Life Sciences 79 (2006) 559 569

www.elsevier.com/locate/lifescie

Quaternary derivatives of granatanol diesters: Potent, ultrashort acting

non-depolarizing neuromuscular relaxants
Laszlo Gyermek , Chingmuh Lee, Young-Moon Cho, Nguyen Nguyen
Harbor-UCLA Medical Center, 1000 W. Carson St., Torrance, California, 90509, USA
Received 28 February 2005; accepted 30 January 2006

Abstract

The purpose of this study was to explore the feasibility of utilizing the granatanol: N-methyl [9-azabicyclo (3.3.1) nonane] 3-alpha-ol as
the terminal group in a series of new bisquaternary azabicycyclic diester-type neuromuscular blocking agents. Fifty two bisquaternary
ammonium derivatives of several dicarboxylic acid esters of granatanol and three similar derivatives of pseudo granatanol have been
investigated for neuromuscular blocking (NMB) potency (ED50 s), onset and recovery of action and for cardiovascular side effects. All
agents were studied first in anesthetized rats, and selected agents were subjected to further pharmacodynamic testing in rabbits, juvenile pigs,
cats, dogs and monkeys. One agent was tested in continuous i.v. infusion mode in comparison with its corresponding tropine diester and the
aminosteroid muscle relaxant, rocuronium. Several new and highly potent NMB granatanol derivatives are described, which are largely
similar in NMB potency to the previously described tropine: N-methyl [8-azabicyclo (3.2.1)] 3-alpha-ol diester derivatives. The majority of
the presently described granatanol derivatives displayed ultrashort onset and duration of actions. In that respect some of these agents proved
to be the fastest and shortest acting non-depolarizing muscle relaxants described so far. On the negative side, many, but not all, granatanol
derivatives produced cardiovascular side effects: e.g. changes in heart rate and blood pressure. Like with the similar tropinyl diester
derivatives, cardiac vagal block was present with the majority of these agents as assessed in the rat, pig and cat. Few glutaryl, fumaryl and
cyclobutane (trans) 1,2-dicarboxylyl granatanol diesters quaternized with disubstituted benzyl halides, bearing p-acyloxy radicals, showed
excellent NMB profile. In these derivatives, however, the rapid decomposition of the p-acyloxy groups leads to formation of toxic quinone
methene metabolites which precludes their further pharmaceutical development. The pseudo granatanol derivatives were less potent in the rat
than the corresponding granatanols and were not further investigated. We conclude that the 9-azabicyclo (3.3.1) nonane (granatane) ring
system can successfully replace the similar 8-azabicyclo (3.2.1) octane (tropane) ring system in building potent, utrashort acting NMB
agents.
2006 Elsevier Inc. All rights reserved.

Keywords: Neuromuscular block; Quaternary ammonium derivatives of granatanol: N-methyl [9-azabicyclo (3.3.1) octane] 3-alpha-ol diesters; Ultrashort acting
myoneural relaxants; Structureactivity relationships of new myoneural relaxants

Introduction
In the search for new, ideal neuromuscular blocking (NMB)
drugs with rapid onset and ultrashort duration of action, various
chemical classes of agents have been developed in the last few
Corresponding author. 25 Oceancrest Ct., Rancho Palos Verdes, CA 90275,
USA. Tel.: +1 310 544 4185; fax: +1 310-544 9985.
E-mail address: laslogy@cox.net (L. Gyermek).
0024-3205/$ - see front matter 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.lfs.2006.01.038
decades (Tuba et al., 2002; Gyermek, 2002, 2005). Since the 9-
azabicyclo (3.3.1) nonane 3-alpha-ol (granatanol) ring is nearly
isosteric with the similarly configured 8-azabicyclo (3.2.1)
octane 3-alpha-ol (tropine) (Fig. 1) that has been extensively
utilized previously in producing ultrashort acting NMB agents
(Gyermek et al., 2002), we synthesized and pharmacologically
evaluated a series of new bisquaternary granatanol diesters.
Based on our previous, favorable experience with disubstituted
benzyl quaternary derivatives of tropinyl diesters (Gyermek et
al., 2002), we placed emphasis in this study on similar
quaternaries of different granatanol diesters. A total of 52
560 L. Gyermek et al. / Life Sciences 79 (2006) 559569

Fig. 1. Three dimensional wire frame models of tropine and granatanol constructed by the MMP program (Chem SW, Fairfield, CA). The two energy minimized
samples (tropine on the left and granatanol on the right) indicate similar molecular volume (e.g. 142.9 and 160.53 for tropine and granatanol, respectively) and shape.
It should be noted that the extension by one CH2 group of the granatane ring is pointing away from the functional groups of the onium group (e.g. the protonated N
atom with its axial CH3 radical and, in case of the quaternary forms, the equatorially oriented aralkyl group (for the schematic representation of the approximate
orientation of substituents on the onium groups see Fig. 2).

such diesters, quaternized with disubstituted or trisubstituted
benzyl groups were studied.
Materials and methods
Chemistry
The synthesis and chemical analysis of the compounds was
reported in US Patents 5,990.124 and 6,376,510 B1. Fourteen
different dicarboxylic acid esters of granatane 3-alpha-ol and
one dicarboxylic acid (e.g. glutaric) ester of granatane 3-beta-ol
have been synthesized first. Nine different disubstituted benzyl
quaternizing groups and three trisubstituted benzyl (e.g. 3,4,5-
triacetoxy benzyl; 3,5-dimethoxy-4-acetoxy benzyl; and 3,5-
dimethoxy-4-propionyl-oxy benzyl) quaternizing groups were
employed to obtain the final products. A total of 44 such
diesters, quaternized with disubstituted benzyl groups, and
8 granatanol diesters quaternized with trisubstituted benzyl
groups were prepared and studied. The general structural
formula of these agents are shown in Fig. 2.

a) 3 alpha ol diesters

Me R1
R1 + +
Me 2 Br - N
N

O O
C R2 C

O O

b) 3 beta ol diesters Me
R1 + R1
+ Me
N N
2 Br -

O O
C R2 C

O O

Fig. 2. General formula of quaternary granatanol diesters where R1 = disubstituted benzyl group; and R2 = alkane, alkene, cycloalkane or aromatic groups.
 L. Gyermek et al. / Life Sciences 79 (2006) 559569
Computational chemistry
Basic, semi-quantitative molecular modeling calculations
were made using the Molecular Modeling Pro (Chem SW,
Fairfield, CA) program only with respect to a) the volume and
shape of the terminal onium groups containing the granatanol,
in comparison with the corresponding tropanering system, b)
the sizes and distribution of electric charges and c) the relative
hydrophilic vs. lipophilic surface areas of these terminal groups.
Inter-onium distance calculations of NMB agents, even with
more elaborate computational methods are largely influenced
by changes in the conformation of flexible linkers, therefore
their calculated values remain ill defined (Lee, 2003). Thus their
use creating useful SAR databases is quite limited. For the
above reason we will not discuss inter-onium lengths vs.
functional parameters in this paper.
Pharmacology
With Institutional Review Board approvals of the Los
Angeles Biomedical Research Institute of HarborUCLA
Medical Center, Torrance, CA and of the National Taiwan Uni-
versity, Taipei, we evaluated the agents as follows: White, adult
male rats were anesthetized with 4050 mg/kg pentobarbital
sodium i.p.; rabbits with 3050 mg/kg pentobarbital sodium i.v.
Adult male and nonpregnant female cats and mongrel dogs were
anesthetized with 2530 mg/kg pentobarbital and 60 mg/kg
alpha chloralose i.p. The trachea was cannulated and artificial
ventilation was given to maintain normocarbia with continual
monitoring of the end tidal CO2 (except in the rat). Juvenile,
Duroc pigs (811 kg) were first anesthetized with halothane (1
2%) in a Lucite box and, after orotracheal intubation,
supplemented with 2030 mg/kg pentobarbital sodium i.v.
Adult Macaca Cyclopis Swinhoe monkeys of both sexes,
between 5.8 and 8kg body weight were sedated with ketamine
7 mg/kg i.m., anesthetized with halothane, and ventilated with a
halothaneO2 mixture. The trachea was intubated; one periph-
eral vein was cannulated trans-cutaneously, and an arterial line
was established in one extremity through a trans-cutaneous
cannula. After the experiments the monkeys were properly
recovered before readmitting them to the holding facility.
Maintenance of anesthesia was provided in most species
with additional (515 mg/kg) incremental doses of pentobarbi-
tal. Neuromuscular blocking drugs were given i.v. via a catheter
inserted into an external jugular or femoral vein. Neuromuscular
responses were monitored electromyographically (EMG) (Lee
et al., 1978) and/or mechanomyographically (MMG) at the
anterior tibial muscle using supramaximal train-of-four (TOF)
stimuli every 12 s. The stimuli were delivered via bipolar needle
electrodes to the sciatic nerve or to its common peroneal branch
via a Grass model S88 Laboratory Stimulator. Arterial blood
pressure was transduced via a cannulated common carotid
artery with the exception of monkeys on which percutaneous
arterial lines were established. For monitoring heart rate
changes, pulse pressure responses were interfaced with a
Beckman Instruments Model 9857-B cardiotachograph. Cardi-
ac vagal block was evaluated in rats, cats and pigs based on the
561
inhibition of the bradycardic responses to peripheral stimulation
of the cut right vagus nerve with 1520 Hz supramaximal
impulses of 0.2 ms duration, for 510 s, delivered every 1.5 min.
Recording of all above responses were made on a Beckman
Instruments Model 6011 polygraph. ED50 doses of neuromus-
cular block were determined by graphic interpolation from
plotting the responses to incremental single doses of each
animal against the doses on a log dose-probit chart. The
incremental doses were given in 5min intervals (except in the
case of few slower and longer acting agents, where appropriate
longer intervals were kept between the incremental doses to
avoid cumulative response). Mean values per group (N = 310)
and standard errors of the mean (SEM) were calculated. Onset
of action (time from the end of the rapid injection of an ED 80
90 NMB dose until maximal NMB) and 2575% twitch
recovery indices were determined to define speed of onset and
recovery. Cardiovascular side effects were defined by the peak
cardiac vagal block (in percent) following these injections. TOF
fade (ratio of T4 over T1 response) was recorded to indicate the
non-depolarizing-type of the NMB. In pigs one agent: #7 was
evaluated in the constant infusion administration mode starting
5min after the full recovery from a single 8090% paralyzing
bolus dose. (For comparison another group of pigs were given
the initial NMB ED80 dose and the maintenance mode of
infusion of rocuronium and the tropine analog of agent #7 were
administered to maintain an 8090% NMB for an hour).
Thereafter NMB was allowed to spontaneously recover.
Maintenance doses for 80% depression of the first twitch of
TOF response were calculated for 10-min epochs; the recovery
indices were determined: a) following the initial single dose and
b) during the spontaneous recovery after the 60 min infusion;
and compared. Selection of several animal models was an
important aspect of the study and it was based on scientific as
well as practical criteria. The primary use of rats for the
screening stage of investigation was based on our favorable,
long term experience with this species and particularly the EMG
recording technology developed by one of us (Lee et al., 1978).
Although a preliminary SAR correlation estimation was
obtainable on this species, the clinically important side effects
and pharmacokinetic profile of selected agents required further,
numerous preclinical experiments utilizing animals, whose
data-profiles can be more projected to man than data obtained
on a small rodent. The present consensus seems to indicate that
for the prediction of an onset and duration of action profile
in humans subhuman primates would be the best appropriate
species. Humanoid apes remain practically uninvestigated.
Among the relatively abundant (and less protected) monkey
species: macacque monkeys have been included in the
preclinical evaluation of all recent, clinically introduced new
NMB agents. Selection of some, intermediate species
between the rat and monkey was necessary because of the
highly variable potency and side effects characteristics
which exist among different animal models. For the study,
particularly of the cardiovascular and autonomic nervous side
effects the cat was often used in the past, but because of its
highly protected status, which is shared by the dog, these
species were utilized in this study only very sparingly with a
562 L. Gyermek et al. / Life Sciences 79 (2006) 559569

Table 1
NMB profile of bisquaternary granatanol diesters in the rat
No. R1N R2 ED50 M/kg Onset min RI min CVB % N
1 3,4-diacetoxy B (CH2) 0.33 (0.1) 0.6 (0.1) 0.55 (0.1) 0 5
2 " (CH2)2 0.19 (0.06) 0.6 (0.05) 0.55 (0.05) 70 (10) 6
3 " (CH2)3 0.21 (0.03) 0.6 (0.05) 0.45 (0.05) 0 8
4 " (CH2)4 0.17 (0.05) 0.5 (0.04) 0.40 (0.05) 14 (3) 10
5 " (CH2)5 0.18 (0.02) 0.45 (0.05) 0.40 (0.05) 0 8
6 " (CH2)6 0.52 (0.1) 0.40 (0.10) 0.35 (0.05) 75 (15) 4
7 " CH=CH 0.38 (0.05) 0.55 (0.03) 0.45 (0.04) 0 8
8 " CCl=CH 0.86 (0.23) 0.4 (0.05) 0.3 (0.05) 5 (2) 4
9 " (CH2)2CH=CH(CH2)2 0.46 (0.16) 0.5 (0.1) 0.45 (0.1) 78 (3) 3
10 " (CH2)S(CH2) 0.19 (0.04) 0.55 (0.1) 0.4 (0.1) 32 (4) 6

11 " 0.15 (0.03) 0.5 (0.05) 0.45 (0.05) 2 (2) 4

12 " 0.19 (0.04) 0.5 (0.05) 0.35 (0.05) 7 (5) 5

13 " 0.21 (0.04) 0.55 (0.10) 0.5 (0.10) 13 (9) 5

14 " CH2 0.23 0.55 0.45 55 2

15 3,4-propionyl-oxy B (CH2)2 0.11 0.5 0.4 50 2

16 " (CH2)3 0.30 (0.03) 0.55 (0.1) 0.4 (0.02) 0 5
17 " (CH2)6 1.04 0.4 0.4 38 2
18 " CH=CH 0.48 (0.05) 0.5 (0.02) 0.35 (0.05) 0 6
19 " CCl=CH 0.73 (0.15) 0.4 (90.05) 0.3 (0.05) 6 (3) 3

20 " 0.43 (0.08) 0.45 (0.05) 0.3 (0.05) 2 (1) 4

21 " 0.33 (0.02) 0.6 (0.05) 0.4 (0.03) 0 3

22 3-methoxy-4-acetoxy B (CH2) 0.86 (0.18) 0.55 (0.1) 0.7 (0.1) 0 5

23 " (CH2)3 0.43 0.7 0.6 25 2
24 " (CH2)5 0.28 0.45 0.45 18 (6) 5
25 " (CH2)6 0.68 0.4 0.35 40 2
26 " CH=CH 0.52 (0.06) 0.6 (0.1) 0.6 (0.1) 6 (4) 6
27 " CCl=CH 0.95 (0.30) 0.6 (0.1) 0.6 (0.02) 13 (5) 3
28 " (CH2)S(CH2) 0.67 (0.15) 0.45 (0.01) 0.46 (0.06) 51 (12) 5

29 " 0.34 0.8 0.7 55 2

30 " CH2 0.35 0.5 0.5 75 2

31 3-chloro-4-acetoxy B (CH2)3 0.29 (0.04) 0.75 (0.13) 0.6 (0.1) 40 (20) 3

32 " (CH2)4 0.45 (0.19) 0.6 (0.01) 0.47 (0.05) 35 (23) 3
33 3-nitro-4-acetoxy B (CH2)3 0.4 1.3 4.0 1

34 2,5-dimethoxy B CH2 0.078 3.0 4.2 50 2

35 2-methoxy, 5-nitro B CH2 0.08 3.0 4.0 1

36 3,5-dimethoxy-4-acetoxy B (CH2)3 0.27 0.8 0.9 30 2

37 " (CH2)5 0.4 (0.1) 0.5 (0.1) 0.85 (0.15) 7 (3) 4
38 " (CH2)6 0.58 0.4 0.5 70 1
39 " (CH2)S(CH2 0.61(0.13) 0.5 (0) 0.63 (0.15) 51 (22) 4
 L. Gyermek et al. / Life Sciences 79 (2006) 559569 563

Table 1 (continued)
40 " CH=CH 0.55 1.0 1.1 0 2

41 " 0.28 (0.005) 0.8 (0.15) 1.3 (0.15) 29 (31) 3

42 " CH2 0.25 0.7 0.65 55 2

43 3,4-dimethoxy-4-propionyl-oxy B (CH2)3 0.52 0.8 0.95 10 2

44 3,4,5-triacetoxy B (CH2)3 0.65 (0.1) 0.6 (0.05) 0.4 (0.07) 0 4
45 " CH=CH 2.8 (0.14) 0.8 (0.15) 0.5 (0.2) 0 3
General structure (see also Fig. 2): R1N GROCOR2OCOGR NR1. 2Br where GR = granatane, No. = Identifying number of compounds, R1N = terminal N
quaternized with di- or trisubstituted benzyl (B) groups. R2 = connecting chain; 2Br = two bromine atoms. CVB = cardiac vagus block at 80% NMB. N = number of
samples.

few, selected agents, which have shown promise in one or two
intermediary species (e.g. the rabbit and juvenile pig).
Results
NMB profile of the agents (Tables 17).
Table 1 lists the results obtained in the rat. Table 2
summarizes the data obtained in the rabbit. Table 3 contains
the results in the pig. Table 4 lists the data in the cat. Table 5
represents results obtained in the dog. Table 6 lists data in the
monkey. Table 7 outlines the results obtained with compound
#7 using the infusion administration mode in the pig model. The
neuromuscular blocking profile has been characterized by NMB
potency (NMB ED50 values), onset of action, recovery index
(2575% twitch response recovery) and cardiovascular side
effects. Peak heart rate changes (percent) were listed with dogs
and monkeys. Table 8 shows the comparison of the results,
obtained within the investigated parameters and on different
species, between granatanol derivatives and the corresponding
tropanol derivatives (Gyermek et al., 2002; Gyermek, 2002).
Table 9 presents the data obtained with three bisquaternary
derivatives (#53, 54 and 55) of the glutaryl ester of pseudo
granatanol (e.g. granatane (3-beta)-ol).
NMB profile
Rat data
The highest potency (e.g. < 0. 2M/kg) has been observed
with agents #2, 4, 5, 10, 11, 12, 15, 34, and 35. These consisted
mostly of the 3,4-diacetoxy benzyl quaternaries. Outstanding
potency was observed, however, with agents #34 and 35, where
the quaternizing groups were 2,5-dimethoxy or 2-methoxy, 5-
nitrobenzyl and the linker contained 36 endomethylene
tetrahydro-phtaloyl diester. (Both these quaternizing moieties

Table 2
NMB profile of quaternary granatanol diesters in the rabbit
No. N1R R2 ED50 M/kg Onset min RI min N
2 3,4-diacetoxy B (CH2)2 0.19 1.0 1.4 1
3 " (CH2)3 0.19 (0.028) 0.65 (0.1) 0.55 (0.05) 4
4 " (CH2)4 0.2 (0.03 0.68 (0.2 ) 0.5 (0.1) 3
5 " (CH2)5 0.31 (0.03) 0.72 (0.2) 0.6 (0.1) 4

11 " 0.14 0.6 0.6 2

14 " CH2 0.41 (0.06) 0.65 (0.05) 0.55 (0.04) 5

16 3,4-dipropionyl-oxy B (CH2)3 0.32 (0.025) 0.6 (0.05) 0.45 (0.03) 4

23 3-methoxy-4-acetoxy B (CH2)3 0.34 0.6 0.6 2

30 " CH2 0.37 (0.06) 0.5 (0.05) 0.5 (0.02) 5

36 3,5 dimethoxy-4-acetoxy B (CH2)3 0.34 (0.04) 0.5 (0.05) 1.2 (0.1) 3

34 2,5-dimethoxy B CH2 0.011 4 9 1

35 2-methoxy-5-nitro B " 0.013 6 (1) >10 3

46 3,4,5-trimethoxy B " 0.095 5 12 1
42 3,5-dimethoxy-4-acetoxy B " 0.46 (0.04) 0.5 (0.05) 0.6 (0.05) 5
43 3,5-dimethoxy-4-propionyl-oxy B (CH2)3 0.81 0.45 0.8 2
General structure (see also Fig. 2): R1N GROCOR2OCOGRN R1. 2Br where GR = granatane; No. = identifying number of compounds; RN = terminal N
quaternized with di- or trisubstituted benzyl (B) groups. R2 = connecting chain; 2Br = two bromine atoms. N = number of samples.
564 L. Gyermek et al. / Life Sciences 79 (2006) 559569

Table 3
NMB profile of quaternary granatanol diesters in the pig
No. R1N R2 ED50 M/kg Onset min RI min CVB % N
1 3,4-diacetoxy B (CH2) 0.88 0.5 0.5 22 2
2 " (CH2)2 0.73 (0.04) 0.65 (0.05) 0.7 (0.1) 76 (5) 5
3 " (CH2)3 0.29 (0.05) 0.8 (0.04) 0.9 (0.1) 40 (8) 6
4 " (CH2)4 0.17 (0.016) 0.7 (0.05) 0.9 (0.05) 51 (15) 3
5 " (CH2)5 0.23 (0.04) 0.7 (0.1) 1.0 (0.1) 48 (11) 3
6 " (CH2)6 0.30 (0.03) 0.6 (0.1) 0.8 (0.2) 80 (15) 3
7 " CH=CH 0.30 (0.04) 0.7 (0.1) 0.7 (0.03) 24 (6) 4
8 " CCl=CH 0.96 0.4 0.3 15 2
9 " (CH2)2CH=CH(CH2)2 0.12 1.0 1.5 100 1
10 " (CH2)S(CH2) 0.2 0.75 1.2 66 1

11 " 0.15 (0.02) 0.8 (0.05) 1.0 (0.05) 15 (5) 4

12 " 0.41 (0.034) 0.9 (0.02) 0.9 (0.1) 50 (8) 3

13 " 0.19 (0.07) 0.6 (0.1) 0.6 (0.1) 50 (8) 3

47 " 0.39 1.1 1.0 70 1

CH2 CH2
15 3,4-dipropionyl-oxy B CH2)2 0.35 0.4 0.4 0 1
48 CH2)4 0.31 (0.03) 0.5 (0.02) 0.4 (0.01) 42 (3) 3
18 " CH=CH 0.47 (0.06) 0.7 (0.1) 0.45 (0.03) 16 (3) 6
19 " CCl=CH 0.61 0.5 0.4 18 2

49 " 0.22 (0.02) 0.5 (0.1) 0.45 (0.03) 18 (4) 4

20 " 0.74 1.0 0.8 46 2

21 " 0.41 0.52 0.45 62 2

22 3-methoxy-4-acetoxy B (CH2) 0.76 (0.016) 0.5 (0.05) 0.65 (0.1) 50 (8) 3

26 " CH=CH 0.43 (0.01) 0.7 (0.1) 1.1 (0.2) 38 (4) 4
27 " CCl=CH 0.9 0.4 0.3 15 1
28 " (CH2)S(CH2) 0.45 0.4 0.9 90 1

50 " 0.23 0.7 1.2 65 1

29 " 0.25 0.5 1.0 57 1

51 3,4,5-triacetoxy B CH2)3 0.59 1.0 0.9 80 1

52 " CH=CH 3.8 0.5 0.5 10 1
40 3,5-dimethoxy-4-acetoxy B CH=CH 0.6 (0.06) 0.7 (0.1) 0.7 (0.1) 14 (1) 5
General structure (see also Fig. 2): R1N GROCOR2OCOGR NR1. 2Br where GR = granatane; No. = identifying number of compounds; R1 = terminal N
quaternized with di- or trisubstituted benzyl (B) groups. R2 = connecting chain; 2Br = two bromine atoms. CVB = cardiac vagus block at 80% NMB. N = number of
samples.

and this dicarboxylic acid linker proved to be favorable with
respect to potency within several previously reported bis
tropinium NMB agents (Gyermek et al., 2002; Gyermek,
2002)). Di- and trisubstituted benzyl quaternary derivatives
with mixed methoxy and acyloxy substitutions usually yielded
compounds with lower potency. Finally compound #45, a
fumaryl ester with 3,4,5-triacetoxy benzyl quaternizing groups
has proven to be of very low potency.
Ultrashort onset and recovery characteristics (e.g. recovery
indices less than 1 min) were observed with the majority of these
granatanol diesters. Characteristically the high potency com-
pounds #34 and 35, lacking the acyloxy component, were both
 L. Gyermek et al. / Life Sciences 79 (2006) 559569 565

Table 4
NMB profile of quaternary granatanol diesters in the cat
No. N1R R2 ED50 M/kg Onset min RI min CVB % N
2 3,4-diacetoxy B (CH2)2 0.11 (0.03) 0.75 (0.1) 0.45 (0.1) 20 3
3 " (CH2)3 0.85 0.7 0.4 62 1
4 " (CH2)4 0.48 0.7 0.5 70 1
6 " (CH2)6 0.81 80 1
7 " CH=CH 0.28 (0.03 0.8 (0.07) 0.45 (0.01) 17 (5) 4
8 " CHCl=CH 0.6 1.0 0.9 50 1

52 " 0.41 0.7 0.45 50 1

11 " 0.23 (0.08) 0.6 (0.05) 0.45 (0.05) 12 (6) a 4

18 3,4-dipropionyl-oxy B CH=CH 0.47 (0.06) 0.6 (0.04) 0.36 (0.01) 18 (5) 4

26 3-methoxy-4-acetoxy B CH=CH 0.36 (0.06) 0.55 (0.02) 0.5 (0.02) 17 (5) 4
51 3,4,5-triacetoxy B CH=CH 0.20 (0.02) 1.0 (0.1) 1.2 (0.2) 11 (5) 3
General structure (see also Fig. 2): R1N GROCOR2OCOGRN R1. 2Br where GR = granatane; No. = identifying number of compound; R1N = terminal N
quaternized with di- or trisubstituted benzyl (B) groups. R2 = connecting chain; 2Br = two bromine atoms. CVB = cardiac vagus block at 80% NMB. N = number of
samples.
a
At 6 ED 80 NMB dose CVB is 36%.

slow and relatively long acting. From the point of view of cardiac
vagal blocking side effect several compounds showed only
very slight effect. For this the malonyl, glutaryl, pimelyl,
fumaryl, and cyclobutane 1,2 dicarboxylyl linkers seem to be
responsible. In contrtast, longer dicarboxylic acid esters (e.g.
compounds #6, 9, 38,) and the 36 endomethylene tetrahy-
drophtalic esters showed higher cardiac vagal blocking potency.
(It is noteworthy that in assessing the cardiac vagal block the rat
model is useful only for preliminary screening, since false
negative responses may occur in this species, which often could
not be reproduced in other species. (e.g. the pig and cat). In
overall, the best NMB profile on this species belonged to agents
#3, 5, 7, 11, 12, 16 and 21. These were all either 3,4-diacetoxy-
or 3,4-dipropionyl-oxy benzyl quaternary derivatives.
Rabbit data
Only the potency, onset and recovery characteristics were
assessed on this species because cardiovascular stability and
reproducibility of heart rate responses on vagus nerve
stimulation are inconsistent in this species. The best four
compounds were #3, 4, 11 and 16. Similar to the rat data,
highest potency was found with two methoxy and mixed
methoxy nitrobenzyl quaternaries (#34 and 35). Like in the rat,
both of these compounds were slow and long acting.
Pig data
Highest potency (e.g. ED50 s of <0.25 M/kg) belonged to
the following compounds: #4, 5, 9, 10, 11, 13, 49 and 50. The
fastest and shortest acting compounds were #8, 19 and 27 (with
chlorophtaloyl ester linkers). These, and several of the 3,4-
dipropionyl-oxy benzyl quaternaries (e.g. #15, 19, 21, 48 and
49) all showed shorter than 1 min onset and recovery indices.
Here the compounds with chlorofumaryl linkers excelled with
an onset of 2030s and recovery indices of 2530s. A
triacetoxy benzyl quaternary derivative (#45) seemed to be also
very short acting (N = 1) but its potency was the weakest among
all the compounds. The pig, being very sensitive to cardiac
vagal block, showed a moderate degree of vagal block even
with the best compounds. Less then 20% cardiac vagal block
occurred with the following compounds: #8, 11, 15, 18, 19, 27,
40, 45 and 49. Of these, three carried chlorofumaryl, and
another three fumaryl diester linkers. Two others (e.g. #11 and
49) were 1,2-cyclobutane dicarboxylyl derivatives and
contained the favorable acyloxy benzyl quaternary groups.

Table 5
NMB profile of bisquaternary granatanol diesters in the dog
No. R1N R2 ED50 M/kg Onset min RI min Heart rate change % N
2 3,4-diacetoxy B (CH2)2 0.17 0.65 0.45 +7 2
3 " (CH2)3 0.27 (0.02) 0.65 (0.05) 0.6 (0.04) +50 (8) 3
4 " (CH2)4 0.17 0.6 0.4 +8 1
5 " (CH2)5 0.3 1.2 0.5 +10 1
6 " (CH2)6 0.29 0.8 0.45 +60 2

11 " 0.18 (0.04) 0.75 (0.1) 0.7 (0.05) +27 (5) 4

General structure (see also Fig. 2): R1N GROCOR2OCOGRN R1. 2Br where GR = granatane; No. = identifying number of compounds; R1N = terminal N
quaternized with di- or trisubstituted benzyl (B) groups. R2 = connecting chain; 2Br = two bromine atoms. N = number of samples.
566 L. Gyermek et al. / Life Sciences 79 (2006) 559569

Table 6
NMB profile of quaternary granatanol diesters in the monkey
No. R1N R2 ED50 M/kg Onset min RI min Heart rate change % N
2 3,4-diacetoxy B (CH2)2 0.03 (0.004) 0.7 (0.02) 0.4 (0.01) 0 5
3 " (CH2)3 0.095 (0.009) 0.7 (0.1) 0.4 (0.02) 0 6
4 " (CH2)4 0.1 (0.006) 0.8 (0.1) 0.4 (0.02) 0 3
5 " (CH2)5 0.17 1.2 0.5 0 1
6 " (CH2)6 0.22 0.9 0.4 0 1
7 CH=CH 0.32 (0.37) 0.76 (0.05) 0.57 (0.02) 0 5

11 " 0.10 (0.025) 0.7 (0.02) 0.4 (0.02) 0 4

18 3,4-dipropionyl-oxy B CH=CH 0.62 (0.06) 0.54 (0.02) 0.36 (0.02) 0 5

26 3-methoxy-4-acetoxy B CH=CH 0.54 (0.028) 0.68 (0.05) 0.63 (0.02) 0 5
General structure (see also Fig. 2): R1N GROCOR2OCOGRN R1. 2Br where GR = granatane; No. = identifying number of compounds; R1N = terminal N
quaternized with di- or trisubstituted benzyl (B) groups. R2 = connecting chain; 2Br = two bromine atoms. N = number of samples.

The best overall NMB profile belonged to compounds #11, 18,
27, 40 and 49. All of them were fumaryl, chlorofumaryl or 1,2-
cyclobutane dicarboxylyl diesters, quaternized with diacyloxy
benzyl or mixed methoxyacyloxybenzyl groups.
Cat data
The three most potent three agents were #2, 11 and 51. The
former two were also highly potent in several other species.
Surprisingly, however, #51 showed very low potency in the rat
and pig (although in the latter species only one animal was
tested). This can be explained by the observation that all the four
fumaryl esters were relatively more potent in the cat that in other
species, including the monkey (see Table 6). The onset of action
is somewhat slower in this species, and ranged between 0.55
(with agent #26) and 1.0min (with the two fumaryl diesters #8
and 51). Recovery indices were very short with all agents (e.g.
between 0.36 and 0.9 min) with the exception of #51 where it
was 1.2 min. Least cardiac vagal block occurred with agents #7,
11, 18, 26 and 51, which reflected the trend observed in the rat
and pig. Highest degree of cardiac vagal block has been found
with agents #4 and 6. The overall best NMB profile belonged to
compounds #2, 7, 11 and 25.
Dog data
The most potent agents were #2, 4 and 11, corresponding
with data obtained in other species. All six agents studied
exhibited rapid (e.g. < 1.2 min) onset and ultrashort (e.g.
<0.7 min) recovery indices. Hovewer, most agents in the
NMB dose level produced increased heart rate, ranging between
7% (agent #2) and 60% (agent #6). Even agent #11, with the
best overall NMB profile in several other species, produced in
the dog significant (e.g. 27%) tachycardia.
Monkey data
Highest potency was observed with agents #2, 3, 4 and 11,
which parallels the data obtained in other species. All agents
displayed rapid onset and particularly ultrashort recovery
indices (e.g. 0.360.63 min). In contrast to the dog no
significant heart rate changes were observed in the NMB dose
range which was lower than in the dog and may explain why

Table 7
Comparison of the NMB profile of identical quaternary derivatives of granatanol (G) and tropanol (T) diesters in different animal species (each number represents the
average value of the entire group)
Diester Species Number of NMB ED50 Diff. (P) Onset min Diff. (P) RI min (SEM) Diff. (P) CVB% Diff. (P)
compounds (SEM) (SEM) (SEM)
G Rat 16 0.294 (0.028) 0.596 (0.033) 0.567 (0.062) 25.5 (6.0)
0.124 (0.07) 0.03 (NS) 0.018 (NS) 2.0 (NS)
T " 16 0.418 (0.061) 0.63 (0.038) 0.582 (0.031) 23.5 (5.5)
G Pig 7 0.44 (0.11) 0.65 (0.057) 0.64 (0.096) 29.1 (8.0)
0.01 (NS) 0.19 (0.02) 0.25 (0.017) 5.1 (NS)
T " 7 0.43 (0.1) 0.83 (0.06) 0.89 (0.06) 34.2 (8.5)
G Cat 2 0.17 0.67 0.46
0.015 (NS) 0.3 (?) 0.2 (NS)
T " 2 0.155 0.97 0.48
G Dog 2 0.21 (0.03) 0.7 (0.07) 0.65 (0.045)
0.06 0.2 0.02
T " 2 0.15 (0.02) 0.9 (0.07) 0.63 (0.05)
G Monkey 3 0.098 (0.016) 0.73 (0.03) 0.4 (0.001)
0.01 (NS) 0.14 0.1
T " 3 0.087 (0.067) 0.87 (0.03) 0.5 (0.1)
 L. Gyermek et al. / Life Sciences 79 (2006) 559569 567

Table 8
Cumulative and cardiovascular side effects profile of compound #7 [N,N (3,4-diacetoxy benzyl) bis granatanium 3 -ol fumarate] in comparison with its tropane 3 -
ol analog and rocuronium in pig i.v. perfusion experiments
Compound Infusion time (min) Number
of tests
0 10 20 30 40 50 60
#7
Systolic BP (mm Hg) 97 (4) 100 (5) 102 (6) 101 (6) 104 (6) 103 (5) 105 (60) 5
HR 125 (8) 126 (6) 130 (4) 128 (6) 128 (5) 127 (4) 127 (4)
VB 0 34 (8) 35 (9) 35 (8) 35 (8) 36 (8) 36 (9)
Single dose NMB ED80 = 0.93 (0.09) M/kg
Average infusion rate = 0.33 (0.05) M/kg/min
RI (b) = 0.7 (0.1) min, RI (a) = 3.2 (0.4) min, RI (a) / RI (b) = 4.57

Tropane 3 -ol analog of #7

Systolic BP (mm Hg) 101 (2) 98 (2) 99 (2) 99 (2) 101 (2) 102 (2) 103 (2) 4
HR 144 (7) 139 (5) 136 (4) 135 (3) 135 (3) 136 (3) 136 (3)
VB 0 37 (6) 39 (6) 40 (6) 42 (6) 42 (5) 42 (6)
Single dose NMB ED80 = 0.44 (0.07) M/kg
Average infusion rate = 0.15 (0.02) M/kg/min
RI (b) = 0.9 (0.1) min, RI (a) = 1.3 (0.1) min, RI (a) / RI (b) = 1.45

Rocuronium
Systolic BP (mm Hg) 91 (6) 91 (6) 91 (6) 92 (5) 92 (6) 92 (6) 94 (5) 6
HR 138 (4) 140 (3) 141 (4) 141 (3) 142 (4) 143 (3) 143 (3)
VB 0 29 (6) 33 (5) 34 (6) 35 (2) 37 (8) 37 (8)
Single dose NMB ED80 = 0.83 (0.06) M/kg
Average infusion rate = 0.13 ( 0.026) M/kg/min
RI (b) = 2.0 (0.1) min, RI (a) 5.8 (0.3) min, RI (a) / RI (b) = 2.9
Values in parentheses = SEM; BP = arterial blood pressure; HR = heart rate; VB = cardiac vagal block in %; RI (b) = recovery index of the NMB following a bolus dose
of ED80, before infusion. RI (a) = recovery index of the NMB after 60 min infusion.

was the CV stability more favorable in the monkey. The other
reason may be the difference in the anesthetic regimen. While
pentobarbital + chloralose anesthesia was used in the dog,
ketamine and halothane were utilized in the monkey.
Discussion
Structural comparison between bisquaternary granatanol and
tropanol diesters
Molecular volume
Expanding the asymmetric 8-azabicyclo (3.2.1) octane
(tropan) ring system to the symmetric 9-azabicyclo (3.3.1)
nonane ring system brings about only minor change in molecular
volume, but makes it more flexible (Fig. 1). This allows changes
in the overall conformation of the ring system and also in the
position of the ring substituents, particularly on the quaternary
nitrogen atom, as referred to earlier (Gyermek et al., 2002).
For instance, with certain tropane quaternary derivatives the
distribution of N-benzyl or p-phenylbenzyl groups favors the N-
equatorial orientation by a ratio of about 68 : 1 vs. the axial N-
methyl group (Nador et al., 1978). In the case of the granatane
ring system, for comparison, the distribution of the hetero-
quaternary pairs are close to 1 : 1. These N-stereoisomeric
differences are interesting to consider with respect to the
presently reported series because the similarity in the pharma-
cologic profile of the quaternary pairs of granatanol vs. tropanol
diesters seems to indicate that in NMB properties the role of
substitutional asymmetry about the quaternary ammonium
group is not significant. This is further underlined by the rather
similar NMB profile of the (prevailing) N-aralkyl anti-anti vs.
the N-aralkyl syn-anti (minority) species of racemic TAAC3
(ORG 95415) (Gyermek et al., 2002). These observations are in
contrast to those obtained with different monoquaternary
atropine N-enantiomers, where marked quantitative differences
were noted either in the antimuscarinic (Engelhardt, 1979;

Table 9
NMB profile of bisquaternary pseudo granatanol diesters in the rat
No. R1N R2 ED50 M/kg Onset min RI min CVB % N
53 3,4-diacetoxy B (CH2)3 0.9 (0.02) 0.5 (0.03) 0.5 (0.06) 0 3
54 3,4-dipropionyl-oxy B 1.2 (0.26) 0.5 (0.05 ) 0.38 (0.05) 0 3
55 3-methoxy-4-acetoxy B 0.8 (0.12) 0.52 (0.03) 0.6 (0.15) 5 3
General structure (see also Fig. 2): R1N GROCOR2OCOGR R1N. 2Br where GR = granatane 3-exo-ol; No. = identifying number of compounds; R1N =
terminal N quaternized with di- or trisubstituted benzyl (B) groups. R2 = connecting chain; 2Br = two bromine atoms. CVB = cardiac vagus block at 80% NMB; N =
number of samples.
568 L. Gyermek et al. / Life Sciences 79 (2006) 559569

Nador et al., 1978) or ganglion blocking (Nador et al., 1978) Note: At an advanced stage of our preclinical animal studies
potencies. with bisquaternary tropanol and granatanol diesters, in 2002
we were informed by our sponsor, Organon Laboratories,
Electric-charge characteristics that the 3,4-diacetoxy benzyl quaternizing group of the
The size of the positive charge of the quaternary N atoms glutaryl bis tropine: TAAC3/ORG-25415, responsible for its
forming the cationic head and the partial del Re charges ultrashort action, in larger doses results in the formation of a
around the onium head as calculated with the MMPro (Chem nephrotoxic metabolite: p-quinone methene. The 14
SW, Fairfield, CA) program were identical in the tropanol and elimination reaction, leading to such metabolite generally
the granatanol diester series. occurs with 4-acyloxy substituted aryl quaternary groups.
Thus all mono- or polysubstituted acyloxy benzyl quatern-
Lipophilicity ary derivatives, where an acyloxy group resides in the para
Calculation of lipophilicity by determining log P values of position, are liable to produce the same or similar toxic
quaternary ammonium compounds is considered unreliable. quinone methenes. This applies to the similar quaternary
Thus we resorted to obtain a semi-quantitative indicator of derivatives of granatanol diesters, representing a major
lipophilicity: the relative proportion of the hydrophilic vs. hindering factor for their further pharmaceutical develop-
lipophilic surface areas of a few representative bis granatanol ment. For the above reasons pharmacokinetic experiments
and tropanol diesters. It was found that the contribution of the
additional methylene group in the granatane ring, to the relative Table 10
size of the lipophilic surface area is negligible. Neuromuscular blocking profile
Onset min RI min N
Inter-onium distance
Rat
The above mentioned factors may be responsible for altering
#7 0.55 0.45 8
NN inter-onium distance, another structural element consid- #11 0.5 0.45 4
ered to be important, at least in the potency and/or side Rocuronium 0.95 1.3 7
effects profile of NMB agents. In the case of agents with a Atracurium 1.5 2.9 5
flexible linker component, as mentioned earlier, even the in Mivacurium 1.8 3.9 8
silico computational model calculations remain unreliable.
Pig
Further, the functional inter-onium distance, e.g. the confor- #7 0.7 0.8 4
mation prevailing at the instant of docking to the receptor, to #11 0.8 1.0 4
the knowledge of the authors, cannot be determined with the Rocuronium 0.95 1.1 7
presently available methods. Mivacurium 1.5 3.5 5
Suxamethonium 1.0 3.3 4

Functional aspects Cat

#7 0.8 0.45 4
To examine SAR relationships the relatively large data base #11 0.6 0.45 4
consisting of 3,4-diacetoxy benzyl quaternary derivatives, and Rocuronium 1.2 2.0 3
Mivacurium 3.5 5.0 6
based on rat, pig, cat and monkey data, has been used. The
compounded NMB profile of the granatanol and tropine series Monkey
has been found to be largely identical as shown in Table 8. The #7 0.76 0.57 5
onset of action was seemingly faster in the case of granatanol #11 0.7 0.4 5
derivatives in all species, yet the difference reached significant Rocuronium 1.9 2.4 6
Mivacurium 3.2 3.5 4
level (P = 0.02) only in the pig. The recovery indices of
granatanol derivatives were also slightly shorter in all species, Comparison of the onset and recovery indices of two granatanol diesters with
except the dog, but the difference reached significant level those of clinically used short acting NMB agents on experimental animals.
Truncated data on clinically introduced agents taken from Gyermek (2002). The
(P = 0.018) again only in the pig. In contrast to the single dose table clearly shows that the two selected granatanol diester derivatives exhibited
administration mode, however, with infusion administration markedly shorter onset of action and showed also faster recovery indices than
the #7 granatanol derivative became more cumulative than either one of the clinically used, reference compounds. These differences were
the corresponding tropanol derivative as indicated in Table 7. the least pronounced on the pig and most remarkable in the monkey. It should be
The most likely explanation for this difference would be the noted that while compounds #7 and 11 were among the most explored
compounds, some of the others, e.g. #18, exhibited even faster and shorter NMB
possibility of the formation of a more active metabolite in the action profile. These data altogether offer convincing evidence that these agents,
case of the garanatanol, than in the tropanol diester series. on experimental animals, are the fastest and shortest acting non-depolarizing
Since the main metabolite with this type of compounds is the NMB agents discovered so far. On the pig they even surpassed the depolarizing-
tertiary diester, it might be possible that the tertiary granatanol type succinylcholine in rapidity of action.
diester is either a more potent NMB agent than the Previously in experimental animals we could demonstrate that the onset and
recovery characteristics of structurally similar bisquaternary tropinyl diesters-
corresponding tertiary tropanol diester, or its speed of further type agents, e.g. TAAC3 (Gyermek et al., 2002) were also faster and shorter than
elimination, e.g. the hydrolysis of the linker, or excretion reported in the literature for the new experimental neuromuscular relaxant
through the kidney is slower. GW280430A (Boros et al., 1999).
 L. Gyermek et al. / Life Sciences 79 (2006) 559569
were not done with the bis granatanol derivatives beyond
the few constant speed-infusion experiments carried out in
pigs and shown in Table 7.
Conclusion
We have found that the granatanol ring system is a suitable
substitute for the tropanol ring in the class of ultrashort acting
neuromuscular relaxants. This finding is in contrast to the
results obtained with few bisquaternary 7-azabicyclo (2.2.1)
heptane 2-ol diesters previously, indicating that these are
usually less potent neuromuscular blocking agents and show
higher degree of CV side effects (Gyermek et al., 2002). We
have demonstrated that in NMB function the symmetrical
granatanol ring system can replace the asymmetric tropane ring
system to which is nearly isosteric and also similar in atomic
charge distribution and degree of lipophilicity. While the molar
NMB potency in case of the granatanol diesters is about
identical to that of the corresponding tropanol diesters, the onset
and recovery profile of several of the granatanol derivatives
seems to be even faster and shorter. We thus can conclude that in
experimental animals these new granatanol diesters are the
fastest and shortest acting non-depolarizing NMB agents
known. Table 10 illustrates this point by listing the onset and
recovery characteristics of two selected granatanol diester
derivatives in comparison with animal data obtained with
clinically introduced short acting non-depolarizing NMB
agents. The above mentioned nephrotoxic quinone methene
metabolite formation precludes their further pharmaceutical
development, unless functionally similar new quaternary
groups, without such propensity, can be introduced into the
basic skeleton of these agents. In terms of new drug
development, in general, the granatane ring system has not
been extensively explored in spite of its facile synthesis. For
example granatan-3 one, the precursor to the corresponding 3-
alpha- or beta-ols can be obtained through a one step
condensation reaction similar to that of tropinone (Robinson,
1917).
The only references to granatanol derivatives in the field of
anticholinergic pharmacology are that of a) Ribbentrop and
Schaumann (1965) who described different, mostly di-
aromatic acid esters and ethers of granatanol as centrally, as
well as peripherally acting antimuscarinic agent, b) Gyorgy et
al. (1970) who reported on the potent central anticholinergic
activity of 2-dimethylamino-9-benzylyloxy bicyclo (3,3,1)
nonane, and c) Hunt and Robinson (1970) describing the
synthesis of tropic acid esters of granatane-3-alpha- and 3-
beta-ols. Other few granatane derivatives were reported as
having adrenergic neuron blocking (Doda et al., 1967) and
anticonvulsant (Ballesteros et al., 1981) actions. According to
our knowledge neuromuscular blocking bis granatanol deri-
vatives were not reported in the literature, therefore our study
represents the first systematic study about this group as NMB
agents.
569
Acknowledgements
The authors wish to express their thanks to Organon NV
(Oss, The Netherlands) for the support of the study during
19982002, and to Dr. Erik Sprengers (Organon NV) and Dr.
Niall Hamilton (Organon Laboratories) for their help and
interest in the study.
References
Ballesteros, P., Avendano, M.C., Canadas, S., Gonzalez, M.P., 1981. Effect of
diphenylhydantoin, granatane-3-spiro-5 hydantoin and leptazol on mouse
brain glutamate dehydrogenase, urea and ammonia levels in brain. Revista
Espanola de Fisiologia 37, 205210.
Boros, E.E., Bigham, E.C., Boswell, G.E., Mook Jr., R.A., Patel, S.S.,
Savarese, J.J., Ray, J.A., Thompson, J.B., Hashim, M.A., Wisowaty, J.C.,
Feldman, P.L., Samano, V., 1999. Bis- and mixed-tetrahydroisoquinolinium
chlorofumarates: new ultra-short-acting nondepolarizing neuromuscular
blockers. Journal of Medicinal Chemistry 42, 206209.
Doda, M., Molnar, J., Gyorgy, L., Nador, K., 1967. Pharmacological analysis of
adrenergic neuron blocking granatane derivatives. Acta Physiologica
Academiae Scientiarum Hungaricae 32, 257266.
Engelhardt, A., 1979. Pharmacology and toxicity of Atrovent. Scandinavian
Journal of Respiratory Diseases 193, 110115.
Gyermek, L., 2002. Structureaction relationships among derivatives of
dicarboxylic acid tropine esters. Pharmacology And Therapeutics 96, 121.
Gyermek, L., 2005. Development of ultrashort acting muscle relaxant agents.
History, research strategies and challenges. Medicinal Research Reviews 25,
610654.
Gyermek, L., Lee, C., Nguyen, N., Cho, Y.M., 2002. Ultrashort acting
neuromuscular blocking agents: 1. Di- and tri-substituted benzyl quaternary
ammonium derivatives of (bis-tropane alpha-ol) diesters. Medicinal
Chemistry Research 111, 316336.
Gyorgy, L., Pfeiffer, A.K., Doda, M., Molnar, J., Nador, K., 1970. A compound
with a strong and specific central anticholinergic activity, 2-dimethylamino-
9-benzylyloxy bicyclo (3,3,1) nonane. Acta Physiologica Academiae
Scientiarum Hungaricae 37, 313318.
Hunt, R.J., Robinson, J.B., 1970. Preparation and characterization of the tropic
acid esters of tropane 3-ol, granatan 3 alpha-ol and granatan 3 beta-ol.
Journal of Pharmacy and Pharmacology 22, 29S (Suppl).
Lee, C., 2003. Conformation, action, and mechanism of action of neuromuscular
blocking muscle relaxants. Pharmacology and Therapeutics 98, 143169.
Lee, C., de Silva, A.J.C., Katz, R.L., 1978. Multipeaking of the compound
electromyographic response to single nerve stimulus: its occurrence and
prevention in an experimental model. British Journal of Anaesthesia 50,
10691073.
Nador, K., Scheiber, P., Karpati, E., et al., 1978. Synthese und Pharmakologie
von N-stereoisomeren quataerneren Tropan Verbindungen. Arzneimittel-
Forschung 21, 13241327.
Ribbentrop, A., Schaumann, W., 1965. Zentrale anticholinergische Wirksamkeit
und Spezifitt einiger derivate des Granatans. Arzneimittel-Forschung 15,
857862.
Robinson, R., 1917. Synthesis of tropinone. Journal of the Chemical Society
(London) 3, 762.
Tuba, Z., Maho, S., Vizi, E.S., 2002. Synthesis and structureactivity
relationship of neuromuscular blocking agents. Current Medicinal Chem-
istry 9, 15071536.