p-Cresol as a reversible acylium ion scavenger in solid-phase peptide synthesis

Article abstract of DOI:10.1021/ja973322k

In this work we have defined the nature of the p-cresol and p-thiocresol adducts generated from acylium ions during HF cleavage, following contemporary Boc/benzyl solid-phase peptide synthesis. Contrary to the results in previous reports, we found that bo

Full text of DOI:10.1021/ja973322k

1410
J. Am. Chem. Soc. 1998, 120, 1410-1420
p-Cresol As a Reversible Acylium Ion Scavenger in Solid-Phase
Peptide Synthesis
Les P. Miranda, Alun Jones, Wim D. F. Meutermans,* and Paul F. Alewood
Contribution from the Centre for Drug Design & DeVelopment, The UniVersity Of Queensland,
Brisbane, Queensland 4072, Australia
ReceiVed September 22, 1997
Abstract: In this work we have defined the nature of the p-cresol and p-thiocresol adducts generated from
acylium ions during HF cleavage, following contemporary Boc/benzyl solid-phase peptide synthesis. Contrary
to the results in previous reports, we found that both p-cresol and p-thiocresol predominantly form aryl esters
under typical cleavage conditions. Initially we investigated a number of small peptides containing either a
single glutamate residue or a C-terminal long-chain amino acid which allowed us to unambiguously characterize
the “scavenged” side products. Whereas, the p-cresol esters are stable at 0 °C they rearrange irreversibly at
higher temperatures (5-20 °C) to form aryl ketones. By contrast, p-thiocresol esters do not undergo a Fries
rearrangement but readily undergo further additions of p-thiocresol to form ketenebisthioacetals and trithio
ortho esters, even at low temperatures. Importantly, we found by LC/MS and FT-ICR MS analysis that peptides
containing p-cresol esters at glutamyl side chains are susceptible to amidation and fragmentation reactions at
these sites during standard mild base workup procedures. The significance of these side reactions was further
demonstrated in the synthesis of neutrophil immobilization factor, a 26-residue peptide, containing four glutamic
acid residues. The side reactions were largely avoided by mild hydrogen peroxide-catalyzed hydrolysis which
converted the p-cresol adducts to the free carboxylic acids in near quantitative yield. The choice of p-cresol
as a reversible acylium ion scavenger when coupled with the simple workup conditions described is broadly
applicable to Boc/benzyl peptide synthesis and will significantly enhance the quality of peptides produced.
Introduction
and scavengers. The most effective and commonly used acid
for this purpose is anhydrous hydrogen fluoride,3b although acids
such as hydrogen bromide4,5 and trifluoromethanesulfonic acid
in trifluoroacetic acid have also been used successfully.6
HF is a highly volatile, nonoxidizing acid with a strong
protonating ability and forms an excellent solvating medium
for peptides. When used in high concentrations, HF removes
all the common protection groups efficiently via an SN1 process,
thereby generating carbocations, which are subsequently trapped
by the added scavenger(s). Several HF-induced side reactions
have been observed and thoroughly studied, for example: (1)
side chain alkylation at tyrosine,7 tryptophan,8,9 methionine,10
and cysteine11,12 residues due to insufficient scavenging of
various carbocations; (2) formation of succinimide at aspartyl
residues, which, upon hydrolysis, ring open to form both
R-(backbone) and â-(side chain) peptides;13 (3) N f O acyl
One of the main challenges in peptide synthesis is to establish
synthetic routes to homogeneous products of defined covalent
structure. Since Merrifield introduced solid-phase peptide
synthesis (SPPS),1,2 new synthetic protocols have continually
been formulated toward greater efficiency in peptide assembly
and cleavage from the solid support. Despite this progress, side
reactions still plague many syntheses, often resulting in low
yields and poor homogeneity of the target peptide, especially
for peptides over 20 residues.
In Boc SPPS3a the protection strategy adheres to the principle
of graded acid lability. Following chain assembly, cleavage
from the solid support and side chain deprotection is generally
performed by treating the resin bound peptides with strong acids
* Author to whom correspondence should be addressed: email
W.Meutermans@mailbox.uq.edu.au.
(1) Merrifield, R. B. Fed. Proc. Am. Soc. Exp. Biol. 1962, 21, 412.
(2) Merrifield, R. B. J. Am. Chem. Soc. 1963, 85, 2149-2154.
(3) (a) Boc SPPS refers to use of tert-butoxycarbonyl and benzyl
temporary and permanent protecting groups, respectively. Abbreviations
used are the following: Bzl, benzyl; CH3CN, acetonitrile; COSY, correlation
spectroscopy; Da, Dalton; DCM, dichloromethane; DIEA, diisopropylethyl-
amine; DMF, N,N-dimethylformamide; ES-MS, electrospray mass spec-
trometry; FT-ICR MS, Fourier transform ion cyclotron resonance mass
spectrometry; H2O2, hydrogen peroxide; HBTU, 2-(1H-benzotriazol-1-yl)-
1,1,3,3-tetramethyluronium hexafluorophosphate; HF, anhydrous hydrogen
fluoride; LC/MS, liquid chromatography mass spectrometry; Merrifield
resin, chloromethylpolystyrene-divinylbenzene (200-400 mesh); MRP14,
migration inhibitory factor related protein 14 kDa; NIF, neutrophil im-
mobilization factor; NMR, nuclear magnetic resonance; Pam, phenylacet-
amidomethyl polystyrene resin; RP-HPLC, reverse-phase high-performance
liquid chromatography; SPPS, solid-phase peptide synthesis; TFA, trifluoro-
acetic acid. Standard IUPAC single and triple letter codes for amino acids
are used throughout. (b) Sakakibara, S.; Shimonishi, Y. Bull. Chem. Soc.
Jpn. 1965, 38, 1412-1413.
(4) Ben-Ishai, D.; Berger, A. J. Org. Chem. 1952, 17, 1564-1570.
(5) Guttmann, S.; Boissonnas, R. A. HelV. Chim. Acta 1959, 42, 2721-
2739.
(6) Yajima, H.; Kiso, Y.; Ogawa, H.; Kawatani, H. J. Chem. Soc. Chem.
Commun. 1974, 107-108.
(7) Erickson, B. W.; Merrifield, R. B. J. Am. Chem. Soc. 1973, 95, 3750.
(8) Wu¨nsch, E.; Jaeger, E.; Kisfaludy, L.; Low, M. Angew. Chem. 1997,
89, 330.
(9) Masiu, Y.; Chino, N.; Sakakibara, S. Bull. Chem. Soc. Jpn. 1980,
53, 464-468.
(10) Noble, R. L.; Yamashiro, D.; Li, C. H. J. Am. Chem. Soc. 1976,
98, 2324-2328.
(11) Tam, J. P.; Merrifield, R. B. In The Peptides: Analysis, Synthesis,
Biology; S. Udenfriend and J. Meienhofer, Eds.; Academic Press: London,
1987; Vol. 9; pp 191-199.
(12) Tam, J. P. In Macromolecular Sequencing and Synthesis: Selected
Methods and Application; Alan R. Liss, Ed.: 1988; pp 153-184.
(13) Baba, T.; Sugiyama, H.; Seto, S. Chem. Pharm. Bull. 1973, 21,
207-209.
S0002-7863(97)03322-2 CCC: $15.00 © 1998 American Chemical Society
Published on Web 02/04/1998

p-Cresol As a ReVersible Acylium Ion ScaVenger
J. Am. Chem. Soc., Vol. 120, No. 7, 1998 1411
Scheme 1. Reactions of ω-Amino Acid Derivatives in HF/p-Cresol (9:1) at 0 and 20 °C
transfer of serine- and threonine-containing peptides from the
backbone nitrogen atom to the side chain oxygen atom on the
serine and threonine residues14-17 (reversed by mild base
treatment); (4) acylation via acylium ion formation on glutamyl
side chains18-20 and on C-terminal ω-amino acids.21 The extent
of these side reactions is largely controlled by the cleavage
conditions, including time, temperature, and HF concentration,
and by the protection groups used in the synthesis. In an effort
to minimize cleavage side reactions, Tam et al.22 proposed the
use of a “low-HF” mixture (HF/dimethyl sulfide/p-cresol, 25:
65:10, v/v) where initial cleavage of most side chain groups
follows an SN2 mechanism. This is then followed by a “high-
HF” treatment to facilitate complete side chain deprotection and
cleavage from the solid support (HF/p-cresol/p-thiocresol, 18:
1:1, v/v). Since most of the carbocation precursors are removed
under low-HF conditions, the number of carbocations produced
during the “high-HF” step would be greatly reduced and, in
principle, the number of side products would be greatly reduced
as well.
Here we report a detailed study on the nature of the p-cresol
and p-thiocresol scavenging of acylium ions and examine their
new role in peptide synthesis as reversible acylium ion
scavengers for improving yields and homogeneity of target
peptides.
Results and Discussion
HF Cleavage of ω-Amino Acids. Long-chain amino acids
at the C-terminus of peptides nearly quantitatively generate
acylium ions in anhydrous HF21 under standard cleavage
conditions. They thus provided an excellent model for our initial
examination of the nature of p-cresol scavenging and its
accompanying side reactions.
We subjected amino acid derivatives 1a to standard HF
cleavage conditions (0 °C, 1 h) in the presence of p-cresol as
scavenger (Scheme 1). In both cases (R ) Merrifield resin or
H), the main component (> 85%) of the residue after HF
evaporation was characterized unambiguously by ES-MS and
1H and 13C NMR as ester 2a. That 2a is indeed an ester and
not the expected ketone is evident from the para-substitution
However, even under these controlled conditions acylium ions
may still be efficiently formed from the unprotected carboxylic
acids.21 If not scavenged, the γ-glutamyl acylium ion ring closes
to the R-nitrogen atom in the backbone thus forming a
pyrrolidinone which is then susceptible to several subsequent
reactions during the workup of the cleavage mixture.18 In the
presence of aromatic scavengers, such as anisole, the acylium
ion is trapped irreversibly through a Friedel-Crafts acylation
generating an aryl ketone.18,23 For long-chain carboxylic acids
separated from the amide backbone by more then two methyl-
enes, acylium ion formation under neat HF cleavage conditions
is nearly quantitative and results in total loss of target peptide.21
Effective Boc solid-phase synthesis of homogeneous peptides
containing glutamyl residues and/or long-chain carboxylic acids
would benefit greatly from a new scavenging mechanism. Such
a mechanism should, on the one hand, rapidly trap the acylium
ion, and on the other hand, the trapped product should be
convertible to the target peptide via a mild chemical procedure
that does not affect other functionalities in the molecule.18
1
pattern of the phenyl ring in the H NMR spectrum and the
carbonyl resonance in the 13C NMR spectrum at 183 ppm. In
0.1 M NH4HCO3, ester 2a was converted to lactam 3a, as
evidenced by ES-MS and RP-HPLC comparison with an
authentic sample. Similarly, aminocaproic acid derivatives 1b
mainly (> 85%) generated ester 2b under standard HF/p-cresol
cleavage conditions.
By contrast, longer cleavage times and higher temperatures
caused substantial formation of ketones 4 via a HF-catalyzed
(14) Shin, K. H.; Sakakibara, S.; Scheinder, W.; Hess, G. P. Biochem.
Biophys. Res. Commun. 1962, 8, 288-293.
(15) Sakakibara, S.; Shin, K. H.; Hess, G. P. J. Am. Chem. Soc. 1962,
84, 4921-4928.
(16) Muir, T. W.; Williams, M. J.; Kent, S. B. H. Anal. Biochem. 1995,
224, 100- 109.
(17) Fujino, M.; Wakimasu, M.; Shinagawa, S.; Kitada, C.; Yajima, H.
Chem. Pharm. Bull. 1978, 26, 539-548.
(18) Feinberg, R. S.; Merrifield, R. B. J. Am. Chem. Soc. 1975, 97, 3485-
3496.
(19) Sano, S.; Kawanishi, S. J. Am. Chem. Soc. 1975, 97, 3480-3484.
(20) Schon, I.; Kisfaludy, L. Int. J. Pept. Protein Res. 1979, 14, 485-
494.
(21) Bednarek, M. A.; Springer, J. P.; Cunningham, B. R.; Bernick, A.
M.; Bodanszky, M. Int. J. Pept. Protein Res. 1993, 42, 10-13.
(22) Tam, J. P.; Heath, W. F.; Merrifield, R. B. J. Am. Chem. Soc. 1983,
105, 5, 6442-6455.
(23) Stewart, J. M.; Young, J. D. In Solid-Phase Peptide Synthesis; Pierce
Chemical Company: 1984; pp 40-47.
(24) Merrifield, R. B. In Peptides: Synthesis, Structure and Applications;
B. Gutte, Ed.; Academic Press: San Diego, 1995; pp 93-169.
(25) Tam, J. P.; Heath, W. F.; Merrifield, R. B. Int. J. Pept. Protein
Res. 1983, 21, 57-65.
(26) Heath, W. F.; Tam, J. P.; Merrifield, R. B. Int. J. Pept. Protein
Res. 1986, 28, 896-897.
(27) Alewood, P. F.; Meutermans, W. D. F. In 23rd European Peptide
Symposium; Escom: Braga, 1994; pp 242-243.
p-Cresol and p-thiocresol are commonly used scavengers for
HF cleavage of protected peptide resins.24 p-Cresol is highly
effective in preventing tyrosine ring benzylation and aspartimide
formation during HF-mediated cleavage,12,25 while p-thiocresol
improves cation scavenging and increases deprotection rates of
the protected forms of tryptophan, arginine or cysteine resi-
dues.11,12,22,26 It is generally believed that these scavengers act
in a fashion similar to anisole with acylium ion trapping leading
to aryl ketones.21,23 However, recent data from our laboratory
led us to believe that the formation of aryl ketones with cresol-
based scavengers did not occur under standard HF conditions
(0 °C, 1 h).27

1412 J. Am. Chem. Soc., Vol. 120, No. 7, 1998
Miranda et al.
Table 1. HF/p-Cresol (9:1) Treatment of ω-Amino Acids 1, Effect
of one molecule of p-thiocresol from 7 (Mr 343). Trithio ortho
esters that contain R protons have been reported to readily
eliminate thio alcohol and generate ketenebisthioacetals.30,31
Thus, we were able to convert the trithio ortho ester 7 to the
ketenebisthioacetals (8) by refluxing an aqueous solution
containing 10 equiv of ZnCl2.31 The 1H and 13C NMR data of
the “double adduct” confirmed structure 8. We also observed
that the trithio ortho ester 7 readily eliminated one p-thiocresol
molecule under standard ES-MS conditions with only a small
MH+ signal observed at 468 m/z with the predominant signal
at 344 m/z. Thus, it was clear that long-chain amino acids and
probably glutamyl side chains are susceptible to formation of
trithio ortho esters and ketenebisthioacetals under standard HF/
p-thiocresol cleavage conditions.
of Cleavage Conditions on the Product Distribution
n
(1) R
temp (°C)
time (h)
2 (%)
4 (%)
5 (%)
3
5
5
3
5
resin
resin
H
H
H
-5
-5
0
20
20
1
2
2
12
12
>85
80
75
<1
<1
<1
15
20
<1
30
<1
<1
2
85
50
a Yields were estimated by integration of the 1H NMR spectrum of
the crude cleavage product.
Scheme 2. Reactions of ω-Amino Acid Derivatives in
HF/p-Thiocresol (9:1)
A proposed mechanism for the formation of ketenebisthio-
acetals and trithio ortho esters from HF treatment of carboxylic
acids is shown in Scheme 3. The initially formed thioester 6
undergoes acid-catalyzed addition of p-thiocresol, followed by
elimination of H2O to the double adduct 8 or substitution by
another p-thiocresol molecule to the trithio ortho ester 7.
Comparison of Hydrolysis Rates of p-Cresol and p-
Thiocresol Esters 9a and 9b. In contrast to the aryl ketone
adducts formed through Friedel-Crafts acylation of anisole,
p-cresol and p-thiocresol esters may permit hydrolysis to the
carboxylic acid and in principle total recovery of the target
peptide. As the p-cresol esters 2 preferentially cyclize to lactam
under hydrolysis conditions, we chose the related model
dipeptide, arginine-6-aminocaproic acid, to define appropriate
conditions for the recovery of the target peptide. Arginine-6-
aminocaproic acid was assembled on Merrifield resin and
subjected to anhydrous HF treatment (1 h, 0 °C) in the presence
of either p-cresol or p-thiocresol. The main components isolated
from the resulting crude residues were characterized as the
p-cresol ester 9a or p-thiocresol ester 9b, respectively (Scheme
4). As expected, an LC/MS profile of the p-thiocresol scav-
enged cleavage product indicated the presence of double and
triple p-thiocresol adducts (data not shown). Only trace amounts
of the free acid were isolated from either of the crude residues.
Hydrolysis of the purified esters 9a and 9b (Scheme 4) was
slow at pH values below 10. However, the highly nucleophilic
peroxide anion efficiently catalyzes the hydrolysis of the esters.32
For instance, at pH 9 and using 1.2 equiv of hydrogen peroxide,
ester hydrolysis was complete in 8 min. Interestingly, the
catalyzed hydrolysis was found to proceed at comparable rates
for both the oxyester 9a and the thioester 9b (data not shown).
In terms of recovery of the carboxylic acid 10 from the
respective esters, there was no advantage in choosing p-
thiocresol over p-cresol as a scavenger.
Fries rearrangement28 of the initially formed esters 2. Ketone
4a was not directly detected as it cyclized spontaneously and
quantitatively to the Schiff’s base 5a. The correct molecular
1
weight, H NMR, and a characteristic 13C NMR signal at 183
ppm confirmed the imine structure over that of the intermediate
ketone. For 1b, LC/MS and 13C NMR analyses of the crude
residue (HF, 20 °C, 12 h) indicated the presence of both ketone
4b and cyclic imine 5b from resonances at 208 and 183 ppm,
respectively. Cyclization of ketone 4b to the Schiff’s base 5b
was driven to completion by base treatment of the crude mixture.
The effect of time and temperature on the product distribution
in the crude residue based on 1H NMR analysis is given in Table
1. All p-cresol esters had rearranged to their corresponding aryl
ketones after standing in HF for 12 h at 20 °C. Through careful
control of the cleavage temperature and time, the Fries re-
arrangement of the p-cresol ester can be largely avoided and
results in nearly full conversion of the acylium ion to the p-cresol
ester. The product distribution was similar for both the resin-
bound and free aminobutyric acid, which indicates that “depro-
tection” of the acid function prior to high-HF treatment (e.g.,
via a low-high HF treatment22) will not prevent acylium ion
formation in the cleavage process.
Similar to the above, HF treatment of compounds 1b at 0 °C
in the presence of p-thiocresol produces thioesters 6 (> 85%)
(Scheme 2). If a mixture of p-cresol and p-thiocresol (1:1) is
used, as is often the case in HF cleavages of peptidyl resins,
thioester 6 is still the predominant product, with only traces of
the O-ester 2 present. The RP-HPLC profile (data not shown)
of the crude cleavage mixtures indicated the presence of other
side products (7 and 8) corresponding to multiple additions of
p-thiocresol to the acylium ion derived from 1b.
HF Cleavage of LTEN-Resin. We have previously re-
ported that during the HF/p-cresol cleavage of the tetrapeptide
LTEN (Mr 475) from phenylacetamidomethyl polystyrene resin
(Pam),33 a similar acylium ion-directed side reaction occurred.34
Application of the low-high cleavage procedure22 did not
significantly alter the product distribution, i.e., the p-cresol
adducts were still formed in similar proportions. The yield of
the p-cresol adduct (Mr 565) increased with longer cleavage
times (2 h) to 50%, after RP-HPLC isolation, but was strongly
Aryl thioesters are not susceptible to Fries rearrangement29
and thus will not generate aryl thioketones in anhydrous HF,
even at higher temperatures or longer cleavage times. To verify
this and further elucidate the nature of the multiple p-thiocresol
adducts, aminocaproic acid on Merrifield resin (1b) was treated
with HF/p-thiocresol (9:1) at 20 °C for 2 h. The major
component of the reaction product (>85%) was characterized
as the trithio ortho ester 7 (Mr 467). A second component, 8
(5%), possessed a molecular weight corresponding to the loss
(30) Gro¨bel, B.-T.; Seebach, D. Synthesis 1977, 357-402.
(31) Simchen, G. In Methoden Der Organischen Chemie: Organische
Schwefel- Verbindungen; D. Klamann, Ed.; Verlag: New York, 1985; Vol.
B and E11.
(32) Kenner, G. W.; Seely, J. H. J. Am. Chem. Soc. 1972, 94, 3259.
(33) Mitchell, A. R.; Erickson, B. W.; Ryaltseo, M. W.; Hodges, R. S.;
Merrifield, R. B. J. Am. Chem. Soc. 1976, 98, 7357-7362.
(34) Alewood, P. F.; Bailey, A.; Brinkworth, R.; Fairlie, D.; Jones, A.
J. Chromatography 1993, 646, 185-192.
(28) Norell, J. B. J. Org. Chem. 1973, 38, 1924.
(29) Aslam, M.; Davenport, K. G.; Stansbury, W. F. J. Org. Chem. 1991,
56, 5955- 5958.

p-Cresol As a ReVersible Acylium Ion ScaVenger
J. Am. Chem. Soc., Vol. 120, No. 7, 1998 1413
Scheme 3. Proposed Mechanism for the Formation of Ketenebisthioacetals and Trithio Ortho Esters under HF/p-Thiocresol
Cleavage Conditions
Scheme 4. H2O2-Catalyzed Hydrolysis of p-Cresol Esters
For example, LC/MS analysis34,37 of ester 11a at pH 8 (0.1 M
NH4HCO3) and at pH 9.2 (Na2CO3/NaHCO3 buffer), shown in
Figure 1, indicated that in addition to the expected target peptide
LTEN (and LTQN from aminolysis with ammonia at pH 8),
three other main hydrolysis products were formed (Mr 243, 232,
and 214). Further LC/MS analysis at different time intervals
(pH 9) revealed the existence of an intermediate with Mr of
457. After 2 days at pH 9, the intermediate had fully converted
to the hydrolysis products mentioned above. LTEN itself is
stable under these basic conditions.
Scheme 5 depicts three possible reaction pathways for ester
11a. On the basis of the observed intermediate and its
conversion products, we believe reaction pathway II is the major
contributor to hydrolysis of ester 11a. Mild basic conditions
promote the early formation of a pyrrolidinone 12 (Mr 457).
These imides react readily following several possible pathways
depending on the buffer used and the pH of the solution.
Intramolecular aminolysis via route A yields diketopiperazine
13 (Mr 214) and pyrrolidinone 14 (Mr 243). Hydrolysis via
route B generates dipeptide 15 (Mr 232) and pyrrolidinone 14.
Direct hydrolysis via route C on the other hand results in
formation of the target LTEN 16a (and in 0.1 M NH4HCO3,
also LTQN 16b). The absence of the dipeptide 15 at pH 8 is
consistent with the proposed reaction scheme; at lower pH (data
not shown) intramolecular aminolysis to diketopiperazine 13
will be preferred over bimolecular hydrolysis (by a hydroxide
ion or ammonia). After 1 h of hydrolysis of ester 11a at pH 9,
the intermediate imide 12 was isolated by RP-HPLC and MS/
MS analysis performed. The fragmentation pattern in the
collision-induced dissociation experiment was consistent with
the proposed structure. LC/MS analysis of a 2-day-old pH 9
solution of this purified imide 12 displayed a similar hydrolysis
profile: hydrolysis to dipeptide 15, diketopiperazine 13, and
some recovery of LTEN 16a. The existence of these side
influenced by the batch of resin used.35a LC/MS, 1H NMR (1D,
COSY), and 13C NMR analyses allowed full characterization
of the ester structure 11a (Scheme 5). MS/MS of the ester
molecular ion displayed a fragmentation pattern consistent with
the proposed position of the ester, i.e., on the glutamyl side
chain. Contrary to the long-chain amino acids (Table 1), no
ketone was observed in the crude product by NMR or LC/MS,
even after longer cleavage times (2 h). Importantly, the
unprotected tetrapeptide LTEN 16a under identical HF cleavage
conditions (HF/p-cresol (9:1) at 0 °C) generated the same
products. Thus, similar to the long-chain acids, side chain
deprotection of the glutamate residue in LTEN prior to high-
HF treatment did not prevent acylium ion adduct formation.
Side Reactions During Rearrangement of N f O Acyl
Migrations. In standard workup procedures, crude cleaved
peptides are often treated with bicarbonate buffers (usually in
the pH range of 7.5-9.5) for 12 h to reverse potential N f O
shifts at threonine or serine residues.23,39,40 In our hands, mild
base treatments of the p-cresol ester 11a gave complex mixtures.
(35) (a) In another experiment using a different batch of Boc-Asn-OCH2-
Pam resin from the same company, the yield of the p-cresol adduct under
the same reaction conditions was significantly lower (10%). (b) Smith, R.
D.; Loo, J. A.; Edmonds, C. G.; Barinaga, C. J.; Udseth, H. R. Anal. Chem.
1990, 62, 882-899.
(36) Smith, R. D.; Barinaga, C. J.; Udseth, H. R. Anal. Chem. 1988, 60,
1948-1952.
(37) Huang, E. C.; Henion, J. D. J. Am. Soc. Mass. Spectrom. 1990, 1,
158-165.
(38) Bruins, A. P.; Covet, T. R.; Henion, J. D. Anal. Chem. 1987, 59,
2642-2646.
(40) Barany, G.; Merrifield, R. B. In The Peptides: Analysis, Synthesis,
Biology; E. Gross and F. Meienhofer, Eds.; Academic Press: New York,
1979; Vol. 2; pp 1- 284.
(39) Fields, G. B.; Tian, Z.; Barany, G. In Synthetic Peptides: A User’s
Guide; G. A. Grant, Ed.; W. H. Freeman: 1992; pp 130-136.

1414 J. Am. Chem. Soc., Vol. 120, No. 7, 1998
Miranda et al.
Scheme 5. Reaction Pathways of Mild Base Hydrolysis of LTEN p-Cresol Ester 11a
products was confirmed by ES-MS and RP-HPLC coelution
experiments with authentic samples.
were stable under these conditions and did not lead to recovery
of the target peptide (data not shown).
Application to Multi-Glu-Containing Peptides: Synthesis
of MRP14(89-114). HF-induced p-cresol ester formation and
subsequent side reactions induced by mild base are not confined
to the LTEN sequence. To further demonstrate the significance
of these findings with respect to obtaining homogeneous product
after HF treatment, we carried out a detailed study on the
synthesis and HF cleavage of MRP14(89-114) (19), also known
as neutrophil immobilization factor (NIF).41b,42 The sequence
contains four glutamyl residues that are sites for HF-induced
acylium ion formation (Table 2).
It is conceivable that the ester 11a partially cyclizes to a six-
membered ring system 17, with molecular weight of 457,
through nucleophilic attack by the nitrogen atom of the (i + 1)
residue. Further hydrolysis of 17 could produce LTEN (16a)
or the side chain connected δ-peptide LTE(N) 18 (Mr 475). The
latter product however was not detected. From the observed
product distributions in several hydrolysates of ester 11a, it is
clear that reaction route II via the imide 12 is the major
contributor in the hydrolysis, thus limiting the significance of
route III, if present at all.
Minimizing Side Reactions During Rearrangement of N
f O Acyl Migrations. Clearly, the use of NH4HCO3 buffer
(pH 8) for 12 h to reverse potential N f O shifts23,39,40 induced
a string of side reactions at p-cresol-esterified glutamyl residues,
generating both side chain amide analogues (i.e., GlufGln) of
the target peptide and several fragments. The purity and yield
of the final product was thus significantly reduced. To avoid
these fragmentation reactions, and recover LTEN, we applied
a H2O2-catalyzed hydrolysis of isolated ester 11a (Figure 2).
At pH 10, treatment with 2 equiv of H2O2 for 5 min produced
the target tetrapeptide 16a with minor amounts of dipeptide 15
formed. After treatment with 25 equiv H2O2 at pH 10 for 5
min, no contaminants were detected and homogeneous 16a was
readily recovered in excellent yield.41a
When p-cresol was replaced by p-thiocresol in the HF
cleavage procedure, the corresponding p-thiocresol ester (Mr
581) was isolated in yields similar to those of the p-cresol ester.
In a mixture of 1:1 p-cresol/p-thiocresol, the thioester 11b was
formed preferentially over the oxyester. Further LC/MS analysis
of both cleavage mixtures (data not shown) indicated again the
presence of “double” (Mr 687) and “triple” (Mr 811) p-thiocresol
adducts. In contrast to the facile peroxide ion-catalyzed
hydrolysis of the cresol esters, the double and triple adducts
Mr 2728.9
ASHEKMHEGDEGPGHHHKPGLGEGTP
19
The assembly of the peptide was performed on Pam resin
using in situ Boc/HBTU activation protocols.43 The peptide
was detached from the resin with anhydrous HF (0 °C, 1 h) in
the presence of 10% p-cresol, p-cresol/p-thiocresol (1:1), or
p-thiocresol. The LC/MS profiles of the three crude products
are shown in Figure 3.
(41) (a) It is not inconceivable that the proposed peroxide-catalyzed
hydrolysis of O-esters could lead to chain cleavage at Ser/Thr residues in
N f O acyl rearranged peptides. To examine this in more detail, we
synthesized and isolated N f O acyl rearranged human immunodeficiency
virus-1 proteinase, HIV-1 PR(81-99) (P. F. Alewood, unpublished results).
The product was then subjected to the described H2O2-catalyzed hydrolysis
conditions. No chain cleavage at Ser/Thr residues in HIV-1 PR(81-99) was
observed by HPLC/MS analysis and pure HIV-1 PR(81-99) was recovered
in quantitative yields. Thus reversal of the N f O acyl migration under
our proposed conditions occurs preferentially to peroxide-catalyzed cleavage.
(b) Odink, K.; Cerletti, N.; Brueggen, J.; Clerc, R. G.; Tarcsay, L.; Zwadlo,
G.; Gerhards, G.; Schlegel, R.; Sorg, C. Nature 1987, 330, 80-82.
(42) Watt, K. W. K.; Brightman, I. L.; Goetzl, E. J. Immunology 1983,
48, 79-86.
(43) Schno¨lzer, M.; Alewood, P.; Jones, A.; Alewood, D.; Kent, S. B.
H. Int. J. Pept. Protein Res. 1992, 40, 180-193.

p-Cresol As a ReVersible Acylium Ion ScaVenger
J. Am. Chem. Soc., Vol. 120, No. 7, 1998 1415
Figure 2. LC/MS analyses of catalyzed hydrolysis of isolated LTEN
p-cresol ester 11a in pH 10 Na2CO3/NaHCO3 buffer for 8 min using
(A) 2 equiv of H2O2 and (B) 25 equiv of H2O2. The products were
separated on a Vydac reversed phase C-18 (5 µm, 300 Å, 0.21 × 25
cm) column using a linear 0-60% buffer B gradient over 30 min at a
flow rate of 150 µL/min and directly infused into the mass spectrometer
using Turbo Ionspray. Full scan mass spectra were acquired over the
mass range of 300-1000 Da with a scan step of 0.1 Da.
Table 2. Molecular Mass Comparison of the p-Cresol Esters of
MRP14(89-114) (19), with Those Observed by LC/MS Analysis of
the Crude Material after HF/p-Cresol (9:1) Treatment of
Resin-Bound MRP14(89-114) Where X Indicates a p-Cresol Ester
on the Glutamyl Side Chain
Figure 1. LC/MS analyses of a buffered solution of p-cresol ester
11a (A) after 2 days in 0.1 M NH4HCO3 buffer (pH 8), (B) after 1 h
in 0.1 M Na2CO3/NaHCO3 buffer (pH 9.2), and (C) after 2 days in 0.1
M Na2CO3/NaHCO3 buffer (pH 9.2). The products were separated on
a Vydac reversed phase C-18 (5 µm, 300 Å, 0.21 × 25 cm) column
using a linear 0-60% buffer B gradient over 60 min at a flow rate of
150 µL/min and directly infused into the mass spectrometer using Turbo
Ionspray. Full scan mass spectra were acquired over the mass range of
200-1000 with a scan step of 0.1 Da.
With p-cresol as the sole scavenger (Figure 3A), four major
side products of the target peptide were detected. Four peaks
displayed an m/z value corresponding to the formation of a single
cresol ester (Mr 2819 (2729 + 90)) at each glutamyl residue,
20a-d (Table 2). All p-cresol esters were separated from the
target peptide and combined represented 21% of the total
products. Importantly, HF/p-cresol treatment of purified MRP14-
(89-114) produced a comparable LC/MS profile. Application
of the low-high HF cleavage procedure did not prevent the
formation of p-cresol and p-thiocresol adducts, since glutamyl
acylium ions can be generated from either the protected or free
glutamyl residues.
Cleavage using HF/p-thiocresol (9:1) significantly increased
the number of side products in the crude cleavage product as
observed by LC/MS analysis (Figure 3B). At least 12 side
products could be readily characterized, four peaks with a Mr
of 2835 corresponding to the four single p-thiocresol adducts
(2729 + 106 Da). Four peaks corresponding to a Mr of 2941
(2729 + 212) indicated the formation of four double p-thiocresol
adducts. On the basis of our earlier results with aminocaproic
acid, we propose that these double p-thiocresol adducts originate
from the formation of a ketenebisthioacetal at each glutamyl
residue rather than the formation of thioesters at two distinct
Glu sites. A further four peaks each displayed m/z values
corresponding to 2941 and 3065. Mr 3065 represents the
molecular weight of the NIF sequence containing a single trithio
ortho ester (2729 + 336). As is the case for the previously
described trithio ortho esters, these triple thiocresol adducts are
expected to release one thiocresol molecule (-124) under Turbo
Ionspray LC/MS conditions.44 Consequently, we believe that
the 2941 and 3065 m/z ions in these LC/MS peaks originate
mainly, if not exclusively, from the trithio ortho ester derivatives
of 19. After RP-HPLC purification of MRP14(89-114) all
cleavage side products combined represented 29% of the total
quantity of isolated material.
When equimolar amounts of p-cresol and p-thiocresol were
used in the HF cleavage (Figure 3C), the yield of side products
increased to 33%. The composition of the side products was
complex with molecular weights corresponding to p-cresol and
p-thiocresol esters (2729 + 90 and 2729 + 106), ketenebisthio-
acetals (2729 + 212), a combination of ketenebisthioacetals and
p-cresol ester (2729 + 302), and trithio ortho esters (2729 +
336).
Mild Base Treatment of Crude MRP14(89-114). Similar
to LTEN, mild base treatment (0.1 M NH4HCO3, pH 8.5) of
(44) Fractions from the preparative RP-HPLC separation of crude
thiocresol adducts were found to have a single mass of 3064 (triple thiocresol
adduct) by ES-MS without Turbo Ionspray. However, when LC/MS
conditions were simulated by employing Turbo Ionspray on those fractions,
both 2942 and 3064 m/z ions were observed.

1416 J. Am. Chem. Soc., Vol. 120, No. 7, 1998
Miranda et al.
Figure 4. (A) Theoretical isotope distribution for the molecular mass
region of a 3:1 mixture of MRP14(89-114) 19 and its single Glu f
Gln mutant. (B) The molecular mass region of the reconstructed FT-
ICR MS spectrum of purified (RP-HPLC) MRP14(89-114) obtained
from mild base treatment (0.1 M NH4HCO3 (pH 8.5) of crude HF/p-
cresol cleaved product.
Figure 3. LC/MS analyses of the crude HF cleavage products of resin-
bound MRP14(89-114) 19 using (A) HF/p-cresol (9:1), (B) HF/p-
thiocresol (9:1), and (C) HF/p-cresol/p-thiocresol (18:1:1). The crude
products were dissolved in aqueous 0.1% TFA (1 mg/mL) and then
loaded directly onto a Vydac reversed phase C-18 (5 µm, 300 Å, 0.21
× 25 cm) HPLC column. The products were separated using a linear
0-60% buffer B gradient over 60 min at a flow rate of 150 µL/min
and directly infused into the mass spectrometer using Turbo Ionspray.
Full scan mass spectra were acquired over the mass range of 400-
2000 with a scan step of 0.1 Da. The symbols in the figure indicate
peaks of the following molecular weights, where M ) 2728.9 Da: (O)
M + 90.1 Da; (b) M + 106.0 Da; (]) M + 212 Da; (*) M + 303 Da;
(4) M + 212 and M + 336 Da.
crude MRP14(89-114) (HF/p-cresol; 9:1) resulted in partial
amidation of the p-cresol esterified Glu residues (i.e., Glu to
Gln) and fragmentation. RP-HPLC separation of our target
peptide from these Glu to Gln mutants was not readily achieved.
ES-MS analysis of the purified product did not provide sufficient
resolution to unambiguously confirm presence of Glu to Gln
mutants (1 Da difference). However, the presence of Gln
mutants at each Glu site was unequivocally confirmed by Edman
sequencing of the MRP14(89-114) separated from the NH4HCO3
solution. Further, the high-resolution FT-ICR MS deconvoluted
spectrum of the same fraction (Figure 4) displayed an isotope
envelope corresponding to a (1:3) mixture of Glu to Gln mutants
over target peptide 19. Even though these impurities were
present in significant quantity, it is conceivable they would have
remained unnoticed after standard workup, purification, and
characterization by RP-HPLC and low-resolution ES-MS analy-
sis.
To examine the fragmentation pathways in detail, the p-cresol
adducts were isolated, combined, treated with 0.1 M NH4HCO3
pH 8.5 for 24 h, and then analyzed by LC/MS (Figure 6). RP-
HPLC analysis demonstrated that the target peptide coeluted
with several smaller fragments. Scheme 6 depicts the predicted
hydrolysis pathways for p-cresol-esterified peptides in a mild
base buffer, including formation of the free acid, amidation (Glu
Figure 5. LC/MS analyses of (A) separated and combined p-cresol
adducts 20a-d; (B) p-cresol adducts treated for 30 min with 5% H2O2
in 0.1 M NaHCO3 (pH 8.5). The solutions were acidified (pH 2) and
were loaded directly onto a Vydac reversed phase C-18 (5 µm, 300 Å,
0.21 × 25 cm) HPLC column. The products were separated using a
linear 0-60% buffer B gradient over 60 min at a flow rate of 150
µL/min and directly infused into the mass spectrometer using Turbo
Ionspray. Full scan mass spectra were acquired over the mass range of
400-2000 with a scan step of 0.1 Da.
to Gln), and fragmentation through an intermediate pyrrolidi-
none. In Table 3 we have listed the corresponding fragments
for peptides 20a-d (fragmentation pathway IIB) and compared
the theoretical molecular weights with the observed masses in
the LC-MS analysis (Figure 6). The experimental data are in
good agreement with the proposed fragmentation route IIB.
Furthermore, the fact that all four p-cresol esters (20a-d) are
readily hydrolyzed eliminates the possibility that they derive
from incomplete Bzl-deprotection at threonine or serine. Treat-
ment of the p-cresol adducts with a NaHCO3/Na2CO3 buffer

p-Cresol As a ReVersible Acylium Ion ScaVenger
J. Am. Chem. Soc., Vol. 120, No. 7, 1998 1417
Scheme 6. General Routes for the Degradation of Peptides Containing p-Cresol-Esterified Glutamyl Residues at pH 8
Table 3. Comparison of Predicted Fragments of Peptides 20a-d Formed after Base Treatment via Pathway IIB with Molecular Masses
Observed in the LC/MS Analysis of the Crude Product Shown in Figure 6
fragments of 20a-da
calcd Mr
found Mr
N-term.b
C-term.b
N-term.
C-term.
N-term.
C-term.
ASH
ASHEKMH
ASHEKMHEGD
ASHEKMHEGDEGPGHHHKPGLG
BKMHEGDEGPGHHHKPGLGEGTP
BGDEGPGHHHKPGLGEGTP
BGPGHHHKPGLGEGTP
BTGP
313.1
838.9
1140.2
2344.5
2415.6
1890.0
1588.7
384.4
313.2 (A)
838.5 (B)
1140.0 (D)
2344.1 (E)
2414.9 (H)
1889.5 (G)
1588.2 (F)
384.0 (C)
a B denotes pyroglutamic acid. b N- and C-term. refer to N- and C-terminal fragments, respectively.
p-thiocresol adduct mixture led to fragmentation according to
route IIB. The double and triple adducts observed in the
original mixture were stable to these conditions.
Efficient Recovery of Target Peptide MRP14(89-114). To
study the feasibility of recovery of target peptide, we initially
investigated the efficacy of the H2O2 catalysis conditions on
the isolated p-cresol and p-thiocresol adducts. The separated
p-cresol esters converted nearly quantitatively to the methionine-
oxidized MRP14(89-114) (19a), as illustrated in the LC/MS
profile (Figure 5B). By comparison the H2O2-catalyzed hy-
drolysis of the separated p-thiocresol adducts (esters, ketene-
bisthioacetals, and trithio ortho esters) was less effective with
the ketenebisthioacetals and trithio ortho esters proving resistant
to H2O2-catalyzed hydrolysis. Further studies on the crude HF/
p-cresol cleavage product of 19 were undertaken. Clean
hydrolysis of the O-esters was observed as shown in Figure 7.
Subsequent reduction of the oxidized methionine was quantita-
tive with N-mercaptoacetamide under a nitrogen atmosphere to
give highly homogeneous MRP14(89-114).
Figure 6. Detailed LC/MS analysis of p-cresol adducts 20a-d treated
with 0.1 M NH4HCO3 (pH 8.5) for 24 h. The solution was loaded
directly onto a Vydac reversed phase C-18 (5 µm, 0.21 × 25 cm) HPLC
column. The products were separated using a linear 0-60% buffer B
gradient over 60 min at a flow rate of 150 µL/min and directly infused
into the mass spectrometer using Turbo Ionspray. Full scan mass spectra
were acquired over the mass range of 400-2000 with a scan step of
0.1 Da. The letters in the figure indicate fragments of 20a-d from
Table 1, while the symbol (O) denotes peaks with a molecular weight
corresponding to the initial single p-cresol adduct.
Conclusion
The advent of chemical ligation techniques has opened new
routes to the chemical synthesis of proteins.45-52 These
advances have largely relied upon successful Boc SPPS to
(pH 9.2) avoided formation of Gln mutants but resulted in a
similar fragmentation pattern.
In a similar fashion the combined p-thiocresol adducts were
subjected to mild base treatment in a 0.1 M NaHCO3 buffer at
pH 8.5 for 24 h and analyzed by LC/MS (data not shown). As
with hydrolysis of the p-cresol adducts base treatment of the
(45) Dawson, P. E.; Muir, T. W.; Clark-Lewis, I.; Kent, S. B. H. Science
1994, 266 776-779.
(46) Liu, C. F.; Tam, J. P. J. Am. Chem. Soc. 1994, 116, 4149-4153.
(47) Dawson, P. E.; Kent, S. B. H. J. Am. Chem. Soc. 1993, 115, 7263-
7266.
(48) Englebretsen, D. R.; Garnham, B. G.; Bergman, D. A.; Alewood,
P. F. Tetrahedron Lett. 1995, 36, 8871-8874.

1418 J. Am. Chem. Soc., Vol. 120, No. 7, 1998
Miranda et al.
Auspep (Melbourne, Australia). p-Cresol, p-thiocresol, hydrogen
peroxide, 6-aminocaproic acid, and 4-aminobutyric acid were purchased
from Aldrich or Fluka (Sydney, Australia). HPLC grade acetonitrile
was purchased from Millipore-Waters (Sydney, Australia). 2-(1H-
Benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate
(HBTU) was purchased from Richelieu Biotechnologies (Quebec,
Canada). Deionized water was used throughout and was prepared with
a Milli-Q water purification system (Millipore-Waters). Screw-cap
glass peptide synthesis reaction vessels (20 mL) with sintered glass
filter frits were obtained from Embell Scientific Glassware (Queensland,
Australia). An all-Kel-F apparatus (Peptide Institute) was used for HF
cleavage. Argon, helium, and nitrogen (all ultrapure grade) were from
BOC Gases (Queensland, Australia). 1H NMR and 13C NMR spectra
were recorded on a Varian 300 MHz Gemini in CD3OD, and chemical
shifts are reported in parts per million (ppm) downfield from (CH3)4Si.
Reversed phase high-performance liquid chromatography (RP-HPLC)
was performed on a Waters 600E solvent delivery system equipped
with a 484 nm absorbance detector and recorded on an Apple Macintosh
computer using Model 600 software (Applied Biosystems Inc.). RP-
HPLC was performed using detection wavelengths of 214 and 230 nm
on Vydac C-18 analytical (5 µm, 0.46 cm × 25 cm) and preparative
C-18 (10 µm, 2.2 cm × 25 cm) columns, respectively. Chromato-
graphic separations were achieved using linear gradients of buffer B
in A (A ) 0.1% aqueous TFA; B ) 90% CH3CN, 10% H2O, 0.09%
TFA) over 60 min at a flow rate of 1 mL/min (analytical) and 8 mL/
min (preparative).
Figure 7. HPLC analysis of crude MRP14(89-114) HF/p-cresol
cleavage product (A) untreated, (B) treated with 5% H2O2 in 0.1 M
NaHCO3 (pH 8.5) for 10 min, and (C) followed by reduction with 15%
acetic acid and 10% N-mercaptoacetamide under N2 at 37 °C for 18 h.
The solutions were acidified (pH 2) and were loaded directly onto a
Vydac reversed phase C-18 (5 µm, 300 Å, 0.21 × 25 cm) HPLC
column. The products were separated using a linear 0-60% buffer B
gradient over 60 min at a flow rate of 1 mL/min.
Mass Spectrometry. Mass spectra were acquired on a PE-Sciex
API-III triple-quadrupole mass spectrometer equipped with an Ionspray
atmospheric pressure ionization source. Samples (10 µL) were injected
into a moving solvent (30 µL/min; 50:50 CH3CN/0.05% TFA) coupled
directly to the ionization source via a fused silica capillary interface
(50 µm i.d. × 50 cm length). Sample droplets were ionized at a positive
potential of 5 kV and entered the analyzer through an interface plate
and subsequently through an orifice (100-120 µm diameter) at a
potential of 80 V. Full scan mass spectra were acquired over the mass
range of 400-2000 Da with a scan step size of 0.1 Da. Molecular
masses were derived from the observed m/z values using the MacSpec
3.3 and Biomultiview 1.2 software packages (PE-Sciex Toronto,
Canada). Calculated theoretical monoisotopic and average masses were
determined using the MacBiospec program (PE-Sciex Toronto, Canada).
LC/MS runs were carried out using a linear gradient on a 140B ABI
dual syringe pump solvent delivery system and a Vydac reversed phase
C-18 (5 µm, 300 Å, 0.21 × 25 cm) column at a flow rate of 150 µL/
min. Samples (typically 5 µL of 1 mg/mL solution) were loaded
directly on the column, and the eluent was directly connected to the
mass spectrometer via a 30 cm, 75 µm i.d. fused silica capillary. The
application of Turbo Ionspray (5 L/min N2 at 500 °C) allowed the
introduction of the total eluent without splitting and loss in sensitivity.
Acquisition parameters were as described above. For tandem mass
spectrometry the molecular ions were bombarded in quadrupole-2 with
argon; the collision cell gas thickness was 3.5 × 1014 atoms/cm2, and
the collision energy was 10-20 eV. Quadrupole-3 was then scanned
to monitor product ions from m/z 50 to just above the parent ion
molecular mass in 3 s with a step size of 0.2 Da. Accurate mass
measurement of compound 5a was performed on a Kratos MS25 mass
spectrometer. Other high-resolution data were obtained on a Bruker
Spectrospin BioAPEX external-ion-source Fourier transform electro-
spray mass spectrometer at a magnetic field of 4.7 T. Typically samples
were dissolved in 50% aqueous methanol containing 1% acetic acid.
Peptide Synthesis. 6-Aminocaproic acid and 4-aminobutyric acid
were Boc protected and coupled to Merrifield resin using standard
procedures.53,54 LTEN and MRP14(89-114) were chemically synthe-
sized stepwise using in situ neutralization/HBTU activation protocols
for Boc chemistry as previously described.43 The synthesis was
performed on Boc-Asn(xanthyl)-OCH2 Pam (or Merrifield resin) and
Boc-Pro-OCH2 Pam resin, respectively. The following amino acid side
chain protection was used: Boc-Asn(xanthyl)-OH, Boc-Asp(O-cyclo-
hexyl)-OH, Boc-Glu(O-cyclohexyl)-OH, Boc-His(Nim-2,4-dinitro-
phenyl)-OH, Boc-Lys-(Nꢀ-2-chlorobenzyloxycarbonyl)-OH, Boc-Ser(O-
deliver highly homogeneous peptides typically 50+ residues.
In this work we have addressed the long-standing “acylium ion”
problem which has plagued many synthetic efforts in the past.
The new chemical insights describe not only the nature of the
problem more fully but have also facilitated the development
of novel chemical procedures that largely overcome the former
difficulties. To synthesize homogeneous peptides containing
long-chain acids or glutamyl residues, we recommend that
p-cresol be utilized as a reversible acylium ion scavenger during
HF cleavage. Where feasible, p-thiocresol should not be used
either alone or in combination with p-cresol. Furthermore, mild
base treatments on crude cleaved peptide (typically employed
to reverse potential N f O shifts) should be replaced by a mild
peroxide-catalyzed hydrolysis step. The implementation of this
novel concept should help facilitate further advances in the
chemical synthesis of highly homogeneous proteins.
Experimental Section
Materials and Methods. Merrifield resin (s.V.:4.1 mmol/g), NR-
tert-butoxycarbonyl (Boc)-L-amino acids, and reagents used during
chain assembly were peptide synthesis grade purchased from Auspep
(Melbourne, Australia), Novabiochem (San Diego, CA), and the Peptide
Institute (Osaka, Japan). Boc-L-amino acid-OCH2 (Pam) resin was
purchased from Applied Biosystems (Foster City, CA). Dichloro-
methane (DCM), diisopropylethylamine (DIEA), N,N-dimethylform-
amide (DMF), and trifluoroacetic acid (TFA) were obtained from
(49) Baca, M.; Muir, T. W.; Schno¨lzer, M.; Kent, S. B. H. J. Am. Chem.
Soc. 1995, 117, 1881-1887.
(50) Liu, C. F.; Tam, J. P. Proc. Nat. Acad. Sci. U.S.A. 1994, 91, 6584-
6588.
(51) Liu, C. F.; Rao, C.; Tam, J. P. J. Am. Chem. Soc. 1996, 118, 307-
312.
(52) Dawson, P. E.; Churchill, M. J.; Ghadiri, M. R.; Kent, S. B. H. J.
Am. Chem. Soc. 1997, 119, 4325-4329.
(53) Carpino, L. A. Acc. Chem. Res 1973, 6, 191-198.
(54) Gisin, B. F. HelV. Chim. Acta 1973, 56, 1476-1482.

p-Cresol As a ReVersible Acylium Ion ScaVenger
J. Am. Chem. Soc., Vol. 120, No. 7, 1998 1419
benzyl)-OH, and Boc-Thr(O-benzyl)-OH. Each residue was coupled
for 10 min, and coupling efficiencies determined by the quantitative
ninhydrin reaction.55 The average yield of chain assembly was 99.75%
for LTEN and 99.85% for MRP14(89-114). Before HF cleavage, the
dinitrophenyl groups were removed by thiolysis (20% â-mercapto-
ethanol, 10% DIEA, and 70% DMF; 2 × 2 h) prior to N-terminal Boc
deprotection with 100% TFA. The resin was then washed with DCM
and dried under nitrogen.
1.10 (m, 2H, CH2); 13C NMR (75 MHz, CD3OD, ppm) δ 142.43,
139.07, 131.92, 131.37, 79.14, 42.52, 42.20, 30.04, 28.79, 27.41, 22.88;
ES-MS Mr 467.0, calcd for C27H33NS3, 467.2 (monoisotopic).
1,1-Bis(4-methylphenyl)sulfanyl-6-aminohex-1-ene (8): 1H NMR
(300 MHz, CD3OD, ppm) δ 7.12 (s, 4H, H-ar), 7.09 (s, 4H, H-ar),
6.17 (t, J ) 6.5 Hz, 1H, H-CdC), 2.91 (m, 2H, CH2-N), 2.46 (m, 2H,
CH2-CdN), 2.31 (s, 3H, CH3), 2,30 (s, 3H, CH3), 1.66 (m, 2H, CH2),
1.50 (m, 2H, CH2); 13C NMR (75 MHz, DMSO-d6, ppm) δ 142.78
(CH), 137.43 (C), 136.58 (C), 131.36 (CH), 129.95 (CH), 129.82 (CH),
129.78 (CH), 129.00 (C), 38.56 (CH2), 30.24 (CH2), 26.66 (CH2), 25.31
(CH2), 20.72 (CH3), 20.64 (CH3); ES-MS Mr 343.1, calcd for C20H25-
NS2, 343.1 (monoisotopic).
HF Cleavage. Peptide resin (300 mg) was treated at the given
temperature with 10 mL of HF/p-cresol (9:1 (v/v)), HF/p-thiocresol
(9:1 (v/v)), or HF/p-cresol/p-thiocresol (18:1:1 (v/v)). After evaporation
of the HF, the crude product was precipitated and washed with cold
diethyl ether (2 × 10 mL), dissolved in 20% aqueous acetic acid (5
mL), and lyophilized after aqueous dilution.
Arginine-6-aminocaproic acid 4′-methylphenyl ester, TFA salt
1
(9a): H NMR (300 MHz, CD3OD, ppm) δ 6.95 (d, 2H, H-ar), 6.76
1
4-Aminobutanoic acid 4′-methylphenyl ester, TFA salt (2a): H
(d, 2H, H-ar), 3.77 (t, 1H, CR), 3.16 (m, 3H, CH2), 3.04 (m, 3H, CH2),
2.42 (t, 2H, CH2), 2.13 (m, 2H, CH3), 1.72 (m, 4H, CH2), 0.47 (m, 2H,
CH2); 13C NMR (75 MHz, CD3OD, ppm) δ 173.52 (CdO), 170.00
(CdO), 158.71 (C-ar), 149.96 (C-ar), 136.70 (C-ar), 130.83 (C-ar),
122.36 (C-ar), 54.11 (CH), 41.68 (CH2), 39.97 (CH2), 39.74 (CH2),
32.17 (CH2), 29.17 (CH2), 25.63 (CH2), 25.56 (CH2), 25.39 (CH2), 20.84
(CH3); ES-MS Mr 377.2, calcd for C19H31N5O3, 377.2 (monoisotopic).
NMR (300 MHz, CD3OD, ppm) δ 7.19 (d, J ) 8.5 Hz, 2H, H-ar),
6.95 (d, J ) 8.5 Hz, 2H, H-ar), 3.05 (t, J ) 7.8 Hz, 2H, CH2-N), 2.73
(t, J ) 7.1 Hz, 2H, CH2 -CdO), 2.33 (s, 3H, CH3), 2.04 (m(5), 2H,
CH2); 13C NMR (75 MHz, CD3OD, ppm) δ 173.03 (CdO), 150.04
(C-i-ar), 136.93 (C-p-ar), 131.00 (C-m-ar), 122.44 (C-o-ar), 40.09
(CH2-N), 31.77 (CH2-CO), 23.84 (CH2), 20.99 (CH3); ES-MS Mr
193.1, calcd for C11H15NO2, 193.1 (monoisotopic).
6-Aminohexanoic acid 4′-methylphenyl ester, TFA salt (2b): 1H
NMR (300 MHz, CD3OD, ppm) δ 7.18 (d, J ) 8.7 Hz, 2H, H-ar),
6.93 (d, J ) 8.5 Hz, 2H, H-ar), 2.94 (t, J ) 7.8 Hz, 2H, CH2-N), 2.60
(t, J ) 7.1 Hz, 2H, CH2 -CdO), 2.32 (s, 3H, CH3), 1.74 (m(7), 4H,
2 × CH2), 1.51 (m, 2H, CH2); 13C NMR (75 MHz, CD3OD, ppm) δ
173.87 (CdO), 150.00 (C-i-ar), 136.69 (C-p-ar), 130.85 (C-m-ar),
122.34 (C-o-ar), 40.56 (CH2-N), 34.61 (CH2-CO), 28.29 (CH2), 26.86
(CH2), 25.34 (CH2), 20.84 (CH3), ES-MS Mr 221.1, calcd for C13H19-
NO2, 221.1 (monoisotopic).
Arginine-6-aminocaprothioic acid S-4′-methylphenyl ester, TFA
1
salt (9b): H NMR (300 MHz, CD3OD, ppm) δ 7.07 (s, 4H, H-ar),
3.75 (t, 1H, CR), 3.11 (m, 4H, CH2), 3.02 (m, 2H, CH2), 2.53 (t, 2H,
CH2), 2.16 (s, 3H, CH3), 1.68 (m, 4H, CH2), 1.48 (m, 2H, CH2); 13C
NMR (75 MHz, CD3OD, ppm) δ 199.07(CdS), 169.92 (CdO), 158.71
(Cꢀ), 141.07 (C-ar), 135.61 (C-ar), 130.98 (C-ar), 125.50 (C-ar), 54.08
(CR), 41.69 (CH2), 41.44 (CH2), 39.75 (CH2), 29.69 (CH2), 26.09 (CH2),
25.39 (CH2), 21.27 (CH3); ES-MS Mr 393.1, calcd for C19H31N5O2S,
393.2 (monoisotopic).
Comparison of Hydrolysis Rates of 9a and 9b. Fifty microliters
of a 10 mg/mL solution of the ester in water was diluted with 450 µL
of a pH 9.2 buffer (Na2CO3/NaHCO3), and 12.5 µL of an 0.3%
hydrogen peroxide solution in water was immediately added. Aliquots
of 50 µL were taken out after specific time intervals (1, 2, 4, and 8
min), and the hydrolysis was quenched with 100 µL of 0.1 M HCl.
The solutions were then analyzed by RP-HPLC.
6-Amino-1-(5′-methyl-2′-hydroxyphenyl)hexan-1-one, TFA salt
(4b): 13C NMR (75 MHz, CD3OD, ppm) δ 207.98 (CdO), 161.22 (C-
ar), 138.31 (C-ar), 131.09 (C-ar), 129.56 (C-ar), 120.40 (C-ar), 118.85
(C-ar), 38.80 (CH2-N), 34.60 (CH2-CdN), 28.43 (CH2), 27.03 (CH2),
24.65 (CH2); ES-MS Mr 221.1, calcd for C13H19NO2, 221.1 (monoiso-
topic).
2-(4,5′-Dihydro-3H-pyrrol-2′-yl)-4′-methylphenol, TFA salt (5a):
1H NMR (300 MHz, CD3OD, ppm) δ 7.613 (d, J ) 2.0 Hz, 1H, H-ar),
7.50 (dd, J1 ) 2.0 Hz, J2 ) 8.6 Hz, 1H, H-ar), 7.03 (d, J ) 8.6 Hz,
1H, H-ar), 4.18 (t, J ) 7.7 Hz, 2H, CH2-N), 3.61 (t, J ) 8.3 Hz, 2H,
CH2-CdN), 2.34 (s, 3H, CH3-Phe), 2.33 (m(5), 2H, CH2-CH2-
CH2); 13C NMR (75 MHz, CD3OD, ppm) δ 183.22 (CdN), 159.84
(C-ar), 140.77 (C-ar), 134.40 (C-ar), 131.62 (C-ar), 117.93 (C-ar),
113.67 (C-ar), 55.02 (CH2-N), 36.38 (CH2-CdN), 20.31 (CH3), 20.23
(CH2-CH2-CH2); ES-MS Mr 175.1, accurate mass EI obsd 175.0997,
calcd for C11H13NO, 175.0997 (monoisotopic).
LTEN glutamyl p-cresol ester (11a): 1H NMR (300 MHz, CD3OD,
ppm) δ 7.18 (d, J ) 8.6 Hz, 2H, H-ar), 6.98 (d, J ) 8.6 Hz, 2H, H-ar),
4.72 (t, 1H, R-Glu), 4.58 (dd, 1H, R-Asn), 4.41 (d, 1H, R-Thr), 4.16
(m, 1H, â-Thr), 3.99 (t, 1H, R-Leu), 2.78 (m, 2H, â-Asn), 2.69 (t, 2H,
γ-Glu), 2.33 (s, 3H, CH3-Ar), 2.29 (m, 1H, â-Glu), 2.02 (m, 1H,
â-Glu), 1.70 (m, 3H, â-Leu + γ-Leu), 1.24 (d, 3H, CH3-Thr), 0.95
(m, 6H, CH3-Leu); 13C NMR (75 MHz, CD3OD, ppm) δ 174.92, 174.17,
173.33, 172.92, 171.87, 171.05, 150.02, 136.66, 130.81, 122.48, 68.42,
60.45, 53.47, 52.93, 50.69, 41.60, 37.64, 31.04, 28.36, 25.39, 23.16,
21.93, 20.84, 20.18; ES-MS Mr 565.4, high-resolution FT-ICR MS
565.2827, calced for C26H39N5O9, 565.2748 (monoisotopic).
LTEN glutamyl p-thiocresol ester (11b): 1H NMR (300 MHz,
CD3OD, ppm) δ 7.34 (s, 4H, H-ar), 4.60 (m, 1H, R-Glu), 4.49 (m, 1H,
R-Asn), 4.45 (m, 1H, R-Thr), 4.12 (m, 2H, R-Leu + â-Thr), 2.85 (m,
2H, â-Asn), 2.79 (m, 2H, γ-Glu), 2.38 (s, 3H, CH3-Ar), 2.27 (m, 1H,
â-Glu), 2.03 (m, 1H, â-Glu), 1.69 (m, 3H, â-Leu + γ-Leu), 1.23 (d,
3H, CH3-Thr), 0.88 (m, 6H, CH3-Leu); 13C NMR (75 MHz, CD3OD,
ppm) δ 203.03, 175.91, 172.91, 171.73, 171.29, 142.00, 135.62, 131.33,
123.95, 68.27, 60.11, 53.19, 52.79, 51.64, 40.94, 39.61, 37.80, 27.80,
24.87, 22.89, 22.01, 21.86, 19.85; ES-MS Mr 581.1, calcd for
C26H39N5O8S, 581.3 (monoisotopic).
(1′-[2′-Hydroxy-5′-methylphenyl])-2-azacyclohept-1-ene, TFA salt
1
(5b): H NMR (300 MHz, CD3OD, ppm) δ 7.41 (s, 1H, H-ar), 7.16
(dd, 1H, Jo ) 8.6 Hz, Jm ) 2.3 Hz, H-ar), 6.68 (d, J ) 8.7 Hz, 1H,
H-ar), 3.81 (m, 2H, CH2), 3.12 (m, 2H, CH2), 2.23 (s, 3H, CH3), 1.94
(m, 2H, CH2), 1.72 (m, 4H, CH2); 13C NMR (75 MHz, CDCl3, ppm)
δ 183.27 (CdN), 173.24 (C-ar), 138.48 (C-ar), 130.90 (C-ar), 129.70
(C-ar), 124.72 (C-ar), 123.49 (C-ar), 46.97 (CH2-NdC), 31.50 (CH2),
28.42 (CH2), 28.10 (CH2), 24.43 (CH2), 20.72 (CH3); ES-MS Mr 203.1,
calcd for C13H17NO, 203.1 (monoisotopic).
6-Aminohexanethioic acid S-4′-methylphenyl ester, TFA salt (6):
1H NMR (300 MHz, CD3OD, ppm) δ 7.26 (s, 4H, H-ar), 2.92 (t, J )
7.5 Hz, 2H, CH2-N), 2.71 (t, J ) 7.2 Hz, 2H, CH2-CdN), 2.37 (s,
3H, CH3), 1.71 (m, 4H, 2 × CH2), 1.45 (m, 2H, CH2); 13C NMR (75
MHz, CD3OD, ppm) δ 199.54 (CdS), 141.20 (C-ar), 135.75 (C-ar),
131.12 (C-ar), 125.81 (C-ar), 43.92 (CH2-N), 40.67 (CH2-CdS), 28
(CH2), 27.6 (CH2), 26.20 (CH2), 21.42 (CH3); ES-MS Mr 236.9, calcd
for C13H19NOS, 237.1 (monoisotopic).
Hydrolysis of the Esters 11a and 11b. The following buffers were
used: (1) 0.1 M NH4HCO3 (pH 8); (2) 0.1 M NaCO3/NaHCO3 buffered
to pH 9.2; (3) 0.1 M NaCO3/NaHCO3 buffered to pH 10. Batches of
500 µg of ester were dissolved in 500 µL of buffer. After the required
time, the pH was adjusted to 2 using TFA and the solution was analyzed
by LC/MS. For the peroxide-catalyzed experiments, hydrogen peroxide
(0.3% in water) was added to the buffered solution.
1,1,1-Tris-(4′-methylphenyl)sulfanyl-6-aminohexane, TFA salt
1
(7): H NMR (300 MHz, CD3OD, ppm) δ 7.51 (d, J ) 7.6 Hz, 6H,
Human MRP14(89-114) (19). Five hundred micrograms of the
crude product (HF/p-cresol (9:1), 0 °C, 1 h) was dissolved in buffer A
and purified by preparative RP-HPLC. ES-MS: Mr 2728.7 (av); high-
resolution FT-ICR MS 2727.3019 (monoisotopic), calcd for
H-ar), 7.17 (d, J ) 7.6 Hz, 6H, H-ar), 2.81 (m, 2H, CH2-N), 2.35 (s,
9H, CH3), 1.75 (m, 2H, CH2), 1.65 (m, 2H, CH2), 1.52 (m, 2H, CH2),
(55) Sarin, V.; Kent, S. B. H.; Tam, J. P.; Merrifield, R. B. Anal. Biochem.
1981, 117, 147-157.
C114H171N38O39S, 2727.2208 (monoisotopic).

1420 J. Am. Chem. Soc., Vol. 120, No. 7, 1998
Miranda et al.
p-Cresol Esters of MRP14(89-114) (20a-d). The crude peptide
(HF/p-cresol (9:1), 0 °C, 1 h) was dissolved in buffer A, and preparative
RP-HPLC was performed. The separated p-cresol esters were then
combined and lyophilized; yield 21% (w/w). ES-MS: Mr 2818.8, calcd
for C121H176N38O39S, 2819.0 (av).
p-Thiocresol Esters of MRP14(89-114). The crude peptide (HF/
p-thiocresol (9:1), 0 °C, 1 h) was dissolved in buffer A, and preparative
RP-HPLC (C-18) was performed. The separated p-thiocresol esters
were then combined and lyophilized; yield 29% (w/w). ES-MS: Mr
2834.5, calcd for C121H176N38O38S2, 2835.1 (av).
Hydrolysis of p-Cresol Esters 20a-d. Five hundred microgram
batches of the MRP14(89-114) p-cresol esters were treated with the
following solutions: (a) 500 µL of 0.1 M NaHCO3 (pH 8.5) for 24 h
at room temperature, (b) 500 µL of 0.1 M NH4HCO3 (pH 8.5) for 24
h at room temperature, (c) 500 µL of 0.1 M NaOH (pH 13) for 1 h at
room temperature, and (d) 500 µL of 0.1 M NaHCO3 (pH 8.5)
containing 25 equiv of H2O2 for 5 min at room temperature. All
reactions were monitored by LC/MS.
Human MRP14(89-114) Glutamyl p-Cresol Ester, NH4HCO3
Treated. Five hundred micrograms of the MRP14(89-114) cresol ester
was then treated with 0.1 M NH4HCO3 (pH 8.5) and purified by RP-
HPLC. ES-MS: Mr 2727.2 (av); high-resolution FT-ICR MS, 2726.3033
(monoisotopic); calcd for C114H171N39O38S, 2726.2368 (monoisotopic).
Human MRP14(89-114) Ketenebisthioacetals. Five hundred
micrograms of the crude product (HF/p-thiocresol (9:1), 0 °C, 1 h)
was dissolved in buffer A and purified by RP-HPLC. ES-MS: Mr
2940.3, calcd for C128H182N38O37S3, 2941.3 (av).
Human MRP14(89-114) Trithio Ortho Ester. Five hundred
micrograms of the crude product (HF/p-thiocresol (9:1), 0 °C, 1 h)
was dissolved in buffer A and purified by preparative RP-HPLC. ES-
MS: Mr 3065.0, high-resolution FT-ICR MS, 3063.3052, calcd for
C135H190N38O37S4, 3063.3037 (monoisotopic).
Reduction of MRP14(89-114)Met94[O] (19a). One milligram of
MRP14(89-114)Met94[O], 19a, was dissolved in a solution containing
15% (v/v) acetic acid and 10% (v/v) N-mercaptoacetamide. The
mixture was placed under a N2 atmosphere and then incubated at 37
°C for 18 h. The conversion to 19 was quantitative as determined by
RP-HPLC analysis. After isolation and lyophilization, the yield was
found to be 80% (w/w).
Acknowledgment. We gratefully acknowledge the assistance
of Rick Willis, The Australian Institute of Marine Science, and
Graham MacFarlane, University of Queensland, for FT-ICR MS
and EI MS measurements, Trudy Bond and Ann Atkins for
amino acid analysis and Edman sequencing. L. P. Miranda was
supported by an Australian postgraduate research award (APRA)
scholarship from the Australian Government.
JA973322K