Non-charged thiamine analogs as inhibitors of enzyme transketolase

Article abstract of DOI:10.1016/j.bmcl.2007.11.098

Inhibition of the thiamine-utilizing enzyme transketolase (TK) has been linked with diminished tumor cell proliferation. Most thiamine antagonists have a permanent positive charge on the B-ring, and it has been suggested that this charge is required for diphosphorylation by thiamine pyrophosphokinase (TPPK) and binding to TK. We sought to make neutral thiazolium replacements that would be substrates for TPPK, while not necessarily needing thiamine transporters (ThTr1 and ThTr2) for cell penetration. The synthesis, SAR, and structure-based rationale for highly potent non-thiazolium TK antagonists are presented.

Full text of DOI:10.1016/j.bmcl.2007.11.098

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 509–512
Non-charged thiamine analogs as inhibitors of enzyme transketolase
Allen A. Thomas,a,* J. De Meese,a Y. Le Huerou,a Steven A. Boyd,a
Todd T. Romoff,a Steven S. Gonzales,a Indrani Gunawardana,a Tomas Kaplan,a
Francis Sullivan,a Kevin Condroski,a Joseph P. Lyssikatos,a Thomas D. Aicher,a
Josh Ballard,a Bryan Bernat,a Walter DeWolf,a May Han,b Christine Lemieux,a
Darin Smith,a Solly Weiler,b S. Kirk Wright,b, Guy Vigersa and Barb Brandhubera
aArray BioPharma Inc., 3200 Walnut Street, Boulder, CO 80301, USA
bAVEO Pharmaceuticals, Inc., 75 Sidney Street, 4th floor, Cambridge, MA 02139, USA
Received 19 September 2007; revised 21 November 2007; accepted 27 November 2007
Available online 3 December 2007
Abstract—Inhibition of the thiamine-utilizing enzyme transketolase (TK) has been linked with diminished tumor cell proliferation.
Most thiamine antagonists have a permanent positive charge on the B-ring, and it has been suggested that this charge is required for
diphosphorylation by thiamine pyrophosphokinase (TPPK) and binding to TK. We sought to make neutral thiazolium replace-
ments that would be substrates for TPPK, while not necessarily needing thiamine transporters (ThTr1 and ThTr2) for cell penetra-
tion. The synthesis, SAR, and structure-based rationale for highly potent non-thiazolium TK antagonists are presented.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Activation of the non-oxidative branch of the pentose
phosphate pathway for ribose-5 phosphate synthesis
has been demonstrated in tumor cells.1 As described in
the preceding paper, the enzyme transketolase (TK)
has a high control coefficient toward ribose synthesis
in the pathway, suggesting a potential role in the treat-
ment of cancer.
Figure 1. Catalysis by thiamine requires thiazolium deprotonation.
Since thiamine is a cofactor of transketolase, and since
the thiamine B-ring is an essential component of cataly-
sis (see Fig. 1, for mechanistic rationale), thiamine ana-
logs possessing B-rings unable to participate in the cycle
have potential to be TK antagonists. In order to test the
hypothesis that a TK inhibitor would result in dimin-
ished tumor cell proliferation, we synthesized various
non-permanently charged thiamine analogs based on lit-
erature compounds (i.e., thiamine thiazolone, deazathi-
amine; details below). In addition, we utilized the
human crystal structure of TK (to be published else-
where) to rationally design novel, potent inhibitors.
We sought non-charged thiamine analogs based on a
hypothesis that they would have cell penetration inde-
pendent of active transport via thiamine transporters,
and that they would have improved pharmacokinetic
properties compared with permanently charged thia-
mine mimetics (see preceding article).
Although there have been previous reports of TK inhib-
itors in the literature,2 the majority of these have a per-
manently charged B-ring and require pyrophosphate to
effectively compete with TPP. It has been stated that the
permanent charge is necessary for recognition by
TPPK.3 In contrast, others have reported that a neutral
thiamine analog, thiamine thiazolone (Table 1, com-
pound 2), may be pyrophosphorylated in the cytosol.4
Considering the critical role of the pyrophosphate group
for binding to TK, this question was one for which we
sought resolution.
Keywords: Transketolase; Thiamine; Pyrophosphokinase; Deazathi-
amine; Pyrophosphate; Thiazolone; Pyrazolothiamine.
*
Corresponding author. Tel.: +1 303 386 1523; fax: +1 303 386
1420; e-mail: athomas@arraybiopharma.com
Present address: Novartis Institutes for BioMedical Research, Inc.,
250 Mass Ave, Cambridge, MA 02139, USA.
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.11.098

510
A. A. Thomas et al. / Bioorg. Med. Chem. Lett. 18 (2008) 509–512
Table 1. SAR for thiamine and thiazolone analogs 1–5
Compound
Structure
Apo-TK
HCT-116 Cell TK TPPK kcat
b
/
Km (nMÀ1 sÀ1
)
a
c
enzymatic Kd (nM) EC50 (lM)
1: Thiamine
6.8 · 106
3.4 · 104
>5.0 · 105
—
1.2 · 10À5
2: Thiamine thiazolone
2.71
3.24
1.0 · 10À5
3: N3P-TT
—
4: Thiamine pyrophosphate (TPP)
170
16
—
—
—
—
5: Thiamine thiazolone pyrophosphate (TTPP)
a Average Kd, nM, for enzymatic assay performed in triplicate. Data were within 2-fold of the mean.
b Average EC50, lM, for cellular assay performed in triplicate. Data were within 4-fold of the mean.
c Average kcat/Km, nMÀ1 À1, for enzymatic assay performed in triplicate. Data were within 2-fold of the mean.
s
To rationally design thiamine mimetics which might re-
tain TPPK activity, we evaluated the published mouse
TPPK crystal structure.5 In the crystal structure, there
are two backbone amide carbonyls pointing in the direc-
tion of the thiazolium cation. One might hypothesize
that this apparent bi-dentate interaction plays an impor-
tant role in conversion of the primary alcohol to the
pyrophosphate. To test this, we explored replacement
of the thiazolium ring with various heterocycles, includ-
ing thiazolone, pyrazole, and thiophene.
thiamine, both as the free alcohol 2 (>100·; Table 1)
and as the pyrophosphate 5 (>10·; Table 1). Also con-
sistent with the observations by Kandiko et al.,4 thia-
mine thiazolone had TPPK activity comparable to
thiamine. Furthermore, we discovered that thiazolone
analogs, including a novel aminopyridine 3 (N3P-TT),
synthesized according to Scheme 1, possessed low
Previously it was reported that replacement of thia-
mine’s B-ring with a neutral thiazolone resulted in dra-
matically greater binding to thiamine-utilizing
enzymes.6 One explanation for this increased potency
was a reduction in desolvation energy associated with
binding of a neutral, hydrophobic species (thiazolone)
versus a cationic one (thiazolium).7
From examination of the human crystal structure of
holo-TK (to be published elsewhere), we hypothesized
that removal of thiamine’s positive charge should not
diminish analog binding since contribution of the cation
to enzyme interaction appears to be negligible. Consis-
tent with the literature, we found that thiamine thiazo-
lone was a significantly better binder to TK than
Scheme 1. Synthesis of N3P-TT (3).9 Reagents and conditions: (a)
K3Fe(CN)6, NaOH; (b) s-butanol, reflux.

A. A. Thomas et al. / Bioorg. Med. Chem. Lett. 18 (2008) 509–512
Table 2. SAR for deazathiamine and related analogs 6–13
511
Compound
R
Apo-TK enzymatic
Kd (lM)
TPPK/Apo-TK
coupled assay Kd (nM)
HCT-116 Cell TK
EC50 (lM)
TPPK kcat/
Km (nMÀ1 sÀ1
a
b
c
d
)
6
7 Deazathiamine
H
CH3
>100
18
—
12
64
60
63
230
—
3.0
52
1.2 · 10À5
3.4 · 10À5
2.4 · 10À5
2.7 · 10À5
8.5 · 10À4
1.0 · 10À6
—
1.2
2.6
0.66
25
8
i-Propyl
c-Propyl
CF3
>500
>500
—
9
10
11
12
13
Ph
NH2
>500
2.2
1.1
>10
8.5
6.3
NHCOCH3
1.1 · 10À6
a Average Kd, lM, for enzymatic assay performed in triplicate. Data were within 2-fold of the mean.
b Average Kd, nM, for coupled assay from at least three experiments. Data were within 2-fold of the mean.
c Average EC50, lM, for cellular assay performed in triplicate. Data were within 4-fold of the mean.
d Average kcat/Km, nMÀ1 sÀ1, for enzymatic assay performed in triplicate. Data were within 2-fold of the mean.
micromolar cellular potency against TK in a human co-
lon cancer cell line (HCT-116).8
amines retained TPPK activity. And in the coupled assay
they demonstrated remarkable potency (Table 2: TPPK/
Apo-TK Kd of 3 nM for 13). Despite improvements in
binding to TK in enzymatic assays, cell potency relative
to the thiazolones and charged thiamine mimetics (pub-
lished in the preceding article) decreased.
A synthesis for another neutral thiamine mimetic, deaza-
thiamine (Table 2, compound 7), was previously de-
scribed in the literature,10 however activity against TK
was not reported. Using a shorter, published route toward
trisubstituted thiophenes,11 we were able to conveniently
synthesize (Scheme 2) several deazathiamine analogs that
had unprecedented in vitro activity as the free alcohols
(Table 2: Apo-TK Kd of 1.1 lM for 13). The increased
hydrophobicity of the thiophene ring versus the thiazo-
lone ring (e.g. compounds 2 and 3) may be a contributing
factor in the tighter TK binding of analogs 7, 12, and 13.12
In order to assess their potential as pyrophosphate pro-
drugs, a coupled TPPK/Apo-TK enzymatic assay was
developed.13 As with thiamine thiazolone, the deazathi-
We hypothesized that replacement of the thiamine thia-
zolium ring with a reversibly charged pyrazole ring, syn-
thesized according to Scheme 3, would maintain TPPK
substrate recognition and might allow recognition by
Scheme 3. Synthesis of pyrazolothiamines.14 Reagents and conditions:
(a) methyl hydrazine, MeOH, 60%; (b) methyl hydrazine, diethyl ether,
61%15; (c) hydrazine, MeOH, diethyl ether; (d) BnCl, NaH, DMF,
100% for steps c and d; (e) DIBAL, THF; (f) MsCl, DIEA, THF, then
LiAlH4, 0 ꢁC, 47–52% for steps e and f; (g) Br2, Na2CO3, CH2Cl2,
À10 ꢁC; (h) t-BuLi, MeOCHO, diethyl ether, À78 ꢁC, 4–41% for steps
g and h; (i) DMSO, KOtBu, t-BuOH, 55 ꢁC; (j) NaOEt, EtOH, reflux,
6–27% for steps i and j; (k) Pd black (1 equiv), 1,4-cyclohexadiene
(100 equiv), EtOH, reflux, 61–76%.
Scheme 2. Synthesis of deazathiamines. Reagents and conditions: (a)
NaHS, 3:1 toluene/water, 10 ꢁC, 1 h; (b) ethyl 3-ethoxyacrylate,
LiHMDS, rt, 18 h, 80%; (c) concd HCl, 2 h, 50%; (d) LiAlH4, THF,
reflux, 18 h, 85%; (e) MnO2, CHCl3, 77%; (f) DMSO, KOtBu,
t-BuOH, 55 ꢁC, 77%; (g) NaOEt, 65 ꢁC, 3 d, 10–50%.

512
A. A. Thomas et al. / Bioorg. Med. Chem. Lett. 18 (2008) 509–512
Table 3. SAR for pyrazolothiamine analogs 14–16
References and notes
Compound Structure
HCT-116
CellTK
TPPK kcat
/
Km (nMÀ1 sÀ1
)
b
1. (a) Cascante, M.; Centelles, J. J.; Veech, R. L.; Lee, W.-N.
P.; Boros, L. G. Nutr. Cancer 2000, 36, 150; (b) Boros, L.
G.; Torday, J. S.; Lim, S.; Bassilian, S.; Cascante, M.; Lee,
W.-N. P. Cancer Res. 2000, 60, 1183.
a
EC50 (lM)
2. (a) Du, M. X.; Sim, J.; Fang, L.; Zheng, Y.; Koh, S.;
Stratton, J.; Pons, J.; Wang, J. J.-X.; Carte, B. J. Biomol.
Screen 2004, 9, 427; (b) Solovjeva, O. N.; Kochetov, G. A.
FEBS Lett. 1999, 462, 246; (c) Boros, L. G. Cancer Res.
1998, 58, 3188; (d) Boros, L. G.; Puigjaner, J.; Cascante,
M.; Lee, W.-N. P.; Brandes, J. L.; Bassilian, S.; Yusuf, F.
I.; Williams, R. D.; Muscarella, P.; Melvin, W. S.;
Schirmer, W. J. Cancer Res. 1997, 57, 4242; (e) Kochetov,
G. A.; Izotova, A. E.; Meshalkina, L. E. Biochem.
Biophys. Res. Commun. 1971, 43, 1198.
14
15
17
—
1.1
1.9 · 10À5
16
4.7
—
3. Voskoboyev, A. I.; Ostrovsky, Y. M. Ann. N.Y. Acad. Sci.
1982, 378, 161.
4. Kandiko, C. T.; Smith, D.; Yamamoto, B. K. Biochem.
Pharmacol. 1988, 37, 4375.
5. Timm, D. E.; Liu, J.; Baker, L.-J.; Harris, R. A. J. Mol.
Biol. 2001, 310, 195.
6. Gutowski, J. A.; Lienhard, G. E. J. Biol. Chem. 1976, 251,
2863.
a Average EC50, lM, for cellular assay performed in triplicate. Data
were within 4-fold of the mean.
s
b Average kcat/Km, nMÀ1 À1, for enzymatic assay performed in trip-
licate. Data were within 2-fold of the mean.
7. Kluger, R.; Gish, G.; Kauffman, G. J. Biol. Chem. 1984,
259, 8960.
thiamine transporters leading to improved cell potency.
Also, pyrazole analogs would be amenable to further
substitutions on the B-ring including extension of func-
tionalities into the substrate-binding region of TK,
which might allow for improved potency. As with the
thiazolones and deazathiamines, the pyrazoles possessed
TPPK activity (Table 3). Unfortunately, they demon-
strated only modest activity in the tumor cell assay
(HCT-116).
8. Though it is possible that the cellular activity of non-
pyrophosphates such as compound 3, N3P-TT, is a result
of their direct interaction with TK, we believe it is much
more likely that they are first converted by TPPK inside
the cell into their pyrophosphate forms. The pyrophos-
phate binding to TK is greater by several orders of
magnitude than the corresponding non-pyrophosphate
analogs (Table 1).
9. Synthesis of starting material N3PT is presented in the
preceding publication, Thomas, et al.
Taken together, these results support our hypothesis
that non-permanently charged thiamine mimetics are
competent substrates for TPPK, and that the resulting
pyrophosphates antagonize the activity of TK in vitro.
For the first time, low micromolar non-pyrophosphate
TK inhibitors were identified in a deazathiamine series.
Despite remarkable potencies in enzymatic assays, cellu-
lar potencies were modest to poor. An explanation for
the discrepancy between the two assays is not obvious,
although one might surmise that these analogs do not
achieve sufficient intracellular concentrations to com-
pete with endogenous thiamine pyrophosphate for TK.
Future work in this area will attempt to improve cell po-
tency by modifying their physiochemical properties and
incorporating functionalities which might enable recog-
nition by thiamine transporters.
10. Hawksley, D.; Griffin, D. A.; Leeper, F. J. J. Chem. Soc.,
Perkin Trans. 1 2001, 144.
11. Fevig, T. L.; Phillips, W. G.; Lau, P. H. J. Org. Chem.
2001, 66, 2493.
12. We would expect entropy differences related to structured
water around the more polar thiazolone ring versus the
thiophene ring. These entropy differences often lead to
increased protein–ligand interactions. Andrews, P. R. In
The Practice of Medicinal Chemistry; Wermuth, C. G.,
Ed.; Academic Press: London, 1996; p 349.
13. Prior to performing the TPPK/Apo-TK assay, each
compound was evaluated for its ability to be pyrophosph-
orylated by TPPK using thiamine as a control substrate.
The extent of substrate pyrophosphorylation relative to
thiamine pyrophosphorylation was determined prior to
competition with TPP in inhibiting TK activity according
to methods described in WO2005/095391.
14. Starting material (Z)-ethyl 5-(benzyloxy)-2-[(dimethyl-
amino)methylene]-3-oxopentanoate was prepared in two
steps from ethyl acetoacetate (50% yield). J. Org. Chem.
1986, 51, 106.
Supplementary data
15. Either N-methylated pyrazole regioisomer can be prepared
in a 3:1 regioselectivity by choice of methanol or diethyl
ether as solvent Menozi, G. et al. J. Heterocycl. Chem.
1987, 24, 1669.
Procedures and analytical data are available for com-
pounds 2, 3, 5, and 6–16. Supplementary data associated
with this article can be found, in the online version, at
doi:10.1016/j.bmcl.2007.11.098.