Tyrosine analogues as alternative substrates for protein tyrosine kinase Csk: Insights into substrate selectivity and catalytic mechanism

Article abstract of DOI:10.1016/S0968-0896(00)00053-5

Protein tyrosine kinases are critical enzymes in cell signal transduction but relatively little is known about the molecular recognition of the tyrosine substrate by these enzymes. Details of tyrosine substrate specificity within the context of a short peptide were investigated for protein tyrosine kinase Csk. It was found that aryl ring functional group substitutions the size of methyl group or smaller were generally well tolerated by the protein tyrosine kinase Csk whereas larger groups caused a decline in substrate efficiency. Extension of the phenol from the peptide backbone by a single methylene was acceptable for phosphorylation whereas removal of a methylene nearly abolished reactivity. Only the L-tyrosine derivative was processed. A negative charge ortho to the phenol hydroxyl was incompatible with substrate reactivity, consistent with previous pH rate profiles which indicated the importance of the neutral phenol. Overall, these studies confirmed the interpretation of a previous linear free energy relationship analysis which suggested that the enzyme followed a dissociative transition state mechanism. Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(00)00053-5

Bioorganic & Medicinal Chemistry 8 (2000) 1263±1268
Tyrosine Analogues as Alternative Substrates for Protein Tyrosine
Kinase Csk: Insights into Substrate Selectivity and Catalytic
Mechanism
Kyonghee Kim, Keykavous Parang, Ontario D. Lau and Philip A. Cole*
Department of Pharmacology and Molecular Sciences, Johns Hopkins University School of Medicine,
7
25 N. Wolfe St., Baltimore, MD 21205, USA
Received 26 October 1999; accepted 17 January 2000
AbstractÐProtein tyrosine kinases are critical enzymes in cell signal transduction but relatively little is known about the molecular
recognition of the tyrosine substrate by these enzymes. Details of tyrosine substrate speci®city within the context of a short peptide were
investigated for protein tyrosine kinase Csk. It was found that aryl ring functional group substitutions the size of methyl group or smaller
were generally well tolerated by the protein tyrosine kinase Csk whereas larger groups caused a decline in substrate eciency. Extension
of the phenol from the peptide backbone by a single methylene was acceptable for phosphorylation whereas removal of a methylene
nearly abolished reactivity. Only the l-tyrosine derivative was processed. A negative charge ortho to the phenol hydroxyl was incom-
patible with substrate reactivity, consistent with previous pH rate pro®les which indicated the importance of the neutral phenol. Overall,
these studies con®rmed the interpretation of a previous linear free energy relationship analysis which suggested that the enzyme followed
a dissociative transition state mechanism. # 2000 Elsevier Science Ltd. All rights reserved.
Introduction
on the aromatic ring, that could enhance or detract from
binding or catalysis.
Protein tyrosine kinases catalyze the transfer of the g-
phosphoryl group from ATP to the phenol side chain of
protein tyrosine residues (Fig. 1). Interest in the biology
of protein tyrosine kinases has intensi®ed in recent
years. These enzymes have emerged as key signalling
switches that can a€ect cell growth and di€erentiation.1
Speci®city of protein kinases for their target substrates
is intimately associated with the regulation and
The catalytic mechanism of protein tyrosine kinase Csk,
an enzyme important in modulating immune system acti-
vation and neuro-development, has been suggested to
involve a dissociative transition state with relatively minor
5
nucleophile involvement. Among the experiments used to
support a dissociative transition state include the analysis
of linear free energy relationships (measurements of bnuc
and bleaving group) using a series of peptides containing
¯uorotyrosine derivatives as substrates.5 The observa-
tion was made that ¯uoro-substitution of the aromatic
ring of the substrate, while signi®cantly altering the
2
dysregulation of cell signal transduction pathways.
a,b
De®ning the characteristics of natural and unnatural
molecules that lead to protein tyrosine kinase active site
recognition should be instructive in the generation of
3
speci®c protein kinase inhibitors. While it is now well-
phenol pK , had minimal impact on recognition and
a
recognized that the nucleotide binding site of protein
kinases can be exploited to generate highly speci®c pro-
tein kinase inhibitors,4 the peptide/protein substrate
binding site could in principle provide an even greater
source of speci®city for the design of protein kinase
inhibitors targeted to an individual protein kinase. Rela-
tively little is known in a general sense about features of
the tyrosine phenol side-chain, or potential substitutions
phosphorylation eciency by Csk. Steric and electronic
impacts of the ¯uorine atoms on enzyme active site bind-
ing were assumed to be minor. In addition, the para-
doxical observation was made that the phenoxide anion
species of the ¯uorotyrosine containing substrates was
found to be less enzymatically active than the corre-
sponding neutral phenol. Electrostatic repulsion of the
anions within the enzyme active site in the transition state
was assumed to be the dominant factor here.
A systematic investigation is lacking of the tolerance of
tyrosine kinases toward unnatural tyrosine derivatives
*
7
Corresponding author. Tel.:+1-410-614-05440; fax:+1-410-614-
717; e-mail: pcole@jhmi.edu
0968-0896/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00053-5

1
264
K. Kim et al. / Bioorg. Med. Chem. 8 (2000) 1263±1268
Figure 3. Tyrosine phenol-lyase (TPL) reaction used to make tyrosine
analogues.
Figure 1. Protein tyrosine kinase reaction.
with varying features of size, polarity, charge and stereo-
chemistry. To extend our understanding of substrate
recognition by protein kinases, we undertook a more
detailed investigation of the e€ects of ring substitution
and phenol positioning on protein tyrosine kinase Csk
catalytic processing, and report the results herein.
tives were used in automated solid phase peptide synth-
esis to generate the desired peptides, puri®ed by reversed
phase HPLC and characterized by electrospray mass
spectrometry. To produce the 3-hydroxymethyltyrosine
(3) containing peptide, it proved necessary to alter the
standard resin-cleavage/deprotection conditions (omit-
ting the anisole from tri¯uoroacetic acid/water/anisole).
Agents such as anisole were shown to undergo acid-
catalyzed attack at the benzylic position of 3 in the pep-
tide. To prepare the salicylic acid tyrosine (10) containing
peptide, prior to cleavage from the resin and deblocking
the peptide containing 3 was oxidized with tetrapropyl
Results and Discussion
A series of tyrosine analogues with varied steric, elec-
tronic and stereochemical features was incorporated in
place of tyrosine into a short, well-characterized Csk
substrate peptide sequence Glu-Asp-Asn-Glu-Tyr-Thr-
7
ammonium perruthenate and mCPBA (Fig. 4). A com-
bination of the MS and NMR data of the puri®ed peptide
was used to con®rm product identity in this case.
�
1
Ala (K ꢀ5 mM, k ꢀ20 min ). The tyrosine deriva-
m
cat
tives used in this study were prepared either using the
enzyme, tyrosine phenol-lyase (2±5, 9), or obtained from
commercially available sources (6±8) (Fig. 2). Tyrosine
phenol-lyase has been found to be generally useful in pre-
paring l-tyrosine derivatives using the corresponding
Each peptide was tested as a protein tyrosine kinase Csk
alternative substrate using an established HPLC assay. To
verify the structural identity of the product of the phos-
phorylation reaction, the reaction products were separated
from the unphosphorylated peptides and the phosphoryl-
ation state was con®rmed by mass spectrometric analysis.
We propose that, given the increased availability and e-
ciency of modern mass spectrometry (MALDI, ESI) in
6
phenols, ketoacid and ammonia (Fig. 3). These synthetic
tyrosine analogues were converted to the corresponding
N-Fmoc derivatives without side-chain protection using
standard conditions. The corresponding Fmoc deriva-
Figure 2. Tyrosine analogues investigated in this work.

K. Kim et al. / Bioorg. Med. Chem. 8 (2000) 1263±1268
1265
Figure 4. Conversion of salicylic alcohol containing peptide to the salicylic acid derivative.
analyzing biopolymers, this should become a routine pro-
cedure for investigators carrying out kinase assays for the
the phenol and the peptide backbone for kinase proces-
sing? (vi) What are the e€ects of introducing a negative
charge ortho to the phenol hydroxy?
®rst time with unnatural tyrosine derivatives, to avoid
artifacts associated with less direct approaches. All pep-
tides which were Csk substrates showed linear kinase
activity versus time and enzyme concentration, allowing
determination of the catalytic eciency, k /K , and the
Steric bulk
Placement of a methyl group ortho to the phenol
hydroxyl as present in 2 had a modest 2.5-fold e€ect on
decreasing catalytic processing eciency compared to
tyrosine and the even bulkier substituent hydroxymethyl
found in 3 decreased activity as a kinase substrate by 4-
fold compared to tyrosine (see Table 1). Meta substit-
ution of the tyrosine aromatic ring with a methyl group
represented in 4 led to minimal impact on processing,
whereas OMe substitution (peptide containing 5) did
cause an 8-fold drop in processing compared to tyr-
osine. Since it was uncertain whether rate reduction
with OMe resulted from the electronic e€ect of oxygen
or the increased steric e€ect due to a larger group, the
ethyl substituted analogue 9 was evaluated. The ethyl
derivative was similar to the OMe analogue as a Csk
substrate, suggesting that electronic e€ects of the oxy-
cat
m
results are shown in Table 1.
The range of substituents allowed us to address several
questions: (i) How would increased steric bulk ortho or
meta to the phenol hydroxyl in¯uence catalysis? (ii) How
would hydrogen bond donating/accepting groups ortho to
the phenol a€ect recognition or catalysis? (iii) Can a
benzylic hydroxyl change phosphorylation site selectivity?
(iv) How is a peptide containing d-tyrosine handled by
the enzyme? (v) How important is the spacing between
Table 1. Catalytic eciency of peptides containing tyrosine ana-
logues as Csk tyrosine kinase substratesa
Tyrosine analogue
Relative kcat/K
m
8
gen atom at the meta position were non-disruptive.
l-Tyrosine (1)
100%
38%
23%
90%
12%
3
3
2
2
-Methyltyrosine (2)
Taken together, these results suggest that the steric
contributions of ¯uorine atoms should have minimal
-Hydroxymethyltyrosine (3)
-Methyltyrosine (4)
-Methoxytyrosine (5)
e€ect on the b
or bleaving group measured in previous
nuc
studies.5 Thus, a dissociative mechanism is still the
favored interpretation of the small bnuc reported pre-
viously. However, kinase processing is somewhat sensitive
to (and disrupted by) substrate steric e€ects for larger
substituents at the ortho position and to a lesser extent
at the meta position of tyrosine in peptide substrates. It
is also noteworthy that no enhancement above the rate
of tyrosine processing was observed with any of the
,8
d-Tyrosine (6)
l-Hydroxyphenylglycine (7)
Homotyrosine (8)
<0.5%
0.75%
26%
21%
<0.5%
2-Ethyltyrosine (9)
3-Carboxytyrosine (10)
aThe relative eciencies (kcat/K
) are referenced to substrate peptide
m
with l-tyrosine and measured as described in Experimental. Estimated
standard errors for the values (given as percent of maximal activity) are
‹20%.

1
266
K. Kim et al. / Bioorg. Med. Chem. 8 (2000) 1263±1268
derivatives evaluated in this study (which contrasts
meta-¯uorotyrosine behavior, which was a 2±3-fold
more ecient substrate compared to tyrosine ). Thus, it
found that both Abl and Src were capable of phos-
phorylating d-tyrosine residues at reasonable enzymatic
rates. This work employed an unnatural C-terminal
5
10
is unlikely that increased stabilization of the transition
state complex can be achieved by enhancing additional
active site interactions compared to a standard tyrosine.
tyrosine linkage making its insight into standard sub-
strate recognition unclear. Investigation of the EGF
receptor kinase revealed that a peptide containing d-
tyrosine (6) was reported to be a fairly ecient sub-
strate, although the phosphorylated peptide product
was not rigorously characterized.12 In the case of Csk
and the standard substrate heptapeptide containing d-
tyrosine (6), no detectable (<200-fold l-tyrosine phos-
phorylation rate) phosphorylation occurred. It is poss-
ible that protein tyrosine kinases are divergent in their
ability to discriminate between d- and l-tyrosine resi-
dues in the context of peptides. This would be somewhat
surprising given the high degree of similarity in the cata-
lytic core of protein tyrosine kinases and further work is
needed with a variety of systems to clarify this issue.
ortho Hydrogen bond donor
The presence of the hydroxymethyl ortho to the phenol
hydroxy in 3 has the capacity to disrupt hydrogen
bonding or augment hydrogen bonding to active site
groups. In particular, 3 is expected to maintain an
internal hydrogen bond between the phenol oxygen and
the CH OH oxygen. The practical e€ect of this group
2
was minimal (less than 2-fold compared to the corre-
sponding methyl analogue 2), suggesting that potential
H-bonding between the hydroxymethyl and the phenol
oxygen or active site groups does not interfere. Alter-
natively, the benzylic hydroxy group could ®ll space
normally occupied by a water molecule in this position
and the net e€ects on catalysis would be minor. It is
worth noting that the benzylic hydroxy group could be
converted into a leaving group if it were protonated by
an active site acidic group. In principle a peptide con-
taining 3 could therefore cause irreversible inactivation
of Csk. However, no time-dependent loss of Csk kinase
activity was detected with this peptide, arguing against
the presence of such acid activation and/or a reactive
enzyme nucleophile nearby. In future work, it will be
worth investigating whether a better leaving group such
as ¯uoride can enhance enzyme-induced alkylation.9
Homotyrosine (8) and hydroxyphenylglycine (7) studies
In a dissociative mechanism, the transition state bond
formation between the nucleophilic hydroxyl and the
ATP g-phosphorus is minimal. It is therefore expected
that the nucleophilic hydroxyl may be positioned quite
far away from the g-phosphorus in the transition
state.13 It might be expected that signi®cant latitude
would be tolerated in the distance between peptide
backbone and attacking phenol. Peptides containing l-
homotyrosine (8) and l-hydroxyphenylglycine (7) were
prepared to examine this issue. Both residues were
acceptable as substrates. The homotyrosine (8) contain-
ing peptide was only 4-fold down compared to tyrosine
as a substrate, but hydroxyphenylglycine (7) was over
Does 3 get phosphorylated only on the phenol?
1
00-fold worse. Thus, positioning of the hydroxy group
The capacity for the benzylic hydroxy function to
receive a phosphate from ATP in the presence of the
standard phenol was unknown. Modelling suggested
that the ortho hydroxymethyl could extend far enough
from the peptide backbone to simulate typical tyrosine
phosphorylation without severe strain. The major pro-
duct of phosphorylation was revealed to be the phenolic
phosphate (UV, NMR). HPLC analysis of the reaction
mixture revealed no evidence (less than 10% of the rate of
phenol phosphorylation) of benzylic site phosphorylation.
This contrasts with the ®ndings of promiscuity observed
in a system reported by Lawrence and colleagues,10
although the C-terminal location of the benzyl hydroxy
group in that instance could be responsible for enhanced
tolerance of unnatural substrates (see below). The higher
in the transition state is important in the enzyme cata-
lyzed reaction. It has been suggested that the basis for
protein tyrosine kinase selectivity versus serine kinases
is primarily related to the distance between the hydroxyl
and the peptide backbone.14 The ®ndings with peptide
containing hydroxyphenylglycine (7) shown to be a
poor Csk substrate would tend to support such a pro-
posal, although orienation could also be important.
How do we rationalize these ®ndings with a dissociative
transition state of enzymatic phosphoryl transfer? Two
broad possibilities include: (i) the binding of the peptide
containing hydroxyphenylglycine (7) to Csk indirectly
causes a distortion in Csk interaction with ATP in the
ternary complex and thus slows leaving group (ADP)
departure; (ii) there is signi®cant entropy loss with
improper nucleophile alignment even in a dissociative
transition state. A detailed discussion concerning the
second possibility has recently been reported in the
nucleoside diphosphate kinase system.15 Further
experimental work will be needed in the Csk system to
discriminate between these mechanistic models and is
the subject of ongoing work in the lab.
pK of the alkyl hydroxy group versus the phenol would be
a
unlikely to explain the decreased reactivity since the b
nuc
for the reaction is near zero.5
D-Tyrosine as substrate
The ability of protein tyrosine kinases to discriminate
between l- and d-tyrosine residues within peptide
substrates has not been well studied. The insulin recep-
tor tyrosine kinase has been reported to be unable to
Electrostatic repulsion
1
1
process a peptide containing d-tyrosine (6), consistent
with expectations about the standard stereospeci®city of
enzyme active sites. In contrast, Lawrence and colleagues
The inability of the salicylic acid (10) containing peptide
to serve as a Csk substrate (<0.5% kinase rate with the

K. Kim et al. / Bioorg. Med. Chem. 8 (2000) 1263±1268
1267
tyrosine containing peptide) provides strong evidence
for the role of electrostatic repulsion in preventing the
processing of the phenoxide anion species. The pK of
a
protection or puri®cation. All other reagents not described
below were obtained from commercial sources in reagent
grade quality and used without further puri®cation.
5
the benzoic acid function is estimated to be approxi-
mately 3 in free solution. While enzyme active sites can
perturb substrate pK values, it is unlikely that Csk
3-Methyltyrosine (2), 3-hydroxymethyltyrosine (3), 2-
methyltyrosine (4), 2-methoxytyrosine (5), and 2-ethyl-
tyrosine (9) were prepared using the enzyme tyrosine phe-
a
would raise the pK more than 1±2 units at this posi-
a
5
6
tion and thus this group should be fully deprotonated
nol-lyase (TPL). Reactions were carried out as previously
described using the recombinant enzyme according to the
under the assay conditions (pH 7.4). While the large loss
in substrate eciency of the carboxylic acid substituted
derivative could in principle be due to steric repulsion,
the relative tolerance of the enzyme toward the peptide
with the hydroxymethyl derivative 3 (4.5-fold rate
reduction compared to peptide with tyrosine) suggests
5
adapted procedures necessary for larger scale work. In
particular, reactions were carried out on a 500 mL scale
and used 50±100 units of TPL. The typical yields were 0.5±
1.5 g of derivative per reaction. Work up and puri®cation
by cation exchange chromatography were carried out as
described previously. Each analogue was characterized
otherwise. The molecular volumes of CH OH versus
2
1
6
1
13
CO H are similar. It can be estimated that the elec-
2
using H NMR, C NMR, and MS and the spectroscopic
data for these compounds are shown below.
trostatic repulsion in the transition state by the carbox-
ylate ion leads to at least a 40-fold e€ect, or a
ÁG52.3 kcal/mol. Assuming a similar electrostatic
repulsion energy for the phenoxide anion in the Csk
transition state, this free energy increase would be
nearly enough to o€set the modest gain in rate (50-fold
assuming bnuc=0.1) expected in a dissociative mechan-
ism with the increased nucleophilicity of the phenoxide
anion. The loss of a hydrogen bond between the sub-
3-Hydroxymethyltyrosine (3).6,18 1H NMR (D O) d 6.77
(s, 1H), 6.71 (d, J=8.4 Hz, 1H), 6.47 (d, J=8.4 Hz), 4.23
2
1
3
(s, 2H), 3.48 (dd, J=6.8, 5.6 Hz, 1H), 2.69 (m, 2H);
C
NMR (D O) d 174.4, 153.0, 131.3, 129.7, 127.1, 117.1,
2
115.5, 59.8, 56.3, 35.8; HRMS calcd for C H NO
(MH+), 212.0923; found, 212.0926.
10
13
4
5
2-Methyltyrosine (4).6 H NMR (D O) d 7.11 (d,
1
strate tyrosine phenol and Asp-314 (ÁGꢀ3 kcal/mol)
2
associated with the phenoxide anion species would also
be expected to contribute to this species being a poor
protein tyrosine kinase substrate.
J=8.0 Hz, 1H), 6.80 (d, J=2.8 Hz, 1H), 6.72 (dd, J=8.0,
2.8 Hz, 1H), 3.82 (dd, 9.0, 5.8 Hz, 1H), 3.25 (dd, J=14.6,
5.8 Hz, 1H), 2.93 (dd, J=14.6, 9.0 Hz, 1H), 2.31 (s, 3H);
1
3
C NMR (D O) d 180.4, 158.0, 139.0, 131.9, 125.8,
2
+
1
18.6, 114.3, 56.70, 36.5, 19.0; MS (MH ) m/z 196.
Conclusion
3
-Methyltyrosine (2).6 1H NMR (D O) d 7.01 (s, 1H),
6.93 (d, J=8.0 Hz, 1H), 6.74 (d, J=8.0 Hz, 1H), 3.58 (dd,
2
This work provides new insights into the accessibility of
a protein tyrosine kinase active site to unnatural tyr-
osine derivatives. Some of the key ®ndings include: (i)
the tyrosine kinase Csk can tolerate substantial steric
bulkiness in the substrate aromatic ring, enhancing the
con®dence of the interpretations of the linear free
energy relationship measurements with the ¯uorotyr-
osine derivatives reported previously; (ii) proper ®t and
alignment for the substrate tyrosine group in the tyr-
osine kinase active site are important factors in catalysis
even in a dissociative transition state; (iii) electrostatic
repulsion is an important consideration in the enhanced
enzyme reactivity of the neutral phenol species of tyr-
osine substrates compared to the corresponding phen-
oxide anions. These studies allow a richer understanding
of the substrate selectivity and mechanism of protein
tyrosine kinases and should enhance the design of pepti-
domimetic tyrosine kinase inhibitors. They might also
provide opportunities in `bumps and holes' com-
J=7.4, 5.4 Hz, 1H), 2.96 (dd, J=13.8, 5.4 Hz, 1H), 2.77
(dd, J=13.8, 7.4 Hz, 1H), 2.14 (s, 3H); 13C NMR (D O)
2
d 179.9, 155.2, 132.2, 128.2, 126.9, 117.0, 115.4, 57.3,
38.5, 15.9; MS (MH ) m/z 196.
+
2-Methoxytyrosine (5).6,19 1H NMR (D O) d 7.03 (d,
J=8.0Hz, 1H), 6.51 (s, 1H), 6.43 (d, J=8.0Hz, 1H), 3.91
2
(dd, J=8.0, 4.4 Hz, 1H), 3.80 (s, 3H), 3.22 (dd, J=14.6,
4.4 Hz, 1H), 2.91 (dd, J=14.6, 8.0, 1H); 13C NMR (D O)
2
d 174.7, 158.9, 157.4, 132.4, 115.1, 107.8, 99.7, 55.6, 55.5,
+
31.2; HRMS calcd for C H NO (MH ), 212.0923;
found, 212.0926.
10
13
4
2-Ethyltyrosine (9).6 1H NMR (D O) d 7.09 (d, J=9Hz,
2
1H), 6.80 (d, J=2 Hz, 1H), 6.70 (dd, J=9, 2 Hz, 1H), 3.81
(dd, J=9, 6 Hz, 1H), 3.27 (dd, J=14, 6 Hz, 1H), 2.91 (dd,
J=14, 9 Hz, 1H), 2.61 (q, J=7.5 Hz, 2H), 1.16 (t,
J=7.5Hz, 3H); 13C NMR (D O) d 175.3, 155.7, 145.8,
2
1
7
plementarity studies that could ultimately allow speci®c
control of cellular signal transduction pathways.
132.2, 125.5, 116.3, 113.7, 56.5, 33.7, 25.4, 15.1; HRMS
calcd. for C H NO (MH+), 210.1130; found 210.1120.
11
15
3
Conversion to the Fmoc derivatives
Experimental
General
The Fmoc derivatives of all tyrosine analogues were
prepared using sodium carbonate and 9-¯uorenylmethyl
N-succinimidyl carbonate in dioxane according to
methods described previously, worked up as previously
stated by extraction and used in peptide synthesis with-
out further puri®cation.5
Fmoc-d-tyrosine, Fmoc-l-homotyrosine and Fmoc-l-
hydroxyphenylglycine were purchased from Novo-
biochem and used in peptide synthesis without further

1
268
K. Kim et al. / Bioorg. Med. Chem. 8 (2000) 1263±1268
Peptide synthesis
a) The peptides incorporating the unnatural tyrosine
analogues were synthesized and puri®ed according to
Acknowledgements
(
K.K. and K.P. contributed equally. We thank the
Winston Foundation for postdoctoral fellowship sup-
port to K.K. P.A.C. thanks the Burroughs Wellcome
Fund, the Irma T. Hirschl Foundation, the Irving
Hansen Foundation and the NIH for ®nancial support.
5
previously described methods. Cleavage and deprotec-
tion of the peptide containing salicyl alcohol function
were carried out in the absence of anisole to enhance the
yield of the desired material. The molecular weight of
each heptapeptide was con®rmed by electrospray mass
spectrometry and all were shown to be >95% pure by
reverse-phase HPLC (C-18 column, H O, CH CN, tri-
2
3
References
¯
uoroacetic acid).
1. Hunter, T. Cell 1995, 80, 225.
2. Shokat, K. M. Chemistry and Biology 1995, 2, 509.
3. Lee, T. R.; Niu, J.; Lawrence, D. S. Biochemistry 1994, 33,
(
b) The salicylic acid containing peptide was prepared
by oxidizing the salicylic alcohol containing peptide in
the solid phase which had been prepared with an N-
terminal Boc group. The resin bound peptide (795 mg;
0.1 mmol scale) was suspended in DMF (10 mL), treated
with 141 mg (0.4 mmol) of tetrapropyl ammonium per-
ruthenate and 4-methylmorpholine N-oxide (1.4 g,
12 mmol) with stirring for 5 h. The reaction mixture was
®
and dried under vacuum overnight. The resultant resin
was treated with mCPBA (77%, 859 mg) in DMF
(
for 28 h at 4 C was washed (DMF, MeOH, CH Cl ) and
4
4
5
1
6
1
6
245.
. Lawrence, D. S.; Niu, J. Pharmacol. Ther. 1998, 77, 81.
. (a) Kim, K.; Cole, P. A. J. Am. Chem. Soc. 1997, 119,
1096. (b) Kim, K.; Cole, P. A. J. Am. Chem. Soc. 1998, 120,
851. (c) Cole, P. A.; Sondhi, D.; Kim, K. Pharmacol. Ther.
999, 82, 219.
. (a) Nagasawa, T.; Utagawa, T.; Goto, J.; Kim, C.-J.; Tani, Y.;
ltered and washed with DMF, CH Cl , MeOH, water
2
2
Kumagai, H.; Yamada, H. Eur. J. Biochem. 1981, 117, 33. (b)
Hebel, D.; Furlano, C.; Phillips, R. S.; Koushik, S.; Creveling, C.
R.; Kirk, K. L. Bioorg. Med. Chem. 1992, 2, 41. (c) Phillips, R.
S.; Fletcher, J. G.; von Tersch, R. L.; Kirk, K. L. Arch. Bio-
chem. Biophys. 1990, 276, 65.
12 mL) and dimethoxyethane (5 mL) and after stirring
ꢁ
ltered. The resultant resin was treated with standard
2
2
7
. (a) Yan, B.; Sun, Q.; Wareing, J. R.; Jewell, C. F. J. Org.
®
Chem. 1996, 61, 8765. (b) Beebe, X., Schore N. E. Kurth, M.
J. J. Am. Chem. Soc. 1992 114 10061.
deprotection/cleavage conditions (5 mL tri¯uoroacetic
acid/0.25 mL water/0.25 mL anisole mixture) and the
peptide isolated by ether precipitation and puri®ed by
reverse-phase HPLC (typical yield of puri®ed peptide
was ꢂ1±2 mg). The molecular weight was con®rmed by
8
. Silverman, R. B. The Organic Chemistry of Drug Design and
Drug Action; Academic: New York, 1992.
. (a) Myers, J. K.; Cohen, J. D.; Widlanski, T. S. J. Am.
9
Chem. Soc. 1995, 117, 11049. (b) Born, T. L.; Myers, J. K.;
Widlanski, T. S.; Rusnack, F. J. Biol. Chem. 1995, 270, 25651.
10. Lee, T. R.; Till, J. H.; Lawrence, D. S.; Miller, W. T. J.
Biol. Chem. 1995, 270, 27022.
1
electrospray MS and H NMR of the puri®ed peptide
revealed the loss of the methylene signal attributed to
the hydroxymethyl protons and appropriate aromatic
shifts characteristic of the salicylic acid derivative.
1
1. Walker, D. H.; Kuppuswamy, D.; Visvanathan, A.; Pike,
L. J. Biochemistry 1987, 26, 1428.
2. Fry, D. W.; McMichael, A.; Singh, J.; Dobrusin, E. M.;
McNamara, D. J. Peptides 1994, 15, 951.
1
Enzyme assays
1
1
1
3. Mildvan, A. S. Proteins: Structure, Function, and Genetics
997, 29, 401.
4. Hubbard, S. R.; Wei, L.; Ellis, L.; Hendrickson, W. A.
Assays were carried out using previously described
methods with quanti®cation of the substrate and phos-
phopeptide product by HPLC integration at 214 nm.5
Brie¯y, the conditions included saturating ATP (2 mM),
Nature 1994, 372, 746.
15. Admiraal, S. J.; Schneider, B.; Meyer, P.; Janin, J.; Veron,
M.; Deville-Bonne, D.; Herschlag, D. Biochemistry 1999, 38,
4701.
1
3
mM DTT, 5 mM MnCl , 60 mM Tris±HCl, pH 7.4,
2
0 C. Electrospray mass spectrometry was used to
ꢁ
1
6. Hansch, C.; Leo, A. Exploring QSAR; ACS: Washington,
DC, 1995, pp 81±84.
7. (a) Belshaw, P. J.; Schoepfer, J. G.; Liu, K.-Q.; Morrison,
K. L.; Schreiber, S. L. Angew. Chem. Int. Ed. Engl. 1995, 34,
con®rm the identity of the phosphopeptide product in
each case. All rates were measured in a linear range with
respect to time and enzyme concentration with <10%
turnover of the limiting substrate. Peptide substrate
concentration ranges were varied from 0 to 4 mM and
the data were ®t to the Michaelis Menten equation
which in most cases showed that the K was >4 mM so
m
that only the kcat/Km could be reliably compared for
each. All experiments were performed at least twice.
1
2
2
129. (b) Hwang, Y. W., Miller, D. L. J. Biol. Chem. 1987,
62, 13081. (c) Shah, K.; Liu Y.; Deirmengian, C.; Shokat, K.
M.; Proc. Natl. Acad. Sci. USA 1997, 94, 3565.
18. Atkinson, M.; Hartley, D.; Lunts, L. H. C.; Ritchie, A. C.
J. Med. Chem. 1974, 17, 248.
19. Ferrini, R.; Glasser, A. Biochem. Pharm. 1964, 13, 798.