TETRAHEDRON LETTERS Pergamon
Tetrahedron Letters 42 (2001) 1855–1858
Solid-phase synthesis of a-hydroxy phosphonates and hydroxystatine amides. Transition-state isosteres derived from resin-bound amino acid aldehydes Roland E. Dolle,* Timothy F. Herpin and Yvonne Class Shimshock Department of Chemistry, Pharmacopeia, Inc., PO Box 5350, Princeton, NJ 08543 -5350, USA Received 18 October 2000; revised 12 January 2001; accepted 15 January 2001
Abstract—Resin-bound N-acylated amino acid aldehydes iv were converted in a single step to a-hydroxy phosphonates vii (Pudovik reaction) and in six-steps to hydroxystatine amides viii, demonstrating the utility of intermediates iv for constructing multiple aspartic acid transition-state isosteres. © 2001 Elsevier Science Ltd. All rights reserved.
The incorporation of stable transition-state isosteres into pseudopeptide templates is an effective strategy for inhibiting aspartic acid proteases.1 Nearly two dozen isosteres have been reported in the literature, synthesized primarily in conjunction with the development of clinically efficacious inhibitors of renin and HIV protease.1,2 Several isosteres, notably statine i,3 hydroxyethylamine ii,4 and the C-2 symmetric diaminodiol iii,5 have served as pharmacophores in chemical libraries. These libraries successfully yielded highly potent and selective inhibitors of HIV protease, human cathepsin D, and malarial plasmepsin II. In seeking a versatile approach to con-
struct encoded combinatorial libraries with broad utility for the discovery of aspartyl protease inhibitors, amino acid aldehydes iv were envisaged as pivotal intermediates from which multiple transitionstate isosteres may be obtained. We recently reported an efficient five-step conversion of b-amino homoallylic alcohols v to the hydroxypropylamine isostere vi on solid support.6 Intermediates v were derived via the addition of allylindium or allylboronic acid pinacolate to resin-bound N-acyl amino acid aldehydes vi. Herein we describe the solid-phase synthesis of ahydroxy phosphonates (Scheme 1) vii and a-hydroxystatines (Scheme 2) viii from iv.
* Corresponding author. Present address: Adolor Corporation, Malvern, PA 19355, USA. 0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 0 - 4 0 3 9 ( 0 1 ) 0 0 0 9 7 - 1
R. E. Dolle et al. / Tetrahedron Letters 42 (2001) 1855–1858
1856 Ph
NO2
Ph
OTBDMS
H2N
Br
H N
Ph O OMe P OMe OH
O
2
a N
CHO
Ph
ref. 6 O
Ph
N R
O 1
3
4: R = 5a: R = H (α-R) 1:1 (85%) 5b: R = H (α-S)
b Ph
Ph O OMe P OMe OH
OMe O N H MeO
O
O OCH2Ph P OCH2Ph OH
N H N
6: α-R/S 1:1 (85%)
Ph
Ph
O
O OCH2Ph P OCH2Ph OH
N H
Ph
7: α-R/S 1:1 (90%)
O
O OCH2CF3 P OCH2CF3 OH
N H
8: α-R/S 3:1 (92%)
9: α-R/S 1:1 (87%)
Scheme 1. Solid-phase synthesis of a-hydroxy phosphonates. Reagents and conditions: (a) 10 equiv. each HP(O)(OMe)2 and Et3N, CH2Cl2, 12 h; (b) hn (365 nm), MeOH, 3 h, 40°C.
Ph OH
O Ph
b
N Ph
CHO O
Ph
1
N H
(a:b 7:1; 70%)
N H
12a
OH
(a:b 7:1; 85%)
Ph OH
O 11
10
Ph
CO2tBu
N H
d
OH 12b
CONHBu
Ph OAc
Ph OR
OH
f
c
17a
N
Ph OH
O Ph
O
Ph OH
O Ph
Ph
O
CO2tBu OH
c
CO2tBu
N
CO2tBu
N
N H
Ph
Ph OH
Ph a
Ph g
CONHBu OH
O
CO2R
N
CONHBu OR
Ph
15: R = Ac 16: R = H
e
O
OAc
13: R = tBu 14: R = H
17b
Ph
O
OH N H
Ph
CONHBu
N H
OH
MeO
18: syn/anti: 7:3 (72%)
Ph Ph
O
OH N H
N H N
OH O
20: syn/anti: 7:3 (76%)
Ph OH
OMe O
O O
OH O
19: syn/anti: 7:1 (77%) Ph OH
O BuHN
N
N H
CONHBu OH
21: syn/anti: 7:1 (65%)
O
OH N H
Ph
H N
OH O
OH Ph
22: syn/anti: 7:4 (75%)
Scheme 2. Solid-phase synthesis of hydroxystatines. Reagents and conditions: (a) 4 equiv. Ph3PCHCO2tBu, THF, 12 h; (b) 0.2 equiv. OsO4, 3 equiv. morpholine N-oxide, acetone–water (1:1), 12 h; (c) hn (365 nm), MeOH, 3 h, 40°C; (d) 30 equiv. (MeCO)2O, cat. DMAP, pyridine, 12 h; (e) 50% TFA–CH2Cl2, 8 h; (f) i. 8 equiv. pentafluoro phenyl trifluoroacetate and 10 equiv. pentafluorophenol, DMF–pyridine (1:1); ii. DMF wash; iii. 3 equiv. BuNH2, DMF, 12 h; (g) 10% H2NNH2–MeOH, 2 h.
R. E. Dolle et al. / Tetrahedron Letters 42 (2001) 1855–1858
Resin-bound aldehyde 1 was prepared via coupling of phenylalaninol-O-t-butyldimethylsilylether 2 to Tentagel resin 3 derivatized with 4-bromomethyl-3-nitrobenzoic acid (photo-labile linker), benzoylation, desilylation, and oxidation with iodoxybenzoic acid as previously described.6 Treatment of 1 with 10 equiv. of a 0.2 M solution of dimethyl phosphite in CH2Cl2 containing 10 equiv. of Et3N (12 h, 25°C; 1–4) followed by photolysis furnished the a-hydroxy dimethylphosphonates 5a,b as a 1:1 mixture of diastereomers. Purity (HPLC) of the crude cleavage product was in excess of 95% and 5a,b were isolated in ca. 45% yield, respectively, from 1.7 Optical purity of the isolated materials was >95% via chiral HPLC indicating racemization at the a-center in 1 was negligible. A survey of the reaction of other amino acid aldehydes (e.g. alanine, leucine, substituted phenyalanine acylated with electron-rich/deficient aroyls, heteroaroyls, substituted acyls) with commercially available dialkyl- and dibenzyl-phosphites gave similarly clean conversions (>85% purity by HPLC, >80% isolated yield) to a-hydroxy phosphonates vii (e.g., 6–9). Diastereomeric ratios ranged from 1:1 to 3:1 (31P NMR) in favor of the a-S stereochemistry (syn with respect to the amino acid side chain), analogous to solution-phase results.8 The only exception being the reaction of the ethylene cyclic phosphite with intermediates iv which proceeded in poor yield (<10%). In contrast to the single step conversion of iv to the hydroxy phosphonate isostere, elaboration of iv to hydroxystatine required a two-carbon homologation and installment of the diol. Two-carbon homologation was straightforward via Wittig condensation of 1 with t-butyl triphenylphosphoranylidene)acetate (4 equiv.) in THF at room temperature for 3 h producing 10.9a Dihydroxylation of 10 using a catalytic amount of OsO4 and NMO afforded diol 11. This reaction occurred in high yield and purity as established by photolytic cleavage of 11, yielding 12a,b as a 7:1 mixture of diastereomers.9b The diol stereochemistry (major diastereomer—diol syn with respect to the amino acid side chain) was tentatively assigned based on literature precedent for this reaction in solution10 and on the basis of biological activity.11 To complete the synthesis of the desired hydroxystatine amide isostere, it was found necessary to protect the diol 11 as its corresponding diacetate diester 13 (30 equiv. Ac2O, pyridine, cat. DMAP) to permit clean ester to amide conversion.12 Treatment of resin-bound ester 13 with 50% TFA– CH2Cl2 for 8 h gave acid 14. Several conventional carboxylate activation methods including the HATU, DIC/HOBt or PyBroP either gave very slow conversions when n-butylamine was added with the activating agent or multiple by-products when the carboxylate was pre-activated. The mixed anhydride derived from isobutyl chloroformate gave somewhat purer product and the formation of the pentafluorophenyl ester with pentafluorophenyl trifluoroacetate13 further improved the coupling. Optimizing the latter method, the last trace of unidentified by-products was removed by adding 2 additional equiv. of pentafluorophenol to the
1857
pentafluorotrifluoroacetate to ensure that the initially formed mixed anhydride was completely converted to the pentafluorophenyl ester before the addition of amine. Under these conditions, the two-step coupling provided clean high yield conversion to amide 15.14 Deprotection of the diacetate (15 to 16) with hydrazine in methanol was rapid requiring 2 h on solid support (<1 h in solution9a). Photolysis of 16 afforded hydroxystatine amide 17a,b as a 7:1 mixture of diastereomers in 70% overall yield from 1. The utility of this chemistry to generate hydroxystatine amides is further exemplified in the solid-phase synthesis of 18–22. In summary, resin-bound amino acid aldehydes are useful intermediates for generating multiple transitionstate isosteres, specifically hydroxy phosphonates vii, hydroxystatine esters and amides viii, and hydroxypropylamines vi.6 Further application of these chemistries in the preparation of encoded combinatorial libraries targeted for aspartic acid proteases will be forthcoming.
References 1. Rich, D. H. In Proteinase Inhibitors; Barrett, A. J., Salvensen, G., Eds.; Elsevier Science: Amsterdam, 1986; Chapter 5. 2. (a) Abdel-Meguid, S. S. J. Med. Res. Rev. 1993, 13, 731; (b) Huff, J. R. J. Med. Chem. 1991, 34, 2305; (c) Greenlee, W. J. Med. Res. Rev. 1990, 10, 173. 3. (a) Carroll, C. D.; Patel, H.; Johnson, T. O.; Guo, T.; Orlowski, M.; He, Z.-M.; Cavallaro, C. L.; Guo, J.; Oksman, A.; Gluzman, I. Y.; Connelly, J.; Chelsky, D.; Goldberg, D. E.; Dolle, R. E. Bioorg. Med. Chem. Lett. 1998, 8, 2315–2320; (b) Carroll, C. D.; Johnson, T. O.; Tao, S.; Lauri, G.; Orlowski, M.; Gluzman, I. Y.; Goldberg, D. E.; Dolle, R. E. Bioorg. Med. Chem. Lett. 1998, 8, 3203–3206; (c) Owens, R. A.; Gesellchen, P. D.; Houchins, B. J.; DiMarchi, R. D. Biochem. Biophys. Res. Commun. 1991, 181, 402. 4. Haque, T. S.; Skillman, A. G.; Lee, C. E.; Habashita, H. G.; Ilya, Y.; Ewing, T. J.; Goldberg, D. E.; Kuntz, I. D.; Ellman, J. A. J. Med. Chem. 1999, 42, 1428. 5. Wang, G. T.; Li, S.; Wideburg, N.; Krafft, G. A. J. Med. Chem. 1995, 38, 2995. 6. Cavallaro, C. L.; Herpin, T.; McGuinness, B. F.; Shimshock, Y. C.; Dolle, R. E. Tetrahedron Lett. 1999, 40, 2741. 7. All new compounds gave physical and spectroscopic data consistent with their structure. Yields reported herein are cleaved, purified yields derived from 1. The isolated yield of 1 is 40% based on theoretical resin loading: see reference 6. The optimized reaction conditions were established based on a survey of up to 30 substrates. 8. Patel, D. V.; Rielly-Gauvin, K.; Ryono, D. E.; Free, C. A.; Rogers, W.; Lynn, W.; Smith, S. A.; DeForrest, J. M.; Oehl, R. S.; Petrillo, Jr., E. W. J. Med. Chem. 1995, 38, 4557. The addition of aldehydes to resin-bound phosphites has been reported: Cao, X.; Mjalli, A. M. M. Tetrahedron Lett. 1996, 37, 6037.
1858
R. E. Dolle et al. / Tetrahedron Letters 42 (2001) 1855–1858
9. (a) Solid-phase transformations in Scheme 2 were first investigated and optimized in solution using substrate 23; (b) Dihydroxylation of 24 afforded 25 as a 7:1 mixture of diastereomers. Attempts to use chiral ligands in the asymmetric osmylation reaction did not significantly change the diastereomeric ratio. 10. Reetz, M. T.; Strack, T. J.; Mutulis, F.; Goddard, R. Tetrahedron Lett. 1996, 37, 9293. 11. The minor diastereomers of 18–22 display inhibitory activity against the aspartyl protease cathepsin D and/or plamepsin II while the major diastereomers are inactive (unpublished observation). These results provide additional support that the amino acid side chain and the adjacent hydroxyl are disposed in an anti relationship in the minor isomer, as this is the stereochemical configuration required for aspartyl protease affinity (see references 1–3).
12. As originally planned, ester 10 would be converted to a,b-unsaturated amide 26 on resin and then dihydroxylated as a final reaction step before cleavage. However, amide 27 was a poor substrate for dihydroxylation (ca. 50% conversion to diol 28 under vigorous conditions) owing to the difficulty in hydrolyzing the intermediate osmate ester. This result was even more pronounced on solid-phase; hence, this route was abandoned in favor of the diol protection–deprotection (diacetate) sequence. 13. Ni, Z.-J.; Maclean, D.; Holmes, C. P.; Murphy, M. M.; Ruhland, B.; Jacobs, J. F.; Gordon, E. M.; Gallop, M. A. J. Med. Chem. 1996, 39, 1601. 14. (a) In some instances, partial de-acetylation was also observed during amide formation; (b) A broad survey of amines was conducted in this coupling step, anilines were noted as poor partners.
. .