6742
E. Zambon et al. / Tetrahedron 64 (2008) 6739–6743
3-(2-aminophenyl)propionate13 and methyl 2-aminomethyl-
benzoate18 are characterized by EM values about two order of
magnitude higher, while the intramolecular aminolysis of
solution that was kept at 80 ꢀC in an oil bath. In a typical kinetic run
the initial concentration of gabapentin is 0.0029 M. All reactions
were studied at a fixed ionic strength (
propriate time intervals samples of 50
quenched in an ice-water bath. Each sample of 20
2 mL of borate buffer (0.1 M, pH 9.0) in fluorimeter cuvettes and
these solutions were treated with 100 L of acetone solution of
m
¼0.5 M with NaCl). At ap-
L were withdrawn and
L was added to
substituted phenyl
g
-dimethylaminobutyrates,19 a less constrained
m
substrate, shows EM values comparable to that obtained in the
cyclization of gabapentin. Therefore, we can assume that the high
effective molarity of the amino group is important in determining
the observed reactivity of gabapentin although this effect is limited
by the relatively high conformational flexibility respect to more
constrained and pre-organized system.
m
m
fluorescamine (10 mM). The decrease in the concentration of
gabapentin was followed by measuring the fluorescence emission
at 480 nm. The concentration of gabapentin was calculated using
a calibration curve obtained by measuring the emission intensity of
solution at known concentration of gabapentin. The linearity range
3. Conclusions
of the assay was determined (up to a final concentration of 30 mM)
The intramolecular amide bond formation of gabapentin in
and the proper dilution in order to operate in the linearity range
was used.
aqueous solution is very slow at room temperature, t1/2y1 day
at 80 ꢀC and pH
2
whereas the cyclization of 3-(2-amino-
Sulfate buffer, made by mixing the proper amount of NaHSO4
and Na2SO4, was used for reactions at pH 2.24 and 3.29, acetate
buffer was used for reactions at pH 4.19 and 5.15, phosphate buffer
was used for reactions at pH 6.21, 7.16, and 8.15, and borate buffer
was used for reactions from pH 8.15 to 11.15. In each kinetic run the
pH of the reaction mixture was checked before and after the
reaction and no change in pH was observed. The rates of the re-
actions from pH 2.24 to 9.10 were determined at three different
concentrations of buffers (0.025, 0.060, and 0.10 M) at a fixed pH,
while the rate constants of the reactions from pH 9.80 to 11.15 were
determined at two different concentrations of borate buffer (0.05
and 0.10 M). When a buffer effect was observed (pH<9.80) the
first order buffer-independent rate constant k0 was obtained from
the intercept of plots of kobs versus the total buffer concentration;
when no buffer effect was observed (pH>9.80) the first-order ki-
netic constants reported are average values of different kinetic
runs.
phenyl)propionic acid at 39 ꢀC occurs with an half life of about 3.5 h
at pH 2.24.8 Responsible for this is probably the less rigid structure
of the substrate, which results in a lower effective molarity of the
amino group and the formation of a five-membered ring, more
strained than a six-membered ring. The inertness of gabapentin to
cyclization reaches a maximum at pH values between 5 and 7. The
process presents the maximum rate at basic pH values above pH
9.80 and this reactivity, not observed in the case of 3-(2-amino-
phenyl)propionic acid, is related to the higher basicity (pKa¼10.57
at 25 ꢀC) and nucleophilicity of the aliphatic NH2 group. The most
surprisingly outcome of our studies is the decrease of reactivity on
increasing the amino acid concentration (at pH 10.45). We can
speculate that increasing the gabapentin concentration up to the
solid state the cyclization may be very much depressed or even
absent and we are actively investigating on this important finding.
4. Experimental section
4.1. General
The experiments in D2O were performed from pD 2.24 to 11.15 at
buffer concentration 0.1 M. The pD value were measured in the
usual way using a standard pH meter and applying the following
correction: pD¼pHmeasuredþ0.4.20
Fluorescence measurements were carried out on a Perkin–
Elmer Luminescence Spectrometer LS 50B using fluorimeter cu-
vettes in polymethacrylate (lexcitation¼390 nm, lemission¼480 nm,
scan speed¼100 nm/min, scan range¼430–530 nm, excitation and
emission slit¼3.5 nm). NMR experiments were performed with
a Jeol 400 MHz NMR instrument operating at 399.78 MHz for 1H
and at 100.53 MHz for 13C and equipped with a variable tempera-
ture unit. All kinetic experiments were performed at 80 ꢀC. The pH
values of the reaction mixtures were determined at 80 ꢀC with a pH
meter Basic 20 equipped with a Crison glass electrode. The glass
electrode was standardized with buffer solution thermostatted at
80 ꢀC following the indications of the producer. The apparent dis-
sociation constants of gabapentin were determined at 25, 35, and
At least four independent experiments were carried out for each
reaction and the reported kinetic constants are average values.
4.2.2. NMR method
Experiments were performed at pH 10.45, with concentrations
of gabapentin >0.10 M. The experiments were carried out keeping
at 80 ꢀC in aqueous solution (5% D2O, pH 10.45, [borate]¼0.10 M,
m¼0.5 M with NaCl) with the concentration of gabapentin reported
in Table 1 in NMR tubes equipped with airtight screw caps. The
reactions have been monitored at time intervals of 30 min. The
decrease of the concentration of gabapentin was followed by
measuring the 1H NMR integrated area of the methylenic reso-
nances at 3.01 and 3.58 ppm. The increase of the concentration of
lactam was followed by measuring the 1H NMR integrated area of
the correspondent peaks at 2.86 and 3.83 ppm. The concentration
of gabapentin Cg was calculated by equation: Cg¼(a/(aþb))C0 where
a and b are the areas of gabapentin and lactam peaks, respectively,
C0 is the initial concentration of gabapentin. At least four in-
dependent experiments were carried out for each reaction and the
reported kinetic constants are average values.
45 ꢀC by potentiometric titration using
DMS Titrino equipped with Metrohm glass electrode
U Metrohm pH meter 716
a
U
6.0204.100. Authentic samples of gabapentin and 2-aza-spi-
ro[4,5]decan-3-one were synthesized by the Fine Chemicals Busi-
ness Unit of Zambon Group S.p.A. Both compounds have been
prepared with the high purity required for pharmaceutical
drugs and have been fully characterized (1H NMR, 13C NMR,
mass-spectrometry, and elemental analysis). Commercial reagents
were purchased from standard chemical suppliers. All experiments
were conducted with bidistilled water or deuterated water (99.9%
CIL).
4.3. Determination of pKa1 and pKa2
The pKa1 and pKa2 of gabapentin were determined at 25, 35, and
45 ꢀC by potentiometric titration. An aqueous gabapentin solution
(50.6 mL, 1.18 mM) containing 1 equiv of HCl and having an
ionic strength of 0.10 M (NaCl) were titrated with 1.5 mL of
1.016ꢁ10ꢂ1 M NaOH standard solution. The pKa1 and pKa2 values at
80 ꢀC were obtained by linear extrapolation of pKa values versus
temperature.
4.2. Kinetic measurements
4.2.1. Fluorimetric method
Kinetic experiments were carried out by addition of 100
a stock solution of gabapentin (50 mg/mL) to 10 mL of the buffered
mL of