Rhodium-Catalyzed Desymmetrization of meso -Glutaric Anhydrides to Access Enantioenriched anti, anti -Polypropionates

Article abstract of DOI:10.1055/s-0036-1591488

An expedient desymmetrization of 3,5-dimethyl-4-alkoxyglutaric anhydrides to access anti, anti -polypropionates is described. The previously unknown anhydrides are rapidly assembled from readily available precursors. A Rh(I)· t -BuPHOX catalyst system was found to provide good yield and high selectivities. With these conditions, the trisubstituted anhydrides were desymmetrized with various alkyl zinc reagents to provide synthetically useful enantioenriched anti,anti -2,4-dimethyl-3-hydroxy-δ-ketoacids. An identical catalyst system also affords access to syn, syn -stereotriads as well as a partial kinetic resolution of a chiral anhydride.

Full text of DOI:10.1055/s-0036-1591488

SYNLETT
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6
© Georg Thieme Verlag Stuttgart · New York
2017, 28, A–D
letter
A
en
B. M. Cochran et al.
Letter
Synlett
Rhodium-Catalyzed Desymmetrization of meso-Glutaric Anhy-
drides to Access Enantioenriched anti,anti-Polypropionates
Brian M. Cochrana
Daniel D. Hendersona
Scott M. Thullena,b
O
O
O
OR'
O
Rh(I) PHOX
XZn
HO2C
Me
Me
R
Me
Me
R
OR'
Tomislav Rovis*a,b
0
0
0
0
0-
0
0
1
6-
2
8
7
8-
6
6
9
55–89% yield
78–96% ee
single diastereomer
11 examples
a Department of Chemistry, Colorado State University, Fort
Collins, Colorado 80523, USA
b Department of Chemistry, Columbia University, New York,
New York 10027, USA
tr2504@columbia.edu
Received: 22.08.2017
Accepted after revision: 13.09.2017
Published online: 17.10.2017
OH
O
i-Pr
Me
OH
Me
HO
O
O
H
Me H
OH
Me
DOI: 10.1055/s-0036-1591488; Art ID: st-2017-b0639-l
Me
Me
O
Me
OH
O
Me
Abstract An expedient desymmetrization of 3,5-dimethyl-4-alkoxy-
glutaric anhydrides to access anti,anti-polypropionates is described. The
previously unknown anhydrides are rapidly assembled from readily
available precursors. A Rh(I)·t-BuPHOX catalyst system was found to
provide good yield and high selectivities. With these conditions, the tri-
substituted anhydrides were desymmetrized with various alkyl zinc re-
agents to provide synthetically useful enantioenriched anti,anti-2,4-di-
methyl-3-hydroxy-δ-ketoacids. An identical catalyst system also affords
access to syn,syn-stereotriads as well as a partial kinetic resolution of a
chiral anhydride.
OH
O
Me
O
HO
O
Me
Me
Me
ionomycin
Me
Me
Me
Me
dolabriferol B
Me
OH OH OH
O
O
OH
Key words anhydride, desymmetrization, polypropionate, rhodium,
zinc
H
H
Me
Me
Me
Me
Me
Me
OH
Me
Polypropionate natural products possess a daunting
synthetic complexity that has long fascinated the synthetic
community. The best known polypropionates are likely the
macrolide antibiotics which possess distinctive 1,3-dimeth-
yl moieties. Other polypropionates, such as dolabriferol,1
ionomycin,2 and zincophorin3 also contain a similar an-
ti,anti-1,3-dimethyl-2-hydroxy motif highlighted in Figure
1. Commonly utilized approaches to synthesize compounds
like those shown in Figure 1 include aldol and crotylation
reactions, although many strategies have been reported and
reviewed.4,5 In nature, these molecules are assembled from
propionate and acetate monomers by the polyketide syn-
thase (PKS) family of enzymes.6,7 The almost limitless syn-
thetic complexity arises from enzyme-promoted Claisen
condensations, followed by partial or complete reduction of
the resultant ketones. While nature can execute lengthy it-
erative syntheses with evolved efficiency, experimental
syntheses of compounds containing several polypropionate
segments can be lengthy and low-yielding ordeals.8
Me
zincophorin
Figure 1 Natural products with the anti,anti-stereotriad
Desymmetrization of a meso starting material contain-
ing latent stereocenters establishes multiple stereocenters
in a single operation,9 but it is rarely used for the construc-
tion of polypropionate fragments.10,2c We have described Ni,
Pd, and Rh catalysts for the cross-coupling of cyclic anhy-
drides, culminating in a Rh-catalyzed asymmetric de-
symmetrization of meso-3,5-dimethylglutaric anhydride
with alkyl zinc reagents.11–13 The valuable syn-1,3-dimethyl
deoxypolypropionate stereodiad was thereby expeditiously
accessed. Application of this methodology to hydroxylated
glutaric anhydride 1 should result in access to polypropio-
nates with oxygenation as shown in Scheme 1. However,
the requisite anhydride 1 was not known, and the route
used to access the deoxyanhydride was not amenable to
substitution at that position. Furthermore, it was not obvi-
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 28, A–D

B
B. M. Cochran et al.
Letter
Synlett
ous that additional steric bulk would be tolerated or the
conditions mild enough to prevent elimination of the gen-
erated product. Herein, we disclose the successful realiza-
tion of this goal.
R
OH OR OH
O
i. Swern
i. 9-BBN,THF
ii. Pinnick
iii. TFAA
ii. NaOH, H2O2
Me
Me
Me
Me
76%, > 20:1:1 dr
71%, 3.5:1 dr
3a R = TBS
3b R = Bn
4a R = TBS
4b R = Bn
85%
66%
Previous Work:
O
O
O
O
O
O
O
[Rh(nbd)Cl]2
62–85%
85–95% ee
HO2C
(S)-t-BuPHOX
Me
Me
Me
Me
Me
Me
Me
R
R
OR
RCH2ZnX
i. BF3⋅OEt2
1a R = TBS
1b R = Bn
1c R = Ac
1d R = Bz
This Work:
ii Ac2O, 88%
ii Bz2O, 43%
O
O
O
[Rh(nbd)Cl]2
(S)-t-BuPHOX
OR'
O
HO2C
Scheme 2 Synthesis of meso-glutaric anhydrides
Me
Me
RCH2ZnX
Me
OR'
1
2
ly undemanding ligand is used (Table 1, entry 3). Gratify-
ingly, when the size of the protecting group on the alcohol
was reduced to an acetate (1c), ketoacid 2c was isolated in
moderate yield and good enantioselectivity (Table 1, entry
3). Notably, when the diene ligand of the rhodium precata-
lyst was changed from cyclooctadiene to norbornadiene,
the product yield and enantioselectivity increased (Table 1,
entry 4).15,16 Desymmetrization of anhydride 1d and 1b (Bz
and Bn; Table 1, entries 5 and 6) also provides high yields
and enantioselectivities of ketoacids 2d and 2b.17 Due to the
inherent stability of the Bn protecting group vs the Ac/Bz
group,18 anhydride 1b was used to further explore the
scope.
O
PPh2
N
(S) t-BuPHOX
t-Bu
Scheme 1 Extension of anhydride desymmetrization for anti,anti frag-
ments
After surveying a number of approaches, we found that
the 4-hydroxy-3,5-dimethyl glutaric anhydride desymme-
trization precursor was readily accessed by a three-step
process from dienes 3a,b (Scheme 2). Utilizing hydrobora-
tion conditions developed by Harada,14 O-TBS-protected di-
ene 3a was selectively oxidized to the anti-anti-diol using
9-BBN followed by hydrogen peroxide in >20:1:1 dr
(Scheme 2). An O-Bn analogue 3b was also oxidized to the
diol under the same conditions with moderate selectivity
(3.5:1 dr). Although accessing the anhydride from the diol
appears trivial, the intermediacy of the lactol after the first
oxidation and its inevitable oxidation to the lactone compli-
cated the approach; the lactone proved resistant to further
oxidation. Stepwise oxidation of the diol to the diacid was
required to prevent lactol/lactone formation. To this end, a
Swern followed by a Pinnick oxidation proceeded smoothly
for both O-protected substrates. Cyclization of the diacid
using trifluoroacetic anhydride gave the desired anhydrides
in high yield over three steps. By this sequence, multigram
quantities of anhydrides 1a and 1b were synthesized. Fur-
thermore, the TBS group of 1a was manipulated to an Ac
(1c) or a Bz (1d) in a two-step process utilizing BF3·OEt2 fol-
lowed by the appropriate acylating agent.
Table 1 Screen of O-Protecting Groups
O
O
O
catalyst (5 mol%)
(S) t-Bu-PHOX (10 mol%)
MeZnBr (1.7 equiv)
OR
O
HO2C
Me
Me
Me
Me
Me
THF
OR
1a–d
2a–d
Entry
R
Catalyst
Yield (%)
ee (%)
1
2
3
4
5
6
1a TBS
1a TBS
1c Ac
1c Ac
1d Bz
1b Bn
[Rh (COD)Cl]2
[Rh(nbd)Cl]2
[Rh (COD)Cl]2
[Rh(nbd)Cl]2
[Rh(nbd)Cl]2
[Rh(nbd)Cl]2
decomp.
decomp.
47
–
–
84
91
86
91
90
50
98
An examination of the nucleophile scope beyond com-
mercially available dialkyl zinc and alkyl zinc halide re-
agents was initially met with poor reactivity (Table 2). Dioc-
tyl zinc was generated in situ from 1-iodooctane and dieth-
yl zinc and when implemented using the above reaction
conditions, ketoacid 5 was formed in 21% yield (Table 2, en-
try 1). Postulating that incomplete formation of the dioctyl
zinc reagent results in the low yield, catalytic copper iodide
was added during its formation19 and the ketoacid yield in-
creased to 47% (Table 2, entry 2). Concurrently with the de-
The trisubstituted anhydrides 1a–d were subjected to
Rh(I)/PHOX/MeZnBr cross-coupling according to conditions
previously developed for 3,5-dimethylgultaric anhydride
(Table 1).12 Anhydride 1a failed to deliver ketoacid 2a under
these conditions and only decomposition products were
detected (Table 1, entries 1 and 2). Presumably, the steric
bulk of the TBS group on the anhydride prevents the oxida-
tive insertion of the rhodium complex even when a sterical-
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 28, A–D

C
B. M. Cochran et al.
Letter
Synlett
velopment of this methodology, our laboratory successfully
explored hydroheteroarylation using rhodium(I) acetate
complexes.20 We envisioned a rhodium(I) acetate complex
may undergo faster transmetalation with the dioctylzinc
nucleophile as compared to a halogenated precatalyst. Add-
ing 10 mol% Zn(OAc)2to the current reaction conditions
drastically improved the yield of the ketoacid to 89% (Table
2, entry 3) presumably by making the rhodium acetate
complex in situ. Indeed, when subjecting the isolated rhodi-
um acetate norbornadiene complex to the reaction in the
absence of Zn(OAc)2 (Table 2, entry 4), the ketoacid was iso-
lated with a comparable yield.
[Rh(nbd)OAc]2
O
O
O
OBn
O
R
(S)-t-BuPHOX
XO2C
Me
Me
R2Zn or RZnBr
THF, 25 °C
Me
Me
OBn
X = H, Me after CH2N2
OBn
O
R = Me, 2a*
R = n-octyl, 5 89%, 91% ee
R = Et, 6 84%, 86% ee
84%, 92% ee
HO2C
R
Me
Me
OBn
O
OBn O
MeO2C
Cl
MeO2C
OAc
Me
Me
Me
8 70%, 92% ee
OBn
Me
7* 84%, 96% ee
OBn
Table 2 Desymmetrization Optimization
O
O
(n-C8H17)2Zn
catalyst
MeO2C
CO2Et
MeO2C
O
O
O
OBn
O
(S)-t-BuPHOX
Me
9 73%, 94% ee
OBn
Me
Me
Me
HO2C
n-C8H17
10 83%, 86% ee
THF, 25 °C
Me
Me
Me
Me
OBn
F
OMe
O
OBn O
MeO2C
MeO2C
Entrya
Additive (mol%)b Catalyst (mol%)
Yield (%)
Me
Me
Me
Me
1
2
3
4
none
[Rh(nbd)Cl]2 (5)
[Rh(nbd)Cl]2 (5)
21
47
89
83
11 78%, 85% ee
12 80%, 88% ee
CuI (0.3)
CuI (0.3)
CuI (0.3)
[Rh(nbd)Cl]2 (5) + Zn(OAc)2 (10)
[Rh(nbd)OAc]2 (5)
Scheme 3 Scope of alkyl zinc reagents. All reactions performed on 0.2
mmol scale with 5 mol% [Rh] dimer, 10 mol% t-BuPHOX and 1.7 equiv
R2Zn or RZnX in THF (0.07 M) at r.t. for 48 h. *[Rh(nbd)Cl]2 used in place
of [Rh(nbd)OAc]2.
a All reactions performed with 5 mol% [Rh], 10 mol% PHOX, and 1.7 equiv
Oct2Zn in THF (0.07 M) at 25 °C for 48 h.
b Additive used in formation of Oct2Zn from OctI and Et2Zn.
[Rh(nbd)Cl]2 (5 mol%)
(S)-t-BuPHOX (10 mol%)
O
O
O
With these optimized conditions, an array of alkyl zinc
reagents were effectively used in the desymmetrization of
anhydride 1b (Scheme 3). Small and long-chain alkyl
groups were successfully added to give ketoacids 2a, 5, and
6 in good yield and good ee. The reaction tolerates chloride
(7) and ester (8 and 9) functionalities having minimal im-
pact on yield and ee. Benzyl groups with electron-donating
and electron-withdrawing substituents (10–12) were also
efficient in the desymmetrization reaction.
OAc
O
66% yield
78% ee
HO2C
Me
Me
Me
Me
Me2Zn
THF, 25 °C
Me
Me
Me
OAc
13
14
[Rh(nbd)Cl]2 (5 mol%)
(S)-t-BuPHOX (10 mol%)
O
O
O
OAc
O
55% yield
4:1 dr
69% ee (major)
HO2C
Me2Zn
THF, 25 °C
Me
Me
Me
16
OAc
15
In addition to using this methodology to synthesize an-
ti,anti-polypropionates, we briefly explored the desymme-
trization of anhydrides to form syn,syn- and syn,anti-poly-
propionates with encouraging results (Scheme 4).21 meso-
Anhydride 13 successfully reacted with dimethylzinc utiliz-
ing early desymmetrization conditions to give ketoacid 14
in moderate yield and enantioselectivity. Additionally, race-
mic anhydride 15 was desymmetrized with dimethyl zinc
in an enantio- and diastereoselective manner to give keto-
acid 16 in 4:1 dr and 69% ee (major).22 These unoptimized
results are encouraging leads for further optimization.
In summary, we have demonstrated new methodology
for the convergent synthesis of the anti,anti-polypropionate
ketoacids utilizing a Rh(I)-catalyzed desymmetrization of
anhydrides 1b–d with primary alkyl zinc reagents and a t-
Scheme 4 Accessing syn,syn- and syn,anti-adducts
Bu-PHOX ligand.23,24 The anhydride precursors are readily
accessed from a single diene. Through optimization, a Rh(I)
acetate precatalyst was identified as superior vs a Rh(I)
chloride precatalyst in the reaction. With this more reactive
system, a variety of zinc nucleophiles were successfully
used to desymmetrize anhydride 1b in high yields and en-
antioselectivity to give synthetically useful anti,anti-poly-
propionates.
Funding Information
We thank the National Science Foundation for support of this work.
)(
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 28, A–D

D
B. M. Cochran et al.
Letter
Synlett
Supporting Information
(16) Johnson, J. B.; Rovis, T. Angew. Chem. Int. Ed. 2008, 47, 840.
(17) Absolute stereochemistry was determined by single crystal
analysis of a sulfoxide derivative of 2b. See Supporting Informa-
tion.
Supporting information for this article is available online at
https://doi.org/10.1055/s-0036-1591488.
S
u
p
p
ortiInfogrmoaitn
S
u
p
p
ortioInfgrmoaitn
(18) Over long reaction times or higher temperatures, the acetate
and benzoate groups eliminate.
References and Notes
(19) Knochel, P.; Singer, R. D. Chem. Rev. 1993, 93, 2117.
(20) Filloux, C. M.; Rovis, T. J. Am. Chem. Soc. 2015, 137, 508.
(21) Anhydrides 13 and 15 were synthesized in a similar manner as
1c and 1d utilizing alternative hydroboration conditions devel-
oped by Harada (see ref. 14). See Supporting Information.
(22) Determination of a krel selectivity factor in this reaction was
frustrated by anhydride hydrolysis and decomposition on
workup.
(1) (a) Ciavatta, M. L.; Gavagnin, M.; Puliti, R.; Cimino, G.; Martinez,
E.; Ortea, J.; Mattia, C. A. Tetrahedron 1996, 52, 12831.
(b) Currie, R. H.; Goodman, J. M. Angew. Chem. Int. Ed. 2012, 51,
4695. (c) Karagiannis, A.; Diddi, N.; Ward, D. E. Org. Lett. 2016,
18, 3794.
(2) (a) Evans, D. A.; Dow, R. L.; Shih, T. L.; Takacs, J. M.; Zahler, R.
J. Am. Chem. Soc. 1990, 112, 5290. (b) Hanessian, S.; Cooke, N. G.;
DeHoff, B.; Sakito, Y. J. Am. Chem. Soc. 1990, 112, 5276.
(c) Lautens, M.; Colucci, J. T.; Hiebert, S.; Smith, N. D.; Bouchain,
G. Org. Lett. 2002, 4, 1879. (d) Gao, Z.; Li, Y.; Cooksey, J. P.;
Snaddon, T. N.; Schunk, S.; Viseux, E. M. E.; McAteer, S. M.;
Kocienski, P. J. Angew. Chem. Int. Ed. 2009, 48, 5022.
(3) For recent contributions, see: (a) Kasun, Z. A.; Gao, X.; Lipinski,
R. M.; Krische, M. J. J. Am. Chem. Soc. 2015, 137, 8900. (b) Chen,
L.-A.; Ashley, M. A.; Leighton, J. L. J. Am. Chem. Soc. 2017, 139,
4568.
(4) Li, J.; Menche, D. Synthesis 2009, 2293.
(5) Gao, X.; Han, H.; Krische, M. J. J. Am. Chem. Soc. 2011, 133,
12795.
(6) Weissman, K. J. In Methods in Enzymology, Vol. 459; Elsevier
2009, 3–16.
(7) Katz, L. In Methods in Enzymology, Vol. 459; Elsevier 2009, 113–
142.
(8) Recent methodological breakthroughs are focused on circum-
venting this problem. See references 3 and 5 and references
therein.
(9) Willis, M. C. J. Chem. Soc., Perkin Trans. 1 1999, 1765.
(10) Mohr, P.; Waespe-Šarčević, N.; Tamm, C. Helv. Chim. Acta 1983,
66, 2501.
(11) (a) Bercot, E. A.; Rovis, T. J. Am. Chem. Soc. 2002, 124, 174.
(b) Bercot, E. A.; Rovis, T. J. Am. Chem. Soc. 2004, 126, 10248.
(c) Johnson, J. B.; Rovis, T. Acc. Chem. Res. 2008, 41, 327.
(d) Stache, E. E.; Rovis, T.; Doyle, A. G. Angew. Chem. Int. Ed.
2017, 56, 3679.
(12) Cook, M. J.; Rovis, T. J. Am. Chem. Soc. 2007, 129, 9302.
(13) Johnson, J. B.; Cook, M. J.; Rovis, T. Tetrahedron 2009, 65, 3202.
(14) Harada, T.; Matsuda, Y.; Wada, I.; Uchimura, J.; Oku, A. J. Chem.
Soc., Chem. Commun. 1990, 21.
(23) General Procedure
To a 5 mL flask was added [Rh(nbd)Cl]2 (5 mol%) and (S)-t-
BuPHOX (10 mol%). In some cases, Zn(OAc)2(10 mol%) was also
added. The flask was capped with a septum, placed under
vacuum, and the atmosphere was replaced with argon. The
solids were dissolved in THF (0.4 M vs anhydride) and freshly
prepared organozinc reagent (see below; 0.2 M in THF, 1.7
equiv) was added by syringe producing a deep red solution. The
appropriate anhydride (1 equiv) was dissolved in THF (0.25 M)
and was added to the reaction by syringe taking care to remove
all air from the headspace. The reaction was stirred at r.t. for
16–24 h, quenched with 1 M HCl and extracted in to EtOAc (3×).
(24) Representative Example
(2R,3R,4S)-3-(Benzyloxy)-2,4-dimethyl-5-oxohexanoic Acid
(2b)
Following the general procedure on 1 g scale of the anhydride
and using MeZnBr, a colorless oil was isolated (0.91 g, 87%).
Conversion into the methyl ester utilizing diazomethane
allowed for enantiomeric excess determination: Chiralpak IA
column eluting with 99:1 hexanes/isopropanol, eluting at 1.0
mL/min, showing 92% ee, with the major enantiomer eluting at
19.26 min and the minor at 22.74 min. Rf = 0.26 (59:40:1, hex-
20
ane/EtOAc/AcOH). [α]D = –6.5 (c 0.011 g/mL, CH2Cl2). IR (thin
film): νmax = 3090, 3032, 2982, 2886, 1738, 1711, 1456, 1379,
1190, 1067, 738, 699 cm–1 1H NMR: (300 MHz, CDCl3): δ =
.
11.12 (br, 1 H), 7.24–7.18 (5 H, m), 4.53 (1 H, d, J = 11.2 Hz), 4.45
(1 H, d, J = 11.2 Hz), 3.96 (1 H, dd, J = 4.0, 8.0 Hz), 2.90 (1 H,
quint, J = 7.2 Hz), 2.78 (1 H, dq, J = 4.4, 7.2 Hz), 2.17 (3 H, s), 1.25
(3 H, d, J = 7.2 Hz), 1.07 (3 H, d, J = 7.2 Hz). 13C NMR (101 MHz,
CDCl3): δ = 211.1, 179.1, 137.6, 128.4, 127.8, 82.7, 74.4, 49.2,
41.6, 30.3, 12.9, 12.6. LRMS (ESI, pos.): m/z calcd for C15H20O4Na
[M + Na]+: 287.30; found: 287.1.
(15) See Supporting Information for additional details.
© Georg Thieme Verlag Stuttgart · New York — Synlett 2017, 28, A–D