Regioselective synthesis of peptidic derivatives and glycolamidic esters of Methotrexate

Article abstract of DOI:10.1016/j.tet.2004.11.049

Convenient methods for regioselective syntheses of Methotrexate peptidic conjugates are described. Solid phase synthesis for derivatives of Methotrexate containing an amide bond has been applied and showed higher efficiency than liquid phase synthesis. Synthetic pathways for regioselective preparation of Glycolamidic ester derivatives of Methotrexate were also developed using 4-amino-4-deoxy-N¹⁰-methylpteroic acid as starting material. These ester bonds were obtained either in solution or by using a combined liquid and solid phase synthesis.

Full text of DOI:10.1016/j.tet.2004.11.049

Tetrahedron 61 (2005) 803–812
Regioselective synthesis of peptidic derivatives and glycolamidic
esters of Methotrexate
a,
Cedric Castex, Christophe Lalanne, Patrick Mouchet, Marc Lemaire and Roger Lahana
a
a
b
a
*
´
a
ˆ
Syntem, Parc Scientifique Georges Besse, 30035 Nımes, France
` ´ ˆ
Laboratoire de Catalyse et Synthese Organique, Universite Claude Bernard Lyon 1, UMR 5181, CPE Bat. 308, 43 boulevard du 11,
b
Novembre 1918, 69622 Villeurbanne Cedex, France
Received 28 July 2004; revised 17 November 2004; accepted 18 November 2004
Available online 8 December 2004
Abstract—Convenient methods for regioselective syntheses of Methotrexate peptidic conjugates are described. Solid phase synthesis for
derivatives of Methotrexate containing an amide bond has been applied and showed higher efficiency than liquid phase synthesis. Synthetic
pathways for regioselective preparation of Glycolamidic ester derivatives of Methotrexate were also developed using 4-amino-4-deoxy-N10-
methylpteroic acid as starting material. These ester bonds were obtained either in solution or by using a combined liquid and solid phase
synthesis.
q 2004 Elsevier Ltd. All rights reserved.
1. Introduction
The efficacy of MTX is hampered by its very short plasma
half-life. It is administrated in relatively high doses, which
often leads to drug resistance and causes non-specific
toxicities in normal cells.1 Resistance to MTX is the result
of different biological phenomena in the target cell.
Deficiency in the reduced folate transporter, in folyl
polyglutamate synthetase (FPGS) or over-expression of
gamma-glutamyl hydrolase (GGS) and efflux proteins at the
surface of the cell (MRP proteins), can prevent MTX to
enter into and/or stay inside the cell. In the same time, over-
expression and/or mutation of DHFR can prevent MTX to
exert a sufficient cytotoxic activity.4–6 Another important
side effect of MTX which has been clearly identified, is its
huge chronic toxicity for the central nervous system when
the drug is used concomitantly with radiations.7
Methotrexate (MTX) (Scheme 1), a folic acid antagonist,
has been extensively used in cancer chemotherapy since
1948.1 This drug acts as an inhibitor of dihydrofolate
reductase (DHFR), an essential enzyme in the biosynthesis
of thymidylate, which is required for DNA replication.2
MTX is also subsequently used in the treatment of non-
neoplasmic diseases as an anti-inflammatory and/or an
immunosuppressive drug.3 The mechanism of action of
MTX is well documented. It is brought into the cell by the
reduced folate transporter, and it is then polyglutamylated in
order to prevent efflux from the cell and to exert its cytotoxic
activity by blocking the synthesis of the N5, N10-
methylenetetrahydrofolate.3
Scheme 1. Structures of MTX and APA.
Keywords: Methotrexate; Peptidic conjugate; Regioselective synthesis; Solid phase synthesis; Vectorisation.
* Corresponding author. Tel.: C33 466 048 666; fax: C33 466 048 667; e-mail: ccastex@syntem.com
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.11.049

804
C. Castex et al. / Tetrahedron 61 (2005) 803–812
To overcome these problems, many experimental efforts
have been made. Among them, modifications of MTX are
the most employed. g-tert-Butyl ester, g-hydrazide, g-n-
butylamide and g-benzylamide of MTX were synthesized in
order to modify the biological behaviour of the drug.8 Other
compounds with amino acids and various esters on the two
carboxylic acid functions were also prepared with different
strategies, the most current one involving the use of
4-amino-4-deoxy-N10-methylpteroic acid (APA) (Scheme1)
as a building block.1,9–11 These transformations were
performed in order to modify the lipophilicity of MTX
and/or to study the cytotoxicity of the prodrug activated by
various enzymes. In all cases, either the regioselective
synthesis described involve numerous steps, or the final
products are obtained after purification of the formed
isomers.
Pignatello and co-workers have described the synthesis of
several a- and g-monosubstituted and a,g-disubstituted
lipoamino acid conjugates of MTX coupled with a
glycolamidic ester.20 Once again, the monosubstituted
products are obtained as secondary products of the
disubstituted ones, and the yields do not exceed 15% for
these compounds.
Previous work in our laboratory led to the discovery of
peptides derived from the Protegrin PG-1 peptide, which
belongs to the family of b-stranded antimicrobial pep-
tides.21 These peptides have been shown to translocate
efficiently through biological membranes, thus providing
the basis for the development of new peptide-conjugated
drugs that cross-cell membranes. In a precedent work,22 the
authors have shown that vectorized doxorubicin bypasses
the P-glycoprotein pump at the luminal site of the blood-
brain barrier. Moreover, using these peptidic vectors with
doxorubicin is efficient in overcoming multidrug resistance.
In the present paper, we report the synthesis of MTX
peptidic conjugates using solid phase synthesis for the
conjugate with amide bonds between the vector and the
drug. A regioselective liquid phase synthesis will also be
presented for conjugates containing a glycolamidic ester
bond (Scheme 2).
Instead of modifying the core structure of MTX, vectorisa-
tion is an alternative strategy to increase the efficacy of
MTX. MTX has been attached to various vectors like
monoclonal antibodies,12,13 proteins,14 polymers15,16 and
hormonal peptides.17–19 All the synthesis are performed in
liquid phase with or without protecting groups, leading,
respectively, to a single product with a relatively low
yield, or to a mixture of regioisomers. More recently,
Scheme 2. Target molecules and general strategies used.

C. Castex et al. / Tetrahedron 61 (2005) 803–812
805
2. Results and discussion
affecting the biological activity of the drug. In order to
verify these hypotheses, a-MTX-SynB3 3 and g-MTX-
SynB3 4 have been synthesized with yields of 50 and 66%,
respectively. After classical peptidic synthesis using Fmoc/
tBu strategy, APA was introduced using PyBOP as an
activator. The two position isomers were obtained by using
commercially available, correctly protected glutamic acid
residues. No secondary product other than peptidic
elongation ones was observed following cleavage and
deprotection of the conjugate, thus showing that MTX
remains stable upon drastic acidic treatment.
2.1. Solid phase synthesis of MTX conjugates
In order to selectively synthesize a- or g-conjugates of
MTX and peptide carriers, a method using solid phase
synthesis has been used. Previous work had already been
described for the synthesis of MTX-peptides conjugates, but
the strategy used by the authors involved liquid phase
synthesis of a protected MTX.17 Use of solid phase
synthesis for MTX conjugates has already been described,
but in the case of MTX lipoamino acids conjugates.20
Using this solid phase synthesis strategy, conjugates with a
peptidic spacer between MTX and the vector were also
prepared (Scheme 4). A tri-glutamic acid spacer was
introduced in order to enhance the biological activity of
the drug. Indeed, the mechanism of action of MTX includes
a step of g-polyglutamylation by folylpolyglutamyl
synthase.3 This step could allow MTX not to be effluxed
across the cell membrane by multidrug related proteins
(MRP), hence increasing its intracellular half-life. Further-
more, this polyglutamylated form of MTX has been proven
In the present work, two conjugates without any spacer
(Scheme 3) have been synthesized. They differ in the MTX
position of the vector coupling. According to the literature,
the free a-carboxylic acid function of MTX seems to be
important for the stability of the complex between the drug
and DHFR.24 In contrast, the g-carboxylic acid function of
MTX lies outside of the binding pocket of DHFR. As this
function does not seem to be involved in the stability of the
complex, it may be possible to couple the vector without
Scheme 3. Solid phase synthesis of compounds a-MTX-SynB3 3 and g-MTX-SynB3 4.

806
C. Castex et al. / Tetrahedron 61 (2005) 803–812
Scheme 4. Solid phase synthesis of compounds g-MTX-(Glu)3-SynB3 5 and g-MTX-(b-Ala)2-SynB3 6.
to inhibit DHFR with the same efficacy than MTX, but has
also an increased affinity for certain folate-dependent
enzymes such as thymidylate synthase.
because of competitive reactions of the free carboxylic acid
functions.
The second derivative 6 synthesized using this strategy
included a di-b-alanyl peptidic spacer. This product was
designed for two reasons. Firstly, the presence of the spacer
allows MTX and the peptide vector not to be too closed one
to each over, which could hamper the formation of the
complex between MTX and DHFR. Secondly, it has been
shown in the literature that a-MTX-peptides are cleaved by
many carboxypeptidases and that the rate of cleavage
depends on the nature of the amino acid bearing MTX.10,25
The conjugate 5 was obtained with a yield of 46%. In order
to compare solid and liquid phase syntheses, an attempt was
made to synthesize this compound with a final liquid
phase step for the introduction of APA. The yield of the
reaction was only 33%, showing the higher efficiency of
solid phase synthesis. Moreover, the conversion of the
reaction was only about 50%, hence decreasing the yield of
the synthesis. A better conversion could not be reached

C. Castex et al. / Tetrahedron 61 (2005) 803–812
807
It was also shown that only a-MTX-Arg is cleaved in human
serum by carboxypeptidase A.10 Arg being the N-terminal
amino acid of the peptidic vector used in the present study
and although the conjugates synthesized were essentially
g-isomers, this new conjugate was then synthesized to
compare the biological activity of our products depending
on the nature of the amino acid bearing MTX. This
conjugate was obtained with a yield of 66%.
2.2. Glycolamidic ester conjugates of MTX
The difference in the biological and hydrolytic stability
between amide and ester bonds prompted us to synthesize
conjugates with an ester bond between the vector and MTX.
Previous work in the laboratory with benzylpenicillin had
already shown the usefulness of a glycolamidic ester linker
for peptidic conjugation.26 Furthermore, to our knowledge,
g-glycolamidic esters of MTX have only been obtained
with low yields after separation from the diesters and the
a-isomer.20 Consequently, we elaborated a regioselective
synthesis of g-glycolamidic esters of MTX.
Two different synthetic routes have been investigated. The
first one had the advantage to introduce APA during the last
steps of the synthesis. There was, however, an important
risk of pyroglutamate formation during deprotection of the
amine function of the glutamic acid before coupling with
APA. Indeed, g-ester of glutamic acid is well known to
undergo rapid lactamization under certain conditions.27
Simplification of the synthesis scheme being, however,
attractive, the synthesis was undertaken since it has been
reported that the glycolamidic ester link was stable toward
hydrolysis during prolonged time.20 Cesium salt of Fmoc-
Glu-OtBu was formed in solution in methanol and was
reacted with 2,3,4-trichlorophenol bromoacetate28 in
solution in DMF. Without further purification, the peptide
was added in the presence of DIEA to yield the desired
g-glycolamidic ester (yield: 31%, 3 steps). Deprotection of
the amine function was then attempted. Unfortunately,
deprotection using excess of piperidine in dry DMF only
yielded the pyroglutamic compound and the hydroxy-acetic
peptide. Decreasing both the amount of base and the
temperature did not product any effect since the cleavage of
the Fmoc protective group was immediately followed by the
reaction of lactamization.
Scheme 5. Glycolamidic ester conjugate of MTX: final synthetic route.
identify the products, we synthesized the glycolamidic
conjugate with D-H-Glu(OH)-OtBu as starting material.
Comparison of the final products with those obtained using
L-H-Glu(OH)-OtBu proved that the two isomers were not
two diastereoisomers as initially supposed. This suggests
that they could actually be regioisomers. Cheung and co-
workers have described the synthesis of N-(L-a-aminoacyl)
derivatives of MTX.29 The compounds were obtained by
reaction of the di-tert-butyl ester of MTX with N-Boc-
amino acid to yield mono and di substituted derivatives of
MTX. Deprotection of the Boc group in acidic medium
yielded only the 2-acyl products, the 4-acyl ones leading to
decomposition products. Based on these observations, the
two compounds obtained in our synthesis were reacted with
an excess of benzylamine in the presence of PyBOP and
DIEA (Scheme 6). The main product of the synthesis
reacted with only 1 equiv of amine whereas the secondary
product reacted with 2 equiv of benzylamine (observed in
MALDI-TOF spectrometry). These findings suggest that the
secondary product could possess two free carboxylic acid
functions. The hypothesis of regioisomerism seems there-
fore to be confirmed. The secondary product was certainly
the result of a nucleophilic attack of the bromoacetyl peptide
by the two nitrogen of MTX in place of the carboxylate
alkylated as cesium salt (Scheme 6).
Because of the difficulties encountered, we decided to use
another synthetic scheme (Scheme 5). H-Glu(OH)-OtBu
was reacted with APA and HATU as an activator in DMSO
to yield a-tertiobutyl-MTX 7 (yield 73%). Cesium salt of 7
formed in situ was then reacted with bromoacetyl peptide 8
obtained by reaction of bromoacetic acid with peptidyl
resin, followed by deprotection and cleavage of the peptide
from the resin. Compound 9 was obtained with a 51% yield
as a single product as confirmed by HPLC. Cleavage of
tertiobutyl protecting group was performed in a TFA/TIPS
solution (95/5; v/v) and yielded the final product as a double
peak in HPLC. After isolation of the two products using
preparative HPLC, MALDI-MS analysis proved that the
products were isomer with the same mass.
Moreover, the relatively low yield of the final deprotection
step (42%) could be explained by the degradation as
already observed by Cheung and co-workers.29 Separation
of the products in analytical HPLC only occurred after
We first thought that a racemisation of the chiral glutamic
acid carbon had occurred during the esterification step. To

808
C. Castex et al. / Tetrahedron 61 (2005) 803–812
Scheme 6. Hypothesis for formation and evolution of regioisomers formed during esterification reaction.
cleavage of the tertio-butyl group. Decomposition of an
alkyl derivative of MTX in acidic medium could explain
the difficulties encountered during this final step, since it
has been observed that 4-acyl derivatives of MTX
decomposed under these conditions, whereas 2-acyl ones
were stable.
was made using a solution containing 1% TFA in DCM. The
yield of the two steps was 78%. Once again, comparison
between liquid and solid phase synthesis proved that solid
phase synthesis was more efficient for the introduction of
APA. The same reaction was made using a peptide already
cleaved from the resin and the yield for the APA coupling
step was 38%. This low yield obtained was not due to the
presence of secondary products in the crude conjugate but to
an important quantity of APA which did not react under the
conditions used. The rest of the synthesis was the same as
described for the previous glycolamidic ester conjugate.
Formation of the ester bond was made using cesium
hydrogencarbonate in DMF and compound 14 was obtained
with a yield of 30%. Deprotection of the tertiobutyl
protecting groups was then achieved to give compound 15
(yield: 47%). As expected, two final products with the same
mass were identified, suggesting that regioisomers were
once again present. It is also interesting to notice that the
yields of the reactions with the tri-glutamic acids were
below those with other derivatives in the same reaction
conditions. Several hypotheses could be made as for
example the steric hindrance of the numerous tertiobutyl
group.
Another compound containing a glycolamidic ester
bond has finally been synthesized (Scheme 7,
compound 15).
This compound has been designed to be efficient against
MTX resistant cell lines by taking advantage of the lability
of the ester bond on the one hand, and on the other hand by
delivering to the cell an active form of MTX. Indeed, the
presence of the tri-glutamyl residue could bypass the
resistant form of the cell where FPGS is deficient or GGS
is overexpressed. The synthesis implies the use of a resin
cleavable in mild acidic conditions in order to keep the
protection of the carboxylic function on the glutamic acid
residues. Introduction of the four correctly protected
glutamic acids was made with a yield of 38%. APA was
then coupled to the peptidyl resin and cleavage of the resin

C. Castex et al. / Tetrahedron 61 (2005) 803–812
809
Scheme 7. Synthetic route for glycolamidic ester conjugate of tri-glutamylated MTX.
3. Conclusion
demonstrates that yields obtained with solid phase
synthesis are higher than those obtained in solution. To
our knowledge, no other solid phase synthesis for MTX
peptide conjugates was previously reported. The selective
method used for MTX lipoamino conjugates20 has been
successfully applied to peptidic derivatives. A combined
The present study details a convenient method using solid
phase synthesis for the regioselective preparation of a- and
g-MTX peptide conjugates with APA as starting material.
Comparison with liquid phase synthesis clearly

810
C. Castex et al. / Tetrahedron 61 (2005) 803–812
solid and liquid phase synthesis has been developed to obtain
glycolamidic conjugates of MTX in a regioselective manner.
They were synthesized with good yields unlike previously
reported cases.20 These conjugates were designed in order to
be efficient against MTX resistant cell lines and to increase the
delivery of MTX through the blood brain barrier (BBB). We
have used a member of the SynB peptide family which have
been proved to transport effectively several drug-like
molecules through the BBB.26,30 Toward this goal, several
studies are on going to explore the cytotoxic potency of
these conjugates and the results will be published later.
HPLC and MALDI-MS. The peptide sequence was SynB3
(H-RRLSYSRRRF-NH2, Mw 1395 g molK1).
4.2.2. General method for the coupling of APA with
peptidyl resin. The last residue (APA) for solid phase
synthesis of the conjugates with MTX was introduced using
a manual apparatus. Briefly, starting from appropriate
peptidyl resin, APA (4 equiv) was introduced using
PyBOP (4 equiv) and DIEA (4 equiv) in a DMSO solution.
After 4 h of reaction, the resin was washed with DMSO,
DMF, DCM and dried under reduced pressure for 1 h.
Cleavage and deprotection of the conjugate were performed
during 3 h using a solution of TFA/H2O/triisopropylsilane
(TIPS)/phenol (88/5/4/5; v/v/v/m). 10 mL of deprotection
solution were used for 60 mmol of peptide (theoretical).
After precipitation of the compound with diethyl ether, the
crude conjugate was purified using a C18 reversed-phase
preparative HPLC (from 5 to 60% of buffer B in 60 min).
The fractions containing the desired product were pooled
and lyophilized to give the conjugate as a yellow powder.
Yields and analytical characteristics of each compound are
given below.
4. Experimental
4.1. Materials
N,N-Diisopropylethylamine (DIEA), bromoacetic acid,
2,3,4-trichlorophenol, anhydrous methanol and piperidine
were bought from Fluka. Dichloromethane (DCM),
dimethylformamide (DMF), acetonitrile, diethyl ether,
N,N-diisopropylcarbodiimide (DIPCDI), 4-amino-4-deoxy-
N10-methylpteroic acid and acetic acid were purchased from
Sigma-Aldrich. 4-Dimethylaminopyridine (DMAP) and
benzotriazole-1-yl-oxy-tris-pyrrolidino-phosphonium
hexafluorophosphate (PyBOP) were obtained from Nova-
biochem. Cesium hydrogencarbonate was purchased from
Acros Organics. The appropriate protected amino acids for
solid phase synthesis and 1-hydroxybenzotriazole (HOBT)
were bought from Senn Chemicals, except for Fmoc-
Glu(OH)-OtBu which was from Novabiochem. The resins
used for solid phase synthesis were a Rink Amide Novagel
HL substituted at 0.86 mmol/g and a HMPB-MBHA resin
substituted at 0.82 mmol/g from Novabiochem. Prior to use,
4.2.3. N0[4-[N-[(2,4-Diamino-6-pteridinyl)-methyl]-N-
methylamino]benzoyl]-L-glutamic acid a-(Arg-Arg-
Leu-Ser-Tyr-Ser-Arg-Arg-Arg-Phe-NH2)-amide [3]
(a-MTX-SynB3). Yield: 50%; HPLC: trZ6.24 min (using
a separative gradient from 5 to 95% of buffer B in 15 min);
MALDI-MS: m/z [DHB] 1832.1 (MCHC), 1854.6 (MC
NaC), 1870.8(MCKC). Calcdexact mass: 1831.45 g molK1
.
4.2.4. N0[4-[N-[(2,4-Diamino-6-pteridinyl)-methyl]-N-
methylamino]benzoyl]-L-glutamic acid g-(Arg-Arg-
Leu-Ser-Tyr-Ser-Arg-Arg-Arg-Phe-NH2)-amide [4]
(g-MTX-SynB3). Yield: 66%; HPLC: trZ6.26 min
(using a separative gradient from 5 to 95% of buffer B in
15 min); MALDI-MS: m/z [DHB] 1832.6 (MCHC), 1855.9
(MCNaC), 1871.2 (MCKC). Calcd exact mass:
˚
DMF was dried 24 h on molecular sieves 3 A (extruders
1.6 mm; Prolabo).
Analytical HPLC was performed on a Beckman Gold 32
Karat System, with a Waters Symmetry Shield C18 column
(250!4.6 mm; 5 m) using the following solvent system AZ
H2O/0.1% TFA and BZacetonitrile/0.1% TFA. Waters
preparative system equipped with Prep LC Novapack C18
1831.45 g molK1
.
4.2.5. N0[4-[N-[(2,4-Diamino-6-pteridinyl)-methyl]-N-
methylamino]benzoyl]-L-glutamic acid g-(Glu-Glu-Glu-
Arg-Arg-Leu-Ser-Tyr-Ser-Arg-Arg-Arg-Phe-NH2)-
amide [5] (g-MTX-[Glu]3-SynB3). Yield: 47%; HPLC:
trZ6.33 min (using a separative gradient from 5 to 95% of
buffer B in 15 min); MALDI-MS: m/z [DHB] 2218.1 (MC
˚
cartridges (40!10 mm; 6 m; 60 A) was used for preparative
HPLC with the same solvent system. MALDI-TOF analysis
was performed on a Voyager Elite XL System spectrometer
(Perseptive Biosystems) and LC/MS/MS mass analysis was
achieved on a quadripolar MicroMass Quattro Ultima
spectrometer.
HC). Calcd exact mass: 2217.45 g molK1
.
4.2.6. N0[4-[N-[(2,4-Diamino-6-pteridinyl)-methyl]-N-
methylamino]benzoyl]-L-glutamic acid g-(b-Ala-b-Ala-
Arg-Arg-Leu-Ser-Tyr-Ser-Arg-Arg-Arg-Phe-NH2)-
amide [6] (g-MTX-[b-Ala]2-SynB3). Yield: 65%; HPLC:
trZ6.16 min (using a separative gradient from 5 to 95% of
buffer B in 15 min); MALDI-MS: m/z [DHB] 1975.5 (MC
4.2. Synthesis
Yields calculated for the following reactions take into
account TFA salts. One TFA salt was used for MTX and one
per Arg residue on the peptide.
HC). Calcd exact mass: 1973.75 g molK1
.
4.2.1. Solid phase synthesis. Peptides were assembled by
conventional solid-phase chemistry on a Pioneer automatic
synthesizer (Perseptive Biosystems) using a 9-fluorenyl-
methoxycarbonyl/tertiobutyl (Fmoc/tBu) protection
scheme.23 The crude peptides and conjugates were purified
on C18 reversed-phase preparative HPLC after trifluoro-
acetic acid (TFA) cleavage/deprotection. The purity of
compounds was assessed by C18 reversed-phase analytical
4.2.7. N0[4-[N-[(2,4-Diamino-6-pteridinyl)-methyl]-N-
methylamino]benzoyl]-L-glutamic acid a-tert-butyl-
ester [7] (a-MTX-OtBu). DIEA (42 mL; 0.264 mmol) and
O-(7-Azabenzotriazol-1-yl)-N,N,N0,N0-tetramethyluronium
hexafluorophosphate (HATU, 60 mg; 0.158 mmol) were
added to a solution of APA (50 mg; 0.132 mmol) in 1 mL of
DMSO. After 25 min of stirring, 26.8 mg (0.132 mmol) of

C. Castex et al. / Tetrahedron 61 (2005) 803–812
811
H-Glu(OH)-OtBu solubilised in 300 mL DMSO were added.
The solution was stirred at room temperature for 3 h and
diethyl ether was added. The yellow precipitate was re-
suspended in dry DMF and precipitated again by addition of
diethyl ether. The crude MTX ester was purified on a C18
reverse-phase preparative HPLC (from 5 to 60% of buffer B
in 60 min). The fractions containing the desired product
were pooled and lyophilized to give 60 mg (yield 73%) of a
98% pure yellow powder; HPLC: trZ8.18 min (using a
separative gradient from 5 to 95% of buffer B in 15 min);
MALDI-MS: m/z [DHB] 511.2 (MCHC), 533.2 (MC
5% (v/v). After 30 min of stirring, TFA was evaporated to
dryness under reduced pressure, and the yellow film
obtained was dissolved in 5 mL of a solution containing
H2O (TFA 0.1%; v/v)/acetonitrile (TFA 0.08%; v/v) (90/10;
v/v). The crude product was then applied on a C18 reversed-
phase preparative HPLC (from 5 to 60% of buffer B in
60 min). The fractions containing the conjugate were pooled
and lyophilized to yield 25 mg of a 98% pure yellow powder
(yield 42%); HPLC: trZ6.91 min (using a separative
gradient from 5 to 95% of buffer B in 15 min); MALDI-
MS: m/z [HABA] 1890.5 (MCHC), 1930.7 (MCKC).
NaC), 549.1 (MCKC). Calcd exact mass: 510.56 g molK1
;
Calcd exact mass: 1888.56 g molK1
.
1H NMR (DMSO-d6): dZ12, 14 ppm (s, 1H, –COOH), dZ
8.72 ppm (s, 1H, –CH]N– pteridine), dZ8.20 ppm (d,
1H, JZ7.5 Hz, –NH–), dZ7.72 ppm (d, 2H, JZ8.9 Hz,
CH benzyl), dZ6.80 ppm (d, 2H, JZ8.9 Hz, CH benzyl),
dZ4.87 ppm (s, 2H, –CH2–NMe), dZ4.27 ppm (m, 1H,
–CH–COOtBu), dZ3.45 ppm (s, 4H, –NH2 pteridine), dZ
3.25 ppm (s, 3H, CH3–N–), dZ2.32 ppm (t, 2H, JZ7.5 Hz,
–CH2–COOH), dZ1.98 ppm (m, 2H, –CH2–CH–COOtBu),
dZ1.39 ppm (s, 9H, –C(CH3)3).
4.2.11. H-Glu[Glu(OtBu)-Glu(OtBu)-Glu(OtBu)-HMPB-
MBHA resin]-OtBu [11]. First amino acid was introduced
on HMPB-MBHA resin using a manual apparatus. Briefly,
519 mL (3.435 mmol) of DIPCDI in solution in 2 mL of dry
DCM were added drop wise to 2.9 g (6.87 mmol) of Fmoc-
Glu(OtBu)-OH solubilised in 29 mL of dry DCM. After
30 min of stirring, DCM was evaporated to dryness and the
white residue was resuspended in 30 mL of DMF. DMAP
(0.168 g, 1.374 mmol) and the resin (1.676 g, 1.374 mmol)
were added and the solution was stirred slowly for 3 h. The
resin was filtrated, washed successively with DMF (three
times), DCM (three times) diethyl ether and dried under
vacuum. 2.22 g (yield 95%) of functionalized resin were
obtained. The other amino acids were introduced using
automatic solid phase synthesizer as previously described.
Yield of the synthesis was evaluated at 38% using automatic
UV titration of fluorene nucleus released.
4.2.8. Bromoacetyl-Arg-Arg-Leu-Ser-Tyr-Ser-Arg-Arg-
Arg-Phe-NH2 [8]. The compound was synthesized using
solid phase synthesis. Briefly, starting from SynB3 peptidyl
resin, bromoacetyl was introduced using bromoacetic acid
(4 equiv), PyBOP (4 equiv) and DIEA (4 equiv) in solution
in DMF. After 4 h of reaction, the resin was washed with
DMF, dichloromethane (DCM) and dried under reduced
pressure for 1 h. Cleavage, deprotection and purification of
the modified peptide were performed as previously
described. Yield: 40% (including peptide synthesis);
HPLC: trZ6.17 min (using a separative gradient from 5 to
95% of buffer B in 15 min); MALDI-MS: m/z [DHB]
1518.1 (MCHC), 1540.1 (MCNaC), 1557.9 (MCKC).
4.2.12. N0[4-[N-[(2,4-Diamino-6-pteridinyl)-methyl]-N-
methylamino]benzoyl]-L-glutamic acid a-tert-butyl-
ester, g-(Glu(OtBu)-Glu(OtBu)-Glu(OtBu)-HMPB-
MBHA-resin)-amide [12]. APA (64 mg, 168 mmol) was
introduced on 500 mg (130 mmol) of peptidyl resin 11 as
previously described using HATU (63.8 mg, 168 mmol) and
DIEA (38 mL, 260 mmol) in solution in DMSO. After
stirring slowly during 4 h, the resin was washed with
DMSO, DMF, DCM and dried under reduced pressure for 1 h.
Calcd exact mass: 1516.7 g molK1
.
4.2.9. N0[4-[N-[(2,4-Diamino-6-pteridinyl)-methyl]-N-
methylamino]benzoyl]-L-glutamic acid a-tert-butyl,g-
(glycolamido-Arg-Arg-Leu-Ser-Tyr-Ser-Arg-Arg-Arg-
Phe-NH2)-ester [9] (g-MTX-[GEL-SynB3]-OtBu). 25 mg
(0.049 mmol) of a-MTX-OtBu were dissolved in 125 mL of
dry DMF. Solutions containing, respectively, 47.5 mg
(0.245 mmol) of Cesium hydrogencarbonate in 50 mL of
dry DMF and 123 mg (0.058 mmol) of bromoacetyl-SynB3
in 200 mL of dry DMF were successively added. The
solution was stirred for 48 h at room temperature and after
addition of 3 mL of dry DMF, the crude solution was
applied on a C18 reversed-phase preparative HPLC (from 5
to 60% of buffer B in 60 min). The fractions containing the
desired product were pooled and lyophilized to yield
65.2 mg of conjugate (yield: 51%) of a 85% pure yellow
powder; HPLC: trZ7.11 min (using a separative gradient
from 5 to 95% of buffer B in 15 min); MALDI-MS: m/z
[HABA] 1946.3 (MCHC), 1986.3 (MCKC). Calcd exact
4.2.13. N0[4-[N-[(2,4-Diamino-6-pteridinyl)-methyl]-N-
methylamino]benzoyl]-L-glutamic acid a-tert-butyl-
ester, g-(Glu(OtBu)-Glu(OtBu)-Glu(OtBu))-amide [13].
Peptidyl resin 12 was suspended in 178 mL of DCM
containing 1.8 mL of TFA. Slow stirring was maintained for
30 min and the resin was filtrated. The operation was
repeated twice and the filtrates were pooled, neutralised by
slow addition of ammonium hydroxyde and evaporated to
dryness. The yellow residue was purified on a C18 reversed-
phase preparative HPLC (from 5 to 60% of buffer B in
60 min). The fractions containing the desired product were
pooled and lyophilized to give 120 mg (yield 78% for two
steps) of a 98% pure yellow powder.
mass: 1944.56 g molK1
.
HPLC: trZ10.25 min (using a separative gradient from 5 to
95% of buffer B in 15 min); MALDI-MS: m/z [DHB]
1066.4 (MCHC), 1089.4 (MCNaC), 1104.3 (MCKC).
4.2.10. N0[4-[N-[(2,4-Diamino-6-pteridinyl)-methyl]-N-
methylamino]benzoyl]-L-glutamic acid g-(glycolamido-
Arg-Arg-Leu-Ser-Tyr-Ser-Arg-Arg-Arg-Phe-NH2)-
ester [10] (g-MTX-GEL-SynB3). 65.2 mg (0.0315 mmol)
of g-MTX-[GEL-SynB3]-OtBu were suspended in 7 mL of
a solution composed of TFA 95%/Triisopropylsilane (TIPS)
Calcd exact mass: 1066.23 g molK1
.
4.2.14. N0[4-[N-[(2,4-Diamino-6-pteridinyl)-methyl]-N-
methylamino]benzoyl]-L-glutamic acid a-tert-butyl-
ester,g-(Glu(OtBu)-Glu(OtBu)-Glu(OtBu)-glycolamido-

812
C. Castex et al. / Tetrahedron 61 (2005) 803–812
5. Bleyer, W. A. Cancer 1978, 41, 36–51.
Arg-Arg-Leu-Ser-Tyr-Ser-Arg-Arg-Arg-Phe-NH2)-
amide [14] (g-MTX-[Glu(OtBu)3-GEL-SynB3]-OtBu).
Glycolamidic ester bond was formed as described for
compound 7. 119 mg (0.112 mmol) of compound 13 were
dissolved in 300 mL of dry DMF. Solutions containing,
respectively, 119 mg (0.56 mmol) of Cesium hydrogen-
carbonate in 100 mL of dry DMF and 280 mg (0.134 mmol)
of bromoacetyl-SynB3 in 200 mL of dry DMF were
successively added. The solution was stirred for 48 h at
room temperature and after addition of 3 mL of DMF, the
crude solution was applied on a C18 reversed-phase
preparative HPLC (from 5 to 60% of buffer B in 60 min).
The fractions containing the desired product were pooled
and lyophilized to yield 108 mg of conjugate (yield: 30%) of
a 90% pure yellow powder; HPLC: trZ8.15 min (using a
separative gradient from 5 to 95% of buffer B in 15 min);
MALDI-MS: m/z [HABA] 2502.9 (MCHC). Calcd exact
¨
6. Bertino, J. R.; Goker, E.; Gorlick, R.; Li, W. W.; Banerjee, D.
Stem Cells 1996, 14, 5–9.
7. Shapiro, W. R.; Allen, J. C.; Horten, B. C. Clin. Bull. 1980, 10,
49–52.
8. Rosowsky, A.; Forsch, R. A.; Uren, J.; Wick, M. J. Med.
Chem. 1981, 24, 1450–1455.
9. Rosowsky, A.; Forsch, R. A.; Yu, C. S.; Lazarus, H.;
Beardsley, G. P. J. Med. Chem. 1984, 27, 605–609.
10. Kuefner, U.; Lohrmann, U.; Montejano, Y.; Vitols, K. S.;
Huennekens, F. M. Adv. Enzyme Regul. 1988, 27, 3–13.
11. Piper,J.R.;Montgomery, J.A.J.Med.Chem. 1982,25, 182–187.
12. Kralovec, J.; Spencer, G.; Blair, A. H.; Mammen, M.; Singh,
M.; Ghose, T. J. Med. Chem. 1989, 32, 2426–2431.
¨ ¨
13. Hellstrom, I.; Hellstrom, K. E.; Siegall, C. B.; Trail, P. A. Adv.
Pharmacol. 1995, 33, 349–388.
14. Boratynski, J.; Opolski, A.; Wietrzyk, J.; Gorski, A.;
Radzikowski, C. Cancer Lett. 2000, 148, 189–195.
15. Subr, V.; Strohalm, J.; Hirano, T.; Ito, Y.; Ulbrich, K.
J. Controlled Release 1997, 49, 123–132.
mass: 2501.93 g molK1
.
4.2.15. N0[4-[N-[(2,4-Diamino-6-pteridinyl)-methyl]-N-
methylamino]benzoyl]-L-glutamic acid g-(Glu-Glu-Glu-
glycolamido-Arg-Arg-Leu-Ser-Tyr-Ser-Arg-Arg-Arg-
Phe-NH2)-amide [15] (g-MTX-Glu3-GEL-SynB3).
108 mg (0.035 mmol) of (g-MTX-[Glu(OtBu)3-GEL-
SynB3]-OtBu) 14 were suspended in 9.6 mL of a solution
composed of TFA 95%/Triisopropylsilane (TIPS) 5% (v/v).
After stirring 30 min, TFA was evaporated to dryness under
reduced pressure, and the yellow film obtained was
dissolved in 5 mL of a solution containing H2O (TFA
0.1%; v/v)/acetonitrile (TFA 0.08%; v/v) (90/10; v/v). The
crude product was then purified on a C18 reversed-phase
preparative HPLC (from 5 to 60% of buffer B in 60 min).
The fractions containing the conjugate were pooled and
lyophilized to give 48.5 mg (yield: 47%) of a 98% pure
yellow powder; HPLC: trZ5.36 min (using a separative
gradient from 5 to 95% of buffer B in 15 min); MALDI-MS:
m/z [HABA] 2277.8 (MCHC), 2300.8 (MCNaC), 2316.7
¨
16. Riebeseel, K.; Biedermann, E.; Loser, R.; Breiter, N.;
Hanselmann, R.; Mu¨lhaupt, R.; Unger, C.; Kratz, F.
Bioconjugate Chem. 2002, 13, 773–785.
17. Nagy, A.; Szoke, B.; Schally, A. V. Proc. Natl. Acad. Sci.
U.S.A. 1993, 90, 6373–6376.
¨ ¨
18. Janaky, T.; Juhasz, A.; Rekasi, Z.; Serfoso, P.; Pinski, J.;
Bokser, L.; Srkalovic, G.; Milovanovic, S.; Redding, T. W.;
Halmos, G.; Nagy, A.; Schally, A. V. Proc. Natl. Acad. Sci.
U.S.A. 1992, 89, 10203–10207.
19. Szepeshazi, K.; Schally, A. V.; Juhasz, A.; Nagy, A.; Janaky,
T. Anti-Cancer Drugs 1992, 3, 109–116.
20. Pignatello, R.; Spampinato, G.; Sorrenti, V.; Di Giacomo, C.;
Vicari, L.; McGuire, J. J.; Russell, C. A.; Puglisi, G.; Toth, I.
Eur. J. Pharm. Sci. 2000, 10, 237–245.
21. Drin, G.; Temsamani, J. Biochim. Biophys. Acta 2002, 1559,
160–170.
(MCKC). Calcd exact mass: 2277.47 g molK1
.
22. Mazel, M.; Clair, P.; Rousselle, C.; Vidal, P.; Scherrmann,
J.-M.; Mathieu, D.; Temsamani, J. Anti-Cancer Drugs 2001,
12, 107–116.
23. Atherton, E.; Sheppard, R. C. In Solid Phase Peptide
Synthesis: a Practical Approach; IRL: Oxford, 1989.
24. Bolin, J. T.; Filman, D. J.; Matthews, D. A.; Hamlin, R. C.;
Kraut, J. J. Biol. Chem. 1982, 257, 13650–13662.
25. Kuefner, U.; Lorhmann, U.; Montejano, Y.; Vitols, K. S.;
Huennekens, F. M. Biochemistry 1989, 28, 2288–2297.
26. Rousselle, C.; Clair, P.; Temsamani, J.; Scherrmann, J.-M.
J. Drug Target. 2002, 10, 309–315.
Acknowledgements
The authors wish to thank Dr. P. Clair and Dr. J. Temsamani
for permanent interest and helpful advice.
References and notes
27. Wakamiya, T.; Yoshida, D.; Mizuki, Y.; Hanatani, K.;
Yamaguchi, Y.; Shimamoto, T. Study on Solid-Phase
Synthesis of Compounds, Containing Hydroxyamino Acids or
Saccharides, by Taking Advantage of the Characteristics of
GlutamicAcid. PeptideScience; Aoyagi, H., Ed.; 2001, 391–394.
28. Chavanieu, A.; Ceccato, M.; Chenu, J.; Mendre, C.; Calas, B.
Protein Peptide Lett. 1994, 1, 50–53.
1. Huennekens, F. M. Adv. Enzyme Regul. 1997, 37, 77–92.
2. Hitchings, G. H.; Smith, S. L. Adv. Enzyme Regul. 1980, 18,
349–371.
3. Genestier, L.; Paillot, R.; Quemeneur, L.; Izeradjene, K.;
Revillard, J. P. Immunopharmacology 2000, 47, 247–257.
4. Gorlick, R.; Cole, P.; Banerjee, D.; Longo, G.; Wei Li, W.;
Hochhauser, D.; Bertino, J. R. Mechanisms of Methotrexate.
Resistance in Acute Leukemia. Decreased Transport and
Polyglutamylation. In Drug Resistance in Leukemia and
Lymphoma III; Kaspers, G. J. et al, Ed.; Plenum: New York,
1999; pp 543–550.
29. Cheung, H. T. A.; Boadle, D. K.; Tran, T. Q. Heterocycles
1989, 28, 751–758.
30. Rousselle, C.; Clair, P.; Lefauconnier, J.-M.; Kaczorek, M.;
Scherrmann, J.-M.; Temsamani, J. Mol. Pharmacol. 2000, 57,
679–686.