Identification of potent type I MetAPs inhibitors by simple bioisosteric replacement. Part 2: SAR studies of 5-heteroalkyl substituted TCAT derivatives

Article abstract of DOI:10.1016/j.bmcl.2005.06.005

Systematic SAR studies on the thiazole ring 5-substituent of TCAT derivatives revealed that the introduction of a β-alkoxy or an amino group enhanced the inhibitory activity significantly. The present compounds are representative of specific Co(II)-MetAP1 inhibitors. Before the physiologically relevant metal ions for MetAPs are established, these small molecular compounds could be used as tools for detailed biological studies.

Full text of DOI:10.1016/j.bmcl.2005.06.005

Bioorganic & Medicinal Chemistry Letters 15 (2005) 4130–4135
Identification of potent type I MetAPs inhibitors by simple
bioisosteric replacement. Part 2: SAR studies
of 5-heteroalkyl substituted TCAT derivatives
Yong-Mei Cui,a Qing-Qing Huang,a Jie Xu,a Ling-Ling Chen,a Jing-Ya Li,a
Qi-Zhuang Ye,a,b Jia Lia,* and Fa-Jun Nana,*
aChinese National Center for Drug Screening, Shanghai institute of Materia Medica, Shanghai Institutes for Biological
Sciences, Graduate School of the Chinese Academy of Sciences, Chinese Academy of Sciences, 189 Guoshoujing Road,
Zhangjiang Hi-Tech Park, Shanghai 201203, PR China
bHigh-Throughput Screening Laboratory, University of Kansas, Lawrence, KS 66047, USA
Received 17 March 2005; revised 27 May 2005; accepted 2 June 2005
Available online 6 July 2005
Abstract—Systematic SAR studies on the thiazole ring 5-substituent of TCAT derivatives revealed that the introduction of a b-alk-
oxy or an amino group enhanced the inhibitory activity significantly. The present compounds are representative of specific Co(II)-
MetAP1 inhibitors. Before the physiologically relevant metal ions for MetAPs are established, these small molecular compounds
could be used as tools for detailed biological studies.
Ó 2005 Elsevier Ltd. All rights reserved.
The methionine aminopeptidases (MetAPs) are a novel
class of dinuclear metalloprotease responsible for
removal of the initiator N-terminal methionine residue
of nascent proteins.1 It is widely found in prokaryotic
and eukaryotic cells and exists in two forms: type I
(MetAP1) and type II (MetAP2).2 The removal of
methionine represents a critical step in the maturation
of proteins for proper function, targeting, and eventual
degradation.3–6 MetAPs present good targets for new
antibiotic drug discovery because of their important
physiological functions.7–9 Moreover, MetAPs have
been shown, biochemically and structurally, to be the
molecular target of the antiangiogenesis agent fumagil-
lin and its derivatives.10 And inhibitors of MetAPs offer
hope as new treatments for bacterial and fungal infec-
tions and cancers.11
compounds demonstrated that the introduction of
b-methoxy to the 5-alkyl-substituted compounds
improved the inhibitory activity against the enzymes
dramatically. These series of b-methoxy-containing
compounds interested us particularly, because of their
prominent potency against EcMetAP1, as well as
the structural feature of b-methoxy-containing alkyl
substituents at the 5-position, which stimulated us to
investigate the effect of the heteroalkyl group at the
5-position on the inhibition of MetAP1s. In this study,
we report the synthesis and evaluation of a series of
heteroalkyl-containing TCAT derivatives (Fig. 1).
Initially, we synthesized a series of b-alkoxy-containing
compounds from the corresponding a,b-unsaturated
aldehyde using a similar method with the syntheses of
In the preceding paper, we obtained a new series of
potent MetAP1 inhibitors through simple bioisosteric
replacement from the PCAT series of compounds.12 Pre-
liminary systematic SAR studies of these TCAT series
Keywords: Type I MetAPs inhibitors; Bioisosteric replacement.
*
Corresponding authors. Tel.: +86 21 50800954; fax: +86 21 50800721
(F.-J.N.); tel.: +86 21 50801313x132; fax: +86 21 50800721 (J.L.);
e-mail addresses: fjnan@mail.shcnc.ac.cn; jli@mail.shcnc.ac.cn
Figure 1.
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.06.005

Y.-M. Cui et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4130–4135
Table 1. Inhibition of EcMetAP1 and ScMetAP1a
4131
3a and 3b. As shown in Scheme 1, the Michael addition
reaction took place in the first Darzens reaction of ethyl
dichloroacetate and the a,b-unsaturated aldehyde with
NaOMe or NaOEt in Et2O, and the resulted intermedi-
ates reacted with thiourea to form 2-aminothiazole-4-car-
boxylate 4a, 4c–g.13 Replacement of the 2-amino-group
with hydrogen and further basic hydrolysis gave 6a, 6c–
g, followed by condensation with 2-aminothiazole in the
presence of EDC in DMF afforded 7a, 7c–g. Similarly
compound 7b was synthesized from 3-benzyloxy-
propionaldehyde.
Compound
R
IC50 (lM)
EcMetAP1 ScMetAP1
1b
2
—
—
5.0 0.8
0.11 0.02
7.0 0.1
2.26 0.38
3a
3b
0.074 0.008
0.089 0.006
0.88 0.08
0.50 0.05
As shown in Table 1, all 7-alkoxy derivatives 7a–g
showed good inhibition of EcMetAP1 with IC50 values
less than 100 nM except 7a. In particular, the ethoxy
derivative 7e exhibited potent inhibition activity for
7a
7b
0.15 0.02
0.09 0.02
6.43 0.57
4.25 1.47
EcMetAP1 (IC50 = 28 nM) and ScMetAP1 (IC50
=
150 nM). Although the introduction of methoxy (7a)
or benzyloxy (7b) to the b-position had little effect on
the activity against the enzymes, the effect of adding a
methyl group to the end of 7a was striking (3a vs 7a).
However, an additional methyl group (7c) decreased
the ScMetAP1 activity and improved the EcMetAP1
selectivity. Both straight chain and branched chain alkyl
derivatives (7d and 3b, respectively) showed good activ-
ity against two enzymes. In addition, changing the meth-
oxy to ethoxy increased the activity (7e vs 3b). The
results may suggest that besides the importance of the
oxygen of the alkoxy, these alkyl groups are reaching
out to fill a hydrophobic pocket on the active site of
the enzymes. This hydrophobic pocket was able to
accommodate medium-sized alkyl groups such as isobu-
tyl and cyclohexyl (7f and g, respectively).
7c
7d
7e
0.069 0.005
0.10 0.01
3.21 0.47
0.45 0.05
0.15 0.02
0.028 0.001
7f
0.09 0.01
0.50 0.07
0.40 0.01
7g
0.054 0.010
7h
7i
0.043 0.002
0.023 0.002
0.44 0.06
0.31 0.07
On the basis of the above analyses, we became interested
in the syntheses of analogues containing b-methoxy and
aryl groups at the 5-position of TCAT. However, at-
tempts to prepare analogous compounds by the conden-
sation of ethyl dichloroacetate with cinnamaldehyde
7j
0.035 0.006
0.057 0.003
0.57 0.08
0.39 0.08
7k
7l
0.037 0.008
0.034 0.002
0.033 0.002
0.34 0.05
0.27 0.01
0.57 0.13
7m
7n
7o
0.09
0
0.52 0.08
7p
0.17 0.01
0.097 0.020
0.061 0.008
0.036 0.004
0.066 0.004
0.23 0.05
7.53 1.22
0.75 0.04
0.45 0.02
0.45 0.03
7q
7r
7h(R)
7h(S)
Scheme 1. Reagents: (a) NaOMe or NaOEt, Et2O; (b) NH2CSNH2,
MeOH, reflux; (c) NaNO2, H3PO2; (d) LiOH, MeOH–H2O; (e) 2-
aminothiazole, DCC, HOBt, DMF.
a Assays were performed as previously described.12a
b See Ref. 12a.

4132
Y.-M. Cui et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4130–4135
were unsuccessful, and an unexpected mixture of prod-
ucts was obtained. Therefore, b-methoxy aldehyde was
first prepared, as shown in Scheme 2. Treatment of the
corresponding aldehyde with allylmagnesium bromide
gave homoallylic alcohol 8h–p in high yields. This prod-
uct was protected as the methyl ether 9h–p, which was
then converted into the corresponding aldehyde 10h–p
by ozonolysis in MeOH–CH2Cl2 solution followed by
reductive treatment with Me2S. The resulting b-methoxy
aldehydes were further converted to the corresponding
compounds 7h–p by using the same procedure shown
in Scheme 1.
Scheme 3. Reagents: (a) LiOH, MeOH–H2O; (b) 2-aminothiazole,
DCC, HOBt, DMF.
product was protected as methyl ether 14R and then
desilicated by TBAF in THF giving the alcohol 15R.
Swern oxidation of 15R gave (R)-3-methoxy-3-phenyl-
propionaldehyde 16R, which could be converted into
7h(R) using the method described in Scheme1. Com-
pound 7h(S) was obtained similarly, except employing
D-(À)-diethyl tartrate as the chiral ligand.
As expected, all these nine compounds (7h–p) showed
good inhibitory activity against both EcMetAP1 and
ScMetAP1. The effects of substitution on the phenyl
group derived are summarized in Table 1. In general,
substituents on the phenyl ring are well tolerated,
regardless of their electronic properties, and a number
of very potent derivatives were obtained with IC50 val-
ues less than 50 nM against EcMetAP1. Except 7p, the
Cl- and F-substituted derivatives (7k–m, 7n–o) showed
good inhibitory activity against EcMetAP1. Although
not so obvious, all these compounds showed good inhi-
bition activity for ScMetAP1 (IC50 < 0.57 lM).
The results are shown in Table 1, from which we can see
that there is a nearly 2-fold activity difference between
the two stereoisomers of 7h(R) and 7h(S) against EcMe-
tAP1, although they showed similar activity against
ScMetAP1. These differences may reflect the subtle dif-
ferences in the active sites of EcMetAP1 and ScMetAP1.
We also synthesized two typical a-methoxy-containing
derivatives 7q and r. Both could be prepared by the
corresponding brominated ester with replacement by
methoxide anion, hydrolysis, and condensation with 2-
aminothiazole (Scheme 3). As shown in Table 1, 7q
showed similar activity against the enzymes to 7a, and
7r showed weaker activity than 7h, which showed that
the b-methoxy derivatives may fit better in the pockets
of the enzymes.
To understand the subtle differences between the
enzymes and obtain some more potent and selective
inhibitors, we considered other hydrogen-bonding deriv-
atives. Starting from compound 8a, a hydrogen-bonding
acceptor alcohol, we synthesized a series of its deriva-
tives, such as the ester, sulfonamide and the results are
shown in Table 2. Compounds 8b–h were obtained by
acylation of 8a, which was obtained by debenzylation
of 7b with the corresponding carboxylic chlorides under
basic conditions (Scheme 5).
To understand further the subtle stereochemical require-
ment of protein to small molecular inhibitors, because
its X-ray structure is currently unavailable, we synthe-
sized the enantiomers of 7h, as shown in Scheme 4.
The optical isomers of 3-methoxy-3-phenylpropionalde-
hyde 16 were prepared from cinnamyl alcohol in six
steps. Catalytic asymmetric epoxidation of cinnamyl
alcohol using L-(+)-diethyl tartrate (DET) gave (À)-
(2S,3S)-2,3-epoxycinnamyl alcohol 11R.14 Regioselec-
tive reduction of 11R with Red-Al in DME gave
(R)-3-phenyl-1,3-dihydroxypropane 12R in high yield.15
Treatment of the crude diol 12R with 1.1 equiv
TBDMSCl led to monosilicified compound 13R. This
Except for compounds 8c and 8f, most of these com-
pounds showed moderate or weak activity against
ScMetAP1, and were selective EcMetAP1 inhibitors.
Although 8a showed weak activity against the enzymes,
its ester derivatives showed increased EcMetAP1 activi-
ty with increasing bulk of the end substituents (8a–e),
especially the aryl containing group (8d and 8e). Howev-
er, insertion of an ether and a methylene spacer (8f) or
increasing the bulk of the phenyl ring (8g) reduced the
EcMetAP1 activity. In addition, sulfonylation of the
alcohol oxygen (8h) dramatically reduced the EcMe-
tAP1 activity.
Scheme 2. The syntheses of compounds 7h–p.

Y.-M. Cui et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4130–4135
4133
Scheme 4. The syntheses of enantiomers of 7h.
Table 2. Inhibition of EcMetAP1 and ScMetAP1a
Compound
R
IC50 (lM)
EcMetAP1
ScMetAP1
2.26 0.38
10.39 4.78
Scheme 5. The syntheses of 8a–h.
2
—
0.11 0.02
0.47 0.10
8a
Finally, we synthesized several amino-containing deriv-
atives. As shown in Scheme 6, compounds 9a–d were
prepared by condensation of the anion of a 2-amino-
thiazole with the corresponding 5-substituted-thiazole-
4-carboxylic acid methyl ester in THF.
8b
8c
8d
0.31 0.06
0.14 0.01
17.92 3.31
0.80 0.05
3.08 0.43
0.041 0.011
As shown in Table 2, all these four compounds 9a–d
showed good inhibition of EcMetAP1 with IC50 values
less than 60 nM. The introduction of pyrrolidinyl (9a)
or piperidinyl (9b) to the a-position of 5-methyl of
TCAT dramatically increased the activity, especially
for compound 9b, which is the best inhibitor described
in the literature to date with IC50 values of 10 nM for
EcMetAP1 and 75 nM for ScMetAP1. However, inser-
tion of an additional methylene spacer (9c) or oxygen
atom to the piperidine (9d) decreased the activity. This
may be due to the appropriate size and shape of the
8e
8f
0.033 0.011
0.13 0.01
6.04 0.51
0.55 0.08
8g
0.16 0.01
1.56 0.43
8h
9a
1.53 0.11
3.72 0.26
0.025 0.001
0.35 0.007
9b
9c
9d
0.010 0.001 0.075 0.007
0.045 0.005
0.055 0.003
0.41 0.04
1.76 0.28
Scheme 6. Reagents and conditions: (a) NaH, 2-aminothiazole; THF,
reflux.
a Assays were performed as previously described.12a

4134
Y.-M. Cui et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4130–4135
pockets in which MetAP1 accommodates its inhibitor.
Therefore, we did not attempt to prepare analogues con-
taining substituents of increased chain length.
still depend on the exact match of in vitro data and
their in vivo antibacterial activity.
Although the in vitro potency could not match the in
vivo activity, the present compounds are representative
of specific Co(II)–MetAP1 inhibitors. Before the physi-
ologically relevant metal ions for MetAPs are estab-
lished, these small molecular compounds could be used
as tools for detailed biological studies.
Of the above TCAT inhibitors, we tested several typi-
cal compounds for antibacterial activity against Staph-
ylococcus aureus, Enteropathogenic Escherichia coli,
and Pseudomonas aeruginosa, and the activity is poor
(MIC > 16 lg/mL, data not shown). We also tested
the compounds more active in inhibiting ScMetAP1
for antifungal activity. Most of the inhibitors tested
showed no antifungal activity against Candida albicans,
Aspergillus niger, Trichophyton rubrum, Saccharomyces
cerevisiae, Epidermophyton floccosum, and Microsporum
canis (MIC > 128 lg/mL, data not shown). Although
we have obtained very potent inhibitors of S. cerevisiae
MetAP1 through rational structural modification, it
seemed that there was little relationship between the
in vitro inhibitory potencies for the Co(II)–ScMetAP1
and in vivo efficacy in antifungal activity. One possibil-
ity is that the poor antifungal activity may be derived
from poor penetration of the fungus wall. However,
in the case of these in vitro Co(II)–ScMetAP1 inhibi-
tors, the lack of antifungal activity might also be relat-
ed to the metal ion present in physiological situations.
The physiological metal ions for MetAPs have not
been established and are controversial at the moment.
MetAPs of bacteria and yeast have been categorized
as cobalt-dependent metalloenzymes, based on the
observations that the purified enzymes show highest
activity in the presence of cobalt as compared with that
of other divalent metal ions.2 A more recent study
showed that Zn(II) was a superior cofactor to Co(II)
for yeast MetAP1 because Co(II) did not stimulate
yeast MetAP1 activity in the presence of physiological
concentrations of reduced glutathione.16 Another study
demonstrated that the physiologically relevant metal
ion for E. coli MetAP1 was probably Fe(II), on the ba-
sis of whole cell metal analyses.17 Most recently, DÕsou-
za et al. showed the kinetic and structural
characterization of manganese-loaded MetAPs from
E. coli and the hyperthermophilic archaeon Pyrococcus
furiosis, and implicated manganese as a metal cofactor
for MetAPs.18 Identification of physiological metal
cofactors for MetAPs is critical for discovery of small
molecule therapeutic inhibitors because their potency
may vary for different metal ions. In a study on yeast
MetAP1 that argues for Zn(II) as the cofactor, manga-
nese is also shown to induce enzyme activity both in
the absence and in the presence of glutathione. Manga-
nese is shown to be concentrated in bacteria over-
expressing E. coli MetAP1 by a factor of 2.2, similar
to the increase of iron, and it also induced E. coli Me-
tAP1 enzyme activity.2 More recently, Ye et al. report-
ed the discovery and characterization of two groups of
potent and highly metallo form-selective inhibitors of
the Co(II)-form, and of the Mn(II)-form, and an X-
ray structure of a di-Mn(II)-form of E. coli MetAP
complexed with the Mn(II)-form selective inhibitor
was also obtained.19 These results partially supported
the contention that Mn(II) ions could be physiological
metal cofactors for MetAPs. However, the relevant de-
tails of physiological metal cofactors for MetAPs will
In summary, we obtained a new series of potent Me-
tAP1 inhibitors through simple bioisosteric replacement
from the PCAT series of compounds. The detailed SAR
showed that these TCAT series of compounds showed
different activity and selectivity compared with the cor-
responding PCAT compounds. These differences may
reflect subtle differences in the active sites of EcMetAP1
and ScMetAP1. Further efforts in modifying these and
other lead structures with the aim of improving potency
as well as specificity in vitro, and efficacy in vivo, are in
progress.
Acknowledgments
This work was supported by the National Natural Sci-
ence Foundation of China Grants 30271528 (F.-J.N.),
the Qi Ming Xing Foundation of Shanghai Ministry of
Science and Technology Grant 02QB14013 (F.-J.N.).
References and notes
1. Bradshaw, R. A.; Brickey, W. W.; Walker, K. W. Trends
Biochem. Sci. 1998, 23, 263.
2. Arfin, S. M.; Kendall, R. L.; Hall, L.; Weaver, L. H.;
Stewart, A. E.; Matthews, B. W.; Bradshaw, R. A. Proc.
Natl. Acad. Sci. U.S.A. 1995, 92, 7714.
3. Lowther, W. T.; Matthews, B. W. Chem. Rev. 2002, 102,
4581.
4. Prchal, J. T.; Cashman, D. P.; Kan, Y. W. Proc. Natl.
Acad. Sci. U.S.A. 1986, 83, 24.
5. Boissel, J. P.; Kasper, T. J.; Shah, S. C.; Malone, J. I.;
Bunn, H. F. Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 8448.
6. Tso, J. Y.; Hermodson, M. A.; Zalkin, H. J. Biol. Chem.
1982, 257, 3532.
7. Chang, S. Y.; McGary, E. C.; Chang, S. J. Bacteriol. 1989,
171, 4071.
8. Miller, C. G.; Kukral, A. M.; Miller, J. L.; Movva, N. R.
J. Bacteriol. 1989, 171, 5215.
9. Li, X.; Chang, Y. H. Proc. Natl. Acad. Sci. U.S.A. 1995,
92, 12357.
10. (a) Griffith, E. C.; Su, Z.; Niwayama, S.; Ramsay, C. A.;
Chang, Y. H.; Liu, O. J. Proc. Natl. Acad. Sci. U.S.A.
1998, 95, 15183; (b) Sin, N.; Meng, L.; Wang, M. Q.; Wen,
J. J.; Bornmann, W. G.; Crews, C. M. Proc. Natl. Acad.
Sci. U.S.A. 1997, 94, 6099.
11. Vaughan, M. D.; Sampson, P. B.; Honek, J. F. Curr. Med.
Chem. 2002, 9, 385.
12. (a) Luo, Q. L.; Li, J. Y.; Liu, Z. Y.; Chen, L. L.; Li, J.;
Qian, Z.; Shen, Q.; Li, Y.; Livingston, G. H.; Ye, Q. Z.;
Nan, F. J. J. Med. Chem. 2003, 46, 2631; (b) Luo, Q. L.;
Li, J. Y.; Liu, Z. Y.; Chen, L. L.; Li, J.; Ye, Q. Z.; Nan, F.
J. Bioorg. Med. Chem. Lett. 2005, 15, 635; (c) Luo, Q. L.;
Li, J. Y.; Chen, L. L.; Li, J.; Ye, Q. Z.; Nan, F. J. Bioorg.
Med. Chem. Lett. 2005, 15, 639.

Y.-M. Cui et al. / Bioorg. Med. Chem. Lett. 15 (2005) 4130–4135
4135
13. Barton, A.; Breukelman, S. P.; Kaye, P. T.; Denis
Meakins, G.; Morgan, D. J. J. Chem. Soc. Perkin. Trans.
1 1982, 159.
14. Gao, Y.; Hanson, R. M.; Klunder, J. M.; Ko, S. Y.;
Masamune, H.; Sharpless, K. B. J. Am. Chem. Soc. 1987,
109, 5765.
16. Walker, K. W.; Bradshaw, R. A. Protein Sci. 1998, 7,
2684.
17. DÕsouza, V. M.; Holz, R. C. Biochemistry 1999, 38, 11079.
18. DÕsouza, V. M.; Swierczek, S. I.; Cosper, N. J.; Meng, L.;
Ruebush, S.; Copik, A. J.; Scott, R. A.; Holz, R. C.
Biochemistry 2002, 41, 13096.
15. Gao, Y.; Sharpless, K. B. J. Org. Chem. 1988, 53,
4081.
19. Ye, Q. Z.; Xie, S. X.; Huang, M.; Huang, W. J.; Lu, J. P.;
Ma, Z. Q. J. Am. Chem. Soc. 2004, 126, 13940.