Development of novel 4-aminopyridine derivatives as potential treatments for neurological injury and disease

Article abstract of DOI:10.1016/j.ejmech.2005.04.017

The amine position of the K⁺ channel blocker 4-aminopyridine was functionalized to form amide, carbamate and urea derivatives in an attempt to identify novel compounds which restore conduction in injured spinal cord. Eight derivatives were tested in vitro, using a double sucrose gap chamber, for the ability to restore conduction in isolated, injured guinea pig spinal cord. The methyl, ethyl and t-butyl carbamates of 4-aminopyridine induced an increase in the post injury compound action potential. The methyl and ethyl carbamates were further tested in an in vivo model of spinal cord injury. These results represent the first time that 4-aminopyridine has been derivatized without losing its ability to restore function in injured spinal cord tissue.

Full text of DOI:10.1016/j.ejmech.2005.04.017

European Journal of Medicinal Chemistry 40 (2005) 908–917
www.elsevier.com/locate/ejmech
Original Article
Development of novel 4-aminopyridine derivatives as potential treatments
for neurological injury and disease
a,
b
b
b
Daniel T. Smith *, Riyi Shi , Richard B. Borgens , Jennifer M. McBride ,
c
a
Kevin Jackson , Stephen R. Byrn
a
Department of Industrial and Physical Pharmacy, Purdue University, 575 Stadium Mall Drive, West Lafayette, IN 47907, USA
Center for Paralysis Research, School of Veterinary Medicine, School of Engineering, Purdue University, West Lafayette, IN 47907, USA
Section of Neurosurgery, Wishard Memorial Hospital, Indiana University Medical Center, 1001 West 10th Street, Indianapolis, IN 46202, USA
Received 19 November 2004; received in revised form 23 March 2005; accepted 4 April 2005
Available online 01 August 2005
b
c
Abstract
+
The amine position of the K channel blocker 4-aminopyridine was functionalized to form amide, carbamate and urea derivatives in an
attempt to identify novel compounds which restore conduction in injured spinal cord. Eight derivatives were tested in vitro, using a double
sucrose gap chamber, for the ability to restore conduction in isolated, injured guinea pig spinal cord. The methyl, ethyl and t-butyl carbamates
of 4-aminopyridine induced an increase in the post injury compound action potential. The methyl and ethyl carbamates were further tested in
an in vivo model of spinal cord injury. These results represent the first time that 4-aminopyridine has been derivatized without losing its ability
to restore function in injured spinal cord tissue.
©
2005 Elsevier SAS. All rights reserved.
Keywords: 4-Aminopyridine; 4-Aminopyridine derivatives; Spinal cord injury; Potassium channel blocker; Demyelination; Double sucrose gap chamber
1
. Introduction
myelin damage. Unfortunately, there are relatively few com-
pounds that are known to block voltage-gated potassium chan-
nels. The most commonly cited examples are the aminopy-
ridines (Fig. 1), tetraethylammonium (TEA) salts and a family
of polypeptide toxins isolated from the venoms of scorpions,
snakes, bees and anemones [3–8]. Of these compounds,
The majority of clinical spinal cord injuries (SCI) do not
involve severing the spinal cord but instead result from the
compression or contusion of the cord, leaving variable
amounts of the spinal cord intact post injury [1]. In the sur-
viving spinal cord, the mechanical insult leads to the destruc-
tion of the nerve’s myelin sheath at the site of injury, expos-
4
-aminopyridine (4-AP) has shown the greatest potential as a
treatment of SCI.
+
+
ing paranodal K channels. The exposed K channels provide
a pathway for the exodus of K ions during neuronal activity,
+
Study of 4-AP as a potential treatment for SCI was initi-
ated in the late 1980s [9] and, more recently, it has moved
into human clinical trials [10–16]. Behavioral improvements
attributed to 4-AP treatment include: voluntary motor con-
trol, ambulation, continence, respiratory control along with
decreases in spasticity and idiopathic pain [10–15,17–21].
However, the positive improvements induced by 4-AP are
variable in their expression and may evoke serious side effects
such as mood swings along with a heightened state of anxi-
ety and seizures, both mild and severe [11,12,20,22,23]. It
has become clear that the maximal safe dose is insufficient to
produce the maximal beneficial functional recovery in many
cases [24].
which effectively quenches the nerve impulse as it reaches
the section of damaged spinal cord. The overall effect is essen-
tially the same as if the nerve were severed: no conduction
occurs through the site of injury [1,2].
Fast voltage-gated (A current) potassium channel block-
ers have been found to be useful tools in counteracting the
effects of demyelination in SCI. By blocking the potassium
ions’ route of exodus from the cell, it is possible to permit
nerve impulse conduction to effectively bridge the area of
*
Corresponding author. Tel.: +1 765 494 0513; fax: +1 765 494 6545.
E-mail address: dtsmith@pharmacy.purdue.edu (D.T. Smith).
0223-5234/$ - see front matter © 2005 Elsevier SAS. All rights reserved.
doi:10.1016/j.ejmech.2005.04.017

D.T. Smith et al. / European Journal of Medicinal Chemistry 40 (2005) 908–917
909
Fig. 1
. The aminopyridines. 2-Aminopyridine (2-AP, 1), 3-aminopyridine (3-AP, 2), 4-aminopyridine (4-AP, 3), 3,4-diaminopyridine (3,4-DAP, 4), and
4-aminoquinoline (4-AQ, 5).
This behavioral “cap”, in a addition to the fact that very
tant. In that way, the two features most pertinent to 4-AP’s
biological activity would be maintained.
few 4-AP derivatives have been tested as alternatives (all
unsuccessfully [25–30]), led us to investigate the chemical
derivatization of 4-AP as a means of generating novel com-
pounds that restore conduction in injured spinal cord, possi-
bly leading to an alternative therapy for treating conditions
that result from nerve demyelination.
Fig. 2 illustrates the range of compounds tested in this
study. 4-(Dimethylamino)pyridine (DMAP, 6) was chosen as
an extension of the previous work using 4-aminopyrinine
methiodide (4-AMPI) [26–28,30,32]. The amide, carbamate
and urea functional groups were seen as a means of increas-
ing the available hydrogen bonding groups while maintain-
ing at least some of the pyridine–pyridinium equilibrium.
Compounds 7–11 and 13 were synthesized from 4-AP via
standard synthetic methodology. Cyclic urea 12 was synthe-
sized from 3,4-diaminopyridine. DMAP (6) was purchased
2
. Chemistry
At physiological pH, the basic nature of 4-AP (pK = 9.4)
a
(Aldrich, Milwaukee, WI.) and used as received.
[
31] leads to an equilibrium between the neutral and proto-
nated forms that strongly favors the pyridinium ion (~99%).
It has been shown that the neutral from is responsible for
crossing the cell’s membrane while the protonated form
blocks the channel from the cytoplasmic side of the cell [3,26–
3
. Bioevaluation
Each of the derivatives was tested in vitro to determine its
2
4
8,32–37]. We feel that the hydrogen bonding ability of
-AP’s protonated form, which is strengthened by its ability
ability to restore conduction in isolated, injured guinea pig
spinal cord. This was done using a double sucrose gap isola-
tion and recording chamber [40,41]. Strips of spinal cord white
matter were placed in the chamber then allowed to stabilize
for 1 h in order to identify a characteristic response to elec-
trical stimulation. A stretch injury caused the loss of conduc-
tion, which improved slightly with time. Once a stable “recov-
ery” compound action potential (CAP) was achieved (the
baseline), the injured cord was exposed to a solution of the
test compound. The amount of restored conduction was deter-
mined by noting the percentage increase of CAP amplitude
over the baseline.
to delocalize the charge throughout the molecule, is respon-
sible for the drug–channel interaction. This assertion is sup-
ported by the theoretical work performed by Niño et al.
[38,39].
In identifying derivatives of 4-AP to synthesize, func-
tional groups where chosen that would not completely elimi-
nate the pyridine–pyridinium equilibrium responsible for both
transport and biological activity. The addition of functional-
ity that was capable of some amount of hydrogen bonding,
either as a donor or acceptor, was also considered to be impor-
Fig. 2. 4-AP derivatives tested in this study.

910
D.T. Smith et al. / European Journal of Medicinal Chemistry 40 (2005) 908–917
For determining activity in vivo, a normal regimen of soma-
but two animals these “flat-line” recordings were character-
istic of the 1 week recordings as well, indicating there was
little evidence for spontaneous recovery of conduction in any
of the guinea pigs 1 week after surgery. In the two excep-
tions, a small and very early arriving EP (approximately 25 ms
latency) was noted, however, the magnitude of this peak was
barely detectable above baseline and it was not dependably
evoked. Therefore, gavage and further recordings were car-
ried out on these animals and their data are pooled with the
rest. Gavage of 4-AP and carbamates 9 and 10 was unevent-
ful; however, we were unable to carry out this procedure using
the t-butyl carbamate 11. Aqueous solubility proved to be
problematic with this carbamate and the initial attempts at
producing a solution of known concentration for oral inser-
tions were unsuccessful. As a result, we can provide only the
preliminary in vitro data below for comparison purposes.
4-AP produced a strong and dependable enhancement of
EPs within 30–60 min after insertion as observed by a return
of an early arriving EP (over baseline). Two of the 10 animals
failed to respond to 4-AP by 4 h after gavage. There were no
failures to respond when using carbamates 9 and 10.
Even though the sample size was small, we were encour-
aged by an apparent enhancement of EP recovery by methyl
carbamate 9 (Fig. 4). Thus, some statistical comparison
between these groups is provided, however, we caution that
these data should be considered as preliminary since place-
ment and replacement of either stimulating or recording elec-
trodes does affect the size and duration of EP measured over
the sensory cortex. All data were normalized by dividing the
area (expressed as pixels) beneath recovered EP by the same
pre-injury data.
tosensory evoked potential (SSEP) testing was carried out in
the intact animal. A standardized injury to the guinea pig spi-
nal cord was induced surgically. Upon recovery from surgery
and stabilization of the injury (~1 week), a baseline ability of
the spinal cord to conduct an impulse from the tibal nerve of
the hind limb to the contralateral cortex was determined. The
animal was dosed with the test compound by gavage and the
amount of increase in the evoked potential (EP) over baseline
was determined.
4
. Results
4
.1. In vitro testing of conduction through guinea pig
spinal cord
A total of eight 4-AP derivatives have been tested in vitro
in order to determine their ability to restore conduction in
stretched spinal cord. DMAP (6), the amides (7 and 8) and
the ureas (12 and 13) all failed to produce any enhancement
in the recovery CAP amplitude at the same concentration that
is most effective for 4-AP (100 µM). The negative results are
not shown here.
Three compounds, N-(4-pyridyl) methyl carbamate (9);
N-(4-pyridyl) ethyl carbamate (10) and N-(4-pyridyl) t-butyl
carbamate (11), demonstrated the ability to restore conduc-
tion through injured spinal cord. Pilot trials with these three
carbamates established that working concentrations equal to
or below that of 4-AP were capable of inducing a reproduc-
ible recovery of conduction in the double sucrose gap record-
ing chamber. In this type of experiment, 4-AP has been shown
to induce a recovery of about 25% at 100 µM [42]. For methyl
carbamate 9, a concentration of 100 µM produced an increase
in the amplitude of the recovered CAP averaging
Given these precautions, it is likely that carbamate 9 sig-
nificantly enhanced the recovery of EP compared to 4-AP at
1 week post injury. Both two-wayANOVA and a straight for-
ward Student’s t-test confirmed a significant increase
3
0.6 ± 11.7% (n = 4). For ethyl carbamate 10, the increase
was 18.6 ± 8.7% (n = 8) at the same concentration. The
increases in amplitude were statistically significant (P ≤ 0.04).
t-Butyl carbamate 11, on the other hand, was found to be toxic
to white matter at 100 µM when added to the bathing solu-
tion. Toxicity in these in vitro trials was defined as the sig-
nificant reduction of the CAP after exposure to the test com-
pound. It was found, however, that at a concentration of 1 µM,
carbamate 11 reproducibly increased the recovery CAP ampli-
tude by 15.4 ± 3.4% (n = 5) in the absence of toxicity. This
increase was very significant relative to the pre-drug CAPs
(
ANOVA = 0.02; two-tailed Student’s T = 0.47. Given that
the latter test is more conservative relative to small sample
size we only provide this comparison in Table 1). Even more
striking was the enhancement of EPs by methyl carbamate 9
compared to the structurally similar ethyl carbamate 10
(Table 1). The latter carbamate did not meet significance com-
pared to 4-AP data.
5
. Discussion
(
P ≤ 0.002).
5
.1. Derivative design
For all three carbamates, a 1 h wash returned the enhanced
CAP amplitude to pre-drug levels. Fig. 3 provides examples
of the enhanced CAP electrical records for each of these three
drugs and graphically displays the data. Based on the encour-
aging results seen in the in vitro studies, it was decided to
move the three carbamates to in vivo testing.
It was felt that any potential analogue designed would have
to account for two features salient to 4-AP. First, the acid–
base equilibrium inherent to 4-AP would have to be main-
tained to some extent as it is directly responsible for both the
transported species and the active species. Second, at least
some of 4-AP’s hydrogen bonding ability would need to be
maintained in order to bind with the channel. Chemically,
functionalizing the amine nitrogen was the simplest course
of action that would meet the requirements.
4.2. In vivo testing in guinea pig
Post surgery, SSEPs were eliminated in all animals by the
time electrical records could be accomplished (< 1 h). In all

D.T. Smith et al. / European Journal of Medicinal Chemistry 40 (2005) 908–917
911
Fig. 3. Responses of recovered compound nerve impulses to three 4-AP derivatives. In A, the first electrical record shows the recovered CAP prior to the
addition of ethyl carbamate 10 (the initial small wave form is the stimulus artifact). The second record was taken about 1/2 h after the addition of carbamate 10.
Note the ~ 20% increase in the amplitude of the CAP. The third record shows that the amplitude of this enhanced CAP has fallen back to pre-drug levels after
removal of the drug and a 30 min wash. B shows a similar set of electrical records obtained after introduction of methyl carbamate 9. Again, the increase in
amplitude in the presence of the drug is obvious. This enhancement was only sustained during the administration of the drug and approximate pre-drug levels
were measured after a washing out of the drug. C depicts a similar set of records to A and B, using t-butyl carbamate 11. D provides a histogram of all data.
Carbamates 9 and 10 statistically enhanced CAP amplitudes at 100 µM concentration (P ≤ 0.04; one asterisk), while carbamate 11 statistically improved the
CAP at 1 µM compared to pre-drug amplitudes (P ≤ 0.002; two asterisks).

912
D.T. Smith et al. / European Journal of Medicinal Chemistry 40 (2005) 908–917
lack of activity [26], the replacement of a hydrogen bonding
donor with a non-polar, non-hydrogen bonding methyl group
could also explain the lack of activity.
Similar to 4-AMPI, DMAP demonstrated no ability to
restore conduction in injured spinal cord, indicating that com-
plete methylation of either “hemisphere” of 4-AP is detrimen-
tal to its activity. Unfortunately, these results do not help to
clarify the causes of the reduced biological activity, as either
the steric argument or the hydrogen bonding argument could
be invoked. However, in light of the restorative effects
observed with several sterically larger 4-AP derivatives, we
are inclined to believe that changes in hydrogen bonding abil-
ity is of greater importance than the increase in steric size of
either methylated derivative vs. 4-AP.
In addition to alkylating 4-AP, we were also interested in
testing several acylated derivatives.Acylation of 4-AP’s amine
group would remove a hydrogen bond donor, however, the
introduced oxygen atom(s) might serve as a hydrogen bond
acceptor, interacting with the channels amino acid backbone
and offsetting the loss of the amine proton. This course of
derivatization was expected to have a significant effect on the
molecule’s pK and the pyridine–pyridinium equilibrium. For
a
example, the pK s of the carbamates have been calculated to
a
be between 6.0 and 6.4 [43]. This reduction in pK would
Fig. 4
.
Recovery of spinal cord conduction subsequent to oral administra-
a
tion of N-(4-pyridyl) methyl carbamate (9). An unedited electrical record of
a normal SSEP is shown in A. The waveforms were recorded by electrodes
located over the sedated guinea pig’s sensory cortex after electrical stimula-
tion of the tibial nerve of the hind paw. The bottom three records are indivi-
dual traces of 200 repetitive stimulations of the nerve as discussed in Section
shift the equilibrium in favor of the neutral pyridine, decreas-
ing the amount of pyridinium ion in solution from 99% to
between 4% and 9%.
Easily prepared amides 7 and 8, along with ureas 12 and
13 were chosen as representative members of their respective
classes of 4-AP derivatives. We also chose to synthesize the
three carbamates 9–11. While the carbamates represented a
third class of 4-AP derivative, they also offered the ability to
vary the steric bulk of the carbamate alkyl group, providing
the ability to probe the nature, if any, of the steric require-
ments within the active site.
7. The dotted line (SA) marks the stimulus artifact. An early arriving cortical
potential is shown, approximately 30 ms after stimulation (marked P 1) in
the uninjured guinea pig (A). In B, responses to the oral ingestion of the test
compound are shown in the same animal providing the record in A. The top
record was made 1 h after a crush injury to the midthoracic spinal cord as
described in the text. Note the complete elimination of the cortical potential
after stimulation of the tibial nerve of the hind paw. One week later (second
trace) this failure to conduct impulses through the SCI has changed little.
The bottom record was made 1 h after gastric tube administration of N-(4-
pyridyl) methyl carbamate (9). Note the rapid establishment of a nearly nor-
mal cortical potential (P 1).
The amides (7 and 8) and ureas (12 and 13) induced no
improvement of the recovery CAPs in isolated spinal cords
and, as a result, no in vivo testing was performed with these
compounds. In light of the positive results obtained with the
carbamates, the lack of activity from these derivatives is puz-
zling, as they would seem to fit the structural requirements
set forth equally as well as the carbamates. In particular, the
inability of ureas 12 and 13 to restore conduction is curious.
These compounds represent only slight modifications of two
Table 1
Summary of in vivo results
Test
drug
n
Mean
S.D. S.E.M.
31.0 9.8
Range Statistic
4-AP
10
5
40.0
0–100 4-AP vs. 9,
P = 0.47
Methyl
carba-
mate (9)
Ethyl
carba-
mate
78.0
14.2 5.8
58–100
9
vs.
10,
+
well known K channel blockers (3,4-DAP and 4-AP, respec-
P = 0.003
tively) with only a carbonyl group inserted between amine
groups. At this time, there is no clear explanation as to why
the carbamates exhibit a restorative effect while the amides
and ureas do not.
4
41.5
13.2 6.6
27–55 4-AP vs. 10,
P = 0.93
(10)
The most promising results were obtained with the three
carbamate derivatives of 4-AP (9–11). All three compounds
induced an increase in CAP above the pre-drug level and did
so, at the appropriate concentration, without any indication
of toxicity. These three compounds are the first derivatives of
4-AP to demonstrate the ability to restore conduction in
injured spinal cord. Methyl and ethyl carbamates 9 and 10
DMAP (6) was seen as an ideal a probe to explore the
importance of the amine hydrogens to 4-AP’s biological activ-
ity. A number of published reports indicate that methylation
of the pyridine nitrogen (4-AMPI) [26–30] effectively
removes 4-AP’s channel blocking ability. While “severe”
steric constraints at the active site were used to explain the

D.T. Smith et al. / European Journal of Medicinal Chemistry 40 (2005) 908–917
913
both produced similar recoveries to those previously observed
with 4-AP and did so at the same concentration (100 µM).
The toxicity observed at 100 µM with t-butyl carbamate 11
was unexpected considering its structural similarity with the
methyl and ethyl derivatives. Equally surprising was the level
of restored conduction once the concentration was reduced
On the other hand, such a recovery over the flat-line within
minutes after drug ingestion does suggest a marked recovery
of conductance. Even given the small sample size used to
screen these first compounds as a proof of concept, it is very
likely that methyl carbamate 9 will outperform 4-AP in live
animal testing, and even more certain that this compound will
be more effective than its close analog, ethyl carbamate 10.
While the results detailed herein are an effective proof of
concept that derivatized 4-AP can restore conduction in
injured spinal cord, a number of practical question remain to
be answered. More exploratory work with t-butyl carbamate
11 is required. Early attempts to dose guinea pigs with this
compound suffered from this compound’s relatively low water
solubility. Efforts are continuing to develop an acceptable dos-
age regimen in order to determine whether this derivative is
as active in the animal as it is in vitro. As demonstrated by
carbamate 11, the dose–response relationships of the carbam-
ates needs to be developed in order to determine their opti-
mal dosage so these derivatives’ effectiveness as a treatment
of SCI can be more closely compared with each other and
4-AP. Work is ongoing in order to provide answers to these
questions and the results of will be reported in due course.
100-fold, which was equal to ethyl carbamate 10 and almost
as much as 4-AP itself.
The source of the t-butyl carbamate’s lower effective con-
centration has not yet been determined. The carbamate deriva-
tives were designed to mimic and possibly improve upon
4-AP’ binding. While improved drug–channel interaction cer-
tainly could explain the lower effective concentration of 11,
other explanations can not be ruled out at this time. These
include: a lower pK than 4-AP, increased lipophilicity or a
a
slightly different binding site within the channel. The reduc-
tion in pK shifts the pyridine–pyridinium equilibrium towards
a
the neutral pyridine form, which is likely to increase the rate
of transport and lead to higher intracellular concentrations of
the drug. Similarly, increasing lipophilicity vs. 4-AP should
also lead to an increase in the intracellular concentration.
Finally, carbamate 11 might bind to a different set of amino
acids, altering its activity vs. 4-AP.
A quick stability study was used to determine whether the
carbamate derivatives might be “decomposing” back to 4-AP
under the experimental conditions, leading to false positive
results.A solution was made of each carbamate in both Krebs
solution (the in vitro dosing medium) and simulated human
gastric fluid. The solutions were analyzed by TLC after 24 h
at r.t. to determine whether any decomposition had occurred.
No changes were apparent in any of the samples. While these
qualitative results are not unequivocal proof that the carbam-
ates are the active species, they do suggest that the carbam-
ates are stable to the dosing conditions and are likely to be
the active species, particularly in the in vitro experiments.
For all three carbamates, it is proposed that they restore
6. Conclusion
In summary, several derivatives of 4-AP have been synthe-
sized and tested in isolated spinal cord and whole animal mod-
els of SCI in order to determine their ability to restore con-
duction in injured nerves. The effects of alkylation and
acylation of 4-AP’s amine nitrogen were explored. The alky-
lated compound did not exhibit any detectable ability to
restore conduction in vitro, similar to the results previously
reported for 4-AMPI. Somewhat surprisingly, the amide and
urea derivatives did not demonstrate the ability to restore con-
duction to isolated, injured spinal cord. However, the methyl,
ethyl and t-butyl carbamates 9–11 all allowed recovery of con-
duction through injured spinal cord at or below 4-AP’s opti-
mal concentration, with the t-butyl carbamate exhibiting a
similar level of restoration at 1% of the concentration required
for 4-AP. These results represent the first successful attempt
at synthesizing a 4-AP derivative that maintains 4-AP’s abil-
ity to restore conduction in injured spinal cord and suggests
the possibility of developing additional molecules that could
be useful in circumventing the behavioral cap seen when using
+
conduction in injured spinal cord through a K channel block-
ade mechanism. The structural similarity between these
derivatives and 4-AP along with the similarity in restoration
of CAPs suggests to us that a similar mechanism is in effect.
True “proof” of a channel blockade mechanism will need to
be determined through the use of patch clamp electrophysi-
ology. Efforts to further understand the mechanisms by which
these derivatives restore conduction are currently underway
and will be reported in due course.
4
-AP as a treatment of conditions that result from demyeli-
5.2. In vivo testing
nation.
In vivo recording of SSEPs provide data supportive of the
of test compounds’ ability to restore conduction through the
spinal cord lesion. As mentioned, one should be cautious to
not over interpret the quantitative data. For example, a recov-
ery to 100% of the normal pre-injury recording should not be
interpreted as evidence of an absolute recovery of normal EPs
conducted through the lesion since the amplitude and dura-
tion of these waveforms are sensitive to electrode placement.
7. Experimental protocols
7.1. Chemistry
DMAP (6) was purchased from Sigma-Aldrich, Milwau-
kee, WI. Amides 7 [44] and 8 [45], along with ureas 12 [46]
and 13 [47], were synthesized using published methods. 4-AP

914
D.T. Smith et al. / European Journal of Medicinal Chemistry 40 (2005) 908–917
was purchased from Richman Chemical Co., Lower Gw-
ynedd, PA. All reagents were used as received without fur-
ther purification. Melting points were determined in capil-
lary tubes using a Thomas Hover melting point apparatus and
are uncorrected. NMR spectra were obtained on a Bruker
ARX-300 instrument using the indicated solvent.
7.2. Biology
In vitro and in vivo testing in guinea pigs with SCI was
used to test for a given compound’s ability to restore conduc-
tion in injured spinal cord white matter. 4-AP was used as a
comparator compound. These procedures, along with con-
struction of the double sucrose gap chamber, have been thor-
oughly described previously [40–42,48–50] and were car-
ried out in accordance with policies set forth by the United
States’ National Institutes of Health. Herein are brief but spe-
cific details of the double sucrose gap recording, SSEP and
surgery regimes.
7.1.1. Synthesis of N-(4-pyridyl) methyl carbamate (9)
To a solution of 4-aminopyridine (15.0 g, 160 mmol) in
CH Cl (200 ml) at 0 °C was added triethylamine (30.0 ml,
2
2
212 mmol) then methyl chloroformate (15.0 ml, 192 mmol).
The resulting mixture was allowed to warm to r.t. overnight
then was concentrated. The solid products were slurried with
saturated aqueous NaHCO , stirred for 1 h, concentrated and
3
7.2.1. In vitro testing: double sucrose gap recording
Fig. 5 shows the double sucrose gap recording chamber.
The entire spinal cord of a guinea pig is quickly dissected
from the deeply anesthetized animal, cut into 40–45 mm strips
of predominately white matter and placed in aerated Krebs
solution. There are three large compartments which contain
different bathing media: a central one in which oxygenated
Krebs solution is continuously pumped and withdrawn (by
aspiration) and two end chambers filled with isotonic KCl.
The length of the spinal cord spans all three chambers along
with two small reservoirs filled with isotonic sucrose which
is continuously perfused at a rate of 1 ml/min. Stimulation at
one end of the spinal cord produces CAPs that are conducted
through the white matter to be recorded at the other end of
the chamber.
dried under vacuum. The crude product was slurried with hot
MeOH (500 ml) for 1 h, filtered and the filtrate was concen-
trated. The crude carbamate was recrystallized from water to
give 15.9 g (66% yield) of pure methyl carbamate 9.
1
M.p. = 168–170 °C. H NMR (300 MHz, MeOH-d ) d 8.49
4
(
d, J = 6.5 Hz, 2H, Ar); 7.67 (d, J = 6.5 Hz, 2H, Ar); 5.22 (br
13
s, 1H, N–H); 3.94 (s, 3H, OMe). C NMR (75 MHz, MeOH-
d4) d 156.0 (C=O); 151.0 (Ar); 149.2 (Ar); 114.3 (Ar); 53.3
(
OMe). Anal. C H N O (MW = 152.15).
7 8 2 2
7.1.2. Synthesis of N-(4-pyridyl) ethyl carbamate (10)
To a solution of 4-aminopyridine (20.0 g, 212 mmol) in
CH Cl (200 ml) at 0 °C was added triethylamine (30.0 ml,
2
2
212 mmol) and ethyl chloroformate (20.3 ml, 212 mmol).
The resulting mixture was allowed to warm to r.t. overnight
then was concentrated. The solid products were slurried with
Typically, the spinal cord is allowed to stabilize in the
recording chamber for about 1 h in order to produce a char-
acteristic response to electrical stimulation. Subsequently the
saturated aqueous NaHCO , stirred for 1 h, concentrated and
3
dried under vacuum. The crude product was slurried with hot
MeOH (500 ml) for 1 h, filtered and the filtrate was concen-
trated. The crude carbamate was recrystallized from
toluene/hexanes to give 31.8 g (90% yield) of pure ethyl car-
1
bamate 10. M.p. = 127–128 °C. H NMR (300 MHz, CDCl )
3
d 10.02 (s, 1H, N–H); 8.56 (d, J = 6.2 Hz, 2H, Ar); 7.63 (d,
J = 6.2 Hz, 2H, Ar); 4.33 (q, J = 7.1 Hz, 2H, OCH ); 1.38 (t,
2
13
J = 7.1 Hz, 3H, Me). C NMR (75 MHz, CDCl ) d 154.3
3
(
C=O); 150.2 (Ar); 147.3 (Ar); 113.3 (Ar); 61.8 (OCH ); 14.8
2
(
Me). Anal. C H N O (MW = 166.18).
8
10 2 2
7.1.3. Synthesis of N-(4-pyridyl) t-butyl carbamate (11)
To a solution of 4-aminopyridine (50.0 g, 531 mmol) in
triethylamine/CH Cl (1:1, 200 ml) at 0 °C was slowly added
2
2
a solution of di-t-butyl-dicarbonate (116 g, 531 mmol) in
CH Cl (150 ml). The resulting mixture was allowed to warm
2
2
Fig. 5. Double sucrose gap recording and isolation chamber. The isolated
to r.t. overnight then was concentrated. The crude product
was taken up in hot EtOAc, filtered and precipitated with hex-
anes. The precipitate was collected by filtration, washed with
hexanes and dried under vacuum to give 91.0 g (88% yield)
spinal cord (stippled band) is shown mounted in the chamber, with the cen-
tral well continuously perfused with oxygenated Krebs solution (similar to
extracellular fluid). The test drugs were added to this chamber. The two ends
of the spinal cord lie in separate wells filled with isotonic KCl (similar to
intracellular fluid) divided from the central well by narrow gaps filled with
isotonic sucrose solution. Electrodes were formed from Ag/AgCl wires.
Action potentials were generated through a pair of electrodes at the left hand
sucrose gap, conducted through the central part of the spinal cord and recor-
ded by another pair of electrodes in the right-hand gap by conventional bridge
amplification techniques. Injury to the cord is performed at its approximate
mid-point, within the central chamber.
1
of pure t-butyl carbamate 11. M.p. = 147–148 °C. H NMR
(
2
300 MHz, CDCl ) d 8.97 (s, 1H, N–H); 8.39 (d, J = 5.2 Hz,
H, Ar); 7.41 (d, J = 5.2 Hz, 2H, Ar); 1.46 (s, 9H, t-Bu). C
3
1
3
NMR (75 MHz, CDCl ) d 153.1 (C=O); 150.3 (Ar); 147.3
3
(
Ar); 113.0 (Ar); 81.5 (O–C–R ); 28.6 (Me). Anal.
3
C H N O (MW = 194.23).
10 14 2 2

D.T. Smith et al. / European Journal of Medicinal Chemistry 40 (2005) 908–917
915
cord was injured by stretching which is described in detail in
previous publications [41,51]. The injury induces a transi-
tory loss in CAP propagation across the lesion, which gener-
ally improves with time. Once spontaneous recovery pro-
duces a stable “recovery CAP”, drug is added to the medium
bathing the central chamber. Recording of CAPs is continu-
ous, however about 0.5 h is required for the drug induced
changes in amplitude to stabilize. The response to the drug is
reported as an increase (or decrease) in the “pre-drug” recov-
ered potential (which is normalized to 100%). Subsequently
the drug is washed out of the chamber, and the cords lesion
exposed to fresh aerated Krebs solution for approximately
1
h prior to obtaining “post-drug” electrical recordings.
7.2.2. In vivo testing
Large (>350 g) adult female guinea pigs were sedated (see
Section 7.2.3 for details) and a normal regimen of SSEP test-
ing was carried out in the intact animal. This included record-
ing EPs from the sensorimotor cortex in response to stimula-
tion of the median nerve of the contralateral fore limb and the
median nerve of the contralateral hind limb. After these base-
line records were obtained, animals were deeply anesthe-
tized, the spinal cord surgically exposed and crushed by a
standardized technique. Within an hour, a second set of elec-
trical records was obtained, demonstrating the loss of con-
duction of EP conduction through the fresh lesion. It is impor-
tant to first establish that the neural circuit above the level of
the crush injury was intact and that normal SSEPs could be
recorded subsequent to medial nerve stimulation in this ani-
mal, which constitutes an internal control on SSEP recording
procedures (Fig. 6A).
After these records were obtained, the animals recovered
from anesthesia and were maintained for 1 week. Another set
of EPs were recorded to establish the extent of the remaining
conduction deficit ~ 1 week post injury (Fig. 6D). After these
records were taken, animals were administered either 4-AP,
which was used as a standard, or one of the derivatives by
gavage. The effect that the drugs had on conduction was then
monitored at t = 30 min, 1 h, and then hourly up to 4 h. After
the period of recording was concluded, the animals were
allowed to recover and then were returned to their cages. A
final set of records was obtained the next day (~18 h later)
then the animal was euthanized.
Fig. 6
.
Measurement of SSEPs in the sedated guinea pig. Record A is an
unedited electrical record of early arriving EPs ascending the spinal cord in
response to stimulation of the tibial nerve (peak 1; P 1). The bottom three
traces are individual trains of the 200 stimulations. The single trace above,
along with records B–D, is the averaged waveform. This record was obtai-
ned from a sedated guinea pig prior to spinal cord compression. In B, the
waveform, taken 1 h after spinal cord compression, shows a control stimu-
lation of the medial nerve of the forelimb in the same animal. Note the large
amplitude of the early arriving EP. In C, the complete loss of SSEP conduc-
tance through the spinal cord lesion is revealed subsequent to stimulation of
the tibial nerve of the hind limb at the same time and in the same animal.
This complete loss in the ability of ascending volleys of CAPs to be conduc-
ted through damaged spinal cord white matter is still obvious l week later in
this animal (D). Such records are characteristic of all animals used in this
report. The drawing of the guinea pig shows the control circuit activated by
stimulation of the medial nerve (in green). The neural circuit (in blue) acti-
vated by stimulation of the tibial nerve of the hind limb is shown to be inter-
rupted by damage to the spinal cord.
7.2.3. Surgery
The same method of producing a standardized compres-
sion lesion to the exposed guinea pig spinal cord has been
employed in these labs for over a decade.A constant displace-
ment injury is favored over a constant impact as a means to
reduce the variability of behavioral and anatomical conse-
quences of the injury. For details, the interested reader is
referred to previously published procedures (see [24,49]).
Briefly:Adult guinea pigs (> ~ 350 g; Hartley Strain) were
anesthetized by an intramuscular injection of 100 mg/kg ket-
amine HCl and 20 mg/kg xylazine. A dorsal laminectomy
procedure exposed the spinal cord at about the 12th thoracic
vertebral level (T12) to the first lumbar level (L1). The
exposed cord was crushed using a specially modified forceps
possessing a détente. To immobilize animals for electrical
records, a more gentle sedation was produced by intramus-

916
D.T. Smith et al. / European Journal of Medicinal Chemistry 40 (2005) 908–917
3
cular injection (0.1 cm sodium pentobarbital, 50 mg/ml). At
the end of the study, while the animals were under anesthe-
sia, the guinea pigs were euthanized by increasing the anes-
thetic dose significantly, followed by perfusion/fixation (glu-
taraldehyde in phosphate buffered Ringer’s solution).
pass the tube. The feeding needle is advanced between the
incisors and the beginning of the molars, which initiates a
swallowing or gag reflex, facilitating advancement of the feed-
ing needle to the stomach. Prior to this operation, the correct
needle (sized relative to the size of the guinea pig) is marked
by placing its end adjacent the animal’s last rib and marking
the proximal portion of the needle near the tip of the nose.
This provides an estimate of the required length to advance
to the stomach during gavage. The needle can be connected
to either a syringe or an aspirator allowing material to be intro-
duced into, or withdrawn from, the stomach.
7.2.4. SSEPs
It is the loss of nerve impulse conduction through the white
matter of the spinal cord lesion that is associated with the
catastrophic deficits in behavior observed in SCI [1]. These
volleys of compound nerve impulses ascending and descend-
ing the spinal cord are associated with numerous axons and
synapses, and are referred to as EPs when stimulated syn-
chronously by electrical activation of a compound nerve of
the lower or upper limbs (in the SSEP) or activation of the
cortex (during motor EP recording, or MEPs, not performed
here). This form of stimulation of largely ascending impulses
and then recording them at the contralateral sensorimotor cor-
tex of the brain is referred to as SSEP testing.
Stimulation of the tibial nerve of the hind limb was accom-
plished using a pair of needle electrodes inserted near the
nerve at the popliteal notch of the knee. Similar SSEPs were
evoked by stimulation of the median nerve of the forelimb
with pair of stimulation electrodes. Recording electrodes were
located in the scalp covering the contralateral cortex region,
with an indifferent electrode usually located in the pinna of
the ear.
A complete electrical record for one period of investiga-
tion involved stimulating the nerve 200 times in a train (2 mA
square wave, 200 µs duration at 3 Hz). Three to four sets of
these records were then averaged to produce a single wave-
form for quantitative study. Recording and stimulation used
a Nihon Kohden Neuropak 4 and a Power Mac G-3 com-
puter.
Quantitative evaluation involved measurements of latency
from the initiation of stimulation (noted by the stimulus arti-
fact) to the initiation of the EP, however, the most reliable
and informative comparative data involved a determination
of the area beneath the EP in pixels using IPLab™ spectrum
software (Scanalytics, Farifax, VA). These areas beneath the
peak waveform (that is above baseline) were normalized by
dividing the post injury (or post treatment) EP by the area of
the initial EP recorded in the normal animal. If all (100%)
nerve fibers normally fired in synchrony by the stimulation
regimen before injury were theoretically recruited into con-
duction after a treatment, the average SSEPs should thus reach
unity (1.0).
A starting dosage for guinea pigs was estimated by work-
ing within the range of 4-AP given in clinical cases of paraple-
gia in dogs (0.5–1.0 mg/kg body weight) [24]. In the case of
an approximately 500 g guinea pig, this would result in a
total dosage of about 0.25 mg. A stock solution for gavage
3
was prepared where 0.2 cm of an aqueous solution con-
tained 0.2 mg of drug. This allowed the relative standardiza-
tion of total concentration given to animals, and was facili-
tated by the lack of significant variation in their weights (for
the 10 animals that were tested with 4-AP the mean weight
was 421.5 ± 20.9 g; for methyl carbamate 9 the mean weight
was 411.6 ± 23 g; P ≥ 0.5). The effective dosage for 4-AP
was sufficient to produce an improvement in the electrophysi-
ological record equal to 0.47 ± 0.04 mg/kg. By comparison,
the effective dosage for ethyl carbamate 10 appeared to be
slightly lower (0.37 ± 0.2 mg/kg) but this difference was not
statistically different given the small sample size (P ≥ 0.05).
Acknowledgements
We appreciate the expert technical aid of Debra Bohnert,
Brent Dawson andAliciaAltheizer. This work was supported
by financial support of the Center for Paralysis Research from
the State of Indiana.
References
[
1] R.B. Borgens, Restoring Function to the Injured Human Spinal Cord
Monograph), Springer-Verlag, Heidelburg, 2003.
(
[2] I. Wickelgren, Science 297 (2002) 178.
[3] C.M. Armstrong, A. Loboda, Biophys. J. 81 (2001) 895.
[
[
[
4] M. Crest, E. Beraud-Juven, M. Gola, Perspect. Drug Discov. Des.
5–16 (1999) 333.
5] F.S.P. Chen, D. Fedida, Perspect. Drug Discov. Des. 15–16 (1999)
27.
6] T. Baukrowitz, G. Yellen, Proc. Natl. Acad. Sci. USA 93 (1996)
13357.
1
2
7
.2.5. Drug administration
Drugs were administered by gavage using the following
[7] A.P. Southan, D.G. Owen, Br. J. Pharmacol. 122 (1997) 335.
[
8] K.M. Giangiacomo, J. Gabriel, V. Fremont, T.J. Mullmann, Perspect.
Drug Discov. Des. 15–16 (1999) 167.
methods. All tested drugs were introduced, in solution form,
into the guinea pig stomach using a round tipped feeding
needle (either 18 gauge, 55 mm length or 16 gauge, 75 mm
length; Fine Science Tools). Gavage is only carried out on
the sedated animal. In guinea pigs, the soft palate is continu-
ous with the base of the tongue with only a small opening to
[
[
9] A.R. Blight, J.A. Gruner, J. Neurol. Sci. 82 (1987) 145.
10] R.R. Hansebout, A.R. Blight, S. Fawcett, K. Reddy, J. Neurotrauma
1
0 (1993) 19.
[
11] P.J. Potter, K.C. Hayes, J.T.C. Hsieh, G.A. Delaney, J.L. Segal, Spinal
Cord 36 (1998) 147–155.

D.T. Smith et al. / European Journal of Medicinal Chemistry 40 (2005) 908–917
917
[
[
12] P.J. Potter, D.C. Hayes, J.L. Segal, J.T.C. Hsieh, S.R. Brunneman,
[31] A. Albert, R. Goldacre, J. Phillips, J. Chem. Soc. (1948) 2240.
[32] P.K. Wagoner, G.S. Oxford, Biophys. J. 58 (1990) 1481.
[33] B. Robertson, D.G. Owen, Br. J. Pharmacol. 109 (1993) 725.
[34] N.A. Castle, S.R. Fadous, D.E. Logothetis, G.K. Wang, Mol. Pharma-
col. 46 (1994) 1175.
[35] N.A. Castle, S.R. Fadous, D.E. Logothetis, G.K. Wang, Mol. Pharma-
col. 45 (1994) 1242.
[36] R. Bouchard, D. Fedida, J. Pharmacol. Exp. Ther. 275 (1995) 864.
[37] H. Zhang, B. Zhu, J.-A. Yao, G.-N. Tseng, J. Pharmacol. Exp. Ther.
287 (1998) 332.
G.A. Selaney, D.S. Tierney, D. Mason, J. Neurotrauma 10 (1998) 837.
13] J. Qiao, K.C. Hayes, J.T.C. Hsieh, P.J. Potter, G.A. Delaney, J. Neu-
rotrauma 14 (1997) 135–149.
14] J.L. Segal, S.R. Brunnemann, Pharmacotherapy 17 (1997) 415.
15] K.C. Hayes, Restor. Neurol. Neurosci. 6 (1994) 259–270.
16] Acorda Therapeutics. Acorda Therapeutics Reports Results of
Fampridine-SR Clinical Trials (Press Release). April 14, 2004.
17] E.F. Targ, J.D. Kocsis, Brain Res. 328 (1985) 358.
[
[
[
[
[
[
[
18] K.C. Hayes, A.R. Blight, P.J. Potter, Paraplegia 31 (1993) 216.
19] S.G. Waxman, J.M. Ritchie, Ann. Neurol. 33 (1993) 121.
20] J.L. Segal, M.S. Pathak, J.P. Hernandes, Pharmacotherapy 19 (1999)
[38] A. Niño, C. Muñoz-Caro, Biophys. Chem. 91 (2001) 49.
[39] A. Niño, C. Muñoz-Caro, R. Carbó-Dorca, X. Gironés, Biophys.
Chem. 104 (2003) 417.
7
13.
[
21] I. Grijalva, G. Guízar-Sahagún, G. Castañeda-Hernández, D. Mino,
H. Maldonado-Julián, G. Vidal-Cantú, A. Ibarra, O. Serra, H. Sal-
gado-Ceballos, R. Arenas-Hernández, Pharmacotherapy 23 (2003)
[40] R. Shi, A.R. Blight, J. Neurophysiol. 76 (1996) 1572.
[41] J.M. Jensen, R. Shi, J. Neurophysiol. 90 (2003) 2334.
[42] R. Shi, A.R. Blight, Neuroscience 77 (1997) 553.
[43] Calculated using Advanced Chemistry Development (ACD/Labs)
Software Solaris V4.67 (1994–2005 ACD/Labs) and reported on Sci-
Finder Scholar.
8
23.
[
[
22] C.M. Stork, Clin. Toxicol. 32 (1994) 583.
23] M.A. van der Bruggen, B.M. Huisman, H. Beckerman, F.W. Bertels-
mann, C.H. Polman, G.F. Lankhorst, J. Neurol. 248 (2001) 665.
24] A.R. Blight, J.P. Toombs, M.S. Bauer, W.R. Widmer, J. Neurotrauma 8
[44] S. Ghosh,A. Krishnan, P.K. Das, S. Ramakrishnan, J.Am. Chem. Soc.
125 (2003) 1602.
[
[
(1991) 103.
[45] T. Kato, Y. Yamamoto, S. Takeda, Yakugaku Zasshi 93 (1973) 1034.
[46] N.A. Meanwell, S.Y. Sit, J. Gao, H.S. Wong, O. Gao, D.R. St. Laurent,
N. Balasubramanian, J. Org. Chem. 60 (1995) 1565.
[47] A.L. Kovalenko, Y.V. Serov, A.A. Nikonov, I.V. Tselinskii, Z. Org.
Khim. 27 (1991) 2074.
[48] R. Shi, R.B. Borgens, J. Neurophysiol. 81 (1999) 2406.
[49] R.B. Borgens, R. Shi, T.D. Bohner, J. Exp. Biol. 205 (2002) 1.
[50] L.J. Moriarty, B.S. Duerstock, C.L. Bajaj, K. Lin, R.B. Borgens, J.
Neurol. Sci. 155 (1998) 121.
25] A.D. Hertog, P. Biessels, J. Van den Akker, S. Agoston, A. Horn, Eur.
J. Pharmacol. 142 (1987) 115.
[26] J.R. Howe, J.M. Ritchie, J. Physiol. 433 (1991) 183.
[27] J.K. Hirsh, F.N. Quandt, J. Pharmacol. Exp. Ther. 267 (1993) 604.
[28] G.J. Stephens, J.C. Garratt, B. Robertson, D.G. Owen, J. Physiol. 477
(1994) 187.
[
[
29] J.W. Wegener, A. Peiter, S.R. Sampson, H. Nawrath, J. Cardiovasc.
Pharmacol. 32 (1998) 134.
30] G.E. Kirsch, T. Narahashi, J. Pharmacol. Exp. Ther. 226 (1983) 174.
[51] R. Shi, J.D. Pryor, Neuroscience 110 (2002) 765.