Synthesis of the N-(tert-butyloxycarbonyl)-O-triisopropylsilyl-d- pyrrolosamine glycal of lomaiviticins A and B via epimerization of l-Threonine

Article abstract of DOI:10.1016/j.tetlet.2010.06.028

An efficient synthesis of the N-(tert-butyloxycarbonyl)-O- triisopropylsilyl-d-pyrrolosamine glycal of lomaiviticin A (1) and lomaiviticin B (2) is described. The synthesis is highlighted by the epimerization of the l-threonine-derived oxazolidine 10 to oxazolidine 11. This key epimerization reaction, which serves to establish the correct relative configuration of the carbohydrate unit, was made possible only after conformational analysis indicated that substituted oxazolidines may adopt conformations that preclude enolization.

Full text of DOI:10.1016/j.tetlet.2010.06.028

Tetrahedron Letters 51 (2010) 4310–4312
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Synthesis of the N-(tert-butyloxycarbonyl)-O-triisopropylsilyl-D-pyrrolosamine
glycal of lomaiviticins A and B via epimerization of L-Threonine
*
William J. Morris, Matthew D. Shair
Department of Chemistry and Chemical Biology, Harvard University, 12 Oxford Street Cambridge, MA 02138, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
An efficient synthesis of the N-(tert-butyloxycarbonyl)-O-triisopropylsilyl-D-pyrrolosamine glycal of
Received 21 May 2010
Accepted 7 June 2010
Available online 12 June 2010
lomaiviticin A (1) and lomaiviticin B (2) is described. The synthesis is highlighted by the epimerization
of the -threonine-derived oxazolidine 10 to oxazolidine 11. This key epimerization reaction, which
L
serves to establish the correct relative configuration of the carbohydrate unit, was made possible only
after conformational analysis indicated that substituted oxazolidines may adopt conformations that pre-
clude enolization.
Ó 2010 Elsevier Ltd. All rights reserved.
In 2001, He and co-workers reported the isolation and charac-
terization of lomaiviticin A (1) and lomaiviticin B (2) (Fig. 1).1
These molecules are potent growth inhibitors against 24 cultured
human cancer cell lines (GI50 = 0.01–98 ng/ml). The cytotoxicity
patterns of 1 and 2 in a 24 cell line panel of human cancer cells
are unique, suggesting that they have novel mechanisms of action.
In addition to their potent activity in cells, 1 and 2 are unprec-
edented C2-symmetric structures. They share identical core struc-
tures, but lomaiviticin A is glycosylated at C3 and C30 while the
C3 and C30 carbinols of lomaiviticin B are engaged as ketals with
C1 and C10. The C4 and C40 carbinols of 1 and 2 are glycosylated
with rare N,N-dimethylpyrrolosamine carbohydrates. Both 1 and
2 possess a diazobenzofluorene ring system that evokes compari-
sons to the kinamycin family of natural products.2 Progress toward
the synthesis of 1 and 2 has been reported,3 including our approach
to the central ring system of lomaiviticin A using a stereoselective
oxidative enolatedimerization of a 7-oxanorbornanone.4
Recently, the synthesis of the N,N-dimethylpyrrolosamine car-
bohydrate found in both 1 and 2 has been addressed by our group5
as well as Herzon and co-workers.6 In this Letter, we describe an
alternative synthesis of the N,N-dimethylpyrrolosamine sugar that
utilizes an interesting and useful epimerization reaction.
Our initial synthesis plan is outlined in Scheme 1. We targeted a
suitably protected glycal that could ultimately be converted to a
glycosyl donor. The retrosynthetic analysis began from glycal 3
which would be obtained via cycloisomerization of 4. Alkynol 4
would be accessed from methyl ester 5, which could be derived
that several methods are available for its preparation.8 Despite
the availability of these methods, we were interested in devising
a more efficient strategy to access this important amino acid. Spe-
cifically, we sought to develop a strategy where
L-threonine 7 could
be epimerized at the amino stereocenter to provide the desired
D-allo-threonine configuration since 7 is readily available (Scheme 2).
Our revised plan was to start the synthesis route outlined in
Scheme 1 with
reomer 6. We surmised that the enolate of the
L
-threonine (7) instead of its more expensive diaste-
-threonine-derived
L
oxazolidine 8 would be protonated from the si-face, providing the
desired configuration at the amino stereocenter. This epimeriza-
tion strategy offered two distinct advantages over the approaches
previously reported in the literature. First, our synthesis would be-
gin from 7, a cheap and readily available starting material. Second,
this strategy provides an alternative to undertaking a separate syn-
thesis to procure multigram quantities of
lizing an intermediate in our proposed route to the target glycal 3.
Toward this end, -threonine was readily converted to oxazoli-
D-allo-threonine by uti-
L
dine 8 (Scheme 3). Initially, we chose to carry out a control exper-
iment to test the feasibility of the approach outlined in Scheme 2.
Oxazolidine 8 was treated with LDA at ꢀ78 °C followed by expo-
sure to MeI. The purpose of using MeI in this control experiment
was twofold. While serving to confirm the facial selectivity of the
alkylation (and ultimately, the protonation), this experiment
would also allow us to unambiguously confirm if enolization was
achieved.9 Surprisingly, after oxazolidine 8 was treated with LDA
at ꢀ78 °C followed by MeI the starting material was recovered
unchanged.
Though we acknowledged the possibility that the enolate was
too hindered to be alkylated with MeI, we considered this scenario
to be unlikely.10 It seemed more probable that the enolate had not
been formed. This observation can be rationalized by considering
possible conformations of oxazolidine 8 (Fig. 2). An important
consideration to the following analysis is that the carbamate will
from the amino acid
An initial challenge to this synthesis plan was the limited com-
mercial availability of
-allo-threonine 6.7 Given the potential util-
D-allo-threonine (6).
D
ity of this amino acid in organic synthesis, it was not surprising
* Corresponding author. Tel.: +1 617 495 5008; fax: +1 617 496 4591.
E-mail address: shair@chemistry.harvard.edu (M.D. Shair).
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.06.028

W. J. Morris, M. D. Shair / Tetrahedron Letters 51 (2010) 4310–4312
OH
4311
NMe2
O
OH
O
Me
1. SOCl2, MeOH
Me
OH
Me
OMe
NHBoc
Me
OH
O
2. (Boc)2O, NEt3
CH2Cl2
O
OH
OMe
HO
NH2
N
9
7
O
96%, 2 steps
N
Et
O
O
O
O
4
1
5
H
MeO OMe
Me Me
2
HO
Me
O
6'
1'
2'
4'
1. LDA, –78 °C
2. MeI, 23 °C
OH
No
Reaction
H
O
O
O
OMe
O
O
BF3•OEt2,
CH2Cl2
NBoc
Et
O
N
Me
O
OH
8
Me
N
MeO
HO
80%
O
HO
1
Me
Scheme 3. Synthesis of epimerization precursor.
Me
Me2N
possess double bond character. In conformer I, the hydrogen
a to
NMe2
OH
Me
the ester is placed in plane with the carbamate in order to mini-
mize allylic strain. This forces the ester into a pseudo-axial position
and introduces an unfavorable syn-pentane interaction between
the ester and the methyl substituent cis to the ester. An alternative
conformer (II) is one in which the ester is placed in a pseudo-equa-
torial position, thereby alleviating the syn-pentane interaction.
However, in order to minimize unfavorable steric interactions with
the carbamate, the ester rotates about the C–C bond (shown in
O
N
N
HO
O
H
Et
O
O
O
H
OH
HO
OH
HO
red). As a consequence of this bond rotation, the
a-hydrogen is
O
O
O
no longer stereoelectronically aligned for deprotonation.
This conformational analysis indicates that enolization will only
take place if conformer I could be accessed. Using this analysis as a
guide we sought to redesign the substrate such that the syn-pen-
tane interaction would be alleviated. We rationalized this would
be done most effectively by replacing the methyl group cis to the
methyl ester with a proton.
Et
O
N
OH
N
2
O
HO
Me
Me2N
Figure 1. Lomaiviticin A (1) and lomaiviticin B (2).
Toward that end, methyl ester 9 was treated with pivaldehyde
under acidic conditions to afford oxazolidine 10 (Scheme 4). With
the syn-pentane interaction now removed, we were pleased to dis-
cover that exposure of 10 to LDA followed by a reverse quench
with AcOH/MeOH afforded the fully epimerized product in
quantitative yield.11 This result strongly supports the assertion
Me
OTIPS
O
OH OTIPS
BocHN
Me
O
OMe
Me
OH
NBoc
Me
Me
NHBoc
O
3
5
4
H
Me
CO2Me
Me
O
OH
O
O
Me
OMe
O
Me
O
N
N
Me
t-BuO
H
Me
H
H
t-BuO
Me
NH2
O
(I)
(II)
6
Scheme 1. Retrosynthetic analysis.
Figure 2. Conformational analysis of oxazolidine 8.
Me
O
OH
O
pPTS, pivaldehyde
O
OMe
Me
OMe
Me
O
triethylformate
OH
O
NBoc
3 steps
NHBoc
1. LDA
2. H+ quench
PhMe, 90 °C
71%, dr 15:1
t-Bu
9
10
O
OMe
Me
OH
NBoc
Me
NH2
7
8
Me
Me
O
Me
O
LDA, –78 °C
O
OMe
O
then
OMe
NBoc
AcOH, MeOH
NBoc
t-Bu
Me
11
5
quantitative
Me
Scheme 2. Proposed epimerization of
L
-threonine.
Scheme 4. Synthesis of revised epimerization substrate.

4312
W. J. Morris, M. D. Shair / Tetrahedron Letters 51 (2010) 4310–4312
3. (a) Nicolaou, K. C.; Denton, R. M.; Lenzen, A.; Edmonds, D. J.; Li, A.; Milburn, R.
Me
Me
O
O
1. MeNH(OMe)•HCl
R.; Harrison, S. T. Angew. Chem., Int. Ed. 2006, 45, 2076–2081; (b) Zhang, W.;
Baranczak, A.; Sulikowski, G. A. Org. Lett. 2008, 10, 1939–1941; (c) Nicolaou, K.
C.; Nold, A. N.; Li, H. Angew. Chem., Int. Ed. 2009, 48, 5860–5862.
4. Krygowski, E. S.; Murphy-Benenato, K.; Shair, M. D. Angew. Chem., Int. Ed. 2008,
47, 1680–1684.
i-PrMgCl, THF, –20 °C
O
O
OMe
NBoc
NBoc
11
2. Ethynyl Grignard,
THF, 0 °C to 23 °C
t-Bu
t-Bu
12
5. Moris, W. J.; Shair, M. D. Org. Lett. 2008, 11, 9–12.
6. Gholap, S. L.; Woo, C. M.; Ravikumar, P. C.; Herzon, S. B. Org. Lett. 2009, 11,
4322–4325.
68%, 2 steps
7. Recently, the production of
Aldrich.
D-allo-threonine has been discontinued by Sigma–
Me
OTIPS
1. NaBH4, MeOH, 0 °C
AcOH
8. (a) Evans, D. A.; Sjogren, E. B.; Weber, A. E.; Conn, R. E. Tetrahedron Lett. 1987,
28, 39–42; (b) Genet, J. P.; Juge, S.; Mallart, S. Tetrahedron Lett. 1988, 29, 6765–
6768; (c) Pons, D.; Savignac, M.; Genet, J. P. Tetrahedron Lett. 1990, 31, 5023–
5026; (d) Beaulieu, P. L. Tetrahedron Lett. 1991, 32, 1031–1034; (e) Blaskovich,
M. A.; Lajoie, G. A. Tetrahedron Lett. 1993, 34, 3837–3840; (f) Fischer, P. M.;
Sandosham, J. Tetrahedron Lett. 1995, 36, 5409–5412; (g) Shao, H.; Goodman, M.
J. Org. Chem. 1996, 61, 2582–2583; (h) Lloyd-Williams, P.; Sanchez, A.; Carulla,
N.; Ochoa, T.; Giralt, E. Tetrahedron 1997, 53, 3369–3382; (i) Albanese, D.;
Landini, D.; Lupi, V.; Penso, M. Eur. J. Org. Chem. 2000, 1443–1449.
O
2. TIPSOTf, CH2Cl2
0 °C
THF/H2O
NBoc
65 °C
58%
t-Bu
13
88%, 2 steps
dr 4:1
OTIPS
OH OTIPS
BocHN
Me
(Ph3P)3RhCl
9. The product resulting from protonation of the enolate from the re-face would
be identical to the starting material. Therefore, we would not be able to
Me
unambiguously determine whether this product was
protonation or failure to achieve enolate formation.
a result of re-face
DMF, 80 °C
76%
O
NHBoc
3
4
10. Similar serine-derived oxazolidines have been alkylated with MeI: Brunner,
M.; Saarenketo, P.; Straub, T.; Rissanen, K.; Koskinen, A. M. P. Eur. J. Org. Chem.
2004, 3879–3883.
Scheme 5. Synthesis of glycal 3.
11. The resulting oxazolidine 11 was cleaved under acidic conditions. The 1H NMR
of the resulting b-hydroxy-a
-amino ester matched the known 1H NMR of the of
N-Boc-
L-allo-threonine methyl ester:
that enolization was precluded because of unfavorable non-cova-
lent interactions described in Figure 2.
Me
O
After the successful epimerization12 of 10, oxazolidine 11 was
converted to a Weinreb amide followed by treatment with ethynyl
Grignard (Scheme 5). The ynone 12 was reduced to the corre-
sponding propargyl alcohol with NaBH4.13 The resulting carbinol
was protected as a TIPS ether and the oxazolidine was cleaved un-
der acidic conditions. The resulting alkynol underwent cycloiso-
merization14 in the presence of Wilkinson’s catalyst to provide
glycal 3 in 76% yield.15
OH
O
AcOH, THF
O
OMe
Me
OMe
NHBoc
H2O, 65 °C
50%
NBoc
t-Bu
11
Ishizuka, T.; Kunieda, T. Tetrahedron Lett. 1987, 28, 4185–4188.
12. Experimental procedure for epimerization of oxazolidine 10: To a solution of
diisopropylamine (20 ml, 143 mmol) in THF (252 ml) at ꢀ78 °C was added
nBuLi (53 ml, 126 mmol) dropwise. The resulting LDA solution was allowed to
stir at ꢀ78 °C for 10 min. A solution of oxazolidine 10 (15.5 g, 57.2 mmol) in
THF (114 ml) was added to the cooled solution (ꢀ78 °C) of LDA dropwise over
20 min. The resulting pale yellow solution was allowed to stir at ꢀ78 °C for 2 h.
The enolate solution was transferred to a rapidly stirring mixture of AcOH
(78 ml) in MeOH (200 ml) at ꢀ78 °C and allowed to stir for 5 min. The reaction
mixture was diluted with EtOAc (200 ml), washed with saturated NaHCO3
(3 ꢁ 100 ml), and brine. The organic layer was dried over Na2SO4, filtered, and
concentrated to afford oxazolidine 11 (15.5 g, quantitative). Characterization
data for 11: 1H NMR (600 MHz, CDCl3) d 4.94 (s, 1H), 4.57–4.55 (d, J = 7.1 Hz,
1H), 4.28–4.24 (ddd, J = 6.5, 7.1, 12.7 Hz, 1H), 3.76 (s, 3H), 1.49 (s, 9H), 1.35–
1.34 (d, J = 6.5 Hz, 3H), 1.04 (s, 9H); 13C NMR (125 MHz, CDCl3) d 170.4, 155.5,
96.9, 74.3, 64.1, 51.8, 36.8, 28.4, 26.5, 26.3, 16.0. HRMS (ESI) Mass calcd for
In conclusion, we have developed an efficient synthesis of the
N-(tert-butyloxycarbonyl)-O-triisopropylsilyl-
glycal of lomaiviticin A (1) and lomaiviticin B (2). Our synthesis is
highlighted by the epimerization of the -threonine derived oxazol-
idine 10 to oxazolidine 11, which possesses the desired relative
configuration. This epimerization reaction was made possible only
after control experiments indicated that substituted oxazolidines
may adopt conformations that preclude enolization. Glycal 3 will
ultimately be converted to a suitable glycosyl donor for the glyco-
sylation of the aglycones of 1 and 2. These results will be reported
in due course.
D-pyrrolosamine
L
C
15H27NO5 [M+Na]+, 324.1781. Found 324.1793.
13. The stereochemistry of the NaBH4 reduction was confirmed by treating the
resulting propargyl alcohol with TFA in CH2Cl2:
Acknowledgments
Ha
Me
OH
Me
O
TFA, CH2Cl2
BocHN
Hb
O
t-Bu
O
Financial support for this project was provided by the
NIH (CA125240), Merck Research Laboratories, Novartis, and
AstraZeneca.
23 °C
60%
NBoc
H
Ja-b = 1.3 Hz
t-Bu
Villard, R.; Fotiadu, F.; Buono, G. Tetrahedron: Asymmetry 1998, 9, 607–611.
14. Trost, B. M.; Rhee, Y. H. J. Am. Chem. Soc. 2003, 125, 7482–7483.
References and notes
15. Characterization data for 3: 1H NMR (600 MHz, CDCl3)
d 6.35–6.34 (d,
J = 6.3 Hz, 1H), 4.84–4.83 (m, 1H), 4.67–4.65 (m, 1H), 4.20–4.18 (ddd, J = 3.4,
9.7, 11.6 Hz 1H), 4.17–4.16 (br s, 1H), 3.70 (m, 1H), 1.47 (s, 9H), 1.46–1.45
(d, J = 6.4 Hz, 3H), 1.14–1.13 (m, 3H), 1.10–1.09 (d, J = 6.3 Hz, 18H); 13C NMR
(125 MHz, CDCl3) d155.1, 143.3, 103.0, 79.6, 74.0, 66.1, 55.5, 28.5, 18.2, 17.9,
12.5. HRMS (ESI) Mass calcd for
408.2533.
1. He, H.; Ding, W.-D.; Bernan, V. S.; Richardson, A. D.; Ireland, C. M.; Greenstein,
M.; Ellestad, G. A.; Carter, G. T. J. Am. Chem. Soc. 2001, 123, 5362–5363.
2. (a) Gould, S. J.; Tamayo, N.; Melville, C. R.; Cone, M. C. J. Am. Chem. Soc. 1994,
116, 2207–2208; (b) Mithani, S.; Weeratunga, G.; Taylor, N. J.; Dmitrienko, G. I.
J. Am. Chem. Soc. 1994, 116, 2209–2210; (c) Marco-Contrelles, J.; Molina, M. T.
Curr. Org. Chem. 2003, 7, 1433–1442.
C
20H39NO4Si [M+Na]+, 408.2540. Found