Preliminary in vitro studies on two potent, water-soluble trimethoprim analogues with exceptional species selectivity against dihydrofolate reductase from Pneumocystis carinii and Mycobacterium avium

Article abstract of DOI:10.1016/j.bmcl.2003.12.103

2,4-Diamino-5-[3′,4′-dimethoxy-5′-(5-carboxy-1-pentynyl)] benzylpyrimidine (6) and 2,4-diamino-5-[3′,4′-dimethoxy-5′-(4- carboxyphenylethynyl)benzylpyrimidine (7) were synthesized from 2,4-diamino-5-(5′-iodo-3′,4′-dimethoxybenzyl)pyrimidine (9) via a Sonogashira reaction with appropriate acetylenic esters followed by saponification, and were tested as inhibitors of dihydrofolate reductase (DHFR) from Pneumocystis carinii (Pc), Toxoplasma gondii (Tg), Mycobacterium avium (Ma), and rat in comparison with the widely used antibacterial agent 2,4-diamino-5-(3′,4′,5′-trimethoxybenzyl)pyrimidine (trimethoprim, TMP). The selectivity index (SI) for each compound was calculated by dividing its 50% inhibitory concentration (IC₅₀) against rat DHFR by its IC₅₀ against Pc, Tg, or Ma DHFR. The IC ₅₀ of 6 against Pc DHFR was 1.0 nM, with an SI of 5000. Compound 7 had an IC₅₀ of 8.2 nM against Ma DHFR, with an SI of 11000. By comparison, the IC₅₀ of TMP was 12000 nM against Pc, 300 nM against Ma, and 180000 against rat DHFR. The potency and selectivity values of 6 and 7 were not as high against Tg as they were against Pc or Ma DHFR, but nonetheless exceeded those of TMP. Because of the outstanding selectivity of 6 against Pc and of 7 against Ma DHFR, these novel analogues may be viewed as promising leads for further structure-activity optimization.

Full text of DOI:10.1016/j.bmcl.2003.12.103

Bioorganic & Medicinal Chemistry Letters 14 (2004) 1811–1815
Preliminary in vitro studies on two potent, water-soluble
trimethoprim analogues with exceptional species selectivity
against dihydrofolate reductase from Pneumocystis carinii
and Mycobacterium avium
Ronald A. Forsch,a Sherry F. Queenerb and Andre Rosowskya,*
aDana-Farber Cancer Institute and Department of Biological Chemistry and Molecular Pharmacology,
Harvard Medical School, Boston, MA02115, USA
bDepartment of Pharmacology and Toxicology, Indiana University School of Medicine, Indianapolis, IN 46202, USA
Received 22 September 2003;accepted 4 December 2003
Abstract—2,4-Diamino-5-[30,40-dimethoxy-50-(5-carboxy-1-pentynyl)]benzylpyrimidine (6) and 2,4-diamino-5-[30,40-dimethoxy-50-
(4-carboxyphenylethynyl)benzylpyrimidine (7) were synthesized from 2,4-diamino-5-(50-iodo-30,40-dimethoxybenzyl)pyrimidine (9)
via a Sonogashira reaction with appropriate acetylenic esters followed by saponification, and were tested as inhibitors of dihy-
drofolate reductase (DHFR) from Pneumocystis carinii (Pc), Toxoplasma gondii (Tg), Mycobacterium avium (Ma), and rat in com-
parison with the widely used antibacterial agent 2,4-diamino-5-(30,40,50-trimethoxybenzyl)pyrimidine (trimethoprim, TMP). The
selectivity index (SI) for each compound was calculated by dividing its 50% inhibitory concentration (IC50) against rat DHFR by
its IC50 against Pc, Tg, or Ma DHFR. The IC50 of 6 against Pc DHFR was 1.0 nM, with an SI of 5000. Compound 7 had an IC50 of
8.2 nM against Ma DHFR, with an SI of 11000. By comparison, the IC50 of TMP was 12000 nM against Pc, 300 nM against Ma,
and 180000 against rat DHFR. The potency and selectivity values of 6 and 7 were not as high against Tg as they were against Pc or
Ma DHFR, but nonetheless exceeded those of TMP. Because of the outstanding selectivity of 6 against Pc and of 7 against Ma
DHFR, these novel analogues may be viewed as promising leads for further structure–activity optimization.
# 2004 Elsevier Ltd. All rights reserved.
The well-known antimicrobial agent trimethoprim
(TMP, 1)1a is one the most widely prescribed anti-
microbial drugs in the world. As an ingredient in cotri-
moxazole,1b a two-drug combination also containing
sulfamethoxazole, TMP is used extensively for the pro-
phylaxis and/or therapy of potentially life-threatening
opportunistic infections in patients with AIDS.2
Although the role of such prophylaxis is less widely
accepted outside the AIDS setting, there is nonetheless
accruing evidence that it can be useful in the clinical
management of cancer patients with certain hematolo-
gic malignancies, whose immune system is temporarily
compromised as a result of whole-body radiotherapy or
the administration of high doses of steroids (e.g., for
brain tumors), or who are treated with immunosup-
pressive drugs like cyclosporin as part of an organ
transplantation regimen.3 Three of the opportunistic
diseases prevalent among AIDS patients, and often the
first telltale sign of HIV-1 infection, are pneumocystis
pneumonia, toxoplasmic encephalitis, and disseminated
infection by Mycobacterium avium complex (MAC).
Pneumocystis pneumonia is caused by the fungal
organism Pneumocystis carinii, whereas cerebral toxo-
plasmosis generally results from latent activation of
Toxoplasma gondii spores typically transmitted to
humans earlier in life by domestic pets or farm animals.
Several recent examples of the clinical efficacy of
prophylaxis with cotrimoxazole to prevent P. carinii
and T. gondii infections in the non-AIDS setting are of
interest.4
* Corresponding author. Tel.: +1-617-632-3117;fax: +1-617-632-
2410;e-mail: andre_rosowsky@dfci.harvard.edu
0960-894X/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.12.103

1812
R. A. Forsch et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1811–1815
Trimethoprim acts by blocking the action of dihy-
drofolate reductase (DHFR),5 which catalyzes hydride
transfer from NADPH to dihydrofolate. The resulting
tetrahydrofolate product is a co-factor in the biosynth-
esis of the purine and pyrimidine building blocks
required for RNA and DNA synthesis. The 5,10-meth-
ylene derivative tetrahydrofolate is the natural substrate
for a second key enzyme, thymidylate synthase, whose
role is to convert 20-deoxyuridylate (dUMP) to 20-deoxy-
thymidylate (dTMP) for subsequent incorporation into
DNA. The other product of this reaction is dihy-
drofolate, and the latter has to be reduced back to tet-
rahydrofolate by DHFR in order for the process to
continue. In the context of antimicrobial chemotherapy,
the potency of TMP is generally not sufficient by itself
to bring the number of infectious organisms down to
a low enough level to assure success. For this reason, a
sulfa drug is co-administered to block a second enzyme,
dihydropteroate synthetase, which is used for the de
novo synthesis of folates by many lower organisms
including P. carinii, T. gondii, and M. avium, but is
absent in mammals.6 At the present time the only sulfa
drug used clinically in combination with TMP, other
than sulfamethoxazole, is dapsone.7 A major drawback
to the use of TMP–sulfa combinations in immuno-
compromised patients is the occurrence of hypersen-
sitivity reactions to the sulfa drug that can be severe
enough to require discontinuation of treatment when
the side effects cannot be managed with steroids,
especially since these can actually exacerbate the infec-
tion.8 More potent DHFR inhibitors than TMP, such
as trimetrexate (TMX, 2)9 or piritrexim (PTX, 3),10 do
not require potentiation by a sulfa drug. However,
because these potent antifolates, unlike TMP, are not
selective for non-mammalian versus mammalian DHFR
they can produce dangerous hematologic side effects.
Thus they require administration of a protective or
‘rescue’ agent in the form of leucovorin (LV, 5-for-
myltetrahydrofolate) in order to be clinically useful.
While the TMX-LV combination has aroused clinical
interest for the treatment of acute pneumocystis pneu-
monia in AIDS patients who cannot tolerate sulfa drugs
(or other popular agents such as aerosolized pentami-
dine), even more vigilant clinical monitoring needs to be
practiced with this regimen than with TMP–sulfa-
methoxazole or TMP–dapsone. Thus DHFR inhibitors
that are more potent than TMP and do not require co-
administration of a sulfa drug, while being safe to use in
humans without LV, are of considerable therapeutic
interest.
As part of a larger study aimed at the discovery of
newer DHFR inhibitors that fulfill the elusive criterion
of blending into a single molecule the selectivity of TMP
and the potency of TMX or PTX, we recently synthe-
sized and tested a number of 2,4-diamino-5-(20,50-di-
substituted benzyl)pyrimidines containing a COOH
group linked to the benzyl ring via a spacer.11À13 The
20,50-substitution pattern was chosen to differentiate
these compounds from TMP (selective but not potent)
and bring them closer in structure to piritrexim (potent
but not selective). Two of the more interesting com-
pounds in this group were the 50-(5-carboxy-1-pentynyl)
analogue 411 and the 50-(4-carboxyphenylethynyl) ana-
logue 5.12
Armed with knowledge of the pioneering work of
Kuyper et al. on TMP analogues with a carboxyalkyl
group on the 50-oxygen as inhibitors of Eschericia coli
DHFR,14 we could not pass up the opportunity to
examine the effect of 30,40-dimethoxy-50-(5-carboxy-1-
pentynyl)- and 50-(4-carboxyphenylethynyl) substitution
on Pc, Tg, and Ma DHFR inhibition. Thus, extending
the method we had already followed to prepare 4 and 5,
we synthesized the 30,40,50-trisubstituted analogues 6 and
7 from 5-iodo-3,4-dimethoxybenzaldehyde (8).15 As
shown in Scheme 1, condensation with 3-morpholino-
propionitrile in the presence of NaOEt, followed by
aniline hydrochloride, and finally guanidine16 afforded
the diaminopyrimidine 9. A Pd(II)-catalyzed Sonoga-
shira coupling reaction17 was then used to obtain alkyne
esters 10 and 11 from 9 and benzyl 5-hexynoate and
methyl 4-ethynylbenzoate, respectively. Hydrolysis of
the esters with NaOH in DMSO followed by pre-
parative HPLC on C18 silica gel yielded the desired
acids 6 and 7. The identity and purity of the products
was established by microchemical analysis as well as
1
from their IR and H NMR spectra.
.
Scheme 1. (a) (i) 3-morpholinopropionitrile, NaOEt;(ii) PhNH 2 HCl;(iii) H 2NC(¼NH)NH2;(b) RC ꢀCH, (Ph3P)2PdCl2, (Ph3P)3CuBr, Et3N;
(c) NaOH, DMSO.

R. A. Forsch et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1811–1815
1813
Table 1. Inhibition of Pc, Tg, Ma, and rat DHFR by 2,4-diamino-5-(3’,4’-dimethoxy-5’-substituted benzyl]pyrimidines with a carboxyalkynyl or
carboxyphenylalkynyl group in the side chain
IC50 (nM)a
Ma
Selectivity Index (SI)b
Tg
Cmpd
Pc
Tg
rat
Pc
Ma
4
28
(25–30)
1300
(0.65–2.5)
1.0
32
(27–37)
340
(0.20–0.55)
34
7.8
(6.4–9.5)
3.7
(0.0029–0.0046)
2.4
2200
(1900–2500)
8200
(7100–9300)
5000
79
(63–100)
6.3
(2.8–14)
5000
69
(52–91)
24
(13–47)
150
280
(200–380)
2200
(1500–3200)
2100
5
6
(0.82–1.2)
1200
(30–39)
420
(0.0021–0.0027)
8.2
(4300–5800)
88000
(3600–7100)
73
(110–190)
210
(2000–2800)
11000
7
(990–1400)
13000
(10000–16000)
47
(34–66)
13
(9.0–17)
(300–590)
2800
(2400–3300)
16
(8–30)
4.3
(4.0–4.6)
(6.9–9.8)
300
(260–350)
1.5
(1.3–1.7)
0.61
(0.53–0.70)
(57000–140000)
180000
(160000–210000)
8.0
(7.0–9.2)
3.3
(2.9–3.9)
(41–140)
14
(10–20)
0.017
(0.11–0.27)
0.26
(0.17–0.42)
(97–470)
65
(48–87)
0.50
(0.23–1.2)
0.76
(0.63–0.97)
(5800–20000)
610
(460–810)
5.3
(4.1–7.1)
5.4
(4.1–7.2)
TMPc
TMXd
PTXe
a Numbers in parentheses are the 95% confidence intervals, rounded off to two significant figures and are based on IC50 values similarly rounded off
to two figures. The difference in IC50 between rat liver DHFR and each of the parasite enzymes was statistically significant at P<0.01 (Welsh’s
t-test). Data for 4 and 5 shown for comparison are from refs. 13 and 14 respectively.
b SI=IC50(rat DHFR)/IC50(Pc, Tg, or Ma DHFR). Numbers in parentheses are rounded off to two figures, and represent a range calculated by
dividing the lower end of the 95% confidence range for the IC50 against rat DHFR by the upper end of the 95% confidence range for the IC50
against Pc, Tg, or Ma DHFR.
c TMP=2,4-diamino-5-(30,40,50-trimethoxybenzylpyrimidine);data given for comparison are from ref. 14.
d TMX=2,4-diamino-5-methyl-6-(30,40,50-trimethoxyanilino)-methylquinazoline.
e PTX=2,4-diamino-5-methyl-6-(20,50-dimethoxybenzyl)pyrido[2,3-d]pyrimidine;data given for comparison are from ref. 14.
The ability of 6 and 7 to inhibit Pc, Tg, Ma, and rat
DHFR was determined by the standardized method in
our previous work.18 The assay is based on measure-
ment of the change in absorbance at 340 nm when
dihydrofolate is reduced to tetrahydrofolate in the pre-
sence of NADPH. The results are presented in Table 1.
With regard to phenylalkynyl substitution, the IC50 of 7
against Pc and Tg DHFR was not substantially differ-
ent from that of 5. However binding to Ma DHFR was
enhanced ca. 2-fold while binding to rat DHFR
was decreased 11-fold. As a consequence, the selectivity
index of 7 increased 5-fold to 11,000, making this the
most selective TMP analogue we have tested to date
against the Ma enzyme.21
Whereas the IC50 of 4 against Pc and Ma DHFR was
previously found to be 28 nM and 7.8 nM, respectively,
the corresponding values for the 30,40,50-trisubstituted
analogue 6 proved to be ca. 3-fold lower (Table 1),
indicating that the latter structure is more favorable for
binding to these enzymes. More significantly, binding to
rat DHFR was also decreased ca. 2-fold, resulting in a
>10-fold improvement in selectivity against both
enzymes. By comparison, the IC50 against Tg DHFR
was unaffected, and selectivity was only marginally
enhanced. The 5000-fold selectivity of 6 against Pc ver-
sus rat DHFR is considerably higher than we have
encountered with any of our 20,50-disubstituted inhibi-
tors with a terminal COOH group in the side chain,
rivaling only that of the dicarboxylic acid 12, which is
claimed to have 4300-fold selectivity for Pc versus
human DHFR in a patent by Kompis et al.19 Com-
pound 12 is structurally related to 2,4-diamino-5-(40-
bromo-30,50-dimethoxybenzyl)pyrimidine (brodimoprim,
BMP), which has an overall spectrum of antibacterial
activity not dissimilar from that of TMP but displays
more favorable cell penetration and pharmacokinetics.20
It should be noted that 6 was 12,000 times more potent
than TMP against Pc DHFR, and that its potency
actually exceeded that of TMX and PTX by a factor of
at least 10. However, in contrast to the latter drugs,
which are much more potent against rat than Pc DHFR
and thus may be said to be ‘reverse-selective’, 6 shows a
level of binding selectivity at least 300-fold greater than
that of TMP. A further advantage of both 6 and 7 over
TMP, TMX, and PTX is that they are soluble in water
at physiologic pH, and thus are more suitable for par-
enteral administration.26
The fact that a DHFR inhibitor is potent against the
enzyme does not, of course, mean that it will have an
antifolate effect on intact cells. To address this issue, the
effect of 6 on proliferation of Pc cells was compared
with that of TMP in a luciferase-based bioluminescence
assay of ATP.23À25 Treatment of three different Pc iso-
lates (from immunosuppressed rats) with 100 mg/mL
(250 mM) of 6 in RPMI 1640 medium containing 20%
calf serum for 72 h resulted in a mean T/C (treated/
control) value of 40%, whereas the T/C for 100 mg/mL
(180 mM) of TMP under the same conditions was 22%.
As would be expected, lower drug concentrations (e.g.,
10 mg/mL) or shorter treatment times (e.g., 48 h) were
less effective for both drugs. Thus, it was gratifying that,
while its molar potency against intact Pc was slightly

1814
R. A. Forsch et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1811–1815
lower that of TMP, 6 nonetheless had the ability to
enter Pc cells despite the presence of a free COOH group.
Lack, E. E.;Shelhamer, J. H.;Balis, F.;Walker, W.;
Kovacs, J. A.;Lane, H. C.;Masur, H. N. Eng. J. Med.
1987, 317, 978. (b) Sattler, F. R.;Frame, P.;Davis, R.;
Nichols, L.;Shelton, B.;Akil, B.;Baughman, R.;Hugh-
lett, C.;Weiss, W.;Boylen, C. T.;van der Horst, C.;
Black, J.;Powderly, W.;Steigbigel, R. T.;Leedom, J. M.;
Masur, H.;Feinberg, J.;Benoit, S.;Eyster, E.;Gocke, D.;
Beck, K.;Lederman, M.;Phari, J.;Reichman, R.;Sacks,
H. S.;Soiero, R. J. Infect. Dis. 1994, 170, 165.
In summary, our results indicate that, where binding
and selectivity for Pc DHFR are concerned, 50-(5-car-
boxy-1-pentynyl) substitution is more favorable when
the benzyl ring is 30,40,50-trisubstituted than when it is
20,50-disubstituted. Where binding to Ma DHFR is
concerned, both 50-(5-carboxy-1-pentynyl) and 50-(4-
carboxyphenylethynyl) substitution are more conducive
to binding and selectivity when the benzyl ring is
30,40,50-trisubstituted. Compounds 6 and 7 have the
highest species selectivity for Pc and Ma DHFR
respectively, of any 2,4-diamino-5-benzylpyrimidine
inhibitors of these enzymes we have synthesized and
tested to date. For this reason these novel analogues
may be viewed as novel leads for further structure–
activity optimization of DHFR binding, and ultimately
cell penetration, tissue distribution, metabolism, and
pharmacokinetics.
10. (a) Kovacs, J. A.;Allegra, C. J.;Swan, J. C.;Drake,
J. C.;Parrillo, J. E.;Chabner, B. A.;Masur, H.
Anti-
microb. Agents Chemother. 1988, 32, 430. (b) Falloon, J.;
Allegra, C. A. J.;Kovacs, J.;O’Neill, D.;Ogata-Arakaki,
D.;Feuerstein, I.;Polis, M.;Davey, R.;Lane, H. C.;
LaFon, S.;Rogers, M.;Zunich, K.;Turlo, J.;Tuazon, D.;
Parenti, D.;Simon, G.;Mazur, H. Clin. Res. 1990, 38,
361A.
11. Rosowsky, A.;Forsch, R. A.;Queener, S. F.
Chem. 2002, 45, 233.
12. Rosowsky, A.;Forsch, R. A.;Queener, S. F.
J. Med.
J. Med.
Chem. 2003, 46, 1726. This paper may be consulted for a
comprehensive bibliography of our past work on DHFR
inhibitors as potential drugs against opportunistic infec-
tions.
13. Rosowsky, A.;Forsch, R. A.;Sibley, C. H.;Inderlied, C.
B.;Queener, S. F. Manuscript submitted.
Acknowledgements
14. Kuyper, L. F.;Roth, B.;Baccanari, D. P.;Ferone, R.;
Beddell, C. R.;Champnesss, J. N.;Stammers, D. K.;
Dann, J. G.;Norrington, F. E.;Baker, D. J.;Goodford,
P. J. J. Med. Chem. 1985, 28, 303.
15. Calas, M.;Barbier, A.;Giral, L.;Balmayer, B.;Despaux,
E. Eur. J. Med. Chem. 1982, 17, 497.
This work was supported by NIH grant RO1-AI29904
(to A.R.) from the National Institute of Allergy and
Infectious Diseases (NIAID).
16. Stuart, A.;Paterson, T.;Roth, B.;Aig, E. J. Med. Chem.
1983, 26, 667.
References and notes
17. Sonogashira, K.;Tohda, Y.;Hagihara, B.
Lett. 1975, 4467.
Tetrahedron
1. (a) Roth, B.;Falco, E. A.;Hitchings, G. H.;Bushby,
S. R. M. J. Med. Chem. 1962, 5, 1103. (b) Bushby,
S. R. M.;Hitchings, G. H. Br. J. Pharmacol. Chemother.
1968, 33, 72.
2. (a) Fishman, J. A. Antimicrob. Agents Chemother. 1998,
42, 1309. (b) Kovacs, J. A.;Gill, V. J.;Meshnick, S.;
Masur, H. JAMA 2001, 286, 2450.
3. Kovacs, J. A.;Hiemenz, J. W.;Macher, A. M.;Stover,
D.;Murray, H. W.;Shelhamer, J.;Lane, H. C.;Urma-
cher, C.;Honig, C.;Longo, D. L.;Parker, M. M.;
Natanson, C.;Parrillo, J. E.;Fauci, A. S.;Pizzo, P. A.;
Masur, H. Ann. Intern. Med. 1984, 100, 663.
4. (a) Henson, J. W.;Jalaj, J. K.;Walker, R. W.;Stover,
D. E.;Fels, A. O. Arch. Neurol. 1991, 48, 406. (b) Oken,
M. M.;Pomeroy, C.;Weidorf, D.;Bennett, J. M. Am. J.
Med. 1996, 100, 624. (c) Mehta, J.;Powles, R.;Singhal,
18. Rosowsky, A.;Cody, V.;Galitsky, N.;Fu, H.;Papoulis,
A.;Queener, S. F. J. Med. Chem. 1999, 42, 4853.
19. Kompis, I.;Blaney, J. M.;Marlow, C. K. International
Patent WO 92/08461 (29 May 1992), assigned to Protos
Corporation, Emeryville, CA.
20. (a) Braunsteiner, A. R.;Finsinger, F. J. Chemother. 1993,
5, 507. (b) Braga, P. C.;Dal Sasso, M.;Maci, S.;Bon-
diolotti, G.;Fonti, E.;Reggio, S.
Chemother. 1996, 40, 2392.
Antimicrob. Agents
21. After completing the synthesis and testing of 7 we dis-
covered that this structure was claimed as part of a
patent on TMP analogues as antimicrobials against
gram-negative bacteria, and P. carinii.22 However, only
the synthesis of this compound was described in the
patent, and no data on binding to either Pc or Ma were
disclosed.
S.;Riley, U.;Treleaven, J.;Catovsky, D.
Leuk. Lym-
22. Guerry, P.;Jolidon, S.;Masciadri, R.;Salder, H.;Then,
R. U.S. Patent 5,763,450 (9 June 1998), assigned to Hoff-
mann-La Roche, Inc., Nutley, NJ.
23. (a) Chen, F.;Cushion, M. T. J. Clin. Microbiol. 1994, 32,
2791. (b) Cushion, M. T.;Chen, F.;Kloepfer, N. Anti-
microb. Agents Chemother. 1997, 41, 379.
24. We are grateful to Dr. Melanie Cushion, Dept. of the VA
Medical Center, University of Cincinnati, for kindly per-
forming these assays.
25. Although assays using 6 against intact Ma cells have not
yet been performed, it may be noted that Dr. Clark
Inderlied (Children0s Hospital Medical Center, Los
Angeles, CA) recently found that the 20,50-disubstituted
analogue 4 inhibits the growth of human isolates of this
organism in culture at concentrations of 4-32 mg/mL
depending on the isolate, whereas TMP was inactive even
at 64 mg/mL.13
phoma 1997, 26, 83. (d) Torre-Cisneros, J.;de la Mata,
M.;Lopez-Ciller, P.;Sanchez-Guijo, P.;Mino, G.;Pera,
C. Transplantation 1996, 62, 1519. (e) Baden, L. R.;Katz,
J. T.;Franck, L.;Tsang, S.;Hall, M.;Rubin, R. H.;
Jarcho, J. Transplantation 2003, 75, 339.
5. Schweitzer, B. I.;Dicker, A. P.;Bertino, J. R. FASEB J.
1990, 4, 2441.
6. Kovacs, J. A.;Allegra, C. J.;Beaver, J.;Boarman, D.;
Lewis, M.;Parillo, J. E.;Chabner, B.;Masur, H. J. Infect.
Dis. 1989, 160, 312.
7. Medina, I.;Mills, J.;Leoung, G.;Hopewell, P. C.;Lee,
B.;Modin, G.;Benowitz, W. C. B. N. Engl. J. Med. 1990,
323, 776.
8. Bayard, P. J.;Berger, T. G.;Jacobson, M. A. J. Acquir.
Immune Def. Syndr. 1992, 5, 1237.
9. (a) Allegra, C. J.;Chabner, B. A.;Tuazon, C. U.;Ogata-
Arakaki, D.;Baird, B.;Drake, J. C.;Simmons, J. T.;

R. A. Forsch et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1811–1815
1815
26. The relative ability of water-soluble TMP analogues con-
taining a COOH group in the side chain to enter cells can
be expected to vary substantially depending on the target
organism. For example, Kuyper et al.14 reported that 2,4-
diamino-5-[30,40 -dimethoxy-50 -(5-carboxypentyloxy)]pyr-
imidine, a structural analogue of 6, had nearly the same
activity as TMP against S. aureus, but was 300-fold less
active than TMP against E. coli organisms in culture even
though it had 50-fold greater affinity for E. coli DHFR in
a cell-free assay.