Degradation kinetics of gonadorelin in aqueous solution

Article abstract of DOI:10.1021/js9601187

The degradation kinetics of gonadorelin were investigated systematically with reversed-phase high-performance liquid chromatography. The stability- indicating properties of this system were checked with photodiode array detection and by comparison with capillary zone electrophoretic analysis. Influences of gonadorelin concentration, pH, temperature, buffer ions, and ionic strength on the degradation kinetics were studied. The pH-Iog κ(abs) profile can be divided into three pads, a proton, a solvent, and a hydroxyl- catalyzed section, with different degradation products. These degradation products were characterized by mass using LC-MS. Gonadorelin is most stable at pH 5-5.5 with a half-life of 70 days at 70 °C. The overall degradation rate constant as a function of the temperature under acidic and alkaline conditions obeys the Arrhenius equation. The gonadorelin concentration and the concentrations of acetate, phosphate, borate, and carbonate buffer have no influence on the decomposition rate of the κ(abs). Increasing ionic strength led to higher κ(abs) at pH 2 and lower κ(abs) at pH 9, but influences were relatively small.

Full text of DOI:10.1021/js9601187

+
+
Degradation Kinetics of Gonadorelin in Aqueous Solution
M
W
ARNIX A. HOITINKX, JOS H. BEIJNEN, AUKE
BULT, OEDS A. G. J. VAN DER
HOUWEN, JACK
N
IJHOLT AND
,
ILLY J. M. UNDERBERG
Received March 6, 1996, from the Department of Pharmaceutical Analysis, Faculty of Pharmacy, Utrecht University, Sorbonnelaan 16,
NL-3584 CA, Utrecht, The Netherlands.
Final revised manuscript received July 15, 1996 .
Accepted for
publication July 18, 1996X.
Experimental Section
Abstract
0 The degradation kinetics of gonadorelin were investigated
systematically with reversed-phase high-performance liquid chromatog-
raphy. The stability-indicating properties of this system were checked
with photodiode array detection and by comparison with capillary zone
electrophoretic analysis. Influences of gonadorelin concentration, pH,
temperature, buffer ions, and ionic strength on the degradation kinetics
Ch em ica lssGonadorelin acetate was obtained from Sigma Chemi-
cal Company, St. Louis, MO. All other chemicals used were of
analytical grade, and deionized water was applied throughout the
study.
RP -HP LC System sThe liquid chromatograph consisted of a
Waters Model 510 pump (Waters Associates, Milford, MA), a Gilson
231 sample injector with a 20 µL loop, a Gilson 401 dilutor (both from
Gilson, Villiers, France), and an Applied Biosystems 785A program-
mable absorbance detector (Separations, H. I. Ambacht, The Neth-
erlands). The column used was a LiChrospher 100 RP-18 (5 µm) 125
× 4 mm i.d. (Merck, Darmstadt, Germany). The mobile phase
consisted of 16% acetonitrile (w/w) with 0.1% trifluoroacetic acid (v/
v), the flow was set at 1 mL/min, and detection was performed at
214 nm. Chromatograms were recorded on a BD 40 recorder (Kipp
& Zonen, Emmen, The Netherlands), and peak heights were measured
for quantitation.
Ca p illa r y Zon e Electr op h or esis (CZE)sAll CZE experiments
were performed on a Prince injection system (Lauerlabs, Emmen, The
Netherlands) equipped with an 80 cm 75 µm i.d. fused silica capillary
(Bester, Amstelveen, The Netherlands) and an Applied Biosystems
785A programmable absorbance detector (Separations). Detection
was performed 60 cm from the injector at 214 nm, and electrophero-
grams were recorded on a BD 40 recorder (Kipp & Zonen). Acetate
buffer, 200 mM, pH 4 was used as running buffer. Injection was
performed during 6 s at 40 mbar overpressure, and analysis took place
at 20 kV with a typical current of 50 µA.
p H Mea su r em en tsAll pH measurements were performed on a
Consort P514 meter (Salm & Kipp, Breukelen, The Netherlands)
equipped with a Slim-trode electrode (Hamilton, Darmstadt, Ger-
many). The pH was measured before and after degradation. All pH
values were determined at the appropriate degradation temperature.
Liqu id Ch r om a togr a p h y-Ma ss Sp ectr om etr y (LC-MS)sThe
liquid chromatograph consisted of a Waters model 510 pump, a
Waters UK6 injector (both from Waters Associates) and an Applied
Biosystems 785A programmable absorbance detector (Separations).
The column used was a Superspher 100 RP-18 (5 µm) 119 × 2 mm
i.d. ( Merck). The mobile phase consisted of 16% acetonitrile (w/w)
with 0.1% trifluoroacetic acid (v/v), the flow was set at 0.1 mL/min,
and detection was performed at 214 nm. Chromatograms were
recorded on a BD 40 recorder (Kipp & Zonen).
This LC system was directly coupled to a VG platform II mass
spectrometer (Fisons, Altrincham, U.K.), operated in the positive ion
mode. The scanned mass range was 100-1500 m/ z, scan time was
set at 3 s, and low and high mass resolution were 12.5 and 15
(instrumental units), respectively. Data aqcuisition and calculation
of the m/ z ratios was done with MassLynx version 2.0 (Fisons).
Samples:
were studied. The pH
proton, solvent, and
degradation products. These degradation products were characterized
by mass using LC MS. Gonadorelin is most stable at pH 5 5.5 with a
half-life of 70 days at 70 C. The overall degradation rate constant as
−
log kobs profile can be divided into three parts, a
a
a
hydroxyl-catalyzed section, with different
−
−
°
a function of the temperature under acidic and alkaline conditions obeys
the Arrhenius equation. The gonadorelin concentration and the concen-
trations of acetate, phosphate, borate, and carbonate buffer have no
influence on the decomposition rate of the analyte. Increasing ionic
strength led to higher kobs at pH 2 and lower kobs at pH 9, but influences
were relatively small.
Introduction
Gonadorelin, or luteinizing hormone-releasing hormone
(LH-RH, Figure 1), is a peptide consisting of 10 amino acid
residues. Since the isolation and identification of gonadorelin
in 1971,1 a lot of research in the therapeutic field has been
done and structural analogues have been developed with
considerably higher potency than gonadorelin itself. Several
analogues are now being used in therapy for, for example,
prostate cancer, endometriosis, and breast cancer.2
Limited research, however, was performed on the degrada-
tion of gonadorelin and its analogues.3-9 Helm et al.3 inves-
tigated the kinetics of gonadorelin and triptorelin between pH
2 and 8, but the pH range was too small to characterize the
different pH-dependent degradation reactions. The same
holds for the gonadorelin study of Powell et al.4 and the
nafarelin study of J ohnson et al.5 Strickley et al.6 investigated
systematically the degradation of RS-26306, a gonadorelin
antagonist, and found that the degradation reactions can be
divided into a proton-catalyzed deamidation, a solvent-
catalyzed hydrolysis, and a hydroxyl-catalyzed epimerization.
The purpose of this study was to extend insight into the
degradation mechanisms and the influences of pH, tempera-
ture, buffer ions, and ionic strength on the reaction kinetics.
This study will be used as a reference in the stability research
of other LH-RH analogues. Because of the relatively small
chain of 9-10 amino acids and structural similarities, LH-
RH analogues are suitable to investigate the structure-
chemical stability relationships of peptides in general. The
results will extend the knowledge about the chemical degra-
dation of amino acid residues and can be used to predict the
chemical stability of peptides.
(1) 250 µg/mL gonadoreline, 0.01 M HClO4 (pH 2)
(2) 500 µg/mL gonadoreline, 25 mM acetate buffer pH 5
(3) 250 µg/mL gonadoreline, 25 mM borate buffer pH 9
Five microliters of each sample was injected into the LC-MS
system, and the m/ z ratios of P1-P4 were determined.
Deter m in a tion of th e p Ka of th e Tyr osin e Resid u esThe pKa
of the tyrosine residue was determined spectrophotometrically on a
UV 140 double-beam spectrophotometer (Shimadzu, Tokyo, J apan)
at 70 °C. To set the pH of the solutions, 10 mM carbonate buffers
X Abstract published in Advance ACS Abstracts, September 1, 1996.
© 1996, American Chemical Society and
American Pharmaceutical Association
S0022-3549(96)00118-9 CCC: $12.00
Journal of Pharmaceutical Sciences / 1053
Vol. 85, No. 10, October 1996

+
+
Figure 1sChemical structure of native gonadorelin.
(pH 8-12) were used and absorbance was measured at 240 nm. All
experiments were performed under nitrogen to prevent oxidation of
gonadorelin.
F lu or escen ce Exp er im en tssThe fluorescence experiments were
performed on the RP-HPLC system described before, which was
expanded with a J asco Model 821-FP spectrofluorometric detector
(J apan Spectroscopic Co., Hachioji City, J apan). The excitation and
emission wavelength were set at 278 and 348 nm, respectively.
Sta n d a r d Degr a d a tion Con d ition ssDegradation experiments
were performed in flame-sealed 1 mL glass ampules, which were kept
in a thermostated water bath at the appropiate temperature. Sam-
pling of the ampules was done over a range of 2 times the estimated
half-life. For each degradation curve 10 measurement points were
used and all experiments were performed in duplicate. The gona-
dorelin concentration at pH <6 was 10 µg/mL, and above pH 6, 20
µg/mL. Standard degradation conditions were 70 °C, an ionic strength
of 0.1 M, adjusted with sodium perchlorate, and a buffer concentration
of 25 mM.
Rea ction Or d er a n d Sta n d a r d Devia tion in kobssThe reaction
order and standard deviation in kobs were determined in acidic and
alkaline solution. At pH 2.0 and pH 8.4, perchloric acid and 25 mM
borate buffer were used, respectively.
In flu en ce of th e Gon a d or elin Con cen tr a tion on kobssThe
influence of the gonadorelin concentration was investigated in acidic
and alkaline solution. At pH 2.1 and pH 8.5, perchloric acid and 25
mM borate buffer were used, respectively. The gonadorelin concen-
tration was varied between 10 and 25 µg/mL.
In flu en ce of th e p H on kobssThe influence of pH on kobs was
determined between pH 1 and 12. Buffers used were perchloric acid
below pH 3, acetate in the pH range 4-5.5, phosphate in the pH range
6-7, borate in the pH range 7.5-9, carbonate in the pH range 9-10,
and sodium hydroxide above pH 10.
Ar r h en iu s P lot sThe influence of the temperature on kobs was
determined at at pH 2.1, pH 5.1, and pH 8.4 using perchloric acid
solutions, 25 mM acetate buffer, and 25 mM borate buffer, respec-
tively. At pH 2.1 and pH 8.4 the reaction rate constant was
determined at 37, 50, 60, 70, and 80 °C, and at pH 5.1 the temperature
was varied between 70 and 90 °C.
In flu en ce of Bu ffer Ion s on kobssThe influence of buffer ions
on kobs was determined for acetate, phosphate, borate, and carbonate
by performing experiments at buffer concentrations between 10 and
200 mM at constant pH. For acetate buffer solutions, the pH was
5.0; for phosphate, 6.2; for borate, 8.5; and for carbonate, 9.3. The
experiments with phosphate buffer were performed at 80 °C, all other
experiments were performed at 70 °C.
Figure 2
column (
by solid lines (
dashed lines (− − −). Degradation conditions: pH 9, gonadorelin concentration
10 g/mL, ionic strength 0.1, 25 mM borate buffer, and 70 C.
s
Degradation curve of gonadorelin obtained with a LichroSpher RP-18
0) and a LichroSpher RP-8 column (
+
). The first-order fit is represented
s), and the experimental curve after more than 2 half-lives by
µ
°
relationships were found between the log of the acetonitrile
concentration and the capacity factor of gonadorelin. An
acetonitrile percentage of 16% w/w was selected for all further
analyses with a capacity factor of 7-9 for gonadorelin.
A carryover of approximately 10% was noticed. This is most
likely due to adsorption of gonadorelin to the autosampler
tubing. By rinsing the autosampler with a 20% w/w aceto-
nitrile/water mixture instead of water, the injection device was
adequately cleaned between two runs and carryover was
diminished. As a result, the calibration curve was linear (R
> 0.99) in the 5-25 µg/mL range of interest, however, with a
negative intercept due to the loss of LH-RH in the autosam-
pler tubing. The injection reproducibility at a concentration
of 15 µg/mL was 3% (n ) 10).
The RP-HPLC system was checked for its stability-indicat-
ing properties. Degradation curves were produced at pH 2
and 9. The acidic samples gave a good first order fit (R <
-0.99), but the alkaline samples resulted in a nonlinear curve,
indicating either non-first-order kinetics or non-stability-
indicating RP-HPLC conditions. Experiments, in which the
parent peak was degraded to less than 2% of the original
concentration, showed that one of the alkaline degradation
products practically coeluted with the parent.
In flu en ce of Ion ic Str en gth on kobssThe ionic strength (µ) was
varied between 0.01 and 1 M using sodium perchlorate while the pH
and buffer concentration were kept constant. At pH 2.0 and pH 9.0
perchloric acid solutions and 25 mM borate buffer were used,
respectively.
Results and Discussion
An a lytica l P r oced u r essAn RP-HPLC system was devel-
oped for analyzing the LH-RH samples. Acetonitrile was
chosen as the modifier, and 0.1% v/v trifluoroacetic acid was
added as acidifying and ion-pairing reagent. Initially, the
stationary phase was a Lichrosorb RP-8 column. Nonlinear
To solve this separation problem, the RP-HPLC conditions
were further optimized. At pH 9 degraded samples were
analyzed on an RP-8 and an RP-18 column. The kinetic
1054 / Journal of Pharmaceutical Sciences
Vol. 85, No. 10, October 1996

+
+
Table 1
s
Influence of the Gonadorelin Concentration on the Degradation
Table 2sm/z Ratios of the Degradation Products of Gonadorelin
Rate Constant kobs
M
+
(m/z)
2H+
M
+
(m/z)
H+
Theor M
Value (m/z)
+
H+
Gonadorelin
Concn (
kobs at pH 2.1
kobs at pH 8.5
Product
1
1
µg/mL)
(10-6 s-
)
(10-6 s-
)
Gonadorelin
592.0
592.2
592.0
374.6
1182.8
1183.7
1182.8
748.7
1182.6
1183.6
1182.6
748.4
10
3.2
2.9
3.0
3.1
±
±
±
±
0.2
9.4
8.2
8.3
8.2
±
±
±
±
0.4
P1
P2
P3
P4
15
20
25
0.2
0.2
0.2
0.3
0.3
0.3
453.5
453.2
results are shown in Figure 2. There is a significant difference
between the performance of the two columns, as an interfering
degradation product peak coeluted with the parent peak when
the RP-8 column was used, but was separated on the RP-18
column.
providing the reaction rate constant kobs as the slope of plots
of the natural logarithm of the residual drug concentration
versus time.10 The mean kobs values at pH 2.0 and 8.4 were
found to be (3.7 ( 0.26) × 10-6 s-1 (n ) 6) and (7.9 ( 0.33) ×
10-6 s-1 (n ) 6), respectively. This is acceptable for kinetic
analysis.
In flu en ce of th e Con cen tr a tion on kobssFrom Table 1
it can be concluded that there is no significant influence of
the gonadorelin concentration on the kobs at pH 2.1 and 8.5 in
the studied range of 10-25 µg/mL. These results confirm the
first-order reaction kinetics and indicate that there is no
interaction between the gonadorelin molecules at concentra-
tions below 25 µg/mL.
Degr a d a tion P r od u ctssTypical chromatograms of de-
graded gonadorelin are shown in Figure 3. The m/ z ratios
of the peaks P1-P4 were determined using LC-MS, and the
results are depicted in Table 2. The MS spectrum example
of P2 in Figure 4 gives an impression of the sensitivity and
accuracy of the LC-MS system.
Under acidic conditions (pH < 3) there is one major
degradation product, P1, which is probably deamidated go-
nadorelin.7,11 In the pH range 4-6 solvent catalysis plays the
most important role. Hydrolysis of the peptide bond on the
amino site of the serine residue gives rise to P3, the hep-
tapeptide Ser-Tyr-Gly-Leu-Arg-Pro-Gly-NH2, and P4, the tri-
peptide pyroGlu-His-Trp.6-9 Above pH 7 hydroxyl-catalyzed
degradation emerges as the most important pathway, which
results in products P1 and P2. P2 is probably the D-Ser4
gonadorelin epimer,7,11,12 and on the basis of the capacity
For all other experiments an RP-18 column was used. The
range over which the degradation was studied was kept less
than 2 half-lives, because under alkaline conditions some
degradation products elute close to the parent peak and
further degradation disturbs the base-line separation. This
base-line disturbance is also responsible for the deviation from
pseudo-first-order kinetics as depicted in Figure 2.
Further evaluation of the RP-HPLC system was performed
with a photodiode array detector, which did not indicate any
change in the UV spectrum of the parent peak during the
course of the degradation process. The degradation products
of gonadorelin at pH 2 and 9 have the same UV spectrum as
the parent peak, indicating that the chromophores are stable.
The photodiode array detector therefore gives no information
about the stability-indicating properties of the RP-HPLC
system.
A CZE system was used to confirm the separation of the
products of acidic and alkaline degradation on RP-HPLC. Both
RP-HPLC and CZE gave the same kobs for the acidic degrada-
tion, but CZE detected fewer peaks and a smaller kobs for the
alkaline degradation. The separation power of CZE appears
to be less than that of RP-HPLC.
Rea ction Or d er a n d Sta n d a r d Devia tion in kobssIn the
pH range 1-12 the gonadorelin concentration decreased over
the first 2 half-lives according to pseudo-first-order kinetics,
Figure 3sRP-HPLC chromatograms of the degradation of gonadorelin at (A) pH 2.0, (B) pH 5.3, and (C) pH 8.4.
Journal of Pharmaceutical Sciences / 1055
Vol. 85, No. 10, October 1996

+
+
Figure 5
s
pH
−
log kobs profile of gonadorelin. Degradation conditions: pH range
1−
12 (70 °C), ionic strength 0.1, T ) 70 °C, buffer concentration 25 mM.
H+
AH3
M0 ) k
(2)
(3)
(4)
(5)
2+
+
M1 ) kSAH 2+ + kH +K
AH2
1
3
+
OH-
AH3
M2 ) kSAH +K + kAHHK1K2 + k 2+K
1
W
2
+
-
M3 ) kAS HK1K2 + kHA-K K K + kOH +K K
1
2
3
AH2
1
W
-
M4 ) kSA-K K K + kOH K1K2KW
(6)
(7)
1
2
3
AH
-
M5 ) kOAH K1K2K3KW
-
Figure 4sRaw single-scan mass spectrum of degradation product P2.
Because of the limited amount of gonadorelin available, the
pKa values could not be determined with titration techniques.
However, the pKa value of tyrosine could be determined
spectrophotometrically and was 9.5 ( 0.1 at 70 °C.
The pKa of the histidine residue can be calculated using the
tyrosine pKa of 9.5 with eq 1 reduced by deleting the terms
M5/[H+]4 and K1K2K3/[H+]3. This is allowed because in this
pH range they do not contribute significantly. The kinetically
determined pKa is 5.8 ( 0.4, in accordance with the literature
value of 6.1.14 The relatively large standard deviation in the
pKa of histidine is caused by two factors: firstly, the inflection
in the pH profile due to the histidine residue is not very
pronounced; and secondly, the amount of data obtained in this
area is limited.
factor and the m/ z ratio, P1 is deamidated gonadorelin, the
same as the degradation product of the proton-catalyzed
reaction.
In flu en ce of th e p H on kobssThe pH-log kobs profile is
shown in Figure 5. Data were fitted according to a model
earlier developed by Van der Houwen et al.,13 which accounts
for changes in kobs caused by protolytic equilibria. Depending
on the degree of (de)protonation, different species are defined
+
by their own micro reaction constants kS, kH , and kOH-. Micro
reaction constants of kinetically indistinguishable reactions
are combined to macro reaction constants Mx. Gonadorelin
has three ionizable functions in the pH range of interest:
imidazole (His), phenol (Tyr), and guanidine (Arg). This
results in four species: AH32+, AH2+, AH, and A-. The pH-
log kobs profile should therefore be fitted using eq 1 with six
macro constants.11
The values for M0 to M4 of the fit can be found in Table 3.
In the pH range 1-3 there is specific acid catalysis with a
H+
AH3
k
2+ value of 4.0 × 10-4 M-1 s-1 (70 °C) derived from M0. At
H+
AH3
80 °C k 2+ is 9.0 × 10-4 M-1 s-1, which is in agreement with
1.0 × 10-3 M-1 s-1 (80 °C) found by J ohnson et al.5 for
nafarelin. Nafarelin also contains a glycinamido group, which
is likely to undergo deamidation. Between pH 4 and 5.5 a
minimum plateau is observed in which the solvent-catalyzed
M2
M3
M4
M5
M0[H+] + M1 +
+
+
+
[H+] [H+]2 [H+]3 [H+]4
kobs
)
K1
K1K2 K1K2K3
1 +
+
+
[H+] [H+]2
[H+]3
reaction predominates. If the assumption is made that
H+
AH2
(1)
k
+K1 is relatively small, based on the fact that deamidated
gonadorelin is absent, kSAH can be calculated as 1.0 × 10-7
2+
K1, K2, and K3 represent the dissociation constants of histidine,
tyrosine, and arginine, respectively, and M0 to M5 represent
s-1. The other micro reacti3on constants cannot be determined
because M2 to M4 represent combinations of kinetically
1056 / Journal of Pharmaceutical Sciences
Vol. 85, No. 10, October 1996

+
+
Table 3
s
Macro Reaction Constants Extracted from the pH
−
log kobs
without interference of protolytic or prototropic equilibria still
leads to a slope in the pH-log kobs profile of 1.13
Ar r h en iu s P lotsThe influence of temperature on reaction
rate constants is given by the Arrhenius equation (eq 8),
Profile of Gonadorelin
Macro Reactn
Const
Macro Reactn
Const
Log Mx
Log Mx
Ea
M0
M1
M2
−
3.44
±
±
±
0.05
0.08
0.1
M3
M4
−
−
19.36
29.30
±
±
0.03
0.05
ln kobs ) ln A -
(8)
−7.00
12.2
RT
−
in which A represents the frequency factor, Ea is the energy
of activation, R is the gas constant, and T is the absolute
temperature. The data obtained were fitted with this equa-
tion, and the results are shown in Table 4.
The chromatograms in Figure 3 indicate that at pH 2 only
one relevant product is formed, probably through a proton-
catalyzed degradation reaction. The linearity of the curve and
the similarity of the chromatograms at 37 and 80 °C indicate
that the reaction mechanism is the same at both tempera-
tures. The parameters from Table 4 can therefore be ap-
pointed to a single reaction, most likely the hydrolysis of the
glycinamido group.
The Arrhenius fit at pH 8.4 concerns two degradation
products P1 and P2. From Figure 6 it can be seen that at
lower temperature P1 is relatively more present. In fact, after
30% degradation the ratio P1/P2 has nearly doubled at 37 °C
compared to 80 °C. This indicates that P1 and P2 are the
result of two parallel reactions with different energies of
activation. Indeed, it is not likely that the D-Ser4 epimer P2
and deamidated Gly10 gonadorelin P1 result from one reaction
with two products. Therefore, it can be concluded that the
energy of activation of the reaction resulting in P1 is lower
than that of the reaction resulting in P2. Two degradation
reactions with different energies of activation often do not
result in linear Arrhenius fits, but Table 4 shows that in this
case the line is reasonably straight (R ) -0.998), indicating
that the difference in the energies of activation of the two
reactions is relatively small.
Table 4
sInfluence of the Temperature on the Degradation Rate Constant
kobs of Gonadorelin
1
pH
Ea (J/mol)
Ea (cal/mol)
A (s-
)
R
2.1
5.1
8.4
7.6
8.8
1.2
×
×
×
104
104
105
1.8
2.1
2.9
×
×
×
104
104
104
1.2
2.7
1.4
×
×
×
106
106
1013
−0.998
−0.996
−0.998
indistinguishable reaction constants, predominantly hydroxyl-
catalyzed epimerization.7,11,12
Linear regression using eq 1 with fixed pKas at 5.8, 9.5,
and 1214 results in values for M0 to M4 similar to those in
Table 3. For M5 a value of (-1.8 ( 4.3) × 10-42 was calculated.
This indicates that M5 does not contribute within the pH range
studied.
The influence of pH on the degradation kinetics of gona-
dorelin has also been determined by Powell et al.4 However,
their data were fitted with a model which does not account
for protolytic equilibria. On the basis of this limited model,
it is concluded that several specific base-catalyzed reactions
are involved because the slope of the alkaline part of the pH-
k
obs profile is not +1. However, this relationship is not linear
as a result of the protolytic/prototropic equilibria and it is
therefore not correct to calculate a slope. Furthermore, a
combination of specific base-catalyzed reactions in a pH range
Figure 6
s
RP-HPLC chromatograms of partly (33%) degraded gonadorelin at (A) 37
g/mL.
°
C and (B) 80
°
C. Degradation conditions: pH 8.4, ionic strength 0.1, gonadorelin
concentration 20
µ
Journal of Pharmaceutical Sciences / 1057
Vol. 85, No. 10, October 1996

+
+
Table 5
s
Influence of Buffer Ions on the Degradation Rate Constant kobs of Gonadorelin
Acetate Buffer, pH 5.0 Phosphate Buffer, pH 6.2 Borate Buffer, pH 8.5
Concn (mM) Concn (mM)]
kobs (10-7 s- kobs (10-6 s-
Carbonate Buffer, pH 9.3
1
1
1
1
Concn (mM)
kobs (10-7 s-
)
)
)
Concn (mM)]
kobs (10-5 s-
)
25
50
75
1.16
1.06
1.03
1.15
±
±
±
±
0.05
10
25
55
88
2.7
2.4
2.3
2.8
±
±
±
±
0.1
0.1
0.1
0.1
25
50
75
100
125
150
175
200
7.9
8.4
9.2
±
±
±
±
±
±
±
±
0.3
0.4
0.4
0.4
0.5
0.4
0.4
0.4
20
25
40
50
4.9
5.3
4.8
5.0
4.9
±
±
±
±
±
0.2
0.2
0.2
0.2
0.2
0.04
0.04
0.05
100
9.2
11.4
10.0
8.8
62.5
9.4
Table 6sInfluence of the Ionic Strength on the Degradation Rate
At pH 5.3, the found energy of activation of 88 kJ /mol is
Constant kobs of Gonadorelin
roughly comparable with the energies found in the litera-
ture: 96 kJ /mol for gonadorelin,3 101 kJ /mol for gonadorelin,8
88 kJ /mol for nafarelin,5 92 kJ /mol for triptorelin,3 and 96 kJ /
mol for RS-26306.6 This indicates a similar solvent-catalyzed
degradation for the mentioned LH-RH analogues.
pH 2.0
Ionic
pH 9.0
Ionic
Strength
1
1
Strength
µ
(M)
kobs (10-6 s-
)
µ
(M)
kobs (10-5 s-
)
The temperature also influences the dissociation equilibria
of gonadorelin and therefore the reaction rate constant. The
energies of activation at pH 5.3 and 8.4 should be considered
as apparent energies of activation.
0.01
0.10
0.25
0.40
0.55
0.70
2.5
2.8
4.4
4.2
4.8
5.8
±
±
±
±
±
±
0.2
0.2
0.3
0.3
0.3
0.4
0.05
0.1
0.4
0.6
0.8
1.0
2.3
3.6
3.6
3.4
2.6
2.7
±
±
±
±
±
±
0.1
0.2
0.2
0.1
0.1
0.1
In flu en ce of Bu ffer Ion ssIn Table 5 some of the results
of the experiments concerning the influence of buffer ions on
the degradation rate are presented. All buffer ions are free
of catalytic influences. This is in disagreement with the
findings of Helm et al.,3 who found a catalytic effect of the
phosphate buffer at pH 5. In our study no influence of
phosphate buffer ions was found in the pH range 6-8, and
-
2-
therefore it is not likely that H2PO4 or HPO4 has any
relevant catalytic effect.
In flu en ce of Ion ic Str en gth sThe influence of ionic
strength on degradation rate constants of charged molecules
is given by eq 9,10
x
µ
log kobs ) log k0 + 1.02zazb
(9)
in which k0 is the apparent reaction rate constant at an ionic
strength of 0, za and zb are the charges of the reacting
molecules, and µ is the ionic strength. One of the two
reactants is gonadorelin, the other H+ or OH-. The overall
His
charge of gonadorelin is somewhere between +2 (pH < pKa
- 2) and -1 (pH > pKaArg + 2). In acidic solution the repulsive
electrostatic force between the reactants is reduced at high
µ, which results in a higher kobs
. In alkaline solution the
attractive electrostatic force is reduced at high µ, leading to
a lower kobs. If the transition state is considered, the effect of
the ionic strength on the transition state is expected to be
much stronger at pH 2 than at pH 9 because the developing
charge is +3 and +0.2, respectively.
From eq 9 and the charge of the reacting molecules one can
derive that the slope of the xµ-log k plot at pH 2.0 and pH
9.0 should be +2.04 and about -0.8, respectively, assuming
His
Tyr
Arg
that the pKa is 5.8, pKa
is 9.5, and pKa
is 12.
Figure 7
s
RP-HPLC chromatogram of the degradation of gonadorelin at µ ) 1,
pH 2, and 5
°
C.
The results from experiments with the ionic strength at pH
2 and 9 are depicted in Table 6. If data at pH 2 are fitted,
the resulting slope is +1.14 (R ) 0.96). Because of its small
deviation from the slope value for a singly charged molecule
(+1.02), the assumption is justified that the distance between
the reactive center at the carboxyl terminal end of the peptide
and the histidine residue is relatively large compared to the
distance between the carboxyl terminal end and the arginine
residue, resulting in a minimal influence of the charge of the
histidine.
However, it should be kept in mind that the ionic strength
also influences protolytic and prototropic equilibria. At pH
9, this might cover up the effect caused by changing activity
coefficients.
In Figure 7 it can be seen that at pH 2 another degradation
mechanism is promoted by high ionic strengths, resulting in
two products P5 and P6 with roughly equal peak areas.
Under the experimental conditions, the gonadorelin concen-
tration is not correlated with the degradation time. Fluores-
cence experiments indicate that P5 and P6 have a modified
If data at pH 9 are fitted, no correlation is observed (R )
-0.024, slope) -0.01), indicating no effect of µ on kobs
.
1058 / Journal of Pharmaceutical Sciences
Vol. 85, No. 10, October 1996

+
+
2. Fraser, H. M.; Baird, D. T. Bailliere’s Clinical Endocrinol. Metab.
1987, 1, 43-70.
3. Helm, V. J .; Mu¨ller, B. W. Pharm. Res. 1990, 7, 1253-1256.
4. Powell, M..; Sanders, L. M.; Rogerson, A.; Si, V. Pharm. Res.
1991, 8, 1258-1263.
tryptophan residue, since the fluorescence of this moiety has
disappeared. Further research is ongoing to elucidate the
chemical structures of these compounds.
5. J ohnson, D. M.; Pritchard, R. A.; Taylor, W. F.; Conley, D.;
Conclusions
Zuniga, G.; McGreevy, K. G. Int. J . Pharm. 1986, 31, 125-129.
The RP-HPLC system with acetonitrile and trifluoroacetic
acid is simple and effective. One of the main advantages of
this system is the possibility of coupling it to a mass
spectrometer. The RP-18 column gives better results in
separating the degradation products of gonadorelin than the
RP-8 column. Whether this is applicable to all peptides cannot
be concluded yet.
The degradation of gonadorelin can be divided into three
types of mechanisms: proton-, solvent-, and hydroxyl-
catalyzed mechanisms. At pH below 4, proton-catalyzed
deamidation dominates. In the pH range 4-6 solvent-
catalyzed hydrolysis of the peptide backbone at the N-terminal
site of the serine residue plays the most important role.
Hydroxyl-catalyzed epimerization and deamidation becomes
dominant at pH > 6.
6. Strickley, R. G.; Brandl, M.; Chan, K. W.; Straub, K.; Gu, L.
Pharm. Res. 1990, 7, 530-536.
7. Motto, M. G.; Hamburg, P. F.; Graden, D. A.; Shaw, C. J .; Cotter,
M. L. J . Pharm. Sci. 1991, 80, 419-423.
8. Oyler, A. R.; Naldi, R. E.; Lloyd, J . R.; Graden, D. A.; Shaw, C.
J .; Cotter, M. L. J . Pharm. Sci. 1991, 80, 271-275.
9. Okada, J .; Seo, T.; Kasahara, F.; Takeda, K.; Kondo, S. J . Pharm.
Sci. 1991, 80, 167-170.
10. Martin, A.; Swarbrick, J .; Cammarata, A. Physical Pharmacy;
Lea & Febiger: Philadelphia,1983; pp 352-398.
11. Manning, M. C.; Patel, K.; Borchardt, R. T. Pharm. Res. 1989,
6, 903-917.
12. Nishi, K.; Ito, H.; Shinagawa, S.; Hatanaka, C.; Fujino, M.;
Hattori, M. Pept. Chem. 1979, 175-180.
13. Van der Houwen, O. A. G. J .; Beijnen, J . H.; Bult, A.; Underberg,
W. J . M. Int. J . Pharm. 1988, 45, 181-188.
14. Handbook of Chemistry and Physics, 64th ed.; Weast, R. C., Ed.;
A new type of degradation was found in acidic solution at
high ionic strength and low temperatures (5 °C), which
probably involves changes in the tryptophan residue.
CRC Press Inc.: Boca Raton, FL, 1984; p C-719.
Acknowledgments
References and Notes
1. Schally, A. V.; Arimura, A.; Kastin, A. J .; Matsuo, H.; Baba, Y.;
Redding, T. W.; Nair, R. M. G.; Debeljuk, L. Science 1971, 173,
1036-1038.
The authors wish to thank Mr. Ed Hop for his valuable assistance
with the LC-MS experiments.
J S9601187
Journal of Pharmaceutical Sciences / 1059
Vol. 85, No. 10, October 1996