Integrating fragment assembly and biophysical methods in the chemical advancement of small-molecule antagonists of IL-2: An approach for inhibiting protein-protein interactions

J . Med. Chem. 2004, 47, 3111-3130

3111

In tegr a tin g F r a gm en t Assem bly a n d Biop h ysica l Meth od s in th e Ch em ica l

Ad va n cem en t of Sm a ll-Molecu le An ta gon ists of IL-2: An Ap p r oa ch for

In h ibitin g P r otein -P r otein In ter a ction s^†

Brian C. Raimundo,* J ohan D. Oslob, Andrew C. Braisted, J ennifer Hyde, Robert S. McDowell, Mike Randal,

Nathan D. Waal, J ennifer Wilkinson, Chul H. Yu, and Michelle R. Arkin*

Sunesis Pharmaceuticals Inc., 341 Oyster Point Boulevard, South San Francisco, California 94080

Received J anuary 9, 2004

Fragment assembly has shown promise for discovering small-molecule antagonists for difficult

targets, including protein-protein interactions. Here, we describe a process for identifying a

60 nM inhibitor of the interleukin-2 (IL-2)/IL-2 receptor (IL-2RR) interaction. By use of

fragment-based approaches, a compound with millimolar affinity was evolved to a hit series

with low micromolar activity, and these compounds were optimized into a lead series with

nanomolar affinity. Fragment assembly was useful not only for hit identification, but also for

lead optimization. Throughout the discovery process, biophysical methods and structural biology

demonstrated that compounds bound reversibly to IL-2 at the IL-2 receptor binding site.

In tr od u ction

screening of a very large virtual space while synthesiz-

ing a relatively small number of molecules and may

therefore have an advantage over traditional screening

approaches when hits are rare. Furthermore, since

protein-protein interfaces tend to cover a large area of

the protein surface, small molecule inhibitors might

need to make a number of key binding interactions.

Recently, it has been shown that, as the number of

interactions needed for productive binding increases, the

chances of finding a complimentary ligand-protein pair

decreases dramatically; thus, complex binding sites are

well-suited for a fragment-based approach.²⁴

Interleukin-2 (IL-2) is a key mediator of the T-helper

1 (Th1) immune response.^1-5Binding of IL-2 to the

trimeric IL-2 receptor (IL-2R)sconsisting of R (IL-2RR),

â (IL-2Râ), and γ chains (IL-2Rγ)scauses proliferation

and clonal expansion of activated T cells. Abnormal Th1

immune responses have been implicated in graft rejec-

tion and other autoimmune diseases,^6,7and current

therapies target IL-2 production or the IL-2 signaling

pathway.^8,9Clinical data with antibodies directed against

IL-2RR suggest that specific inhibition of the IL-2/IL-

2RR interaction targets the Th1 response without caus-

ing the toxicity observed for more general immunomod-

ulators.^10,11Therapeutic antibodies have several draw-

backs, however, including high cost-of-goods and lack

of oral bioavailability. A small-molecule inhibitor of the

IL-2/IL-2RR interaction could offer a significant im-

provement in immunosuppressive therapy. However,

small molecule inhibitors of such protein-protein in-

teractions have been difficult to identify.^12-16This

manuscript describes our research with IL-2, which we

consider to be both an important therapeutic target for

immune disease and a training ground for developing

approaches toward small-molecule drug discovery at

protein-protein interfaces.

Fragment-assembly approaches have been applied

toward inhibitor discovery and optimization,^17-23and

we have been interested in using these methods for

protein-protein interactions. First, small (<200 Da)

organic molecules are screened for binding or inhibition

against multiple subsites of a target. In a second step,

selected fragments are chemically linked or otherwise

optimized into a compound with sufficient activity to

serve as a starting point for medicinal chemistry.

Fragment assembly might be particularly useful for

developing inhibitors of difficult targets such as protein-

protein interactions. Fragment-based methods allow

An important challenge with fragment assembly is

the identification of the fragments, since they often have

low (millimolar) affinity for the target. Methods to

discover and validate fragments include NMR,^18,19,25

X-ray crystallography,^22,26functional screening,²⁷sur-

face plasmon resonance (SPR),^28,29and Tethering. (The

term Tethering is a service mark of Sunesis Pharma-

ceuticals Inc. for its fragment-based drug discovery.)^23,30

In Tethering, fragments are selected based on disulfide-

exchange between a cysteine-containing protein and

disulfide-modified fragments (Figure 1). Each of these

approaches has unique advantages, and we have used

a number of methods in tandem to evolve inhibitors of

IL-2.

IL-2 is one of the few protein hormones for which a

small-molecule inhibitor has been described. Hoffman

LaRoche has shown by NMR that the low micromolar

inhibitor 1 (Figure 2) binds to IL-2 at the IL-2RR

binding site.^31,32Recently, we published the X-ray

structure of 1 bound to IL-2 and characterized the

binding site by Tethering.³³The small-molecule binding

site seen in Figure 2 contains two subsites. To the right

is a shallow pocket containing the acidic residue Glu

62 that makes two hydrogen bonds with the guanidine

moiety of 1; to the left is a malleable, hydrophobic

groove, created by the residues Arg 38, Leu 72, Lys 35,

Met 39, and Lys 76, that binds to the biaryl portion of

the compound. 1 can be seen as a merger of two

* To whom correspondence should be addressed: raimundo@

sunesis.com; mra@sunesis.com.

^†We dedicate this manuscript to Miles Braisted and J oelle Morrow.

Published on Web 05/07/2004

3112 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12

Raimundo et al.

F igu r e 1. The Tethering method of fragment discovery is used

to identify fragments that bind to a selected region of IL-2. In

the first step, a cysteine mutation is placed near the binding

site for IL-2RR. The protein is then interrogated with a pool

of fragments that contain disulfide bonds, under conditions

that favor disulfide exchange. After equilibrium has been

reached, the compound/protein mixture is analyzed by mass

spectrometry, which identifies the protein-small molecule

conjugates. The highly reducing conditions ensure that only

compounds with intrinsic affinity for the protein surface will

be captured. In a separate step, the fragment can be prepared

without a disulfide handle and tested for noncovalent binding

to IL-2.

fragmentssa biaryl acetylene amino acid (e.g., 2) and

a piperidinyl guanidine acetic acid (e.g., 3)sjoined by

an amide bond. We chose to use this compound to

develop tight-binding inhibitors of IL-2 using a frag-

ment-based approach. The first step of our strategy was

to determine whether fragments derived from 1 could

be identified by various discovery approaches. The

second step was to use these fragments to develop a new

compound series that was chemically progressable and

bound analogously to 1. Finally, Tethering was used to

identify an additional subsite adjacent to the known

binding site of 1. Incorporating fragments that target

this subsite into our new series generated a molecule,

33i, that inhibits IL-2/IL-2RR binding with an IC₅₀of

60 nM and shows activity in a cell-based assay. Here

we describe a highly integrated drug-discovery approach

that combined fragment assembly, medicinal chemistry,

biophysics, and structural biology to identify potent

inhibitors of this important protein-protein interaction.

F igu r e 2. Fragments derived from 1. Top, structure of 1

bound to IL-2 from X-ray crystallography (protein data bank

ID code 1M48).³³1 shown in green sticks, with IL-2 shown in

white ribbon. IL-2 residues comprising the binding site are

shown in white sticks (oxygen atoms shown in red, nitrogen

in blue, sulfur in orange). Bottom, structure of 1 and composite

fragments (2, 3). 2 was synthesized as the acid for solubility

purposes; the acid derivative of 1 has the same activity as 1

itself (unpublished observations). Listed IC₅₀data are average

values from scintillation proximity assay (SPA); the listed

K_d,SPRvalue is from SPR measured at equilibrium for several

concentrations of fragment (see Experimental Section).

step process of amine deprotection, guanidine formation,

and guanidine deprotection was carried out in the same

vial with removal of solvent in vacuo after each trans-

formation. After final deprotection, each member of the

library was purified by RP-HPLC.

Ch em istr y

P h en yl Rin g Su bstitu tion . The most successful

hydrophobic amine from Scheme 2 was a tricycle

containing linked piperidine, pyrazole, and phenyl rings.

Substitution of the phenyl ring was altered by synthe-

sizing pyrazole triflate, 12, and coupling it with substi-

tuted aryl boronic acids (Scheme 3). Starting with Boc-

piperidine 9, C-acylation with methyl cyanoformate

provided 10. Cyclization of 10 using hydrazine yielded

pyrazolone 11. Formation of the triflate and then Boc-

protection of the pyrazole nitrogen provided 12. Aryl

boronic acids were coupled to the pyrazole using Suzuki

reaction conditions and removal of the Boc-groups

yielded piperidine 14. Coupling aryl boronic acids

with the triflate required Boc-protection of the pyrazole

ring. Significant hydrolysis of the triflate was observed

if the pyrazole ring was left unprotected. Transforma-

Bia r yl Acetylen e Dip ep tid es. The first step in new

lead generation was to replace the piperidinyl guanidine

moiety on 1. Starting with the biaryl acetylene 4, a

dipeptide guanidine fragment was introduced by stan-

dard peptide chemistry with Boc-amino acids (Scheme

1). Guanidine formation of 6 using N,N′-bis-Boc-1-

guanylpyrazole reagent, followed by Boc-deprotection

and purification, led to 7a -m .

Libr a r y Id en tifyin g New Hyd r op h obic F r a g-

m en ts. To replace the biaryl moiety, the chemical

sequence from Scheme 1 was inverted, starting with a

preferred dipeptide fragment (AA2 ) ^D-cyclohexylgly-

cine) as an ‘anchor’ and incorporating ∼100 lipophilic

amines (Scheme 2). The activated ester 8 was used in a

solution phase, parallel synthesis approach. A three-

IL-2 Inhibitors Small-Molecule Antagonists of IL-2

J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12 3113

Sch em e 1. Synthesis of Dipeptide Linkers for IL-2 Binding Molecules^a

a

(a) Boc-glycine, EDC, HOBT, Et₃N in CH₂Cl₂; (b) HCl in dioxane; (c) Boc amino acid, EDC, HOBT, Et₃N in CH₂Cl₂; (d) N,N′-bis-Boc-

1-guanylpyrazole, Et₃N, MeOH; (e) TFA in CH₂Cl₂.

Sch em e 2. Synthesis of Hydrophobic Fragment

Other tricyclic fragments contained replacements for

the pyrazole ring (Scheme 4). The phenyl analogue 19

was synthesized from the commercially available pip-

eridine phenol by protection, conversion to the triflate,

and then Suzuki coupling. For oxadiazole 21, hydroxy-

benzamidine 20 was condensed with the acid fluoride

of Boc-isonipecotic acid to yield the desired tricycle.

Completion of each analogue (22-26) followed the

procedure described for the parallel library.

Library^a

a

(a) RR′NH, Et₃N; (b) HCl in dioxane; (c) N,N′-bis-Boc-1-

guanylpyrazole, Et₃N, MeOH; (d) TFA/CH₂Cl₂(1:1).

In cor p or a tion of Teth er in g-Der ived F r a gm en ts

a n d Syn th esis of 33i. As shown in Scheme 5, frag-

ments identified from Tethering (see below) were in-

troduced through an ether linkage at the 4-position of

the phenyl ring (relative to the pyrazole). At this stage,

the original route for the synthesis of N-methylated

pyrazoles was modified such that the 2-methylated

pyrazole was obtained preferentially, as follows: Alkyne

29 was synthesized from an aldehyde derivative of

isonipecotic acid by treatment with Bestmann’s reagent.

Palladium-mediated coupling of 29 and acid chloride 27

provided the alkynone which was treated with methyl

hydrazine to yield the 2-methylated pyrazole as a single

regioisomer. Deprotection of the phenol with TBAF

yielded 32. Incorporation of the fragments identified by

Tethering was accomplished using standard alkylation

or Mitsunobu conditions. Finally, the guanidine was

installed as described above.

tion into the final compounds 15b-k followed the same

synthetic route as described for the general library

(Scheme 2).

P yr a zole R in g Su b st it u t ion /R ep la cem en t . As

shown in Scheme 4, a methyl group was introduced at

the three unsubstituted sites of the pyrazole by cycliza-

tion of diketone intermediates with the appropriate

hydrazine. The synthesis of the N-methylated pyrazoles

began by C-acylation of ketone 9. Treatment of the

diketone with methyl hydrazine provided the methyl

pyrazoles 16a ,b as a mixture of regioisomers that

were separated by chromatography. A series of NMR

experiments were used to characterize each of the regio-

isomers. The C-methylated pyrazole, 18, was synthe-

sized by a similar sequence starting with ethyl ketone

17.

Sch em e 3. Synthesis of Substituted Phenyl Rings^a

a

(a) LHMDS, MeOCOCN, THF; (b) H₂NNH₂, EtOH, 70 °C; (c) PhN(Tf)₂, Et₃N, CH₂Cl₂; (d) Boc₂O, Et₃N, DMAP, CH₂Cl₂; (e) ArB(OR)₂,

PdCl₂(PPh₃)₂, K₂CO₃, DMF/dioxane/H₂O; (f) HCl in dioxanes; (g) 8, Et₃N, CH₂Cl₂; (h) N,N′-bis-Boc-1-guanylpyrazole, Et₃N, MeOH; (i)

TFA/CH₂Cl₂.

3114 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12

Raimundo et al.

Sch em e 5. Synthesis of Ring-Addition Library^a

Sch em e 4. Synthesis of Pyrazole Substitutions/

Replacements^a

a

(a) Br₂, CH₂Cl₂; (b) TIPSCl, Et₃N, THF; (c) n-BuLi, CO₂, THF;

(d) (COCl)₂, DMF, CH₂Cl₂; (e) 28, K₂CO₃, MeOH; (f) HCl in

dioxanes; (g) Boc-glycine, EDC, HOBT, Et₃N, CH₂Cl₂; (h) Boc-D-

leucine, EDC, HOBT, Et₃N, CH₂Cl₂; (i) CuI, Et₃N, Cl₂(Ph₃P)₂Pd,

toluene; (j) MeHNNH₂, EtOH; (k) TBAF, THF; (l) i. RX, Cs₂CO₃,

DMSO, 65 °C or ROH, DEAD, PhP₃, THF; ii. LiOH, THF, H₂O,

60 °C; (m) N,N′-bis-Boc-1-guanylpyrazole, Et₃N, MeOH; (n) TFA

in CH₂Cl₂.

HSQC NMR; in the case of 2, precipitation was observed

when compound and protein were combined at the

millimolar concentrations required by the method. A

combination of high-concentration screening and Teth-

ering was therefore needed to identify the two fragments

comprising 1.

a

(a) LHMDS, methyl 2,3-dichlorobenzoate, THF; (b) MeNHNH₂,

MeOH; (c) H₂NNH₂, MeOH; (d) (Boc)₂O; (e) Tf₂NPh, Et₃N, CH₂Cl₂;

(f) 2,4-dichlorophenylboronic acid, Pd(PPh₃)₄, K₂CO₃, dioxane; (g)

H₂NOH, MeOH; (h) TFFH, Et₃N, DMF; (i) 20, THF; (j) HCl,

dioxane; (k) 8, Et₃N, CH₂Cl₂; (l) N,N′-bis-Boc-1-guanylpyrazole,

Et₃N, MeOH; (m) TFA in CH₂Cl₂.

Rep la cem en t of P ip er id in yl Acetic Acid Lin k er .

Having determined the binding properties of the two

compounds comprising the inhibitor 1, these fragments

were used to identify a new series of inhibitors. The

critical binding elements, the guanidine and biaryl-

acetylene, were retained, and chemistry was focused on

identifying linkers to replace the piperidine acetic acid.

Ease of synthesis, the ability to generate analogues, and

retention of the approximate length of the piperidine

acetic acid were three major criteria used for prioritizing

linker ideas. A dipeptide linker satisfied these criteria

while offering the opportunity to alter the spatial

relationship of the guanidine and biaryl acetylene.

A matrix of 18 compounds was synthesized as an

initial set (Scheme 1) to probe conformation and linker

length. Glycine, sarcosine, and â-alanine were incorpo-

rated at AA1 and six amino acids (glycine, alanine,

D-alanine, sarcosine, proline, D-proline) were positioned

at AA2. Representative data from the matrix are shown

in Table 1. From this set, only one compound, 7c, had

an IC₅₀< 300 µM; equally important, the Hill slope for

the inhibition curve was ∼1, in sharp contrast to the

Resu lts

Id en tifyin g Sta r tin g F r a gm en t fr om 1. The frag-

ments 2 and 3 were derived from 1, and the binding

properties of each component were characterized (Fig-

ures 2, 3). Compound 2 inhibits the IL-2/IL-2RR inter-

action with an inhibition constant (IC₅₀) of 2.2 mM.

Binding of these compounds to IL-2 was monitored by

SPR (see below for description of method). SPR data

were in agreement with the inhibition assay and

indicated a K_dof approximately 1.5 mM for compound

2. By contrast, no binding or inhibitory activity was

observed for 3 at concentrations up to 10 mM. Using

Tethering, we were able to detect covalent binding of a

disulfide derivative of 3 to the Lys43Cys mutant of IL-2

(K43C-IL-2; Figure 3).³³Fragment 2 was also deriva-

tized with a disulfide-containing linker and tested for

binding to K43C-IL-2. No conjugation was observed,

however, suggesting that either the disulfide linker or

the protein was not in a proper conformation for

binding. Neither compound was detected by ¹H-¹⁵N

IL-2 Inhibitors Small-Molecule Antagonists of IL-2

J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12 3115

Ta ble 2. Effect of Terminal-Ring Substitution on Inhibition of

IL-2/IL-2rR

compound

R

IC₅₀(µM)^a

15a

15b

15c

15d

15e

15f

15g

15h

15i

2,4-Cl₂

H

2-Cl

10

120

40

130

10

110

50

150

130

30

3-Cl

2,3-Cl₂

3,4-Cl₂

2-Me

3-Me

2,3-Me₂

2-Cl,4-OH

2-Cl,4-OMe

15j

15k

50

a

Values determined by SPA. n > 2, Hill slopes ) 1 ( ∼0.2 for

all compounds.

inhibition. This information, with the support of ad-

ditional biophysical characterization (see below), sug-

gested that 7c, although a weak inhibitor, functioned

by blocking the IL-2RR site and was worthy of further

optimization.

The synthetic flexibility of the dipeptide framework

facilitated the rapid advancement of 7c (Table 1).

Neither polar side chains nor amino acids with the

natural L-configuration were tolerated at the AA2

position. However, lipophilic, D-amino acids increased

compound activity, with D-tert-butyl alanine (7l) giving

a 30-fold increase in activity relative to D-alanine (7c).

Replacement of glycine at the AA1 position or methyl-

ation of the two backbone nitrogen atoms did not

improve activity. Overall, the SAR was well-defined, and

within a set of ∼15 compounds a compound with an IC₅₀

< 10 µM had been identified.

F igu r e 3. Top, Inhibition of IL-2/IL-2RR by fragments 2 (filled

circles) and 3 (filled squares) shown by SPA (see Experimental

Section). 2 has an IC₅₀) 2.2 mM and a Hill slope ∼ 1. Bottom,

disulfide derivatives of 2 and 3. Tethering at K43C-IL-2

captured 3SS but not 2SS.

Ta ble 1. Effect of Amino Acid Substitution on Inhibition of

IL-2/IL-2RR

Rep la cem en t of Hyd r op h obic F r a gm en t. In the

third step toward discovering a lead series, alternate

hydrophobic fragments were identified through synthe-

sis of a solution-phase parallel library. Approximately

100 diverse primary and secondary amines, containing

at least one aromatic ring and within a molecular

weight range of 100-300, were coupled to the dipeptide

guanidine fragment (Scheme 2). From this library, only

a few compounds had an IC₅₀< 200 µM and all

compounds in this active set incorporated bicyclic

aromatic amines. Simple substituted benzyl or phen-

ethylamines showed little or no inhibition of IL-2/IL-

2RR, consistent with the hypothesis that activity re-

quires occupation of both the guanidine-binding site and

the lipophilic cleft. The most potent compound from the

library, 15a (Table 2), incorporated a tricyclic fragment

containing a piperidine, aryl-pyrazole. The activity of

15a (IC₅₀) 10 µM) and the synthetic versatility of the

tricyclic framework made this compound a starting

point for more detailed optimization.

compound

AA2

glycine

IC₅₀(µM)^a

slope^b

7a

7b

7c

7d

7e

7f

7g

7h

7i

>2000

∼2000

280

1200

1000

720

80

>2

1.4

3.5

5.6

2.2

0.9

n.a.^d

0.8

1.2

1.0

1.4

1.8

alanine

D-alanine

sarcosine^c

proline

D-proline

D-valine

valine

>2000

D-Cha⁵

20

37

26

9

7j

7k

7l

D-Chg⁶

D-Isoleucine

t-Bu-alanine

glutamine

7m

900

a

IC₅₀values determined by ELISA, n g 2; standard errors are

b

within 2-fold of the reported value. Hill slope in ELISA; a value

of 1 is expected for single-site binding. ^cN-methyl glycine. n.a.

d

) not applicable. ^eCyclohexylalanine. ^fCyclohexylglycine.

Op tim iza tion of th e Hyd r op h obic F r a gm en t. To

develop the SAR around library compound 15a , we first

investigated substitution around the phenyl ring (Table

2). Removing both chlorine atoms from the ring reduced

activity more than 10-fold (15b), and this activity could

other 17 compounds synthesized. Compounds 7a ,b,d -f

had a Hill slope of .1, suggesting an artifactual

componentssuch as aggregation or precipitationsto the

3116 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12

Raimundo et al.

Ta ble 3. Effect of Pyrazole Methylation on Inhibition of

IL-2/IL-2RR

Ta ble 4. Biophysical Characterization of Binding of

Compounds to IL-2^a

compd^bIC₅₀(µM)^cK_d,SPR(µM)^dK_d,AU(µM)^{e 15}N,¹H, HSQC^f

1

7

280

26

9

10

19

150

70

15

18

7

20

400

50

+

7c

7k

7l

15a

15e

+

n.t.

+

4

n.t.

8^g

a

Measurements are typically within 2-fold of the reported value.

See Figure 2 and Tables 1 and 2 for compound structures. ^cIC₅₀

b

d

values from ELISA, n g 7. Determined by equilibrium binding

as a function of compound concentration, see Experimental Section

for details. ^eAU ) analytical ultracentrifugation at sedimentation

equilibrium; see Experimental Section for details. n.t., not tested.

^fEvidence for binding at the IL-2 hot spot by NMR, “+” indicates

characteristic shifts observed for E62 and F42, among other

g

residues. Determined for close analogue, (AA2 ) D-isoleu).

significant loss in activity was observed when the

methyl group was introduced at the 2- or the 4-position

(23, 24); in particular, compound 24 showed a 400-fold

loss in potency. Replacing the pyrazole with a phenyl

or oxadiazole rings also resulted in a significant reduc-

tion in activity, suggesting that binding to IL-2 is highly

sensitive to the shape of the tricycle.

Bin d in g of Com p ou n d s to IL-2. Throughout the

chemical optimization process, compounds were ana-

lyzed by SPR, analytical ultracentrifugation (AU), and

NMR (Table 4; Figures 4-6). Experiments were de-

signed to test whether the inhibition of IL-2/IL-2RR

correlates with binding of compounds to IL-2 and to

determine the reversibility and stoichiometry of binding

at the IL-2RR binding site.

Binding of the compounds to IL-2 was characterized

by SPR. By this method, a surface is labeled with IL-2

and the compounds are flowed over the chip’s surface;

binding of compounds causes a change in mass on the

chip, which is detected by the change in reflectance

units (RU). The kinetics of binding can be determined

by monitoring the change in the SPR signal as a

function of time. Each IL-2 inhibitor tested displayed

rapid binding and dissociation kinetics, indicating that

the compounds bind reversibly to IL-2 (Figure 4). The

binding affinity and stoichiometry can also be deter-

mined by monitoring the RU as a function of compound

concentration, since the RU value is directly propor-

tional to the number of molecules binding to the surface

of the chip (Figure 4). The compounds shown in Table

4 bind with 1:1 stoichiometry to IL-2; furthermore, the

dissociation constants measured by SPR give the same

rank-ordering of compounds as the inhibition assays,

although the K_dvalues tend to be slightly higher than

the IC₅₀values. Accuracy of the SPR data can be

affected by nonspecific binding, which we did observe

for this series of compounds at concentrations above

∼500 µM (See Experimental Section).

a

Values determined by SPA. Hill slopes ) 1 ( ∼0.2 for all

compounds.

not be regained by replacement with other substituents.

Introducing small substituents (methyl, chloro, fluoro)

at the 2-position (15c, 15g) improved activity ∼3-fold,

while single substitution at the 3- or 4-position were

nonproductive (15d , 15h ). Activity was significantly

decreased when larger substituents were introduced at

any position (data not shown), and only the 4-position

could tolerate heteroatom substitution (15j). Maintain-

ing an IC₅₀value below 10 µM required incorporation

of a chlorine atom at the 2-position as well as a second

chlorine atom at either the 3- or 4-position. Interest-

ingly, a modest change of substituents, such as replace-

ment of the chlorine atoms with methyl groups (15i)

resulted in significant loss of activity. The strong

preference for dichloro substitution suggests that po-

larizable hydrophobic groups make favorable interac-

tions with the protein surface and/or that these groups

stabilize the bioactive conformation.

Dissociation constants for small molecule/IL-2 inter-

actions were also determined by equilibrium AU (Table

4 and Figure 5).³⁴Varying ratios of compound and

protein are spun in an analytical ultracentrifuge until

sedimentation equilibrium is reached. The sedimenta-

tion curves are then measured by absorbance spectros-

copy at wavelengths where the compound and/or protein

absorb light. The shapes of the sedimentation curves

determine the molecular weight of the protein, and the

absorbance profile for the small molecule permits cal-

The central pyrazole ring was also a site of early

chemical exploration (Table 3). A methyl group was

introduced at the three unsubstituted positions of the

ring in order to identify sites for further elaboration and

to understand the conformational preference of the

tricycle. The activity dropped only 2-fold when a methyl

group was installed at the 1-position (22), but a more

IL-2 Inhibitors Small-Molecule Antagonists of IL-2

J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12 3117

F igu r e 4. SPR measurement for the weak inhibitor 7c. Top,

sensorgram with background subtraction for 7c binding to a

CM5 chip labeled with 1700 RU of IL-2. The dotted line shows

the RU of 7c expected for 1:1 binding. The slope observed in

the plateau region for 1000 µM 7c suggests nonspecific binding

at this high concentration. Bottom, dose-response curve for

the same SPR data. The binding RU at 100 s are plotted as a

function of concentration of compound, and the resulting curve

is fit by nonlinear regression. For this data set, K_d,SPRis

calculated to be 150 µM.

F igu r e 5. Determination of K_dby sedimentation equilibrium

AU for 7c and 1. Top, sedimentation equilibrium curves for

three ratios of 7c:IL-2 at 25K RPM. Note that as concentration

of protein is raised, the binding of compound increases,

resulting in increased curvature. Inset, result of fitting AU

data to determine K_d.³⁴The minimum in the parabola gives

the negative-log of the dissociation constant pK. For this data

set, pK ) 3.4, K_d) 400 µM. Bottom, sedimentation equilibrium

curves for three ratios of 1:IL-2 at 25K RPM. Inset, result of

fitting AU data to determine K_d. For this data set, pK ) 4.7,

K_d) 20 µM. Compound concentration is 25 µM; protein

concentrations are shown on the plots. Sedimentation curves

are an average of five scans, measured in 0.003 mm incre-

ments. Extinction coefficients at ∼300 nm: IL-2, 1000 M^-1

cm^-1; 7c, 22 000 M^-1cm^-1; 1, 13 000 M^-1cm^-1. For experi-

mental details, see Experimental Section.

culation of the K_dfor the interaction. AU can also

identify compounds that cause the protein to aggregate;

the molecular weight of IL-2 was unchanged in the

presence of inhibitors, indicating the absence of ag-

gregation. From Figure 5, one can observe that the

weakly bound compound 7c (K_d) 400 µM by AU) shows

less curvature than the more tightly bound compound

1 (K_d) 20 µM by AU). As shown in Table 4, K_dvalues

obtained by AU provide the same rank-order of com-

pounds as do the IC₅₀and SPR measurements.

¹H-¹⁵N HSQC NMR spectroscopy was used to deter-

mine the binding site of 7c and subsequent members

in this chemical class. Figure 6 shows that the sub-

millimolar compound 7c, the low-micromolar inhibitor

7l, and the tricyclic compound 15a bind at the IL-2RR

site in a manner similar to 1.^31,32As shown in Figure 2,

Glu 62 and Phe 42 make direct contact with the small

molecule; thus, the position of their NMR resonances

serve as diagnostics for binding to the IL-2RR site.

Moreover, the degree to which these resonances shift

correlates with binding affinity of compound. Even the

weakly bound compound 7c induces a shift in Glu 62,

indicating that molecules with an affinity in the 300 µM

range demonstrate specific binding at this site. Although

all compounds bind at the same site, subtle differences

in the NMR shifts of the dipeptide series (15a , 7l) versus

1 indicate that the details of the binding contacts may

vary, particularly in the region of the Glu 62-guanidine

interaction.

X-r a y Str u ctu r e of 15e. The X-ray structure of 15e

bound to IL-2 confirms that this compound binds in a

manner very similar to 1 (Figure 7),^33,35though differ-

ences are observed in the guanidine-Glu 62 interaction.

The guanidine of 15e makes one hydrogen bond with

Glu 62, while 1 makes two hydrogen bonds with this

residue. The cyclohexyl group sits in a hydrophobic cleft

formed by the side chains of Tyr 45, Lys 43, and Thr

111. This interaction explains the strong preference for

lipophilic D-amino acids at AA2 and represents a new

binding interaction not observed in the structure of 1

bound to IL-2. The dipeptide linker, in contrast to the

piperidine acetic acid of 1, makes minimal contact with

the protein surface. However, in the hydrophobic site,

the tricyclic fragment complements the lipophilic cleft,

with each ring contacting the protein surface. Consistent

with the SAR, methyl-group substitution is tolerated at

the solvent-exposed 1-position of the pyrazole, but

neither a methyl group nor a heteroatom can be accom-

modated at the buried 4-position (Table 3). Around the

3118 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12

Raimundo et al.

F igu r e 6. ¹H-¹⁵N HSQC NMR spectra of IL-2 in the presence

1

and absence of small-molecule inhibitors. Left, H-¹⁵N cross-

peak corresponding to Glu 62 (E62). Right, ¹H-¹⁵N cross-peak

corresponding to Phe 42 (F42). NMR samples contain 200 µM

IL-2 uniformly labeled with ¹⁵N and 1% DMSO, in a buffer of

10 mM sodium acetate, pH 4.5. Spectra were measured at 40

°C in the presence of the following compounds: IL-2 only

(black), 500 µM 1 (blue), 2 mM 7c (magenta), 500 µM 7l

(green), and 500 µM 15a (red). In each case, the magnitude of

the cross-peak shift correlates with binding affinity. Note that

even the weak inhibitor 7c causes Glu 62 and Phe 42 cross-

peaks to shift, suggesting that this compound binds in the

same site as 1, 7l, and 15a . The placement of these residues

in the small-molecule binding site of IL-2 is shown in Figures

2 and 7.

F igu r e 8. Position of Leu 72 and sample hit obtained at this

residues by Tethering. Top, structure of 15e (cyan) bound to

IL-2 (white ribbon). Leu72Cys was used to identify fragments

in the adaptive region adjacent to the binding site for 15e.

Bottom, strongly selected fragment from Leu72Cys. Statistical

analysis shows a strong selection for small, aromatic, acidic

fragments at this site.³⁶Modeling of the salicylic acid fragment

from Leu72Cys suggested that acidic fragments could be

appended to 15e at the 4-position through a two-atom linker.

mutants of IL-2 were constructed and screened against

a library of 7000 disulfide-containing fragments. Al-

though these mutants were distributed around the two

subsites, most of the hits were obtained from mutants

targeting the lipophilic region. Aromatic acids were

preferentially selected within this region; one example,

a salicylate from Leu72Cys is shown in Figure 8.

Inspection of Leu 72 in the 15e/IL-2 cocrystal struc-

ture suggested that these aryl carboxylic acids bind in

the hydrophobic adaptive region adjacent to the binding

site of 15e.³⁶Early SAR of the phenyl pyrazoles (Table

2) demonstrated that an ether was tolerated at the

4-position of the phenyl ring, thus providing a syntheti-

cally attractive handle and an appropriate vector for

targeting this adaptive region. We first constructed a

template compound that incorporated the preferred

substitutions (2,3-dichloro substitution of the phenyl

ring, 2-methylation of the pyrazole, and D-leucine at

AA2) into one molecule. The resulting compound, 33c

(Table 5), had an IC₅₀value of 3.5 µM and was used as

the basis for further libraries.

F igu r e 7. Structure of 15e bound to IL-2 from X-ray crystal-

lography (Protein Data Bank ID code 1PW6).³⁵15e is shown

in stick representation (carbon atoms in cyan, nitrogen in blue,

oxygen in red, chlorine in pink); IL-2 is shown as a white

surface. IL-2 residues within six angstroms of the compound

are shown in stick representation (carbon atoms in white,

nitrogen in blue, oxygen in red, sulfur in orange). Hydrogen

bonds between the compound and protein are shown with

yellow dashed lines (guanidine to Glu 62 and carbonyl oxygen

to backbone nitrogen of Lys 43). Labeled IL-2 residues are

discussed in the text and Figure 6. 15e binds in the IL-2RR

binding site determined by site directed mutagenesis.^31,54,55

In general, merging aryl acids onto compound 33c

resulted in a significant enhancement of activity (Table

5).³⁶A carboxylate functionality was critical to achieving

submicromolar activity, although its exact placement

about the aromatic ring was found to be less important.

Addition of a benzoic acid fragment resulted in an IC₅₀

of 200 nM (33f). Removing the carboxylate (33d ) or

masking it as the methyl ester (33e) reduced activity

by more than 100-fold. Shifting the carboxylate to the

meta-position (33g) or extending the carboxylate by one

methylene unit (33h ) resulted in a modest loss of

activity. Of the 20 compounds synthesized, eight had

an IC₅₀< 1 µM. The most potent compound, 33i,

phenyl ring, both chlorine atoms bury into the protein

surface, explaining the preference for small, lipophilic

substituents at the 2- and 3-positions. The costructure

of 15e bound to IL-2 thus corroborates the SAR and

biophysical data.

In cor p or a tion of F r a gm en ts fr om Teth er in g. We

applied Tethering to identify fragments that could

complement the IL-2RR binding site and assist in

compound optimization.^29,33,36Ten individual cysteine

incorporated a furanoic acid moiety and had an IC₅₀)

60 nM. Thus, Tethering provided information about the

preferred binding elements at the adaptive region that,

in turn, directed the chemical optimization. Synthesiz-

IL-2 Inhibitors Small-Molecule Antagonists of IL-2

J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12 3119

Ta ble 5. Effect of Addition of Aryl Carboxylic Acids on

Inhibition of IL-2/IL-2RR

F igu r e 9. Structure of 33i bound to IL-2 from X-ray crystal-

lography (Protein Data Bank ID code 1PY2).³⁵33i is shown

in stick representation (carbon atoms in pink, nitrogen in blue,

oxygen in red, chlorine in dark pink); IL-2 is shown in surface

representation, and residues within 6 Å of 33i are shown in

stick form (carbon atoms in white, nitrogen in blue, oxygen in

red, sulfur in orange). Hydrogen bonds between the compound

and protein are shown with yellow dashed lines (guanidine to

Glu 62, carbonyl oxygen to the backbone nitrogen atom of Lys

43). 33i binds similarly to 15e (Figure 7), at the binding site

for IL-2RR determined by site-directed mutagenesis.^31,54,55Note

that the furanoic acid binds in a groove formed by the

positively charged residues Arg 38 and Lys 35. Labeled

residues are discussed in the text.

a

Values determined by SPA. Hill slopes ) 1 ( ∼0.2 for all

compounds.

F igu r e 10. Inhibition of IL-2 dependent, STAT-5 phospho-

rylation in the presence of 33i. STAT-5 phosphorylation was

monitored by Western blot, using an anti-phosphotyrosine

antibody, an alkaline-phosphatase-labeled secondary antibody,

and a luminescent substrate. Luminescence was recorded on

film and quantified by densitometry, and reduction in band

intensity was plotted as a function of compound concentration.

Data shown are an average of three separate experiments.

ing a small set of compounds defined by the Tethering

data led to a 30-fold increase in activity.

X-r a y Str u ctu r e of 33i. The X-ray crystal structure

of 33i bound to IL-2, shown in Figure 9, supports the

design strategy and explains the observed SAR.³⁵As

expected, 33i binds analogously to 15e (Figure 7). One

striking difference in the structure of 33i is the guani-

dine moiety, which adopts a conformation very similar

to that of 1 (Figure 2) and makes two hydrogen bonds

to Glu 62. The greater flexibility of the D-leucine side

chain relative to the D-cyclohexyl glycine of 15e probably

allows for this reorientation of the guanidine. There are

also minor differences in the orientation of the tricycle

within the hydrophobic cleft. The furanoic acid moiety

is bound in a shallow pocket next to the hydrophobic

cleft and is surrounded by the positively charged side

chains of Arg 38 and Lys 35. These side chains are not

well-ordered in the X-ray structure, and there are no

apparent hydrogen bonds between the carboxylate and

the protein. Rather, it appears that the increased

affinity seen for 33i is due to the complimentary

electrostatic potentials of the negatively charged acid

and the positively charged surface of IL-2 in this region

(unpublished observations). This notion is supported by

the observed SAR, since the specific placement of the

acidic group has only a minor effect on the activity of

compounds in this series.

Biologica l P r op er ties of 33i. 33i was tested for IL-2

antagonism in a cell-based assay. Human IL-2 binds to

mouse IL-2R expressed on the surface of the murine T

cell line CTLL-2. IL-2 activates intracellular signaling,

which leads to phosphorylation of the transcription

factor STAT5.³⁷In the presence of 33i, phosphorylation

of STAT5 is reduced in a dose-dependent manner with

an EC₅₀) 3 µM (Figure 10). In the same assay, 1 shows

little inhibition of STAT5 phosphorylation at 100 µM

concentration, 33a gives an EC₅₀∼ 25 µM, and soluble

IL-2RR has an EC₅₀∼ 30 nM (3-10-fold weaker than

observed in SPA and ELISA). As a control, the inhibition

of IL-15 signaling was measured. IL-15 signals through

a trimeric receptor that includes the â and γ chains of

IL-2R, but the IL-15-specific chain IL-15RR;³therefore,

compounds that bind to IL-2 at the IL-2RR site should

3120 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12

Raimundo et al.

have no effect on IL-15 signaling. As expected, 33i and

IL-2RR were inactive against IL-15-mediated phospho-

rylation of STAT5.

interaction. From there, efficient optimization of the

dipeptide linker further boosted affinity by incorporat-

ing a new binding element (hydrophobic D-amino acid).

Throughout the process, these low-affinity compounds

demonstrated the correct behavior profile, including

inhibition curves with Hill slopes close to 1, solubilities

in excess of the IC₅₀value, and evidence for reversible,

low stoichiometry binding. The process of evolving

fragments to micromolar hits was successful because the

recognition requirements and shape of the binding site

were known, the chemical strategy allowed rapid paral-

lel synthesis, and early characterization ensured that

weak inhibitors bound in the desired site on IL-2.

The in vitro and in vivo pharmacokinetic properties

of 33i were also analyzed. In the presence of human

liver microsomes, the compound was found to be stable,

with 80% ((1%, n ) 2) remaining after 60-min incuba-

tion at 37 °C. The pharmacokinetic properties in mice

were also favorable, with the compound showing a 2.1

h terminal half-life, a clearance rate of 13.6 mL/min/

kg, and a volume of distribution at steady state (V_ss) of

1.1 L/kg.

Discu ssion

Com p ou n d Ad va n cem en t t h r ou gh Tet h er in g.

Fragment discovery tools are typically used to identify

new starting points for chemistry,^18-20,22,23but they can

also be helpful for optimizing a chemical lead series.^21,40

Here, Tethering identified a class of fragments that

bound in a new site adjacent to the known ligand-

binding site. Incorporation of these fragments into the

existing lead series led to a dramatic increase in binding

affinity.

F r a gm en t Ch a r a cter iza tion . To fulfill the potential

of fragment assembly, robust methods must be devel-

oped for identifying the fragments, characterizing their

binding, and linking them productively. As a proof-of-

concept for fragment identification, we began with a

known small-molecule ligand and examined its compo-

nent fragments for binding to IL-2. These two fragments

represent two different classes of molecules. Compound

2 is hydrophobic and binds in an adaptive region of the

protein, while compound 3 is hydrophilic and binds in

a shallow, preformed pocket utilizing a hydrogen bond

network. Given these structural differences, it is not

surprising that different fragment screening methods

are needed to identify these compounds.

Fragment 2 has millimolar activity in functional

screens (SPA, ELISA) and in a binding assay (SPR) but

was not captured by Tethering at K43C-IL-2. This result

is consistent with the structural data on IL-2; without

ligands or with disulfide-bound fragments,^29,33,38,39Phe

42 is found in an “up” conformation, acting as a wall

between the hydrophilic and hydrophobic subsites.

Thus, a fragment bound at Lys43Cys would be pre-

vented from binding the hydrophobic site. Tethering

does capture numerous fragments when the cysteine

mutant is placed on the other side of Phe 42, and these

hits recapitulate the hydrophobic, aromatic properties

of the biaryl acetylene.^29,33

In contrast to 2, fragment 3 did not show activity in

functional assays, suggesting that its binding affinity

is weaker than that of 2. However, 3 was captured by

Tethering at K43C-IL-2. Binding of fragments in the

hydrophilic site requires satisfying a number of hydro-

gen bonds within certain geometric constraints. Fur-

thermore, highly soluble fragments have little energy

to release by binding to the protein, and thus inherent

affinities may be low. Tethering facilitates binding of

such fragments by augmenting the affinity through

formation of a reversible covalent bond. In general, we

find that a combination of fragment discovery methods

is more successful in finding a range of fragment

structures than any one method alone.

Evolvin g F r a gm en ts to Hits. Often, the most

difficult step in fragment assembly is advancing com-

pounds from millimolar to micromolar affinity. For IL-

2, the starting point was a hydrophobic fragment with

millimolar affinity, and the parameters that defined the

adjacent hydrophilic binding site were known. Within

these boundaries and using a modular chemical ap-

proach, we readily attained a 10-fold improvement in

activity and recaptured the Glu 62-guanidine binding

Tethering greatly facilitates lead optimization by

providing information on the location of the fragments’

binding site.²³Although it was difficult to model the

binding modes of fragments selected in the flexible

hydrophobic region, it was straightforward to estimate

the distance between the disulfide-bound fragments and

the terminal phenyl ring of the tricyclic compound given

the relatively fixed position of Leu 72.^33,36The design

strategy was supported by the activity data, since all of

the aromatic acids attached by an ether linker to 33c

led to improved potency. It is noteworthy that early SAR

around the tricycle did not uncover additional binding

affinity. This may not be surprising, given that an

activity boost required a large step in compound com-

plexity, with a carboxylic acid and aromatic ring prop-

erly connected to the scaffold. Fragment-based ap-

proaches can be especially valuable for identifying

complex binding elements adjacent to known binding

sites. Tethering provides a useful way to interrogate

these novel subsites without requiring detailed knowl-

edge of the protein conformation.

IL-2 a s Test Ca se for P r otein -P r otein In ter a c-

tion s. Historically, protein-protein interactions have

been difficult to inhibit with small, organic molecules,

although some successes have been reported.^12-16One

likely reason for this general intractability is the

structure of protein surfaces; in crystal structures of

protein-protein pairs, the interfaces tend to be rather

flat and lacking the deep pockets and grooves seen in

more traditional drug targets such as enzymes and

G-coupled protein receptors.^41-43On the other hand,

work with IL-2 and other proteins indicates that these

protein surfaces can be highly adaptive^29,33,44-48and

might therefore contain transient structures that are

amenable to small-molecule binding.

The importance of IL-2 adaptivity in small-molecule

recognition is highlighted by the structures of 15e and

33i bound to IL-2. The surface of IL-2 undergoes

numerous structural adaptations that are only observed

when small molecules are bound. For example, Phe 42

is always found in an up-position in the absence of 1,

15e, and 33i.^{29,33,35,38,39}In the presence of these ligands,

IL-2 Inhibitors Small-Molecule Antagonists of IL-2

J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12 3121

Applied Biosystems Qstar Pulsar-i. Flash column chromatog-

raphy was performed using EM Science silica gel 60 (230-

400 mesh). Elemental analyses were performed by Robertson

Microlit Laboratories, Inc., Madison, NJ . Analytical thin layer

chromatograms (TLCs) were run using glass thin layer plates

coated with silica gel as supplied by E. Merck and were

visualized by viewing under 254 nm UV or by exposure to a

potassium permanganate solution in ethanol. Analytical HPLC

was conducted on an Agilent 1100 LC/MS system using a C-18

Synergi Hydro-RP 4.6 mm × 150 mm (Phenomenex) at a flow

rate of 1.5 mL per minute. The mobile phase consisted of water

(A) and acetonitrile (B) each containing 0.1% trifluoroacetic

acid by volume. Two methods were employed to assess the

purity of compounds. In Method I, the gradient began at 5%

B and ramped linearly to 100% B over 25 min. In Method II,

the mobile phase was kept isocratic at 35% B for 24 min and

then ramped up to 100% B over 1 min. UV absorbance was

monitered at 220 nm and 254 nm. Preparative reverse phase

high-pressure liquid chromatography (RP-HPLC) was carried

out on a Gilson HPLC fitted with a Waters Nova-Pak C-18

(25 × 100 mm) column eluting at 25 mL/min employing a

gradient of acetonitrile:water (each containing 0.1% TFA) over

10 min and holding at 100% water for 3 min. Solvents and

reagents were obtained from Aldrich Chemical Co., Lancaster,

Nova Biochem, Fluka, Bionet, or Advanced ChemTech and

used as received unless otherwise stated.

Phe 42 rests against the protein surface, where it

provides a binding surface for the small molecule. There

are also small rearrangements around the hydrophilic

site upon binding of the lipophilic D-amino acid in 15e.

The hydrophobic region also adapts upon small-molecule

binding, maintaining a similar conformation when

bound with ligands 1 and 15e. In the costructure of 33i,

there appears to be a further organization of the site

occupied by the furanoic acid,³⁵and this conformation

could maximize the electrostatic interactions between

the acid and basic side chains on IL-2. Thus, at each

stage in the small-molecule discovery process, the

structure of IL-2 is seen to change in response to new

binding elements introduced onto the molecule. Iden-

tifying and targeting adaptive regions of structure might

therefore be a valuable approach for inhibitor discovery

at protein interfaces.

Another reason that few drug leads have emerged for

protein-protein targets could be the prevalence of

screening hits that inhibit by an unacceptable mecha-

nism. While screening artifacts can arise for any target,

they might be more difficult to characterize for protein-

protein systems, where powerful tools such as enzyme

kinetics are not available. In characterizing inhibitors

of IL-2, therefore, we considered it important to dem-

onstrate that inhibition resulted from single-site binding

at the IL-2RR binding site. It is noteworthy that

hydrogen bond interactions between Glu 62 and a

guanidine moiety appear to be critical for IL-2 inhibition

by small molecules. By enforcing a particular geometry,

hydrogen-bonding interactions set the register for the

more malleable hydrophobic binding interactions and

increase the likelihood that the compound will bind at

only one site. Also, the addition of charged elements in

the compound increase its solubility and decrease the

likelihood of aggregation, precipitation, or denatur-

ation.^49,50From this perspective, the IL-2 binding site

seems to be particularly well designed for small-

molecule inhibitors. Nevertheless, the lessons learned

through this small-molecule discovery effort should be

generally applicable to other protein-protein targets.

(S )-2-(2-Me t h oxy-a ce t yla m in o)-3-(4-p h e n yle t h yn yl-

p h en yl)-p r op ion ic Acid (2). To a mixture of (S)-2-amino-3-

(4-phenylethynyl-phenyl)-propionic acid methyl ester hydro-

chloride³¹(4, 90 mg, 0.285 mmol) and pyridine (140 µL, 1.73

mmol) in dichloromethane (1.0 mL) was added methoxyacetyl

chloride (58 µL, 0.626 mmol) dropwise. After 3 h at ambient

temperature, the reaction was diluted with ethyl acetate and

washed sequentially with 1 N HCl (2×), saturated sodium

bicarbonate (2×), and brine (2×). The organic layer was dried

(Na₂SO₄) and concentrated. The crude residue was purified

by flash column chromatography (SiO₂: gradient 0 to 60%

ethyl acetate in hexane) to yield 70 mg (70%) of a film. ¹H

NMR (400 MHz, CDCl₃) δ 7.52 (m, 2 H), 7.46 (d, 2 H, J ) 8.0

Hz), 7.35 (m, 3 H), 7.12 (d, 2 H, J ) 8.0 Hz), 6.96 (d, 1 H, J )

8.2 Hz), 4.93 (m, 1 H), 3.89 (s, 2 H), 3.72 (s, 3 H), 3.37 (s, 3 H),

3.18 (dd, 1 H, J ₁) 14.1 Hz, J ₂) 5.9 Hz), 3.10 (dd, 1 H, J ₁

)

14.0 Hz, J ₂) 5.9 Hz); TLC (SiO₂: 50% ethyl acetate in

hexane): R_f) 0.18; ES (+) MS m/e ) 352 (M + 1).

To a solution containing the amide (53 mg, 0.150 mmol) in

THF (0.84 mL) was added aqueous LiOH (0.80 M, 0.56 mL,

0.45 mmol). The reaction was stirred for 3 h followed by the

addition of diethyl ether (3.0 mL) and saturated sodium

bicarbonate. The layers were separated, and the organic phase

was extracted with 2.0 mL saturated sodium bicarbonate. The

combined aqueous layers were acidified with concentrated HCl

and extracted with dichloromethane four times. The organic

phases were combined, dried (MgSO₄), and concentrated. The

crude product was purified by reverse phase HPLC. The

fractions containing pure compound were combined and

concentrated. The residue was lyophilized to yield 10.9 mg

Con clu sion s

This report describes the evolution of a compound

from millimolar affinity to nanomolar affinity. Often,

the most difficult step in fragment assembly technology

is advancing the molecules from the millimolar to

micromolar stage. In the case of IL-2, this advancement

was accomplished by linking two key binding elements

with a chemically and structurally modular linker. The

chemical series thus identified demonstrated attractive

binding properties but reached an affinity plateau in

the low micromolar range. This plateau was surpassed

by incorporating fragments identified through Tether-

ing. Using fragments in both the early and later stages

of compound design efficiently improved compound

potency and accessed new chemical space.

1

(22%) of 2 a white powder. H NMR (400 MHz, CDCl₃) δ 7.49

(m, 4 H), 7.33 (m, 3 H), 7.17 (d, 2 H, J ) 7.9 Hz), 7.01 (d, 1 H,

J ) 8.0 Hz), 4.93 (app q, 1 H, J ) 6.5 Hz), 3.91 (s, 2 H), 3.36

(s, 3 H), 3.27 (dd, 1 H, J ₁) 14.1 Hz, J ₂) 5.4 Hz), 3.16 (dd, 1

H, J ₁) 14.0 Hz, J ₂) 6.5 Hz); TLC (SiO₂: 2% AcOH and 8%

MeOH in CHCl₃): R_f) 0.15; Anal. (C₂₀H₁₉NO₄) C, H, N.

(S)-2-(2-Am in o-a cetyla m in o)-3-(4-p h en yleth yn yl-p h en -

yl)-p r op ion ic Acid Meth yl Ester Hyd r och lor id e (5). To a

solution of 4 (0.25 g, 0.8 mmol) in dichloromethane (6 mL) were

added N-Boc-glycine (0.18 g, 1.0 mmol), EDC (0.19 g, 1.0

mmol), HOBt (0.15 g, 1.0 mmol), and triethylamine (0.28 mL,

2.0 mmol). The mixture was stirred overnight and then was

partitioned between dichloromethane and water. The aqueous

layer was washed with dichloromethane (2×), and the organic

layer was washed with 1 N HCl and saturated NaHCO₃. The

organic layer was dried over Na₂SO₄, filtered, and concentrated

in vacuo to provide the desired amide. The crude residue was

dissolved in HCl/dioxane (4.0 N, 5 mL), stirred for 45 min, and

Exp er im en ta l Section

Gen er a l Meth od s. ¹H NMR spectra were recorded on a

Bruker 400 MHz spectrometer with chemical shifts reported

in units of parts per million (ppm). Optical rotations were

determined with a J asco model P-1020 polarimeter. ESI mass

spectra were obtained on a HP 1100MSD mass spectrometer.

High-resolution mass spectra (HRMS) were determined on a

3122 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12

Raimundo et al.

then concentrated in vacuo to afford 350 mg (over theory) of a

solid. H NMR (400 MHz, CD₃OD) δ 7.5 (m, 4 H), 7.37 (m, 3

H), 7.2 (m, 2H), 4.8 (m, 1 H), 3.7 (s, 3 H), 3.6 (m, 2 H), 3.3 (m,

1 H), 3.15 (m, 1 H); ES (+) MS m/e ) 338 (M + 1).

the procedures described for the synthesis of 7c but using

N-Boc-^D-proline in place of N-Boc-D-alanine, 7f was prepared

as a white powder. ¹H NMR (400 MHz, CD₃OD) δ 8.43 (d, J )

7.8, 2 H), 7.42-7.49 (m, 3 H), 7.22 (d, J ) 7.9, 2 H), 4.72 (dd,

J ) 8.0, 5.6, 1 H), 4.47 (d, J ) 6.4, 1 H), 3.96 (d, J ) 16.8, 1

H), 3.84 (d, J ) 16.8, 1 H), 3.59 (s, 3 H), 3.46-3.49 (m, 1 H),

3.42-3.45 (m, 1 H), 3.18 (dd, J ) 13.7, 5.5, 1 H), 3.00 (dd, J )

13.7, 8.3, 1 H), 2.32-2.38 (m, 1 H), 2.16-2.24 (m, 1 H), 1.93-

2.07 (m, 2 H). ES (+) MS m/e ) 476 (M + 1).

2S-[(2R)-(2-Gu a n id in o-3-m eth yl-bu tyr yla m in o)-a cetyl-

am in o]-3-(4-ph en yleth yn yl-ph en yl)-pr opion ic Acid Meth -

yl E st er Tr iflu or oa cet a t e (7g). Following the procedures

described for the synthesis of 7c but using N-Boc-^D-valine in

place of N-Boc-^D-alanine, 7g was prepared as a white powder.

¹H NMR (400 MHz, CD₃OD) δ 7.20-7.55 (m, 9 H), 4.72 (dq, J

) 7.8, 5.5, 1 H), 3.90-4.00 (m, 3 H), 3.73 (s, 3 H), 3.19 (dd, J

) 13.8, 5.3, 1 H), 3.01 (dd, J ) 8.5, 5.2, 1 H), 2.20-2.25 (m, J

) 12.8, 6.5, 1 H), 0.82 (d, J ) 6.7, 3 H), (0.74 (d, J ) 6.7, 3 H).

ES (+) MS m/e ) 478 (M + 1).

1

(2S)-[2-((2R)-Am in o-p r op ion yla m in o)-a cetyla m in o]-3-

(4-p h en yleth yn yl-p h en yl)-p r op ion ic Acid Meth yl Hyd r o-

ch lor id e (6c). To a solution of 5 (0.1 g, 0.27 mmol) in

dichloromethane (2 mL) were added N-Boc-^D-alanine (66 mg,

0.35 mmol), EDC (67 mg, 0.35 mmol), HOBt (53 mg, 0.35

mmol), and triethylamine (0.1 mL, 0.7 mmol). The mixture

was stirred overnight and then was partitioned between

dichloromethane and water. The aqueous layer was washed

with dichloromethane, and the organic layer was washed with

1 M HCl and saturated NaHCO₃. The organic layer was dried

over Na₂SO₄, filtered, and concentrated in vacuo. The crude

residue was dissolved in HCl/dioxane (4.0 N, 5 mL), stirred

for 30 min, and then concentrated in vacuo to provide 120 mg

(100%) of the corresponding amine hydrochloride salt.

(2S)-[2-((2R)-Gu an idin o-pr opion ylam in o)-acetylam in o]-

3-(4-p h en yleth yn yl-p h en yl)-p r op ion ic Acid Meth yl Ester

Tr iflu or oa ceta te (7c). To a solution of 6c (75 mg, 0.17 mmol)

in MeOH (2.5 mL) were added (tert-butoxycarbonylimino-

pyrazol-1-yl-methyl)-carbamic acid tert-butyl ester (77 mg, 0.25

mmol) and triethylamine (45 µL, 0.32 mmol). The mixture was

stirred overnight and then was partitioned between dichlo-

romethane and water. The aqueous layer was washed with

dichloromethane (2×), and the organic layer was washed with

1 N HCl, dried over Na₂SO₄, filtered, and concentrated in

vacuo. The residue was dissolved in dichloromethane (1.5 mL)

and TFA (1.5 mL) was added. The solution was stirred for 2 h

and then concentrated to an oil. Purification by RP HPLC

followed by lyophilization yielded 7c (46 mg, 48%) as a white

solid. ¹H NMR (400 MHz, CD₃OD) δ 8.45 (d, J ) 7.5, 1 H),

8.34 (m, 1 H), 7.61 (d, J ) 7.4, 1 H), 7.47 (m, 4 H), 7.35 (m, 3

H), 7.23 (d, J ) 7.9, 2 H), 4.74 (m, 1 H), 4.2 (m, 1 H), 3.9 (m,

2 H), 3.8 (s, 3H), 3.19 (dd, J ) 13.8, J ) 5.5, 1 H), 3.01 (dd, J

) 13.7, 8.4, 1 H), 1.44 (d, J ) 6.9, 3 H); ES (+) MS m/e ) 450

(M + 1); Anal. (C₂₄H₂₇N₅O₄‚1.6CF₃CO₂H) C, H, N.

2S-[(2S)-(2-Gu a n id in o-3-m eth yl-bu tyr yla m in o)-a cetyl-

am in o]-3-(4-ph en yleth yn yl-ph en yl)-pr opion ic Acid Meth -

yl Ester Tr iflu or oa ceta te (7h ). Following the procedures

described for the synthesis of 7c but using N-Boc-valine in

place of N-Boc-^D-alanine, 7h was prepared as a white powder.

ES (+) MS m/e ) 478 (M + 1).

2S-[2-((2R)-3-Cycloh exyl-2-gu an idin o-pr opion ylam in o)-

a cetyla m in o]-3-(4-p h en yleth yn yl-p h en yl)-p r op ion ic Acid

Meth yl Ester Tr iflu or oa ceta te (7i). Following the proce-

dures described for the synthesis of 7c but using N-Boc-^D-

cyclohexylalanine in place of N-Boc-^D-alanine, 7i was prepared

as a white powder. ES (+) MS m/e ) 532 (M + 1).

(2S)-[2-((2R)-Cycloh exyl-2-gu a n id in o-a cet yla m in o)-

a cetyla m in o]-3-(4-p h en yleth yn yl-p h en yl)-p r op ion ic Acid

Meth yl Ester Tr iflu or oa ceta te (7j). Following the proce-

dures described for the synthesis of 7c but using N-Boc-^D-

cyclohexylglycine in place of N-Boc-^D-alanine, 7j was prepared

1

as a white powder. H NMR (400 MHz, CD₃OD) δ 8.4 (d, J )

7.7, 1 H), 8.28 (m, 1 H), 7.49 (m, 4 H), 7.35 (m, 3 H), 7.23 (d,

J ) 7.9, 2 H), 4.72 (m, 1 H), 3.9 (m, 3 H), 3.7 (s, 3 H), 3.19 (dd,

J ) 13.8, J ) 5.3, 1 H), 3.01 (dd, J ) 13.7, 8.3, 1 H), 1.75 (m,

6 H), 1.2 (m, 5 H). ES (+) MS m/e ) 518 (M + 1).

(S)-2-[2-(2-Gu a n id in o-a cetyla m in o)-a cetyla m in o]-3-(4-

p h en yleth yn yl-p h en yl)-p r op ion ic Acid Meth yl Ester Tr i-

flu or oa ceta te (7a ). Following the procedures described for

the synthesis of 7c but using N-Boc-glycine in place of N-Boc-

2S-[2-((2R)-2-Gu a n id in o-3-m et h yl-p en t a n oyla m in o)-

a cetyla m in o]-3-(4-p h en yleth yn yl-p h en yl)-p r op ion ic Acid

Meth yl Ester Tr iflu or oa ceta te (7k ). Following the proce-

dures described for the synthesis of 7c but using N-Boc-^D-

leucine in place of N-Boc-^D-alanine, 7k was prepared as a

1

D-alanine, 7a was prepared as a white powder. H NMR (400

MHz, CD₃OD) δ 7.5 (m, 4 H), 7.3 (m, 3 H), 7.17 (m, 1 H), 4.73

(m, 1 H), 3.95 (s, 2 H), 3.9 (m, 2 H), 3.8 (s, 3 H), 3.20 (dd, J )

13.5, J ) 5.4, 1 H), 3.01 (dd, J ) 13.8, 8.7, 1 H). ES (+) MS

m/e ) 436 (M + 1); HRMS (TOF) calcd for C₂₃H₂₆N₅O₄(M +

H)⁺436.1985, found 436.1974. HPLC: Purity g98%.

(2S)-[2-((2S)-Gu an idin o-pr opion ylam in o)-acetylam in o]-

3-(4-p h en yleth yn yl-p h en yl)-p r op ion ic Acid Meth yl Ester

Tr iflu or oa ceta te (7b). Following the procedures described

for the synthesis of 7c but using N-Boc-alanine in place of

N-Boc-^D-alanine, 7b was prepared as a white powder. ¹H NMR

(400 MHz, CD₃OD) δ 8.44 (d, J ) 7.9, 1 H), 8.33 (m, 1 H), 7.61

(d, J ) 7.5, 1 H), 7.47 (m, 4 H), 7.35 (m, 3 H), 7.23 (d, J ) 7.9,

2 H), 4.75 (m, 1 H), 4.15 (m, 1 H), 4.0 (m, 1 H), 3.9 (m, 1 H),

3.8 (s, 3 H), 3.19 (dd, J ) 13.8, J ) 5.2, 1 H), 3.0 (dd, J ) 13.8,

8.9, 1 H), 1.44 (d, J ) 6.9, 3 H). ES (+) MS m/e ) 450 (M + 1);

Anal. (C₂₄H₂₇N₅O₄‚1.5CF₃CO₂H) C, H, N.

1

white powder. H NMR (400 MHz, CD₃OD) δ 8.41 (d, J ) 7.7,

1 H), 8.31 (s, 1 H), 7.47-7.49 (m, 2 H), 7.44 (d, J ) 7.8, 2 H),

7.36-7.43 (m, 3 H), 7.23 (d, J ) 7.0, 2 H), 4.71-4.75 (m, 1 H),

3.85-3.96 (m, 3 H), 3.73 (s, 3 H), 3.19 (dd, J ) 13.8, 5.4, 1 H),

3.02 (dd, J ) 13.8, 8.3, 1 H), 1.93-1.97 (m, 1 H), 1.48-1.55

(m, 1 H), 1.12-1.22 (m, 1 H), 0.77-0.94 (m, 6 H). ES (+) MS

m/e ) 492 (M + 1).

2S-[2-((2R)-2-Gu an idin o-4,4-dim eth yl-pen tan oylam in o)-

a cetyla m in o]-3-(4-p h en yleth yn yl-p h en yl)-p r op ion ic Acid

Meth yl Ester Tr iflu or oa ceta te (7l). Following the proce-

dures described for the synthesis of 7c but using N-Boc-^D-tert-

butylalanine in place of N-Boc-^D-alanine, 7l was prepared as

a white powder. ES (+) MS m/e ) 506 (M + 1).

(2S)-{2-[2-(N-Meth yl-gu a n id in o)-a cetyla m in o]-a cetyl-

am in o}-3-(4-ph en yleth yn yl-ph en yl)-pr opion ic Acid Meth -

yl Ester (7d ). Following the procedures described for the

synthesis of 7c but using N-Boc-sarcosine in place of N-Boc-

D-alanine, 7d was prepared as a white powder. ES (+) MS m/e

) 450 (M + 1).

2S-[2-((2R)-4-Ca r ba m oyl-2-gu a n id in o-bu tyr yla m in o)-

a cetyla m in o]-3-(4-p h en yleth yn yl-p h en yl)-p r op ion ic Acid

Meth yl Ester Tr iflu or oa ceta te (7m ). Following the proce-

dures described for the synthesis of 7c but using N-Boc-^D-

glutamine in place of N-Boc-^D-alanine, 7m was prepared as a

white powder. ES (+) MS m/e ) 507 (M + 1).

(2S)-{2S-[(1-Ca r b a m im id oyl-p yr r olid in e-2-ca r bon yl)-

a m in o]-a cetyla m in o}-3-(4-p h en yleth yn yl-p h en yl)-p r op i-

on ic Acid Meth yl Ester Tr iflu or oa ceta te (7e). Following

the procedures described for the synthesis of 7c but using

N-Boc-proline in place of N-Boc-^D-alanine, 7e was prepared

as a white powder. ES (+) MS m/e ) 476 (M + 1).

(2S)-{2S-[(1-Ca r b a m im id oyl-p yr r olid in e-2-ca r b on yl)-

a m in o]-a cetyla m in o}-3-(4-p h en yleth yn yl-p h en yl)-p r op i-

on ic Acid Meth yl Ester Tr iflu or oa ceta te (7f). Following

(R)-(2-ter t-Bu toxyca r bon yla m in o-2-cycloh exyl-a cety-

la m in o)-a cetic Acid P en ta flu or op h en yl Ester (8). To

N-Boc-^D-cyclohexylglycine (10.0 g, 38.9 mmol) in dichlo-

romethane (100 mL) were added EDC (8.2 g, 42.8 mmol), HOBt

monohydrate (6.5 g, 42.8 mmol) and triethylamine (11.9 mL,

85.5 mmol). Glycine methyl ester hydrochloride (5.8 g, 46.6

mmol) was added, and the reaction mixture was stirred

overnight at room temperature. The mixture was diluted with

dichloromethane (60 mL) and partitioned with water (70 mL).

IL-2 Inhibitors Small-Molecule Antagonists of IL-2

J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12 3123

The organic layer was separated, and the aqueous layer was

extracted with dichloromethane (2 × 60 mL). The combined

organic layer was washed with 1 N HCl (60 mL) and saturated

NaHCO₃(60 mL) and dried over Na₂SO₄. The solvent was

evaporated, and the residue was dissolved in THF/water (3:1,

150 mL). To the resulting solution was then added lithium

hydroxide (2.0 g, 83.5 mmol). The reaction mixture was stirred

at room-temperature overnight and then acidified using 1 N

HCl (100 mL). The mixture was extracted with ethyl acetate

(3 × 100 mL), and the combined organic extracts were dried

over Na₂SO₄and concentrated in vacuo. The residue was

dissolved in THF (120 mL), and to the resulting solution was

added pyridine (2.8 mL, 34.0 mmol) followed by pentafluo-

rophenyl trifluoroacetate (5.8 mL, 34.0 mmol). The reaction

mixture was stirred at room temperature for 2 h, and then

the solvent was removed under reduced pressure. The residue

was dissolved in ethyl acetate (100 mL) and was washed with

1 N HCl (70 mL) and saturated NaHCO₃(70 mL). The organic

layer was dried over Na₂SO₄and concentrated in vacuo to

afford 13.5 g (72%) of 8. ¹H NMR (400 MHz, CDCl₃) 6.67 (br s,

1 H), 5.00 (br s, 1 H), 4.41 (m, 2 H), 3.98 (m, 1 H), 1.73-1.64

(m, 6 H), 1.43 (s, 9 H), 1.28-0.98 (m, 5 H). ES (+) MS m/e )

425 (M - 55).

(R)-2-Cycloh exyl-N-(2-{4-[5-(2,4-d ich lor o-p h en yl)-2H -

p yr a zol-3-yl]-p ip er id in -1-yl}-2-oxo-et h yl)-2-gu a n id in o-

a ceta m id e (15a ). Gen er a l Libr a r y P r otocol. To 4-[5-(2,4-

dichloro-phenyl)-2H-pyrazol-3-yl]-piperidine (44 mg, 0.15 mmol)

in CH₂Cl₂(1 mL) was added ester 8 (96 mg, 0.20 mmol). The

reaction was stirred at room temperature for 3 h, and then

the solvent was removed under reduced pressure to yield the

desired amide, which was used without purification.

The resulting amide was dissolved in HCl (1 mL, 4 M in

1,4-dioxane) and stirred at room temperature for 30 min. The

solvent was removed under reduced pressure to provide 80 mg

of the amine hydrochloride salt.

starting material (∼1 h). The reaction mixture was cooled to

ambient temperature, and the resulting precipitate (11.5 g)

was collected by filtration. The filtrate was concentrated,

redissolved in hot ethanol, and allowed to reach room tem-

perature. After standing for several hours, a second crop (4.3

g) was collected and when combined provided 15.8 g (54% from

9) of an off-white solid. ¹H NMR (400 MHz, CD₃OD) δ 4.11

(m, 2 H), 3.30 (m, 1 H), 2.86 (br s, 2 H), 2.73 (m, 2 H), 1.88 (m,

2 H), 1.53 (m, 2 H), 1.46 (s, 9 H); TLC (SiO₂: 10% methanol in

dichloromethane): R_f) 0.27; ES (+) MS m/e ) 268 (M + 1).

4-(1-ter t-Bu toxyca r bon yl-5-tr iflu or om eth a n esu lfon y-

loxy-1H-p yr a zol-3-yl)-p ip er id in e-1-ca r boxylic a cid ter t-

bu tyl Ester (12). (Regiochemistry of Boc-addition not deter-

mined.) A solution of 11 (11.5 g, 43.0 mmol) and N-phenyltri-

fluoromethanesulfonimide (19.9 g, 55.9 mmol) in anhydrous

dichloromethane (150 mL) under nitrogen at 0 °C was treated

with triethylamine (60 mL, 430 mmol). The mixture was

allowed to reach ambient temperature. After 1 h, LC-MS

indicated complete consumption of 11, and the solvent was

removed under reduced pressure. Purification of the crude

residue by flash column chromatography (SiO₂: gradient 0 to

40% ethyl acetate in hexane) yielded 14.8 g (86%) of a white

1

solid. H NMR (400 MHz, CDCl₃) δ 5.95 (s, 1 H), 4.13 (br s, 2

H), 2.82 (m, 3 H), 1.94 (m, 2 H), 1.61 (m, 2 H), 1.46 (s, 9 H);

TLC (SiO₂: 50% ethyl acetate in hexane): R_f) 0.45; ES (+)

MS m/e ) 400 (M + 1).

To a solution of the triflate (14.8 g, 36.9 mmol) in dichlo-

romethane (35 mL) were added a mixture of di-tert-butyl

dicarbonate (12.1 g, 55.6 mmol), DMAP (0.45 g, 3.71 mmol),

and triethylamine (75 mL) in dichloromethane (70 mL). After

1 h, HPLC indicated complete conversion to the protected

material. The solvent was removed under reduced pressure,

and the crude residue was purified by flash column chroma-

tography (SiO₂: gradient 0 to 10% ethyl acetate in hexane) to

yield 16.9 g (91%) 12 as a white solid. ¹H NMR (400 MHz,

CDCl₃) δ 6.08 (s, 1 H), 4.23 (br s, 2 H), 3.47 (m, 1 H), 2.81 (m,

2 H), 2.00 (m, 2 H), 1.64 (s, 9 H), 1.49 (m, 2 H), 1.47 (s, 9 H);

TLC (SiO₂: 20% ethyl acetate in hexane): R_f) 0.36; ES (+)

MS m/e ) 522 (M + 23). HRMS (TOF): exact mass calcd for

To the amine HCl salt (80 mg, 0.15 mmol) in MeOH (1 mL)

and triethylamine (62 µL, 0.45 mmol) was added N,N′-bis-Boc-

1-guanylpyrazole (71 mg, 0.23 mmol). The reaction mixture

was stirred at room-temperature overnight. The solvent was

removed under reduced pressure to yield the bis-Boc-protected

guanidine which was used without purification.

C

₁₀H₁₃F₃N₃O₅S (M - t-Bu - Boc + H)⁺344.0528, found

344.0579. Anal. (C₁₉H₂₈F₃N₃O₇S) C, H, N.

The protected guanidine (110 mg, 0.15 mmol) was dissolved

in TFA/CH₂Cl₂(1 mL, 1:1) and stirred at room temperature

for 3 h. The solvent was removed under reduced pressure to

afford the desired product as the trifluoroacetate salt, which

was purified via reverse-phase HPLC to provide 15a as a white

powder (12 mg, 12% over four steps. ¹H NMR (400 MHz, CD₃-

OD) δ 7.64-7.54 (m, 2 H), 7.40 (m, 1 H), 6.56 (s, 1 H), 4.59 (d,

1 H, J ) 13.4 Hz), 4.32-3.97 (m, 4 H), 3.32 (m, 1 H), 3.07 (m,

1 H), 2.88 (m, 1 H), 2.13-1.62 (m, 10 H), 1.36-1.11 (m, 5 H).

Anal. (C₂₅H₃₃Cl₂N₇O₂‚1.5CF₃CO₂H) C, H, N.

4-[5-(2,3-Dich lor o-p h en yl)-2H-p yr a zol-3-yl]-p ip er id in e-

1-ca r boxylic a cid ter t-bu tyl Ester (13e). Su zu k i Cou -

p lin gs w ith 12. Gen er a l P r oced u r e A. A mixture of 12 (300

mg, 0.601 mmol), 2,3-dichlorophenylboronic acid (143 mg, 0.75

mmol), Pd(PPh₃)₄(69 mg, 0.0601 mmol), and aqueous potas-

sium carbonate (2 M, 0.90 mL, 1.80 mmol) in dioxane (2.1 mL)

was heated at 102-103 °C (oil bath temperature) in a sealed

vial. After 3 h, the mixture was cooled to room temperature,

and the phases were separated. The organic layer was dried

(Na₂SO₄), filtered through a plug of Celite, and concentrated.

The crude residue was purified by flash column chromatog-

raphy (SiO₂: 0 to 50% ethyl acetate in hexane) to yield 106

mg (45%) of 13e as a white powder. ¹H NMR (400 MHz, CDCl₃)

δ 7.51 (d, 1 H, J ) 7.8 Hz), 7.47 (dd, 1 H, J ₁) 8.0 Hz, J ₂) 1.4

Hz), 7.23 (app t, 1 H, J ) 7.9 Hz), 6.50 (s, 1 H), 4.21 (br s, 2

H), 2.86 (m, 3 H), 2.02 (d, 2 H, J ) 11.4 Hz), 1.69 (app q, 2 H,

J ) 11.4 Hz), 1.50 (s, 9 H); TLC (SiO₂: 50% ethyl acetate in

hexane): R_f) 0.27; ES (+) MS m/e ) 418 (M + 23). HRMS

(TOF): exact mass calcd for C₁₅H₁₆Cl₂N₃O₂(M - t-Bu + H)⁺

340.0620, found 340.0646.

4-(2-Met h oxyca r bon yl-a cet yl)-p ip er id in e-1-ca r b oxyl-

ic Acid ter t-Bu tyl Ester (10). To a solution of 4-acetyl-

piperidine-1-carboxylic acid tert-butyl ester⁵¹(9, 25 g, 110

mmol) in THF (435 mL) at -78 °C was added dropwise lithium

bis(trimethylsilyl)amide (1.0 M in THF, 220 mL, 220 mmol).

To the resulting mixture was added methyl cyanoformate (15.7

mL, 198 mmol). After 30 min at -78 °C, the reaction was

warmed to room temperature, diluted with diethyl ether, and

then quenched with water. The mixture was extracted with

water, 0.5 N aqueous HCl, and brine. The organic layer was

dried over Na₂SO₄and concentrated to yield 33.4 g (over

theory) of a colorless oil. A pure sample of 10 was obtained by

flash column chromatography (SiO₂: gradient 0 to 40% ethyl

4-[5-(2,3-Dich lor o-p h e n yl)-2H -p yr a zol-3-yl]-p ip e r i-

d in e Hyd r och lor id e (14e). Dep r otection of N-Boc P ip -

er id in e Der iva tives. Gen er a l P r oced u r e B. To a solution

of 13e (90 mg, 0.227 mmol) in dichloromethane (0.5 mL) was

added 4 N HCl in dioxane (4.0 mL). The mixture was stirred

at ambient temperature for 2 h and was then diluted with

diethyl ether (4.0 mL). The resulting white precipitate was

collected by filtration, washed with diethyl ether (3 × 4 mL),

and dried under high-vacuum to yield 48 mg (72%) of a white

powder. ¹H NMR (400 MHz, DMSO-d₆) δ 9.18 (br s, 1 H), 8.95

(br s, 1 H), 7.66 (m, 2 H), 7.42 (app t, 1 H, J ) 7.9 Hz), 6.55 (s,

1 H), 3.31 (d, 2 H, J ) 12.4 Hz), 3.02 (m, 3 H), 2.14 (d, 2 H, J

1

acetate in hexane). H NMR (400 MHz, CDCl₃) δ 4.10 (br s, 2

H), 3.73 (s, 3 H), 3.50 (s, 2 H), 2.77 (m, 2 H), 2.60 (m, 1 H),

1.83 (m, 2 H), 1.57 (m, 2 H), 1.44 (s, 9 H); TLC (SiO₂: 50%

ethyl acetate in hexane): R_f) 0.34; ES (+) MS m/e ) 286 (M

+ 1).

4-(5-Oxo-4,5-dih ydr o-1H-pyr azol-3-yl)-piper idin e-1-car -

boxylic Acid ter t-Bu tyl Ester (11). A mixture of the keto-

ester 10 (32.8 g) obtained in the previous step and hydrazine

hydrate (5.58 mL, 115 mmol) in ethanol (120 mL) was heated

at 70 °C until TLC indicated complete consumption of the

3124 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12

Raimundo et al.

) 12.7 Hz), 1.86 (q, 2 H, J ) 11.8 Hz); HRMS (TOF): exact

mass calcd for C₁₄H₁₆Cl₂N₃(M + H)⁺296.0721, found 296.0746.

3,4-dichlorophenylboronic acid instead of 2,3-dichlorophenyl-

boronic acid. H NMR (400 MHz, CD₃OD) δ 7.68 (d, 1 H, J )

1

8.3 Hz), 7.58-7.49 (m, 2H), 6.58 (s, 1H), 4.59 (d, 1 H, J ) 13.2

Hz), 4.35-3.96 (m, 4 H), 3.28 (t, 1 H, J ) 13.2 Hz), 3.05 (t, 1

H, J ) 11.0 Hz, 2.88 (t, 1 H, J ) 12.2 Hz), 2.13-1.67 (m, 10

H), 1.37-1.10 (m, 5 H). ES (+) MS m/e ) 536 (M + 1).

(R)-2-Cycloh exyl-2-gu a n id in o-N-{2-oxo-2-[4-(5-o-tolyl-

2H-p yr a zol-3-yl)-p ip er id in -1-yl]-eth yl}-a ceta m id e Tr if-

lu or oa ceta te (15g). The title compound was prepared from

12 according to General Procedures A-D using o-tolylboronic

acid (100 mg, 0.20 mmol) instead of 2,3-dichlorophenylboronic

acid. White powder (19.1 mg, 16% from 12). ¹H NMR (400

MHz, CD₃OD) δ 7.49 (d, 1 H, J ) 8.3 Hz), 7.38 (d, 1 H, J ) 6.8

Hz), 7.29 (m, 2 H), 7.24 (m, 1 H), 6.37 (s, 1 H), 4.58 (d, 1 H, J

) 13.3 Hz), 4.40-4.03 (m, 2 H), 4.00 (br s, 2 H), 3.25 (app t, 1

H, J ) 13.2 Hz), 3.07 (m, 1 H), 2.87 (app t, 1 H, J ) 12.4),

2.37 (s, 3 H), 2.09 (app t, 2 H, J ) 14.0 Hz), 1.90-1.68 (m, 8

H), 1.34-1.11 (m, 5 H); Anal. (C₂₆H₃₇N₇O₂‚1.5CF₃CO₂H) C, H,

N.

(R)-[Cycloh exyl-(2-{4-[5-(2,3-d ich lor o-p h en yl)-2H-p yr a -

zol-3-yl]-p ip er id in -1-yl}-2-oxo-eth ylca r ba m oyl)-m eth yl]-

ca r ba m ic Acid ter t-Bu tyl Ester . Acyla tion of 4-Ar yl

P ip er id in e Der iva tives. Gen er a l P r oced u r e C. To a

mixture of 14e (50 mg, 0.150 mmol) and triethylamine (0.5

mL) in dichloromethane (2.0 mL) was added 72 mg (0.150

mmol) of 8. The resulting solution was stirred at ambient

temperature until LC/MS indicated complete conversion to

product. The solvent was then removed under reduced pres-

sure to yield a crude residue that was used directly in the

following step. ES (+) MS m/e ) 592 (M + 1).

(R)-2-Cycloh exyl-N-(2-{4-[5-(2,3-d ich lor o-p h en yl)-2H -

p yr a zol-3-yl]-p ip er id in -1-yl}-2-oxo-et h yl)-2-gu a n id in o-

a ceta m id e Tr iflu r oa ceta te (15e). Dep r otection /Gu a n id -

in yla tion /Dep r otection Sequ en ce. Gen er a l P r oced u r e D.

The crude residue obtained in the previous step (∼0.150 mmol

of 28) was dissolved in HCl (4.0 N in dioxane, 1 mL) and was

stirred at room temperature for 1 h. The mixure was concen-

trated in vacuo, and the residual crude amine hydrochloride

salt was taken up in methanol (2.0 mL). To the resulting

solution was added triethylamine (0.5 mL) followed by N,N′-

bis-Boc-1-guanylpyrazole (62 mg, 0.20 mmol). The reaction

mixture was stirred at room-temperature overnight. The

solvent was removed under reduced pressure, and the residue

was treated with TFA/dichloromethane (1:1, 1 mL) at room

temperature. When LC/MS indicated complete deprotection,

the mixture was concentrated to yield a crude product that

was purified by RP HPLC. The fractions containing pure

compound were combined and concentrated. The residue was

lyophilized under high-vacuum to yield 24 mg (25% from 14e)

(R)-2-Cycloh exyl-2-gu a n id in o-N-{2-oxo-2-[4-(5-m -tolyl-

2H-p yr a zol-3-yl)-p ip er id in -1-yl]-eth yl}-a ceta m id e (15h ).

The title compound was prepared from 12 according to General

Procedures A-D using m-tolylboronic acid instead of 2,3-

dichlorophenylboronic acid. White powder (10 mg, 2.8% from

1

12). H NMR (400 MHz, CD₃OD) δ 7.50-7.56 (m, 3 H), 7.31

(t, 1 H, J ) 7.7 Hz), 7.18 (d, 1 H, J ) 7.3 Hz), 6.52 (s, 1 H),

4.61 (m, 1 H), 3.98-4.38 (m, 4 H), 3.31 (m, 1 H), 3.04 (m, 1

H), 2.89 (m, 1 H), 2.40 (s, 3 H), 2.09 (m, 2 H), 1.67-1.84 (m, 8

H), 1.14-1.37 (m, 5 H). Anal. (C₂₆H₃₇N₇O₂‚2.5CF₃CO₂H) C, H,

N.

(R)-2-Cycloh exyl-N-(2-{4-[5-(2,3-d im eth yl-p h en yl)-2H-

p yr a zol-3-yl]-p ip er id in -1-yl}-2-oxo-et h yl)-2-gu a n id in o-

a ceta m id e Tr iflu r oa ceta te (15i). The title compound was

prepared from 12 according to General Procedures A-D using

2,3-dimethylphenylboronic acid instead of 2,3-dichlorophenyl-

1

of 15e as a white powder. H NMR (400 MHz, CD₃OD) δ 7.56

(d, 1 H, J ) 1.5 Hz), 7.52 (d, 1 H, J ) 1.5 Hz), 7.38 (t, 1 H, J

) 8.0 Hz), 6.52 (s, 1 H), 4.57 (d, 1 H, J ) 14.15 Hz), 4.33-3.93

(m, 4 H), 3.26 (t, 1 H, J ) 13.6), 3.06 (app t, 1 H, J ) 11.8 Hz),

2.87 (app t, 1 H, J ) 12.6 Hz), 2.08 (app t, 2 H, J ) 14.8 Hz),

1

boronic acid. H NMR (400 MHz, CD₃OD) δ 7.22-7.12 (m, 3

H), 6.25 (s, 1 H), 4.59 (d, 1 H, J ) 13.2 Hz), 4.34-3.96 (m, 4

H), 3.28 (t, 1 H, J ) 13.2 Hz), 3.10-3.02 (m, 1 H), 2.89 (t, 1 H,

J ) 13.2 Hz), 2.34 (s, 3 H), 2.26 (s, 3 H), 2.14-1.67 (m, 10 H),

1.36-1.11 (m, 5 H). ES (+) MS m/e ) 494 (M + 1).

1.85-1.63 (m, 8 H), 1.36-1.09 (m, 5 H). Anal. (C₂₅H₃₃

Cl₂N₇O₂‚1.2CF₃CO₂H) C, H, N.

-

(R)-2-Cycloh exyl-2-gu a n id in o-N-{2-oxo-2-[4-(5-p h en yl-

2H -p yr a zol-3-yl)-p ip er id in -1-yl]-et h yl}-a cet a m id e Tr i-

flu or oa ceta te (15b). The title compound was prepared from

12 according to General Procedures A-D using phenylboronic

acid instead of 2,3-dichlorophenylboronic acid. White powder

(10 mg, 2.9% from 12). ¹H NMR (400 MHz, CD₃OD) δ 7.72

(m, 2 H), 7.50-7.33 (m, 3 H), 6.53 (s, 1 H), 4.59 (d, 1 H, J )

12.1 Hz), 4.34-3.99 (m, 4 H), 3.28 (m, 1 H), 3.04 (m, 1 H),

2.89 (m, 1 H), 2.06-2.12 (m, 2 H), 1.68-1.93 (m, 8 H), 1.14-

1.37 (m, 5 H). Anal. (C₂₅H₃₅N₇O₂‚2.35CF₃CO₂H) C, H, N.

(R)-N-(2-{4-[5-(2-Ch lor o-4-h ydr oxy-ph en yl)-2H-pyr azol-

3-yl]-p ip er id in -1-yl}-2-oxo-et h yl)-2-cycloh exyl-2-gu a n i-

d in o-a ceta m id e Tr iflu or oa ceta te (15j). The title compound

was prepared from 12 according to General Procedures A-D

using 2-chloro-4-hydroxyphenylboronic acid instead of 2,3-

dichlorophenylboronic acid. ES (+) MS m/e ) 517 (M + 1).

(R)-N-(2-{4-[5-(2-Ch lor o-4-m eth oxy-ph en yl)-2H-pyr azol-

3-yl]-p ip er id in -1-yl}-2-oxo-et h yl)-2-cycloh exyl-2-gu a n i-

d in o-a ceta m id e Tr iflu or oa ceta te (15k ). The title compound

was prepared from 12 according to General Procedures A-D

using 2-chloro-4-methoxyphenylboronic acid instead of 2,3-

dichlorophenylboronic acid. ES (+) MS m/e ) 531 (M + 1).

4-[5-(2,3-Dich lor op h en yl)-1-m et h yl-1H -p yr a zol-3-yl]-

p ip er id in e-1-ca r boxylic Acid ter t-Bu tyl Ester (16a ) a n d

4-[5-(2,3-Dich lor op h en yl)-2-m eth yl-2H-p yr a zol-3-yl]-p ip -

er id in e-1-ca r boxylic Acid ter t-Bu tyl Ester (16b). To a

solution of 4-acetyl-piperidine-1-carboxylic acid tert-butyl es-

ter⁵¹(9, 1.8 g, 8.0 mmol) in 1,2-dimethoxyethane (30 mL) at

-78 °C under nitrogen was added dropwise lithium bis-

(trimethylsilyl)amide (1.0 M in THF, 16 mL, 16 mmol). The

reaction was warmed to room temperature, stirred for 30 min,

and then cooled again to -78 °C. To the mixture was added

methyl 2,3-dichlorobenzoate (2.0 g, 9.5 mmol), and the result-

ing solution was refluxed overnight. The reaction mixture was

cooled to room temperature, treated with 1.0 M HCl (15 mL),

and extracted with ethyl acetate (3 × 30 mL). The combined

organic extracts were dried over Na₂SO₄and concentrated to

yield 3.6 g an amber oil. Purification of the crude material by

flash column chromatography (SiO₂: 20% ethyl acetate in

hexanes) provided 1.5 g (48%) of the diketone intermediate.

To a solution of the diketone (300 mg, 0.75 mmol) in

methanol (3 mL) was added methylhydrazine (80 µL, 1.5

mmol). The resulting mixture was stirred for 1 h at 50 °C and

then concentrated under reduced pressure. Purification of the

(R)-N-(2-{4-[5-(2-Ch lor o-p h en yl)-2H-p yr a zol-3-yl]-p ip -

er id in -1-yl}-2-oxo-eth yl)-2-cycloh exyl-2-gu a n id in o-a ceta -

m id e Tr iflu or oa ceta te (15c). The title compound was pre-

pared from 12 according to General Procedures A-D using

2-chlorophenylboronic acid instead of 2,3-dichlorophenylbo-

1

ronic acid. White powder (11 mg, 3% from 12). H NMR (400

MHz, CD₃OD) δ 7.59 (m, 1 Η), 7.51 (m, 1 H), 7.38 (m, 2 H),

6.57 (s, 1 H), 4.60 (m, 1 H), 4.39-3.95 (m, 4 H), 3.33 (m, 1 H),

3.07 (m, 1 H), 2.87 (m, 1 H), 2.11 (m, 2 H) 1.99-1.60 (m, 8 H),

1.37-1.13 (m, 5 H). HRMS (TOF) calcd for C₂₅H₃₅ClN₇O₂(M

+ H)⁺500.2541, found 500.2558. HPLC: Purity g98%.

(R)-N-(2-{4-[5-(3-Ch lor o-p h en yl)-2H-p yr a zol-3-yl]-p ip -

er id in -1-yl}-2-oxo-eth yl)-2-cycloh exyl-2-gu a n id in o-a ceta -

m id e Tr iflu or oa ceta te (15d ). The title compound was

prepared from 12 according to General Procedures A-D using

3-chlorophenylboronic acid instead of 2,3-dichlorophenylbo-

ronic acid. White powder (15 mg, 4.1% from 12). ¹H NMR (400

MHz, CD₃OD) δ 7.78-7.67 (m, 2 H), 7.50-7.33 (m, 2 H), 6.57

(s, 1 H), 4.58 (m, 1 H), 4.34-4.00 (m, 4 H), 3.32 (m, 1 H), 3.05

(m, 1 H), 2.89 (m, 1 H), 2.09-1.71 (m, 10 H), 1.36-1.11 (m, 5

H). Anal. (C₂₅H₃₄ClN₇O₂‚1.8CF₃CO₂H) C, H, N.

(R)-2-Cycloh exyl-N-(2-{4-[5-(3,4-d ich lor o-p h en yl)-2H -

p yr a zol-3-yl]-p ip er id in -1-yl}-2-oxo-et h yl)-2-gu a n id in o-

a ceta m id e Tr iflu or oa ceta te (15f). The title compound was

prepared from 12 according to General Procedures A-D using

IL-2 Inhibitors Small-Molecule Antagonists of IL-2

J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12 3125

residue by flash chromatography (SiO₂: 10 to 40% ethyl

acetate in hexanes) provided 81 mg (26%) of regioisomer 16a

and 23 mg (7%) of regioisomer 16b. Ma jor r egioisom er

concentrated under reduced pressure, and purified via flash

chromatography (SiO₂: 25% ethyl acetate in hexanes) to afford

3.1 g (67% after two steps) of the triflate as a clear oil.

(16a ): ¹H NMR (400 MHz, DMSO-d₆) δ 7.77 (dd, 1 H, J ₁

)

To the triflate (300 mg, 0.73 mmol) and 2,4-dichlorophenyl-

boronic acid (153 mg, 0.81 mmol) in 1,4-dioxane (3 mL) was

added aqueous potassium carbonate (733 µL, 1.5 mmol, 2 M)

followed by tetrakis(triphenylphosphine)palladium (30 mg,

0.026 mmol). The reaction mixture was refluxed overnight,

diluted with ethyl acetate (8 mL), and washed with 1 M HCl

(8 mL). The aqueous layer was extracted with ethyl acetate

(2 × 8 mL), and the combined organic layer was washed with

saturated NaHCO₃solution, dried over Na₂SO₄and concen-

trated in vacuo. The resulting residue was purified via flash

chromatography (SiO₂: 25% ethyl acetate in hexanes) to afford

130 mg (44%) of 19. ¹H NMR (400 MHz, CDCl₃) δ 7.49 (s, 1H),

7.40-7.34 (m, 1H), 7.31-7.20 (m, 5H), 4.32-4.20 (m, 2H),

2.89-2.65 (m, 3H), 1.91-1.83 (m, 2H), 1.72-1.58 (m, 2H), 1.48

(s, 9H).

25. The title compound was prepared according to General

Procedures B-D but using 19 instead of 13e. ¹H NMR (400

MHz, CD₃OD) δ 7.55 (br s, 1 H), 7.42-7.20 (m, 6 H), 4.64 (br

s, 1 H), 4.35-3.92 (m, 4 H), 3.24 (br s 1 H), 2.96-2.72 (m, 2

H), 1.99-1.63 (m, 10 H), 1.37-1.07 (m, 5 H). ES (+) MS m/e

) 544 (M + 1).

4-[3-(2,4-Dich lor op h en yl)-[1,2,4]oxa d ia zol-5-yl]-p ip er i-

d in e-1-ca r boxylic a cid ter t-Bu tyl Ester (21). To 2,4-dichlo-

robenzonitrile (200 mg, 1.2 mmol) in MeOH (5 mL) was added

hydroxylamine (85 mg, 1.3 mmol, 50wt % in H₂O). The reaction

mixture was heated at 50 °C for 30 min and then refluxed for

1 h. The mixture was cooled to room temperature, and the

solvent was removed under reduced pressure. The resulting

residue was then dissolved in ethyl acetate (5 mL), washed

with water (5 mL), dried over Na₂SO₄, and concentrated in

vacuo to yield 272 mg (>100%) of 20 which was used without

further purification.

To piperidine-1,4-dicarboxylic acid mono-tert-butyl ester (91

mg, 0.40 mmol) in N,N-dimethylformamide (1 mL) with

diisopropylethylamine (83 µL, 0.48 mmol) was added tetrafluo-

roformamidinium hexafluorophosphate (116 mg, 0.44 mmol).

The reaction was stirred at room temperature for 15 min, and

then a solution of amidoxime 20 (90 mg, 0.44 mmol) in THF

(1 mL) was added. The resulting mixture was stirred for 3 h

at 110 °C and then was diluted with ethyl acetate (3 mL),

washed with water (1 × 3 mL), and washed with 1 M HCl (1

× 3 mL). The organic layer was dried over Na₂SO₄and

concentrated under reduced pressure to afford 205 mg of 21

which was used without further purification. ES (+) MS m/e

) 341 (M - 56).

7.6 Hz, J ₂) 1.9 Hz), 7.50-7.43 (m, 2 H), 6.24 (s, 1 H), 3.97

(br s, 2 H), 3.57 (s, 3 H), 2.92-2.76 (m, 3 H), 1.90 (d, 2 H, J )

11.1 Hz), 1.47 (app qd, 2 H, J ₁) 12.0 Hz, J ₂) 3.3 Hz), 1.41

(s, 9 H). ES (+) MS m/e ) 410 (M + 1). Anal. (C₂₀H₂₅Cl₂N₃O₂)

C, H, N. Min or r egioisom er (16b): ¹H NMR (400 MHz,

DMSO-d₆) δ 7.69 (dd, 1 H, J ₁) 7.8 Hz, J ₂) 1.6 Hz), 7.60 (dd,

1 H, J ₁) 8.0 Hz, J ₂) 1.5 Hz), 7.38 (app t, 1 H, J ) 7.9 Hz),

6.58 (s, 1 H), 4.17 (br s, 2 H), 3.83 (s, 3 H), 2.94-2.78 (m, 3 H),

1.89 (d, 2 H, J ) 10.8 Hz), 1.43 (m, 2 H), 1.42 (s, 9 H). ES (+)

MS m/e ) 410 (M + 1). Anal. (C₂₀H₂₅Cl₂N₃O₂) C, H, N.

(R)-2-Cycloh e xyl-N-(2-{4-[5-(2,3-d ich lor o-p h e n yl)-2-

m et h yl-2H -p yr a zol-3-yl]-p ip er id in -1-yl}-2-oxo-et h yl)-2-

gu a n id in o-a ceta m id e Tr iflu or oa ceta te (22). The title com-

pound was prepared according to General Procedures B-D but

using 16a (62 mg, 0.15 mmol) instead of 13e. White powder

1

(11 mg, 10% from 16a ). H NMR (400 MHz, CD₃OD) δ 7.60-

7.50 (m, 2 H), 7.34 (t, 1 H, J ) 7.9 Hz), 6.53 (s, 1 H), 4.65 (d,

1 H, J ) 13.7 Hz), 4.35-3.89 (m, 7 H), 3.31 (m, 1 H), 3.13 (m,

1 H), 2.88 (m, 1 H), 2.07 (m, 2 H), 1.92-1.57 (m, 8 H), 1.47-

1.11 (m, 5 H). HRMS (TOF) calcd for C₂₆H₃₆Cl₂N₇O₂(M + H)⁺

548.2308, found 548.2319. HPLC: Purity g98%.

(R)-2-Cycloh e xyl-N-(2-{4-[5-(2,3-d ich lor o-p h e n yl)-1-

m et h yl-1H -p yr a zol-3-yl]-p ip er id in -1-yl}-2-oxo-et h yl)-2-

gu a n id in o-a ceta m id e Tr iflu or oa ceta te (23). The title com-

pound was prepared according to General Procedures B-D but

using 16b (62 mg, 0.15 mmol) instead of 13e. White powder

(10 mg, 10% from 16b). ¹H NMR (400 MHz, CD₃OD) δ 7.71

(d, 1 H, J ) 8.0 Hz), 7.47-7.36 (m, 2 H), 6.24 (s, 1 H), 4.58 (d,

1 H, J ) 12.0 Hz), 4.33-4.06 (m, 2 H), 3.99 (m, 2 H), 3.65 (s,

3 H), 3.25 (m, 1 H), 3.02-2.87 (m, 2 H), 2.11-1.63 (m, 10 H),

1.37-1.14 (m, 5 H). HRMS (TOF) calcd for C₂₆H₃₆Cl₂N₇O₂(M

+ H)⁺548.2308, found 548.2318. HPLC: Purity g98%.

4-[5-(2,3-Dich lor o-p h en yl)-4-m et h yl-2H -p yr a zol-3-yl]-

p ip er id in e-1-ca r boxylic Acid ter t-Bu tyl Ester (18). To a

solution of 4-propionyl-piperidine-1-carboxylic acid tert-butyl

ester 17 (0.37 g, 1.53 mmol) in DMF (5 mL) was added NaH

(60% in mineral oil, 0.12 g, 3.07 mmol)). The reaction was

stirred for 15 min followed by addition of methyl 2,3-dichlo-

robenzoate (0.47 g, 2.3 mmol). After being stirred overnight,

the reaction mixture was diluted with EtOAc and H₂O. The

layers were separated, and the aqueous layer was extracted

with EtOAc. The organic layer was dried over MgSO₄and

concentrated under reduced pressure. The crude material was

purified by flash chromatography (SiO₂: 5% EtOAc in hexanes)

to afford the diketone which was used directly in the next step.

To the diketone in EtOH (3 mL) was added hydrazine (0.5

mL). After 30 min, the reaction was concentrated under

reduced pressure and purified by flash chromatography

(SiO₂: 33% EtOAc in hexanes) to afford 18. ES (+) MS m/e )

411 (M + 1).

(R)-2-Cycloh e xyl-N-(2-{4-[5-(2,3-d ich lor o-p h e n yl)-4-

m et h yl-2H -p yr a zol-3-yl]-p ip er id in -1-yl}-2-oxo-et h yl)-2-

gu a n id in o-a ceta m id e Tr iflu or oa ceta te (24). The title com-

pound was prepared according to General Procedures B-D but

using 18 instead of 13e. ES (+) MS m/e ) 549 (M + 1).

4-[3-(2,4-Dich lor oph en yl)-ph en yl]-piper idin e-1-car box-

ylic Acid ter t-Bu tyl Ester (19). To 3-piperidin-4-yl phenol

(2.0 g, 11.3 mmol) in dichloromethane (50 mL) was added

triethylamine (1.7 mL, 12.4 mmol) followed by Boc-anhydride

(2.7 g, 12.4 mmol). The reaction mixture was stirred at room

temperature for 2 days and was then washed with 1 M HCl (1

× 35 mL). The organic layer was dried over Na₂SO₄and

concentrated in vacuo to provide 3.4 g (>100%) of crude

material that was used without further purification.

(R)-2-Cycloh exyl-N-(2-{4-[3-(2,4-dich lor o-ph en yl)-[1,2,4]-

oxa d ia zol-5-yl]-p ip er id in -1-yl}-2-oxo-eth yl)-2-gu a n id in o-

a ceta m id e Tr iflu or oa ceta te (26). The title compound was

prepared according to General Procedures B-D but using 21

1

(164 mg, 0.40 mmol) instead of 13e. H NMR (400 MHz, CD₃-

OD) δ 7.94 (d, 1 H, J ) 8.3 Hz), 7.71 (s, 1 H), 7.53 (d, 1 H, J

) 8.3 Hz), 4.65 (s, 1 H), 4.32-3.95 (m, 4 H), 3.52-3.34 (m, 2

H), 3.01 (t, 1 H, J ) 12.2 Hz), 2.25 (t, 2 H, J ) 14.7 Hz), 1.87-

1.70 (m, 8 H), 1.37-1.10 (m, 5 H). ES (+) MS m/e ) 538 (M +

1).

2,3-Dich lor o-4-t r iisop r op ylsila n yloxy-b en zoyl Ch lo-

r id e (27). To a solution of 2,3-dichlorophenol (10.0 g, 61.3

mmol) in dichloromethane (50 mL) was added bromine (4.11

mL, 79.8 mmol) dropwise. After 2 h, HPLC indicated complete

consumption of the starting material. The reaction mixture

was slowly poured into a 10% aqueous sodium thiosulfate

solution. The phases were separated, and the aqueous layer

was extracted with dichloromethane (3×). The combined

organic layers were dried over Na₂SO₄and concentrated.

Purification of the crude residue by flash column chromatog-

raphy (SiO₂: 0 to 20% ethyl acetate in hexane) yielded 7.86 g,

53%) of a white solid. Spectroscopic data is identical to that of

commercially available 4-bromo-2,3-dichloro-phenol from Chem-

Service Inc. ¹H NMR (400 MHz, CDCl₃) δ 7.44 (d, 1 H, J ) 8.9

Hz), 6.86 (d, 1 H, J ) 8.9 Hz), 5.72 (s, 1 H); TLC (SiO₂: 20%

ethyl acetate in hexane): R_f) 0.29.

To the phenol (3.1 g, 11.3 mmol) in dichloromethane (50 mL)

was added triethylamine (1.7 mL, 12.4 mmol) followed by

N-phenyltrifluoromethanesulfonimide (4.4 g, 12.4 mmol). The

reaction was stirred overnight at room temperature and was

then washed sequentially with 1 M HCl (1 × 35 mL) and 1 M

NaOH (1 × 35 mL). The organic layer was dried over Na₂SO₄,

3126 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12

Raimundo et al.

To a solution of 4-bromo-2,3-dichloro-phenol (11.7 g, 48.4

mmol) in dry THF (100 mL) was added triethylamine (7.4 mL,

53.2 mmol) followed by triisopropylsilyl chloride (11.4 mL, 53.2

mmol). The reaction was stirred for 1 h at room temperature.

Water (250 mL) and ethyl acetate (250 mL) were added and

the layers separated. The aqueous layer was washed with

additional ethyl acetate (200 mL). The organic layers were

combined, dried over MgSO₄, filtered through a short plug of

silica gel, concentrated under reduced pressure, and dried

under high vacuum at 50 °C for 48 h to yield 19.0 g (99%) of

To N-Boc-^D-leucine (6.6 g, 28.7 mmol) in dichloromethane

(120 mL) were added EDC (6.6 g, 34.4 mmol), HOBt mono-

hydrate (4.6 g, 34.4 mmol), and triethylamine (9.6 mL, 68.9

mmol). To the activated acid solution was added the amine

(5.8 g, 28.7 mmol), and the reaction was stirred at room

temperature for 4 h. The reaction mixture was partitioned with

water (100 mL) and separated. The aqueous layer was

extracted with dichloromethane (2 × 100 mL), and the

combined organic layer was washed with 1 M HCl (100 mL)

and saturated NaHCO₃(100 mL) and dried over Na₂SO₄. The

solvent was removed under reduced pressure to provide the

crude amide, and purification by flash chromatography (SiO₂:

25% ethyl acetate in hexanes) provided 6.6 g (61% over four

1

a viscous oil. H NMR (400 MHz, CDCl₃) δ 7.36 (d, 1 H, J )

8.9 Hz), 6.72 (d, 1 H, J ) 8.9 Hz), 1.32-1.27 (m, 3 H), 1.10 (d,

18 H, J ) 7.7 Hz). Anal. (C₁₅H₂₃BrCl₂OSi) C, H, N.

1

To a -78 °C solution of the silyl ether (15.4 g, 38.7 mmol)

in anhydrous THF (385 mL) under nitrogen was added

n-butyllithium (1.57 M in hexane, 24.6 mL, 38.7 mmol)

dropwise. After 10 min at -78 °C, carbon dioxide was bubbled

through the solution for approximately 10 min. The mixture

was allowed to reach room temperature and was quenched by

careful addition of water (310 mL) followed by aqueous HCl

(1.0 N, 38.7 mL, 38.7 mmol). The mixture was extracted with

diethyl ether (3×), and the combined organic layer was dried

over MgSO₄and concentrated. Purification of the crude residue

by flash column chromatography (SiO₂: 0 to 70% ethyl acetate

in hexane) yielded 6.65 g (47%) of the acid as a white solid.

¹H NMR (400 MHz, CDCl₃) δ 7.88 (d, 1 H, J ) 8.8 Hz), 6.89

(d, 1 H, J ) 8.8 Hz), 1.32-1.27 (m, 3 H), 1.15 (d, 18 H, J ) 7.4

Hz); TLC (SiO₂: 30% ethyl acetate in hexane): R_f) 0.39; ES

(+) MS m/e ) 363 (M + 1). Anal. (C₁₆H₂₄Cl₂O₃Si) C, H, N.

To a solution of the acid (6.45 g, 17.8 mmol) in anhydrous

dichloromethane (120 mL) under nitrogen was added DMF

(1.38 mL, 17.8 mmol) followed by dropwise addition of oxalyl

chloride (2.01 mL, 23.1 mmol). After 10 min, HPLC indicated

complete consumption of the starting material, and 1,2-

dichloroethane was added. The solvent was removed in vacuo,

and the excess oxalyl chloride was removed by coevaporation

with 1,2-dichloroethane. The residue was dried under high

vacuum and was then used without further purification.

4-Eth yn yl-p ip er id in e-1-ca r boxylic Acid ter t-Bu tyl Es-

ter (29). To a heterogeneous solution of 4-formyl-piperidine-

1-carboxylic acid tert-butyl ester⁵²(23.7 g, 0.11 mol) and

potassium carbonate (30.7 g, 0.22 mol) in MeOH (1.0 L) was

added dropwise dimethyl-1-diazo-2-oxopropylphosphonate⁵³

(28, 21.3 g, 0.11 mol) in MeOH (100 mL). The resulting mixture

was stirred for 3 h and concentrated under reduced pressure.

The residue was diluted with diethyl ether (500 mL) and 5%

aqueous NaHCO₃(700 mL). The layers were separated, and

the aqueous layer was washed with diethyl ether (2 × 500 mL).

The organic layer was dried over MgSO₄and concentrated

under reduced pressure to afford crude 29 (22.4 g, 96%). ¹H

NMR (CDCl₃) δ 3.67 (m, 2 H), 3.20-3.15 (m, 2 H), 2.58 (m, 1

H), 2.09 (s, 1 H), 1.77-1.75 (m, 2 H), 1.60-1.57 (m, 2 H), 1.45

(s, 9 H). ES (+) MS m/e ) 154 (M - t-Bu + 2H)⁺.

steps). H NMR (400 MHz, CDCl₃) δ 7.04 (br s, 1H), 4.91 (br

s, 1H), 4.20 (br s, 1H), 4.04 (m, 2H), 3.08 (m, 1H), 3.56 (m,

2H), 3.28 (m, 1H), 2.73 (br s, 1H), 2.14 (s, 1H), 1.80 (m, 2H),

1.68 (m, 4H), 1.48 (m, 1H), 1.44 (s, 9H), 0.94 (m, 6H). ES (+)

MS m/e ) 324 (M - 55). Anal. (C₂₀H₃₃N₃O₄) C, H, N.

(R)-[1-(2-{4-[5-(2,3-Dich lor o-4-t r iisop r op ylsila n yloxy-

p h en yl)-2-m eth yl-2H-p yr a zol-3-yl]-p ip er id in -1-yl}-2-oxo-

et h ylca r b a m oyl)-3-m et h yl-b u t yl]-ca r b a m ic Acid ter t-

Bu tyl Ester (31). A round-bottom flask containing 27 (∼17.8

mmol), CuI (0.17 g, 0.89 mmol), PdCl₂(PPh₃)₂(0.63 mg, 0.89

mmol), and 30 (6.75 g, 17.8 mmol) was flushed with nitrogen

for several minutes and then charged with a degassed solution

of triethylamine (4.97 mL, 35.6 mmol) in toluene (120 mL).

The reaction was monitored by HPLC, and the solvent was

removed when 27 had been completely consumed. The crude

residue was immediately used in the following step. ES (+)

MS m/e ) 724 (M + 1).

The crude material and methylhydrazine (9.47 mL, 178

mmol) in ethanol (120 mL) was stirred at ambient temperature

for 1.5 h. The reaction mixture was concentrated and then

redissolved in dichloromethane. The organic phase was washed

with water, and the resulting aqueous layer was extracted with

dichloromethane (2×). The combined organic layers were dried

over Na₂SO₄and concentrated. The crude residue was purified

by flash column chromatography (SiO₂: 50 to 100% ethyl

acetate in hexane) to yield 7.0 g (52%) of 31 as a single

regioisomer. ¹H NMR (400 MHz, CDCl₃) δ 7.47 (d, 1 H, J )

8.6 Hz), 7.09 (br s, 1 H), 6.84 (d, 1 H, J ) 8.6 Hz), 6.43 (s, 1

H), 4.93 (br s, 1 H), 4.73 (d, 1 H, J ) 12.1 Hz), 4.28-4.13 (m,

3 H), 3.88 (s, 3 H), 3.85 (m, 1 H), 3.17 (app t, 1 H, J ) 11.9

Hz), 2.84 (m, 1 H), 2.77 (app t, 1 H, J ) 12.9 Hz), 2.03 (m, 2

H), 1.66-1.59 (m, 4 H), 1.45 (m, 1 H), 1.43 (s, 9 H), 1.33 (m, 3

H), 1.12 (m, 18 H), 0.93 (m, 6 H); TLC (SiO₂: ethyl acetate):

R_f) 0.19; ES (+) MS m/e ) 752 (M + 1). Anal. (C₃₇H₅₉Cl₂N₅O₅-

Si) C, H, N.

(R)-[1-(2-{4-[5-(2,3-Dich lor o-4-h yd r oxy-p h en yl)-2-m eth -

yl-2H-pyr azol-3-yl]-piper idin -1-yl}-2-oxo-eth ylcar bam oyl)-

3-m eth yl-bu tyl]-ca r ba m ic Acid ter t-Bu tyl Ester (32). To

a 0 °C solution of 31 (7.0 g, 9.3 mmol) in anhydrous THF (50

mL) was added tetrabutylammonium fluoride (1.0 M in THF,

14 mL, 14 mmol) dropwise. The reaction mixture was stirred

for 30 min and then partitioned between water (200 mL) and

ethyl acetate (200 mL). The aqueous layer was washed with

ethyl acetate (3 × 200 mL), and the combined organic layer

was dried (MgSO₄) and concentrated under reduced pressure.

The crude residue was purified by flash column chromatog-

raphy (SiO₂: 0-10% methanol in dichloromethane) to yield

(R)-{1-[2-(4-Eth yn yl-piper idin -1-yl)-2-oxo-eth ylcar bam -

oyl]-3-m eth yl-bu tyl}-ca r ba m ic a cid ter t-Bu tyl Ester (30).

A solution of 24 (6.0 g, 28.7 mmol) in HCl/dioxane (4.0 N, 50

mL) was stirred at room temperature for 30 min. The solvent

was removed under reduced pressure to afford the desired

amine (4.2 g, 100%) as the hydrochloride salt.

To N-Boc-glycine (5.0 g, 28.7 mmol) in dichloromethane (120

mL) were added EDC (6.6 g, 34.4 mmol), HOBt monohydrate

(4.6 g, 34.4 mmol), and triethylamine (9.6 mL, 68.8 mmol).

The piperidine amine HCl salt (4.2 g, 28.7 mmol) was added,

and the reaction was stirred at room-temperature overnight.

The reaction mixture was partitioned with water (100 mL) and

separated. The aqueous layer was extracted with dichlo-

romethane (2 × 100 mL), and the combined organic layer was

washed with 1 M HCl (100 mL) and saturated NaHCO₃(100

mL) and dried over Na₂SO₄to provide the desired amide.

1

5.0 g (90%) of an off-white solid. H NMR (400 MHz, CD₃OD)

δ 7.35 (d, 1 H, J ) 8.5 Hz), 7.14 (d, 1 H, J ) 8.8 Hz), 6.40 (s,

1 H), 4.61 (d, 1 H, J ) 12.8 Hz), 4.16-4.05 (m, 4 H), 3.88 (s,

3 H), 3.25 (m, 1 H), 3.06 (m, 1 H), 2.83 (m, 1 H), 2.01 (m, 2 H),

1.71-1.53 (m, 6 H), 1.44 (s, 9 H), 2.83 (m, 6 H); TLC (SiO₂:

5% methanol in dichloromethane): R_f) 0.10; ES (+) MS m/e

) 596 (M + 1). HRMS (TOF) calcd for C₂₈H₄₀Cl₂N₅O₅(M +

H)⁺596.2406, found 596.2432. Anal. (C₂₈H₃₉Cl₂N₅O₅) C, H, N.

2R-Gu a n id in o-4-m eth yl-p en ta n oic Acid (2-{4-[5-(2,3-

d ich lor o-4-h yd r oxy-p h en yl)-2-m et h yl-2H -p yr a zol-3-yl]-

piper idin -1-yl}-2-oxo-eth yl)-am ide Tr iflu or oacetate (33b).

The title compound was prepared according to General Pro-

cedure D using 32. ES (+) MS m/e ) 538 (M + 1).

To the amide (7.7 g, 28.7 mmol) was added HCl/dioxane (4.0

N, 50 mL), and the reaction was stirred at room temperature

for 30 min. The solvent was removed under reduced pressure

to give the desired amine (5.8 g, 100%) as the hydrochloride

salt.

IL-2 Inhibitors Small-Molecule Antagonists of IL-2

J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12 3127

2,3-Dich lor o-4-br om oa n isole (34). A mixture of 4-bromo-

2,3-dichlorophenol (637 mg, 2.63 mmol), iodomethane (0.660

mL, 10.5 mmol), and Cs₂CO₃(1.72 g, 5.26 mmol) in DMF (8.0

mL) was stirred at 70 °C for 3 h. After being cooled to ambient

temperature, the mixture was diluted with water and ex-

tracted with dichloromethane (3×). The combined organic

phases were washed with brine, dried (MgSO₄), and concen-

trated. Purification by flash chromatography (SiO₂: 5% EtOAc

in hexanes) provided 34 as a white solid (520 mg, 77%). ¹H

NMR (400 MHz, CDCl₃) δ 7.49 (d, 1 H, J ) 9.0 Hz), 6.75 (d, 1

H, J ) 9.0 Hz), 3.90 (s, 3 H); TLC (SiO₂: 5% EtOAc in

hexanes): R_f) 0.35.

6.88 (d, 1 H, J ) 8.8 Hz), 6.39 (s, 1 H), 6.00 (m, 1 H), 5.37 (dd,

1 H, J ₁) 17.2 Hz, J ₂) 1.3 Hz), 5.25 (d, 1 H, J ) 10.4 Hz),

5.17 (s, 2 H), 5.10 (br s, 1 H), 4.78 (m, 2 H), 4.66 (d, 1 H, J )

12.0 Hz), 4.18 (br s, 1 H), 4.06 (m, 2 H), 3.90 (br s, 1 H), 3.83

(s, 3 H), 3.13 (app t, 1 H, J ) 12.0 Hz), 2.86 (m, 1 H), 2.73

(app t, 1 H, J ) 12.9 Hz), 1.99 (app t, 2 H, J ) 9.9 Hz), 1.63-

1.50 (m, 4 H), 1.46 (m, 1 H), 1.40 (s, 9 H), 0.90 (m, 6 H); TLC

(SiO₂: 5% methanol in dichloromethane): R_f) 0.25; ES (+)

MS m/e ) 770 (M + 1).

(R)-3-[4-(5-{1-[2-(2-ter t-Bu toxyca r bon yla m in o-4-m eth -

yl-p en ta n oyla m in o)-a cetyl]-p ip er id in -4-yl}-1-m eth yl-1H-

p yr a zol-3-yl)-2,3-d ich lor o-p h en oxym eth yl]-ben zoic Acid

(39). To a solution of 38 (719 mg, 0.933 mmol) and Pd(PPh₃)₄

(54.0 mg, 0.0466 mmol) in THF (4.0 mL) under nitrogen was

added morpholine (407 µL, 4.66 mmol). The resulting mixture

was stirred at ambient temperature for 1 h. The solvent was

then removed in vacuo, and the crude residue was purified by

flash column chromatography (0.5% AcOH and 0-5% metha-

nol in dichloromethane) to yield 577 mg (85%) of a tan solid.

¹H NMR (400 MHz, CDCl₃) δ 9.31 (br s, 1 H), 8.13 (s, 1 H),

8.02 (d, 1 H, J ) 7.7 Hz), 7.66 (d, 1 H, J ) 7.6 Hz), 7.54 (d, 1

H, J ) 8.7 Hz), 7.44 (app t, 1 H, J ) 7.6 Hz), 7.13 (d, 1 H, J

) 7.1 Hz), 6.89 (d, 1 H, J ) 8.9 Hz), 6.39 (s, 1 H), 5.32 (br s,

1 H), 5.16 (s, 2 H), 4.70 (d, 1 H, J ) 12.1 Hz), 4.28-4.13 (m,

3 H), 3.88 (m, 1 H), 3.85 (s, 3 H), 3.15 (app t, 1 H, J ) 12.1

Hz), 2.84 (m, 1 H), 2.78 (app t, 1 H, J ) 12.9 Hz), 1.99 (m, 2

H), 1.66-1.41 (m, 5 H), 1.40 (s, 9 H), 0.91 (m, 6 H); TLC

(SiO₂: 0.5% AcOH 5% methanol in dichloromethane): R_f)

0.18; ES (+) MS m/e ) 730 (M + 1). Anal. (C₃₆H₄₅Cl₂N₅O₇) C,

H, N.

2,3-Dich lor o-4-m eth oxy-ben zoic Acid (35). Following the

procedure described for the synthesis of 27 but using 34 (400

mg, 1.56 mmoL) in place of 4-bromo-2,3-dichloro-1-triisopro-

pylsilyloxybenzene, 35 was prepared as a white powder (280

1

mg, 81%). H NMR (400 MHz, CDCl₃) δ 8.00 (d, 1 H, J ) 8.9

Hz), 6.92 (d, 1 H, J ) 8.9 Hz), 3.99 (s, 3 H); TLC (SiO₂: 2%

AcOH and 50% ethyl acetate in hexanes): R_f) 0.21.

(R)-[1-(2-{4-[5-(2,3-Dich lor o-4-m eth oxy-ph en yl)-2-m eth -

yl-2H-pyr azol-3-yl]-piper idin -1-yl}-2-oxo-eth ylcar bam oyl)-

3-m eth yl-bu tyl]-ca r ba m ic Acid ter t-Bu tyl Ester (36).

Following the procedures described for the synthesis of 27 and

31 but using 35 (140 mg, 1.56 mmoL), 36 was prepared as a

1

tan solid. H NMR (400 MHz, CD₃OD) δ 7.52 (d, 1 H, J ) 8.7

Hz), 7.11 (d, 1 H, J ) 8.8 Hz), 6.45 (s, 1 H), 4.64 (d, 1 H, J )

12.9 Hz), 4.18-3.99 (m, 4 H), 3.95 (s, 3 H), 3.91 (s, 3 H), 3.28

(app t, 1 H, J ) 12.8 Hz), 3.10 (m, 1 H), 2.85 (app t, J ) 11.8

Hz), 2.04 (m, 2 H), 1.73 (m, 2 H), 1.59 (m, 3 H), 1.46 (s, 9 H),

0.98 (d, 3 H, J ) 7.0 Hz), 0.96 (d, 3 H, J ) 7.1 Hz). TLC (SiO₂:

5% MeOH in dichloromethane): R_f) 0.13; ES (+) MS m/e )

610 (M + 1). HRMS (TOF) calcd for C₂₄H₃₄Cl₂N₅O₃(M - Boc

+ H)⁺510.2039, found 510.2103.

(R)-3-[2,3-Dich lor o-4-(5-{1-[2-(2-gu a n id in o-4-m et h yl-

p en t a n oyla m in o)-a cet yl]-p ip er id in -4-yl}-1-m et h yl-1H -

p yr a zol-3-yl)-p h en oxym eth yl]-ben zoic Acid Tr iflu or o-

a ceta te (33g). The title compound was prepared according

to General Procedure D using 39 (120 mg, 0.164 mmol). White

(R)-2-Gu a n id in o-4-m eth yl-p en ta n oic Acid (2-{4-[5-(2,3-

d ich lor o-4-m eth oxy-p h en yl)-2-m eth yl-2H-p yr a zol-3-yl]-

piper idin -1-yl}-2-oxo-eth yl)-am ide Tr iflu or oacetate (33c).

The title compound was prepared according to General Pro-

cedure D using 36. White powder, 13.6 mg (3.2% overall from

1

powder, 30.2 mg (23% from 39). H NMR (400 MHz, CD₃OD)

δ 8.15 (s, 1 H), 7.99 (d, 1 H, J ) 7.7 Hz), 7.72 (d, 1 H, J ) 7.6

Hz), 7.52-7.47 (m, 2 H), 7.15 (d, 1 H, J ) 8.8 Hz), 6.41 (s, 1

H), 5.28 (s, 2 H), 4.62 (d, 1 H, J ) 13.0 Hz), 4.30-4.01 (m, 2

H), 3.88 (s, 3 H), 3.27 (m, 1 H), 3.07 (m, 1 H), 2.84 (m, 1 H),

1.99 (app t, 2 H, J ) 14.2 Hz), 1.76-1.68 (m, 4 H), 1.51 (m, 1

H), 0.98 (m, 6 H); Anal. (C₃₂H₃₉Cl₂N₇O₅‚CF₃CO₂H) C, H, N.

1

35). H NMR (400 MHz, CD₃OD) δ 7.51 (d, 1 H, J ) 8.7 Hz),

7.09 (d, 1 H, J ) 8.8 Hz), 6.42 (s, 1 H), 4.61 (d, 1 H, J ) 12.9

Hz), 4.30-3.97 (m, 4 H), 3.93 (s, 3 H), 3.89 (s, 3 H), 3.30 (m,

1 H), 3.07 (m, 1 H), 2.84 (m, 1 H), 2.03 (app t, 2 H, J ) 13.9

Hz), 1.76-1.68 (m 4 H), 1.51 (m,1 H), 0.99 (m, 6 H); Anal.

(C₂₅H₃₅Cl₂N₇O₃‚1.5CF₃CO₂H) C, H, N.

(R)-2-Gu a n id in o-4-m eth yl-p en ta n oic Acid (2-{4-[5-(4-

b en zyloxy-2,3-d ich lor o-p h en yl)-2-m et h yl-2H -p yr a zol-3-

yl]-p ip er id in -1-yl}-2-oxo-eth yl)-a m id e (33d ). The title com-

pound was prepared from 31 according to General Procedures

D and E using benzylbromide in place of 37. White powder (4

3-Ch lor om eth yl-ben zoic Acid Allyl Ester (37). To a

mixture of 3-(chloromethyl)benzoic acid (1.00 g, 5.85 mmol),

EDC (1.80 g, 9.40 mmol), and DMAP (75 mg, 0.61 mmol) in

dichloromethane (20 mL) was added allyl alcohol (600 µL, 8.82

mmol). After 2 h at ambient temperature, LC/MS indicated

complete conversion. The solution was concentrated and taken

up in ethyl acetate. The mixture was washed with aqueous 1

N HCl (3×), aqueous sat. NaHCO₃(1×), and brine (1×). The

organic phase was dried (MgSO₄) and concentrated to yield

796 mg (64%) of a colorless liquid. ¹H NMR (400 MHz, CDCl₃)

δ 8.06 (s, 1 H), 8.00 (d, 1 H, J ) 7.8 Hz), 7.57 (d, 1 H, J ) 7.6

Hz), 7.43 (app t, 1 H, J ) 7.7 Hz), 6.03 (m, 1 H), 5.40 (d, 1 H,

1

mg, 4.5% from 32). H NMR (400 MHz, CD₃OD) δ 7.69-7.33

(m, 6 H), 7.17 (d, 1 H, J ) 8.8 Hz), 6.45 (s, 1 H), 5.36 (s, 2 H),

4.66 (d, 1 H, J ) 13.2 Hz), 4.33-4.00 (m, 4 H), 3.91 (s, 3 H),

3.28 (m, 1 H), 3.12 (m, 1 H), 2.87 (m, 1 H), 2.06 (m, 2 H), 1.75-

1.55 (m, 5 H), 1.03-0.99 (m, 6 H). Anal. (C₃₁H₃₉Cl₂N₇O₃‚2CF₃-

CO₂H) C, H, N.

(R)-4-[4-(5-{1-[2-(2-ter t-Bu toxyca r bon yla m in o-4-m eth -

yl-p en ta n oyla m in o)-a cetyl]-p ip er id in -4-yl}-1-m eth yl-1H-

p yr a zol-3-yl)-2,3-d ich lor o-p h en oxym eth yl]-ben zoic Acid

Meth yl Ester (40). The title compound was prepared accord-

ing to General Procedure E using 32 (75 mg, 0.120 mmol),

4-bromomethyl-benzoic acid methyl ester (43 mg, 0.190 mmol),

and K₂CO₃(35 mg, 0.250 mmol) in DMF (0.5 mL). The crude

product (76 mg, 85%) was used without further purification.

ES (+) MS m/e ) 744 (M + H).

(R)-4-[2,3-Dich lor o-4-(5-{1-[2-(2-gu a n id in o-4-m et h yl-

p en t a n oyla m in o)-a cet yl]-p ip er id in -4-yl}-1-m et h yl-1H -

p yr a zol-3-yl)-p h en oxym eth yl]-ben zoic Acid Meth yl Es-

ter Tr iflu or oa ceta te (33e). The title compound was prepared

according to General Procedure D using 40 (76 mg, 0.10 mmol).

White powder, 4 mg (4.9% overall from 40). ¹H NMR (400 MHz,

CDCl₃) δ 8.07 (d, 2 H, J ) 8.3 Hz), 7.63 (d, 2 H, J ) 8.3 Hz),

7.51 (d, 1 H, J ) 8.7 Hz), 7.17 (d, 1 H, J ) 8.8 Hz), 6.46 (s, 1

H), 5.34 (s, 2 H), 4.65 (d, 1 H, J ) 13.2 Hz), 4.29-4.02 (m, 4

H), 3.93 (s, 3 H), 3.91 (s, 3 H), 3.30 (m, 1 H), 3.30-3.10 (m, 2

H), 2.88 (m, 1 H), 2.06 (m, 1 H) 1.78-1.71 (m 4 H), 1.50 (m, 1

J ) 17.2 Hz), 5.28 (d, 1 H, J ) 10.4 Hz), 4.81 (dd, 2 H, J ₁

)

5.6 Hz, J ₂) 0.8 Hz), 4.60 (s, 2 H); ES (+) MS m/e ) 211 (M +

1).

(R)-3-[4-(5-{1-[2-(2-ter t-Bu toxyca r bon yla m in o-4-m eth -

yl-p en ta n oyla m in o)-a cetyl]-p ip er id in -4-yl}-1-m eth yl-1H-

p yr a zol-3-yl)-2,3-d ich lor o-p h en oxym eth yl]-ben zoic Acid

Allyl Ester (38). Gen er a l Alk yla tion P r oced u r e E. A

mixture of 32 (700 mg, 1.17 mmol), 37 (247 mg, 1.17 mmol),

and Cs₂CO₃(762 mg, 2.34 mmol) in DMF (7.0 mL) was stirred

at 70 °C for 3 h. After being cooled to ambient temperature,

the mixture was diluted with water and extracted with

dichloromethane (3×). The combined organic phases were

washed with brine, dried (MgSO₄), and concentrated. The

crude residue was purified by flash column chromatography

(SiO₂: 0 to 5% methanol in dichloromethane) to yield 807 mg

(90%) of a white foam. ¹H NMR (400 MHz, CDCl₃) δ 8.10 (s, 1

H), 7.99 (d, 1 H, J ) 7.7 Hz), 7.65 (d, 1 H, J ) 7.5 Hz), 7.53 (d,

1 H, J ) 8.7 Hz), 7.43 (app t, 1 H, J ) 7.7 Hz), 7.14 (br s, 1 H),

3128 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12

Raimundo et al.

H), 1.01 (m, 6 H). HRMS (TOF) calcd for C₃₃H₄₂Cl₂N₇O₅(M +

H)⁺686.2624, found 686.2614. HPLC: Purity g98%.

6.44 (s, 1 H), 5.24 (s, 2 H), 4.55-4.50 (m, 1 H), 4.16-4.05 (m,

4 H), 3.89 (s, 3 H), 3.87 (s, 3 H), 3.29 (m, 1 H), 3.12 (m, 1 H),

2.88 (m, 1 H), 2.02-2.01 (m, 2 H), 1.72-1.69 (m, 2 H), 1.60-

1.53 (m, 3 H), 1.44 (s, 9 H), 0.96-0.89 (m, 6 H). ES (+) MS

m/e ) 734 (M + 1).

To the methyl ester (814 mg, 1.10 mmol) in THF (30 mL)

was added 1.0 M LiOH (1.10 mL, 1.10 mmol). The resulting

mixture was stirred at room-temperature overnight and was

then acidified with 1 N HCl to pH 3-4. The aqueous layer

was extracted with ethyl acetate (2 × 50 mL). The combined

organic phases were washed with brine, dried (MgSO₄), and

concentrated under reduced pressure to yield 0.764 g (96%) of

43 as a colorless oil. ES (+) MS m/e ) 720 (M + 1).

(R)-4-[4-(5-{1-[2-(2-ter t-Bu toxyca r bon yla m in o-4-m eth -

yl-p en ta n oyla m in o)-a cetyl]-p ip er id in -4-yl}-1-m eth yl-1H-

p yr a zol-3-yl)-2,3-d ich lor o-p h en oxym eth yl]-ben zoic Acid

(41). To a solution of 40 (75 mg, 0.12 mmol) in THF (0.5 mL)

was added aqueous LiOH (1 M, 2.0 mL, 2.0 mmol). The

resulting mixture was heated at 60 °C for 18 h. The reaction

mixture was cooled, concentrated under reduced pressure, and

purified by RP HPLC to afford 36 mg (39%) of a film. ES (+)

MS m/e ) 730 (M + 1).

(R)-4-[2,3-Dich lor o-4-(5-{1-[2-(2-gu a n id in o-4-m et h yl-

p en t a n oyla m in o)-a cet yl]-p ip er id in -4-yl}-1-m et h yl-1H -

p yr a zol-3-yl)-p h en oxym eth yl]-ben zoic Acid Tr iflu or o-

a ceta te (33f). The title compound was prepared according to

General Procedure D using 41 (36 mg, 0.05 mmol). White

powder, 4 mg (12% from 41). ¹H NMR (400 MHz, CDCl₃) δ

8.07 (d, 2 H, J ) 8.2 Hz), 7.62 (d, 2 H, J ) 8.1 Hz), 7.51 (d, 1

H, J ) 8.7 Hz), 7.18 (d, 1 H, J ) 8.9 Hz), 6.46 (s, 1 H), 5.35 (s,

1 H), 4.21-4.00 (m, 4 H), 3.92 (s, 3 H), 3.15 (m, 1 H), 2.88 (m,

1 H), 2.07 (m, 2 H), 1.75-1.55 (m, 6 H), 1.03-0.99 (m, 6 H).

HRMS (TOF) calcd for C₃₂H₄₀Cl₂N₇O₅(M + H)⁺672.2468,

found 672.2441. HPLC: Purity g98%.

{4-[2,3-Dich lor o-4-(5-{1-[2R-(2-gu an idin o-4-m eth yl-pen -

ta n oyla m in o)-a cetyl]-p ip er id in -4-yl}-1-m eth yl-1H-p yr a -

zol-3-yl)-p h en oxym eth yl]-p h en yl}-a cetic Acid Tr iflu o-

r oa ceta te (33h ). To 32 (60 mg, 0.10 mmol) in dimethylform-

amide (500 µL) were added methyl (4-bromomethyl)phenylac-

etate (29 mg, 0.12 mmol) and potassium carbonate (28 mg,

0.20 mmol). The reaction mixture was stirred for 3 h at 65 °C

and then diluted with ethyl acetate and washed with water.

The aqueous layer was extracted with ethyl acetate (3 × 2 mL),

and the combined organic layer was dried over Na₂SO₄and

concentrated under reduced pressure to obtain the ether.

The ether was dissolved in a solution of tetrahydrofuran and

water (500 µL, 3:1) and lithium hydroxide (23 mg, 1 mmol)

was added. The reaction mixture was stirred at room temper-

ature for 3 h and then was acidified using aqueous hydrochloric

acid (3 mL, 1.0 M). The mixture was extracted with ethyl

acetate (3 × 4 mL). The combined organic layer was dried over

Na₂SO₄and concentrated under reduced pressure to afford the

acid.

(R)-5-[2,3-Dich lor o-4-(5-{1-[2-(2-gu a n id in o-4-m et h yl-

p en t a n oyla m in o)-a cet yl]-p ip er id in -4-yl}-1-m et h yl-1H -

p yr a zol-3-yl)-p h en oxym eth yl]-fu r a n -2-ca r boxylic Acid

Tr iflu or oa ceta te (33i). The title compound was prepared

according to General Procedure D using 33. White powder, 400

mg (41% overall for five steps from 51). The regiochemical

assignment was verified by NOE spectrocscopy. (The positions

of the N-methyl substituent and of the O-benzyl derivative

were confirmed by NOE difference spectroscopy (400 MHz,

CD₃OD). Selected enhancements.) [R]_D) 13.1° (c ) 1.2,

MeOH). ¹H NMR (400 MHz, DMSO-d₆) δ 8.10 (s, 1 H), 7.85

(d, 1 H, J ) 10.6 Hz), 7.66 (d, 1 H, J ) 8.8 Hz), 7.38 (d, 1 H,

J ) 9.0 Hz), 7.22 (d, 1 H, J ) 3.6 Hz), 6.81 (d, 1 H, J ) 3.5

Hz), 6.49 (s, 1 H), 5.32 (s, 2 H), 4.48 (d, 1 H, J ) 10.8 Hz), 4.25

(br s, 1 H), 4.06 (m, 2 H), 3.86 (m, 1 H), 3.81 (s, 3 H), 3.17 (app

t, 1 H, J ) 14.7 Hz), 3.09 (app t, 1 H, J ) 11.4 Hz), 2.75 (app

t, 1 H, J ) 12.4 Hz), 1.94 (d, 2 H, J ) 10.8 Hz), 1.58-1.41 (m,

1

4 H), 1.37 (m, 1 H), 0.90 (m, 6 H). H NMR (400 MHz, CD₃-

OD) (No NOEs between the dichlorophenol moiety and the

pyrazole N-Me substituent were observed. Enhancements

between the pyrazol/piperidine fragment and the aryloxy-

methyl-furan portion were also absent.) δ 7.55 (d, 1 H, J )

8.7 Hz, H-5), 7.26 (d, 1 H, J ) 8.8 Hz, H-4), 7.21 (d, 1 H, J )

3.4 Hz, H-1), 6.71 (d, 1 H, J ) 3.5 Hz, H-2), 6.46 (s, 1 H, H-6),

5.26 (s, 2 H, H-3), 4.79 (d, 1 H, J ) 13.2 Hz, H-9ax′), 4.34-

4.02 (m, 4 H, H-9ax′′, COCH₂NHCO, and H-10), 3.92 (s, 3 H,

NCH₃), 3.29 (m, 1 H, H-9eq′), 3.12 (m, 1 H, H-7), 2.88 (m, 1 H,

H-9eq′′), 2.06 (app t, 2 H, J ) 11.2 Hz, H-8eq), 1.75-1.60 (m,

5 H, 2 H-8ax, (Me)₂CHCH₂), 1.01 (m, 6 H, CH(CH₃)₂). Anal.

(C₃₀H₃₇Cl₂N₇O₆‚CF₃CO₂H) C, H, N.

The title compound was prepared according to General

Procedure D. ES (+) MS m/e ) 688 (M + 1).

5-Hyd r oxym eth yl-fu r a n -2-ca r boxylic Acid Meth yl Es-

ter (42). To a solution of 5-formyl-2-furancarboxylic acid

(commercially available from TCI America) (0.28 g, 2.0 mmol)

in benzene/methanol (5:1, 5 mL) at room temperature was

added (trimethylsilyl)diazomethane (2.0 M in hexanes, 1.0 mL,

2.0 mmol). The resulting solution was stirred at room tem-

perature for 1 h and concentrated under reduced pressure. The

residue was dissolved in methanol (2 mL) and treated with

sodium borohydride (83 mg, 2.2 mmol). After 2 h, the reaction

mixture was partitioned between water (20 mL) and ethyl

acetate (20 mL). The aqueous layer was washed with ethyl

acetate (2 × 20 mL), and the combined organic layer was dried

over MgSO₄and concentrated under reduced pressure to afford

42 (240 mg, 77%) as a white solid. ¹H NMR (400 MHz, CDCl₃)

δ 7.13 (d, 1 H, J ) 3.4 Hz), 6.41 (d, 1 H, J ) 3.4 Hz), 4.67 (s,

2 H), 3.89 (s, 3H). ES (+) MS m/e ) 157 (M + 1).

IL-2/IL-2Rr In h ibition Assa ys. IL-2 and IL2RR were

cloned and expressed as previously described.³³Both enzyme-

linked immunosorbent assay (ELISA) and scintillation prox-

imity assay (SPA) assay were used to measure compound

inhibition of IL-2/IL-2RR. Using two assays served as an initial

check for artifacts. In general, SPA is the preferred method

for monitoring weak binders because it is less artifact-prone.

However, data are the same within experimental error for

ELISA and SPA for compounds with IC₅₀values <1 mM.

For ELISA, 15 nM biotinylated IL-2RR was immobilized in

the wells of a streptavidin (Pierce Biotechnology, Rockford, IL)-

coated Maxisorp 96-well plate (Corning Life Sciences). Three-

fold serial dilutions of compounds were prepared in DMSO,

added to a solution of 5 nM IL-2 (2% DMSO final) in

Superblock (Pierce Biotechnology, Rockford, IL) with 0.01%

TWEEN 20, and incubated with the immobilized IL-2RR. After

unbound IL-2 was removed by washing, bound IL-2 was

measured with 0.65 nM of an anti-IL-2 antibody labeled with

horseradish peroxidase (HRP, Pierce Biotechnology, Rockford,

IL) followed by addition of a colorometric substrate for HRP

(TMB, Pierce Biotechnology, Rockford, IL). The inhibition of

(R)-5-[4-(5-{1-[2-(2-ter t-Bu toxyca r bon yla m in o-4-m eth -

yl-p en ta n oyla m in o)-a cetyl]-p ip er id in -4-yl}-1-m eth yl-1H-

p yr a zol-3-yl)-2,3-d ich lor o-p h en oxym eth yl]-fu r a n -2-ca r -

boxylic Acid (43). To a solution of 32 (740 mg, 1.2 mmol), 42

(240 mg, 1.5 mmol), and triphenylphosphine (440 mg, 1.7

mmol) in CH₂Cl₂(8 mL) at room temperature was added

diethyl azodicarboxylate (0.27 mL, 1.7 mmol). The resulting

solution was stirred for 16 h at room temperature and was

then partitioned between water and ethyl acetate. The organic

layer was dried (Na₂SO₄) and concentrated. The crude residue

was purified by flash chromatography (SiO₂: 100% ethyl

acetate, then 0-5% MeOH in dichloromethane) to yield 814

1

mg (89%) of a solid. H NMR (400 MHz, CD₃OD) δ 7.51 (d, 1

H, J ) 8.7 Hz), 7.23-7.21 (m, 2 H), 6.70 (d, 1 H, J ) 3.5 Hz),

IL-2 Inhibitors Small-Molecule Antagonists of IL-2

J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12 3129

bound IL-2 was plotted as a function of compound concentra-

tion, and the inhibition constant (IC₅₀) was determined by

nonlinear regression analysis (Kaleidagraph, Synergy Soft-

ware, Reading, PA).

in mass of the protein conjugate. Active fragments were

confirmed by Tethering with 4 mM âME and 200 µM of the

individual compound.

Cell-Ba sed Assa y. Compound activity was determined by

inhibition of STAT5 phosphorylation. Cells from the mouse

T-cell line CTLL-2 (gift from DNAX, Palo Alto, CA) were

stimulated with human IL-2, which binds to mouse IL-2R and

causes phosphorylation of the transcription factor STAT5.

Cells were grown to approximately 1 × 10⁶cells/mL in RPMI/

IL-2 then grown without IL-2 overnight. Two-fold serial

dilutions of compounds were prepared in DMSO and added to

the cells in media supplemented with 0.1 ng/mL IL-2. The cell/

compound mixture was incubated for 30 min at 37 °C. The

amount of STAT5 phosphorylation was then measured by

western blot, using phospho-STAT5 mAb (Upstate Biotech-

nology, Lake Placid, NY) as a primary antibody and HRP-

conjugated Rabbit anti-Mouse antibody (Zymed, South San

Francisco, CA) as a secondary antibody. Proliferating cell

nuclear antigen (PCNA) (BD Biosciences Pharmingen, San

Diego, CA) was used as a control for cell number, blot transfer,

and the like. “Fractional inhibition” of STAT-5 phosphorylation

was measured by densitometry of the phospho-STAT5 band

and normalized by the intensity of the PCNA band. Selectivity

studies monitored STAT5 phosphorylation in response to IL-

15, using the analogous assay.

Meta bolic Sta bility a n d P h a r m a cok in etics. The meta-

bolic stability of 33i was characterized by MDS Pharma

Services (Bothell, WA). Pooled human liver microsomes (n )

15, 0.5 mg/mL; XenoTech, Lenexa, KS), an NADPH regenerat-

ing system, and 1 µM compound were incubated at 37 °C. The

percent of compound remaining was determined by LC/MS/

MS. For pharmacokinetic measurements of 33i, injections were

performed at Northview Pacific Laboratories, Inc. (Hercules,

CA), and analysis was performed in-house. Compound was

formulated at 2.5 mg/mL compound in PEG 400. Groups of

three male CD-1 mice per time point were treated via

intravenous bolus injection (lateral tail vein) at a dose of 10

mg/kg. At serial time points, from 0.5 to 8 h, animals were

sacrificed and blood was collected by cardiac puncture into

sodium-EDTA tubes. Plasma was prepared by centrifugation

and stored at -20 °C until analysis. Samples were analyzed

using a nonvalidated LC-MS/MS method, using a plasma

standard curve with appropriate quality controls for quanti-

tation. Back-calculated standards did not deviate by more than

(15% from nominal value. The average concentration (stan-

dard deviation) in µg/mL were as follows: 0.5 h, 7.1(3.1); 1 h,

3.6(1.6); 1.5 h, 1.4(0.5); 2 h, 1.1(0.3); 3 h, 0.46(0.17); 4 h, 0.43-

(0.13); 6 h, 0.16(0.06); 8 h, 0.1(0.1). Pharmacokinetic analysis

was based on mean plasma concentration-time profiles.

Noncompartmental analysis was performed using WinNonlin

4.0 (Pharsight Corp., Mountain View, CA).

For SPA, 3-fold serial dilutions of compounds were prepared

in DMSO and added to 10 nM IL-2 labeled with tritiated

propionic acid in 100 µL of Superblock. This solution was then

added to biotinylated IL-2RR bound to streptavidin-labeled

scintillant-containing beads (0.7 mg/mL beads; Amersham

Biosciences, Piscataway, NJ ). Scintillation due to IL-2 bound

to IL-2RR was read on a Trilux scintillation counter (Applied

Biosystems, Foster City, CA). IC₅₀values were determined

from curves of scintillation counts versus compound concentra-

tion using a nonlinear regression analysis.

Su r fa ce P la sm on Reson a n ce (SP R). Wild-type IL-2 was

immobilized on a CM5 chip (Biacore, Uppsala, Sweden) using

standard amine coupling. Between 4000 and 8000 RU (refrac-

tive units) of IL-2 were immobilized; as a control, protein

tyrosine phosphatase (PTP-1B) was immobilized on a separate

channel. Two-fold serial dilutions of compounds were prepared

in DMSO and diluted in phosphate-buffered saline (PBS) with

0.05% azide, 1% DMSO. These were injected over the IL-2/

PTP-1B labeled chip at a rate of 40 µL/min at 25 °C for 1 min.

The stoichiometry of binding was estimated by eq 1 where MW

is the molecular weight. The dissociation constant (K_d,SPR) for

each compound was determined by plotting the RU at the

plateau of the binding curve versus the compound concentra-

tion; data were fit by nonlinear regression (Kaleidagraph).

stoichiometry ) RU_ligand/MW_ligand* MW_protein/RU_protein(1)

Sed im en ta tion Equ ilibr iu m An a lytica l Ultr a cen tr ifu -

ga tion (AU). AU measurements employed the Beckman XLA,

outfitted with six-sector cells and an eight-cell rotor (Beckman

Coulter, Palo Alto, CA). The experiment and fitting algorithm

have been described in detail.³⁴Briefly, compound and/or

protein solutions were prepared by diluting DMSO stocks in

phosphate-buffered saline to a final DMSO concentration of

1% v/v. Appropriate concentrations were chosen based on

extinction coefficients and expected affinities. Samples were

then measured at the following three speeds: 3K RPM scans

provided extinction coefficients (via Beer’s law), 25 K RPM

equilibrium scans (taken after 20-24 h) provided sedimenta-

tion curves for data-fitting, and 50 K RPM equilibrium scans

provided an experimental measurement of the baseline (“back-

ground” absorbance at the meniscus in protein-only samples).

For each K_ddetermination, three ratios of compound/protein

were measured simultaneously, and the three data sets were

fit globally using the “Hetfitter” analysis package³⁴and a

commercially available plotting program (Igor Pro, Wavem-

etrics, Lake Oswego, OR).

¹H -¹⁵N h et er on u clea r Sin gle-Qu a n t u m Coh er en ce

(HSQC) NMR. Uniformly ¹⁵N-labeled IL-2 was prepared in

minimal media containing ¹⁵NH₄Cl and exchanged into 10 mM

NaOAc pH 4.5 (NAP5, Pierce Biotechnology, Rockford, IL).

Samples were prepared to contain 150-200 µM protein, 500-

2000 µM compound, and 2% DMSO. ¹H-¹⁵N HSQC spectra

were recorded on a 400-MHz Bruker Avance spectrometer

(Bruker Biospin, Billerica, MA) at 39 °C. Significant shifts in

the NMR spectra were observed for residues throughout the

IL-2RR binding site and in the extended loops. Diagnostic

resonances for binding at the IL-2RR site include those due to

Glu 62 (which makes hydrogen bonds with the compounds)

and Phe 42 (which undergoes large conformational changes

upon binding).^31,33

Ack n ow led gm en t. The authors would like to thank

J . Silverman, U. Hoch, T. Lac, and S. Prabhu for

pharmacokinetic analysis, L. Taylor for protein purifica-

tion, T. O’Brien for development of the STAT5 phos-

phorylation assay, M. Cancilla and S. Evarts for high-

resolution MS analysis, and S. Lam, T. Webb, and A.

Hsi for compound purification. We also thank E. M.

Gordon and J . A. Wells for insightful advice.

Su p p or tin g In for m a tion Ava ila ble: Elemental analyses

data. This material is available free of charge via the Internet

at http://pubs.acs.org.

Teth er in g. Identifying fragments that covalently modify

the Leu72Cys mutant of IL-2 has been described elsewhere.^33,36

Briefly, 10 µM L72C-IL-2 in 100 mM HEPES buffer, pH 7.4,

was combined with 4 mM âME and 2 mM total disulfide-

containing compounds (200 µM each of 10 compounds) at a

final DMSO concentration of 2%. Mixtures were allowed to

come to equilibrium (2-8 h) at room temperature and then

were analyzed by liquid chromatography/mass spectrometry

(LC/MS). Hits were identified by the corresponding increase

Refer en ces

(1) Rubin, L. A.; Nelson, D. L. The Soluble Interleukin-2 Receptor:

Biology, Function, and Clinical Application. Ann. Intern. Med.

1990, 113, 619-627.

(2) Nelson, B. H.; Willerford, D. M. Biology of the Interleukin-2

Receptor. Adv. Immunol. 1998, 70, 1-81.

(3) Waldmann, T. A.; Dubois, S.; Tagaya, Y. Contrasting Roles of

Il-2 and Il-15 in the Life and Death of Lymphocytes: Implica-

tions for Immunotherapy. Immunity 2001, 14, 105-110.

3130 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 12

Raimundo et al.

(4) Nelson, B. H. Interleukin-2 Signaling and the Maintenance of

Self-Tolerance. Curr. Dir. Autoimmun. 2002, 5, 92-112.

(5) Church, A. C. Clinical Advances in Therapies Targeting the

Interleukin-2 Receptor. Q. J . Med. 2003, 96, 91-102.

(6) Le Moine, A.; Goldman, M.; Abramowicz, D. Multiple Pathways

to Allograft Rejection. Transplantation 2002, 73, 1373-1381.

(7) Singh, V. K.; Mehrotra, S.; Agarwal, S. S. The Paradigm of Th1

and Th2 Cytokines: Its Relevance to Autoimmunity and Allergy.

Immunol. Res. 1999, 20, 147-161.

(8) Schreiber, S. L.; Crabtree, G. R. The Mechanism of Action of

Cyclosporin a and Fk506. Immunol. Today 1992, 13, 136-142.

(9) Masri, M. A. The Mosaic of Immunosuppressive Drugs. Mol.

Immunol. 2003, 39, 1073-1077.

(10) Berard, J . L.; Velez, R. L.; Freeman, R. B.; Tsunoda, S. M. A

Review of Interleukin-2 Receptor Antagonists in Solid Organ

Transplantation. Pharmacotherapy 1999, 19, 1127-1137.

(11) Waldmann, T. A.; O’Shea, J . The Use of Antibodies against the

Il-2 Receptor in Transplantation. Curr. Opin. Immunol. 1998,

10, 507-512.

(12) Toogood, P. L. Inhibition of Protein-Protein Association by Small

Molecules: Approaches and Progress. J . Med. Chem. 2002, 45,

1-16.

(13) Berg, T. Modulation of Protein-Protein Interactions with Small

Organic Molecules. Angew. Chem,. Int. Ed. 2003, 42, 2462-2481.

(14) Cochran, A. G. Antagonists of Protein-Protein Interactions.

Chem. Biol. 2000, 7, R85-94.

(15) Gadek, T. R.; Nicholas, J . B. Small Molecule Antagonists of

Proteins. Biochem. Pharmacol. 2003, 65, 1-8.

(16) Arkin, M. R.; Wells, J . A. Small-Molecule Inhibitors of Protein-

Protein Interactions: Advancing Towards the Dream. Nat. Rev.

Drug Discovery 2004, in press.

(17) Nakamura, C. E.; Abeles, R. H. Mode of Interaction of Beta-

Hydroxy-Beta-Methylglutaryl Coenzyme a Reductase with Strong

Binding Inhibitors: Compactin and Related Compounds. Bio-

chemistry 1985, 24, 1364-1376.

(18) Hajduk, P. J .; Meadows, R. P.; Fesik, S. W. Nmr-Based Screening

in Drug Discovery. Q. Rev. Biophys. 1999, 32, 211-240.

(19) Fejzo, J .; Lepre, C. A.; Peng, J . W.; Bemis, G. W.; Ajay et al.

The Shapes Strategy: An Nmr-Based Approach for Lead Gen-

eration in Drug Discovery. Chem. Biol. 1999, 6, 755-769.

(20) Boehm, H. J .; Boehringer, M.; Bur, D.; Gmuender, H.; Huber,

W.; et al. Novel Inhibitors of DNA Gyrase: 3d Structure Based

Biased Needle Screening, Hit Validation by Biophysical Meth-

ods, and 3d Guided Optimization. A Promising Alternative to

Random Screening. J . Med. Chem. 2000, 43, 2664-2674.

(21) Liu, G.; Huth, J . R.; Olejniczak, E. T.; Mendoza, R.; DeVries, P.

et al. Novel P-Arylthio Cinnamides as Antagonists of Leukocyte

Function-Associated Antigen-1/Intracellular Adhesion Mole-

cule-1 Interaction. 2. Mechanism of Inhibition and Structure-

Based Improvement of Pharmaceutical Properties. J . Med.

Chem. 2001, 44, 1202-1210.

(22) Carr, R.; J hoti, H. Structure-Based Screening of Low-Affinity

Compounds. Drug Discovery Today 2002, 7, 522-527.

(23) Erlanson, D. A.; Wells, J . A.; Braisted, A. C. Tethering: Frag-

ment-Based Drug Discovery. Annu. Rev. Biophys. Biomol. Struct.

2004, 33, in press.

(24) Hann, M. M.; Leach, A. R.; Harper, G. Molecular Complexity

and Its Impact on the Probability of Finding Leads for Drug

Discovery. J . Chem. Inf. Comput. Sci. 2001, 41, 856-864.

(25) J ahnke, W.; Florsheimer, A.; Blommers, M. J .; Paris, C. G.;

Heim, J .; et al. Second-Site NMR Screening and Linker Design.

Curr. Top Med. Chem. 2003, 3, 69-80.

(26) Nienaber, V. L.; Richardson, P. L.; Klighofer, V.; Bouska, J . J .;

Giranda, V. L.; et al. Discovering Novel Ligands for Macromol-

ecules Using X-ray Crystallographic Screening. Nat. Biotechnol.

2000, 18, 1105-1108.

(27) Maly, D. J .; Choong, I. C.; Ellman, J . A. Combinatorial Target-

Guided Ligand Assembly: Identification of Potent Subtype-

Selective C-Src Inhibitors. Proc. Natl. Acad. Sci. U.S.A. 2000,

97, 2419-2424.

(28) Vetter, D. Chemical Microarrays, Fragment Diversity, Label-

Free Imaging by Plasmon Resonancesa Chemical Genomics

Approach. J . Cell Biochem. Suppl. 2002, 39, 79-84.

(29) Hyde, J .; Braisted, A. C.; Randal, M.; Arkin, M. R. Discovery

and Characterization of Cooperative Ligand Binding in the

Adaptive Region of Interleukin-2. Biochemistry 2003, 42, 6475-

6483.

(30) Erlanson, D. A.; Braisted, A. C.; Raphael, D. R.; Randal, M.;

Stroud, R. M.; et al. Site-Directed Ligand Discovery. Proc. Natl.

Acad. Sci. U.S.A. 2000, 97, 9367-9372.

(32) Emerson, S. D.; Palermo, R.; Liu, C. M.; Tilley, J . W.; Chen, L.;

et al. NMR Characterization of Interleukin-2 in Complexes with

the Il-2ralpha Receptor Component, and with Low Molecular

Weight Compounds That Inhibit the Il-2/Il-Ralpha Interaction.

Protein Sci. 2003, 12, 811-822.

(33) Arkin, M. R.; Mike, R.; Delano, W. L.; Hyde, J .; Luong, T. N. et

al. Binding of Small Molecules to an Adaptive Protein: Protein

Interface. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 1603-1608.

(34) Arkin, M.; Lear, J . D. A New Data Analysis Method to Determine

Binding Constants of Small Molecules to Proteins Using Equi-

librium Analytical Ultracentrifugation with Absorption Optics.

Anal.Biochem. 2001, 299, 98-107.

(35) Thanos, C.; Randal, M.; Wells, J . A. Potent Small-Molecule

Binding to a Dynamic Hot Spot on Il-2. J . Am. Chem. Soc. 2003,

125, 15280-15281.

(36) Braisted, A. C.; Oslob, J . D.; Delano, W. L.; Hyde, J .; McDowell,

R. S.; et al. Discovery of a Potent Small Molecule Il-2 Inhibitor

through Fragment Assembly. J . Am. Chem. Soc. 2003, 125,

3714-3715.

(37) J ohnston, J . A.; Bacon, C. M.; Finbloom, D. S.; Rees, R. C.;

Kaplan, D.; et al. Tyrosine Phosphorylation and Activation of

Stat5, Stat3, and J anus Kinases by Interleukins 2 and 15. Proc.

Natl. Acad. Sci. U.S.A. 1995, 92, 8705-8709.

(38) McKay, D. B. Unraveling the Structure of Interleukin-2: Reply.

Science 1992, 257, 412-413.

(39) Mott, H. R.; Baines, B. S.; Hall, R. M.; Cooke, R. M.; Driscoll, P.

C.; et al. The Solution Structure of the F42a Mutant of Human

Interleukin 2. J . Mol. Biol. 1995, 247, 979-994.

(40) Hajduk, P. J .; Gomtsyan, A.; Didomenico, S.; Cowart, M.;

Bayburt, E. K.; et al. Design of Adenosine Kinase Inhibitors from

the Nmr-Based Screening of Fragments. J . Med. Chem. 2000,

43, 4781-4786.

(41) Bogan, A. A.; Thorn, K. S. Anatomy of Hot Spots in Protein

Interfaces. J . Mol. Biol. 1998, 280, 1-9.

(42) Lo Conte, L.; Chothia, C.; J anin, J . The Atomic Structure of

Protein-Protein Recognition Sites. J . Mol. Biol. 1999, 285,

2177-2198.

(43) Wells, J . A.; de Vos, A. M. Hematopoietic Receptor Complexes.

Annu. Rev. Biochem. 1996, 65, 609-634.

(44) Atwell, S.; Ultsch, M.; De Vos, A. M.; Wells, J . A. Structural

Plasticity in a Remodeled Protein-Protein Interface. Science

1997, 278, 1125-1128.

(45) DeLano, W. L.; Ultsch, M. H.; de Vos, A. M.; Wells, J . A.

Convergent Solutions to Binding at a Protein-Protein Interface.

Science 2000, 287, 1279-1283.

(46) Ma, B.; Shatsky, M.; Wolfson, H. J .; Nussinov, R. Multiple

Diverse Ligands Binding at a Single Protein Site: A Matter of

Preexisting Populations. Protein. Sci. 2002, 11, 184-197.

(47) Sundberg, E. J .; Mariuzza, R. A. Luxury Accommodations: The

Expanding Role of Structural Plasticity in Protein-Protein

Interactions. Structure Fold. Des. 2000, 8, R137-142.

(48) Freire, E. The Propagation of Binding Interactions to Remote

Sites in Proteins: Analysis of the Binding of the Monoclonal

Antibody D1.3 to Lysozyme. Proc. Natl. Acad. Sci. U.S.A. 1999,

96, 10118-10122.

(49) McGovern, S. L.; Helfand, B. T.; Feng, B.; Shoichet, B. K. A

Specific Mechanism of Nonspecific Inhibition. J . Med. Chem.

2003, 46, 4265-4272.

(50) McGovern, S. L.; Caselli, E.; Grigorieff, N.; Shoichet, B. K. A

Common Mechanism Underlying Promiscuous Inhibitors from

Virtual and High-Throughput Screening. J . Med. Chem. 2002,

45, 1712-1722.

(51) Iyobe, A.; Uchida, M.; Kamata, K.; Hotei, Y.; Kusama, H. et al.

Studies on New Platelet Aggregation Inhibitors 1. Synthesis of

7-Nitro-3,4-Dihydroquinoline-2(1H)-one Derivatives. Chem.

Pharm. Bull. (Tokyo) 2001, 49, 822-829.

(52) Klein, S. I.; Molino, B. F.; Czekaj, M.; Gardner, C. J .; Chu, V.;

et al. Design of a New Class of Orally Active Fibrinogen Receptor

Antagonists. J . Med. Chem. 1998, 41, 2492-2502.

(53) Muller, S.; Liepold, B.; Roth, G. J .; Bestmann, H. J . An Improved

One-Pot Procedure for the Synthesis of Alkynes from Aldehydes.

Synlett 1996, 521-522.

(54) Sauve, K.; Nachman, M.; Spence, C.; Bailon, P.; Campbell, E.;

et al. Localization in Human Interleukin 2 of the Binding Site

to the Alpha Chain (P55) of the Interleukin 2 Receptor. Proc.

Natl. Acad. Sci. U.S.A. 1991, 88, 4636-4640.

(55) Zurawski, S. M.; Vega, F., J r.; Doyle, E. L.; Huyghe, B.; Flaherty,

K.; et al. Definition and Spatial Location of Mouse Interleukin-2

Residues That Interact with Its Heterotrimeric Receptor. Embo

J . 1993, 12, 5113-5119.

(31) Tilley, J . W.; Chen, L.; Fry, D. C.; Emerson, S. D.; Powers, G.

D.; et al. Identification of a Small Molecule Inhibitor of the Il-

2/Il-2ralpha Receptor Interaction Which Binds to Il-2. J . Am.

Chem. Soc. 1997, 119, 7589-7590.

J M049967U

Article Doi

Integrating fragment assembly and biophysical methods in the chemical advancement of small-molecule antagonists of IL-2: An approach for inhibiting protein-protein interactions

DOI: 10.1021/jm049967u

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jm049967u

Full text of DOI:10.1021/jm049967u

Products guided by the article

R&D Labs maybe for 550377-06-5

Relevant to this article

Hot Product