Investigating protein-ligand interactions with a mutant FKBP possessing a designed specificity pocket

Article abstract of DOI:10.1021/jm9904396

Using structure-based design and protein mutagenesis we have remodeled the FKBP12 ligand binding site to include a sizable, hydrophobic specificity pocket. This mutant (F36V-FKBP) is capable of binding, with low or subnanomolar affinities, novel synthetic ligands possessing designed substituents that sterically prevent binding to the wild-type protein. Using binding and structural analysis of bumped compounds, we show here that the pocket is highly promiscuous - capable of binding a range of hydrophobic alkyl and aryl moieties with comparable affinity. Ligand affinity therefore appears largely insensitive to the degree of occupancy or quality of packing of the pocket. NMR spectroscopic analysis indicates that similar ligands can adopt radically different binding modes, thus complicating the interpretation of structure-activity relationships.

Full text of DOI:10.1021/jm9904396

J . Med. Chem. 2000, 43, 1135-1142
1135
In vestiga tin g P r otein -Liga n d In ter a ction s w ith a Mu ta n t F KBP P ossessin g a
Design ed Sp ecificity P ock et
Wu Yang, Leonard W. Rozamus, Surinder Narula, Carl T. Rollins, Ruth Yuan, Lawrence J . Andrade,
Mary K. Ram, Thomas B. Phillips, Marie Rose van Schravendijk, David Dalgarno,
Tim Clackson,* and Dennis A. Holt*
ARIAD Gene Therapeutics, 26 Landsdowne Street, Cambridge, Massachusetts 02139-4234
Received August 25, 1999
Using structure-based design and protein mutagenesis we have remodeled the FKBP12 ligand
binding site to include a sizable, hydrophobic specificity pocket. This mutant (F36V-FKBP) is
capable of binding, with low or subnanomolar affinities, novel synthetic ligands possessing
designed substituents that sterically prevent binding to the wild-type protein. Using binding
and structural analysis of bumped compounds, we show here that the pocket is highly
promiscuousscapable of binding a range of hydrophobic alkyl and aryl moieties with comparable
affinity. Ligand affinity therefore appears largely insensitive to the degree of occupancy or
quality of packing of the pocket. NMR spectroscopic analysis indicates that similar ligands
can adopt radically different binding modes, thus complicating the interpretation of structure-
activity relationships.
In tr od u ction
tial limitations, we designed new ligands which have
greatly reduced affinity for wild-type FKBP12, but
which possess high affinity for minimally mutated
FKBP domains of the chimeric receptor proteins.9
Traditionally, our understanding of how small mol-
ecules bind protein targets has come from structure-
activity relationships, derived from the binding of series
of iteratively modified compounds to a single protein.
These studies have revealed the importance of features
such as shape complementarity, hydrogen bonding, and
ligand preordering.1 More recently, the ability to specif-
ically mutate proteins using recombinant DNA technol-
ogy has allowed an extension of this approach: both the
protein target and its small molecule ligand can be
modified, for example, to insert a cavity into the ligand
The process of metamorphosing both target protein
and ligand to achieve a novel pattern of selectivity began
with the aid of structural modeling. These studies
suggested that pyramidalization and appropriate sub-
stitution of C9 of a typical pipecolyl R-ketoamide ligand
(e.g., 117) would introduce a sterically encumbering
“bump” that would prevent binding to wild-type FKBP,
but which might be accommodated by a compensatory
“hole” created in the protein by a single mutation of
Phe36 to a smaller hydrophobic amino acid (Figure 1).10
Indeed, compound 2S was shown to have exceptional
affinity for F36V-FKBP with a Kd of ca. 100 pM while
having negligible affinity for wild-type FKBP12.9 Sub-
sequent structure determination of the F36V-FKBP/2S
complex by X-ray crystallography confirmed the pre-
dicted binding mode, showing that the C9 ethyl sub-
stituent (the bump) partially, but not completely, fills
the 90 Å3 hole created by the F36V mutation (Figure
2).9
binding site and design
a complementary ligand
“bump”.2-10 Analysis of such remodeled complexes
might provide new insights into the factors important
to binding affinity and specificity and may also guide
attempts to design drugs for naturally “mutated” pro-
teins, such as homologous family members11,12 or protein
targets that mutate during the evolution of microbial
drug resistance.
We have been remodeling the high-affinity interface
between FKBP12 and relatively simple ligands to
develop improved chemical inducers of dimerization
(CIDs)13 for directing protein-protein association inside
cells. Bivalent dimers of wholly synthetic ligands which
are capable of simultaneously binding two FKBP mol-
ecules, and thus able to dimerize or oligomerize chimeric
proteins containing FKBP domains, have been used to
control signal transduction and gene expression events
in genetically engineered cells.14-16 Because these com-
pounds bind equally well to endogenous FKBP12 and
to their intended FKBP-fusion targets, they hypotheti-
cally could interfere with the normal physiological
functions of FKBP. In addition, these compounds might
suffer reduced potency due to sequestration by the
ubiquitous wild-type protein. To overcome these poten-
In this study we set out to probe molecular recognition
in this engineered binding site by determining the
specificity of the F36V hole for various bumps. We
prepared a series of R-substituted trimethoxyphenylace-
tamide and phenylacetamide derivatives with bumps
ranging in size from hydrogen (no bump) to cyclohexyl
and analyzed their binding to F36V-FKBP by using a
combination of affinity measurements and NMR struc-
tural studies.
Resu lts a n d Discu ssion
Tr im eth oxyp h en yla ceta m id e Ser ies. We first ex-
amined binding of the trimethoxyphenylacetamide ser-
ies to the mutated protein. Initial molecular modeling
* To whom correspondence should be addressed. Phone: 617-494-
0400. Fax: 617-225-2589. E-mail: holtd@tiac.net, clackson@ariad.com.
10.1021/jm9904396 CCC: $19.00 © 2000 American Chemical Society
Published on Web 02/25/2000

1136 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 6
Yang et al.
F igu r e 2. X-ray crystal structure of 2S bound to F36V-FKBP.9
Portions of the calculated solvent accessible surfaces of the
protein and ligand are shown as solid green surface and orange
mesh, respectively.
F igu r e 1. X-ray crystal structure of 1 bound to FKBP12.17
cut-away view of the calculated solvent accessible surface of
the protein in the ligand binding region is shown in green.
A
studies suggested that C9 alkyl substituents in the S
configuration would more optimally engage the hole
created by the F36V mutation than those of the R
configuration. This expectation was born out as 2S
exhibits a 100-fold higher affinity than its C9-epimeric
analogue (2R, Table 1).9 That compound 2R retains
reasonable affinity for F36V-FKBP, while having no
detectable affinity for the wild-type protein, suggests
that the R-configured ethyl bump is still accommodated
by the engineered protein hole, albeit at an energetic
cost of nearly 3 kcal/mol. NMR structure determinations
of the 2S and 2R complexes with F36V-FKBP revealed
similar ligand conformations relative to each other
(Figure 3, 2S and 2R) and to that observed previously9
in the crystal structure of 2S (Figure 2). The pipecolyl
and trimethoxyphenyl rings of 2R and 2S are nearly
superimposable. The latter occupies a natural pockets
defined by residues Asp37, His87, and Ile90swhich also
binds the tetrahydropyranyl (THP) moiety of FK506 and
rapamycin. Whereas the ethyl group of 2S has the most
direct approach into the F36V hole, Figure 3 illustrates
that the C9 ethyl substituent of 2R is also able to engage
the hole. Presumably, reduced contact of the R-ethyl
group with the hydrophobic hole and/or a degree of
steric repulsion with Tyr26 leads to the reduced affinity
of 2R relative to 2S.
analogue (8S) is reduced approximately 2-fold which is
still some 15-fold better than the unsubstituted (un-
bumped) analogue (7). Incorporation of isopropyl as the
bump (10S) results in a 4-fold loss in affinity relative
to the ethyl (9S), similar to the difference observed
between 2S and 5S, possibly owing to unfavorable steric
interactions of the branched isopropyl side chain at the
mouth of the F36V hole. Model building suggests that
the engineered hole is capable of accommodating a
cyclohexyl substituent with only minimal distortion of
the protein. Although similar steric clashes are likely
to occur for cyclohexyl as for isopropyl, the additional
hydrophobic interactions available to the cyclohexyl
presumably accounts for a 2-fold binding advantage
relative to the isopropyl analogue and restores the
affinity of 11S to within 1.5-fold of the ethyl analogue
(9S). As observed for the S series of trimethoxyphenyl
congeners, the affinities of the phenyl analogues are
largely independent of the size of the bump. Only a
4-fold range of IC50s is seen in the methyl to cyclohexyl
series. At the extreme, incorporation of phenyl as the
bump (12) results in only a 7-fold reduction in affinity
in comparison to 9S.
Examination of the R-congeneric series reveals a
remarkably different structure-activity trend. Whereas
the R-methyl analogue (8R) and R-ethyl analogue (9R)
are both less active than their S-configured analogues
(8S and 9S)sas the original bump-hole design had
predictedsa reversal in the stereochemical preference
is seen for the R-isopropyl and R-cyclohexyl analogues.
Both 10R and 11R exhibit roughly 6-fold higher affinity
than their corresponding S analogues. The explanation
for this reversal is apparent from the NMR structures
Recognizing that some 50-60 Å3 of the F36V hole
remains unoccupied while coordinated with 2S, a series
of analogues with progressively larger bumps was
prepared in an attempt to enhance affinity by more
completely filling the engineered hole. In fact, the
binding of the S-trimethoxyphenyl analogues appears
to be remarkably insensitive to modest changes to the
bump (Table 1). The corresponding R analogues show
somewhat greater enhancements to binding with in-
creasing bump size. The R-allyl compound (3R) exhibits
the highest affinity of the subset having 10-fold higher
affinity than the R-ethyl compound (2R). In keeping
with the original model of binding, for each diastereo-
meric pair the S analogue is preferred, implying that
the overall binding mode of 3-6 is the same as 2.
P h en yla ceta m id e Ser ies. Previously, we showed
that C9 bump moieties could be combined with several
suitable “bottom” groups in place of trimethoxyphenyl.
The phenyl for trimethoxyphenyl substitution results
in a loss of affinity (20-fold for 2S vs 9S), but specificity
is retained.9 Therefore, we also examined the specificity
of bump moieties to F36V-FKBP in the more readily
prepared phenyl congeneric series. In the S series, the
ethyl analogue (9S) exhibits the highest affinity for
F36V-FKBP (Table 2). The affinity of the methyl
Ta ble 1. Trimethoxyphenylacetamide Series: Binding
Affinities (IC50s, nM) for F36V-FKBP
R series RS
H RR ) bump
)
S series RS )
bump RR ) H
compd no.
bump
Et
allyl
n-Pr
i-Pr
c-PrCH2
2 (R,S)
3 (R,S)
4 (R,S)
5 (R,S)
6 (R,S)
187
19
43
65
126
1.8
1.5
3.8
5.8
4.4

FKBP12 Ligand Binding Site with Specificity Pocket
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 6 1137
F igu r e 3. NMR-derived structural models of 2S, 2R, 9S, 9R, 10S, and 10R bound to F36V-FKBP. Carbon atoms of the ligands
are shown in yellow; C9 hydrogen atoms are shown in cyan. Portions of the calculated solvent accessible surfaces are shown as
solid green surfaces.
Ta ble 2. Phenylacetamide Series: Binding Affinities (IC50s,
nM) for F36V-FKBP
Sch em e 1
R series RS
H RR ) bump
)
S series RS )
bump RR ) H
compd no.
bump
(methyl < ethyl < isopropyl ≈ cyclohexyl) to bind in the
larger THP-binding pocket while the phenyl group is
engaging the engineered hole. This SAR trend for alkyl
groups occupying the THP-binding pocket is identical
to that reported for simple, unbumped FKBP12 ligands.18
The affinity of the C9-symmetrical diphenyl analogue
(12) for F36V-FKBP suggests that the THP-binding
pocket prefers occupation by the branched alkyls, iso-
propyl and cyclohexyl, over phenyl by some 15- to 20-
fold (12 vs 10R and 11R), whereas the engineered F36V
hole has a lower preference for those same branched
alkyls over phenyl by a factor of only 2- to 4-fold (12 vs
10S and 11S). Thus, in comparing 10R vs 10S or 11R
vs 11S, F36V-FKBP preferentially binds the R ana-
logues in order to place the alkyl groups in the THP-
binding pocket and the phenyl moiety in the engineered
hole.
Syn t h esis. The pipecolyl trimethoxyphenylacet-
amides of Table 1 were prepared by conventional means
as illustrated in Scheme 1. Enantiomerically pure
diarylpropanol A was previously described. Following
coupling of B and C, diastereomers were separated
chromatographically prior to final tert-butyl ester depro-
tection. The analogous phenylacetamides of Table 2
were prepared similarly.
7
H
Me
Et
i-Pr
c-Hx
Ph
1530
273
8 (R,S)
9 (R,S)
10 (R,S)
11 (R,S)
12
590
133
19
94
40
150
66
12
of ethyl analogues 9R and 9S and isopropyl analogues
10R and 10S (Figure 3). The R series bind in a nearly
identical fashion as the S series with the exception of a
120° torsional rotation about the C8-C9 bond. For the
R compounds this leads to a positioning of the phenyl
C9 substituent into the F36V engineered hole and
allows the variable bump substituent to take station in
the now-vacated THP-binding pocket. Thus, despite the
apparent inconsistency in SAR, the R- and S-phenyl-
acetamide derivatives adopt a common orientation
where the C9 proton is always pointing toward Phe99.
That 2R, and presumably other R-trimethoxyphenyl-
acetamide analogues, adopts a different binding modes
one where the C9 hydrogen is pointed toward Val36
(Figure 3)ssuggests that the trimethoxyphenyl moiety
is too large to enter the engineered hole and remains
fixed in the THP-binding pocket. The divergent SAR for
the R series of phenylacetamide analogues (8-11)
reflects the different propensities for the alkyl groups

1138 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 6
Yang et al.
Con clu sion s
built from the FKBP ligands described here may be
more efficient at activating transcription or signal
transduction events, by virtue of their specific interac-
tion with target FKBP fusion proteins and not endog-
enous FKBP.9,15,16 Our results provide guidance for the
reengineering of other binding partners for dissecting
and controlling complex biological pathways. In addi-
tion, proteins with binding sites containing engineered
cavities can be thought of as model systems for natural
protein drug targets that closely resemble other, non-
target proteins (such as homologues within a protein
family11,12). Our results imply that specific, high-affinity
ligands for such natural variants may be more easily
obtained than was previously appreciated.
At the extremes of classic protein-ligand recognition
paradigms are the lock-and-key and induced-fit models.
In each case, whether the protein is assumed to be quite
rigid as in the former or to have greater plasticity as in
the latter, a highly complementary packing of protein
and ligand has been thought to be characteristic of high-
affinity interactions. In this and a preceding study we
have shown that such precision in fit is not required,
at least for the hydrophobic interface we studied. In this
example, a single F36V mutation in FKBP12 opens a
cavity of approximately 90 Å3 at the base of the natural
binding site. This engineered hole, designed as a speci-
ficity pocket, is remarkably stable and retains its shape
even when unoccupied (as shown by a crystal structure
of the protein bound to FK50619). While the binding of
ligands possessing specificity elements (bumps) relies
on the presence of this hole, complete filling of the hole
by the bump is not required for high-affinity interac-
tions. As a consequence, this specificity pocket appears
to be surprisingly promiscuous in the accommodation
of nonpolar ligand motifs ranging in size from methyl
to cyclohexyl.
Despite examples to the contrary,20 a nearly unavoid-
able assumption in the interpretation of structure-
activity relationships is that common ligand cores result
in common binding modes. Here, we studied ligand
binding preferences of two pockets within a binding
sitesone natural (THP) and the other an engineered
cavity. Unexpectedly, we found that the binding mode
of compounds can “flip”, positioning the moiety targeted
for the engineered cavity in the THP-binding pocket and
vice versa. This observation underscores the need for
caution and the utility of spectroscopic verification in
protein-binding structure-activity studies of congeneric
small molecules.
Although F36V-FKBP is the first remodeled protein-
small molecule interface to be structurally elucidated,
promiscuous recognition and a toleration of imperfectly
packed nonpolar interactions may prove to be general
features of such interfaces.10 Models of novel pairings
between cyclophilin and cyclosporin7,8,21 and between
tyrosine kinases and ATP2-6 generally indicate that
unfilled cavities would remain after ligand binding,
unless protein structural adjustments occurred.10 One
study of “bumped” pyrazolo[3,4-d]pyrimidine analogues
binding to mutated tyrosine kinases found a variety of
nonpolar cyclic substituents were tolerated in the
engineered hole with similar affinities, albeit in the
micromolar range.6 Structural studies similar to those
described here will be required to determine whether
this behavior involves protein adjustments or changes
in binding mode. Nevertheless, at least for the hydro-
phobic cavities engineered to date, it is clear that tightly
packed complementarity is not a prerequisite for high-
affinity binding.
Exp er im en ta l Section
(1R)-3-(3,4-Dim eth oxyp h en yl)-1-[3-(t-bu toxyca r bon yl-
m eth oxy)p h en yl]-1-p r op yl (2S)-1-(9-F lu or en ylm eth oxy-
ca r bon yl)-2-p ip er id in eca r boxyla te. A solution of (R)-1-(3-
(t-butoxycarbonylmethoxy)phenyl)-3-(3,4-dimethoxyphenyl)
propan-1-ol15,16 (A, 3.1 g, 7.7 mmol) in CH2Cl2 (40 mL) was
treated with N-Fmoc-L-pipecolic acid (3.0 g, 8.5 mmol) followed
by 1,3-dicyclohexyl carbodiimide (DCC, 1.9 g, 9.2 mmol) and
4-(dimethylamino)pyridine (DMAP, 560 mg, 4.6 mmol) under
a nitrogen atmosphere. The resulting white suspension was
allowed to stir overnight at ambient temperature. The reaction
mixture was then filtered, evaporated, and flash chromato-
graphed (silica gel, 15 to 20% EtOAc/hexanes) to afford 4.7 g
(83%) of the Fmoc-protected pipecolyl ester as a white foam:
IR (neat) 2939, 1746, 1702, 1516, 1451, 1258, 1154, 760, 740
1
cm-1; H NMR (CDCl3, 300 MHz) 7.73 (m, 2H), 7.59 (t, J )
6.6 Hz, 1H), 7.16-7.49 (m, 6H), 6.94 (d, J ) 7.6 Hz, 1H), 6.89
(s, 1H), 6.72-6.82 (m, 2H), 6.62 (m, 2H), 5.76 (br s, 1H), 5.02
(d, J ) 3.7 Hz, 1H), 4.25-4.49 (m, 5H), 4.07-4.14 (m, 1H),
3.83 (s, 6H), 3.14 (t, J ) 11.1 Hz, 1H), 2.46-2.54 (m, 2H), 2.16-
2.33 (m, 2H), 2.00-2.07 (m, 1H), 1.68-1.78 (m, 4H), 1.46 (s,
9H), 1.39-1.56 (m, 1H); 13C NMR (CDCl3, 75 MHz) 174.50,
171.33, 168.28, 158.48, 147.73, 144.30, 142.12, 133.90, 130.07,
128.07, 127.45, 125.48, 120.50, 120.35, 114.34, 113.66, 112.12,
111.74, 82.74, 76.82, 76.59, 68.20, 66.16, 56.32, 56.20, 47.63,
38.44, 31.98, 31.54, 28.42, 27.23, 25.18, 21.20; MS (FAB) (M
+ Na)+ 758.69.
(1R)-3-(3,4-Dim eth oxyp h en yl)-1-[3-(t-bu toxyca r bon yl-
m eth oxy)p h en yl]-1-p r op yl (2S)-2-P ip er id in eca r boxyla te
(B). A solution of the above Fmoc-protected pipecolyl ester (833
mg, 1.13 mmol) in CH2Cl2 (30 mL) was treated with piperidine
(1.12 mL, 11.3 mmol), and the mixture was stirred overnight
at ambient temperature. The reaction mixture was concen-
trated and flash chromatographed (silica gel, 50 to 100%
EtOAc/hexanes) to afford the free amine (569 mg, 98%) as a
white sticky foam: IR (neat) 2935, 1748, 1590, 1516, 1453,
1153, 1030, 733 cm-1; 1H NMR (CDCl3, 300 MHz) (mixture of
rotamers, data for trans rotamer) 7.28 (t, J ) 7.9 Hz, 1H), 6.98
(d, J ) 7.7 Hz, 1H), 6.93 (s, 1H), 6.84 (m, 2H), 6.71 (d, J ) 8.3
Hz, 1H), 6.69 (s, 1H), 5.77 (dd, J ) 6.3, 6.8 Hz, 1H), 4.55 (s,
2H), 3.91 (s, 6H), 3.42 (m, 1H), 3.33 (s, 1H), 3.01 (m, 1H), 2.39-
2.63 (m, 3H), 2.11-2.27 (m, 1H), 2.05-2.09 (m, 1H), 1.92 (m,
1H), 1.54 (s, 9H), 1.54-1.74 (m, 4H); 13C NMR (CDCl3, 75
MHz) 173.42, 168.29, 158.34, 149.20, 147.64, 142.54, 134.06,
129.92, 120.49, 120.19, 114.11, 113.54, 111.99, 111.60, 82.74,
75.21, 66.05, 61.45, 56.30, 56.21, 48.57, 38.55, 31.63, 29.41,
28.44, 25.70, 22.56; MS (FAB) (M + H)+ 514.30.
Gen er a l P r oced u r e for 2-Alk yl Tr im eth oxyp h en yla ce-
tic Acid Syn th esis (C). Sodium bis(trimethylsilyl)amide (40
mL, 1.0 M solution in THF) was added slowly to a stirring
solution of 3,4,5-trimethoxyphenylacetic acid (4.0 g, 17.7 mmol)
in 20 mL of dry THF at 0 °C under nitrogen. After 15 min,
alkyl bromide or iodide (21.3 mmol) was added. The solution
was allowed to warm to room temperature. After 12 h, the
mixture was diluted with 100 mL of ethyl acetate and acidified
by careful addition of 60 mL of 1.0 N HCl. The layers were
separated, and the aqueous phase was extracted with an
Reengineered (orthogonal) protein-small molecule
pairs are potentially powerful tools for a variety of
biological problems because they allow an otherwise
nonspecific small molecule to be restricted to a single
(engineered) target protein inside cells.10 Thus, the
consequences of an inhibition,3,6,8 activation,9,22 or en-
zyme turnover event2,5,23 can be studied and/or con-
trolled in isolation. For example, chemical dimerizers

FKBP12 Ligand Binding Site with Specificity Pocket
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 6 1139
additional 50 mL of ethyl acetate. The organic portions were
combined and washed with water (2 × 10 mL) and brine (10
mL), dried over anhydrous MgSO4, concentrated, and subjected
to flash chromatography (2.5% AcOH in 1:1 EtOAc/hexane)
to yield the desired acid as a white solid.
694.3197; HPLC retention time ) 10.89 min (normal phase),
7.24 min (reverse phase).
Com p ou n d 2R . IR (neat) 2939, 1738, 1590, 1514, 1454,
1
1240, 1129, 1017, 913, 731 cm-1; H NMR (CDCl3, 300 MHz)
(mixture of rotamers) 7.25 (t, J ) 7.8 Hz, 1H), 6.70-6.90 (m,
6H), 6.45 (s, 2H), 5.66-5.70 (m, 1H), 5.53 (d, J ) 4.5 Hz, 1H),
4.59-4.72 (m, 2H), 3.86 (s, 3H), 3.85 (s, 3H), 3.84 (s, 3H), 3.83
(s, 6H), 3.47-3.79 (m, 2H), 3.23-3.32 (m, 1H), 2.52-2.71 (m,
2H), 2.00-2.31 (m, 4H), 1.52-1.75 (m, 4H), 1.28-1.46 (m, 2H),
0.83 (t, J ) 7.3 Hz, 3H); 13C NMR (CDCl3, 75 MHz) 174.3,
172.0, 171.0, 158.6, 154.0, 149.6, 148.1, 142.7, 137.6, 136.3,
134.0, 130.3, 120.8, 120.2, 116.2, 112.3, 112.0, 110.9, 105.5,
77.8, 77.1, 66.0, 61.5, 56.9, 56.6, 56.5, 53.2, 51.6, 44.0, 38.7,
32.0, 28.4, 27.5, 25.4, 21.5, 12.9; HRMS (FAB) (M + H)+ calc.
694.3227, meas. 694.3245; HPLC retention time ) 13.64 min
(normal phase), 7.09 min (reverse phase).
1
2-Eth yl-(3,4,5-tr im eth oxyp h en yl)a cetic Acid . H NMR
(CDCl3, 300 MHz) 6.53 (s, 2H), 3.85 (s, 6H), 3.83 (s, 3H), 3.39
(t, J ) 7.6 Hz, 1H), 2.14-2.04 (m, 1H), 1.85-1.75 (m, 1H),
0.93 (t, J ) 7.2 Hz, 3H); MS (ES)- 254.
2-Allyl-(3,4,5-tr im eth oxyp h en yl)a cetic Acid . 1H NMR
(CDCl3, 300 MHz) 6.54 (s, 2H), 5.81-5.67 (m, 1H), 5.11 (d, J
) 15.6 Hz, 1H), 5.07 (d, J ) 6.9 Hz, 1H), 3.85 (s, 6H), 3.83 (s,
3H), 3.57 (dd, J ) 7.0, 8.5 Hz, 1H), 2.85-2.75 (m, 1H), 2.56-
2.47 (m, 1H); MS (FAB) (M - H)- 265.14.
2-P r op yl-(3,4,5-tr im eth oxyp h en yl)a cetic Acid . 1H NMR
(CDCl3, 300 MHz) 6.54 (s, 2H), 3.85 (s, 6H), 3.83 (s, 3H), 3.48
(t, J ) 7.6 Hz, 1H), 2.07-2.01 (m, 1H), 1.78-1.71 (m, 1H),
1.37-1.26 (m, 2H), 0.98-0.86 (m, 3H); MS (FAB) (M - H)-
267.09.
Com p ou n d 3S. IR (neat) 2939, 1739, 1590, 1514, 1453,
1240, 1127 cm-1 1H NMR (CDCl3, 300 MHz) (mixture of
;
rotamers) 7.19 (t, J ) 7.8 Hz, 1H), 6.84-6.66 (m, 6H), 6.44
(brs, 1H), 6.33 (s, 2H), 5.84-5.71 (m, 1H), 5.56 (dd, J ) 5.1,
8.5 Hz, 1H), 5.47 (d, J ) 5.6 Hz, 1H), 5.04 (d, J ) 18.3 Hz,
1H), 4.99 (d, J ) 10.1 Hz, 1H), 4.65 (AB, J ) 16.4 Hz, 2H),
3.85 (s, 3H), 3.84 (s, 3H), 3.92-3.67 (m, 2H), 3.78 (s, 3H), 3.61
(s, 6H), 2.92-2.76 (m, 2H), 2.65-2.30 (m, 4H), 2.21-1.95 (m,
2H), 1.76-1.42 (m, 5H); 13C NMR (CDCl3, 75 MHz) 173.2,
171.8, 170.6, 158.4, 153.6, 149.4, 147.8, 142.8, 137.1, 136.8,
134.5, 133.8, 129.9, 120.6, 119.8, 116.8, 115.4, 112.2, 111.8,
110.8, 105.4, 76.8, 65.7, 61.1, 56.5, 56.3, 52.7, 49.8, 43.9, 39.7,
38.7, 32.0, 27.4, 25.6, 21.2; HRMS (FAB) (M + H)+ calc.
706.3227, meas. 706.3249; HPLC retention time ) 10.91 min
(normal phase), 10.24 min (reverse phase).
2-Isop r op yl-(3,4,5-tr im eth oxyp h en yl)a cetic Acid . 1H
NMR (CDCl3, 300 MHz) 6.56 (s, 2H), 3.85 (s, 6H), 3.83 (s, 3H),
3.06 (d, J ) 10.7 Hz, 1H), 2.33-2.23 (m, 1H), 1.08 (d, J ) 6.5
Hz, 3H), 0.75 (d, J ) 6.7 Hz, 3H); MS (FAB) (M - H)- 267.09.
2-Cyclop r op ylm et h yl-(3,4,5-t r im et h oxyp h en yl)a cet ic
Acid . 1H NMR (CDCl3, 300 MHz) 11.32 (brs, 1H), 6.56 (s, 2H),
3.85 (s, 6H), 3.83 (s, 3H), 3.60 (t, J ) 7.5 Hz, 1H), 1.95-1.68
(m, 2H), 0.43 (m, 2H), 0.13-0.02 (m, 2H).
Gen er a l P r oced u r e for Liga n d Syn th esis. A solution of
the above amine (484 mg, 0.94 mmol) in CH2Cl2 (10 mL) was
treated with the appropriate phenylacetic acid analogue (1.9
mmol) followed by PyBroP (878 mg, 1.9 mmol) and diisopro-
pylethylamine (819 mL, 4.7 mmol) under a nitrogen atmos-
phere at ambient temperature. The resulting solution was
allowed to stir overnight. The reaction mixture was then
concentrated and flash chromatographed (silica gel, EtOAc/
hexanes/CH2Cl2) to afford the penultimate tert-butyl ester-
protected pipecolyl phenylacetamide as a white foam. A
solution of the tert-butyl ester (0.48 mmol) in CH2Cl2 (4 mL)
was treated with trifluoroacetic acid (0.74 mL, 9.6 mmol), and
the mixture was stirred at ambient temperature for 2 h. The
reaction mixture was then diluted with toluene (50 mL),
concentrated, and flash chromatographed (silica gel, EtOAc
with 1% acetic acid) to afford the acid as a white solid.
The diastereomeric tert-butyl ester precursors of compounds
8R and 8S and of 9R and 9S were prepared independently
from enantiomerically pure R-methylphenylacetic acids or
R-ethylphenylacetic acids, respectively. The diastereomeric
pairs of all other tert-butyl ester precursors were prepared as
mixtures from the corresponding racemic trimethoxyphen-
ylactic acids or phenylacetic acids and separated by flash
chromatography (as specified above) or by normal phase HPLC
prior to TFA deprotection. The stereochemistries of each pair
of diastereomers were assigned by comparison of their NMR
spectra with those of 8 and 9.
Com p ou n d 3R . IR (neat) 2937, 1737, 1590, 1512, 1453,
1239, 1128 cm-1 1H NMR (CDCl3, 300 MHz) (mixture of
;
rotamers) 7.26 (t, J ) 8.9 Hz, 1H), 6.94-6.68 (m, 7H), 6.47 (s,
2H), 5.79-5.67 (m, 2H), 5.54 (d, J ) 4.4 Hz, 1H), 5.00 (dd, J
) 1.4, 17.1 Hz, 1H), 4.93 (d, J ) 10.2 Hz, 1H), 4.65 (s, 2H),
3.86-3.61 (m, 17H), 3.22 (dt, J ) 2.3, 12.0 Hz, 1H), 2.86-
2.81 (m, 1H), 2.79-2.33 (m, 3H), 2.31-2.04 (m, 3H), 1.63-
1.30 (m, 5H); 13C NMR (CDCl3, 75 MHz) 173.2, 172.1, 170.9,
158.4, 153.8, 149.4, 147.8, 142.4, 137.5, 136.5, 135.6, 133.8,
130.1, 120.6, 120.1, 116.9, 115.3, 112.2, 112.0, 111.9, 105.3,
76.7, 65.6, 61.3, 56.7, 56.3, 53.0, 49.7, 43.8, 39.3, 38.5, 32.0,
27.1, 25.1, 23.0, 21.2; HRMS (FAB) (M + H)+ calc. 706.3227,
meas. 706.3210; HPLC retention time ) 13.59 min (normal
phase), 10.11 min (reverse phase).
Com p ou n d 4S. IR (neat) 2938, 1739, 1590, 1514, 1454,
1
1240, 1128, 1027 cm-1; H NMR (CDCl3, 300 MHz) (mixture
of rotamers) 9.53 (brs, 1H), 7.16 (t, J ) 7.9 Hz, 1H), 6.86-
6.64 (m, 6H), 6.30 (s, 2H), 5.53 (dd, J ) 5.0, 8.4 Hz, 1H), 5.46
(d, J ) 4.8 Hz, 1H), 4.63 (AB, J ) 16.4 Hz, 2H), 3.83 (s, 3H),
3.82 (s, 3H), 3.79-3.64 (m, 2H), 3.76 (s, 3H), 3.59 (s, 6H), 2.85
(t, J ) 13.1 Hz, 1H), 2.65-2.45 (m, 2H), 2.31 (d, J ) 13.1 Hz,
1H), 2.17-1.95 (m, 3H), 1.72-1.51 (m, 4H), 1.44-1.24 (m, 4H),
0.88 (t, J ) 7.3 Hz, 3H); 13C NMR (CDCl3, 75 MHz) 173.9,
171.9, 170.6, 158.4, 153.6, 149.4, 147.8, 142.8, 136.4, 135.4,
133.8, 129.9, 120.6, 119.9, 115.4, 112.2, 111.8, 110.8, 105.3,
76.7, 65.8, 61.1, 56.3, 52.7, 49.4, 43.9, 38.6, 37.7, 31.8, 27.4,
25.6, 21.4, 14.4; HRMS (FAB) (M + H)+ calc. 708.3384, meas.
708.3365; HPLC retention time ) 9.55 min (normal phase),
10.75 min (reverse phase).
Compound homegeneity was assessed utilizing both normal
and reverse phase HPLC analysis: normal phase HPLC (250
mm × 4.6 mm Chiralcel OD column, 35% 2-propanol:65%
hexanes with 0.2% formic acid, 1.5 mL/min, ambient temper-
ature, detection at 254 nm); reverse phase HPLC (250 mm ×
4.6 mm Vydac 218TP54 column, gradient elution at 1 mL/
min: 1 min- 55% A (90% CH3CN/H2O with 0.1% TFA), 45% B
(H2O with 0.1% TFA); 31 min- 80% A; 32 min- 100% A, column
temperature of 60 °C, detection at 254 nm).
Com p ou n d 4R . IR (neat) 2936, 1738, 1590, 1514, 1455,
1330, 1240, 1128, 1012 cm-1 1H NMR (CDCl3, 300 MHz)
;
(mixture of rotamers) 9.98 (brs, 1H), 7.23 (t, J ) 7.7 Hz, 1H),
6.91-6.65 (m, 6H), 6.45 (s, 2H), 5.71 (dd, J ) 5.4, 7.9 Hz, 1H),
5.53 (d, J ) 4.4 Hz, 1H), 4.62 (s, 2H), 3.83 (s, 6H), 3.80 (s,
9H), 3.75-3.59 (m, 2H), 3.19 (dt, J ) 13.0, 2.1 Hz, 1H), 2.61-
2.50 (m, 2H), 2.28-2.17 (m, 2H), 2.10-1.97 (m, 2H), 1.70-
1.17 (m, 8H), 0.84 (t, J ) 7.3 Hz, 3H); 13C NMR (CDCl3, 75
MHz) 173.9, 172.0, 170.9, 158.4, 153.8, 149.3, 147.8, 142.4,
137.3, 136.4, 133.9, 130.1, 120.6, 120.0, 115.2, 112.2, 112.1,
111.9, 105.3, 76.9, 65.6, 61.2, 56.7, 56.3, 52.9, 49.2, 43.9, 39.3,
38.4, 37.3, 31.7, 27.1, 25.2, 21.3, 14.4; HRMS (FAB) (M + H)+
calc. 708.3384, meas. 708.3382; HPLC retention time ) 12.70
min (normal phase), 11.27 min (reverse phase).
Com p ou n d 2S. IR (neat) 2939, 1741, 1591, 1515, 1454,
1241, 1128, 1027, 731 cm-1 1H NMR (CDCl3, 300 MHz)
;
(mixture of rotamers) 7.19 (t, J ) 7.9 Hz, 1H), 6.67-6.96 (m,
6H) 6.24 (s, 2H), 5.48-5.50 (m, 2H), 4.60-4.74 (m, 2H), 3.86
(s, 3H), 3.85 (s, 3H), 3.78 (s, 3H), 3.61-3.93 (m, 2H), 3.55 (s,
6H), 2.85-2.93 (m, 1H), 2.56-2.68 (m, 2H), 1.98-2.36 (m, 4H),
1.63-1.77 (m, 4H), 1.26-1.46 (m, 2H), 0.90 (t, J ) 7.3 Hz, 3H);
13C NMR (CDCl3, 75 MHz) 174.2, 171.4, 170.6, 158.7, 153.7,
143.2, 135.1, 133.9, 130.0, 120.8, 120.0, 116.4, 112.3, 112.0,
105.4, 77.8, 66.2, 61.3, 56.5, 52.9, 51.7, 39.0, 32.1, 28.7, 27.8,
25.7, 21.4, 12.0; HRMS (FAB) (M + H)+ calc. 694.3227, meas.

1140 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 6
Com p ou n d 5S. IR (neat) 2938, 1737, 1590, 1513, 1455,
Yang et al.
3.94 (q, J ) 6.7 Hz, 1H), 3.84 (s, 6H), 3.76 (d, J ) 12.6 Hz,
1H), 2.75 (t, J ) 11.5 Hz, 1H), 2.62-2.46 (m, 2H), 2.34-2.02
(m, 3H), 1.99-1.27 (m, 5H), 0.88 (m, 3H); 13C NMR (CDCl3,
75 MHz) 174.1, 172.2, 170.7, 158.2, 149.4, 147.8, 142.4, 141.5,
134.0, 130.0, 129.3, 127.8, 127.4, 127.2, 120.6, 120.2, 115.4,
112.3, 111.9, 76.4, 65.7, 56.4, 56.3, 52.8, 44.1, 43.8, 38.4, 32.0,
27.2, 25.6, 21.3, 21.2; HRMS (FAB) (M + H)+ calc. 590.2754,
meas. 590.2745; HPLC retention time ) 11.03 min (normal
phase), 7.38 min (reverse phase).
1
1240, 1128, 1025 cm-1; H NMR (CDCl3, 300 MHz) (mixture
of rotamers) 7.16 (t, J ) 8.0 Hz, 1H), 6.94-6.45 (m, 6H), 6.33
(s, 2H), 5.51 (dd, J ) 5.0, 8.5 Hz, 1H), 5.46 (d, J ) 4.1 Hz,
1H), 4.63 (AB, J ) 16.3 Hz, 2H), 3.85-3.73 (m, 13H), 3.60 (s,
3H), 3.27 (d, J ) 9.2 Hz, 1H), 2.86 (t, J ) 13.0 Hz, 1H), 2.61-
2.26 (m, 4H), 2.16-1.92 (m, 2H), 1.74-1.26 (m, 5H), 1.00 (d,
J ) 6.2 Hz, 3H), 0.70 (d, J ) 6.7 Hz, 3H); 13C NMR (CDCl3, 75
MHz) 173.4, 171.0, 170.0, 158.1, 153.0, 149.0, 147.5, 142.6,
136.9, 133.4, 133.2, 129.6, 120.3, 119.6, 115.5, 111.8, 111.5,
109.5, 105.9, 76.4, 65.7, 60.9, 56.6, 56.1, 52.4, 43.6, 38.3, 31.9,
31.5, 27.3, 25.4, 22.3, 20.9, 20.4; HRMS (FAB) (M + H)+ calc.
708.3384, meas. 708.3361; HPLC retention time ) 7.69 min
(normal phase), 8.27 min (reverse phase).
Com p ou n d 8R . IR (neat) 2937, 1737, 1604, 1516, 1452,
1238, 1158, 10278, 701 cm-1 1H NMR (CDCl3, 300 MHz)
;
(mixture of rotamers) 9.73 (brs, 1H), 7.33-7.08 (m, 6H), 6.91-
6.68 (m, 6H), 5.69 (dd, J ) 5.1, 8.3 Hz, 1H), 5.54 (d, J ) 4.1
Hz, 1H), 4.64 (d, J ) 4.4 Hz, 2H), 3.89 (q, J ) 6.6 Hz, 1H),
3.85 (s, 3H), 3.83 (s, 3H), 3.75 (d, J ) 12.3 Hz, 1H), 3.23 (dt,
J ) 2.3, 12.9 Hz, 1H), 2.71-2.51 (m, 2H), 2.30-2.17 (m, 2H),
2.12-2.03 (m, 1H), 1.71-1.27 (m, 8H); 13C NMR (CDCl3, 75
MHz) 174.6, 172.1, 170.9, 158.5, 149.4, 147.9, 142.5, 142.3,
133.9, 130.1, 129.4, 127.6, 127.2, 120.6, 120.0, 115.8, 112.3,
111.9, 111.1, 76.8, 65.8, 56.4, 56.3, 53.0, 44.0, 43.7, 38.5, 31.8,
27.2, 24.9, 21.3, 20.9; HRMS (FAB) (M + H)+ calc. 590.2754,
meas. 590.2776; HPLC retention time ) 21.02 min (normal
phase), 13.58 min (reverse phase).
Com p ou n d 5R . IR (neat) 2939, 1736, 1590, 1514, 1458,
1239, 1129, 1021 cm-1 1H NMR (CDCl3, 300 MHz) (single
;
enantiomer, mixture of rotamers) 7.24 (t, J ) 7.9 Hz, 1H),
6.91-6.67 (m, 6H), 6.50 (s, 2H), 5.71 (dd, J ) 5.0, 8.0 Hz, 1H),
5.52 (d, J ) 4.3 Hz, 1H), 4.64 (s, 2H), 4.69-4.57 (br, 1H), 4.00
(d, J ) 12.0 Hz, 1H), 3.85-3.60 (m, 6H), 3.32-3.28 (m, 2H),
2.65-2.56 (m, 2H), 2.38-2.05 (m, 4H), 1.65-1.27 (m, 4H), 0.99
(m, 1H), 0.94 (d, J ) 6.3 Hz, 3H), 0.68 (d, J ) 6.6 Hz, 3H); 13
C
NMR (CDCl3, 75 MHz) 173.7, 171.8, 170.5, 158.1, 153.3, 149.0,
147.5, 142.1, 137.1, 134.7, 133.5, 129.7, 120.2, 119.6, 115.4,
111.8, 111.5, 110.7, 105.3, 76.3, 65.7, 60.9, 56.0, 55.9, 52.6, 43.7,
38.1, 32.2, 31.5, 26.7, 24.9, 22.1, 20.9, 20.4; HRMS (FAB) (M
+ H)+ calc. 708.3384, meas. 708.3402; HPLC retention time
) 11.21 min (normal phase), 8.32 min (reverse phase).
Com p ou n d 9S. IR (neat) 2938, 1735, 1596, 1516, 1452,
1237, 1158, 1026, 701 cm-1 1H NMR (CDCl3, 300 MHz)
;
(mixture of rotamers) 9.66 (brs, 1H), 7.28-7.13 (m, 7H), 7.00-
6.62 (m, 5H), 5.69 (t, J ) 6.6 Hz, 1H), 5.54 (brs, 1H), 4.59 (s,
2H), 3.87 (s, 6H), 3.70 (t, J ) 7.0 Hz, 1H), 2.80-2.09 (m, 7H),
1.98-1.17 (m, 7H), 0.90 (m, 3H); 13C NMR (CDCl3, 75 MHz)
173.4, 172.3, 170.7, 158.1, 149.4, 147.8, 142.4, 139.7, 134.0,
130.0, 129.4, 129.1, 128.5, 127.3, 120.7, 120.3, 115.1, 112.4,
111.9, 76.4, 65.6, 56.4, 56.3, 52.7, 51.1, 44.1, 38.4, 31.7, 28.5,
27.2, 25.7, 21.3, 12.7; HRMS (FAB) (M + Na) calc. 626.2730,
meas. 626.2733; HPLC retention time ) 9.12 min (normal
phase), 9.64 min (reverse phase).
Com p ou n d 6S. IR (neat) 2938, 1740, 1590, 1514, 1453,
1
1239, 1126, 1026 cm-1; H NMR (CDCl3, 300 MHz) (mixture
of rotamers) 10.48 (brs, 1H), 7.18 (t, J ) 7.9 Hz, 1H), 6.84-
6.65 (m, 6H), 6.35 (s, 2H), 5.56 (dd, J ) 5.0, 8.4 Hz, 1H), 5.49
(d, J ) 4.5 Hz, 1H), 4.66 (AB, J ) 18.7 Hz, 2H), 3.85 (s, 3H),
3.84 (s, 3H), 3.83-3.80 (m, 2H), 3.78 (s, 3H), 3.61 (s, 6H), 2.88
(dt, J ) 2.5, 13.4 Hz, 1H), 2.64-2.49 (m, 2H), 2.33 (d, J )
13.4 Hz, 1H), 2.18-1.93 (m, 3H), 1.76-1.42 (m, 6H), 0.70-
0.66 (m, 1H), 0.42-0.39 (m, 2H), 0.07-0.02 (m, 2H); 13C NMR
(CDCl3, 75 MHz) 173.9, 172.4, 170.6, 158.4, 153.5, 149.4, 147.8,
142.8, 137.1, 135.2, 133.8, 129.9, 120.6, 119.9, 115.4, 112.2,
111.8, 110.9, 105.4, 76.7, 65.7, 61.1, 56.3, 52.7, 49.9, 43.9, 40.5,
38.6, 31.8, 27.4, 25.6, 21.2, 9.7, 5.2, 4.9; HRMS (FAB) (M +
H)+ calc. 720.3384, meas. 720.3417; HPLC retention time )
10.66 min (normal phase), 9.02 min (reverse phase).
Com p ou n d 9R . IR (neat) 2937, 1737, 1596, 1516, 1451,
1237, 1158, 1026, 701 cm-1 1H NMR (CDCl3, 300 MHz)
;
(mixture of rotamers) 10.15 (s, 1H), 7.32-7.21 (m, 6H), 7.09-
6.57 (m, 6H), 5.73 (dd, J ) 5.4, 7.9 Hz, 1H), 5.56 (s, 1H), 4.65
(s, 2H), 3.85 (s, 3H), 3.83 (s, 3H), 3.64 (t, J ) 7.1 Hz, 1H), 3.20
(t, J ) 12.7 Hz, 1H), 2.69-2.51 (m, 2H), 2.27-2.03 (m, 4H),
1.86-1.27 (m, 7H), 0.95-0.81 (m, 3H); 13C NMR (CDCl3, 75
MHz) 176.9, 174.0, 172.3, 158.4, 149.4, 147.8, 142.4, 140.5,
133.9, 130.1, 129.1, 128.2, 127.3, 120.6, 120.1, 115.3, 112.0,
111.9, 76.7, 65.7, 56.3, 56.2, 52.9, 51.2, 43.8, 38.5, 31.9, 28.2,
27.0, 25.6, 21.3, 12.6; HRMS (FAB) (M + Na) calc. 626.2730,
meas. 626.2700; HPLC retention time ) 15.67 min (normal
phase), 10.78 min (reverse phase).
Com p ou n d 6R . IR (neat) 2937, 1738, 1590, 1514, 1453,
1
1239, 1127, 1015 cm-1; H NMR (CDCl3, 300 MHz) (mixture
of rotamers) 9.58 (brs, 1H), 7.24 (t, J ) 7.8 Hz, 1H), 6.92-
6.67 (m, 6H), 6.49 (s, 2H), 5.71 (dd, J ) 5.3, 8.0 Hz, 1H), 5.55
(d, J ) 3.8 Hz, 1H), 4.62 (s, 2H), 3.94-3.60 (m, 17H), 3.26 (t,
J ) 11.0 Hz, 1H), 2.67-2.55 (m, 2H), 2.31-2.16 (m, 2H), 2.10-
1.86 (m, 2H), 1.63-1.25 (m, 5H), 0.91-0.82 (m, 1H), 0.59 (m,
1H), 0.35-0.32 (m, 2H), 0.01-0.00 (m, 2H); 13C NMR (CDCl3,
75 MHz) 173.7, 172.3, 170.5, 158.0, 153.6, 148.9, 147.4, 142.1,
137.0, 135.9, 133.4, 129.7, 120.2, 119.7, 115.0, 111.8, 111.5,
105.0, 76.4, 65.5, 60.9, 56.3, 55.9, 52.7, 49.6, 43.6, 39.7, 38.1,
31.4, 29.7, 26.8, 24.7, 9.2, 4.8, 4.4; HRMS (FAB) (M + H)+ calc.
720.3384, meas. 720.3401; HPLC retention time ) 13.25 min
(normal phase), 8.75 min (reverse phase).
Com p ou n d 10S. IR (neat) 2956, 1733, 1594, 1516, 1453,
1236, 1158, 1026 cm-1; H NMR (CDCl3, 300 MHz) (mixture
1
of rotamers) 7.29-6.55 (m, 12H), 5.60 (dd, J ) 5.3, 8.2 Hz,
1H), 5.47 (d, J ) 4.6 Hz, 1H), 4.60 (d, J ) 2.9 Hz, 2H), 3.95 (d,
J ) 13.8 Hz, 1H), 3.84 (s, 6H), 3.37 (d, J ) 9.6 Hz, 1H), 2.80
(t, J ) 12.0 Hz, 1H), 2.58-2.24 (m, 4H), 2.09-1.85 (m, 2H),
1.69-1.37 (m, 4H), 1.02 (d, J ) 6.4 Hz, 3H), 0.67 (d, J ) 6.7
Hz, 3H); 13C NMR (CDCl3, 75 MHz) 173.0, 171.5, 170.2, 157.7,
148.9, 147.4, 142.2, 138.1, 133.6, 129.6, 128.7, 128.5, 128.1,
127.0, 120.3, 119.8, 115.0, 111.9, 111.4, 75.9, 65.3, 56.4, 56.0,
55.9, 52.3, 43.7, 37.9, 32.1, 31.3, 27.0, 25.4, 22.2, 21.0, 20.3;
HRMS (FAB) (M + H)+ calc. 618.3067, meas. 618.3088; HPLC
retention time ) 8.22 min (normal phase), 12.59 min (reverse
phase).
Com p ou n d 7. IR (neat) 2938, 1736, 1593, 1516, 1452, 1237,
1156, 1027, 732, 698 cm-1 1H NMR (CDCl3, 300 MHz)
;
(mixture of rotamers) 7.26-7.20 (m, 6H), 6.91-6.77 (m, 4H),
6.70-6.64 (m, 3H), 5.68 (dd, J ) 5.1, 6.0 Hz, 1H), 5.48 (s, 1H),
4.59 (AB, J ) 16.4 Hz, 2H), 3.85 (s, 6H), 3.77-3.73 (m, 3H),
3.24 (t, J ) 11.7 Hz, 1H), 2.65-2.50 (m, 2H), 2.32 (d, J ) 12.4
Hz, 1H), 2.24-2.03 (m, 3H), 1.69-1.56 (m, 3H), 1.33-1.25 (m,
2H); 13C NMR (CDCl3, 75 MHz) 172.1, 171.9, 170.7, 158.5,
149.4, 147.9, 142.4, 134.9, 133.9, 130.1, 129.1, 129.0, 127.2,
120.6, 120.0, 115.7, 112.3, 111.9, 111.4, 76.8, 65.8, 56.4, 56.3,
52.8, 44.5, 41.4, 38.4, 31.8, 27.2, 25.4, 21.1; HRMS (FAB) (M
+ H)+ calc. 576.2597, meas. 576.2623; HPLC retention time
) 18.67 min (normal phase), 7.80 min (reverse phase).
Com p ou n d 10R. IR (neat) 2937, 1737, 1594, 1515, 1452,
1236, 1158, 1026 cm-1; H NMR (CDCl3, 300 MHz) (mixture
1
of rotamers) 8.13 (brs, 1H), 7.27-7.22 (m, 6H), 6.92-6.67 (m,
6H), 5.72 (dd, J ) 5.2, 8.0 Hz, 1H), 5.52 (d, J ) 4.3 Hz, 1H),
4.59 (d, J ) 10.9 Hz, 2H), 3.97 (d, J ) 10.9 Hz, 1H), 3.86 (s,
3H), 3.84 (s, 3H), 3.39 (d, J ) 9.7 Hz, 1H), 3.26 (t, J ) 11.8
Hz, 1H), 2.70-2.52 (m, 2H), 2.48-2.35 (m, 1H), 2.26-2.16 (m,
2H), 2.13-2.03 (m, 1H), 1.59-1.42 (m, 4H), 0.96 (d, J ) 6.4
Hz, 3H), 0.65 (d, J ) 6.7 Hz, 3H); 13C NMR (CDCl3, 75 MHz)
173.8, 171.6, 170.4, 158.0, 149.0, 147.4, 142.1, 139.0, 133.5,
129.7, 128.6, 128.5, 127.0, 120.2, 119.7, 115.4, 111.8, 111.4,
110.6, 76.3, 65.4, 56.3, 55.9, 55.8, 52.5, 43.6, 38.1, 32.0, 31.4,
Com p ou n d 8S. IR (neat) 2936, 1736, 1598, 1516, 1450,
1
1238, 1158, 1027 cm-1; H NMR (CDCl3, 300 MHz) (mixture
of rotamers) 10.17 (brs, 1H), 7.32-7.14 (m, 6H), 6.96-6.67 (m,
6H), 5.69 (dd, J ) 5.2, 7.6 Hz, 1H), 5.52 (s, 1H), 4.58 (s, 2H),

FKBP12 Ligand Binding Site with Specificity Pocket
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 6 1141
26.7, 24.8, 22.2, 20.9, 20.4; HRMS (FAB) (M + H)+ calc.
618.3067, meas. 618.3045; HPLC retention time ) 12.00 min
(normal phase), 12.44 min (reverse phase).
4-6 h at 25 °C. For purification, frozen cell paste was rapidly
thawed into 5 volumes of chilled lysis buffer (20 mM Tris/Tris-
HCl pH 7.4, containing 5 mM DTT, 5 mM EDTA, and 1 mM
of the protease inhibitor AEBSF) (Calbiochem, San Diego, CA).
The lysate was rapidly refrozen in an ethanol/dry ice bath and
rapidly rethawed on a chilled water bath. The suspension was
then centrifuged at 25000g for 15 min at 4 °C. After the
solution was filtered through a 0.45 mm membrane, the
clarified supernatant was loaded onto a polyethyleneimine
column (3.3 mL of resin/g of cell paste) (Bakerbond PEI, VWR
Scientific Products, Edison, NJ ) and then a Macro-Prep High
Q column (0.6 mL of resin/g of cell paste) (BioRad Laboratories,
Hercules, CA), both columns having been equilibrated in lysis
buffer. The flow-through was collected, concentrated on an
Amicon YM3 membrane to 10-20 mg/mL, filtered through a
0.2 mm filter, and further purified by gel filtration on a 2.6 ×
90 cm Sephacryl S100HR column (Pharmacia Biotech, Piscat-
away, NJ ) equilibrated with 50 mM potassium phosphate
buffer pH 6.5, containing 100 mM NaCl, 1 mM EDTA, 0.02%
NaN3, and 5 mM DTT, at a flow-rate of 1 mL/min. F36V-FKBP
eluted at 345 mL, and the yield was approximately 8.5 mg/L
culture.
NMR Meth od s a n d Str u ctu r e Deter m in a tion . The
sequential NMR assignments of the protein were carried out
using standard multidimensional heteronuclear NMR methods
on a 1:1 double-labeled F36V-FKBP/ligand complex.25 The two-
dimensional 1H-15N HSQC26 spectrum of uncomplexed double-
labeled F36V-FKBP showed many broad peaks which sharp-
ened upon adding ligand, suggesting that the uncomplexed
protein may exist in some form of oligomerization equilibrium.
However, F36V-FKBP migrated identically to WT FKBP in
HPLC gel filtration on a TSK G2000SW column (7.5 × 300
mm, 5 mm particle size) (Tosohaas Inc., Montgomeryville, PA).
By this criterion, F36V-FKBP was 97% monomeric following
NMR analysis at 1-2 mM concentration at 30 °C, then dilution
to 0.2 mM for injection on the TSK column. Therefore, any
oligomerization phenomenon that might be responsible for line
broadening of the uncomplexed protein in NMR appears to be
of low affinity and rapidly reversible. Nevertheless, the effect
was such that the sequential NMR assignment of the uncom-
plexed protein could not be completed for all of the residues.
Ligand-protein and ligand-ligand NOEs were obtained
using isotope-filtered NMR methods.27 The structures were
calculated using a slow cooling simulated annealing docking
procedure within X-PLOR in which protein-ligand and ligand-
ligand NOEs (typically 60-70 and 20-25, respectively, for
each complex) were used.28 All the heavy atoms of the protein,
with the exception of several side chains which showed NOEs
to ligand, were held rigid. The starting coordinates were from
the X-ray crystal structure of the F36V-FKBP/2S complex.9
All the waters were removed, as the binding cavity does not
contain any water molecules. A family of 15 structures, usually
from a pool of 100-200 structures, which had minimum energy
and showed no violations of the experimental NMR data and
covalent geometries was selected. The rmsd was calculated for
the heavy atoms of the ligand. The structure with a minimum
violation energy among a family of structures was selected for
comparison and overlay.
Com p ou n d 11S. IR (neat) 2999, 1734, 1594, 1516, 1450,
1158, 1028, 700 cm-1; 1H NMR (CDCl3, 300 MHz) (mixture of
rotamers) 7.26-7.12 (m, 6H), 6.98-6.56 (m, 6H), 5.58 (dd, J
) 4.8, 8.0 Hz, 1H), 5.47 (s, 1H), 4.61 (AB, J ) 15.4 Hz, 2H),
3.96 (d, J ) 13.6 Hz, 1H), 3.85 (s, 6H), 3.46 (d, J ) 9.7 Hz,
1H), 2.82 (t, J ) 12.2 Hz, 1H), 2.59-2.41 (m, 2H), 2.32-1.92
(m, 5H), 1.89-1.42 (m, 6H), 1.26-0.67 (m, 8H); 13C NMR
(CDCl3, 75 MHz) 173.0, 170.2, 157.7, 148.9, 147.4, 142.2, 137.6,
133.6, 129.6, 128.8, 128.5, 126.9, 120.3, 119.8, 115.2, 111.9,
111.4, 111.1, 75.9, 56.0, 55.1, 52.4, 43.7, 41.2, 37.9, 32.8, 31.6,
30.7, 27.0, 26.5, 26.2, 25.5, 21.0; HRMS (FAB) (M + H)+ calc.
658.3380, meas. 658.3398; HPLC retention time ) 8.19 min
(normal phase), 17.24 min (reverse phase).
Com p ou n d 11R. IR (neat) 2931, 1737, 1595, 1516, 1449,
1236, 1158, 1028, 701 cm-1 1H NMR (CDCl3, 300 MHz)
;
(mixture of rotamers) 7.27-7.16 (m, 6H), 6.90-6.48 (m, 6H),
5.68 (dd, J ) 4.6, 7.9 Hz, 1H), 5.51 (s, 1H), 4.69 (AB, J ) 16.8
Hz, 2H), 3.96 (d, J ) 12.8 Hz, 1H), 3.87 (s, 3H), 3.86 (s, 3H),
3.46 (d, J ) 9.7 Hz, 1H), 3.30 (t, J ) 12.3 Hz, 1H), 2.68-2.53
(m, 2H), 2.27-2.08 (m, 4H), 1.82-1.44 (m, 7H), 1.26-1.06 (m,
4H), 0.98-0.68 (m, 4H); 13C NMR (CDCl3, 75 MHz) 173.9,
170.4, 158.1, 149.0, 147.5, 142.2, 138.4, 133.5, 129.7, 128.6,
128.5, 126.9, 120.2, 119.6, 115.9, 111.8, 111.4, 109.7, 76.4, 55.9,
55.1, 52.6, 43.6, 41.1, 38.1, 32.7, 31.5, 30.7, 26.8, 26.5, 26.1,
24.9, 20.9; HRMS (FAB) (M + H)+ calc. 658.3380, meas.
658.3349; HPLC retention time ) 10.47 min (normal phase),
18.25 min (reverse phase).
Com p ou n d 12. IR (neat) 2937, 1735, 1599, 1515, 1453,
1236, 1157, 1025, 700 cm-1 1H NMR (CDCl3, 300 MHz)
;
(mixture of rotamers) 7.33-7.07 (m, 11H), 6.95-6.67 (m, 6H),
6.17 (brs, 1H), 5.71 (dd, J ) 5.1, 8.0 Hz, 1H), 5.53 (d, J ) 3.4
Hz, 1H), 5.24 (s, 1H), 4.55 (AB, J ) 16.6 Hz, 2H), 3.85 (s, 7H),
3.18 (t, J ) 12.0 Hz, 1H), 2.66-2.56 (m, 2H), 2.35-2.08 (m,
3H), 1.70-1.53 (m, 3H), 1.49-1.26 (m, 2H); 13C NMR (CDCl3,
75 MHz) 172.4, 171.3, 170.4, 158.0, 149.0, 147.5, 142.1, 139.1,
138.8, 133.5, 129.1, 128.6, 128.5, 127.0, 120.2, 119.6, 115.5,
111.8, 111.4, 76.3, 65.2, 56.0, 55.9, 52.7, 44.0, 38.2, 31.5, 27.0,
25.0, 20.8; HRMS (FAB) (M + H)+ calc. 652.2910, meas.
652.2934; HPLC retention time ) 14.30 min (normal phase),
13.73 min (reverse phase).
F KBP Bin d in g Assa y. The affinities of synthetic ligands
for FKBP12 were determined using a competitive assay based
on fluorescence polarization (FP). Full details of the assay will
be provided elsewhere (C. T. Rollins, E. Laborde, D. Holt, and
T. Clackson, in preparation). Briefly, a fluorescent probe was
prepared by coupling 4′-(aminomethyl)-fluorescein (Molecular
Probes) to the modified allyl group of FK506 in a manner
similar to the preparation of FK1012.14 Recombinant human
FKBP12 was expressed and purified by standard methods.24
In the wells of a Dynatech microfluor plate, subsaturating
concentrations of FKBP12 or F36V-FKBP were incubated with
2.5 nM probe and serial dilutions of competitive ligand in FP
buffer (50 mM potassium phosphate pH 7.8/150 mM NaCl/
100 µg/mL bovine gamma globulin/1% EtOH). After equilibra-
tion for 30 min in the dark, fluorescence polarization was read
on a J olley FPM-2 (J olley Consulting and Research, Inc.,
Grayslake, IL). The increase in polarization of the probe upon
binding protein was used as a direct readout of percent probe
bound, compared to controls containing no competitor (100%)
and no protein (0%), and the concentration of competitor
resulting in 50% binding (IC50) was determined by a nonlinear
least-squares fit to a four-parameter equation.
Ack n ow led gm en t. We thank Don Cullis and Bon-
nie Marmor for technical assistance in preparative
chemistry and mass spectrometry.
Su p p or tin g In for m a tion Ava ila ble: Tables of atomic
coordinates for F36V-FKBP complexes with 2R, 2S, 9R, 9S,
10R, and 10S (PDB). This material is available free of charge
via the Internet at http://pubs.acs.org.
P r ep a r a tion of Un ifor m ly [13C]- a n d
[
15N]-La beled
F KBP for NMR. Human F36V-FKBP with no hexahistidine-
tag was produced in Escherichia coli BL21(DE3) from vector
pET20b (Novagen Inc., Madison, WI).9 Cells were grown at
37 °C in M9 medium containing 200 mg/L ampicillin, with
U-[13C]-D-glucose (2.9 g/L) and 15NH4Cl (1 g/L) (Cambridge
Isotope Laboratories, Andover, MA), as the only carbon and
nitrogen source, and they were induced at an A600 of ap-
proximately 0.8 with 1 mM isopropyl-â-D-thiogalactoside for
Refer en ces
(1) Babine, R. E.; Bender, S. L. Molecular Recognition of Protein-
Ligand Complexes: Applications to Drug Design. Chem. Rev.
1997, 97, 1359-1472.
(2) Shah, K.; Liu, Y.; Deirmengian, C.; Shokat, K. M. Engineering
Unnatural Nucleotide Specificity for Rous Sarcoma Virus Ty-
rosine Kinase to Uniquely Label Its Direct Substrates. Proc.
Natl. Acad. Sci. U.S.A. 1997, 94, 3565-3570.

1142 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 6
Yang et al.
(3) Bishop, A. C.; Kung, C.-y.; Shah, K.; Witucki, L.; Shokat, K. M.;
Liu, Y. Generation of Monospecific Nanomolar Tyrosine Kinase
Inhibitors via a Chemical Genetic Approach. J . Am. Chem. Soc.
1999, 121, 627-631.
(4) Liu, Y.; Shah, K.; Yang, F.; Witucki, L.; Shokat, K. M. A
Molecular Gate Which Controls Unnatural ATP Analogue
Recognition by the Tyrosine Kinase v-Src. Bioorg. Med. Chem.
1998, 6, 1219-1226.
(5) Liu, Y.; Shah, K.; Yang, F.; Witucki, L.; Shokat, K. M. Engineer-
ing Src Family Protein Kinases with Unnatural Nucleotide
Specificity. Chem. Biol. 1998, 5, 91-101.
(6) Bishop, A. C.; Shah, K.; Liu, Y.; Witucki, L.; Kung, C.-y.; Shokat,
K. M. Design of Allele-Specific Inhibitors to Probe Protein Kinase
Signaling. Curr. Biol. 1998, 8, 257-266.
(7) Belshaw, P. J .; Schoepfer, J . G.; Liu, K.-Q.; Morrison, K. L.;
Schreiber, S. L. Rational Design of Orthogonal Receptor-Ligand
Combinations. Angew. Chem., Int. Ed. Engl. 1995, 34, 2129-
2132.
(8) Belshaw, P. J .; Schreiber, S. L. Cell-Specific Calcineurin Inhibi-
tion by a Modified Cyclosporin. J . Am. Chem. Soc. 1997, 119,
1805-1806.
FKBP12 Ligands for the Regulated Dimerization of Proteins.
Bioorg. Med. Chem. 1998, 6, 1309-1335.
(17) Holt, D. A.; Luengo, J . I.; Yamashita, D. S.; Oh, H.-J .; Konialian,
A. L.; Yen, H.-K.; Rozamus, L. W.; Brandt, M.; Bossard, M. L.;
Levy, M. A.; Eggleston, D. S.; Liang, J .; Schultz, L. W.; Stout,
T. J .; Clardy, J . Design, Synthesis, and Kinetic Evaluation of
High-Affinity FKBP Ligands and the X-ray Crystal Structures
of Their Complexes with FKBP12. J . Am. Chem. Soc. 1993, 115,
9925-9938.
(18) Holt, D. A.; Konialian-Beck, A. L.; Oh, H.-J .; Yen, H.-K.;
Rozamus, L. W.; Krog, A. J .; Erhard, K. F.; Ortiz, E.; Levy, M.
A.; Brandt, M.; Bossard, M. J .; Luengo, J . I. Structure-Activity
Studies of Synthetic FKBP Ligands as Peptidyl-Prolyl Isomerase
Inhibitors. Bioorg. Med. Chem. Lett. 1994, 4, 315-320.
(19) Hatada, M., unpublished results.
(20) Boehm, H.-J .; Klebe, G. What Can We Learn From Molecular
Recognition in Protein-Ligand Complexes for the Design of New
Drugs? Angew. Chem., Int. Ed. Engl. 1996, 35, 2588-2614.
(21) Pierce, A. C.; J orgensen, W. L. Computational Binding Studies
of Orthogonal Cyclosporin-Cyclophilin Pairs. Angew. Chem., Int.
Ed. Engl. 1997, 36, 1466-1469.
(22) Redfern, C. H.; Coward, P.; Degtyarev, M. Y.; Lee, E. K.; Kwa,
A. T.; Hennighausen, L.; Bujard, H.; Fishman, G. I.; Conklin,
B. R. Conditional Expression and Signaling of a Specifically
Designed Gi-Coupled Receptor in Transgenic Mice. Nat. Bio-
technol. 1999, 17, 165-169.
(23) Smith, G. K.; Banks, S.; Blumenkopf, T. A.; Cory, M.; Hum-
phreys, J .; Laethem, R. M.; Miller, J .; Moxham, C. P.; Mullin,
R.; Ray, P. H.; Walton, L. M.; Wolfe, L. A., III Toward Antibody-
Directed Enzyme Prodrug Therapy with the T268G Mutant of
Human Carboxypeptidase A1 and Novel In Vivo Stable Prodrugs
of Methotrexate. J . Biol. Chem. 1997, 272, 15804-15816.
(24) Wiederrecht, G.; Hung, S.; Chan, H. K.; Marcy, A.; Martin, M.;
Calaycay, J .; Boulton, D.; Sigal, N.; Kincaid, R. L.; Siekierka, J .
J . Characterization of High Molecular Weight FK-506 Binding
Activities Reveals a Novel FK-506-Binding Protein as Well as a
Protein Complex. J . Biol. Chem. 1992, 267, 21753-21760.
(25) Muhandiram, D. R.; Kay, L. E. Gradient-Enhanced Triple-
Resonance Three-Dimensional NMR Experiments with Im-
proved Sensitivity. J . Magn. Reson., Ser. B 1994, 103, 203-216.
(26) Kay, L. E.; Keifer, P.; Saarinen, T. Pure Absorption Gradient
Enhanced Heteronuclear Single Quantum Correlation Spectros-
copy with Improved Sensitivity. J . Am. Chem. Soc. 1992, 114,
10663-10665.
(9) Clackson, T.; Yang, W.; Rozamus, L. W.; Hatada, M.; Amara, J .
F.; Rollins, C. T.; Stevenson, L. F.; Magari, S. R.; Wood, S. A.;
Courage, N. L.; Lu, X.; Cerasoli, F., J r.; Gilman, M.; Holt, D. A.
Redesigning an FKBP-Ligand Interface to Generate Chemical
Dimerizers with Novel Specificity. Proc. Natl. Acad. Sci. U.S.A.
1998, 95, 10437-10442.
(10) Clackson, T. Redesigning Small Molecule-Protein Interfaces.
Curr. Opin. Struct. Biol. 1998, 8, 451-458.
(11) Kurumbail, R. G.; Stevens, A. M.; Gierse, J . K.; McDonald, J .
J .; Stegeman, R. A.; Pak, J . Y.; Gildehaus, D.; Miyashiro, J . M.;
Penning, T. D.; Seibert, K.; Isakson, P. C.; Stallings, W. C.
Structural Basis for Selective Inhibition of Cyclooxygenase-2 by
Antiinflammatory Agents. Nature 1996, 384, 644-648.
(12) Gierse, J . K.; McDonald, J . J .; Hauser, S. D.; Rangwala, S. H.;
Koboldt, C. M.; Seibert, K. A Single Amino Acid Difference
Between Cyclooxygenase-1 (COX-1) and -2 (COX-2) Reverses the
Selectivity of COX-2 Specific Inhibitors. J . Biol. Chem. 1996,
271, 15810-15814.
(13) Clemons, P. A. Design and Discovery of Protein Dimerizers.
Curr. Opin. Chem. Biol. 1999, 3, 112-115.
(14) Spencer, D. M.; Wandless, T. J .; Schreiber, S. L.; Crabtree, G.
R. Controlling Signal Transduction with Synthetic Ligands.
Science 1993, 262, 1019-1024.
(15) Amara, J . F.; Clackson, T.; Rivera, V. M.; Guo, T.; Keenan, T.;
Natesan, S.; Pollock, R.; Yang, W.; Courage, N. L.; Holt, D. A.;
Gilman, M. A Versatile Synthetic Dimerizer for the Regulation
of Protein-Protein Interactions. Proc. Natl. Acad. Sci. U.S.A.
1997, 94, 10618-10623.
(16) Keenan, T.; Yaeger, D. R.; Courage, N. L.; Rollins, C. T.; Pavone,
M. E.; Rivera, V. M.; Yang, W.; Guo, T.; Amara, J . F.; Clackson,
T.; Gilman, M.; Holt, D. A. Synthesis and Activity of Bivalent
(27) Ikura, M.; Bax, A. Isotope-Filtered 2D NMR of a Protein-Peptide
Complex: Study of a Skeletal Muscle Myosin Light Chain Kinase
Fragment Bound to Calmodulin. J . Am. Chem. Soc. 1992, 114,
2433-2440.
(28) Brunger, A. X-PLOR, Version 3.1. A System for X-ray Crystal-
lography and NMR; Yale University Press: New Haven, CT,
1992.
J M9904396