8-Aminobicyclo[3.2.1]octanes: Synthesis and anti-viral activity

Article abstract of DOI:10.1016/S0040-4039(01)01593-3

A series of 8-chlorobicyclo[3.2.1]oct-6-enes has been prepared from cyclohexenyl chlorides and simple alkynes via a one-step [3+2] cycloaddition methodology, and then converted into the corresponding bicyclic amines. Against influenza-A virus a number of

Full text of DOI:10.1016/S0040-4039(01)01593-3

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 7503–7507
8-Aminobicyclo[3.2.1]octanes: synthesis and anti-viral activity
J. A. Miller,* G. M. Ullah, G. M. Welsh and A. C. Hall
Department of Medicinal Chemistry, Wellcome Research Laboratories, Langley Court, Beckenham, Kent BR3 3BS, UK
Received 14 June 2001; accepted 24 August 2001
Abstract—A series of 8-chlorobicyclo[3.2.1]oct-6-enes has been prepared from cyclohexenyl chlorides and simple alkynes via a
one-step [3+2] cycloaddition methodology, and then converted into the corresponding bicyclic amines. Against influenza-A virus
a number of these showed in vitro activity, and amine (2) was comparable to amantadine in vivo. Good activity against
influenza-B virus was less common, and none of the amines showed high potency against both viruses. Amines (20) and (23)
showed significant activity versus respiratory syncytial virus. © 2001 Elsevier Science Ltd. All rights reserved.
The discovery of the therapeutic and prophylactic prop-
erties of amantadine nearly 40 years ago led to immense
interest in the synthesis and testing of lipophilic, bulky
amines as potential anti-influenza agents,1 despite the
fact that nothing was known then about the biological
target of these compounds.2 Moreover, although aman-
tadine and its analogue rimantadine have had some
success as influenza therapies, they have a number of
drawbacks, notably CNS side-effects and a lack of
activity against clinically relevant strains of influenza-B
virus.1
discovery6 of a direct [3+2] cycloaddition route to 8-
chlorobicyclo[3.2.1]oct-6-enes gave us the opportunity
to synthesise a range of derived bicyclic amines, typified
by the generic structures A and B. Our broad objective
was to find compounds that were orally active against
both influenza-A and influenza-B, the clinically impor-
tant influenza viruses. This report outlines the synthesis
of compounds of types A and B, as well as two
compounds of type C, and describes their anti-viral
properties against influenza-A and -B, and respiratory
syncytial virus (RSV).
Previous study of structure–activity relationships (SAR)
in the amantadine series revealed that introduction of
polar functionality beyond the basic amino group and/
or addition of bulky ligands to the adamantane frame-
work generally led to loss of activity.1,3 Moreover, these
SAR features extended to other amino substituted
bridged bicyclic systems, such as bicyclo[2.2.2]octanes4
and bicyclo[2.2.1]heptanes.5 However, the synthetic
routes to these analogues were often tedious, and were
not always amenable to variation in substituents. The
Given the simple SAR already established for aman-
tadine and its congeners, our first target was to prepare
and test a few highly symmetrical bicyclo[3.2.1] amines
which were roughly the same size, shape and polarity as
amantadine, and compounds 1 and 2 were chosen as
representatives of types A and B, respectively. Of these,
compound 2 was particularly striking because the two
methyl groups at C-6 and C-7 are tucked under the
bridge and ensure an amantadine-like shape, all com-
Keywords: allyl cation; [3+2] cycloaddition; 8-aminobicyclo[3.2.1]octanes and 6-enes; anti-viral activity versus influenza-A and -B; respiratory
syncytial virus.
* Corresponding author. Present address: Protherics PLC, Beechfield House, Lyme Green Business Park, Macclesfield, Cheshire SK11 0JL, UK.
Tel.: +0044 (0)1625 500555; fax: 0044 (0)1625 500666; e-mail: allen.miller@protherics.com
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01593-3

7504
J. A. Miller et al. / Tetrahedron Letters 42 (2001) 7503–7507
pounds have the C-8 bridge arbitrarily depicted as ‘up’.
We were therefore, highly encouraged when 1 was
found to have an IC50 of 12 mM, and its saturated
analogue 2 had an IC50 of 1–3 mM against influenza-A
in vitro. On this basis, we embarked on a programme
of synthesis of analogues of 1 and 2.
epimer of the amine 2. As shown in Scheme 3, and
explained above, inversion at C-8 could only be
achieved after removal of the C-6 to C-7 double bond,
and this was followed by tosylate displacement by azide
ion to give the inverted (3J=1.5 Hz for C8-H) azide (IR
2090 cm−1). The starting material for this sequence,
made by hydrolysis of the corresponding chloride,7 was
also used in the preparation of bridge amides via the
Ritter reaction, see also Scheme 3. Other amides were
made by direct acylation of the amine 1.
The general synthesis of type A and type B amines is
outlined in Scheme 1. The [3+2] cycloaddition (the term
indicates a structural feature, and has no mechanistic
implications) was achieved using zinc chloride, in
accord with the original conditions6 and the adduct
yields were generally in the moderate to good range
(55–80%). When there is the possibility of forming two
regiomeric cycloadducts, appropriate choice of R6 and
R7 (e.g. H and Ph; or Me and Ph; or Me and SPh,
respectively) results in only one isomer being observed.
A further feature of the cycloadducts is that the bridge
chlorine is always oriented exo to the double bond, i.e.
it lies over the cyclohexane ring in the adduct. This has
been observed consistently6,7 and been commented
upon elsewhere.8 Moreover, provided that the 6–7 dou-
ble bond is still in place when the bridge chlorine is
replaced, nucleophilic substitution reactions of the chlo-
rine at the bridge all result in retention of configura-
tion, so that in the sequence C8-Cl“C8-N3“C8-NH2,
as in Scheme 1, the stereochemistry of the azide and
amino groups is assured. The relative stereochemistry at
the bridge is readily assigned from the vicinal coupling
constant of about 5 Hz between the bridge methine
proton and those at C-1 and C-5.7,9 For our type B
targets, the double bond was finally removed using
catalytic hydrogenation, which always occurred from
the same face as the bridge, so that groups R6 and R7
were placed endo in the cyclopentane ring. It was this
consideration which led to 2 as the preferred target for
early synthesis.
The same alcohol was also the starting material for the
synthesis of rimantadine-like analogues of 2, in which
the basic amino group is moved out from the bridge by
one carbon. As illustrated for amine 6 in Scheme 4, the
chain extension was achieved by an Emmons–Horner-
like sequence,10 followed by hydrolysis of the intermedi-
ate enol–ethers under acidic conditions. The final
product, 6, was isolated and tested as a diastereoiso-
meric mixture.
Scheme 2. Reagents and conditions: (i) W-2 Raney nickel,
EtOH (reflux); (ii) HCl, H2O, EtOH (reflux).
The above issues were common to all of the bicy-
clo[3.2.1]octanes or oct-6-enes, but the functional group
manipulations were otherwise routine and not worthy
of detailed comment, apart from the few examples
discussed below. After 1 and 2 had been identified as
leads, there was a clear need to make analogues in
which one or both of the methyl groups was removed,
and this was achieved using reductive de-sulphurisation
to yield compounds 3 and 4, as in Scheme 2. One or
more phenylsulphenyl groups in the alkyne improves
the [3+2] cycloaddition yield noticeably, and reductive
removal is also efficient, such that this route to 3 and 4
is preferred to routes using ethyne or propyne as the
alkyne partner. Another obvious target was 5, the C-8
Scheme 3. Reagents and conditions: (i) 5% Pd–C, H2 (50 atm),
EtOH (50°C); (ii) p-toluenesulphonyl chloride, pyridine (40–
60°C); (iii) NaN3, HMPA (80°C); (iv) LAH, Et2O (−10–
22°C); (v) 57% H2SO4, RCN, CHCl3 (0–22°C); (vi) (RCO)2O,
DMAP, pyridine.
Scheme 1. Reagents and conditions: (i) ZnCl2, CH2Cl2; (ii) ZnCl2, NaN3; (iii) LAH, Et2O (−10 to 22°C); (iv) Pd–C, H2, AcOH.

J. A. Miller et al. / Tetrahedron Letters 42 (2001) 7503–7507
7505
Scheme 4. Reagents and conditions: (i) pyridinium chlorochromate, CH2Cl2; (ii) Ph2P(O)CH(Me)OMe, LDA, THF; (iii) HCl,
MeOH; (iv) NH2OH, HCl, pyridine, EtOH (reflux); (v) PtO2, H2 (50 atm), AcOH (50°C).
All compounds were initially screened for activity
against influenza-A using a standard plaque inhibition
assay. Active compounds11 were then assayed quantita-
tively by plaque reduction and an IC50 was assigned, as
shown in the Table 1. As is general with anti-viral
assays, any effects of compounds on host cell (usually
MDCK cells) growth were also monitored, in order
that anti-viral activity was not confused with host cell
toxicity. With respect to anti-influenza activity, it was
quickly confirmed that the established SAR, referred to
above, also applied to the bicyclo[3.2.1]octanes. For
example, in the B series, placement of a tertiary OH
group at C-7 (see entry 8) caused complete loss of
activity. Moreover, for both type A and B (entries 9
and 10, respectively), a simple replacement of the 6-Me
by 6-Ph abolished activity. Similarly, introduction of a
Table 1. Anti-viral activity of bicyclo[3.2.1]octane amines against enveloped viruses
Compd.
no.
Structure
typea
Structure detailsb
Anti-viral activityc
Toxicity
ratingd
R1
R2a R2e
R3a R3e
R4a R4e
R6
R7
R8
Flu A
Flu B
RSV
1
2
3
4
5
6
7
8
A
B
B
B
C
B
C
B
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
Me
H
Me
Me
Me
H
Me
Me
Me
NH2
NH2
NH2
NH2
12.5
1.0–3.0
4.0
30.1
5.3
2.25 \50
1.2
\100
I
I
I
I
N
N
N
N
N
S
55
68
20.0
\100
\100
H
Me
Me
Me
H
NH2
34.0
CH(Me)NH2
CH(Me)NH2
NH2
I
–
I
–
T
N
I
9
A
B
B
A
B
B
A
B
B
B
B
A
A
A
A
A
A
B
A
A
A
A
B
A
B
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
H
iPr
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
H
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
H
H
H
H
H
H
H
H
H
H
H
iPr
H
H
H
H
H
H
H
H
H
H
H
H
Me
iPr
iPr
iPr
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
H
H
Me
H
Ph
Ph
Et
Et
Et
H
Me
Me
Me
Et
Et
H
NH2
NH2
NH2
NH2
NH2
NH2
NH2
NH2
NH·Et
NH·Et
NHCH2Ph
NH2
NH2
NH2
NH2
NH2
NH2
NH2
NH2
NH2
NH2
NH2
NH2
NH2
NH2
\25
\25
17.9
I
I
I
I
I
I
I
I
N
N
N
T
T
N
N
S
N
N
S
N
N
S
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
\100
I
I
I
I
I
H
Me
Me
Me
Me
H
H
Me
H
\25
\25
13.4
15.1 \100
Me
Me
Me
H
Me
Me
Me
Me
Me
Et
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Et
Me
Me
Me
Me
Me
Me
Me
Me
Me
12.4 \100
64.3
B25
\100
10
61.9 \100
\50 \50
15.9
I
30.4
\20
\20
\20
\20
\20
67.0
53
92
53
36
$30
$15
\50
\20
\20
H
Me
Me
H
Me
H
8.6
S
I
\20
\20
\100
\100
58
\12.5
\25
–
T
N
S
S
S
T
T
S
T
S
$20
\50
34
37
\25
45.1
61.6
35
iPr
iPr
iPr
Me
iPr
H
H
H
Me
H
H
Me
Me
H
H
H
12
33
51
34.3
H
H
H
iPr
H
\100
\25
53
a Where A, B and C are as in main text, and C has epimeric configuration at C-8.
b Where superscripts a and e refer to axial or equatorial orientation (respectively) of R on bicyclooctanes or octenes.
c IC50, minimum inhibitory concentration (mM) required to reduce viral plaques by 50% (I=inactive).
d Qualitative assessment from plaque inhibition experiments (N=non toxic; S=slightly toxic; T=toxic).

7506
J. A. Miller et al. / Tetrahedron Letters 42 (2001) 7503–7507
6-Et group in type B (entry 11) resulted in a drop of
about 10-fold, whilst a further identical substitution at
C-7, in either series (entries 12 and 13), again abolished
activity. In the same vein, substitution of Me for H at
C-1 (entries 14–16) or Me2 for H2 at C-4 (entry 21)
resulted in a significant reduction in potency. From
these and other data in the Table 1, it became clear that
the symmetry, size and shape of the lead compound 2
was close to optimal, at least for influenza-A. Even the
C-8 epimeric amine (entry 5) showed a reduction in
activity. We also looked at the effect of removing the
Me groups from 2, and found that loss of the C-6 Me
(entry 3) was tolerated, but loss of both Me groups
(entry 4) led to up to a thirty-fold drop in potency.12 In
the rimantadine-type compounds (entries 6 and 7),
potency was close to that of 2, but there was clear
evidence of toxicity to the host.
[3+2] Cycloaddition procedure—preparation of 8-chloro-
6,7-dimethylbicyclo[3.2.1]oct-6-ene: A solution of 3-
chlorocyclohexene (2.0 g, 17 mmol) in dry
dichloromethane (10 ml) was added dropwise to a
stirred suspension of zinc chloride (3.5 g, 25.73 mmol)
in a solution of but-2-yne (1.6 ml, 20.4 mmol) in dry
dichloromethane (50 ml), at 0°C under nitrogen. After
1 h, the reaction mixture was allowed to warm up to
room temperature and stirred for a further 3 h. Water
(50 ml) was cautiously added and the aqueous layer
extracted with chloroform (3×20 ml). The combined
organic extracts were washed with water, sodium
hydrogen carbonate solution and dried (MgSO4). After
filtering the dried extracts and evaporating the solvent,
the crude product was purified by flash column chro-
matography, eluting with light petroleum (bp 40–60°C),
to yield 8-chloro-6,7-dimethylbicyclo[3.2.1]oct-6-ene
(1.72 g, 58.9%) as a light yellow oil. Key data included
1H NMR (CDCl3): l 1.54 (bs, C-6 and C-7 Me), and
4.17 (t, J=5 Hz, 1H, CHCl) ppm. IR (neat) showed
strong, diagnostic band15 for the chlorocyclopentene at
830 cm−1.
Further screening was also undertaken against influ-
enza-B, and against RSV, which is now recognised to
be the agent responsible for a range of severe, flu-like
conditions that frequently affect very young children.13
Against influenza-B, we confirmed genuine activity
(IC50 520 mM) in only two compounds, 15 and 5. The
RSV screen identified compounds 20 and 23 as the
most potent inhibitors, with IC50s of 8 and 16 mM,
respectively, although 5 was a weaker inhibitor at 34
mM and had the distinction of being the only com-
pound with an IC50 of below 40 mM against all three
viruses. From 20 and 23, it looks as if RSV is tolerant
of substitution at C-1, C-3 and C-4, although further
examples would be required to substantiate this. The
bicyclic amides (see Scheme 3, but not detailed in the
Table 1), were universally inactive, and often quite
toxic. A small number of N-alkyl derivatives of both A
and B series was made and these showed activity,
provided that the amine remained basic, and that the
alkyl group was small, e.g. N-ethyl (entries 17 and 18)
and not N-benzyl (entry 19).
General procedure for exchange at C-8—preparation of
8-azido-6,7-dimethylbicyclo[3.2.1]oct-6-ene:
Sodium
azide (7.54 g, 0.116 mol) and reagent grade anhydrous
zinc chloride (11.8 g, 0.087 mol) were added in
sequence to
a
stirred solution of 8-chlorobicy-
clo[3.2.1]oct-6-ene (10.02 g, 0.058 mol) in dry
dichloromethane (150 ml) under nitrogen at room tem-
perature, and the suspension stirred for a further seven
days. Water (500 ml) was then added and the organic
phase was separated, before being combined with chlo-
roform (3×100 ml) washings of the aqueous phase. The
combined organic extracts were then dried (MgSO4)
and filtered, and the solvents evaporated on a rotary
evaporator. After careful flash chromatography, using
light petroleum (bp 40–60°C), 8-azidobicyclo[3.2.1]oct-
6-ene (6.6 g, 63%) was isolated as a clear colourless oil.
1
Key data included H NMR (CDCl3): l 3.82 (t, J=5
Hz, 1H, CHN3) ppm. IR (neat) showed a very strong
azide absorption at 2095 cm−1.
Acknowledgements
The authors are grateful to Dr. Margaret Tisdale for all
anti-viral assay data and for valued advice.
The lead compounds from the in vitro influenza screens
(entries 2 and 15, for flu-A and flu-B, respectively) were
assessed in vivo, in standard mouse models. Compound
15 was found to be essentially inactive versus influenza-
B/Lee. In tests run under the same in vivo conditions,
compound 2 was equiactive orally with amantadine
against influenza-A/Sweden, influenza-A/Bell and influ-
enza-A/Okuda, and inactive against two amantadine
resistant strains of the latter two viruses. These data are
suggestive that 2 and amantadine have a similar mode
of action and potency, a situation more recently
described for BL1743.14 Cross-resistance with aman-
tadine is a serious flaw in a potential anti-flu drug and
is sufficient to deter further development of lead com-
pounds, such as 2.1b
References
1. (a) Oxford, J. S. In NATO ASI Series A, Life Sciences; de
Clerq, E.; Walker, R. T., Eds. Targets for the Design of
Antiviral Agents; Plenum Press, 1983; Vol. 73, pp. 159–
176; (b) Meanwell, N. A.; Krystal, M. Drug Discovery
Today 1996, 1, 316–324 (Part 1) and 1996, 1, 388–397
(Part 2).
2. Blocking the ion channel of the M-2 protein is the main
role. See: (a) Hay, A. J.; Wolstenholme, A. J.; Skehel, J.
J.; Smith, M. H. EMBO J. 1985, 4, 3021–24; (b) Hay, A.
J. Semin. Virol. 1992, 3, 21–30; (c) Ref. 1b, p. 321.

J. A. Miller et al. / Tetrahedron Letters 42 (2001) 7503–7507
7507
3. (a) Swallow, D. L. Progr. Med. Chem. 1971, 8, 119–165;
(b) Hoffman, C. E. Antibiot. Chemother. 1980, 27, 233–
250.
4. Whitney, J. G.; Gregory, W. A.; Kauer, J. C.; Roland, J.
R.; Snyder, J. A.; Benson, R. E.; Hermann, E. C. J. Med.
Chem. 1970, 13, 254.
9. Piers, E.; Britton, E. W.; Geraghty, M. B.; Keziere, R. J.;
Kido, F. Can. J. Chem. 1975, 53, 2838–2848.
10. Maleki, M.; Miller, J. A.; Lever, Jr., O. W. Tetrahedron
Lett. 1981, 22, 365–368.
11. All compounds were characterised by appropriate analyt-
ical and spectroscopic data.
5. Martinez, A. G.; Vilar, E. T.; Fraile, A. G.; de la Moya
Cerero, S.; Herrero, M. E. R.; Ruiz, P. M.; Subramanian,
L. R.; Gancedo, A. G. J. Med. Chem. 1995, 38, 4474–
4477.
6. Gray, B. D.; McMillan, C. M.; Miller, J. A.; Moore, M.
Tetrahedron Lett. 1987, 28, 235–238.
12. After this study was started, the weak anti-influenza-A
properties of (unsubstituted) 8-aminobicyclo[3.2.1]octane
were discovered in a US Patent (3,439,101 to SmithKline
and French, 9/8/66). 3-Spiropyrolidino bicyclo[3.2.1]-
octanes have been reported as more potent.1,3
13. Stott, E. J.; Taylor, G. Arch. Virol. 1985, 84, 1–52.
14. Kurtz, S.; et al. Antimicrob. Agents. Chemother. 1995, 39,
2204–2209.
7. Miller, J. A.; Ullah, G. M.; Welsh, G. M.; Mallon, P.
Tetrahedron Lett. 2001, 42, 2729–2731.
8. Hess, B. A. J. Am. Chem. Soc. 1971, 93, 1000–1003. The
same behaviour is well known for 7-X norbornenes.
15. Miller, J. A.; Pugh, A. W.; Ullah, G. M.; Welsh, G. M.
Tetrahedron Lett. 2001, 42, 955–959.