Discovery of CP-690,550: A potent and selective janus kinase (JAK) inhibitor for the treatment of autoimmune diseases and organ transplant rejection

8468 J. Med. Chem. 2010, 53, 8468–8484

DOI: 10.1021/jm1004286

Discovery of CP-690,550: A Potent and Selective Janus Kinase (JAK) Inhibitor for the Treatment of

Autoimmune Diseases and Organ Transplant Rejection

Mark E. Flanagan,*^,†Todd A. Blumenkopf,^‡William H. Brissette,^§Matthew F. Brown,^†Jeffrey M. Casavant,^†Chang Shang-Poa,

Jonathan L. Doty,^{^}Eileen A. Elliott,^†Michael B. Fisher,^#Michael Hines,^†Craig Kent,^†Elizabeth M. Kudlacz,^‡

Brett M. Lillie,^†Kelly S. Magnuson,^‡Sandra P. McCurdy,^†Michael J. Munchhof,^†Bret D. Perry,^†Perry S. Sawyer,^†

Timothy J. Strelevitz,^†Chakrapani Subramanyam,^†Jianmin Sun,^†David A. Whipple,⁴and Paul S. Changelian^r

^†Groton Laboratories, Pfizer Global Research & Development, Eastern Point Road, Groton, Connecticut 06340, United States,

^‡Pfizer Global Research & Development, 50 Pequot Avenue, New London, Connecticut 06320, United States, ^§Yale University, New Haven,

Connecticut 06520, United States, Department of Statistics, University of Connecticut, 215 Glenbrook Road U-4120, Storrs, Connecticut 06269,

United States, ^{^}Hill Road, East Lyme, Connecticut 06333, United States, ^#Boehringer Ingelheim Pharmaceuticals, Inc., 900 Ridgebury Road,

P.O. Box 368, Ridgefield, Connecticut 06877, United States, ⁴Del Shankel Structural Biology Center, University of Kansas, 2034 Becker Drive,

Lawrence, Kansas 66047, United States, and ^rLycera Corporation, 46701 North Commerce Center Drive, Plymouth, Michigan 48170,

United States

Received April 7, 2010

There is a critical need for safer and more convenient treatments for organ transplant rejection and

autoimmune disorders such as rheumatoid arthritis. Janus tyrosine kinases (JAK1, JAK3) are expressed

in lymphoid cells and are involved in the signaling of multiple cytokines important for various T cell

functions. Blockade of the JAK1/JAK3-STAT pathway with a small molecule was anticipated to provide

therapeutic immunosuppression/immunomodulation. The Pfizer compound library was screened against

the catalytic domain of JAK3 resulting in the identification of a pyrrolopyrimidine-based series of

inhibitors represented by CP-352,664 (2a). Synthetic analogues of 2a were screened against the JAK

enzymes and evaluated in an IL-2 induced T cell blast proliferation assay. Select compounds were evalu-

ated in rodent efficacy models of allograft rejection and destructive inflammatory arthritis. Optimization

within this chemical series led to identification of CP-690,550 1, a potential first-in-class JAK inhibitor

for treatment of autoimmune diseases and organ transplant rejection.

Introduction

The Janus tyrosine kinases (JAKs)^ahave generated great

interest in recent years as therapeutic targets due to the unique

role that these enzymes play as gatekeepers in the signal trans-

duction process.^3-5Named after the Roman god of gates and

doors (“Janus”), these four enzymes (JAK1, JAK2, JAK3,

and Tyk2) control signaling of numerous cytokines and there-

fore play a central role in acquired and innate immunity and

hematopoiesis.⁶JAK3 expression is limited to lymphoid cells

where it associates with the common γ chain (γ_c) of the IL-2

family of receptors (Figure 1).⁷JAK1 is broadly expressed,

including expression in lymphoid cells as well as the nervous

system, where it associates with β-chains of the IL-2 family of

receptors, as well as numerous other cytokine receptors. Upon

cytokine binding, receptor assembly brings JAK1 and JAK3

in close proximity such that they undergo autophosphoryla-

tion and phosphorylation of the β-chain C-terminus. Ulti-

mately, this leads to the activation of signal transducers and

activators of transcription (STAT) proteins, a critical step in

the relay of messages to the nucleus.⁸Through this process

JAK1 and JAK3 control the signaling pathways of multiple

cytokines (IL-2, -4, -7, -9, -15, and -21) important for various

T cell functions including development, activation, and homeo-

stasis.⁹The criticality of these cytokines and the subsequent

transduction of their signals for proper T cell function are

evident from mutations of JAK3 or γ_c, both of which disrupt

cytokine signaling and lead to a virtually identical immuno-

deficient phenotype [severe combined immunodeficiency

Modulation of immune function has been a successful strategy

for treating transplant rejection and autoimmune diseases.¹

However, current therapies also possess significant side effects

resulting from the ubiquitous expression of their targets of

action.²Consequently, researchers continue to search for new

biological targets, such as cytokine signaling pathways (one

such pathway is shown in Figure 1) that may offer more

selective immunosuppressive therapies to treat these conditions.

*To whom correspondence should be addressed. Telephone: 860-441-

0205. Fax: 860-715-4693. E-mail: mark.e.flanagan@pfizer.com.

^aAbbreviations: γ_c, γ common chain; ACR, American College of

Rheumatology; ADME, absorption, distribution, metabolism, and

excretion; 9-BBN, 9-borabicyclo(3.3.1)nonane;BBO, broadbandobserve;

BID, twice daily; CL_R, renal clearance; COSY, correlation spectroscopy;

CYP, cytochrome P₄₅₀; DMF, dimethylformamide; DMSO, dimethyl

sulfoxide; Fu, fraction unbound; GM-CSF, granulocyte macrophage-

colony stimulating factor; GST, glutathione S-transferase; HSA, high-

speed analoging; HFF, human foreskin fibroblast; HHT, heterotopic

cardiac transplant; HLM, human liver microsome; HMBC, heteronuclear

multiple bond coherence; HPLC, high-performance liquid chromatogra-

phy; HRMS, high-resolution mass spectrometry; HU03, human erythro-

leukemia cells; iv, intravenous; JAKs, Janus tyrosine kinases; LC/MS/MS,

liquid chromatography/tandem mass spectrometry; LRMS, low-resolution

mass spectrometry; MMF, mycophenolate mofetil; NADPH, nicotinamide

adenine dinucleotide phosphate; NOESY, nuclear Overhauser enhance-

ment spectroscopy; po, orally; RA, rheumatoid arthritis; SAR, structure-

activity relationships; SCID, severe combined immunodeficiency; SCX,

strong cation exchange; STAT, signal transducers and activators of tran-

scription; THF, tetrahydrofuran; Vd, volume of distribution.

r

pubs.acs.org/jmc

Published on Web 11/24/2010

2010 American Chemical Society

Article

Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24 8469

cell homeostasis,¹²which it was believed could lead to un-

toward effects such as anemia. Compounds were also evaluated

in an IL-2 induced T cell blast proliferation assay, which

provides a functional measure of the ability of compounds to

inhibit JAK1 and/or JAK3 activities. Additionally, since the

lead compound (2a) had a short human liver microsome

(HLM) half-life (14 min), new analogues were routinely

screenedfor metabolicstabilityusinghuman livermicrosomes

for half-life determination.

Cellular specificity assays were also employed for com-

pounds of interest. A human foreskin fibroblast (HFF) assay

was used to determine nonspecific inhibition of cellular pro-

liferation outside of the hematopoietic compartment, and a

granulocyte macrophage-colony stimulating factor (GM-

CSF) induced JAK2 cellular proliferation assay (HU03 cell

line) was used to evaluate compounds for JAK2 functional

activity. Consequently, a ratio of HU03/IL2-blast activity for

a given analogue provided a measure of JAK2 activity relative

to JAK1/3. Finally, the most significant in vivo hurdle for an

experimental immunosuppressive drug was to prolong sur-

vival of a transplantedallogeneicorgan in apreclinicalspecies.

Therefore, select compounds were evaluated in a murine

vascularized heterotopic cardiac transplant (HTT) model¹⁴

and a neovascularized transplant model.¹⁸

At the outset of the program the goal was to identify a

prototype molecule that had a potency of e10 nM against the

JAK3 kinase and e100 nM in the IL-2 blast cellular assay,

greater than 100-fold selectivity as measured by a comparison

of enzyme or cellular data for JAK3 relative to JAK2, HLM

half-life nominally of g60 min, and improved overall physico-

chemical characteristics relative to 2a.

Figure 1. Overview of cytokine signaling through the JAK/STAT

pathway. Reprinted by permission from Macmillan Publishers Ltd.:

Results

Improved Cellular Potency and the Likely Role of JAK1.

Early structure-activity relationships (SAR) revealed that

compounds possessing cyclic, lipophilic, amino headgroups

provided the best JAK3 enzyme potency (e.g., 2a-i, Table 1).

Consequently, efforts were focused on optimization of this

group from the pyrrolopyrimidine heterocycle in order to

achieve the program objectives.

Figure 2. Structure of 1 and HTS lead 2a (CP-352,664).

(SCID)].^10-12Upto two-thirds of all hereditary casesof SCID

are believed to occur as a result of mutations in one of these

two proteins;¹³up to 14% of all hereditary cases of SCID are

estimated to be due to JAK3 mutations.¹¹

We recognized the potential utility for blockade of this

pathway in order to achieve therapeutically desirable immuno-

suppression/immunomodulation and initiated a program to

identify a selective small molecule inhibitor of JAK3. This

led to the identification of the development compound

CP-690,550 1 (Figure 2), which has demonstrated efficacy in

preclinical and clinical studies of autoimmune diseases, includ-

ing rheumatoid arthritis (RA) and chronic plaque psoriasis,

and renal allograft rejection.^14-17We describe herein the

medicinal chemistry strategy that led to the discovery of this

chemical entity.

Our initial efforts were focused on JAK3 as a target for

immunosuppression because its expression is limited to lym-

phoid cells. To identify lead matter, the Pfizer compound

library (∼400000 compounds, ca. 1996) was screened against

the JAK3 catalytic domain as a recombinant glutathione S-

transferase (GST) fusion protein, resulting in the identification

of a pyrrolo[2,3-d]pyrimidine-containing lead series represented

by 2a (CP-352,664) (210 nM vs JAK3 kinase). Follow-on

analogues of 2a were initially screened against the JAK1,

JAK2, and JAK3 enzymes. Inhibition of JAK2 was assessed

early in the screening cascade because of its essential role in

hematopoiesis, including epo receptor signaling and red blood

It was anticipated that inhibition of JAK1 could play a

role in cellular potency either additively or synergistically with

JAK3, due to the manner by which these enzymes operate

together at the IL-2 receptor. The first indication of this came

from an empirical observation relating amino headgroup

structureto IL-2blast(cell) potency. Asillustrated inTable1,

it was observed that compounds possessing an N-methyl-

cycloalkyl headgroup motif exhibited an improvement in cell

potency. This was unexpected based solely on consideration

of JAK3 potency, which remained approximately the same

as compared with earlier compounds (e.g., 2a-c). However,

the N-methyl-cycloalkyl analogues demonstrated greater potency

against JAK1 (e.g., 2e-i), suggesting that the combination of

JAK1 and JAK3 (or JAK1 alone) may be key to realizing the

improved cellular potencies observed. The physical property

(clogP) distribution across these compounds (2) and the fact

that none were crystalline when initially tested make other

explanations for these cellular potency differences, such as

solubility or plasma protein binding, less likely. The role

of JAK1 is further supported by the fact that 1 in a solu-

tion phase kinase assay was more potent against JAK1

than measured in the ELISA assay (IC₅₀=3.2, 4.1, and

1.6 for JAK1, JAK2, and JAK3, respectively).¹⁹This latter

8470 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24

Flanagan et al.

Table 1. Headgroup Structure and JAK Enzyme Potencies of Compounds 2a-2i^a

^aCell potencies were determined using an IL-2 induced T cell blast proliferation assay. Metabolic stabilities of 2a-2i were determined by incubations

with human liver microsomes (HLMs) and reported by half-life.

point will be explored in more detail in the Discussion and

Conclusions.

kinase activity (see also Supporting Information Table 1).

Although a breakthrough in combining metabolic stability

with improved cellular potency was not realized, other

important SAR features were revealed from this HSA study.

Compounds 6a-c indicated a potency advantage for methyl

substitutions at the C2⁰and/or C5⁰positions of a cyclohexane

ring; the most potent compound identified was 6c, which

exhibited a 2⁰,5⁰-dimethyl substitution pattern on the cyclo-

hexyl ring. At 20 nM potency against JAK3, even as a

mixture of at least three isomers (prepared from the racemic

ketone), this material represented an 8-fold improvement in

JAK3 inhibition compared with 2f (unsubstituted cyclo-

hexyl). Although the cellular potency for these did not reflect

this 8-fold difference, JAK1 enzyme data were not deter-

mined for these HSA compounds, which may have helped to

explain this discrepancy. As shown for the other compounds

in Figure 3, a small substituent at C2⁰(methyl or ethyl)

provided the most significant potency boost. Although a

small substituent at this position was clearly advantageous,

larger groups were not well tolerated (e.g., 6e-h). While

While the N-methyl-cycloalkyl analogues (2e-i) demon-

strated improved whole-cell activity compared with the lead

(2a), the JAK2/JAK3 selectivity did not improve. Further-

more, these analogues showed no improvement in metabolic

stability as indicated by their HLM half-lives. Therefore,

many issues remained, but further investigation was war-

ranted based on the association of this SAR with improved

cellular potency. We expanded on this SAR through the use

of high-speed analoging (HSA) as illustrated in Figure 3.

Reductive aminations with methylamine were conducted

with cyclic ketones containing various functionalities. The

resulting cyclic amines (4) were then coupled to the 4-chloro-

pyrrolopyrimidine (5), purified, and tested. In all, approxi-

mately 100 compounds (6) were targeted in this manner.

Polar features were included in this array in an effort to

improve metabolic stability and physical properties. Unfor-

tunately, the randomized placement of these polar groups in

this experiment resulted only in compounds exhibiting reduced

Article

Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24 8471

Figure 3. Expansion of structure-activity relationships using high-speed analoging around the N-methyl-cycloalkyl headgroup motif leading

to compounds 6a-h.

Table 2. HSA Follow-Up Compounds Derived from Carvone Analogue 12

SAR at the C2⁰position appeared to be restricted to smaller

groups, HSA follow-up compounds derived from the natural

product carvone (described in the next section) indicated that

the C5⁰position would accommodate larger groups (Table 2).

Optimizing Interactions through the Use of Natural Pro-

ducts. One follow-up strategy employed to capitalize on

learnings from the HSA work and potentially to simplify

the stereochemical issues associated with the 6c mixture was

to take advantage of natural products that offered access to

similarly substituted ring systems as exhibited by 6c. One

such natural product was carvone (7, Figure 4). Carvone is a

member of the terpenoid family, and in addition to possessing

the desired substitution around the cyclohexane ring, it also

offered the advantage of being commercially available in

both the R-(-) and S-(þ) stereochemical configurations.²⁰

Therefore, as illustrated in Figure 4, reduction of the carvone

double bond with L-selectride, followed by reductive amina-

tion with methylamine, generated a mixture of the amine

isomers 9. Reacting mixture 9 with 5 resulted in new mixtures

that were approximately 1:1 consisting of the all-cis and all-

equatorial products. Remarkably, the mixture derived from

S-(þ)-carvone (11) was approximately 300 times more

potent against JAK3 kinase than the corresponding mixture

derived from R-(-)-carvone (10). Separation and analysis of

8472 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24

Flanagan et al.

Figure 4. Generation of carvone-derived analogues resulting in the identification of milestone compound 12.

Table 3. Moving from the Carvone Analogues to a Piperidine-Containing Headgroup Scaffold^a

^aThe main subseries investigated were sulfonamides, ureas, amines, and amides (representative examples).

the components of 11 revealed that the all-cis compound 12

was the most potent isomer (JAK3 IC₅₀=2 nM; IL-2 blast

IC₅₀= 50 nM), therefore establishing the optimal relative

and absolute stereochemistry around the cyclohexane ring to

be the all-cis configuration derived from S-(þ)-carvone.

Importantly, 12 represented a milestone for the program,

as it was the first analogue tested that possessed both JAK3

kinase and cellular potencies consistent with program objec-

tives. With greater hydrophobicity (a higher clogP) than 6c,

it was not surprising that microsomal stability for 12 was

poor (HLM t_1/2= 14 min). Furthermore, aqueous solubility

(1.3 μg/mL) and rat bioavailability (% F ∼ 7) were less than

desirable.

Attempts to address these issues from this carvone scaffold

resulted in some potent analogues (e.g., Table 2); however, it

was ultimately determined that the lipophilicity of the core

cyclohexane ring of this series made it difficult to address

pharmacokinetic issues while maintaining other desirable

attributes. Furthermore, many of the synthetic manipulations

from the isopropenyl group of 12 resulted in a new stereo-

center, which complicated analogue generation and scale-up.

Consequently, our strategy shifted toward identifying more

polar headgroups capable of providing analogues in more

drug-like space (clogP e 3).

Improved Properties: The “Piperidine Series” and 1. To

address issues with the carvone analogues, we moved to a

piperidine headgroup scaffold as illustrated in Table 3. The

advantages of adding a heteroatom into the cyclohexane

ring were several. First, this change lowered the clogP of the

target compounds compared with similar carvone-derived

analogues (generally greater than one logP unit). Second,

incorporating a nucleophilic nitrogen at this position in the

ring (at what would be the C5⁰position of a cyclohexyl)

greatly facilitated analogue generation by allowing attachment

of groups through a variety of linkages. Finally, by preparing

compounds in this manner, it was possible to avoid the intro-

duction of the additional stereocenters generated through

modification of the carvone isopropenyl group. Although

the piperidine had two stereocenters that required being

appropriately defined and set, this was eventually accom-

plished very efficiently by classical resolution (patent WO

096909, 2002).²¹

Article

Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24 8473

Table 4. Structure-Activity Relationships for the Amide Subseries of Piperidine-Containing Compounds

Given the tight SAR observed for the carvone-derived

analogues, the fundamental question at the outset of the

piperidine approach was whether it would be possible to

introduce a heteroatom into this ring and still maintain

desirable enzyme and cellular potency. To address this question,

a variety of groups were appended to the piperidine. The

main subseries investigated were sulfonamides, ureas, amines

(reductive amination products), and amides (Table 3). Although

the SAR varied across these subseries, the important general

finding was that good JAK3 enzyme potency was achievable

from this scaffold. Sulfonamides provided good JAK3 in-

hibition and encouraging selectivity relative to JAK2, but

cell potency was found to be very difficult to achieve. Several

urea analogues exhibited very good potency (both enzyme

and cell); however, finding the right balance of other properties

proved difficult. Several analogues with alkyl group sub-

stitutions on the piperidine nitrogen provided excellent

enzyme potency but required electron-withdrawing groups

(benzylic) in order to reduce the pK_aof the piperidine nitrogen

(ionization in the headgroup typically resulted in significant

loss of activity). These lipophilic amines generally exhibited

high plasma protein binding and significant HLM metabolism.

The amide subseries was ultimately successful (Table 4).

One early trend observed with these amide-containing com-

pounds was that smaller substituents generally conferred

better enzyme potency; furthermore, the cell potency tracked

with an amalgamation of JAK3 and JAK1 enzymatic inhibi-

tions. Importantly, these compounds also followed a trend

for lower clogPs (e2) translating to improved metabolic

stability. As a consequence, the cyanoacetamide compound

8474 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24

Flanagan et al.

Figure 5. Preparation of both enantiomers of 1 from intermediates 19a and 19b, separated by chiral HPLC. Single crystal X-ray structure of

urea analogue 20, prepared from 19a, provided unambiguous confirmation of the 3R,4R stereochemical configuration for 1.

18m was prepared, as it was anticipated that the small, polar

nature of this side chain would intersect both of these SAR

patterns and impart improved potency and metabolic stabil-

ity. This is precisely what was observed, and with 18m, for the

first time, all in vitro program objectives had been met, with

the exception of selectivity for JAK3 over JAK2, which was

still measured to be only 20-fold.

dosing of 1 resulted in moderate to high plasma clearances,

which were predicted from microsome incubations (t_1/2= 68

and 30, respectively). In monkeys, plasma clearances were

low to moderate, consistent with the half-life observed in monkey

liver microsomal incubations (>100 min). Low to moderate

volumes of distribution were also observed, resulting in half-

lives of 0.6-2.1 h (Table 5). Oral dosing of the crystalline

citrate salt resulted in oral bioavailabilities that were consistent

with the expected magnitude in first-pass hepatic metabolism,

estimated from the iv clearances. Plasma protein binding

for 1 was very similar in rats, dogs, monkeys, and humans

[human fraction unbound (Fu) = 24%]. Preliminary in vitro

studies predicted minimal risk with respect to unwanted drug-

drug interactions for multiple cytochrome P₄₅₀(CYP) iso-

forms (IC₅₀s > 52 μM for CYPs 1A2, 2C19, 2D6, and 3A4).

Studies performed in Caco-2 cell monolayers predicted that 1

(at 10 μΜ) should not be a substrate for efflux transporters

and should be absorbed transcellularly by passive diffusion

with a predicted human absorption of >70%.

In contrast to the moderate to high clearances observed in

preclinical species, human hepatic metabolism of 1 was pre-

dicted to be low based on studies conducted in human liver

microsomes and hepatocytes (t_1/2>120 min). Consequently,

the low clearance of the citrate salt in humans was predicted

to occur via a combination of renal excretion of unchanged

drug and low hepatic metabolic clearance. Allometric scaling

of the renal clearance (CL_R) in animals (Table 5) yielded a

predicted human CL_Rof 1.2 mL min^-1kg^-1, which was then

incorporated with an estimate of low metabolic clearance

(e2.5 mL min^-1kg^-1). Volume of distribution predictions

yielded an average Vdss in human of 1.9 L/kg. Based on the

Vdss and CL predictions, the predicted terminal elimination

half-life in human was ∼7 h. Therefore, although 1 was meta-

bolized in preclinical species, the low metabolic turnover observed

with human liver microsomes and hepatocytes provided

confidence that clearance would occur predominantly via a

combination of renal excretion of unchanged drug and low

hepatic metabolic clearance, and pharmacokinetics sufficient

to support qd or bid dosing of 1 would be observed in

humans. This prediction has largely held true in the clinic

Separation of the enantiomers of 18m was not successful

by chiral high-performance liquid chromatography (HPLC);

however, the benzyl-protected precursor of this material (19)

was separable by this method (Figure 5). The two purified

components 19a and 19b were therefore carried forward

separately producing 1 and its enantiomer 1a. Evaluation

of these two compounds revealed that 1 had an IC₅₀of 1 nM

against JAK3, which was approximately 40 times more potent

than 1a. Eventually all four stereoisomers of 18m were pre-

pared and studied in our laboratories and other laboratories.²²

With compound 1 being the most potent isomer it was anti-

cipated that its absolute stereochemical configuration would

reflect that of the most potent carvone-derived analogue 12.

Final stereochemical confirmation for 1 came from the evalua-

tion of a related urea analogue 20 (Figure 5). Compound 20

was crystalline as the free base, and since the stereochemical

configuration of the pyrrolidinol moiety was known (derived

from (S)-3-hydroxypyrrolidine), the corresponding single

crystal X-ray structure of 20 provided unambiguous assignment

of the 3R,4R stereochemical configuration for 1. Corroboration

of this stereochemical assignment has come from published

cocrystal structures of 1 with JAK family members.^23,24

In order to proceed to development studies, a suitable solid

form of 1 needed to be identified. To do this, the free base

of 1 was screened against a panel of acids, resulting in the

identification of the crystalline citrate salt. This material

(citrate salt) was ultimately selected for development

based on its desirable characteristics of good solubility

(>4mg/mLinwater),high melting point onset (>200 ꢀC), lack

of hygroscopicity, and high degree of crystallinity (patent

WO 048162, 2003).²⁵

Preclinical Pharmacokinetics. 1 is being developed as the

crystalline citrate salt. In rats and dogs, intravenous (iv)

Article

Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24 8475

Table 5. Preclinical Pharmacokinetic (PK) Summary for 1

iv PK summary

po PK summary

species

CL (mL min^-1kg^-1

)

CL_R(mL min^-1kg^-1

)

Vdss (L/kg)

t_1/2(h)

Fu (%)

dose (mg/kg)

C_max(ng/mL)

F (%)

rat

dog

62

19

18

6.2

1.9

2.3

2.6

1.8

1.7

0.6

1.2

2.1

23

19

26

10

5

442

1020

790

27

78

48

monkey

5

(oral t_1/2= 2-5 h) with the exception that some hepatic

metabolism of 1 was in fact observed via CYPs 3A4/5 and

2C19.^26-28

center study, 1 dosed at 15 and 30 mg bid was compared with

tacrolimus in 61 patients receiving de novo kidney transplants.³³

In the latter study, immune suppression in the 1treatment groups

was supported by prevention of rejection and by evidence of

infection. The effects of JAK2 may have manifested in the 1

groups as a trend toward more frequent anemia and neutro-

penia. This study suggests that relatively low rates of acute

rejection are achievable with 1-based regimens and that JAK

inhibition provides a novel, potentially viable means to prevent

allogeneic response in allograft recipients.

1 has also demonstrated efficacy in therapeutic models of

destructive, inflammatory arthritis in rodents.³⁴Encouraged

by these data, 1 entered development as a disease-modifying

antirheumatic drug (DMARD) for the treatment of RA.

A 6-week proof-of-concept study in 264 adult patients with

active RA, who had an inadequate response to methotrexate,

etanercept, infliximab, or adalimumab, compared 1 (5, 15,

and 30 mg bid) with placebo. All three doses resulted in

significant improvement versus placebo (p<0.0001) in the

primary end point, American College of Rheumatology

(ACR)20.³⁵In this study, the most common adverse events

reported were headache and nausea. The risk of infection

over 6 weeks appeared to be slightly higher for 1 15 and 30 mg

bid than placebo or 1 5 mg bid. Dose-dependent decreases in

absolute neutrophil count and a dose response in mean

change (decrease) from baseline in hemoglobin were ob-

served. Small increases (0.004-0.06 mg/dL mean) in mean

serum creatinine compared with baseline and with placebo

and dose-dependent increases in total, high-density lipoprotein-

and low-density lipoprotein-cholesterol were observed in all

1 dose groups.³⁵

Discussion and Conclusions

Following its identification as a lead molecule, the in vitro

pharmacology of 1 has been studied extensively. The results of

these studies show 1 to be a potent inhibitor of the JAK kinase

family, with a high degree of selectivity within the human

kinome.²⁹Structural and thermodynamic characterization of

1 complexed with JAK family members provides insights into

this high degree of kinome selectivity.²⁴

Previous reports of kinase activity describe 1 as being 20-

fold selective relative to JAK2 and 112-fold selective relative

to JAK1.¹⁴More recently, in solution-phase caliper assays 1

exhibited low nanomolar “PAN” inhibitory potency against

JAK1, JAK2, and JAK3 and to a lesser extent Tyk2 (IC₅₀=

3.2, 4.1, 1.6, and 34.0 for JAK1, JAK2, JAK3, and Tyk2,

respectively).¹⁹

In cellular assays 1 exhibited 11 nM potency in the IL-2

blast proliferation assay (JAK1 and JAK3), 324 nM potency

in the HU03 cellular specificity assay (JAK2), and >10000

nM activity in the HFF assay (nonspecific effects on cellular

proliferation).¹⁴While the JAK kinase activities for 1 changed

with assay conditions, the cellular potencies measured are consis-

tent with recent results of a human whole blood assay indi-

cating that γ_ccytokine-dependent activation (driven by JAK1

and JAK3) was potently inhibited by 1, exhibiting functional

selectivity over GM-CSF-dependent (JAK2-driven) activation

of the pathway.^19,30

These data confirm that the desired 100-fold selectivity for

JAK3 relative to the other JAKs was ultimately not accom-

plished with 1; however, inhibition of JAK1 and JAK2 may

actually enhance the efficacy of 1 for certain indications, such

as RA, by impeding the signaling of proinflammatory cytokines,

such as IL-6 (IL-6 signals through JAK1/JAK2). It is there-

fore possible that a near-optimal balance of JAK kinase inhibi-

tions may have fortuitously been achieved with 1, benefiting

efficacy across multiple therapeutic indications while main-

taining clinically manageable functional selectivity for JAK2.

Clinical Development. 1 is currently in clinical develop-

ment for RA (phase III), prevention of acute (renal) allograft

rejection (phase II), psoriasis (oral and topical; phase II),

Crohn’s disease (phase II), ulcerative colitis (phase II), and

dry eye disease (topical; phase I/II).

In preclinical studies of inhibition of allograft rejection 1,

dosed as monotherapy or in combination with mycopheno-

late mofetil (MMF), demonstrated efficacy in rodents and in

nonhuman primates; prolonged survival was observed in a

dose-responsive fashion.^14,31Prevention of allograft rejection

has also been demonstrated in a rat model of aorta transplan-

tation³²and in various other murine models.¹⁸

This study was followed by two large phase IIb placebo-

controlled, dose-ranging studies, one in 509 RA patients with

an inadequate response to stable background methotrexate

(study A3921025; NCT00413660^36,37) and one in 384 patients

with an inadequate response to prior DMARDs, who had

discontinued all traditional and biologic DMARDs other

than antimalarials (study A3921035; NCT00550446^38,39).

Study A3921035 also included adalimumab dosed 40 mg

subcutaneously every other week for six injections, followed

by 1 5 mg bid for another 12 weeks.

In the analysis of the primary end points in studies A3921025

and A3921035, 1 [3-15 mg bid; 20 mg qd (A3921025 only)]

was significantly effective for the primary end point, ACR20

response at 12 weeks compared with placebo. Adverse events

in both studies were similar to those observed in other phase

II studies; in addition, a slight increase in the rate of alanine

aminotransferase increases >3 times the upper limit of

normal was observed in patients dosed with 1 15 mg bid in

study A3921025.³⁶

Phase III studies of 1 (5 and 10 mg bid) for the treatment of

RA, a phase II study (2, 5, and 15 mg bid) for psoriasis, and a

phase I/II study (topical solution) are currently ongoing.

In conclusion, the JAK1/3-STAT signaling pathway has

proven to be a highly potent and productive target for selec-

tive immunosuppression/immunomodulation. The utility of

exploiting this mechanism is evident from the preclinical and

Following the encouraging findings in preclinical transplant

models, 1 was evaluated in renal allograft recipients. The toler-

ability profile of 1 coadministered with MMF was evaluated in

a phase I dose-escalation study.²⁶In a pilot, phase IIa multi-

8476 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24

Flanagan et al.

clinical efficacy observed with 1. JAK3 was the original

target due to its expression being limited to lymphoid cells

and the immunosuppressed phenotype of patients possessing

mutations in this protein. While the targeted 100-fold selec-

tivity for JAK3 relative to JAK2 was ultimately not achieved,

1 is highly selective for the JAK family of kinases. Further-

more, in cellular assays, 1 has demonstrated functional

selectivity for JAK1/3 versus JAK2, which importantly has

proven to be manageable clinically in terms of unwanted

JAK2 effects, such as anemia, aided by the favorable pharma-

cokinetics of this compound.

To find this chemical entity, a standard high-throughput

screening protocol was initiated, which resulted in the iden-

tification of 2a. With this lead, a strategy was pursued in

which the basic medicinal chemistry principles of low molec-

ular weight and logP were observed, eventually leading to a

highly soluble, highly ligand and lipid efficient inhibitor in 1.

From the outset of the program the issues of potency (both

enzyme and cell), metabolic stability, and selectivity were

engaged simultaneously. High-speed analoging was performed

to screen chemical space quickly, resulting in new substitution

patterns associated with potency improvements. Natural

products were also used as chemical building blocks to under-

stand target binding better and to optimize interactions.

Finally, a scaffold shift to a piperidine-containing headgroup

resulted in an improvement in drug-like properties. Ultimately,

however, it was the introduction of the cyanoacetamide side

chain that brought the optimal balance of potency, chemical

properties, and pharmacokinetic attributes that has posi-

tioned 1 to become a novel, first-in-chemotype, and potential

first-in-class JAK inhibitor for the treatment of autoimmune

diseases and organ transplant rejection.

human recombinant IL-2 (R&D Systems, catalog no. 202-IL).

After 1 week the cells are IL-2 dependent and can be maintained

for up to 3 weeks by feeding twice weekly with equal volumes of

media plus 100 units/mL IL-2.

To assay for a test compound’s ability to inhibit IL-2 depen-

dent T cell proliferation, IL-2 dependent cells are washed three

times, resuspended in media, and then plated (50000 cells per

well/0.1 mL) in a flat-bottom 96-well microtiter plate (Falcon

catalog no. 353075). From a 10 mM stock of test compound in

dimethyl sulfoxide (DMSO), serial 2-fold dilutions of com-

pound are added in triplicate wells starting at 10 μM. One row

serves as 100% proliferation control and is incubated with

DMSO alone. After 1 h, 10 units/mL IL-2 is added to each test

well, and plates are then incubated at 37 ꢀC and 5% CO₂for

72 h. Plates are then pulsed with [³H]thymidine (0.5 μCi/well;

NEN catalog no. NET-027A) and incubated for an additional

18 h. Culture plates are harvested with a 96-well plate harvester,

and the amount of [³H]thymidine incorporated into proliferat-

ing cells is determined by counting on a Packard Top Count

scintillation counter. Data are analyzed by plotting the percent

inhibition of proliferation versus the concentration of test com-

pound. An IC₅₀value (micromolar) is determined from this plot.

Inhibition of Human Erythroleukemia Cells (HU03 Assay).

This screen is designed to assess the ability of compounds to

inhibit GM-CSF-driven proliferation of the human erythro-

leukemia cell line HU03. GM-CSF receptor is dependent upon the

protein tyrosine kinase JAK2 for signaling. This screen assesses

the cellular activity of compounds that inhibit this kinase. HU03

are maintained [RPMI þ 10% FCS þ P/S þ 25 ng/mL GM-CSF

(Sargramostim Leukine, 250 μg; Immunex Corp.)] at 3.5 ꢀ

10⁵-1.0ꢀ10⁶cells/mL. The cells are fed three times per week

depending upon the cell count. To assay for a test compound’s

ability to inhibit GM-CSF-dependent HU03 cell proliferation,

cells are washed three times, resuspended in media, and then

plated (2500 cells per well/0.1 mL) in a flat-bottom 96-well

microtiter plate (Falcon catalog no. 353075). From a 10 mM

stock of test compound in DMSO, serial 2-fold dilutions of

compound are added in triplicate wells starting at 10 μM. One

row serves as 100% proliferation control and is incubated with

DMSO alone. After 1 h, 50 μL of 10 units/mL GM-CSF (final)

in media is added to all wells. Plates are then incubated at 37 ꢀC

and 5% CO₂for 72 h. Cells are pulsed with [³H]thymidine and

harvested, and percent inhibition is determined in the same

fashion as the IL-2 dependent T cell proliferation assay.

Inhibition of Human Foreskin Fibroblast Proliferation (HFF

Assay). This screen measures nonspecific inhibition of cellular

proliferation. HFFs are obtained from explants of neonatal

human foreskin tissue and cultured in media (DMEM with 10%

heat-inactivated fetal bovine serum and 1% P/S). Primary cultures

are grown in media in tissue culture flasks and split prior to con-

fluency using trypsin-EDTA (Gibco). Fibroblasts are grown in

T-175 tissue culture flasks and split 1:5 every 3-4 days. HFFs

are used in this assay from passage 5 to 16. To assay for a test

compound’s ability to inhibit HFF proliferation, cells are trypsin-

ized and washed three times, resuspended in media, and then

plated (2500 cells per well/0.15 mL) in a flat-bottom 96-well

tissue culture plate (Falcon catalog no. 353075). From a 10 mM

stock of test compound in DMSO, serial 2-fold dilutions of com-

pound are added in triplicate wells starting at 10 μM. One row is

plated with DMSO only and serves as the 100% proliferation

control. Cells are pulsed with [³H]thymidine and harvested, and

percent inhibition is determined in the same fashion as the IL-2

dependent T cell proliferation assay.

Experimental Section

Enzyme Assays. The JAK1, JAK2, and JAK3 kinase assays

utilize a protein expressed in baculovirus-infected SF9 cells (a

fusion protein of GST and the catalytic domain of human JAK

enzyme) purified by affinity chromatography on glutathione-

Sepharose. The substrate for the reaction was polyglutamic

acid-tyrosine [PGT (4:1), Sigma catalog no. P0275], coated onto

Nunc Maxi Sorp plates at 100 μg/mL overnight at 37 ꢀC. The

plates were washed three times, and JAK enzyme was added to

the wells, which contained 100 μL of kinase buffer (50 mM

HEPES, pH 7.3, 125 mM NaCl, 24 mM MgCl₂) þ ATP þ 1 mM

sodium orthovanadate). After incubation at room tempera-

ture for 30 min, the plates were washed three times. The level

of phosphorylated tyrosine in a given well was determined

by standard ELISA assay utilizing an anti-phosphotyrosine

antibody.

Cell Assays. Inhibition of Human IL-2 Dependent T Cell Blast

Proliferation (IL-2 Blast Assay). This screen measures the

inhibitory effect of compounds on IL-2 dependent T cell blast

proliferation in vitro. Since signaling through the IL-2 receptor

requires JAK3, cell active inhibitors of JAK3 should inhibit IL-2

dependent T cell blast proliferation. The cells for this assay are

isolated from fresh human blood. After separation of the mono-

nuclear cells using Accuspin System-Histopaque-1077 (Sigma

catalog no. A7054), primary human T cells are isolated by negative

selection using Lympho-Kwik T (One Lambda, Inc., catalog no.

LK-50T). T cells are cultured at (1-2) ꢀ 10⁶/mL in media

[RPMI þ 10% heat-inactivated fetal calf serum (Hyclone catalog

no. A-1111-L) þ 1% penicillin/streptomycin (P/S; Gibco)] and

induced to proliferate by the addition of 10 μg/mL PHA (Murex

Diagnostics, catalog no. HA 16). After 3 days at 37 ꢀC in 5%

CO₂, cells are washed three times in media and resuspended to a

density of (1-2) ꢀ 10⁶cells/mL in media plus 100 units/mL

In Vitro Pharmacokinetic Assays. In vitro absorption, distri-

bution, metabolism, and excretion (ADME) and in vivo studies

were carried out in accordance with previously published methods.⁴⁰

Microsome Incubations. Incubations of 1 (1 μM) were carried

out in replicate with microsomes (CYP concentration =

0.25 μM) in 0.1 M potassium phosphate buffer (pH = 7.4) at

37 ꢀC. The reaction mixture was prewarmed at 37 ꢀC for 2 min

Article

Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24 8477

before adding nicotinamide adenine dinucleotide phosphate

(NADPH) (1.2 mM). The final incubation volume was 600 mL.

Aliquots (75 mL) of the reaction mixture at t = 0, 2, 5, 15, and

30 min were added to acetonitrile (200 mL) containing indo-

methacin as internal standard (50 ng/mL), and the samples were

centrifuged at 2500g for 5 min before liquid chromatography/

tandem mass spectrometry (LC/MS/MS) analysis. For control

experiments, NADPH was omitted from these incubations.

Hepatocyte Incubations. Stability of 1 in hepatocytes was

examined as follows. Cryopreserved hepatocytes from preclini-

cal species and human were thawed and suspended in Williams’

E media supplemented with 24 mM sodium bicarbonate and

10% fetal bovine serum at 0.5 ꢀ 10⁶viable cells/mL for clearance

prediction. 1 (1 μM) for clearance predictions was incubated

with hepatocytes at 37 ꢀC for 4 h with gentle agitation. A gas

mixture of oxygen/carbon dioxide (95:5) maintained at about

2.5 kPa for about 5 s was passed through this mixture at every 1 h

of incubation. Flasks were corked immediately after gassing.

Freezing the incubation aliquots in liquid nitrogen terminated

the reactions. Following extraction, the samples were then ana-

lyzed by LC/MS/MS.

Plasma Protein Binding. The binding of 1 to rat, dog, monkey,

and human plasma proteins was determined using the equilibri-

um dialysis method. 1 was added to fresh plasma from animals or

human. 1 concentrations examined are pharmacologically rele-

vant based on in vivo efficacy studies in animals. The plasma

samples (1 mL) were subjected to equilibrium dialysis (n = 6 or 8)

using Spectra Por dialysis membranes (VWR, West Chester, PA,

USA) with a molecular mass cutoff of 12.4 kDa. The resulting

plasma samples were subjected to dialysis against isotonic phos-

phate buffer (pH 7.4) for 6 h. Following equilibration, the volume

of the plasma and buffer samples removed from each equilibra-

tion cell was noted. Aliquots of dialyzed plasma (100 mL) and

buffer (100 mL) were transferred to 96-well blocks, and acetoni-

trile (200 mL) containing indomethacin as internal standard

(50 ng mL^-1) was added to each well. Each matrix was normal-

ized to the other by the addition of equal volumes of the opposite

matrix (i.e., plasma to buffer and buffer to plasma). Following

extraction, the samples were then analyzed by LC/MS/MS.

Drug-Drug Interactions. The ability of a compound to act as

a competitive inhibitor of the five major CYPs, namely,

CYP1A2, 2C9, 2C19, 2D6, and 3A4, was examined as follows.

Phenacetin (50 mM) (a probe substrate for CYP1A2), diclo-

fenac (10 mM) (a probe substrate for CYP2C9), S-mephenytoin

(50 mM) (a probe substrate for CYP2C19), bufuralol (10 mM)

(a probe substrate for CYP2D6), testosterone (50 mM), or

midazolam (5 mM) (probe substrates for CYP3A4) was incu-

bated with human liver microsomes (CYP concentration =

0.25 mM), 3.3 mM MgCl₂, and 1.3 mM NADPH in a total

volume of 0.6 mL of 100 mM phosphate buffer, pH 7.4, in the

presence and absence of test compound at a concentration range

of 0-50 mM. The reactions were initiated by the addition of

NADPH, and incubations were conducted in a shaking water

bath at 37 ꢀC for 10 min, in the case of diclofenac, bufuralol,

testosterone, and midazolam, or at 30 min, in the case of phena-

cetin and S-mephenytoin. All reactions were terminated by the

addition of acetonitrile (200 mL), and the samples were centrifuged

at 2500g for 5 min before LC/MS/MS analysis. Procedures for

the quantitation of metabolites of probe P450 substrates used in

the present study have been described in detail by Obach.⁴¹Specific

CYP enzyme inhibitors, ketoconazole (CYP3A4), quinidine

(CYP2D6), sulfaphenazole (CYP2C9), (þ)-N-3-benzylnirvanol

(CYP2C19), and furafylline (CYP1A2), were used as positive

controls in the competitive inhibition studies.

used between passages 34 and 41. Complete culture medium was

removed from both the apical and basolateral compartments,

and the monolayer was preincubated with prewarmed apical

(0.5 mL) or basolateral (1.5 mL) buffer for 0.5 h at 37 ꢀC in a

shaking water bath. The apical buffer consisted of Hank’s balanced

salt solution, 25 mM d-glucose monohydrate, 20 mM 2-(N-mor-

pholino)ethanesulfonic acid, 1.25 mM CaCl₂, and 0.5 mM MgCl₂

(pH = 6.5). The basolateral buffer consisted of Hank’s balanced

salt solution, 25 mM d-glucose monohydrate, 20 mM N-(2-

hydroxyethyl)piperazine-N⁰-2-ethanesulfonic acid, 1.25 mM

CaCl₂, and 0.5 mM MgCl₂(pH = 7.4). At the end of the pre-

incubation, the media were removed and one (at concentrations

of 1, 10, and 50 mM) or blank buffer was added. The effects of

P-glycoprotein inhibitors on transport of 1 (1 mM) in both

apical to basolateral (A to B) and basolateral to apical (B to A)

directions were assessed by inclusion of verapamil (100 mM) or

CP-100,356 (10 mM) in both donor and acceptor well solutions.

To assess the integrity of the monolayer, the flux of [¹⁴C]mannitol

from apical to basolateral was determined following incubation

with 1. Inserts with a flux greater than 2%/h were excluded

from the study. Flux rate (F, mass/time) was calculated from the

slope of the appearance of 1 on the receiver side, and the

apparent permeability (P_app) was calculated from the equation

P_app(cm/s) = (FV_D)/(SC₀), where S is surface area of the

monolayer (0.83 cm²), V_Dis the donor volume, and C₀is the

initial concentration in the donor compartment. Incubations

were conducted in duplicate. Following extraction, the samples

were then analyzed by LC/MS/MS.

In Vivo Pharmacokinetic Studies. The Pfizer Institutional

Animal Care and Use Committee approved all procedures.

Male and female Sprague-Dawley rats (220-250 g), beagle dogs

(9-13 kg), and cynomolgus monkeys (4-6 kg) were used for

these studies. All animals were fasted overnight before dosing. 1

was administered via the jugular vein (rats) or via the brachial

vein (dogs and monkeys). 1 was administered iv at 3 mg/kg and

orally (po) at 10, 30, and 100 mg/kg in rats (administered in 20%

cremophor/10% ethanol/70% water), 3 mg/kg iv and 5 mg/kg

po in monkeys (administered in glycerol formal), and 3 mg/kg iv

and 5 mg/kg po in dogs (administered in glycerol formal). For all

po studies, 1 was given as a suspension in 0.5% methylcellulose.

Blood samples were taken, via the jugular vein (rats and dogs) or

femoral vein (monkeys), in series at appropriate time intervals,

and urine samples were also collected from rats for the length of

the entire study (0-24 h). All samples were kept frozen until

analysis. Following extraction, the samples were then analyzed

by LC/MS/MS.

Compound Preparation and Characterization. Routine ¹H

nuclear magnetic resonance (NMR) and ¹³C NMR spectra were

recorded on a Varian Inova 400 MHz spectrometer unless otherwise

specified. NMR structure elucidation experiments were collected

using a Bruker DMX-500 spectrometer equipped with a Bruker

5 mm broad-band observe (BBO) probe at 300 K. Standard

pulse sequences and phase cycling were used for all correlation

spectroscopy (COSY), heteronuclear single-quantum coherence

(HSQC), heteronuclear multiple bond coherence (HMBC), and

nuclear Overhauser enhancement spectroscopy (NOESY) spectra.

Low-resolution mass spectral data were collected using a Waters

Micromass ZMD (electrospray ionization, chromatography on

a Varian Polaris 5 C-18 column with acetonitrile and 0.1%

formic acid aqueous gradient eluant). Compounds were deter-

mined to be g95% pure by HPLC [C18, 0% acetonitrile in water

(0.1% trifluoroacetic acid) to 60% acetonitrile, UV 220 nm].

Reagents and solvents were obtained from commercial sources

unless otherwise noted. All reactions were run under nitrogen

unless otherwise noted. Chiral separations were conducted

using a Voyager instrument with a Chiralpak AD column and

a mobile phase of ethanol/hexane or isopropanol/hexane. X-ray

Caco-2 Cell Permeability.⁴²Caco-2 cells, obtained from the

American Type Culture Collection (Rockville, MD), were seeded

in 24-well Falcon multiwell plates (polyethylene terephthalate

membranes, pore size 1.0 mm) at 4.0ꢀ10⁴cells per well. The

cells were grown on filter supports for 3 weeks. Permeability

studies were performed after 15-17 days of culture. Cells were

˚

data were collected at 1 A (maximum sin Θ/λ=0.5) on a Siemens

R4RA/v diffractometer at room temperature.

8478 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24

Flanagan et al.

Compound Synthesis. 3-((3R,4R)-4-Methyl-3-(methyl(7H-

pyrrolo[2,3-d]pyrimidin-4-yl)amino)piperidin-1-yl)-3-oxopropane-

nitrile (1) Monocitrate. A mixture of 4-chloro-7H-pyrrolo[2,3-d]-

pyrimidine (5) (2.4 g, 15.9 mmol), prepared by the method of

Davoll,⁴³(3R,4R)-1-benzyl-N,4-dimethylpiperidin-3-amine (1.7 g,

7.95 mmol), prepared by the method of Brown Ripin²¹and

resolved as described in WO 02/096909, and 10 mL of triethylamine

was heated in a sealed tube at 100 ꢀC for 4 days. After cooling to

room temperature and concentration under reduced pressure, the

residue was purified by flash chromatography (silica; 3% metha-

nol in dichloromethane) affording 1.0 g (38%) of (3R,4R)-1-

benzyl-4-methylpiperidin-3-yl)methyl(7H-pyrrolo[2,3-d]pyrimidin-

4-yl-amine (19a) as a colorless oil. Alternative coupling conditions

can be found in WO 02/096909. Low-resolution mass spectrom-

etry (LMRS): m/z 336.1 (MH^þ). ¹H NMR (400 MHz) (CDCl₃)

δ: 0.95(3H, d,J=6.8 Hz), 1.61-1.75 (1 H, m), 1.81 (1 H, br s), 2.37

(2 H, br s), 2.82-2.95(1H, m), 3.54 (2H, s), 3.61(3H, s), 5.21 (1H,

br s), 6.58 (1 H, d, J = 4.0 Hz), 7.18-7.40 (5 H, m), 8.25 (1 H, s).

Alternatively, compound 19 was prepared by coupling racemic

1-benzyl-N,4-dimethylpiperidin-3-amine with 5. Compound 19

was then separated by chiral HPLC (Chiralpak AD; 85:15 hexanes/

ethanol); 19a had a retention time of 7.43 min, and 19b had a

retention time of 6.90 min. Compound 1 was then prepared from

19a, and compound 1a was prepared from 19b as described below.

To a solution of 79 g of (3R,4R)-1-benzyl-4-methylpiperidin-

3-yl)methyl(7H-pyrrolo[2,3-d]pyrimidin-4-yl)amine dissolved in

2 L of ethanol was added 79 g of 20% palladium hydroxide on

carbon (50% water by weight), and the mixture was agitated

under an atmospheric pressure of 50 psi hydrogen for 3 days

[conducting the hydrogenolysis at elevated temperature (50-70 ꢀC)

significantly decreases reaction times]. After the catalyst was

removed by filtration through Celite, 51 g of cyanoacetic acid

2,5-dioxopyrrolidin-1-yl ester was added to the ethanolic solu-

tion, and the resulting mixture was stirred at room temperature

for 1 h, at which time the ethanol was removed under reduced

pressure. The residue was redissolved in 1.0 L of dichloro-

methane, and the solution was sequentially washed with 0.6 L

of saturated aqueous sodium bicarbonate and 0.4 L of saturated

sodium bicarbonate. The combined aqueous layers were back-

washed with 0.4 L of 30 dichloromethane, and the dichloro-

methane layers were combined, dried over magnesium sulfate,

filtered, and concentrated in vacuo affording 61 g of amber oil.

This material was then redissolved in 2.1 L of acetone, and the

solution was heated to 40 ꢀC. Finely ground citric acid (37 g) was

added slowly (as a solid) to the solution. The mixture continued

stirring at 40 ꢀC for 2 h (granulation was complete). After being

cooled to room temperature, the solids were collected by filtra-

tion, washed with acetone, and dried in vacuo affording 78.5 g of

3-(3R,4R)-4-methyl-3-(methyl(7H-pyrrolo[2,3-d]pyrimidin-4-yl)-

amino)piperidin-1-yl)-3-oxopropanenitrile monocitrate (1 mono-

citrate) (66% from ((3R,4R)-1-benzyl-4-methylpiperidin-3-

yl)methyl(7H-pyrrolo[2,3-d]pyrimidin-4-yl)amine) as a slightly

off-white crystalline solid. This material was then combined

with 750 mL of 1:1 v/v ethanol/water, and the slurry was agitated

overnight. The solids were filtered and dried to afford 69 g of 1

monocitrate as a white crystalline solid (mp = 201 dec). LRMS:

reflux for 16 h. After being cooled to room temperature the

reaction mixture was washed with 10% CuSO₄and saturated

brine and dried over MgSO₄. Following filtration and concen-

tration under reduced pressure the crude product (3.42 g) was

purified by column chromatography (silica gel; 1:1 ethyl acetate/

hexanes) affording 1.06 g (28%) of 2a as an off-white solid.

LRMS: m/z 291 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ:

1.21-1.45 (3 H, m), 1.67 (4 H, br s), 1.81-1.99 (1 H, m), 2.33

(2 H, dd, J = 32.0 Hz, J = 11.6 Hz), 3.62 (1 H, s), 4.86 (1 H, br s),

7.00-7.18 (2 H, m), 7.19-7.35 (3 H, m), 8.35 (1 H, d, J =

6.8 Hz). High-resolution mass spectrometry (HRMS): calcd for

C₁₈H₁₈N₄(MH^þ) 291.1604, found 291.1613.

4-(Piperidin-1-yl)-7H-pyrrolo[2,3-d]pyrimidine (2b). To a stirred

solution of 5 (0.5 g/3.3 mmol) dissolved in 5 mL of tert-butanol

was added 1.1 g (13 mmol) of piperidine, and the resulting

mixture was heated to 85 ꢀC for 3 h. After being cooled to room

temperature the pH of the reaction mixture was adjusted to ∼1

with 1 N HCl (aqueous) and washed 2 ꢀ 20 mL with ether. The

aqueous layer was then basified with 1 N NaOH precipitating

the product, which was collected by filtration and dried in a

desiccator affording 395 mg (60%) of 2b as a white solid. LRMS:

m/z 203.1. ¹H NMR (400 MHz) (CDCl₃) δ: 1.69 (6 H, s), 3.88

(4 H, s), 6.47 (1 H, d, J = 3.2 Hz), 7.06 (1 H, d, J = 3.2 Hz), 8.32

(1 H, s), 11.65 (1 H, br s). ¹³C NMR (400 MHz) (CDCl₃) δ: 24.8,

25.9, 47.3, 101.5, 102.9, 120.3, 150.9, 152.2, 157.3. HRMS: calcd

for C₁₁H₁₅N₄(MH^þ) 203.1291, found 203.1296.

Compounds 2c through 2i were prepared in an analogous

manner to that described for 2b.

4-(2,5-Dimethylpyrrolidin-1-yl)-7H-pyrrolo[2,3-d]pyrimidine (2c).

LRMS: m/z 217.3 (MH^þ). ¹H NMR (400 MHz) (CD₃OD)

δ: 1.43 (6 H, d, J = 6.3 Hz), 1.82-1.87 (2 H, m), 2.14-2.17

(2 H, m), 4.50-4.53 (2 H, m), 6.55 (1 H, d, J = 3.5 Hz), 7.07

(1 H, d, J = 3.5 Hz), 8.07 (1 H, s). ¹³C NMR (400 MHz)

(CD₃OD) δ: 21.2, 31.4, 55.1, 101.7, 102.5, 120.5, 150.2, 150.7, 155.0.

4-(Octahydro-1H-indol-1-yl)-7H-pyrrolo[2,3-d]pyrimidine (2d).

LRMS: m/z 243.3 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ:

1.18-1.43 (3 H, m), 1.55 (1 H, d, J = 10.8 Hz), 1.62-1.95 (4 H,

m), 2.16 (1 H, s), 2.26 (1 H, br s), 2.40 (1 H, br s), 3.82 (1 H, d, J =

7.6 Hz), 3.95 (1 H, d, J = 8.8 Hz), 4.38 (1 H, br s), 6.50 (1 H, s),

7.04 (1 H, s), 8.30 (1 H, s).

N-Cyclopentyl-N-methyl-7H-pyrrolo[2,3-d]pyrimidin-4-amine

(2e). LRMS: m/z 217.2 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ:

1.58-2.05 (9 H, m), 3.22 (3 H, s), 5.31 (1 H, t, J = 7.6 Hz), 6.56

(1 H, d, J = 3.6 Hz), 7.02 (1 H, d, J = 3.6 Hz), 8.29 (1 H, s).

N-Cyclohexyl-N-methyl-7H-pyrrolo[2,3-d]pyrimidin-4-amine

(2f). LRMS: m/z 231.2 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ:

1.07-1.28 (2 H, m), 1.39-1.62 (3 H, m), 1.73 (2 H, d, J = 13.6

Hz), 1.78-1.95 (4 H, m), 3.22 (3 H, s), 4.68 (1 H, br s), 6.53 (1 H,

d, J = 3.2 Hz), 7.02 (1 H, d, J = 3.6 Hz), 8.30 (1 H, s). ¹³C NMR

(400 MHz) (CDCl₃) δ: 25.7, 25.9, 29.7, 30.4, 30.9, 55.4, 102.0,

102.8, 119.8, 150.9, 151.8, 157.5. HRMS: calcd for C₁₃H₁₉N₄

(MH^þ) 231.1604, found 231.1603.

N-Methyl-N-(4-methylcyclohexyl)-7H-pyrrolo[2,3-d]pyrimidin-

1

4-amine (2g). LRMS: m/z 245.2 (MH^þ). H NMR (400 MHz)

(CDCl₃) δ: 0.91 (1.5 H, d, J = 6.4 Hz), 1.02 (1.5 H, d, J = 7.2

Hz), 1.17 (2 H, q, J = 10.0 Hz), 1.38 (2 H, br s), 1.60 (3 H, t, J =

14.0 Hz), 1.70-2.03 (3 H, m), 3.20 (1.5 H, s), 3.24 (1.5 H, s), 4.64

(1 H, br s), 6.51 (1 H, t, J = 3.2 Hz), 7.03 (1 H, d, J = 3.2 Hz),

8.27 (1 H, s).

N-Cycloheptyl-N-methyl-7H-pyrrolo[2,3-d]pyrimidin-4-amine

(2h). LRMS: m/z 245.5 (MH^þ). ¹H NMR (400 MHz) (CD₃OD)

δ: 1.40-1.83 (12 H, m), 3.09 (3 H, s), 3.21 (1 H, s), 6.50 (1 H, d,

J = 3.6 Hz), 6.96 (1 H, d, J = 3.2 Hz), 7.97 (1 H, s). ¹³C NMR

(400 MHz) (CD₃OD) δ: 26.4, 28.8, 32.1, 33.4, 58.8, 103.3, 104.0,

121.5, 151.7, 151.9, 158.2. HRMS: calcd for C₁₄H₂₁N₄(MH^þ)

245.1761, found 245.1759.

1

m/z 313.2 (MH^þ). H NMR (400 MHz) (D₂O) δ HOD: 0.92

(2 H, d, J = 7.2 Hz), 0.96 (1 H, d, J = 7.6 Hz), 1.66 (1 H, m), 1.80

(1 H, m), 2.37 (1 H, m), 2.58 (2 H, 1/2 ABq, J = 15.4 Hz), 2.70 (2

H, 1/2 ABq, J = 15.4 Hz), 3.23 (2 H, s), 3.25 (1 H, s), 3.33 (1 H,

m), 3.46 (1 H, m), 3.81 (4 H, m), 4.55 (1 H, m), 6.65 (1 H, d, J =

3.2 Hz), 7.20 (1 H, t, J = 3.2 Hz), 8.09 (1 H, m). Anal. Calcd for

C₂₂H₂₈N₆O₈: C, 52.38; H, 5.59; N, 16.66. Found: C, 52.32; H,

5.83; N, 16.30. For additional characterization of the monoci-

trate salt of 1 see WO 03/048162.

9-(7H-Pyrrolo[2,3-d]pyrimidin-4-yl)-2,3,4,4a,9,9a-hexahydro-

1H-carbazole (2a). To a stirred solution of 5 (2 g/0.013 mol) in

100 mL of dimethylformamide was added 3.3 g (0.013 mol) of

2,3,4,4a,9,9a-hexahydro-1H-carbazole hydrobromide followed

by 2.1 mL of pyridine, and the resulting mixture was stirred at

N-Cyclooctyl-N-methyl-7H-pyrrolo[2,3-d]pyrimidin-4-amine

(2i). LRMS: m/z 259.3 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ:

1.46-1.85 (14 H, m), 3.19 (3 H, s), 5.03 (1 H, br s), 6.56 (1 H, d,

J = 3.2 Hz), 7.00 (1 H, d, J =3.2 Hz), 8.29 (1 H, s), 9.79 (1 H, br s).

Article

HSA Protocol. Ketones (3) (1 mmol) were added to tarred

Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24 8479

25.5, 28.8, 32.2, 33.0, 33.9, 33.9, 34.2, 34.4, 36.2, 37.9, 42.3, 54.3,

59.2, 105.7, 105.8, 106.7, 124.1, 154.2, 154.2, 154.4, 161.2, 161.4.

HRMS: calcd for C₁₄H₂₁N₄(MH^þ) 245.1761, found 245.1756.

Compounds 6b through 6h were prepared in an analogous

manner to that described for 6a. Compounds 6d, 6e, 6g, and 6h

were characterized by LRMS and HPLC purity as described

above.

N-Methyl-N-(2-methylcyclohexyl)-7H-pyrrolo[2,3-d]pyrimidin-

4-amine (6b) (∼3:1 Mixture of Diastereomers). LRMS: m/z 245.5

(MH^þ). ¹H NMR (400 MHz) (CD₃OD) δ: 0.66 (0.75 H, d, J =

7.0 Hz), 0.92 (2.25 H, d, J = 7.0 Hz), 0.98-1.24 (1 H, m),

1.25-1.42 (2 H, m), 1.43-1.55 (2 H, m), 1.56-1.82 (2 H, m),

1.90 (1 H, sp q, 2.29 (1 H, br s), 3.15 (2.25 H, s), 3.21 (0.75 H, s),

4.52-4.60 (1 H, m), 6.48 (1 H, d, J = 3.6 Hz), 6.97 (1 H, d, J =

3.2 Hz), 7.98 (0.25 H, s), 8.00 (0.75 H, s). ¹³C NMR (400 MHz)

(CD₃OD) δ: 13.5, 19.4, 20.9, 26.3, 26.9, 27.0, 27.9, 33.5, 34.0,

34.1, 36.0, 36.4, 59.7, 103.4, 104.0, 121.6, 151.4, 151.7, 158.6,

159.4. HRMS: calcd for C₁₄H₂₁N₄(MH^þ) 245.1761, found

245.1754.

4 mL vials and dissolved in 2 mL of methanol containing 4 mmol

of methylamine. Acetic acid (120 mg/2 mmol as a 4 M solution in

dichloroethane) was added followed by 1 g of borohydride

polymer support (Amberlite IRA-400, 2.5 mmol BH₄/g resin),

and the capped vials were agitated for 18 h at room temperature.

The reaction mixtures were then transferred to filters to remove

the resin, the resin was washed 2 ꢀ 3 mL with methanol, and the

samples were concentrated affording crude amines 4, which

were carried on without additional purification. To these vials

was added 5 (0.5 mmol) as a 0.5 M solution in tetrahydrofuran

(THF) followed by 3 mL of a 10% solution of dichloroethane in

tert-butanol and 1 M in diisopropylethylamine. The resulting

capped mixtures were agitated at 80 ꢀC for 40 h and then

concentrated in a vacuum oven at 35 ꢀC. The crude products

were dissolved in 3 mL of 10% methanol in dichloromethane

and transferred to 6 cm³strong cation-exchange (SCX) columns

that had been preconditioned with 3 column volumes of metha-

nol. The columns were then washed 3ꢀ5 mL with methanol,

and the product was eluted with 5 mL of 0.1 N ammonia

affording test compounds 6 after drying in a vacuum oven at

35 ꢀC. QC was performed on all final compounds by LRMS and

HPLC [C18, 5 μm, 19 ꢀ 100 mm; solvent A, 0.1% trifluoroacetic

acid in water (v/v); solvent B, 0.1% trifluoroacetic acid in

acetonitrile (v/v)].

N-(2,5-Dimethylcyclohexyl)-N-methyl-7H-pyrrolo[2,3-d]pyri-

midin-4-amine (6c) (Mixture of Diastereomers). LRMS: m/z

1

259.1 (MH^þ). H NMR (400 MHz) (CD₃OD) δ: 0.68 (1 H, d,

J = 6.8 Hz), 0.82 (1 H, d, J = 7.0 Hz), 0.84 (2 H, d, J = 6.0 Hz),

0.88 (1 H, d, J = 7.2 Hz), 1.01-1.22 (1 H, m), 1.22-1.41 (1 H,

m), 1.42-1.62 (2 H, m), 1.63-1.76 (1 H, m), 1.77-1.89 (0.5 H,

m), 2.05 (0.5 H, d, J = 12.2), 2.15 (0.5 H, br s), 2.26 (0.5 H, br s),

3.02 (0.5 H, s), 3.08 (0.5 H, s), 3.14 (1.5 H, s), 3.18 (1.5 H, s),

4.55-4.62 (1 H, m), 6.42-6.54 (1 H, m), 6.92-6.99 (1 H, m),

7.96-8.07 (1 H, m). ¹³C NMR (400 MHz) (CD₃OD) δ: 22.7,

23.0, 28.4, 29.0, 29.7, 29.8, 30.2, 31.1, 32.3, 33.0, 33.2, 33.6, 33.9,

34.0, 34.7, 34.8, 34.8, 35.4, 35.7, 35.8, 36.2, 36.9, 39.3, 48.4, 48.6,

48.8, 49.0, 49.2, 49.5, 49.7, 54.1, 59.4, 103.4, 104.1, 104.1, 121.5,

121.5, 151.6, 151.7, 151.8, 151.9, 152.0, 158.7, 159.2, 159.4.

HRMS: calcd for C₁₅H₂₃N₄(MH^þ) 259.1917, found 259.1917.

N-(2-Isopropyl-5-methylcyclohexyl)-N-methyl-7H-pyrrolo[2,3-d]-

pyrimidin-4-amine (6f) (∼3:1 Mixture of Diastereomers). LRMS:

m/z 287.4 (MH^þ). ¹H NMR (400 MHz) (CD₃OD) δ: 0.55 (1 H,

br s), 0.71 (3 H, d, J = 7.0 Hz), 0.79 (1 H, d, J = 7.0 Hz), 0.91

(3 H, d, J = 6.4 Hz), 0.97 (3 H, d, J = 4.0 Hz), 1.38-1.60 (3 H,

m), 1.61-1.73 (2 H, m), 1.82-2.00 (2 H, m), 3.21 (1 H, s), 3.21

(3 H, s), 4.62 (1 H, br s), 6.49 (0.75 H, d, J = 2.8 Hz), 6.55 (0.25

H, s), 6.99 (1 H, s), 7.99 (0.25 H, s), 8.01 (0.75 H, s). ¹³C NMR

(400 MHz) (CD₃OD) δ: 16.5, 21.5, 21.9, 22.7, 23.0, 24.9, 26.0,

28.1, 28.4, 31.8, 33.4, 34.1, 34.6, 35.6, 36.6, 39.4, 43.3, 46.1, 48.4,

48.6, 48.8, 49.0, 49.3, 49.7, 49.7, 60.9, 103.7, 104.0, 122.0, 150.4,

150.6, 150.9, 158.0. HRMS: calcd for C₁₇H₂₇N₄(MH^þ) 287.2230,

found 287.2231.

(5S)-2-Methyl-5-(prop-1-en-2-yl)cyclohexanone (8). A solution

of S-(þ)-carvone (7) (2.05 g/13.6 mmol) dissolved in 14 mL of dry

THF was cooled to -70 ꢀC, and to this mixture was added

dropwise 14 mL of a 1.0 M solution of L-selectride in THF. Once

the addition was complete, the reaction mixture stirred at -70 ꢀC

for 2 h, at which time the reaction was warmed to 0 ꢀC and 40 mL

of a 10% solution of sodium hydroxide was added dropwise with

stirring over a 10 min period. Hydrogen peroxide (30%, 28 mL)

was then added dropwise over a 10 min period, and the mixture

was warmed to room temperature, transferred to a separatory

funnel, and extracted 3 ꢀ 100 mL with hexanes. The organic

extracts were washed 2 ꢀ 100 mL with water, 2 ꢀ 100 mL with 1.5 M

sodium bisulfite, and 50 mL with saturated brine. The organic

layer was then dried over MgSO₄, filtered through Celite, and

concentrated to dryness in vacuo. The crude product was then

purified by flash chromatography (silica gel; 95:5 hexanes/ethyl

acetate) affording 1.08 g (52%) of compound 8 as a colorless oil.

LRMS: m/z 153.4 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ: 1.00

(2.5 H, d, J = 6.4 Hz), 1.03(0.5 H, d, J = 6.8 Hz), 1.30-1.38 (1 H,

m), 1.60(1H,d,J=9.2 Hz), 1.69 (3 H, s), 1.84 (1 H, d, J=25.6 Hz),

2.05-2.12 (1 H, m), 2.98-2.43 (4 H, m), 4.69 (2 H, d, J = 10.4 Hz).

(5S)-N,2-Dimethyl-5-(prop-1-en-2-yl)cyclohexanamine (9). To a

250 mL round-bottom flask was charged 3.21 g (15.1 mmol) of

Preparation of 4a through 4h was conducted in an analogous

manner to that described in the HSA Protocol starting with

ketones 3a through 3h. Some amine intermediates (4) were

carried on without characterization.

N,3-Dimethylcyclohexanamine (4a) (Formate). ¹H NMR (400

MHz) (CDCl₃) δ: 0.97 (3 H, d, J = 7.2 Hz), 1.23 (1 H, dd, J =

116 Hz, J = 12.4 Hz), 1.25 (1 H, br s), 1.52-1.73 (4 H, m),

1.78-1.91 (2 H, m), 1.98-2.15 (1 H, m), 2.59 (3 H, s), 3.12 (1 H,

q, J = 3.6 Hz), 8.61 (1 H, br s). HRMS: calcd for C₈H₁₇N

(MH^þ) 128.1434, found 128.1433.

N,2-Dimethylcyclohexanamine (4b) (Formate). ¹H NMR (400

MHz) (CDCl₃) δ: 0.89-1.08 (3 H, m), 1.22 (2 H, br s), 1.28-1.81

(5 H, m), 2.01 (1 H, br s), 2.23 (1 H, br s), 2.68 (3 H, s), 2.91 (1 H,

d, J = 9.2 Hz), 8.47 (1 H, s). HRMS: calcd for C₈H₁₇N (MH^þ)

128.1434, found 128.1432.

N,2,5-Trimethylcyclohexanamine (4c) (Formate). ¹H NMR

(400 MHz) (CDCl₃) δ TMS: 0.89 (3 H, d, J = 6.0 Hz), 0.94 (3

H, d, J = 7.4 Hz), 0.95-1.12 (2 H, m), 1.21-1.63 (3 H, m), 1.75

(1 H, dd, J = 28.0 Hz, J = 12.4 Hz), 2.74 (1 H, br s), 2.81 (3 H, s),

2.92 (1 H, d, J = 10.8 Hz), 8.46 (1 H, br s). HRMS: calcd for

C₉H₁₉N (MH^þ) 142.1590, found 142.1590.

2-Isopropyl-N,5-dimethylcyclohexanamine (4f) (Formate). ¹H

NMR (400 MHz) (CDCl₃) δ: 0.61-0.78 (9 H, m), 0.82 (1 H, d,

J = 6.4 Hz), 0.91 (1 H, t, J = 12.8 Hz), 1.08-1.32 (1 H, m),

1.43-1.65 (2 H, m), 1.84 (1 H, d, J = 14.8 Hz), 2.41 (3 H, s), 3.14

(1 H, s), 8.32 (1 H, s). HRMS: calcd for C₁₁H₂₃N (MH^þ)

170.1903, found 170.1902.

N-Methyl-N-(3-methylcyclohexyl)-7H-pyrrolo[2,3-d]pyrimidin-

4-amine (6a). To a slurry of 5 (50 mg/0.32 mmol) in 1 mL of tert-

butanol was added 45 mg (0.35 mmol) of 4a followed by 84 mg

(0.65 mmol) of diisopropylethylamine, and the resulting mixture

was heated to 80 ꢀC for 24 h. The reaction mixture was then

cooled to room temperature and diluted with 25 mL of dichloro-

methane. The solution was washed with 1 N HCl and then with

saturated NaHCO₃, dried over Na₂SO₄, and concentrated to dryness

in vacuo. The crude product was then purified by preparative

thin-layer chromatography (TLC) (silica gel; 2:1 hexanes/ethyl

acetate) affording 20 mg of 6a (25%) as an ∼2:1 mixture of

diastereomers. LRMS: m/z 245.2 (MH^þ). ¹H NMR (400 MHz)

(CD₃OD) δ: 0.68-0.80 (1 H, m), 0.84 (1 H, d, J = 7.2 Hz), 1.03

(2 H, d, J = 6.8 Hz), 1.18 (1 H, q, J = 12.2 Hz), 1.31-1.68 (5 H,

m), 1.69-1.82 (1 H, m), 2.08 (1 H, br s), 3.06 (2 H, s), 3.09 (1 H,

s), 3.21 (1 H, s), 4.55 (1 H, br s), 6.46 (0.3 H, d, J = 3.6 Hz), 6.48

(0.7 H, d, J = 2.0 Hz), 6.97 (1 H, t, J = 2.0 Hz), 7.98 (0.3 H, s),

7.99 (0.7 H, s). ¹³C NMR (400 MHz) (CD₃OD) δ: 21.1, 23.7,

8480 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24

Flanagan et al.

sodium triacetoxyborohydride and 1.08 g (7.1 mmol) of 8

dissolved in 70 mL of THF, to this mixture was added 3.6 mL

of 2.0 M solution of methylamine in THF, and the resulting

mixture was stirred at room temperature for 16 h, at which time

the reaction was quenched upon addition of 75 mL of water. The

resulting mixture was extracted 3ꢀ100 mL with chloroform, and

the combined organic extracts were dried over MgSO₄, filtered

through Celite, and concentrated to dryness in vacuo. The crude

product was then purified by flash chromatography (silica gel;

6:4 dichloromethane/hexanes to 9:1 dichloromethane/methanol)

affording 9 as a colorless oil. LRMS: m/z 168.5 (MH ^þ). ¹H

NMR (400 MHz) (CDCl₃) δ: 0.79 (2 H, d, J = 7.2 Hz), 0.87 (1 H,

d, J = 6.8 Hz), 1.14 (1 H, t, J = 4.8 Hz), 1.21-1.55 (3 H, m),

1.56-1.73 (2 H, m), 1.65 (3 H, s), 1.80-2.20 (2 H, m), 2.34 (1 H,

s), 2.35 (2 H, s), 2.47 (0.6 H, d, J = 12 Hz), 2.57 (0.3 H, d, J = 3.2

Hz), 4.61 (2 H, d, J = 3.6 Hz).

N-Methyl-N-((5S)-2-methyl-5-(prop-1-en-2-yl)cyclohexyl)-7H-

pyrrolo[2,3-d]pyrimidin-4-amine (Diastereomeric Mixture 11). A

sealed tube was charged with 119 mg (0.78 mmol) of 5, 233 mg

(1.02 mmol) of 9, 320 mg (3.16 mmol) of triethylamine, and 3 mL

of tert-butanol, and the capped mixture was stirred at 100 ꢀC for

48 h. After being cooled to room temperature the mixture was

diluted with 30 mL of ether, transferred to a separatory funnel,

and washed 2ꢀ20 mL with 1 N HCl, and the aqueous layers

were combined and adjusted to pH 12 with 2 N KOH. The result-

ing basic layer was extracted 3 ꢀ 100 mL with dichloromethane,

and the combined dichloromethane extracts were dried over

MgSO₄, filtered through Celite, and concentrated under reduced

pressure affording a yellow oil. The crude product was then

purified by flash chromatography (silica gel; 7:3 ethyl acetate/

hexanes), and the fraction with an R_f= 0.16 was collected and

concentrated to dryness in vacuo affording 10.4 mg of com-

pound 11 as a white solid. LRMS: m/z 285.3 (MH ^þ). ¹H NMR

(400 MHz) (CDCl₃) δ: 0.82 (1.5 H, d, J = 6.4 Hz), 1.00 (1.5 H, d,

J = 7.2 Hz), 1.40-1.48 (1 H, m), 0.65 (0.5 H, d, J = 13.2 Hz),

0.70 (0.5 H, d, J = 13.6 Hz), 1.69-1.95 (4 H, m), 1.72 (1.5 H, s),

1.75 (1.5 H, s), 2.05-2.22 (1 H, m), 2.45 (1 H, br s), 3.21 (1 H, s),

3.28 (3 H, s), 4.70 (1 H, d, J = 12.4 Hz), 4.73 (1 H, d, t, J = 2.8

Hz), 4.77 (1 H, br s), 6.51 (0.5 H, d, J = 3.2 Hz), 6.55 (0.5 H, s),

7.02 (0.5 H, d, J = 3.6 Hz), 7.03 (0.5 H, d, J = 3.6 Hz), 8.26 (0.5

H, s), 8.28 (0.5 H, s).

Diastereomeric mixture 10 was prepared in an analogous

manner to that described for 11 beginning with R-(-)-carvone.

N-Methyl-N-((1S,2R,5S)-2-methyl-5-(prop-1-en-2-yl)cyclo-

hexyl)-7H-pyrrolo[2,3-d]pyrimidin-4-amine (12). Compound 12

was isolated from diastereomeric mixture 11 by chiral HPLC

(Chiralpak AD; 9:1 hexanes/isopropanol þ 0.1% DEA mod-

ifier). The retention time for 12 was 8.26 min; its diastereomer

came off at 13.26 min. The peaks were collected and concen-

trated to dryness in vacuo. Physical data for 12: LRMS: m/z

285.5 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ: 1.01 (3 H, d, J =

6.8 Hz), 1.42 (sp t, J = 9.2 Hz), 1.58 (1 H, br d, J = 12.8 Hz),

1.63-1.94 (4 H, m), 1.75 (3 H, s), 2.11 (1 H, t, J = 11.6 Hz), 2.45

(1 H, br s), 3.30 (3 H, s), 4.73 (2 H, s), 4.77 (1 H, br s), 6.53 (1 H, d,

J = 3.6 Hz), 7.04 (1 H, d, J = 3.6 Hz), 8.27 (1 H, s). 1D and 2D

NMR structural assignments for 12 were consistent with the all-

cis relative stereochemistry around the cyclohexane ring.

N-Methyl-N-((1S,2R,5S)-2-methyl-5-(prop-1-en-2-yl)cyclo-

hexyl)-7-(phenylsulfonyl)-7H-pyrrolo[2,3-d]pyrimidin-4-amine (13).

To 5 mL of dimethylformamide (DMF) was added 217 mg (5.43

mmol) of sodium hydride and after cooling to 0 ꢀC was added

1.02 g (3.59 mmol) of 12 dissolved in 5 mL of DMF dropwise over a

10-min period. After bubbling ceased, 685 μL of phenylsulfonyl

chloride was added, and the mixture was brought to room tem-

perature and stirred for 1.5 h, at which time the reaction was

quenched upon addition of 10 mL of water. The resulting mixture

was then extracted 3 ꢀ 10 mL with dichloromethane, and the

combined extracts were washed 2 ꢀ with water and 2 ꢀ with

saturated brine, dried over MgSO₄, filtered, and concentrated

under reduced pressure producing 1.1 g of a yellow oil. The oil

was purified by flash chromatography (silica gel; 9:1 hexanes/

ethyl acetate) affording 748 mg (50%) of 13 as a white solid.

LRMS: m/z 425.3 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ: 0.94

(3 H, d, J = 7.2 Hz), 1.39 (1 H, dd, J = 13.2 Hz, J = 3.6 Hz),

1.49-1.65 (2 H, m), 1.72 (3 H, s), 1.75-1.86 (2 H, m), 2.06 (1 H,

t, J = 12.0 Hz), 2.33 (1 H, br s), 3.18 (3 H, s), 4.64 (1 H, br s), 4.71

(2 H, s), 6.58 (1 H, d, J = 3.6 Hz), 7.42 (1 H, d, J = 4.4 Hz), 7.47

(2 H, d, J = 8.0 Hz), 7.56 (1 H, t, J = 2.0 Hz), 8.15 (2 H, d, J =

8.8 Hz), 8.37 (1 H, s).

(R)-2-((1S,3S,4R)-4-Methyl-3-(methyl(7-(phenylsulfonyl)-7H-

pyrrolo[2,3-d]pyrimidin-4-yl)amino)cyclohexyl)propan-1-ol (14).

To a stirred solution of 13 (745 mg/1.76 mmol) in 5 mL of THF

was added 7.0 mL (3.50 mmol) of 9-borabicyclo(3.3.1)nonane

(9-BBN) (0.5 M in THF; Aldrich), and the resulting solution was

heated to reflux for 18 h (reaction incomplete by TLC). Addi-

tional 9-BBN (7.0 mL) was added with continued refluxing for

2 h, at which time the reaction was cooled to room temperature,

and water (10 mL) slowly added. The reaction mixture was

partitioned between saturated NaHCO₃and dichloromethane,

and the organic layer was washed 2 ꢀ with water and with

saturated brine, dried over MgSO₄and concentrated under

reduced pressure. The resulting crude oil was then purified by

flash chromatography (silica gel; 1:1 hexanes/ethyl acetate) afford-

ing 428 mg (55%) of 14 as a white solid. LRMS: m/z 443.1

1

(MH^þ). H NMR (400 MHz) (CDCl₃) δ: 0.85-0.97 (6 H, m),

1.20-1.29 (1 H, m), 1.47 (1 H, d, J = 16.4 Hz), 1.55-1.64 (4 H,

m), 1.65-1.80 (1 H, m), 2.33 (1 H, br s), 3.19 (3 H, s), 3.44-3.56

(1 H, m), 3.58-3.63 (1 H, m), 4.59 (1 H, br s), 6.58 (1 H, d, J =

3.6 Hz), 7.43 (1 H, d, J = 4.0 Hz), 7.49 (2 H, d, J = 8.0 Hz), 7.57

(1 H, t, J = 2.0 Hz), 8.16 (2 H, d, J = 7.2 Hz), 8.38 (1 H, s).

(R)-2-((1S,3S,4R)-4-Methyl-3-(methyl(7H-pyrrolo[2,3-d]pyri-

midin-4-yl)amino)cyclohexyl)propan-1-ol (16a). To a stirred so-

lution of 14 (400 mg/0.90 mmol) in 10 mL of THF at room

temperature was added ∼10 mL of a solution made by dissolving

sodium hydroxide pellets in methanol. The resulting mixture

was stirred until the reaction was complete by TLC (∼3 h). To

the reaction was then added 50 mL of water, and the mixture was

extracted 3 ꢀ with ethyl acetate. The combined extracts were

washed 2 ꢀ with water and saturated brine, dried over MgSO₄,

filtered, and concentrated to dryness in vacuo, affording 196 mg

(72%) of 16a as a white solid. This material could also be pre-

pared by direct hydroboration of compound 12 using 9-BBN.

LRMS: m/z 303.5 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ: 0.81

(1.5 H, d, J = 6.4 Hz), 0.86-1.00 (4.5 H, m), 1.14-1.26 (1 H, m),

1.37-2.49 (5 H, m), 1.55-2.64 (5 H, m), 3.21 (1 H, br s),

3.44-3.59 (1 H, m), 3.60-3.63 (1 H, m), 4.70 (1 H, br s), 6.48

(1 H, d, J = 3.6 Hz), 6.56 (1 H, br s), 7.00 (1 H, d, J = 3.6 Hz),

8.22 (0.5 H, s), 8.44 (0.5 H, s).

N,N-Diethyl-2-((R)-2-((1S,3S,4R)-4-methyl-3-(methyl(7H-

pyrrolo[2,3-d]pyrimidin-4-yl)amino)cyclohexyl)propoxy)acetamide

(16b). To a slurry of sodium hydride (7.0 mg, 60% dispersion) in

0.5 mL of DMF at 0 ꢀC was added 49 mg (0.112 mmol) of 14

dissolved in 0.5 mL of DMF, the resulting mixture stirred at 0 ꢀC

for 30 min, at which time 15 μL (0.113 mmol) of 2-chloro-N,N-

diethylacetamide (Aldrich) was added, and the resulting solu-

tion was stirred at 0 ꢀC for 4 h, at which time 10 mL of water was

added. This mixture was then extracted 3ꢀ15 mL with ethyl

acetate, and the combined extracts were washed with 10 mL of

water and then 10 mL of saturated brine, dried over MgSO₄,

filtered, and concentrated under reduced pressure affording

45 mg of intermediate 15b. This material (15b) was then dis-

solved in 1.0 mL of THF, and to this solution was added 1.0 mL

of a solution made by dissolving NaOH pellets in methanol. The

resulting mixture was stirred until the reaction was complete by

TLC (∼30 min). To the reaction was then added 10 mL of water,

and the mixture was extracted 3 ꢀ with ethyl acetate. The

combined extracts were washed 2ꢀ with water and saturated

brine, dried over MgSO₄, filtered, and concentrated to dryness

in vacuo, affording 20 mg (61%) of 16b as a colorless oil. LRMS:

m/z 416.4 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ: 0.78-0.92

Article

Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24 8481

(1 H, m), 0.93-0.96 (1 H, m), 0.97 (3 H, d, J = 7.6 Hz), 1.10 (3 H,

t, J = 7.2 Hz), 1.16-1.20 (3 H, m), 1.24 (1 H, s), 1.43 (2 H, br s),

1.52-1.63 (4 H, m), 1.75 (2 H, q, J = 4.0 Hz), 3.15-2.28 (1 H,

m), 3.30 (3 H, s), 3.31 (2 H, s), 3.32-3.43 (2 H, m), 4.69 (2 H, br

s), 5.01 (4 H, q, J = 11.6 Hz), 6.52 (1 H, d, J = 3.6 Hz), 7.01 (1 H,

d, J = 3.6 Hz), 8.27 (1 H, s).

N-Methyl-2-((R)-2-((1S,3S,4R)-4-methyl-3-(methyl(7H-pyrrolo-

[2,3-d]pyrimidin-4-yl)amino)cyclohexyl)propoxy)acetamide (16c).

LRMS: m/z 374.1 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ:

0.78-1.03 (6 H, m), 1.18-1.84 (9 H, m), 3.27 (1.5 H, s), 3.32 (1.5

H, s), 3.49-3.59 (1 H, m), 3.60-3.69 (1 H, m), 3.91-4.06 (1 H,

m), 4.21 (1 H, q, J = 4.0 Hz), 4.72 (1 H, br s), 6.45 (1 H, br s), 6.47

(0.5 H, d, J = 12.0 Hz), 6.52 (0.5 H, d, J = 2.8 Hz), 6.92 (0.5 H,

d, J = 3.6 Hz), 6.99 (0.5 H, d, J = 3.6 Hz), 8.23 (0.5 H, s), 8.31

(0.5 H, s).

N-Benzyl-2-((R)-2-((1S,3S,4R)-4-methyl-3-(methyl(7H-pyrrolo-

[2,3-d]pyrimidin-4-yl)amino)cyclohexyl)propoxy)acetamide (16d).

LRMS: m/z 450.1 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ:

0.78-1.03 (6 H, m), 1.37-1.82 (6 H, m), 2.41 (2 H, br s), 3.23

(3 H, s), 3.26 (1 H, s), 3.29 (1 H, s), 3.44-3.53 (1 H, m),

3.59-3.68 (1 H, m), 4.20 (1 H, q, J = 7.2 Hz), 4.37 (1 H, d,

J = 6.4 Hz), 4.45 (1 H, dd, J = 6.4 Hz, J = 4.4 Hz), 4.71 (2 H, d,

J = 5.2 Hz), 6.44 (1 H, br s), 6.55 (1 H, d, J = 3.6 Hz), 6.95 (1 H,

d, J = 3.6 Hz), 7.19-7.35 (5 H, m), 8.24 (1 H, s).

room temperature for 18 h. The reaction mixture was then

diluted with dichloromethane, washed with saturated NaHCO₃,

dried over MgSO₄, filtered, and concentrated to dryness in vacuo.

The intermediate phenyl (3-methylisoxazol-5-yl)carbamate was

carried on without further purification by forming a slurry in

tert-butanol (2 mL) and adding 50 mg (0.20 mmol) of cis-N-

methyl-N-(4-methylpiperidin-3-yl)-7H-pyrrolo[2,3-d]pyrimidin-

4-amine; the resulting reaction mixture was heated to 80 ꢀC for

48 h. After being cooled to room temperature the reaction was

concentrated under reduced pressure, redissolved in dichloro-

methane, washed with saturated NaHCO₃, dried over MgSO₄,

filtered through Celite, and concentrated to dryness in vacuo.

The crude product was then purified by preparative TLC (silica

gel; 10:1 dichloromethane/methanol) affording 7 mg of pure 17b

after two chromatographies. LRMS: m/z 386.2 (MH^þ). ¹H

NMR (400 MHz) (CD₃OD) δ: 1.12 (3 H, d, J = 7.2 Hz),

1.70-2.00 (4 H, m), 2.30 (3 H, s), 2.45-2.56 (1 H, m), 3.44 (3 H,

s), 3.67 (2 H, t, J = 6.0 Hz), 3.88-4.02 (1 H, m), 4.94 (1 H, br s),

6.62 (1 H, s), 6.71 (1 H, d, J = 3.6 Hz), 7.16 (1 H, d, J = 3.6 Hz),

8.15 (1 H, s). ¹³C NMR (400 MHz) (CD₃OD) δ: 14.7, 18.8, 32.0,

33.3, 35.6, 42.6, 45.6, 104.0, 104.2, 107.7, 122.6, 155.8, 164.7,

164.9. HRMS: calcd for C₁₈H₂₄N₇OS 386.1758 (MH^þ), found

386.1751.

cis-N-(1-(2,3-Difluorobenzyl)-4-methylpiperidin-3-yl)-N-methyl-

7H-pyrrolo[2,3-d]pyrimidin-4-amine (17c). To a stirred solution

of cis-N-methyl-N-(4-methylpiperidin-3-yl)-7H-pyrrolo[2,3-d]pyri-

midin-4-amine (50 mg/0.20 mmol) and 156 μL (1.43 mmol) of

2,3-difluorobenzaldehyde in 3 mL of anhydrous methanol was

added 20 mg (0.31 mmol) of sodium cyanoborohydride, and the

resulting mixture was stirred at room temperature for 3 h, at

which time a second portion of sodium cyanoborohydride was

added, and the resulting mixture was stirred for 72 h. The

reaction was quenched upon addition of 2 drops of 1 N NaOH,

and the reaction was concentrated under reduced pressure.

Water was added, the mixturewas extracted3 ꢀ withchloroform,

and the combined chloroform extracts were dried over MgSO₄,

filtered through Celite, and concentrated to dryness in vacuo.

The crude product was then purified by flash chromatography

(silica gel; 4% methanol in chloroform) affording 24 mg (32%)

of 17c as a white solid. LRMS: m/z 372.3 (MH^þ). ¹H NMR (400

MHz) (CDCl₃) δ: 0.93 (3 H, d, J = 7.2 Hz), 1.60-1.85 (4 H, m),

2.30 (1 H, br s), 2.62-2.83 (2 H, m), 2.87-2.96 (1 H, m), 3.59

(2 H, s), 3.61 (3 H, s), 5.22 (1 H, br s), 6.58 (1 H, d, J = 4.0 Hz),

6.95-7.08 (3 H, m), 7.13 (1 H, t, J = 5.2 Hz), 8.22 (1 H, s).

HRMS: calcdfor C₂₀H ₂₄N₅F₂(MH^þ) 372.1994, found372.1985.

cis-2-Hydroxy-1-(4-methyl-3-(methyl(7H-pyrrolo[2,3-d]pyri-

midin-4-yl)amino)piperidin-1-yl)ethanone (17d). To a stirred solution

of cis-N-methyl-N-(4-methylpiperidin-3-yl)-7H-pyrrolo[2,3-d]pyri-

midin-4-amine (200 mg/0.82 mmol) and 163 mg (0.98 mmol) of

2-(benzyloxy)acetic acid in 10 mL of dichloromethane at 0 ꢀC

was added 188 mg (0.98 mmol) of 1-ethyl-3-(3-dimethyamino-

propyl)carbodiimide followed by 26 mg (0.21 mmol) of 4-di-

methylaminopyridine, and the resulting mixture was brought to

room temperatureand stirred for 24 h, at which time the reaction

mixture was partitioned between dichloromethane and satu-

rated NaHCO₃. The dichloromethane layer was dried over

MgSO₄, filtered through Celite, and concentrated to dryness

in vacuo affording crude 2-(benzyloxy)-1-(4-methyl-3-(methyl-

(7H-pyrrolo[2,3-d]pyrimidin-4-yl)amino)piperidin-1-yl)ethanone,

which was purified by flash chromatography (silica gel; 4%

methanol in dichloromethane) affording 170 mg (53%) of this

(R)-2-((1S,3S,4R)-4-Methyl-3-(methyl(7H-pyrrolo[2,3-d]pyri-

midin-4-yl)amino)cyclohexyl)propyl Diethylcarbamate (16e). LRMS:

m/z 402.5 (MH^þ). ¹H NMR(400 MHz) (CDCl₃) δ:0.72-1.00 (6 H,

m), 1.18 (1 H, t, J = 1.2 Hz), 1.22 (3H, s), 1.23 (3H, s), 1.39-1.82 (6

H, m), 3.18-3.35 (7 H, m), 3.43-3.66 (1 H, m), 3.87-4.03 (1 H, m),

4.14 (1 H, dd, J = 5.6 Hz, J = 25 Hz), 0.67 (1 H, br s), 6.44 (1 H, br

s), 6.82 (1 h, t, J = 4.4 Hz), 6.97 (1 H, d, J = 3.6 Hz), 8.25 (1 H, t,

J = 6.2 Hz).

cis-N-Methyl-N-(4-methyl-1-(2,2,2-trifluoroethylsulfonyl)-

piperidin-3-yl)-7H-pyrrolo[2,3-d]pyrimidin-4-amine (17a). Com-

pound 19 was deprotected as described above; however, the inter-

mediate hydrogenolysis product cis-N-methyl-N-(4-methyl-

piperidin-3-yl)-7H-pyrrolo[2,3-d]pyrimidin-4-amine was isolated

by filtration through Celite and concentration under reduced

pressure. LRMS: m/z 246.0 (MH^þ). ¹H NMR (400 MHz)

(CD₃OD) δ: 1.00-1.13 (5 H, m), 1.80 (1 H, br s), 1.95 (1 H, br

s), 2.43 (1 H, br s), 3.07-3.18 (1 H, m), 3.38 (1 H, dd, J = 7.6 Hz,

J = 4.8 Hz), 3.41 (3 H, s), 3.52 (1 H, q, H = 7.2 Hz), 3.61 (1 H, q,

J = 8.8 Hz), 6.74 (1 H, br s), 7.11 (1 H, d, J = 3.2 Hz), 8.35 (1 H,

br s). ¹³C NMR (400 MHz) (CD₃OD) δ: 14.2, 29.4, 32.3, 38.5,

41.7, 43.9, 57.4, 104.3, 104.6, 123.2, 148.9, 149.5, 158.1. HRMS:

calcd for C₁₃H ₂₀N₅(MH^þ) 246.1713, found 246.1715. To this

material (30 mg/0.122 mmol) dissolved in 1.0 mL of dichloro-

methane containing 0.1 mL of pyridine was added 20 μL (0.181

mmol) of 2,2,2-trifluoroethanesulfonyl chloride, and the result-

ing mixture was stirred at room temperature for 16 h. The

reaction mixture was then partitioned between dichloromethane

and saturated sodium bicarbonate, the aqueous layer was back-

washed 2 ꢀ with dichloromethane, and the combined organic

layers dried over sodium sulfate and concentrated to dryness

in vacuo. The crude product was then purified by preparative

TLC (silica gel; 20:1 dichloromethane/methanol) affording 9 mg

(19%) of 17a as a white powder. LRMS: m/z 392.1 (MH^þ). ¹H

NMR (400 MHz) (CDCl₃) δ: 1.03 (3 H, d, J = 6.8 Hz), 1.78-1.96

(2 H, m), 2.35 (1 H, br s), 3.55 (1 H, dd, J = 14.8 Hz, J = 4.0 Hz),

3.58 (3 H, s), 3.61-3.75 (1 H, s), 3.75 (2 H, q, J = 9.6 Hz), 3.83

(1 H, dd, J = 8.2 Hz, J = 5.6 Hz), 5.39 (1 H, d, J = 4.0 Hz), 6.61

(1 H, d, J = 3.2 Hz), 7.11 (1 H, d, J = 3.6 Hz), 8.23 (1 H, s). ¹³C

NMR (400 MHz) (CDCl₃) δ: 15.2, 30.8, 32.0, 35.9, 43.9, 47.1,

52.4, 52.5, 103.3, 120.3, 121.6, 123.1, 145.8, 158.2. HRMS: calcd

for C₁₅H₂₁F₃N₅O₂S (MH^þ) 392.1363, found 392.1369.

1

pure intermediate. LRMS: m/z 394.2 (MH^þ). H NMR (400

MHz) (CD₃OD) (amide rotamers) δ: 1.06 (1.5 H, d, J = 7.2 Hz),

1.09 (1.5 H, d, J = 7.2 Hz), 1.61-1.74 (1 H, m), 1.80-1.91 (1 H,

m), 2.40-2.52 (1 H, m), 3.31 (1.5 H, s), 3.35 (1.5 H, s), 3.37-6.64

(1 H, m), 3.72-3.99 (2 H, m), 4.09 (0.5 H, 1/2 AB q, J = 12.8

Hz), 4.25 (0.5 H, 1/2 AB q, J = 13.2 Hz), 4.27 (2 H, s), 4.27 (1 H,

d, J = 4.8 Hz), 4.59 (2 H, s), 4.83-5.03 (1 H, m), 6.56 (0.5 H, d,

J = 3.2 Hz), 6.61 (0.5 H, d, J = 3.2 Hz), 7.00-7.15 (3 H, m),

7.25-7.41 (3 H, m), 8.07 (0.5 H, s), 8.08 (0.5 H, s). HRMS: calcd

cis-4-Methyl-3-(methyl(7H-pyrrolo[2,3-d]pyrimidin-4-yl)amino)-

N-(3-methylisothiazol-5-yl)piperidine-1-carboxamide (17b). To a

stirred solution of 3-methylisoxazol-5-amine (100 mg/1.00

mmol) in 2 mL of acetonitrile was added 103 mg (0.66 mmol)

of phenylchloroformate, and the resulting mixture was stirred at

8482 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24

Flanagan et al.

for C₂₂H₂₈N₅O₂(MH^þ) 394.2238, found 394.2239. To this

compound (1.5 g, 3.81 mmol) dissolved in 75 mL of ethanol was

added 2.6 g (1.9 mmol) of 20% Pd(OH)₂on carbon (50% water),

and the resulting mixture was agitated under a 50 psi atmosphere

of hydrogen gas for 48 h (catalyst was recharged once during this

time frame). The reaction mixture was filtered through Celite

and concentrated to dryness to produce 720 mg of crude

product, which was purified by flash chromatography (silica

gel; 10% methanol in dichloromethane) affording 480 mg (42%)

of compound 17d as a colorless glass. LRMS: m/z 304.2

(MH^þ). ¹H NMR (400 MHz) (CDCl₃) (amide rotomers) δ:

1.00-1.13 (3 H, m), 1.61-1.77 (1 H, m), 1.82-1.98 (1 H, m),

3.22-3.47 (2 H, m), 3.41 (3 H, s), 3.65 (1 H, q, J = 7.2 Hz),

3.71-3.85 (2 H, m), 4.09-4.28 (1 H, m), 4.22 (2 H, s),

5.05-5.22 (1 H, m), 6.52 (1 H, s), 7.05 (0.5 H, d, J = 3.2 Hz),

7.08 (0.5 H, d, J = 3.2 Hz), 8.25 (0.5 H, s), 8.27 (0.5 H, s), 10.80

(1 H, br s). ¹³C NMR (400 MHz) (CDCl₃) δ: 14.0, 14.9, 30.4,

31.4, 31.8, 32.2, 34.4, 35.1, 40.0, 40.6, 42.6, 44.9, 52.8, 53.0,

60.0, 102.3, 102.4, 103.0, 120.2, 120.4, 150.4, 150.6, 157.7,

157.8, 170.3, 170.4. HRMS: calcd for C₁₅H₂₂N₅O₂(MH^þ)

304.1767, found 304.1764.

cis-(5-Chloro-2,3-dihydrobenzofuran-2-yl)(4-methyl-3-(methyl-

(7H-pyrrolo[2,3-d]pyrimidin-4-yl)amino)piperidin-1-yl)methanone

(18a). To a stirred solution of cis-N-methyl-N-(4-methylpiperidin-

3-yl)-7H-pyrrolo[2,3-d]pyrimidin-4-amine (50 mg/0.20 mmol)

and 5-chloro-2,3-dihydrobenzofuran-2-carboxylic acid (50 mg/

0.25 mmol) in 6 mL of dimethylformamide cooled to 0 ꢀC was

added 40 mg (0.21 mmol) of 1-ethyl-3-(3-dimethylaminopro-

pyl)carbodiimide hydrochloride, and the resulting mixture was

warmed to room temperature and stirred for 16 h, at which point

the reaction mixture was partitioned between dichloromethane

and saturated NaHCO₃. The aqueous layer was extracted a

second time with dichloromethane, and the combined dichloro-

methane layers were dried over MgSO₄, filtered through Celite,

and concentrated to dryness in vacuo. The crude product was

then purified by preparative TLC (silica gel; 19:1 dichloro-

methane/methanol) affording 22 mg (26%) of 18a as a colorless

oil (∼3:1 mixture of diastereomers). LRMS: m/z 426.3 (MH^þ).

¹H NMR (400 MHz) (CDCl₃) δ: 0.85-1.00 (0.75 H, m),

1.05-1.18 (2.25 H, m), 1.83-2.13 (1 H, m), 2.43-2.71 (1 H,

m), 3.15-3.42 (3 H, m), 3.44 (0.75 H, s), 3.48 (2.25 H, s), 3.53-

4.27 (4 H, m), 5.21 (1 H, br s), 5.38 (1 H, dd, J = 7.6 Hz, J = 2.4

Hz), 5.46 (1 H, dd, J = 7.6 Hz, J = 2.4 Hz), 6.39-6.78 (2 H, m),

6.90-7.36 (3 H, m), 8.23-8.35 (1 H, m).

(MH^þ). ¹H NMR (400 MHz) (CD₃OD) δ: 0.81-0.93 (1 H, m),

1.02-1.15 (2 H, m), 1.60-1.75 (1 H, m), 1.82-1.94 (1 H, m),

2.00-2.52 (5 H, m), 3.03-3.12 (1 H, m), 3.27 (2 H, s), 3.39 (1 H,

s), 3.46-3.63 (2 H, m), 3.68-3.73 (1 H, m), 3.80-4.02 (2 H, m),

4.95 (1 H, br s), 6.59-6.60 (1 H, m), 7.05-7.07 (1 H, m), 8.06 (0.7

H, s), 8.09 (0.3 H, s).

(4-Methyl-3-(methyl(7H-pyrrolo[2,3-d]pyrimidin-4-yl)amino)-

piperidin-1-yl)(thiophen-3-yl)methanone (18e). LRMS: m/z 356.3

1

(MH^þ). H NMR (400 MHz) (CDCl₃) δ: 0.95-1.15 (4 H, m),

1.22-1.34 (1 H, m), 1.68 (2 H, br s), 3.40 (3 H, s), 3.63 (2 H, br s),

3.85 (2 H, t, J = 8.8 Hz), 5.13 (1 H, br s), 6.54 (1 H, d, J =

3.6 Hz), 7.05 (1 H, d, J = 4.0 Hz), 7.19 (1 H, d, J = 5.6 Hz), 7.30

(1 H, br s), 7.53 (1 H, s), 8.28 (1 H, s).

(4-Methyl-3-(methyl(7H-pyrrolo[2,3-d]pyrimidin-4-yl)amino)-

piperidin-1-yl)(tetrahydrofuran-3-yl)methanone (18f). LRMS: m/z

344.2 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ: 1.00-1.04 (3 H,

m), 1.78 (1 H, br s), 1.75-1.80 (1 H, m), 1.81-2.51 (3 H, m),

3.10-3.34 (2 H, m), 3.23 (3 H, s), 3.40-3.91 (6 H, m), 4.05 (1 H, sp

1/2 ABq), 5.00 (1 H, 1/2 ABq, J = 60 Hz), 6.44 (1 H, d, J = 10.4

Hz), 7.00 (1 H, dd, J = 16.0 Hz, J = 3.2 Hz), 8.21 (0.5 H, s), 8.25

(0.5 H, s), 11.58 (0.5 H, s), 11.80 (0.5 H, s). ¹³C NMR (400 MHz)

(CDCl₃) δ: 13.9, 14.0, 14.7, 14.9, 30.0, 30.1, 30.3, 30.7, 31.8, 31.9,

32.0, 32.2, 32.3, 34.3, 35.1, 35.2, 39.4, 39.6, 41.2, 41.3, 41.5, 41.5,

42.1, 42.4, 46.7, 46.9, 53.3, 53.7, 68.4, 68.5, 70.5, 70.6, 70.7, 102.0,

102.1, 103.0, 103.1, 120.2, 120.6, 150.6, 151.9, 152.0, 157.7, 157.8,

171.3, 171.7. HRMS: calcd for C₁₈H₂₆N₅O₂(MH^þ) 344.2081,

found 344.2075.

cis-Cyclopentyl(4-methyl-3-(methyl(7H-pyrrolo[2,3-d]pyrimidin-

4-yl)amino)piperidin-1-yl)methanone (18g). LRMS: m/z 342.2

1

(MH^þ). H NMR (400 MHz) (CDCl₃) δ: 0.95-1.05 (3 H, m),

1.38-1.88 (7 H, m), 2.37-2.38 (2 H, m), 2.79-2.89 (2 H, m),

3.29 (1.5 H, s), 3.35 (1.5 H, s), 3.52-3.68 (4 H, m), 3.82 (0.5 H, q,

J = 7.6 Hz), 4.09 (0.5 H, d, J = 10.4 Hz), 4.93 (0.5 H, br s), 5.05

(0.5 H, br s), 6.44 (0.5 H, s), 6.46 (0.5 H, s), 6.98 (0.5 H, d, J = 2.8

Hz), 7.03 (0.5 H, d, J = 2.8 Hz), 8.21 (0.5 H, s), 8.25 (0.5 H, s),

11.26 (0.5 H, br s), 11.47 (0.5 H, br s). ¹³C NMR (400 MHz)

(CDCl₃) δ: 14.2, 14.9, 16.0, 30.1, 30.1, 30.4, 30.7, 32.0, 32.1, 32.5,

34.3, 35.1, 39.3, 41.1, 41.4, 42.1, 42.5, 46.9, 53.3, 102.1, 102.3,

102.9, 120.0, 120.4, 150.7, 151.9, 157.8, 174.6, 175.0. HRMS:

calcd for C₁₉H₂₈N₅O (MH^þ) 342.2288, found 342.2280.

1-(4-Methyl-3-(methyl(7H-pyrrolo[2,3-d]pyrimidin-4-yl)amino)-

piperidin-1-yl)-2-(methylsulfonyl)ethanone (18h). LCMS: m/z

1

366.2 (MH^þ). H NMR (400 MHz) (CD₃OD) δ: 1.05 (2 H, d,

J = 6.8 Hz), 1.11 (1 H, d, J = 7.2 Hz), 1.61-1.95 (2 H, m),

2.06-2.15 (1 H, m), 1.40-1.53 (1 H, m), 3.11 (2 H, s), 3.13 (1 H,

s), 3.59-3.68 (1 H, m), 3.80-4.09 (2 H, m), 4.35-4.63 (2 H, m),

4.99-5.12 (1 H, m), 6.62 (1 H, d, J = 3.6 Hz), 7.08-7.89 (1 H,

m), 8.10 (0.6 H, s), 8.15 (0.4 H, s).

Compounds 18b through 18m were prepared in an analogous

manner to that described for 17a. The NMR data for these

indicate the presence of amide rotamers.

1-Benzyl-3-(4-methyl-3-(methyl(7H-pyrrolo[2,3-d]pyrimidin-

4-yl)amino)piperidine-1-carbonyl)pyrrolidin-2-one (18b). Liquid

chromatography mass spectrometry (LCMS): m/z 447.4 (MH^þ).

¹H NMR (400 MHz) (CD₃OD) δ: 0.81-1.26 (4 H, m), 1.54-1.65

(1 H, m), 1.73-1.92 (1 H, m), 2.40 (1 h, br s), 2.52-2.80 (2 H, m),

3.20 (1.5 H, s), 3.25 (1.5 H, s), 3.29-3.98 (6 H, m), 4.31-4.52

(2 H, m), 6.54-6.64 (1 h, m), 7.04-7.07 (1 H, m), 7.15-7.33

(5 H, m), 8.04-8.06 (1 H, m).

cis-2-Cyclopentyl-1-(4-methyl-3-(methyl(7H-pyrrolo[2,3-d]pyri-

midin-4-yl)amino)piperidin-1-yl)ethanone (18c). Amide rotamers.

LRMS: m/z 356.2 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ: 1.01

(3 H, s), 0.90-1.18 (2 H, m), 1.37-1.89 (7 H, m), 1.67 (2 H, s),

2.15-2.50(4H, m), 3.30(1.5H, s), 3.35(1.5H, s), 3.47(1H, s), 3.77

(2 H, br s), 3.78 (0.5 H, t, J = 6.8 Hz), 4.07 (0.5 H, d, J = 11.6 Hz),

4.95 (0.5 H, s), 5.05 (0.5 H, s), 6.46 (1 H, s), 6.97 (1 H, d, J =

14.0 Hz), 8.22 (1 H, d, J = 11.2 Hz), 10.15 (0.5 H, br s), 10.22 (0.5

H, br s). ¹³C NMR (400 MHz) (CDCl₃) δ: 14.1, 14.9, 24.9, 25.0,

30.7, 31.9, 31.9, 32.4, 32.7, 32.8, 34.3, 35.0, 36.6, 36.8, 39.0, 39.4,

39.5, 41.7, 42.7, 47.1, 53.3, 102.3, 102.4, 102.8, 119.7, 120.1, 150.8,

150.9, 151.7, 152.0, 157.8, 157.9, 171.4, 171.6. HRMS: calcd for

C₂₀H₃₀N₅O (MH^þ) 356.2445, found 356.2438.

cis-2-Cyclopropyl-1-(4-methyl-3-(methyl(7H-pyrrolo[2,3-d]pyri-

midin-4-yl)amino)piperidin-1-yl)ethanone (18i). LRMS: m/z 328.2

(MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ: -0.01-0.04 (2 H, m),

0.24-0.41 (2 H, m), 0.78-0.94 (3 H, m), 1.48 (1 H, br s), 1.59-1.76

(1 H, m), 1.94-2.38 (4 H, m), 3.16 (1.5 H, s), 3.20 (1.5 H, s), 3.31

(1 H, d, J = 4.8 Hz), 3.39-3.58 (2 H, m), 3.59-3.68 (0.5 H, m),

3.93 (0.5 H, d, J = 9.6 Hz), 4.82 (0.5 H, br s), 4.92 (0.5 H, br s), 6.31

(0.5 H, s), 6.32 (0.5 H, s), 6.85 (0.5 H, d, J = 3.2 Hz), 6.89 (0.5 H, d,

J = 2.8 Hz), 8.07 (0.5 H, s), 8.10 (0.5 H, s), 11.17 (0.5 H, br s), 11.37

(0.5 H, br s). ¹³C NMR (400 MHz) (CDCl₃) δ: 4.5, 4.7, 7.3, 7.4,

14.1, 14.9, 30.6, 31.9, 32.0, 32.4, 34.3, 35.0, 38.7, 38.9, 39.1, 41.8,

42.8, 47.1, 53.2, 53.5, 102.1, 102.3, 102.9, 120.1, 120.4, 150.7, 151.9,

152.0, 157.8, 157.9, 171.3, 171.5. HRMS: calcd for C₁₈H₂₆N₅O

(MH^þ) 328.2132, found 328.2124.

cis-3,3,3-Trifluoro-1-(4-methyl-3-(methyl(7H-pyrrolo[2,3-d]-

pyrimidin-4-yl)amino)piperidin-1-yl)propan-1-one (18j). LRMS:

m/z 356.2 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ TMS: 0.97

(1.5 H, d, J = 7.2 Hz), 1.08 (1.5 H, d, J = 7.2 Hz), 1.60-1.70

(1 H, m), 1.79-1.94 (1 H, m), 2.37-2.50 (1 H, m), 3.14-3.37

(2 H, m), 3.20 (1.5 H, s), 3.35 (1.5 H, s), 3.40-3.62 (2 H, m), 3.72

(1 H, q, J = 12.8 Hz), 3.70-3.86 (0.5 H, m), 4.06 (0.5 H, d, J =

8.8 Hz), 5.05 (1 H, br s), 6.46 (0.5 H, d, J = 3.6 Hz), 6.47 (0.5 H,

3-(4-Methyl-3-(methyl(7H-pyrrolo[2,3-d]pyrimidin-4-yl)amino)-

piperidine-1-carbonyl)cyclopentanone (18d). LCMS: m/z 356.4

Article

Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24 8483

d, J = 5.2 Hz), 6.99 (0.5 H, d, J = 3.2 Hz), 7.04 (0.5 H, d, J =

3.2 Hz), 8.21 (0.5 H, s), 8.24 (0.5 H, s), 10.96 (1 H, br s), 11.15 (0.5

H, br s). ¹³C NMR (400 MHz) (CDCl₃) δ: 13.1, 13.6, 29.5, 29.9,

30.5, 30.8, 30.9, 33.3, 34.0, 36.9, 37.2, 37.5, 38.3, 41.4, 42.8, 46.3,

49.7, 51.7, 52.5, 101.0, 101.3, 101.9, 102.0, 119.1, 119.5, 121.8,

124.5, 149.6, 149.7, 150.9, 151.0, 156.7, 160.6, 161.0, 206.1.

HRMS: calcd for C₁₆H₂₁F₃N₅O (MH^þ) 356.1693, found 356.1689.

2,2-Dimethyl-3-(4-methyl-3-(methyl(7H-pyrrolo[2,3-d]pyri-

midin-4-yl)amino)piperidin-1-yl)-3-oxopropanenitrile (18k). LRMS:

m/z 341.3 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ: 1.09 (3 H, d,

J = 7.2 Hz), 1.63 (6 H, s), 1.79 (1 H, br s), 1.99-2.16 (1 H, m),

2.51-2.57 (1 H, m), 3.28-3.42 (1 H, m), 3.42 (3 H, s), 3.88 (2 H, br

s), 4.07 (1 H, d, J = 8.4 Hz), 5.15 (1 H, br s), 6.57 (1 H, d, J = 3.6

Hz), 7.09 (1 H, d, J = 3.2 Hz), 8.28 (1 H, s).

2-Methyl-3-(4-methyl-3-(methyl(7H-pyrrolo[2,3-d]pyrimidin-

4-yl)amino)piperidin-1-yl)-3-oxopropanenitrile (18l). LCMS: m/z

327.3 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ: 1.03-1.11 (3 H,

m), 1.51-1.56 (3 H, m), 1.61 (1 H, br s), 1.78-1.96 (1 H, m), 2.08

(1 H, br s), 3.34 (1 H, s), 3.39 (2 H, s), 3.40-3.59 (1 H, m),

3.60-3.81 (2 H, m), 3.81-4.10 1.5 H, m), 4.21 (1/2 H, d, J = 13.2

Hz), 5.03 (1/2 H, br s), 5.10 (1/2 H, br s), 5.20 (1 H, br s), 6.51

(1/3 H, d, J = 3.2 Hz), 6.55 (2/3 H, d, J = 3.6 Hz), 7.05 (1 H, d,

J = 3.6 Hz), 8.23 (1/3 H, s), 8.25 (2/3 H, s).

147.4, 157.6, 163.2. HRMS: calcd for C₁₈H₂₇N₆O₂(MH^þ)

359.2190, found 359.2188.

Acknowledgment. This research was sponsored by Pfizer

Inc. The authors gratefully acknowledge Dr. Gary Chan MD

and Dr. Samuel Zwillich MD for helpful conversations and

for providingsummaries of clinicalresults for1 citrate relating

to the prevention of renal allograft rejection and treatment of

rheumatoid arthritis, respectively. The authors also acknowl-

edge Daniel Virtue and Melissa Rockwell for chiral separa-

tions performed, Matthew Teague for high-resolution mass

spectra, and Jon Bordner and Ivan Samardjiev for the single

crystal X-ray structure of compound 20. The support pro-

vided by Martin Goulding at Complete Medical Communica-

tions, funded by Pfizer Inc., consisted solely of manuscript

editing and formatting, and no contribution was made to

editorial content.

Supporting Information Available: Additional HSA com-

pounds (6c, S1-S63), with initial JAK3% inhibition data. This

material is available free of charge via the Internet at http://

pubs.acs.org.

cis-3-(4-Methyl-3-(methyl(7H-pyrrolo[2,3-d]pyrimidin-4-yl)-

amino)piperidin-1-yl)-3-oxopropanenitrile (18m). LRMS: m/z

313.3 (MH^þ). ¹H NMR (400 MHz) (CDCl₃) δ: 1.08 (1.5 H, d,

J = 7.2 Hz), 1.10 (1.5 H, d, J = 7.2 Hz), 1.63-2.09 (2 H, m),

2.42-2.58 (1 H, m), 3.26-3.68 (2 H, m), 3.41 (1.5 H, s), 3.46 (1.5

H, s), 3.56 (2 H, s), 3.70-3.92 (1.5 H, m), 4.09 (0.5 H, dd, J = 8.8

Hz, J = 4.4 Hz), 5.16 (1 H, br s), 6.57 (1 H, t, J = 3.6 Hz), 7.10

(1 H, t, J = 3.6 Hz), 8.26 (1 H, s). HRMS: calcd for C₁₆H₂₁N₆O

(MH^þ) 313.1771, found 313.1776.

References

(1) Stock, P. G. The year in review—ATC 2002. Am. J. Transplant.

2003, 3, 373–380.

(2) Hong, J. C.; Kahan, B. D. Immunosuppressive agents in organ

transplantation: past, present, and future. Semin. Nephrol. 2000,

20, 108–125.

(3) Darnell, J. E. J.; Kerr, I. M.; Stark, G. R. Jak-STAT pathways and

transcriptional activation in response to IFNs and other extra-

cellular signaling proteins. Science 1994, 264, 1415–1421.

(4) Ghoreschi, K.; Laurence, A.; O’Shea, J. J. Selectivity and ther-

apeutic inhibition of kinases: to be or not to be? Nat. Immunol.

2009, 10, 356–360.

((S)-3-Hydroxypyrrolidin-1-yl)((3R,4R)-4-methyl-3-(methyl-

(7H-pyrrolo[2,3-d]pyrimidin-4-yl)amino)piperidin-1-yl)methanone

(20). Toa stirred solutionof (S)-pyrrolidin-3-ol (1.0 g/11.5 mmol)

dissolved in 15 mL of dichloromethane and cooled to 0 ꢀC was

added 2.3 g (11.5 mmol) of 4-nitrophenylchloroformate followed

by 1.6 mL (11.5 mmol) of triethylamine, and the resulting mixture

was stirred at 0 ꢀC for 1.5 h. The reaction mixture was then parti-

tioned between water and chloroform, the aqueous layer was

extracted 3 ꢀ 50 mL with chloroform, and the combined chloro-

form layers were dried over MgSO4, filtered through Celite, and

concentrated to dryness in vacuo affording 2.9 g (quant) of the

intermediate (S)-4-nitrophenyl 3-hydroxypyrrolidine-1-carboxy-

late as a yellow solid. LRMS: m/z 253.2 (MH^þ). ¹H NMR

(400 MHz) (CDCl₃) δ: 1.97-2.16 (2 H, m), 3.48-3.80 (4 H, m),

4.56-4.60 (1 H, m), 7.33 (2 H, dd, J = 6.8 Hz, J = 2.4 Hz), 8.25

(5) Ghoreschi, K.; Laurence, A.; O’Shea, J. J. Janus kinases in immune

cell signaling. Immunol. Rev. 2009, 228, 273–287.

(6) Johnston, J. A.; Bacon, C. M.; Riedy, M. C.; O’Shea, J. J. Signaling

by IL-2 and related cytokines: JAKs, STATs and relationship to

immunodeficiency. J. Leukocyte Biol. 1996, 60, 441–452.

(7) Oakes, S. A.; Candotti, F.; Johnston, J. A.; Chen, Y. Q.; Ryan, J. J.;

Taylor, N.; Liu, X.; Henninghausen, L.; Notarangelo, L. D.; Paul,

W. E.; Blaese, R. M.; O’Shea, J. J. Signaling via IL-2 and IL-4 in

JAK3-deficient severe combined immunodeficiency lymphocytes:

JAK3-dependent and independent pathways. Immunity 1996, 5,

605–615.

(8) Chen, M.; Cheng, A.; Chen, Y. Q.; Hymel, A.; Hanson, E. P.;

Kimmel, L.; Minami, Y.; Taniguchi, T.; Changelian, P. S.; O’Shea,

J. J. The amino terminus of JAK3 is necessary and sufficient for

binding to the common gamma chain and confers the ability to

transmit interleukin 2-mediated signals. Proc. Natl. Acad. Sci. U.S.A.

1997, 94, 6910–6915.

(2 H, d, J = 8.8 Hz). HRMS: calcd for C₁₁H₁₃N O₅(MH^þ)

2

253.0819, found 253.0817. Separately, N-methyl-N-((3R,4R)-4-

methylpiperidin-3-yl)-7H-pyrrolo[2,3-d]pyrimidin-4-amine (50 mg/

0.20 mmol), prepared from 19a as described above, dissolved in

5 mL of tert-butanol, was charged to a pressure tube along with

76 mg (0.30 mmol) of (S)-4-nitrophenyl 3-hydroxypyrrolidine-

1-carboxylate and 0.5 mL of triethylamine, and the sealed tube

then heated at 100 ꢀC for 48 h. After being cooled to room tem-

perature the reaction mixture was partitioned between chloro-

form and saturated NaHCO ₃. The aqueous layer was extracted

3 ꢀ 30 mL with chloroform, and the combined chloroform

layers were dried over Na₂SO₄, filtered through Celite, and

concentrated to dryness in vacuo. The crude product was then

purified by flash chromatography (silica gel; 10-15% methanol

in chloroform). The product was then triturated with chloro-

form/hexanes to yield 24 mg (33%) of 20 as an off-white solid

(9) Hofmann, S. R.; Ettinger, R.; Zhou, Y. J.; Gadina, M.; Lipsky, P.;

Siegel, R.; Candotti, F.; O’Shea, J. J. Cytokines and their role in

lymphoid development, differentiation and homeostasis. Curr.

Opin. Allergy Clin. Immunol. 2002, 2, 495–506.

(10) Macchi, P.; Villa, A.; Giliani, S.; Sacco, M. G.; Frattini, A.; Porta,

F.; Ugazio, A. G.; Johnston, J. A.; Candotti, F.; O’Shea, J. J.

Mutations of Jak-3 gene in patients with autosomal severe com-

bined immune deficiency (SCID). Nature 1995, 377, 65–68.

(11) Noguchi, M.; Rosenblatt, H. M.; Filipovich, A. H.; Adelstein, S.;

Modi, W. S.; McBride, O. W.; Leonard, W. J. Interleukin-2

receptor gamma chain mutation results in X-linked severe com-

bined immunodeficiency in humans. Cell 1993, 73, 147–157.

(12) Russell, S. M.; Tayebi, N.; Nakajima, H.; Riedy, M. C.; Roberts,

J. L.; Aman, M. J.; Migone, T. S.; Noguchi, M.; Markert, M. L.;

Buckley, R. H.; O’Shea, J. J.; Leonard, W. J. Mutation of Jak3 in a

patient with SCID: essential role of Jak3 in lymphoid development.

Science 1995, 270, 797–800.

1

after drying. LRMS: m/z 359.2 (MH^þ). H NMR (400 MHz)

(CDCl₃) δ: 1.03 (3 H, d, J = 6.8 Hz), 1.60 (1 H, d, J = 10.8 Hz),

1.93 (2 H, br s), 2.45 (1 H, t, J = 5.2 Hz), 3.22 (1 H, t, J = 10.8 Hz),

3.36 (3 H, s), 3.37-3.71 (8 H, m), 4.44 (1 H, s), 4.94 (1 H, br s),

6.46 (1 H, s), 7.04 (1 H, d, J = 3.6 Hz), 8.20 (1 H, s), 11.81 (1 H,

br s). ¹³C NMR (400 MHz) (CDCl3) δ: 13.8, 31.2, 31.5, 33.9,

34.9, 43.1, 45.5, 46.2, 54.8, 56.5, 70.4, 102.8, 102.9, 121.1, 147.3,

(13) Pesu, M.; Candotti, F.; Husa, M.; Hofmann, S. R.; Notarangelo,

L. D.; O’Shea, J. J. Jak3, severe combined immunodeficiency, and a

new class of immunosuppressive drugs. Immunol. Rev. 2005, 203,

127–142.

(14) Changelian, P. S.; Flanagan, M. E.; Ball, D. J.; Kent, C. R.;

Magnuson, K. S.; Martin, W. H.; Rizzuti, B. J.; Sawyer, P. S.;

8484 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 24

Flanagan et al.

Perry, B. D.; Brissette, W. H.; McCurdy, S. P.; Kudlacz, E. M.;

Conklyn, M. J.; Elliott, E. A.; Koslov, E. R.; Fisher, M. B.;

Strelevitz, T. J.; Yoon, K.; Whipple, D. A.; Sun, J.; Munchhof,

M. J.; Doty, J. L.; Casavant, J. M.; Blumenkopf, T. A.; Hines, M.;

Brown, M. F.; Lillie, B. M.; Subramanyam, C.; Shang-Poa, C.;

Milici, A. J.; Beckius, G. E.; Moyer, J. D.; Su, C.; Woodworth,

T. G.; Gaweco, A. S.; Beals, C. R.; Littman, B. H.; Fisher, D. A.;

Smith, J. F.; Zagouras, P.; Magna, H. A.; Saltarelli, M. J.; Johnson,

K. S.; Nelms, L. F.; Des Etages, S. G.; Hayes, L. S.; Kawabata,

T. T.; Finco-Kent, D.; Baker, D. L.; Larson, M.; Si, M. S.;

Paniagua, R.; Higgins, J.; Holm, B.; Reitz, B.; Zhou, Y. J.; Morris,

R. E.; O’Shea, J. J.; Borie, D. C. Prevention of organ allograft

rejection by a specific Janus kinase 3 inhibitor. Science 2003, 302,

875–878.

Assessment of three human in vitro systems in the generation of

major human excretory and circulating metabolites. Chem. Res.

Toxicol. 2009, 22, 357–368.

(29) Karaman, M. W.; Herrgard, S.; Treiber, D. K.; Gallant, P.;

Atteridge, C. E.; Campbell, B. T.; Chan, K. W.; Ciceri, P.; Davis,

M. I.; Edeen, P. T.; Faraoni, R.; Flloyd, M.; Hunt, J. P.; Lockhart,

D. J.; Milanov, Z. V.; Morrison, M. J.; Pallares, G.; Patel, H. K.;

Pritchard, S.; Wodicka, L. M.; Zarrinkar, P. P. A qualitive analysis

of kinase inhibitor selectivity. Nat. Biotechnol. 2008, 26, 127–132.

(30) Pfizer Inc. Data on file. 2009.

(31) Borie, D. C.; Larson, M. J.; Flores, M. G.; Campbell, A.; Rousvoal,

G.; Zhang, S.; Higgins, J. P.; Ball, D. J.; Kudlacz, E. M.; Brissette,

W. H.; Elliott, E. A.; Reitz, B. A.; Changelian, P. S. Combined use

of the JAK3 inhibitor CP-690,550 with mycophenolate mofetil to

prevent kidney allograft rejection in nonhuman primates. Trans-

plantation 2005, 80, 1756–1764.

(15) West, K. CP-690550, a JAK3 inhibitor as an immunosuppressant

for the treatment of rheumatoid arthritis, transplant rejection,

psoriasis and other immune-mediated disorders. Curr. Opin. Invest.

Drugs 2009, 10, 491–504.

(32) Rousvoal, G.; Si, M. S.; Lau, M.; Zhang, S.; Berry, G. J.; Flores,

M. G.; Changelian, P. S.; Reitz, B. A.; Borie, D. C. Janus kinase 3

inhibition with CP-690,550 prevents allograft vasculopathy. Trans-

plant Int. 2006, 19, 1014–1021.

(16) Sorbera, L. A.; Serradell, N.; Bolos, J.; Rosa, E.; Bozzo, J. CP-

690,550. Drugs Future 2007, 32, 674–680.

(17) Wrobleski, S. T.; Pitts, W. J. Chapter 12 Advances in the discovery

of small molecule JAK3 inhibitors. Annu. Rep. Med. Chem. 2009,

44, 247–264.

(33) Busque, S.; Leventhal, J.; Brennan, D.; Steinberg, S.; Klintmalm,

G.; Shah, T.; Mulgaonkar, S.; Bromberg, J.; Vincenti, F.; Hariharan,

S.; Slakey, D.; Peddi, V. R.; Fisher, R. A.; Lawendy, N.; Wang, C.;

Chan, G. Calcineurin-inhibitor-free immunosuppression based on

the JAK inhibitor CP690,550: a pilot study in de novo kidney

allograft recipients. Am. J. Transplant. 2009, 9, 1936–1945.

(34) Milici, A. J.; Kudlacz, E. M.; Audoly, L.; Zwillich, S.; Changelian, P.

Cartilage preservation by inhibition of Janus kinase 3 in two rodent

models of rheumatoid arthritis. Arthritis Res. Ther. 2008, 10, R14.

(35) Kremer, J. M.; Bloom, B. J.; Breedveld, F. C.; Coombs, J. H.;

Fletcher, M. P.; Gruben, D.; Krishnaswami, S.; Burgos-Vargas, R.;

Wilkinson, B.; Zerbini, C. A. F.; Zwillich, S. H. The safety and

efficacy of a JAK inhibitor in patients with active rheumatoid

arthritis: results of a double-blind, placebo-controlled phase IIa

trial of three dose levels of CP-690,550 versus placebo. Arthritis

Rheum. 2009, 60, 1895–1905.

(36) Kremer, J. M.; Cohen, S.; Wilkinson, B.; Connell, C; French, J;

Gomez Reino, J; Gruben, D.; Kanik, K; Krishnaswami, S.;

Pascual-Ramos, V; Wallenstein, G; Zwillich, S. The oral JAK

inhibitor CP-690,550 (CP) in combination with methotrexate

(MTX) is efficacious, safe and well tolerated in patients with active

rheumatoid arthritis (RA) with an inadequate response to metho-

trexate alone. Arthritis Rheum. 2008, 58, 4030.

(18) Kudlacz, E.; Perry, B.; Sawyer, P.; Conklyn, M.; McCurdy, S.;

Brissette, W.; Flanagan, A.; Changelian, P. The novel JAK-3

inhibitor CP-690550 is a potent immunosuppressive agent in

various murine models. Am. J. Transplant. 2004, 4, 51–57.

(19) Meyer, D. M.; Jesson, M. I.; Li, W.; Elrick, M. M.; Funckes-

Shippy, C. L.; Warner, J. D.; Gross, C. J.; Dowty, M. E.; Ramaiah,

S. K.; Hirsch, J. L.; Saabye, M. J.; Barks, J. L.; Kishore, N.; Morris,

D. L. Anti-inflammatory activity and neutrophil reductions

mediated by JAK1 and JAK3 inhibition in the rat AIA model.

J. Inflammation (London) 2010, 7, 41.

(20) Simonsen, J. The Terpenes; Cambridge University Press: Cambridge,

1953; pp 394-408.

(21) Brown Ripin, D. H.; Abele, S.; Cai, W.; Blumenkopf, T.; Casavant,

J. M.; Doty, J. L.; Flanagan, M.; Koecher, C.; Laue, K. W.;

McCarthy, K.; Meltz, C.; Munchhof, M. J.; Pouwer, K.; Sha, B.;

Sun, J.; Teixeira, J.; Vries, T.; Whipple, D. A.; Wilcox, G. Devel-

opment of a scaleable route for the production of cis-N-benzyl-3-

methylamino-4-methylpiperidine. Org. Process Res. Dev. 2003, 7,

115–120.

(22) Jiang, J. K.; Ghoreschi, K.; Deflorian, F.; Chen, Z.; Perreira, M.;

Pesu, M.; Smith, J.; Nguyen, D. T.; Liu, E. H.; Leister, W.;

Costanzi, S.; O’Shea, J. J.; Thomas, C. J. Examining the chirality,

conformation and selective kinase inhibition of 3-((3R,4R)-4-

methyl-3-(methyl(7H-pyrrolo[2,3-d]pyrimidin-4-yl)amino)piperidin-

1-yl)-3-oxopropanenitrile (CP-690,550). J. Med. Chem. 2008, 51,

8012–8018.

(37) Kremer, J. M.; Cohen, S.; Wilkinson, B.; Gruben, D.; Wallenstein,

G; Kanik, K; Zwillich, S. H. Safety and efficacy after 24 week (WK)

dosing of the oral JAK inhibitor CP-690,550 (CP) in combination

with methotrexate (MTX) in patients (PTS) with active rheumatoid

arthritis (RA). Arthritis Rheum. 2009, 60 (Suppl. 10), S719, 1925

(abstract).

(23) Williams, N. K.; Bamert, R. S.; Patel, O.; Wang, C.; Walden, P. M.;

Wilks, A. F.; Fantino, E.; Rossjohn, J.; Lucet, I. S. Dissecting

specificity in the Janus kinases: the structures of JAK-specific

inhibitors complexed to the JAK1 and JAK2 protein tyrosine

kinase domains. J. Mol. Biol. 2009, 387, 219–232.

(24) Chrencik, J. E.; Patny, A.; Leung, I. K.; Korniski, B.; Emmons,

T. L.; Hall, T.; Weinberg, R. A.; Gormley, J. A.; Williams, J. M.;

Day, J. E.; Hirsch, J. L.; Kiefer, J. R.; Leone, J. W.; Fischer, H. D.;

Sommers, C. D.; Huang, H. C.; Jacobsen, E. J.; Tenbrink, R. E.;

Tomasselli, A. G.; Benson, T. E. Structural and thermodynamic

characterization of the TYK2 and JAK3 kinase domains in com-

plex with CP-690550 and CMP-6. J. Mol. Biol. 2010, 400, 413–433.

(25) World Intellectual Property Organization. (WO/2003/048162)

Novel Crystalline Compound. http://www.wipo.int/pctdb/en/wo.

jsp?wo=2003048162&IA=IB2002004948&DISPLAY=DOCS.

Accessed 3 August 2010.

(38) Kanik, K.; Fleischmann, R.; Cutolo, M.; Genovese, M.; Lee, E.;

Sadis, S.; Connell, C.; Gruben, D.; Krishnaswami, S.; Wallenstein,

G.; Wilkinson, B.; Zwillich, S. Phase 2b dose ranging monotherapy

study of the oral JAK inhibitor CP-690,550 (CP) or adalimumab

(ADA) vs placebo (PBO) in patients (pts) with active rheumatoid

arthritis (RA) with an inadequate response to DMARDs. Ann.

Rheum. Dis. 2009, 68 (Suppl. 3), 123, OP-0159 (abstract).

(39) Fleischmann, R. M.; Genovese, M. C.; Gruben, D.; Kanik, K;

Wallenstein, G; Wilkinson, B.; Zwillich, S. H. Safety and efficacy

after 24 week (wk) dosing of the oral JAK inhibitor CP-690,550

(CP) as monotherapy in patients (pts) with active rheumatoid

arthritis (RA). Arthritis Rheum. 2009, 60 (Suppl. 10), S718, 1924

(abstract).

(40) Kalgutkar, A. S.; Choo, E.; Taylor, T. J.; Marfat, A. Disposition of

CP-671,305, a selective phosphodiesterase 4 inhibitor in preclinical

species. Xenobiotica 2004, 34, 755–770.

(26) van Gurp, E.; Weimar, W.; Gaston, R.; Brennan, D.; Mendez,

R.; Prusch, J.; Swan, S.; Pescovitz, M. D.; Ni, G.; Wang, C.;

Krishnaswami, S.; Chow, V.; Chan, G. Phase 1 dose-escalation

study of CP-690-550 in stable renal allograft recipients: preliminary

findings of safety, tolerability, effects on lymphocyte subsets and

pharmacokinetics. Am. J. Transplant. 2008, 8, 1711–1718.

(27) Lawendy, N.; Krishnaswami, S.; Wang, R.; Gruben, D.; Cannon,

C.; Swan, S.; Chan, G. Effect of CP-690,550, an orally active janus

kinase inhibitor, on renal function in healthy adult volunteers.

J. Clin. Pharmacol. 2009, 49, 423–429.

(41) Obach, R. S. Inhibition of human cytochrome P450 enzymes by

constituents of St. John’s wort, an herbal preparation used in the

treatment of depression. J. Pharmacol. Exp. Ther. 2000, 294, 88–95.

(42) Kalgutkar, A. S.; Hatch, H. L.; Kosea, F.; Nguyen, H. T.; Choo,

E. F.; McClure, K. F.; Taylor, T. J.; Henne, K. R.; Kuperman,

A. V.; Dombroski, M. A.; Letavic, M. A. Preclinical pharmaco-

kinetics and metabolism of 6-(4-(2,5-difluorophenyl)oxazol-5-yl)-

3-isopropyl-[1,2,4]-triazolo[4,3-a]pyridine, a novel and selective

p38alpha inhibitor: identification of an active metabolite in pre-

clinical species and human liver microsomes. Biopharm. Drug Dispos.

2006, 27, 371–386.

(28) Dalvie, D.; Obach, R. S.; Kang, P.; Prakash, C.; Loi, C. M.; Hurst,

S.; Nedderman, A.; Goulet, L.; Smith, E.; Bu, H. Z.; Smith, D. A.

(43) Davoll, J. Pyrrolo[2,3-d]pyrimidines. J. Chem. Soc. 1960, 82, 131–138.

Article Doi

Discovery of CP-690,550: A potent and selective janus kinase (JAK) inhibitor for the treatment of autoimmune diseases and organ transplant rejection

DOI: 10.1021/jm1004286

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jm1004286

Full text of DOI:10.1021/jm1004286

Products guided by the article

R&D Labs maybe for 55794-87-1

Relevant to this article

Hot Product