Synthesis and biological evaluation of 9-thia-5,10-dideazafolic acid

Article abstract of DOI:10.1002/jhet.5570390541

The folate analogue, 9-thia-5,10-dideazafolic acid (3b), was obtained in an efficient two-step procedure in an overall yield of 60%. The previously unknown intermediate dimethyl-thiocarbamic acid S-(2-amino-3,4-dihydo-4-oxo-pyrido[2,3-d] pyrimidin-6-yl) e

Full text of DOI:10.1002/jhet.5570390541

†
Sep-Oct 2002
1097
Synthesis and Biological Evaluation of 9-Thia-5,10-dideazafolic Acid
Mark Wall and Stephen J. Benkovic*
The Pennsylvania State University, Department of Chemistry
414 Wartik Laboratory, University Park, PA, 16802
Received April 10, 2002
The folate analogue, 9-thia-5,10-dideazafolic acid (3b), was obtained in an efficient two-step procedure in
an overall yield of 60%. The previously unknown intermediate dimethyl-thiocarbamic acid S-(2-amino-3,4-
dihydo-4-oxo-pyrido[2,3-d]pyrimidin-6-yl) ester (5) was prepared via the condensation of 2,6-diamino-3H-
pyrimidin-4-one and S-(2-malonaldehyde)-1,1,3,3-tetramethylthiouronium bromide (4). Compound 5, in a
one pot procedure, was deprotected using sodium hydroxide and then coupled to diethyl N-[(4-
chloromethyl)benzoyl]-L-glutamate, followed by saponification of the ethyl esters to give the 9-thia-5,10-
dideazafolic acid (3b). Compound 3b was a potent inhibitor of human 5-aminoimidazole-4-carboxamide
ribonucleotide transformylase (K of 8 ± 5 µM) and showed no inhibition of human glycinamide ribonu-
i
cleotide transformylase at concentrations as high as 50 µM. Compound 3b was screened by the National
Cancer Institute Developmental Therapeutics Program against 60 human tumors and was found to be active
against a leukemia RPMI-8226 cell line where the LC was 1 µM.
50
J. Heterocyclic Chem., 39, 1097 (2002).
The discovery that folic acid analogs such as methotrex-
prepared in this series are 10-thia-5-deazafolic acid (1b)
[4], and N -substituted analogs of 5-deazaisofolic acid
(3a) and 5-deaza-5,6,7,8-tetrahydroisofolic acid (2b) [5].
9
ate are useful therapeutic agents in the treatment of neo-
plastic and inflammatory diseases has prompted the syn-
thesis of numerous folate derivatives. Of particular inter-
est are derivatives of 5-deazafolic acid (1a), which have
shown antitumor activity via the selective inhibition of
glycinamide ribonucleotide transformylase (GAR Tfase),
the first of two enzymes requiring folate in de novo purine
biosynthesis. One of the more promising of these analogs,
5,10-dideaza-5,6,7,8-tetrahydrofolic acid (lometrexol)
(2a), showed broad spectrum antitumor activity, was
active against methotrexate resistant cell lines and is now
in clinical trials [1-3]. Other analogs that have been
The most potent of these compounds (2b) had an IC of 1
50
µM against MCF-7 cells, which is 100 fold less potent than
2a. In an effort to further explore the therapeutic potential
of 5-deaza analogues of folic acid we have prepared and
determined the biological properties of 9-thia-5,10-
dideazafolic acid (3b).
The synthetic route used to prepare 9-thia-5,10-dideaza
folic acid is shown in Scheme 1. The compound was pre-
pared in a very efficient manner from readily available
intermediates in two steps in overall yield of 60%. The key
starting material, S-(2-malonaldehyde)-1,1,3,3-tetram-
ethylthiouronium bromide (4), obtained from the reaction
of 1,1,3,3-tetramethylthiourea and 2-bromomalonalde-
hyde, crystallized from solution as the analytically pure
bromide salt. The use of tetramethylthiourea was neces-
sary in order to prevent intramolecular cyclization to the
thiazole [6]. The synthesis of 6-substituted-5-deaza-
purines via the condensation of 2,6-diamino-3H-pyrim-
idin-4-one with malonaldehydes has been reported for
2-nitromalonaldehyde [7], triformylmethane (2-formyl-
malonaldehyde) [8] and 2-bromomalonaldehyde [9]. In a
similar manner malonaldehyde 4 condensed readily with
2,6-diamino-3H-pyrimidin-4-one under acidic conditions
to give compound 5 in 85% yield. This is the first example
of a thiomalonaldehyde equivalent in this reaction, which
under aqueous acid conditions the tetramethylthiouronium
group is hydrolyzed to the N,N-dimethylcarbamate pro-
tected thiol. Compound 5 was very insoluble and we were
not able to find a suitable solvent for recrystallization,
however, the compound was of sufficient purity for use in
the next step. In a one-pot procedure, the dimethylcarba-
mate group of 5 was first hydrolyzed in refluxing 1 N
sodium hydroxide to give the sulfide, which was then

1098
M. Wall and S. J. Benkovic
Vol. 39
13
2H), 2.97 (s, 12H); C NMR (D O): δ 186.6, 176.4, 107.2, 43.4;
coupled to 6, followed by saponification of the ethyl esters
and acidification of the reaction mixture to give the target
compound, 10-thia-5,10-dideazafolic acid (3b), in 73%
yield. Attempts to reduce the ring using catalytic hydro-
genation over platinum oxide or 10 % palladium on carbon
were unsuccessful.
2
+
MALDI ms: m/z 203 (M ).
Anal. Calcd. for C H N O SBr: C, 33.93; H, 5.34; N, 9.89.
8
15 2 2
Found: C, 34.14; H, 5.48; N, 9.66.
Dimethyl-thiocarbamic Acid S-(2-Amino-3,4-dihydo-4-oxo-
pyrido[2,3-d]pyrimidin-6-yl) Ester (5).
The compound (3b) was assayed for inhibition of both
folate requiring enzymes in de novo purine biosynthesis:
GAR Tfase and 5-aminoimidazole-4-carboxamide ribonu-
cleotide transformylase (AICAR Tfase). Compound 3b
A mixture of 0.6 g of (4.7 mmol) of 2,6-diamino-3H-pyrim-
idin-4-one and 1.4 g (4.9 mmol) of S-(2-malonaldehyde)-1,1,3,3-
tetramethylthiouronium bromide in 30 mL of ethanol and 5 mL
of 12 N hydrochloric acid was refluxed for 1 hour. The solution
was cooled to 25 °C, concentrated and the residue triturated with
5% sodium hydrogencarbonate. The precipitate was collected by
filtration, washed with water and ethanol, and dried under vac-
uum at 60 °C to give 1.1 g (85 %) of a yellow-brown solid, mp
>250 °C. This compound could not be purified and fully charac-
terized due to its insolubility, however it was of sufficient purity
was a potent inhibitor of human AICAR Tfase (K of 8 ± 5
i
µM) with an affinity 5 times greater than its cofactor 10-
formyltetrahydrofolate [10]. However, there was no inhi-
bition of human GAR Tfase at concentrations as high as 50
µM. Compound 3b was screened by the National Cancer
Institute Developmental Therapeutics Program against 60
human tumor cell lines. The compound was found to be
most active against a leukemia RPMI-8226 cell line where
1
for use in the preparation of 3b, H NMR (DMSO): δ 11.61 (br s,
1H, N-H), 8.51 (d, 1H, H-7, J = 2.5 Hz), 8.15 (d, 1H, H-5,
7,5
J
= 2.5 Hz), 7.08 (br s, 2H, NH ), 3.07 (s, 3H), 2.94 (s, 3H);
5,7
13
2
C (d-TFA): δ 169.98, 162.13, 157.28, 156.68, 154.44, 152.67,
the LC was 1 µM.
50
+
125.96, 116.28, 39.23, 38.78; FAB ms: m/z 266 (MH ).
Diethyl N-[(4-Chloromethyl)benzoyl]-L-glutamate (6).
EXPERIMENTAL
This compound was prepared in a manner similar to that for the
preparation of diethyl N-[(4-bromomethyl)benzoyl]-L-glutamate
[12]. To a solution of 0.9 g (4.8 mmol) of 4-(chloromethyl)ben-
zoyl chloride in 2 mL of dichloromethane, cooled to 0 °C, was
added dropwise a solution of 1.2 g (5.0 mmol) of diethyl L-gluta-
mate hydrochloride and 1.4 mL (10 mmol) of triethylamine in 3
mL of dichloromethane. The mixture was stirred for 1 hour at 0
°C and then 2 hours at 25 °C. The solution was diluted with
dichloromethane and washed with 0.1 N hydrochloric acid and
then brine. The solution was dried (sodium sulfate) and evapo-
Melting points were taken on a Mel-Temp II in an open capil-
1
13
lary and are uncorrected. The H and C NMR spectra were
taken on a Bruker AMX-360 spectrophotometer, MALDI and
APCI mass spectra were obtained on a Perseptive Biosystems
(Framingham, MA), and FAB mass spectra on a Kratos
Analytical (Manchester, UK). Elemental analysis was performed
by Atlantic Microlab, Inc., Norcross, Georgia.
S-(2-Malonaldehyde)-1,1,3,3-tetramethylthiouronium Bromide (4).
1
To a solution of 1.5 g (10 mmol) of 2-bromomalonaldehyde
[11] in 20 mL of acetone at 25 °C was added 1.3 g (10 mmol) of
1,1,3,3-tetramethylthiourea. The solution was stirred at 25 °C
and after approximately 5 minutes a precipitate was observed.
Stirring was continued for an additional 30 minutes and the white
precipitate was collected by filtration and washed with acetone to
rated to give 1.5 g (85%) of a white solid, mp 110 °C; H NMR
(CDCl ): δ 7.80 (d, 2H, 2',6', J = 8.2 Hz), 7.45 (d, 2H, 3',5', J =
3
o
o
8.2 Hz), 7.05 (d, 1H, NH, J = 9.8 Hz), 4.76 (m, 1H, glutamate α
H), 4.59 (s, 2H, CH Cl), 4.22 (m, 2H, -OCH CH ), 4.09 (m, 2H,
2
2
3
-OCH CH ), 2.10-2.58 (m, 4H, glutamate β and γ H's), 1.28 (t,
2
3
3H, -OCH CH J = 7.3 Hz), 1.20 (t, 3H, -OCH CH ,J = 7.3 Hz);
2
3
2
3
1
13
give 2.7 g (96%) of 4, mp 147-148 °C; H NMR (D O): δ 8.66 (s,
C NMR (CDCl ): δ 173.27, 171.87, 166.43, 141.05, 133.56,
2
3

Sep-Oct 2002
Synthesis and Biological Evaluation of 9-Thia-5,10-dideazafolic Acid
1099
REFERENCES AND NOTES
128.66, 127.52, 61.71, 60.81, 52.40, 45.31, 30.44, 27.05, 14.11,
14.08; APCI ms: m/z 356 (MH ).
+
Anal. Calcd. for C H NO Cl: C, 57.38; H, 6.23; N, 3.94.
Found: C, 57.23; H, 6.15; N, 4.05.
17 22
5
*
To whom correspondence should be addressed. Phone
814-865-2882. FAX 814-865-2973.
9-Thia-5,10-dideazafolic Acid (3b).
[1] E. C. Taylor, P. J. Harrington, S. R Fletcher, G. P
Beardsley, and R. G. Moran, J. Med. Chem., 23, 921 (1985).
[2] G. P Beardsley, E. C. Taylor, G. B. Grindley, and R. G.
Moran, Chemistry and Biology of Pteridines. B. A. Cooper and E. A.
Whitehorn, eds, New York, NY, Walter de Gruyter. 953, 1986.
[3] S. Laohavinij, S. R. Wedge, M. J. Lind, N. Bailey, A.
Humphreys, M. Proctor, F. Chapman, D. Simmons, A. Oakley, L.
Robson, L. Gumbrell, G. A. Taylor, H. D. Thomas, A. V. Boddy, D.
R. Newell and A. H. Calvert, Invest. New Drugs, 14, 325 (1996).
[4] S. K. Singh and J. B. Hynes, J. Heterocyclic Chem., 28,
977 (1991).
A solution of 0.27 g (1.0 mmol) of 5 in 2.8 mL of degassed 1 N
sodium hydroxide was refluxed for 1.5 hours. The solution was
cooled to 25 °C and a solution of 0.39 g (1.1 mmol) of diethyl N-
[(4-chloromethyl)benzoyl]-L-glutamate in 1.5 mL of ethanol was
added. The reaction mixture was stirred for 20 minutes at 25 °C
and then 1.5 mL of 1 N sodium hydroxide was added and stirring
continued for 10 hours. The pH was adjusted to 3 with 6 N
hydrochloric acid and the resulting precipitate was collected by
filtration and washed with water, ethanol, and diethyl ether and
dried under vacuum at 60 °C to give 0.33 g (73 %) of 3b, mp
[5] S. K. Singh, I. K. Dev, D. S. Duch, R. Ferone, G. K. Smith,
J. H. Freisheim and J. B. Hynes, J. Med. Chem., 34, 606 (1991).
[6] P. V. Plazzi, F. Bordi, C. Silva, G. Morini and P. L.
Catellani, Farmaco, 44, 1011 (1989).
[7] R. Bernetti, F. Mancini and C. C. Price, J. Org. Chem., 27,
2863 (1962).
>250 °C; uv: (pH 7.5): λ max 285 nm (ε 17,600), 333 nm (ε
1
4,700); H NMR (0.1 N NaOD): δ 8.31 (d, 1H, H-7, J = 2.3
7,5
Hz), 8.28 (d, 1H, H-5, J = 2.3 Hz), 7.70 (d, 2H, J = 6.2 Hz),
5,7
7.28 (d, 2H, J = 6.2 Hz), 4.33 (dd, 1H, J = 9.0 Hz, J = 4.5 Hz),
4.16 (s, 2H), 2.32 (dd, 2H, J = 7.9 Hz, J = 7.9 Hz), 2.12 (m, 2H);
13
C NMR (0.1 N NaOD): δ 182.66, 179.26, 174.61, 171.39,
[8] C. Temple, Jr., R. D. Elliot, and J. A. Montogomery, J.
Org. Chem., 47, 761 (1982).
160.97, 158.32, 142.27, 141.71, 132.70, 129.58, 127.73, 122.25,
112.28, 56.35, 40.37, 34.62, 28.73; MALDI ms: m/z 458 (MH ).
+
[9] E. C. Taylor and C. Yoon, Syn. Commun., 18, 1187 (1988).
[10] M. Wall, J. H. Shim and S. J. Benkovic, Biochemistry, 39,
11303 (2000).
[11] S. Trofimenko, J. Org. Chem., 28, 3243 (1963).
[12] S. J. Yan, L. T. Weinstock and C. C. Cheng, J.
Heterocyclic Chem., 16, 541 (1979).
Anal. Calcd. for C H N O S•1.3H O: C, 49.95; H, 4.53; N,
14.56. Found: C, 50.19; H, 4.56; N, 14.26.
20 19
5
6
2
Acknowledgement.
PHS Grant GM24129-21 from the National Institute of Health
(S.J.B.) supported this work.