Cyclic nitrone free radical traps: Isolation, identification, and synthesis of 3,3-dimethyl-3,4-dihydroisoquinolin-4-ol N-oxide, a metabolite with reduced side effects

Article abstract of DOI:10.1021/jm960244n

A C-4 hydroxylated metabolite (2, 3,3-dimethyl-3,4-dihydroisoquinolin- 4-ol N-oxide) of the previously described cyclic nitrone free radical trap 1 (3,3-dimethyl-3,4-dihydroisoquinoline N-oxide, a cyclic analog of phenyl- tert-butylnitrone (PBN)) was isol

Full text of DOI:10.1021/jm960244n

J . Med. Chem. 1996, 39, 4997-5004
4997
Cyclic Nitr on e F r ee Ra d ica l Tr a p s: Isola tion , Id en tifica tion , a n d Syn th esis of
3,3-Dim eth yl-3,4-d ih yd r oisoqu in olin -4-ol N-Oxid e, a Meta bolite w ith Red u ced
Sid e Effects
Craig E. Thomas,*,† Patrick Bernardelli,‡ S. Marc Bowen, Stephen F. Chaney, Dirk Friedrich,
David A. J anowick,§ Bryan K. J ones, Frederick J . Keeley, J ohn H. Kehne,| Bert Ketteler, David F. Ohlweiler,
Leo A. Paquette,‡ David J . Robke, and Thomas L. Fevig
Hoechst Marion Roussel, Inc., 2110 East Galbraith Road, Cincinnati, Ohio 45215, and Evans Chemical Laboratories, The
Ohio State University, Columbus, Ohio 43210
Received March 28, 1996X
A C-4 hydroxylated metabolite (2, 3,3-dimethyl-3,4-dihydroisoquinolin-4-ol N-oxide) of the
previously described cyclic nitrone free radical trap 1 (3,3-dimethyl-3,4-dihydroisoquinoline
N-oxide, a cyclic analog of phenyl-tert-butylnitrone (PBN)) was isolated, identified, and
synthesized. The metabolite (2), though a less potent antioxidant than 1 in an in vitro lipid
peroxidation assay, showed greatly reduced acute toxicity and sedative properties. Several
analogs of 2 were prepared in attempts to improve on its weak antioxidant activity while
retaining the desirable side effect profile. Effective structural changes included replacement
of the gem-dimethyl moiety with spirocycloalkane groups and/or oxidation of the alcohols to
the corresponding ketones. All of the analogs were more lipophilic (log k′w) and more active in
the standard lipid peroxidation assay than 2. In addition, some of the compounds were able
to protect cerebellar granule cells against oxidative damage (an in vitro model of oxidative
brain injury) with IC50 values well below the value of the lead compound 1. The ketones, as
predicted, were much more potent than 2 (and 1) in both of the above assays (up to ca. 200-
fold). However, only compounds with a hydroxyl or an acetate group at C-4 displayed
significantly reduced acute toxicity and sedative properties relative to those of 1. Importantly,
the diminishment of toxicity and sedation were not the result of a lack of brain penetration as
both 2 and the corresponding ketone (3,3-dimethyl-3,4-dihydro-3H-isoquinolin-4-one N-oxide)
achieved equal or greater brain levels than those of 1 when administered to rats ip.
In tr od u ction
Free radical mediated oxidation of cellular macromol-
ecules (lipids, proteins, DNA, etc.) has been implicated
in a number of disease states, including stroke and head
trauma.1 For example, during stroke, biochemical
F igu r e 1.
changes arising from ischemia alter the cell such that
compounds by adding to the nitrone double bond. The
subsequent reperfusion induces a large flux of reactive
resulting products, though still free radicals (nitroxides),
free radicals, such as superoxide anion and hydroxyl
are much less reactive. They cannot propagate a radical
radical. These radicals, acting mainly through initiation
chain reaction, and they have sufficient lifetimes to
of chain reactions, can damage the aforementioned
diffuse from the site at which they are generated,
macromolecules to such an extent that neuronal cell
thereby presumably preventing concentrated, debilitat-
death ultimately occurs, with resultant neurologic im-
pairment.
ing tissue damage. Indeed, PBN has been shown to
reduce significantly neuronal cell loss and neurologic
deficits in gerbil4 and rat5,6 models of stroke.
One approach to the treatment of stroke, then, is to
disrupt destructive radical chain reactions by intercept-
ing chain initiators and/or chain carriers with free
radical traps (antioxidants). For example, prolonged
pretreatment of animals with the lipophilic antioxidant
R-tocopherol can ameliorate stroke-induced neurologic
damage.2 We are particularly interested in trapping
radicals with nitrones,3 typified by phenyl-tert-butylni-
trone (PBN, Figure 1). Free radicals react with these
The low in vitro potency of PBN (the IC50 in an in
vitro lipid peroxidation assay is 14.3 mM) and the
requirement of a 100-300 mg/kg body weight dose to
achieve efficacy in in vivo studies led us to develop a
series of cyclic nitrones,3 as outlined in the preceding
paper in this issue. The simplest member of the series,
compound 1 (Figure 1), was already 8-fold more potent
than PBN in the in vitro lipid peroxidation assay.
Further refinements ultimately afforded compounds up
to 650-fold more potent (IC50 ) 22 µM). During the
course of this work, several structure-activity relation-
ships were revealed. For example, a correlation was
found between lipophilicity (clog P or log k′w)7 and in
vitro potency. That is, the more lipophilic compounds
were generally more potent antioxidants. Second, we
found that incorporation of an additional antioxidant
moiety (o,o′-dimethylphenol) into the molecule increased
* To whom correspondence should be addressed. Tel.: 317-277-
8030. FAX: 317-277-4783.
† Current address: Eli Lilly and Company, Lilly Research Labora-
tories, Drop Code GL 45, 2001 W Main St, Greenfield, IN 46140.
‡ The Ohio State University.
§ Current address: Abbott Laboratories, 100 Abbott Park Rd.,
Abbott Park, IL 60064-3500.
| Current address: Neurogen Corporation, 35 Northeast Industrial
Rd., Branford, CT 06405.
X Abstract published in Advance ACS Abstracts, November 15, 1996.
S0022-2623(96)00244-0 CCC: $12.00 © 1996 American Chemical Society

4998 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 25
Thomas et al.
(3599, 3402, 1548, and 1172 cm-1) and UV (maximum
at 305 nm) spectra also indicated that a hydroxyl group
was present and that the nitrone functionality was
retained. Similar to the parent compound 1, the 1H
NMR spectrum of compound 2 showed four aromatic
protons (δ 7.49, br d, J ) 7.5 Hz; 7.46-7.38, AB-m, 2
H; 7.24, br d, J ) 7.5 Hz), consistent with an unaltered
1,2-disubstituted benzene partial structure, and a far-
F igu r e 2. Metabolites of compound 1.
the activity. Perhaps most importantly, it was observed
that all of the compounds tested in animal models
produced side effects.8 These ranged from transient
sedation at moderate doses to rapid death at very high
doses. Although the severity of the sedative effect
varied among different compound classes, the overall
result was that an acceptable therapeutic index was not
attained. Indeed, disappointingly, the most active
compounds also seemed to be the most toxic.
The goal of the present study was to prepare a potent
antioxidant which could readily penetrate the brain
without causing sedation or other undesirable side
effects. We describe herein our efforts to isolate,
identify, and synthesize the major metabolite of 1. We
also report the synthesis of several related compounds
and the evaluation of all of these compounds as free
radical traps (antioxidants) for the treatment of stroke.
1
ther downfield resonance (δ 7.95, br s, J C,H ) ca. 180
Hz), consistent with an unaltered nitrone functionality.
In addition, an aliphatic methine proton (δ 4.64, br s,
1J C,H ) ca. 145 Hz) and two diastereotopic methyl
groups (δ 1.57, s, 3 H and 1.41, s, 3 H) were observed,
indicating that the metabolite was derived from 1 by
introduction of a hydroxyl group at the benzylic position.
This was supported by observation of four aliphatic
carbons with appropriate chemical shifts and multiplici-
ties (δ 74.9, d, 71.5, s, 23.2, q, and 19.0, q) in the 13C/
APT NMR spectra. Furthermore, treatment of com-
pound 2 with acetic anhydride and (N,N-dimethylamino)-
pyridine in the NMR tube led to a substantial downfield
shift of the aliphatic methine proton (from 4.64 to 5.89
ppm) and the appearance of an additional methyl group
(δ 2.02, s, 3 H), consistent with transformation of a
secondary alcohol into its acetate. Also consistent with
structure 2 were NOED experiments in which NOE’s
between the methine proton (δ 4.64) and the proximal
aromatic ortho proton (δ 7.49), as well as the syn-
oriented methyl group (δ 1.41), were observed. The
structural assignment was ultimately confirmed by
synthesis of racemic 2. We note that isolated metabolite
2 was formed as a ca. 90/10 mixture of enantiomers as
determined by chiral HPLC (see the Supporting Infor-
mation). Because of the relatively poor antioxidant
activity of the compound (see below, Table 1) compared
with subsequently prepared analogs, we made no at-
tempts to separate or identify the absolute configuration
of the two enantiomers.
Resu lts a n d Discu ssion
The search for nonsedating nitrones led us to examine
the metabolism of 1 because of the unusual nature of
the sedation (rapid onset and rapid decline) observed
with this compound. When 1 was administered to mice
and the serum was analyzed at various time points by
HPLC, it was observed that the decline in sedation
coincided with the appearance of a major metabolite.
The fact that the antioxidant activity persisted long
after sedation had subsided led to the speculation that
this metabolite retained the antioxidant activity of the
parent, but did not cause sedation.
Isolation an d Iden tification of Metabolites. Spra-
gue-Dawley rats were treated with 1, intraperitoneally,
and three components (2, 3, and recovered 1, Figure 2)
were subsequently isolated from the serum, as described
in the Experimental Section. Compound 1 was identi-
fied by comparison to an authentic sample, whereas the
structures of 2 and 3 were assigned as follows.
Compound 3 could be readily identified by an analysis
of the following spectral data. A molecular ion of 191
and a proton NMR spectrum similar to that of 2 in the
aliphatic and aromatic regions again indicated C-4
hydroxylation. In particular, an aliphatic methine
1
proton (δ 4.58, br s, J C,H ) ca. 145 Hz) was observed,
The mass spectrum of compound 2 exhibited a mo-
lecular ion of 191, 16 mass units higher than the parent,
suggesting that hydroxylation had occurred. The IR
which exhibited a substantial downfield shift (from 4.58
to 5.83 ppm) upon acetylation as before. However, the
absence of the nitrone moiety could be inferred from the
Ta ble 1. Correlation of in Vitro Antioxidant Activity with log k′w and Ability To Maintain Viability of Neuronal Cells Subject to
Oxidative Stress
IC50 (µM)
inhib of liposomal peroxidationb
inhib of neuronal peroxidationc
maintenance of cell viabilityc
a
compd
log k′w
PBN
1
2
6b
6c
7a
7b
7c
8
14300
1670
5500
1040
539
940
157
27
2090
2580d (307)
2600d (292)
1.90
0.91
1.67
2.10
1.65
2.05
2.60
1.95
307/538e
292/490e
1900 (307)
771 (538)
520 (538)
830 (307)
102 (538)
37 (538)
1500 (292)
817 (490)
588 (490)
882 (292)
105 (490)
39 (490)
839 (307)
660 (292)
a
These values represent single determinations. The r2 values for linear regression analyses of the raw chromatographic data were
b
>0.999 in each case. See the Experimental Section of the preceding paper in this issue. Liposomes were prepared from soybean
phosphatidylcholine. Oxidation was initiated and oxidation products quantitated as described previously (see ref 17). Compound 1 was
included as an internal standard in the lipid peroxidation assay of test compounds. If the value obtained for 1 varied by more than 15%
from the value given above, the experiment was excluded. The value given for PBN was reported earlier (see ref 7). c See the Experimental
d
Section for a description of the cerebellar granule cell experiments. These PBN data were not obtained in triplicate runs (n ) 1) as were
those for the other compounds. e Compound 1 was used as an internal standard. The value obtained for 1 in a run on a test compound
is given in parentheses after the value obtained for the test compound.

Cyclic Nitrone Free Radical Traps
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 25 4999
Sch em e 1a
a
Reagents: (a) NaOMe, MeOH; H2SO4; (b) Al/Hg, Et2O/H2O; (c) 2 N HCl, THF; (d) (COCl)2/DMSO, Et3N, CH2Cl2.
Sch em e 2a
lack of the characteristic UV absorption maximum at
305 nm and the absence of the C-1 proton (singlet at δ
7.95 in 2). Instead, an IR band (1664 cm-1), the
downfield shift of one of the aromatic ortho protons
(doublet at δ 8.08), and the presence of an exchangeable
proton (broad singlet at δ 5.93) were consistent with the
a
Reagents: (a) Ac2O, DMAP, Et3N, CH2Cl2.
lactam structure 3. Although this compound was not
synthesized because of its low abundance and likely lack
of antioxidant activity, all spectral data obtained on the
isolated metabolite are consistent with the assigned
structure. We did not determine the enantiomeric
composition of 3.
Compound 2 is much more abundant in vivo than 3
and retains the nitrone functionality and was therefore
more likely to be the putative nonsedating antioxidant.
In order to obtain sufficient material for in vivo testing
and to confirm the structural assignment, the synthesis
of 2 was undertaken.
structure, but differed somewhat from those exhibited
by the material isolated from rat serum (see the
Experimental Section).12 The identity of the two samples
was therefore further confirmed by chiral HPLC as
mentioned above and by acetylation of each sample. In
the former case, the isolated sample gave two major
peaks (ca. 10/90 ratio) with retention times of 9.3 and
11.4 min, respectively. Synthetic (racemic) 2 gave two
peaks of equal intensity, also at 9.3 and 11.4 min (see
the Supporting Information). Upon acetylation, the
signal for the methine proton of the isolated sample
shifts from δ 4.64 to 5.89 ppm as described above,
whereas the corresponding signal for synthetic 2 shifts
from δ 4.58 to 5.89 ppm. The two acetylated samples
also have the same Rf on TLC. We consider the
structure assignment of this simple molecule and the
identity of the two samples to be completely unambigu-
ous.
Syn th esis of th e Ma jor Meta bolite a n d Rela ted
Com p ou n d s. In addition to 2, some more lipophilic
analogs were targeted because of the lipophilicity/
activity correlation mentioned earlier. Oxidation of 2
to the corresponding ketone was especially interesting
since this simple modification would simultaneously
increase the lipophilicity, remove the stereocenter, and
introduce an electron-withdrawing group in conjugation
with the nitrone, rendering it more reactive toward
radicals. Acylation of the hydroxyl group was expected
to increase lipophilicity but have little additional effect.
Finally, replacement of the gem-dimethyl moiety with
a spirocycloalkyl group in both the alcohol and ketone
series was pursued since this approach had been shown
to increase potency in other series and would require
only a trivial modification of the synthetic scheme.
The target molecules were prepared in three or four
steps starting from various nitroalkanes and o-phthala-
ldehyde (Scheme 1). Reaction of these two substrates
in the presence of sodium methoxide,9 followed by
acidification, gave the cyclic acetals 4 (as ca. 1:1
mixtures of cis and trans isomers). The crude mixtures
were obtained in quantitative yield and could be carried
on to the next step as such with no loss in yield.
The nitro acetals 4 were reduced to hydroxylamines
5 by treatment with aluminum amalgam in ether/water
according to a literature procedure.10 Purified hydroxy-
lamines 5 were obtained in 42-53% yield (from o-
phthalaldehyde) after separation from variable amounts
of recovered 4 and the over-reduced amino compounds.
Treatment of hydroxylamino acetals 5 with aqueous HCl
in THF provided the hydroxy nitrones 2 and 6b,c
cleanly and rapidly.11 The crude products could be
easily purified by crystallization or chromatography
(52-67% yield; these yields may have been limited by
the relatively high water solubility of the products).
Spectroscopic data (1H NMR, 13C NMR, IR, MS, UV)
for synthetic 2 were fully consistent with the assigned
Hydroxy nitrones 2 and 6b,c were oxidized to the
corresponding ketones 7a -c under Swern13 conditions
(60-74% yield), whereas only 2 was converted into the
corresponding acetate, 8 (Scheme 2, 38% yield).
The syntheses described above afforded ample quan-
tities of 2 and 6-8 for in vitro and preliminary in vivo
testing. However, in attempting to scale-up the syn-
theses further, difficulties were encountered in repro-
ducing the reduction step (4 to 5). For unknown
reasons, over-reduction to the corresponding amine (9c,
Scheme 3) occasionally dominated the desired process.
In order to circumvent this problem, the following
modified scheme was developed. The nitro acetal 4c
was deliberately converted into the amine 9c, which,
upon treatment with acid, cyclized to imine 10c. Oxida-
tion of the imine with Oxone afforded oxaziridine 11c
very rapidly and cleanly.14 Attempted isomerization of
the oxaziridine to the nitrone by a literature procedure
(40% H2SO4 in MeOH, room temperature15a) afforded
mainly the undesired amide (not shown). However,
isomerization of 11c with p-TsOH in MeOH in the
dark15b gave 6c in good yield. Although this modified
route is somewhat longer than the original one, it
proceeds in good overall yield and, in principle, would
allow the use of mercury to be avoided.
Eva lu a tion of 2 a n d Rela ted Com p ou n d s. The
target compounds 2, 6b,c, 7a -c, and 8 were tested in
the lipid peroxidation assay, and their relative lipophi-
licity was assessed by a chromatographic method (log
k′w, Table 1).16 Their IC50 values for prevention of

5000 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 25
Thomas et al.
Sch em e 3a
a
Reagents: (a) Al/Hg, Et2O/H2O; (b) 10% HCl, THF; (c) Oxone, K2CO3, tol/H2O; (d) p-TsOH, MeOH.
Ta ble 2. Acute Toxicity of Test Compounds at High Doses
Ta ble 3. Comparative in Vitro Activity and Brain
Penetrability of PBN, 1, the Major Metabolite 2, and Ketone 7a
compd 200 mg/kg (deaths/5 mice)a 400 mg/kg (deaths/5 mice)a
brain levelsb
inhib of lipid peroxidationa
PBN
1
2
6b
6c
7a
7b
7c
8
1b
0
0
0
0
0
0
0
0
4b
5
0
0
0
1
4
5
0
compd
PBN
IC50 (µM)
time, min
µg/g wet wt
14300
30
60
30
60
30
60
30
60
51 ( 3c
38 ( 5c
109 ( 34
77 ( 21
142 ( 6
127 ( 5
103 ( 22
67 ( 7
1
1670
5500
940
2
7a
a
b
Compounds were administered intravenously. All of these
animals suffered convulsions.
a
Liposomes were prepared from soybean phosphatidylcholine.
Oxidation was initiated and quantitated as described previously
b
(ref 17). Rats were dosed with 1, 2, or 7a at 200 mg/kg body
weight (ip) and then sacrificed at the indicated time points. The
brains were removed, homogenized, and extracted with CHCl3/
MeOH. Concentrations of 1, 2, and 7a , respectively, were then
determined by HPLC analysis using authentic samples of the
compounds as internal standards. Values represent average (
SEM, n ) 5. c These values are approximate and were taken from
the literature. See ref 18.
to be less active in vitro than 1, instead it is more so.
Likewise, ketone 7b has a log k′w value comparable to
that of 1, yet is 10-fold more potent. Among the ketones,
however, the lipophilicity/activity correlation holds (see
compounds 7a -c).
When the compounds were tested in a whole cell
assay for their ability to protect cultured cerebellar
granule cell neurons from radical-induced oxidation, the
spirocyclic ketones (7b and 7c) were substantially more
effective than 1 and one of the alcohols (6c) was
comparable (Table 1). The data for the cell-based assays
were obtained in two separate experiments run ap-
proximately 6 months apart. Although we do not fully
understand the rather large difference in the IC50 values
for 1 in the two experiments, these are primary cultures,
and changes in the diet of the animals, seasonal
variation, etc. can lead to anomalous results in studies
of this type. Within each experiment, however, the IC50
values were derived from several concentrations of
nitrone run in triplicate, and very good reproducibility
was observed. Compound 1 was included as an internal
standard in each experiment so that valid comparative
data could be obtained.
It is readily apparent that the majority of the IC50
values for antioxidant activity agree well with those for
maintenance of cell viability. Thus, the inhibition of
lipid oxidation by the cyclic nitrones translates to
neuroprotection in this in vitro model of oxidative injury.
The low lipophilicity of 2 would be expected to result in
poor cellular uptake, so it is perhaps not surprising that
this compound is the least effective in this assay.
As stated previously, a major goal of this study was
to find compounds with improved side effect profiles
(relative to 1) and the ability to enter the brain. The
results of an acute toxicity study, given in Table 2, show
that alcohols 2 and 6b,c, and acetate 8 are indeed much
less toxic than 1. No deaths occurred when these
F igu r e 3. Effect of 1 and 2 on spontaneous locomotor activity
in rats. Open squares ) 2, closed squares ) 1. For 1, n ) 24
for the control group and n ) 6 for each dose group except 6.3
mg/kg (n ) 12). For 2, n ) 12 for the control group and n )
6 for each dose group. Data points exhibiting significant
differences from control (p < 0.05, ANOVA, followed by
Dunnett’s multiple comparison analysis) are noted with an
asterisk. See the Experimental Section for a description of
the experiment and the Supporting Information for raw data
with SEM.
oxidation of neuronal lipids were determined,17 and
their ability to maintain viability of neurons subjected
to oxidative stress in vitro was measured (Table 1). In
addition, a crude assessment of the toxicity of the
compounds was obtained by determining survival rates
of mice dosed intravenously at either 200 or 400 mg/kg
body weight (Table 2). Finally, the sedative effects of
compounds 1 and 2 were measured (Figure 3), and the
brain penetrability of 1, 2, and the ketone analog 7a
were quantitated (Table 3).18
The data in Table 1 (with values for PBN and 1
included for comparison) show the expected trend that
increased lipophilicity generally correlates with im-
proved in vitro antioxidant activity. The metabolite 2,
the most polar compound, is less active than the parent
compound 1, but is still more than twice as active as
the acyclic nitrone PBN. Also, the prediction that the
keto analogs would be more potent free radical traps
than the corresponding methylene or hydroxy com-
pounds is apparently correct. Compare, for example,
compounds 1 and 7a in Table 1. On the basis solely of
the lipophilicity/activity correlation, the ketone 7a ought

Cyclic Nitrone Free Radical Traps
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 25 5001
compounds were given to mice at 400 mg/kg (iv), a dose
at which 1 is always lethal. The ketone analogs 7,
however, display acute toxicity that seems to be cor-
related with lipophilicity. The most polar ketone,
dimethyl analog 7a , does show significantly reduced
acute toxicity at the highest dose, but the more lipophilic
spirocyclopentane (7b) is only marginally less lethal
than 1 and the most lipophilic ketone (7c) is indistin-
guishable from 1 in this assay.
and, if so, to determine whether an acceptable thera-
peutic index has been obtained. The former compound
has an activity profile similar to that of 1, but with
sharply reduced side effects, whereas the latter, though
it causes sedation, is much more potent than 1. The
results of these studies will be reported in due course.
Exp er im en ta l Section
Gen er a l. General experimental methods, including deter-
mination of log k′w values and biochemical methods, were
performed as described in the preceding paper in this issue.
P r otection of Cer ebella r Gr a n u le Cells a ga in st Oxid a -
tive Da m a ge. Cerebellar granule cell cultures were prepared
from 8-day-old rats as previously described.19 Briefly, 8-10
cerebella were removed and placed in a Krebs-Ringer bicar-
bonate medium supplemented with BSA and MgSO4. Cer-
ebella were finely chopped and digested with a trypsin/Krebs-
Ringer solution. Cells were then dispersed by trituration in
Krebs-Ringer containing DNase, MgSO4, and trypsin inhibitor
followed by plating at a density of 1 × 106 cells/well in poly-
L-lysine-coated 12-well dishes in MEM with 10% fetal bovine
serum/KCl/glutamine/gentamicin. Medium was replaced at 24
h and cytosine arabinoside added. Experiments were con-
ducted with cells 8-10 days in vitro.
For oxidation studies the medium was removed and replaced
with Na+-free Locke’s solution (154.6 mM N-methyl-D-glucam-
ine, 5.6 mM KCl, 2.3 mM CaCl2, 1 mM MgCl2, 3.6 mM
NaHCO3, and 5 mM HEPES, pH 7.3) with the omission of
glucose. The nitrones were added in either Locke’s solution
or 20% â-cyclodextrins and allowed to incorporate into the cells
for 30 min. At this time, 20 µL of a 5 mM stock solution of
ferrous chloride was added for a final concentration of 100 µM.
After 45 min, the medium was removed and added to 1.5 mL
of TBA/TCA (2:1) with 25 µL of 2% BHT, and thiobarbituric
acid reactive substances (TBARS) were determined as de-
scribed previously.17 The absorbance of the control cells (no
iron) was subtracted from that of the iron-treated cells, and
this value was taken as the denominator to determine the
concentration of nitrone required to inhibit oxidation by 50%.
After removal of the Locke’s, fresh MEM media was added
to the cells along with 100 µL of MTT. After 4 h, 1 mL of cold
2-propanol/0.04 N HCl was added, and the cells were scraped,
mixed well, and transferred to 13 × 100 mm glass test tubes.
The absorbance resulting from mitochondrial reduction of the
MTT (570 nm minus 630 nm) was measured as an assessment
of viability. The percent cell death was determined by
comparison to the absorbance of cells to which no iron had
been added. The difference between the iron-treated and
The compounds were also evaluated for their tendency
to cause sedation. Qualitative observations from several
in vivo experiments suggested that alcohols 2 and 6b,c
and acetate 8 were essentially nonsedating (though
some sluggishness was observed at the very highest
doses, e.g. 400 mg/kg), whereas the ketones 7a -c
caused significant sedation, but of a different nature
than that observed with 1. With the ketones, the onset
of sedation is slower, the duration is longer, and the
intensity is somewhat diminished. Only the metabolite
2 had its sedative effect measured quantitatively (in-
hibition of locomotor activity) in a head-to-head com-
parison with 1. The results are shown in Figure 3.
Linear regression analysis of these data yielded an ED50
value of 13.9 mg/kg for 1. This value falls within the
range of doses which elicited activity statistically dif-
ferent from that of the control. None of the data points
obtained for 2 were statistically different from those of
the control, and the ED50 for 2 was therefore estimated
to be >200 mg/kg. Though less impressive than the
qualitative observations, this single experiment never-
theless demonstrates that the metabolite 2 is, at the
very least, a substantially less potent sedative than 1,
in agreement with the original hypothesis. It should
be noted that the protocol (see the Experimental Sec-
tion) for this experiment did not allow measurements
to be taken during the peak period of sedation observed
with 1. Hence, the disparity between the sedative
properties of 1 and 2 may be greater than that indicated
by Figure 3.
Finally, the data shown in Table 3 suggest that the
diminished toxicity and sedation exhibited by 2 is not
due to poor brain penetration, since brain levels of the
metabolite were equal to, or higher than, those of the
parent at both time points tested. At least one of the
ketone analogs, 7a , also seems to enter the brain
readily. At this time we do not have a convincing
explanation for the dramatic differences in sedative
properties and lethality among these series of com-
pounds.
control cells was taken as 100% in order to calculate the IC50
.
In h ib it ion of Sp on t a n eou s Locom ot or Act ivit y in
Ra ts. Adult male CD rats (125-150 g) were placed in a colony
room which was on a 14:10 light/dark cycle (lights on at 6:00
a.m.) and maintained at 23-26 °C. The animals were housed
four per cage with free access to food and water. All animals
were acclimated for at least 1 week from the date of receipt
before beginning the experiments.
Compounds 1 and 2 were prepared as suspensions in
distilled water containing 1% Tween 80 and given intraperi-
toneally (ip).
Con clu sion s
A hydroxylated metabolite of the previously described
cyclic nitrone free radical trap 1 was isolated, identified,
synthesized, and tested for its ability to inhibit lipid
peroxidation. The metabolite (2), though less potent in
this assay, showed greatly diminished toxic and sedative
effects. Several analogs of 2 were prepared in attempts
to increase the potency while retaining the desirable
side effect profile. All of the compounds were more
lipophilic (log k′w) and more active in vitro than 2. The
ketone analogs, as predicted, were much more potent
than 2 in vitro (up to ca. 200-fold), but showed relatively
little improvement over 1 with respect to side effects.
The two spirocyclohexyl compounds, alcohol 6c and
ketone 7c, are undergoing further evaluation to see if
they are neuroprotective in an animal model of stroke
The apparatus used to measure locomotor activity was a
photocell-based Optivarimex “Autotrack” system (Columbus
Instruments, Columbus, OH). The measure of forward loco-
motion used was the “distance traveled” (DT) parameter as
defined by the Autotrack software. Four beam breaks were
designated as the criterion for registering as one count of
forward locomotion. Rats were injected ip with various doses
of 1 and 2 and placed singly into one of eight clear Plexiglas
boxes (16 × 16 × 8 in.) that were placed on laboratory
benchtops and allowed to acclimatize for 30 min. Assessments
of general depressant potential, as measured by a reduction
of spontaneous locomotor activity, were made as follows: At
the end of the 30 min acclimation period, each rat in a
Plexiglas box was placed into the test chamber and tested for
30 min. ED50 values were then obtained using linear regres-
sion on the cumulative DT scores.20 The raw data for this
experiment are in the Supporting Information.

5002 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 25
Thomas et al.
Isola tion of Com p ou n d s 2 a n d 3 fr om Ra t Ser u m . Ten
male Sprague-Dawley rats (400-410 g) were dosed with 1
(75 mg/kg, intraperitoneally) at time 0 and 1 h. At time 3.5
h, the animals were anesthetized with CO2, and the blood was
withdrawn by cardiac puncture and allowed to clot on wet ice.
The serum was collected and extracted with 2 volumes of
CHCl3/MeOH (2:1). After evaporation of the solvent, the
residue was reconstituted in 300 µL of 5% MeOH/H2O and
purified by reversed phase HPLC (CH3CN/H2O gradient),
whereupon three components were collected from multiple
injections. The last compound to elute was identified as
unchanged 1 (retention time 9.3 min, 4 mg) by comparison
with an authentic sample. The earlier eluting compounds
were assigned structures 2 (retention time 4.9 min, 3 mg) and
3 (retention time 6.5 min, 0.3 mg) on the basis of the following
spectral data. 13C multiplicities were determined from APT
or DEPT spectra, and one-bond proton-carbon couplings
(1J C,H) were estimated by examining 13C satellites in the 1H
NMR spectra. All other spectral data are reported according
to the specifications described in the preceding paper in this
issue.
Compound 2: 1H NMR (CDCl3, ca. 2 mg/mL)12 7.95 (br s, 1,
1J C,H ) ca. 180), 7.49 (br d, 1, J ) 7.5), 7.46-7.38 (AB m, 2),
7.24 (br d, 1, J ) 7.5) 4.64 (br s, 1, 1J C,H ) ca. 145), 1.57 (s, 3),
1.41 (s, 3) (OH not observed); 13C NMR (CDCl3) 132.6 (C), 130.8
(CH), 129.4 (CH), 127.5 (CH), 126.2 (CH), 74.9 (CH), 71.5 (C),
23.2 (CH3), 19.0 (CH3) (only five of the expected seven
resonances in the olefinic region were clearly detectable,
presumably due to line broadening caused by traces of DCl
present in the solvent); IR (CDCl3) 3599, 3402 (br), 1548, 1172,
1033; MS (EI, 70 eV) m/z 191 (M+, base peak), 174, 145, 130.
Compound 3: 1H NMR (CDCl3, ca. 0.2 mg/mL) 8.08 (d, 1, J
) 7.5), 7.60 (dd, 1, J ) 7.5, 7.5), 7.51 (d, 1, J ) 7.5), 7.47 (dd,
1, J ) 7.5, 7.5), 5.93 (exchangeable, br s, 1), 4.58 (br s, 1, 1J C,H
) ca. 145), 1.33 (s, 3), 1.31 (s, 3); IR (CDCl3) 3400 (br), 1664;
MS (EI, 70 eV) m/z 191 (M+), 175, 153, 105, 77 (base peak).
1-Meth oxy-3-(1-m eth yl-1-n itr oeth yl)-1,3-dih ydr oisoben -
zofu r a n (4a ). Sodium metal (12.4 g, 0.539 mol) was added
to methanol (MeOH, 1 L) at 10 °C over 90 min. When a
homogeneous solution was obtained, the cold water bath was
removed and 2-nitropropane (256 mL, 2.85 mol) was added,
followed by o-phthalaldehyde (120 g, 0.895 mol). The resulting
solution was stirred at room temperature overnight. The
solution was brought to pH 2 by adding 1 N H2SO4. White
solids precipitated. The mixture was filtered, and the filter
cake was washed with MeOH and discarded. The filtrate was
stirred at room temperature for 3 h and then made basic by
adding 3 N NaOH. The solution was concentrated in vacuo
to remove the MeOH. The resulting aqueous solution was
extracted twice with ether. The combined organic layers were
washed once with water, dried (MgSO4), and evaporated.
Kugelrohr distillation of remaining solvent at 50 °C (2 mmHg)
left 195 g (106% of theoretical, 86% purity by GC) of a brown
liquid which was used as such in the next step. The ratio of
diastereoisomers was 1:1 (1H NMR). A portion of the crude
material was purified by FC (9:1 cyclohexane/EtOAc) to give
pure 4a as a pale yellow oil: 1H NMR (CDCl3) 7.40-7.35 (m,
3), 7.15-7.10 (m, 1), 6.25 and 5.90 (isomer I, d and dd, 1 total,
J ) 2.4 and 2.4, 0.6, respectively), 6.01 and 5.72 (isomer II, s
and d, 1 total, J ) 0.6), 3.58 and 3.37 (isomers I and II,
respectively, 2 s, 3 total), 1.57 and 1.56 and 1.55 and 1.48 (4
s, 6 total); 13C NMR (CDCl3) 138.8, 138.5, 137.6, 129.8, 129.7,
129.2, 129.1, 123.4, 122.1, 107.3, 107.0, 90.5, 86.8, 56.1, 54.0,
22.5, 22.1, 21.7, 21.0; IR (film) 1543, 1464, 1398, 1373, 1348,
1113, 1094, 1026, 974, 756; MS (CI/CH4, 120 eV) m/z 236 (M
- H)+, 219, 206, 191, 175, 159, 149 (base peak), 131, 118, 91,
73. Anal. (C12H15NO4) C, H, N.
oil (4.6 g, 88%). Purification by FC (1:1 cyclohexane/EtOAc)
gave recovered starting material (0.48 g, 10%), and the
hydroxylamine 5a as a pale green glass (2.24 g, 43%): 1H NMR
(CDCl3) 7.40-7.30 (m, 4), 6.28 and 5.55 (isomer I, 2 d, 1 total,
J ) 2.4), 6.03 and 5.41 (isomer II, 2 s, 1 total), 3.62 and 3.34
(isomers I and II, respectively, 2 s, 3 total), 1.32 and 0.88
(isomer II, 2 s, 3 total), 1.27 and 0.80 (isomer I, 2 s, 3 total);
13C NMR (CDCl3) for one isomer, 140.1, 138.2, 129.2, 128.1,
123.2, 122.3, 106.51, 85.6, 61.1, 53.2, 20.4, 19.1; for other
isomer, 139.7, 138.5, 129.1, 128.0, 123.1, 122.3, 107.0, 60.3,
56.1, 20.9, 19.4; IR (CHCl3) 2980, 2934, 2907, 2891, 1375, 1111,
1092, 1015, 974, 752; MS (CI/CH4, 120 eV) m/z 224 (M + H)+,
206, 192, 176, 174, 159, 149, 147, 135, 119 (base peak), 118,
102, 91, 74.
3,3-Dim eth yl-3,4-d ih yd r oisoqu in olin -4-ol N-Oxid e (2).
To a solution of 5a (7.1 g, 31.8 mmol) in THF (20 mL) was
added 2 N HCl (10 mL), and the resulting solution was stirred
for 45 min at room temperature. More 2 N HCl (10 mL) was
added, and the solution ws stirred for 30 min. The solution
was then slowly poured into a saturated aqueous NaHCO3
solution and extracted five times with EtOAc (at this point,
TLC analysis of the aqueous layer showed that some desired
product still remained, but it was not recovered). The com-
bined organic layers were dried (MgSO4), filtered, and evapo-
rated to obtain 6.1 g of a yellow oil. Recrystallization from
EtOAc/cyclohexane gave 3.45 g (57%) of cream-colored crystals,
mp 134-136 °C. A second crop (0.62 g, 10%) was obtained by
evaporating the mother liquor and recrystallizing the residue
from hexane/CH2Cl2, bringing the total yield to 67%: 1H NMR
1
(CDCl3, ca. 20 mg/mL) 7.62 (s, 1, J C,H ) ca. 180), 7.46 (m, 1),
1
7.38-7.30 (AB m, 2), 7.10 (m, 1), 4.58 (d, 1, J ) 6.3, J C,H
)
ca. 145, CHOH), 3.93 (exchangeable, d, 1, J ) 6.3, CHOH),
1.47 (s, 3), 1.36 (s, 3); 13C NMR (CDCl3) 132.8 (C), 132.7 (CH),
129.8 (CH), 128.9 (CH), 127.3 (CH), 126.6 (C), 125.2 (CH), 74.7
(CH), 71.6 (C), 23.3 (CH3), 19.0 (CH3); IR (CHCl3) 3605, 3403
(br), 1549, 1173, 1031; MS (EI, 70 eV) m/z 191 (M+, base peak),
174, 145, 130. Anal. (C11H13NO2) C, H, N.
3,3-Dim eth yl-3,4-d ih yd r o-3H-isoqu in olin -4-on e N-Ox-
id e (7a ). To a solution of 2 (3.22 g, 16.8 mmol) in CH2Cl2 (150
mL) was added dimethyl sulfoxide (DMSO, 23.8 mL, 336
mmol). The resulting solution was cooled to -45 °C. Oxalyl
chloride ((COCl)2, 11.4 mL, 131 mmol) was added over 10 min,
such that the internal temperature remained below -40 °C.
The mixture was stirred and maintained between -55 and
-40 °C for 2 h. iPr2NEt (44 mL, 250 mmol) was added over
15 min, such that the internal temperature remained below
-50 °C. The reaction mixture was then allowed to warm to
room temperature, whereupon it was poured into water and
extracted with CH2Cl2 (2×). The organic phase was washed
with brine, dried (MgSO4), filtered, and evaporated to give a
yellow oil. The material was filtered through silica gel (EtOAc)
and crystallized from cyclohexane/EtOAc to furnish 7a as a
yellow powder (2.0 g, 63%): 1H NMR (CDCl3) 8.07 (d, 1, J )
7.8), 7.86 (s, 1), 7.69 (dt, 1, J ) 7.6, 1.3), 7.49 (dt, 1, J ) 7.6,
1.0), 7.31 (d, 1, J ) 7.8), 1.74 (s, 6); IR (KBr) 3048, 2996, 1680,
1601, 1555, 1487, 1377, 1366, 1300, 1281, 1244, 1179, 891, 872,
758, 660; MS (EI, 70 eV) m/z 191, 189 [(M+), base peak], 172,
158, 145, 144, 130, 115, 104, 89, 77, 63, 51. Anal. (C11H11
NO2) C, H, N.
-
1-Met h oxy-3-(1-n it r ocyclop en t yl)-1,3-d ih yd r oisob en -
zofu r a n (4b). Condensation of nitrocyclopentane (5.00 g, 40.0
mmol) with o-phthalaldehyde (3.76 g, 28.0 mmol) in the
presence of NaOMe (10 mmol) was carried out as described
above for compound 4a . The resulting pale green oil (7.21 g,
98%) was pure enough for use in the next step. The following
data were obtained on a ca. 1:1 mixture of the cis and trans
diastereoisomers: 1H NMR (CDCl3) 7.50-7.25 (m, 3), 7.10-
7.00 (m, 1), 6.16 and 5.91 (isomer I, 2 d, 1 total, J ) 2.3), 5.93
and 5.80 (isomer II, s and d, respectively, 1 total, J ) 0.7),
3.49 and 3.30 (isomers I and II, respectively, 2 s, 3 total), 2.50-
2.35 (m, 1), 2.30-1.90 (m, 3), 1.75-1.50 (m, 4); MS (CI/CH4,
120 eV) m/z 236 (M - H)+, 219, 206, 191, 175, 159, 149 (base
peak), 131, 118, 91, 73; (C12H15NO2) C, H, N.
N-[1-(3-Meth oxy-1,3-d ih yd r oisoben zofu r a n -1-yl)-1-m e-
th yleth yl]h yd r oxyla m in e (5a ). Aluminum amalgam (Al-
Hg) was prepared10 from aluminum foil (Reynolds, 1.29 g,
0.048 mol) and added to ether (100 mL) and water (0.6 mL,
33 mmol) in a three-necked flask. A solution of 4a (4.8 g, 23.0
mmol) in ether (50 mL) was then added to the stirred mixture
from a dropping funnel at a rate to maintain a vigorous reflux.
The bubbling which occurred initially subsided within 30 min.
The mixture was filtered and the filtrate washed twice with 2
N NaOH, dried (MgSO4), and evaporated to obtain a pale green
N-[1-(3-Meth oxy-1,3-d ih yd r oisoben zofu r a n -1-yl)-1-cy-
clop en tyl]h yd r oxyla m in e (5b). The nitro acetal 4b from
the previous reaction (7.21 g, 27.5 mmol) was reduced with
Al-Hg (from 3.31 g of aluminum foil) as described above for

Cyclic Nitrone Free Radical Traps
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 25 5003
compound 5a . Purification by FC (1:1 EtOAc/hexane) provided
recovered starting material (3.65 g, 35%) and the oily hydroxy-
lamine 5b (4.14 g, 42%): 1H NMR (CDCl3) 7.40-7.30 (m, 4),
6.28 and 5.55 (isomer I, 2 d, 1 total, J ) 2.4), 6.03 and 5.41
(isomer II, 2 s, 1 total), 3.62 and 3.34 (isomers I and II,
respectively, 2 s, 3 total), 1.32 and 0.88 (isomer II, 2 s, 3 total),
1.27 and 0.80 (isomer I, 2 s, 3 total); MS (CI/CH4, 120 eV) m/z
250 (M + H)+, 248, 246, 218, 200, 185, 172, 149, 135, 119, 100
(base peak), 84, 67.
(CDCl3) 132.4, 131.6, 129.6, 129.3, 128.7, 126.9, 125.1, 74.2,
69.7, 32.0, 26.2, 25.0, 22.6, 22.1; IR (KBr) 3408, 3073, 3052,
2980, 2938, 2926, 2857, 1593, 1553, 1454, 1414, 1260, 1235,
1179, 1161, 1107, 1049, 1030, 912, 851, 764, 613; MS (CI/CH4,
120 eV) m/z 232 [(M + H)+, base peak], 214, 198, 183. Anal.
(C14H17NO2) C, H, N.
3,4-Dih yd r o-3H -isoq u in olin -4-on e -3-sp ir ocycloh e x-
a n e N-Oxid e (7c). To a solution of (COCl)2 (0.50 mL, 5.73
mmol) in CH2Cl2 (15 mL) at -78 °C was added a solution of
DMSO (1 mL, 14.1 mmol) in CH2Cl2 (5 mL). The resulting
solution was stirred for 5 min at -78 °C. Compound 6c (1.16
g, 5.00 mmol) was dissolved in warm DMSO, and then the
solution was allowed to cool to room temperature. This
solution was added to the reagent solution at a rate such that
the internal temperature remained below -40 °C. The
mixture was stirred and maintained in the -78 °C bath for
15 min and then treated with a solution of Et3N (3.5 mL, 25
mmol) in CH2Cl2 (7 mL) at a rate such that the internal
temperature remained below -50 °C. The mixture was stirred
and maintained in the -78 °C bath for 15 min and then
allowed to warm to room temperature. The reaction mixture
was poured into water and extracted with CH2Cl2 (2×). The
organic phase was washed with brine, dried (MgSO4), filtered,
and evaporated. The residue was crystallized from hexane/
EtOAc to furnish 7c as yellow needles (0.85 g, 74%): mp 92-
93 °C; 1H NMR (CDCl3) 8.03 (d, 1, J ) 7.7), 7.89 (s, 1), 7.66 (t,
1, J ) 7.6), 7.48 (t, 1, J ) 7.6), 7.28 (d, 1, J ) 7.5), 2.55-2.35
(m, 2), 2.15-1.65 (m, 6), 1.60-1.35 (m, 2); 13C NMR (CDCl3)
197.0, 135.3, 132.2, 129.6, 127.2, 125.4, 125.3, 111.1, 80.2, 31.9,
24.0, 21.3; IR (KBr) 3441, 3040, 2942, 2884, 2868, 2845, 1694,
1599, 1553, 1447, 1366, 1319, 1281, 1258, 1182, 1157, 882, 752,
696, 660, 637; MS (EI, 70 eV) m/z 229 (M+), 213, 212 (base
peak), 188, 184, 174, 158, 132, 129, 102, 89, 76, 63, 51, 41.
Anal. (C14H15NO2) C, H, N.
4-Acetoxy-3,3-d im eth yl-3,4-d ih yd r oisoqu in olin e N-Ox-
id e (8). To a solution of 2 (3.3 g, 17 mmol) in CH2Cl2 (100
mL) was added Et3N (3.1 mL, 22 mmol), 4-(dimethylamino)-
pyridine (210 mg, 1.7 mmol), and acetic anhydride (1.8 mL,
19 mmol). The mixture was stirred for 1 h at room temper-
ature, then poured into water, and extracted with CH2Cl2 (2×).
The organic phase was dried (MgSO4), filtered, and evaporated
to give a yellow paste. This was purified by FC (9:1 CH2Cl2/
acetone) to afford 1.83 g of a pale yellow solid. Recrystalliza-
tion from cyclohexane/EtOAc provided 8 as cream-colored
crystals (1.53 g, 38%): 1H NMR (CDCl3) 7.76 (s, 1), 7.45-7.25
(m, 3), 7.20-7.15 (m, 1), 5.89 (s, 1), 2.02 (s, 3), 1.57 (s, 3), 1.36
(s, 3); IR (KBr) 3048, 2986, 2936, 1734, 1593, 1553, 1454, 1375,
1287, 1240, 1211, 1018, 978, 964, 770; MS (EI, 70 eV) m/z 233
[(M+), base peak], 191, 190, 174, 156, 143, 130, 115, 91, 89,
77, 63, 51, 43. Anal. (C13H15NO3) C, H, N.
3,4-Dih yd r oisoq u in olin -4-ol-3-sp ir ocyclop en t a n e N-
Oxid e (6b). Hydroxylamine 5b (4.14 g, 16.7 mmol) was
converted into nitrone 6b according to the procedure described
above for 2. The product (1.87 g, 52%) was obtained as a white
solid after FC (1:1 EtOAc/hexane, then EtOAc): mp 141-143
1
°C; H NMR (CDCl3) 7.67 (s, 1), 7.45-7.30 (m, 3), 7.15-7.10
(m, 1), 4.52 (d, 1, J ) 7.3), 3.98 (d, 1, J ) 7.3), 2.70-2.55 (m,
1), 2.15-2.05 (m, 1), 2.00-1.50 (m, 6); 13C NMR (CDCl3) 132.7,
132.5, 129.6, 129.1, 127.8, 127.2, 125.2, 82.2, 74.3, 36.6, 30.5,
26.6, 25.9; IR (KBr) 3397, 3385, 3351, 3196, 3117, 3067, 3000,
2959, 2872, 1595, 1561, 1452, 1397, 1254, 1240, 1171, 1119,
1101, 1063, 1030, 772; MS (EI, 70 eV), m/z 218, 217 (M+), 200
(base peak), 176, 170, 142, 130, 115, 104, 89, 77, 51, 41. Anal.
(C13H15NO2) C, H, N.
3,4-Dih yd r o-3H -isoq u in olin -4-on e -3-sp ir ocyclop e n -
ta n e N-Oxid e (7b). Compound 6b (1.09 g, 5.02 mmol) was
oxidized with DMSO (1.0 mL, 14.1 mmol), (COCl)2 (0.5 mL,
5.73 mmol), and Et3N (3.5 mL, 25 mmol) according to the
procedure described above for 7a . The crude product was
purified by two crystallizations from hexane/EtOAc to furnish
1
7b as a yellow solid (0.65 g, 60%): mp 107-108 °C; H NMR
(CDCl3) 8.06 (d, 1, J ) 7.8), 7.88 (s, 1), 7.68 (t, 1, J ) 7.6), 7.47
(t, 1, J ) 7.6), 7.30 (d, 1, J ) 7.8), 2.55-2.45 (m, 2), 2.35-1.90
(m, 6); 13C NMR (CDCl3) 197.9, 135.6, 132.1, 132.0, 129.6,
127.2, 125.7, 125.0, 86.9, 40.4, 27.8; IR (KBr) 3441, 2976, 2945,
2870, 1682, 1595, 1555, 1485, 1360, 1323, 1281, 1252, 1181,
893, 855, 756, 662; MS (EI, 70 eV) m/z 215 (M+), 198 (base
peak), 174, 170, 152, 130, 127, 103, 89, 76, 63, 41. Anal.
(C13H13NO2) C, H, N.
1-Meth oxy-3-(1-n itr ocycloh exyl)-1,3-d ih yd r oisoben zo-
fu r a n (4c). Nitrocyclohexane (12.92 g, 100 mmol) was
condensed with o-phthalaldehyde (8.38 g, 60.0 mmol) as
described above for compound 4a . After final drying on a
vacuum pump, a yellow oil was obtained (17.29 g, 100%). The
following data were obtained on a ca. 1:1 mixture of the cis
and trans diastereoisomers: 1H NMR (CDCl3) 7.45-7.35 (m,
3), 7.20-7.10 (m, 1), 6.26 and 5.55 (isomer I, 2 d, 1 total, J )
2.7, and 2.3, respectively), 5.98 and 5.38 (isomer II, 2 s, 1 total),
3.59 and 3.36 (isomers I and II, respectively, 2 s, 3 total), 2.59
(m, 2), 2.22 (m, 2), 1.95-1.10 (m, 6).
N-[1-(3-Meth oxy-1,3-d ih yd r oisoben zofu r a n -1-yl)-1-cy-
cloh exyl]h yd r oxyla m in e (5c). Nitro acetal 4c (14.24 g, 51.4
mmol) was reduced with Al-Hg (from 6 g of aluminum foil)
as described above for compound 5a . Purification by FC (1:1
EtOAc/hexane) gave the hydroxylamine 5c as a pale yellow
oil (7.19 g, 53%). The following data were obtained on a ca.
1:1 mixture of cis and trans isomers: 1H NMR (CDCl3) 7.45-
7.30 (m, 4), 6.28 and 5.55 (isomer I, 2 d, 1 total, J ) 2.4), 6.03
and 5.39 (isomer II, 2 s, 1 total), 3.63 and 3.31 (isomers II and
I, respectively, 2 s, 3 total), 1.85-1.40 (m, 8), 1.20-1.05 (m,
2); MS (CI/CH4, 120 eV) m/z 264 (M + H)+, 262, 246, 230, 214,
199, 171, 150, 149, 135, 118, 114 (base peak), 96, 84.
Alter n a te Syn th esis of 6c. N-[1-(3-Meth oxy-1,3-d ih y-
d r oisoben zofu r a n -1-yl)-1-cycloh exyl]a m in e (9c).21 Nitro
compound 4c (34.4 g, 124 mmol) was reduced with Al-Hg
(from 7.6 g of Al foil, 0.283 mol) as described above for
compound 5a . The crude product was purified by FC (1:1
hexane/EtOAc) to give recovered 4c (5.9 g, 17%) and amine
9c (21.5 g, 70%) as a mixture of diastereoisomers: 1H NMR
(CDCl3) 7.40-7.30 (m, 4), 6.24 and 5.10 (isomer I, 2 d, 1 total,
J ) 2.2), 5.94 and 5.02 (isomer II, 2 s, 1 total), 3.58 and 3.31
(isomers II and I, respectively, 2 s, 3 total), 1.62-1.18 (m, 10);
IR (CHCl3) 3691, 3597, 3080-2860, 1601, 1508, 1451, 1376,
1087, 1011; 13C NMR (CDCl3) 139.7, 139.3, 138.9, 138.7, 128.8,
128.7, 127.9, 127.9, 123.0, 122.9, 106.6, 106.3, 91.1, 90.7, 54.8,
53.8, 53.1, 35.0, 33.7, 33.5, 33.4, 25.9, 25.8, 21.4, 21.3, 21.2,
21.1.
3,4-Dih yd r oisoqu in olin -4-ol-3-sp ir ocycloh exa n e N-Ox-
id e (6c). To a solution of 5c (7.19 g, 27.3 mmol) in THF (100
mL) was added 10% HCl (50 mL), and the resulting solution
was stirred for 20 min at room temperature. The solution was
then slowly poured into saturated aqueous NaHCO3 solution
and extracted three times with EtOAc. The combined organic
layers were dried (MgSO4), filtered, and concentrated, where-
upon a beige solid precipitated. This was collected and washed
with hexane to furnish 3.36 g (53%) of pure product. The
filtrate was evaporated, and the residue crystallized from
EtOAc/hexane to give a second crop (0.72 g, 11%) of product,
bringing the total yield to 64%: mp 195-197 °C; 1H NMR
(CDCl3) 7.65 (s, 1), 7.45-7.30 (m, 3), 7.20-7.10 (m, 1), 4.93
(d, 1, J ) 7.3), 3.32 (d, 1, J ) 7.3), 2.47 (td, 1, J ) 16.0, 4.9),
2.25-2.15 (m, 1), 2.00-1.85 (m, 1), 1.80-1.30 (m, 7); 13C NMR
3,4-Dih yd r oisoqu in olin -4-ol-3-sp ir ocycloh exa n e (10c).
Amine 9c (5.10 g, 20.6 mmol) was dissolved in THF (70 mL)
and treated with 10% HCl. The mixture was stirred overnight
at room temperature, then poured into saturated NaHCO3
solution, and extracted with EtOAc (3×). The organic phase
was dried (MgSO4), filtered, and concentrated to give 4.31 g
(97%) of imine 10c as a white solid: mp 120-123 °C (EtOAc/
1
hexanes); H NMR (CDCl3) 8.12 (s, 1), 7.40-7.20 (m, 4), 4.41
(s, 3), 4.11 (br s, 1), 1.85-1.25 (m, 10); 13C NMR (CDCl3) 157.2,
137.0, 131.4, 128.3, 127.7, 127.0, 126.3, 70.5, 60.5, 33.4, 31.1,
25.6, 21.7, 21.3; IR (CHCl3) 3300, 3076-2859, 2209, 1628,
1577, 1454, 1376, 1218, 1013.

5004 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 25
Thomas et al.
(8) Unpublished results.
Con ver sion of Im in e 10c in to Nitr on e 6c. The imine
10c (10.13 g, 47.0 mmol) from the previous reaction was
dissolved in warm toluene (300 mL), and an aqueous solution
of K2CO3 (52.1 g, 377 mmol, in 250 mL of H2O) was added. To
the resulting vigorously stirred mixture was added dropwise
a solution of Oxone (34.7 g, 56.5 mmol) in H2O (250 mL). After
the addition was complete, the mixture was stirred for 15 min
and then the layers were separated. The aqueous phase was
extracted with EtOAc, and the combined organic phase was
washed with 10% Na2SO3 solution, dried (MgSO4), filtered, and
evaporated to furnish the intermediate oxaziridine 11c (11.1
g, quant) as a semisolid containing a trace of toluene. This
material was dissolved in MeOH (500 mL) and treated with a
catalytic amount of p-toluenesulfonic acid (2.25 g, 11.82 mmol).
The mixture was stirred at room temperature in the dark for
1.5 days. The reaction mixture was then poured into saturated
NaHCO3 solution and extracted first with CH2Cl2 (4×) and
then with 50% MeOH/CH2Cl2 (3×). The extracts were com-
bined, dried (MgSO4), filtered, and evaporated. The resulting
solid was dissolved in MeOH and diluted with EtOAc to induce
crystallization. Two crops of yellowish crystals (6c) were
collected to give a total yield of 7.42 g (68%) from 10c. The
spectral data for this material matched those given above for
6c.
(9) Marquard, F.-H.; Edwards, S. Reductive Synthesis of R,R-
Dimethylphenethylamine. J . Org. Chem. 1972, 37, 1861-1863.
(10) Calder, A.; Forrester, A. R.; Hepburn, S. P. 2-Methyl-2-nitroso-
propane and its Dimer. Organic Syntheses; Wiley: New York,
1988; Collect. Vol. VI, pp 803-806.
(11) Treatment of 4a with Zn/HOAc, known conditions for intermo-
lecular reactions of aldehydes with nitro compounds to give
nitrones (Hinton, R. D.; J anzen, E. G. Synthesis and Charac-
terization of Phenyl-Substituted C-Phenyl-N-tert-butylnitrones
and some of Their Radical Adducts. J . Org. Chem. 1992, 57,
2646-2651), failed to afford 2.
(12) A reviewer expressed concern about the discrepancies between
the NMR chemical shifts reported for the isolated metabolite 2
and synthetic 2 (see the Experimental Section). We note that
1H NMR spectra of synthetic 2 obtained in CDCl3 at different
concentrations and/or using different batches of solvent show
significant variations in chemical shifts and line widths, par-
ticularly in the aromatic/olefinic region. These variations are
presumably caused by small amounts of acidic impurities,
specifically DCl, which is typically present in this solvent unless
removed immediately prior to use. This is illustrated by a series
of spectra of synthetic 2 in CDCl3 under different conditions.
These spectra, and a spectrum of the isolated metabolite 2 for
comparison, are included in the Supporting Information. Similar
variations have also been observed for the parent compound 1.
This is in fact a very common phenomenon that can be observed
for many compounds containing basic functionality which can
be protonated and can engage in intra- or intermolecular
tautomeric equilibria.
Ack n ow led gm en t. We thank Dr. Teng-Man Chen
for performing the chiral HPLC analyses, Dr. Michael
P. J ohnson for help with statistical analysis, and Drs.
Robert A. Farr and Robert J . Cregge for helpful discus-
sions.
(13) Omura, K.; Swern, D. Oxidation of Alcohols by “Activated”
Dimethyl Sulfoxide. A Preparative, Steric and Mechanistic
Study. Tetrahedron 1978, 34, 1651-1660.
(14) This is the same procedure used to prepare Davis’ oxaziridine.
See: Davis, F. A.; Chattopadhyay, S.; Towson, J . C.; Lal, S.;
Reddy, T. Chemistry of Oxaziridines. 9. Synthesis of 2-Sulfonyl-
and 2-Sulfamyloxaziridines Using Potassium Peroxymonosulfate
(Oxone). J . Org. Chem. 1988, 53, 2087-2089.
Su p p or tin g In for m a tion Ava ila ble: Chiral HPLC chro-
matograms for isolated and synthetic 2, additional 1H NMR
1
data for synthetic 2, the H NMR spectrum of isolated 2, and
raw data for Figure 3 (10 pages). Ordering information is
given on any current masthead page.
(15) (a) Ogata, Y.; Sawaki, Y. Peracid Oxidation of Imines. Kinetics
and Mechanism of Competitive Formation of Nitrones and
Oxaziranes from Cyclic and Acyclic Imines. J . Am. Chem. Soc.
1973, 95, 4692-4698. (b) For a UV light-promoted isomerization
of oxaziridines to amides, see Duhamel, P.; Goument, B.;
Plaquevent, J .-C. A Formal Synthesis of Aspartame via the
Oxaziridine-Amide Rearrangement. Tetrahedron Lett. 1987, 28,
2595-2596.
(16) log k′w values were measured according to a literature method,
as described briefly in the Experimental Section of the preceding
paper in this issue. See, for example: (a) Braumann, T.
Determination of Hydrophobic Parameters by Reversed-Phase
Liquid Chromatography: Theory, Experimental Techniques, and
Application in Studies on Quantitative Structure-Activity Re-
lationships. J . Chromatogr. 1986, 373, 191-225. (b) Hsieh, M.
M.; Dorsey, J . G. Bioavailability Estimation by Reversed-Phase
Liquid Chromatography: High Bonding Density C-18 Phases
for Modeling Biopartitioning Processes. Anal. Chem. 1995, 67,
48-57. (c) Kugel, C.; Heintzelmann, B.; Wagner, J . Determi-
nation of Distribution Coefficients for Some 5-HT3 Receptor
Antagonists by Reversed-Phase High-Performance Liquid Chro-
matography. J . Chromatogr. A 1994, 667, 29-35.
Refer en ces
(1) (a) Floyd, R. A. Role of Oxygen Free Radicals in Carcinogenesis
and Brain Ischemia. FASEB J . 1990, 4, 2587-2597. (b) Kontos,
H. A. Oxygen Radicals in CNS Damage. Chem-Biol. Int. 1989,
72, 229-255.
(2) Hara, H.; Kato, H.; Kogure, K. Protective Effect of R-Tocopherol
on Ischemic Neuronal Damage in the Gerbil Hippocampus.
Brain Res. 1990, 510, 335-338.
(3) (a) Bernotas, R. C.; Thomas, C. E.; Carr, A. A.; Nieduzak, T. R.;
Adams, G.; Ohlweiler, D. F.; Hay, D. A. Synthesis and Radical
Scavenging Activity of 3,3-Dialkyl-3,4-Dihydroisoquinoline 2-Ox-
ides. Bioorg. Med. Chem. Lett. 1996, 6, 1105-1110. (b) Ber-
notas, R. C.; Adams, G.; Carr, A. A. Synthesis of Benzazepine-
based Nitrones as Radical Traps. Tetrahedron 1996, 52, 6519-
6526. (c) Thomas, C. E.; Ohlweiler, D. F.; Kalyanaraman, B.
Multiple Mechanisms for Inhibition of Low Density Lipoprotein
Oxidation by Novel Cyclic Nitrone Spin Traps. J . Biol. Chem.
1994, 269 (45), 28055-28061.
(4) (a) Thomas, C. E.; Carney, J . M.; Bernotas, R. C.; Hay, D. A.;
Carr, A. A. In Vitro and In Vivo Activity of a Novel Series of
Radical Trapping Agents in Model Systems of CNS Oxidative
Damage. Ann. N.Y. Acad. Sci. 1994, 738, 243-249. (b) Oliver,
C. N.; Starke-Reed, P. E.; Stadtman, E. R.; Liu, G. J .; Carney,
J . M.; Floyd, R. A. Oxidative Damage to Brain Proteins, Loss of
Glutamine Synthetase Activity, and Production of Free Radicals
During Ischemia/Reperfusion-Induced Injury to Gerbil Brain.
Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 5144-5147. (c) Yue, T.-
L.; Gu, J .-L.; Lysko, P. G.; Cheng, H.-Y.; Barone, F. C.; Feuer-
stein, G. Neuroprotective Effects of Phenyl-t-butyl-nitrone in
Gerbil Global Brain Ischemia and in Cultured Rat Cerebellar
Neurons. Brain Res. 1992, 574, 193-197 and references cited
therein.
(5) Cao, X.; Phillis, J . W. R-Phenyl-tert-butyl Nitrone Reduces
Cortical Infarct and Edema in Rats Subjected to Focal Ischemia.
Brain Res. 1994, 644, 267-292.
(6) Zhao, Q.; Pahlmark, K.; Smith, M.-L.; Siesjo, B. K. Delayed
Treatment with the Spin Trap R-Phenyl-N-tert-butyl Nitrone
(PBN) Reduces Infarct Size Following Transient Middle Cerebral
Artery Occlusion in Rats. Acta Physiol. Scand. 1994, 152, 349-
350.
(7) Thomas, C. E.; Ohlweiler, D. F.; Carr, A. A.; Nieduzak, T. R.;
Hay, D. A.; Adams, G.; Vaz, R.; Bernotas, R. C. Characterization
of the Radical Trapping Activity of a Novel Series of Cyclic
Nitrone Spin Traps. J . Biol. Chem. 1996, 271, 3097-3104.
(17) Thomas, C. E.; McLean, L. R.; Parker, R. A.; Ohlweiler, D. F.
Ascorbate and Phenolic Antioxidant Interactions in Prevention
of Liposomal Oxidation. Lipids 1992, 27, 543-550.
(18) For brain penetrability of PBN, see: Chen, G.; Griffin M.; Poyer,
J . L.; McCay, P. B. HPLC Procedure for the Pharmacokinetic
Study of the Spin-Trapping Agent, R-Phenyl-N-tert-Butyl Ni-
trone (PBN). Free Radical Biol. Med. 1990, 9, 93-98.
(19) Levi, G.; Aloisi, F.; Ciotti, M. T.; Thangnipon, W.; Kingbury, A.;
Balazs, R. Preparation of 98% Pure Cerebellar Granule Cell
Cultures. In A Dissection and Tissue Culture Manual of the
Nervous System; Shahar, A., de Vellis J ., Vernadakis, A., Haber,
B., Eds.; Alan R. Liss, Inc.: New York, 1989; pp 211-214.
(20) Finney, D. J . Statistical Method in Biological Assay, 3rd ed.;
MacMillan: New York, 1978.
(21) Although this procedure is ostensibly the same as that given
for the synthesis of hydroxylamine 5c, it was carried out in a
different laboratory and on several occasions gave mainly the
amino compound 9c described here. Some of this amine (9c)
was formed in the preparation of 5c described above as well,
but it was not isolated.
J M960244N