Analogues of the potent nonpolyglutamatable antifolate N(α)-(4-amino-4- deoxypteroyl)-N(δ)-hemiphthaloyl-L-ornithine (PT523) with modifications in the side chain, p-aminobenzoyl moiety, or 9,10-bridge: Synthesis and in vitro antitumor activity

Article abstract of DOI:10.1021/jm990630f

Seven N(α)-(4-amino-4-deoxypteroyl)-N(δ)-hemiphthaloyl-L-ornithine (2, PT523) analogues were synthesized by modifications of the literature synthesis of the corresponding AMT (1) analogues and were tested as inhibitors of tumor cell growth. In growth assa

Full text of DOI:10.1021/jm990630f

1620
J . Med. Chem. 2000, 43, 1620-1634
An a logu es of th e P oten t Non p olyglu ta m a ta ble An tifola te
Nr-(4-Am in o-4-d eoxyp ter oyl)-Nδ-h em ip h th a loyl-L-or n ith in e (P T523) w ith
Mod ifica tion s in th e Sid e Ch a in , p-Am in oben zoyl Moiety, or 9,10-Br id ge:
Syn th esis a n d in Vitr o An titu m or Activity
Andre Rosowsky,* J oel E. Wright, Chitra M. Vaidya, Ronald A. Forsch, and Henry Bader
Dana-Farber Cancer Institute and Department of Biological Chemistry and Molecular Pharmacology,
Harvard Medical School, 44 Binney Street, Boston, Massachusetts 02115
Received December 28, 1999
Seven NR-(4-amino-4-deoxypteroyl)-Nδ-hemiphthaloyl-L-ornithine (2, PT523) analogues were
synthesized by modifications of the literature synthesis of the corresponding AMT (1) analogues
and were tested as inhibitors of tumor cell growth. In growth assays against cultured CCRF-
CEM human leukemic cells exposed to drug for 72 h, the IC50 values of analogues in which N10
was replaced by CH2 and CHMe were found to be 0.55 ( 0.07 and 0.63 ( 0.08 nM, and thus
these analogues are more potent than 1 (IC50 ) 4.4 ( 1.0 nM) or 2 (IC50 ) 1.5 ( 0.39 nM). The
10-ethyl-10-deaza analogue of 2 (IC50 ) 1.2 ( 0.25 nM) was not statistically different from 2
but was more potent than edatrexate, the 10-ethyl-10-deaza analogue of 1, which had an IC50
of 3.3 ( 0.36 nM. In contrast, the analogue of 2 with both an ethyl and a CO2Me group at the
10-position had an IC50 of 54 ( 4.9 nM, showing this modification to be unfavorable. The
4-amino-1-naphthoic acid analogue of 2 had an IC50 of 1.2 ( 0.22 nM, indicating that
replacement of the p-aminobenzoic acid (pABA) moiety does not diminish cytotoxicity. The
analogues in which the (CH2)3 side chain was replaced by slightly longer CH2SCH2 and (CH2)2-
SCH2 groups gave IC50 values of 4.4 ( 1.1 and 5.0 ( 0.56 nM and thus were somewhat less
potent than the parent molecule. However the analogues in which the aromatic COOH group
was at the meta and para positions of the phthaloyl ring had IC50 values of 7.5 ( 0.47 and 55
( 0.07 nM, confirming the low potency we had previously observed with these compounds
against other cell lines. Overall, the results in this study support the conclusion that, while
the position of the phthaloyl COOH group and the length of the amino acid side chain in 2 are
important determinants of cytotoxic potency, changes in the pABA region and 9,10-bridge are
well-tolerated and can even increase potency.
In tr od u ction
in plasma and tissues. In the presence of acid, on the
other hand, 2 is extensively cleaved to the weak DHFR
inhibitor NR-(4-amino-4-deoxypteroyl)-L-ornithine.8,9 The
high potency of 2 against cultured cells is believed to
be due to a combination of two factors that work
together in its favor: (a) an unusually strong interaction
with DHFR, a process in which the phthaloyl group is
believed to play a specific and important structural
role,10,11 and (b) an unusually efficient utilization of the
reduced folate carrier (RFC) for influx into cells.12-14
Analogues of the classical dihydrofolate reductase
(DHFR) inhibitor aminopterin (AMT, 1; cf. Chart 1) in
which Nδ-hemiphthaloyl-L-ornithine replaces the glutam-
ic acid side chain exhibit potent in vitro antitumor
activity despite the fact that, unlike the glutamate
derivatives, they are incapable of being metabolized to
more tightly enzyme-bound and less rapidly effluxed
γ-polyglutamates (reviewed in refs 1 and 2). Detailed
investigations of the mode of action of NR-(4-amino-4-
deoxypteroyl)-Nδ-hemiphthaloyl-L-ornithine (PT523, 2;
cf. Chart 1), the first compound synthesized in this
series,3,4 established that the drug is a potent inhibitor
of both purine and pyrimidine biosynthesis,5 that it is
taken up, at least in part, via the reduced folate carrier
(RFC),5 and that its activity is associated with rapid and
persistent depletion of cellular tetrahydrofolate cofac-
tors, both in cultured cells5 and in mice.6 It is a
moderately good substrate for partially purified human
hepatic aldehyde oxidase,7 which converts it to a 7-hy-
droxy derivative. Although some 7-hydroxylation has
indeed been observed in tumor-bearing mice,6 the amide
bond in the hemiphthaloyl group appears to be stable
Because of the novel mode of action of 2 as a water-
soluble nonpolyglutamatable DHFR inhibitor, a sys-
tematic program of synthesis of second-generation
analogues was undertaken in our laboratory with a view
to identifying structure-activity correlations among this
class of antifolates. Previous papers have described
analogues 3-14, in which (a) the side chain was either
lengthened or shortened (3-5),15,16 (b) the carboxyl
group in the phthaloyl moiety was moved from the ortho
to the meta or para position (6, 7),16 (c) ring substitution
was introduced into the 4-aminobenzoic acid (pABA)
moiety (8),16 and (d) the B-ring was modified at N5 and/
or N8 (9-13).17 In the present paper we report the
synthesis of additional analogues (15-21) in which N10
is replaced by carbon (15-18), the pABA moiety is
replaced by 4-amino-1-naphthoic acid (19), or the side
* To whom correspondence should be addressed. Tel: 617-632-3117.
Fax: 617-632-2410. E-mail: andre_rosowsky@dfci.harvard.edu.
10.1021/jm990630f CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/31/2000

Analogues of the Nonpolyglutamatable Antifolate PT523
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1621
Ch a r t 1
Ch a r t 2
chain contains a sulfur atom in the form of cysteine or
homocysteine (20, 21). Compound 17 may be viewed as
an analogue of 10-ethyl-10-deazaaminopterin (edatrex-
ate),18 which has been undergoing clinical trial for the
past several years.19 The structures of the compounds
3-21 are shown in Chart 2.
to others using 26‚HBr,20,23 25 was generated with
potassium tert-butoxide. The ylide was readily detected
by its dark-red color, which began to fade immediately
upon addition of the aldehyde. The resulting olefin 28,
assumed (but not formally proved) to be a mixture of
cis and trans isomers, was reduced catalytically to a
mixture of 9,10-dihydro derivatives and the B-ring was
reoxidized with H2O2 as described.23 The overall yield
of phthalimide ester 29 from 27 was 54%. Treatment
of 29 with NaOH in warm MeOH for 1.75 h afforded 15
in 53% yield (29% overall). As in our previous experience
with this type of chemistry, the phthalimide esters 28
and 29 were amenable to purification by chromatogra-
phy on silica gel. When 29 was rigorously purified before
final hydrolysis, 15 could be obtained in an analytically
pure state without further chromatography.
Ch em istr y
Three different approaches to the synthesis of 10-
deaza analogues of 2 were investigated (Schemes 1-4).
In general, these approaches are patterned after routes
used previously to obtain the corresponding glutamate
analogues.20-25 Thus, as shown in Scheme 1, coupling
of 4-formylbenzoic acid (22) to methyl 2-L-amino-5-
phthalimidopentanoate hydrochloride (23‚HCl)17 via the
mixed carboxylic anhydride (MCA) method afforded
aldehyde 24. The latter was used in a Wittig reaction
with the putative 2,4-bis(phosphazene) ylide 25,26,27
arising either from pre-formed 2,4-diamino-6-(bromo-
methyl)pteridine hydrobromide (26‚HBr)28 or in situ
from 2,4-diamino-6-(hydroxymethyl)pteridine (27)29 as
we described earlier.30 In the present study, in contrast
An alternate route to 28 was also investigated (Scheme
2), involving a Wittig reaction between aldehyde 24 and
the ylide generated from 2-amino-3-cyanopyrazin-5-
ylmethyltriphenylphosphonium chloride (30).31 To our
knowledge, this was the first reported attempt to use
30 as an intermediate in the synthesis of 2,4-diamino-

1622 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Rosowsky et al.
Sch em e 1
Sch em e 2
10-deazapteridines with an amino acid side chain. The
dark-red ylide 31 formed very rapidly in the presence
of potassium tert-butoxide,32 and its reaction with the
aldehyde appeared to proceed somewhat more rapidly
than that of 25. After flash chromatography on silica
gel, olefin 32 was isolated in 40% yield. Like the
pteridine 28 derived from ylide 25, this product was
assumed to be a mixture of the cis and trans olefins,
but again the isomer ratio was not studied. Ring closure
to a pteridine was performed by heating for ca. 4 days
with guanidine in refluxing MeOH, and the product was
purified by silica gel chromatography. Somewhat un-
expectedly, microchemical analysis of the material
isolated in this particular run suggested that it was not
28 but rather the monoammonium salt of the ring-
opened structure 33, whose catalytic reduction, reoxi-
dation with H2O2, and treatment with NaOH afforded
15. Cleavage of the phthalimide ring of 28 had appar-
ently occurred when the product was dissolved and left
to stand in aqueous ammonia at pH 9. Thus a way had
been serendipitously found by which the pthalimide ring
could be opened without cleaving the methyl ester.
However, even though 15 could be made via pyrazine
ylide 31, this approach was judged to be less satisfactory
than the one using pteridine ylide 25.
The third route we utilized to obtain 15 (Scheme 3)
was via MCA coupling of 23‚HCl with 4-amino-4-deoxy-
10-deazapteroic acid (34). The latter was made from
2,4,5,6-tetraaminopyrimidine by an adaptation of the
five-step sequence described in the literature.21 Briefly,

Analogues of the Nonpolyglutamatable Antifolate PT523
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1623
Sch em e 3
Sch em e 4
3-methoxy-1-propyne (35) was converted sequentially to
1-methoxyallene (36) and 1-chloro-4-methoxy-2-butene
(37), the latter was used to C-alkylate the dilithio
derivative of 4-toluic acid, and the resulting enol ether
38 was subjected to bromination followed by acidolysis
to obtain the key intermediate 2-bromo-4-(4-carboxy-
phenyl)butanal (39). The entire sequence was carried
out in a continuous operation with each intermediate
being used directly in the next reaction. The yield of 29
obtained from 39, after chromatography, was 25%.
Hydrolysis of 29 to 15 was carried out, in this case, with
NaOH as the base and DMSO as the solvent. However,
while it accelerated the reaction in comparison with
MeOH (5 min versus 1.75 h), the use of DMSO did not
substantially change the yield, which remained around
50-60%. Because the overall yield of 34 from p-toluic
acid, in our hands, was <1%, we concluded that the
more satisfactory route to 15 was via Scheme 1.
Having successfully used MCA coupling to prepare
15, we also chose this method to obtain the 10-methyl-
10-deaza analogue 16 (Scheme 4). The requisite inter-
mediate 4-amino-4-deoxy-10-methyl-10-deazapteroic acid
(40) was previously synthesized by DeGraw and co-
workers21 from 36 and the dilithio derivative of 4-eth-
ylbenzoic acid, but in the present work 40 was pre-
pared via a slight modification of the more recent
method used by the same authors to obtain 10-propar-
gyl-10-deazaaminopterin.24 Although detailed proce-
dures were published only for the 10-propargyl ana-
logue, this method was described in a review article
as being generally applicable to the synthesis of other
10-alkyl derivatives.25 Briefly, dimethyl homotereph-

1624 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Rosowsky et al.
Sch em e 5
thalate (41) was C-alkylated with methyl iodide in the
presence of NaH, and the resulting diester 42 (56%
yield) was re-alkylated, this time with 26‚HBr, to ob-
tain 43 (29% yield). After saponification to 44 (73%
yield), selective removal of the benzylic carboxyl was
accomplished in 59% yield by heating the diacid in
DMSO at 135 °C. The overall four-step yield of 40 by
this route was approximately 10%. Mixed anhydride
coupling to 23‚HCl, followed by treatment with NaOH
in DMSO, gave 45 (19%) and 16 (61%), respectively. As
in the case of 15, rigorous purification of the penulti-
mate product 45 on a silica gel column allowed the final
hemiphthaloyl derivative to be obtained in analytically
pure state without further chromatography. The amount
of 16 obtained via this sequence was again quite low,
though it should noted that, since we desired only to
prepare enough material for small-scale in vitro testing,
the yield in each step of the process was not optimized.
In agreement with the reported synthesis of the corre-
sponding glutamate analogues,21,24 45 and 16 were
isolated as unresolved 10R/10S mixtures. The presence
of the two isomers was obvious from 500-MHz 1H NMR
spectra of these compounds, which featured two closely
spaced singlets of equal intensity for the pteridine C-7
proton. However the C-7 proton in both 29 and 15 was
a singlet.
An attempt was next made to synthesize the 10-ethyl
analogue 17 by the same method as 16. Thus, 41 was
C-alkylated with ethyl iodide in the presence of NaH to
obtain adduct 46 (Scheme 4). Further reaction with 26‚
HBr afforded 47, but when the latter compound was
subjected to the saponification conditions we had used
to prepare 40 (i.e., NaOH in DMSO at room tempera-
ture), the desired diacid 48 was not formed cleanly. This
was confirmed by the 1H NMR spectrum, which con-
tained a sharp singlet at 3.6 ppm and no signal at 3.9
ppm, the latter being the chemical shift we had noted
for the aromatic CO2Me group in precursor 47. Thus, it
was evident that saponification had produced mainly
49 rather than 48. Final confirmation of the structure
of 49 came from the fact that MCA coupling with 23‚
HCl, followed by the usual mild treatment with NaOH
in DMSO, led to a product whose elemental analysis
revealed a higher than expected C/N ratio and whose
1H NMR spectrum still contained a sharp aliphatic ester
singlet at δ 3.59. On the basis of these results we
characterized the coupling product as 50 and the
eventual hydrolysis product as 18. To our knowledge,
18 is the first and only example of a 10-deaza antifolate
with an ester group on C-10.
It should be noted that a number of experiments were
done to see if hydrolysis of 47 would succeed under more
vigorous conditions. Higher concentrations of NaOH,
higher reaction temperatures, and solvents other than
DMSO (e.g., refluxing 2-methoxyethanol) were tried, but
to our surprise, the product always contained varying
amounts of unreacted starting material. Moreover there
were also more byproducts, possibly resulting from
cleavage of the 4-amino group and rupture of the
pyrimidine ring. Another method of ester cleavage
involving the use of cyanide ion as a nucleophile was
abandoned when it was found to lead to a complex
mixture of products, possibly reflecting the known
susceptibility of pteridines to cyanide attack at the C-7
position.33 Other strong nucleophiles were not investi-
gated.
Faced with this unexpected setback, we then decided
to return to the Wittig approach as a route to 17, and
for this purpose needed to make a quantity of 4-(meth-
oxycarbonyl)propiophenone (51; Scheme 5), which Piper
and co-workers had used in their synthesis of edatrex-
ate.23 However, because we found in pilot experiments
that the published routes to the carboxylic acid precur-
sor of 51 from p-methylpropiophenone gave low and/or
irreproducible yields for the oxidation of the methyl
group with KMnO434 or with O2 in the presence of nickel

Analogues of the Nonpolyglutamatable Antifolate PT523
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1625
Sch em e 6
benzoate,35 affording mainly terephthalic acid rather
than 4-carboxypropiophenone, we developed a new and
more reliable synthesis of 51 from the commercially
available starting material 4-bromopropiophenone (52).
As shown in Scheme 5, treatment of 52 with NaCN in
DMF afforded the nitrile 53, which on hydrolysis with
NaOH in 2-methoxyethanol was converted to carboxylic
acid 54. Neutral esterification with MeI and i-Pr2NEt
in DMF completed the synthesis. The overall three-step
yield of 51 by this route was 55%. Wittig condensation
of 51 and 26‚HBr in the presence of n-Bu3P, followed
by catalytic hydrogenation and reoxidation with H2O2
as described,23 yielded 55 and 56, respectively. Several
attempts to use Ph3P in place of n-Bu3P in the Wittig
reaction were unsuccessful, probably because the ylide
25 is too sterically hindered to react with ketone 51.
Saponification of 56, followed as usual by MCA coupling
with 23‚HCl and brief treatment with Ba(OH)2 in
aqueous EtOH, afforded 57, 58, and 17, respectively.
The synthesis of the naphthyl analogue 19 is shown
in Scheme 6 and was patterned after that of the
corresponding glutamate.27 Thus, 4-nitro-1-naphthoyl
chloride (59) was prepared from 4-nitro-1,8-naphthoic
anhydride and condensed with 23‚HCl to obtain the
phthalimide ester 60 with an overall yield of 32%.
Catalytic reduction of the nitro group, using ammonium
formate as the H donor in the presence of 10% Pd-C,
yielded the amine 61, which was N-alkylated with 26‚
HBr in the presence of Na2CO3 to obtain 62 (51% yield).
Compounds 60 and 61 were purified by recrystallization
and 62 by chromatography on silica gel. Hydrolysis of
the methyl ester with concomitant opening of the
phthalimide ring to form 19 (71% yield) was accom-
plished by stirring with Ba(OH)2 in aqueous EtOH at
ornithine analogues into cells by the RFC is believed to
reflect the presence in these compounds of two COOH
groups, and the ring-closed form would have the same
microchemical analysis as the ring-opened form with
one less molecule of H2O of solvation.
The cysteine and homocysteine derivatives 20 and 21
were synthesized as shown in Scheme 7, starting from
commercially available methyl L-cysteinate hydrochlo-
ride (63‚HCl) and D,L-2-aminobutyrolactone (64), re-
spectively. The racemic lactone was used for reasons of
economy, and it was assumed, by analogy with the well-
known precedent of D- versus L-methotrexate,36 that the
product would have about one-half the activity of the
pure L-enantiomer. Sequential reaction of 63‚HCl with
di-tert-butyl dicarbonate in the presence of Et3N af-
forded the Boc derivative 65 (61%), which on further
reaction with N-bromomethylphthalimide in the pres-
ence of i-Pr2NEt was converted to the phthalimide 66
(87%). Cleavage of the Boc group with p-TsOH in
refluxing toluene afforded the tosylate salt 67‚TsOH
(86%). The overall yield of 67‚TsOH from 63‚HCl was
46%. Coupling of 67‚TsOH to 4-amino-4-deoxy-N10
-
formylpteroic acid (68) by the MCA method yielded the
phthalimide ester 69. The usual mild treatment with
NaOH in DMSO then gave a mixture whose HPLC (C18
silica gel, 8% MeCN in 0.1 M NH4OAc, pH 6.9) revealed
two principal peaks in a ratio of ca. 2:1. Lyophilization
of the major product and reprecipitation from dilute
NH4OH with AcOH afforded 20 with an overall yield
10% based on 67‚TsOH. The identity of the minor
product was not studied. For the synthesis of 21, the
key intermediate methyl D,L-2-amino-5-[(2-phthalimi-
domethyl)thio]butanoate (70) was synthesized via a
sequence recently used by us in another project.37
Briefly, methyl D,L-2-amino-4-bromobutanoate hydro-
chloride (71‚HCl) was prepared from 64 by acidolysis
(HBr/AcOH) and esterification (MeOH/SOCl2), and was
condensed with N-mercaptomethylphthalimide in the
presence of K2CO3 to obtain 70. MCA coupling of 68 and
70 then gave the phthalimide 72, As with other phthal-
imide esters described above, 72 was purified to homo-
geneity by silica gel column chromatography, whereas
21 was purified by reversed-phase preparative HPLC.
The combined two-step yield of 21 from 70 was 19%.
The presence of a hemiphthaloyl group in the final
1
room temperature for 24 h. A notable feature of the H
NMR spectrum of 19, confirming the presence of the
hemiphthaloyl group, was the chemical shift of the
ornithine δ-CH2, which was at 3.19 ppm as compared
with 3.59 ppm in the ring-closed compound 62. The
same upfield displacement was likewise noted in the 1H
NMR spectra of the 10-deaza analogues 15-18 relative
to their ring-closed precursors 29, 45, 50, and 58 (see
Experimental Section), thus proving that all these
compounds are dicarboxylic acids. It is important that
this be clearly established in every case, inasmuch as
the very efficient active transport of the hemiphthaloyl-
1
product was confirmed by its H NMR spectrum, which

1626 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Rosowsky et al.
Sch em e 7
revealed an upfield shift of ca. 0.3 ppm for the SCH2N
singlet in the ring-opened structure.
SCC25 head-and-neck squamous cell carcinoma can be
extended to CCRF-CEM cells in suspension.
With regard to the new compounds reported here,
replacement of N10 in 2 by CH2 (15) or CHMe (16) led
to a ca. 3-fold increase in potency. However, 17 was
slightly less active than 15 or 16 and thus was not
statistically different from 2. The 10-ethyl-10-methoxy-
carbonyl analogue 18 had an IC50 of 54 ( 4.9 nM. Thus,
this modification of the 10-position in 15 was clearly
unfavorable.
It may be noted that 10-deazaaminopterin and edatr-
exate are reported38 to be essentially like aminopterin
in their growth inhibitory activity against Manca cells,
which are likewise a human leukemic cell line but are
of B-cell rather than T-cell lineage. The reported IC50
values for aminopterin and edatrexate against these
cells are 1.4 ( 0.3 and 0.9 ( 0.2 nM, respectively, and
thus are ca. 3-fold lower than our values for these
compounds against CCRF-CEM cells. Whether this is
due to a difference in sensitivity to DHFR inhibitors
between B-cells and T-cells or to small differences in
assay methodology is unknown. It is clear from our
results, however, that the effect of replacing nitrogen
by carbon at the 10-position on cell growth inhibition is
qualitatively similar in the hemiphthaloylornithine
derivatives as it is in glutamate derivatives even though
only the latter can form polyglutamates.
Biologica l Activity
The seven new analogues of 2 described in this paper
were tested as inhibitors of the growth of cultured
CCRF-CEM cells exposed continuously to drug for 72
h, a period of time that allows the nonpolyglutamatable
analogues to be compared to classical DHFR inhibitors
under conditions where the latter drugs have ample
time to form polyglutamates in the cell. As in our
previous work on second-generation analogues of 2, the
goal of this study was to compare the biological activities
of these nonpolyglutamatable compounds with those of
classical glutamate analogues, whose effect on cell
growth is expected to reflect not only their influx and
subsequent binding to DHFR but also their ability to
form polyglutamates of different chain length and the
affinity of these polyglutamates for enzymes other than
DHFR. In contrast, inhibition of cell growth by non-
polyglutamatable analogues is assumed to be mecha-
nistically more straightforward in that it probably
reflects only influx and DHFR binding.
The IC50 values of 15-21 as inhibitors of CCRF-CEM
cell growth are given in Table 1, along with comparison
data for aminopterin, 2, edatrexate, and four additional
compounds (3-7) we had previously tested against other
human tumor cell lines.15,16 As may be seen from the
results, decreasing the number of CH2 groups in the
amino acid side chain (3, 4) and increasing this number
(5) both gave a decrease in potency. Moving the COOH
group to the meta and para positions (6, 7) was likewise
detrimental, particularly in the case of 7, whose IC50
was 37-fold higher than that of 2. These results dem-
onstrated that the in vitro structure-activity patterns
we had noted earlier15,16 with these compounds using
monolayer cultures of A549 non-small-cell lung and
Because of the chiral nature of C-10 in 16 and 17,
these compounds can exist as mixtures of diastereomers,
as is also true for 10-methyl-10-deazaaminopterin and
1
edatrexate.22 Our 500-MHz H NMR spectra indicate
that both diastereomers are in fact present in 16 and
17 in roughly equal proportions, so that the IC50 values
in Table 1 actually represent the averaged contributions
of the 10R and 10S isomers. In the case of 10-methyl-
10-deazaaminopterin and edatrexate, the 10R and 10S
isomers have been synthesized in stereochemically pure

Analogues of the Nonpolyglutamatable Antifolate PT523
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1627
Ta ble 1. Growth Inhibition of CCRF-CEM Human Leukemic Lymphoblasts by PT523 Analogues
compd
X
R
Y
m
IC50 (nM)a
AMT (1)
PT523 (2)
edatrexate
3
4
5
6
NH
NH
CHEt
NH
NH
NH
NH
NH
CH2
CHMe
CHEt
C(Et)CO2Me
NH
H
H
H
H
H
H
H
H
H
H
H
H
b
-
4.4 ( 1.0 (0.34)
1.5 ( 0.39 (1)c
(CH2)3
b
CH2
ortho
-
3.3 ( 0.36 (0.45)
5.1 ( 0.58 (0.29)
5.5 ( 1.6 (0.27)
2.9 ( 0.97 (0.52)
7.5 ( 0.47 (0.20)
55 ( 2.0 (0.027)
0.53 ( 0.07 (2.8)
0.63 ( 0.08 (2.4)e
1.2 ( 0.25 (1.3)e
54 ( 4.9 (0.028)e
1.2 ( 0.22 (1.3)
4.4 ( 1.1 (0.34)
5.0 ( 0.56 (0.30)
ortho
ortho
ortho
meta
para
ortho
ortho
ortho
ortho
ortho
ortho
ortho
(CH2)2
(CH2)3
(CH2)3
(CH2)3
(CH2)3
(CH2)3
(CH2)3
(CH2)3
(CH2)3
CH2SCH2
(CH2)2SCH2
7d
15
16
17d
18d
19
20
21f
(CHdCH)2
H
H
NH
NH
a
Cells were treated continuously with drug as described in the Experimental Section. Unless otherwise noted, each result is the mean
( SD (n ) 3). Numbers in parentheses are normalized relative to PT523 ()1.0). Glutamate side chain. c For 2, n ) 10. For 7, 17 and
b
d
18, n ) 6. e Tested as unresolved 10R,10S mixtures. f Tested as the unresolved D,L mixture.
form,22 but the absolute configuration at C-10 was not
established and the diastereomers of each compound
were simply referred to as l,L and d,L isomers. However
it was noted in growth inhibition assays against L1210
cells that the d,L isomers of 10-methyl-10-deazaami-
nopterin and edatrexate were 1.7- and 3.5-fold more
potent than the respective l,L isomers. If we assume for
the sake of simplicity that the d,L and l,L isomers of 16
and 17 are present in a 1:1 ratio and that their potencies
differ by the same amount as those of the corresponding
glutamates, a straightforward calculation leads to a
predicted IC50 of 0.48 nM for d,L-16 and 0.54 nM for
d,L-17 (i.e., essentially the same as the IC50 of 15).
However, since we do not actually know that the same
differences in activity would be obtained for CCRF-
CEM cells as for L1210, these values must be considered
as merely estimates. While it would have been possible
in principle to synthesize the d,L isomers of 16 and 17
from the d,L isomers of 39 and 55, the laborious
resolution of the latter compounds was not deemed
worthwhile.
The IC50 of the naphthalene analogue 19 was found
to be 1.2 ( 0.22 nM, a value 3.7-fold lower than that of
1 but not distinguishable from that of 2. Although an
IC50 for the corresponding glutamate analogue against
CCRF-CEM cells is unknown, the IC50 of this compound
against another human cell line, HL-60 promyelocytic
leukemia, is reported to be 4.6 ( 1.0 nM as compared
with 8.1 ( 1.0 nM for methotrexate.38 Based on this
indirect comparison, it appears that 19 may be a little
more potent than its glutamate analogue. An interesting
characteristic of the glutamate analogue that makes it
particularly interesting is that it lacks detectable sub-
strate activity for folylpolyglutamate synthetase from
CCRF-CEM cells and thus can be viewed as a nonpoly-
glutamatable antifolate even though it contains a
glutamate side chain. Its greater potency relative to
methotrexate, like that of 2, is probably due to improved
transport and DHFR binding.
The sulfur analogues 20 and 21 were of interest to
test because of their relationship to the previously
synthesized analogues of 2 with CH2, (CH2)2, and (CH2)4
side chains in place of (CH2)3.15,16 It is a commonly held
view that the S atom is bioisosterically equivalent to a
CHdCH group (i.e., it is larger than a CH2 group but
smaller than a CH2CH2 group). On this basis we
expected the side chain of 20 to be slightly shorter than
that of the previously described lysine analogue 5,
whereas the side chain of 21 would be longer. This in
turn would lead to altered distances between the R and
aromatic COOH groups which could affect biological
activity. As shown in Table 1, the IC50 of 20 was 4.4 (
1.1 nM as compared with 2.9 ( 0.97 nM for 5 and 1.5 (
0.39 nM for 2, but unfortunately the difference between
20 and 5 was not large enough to allow us to draw any
conclusion as to whether the small change in chain
length embodied in structure 20 is significant in terms
of cell growth inhibition. However it appeared from the
results that 20 has about the same potency as amino-
pterin but is ca. 3-fold less potent than 2. With respect
to 21, whose IC50 was 5.0 ( 0.56 nM, it may be noted
that the sample tested was a D,L mixture, since the
starting material was racemic. Since the IC50 values of
L- and D-methotrexate against CCRF-CEM cells are
reported to be 13 and >1000 nM,36 respectively, it is
reasonable to assume that the IC50 of the L form of 21
is as low as one-half the IC50 of the D,L mixture, or 2.5
nM. This would make it more potent than either 3 or 4,
as well as more potent than 20, but still not as potent
as 2. We have previously noted that aminopterin
analogues with nine and ten CH2 groups in the side
chain are more cytotoxic than the parent drug, at least
against L1210 mouse leukemia.39 Thus it is possible that
similar elongation of the side chain of 2 to nine or more
CH2 groups would similarly have increased the potency
of 2.
In summary, the results in this study support the
conclusion that, while the position of the phthaloyl

1628 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Rosowsky et al.
of t-BuOK in DMAC was prepared by adding KH (35% w/w in
mineral oil; 105 mg, calculated to contain 36 mg, 0.9 mmol) in
t-BuOH, evaporating the solution to dryness, and taking up
the residue in dry DMAC (5 mL). The slightly cloudy solution
was added to a stirred suspension of phosphonium salt 30 (215
mg, 0.5 mmol)31 in dry DMAC (5 mL) at room temperature.
The intense red color of the ylide formed immediately and
lightened upon addition of the aldehyde 24 (207 mg, 0.5 mmol).
The reaction mixture was stirred at room temperature for 26
h, the solvent was removed by rotary evaporation (vacuum
pump, receiver cooled in dry ice/acetone), and the residue was
redissolved in CHCl3 (100 mL). The solution was washed with
H2O (2 × 50 mL), dried (MgSO4), concentrated to a small
volume, and chromatographed on flash-grade silica gel (11 g,
1.5 × 12 cm) with 98:2 CHCl3-MeOH as the eluent. Homo-
geneous fractions giving a yellow TLC spot with Rf 0.29 (silica
gel, 95:5 CHCl3-MeOH) were pooled and evaporated to obtain
32 as a yellow solid (106 mg, 40%). Recrystallization from a
mixture of CH2Cl2 and Et2O afforded light-yellow crystals: mp
116-117 °C; IR (KBr) ν 3420, 3315, 3220, 2940, 2215, 1770,
1745, 1640, 1610, 1565, 1520, 1490, 1360 cm-1; UV λmax (95%
COOH group and the length of the amino acid side chain
in 2 are important determinants of cytotoxic potency,
changes in the pABA region and 9,10-bridge are well-
tolerated and in some cases can even produce a sever-
alfold increase in cytotoxic potency. Whether these
increases are due to tighter DHFR binding, more
efficient utilization of the RFC for influx into cells, or a
combination of these factors is currently being investi-
gated.
Exp er im en ta l Section
IR spectra were obtained on a Perkin-Elmer model 281
double-beam spectrophotometer and UV spectra on a Varian
model 210 instrument. For the sake of brevity, only IR peaks
with wavenumbers greater than 1200 cm-1 are reported, and
1
very weak peaks and shoulders are omitted. H NMR spectra
were recorded at 60 MHz with a Varian model EM360L
instrument using Me4Si as the reference or at 500 MHz with
a Varian model ML500 instrument. The very broad amide NH
signal in several of the intermediates (e.g., 24) whose 1H NMR
spectrum was recorded at 60 MHz was indistinguishable from
the baseline and thus is not reported. Analytical TLC was
carried out on fluorescent Whatman MK6F silica gel-coated
glass slides, and 254-nm UV illumination was used to visualize
spots. Column chromatography was on Baker silica gel (regu-
lar grade, 60-200 mesh; flash grade, 40-µm particle size) or
on Whatman DE-52 preswollen DEAE-cellulose. Moisture-
sensitive reactions were carried out in solvents that were of
Sure-Seal grade (Aldrich, Milwaukee, WI) or had been stored
over Linde 4A molecular sieves. In the case of THF, the solvent
was distilled from sodium benzophenone ketyl. HPLC separa-
tions were on C18 silica gel radial compression cartridges
(Millipore, Milford, MA; analytical, 5-µm particle size, 5 × 100
mm; preparative, 15-µm particle size, 25 × 100 mm). Solids
were generally dried over P2O5 at 50-80 °C in a vacuum oven
or Abderhalden apparatus. Melting points (not corrected) were
determined on a Fisher-J ohns hot-stage microscope or in open
Pyrex capillary tubes in a Mel-Temp apparatus (Cambridge
Laboratory Devices, Cambridge, MA). Methyl 2-L-amino-5-
phthalimidopentanoate hydrochloride (23‚HCl),16 2,4-diamino-
6-(bromomethyl)pteridine hydrobromide (26‚HBr),27 2,4-diamino-
6-(hydroxymethyl)pteridine (27),29 4-amino-4-deoxy-N10-for-
mylpteroic acid (68),3 and methyl S-(phthalimidomethyl)-D,L-
homocysteinate hydrochloride (70‚HCl)37 were prepared as
previously described. Starting materials 22, 35 and 52 as well
as other reagents and chemicals were purchased from Aldrich,
Milwaukee, WI, or Fisher, Boston, MA. Microanalyses were
performed by Robertson Laboratory, Madison, NJ , or Quan-
titative Technologies, Whitehouse, NJ , and were within (0.4%
of theory unless otherwise indicated.
Meth yl 2-L-[N-(4-F or m ylben zoyl)a m in o]-5-p h th a lim i-
d op en ta n oa te (24). i-BuOCOCl (262 mg, 1.93 mmol) was
added to a stirred solution of 22 (289 mg, 1.93 mmol) and Et3N
(428 mg, 4.24 mmol) in dry DMF (10 mL). The reaction mixture
was stirred at room temperature for 20 min and 23‚HCl (548
mg, 1.75 mmol) was added in portions. After another 30 min
of being stirred at room temperature, the clear solution was
evaporated to dryness. The residue was taken up in CHCl3
and the solution extracted consecutively with 0.1 N HCl, H2O,
and saturated aqueous NaHCO3, then dried (MgSO4) and
evaporated. The product was chromatographed on silica gel
(flash grade, 55 g, 20.5 × 3 cm) with CHCl3 as the eluent.
Fractions giving a single TLC spot with Rf 0.53 (silica gel, 95:5
CHCl3-MeOH) were pooled and evaporated to obtain 24 as
colorless crystals (400 mg, 56%): mp 166-167 °C; IR (KBr) ν
3460, 3320, 2950, 2850, 1770, 1745, 1690-1710, 1680, 1640,
1570, 1530, 1500, 1460, 1450, 1430, 1400, 1380, 1350, 1330,
1300 cm-1; 1H NMR (CDCl3) δ 1.6-2.3 (m, 4H, â- and γ-CH2),
3.75 (s, 5H, OMe and δ-CH2), 4.8 (m, 1H, R-CH), 7.7-7.9
(complex m, 8H, aromatic protons), 10.1 (s, 1H, CHdO). Anal.
(C22H20N2O6) C, H, N.
1
EtOH) 233, 242, 326 nm; H NMR (CDCl3) δ 1.7-2.2 (m, 4H,
â- and γ-CH2), 3.75 (t, 2H, δ-CH2), 3.76 (s, 3H, OMe), 4.87 (m,
1H, R-CH), 5.25 (s, 2H, NH2), 6.84 (d, 1H, vinyl H), 7.07 (d,
1H, vinyl H), 7.5-7.9 (m, 8H, aromatic), 8.27 (s, 1H, pyrazine
6-H). Anal. (C18H24N6O5‚0.4H2O) C, H, N.
Meth yl 2-L-[4-[2-(2,4-Dia m in op ter id in -6-yl)vin yl]ben -
zoyl]a m in o-5-p h th a lim id op en ta n oa te (28). Br2 (240 mg,
1.5 mmol) was added dropwise to a stirred solution of Ph3P
(393 mg, 1.5 mmol) in DMAC (5.5 mL, dried over molecular
sieves) while keeping the temperature at 0-5 °C. After 5 min,
2,4-diamino-6-hydroxymethylpteridine (27) (97 mg, 0.5 mmol)
was added and stirring was continued at room temperature
for 20 h. Additional DMAC (4 mL) and Ph3P (138 mg, 0.53
mmol) were then added and the solution was heated at 64-
66 °C for 2 h. In a separate flask, a solution of t-BuOK was
prepared in t-BuOH (3 mL) by adding KH (35% w/w in mineral
oil, 458 mg; calculated to contain 160 mg, 4 mmol), and when
H2 evolution ceased the alcohol was evaporated under reduced
pressure and replaced with dry DMAC (3 mL). The contents
of the first flask were added to the second flask with the help
of an additional small volume of DMAC to effect transfer.
Aldehyde 24 (207 mg, 0.5 mmol) was then added, whereupon
partial decolorization occurred. After 15 min the reaction
mixture was diluted with dry THF (5 mL), stirring at room
temperature was continued for 3.5 days, and the reaction was
quenched with glacial AcOH (0.5 mL). The dark reaction
mixture was filtered, the filter cake was washed with MeOH,
and the combined filtrate and wash solution were evaporated
to dryness under reduced pressure. The residue was dispersed
in benzene (10 mL) by sonication and was filtered. The solid
was triturated with CHCl3 (75 mL), and the CHCl3 extract
was decanted from an insoluble black tar and washed with
H2O (75 mL). The aqueous layer was re-extracted with CHCl3
(50 mL), and the combined CHCl3 extracts were washed again
with H2O (75 mL), dried over MgSO4, and evaporated. The
material which formed at the CHCl3-H2O interface was
collected and dried in vacuo at 60 °C over P2O5 to obtain a
pale-yellow solid (58 mg) whose TLC contained a strong yellow
spot (Rf 0.87, silica gel, 28:12:1 CHCl3-MeOH-28% NH4OH).
The residue after evaporation of the CHCl3 layer was taken
up by sonication in warm 28:12:1 CHCl3-MeOH-28% NH4-
OH. A total of 25 mL was required to dissolve the sample. The
solution was applied onto a column of silica gel (flash grade,
20 g, 22 × 1.5 cm), which had been prepared with 85:15:1
CHCl3-MeOH-28% NH4OH and was eluted with the same
solvent. Homogeneous fractions (TLC: Rf 0.25, silica gel 9:1
CHCl3-MeOH) were pooled and evaporated to obtain 28 as a
pale-yellow solid (102 mg), thus bringing the total yield to 160
mg (54%). A sample for microanalysis, assumed to contain cis
and trans isomers, was obtained by recrystallization from a
CHCl3-Me2CO-Et2O mixture and melted at 148-155 °C: IR
(KBr) ν 3380, 3200, 2950, 1770, 1740, 1710, 1640-1620 br,
Meth yl 2-L-[4-[2-(2-Am in o-3-cya n op yr a zin -2-yl)vin yl]-
ben zoyl]a m in o-5-p h th a lim id op en ta n oa te (32). A solution
1580, 1560, 1500, 1440, 1400, 1365, 1310 cm-1 1H NMR
;

Analogues of the Nonpolyglutamatable Antifolate PT523
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1629
(DMSO-d6) δ 1.6-1.9 (m, 4H, δ- and γ-CH2), 3.61 (m, 5H, OMe
and δ-CH2), 4.45 (m, 1H, R-CH), 6.82 (s, 2H, NH2), 6.6-8.5
(complex m, 10H, vinyl and aromatic H), 8.73 (d, 0.33H, 7-H
of one vinyl isomer), 8.85 (s, 0.66H, 7-H of other vinyl isomer).
Anal. (C19H26N9O6‚1.6H2O) C, H, N.
NaOH, warmed back to 50 °C, and treated dropwise with a
solution of crude 39 (calculated to contain 8.5 g, 0.03 mol) in
AcOH (50 mL). An insoluble gum formed during the reaction.
Stirring at 50 °C was continued for 1.5 h, and the supernatant
was poured off from the gum, left to cool to room temperature,
and treated with small portions of freshly prepared 5 N KI3
solution until no more color change was seen. The mixture was
then left in the refrigerator for 24 h, and the solid was
collected, washed with H2O and Et2O, and stirred for 2 h in
H2O (60 mL) containing 28% NH4OH (1.5 mL). Filtration and
acidification of the filtrate with 10% AcOH led to precipitation
of 34 as a yellow solid, which was collected, washed with H2O,
and dried; yield 0.23 g (1% from p-toluic acid). This material
was used directly for subsequent coupling to 23‚HCl (vide
infra).
Meth yl Nr-[4-[2-(2,4-Dia m in op ter id in -6-yl)vin yl]ben -
zoyl]-Nδ-h em ip h th a loyl-L-or n ith in a te (33). Metallic Na (14
mg, 0.6 mmol) was added to MeOH (2.5 mL) and the solution
was added to a solution of guanidine hydrochloride (57 mg,
0.6 mmol) in MeOH (15 mL). Amino nitrile 32 (106 mg, 0.2
mmol) was then added, and after being stirred under reflux
for 43 h the mixture was poured into H2O (100 mL). The
solution was adjusted to pH 3.9 with 10% AcOH and left at 5
°C for 2 days. The precipitate was collected by centrifugation,
washed twice with H2O, and dried on a lyophilizer to obtain a
yellow powder (115 mg). The crude product was chromato-
graphed on flash-grade silica gel (5 g, 1.0 × 15.5 cm) with 5:4:1
CHCl3-MeOH-28% NH4OH as the eluent, and homogeneous
fractions (TLC: Rf 0.56, silica, 5:4:1 CHCl3-MeOH-28% NH4-
OH) were pooled and evaporated. The residue was redissolved
in dilute ammonia at pH 9-10, and the pH was adjusted to
4.3 with 10% AcOH to obtain a gelatinous orange precipitate
which became more granular after the acidified mixture was
kept overnight at 5 °C. The solid was collected by centrifuga-
tion, washed with H2O, and dried on a lyophilizer to obtain a
yellow solid (45 mg, 33%) whose elemental analysis was
consistent with a hydrated monoammonium salt of 33: mp
210-214 °C; IR (KBr) ν 3400 br, 2925, 1630, 1550, 1530, 1500,
1440, 1370, 1300, 1270-1240 cm-1; UV λmax (95% EtOH) 323,
334 infl, 406 nm. Anal. (C29H28N8O6‚NH3‚4H2O) C, H, N.
Meth yl 2-L-[4-[2-(2,4-Dia m in op ter id in -6-yl)eth yl]ben -
zoyl]a m in o-5-p h th a lim id op en ta n oa te (29). Meth od A. A
solution of 28 (150 mg, 0.25 mmol) in glacial AcOH (30 mL)
was shaken in the presence of PtO2 (30 mg) under H2 (initial
pressure 15 lb/in.2) for 19 h. After filtration of the catalyst
through Celite, the solvent was evaporated to dryness with
the help of EtOH to coevaporate the last traces of entrapped
AcOH. The residue was taken up in a mixture of absolute
EtOH (40 mL) and 1 N HCl (1 mL), and the solution was
treated with 30% H2O2 (0.5 mL) and left in an open flask for
1.75 h. Excess Et3N was added, the solution was evaporated
to dryness under reduced pressure, the residue was taken up
in a mixture of absolute EtOH and CHCl3, the solution was
evaporated again, and the final residue was dissolved in a
small volume of 95:5 CHCl3-MeOH and applied onto a column
of flash grade silica gel (6 g, 15 × 1.0 cm) which was prepared
and eluted with the same solvent mixture. Homogeneous
fractions (TLC: Rf 0.32, blue-fluorescent spot, 9:1 CHCl3-
MeOH) were pooled and evaporated to a yellow powder 29 (124
mg, 81%). The analytical sample, prepared by recrystallization
of a small portion of this solid from a mixture of MeOH and
i-PrOH, melted at 162 °C dec (softening above 131 °C): IR
(KBr) ν 3410, 2930, 2940, 2740, 2680, 2500, 1730, 1710, 1635,
For further elaboration to 15, this material was combined
with 28 and subjected to catalytic hydrogenation, reoxidation,
and alkaline hydrolysis as described for the synthesis of 15
via the ring-closed intermediate 29 (vide infra).
4-[2-(2,4-Dia m in op ter id in -6-yl)eth yl]ben zoic Acid (34).
Step 1. 3-Methoxypropyne (35; 20 g, 0.28 mol) was added to
t-BuOK (0.9 g, 0.008 mol) which had been dried in vacuo at
60 °C over P2O5. The mixture was refluxed under N2 at 70 °C
for 4 h, and allene 36 (17 g, 83%) was isolated by short-path
distillation into a receiver cooled in dry ice. A solution of
gaseous HCl (2.7 g, 0.074 mol) in dry Et2O (45 mL) was added
dropwise under N2 to a solution of 36 (5.2 g, 0.074 mol) in dry
Et2O (26 mL) at -78 °C. The reaction mixture was stirred at
-78 °C for 30 min, and the resulting solution of 37 was kept
at 0-5 °C for 40 h until it could be used in the next step.
1560-1540, 1500, 1480, 1490, 1400 cm-1 1H NMR (DMSO-
;
d6) δ 1.66-1.78 (m, 4H, â- and γ-CH2), 3.2-3.3 (m, 4H, bridge
CH2CH2), 3.59 (s, 5H, δ-CH2 and COOMe), 4.4 (m, 1H, R-CH),
6.2 (br s, 2H, NH2), 7.3 (m, 2H, 3′- and 5′-H), 7.6 (bs s, 2H,
NH2), 7.7 (m, 2H, 2′- and 6′-H), 7.8 (m, 4H, phthaloyl ring
protons), 8.52 (s, 1H, pteridine 7-H), 8.6 (d, 1H, CONH). Anal.
(C29H28N8O5‚2H2O) C, H, N.
Meth od B. A mixture of 34 (270 mg, 0.87 mmol) and Et3N
(175 mg, 1.74 mmol) in dry DMF (15 mL) was stirred at 80 °C
until all the solid dissolved. The solution was cooled to 0-5
°C, treated dropwise with i-BuOCOCl (238 mg, 1.74 mmol) and
stirred for 1.5 h. To the solution were then added 23‚HCl (540
mg, 1.74 mg) and Et3N (175 mg, 1.74 mmol). Stirring was
continued at 0-5 °C for 2 h, then at room temperature for 24
h, and the reaction mixture was evaporated to dryness under
reduced pressure. The residue was dissolved in a minimum of
9:1 CHCl3-MeOH, and the solution was applied onto a column
of flash-grade silica gel (20 g, 2 × 20 cm), which was eluted
with 97:3 CHCl3-MeOH. Fractions giving a single TLC spot
with Rf 0.32 (silica gel, 9:1 CHCl3-MeOH) were pooled and
evaporated, the residue was redissolved in a minimal volume
of 9:1 CHCl3-MeOH, and the solution was added dropwise to
excess ether. The precipitate was filtered, washed with Et2O,
and dried at room temperature in vacuo to obtain 29 as a
yellow powder (119 mg, 25%). Melting points and IR and 1H
NMR spectra of the products obtained by methods A and B
were indistinguishable and are given above.
Nr-[4-[2-(2,4-Dia m in op t er id in -6-yl)et h yl]b en zoyl]-Nδ-
h em ip h th a loyl-L-or n ith in e (15). Meth od A. A solution of
29 (44 mg, 0.072 mmol) in warm MeOH (30 mL) was cooled to
room temperature, treated with 1 N NaOH (0.75 mL), and
stirred for 1.75 h. The pH was adjusted to pH 5.2 with 10%
AcOH, and the mixture was left at 5 °C for 1 h and filtered.
The solid was washed with H2O (2 mL) and dried on a
lyophilizer to obtain 15 as a pale-yellow solid (24 mg, 53%):
mp >210 °C dec (darkening above 175 °C); HPLC 15 min (C18
Step 2. A solution of n-BuLi in hexanes (87.5 mL, calculated
to contain 0.14 mol) was added dropwise under N2 to a stirred
solution of i-Pr2NH (14 g, 0.14 mol, freshly distilled from NaH)
in dry THF (205 mL) while maintaining the temperature at
0-5 °C. After 30 min, a solution of p-toluic acid (9.5 g, 0.07
mol) in dry THF (40 mL) was added. The dark-red mixture
was stirred at 0-5 °C for 3.5 h and after being kept in the
refrigerator for 22 h was treated dropwise at 0-5 °C under
N2 with the solution of 37 prepared above. The red color faded
gradually over 2 h, whereupon the solvents were evaporated
and the residue was partitioned between Et2O (150 mL) and
H2O (200 mL). The aqueous layer was chilled and adjusted to
pH 8-9 with dry ice. To the resulting solution of enol ether
38 were then added CH2Cl2 (200 mL), followed dropwise by a
1 M solution of Br2 in CH2Cl2 while maintaining a pH of 7-8
with solid NaHCO3. Addition was continued until persistence
of the red color of Br2 was noted, whereupon the mixture was
acidified to pH 2 with 6 N HCl, and the organic layer was
separated. The aqueous layer was washed with CH2Cl2, and
the combined organic layers were dried (MgSO4) and evapo-
rated to a semisolid (13 g, 61%) whose 1H NMR showed it to
be a 62:38 mixture of the desired R-bromo aldehyde 39 and
unchanged p-toluic acid. The mixture was used in the next
step without further purification.
Step 3. A mixture of 2,4,5,6-tetraaminopyrimidine sulfate
(7.2 g, 0.03 mol) and BaCl2‚2H2O (7.3 g, 0.03 mol) in H2O (145
mL) was stirred at room temperature for 1.5 h, then warmed
to 70 °C in a water bath, and filtered while hot. The filtrate
was cooled to room temperature, adjusted to pH 3-5 with 10%

1630 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Rosowsky et al.
silica gel, 10% MeCN in 0.1 M NH4OAc, pH 6.8, 1 mL/min);
IR (KBr) ν 3400 br, 2930, 2860, 1645 br, 1565-1535, 1505,
1450, 1400 cm-1; UV λmax (pH 7.4) 230 infl (ꢀ 23200), 255
immersed in an oil bath preheated to 135 °C. The flask was
fitted with a reflux condenser and kept in the bath while the
progress of the reaction was monitored by HPLC (C18 silica
gel, 25% MeCN in 0.1 M NH4OAc, pH 7.5, 1 mL/min). The
peak eluting at 8 min, corresponding to 43, was gradually
replaced by new peak eluting at 18 min. After 6 h the reaction
was stopped and the solution was cooled to room temperature
and diluted with H2O. The precipitate was fitered, washed with
H2O, and redissolved in 10% NH4OH. The pH was adjusted
to 4.0 with 10% AcOH while cooling in ice, and the precipitate
was collected, washed with H2O, and dried in vacuo at 60 °C
to obtain 40 as a yellow powder (15 mg, 57%) which was
combined with another larger batch and used directly in the
next reaction: mp >200 °C dec; IR (KBr) ν 3450, 3200, 2960,
1630, 1555, 1450, 1375, 1260 cm-1; 1H NMR (DMSO-d6) δ 1.24
(d, 3H, CHMe), 3.1 (m, 2H, CH2), 3.4 (m, 1H, CHMe), 6.5 (br
s, 2H, NH2), 7.3 (m, 2H, 3′- and 5′-H), 7.8 (m, 2H, 2′- and 6′-
H), 8.4 (s, 1H, pteridine 7-H).
1
(29500), 280 infl (5700), 370 (7100) nm; H NMR (DMSO-d6)
δ 1.4-1.6 (m, 4H, â- and γ-CH2), 2.8-3.2 (m, 6H, CH2CH2
bridge and δ-CH2), 4.36 (m, 1H, R-CH), 6.55 (m, 2H, NH2,
exchangeable with D2O), 7.32 (m, 2H, 3′- and 5′-H), 7.34-7.60
(m, 4H, phthaloyl H), 7.6 (m, 2H, NH2, exchangeable with
D2O), 7.77 (m, 2H, 2′- and 6′-H), 8.28 (t, 1H, phthaloyl CONH,
exchangeable with D2O), 8.46 (d, 1H, benzoyl CONH), 8.49 (s,
1H, pteridine 7-H).
Meth od B. A solution of 29 (50 mg, 0.09 mmol) in DMSO
(0.5 mL) was treated with 2 N NaOH (0.13 mL) for 5 min at
room temperature. After exactly 5 min, the solution was
diluted with 7.5 mL of H2O, cooled to 5 °C, and adjusted to
pH 4.3 with 10% AcOH. The precipitated solid was filtered,
washed with H2O, and dried in vacuo at 50 °C to obtain 15 as
a pale-yellow solid (30 mg, 58%): mp >220 °C, darkening
above 175 °C. Melting points and IR and 1H NMR spectra of
the product obtained via methods A and B were indistinguish-
able and are listed above. Anal. (C28H28N8O6‚0.5CH3COOH‚
0.9H2O) C, H, N.
Meth yl 2-L-[4-[2-(2,4-Dia m in op ter id in -6-yl)p r op yl]ben -
zoyl]a m in o-5-p h t h a lim id op en t a n oa t e (45). St ep 1. Di-
methyl homoterephthalate (41) (6 g, 28.8 mmol) was added to
a suspension of NaH (60% w/w in mineral oil; 1.27 g, calculated
to contain 762 mg, 31.7 mmol) in dry THF (150 mL) at 0 °C.
After 1 h of stirring, MeI (4.5 g, 31.7 mmol) was added, and
stirring was continued at 0 °C for 30 min and then at room
temperature for 16 h. The reaction mixture was treated with
50% AcOH (2.9 mL) and poured into H2O (300 mL). The
product was extracted into Et2O (3 × 100 mL), the combined
extracts were dried (MgSO4) and evaporated, and the residue
was purified by flash chromatography on silica gel. The column
was packed with 95:5 hexanes-EtOAc and then eluted se-
quentially with 95:5, 94:6, and finally 93:7 hexanes-EtOAc.
Fractions giving a single TLC spot (Rf 0.33, silica gel, 92:8
hexanes-EtOAc) were pooled and concentrated to obtain 42
as a colorless oil which was used directly in the next reaction
(3.56 g, 56%): 1H NMR (CDCl3) δ 1.55 (d, 3H, CHMe), 3.6 (m,
4H, COOMe and CHMe), 3.9 (s, 3H, aromatic COOMe), 7.5-
8.1 (m, 4H, aromatic protons).
Step 2. A solution of diester 42 (2.40 g, 10.8 mmol) in dry
DMF (10 mL) was added dropwise to a stirred suspension of
NaH (60% w/w in mineral oil; calculated to contain 259 mg,
10.8 mmol) in dry DMF (10 mL) precooled to -5 °C. The
reaction mixture was stirred at 0 °C for 5 min, then cooled to
-35 °C, and treated dropwise with a solution of 26‚HBr (1.4
g, 3.6 mmol) in dry DMF (20 mL), while keeping the temper-
ature between -45 and -35 °C during addition. The temper-
ature was allowed to rise to 10 °C and stirring was continued
for 2.5 h. The pH was adjusted to 7.0 with dry ice, and the
DMF was removed by rotary evaporation (vacuum pump, dry
ice/acetone). The residue was washed with EtOAc and H2O
and dried in vacuo at 60 °C to obtain 43 as a yellowish-brown
powder (1.22 g, 29%) which was used without further purifica-
tion: mp 170 °C, darkening above 155 °C; IR (KBr) ν 3450,
3310, 3150, 2950, 1720, 1625, 1560, 1440, 1280 cm-1; 1H NMR
(DMSO-d6) δ 1.5 (s, 3H, CHMe), 3.6 (m, 5H, CH2 and benzylic
COOMe), 3.8 (s, 3H, aromatic COOMe), 7.5-8.8 (m, 4H,
aromatic protons), 8.35 (s, 1H, pteridine 7-H).
Step 5. A mixture of 40 (296 mg, 0.91 mmol) and Et3N (184
mg, 1.82 mmol) was dissolved in dry DMF (15 mL) with slight
warming. The solution was cooled to 0-5 °C, i-BuOCOCl (249
mg, 1.82 mmol) was added, and the reaction mixture was
stirred for 1.5 h. Phthalimide 23‚HCl (567 mg, 1.82 mmol) was
then added, followed by another portion of Et3N (184 mg, 1.82
mmol). After another 2 h at 0-5 °C, the reaction mixture was
stirred at room temperature for 24 h and concentrated to
dryness by rotary evaporation (vacuum pump, receiver cooled
in dry ice/acetone). The residue was dissolved in CHCl3, the
solution washed with H2O, and the organic layer dried
(MgSO4) and chromatographed on silica gel (20 g, 2 × 20 cm)
with 98:2 followed by 97:3 CHCl3-MeOH as the eluents.
Fractions giving a single yellow spot with Rf 0.37 (silica gel,
9:1 CHCl3-MeOH) were pooled and evaporated to dryness, the
residue was taken up in a minimal volume of 9:1 CHCl3-
MeOH, and the solution was added to excess Et2O. The precipi-
tated solid was filtered, washed with Et2O, and dried in vacuo
at 60 °C to obtain 45 as a yellow powder (100 mg, 19%): mp
135 °C, darkening above 115 °C; IR (KBr) ν 3420, 2950, 1710,
1620, 1560, 1540, 1500, 1440, 1400, 1350 cm-1 1H NMR
;
(DMSO-d6) δ 1.25 (m, 3H, CHMe), 1.60-1.77 (m, 4H, â- and
γ-CH2), 3.08 (m, 2H, 9-CH2), 3.50 (m, 1H, CHMe), 3.57 (m,
5H, δ-CH2 and OMe), 4.40 (m, 1H, R-CH), 6.5 (br s, 2H, NH2),
7.33 (m, 2H, 3′- and 5′-H), 7.77 (m, 2H, 2′- and 6′-H), 7.82-
7.84 (m, 4H, phthaloyl ring protons), 8.30 (d, 1H, pteridine
7-H), 8.57 (d, 1H, CONH). Anal. (C30H30N8O5‚H2O) C, H, N.
Nr-[4-[2-(2,4-Dia m in op ter id in -6-yl)p r op yl]ben zoyl]-Nδ-
h em ip h th a loyl-L-or n ith in e (16). Ester 45 (50 mg, 0.086
mmol) was dissolved in DMSO (0.5 mL), 1 N NaOH (0.26
mL, 0.26 mmol) was added, and the reaction mixture was
stirred at room temperature for exactly 5 min and then diluted
with H2O (8 mL). The pH was adjusted to 4.0 with 10%
AcOH, and the precipitate was filtered, washed with H2O,
and dried in vacuo at 60 °C to obtain 16 as a pale-yellow
powder (31 mg, 61%): mp >190 °C dec; HPLC 15 min (C18
silica gel, 0.1 M NH4OAc, pH 6.8, 1 mL/min); IR (KBr) ν 3420,
2950, 1640, 1550, 1500, 1450, 1400, 1360, 1300 cm-1; 1H NMR
(DMSO-d6) δ 1.25 (d, 3H, CHMe), 1.56 (m, 2H, â-CH2), 1.8 (m,
2H, γ-CH2), 3.08 (m, 2H, 9-CH2), 3.19 (m, 2H, δ-CH2), 3.47
(m, 1H, CHMe), 4.35 (m, 1H, R-CH), 6.6 (br s, 2H, NH2), 7.33
(m, 2H, 3′- and 5′-H), 7.47-7.76 (m, 5H, phthaloyl ring pro-
tons and phthaloyl CONH), 7.77 (m, 2H, 2′- and 6′-H), 8.37
(d, 1H, pteridine 7-H), 8.4 (d, 1H, CONH). Anal. (C29H30N8O6‚
H2O) C, H, N.
Meth yl 2-L-[4-[1-(2,4-Dia m in op ter id in -6-yl)-2-(m eth -
oxyca r bon yl)-2-bu tyl]ben zoyl]a m in o-5-p h th a lim id op en -
ta n oa te (50). Step 1. Dimethyl homoterephthalate (41; 9 g,
43 mmol) was added to a suspension of NaH (60% w/w in
mineral oil; 3.2 g, calculated to contain 1.9 g, 48 mmol) in dry
THF (225 mL) at 0 °C. After 1 h of stirring, ethyl iodide (7.4
g, 48 mmol) was added, and stirring was continued at 0 °C for
1 h and then at room temperature for 16 h. The reaction
mixture was treated with 50% AcOH (4.4 mL) and poured into
H2O (300 mL). The product was extracted into Et2O (3 × 125
mL), the combined extracts were dried (MgSO4) and evapo-
Step 3. Diester 43 (1 g, 2.5 mmol) was dissolved in DMSO
(5 mL) with slight warming and 5 N NaOH (5 mL, 25 mmol)
was added. The reaction was stirred at room temperature for
5 min, diluted with H2O (50 mL), and adjusted to pH 4.0 with
10% AcOH while cooling in an ice bath. The precipitate was
filtered, washed with H2O, and dried in vacuo at 60 °C to
obtain 44 as a yellow powder (670 mg, 73%): mp >220 °C dec;
IR (KBr) ν 3350, 2950, 1640, 1560, 1380, 1250 cm-1; 1H NMR
(DMSO-d6) δ 1.6 (s, 3H, CHMe), 3.6 (m, 2H, CH2), 6.5 (br s,
2H, NH2), 7.5-8.0 (m, 4H, aromatic protons), 8.4 (s, 1H,
pteridine 7-H).
Step 4. A portion of the foregoing solid (30 mg, 0.08 mmol)
was dissolved in DMSO (0.2 mL) and the mixture was

Analogues of the Nonpolyglutamatable Antifolate PT523
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1631
rated, and the residue was purified on a silica gel column
packed with 95:5 heptane-EtOAc and eluted with 95:5, 94:6,
and finally 93:7 heptane-EtOAc. Fractions giving a single
TLC spot (Rf 0.25, silica gel, 92:8 heptane-EtOAc) were pooled
and concentrated to obtain 46 as a colorless oil which was used
directly in the next step (5.2 g, 51%): 1H NMR (CDCl3) δ 0.9
(t, 3H, CH2CH3), 2.0 (m, 2H, CH2CH3), 3.6 (m, 4H, aliphatic
COOMe and CHEt), 3.9 (s, 3H, aromatic COOMe), 7.4-8.0 (m,
4H, aromatic protons).
Step 2. A solution of diester 46 (5.2 g, 22 mmol) in dry DMF
(18 mL) was added dropwise to a stirred suspension of NaH
(60% w/w in mineral oil, calculated to contain 880 mg, 22
mmol) in dry DMF precooled to -5 °C. The reaction mixture
was stirred at 0 °C for 5 min, then cooled to -35 °C, and
treated dropwise with a solution of 26‚HBr (2.9 g, 7.3 mmol)
in dry DMF (38 mL) while keeping the temperature between
-45 and -35 °C during addition. The temperature was allowed
to rise to 10 °C and stirring was continued for 2.5 h. Excess
NaH was destroyed by addition of dry ice, and the DMF was
removed by rotary evaporation (vacuum pump, dry ice/
acetone). The residue was partitioned between EtOAc and
H2O, and the organic layer was dried (MgSO4) and evaporated
to a yellowish-brown powder (47; 2.14 g, 71%) which was used
without further purification: 1H NMR (DMSO-d6) δ 0.9 (t, 3H,
CH2CH3), 2.0 (m, 2H, CHCH2CH3), 3.6 (m, 4H, aliphatic
COOMe and CHCH2), 3.9 (s, 3H, aromatic COOMe), 6.5 (br s,
2H, NH2), 8.3 (s, 1H, pteridine 7-H), 7.4-8.0 (m, 4H, aromatic
protons).
then diluted with H2O (8 mL) and adjusted to pH 4.0 with
10% AcOH. The precipitate was filtered, washed with H2O,
and dried in vacuo at 60 °C to obtain 18 as a pale-yellow
powder (31 mg, 61%): mp >190 °C dec, darkening above 170
°C; HPLC 30 min (C18 silica gel, 12% MeCN in 0.1 M NH4-
OAc, pH 6.8, 1 mL/min); IR (KBr) ν 3420, 2950, 1715, 1640,
1550, 1500, 1450, 1360, 1300, 1220 cm-1; 1H NMR (DMSO-d6)
δ 0.75 (m, 3H, CH2CH3), 1.58 (m, 2H, â-CH2), 1.89-1.95 (m,
4H, CH2CH3 and γ-CH2), 3.2 (m, 2H, δ-CH2), 3.55 (m, 2H,
9-CH2), 3.59 (s, 3H, aliphatic COOMe), 4.38 (m, 1H, R-CH),
6.5 (br s, 2H, NH2), 7.33 (m, 2H, 3′- and 5′-H), 7.33-7.78 (m,
4H, phthaloyl ring protons), 7.84 (m, 2H, 2′- and 6′-H), 8.15
(d, 1H, pteridine 7-H), 8.7 (d, 1H, CONH); MS m/z 659 (M +
1). Anal. (C32H32N8O8‚1.5H2O) C, H, N.
4-(Meth oxyca r bon yl)p r op iop h en on e (51) Step 1. A
mixture of p-bromopropiophenone (52; 5 g, 23 mmol) and
CuCN (2.4 g, 27 mmol) in dry DMF (6 mL) was refluxed for 4
h, then poured into H2O (15 mL) containing FeCl3 (9.2 g) and
12 N HCl (2.3 mL). The mixture was kept in a water bath at
70 °C for 20 min and left at room temperature overnight. The
precipitated solid was filtered and dissolved in toluene, and
the solution was washed with 1 N HCl, H2O, and 10% NaOH.
The organic layer was dried (MgSO4) and evaporated to obtain
4-propionylbenzonitrile (53; 2.9 g, 79%).
Step 2. The foregoing nitrile (5 g, 31 mmol) was dissolved
in 2-methoxyethanol (100 mL), 2 N NaOH (100 mL) was added,
the solution was heated in an oil bath at 100 °C for 15 h, and
the reaction mixture was cooled in ice and adjusted to pH 7-8
with 12 N HCl. The precipitate was filtered, washed with H2O,
and redissolved in dilute ammonia at pH 8-9. A small amount
of undissolved material was filtered off, and the filtrate was
adjusted to pH 2 with 12 N HCl. The solid was collected,
washed, and dried in vacuo to obtain 4-carboxypropiophenone
(54; 4.4 g, 78%): mp 172-178 °C dec, softening above 165 °C
(lit.23 mp 157-158 °C, lit.40 mp 176-178 °C).
Step 3. A solution of 54 (4.4 g, 25 mmol) in dry DMF (30
mL) was treated with i-Pr2NEt (4.1 g, 32 mmol) and MeI (4.9
g, 35 mmol), and the mixture was stirred at room temperature
for 40 h. The solvent was removed by rotary evaporation
(vacuum pump, dry ice/acetone), and the residue was stirred
with 3% Na2CO3 (40 mL), filtered, washed with H2O, dried in
vacuo, and recrystallized from MeOH to obtain 51 as a
colorless solid (4.2 g, 89%): mp 80-81 °C (lit.23 mp 80-81 °C).
This product was used directly in the next reaction.
Step 3. Diester 47 (2 g, 4.9 mmol) was dissolved in DMSO
(9 mL) and 5 N NaOH (9.8 mL, 49 mmol) was added. The
solution was stirred at room temperature for 5 min, diluted
with H2O (50 mL), then cooled in ice and adjusted to pH 4.0
with 10% AcOH. The precipitate was filtered, washed with
H2O, and dried in vacuo at 60 °C to obtain 49 as a yellow
powder (1.12, 60%) which was used without further purifica-
tion: mp 140 °C, darkening above 120 °C; IR (KBr) ν 3350,
2950, 1710, 1610, 1560, 1540, 1500, 1450, 1390, 1360 cm-1
;
1H NMR (DMSO-d6) δ 0.75 (m, 3H, CH2CH3), 1.90 (m, 2H,
CHCH2CH3), 3.61 (s, 3H, aliphatic COOMe), 7.33 (m, 2H, 3′-
and 5′-H), 7.77 (m, 2H, 2′- and 6′-H), 8.18 (d, 1H, pteridine
7-H); MS m/z 397 (M + 1). More vigorous hydrolysis conditions
were tried in order to convert 47 to 48 but led to complex
mixtures of difficultly separable products.
Step 4. Acid 49 (280 mg, 0.70 mmol) was dissolved in dry
DMF (15 mL) with gentle warming, and the solution was
cooled to 0-5 °C and treated dropwise with i-BuOCOCl (226
mg, 1.66 mmol) followed by Et3N (168 mg, 1.66 mmol). After
1.5 h of stirring, 23‚HCl (510 mg, 1.66 mmol) was added, and
stirring was resumed at 0-5 °C for 2 h and continued at room
temperature for 24 h. The solvent was removed by rotary
evaporation (vacuum pump, receiver cooled in dry ice), and
the residue was taken up in CHCl3. The solution was washed
with H2O, dried (MgSO4), and applied onto a silica gel column
(20 g, 2 × 20 cm) which was eluted successively with 98:2 and
97:3 CHCl3-MeOH. Fractions giving a single TLC spot with
Rf 0.4 (silica gel, 9:1 CHCl3-MeOH) were pooled and evapo-
rated, and the residue was taken up in a minimum of 95:5
CHCl3-MeOH. The solution was added to a large excess of
Et2O, and the precipitated solid was filtered, washed with
Et2O, and dried in vacuo at 60 °C to obtain 50 as a yellow
solid (87 mg, 20%): mp 140 °C, darkening above 120 °C; IR
(KBr) ν 3350, 2950, 1710, 1610, 1560, 1540, 1500, 1450, 1390,
1360, 1220 cm-1; 1H NMR (DMSO-d6) δ 0.75 (m, 3H, CH2CH3),
1.60-1.80 (m, 4H, â- and γ-CH2), 1.95 (m, 2H, CH2CH3), 3.56-
3.61 (m, 10H, δ-CH2, 9-CH2, two COOMe groups), 4.42 (m, 1H,
R-CH), 6.5 (br s, 2H, NH2), 7.33 (m, 2H, 3′- and 5′-H), 7.77 (m,
2H, 2′- and 6′-H), 7.82-7.84 (m, 4H, phthaloyl protons), 8.18
(d, 1H, pteridine 7-H), 8.7 (d, 1H, CONH). Anal. (C33H34N8O7‚
0.8H2O) C, H, N.
Met h yl 2-L-[4-[1-(2,4-Dia m in op t er id in -6-yl)-2-b u t yl]-
b en zoyl]a m in o-5-p h t h a lim id op en t a n oa t e (58). St ep 1.
n-Bu3P (1.82 g, 3 mmol) was added to a solution of 26‚HBr (1
g, 3 mmol) in dry DMSO (35 mL) at 55 °C, and the mixture
was stirred under N2 for 30 min and then cooled back to room
temperature. Keto ester 51 (573 mg, 3 mmol) was added,
followed by NaH (60% in mineral oil, calculated to contain 240
mg, 6 mmol). The mixture turned red as H2 was evolved.
Stirring at room temperature was continued for 40 h, and the
solvent was removed by rotary evaporation (vacuum pump,
dry ice/acetone). The residue was stirred in Et2O, and the
orange solid was filtered, washed with H2O, and dried to obtain
55 as a yellow solid (650 mg, 62%) which was used directly in
the next step: mp >235 °C, darkening above 200 °C; MS m/e
351 (M + 1); IR (KBr) ν 3450, 3350, 3200, 2950, 1720, 1615,
1565, 1555, 1440, 1350, 1285 cm-1 1H NMR (DMSO-d6) cis
;
isomer δ 1.04 (t, CH2CH3), 2.57 (q, CH2CH3), 3.84 (s, COOMe),
6.64 (s, vinyl H), 7.34 (d, 3′ and 5′-H), 7.75 (d, 2′- and 6′-H),
8.06 (s, pteridine 7-H), trans isomer δ 1.10 (t, CH2CH3), 3.16
(q, CH2CH3), 3.86 (s, COOMe), 6.98 (s, vinyl H), 7.95 (d, 3′-
and 5′-H), 7.98 (d, 2′- and 6′-H), 8.79 (s, pteridine 7-H).
Step 2. PtO2 (0.16 g) was added to a solution of 55 (600
mg, 1.7 mmol) in glacial AcOH (85 mL) and the mixture was
hydrogenated in a Parr apparatus at (initial pressure 45 lb/
in.2) for 18 h. The reaction mixture was filtered through Celite
and the filtrate treated with 3% H2O2 (2.1 mL) and stirred for
5 h. A second portion of 3% H2O2 (1 mL) was added, and
stirring was resumed for another 15 h. The solution was
concentrated to a volume of 10 mL under reduced pressure,
and H2O (60 mL) was added, followed by dropwise addition
Nr-[4-[1-(2,4-Diam in opter idin -6-yl)-2-m eth oxycar bon yl-
2-bu tyl]ben zoyl]-Nδ-h em ip h th a loyl-L-or n ith in e (18). Di-
ester 50 (50 mg, 0.076 mmol) was dissolved in DMSO (0.5 mL),
1 N NaOH (0.26 mL, 0.26 mmol) was added, and the reaction
mixture was stirred at room temperature for exactly 5 min,

1632 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Rosowsky et al.
28% NH4OH to pH 7.5. The precipitate was filtered, washed
with H2O, and dried in vacuo to obtain 56 as a brown solid
(510 mg, 85%) which was used directly in the next step: mp
>215 °C dec; MS m/e 353 (M + 1); IR (KBr) ν 3450, 3350, 2950,
1720, 1620, 1585, 1560, 1440, 1280 cm-1; 1H NMR (DMSO-d6)
δ 0.73 (t, 3H, CH2CH3), 1.66 (m, 2H, CH2CH3), 3.05-3.18 (m,
3H, 9-CH2 and CHEt), 3.79 (s, 3H, COOMe), 6.48 (br s, 2H,
NH2), 7.33 (m, 2H, 3′- and 5′-H), 7.80 (m, 2′- and 6′-H), 8.32
(s, 1H, pteridine 7-H). The CONH proton and other NH2
protons were not observed.
Step 3. A solution of 56 (500 mg, 1.42 mmol) in DMSO (15
mL) was treated with 5 N NaOH (2.8 mL) and stirred at room
temperature for 10 min. The reaction mixture was diluted to
100 mL with H2O, adjusted to pH 4.0 with 10% AcOH, and
refrigerated overnight. The precipitate was filtered, washed
with H2O, and dried to obtain 57 as a brown solid (325 mg,
68%) which was used directly in the next step: mp >200 °C
dec; MS m/e 339 (M + 1); IR (KBr) ν 3420, 3200, 2980, 2910,
1620, 1555, 1440, 1370, 1260 cm-1; 1H NMR (DMSO-d6) δ 0.72
(t, 3H, CH2CH3), 1.66 (m, 2H, CH2CH3), 3.08 (m, 3H, 9-CH2
and CHEt), 6.49 (br s, 2H, NH2), 7.31 (m, 2H, 3′- and 5′-H),
7.79 (m, 2H, 2′- and 6′-H), 8.32 (s, 1H, pteridine 7-H).
Step 4. Et3N (363 mg, 3.6 mmol) and i-BuOCOCl (125 mg,
0.9 mmol) were added to a stirred solution of 57 (300 mg, 0.89
mmol) in dry DMF (20 mL), followed 30 min later by 23‚HCl
(509 mg, 1.57 mmol). After 18 h the solvent was removed by
rotary evaporation (vacuum pump, dry ice/acetone), and the
residue was taken up in CHCl3. The solution was washed with
H2O, and the organic layer was dried (MgSO4) and evaporated
to dryness. The residue was redissolved in 9:1 CHCl3-MeOH
and chromatographed on silica gel (flash grade, 20 g, 2 × 20
cm) with 98:2 and 97:3 CHCl3-MeOH as the eluents. TLC
homogeneous fractions (Rf 0.38, silica gel, 9:1 CHCl3-MeOH)
were pooled and evaporated, and the residue was redissolved
in a minimum volume of 9:1 CHCl3-MeOH. The solution was
poured into excess Et2O, and the precipitate was filtered,
washed with Et2O, and dried in vacuo at 60 °C to obtain 58 as
a yellow solid (152 mg, 29%): mp 155 °C, softening above 120
°C; IR (KBr) ν 3400, 2950, 1710, 1620, 1550, 1535, 1500, 1440,
1355, 1210 cm-1; 1H NMR (DMSO-d6) δ 0.72 (t, 3H, CH2CH3),
1.61-1.78 (m, 6H, â- and γ-CH2, CH2CH3), 3.06-3.17 (m, 3H,
9-CH2 and CHEt), 3.59 (m, 5H, δ-CH2 and COOMe), 4.38 (m,
1H, R-CH), 6.5 (br s, 2H, NH2), 7.28 (m, 2H, 3′- and 5′-H), 7.5
(br s, 2H, NH2), 7.68 (m, 2H, 2′- and 6′-H), 7.81-7.85 (m, 4H,
phthalimide protons), 8.31 (d, 1H, pteridine 7-H), 8.57 (d, 1H,
CONH). Anal. (C31H32N8O5‚H2O) C, H, N.
Nr-[4-[1-(2,4-Dia m in op t er id in -6-yl)-2-b u t yl]b en zoyl]-
Nδ-h em ip h th a loyl-L-or n ith in e (17). A solution of 58 (20 mg,
0.034 mmol) was dissolved in EtOH (10 mL) containing a few
drops of H2O. To the solution were then added Ba(OH)2‚2H2O
(21 mg, 0.067 mmol) followed 24 h later by NH4HCO3 (53 mg,
0.8 mmol) and the mixture stirred for another 30 min. The
reaction mixture was filtered to remove the BaCO3, the EtOH
was removed on the rotary evaporator, and the remaining
aqueous solution was adjusted to pH 4.0 with 10% AcOH and
refrigerated overnight. The solid was isolated by centrifuga-
tion, washed with H2O, and dried in a lyophilizer and a
vacuum oven at 60 °C to obtain 17 as a yellow solid (14 mg,
70%): mp >200 °C dec; HPLC 26 min (C18 silica gel, 12%
MeCN in 0.1 M NH4OAc, pH 6.8, 1 mL/min); IR (KBr) ν 3400,
2950, 1635, 1530, 1500, 1440, 1370, 1300, 1210 cm-1; 1H NMR
(DMSO-d6) δ 0.7 (t, 3H, CH2CH3), 1.55-1.89 (m, 6H, â- and
γ-CH2, CH2CH3), 3.06-3.26 (m, 5H, 9-CH2, CHEt, and δ-CH2),
4.35 (m, 1H, R-CH), 6.5 (br s, 2H, NH2), 7.29 (m, 2H, 3′- and
5′-H), 7.38-7.52 (m, 3H, phthaloyl 3-, 4-, and 5-protons), 7.63-
7.75 (m, 3H, 3′- and 5′-H, phthaloyl 6-H), 8.27 (m, 1H,
phthaloyl CONH), 8.31 (d, 1H, pteridine 7-H), 8.47 (d, 1H,
benzoyl CONH). Anal. (C30H32N8O6‚0.1CH3COOH‚3.2H2O) C,
H, N.
reaction mixture was refluxed for 50 h and cooled. The
precipitate was filtered, washed with H2O, and suspended in
12 N HCl (85 mL). The mixture was refluxed for 3 h, cooled,
and filtered. The solid was washed with H2O and recrystallized
from AcOH to obtain brown crystals of 4-nitro-1-naphthoic acid
(5.9 g, 66%): mp 218-219 °C (lit.41 mp 225-226 °C). The crude
acid was combined with SOCl2 (25 mL) and 2 drops of dry
DMF, and the reaction mixture was refluxed for 20 h. Excess
SOCl2 was removed by rotary evaporation and the residue was
recrystallized from 1:1 hexanes-Et2O to obtain the crude acid
chloride 59 (4.5 g, 69%). Without further purification the acid
chloride (3 g, 0.013 mol) and 23‚HCl (4 g, 0.013 mol) were
combined in CH2Cl2 (100 mL). The mixture cooled in an ice
bath while Et3N (2.63 g, 0.026 mol) was added dropwise with
stirring. The solids dissolved, and stirring was continued for
1 h at 0 °C and for 20 h at room temperature. The reaction
mixture was washed successively with 0.1 N HCl, H2O, and
saturated aqueous NaHCO3, then dried (MgSO4) and evapo-
rated. Recrystallization from absolute EtOH afforded 60 as
pale-yellow crystals (4.41 g, 71%; 32% overall from 4-nitro-
1,8-naphthoic anhydride): mp 144-145 °C; IR (KBr) ν 3250,
1740, 1705, 1650, 1520, 1400, 1330, 1250, 1200 cm-1; 1H NMR
(DMSO-d6) δ 1.75-1.80 (m, 4H, â- and γ-CH2), 3.61-3.63 (m,
2H, δ-CH2), 3.71 (s, 3H, OMe), 4.5 (m, 1H, R-CH), 7.6 (d, 1H,
naphthoyl 5-H), 7.69-7.85 (m, 6H, naphthoyl 2- and 8-H, 4
phthalimide ring protons), 8.24-8.32 (m, 3H, naphthoyl 6- and
7-H, CONH), 9.1 (d, 1H, naphthoyl 3-H). Anal. (C25H21N3O7)
C, H, N.
Meth yl 2-L-(4-Am in o-1-n a p h th oyl)a m in o-5-p h th a lim i-
d op en ta n oa te (61). Ammonium formate (2.44 g, 0.039 mmol)
and 10% Pd-C (500 mg) were added to a suspension of 60 (4
g, 8.4 mmol) in MeOH. After being stirred at room temperature
for 3.5 h, the reaction mixture was filtered through a bed of
Celite, the filtrate was evaporated to dryness, and the residue
was stirred in cold H2O, filtered, and dried in vacuo at 60 °C
to obtain 61 (0.95 g, 25%): mp 157-160 °C after recrystalli-
zation from EtOAc-cyclohexane; IR (KBr) ν 3435, 2950, 1730,
1710, 1645, 1575, 1515, 1400, 1210 cm-1; 1H NMR (DMSO-d6)
δ 1.74-1.83 (m, 4H, â- and γ-CH2), 3.32-3.68 (m, 5H, δ-CH2
and OMe), 4.4 (m, 1H, R-CH), 6.1 (br s, 2H, NH2), 6.6 (d, 1H,
naphthoyl 3-H), 7.37-7.44 (m, 3H, naphthoyl 2-, 6-, and 7-H),
7.82-7.83 (m, 4H, phthaloyl ring protons), 8.08 (m, 1H,
CONH), 8.38 (m, 1H, naphthoyl 5-H), 8.43 (m, 1H, naphthoyl
8-H). Anal. (C25H23N3O5‚0.2H2O) C, H, N.
Meth yl 2-L-[4-[N-[(2,4-Dia m in op ter id in -6-yl)m eth yl]-
am in o]-1-n aph th oyl]am in o-5-ph th alim idopen tan oate (62).
A solution of 61 (445 mg, 1 mmol) and 26‚HBr (360 mg, 0.9
mmol) in dry DMF (5 mL) was stirred at room temperature
under N2 for 5 days in a flask wrapped in aluminum foil. The
reaction mixture was then added slowly with constant stirring
to a solution of Na2CO3 (110 mg) in H2O (50 mL). The yellow
precipitate was filtered, washed with H2O, dried in a vacuum
oven, and chromatographed on a silica gel column (30 g, 2 ×
28 cm) that was packed and eluted with 91:9 CHCl3-MeOH.
Fractions giving a single TLC spot (Rf 0.36, silica gel, 9:1
CHCl3-MeOH) were pooled and evaporated, the residue was
redissolved in a minimal volume of 9:1 CHCl3-MeOH, and
the solution was poured into a large volume of Et2O. The
precipitate was filtered, washed with Et2O, and dried in vacuo
to obtain 62 as a yellow solid (0.28 g, 51%): mp 165 °C,
darkening above 150 °C; IR (KBr) ν 3430, 2950, 1710, 1610,
1580, 1520, 1440, 1390, 1350, 1250 cm-1; 1H NMR (DMSO-d6)
δ 1.71-1.76 (m, 4H, â- and γ-CH2), 3.59 (m, 2H, δ-CH2), 3.63
(s, 3H, OMe), 4.45 (m, 1H, R-CH), 4.67 (m, 2H, 9-CH2), 6.52
(d, 1H, naphthoyl 3-H), 6.62 (br s, 2H, NH2), 7.41 (d, 1H, 10-
NH), 7.47-7.49 (m, 2H, naphthoyl 6- and 7-H), 7.71 (br s, 2H,
NH2), 7.80-7.83 (m, 4H, phthaloyl ring protons), 8.29-8.32
(m, 2H, naphthoyl 5- and 8-H), 8.46 (d, 1H, CONH), 8.69 (s,
1H, pteridine 7-H). Anal. (C27H29N9O5) C, H, N.
Met h yl 2-L-(4-Nit r o-1-n a p h t h oyl)a m in o-5-p h t h a lim i-
d op en ta n oa te (60). 4-Nitro-1,8-naphthoic anhydride (10 g,
0.041 mol) and a solution of HgO (10 g, 0.046 mol) in a mixture
of H2O (25 mL) and AcOH (8 mL) were added with stirring to
a solution of NaOH (5.4 g, 0.67 mol) in H2O (200 mL). The
Nr-2-L-[4-[N-[(2,4-Dia m in opter idin -6-yl)m eth yl]a m in o]-
1-n a p h th oyl]-Nδ-h em ip h th a loyl-L-or n ith in e (19). Ester 62
(50 mg, 0.081 mmol) was added to EtOH (25 mL) containing
a few drops of H2O, the mixture was solubilized with the help
of a sonication bath, and a solution of Ba(OH)2 (51 mg, 0.16

Analogues of the Nonpolyglutamatable Antifolate PT523
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1633
mmol) in H2O (25 mL) was added. Stirring was continued for
24 h, solid NH4HCO3 (13 mg, 0.16 mmol) was added, and
stirring was continued for another 30 min. The BaCO3
precipitate was filtered and the EtOH removed by rotary
evaporation. The pH was adjusted to 4.0 with 10% AcOH, the
mixture was kept in the refrigerator overnight, and the solid
was filtered, washed with H2O, and dried to obtain 19 as a
yellow solid (36 mg, 71%): mp >200 °C dec, darkening above
180 °C; HPLC 15 min (C18 silica gel, 10% MeCN in 0.1 M NH4-
OAc, pH 6.8, 1 mL/min; IR (KBr) ν 3420, 2950, 1640, 1570,
raphy (13 g, 2 × 9 cm) with 10:1 CHCl3-MeOH as the eluent.
Fractions showing a TLC spot at Rf 0.2 (silica gel, 9:1 CHCl3-
MeOH) were combined and diluted with Et2O until a precipi-
tate formed. The solid was filtered and dried in vacuo at 65
°C over P2O5 to obtain the intermediate phthalimide ester 69
1
(203 mg, >100%). Although the H NMR spectrum contained
peaks consistent with the presence of some residual Et3N‚HCl,
this solid was used without further purification.
Step 2. A solution of crude 69 (200 mg, assumed to contain
the theoretical 0.325 mmol) was dissolved in DMSO (4 mL)
and the solution was treated with 2 N NaOH (1.5 mL, 3 mmol).
After exactly 2 min of stirring at room temperature, the
solution was diluted with H2O (50 mL) and adjusted to pH
7.5 with saturated aqueous (NH4)2SO4. Analytical HPLC (8%
MeCN in 0.1 M NH4OAc, pH 6.9, 1.0 mL/min) showed two
peaks with retention times of 15 and 25 min, respectively, and
a peak height ratio of ca. 2:1. Treatment of an aliquot of the
solution with excess 2 N NaOH at room temperature for 15
min did not alter the ratio. The slower-moving major product
was isolated by preparative HPLC using the same eluent
system. Homogeneous fractions were pooled and freeze-dried,
and the residue was dissolved in dilute ammonia. Acidification
with AcOH yielded a precipitate, which was filtered and dried
in a lyophilization apparatus to obtain 20 as a yellow solid
(24 mg, 7% for two steps): mp >200 °C dec; IR (KBr) ν 3340
br, 1710, 1640, 1600, 1540, 1510, 1445, 1405, 1380, 1325, 1260
cm-1. The identity of the minor product was not investigated.
Anal. (C26H25N9O6S‚CH3COOH‚1.5H2O) C, H, N, S.
N-[4-[(2,4-Dia m in o-6-p ter id in -6-yl)m eth yla m in o]ben -
zoyl]-S-(2-ca r boxyben za m id om eth yl)-D,L-h om ocystein e
(21). Using essentially the same procedure as in the preceding
experiment, 70‚HCl (172 mg, 0.5 mmol) and 68 (170 mg, 0.5
mmol) were converted into phthalimide ester 72. After partial
purification by flash chromatography on a silica gel (13 g, 2 ×
9 cm) with 9:1 CHCl3-MeOH as the eluent, the coupling
product was treated with NaOH in DMSO. Preparative HPLC
(C18 silica gel, 10% MeCN in 0.1 M NH4OAc, pH 6.9) afforded
21 as a yellow solid (57 mg, 19% for two steps): mp >200 °C
dec; HPLC 11 min (C18 silica gel, 10% MeCN in 0.1 M NH4-
OAc, pH 6.9, 1.0 mL/min); IR (KBr) ν 3330 br, 3070, 2930,
1635, 1605, 1535, 1505, 1445, 1400, 1325, 1260 cm-1; 1H NMR
(DMSO-d6 + 1 drop D2O) δ 2.17 (m, 2H, â-CH2), 2.73 (m, 2H,
γ-CH2) (partial overlap by DMSO signal), 4.47 (m, 5H, 9-CH2,
R-CH, and SCH2N), 6.70 (d, J ) 8 Hz, 2H, 3′- and 5′-H), 7.63
(m, 6H, 2′-H, 6′-H, and phthaloyl protons), 8.77 (s, 1H,
pteridine 7-H). Anal. (C27H27N9O6S‚2H2O‚0.5CH3COOH) C, H,
N, S.
Cell Gr ow th In h ibition Assa y. CCRF-CEM human leu-
kemic lymphoblasts (American Type Culture Collection, Rock-
ville, MD) were maintained in suspension culture in 175-cm2
vented flasks at 37 °C in a 5% CO2 humidified atmosphere in
RPMI 1640 medium (Fisher, Boston, MA) containing 10% fetal
bovine serum, 2 mM L-glutamine, 100 U/mL penicillin, and
100 µg/mL streptomycin (Sigma, St. Louis, MO). In the growth
inhibition assays, cells (5.5 × 104/mL) were incubated for 72
h in 24-well microtiter plates containing different concentra-
tions of drugs (0.1-1000 nM in half-log increments) or in the
absence of drug (negative controls). A 0.5-mL aliquot from each
well was suspended in Isoton II (Fisher, Boston, MA) and cell
numbers were determined with a hemocytometer after stain-
ing with trypan blue or eletronically with a Coulter model ZBI
counter gated for a cell diameter of 10-30 µm. In the latter
procedure, background values were obtained by counting cells
that had been treated for 72 h with 1 µM PT523, a concentra-
tion ca. 1000-fold higher than the IC50. Replicate validation
trials showed that these cells were completely permeable to
trypan blue and that >98% of them had a diameter of <10
µm. The survival fraction (SF) was calculated from the
formula: SF ) (treated cell number - background)/(control
cell number - background). A semilog plot of drug concentra-
tion versus SF was generated with Microsoft Excel software,
and the IC50 was determined by interpolation. Each assay
was performed at least three times, and the results were
averaged.
1
1520, 1440, 1370 cm-1; H NMR (DMSO-d6) δ 1.61-1.63 (m,
4H, â- and γ-CH2), 3.19 (m, 2H, δ-CH2), 4.36 (m, 1H, R-CH),
4.66 (m, 2H, 9-CH2), 6.52 (d, 1H, naphthoyl 3-H), 6.53 (br s,
2H, NH2), 7.34 (m, 1H, naphthoyl 2-H), 7.37 (d, 1H, 10-NH),
7.34-7.51 (m, 5H, naphthoyl 6-H and 7-H, and 3 phthaloyl
ring protons), 7.69 (br s, 2H, NH2), 7.71 (m, 1H, 1 phthaloyl
ring proton), 8.25-8.39 (m, 4H, naphthoyl 5′- and 8′-H,
naphthoyl and phthaloyl CONH), 8.69 (s, 1H, pteridine 7-H).
Anal. (C31H29N9O6‚0.4CH3COOH‚H2O) C, H, N.
Meth yl S-(P h th alim idom eth yl)cystein ate Tosylate Salt
(67‚TsOH). Step 1. A suspension of L-cysteine methyl ester
hydrochloride (63‚HCl; 1.08 g, 6.30 mmol) in EtOAc (30 mL)
was treated with di-tert-butyl dicarbonate (1.37 g, 6.30 mmol)
followed by Et3N (0.657 g, 904 µL, 6.50 mmol). After being
refluxed for 20 h, the reaction mixture was cooled and
extracted with 3% citric acid, H2O, 5% NaHCO3, and finally
H2O. Evaporation to dryness and column chromatography on
silica gel (flash grade, 20 g, 2 × 15 cm) with 2:1 hexanes-
EtOAc as the eluent afforded the Boc derivative 65 as an oil
which was used directly in the next step (906 mg, 61%): Rf
1
0.5 (silica gel, 2:1 hexanes-EtOAc); H NMR (CDCl3) δ 1.43
(s, 9H, t-Bu), 2.97 (m, 2H, CH2), 3.75 (s, 3H, OMe), 4.54 (m,
1H, R-CH).
Step 2. A solution of the foregoing oil (306 mg, 1.30 mmol)
in MeOH (10 mL) was treated with N-(bromomethyl)phthal-
imide (313 mg, 1.30 mmol) and i-Pr2NEt (268 mg, 226 µL, 1.30
mmol), and the mixture was refluxed for 30 min, cooled, and
washed with 3% citric acid. Evaporation of the solvent and
column chromatography on silica gel (flash grade, 20 g, 2 ×
15 cm) with 2:1 hexanes-EtOAc as the eluent afforded 66 as
a gum (448 mg, 87%) which was used directly in the next
step: 1H NMR (CDCl3) δ 1.42 (s, 9H, t-Bu), 3.15 (m, 2H,
â-CH2), 3.74 (s, 3H, OMe), 4.65 (m, 1H, R-CH), 4.78 (s, 2H,
SCH2N), 7.83 (m, 4H, aromatic protons). When the reaction
was carried out in EtOH at 65 °C for 15 min the yield was
only 37%, suggesting a marked solvent effect on the rate.
Step 3. A solution of 66 (520 mg, 1.32 mmol) in toluene (25
mL) was treated with p-TsOH‚H2O (257 mg, 1.35 mmol),
refluxed for 10 min, and cooled to room temperature until a
precipitate formed. The solid was filtered, washed with
toluene, and dried in vacuo at 65 °C over P2O5 to obtain 67‚
TsOH as a white solid (527 mg, 86%; 46% overall from 63‚
HCl): mp 160-162 °C; IR (KBr) ν 3460, 1755, 1710, 1610,
1520, 1510, 1465, 1440, 1415, 1385, 1310, 1285, 1215 cm-1
;
1H NMR (DMSO-d6) δ 2.30 (s, 3H, Me of TsOH), 3.21 (m, 2H,
â-CH2), 3.72 (s, 4H, OMe, partially overlapped by H2O signal),
4.35 (m, 1H, R-CH), 4.78 (s, 2H, SCH2N), 7.13 (d, J ) 8 Hz,
2H, 3- and 5-H of TsOH), 7.53 (d, J ) 8 Hz, 2H, 2- and 6-H of
TsOH), 7.93 (s, 4H phthalimide protons), 8.52 (br m, 3H,
NH3+). Anal. (C20H22N2O7S2‚0.5H2O) C, H, N, S.
N-[4-[(2,4-Dia m in o-6-p ter id in -6-yl)m eth yla m in o]ben -
zoyl]-S-(2-car boxyben zam idom eth yl)-L-cystein e (20). Step
1. Et3N (50 mg, 69 µL, 0.5 mmol) and i-BuOCOCl (68 mg, 65
µL, 0.5 mmol) were added to a stirred suspension of 4-amino-
4-deoxy-N10-formylpteroic acid (68; 170 mg, 0.5 mmol) in dry
DMF (10 mL) at room temperature. After 5 min, when the
solution became clear, 67‚TsOH (233 mg, 0.33 mol) and a
second portion of Et3N (50 mg, 69 µL, 0.5 mmol) were added.
Another set of stepwise additions was carried out (23 µL of
Et3N, 22 µL of i-BuOCOCl, and 77 mg of 67‚TsOH), and as
soon as this addition cycle was completed the solvent was
evaporated under reduced pressure and the residue was
swirled with Et2O. The Et2O was decanted and the insoluble
material was partially purified by flash silica gel chromatog-

1634 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Rosowsky et al.
(18) Grant, S. C.; Kris, M. G.; Young, C. W.; Sirotnak, C. W.
Edatrexate: an antifolate with antitumor activity: a review.
Cancer Invest. 1993, 11, 36-45.
(19) Takimoto, C. H.; Allegra, C. J . New antifolates in clinical
development. Oncology 1995, 9, 649-656.
(20) Piper, J . R.; Montgomery, J . A. Convenient synthesis of 10-
deazaaminopterin via a pteridine ylide. J . Med. Chem. 1980, 23,
320-321.
(21) DeGraw, J . I.; Brown, V. H.; Tagawa, H.; Kisliuk, R. L.;
Gaumont, Y.; Sirotnak, F. M. Synthesis and antitumor activity
of 10-alkyl-10-deazaaminopterins. A convenient synthesis of 10-
deazaaminopterin. J . Med. Chem. 1982, 25, 1227-1230.
(22) DeGraw, J . I.; Christie, P. H.; Tagawa, H.; Kisliuk, R. L.;
Gaumont, Y.; Schmid, F. A.; Sirotnak, F. M. Synthesis and
biological activity of resolved C-10 diastereomers of 10-methyl-
and 10-ethyl-10-deazaaminopterin. J . Med. Chem. 1986, 29,
1056-1061.
(23) Piper, J . R.; J ohnson, C. A.; Otter, G. M.; Sirotnak, F. M.
Synthesis and antifolate evaluation of 10-ethyl-5-methyl-5,10-
dideazaaminopterin and an alternative synthesis of 10-ethyl-
10-deazaaminopterin (Edatrexate). J . Med. Chem. 1992, 35,
3002-3006.
Ack n ow led gm en t. This work was supported in part
by Grants CA25394 and CA70349 from the National
Cancer Institute (DHHS). Technical assistance in ob-
taining cytotoxicity data against CCRF-CEM cells was
provided by Dr. Ying-Nan Chen.
Refer en ces
(1) Rosowsky, A. PT523 and other aminopterin analogues with a
hemiphthaloyl-L-ornithine side chain: exceptionally tight-bind-
ing inhibitors of dihydrofolate reductase which are transported
by the reduced folate carrier but cannot form polyglutamates.
Curr. Med. Chem. 1999, 6, 329-352.
(2) Rosowsky, A. Development of nonpolyglutamatable DHFR in-
hibitors. In Antifolate Drugs in Cancer Therapy; J ackman, A.
L., Ed.; Humana Press: Totowa, NJ , 1999; pp 59-100.
(3) Rosowsky, A.; Bader, H.; Cucchi, C. A.; Moran, R. G.; Kohler,
W.; Freisheim, J . H. Nδ-Acyl-NR-(4-amino-4-deoxypteroyl)-L-
ornithine derivatives: synthesis and in vitro antitumor activity.
J . Med. Chem. 1988, 26, 1332-1337.
(4) Rosowsky, A.; Bader, H.; Forsch, R. A. Synthesis of the folylpoly-
glutamate synthetase inhibitor NR-pteroyl-L-ornithine and its Nδ-
benzoyl and Nδ-hemiphthaloyl derivatives, and an improved
synthesis of NR-(4-amino-4-deoxypteroyl)-Nδ-hemiphthaloyl-L-
ornithine. Pteridines 1989, 1, 91-98.
(5) Rhee, M. S.; Galivan, J .; Wright, J . E.; Rosowsky, A. Biochemical
studies on PT523, a potent nonpolyglutamatable antifolate, in
cultured cells. Mol. Pharmacol. 1994, 45, 783-791.
(6) Wright, J . E.; Pardo, M.; Tretyakov, A.; Alperin, W. L.; Trites,
D.; Rosowsky, A. Pharmacokinetics, antifolate activity and tissue
distribution of PT523 in SCC VII tumor bearing mice. Cancer
Chemother. Pharmacol. 1998, 42, 300-306; cf. erratum: Cancer
Chemother. Pharmacol. 1999, 43, 522.
(7) Rosowsky, A.; Wright, J . E.; Holden, S. A.; Waxman, D. J .
Influence of lipophilicity and carboxyl group content on the rate
of hydroxylation of methotrexate derivatives by aldehyde oxi-
dase. Biochem. Pharmacol. 1990, 40, 851-857.
(8) Rosowsky, A.; Freisheim, J . H.; Moran, R. G.; Solan, V. C.; Bader,
H.; Wright, J . E.; Radike-Smith, M. Methotrexate analogues. 26.
Inhibition of dihydrofolate reductase and folylpolyglutamate
synthetase activity and in vitro tumor cell growth by metho-
trexate and aminopterin analogues containing a basic amino acid
side-chain. J . Med. Chem. 1986, 29, 655-660.
(9) McGuire, J . J .; Bolanowska, W. E.; Piper, J . A. Structural
specificity of inhibition of human folylpolyglutamate synthetase
by ornithine-containing folate analogues. Biochem. Pharmacol.
1988, 37, 3931-3939.
(10) J ohnson, J . M.; Meiering, E. M.; Wright, J . E.; Pardo, J .;
Rosowsky, A.; Wagner, G. NMR solution structure of the
antitumor compound PT523 and NADPH in the ternary complex
with human dihydrofolate reductase. Biochemistry 1997, 36,
4399-4411.
(11) Cody, V.; Galitsky, N.; Luft, J . R.; Pangborn, W.; Rosowsky, A.;
Blakley, R. L. Comparison of two independent crystal structures
of human dihydrofolate reductase ternary complexes with
NADPH and the very tight-binding inhibitor PT523. Biochem-
istry 1997, 36, 13897-13903.
(12) Westerhof, G. R.; Schornagel, J . H.; Kathmann, I.; J ackman, A.
L.; Rosowsky, A.; Forsch, R. A.; Hynes, J . B.; Boyle, F. T.; Peters,
G. J .; Pinedo, H. M.; J ansen, G. Carrier- and receptor-mediated
transport of folate antagonists targeting folate-dependent en-
zymes: correlates of molecular-structure and biological activity.
Mol. Pharmacol. 1995, 48, 459-471.
(13) Chen, G.; Wright, J . E.; Rosowsky, A. Dihydrofolate reductase
binding and cellular uptake of nonpolyglutamatable anti-
folates: correlates of cytotoxicity toward methotrexate sensi-
tiveand resistant human head and neck squamous carcinoma
cells. Mol. Pharmacol. 1995, 45, 758-765.
(14) Wright, J . E.; Vaidya, C. M.; Chen, Y.-N.; Rosowsky, A. Efficient
utilization of the reduced folate carrier in CCRF-CEM human
leukemic lymphoblasts by the potent antifolate NR-(4-amino-4-
deoxypteroyl)-Nδ-hemiphthaloyl-L-ornithine (PT523) and its B-
ring analogues. Biochem. Pharmacol., in press.
(15) Rosowsky, A.; Bader, H.; Wright, J . E.; Keyomarsi, K.; Matherly,
L. S. Synthesis and biological activity of Nω-hemiphthaloyl-R,ω-
diaminoalkanoic acid analogues of aminopterin and 3′,5′-dichlo-
roaminopterin. J . Med. Chem. 1994, 37, 2167-2174.
(16) Rosowsky, A.; Vaidya, C. M.; Bader, H.; Wright, J . E.; Teicher,
B. A. Analogues of NR-(4-amino-4-deoxypteroyl)-Nδ-hemiphtha-
loyl-L-ornithine (PT523) modified in the side chain: synthesis
and biological evaluation. J . Med. Chem. 1997, 40, 286-299.
(17) Rosowsky, A.; Wright, J . E.; Vaidya, C. M.; Bader, H.; Forsch,
R. A.; Mota, C. E.; Pardo, J .; Chen, C. S.; Chen, Y.-N. Synthesis
and potent antifolate activity and cytotoxicty of B-ring deaza
analogues of the nonpolyglutamatable dihydrofolate reductase
inhibitor NR-(4-amino-4-deoxypteroyl)-Nδ-hemiphthaloyl-L-orni-
thine (PT523). J . Med. Chem. 1998, 41, 5310-5319.
(24) DeGraw, J . I.; Colwell, W. T.; Piper, J . R.; Sirotnak, F. M.
Synthesis and antitumor activity of 10-propargyl-10-deazaami-
nopterin. J . Med. Chem. 1993, 36, 2228-2231.
(25) For an excellent review of the various known routes to 10-
deazaaminopterin and its 10-substituted derivatives, see: De-
Graw, J . I.; Colwell, W. T.; Piper, J . R.; Sirotnak, F. M.; Smith,
L. R. New analogues of methotrexate in cancer and arthritis.
Curr. Med. Chem. 1995, 2, 630-653.
(26) Ellard, J . A. Synthesis of methotrexate. U.S. Patent 4,080,325,
March 21, 1978; Chem. Abstr. 1978, 88, 38149k.
(27) Piper, J . R.; J ohnson, C. A.; Maddry, J . A.; Malik, N. D.; McGuire,
J . J .; Otter, G. M.; Sirotnak, F. M. Studies on analogues of
classical antifolates bearing the naphthoyl group in place of
benzoyl in the side chain. J . Med. Chem. 1993, 36, 4161-4171.
(28) Piper, J . R.; Montgomery, J . A. A convenient synthesis of
aminopterin and homologues via 6-(bromomethyl)-2,4-diami-
nopteridine hydrobromide. J . Med. Chem. 1974, 11, 379-380.
(29) Boyle, P. H.; Pfleiderer, W. Synthesis and reactivity of 2,4-
diamino-6-(hydroxymethyl)-pteridine. Chem. Ber. 1980, 113,
1514-1523.
(30) Rosowsky, A.; Freisheim, J . H.; Bader, H.; Forsch, R. A.; Susten,
S. S.; Cucchi, C. A.; Frei, E., III. Methotrexate analogues. 25.
Chemical and biological studies on the tert-butyl esters of
methotrexate and aminopterin. J . Med. Chem. 1985, 28, 660-
667.
(31) Taylor, E. C.; Kobayashi, T. Pteridines. XXXII. 2-Amino-3-cyano-
5-chloromethylpyrazine-1-oxide and its conversion to 6-alkenyl-
substituted pteridines. J . Org. Chem. 1973, 38, 2817-2820.
(32) The CH2 group in 30 is acidic enough to form an ylide even with
Et3N (see ref 31).
(33) Dawson, M. I.; O’Krongly, D.; Hobbs, P. D.; Barrueco, J . R.;
Sirotnak, F. M. Synthesis of the 7-hydroxy metabolites of
methotrexate and 10-ethyl-10-deazaaminopterin. J . Pharm. Sci.
1987, 76, 635-638.
(34) Bergmann, E. D.; Blum, J . p-Vinylbenzoic and p-vinylpheny-
lacetic acids. J . Org. Chem. 1959, 24, 549-551.
(35) Sergeev, P. G.; Sladkov, A. M. Autoxidation of substituted alkyl
homologues of benzene. III. Preparation of p-acetylbenzoic acid.
J . Gen. Chem. U.S.S.R. 1957, 27, 893-894.
(36) Cramer, S. M.; Schornagel, J . H.; Kalghatgi, K. K.; Bertino, J .
R.; Harvath, C. Occurrence and significance of D-methotrexateas
a contaminant of commercial methotrexate. Cancer Res. 1984,
44, 1843-1846.
(37) Forsch, R. A.; Wright, J . E.; Rosowsky, A. Synthesis of γ-sulfinic
and γ-nitro analogues of 5-deazatetrahydrofolic acid. Heterocycles
1999, 51, 1789-1805.
(38) Samuels, L. L.; Moccio, D. M.; Sirotnak, F. M. Similar differential
for total polyglutamylation and cytotoxicity among various folate
analogues in human and murine tumor cells in vitro. Cancer
Res. 1985, 45, 1488-1495.
(39) Rosowsky, A.; Bader, H.; Kohler, W.; Freisheim, J . H.; Moran,
R. G. Methotrexate analogues. 34. Replacement of the glutamate
moiety in methotrexate and aminopterin by long- chain 2-ami-
noalkanedioic acids. J . Med. Chem. 1988, 31, 1338-1344.
(40) Bowden, K.; Shaw, M. J . Transmission of polar effects. Part XIII.
The polar effects of para-acyl and -aroyl groups. J . Chem. Soc.
1971, 161-162.
(41) Leuck, G. J .; Perkins, R. P.; Whitmore, F. C. The mercuration
of naphthalic acids. J . Am. Chem. Soc. 1929, 34, 1831-1836.
J M990630F