Evaluation of short-tether bis-THA AChE inhibitors. A further test of the dual binding site hypothesis

Article abstract of DOI:10.1016/S0968-0896(98)00213-2

To provide a further test of the dual binding site hypothesis proposed for acetylcholinesterase (AChE) inhibitor heptylene-linked bis-(9-amino-1,2,3,4-tetrahydroacridine) A7A, short-tether (ethylene-hexylene) homologs A2A-A6A were prepared. En route to these compounds, convenient and scaleable syntheses of useful pharmaceutical intermediate 9-chloro-1,2,3,4-tetrahydroacridine 3 and A7A were developed. AChE and butyrylcholinesterase (BChE) inhibition assays of A2A-A10A confirm that a seven methylene tether (A7A) optimizes AChE inhibition potency and AChE/BChE selectivity. Finally, these studies indicate that simultaneous binding of alkylene-linked 9-amino-1,2,3,4-tetrahydroacridine dimers to the catalytic and peripheral sites of AChE is possible with a tether length as short as 5 methylenes. Copyright (C) 1999 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(98)00213-2

Bioorganic & Medicinal Chemistry 7 (1999) 351±357
Evaluation of Short-tether Bis-THA AChE Inhibitors.
A Further Test of the Dual Binding Site Hypothesis
Paul R. Carlier,a,* Yi Fan Han,b,* Ella S.-H. Chow,a Crystal P.-L. Li,a Hong Wang,b
Thuy Xuan Lieu,a Hau Sum Wonga and Yuan-Ping Pangc
aDepartment of Chemistry, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong
bDepartment of Biochemistry, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong
cMayo Cancer Center, Department of Pharmacology, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA
Received 20 August 1998; accepted 9 October 1998
AbstractÐTo provide a further test of the dual binding site hypothesis proposed for acetylcholinesterase (AChE) inhibitor heptyl-
ene-linked bis-(9-amino-1,2,3,4-tetrahydroacridine) A7A, short-tether (ethylene±hexylene) homologs A2A±A6A were prepared. En
route to these compounds, convenient and scaleable syntheses of useful pharmaceutical intermediate 9-chloro-1,2,3,4-tetra-
hydroacridine 3 and A7A were developed. AChE and butyrylcholinesterase (BChE) inhibition assays of A2A±A10A con®rm that a
seven methylene tether (A7A) optimizes AChE inhibition potency and AChE/BChE selectivity. Finally, these studies indicate that
simultaneous binding of alkylene-linked 9-amino-1,2,3,4-tetrahydroacridine dimers to the catalytic and peripheral sites of AChE is
possible with a tether length as short as 5 methylenes. # 1999 Elsevier Science Ltd. All rights reserved.
Introduction
is important not only from the perspective of drug
development, but as a further physical test of the dual
binding site hypothesis. If simultaneous binding to the
catalytic and peripheral sites is responsible for the
enhanced AChE inhibition potency of A7A, there
should be a well-de®ned minimum tether length, beyond
which dual site binding is impossible and AChE IC50
rises dramatically.
9-Amino-1,2,3,4-tetrahydroacridine (THA) dimer A7A is
a promising drug candidate for the palliative treatment
of Alzheimer's disease, due to its remarkable enhance-
ment in acetylcholinesterase (AChE) inhibition potency
and AChE/butyrylcholinesterase (BChE) selectivity
relative to THA itself (Table 1).1 The synthesis and
evaluation of A7A was predicated on computational
studies which indicated anity of THA to both the
catalytic site of AChE and to aromatic amino acid resi-
dues Trp279, Tyr70, Phe290 (the `peripheral site') near the
mouth of the active site gorge.2 A series of four homo-
logous THA dimers (AnA) was screened, which showed
a monotonic increase in AChE inhibition potency and
AChE/BChE selectivity from A10A to A7A (Table 1).
Chemical synthesis
Application of the previously described THA alkylation
protocol3 to the synthesis of desired homologs A2A-
A6A was not successful. In the case of A6A the desired
product could be obtained in low yield (37%), but was
accompanied by signi®cant amounts of the intramole-
cular cyclization product 1[n=6] and elimination pro-
duct 2[n=6] (Scheme 1).
A series of THA±toluene heterodimers (AnT, n=3 9)
was also evaluated, which indicated that a spacing of 7
methylene units optimized both potency and selectivity
(Table 2).
Attempts to increase the yield of the A6A product by
variation of the base, reaction temperature, concentra-
tion and solvent were fruitless. For n=3 5, the cycli-
Based on the AnT results, it was predicted that hepty-
lene-linked bis-THA (A7A) is the optimum THA
dimer.1 However to date, this conjecture remains
unproven. Demonstration that A7A is superior to
homologous `short-tether' THA dimers (i.e. AnA, n<7)
zation
and
elimination
pathways
completely
overwhelmed the desired bimolecular substitution reac-
tion, to the point that none of the desired AnA products
could be obtained.
To overcome these diculties we explored reaction of 9-
chloro-1,2,3,4-tetrahydroacridine 3 with 1,n-diamines.
Similar strategies have been successfully applied to the
*Corresponding authors.
0968-0896/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(98)00213-2

352
P. R. Carlier et al. / Bioorg. Med. Chem. 7 (1999) 351±357
Table 1. AChE inhibition potency and selectivity of A7A±A10A and
THAa
Table 2. AChE inhibition potency and selectivity of THA±toluene
heterodimers AnTa
n
AChE IC50 (nM)
Selectivity for AChEb
3
4
5
6
7
8
9
1,400‹140
1,500‹190
610‹23
390‹12
210‹19
nd
0.4
1.4
2.6
11.4
0.5
0.4
20,000‹950
27,000‹4,000
n
AChE IC50 (nM)
Selectivity for AChEb
aData taken from reference 1.
7
8
0.40‹0.025
0.66‹0.20
0.77‹0.11
3.1‹0.75
590‹37
980
520
250
140
0.1
bSelectivity for AChE is de®ned as IC50(BChE)/IC50(AChE).
9
10
THA
ciently high temperature, gave excellent conversion, and
.
facilitated direct crystallization of pure A7A 2HCl (>98
area% by HPLC) from the reaction mixture in 89%
yield during aqueous workup (Scheme 2).
aData taken from reference 1.
bSelectivity for AChE is de®ned as IC50(BChE)/IC50(AChE).
synthesis of alkylene-linked bis-9-aminoacridines4 and
bis-4-aminoquinolines,5 and 3 has been previously sug-
gested3 as a possible starting material for A7A. Reaction
of 3 and 1,7-heptanediamine in the absence of solvent
proceeded to low conversion, despite the application of
mild pressure (150 ꢀC, sealed tube). Non-polar, dipolar
aprotic, and protic solvents were then explored, at
atmospheric pressure and in sealed tube reactions. The
optimum system proved to be re¯uxing 1-pentanol6 at
atmospheric pressure for 40 h, which provided a su-
This amination procedure is easily carried out on a large
scale, in contrast to the previously reported synthesis.
Synthesis of A7A by alkylation of THA3 occurs in high
yield (87%), but requires a careful chromatographic
puri®cation step to remove both residual THA and
cyclization product 1[n=7].7 We have con®rmed that
unless these materials are removed by chromatography,
neither A7A nor A7A 2HCl can be isolated by crystal-
lization. This constraint renders scale-up of the alkyla-
tion reaction dicult.
.
Scheme 1. Synthesis of THA dimers AnA by alkylation of THA.
Scheme 2. Synthesis of THA dimers AnA by amination of 3.

P. R. Carlier et al. / Bioorg. Med. Chem. 7 (1999) 351±357
353
The amination strategy was then applied to synthesis of
the desired short-tether homologs. Moderate to good
Tables 1 and 3). To provide an additional check, enzyme
kinetic studies were performed in rat cortex for A7A
and THA. Both drugs exhibited mixed inhibition, and
the rat cortex KI values (A7A 2.6 nM, THA 104 nM)
matched the previously determined human cerebellum
KI values (A7A 1.4 nM, THA 80 nM)1 rather closely.
.
yields were obtained; bis-hydrochloride salts AnA 2HCl
precipitated directly from the reaction mixture upon
cooling, or upon addition of aqueous HCl. In the course
of developing the protocol described above, we also
developed an improved synthesis of 3. Previously
reported syntheses of 3 proceed via chlorination of the
corresponding 1,2,3,4-tetrahydro-9-acridone 4, typically
in 80±85% yield. Compound 4 is commercially avail-
able, but is moderately expensive. The most practical
synthesis of 4 reported thus far is a modi®ed von Nie-
mentoski reaction of methyl anthranilate and cyclohex-
anone, which proceeds in only 50% yield.8,9 We have
found that isolation of the von Niementoski product 4
is not necessary; under milder conditions (toluene
re¯ux) anthranilic acid and cyclohexanone provide
spiro-carbinolamine ester 510 in 93% yield (Scheme 3).
Review of the data for the complete series of AnA
dimers reveals that as tether length is decreased below 7
methylenes, performance deteriorates (Table 3).
Thus, as anticipated, a tether of 7 methylene units is
indeed optimum for both AChE inhibition potency and
selectivity. Interestingly, new short-tether THA dimers
A5A and A6A still display considerable enhancement in
AChE inhibition potency relative to THA itself
(Table 3), a result which suggests that simultaneous dual
site binding can still occur at these shorter tether
lengths. However, as the tether length decreases further
(n<5), the AChE IC50 increases rapidly (Fig. 1). A4A
and A3A have nearly equivalent potency to THA, sug-
gesting that simultaneous dual site binding is no longer
possible. A2A is considerably less potent than THA,
possibly as a consequence of destabilizing non-bonded
interactions at the bottom of the narrow active site
gorge of AChE.
Direct treatment of 5 with 4 equivalents of phosphorus
oxychloride provided 3 in high yield (96%), for an
overall yield of 89%. This procedure has been carried
out on up to a 60 g scale. Because 3 has also proven
useful in the synthesis of endothelin-A receptor ago-
nists,11 and chemo- and radiosensitizers for cancer
therapy,12 we believe the convenient procedure outlined
above will facilitate further exploration of medicinal
applications of the 1,2,3,4-tetrahydroacridin-9-yl unit.
Thus the anticipated dramatic increase in AChE IC50
with decreasing tether length is indeed observed.
AChE and BChE assay results and discussion
In contrast, inspection of the BChE data for A2A to
A10A reveals no clear trend in inhibition potency
(Table 3, Fig. 1). Whereas AChE IC50 values of A2A-
A10A span a 500-fold range, BChE IC50 values vary
over only a 3-fold range and at best (A2A, A8A) only
equal that of THA itself. These observations are con-
sistent with the lack of a `peripheral site' in BChE,
engendered by the substitution of nonaromatic residues
(Ala279, Asn70, Gln121) in BChE at positions corre-
sponding to the peripheral site of AChE. The fact that
BChE IC50 values do not detectably increase at the
shortest tether lengths (n=2±3) suggests that BChE is
better able than AChE to accomodate steric bulk near
the catalytic site. Such a result is consistent with the fact
.
.
Drugs A2A 2HCl±A10A 2HCl possessed adequate aqu-
eous solubility (ꢁ10 mg/mL) and were assayed for
AChE (rat cortex) and BChE (rat serum) inhibition
potency (Ellman method13) without the need for the
ethanol cosolvent used in the previous study of A7A.
The IC50 values obtained for THA and A7A±A10A dif-
fer somewhat from that previously obtained, although
the same trends are observed: A7A is optimum and
considerably more potent and selective than THA (cf.
Table 3. AChE and BChE inhibitory potency of THA dimers
.
AnA 2HCl and THA HCl
.
Drug
AChE
IC50 (nM)a
BChE
IC50 (nM)b
Selectivity
for AChEc
.
.
A2A 2HCl
711‹25
254‹55
157‹23
28‹5
3.8‹0.4
1.5‹0.3
7.8‹0.9
31‹3
102‹4
152‹17
252‹9
329‹21
119‹6
149‹23
105‹13
155‹25
167‹12
92‹2
0.14
0.60
1.60
11.7
31.6
99.4
13.5
4.9
A3A 2HCl
.
A4A 2HCl
.
A5A 2HCl
.
A6A 2HCl
.
A7A 2HCl
.
A8A 2HCl
.
A9A 2HCl
.
A10A 2HCl
.
THA HCl
40‹6
223‹11
4.2
0.4
Scheme 3. Improved Synthesis of 3. a) 1.2 equiv. cyclohexanone,
2.0 equiv. P2O5, 2.0 equiv. N,N-dimethylcyclohexylamine, 170 ꢀC
(50%). b) 10 equiv. POCl3, re¯ux (85%). c) 1.2 equiv. cyclohexanone,
toluene, Dean±Stark Trap, re¯ux (93%). d) 4 equiv. POCl3, re¯ux
(96%).
aAssay performed using rat cortex homogenate, in the presence of
ethopropazine as a speci®c BChE inhibitor.
bAssay performed using rat serum, in the presence of BW284c51 as a
speci®c AChE inhibitor.
cSelectivity for AChE is de®ned as IC50(BChE)/IC50(AChE).

354
P. R. Carlier et al. / Bioorg. Med. Chem. 7 (1999) 351±357
Figure 1. AChE and BChE inhibition by THA dimers AnA as a function of tether length n (n=number of methylene units, IC50 values from
Table 3). Where no error bars are shown, the error is smaller than the symbol.
ꢀ
ꢀ
1
that BChE requires a larger active site gorge to accom-
odate its natural substrate BCh.14
crystals, mp 143.4±145 C (lit. 155 C10). H NMR and
IR data were consistent with the literature.10
1H NMR (CDCl3): d 1.45±1.55 (m, 4H), 1.78±1.81 (m,
4H), 2.08±2.12 (m, 2H), 4.71 (s, 1H), 6.76 (apparent d,
1H, J=8.3), 6.90 (apparent t, 1H, 7.5 Hz ),7.39 (ddd,
1H, J=1.5, 7.3, 8.1 Hz), 7.906 (dd, 1H, J=1.3, 7.9 Hz);
13C NMR (CDCl3): 22.02, 24.81, 35.59, 90.14, 113.38,
116.26, 119.86, 130.16, 135.28, 145.26, 163.85; MS
Conclusion
This study has provided greatly improved and con-
venient syntheses of 9-chloro-1,2,3,4-tetrahydroacridine
3 and heptylene-linked bis-THA A7A. Short-tether
homologs A2A±A6A have been prepared for the ®rst
time, and AChE and BChE inhibition studies demon-
strate that a tether of 7 methylene units indeed opti-
mizes both AChE inhibition potency and selectivity,
consistent with earlier projections. Finally, when the
tether is too short (<5 methylenes) simultaneous bind-
ing to both the catalytic and peripheral sites of AChE
(`dual-site binding') appears to be impossible.
1
(CI+): 218 (M+1) IR: 3290, 2942, 1682, 1488 cm
9-Chloro-1,2,3,4-tetrahydroacridine (3). Spiro-carbinola-
mine ester 5 (67.0 g, 309 mmol) was added in portions to
phosphorus oxychloride (120 mL, 1.29 mol) (caution:
exotherm). The mixture was heated to re¯ux (oil bath
temperature 120 ꢀC) for 2 h. After cooling to room
temperature the reaction was cautiously added dropwise
to KOH (540 g) in ice water (1000 mL), with rapid stir-
ring. After the addition was complete a yellow solid had
precipitated and the supernatant pH was 10±11.
Methylene chloride (1500 mL) was added to dissolve the
solid, and the aqueous layer was extracted with CH2Cl2
(3Â1000 mL). The combined organic extracts were dried
(MgSO4), ®ltered and the solvent removed in vacuo to
provide 64.0 g (96%) of 3 as a yellow solid, mp 66.5±
67.7 ꢀC (lit. 65±67 ꢀC).15 1H NMR data matched the lit-
erature.15 HPLC indicated 100.0 area % purity (reten-
tion time 20.6 min).
Experimental
THA was obtained from Sigma. 1H and 13C NMR
spectra were recorded at 300 and 75 MHz, respectively.
Chemical ionization (CI) mass spectra were acquired
using CH4 as the reagent gas. Elemental analysis was
performed by the Shanghai Institute of Organic Chem-
istry, Chinese Academy of Sciences (Shanghai, P.R.C.).
Melting points were determined with a Electrothermal
9100 melting point apparatus and are uncorrected.
HPLC analysis was performed on a Lichrosorb1 RP
Select B column (0.4Â25 cm, 5 mm), detection at 245 nm,
¯ow rate 1.0 mL/min, with gradient elution. Elution
program: 90% A at 0 min, 10% A at 30 min (Bu€er
A=0.1% (v/v) aqueous CF3CO2H, Bu€er B=1/4 (v/v)
Bu€er A/CH3CN).
1H NMR (CDCl3): d 1.91±2.01 (m, 4H), 3.03 (t, 2H,
J=5.8 Hz), 3.14 (t, 2H, J=5.6 Hz), 7.55 (ddd, 1H,
J=1.2, 7.0, 8.3 Hz), 7.68 (ddd, 1H, J=1.4, 6.9, 8.4 Hz),
7.99 (apparent d, 1H, J=8.4 Hz), 8.17 (dd, 1H, J=1.00,
8.4 Hz); 13C NMR (CDCl3): d 22.57, 22.61, 27.45, 34.13,
123.63, 125.33, 126.41, 128.57, 128.81, 129.19, 141.39,
146.62, 159.45; MS (CI+): 218 (M+1).
Spiro[2H-3,1-benzoxazine-2,10-cyclohexan]-4(1H)-one (5).
Anthranilic acid (50.0 g, 364 mmol) was dissolved in
100 ml of toluene, and combined with cyclohexanone
(45.3 ml, 437 mmol). The mixture was heated to re¯ux
(Dean±Stark Trap) until one equivalent of water
(6.8 mL) was removed (2 h). Short needle-like crystals
formed on cooling to room temperature. The mixture
was further cooled to 0 ꢀC, the solid collected by ®ltra-
tion, washed with toluene (50 mL) and ethanol (50 mL),
and dried in vacuo, a€ording 73.2 g (93%) of 5 as white
.
Heptylene-linked bis-THA bis-HCl salt (A7A 2HCl). 3
(5.0 g, 23.1 mmol) and 1,7-diaminoheptane (1.5 g,
11.5 mmol) were dissolved in 30 mL of 1-pentanol, and
heated to re¯ux (oil bath temperature 160 ꢀC) for 40 h.
After cooling to room temperature, the mixture was
poured into a 1-L beaker, and 2N HCl (600 mL) was
added to it to precipitate the product. The solid was
collected by suction ®ltration, washed with toluene

P. R. Carlier et al. / Bioorg. Med. Chem. 7 (1999) 351±357
355
.
(50 mL) and dried in vacuo to a€ord 5.82 g (89%) of
Analysis: Calc. for C29H34N4Cl2 2.5H2O: C, 62.81; H,
7.09; N, 10.10; Found: C, 62.87; H, 7.53; N, 9.88.
ꢀ
.
A7A 2HCl as a white solid (mp 107±109.4 C), which
was pure by 1H NMR spectroscopy except for trace
residual toluene (<0.8 wt%). HPLC analysis revealed
98.2 area% purity (retention time 24.2 min).
.
Butylene-linked bis-THA bis-hydrochloride (A4A 2HCl).
1,4-Diaminobutane (213 mL, 2.12 mmol) and 3 (1.0 g,
4.67 mmol) were combined as above, to yield 723 mg of
A4A 2HCl (65%), which was pure by H and 13C NMR
1
.
Recrystallization from warm 10% (v/v) aqueous MeOH
.
(50 mL) a€orded 4.72 g (72%) of A7A 2HCl, increasing
analysis. Recrystallization from 10% MeOH±H2O
a€orded 586 mg (53%) of the desired product as o€-
white crystals. 1H NMR (D2O): d 1.63 (s, br, 12H), 2.03
(s, br, 4H), 2.48 (s, br, 4H), 3.65 (s, br, 4H), 7.14±7.20
(m, 4H), 7.56 (d, J=7.8, 2 H), 7.62 (t, J=8.8, 2H); 13C
NMR (CD3OD): d 22.74, 23.86, 25.97, 29.13, 30.48,
49.07, 114.03, 118.08, 121.49, 127.04, 127.19, 134.70,
140.86, 153.00, 158.19; MS (CI+): Calc. for C30H34N4:
450. Found: 451 (M+1). mp: 271 ꢀC (decomp.). HPLC:
20.9 min, 99.8 area% Elemental Analysis: Calc. for
the HPLC area% purity slightly (to 99.2%) and the
melting point dramatically (mp 145.7±148.1 ꢀC). Large
Scale Synthesis: 3 (25.0 g, 115 mmol) and 1,7-diamino-
heptane (7.50 g, 57.6 mmol) were reacted as above, giv-
.
ing after recrystallization 21.7 g (67%) of A7A 2HCl
(mp 148.3±150 ꢀC, 99.8 area% purity by HPLC, ele-
mental analysis (below) matches dihydrate).
1H NMR (D2O): d 1.38 (s, br, 6H), 1.66±1.79 (m, 12H),
2.34 (s, br, 4H), 2.71 (s, br, 4H), 3.61 (t, 4H, J=6.8 Hz),
7.22 (apparent t, 2H, J=7.8 Hz), 7.30 (apparent d, 2H,
J=8.4 Hz), 7.56 (apparent t, 2H, J=7.7 Hz), 7.80
(apparent d, 2H, J=8.7 Hz) [HOD peak as reference at
4.8 ppm.]; 13C NMR (CD3OD): d 22.67, 23.81, 25.74,
28.44, 30.14, 30.75, 32.29, 49.90, 113.65, 117.85, 120.96,
127.14, 127.31, 134.90, 140.61, 152.52, 158.76; MS
(CI+): Calc. for C33H41N4 (M+1): 493. Found: 493.
.
C30H36N4Cl2 0.5 H2O: C, 67.66; H, 7.00; N, 10.52.
Found: C, 67.78; H,6.89; N, 10.60.
.
Pentylene-linked bis-THA bis-hydrochloride (A5A 2HCl).
1,5-Diaminopentane (271 mL, 2.31 mmol) and 3 (1.0 g,
4.62 mmol) were combined as above, to yield 834 mg of
A5A 2HCl (78%), which was pure by H and 13C NMR
analysis. Recrystallization from 10% MeOH±H2O
a€orded 677 mg (54.6%) of the desired product as o€-
white crystals. H NMR (D2O): d 1.41±1.57 (m, 2H),
1.61±1.81 (m, 12H), 2.28 (s, br, 4H), 2.61 (s, br, 4H),
3.63±3.73 (m, 4H), 7.18 (t, J=7.8 Hz, 2H), 7.26 (d,
J=8.4 Hz, 2H), 7.59 (t, J=7.7 Hz, 2H), 7.75 (d,
J=8.7 Hz, 2H); 13C NMR (CD3OD): d 23.12, 23.94,
25.46, 25.98, 31.88, 31.96, 49.58, 115.05, 119.37, 124.11,
125.91, 126.07, 132.058, 142.81, 155.48, 156.17; MS
(CI+): Calc. for C31H36N4: 464. Found: 465 (M+1).
mp: 241 ꢀC (decomp). HPLC: 22.4 min, 98.7 area%.
1
.
.
Elemental Analysis: Calc. for C33H42N4Cl2 2H2O: C,
65.88; H, 7.71; N, 9.31; Cl, 11.78. Found: C, 66.03; H,
7.88; N, 9.31; Cl, 11.80.
1
.
Ethylene-linked bis-THA bis-hydrochloride (A2A 2HCl).
Ethylene diamine (153 mg, 2.55 mmol) and 3 (1.0 g,
4.64 mmol) were combined as above, to yield 828 mg of
1
A2A 2HCl (66%), which was pure by H and 13C NMR
.
analysis. Recrystallization from 10% MeOH±H2O a€or-
ded 667 mg (53%) of the desired product as o€-white
1
.
crystals. H NMR (D2O): d 1.61 (s, br, 8H), 2.13 (s, br,
Elemental Analysis: Calc. for C31H38N4Cl2 2.1 H2O: C,
64.71; H, 7.39; N, 9.74. Found: C, 64.38; H, 6.99; N,
9.90.
4H), 2.41 (s, br, 4H), 3.89 (s, br, 4H), 7.07 (d, J=8.4 Hz,
2H), 7.14 (t, J=7.7 Hz, 2H), 7.41 (d, J=8.6 Hz, 2H),
7.53 (t, J=7.7 Hz, 2H); 13C NMR (D2O): d 22.23, 23.34,
26.53, 29.94, 49.25, 114.39, 117.02, 120.61, 124.76,
127.51, 135.06, 138.36, 153.18, 158.27 ppm; MS (CI+):
Calc. for C28H30N4: 422. Found: 423 (M+1); mp:
274 ꢀC (decomp.); HPLC: 18.8 min, 96.0 area%. Ele-
.
Hexylene-linked bis-THA bis-hydrochloride (A6A 2HCl).
1,6 Diaminohexane (2.68 g, 23.0 mmol) and 3 (10.0 g,
46.08 mmol) were combined as above, to yield 11.06 g
(87%) A6A 2HCl, which was pure by H and 13C NMR
analysis. Recrystallization from 10% MeOH±H2O
a€orded 8.94 g (71%) of the desired product as o€-
white crystals. H NMR (D2O): d 1.41 (s, br, 4H), 1.71
1
.
.
mental Analysis: Calc. for C28H32N4Cl2 4.1H2O: C,
59.07; H, 7.12; N, 9.84. Found: C, 58.69; H, 7.14; N,
9.71.
1
(m, 12H), 2.23 (s, br, 4H), 2.61 (s,br, 4H), 3.58 (t,
J=6.7 Hz, 4H), 7.21 (m, 4H), 7.55 (apparent t,
J=7.7 Hz, 2Hz), 7.74 (apparent d, J=8.6 Hz, 2H); 13C
NMR (CD3OD): d 22.61, 23.76, 25.73, 28.19, 30.09,
32.21, 113.62, 117.80, 120.88, 127.12, 127.27, 134.86,
140.50, 152.45, 158.71; MS (CI+): Calc. for C32H38N4:
478. Found: 479 (M+1). mp: 153.3±154.6 ꢀC. HPLC:
23.6 min, 99.5 area%. Elemental Analysis: Calc. for
.
Propylene-linked bis-THA bis-hydrochloride (A3A 2HCl).
1,3-Diaminopropane (209 mL, 2.51 mmol) and 3 (1.1 g,
5.07 mmol) were combined as above, to yield 821 mg of
1
A3A 2HCl (65%), which was pure by H and 13C NMR
.
analysis. Recrystallization from 10% MeOH±H2O a€or-
ded 657 mg (52%) of the desired product as o€-white
1
crystals. H NMR (CD3OD): d 1.7±1.9 (m, 8H), 2.29
.
(quintet, J=6.0 Hz, 2H), 2.4±2.5 (m, 4H), 2.75±2.85 (m,
4H), 4.02 (t, J=6.0 Hz, 3H), 7.47 (ddd, J=1.5, 6.9,
8.6 Hz, 2H), 7.72 (dd, J=1.2, 8.5 Hz, 2H), 7.80 (ddd,
J=1.29, 7.0, 8.3 Hz, 2H), 8.16 (apparent d, J=8.6 Hz,
2H); 13C NMR (CD3OD): d 22.81, 23.76, 25.83, 30.37,
33.16, 46.98, 114.34, 117.69, 121.64, 127.95, 127.71,
135.06, 140.97, 152.90, 159.02; MS (CI+): Calc. for
C29H32N4: 436. Found: 437 (M+1). mp: 269 ꢀC
(decomp.). HPLC: 19.3 min, 99.9 area%. Elemental
C32H40N4Cl2 2H2O: C, 65.41; H, 7.55; N, 9.53. Found:
C, 65.76; H, 7.51; N, 9.48.
.
Octylene-linked bis-THA bis-hydrochloride (A8A 2HCl).
Using the previously described THA alkylation proto-
col,3 THA (1.0 g, 3.96 mmol) and powdered KOH
(1.55 g, 27.7 mmol) were dissolved in 20 mL of DMSO,
and placed under nitrogen. This mixture was heated
periodically with a heat gun for 30 min, avoiding

356
P. R. Carlier et al. / Bioorg. Med. Chem. 7 (1999) 351±357
.
Analysis: Calc. for C36H48N4Cl2 2.3H2O: C, 66.61; H,
8.17; N, 8.63. Found: C, 66.66; H, 8.15; N, 8.56.
boiling. After cooling to room temperature, 1,8-dibro-
mooctane (364 mL, 1.98 mmol) was added by syringe.
The homogeneous brown solution was stirred for 30 h
at room temperature, quenched by pouring into 100 mL
of H2O, and extracted with ethyl acetate (4Â50 mL).
The combined extracts were washed with H2O
(2Â30 mL), dried (MgSO4), ®ltered and the solvent
removed in vacuo to give a brown liquid; puri®cation by
¯ash chromatography (8% MeOH±CH2Cl2 with 7 mL
conc. NH3 per L) gave 805 mg of A8A (80%), which
was pure by 1H and 13C NMR analysis. The bis-
HCl salt was prepared in methanol by treatment with
HCl gas. Recrystallization from 10% MeOH±H2O
a€orded 598 mg (52%) of the desired product
Enzyme preparation and AChE/BChE inhibition studies.
AChE and BChE enzyme preparations were prepared
from cortex and serum respectively of decapitated rats.
Frontal cortex (brain dissected on ice) was homogenized
in sodium phosphate bu€er (39 vol. 75 mM, pH 7.4).
Rat serum was obtained by centrifugation of blood
(3500Âg, 10 min.). The cholinesterase assays were per-
formed using the colorimetric method of Ellman,13 with
minor modi®cation. For determination of AChE inhi-
bition, cortex homogenate was preincubated for 5 min
with ethopropazine (0.1 mM), a selective inhibitor of
BChE. Similarly, for determination of BChE inhibition,
serum was preincubated with BW284c51 (0.01 mM), a
selective inhibitor of AChE. A mixture of 4 mL con-
taining acetylthiocholine iodide (0.3 mM) or butrylthio-
choline iodide (0.4 mM), 1 mL sodium phosphate bu€er
(0.1 mM, pH 7.4), a solution of the compound being
tested (0.1 mL), and homogenate or serum (0.1 mL) was
incubated at 37 ꢀC for 8 min. The reaction was termi-
nated by the addition of sodium dodecyl sulfate (3% w/
v, 1 mL), after which the 5.50-dithio-bis(2-nitrobenzoic
acid) indicator (0.2% w/v, 1 mL) was added. Enzyme
activity was determined by measuring the absorbance at
420 nm after 10 min, relative to the drug-free control.
Triplicate measurements were performed at typically a
total of 8 drug concentrations; IC50 values were deter-
1
.
A8A 2HCl as yellow crystals. H NMR (D2O): d 1.33 (s,
br, 8H), 1.64 (s, br, 4H), 1.78 (s, br, 8H), 2.27 (s, br,
4H), 2.68 (s, br, 8H), 3.53 (t, J=7.1 Hz, 4H), 7.18
(apparent t, J=7.9 Hz, 2H), 7.24 (apparent d,
J=8.4 Hz, 2H), 7.52 (apparent t, J=7.7 Hz, 2H), 7.71
(apparent d, J=8.6 Hz, 2H); 13C NMR (CD3OD): d
22.61, 23.70, 25.56, 28.13, 30.16, 30.60, 32.08, 113.64,
117.53, 127.19, 135.06, 140.40, 152.54, 158.43; MS
(CI+): Calc. for C38H42N4: 506, Found: 507 (M+1);
mp: 148.7±150.9 ꢀC; HPLC: 25.9 min., 98.4 area%; Ele-
.
mental Analysis: Calc. for C34H44N4Cl2 2.1H2O: C,
66.13; H, 7.87; N, 9.07. Found: C, 65.84; H, 7.56; N,
9.39.
.
Nonylene-linked bis-THA bis-hydrochloride (A9A 2HCl).
Use of 1,9-dibromononane in the A8A procedure
.
described above provided 585 mg of A9A 2HCl.
mined from a plot of Enzyme Activity versus
.
log[drug]. The AChE Ki values of A7A 2HCl and
.
THA HCl, and modes of inhibition were determined
using Lineweaver±Burke plots.
Recrystallization from 10% MeOH±H2O a€orded
414 mg (35%) of the desired product as yellow crystals.
1H NMR (D2O): d 1.30±1.35 (m, 10H), 1.63±1.65 (m,
4H), 1.81 (s, br, 8H), 2.29 (s, br, 4H), 2.73 (s, br, 4H),
3.52 (t, J=7.0 Hz, 4H), 7.20 (apparent t, J=7.8 Hz,
2H), 7.23 (apparent d, J=8.4 Hz, 2H), 7.55 (apparent t,
J=7.7 Hz, 2H), 7.71 (apparent d, J=8.6 Hz, 2H); 13C
NMR (CD3OD): d 22.40, 23.50, 25.39, 28.06, 29.93,
30.50, 30.77, 31.96, 113.41, 117.33, 120.81, 126.99,
134.82, 140.20, 152.26, 158.26; MS (CI (NH3)): Calc.
for C35H44N4: 520. Found: 519 (M 1); HPLC:
26.5 min, 96.1 area%; Elemental Analysis: Calc. for
Acknowledgements
The authors thank Dr. Bizeng Zhan for HPLC analyses,
and the Research Grants Council (H. K.) for ®nancial
support of this work (Competitive Earmarked Research
Grant HKUST 6156/97M).
References and Notes
.
1. Pang, Y.-P.; Quiram, P.; Jelacic, T.; Hong, F.; Brimijoin, S.
J. Biol. Chem. 1996, 271, 23646.
2. Pang, Y.-P.; Kozikowski, A. P. J. Comput.-Aided Mol.
Design 1994, 8, 669.
3. Pang, Y.-P.; Hong, F.; Quiram, P.; Jelacic, T.; Brimijoin, S.
J. Chem. Soc. Perkin Trans. I 1997, 171.
4. Galanakis, D.; Davis, C. A.; Herrero, B. D. R.; Ganellin,
C. R.; Dunn, P. M.; Jenkinson, D. H. J. Med. Chem. 1995, 38,
595.
C35H46N4Cl2 2H2O: C, 66.76; H, 8.00; N, 8.90. Found:
C, 66.77; H, 7.62; N, 9.30.
.
Decylene-linked bis-THA bis-hydrochloride (A10A 2HCl).
1,10-Diaminodecane (397 mg, 2.30 mmol) and 3 (1.0 g,
4.61 mmol) were combined as above, to yield 909 mg of
A10A 2HCl (65%), which was pure by 1H and 13C
.
NMR analysis. Recrystallization from 10% MeOH±
H2O a€orded 416 mg (30%) of the desired product as
5. Galanakis, D.; Davis, C. A.; Ganellin, C. R.; Dunn, P. M.
J. Med. Chem. 1996, 39, 359.
1
o€-white crystals H NMR (CD3OD): d 1.31±1.40 (m,
6. Ganellin previously reported the ecacy of re¯uxing 1-
pentanol solvent for the amination of 4-chloroquinoline by
1,n-diamines (ref. 5). Use of 1-butanol in amination reactions
of 1 resulted in signi®cantly longer reaction times.
7. Order of elution on silica gel (Rf): 1[n=7] (0.47), A7A
(0.42), THA (0.32), in 0.7/9.9/89.4 (v/v) conc. NH4OH/
MeOH/CH2Cl2 (optimized eluent).
8. Finlander, P.; Fischer, H. P.; Pedersen, E. B. Heterocycles
1985, 23, 1437.
9. Other higher yielding syntheses have been reported, but
depend on commercially unavailable starting materials: a)
12H), 1.81 (quintet, J=7.3 Hz, 4H), 1.90±1.99 (m, 8H),
2.68 (t, J=5.7 Hz, 4H), 3.01 (t, J=5.4 Hz, 4H), 3.93 (t,
J=7.4 Hz, 4H), 7.57 (ddd, J=1.6, 6.9, 8.6 Hz, 2H), 7.76
(dd, J=1.3, 8.6 Hz, 2H), 7.83 (ddd, J=1.2, 6.9, 8.3 Hz,
2H), 8.37 (apparent d, J=8.7 Hz, 2H); 13C NMR
(CD3OD): d 22.41, 23.54, 25.44, 28.76, 29.88, 30.84,
31.06, 32.10, 49.71, 113.40, 117.62, 120.69, 126.85,
127.04, 134.63, 140.38, 152.26, 158.54; mp: 128.6±
130.0 ꢀC; MS(CI+): Calc. for C36H46N4: 534. Found:
535 (M+1); HPLC: 27.8 min, 96.2 area%; Elemental

P. R. Carlier et al. / Bioorg. Med. Chem. 7 (1999) 351±357
357
Chong, R. J.; Siddiqui, M. A.; Snieckus, V. Tetrahedron Lett.
1986, 27, p. 5323 (70% yield). b) Reference 11 (55% yield).
10. Badger, G. M.; Hinde, R. W.; Matthews, W. E.; Spots-
wood, T. M. Australian J. Chem. 1963, 16, 732.
11. Cheng, X.-M.; Lee, C.; Klutchko, S.; Winters, T.; Rey-
nolds, E. E.; Welch, K. M.; Flynn, M. A.; Dougherty, A. M.
Bioorg. Med. Chem. Lett. 1996, 6, 2999.
13. Ellman, G. L.; Courtney, K. D.; Andres, V. J.; Feath-
erstone, R. M. Biochem. Pharm. 1961, 7, 88.
14. Harel, M.; Sussman, J. L.; Krejci, E.; Bon, S.; Chanal, P.;
Massoulie, J.; Silman, I. Proc. Natl. Acad. Sci. USA 1992, 89,
10827.
15. Michalson, E. T.; D'Andrea, S.; Freeman, J. P.; Szmusz-
kovicz, J. Heterocycles 1990, 30, 415.
12. Papadopoulou, M. V.; Rosenzweig, H. S.; Doddi, M.;
Bloomer, W. D. Oncol. Res. 1994, 6, 439.