An efficient and general method for the synthesis of α,ω- difunctional reduced polypropionates by Zr-catalyzed asymmetric carboalumination: Synthesis of the scyphostatin side chain

Article abstract of DOI:10.1002/anie.200353429

A decrease in the number of linear synthetic steps and an increase in efficiency have been realized in the synthesis of the side chain of scyphostatin. An efficient and selective synthesis of α,ω- difunctional reduced polypropionates from methyl 3-hydroxy-2-methylpropionate through the use of a Zr-catalyzed asymmetric carboalumination has been developed (see scheme, TBS = tert-butyldimethylsilyl).

Full text of DOI:10.1002/anie.200353429

Angewandte
Chemie
Synthetic Methods
An Efficient and General Method for the
Synthesis of a,w-Difunctional Reduced
Polypropionates by Zr-Catalyzed Asymmetric
Carboalumination:Synthesis of the Scyphostatin
Side Chain**
Ze Tan and Ei-ichi Negishi*
We recently reported an efficient and general method for the
synthesis of reduced polypropionates with a single hetero-
function[1] through the application of Zr-catalyzed asymmet-
ric carboalumination.[2,3] In view of the large number of
complex natural products that are of medicinal and biological
interest, such as scyphostatin (1),[4] ionomycin,[5] doliculide,[6]
and borrelidin,[7] we thought it worthwhile to search for a
related method for the synthesis of terminally differentiated
a,w-difunctional reduced polypropionates through the use of
Zr-catalyzed asymmetric carboalumination.[1–3] We report
herein one such method involving 1) just one relatively
inexpensive and enantiomerically pure (ꢀ 99% ee) methyl
(R)- or (S)-3-hydroxy-2-methylpropionate and 2) a catalytic
and reagent-controlled asymmetric carbometalation[1–3] with
Me3Al and either enantiomer of [ZrCl2(nmi)2] (nmi = 1-
neomenthylindenyl; see 2).[8] We also report the application
of this method to the synthesis of the scyphostatin side chain.
Preparation of 3 (Scheme 1) started with protection of
methyl (S)-3-hydroxy-2-methylpropionate (ꢀ 99% ee; TCI,
[*] Dr. Z. Tan, Prof. E.-i. Negishi
Herbert C. Brown Laboratories of Chemistry
Purdue University
560 Oval Drive, West Lafayette, IN 47907-2084 (USA)
Fax: (+1)765-494-0239
E-mail: negishi@purdue.edu
[**] We thank the National Institutes of Heath (grant no. GM36792) and
Purdue University for their support of this research, and Boulder
Scientific Co. for assistance in the procurement of Zr compounds.
The collaborative efforts of B. Liang, T. Novak, and M. Magnin-
Lachaux are acknowledged.
Angew. Chem. Int. Ed. 2004, 43, 2911 –2914
DOI: 10.1002/anie.200353429
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2911

Communications
(À)-[ZrCl2(nmi)2] led to the formation of a 1:8
mixture of the same two isomers, with a
combined yield of 72%. This d.r. value of 1:8
corresponds to 89% stereoselectivity at C2. The
observed modest preference for the formation
of syn-2,4-dimethyl-1-alkanols over that of anti-
2,4-dimethyl-1-alkanols appears to be a general
trend in Zr-catalyzed asymmetric carboalumi-
nation, although the corresponding reaction of
the TBDPS-protected 3b did not show any clear
sign of internal asymmetric induction (Table 1).
To probe the extent of internal asymmetric
induction, we treated 3a and 3b with Me3Al in
the presence of [ZrCl2(ind)2] (ind = indenyl).
Yields of 73% and 78% were obtained from 3a
and 3b after oxidation, and the products exhib-
ited 2S,4R/2R,4R ratios of 1.12:1 and 1.1:1,
Scheme 1. Synthesis of TBS- or TBDPS-protected (R)-2-methyl-4-penten-1-ol (3) from methyl
(S)-3-hydroxy-2-methylpropionate. TBS=tert-butyldimethylsilyl, TBDPS=tert-butyldiphenylsilyl,
DMF=dimethylformamide, DIBAH=diisobutylaluminium hydride, THF=tetrahydrofuran,
dpephos=bis[2-(diphenylphosphanyl)phenyl] ether.
Tokyo, Kasei Kogyo Co., Ltd) by treatment with either
TBSCl or TBDPSCl in the presence of imidazole and DMF
(ca. 95% yield). Reduction by treatment with DIBAH,
followed by iodination led to the TBS- or TBDPS-protected
(S)-3-iodo-2-methyl-1-propanol. This compound was con-
verted into the corresponding zinc reagent in situ by treat-
ment with tBuLi (2.1 equiv) in diethyl ether at À788C[9a] and
then with dry ZnBr2 (0.65 molarequiv) at À78–08C. The
respectively, which suggests that there is indeed a minor
preference for the formation of the syn isomer. Chromato-
graphic separation of the minor isomer (2R,4R)-4 from
(2S,4R)-4 was readily achieved in one operation on silica gel
(EtOAc/hexanes, 1:50). (2S,4R)-4a (Z = TBS; d.r. ꢀ 40:1) and
(2S,4R)-4b (Z = TBDPS; d.r. ꢀ 37:1) were obtained in yields
of 55 and 61%, respectively, based on the starting alkene 3.
These yields correspond to 73% and 77% recovery in the
chromatographic separation. Since d.r. values of 13:1 and 10:1
indicate maximum possible recovery rates of 93% and 91%,
respectively, there seems to be room for further improvement
in the recovery in view of the ease of diastereomeric
separation in these cases.
The high enantiomeric purity of 3 (ꢀ 98% ee) guarantees
an overall ee value at least this high, regardless of the extent
of stereoselection in the conversion of 3 into 4. Statistical
enantiomeric amplification, which is nothing more than a
manifestation of the kinetic mass action law, would further
elevate the overall ee value after each additional asymmetric
stereogenic step. For example, the overall ee values estab-
lished on the basis of the mass action law for asymmetric
reaction of compounds of 98.0% ee in the absence of internal
asymmetric induction may reliably be predicted as shown in
Table 2. The overall ee value of the products 4a and 4b may
be estimated from Table 2 to be 99.6% or more. HPLC
analysis of the urethanes derived from 4a and (+)- and (À)-1-
=
organozinc intermediate was vinylated with BrCH CH2
(3 equiv) in THF/diethyl ether[9b–d] in the presence of
[PdCl2(dpephos)][10] (2–5 mol%) at 238C to give either 3a
(Z = TBS) or 3b (Z = TBDPS) in a yield of 84 or 88%,
respectively. The overall yields of 3a and 3b from methyl (S)-
3-hydroxy-2-methylpropionate over four steps were 57% and
67%, respectively.
Treatment of 3a with Me3Al (3 equiv), methylaluminox-
ane (MAO; 1.2 equiv), and (+)-[ZrCl2(nmi)2] (5 mol%) in
CH2Cl2 at 08C for 20 h produced, after oxidation with O2, a
13:1 mixture of (2S,4R)- and (2R,4R)-4 in 75% combined
yield (Table 1). The observed diastereomeric ratio (d.r.) of
13:1 suggests that the enantioselectivity at C2 would be 93%,
were it not for the pre-existing asymmetric carbon center at
C4. This value, which is somewhat higher than the 85–90%
enantioselectivity range observed with achiral alkenes, sug-
gests a modest level of favorable internal asymmetric
induction. The corresponding reaction in the presence of
Table 1: Conversion of 3 into 5-siloxy-2,4-dimethyl-1-pentanols (4).
3
Z
[ZrCl2(nmi)2]
Combined
yield [%][a]
2S,4R/2R,4R
Yield of 4
after chromatography [%]
2S,4R/2R,4R
before chromatography[b]
after chromatography[b]
3a
3a
3b
3b
TBS
TBS
TBDPS
TBDPS
(+)
(À)
(+)
(À)
75
72
82
81
13/1
1/8
10/1
1/10
55 (31 over 5 steps)
ꢀ40/1
[c]
[c]
61 (41 over 5 steps)
ꢀ37/1
[c]
[c]
[a] Combined yield of 4 before chromatography. [b] Determined by 13C NMR spectroscopy. [c] Not purified by chromatography.
2912
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 2911 –2914

Angewandte
Chemie
Table 2: Estimation of overall ee values and maximum attainable yields
for asymmetric reactions of chiral compounds of 98.0% ee.
(2S,4R,6R)-5a into (2S,4R,6R,8R)-6a (d.r. > 65:1) in 40%
combined yield over three steps. The diastereomeric ratios
observed for the products of the Zr-catalyzed asymmetric
carboalumination itself (step C) were 10.5:1 and 12:1
(Scheme 2).
Asymmetric
induction
de [%] (d.r.)
Max. possible
yield [%]
Overall
ee [%][b]
R/S or S/R[a]
Scyphostatin (1), a potent inhibitor of neutral sphingo-
myelinase,[4a] was first reported in 1997 but its total synthesis
does not appear to have been reported.[11] The structure of
this compound consists of a C20 carboxylic acid side chain and
a bicyclic amine core fragment. The side chain has been
synthesized once before in the form of ethyl ester 7 in nine
steps (six isolation steps) from (2S,4R)-4b and in 15%
yield.[4b] However, the preparation of (2S,4R)-4b from diethyl
2-methylmalonate and ethyl a-bromoisobutyrate by enzyme-
catalyzed monoacetylation of meso-2,4-dimethyl-1,5-pen-
tanediol required seven additional steps and was achieved
in a mere 4% yield. Thus, the linear 16-step synthesis of 7
proceeded in 0.6% total yield. Use of the 5-step synthesis of
(2S,4R)-4b (d.r. ꢀ 37:1) in 41% total yield (Table 1) would
alone amount to a tenfold increase in the yield of 7.
50/50
80/20
85/15
90/10
95/05
100/0
00.0 (1:1)
58.8 (3.9:1)
68.6 (5.4:1)
78.4 (8.3:1)
88.2 (9.4:1)
100.0 (¥)
50.0
79.4
84.3
89.2
94.1
100.0
98.0
99.5
99.6
99.8
99.9
100.0
[a] The R/S or S/R ratio at the stereogenic center. [b] Statistically
estimated.
(a-naphthyl)ethyl isocyanates provides experimental evi-
dence of ee values of more than 99%. In these cases, however,
the maximum possible product yields are in practice much
more important. If yields of at least 80% are assumed to be
desirable for practical purposes, asymmetric reactions with
stereoselectivities at the second stereogenic center of 80% or
higher would suffice to attain this practical goal. Such a
stereoselectivity corresponds to over 4:1 d.r. and the gener-
ation of chiral compounds of over 99% ee. In short, we have
developed an efficient, selective, and practical five-step
protocol for the synthesis of 2,4-dimethyl-1-alkanols that is
catalytic in chiral auxiliaries (Table 1). Synthesis of (2S,4R)-
4a and (2R,4R)-4b from methyl (S)-3-hydroxy-2-methylpro-
pionate was achieved with total yields over five steps of 31%
and 41%, respectively, 97–98% diastereomeric purity, and at
least 99.6% ee (estimated).
We sought a convergent synthetic scheme to make the
synthesis of 7 even more efficient. We developed a synthesis
whose longest linear sequence is 11 steps (Scheme 3). In
Conversion of 2,4-dimethyl-1-alkanols into 2,4,6-tri-
methyl-1-alkanols and higher reduced polypropionates can
be achieved by application of an iterative three-step proto-
col[1] consisting of 1) iodination (A), 2) Pd-catalyzed vinyl-
ation (B), and 3) Zr-catalyzed methylalumination (C,
Scheme 2). (2S,4R)-4a was converted into (2S,4R,6R)-5a
(d.r. > 30:1) in 36% combined yield over three steps, and
iteration of the same three-step protocol converted
Scheme 3. Convergent synthesis of the scyphostatin side chain (7) by
Zr-catalyzed asymmetric carboalumination. 1) TBDPSCl (1.2 equiv),
imidazole, DMF; 2) a) 5 mol% (À)-[ZrCl2(nmi)2], MAO (1.2 equiv),
Et3Al (3 equiv), 08C; b) aq HCl; 3) TBAF (1.5 equiv), THF; 4) Swern
oxidation; 5) PPh3 (2 equiv), CBr4 (2 equiv), Zn dust (2 equiv), CH2Cl2,
08C; 6) a) nHexLi (2.05 equiv), À78–08C; b) MeI (3 equiv);
7) a) [HZrCp2Cl] (1.5 equiv), THF, 608C; b) I2 (1.5 equiv), 08C;
8) 5 mol% [PdCl2(dpephos)], vinyl bromide (3 equiv), 238C; 9) TBAF
(1.5 equiv); 10) cat. TPAP, NMO, CH2Cl2; 11) LDA, 08C. TBAF=tetra-
butylammonium fluoride, TPAP=tetrapropylammonium perruthenate,
NMO=4-methylmorpholine N-oxide, LDA=lithium diisopropylamide.
Scheme 2. Synthesis of trimethyl- and tetramethyl-1-alkanols from
(2S,4R)-4a through application of a three-step protocol. A, B, and C
are defined in Scheme 1 and Table 1.
Angew. Chem. Int. Ed. 2004, 43, 2911 –2914
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2913

Communications
addition to (2S,4R)-4a, (2E,4R)-2-iodo-4-methyl-2-hexene
of the scyphostatin side chain in the form of its ethyl ester 7
was achieved in 11 linear steps (nine isolation steps) with
19% overall yield. These results should prove to be useful for
exploitation of 7 in the total synthesis of scyphostatin.
(8) and the previously employed 6-P-substituted (2E,4E)-
hexa-2,4-dienoic acid ethyl ester[4b] (10) were used as key
intermediates. The preparation of 8 started with protection of
allyl alcohol with TBDPSCl. Asymmetric ethylalumination of
the TBDPS-protected allyl alcohol with Et3Al (3 equiv),
MAO (1 equiv), and 5 mol% (À)-[ZrCl2(nmi)2], followed by
protonolysis and desilylation by treatment with TBAF gave
(R)-2-methyl-1-butanol in 92% ee and 65% combined yield
over three steps. No enantiomeric separation was carried out.
The crude (R)-2-methyl-1-butanol was 1) oxidized under
Swern conditions,[12] 2) subjected to the Corey–Fuchs reac-
tion[13] to give (R)-1,1-dibromo-3-methyl-1-pentene, and
3) converted into (R)-4-methyl-2-hexyne. Hydrozirconation/
iodinolysis of this compound provided (2E,4R)-2-iodo-4-
methyl-2-hexene (8) in more than 98% isomeric purity and
71% yield. The total yield of 8 from allyl alcohol over seven
steps (six isolation steps) was 30%.
Received: November 28, 2003
Revised: February 20, 2004 [Z53429]
Keywords: carboalumination · enantioselectivity ·
.
homogeneous catalysis · synthesis design · zirconium
[1] E. Negishi, Z. Tan, B. Liang, T. Novak, Proc. Natl. Acad. Sci.
USA, 2004, 101, 5782.
[2] a) D. Y. Kondakov, E. Negishi, J. Am. Chem. Soc. 1995, 117,
10771; b) D. Y. Kondakov, E. Negishi, J. Am. Chem. Soc. 1996,
118, 1577.
[3] a) S. Huo, E. Negishi, Org. Lett. 2001, 3, 3253; b) S. Huo, J. Shi,
E. Negishi, Angew. Chem. 2002, 114, 2245; Angew. Chem. Int.
Ed. 2002, 41, 2141.
Iodination of (2S,4R)-4a produced the desired iodide in
89% yield. This iodide was treated first with tBuLi (2.1 equiv)
in diethyl ether at À788C and then with dry ZnBr2
(0.65 molarequiv). The alkyl zinc derivative thus generated
was cross-coupled with 8 in the presence of [PdCl2(dpephos)]
(5 mol%).[10] The coupling proceeded in 94% yield based on
the amount of iodide used. After desilylation with TBAF,
column chromatography on silica gel (ethyl acetate/hexanes,
1:25) provided the desired compound 9 in more than 98%
isomeric purity and 87% yield. Compound 9 was thus
synthesized in 25% overall yield and the longest linear
synthetic sequence consisted of nine steps starting with allyl
alcohol. An additional six steps were needed for the synthesis
of the TBS-protected 5-iodo-2,4-dimethyl-1-pentanol in 37%
overall yield. Since the previously reported synthesis of 7
employed 9 as a key intermediate, a formal synthesis of 7 was
complete at this point. For a variety of reasons, we never-
theless decided to complete the synthesis of 7 by following the
reported procedure involving 1) oxidation of 9 with TPAP
and NMO and 2) olefination of the resultant aldehyde with
10, which was prepared from methyl (E)-4-bromocrotonate in
55% yield over four steps.[4b] These two processes were
completed in 78% combined yield. Our convergent and
efficient synthesis of 7 therefore occurred in 19% yield. The
longest linear sequence of the synthesis involved 11 steps. The
1H and 13C NMR spectra of the product were not only
consistent with the assigned structure but also agreed well
with those reported by Hoye and Tennakoon.[4b]
[4] a) For the isolation procedure, see: F. Nara, M. Tanaka, T.
Hosoya, K. Suzuki-Konagai, T. Ogita, J. Antibiot. 1999, 52, 525;
b) for a synthesis of the side chain, see: T. R. Hoye, M. A.
Tennakoon, Org. Lett. 2000, 2, 1481.
[5] a) S. Hanessian, N. G. Cooke, B. Dehoff, Y. Sakito, J. Am. Chem.
Soc. 1990, 112, 5276; b) D. A. Evans, R. L. Dow, T. L. Shih, J. M.
Takacs, R. Zahler, J. Am. Chem. Soc. 1990, 112, 5290; c) M.
Lautens, J. T. Colucci, S. Hiebert, N. D. Smith, G. Bouchain, Org.
Lett. 2002, 4, 1879.
[6] A. K. Ghosh, C. Liu, Org. Lett. 2001, 3, 635.
[7] a) M. O. Duffey, A. LeTiran, J. P. Morken, J. Am. Chem. Soc.
2003, 125, 1458; b) B. G. Vong, S. Abraham, A. X. Xiang, E. A.
Theodorakis, Org. Lett. 2003, 5, 1617; c) S. Hanessian, Y. Yang, S.
Giroux, V. Mascitti, J. Ma, F. Raeppel, J. Am. Chem. Soc. 2003,
125, 13784.
[8] G. Erker, M. Aulback, M. Knickmeier, D. Wingbermꢀhle, C.
Krꢀger, M. Nolte, S. Werner, J. Am. Chem. Soc. 1993, 115, 4590.
[9] a) E. Negishi, D. R. Swanson, C. J. Rousset, J. Org. Chem. 1990,
55, 5406; b) E. Negishi, L. F. Valente, M. Kobayashi, J. Am.
Chem. Soc. 1980, 102, 3298; c) S. Gagneur, J. L. Montchamp, E.
Negishi, Organometallics 2000, 19, 2417; d) E. Negishi, S.
Gagneur in Handbook of Organopalladium Chemistry for
Organic Synthesis (Ed.: E. Negishi), Wiley-Interscience, New
York, 2002, chapter III.2.11.1, pp. 597–618.
[10] M. Kranenburg, Y. E. M. van der Burgt, P. C. J. Kamer,
P. W. N. W. van Leeuwen, K. Goubitz, J. Fraanje, Organometal-
lics 1995, 14, 3081.
[11] Note added in proof: After submission of this paper, we were
informed of an oral report of a total synthesis of scyphostatin.
We thank Prof. Katoh for providing us with this information (M.
Inoue, W. Yokota, M. G. Murugesh, T. Izuhara, T. Katoh, 45th
Symposium on the Chemistry of National Products, October 6–8,
2003, Kyoto, Japan).
In conclusion, 1) an efficient and potentially general
method for the synthesis of a,w-difunctional reduced poly-
propionates through the use of Zr-catalyzed asymmetric
carboalumination has been developed, 2) catalysis with
[12] K. Omura, D. Swern, Tetrahedron 1978, 34, 1651.
[13] E. J. Corey, P. L. Fuchs, Tetrahedron Lett. 1972, 3769.
À
respect to a chiral auxiliary for asymmetric C C bond
formation, coupled with high overall efficiency makes the
method reported herein unique among the known methods
for the synthesis of a,w-difunctional reduced polypropionates,
3) the use of commercially available and relatively inexpen-
sive methyl (R)- and (S)-3-hydroxy-2-methylpropionates with
ee values of at least 98% makes enantiomeric separation
unnecessary for the synthesis of reduced polypropionates of
more than 98% ee, and 4) a convergent and efficient synthesis
2914
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 2911 –2914