Regio- and stereoselective ring openings of 3-aza-2-oxabicyclo[2.2.1]hept-5-ene systems with copper catalyst-modified grignard reagents: Application to the synthesis of an inhibitor of 5-lipoxygenase

Article abstract of DOI:10.1021/jo016275u

Treatment of acylnitroso hetero Diels-Alder cycloadducts 2 with organomagnesium reagents in the presence of a catalytic amount of copper induces ring opening to afford predominantly monocyclic anti-1,2-hydroxamic acids 12. Alkylmagnesium reagents were found to give superior regio- and stereoselectivities compared with vinyl and arylmagnesium reagents. This cycloadduct ring opening methodology was applied to the synthesis of a unique cyclopentenyl hydroxamic acid-based inhibitor of 5-lipoxygenase.

Full text of DOI:10.1021/jo016275u

J . Org. Chem. 2002, 67, 4115-4121
4115
Regio- a n d Ster eoselective Rin g Op en in gs of
3-Aza -2-oxa bicyclo[2.2.1]h ep t-5-en e System s w ith Cop p er
Ca ta lyst-Mod ified Gr ign a r d Rea gen ts: Ap p lica tion to th e
Syn th esis of a n In h ibitor of 5-Lip oxygen a se
Matthew D. Surman, Mark J . Mulvihill, and Marvin J . Miller*
Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, Indiana 46556
Marvin.J .Miller.2@nd.edu
Received November 8, 2001
Treatment of acylnitroso hetero Diels-Alder cycloadducts 2 with organomagnesium reagents in
the presence of a catalytic amount of copper induces ring opening to afford predominantly monocyclic
anti-1,2-hydroxamic acids 12. Alkylmagnesium reagents were found to give superior regio- and
stereoselectivities compared with vinyl and arylmagnesium reagents. This cycloadduct ring opening
methodology was applied to the synthesis of a unique cyclopentenyl hydroxamic acid-based inhibitor
of 5-lipoxygenase.
In tr od u ction
In our investigation of cycloadduct ring opening reac-
tions through C-O bond cleavage, we desired to induce
ring opening in conjunction with C-C bond formation.
In this way, the carbon framework of the molecules could
be expanded directly from the cycloadduct. Thus, the
synthetic versatility of the cycloadducts would be greatly
enhanced. One method that proved to be successful for
this transformation involved the treatment of cycload-
ducts with palladium(0) and a carbon nucleophile such
as dimethyl malonate9a or methyl nitroacetate.9c Another
intriguing possibility was the use of organometallic
carbon nucleophiles, such as Grignard reagents. Initial
inspection of the N-acetyl cycloadducts might arouse a
certain level of skepticism, however, as the cycloadducts
are, in fact, bicyclic forms of Weinreb amides.10 Therefore,
3-Aza-2-oxabicyclo[2.2.1]hept-5-enes (2), derived from
the hetero Diels-Alder reaction between transient acylni-
troso species and cyclopentadiene,1 continue to hold a
tremendous amount of synthetic potential as intermedi-
ates in the construction of a variety of biologically
interesting molecules. Most commonly, the acylnitroso
Diels-Alder cycloadducts (2) are elaborated through
cleavage of the N-O bond to form 1,4-aminocyclopen-
tenols (3) (Scheme 1). These amino alcohols are highly
valuable intermediates in the synthesis of carbocyclic
nucleosides,2 prostaglandins,3 and other natural prod-
ucts.4 Alternatively, the cycloadducts (2) can be opened
through oxidative cleavage of the olefin to form diacids
(4),5 which can be carried on to give novel amino acids.
A less commonly employed approach to the elaboration
of acylnitroso Diels-Alder cycloadducts involves the
cleavage of the C-O bond. This approach is of particular
interest because it generates a hydroxamic acid, which
is an important functionality found in a wide range of
biologically active compounds.6 We and others have
shown that this ring opening can be induced by Brønsted
acid,7 Lewis acids,7c,8 or palladium(0)9,7c to selectively
provide syn-1,2-, anti-1,4-, or syn-1,4-disubstituted cyclo-
pentenes (5-7), respectively.
(6) For accounts of the synthesis and biological importance of related
N-hydroxy amino acids, see: (a) Ottenheijm, H. C. J .; Herscheid, J .
D. M. Chem. Rev. 1986, 86, 697. (b) Chimiak, A.; Mitewska, M. J . Prog.
Chem. Org. Nat. Prod. 1988, 53, 203. (c) Coutts, R. T. Can. J . Pharm.
Sci. 1967, 27, 1. (d) Kehl, H., Ed. Chemistry and Biology of Hydroxamic
Acids; Karger: Basel, 1982. As antibacterial agents, see: (e) Miller,
M. J . Acc. Chem. Res. 1986, 19, 49. (f) Neilands, J . B.; Valenta, J . R.
In Metal Ions in Biological Systems; Sigel, H., Ed.; Marcel Dekker:
New York, 1985; Vol. 19, Chapter 11. (g) Rogers, H. J . In Iron Transport
in Microbes, Plants and Animals; Winkelmann, G., van der Helm, D.,
Neilands, J . B., Eds.; VCH: Weinheim, New York, 1987; Chapter 13.
(h) Griffiths, E. In Iron and Infection; Bullen, J . J ., Griffiths, E., Eds.;
Wiley: New York, 1987; Chapter 3. As antifungal agents, see: (i)
Kaczka, E. A.; Gitterman, C. O.; Dulaney, E. L.; Falkers, K. Biochem-
istry 1962, 1, 340. (j) Young, C. W.; Schachetman, C. S.; Hodas, S.;
Bolis, M. C. Cancer Res. 1967, 27, 535. As anticancer agents, see: (k)
Neilands, J . B. Bacteriol. Rev. 1957, 21, 101. As specific enzyme
inhibitors, see: (l) Umezawa, H.; Aoyagi, T.; Ogawa, K.; Obata, T.;
Iinuma, H.; Naganawa, H.; Hamada, M.; Takekuchi, T. J . Antibiot.
1985, 38, 1815. (m) Stewart, A. O.; Martin, J . G. J . Org. Chem. 1989,
54, 1221. (n) Staszak, M. A.; Doecke, C. W. Tetrahedron Lett. 1994,
35, 6021. (o) Currid, P.; Wightman, R. H. Nucleosides Nucleotides 1997,
16, 115. As constituents in siderophores, see: (p) Miller, M. J . Chem.
Rev. 1989, 89, 1563. (q) Roosenberg, J . M.; Lin, Y. M.; Lu, Y.; Miller,
M. J . Curr. Med. Chem. 2000, 7, 159.
(1) (a) Vogt, P. F.; Miller, M. J . Tetrahedron 1998, 54, 1317. (b)
Kirby, G. W. Chem. Soc. Rev. 1977, 6, 1. (c) Kirby, G. W.; Nazeer, M.
Tetrahedron Lett. 1988, 29, 6173. (d) Miller, A.; McC. Patterson, T.;
Procter, G. Synlett 1989, 1, 32. (e) Miller, A.; Procter, G. Tetrahedron
Lett. 1990, 30, 1041. (f) Mulvihill, M. J .; Gage, J . L.; Miller, M. J . J .
Org. Chem. 1998, 63, 3357.
(2) (a) Ritter, A. R.; Miller, M. J . J . Org. Chem. 1994, 59, 4602. (b)
Zhang, D.; Miller, M. J . J . Org. Chem. 1998, 63, 755. (c) Trost, B. M.;
Van Vranken, D. L. J . Am. Chem. Soc. 1993, 115, 444. (d) King, S. B.;
Ganem, B. J . Am. Chem. Soc. 1991, 113, 5089. (e) Ledford, B. E.;
Carreira, E. M. J . Am. Chem. Soc. 1995, 117, 11811. (f) Shireman, B.
T.; Miller, M. J . Tetrahedron Lett. 2000, 41, 9537.
(3) Lin, C.-C.; Wang, Y.-C.; Hsu, J .-L.; Chiang, C.-C.; Su, D.-W.; Yan,
T.-H. J . Org. Chem. 1997, 62, 3806.
(4) Nakagawa, H.; Sugahara, T.; Ogasawara, K. Org. Lett. 2000, 2,
3181.
(5) (a) Ritter, A. R.; Miller, M. J . Tetrahedron Lett. 1994, 35, 9379.
(b) Heinz, L. J .; Lunn, W. H. W.; Murff, R. E.; Paschal, J . W.; Spangle,
L. A. J . Org. Chem. 1996, 61, 4838. (c) Shireman, B. T.; Miller, M. J .
J . Org. Chem. 2001, 66, 4809.
(7) (a) Muxworthy, J . P.; Wilkinson, J . A.; Procter, G. Tetrahedron
Lett. 1995, 36, 7535. (b) Muxworthy, J . P.; Wilkinson, J . A.; Procter,
G. Tetrahedron Lett. 1995, 36, 7541. (c) Mulvihill, M. J .; Surman, M.
D.; Miller, M. J . J . Org. Chem. 1998, 63, 4874.
(8) Surman, M. D.; Miller, M. J . J . Org. Chem. 2001, 66, 2466.
(9) (a) Miller, A.; Procter, G.; Tetrahedron Lett. 1990, 31, 1043. (b)
Cowart, M.; Bennett, M. J .; Kerwin, J . F., J r. J . Org. Chem. 1999, 64,
2240. (c) Surman, M. D.; Miller, M. J . Org. Lett. 2001, 3, 519.
(10) Hahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815.
10.1021/jo016275u CCC: $22.00 © 2002 American Chemical Society
Published on Web 05/14/2002

4116 J . Org. Chem., Vol. 67, No. 12, 2002
Surman et al.
Sch em e 1
Sch em e 2
Sch em e 3
one might expect the nucleophile to attack the carbonyl
carbon, giving rise to a ketone (8) after acidic workup
(Scheme 2, path A). There are, however, several other
potential electrophilic sites on the cycloadducts. One
could envision direct nucleophilic displacement of either
the oxygen (path B) or nitrogen (path B′) to give rise to
anti-1,4-disubstituted hydroxamic acids (10) or hydrox-
amates (11). Alternatively, an SN2′ attack might also be
possible to indirectly displace either the oxygen (path C)
or nitrogen (path C′) and give rise to anti-1,2-products
(12 and 13). In order for the desired transformation to
be synthetically useful, two criteria must be met. First,
one of the many possible reaction pathways must be
favored over the others, and second, that pathway must
not be path A.
First, of all the possible reaction pathways, only two
products were isolated, and second, none of the products
arising from attack at the carbonyl were detected.
Therefore, attention was turned to optimization of the
reaction conditions. Copper is often used to affect the
reactivity of organomagnesium reagents. Thus, the reac-
tion was repeated with the addition of a catalytic amount
(10 mol %) of copper(II) chloride.12,13 A dramatic effect
on the yield and selectivity of the reaction was observed,
providing an 89% yield of a 7:3:1 ratio of anti-1,2-:anti-
1,4-:syn-1,4-hydroxamic acids (14a :15a :16a ) (Table 1,
entry 2).14 The major anti-1,2-product could be isolated
cleanly by silica gel chromatography. Therefore, through
the addition of copper, not only was the yield significantly
increased, the reaction pathway leading to the separable
anti-1,2-product 14a also was shown to be favored. Thus,
according to the original criteria, the reaction could be
considered a synthetically useful route to anti-1,2-cyclo-
pentenyl hydroxamic acids (14).
Resu lts a n d Discu ssion
Our investigation began with the treatment of N-acetyl
cycloadduct 2a 11 with vinylmagnesium bromide. A low
yield (11%) of an essentially 1:1 mixture of anti-1,2-:anti-
1,4-hydroxamic acids (14a :15a ) was obtained (Table 1,
entry 1). While the yield and selectivity were not par-
ticularly impressive, two encouraging points were noted.
While encouraged by the result of the addition of
copper(II), questions remained as to the role of the copper
in the reaction. Our previous studies of Lewis acid-
mediated ring openings of acylnitroso Diels-Alder cyclo-
(12) Mu¨ller, P.; Nury, P. Org. Lett. 1999, 1, 439.
(11) For the synthesis of cycloadducts 2a , 2c, and 2d , see: Mulvihill,
M. J .; Gage, J . L.; Miller, M. J . J . Org. Chem. 1998, 63, 4874. For the
synthesis of cycloadduct 2b, see: Mulvihill, M. J .; Surman, M. D.;
Miller, M. J . J . Org. Chem. 1998, 63, 4874.
(13) (a) Ito, M.; Matsuumi, M.; Murugesh, M. G.; Kobayashi, Y. J .
Org. Chem. 2001, 66, 5881. (b) For related Cu-mediated SN2′ reactions,
see: Curran, D. P.; Chen, M.-H.; Leszczweski, D.; Elliott, R. L.;
Rakiewicz, D. M. J . Org. Chem. 1986, 51, 1612 and references therein.

Ring Openings of 3-Aza-2-oxabicyclo[2.2.1]hept-5-enes
J . Org. Chem., Vol. 67, No. 12, 2002 4117
Ta ble 1. Rin g Op en in gs of Cycloa d d u ct 2 w ith Or ga n om eta llic Rea gen ts
entry
R1
Nuc
R2
metal
none
CuCl2
FeCl3
Ga(acac)3
CuCl
none
products
yield, %
product ratios, 14:15:16
1:1:trace
7:3:1
complex mixture
complex mixture
3:3:1
2:1:0
1
2
3
4
5
6
7
8
CH3
CH3
CH3
CH3
vinylMgBr
vinylMgBr
vinylMgBr
vinylMgBr
vinylMgBr
Bu2CuLi
PhMgBr
PhMgBr
vinylMgBr
EtMgBr
CH2CH
CH2CH
CH2CH
CH2CH
CH2CH
CH3(CH2)3
C6H5
14a -16a
14a -16a
14a -16a
14a -16a
14a -16a
14b-16b
14c-16c
14c-16c
14d -16d
14e-16e
14e-16e
14f-16f
14e-16e
11
89
-
-
CH3
77
73
-
PhCH2
PhCH2
PhCH2
t-BuO
t-BuO
t-BuO
t-BuO
t-BuO
none
no hydroxamic acid products
C6H5
CuCl2
CuCl2
CuCl2
none
CuCl2
MgBr2‚OEt2
96
71
93
87
50
89
1.9:1.6:1
8.4:3:1
18:2:1
38:2:1
3.5:1:0
5.7:3:1
9
CH2CH
CH3CH2
CH3CH2
C6H5
10
11
12
13
EtMgBr
PhMgBr
EtMgBr
CH3CH2
Ta ble 2. Rin g Op en in gs of Cycloa d d u ct 2 w ith Lew is Acid s
entry
R1
Nuc
R2
Lewis acid
products
yield, %
product ratios, 14:15:16
1
2
3
4
5
CH2Ph
CH2Ph
CH2Ph
t-BuO
t-BuO
MeOH
MeOH
MeOH
MeOH
MeOH
OCH3
OCH3
OCH3
OCH3
OCH3
CuCl2
FeCl3
Ga(acac)3
CuCl2
FeCl3
14g-16g
14g-16g
14g-16g
14g-16g
14g-16g
78
74
-
41
72
1:14:5
1:7:1.7
no reaction
1:2.4:1.4
1:1.8:1.2
adducts (2) showed that the cycloadducts were suscep-
tible to ring opening with copper(II) chloride (Table 2,
entry 1).8 Therefore, the copper could have acted as a
Lewis acid and assisted in the opening of the cyclo-
adducts. To test this hypothesis, reactions were run in
which the copper(II) chloride was replaced with other
Lewis acids. Iron(III) chloride and gallium(III) acetyl-
acetonate were chosen as being toward opposite ends of
the spectrum of reactivity with acylnitroso Diels-Alder
cycloadducts. In alcohol solvent, iron(III) chloride affords
a significantly more facile ring opening than copper(II)
chloride and gallium(III) acetylacetonate is essentially
ineffective at inducing any cycloadduct ring opening
(Table 2, entries 2 and 3). When N-acetyl cycloadduct 2a
was treated with vinylmagnesium bromide and each of
the alternative Lewis acids, the reactions were unsuc-
cessful, giving rise to complex mixtures of products (Table
1, entries 3 and 4). From this brief study, the copper(II)
chloride did not appear to act as a simple Lewis acid.
Another possible role of the copper was in a reaction with
the vinylmagnesium bromide to form an organocopper
reagent. If this was the case, then a copper(I) salt should
also effectively catalyze the ring opening reaction. Indeed,
treatment of N-acetyl cycloadduct 2a with vinylmagne-
sium bromide and copper(I) chloride (10 mol %) did
provide hydroxamic acid products in good yield (77%)
(Table 1, entry 5). However, the selectivity of the reaction
suffered significantly, giving a 3:3:1 ratio of anti-1,2-:anti-
1,4-:syn-1,4-hydroxamic acids (14a :15a :16a ). These stud-
ies indicated that the role of the copper(II) chloride was
a bit more complex than just a simple Lewis acid or in
the catalytic generation of the organocopper reagent.
However, further evidence of an organocopper reagent
being responsible for the cycloadduct ring opening came
from the treatment of N-phenylacetyl cycloadduct 2b
with preformed dibutylhomocuprate. The reaction re-
sulted in a 73% yield of a 2:1 mixture of anti-1,2-:anti-
1,4-hydroxamic acids (14b:15b) (Table 1, entry 6).
Aryl Grignard reagents were found to offer poorer
product selectivity than vinylmagnesium bromide. Not
surprisingly, when N-phenylacetyl cycloadduct 2b was
initially reacted with phenylmagnesium bromide in the
absence of any copper salts, no hydroxamic acid products
were formed (Table 1, entry 7). Once again, however, the
addition of a catalytic amount of copper(II) chloride had
a dramatic effect on the outcome of the reaction. When
copper was added, the reaction provided a 96% yield of
1.9:1.6:1 ratio of anti-1,2-:anti-1,4:syn-1,4-hydroxamic
acids (14c:15c:16c) (Table 1, entry 8).
The copper-catalyzed reactions of N-carbamate cyclo-
adducts with organomagnesium reagents also were in-
vestigated. Previous studies showed that Lewis acid-
mediated ring openings of N-carbamate cycloadducts
provided compromised product selectivities in comparison
with N-acetyl cycloadducts (Table 2, entries 4 and 5).8
Therefore, we were pleased to find that reaction of N-Boc
cycloadduct 2c with vinylmagnesium bromide and cop-
per(II) chloride gave a comparable yield (71%) and
selectivity [8.4:3:1 ratio of anti-1,2-:anti-1,4-:syn-1,4-
hydroxamic acids (14d :15d :16d )] to the treatment of
N-acetyl cycloadduct under the same reaction conditions
(Table 1, entry 9). This improved selectivity might be due
to the decreased leaving group abilities of the N-hydroxy-
carbamate relative to the hydroxamate. Improved yields
(14) The syn-1,4- and anti-1,4-stereochemical assignments were
based upon the 1H NMR coupling pattern. The C(5) methylene protons
of 1,4-disubstituted cyclopentene systems have a very characteristic
coupling pattern. The C(5) protons of syn-1,4-aminocyclopentenols have
a characteristic overlapping ddd pattern, with approximate J values
of 3.9, 3.9, and 14.7 Hz and 7.8, 7.8, and 14.7 Hz for the cis- and trans-
methylene protons, respectively. There is normally a difference in
chemical shift ranging from 0.8 to 1.3 ppm between each respective
C(5) proton. The C(5) protons of anti-1,4-aminocyclopentenols have a
characteristic ddd pattern, with approximate J values of 3.9, 7.1, and
13.6 Hz with a smaller chemical shift, ranging from 0.2 to 0.35 ppm
between the two C(5) protons. The anti-1,2-stereochemical assignment
was proven through the synthesis of known anti-2-methylcyclopentyl-
amine hydrochloride (ii) from cycloadduct 2c. Cycloadduct 2c was
treated with methylmagnesium bromide and a catalytic amount of
copper(II) chloride to give anti-1,2-hydroxamic acid i. Reduction of the
N-O bond followed by reduction of the cyclopentenyl olefin and
removal of the Boc group provided anti-2-methylcyclopentylamine
hydrochloride (ii). The melting point of ii (196-197 °C) was consistent
with the literature value (196-198 °C).15 The melting point of syn-2-
methylcyclopentylamine hydrochloride is 256-257 °C.16

4118 J . Org. Chem., Vol. 67, No. 12, 2002
Surman et al.
Ta ble 3. 5-LO In h ibitor y Activity of Com p ou n d s 18 a n d
19a
and selectivities could be achieved by substituting the
vinylmagnesium bromide with an alkylmagnesium re-
agent. Treatment of N-Boc cycloadduct 2c with ethyl-
magnesium bromide resulted in a 93% yield of an 18:2:1
ratio of anti-1,2-:anti-1,4-:syn-1,4-hydroxamic acids (14e:
15e:16e) (Table 1, entry 10). Interestingly, when the
N-carbamate cycloadducts were used, the use of catalytic
copper was not always necessary. For example, when
N-Boc cycloadduct 2c was treated with ethylmagnesium
bromide in the absence of any copper salts, a 38:2:1 ratio
of anti-1,2-:anti-1,4-:syn-1,4-hydroxamic acids (14e:15e:
16e) was obtained in 87% yield (Table 1, entry 11).
Furthermore, the reactions of N-carbamate cycloadducts
often appeared to be influenced more by the Lewis acidity
of the reaction media than by the specific presence of
copper(II). When N-Boc cycloadduct 2c was treated with
ethylmagnesium bromide and 0.5 equiv of MgBr2‚OEt2,
a reduction in selectivity was observed. The reaction gave
an 89% yield of a 5.7:2:1 ratio of anti-1,2-:anti-1,4-:syn-
1,4-hydroxamic acids (14e:15e:16e) (Table 1, entry 13).
As seen previously with N-phenylacetyl cycloadduct 2b,
reaction of N-Boc cycloadduct 2c with arylmagnesium
reagents resulted in compromised selectivity. Treatment
of N-Boc cycloadduct 2c with phenylmagnesium bromide
and copper(II) chloride resulted in a 50% yield of a 3.5:1
ratio of anti-1,2-:anti-1,4-hydroxamic acids (14f:15f) (Table
1, entry 12).
The results of these studies showed that acylnitroso
Diels-Alder cycloadducts 2 could be opened with orga-
nomagnesium reagents in the presence of a catalytic
amount of copper to provide hydroxamic acids in up to
93% yield and with selectivities of up to 18:2:1 for the anti-
1,2-:anti-1,4-:syn-1,4-products (14:15:16). This constitutes
a novel and synthetically useful cycloadduct ring opening
and allows for the direct elaboration of the cyclopentene
carbon framework while regenerating a hydroxamic acid
moiety.
As mentioned previously, hydroxamic acids represent
an important functionality found in a variety of biologi-
cally active compounds.6 The activity of these compounds
often relies on the hydroxamic acid’s ability to effectively
bind metals such as iron(III), nickel(II), and zinc(II).17
Thus, hydroxamates often act as inhibitors of metal-
containing enzymes such as 5-lipoxygenase.18 5-Lipoxy-
genase (5-LO) is an iron(III) containing enzyme that is
the first in a cascade of enzymes that metabolize arachi-
donic acid (17) into inflammatory-mediating leukotrienes.
Elevated levels of leukotrienes have been associated with
a variety of disease states including asthma, rheumatoid
arthritis, inflammatory bowel disease, psoriasis, and
allergy. Compounds that inhibit 5-lipoxygenase show
promise in treating such inflammatory diseases.19 Many
known inhibitors of 5-LO contain hydroxamic acids,
which are thought to bind to the iron(III) within the
enzyme. These inhibitors also typically contain an aro-
matic portion, which is thought to interact with the
IC50
nM
,
entry compd
5-LO source
1
2
18
19a
rat polymorphonuclear leukocytes
human peripheral blood mononuclear cells 51
46
hydrophobic regions of the enzyme. An example of a
potent class of such 5-LO inhibitors is the 3,4-dihydro-
naphthyl hydroxamic acids.20 Extensive SAR studies
revealed that a phenoxy group at the 5 position of the
3,4-dihydronaphthalene ring produced the most potent
compounds.
We envisioned using our newly developed acylnitroso
Diels-Alder cycloadduct ring opening methodology with
copper-catalyzed organomagnesium reagents to produce
analogues of known 5-LO inhibitors. Importantly, these
analogues would be accessible in a single synthetic
transformation from the cycloadduct. To demonstrate our
vision, we designed an analogue of potent inhibitor
5-phenoxy-3,4-dihydronaphthyl hydroxamic acid 18 (IC50
) 46 nM against 5-LO in intact rat polymorphonuclear
leukocyts, Table 3, entry 1). Figure 1 shows how hydrox-
amic acid 18 was thought to sit in the active site of the
5-LO enzyme in alignment with the natural substrate,
arachidonic acid (17). While the hydroxamic acid moiety
was bound to the iron(III), the aromatic portions of 18
were proposed to overlap with the sites of the enzyme
that interact with the C5-C15 portions of arachidonic
acid (17). Similarly, proposed analogue 19a contains
hydroxamic acid and aromatic portions that could inter-
act with the 5-lipoxygenase enzyme in a comparable
fashion.
Formation of analogue 19 was anticipated to involve
treatment of N-methyl carbamate cycloadduct 2d with
an appropriate organomagnesium reagent (Scheme 4).
The required bromide precursor (21) was obtained after
treatment of 3-phenoxybenzyl alcohol (20) with carbon
tetrabromide and triphenylphosphine.21 Accordingly, bro-
mide 21 was converted to the corresponding organomag-
nesium reagent and used to open cycloadduct 2d in the
presence of a catalytic amount of copper(II) chloride.
Initial attempts using THF as the solvent were unsuc-
cessful, leading only to decomposition and polymerization
products. However, when the solvent was changed to
diethyl ether, a 60% yield of a 8.8:4:1 ratio of anti-1,2-:
anti-1,4-:syn-1,4-hydroxamic acids (19a :19b:19c) was
obtained.
The desired analogue 19a was isolated and submitted
for testing against human 5-LO enzyme.22 The test
results indicated complete (100%) enzyme inhibition at
the testing concentrations of 10, 5, and 1 µM and an IC50
value of 51 nM (Table 3, entry 2). Thus, compound 19a
was shown to be a potent inhibitor of 5-LO with compa-
rable activity to known inhibitor 18. One could envision
rapidly generating libraries of such cyclopentenyl hy-
droxamic acid-based 5-LO inhibitors from acylnitroso
Diels-Alder adducts and a variety of aromatic Grignard
reagents.
(15) Brown, H. C.; Kim, K.-W.; Cole, T. E.; Singaram, B. J . Am.
Chem. Soc. 1986, 108, 6761.
(16) Wiehl, W.; Frahm, A. W. Chem. Ber. 1986, 119, 2668.
(17) (a) Dixon, N. E.; Hinds, J . A.; Fihelly, A. K.; Gazzola, C.; Winzor,
D. J .; Blakeley, R. L.; Zerner, B. Can. J . Biochem. 1980, 58, 1323. (b)
Nagarajan, K.; Fishbein, W. N. Fed. Proc. 1977, 36, 700. (c) Michaelides,
M. R.; Curtin, M. L. Curr. Pharm. Des. 1999, 5, 787.
Con clu sion s
Acylnitroso Diels-Alder cycloadducts 2 were shown to
be susceptible to ring opening with organomagnesium
(18) (a) Corey, E. J .; Cashman, J . R.; Kanter, S. S.; Corey, D. R. J .
Am. Chem. Soc. 1984, 106, 1503. (b) Summers, J . B.; Gunn, B. P.;
Mazdiyasni, H.; Goetze, A. M.; Young, P. R.; Bouska, J . B.; Dyer, R.
D.; Brooks, D. W.; Carter, G. W. J . Med. Chem. 1987, 30, 2121.
(19) Young, R. N. Eur. J . Med. Chem. 1999, 34, 671.
(20) Yatabe, T.; Kawai, Y.; Oku, T.; Tanaka, H. Chem. Pharm. Bull.
1998, 46, 966.
(21) Mandal, A. K.; Mahajan, S. W. Tetrahedron Lett. 1985, 26, 3863.
(22) Biological testing was performed by MDS Pharma Services.

Ring Openings of 3-Aza-2-oxabicyclo[2.2.1]hept-5-enes
J . Org. Chem., Vol. 67, No. 12, 2002 4119
F igu r e 1.
1,2-product (14a ) as a white solid: mp 87-88 °C; IR (TF) 3127,
Sch em e 4
1
2764, 2675, 1562, 1474, 1421, 1242, 1187, 933, 725; H NMR
(300 MHz, DMSO-d6) δ 1.98 (s, 3H), 2.43 (d, J ) 6.9 Hz, 2H),
3.45 (m, 1H), 4.88 (m, 1H), 4.96 (d, J ) 9.6 Hz, 1H), 5.05 (d, J
) 17.4 Hz, 1H), 5.58 (d, J ) 4.5 Hz, 1H), 5.72 (m, 1H), 5.79
(dd, J ) 7.2, 9.9 Hz, 1H), 9.58 (s, 1H); 1H NMR (500 MHz,
DMSO-d6) δ 1.99 (s, 3H), 2.43 (bs, 2H), 3.44 (m, 1H), 4.88 (dm,
J ) 7.5 Hz, 1H), 4.96 (d, J ) 9.5 Hz, 1H), 5.05 (d, J ) 17.0 Hz,
1H), 5.58 (dm, J ) 4.5 Hz, 1H), 5.72 (m, 1H), 5.78 (m, 1H),
9.60 (s, 1H); 13C NMR (125 MHz, DMSO-d6) δ 20.77, 34.04,
50.59, 59.63, 114.67, 129.43, 131.80, 139.87, 170.50; HRMS
(FAB) calcd for C9H13NO2 (M + H)+ 168.1025, found 168.1036.
1
An ti-1,4-p r od u ct (15a ): H NMR (300 MHz, DMSO-d6) δ
1.75 (ddd, J ) 4.8, 8.7, 13.2 Hz, 1H), 1.96 (s, 1H), 2.04 (ddd, J
) 4.8, 8.4, 13.2 Hz, 1H), 3.44 (bm, 1H), 4.92 (dm, J ) 9.9 Hz,
1H), 5.03 (dm, J ) 17.1 Hz, 1H), 5.50 (bs, 1H), 5.59 (m, 1H),
5.71 (m, 1H), 5.86 (dm, J ) 5.4 Hz, 1H), 9.39 (s, 1H
reagents in the presence of a catalytic amount of copper
to provide hydroxamic acids in up to 93% yield and with
selectivities of up to 18:2:1 for the anti-1,2-:anti-1,4-:syn-
1,4-products (14:15:16). This novel ring opening meth-
odology was shown to have synthetic utility in the
construction of a unique cyclopentenyl hydroxamic acid
that demonstrated potent 5-LO inhibitory activity.
1,2-a n ti-Hyd r oxa m ic Acid (14b). A solution of butyl-
lithium (0.43 mL, 2.3 M) in THF was added over the course of
10 min to a suspension of CuI (94 mg, 0.491 mmol) in THF (3
mL) at 0 °C. The resulting dark mixture was stirred for 30
min and then treated with a solution of N-phenylacetyl
cycloadduct 2b11 (105 mg, 0.489 mmol) in THF (3 mL). After
the mixture was stirred for 30 min at 0 °C, the ice bath was
removed, and the mixture was stirred at room temperature
for 45 min. The reaction mixture was partitioned between a
saturated solution of NH4Cl and ethyl acetate, and the organic
phase was washed with the NH4Cl solution, a saturated
solution of Na2EDTA, and brine. The combined aqueous layers
were extracted with ethyl acetate, and the combined organics
were dried with Na2SO4 and concentrated under reduced
pressure. Flash chromatography (silica gel; eluted with hex-
anes-ethyl acetate, 2:1) provided 97 mg (73%) of a 2:1 mixture
of hydroxamic acids 14b:15b as an oil. Chromatography (silica
gel; eluted with 1% AcOH, 1%MeOH-CH2Cl2) provided anti-
1,2-product (14b) as an oil: IR (TF) 3176, 2960, 2921, 2862,
1597, 1494, 1450, 1268, 1234, 1175, 1032 cm-1; 1H NMR (300
MHz, DMSO-d6) δ 0.81 (t, J ) 6.3 Hz, 3H), 1.23 (m, 6H), 2.41
(bd, J ) 6.9 Hz, 2H), 2.77 (m, 1H), 3.66 (d, J ) 14.7 Hz, 1H),
3.76 (d, J ) 15 Hz, 1H), 4.75 (dd, J ) 7.2 Hz, 1H), 5.63 (s,
2H), 7.25 (m, 5H), 9.64 (s, 1H); 13C NMR (75 MHz, DMSO-d6)
δ 13.87, 22.25, 28.93, 33.44, 34.31, 38.80, 46.93, 59.65, 126.16,
128.03, 128.10, 129.30, 133.54, 135.94, 170.66; HRMS (FAB)
calcd for C17H24NO2 (M + H)+ 274.1808, found 274.1827.
Exp er im en ta l Section
Gen er a l Meth od s. Melting points were taken on a Thomas-
Hoover capillary melting point apparatus and are uncorrected.
1H NMR and 13C NMR spectra were obtained on a Varian 300
spectrometer and were referenced to residual DMSO or CDCl3.
Infrared spectra were recorded on a Perkin-Elmer Paragon
1000 FT-IR spectrometer, and TF refers to thin film. Analytical
TLC was carried out using Merck aluminum-backed 0.2 mm
silica gel 60 F-254 plates. Column chromatography was
conducted using Merck silica gel 60 (230-400) mesh.
All reactions were periodically monitored by TLC and
worked up after the complete consumption of starting materi-
als unless specified otherwise. Anhydrous tetrahydrofuran was
freshly distilled from sodium and stored under argon. All
purchased reagents were of reagent grade quality and were
used without further purification.
All product ratios were determined by 1H NMR spectroscopy
of the crude reaction mixtures. The major products (14a , 14b,
14c, 14e, 21, and 19) were isolated and fully characterized as
follows. The minor products (15a , 15b, 15c, 16c, and 19b) and
major products 14d and 14f were characterized from the crude
reaction mixtures.
1,2-a n ti-Hyd r oxa m ic Acid (14a ). A solution of N-acetyl
cycloadduct 2a 11 (112 mg, 0.803 mmol) in THF (2 mL) was
cooled to 0 °C in an ice bath and treated with CuCl2 (11 mg,
0.0833 mmol). The resulting mixture was slowly treated with
a 1 M solution of vinylmagnesium bromide (1.0 mL, 1.0 mmol)
in THF over the course of 15 min and stirred under argon for
1 h. The mixture was diluted with EtOAc and washed with 1
M HCl. The aqueous layer was extracted with EtOAc, and the
combined organic layers were washed with a saturated solu-
tion of Na2EDTA, dried over Na2SO4, filtered, and concentrated
under reduced pressure. Flash chromatography (silica gel;
eluted with 3% MeOH-CH2Cl2) provided 120 mg (89%) of a
7:3:1 mixture of hydroxamic acids 14a :15a :16a as an orange
solid. Recrystallization from EtOAc and hexanes provided anti-
1
An ti-1,4-p r od u ct (15b): H NMR (300 MHz, DMSO-d6) δ
0.86 (t, J ) 6.6 Hz, 1H), 1.26 (m, 6H), 1.59 (m, 1H), 1.97 (ddd,
J ) 4.2, 8.1, 13.2 Hz, 1H), 2.76 (bs, 1H), 3.68 (s, 2H), 5.49 (m,
2H), 5.94 (m, 1H), 7.24 (m, 5H), 9.46 (s, 1H).). HRMS (FAB)
calcd for C17H23NO2 (M + H)+ 274.1807, found 274.1807.
1,2-a n ti-Hyd r oxa m ic Acid (14c). A solution of N-phenyl-
acetyl cycloadduct 2b11 (106 mg, 0.493 mmol) in THF (2 mL)
was cooled to 0 °C in an ice bath and treated with CuCl2 (7
mg, 0.0528 mmol). The resulting mixture was slowly treated
with a 1 M solution of phenylmagnesium bromide (1.0 mL,
1.0 mmol) in THF over the course of 15 min and stirred under
argon for 1 h. The mixture was diluted with EtOAc and washed
with 1 M HCl. The aqueous layer was extracted with EtOAc,
and the combined organic layers were washed with a saturated
solution of Na2EDTA, dried over Na2SO4, filtered, and con-
centrated under reduced pressure. Flash chromatography
(silica gel; eluted with 3% MeOH-CH2Cl2) provided 140 mg

4120 J . Org. Chem., Vol. 67, No. 12, 2002
Surman et al.
(96%) of a 1.9:1.6:1 mixture of hydroxamic acids 14c:15c:16c
as an oil. Chromatography (silica gel; eluted with 5% EtOAc-
CH2Cl2) followed by recrystallization from EtOAc-hexanes
provided 14c as a white solid: mp 123-124 °C; IR (TF) 3153,
3061, 3030, 2885, 1600, 1494, 1453, 1236, 1176, 1031, 699; 1H
NMR (300 MHz, DMSO-d6) δ 2.53 (m, 2H), 3.67 (d, J ) 15.3
Hz, 1H), 3.77 (d, J ) 15.9 Hz, 1H), 4.03 (m, 1H), 4.96 (dd, J )
7.5 Hz, 1H), 5.73 (dm, J ) 5.7 Hz, 1H), 5.88 (dm, J ) 6.3 Hz,
1H), 7.18 (m, 5H), 9.87 (s, 1H); 13C NMR (75 MHz, DMSO-d6)
δ 34.43, 38.82, 52.02, 63.09, 126.18, 126.29, 127.10, 128.03,
128.33, 129.31, 129.83, 132.60, 135.76, 143.55, 171.03; HRMS
(FAB) calcd for C19H20NO2 (M + H)+ 294.1494, found 294.1483.
solution of Na2EDTA, dried over Na2SO4, filtered, and con-
centrated under reduced pressure. Flash chromatography
(silica gel; eluted with hexanes-ethyl acetate, 2:1) provided
70 mg (50%) of a 3.5:1 mixture of hydroxamic acids 14f:15f as
an oil: 1H NMR (300 MHz, DMSO-d6) δ 1.16 (s, 9H), 2.53 (dm,
J ) 7.2 Hz, 2H), 3.98 (m, 1H), 4.55 (dd, J ) 8.1 Hz, 1H), 5.65
(m, 1H), 5.84 (m, 1H), 7.22 (m, 5H), 9.23 (s, 1H); HRMS (FAB)
calcd for C16H21NO3 (M + H)+ 276.15997, found 276.15933.
3-P h en oxyben zyl Br om id e (21). A solution of CBr4 (1.09
g, 3.28 mmol) in THF (17 mL) was cooled to 0 °C in an ice
bath and treated with 3-phenoxybenzyl alcohol (0.44 mL, 2.52
mmol). The resulting mixture was slowly treated with a
solution of PPh3 in THF (8 mL) over the course of 5 min and
stirred under argon at 0 °C for 1 h and at room temperature
for 20 h. The mixture was concentrated under reduced pres-
sure. Flash chromatography (silica gel; eluted with 100%
hexanes to hexanes-ethyl acetate, 1:1) provided 452 mg (68%)
of 3-phenoxybenzyl bromide (21) as an oil: IR (TF) 3039, 1584,
1487, 1445, 1258, 1220, 1162, 959, 886, 778, 692 cm-1; 1H NMR
(300 MHz, CDCl3) δ 4.46 (s, 2H), 6.96 (m, 1H), 7.05 (m, 3H),
7.16 (m, 2H), 7.35 (m, 3H); 13C NMR (75 MHz, CDCl3) δ 32.87,
118.55, 119.07, 119.13, 123.57, 123.65, 129.81, 130.05, 139.55,
156.66, 157.55; HRMS (FAB) calcd for C13H12BrO (M + H)+
168.1025, found 168.1036.
1
An ti-1,4-p r od u ct (15c): H NMR (300 MHz, DMSO-d6) δ
1.88 (ddd, J ) 5.1, 8.1, 13.5 Hz, 1H), 2.35 (ddd, J ) 4.2, 8.7,
12.9 Hz, 1H), 3.73 (s, 2H), 4.07 (bm, 1H), 5.67 (bm, 1H), 5.74
(m, 1H), 6.02 (m, 1H), 7.24 (m, 10H), 9.65 (s, 1H). HRMS (FAB)
calcd for C19H19NO2 (M + H)+ 294.1494, found 294.1489.
Syn -1,4-p r od u ct (16c): 1H NMR (300 MHz, DMSO-d6) δ
1.73 (overlapping ddd, J ) 7.8, 7.8, 12.9 Hz, 1H), 2.58
(overlapping ddd, J ) 8.1, 8.1, 12.3 Hz, 1H), 3.73 (s, 2H), 3.80
(m, 1H), 5.59 (bm, 1H), 5.74 (m, 1H), 5.90 (m, 1H), 7.24 (m,
10H), 9.67 (s, 1H).
1,2-a n ti-Hyd r oxa m ic Acid (14d ). A solution of N-Boc
cycloadduct 2c11 (140 mg, 0.711 mmol) in THF (2 mL) was
cooled to 0 °C in an ice bath and treated with CuCl2 (10 mg,
0.0758 mmol). The resulting mixture was slowly treated with
a 1 M solution of vinylmagnesium bromide (1.0 mL, 1.0 mmol)
in THF over the course of 15 min and stirred under argon for
1 h. The mixture was diluted with EtOAc and washed with 1
M HCl. The aqueous layer was extracted with EtOAc, and the
combined organic layers were washed with a saturated solu-
tion of Na2EDTA, dried over Na2SO4, filtered, and concentrated
under reduced pressure. Flash chromatography (silica gel;
eluted with 1% MeOH-CH2Cl2) provided 114 mg (71%) of a
8.4:3:1 mixture of hydroxamic acids 14d :15d :16d as an oil:
1H NMR (300 MHz, DMSO-d6) δ 1.39 (s, 9H), 2.43 (dm, J )
8.1 Hz, 2H), 3.43 (m, 1H), 4.46 (dd, J ) 8.1 Hz, 1H), 5.01 (m,
2H), 5.67 (m, 3H), 9.10 (s, 1H); 13C NMR (75 MHz, DMSO-d6)
δ 27.95, 34.02, 50.55, 63.82, 79.52, 114.68, 129.37, 131.75,
140.17, 155.78.
N-Hyd r oxyca r ba m a te 19. A mixture of freshly ground
magnesium (105 mg, 4.33 mmol) and a catalytic amount of
HgCl2 in diethyl ether (1 mL) was treated with a solution of
bromide 21 (518 mg, 1.97 mmol) in diethyl ether (3 mL) and
refluxed in an oil bath (50 °C) for 30 min. A mixture of
N-methyl carbamate cycloadduct 2d 11 (119 mg, 0.766 mmol)
and CuCl2 (11 mg, 0.084 mmol) in diethyl ether (2.5 mL) was
cooled in an ice bath and treated with the Grignard reagent
(2.0 mL, 0.983 mmol). After stirring in the ice bath for 1.5 h,
the reaction mixture was diluted with ethyl acetate and
quenched with 1 M HCl. The aqueous phase was extracted
with ethyl acetate, and the combined organics were washed
with a saturated solution of Na2EDTA, dried over Na2SO4,
filtered, and concentrated under reduced pressure. Chroma-
tography (silica gel; eluted with hexanes-ethyl acetate, 2:1)
provided 88 mg (34%) of anti-1,2-product 19a as a clear oil, as
well as 68 mg (26%) of a 1:4:1 ratio of anti-1,2-:anti-1,4-:syn-
1,4-hydroxamic acid products (19) as a clear oil. anti-1,2-
product 19a : IR (TF) 3272, 3061, 2921, 1695, 1582, 1487, 1447,
1,2-a n ti-Hyd r oxa m ic Acid (14e). A mixture of freshly
ground magnesium (253 mg, 10.4 mmol) and a catalytic
amount of HgCl2 in THF (6.7 mL) was treated with bromo-
ethane (0.5 mL, 6.70 mmol) and refluxed in an oil bath (50
°C) for 20 min. The resulting gray mixture was allowed to cool
to room temperature. A mixture of N-Boc cycloadduct 2c11 (102
mg, 0.516 mmol) and CuCl2 (7 mg, 0.084 mmol) in diethyl ether
(2 mL) was cooled in an ice bath and slowly treated with the
freshly prepared ethylmagnesium bromide (0.8 mL, 0.8 mmol)
over the course of 15 min. After stirring in the ice bath for 1
h, the reaction mixture was diluted with ethyl acetate and
quenched with 1 M HCl. The aqueous phase was extracted
with ethyl acetate, and the combined organics were washed
with a saturated solution of Na2EDTA, dried over Na2SO4,
filtered, and concentrated under reduced pressure. Chroma-
tography (silica gel; eluted with hexanes-ethyl acetate, 2:1)
provided 110 mg (93%) of an 18:2:1 ratio of hydroxamic acids
14e:15e:16e as a clear oil. Chromatography (silica gel; eluted
with 3% MeOH-CH2Cl2) provided 1,2-anti product (14e) as
an oil: IR (TF) 3225, 2966, 2932, 16.94, 1456, 1368, 1250, 1165,
1340, 1252, 1109, 946, 755, 692 cm-1 1H NMR (300 MHz,
;
DMSO-d6) δ 2.38 (m, 2H), 2.52 (dd, J ) 8.4, 13.5 Hz, 1H), 2.75
(dd, J ) 6.0, 13.5 Hz, 1H), 3.10 (m, 1H), 3.56 (s, 3H), 4.43 (m,
1H), 5.49 (dm, J ) 2.1, 6.3 Hz, 1H), 5.60 (dm, J ) 2.1, 6.3 Hz,
1H), 6.80 (dm, J ) 7.5 Hz, 1H), 6.86 (m, 1H), 6.97 (m, 3H),
7.10 (tm, J ) 7.2 Hz, 1H), 7.27 (t, J ) 7.8 Hz, 1H), 7.36 (m,
1
2H), 9.27 (s, 1H); H NMR (300 MHz, CDCl3) δ 2.50 (m, 1H),
2.60 (m, 1H), 2.62 (dd, J ) 8.7, 13.5 Hz, 1H), 2.82 (dd, J )
6.0, 13.5 Hz, 1H), 3.29 (dm, J ) 6.3 Hz, 1H), 3.68 (s, 3H), 4.47
(q, J ) 7.5 Hz, 1H), 5.58 (m, 1H), 5.67 (m, 1H), 6.83-7.35 (m,
9H); 13C NMR (75 MHz, DMSO-d6) δ 34.10, 38.78, 48.10, 52.33,
62.48, 116.30, 118.28, 119.52, 123.15, 124.38, 128.75, 129.64,
129.95, 132.45, 142.15, 156.29, 156.83, 156.90; 13C NMR (75
MHz, CDCl3) δ 34.45, 39.85, 48.51, 53.30, 63.39, 116.58, 118.62,
119.71, 123.03, 124.06, 128.66, 129.43, 129.67, 132.59, 142.09,
157.04, 157.38, 157.56; HRMS (FAB) calcd for C20H22NO4 (M
+ H)+ 340.1549, found 340.1546.
1
1110, 865, 757, 712 cm-1; H NMR (300 MHz, DMSO-d6): δ
An ti-1,4-p r od u ct 19b: 1H NMR (300 MHz, DMSO-d6) δ
1.67 (ddd, J ) 4.8, 8.7, 13.5 Hz, 1H), 1.90 (ddd, J ) 4.5, 8.1,
13.2 Hz, 1H), 2.51 (dd, J ) 7.8, 13.2 Hz, 1H), 2.64 (dd, J )
6.9, 13.2 Hz, 1H), 3.03 (m, 1H), 3.58 (s, 3H), 5.05 (m, 1H), 5.55
(m, 1H), 5.83 (m, 1H), 6.83 (m, 2H), 6.97 (dm, J ) 7.5 Hz, 3H),
7.11 (t, J ) 7.5 Hz, 1H), 7.28 (t, J ) 7.8 Hz, 1H), 7.36 (m, 2H).
5-LO In h ibition Testin g P r oced u r e.22 Human peripheral
blood mononuclear cells were isolated through a Ficoll-Paque
density gradient. Compound 19a (10 µM) was incubated with
A 23187 (30 µM)-stimulated human peripheral blood mono-
nuclear (≈5 × 106 cells/mL) in HBBS for 15 min at 37 °C.
Following neutralization with NaOH and centrifugation, the
supernatant LTB4 was measured using an EIA kit (Assay
Design Inc.).
0.87 (t, J ) 7.5 Hz, 3H), 1.34 (m, 2H), 1.40 (s, 9H), 2.40 (d, J
) 8.1 Hz, 2H), 2.71 (m, 1H), 4.34 (q, J ) 7.5 Hz, 1H), 5.60 (s,
2H), 8.98 (s, 1H); 13C NMR (75 MHz, DMSO-d6) δ 11.25, 26.05,
28.01, 34.32, 48.13, 62.57, 79.30, 128.22, 133.07, 155.54; HRMS
(FAB) calcd for C12H22NO3 (M + H)+ 228.1601, found 228.1612.
1,2-a n ti-Hyd r oxa m ic Acid (14f). A solution of N-Boc
cycloadduct 2c11 (101 mg, 0.512 mmol) in THF (2 mL) was
cooled to 0 °C in an ice bath and treated with CuCl2 (8 mg,
0.0565 mmol). The resulting mixture was slowly treated with
a 1 M solution of phenylmagnesium bromide (1.0 mL, 1.0
mmol) in THF over the course of 15 min and stirred under
argon for 1 h. The mixture was diluted with EtOAc and washed
with 1 M HCl. The aqueous layer was extracted with EtOAc,
and the combined organic layers were washed with a saturated

Ring Openings of 3-Aza-2-oxabicyclo[2.2.1]hept-5-enes
J . Org. Chem., Vol. 67, No. 12, 2002 4121
Ack n ow led gm en t. We gratefully acknowledge the
NIH for partial financial support of this research, the
University of Notre Dame for support in the form of a
J . Peter Grace fellowship, the Lizzadro Magnetic Reso-
nance Research Center at Notre Dame for NMR facili-
ties, Dr. W. Boggess and N. Sevova for mass spectrom-
etry facilities, and Maureen Metcalf for assistance with
the manuscript.
Su p p or tin g In for m a tion Ava ila ble: NMR data for prod-
ucts 14a -c, 14e, 21, and 19. This material is available free of
charge via the Internet at http://pubs.acs.org.
J O016275U