4′-O-[2-(2-fluoromalonyl)]-L-tyrosine: A phosphotyrosyl mimic for the preparation of signal transduction inhibitory peptides

Article abstract of DOI:10.1021/jm950621g

Development of phosphotyrosyl (pTyr) mimetics which are stable to protein-tyrosine phosphatases (PTPs), yet can retain biological potency when incorporated into peptides, is an active area of drug development. Since a majority of pTyr mimetics derive thei

Full text of DOI:10.1021/jm950621g

J . Med. Chem. 1996, 39, 1021-1027
1021
Articles
4′-O-[2-(2-F lu or om a lon yl)]-L-tyr osin e: A P h osp h otyr osyl Mim ic for th e
P r ep a r a tion of Sign a l Tr a n sd u ction In h ibitor y P ep tid es1
Terrence R. Burke, J r.,*,† Bin Ye,† Miki Akamatsu,† Harry Ford, J r.,† Xinjian Yan,† Hemanta K. Kole,‡
Gert Wolf,§ Steven E. Shoelson,§ and Peter P. Roller†
Laboratory of Medicinal Chemistry, Building 37, Room 5C06, Division of Basic Sciences, National Cancer Institute,
National Institutes of Health, Bethesda, Maryland 20892, Laboratory of Clinical Physiology, Diabetes Unit,
Gerontology Research Center, National Institute on Aging, National Institutes of Health, Baltimore, Maryland 21224, and
J oslin Diabetes Center & Department of Medicine, Harvard Medical School, Boston, Massachusetts 02215
Received August 22, 1995X
Development of phosphotyrosyl (pTyr) mimetics which are stable to protein-tyrosine phos-
phatases (PTPs), yet can retain biological potency when incorporated into peptides, is an active
area of drug development. Since a majority of pTyr mimetics derive their “phosphofunctionality”
from phosphorus-containing moieties, such as phosphonates, evolution of new inhibitors and
modes of prodrug derivatization have been restricted to chemistries appropriate for phosphorus-
containing moieties. A new, nonphosphorus-containing pTyr mimetic has recently been
reported, L-O-(2-malonyl)tyrosine (OMT, 5), which can be incorporated into peptides that exhibit
good PTP and Src homology 2 (SH2) domain inhibitory potency. For phosphonate-based pTyr
mimetics such as phosphonomethyl phenylalanine (Pmp, 2), introduction of fluorines R to the
phosphorus has provided higher affinity pTyr mimetics. This strategy has now been applied
to OMT, and herein is reported 4′-O-[2-(2-fluoromalonyl)]-L-tyrosine (FOMT, 6), a new fluorine-
containing nonphosphorus pTyr mimetic. Incorporation of FOMT into appropriate peptides
results in good inhibition of both PTP and SH2 domains. In an assay measuring the inhibition
of PTP 1B-mediated dephosphorylation of phosphorylated insulin receptor, the peptide Ac-D-
A-D-E-X-L-amide exhibited a 10-fold enhancement in inhibitory potency for X ) FOMT (19)
(IC50 ) 1 µM) relative to the unfluorinated peptide, X ) OMT (18) (IC50 ) 10 µM). Molecular
modeling indicated that this increased affinity may be attributable to new hydrogen-bonding
interactions between the fluorine and the enzyme catalytic site, and not due to lowering of
pKa values. In a competition binding assay using the p85 PI 3-kinase C-terminal SH2 domain
GST fusion construct, the inhibitory peptide, Ac-D-X-V-P-M-L-amide, showed no enhancement
of inhibitory potency for X ) FOMT (22) (IC50 ) 18 µM) relative to the unfluorinated peptide,
X ) OMT (21) (IC50 ) 14 µM). The use of FOMT would therefore appear to have particular
potential for the development of PTP inhibitors.
Modulators of cell-signaling pathways promise new
approaches toward therapeutic design, with the devel-
opment of inhibitors of protein-tyrosine kinase (PTK)
cascades currently being one important area of inves-
tigation.2,3 Central to signaling through PTK-dependent
pathways is the generation (via PTKs), utilization (via
Src homology 2 (SH2) domains), and destruction (by
protein-tyrosine phosphatases, PTPs) of phosphotyrosyl
(pTyr, 1) residues within key proteins. Additionally,
recent reports of pTyr-dependent protein-protein bind-
ing not involving SH2 domains4-7 adds yet another
dimension to this signaling scheme. While significant
work has been reported on PTK inhibitors,3,8,9 develop-
ment of SH2 domain and PTP inhibitors has not been
well represented.
ment of respective inhibitors, and we have previously
reported the synthesis of constrained pTyr mimetics as
potential SH2 domain inhibitors. However, one limita-
tion of inhibitors based on either tyrosine phosphate or
related phenyl phosphates is their potential susceptibil-
ity to phosphorolysis and inactivation by PTPs. To
overcome this liability, the use of benzylphosphonate-
containing structures, in which the phosphate ester
oxygen has been replaced by a chemically and enzy-
matically stable methylene unit, have been used. In the
case of pTyr (1), the resulting (phosphonomethyl)-
phenylalanine (Pmp; 2)10 has been successfully substi-
tuted into peptides which are antagonists of both SH2
domain binding11-13 and PTPs.14-16 Modification of
Pmp by substitution of fluorine at the methylene bridge
to give monofluoroPmp (FPmp, 3)17 or difluoroPmp (F2-
Pmp, 4)17 resulted in enhanced potency against both
SH2 domains12,18,19 and PTPs.16
Since the pTyr pharmacophore is a defining determi-
nant for both of these interactions, basing the design of
SH2 domain and PTP inhibitors on the structure of
tyrosine phosphate is one approach toward the develop-
In exploring fresh directions for inhibitor design, non-
phosphorus-containing pTyr mimetics are of interest,
since they could afford new avenues for prodrug delivery
and could potentially display altered biological profiles.
We have recently reported one such pTyr mimetic, 4′-
† Laboratory of Medicinal Chemistry, NCI.
‡ Laboratory of Clinical Physiology, NIA.
§ J oslin Diabetes Center.
X Abstract published in Advance ACS Abstracts, February 1, 1996.
This article not subject to U.S. Copyright. Published 1996 by the American Chemical Society

1022 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 5
Burke et al.
Resu lts a n d Discu ssion
P TP In h ibition . In order to examine the ability of
FOMT-containing peptides to inhibit PTPs, the peptide
Ac-D-A-D-E-[FOMT]-L-amide (19) was prepared, based
on the corresponding pTyr-containing sequence, which
has been shown to be a high-affinity hexamer substrate
of PTP 1.15 The OMT-substituted peptide 18 has
previously been shown to inhibit the PTP 1B-mediated
dephosphorylation of endogenously phosphorylated in-
sulin receptor with an IC50 value of 10 µM.22 Under
the same assay conditions, FOMT-containing peptide 19
exhibited an IC50 value of 1 µM, representing a 10-fold
increase in affinity relative to its unfluorinated coun-
terpart 18 (Figure 1).
O-(2-malonyl)-L-tyrosine (OMT, 5), in which the phos-
phate group has been replaced by a malonate struc-
ture.20 Peptides bearing this new analogue exhibit
reasonable inhibitory potency in both SH2 domain21 and
PTP22 assays. In order to explore structural variants
of OMT which may modify its potency or specificity, we
took note of the previous dramatic effect on both SH2
domain and PTP inhibition when fluorine was intro-
duced into Pmp. We therefore designed a modified OMT
which contains a fluorine at the malonyl R-carbon.
Herein we report the synthesis of this analogue, 4′-O-
[2-(2-fluoromalonyl)]-L-tyrosine (FOMT, 6), as its NR-
Fmoc-O,O-di-tert-butyl derivative (7), which is suitably
protected for solid-phase peptide synthesis of L-FOMT
peptides. (Note: Although these derivatives may be
more properly termed, “2-O-tyrosinyl)malonate ethers,”
we have adapted an alternate naming system which is
consistent with previous nomenclature applied to the
unfluorinated analogues.21,22) We include the synthesis
of FOMT-containing peptides and examine their inhibi-
tory potency in SH2 domain and PTP assays.
SH2 Dom a in In h ibition . The “pY-V-P-M-L” se-
quence has been extensively studied as an inhibitory
motif directed against the PI 3-kinase p85 C-terminal
SH2 domain, and it has served as a benchmark se-
quence in our previous studies on the SH2 domain
inhibitory potency of both linear12 and cyclic18,19 peptides
containing phosphonate-based pTyr mimetics. More
recently Ac-D-[L-OMT]-V-P-M-L-amide (21) has been
examined for inhibitory potency in a p85 C-terminal
binding assay and has been found to have an IC50 value
of 14 µM.21 This is on the order of inhibitory potencies
observed for L-OMT-containing peptides directed against
the Src SH2 domain (IC50 ) 25 µM) and the SH-PTP2
SH2 domain (IC50 ) 22 µM).21 The present study
examined the effect of fluorine substitution in a peptide
containing OMT. Analogous fluorine substitution in
“pY-V-P-M-L” sequences containing phosphonate-based
pTyr mimetics resulted in a doubling of potency for each
fluorine added, with the difluorinated “A-D-[L-F2Pmp]-
V-P-M-L-amide” having a potency equivalent to the
parent pTyr-containing peptide 20 (IC50 ≈ 0.15 µM).12
In contrast to the enhancing effect of fluorination on
the SH2 domain inhibitory potency of peptides bearing
phosphonate-based pTyr mimetics, fluorination of OMT
had no effect on inhibitory potency as indicated by p85
C-terminal SH2 domain binding assays using the L-
FOMT-containing peptide 20 (IC50 ) 18.3 µM; Figure
2).
Syn th esis
Mod el F lu or in a tion . The 6-bromo-2-[[2-(2-fluoro-
malonyl)]oxy]naphthalene (13) was prepared as a model
compound to examine fluorination conditions. It was
also subsequently used for the determination of pKa
values. The 6-bromo substituent of 13 facilitated crys-
tallization of products relative to the unbrominated
species. Refluxing a benzene solution of 6-bromo-2-
naphthol (8) and di-tert-butyl R-diazomalonate23 (9) in
the presence of rhodium diacetate24 provided the ma-
lonyl adduct 10, with subsequent TFA-catalyzed hy-
drolysis of the tert-butyl protecting groups affording the
free diacid 12 (Scheme 1). Electrophilic fluorination of
10 using sodium bis(trimethylsilyl)amide and N-fluo-
robenzenesulfonimide provided the 2-fluoromalonyl prod-
uct 11, which was hydrolyzed to the free diacid 13.
F lu or in a tion of P r otected OMT. Preparation of
the final NR-Fmoc-O,O-di-tert-butyl-protected deriva-
tive 7 was achieved by initially applying the above
outlined electrophilic fluorination on the protected OMT
derivative 15 to yield 16 (Scheme 2). Compound 15
was obtained from the corresponding methyl NR-Fmoc-
tyrosinate (14) as previously described.20,21 Crucial to
the success of the fluorination was the stability of the
base-labile N-Fmoc protecting group to the strongly
alkaline bis(trimethylsilyl)amide. Equally important
was the stability of both the Fmoc group and the
R-fluoromalonyl ester groups to the lithium hydroxide-
catalyzed hydrolysis of the tyrosinate methyl ester to
the final free amino acid 7.25 The steric bulk afforded
by the tert-butyl groups contributed to the stability of
the R-fluoromalonyl ester groups, as the corresponding
n-butyl-protected R-fluoromalonyl diester underwent
hydrolysis of one ester group under similar conditions.26
Effect of F lu or in a tion on Ion iza tion Con sta n ts.
The phosphorus-containing pTyr mimetic, Pmp (2),
provides two geminal sites for introduction of fluorine
at the phosphonate R-methylene bridge. The corre-
sponding difluorinated phosphonate, F2Pmp (4), when
incorporated into a peptide, has dramatically enhanced
inhibitory potency against the tyrosine phosphatase
PTP 1B16 and appreciably increased binding potency in
SH2 domain binding assays.12 The potentiating effects
of fluorine in these cases may be at least partially
attributable to two factors: (1) the reintroduction of
hydrogen-bonding capability through the fluorine atom
at the phosphonate R-methylene which was originally
present in the parent phosphate and (2) a reduction in
second phosphonate ionization constant (pKa ).12 The
2
effect on pKa values is of particular note. An aryl
2
phosphate has two ionizable hydroxyls, each with a
distinct pKa value. The first ionization constant is lower

4′-O-[2-(2-Fluoromalonyl)]-L-tyrosine
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 5 1023
Sch em e 1
Sch em e 2
Ta ble 1. Comparison of pKa values
2
inhibitors (such as Pmp) by fluorination is the increase
in the population of di-ionized species which results
because changes in pKa values occur at or near physi-
F igu r e 1. Effect of FOMT-containing peptide 19 on recom-
binant PTP 1B-mediated dephosphorylation of insulin recep-
tors. Results are the mean ( SE of three separate experiments
performed in duplicate.
2
ological pH. In the case of OMT (5), a single malonyl
R-hydrogen exists, limiting fluorination to the mono-
fluoro species (FOMT, 6). As with phosphonate-based
pTyr mimetics, physiological effects of fluorination could
arise from increased hydrogen-bonding interactions and/
or from a lowering of the malonate pKa value.
2
To examine the possible contribution changes in pKa
values may have on differential potencies of OMT- and
FOMT-containing peptides, pKa values were determined
for compounds 12 and 13, which represent structurally
simplified analogues of OMT and FOMT, respectively
(Table 1). Compound 12 exhibits a pKa value ap-
2
proximately 2 units lower than unsubstituted malonic
acid, with monofluorinated compound 13 showing a
reduction in its pKa value by yet another unit. For both
2
F igu r e 2. Competitive binding curve for the peptide Ac-D-
FOMT-V-P-M-L-amide (22) in an assay using the p85 PI
3-kinase C-terminal SH2 domain. Results are the mean ( SE
of duplicate experiments.
12 and 13 however, pKa values are greater than 3 log
2
units below physiological pH, and these compounds
would exist nearly exclusively as the di-ionized species.
For this reason, effects of fluorination on pKa values
would not be expected to affect the biological potency
of 13 or similar R-fluoro 2-malonate-containing com-
pounds such as FOMT.
than the second ionization constant, which occurs at a
pKa value of approximately 6.2. Therefore at physi-
2
ological pH the phosphate exists predominately in the
di-ionized form. Benzylphosphonic acid has a pKa value
Molecu la r Mod elin g. In order to better explain the
10-fold enhancement in the PTP 1B interaction of the
FOMT-containing peptide 19 as compared with the
nonfluorinated OMT counterpart 18, a computer-as-
sisted molecular-modeling study were undertaken. For
modeling the effect of fluorination on the binding of
OMT versus FOMT to the catalytic site of PTP 1B, we
utilized the recently solved X-ray structure of PTP 1B
bearing a [1,1-difluoro-1-(2-naphthyl)methyl]phosphonic
2
of approximately 7.7 and therefore has an appreciable
population of monoionized species at physiological pH.
Substitution of fluorines at the R-methylene of ben-
zylphosphonic acid reduces its pKa value by approxi-
2
mately one unit per fluorine added, with the monofluoro-
and difluorophosphonates exhibiting pKa values of 6.6
2
and 5.7, respectively.27 Central to the enhancement of
inhibitory potency of benzylphosphonic acid-containing

1024 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 5
Burke et al.
these latter compounds, this potentiation could be
ascribed to the ability of R-fluorines to both reduce the
phosphonate pKa value and reintroduce hydrogen-
2
bonding interactions similar to the parent tyrosine
phosphate (1). In studies presented here, we demon-
strate that relative to an OMT-containing peptide, the
corresponding FOMT-containing peptide exhibited a 10-
fold enhancement in inhibitory potency against PTP 1B.
Determination of pKa values with model compounds
2
indicated that while addition of a fluorine reduced the
pKa value, any biological effect from this would not be
2
appreciable at physiological pH, where complete ioniza-
tion of malonyl carboxyls on both OMT and FOMT
would occur. However, molecular-modeling studies
which simulated the binding of OMT and FOMT to PTP
1B indicated that addition of the fluorine atom could
introduce new hydrogen bonding with the catalytic site.
Therefore, a hydrogen-bonding effect and not a pKa
2
effect could at least partially account for the observed
increase in affinity of FOMT-containing 19 relative to
unfluorinated 18.
F igu r e 3. Structure of CHF(CO2)2 bound within the PTP 1B
catalytic site. The overall geometry of binding was based on
the X-ray structure of an aryl difluorophosphonate inhibitor
complexed to PTP 1B. Key hydrogen bonding interactions are
shown by dotted lines. A bold dotted line designates an
important F-H interaction with Phe182.
A study designed to examine the efficacy of FOMT
as a pTyr mimetic in SH2 domain-directed peptides was
also undertaken. In this assay FOMT-containing pep-
tide 22 showed no increase in binding potency relative
to its unfluorinated counterpart 21 using a p85 PI
3-kinase C-terminal binding system. Unlike phospho-
nate-based pTyr mimetics where R-fluorination in-
creased SH2 domain binding potency, there appears to
be no advantage gained by R-fluorination of OMT.
acid ligand bound, which approximates the binding of
a tyrosine phosphate moiety.28 Recently the X-ray
structure has also been reported of Ac-D-A-D-E-pY-L-
amide bound to a deactivated PTP 1B mutant (the
catalytic Cys residue has been substituted with a Ser
residue).29 Replacement of the difluorophosphonate in
the X-ray structure with a malonate moiety [CH2(CO2)2]
and subsequent minimization showed that although the
malonate structure occupies approximately 12% more
volume than the phosphonate, it is easily accommodated
without significant disruption of the enzyme catalytic
site geometry.22 Introduction of a fluorine atom onto
the malonate R-methylene approximates the corre-
sponding structural change in going from OMT to
FOMT. As shown in Figure 3, the R-fluoro atom is
suitably situated to engage in hydrogen-bonding inter-
actions with the Asp181-Phe182 amide nitrogen. Al-
though the observed distance (2.7 Å) is not optimal for
maximum hydrogen bonding, reminimization of the
enzyme following introduction of the fluorine would be
expected to shorten the distance. The binding energy
gained by the F-H hydrogen bonding could contribute
to some of the 10-fold enhancement in binding observed
for the FOMT-containing peptide 19 relative to the
unfluorinated 18.
Su m m a r y. The O-(2-malonyl)tyrosyl residue (OMT,
5) has recently demonstrated usefulness as a pTyr
mimetic in peptides which are inhibitory to both PTPs
and SH2 domain binding interactions. As a non-
phosphorus-based pTyr mimetic, OMT may provide
alternative approaches for developing both selective and
bioavailable inhibitors of protein-tyrosine kinase-de-
pendent signaling pathways. A first study in preparing
modified OMT analogues was undertaken in which a
fluorine atom was introduced onto the malonyl R-me-
thylene, providing the new analogue FOMT (6). In the
phosphonate-based series of pTyr mimetics (i.e., Pmp,
2), analogous addition of R-fluorines (F2Pmp, 4) en-
hanced both PTP and SH2 domain binding potency. For
Exp er im en ta l Section
Syn th esis. Petroleum ether was of the boiling range 35-
60 °C, and removal of solvents was performed by rotary
evaporation under reduced pressure. Silica gel filtration was
carried out using TLC grade silica gel (5-25 µm, Aldrich; 6.5
cm diameter × 4 cm high). Melting points were determined
on a Mel Temp II melting point apparatus and are uncorrected.
Elemental analyses were obtained from Atlantic Microlab Inc.,
Norcross, GA, and are within 0.4% of theoretical values unless
otherwise indicated. Fast atom bombardment mass spectra
(FABMS) were acquired with a VG Analytical 7070E mass
spectrometer under the control of a VG 2035 data system.
Where indicated, FABMS were run using a matrix of 3-ni-
trobenzyl alcohol (NBA). 1H NMR data were obtained on
Bruker AC250 (250 MHz) or, if indicated, AMX500 (500 MHz)
instruments and are reported in ppm relative to TMS and
referenced to the solvent in which they were run. Anhydrous
solvents were obtained commercially and used without further
drying.
6-Br om o-2-[[O,O′-d i-ter t-bu tyl-2-m a lon yl]oxy]n a p h th a -
len e (10). A mixture of 6-bromo-2-naphthol (2.23 g, 10.0
mmol), di-tert-butyl R-diazomalonate23 (2.90 g, 12.0 mmol), and
rhodium(II) acetate dimer (196 mg, 0.44 mmol) in anhydrous
benzene (100 mL) was stirred at reflux overnight. The mixture
was cooled to room temperature, filtered through Celite, taken
to dryness, and crystallized from petroleum ether (cooling to
-78 °C). The resulting solid was triturated with petroleum
ether at room temperature and then dried, yielding white
crystals (2.85 g) which contained approximately 25% of an
impurity (by NMR). The material was recrystallized from
hexanes to provide 10 as a white crystalline solid (893 mg,
20%): mp 120 °C soften, 127-128 °C; 1H NMR (CDCl3) δ 7.92
(d, 1H, J ) 2 Hz), 7.69 (d, 1H, J ) 9.1 Hz), 7.57 (d, 1H, J )
8.7 Hz), 7.50 (dd, 1H, J ) 1.8, 8.6 Hz), 7.32 (dd, 1H, J ) 2.0,
9.0 Hz), 7.09 (d, 1H, J ) 2.5 Hz), 5.12 (s, 1H), 1.50 (2, 18H).
Anal. (C21H25BrO5) C,H.
6-Br om o-2-[[O,O′-d i-ter t-bu tyl-2-(2-flu or om a lon yl)]ox-
y]n a p h th a len e (11). To a solution of 10 (180 mg, 0.5 mmol)
in anhydrous THF (3 mL) at -78 °C under argon was added
sodium bis(trimethylsilyl)amide, 1.0 M in THF (550 µL), and

4′-O-[2-(2-Fluoromalonyl)]-L-tyrosine
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 5 1025
the mixture was stirred at -78 °C (40 min). A solution of
N-fluorobenzenesulfonimide (Allied Signal, Inc.) (173 mg, 0.55
mmol) in anhydrous THF (2 mL) was added dropwise, and the
reaction was continued first at -78 °C (4 h) and then at
ambient temperature (20 min). The resulting white suspen-
sion was diluted with saturated aqueous ammonium chloride
(50 mL), extracted with CHCl3 (2 × 50 mL), dried (MgSO4),
and taken to dryness, yielding a yellow oil (206 mg). Chro-
matographic purification (4.5 cm diameter × 2.5 cm high
column using a stepwise gradient of CHCl3 (0-75%) in
petroleum ether) provided pure 11 (TLC Rf ) 0.70, CHCl3) as
a colorless syrup (109 mg, 58%), with additional material being
collected as a mixture of 11 and it mono-tert-butyl ester (90
mg): 1H NMR (CDCl3) δ 7.90 (d, 1H, J ) 2 Hz), 7.64 (d, 1H, J
) 9.0 hz), 7.57 (d, 1H, J ) 8.8 Hz), 7.51 (br s, 1H), 7.47 (dd,
1H, J ) 2.2, 9.0 Hz), 7.31 (dd, 1H, J ) 2.3, 8.9 Hz), 1.32 (s,
18H).
6-Br om o-2-(2-m a lon yloxy)n a p h th a len e (12). To 10 (180
mg, 0.5 mmol) was added neat TFA (1.0 mL), and the mixture
was stirred at room temperature (1 h). The resulting white
suspension was diluted with petroleum ether (2 mL) and
filtered, and the filter cake was washed with several portions
of petroleum ether, providing 12 as white crystals (120 mg,
98%): An analytical sample was obtained by hplc purification
(Vydac C18 peptide and protein semiprep column; A ) 0.05%
aqueous TFA, B ) 0.5% TFA in acetonitrile; isocratic, 40% B;
retention K′ ) 1.5): mp 152 °C (shrink), 215-223 °C (note: if
the temperature is increased rapidly, decomposition with
evolution of gas occurs at approximately 170 °C); 1H NMR (500
MHz; H2O:D2O) δ 8.08 (br s, 1H), 7.84 (dd, 1H, J ) 3.5, 9.0
Hz), 7.75 (dd, 1H, J ) 4.3, 8.6 Hz), 7.62 (d, 1H, J ) 8.8 hz),
7.35 (dd, 1H, J ) 2.1, 9.0 Hz), 7.22 (br s, 1H), 5.37 (s, 1H);
FABMS (NBA, -ve) m/z 323 (M - H)-. Anal. (C13H9BrO5‚3/
4H2O) C,H.
6-Br om o-2-[[2-(2-flu or om a lon yl)]oxy]n a p h th a len e (13).
A mixture of 11 contaminated with a little of its mono-tert-
butyl ester (199 mg total, 00.5 mmol theoretical) was stirred
at room temperature with 2 mL of 90% TFA (1 h). The
mixture was then taken to dryness under high vacuum, and
the resulting colorless syrup was purified by silica gel chro-
matography (CHCl3 followed by EtOAc) to provide a colorless
syrup (125 mg), which was crystallized from petroleum ether:
CHCl3 to give 13 (TLC Rf ) 0.06, CHCl3) as white crystals (76
mg, 42%), contaminated with a small amount of impurity. An
analytical sample was prepared by recrystallization from
CHCl3 to provide crystals, mp 141-142 °C. Further purifica-
tion by HPLC was performed (Vydac C18 peptide and protein
semiprep column; A ) 0.05% aqueous TFA, B ) 0.5% TFA in
acetonitrile; linear gradient, 5% B to 95% B over 30 min;
retention K′ ) 2.3): 1H NMR (DMSO-d6) δ 8.23 (d, 1H, J )
1.9 Hz), 7.96 (d, 1H, J ) 9.1 Hz), 7.90 (d, 1H, J ) 8.9 Hz),
7.65 (dd, 1H, J ) 1.9, 8.8 Hz), 7.64 (d, 1H, J ) 2.3 Hz), 7.39
(dd, 1H, J ) 2.3, 8.9 Hz); FABMS (NBA, -ve) m/z 341 (M -
H)-. Anal. (C13H8BrFO5‚1/2H2O) C,H.
(m, 1H), 3.00 (d, 2H, J ) 5.5 Hz), 1.34 (s, 18H). Anal. (C36H40
-
FNO9) C, H, N.
Nr-F m oc-4-O-[O′,O′′-d i-ter t-bu tyl-2-(2-flu or om a lon yl)]-
L-tyr osin e (7). To a solution of 16 (300 mg, 0.46 mmol) in
THF (4.6 mL) at 0 °C was added a solution of 0.2 N LiOH (4.6
mL, 0.92 mmol) and the reaction mixture stirred at 0 °C (30
min). An additional portion of ice-cold 0.2 N LiOH (2.3 mL,
0.46 mmol) was added, and the reaction continued at 0 °C (30
min). The mixture was then partitioned between ice-cold 0.2
N HCl (100 mL) and ethyl acetate (2 × 50 mL), dried (MgSO4),
and taken to dryness, yielding 7 as a white foam (260 mg,
89%): 1H NMR (CDCl3) δ 7.77 (d, 2H, J ) 7.4 Hz), 7.56 (d,
2H, J ) 7.3 Hz), 7.40 (t, 2H, J ) 7.3 Hz), 7.32 (t, 2H, J ) 7.4
Hz), 7.12 (d, 2H, J ) 8.3 Hz), 7.04 (d, 2H, J ) 8.3 Hz), 5.10 (d,
1H, J ) 8.0 Hz), 4.69-4.58 (m, 1H), 4.51-4.40 (m, 2H), 4.20
(t, 2H, J ) 6.7 Hz), 3.73 (m, 1H), 3.21-3.02 (m, 2H), 1.39 (s,
18H); FABMS m/z 634 (M + H)+. Anal. (C35H38FNO9) C, H,
N.
P ep tid e Syn th esis. The pTyr mimetic X ) L-FOMT (6)
was incorporated into peptides 19 and 22 using the amino acid
NR-F m oc-4-O-[O′,O′′-di-ter t-bu t yl-2-(2-flu or om a lon yl)]t y-
rosine (7) and solid-phase synthesis with Fmoc chemistry.
Fmoc derivatives of standard amino acids were obtained from
Bachem Califormia (Torrence, CA) or from Millipore/Milligen
(Bedford, MA). Side-chain protection was as follows: Asp(tBu)
and Glu(tBu), and Asn and Gln were unprotected. The
peptides were prepared using PAL amide resin,30 with DIPCDI/
HOBT as coupling reagents in DMF (coupling time 1 h, except
the FOMT amino acid 7 which was coupled for 39 h (peptide
19) and for 17 h (peptide 22)). Fmoc deprotection was
conducted using 20% piperidine/DMF (2 min, then 12 min).
The resin-bound protected peptide was N-terminally acetylated
with 10% 1-acetylimidazole/DMF (2 × 1.5 h room tempera-
ture). Resin cleavage and side-chain deprotection was done
in one step with TFA containing 5% each (v/v) of ethanedithiol,
m-cresol, thioanisole, and H2O (0.5 h at 4 °C, then 2.5 h at
room temperature). To isolate the peptides, two-thirds of the
reagent mixture from the deprotection reaction was evaporated
under N2, and the mixture was triturated with ice-cold Et2O.
The precipitated crude peptides were purified to homogeneity
by reverse phase HPLC. Conditions: Vydac C18 column (10
× 250 mm); solvent gradient 0.05% TFA in H2O, B 0.05% TFA
in 90% acetonitrile in H2O, gradient as indicated below; flow
rate 2.5 mL/min; UV detector 220 nm. Mass spectra (fast atom
bombardment, unit resolution, glycerol matrix, positive and/
or negative ion mode) were performed on a VG Analytical
7070E-HF mass spectrometer. Amino acid analysis (6 N HCl,
110 °C, 24 h) was carried out at the Protein Structure
Laboratory, University of California, Davis, CA.
Ac-Asp-Ala-Asp-Glu -(L-FOMT)-Leu -am ide (19): RPHPLC
retention time 14.7 min (gradient 10-45% B over 25 min);
FABMS (M - H)- 884.4 (calc 884.4); amino acid analysis Asp
1.92 (2), Glu 1.06 (1), Ala (0.94 (1), Leu 1.08 (1). FOMT was
not analyzed.
Met h yl Nr-F m oc-4-O-[O′,O′′-d i-ter t-b u t yl-2-(2-flu or o-
m a lon yl)]-L-tyr osin a te (16). To a solution of methyl NR-
Fmoc-4-O-[O′,O′′-di-tert-butyl-2-malonyl]tyrosinate (15)21 (1.26
g, 2.0 mmol) in anhydrous THF (10 mL) at -78 °C under argon
was added sodium bis(trimethylsilyl)amide, 1.0 M in THF (2.0
mL, 2.0 mmol), dropwise over 5 min, and then the reaction
mixture was stirred at -78 °C (15 min). A solution of
N-fluorobenzenesulfonimide (Allied Signal Corp.) (630 mg, 2.0
mmol) in anhydrous THF (5 mL) was added dropwise, and
then the reaction mixture was stirred at -78 °C (1 h). The
reaction was diluted with saturated ammonium chloride (100
mL), extracted with ethyl acetate (2 × 50 mL), washed with
saturated ammonium chloride (50 mL), dried (MgSO4), and
taken to dryness, yielding a white foam (1.33 g). Purification
by silica gel chromatography (ethyl acetate in petroleum ether;
10:90 followed by 20:80) provided 16 (TLC Rf ) 0.78, ethyl
acetate:CHCl3, 1:4) as a white foam (1.0 g, 77%). Crystalliza-
Ac-Asp-(L-FOMT)-Val-P r o-Met-Leu -am ide (22): RPHPLC
retention time 12.5 min (gradient: 25-45% B over 20 min);
FABMS (M - H)- 896.4 (calc 896.4); amino acid analysis Asp
1.00 (1), Pro 1.00 (1), Val 1.00 (1), Met 0.97 (1), Leu 1.03 (1).
FOMT was not analyzed.
P TP In h ibition Assa y. The ability of FOMT hexamer 19
to inhibit the PTP 1B-mediated dephosphorylation of phos-
phorylated insulin receptor was measured as previously
described.22 In summary, Chinese hamster ovary (CHO) cell
line transfected with an expression plasmid encoding the
normal human insulin receptor [CHO/HIRc] was used in this
study and was a generous gift from Dr. Morris F. White, J oslin
Diabetes Center, Boston, MA. The cells were maintained in
F-12 medium containing 10% fetal bovine serum and were
cultured to confluence. Membranes from these cultured CHO/
HIRc cells were isolated and solubilized with Triton X-100,
essentially as described.31 Partially purified insulin receptors
from solubilized membranes were obtained after passing
through a wheat germ agglutin (WGA) column.32 WGA-
purified human insulin receptors were autophosphorylated
with [γ-32P]ATP as previously described,31 and this 32P-labeled
1
tion from methanol gave white needles: mp 113-114 °C; H
NMR (CDCl3) δ 7.71 (d, 2H, J ) 7.4 Hz), 7.51 (d, 2H J ) 7.3
Hz), 7.38-7.23 (m, 4H), 7.05 (d, 2H, J ) 8.3 Hz), 6.92 (d, 2H,
J ) 8.4 Hz), 5.10 (d, 1H, J ) 8.0 Hz), 4.59-4.44 (m, 1H), 4.43-
4.28 (m, 2H), 4.14 (t, 1H, J ) 6.7 Hz), 3.63 (s, 3H), 3.58-3.52

1026 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 5
Burke et al.
insulin receptor was used as substrate for the assay of PTP
1B activity, essentially following the method described.16
Refer en ces
(1) A prelinary account of this work as been presented: Burke, T.
R., J r.; Ye, B.; Akamatsu, M.; Yan, X.; Kole, H. K.; Wolf, G.;
Shoelson, S. E.; Roller, P. P. Non phosphorus-containing phos-
photyrosyl mimetics amenable to prodrug derivatization and
their use in solid-phase peptide synthesis of SH2 domain and
phosphatase inhibitors. 209th National American Chemical
Society Meeting, Anaheim, CA, April 2-7, 1995; MEDI 14.
(2) Boutin, J . A. Tyrosine protein kinase inhibition and cancer. Int.
J . Biochem. 1994, 26, 1203-1226.
(3) Levitzki, A.; Gazit, A. Tyrosine kinase inhibition: An approach
to drug development. Science 1995, 267, 1782-1788.
(4) Kavanaugh, W. M.; Williams, L. T. An alternative to SH2
domains for binding tyrosine-phosphorylated proteins. Science
1994, 266, 1862-1865.
(5) Bork, P.; Margolis, B. A phosphotyrosine interaction domain.
Cell 1995, 80, 693-694.
(6) Gustafson, T. A.; He, W. M.; Craparo, A.; Schaub, C. D.; Oneill,
T. J . Phosphotyrosine-dependent interaction of SHC and insulin
receptor substrate 1 with the NPEY motif of the insulin receptor
via a novel non-SH2 domain. Mol. Cell. Biol. 1995, 15, 2500-
2508.
(7) Eck, M. J . A new flavor in phosphotyrosine recognition. Struc-
ture 1995, 3, 421-424.
(8) Burke, T. R., J r. Protein-tyrosine kinase inhibitors. Drugs
Future 1992, 17, 119-131.
(9) Chang, C. J .; Geahlen, R. L. Protein-Tyrosine Kinase Inhibition
- Mechanism-Based Discovery of Antitumor Agents. J . Nat.
Prod. Lloydia 1992, 55, 1529-1560.
(10) Marseigne, I.; Roques, B. P. Synthesis of new amino acids
mimicking sulfated and phosphorylated tyrosine residues. J .
Org. Chem. 1988, 53, 3621-3624.
(11) Domchek, S. M.; Auger, K. R.; Chatterjee, S.; Burke, T. R., J r.;
Shoelson, S. E. Inhibition of SH2 Domain/Phosphoprotein As-
sociation by a Nonhydrolyzable Phosphonopeptide. Biochemistry
1992, 31, 9865-9870.
(12) Burke, T. R., J r.; Smyth, M. S.; Otaka, A.; Nomizu, M.; Roller,
P. P.; Wolf, G.; Case, R.; Shoelson, S. E. Nonhydrolyzable
phosphotyrosyl mimetics for the preparation of phosphatase-
resistant SH2 domain inhibitors. Biochemistry 1994, 33, 6490-
6494.
(13) Gilmer, T.; Rodriquez, M.; J ordan, S.; Crosby, R.; Alligood, K.;
Green, M.; Kimery, M.; Wagner, C.; Kinder, D.; Charifson, P.;
Hassell, A. M.; Willard, D.; Luther, M.; Rusnak, D.; Sternbach,
D. D.; Mehrotra, M.; Peel, M.; Shampine, L.; Davis, R.; Robbins,
J .; Patel, I. R.; Kassel, D.; Burkhart, W.; Moyer, M.; Bradshaw,
T.; Berman, J . Peptide inhibitors of src SH3-SH2-phosphoprotein
interactions. J . Biol. Chem. 1994, 269, 31711-31719.
(14) Chatterjee, S.; Goldstein, B. J .; Csermely, P.; Shoelson, S. E. In
Peptides: Chemistry and Biology; Rivier, J . E., Smith, J . A., Eds.;
Escom Science Publishers: Leiden, Netherlands, 1992; pp 553-
555.
(15) Zhang, Z. Y.; Maclean, D.; Mcnamara, D. J .; Sawyer, T. K.;
Dixon, J . E. Protein Tyrosine Phosphatase Substrate Specificity
- Size and Phosphotyrosine Positioning Requirements in Peptide
Substrates. Biochemistry 1994, 33, 2285-2290.
(16) Burke, T. R., J r.; Kole, H. K.; Roller, P. P. Potent inhibition of
insulin receptor dephosphorylation by a hexamer peptide con-
taining the phosphotyrosyl mimetic F(2)Pmp. Biochem. Biophys.
Res. Commun. 1994, 204, 129-134.
(17) Burke, T. R., J r.; Smyth, M.; Nomizu, M.; Otaka, A.; Roller, P.
P. Preparation of fluoro- and hydroxy-4-phosphonomethyl-D,L-
phenylalanine suitably protected for solid-phase synthesis of
peptides containing hydrolytically stable analogues of O-phos-
photyrosine. J . Org. Chem. 1993, 58, 1336-1340.
(18) Burke, T. R., J r.; Nomizu, M.; Otaka, A.; Smyth, M. S.; Roller,
P. P.; Case, R. D.; Wolf, G.; Shoelson, S. E. Cyclic peptide
inhibitors of phosphatidylinositol 3-kinase p85 SH2 domain
binding. Biochem. Biophys. Res. Commun. 1994, 201, 1148-
1153.
(19) Roller, P. P.; Otaka, A.; Nomizu, M.; Smyth, M. S.; Barchi, J . J .,
J r.; Burke, T. R., J r.; Case, R. D.; Wolf, G.; Shoelson, S. E.
Norleucine as a replacement for methionine in phosphatase-
resistant linear and cyclic peptides which bind to p85 SH2
domains. Bioorg. Med. Chem. Lett. 1994, 1879-1882.
(20) Ye, B.; Burke, T. R., J r. L-O-Malonyltyrosine (L-OMT) a New
Phosphotyrosyl Mimetic Suitably Protected for Solid-Phase
Synthesis of Signal Transduction Inhibitory Peptides. Tetra-
hedron Lett. 1995, 36, 4733-4736.
(21) Ye, B.; Akamatsu, M.; Wolf, G.; Shoelson, S. E.; Wolf, G.;
Giorgetti-Peraldi, S.; Yan, X.; Roller, P. P.; Burke, T. R., J r.
L-O-(2-Malonyltyrosine: A new phosphotyrosyl mimetic for the
preparation of Src homology 2 domain inhibitory peptides. J .
Med. Chem. 1995, 38, 4270-4275.
SH2 Dom a in Bin d in g Assa ys. Details of the SH2 domain
competition assay have been published previously.33 In this
study the relative p85 PI 3-kinase C-terminal SH2 domain
affinity was determined for the FOMT-containing peptide 22
vs a high-affinity phosphopeptide ligand. The glutathione
S-transferase (GST)/SH2 domain fusion protein was paired
with [125I]Bolton-Hunter-radiolabeled phosphopeptide corre-
sponding to the sequence derived from IRS-1 pY628, GNGD-
pYMPMSPKS,33 and varying concentrations of unlabeled 20
were added as competitors. An underline denotes the position
of the [125I]Bolton-Hunter-modified lysine. GST/SH2 domain
fusion proteins (0.5-1.0 µM, estimated by Bradford assay), 35
fmol of HPLC-purified, [125I]Bolton-Hunter-treated phospho-
peptide (67 nCi), and varying concentrations of 22 were
combined in 150 µL total volume of 20 mM Tris-HCl, 250 mM
NaCl, 0.1% bovine serum albumin, 10 mM dithiothreitol, pH
7.4, and vortexed. Glutathione-agarose (50 µL of a 1:10
aqueous slurry, Molecular Probes) was added, and the samples
were incubated overnight at 22 °C with constant mixing.
Following centrifugation for 5 min at 12000g, supernatant
solutions were removed by aspiration, and [125I]radioactivity
associated with the unwashed pellets was determined with a
γ-counter.
p Ka Deter m in a tion . Malonic acid was obtained from
Aldrich Chemical Co. and was used as a reference compound.
Deionized H2O (HPLC grade) and a standard solution of NaOH
(0.1 N) were purchased from Fischer Scientific. A Beckman
Model Φ45 pH meter with a combination epoxy electrode was
used to measure pH. The test compounds malonic acid, 12,
and 13 were accurately weighed (approximately 0.01 mmol)
and dissolved in 10.0 mL of deionized H2O using a magnetic
stirrer. Each solution was titrated by dispensing 10 µL
aliquots of 0.1 N NaOH using a 20 µL pipetman. The pH was
measured after each addition. The pH profiles for each
analogue were then plotted and used to determine titration
equivalents. Titration equivalents are obtained at the maxima
of the first derivative of the titration curve. The log of the
ionization constant (pKa) was determined by treating poten-
tiometric titration data according to the method of Noyes. This
method is used for compounds containing more than one
titratable group with pKa values differing by less than 2.7
units.34
Molecu la r Mod elin g. Modeling was conducted as previ-
ously described.22 The structure of a malonate group [CH2-
(CO2)2], complexed within the catalytic site of the PTP 1B
enzyme, was first defined using the geometry of the binding
site and mode of binding of a phosphonate derived from X-ray
crystallographic data of the difluorophosphonate-containing
inhibitor [1,1-difluoro-1-(2-naphthyl)methyl]phosphonic acid
bound within the PTP 1B catalytic site.28 The model was then
minimized by an ab initio method using a 3-21G basis set on
Cray and CONVEX mainframe computers using GAUSSIAN
92.35 In minimizing the complex of the PTP 1B with the
malonate structure, not only the geometry parameters and
position of the malonate were optimized but also the geometry
of the enzyme structure within the binding site during the first
50 h of CPU time. In order to examine potential alterations
in binding interactions afforded by introduction of a fluorine
onto the malonyl R-methylene, one hydrogen was replaced by
a fluorine atom, with corresponding changing of the C-F bond
length. The results of this modification are shown in Figure
2. From molecular mechanics (CHARMm), binding of benzene-
OCF(CO2)2 to PTP 1B is 2 kcal/mol stronger than benzeneOCF-
(CO2)2.
Ack n ow led gm en t. Appreciation is expressed to Ms.
Pamela Russ and Dr. J ames Kelley of the LMC for mass
spectral analysis. S.E.S. is the recipient of a Burroughs
Welcome Fund Scholar Award in Experimental Thera-
peutics.
(22) Kole, H. K.; Ye, B.; Akamatsu, M.; Yan, X.; Barford, D.; Roller,
P. P.; Burke, T. R., J r. Protein-tyrosine phosphatase inhibition
by a peptide containing the phosphotyrosyl mimetic, O-malo-
nyltyrosine (OMT). Biochem. Biophys. Res. Commun. 1995, 209,
817-822.

4′-O-[2-(2-Fluoromalonyl)]-L-tyrosine
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 5 1027
(23) Ledon, H. An improved preparation of alpha-diazocarbonyl
compounds. Synthesis 1974, 347-348.
(24) Corey, S. D.; Pansegrau, P. D.; Walker, M. C.; Sikorski, J . A.
EPSP synthase inhibitor design III. Synthesis & evaluation of
(30) Albericio, F.; Kneib-Cordonier, N.; Biancalana, S.; Gera, L.;
Masada, R. I.; Hudson, D.; Barany, G. Preparation and applica-
tion of the 5-(4-(9-fluorenylmethoxycarbonyl)aminomethyl-3,5-
dimethoxyphenoxy)-valeric acid (PAL) handle for the solid-
phasesynthesis of C-terminal peptide amides under mild
conditions. J . Org. Chem. 1990, 55, 3730-3743.
(31) Liotta, A. S.; Kole, H. K.; Fales, H. M.; Roth, J .; Bernier, M. A
synthetic tris-sulfotyrosyl dodecapeptide analogue of the insulin
receptor 1146-kinase domain inhibits tyrosine dephosphoryl-
ation of the insulin receptor in situ. J . Biol. Chem. 1994, 269,
22996-23001.
(32) Brillon, D. J .; Henry, R. R.; Klein, H. H.; Olefsky, J . M.;
Freidenberg, G. R. Endocrinology 1988, 123, 1837-1847.
(33) Piccione, E.; Case, R. D.; Domchek, S. M.; Hu, P.; Chaudhuri,
M.; Backer, J . M.; Schlessinger, J .; Shoelson, S. E. Phosphati-
dylinositol 3-kinase p85 SH2 domain specificity defined by direct
phosphopeptide/SH2 domain binding. Biochemistry 1993, 32,
3197-3202.
(34) Noyes. Physik Chem. 1893, 11, 495.
(35) Gaussian 92. Gaussian, Inc., Carnegie Office Park, Building 6,
Pittsburgh, PA 15106.
a
new 5-oxamine acid analog of EPSP which incorporates
amalonate ether as a 3-phosphate mimic. Bioorg. Med. Chem.
Lett. 1993, 3, 2857-2862.
(25) Burke, T. R., J r.; Smyth, M. S.; Otaka, A.; Roller, P. P. Synthesis
of 4-Phosphono(Difluoromethyl)-D,L-Phenylalanine and N-Boc
and N-Fmoc Derivatives Suitably Protected for Solid-Phase
Synthesis of Nonhydrolyzable Phosphotyrosyl Peptide Ana-
logues. Tetrahedron Lett. 1993, 34, 4125-4128.
(26) Unpublished results.
(27) Smyth, M. S.; Ford, H., J r.; Burke, T. R., J r. A general method
for the preparation of benzylic alpha, alpha-difluorophosphonic
acids; non-hydrolyzable mimetics of phosphotyrosine. Tetrahe-
dron Lett. 1992, 33, 4137-4140.
(28) Burke, T. R., J r.; Barford, D., manuscript in preparation.
(29) J ia, Z. C.; Barford, D.; Flint, A. J .; Tonks, N. K. Structural basis
for phosphotyrosine peptide recognition by protein tyrosine
phosphatase 1B. Science 1995, 268, 1754-1758.
J M950621G