Adamantane and Nipecotic Acid Derivatives as Novel β-Turn Mimics

Bioorgonic & A4edicinnl Chemistry Lmers, Vol. 4, No. 11, pp. 1361-1364, 1994 Copyright 0 1994 Elsevier Science Ltd Printed in Great Britain. All rights te.wwd 0960-894X194 $7.00+0.00

0960-894X(94)00156-1

Adamantane and Nipecotic Acid Derivatives as Novel P-Turn Mimics William J. Hoekstra, Jeffery B. Press, Mary Pat Bonner, Patricia Andrade-Gordon and Patricia M. Keane The R. W. Johnson Pharmaceutical Research Institute, Spring House, Pennsylvania 19477

Kathleen A. Durkin and Dennis C. Liotta Emory University, Atlanta, Georgia 30322

Kevin H. Mayo Thomas Jefferson University, Philadelphia, Pennsylvania 19107

Abstract 1,3-Adamantanedhnine and nipecotic acid were. used as scaffolds in designing (hm peptide mimetic de.rivatives 3 and 5. respectively, targeted as fibrinogen-GPIBiIL4 antagonists. The design focused on the 406409 KQAG region of the y-chain of fibrinogen which appears as a p-turn in solution NMR stNctures of Fg-7385-411.

As part of our program in cardiovascular agents, we became interested in novel fibrinogen receptor antagonists which might have utility as platelet aggregation inhibitors. While several laboratories are studying RGD mimetics designed from segments of fibrinogen’s a-chain, we are unaware of any examples of y-chain mimics under investigation.1

This is especially surprising since the RGD sequence is present in a large

number of blood adhesive proteins (hence RGD mimetics are potentially unselective), while the y-chain is unique to fibrinogen and thus offers the possibility of improved selectivity.2 Since the 385-411 sequence of the y-chain (derived from BrCN degradation of fibrinogen) has good inhibition of fibrinogen binding to immobilized GPIIb/IIIa (ICso = 3.4 nIvI) as well as platelet antiaggregatoty activity, we undertook a research program investigating structural mimics of this region. A thorough NMR investigation of ~385-411 at pH 3.5 and 6.0 in 20% TIE/water

solution gave rise to families of solution structure8 with secondary structural

features in the 400-411 binding domain.

Several families of these structures show a tight turn region

encompassing KQAG (406-409) without the commonly observed i to i+3 hydrogen bond. The average CaCo distances between the i and i+3 residues in these turn regions ranged between 4.2 and 4.7 A, suggesting a type II g-turn (1).3,4 Inthis Letter, we describe one aspect of our research project that addressed the design and synthesis of mimics of this KQAG g-turn region.5

cl_di

Q

0

CH,

A

G

;

\/cQH 0)

C&to

KQAG(D) region of y-chain of fibrinogen [406-409 (410)]

1361

C& = 4.7A

1362

W. I. HOEKSTRA ebAI.

One of our design goals was to utilize commercially scaffolding

so that additional

aminoadamantane SIBYL

predicted

availability

available compounds

be conveniently

2 was initially proposed

of precursor

that of a Type II p-turn.

with requisite functionality

introduced

(observed

it was determined

that amino acid replacement

as

The 3-

replacement.

preliminary

search), agreeing well with the experimentally

Furthermore,

as necessary.

as a “QA” dipeptide

1,3_adamantanediamine,

the Cct~ to Ccc~ distance to be in the range of 4.5-4.9A

during a systematic

relationship

could

unit of model compound

addition to the commercial populations

side chains

calculations

In using

in several conformer

determined

4.7A and close to

from the y400-411 peptide structure-activity

of K406 and Da10 greatly reduced the biological

activity of y400-

41 1.*t6 As an additional design requirement,

we decided to incorporate the K residue as well as the side chain

of the flanking D residue into our mimetics.

The Q and A comer residues of the peptide turn may function as

turn-inducing

residues

binding-critical

which enforce

a conformational

K and D side chains.

Presumably,

constraint

securing

this conformational

the spatial orientation

constraint

of the

would be provided by the

synthetic scaffold. After synthesis

of a few adamantane

some activity in our biochemical platelet

aggregation

rigorously

@ 20 PM).*

using modelling

over 50 picoseconds). 4.7& was calculated

we were gratified to find that adamantane

Prior to undertaking

techniques

A &K

derivatives,

assays (44% inhibition of fibrinogen which included

an extensive molecular

to Cam distance of 7.2-7.7&

for a number of conformational

binding synthesis

dynamics

considerably

families.

37 afforded

@ 50 FM, 40% inhibition program,

simulations

of

3 was examined (heating to 900°K

longer than our target distance of

This elongated

distance may arise from side

chain flexibility and may be the cause of the relatively low potency of 3.

NHBoc

H,N(CH~).I 2, avg. CctK to C,, We next examined simulations structure.

narrow conformer Compound

diminished

for lysine,

9-amino-

and nipecotic acid derivatives

4 had the best fit with a calculated

population distribution

mimetics

to the NMR-derived

examination

1-fluorenecarboxylic

led to the prediction

Cct~ to C,c

acid,

that the 3-S-(+)-

distance of 5.7A and a very

of 5.0-7.0 A.

only modest biological

activity of 24% inhibition of fibrinogen

the B-turn of 4 in fact mimics the $06-409 activity of 4 is suggestive

KQAG

of 1,8-naphthalenediamine,

4 meets our design goals of creating a p-turn mimic from commercially

I), but produces

chain of fibrinogen

available scaffolds using molecular dynamics

2_aminoperimidine,

pseudopelletierine

of nipecotamide

commercially

fit of these potential

the L-configuration

5_aminophthalimide,

isomannide,

Assuming

eight bis-functionalized,

maintaining

tetrahydropyrimidone,

(Scheme

3, avg. CaK to CaG = 7.7A

in an effort to find an improved While

diastereomer

= 4.9A

(CH,hNH, 4,n=l 5,n=2

of the importance

region as predicted of precise positioning

available materials binding at 50 PM.

by our computer

models,

the

of the D410 side chain in the y

(vi& s~pdpra). Extension of the C-3 side chain of nipecotamide

4 by one methylene unit (p

Adamantane

alanine)

provided

compound

dramatic improvement

and nipecotic

5 (5 is a 1:l mixture of diastereomers

in biological

1363

acid derivatives

by NMR and HPLC), which showed

activity (vis-a-vis 4) with an ICso of 0.074 uM in fibrinogen binding and

86% inhibition of platelet aggregation at 50 pM. * This effect may be due to an improved geometric alignment of the carboxy terminus of 5 with the native peptide. H

H

Scheme 1

OH

a, b

0CH3 l+f v H

0

-&

92%

lHCI

N 0 +

H

H

H

d, :

N, (CH$OBn

%

53% n=l 92% n=2

NHAc “H (CH.&NHBoc

H

& N 0

c

0

N\(CH$OH

46% n=l -& 0 53% n=2

0

NHAc //H (CH*)‘,NHBoc

N

0 NHAc “H

+

4 ,.,=,

(CHd4NH2

5 n=2

7

(a) Ac-L-Lys(Boc)-OH, BOP-CVNMM, CH2C12; (b) LIOH, aq. THF; (c) H2N(CH&C02Bn, EDC, NMM; (d) H.JPd-C, THF, aq. AcOH; (e) TFA or aq. HCI The successful

use of adamantane

new tools to the field of peptide mimics. our specific fibrinogen

therapeutic ychain.

Furthermore,

targets by facile modification

Further modifications

of GPIIb/BIa antagonists References

and, more importantly,

nipecotic

acid as p-turn scaffolds

we have demonstrated

adds two

the utility of these mimetics for

of the region representing

the ~410 side chain of the

and more detailed biological evaluation of this nipecotamide

series

are the subject of a future publication.1°

and Notes

1.

Blackburn, B. K.; Gadek, T. R. Ann. Rep. Med. Chem. 1993,28,79-88.

2.

Kloczewiak,

3.

Fan, F.; Kloczewiak,

M.; Timmons,

S.; Bednarek, M. A.; Sakon, M.; Hawiger, J. Biochemistry

1989, 28,2915-

2919. An observed

p-turn in

the QAGD sequence (407-410) of the shorter y392-411 has been reported with 60% incidence

M.; Mayo, K. H., Biochemistry,

at pH 5.2

only (M. Blumenstein

et al, Biochemistry

1992,31,

submitted for publication.

10692).

We have made NMR determinations

showing that the series of KQAG p-turns exists in the longer y385-411 at the two disclosed pII’s. 4.

Because the NMR-derived conformation, Preliminary

solution structure of ~85-411

transferred-NOE experiments

studies of the peptide

demonstrate

may not be indicative of its receptor binding “bound” to GPIIb/IIIa

that with high enough peptide

are currently

to receptor

ratios, ~385-411 is

clearly interacting with its receptor, as indicated by the build-up of strong peptideheceptor studies should lead to structural families which are more representative the peptide.

underway.

NOES. These

of the bioactive conformation

of

W. J. HOEKSTRA et al.

1364

5.

For a recent review of /&turn mimetics, see Olson, G. L.; Bolin, D. R.; Bonner, M. P.; BBS, M.; Cook, C. M.; Fry, D. C.; Graves, B. J.; Hatada, M.; Hill, D. E.; Kahn, M.; Madison, Sarabu, R.; Sepinwall, J.; Vincent, G. P.; Voss, M. E. J. Med. Chem. 1993,36,

6.

Hoekstra,

W. J.; Bonner,

Evangel&o,

M. P.; Andrade-Gordon,

V. S.; Rusiecki, 3039-3049.

P.; Press, J. B.; Keane,

M. F.; Mayo, K. H.; Fan, F.; Kloczewiak,

V. K.;

P. M.; Tomko,

K. A.;

M.; Durkin, K. A.; Liotta, D. C. 207h ACS

Meeting, Abstract 214, San Diego, CA, 1994. 7.

Compound

3, a tan powder, was prepared from 3-amino- 1-adamantane-N-glycine

Boc-L-Lys(Z)-OSu

by standard solution phase peptide synthesis:9

benzyl ester and N-

lH NMR (DMSO-d6)

6 7.48 (br. s,

lH), 6.64 (d, J=7, lH), 3.86 (m, lH), 3.03 (br. s, 4H), 2.62 (m, IH), 2.03 (m, 2 H), 1.99 (m, lH), 1.7-1.9 (m, lOH), 1.62 (br. s, 2H), 1.4-1.6 (m, 7H), 1.37 (s, 9H), 1.26 (m, 3H); MS m/e 453 (MH+); [a]25D -26.25” (c 0.08, MeOH).

Anal. calcd. for C23H40N405*2C2H402*H20:

Found: C, 54.67; H, 8.20; N, 9.73. follows: To a suspension

3-Amino-1-adamantane-N-glycine

of 1,3-adamantanediamine

dihydrochloride

C, 54.90; H, 8.53; N, 9.48. benzyl ester was prepared

mL) at RT was added NaH (5.64g, 0.19 mol, 80% mineral oil suspension). 55°C for 2 h, cooled to RT, and treated with benzyl2-bromoacetate 1 h period.

as

(15.0 g, 0.063 mol) and DMF (300 The mixture was warmed at

(14.4g, 0.063 mol) dropwise over a

The mixture was stirred for 18 h at RT, diluted with sat’d NH4CI (50 mL), sat’d NaHC03

(150 mL) and CH2C12 (200 mL). The layers were separated, and the aqueous layer was extracted with CH2C12 (100 mL). MgS04,

The combined

and evaporated

iPrOWCH2C12)

organic layers were washed with water (150 mL), filtered through

to an oil. The oil was purified by flash chromatography

to give the title compound

(300 MHz, DMSO-d6)

(l%NH40WlO-40%

(13.0 g, 65%) as a white powder: mp 172-174°C; lH NMR

6 8.05 (m, 3H), 7.40 (m, 5H), 5.15 (s, 2H), 3.37 (s, 2H), 2.18 (s, 2H), 1.2-1.9 (m,

12H); MS m/e 315 (MH+). 8.

In vitro biological

methods:

purified GPIIb/IIIa concentration)

and the plate washed extensively.

incubated

at RT for 15 min.

test compounds

buffer.*

standard: Merck L-700462, IC5u = 0.001 pM.

sodium citrate.

The absorbance

Bodamzky,

by centrifugation.

for 3 min after addition thrombin, O.lunitAnL.

by increase

(10 nM, final

reagent

is added and

in light transmission

of compound-treated

RGDS, ICso = 0.9 @I. in tubes containing

PRP is gel-filtered Aggregation

0.13M through

is monitored

in a

Percentage platelet aggregation vs. control-treated

platelet

Peptide standard: RGDS, IC5o = 30.0 pM. M.; Bodanszky,

A. The Practice of Peptide Synthesis, Springer-Verlag:

Hoekstra, W. J.; Bonner, M. P.; Andrade-Gordon,

(Received

standard:

normal donors is collected

Platelet rich plasma (PRP) is collected

Press, J. B.; Mayo, manuscript

Peptide

2B, and platelet count is adjusted to 2~10~ platelets/ sample.

is calculated concentrate.

is read at 490 nM.

Blood from drug-free,

BIODATA aggregometer

fibrinogen

Vecta Stain HRP-Biotin-Avidin

Nonpeptide

Azgrenation.

Biotinylated

and left at RT for 2-4 h. The solution is

The wells are allowed to develop for 3-5 min at RT after addition of a

developing

Sepharose

10.

in a TiterTek 96-well plate.

is added to wells containing

discarded

Platelet

9.

&&$ Phase Purified Glvco protein IIb/IIIa Binding Assav. RGD-affinity

is immobilized

K. H.; Fan, F.; Kloczewiak,

P.; Keane, P. M.; Tomko, K. A.; Evangelism M.; Liotta, D. C.: Durkin,

in preparation.

in USA 24 February

1994; accepted

New York, 1984.

25 April 1994)

M. F.;

K. A. J. Med. Chem.,