Degradation kinetics of L-glutamine in aqueous solution

Article abstract of DOI:10.1016/S0928-0987(99)00047-0

The degradation kinetics of L-glutamine (Gln) in aqueous solution was studied as a function of buffer concentration, pH and temperature. Stability tests were performed using a stability-indicating high-performance liquid chromatographic assay. The degradation product of Gln was 5-pyrrolidone-2-carboxylic acid. The reaction order for Gln in aqueous solution followed pseudo-first-order kinetics under all experimental conditions. The maximum stability of Gln was observed in the pH range from 5.0 to 7.5. The pH-rate profile described by specific acid-base catalysis and hydrolysis by water molecules agreed with the experimental results. Arrhenius plots showed the temperature dependence of Gln degradation, and the apparent activation energy at pH 6.41 was determined to be 9.87x10⁴ J mol^-1. Copyright (C) 1999 Elsevier Science B.V.

Full text of DOI:10.1016/S0928-0987(99)00047-0

European Journal of Pharmaceutical Sciences 9 (1999) 75–78
www.elsevier.nl/locate/ejps
Degradation kinetics of L-glutamine in aqueous solution
*
Kanji Arii , Hideyuki Kobayashi, Toshiya Kai, Yukifumi Kokuba
Preclinical Development Laboratories, Hoechst Marion Roussel Ltd., 1541A-3 Nagahara, Yasu-cho, Yasu-gun, Shiga 520-2304, Japan
Received 16 November 1998; received in revised form 1 June 1999; accepted 30 June 1999
Abstract
The degradation kinetics of
L
-glutamine (Gln) in aqueous solution was studied as a function of buffer concentration, pH and
temperature. Stability tests were performed using a stability-indicating high-performance liquid chromatographic assay. The degradation
product of Gln was 5-pyrrolidone-2-carboxylic acid. The reaction order for Gln in aqueous solution followed pseudo-first-order kinetics
under all experimental conditions. The maximum stability of Gln was observed in the pH range from 5.0 to 7.5. The pH–rate profile
described by specific acid–base catalysis and hydrolysis by water molecules agreed with the experimental results. Arrhenius plots showed
4
21
the temperature dependence of Gln degradation, and the apparent activation energy at pH 6.41 was determined to be 9.87310 J mol
1999 Elsevier Science B .V . All rights reserved.
.

Keywords: L-Glutamine; Degradation; 5-Pyrrolidone-2-carboxylic acid
1
. Introduction
any precise kinetic study concerning the degradation of
Gln in aqueous solution. The purpose of this study was to
investigate the effects of pH and temperature on the
stability of Gln using a stability-indicating high-perform-
ance liquid chromatographic (HPLC) assay method.
L-Glutamine (Gln) is an important substrate for renal
ammonia production (Welbourne, 1987), hepatic
gluconeogenesis (Kimura et al., 1988), modulation of
muscle protein turnover (Wu and Thompson, 1990), and is
a metabolic fuel for enterocytes (Horvath et al., 1996;
Wilmore and Shabert, 1998). During total parenteral
nutrition (TPN), changes in the morphology and function
of the small intestine occur (Li et al., 1994), and the risk of
bacteremia and endotoxemia increase (Frankel et al., 1993;
Bai et al., 1996). The addition of Gln to TPN solution
maintains the integrity of the gut mucosa in postoperative
patients (Hulst et al., 1993; Morlion et al., 1998). Former-
ly, Gln was thought to be ‘non-essential’ and was not
included in TPN solution because of its relative instability
in solution. Several studies under simulated usage patterns
have shown that Gln is stable if the correct pharmaceutical
techniques are followed. At a pH just below neutrality, the
concentration of Gln in TPN solution decreased by 5% per
day at 378C (Souba et al., 1985). At room temperature, Gln
degradation was 0.7–0.9% per day in different parenteral
solutions (Khan et al., 1991), at ¯48C, it was minimal
2. Experimental procedures
2.1. Materials
Gln (.99.0%, reagent grade) was purchased from
Takara Kohsan (Tokyo, Japan). All other chemicals used
in this experiment were of analytical or HPLC grade. The
water was double-distilled.
2.2. Stability studies of Gln
In the stability studies, several different pH values of
citrate buffer (pH 1.21–6.41) and borate buffer (pH 7.39–
11.01) were used. The pH values were measured at the
experimental temperature using a pH meter equipped with
a combination electrode. The concentration of the total
citrate buffer was 13.8–110 mM and that of total borate
buffer was 25–140 mM. The ionic strength was adjusted to
(
Wilmore, 1994). However, there has been no report of
0
.5 by using potassium chloride. Weighed amounts of Gln
*
Corresponding author. Tel.: 181-77-588-2075.
E-mail address: kanji-arii@sam.hi-ho.ne.jp (K. Arii)
were dissolved in the buffer solutions preheated to a
0
928-0987/99/$ – see front matter
 1999 Elsevier Science B .V . All rights reserved.
PII: S0928-0987(99)00047-0

7
6
K. Arii et al. / European Journal of Pharmaceutical Sciences 9 (1999) 75–78
desired temperature to produce a final Gln concentration of
termined potentiometrically by measuring the pH solutions
2
1
21
0
.2 mg ml . Aliquots (200 ml) of the solutions were
prepared by mixing Gln solution with 0.1 mol l hydro-
2
1
poured into 200-ml glass vials and sealed. The sealed
solutions were stored in a constant temperature water bath
at 708C. At pH 6.41, the samples were studied at four
different temperatures (40, 50, 60 and 708C). Aliquots of
the solution were withdrawn at appropriate time intervals
and assayed immediately.
chloric acid or 0.1 mol l sodium hydroxide solution at
24
708C. The initial concentration of Gln was 5.0310 mol
per 50 ml. The same method has been used in previous
research (Takeuchi et al., 1995). The dissociation constants
of Gln, pK and pK , were 2.21 and 9.01, respectively.
1
2
2
.3. Analytical procedures
3
. Results and discussion
The HPLC system consists of a solvent delivery pump
(
model 655-15, Hitachi, Tokyo, Japan), a high-pressure
sampling valve (model 655, Hitachi), and an integrator
model 655-71B, Hitachi). Column effluents were moni-
3.1. Degradation product
(
A typical chromatogram of the reaction mixture in 27.5
mM citrate buffer of pH 6.41 is shown in Fig. 1. The peak
designated as 2 is the parent compound Gln, and that
designated as 1 is the internal standard L-asparagine. There
have been several reports on the degradation products of
Gln. During storage, Gln hydrolyzed to L-glutamic acid
(Glu) and ammonia (Heller et al., 1967), while at higher
temperature, the amino-group of Gln was labile and
yielding pGlu (Stehle et al., 1984). In this study, the
OPA-reactive substance such as Glu (retention time53.2
min) was not observed. This finding indicated that hydro-
lytic deamination of Gln to Glu was not the cause of Gln
loss. In the analysis of reaction mixtures pH 1.93, 6.41 and
tored with a fluorimeter (model RF-530, Shimazu, Kyoto,
Japan) and the excitation and emission wavelength were
set to 330 and 470 nm, respectively. The separation was
carried out using a C18 column (20034 mm I.D., Nucleosil
5
C , Macherey-Nagel, Duren, Germany). The mobile
18
phase was a 65:35 mixture (v/v) of 12.5 mM di-sodium
hydrogenphosphate and methanol. The mobile phase was
2
1
delivered at a constant rate of 1.0 ml min . The remain-
ing Gln was determined by a pre-column derivatization
with o-phthalaldehyde (OPA) (Lindroth and Mopper,
1
979). The concentration of Gln was determined by a
method of peak area ratio compared to the peak area ratio
of samples of standard solutions from the calibration curve
2
4
11.01, the decrease of Gln was 6.32310 M (708C, 0.6
2
4
24
(
internal standard: L-asparagine).
h), 4.78310 M (708C, 10 h) and 5.71310 M (708C, 3
h), respectively, and the formation of pGlu was 6.183
2
4
24
24
2
.4. Degradation product of Gln
10 , 4.72310 , and 5.60310 M, respectively. There
was close agreement between the loss of Gln and the
The degradation product of Gln was isolated by the
HPLC system (solvent delivery pump: model 655-15,
sampling valve: model 655, integrator: model 655-71B,
Hitachi). The mobile phase (0.1% perfluorobutyric acid
2
1
solution) was delivered at a constant rate of 0.7 ml min
and the column eluent was monitored at 210 nm (UV-
detector: model 655, Hitachi). The mass spectra of the
degradation product was recorded by a Frit FAB-MS
spectrometry system (model MS-LX2000, JEOL, Tokyo,
Japan) via a direct inlet. The FAB-MS matrix was metha-
nol containing 3% of glycerol and the flow rate was 0.2 ml
2
1
min . The FAB-MS measurement was done at a xenon
acceleration voltage of 3 kV and an ion acceleration
voltage of 2 kV. The degradation product of Gln wa1s
identified as 5-pyrrolidone-2-carboxylic acid (pGlu) (MH
ion at m/z 130) by the authentic compound.
The concentration of pGlu was determined by a pre-
column derivatization with 4-nitrophenacyl bromide. The
method used was established in earlier research (Bousquet
et al., 1983).
2
.5. Measurement of dissociation constants of Gln
Fig. 1. A typical chromatogram of a partially degraded sample of Gln at
The apparent dissociation constants of Gln were de-
708C in buffer solution. (1) internal standard, (2) Gln.

K. Arii et al. / European Journal of Pharmaceutical Sciences 9 (1999) 75–78
77
appearance of pGlu. Extensive loss of Gln was the result of
the intramolecular cyclization to pGlu.
3
.2. Order of reaction
The time-course of the concentration of Gln at various
pH, at constant temperature (708C) and constant ionic
strength (0.5) is shown in Fig. 2. In all determinations,
there was a linear relationship between storage time and
logarithmic remaining percent of Gln, and the degradation
of Gln followed pseudo-first-order reaction kinetics in
aqueous solution. The observed first-order rate constant,
Kobs, was determined from the slope of the graph by the
method of least squares. The first-order plots under all
experimental conditions yielded coefficients of over r52
0
.99.
3
.3. pH–rate profile of Gln
Fig. 3 shows the effects of citrate and borate on the
Fig. 3. Plots of the Kobs versus the total buffer concentration at 708C and
m50.5 for the degradation of Gln. pH values: (j) 1.21, (d) 1.93, (m)
degradation of Gln at various pH. Kobs increased linearly
with the increase of buffer concentrations; thus, K, the
apparent rate constant which depends purely on pH, is
obtained by the extrapolation to zero buffer concentration.
Fig. 4 shows the pH–rate profile constructed from the
buffer-free first-order rate constants and the pH values. The
observed rate in this pH–rate profile was a summation of a
series of rates of catalytic reactions by the hydrogen and
hydroxyl ions and water molecules. In the isoelectric
regions, Gln is more stable and the degradation rate
constant approaches a plateau. Therefore, the degradation
of neutral species of Gln is due mainly to hydrolysis by
water molecules (Fig. 4a). The degradation of Gln is rapid
2
9
.87, (.) 3.93, (♦) 4.80, (\) 5.61, (ȣ) 6.41, (^) 7.39, (h) 8.31, (s)
.31, (s dot) 11.01.
in acidic regions, and the rate decreases with an increase in
pH, which indicates specific hydrogen-ion catalysis (Fig.
4
d) and water hydrolysis (Fig. 4b) of the cationic species
of Gln. In the alkaline regions, the rate indicates specific
hydroxide-ion catalysis (Fig. 4e) and water hydrolysis
(
Fig. 4c) of the anionic species of Gln. Similar degradation
Fig. 4. pH–rate profile for the degradation of Gln in aqueous solution at
08C. The points are the experimental values, and the lines are the
7
theoretical curves. (a) Water-catalyzed degradation of neutral species, (b)
water-catalyzed degradation of cationic species, (c) water-catalyzed
degradation of anionic species, (d) hydrogen-ion-catalyzed degradation of
cationic species, and (e) hydroxide-ion-catalyzed degradation of anionic
species.
Fig. 2. Apparent first-order plots for the degradation of Gln at 708C in
buffer solution. pH values: (j) 1.21, (d) 1.93, (m) 2.87, (.) 3.93, (♦)
4
.80, (\) 5.61, (ȣ) 6.41, (^) 7.39, (h) 8.31, (s) 9.31, (s dot) 11.01.

7
8
K. Arii et al. / European Journal of Pharmaceutical Sciences 9 (1999) 75–78
Bai, M.X., Jiang, Z.M., Liu, Y.W., Wang, W.T., Li, D.M., Wilmore, D.W.,
profiles were obtained for L-alanyl-L-glutamine (Arii et al.,
998) and orbifloxacin (Morimura et al., 1995). Thus, the
overall velocity is equal to the sum of the rates of these
reactions, as shown in the following equation
1
7
996. Effects of alanyl-glutamine on gut barrier function. Nutrition 12,
93–796.
1
Bousquet, E., Guarcello, V., Morale, M.C., Rizza, V., 1983. Analysis of
-pyrrolidone-2-carboxylate ester by reverse phase high-performance
5
liquid chromatography. Anal. Biochem. 131, 135–140.
Frankel, W.L., Zhang, W., Afonso, J., Klurfeld, D.M., Don, S.H., Laitin,
E., Deaton, D., Furth, E.E., Pietra, G.G., Naji, A., Rombeau, J.L.,
K 5 K
9
? [a ] ? f 1 K
9
? f 1 K
? f 1 K99 ? f3
H O 2 H O
2 2
H
H
1
H O
1
2
1
K
O99H ? [aOH] ? f3
(1)
1
993. Glutamine enhancement of structure and function in transplanted
small intestine in the rat. J. Parenter. Enter. Nutr. 17, 47–55.
Harned, H.S., Hamer, W.J., 1933. The ionization constant of water and the
dissociation of water in potassium chloride solutions from electromo-
tive forces of cells without liquid junction. J. Am. Chem. Soc. 55,
2194–2206.
where K is the second-order rate constant for the
9
H
hydrogen-ion-catalyzed degradation of the cationic species;
9O9H is the second-order rate constant for hydroxide-ion-
K
catalyzed degradation of the anionic species; K
9
, K
H O H O
2 2
and K99 are the first-order rate constants for the water-
Heller, L., Becher, A., Beck, A., M u¨ ller, F., 1967. On the problem of
utilization of infused amino acid solutions. Klin. Wochenschr. 45,
H O
2
catalyzed degradation of cationic, neutral and anionic
species, respectively; and aH is the hydrogen-ion activity
and aOH is the hydroxyl-ion activity. By introducing the
fractions of Gln in the cationic ( f ), neutral ( f ) and
3
17–318.
Horvath, K., Jami, M., Hill, I.D., Papadimitriou, J.C., Magder, L.S.,
Chanasongcram, S., 1996. Isocaloric glutamine-free diet and the
morphology and function of rat small intestine. J. Parenter. Enter.
Nutr. 20, 128–134.
Hulst, R.R.W.J., Kreel, B.K., Meyenfeldt, M.F., Brummer, R.J.M.,
Arends, J.W., Deutz, N.E.P., Soeters, P.B., 1993. Glutamine and the
preservation of gut integrity. Lancet 341, 1363–1365.
Khan, K., Hardy, G., McElroy, B., Elia, M., 1991. The stability of
L-glutamine in total parenteral nutrition solutions. Clin. Nutr. 10,
193–198.
1
2
anionic ( f ) forms, the dissociation constants of Gln, K
3
1
and K , and the dissociation constant of water, K 512.80
2
W
(
Harned and Hamer, 1933), the following overall rate
expression, Eq. (2), was obtained:
3
2
K 5 K
9
? [a ] 1 K
9
? [a ] 1 K
? [a ] ? K 1 K99
H O H 1 H O
2 2
H
H
H O
H
2
Kimura, R.E., LaPine, T.R., Johnston, J., Ilich, J.Z., 1988. The effect of
fasting on rat portal venous and aortic blood glucose, lactate, alanine,
and glutamine. Pediatr. Res. 23, 241–244.
Li, J., Langkamp-Henken, B., Suzuki, K., Stahlgren, L.H., 1994.
Glutamine prevents parenteral nutrition-induced increases in intestinal
permeability. J. Parenter. Enter. Nutr. 18, 303–307.
Lindroth, P., Mopper, K., 1979. High performance liquid chromatographic
determination of subpicomole amounts of amino acids by precolumn
fluorescence derivatization with o-phthaldialdehyde. Anal. Chem. 51,
2
?
K ? K 1 K99 ? K ? K ? K ? [a ]/([a ] 1 K ? [a ]
1
2
OH
1
2
W
H
H
1
H
1
K ? K )
(2)
1
2
The rate constants were estimated by the least-squares
method by fitting Eq. (2) to the experimental rate constants
2
1
21
21
at various pH: K
9
51.45310 M
h
, K99 53.52 M
OH
21 21
H
2
1
21
h
, K
9
51.29 h , K99 51.09310
h
, K
H O
2
5
H O
H O
2
2
2
2
21
1
667–1674.
4
.87310
h . The calculated values are in good fit with
Morimura, T., Ohno, T., Matsukura, H., Nobuhara, Y., 1995. Degradation
kinetics of the new antibacterial fluoroquinolone derivative, orbiflox-
acin, in aqueous solution. Chem. Pharm. Bull. 43, 1052–1054.
Morlion, B.J., Stehle, P., Wachtler, P., Siedhoff, H.P., Koller, M., Konig,
W., F u¨ rst, P., Puchstein, C., 1998. Total parenteral nutrition with
glutamine dipeptide after major abdominal surgery: a randomized,
double-blind, controlled study. Ann. Surg. 227, 302–308.
the experimental data (r50.99) and this equation
adequately describes the degradation kinetics in the pH
ranges studied.
3
.4. Dependence of degradation rate on temperature
Souba, W.W., Smith, R.J., Wilmore, D.W., 1985. Glutamine metabolism by
the intestinal tract. J. Parenter. Enter. Nutr. 9, 608–617.
The temperature dependence of Gln degradation in
buffer solution at pH 6.41 (27.5 mM citrate) was studied in
the temperature range of 40–708C. Arrhenius plots of log
rate versus the reciprocal of the absolute temperature were
constructed to describe the results of this investigation.
Linearity (r50.99) of the regression line in this plot was
observed in the temperature range of 40–708C. The
activation energy for degradation was determined to be
¨
Stehle, P., Pfaender, P., Furst, P., 1984. Isotachophoretic analysis of a
synthetic dipeptide L-alanyl-L-glutamine. Evidence for stability during
heat sterilization. J. Chromatogr. 294, 507–512.
Takeuchi, Y., Sunagawa, M., Isobe, Y., Hamazume, Y., Noguchi, T., 1995.
Stability of a 1b-methylcarbapenem antibiotic, meropenem (SM-7338)
in aqueous solution. Chem. Pharm. Bull. 43, 689–692.
Welbourne, T.C., 1987. Interorgan glutamine flow in metabolic acidosis.
Am. J. Physiol. 253, F1069–F1076.
Wilmore, D.W., 1994. Glutamine and the gut. Gastroenterology 107,
1885–1886.
Wilmore, D.W., Shabert, J.K., 1998. Role of glutamine in immunologic
responses. Nutrition 14, 618–626.
4
21
9
.87310 J mol
from the slope of this plot and the
21
frequency factor (log A) was determined to be 31.5 h
.
Wu, G., Thompson, J.R., 1990. The effect of glutamine on protein
turnover in chick skeletal muscle in vitro. Biochem. J. 265, 593–598.
References
Arii, K., Kai, T., Kokuba, Y., 1998. Degradation kinetics of L-alanyl-L-
glutamine and its derivatives in aqueous solution. Eur. J. Pharm. Sci.
7
, 107–112.