Polyamine analogue regulation of NMDA MK-801 binding: A structure-activity study

Article abstract of DOI:10.1021/jm960545x

A series of analogues and homologues of spermine were synthesized, and their impact on MK-801 binding to the N-methyl-D-aspartate (NMDA) receptor was evaluated. These tetraamines encompass both linear and cyclic compounds. The linear molecules include norspermine, N¹,N¹¹-diethylnorspermine, N¹,N¹²-bis(2,2,2-trifluoroethyl)spermine, homospermine, and N¹,N¹⁴-diethylhomospermine. The cyclic tetraamines consist of the piperidine analogues N¹,N³-bis(4-piperidinyl)-1,3-diaminopropane, N¹,N⁴-bis(4-piperidinyl)-1,4-diaminobutane, N¹,N⁴-bis(4-piperidinylmethyl)-1,4-diaminobutane, and N¹,N⁴-bis[2-(4-piperidinyl)ethyl]-1,4-diaminobutane and the pyridine analogues N¹,N³-bis(4-pyridyl)-1,3-diaminopropane, N¹,N⁴-bis(4-pyridyl)-1,4-diaminobutane, N¹,N⁴-bis(4-pyridylmethyl)-1,4-diaminobutane, and N¹,N⁴-bis[2-(4-pyridyl)-ethyl]-1,4-diaminobutane. This structure-activity set makes it possible to establish the importance of charge, intercharge distance, and terminal nitrogen substitution on polyamine-regulated MK-801 binding in the NMDA channel. Four families of tetraamines are included in this set: norspermines, spermines, homospermines, and tetraazaoctadecanes. Calculations employing a SYBYL modeling program revealed that the distance between terminal nitrogens ranges between 12.62 and 19.61 ?. The tetraamines are constructed such that within families cyclics and acyclics have similar lengths but different nitrogen pK_a's and thus different protonation, or charge, states at physiological pH. The pK_a values for all nitrogens of each molecule and its protonation state at physiological pH are described. The modifications at the terminal nitrogens include introduction of ethyl and β,β,β-trifluoroethyl groups and incorporation into piperidinyl or pyridyl systems. The studies clearly indicate that polyamine length, charge, and terminal nitrogen substitution have a significant effect on how the tetraamine regulates MK-801 binding to the NMDA receptor. Thus a structure-activity basis set on which future design of MK-801 agonists and antagonists can be based is now available.

Full text of DOI:10.1021/jm960545x

J . Med. Chem. 1996, 39, 5257-5266
5257
P olya m in e An a logu e Regu la tion of NMDA MK-801 Bin d in g: A
Str u ctu r e-Activity Stu d y
Raymond J . Bergeron,* William R. Weimar, Qianhong Wu, Yang Feng, and J ames S. McManis
Department of Medicinal Chemistry, University of Florida, Box 100485, Gainesville, Florida 32610-0485
Received J uly 12, 1996X
A series of analogues and homologues of spermine were synthesized, and their impact on MK-
801 binding to the N-methyl-D-aspartate (NMDA) receptor was evaluated. These tetraamines
encompass both linear and cyclic compounds. The linear molecules include norspermine, N1,N11-
diethylnorspermine, N1,N12-bis(2,2,2-trifluoroethyl)spermine, homospermine, and N1,N14-di-
ethylhomospermine. The cyclic tetraamines consist of the piperidine analogues N1,N3-bis(4-
piperidinyl)-1,3-diaminopropane, N1,N4-bis(4-piperidinyl)-1,4-diaminobutane, N1,N4-bis(4-
piperidinylmethyl)-1,4-diaminobutane, and N1,N4-bis[2-(4-piperidinyl)ethyl]-1,4-diaminobutane
and the pyridine analogues N1,N3-bis(4-pyridyl)-1,3-diaminopropane, N1,N4-bis(4-pyridyl)-1,4-
diaminobutane, N1,N4-bis(4-pyridylmethyl)-1,4-diaminobutane, and N1,N4-bis[2-(4-pyridyl)-
ethyl]-1,4-diaminobutane. This structure-activity set makes it possible to establish the
importance of charge, intercharge distance, and terminal nitrogen substitution on polyamine-
regulated MK-801 binding in the NMDA channel. Four families of tetraamines are included
in this set: norspermines, spermines, homospermines, and tetraazaoctadecanes. Calculations
employing a SYBYL modeling program revealed that the distance between terminal nitrogens
ranges between 12.62 and 19.61 Å. The tetraamines are constructed such that within families
cyclics and acyclics have similar lengths but different nitrogen pKa’s and thus different
protonation, or charge, states at physiological pH. The pKa values for all nitrogens of each
molecule and its protonation state at physiological pH are described. The modifications at the
terminal nitrogens include introduction of ethyl and â,â,â-trifluoroethyl groups and incorpora-
tion into piperidinyl or pyridyl systems. The studies clearly indicate that polyamine length,
charge, and terminal nitrogen substitution have a significant effect on how the tetraamine
regulates MK-801 binding to the NMDA receptor. Thus a structure-activity basis set on which
future design of MK-801 agonists and antagonists can be based is now available.
In tr od u ction
Recently a great deal of attention has been focused
on the N-methyl-D-aspartate (NMDA) receptor as a
target for the development of central nervous system
(CNS) related therapeutics. The NMDA subtype of the
glutamate receptor is a ligand-gated ion channel in-
volved in excitatory neurotransmission in the mam-
malian CNS.1 Activation of the NMDA receptor-
channel complex has been implicated in several phys-
iological phenomena important to higher order CNS
functions including long-term potentiation2 and neu-
ronal plasticity.3 Overstimulation of this receptor re-
sults in an inflow of Ca2+ and neuronal excitotoxicity
important in the pathophysiology of epilepsy4 and
F igu r e 1. Diagram depicting the mammalian N-methyl-D-
aspartate (NMDA) receptor complex and some of the ligand
binding sites important to receptor function.
ischemia-induced neuron death.5 This ligand-gated
ionotropic glutamate receptor is subject to complex
regulation by a number of ligands, some of which are
illustrated in Figure 1.6 Separate regulatory sites
include the binding site (2) for the agonist L-glutamate;
a high-affinity binding site (3) for the obligatory coago-
nist, glycine;7,8 a site (4) where Zn2+ acts to allosterically
inhibit the agonist-induced response independently of
membrane potential;9,10 a site(s) (5) within the chan-
nel11,12 where Mg2+ and phencyclidine (PCP), (+)-5-
methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5,10-
imine (dizocilpine, MK-801), N-(1-[2-thienyl]cyclohexyl)-
piperidine (TCP), and ketamine bind13-15 to produce a
voltage-dependent open channel block; and a distinct
binding site (1) for the polyamines, spermine and
spermidine, which modulate NMDA receptor func-
tion.16,17
Spermine and spermidine serve as potent allosteric
effectors of the NMDA receptor. Both biochemical and
electrophysiological studies support a “glycine indepen-
dent” stimulation at saturating glycine concentrations
and mild glycine dependent stimulation, in which a
small spermine-dependent increase in glycine binding
has been demonstrated.
Initial studies on the impact of these polyamines on
dizocilpine binding to the NMDA receptor revealed a
kind of biphasic polyamine action, which was agonistic
* To whom correspondence should be addressed.
X Abstract published in Advance ACS Abstracts, December 1, 1996.
S0022-2623(96)00545-6 CCC: $12.00 © 1996 American Chemical Society

5258 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26
Bergeron et al.
Sch em e 1.a Synthesis of Homospermine (HSPM) (4)
at low concentration and antagonistic at high concen-
tration.6,18 Electrophysiological studies have suggested
that the voltage dependent inhibition by spermine, or
antagonistic effect, probably involves an open channel
block and/or a screening of surface charges near the
opening of the ion channel.19 Thus it seems that
potentiation and inhibition at least by spermine are
consistent with two separate binding sites on the NMDA
receptor-channel complex.20
Terminally N-dialkylated analogues and homologues
of the naturally occurring tetraamine spermine exhibit
a number of therapeutically useful properties. In fact
phase I clinical trials are ongoing with DENSPM as an
antineoplastic, and DEHSPM is in phase II clinical
trials as an agent for the treatment of AIDS-related
diarrhea (ARD).
The N-alkylated spermine analogues and homologues
exhibit antineoplastic activity against a number of
murine and human tumor lines both in vitro and in
vivo.21,22 These compounds have been shown to utilize
the polyamine transport apparatus for incorporation,23,24
to deplete polyamine pools,25 to drastically reduce the
level of ODC26,27 and AdoMetDC activities,28 and in
some cases to upregulate SSAT.29-33 Interestingly, very
small structural alterations in these polyamine ana-
logues can result in marked differences in their biologi-
cal activities.24 For example DENSPM stimulates SSAT
by 1200-fold in MALME-3 cells, whereas DEHSPM
stimulates SSAT by only 30-fold.30,31,34 Differences in
the behavior of these analogues are also reflected in vivo
in terms of the therapeutic windows.35,36
Interestingly, just as with the antineoplastics, small
alterations in the polyamine backbone can have a
tremendous impact on antitransit properties of the
analogues. For example, DENSPM is 10 times less
effective than DEHSPM as an antidiarrheal. This has
been demonstrated in a number of animal models and
recently in the clinic against ARD.37
In an earlier study we were able to demonstrate that
a number of terminally dialkylated polyamine analogues
had a significant impact on binding of MK-801 to the
NMDA receptor in isolated rat brain cortical vesicles.6
Of the analogues studied we were able to show a graded
biphasic behavior. While none of the compounds inves-
tigated was as agonistic as spermine itself, they nev-
ertheless all exhibited some agonistic behavior. This
of course could translate to a potential toxicity problem
if a polyamine analogue as a drug accumulates in the
CNS to any significant extent.
In view of the biphasic behavior of the analogues
investigated in the earlier study, we have elected to
explore in a more systematic way how structural
alterations of the polyamine analogues affect NMDA
MK-801 binding. Clearly an understanding of such an
SAR would help to facilitate the removal of potential
neurotoxicities. Thus in the current study we evaluate
the effects of chain length, charge, and terminal nitro-
gen substitution on the NMDA MK-801 binding proper-
ties.
a
Reagents: (a) N-(4-bromobutyl)phthalimide (2 equiv)/NaH/
DMF; (b) (NH2)2/EtOH; (c) 30% HBr in HOAc/PhOH/CH2Cl2 then
HCl.
tetraamine 2. Removal of the imide protecting groups
of 2 with hydrazine hydrate in refluxing ethanol re-
sulted in N5,N10-bis(mesitylenesulfonyl)homospermine
(3). The internal sulfonyl groups were cleaved using
30% HBr in HOAc/PhOH to give HSPM (4) as its
tetrahydrochloride salt. This method is a viable alter-
native to the literature preparation of HSPM,39 in which
the internal nitrogens were masked as their N-benzy-
lamines.
The aliphatic cyclic tetraamines, N1,N3-bis(4-piperidi-
nyl)-1,3-diaminopropane [PIP(3,3,3)] (9) and N1,N4-bis-
(4-piperidinyl)-1,4-diaminobutane [PIP(3,4,3)] (10), were
also prepared utilizing the mesitylenesulfonyl amine
protecting group and segmenting reagents (Scheme
2).38,40 4-Amino-1-benzylpiperidine was converted to
4-aminopiperidine dihydrochloride (5)41 employing mild
catalytic reduction conditions (Pd-C/methanolic HCl).
Treatment of 5 with mesitylenesulfonyl chloride under
biphasic conditions (NaOH/CH2Cl2) furnished N,N′-bis-
(mesitylenesulfonyl)-4-aminopiperidine (6). Alkylation
of 6 with 1,3-dichloropropane (0.5 equiv)/NaH/DMF
provided the tetrasulfonamide 7, which was cleanly
converted with 30% HBr in HOAc (PhOH/CH2Cl2) to
bicyclic DENSPM analogue 9. Reaction of 6 with 1,4-
diiodobutane (0.5 equiv)/NaH/DMF gave masked poly-
amine 8, which was deprotected as before to external
dipiperidinyl DESPM analogue 10.
N1,N3-Bis(4-pyridyl)-1,3-diaminopropane [PYR(3,3,3)]
(11) and N1,N4-bis(4-pyridyl)-1,4-diaminobutane [PYR-
(3,4,3)] (12), DENSPM and DESPM analogues, respec-
tively, in which the terminal ethyl groups have been tied
back into the polyamine chain and the rings aromatized,
have been synthesized by heating the hydrochloric acid
salts of 4-bromopyridine and the appropriate diamine
above 200 °C (Scheme 3).42
Molecu la r Geom etr y. In order to assess the sig-
nificance of chain length in the polyamine regulation
of MK-801 binding to the NMDA receptor, the distances
between internal and external nitrogens were deter-
mined for each tetraamine using a SYBYL molecular
modeling program (Table 1). Two distances were cal-
culated: NR-Nâ, the distance between the external
Syn th esis. Homospermine (HSPM), the predomi-
nant metabolic product from N1,N14-diethylhomosper-
mine (DEHSPM),36 was synthesized for testing in the
NMDA receptor according to Scheme 1. N1,N4-Bis-
(mesitylenesulfonyl)putrescine (1)38 was converted to its
dianion (NaH/DMF) and alkylated with N-(4-bromobu-
tyl)phthalimide (2 equiv), generating fully protected

Regulation of NMDA MK-801 Binding
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26 5259
Sch em e 2.a Synthesis of DIPIP(3,3,3) (9) and DIPIP(3,4,3) (10)
a
Reagents: (a) H2/10% Pd-C/CH3OH/HCl; (b) mesitylenesulfonyl chloride (2 equiv)/NaOH/CH2Cl2; (c) 1,3-dichloropropane (0.5 equiv)/
NaH/DMF; (d) 1,4-diiodobutane (0.5 equiv)/NaH/DMF; (e) 30% HBr in HOAc/PhOH/CH2Cl2 then HCl.
Sch em e 3.a Synthesis of DIPYR(3,3,3) (11) and DIPYR(3,4,3) (12)
Ta ble 1. SYBYL-Generated Distances between Nitrogens on
Polyamine Analogues
However, the turbidity completely cleared just after the
first equivalent of acid had been added during the
polyamines
NR-Nâ (Å)
NR-NR′ (Å)
titration (ca. pH 9.3), and exactly 2 equiv of acid were
consumed prior to pH 8. There were only these two
titratable groups in the pH range 2-12. The titration
curve itself was very well-fitted to the theoretical with
pKa’s of 9.55 and 9.14 as indicated; if there is any error
due to solubility problems, these values are artificially
low. The conclusion that PYR(3,3,3) exists almost
exclusively as a bispyridinium dication at pH 7.4 with
the charges separated by the entire length of the
molecule is corroborated by previous determination of
the pKa’s in 4-aminopyridine43 where the pyridine
nitrogen is quite basic (pKa1 ) 9.15) and the primary
amine is too weak to be titrated in aqueous solution
(pKa2 ) -6.60).
NSPM(3,3,3)
DENSPM(3,3,3)
PIP(3,3,3)
5.028
5.038
4.333
4.177
5.029
5.038
5.006
4.368
4.177
6.361
6.370
5.377
5.020
6.777
6.506
15.093
15.110
13.164
12.617
16.406
16.425
16.352
14.089
13.863
18.978
18.994
17.035
16.275
19.609
18.859
PYR(3,3,3)
SPM(3,4,3)
DESPM(3,4,3)
FDESPM(3,4,3)
PIP(3,4,3)
PYR(3,4,3)
HSPM(4,4,4)
DEHSPM(4,4,4)
PIP(4,4,4)
PYR(4,4,4)
PIP(5,4,5)
PYR(5,4,5)
nitrogen and the first internal nitrogen and NR-NR′, the
distance between the external nitrogens. In this cal-
culation the charges set on the nitrogens were those
calculated for the predominant species at physiological
pH (Table 2). These cationic species were set in their
lowest energy state, and the distances were measured.
On the basis of previous studies carried out on various
4-amino-substituted pyridines,43 we made the assump-
tion that PYR(3,3,3) and PYR(3,4,3) are dications from
protonation of the pyridine nitrogens.
P r oton a tion Sta te Ca lcu la tion s. One of the criti-
cal structure-activity issues is related to the signifi-
cance of polyamine charge on polyamine regulation of
MK-801 binding in the NMDA channel. Thus we
designed a series of tetraamines which we anticipated
would have similar dimensions but different protonation
states, e.g., DESPM vs FDESPM or PIP(3,4,3) vs PYR-
(3,4,3) (Table 2). Polyamine hydrochlorides were first
converted to free amines by addition of base, and
titrations were carried out in 0.1 N KCl. The pKa’s were
estimated from a computer fit of the titration curves
utilizing an MINSQ.2.8 nonlinear parameter estimation.
The fraction of various protonated species was calcu-
lated using the same program in the simulation mode
given the measured pKa values. However, in the case
of PYR(3,4,3) insolubility of the free base at high pH
precluded pKa measurement. In the case of PYR(3,3,3),
addition of base resulted in a very fine dispersion.
Nor sp er m in e a n d An a logu es. Like spermine, nor-
spermine (NSPM) is biphasic in its impact on MK-801
binding but has a more protracted concentration range
in which it is agonistic (Figure 2a). The agonistic phase
begins at 0.5 µM, peaks at 10 µM with a 33% binding
increase and approaches an antagonistic frame around
1 mM. The agonistic behavior lasts over a 1 mM span.
Although NSPM never really becomes an antagonist
over the concentration evaluated, it is clearly moving
in that direction. DENSPM also exhibits biphasic
activity, but relative to NSPM the curve is shifted to
the left. The agonistic behavior, albeit somewhat weak,
begins at 0.1 µM, peaks at 1 µM with a 13% binding
enhancement, and moves into an antagonistic mode
above 2.0 µM. Thus the agonistic phase spans a 1.9 µM
range. At 25 µM DENSPM, MK-801 binding is reduced
by 50%. We shall refer to this as the antagonistic IC50
value (A IC50). The piperidine analogue PIP(3,3,3) is
the most potent agonist we have identified, increasing
MK-801 by 100%. Its agonistic behavior begins at 0.05
µM, peaks at 500 µM, and begins to fall. Its antagonistic
concentration curve is shifted to the right relative to
NSPM. It seems clear that if indeed a concentration
can be attained in which the tetraamine PIP(3,3,3) is
an antagonist, it would require high millimolar concen-
trations. Finally, the pyridine norspermine analogue,
PYR(3,3,3), while it may demonstrate minor, if any,
agonistic behavior at low concentrations, is very similar

5260 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26
Bergeron et al.
Ta ble 2
to DENSPM in its antagonistic behavior, which begins
around 0.05 µM with a left shift in the concentration vs
binding curve. It has an A IC50 of 11 µM.
tion with an A IC50 of 8 µM and a clear left shift in the
curve relative to SPM. Finally, over the concentrations
studied FDESPM has little if any activity, similar to
putrescine (PUT).17 In both FDESPM and PUT the two
point charges at a pH of 7.4 are separated by four
methylenes.
Sp er m in e a n d An a logu es. Spermine (SPM) has
clear biphasic behavior with an agonistic segment
beginning at 0.1 µM and peaking at 11 µM with a 35%
enhancement in MK-801 binding (Figure 2b). The
antagonistic behavior begins at 300 µM. Thus the
agonistic mode continues through a 300 µM concentra-
tion range. DESPM also shows biphasic activity, al-
though not nearly as pronounced as the parent com-
pound. The agonistic onset begins at 0.1 µM and lasts
over a 1.9 µM range with a peak at 1 µM. The
antagonistic phase begins at 2.0 µM with an A IC50 at
10 µM.
Initially PIP(3,4,3) behaves very much like SPM with
onset of agonism beginning at 0.8 µM and peaking at
10 µM with a 45% increase in MK-801 binding. Its
curve is thus shifted to the right relative to SPM.
However, agonistic behavior spans at least into the
millimolar range. In contrast PYR(3,4,3) is essentially
a pure antagonist manifested at a very low concentra-
Hom osp er m in e a n d Hom ologu es. Homospermine
(HSPM) shows much less of an exaggerated biphasic
behavior than spermine (Figure 2c). Agonism begins
at 0.5 µM and peaks at 7.0 µM with a 13% increase in
MK-801 binding; antagonism begins at 10 µM. Thus
the agonistic fragment of the curve persists over a 9.5
µM range. Of the three parent linear tetraamines,
HSPM is clearly the most effective antagonist, with an
80 µM A IC50. DEHSPM behaves like HSPM in its
biphasic activity except that the curve is shifted to the
left. The agonistic portion of the curve begins at 0.1
µM and peaks at 1 µM with a 13% increase in MK-801
binding. The antagonistic phase begins at 2 µM, with
an A IC50 of 10 µM. Thus the agonistic phase runs over
a 1.9 µM range. Of the HSPM analogues, it is the most
potent antagonist. The homospermine piperidine ana-

Regulation of NMDA MK-801 Binding
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26 5261
logue PIP(4,4,4) response curve is shifted to the right
of HSPM itself with the agonistic component beginning
at 0.5 µM and peaking at 8 µM with a 30% enhanced
binding. The antagonistic phase begins at 12 µM. Thus
the agonistic phase encompassed an 11.5 µM range with
moderate antagonistic behavior beyond this point. The
pyridine analogue PYR(4,4,4) is a weak agonist over the
entire range studied beginning at 0.1 µM. However,
MK-801 binding is only increased by 16% at 100 µM.
The PYR(5,4,5) analogue is a pure agonist in the
concentration range studied with a 50% binding en-
hancement at 100 µM with a clear left shift in the curve
relative to HSPM. The agonistic behavior of this
compound begins at very low concentration, <0.05 µM.
Interestingly the corresponding piperidine analogue
PIP(5,4,5) is weakly biphasic with a slightly left shift
and an agonistic component beginning at 0.1 µM and
peaking at 1 µM with a 10% MK-801 binding enhance-
ment. The compound then moves into a potent antago-
nistic phase at 8 µM with a 50 µM A IC50. The agonistic
component lasts only through a 7.9 µM range.
Discu ssion
Earlier studies in our laboratories6 and others have
clearly demonstrated that polyamines16,17 and polyamine
analogues44 have a profound impact on the NMDA
receptor. Probably the most thoroughly studied of the
polyamines, spermine, a naturally occurring linear
tetraamine, has been shown to exhibit concentration
dependent biphasic activity at the level of MK-801
binding in the NMDA channel. The inhibitory and
potentiation properties of this tetraamine involve two
distinct polyamine binding sites. At saturating concen-
trations of glycine, low concentrations of spermine
produce a “glycine independent” increase in NMDA-
induced whole cell currents and a voltage dependent
inhibition at higher spermine concentrations. However,
studies with various cloned NMDA heteromeric recep-
tors strongly suggest that spermine-induced glycine
dependent stimulation is mechanistically different from
glycine independent stimulation.45-48 It seems likely
that the voltage dependent inhibition by spermine
involves an open channel block. However, a still linger-
ing alternative explanation involves the screening of
surface charge on the NMDA receptor by the polyamine.19
An earlier structure-activity investigation of the
effects of polyamine analogues on MK-801 binding in
the NMDA channel suggested the key parameter in
voltage dependent inhibition is the length of the
polyamine chain rather than the nature of the terminal
alkyl substituents.44 The study included a series of
ethylated spermine homologues and analogues, both
tetraamines and pentaamines. However, while the
results were very interesting, the program only focused
on terminally ethylated polyamines, e.g., DENSPM,
DEHSPM, and did not include the corresponding nor-
spermine and homospermine controls. This exclusion
makes it somewhat difficult to assess the significance
of the alkyl groups within a family of particular chain
length. For example, if polyamine chain length is
indeed the most important structural feature in voltage
block, then it becomes problematic to assess the contri-
bution, if any, of terminal alkyl groups on norspermine
when comparing the voltage blocking properties of an
alkylated norspermine with those of an alkylated ho-
mospermine. The significance of these terminal alkyl
F igu r e 2. Modulation of [3H]-(+)-5-methyl-10,11-dihydro-5H-
dibenzo[a,d]cyclohepten-5,10-imine ([3H]MK-801, 2 nM) bind-
ing to extensively washed rat cortical membranes (ca. 300 µg
of protein/mL) in the presence of excess L-glutamate (100 µM)
and glycine (100 µM) by the tetraamines, norspermine (2a ),
spermine (2b), and homospermine (2c) and the corresponding
terminally N-ethylated analogues, and bis(4-piperidinyl/py-
ridyl) homologues. Each experiment was done in triplicate for
each concentration. Data are presented relative to “100%
control” binding observed in the presence of L-Glu and Gly
coagonists, but absent added polyamine modulator. Standard
deviation of the “100% control” binding was typically about
(3%. Data are plotted as means ( standard deviation of four
separate experiments (N ) 4) for each tetraamine and each
tetraamine analogue and homologue.

5262 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26
Bergeron et al.
groups becomes particularly important in the design of
polyamine analogue therapeutics which could impact on
the NMDA receptor as either agonists or antagonists.6
We have already demonstrated that polyamine ana-
logues can be substantially less toxic than the parent
amine.27 We have also demonstrated that it is possible
to metabolically program a polyamine analogue in such
a way as to substantially reduce toxicity associated with
its metabolites.37
Substitution of the ethyl group of DESPM with a
â,â,â-trifluoroethyl group reduces the pKa of the two
terminal nitrogens and thus the fraction of tetracation
from 88% to essentially 0% while raising the level of
dication from 0 to 99%. A similar situation exists for
PIP(4,4,4) and the corresponding pyridyl compound
PYR(4,4,4). The former tetraamine is 92% in the
tetracationic state and nearly 0% in the dicationic state,
while the latter compound is 63% dication and 36%
monocation with no tetracationic form at physiologic pH.
A similar scenario applies to the PIP(5,4,5) and the
PYR(5,4,5). The PIP compound is 98% in the tetraca-
tionic state with no dicationic species. The PYR(5,4,5)
presents almost exclusively as the dication (95%). The
difference in the fraction of monocationic and dicationic
species between PYR(4,4,4) and PYR(5,4,5) is likely
related to the fact that the PYR(4,4,4) has only one
methylene separating the two central nitrogens from
each electron-withdrawing pyridine ring, while PYR-
(5,4,5) has two such insulators.
Thus a consideration of how structural modifications
affect a linear polyamine’s impact on the NMDA recep-
tor could be critical in the design of polyamine ana-
logues. In this regard we have adopted a systematic
approach to assessing the effects of charge, polyamine
chain length, and terminal alkyl group size on the
ability of polyamines to agonize or antagonize MK-801
binding.
We have chosen to evaluate a group of spermine
homologues and analogues including norspermines
(3,3,3), spermines (3,4,3), homospermines (4,4,4), and
tetraazaoctadecanes (5,4,5). The terminal nitrogen
substituents include ethyl and â,â,â-trifluoroethyl, as
well as incorporation of the terminal nitrogens into a
piperidine or pyridine ring.
The most interesting pyridine analogues are PYR-
(3,3,3) and PYR(3,4,3). In both cases there is no
insulator between the internal nitrogens and the aro-
matic ring. PYR(3,3,3) is almost exclusively in the
dicationic state. On the basis of literature studies of
similar molecules, it is likely that only the pyridine
nitrogens of both analogues are protonated.43 Because
of this one would anticipate that PYR(3,3,3) and PYR-
(3,4,3) would behave much like the corresponding long-
chain diamines 1,11-diaminoundecane and 1,12-diami-
nododecane, respectively.
Ch a r ge. Four structural alterations at the terminal
nitrogens were used to control the charge, or protonation
state, of the polyamine analogues (Table 2). The
alterations were all made at the primary terminal
nitrogen. The terminal primary amines of norspermine,
spermine, or homospermine were either ethylated to
DENSPM, DESPM, DEHSPM; incorporated into a pi-
peridyl ring system PIP(3,3,3), PIP(3,4,3), PIP(4,4,4); or
incorporated into a pyridyl system PYR(3,3,3), PYR-
(3,4,3), PYR(4,4,4). The terminal nitrogens of spermine
were also â,â,â-trifluoroethylated. Finally in the tet-
raazaoctadecanes the terminal nitrogen was incorpo-
rated into either a piperidyl or pyridyl ring.
The importance of charge to the tetraamine’s biologi-
cal activity is quite clear from the studies. In the
norspermine series the best comparison is made be-
tween PIP(3,3,3) and PYR(3,3,3). The PIP(3,3,3) tet-
raamine is the most potent MK-801 binding agonist we
have identified, while the aromatic tetraamine is a very
potent antagonist much like 1,10-diaminodecane.49,50
Again this is simply because PYR(3,3,3) has a charge
disposition very much like the true diamine. In the
spermine series the effect of substituting the ethyls of
DESPM with â,â,â-trifluoroethyls is profound. The
resulting FDESPM, a dicationic species at physiologic
pH, is nearly devoid of any activity over the concentra-
tion range studied, while DESPM exhibits biphasic
activity with a potent antagonistic component. In this
tetraamine series the NMDA agonist and antagonist
relationship between the PIP(3,4,3) and PYR(3,4,3)
analogues is similar to that seen with PIP(3,3,3) and
PYR(3,3,3). Thus the tetracation PIP(3,4,3) is a potent
agonist albeit somewhat less active than the shorter
PIP(3,3,3). The PYR(3,4,3) analogue is the most potent
antagonist we have identified. Again, because of the
terminal position of the charges in this molecule at
physiological pH, this tetraamine behaves like 1,12-
diaminododecane.49,50
The trends in the pKa values are in keeping with
simple electrostatics (Table 2). In going from norsper-
mine, a tetraamine in which all of the nitrogens are
separated by three methylenes, through spermine, in
which the terminal and internal nitrogens are separated
by three methylenes and the internal nitrogens by four
methylenes, to homospermine, in which successive
nitrogens are separated by four methylenes, the fraction
of tetracation at physiological pH goes from 54 to 76 to
85%. In each instance fixing an ethyl group to both of
the external nitrogens increases the fraction of tetra-
cation by about 10%. Interestingly in going from the
terminal secondary amines DENSPM, DESPM, and
DEHSPM to the corresponding secondary piperidyl
systems, the fraction of +4 cation drops in each case
for PIP(3,3,3), PIP(3,4,3), and PIP(4,4,4). The longer the
molecule the smaller the decrease. This is in keeping
with the idea that folding the ethyl groups of DENSPM,
DESPM, and DEHSPM back into a piperidyl system
brings the terminal secondary nitrogens even closer to
the internal nitrogens. Thus in the tetracation the
charges would be closer together for the piperidyl
systems, an electrostatically less favorable situation. As
predicted this effect becomes less significant the longer
the methylene insulator between nitrogens, e.g., the
decrease in the +4 cation ratio between DENSPM/PIP-
(3,3,3) and DEHSPM/PIP(4,4,4).
In the homospermine family the significance of charge
in MK-801 binding is further underscored. PIP(4,4,4)
is biphasic, while the corresponding pyridyl compound
PYR(4,4,4) has no antagonistic properties over the range
investigated but is weakly agonistic. Recall that at
physiological pH this PYR(4,4,4) has a fairly substantial
fraction of monocation (36%) with 63% in the form of
the dication. The PIP(5,4,5) compound is slightly bi-
phasic but primarily an antagonist. However and more

Regulation of NMDA MK-801 Binding
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26 5263
notable is the fact that the PYR analogue is a pure
agonist and its agonistic properties relative to PIP(5,4,5)
are greater than those of PYR(4,4,4) relative to its
analogue PIP(4,4,4). Recall that PYR(5,4,5) has a much
higher fraction of dication than PYR(4,4,4) at physi-
ological pH.
PIP compounds, with agonistic activity of PIP(3,3,3) >
PIP(3,4,3) > PIP(4,4,4) > PIP(5,4,5). No such relation-
ship between the concentration span of antagonism and
PYR analogue structure is apparent.
The onset of antagonism also correlates inversely with
the length of the tetraamine for both the parent
polyamines and the piperidyl analogues. However, both
PYR(3,3,3) and PYR(3,4,3) have very low antagonism
onset values, while the corresponding PYR(4,4,4) and
PYR(5,4,5) systems are pure agonists.
Nitr ogen Su bstitu tion . For norspermine, sper-
mine, and homospermine the overall trend in terminal
nitrogen substitution seems clear. Addition of an ethyl
group to the parent amine shifts the dose-response
curves to the left, rendering the tetraamine better
antagonists. However, incorporation of the terminal
nitrogens into a piperidine ring system shifts the dose-
response curves to the right, resulting in compounds
which are better agonists. For both PYR(3,3,3) and
PYR(3,4,3), where the terminal pyridine nitrogens are
protonated, the curves shift to the left, resulting in
compounds that are potent antagonists. In the PYR-
(4,4,4) case without terminal protonation at physiologi-
cal pH, the curve is shifted right to a more agonistic
direction. It is unfortunate that we cannot speak to the
impact of PYR or PIP on the 5,4,5 system, but the parent
tetraamine was not available.
Ch a in Len gth . In designing these molecules a
BIOSYM molecular modeling program was employed.
The synthetic candidates were set in their potential
minimum, and the through-space distances between the
respective NR-Nâ nitrogens, the external nitrogen and
closest neighboring nitrogen, were measured along with
the NR-NR′ value the distance between the two external
nitrogens (Table 1).
These calculations were carried out on the basis of
the following assumptions about the charged state of
the tetraamines at physiological pH (Table 2). On the
basis of the pKa studies, all of the parent amines are
largely in the +4 charge state as are their ethylated
analogues. PIP(3,3,3) is assumed to be predominately
in the +3 cationic and PIP(3,4,3), PIP(4,4,4), and PIP-
(5,4,5) largely in the +4 cationic state. FDESPM and
the pyridyl compounds are in the +2 cationic state.
FDESPM is assumed to have a single charge on each
of the central nitrogens. The pyridine compounds PYR-
(3,3,3) and PYR(3,4,3) are assumed to have a single
charge on each of the terminal nitrogens.43 While there
was nothing unexpected about the results of the calcu-
lations, a few brief comments are in order. In each
family of tetraamines-norspermine, spermine, homo-
spermine, and tetraazaoctadecanesthe NR-NR′ dis-
tances consistently decreased in going from the ethy-
lated compounds through the piperidyl to pyridyl
analogues. The tetraamines varied in length from
12.617 Å for PYR(3,3,3) to 19.609 Å for PIP(5,4,5).
Finally DESPM and FDESPM are nearly identical in
terms of their NR-NR′ and NR-Nâ values.
There is indeed a correlation between tetraamine
length and impact on MK-801 binding. Looking at the
concentration required to reduce MK-801 binding by
50%, antagonistic IC50 (A IC50), homospermine is sig-
nificantly more effective than either norspermine or
spermine. Although PYR(3,3,3) and PYR(3,4,3) are
shorter molecules than PYR(4,4,4) and PYR(5,4,5), it is
critical to recall that when considering the overall length
of the molecule, the important distance is between
external cations. Both PYR(3,3,3) and PYR(3,4,3) are
protonated on the terminal pyridine nitrogens at physi-
ological pH, unlike PYR(4,4,4) and PYR(5,4,5), which
are protonated on the two internal nitrogens. Thus
from a length perspective PYR(3,3,3) and PYR(3,4,3) are
more like 1,11-diaminoundecane and 1,12-diaminodode-
cane,49,50 while PYR(4,4,4) and PYR(5,4,5) are both like
disubstituted putrescine. This thus explains the im-
pressive antagonistic behavior of PYR(3,3,3) and PYR-
(3,4,3) relative to the lack of antagonism with PYR(5,4,5)
and PYR(4,4,4). Thus it seems like these observations
are in keeping with earlier suggestions that the length
of the tetraamine is a major component contributing to
its antagonistic behavior.
Con clu sion
It seems clear that the effectiveness with which the
tetraamines described in the study antagonize MK-801
binding in the NMDA channel is related to their length
and charge state. For example, when comparing tet-
racations at physiological pH, the longer molecules are
the best antagonists. It is not enough to compare
tetraamines unless the charge distribution is the same.
Holding the geometry of the molecule essentially con-
stant but changing the pKa of the nitrogens and thus
the protonation state has a profound effect on the
tetraamine’s activity. For example, FDESM is virtually
identical to the potent tetraamine analogue, DESPM,
except that the pKa’s of the two external nitrogens are
changed from strongly to weakly basic so that FDESM
is charged only at the two internal nitrogens, resembling
putrescine, and like putrescine is completely devoid of
activity. In another example, PIP(3,3,3) has substantial
tetracationic character at physiological pH and behaves
as a short tetracation like norspermine, an agonist over
a wide concentration range. However, the correspond-
ing PYR(3,3,3) analogue is only charged at the terminal
pyridinium nitrogens, resembling a long-chain diamine
like 1,11-diaminoundecane. Not surprisingly, PYR-
(3,3,3) is a potent antagonist of MK-801 binding as is
the true diamine.
Exp er im en ta l Section
DENSPM, DESPM, FDESPM, DEHSPM, PIP(4,4,4), PYR-
(4,4,4), PIP(5,4,5), and PYR(5,4,5) were previously prepared
in these laboratories.21,38,40 NSPM, SPM, and other reagents
were purchased from Aldrich Chemical Co. Reactions using
hydride reagents were run in distilled DMF under a nitrogen
atmosphere. Fisher Optima grade solvents were routinely
used, and organic extracts were dried with sodium sulfate
unless otherwise indicated. Silica gel 32-63 (40 µM “flash”)
from Selecto, Inc. (Kennesaw, GA) was used for column
chromatography. Melting points were determined on a Fisher-
J ohns melting point apparatus and are uncorrected. Proton
The agonistic concentration span is also inversely
related to the length of the parent tetraamines. For
example norspermine is agonistic over the largest
concentration range, followed by spermine and homo-
spermine. The same inverse relationship holds for the

5264 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26
Bergeron et al.
1
NMR spectra were run at 300 MHz in CDCl3 (not indicated)
or D2O with chemical shifts given in parts per million down-
field from tetramethylsilane or 3-(trimethylsilyl)propionic-
2,2,3,3-d4 acid, sodium salt, respectively. Coupling constants
(J ) are in hertz. FAB mass spectra were run in a glycerol/
trifluoroacetic acid matrix. Elemental analyses were per-
formed by Atlantic Microlabs, Norcross, GA.
solid: mp 172-174 °C; H NMR δ 1.3-1.7 (m, 12 H), 2.23 (s,
6 H), 2.55 (s, 12 H), 3.07-3.18 (m, 8 H), 3.54 (t, 4 H), 6.89 (s,
4 H), 7.68-7.86 (m, 8 H). Anal. (C46H54N4O8S2) C, H, N.
N5,N10-Bis(m esitylen esu lfon yl)h om osp er m in e (3). Hy-
drazine monohydrate (1.5 mL, 31 mmol) was added to a
suspension of 2 (9.32 g, 10.9 mmol) in absolute EtOH (120 mL),
and the mixture was heated at reflux for 2 h. After being
cooled to room temperature, the mixture was filtered, and the
white solid was washed with EtOH (2×). The filtrate was
concentrated to furnish 3, which was used directly in the next
step. Anal. (C30H50N4O4S2) C, H, N.
Com p u ter Mod elin g.40 The distances between nitrogens
on polyamine analogues were taken from their most stable
conformations. The search for low-energy conformations of
polyamine analogues was done using BIOSYM program 2.3.0
running on a Silicon Graphics Indigo2 workstation, and the
results are similar to those using the conformational search
program supported by Tripos Associates Inc. in SYBYL 6.02.
P oten tiom etr ic Mea su r em en ts.40 The amine tetrahy-
drochloride (0.300 mmol) was dissolved in carbon dioxide free
0.1 M KCl (20.0 mL) and converted to the free base by addition
of a slight excess of 1 N NaOH (1.25 mequiv). This solution
was then immediately titrated with 1 N HCl using a Radiom-
eter-Copenhagen DTS833 digital titration system. The pKa’s
were estimated from a computer fit of the titration data using
MINSQ v. 2.0 nonlinear parameter estimation software
(Micromath Scientific Software, Salt Lake City, UT). The
relative abundance of each charged species present was
calculated by MINSQ in the simulation mode given the four
pKa values.
Hom osp er m in e Tetr a h yd r och lor id e (HSP M) (4). HBr
in HOAc (30%, 50 mL) was added to a solution of 3 (7.28 g,
12.2 mmol) and phenol (46 g, 0.49 mol) in CH2Cl2 (100 mL) at
0 °C. After the mixture was stirred at room temperature
overnight, water (100 mL) was added, and the layers were
separated. The organic layer was washed with water (50 mL),
and the combined aqueous portion was extracted with CH2-
Cl2 (3×). The aqueous portion was evaporated under high
vacuum, treated at 0 °C with 2 N NaOH (10 mL) and 50%
NaOH (25 mL), and extracted with CHCl3 (8 × 50 mL). After
solvent removal, the residue was taken up in EtOH (50 mL),
acidified with concentrated HCl (10 mL), and evaporated.
Recrystallization from aqueous EtOH gave 3.7 g of 4 (80%) as
white crystals: 1H NMR (D2O) δ 1.71-1.86 (m, 12 H), 3.01-
3.17 (m, 12 H). Anal. (C12H34Cl4N4) C, H, N.
Mem br a n e P r ep a r a tion a n d Bin d in g Assa y.6 The
procedure for measuring specific binding of [3H]MK-801 to the
N-methyl-D-aspartate receptor complex was a modification of
the method of Ransom and Stec.16 Cerebral cortices from
young male Sprague-Dawley rats (175-200 g) were homog-
enized with 10 volumes ice-cold 0.32 M sucrose using a motor-
driven glass/Teflon homogenizer. The homogenate was cen-
trifuged at 1000g for 10 min. The 1000g pellet was discarded,
and the supernatant was centrifuged at 18000g for 20 min.
After the supernatant was removed, the pellet was resus-
pended in 10 volumes of buffer A (5 mM Tris‚HCl, pH 7.7 at
4 °C) and centrifuged at 8000g for 20 min. The supernatant
and upper buffy coat of the pellet were combined and centri-
fuged at 50000g for 20 min. The pellet was then resuspended
in 10 volumes of buffer A and homogenized using high-
intensity ultrasound and centrifuged at 50000g, discarding the
supernatant. The pellet was washed in this manner an
additional three times and stored as a frozen suspension at
-80 °C for at least 18 h, but no longer than 2 weeks before
use.
For binding experiments, frozen membranes were thawed,
pelleted at 50000g for 20 min, and washed as described above,
except that 20 volumes of buffer A were used for resuspension,
for a total of five times. The final pellet was resuspended in
buffer B (5 mM Tris‚HCl, pH 7.5 at 23 °C). The binding assay
mixture was 1.00 mL of buffer B containing 200-300 µg of
membrane protein (Lowry method51), 100 µM L-glutamate, 100
µM glycine, 2 nM [3H]MK-801, and tetraamine at the following
concentrations: 0 (“100% + L-Glu, + Gly control”) and 0.05,
0.1, 0.5, 1, 5, 10, 50, 100, 500, and 1000 µM. Nonspecific
binding was determined using 100 µM MK-801.
Binding assays were performed in triplicate at 23 °C for 1
h and were terminated by filtration through Whatman GF/B
glass fiber filters followed by three 4.0 mL rinses of ice-cold
buffer B using a Brandel M-48 cell harvester. Each tetraamine
was tested on at least three different preparations of rat
cortical membranes, always with comparable results.
N,N′-Bis(4-ph th alim idobu tyl)-N,N′-bis(m esitylen esu lfo-
n yl)-1,4-bu ta n ed ia m in e (2). Sodium hydride (60%, 1.72 g,
43.0 mmol) was introduced into a solution of 138 (7.59 g, 16.8
mmol) in DMF (125 mL) at 0 °C. After the mixture was stirred
at room temperature for 35 min, N-(4-bromobutyl)phthalimide
(10.20 g, 36.2 mmol) was added. The mixture was stirred at
room temperature for 2 h and at 70 °C for 19 h and cooled,
and the reaction was quenched with EtOH. Following removal
of solvent under high vacuum, water was added, and product
was extracted with EtOAc (3×). The combined organic layers
were washed with water (2×) and brine, concentrated, and
recrystallized from EtOAc to give 9.43 g (66%) of 2 as a white
N,N′-Bis(m esitylen esu lfon yl)-4-a m in op ip er id in e (6). A
solution of mesitylenesulfonyl chloride (14.86 g, 67.9 mmol)
in CH2Cl2 (125 mL) was added to 541 (5.44 g, 34.2 mmol) in 1
N NaOH (150 mL) at 0 °C. After addition was complete, the
biphasic mixture was stirred overnight from 0 °C to room
temperature. Water (100 mL) and CHCl3 (200 mL) were
added, the layers were separated, and the aqueous portion was
extracted further with CHCl3 (2 × 50 mL). The combined
organic phase was washed with 100 mL of H2O, 1 N HCl, and
H2O and evaporated in vacuo. Recrystallization from CHCl3/
EtOAc afforded 11.56 g (73%) of 6 as a crystalline white solid:
mp 220-221 °C; NMR δ 1.35-1.83 (m, 4 H), 2.29 and 2.30 (2
s, 6 H), 2.56 and 2.62 (2 s, 12 H), 2.73-2.84 (m, 2 H), 3.17-
3.49 (m, 3 H), 4.58 (d, 1 H, J ) 8), 6.92 and 6.94 (2 s, 4 H).
Anal. (C23H32N2O4S2) C, H, N.
Tetr a k is(m esitylen esu lfon yl)-N1,N3-bis(4-p ip er id in yl)-
1,3-d ia m in op r op a n e (7). Sodium hydride (60%, 0.55 g, 14
mmol) was added to 6 (4.65 g, 10 mmol) in DMF (80 mL) at 0
°C. The suspension was stirred at room temperature until a
clear solution was obtained. 1,3-Dichloropropane (0.65 g, 5.75
mmol) was added, and the reaction mixture was heated at 85
°C overnight. Additional 60% NaH (0.2 g, 5 mmol) was added
with ice cooling, and then the mixture was stirred for 2 h,
followed by addition of more 1,3-dichloropropane (0.25 g, 2.2
mmol). The reaction mixture was heated at 85 °C for an
additional 5 h. Solvent was removed in vacuo, and the residue
was dissolved in CH2Cl2 (200 mL), which was washed with
H2O (4 × 50 mL) and dried over MgSO4. The solvent was
concentrated to about 50 mL and loaded onto a silica gel flash
chromatography column, with 30% EtOAc/hexane as eluant,
to provide 4.0 g (40%) of 7 as a white solid: mp 215 °C; 1H
NMR δ 1.30-1.65 (m, 4 H), 2.20-2.40 (m, 2 H), 2.40-2.80 (m,
28 H), 2.82-3.02 (m, 4 H), 3.40-3.66 (m, 6 H), 6.80-7.10 (m,
6 H). Anal. (C49H68N4O8S4) C, H, N.
N1,N3-Bis(4-p ip er id in yl)-1,3-d ia m in op r op a n e Tetr a h y-
d r och lor id e [P IP (3,3,3)] (9). HBr in acetic acid (30%, 75
mL) was added to a solution of 7 (3.82 g, 3.94 mmol) and
phenol (14.2 g, 0.15 mmol) in CH2Cl2 (70 mL) at 0 °C. The
reaction mixture was stirred overnight from 0 °C to room
temperature. Distilled water (120 mL) was added at 0 °C,
followed by extraction with CH2Cl2 (3 × 100 mL). The aqueous
layer was evaporated under high vacuum. The residue was
mixed with absolute ethanol (200 mL) to promote precipitation.
The solid was filtered and redissolved in water (10 mL), which
was basified with 1 N NaOH (12 mL) and 50% NaOH (20 mL)
with ice cooling. The solution was extracted with CHCl3 (10
× 50 mL). After CHCl3 extracts were evaporated, the residue
was taken up in ethanol (200 mL), acidified with concentrated
HCl to a pH of 1, and filtered. Recrystallization from aqueous
ethanol provided 1.26 g (83%) of 9 as a white solid: 1H NMR

Regulation of NMDA MK-801 Binding
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26 5265
(6) Bergeron, R. J .; Weimar, W. R.; Wu, Q.; Austin, J . K., J r.;
McManis, J . S. Impact of Polyamine Analogues on the NMDA
Receptor. J . Med. Chem. 1995, 38, 425-428.
(D2O) δ 1.66-2.00 (m, 4 H), 2.02-2.16 (m, 2 H), 2.40 (d, 4 H,
J ) 4.7), 3.12-3.30 (m, 8 H), 3.51-3.63 (m, 6 H). Anal.
(C13H32Cl4N4) C, H, N.
(7) J ohnson, J . W.; Ascher, P. Glycine Potentiates the NMDA
Response in Cultured Mouse Brain Neurons. Nature 1987, 325,
529-531.
Tetr a k is(m esitylen esu lfon yl)-N1,N4-bis(4-p ip er id in yl)-
1,4-d ia m in obu ta n e (8). Sodium hydride (80%, 0.93 g, 31
mmol) was added to 6 (11.24 g, 24.2 mmol) in DMF (145 mL)
at 0 °C. The suspension was stirred at room temperature for
45 min. 1,4-Diiodobutane (1.5 mL, 11 mmol) was added, and
the reaction mixture was heated at 70 °C for 18 h. Additional
80% NaH (0.35 g, 12 mmol) was added with ice cooling, and
then the mixture was stirred for 25 min, followed by addition
of more 1,4-diiodobutane (0.6 mL, 4.5 mmol). The reaction
mixture was heated as before, cooled, and quenched with
absolute EtOH (10 mL). Solvent was removed in vacuo, and
the residue was treated with water (100 mL) and extracted
with CHCl3 (3×). The combined organic extracts were washed
with 100 mL of 1% Na2SO3, H2O, and brine, and the solvent
was removed. Purification of crude product on a flash silica
gel chromatography column, with 30% EtOAc/hexane as
eluant, gave 4.59 g (39%) of 8 as a white solid: mp 206-207
°C; 1H NMR δ 1.23-1.34 (m, 4 H), 1.6-1.8 (m, 8 H), 2.30, 2.55,
and 2.59 (3 s, 36 H), 2.60-2.76 (m, 2 H), 2.97-3.06 (m, 4 H),
3.4-3.7 (m, 6 H), 6.95 (s, 8 H). Anal. (C50H70N4O8S4) C, H,
N.
(8) Reynolds, I. J .; Murphy, S. N.; Miller, R. J . [3H]-Labeled MK-
801 Binding to the Excitatory Amino Acid Receptor Complex
from Rat Brain is Enhanced by Glycine. Proc. Natl. Acad. Sci.
U.S.A. 1987, 84, 7744-7748.
(9) Peters, S.; Koh, J .; Choi, D. W. Zinc Selectively Blocks the Action
of N-Methyl-D-aspartate on Cortical Neurons. Science 1987, 236,
589-593.
(10) Westbrook, G. L.; Mayer, M. L. Micromolar Amounts of Zn+2
Antagonize NMDA and GABA Responses on Hippocampal
Neurons. Nature 1987, 328, 640-643.
(11) Mayer, M. L.; Westbrook, G. L.; Guthrie, P. B. Voltage-dependent
Block by Mg2+ of NMDA Responses in Spinal Cord Neurons.
Nature 1984, 309, 261-263.
(12) Nowak, L.; Bregestovski, P.; Ascher, P.; Herbert, A.; Prochiantz,
A. Magnesium Gates Glutamate-activated Channels in Mouse
Central Neurons. Nature 1984, 307, 462-465.
(13) Wong, E. H. F.; Kemp, J . A.; Priestly, T.; Knight, A. R.; Woodruff,
G. N.; Iversen, L. L. The Anticonvulsant MK-801 is a Potent
N-Methyl-D-aspartate Antagonist. Proc. Natl. Acad. Sci. U.S.A.
1986, 83, 7104-7108.
(14) Huettner, J . E.; Bean, B. P. Block of N-Methyl-D-aspartate-
activated Current by the Anticonvulsant MK-801: Selective
Binding to Open Channels. Proc. Natl. Acad. Sci. U.S.A. 1988,
85, 1307-1311.
(15) Fagg, G. E. Phencyclidine and Related Drugs Bind to the
Activated N-Methyl-D-aspartate Receptor-channel Complex in
Rat Brain Membranes. Neurosci. Lett. 1987, 6, 221-227.
(16) Ransom, R. W.; Stec, N. L. Cooperative Modulation of [3H]MK-
801 Binding to the N-Methyl-D-Aspartate Receptor Ion Channel
by Glutamate, Glycine, and Polyamines. J . Neurochem. 1988,
51, 830-836.
(17) Williams, K.; Romano, C.; Molinoff, P. B. Effects of Polyamines
on the Binding of [3H]MK-801 to the N-Methyl-D-aspartate
Receptor: Pharmacological Evidence for the Existence of a
Polyamine Recognition Site. Mol. Pharmacol. 1989, 36, 575-
581.
(18) Williams, K.; Dawson, V. L.; Romano, C.; Dichter, M. A.;
Molinoff, P. B. Characterization of Polyamines Having Agonist,
Antagonist, and Inverse Agonist Effects at the Polyamine
Recognition Site of the NMDA Receptor. Neuron 1990, 5, 199-
208.
(19) Benveniste, M.; Mayer, M. L. Multiple Effects of Spermine on
N-Methyl-D-aspartic Acid Receptor Responses of Rat Cultured
Hippocampal Neurones. J . Physiol. London 1993, 464, 131-163.
(20) Lynch, D. R.; Lawrence, J . J .; Lenz, S.; Anegawa. N. J .; Dichter,
M.; Pritchett, D. B. Pharmacological Characterization of Het-
erodimeric NMDA Receptors Composed of NR 1a and 2B
Subunits: Differences with Receptors Formed from NR 1a and
2A. J . Neurochem. 1995, 64, 1462-1468.
N1,N4-Bis(4-p ip er id in yl)-1,4-d ia m in obu ta n e Tetr a h y-
d r och lor id e [P IP (3,4,3)] (10). HBr in acetic acid (30%, 150
mL) was added to a solution of 8 (7.62 g, 7.75 mmol) and
phenol (28.81 g, 0.306 mol) in CH2Cl2 (125 mL) at 0 °C. The
reaction mixture was stirred for 1 day at room temperature
and worked up by the procedure of 4. Recrystallization from
aqueous ethanol provided 1.953 g (63%) of 10 as a white
solid: 1H NMR (D2O) δ 1.72-1.96 (m, 8 H), 2.34-2.45 (m, 4
H), 3.05-3.22 (m, 8 H), 3.45-3.66 (m, 6 H); HRMS calcd for
C14H31N4 255.2549 (free amine, M + H), found 255.2543 (M +
H). Anal. (C14H34Cl4N4‚H2O) C, H, N.
N1,N3-Bis(4-pyr id yl)-1,3-d ia m in op r opa n e Dih yd r och lo-
r id e [P YR(3,3,3)] (11). 1,3-Diaminopropane dihydrochloride
(3.16 g, 21.5 mmol) and 4-bromopyridine hydrochloride (10.91
g, 56.1 mmol) were heated at 229 °C for 3.5 h. The flask was
cooled in ice water, NaOH (1 N, 20 mL) and NaOH (50%, 10
mL) were added with ice cooling, and the mixture was
extracted with CHCl3 (150 mL, 2 × 50 mL, 8 × 25 mL). The
organic extracts were filtered, and the oil was purified by silica
gel flash column chromatography with 1.6% concentrated NH4-
OH/CH3OH as eluant. The free amine was dissolved in EtOH
and acidified with concentrated HCl (2 mL). Evaporation of
solvents and recrystallization from aqueous ethanol provided
1.12 g (17%) of 11 as a white solid: 1H NMR (D2O) δ 2.06
(quintet, 2 H, J ) 7), 3.49 (t, 4 H, J ) 7), 6.80-6.86 (m, 4 H),
7.94-8.01 (m, 4 H). Anal. (C13H18Cl2N4) C, H, N.
(21) Bergeron, R. J .; Neims, A. H.; McManis, J . S.; Hawthorne T. R.;
Vinson, J . R. T.; Bortell, R.; Ingeno, M. J . Synthetic Polyamine
Analogues as Antineoplastics. J . Med. Chem. 1988, 31, 1183-
1190.
(22) Porter, C. W.; Miller, J .; Bergeron, R. J . Aliphatic Chain Length
Specificity of the Polyamine Transport System in Ascites L1210
Leukemia Cells. Cancer Res. 1984, 44, 126-128.
(23) Porter, C. W.; Cavanaugh, P. F., J r.; Stolowich, N.; Ganis, B.;
Kelly, E.; Bergeron, R. J . Biological Properties of N4- and N1,N8-
Spermidine Derivatives in Cultured L1210 Leukemia Cells.
Cancer Res. 1985, 45, 2050-2057.
(24) Bergeron, R. J .; Hawthorne, T. R.; Vinson, J . R. T.; Beck, D. E.,
J r.; Ingeno, M. J . Role of the Methylene Backbone in the
Antiproliferative Activity of Polyamine Analogues on L1210
Cells. Cancer Res. 1989, 49, 2959-2964.
N1,N4-Bis(4-p yr id yl)-1,4-d ia m in obu ta n e Dih yd r och lo-
r id e [P YR(3,4,3)] (12). 1,4-Diaminobutane dihydrochloride
(4.10 g, 25.4 mmol) and 4-bromopyridine hydrochloride (10.02
g, 51.5 mmol) were heated at 226 °C for 4 h. The flask was
cooled, additional 4-bromopyridine hydrochloride (4.11 g, 21.1
mmol) was introduced, and the reactants were heated at 226
°C for 3 h. Isolation and purification by the method of 11 gave
1.46 g (18%) of 12 as a white solid: 1H NMR (D2O) δ 1.73-
1.85 (m, 4 H), 3.35-3.47 (m, 4 H), 6.78-6.85 (m, 4 H), 7.93-
8.02 (m, 4 H). Anal. (C14H20Cl2N4) C, H, N.
(25) Porter, C. W.; Pegg, A. E.; Ganis, B.; Madhubala, R.; Bergeron,
R. J . Combined Regulation of Ornithine and S-Adenosylme-
thionine Decarboxylases by Spermine and the Spermine Ana-
logue N1,N12-Bis(ethyl)spermine. Biochem. J . 1990, 268, 207-
212.
Ack n ow led gm en t. Financial support was provided
by SunPharm Corp., J acksonville, FL, and National
Institutes of Health, Grant No. CA37606.
(26) Pegg, A. E.; Madhubala, R.; Kameji, T.; Bergeron, R. J . Control
of Ornithine Decarboxylase Activity in R-Difluoromethylorni-
thine-Resistant L1210 Cells by Polyamines and Synthetic
Analogues. J . Biol. Chem. 1988, 263, 11008-11014.
(27) Porter, C. W.; McManis, J .; Casero, R. A.; Bergeron, R. J .
Relative Abilities of Bis(ethyl) Derivatives of Putrescine, Sper-
midine, and Spermine to Regulate Polyamine Biosynthesis and
Inhibit L1210 Leukemia Cell Growth. Cancer Res. 1987, 47,
2821-2825.
(28) Pegg, A. E.; Wechter, R.; Pakala, R.; Bergeron, R. J . Effect of
N1, N12-Bis(ethyl)spermine and Related Compounds on Growth
and Polyamine Acetylation, Content and Excretion in Human
Colon Tumor Cell. J . Biol. Chem. 1989, 264, 11744-11749.
Refer en ces
(1) Lipton, S. NMDA Receptors - Their Role in CNS Excitatory
Neurotransmission. Trends Neurosci. 1993, 16, 527-532.
(2) Collingridge, G. L.; Bliss, T. V. P. NMDA Receptors - Their Role
in Long-term Potentiation. Trends Neurosci. 1987, 10, 288-293.
(3) Brewer, G. J .; Cotman, C. W. NMDA Receptor Regulation of
Neuronal Morphology in Cultured Hippocampal Neurons. Neu-
rosci. Lett. 1989, 99, 268-273.
(4) Dichter, M. A.; Ayala, G. F. Cellular Mechanisms of Epilepsy:
A Status Report. Science 1987, 237, 157-164.
(5) Choi, D. W. Glutamate Neurotoxicity and Diseases of the
Nervous System. Neuron 1988, 1, 623-634.

5266 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26
Bergeron et al.
(29) Casero, R. A., J r.; Celano, P.; Ervin, S. J .; Porter, C. W.;
Bergeron, R. J .; Libby, P. R. Differential Induction of Spermidine/
Spermine N1-Acetyltransferase in Human Lung Cancer Cells by
the Bis(ethyl)polyamine Analogues. Cancer Res. 1989, 49, 3829-
3833.
(30) Libby, P. R.; Henderson, M.; Bergeron, R. J .; Porter, C. W. Major
Increases in Spermidine/Spermine-N1-Acetyltransferase Activity
by Spermine Analogues and Their Relationship to Polyamine
Depletion and Growth Inhibition in L1210 Cells. Cancer Res.
1989, 49, 6226-6231.
(31) Libby, P. R.; Bergeron, R. J .; Porter, C. W. Structure-Function
Correlations of Polyamine Analog-Induced Increases in Sper-
midine/Spermine Acetyltransferase Activity. Biochem. Pharma-
col. 1989, 38, 1435-1442.
(32) Porter, C. W.; Ganis, B.; Libby, P. R.; Bergeron, R. J . Correlations
Between Polyamine Analog-Induced Increases in Spermidine/
Spermine N1-Acetyltransferase Activity, Polyamine Pool Deple-
tion, and Growth Inhibition in Human Melanoma Cell Lines.
Cancer Res. 1991, 51, 3715-3720.
(33) Fogel-Petrovic, M.; Shappell, N. W.; Bergeron, R. J .; Porter, C.
W. Polyamine and Polyamine Analog Regulation of Spermidine/
Spermine N1-Acetyltransferase in MALME-3M Human Mela-
noma Cells. J . Biol. Chem. 1993, 268, 19118-19125.
(34) Shappell, N. W.; Fogel-Petrovic, M. F.; Porter, C. W. Regulation
of Spermidine/spermine N1-Acetyltransferase by Intracellular
Polyamine Pools-Evidence for a Functional Role in Polyamine
Homeostasis. FEBS Lett. 1993, 321, 179-183.
(35) Bergeron, R. J .; Weimar, W. R.; Luchetta, G.; Streiff, R. R.;
Wiegand, J .; Perrin, J .; Schreier, K. M.; Porter, C.; Yao, G. W.;
Dimova, H. Metabolism and Pharmacokinetics of N1,N11-Dieth-
ylnorspermine. Drug Metab. Dispos. 1995, 23, 1117-1125.
(36) Bergeron, R. J .; Weimar, W. R.; Luchetta, G.; Sninsky, C. A.;
Wiegand, J . Metabolism and Pharmacokinetics of N1,N14-Dieth-
ylhomospermine. Drug Metab. Dispos. 1995, 23, 334-343.
(37) Bergeron, R. J .; Yao, G. W.; Yao, H.; Weimar, W. R.; Sninsky,
C. A.; Raisler, B.; Feng, Y.; Wu, Q.; Gao, F. Metabolically
Programmed Polyamine Analogue Antidiarrheals. J . Med. Chem.
1996, 39, 2461-2471.
(38) Bergeron, R. J .; McManis, J . S.; Liu, C. Z.; Feng, Y.; Weimar,
W. R.; Luchetta, G. R.; Wu, Q.; Ortiz-Ocasio, J .; Vinson, J . R.
T.; Kramer, D.; Porter, C. Antiproliferative Properties of
Polyamine Analogues: A Structure-Activity Study. J . Med.
Chem. 1994, 37, 3464-3476.
(39) Samejima, K.; Takeda, Y.; Kawase, M.; Okada, M.; Kyogoku, Y.
Synthesis of 15N-Enriched Polyamines. Chem. Pharm. Bull.
1984, 32, 3428-3435.
(41) Shirahata, A.; Morohohi, T.; Fukai, M.; Akatsu, S.; Samejima,
K. Putrescine or Spermidine Binding Site of Aminopropyltrans-
ferases and Competitive Inhibitors. Biochem. Pharmacol. 1991,
41, 205-212.
(42) Bailey, D. M.; DeGrazia, C. G.; Hoff, S. J .; Schulenberg, P. L.;
O’Connor, J . R.; Paris, D. A.; Slee, A. M. Bispyridinamines: A
New Class of Topical Antimicrobial Agents as Inhibitors of
Dental Plaque. J . Med. Chem. 1984, 27, 1457-1464.
(43) Bellobono, I. R.; Favini, G. Ultraviolet Absorption and Protona-
tion Equilibria of Amino- and Nitro-substituted Pyridines. J .
Chem. Soc. B 1971, 2034-2037.
(44) Igarashi, K.; Williams, K. Antagonist Properties of Polyamines
and Bis(ethyl)polyamines at N-Methyl-D-aspartate Receptors.
J . Pharmacol. Exp. Ther. 1995, 272, 1101-1109.
(45) Durand, G. M.; Bennett, M. V.; Zukin, R. S. Splice Variants of
the N-Methyl-D-aspartate Receptor NR1 Identify Domains
Involved in Regulation by Polyamines and Protein Kinase C.
Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 6731-6735.
(46) Lynch, D. R.; Anegawa, N. J .; Verdoorn, T.; Pritchett, D. B.
N-Methyl-D-aspartate Receptors: Different Subunit Require-
ments for Binding of Glutamate Antagonists, Glycine Antago-
nists, and Channel-blocking Agents. Mol. Pharmacol. 1994, 45,
540-545.
(47) Williams, K.; Zappia, A. M.; Pritchett, D. B.; Shen, Y. M.;
Molinoff, P. B. Sensitivity of the N-Methyl-D-aspartate Receptor
to Polyamines is Controlled by NR2 Subunits. Mol. Pharmacol.
1994, 45, 803-809.
(48) Gallagher, M. J .; Huang, H.; Pritchett, D. B.; Lynch, D. R.
Interactions Between Ifenprodil and the NR2B Subunit of the
N-Methyl-D-aspartate Receptor. J . Biol. Chem. 1996, 271, 9603-
9611.
(49) Woodward, J . J .; Cueto, S. Effects of the Putative Polyamine
Antagonists Diethylenetriamine and 1,10-Diaminodecane on
N-Methyl-D-aspartic Acid-stimulated [3H]Norepinephrine Re-
lease from Rat Hippocampal Slices. J . Pharmacol. Exp. Ther.
1993, 266, 563-569.
(50) Subramaniam, S.; Donevan, S. D.; Rogawski, M. A. Hydrophobic
Interactions of n-Alkyl Diamines with the N-Methyl-D-aspartate
Receptor: Voltage-dependent and -independent Blocking Sites.
Mol. Pharmacol. 1994, 45, 117-124.
(51) Lowry, O. H.; Rosebrough, N. J .; Farr, A. L.; Randall, R. J .
Protein Measurement with the Folin Phenol Reagent. J . Biol.
Chem. 1951, 193, 265-275.
(40) Bergeron, R. J .; McManis, J . S.; Weimar, W. R.; Schreier, K. M.;
Gao, F.; Wu, Q.; Ortiz-Ocasio, J .; Luchetta, G. R.; Porter, C.;
Vinson, J . R. T. The Role of Charge in Polyamine Analogue
Recognition. J . Med. Chem. 1995, 38, 2278-2285.
J M960545X