Stereoselective synthesis of the halaven C14-C26 fragment from D-quinic acid: Crystallization-induced diastereoselective transformation of an α-methyl nitrile

Article abstract of DOI:10.1002/anie.201501143

Crystallization-induced diastereoselective transformation (CIDT) of an α-methyl nitrile completes an entirely non-chromatographic synthesis of the halichondrin B C14-C26 stereochemical array. The requisite α-methyl nitrile substrate is derived from D-quinic acid through a series of substrate-controlled stereoselective reactions via a number of crystalline intermediates that benefit from a rigid polycyclic template. Therefore, all four stereogenic centers in the Halaven C14-C26 fragment were derived from the single chiral source D-quinic acid.

Full text of DOI:10.1002/anie.201501143

Angewandte
Chemie
International Edition: DOI: 10.1002/anie.201501143
German Edition: DOI: 10.1002/ange.201501143
Asymmetric Synthesis
Stereoselective Synthesis of the Halaven C14–C26 Fragment from
d-Quinic Acid: Crystallization-Induced Diastereoselective
Transformation of an a-Methyl Nitrile**
Francis Belanger, Charles E. Chase, Atsushi Endo, Francis G. Fang,* Jing Li, Steven R. Mathieu,
Annie Z. Wilcoxen, and Huiming Zhang
Abstract: Crystallization-induced diastereoselective transfor-
mation (CIDT) of an a-methyl nitrile completes an entirely
non-chromatographic synthesis of the halichondrin B C14–
C26 stereochemical array. The requisite a-methyl nitrile
substrate is derived from d-quinic acid through a series of
substrate-controlled stereoselective reactions via a number of
crystalline intermediates that benefit from a rigid polycyclic
template. Therefore, all four stereogenic centers in the Halaven
C14–C26 fragment were derived from the single chiral source
d-quinic acid.
T
he important role of total synthesis in the discovery and
development of novel medicines was highlighted by the
recent approval of Halaven (1),[1] a fully synthetic analogue of
the marine natural product halichondrin B (2),[2,3] for treat-
ment of certain patients with advanced breast cancer
(Figure 1). A characteristic feature of the manufacturing
process for two of the three fragments (C1–C13 and C27–
C35) was the employment of stereocontrol on cyclic carbo-
hydrate templates to enable the discovery of crystallization-
based routes.[4] The third fragment 3 (C14–C26), although not
the most structurally complex, presents a distinct challenge
within the overall synthesis owing to the presence of multiple
non-contiguous acyclic stereogenic centers.[5] Reconfiguration
of 3 as cyclic iodoether 4 could provide a conceptually
Figure 1. Synthesis strategy towards the Halaven C14–C26 fragment.
Ms=methanesulfonyl, NIS=N-iodosuccinimide, Pv=pivaloyl, Tf=tri-
fluoromethanesulfonyl.
[*] Dr. C. E. Chase, Dr. A. Endo, Dr. F. G. Fang, Dr. J. Li, S. R. Mathieu
Next Generation Systems
different approach to 3 wherein the acyclic stereogenic
centers are established on a bicyclic template.[6] It was
hoped that a rigid template would impart a high degree of
innate crystallinity to the intermediates, thereby allowing
crystallization, rather than chromatography, to be the key
quality control technology. Proof of concept on the viability of
a bicyclic iodoether template as a synthetic precursor was
quickly demonstrated by stereoselective iodoetherification of
3 to provide 4 and zinc-mediated reductive elimination of 4 to
regenerate 3. Whereas the cis-fused bicyclic template in 4
clearly suggests methods to relate the configurations at the
C17, C20, and C23 positions, establishment of the fourth
ancillary stereogenic center (C25) would require an acyclic
stereocontrol strategy. Furthermore, a convenient chiral
origin for the bicyclic system would need to be identified.
Herein, we report the successful realization of these goals
and: 1) the use of d-quinic acid as a readily available chiral
source for all of the stereogenic centers in 3, 2) the first
example of a crystallization-induced diastereoselective trans-
Eisai Inc. 4 Corporate Drive, Andover, MA 01810 (USA)
E-mail: Frank_Fang@eisai.com
F. Belanger
Current address: PCAS BioMatrix, 725 Trotter St., Saint-Jean-sur-
Richelieu, Quebec J3B 8 J8 (Canada)
Dr. A. Z. Wilcoxen
Current address: Amgen
One Kendall Square, 360 Binney St. Cambridge, MA 02142 (USA)
H. Zhang
Current address: Johnson Matthey Pharma Services
75 Patton Rd. Devens, MA 01434 (USA)
[**] We thank our analytical chemistry colleagues Jian Del Vecchio,
Thomas Noland, Kristen Lacasse, and Noritsugu Maki for HPLC
method development, process research colleagues Robert Giusto
and William Powazinik for calorimetry, and the chemical develop-
ment department for pilot plant production. We thank Dr. Richard
Staples (Crystallographic Resources, Inc.) for X-ray structural
analysis.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201501143.
Angew. Chem. Int. Ed. 2015, 54, 1 – 5
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
formation[7] of an a-methyl nitrile, and 3) crystallization-
based stereochemical quality control benefiting from a rigid
bridged polycyclic system.
d-Quinic acid (5) was transformed into the crystalline
diacetate 6 by lactonization,[8] partial reduction, and perace-
tylation (Scheme 1). Lewis acid mediated C glycosidation of 6
with commercially available 3-trimethylsilyl-methyl-4-pente-
Scheme 2. Crystallization-induced diastereoselective transformation.
diastereomer) with a catalytic amount of potassium tert-
pentoxide equilibrated and precipitated 11 with 12:1 d.r.
(Scheme 2).
With all four stereogenic centers, C17, C20, C23, and C25,
firmly established, unraveling the quinic acid cyclohexane
ring (highlighted in red) was accomplished as depicted in
Scheme 3. Cyclohexylidene cleavage (step l, 11!12), regio-
Scheme 1. Synthesis of a-methyl nitrile 11 from d-quinic acid (5).
Reagents and conditions: a) cyclohexanone, p-TsOH·H2O, toluene,
reflux, 95%; b) TMSCl, imidazole, toluene, 208C; c) DIBAL, toluene,
ꢀ788C; d) 1) AcOH, H2O, DMAP, THF, 208C, 2) Ac2O, NEt3, 208C,
72% (3 steps); e) 7, BF3·OEt2, Tf2O, CH3CN, 208C; f) NaOMe, THF/
MeOH, 108C; g) LiAlH4, THF, 158C; h) MsCl, NEt3, THF, 258C;
i) KCN, EtOH/H2O, 708C, 30% from 6; j) 1) KHMDS (1.1 equiv), MeI,
toluene/THF, ꢀ758C, 2) KHMDS (0.15 equiv), ꢀ758C; k) KHMDS
(0.60 equiv), toluene, ꢀ708C (recycling of filtrates), 72% from 10.
Steps a, d, f, h, i, and j/k gave crystalline products. DIBAL=diisobutyl-
aluminum hydride, DMAP=4-dimethylaminopyridine, KHMDS=po-
tassium hexamethyldisilazane, p-TsOH=para-toluenesulfonic acid,
TMS=trimethylsilyl.
noate (7),[3a,9] followed by sodium methoxide treatment,
stereoselectively established the C20 (single isomer) and
C23 (9:1 d.r.) stereogenic centers of polycyclic pyran 8. Ester
8 was homologated to nitrile 10 by reduction (step g, 8!9),
activation (step h), and cyanide displacement (step i). Crys-
tallization of 10 enhanced the C23 stereochemical quality to
> 100:1. Therefore, a crystalline intermediate has been
rapidly assembled with essentially complete control of three
of the four stereogenic centers (C17, C20, and C23) of 3.
LDA treatment of 10 followed by quenching with methyl
iodide afforded a 1:1 mixture of a-methyl nitrile 11 (m.p.
1238C) and the C25 epimer 11’ (m.p. 828C). Fortuitously, the
desired isomer 11 exhibited greater crystallinity. Mindful of
the results of Fleming et al.,[10] KHMDS was investigated as
an alternative base in the hope that a larger counterion would
enhance chelation to the pyran oxygen atom. Encouragingly,
the selectivity improved to 2:1 (step j, part 1). Moreover,
in situ treatment with a catalytic amount of the base enhanced
the selectivity to > 4:1. Crystallization of the crude mixture
provided 11 with a diastereomeric ratio of 20:1–30:1 at the
C25 carbon atom (step j, part 2). The filtrates (1:2 d.r., 11’ as
the major diastereomer) were re-equilibrated and crystallized
to provide additional 11 (step k).
ꢀ
Scheme 3. Reagents and conditions: l) PhNH NH2·HCl, TsOH·H2O,
=
MeOH, 558C, 99%; m) BrC( O)CMe2(OAc), CH3CN, cat. H2O, 08C;
n) DBU, toluene, 1008C, 47% (2 steps); o) 1) O3, isopropyl acetate/
MeOH, ꢀ308C, 2) NaBH4, ꢀ208C, 3) K2CO3, RT, 4) NaIO4, THF/H2O,
RT, 90%; p) (MeO)2POCH2CO2Me, (iPr)2NEt, LiCl, CH3CN, 188C,
98%; q) H2, Pd/C, (iPr)2NEt, EtOAc, RT, 89%; r) 1) Tf2O, 2,6-lutidine,
MTBE, 08C, 2) NaI, DMF, RT, 81%; s) LiBH4, MeOH, THF/toluene,
138C; t) Zn, AcOH, THF/H2O, 08C, 99% (2 steps); u) 1) HCl, 2-
PrOH/toluene, 258C, 2) H2O, 458C, 94%; v) TBDPSCl, imidazole,
DMF, 238C; w) HNMe(OMe)·HCl, Me3Al, CH2Cl2/toluene, 08C;
x) TBSCl, imidazole, DMF, 238C; y) MeMgCl, THF, 08C; z) Tf2NPh,
KHMDS, THF/toluene, ꢀ208C; aa) HCl, 2-PrOH/MeOH, RT, 49%
from 18. Steps l, m, n, o, and r gave crystalline products. DBU=1,8-
diazabicyclo[5.4.0]undec-7-ene, TBDPS=tert-butyldiphenylsilyl,
TBS=tert-butyldimethylsilyl.
selective bromoacetate formation[11] (step m, 12!13), and
dehydrobromination (step n) afforded crystalline allylic ace-
tate 14. Ozonolysis of 14 (step o, part 1), borohydride quench
(step o, part 2), deacetylation (step o, part 3), and periodate
cleavage (step o, part 4) provided crystalline lactol 15 from
a single-pot sequence.
The results suggested the possibility of a crystallization-
induced diastereoselective transformation as a method to
control the C25 stereogenic center. In the event, treatment of
the epimeric mixture (11/11’, d.r. = 1:1.6, 11’ as the major
2
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Horner–Wadsworth–Emmons reaction of 15 under
Roush–Masamune conditions[12] afforded the a,b-unsaturated
ester 16 (step p). Catalytic hydrogenation (step q) followed by
triflate-mediated iodination of the neopentyl alcohol (step r,
presumably assisted by the neighboring pyran oxygen atom)
afforded iodoester 17 as a crystalline solid, wherein all
stereogenic centers (C17, C20, C23, and C25) were controlled
to d.r. > 1000:1. Reduction of iodoester 17 with LiBH4
(step s), followed by a zinc-mediated Vasella fragmentation
(step t),[13] proceeded smoothly under mild conditions to
furnish the C19 exocyclic olefin while liberating the C23
hydroxy group. Anhydrous HCl treatment to form a cyclic
imidate hydrochloride followed by hydrolytic lactonization
provided 18 (step u). Lactone 18 was converted into known
Weinreb amide 20[4d] by a sequence consisting of protection of
the C14 free alcohol (step v), lactone opening to give the
Weinreb amide (step w), and protection of the C23 hydroxy
functional group (step x). The previously described three-step
process (steps y–aa)[4d] provides the C14–C26 fragment 3a
with a quality suitable for downstream processing.
translating a natural product synthesis to a practical manu-
facturing process is the importance of crystallization in the
design strategy. A classical cyclic stereocontrol approach,
although sometimes lengthier, often benefits from a great
number of crystalline intermediates.
Keywords: antitumor agents · asymmetric synthesis ·
chiral pool · green chemistry · total synthesis
[1] For the discovery and development of 1, see: a) B. A. Littlefield,
M. H. Palme, B. M. Seletsky, M. J. Towle, M. J. Yu, W. Zheng,
World Patent, 9965894, 1999; b) M. J. Towle et al., Cancer Res.
2001, 61, 1013; c) W. Zheng et al., J. Bioorg. Med. Chem. Lett.
2004, 14, 5551; d) M. J. Yu, Y. Kishi, B. A. Littlefield in
Anticancer Agents from Natural Products, 1st ed. (Eds.: G. M.
Cragg, D. G. I. Kingston, D. J. Newman), Taylor & Francis
Group CRC Press, Boca Raton, FL, 2005, pp. 241 – 265; e) L. T.
Vahdat, B. Pruitt, C. J. Fabian, R. R. Rivera, D. A. Smith, E. Tan-
Chiu, J. Wright, A. R. Tan, N. A. DaCosta, E. Chuang, J. Smith,
J. OꢀShaughnessy, D. E. Shuster, N. L. Meneses, K. Chandra-
wansa, F. Fang, P. E. Cole, S. Ashworth, J. L. Blum, J. Clin.
Oncol. 2009, 27, 2954; f) M. J. Yu, Y. Kishi, B. A. Littlefield in
Anticancer Agents from Natural Products, 2nd ed. (Eds.: G. M.
Cragg, D. G. I. Kingston, D. J. Newman), Taylor & Francis
Group CRC Press, Boca Raton, FL, 2012, pp. 317 – 345; g) P. A.
Kaufman, A. Awada, C. Twelves, L. Yelle, E. A. Perez, G.
Velikova, M. S. Olivo, Y. He, C. E. Dutcus, J. Cortes, J. Clin.
Oncol. 2015, 33, 594.
A comparison of the operational efficiency of the
previous, more convergent manufacturing process (20 total
steps, 1% overall yield for the longest linear sequence of
13 steps)[4d] versus the new route towards the C14–C26
fragment (27 steps, 2% overall yield) is provided in
Figure 2, specifically highlighting raw-material cost and
waste minimization (aspects of “green chemistry”). A 78%
[2] For the discovery of the halichondrins, see: a) D. Uemura, K.
Takahashi, T. Yamamoto, C. Katayama, J. Tanaka, Y, Okumura,
Y. Hirata, J. Am. Chem. Soc. 1985, 107, 4796; b) Y. Hirata, D.
Uemura, Pure Appl. Chem. 1986, 58, 701.
[3] For the total synthesis of 2, see: a) T. D. Aicher, K. R. Buszek,
F. G. Fang, C. J. Forsyth, S. H. Jung, Y. Kishi, M. C. Matelich,
P. M. Scola, D. M. Spero, S. K. Yoon, J. Am. Chem. Soc. 1992,
114, 3162; for the total synthesis of norhalichondrin B, see:
b) K. L. Jackson, J. A. Henderson, H. Motoyoshi, A. J. Phillips,
Angew. Chem. Int. Ed. 2009, 48, 2346; Angew. Chem. 2009, 121,
2382; for the total synthesis of halichondrin C, see: c) A.
Yamamoto, A. Ueda, P. Brꢁmond, P. S. Tiseni, Y. Kishi, J. Am.
Chem. Soc. 2012, 134, 893; for a review on the synthetic work
towards the halichondrins, see: d) K. L. Jackson, J. A. Hender-
son, A. J. Phillips, Chem. Rev. 2009, 109, 3044, and references
therein.
Figure 2. Impact of the crystallization-based synthesis of the C14–C26
fragment on raw material cost and waste reduction.
[4] a) B. Austad, C. E. Chase, F. G. Fang, World Patent, 118565,
2005; b) H. Chiba, K. Tagami, J. Synth. Org. Chem. Jpn. 2011, 69,
600; for the synthesis of the C1 – C13 fragment, see: c) C. E.
Chase, F. G. Fang, B. M. Lewis, M. J. Schnaderbeck, G. D.
Wilkie, X. Zhu, Synlett 2013, 323; for the synthesis of the C27 –
C35/C14 – C35 fragments, see: d) B. C. Austad et al., Synlett
2013, 327; for the assembly of 1, see: e) B. C. Austad et al.,
Synlett 2013, 333.
[5] For new syntheses of the Halaven C14–C35 fragment and the
halichondrin C14–C38 fragment, see: a) D.-S. Kim, C. G. Dong,
J. T. Kim, H. Guo, J. Huang, P. S. Tiseni, Y. Kishi, J. Am. Chem.
Soc. 2009, 131, 15636; b) C. G. Dong, J. A. Henderson, Y.
Kaburagi, T. Sasaki, D. S. Kim, J. T. Kim, D. Urabe, H. Guo, Y.
Kishi, J. Am. Chem. Soc. 2009, 131, 15642.
cost improvement and 80% waste reduction were achieved
primarily by eliminating the vast quantities of solvents
required for extensive chromatographic purification. More-
over, stereochemical quality control at all four stereogenic
centers was significantly facilitated by relying on crystalliza-
tion rather than chromatography for purification.
In conclusion, a new synthesis of the halichondrin/
Halaven C14–C26 fragment 3a has been established, which
features a number of substrate-controlled transformations
starting from readily available d-quinic acid (5) as the single
source of chirality. A novel feature of the overall route, the
stereoselective formation of a-methyl nitrile 11 in a crystal-
lization-induced diastereoselective transformation, served as
a cornerstone for crystallization-based quality control for the
entire sequence. A perhaps underappreciated aspect of
[6] For the use of cyclic haloethers for stereoselective synthesis, see:
a) R. B. Woodward, F. E. Bader, H. Bickel, A. J. Frey, R. W.
Kierstead, J. Am. Chem. Soc. 1956, 78, 2023; b) E. J. Corey, R.
Noyori, Tetrahedron Lett. 1970, 11, 311.
[7] N. G. Anderson, Org. Process Res. Dev. 2005, 9, 800.
Angew. Chem. Int. Ed. 2015, 54, 1 – 5
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
[8] G. Ulibarri, W. Nadler, T. Skrydstrup, H. Audrain, A. Chiaroni,
C. Riche, D. S. Grierson, J. Org. Chem. 1995, 60, 2753 – 2761.
[9] D. P. Stamos, Y. Kishi, Tetrahedron Lett. 1996, 37, 8643.
[10] F. F. Fleming, B. C. Shook, Tetrahedron 2002, 58, 1.
[11] a) A. R. Mattocks, J. Chem. Soc. 1964, 1918; b) A. R. Mattocks,
J. Chem. Soc. 1964, 4840.
[12] M. A. Blanchette, W. Choy, J. T. Davis, A. P. Essenfeld, S.
Masamune, W. R. Roush, T. Sakai, Tetrahedron Lett. 1984, 25,
2183.
[13] B. Bernet, A. Vasella, Helv. Chim. Acta 1979, 62, 1990.
Received: February 5, 2015
Published online: && &&, &&&&
4
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Asymmetric Synthesis
F. Belanger, C. E. Chase, A. Endo,
F. G. Fang,* J. Li, S. R. Mathieu,
A. Z. Wilcoxen, H. Zhang
&&&& — &&&&
Stereoselective Synthesis of the Halaven
C14–C26 Fragment from d-Quinic Acid:
Crystallization-Induced
Diastereoselective Transformation of an
a-Methyl Nitrile
Crystallization-induced diastereoselective
transformation of an a-methyl nitrile
completes an entirely non-chromato-
graphic synthesis of the halichondrin B
C14–C26 stereochemical array. All four
stereogenic centers in the Halaven C14–
C26 fragment were assembled on a crys-
talline polycyclic template derived from
the single chiral source D-quinic acid.
Angew. Chem. Int. Ed. 2015, 54, 1 – 5
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!