Applications and limitations of the I2-mediated carbamate annulation for the synthesis of piperidines: Five-versus six-membered ring formation

Article abstract of DOI:10.1021/jo401512h

A protecting-group-free synthetic strategy for the synthesis of piperidines has been explored. Key in the synthesis is an I₂-mediated carbamate annulation, which allows for the cyclization of hydroxy-substituted alkenylamines into piperidines, pyrrolidines, and furans. In this work, four chiral scaffolds were compared and contrasted, and it was observed that with both d-galactose and 2-deoxy-d-galactose as starting materials, the transformations into the piperidines 1-deoxygalactonorjirimycin (DGJ) and 4-epi-fagomine, respectively, could be achieved in few steps and good overall yields. When d-glucose was used as a starting material, only the furan product was formed, whereas the use of 2-deoxy-d-glucose resulted in reduced chemo- and stereoselectivity and the formation of four products. A mechanistic explanation for the formation of each annulation product could be provided, which has improved our understanding of the scope and limitations of the carbamate annulation for piperidine synthesis.

Full text of DOI:10.1021/jo401512h

Article
pubs.acs.org/joc
Applications and Limitations of the I2‑Mediated Carbamate
Annulation for the Synthesis of Piperidines: Five- versus Six-
Membered Ring Formation
Hilary M. Corkran,†,‡ Stefan Munneke,† Emma M. Dangerfield,†,‡ Bridget L. Stocker,*,†,‡
and Mattie S. M. Timmer*,†
†School of Chemical and Physical Sciences, Victoria University of Wellington, P.O. Box 600, Wellington, New Zealand
‡Malaghan Institute of Medical Research, P.O. Box 7060, Wellington, New Zealand
S
* Supporting Information
ABSTRACT: A protecting-group-free synthetic strategy for
the synthesis of piperidines has been explored. Key in the
synthesis is an I2-mediated carbamate annulation, which allows
for the cyclization of hydroxy-substituted alkenylamines into
piperidines, pyrrolidines, and furans. In this work, four chiral
scaffolds were compared and contrasted, and it was observed
that with both D-galactose and 2-deoxy-D-galactose as starting materials, the transformations into the piperidines 1-
deoxygalactonorjirimycin (DGJ) and 4-epi-fagomine, respectively, could be achieved in few steps and good overall yields. When
D-glucose was used as a starting material, only the furan product was formed, whereas the use of 2-deoxy-D-glucose resulted in
reduced chemo- and stereoselectivity and the formation of four products. A mechanistic explanation for the formation of each
annulation product could be provided, which has improved our understanding of the scope and limitations of the carbamate
annulation for piperidine synthesis.
mic effect in streptozotocin-induced diabetic mice,7 while
fagomine isomer 4 inhibits lysosomal α-galactosidase A activity
INTRODUCTION
■
Iminosugars have enormous therapeutic potential in the
in Fabry lymphoblasts.8
treatment of a number of diseases, such as cancer, diabetes,
Given the interesting biological activity of piperidines, we
and lysosomal storage disorders.1 Notable members of the
became interested in extending our previously reported
piperidine iminosuguar family include N-hydroxyethyl-DNJ
protecting-group-free strategy for the synthesis of pyrroli-
(Miglitol, 1), which has been approved for the treatment of
dines9−11 to the synthesis of piperidines.12 To this end, we
non-insulin-dependent diabetes,2 and 1-deoxygalactonojirimy-
proposed that the target piperidines I could be attained
cin (DGJ, 2), recently used in a phase-2 clinical trial for the
following the hydrolysis of the intermediate cyclic carbamate II,
treatment of Fabry’s disease (Figure 1).3 The corresponding 2-
itself prepared via the I2-mediated cyclization of alkenylamine
deoxypiperidines, fagomine (3)4,5 and the unnatural 4-epi-
precursor III (Scheme 1). Alkenylamine III would in turn be
isomer 4, also exhibit promising biological activities. Fagomine
prepared from methyl iodoglycoside IV via our recently
was found to have activity against mammalian gut α-
reported protecting-group-free Vasella reductive amination
glucosidases and β-galactosidase6 and a potent antihyperglyce-
methodology.13 Key in this approach was the provision that
when R1 = OH, alkenylamine III should cyclize to form the
desired piperidine (N-cyclization) rather than undergoing O-
cyclization, which would lead to the undesired amino-
methylfuran.
To explore the potential of our annulation methodology for
the synthesis of piperidines, we first set out to prepare 1-
deoxygalactonojirimycin (DGJ, 2). As previously described,14
alkenylamine 5, itself available in three steps from D-galactose,
was subjected to an I2-mediated carbamate annulation to give,
under optimized conditions, piperidine isomers 6a and 6b and
furan 6c in a 3:1:1 ratio, respectively (Scheme 2). Whereas our
previously reported pyrrolidine annulation methodology
Received: July 12, 2013
Figure 1. Piperidines.
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jo401512h | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 1. Retrosynthetic Analysis for the Synthesis of Piperidines
amine 10, again in excellent yield (93%) with only minor traces
of the secondary amine (<5%).15 Alkenylamine 10 was then
treated with I2 and NaHCO3 in anticipation that carbamate 11
would be formed. Though TLC analysis revealed the smooth
transformation into a higher-running product (Rf = 0.23,
DCM/EtOH/MeOH/30% aq. NH3, 5/2/2/1 v/v/v/v), 13C
NMR analysis of the crude reaction mixture revealed the
presence of an iodomethylene carbon (−0.3 ppm, −CH2I),
thus indicating that the desired carbamate had not been
formed. Furthermore, an aminomethylene signal at 39.4 ppm
and four oxymethine signals (82−76 ppm) were observed,
which indicated the presence of aminomethylfuran 13 rather
than piperidine 12, both of which satisfy the HRMS result ([M
+ H]+ = 273.9938). Attempts to purify the material to allow for
further characterization proved futile, as Dowex-H+ or silica gel
chromatography led to degradation of the product. Full
characterization was therefore achieved after acetylation of
Scheme 2. Synthesis of DGJ (2)
1
the crude product with acetic anhydride and pyridine. The H
proceeded in 18 h at room temperature and with >95:5
diastereoselectivity in favor of the cis product,9−11 cyclization to
the piperidine framework was much slower and resulted in
lower stereoselectivity. Hydrolysis of the carbamate and
purification by silica gel flash column chromatography,
however, allowed 2 to be prepared in a respectable 40% yield
(from 5) and in five steps in total. Given this, we were keen to
explore whether a similar methodology could be used for the
preparation of other piperidines. The results of our studies in
this area are reported herein.
NMR spectral data clearly revealed an NH proton at 5.84 ppm
with a COSY correlation to both H1a and H1b and no
correlation to H5. In addition, the HMBC correlations between
both H1 protons and the NHAc carbonyl were also present. As
acetylation of pyran 12 would protect the amine, leaving no
observable NH proton, these observations allowed the product
of the acetylation reaction to be determined as furan 14.
Following on from this, the product of the I2-mediated
annulation reaction could then be assigned as furan 13.
In an attempt to encourage attack of the nitrogen atom and
formation of the desired carbamate 11 during the annulation
reaction, alkenylamine 10 was treated with a stronger base,
Na2CO3, to favor formation of the free amine. After the
reaction mixture was stirred at room temperature overnight, no
starting material remained; concentration and NMR analysis of
the crude reaction mixture, however, revealed that the carbamic
acid derivative of furan 13, formed via CO2 absorption onto the
amine,16,17 had been prepared. Numerous other changes were
RESULTS AND DISCUSSION
■
Following our successful synthesis of DGJ, we then explored
the potential of our I2-mediated annulation strategy for the
synthesis of 1,5-dideoxy-1,5-imino-L-iditol (7) (Scheme 3). To
this end, methyl glucopyranoside 8 was subjected to
triphenylphosphine, imidazole, and iodine to give the
corresponding iodosugar 9 in 98% yield.13 Subsequent Vasella
reductive amination led to the smooth formation of alkenyl-
Scheme 3. Formation of an Undesired Iodofuranoside by I2-Mediated Annulation
B
dx.doi.org/10.1021/jo401512h | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 4. Formation of Linear Alkenylamine 24
then made to the reaction conditions, including changes to the
temperature (0 °C to reflux) and the reagent concentrations.
However, under all of the conditions attempted, there was no
observable formation of the desired carbamate 11 as
determined by 1H NMR analysis of the crude reaction mixture,
and furan products always appeared to dominate. Though this
result was disappointing, the inherent problem of controlling
the chemoselectivity during protecting-group-free syntheses is
well-known and thus explains why the development of
protecting groups, though undesirable in terms of atom
economy, still remains an important objective for the synthetic
chemist.18
Having observed competing O-cyclization in an attempt to
prepare 1,5-dideoxy-1,5-imino-L-iditol (7) and to a lesser extent
DGJ (2), we thus proposed the synthesis of members of the
fagomine family. Fagomine analogues retain interesting bio-
logical properties but are devoid of the hydroxyl functionality at
C2 (R1 = H; Scheme 1), and hence, the synthesis would be
more readily accomplished. Accordingly, we first set out to
synthesize 4-epi-fagomine (4) (Scheme 4). To this end, 2-
deoxy-D-galactose (17) was first subjected to a Fischer
glycosylation using AcCl and MeOH to give the corresponding
methyl glycoside 18 in modest (40%) yield, with a second
major product being the kinetically favored furan. Methyl
glycoside 18 was then treated with I2, triphenylphosphine, and
imidazole in THF. Under these conditions, however, only
bicyclic pyranoside 19 was observed,19 a consequence of
intramolecular substitution of the activated phosphonium
group at C6 by the hydroxyl at the 3-position. Indeed, a
similar observation was seen en route to the synthesis of DGJ,14
and given the tendency for this intramolecular cyclization in
iodination of methyl glycosides of the galacto configuration, we
resorted to the installation of an isopropylidine protecting
group at the 3- and 4-positions via the treatment of 2-deoxy-D-
galactose (17) with MeOH, acetone, dimethoxypropane, and
AcCl. Conveniently, these conditions led to the concomitant
installation of the methyl group at the anomeric position to give
galactoside 20 in one step. Galactoside 20 was then subjected
to the standard iodination conditions to give the required
primary iodide 21 in 71% yield (from 17), and the
isopropylidine group was then removed to give the desired
methyl iodoglycoside 22 in 94% yield. Here it is also interesting
to note that 20 can be prepared from D-galactal (23) by
protonation of the alkene with TsOH followed by nucleophilic
attack of methanol.20 Under these conditions, pyranose 20 was
formed in ∼78% yield with small amounts (ca. 10%) of methyl
5,6-O-isopropylidene-D-galactofuranoside, which was more
readily separable following conversion of 20 into iodide 21.
With primary iodide 22 in hand, we then applied the Vasella
reductive amination conditions and obtained alkenylamine 24
in excellent yield (99%). Alkenylamine 24 was then subjected
to a solution of I2 and NaHCO3 in H2O to affect the desired
carbamate annulation reaction (Table 1). Again, cyclization to
the desired six-membered ring proved to be sluggish, and no
appreciable reaction was observed after alkenylamine 24 was
stirred in NaHCO3 (sat. aq.) in the presence of excess I2 (10
equiv) at room temperature (ca. 20 °C) for 7 days (entry 1).
When the solution was heated to 50 °C, however, complete
conversion of the starting material was observed after 3 days
(entry 2), and following filtration, carbamates 25a and 25b
1
were observed in a 3:1 ratio, as determined by H NMR
analysis of the crude reaction mixture. Lowering the reaction
temperature to 40 °C resulted in an increased reaction time (5
days) but led to improved diastereoselectivity in favor of the cis
isomer 25a (25a:25b = 9:1; entry 3). Remarkably, a
diastereoselectivity of >95:5 was achieved when the reaction
was performed at 35 °C with a reaction time of 5 days (entry
4). Indeed, using the latter conditions, carbamate 25a was
C
dx.doi.org/10.1021/jo401512h | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
(27a) was converted to the methyl glycoside (27b) under
Fischer glycosylation conditions, and subsequent treatment
with iodine, triphenylphosphine, and imidazole at 75 °C for 1.5
h gave primary iodide 28 in good overall yield (64% for two
steps).26,27 The reaction times and temperatures were critical to
avoid overiodination or incomplete consumption of starting
material. Iodide 28 was then treated with Zn, NaCNBH3, NH3,
and NH4OAc according to the standard Vasella reductive
amination protocol to effect its smooth transformation into
alkenylamine 29 in good yield (86%).
Table 1. Diastereoselectivity of Carbamate Annulation
a
entry
T (°C)
t (days)
25a:25b
1
2
3
4
20
50
40
>7
3
no reaction
3:1
9:1
Alkenylamine 29 was then subjected to a variety of
annulation conditions (Table 2). As previously highlighted,
slightly warmer temperatures were required for the annulation
of alkenylamine 24, and because of this, alkenylamine 29 was
first treated with I2 and NaHCO3 (sat. aq.) at 40 °C (entry 1).
These conditions led to a complex mixture of four carbamate
products later identified as pyrrolidine 30, the two piperidine
diastereomers 31a and 31b, and iodide 32 (vide infra), with
iodide 32 being the major product in this instance. In an
attempt to limit the formation of so many carbamate products,
the reaction temperature was then lowered and the number of
equivalents of NaHCO3 was reduced (entry 2). After the
reaction mixture was stirred at room temperature for 9 days,
however, no product formation was observed. Next, iodine was
added in portions over the course of the reaction while the
reaction mixture was stirred at room temperature until all of the
starting material was consumed, as gauged by TLC analysis
(entry 3). While these conditions invoked annulation reactions
and a slight improvement in the ratio of the piperidine
carbamate 31a was observed, iodide 32 was once again the
major product and all four carbamates were still formed. The
reaction was then attempted at room temperature, again using a
saturated solution of NaHCO3 and I2 (5 equiv) added in one
portion at the start of the reaction (entry 4). Under these
conditions, the formation of iodide 32 was reduced and
pyrrolidine 30 was formed with modest selectivity in a
30:31a:31b:32 ratio of 4:2:2:3. The role of the halide in
influencing the ratio of products was then investigated through
the use of Br2 (entry 5); however, this proved futile, with
degradation being observed. Further changes were then made
to the number of equivalents of reagents and the temperature at
which the reaction was performed, but in all instances complex
mixtures of products were observed, and the best yield of any
single product apart from iodide 32 was still found to be that of
pyrrolidine 30 (as highlighted in entry 4). Although the yield
was disappointing (10−30%), pyrrolidine 30 could nevertheless
be isolated as a pure compound following repeated flash silica
gel column chromatography (eluting in 2% MeOH in EtOAc)
and reversed-phase column chromatography (C8, H2O). Here
the range of yields reflects the difficulty of this chromatographic
separation. Piperidines 31a and 31b, however, could not be
separated from one another and were isolated in a poor
combined yield (ca. 20%). Attempts to separate the products at
a later stage in the synthesis (i.e., following hydrolysis) also did
not lead to any overall improvement in the isolated yields.
To confirm the structures of pyrrolidine 30 and alkyl iodide
32, extensive 1D and 2D NMR spectral analysis was
undertaken. For pyrrolidine 30, an HMBC correlation was
observed between H4 and C1, thus enabling the 5,6-bicyclic
ring system of 30 to be established. Unfortunately, overlapping
signals in the proton spectra meant that the configurations of
the stereocenters could not be determined unequivocally.
However, NMR analysis of the acetylated derivative 33 proved
5
b
35
5
>95:5
a
1
As determined by H NMR analysis of the crude reaction mixture.
b
25a was isolated in 86% yield.
isolated in 86% yield following silica gel column chromatog-
raphy.
To complete the synthesis, carbamate 25a was then
hydrolyzed via treatment with NaOH in EtOH at reflux
(Scheme 5). This resulted in the smooth formation of 4-epi-
Scheme 5. Hydrolysis of cis-Carbamate 25a
fagomine (4) in 92% yield, thus completing this remarkably
efficient six-step synthesis with an overall yield of 52%. 4-epi-
Fagomine has been prepared on several other occasions.21−24
The first synthesis required 10 steps from 321 using the method
of Heiker and Schueller22 (no yield reported), and the synthesis
with the best reported yield (28%) proceeded via an
asymmetric route from a benzyl-protected galactal.25 Although
the need for a protecting group here was unavoidable because
of the inherent reactivity of the galacto configuration, protecting
group manipulation added just one step to the synthesis, and an
expedient synthesis of 4 was achieved with the best yield to
date.
To further demonstrate the potential of our strategy for the
synthesis of fagomine diastereomers, we then set about to
synthesize 5-epi-fagomine (26) (Scheme 6). 2-Deoxy-D-glucose
Scheme 6. Synthesis of Alkenylamine 29 Derived from 2-
Deoxy-D-glucose
D
dx.doi.org/10.1021/jo401512h | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Table 2. Carbamate Annulation Reaction for Alkenylamine 29
a
entry
conditions
products (30:31a:31b:32)
1
2
3
4
5
I2 (5 equiv), NaHCO3 (sat.), 40 °C, 18 h
3:2:2:8
I2 (5 equiv), NaHCO3 (1.5 equiv), rt, 9 days
I2 (3 × 1.5 equiv added over 22 days), NaHCO3 (sat.), rt
I2 (5 equiv), NaHCO3 (sat.), rt, 5 days
starting material
3:2:1:7
b
4 :2:2:3
Br2 (5 equiv), NaHCO3 (sat.), rt, 3 days
decomposition
a
b
1
As determined by H NMR analysis of the crude reaction mixture 30 was isolated as a pure compound in 10−30% yield following column
chromatography
to be more beneficial. The relative configurations of H5, H4,
and H6b were apparent through two large trans-diaxial
couplings (J4,5 = J5,6b = 9 Hz), and the coupling constant
between H5 and H6a (J5,6a = 4.9 Hz) indicated an axial/
equatorial relationship between these two protons. As the
stereochemistry at C3 is known, NOEs were then used to
determine the stereochemistries of the remaining centers.
NOEs were observed between H5 and H3, H3 and H2a, H2a
and H1a, and H6a and H5 (Scheme 7), indicating these
reported,24,29 once again with some variation due to slight
differences in pH.30
Scheme 8. Hydrolysis of Piperidine Carbamates To Give
Fagomine (3) and 5-epi-Fagomine (26)
Scheme 7. Synthesis of Previously Undisclosed
Trihydroxylated Pyrrolidine 34 and NOE Correlations for
Acetylated Carbamate 33
Confirmation of the structure of iodide 32 proved to be more
difficult. No HMBC correlation was observed between the
carbonyl and H3, and accordingly, the acetylated adduct 35
(Scheme 9) was prepared to aid with structural elucidation.
Indeed, upon acetylation, downfield shifts of the resonances for
protons H4 (5.34 ppm) and H5 (5.14 ppm), which were
previously at ca. 3.76 ppm in the starting material, were
observed. Proton H3, which was already at a relatively high
chemical shift, did not move significantly upon acetylation. An
HMBC correlation between H1 and the carbonyl was observed,
though unfortunately, again no HMBC correlation was
observed between the carbamate carbonyl and H3. As the
stereochemistry of compound 32 also needed to be confirmed,
we proposed that the synthesis of the corresponding alkenyl-
amine (Scheme 9) would allow for confirmation of the
structure of 32, including the chirality of all stereogenic
centers. To this end, iodide 32 was first converted to the
corresponding acetonide 36, and on the basis of a slightly
protons to be on the bottom face of the molecule, while NOEs
between H2b and H4 and between H4 and H6b identified
these protons to be on the opposite face, thereby allowing for
the assignment of the L-allo stereochemistry. Pyrrolidine 30 was
then treated with NaOH in EtOH, which resulted in the
smooth hydrolysis of the carbamate and allowed for the
synthesis of the previously undisclosed trihydroxylated
pyrrolidine 34 in 5−15% yield over five steps. The enantiomer
of 34 has been prepared once previously.28 Comparison of
NMR data showed minor differences in the chemical shifts,
which can be explained by the use of a different solvent and
differences in pH (34 was treated with conc. DCl to ensure that
the amine was fully protonated). The optical rotation of 34 was
measured as [α]2D6 = −1.8 (c = 0.41, MeOH), while that of the
enantiomer was of the opposite sign and slightly larger in
magnitude, reported as [α]2D2 = +4.7 (c = 4.0, MeOH), which
can be explained by the differences in concentration and
protonation state of the two compounds. At this stage, the
piperidine carbamates 31a and 31b were also hydrolyzed to
give 5-epi-fagomine (26) and fagomine (3), which were
separated following flash silica gel column chromatography
(Scheme 8). The spectral data were similar to those previously
1
smaller Δδ of the acetonide methyl group proton signals (Δδ H
= 0.039 ppm)31 and diagnostic δ C ≈ 27 ppm for both methyl
13
groups (δ C = 27.0 ppm and δ C = 27.6 ppm),32,33 this allowed
for the relative stereochemistries of the 4- and 5-substituents to
be determined as trans in the acetonide five-membered ring,
indicating a syn relationship in iodide 32. Opening of the
acetonide via oxidative addition of zinc and subsequent
reductive ring opening then provided the truncated carbamate
37 in good yield. Finally, hydrolysis of the carbamate in 37 via
treatment with NaOH under reflux for 4 h gave the desired
13
13
1
alkenylamine 29. Comparison of the H NMR and 13C NMR
spectra of the synthesized alkenylamine with those of (+)-29
revealed them to be identical, and the optical rotations were of
the same sign and comparable magnitude {[α]2D7 = +37.5 (c =
1.0, MeOH) and [α]2D7 = +31.8 (c = 0.2, MeOH)}, thus
E
dx.doi.org/10.1021/jo401512h | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 9. Determination of the Structure of Iodocarbamate 32 via Transformation into Alkenylamine (+)-29
Figure 2. Six-membered iodocyclization transition states for 2-deoxy sugar series.
establishing the structure of the synthesized alkenylamine to be
that of the starting material, (+)-29. From this information and
the previously deduced trans relationship in acetonide 36, the
starting iodocarbamate 32 was thus determined to be
(3R,4R,5R), as depicted. To confirm our methodology,
alkenylamine (+)-29 was also synthesized from the acetylated
adduct 35 by way of reductive elimination of 5-OAc (→38)
and subsequent carbamate hydrolysis.
Taken as a whole, the range of products formed en route to
the synthesis of the target piperidines highlights the inherent
problems of chemoselectivity in protecting-group-free chem-
istry. While we have previously illustrated that the protecting-
group-free synthesis of pyrrolidines proceeds relatively
smoothly with good to excellent cis selectivity for the carbamate
annulation,34 the present results illustrate that this method-
ology does not always translate readily to the synthesis of
piperidines. The first factor to be considered when applying the
carbamate annulation methodology to the synthesis of
piperidines is O- versus N-cyclization. As illustrated when D-
galactose was used as the starting material en route to the
synthesis of DGJ,14 it is possible to favor formation of the
desired pyran. This observation is remarkable in its own right
given the general preference for five- versus six-membered ring
formation. Indeed, when D-glucose was used as the starting
material, only the furan was observed.
To eliminate formation of the furan as a competing reaction,
the synthesis of piperidines devoid of the hydroxyl functionality
at C2 can then be attempted using the corresponding 2-deoxy
sugars as starting materials. Here, the chirality of the
alkenylamine is crucial in determining the number of products
formed during the annulation reaction. While the alkenylamine
derived from 2-deoxy-D-galactose was transformed into the 4-
epi-fagomine carbamate scaffold with excellent chemo- and
diastereoselectivity (cis:trans > 95:5), subjection of the
corresponding alkenylamine derived from 2-deoxy-D-glucose
to the annulation conditions resulted in poor selectivity and the
formation of four products despite much effort to optimize this
reaction to favor piperidine formation. The difference in
selectivity for the two annulation reactions can be explained by
considering the corresponding six-membered iodocyclization
transition states, wherein, according to the seminal work of
Chamberlin and co-workers,35 the allylic hydroxyl has a
stereodirecting effect in the formation of both five- and six-
membered rings. Here, Chamberlin proposed that the
unfavorable C−O σ* to CC π orbital overlap can be
minimized by an OH-in-plane (with the double bond)
transition state. In the case of our I2-mediated carbamate
annulations, this is illustrated in Figure 2, wherein transition
state I for the 2-deoxy-D-galactose series depicts the major OH-
in-plane conformer while II depicts the unfavorable H-in-plane
conformer. Similarly, the OH-in-plane transition state III and
the H-in-plane transition state IV are also given for the 2-
deoxy-D-glucose series.
In addition to the conformation of the allylic hydroxyl, for
six-membered rings the relative axial or equatorial orientation
of the hydroxyl substituents needs to be considered. For the 2-
deoxy-D-galactose series, both the OH-in-plane transition state
I and the H-in-plane transition state II have one axial and one
equatorial hydroxyl group, and as a consequence, the OH-in-
plane transition state is lower in energy because there is no
unfavorable C−O σ* to CC π orbital overlap. For the 2-
deoxy-D-glucose series however, the OH-in-plane transition
state III has the two hydroxyl substituents placed in an
unfavorable axial orientation, thereby increasing the overall
energy of this transition state and making it closer in energy to
transition state IV, which, while having the unfavorable H-in-
plane conformation, has the hydroxyl substituents in a more
favorable equatorial orientation. In summary, the 2-deoxy-D-
galactose series thus leads to the preferential formation of the
cis-piperidine carbamate, while the 2-deoxy-D-glucose series,
F
dx.doi.org/10.1021/jo401512h | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Scheme 10. Proposed Mechanism for the Formation of Pyrrolidine 30, Piperidines 31a and 31b, and Iodide 32
with iodocyclization transition states of similar energy, leads to
a variety of products.
molecular or intermolecular nucleophilic ring opening. It is also
conceivable that formation of a cyclic carbamate at the 3-
position takes place prior to halohydrin formation43 and that
this provides an alternative route to 32. The possibility that 32
is formed via a Payne rearrangement44 of the intermediate
epoxide is unlikely, as this reaction would require strongly basic
conditions.45 Such a cyclic carbamate intermediate, however,
does not add to a conceivable mechanism for the formation of
either 31a or 30.
To explain the formation of pyrrolidine 30, a mechanism that
supports the inversion of the C4 and C5 chiral centers is
required, and thus an iodoepoxide (pathway C) is proposed as
a key intermediate.42,46 Nucleophilic 5-exo-tet ring opening of
the epoxide followed by the addition of CO2 then allows for the
formation of 30.47 Finally, the formation of trans-piperidine
31b can be envisioned to occur via SN2 attack of the
diastereomeric iodine complex (pathway D), with the iodine
complex itself arising from transition state IV (Figure 2).
Subsequent addition of CO2 to the intermediate cyclic
iodoamine then gives the piperidine. Here it should also be
noted that attack at the 5-position of the intermediate epoxide
discussed in pathway C (6-endo-trig) could also lead to 31b, but
this is less favored than the corresponding 5-exo-trig cyclization
(→31b),47 especially in the absence of a bulky group α to the
nitrogen.48
To gain further insight into how to better design starting
materials to favor the formation of the desired piperidine or
pyrrolidine, reaction mechanisms have been proposed to
explain the formation of the four products following the
subjection of alkenylamine 29 to the carbamate annulation
conditions (Scheme 10). First, attack of iodine from the top
face of alkenylamine 29 (via transition state III in Figure 2)
leads to an intermediate that can be transformed into piperidine
31a, iodide 32, and pyrrolidine 30. Here, pathway A involves
amine attack on the iodine complex at C5 followed by the
addition of CO2 to give cis-carbamate 31a in accordance with
our previously published annulation methodology.34 The
synthesis of iodocarbamate 32, which contains a chiral
backbone with a C5 stereocenter that differs from the parent
alkenylamine 29, is proposed to occur via SN2 attack of water
on the iodine complex (pathway B) followed by the addition of
CO2 to the linear iodoamine, which itself can be formed via the
dissociation of carbonic acid. While there is no literature
precedent for the reaction of terminal unsubstituted allylic
alcohols with iodine, a number of halohydrin reactions have
been performed on similar substrates with both the substitution
pattern and the accessibility of the halonium ion effecting
regio-36−38 and stereochemical39,40 outcomes. In the most
closely related system involving hydrobromination, a variety of
Taken together, a number of factors should thus be
considered before using the carbamate annulation reaction as
a key step in the synthesis of piperidines containing multiple
reactive sites. First, the orientations of the ring substituents
affect the relative energies of the OH-in-plane or H-in-plane
transition states, and where there is no obvious low-energy
conformer, this can lead to a variety of products. One could
thus envisage the use of judiciously placed bulky substituents to
favor one conformer over the other, while the positioning of a
products were observed.41,42 In our case, however, H NMR
1
analysis revealed no product that could have arisen from
opening of the iodine complex by water at the terminal carbon
or from attack on the diastereomeric iodine complex. The
observed regioselectivity can be explained by Markovnikov’s
rule, while the stereoselectivity is more difficult to rationalize
but is correlated to the relative reaction rates for formation of
the intermediate iodine complex and the subsequent intra-
G
dx.doi.org/10.1021/jo401512h | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Hz, 1H, H2), 3.47 (ddd, J1a,NH = 6.2, J1a,2 = 7.3, J1a,1b = 14.0 Hz, 1H,
H1a), 3.27 (td, J1b,NH = J1b,2 = 6.2, J1a,1b = 14.0 Hz, 1H, H1b), 3.21 (dd,
J5,6a = 6.3, J6a,6b = 9.7 Hz, 1H, H6a), 3.17 (dd, J5,6b = 8.8, J6a,6b = 9.7 Hz,
1H, H6b), 2.16 (s, 3H, Ac), 2.15 (s, 3H, Ac), 2.00 (s, 3H, NHAc); 13C
NMR (125 MHz, CDCl3) δ 170.5 (CO, NHAc), 170.1 (CO,
Ac), 169.4 (CO, Ac), 79.6 (C5), 78.9 (C2), 76.4 (C3), 76.1 (C4),
38.1 (C1), 23.5 (CH3, NHAc), 20.9 (CH3, Ac), 20.8 (CH3, Ac), −0.6
(C6); HRMS(ESI) m/z calcd for [C12H18O6NI + Na]+ 422.0076,
found 422.0076.
protecting group at the 4-position of the alkenylamine could be
used to inhibit formation of the intermediate epoxide and
hence the associated pyrrolidine products (e.g., 30). We have
previously shown that the carbamate annulation can occur in
the presence of protecting groups,49,50 which supports the
validity of installing one or more protecting groups to control
the chemoselectivity of the annulation when required.
CONCLUSION
Methyl 3,6-Anhydro-2-deoxy-α-D-galactopyranoside (19).
To a solution of methyl glycoside 18 (28 mg, 0.16 mmol) in dry
THF (3 mL) were added PPh3 (83 mg, 0.32 mmol), imidazole (43
mg, 0.63 mmol), and I2 (80 mg, 0.32 mmol). The mixture was refluxed
overnight (18 h) and then cooled and concentrated. The resulting oil
was purified by flash column chromatography (petroleum ether/
EtOAc, 5/1 to 1/1 v/v) to give bicyclic galactoside 19. Rf = 0.3
(EtOAc/MeOH, 98/2 v/v); 1H NMR (500 MHz, D2O) δ 4.69 (d, J1,2b
= 6.2 Hz, 1H, H1), 4.31−4.32 (m, 2H, H3 and H6a), 4.27−4.28 (m,
■
We have presented two highly efficient syntheses of the
biologically important piperidines DGJ (2) and 4-epi-fagomine
(4) in good overall yields. Key in these reactions is the use of an
I2-mediated carbamate annulation. The use of the carbamate
annulation for the synthesis of other piperidines, however, was
less successful, as either O-cyclization or a series of competing
intramolecular reactions occurred. To this end, trihydroxypyr-
rolidine 34 was prepared in five steps and modest overall yield
(up to 15%). These studies highlight the limitations of using
the carbamate annulation for the preparation of piperidines in
protecting-group-free syntheses, although the work also points
to the potential of this methodology for the synthesis of
interesting chiral scaffolds when either protected or suitably
functionalized alkenylamines are used as starting materials.
1H, H5), 4.02 (d, J3,4 = 1.8 Hz, 1H, H4), 3.98 (dd, J5,6b = 3.4, J6a,6b
=
9.9 Hz, 1H, H6b), 3.40 (s, 3H, OMe), 2.19 (dd, J2a,3 = 4.6, J2a,2b = 14.5
Hz, 1H, H2a), 2.02 (ddd, J2b,3 = 1.6, J1,2b = 6.2, J2a,2b = 14.5 Hz, 1H,
H2b); 13C NMR (125 MHz, D2O) δ 97.2 (C1), 77.8 (C3), 77.3 (C5),
74.8 (C4), 70.2 (C6), 55.7 (OMe), 37.6 (C2); HRMS(ESI) m/z calcd
for [C7H12O4 + Na]+ 183.0633, found 183.0631.
Methyl 2-Deoxy-3,4-O-isopropylidene-α/β-D-galactopyrano-
side (20). From 17. AcCl (180 μL) was added to a solution of 2-
deoxy-D-galactose (17) (200 mg, 0.61 mmol) in MeOH/acetone/
dimethoxypropane (6 mL, 2/2/2 v/v/v), and the solution was stirred
for 18 h at room temperature. The reaction was quenched with
Dowex−OH−, and the mixture was filtered and concentrated. The
resulting oil, 20, was used without purification.
From 23. To a solution of D-galactal 23 (200 mg, 1.36 mmol) in
MeOH (8 mL) and acetone (8 mL) was added pTsOH·H2O (24 mg,
0.14 mmol), and the reaction was stirred at room temperature for 18 h.
The reaction mixture was then neutralized by the addition of Dowex−
OH−, filtered, and concentrated in vacuo. The crude reaction product,
20, was used in the iodination reaction without further purification.
Rf = 0.13 (petroleum ether/EtOAc, 1/1 v/v); 1H NMR (500 MHz,
CDCl3) δ 4.86 (dd, J1,2a = 5.1, J1,2b = 6.0 Hz, 1H, H1), 4.47 (ddd, J2b,3
= 4.1, J2a,3 = 5.1, J3,4 = 6.9 Hz, 1H, H3), 4.16 (dd, J4,5 = 1.7, J3,4 = 6.9
Hz, 1H, H4), 3.89 (dd, J5,6a = 5.6, J6a,6b = 10.6 Hz, 1H, H6a), 3.80
(ddd, J4,5 = 1.7, J5,6b = 4.3, J5,6a = 5.6 Hz, 1H, H5), 3.77 (dd, J5,6b = 4.3,
J6a,6b = 10.6 Hz, 1H, H6b), 3.37 (s, 3H, OMe), 2.17 (dt, J2a,2b = 14.8,
EXPERIMENTAL SECTION
■
General Procedures. Unless otherwise stated, all reactions were
performed under atmospheric air. THF was distilled from LiAlH4 prior
to use. Pyridine was distilled over KOH prior to use. All chemicals
obtained from commercial suppliers were used without further
purification. Zn dust was activated by the careful addition of
concentrated H2SO4 followed by decantation, washing with EtOH
(3×) and hexanes (3×), and storage under dry hexanes. All solvents
were removed by evaporation under reduced pressure. Reactions were
monitored by TLC analysis on silica gel-coated plastic sheets with
detection by coating with 20% H2SO4 in EtOH followed by charring at
ca. 150 °C, by coating with a solution of ninhydrin in EtOH followed
by charring at ca. 150 °C, or by coating with a solution of 5% KMnO4
and 1% NaIO4 in H2O followed by heating. 1H and 13C chemical shifts
(δ) were internally referenced to the residual solvent peak. NMR peak
assignments are based on 2D NMR experiments (COSY, HSQC, and
HMBC). High-resolution electrospray ionization mass spectrometry
[HRMS(ESI)] was performed on a Q-TOF instrument.
J1,2a = J2a,3 = 5.1 Hz, 1H, H2a), 1.73 (ddd, J2b,3 = 4.1, J1,2b = 6.0, J2a,2b
=
14.8 Hz, 1H, H2b), 1.47 (s, iPr), 1.31 (s, iPr); 13C NMR (75 MHz,
CDCl3) 109.2 (iPr-C(CH3)2) 97.4 (C1), 73.4 (C4), 70.5 (C3), 68.5
(C5), 62.8 (C6), 55.0 (OMe), 30.8 (C2), 26.8, 25.3 (iPr-C(CH3)2);
HRMS(ESI) m/z calcd for [C10H18O5 + Na]+ 241.1052, found
241.1046.
1-Amino-2,5-anhydro-1,6-dideoxy-6-iodo-L-iditol (13). To a
solution of alkenylamine 1013 (310 mg, 1.69 mmol) in sat. aq.
NaHCO3 (7 mL) was added I2 (645 mg, 2.54 mmol), and the solution
was stirred for 7 days. The reaction mixture, a crude oil, was
concentrated and used without further purification. Rf = 0.23 (DCM/
EtOH/MeOH/30% aq. NH3, 5/2/2/1 v/v/v/v); 1H NMR (500 MHz,
D2O) δ 4.45 (dt, J4,5 = 3.1, J5,6a = J5,6b = 7.3 Hz, 1H, H5), 4.40 (ddd,
Methyl 2,6-Dideoxy-6-iodo-3,4-O-isopropylidene-α/β-D-gal-
actopyranoside (21). From 17. To a solution of 20 (0.61 mmol) in
dry THF (7 mL) were added PPh3 (456 mg, 1.74 mmol), imidazole
(180 mg, 2.62 mmol), and I2 (442 mg, 174 mmol). The mixture was
refluxed for 18 h and then cooled and concentrated. The syrup was
redissolved in ethyl acetate (20 mL), washed with sat. aq. Na2S2O3 (2
× 15 mL) and then brine (20 mL), dried over MgSO4, filtered, and
concentrated. The resulting oil was purified by flash column
chromatography (petroleum ether/EtOAc, 15/1 v/v) followed by
crystallization from hexanes to give iodogalactoside 21 as white crystals
(142 mg, 0.434 mmol, 71% from 17).
J2,3 = 3.8, J1a,2 = 4.4, J1b,2 = 5.6 Hz, 1H, H2), 4.38 (dd, J3,4 = 1.3, J2,3
=
3.8 Hz, 1H, H3), 4.28 (dd, J3,4 = 1.3, J4,5 = 3.1 Hz, 1H, H4), 3.27 (dd,
J5,6a = 7.3, J6a,6b = 9.8 Hz, 1H, H6a), 3.26 (dd, J5,6b = 7.3, J6a,6b = 9.8 Hz,
1H, H6b), 3.258 (dd, J1a,2 = 4.4, J1a,1b = 13.6 Hz, 1H, H1a), 3.23 (dd,
J1b,2 = 5.6, J1a,1b = 13.6 Hz, 1H, H1b); 13C NMR (125 MHz, D2O) δ
81.7 (C5), 77.5 (C3), 76.52, 76.50 (C2 and C4), 39.4 (C1), 0.6 (C6);
HRMS(ESI) m/z calcd for [C6H12O3NI + H]+ 273.9935, found
273.9938.
1-Acetamido-3,4-di-O-acetyl-2,5-anhydro-1,6-dideoxy-6-
iodo-L-iditol (14). A solution of iditol 13 was stirred overnight at rt in
a solution of acetic anhydride (4 mL) and pyridine (4 mL). The
reaction mixture was concentrated, dissolved in EtOAc, washed with
sat. aq. NaHCO3 and brine, dried over MgSO4, filtered, and
From 23. Iodination of 20 performed as above yielded iodogalacto-
side 21 as white crystals (241 mg, 0.73 mmol, 54% over two steps from
23).
Rf = 0.77 (petroleum ether/EtOAc, 1/1 v/v); Mp 91−92 °C; [α]D21
= +72.4 (c = 1.0, CHCl3); IR (film) 2982, 2936, 2912, 2833, 1439,
1
1373, 1246, 1214, 1186, 1088, 1047, 969, 880, 738 cm−1; H NMR
concentrated in vacuo to give acetylated iditol 14 as a colorless oil.
1
(500 MHz, CDCl3) δ 4.86 (dd, J1,2a = 5.4, J1,2b = 6.6 Hz, 1H, H1), 4.49
(ddd, J2b,3 = 3.7, J2a,3 = 4.6, J3,4 = 7.3 Hz, 1H, H3), 4.26 (dd, J4,5 = 2.0,
J3,4 = 7.3 Hz, 1H, H4), 3.86 (ddd, J5,4 = 2.0, J5,6a = 5.8, J5,6b = 8.1 Hz,
1H, H5), 3.47 (s, 3H, OMe), 3.34 (dd, J5,6a = 5.8, J6a,6b = 10.2 Hz, 1H,
Rf = 0.12 (EtOAc); H NMR (500 MHz, CDCl3) δ 5.84 (t, J1a,NH
=
J1b,NH = 6.2 Hz, 1H, NH), 5.38 (dd, J3,4 = 1.4, J4,5 = 4.0 Hz, 1H, H4),
5.27 (dd, J3,4 = 1.4, J2,3 = 3.4 Hz, 1H, H3), 4.52 (ddd, J4,5 = 4.0, J5,6a
6.3, J5,6b = 8.8 Hz, 1H, H5), 4.30 (ddd, J2,3 = 3.4, J1b,2 = 6.2, J1a,2 = 7.3
=
H
dx.doi.org/10.1021/jo401512h | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
H6a), 3.28 (dd, J5,6b = 8.1, J6a,6b = 10.2 Hz, 1H, H6b), 2.24 (ddd, J2a,3
4.6, J1,2a = 5.4, J2a,2b = 15.1 Hz, 1H, H2a), 1.64 (ddd, J2b,3 = 3.7, J1,2b
6.6, J2a,2b = 15.1 Hz, 1H, H2b), 1.48 (s, 3H, iPr), 1.34 (s, 3H, iPr); 13
NMR (75 MHz, CDCl3) 109.3 (iPr-C(CH3)2) 97.6 (C1), 73.4 (C4),
70.9 (C3), 70.0 (C5), 55.2 (OMe), 30.2 (C2), 26.6, 25.2 (iPr-
C(CH3)2), 3.2 (C6); HRMS(ESI) m/z calcd for [C10H17O4I + Na]+
351.0069, found 351.0072.
Methyl 2,6-Dideoxy-6-iodo-α/β-D-galactopyranoside (22).
To a solution of iodogalactoside 21 (237 mg, 0.79 mmol) in MeOH
(10 mL), at 0 °C was added AcCl (200 μL). The reaction mixture was
stirred at room temperature for 5 days, after which the reaction was
quenched with Dowex−OH− and the mixture was concentrated in
vacuo. The product was purified via gradient flash column
chromatography (petroleum ether/EtOAc, 3/1 to 1/1 v/v) to provide
galactoside 22 as a colorless oil (200 mg, 0.69 mmol, 94%). Rf = 0.11
(EtOAc); [α]2D1 = +0.11 (c = 1.0, MeOH); IR (film) 3292, 2925, 2854,
1735, 1457, 1209, 1130, 1034, 668 cm−1; 1H NMR (500 MHz,
=
=
C
1653, 1398, 1145, 1100, 1046, 1022 cm−1; 1H NMR (600 MHz, D2O)
δ 3.92 (bs, 1H, H4), 3.76 (ddd, J = 2.8, J = 8.4 Hz, J = 10.5 Hz, 1H,
H3), 3.69 (dd, J = 5.8, J = 11.7 Hz, 1H, H6a), 3.65 (dd, J = 7.6, J =
11.6 Hz, 1H, H6b), 3.17 (dt, J = 3.6, J = 12.7 Hz, 1H, H1a), 2.92 (app
t, J = 7.1 Hz, 1H, H5), 2.73 (m, 1H, H1b), 1.74−1.79 (m, 2H, H2a
and H2b); 13C NMR (150 MHz, D2O) δ 67.4 (C3), 65.8 (C4), 60.0
(C6), 59.4 (C5), 42.1 (C1), 24.0 (C2); HRMS(ESI) m/z calcd for
[C6H13NO3 + H]+ 148.0974, found 148.0974.
Methyl 2-Deoxy-α/β-D-glucopyranoside (27b). To a solution
of AcCl (240 μL) in MeOH (240 mL) was added 2-deoxy-D-glucose
(27a) (1200 mg, 7.3 mmol), and the mixture was stirred at room
temperature for 18 h. The reaction mixture was then neutralized by the
addition of Dowex−OH−, filtered, and concentrated. The resulting oil
was used without further purification. Methyl 2-deoxy-α/β-D-
glucopyranoside (27b) was obtained as a colorless oil in an 8:1 α:β
ratio.
Data for the α isomer: Rf = 0.27 (10% MeOH in EtOAc); 1H NMR
CD3OD) δ 4.71 (d, J1,2a = 3.6 Hz, 1H, H1), 3.81 (ddd, J3,4 = 3.2, J2b,3
4.1, J2b,3 = 12.6 Hz, 1H, H3), 3.78−3.72 (m, 2H, H4 and H5), 3.31 (s,
3H, OMe), 3.30−3.22 (m, 2H, H6a and H6b), 1.82 (td, J2a,2b = J2a,3
=
(600 MHz, D2O) δ 4.86 (d, J1,2b = 3.5 Hz, 1H, H1), 3.83 (dd, J5,6a
2.1, J6a,6b = 12.0 Hz, 1H, H6a), 3.79−3.87 (m, 1H, H3), 3.73 (dd, J5,6b
= 5.5, J6a,6b = 12.0 Hz, 1H, H6b), 3.56 (ddd, J5,6a = 2.1, J5,6b = 5.5, J4,5
=
=
=
12.6, J1,2a = 3.6 Hz, 1H, H2a), 1.62 (dd, J2b,3 = 5.1, J2a,2b = 12.6 Hz, 1H,
H2b); 13C NMR (125 MHz, CD3OD) δ 100.3 (C1), 72.9 (C4), 70.5
(C5), 66.7 (C3), 55.5 (OCH3), 33.1 (C2), 4.5 (C6); HRMS(ESI) m/z
calcd for [C7H13O4I + Na]+ 310.9756, found 310.9756.
9.6 Hz, 1H, H5), 3.32 (s, 3H, OMe), 3.31 (m, 1H, H4), 2.10 (dd, J2a,3
= 5.3, J2a,2b = 13.5 Hz, 1H, H2a), 1.67 (ddd, J1,2b = 3.5, J2b,3 = 12.0, J2a,2b
= 13.5 Hz, 1H, H2b); 13C NMR (150 MHz, D2O) δ 98.1 (C1), 71.9
(C5), 70.8 (C4), 68.0 (C3), 60.5 (C6), 54.2 (OMe), 36.4 (C2);
HRMS(ESI) m/z calcd for [C7H14O5 + Na]+ 201.0733, found
201.0735.
(3R,4S)-1-Aminohex-5-ene-3,4-diol (24). To a solution of
iodogalactoside 22 (58 mg, 0.21 mmol) in a saturated solution of
NH4OAc in EtOH (3 mL) were added activated Zn (36 mg, 0.55
mmol), NaCNBH3 (21 mg, 0.33 mmol), and 30% aq. NH3 (1 mL).
The mixture was stirred at reflux for 18 h, filtered, and concentrated in
vacuo. The mixture was loaded on a Dowex-H+ column, eluted in 15%
aq. NH3, and concentrated in vacuo. Acidification with 1 M HCl
yielded alkenylamine 24, a white solid, as the HCl salt (35 mg, 0.21
mmol, 99%). Rf = 0.11 (DCM/EtOH/MeOH/30% aq. NH3, 5/2/2/1
v/v/v/v); [α]2D5 = −114 (c = 0.1, EtOH); IR (film) 3370, 3321, 2962,
Data for the β isomer: Rf = 0.27 (10% MeOH in EtOAc); 1H NMR
(600 MHz, D2O) δ 4.59 (dd, J1,2a = 1.8, J1,2b = 9.9 Hz, 1H, H1), 3.89
(dd, J5,6a = 2.0, J6a,6b = 12.0 Hz, 1H, H6a), 3.65−3.72 (m, 2H, H3 and
H6b), 3.48 (s, 3H, OMe), 3.32 (m, 1H, H5), 3.21 (t, J3,4 = J4,5 = 9.6
Hz, 1H, H4), 2.21 (ddd, J1,2a = 1.8, J2a,3 = 5.1, J2a,2b = 12.9 Hz, 1H,
H2a), 1.43 (ddd, J1,2b = 9.9, J2b,3 = 12.0, J2a,2b = 12.9 Hz, 1H, H2b); 13
C
NMR (150 MHz, D2O) δ 100.6 (C1), 75.9 (C5), 70.9 (C4), 70.2
(C3), 60.8 (C6), 56.4 (OMe), 38.0 (C2); HRMS(ESI) m/z calcd for
[C7H14O5 + Na]+ 201.0739, found 201.0735.
1
2836, 1642, 1504, 1435, 1378, 1025, 652 cm−1; H NMR (500 MHz,
D2O) δ 5.87 (ddd, J4,5 = 6.2, J5,6‑cis = 10.3, J5,6‑trans = 17.1 Hz, 1H, H5),
5.30 (d, J5,6‑trans = 17.1 Hz, 1H, H6-cis), 5.27 (d, J5,6‑cis = 10.3 Hz, 1H,
Methyl 2,6-Dideoxy-6-iodo-α-D-glucoside (28). To a solution
of methyl 2-deoxy-D-glucoside (70.1 mg, 0.4 mmol) in dry THF (4
mL) under an atmosphere of argon were added imidazole (81.6 mg,
1.2 mmol), PPh3 (157 mg, 0.6 mmol), and I2 (152 mg, 0.6 mmol).
The reaction mixture was heated to 75 °C for 1.5 h and then cooled
and concentrated. The product was purified by silica gradient flash
column chromatography (petroleum ether/EtOAc, 2/1 v/v) and then
reversed-phase HP20 chromatography (MeOH/water, 1/5 v/v) to
give methyl 2,6-dideoxy-6-iodo-α-D-glucopyranoside (28) as a color-
less oil (74.1 mg, 0.26 mmol, 64% over two steps from 2-deoxyglucose
27). Rf = 0.42 (1% MeOH in EtOAc); [α]2D3 = +84.0 (c = 1.0, CHCl3)
(lit [α]2D7 = +97, c = 0.9 in chloroform51); IR (film) 3387, 2934, 1442,
1377, 1211, 1128, 1044, 966, 938 cm−1; 1H NMR (500 MHz, CDCl3)
δ 4.83 (d, J1,2b = 3.4 Hz, 1H, H1), 3.97 (ddd, J2a,3 = 5.1, J3,4 = 8.7, J2b,3
= 11.7 Hz, 1H, H3), 3.57−3.62 (m, 1H, H6a), 3.40 (s, 3H, OMe),
3.35−3.41 (m, 2H, H5 and H6b), 3.24−3.30 (m, 1H, H4), 2.16 (ddd,
J1,2a = 1.0, J2a,3 = 5.1, J2a,2b = 12.9 Hz, 1H, H2a), 1.71 (ddd, J1,2b = 3.2,
J2b,3 = 11.7, J2a,2b = 12.9 Hz, 1H, H2b); 13C NMR (125 MHz, CDCl3)
δ 98.7 (C1), 77.6 (C4), 70.4 (C5), 69.1 (C3), 55.3 (OMe), 37.8 (C2),
7.9 (C6); HRMS(ESI) m/z calcd for [C7H13O4I + Na]+ 310.9756,
found 310.9751.
(3R,4R)-1-Amino-hex-5-ene-3,4-diol Hydrochloride (29). To a
solution of methyl 2,6-dideoxy-6-iodo-α-D-glucoside (28) (111 mg,
0.39 mmol) in a saturated solution of NH4OAc in EtOH (7.3 mL)
were added activated Zn (450 mg, 6.9 mmol), NaCNBH3 (85 mg, 1.3
mmol), and 30% aq. NH3 (2.9 mL). The mixture was stirred at reflux
for 4.5 h, cooled to room temperature, filtered through Celite, and
concentrated under reduced pressure. HCl (30 mL, 2 M) was added,
and AcOH was removed by evaporation. Coevaporation with water/
EtOH was performed to remove traces of AcOH, after which an
aqueous solution of NaOH (40 mL, 1.25 M) was added and the NH3
was evaporated. The resulting solid was filtered and washed with
EtOH to remove the bulk of the insoluble NaCl salt, and the filtrate
and combined washings were concentrated and then dry-loaded on
silica for column chromatography (DCM/EtOH/MeOH/35% aq.
H6-trans), 4.03 (dd, J3,4 = 4.6, J4,5 = 6.2 Hz, 1H, H4), 3.69 (ddd, J2a,3
3.2, J3,4 = 4.6, J2b,3 = 9.6 Hz, 1H, H3), 2.85 (ddd, J1a,2b = 5.9, J1a,2a = 8.5,
J1a,1b = 12.8 Hz, 1H, H1a), 2.77 (ddd, J1b,2a = 7.0, J1b,2b = 8.0, J1a,1b
=
=
12.8 Hz, 1H, H1b), 1.72 (dddd, J2a,3 = 3.2, J1b,2a = 7.0, J1a,2a = 8.5, J2a,2b
= 17.5 Hz, 1H, H2a), 1.55 (dddd, J1a,2b = 5.9, J1b,2b = 8.0, J2b,3 = 9.6,
J2a,2b = 17.5 Hz, 1H, H2b); 13C NMR (75 MHz, D2O) δ 135.5 (C5),
118.0 (C6), 75.4 (C4), 71.6 (C3), 37.1 (C1), 28.9 (C2); HRMS(ESI)
m/z calcd for [C6H13O2N + H]+ 132.1025, found 132.1022.
1-N,6-O-Carbonyl-4-epi-fagomine (25a). To a solution of
alkenylamine hydrochloride 24 (17 mg, 0.13 mmol) in saturated
NaHCO3 (aq) (0.6 mL) was added I2 (32 mg, 0.25 mmol). The
solution was stirred for 5 days at 35 °C and then filtered and
concentrated in vacuo. Purification by flash chromatography (EtOAc/
MeOH, 99/1 v/v) yielded carbamate 25a as a white solid (15.1 mg,
0.0873 mmol, 86%). [α]1D9 = +4.0 (c = 1, MeOH); IR (film) 3338,
2945, 2833, 2360, 2342, 1653, 1449, 1417, 1114, 1021 cm−1; 1H NMR
(500 MHz, D2O) δ 4.48 (t, J5,6 = J6a,6b = 9.0 Hz, 1H, H6a), 4.37 (dd,
J5,6 = 5.5, J6a,6b = 9.0 Hz, 1H, H6b), 4.03 (ddd, J4,5 = 1.8, J5,6b = 5.5, J5,6a
= 9.0 Hz, 1H, H5), 3.89 (m, 1H, H4), 3.86 (dd, J3,4 = 2.5, J2,3 = 5.5 Hz,
1H, H3), 3.76 (ddd, J1a,2a = 1.8, J1a,2b = 5.5, J1a,1b = 13.5 Hz, 1H, H1a),
3.06 (app dt, J1b,2a = J1b,2b = 4.5, J1a,1b = 13 Hz, 1H, H1b), 1.70−1.80
(m, 2H, H2a and H2b); 13C NMR (125 MHz, D2O) δ 159.2 (CO)
68.6 (C3), 67.9 (C4), 64.1 (C6), 57.0 (C5), 38.5 (C1), 25.1 (C2);
HRMS(ESI) m/z calcd for [C7H11NO4 + Na]+ 196.0586, found
196.0590.
4-epi-Fagomine (4). To a solution of carbamate 25a (11 mg, 0.06
mmol) in EtOH (2 mL) was added NaOH (32 mg, 0.80 mmol). The
solution was stirred at reflux for 2 h and then cooled and purified
directly using a Dowex-H+ column. The product was eluted in 5 to
15% aq. NH3 and concentrated in vacuo. Purification by flash
chromatography (DCM/EtOH/MeOH/30% aq. NH3, 25/2/2/1 to
5/2/2/1 v/v/v/v) yielded piperidine 4 as a white solid (8.1 mg, 0.055
mmol, 92%). [α]1D9 = +11.8 (c = 0.1, MeOH); IR (film) 3286, 2923,
I
dx.doi.org/10.1021/jo401512h | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
NH3, 55/2/2/1 to 5/2/2/1 v/v/v/v). Alkenylamine 29 was obtained
as the HCl salt after the addition of aq. HCl (1.2 M) and
concentration to obtain a white solid (55.5 mg, 0.33 mmol, 86%).
Rf = 0.16 (DCM/EtOH/MeOH/35% aq. NH3, 5/2/2/1 v/v/v/v);
[α]2D7 = +37.5 (c = 1.0, MeOH); IR (film) 3336, 2924, 1621, 1505,
compound was dry-loaded on silica before gradient column
chromatography (petroleum ether to petroleum ether/EtOAc, 1/1
v/v). Acetylated carbamate 33 was recovered as a white solid. Rf = 0.44
(5% MeOH in EtOAc); [α]2D7 = +19.9 (c = 0.1, MeOH); IR (film)
1
2957, 2925, 1741, 1711, 1430, 1367, 1238, 1070, 754 cm−1; H NMR
1
1468, 1396, 1312, 1255, 1122, 1020, 997, 931, 854 cm−1; H NMR
(500 MHz, CDCl3) δ 5.16 (dt, J3,4 = 6.5, J2a,3 = J2b,3 = 7.5 Hz, 1H, H3),
5.06 (td, J5,6a = 4.9, J4,5 = J5,6b = 9.0 Hz, 1H, H5), 4.32 (dd, J5,6a = 4.9,
J6a,6b = 11.0 Hz, 1H, H6a), 4.05 (dd, J5,6b = 9.0, J6a,6b = 11.0 Hz, 1H,
H6b), 3.79 (dt, J1a,2a = J1a,2b = 7.5, J1a,1b = 11.0 Hz, 1H, H1a), 3.50−
(500 MHz, D2O) δ 5.87 (ddd, J4,5 = 6.7, J5,6‑cis = 10.5, J5,6‑trans = 17.3
Hz, 1H, H5), 5.34 (d, J5,6‑trans = 17.3 Hz, 1H, H6-trans), 5.28 (d, J5,6‑cis
= 10.5 Hz, 1H, H6-cis), 4.03 (dd, J4,5 = 6.7, J3,4 = 6.8 Hz, 1H, H4), 3.70
(ddd, J2a,3 = 2.9, J2b,3 = 5.4, J3,4 = 6.8 Hz, 1H, H3), 3.13 (m, 2H, H1),
1.89 (m, 1H, H2a), 1.77 (m, 1H, H2b); 13C NMR (125 MHz, D2O) δ
136.1 (C5), 118.1 (C6), 75.4 (C4), 71.7 (C3), 37.2 (C1), 29.4 (C2);
HRMS(ESI) m/z calcd for [C6H13O2N + H]+ 132.1025, found
132.1031.
3.57 (m, 2H, H1b and H4), 2.32 (dt, J1a,2a = J1b,2a = J2a,3 = 7.5, J2a,2b
=
14.5 Hz, 1H, H2a), 2.11 (s, 3H, Me), 2.08 (s, 3H, Me), 1.88 (dt, J1a,2b
1
= J1b,2b = J2b,3 = 7.5, J2a,2b = 14.5 Hz, 1H, H2b); H NMR (300 MHz,
CD3OD) δ 5.19 (dd, J3,4 = 7.2, J2b,3 = 7.9 Hz, 1H, H3), 5.12 (dd, J5,6a
=
4.7, J4,5 = J5,6b = 8.7 Hz, 1H, H5), 4.32 (dd, J5,6a = 4.7, J6a,6b = 10.8 Hz,
1H, H6a), 4.11 (dd, J5,6b = 8.7, J6a,6b = 10.8 Hz, 1H, H6b), 3.66 (dt,
Iodocyclization/Carbamate Annulation. To a solution of
alkenylamine hydrochloride 29 (240 mg, 1.4 mmol) in saturated aq.
NaHCO3 (7.2 mL) (made fresh immediately prior to use) were added
I2 (1900 mg, 7.5 mmol) and NaHCO3 (1400 mg, 17 mmol). The
solution was stirred at room temperature for 5 days, filtered, and
concentrated under reduced pressure. The products, all as white solids,
were separated by repeated use of silica plug (MeOH in EtOAc, 0−
10% v/v) and reversed-phase columns (octyl-bonded end-capped silica
beads) eluting in H2O (for carbamates 30, 31a, and 31b) and MeOH
(for iodide 32). Yield of 30 = 10−30%.
J1a,2a = J1a,2b = 8.0, J1a,1b = 11.1 Hz, 1H, H1a), 3.62 (dd, J3,4 = 7.2, J4,5
=
8.7 Hz, 1H, H4), 3.49 (ddd, J1b,2b = 4.7, J1b,2a = 9.1, J1a,1b = 11.1 Hz,
1H, H1b), 2.34 (dddd, J1b,2b = 4.7, J2b,3 = 7.9, J1a,2b = 8.0, J2a,2b = 12.9
Hz, 1H, H2a), 2.07 (s, 3H, Me), 2.05 (s, 3H, Me), 1.88 (ddt, J1a,2b
=
J2b,3 = 7.9, J1b,2b = 9.1, J2a,2b = 12.9 Hz, 1H, H2b); 13C NMR (125
MHz, CDCl3) δ 170.1 (CO, Ac), 169.9 (CO, Ac), 151.6 (C
O), 75.2 (C3), 66.7 (C6), 65.7 (C5), 63.2 (C4), 44.9 (C1), 28.8 (C2),
20.9 (CH3, Ac), 20.8 (CH3, Ac); HRMS(ESI) m/z calcd for
[C11H15O6N + Na]+ 280.0797, found 280.0791.
1-N,6-O-Carbonyl-1,2,4-trideoxy-1,4-imino-L-allitol (30). [α]2D5
=
1,2,4-Trideoxy-1,4-imino-L-allitol (34). Carbamate (30) (10.6
mg, 0.061 mmol) was dissolved in EtOH (1 mL), and NaOH (26 mg,
0.6 mmol) was added. The solution was refluxed for 2 h, neutralized
with aq. HCl (1 M), and loaded on silica for chromatography (DCM/
EtOH/MeOH/35% aq. NH3, 55/2/2/1 to 15/2/2/1 v/v/v/v).
Product 34 was obtained as a white solid (8.0 mg, 0.054 mmol,
89%). Rf = 0.19 (DCM/EtOH/MeOH/35% aq. NH3, 5/2/2/1 v/v/v/
v); [α]2D6 = −1.8 (c = 0.41, MeOH); IR (film) 3309, 2956, 2927, 1629,
−0.5 (c = 0.4, MeOH); IR (film) 3390, 3331, 2965, 2914, 1719, 1634,
1517, 1477, 1455, 1133, 1110, 1070 cm−1; 1H NMR (500 MHz, D2O)
δ 4.19−4.27 (m, 2H, H3 and H6a), 3.94−4.03 (m, 2H, H5 and H6b),
3.47 (dd, J1,2a = 6.1, J1,2b = 9.0 Hz, 2H, H1), 3.28 (t, J4,5 = J3,4 = 7.7 Hz,
1H, H4), 2.24 (dq, J1,2a = J2a,3 = 6.1, J2a,2b = 12.6 Hz, 1H, H2a), 1.80
(dq, J1,2b = J2b,3 = 9.0, J2a,2b = 12.6 Hz, 1H, H2b); 13C NMR (150 MHz,
D2O) δ 154.4 (CO), 74.0 (C3), 69.2 (C6), 64.8 (C4), 64.2 (C5),
44.2 (C1), 30.2 (C2); HRMS(ESI) m/z calcd for [C7H11O4N + Na]+
196.0586, found 196.0579.
1420, 1089, 1041 cm−1; 1H NMR (500 MHz, D2O) δ 4.33 (dt, J2b,3
=
J3,4 = 4.3, J2a,3 = 6.2 Hz, 1H, H3), 3.78 (dd, J4,5 = 4.3, J5,6 = 10.3 Hz,
1H, H5), 3.44−3.52 (m, 2H, H6), 3.35 (t, J3,4 = J4,5 = 4.3 Hz, 1H, H4),
3.21 (dd, J1,2b = 6.3, J1,2a = 8.1 Hz, 2H, H1), 2.02 (ddd, J2a,3 = 6.2, J1,2a
= 8.1, J2a,2b = 14.2 Hz, 1H, H2a), 1.78 (ddd, J2b,3 = 4.3, J1,2b = 6.3, J2a,2b
= 14.2 Hz, 1H, H2b); 13C NMR (125 MHz, D2O) δ 69.4 (C3), 68.4
(C5), 67.1 (C4), 62.5 (C6), 44.1 (C1), 32.6 (C2); HRMS(ESI) m/z
calcd for [C6H13O3N + H]+ 148.0974, found 148.0978.
5-epi-Fagomine (26) and Fagomine (3). Carbamates 31a and
31b (11.7 mg, 0.07 mmol) were dissolved in EtOH (1 mL), and
NaOH (80 mg, 2 mmol) was added. The solution was refluxed for 2 h,
neutralized with aq. HCl (1 M), and concentrated in vacuo.
Purification by gradient silica column chromatography (DCM/
EtOH/MeOH/35% aq. NH3, 45/2/2/1 to 15/2/2/1 v/v/v/v) gave
the products 3 and 26 as a white solid and a mixture of diastereomers
(12.4 mg, 0.07 mmol, quant.).
1-N,6-O-Carbonyl-5-epi-fagomine (31a) and 1-N,6-O-Carbon-
ylfagomine (31b). IR (film) 3409, 2948, 2840, 1632, 1462, 1403,
1092, 1010 cm−1.
Data for the major isomer (31a): 1H NMR (500 MHz, D2O) δ 4.53
(t, J5,6a = J6a,6b = 9.0 Hz, 1H, H6a), 4.34 (dd, J5,6b = 6.0, J6a,6b = 9.0 Hz,
1H, H6b), 4.23 (ddd, J4,5 = 2.5, J5,6b = 6.0, J5,6a = 9.0 Hz, 1H, H5), 4.06
(dd, J3,4 = 3.2, J2a,3 = 6.3 Hz, 1H, H3), 3.73 (m, 1H, H4), 3.56 (dd,
J1a,2a = 6.3, J1a,1b = 13.5 Hz, 1H, H1a), 3.23 (td, J1b,2a = 3.7, J1a,1b = J1b,2b
= 13.5 Hz, 1H, H1b), 1.95−2.02 (m, 1H, H2a), 1.64 (d, J2a,2b = 15.0
Hz, 1H, H2b); 13C NMR (125 MHz, D2O) δ 159.2 (CO), 66.5
(C3), 66.5 (C4), 64.1 (C6), 53.2 (C5), 35.2 (C1), 24.4 (C2);
HRMS(ESI) m/z calcd for [C7H11O4N + Na]+ 196.0586, found
196.0587.
1
Data for the minor isomer (31b): H NMR (500 MHz, D2O) δ
Data for 5-epi-fagomine (26): Rf = 0.24 (DCM/EtOH/MeOH/35%
aq. NH3, 5/2/2/1 v/v/v/v); [α]2D9 = −8.9 (c = 0.63, MeOH); IR
(film) 3364, 2970, 1425, 1366, 1217, 1079 cm−1; 1H NMR (500 MHz,
D2O) δ 4.02 (m, 1H, H3), 3.93 (m, 1H, H4), 3.85 (dd, J5,6a = 4.9, J6a,6b
= 12.0 Hz, 1H, H6a), 3.80 (dd, J5,6b = 8.5, J6a,6b = 12.0 Hz, 1H, H6b),
3.54 (dd, J5,6a = 4.9, J5,6b = 8.5 Hz, 1H, H5), 3.20−3.31 (m, 2H,
H1a,b), 2.16−2.24 (m, 1H, H2a), 1.84 (d, J2a,2b = 14.5 Hz, 1H, H2b);
13C NMR (125 MHz, D2O) δ 65.8 (C4), 64.8 (C3), 59.5 (C6), 55.8
(C5), 38.9 (C1), 23.6 (C2); HRMS(ESI) m/z calcd for [C6H13O3N +
H]+ 148.0974, found 148.0968.
4.53 (t, J5,6a = J6a,6b = 9.0 Hz, 1H, H6a), 4.31 (dd, J5,6b = 5.0, J6a,6b = 9.0
Hz, 1H, H6b), 3.68−3.71 (m, 1H, H1a), 3.67 (ddd, J5,6b = 5.0, J5,6a
=
9.0, J4,5 = 9.5 Hz, 1H, H5), 3.62 (ddd, J2b,3 = 5.0, J3,4 = 9.5, J2a,3 = 11.5
Hz, 1H, H3), 3.34 (t, J3,4 = J4,5 = 9.4 Hz, 1H, H4), 3.03 (td, J1b,2a = 3.2,
J1a,1b = J1b,2b = 13.0 Hz, 1H, H1b), 1.92−1.98 (m, 1H, H2a), 1.51 (ddt,
J2b,3 = 5.0, J1a,2b = 11.7, J1b,2b = J2a,2b = 13.0 Hz, 1H, H2b); 13C NMR
(125 MHz, D2O) δ 158.7 (CO), 74.5 (C4), 71.4 (C3), 66.5 (C6),
57.6 (C5), 38.7 (C1), 30.8 (C2); HRMS(ESI) m/z calcd for
[C7H11O4N + Na]+ 196.0586, found 196.0587.
1-Amino-1-N,3-O-carbonyl-1,2,6-trideoxy-6-iodo-L-gulitol (32). Rf
= 0.26 (10% MeOH in EtOAc); [α]2D6 = −8.9 (c = 0.67, MeOH); IR
(film) 3367, 2928, 1680, 1488, 1457, 1300, 1223, 1108, 1021, 523
Data for fagomine (3): Rf = 0.33 (DCM/EtOH/MeOH/35% aq.
NH3, 5/2/2/1 v/v/v/v); [α]2D9 = +8.4 (c = 0.27, MeOH); IR (film)
3367, 2930, 1461, 1271, 1073 cm−1; 1H NMR (500 MHz, D2O) δ 3.93
1
cm−1; H NMR (500 MHz, D2O) δ 4.50 (dt, J = 3.3, J = 9.9 Hz, 1H,
(dd, J5,6a = 3.2, J6a,6b = 12.5 Hz, 1H, H6a), 3.88 (dd, J5,6b = 5.5, J6a,6b
12.5 Hz, 1H, H6b), 3.71 (ddd, J2a,3 = 4.9, J3,4 = 9.0, J2b,3 = 11.7 Hz, 1H,
H3), 3.52 (t, J3,4 = J4,5 = 9.0 Hz, 1H, H4), 3.44 (ddd, J1a,2a = 2.5, J1a,2b
=
H3), 3.72−3.81 (m, 2H, H4 and H5), 3.44 (dd, J = 3.9, J = 10.5 Hz,
1H, H6a), 3.29−3.37 (m, 3H, H1a,b and H6b), 1.90−2.01 (m, 2H,
H2a,b); 13C NMR (125 MHz, D2O) δ 154.7 (CO), 76.1 (C3), 71.8,
68.1 (C4 and C5), 35.8 (C1), 19.6 (C2), 6.2 (C6); HRMS(ESI) m/z
calcd for [C7H12O4NI + H]+ 301.9889, found 301.9888.
3,5-Di-O-acetyl-1-N,6-O-carbonyl-1,2,4-trideoxy-1,4-imino-L-
allitol (33). Carbamate (30) (124.7 mg, 0.7 mmol) was subjected to
dry pyridine (1.0 mL) and acetic anhydride (1.0 mL) at rt under an
argon atmosphere overnight (14 h). After concentration, the crude
=
4.1, J1a,1b = 13.5 Hz, 1H, H1a), 3.12−3.16 (m, 1H, H5), 3.10 (dd, J1b,2a
= 2.5, J1b,2b = J1a,1b = 13.5 Hz, 1H, H1b), 2.21 (ddd, J = 2.5, J2a,3 = 4.9,
J2a,2b = 13.5 Hz, 1H, H2a), 1.72 (ddt, J1a,2b = 4.1, J2b,3 = 11.7, J1a,2b
=
J2a,2b = 13.5 Hz, 1H, H2b); 13C NMR (125 MHz, D2O) δ 70.4 (C3),
69.5 (C4), 60.0 (C5), 57.6 (C6), 41.8 (C1), 28.5 (C2); HRMS(ESI)
m/z calcd for [C6H13O3N + H]+ 148.0974, found 148.0968.
J
dx.doi.org/10.1021/jo401512h | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
4,5-Di-O-acetyl-1-amino-1-N,3-O-carbonyl-1,2,6-trideoxy-6-
iodo-L-gulitol (35). Iodide 32 (20.4 mg, 0.07 mmol) was subjected to
dry pyridine (0.5 mL) and acetic anhydride (0.5 mL) at room
temperature under an argon atmosphere overnight (9 h). After
concentration and repeated coevaporation with toluene to remove
pyridine, the crude compound was dry-loaded on silica before gradient
column chromatography (petroleum ether to petroleum ether/EtOAc,
1/1 v/v). Acetylated iodide 35 was recovered as a white solid in 48%
yield (12.5 mg, 0.03 mmol). Rf = 0.44 (10% MeOH in EtOAc); [α]D25
= +6.6 (c = 0.23, CHCl3); IR (film) 3266, 3056, 2927, 1742, 1716,
1373, 1266, 1219, 1117, 1059, 1024, 598.7 cm−1; 1H NMR (500 MHz,
CDCl3) δ 5.51 (br s, 1H, NH), 5.34 (dd, J3,4 = 3.7, J4,5 = 5.5 Hz, 1H,
concentrated before coevaporation with H2O and toluene to remove
traces of AcOH. The crude product was purified by silica column
chromatography (petroleum ether to petroleum ether/EtOAc, 1/3 v/
v) to give 38 as a white solid (5.5 mg, 0.03 mmol, quant.). Rf = 0.38
(10% MeOH in EtOAc); [α]D25 = −11.0 (c = 0.13, CHCl3); IR (film)
3278, 2925, 2854, 1703, 1489, 1459, 1424, 1372, 1294, 1227, 1104,
1023, 950, 763 cm−1; 1H NMR (500 MHz, CDCl3) δ 5.88 (ddd, J4,5
=
6.5, J5,6‑cis = 10.5, J5,6‑trans = 17.0 Hz, 1H, H5), 5.51 (br s, 1H, NH), 5.44
(d, J4,5 = 6.5 Hz, 1H, H4), 5.43 (d, J5,6‑trans = 17.0 Hz, 1H, H6-trans),
5.37 (d, J5,6‑cis = 10.5 Hz, 1H, H6-cis), 4.39 (ddd, J = 2.5, J = 5.5, J2b,3
=
10.5 Hz, 1H, H3), 3.35−3.45 (m, 2H, H1a,b), 2.13 (s, 3H, Me), 1.94−
2.01 (m, 1H, H2a), 1.86 (dtd, J = 6.5, J = J2b,3 = 10.5, J2a,2b = 14.0 Hz,
1H, H2b); 13C NMR (125 MHz, CDCl3) δ 170.0 (CO, Ac), 153.8
(CO), 131.5 (C5), 120.4 (C6), 77.5 (C3), 74.5 (C4), 39.0 (C1),
22.5 (C2), 21.2 (CH3, Ac); HRMS(ESI) m/z calcd for [C9H13O4N +
H]+ 200.0923, found 200.0918.
(3R,4R)-6-Aminohex-1-ene-3,4-diol Hydrochloride (29). Ace-
tylated alkene carbamate 38 (5.7 mg, 0.03 mmol) was dissolved in
EtOH (1 mL), and NaOH (80 mg, 2 mmol) was added. The solution
was refluxed for 1 h, neutralized with aq. HCl (1.2 M), and then
concentrated in vacuo. Purification by gradient silica column
chromatography (DCM/EtOH/MeOH/35% aq. NH3, 65/2/2/1 to
5/2/2/1 v/v/v/v) gave the alkenylamine, which was converted into
the HCl salt 29 as a white solid (3.3 mg, 0.02 mmol, 69%) by the
addition of HCl (1.2 M, aq.) and concentration. Rf = 0.16 (DCM/
EtOH/MeOH/35% aq. NH3, 5/2/2/1 v/v/v/v); [α]2D7 = +31.8 (c =
0.2, MeOH).
H4), 5.14 (dd, J5,6a = 4.5, J4,5 = J5,6b = 5.5 Hz, 1H, H5), 4.47 (d, J2b,3
=
10.8 Hz, 1H, H3), 3.46 (dd, J5,6a = 4.5, J6a,6b = 11.4 Hz, 1H, H6a),
3.37−4.43 (m, 2H, H1a,b), 3.36 (dd, J5,6b = 5.5, J6a,6b = 11.4 Hz, 1H,
H6b), 2.16 (s, 3H, 4-OAc), 2.13 (s, 3H, 5-OAc), 2.00 (d, J2a,2b = 14.2
Hz, 1H, H2a), 1.79−1.89 (m, 1H, H2b); 13C NMR (125 MHz,
CDCl3) δ 170.2 (CO, 4-Ac), 169.9 (CO, 5-Ac), 152.9 (CO),
75.0 (C3), 73.4 (C4), 70.4 (C5), 38.9 (C1), 22.8 (C2), 20.9 (CH3,
Ac), 20.6 (CH3, Ac), 3.3 (C6); HRMS(ESI) m/z calcd for
[C11H16O6NI + H]+ 386.0101, found 386.0094.
1-Amino-1-N,3-O-carbonyl-1,2,6-trideoxy-6-iodo-4,5-O-iso-
propylidene-L-gulitol (36). Iodide 32 (8.8 mg, 0.3 mmol) was
dissolved in acetone (0.45 mL), dimethoxypropane (0.45 mL), and
MeOH (0.25 mL). A 100 μL aliquot of pTsOH·H2O in acetone (2.85
mg/mL) was added, and the reaction mixture was stirred at room
temperature overnight (17 h). The mixture was then neutralized with
Dowex−OH−, filtered, and concentrated to give a white solid 36 (7.2
mg, 0.02 mmol, 72%) that was used without further purification. Rf =
0.49 (10% MeOH in EtOAc); [α]2D3 = −23.0 (c = 0.48, CHCl3); IR
(film) 3282, 2985, 2930, 1700, 1487, 1456, 1421, 1372, 1294, 1239,
ASSOCIATED CONTENT
■
S
* Supporting Information
NMR spectra (1H and 13C) for all new compounds. This
material is available free of charge via the Internet at http://
pubs.acs.org.
1
1214, 1107, 1069, 554 cm−1; H NMR (500 MHz, CDCl3/CD3OD
95/5) δ 4.40 (app dt, J3,4 = 2.2, J2a,3 = 10.0 Hz, 1H, H3), 4.20 (dt, J5,6a
= J5,6b = 4.7, J4,5 = 7.5 Hz, 1H, H5), 3.83 (dd, J3,4 = 2.2, J4,5 = 7.5 Hz,
1H, H4), 3.41 (ddd, J1a,2b = 3.7, J1a,2a = 5.8, J1a,1b = 11.0 Hz, 1H, H1a),
3.36 (m, 1H, NH), 3.31−3.35 (m, 1H, H1b), 3.31 (dd, J5,6a = 4.7, J6a,6b
= 10.8 Hz, 1H, H6a), 3.28 (dd, J5,6b = 4.7, J6a,6b = 10.8 Hz, 1H, H6b),
2.06 (dtd, J1a,2a = 5.8, J1b,2a = J2a,3 = 10.0, J2a,2b = 14.0 Hz, 1H, H2a),
1.98 (m, 1H, H2b), 1.43 (s, 3H, Me), 1.40 (s, 3H, Me); 13C NMR
(125 MHz, CDCl3/CD3OD 95/5) δ 154.4 (CO), 110.7 (iPr-
C(CH3)2), 82.4 (C4), 75.1 (C3), 74.6 (C5), 38.7 (C1), 27.6, 27.0
(iPr-C(CH3)2), 23.3 (C2), 6.1 (C6); HRMS(ESI) m/z calcd for
[C10H16O4NI + H]+ 342.0202, found 342.0201.
AUTHOR INFORMATION
Corresponding Authors
■
*E-mail: bridget.stocker@vuw.ac.nz.
*E-mail: mattie.timmer@vuw.ac.nz. Telephone: + 64 4 463
6529. Fax: + 64 4 463 5237.
Notes
The authors declare no competing financial interest.
D-threo-1-Amino-1-O,3-N-carbonyl-1,2,5,6-tetradeoxyhex-5-
enose (37). Iodoacetonide 36 (6.8 mg, 0.02 mmol) was subjected to
DCM (0.3 mL), AcOH (0.1 mL), and zinc (27.6 mg, 0.2 mmol) and
stirred at 40−55 °C for 12 days. The reaction mixture was filtered
through silica and concentrated to remove zinc. The crude product
was purified by silica column chromatography (petroleum ether to
petroleum ether/EtOAc, 1/3 v/v), yielding 37 as a white solid (2.4
ACKNOWLEDGMENTS
■
The authors acknowledge the financial contributions of Victoria
University of Wellington (Postgraduate Scholarship and Curtis-
Gordon Research Scholarship, H.M.C.) and the Royal Society
of New Zealand Marsden Fund (M.S.M.T.).
mg, 0.015 mmol, 77%). Rf = 0.22 (10% MeOH in EtOAc); [α]D29
=
−12.8 (c = 0.16, MeOH); IR (film) 3373, 2926, 2856, 1682, 1489,
1459, 1294, 1108, 1071, 1028, 946 cm−1; 1H NMR (500 MHz, D2O) δ
5.92 (ddd, J4,5 = 6.6, J5,6‑cis = 10.5, J5,6‑trans = 17.3 Hz, 1H, H5), 5.40
(app dt, J4,6‑trans = 1.2, J5,6‑trans = 17.3 Hz, 1H, H6-trans), 5.32 (app dt,
REFERENCES
■
(1) For some recent reviews outlining the biological activity of
azasugars, see: (a) Asano, N.; Nash, R. J.; Molyneux, R. J.; Fleet, G. W.
J. Tetrahedron: Asymmetry 2000, 11, 1645−1680. (b) Butters, T. D.;
Dwek, R. A.; Platt, F. M. Curr. Top. Med. Chem. 2003, 3, 561−574.
J4,6‑cis = 1.2, J5,6‑cis = 10.5 Hz, 1H, H6-cis), 4.35 (ddd, J2a,3 = 3.0, J3,4
=
5.7, J2b,3 = 10.5 Hz, 1H, H3), 4.24 (ddt, J4,6‑trans = J4,6‑cis = 1.2, J3,4 = 5.7,
J4,5 = 6.6 Hz, 1H, H4), 3.33 (ddd, J1a,2a = 3.0, J1a,2b = 6.9, J1a,1b = 12.0
Hz, 1H, H1a), 3.32 (ddd, J1b,2a = 4.5, J1b,2b = 10.5, J1a,1b = 12.0 Hz, 1H,
H1b), 1.99 (ddt, J2a,3 = J1a,2a = 3.0, J1b,2a = 4.5, J2a,2b = 14.1 Hz, 1H,
H2a), 1.84 (dtd, J1a,2b = 6.9, J1b,2b = J2b,3 = 10.5, J2a,2b = 14.1 Hz, 1H,
H2b); 13C NMR (150 MHz, D2O) δ 156.8 (CO), 134.9 (C5),
118.4 (C6), 80.2 (C3), 73.5 (C4), 37.8 (C1), 21.7 (C2); HRMS(ESI)
m/z calcd for [C7H11O3N + H]+ 158.0817, found 158.0805.
D-threo-4-O-Acetyl-1-amino-1-O,3-N-carbonyl-1,2,5,6-tetra-
deoxyhex-5-enose (38). Iodide 35 (10.7 mg, 0.03 mmol) was
dissolved in THF (0.3 mL), and H2O (0.06 mL), AcOH (0.06 mL),
and zinc (62 mg, 0.6 mmol) were added. The resulting mixture was
then stirred at room temperature for 1 h. The reaction mixture was
filtered through Celite, washed with DCM and EtOH, and
(c) Greimel, P.; Spreitz, J.; Stutz, A. E.; Wrodnigg, T. M. Curr. Top.
̈
Med. Chem. 2003, 3, 513−523. (d) Borges de Melo, E.; da Silveira
Gomes, A.; Carvalho, I. Tetrahedron 2006, 62, 10277−10302.
(e) Merino, P.; Delso, I.; Marca, E.; Tejero, T.; Matute, R. Curr.
Chem. Biol. 2009, 3, 253−271. (f) Nash, R. J.; Kato, A.; Yu, C.-Y.;
Fleet, G. W. J. Future Med. Chem 2011, 3, 1513−1521. (g) Lopez, O.;
Merino-Montiel, P.; Martos, S.; Gonzalez-Benjumea, A. Carbohydr.
Chem. 2012, 38, 215−262. (h) Alonzi, D. S.; Butters, T. D. Chimia
2011, 65, 35−39. (i) Iminosugars: From Synthesis to Therapeutic
Applications; Compain, P., Martin, O. R., Eds.; Wiley: Chichester, U.K.,
2007.
(2) Segal, P.; Feig, P. U.; Schernthaner, G.; Ratzmann, K. P.; Rybka,
J.; Petzinna, D.; Berlin, C. Diabetes Care 1997, 20, 687−691.
K
dx.doi.org/10.1021/jo401512h | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
(3) See: A Study of AT1001 in Patients with Fabry Disease. http://
www.clinicaltrials.gov/ct/gui/show/NCT00231036 (accessed July 7,
2013).
(4) Lillelund, V. H.; Jensen, H. H.; Liang, X.; Bols, M. Chem. Rev.
2002, 102, 515−554.
(5) Meyers, A. I.; Andres, C. J.; Resek, J. E.; Woodall, C. C.;
McLaughlin, M. A.; Lee, P. H.; Price, D. A. Tetrahedron 1999, 55,
8931−8952.
(6) Kato, A.; Asano, N.; Kizu, H.; Matsui, K.; Watson, A. A.; Nash, R.
J. J. Nat. Prod. 1997, 60, 312−314.
(7) Nojima, H.; Kimura, I.; Chen, F.-J.; Sugihara, Y.; Haruno, M.;
Kato, A.; Asano, N. J. Nat. Prod. 1998, 61, 397−400.
(8) Fan, J.-Q.; Ishii, S.; Asano, N.; Suzuki, Y. Nat. Med. 1999, 5, 112−
115.
(9) Dangerfield, E. M.; Timmer, M. S. M.; Stocker, B. L. Org. Lett.
2009, 11, 535−538.
(10) Dangerfield, E. M.; Gulab, S. A.; Plunkett, C. H.; Timmer, M. S.
M.; Stocker, B. L. Carbohydr. Res. 2010, 345, 1360−1365.
(11) Dangerfield, E. M.; Plunkett, C. H.; Stocker, B. L.; Timmer, M.
S. M. Molecules 2009, 14, 5298−5307.
(12) For recent reviews concerning the synthesis of azasugars, see:
́
(a) Pearson, M. S. M.; Mathe-Allainmat, M.; Fargeas, V.; Lebreton, J.
Eur. J. Org. Chem. 2005, 2159−2191. (b) Afarinkia, K.; Bahar, A.
Tetrahedron: Asymmetry 2005, 16, 1239−1287. (c) Davis, B. G.
Tetrahedron: Asymmetry 2009, 20, 652−671. (d) Stocker, B. L.;
Dangerfield, E. M.; Win-Mason, A. L.; Haslett, G. W.; Timmer, M. S.
M. Eur. J. Org. Chem. 2010, 1615−1637. (e) Wijdeven, M. A.;
Willemsen, J.; Rutjes, F. P. J. T. Eur. J. Org. Chem. 2010, 2831−2844.
(13) Dangerfield, E. M.; Plunkett, C. H.; Win-Mason, A. L.; Stocker,
B. L.; Timmer, M. S. M. J. Org. Chem. 2010, 75, 5470−5477.
(14) Timmer, M. S. M.; Dangerfield, E. M.; Cheng, J. M. H.; Gulab,
S. A.; Stocker, B. L. Tetrahedron Lett. 2011, 52, 4803−4805.
(15) For a full discussion of the factors influencing the protecting-
group-free reductive amination methodology, see: (a) Miriyala, B.;
Bhattacharyya, S.; Williamson, J. S. Tetrahedron 2004, 60, 1463−1471.
(b) Pelter, A.; Rosser, R. M.; Mills, S. J. Chem. Soc., Perkin Trans. 1
1984, 717−720. (c) Borch, R. F.; Bernstein, M. D.; Durst, H. D. J. Am.
Chem. Soc. 1971, 93, 2897−2904. (d) Gomez, S.; Peters, J. A.; van der
Waal, J. C.; van den Brink, P. J.; Maschmeyer, T. Appl. Catal., A 2004,
261, 119−125. (e) Gomez, S.; Peters, J. A.; Maschmeyer, T. Adv.
Synth. Catal. 2002, 344, 1037−1057. (f) Baxter, E.; Reitz, A. Org. React.
2002, 59, 1−714.
A.; Aerts, J. M. F. G.; Overkleeft, H. S. ACS Med. Chem. Lett. 2011, 2,
519−522.
(25) Kumari, N.; Reddy, B. G.; Vankar, Y. D. Eur. J. Org. Chem. 2009,
160−169.
(26) Yorimitsu, H.; Murakami, Y.; Takamatsu, H.; Nishimura, S.;
Nakamura, E. Chem.Asian J. 2007, 2, 57−65.
(27) Skaanderup, P. R.; Poulsen, C. S.; Hyldtoft, L.; Jorgensen, M. R.;
Madsen, R. Synthesis 2002, 1721−1727.
(28) Swamy, N. K.; Pyne, S. G. Synth. Commun. 2011, 41, 2435−
2445.
(29) Bates, R. W.; Ng, P. S. Tetrahedron Lett. 2011, 52, 2969−2971.
(30) Jiangseubchatveera, N.; Bouillon, M. E.; Liawruangrath, B.;
Liawruangrath, S.; Nash, R. J.; Pyne, S. G. Org. Biomol. Chem. 2013, 11,
3826−3833.
(31) Lombardo, M.; Morganti, S.; Trombini, C. J. Org. Chem. 2003,
68, 997−1006.
(32) Dana, G.; Danechpajouh, H. Bull. Soc. Chim. Fr. 1980, 395−399.
(33) Bock, M.; Buntin, K.; Muller, R.; Kirschning, A. Angew. Chem.,
̈
Int. Ed. 2008, 47, 2308−2311.
(34) Stocker, B. L.; Win-Mason, A. L.; Timmer, M. S. M. Carbohydr.
Res. 2012, 356, 163−171.
(35) Chamberlin, A. R.; Mulholland, R. L., Jr.; Kahn, S. D.; Hehre, W.
J. J. Am. Chem. Soc. 1987, 109, 672−677.
(36) Song, W.-J.; Suh, D.-Y.; Kang, Y.-H.; No, K.; Han, B. H. Bull.
Korean Chem. Soc. 1998, 19, 593−595.
(37) Ohta, H.; Sakata, Y.; Takeuchi, T.; Ishii, Y. Chem. Lett. 1990,
733−736.
(38) Masuda, H.; Takase, K.; Nishio, M.; Hasegawa, A.; Nishiyama,
Y.; Ishii, Y. J. Org. Chem. 1994, 59, 5550−5555.
(39) Giese, B.; Bartmann, D. Tetrahedron Lett. 1985, 26, 1197−1200.
(40) Chamberlin, A. R.; Dezube, M.; Dussault, P.; McMills, M. C. J.
Am. Chem. Soc. 1983, 105, 5819−5825.
(41) Chamberlin, A. R.; Mulholland, R. L., Jr. Tetrahedron 1984, 40,
2297−2302.
(42) Midland, M. M.; Halterman, R. L. J. Org. Chem. 1981, 46,
1227−1229.
(43) Peeters, A.; Ameloot, R.; De Vos, D. E. Green Chem. 2013, 15,
1550−1557.
(44) Payne, G. B. J. Org. Chem. 1962, 27, 3819−3822.
(45) Hanson, R. M. Org. React. 2002, 60, 1−156.
(46) For literature precedent for the formation of iodoexpoxides; see:
(a) Winstein, S.; Goodman, L. J. Am. Chem. Soc. 1954, 76, 4373−4378.
(b) Lindgren, B. O.; Svahn, C. M. Acta Chem. Scand. 1970, 24, 2699−
2704. (c) Ganem, B. J. Am. Chem. Soc. 1976, 98, 858−859. (d) Santelli,
M.; Viala, J. Tetrahedron Lett. 1977, 18, 4397−4400. (e) Evans, R. D.;
Magee, J. W.; Schauble, J. H. Synthesis 1988, 862−868. (f) Brunel, Y.;
Roussseau, G. Synlett 1995, 323−324.
(16) McGhee, W.; Riley, D.; Christ, K.; Pan, Y.; Parnas, B. J. Org.
Chem. 1995, 60, 2820−2830.
(17) Hampe, E. M.; Rudkevich, D. M. Tetrahedron 2003, 59, 9619−
9625.
(47) Baldwin, J. E. J. Chem. Soc., Chem. Commun. 1976, 734−736.
(48) Wipf, P.; Stephenson, C. R. J.; Walczak, M. A. A. Org. Lett. 2004,
6, 3009−3012.
(49) Win-Mason, A. L.; Dangerfield, E. M.; Tyler, P. C.; Stocker, B.
L.; Timmer, M. S. M. Eur. J. Org. Chem. 2011, 4008−4014.
(50) Win-Mason, A. L.; Jongkees, S. A. K.; Withers, S. G.; Tyler, P.
C.; Timmer, M. S. M.; Stocker, B. L. J. Org. Chem. 2011, 76, 9611−
9621.
(18) For some examples of recently developed protecting groups,
see: (a) Chmielewski, M. K. Tetrahedron Lett. 2012, 53, 666−669.
(b) Li, Y.; Roy, B.; Liu, X. Chem. Commun. 2011, 47, 8952−8954.
́
(c) Codee, J. D. C.; Ali, A.; Overkleeft, H. S.; van der Marel, G. A. C.
R. Chim. 2011, 14, 178−193. (d) Hagen, V.; Kilic, F.; Schaal, J.;
Dekowski, B.; Schmidt, R.; Kotzur, N. J. Org. Chem. 2010, 75, 2790−
2797. (e) Ali, A.; Van den Berg, R. J. B. H. N.; Overkleeft, H. S.;
́
Filippov, D. V.; van der Marel, G. A.; Codee, J. D. C. Tetrahedron Lett.
(51) Durette, P. L. Synthesis 1980, 1037−1038.
2009, 50, 2185−2188. (f) Timmer, M. S. M.; Stocker, B. L.;
Northcote, P. T.; Burkett, B. A. Tetrahedron Lett. 2009, 50, 7199−
7204.
(19) Foster, A. B.; Overend, W. G.; Stacey, M.; Vaughan, G. J. Chem.
Soc. 1954, 3367−3377.
(20) Toyokuni, T.; Cai, S.; Dean, B. Synthesis 1992, 1236−1238.
(21) Asano, N.; Oseki, K.; Kizu, H.; Matsui, K. J. Med. Chem. 1994,
37, 3701−3706.
(22) Heiker, F.-R.; Schueller, A. M. Carbohydr. Res. 1990, 203, 314−
318.
(23) Takahata, H.; Banba, Y.; Ouchi, H.; Nemoto, H.; Kato, A.;
Adachi, I. J. Org. Chem. 2003, 68, 3603−3607.
(24) van den Berg, R. J. B. H. N.; Wennekes, T.; Ghisaidoobe, A.;
Donker-Koopman, W. E.; Strijland, A.; Boot, R. G.; van der Marel, G.
L
dx.doi.org/10.1021/jo401512h | J. Org. Chem. XXXX, XXX, XXX−XXX