Efficient sialyltransferase inhibitors based on transition-state analogues of the sialyl donor

Full text of DOI:10.1002/(SICI)1521-3773(19981102)37:20<2893::AID-ANIE2893>3.0.CO;2-W

COMMUNICATIONS
measured in 0.1n KOAc/methanol and 1.0n NH4OAc/methanol solutions
at 258C and 60000 rpm; both protocols gave similar results. In pure
methanol the sample sedimented very slowly as a result of electrostatic
repulsion. The partial specific volume of 0.6305 mLg 1 was used.[16]
Efficient Sialyltransferase Inhibitors Based
on Transition-State Analogues of the Sialyl
Donor**
Bernd Müller, Christoph Schaub, and
Richard R. Schmidt*
Received: October 7, 1997
Revised version: June 29, 1998 [Z11007IE]
German version: Angew. Chem. 1998, 110, 3058 ± 3061
Sialic acid containing epitopes are involved in important
biological processes, such as cell adhesion and inflammation.[1]
There is also a correlation between the sialyl content of
glycoconjugates and the malignancy of tumor cells.[2] Fur-
thermore, differences in sialylation type between tumor cells
and normal cells were found;[3, 4] for instance, although N-
glycolylneuraminic acid has not been observed thus far on the
surface of normal human cells, 30 ± 50% of tumor cells of
different origin contain this compound in small amounts.[3]
Recently, an interesting correlation between a(2-6)-sialyla-
tion of N-acetyllactosamine and B lymphocyte activation and
immune function was reported, which could find medicinal
application.[5] Therefore, to study the influence of sialyl
residues in biological systems it is highly desirable to develop
efficient inhibitors for sialyltransferases.
Keywords: N ligands ´ polymers ´ supramolecular chemistry
´ thin films
[1] a) J.-M. Lehn, Supramolecular Chemistry, VCH, Weinheim, 1995; b) C.
Floriani, A. E. Merbach in Perspectives in Coordination Chemistry
(Ed.: A. F. Williams), VCH, Weinheim, 1992.
[2] V. Balzani, F. Scandola, Supramolecular Photochemistry, Ellis Hor-
wood, New York, 1991.
[3] Reviews: a) A. Ulman, An Introduction to Organic Thin FilmsÐFrom
Langmuir ± Blodgett to Self-Assembly, Academic Press, San Diego,
CA, 1991; b) J. D. Swalen, D. L. Allara, J. D. Andrade, E. A.
Chandross, S. Garoff, J. Israelachvili, T. J. McCarthy, R. Murray,
R. F. Pease, J. F. Rabolt, K. J. Wynne, H. Yu, Langmuir 1987, 3, 932 ±
950; c) A. R. Bishop, R. G. Nuzzo, Curr. Opin. Coll. Interf. Sci. 1996, 1,
127 ± 136; d) H. Kuhn, D. Möbius in Investigations of Surfaces and
Interfaces, Part B Methods of Chemistry Series, Vol. IXB (Eds.: B. W.
Rossiter, R. C. Baetzold), 2nd ed., Wiley, New York, 1993, pp. 375 ±
542.
The various sialyltransferases employ, independent of their
source and their acceptor specificity, cytidine monophosphate
N-acetylneuraminic acid (CMP-Neu5Ac, Scheme 1) as sialyl
[4] G. Decher in Comprehensive Supramolecular Chemistry, Vol. 9 (Eds.:
J. L. Atwood, J. E. D. Davies, D. D. MacNicol, F. Vögtle), Pergamon,
Oxford, 1996, pp. 507 ± 528.
[5] T. E. Mallouck, H.-N. Kim, P. J. Ollivier, S. W. Keller in Comprehen-
sive Supramolecular Chemistry, Vol. 7 (Eds.: J. L. Atwood, J. E. D.
Davies, D. D. MacNicol, F. Vögtle), Pergamon, Oxford, 1996,
pp. 189 ± 217.
[6] a) E. C. Constable, A. M. W. Cargill Thompson, D. A. Tocher in
Supramolecular Chemistry (Eds.: V. Balzani, L. De Cola), Kluwer,
Dordrecht, 1992, pp. 219 ± 233; b) E. C. Constable, Pure Appl. Chem.
1996, 68, 253 ± 260.
[7] Other examples of coordination polyelectrolytes: a) U. Velten, B.
Lahn, M. Rehahn, Macromol. Chem. Phys. 1997, 198, 2789 ± 2816;
b) E. C. Constable, A. J. Edwards, D. Phillips, P. R. Raithby, Supramol.
Chem. 1995, 5, 93 ± 95.
[8] Examples of thin films containing metal ions: a) G. Caminati, D. Berti,
G. Gabrielli, S. Leporatti, R. Rolandi, M. G. Ponzi-Bossi, B. Yang,
Thin Solid Films 1996, 284 ± 285, 181 ± 186; b) H. Huesmann, C. A.
Bignozzi, M. T. Indelli, L. Oavanin, M. A. Rampi, D. Möbius, Thin
Solid Films 1996, 285, 62 ± 65; c) T. R. Vierheller, M. D. Foster, A.
Schmidt, K. Mathauer, W. Knoll, G. Wegner, S. Statija, C. F.
Maijkrzak, Langmuir 1997, 13, 1712 ± 1717; d) M. Maskus, H. D.
Abruna, Langmuir 1996, 12, 4455 ± 4462; e) Y. Liang, R. H. Schmehl,
J. Chem. Soc. Chem. Commun. 1995, 1007 ± 1008.
[9] E. C. Constable, Adv. Inorg. Chem. Radiochem. 1986, 30, 69 ± 121.
[10] a) G. D. Storrier, S. B. Colbran, D. C. Craig, J. Chem. Soc. Dalton
Trans. 1997, 3011 ± 3028; b) E. C. Constable, A. M. W. C. Thompson, J.
Chem. Soc. Dalton Trans. 1992, 3467 ± 3475.
[11] J. B. Lambert, H. F. Shurvell, D. A. Lightner, R. G. Cooks, Introduc-
tion to Organic Spectroscopy, Macmillan, New York, 1987, pp. 249 ±
268.
[12] E. C. Constable, A. M. W. Cargill Thompson, J. Chem. Soc. Dalton
Trans. 1992, 3467 ± 3475.
Scheme 1. Mechanism of the sialylation.
[*] Prof. Dr. R. R. Schmidt, Dipl.-Chem. B. Müller,
Dipl.-Chem. C. Schaub
[13] W. Spahni, G. Calzaferri, Helv. Chim. Acta 1984, 67, 450 ± 454.
[14] a) G. Decher, Y. Lvov, J. Schmitt, Thin Solid Films 1994, 244, 772 ± 777;
b) G. Decher, J. D. Hong, J. Schmitt, Thin Solid Films 1992, 210/211,
831 ± 835, c) Y. Lvov, F. Essler, G. Decher, J. Phys. Chem. 1993, 97,
13773 ± 13777.
Fakultät für Chemie der Universität, Fach M725
D-78457 Konstanz (Germany)
Fax: (‡49)7531-88-3135
[**] This work was supported by the Deutsche Forschungsgemeinschaft
and the Fonds der Chemischen Industrie. C.S. gratefully acknowl-
edges a fellowship from the Fonds der Chemischen Industrie. We are
grateful to Dr. A. Geyer for his help in the structural assignments by
NMR experiments.
[15] a) A. Asmussen, H. Riegler, J. Chem. Phys. 1996, 104, 8159 ± 8164;
b) electron densities were taken from T. P. Russel, Mater. Sci. Rep.
1990, 5, 171 ± 271.
[16] The density of [Fe(tpy)2]2‡ was used in the calculations: A. T. Baker,
H. A. Goodwin, Aust. J. Chem. 1985, 38, 207 ± 214.
Angew. Chem. Int. Ed. 1998, 37, No. 20
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3720-2893 $ 17.50+.50/0
2893

COMMUNICATIONS
donor.[6] Analogues derived from the transition state of the
proposed reaction course (CMP-Neu5Ac= in Scheme 1)[7, 8]
could exhibit high enzyme affinity and thus lead to partic-
ularly efficient inhibitors. So far, only a few donor and
acceptor analogues (ˆsubstrate analogues) which serve as
sialyltransferase inhibitors have been reported;[7±13] however,
their potency is only in the micromolar range and thus close to
that of the substrate CMP-Neu5Ac, which binds to a(2-6)-
sialyltransferase from rat liver (EC 2.4.99.1) with KM ˆ 46 mm
(Table 1).[7] Therefore, we have initiated a program on the
To investigate the first question, compounds 3 and 4 were
prepared (Scheme 2). To this end, benzaldehyde and furfural
were transformed with dibenzyl phosphonate into the corre-
sponding a-hydroxyphosphonates[16] (racemic mixtures).
Table 1. Affinity of CMP-Neu5Ac (KM) to a(2-6)-sialyltransferase of rat
liver and inhibition constants (Ki) of transition-state analogues (R)-2, 3h,
3l, 4h, 4l, (R)-6, (S)-6, (R)-7, (S)-7, (E)-8, and (Z)-9.[a]
KM [mm]
Ki [mm]
Ref.
CMP-Neu5Ac
(R)-2
3h
3l
4h
46 Æ 7
±
[7]
[8]
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
0.35 Æ 0.05
0.20 Æ 0.05
1.0 Æ 0.2
0.28 Æ 0.06
1.0 Æ 0.3
10 Æ 2
4l
(R)-6
(S)-6
(R)-7
(S)-7
(E)-8
(Z)-9
7 Æ 2
15 Æ 3
23 Æ 5
±
[8]
±
6.0 Æ 0.5
0.04 Æ 0.008
[a] For details see ref. [7a].
synthesis of transition-state analogues.[7, 14] Compounds 1 and
2 were derived from the transition-state model CMP-Neu5-
Ac= (Scheme 1), which is deduced from the recently support-
ed SN1-type mechanism;[7, 8, 15] they were considered as
Scheme 2. Synthesis of 3 and 4. Bn ˆ benzyl.
Their condensation with cytidine-phosphitamide 5[17] in the
presence of tetrazole followed by oxidation with tert-butyl
hydroperoxide and base-catalyzed cleavage of the cyanoethyl
group afforded the target molecules 3h, 3l and 4h, 4l
(Table 2) after hydrogenolytic debenzylation, deacylation,
chromatography over RP-18 silica gel (Et3NH ´ HCO3 buffer
‡
as eluent), and ion exchange with IR-120 (Na ); the
diastereoisomers were separated based on their different Rf
values (high, low); the configuration of the new stereogenic
center was not assigned. Compounds 6 and 7, in which the
phosphonate residue is replaced by a carboxylate group, can
be readily obtained from methyl (R)- and (S)-mandelate and
methyl (R)- and (S)-phenyllactate in a similar fashion
(Scheme 3, Table 2).
As can be seen in Table 1, phenyl derivative 3h exhibits an
even higher binding affinity to a(2-6)-sialyltransferase than
does neuraminyl derivative (R)-2, and furyl derivative 4h is
also a very good competitive inhibitor.[18] Apparently, the
configuration at the carbon atom bearing the CMP moiety has
no dramatic influence on inhibition (Ki values: 3h:3l ꢀ 1:5;
4h:4l ꢀ 1:4). Replacement of the phosphonate group by a
carboxylate group as in (R)- and (S)-6 and in (R)- and (S)-7
decreases the binding affinity by about one order of magni-
tude; again the configuration at the carbon atom bearing the
potential inhibitors because the distance between the ano-
meric center (C-2 of the Neu5Ac residue) and the leaving
group (CMP) is increased, and thus similar to the distance in
the transition-state model (CMP-Neu5Ac=). In particular
compound 2[8] was of interest, because it contains two negative
charges separated by five bonds, as in CMP-Neu5Ac. In
addition, the anomeric center is trigonal planar, thus favoring
a conformation assumed for a SN1-type transition state. This
concept turned out to be very fruitful, because (R)-2 is a very
potent inhibitor which binds in the nanomolar range to a(2-
6)-sialyltransferase (Table 1).[8] This result led to two ques-
tions: 1) Is the neuraminyl residue as shown in (R)-2 required
at all or can it be replaced by a simple aryl, hetaryl, or related
moiety, thus making these type of compounds readily
accessible? 2) Will a different geometry of the substituents
around the anomeric center increase the binding affinity?
2894
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3720-2894 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1998, 37, No. 20

COMMUNICATIONS
Table 2. Selected physical data of 3h, 3l, 4h, 4l, (R)-6, (S)-6, (R)-7, (S)-7,
and (Z)-9.[a]
3h: HPLC (2% CH3CN, 10 mLmin 1): tR ˆ 10.1 min; [a]2D0
ˆ
37.9 (c ˆ
1
0.8, H2O); H NMR: d ˆ 3.38 (brd, 1H; 4'-H), 3.60 (m, 1H; 5a'-H), 3.81 ±
3.95 (m, 3H; 2'-, 3'-, 5b'-H), 5.04 (dd, 3J(1'',P) ˆ 13.0 Hz, 2J(1'',P) ˆ 11.0 Hz,
1H; 1''-H); MS: m/z ˆ 494 [M 3Na‡2H] , 559.30 for C16H18N3Na3O11P2
3l: HPLC (2% CH3CN, 10 mLmin 1): tR ˆ 12.5 min; [a]2D0 ˆ ‡23.3 (c ˆ 0.8,
H2O); 1H NMR: d ˆ 3.66 ± 3.95 (m, 5H; 2'-, 3'-, 4'-, 5a,b'-H), 5.02 (dd,
3J(1'',P) ˆ 14.0 Hz, 2J(1'',P) ˆ 10.0 Hz, 1H; 1''-H); MS: m/z ˆ 494 [M
3Na‡2H] , 559.30 for C16H18N3Na3O11P2
4h: HPLC (1% CH3CN, 8 mLmin 1): tR ˆ 15.7 min; 1H NMR: d ˆ 3.5 ± 4.0
(m, 5H; 2'-, 3'-, 4'-, 5a,b'-H), 5.04 (dd, 3J(1'',P) ˆ 15 Hz, 2J(1'',P) ˆ 9.2 Hz,
3
1H; 1''-H); 31P NMR: d ˆ 0.73 (d, J(P,P) ˆ 33.4 Hz; phosphate), 12.56 (d;
phosphonate); MS: m/z ˆ 486 [M 2Na‡H H2O] , 504 [M 2Na‡H] ,
525 [M Na] , 549.2 for C14H16N3Na3O12P
4l: HPLC (1% CH3CN, 8 mLmin 1): tR ˆ 19.9 min; 1H NMR: d ˆ 3.65 ±
3
3.82 (m, 2H; 5a,b'-H), 3.9 ± 4.0 (m, 3H; 2'-, 3'-, 4'-H), 5.03 (dd, J(1'',P) ˆ
15.1 Hz, 2J(1'',P) ˆ 9.5 Hz, 1H; 1''-H); 31P NMR: d ˆ 1.07 (d, 3J(P,P) ˆ
33.4 Hz; phosphate), 12.49 (d; phosphonate); MS: m/z ˆ 486 [M
2Na‡H H2O] , 504 [M 2Na‡H] , 525 [M Na] , 549.2 for
C14H16N3Na3O12P
(R)-6: HPLC (3% CH3CN, 8 mLmin 1): tR ˆ 13.67 min; [a]2D0
ˆ
41 (c ˆ
1.09, H2O); 1H NMR: d ˆ 3.55 ± 3.64 (m, 2H; 5a,b'-H), 3.80 ± 3.92 (m, 3H;
2'-, 3''-, 4'-H), 5.19 (d, 3J(1'',P) ˆ 8.9 Hz, 1H; 1''-H); 31P NMR: d ˆ 0.13 (s;
phosphate); MS: m/z ˆ 455 [M 2Na‡H] , 501.3 for C17H18N3Na2O10P
(S)-6: HPLC (3% CH3CN, 8 mLmin 1): tR ˆ 21.32 min; [a]2D0 ˆ ‡37.6 (c ˆ
0.5, H2O); 1H NMR: d ˆ 3.70 ± 3.91 (m, 5H; 2'-, 3'-, 4'- 5a,b'-H), 5.16 (d,
3J(1'',P) ˆ 9.1 Hz, 1H; 1''-H); 31P NMR: d ˆ 0.20 (s; phosphate); MS:
m/z ˆ 455 [M 2Na‡H] , 501.3 for C17H18N3Na2O10P
Scheme 3. Synthesis of 6 and 7.
(R)-7: TLC: Rf ˆ 0.81; [a]2D0 ˆ ‡5.5 (c ˆ 1, H2O); H NMR: d ˆ 2.80 ± 2.97
1
(m, 2H; 2a,b''-H), 3.50 ± 3.58 (m, 1H; 5a'-H), 3.80 ± 3.94 (m, 4H; 2'-, 3'-, 4'-,
5b'-H), 4.38 ± 4.48 (m, 1H; 1''-H); 31P NMR: d ˆ 0.07 (s; phosphate); MS:
m/z ˆ 470 [M 2Na‡H] , 515.07 for C18H20N3Na2O10P
(S)-7: TLC: Rf ˆ 0.81; [a]D20
ˆ
9.4 (c ˆ 1, H2O); 1H NMR: d ˆ 2.81 (dd;
J(2a'',2b'') ˆ 14.2 Hz, J(2a'',1'') ˆ 7.0 Hz, 1H; 2a''-H), 2.89 (dd, J(2b'',1'') ˆ
5.3 Hz, 1H; 2b''-H), 3.60 ± 3.74 (m, 2H; 5a,b'-H), 3.88 ± 3.96 (m, 2H; 2'-,3'-
H), 4.00 (dd, J(4',3') ˆ J(4',5a') ˆ 4.5 Hz, 1H; 4'-H), 4.38 (ddd, J(1'',P) ˆ
8.4 Hz, 1H; 1''-H); 31P NMR: d ˆ 0.04 (s; phosphate); MS: m/z ˆ 470
[M 2Na‡H] , 515.07 for C18H20N3Na2O10P
Scheme 4. Formation of 6 with accomodation of the configurational
difference.
(Z)-9: HPLC (2% CH3CN, 10 mLmin 1): tR ˆ 14.4 min; 1H NMR
(600 MHz, D2O): d ˆ 1.89 (s, 3H; NHAc), 3.42 (dd, J(7'',8'') ˆ 9.1 Hz,
J(7'',6'') ˆ 1.9 Hz, 1H; 7''-H), 3.51 (dd, 2J(9a'',9b'') ˆ 11.9 Hz, J(9a'',8'') ˆ
7.0 Hz, 1H; 9a''-H), 3.75 (dd, J(9b'',8'') ˆ 2.8 Hz, 1H; 9''b-H), 3.84 (dd,
J(6'',5'') ˆ 9.8 Hz, 1H; 6''-H), 3.91 (m, 1H; 8''-H), 4.13 (m, 2H; 3',4'-H),
4.15 ± 4.20 (m, 3H; 2'-, 5a,b'-H), 4.73 (ddd, J(5'',4'') ˆ J(5'',3'') ˆ 1.9 Hz, 1H;
5''-H), 5.76 (ddd, J(4'',3'') ˆ 10.4 Hz, J(4'',P) ˆ 2 Hz, 1H; 4''-H), 5.85 (d,
J(1',2') ˆ 4.2 Hz, 1H; 1'-H), 6.50 (ddd, J(3'',P) ˆ 2.8 Hz, 1H; 3''-H); 31P
NMR (242.5 MHz, D2O): d ˆ 3.6 (brs; phosphate), 3.6 (brs; phospho-
nate); MS: m/z ˆ 644 [M 3Na‡2H] , 710.42 for C20H27N4 Na3O16P2
target compound (Z)-9 the orientation of the two decisive
moietiesÐthe phosphonate and the CMP moietyÐhave the
desired different geometry, and the pyran ring generated from
the neuraminyl residue will be flatter than in (R)-2 and thus
more similar to the benzene ring in 3h. The synthesis of (Z)-9
could be readily accomplished (Scheme 5): Reaction of
aldehyde 10[8] with diallyl phosphonate[19] gave a-hydroxy-
phosphonate 11 (as a mixture of diastereomers (R,S)-11),
which reacted by condensation with 5 as described above to
diastereomers (R,S)-12. Base-supported (DBU) elimination
of acetic acid followed by deallylation with [Pd(PPh3)4] in the
presence of dimedone and base-catalyzed deacylation afford-
ed target molecule (Z)-9 only (Table 2). This compound
exhibited very potent competitive inhibition of a(2-6)-sialyl-
transferase (Ki ˆ 40 nm, Table 1), and its affinity to a(2-6)-
sialyltransferase[20] is three orders of magnitude higher than
that of the natural substrate CMP-Neu5Ac; this supports the
conceptual approach discussed here.
[a] 1H NMR: 250 MHz, D2O, unless indicated otherwise. 31P NMR:
161.7 MHz, D2O, unless indicated otherwise. MS: MALDI, negative mode,
matrix ATT. HPLC: RP-18, 1 ± 3% CH3CN, 0.05m Et3NH ´ HCO3. TLC:
ethyl acetate/methanol/1m CH3CO2NH4 1/1/1.
CMP moiety has no major influence on the Ki values. Thus,
binding as depicted for the examples for (R)- and (S)-6 in
Scheme 4 is supported, which could accomodate the config-
urational difference.
The high affinity of 3h (which contains a flat benzene ring),
the minor influence of the carbon atom bearing the CMP
group, and the unexpected high affinity of elimination product
(E)-8 (Scheme 5; obtained during the synthesis of (R)-2 by
deacetoxyphosphonylation)[8] led us to combine the positive
effects of (R)-2, 3h and (E)-8 and to introduce a phosphonate
residue at the methylene group of 8. Thus, in the derived
Potent sialyltransferase inhibition can be based on flat
pyranosyl ring mimics which possess a methyl or a methylene
carbon atom bearing a phosphonate and a CMP residue, as
clearly exhibited for 1, (R)-2, 3h, (E)-8, and (Z)-9. Thus,
further details of the transition-state geometry in the enzy-
matic sialyl transfer can be deduced, which provides leads for
the extension of this research.
Angew. Chem. Int. Ed. 1998, 37, No. 20
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3720-2895 $ 17.50+.50/0
2895

COMMUNICATIONS
gel (water/ethanol 3/1) and concentrated in vacuo. Then an
aqueous ammonia solution (30%, 3 mL) was added in order to
remove the acetyl groups. The solution was concentrated under
reduced pressure and then dissolved in water (1 mL) and
‡
stirred with IR-120 (Na ). Filtration and subsequent addition
of ethanol (8 mL) afforded (Z)-9 (yield 24 mg, 80%).
Received: May 29, 1998 [Z11914IE]
German version: Angew. Chem. 1998, 110, 3021 ± 3024
Keywords: enzyme inhibitors ´ neuraminic acids ´
phosphonates ´ transition states
[1] a) A. Rosenberg, Biology of Sialic Acids, Plenum, New
York, 1995; b) F. DallꢁOlio, N. Malagolini, F. Serafini-
Cessi, Int. J. Cancer 1992, 50, 325 ± 330; c) J. Morgenthaler,
W. Kemmner, R. Brossmer, Biochem. Biophys. Res.
Commun. 1990, 171, 860 ± 866.
[2] a) R. J. Bernacki, Science 1977, 195, 557 ± 580; b) Q. Zhou,
S. Hakomori, K. Kitamura, Y. Igarashi, J. Biol. Chem.
1994, 269, 1959 ± 1965; c) J. W. Dennis in Cell Surface
Carbohydrates and Cell Development (Ed.: M. Fukuda),
CRC, Boca Raton, FL, USA, 1991, pp. 174 ± 175; d) W.
Kemmner, D. Kruck, P. M. Schlag, Clin. Exp. Metastasis
1994, 12, 245 ± 254.
[3] a) T. Kawai, A. Kato, H. Higashi, S. Kato, M. Naiki,
Cancer Res. 1991, 51, 1242 ± 1246; b) F.-G. Hanisch, B.
Witter, G. Crombach, W. Schänzer, G. Uhlenbruck,
Tumor-Associated Antigens, Oncogenes, Receptors, Cyto-
kines in Tumor Diagnosis and Therapy at the Beginning of
the Nineties. Cancer of the Breast-State and Trends in
Diagnosis and Therapy (Ed.: R. Klapder), Zuckersch-
werdt, München, 1992, pp. 367 ± 370.
[4] a) E. Sjoberg, A. Manzi, K. Khoo, A. Varki, J. Biol. Chem.
1992, 267, 16200 ± 16211; b) T. Kageshita, S. Hirai, T.
Kimura, Cancer Res. 1995, 55, 1748 ± 1751.
Scheme 5. Synthesis of (Z)-9. All ˆ allyl, DBU ˆ 1,8-diazabicyclo[5.4.0]undec-7-ene.
[5] T. Hennet, D. Chui, J. C. Paulson, J. D. Marth, Proc. Natl.
Acad. Sci. USA 1998, 95, 4504 ± 4509.
[6] A. Harduin-Lepers, A. Recchi, P. Delaunnoy, Glyco-
biology 1995, 5, 741 ± 758.
Experimental Section
[7] a) C. Schaub, B. Müller, R. R. Schmidt, Glycoconjugate J. 1998, 15,
345 ± 354; b) C. Schaub, R. R. Schmidt, Abstr. Pap. 2nd Eur. Conf.
Carbohydr. Mimics (La Garda, Italy), 1996, p. C10.
[8] F. Amann, C. Schaub, B. Müller, R. R. Schmidt, Chem. Eur. J. 1998, 4,
1106 ± 1115.
A) Protocol for the synthesis of a-hydroxyphosphonates: The aldehyde
(1 equiv) was dissolved in a small amount of CH2Cl2, and then the
phosphonic acid diester (2 equiv) and a few drops of NEt3 were added. The
solution was stirred for 18 h, concentrated, and purified by chromatography
on silica gel.
[9] B. Müller, T. J. Martin, C. Schaub, R. R. Schmidt, Tetrahedron Lett.
1998, 39, 509 ± 512.
B) Protocol for the condensation with 5: Compound 5 (1.5 equiv) and a-
[10] a) W. Korytnyk, N. Angelino, W. Klohs, R. J. Bernacki, Eur. J. Med.
Chem. 1980, 15, 77 ± 84; b) I. Kijima-Suda, Y. Migamata, S. Toyoshima,
M. Itoh, T. Osawa, Cancer Res. 1986, 46, 858 ± 862; c) S. H. Khan, K. L.
Matta in Glycoconjugates, Composition, Structure, Function (Eds.:
H. J. Allen, E. C. Kisailus), Dekker, New York, 1992, pp. 361 ± 378,
and references therein.
[11] a) H. Hashimoto, K. Sato, T. Wakabayashi, H. Kodama, Y. Kajihara,
Carbohydr. Res. 1993, 247, 179 ± 193; b) M. Imamoto, H. Hashimoto,
Tetrahedron Lett. 1996, 37, 1451 ± 1454.
[12] a) R. G. Kleineidam, T. Schmelter, R. T. Schwarz, R. Schauer,
Glycoconjugate J. 1997, 14, 57 ± 66; b) D. H. van den Eijnden, T. J.
Martin, R. R. Schmidt, unpublished results; c) G. Dufner, Disserta-
tion, Universität Konstanz, 1997.
[13] Recently antisense technology has been investigated in order to
decrease sialyltransferase expression: W. Kemmner, K. Hohaus, P. M.
Schlag, FEBS Lett. 1997, 409, 347 ± 350.
[14] Other examples of transition-state analogues of sugar nucleotides:
UDP-Gal: a) K. Frische, Dissertation, Universität Konstanz, 1992;
b) K. Frische, R. R. Schmidt, Bioorg. Med. Chem. Lett. 1993, 3, 1747 ±
1750; GDP-Fuc: c) S. Cai, M. R. Strond, S. Hakomori, T. Toyokuni, J.
Org. Chem. 1992, 57, 6693 ± 6696.
[15] B. A. Horenstein, M. Bruner, J. Am. Chem. Soc. 1996, 118, 10371 ±
10379; b) B. A. Horenstein, M. Bruner, Biochemistry 1998, 37, 289 ±
297, and references therein.
hydroxyphosphonate (1 equiv) were dissolved in CH2Cl2 and evaporated to
dryness. The remaining foam was dissolved under nitrogen in dry CH2Cl2,
and 1H-tetrazole (2 equiv) was added. After the reaction mixture had been
stirred for 3 h, an anhydrous solution of tert-butyl hydroperoxide
(1.5 equiv) was added, and after an additional hour NEt3 (50 equiv) was
added. After 18 h of stirring, the reaction mixture was concentrated at 208C
and purified by chromatography on silica gel (ethyl acetate/methanol 5/1,
1% NEt3).
O-Debenzylation and/or O/N-deacetylation was performed as previously
described[8] to afford 3h, 3l, 4h, 4l, (R)-6, (S)-6, (R)-7, and (S)-7.
C) Synthesis of (Z)-9: The synthesis of (R,S)-11 and (R,S)-12 followed
protocols A and B, respectively. Compound (R,S)-12 (180 mg, 0.16 mmol)
and 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU, 70 mL, 0.48 mmol) were
dissolved in dry THF (2 mL) and heated to 608C; after 2 h additional DBU
(50 mL, 0.33 mmol) was added, and stirring continued for 3 h. Then acetic
anhydride (2 mL) and pyridine (4 mL) were added at room temperature,
and after 15 h the reaction mixture was concentrated and purified by
chromatography on silica gel (ethyl acetate/methanol 5/1, 1% NEt3) to give
protected (Z)-9 (120 mg, 70%); Rf ˆ 0.10 (ethyl acetate/methanol 4/1, 1%
NEt3). For the removal of the protecting groups (Z)-9 (40 mg) was
dissolved in THF (2 mL) and treated with [Pd(PPh3)4] (4.2 mg, 4 mmol) and
dimedone (21 mg, 0.15 mmol), and the mixture stirred for 30 min in the
dark. The reaction mixture was purified by chromatography on RP-18 silica
2896
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3720-2896 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1998, 37, No. 20

COMMUNICATIONS
[16] M. Hoffmann, Synthesis 1988, 62 ± 64.
Crabtreeꢁs catalyst, a cationic iridium complex with a mono-
phosphane and pyridine as ligands.[5] Owing to the remarkable
properties of Crabtreeꢁs catalystÐthat is, the ability to
hydrogenate tri- and tetrasubstituted olefins that do not react
with rhodium± or ruthenium ± phosphane catalystsÐwe de-
cided to study complexes 1 as catalysts for the hydrogenation
of unfunctionalized tri- and tetrasubstituted olefins.
Initial experiments with complex 1a led to encouraging
results. Indeed, the hydrogenation of olefin 3 in CH2Cl2 with
4 mol% of 1a [Eq. (1)] gave a good ee value of 75%, albeit
with moderate conversion (78%, Table 1). Lower catalyst
loadings led to decreased conversion. Preliminary kinetic data
[17] a) Y. Kajihara, K. Koseki, T. Ebata, H. Kodama, H. Matsusita, H.
Hashimoto, Carbohydr. Res. 1994, 264, C1 ± C5; b) Y. Kajihara, K.
Koseki, T. Ebata, H. Kodama, H. Matsusita, H. Hashimoto, J. Org.
Chem. 1995, 60, 5732 ± 5735.
[18] For details of the measurement of the Ki values, see ref. [7a].
[19] Commercially available from Aldrich.
[20] Investigation of inhibition with a recombinant rat liver a(2-3)-
sialyltransferase led to similar results: C. Schaub, Dissertation, in
preparation.
Enantioselective Hydrogenation of Olefins with
Iridium ± Phosphanodihydrooxazole Catalysts
Andrew Lightfoot, Patrick Schnider, and
Andreas Pfaltz*
Enantioselective hydrogenation is one of the most powerful
methods in asymmetric catalysis, with the most versatile
catalysts being rhodium± and ruthenium ± diphosphane com-
plexes.[1] However, with a few exceptions, the range of
substrates is still limited to certain classes of olefins bearing
polar groups that can coordinate to the catalyst. Moreover,
even in standard substrate categories derivatives can be found
that give unsatisfactory enantioselectivities. There are only a
few examples of highly enantioselective hydrogenations of
unfunctionalized olefins.[2, 3] The most impressive results in
this field have been obtained by Buchwald, who used
Brintzingerꢁs chiral titanocene complexes.[3] Although high
ee values were obtained, the low catalyst activity required that
a relatively high catalyst loading (ꢁ5 mol%) was used.
Therefore, the search for new catalysts to fill these methodo-
logical gaps continues.
with (E)-1,2-diphenyl-1-propene as substrate shows a large
initial turnover frequency (TOF) of 41 min (olefin: 0.31m,
catalyst: 0.003m), but as the reaction proceeds the TOF
1
Table 1. Enantioselective hydrogenation of 3 with use of catalysts 1a ± f
[Eq. (1)].
Entry
Cat. (mol%)
Conversion [%]
ee [%]
1
2
3
4
5
6
1a (4)
1b (4)
1c (4)
1d (4)
1e (0.3)
1 f (0.3)
78
98
> 99
57
> 99
> 99
75
90
91
97
70
98
decreases to essentially zero within one hour.[6] Deactivation
has also been observed with Crabtreeꢁs catalyst, which is
explained by the formation of inactive hydride-bridged
trimers.[5] In our case the origin of deactivation is not known,
but traces of water appear to enhance this undesired side
reaction. Thus, the addition of molecular sieves to the above
system gave an increased conversion. Alternatively, running
the reaction under strictly anhydrous conditions, using freshly
dried CH2Cl2 (distilled over CaH2) and standard Schlenk
techniques, allows full conversion to be obtained with 4 mol%
of 1a with the same ee value as before. Analytically pure
product 4 was isolated in essentially quantitative yield by
removal of the solvent and distillation.
We recently reported enantioselective hydrogenation of
imines catalyzed by iridium ± phosphanodihydrooxazole com-
plexes of the type 1 with PF6 as anion.[4] The coordination
environment of the iridium center resembles that in
A systematic study of ligands 2a ± d led to a highly
enantioselective catalyst for the hydrogenation of 3 (Table 1).
Thus, replacement of the isopropyl group of ligand 2a with a
tert-butyl group (2b) increased the ee value to 90%. Alter-
natively, replacement of the diphenylphosphanyl group of 2a
with a bis(o-tolyl)phosphanyl group (2c) increased the ee
value to 91%. The combined use of a tert-butyl group and a
bis(o-tolyl)phosphanyl group (2d) led to the highest enantio-
selectivity (97% ee, 57% conversion with 4 mol% of 1d).
Attempts to further increase the conversion by varying the
reaction conditions and introducing additives, such as iodide,
were met with little success.[7] However, replacement of the
PF6 anion with BARF (tetrakis[3,5-bis(trifluoromethyl)phe-
[*] Prof. Dr. A. Pfaltz, Dr. A. Lightfoot, Dr. P. Schnider
Max-Planck-Institut für Kohlenforschung
Kaiser-Wilhelm-Platz 1, D-45470 Mülheim an der Ruhr (Germany)
Fax: (‡49)208-306-2992
E-mail: pfaltz@mpi-muelheim.mpg.de
Angew. Chem. Int. Ed. 1998, 37, No. 20
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1433-7851/98/3720-2897 $ 17.50+.50/0
2897