Unanticipated acyloxymethylation of sumatriptan indole nitrogen atom and its implications in prodrug design

Article abstract of DOI:10.1002/ardp.200700250

Sumatriptan is a potent and selective 5-HT_1B and 5-HT _1D agonist used in the symptomatic treatment of migraine; it shows poor oral bioavailability ascribed, in part, to its low lipophilicity. In an attempt to develop acyloxymethyl prodrugs of sumatriptan suitable for oral administration, we carried out the reaction of sumatriptan with chloromethyl esters. To our surprise, acyloxymethylation occurred preferentially at the indole nitrogen rather than at sulfonamide nitrogen, reflecting a difference either in product stability or in the nucleophilicities of the indole and sulfonamide anions. The hydrolysis of the corresponding N¹- acyloxymethyl derivatives was studied in aqueous buffers and in human plasma, by HPLC. N¹-Acyloxymethyl derivatives of sumatriptan are rapidly hydrolysed to the chemically stable N¹-hydroxymethylsumatriptan at pH 1-13. Slow formation of the parent drug was observed only at high pH values. Hydrolysis of sumatriptan derivatives is slower in human plasma than in phosphate buffer and also generates N¹-hydroxymethylsumatriptan rather than the parent drug. These results indicate that N¹- acyloxymethyl derivatives of sumatriptan cannot be considered as true prodrugs of sumatriptan.

Full text of DOI:10.1002/ardp.200700250

344
Arch. Pharm. Chem. Life Sci. 2008, 341, 344–350
Full Paper
Unanticipated Acyloxymethylation of Sumatriptan Indole
Nitrogen Atom and its Implications in Prodrug Design
Tiago Rodrigues1, Rui Moreira1, Rita C. Guedes1, Jim Iley2, and Francisca Lopes1
1 iMed.UL, CECF, Faculty of Pharmacy, University of Lisbon, Lisboa, Portugal
2 Department of Chemistry, The Open University, Milton Keynes, UK
Sumatriptan is a potent and selective 5-HT1B and 5-HT1D agonist used in the symptomatic treat-
ment of migraine; it shows poor oral bioavailability ascribed, in part, to its low lipophilicity. In
an attempt to develop acyloxymethyl prodrugs of sumatriptan suitable for oral administration,
we carried out the reaction of sumatriptan with chloromethyl esters. To our surprise, acyloxy-
methylation occurred preferentially at the indole nitrogen rather than at sulfonamide nitrogen,
reflecting a difference either in product stability or in the nucleophilicities of the indole and
sulfonamide anions. The hydrolysis of the corresponding N1-acyloxymethyl derivatives was stud-
ied in aqueous buffers and in human plasma, by HPLC. N1-Acyloxymethyl derivatives of suma-
triptan are rapidly hydrolysed to the chemically stable N1-hydroxymethylsumatriptan at pH 1-
13. Slow formation of the parent drug was observed only at high pH values. Hydrolysis of suma-
triptan derivatives is slower in human plasma than in phosphate buffer and also generates N1-
hydroxymethylsumatriptan rather than the parent drug. These results indicate that N1-acyloxy-
methyl derivatives of sumatriptan cannot be considered as true prodrugs of sumatriptan.
Keywords: Acyloxymethylation / Indole / Prodrugs / Sumatriptan /
Received: December 6, 2007; accepted: January 18, 2008
DOI 10.1002/ardp.200700250
triptan appears to cross the blood-brain barrier only
Introduction
slowly and its distribution in the CNS is poor [2]. More-
over, in-vivo studies also suggest substantial inter-individ-
ual variability in plasma concentrations after oral admin-
istration, reflected by the fact that approximately 30% of
patients report inadequate relief or do not respond to
oral sumatriptan and a significant number of patients
are prone to headache recurrence [5]. In order to address
these shortcomings of sumatriptan, alternative drug for-
mulations, such as subcutaneous injection, have been
developed [2, 7]. Subcutaneous administration is the
most effective and fastest-acting route of administration,
but the resulting adverse event profile is higher than
that of oral sumatriptan [2, 7].
Sumatriptan 1 is a potent and selective serotonin (5-HT)
agonist that activates 5-HT1B and 5-HT1D subtype receptors
used in the symptomatic treatment of migraine [1].
Although representing a significant achievement in
migraine treatment over ergot alkaloids [2], sumatriptan
presents several limitations, particularly in its pharmaco-
kinetic profile, which may be ascribed to its low lipophi-
licity (log P = 0.93) and first-pass effect through the action
of monoamine oxidase type A (MAO-A) [1, 3–6]. For exam-
ple, sumatriptan exhibits poor oral bioavailability (14%),
reaches peak plasma drug concentration (Cmax) at 2.5 h
and has a short half-life (t1/2 of 2 h) [2]. In addition, suma-
The prodrug approach is widely used to improve drug
delivery properties of numerous therapeutic agents by
covalently attaching a pro-moiety to the parent drug that
can be released in vivo, either chemically or enzymati-
cally [8, 9]. Sumatriptan contains a secondary sulfon-
amide group and an indole group that are amenable to
derivatisation into a prodrug. Only a few successful pro-
Correspondence: Francisca Lopes, iMed.UL, CECF, Faculty of Phar-
macy, University of Lisbon, Av. Prof. Gama Pinto, 1690-019 Lisboa, Por-
tugal.
E-mail: fclopes@ff.ul.pt
Fax: +351 2179 464-70
Abbreviation: density functional theory (DFT)
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Arch. Pharm. Chem. Life Sci. 2008, 341, 344–350
Acyloxymethylation of Sumatriptan
345
Figure 2. Structure of compound 5.
chloromethyl ester and sodium hydride in DMF. The
structural assignment of derivatives 4 is based on several
1
spectroscopic data. A characteristic feature of the H-
NMR spectra for these compounds is the resonance of the
NCH2O group that appears as a singlet at 6.2–6.4 ppm,
which is similar to that for the previously reported N-pro-
tected indole, N-pivaloyloxymethylindole, i.e. 6.1 ppm
[19], but shifted downfield when compared to those of N-
acyloxymethyl-N-methylsulfonamides
[9, 15, 20]. The carbon chemical shifts for the NCH2O sig-
(ca.
5.5 ppm)
Figure 1. Structures of compounds 1–4.
nal of 4 are also similar to that of N-pivaloyloxymethylin-
drugs for secondary sulfonamide agents have been dole, i.e. 68 ppm [19]. Further confirmation that acyloxy-
reported [9–16]. From these, N-acyloxymethylsulfon- methylation occurred at the indole moiety came from
amides 2 emerged as potential prodrugs for both secon- the signal of the sulfonamide N-methyl group, which
dary sulfonamides and carboxylic acid agents, displaying appears as a doublet at ca. 2.5 ppm. Finally, nuclear Over-
low reactivity in aqueous buffers yet being rapidly acti- hauser enhancement spectroscopy (NOESY) experiments
vated to the parent drug by human esterases [9, 15]. In revealed interactions through space between the NCH2O
contrast, prodrug derivatives for the indole moiety have protons with C-2 and C-7 indole protons.
been sparsely reported [17]. However, acyloxymethyla-
The preferential acyloxymethylation of the sumatrip-
tion of the related imidazole NH group has been used to tan indole over the sulfonamide is somewhat surprising,
prepare N-acyloxymethyl derivatives of cimetidine 3 [18]. taking into account the acidity differences between the
Both, cimetidine prodrug tautomers, 3a and 3b, are rap- two NH groups in non-aqueous solvents (for example, the
idly activated to the parent drug in human plasma [18].
pKa values in DMSO for indole and MeSO2NH2 are 20.96
Considering the potential utility of N-acyloxymethyla- and 17.5, respectively [21]). This difference would imply
tion in prodrug synthesis, it was therefore necessary to the major anion present in solution is the sulfonamide
assess the applicability of this approach to sumatriptan. anion, by at least a ratio of 1000 : 1. That acyloxymethyla-
Herein, we report that N-acyloxymethylation of suma- tion occurs at the indole nitrogen atom must be attribut-
triptan with chloromethyl esters occurs preferentially at able to one (or both) of two factors (i) the greater nucleo-
the indole nitrogen rather than at sulfonamide nitrogen, philicity of the indole anion, and (ii) the greater stability
leading to derivatives 4. A kinetic study designed with of the substituted indole product. In an attempt to gain
the objective of evaluating the influence of the acyl car- insight into the factors that affect the acyloxymethyla-
rier on the chemical reactivity of N-acyloxymethyl deriv- tion of sumatriptan, the energies for compound 4a and
atives 4 in aqueous buffers and in human plasma is also its N-acyloxymethylsulfonamide counterpart 5, (Fig. 2)
presented. The mechanism of hydrolysis of the sumatrip- were calculated using density functional theory (DFT).
tan derivatives 4 and its implication in prodrug design The energy-minimized structures of 4a and 5 are depicted
are discussed. The structures of compounds 1–4 are pre- in Fig. 3. The DFT calculations indicate, that compound
sented in Fig. 1.
4a is ca. 7.5 kJ mol– 1 more stable than its counterpart 5.
Thus, substitution at the indole nitrogen in the anion of
1 appears to be both kinetically and thermodynamically
driven.
Results and discussion
Synthesis
Hydrolysis in aqueous buffers
Compounds 4 were synthesized, albeit in moderate The hydrolysis of N-acyloxymethyl sumatriptan deriv-
yields, by reacting sumatriptan with the appropriate atives 4 in aqueous buffers proceeds with quantitative
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.archpharm.com

346
T. Rodrigues et al.
Arch. Pharm. Chem. Life Sci. 2008, 341, 344–350
Figure 4. Time courses for N-pivaloyloxymethylsumatriptan 4a
(h) and N- hydroxymethylsumatriptan (0) in the hydrolysis of 4a
at 378C in pH 7.2 phosphate buffer.
Figure 3. Geometries of 4a and 5 using B3LYP/6-31+G(d,p).
Scheme 2. Loss of formaldehyde from indole anion via E1cb-
type mechanism.
consistent with loss of formaldehyde from 6 occurring
via an E1cb-type mechanism (Scheme 2) [23]. The stability
displayed by 6 is consistent both with that of the indole
N-hydroxymethyl derivative of the NK1 receptor antago-
nist PD154075 (5% decomposition over 4 d at pH 9) [17]
and with the observation of Buur et al. that the rate of
decomposition of N-hydroxymethyl derivatives of amides
and imides increases sharply with the acidity of the
parent NH-acidic compound, as indicated by a Bronsted
blg value of –0.8 [24]. Using this structure-reactivity rela-
tionship and a pKa of 16 for the indole NH [22], a half-life
of ca. 40 days can be calculated for the decomposition of
N-hydroxymethylsumatriptan 6 in pH 7.4 buffer at 378C,
a value consistent with the lack of decomposition of 6 at
this pH.
Scheme 1. Hydrolysis of N-acyloxymethyl sumatriptan deriv-
atives 4 in aqueous buffers.
formation of N-hydroxymethyl sumatriptan (6,
Scheme 1) in the pH range 1 to 10, as revealed by HPLC
and ESI-MS analysis of reaction mixtures. For example,
Fig. 4 depicts the time course for the hydrolysis of com-
pound 4a at pH 7.2. Above pH 10, slow conversion of 6
into sumatriptan was observed. These results reflect the
low nucleofugality of the resulting indole anion (a pKa of
16 in aqueous solution can be calculated for the dissocia-
tion of the sumatriptan indole NH group [22]) and are
The pseudo-first-order rate constants, kobs, for the
hydrolysis of derivatives 4a, b were determined in HCl
and NaOH solutions as well as aqueous buffers using sev-
eral acetate-, phosphate- and borate-buffer concentra-
tions. While pivaloyloxymethyl derivative 4a was studied
at 258C, its 4-methoxybenzoyloxymethyl counterpart 4b
was studied at 158C and between pH 1 and pH 9.5, due to
its higher reactivity. For both compounds, no significant
dependence of hydrolysis rates on buffer concentration
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.archpharm.com

Arch. Pharm. Chem. Life Sci. 2008, 341, 344–350
Acyloxymethylation of Sumatriptan
347
Table 1. Rate data for the hydrolysis of N-acyloxymethylsuma-
triptan derivatives 4a, at 258C, and 4b, at 158C, with ionic
strength maintained at 0.5 M with addition of NaClO4. The pKa
value for 4a was determined from the pH-rate profile.
4a
4b
k0/s–1
4.33610–5; 1.88610–4 a)
;
5.86610–4
6.89610–3
4.32610–4 b); 7.13610–4 c)
kH+/M–1s–1
1.42610–2; 5.15610–3 d)
;
2.63610–2 e); 6.15610–2 a)
kHO-/M–1s–1
k9HO-/M–1s–1
pKa
20.5
90.4
–
–
7.85610–2
9.26
a)
378C.
458C.
508C.
158C.
b)
c)
d)
e)
308C.
experimentally determined first-order rate constants
suggests that the degradation pathway presented in
Scheme 1 and Eq. (1) adequately describes the degrada-
tion kinetics of compound 4a. The kinetically deter-
mined pKa value is 9.26, which is close to, but lower than
that for ionisation of the dimethylamino group in suma-
triptan (pKa 9.63 [25]), and thus is also consistent with
incorporation of the acyloxymethyl moiety at the indole
nitrogen atom.
The pH-rate profile for derivative 4b (Fig. 5) is U-shaped
with a broad plateau from pH 3 to pH 8, indicating the
occurrence of specific acid and base catalysis as well as a
non-catalysed reaction according to the rate expression
shown in Eq. (2), where k0, kH+ and kHO- (Table 1) have the
usual meaning and refer to the hydrolysis of the proto-
nated form of 4b. The higher reactivity of 4b is such that
we were unable to follow reactions by the HPLC method
at pH values greater than 9, so were unable to detect reac-
tion of the unprotonated form of this compound.
Figure 5. pH-Rate profiles for compounds 4a at 258C (0) and 4b
at 158C (f).
was observed over a 7- to 10-fold buffer concentration
range.
The influence of pH on the overall rates of hydrolysis
of N-acyloxymethylsumatriptan derivatives 4 is pre-
sented in Fig. 5, where the logarithm of the kobs values is
plotted against the pH. The pH-rate profile for 4a is sig-
moidal, reflecting protonation of the dimethylamino
group. The best computer fit (solid line) to the experi-
mental data for 4a in Fig. 5 was achieved using Equation
(1) and is consistent with reaction pathway presented in
Scheme 1.
½Hþꢁ
ꢀ
kobs ¼ ðk0 þ kH ½Hþ þ kHO ½HO ꢁÞ
þ
ꢀ
ka þ ½Hþꢁ
Ka
ꢀ
ꢀ
þðk9HO ½HO ꢁÞ
ð1Þ
ka þ ½Hþꢁ
þ
ꢀ
þ
ꢀ
kobs ¼ k0 þ kH ½H ꢁ þ kHO ½HO ꢁ
ð2Þ
Equation 1. pH-Rate expression of hydrolysis of N-acyloxyme-
thylsumatriptan derivative 4a.
Equation 2. pH-Rate expression of hydrolysis of N-acyloxyme-
thylsumatriptan derivative 4b.
In Equation (1), where [H+]/(Ka + [H+]) and Ka/(Ka + [H+]) are
the fractions of the protonated and non-protonated
forms of compound 4a, respectively, Ka is the apparent
ionisation constant, k0 is the first-order rate-constant for
the non-catalysed hydrolysis of the protonated form, kH+
is the second-order rate-constant for the specific acid cat-
alysed hydrolysis of the protonated form and kHO- and k9HO-
are the second-order rate-constants for the specific base-
catalysed hydrolysis of the protonated and non-proto-
nated forms, respectively. The values derived are listed in
Table 1. The good agreement between the calculated and
Inspection of the data in Table 1 indicates that the pro-
tonated form of compound 4b hydrolyses via the pH-inde-
pendent pathway at a rate ca. 25 times faster than its piv-
aloyloxymethyl counterpart 4a (assuming that k0 for 4b
increases two times when temperature increases 108C).
Such difference in reactivity, together with the unusually
high k0 values displayed by compounds 4 when compared
with analogous esters derived from simple aliphatic alco-
hols (usually in the range of 10– 10 to 10– 7 s– 1 at 258C;
[26]), is consistent with an SN1-type mechanism that
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.archpharm.com

348
T. Rodrigues et al.
Arch. Pharm. Chem. Life Sci. 2008, 341, 344–350
Table 2. Pseudo-first-order rate constants, kobs, and half-lives, t1/
2, for the hydrolysis of N1-acyloxymethyl derivatives of sumatrip-
tan, 4, in isotonic phosphate buffer solution at pH 7.4 and 80%
human plasma at 378C.
Compound
pH 7.4 Buffer
80% Human plasma
104kobs/s– 1 t1/2/min
104kobs/s– 1 t1/2/min
4a
4b
2.10
22.3
58
5.2
1.83
8.05
65
15
would imply a strained six-membered transition state for
proton transfer.
Hydrolysis in human plasma
The susceptibility of the sumatriptan derivatives 4 to
undergo enzyme-catalysed hydrolysis was studied in vitro
at 378C in pH 7.4 isotonic phosphate buffer containing
80% human plasma. Under the experimental conditions
used, the reaction displayed first-order kinetics for at
least four half-lives and led to the quantitative formation
of N-hydroxymethylsumatriptan 6. The observed half-
lives for the hydrolysis in human plasma together with
those of hydrolysis in pH 7.4 phosphate buffer are pre-
sented in Table 2. From these data, it is possible to con-
clude that hydrolysis of compounds 4 is slower in human
plasma than in phosphate buffer. This might be ascribed
to strong binding of compounds 4 to non-catalytic pro-
teins, thus preventing their hydrolysis, an effect which
has been also reported for other acyloxymethyl deriv-
atives of NH-acidic compounds [28] as well as for several
alkyl esters [32].
Scheme 3. Hydrolysis mechanism explained in the text.
involves departure of the carboxylate leaving group in
the rate-limiting step, followed by trapping of the result-
ing iminium ion by water (Scheme 3, path a). According
to this mechanism, which has been reported for a wide
range of N-acyloxymethyl derivatives of weakly NH-acidic
compounds such as amides [27–29] and sulfonamides
[9, 15, 20], the rates of hydrolysis increase with acidity of
the carboxylic acid (the pKas for pivalic acid and 4-
methoxybenzoic acid are 5.04 and 4.51, respectively [22]).
Further support for the unimolecular mechanism of the
pH-independent hydrolysis of compounds 4 comes from
the temperature dependence of k0 for 4a (Table 1), which
yielded an entropy of activation, DS‡, value of –35.9 J K– 1
mol– 1. Despite being negative, this DS‡ value is close to
zero and thus consistent with a unimolecular mecha-
nism [15, 30].
Conclusion
In contrast to the pH-independent region, the differ-
ence in reactivity between 4a and 4b in the specific-acid
In conclusion, classical N-acyloxymethylation of suma-
triptan leads to alkylation of the indole nitrogen (N1)
catalysed region is negligible. The temperature depend- atom. The N1-acyloxymethyl derivatives of sumatriptan
ence of kH+ for 4a (Table 1) yielded an DS‡ value of –
7.73 J K– 1 mol– 1. This value is also consistent with a disso-
ciative mechanism, identical to that of the pH-independ-
are rapidly hydrolysed, both in aqueous buffers and in
human plasma, to the corresponding alcohol (N1-hydrox-
ymethylsumatriptan), which is stable in neutral and
ent pathway, other than an extra step involving protona- slightly alkaline solutions. Thus, N1-acyloxymethyl deriv-
tion of the substrate at the ester or indole moieties prior atives of sumatriptan cannot be considered as true pro-
to iminium ion formation (Scheme 3, path b). Protona-
drugs of sumatriptan. These results indicate that future
tion at the carbonyl oxygen atom (Scheme 3, A) and imi- prodrug development for sumatriptan and other indole-
nium-ion formation have opposite electronic require- containing drugs based on the derivatisation of indole
ments, and thus the overall effect exerted by the acyloxy- NH group, must take in account the activation chemistry
and whether this can lead to stable N1-hydroxymethylsu-
methyl moiety on the rate of hydrolysis is expected to be
very small, an outcome consistent with the small reactiv- matriptan.
ity difference between 4a and 4b for the acid-catalysed
pathway. Alternatively, N-protonation of the indole
(Scheme 3, B) is likely to be a fast process [31], but it
CECF and the Faculty of Pharmacy acknowledge Fundaꢀ¼o
para a CiÞncia e Tecnologia (FCT) for the project REDE/1518/
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.archpharm.com

Arch. Pharm. Chem. Life Sci. 2008, 341, 344–350
Acyloxymethylation of Sumatriptan
349
which was subjected to column chromatography on silica gel
using dichloromethane/methanol (4 : 1) as eluent. Subsequent
recrystallisation from dichloromethane and ethyl acetate
afforded compounds 4. Alternatively, the sodium salt of suma-
triptan was prepared directly from sumatriptan succinate
(1 mol eq.) with sodium hydride (3 mol eq.).
REM/2005. TR acknowledges FCT for the PhD fellowship SFRH/
BD/30689/2006. The authors thank Drs. Rosꢁrio Bronze and Isa-
bel Joglar for the ESI-MS experiments.
The authors have declared no conflict of interest.
N1-Pivaloyloxymethylsumatriptan 4a
Experimental
White powder, yield: 56%, mp. 161–1648C. IR mmax/cm– 1: 3432,
1
1737, 1317, 1132. H-NMR dH (d6-DMSO): 1.08 (9H, s, CMe3), 2.39
General procedures
(6H, s, NCH3), 2.59 (3H, d, J = 4.8, NCH3), 2.74 (2H, t, J = 7.6, CH2Ar),
2.88 (2H, t, J = 7.6, NCH2), 4.38 (2H, s, CH2SO2), 6.15 (2H, s, NCH2O),
6.87 (1H, q, J = 4.8, NHMe), 7.21 (1H, d, J = 8.4, ArH), 7.32 (1H, s,
ArH), 7.53 (1H, d, J = 8.4, ArH), 7.56 (1H, s, ArH). 13C-NMR dC (d6-
DMSO): 22.8 (CH2Ar), 26.9 (CMe3), 29.7 (NCH3), 38.9 (CMe3), 45.0
(Me2N), 57.7 (CH2SO2), 59.4 (NCH2CH2), 68.6 (NCH2O), 110.0 (CH,
Ar), 114.5 (CCH2SO2), 121.3 (CH2CH2C), 121.4 (CH, Ar), 125.1 (CH,
Ar), 126.3 (CH, Ar), 128.9 (C, Ar), 136.5 (C, Ar), 178.2 (C=O). MS m/z
(MW) = 409.5275 (Calcd. 409.5498).
1H- and 13C-NMR, COSY, HMQC and NOESY spectra were recorded
as CDCl3 or d6-DMSO solutions on a Bruker AM 400 WB (Bruker
Bioscience, Billerica, MA, USA); chemical shifts are given in ppm
and coupling constants, J, are quoted in Hz. The LC–MS system
consisted of a Waters 2695 Separation Module, Waters 2996 pho-
todiode array detector (Waters Corporation, Milford, MA, USA),
Atlantis dC18 5 lm (2.16150 mm) column (Waters) and a Micro-
mass Quattro Micro API spectrometer (Micromass, Manchester,
UK). The high-resolution mass spectra (HRMS) were obtained on
a Bruker MicroTOF ESI-TOF MS system (Bruker). FTIR spectra
were recorded on a Nicolet Impact 400 spectrophotometer (Nico-
let, Madison, WI, USA). Melting points were determined using a
Bock Monoscop M. Instrument and are uncorrected. HPLC was
performed using a system comprising a Merck Hitachi LaChrom
L-7100 pump (Merck, Darmstadt, Germany) coupled to a Shi-
madzu SPD-6AV UV-vis detector (Shimadzu, Tokyo, Japan)), a
Rheodyne 10 mL injector (Rheodyne Europe, Alsbach, Germany),
a Merck Hitachi D-2500A integrator, and a Merck LiChrospherm
100 RP-8 5 lm 12564 mm column. Water was distilled and de-
ionized using a Millipore apparatus (Millipore Iberica S.A.U.,
Madrid, Spain). All chemicals used were of reagent grade, with-
out further purification, except those for kinetic and HPLC stud-
ies, which were of analytical or LiChrosolvm (Merck) grade. Suma-
triptan succinate Ph. Eur. was obtained from SMS Pharmaceuti-
cals Limited (Hyderabad, India).
N1-(4-Methoxy)benzoyloxymethylsumatriptan 4b
Yellow amorphous solid, yield 59%, mp. 125–1288C. IR mmax/cm–1:
3420, 1712, 1321, 1170. 1H-NMR dH (d6-DMSO): 2.45 (6H, s, NCH3),
2.58 (3H, d, J = 4.8, NCH3), 2.84 (2H, t, J = 6.5, CH2Ar), 2.91 (2H, t, J =
6.5, NCH2), 3.81 (3H, s, OCH3), 4.40 (2H, s, CH2SO2), 6.39 (2H, s,
NCH2O), 6.85 (1H, q, J = 4.8, NHMe), 7.02 (2H, d, J = 8.8, ArH) 7.23
(1H, d, J = 8.4, ArH), 7.43 (1H, s, ArH), 7.58 (1H, s, ArH), 7.67 (1H, d,
J = 8.4, ArH), 7.87 (2H, d, J = 8.8, ArH). 13C-NMR dC (d6-DMSO): 22.4
(CH2Ar), 29.5 (NCH3), 44.5 (Me2N), 55.2 (OMe), 57.3 (CH2SO2), 58.9
(NCH2CH2), 68.4 (NCH2O), 110.0 (CH, Ar), 113.1 (CCH2SO2), 121.2
(CH2CH2C), 121.4 (CH, Ar), 125.1 (CH, Ar), 126.8 (CH, Ar), 128.7 (C,
Ar), 136.3 (C, Ar), 165.6 (C=O). MS m/z (MW) = 459.5801 (Calcd.
459.5663).
Hydrolysis in aqueous buffers and ESI-MS
experiments
Procedure for obtaining neutral sumatriptan
The kinetic studies were carried out using HPLC, following the
loss of substrate and formation of N1-hydroxymethylsumatrip-
tan at a wavelength of 230 nm, at different temperatures. The
ionic strength of the buffers was maintained at 0.5 M using
NaClO4. In a typical run, a reaction was initiated by adding 5 lL
aliquot of a 10– 3 M stock solution of substrate in acetonitrile to
an Eppendorf containing 495 lL of thermostated buffer, result-
ing in a final concentration of 10– 5 M. At regular intervals, sam-
ples of the reaction mixture were analysed by direct injection,
except for solutions of pH above 11.5 in which they were previ-
ously neutralized with HCl. The mobile phase consisted of aceto-
nitrile / aqueous buffer (sodium hexanesulfonate 10 mM, phos-
phoric acid 2.5 mM, sodium acetate 2.5 mM) in 60 : 40 (v/v) with
a 1.0 mL min– 1 flow. Pseudo-first-order rate constants for the
hydrolytic degradation were determined from the slope of lin-
ear plots of ln (peak area) vs. time. Hydrolysis in 0.01 M pH 7.4
PBS was carried out at 378C as described above.
Sumatriptan succinate was dissolved in water and sodium
hydrogen carbonate was added until pH 9.5. Sumatriptan was
then extracted with ethyl acetate. Crystallisation from absolute
ethanol afforded 1 as a pale yellow powder; Mp. 169–1718C. IR
1
mmax/cm– 1: 3354, 1312, 1122. H-NMR dH (d6-DMSO): 2.24 (6H, s,
NMe2), 2.51 (2H, t, J = 7.5, CH2Ar), 2.55 (3H, d, J = 4.8, NMe), 2.82
(2H, t, J = 7.5, NCH2), 4.31 (2H, s, CH2SO2), 6.79 (1H, q, J = 4.8, NH),
7.06 (1H, d, J = 8.4, ArH), 7.11 (1H, s, ArH), 7.30 (1H, d, J = 8.4, ArH),
7.49 (1H, s, ArH), 10.78 (1H, s, NH (Ar)). 13C-NMR dC (d6-DMSO): 23.0
(CH2Ar), 28.9 (NCH3), 45.1 (Me2N), 56.7 (CH2SO2), 59.9 (NCH2CH2),
111.1 (CH (Ar)), 112.6 (CCH2SO2), 119.4 (CH2CH2C), 120.6 (CH (Ar)),
122.9 (CH (Ar)), 123.6 (CH (Ar)), 127.2 (Cquat (Ar)), 135.9 (Cquat (Ar)).
Procedure for the synthesis of N-
acyloxymethylsumatriptan derivatives
To a solution of sumatriptan (1 mol eq.) in anhydrous DMF
(1 mL/sumatriptan mol) sodium hydride was added (1.1 mol eq.).
When liberation of hydrogen was complete, a solution of the
appropriate chloromethyl ester [33] was added and the mixture
left reacting at room temperature. When the reaction was com-
plete (TLC) the solvent was removed under vacuum. Dichlorome-
thane (50 mL) was added to the residue and the resulting solu-
tion washed with water and then evaporated to afford crude 4
For selected reactions of 4a, aliquots were analysed by LC-ESI-
MS in positive mode (collision energy 20 V) , which revealed the
disappearance of substrate (m/z 410, 14%, MH+; m/z 308, 100%,
MH+-OCOCMe3) and formation of N1-hydroxymethylsumatriptan
(m/z 326, 32%, MH+; m/z 296, 11%, MH+-CH2OH; m/z 281, 100%,
MH+-NHMe2; m/z 187, 37%, MH+-NHMe2-SO2NHMe; m/z 58, 18%,
+
CH2NMe2 ).
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.archpharm.com

350
T. Rodrigues et al.
Arch. Pharm. Chem. Life Sci. 2008, 341, 344–350
[16] J. Iley, F. Lopes, R. Moreira, J. Chem. Soc. [Perkin Trans II]
2001, 5, 749–753.
Hydrolysis in human plasma
Compounds 4 were incubated at an initial concentration of
5610– 5 M at 378C in human plasma [28] diluted to 80% (v/v)
with pH 7.4 isotonic phosphate buffer. At appropriate intervals,
200 lL aliquots were added to 400 lL of acetonitrile in order to
quench the reaction and deproteinize the plasma. The samples
were centrifuged 15000 rpm and the supernatant was analysed
by HPLC for the presence of the substrate and product of hydrol-
ysis.
[17] Z. Zhu, H.-G. Chen, O. P. Goel, O. H. Chan, et al., Bioorg.
Med. Chem. Lett. 2000, 10, 1121–1124.
[18] A. Buurs, H. Bundgaard, Acta Pharm. Nord. 1991, 3, 51–56.
[19] D. Dhanak, C. B. Reese, J. Chem. Soc. [Perkin Trans I] 1986, 1,
2181–2186.
[20] T. Calheiros, J. Iley, F. Lopes, R. Moreira, Bioorg. Med. Chem.
Lett. 1995, 5, 937–940.
[21] F. G. Bordwell, Acc. Chem. Res. 1988, 21, 456–463.
Computational methodology
Density functional theory (DFT) [34, 35] calculations were carried
out with the Gaussian 03 suite of programs [36]. The structures
of 4a and 5 were optimized without symmetry constraints using
the B3LYP hybrid functional, which is a combination of the
Becke's three-parameter (B3) exchange functional [37] with the
Lee, Yang, and Parr (LYP) correlation functional [38], in conjunc-
tion with B3LYP/6-31+G(d,p) [39, 40] basis set. Frequency calcula-
tions confirmed that all the geometries found were minima on
the respective potential energy surfaces. In the evaluation of
data, the zero-point and thermal energy corrections were added
to the minimum electronic energies.
[22] SPARC (SPARC Performs Automated Reasoning in Chemis-
try) program; http://ibmlc2.chem.uga.edu/sparc/; S. Hilal,
S. W. Karickhoff, L. A. Carreira, Quant. Struct. – Act. Relat.
1995, 14, 348–355.
[23] W. J. Tenn III, J. L. Murphy, J. K. Bim-Merle, J. A. Brown, et
al., J. Org. Chem. 2007, 72, 6075–6083.
[24] A. Buur, H. Bundgaard, E. Falch, Int. J. Pharm. 1985, 24, 43–
60.
[25] E. Fuseau, O. Petricoul, K. H. P. Moore, A. Barrow, T. Ibbot-
son, Clin. Pharmacokinet. 2002, 41, 801–811.
[26] E. K. Euranto, in The Chemistry of Carboxylic Acids and Esters
(Ed.: S. Patai), Wiley, Chichester, 1969, pp. 213.
[27] J. Iley, R. Moreira, E. Rosa, J. Chem. Soc. [Perkin Trans. II]
1991, 563–570.
References
[28] J. Iley, R. Moreira, T. Calheiros, E. Mendes, Pharm. Res.
1997, 14, 1634–1639.
[1] A. J. J. Wood, N. Engl. J. Med. 2002, 346, 257–270.
[2] C. M. Perry, A. Markham, Drugs 1998, 55, 889–922.
[3] A. K. Scott, Clin. Pharmacokinet. 1994, 27, 337–344.
[4] G. L. Plosker, D. McTavish, Drugs 1994, 47, 622–651.
[29] E. Valente, J. R. B. Gomes, R. Moreira, J. Iley, J. Org. Chem.
2004, 69, 3359–3367.
[30] G. Kohnstam, Adv. Phys. Org. Chem. 1967, 5, 121–172.
[5] B. D. Dulꢀry, M. A. Petty, J. Schoun, M. David, N. D. Hue-
[31] D. M. Muir, M. C. Whiting, J. Chem. Soc. [Perkin Trans. II]
1975, 1316–1320.
bert, J. Pharm. Biomed. Anal. 1997, 15, 1009–1020.
[6] S. R. Patel, H. Zhong, A. Sharma, Y. Kalia, Eur. J. Pharm. Bio-
pharm. 2007, 66, 296–301.
[32] H. Bundgaard, A. Buur, S.-C. Chang, V. H. L. Lee, Int. J.
Pharm. 1988, 46, 77–88.
[7] P. Telf–Hansen, P. Vries, P. Saxena, Drugs 2000, 60, 1259–
[33] B. Mudryk, S. Rajaraman, N. Soundararajan, Tetrahedron
Lett. 2002, 43, 6317–6318.
1287.
[8] J. Leppꢁnen, J. Huuskonen, T. Nevalainen, J. Gynther, et al.,
J. Med. Chem. 2002, 45, 1379–1382.
[34] P. Hohenberg, W. Kohn, Phys. Rev. 1964, 136, 864–871.
[35] R. G. Parr, W. Yang in Density Functional Theory of Atoms and
Molecules; Oxford University Press,:Oxford, U. K., 1989.
[9] F. Lopes, R. Moreira, J. Iley, Bioorg. Med. Chem. 2000, 8,
707–716.
[10] J. D. Larsen, H. Bundgaard, Int. J. Pharm. 1987, 37, 87–95.
[36] Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, et al., Gaussian, Inc., Wallingford
CT, 2004.
[11] J. D. Larsen, H. Bundgaard, V. H. Lee, Int. J. Pharm. 1988, 47,
103–110.
[37] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[12] H. Bundgaard, J. D. Larsen, J. Med. Chem. 1988, 31, 2066–
2069.
[38] C. T. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–
789.
[13] J. D. Larsen, H. Bundgaard, Int. J. Pharm. 1989, 51, 27–38.
[39] R. Krishnan, J. S. Binkley, R. Seeger, J. A. Pople, J. Chem.
Phys. 1980, 72, 650–654.
[14] H. Bundgaard, A. Buur, K. T. Hansen, J. D. Larsen, et al., Int.
J. Pharm. 1988, 45, 47–57.
[15] F. Lopes, R. Moreira, J. Iley, J. Chem. Soc. [Perkin Trans II]
1999, 3, 431–440.
[40] T. Clark, J. Chandrasekhar, G. W. Spitznagel, P. v. R.
Schleyer, J. Comput. Chem. 1983, 4, 294–301.
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.archpharm.com