohexane/EtOAc/NEt3, 7/3/0.1) afforded 25, 65%
1
yield. H NMR (CDCl ): δ 7.62ꢀ6.98 (m, 10H); 6.62 (br, 0.5H); 5.27 (s, 2H); 4.96 (br, 0.5H);
3
1
3
3
.43ꢀ3.30 (m, 2H); 3.21ꢀ2.84 (m, 3H); 1.43 and 1.25 (2s, 9H); C NMR (CDCl ): δ 179.9;
3
1
78.6; 158.1; 156.08; 137.7; 136.8; 136.6; 129.1; 128.8; 128.3; 128.1; 127.9; 127.6; 127.0;
126.7; 126.6; 125.4; 121.9; 119.3; 118.9; 111.6; 111.4; 109.9; 81.1; 79.7; 50.0; 46.8; 46.5; 42.6;
4
1.3; 28.4; 28.1; 25.7; 25.2.
2
(R)ꢀ(1,3ꢀDihydroꢀ1,3ꢀdioxoꢀ2Hꢀisoindolꢀ2ꢀyl)ꢀ3ꢀ(1Hꢀindoleꢀ3ꢀyl)ꢀpropanoic
acid,
αꢀ
(aminomethyl)ꢀ (12R): To
a
solution of Nꢀtertꢀbutyloxycarbonylaminomethylꢀ3ꢀNꢀ
benzylindolylpropanoic acid 25 in THF (3 mL/mmol) were added the sodium (5 equiv) and
liquid ammonia at ꢀ70°C. The blue solution was stirred for 1h at ꢀ50°C. After evaporation of
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 55
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
ammonia at room temperature, EtOAc and the saturated aqueous solution of NH Cl were added.
4
Upon extraction with EtOAc, the combined extracts were washed with the saturated aqueous
solution of NaCl, dried over MgSO and then concentrated in vacuo leading to Bocꢀprotected
4
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
amino acid. The compound was obtained after flash column chromatography (Cy/EtOAc/AcOH,
7/3/1) as white powder, 48% yield. The compound was then treated with 1 mL TFA in 1 mL
CH Cl leading to the compound 26. The compound 12 was obtained following the same
2
2
procedure as for compound 2 using the compound 26 as amino acid. The product is isolated by
preparative reverse phase HPLC, on a Symmetry, C8 (250 x 10 mm; 5ꢁm) column with a
multiple gradient of acetonitrile in water (20 to 40% during 5 min, 40 to 45 % during 20 min) at
1
a flow rate of 3 mL/min. Light yellow solid, 11% yield. HPLC Rt = 18.5 min; HꢀNMR (CDCl ):
3
δ (ppm) 7.81 (dd, J = 5.5 Hz, J = 3.1 Hz, 2H) ; 7.71 (dd, J = 5.5 Hz, J = 3.1 Hz, 2H); 7.62 (d, J =
7
.7 Hz, 1H); 7.32 (d, J = 7.9 Hz, 1H); 7.18 (td, J = 8.1 Hz, J = 7.0 Hz, 1H); 7.15 (s, 1H); 7.12
(td, J = 1.08 Hz, J = 8.1 Hz, J = 7.0 Hz, 1H); 4.12 (dd, J = 14.0 Hz, J = 8.4 Hz, 1H); 3.98 (dd, J
13.8 Hz, J = 5.9 Hz, 1H); 3.48 (m, 1H); 3.25 (dd, J = 15.1 Hz, J = 7.7 Hz, 1H); 3.09 (dd, J =
=
1
3
1
5.6 Hz, J = 6.8 Hz, 1H). CꢀNMR (CDCl ): δ (ppm) 176.2; 167.1; 135.1; 133; 130.8; 126.5;
3
126.2; 122.3; 121.5; 121.1; 118.5; 117.7; 110.9; 110.1; 109.9; 43.3; 38.4; 28.7. HRMS calcd for
+
C H N O [MH ]: 349.1183. Found: 349.1181
2
0
16
2
4
2(S)ꢀ(1,3ꢀDihydroꢀ1,3ꢀdioxoꢀ2Hꢀisoindolꢀ2ꢀyl)ꢀ3ꢀ(1Hꢀindoleꢀ3ꢀyl)ꢀpropanoic
acid,
αꢀ
1
(
aminomethyl)ꢀ (12S): HꢀNMR (MeOD): δ (ppm) 7.82 (dd, J = 5.7 Hz, J = 3.1 Hz, 2H); 7.72
(dd, J = 5.5 Hz, J = 3 Hz, 2H); 7.62 (d, J = 7.6 Hz, 1H); 7.32 (d, J = 8.1 Hz, 2H); 7.19 (dd, J =
.8 Hz, J = 6.8 Hz, 1H); 7.17 (s, 1H); 7.12 (dd, J = 8.8 Hz, J = 6.7 Hz, 1H); 7.13 (dd, J = 7.4 Hz,
J = 6.7 Hz, 1H); 4.13 (dd, J = 13.8 Hz, J = 8.3 Hz, 1H); 3.99 (dd, J = 13.8 Hz, J = 5.9 Hz, 1H);
7
13
3
.49 (m, 1H); 3.26 (dd, J = 14.9 Hz, J = 7.7 Hz, 1H); 3.10 (dd, J = 15.1 Hz, J = 6.6 Hz, 1H). Cꢀ
ACS Paragon Plus Environment
24

Page 25 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
NMR (MeOD): δ (ppm) 178.1; 168.1; 133.9; 131.8; 129.1; 123.3; 122.5; 122.1; 119.5; 118.7;
+
111.1; 44.2; 39.4; 32.7. HRMS calcd for C H N O [MH ]: 349.1183. Found: 349.1181
2
0
16
2
4
Nꢀ[2ꢀ(1,3ꢀdihydroꢀ1,3ꢀdioxoꢀ2Hꢀisoindolꢀ2ꢀyl)acetyl]ꢀLꢀtryptophan phenylmethyl ester (27):
NꢀMethyl morpholine (1.2 equiv) and isobutyl chloroformate (1.2 equiv) were successively
added to a solution of Nꢀphthaloylglycine (1.2 equiv) in THF (3 mL) at ꢀ20°C. After an
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
activation period of 10 min, Lꢀtryptophane benzyl ester (250 mg) in THF (2 mL) was added to
above solution, and the resulting solution was stirred for 1h at ꢀ20°C prior to the addition of 5%
NaHCO (20 mL). After 30 min at room temperature, the aqueous phase was extracted with
3
CH Cl (3x50 mL). The combined organic layer were washed with 5 % NaHCO (2x20 mL) and
2
2
3
dried over Na SO . After evaporation under reduced pressure and purification by flash
2
4
1
chromatography (EtOAc/ cyclohexane) (1/9), 27 was obtained as a white solid, 78 % yield; Hꢀ
NMR (CDCl ): δ (ppm) 8.15 (s, 1H); 7.85 (m, 2H); 7.75 (m, 2H); 7.52 (d, J = 7.2 Hz, 1H); 7.37
3
(m, 3H); 7.28 (m, 3H); 7.11 (m, 2H); 6.80 (d, J = 2.3 Hz, 1H); 6.45 (d, J = 7.7 Hz, 1H); 5.11 (t, J
=
12.3 Hz, 2H); 5.0 (m, 1H); 4.36 (d, J = 16.2 Hz, 1H); 4.26 (d, J = 16.2 Hz, 1H); 3.36 (m, 2H).
1
3
CꢀNMR (CDCl ): δ (ppm) 171.3; 167.7; 165.7; 135.9; 135.1; 134.1; 131.9; 128.6; 128.5;
3
127.6; 123.6; 123.3; 122.1; 119.8; 118.5; 111.2; 109.3; 67.4; 53.4; 40.5; 27.4. HRMS: calcd for
+
C H N NaO [MNa ]: 504.1529. Found: 504.1519
2
8
23
3
5
Nꢀ[2ꢀ(1,3ꢀdihydroꢀ1,3ꢀdioxoꢀ2Hꢀisoindolꢀ2ꢀyl)acetyl]ꢀLꢀtryptophan (14): The compound 27
(70 mg) was dissolved in THF (8 mL) and the Pd/C 10% (0.1 equiv) was added. The mixture
was stirred under hydrogen atmosphere for 48h. The solution was filtered from Pd/C and
concentrated in vacuo. The product was purified by preparative reverse phase HPLC, on a
Symmetry, C8 (250 x 10 mm; 5ꢁm) column with a multiple gradient of acetonitrile in water (20
to 40% during 5 min, 40 to 45 % during 20 min) at a flow rate of 3 mL/min. White solid, 60%
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 55
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
1
yield; HPLC Rt = 11.2 min. HꢀNMR (MeOD): δ (ppm) 7.87 (m, 2H); 7.81 (m, 2H); 7.56 (d, J =
6.8 Hz, 1H); 7.31 (d, J = 7 Hz, 1H); 7.12 (s, 1H); 7.06 (m, 2H); 4.72 (m, 1H); 4.35 (d, J = 16.6
Hz, 1H); 4.30 (d, J = 16.6 Hz, 1H); 3.35 (dd, 1H, J = 14.6 Hz, J = 6.6 Hz); 3.26 (dd, 1H, J = 14.1
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
1
3
Hz, J = 7 Hz). CꢀNMR (MeOD): δ (ppm) 133.5; 123.1; 120.7; 117.7; 111.1; 53.8; 48.1; 39.6.
+
HRMS calcd for C H N NaO [MNa ]: 414.1060. Found: 414.1059.
2
1
17
3
5
tertꢀButyl 3ꢀ(((2S,3R)ꢀ2ꢀ(Benzyloxycarbonyl)ꢀ1ꢀ((S)ꢀ1ꢀphenylethyl)pyrrolidinꢀ3ꢀyl)methyl)ꢀ
Hꢀindoleꢀ1ꢀcarboxylate (29): LDA (12 mL, 24 mmol) was added at ꢀ78 °C to a solution of (S)ꢀ
1
benzyl 2ꢀ(butꢀ3ꢀenyl(1ꢀphenylethyl)amino)acetate 28 (6.47 g, 20 mmol) in dry THF (30 mL)
under argon. ZnBr (1.2 M in Et O, 50 mL) was then added at the same temperature. The
2
2
reaction mixture was stirred overnight at room temperature under argon. tertꢀButyl 3ꢀiodoꢀ1Hꢀ
indoleꢀ1ꢀcarboxylate (8.92 g, 26 mmol) in dry THF (10 mL), Pd(OAc) (180 mg, 0.80 mmol),
2
and tBu PꢀHBF (290 mg, 1.0 mmol) were then successively added, and the mixture was stirred
3
4
overnight at room temperature under argon. Et O was added, and the organic layer was washed
2
with NH Cl, dried over MgSO , and concentered in vacuo. The residue was purified by flash
4
4
2
0
chromatography (cyclohexane/ethyl acetate 95:5) to give a yellow oil, 60% yield. [α]D ꢀ37.9 (c
1
1
, CHCl ). H NMR (CDCl ) δ 8.09 (d, 1H, J = 8.0 Hz), 7.38ꢀ7.16 (m, 14H), 5.10 (AB, 2H, J =
3
3
12 Hz), 3.73 (q, 1H, J = 6.8 Hz), 3.55 (d, 1H, J = 7.7 Hz), 3.09 (m, 1H), 2.91 (m, 1H), 2.81ꢀ2.73
13
(
m, 2H), 2.36 (m, 1H), 1.92 (m, 1H), 1.76 (m, 1H), 1.64 (s, 9H), 1.35 (d, 3H, J = 6.8 Hz). C
NMR (CDCl ) δ 173.2, 149.8, 144.5, 135.9, 135.5, 130.5, 128.9, 128.6, 128.4, 127.5, 127.2,
3
1
24.4, 122.8, 122.4, 119.0, 115.3, 83.5, 66.7, 66.0, 61.6, 50.1, 41.8, 30.2, 28.3, 26.4, 22.9.
+
HRMS calcd for C H N O [MH ]: 539.2904. Found: 539.2910.
3
4
38
2
4
(2S, 3R)ꢀ3ꢀ((1ꢀ(tertꢀButoxycarbonyl)ꢀ1Hꢀindolꢀ3ꢀyl)methyl)pyrrolidineꢀ2 carboxylic acid
(
30): 29 (8.056 mmol, 4.34 g) was dissolved in MeOH (60 mL) and Pd/C 10% (1.2 g) was
ACS Paragon Plus Environment
26

Page 27 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
added. The mixture was stirred under 5 bar hydrogen for one week. The solution was filtered
over Celite and concentrated. The residue was purified by flash chromatography (CH Cl /MeOH
2
2
1
9
:1) to give an oil, 79% yield. HꢀNMR (CDCl ): δ (ppm) 8.07 (s, 1H); 7.52 (d, J = 7.6 Hz, 1H);
3
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
7
.38 (s, 1H); 7.24 (t, J = 7.9 Hz, 1H); 7.10 (t, J = 7.3 Hz, 1H); 4.40 (m, 1H); 3.83 (m, 2H); 3.12
(d, J = 12 Hz,1H); 2.99 (m, 1H); 2.67 (t, J = 12 Hz, 1H); 1.98 (m, 1H); 1.87 (m, 1H); 1.61 (s,
13
9H); CꢀNMR (CDCl ): δ (ppm) 172.2; 149.8; 135.6; 130.4; 124.5; 123.5; 122.7; 118.1; 115.4;
3
+
110.1; 83.7; 64.6; 44.8; 40.8; 29.4; 28.3; 24.6. HRMS calcd for C H N O Na [MNa ]:
19
24
2
4
3
67.1628 Found: 367.1628
(
2S,3R)ꢀ3ꢀ((1Hꢀindolꢀ3ꢀyl)methyl)ꢀ1ꢀ(2ꢀ(1,3ꢀdioxoisoindolinꢀ2ꢀyl)acetyl)pyrrolidineꢀ2ꢀ
carboxylic acid (15): The compound 30 (1 mmol, 360 mg) was dissolved in DMF (1 mL) under
N , to the solution were added the HATU (1 equiv) and DIPEA (1 equiv), the mixture was stirred
2
for 20 min and the Nꢀphthaloylglycine (1.2 equiv) was added. After stirring the mixture
overnight, EtOAc (20 mL) was added. The organic layer was washed with saturated NaHCO (2
3
x 20 mL) and 10% aqueous citric acid (2 x 20 mL), dried with Na SO , filtered and concentrated.
2
4
The crude residue was dissolved in TFA (3 mL) and the mixture was stirred at room temperature
for 4 h. TFA was coꢀevaporated with cyclohexane and the crude residue was purified by
preparative reverse phase HPLC, on a Symmetry, C8 (250 x 10 mm; 5ꢁm) column with a
multiple gradient of acetonitrile in water (20 to 40% during 5 min, 40 to 45 % during 20 min) at
a flow rate of 3 mL/min. Compound 15 was obtained as a light yellow solid, 20% yield; HPLC
1
Rt = 13.5 min; HꢀNMR (DMSO): δ (ppm) 7.92 (m, 4H); 7.55 (d, 1H, J = 7.6 Hz); 7.36 (d, 1H, J
=
8.1 Hz); 7.23 (s, 1H); 7.09 (dd, 1H, J = 8.1 Hz, J = 7.8 Hz); 6.99 (dd, 1H, J = 7.8 Hz, J = 7.5
Hz); 4.55 (d, 1H, J = 16.8 Hz); 4.42 (d, 1H, J = 8.1 Hz); 4.40 (d, 1H, J = 17 Hz); 3.85 (t, 2H, J =
8.3 Hz ); 3.02 (dd, 1H, J = 14.2 Hz, J = 4.3 Hz); 2.78 (m, 1H); 2.46 (dd, 1H, J = 14.2 Hz, J = 4.1
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 55
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
13
Hz); 1.87 (m, 1H); CꢀNMR (DMSO): δ (ppm) 171.9; 167.5; 164.2; 136.2; 134.7; 131.5; 126.9;
123.2; 122.9; 120.9; 118.2; 118.1; 112.1; 111.4; 83.5; 62.1; 45; 40.7; 38.4; 31; 28.7. HRMS
+
calcd for C H N O [MH ]: 432.1554. Found: 432.1599
2
4
22
3
5
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
(2S,3R)ꢀ3ꢀ(1Hꢀindolꢀ3ꢀyl)methyl)ꢀ1ꢀ(2ꢀnaphthoyl)pyrrolidineꢀ2ꢀcarboxylic acid (16): 30 (1
mmol, 360 mg) was dissolved in CH Cl (10 mL). NEt (2.5 equiv) and naphthoyl chloride (2
2
2
3
equiv) were added successively. After stirring the mixture overnight at room temperature, the
solvent was evaporated and the residue was dissolved in AcOEt and washed with NH Cl. The
4
crude residue was dissolved in TFA (3 mL) and the mixture was stirred at room temperature for
4
h. TFA was coꢀevaporated with cyclohexane and the crude residue was purified by flash
chromatography on silica gel (CH Cl /MeOH 9:1 with AcOH 0.1%) to afford 16 as a beige solid,
2
2
1
6
3% yield. HPLC Rt = 19.4 min. HꢀNMR (CDCl ): δ (ppm) 8.03 (s, 1H); 7.88 (m, 4H); 7.36 (d,
3
1H, J = 8.1 Hz); 7.23 (s, 1H); 7.09 (dd, 1H, J = 8.1 Hz, J = 7.8 Hz); 6.99 (dd, 1H, J = 7.8 Hz, J =
.5 Hz); 4.55 (d, 1H, J = 16.8 Hz); 4.42 (d, 1H, J = 8.1 Hz); 4.40 (d, 1H, J = 17 Hz); 3.85 (t, 2H,
J = 8.3 Hz ); 3.02 (dd, 1H, J = 14.2 Hz, J = 4.3 Hz); 2.78 (m, 1H); 2.46 (dd, 1H, J = 14.2 Hz, J =
7
1
3
4
.1 Hz); 1.87 (m, 1H); CꢀNMR (CDCl ): δ (ppm) 173.3; 157.8; 135.4; 132.4; 128.6; 128.3;
3
1
28.1; 127.8; 127.6; 126.8; 124.1; 122.7; 122.1; 119.5; 118.7; 112.5; 111.3; 64.4; 50.1; 42.7;
+
31; 27.8. HRMS calcd for C H N O [MH ]: 399.1703. Found: 399.1708
2
5
23
2
3
(
2S,3R)ꢀ3ꢀ(1Hꢀindolꢀ3ꢀyl)methyl)ꢀ1ꢀbenzoylpyrrolidineꢀ2ꢀcarboxylic acid (17): 30 (1.25
mmol, 432 mg) was dissolved in CH Cl (15 mL). NEt (2.5 equiv) and benzoyl chloride (2
2
2
3
equiv) were added successively. After stirring the mixture for 4 h at room temperature, the
solvent was evaporated and the residue was dissolved in AcOEt and washed with NH Cl. The
4
crude residue was dissolved in TFA (3 mL) and the mixture was stirred at room temperature
overnight. TFA was coꢀevaporated with cyclohexane and the crude residue was purified by flash
ACS Paragon Plus Environment
28

Page 29 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
chromatography on silica gel (CH Cl /MeOH 9:1 with AcOH 0.1%) to afford 17 as a beige solid
2
2
1
(387 mg, 67%). HPLC Rt = 21.4 min. HꢀNMR (CDCl ): δ (ppm) 8.13 (d, 1H); 7.66 (d, 1H, J =
3
8.1 Hz); 7.52 (m, 5H); 7.23 (t, 1H, J = 7 Hz); 7.16 (t, 1H, J = 6.8 Hz); 7.12 (s, 1H); 4.59 (d, 1H,
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
J = 6.6 Hz); 3.61 (m, 2H); 3.23 (dd, 1H, J = 14.4 Hz, J = 5.5 Hz); 3.13 (m, 1H); 2.93 (dd, 1H, J =
1
3
1
4.2 Hz, J = 8.3 Hz); 2.09 (m, 1H); 1.74 (m, 1H); CꢀNMR (CDCl ): δ (ppm) 173.7; 171.8;
3
136.3; 134.6; 130.9; 128.5; 127.6; 127.3; 126.5; 122.7; 122.1; 119.6; 118.7; 112.6; 111.3; 64.3;
+
50; 42.8; 31; 27.7. HRMS calcd for C H N O [MH ]: 349.1547. Found: 349.1517
2
1
21
2
3
(2S,3R)ꢀ3ꢀ((1Hꢀindolꢀ3ꢀyl)methyl)ꢀ1ꢀ((2,4ꢀdinitrophenyl)thio)pyrrolidineꢀ2ꢀcarboxylic acid
(18): To a solution of 30 (1 mmol, 360 mg) in 2M NaOH (1 mL) and dioxane (5 mL), 2,4ꢀ
dinitrophenylsulfenyl chloride (1.1 equiv) was added dropwise over 15 min, whilst 2M NaOH (2
mL) was added dropwise with vigorous stirring. The mixture was diluted with H O (50 mL),
2
filtered from precipitated disulfide, overlayered with EtOAc and acidified with 0.5M H SO .
2
4
Upon extraction with EtOAc, the combined extracts were washed twice with H O and then dried
2
over Na SO . The crude residue was dissolved in TFA (3 mL) and the mixture was stirred at
2
4
room temperature for 4 h. TFA was coꢀevaporated with cyclohexane and the crude residue was
purified by HPLC in the same conditions as described above leading to 18 as a yellow solid, 55%
1
yield. HPLC Rt = 21.9 min. HꢀNMR (MeOD): δ (ppm) 9.13 (s, 1H); 8.46 (dd, 1H, J = 21 Hz, J
=
9.4 Hz); 7.76 (dd, 1H, J = 15.6Hz, J = 9.4 Hz); 7.26 (dd, 1H, J = 6.8 Hz, J = 7.4 Hz); 7.15 (dd,
1H, J = 14.6 Hz, J = 7.2 Hz); 6.91 (d, 1H, J = 14.8 Hz); 6.91 (s, 1H); 6.78 (d, 1H, J = 7.8 Hz);
4.41 (dd, 1H, J = 8.9 Hz, J = 9.8 Hz); 3.65 (m, 2H); 2.78 (m, 1H); 2.37 (m, 1H); 2.13 (m, 1H);
13
1
.89 (m, 2H); CꢀNMR (MeOD): δ (ppm) 186.7; 170.4; 152.4; 146.7; 134; 131; 129.6; 128.7;
126.8; 125.7; 122.6; 122.1; 119.5; 110.1; 112.6; 63.9; 45.7; 40.4; 39.6; 30.5; 29.6. HRMS calcd
+
for C H N O SNa [MNa ]: 465.0839. Found: 465.0839
2
0
19
4
6
ACS Paragon Plus Environment
29