nature CHeMICaL BIOLOGY | vol 6 | march 2010 | www.nature.com/naturechemicalbiology 199

brief communicationpuBLIsHed OnLIne: 24 januarY 2010 | dOI: 10.1038/nCHeMBIO.302

Herein we describe the synthesis of highly substituted chromans and isochromans using carbohydrates as starting materials. Our approach makes use of a Pd-catalyzed domino reaction consisting of oxidative addition, followed by two carbopalladation steps and completed by a cyclization to annelate the benzene moiety. The versatility of this route has been demonstrated by a small library of highly substituted chromans and isochromans.

Chroman and isochroman derivatives are important classes of oxygenated heterocycles showing a broad range of biological and pharmaceutical activity1,2. These structural motifs are widespread elements in a plethora of different natural products2–4. Therefore, numerous synthetic strategies for the preparation of these scaffolds have been developed5–7. Among the different methods, reactions of salicylic aldehydes with α,β-unsaturated carbonyl compounds8,9 and transition metal–catalyzed coupling reactions10–14 are very prominent ones. The large majority of these methods use an appropriately func-tionalized aromatic cycle as starting material in order to annelate the pyran system.

In this communication, we report a highly efficient approach to chromans 1a–1n and isochromans 5a, 5b and 5c that uses a suitably substituted 2-bromoglycal (3 and 7) as starting material (Scheme 1). The latter is easily available from corresponding monosaccharides (10) and therefore allows the facile introduction of defined stereocenters on the pyran unit. The benzene moiety is created in a domino reaction15–19 that uses a diyne that is attached to the pyran core20–22. A detailed retrosynthetic analysis of this strategy leading to monosaccharides of type 10 is depicted in Scheme 1. Such a procedure affords the synthesis of highly substituted chroman and isochroman derivatives that are difficult to obtain by other routes. The target molecules can also be considered as hybrids between carbohydrates and aromatic compounds.

We began our studies of the synthesis of chromans by prepar-ing a number of 2-bromoglycals (2a–2n) with different diyne chains attached to the sugar core (Table 1). As carbohydrate precursors, we used hexoses such as D-glucose (2a–2g), L-rhamnose (2h–2j) and D-galactose (2k–2l) as well as the pentose D-arabinose (2m and 2n). The corresponding 2-bromoglycals are available via the respective glycals starting from the native monosaccharides or commercially available glycals in a few steps (Supplementary Results). After install-ing the appropriate protecting group pattern, we attached the diyne chain of type 4 to the remaining hydroxyl group in position 3 by using the typical reaction conditions for propargylic substitution reactions to afford the precursors 2a–2n (Supplementary Results). In order to investigate the scope of this modular approach, we prepared a variety of different monosaccharides and different diyne chains (Table 1).

With these precursors in hand, we screened different Pd cat-alysts and reaction conditions for the desired domino sequence using precursor 2c as a test substrate (Supplementary Results).

We obtained the best results for furnishing the domino product 1c in a mixture of dimethylformamide (DMF), acetonitrile and N-methylpyrrolidone (8:8:1) with Pd(PPh3)4 and diisopropylamine as base using microwave irradiation. The domino reaction itself proceeded in moderate to excellent yields (25% to quantitative yield), tolerating a variety of modifications of the diyne chain. Alkyl, aryl and silyl substituents were tolerated, and even terminal triple bond systems (R′ = H) led to the desired products. Oxygen moieties in the tether (2e and 2h) decreased the yield, whereas further substituents such as geminal ester moieties (2d and 2j) had no influence. However, attempts to extend this reaction to nitrile 2f that would afford a pyridine moiety (commonly built up via Co-mediated cyclizations; see for example ref. 23) attached to the sugar core were not successful. This failure may be attributed either to a coordination of a nitrile or to unfavorable electronic interac-tion with the Pd species. Deprotection reactions using either acid for isopropylidene and benzylidene groups or hydrogenolysis cleav-ing benzyl moieties afforded chromans 11a–11n with the native hydroxyl group pattern of the respective carbohydrates.

domino access to highly substituted chromans and isochromans from carbohydratesMarkus Leibeling, dennis C Koester, Martin pawliczek, svenia C schild & daniel B Werz

Institut für organische und Biomolekulare chemie, Georg-august-Universität Göttingen, Göttingen, Germany. *e-mail: dwerz@gwdg.de

O

O

R′

XO

O

R′

X

O

Br+ +

OH

O

O

Br

X

R′

X

R′

HO

X

R′

Hal

O OH

OH

OH

HO

OH

O

Br

OAc

O

OAc

Chroman Isochroman

1

2

5

743 8

9

10

O

O

Br

X

R′6

(RO)n

(RO)n

(RO)n

(AcO)n

(AcO)n

(AcO)n

(AcO)n

Scheme 1 | Retrosynthetic analysis of chromans and isochromans leading to monosaccharides of type 10 as starting materials. The chroman and isochroman cores are shown in bold.

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

200 nature CHeMICaL BIOLOGY | vol 6 | march 2010 | www.nature.com/naturechemicalbiology

brief communication NaTuRe cHemical biOlOgy dOI: 10.1038/nCHeMBIO.302

Because the formation of the starting materials 2a–2n is performed in DMF under basic conditions, we became interested in testing the notion that the complete synthesis from 3 and 4 to chromans 1a–1n could be streamlined in a one-pot operation. In principle, our experiments have revealed that such an approach

without isolation of the sugar diyne intermediates 2a–2n is possible, as exemplified by the direct transformation of the correspond-ing bromoglycal and propargyl iodide to afford 1a in 42% yield (Supplementary Results); however, higher overall yields (66% over 2 steps) for 1a were obtained by separating the diyne

Table 1 | carbohydrate diyne derivatives, chroman products, deprotected chromans and corresponding yields of the domino sequence and the deprotection

O

O

Br

2

X

R′

(RO)n

Pd(PPh3)4O

O

X

R′

(RO)n

1

Deprotection

(Hydrolysis orhydrogenolysis)

O

O

X

R′

(HO)n

11

HN(iPr)2DMF, MeCN,

NMP120 °C, mw

O

O

O

O

Br

2a

O

RO

O

RO

1a (R,R = CMe2)11a (R = H)

1b (R,R = CMe2)11b (R = H)

1c (R,R = CMe2)11c (R = H)

1d (R,R = CMe2)11d (R = H)

1e (R,R = CMe2)11e (R = H)

1g (R,R = CMe2)11g (R = H)

O

BnO

O

Br

2h

7087

Quant.83

8175

9288

8889

8633

6768

5299

7557

92

No reaction

–c56b

76

– 38b

25b

–c

OO

RO

O

1h (R = Bn)11h (R = H)

1i (R = Bn)11i (R = H)

1j (R = Bn)11j (R = H)

O

O

O

O

O

Br

2b

Ph

O

RO

O

RO

Ph

O

BnO

O

Br

2i

PhO

RO

O

Ph

O

O

O

O

Br

2c

O

RO

O

ROO

BnO

O

Br

2j

Ph

O

RO

O

Ph

O

O

O

O

Br

2d

Ph

CO2Me

CO2Me CO2Me

CO2Me

CO2Me

CO2Me

O

RO

O

RO

Ph

O

O

O

O

Br

2k

Ph

O

RO

O

RO

1k (R,R = CHPh)11k (R = H)

O

O

O

O

Br

O

2e

O

RO

O

ORO

O

O

O

O

Br

2l

Ph

Ph

O

RO

O

RO

1l (R,R = CHPh)11l (R = H)

Ph

O

O

O

O

Br

N

2f

O

HO

O

Br

2m

O

HO

O

1m

O

O

O

O

Br

2g

Me3Si

O

RO

O

RO

SiMe3

O

BnO

O

Br

2n

Ph

O

RO

O

Ph

1n (R = Bn)11n (R = H)

CO2Me

CO2Me

Starting material Product Yield(%)a Starting material Product Yield

(%)a

aThe first value is the yield of the domino reaction; the second value is the yield of the deprotection. bThe yield is given for a two-step procedure starting from 2-bromoglycal and diynyl halide. chydrolysis (of 1g) and hydrogenolysis (of 1h) led to decomposition.

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.

nature CHeMICaL BIOLOGY | vol 6 | march 2010 | www.nature.com/naturechemicalbiology 201

brief communicationNaTuRe cHemical biOlOgy dOI: 10.1038/nCHeMBIO.302

attachment from the Pd-catalyzed reaction. Attempts to perform the diyne attachment by a Pd-catalyzed propargylic substitution to the hydroxyl group—a reaction that is well known for soft nucleophiles24—were not successful.

The catalytic cascade that explains the formation of the chro-mans and isochromans is initiated by the oxidative addition of the Pd0 species into the C-Br bond of the bromosugar (Supplementary Results). A first intramolecular carbopalladation of the resulting Pd species to the triple bond sequenced by a further carbopalla-dation of the remaining alkyne moiety affords a triene. The final cyclization step generating the benzene unit may be regarded as electrocyclic 6π electron ring closure of that triene followed by the subsequent release of the catalytic species. As a final step, we cannot rule out a Heck-type transformation leading to the desired product; however, an anti-dehydropalladation—that is possible due to the resulting aromatization, but rarely observed—would have to occur. Mechanistically, we further sought to determine whether the final ring closure or a β-hydride elimination is faster when alkyl sub-stituents at the triple bond are used. Therefore, we also synthesized derivative 2c with a CD3 instead of a CH3 group (Supplementary Results). If a β-hydride elimination takes place, one would observe a displacement of one deuterium atom by hydrogen. However, our experiments revealed no scrambling, leading to the assumption that the cyclization step to the benzene moiety is faster than a potential β-hydride elimination.

Encouraged by the results shown above, we tried to extend this domino approach to the preparation of isochromans. Because peracetylated 2-bromoglucal and 2-bromogalactal derivatives of type 7 are precursors in the synthesis of chromans 1a–1n, we also used them as starting materials for isochroman synthesis. A Lewis acid–catalyzed Ferrier reaction25 resulting in a shift of the C=C double bond in the six-membered ring furnished the correspond-ing 2-bromoglycals (6a, 6b and 6c) with a diyne chain attached to the anomeric position (Supplementary Results). In all cases the α-anomer was the major product, and only traces of the β-anomer were detected that could easily be removed by silica gel column chromatography. The obtained 2-bromoglycals 6a, 6b and 6c were converted by using the above procedure to the respective isochro-mans 5a, 5b and 5c (Scheme 2). However, in comparison to the domino reaction leading to chromans 1a–1n, yields for the respective isochromans (ranging from 20% to 65%) were only moderate.

In summary, we have achieved a concise and robust route to highly substituted chromans and isochromans by a Pd-catalyzed domino approach. Combined with the use of monosaccharides (and their rich stereochemistry) as starting materials, our synthetic approach allows the generation of complex oligocyclic structures with little effort. The facile variation of the carbohydrate part on the one hand and the diyne part on the other hand leads to a high degree of structural diversity. Our synthetic strategy should give access to many hybrids of carbohydrates and aromatics that may serve as useful candidates for biological and pharmaceutical studies.

received 19 october 2009; accepted 28 november 2009; published online 24 January 2010

references1. Ellis, G.P. & Lockhart, I.M. The Chemistry of Heterocyclic Compounds, Chromenes,

Chromanones, and Chromones (Wiley-VCH, New York, 2007).2. Shen, H.C. Tetrahedron 65, 3931–3952 (2009).3. Nicolaou, K.C. et al. J. Am. Chem. Soc. 122, 9968–9976 (2000).4. Trost, B.M., Shen, H.C., Dong, L., Surivet, J.-P. & Sylvain, C. J. Am. Chem. Soc.

126, 11966–11983 (2004).5. van Lingen, H.L., Zhuang, W., Hansen, T., Rutjes, F.P.J.T. & Jørgensen, K.A Org.

Biomol. Chem. 1, 1953–1958 (2003).6. Zu, L., Zhang, S., Xie, H. & Wang, W. Org. Lett. 11, 1627–1630 (2009).7. Hong, L., Wang, L., Sun, W., Wong, K. & Wang, R. J. Org. Chem. 74, 6881–6884

(2009).8. Shi, Y.-L. & Shi, M. Org. Biomol. Chem. 5, 1499–1504 (2007).9. Liu, K., Chougnet, A. & Woggon, W.-D. Angew. Chem. Int. Ed. 47, 5827–5829

(2008).10. Fukamizu, K., Miyake, Y. & Nishibayashi, Y. J. Am. Chem. Soc. 130, 10498–10499

(2008).11. Yamamoto, Y. & Itonaga, K. Org. Lett. 11, 717–720 (2009).12. Wegner, H.A., Ahles, S. & Neuburger, M. Chem. Eur. J. 14, 11310–11313

(2008).13. Hashmi, A.S.K. et al. Chem. Eur. J. 14, 6672–6678 (2008).14. Tietze, L.F., Burkhardt, O. & Henrich, M. Liebigs Ann./Recueil 887–891

(1997).15. Tietze, L.F., Brasche, G. & Gericke, K.M. Domino Reactions in Organic Synthesis

(Wiley-VCH, Weinheim, Germany, 2006).16. Tietze, L.F. Chem. Rev. 96, 115–136 (1996).17. Enders, D., Huttl, M.R.M., Grondal, C. & Raabe, G. Nature 441, 861–863

(2006).18. Meng, X. et al. Org. Lett. 11, 991–994 (2009).19. Cui, S.-L., Wang, J. & Wang, Y.-G. J. Am. Chem. Soc. 130, 13526–13527

(2008).20. Meyer, F.E. & de Meijere, A. Synlett 777–778 (1991).21. Negishi, E., Harring, L.S., Owczarczyk, Z., Mohamud, M.M. & Ay, M. Tetrahedr.

Lett. 33, 3253–3256 (1992).22. Blond, G., Bour, C., Salem, B. & Suffert, J. Org. Lett. 10, 1075–1078 (2008).23. Zhou, Y., Porco, J.A. Jr. & Snyder, J.K. Org. Lett. 9, 393–396 (2007).24. Tsuji, J., Watanabe, I., Minami, I. & Shimizu, I. J. Am. Chem. Soc. 107, 2196–2198

(1985).25. Ferrier, R.J., Overend, W.G. & Ryan, A.E. J. Chem. Soc. 3667–3670 (1962).

acknowledgmentsWe are grateful to the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie for financial support (Emmy Noether Fellowship and Liebig Fellowship to D.B.W.). D.C.K. thanks the Studienstiftung des deutschen Volkes for his undergraduate fellowship and the Fonds der Chemischen Industrie for his PhD fellowship. S.C.S. acknowledges the Konrad-Adenauer-Stiftung for her undergraduate fellowship. We thank L.F. Tietze (University of Göttingen) for helpful discussions and generous support of our work.

author contributionsD.B.W. designed the project. M.L., D.C.K., M.P. and S.C.S. performed the experiments. The manuscript was written by D.B.W.

Competing interests statementThe authors declare no competing financial interests.

additional informationSupplementary information and chemical compound information is available online at http://www.nature.com/naturechemicalbiology/. Reprints and permissions information is available online at http://npg.nature.com/reprintsandpermissions/. Correspondence and requests for materials should be addressed to D.B.W.

Scheme 2 | Synthesis of isochromans starting from peracetylated 2-bromoglycals. The first step involves an acid-catalyzed Ferrier reaction to afford 6a, 6b and 6c. The second step involves the Pd-catalyzed domino reaction leading to 5a, 5b and 5c. Three isochromans (5a, 5b and 5c) synthesized by this route and the corresponding yields of the domino reaction are depicted. NmP, N-methyl-2-pyrrolidone; DmF, dimethylformamide; mecN, acetonitrile; hN(iPr)2, diisopropylamine; mw, microwave.

O

O

OAcOAc

O

O

OAcOAc Ph

O

O

OAcOAc H

5a (20%) 5b (41%) 5c (65%)

Pd(PPh3)4

5

O

O

(AcO)n R

O

(AcO)n

Br

OAc

R

HO

+

.

CH2Cl2

67 8

O

O

(AcO)n

Br

R

BF3 OEt2

H

HN(i Pr)2DMF, MeCN,

NMP120 °C, mw

© 2

010

Nat

ure

Am

eric

a, In

c. A

ll ri

gh

ts r

eser

ved

.