Kinetics of diketopiperazine formation using model peptides

Article abstract of DOI:10.1021/js970325m

The intramolecular aminolysis of Phe-Pro-p-nitroaniline (Phe-Pro-pNA) to Phe-Pro-diketopiperazine (Phe-Pro-DKP) was studied as a function of pH, temperature, buffer concentration, and buffer species using an HPLC assay that permits simultaneous analysis o

Full text of DOI:10.1021/js970325m

Kinetics of Diketopiperazine Formation Using Model Peptides
CHIMANLALL GOOLCHARRAN AND RONALD T. BORCHARDT*
Contribution from The Department of Pharmaceutical Chemistry, The University of Kansas, Lawrence, Kansas 66047
Received August 15, 1997.
Final revised manuscript received December 15, 1997.
Accepted for publication December 16, 1997.
has been shown to occur at a faster rate than in dipeptides
that contain other C-terminal amino acids.9 In fact, the
composition and configuration of the peptide plays a major
role in determining the rate of the intramolecular ami-
nolysis reaction.11 During peptide synthesis, this intramo-
lecular aminolysis reaction occurs readily and creates
problems in both solution and solid-phase synthesis
techniques.11-16 The cyclization was shown to be catalyzed
by acids17 and bases.3 In addition to the pH of the solution,
the nature of the buffer species may also influence the rate
of DKP formation.18
Although the formation of diketopiperazine has been
shown to be a prevalent side reaction in peptide synthesis,
it has not been studied in any detail as a degradation
pathway of pharmaceutically important peptides and pro-
teins in solution. Recently, several workers observed DKP
formation upon storage of their protein pharmaceuti-
cals.1,6,18,19 This observation led to a renewed interest in
the intramolecular aminolysis reaction as a significant
degradation pathway for peptides and proteins that contain
a proline as the penultimate residue.
In this study, the kinetics of DKP formation from the
dipeptide analogue Phe-Pro-p-nitroaniline (Phe-Pro-pNA)
was investigated in aqueous solution at 37 °C as a function
of pH, temperature, buffer species, and concentration. The
p-nitroaniline amides were selected for these studies
because they undergo DKP formation more rapidly than
amino acid amides (i.e., tripeptides), thus making the
kinetic studies more convenient. In addition, the influence
of primary sequence on the kinetics was addressed. For
the latter study, we synthesized analogues in which the
amino acid preceding the proline residue was varied.
Abstract 0 The intramolecular aminolysis of Phe-Pro-p-nitroaniline
(Phe-Pro-pNA) to Phe-Pro-diketopiperazine (Phe-Pro-DKP) was studied
as a function of pH, temperature, buffer concentration, and buffer
species using an HPLC assay that permits simultaneous analysis of
the disappearance of the starting material and the appearance of
degradation products. The degradation followed pseudo-first-order
kinetics and showed significant dependence on pH. Phosphate (pH
5−8) and glycine (pH 9−10) buffers exhibit general base catalysis.
The pH−rate profile suggested that the rate of Phe-Pro-DKP formation
depends on the degree of ionization of the N-terminal amino group,
with the unprotonated reactant being more reactive than the protonated
form. The pK value of 6.1, determined kinetically, and three
a
microscopic rate constants were adequate to describe the shape of
the pH−rate profile. In the pH range studied, Phe-Pro-DKP was the
only product generated upon degradation of Phe-Pro-pNA. At pH
values between 3 and 8, Phe-Pro-DKP was stable, while at pH less
than 3 and greater than 8 it undergoes hydrolysis to the dipeptide,
Phe-Pro-OH. Sequence inversion, a reaction normally associated with
DKP formation, was not observed. The influence of primary sequence
on the formation of DKP was also investigated using X-Pro-pNA
analogues, where X ) Gly, Ala, Val, Phe, â-cyclohexylalanine, and
Arg. Changing the amino acid preceding the proline residue had a
significant effect on the rate of DKP formation at pH 7.0.
Introduction
Peptides and proteins that possess the N-terminal
sequence in which proline is the penultimate residue
undergo nonenzymatic aminolysis, yielding a diketopip-
erazine (DKP), which arises from the first two amino acids,
and a truncated sequence.1-3 The mechanism of DKP
formation involves nucleophilic attack of the N-terminal
nitrogen on the amide carbonyl between the second and
third amino acids. This intramolecular aminolysis reac-
tion occurs readily in aqueous solution;4-6 it is known to
play an important role in the biosynthetic pathway of
biologically active cyclic dipeptides such as cyclo(His-Pro),
which are found throughout the central nervous system,
peripheral tissues, and body fluids.7 These cyclic dipep-
tides play important roles in neurophysiological function-
ing.8 Diketopiperazines also have implications in the
diagnosis of proteins in fossils and thermal processing of
food.9
It is well-documented in the literature that dipeptide
esters and amides readily cyclize to the DKP and that the
rate of cyclization competes favorably with hydrolysis.4,9,10
Sequence inversion and racemization have also been as-
sociated with diketopiperazine formation and greatly com-
plicate the kinetics of such reactions.10 The cyclization of
dipeptide esters containing a C-terminal proline residue
Materials and Methods
Ma ter ia lssAll t-Boc amino acid derivatives and proline-p-
nitroaniline (Pro-pNA) were purchased from Bachem, Bioscience
Inc., Philadelphia, PA. HPLC grade acetonitrile and trifluoroacetic
acid were obtained from Fisher Scientific (Fairlawn, NJ ). 1-[3-
(Dimethylamino)propyl]-3-ethylcarbodiimide (EDC), 1-hydroxyl-
benzotriazide (HOBT), sodium phosphate monobasic, sodium
phosphate dibasic, and sodium chloride were purchased from
Sigma Chemical Co. (St. Louis, MO). All other chemicals were of
analytical grade and used as received. The water used in all
studies was from a Millipore MILLI-Q water system. The dipep-
tide analogues and their degradation products (Scheme 1) were
manually synthesized using a solution phase peptide synthesis
technique11 utilizing EDC as the activating reagent. The peptide
syntheses were monitored by TLC using Analtech, Inc., silica gel
plates, which were developed using a mixture of hexane (60%),
ethyl acetate (35%), and glacial acetic acid (5%). Plates were
visualized with iodine vapor. Purification was performed using
HPLC on a preparative Rainin Dynamax C18 column (10 µm, 22.5
× 250 mm).
Kin etic Mea su r em en tssThe following buffers at concentra-
tions of 200, 100, and 50 mM were used for the kinetic experi-
ments: pH 1.0, HCl; pH 2.0-3.0, glycine; pH 4.0-5.0, acetate;
pH 5.5-8.0, phosphate; and pH 9.5-10.0, glycine. A constant ionic
strength of 1.0 M was maintained for each buffer by adding the
* Tel: (785) 864-4820. Fax: (785) 864-5736. E-mail: Borchardt@
smissman.hbc.ukans.edu.
© 1998, American Chemical Society and
American Pharmaceutical Association
S0022-3549(97)00325-0 CCC: $15.00
Published on Web 01/30/1998
Journal of Pharmaceutical Sciences / 283
Vol. 87, No. 3, March 1998

Figure 1sTime course for the disappearance of Phe-Pro-pNA (0) and the
appearance of Phe-Pro-DKP (O) and p-nitroaniline (]) at pH 5.0, 0.1 M acetate
buffer (I ) 1.0) at 37 °C.
Scheme 1
appropriate amount of sodium chloride. The buffers were pre-
pared at the experimental temperature and the pH value for a
given solution remained unchanged throughout the investiga-
tion. An Orion (420A) pH meter equipped with an Orion auto-
matic temperature compensation (ATC) electrode was used to
measure the pH ((0.05) of the buffer solutions. All kinetic
experiments were carried out in aqueous buffer solutions at
37 ( 1 °C unless otherwise indicated. The purified peptide
analogues were dissolved in the appropriate buffer solution to yield
an initial concentration of about 1 mM. Aliquots of 200 µL of the
resulting solution were transferred to 250 µL HPLC vials, which
were then sealed and stored in a 37 °C incubator. At various
times, a vial was removed, cooled to room temperature, and
analyzed by HPLC. All reactions were carried out for three or
more half-lives.
HP LC An a lysissThe HPLC system consisted of Rainin Dy-
namax SD-200 pumps, a Dynamax automatic sample injector (Al-
1A), and a Dynamax absorbance detector (UV-1), which were
controlled by PC Acquisition HPLC-PDA (version 1.8) software.
Analysis of the initial peptide and its degradation products was
performed on a Rainin Dynamax 300A C18 column (5 µm, 4.6 ×
250 mm) at room temperature. The solvent systems consisted of
Milli-Q water:acetonitrile:trifluoroacetic acid in the ratios of 95:
5:0.1 (solvent A) and 10:90:0.1 (solvent B). Elution was ac-
complished utilizing an isocratic method of 15% solvent B for 10
min followed by a linear gradient to 80% solvent B in 15 min. The
flow rate was maintained at 1.0 mL/min and the detection
wavelength was 214 nm. The degradation products were identified
by comparing their retention times with those of authentic
samples.
Figure 2sKinetic profile for the degradation of Phe-Pro-pNA (0) at pH 9.0,
I ) 1.0 and 37 °C. On the ordinate: O, Phe-Pro-DKP; ], p-nitroaniline; and
4, Phe-Pro-OH. The points represent experimental data and the lines were
calculated by non-least-squares curve-fitting to eqs 1−3 and the degradation
pathway shown in Scheme 2.
racemization.7,10 In these studies, formation of Pro-Phe-
OH was not observed, indicating that sequence inversion
did not occur. The possibility of racemization of the
dipeptide was not investigated in this study.
The kinetics of the Phe-Pro-pNA degradation were
studied in aqueous buffer solutions at 37 °C over the pH
range 1-10. The disappearance of the Phe-Pro-pNA and
appearance of the degradation products were monitored by
a stability-indicating HPLC method (see Materials and
Methods). Pseudo-first-order kinetics was observed in all
cases, and the rate constants, kobs, for the formation of DKP
were obtained by linear regression of semilogarithmic first-
order plots. Figure 1 shows a typical time course of the
disappearance of Phe-Pro-pNA and the appearance of Phe-
Pro-DKP and p-nitroaniline. In the pH range 3-8, Phe-
Pro-DKP was the only degradation product observed;
formation of Phe-Pro-OH was not seen in this pH range.
However, at pH values below 3 and above 8 detectable
levels of Phe-Pro-OH were observed. The Phe-Pro-OH was
formed after an initial lag phase that corresponded to about
two half-lives of Phe-Pro-pNA (Figure 2). This lag time
suggests that Phe-Pro-OH was possibly formed from Phe-
Pro-DKP rather than from Phe-Pro-pNA. This hypothesis
was tested by modeling the degradation curves (i.e., Figure
Results and Discussion
Kin etics of Dik etop ip er a zin e F or m a tion sThe two
possible pathways for the degradation of Phe-Pro-pNA are
intramolecular aminolysis, yielding the corresponding Phe-
Pro-DKP, and direct hydrolysis, which produces the dipep-
tide Phe-Pro-OH (Scheme 1). Phe-Pro-DKP could then
undergo further hydrolysis at either of the two carbonyl
groups, producing two different dipeptides as shown in
Scheme 1. Other reactions that have been observed to
occur during DKP formation are sequence inversion and
284 / Journal of Pharmaceutical Sciences
Vol. 87, No. 3, March 1998

Scheme 2
2) to the rate equations that describe the reaction pathway
shown in Scheme 2.
A ) A0 exp(-kamt)
(1)
B ) {(A0kam)/(kdkp - kam)}{exp(-kamt) - exp(-kdkpt)}
(2)
C ) A0 +{A0/(kam - kdkp)}{kdkp exp(-kamt) - kam
exp(-kdkpt)} (3)
The disappearance of Phe-Pro-pNA and the appearance
of Phe-Pro-DKP and Phe-Pro-OH were fitted with a cor-
relation coefficient of 0.9954 (Figure 2). In separate
experiments, Phe-Pro-DKP was found to be stable between
pH 3 and 8, but it hydrolyzed to Phe-Pro-OH at pH values
below 3 and above 8 (data not shown). These results
indicate that the Phe-Pro-OH that is formed during the
degradation of Phe-Pro-pNA must be due to the hydrolysis
of Phe-Pro-DKP and not the hydrolysis of Phe-Pro-pNA.
This is consistent with previous investigations that dem-
onstrated that ester and amide functional groups of dipep-
tides undergo intramolecular aminolysis at a much faster
rate than hydrolysis.4,20
Bu ffer Ca ta lysissThe possible catalytic effect of the
buffer was determined by measuring the rate of Phe-Pro-
DKP formation from Phe-Pro-pNA at constant pH, ionic
strength, and temperature while varying the buffer con-
centration. Plots of kobs values versus total buffer concen-
trations, [B]t, were fit to eq 4
Figure 3sA plot of pseudo-first-order rate constants versus total buffer
concentrations for the degradation of Phe-Pro-pNA in aqueous solution at
37 °C.
kobs ) ko + kcat[B]t
(4)
where ko is the observed rate constant at zero buffer
concentration and kcat is the catalytic rate constant.
At pH below 5.0, kobs does not increase with increasing
buffer concentration, suggesting no buffer catalysis (data
not shown). However, at higher pH values (i.e., above 5.0),
there is a significant increase in kobs as the phosphate and
glycine buffer concentrations increase, indicating the pres-
ence of buffer catalysis (Figure 3a,b). The apparent rate
constants of the buffer-catalyzed reaction, kcat, were used
to determine the presence of general acid or general base
catalysis.
stant for the N-terminal amino group, k1 and k2 are
microscopic first-order rate constants for the water or
neutral catalysis of the protonated and unprotonated
species, respectively, and k3 is the second-order rate
constant for specific base catalysis.
ko ) k1[[H]/([H] + Ka)] + (k2 + k3[OH])[ka/(ka + [H])]
(5)
The kinetic rate constants and dissociation constant used
For the glycine buffer, the rate of Phe-Pro-DKP forma-
tion increases linearly with increasing fraction of un-
charged buffer species, suggesting that the uncharged form
of the buffer is the effective catalyst. On the other hand,
the phosphate buffer exhibits a nonlinear relationship
between the apparent rate constant of the buffer catalyzed
reaction, kcat, and the fraction of HPO42- (Figure 4). There
are two possible explanations for this nonlinearity: (i) the
fractional species of Phe-Pro-pNA is changing and the
buffer is acting on different ionic forms or (ii) there is a
change in the rate-determining step of the reaction.21
p H Dep en d en cesThe rate constants at zero buffer
concentration, ko, were obtained from the intercepts of plots
of kobs versus total buffer concentration (i.e., Figure 3). The
pH dependence of the rate of Phe-Pro-DKP formation at
37 °C and I ) 1.0 is shown in Figure 5. A possible
explanation for the unusual pH profile is shown in Scheme
3, where AH2+ represents the protonated reactant and AH
the neutral form. The overall pH-rate profile can be
adequately described by eq 5, where ko is the pH-dependent
first-order constant, Ka is the apparent dissociation con-
to generate the theoretical pH-rate profile, the solid line
in Figure 5, are the following: k1 ) 3.5((0.4) × 10-4 h-1
k2 ) 2.7((0.5) × 10-2 h-1, k3 ) 1.2((0.08) × 103 M-1 h-1
and Ka ) 7.9((0.6) × 10-6 (pKa ) 6.1).
,
,
The shape of the pH-rate profile suggests that the
degradation kinetics and mechanism of Phe-Pro-pNA are
affected by the ionization of the N-terminal amino group.
It is also apparent that the unprotonated reactant is much
more reactive than the protonated form. At or pH below
3, where the reactant exists predominantly in the proto-
nated form, AH2+, the rate of degradation is independent
of pH. From pH 4-6, the rate increases with pH, but it is
not first-order in hydroxide ion (approximate slope ) 0.73).
This may due to the complexity of the mechanism involving
different ionic species and buffer catalysis. Between pH 6
and 8 there is a plateau. In this region, the free N-terminal
amino group is available for cyclization, but the hydroxide
ion concentration is not high enough, so degradation occurs
by the neutral-catalyzed pathway. In the pH region 9-10,
the slope of the pH-rate profile is unit positive, indicating
specific hydroxide ion catalysis. It should be pointed out
Journal of Pharmaceutical Sciences / 285
Vol. 87, No. 3, March 1998

Figure 4sA plot of the catalytic rate constants versus the fraction of buffer
in the base form for Phe-Pro-DKP formation in phosphate buffer at various
pH values.
Figure 6sEyring plot for the formation of Phe-Pro-DKP at pH 2.0, 7.0, and
10.0 at 0.05 M buffer and I ) 1.0.
Table 1sSummary of Rate Constant for Diketopiperazine Formation
of X-Pro-pNA at pH 7.0 (37 °C, I ) 1.0)
X for analogs
(Rate (h-1) ± SD) × 102
Gly
Ala
Val
Phe
0.57 ± 0.15
2.57 ± 0.13
1.16 ± 0.20
5.82 ± 0.24
3.50 ± 0.21
3.73 ± 0.10
â-cyclohexylalanine
Arg
was observed. However, at pH 2.0 and 10.0, Phe-Pro-OH
was also produced, due to the further degradation of Phe-
Pro-DKP. At all pH values, the rate of DKP formation
increased with increasing temperature, obeying the Ar-
rhenius equation. From the Arrhenius plot, apparent
activation energies of 102.8 ( 5.4, 73.1 ( 4.8, and 57.3 (
5.9 kJ mol-1 were calculated for pH 2.0, 7.0, and 10.0,
respectively. The other activation parameters calculated
Figure 5sThe pH−rate profile for the formation of Phe-Pro-DKP in aqueous
solution at 37 °C. The solid line represents the theoretical profile based on
the nonlinear regression fit to eq 5.
from the Eyring plot (Figure 6) were as follows: ∆Hq
)
100.2 ( 7.2 kJ mol-1 and ∆Sq ) 9.3 ( 5.1 J K-1 mol-1 for
pH 2.0, ∆Hq ) 70.4 ( 5.4 kJ mol-1 and ∆Sq ) -46.3 (
10.2 J K-1 mol-1 for pH 7.0, and ∆Hq ) 56.4 ( 6.1 kJ mol-1
and ∆Sq ) -77.3 ( 9.8 J K-1 mol-1 for pH 10.0. The
positive value of entropy of activation at pH 2.0 suggests
an increase in entropy at the transition state. This
increase is due to a larger contribution of the positive
entropy from the expulsion of water molecules of the highly
solvated charged reactant than from the negative entropy
value upon cyclization.
Scheme 3
that the reaction described by the neutral catalysis of the
protonated species is kinetically equivalent to the acid
catalysis of the unprotonated species and, similarly, the
reaction for the neutral catalysis for the unprotonated
species is kinetically equivalent to the base-catalyzed
reaction of the protonated species. From the pH-rate
profile and the fact that the unprotonated form of the
reactant is more reactive than the protonated form, the
water-catalyzed reaction was chosen for both reactions
described above, although no strict basis exists for the
preference of one over the other.
Effect of Tem p er a tu r esThe rate of Phe-Pro-DKP
formation from Phe-Pro-pNA was studied at pH 7.0 (0.05
M phosphate buffer, I ) 1.0) from 37 to 70 °C. Similar
experiments were carried out at pH 2.0 and 10.0 (0.05 M
glycine buffer, I ) 1.0). At all temperatures and pH values,
the disappearance of the dipeptide analogue followed first-
order kinetics. For pH 7.0, at all temperatures, Phe-Pro-
DKP was the only product formed; no hydrolytic product
Effect of P r im a r y Sequ en cesThe effect of the N-
terminal residue on the rate of DKP formation at pH 7.0
was examined by studying the kinetics of degradation of
X-Pro-pNA analogues, where X ) Gly, Ala, Val, Arg, Phe,
and â-cyclohexylalanine. Varying the amino acid located
on the N-terminal side of the proline residue has a
significant effect on the rate of degradation (Table 1). The
differences observed in the rate of DKP formation can be
explained by (i) differences in the pKa values of the terminal
amino group of the analogues, (ii) differences in the steric
bulk, (iii) the ability of the X-Pro peptide bond to undergo
cis-trans isomerization, and/or (iv) the conformational
stability of the resulting DKPs.
The shape of the pH-rate profile around pH 7.0 does
not change as changes are made in the N-terminal amino
acid (data not shown). This suggests that the pKa of the
terminal amino group does not change dramatically and,
therefore, does not influence the rate of DKP formation for
the analogues studied.
286 / Journal of Pharmaceutical Sciences
Vol. 87, No. 3, March 1998

The effect of alkyl and aryl substituents in intramolecu-
lar reactions has been well documented, with the results
suggesting that an increase in the bulkiness of the sub-
stituent increases the rate of cyclization.22-25 This does
not seem to be true for the series studied herein, as the
Ala analogue cyclizes twice as fast as the bulkier Val
analogue. Thus, simple steric bulk cannot be used to
explain the data.
References and Notes
1. Battersby, J . E.; Hancock, W. S.; Canova-Davis, E.; Oeswein,
J .; O’Connor, B. Diketopiperazine formation and N-terminal
degradation in recombinant human growth hormone. Int. J .
Pept. Protein Res. 1994, 44, 215-222.
2. Kertscher, U.; Bienert, M.; Krause, E.; Sepetov, N. F.; Mehlis,
B. Spontaneous chemical degradation of substance P in the
solid phase and in solution. Int. J . Pept. Protein Res. 1993,
41, 207-221.
3. Beyermann, M.; Bienert, M.; Niedrich, H.; Carpino, L. A.;
Sadat-Aalace, D. Rapid continuous peptide synthesis via
FMOC amino acid chloride coupling and 4-(aminomethyl)-
piperidine deblocking. J . Org. Chem. 1990, 55, 721-728.
4. Purdie, J . E.; Benoiton, N. L. Piperazinedione formation from
esters of dipeptides containing glycine, alanine, and sar-
cosine: The kinetics in aqueous solution. J . Chem. Soc.
Perkin Trans. 2 1973, 13, 1845-1852.
Unlike other peptide bonds, those between Pro and its
preceding amino acid (X-Pro) can exist as a mixture of cis
and trans isomers in solution.26 Usually the trans isomer
is favored over the cis in the absence of ordered structure,
but structural constraints can stabilize one isomer over the
other.27 In addition, the cis-trans equilibrium depends on
the nature of the flanking amino acids and on the charge
distribution around the X-Pro-peptide bond.26,28 In DKPs,
the two side chains are located on the opposite sides of the
general plane of the ring, which greatly reduces the strain
applied to the six-member ring. On the basis of space-
filling models of dipeptide esters containing a proline
residue, the cis isomer has been shown from a geometrical
consideration to adapt a conformation in which the attack-
ing N-terminal amino nitrogen atom is in proximity to the
amide carbonyl, favoring ring closure.4 Thus, one would
expect the cis isomer to undergo aminolysis more rapidly
than its trans counterpart. There are no correlations
between the bulkiness of the side chain of the amino acid
and the occurrence of a cis peptide bond.29 Investigations
are currently underway in our laboratory to determine the
ability of these dipeptide analogues to undergo cis-trans
isomerization.
The conformation of the Phe-Pro-DKP in aqueous solu-
tion has been determined,30,31 and it was shown that the
DKP ring adopted a boat-shaped conformation with the Phe
aromatic ring stacked over it. Ciarkowsi et al.32 argued
that the presence of the ring-ring interaction in a cyclic
dipeptide composed of an aromatic amino acid residue
adopted a more energetically favored conformation, in
which the aromatic ring is stacked over the DKP ring, and
that the intramolecular dipole-induced dipole should
stabilize this form. There is also the possibility of an
aromatic-aromatic interaction between the Phe side chain
and the pNA leaving group, which could stabilize the
tetrahedral intermediate and lead to an enhanced rate of
cyclization of Phe-Pro-pNA. To check this hypothesis, the
aromatic Phe residue was replaced with â-cyclohexylala-
nine. As shown in Table 1, the rate decreased by a factor
of approximately 1.6, suggesting that an aromatic-
aromatic interaction may be one of many factors respon-
sible for the intramolecular reaction. Further studies are
needed to elucidate the contribution of these different
factors.
5. Skwierczynski, R. D.; Connors, K. A. Demethylation kinetics
of aspartame and L-phenylalanine methyl ester in aqueous
solution. Pharm. Res. 1993, 10, 1174-1180.
6. Gu, L.; Strickley, R. G. Diketopiperazine formation, hydroly-
sis, and epimerization of the new dipeptide angiotensin-
converting enzyme inhibitor RS-10085. Pharm. Res. 1987,
4, 392-397.
7. Moss, J .; Bundgaard, H. Kinetics and mechanism of the facile
cyclization of histidyl-prolineamide to cyclo(His-Pro) in aque-
ous solution and the competitive influence of human plasma.
J . Pharm. Pharmacol. 1990, 42, 7-12.
8. Miyashita, K.; Murakami, M.; Yamada, M.; Iriuchijima, T.;
Mori, M. Histidyl-proline diketopiperazine. J . Biol. Chem.
1993, 268, 20863-20865.
9. Steinberg, S. M.; Bada, J . L. Peptide decomposition in the
neutral pH region via the formation of diketopiperazines. J .
Org. Chem. 1983, 48, 2295-2298.
10. Steinberg, S. M.; Bada, J . L. Diketopiperazine formation
during investigations of amino acid racemization in dipep-
tides. Science 1981, 213, 544-545.
11. Bodanszky, M.; Martinez, J . In The Peptides: Analysis,
Synthesis, Biology; Gross, E., Meienhoffer, J ., Eds.; Academic
Press Inc.: New York, 1983; Vol. 5, pp 120-123, 193-196.
12. Gerig, J . T.; McLeod, R. S. Attempted synthesis of 2-meth-
ylalanyl-L-prolyl-L-tryptophan. An unexpected result. J . Org.
Chem. 1976, 41, 1653-1655.
13. Gisin, B. F.; Merrifield, R. B. Carboxyl-catalyzed intramo-
lecular aminolysis. A side reaction in solid-phase peptide
synthesis. J . Am. Chem. Soc. 1972, 94, 3102-3106.
14. Mazurov, A. A.; Andronati, S. A.; Korotenko, T. I.; Gorbatyuk,
V. Y.; Shapiro, Y. E. Formation of pyroglutamylglutamine
(or asparagine) diketopiperazine in ‘non- classical’ condi-
tions: A side reaction in peptide synthesis. Int. J . Pept.
Protein Res. 1993, 42, 14-19.
15. Meienhoffer, J . Syntheses of actinomycin and analogues. III.
A total synthesis of actinomycin D (C1) via peptide cyclization
between proline and sarcosine. J . Am. Chem. Soc. 1970, 92,
3771-3777.
16. Fields, G. B.; Noble, R. L. Solid-phase peptide synthesis
utilizing 9- fluorenylmethoxycarbonyl amino acid. Int. J .
Pept. Protein Res. 1990, 35, 16-214.
17. Sepetov, N. F.; Krymsky, M. A.; Ovchinnikov, M. V.; Be-
spalova, Z. D.; Isakova, O. L.; Soueek, M.; Lebl, M. Rear-
rangement, racemization and decomposition of peptides in
aqueous solution. Pept. Res. 1991, 4, 308-313.
18. Straub, J . A.; Akiyama, A.; Parmar, P.; Musso, G. F.
Chemical pathways of the bradykinin analogue, RMP-7.
Pharm. Res. 1995, 12, 305-308.
19. Oyler, A. R.; Naldi, R. E.; Lloyd, J . R.; Graden, D. A.; Shaw,
C. J . Characterization of the solution degradation products
of Histrelin, a gonadotropin releasing hormone (LH/RH)
agonist. J . Pharm. Sci. 1991, 80, 271-275.
20. Radzicka, A.; Wolfenden, R. Rates of uncatalyzed peptide
bond hydrolysis in neutral solution and the transition state
affinities of proteases. J . Am. Chem. Soc. 1996, 118, 6105-
6109.
21. J enks, W. P. In Catalysis in Chemistry and Enzymology;
McGraw-Hill: New York, 1969; pp 163-195.
22. Bruice, T. C.; Bradbury, W. C. The gem effect. The influence
of 3-mono- and 3,3-disubstitution on the rates of solvolysis
of mono-p-bromophenyl glutarate. J . Am. Chem. Soc. 1965,
87, 4846-4850.
23. Higuchi, T.; Eberson, L.; McRae, J . D. Acid anhydride-free
acid equilibria in water in some substituted succinic scid
systems and their interaction with aniline. J . Am. Chem. Soc.
1967, 89, 3001-3004.
24. Borchardt, R. T.; Cohen, L. A. Stereopopulation Control. II.
Rate enhancement of intramolecular nucleophilic displace-
ment. J . Am. Chem. Soc. 1972, 94, 9166-9174.
Conclusions
These results indicate that DKP formation is sensitive
to the pH of the solution, the buffer species present at high
pH, the temperature, and the N-terminal amino acid
residue. The rapid rate at which the intramolecular
reaction occurs at neutral and basic pH suggests that DKP
formation may play an integral role in the degradation of
peptides and proteins that contain a proline residue at
position 2 under these conditions. The results of these
studies will hopefully give scientists involved in the
formulation of therapeutically important peptides and
proteins insight into the multitude of endogenous (e.g.,
primary sequence) and exogenous (e.g., pH, buffer species)
factors that can influence the rate of diketopiperazine
formation in aqueous solution.
Journal of Pharmaceutical Sciences / 287
Vol. 87, No. 3, March 1998

25. Borchardt, R. T.; Cohen, L. A. Stereopopulation Control. III.
Facilitation of intramolecular conjugate addition of the
carboxyl group. J . Am. Chem. Soc. 1972, 94, 9175-9182.
26. Levitt, M. Effect of proline residues on protein folding. J .
Mol. Biol. 1981, 145, 251-263.
31. Budesinsky, M.; Symersky, J .; J ency, J .; VanHecke, J .;
Hosten, N.; Anteunis, M.; Borremans, F. A. M. Cyclo (L-
propyl-L-N-methylphenylalanyl). Conformation in solution
and in the crystal. Int. J . Pept. Protein Res. 1992, 39, 123-
130.
32. Liwo, A.; Ciakowski, J . Origin of the ring-ring interaction
in cyclic dipeptides incorporating an aromatic amino acid.
Tetrahedron Lett. 1985, 26, 1873-1876.
27. Stein, R. L. In Advances in Protein Chemistry; Afinsen, C.
B., Edsall, J . T., Richards, F. M., Eisenberg, D. S., Eds.;
Academic Press, Inc.: New York, 1993; Vol. 44.
28. Ramachandran, G. N.; Mitra, A. K. An explanation for the
rare occurrence of cis peptide units in proteins and polypep-
tides. J . Mol. Biol. 1976, 107, 85-92.
Acknowledgments
29. Stewart, D. E.; Sarkar, A.; Wampler, J . E. Occurrence and
role of cis peptide bonds in protein structures. J . Mol. Biol.
1990, 214, 253-260.
This work was supported by the United States Public Health
Service (GM-088359). We would also like to thank Drs. Richard
Schowen and Suman Patel for their advice.
30. Ciarkowski, J .; Gdaniec, M.; Kolodziejczyk, A.; Liberek, B.;
Borremans, F. A. M.; Anteunis, M. J . O. Conformation of
cyclo-(D-phenylalanyl-trans-4-fluoro-D-prolyl). Int. J . Pept.
Protein Res. 1990, 36, 285-291.
J S970325M
288 / Journal of Pharmaceutical Sciences
Vol. 87, No. 3, March 1998