Tetrahedron Letters. Vol. 34, No. 34. pp. WXG466.1993 hinted in Great Britain

~9~3 s6.00 + .oo Perg- PreM Ltd

Diastereosektive Synthesis of FuncthUxed 2,~DBamho-34ydroxyglutaric Acid Derivatives of Potential Biologfcal Interest from Glycine Derfvathres

M~~7~~~ ~~ have alrdy been found in nsture, and there is sn evergrowing

interest in the synthesis, ph armacology, and conformational properties of these compounds due to their

biological activities ss antibiotics, metal chelators, neurotoxins, and enzyme inhibitors.’ Thus,

a,a~~bsti~~u-~0 acids,2 @- or 7-hydroxy-a-amino acids?” and b&-a-&no acids’ have been the

subject of numerous studies over recent years. In particulsr, Z,~~o-3-hy&oxy~u~c acid derivatives

have turned out promising in the development of new a&umoml agtxtts? Nevertheless, most syntheses thus far

developed for these comgouuds usually use 1,3-dismino-2-proponol ss starting material, therefore invoiving

rather e ADDS to complai disstemom eric mixtures.

On the other hand, kinetically controlled sldol addition of a-met&ted isocysnides to simple carbonyl

compounds provides sn useful route to l3-amino slcoho1s.b Therefore, this methodology should allow for the

disstereoselective synthesis of the i,3dismino-2-propsnol moiety from suitable functionalixed N-protected

a-amino ketones. Howevcz, metalation of ethyl isocyanoscetste with KO’Etu in THF at -78°C followed by

tmatment with ketones la and lb’ under standard reaction conditionsb did not produce the expected oxsxolines

(synthetic equivalents of 2,4dismino-3-hydroxyglutaric acid derivatives) (Scheme 1), but iV-bis(methylthio)-

methylene alanine ethyl ester 3 was instead isolsted after methsnolic quenching, probably ss a ammquence of

the low activation energy for the retro-aldol reaction* promoted by the high stability imparted to the extruded

fragment by sxa-sllylic anion resonsnce’, as well 89 steric compression in sldolates 2.

Scheme 1 .I

However, oxaxolines 6 were obtained upon treatment of equimolecular smounts of ethyl isocyanoacetate and

ketones 1 with met4 alkoxides under protic umditions &heme 2; Table 1). In this media, the retro4lsisen

reaction in this case responsible of the fotmation of 3 and the amesponding ester derivative of the scyl

5463

5464

fragment in the substrate 4 (isolated for lb and lc) was inhibited at the expense of the diasterem&ctive

synthesis of oxazolines 6. It has to be pointed out that, whereas the choice of the alkoxideAIcoho1 system had

scarce influence on the 3:6 ratio in the case of lithium aikoxides, a marked dependence on this parameter was

noticed in the case of potassium alkoxides (Table 1; runs l-4).

r f& Me

Me

4,

Rcc&FT + J--N MeS ’

.

Mes ’ MoRm E

+YC

+c

E+-N=C: +_

- Me8 4, + Rcc&R

\ o’,M’ A

4

1 R

Id R E

a W EW

b hnvl Et4C

c 2+uyi EQC

On the other hand, when the temperature was increased (runs 5-7) a decmase of 3:6 ratio was observed.

Both these statements are in agreement with the minor activation energy needed for the am-ally1 anion 3

extrusion versus aldol reaotion. Kinetic contro1 was also observed in the diastereomeric reaction pathway to 6

Veo the adduct intermediate 2 (Scheme 2) since neither epimerixation nor deuterium incorporation were

observed when diastereomeric mixtures of oxaxolines 6 were treated with RMeO/CD,OD under standard

reaction conditions.

Table 1. lSaction of Etbvl Isocvmemetate with Ketoaes 1 Under Pmic C!on~ns’”

RUII Ketone MR’O R’OH t(*C) 3% ratio ( lR$s,3S: lR,2&3R) ratio”

1 la Li’BuO BuOH 25 45:55 5o:SO

2 la LiMeO MeOH 25 3263 66:34

3 la ICBUO BuOH 25 4555 59:41

4 la KMeo h&OH 25 13:87 84~16

5 1s NaBtO EtOH -23 1684 8218

6 18 NaEtO EtOH 0 595 80:20

7 la N&O EtOH 25 5:95 80:20

8 la TlEtO EtOH 25 18:82 83:17

9 lb KEto EtOH 25 991 9o:IO

10 lc REtO EtOH 20 20:80 88:12

11 lC Li’pro ‘&OH 20 15:85 70:30

Out of the four possible diastereomers for 6, only two of them were obtained. The diastereoselectivities

observed were subject to metal tuning, and the highest diasmmom&c excesses were found with the use of

potassium alkoxides (Table 1; run 4 for la; run 9 for lb, and run 10 for 1~).

Several attemps to hydrolyze the oxaxolines 6b were unsucce$ful because a retroaldol fragmentation of

j3-hydroxyformamides 7, which were originated by oxaxoline ring opening, was the principal reaction

pathway.” However, albeit oxaxolines 6a did scarcely undergo this retroaidol reaction, cyclixation to single

5465

diastcreomezic & and 9a was ~btained,‘~ instead of the expected 2,~n~2,3~m~yl-3-hy~xy~iu~~

die&y1 eater. It has to be pointed out that full cpimeri;Eation to (lR,2Z&3R:ZQR,3&7a and

(lR,2&3s:lS,2&3R)-7b-7~ was observed for 6a and 6bc, nspactively, on comparing the ‘H-NMR (300

MHz) spectra of quenched reaction mixtures before the equilibrium was reached.

QWW 10 (MAa) llb

On the otha ha& rn~yl~o~ of oxaxoline 6b fol~w~ by NaBHe ~ gave rise to oxazolidine

lib, which upon oxalic acid hydroly& cyclixed to 2-methylthiooxaxoIine Mb with good chemical yield. In

this case, no epimerixation of C-3 was observed. Repetition of the N-methylation@IaEH, reduction/oxalic acid

hydrolysis” sequence on Mb afforded the mixture of the two monolactam derivatives of 12b. Theref~, the

herein described methodology provides an access to symmehical and unsymmetrical ~~~~~o~~on

of the lS-diamino-2-propaaol core unit.

Stereochemical and w control in the hydrolysis, new tictionalization reactions, and the

application of this aldol methodology to an enantioselective semis of bisa-amino acid dczivatives are in

progresS.

Aclmowladgem~tm.- This work was made possible by the financial support of the Dire&on General de

Investiga&n Cientifica y Tccnica (Project PB90-0043). E. M.-S. gratefully acknowledges the Ministry of

Education and Science for a F.P.I. grant.

REFERENCESANDNCYI’ES

1.

2.

3.

4.

5.

6.

a) Weinstein, B. Chemistry and Biochemistry of Amino Aria%, Pqtides and Proteins; Mad D&IX

New York. 1983; Vol. VII, pp. 267-357; b) Wagner, I.; Musso, H. Angew. Chem. Int. Ed. Eng. 19S3,22,

816-828~) Enders, D.; Jqelka, U.; Diicker, El. Angew. Chem., Int. Ed. Eng. 1993, 32, 423-425; d)

Wittenberger, S.J.; Baker, W. R.; Dom~er, B. G. Tktmhedwn, 1993.49.1547-1556.

For a review on the synthesis of a_,u-dieubstituted and @-hydroxyu-amino acids see: Williams, RIU.;

&s&esis of ~~~~ Active &Amino A&, Pcqamon PUG%:: Oxford. 1989 and refaces cited therein.

Recent references: a) Mlhausen, T, Groth, U.; sCh6llkopf, U. LMigs Ann. fBem. 1992, 523-526; b)

Zanbergen, P.; Brus$cc, J.; van der Gen, A.; Kuse, Ch. G. Tetrahecibn Asymmetty, 1992,3, X9-774, c)

Harwoad, L. M.; Macro, 3.; Watkin, D.; Williams, C. E.; Wong, F. L. TeblrliedKur Asymmetry, 1992,3,

1127-I 130; d) Ohfune, Y. Ace. Chem. Rex 1992,25,360-366 and r&~cnces cited therein.

a) Ku&da, T.; Ishizuku, T.; Iguchi, T.; Hiroc, M. J. Org. C%em. 19ft8, 53, 3381-3383; b) Bold, G.;

Allmendingcr, T.; Herold, P.; Mocsch, L.; &h&r, W-. P.; Lh&aler, R 0. HeZv. Chim. Acta, 1992, 75,

865-882.

a) Sasamori, H.; Hori,C.; Chiba, K. Jpn. Ko&u;llt T&&o K&o Jp 02,223,591; CZ4 1991, IiS, 84232; b)

Sasamori, H.; Hori, C.; Niitsuma, S.; Chib, K.; .fpn. Kokai Tokkya Koho Jp 03,127,796; ci 199X, 115,

293683.

a) S&illkopf, U.; Ge&art, F.; Hoppe, I.; Harms, R.; Hantke, K.; scheunerormn, K. H.; Eilers, E.; Blume,

E. Lie&p Ann. Chem. 1976, 183-202; b) S&llkopf, U.; Gerhart, F.; SC-, R.; Hoppe, D. Liebigs

Ann. Chem. 1972,116129.

5466

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

Ketonea 1 were synthetixcd hwn N-bi@cthyhhio)mcthylene alanine ethyl csthcr and the corraspoading

acyl halides u olcctrophilca, following literxmrc conditions: Hoppe, D.; Bc&mann, L. L&b&s Ann.

Chem. 1979,2066-2075.

Amctt, M. E.; Fischer, F. T.; Nicholr, M. A.; Rii, A. A. J. Am. Chem. Sot. 1989, III, 748-749.

Hoppe, D. Angew. Chat, In?. Ed. Eng. 1975,19,424426.

To a suepcawion of MR’G (0.38 ma&) in R’OH (0.35 mL) u&r Ar xtmo#wc, a solution of ethyl

isocyanoawtatc (0.38 -1) and l(O.38 nnnol) in R’GH (0.7 mL) wu addal. The rat&on mixture was

st&d for 15 mim~tea at the given mture, quenched with water, and extra&d with CH&l,. The

analysis by ‘H-NMR (300 MHz) of the reaction crude8 showed in all caaea a net cunvemion of initial

ketone. (lR,2,$3S:lR,2S,3R) Oxazoline diastuwm& ratios were determined by ‘H-NMR from the

reaction crodeb&rctheirpu&ication by flash clnwaatognp hy on silica gel.

The &or oxaxoline wan i8olatai in all case6 to allow (lR,2$3S) mccmxtc configurational a6signcmcat.

X-Ray daeamidon far 6~” is in agrccwnt with the results obtained from NOE m on the

mixtures of oxazolinc 6~. The generalization of this procedure to oxaxolinea (r and 6b allowed the

atabliuhmcnt of an identical relative con&u&on for the major isamen of thcae ones. Thcnc wults am

be justified on the basis of a gwmetrical minimization of the diurtereomaic oxaxolims with the MM2

force fieldI (rcwlts to be pabliahed). Minor oxaxolincs were in all cases the (lR,2.$3R) cpimeric

we. Reaction condition~:‘~ la+78 (yield: 4O%)(AcGH (1.8 M)/H.@-EtGH, 2O’=C, 100 min); 6r+& (36%) + h (47%) (AcGH (1.8 M)/H#-EtGH, pH=5-6,5ooC, 20 h.). Reaction conditions:” ib+llb (yields 85%): i) MeGTf, 2 cq. (2h, r.t.); ii) NaBH,, 2 cq. (THFiMeGH: 4/l v/v; 30 min., Ooc). llb-+lob (yields 86%):iii) oxalic acid, 5 cq. (THF/H~O: l/l v/v, 64 h.,SSoC). lOb+l2b(ae a mixture of the two monolactam d&at&s) (as i, ii, iii; yield: 65%). Strwtuml and relative configuration of all these compotmds were tested by ‘H-NMR (300 MHz) and “C-NMR (75

MHz), and combwtion analysis after purificatioz~ by silica gel flash &romatogrxphy.

Crystal data for 6e: C,s&N,?$,, monoclinic, r10.538(6), &11.356(3), c=l8.213(% u-90”,

p-IO4.23(3)0, r590”. V=2113( l)A3, 2=4, D,-1.34 g.cm”, F(OOO)=904, cI(M+p2.74 cm-‘. Dif&action data were mauored on an Enraf-Nonius CAD4 dim opaating in the (O-28 mode with

graphite-monochmmxtai Mo-K, radiation (X30.71069 A) up to -30” from a cry8tal of dxc 0.4x0.4x0.3

mm. 6128 Unique reflexions were acanned and 2140 with 1>2a(I) were conaidercd and used in the

analysis. The intendtier were corwted for LoreI@ and pohwization &cts. The #ructum was aolvcd by

direct methods and Fourier synthcaia using the Multan 80” and the X-Ray 8019 systems and refined by

least aquarcs. The R and kfwtors were of 0.078 nod 0.074.

Mel+ mechanic calculations were carried out using PCMODEL 4.0 sofIwxrc (SsreM Sofhvarc, Inc.,

Bloomington, IN, U.S.A.).

Sch6llkopf, U.; G&art, F.; SohriMa; R.; Hoppe, D. Lie&r. Ann. Ckm. 1972,116129.

a) Mcyun, A. I.; Shipman, M. .I. Org. C&m. 1991,56,7098-7102; b) Meyers, A. I.; Roth, G. P.; Hoycr,

D.; Bamu, B. A.J. Am. Ckm. Sot. 1988,IlO, 46114624.

Main, P.; L&ngor, L.; Woolfoon, M. M.; Gcrmain, G.; Dcclemq, J. P. MULTANBO, Univtity of York,

England and Louvain, Belgium, 1980.

Stewart, J. M.; Kundell, F. A.; Baldwin, J. C. X-Ray 60 @Hem, Ckmputa Science Center, University of

M College Park, 1)85.

(Received in UK 3 May 1993; accepted 1 July 1993)