Synthetic studies toward spiroleucettadine

Article abstract of DOI:10.1016/j.tetlet.2006.03.024

Synthetic hydroxydienone precursors to spiroleucettadine and to an isomer thereof resist cyclization to the orthoamide-type functionality present in the proposed structure of the natural product.

Full text of DOI:10.1016/j.tetlet.2006.03.024

Tetrahedron Letters 47 (2006) 3599–3601
Synthetic studies toward spiroleucettadine
Jonah J. Chang, Bryan Chan and Marco A. Ciufolini*
Department of Chemistry, University of British Columbia, 2036 Main Mall, Vancouver, BC, Canada V6T 1Z1
Received 11 January 2006; accepted 3 March 2006
Abstract—Synthetic hydroxydienone precursors to spiroleucettadine and to an isomer thereof resist cyclization to the orthoamide-
type functionality present in the proposed structure of the natural product.
Ó 2006 Elsevier Ltd. All rights reserved.
Me
N
PAN
O
In 2004, Crews and collaborators described a notewor-
thy marine alkaloid, which they termed spiroleucetta-
dine and to which they assigned structure 1 (Fig. 1) on
the basis of spectroscopic and chiroptical evidence.1
The substance displays antibiotic activity and a highly
novel architecture that renders it a prime candidate for
synthetic work. A noteworthy feature of 1 is a trans-
fused 1-hydroxy-2-oxa-6,8-diazabicyclo[3.3.0]octane core
that sustains an unusual orthoamide-type functionality.
We estimate that this system contains no less than
14 kcal/mol (MM+)2 more strain energy than the cis-
fused analog, 4. As shown in Figure 2, the ortho-
amide-type functionality would be expected to promote
facile equilibration of the natural product with 2 and 3,
through reversible release of either a C–O or a C–N
bond. Either 2 or 3 could then recyclize to form the less
energetic 4. In a like vein, the hypothetical species 3
might cylize to spiroleucettadine isomers 5 (cis or trans).
Yet, the published spectra of the natural product clearly
show the presence of a single compound, implicating a
significant preference for structure 1 over isomers 5, de-
spite the lack of obvious factors favoring such a bias. In
a recent paper, Danishefsky and Li describe results that
Me
N
PAN
HN
[ ? ]
[ ? ]
HN
Me
1
O
O
N
NH
Me
Me
HO
O
O
2
3
[ ? ]
[ ?]
[ ?]
Me
PAN
PAN
O
N
N
O
O
HN
Me-N
O
N
Me
N
H
OH
OH
PAN = p-anisyl
5
4
Figure 2. Possible isomerization pathways for 1.
‘invite caution’ regarding the structure of 1.3 In particu-
lar, the Columbia researchers determined that the imino
N-BOC derivative of 2 shows no apparent inclination to
cyclize to an orthoamide. This prompts us to disclose
observations of our own that (i) are in complete accord
with the foregoing, (ii) reveal that the precursor to 5 also
exists exclusively as the open form, and (iii) cast doubt
on the viability of the species depicted as 2.
The study described herein ignores issues of absolute ste-
reocontrol, focusing exclusively on structural questions.
Consonant with principles expounded in Ref. 3, our ap-
proach also centered on a Wong-type4 oxidative spiro-
cyclization of 6 and in situ hydration of the presumed
intermediate 8, leading directly to ( )-1 (Fig. 3). Thus,
deprotonation of tyrosine derivative 95 and alkylation
of the corresponding enolate with bromide 166 furnished
10. Cleavage of the BOC group set the stage for N-gua-
nidylation with reagent 17.7 Unexpectedly, this step pro-
ceeded with concomitant cleavage of the methyl ester,
providing acid 12, arguably due to inadvertent introduc-
tion of moisture.
OMe
Me
N
O
HN
O
OH
N
Me
1
Figure 1. Proposed structure of spiroleucettadine.
*
Corresponding author. Tel.: +1 604 822 2419; fax: +1 604 822
8710; e-mail: ciufi@chem.ubc.ca
0040-4039/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.03.024

3600
J. J. Chang et al. / Tetrahedron Letters 47 (2006) 3599–3601
assignment of this material again rests on the presence
PAN
PAN
Me
Me
of strong HMBC correlations between the H atoms of
both N-methyl substituents and the guanidine carbon
(double-headed curved arrows, Scheme 1), but the ab-
sence of a correlation with the carbonyl carbon. Thus,
the imino-hydantoin ring had formed with the regio-
isomeric distribution of N-methyl groups relative to 1.
Compound 15 served as a precursor to the plausible
spiroleucettadine isomer 5.
DIB
[ ? ]
N
N
N
N
HN
HN
O
O
OH
O
O
Me
Me
6
7
PAN
Me
N
H2O
1
HN
O
N
Me
8
The preparation of the ‘correct’ isomer of the dimethyl
creatinine segment proceeded as delineated in Scheme
2. Thus, TFA treatment of 12 provided 18, which under-
went selective BOC protection at the imino nitrogen to
furnish 19. This intermediate was N-methylated to give
20, the structural assignment of which again rests upon
the diagnostic HMBC correlations indicated with dou-
ble-headed curved arrows. Full deblocking surrendered
compound 22.
Figure 3. Strategy for the synthesis of 1.
However, ester saponification was inconsequential, in
that subsequent treatment of 12 with excess NaH and
MeI installed a requisite N-methyl substituent and re-
formed the methyl ester (cf. 13, Scheme 1). We note that
attempts to form 13 directly through reaction of 11 with
the N-methyl analog of 178 were unsuccessful, necessi-
tating the implementation of the present sequence.
Exposure of 13 to KOH induced ester saponification
and release of the sulfonyl group, as well as selective
cleavage of the BOC group on the imino-type N atom
of the guanidine. The structure of the emerging 14 rests
upon an HMBC correlation between the H atoms of an
N-methyl group and the carbonyl carbon of the surviv-
ing BOC unit. Treatment of 14 with TFA induced loss of
the BOC group and cyclization to 15. The structural
Consistent with the observations of Danishefsky and Li,
oxidative attack of 22 with PhI(OAc)2 (‘DIB’) in hexa-
fluoroisopropanol (‘HFIP’) furnished an intractable
mixture of products. The same outcome obtained upon
analogous treatment of 21 or of 15. However, oxidation
of 21 with DIB in aqueous HFIP furnished 23 in a low
8% yield after chromatography (Scheme 3). This mate-
rial was reasonably stable and it was purified to homo-
geneity. Consonant with Ref. 3, 23 existed in solution
exclusively as the depicted tautomer. The 13C spectrum9
of 21 exhibited a resonance at 174.4 ppm attributable to
the carbonyl group of the creatinine segment. This sig-
nal appeared at 175.7 ppm in the 13C spectrum of 23
and it was accompanied by a new carbonyl resonance
arising from the dienone at 187.3 ppm.10 Thus, 23 incor-
porates an intact creatinine carbonyl, that is, it does not
exists as a cyclic orthoamide. Indeed, no signals were
apparent near 102 ppm (the chemical shift of the orthoa-
mide carbon in 1): the 13C spectrum was blank between
113.9 and 79.9 ppm.
OMe
COOMe
a
c
Me
N
N-BOC
MeO
Me
G
9
MeO
O
OSO2Ph
10 G = BOC
11 G = H
b
OMe
OMe
Me
N
t-Bu-O
O
N
e
Me
N
BOC
N
BOC–N
R
Me
O
O
R
OSO2Ph
OMe
HN
O
OH
OH
12 R = H
13 R = Me
Me
N
d
14
a
c
OMe
12
Z
N
N
H
O
OSO2Ph
TFA
= HMBC correlations
Me
f
18 Z = H
H
N
b
19 Z = BOC
N
OMe
OMe
Me
N
O
OH
H
15
Me
Me
N
TFA
N
e
H2N
N
BOC
d
N
Br
SMe
N
Me
O
OH
O
O-R
Me
OSO2Ph BOC-N
NH-BOC
22
17
20 R = SO2Ph
21 R = H
16
= HMBC correlations
Scheme 1. Reagents and conditions: (a) LDA, THF, À78 °C, 1 h, then
16, À78 °C, 3 h, 59% (chrom.); (b) 2:1 CH2Cl2–TFA, 0 °C, 10 min,
73% (chrom.); (c) 17, Et3N, DMF, HgCl2, rt, 12 h, 75% (chrom.); (d)
NaH, MeI, DMF, 0 °C to rt, 12 h, 59%; (e) aq KOH, MeOH, reflux,
1 h, 88%; (f) neat TFA, rt, 1 h, 64% (chrom.).
Scheme 2. Reagents and conditions: (a) TFA, rt, 15 min, 98%; (b)
BOC2O, NaHCO3, aq dioxane, rt, 12 h, 82% (chrom.); (c) MeI,
K2CO3, DMF, rt, 15 min, 99%; (d) aq KOH, MeOH, reflux, 1.5 h,
88%; (e) neat TFA, rt, 15 min, 95%.

J. J. Chang et al. / Tetrahedron Letters 47 (2006) 3599–3601
3601
MeO
MeO
Michael-type polymerization, perhaps accounting for
the difficulties we encountered in its isolation.
a
a
21
15
In conclusion, our results reinforce the cautionary note
of Ref. 3 regarding the stability of the orthoamide func-
tion of 1, and by inference, the proposed constitution of
the natural product.
Me
Me
OH
OH
N
N
BOC-N
Me-N
N
N
H
O
O
O
O
Me
23
24
a
Acknowledgements
22
1 or 2
We thank the University of British Columbia, the Can-
ada Research Chair Program, NSERC and Merck
Frosst Canada for support of our research program.
Scheme 3. Reagents and conditions: (a) PhI(OAc)2, aq (CF3)2CHOH,
rt, 15 min, 8% (chrom.) for 23; 95% (chrom.) for 24.
Encouraged by the successful formation of a hydroxy
dienone in mixed organic–aqueous media, we subjected
22 to the action of DIB in HFIP/water, in the hope of
reaching 1—or at least its open tautomer 2. This resulted
in formation of a complex mixture of products. An ESI
mass spectrum of this crude mixture exhibited a signal at
m/z = 370, corresponding to the protonated form of 1 or
2. Unfortunately, all attempts to retrieve the presumed
product met with failure.11
Supplementary data
Supplementary data (experimental procedures and spec-
tral data for all compounds) associated with this article
can be found, in the online version, at doi:10.1016/
j.tetlet.2006.03.024.
References and notes
Spectral data provided as supporting information in
Ref. 1 do not seem entirely inconsistent with alternative
structure 5 for spiroleucettadine. In an effort to produce
5, compound 15 was oxidized with DIB in aqueous
HFIP. Contrary to previous cases, conversion to di-
enone 24 occurred efficiently (95%). The NMR spectra
of 24 were similar, but by no means identical, to those
of 1. Furthermore, 13C NMR spectroscopy once again
ruled out the presence of cyclic orthoamide tautomers.
In fact, the spectrum of 15 exhibited a signal at 184.6
(creatinine C@O) whereas 24 displayed resonances
at 189.6 (dienone C@O) and 187.6 (intact creatinine
C@O).
1. Ralifo, P.; Crews, P. J. Org. Chem. 2004, 69, 9025.
2. Calculations were carried out with the HyperchemÒ
package.
3. Li, C.; Danishefsky, S. J. Tetrahedron Lett. 2006, 47, 385.
4. Wong, Y.-S. Chem. Commun. 2002, 686.
5. Belagali, S. L.; Thankamma, M.; Himaja, M. Indian J.
Chem. 1995, 34B, 45.
6. Prepared by NBS bromination of the corresponding
toluene (74%). See Supplementary data for details.
7. Cf. Powell, D. A.; Ramsden, P. D.; Batey, R. A. J. Org.
Chem. 2002, 68, 2300.
8. Obtained by N-methylation (NaH/MeI, THF) of 17.
9. All NMR spectra discussed herein were recorded in
MeOH-d4 for the purpose of comparison with those of
1, which were recorded in the same solvent.
10. Our NMR data for 23 are identical to those of Ref. 3.
11. Compound 2 may be prepared in 21% yield by reaction of
22 with PhI(OCOCF3)2 in moist MeCN (Ref. 3).
12. Cf. the N-unprotected form of compds. 6–7 in: Canesi, S.;
Bouchu, D.; Ciufolini, M. A. Angew. Chem., Int. Ed. 2004,
43, 4336.
On a speculative note, the striking contrast between the
stability of 24 and the apparent instability of the pre-
sumed hydroxydienone arising upon oxidation of 22
raise questions about the viability of intermediates of
the type 2. Analogy with the behavior of an N-unpro-
tected dienone intermediate in our synthesis of cylindri-
cines12 intimates that 2 is readily predisposed to