Pergamon

0040-4039(95)01215-X

Tetrahedron Letters, Vol. 36, No. 34, pp. 6043-6046, 1995 Elsevier Science Ltd

Printed in Great Britain 0040-4039/95 $9.50+0.00

Beyond Peptide and Nucleic Acid Combinatorial Libraries- Applying Unions of Multicomponent Reactions towards the

Generation of Carbohydrate Combinatorial Libraries

Matthias Goebel, Ivar Ugi*

Lehrstuhl I fiir Organische Chemic und Biochemie Technische Universit~it M0.nchen

Lichtenbergstr. 4, D-85747 Garching, Germany

Abstract: A novel union of multicomponent reactions based on alkylated glycals is introduced and exem- plaric small carbohydrate combinatorial libraries are generated from tri-O-alkyl-D-glucals, trimethylsilyl azide and dibenzyl disulfide.

In our continued efforts concerning efficient one-pot syntheses via the concept of unions of multicom- ponent reactions (MCR) l, we have taken a closer look at the glycals and their rather intriguing chemistry. The latter has been extensively reviewed by numerous authors 2 and applied to a number of valuable synthetic routes 3, especially in the field of oligosaccharide syntheses. Other applications include the synthesis of precursors for aminocyclitol antibiotics, such as the gentamicines, kanamicines and others 4.

With these possibilities of highly diverse chemistry in mind, it should be feasible to construct combinatorial libraries from glycals and suitable electrophiles and nucleophiles resulting in defined substitution patterns at C-l, C-2 and C-3 of the carbohydrate skeleton. So far, only very few examples of parallel synthesized carbohydrate combinatorial libraries are known, e.g. C-1---~6 diglycosides by Armstrong et aI.5. Previous work on carbohydrate libraries has focused more on extraction of carbohydrate mixtures from biological sources 6 or serial syntheses 7.

In the herein described tmion of multicomponent reactions, alkylated glycals react with two nucleophiles and an electrophile to furnish a possible maximum of four distinct types of products (scheme 1) in predictable ratios ~1. Our focus in this paper is on trimethylsilyl azide and various alcohols as nucleophiles and sulfonium complex 2 as electrophile. Especially the 1,3-dideoxy-l,3-diazido sugars obtainable this way are of particular interest, since they are precursors for 1,3-dideoxy-l,3-diamino sugars bearing close resemblance to the pharmacologically highly active 1,3-diaminocyclitols present in many commercially available antibiotics. Sulfonium complex 2 was prepared in analogy to Franck et al.3h by simply mixing dibenzyl disulfide and antimony pentachloride in methylene chloride at -60°C and using the resulting solution as such lz. Glucals 1.1-1.3 were prepared from D-glucal by modification of the Brimacombe protocoll°.

The possible products are first the "direct addition" compounds 15 and 16, resulting from an electro- philic attack of the sulfonium species 2 on glucals 1 in an analogous fashion to the three component addition reaction with arylsulfenyl halides and diaryl disulfides first performed by Franck et al. 3h ("F-3CR"). The relative amount of this set depends heavily on the reaction temperature and the equivalents of electrophile 2 used (table 1, Nos. 2 vs. 3, 5 vs. 6 and 4 vs. 7). Thus, low temperatures and less than one equivalent of 2 favors this set over the set of compounds 8, 9 and 10. As can be seen from table 1, 1-azido products 16

6043

6044

over l-alkoxy products 15 at lower concentrations of 2. The diastereomeric ratios 16b ' 16d are generally 2: l with little dependance on the reaction temperature.

R~O + TMS'Na I

~%jOR = / ~ 16 : TMS-N a, 2

R~0 ~ I ~ 1 .1 .1 .4

+ I [Ee l DI SbCIeE~mSbCi3

2 + RaOR 4

3A - 3C

O

R I O - ~ l R I O ' ~ + R3OR 4

+-RZOR 4

R 1 - _ . RaOR 4 R 1 - .

4.1~(-4.4¢t : X-OR2, Y - H 5.1a-5.4<x : X-OR3, Y - H

4.1~-4.4# : X - H , Y - O R 2 S.1/~-¢4~ : X - H , Y - O R 3

+ R30R 4

<. R=OR' >

+ TMS-Na < - RZOTMS >

RtO.-.-] dJr:taddition

R~b

E 16.1b,d - lS.4b, d : X - O1:~ l§ . lb ,d - 16.4b, d : X - N 3

R20R 4 or R2OTMS, 2

R ~ O - ' ~ 0

R~O

E 14.1b,d - 14.4b, d

l-transalkoxylaffon / addition

R = O ~ - - 11 : R2OR40r R2OTMS, 2

0 1 2 : 2 , 3 ) -

1 3 : TMS-N 3, 2 RIO I

X 6 . 1 A - 6 A A : × ' O R 3 , Y ' H

6 . 1 B - 6 . 4 B : X ' H , Y - O R 3

3-transalkoxylation / R~O ---~ 0 addition

R ~ O ~ 4 ~ x

OR" E

11.1a-d - 11.4e-d : X - OR 2

12,1a-d - 12.4a-d : X - OR s

13.1a-d - 13.4a-d : X - N s

R ' O ~

7.1C-7.4C : X-Na, Y - H

7.1D-7.4D : X - H , Y - N 3

E - BnS, D - BnSSBn ;

1.1; 4.1 - 16.1: R 1 - R 2 - Me 1.2; 4 .2 .18 .2 : R 1 - R 2 - Et 1.3; 4.3 - 16.3: R t - R 2 - Pr 1.4; 4.4 - 16.4: R' - R = - Bn

R ' O - ~ 0

- R ~ O ~

X 7 .1A-TAA : X-N3, Y - H

8 : R20R 4 or R2OTMS, 2

9 : 2 , 3 =

10: TMS-N 3, 2

7.1B-7.48 : X - H , Y - N a

Conflgura~m of subsUtuents: s : C-1 - C-2 - C-3 - axial (=-O-altro)

b : C-1 - C-2 - C-3 - equatorial (.8-D-gluco)

© : C-1 - C-2 - equatodld, C-3 - axiaJ (~-D-allo)

d : C-1 - C-2 - axial, C-3 - equatorial (¢PD-manno)

R,O._--I 3-azldatlon / J O a d d / t / o n

R I O ~ x

N, E e.la-d - e.4a-d : X - O R ~

9.1a-d - 9.4a-d : X - OFT ~

10.1s-d - 10Aa-d : X - N s

31B : R a - DAGIc, R 4 - H 0

3(;; : I~ - DAQIc, R 4 - SnBu3 3D : R 3 - DAQIc, R 4 - SIMea OAGIc - / I

Scheme 1. Possible pathways for glucal reactions with trimethylsilyl azide and alcohol nucleophiles (sources o f intermediates RzOR 4 and R:OTAIS shown in pointed <brackets> )

Second, "l-transalkoxylation / addition" compounds 14 can result from acid catalyzed alkoxide exchange at the anomeric centre in the intermediate pseudoglucals 4, with participation of the liberated

6045

alcohol or its lrimethylsilyl derivative in the F-3 CR of another glycal. Investigation of raw reaction mixtures as well as chromatographic fractions by NMR techniques has so far not led to proof of their occurrence under the given reaction conditions. However, compounds R2OTMS prove to be very reactive as nucleophiles with the 3-azido glycals 7, as shown by the high proportion of reaction products 8. The trend in reactivity can be logically and experimentally deduced for R 2 as beingprim.-alkyl >> sec.-alkyl, for mixtures of I with either 3B or 3C lead mostly to 8 and 3D, the latter TMS derivative not being sufficiently reactive.

Third, "3-transalkoxylatlon / addition" leading to 11, 12 and 13 results from a possible nucleophilic attack of alcohol or a trimethylsilyl derivative thereof at C-3 of pseudoglucals 4 or 5 with allylic rearrange- ment and a net exchange of two alkoxy functions at C-3 in comparing 1 and 6. This set of products has also so far not umambiguously been detected, most likely owing to the instability of 6 concerning revertion to 4 and 5 and conversion to the mixture of 7.

Table No. Gly- Eq. Eq.

cal Azide SbCis 1.4 y 1.20 2.20 1.4 2.01 1.78 1.4 2.01 1.78 1.2 2.04 g) 1.72 1.4 3.35 1.39

1 2 3 4 5 6 7 8 9

"J total eluted sugars after chromato mass-weighted average

1. Variation of Reaction Parameters Eq. Eq.

(BnS)2 2 2.20 1.47 1.77 1.18 1.77 1.18 1.77 1.15 2.93 0.93

1.4 3.35 1,33 2.93 0.89 1.2 2.01 g) 0.90 1.78 0.60 1.4 3.35 i 0,44 2.93 0.30 7.4 _k) 1,76 1.77 1.17

graphy, based on gJ molecular weight of h)

T [°el

-65- -85 -55--61 -39 - -44 -58- -65

-85- -100 -71- -88 -63- -65 -64- -82 -54- -66

Yield a) 1%1

_e)

63 _e)

63 _e)

74 81 _e)

201)

Ratio b) 1 6 b : 1 6 d : 7 c): 4 d ) : 1 0 a : 8 a 50 : _t~ :-~J: 5 0 : - 0 :_0

22 : 9 : - 0 " . 0 : 3 1 : 3 8 12 : 6 : _0 : 0 : 3 9 : 4 3

l lh): 5 : 2 6 : 1 7 1 ) : 2 5 : 1 6 35 : 1 7 : 2 3 : 1 2 : 4 : 9 24 : 1 5 : 1 2 : 8 : 1 5 : 2 6

33 h): 17:23J) :27i ) : -0 :_0

16 : 8 : 6 0 : 1 2 : - 0 : 4 _0

2.10 eq. of MeOH 3A also added additionally 15.2b formed, ratios 16.2b : 15.2b 77 : 23 (no. 7), 50 : 50 (no. 4) respective cited mixture (see also note b))

0 b) mass-% as determined by IH-NMR spectroscopy exclusive formation of a mixture of 5.2ct and 5.21~, ratio 78 : 22 c) (7A + 7B), typically only trace amounts of7C, 7D J) all isomers 7.2 detected, ratio A : B : C : D = 66 : 16 : 14 : 4 d) 4a, typically only trace amounts of 413 k) 2.0 eq. of preformed 3C s added c) not determined ') yield of isolated 9Aa, in addition 6% of cc-anomer of 9.4b 0 not detected within experimental error formed; partial hydrolysis of isopropylidene groups observed

Fourth, the "3-azidation / addition" leading to compounds 8, 9 and 10 accounts for the second major proportion o f products, namely a set of 3-azido-3-deoxy sugars. The reaction may formally be considered as a union of a two component azidation reaction first performed by Heyns and Guthrie 3~'4ab ("HG-2CR") with the F-3CR. Overall, a novel four component reaction results, namely HGF-4CR = HG-2CR u F-3CR. Thus, depending on the alcohol nucleophile employed, either sets of 3-azido monosaccharides or 3 "-azido-glycopy- ranosyl disaccharides may be obtained. However, in the case of sugar alcohols the reaction should be con- ducted in two steps with intermediate facile isolation of the azide, as the yields of disaccharides are otherwise very low (see "l-transalkoxylation / addition"). Interestingly, for glycals 1.4 the yield of 1-alkoxy product 8a is greater than of 1-azido product 10a, owing to the enhanced reactivity of intermediate BnOTMS. In the case of EtOTMS, the reactivity is also enhanced, albeit not to the same extent. Only in the case of 1.2 was an addition product 10.2b from 3-azido-D-glucal unambiguously detectable after chromatography on the basis of its distinct 3JI, 2 and 3J2, 3 coupling constantsl3.In the case of 1.4, IH NMR detection o f 10.4b is complicated by signal overlaps, H-1 being unassignable. Tentative NMR evidence suggests, however, that it is present. The diastereomeric ratio of 10.2a : 10.2b was not assigned, due to signal overlap. In the case of all compounds 8, no other isomers other than 8a were detectable, suggesting a highly stereoselective attack of 2 on 7A and a high selectivity for 7A vs. 713.

Generally, structural assignments were made on the basis of ~H NMR data and confirmed by 13C NMR as well as MS data 13. 2D-TOCSY-NMR spectra were obtained for representative compounds.

A major improvement in the 3-azidations of glycals is that the large excess of reagents used by Heyns and Guthrie may be scaled down by up to two orders of magnitude in achieving the conversion from 1 to 7 (Table 2). Thus, 0.1 eq. of SbCI 5 and 2.1 eq. of azide is enough to smoothly convert 1 to 712. In all cases, 7A

6046

and 7B are the predominant isomers, the maximum ratio for 7A : 7B achievable for R L = R 2 = Et and decreasing for spatially more demanding residues.

Table 2. Azidations o f Glucals No. GiucallProduct I Eq. I Eq.

1 I 7 I Azide [ SbCls 1 1.1 I 7.1 12.13 } 0.11 2 1.2 ] 7.2 2.08 [ 0.11 3 1.3 7.3 2.09 0.11 4 1.4 7.4 2.09 0.11

T [°C] [ Yield a) °6°11 -50 - -60 95 -50 - -60 98 -50 - -60 91

'~ isolated yield of all isomers combined after reaction time of ~J equilibrium value at r.t. 1.5 hours

b) as determined by ~H-NMR spectroscopy

Ratio b,O Ratio b,e) ] Rati oh'd) A : B : C : D A : B A : B

2 5 : 8 : 6 . 3 : 1 3 .1 :1 4 . 8 : 1 e ) 50 :9 : 9 : 1 5 .6 :1 -

5 . 5 : 2 . 2 : 1 . 1 : 1 2 .5 :1 2 .3 :1 1 0 . 8 : 7 . 5 : 2 . 5 : 1 1 .6 :1 I 1 .4 :1

e) equilibrium value after 3 h at +50°C o not determined

As can be envisioned, other electrophile / nucleophile combinations should also work well in furnishing analogous sets of 1,3-substituted glycopyranosyl compounds. We will report on these results.

References and Notes 1. a) Ugi, I.; D6mling, A.; H6rl, W. Endeavour 1994, 18, 115-122; b) Ugi, I.; D6mling, A.; HSrl, W. GITFachz.

Lab. 1994, 38, 430-437. 2. a) Helferich, B. Adv. Carbohydr. Chem. 1952, 7, 209-245; b) Ferrier, R. J. Adv. Carbohydr. Chem. 1965, 20,

67-137; c) Ferrier, R. J. ibid 1969, 24, 199-266; d) Fraser-Reid, B. Acc.Chem. Res. 1985, 18, 347-354. 3. see for example: a) Heyns, K.; Hohlweg, R. Chem. Ber. 1978, 111, 1632-1645; b) Thiem, J.; Karl, H.;

Schwentner, J. Synthesis, 1978, 696-698; c) Trost, B. M.; Shibata, T.; Martin, S. J. J.. Am. Chem. Soc. 1982, 104, 3228-3230; d) Jaurand, G; Beau, J.-M.; Sinai, P. J. Chem. Soc. Chem. Commun. 1982, 701-703; e) Friesen, R. W.; Danishefsky, S. J. J. Am. Chem. Soc. 1989, 111, 6656-6660; f) Halcomb, R. L.; Danishefsky, S. J. ibid. 1989, 111, 6661-6666; g) Wittman, M. D.; Halcomb, R. L.; Danishefsky, S. J. J . .~g. Chem. 1990, 55, 1979-1981; h) Grewal, G.; Kaila, N.; Franck, R. W. J. Org. Chem. 1992, 57, 2084-2092; 1) Ramesh, N. G.; Balasubramanian, K. K. Tetrahedron 1995, 51,255-272.

4. a) Guthrie, R. D.; Irvine, R. W. Carbohydr. Res. 1980, 82, 207-224; b) Guthrie, R. D.; Irvine, R. W. ibid. 1980, 82, 225-236; c) Mallams, A. K.; Wright, J. J. Ger. Often. 2, 747,946 (11.5.1978), CA 89:18031 lw,

5. Armstrong, R. W.; Sutherlin, D. P. Tetrahedron Lett. 1994, 35, 7743-7746. 6. a) Wing, D. R.; Rademacher, T. W.; Field, M. C,; Dwek, R. A.; Schmitz, B.; Thor, G.; Schachner, M.

Glycoconjugate J.. 1992, 9, 293-301; b) Patel, T. P.; Goelz, S. E.; Lobb, R. R.; Parekh, R. B. Biochemistry 1994, 33, 14815-14824.

7. a) Wong, C.-H.; Halcomb, R. L.; Ichikawa, Y.; Kajimoto, T. Angew. Chem. 1995, 107, 453-474 (Part 1); Wong, C.-H.; Halcomb, R. L.; Ichikawa, Y.; Kajimoto, T. ibid. 1995, 107, 569-593 (Part 2);

8. Thiem, J.; Klaftke, W. J Or g. Chem. 1989, 54, 2006-2009. 9. Chmielenski, M.; Fokt, I.; Grodner, J.; Grynkiewicz, G.; Szeja, W ~ Carbohydr. Chem. 1989, 8, 735 10. a) Brimacombe, J. S.; Jones, B. D.; Stacey, M; Willard, J. J. Carbohydr. Res. 1966, 2, 167-; b) Goebel, M.; Ugi,

I. Synthesis 1991, 1095-1098; c) Goebel, M. O-peralkylierte Glycopyranosylamine [...] in Peptidsynthesen mittels stereoselektiver Vierkomponentenkondensationen, Technical University of Munich 1994. d) Extraction of 1.1-1.3 from the raw mixture was accomplished with hexanes, furnishing essentially pure product after in vacuo evaporation of residual DMF.

11. Typical procedure for glucal reactions with 2 and trimethylsilyl azide: Under an inert gas blanket are added the desired nucleophiles in the quantities stated in table 1 to a solution of 3.6 mmol of 1 in dry CH2C12(20 ml) and cooled to the desired reaction temperature. Meanwhile, a solution of 2 is prepared by dissolving the respective quantity of dibenzyl disulfide under an inert gas blanket in CH2CI2 (20 m] / 6.4 mmol) followed by cooling to -60°C and addition of the desired amount of a commercial IM solution of SbC15 in CHIC12. The blood red solution is stirred for 0.5 hours and then added under inert gas to the solution of 1. Stirring is continued for one hour at the same temperature after which sat. NaHCO~ solution (100 ml) is added and the mixture shaken until disappearance of the orange color. The mixture is stirred until evolution of gas ceases, diluted with CH2C12, extracted twice with IN NaOH, then water and dried with Na2SO4. Filtration and evaporation of solvent in vacuo delivers the raw product mixtures which may be fractionated by chromatography (SIO2 or basic A1203, hexanes-ether as eluent).

12. Azidations ofglucals are carried out in analogy to the preceding reference, addition of SbCI 5 to the glueal solution at -60°Crquantities and reaction times according3to table 2, extraction only wit9 water.

13. AnalyticalDqta: H-NMR(360MHz):16.21b:4.56(d, lH J12=9,7, H-I)2.46(dd, lH, J23=ll.l,H-2);16~4b: 4.66 (d 1H JJl_2 = 9.7 H-I) 2.623(dd. IH JJ2.3 = 10.7, H-2)" 16.2~: 5.30 (d IH ~Jl 2 = 2.7, iq-1), 2.93 (~, 1H, ~J2.3 = 4.4 FI-2)~ 16.4d: ~.31 (d,' 1H J, 2 '= 2~5, H-I) 2.96 (d~,' IH, J23 = 4.3~ H-2); I0.2a: 35.17 (d, 1H, J12 = 1.8, H-1)'2.85 (4d 1H J23 = 4.0 iq-2j" 10.4a.: 5.21(d IH Ji 2 = 1.8' H-I), 2.90,(dd, IH, J23 = 4.2, H-2), 10~2b: 4.61 (d, 1H, Ji.2 = 9.7, l-l-I), 2~,2(dd, IH, Jz3 = 11.3, H-2)" ~k2a: 5 24 ( d , . , IH, Jj 2 = 2 . 2 , , H - I ) , ~.89 (dd, IH, ,~3 = 4.0, H-2); 8.4a: 5.23 (d, IH, JiR = 2.2, H-l), 2.86 (dd, 1H, J23 = ~.3,H-2); 9.4a: 5.83 (d, 1H, J~..2 = 3.5, H-I ), 5.14 (bs 1H, H-I), 3.04 (dd, 1H, J23 = 3.5 H-2); 9.4b: 5.79 (d, ]H, Jv,2 = 3.5, H-I'), 5.17 (bs, 1H, H-I)

Support by the Deutsche Forschungsgemeinschaft is gratefully acknowledged

We wish to dedicate this communication to Prof. Hans-Dieter Scharf on the occasion o f his 65th birthday.

(Received in Germany 23 May 1995; accepted 27 June 1995)