Degradation of azaglycinamido residues in model tripeptides derived from goserelin

Article abstract of DOI:10.1002/(SICI)1520-6017(200001)89:1<108::AID-JPS11>3.0.CO;2-A

Three model tripeptides, N-acetyl-Tyr-Pro-azaGly-NH₂ (NYPaG), Tyr-Pro- azaGly-NH₂ (YPaG), and Tyr-Pro-Gly-NH₂(YPG), were subjected to a systematic degradation study to get information about the degradation of the azaglycinamido residue The degradation products were characterized with LC- MS. Main degradation products of NYPaG possess partially or totally eliminated azaglycinamido residues, while YPaG and YPG are exhibit cyclo(Tyr- Pro) formation, a diketopiperazine. The influence of the pH on the degradation rate constant k(obs) was investigated for NYPaG and YPaG in the pH range 0.4-11. An U-shaped profile with an inflexion around pH 9 was found for NYPaG while the degradation rate of YPaG was independent of the pH. NYPaG apparently was subject to proton-, solvent-, and hydroxyl-catalyzed degradation reactions whereas YPaG only underwent solvent-catalyzed reactions. Some influence of acetate and phosphate ions on k(obs) was found for YPaG. Arrhenius plots of NYPaG and YPaG were found to be linear. (C) 2000 Wiley-Liss, Inc.

Full text of DOI:10.1002/(SICI)1520-6017(200001)89:1<108::AID-JPS11>3.0.CO;2-A

Degradation of Azaglycinamido Residues in Model
Tripeptides Derived from Goserelin
MARNIX A. HOITINK, JOS H. BEIJNEN, AUKE BULT, J. MIRJAM A. DAMEN, OEDS A. G. J. VAN DER HOUWEN,
JOHN A. W. KRUIJTZER, MATTHIJS M. TIBBEN, GERARD WIESE, WILLY J. M. UNDERBERG
Department of Pharmaceutical Analysis, Faculty of Pharmacy, Utrecht University, Sorbonnelaan 16,
NL-3584 CA Utrecht, The Netherlands
Received 6 July 1999; accepted 26 October 1999
ABSTRACT: Three model tripeptides, N-acetyl-Tyr-Pro-azaGly-NH2 (NYPaG), Tyr-Pro-
azaGly-NH2 (YPaG), and Tyr-Pro-Gly-NH2(YPG), were subjected to a systematic deg-
radation study to get information about the degradation of the azaglycinamido residue.
The degradation products were characterized with LC-MS. Main degradation products
of NYPaG possess partially or totally eliminated azaglycinamido residues, while YPaG
and YPG are exhibit cyclo(Tyr-Pro) formation, a diketopiperazine. The influence of the
pH on the degradation rate constant kobs was investigated for NYPaG and YPaG in the
pH range 0.4–11. An U-shaped profile with an inflexion around pH 9 was found for
NYPaG while the degradation rate of YPaG was independent of the pH. NYPaG ap-
parently was subject to proton-, solvent-, and hydroxyl-catalyzed degradation reactions
whereas YPaG only underwent solvent-catalyzed reactions. Some influence of acetate
and phosphate ions on kobs was found for YPaG. Arrhenius plots of NYPaG and YPaG
were found to be linear. © 2000 Wiley-Liss, Inc. and the American Pharmaceutical Association
J Pharm Sci 89: 108–114, 2000
However, besides pharmacokinetics, the phar-
maceutical characteristics are also changed. In
studies on the stability of goserelin,2,3 degrada-
tion of the azaglycinamido moiety was observed
at pH > 4. Not all degradation products could be
identified because in goserelin more degradation
pathways occur and the yield of individual prod-
ucts is limited. In order to get more insight into
both qualitative and quantitative degradation
processes of the the azaglycinamido moiety, small
model peptides were synthesized containing this
functional group and these peptides were sub-
jected to a systematic stability study.
INTRODUCTION
Peptides are gaining importance as medicinal
compounds. A traditional disadvantage for thera-
peutical purposes is that these peptides are sen-
sitive to relatively fast hydrolysis by ubiqutuous
human enzymes. To overcome this disadvantage,
peptidomimetics can be developed in which some
or all amino acid residues are replaced by unnatu-
ral derivatives that are less susceptible to enzy-
matic hydrolysis. A good example of a medicinal
peptidomimetic compound is the gonadorelin ana-
logue goserelin, in which the C-terminally located
glycinamido moiety in gonadorelin has been re-
placed by an azaglycinamido moiety. In this way,
the plasma half-life in humans is increased from
15 min to more than 4 h.1
EXPERIMENTAL SECTION
Chemicals
All chemicals used were of analytical grade.
N-Acetyl-Tyrosine-Proline-aza-Glycine-amide
(NYPaG) and Tyrosine-Proline-aza-Glycine-
amide (YPaG) were synthesized. Every coupling
Correspondence to: M. A. Hoitink (E-mail: M.A.Hoitink@
pharm.uu.nl)
Journal of Pharmaceutical Sciences, Vol. 89, 108–114 (2000)
© 2000 Wiley-Liss, Inc. and the American Pharmaceutical Association
108
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 1, JANUARY 2000

DEGRADATION OF AZAGLYCINAMIDO RESIDUES
109
step was followed by an extraction procedure: the
compound was brought into ethylacetate and sub-
sequently extracted with 1 M KHSO4, 1 M
NaHCO3, and a saturated NaCl solution. The eth-
ylacetate solution was dried with anhydrous
Na2SO4 and filtered, and the solvent was re-
moved under vacuum.
YPaG was synthesized by coupling tertiary
butyloxycarbonyl(boc)-tyrosine(benzyl) with pro-
line methyl ester using benzotriazolyl-N-oxy-
tris(dimethylamino)phosphonium hexafluoro-
phosphate (BOP) and diisopropylethylamine
(DIPEA) in dichloromethane (DCM). The meth-
ylester group was hydrolyzed with NaOH in
methanol. This dipeptide was then coupled to
semicarbazide with BOP/DIPEA in DCM. The
benzyl moiety was removed by hydrogenation
with palladium on carbon in methanol. The boc
moiety was eliminated with HCl in ether.
NYPaG was synthesized similarly to YPaG ex-
cept that the boc-tyrosine(benzyl) was first
treated with HCl in ether and then N-acetylated
with acetic anhydride and triethylamine in diox-
ane.
YPaG was purified with a gradient RP-HPLC
procedure using an 250 × 20 mm i.d. econosphere
C8 column (Alltech, Breda, The Netherlands) at a
flow of 11 mL/min. Mobile phases A and B con-
sisted of 1 mL trifluoro acetic acid (TFA)/1,000
mL water and 1 mL TFA/500 mL water/500 mL
methanol, respectively. The gradient was run
from 0 to 10% B in 30 min.
NYPaG was purified with an isocratic RP-
HPLC procedure using an 250 × 10 mm i.d.
Chromspher C18 column (Chrompack, Bergen op
Zoom, The Netherlands) at a flow of 4 mL/min.
The mobile phase consisted of 1.42 mL acetic acid/
0.936 mL 25% w/w ammonia/800 mL water/200
mL methanol. After purification, the methanol
was removed under a nitrogen stream and the
remaining solutions were freeze-dried. Yields
were in the order of magnitude of 20%.
The Netherlands) equipped with a Slim-trode
electrode (Hamilton, Darmstadt, Germany). The
pH was measured before and after degradation.
All pH values were determined at the appropriate
degradation temperature.
Liquid Chromatography-Mass
Spectrometry (LC-MS)
The liquid chromatograph consisted of a LC-10
AD liquid chromatograph (Shimadzu, Tokyo, Ja-
pan) provided with a 7125 injector (Rheodyne, Co-
tati, CA). The column used was a Superspher 100
RP-18 (5 ␮m) 119 × 2 mm i.d. (Merck). As mobile
phase 0.1% v/w TFA in mixtures of methanol and
water was used. For YPaG-containing samples
the mixture consisted of 12% w/w methanol/88%
w/w water, while 16% w/w methanol/84% w/w wa-
ter was used for all other compounds. The flow
was set at 0.1 mL/min.
This LC system was directly coupled to a VG
Platform II mass spectrometer (MicroMass, Al-
trincham, U.K.), operated in the positive ion
mode. Calibration was performed with an aque-
ous 4 mg/mL sodium iodide solution. The scanned
mass range is 100–500 m/z, scan time is set at 1 s,
and low and high mass resolutions are 12.5 and
15 (instrumental units), respectively. Cone volt-
age was 25 V, and the probe temperature was
kept at 120°C. Data acquisition and calculation of
the m/z values was done with MassLynx version
2.3 (MicroMass).
RP-HPLC Systems for NYPaG, YPaG, and YPG
A Gynkotech Model 300 precision pump (Separa-
tions, H.I. Ambacht, The Netherlands) was con-
nected to a ISS 100 sampling system (Perkin-
Elmer Corporation, Norwalk, CT). Mixtures of
acetonitrile or methanol, water, and TFA were
used as mobile phases. For the analysis of
NYPaG-, YPaG-, and YPG-containing samples,
the composition of the mobile phase was 0.1% v/w
TFA in 12% methanol/88% water w/w, 4% metha-
nol/96% water w/w, and 6% acetonitrile/94% wa-
ter, respectively. The column was a Lichrospher
100 RP-18 (5 ␮m) 125 × 4 mm i.d. (Merck, Darm-
stadt, Germany). Detection was performed using
an Applied Biosystems 785A programmable ab-
sorbance detector (Separations, H.I. Ambacht,
The Netherlands) operating at 214 nm, and a 40
␮L sample was injected.
YPG (Tyrosine-Proline-Glycine-amide) was
synthesized at the Netherlands Cancer Institute
(Amsterdam, The Netherlands) utilizing a pep-
tide synthesizer.
The identities of all three peptides were con-
1
firmed with MS and H-NMR, and the purities
were verified with RP-HPLC and found to be
more than 95%.
pH Measurement
All pH measurements were performed on a Con-
sort P514 pH meter (Salm & Kipp, Breukelen,
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 1, JANUARY 2000

110
HOITINK ET AL.
Preparation of Degraded Samples for LC-MS
Influence of the Temperature on kobs of NYPaG
and YPaG
For LC-MS analysis the initial tripeptide concen-
tration was 500 ␮g/ml. For NYPaG, samples were
prepared at pH 2, 5, 9, and 10, for YPaG at pH 2,
5, 7, 9, and 10, and for YPG at pH 9 and 10. A
carbonate buffer was used for pH 10 instead of the
standard buffer. NYPaG pH 5 was degraded at
80°C, all other samples at 70°C. The typical deg-
radation time was two estimated half-lives.
The influence of the temperature on the degrada-
tion of NYPaG and YPaG was determined at pH
1, 5, and 8.7, and at pH 1.5, 5, and 8.6, respec-
tively. The temperature range was 70–90°C for
low and middle pH values, and 50–80°C for pH
8.6–8.7.
RESULTS AND DISCUSSION
Standard Degradation Conditions
The selected model compounds are NYPaG,
YPaG, and YPG, all tripeptides (Figure 1). Since
Degradation experiments were executed in flame-
sealed 1 mL glass ampoules, that were kept in a
thermostated water bath at the appropriate tem-
perature. Sampling of the ampoules with NYPag,
YPaG, and YPG was done over a range of at least
twice the estimated half-life. For each degrada-
tion curve, 10 measurement points were used and
all experiments were performed in duplicate.
Standard degradation conditions were 70°C, a tri-
peptide concentration of 20 ␮g/mL, a buffer con-
centration of 25 mM, and an ionic strength of 0.1,
adjusted with sodium perchlorate. Buffers used
were perchloric acid below pH 2.5, acetate in the
pH range 4–6, phosphate in the pH range 6–7.5,
borate in the pH range 8–9.5, triethylamine in the
pH range 9.5–10, and sodium hydroxide above pH
10. All experiments were done under standard
degradations conditions, unless otherwise stated.
Calibration Curves and Standard Deviation
Calibration curves of NYPaG, YPaG, and YPG
were constructed in the concentration range 5–40
␮g/mL.
Standard deviations in kobs for NYPaG and
YPaG were determined at pH 8.4 and 8.6, respec-
tively. For each determination 6 degradation
curves were used.
Influences of Buffers and pH on the Degradation
Rate Constant kobs of NYPaG, YPaG, and YPG
The influence of acetate and phosphate buffer
ions on kobs of YPaG was determined at pH 4.9
and pH 6.9, respectively. The buffer concentration
range was 25–100 mM and 12.5–50 mM for ac-
etate and phospate, respectively.
The influence of the pH for NYPaG and YPaG
was determined in the pH range 1–11, and that
for YPG in the pH range 8–11.
Figure 1. Chemical structures of the tripeptides
NYPaG, YPaG, and YPG.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 1, JANUARY 2000

DEGRADATION OF AZAGLYCINAMIDO RESIDUES
111
azaglycinamido-containing peptides cannot be
synthesized with commercially available reagents
using a peptide synthesizer, an in-solution syn-
thesis method was developed for NYPaG and
YPaG. The N-terminal residue is tyrosine that
gives the necessary UV absorbance for RP-HPLC
detection and is sufficiently stable in solution.
The second and the third residues in NYPaG and
YPaG are proline and azaglycinamide, respec-
tively, being identical with the two C-terminally
located residues in goserelin. This is likely to en-
sure similar degradation behavior of the model
peptides compared to goserelin. A free N-termi-
nus has both advantages and disadvantages. The
advantages are satisfactory solubility in water
and high sensitivity in an LC-MS system. How-
ever, an important disadvantage is the reactivity,
which might cause interfering degradation reac-
tions. This is why the N-acetylated tripeptide
NYPaG is included in this study. The solubility
and LC-MS sensitivity of NYPaG were still found
to be sufficient for our purposes. YPG is used to
determine whether any unwanted reactions in
YPaG are exclusively the result of the free N-
terminus or if also the azaglycinamidogroup is in-
volved.
was used, operating at 0.1 mL/min. The 0.1 mL/
min represents the best compromise between
analysis time and ion spray mass spectrometer
performance on a 2 mm column. In Table I, the
LC-MS results are presented as m/z values with
their intensities, related to the intensity of the
strongest peak. From these relative intensities,
the amount of product can be estimated under the
condition that the number of basic groups is un-
changed. In all NYPaG samples, degradation
products with m/z values of 321 and 335 are
found, products B and A in Figure 2, respectively,
which are similar to goserelin.2 Since these prod-
ucts are found at all pH values, the mechanism is
probably solvent-catalyzed. Only at pH 2, are m/z
values of 261 and 336 found, and these are likely
cyclo(Tyr-Pro) (Figure 2C) and YPaG, respec-
tively. Apparently, the N-acetyl group undergoes
proton-catalyzed hydrolysis, and the resulting tri-
peptide YPaG undergoes cyclization as the result
of a nucleophilic attack by the free N-terminus on
the carbonyl group of proline. Products with m/z
values of 305 and 360 are found at pH 2, 9, and
10. The identities of these products are unknown,
but the mass difference of m/z 360 with the parent
is identical to that of an unknown product of gos-
erelin. Speculative structures are presented in
Figure 2D and E for m/z values 360 and 305, re-
spectively. In conclusion, the degradation of
NYPaG is mainly located at the azaglycinamido
and this makes NYPaG a good model compound.
The main degradation product of YPaG has a
m/z value of 261, probably cyclo(Tyr-Pro) depicted
in Figure 2C. The free N-terminus is now respon-
sible for a degradation reaction not found for gos-
erelin, and this obscures the observation of the
Identification of the Degradation Products of
NYPaG, YPaG, and YPG
Degradation samples of the tripeptides were pre-
pared at varying pH values and with different
buffers. The samples were analyzed with a
straightforward LC-MS method using TFA as
acidifying and ion-pairing reagent. Instead of a
sensitivity-reducing flow splitter, a 2 mm column
Table I. m/z Values and Relative Intensities of Compounds Resulting From the
Degradation of NYPaG, YPaG, and YPG*
Peptide/pH
Found m/z values (relative intensity)
NYPaG pH 2
MYPaG pH 5
NYPaG pH 9
NYPaG pH 10
YPaG pH 2
YPaG pH 5
YPaG pH 7
YPaG pH 9
YPaG pH 10
YPG pH 9
261(16%),305(3%),321(4%),335(8%),335(5%),336(30%),360(7%),378(100%)
321(42%),335(10%),335(77%),378(100%)
305(7%),321(13%),335(13%),360(14%),378(100%)
305(6%),321(7%),335(9%),360(13%),378(100%)
261(67%),336(100%)
261(33%),318(5%),336(100%)
261(100%),318(36%),336(6%),336(88%)
261(35%),261(22%),318(74%),336(3%),336(100%)
261(34%),261(16%),318(20%),336(<5%),336(100%)
261(25%),335(20%),335(100%),336(35%)
261(2%),335(7%),335(100%),336(24%)
YPG pH 10
* The parent compounds are written in bold.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 1, JANUARY 2000

112
HOITINK ET AL.
values. The free C-terminal -CO-NH-NH2 possi-
bly promotes a solvent-catalyzed epimerization
reaction of the proline residue.
RP-HPLC systems with UV detection, based on
the LC-MS methods described earlier, were devel-
oped for all three model peptides for the determi-
nation of degradation rate constants. Calibration
curves are linear with correlation coefficients of
0.998 and higher. The stability-indicating proper-
ties of the RP-HPLC systems can be derived from
the LC-MS data, since co-eluting products can be
distinguished by their m/z ratios. No interfering
products were found.
All three peptides were found to degrade ac-
cording to (pseudo)first-order kinetics. The ob-
served degradation rate constant kobs was cal-
culated from the slopes of plots of the natural
logarithm of the residual parent compound con-
centration versus degradation time. The relative
standard deviations in kobs of NYPaG and YPaG
were found to be 16% and 9%, respectively. These
values are in the expected order of magnitude.
Figure 2. Degradation products of NYPaG (R1 ס N-
acetyltyrosine) and YPaG (R1 ס tyrosine).
azaglycinamido degradation. One product involv-
ing the azaglycinamido group at m/z 318 can still
be found. It is similar to m/z 360 in degraded
NYPaG samples. A structure as in Figure 2D can
therefore be postulated. The cyclization reaction
of YPaG might be brought about by the azagly-
cinamido group, because it is the leaving group. To
assess the influence of the azaglycinamido moiety
on the mechanism and kinetics of the cyclization,
a third tripeptide was included in the study, YPG.
The difference between YPG and YPaG is substi-
tution of the glycine residue in YPG for the aza-
glycinamido residue in YPaG. If YPG undergoes a
similar cyclization reaction as YPaG, the free N-
terminus is the driving force of this reaction. If
YPG does not undergo this cyclization, the azagly-
cinamido group in YPaG is likely to be involved.
In the degraded YPG samples, cyclic products
with m/z 261 are found. The cyclization reaction is
indeed found, and yields the same product as with
YPaG. The free N-terminus is the driving force in
the cyclization reaction, so in small model pep-
tides the N-terminus can best be blocked through,
for example, N-acetylation. Parallel to cyclization
deamidation of YPG occurs yielding a product
with m/z 336. The conclusion can be drawn that in
the development of model peptides, a C-terminal
primary amido group might easily be a disturbing
factor because of its sensitivity for deamidation.
In contrast with the primary amido groups, sec-
ondary amido groups are not sensitive to hydro-
lysis, as demonstrated for buserelin,3 and are a
better alternative. Epimers of some compounds
were found especially at pH values above 5 with
one exception: the product of NYPaG with a m/z
value of 335 was also found to epimerize at all pH
Influence of the pH on the Degradation Rate
Constant kobs of NYPaG
The pH–log(kobs) data of NYPaG are depicted in
Figure 3. Data were fitted with a model developed
Figure 3. pH–log(kobs) profile of NYPaG at 70°C and
an ionic strength of 0.1.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 1, JANUARY 2000

DEGRADATION OF AZAGLYCINAMIDO RESIDUES
113
Table II. Macro Constants of the pH–log(kobs) Fit of NYPaG Using One- and Two-pKa Models
Ionizable groups
in model
pKa1 (aza-
glycinamide) (tyrosine)
pKa2
log(M0)
log(M1)
log(M2)
log(M3)
log(M4)
2
1
1
−4.9 0.1 −6.69 0.04 −14.0 0.1 −23.4 0.3 −34.0 0.2
8.5 0.2
8.5 0.1
—
9.50a
—
−4.9 0.1 −6.69 0.05 −14.0 0.1 −24.5 0.2
−4.9 0.1 −6.64 0.08 −14.6 0.1 −26.0 0.4
—
—
9.50a
a Fixed.
by Van der Houwen et al.4 which accounts for the
influence of the ionizable groups on kobs. The pKa
value of the side chain of tyrosine in peptides was
determined at 70°C in an earlier study and was
found to be 9.5.5 The macro constants of these fits
are presented in Table II. The resulting fit of an
one pKa model with a value fixed at 9.5 is not
acceptable, the line does not include most points.
If the fit is performed with a second variable pKa
value using eq 1, a reasonable profile is obtained
with a calculated pKa of 8.5.
catalyzed mechanisms. The main degradation
product of YPaG is cyclo(Tyr-Pro), as described
above, and, apparently, this cyclization reaction
is solvent-catalyzed. The variations in the
log(kobs) values are probably induced by the ion-
izable groups in YPaG, the free N-terminus and
the phenolic moiety in tyrosine, and buffer ion
catalysis. The influences of acetate and phosphate
ions on kobs was investigated (Table III). Indeed,
buffer ion influences were found. Since the same
products are found using different buffer ions,
these buffers apparently catalyze the already rel-
evant reaction. Data used in the pH-log(kobs) pro-
file are corrected for this buffer influence. The
pH–log(kobs) profile is fitted as a horizontal line
with a kobs value of 1.52 × 10−5 s−1. The fifth data-
point, which results from extrapolation of several
phosphate buffer concentration to zero buffer con-
M2
H+
M3
M4
M H+ + M +
+
+
͓
0
͔
1
H+ 2 H+ 3
͓
͔
͓
͔
͓
͔
(1)
kobs
=
K1 K1K2
1 +
+
H+
H+ 2
͓
͔
͓
͔
in which M0–M4 represent the macroconstants
and K1 and K2 are the dissociation constants of
the glycinamido and tyrosine residue, respec-
tively. Apparently, a second ionizable group is in-
volved with a pKa of 8.5, presumably located in
the azaglycinamido moiety since the sequence N-
acetyl-Tyrosine-Proline only has one prototropic
function, the phenol moiety in tyrosine. In the
compound semicarbazide, structurally identical
to the azaglycinamido group, an acid group with a
pKa value of 10.8 occurs.6 Because in NYPaG the
extra electron-withdrawing CO group of proline is
present and the degradation temperature is 50°C
higher, the pKa of the azaglycinamido group
might well have a lower value. The fit with a two
pKa model with one pKa fixed at 9.5 results in the
most appropriate fit (Figure 3).
Influence of the pH on the Degradation Rate
Constant kobs of YPaG and YPG
The pH–log(kobs) profile (Figure 4) of YPaG differs
from NYPaG in two ways: (1) The average
log(kobs) values are higher and (2) there is no U-
shape indicating, neither proton- nor hydroxyl-
Figure 4. pH–log(kobs) profiles of (᭡) YPaG and (᭿)
YPG at 70°C a and an ionic strength of 0.1.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 1, JANUARY 2000

114
HOITINK ET AL.
Table III. Influence of Acetate and Phosphate Buffer Ions on kobs of YPaG
Correlation
Buffer concentration
range [mM]
coefficient
Buffer
pH (70°C)
Slope
Intercept
(R2)
Acetate
Phosphate
4.8
6.8
25–100
13–50
1.6 × 10−4 1.7 × 10−5
4.8 × 10−4 4.5 × 10−5
0.98
0.90
centration, is omitted from the calculation of this
horizontal line since there is a large deviation
from most data points (Figure 4). The correction
for the phosphate buffer influence is possibly less
precise. If the limited data of YPG are compared
with the YPaG results, the main feature of the
degradation of YPG is hydroxyl catalysis (Figure
4). YPG, like YPaG, is subject to cyclization but in
contrast with YPaG, YPG also undergoes deami-
dation. In earlier studies the hydroxyl-catalysis of
the deamidation process was established.3 An-
other difference in the degradation of YPG com-
pared to YPaG is the lower kobs between pH 8 and
10. The cyclization reaction of YPG seems to be
slower compared to YPaG and it looks like the aza-
glycinamido residue of YPaG increases the cycliza-
tion rate when compared to glycinamido moiety.
explanations are the influence of the deacetyla-
tion reaction at pH 2 and ionization effects at
pH 8.7.
CONCLUSIONS
NYPaG is a good model peptide for studying the
degradation of the azaglycinamido residue pre-
sent, for example, in goserelin. YPaG is not a good
model, since the the free N-terminus induces a
relatively fast cyclization reaction. Model pep-
tides for the investigation of degradation mecha-
nisms other than cyclization and deamidation
preferably should not have a free N-terminus or
C-terminal primary amido group.
The azaglycinamido group can give rise to the
following degradation reactions: elimination of
CHNO yielding a C-terminal hydrazide deriva-
tive, elimination of the azaglycinamido group,
and H2O elimination. These degradation reac-
tions are specific for the azaglycinamido residue.
Influence of the Temperature on kobs of NYPaG
and YPaG
The Arrhenius equation parameters are pre-
sented in Table IV. The variations in the energies
of activation (Ea) of the degradation of NYPaG are
to be expected since both proton-, solvent-, and
hydroxyl-catalyzed mechanisms are involved and
are not likely to have equal Ea values. However,
for YPaG similar Ea values are likely since there
is only one main solvent-catalyzed degradation
mechanism over the total pH range studied. How-
ever, similar Ea values are not found. Possible
REFERENCES AND NOTES
1. Dutta AS, Furr BJA, Hutchinson FG. 1993. Zola-
dex: Discovery, pharmacodynamics and formula-
tion. Drug News Perspect 6:325–332
2. Hoitink MA, Hop E, Beijnen JH, Bult A, Kettenes-
van den Bosch JJ, Underberg WJM. 1997. Capil-
lary zone electrophoresis as a tool in the stability
research of the luteinising hormone-releasing hor-
mone analogue goserelin. J Chromatogr A 776:
319–327.
3. Hoitink MA, Beijnen JH, Boschma MUS, Bult A,
Hop E, Nijholt J, Versluis C, Wiese G, Underberg
WJM. 1997. Identification of the degradation prod-
ucts of gonadorelin and three analogues in aqueous
solution. Anal Chem 69:4972–4978.
4. Van der Houwen OAGJ, Beijnen JH, Bult A, Un-
derberg WJM. 1988. A general approach to the in-
terpretation of pH degradation profiles. Int J
Pharm 45:181–188.
5. Hoitink MA, Beijnen JH, Bult A, Van der Houwen
OAGJ, Nijholt J, Underberg WJM. 1996. Degrada-
tion kinetics of gonadorelin in aqueous solution. J
Pharm Sci 85:1053–1059
6. 1976. Merck Index, 9th ed. Rahway: Merck & Co. p
1093.
Table IV. Arrhenius Plot Parameters of NYPaG
and YPaG
Energy of
activation
Correlation
coefficient
(R2)
Peptide/pH
[×104 J/mol]
ln(A)
NYPaG pH 1.0
NYPaG pH 5.0
NYPaG pH 8.7
YPaG pH 1.5
YPaG pH 5.0
YPaG pH 8.6
8.1
10.9
9.0
10.3
9.7
15.3
23.1
18.9
25.0
23.1
19.7
0.998
0.994
0.996
1
0.9997
0.9998
8.8
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 1, JANUARY 2000