Preparation and biological assessment of hydroxycinnamic acid amides of polyamines

Phytochemistry 63 (2003) 315–334

www.elsevier.com/locate/phytochem

Preparation and biological assessment of hydroxycinnamic acid

amides of polyamines

Solomon Fixon-Owoo^a, Frederic Levasseur^a, Keith Williams^b, Thomas N. Sabado^b,

Mike Lowe^c, Markus Klose^c, A. Joﬀre Mercier^c, Paul Fields^d, Jeﬀrey Atkinson^a,*

^aDepartment of Chemistry, Brock University, St.Catharines, Ontario, Canada L2S3A1

^bDepartment of Physiology and Pharmacology, SUNY Downstate Medical Center, Brooklyn, NY 11203-2098, USA

^cDepartment of Biological Sciences, Brock University, St. Catharines, Ontario, Canada L2S 3A1

^dCerealResearch Centre, Agricutlure and Agri-Food Canada, Winnipeg, Manitoba, Canada R3T 2M9

Received 17 December 2002; received in revised form 27 January 2003

Abstract

Many plants contain hydroxycinnamic acid conjugates of polyamines that are remarkably similar in general structure to the

acylated polyamines found in spider and wasp toxins. In an eﬀort to determine whether these compounds might play a role in the

chemical defense of plants against arthropod pests we synthesized a variety of analogues of the coumaric (4-hydroxycinnamic) acid

conjugates of di-, tri-, and tetraamines using common protection and acylation strategies. N¹- and N⁸-coumaroyl spermidine were

tested in feeding trials with insect larvae including the European corn borer (Ostrinia nubilalis), the tobacco budworm (Heliothis

verescens) and the oblique banded leaf roller (Choristoneura rosaceana). Antifeedant assays with the rice weevil Sitophilus oryzae

were also performed. Neither the naturally occurring coumaric acid conjugates of polyamines nor their analogues showed notable

toxicity towards insects, despite precautions to maintain these easily oxidized materials in the wet diet. However, more direct

bioassays of these compounds on glutamate dependent neuroreceptors including the deep abdominal extensor muscles of crayﬁsh,

or mammalian NMDA, d2, and AMPA receptors, clearly showed that these compounds were inhibitory. N¹-Coumaoryl spermine,

a dodecyl and a cyclohexyl analogue were especially active at NMDA NR1/NR2B receptors. The latter had an IC₅₀of 300 mM in

the crayﬁsh. N¹-Coumaroyl spermine had an IC₅₀in the crayﬁsh preparation of 70–300 mM and against the mammalian NR1/

NR2B receptor of 38 nM. Structure–activity variations show similar trends of length and hydrophobicity as has been seen pre-

viously with analogues of spider toxins. We conclude from this work that while the coumaric acid polyamine conjugates are active

when directly applied to neuroreceptors, they show no overt toxicity when ingested by insect larvae.

Keywords: Polyamines; Hydroxycinnamic acids; Neurotoxins; Glutamate receptor antagonist

1. Introduction

clear, it is interesting to note their similarity to other

polyamine conjugates found in the venoms of predac-

eous spiders and wasps (Blagbrough et al., 1994;

McCormick and Meinwald, 1993; Quicke and Usher-

wood, 1990; Schafer et al., 1994). The invertebrate

polyamine toxins are usually composed of an aromatic

acid, often derived from tryptophan or tyrosine, and a

long polyamine tail of varied structure. (Fig. 1) These

compounds comprise a portion of the low molecular

weight components in invertebrate venoms that inhibit

neurotransmission and presumably have evolved to

immobilize prey for capture. Some of these toxins are

active at glutamate regulated ion-channels and the

molecules have attracted a great deal of interest from

It has been known for more than 25 years that plants

can conjugate polyamines with various phenolic acids

creating a mixture of mono-, di- and triamides. Some of

the most common conjugates are those that contain the

hydroxycinnamic acids coumaric, ferulic and caﬀeic

acids (Smith et al., 1983). Although the roles of these

conjugates in normal plant physiology are far from

* Corresponding author. Tel.: +1-905-688-555x3967; fax: +1-905-

682-9020.

E-mailaddress: jatkin@brocku.ca (J. Atkinson).

URL: http://chemiris.labs.brocku.ca/ꢀchemweb/faculty/atkinson/

doi:10.1016/S0031-9422(03)00133-X

316

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

2. Results

2.1. Synthesis

The results of structure–activity studies using ana-

logues of the polyamine toxins from spiders have deter-

mined that the aromatic chromophore is an absolute

requirement for activity at the quisqualate class of glu-

tamate receptors (Blagbrough and Usherwood, 1990;

Bruce et al., 1990). With this in mind we are interested

in the mono-coumaroyl (4-hydroxycinnamoyl) con-

jugates of polyamines. Plants are known to produce a

complex mixture of mono-, di-, and tri-conjugates of the

more common polyamines with diﬀering acyl groups

and diﬀering numbers of hydroxyl and methoxyl groups

(Bokern et al., 1995; Meurer-Grimes, 1995; Werner et

al., 1995). For simplicity, we prepared only the mono-

coumaroyl conjugates in keeping with the common

occurrence of a phenolic group in the invertebrate toxins

(McCormick and Meinwald, 1993). With this portion of

the molecule remaining constant, we designed a select

number of analogues to determine whether the chain

length, degree of heteroatom substitution, or extra

hydrophobic substituents in the sidechain would have

any eﬀect on the biological activity assessed at crusta-

cean and mammalian glutamate receptors.

Many of the synthetic procedures that involve poly-

amines incorporate a chemoselective step at some point

and this has prompted the use of a large variety of pro-

tecting group strategies and reagents (Karigiannis and

Papaioannou, 2000). Our requirement of selectively

acylating di- and polyamines was fairly easily met by the

mono-triﬂuoroacetylation of diamines and the selective

mono-triﬂuoracetylation of primary over secondary

amines (Blagbrough and Geall, 1998; Osullivan and

Dalrymple, 1995). Many diamines can be mono-Boc

protected using (Boc)₂O in dioxane (Krapcho and

Kuell, 1990) and we sometimes use this method as in the

preparation of 6 (Scheme 1). The success of this method

relies on the spontaneous precipitation of the di-Boc

amine product and the easy removal of remaining excess

diamine, usually by distillation. When the diamines are

not available in such large excess, or have high boiling

points, then retrieval of the mono-Boc protected amine

is much less eﬃcient and it is better to ﬁrst mono-

triﬂuoroacetylate, carbamoylate the remaining amine

group, then remove the triﬂuoroacetyl protecting group.

For example, in the case of 6, the diamine was readily

available and it could be easily removed by washing the

crude reaction with water and extracting the mono-

protected material (Scheme 1).

Fig. 1. Comparison of acylated polyamine toxins from invertebrates

(PhTX-433 and JSTX-3) with common plant derived coumaric acid

conjugates of spermidine and spermine.

neurophysiologists (Blagbrough et al., 1994; Moya and

Blagbrough, 1995; Quicke and Usherwood, 1990; Scha-

fer et al., 1994; Usherwood and Blagbrough, 1991).

The similarity of the plant derived hydroxycinnamic

acid (i.e coumaric) amides and the invertebrate poly-

amine toxins has been noted before by Blagbrough

(Blagbrough and Usherwood, 1990, 1992) who prepared

and tested N¹-coumaroyl spermine and analogues

against a locust nerve-muscle preparation. We became

interested in this family of compounds after recognizing

that plant protection against insect herbivory can at

least in part be described by the quantity and type of

phenolic compounds present in tissue (Duﬀey and

Stout, 1996). Many chemical analyses of the total phenolic

content of plants, however, liberate the free phenolics

from conjugated forms by hydrolysis with strong acids

and bases (Sharma and Rajam, 1995; Torrigiani et al.,

1987). A great deal of information concerning the

chemical identity of the phenolic compounds is lost

during such a procedure. We have devised a method for

extraction and analysis of hydroxycinnamic acid amides

from plant tissue (Panagabko et al., 2000) and have

shown that N⁸- and N¹-coumaroyl spermidine, shorter

polyamine conjugates than N¹-coumaroyl spermine, are

active at the glutamate dependent neuromuscular junc-

tion of crustaceans (Klose et al., 2002; Mercier et al.,

1998). We have prepared a number of aryl substituted

cinnamic acid amides by methods essentially identical to

those of Hu and Hesse (1996) and in the present study

we expand on this family by preparing coumaric acid

monoamides with a variety of di-, tri- and tetraamines.

Protection of the triamines 11 and 12 could be

achieved directly with (Boc)₂O, whereas treatment of

these amines with ethyl triﬂuoroacetate consistently

provided diacylated materials as the major product

(Scheme 2). Acylation of 14 with 28 (Scheme 3) provided

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

317

the desired 34 in low yields (16%), and also gave a sig-

niﬁcant amount of the 1,3-dicoumaroyl conjugate 34c

(30%). Very little of such diacylation (8%) was

observed to occur with 15.

The tetraamine 19 is commercially available (Scheme 3)

and 23 and 26 could be prepared by dicyanoethylation of

the diaminocyclohexanes 20 and 13. During one attempt

of a reduction of 25 on a fairly high scale (19 g) the

reduction was quite slow, presumably due to an insuﬃ-

cient amount of Raney nickel catalyst and a signifcant by

product was observed and isolated. This material was

identiﬁed to be the half-hydrolyzed primary amide 26a.

Scheme 1. Monoprotection of selected diamines.

Scheme 2. Monoprotection of selected triamines.

Scheme 3. Preparation and derivatization of tetraamines.

318

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

Acylation of the protected amines with (E)-3-(4-ace-

could be acylated directly with 28, and deprotected

under acidic conditions (Schemes 4–6).

toxyphenyl)prop-2-enoyl chloride 28 (Scheme 3) was

straightforward although recovered yields are variable

depending on the amine. While the mono-triﬂuoro-

acetates themselves can be acylated with 28, removal of

the triﬂuoroacetyl group requires treatment with strong

base and we have found this to reduce yields and

quality of ﬁnal phenolic products. It was thus straight-

forward to prepare mono-protected diamines that

The ﬁnal deprotected coumaric acid conjugates were

all hygroscopic solids that showed a tendency to iso-

merize at the coumaroyl double bond if exposed to

room light for even short periods of time. Careful

handling of the hydrochloride salts is required to

obtain accurate weights for feeding trials and stock

solutions.

Scheme 4. Acylation of protection diamines.

Scheme 5. Acylation of selected triamines.

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

319

Scheme 6. Acylation of selected tetraamines.

2.2. Biological activity assessment

2.3. Mammalian glutamate receptors

Glutamate is the neurotransmitter at most neuro-

muscular synapses in crayﬁsh (Atwood, 1982) and gluta-

mate receptor antagonists block excitatory post-synaptic

potentials (EPSPs) in the crayﬁsh muscles used in this

investigation (Mercier et al., 1998). We have previously

reported the inhibitory activity of N⁸-coumaroyl spermi-

dine (Mercier et al., 1998) and N¹-coumaroyl spermidine

(Klose et al., 2002) at the neuromuscular junction of the

deep abdominal extensor muscles of the crayﬁsh Pro-

cambarus clarkii. The naturally occurring conjugates N¹-

and N⁸-coumaroyl spermidine had IC₅₀values of 80 and

50–150 mM respectively. N¹-coumaroyl spermine was less

potent having an IC₅₀of 70–300 mM. A synthetic analo-

gue 43b (Scheme 6) that was designed as an analogue of

N¹-coumaroyl spermine with a more bulky and hydro-

phobic polyamine chain, was tested on this same cray-

ﬁsh preparation. EPSP amplitudes were recorded for 5

min before toxin application and for 10 min after the

toxin was ﬁrst introduced. Introduction of the toxin to

the bathing solution reduced EPSP amplitude within

approximately 2–3 min (depending on the concen-

tration), and the eﬀect reversed rapidly when the toxin

was washed out with saline. Control trials were per-

formed by recording EPSPs for the same time but

without applying the toxin. The response to the toxin

was estimated in each trial by comparing the amplitude

of the EPSP at 10 min of toxin exposure to the average

EPSP amplitude of seven control trials at the corre-

sponding time. The response in each trial was calculated

as a percentage of the control value, and data from

individual trials were used to construct a dose–response

curve (Fig. 2). The IC₅₀value for 43b was estimated at

approximately 300 mM.

Naturally occurring and synthetic conjugates were

also tested against recombinant mammalian glutamate

receptors expressed in Xenopus oocytes. We studied

three classes of mammalian receptors—NMDA,

AMPA, and d receptors. Native NMDA and AMPA

receptors are gated by glutamate, and a family of sub-

units that can assemble to form these receptors has been

Fig. 2. Dose–response curve for the reduction of excitatory post-

synaptic potentials at a crayﬁsh neuromuscular preparation by 43b.

Non-linear regression analysis was performed to obtain a curve of best

ﬁt for the data and produced a sigmoidal curve.

320

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

cloned (Dingledine et al., 1999; Hollmann and Heine-

mann, 1994). Two ‘‘orphan’’ subunits, termed d1 and

d2, which have homology to NMDA and AMPA

receptor subunits have been identiﬁed, although their

function is still unknown as they are not directly gated

by glutamate. A mutation in the d2 subunit creates a

constitutively open cation channel, and this mutation is

responsible for the neurological phenotype of the

Lurcher mouse (Zuo et al., 1997). Recent studies have

suggested that the pore of d2 channels may be similar to

that of AMPA channels (in particular those containing

Q at the Q/R editing site), but there is as yet no detailed

pharmacology of the d2 channels (Wollmuth et al.,

2000). Studies with a number of traditional glutamate

receptor blockers suggest that the pore has properties in

common with both AMPA and NMDA channels (Wil-

liams et al., 2003). For the current studies, we used het-

eromeric NMDA receptors expressed from the NR1a

and NR2B subunits (NR1/NR2B receptors), homomeric

AMPA receptors expressed from the GluR2(Q) subunit,

which contains Q at the Q/R editing site in the M2 loop

region (Hume et al., 1991), heteromeric GluR1/

GluR2(Q) receptors, and d2 receptors expressed from the

constitutively active d2(A654T) mutant (Zuo et al., 1997).

We screened the activities of thirteen compounds at a

concentration of 10 mM at NMDA, AMPA and d2

Table 1

Eﬀects of coumaroyl–amine conjugates on NMDA, AMPA, and d2 channels

Compound

Number

NMDA

NR1/NR2B

% Inhibition

d2

AMPA

GluR2(Q)

% Inhibition

AMPA

GluR1/GluR2(Q)

% Inhibition

d2(A654T)

% Inhibition

29b

8

11

98

8

1

4

0

1

30b

31b

32b

33b

34b

22

1

13

2

11

2

3

0

2

1

12

1

6

34d

59

ꢁ19

10

35b

36b

37b

37

26

12

ꢁ17

4

11

18

8

11

16

9

4

40b

28

0

15

13

43b

95

43

14

52

84

50

78

–

100

N¹-coumaroyl spermine

The percentage inhibition by 10 mM of each compound is shown. Values are mean from 2–4 oocytes. In some cases, the compound produced a small

potentiation of the current (indicated by a negative value).

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

321

receptors. The results are shown in Table 1. Many

compounds had little or no eﬀect on glutamate channels,

but 34d inhibited NMDA receptors by about 60% and

compounds 31b, 43b and N¹-coumaroyl spermine

inhibited NMDA receptors by 95–100%. Compounds

31b and 43b also inhibited AMPA and d2 channels, but

most showed some selectivity for NMDA receptors,

producing a larger inhibition at NMDA than at AMPA

or d2 channels. N¹-Coumaroyl spermine was a potent

blocker of both NMDA and AMPA channels but not of

d2 channels.

jugates by ion-pair HPLC (Panagabko et al., 2000). We

suspected that peroxidase activity in the grain-base diets

was responsible for a large portion of this loss (Con-

verso and Fernandez, 1995) and indeed, peroxidase

activity could be easily assayed (Makinen and Tenovuo,

1982) (data not shown). This loss could be gre_ꢂatly atten-

uated by heating the dry diet material at 120 C for 1 h

prior to making up the wet diet. With this treatment

about half of the conjugate concentration could be

maintained after one day. For full growth studies this

necessitated replacing the diet each day. Even after these

precautions, there was no discernable diﬀerence between

the growth rate, ﬁfth instar larval mass, nor mortality

when compared to control insects raised on diet simi-

larly heat treated but not containing any spermidine

conjugate. Similar experiments were also performed on

tobacco budworm (Heliothis verescens) and the oblique

banded leaf roller (Choristoneura rosaceana) with the

same results. Further trials with adults of the rice weevil

(Sitophilus oryzae) using wheat ﬂour wafers (Xie et al.,

1996) impregnated with 36b, 37b, 40b, 43b, N¹-feruloyl

spermidine, N¹- and N⁸-coumaroyl spermidine at 1.6

and 3.2 mg/200 mg wheat ﬂour also showed no mortality

after 7, 14 or 21 days. However, 43b at the highest dose

tested did cause 50% mortality after 21 days. Further-

more, 3-day tests showed that the conjugates, even 43b,

had no antifeedant (Xie et al., 1996) properties.

The

concentration–inhibition

relationship

for

N¹-coumaroyl spermine at NR1/NR2B NMDA recep-

tors was also studied. N¹-Coumaroyl spermine inhibited

NMDA receptors with an IC₅₀of 38 nM (Fig. 3). The

blocking eﬀects of polyamines are voltage-dependent,

but the potency of N¹-coumaroyl spermine at NR1/

NR2B receptors at ꢁ70 mV is similar to that of the

spider toxins argiotoxin₆₃₆and Agel-505 at NMDA

receptors (Williams, 1993a) and about 10-fold higher

than the polyamine derivatives N¹,N⁴,N⁸-tribenzyl-

spermidine and N¹-dansyl-spermine (Chao et al., 1997;

Igarashi et al., 1997).

2.4. Insect feeding trials

N¹- and N⁸-Coumaroyl spermidine have also been tes-

ted against insects in feeding assays using semi-synthetic

diets doped with varying amounts of the conjugates.

European corn borer (ECB, Ostrinia nubilalis Hubner)

3. Discussion

¨

were raised on a commercial corn meal base diet. Lar-

vae were selected at second instar based on weight and

reared on diet that contained 10–500 mg conjugate per

gram of wet diet. Initial control experiments showed

that the hydroxycinnamic acid polyamine conjugates

did not remain in the diet for more than a day as fol-

lowed by extracting the diet and analyzing for con-

The conjugates of hydroxycinnamic acids with poly-

amines have been found in a wide variety of plants and

have been proposed to have a number of biological

activities including anti-viral to antifungal eﬀects. The

similarity of these plant derived compounds with the

polyamine toxins from invertebrates has been noted

previously (Blagbrough et al., 1992; Blagbrough and

Usherwood, 1990), but little evidence has been given as

to their relevance in chemical ecology. We were inter-

ested ﬁrst in whether hydroxycinnamic acid–polyamine

conjugates such as N¹-coumaroyl spermidine were toxic

to larvae of certain Lepidopteran crop pests. Since we

had signiﬁcant quantities of some analogues on hand

from synthesis, feeding trials were attempted using lar-

vae raised on semi-synthetic diets. Despite much eﬀort

in maintaining these readily oxidized compounds in the

diets, the results of these trials, and those done with the

rice weevil Sytophilus oryzae, showed a near lack of

toxicity except for 43b at higher doses and long expo-

sure. Either these compounds are not toxic to the

organisms tested, or they do not manifest such toxicity

with insects reared on synthetic diets in the laboratory.

This latter caveat is mentioned since there are reliable

indications that such feeding trials miss the complexity

of whole plant tissues and the milieu of enzymes present

Fig. 3. Eﬀects of N¹-coumaroyl spermine on NMDA receptors. The

eﬀects of N¹-coumaroyl spermine were studied on NMDA receptors

expressed in Xenopus oocytes from NR1/NR2B subunits. Receptors

were activated by 10 mM glutamate+10 mM glycine, and oocytes were

voltage-clamped at ꢁ70 mV.

322

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

as an insect feeds (Duﬀey and Stout, 1996). However,

these eﬀects do not usually mask the presence of potent

toxins in the diet and we conclude that these conjugates,

when ingested, are not toxic to the pests.

longer chain spermine conjugates would have been

expected to be more potent given their greater simi-

larity to the spider toxins. A similar assay with 43b

gave an IC₅₀of 300 mM, approximately 4 times that

of N¹-coumaroyl spermine. Both N¹-coumaroyl sper-

mine and 43b are much less eﬀective on the crusta-

cean receptor than the mammalian glutamate

receptors (see below).

The inability of these coumaric acid–polyamine con-

jugates to arrest larval feeding may be due in part to the

fact that they are relatively poor blockers of the gluta-

mate receptors found at arthropod neuromuscular

junctions. IC₅₀values for N⁸-coumaroyl spermidine

(Mercier et al., 1998), N¹-coumaroyl spermidine (Klose

et al., 2002) and 43b (present results) range from 70 to

300 mM. The IC₅₀values for eﬀect of Joro spider toxin

on crustacean neuromuscular synapses is approximately

0.02–0.1 mM (Abe et al., 1983; Aonuma et al., 1998;

Shudo et al., 1987). Aside from such diﬀerences in

potency, it is not known how eﬀectively the toxins are

absorbed through the digestive system, if at all, and this

could be an important reason for the lack of larval

growth inhibition. Spider and wasp toxins, by contrast,

are injected directly into the hemolymph and, thus, elicit

paralysis very rapidly.

Although the hydroxycinnamic acid conjugates that

we studied are relatively poor blockers of arthropod

glutamate receptors, they are eﬀective in blocking

mammalian glutamate receptors. The IC₅₀for N¹-cou-

maroyl spermine was 38 nM at NMDA receptors. The

same compound blocks arthropod glutamate receptors

with an IC₅₀value of about 80 mM, indicating a diﬀer-

ence in potency of more than three orders of magnitude.

Some of this diﬀerence is probably due to the fact that

arthropod glutamate receptors are more closely related

to the AMPA-type, being activated by quisqualate (Abe

et al., 1983; Shinozaki and Shibuya, 1974; Usherwood

and Blagbrough, 1991).

The percent inhibition of mammalian glutamate

receptors by 10 mM of the coumaroyl polyamines is

shown in Table 1. All compounds show selectivity for

the NMDA NR1/NR2B receptor but range from 3 to

100% in the degree of inhibition. Compounds with the

largest separation between the coumaroyl amide nitro-

gen and the terminal (primary) amine tended to have

higher activity as has been noted for the biological

activity against mammalian nicotinic acetylcholine

receptors (Li et al., 2001). Thus compounds 31b, 43b,

and N¹-coumaroyl spermine all inhibited NR1/NR2B

receptors >95% at 10 mM, and were also active at the

AMPA and d receptors. The activity of the dodecyldi-

amine analogue 31b is remarkable given that it has no

internal nitrogens. Counterpoint this with the near

complete inactivity of the oxa-analogues 32b and 33b.

Philanthotoxin analogues that contain oxa-amine side-

chains retain their activity against acetylcholine recep-

tors (nAChR), but are inactive against non-NMDA

receptors expressed in oocytes, although the structure–

activity descriptions in this case are complicated by a

dependency on the acyl group attached to the a-nitro-

gen of the tyrosine moiety of the toxin (Strømgaard et

al., 2000). The high inhibitory activity of 31b suggests

that the terminal amine is more important than the

internal ones and may also reﬂect the role that hydro-

phobic forces have to play in polyamine binding (Cu et

al., 1998) although a similar analogue of philanthotoxin

showed no inhibition of the glutamatergic neuro-

muscular synapse in larvae of the house ﬂy, Musca

domestica (Benson et al., 1992).

A model of mammalian glutamate receptor channels

based on the structure of the potassium channel KscA

from Streptomyces lividans has been proposed. The

model is supported by the results of studies on muta-

tions within the glutamate channel that aﬀect block by

polyamines (Panchenko et al., 2001). The model

includes a glutamate residue in a solvent accessible loop

on the cytoplasmic side of the channel that is required

for block by intracellular polyamines. Some of the long-

chain polyamines such as 31b may bind in such as way

that their polyamine tail passes through the channel to

interact at the intracellular site, but the head group

remains lodged in the channel unable to pass through

the narrow selectivity ﬁlter about halfway through the

channel. Consistent with this are previous observations

that some unsubstituted polyamines or analogs with

small head groups can permeate the channel when

3.1. Structure–activity

The nature of the assay with crayﬁsh meant that only

a few compounds could easily be assessed. This, cou-

pled with the relatively small number of compounds

made in this study, cannot support a complete analysis

of the variation of biological activity with chemical

structure, but a few salient points can be made. The

IC₅₀s of N¹- and N⁸-coumaroyl spermidine for reduc-

tion of crustacean glutamate receptor excitatory post-

synaptic potentials (EPSPs) are 70 and 200 mM

respectively (Klose et al., 2002). These short chain

amine conjugates are much less active than the longer

chain Joro spider toxin JSTX-3 which has an IC₅₀in

the crustacean preparations of 0.02–0.1 mM. Higher

potencies with longer chain acylpolyamine toxins has

been noted previously for philanthotoxins (Anis et al.,

1990; Benson et al., 1993; Bruce et al., 1990). Treating

nerve preparations with 50 mM of N¹-coumaroyl sper-

midine or N¹-coumaroyl spermine gave EPSP reduc-

tions of near equal magnitude. This was unusual, as the

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

323

applied extracellularly, whereas analogs with larger

headgroups are unable to easily permeate the channel

(Chao et al., 1997; Igarashi et al., 1995, 1997).

muscle preparations. However, since these compounds

would necessarily be ingested by herbivorous insects, we

also tested for toxicity towards a variety of lepidopteran

larvae using semi-synthetic diets laced with N¹- and

N⁸-coumaroyl spermidine. There was no evidence of

mortality, growth reduction or, in the case of the rice

weevil Sitophilus oryzae, any activity as an antifeedant.

Since the amount of acylpolyamine compound required

for these insect feeding trials is large, due to oxidative

decomposition and the necessity of replacing diet every

day over a 10 day feeding trial, not all of the com-

pounds could be tested. Only the shorter chain spermi-

dine conjugates were tested in this manner and their

lower inhibitory potency may have meant that little

observable biological activity was to be witnessed.

Alternatively, there is a great deal of diﬀerence between

in vitro cell based assays, where toxins can be readily

applied close to the site of action, and whole animal

trials where multiple barriers exist between ingestion

and a binding event at a remote tissue. In eﬀect, the

highly charged nature of these compounds may make

them completely inactive as ingested toxins, requiring

them to be injected into tissues, something a spider

already practices.

If hydrophobic forces play a key role in polyamines

block, the dodecylamine analogue 31b may take advan-

tage of hydrophobic forces during entry into the channel

pore, whereas the oxa-analogues may be too polar for

favorable binding. Polyamines that have inappropriate

distances between nitrogen atoms in the chain, i.e. hav-

ing too few or too many methylene groups between

cationic sites to ﬁnd matching cluster of tryptophans in

the channel, would also bind poorly. This may explain

the lack of activity of 34b and 37b. Similar results have

been seen with non-acylated polyamines that bind to

NMDA receptors (Subramaniam et al., 1994) where

inhibitory potency was correlated with carbon chain

length and length of methylene spacer between positive

charges. Compounds 35b and 36b have some residual

activity, but are clearly too short.

The requirement for mammalian receptor binding and

inhibition would thus appear to be a 12-atom separa-

tion between the amide nitrogen and the terminal

amine. This is equivalent to a distance of about 16–17 A

(amide N-atom to terminal N-H) in a fully extended

conformation for N¹-couamroyl spermine and 43b. This

does not necessarily mean that this is the conformation

on binding, although NMR studies at varying pH sup-

port a near linear conformation of argiotoxin₆₃₆

(Raditsch et al., 1996) and Monte-Carlo minimization

methods show that the vast majority of low energy

structures populate linear polyamine conformations

(Tikhonov et al., 2000). The lower activity of 40b with

respect to 43b against the NMDA receptor may arise

because of the lack of conformational freedom in the

side chain of 40b due to the 1,2-disposition of the

cyclohexyldiamine moiety in the side chain.

The expectation that the acylpolyamines prepared in

this work are similar enough to spider and wasp toxins

to be active in receptor based assays was clearly met as

evidenced by the strong activities at mammalian glutamate

receptors, especially the NR1/NR2B NMDA receptors.

The structure–activity description largely parallels that

seen with philanthotoxin, argiotoxin, and non-acylated

polyamine analogues at various receptor preparations.

5. Experimental procedures

Curiously, the diacylated compound 34d retains some

activity (although it is stimulatory at d2-receptors). If

this compound binds in a fashion analogous to the

other compounds then there must be enough space in

the channel for the second bulky group to bind. This

may be reasonable to conclude given the ability of

NMDA receptors to bind tribenzylspermidine (Igarashi

et al., 1997; Kashiwagi et al., 2002).

5.1. Physiological studies

Freshwater crayﬁsh (Procambarus clarkii) were

obtained from Atchafalaya Biological Supply Co., Inc.

(Raceland, LA). Crayﬁsh size were 1–3 inches in

length. The crayﬁsh were fed a diet of ‘‘Tender Vittle’’

catfood and kept in circulating, ﬁltered freshwater

tanks. Physiological eﬀects were examined using deep

abdominal extensor muscles of crayﬁsh, as described

by Mercier and Atwood (1989). Brieﬂy, the dorsal

abdominal shell was dissected away with the extensor

muscles intact and was pinned out in a dish containing a

physiological saline based on that of van Harreveld

(1936), but with reduced Ca₂₊and elevated Mg₂₊levels

to prevent muscle twitching. This saline solution con-

tained 205 mM NaCl, 5.3 mM KCl, 6.25 mM

CaCl₂.2H₂O, 12.5 mM MgCl₂, and 5 mM HEPES, and

the pH was adjusted to 7.4 with NaOH. Excitatory

postsynaptic potentials (EPSPs) in the muscle were

elicited by applying electrical stimuli to the nerve using

4. Conclusions

The original hypothesis for this work was that the

plant hydroxycinnamoyl conjugates of polyamines such

as N¹- and N⁸-coumaroyl spermidine and N¹-coumaroyl

spermine may mimic the arthropod toxins and be in

part responsible for the chemical protection of plants.

We have shown that indeed these compounds are active

inhibitors of glutamatergic receptors in crustaceans as

evidenced by the reduction in EPSPs in crayﬁsh extensor

324

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

a suction electrode. EPSPs were recorded from muscle

L1 (Parnas and Atwood, 1966) using intracellular

microelectrodes ﬁlled with 3 M KCl. The nerve was sti-

mulated at a frequency of 0.2 Hz, and responses were

acquired and averaged using a 386-compatible micro-

computer and software designed speciﬁcally for this

purpose (Technical Services Division, Brock University,

St. Catharines, ON). The recording dish had a volume

of approximately 300 ml and was perfused continuously

at a rate of 1 ml/min throughout each experiment. The

toxin was applied by changing the perfusate from the

standard physiological saline to one containing the

toxin. All experiments were performed at room tem-

5.3. Syntheses

N¹-Coumaroyl spermine was prepared following the

published method (Blagbrough and Geall, 1998) as were

N¹- and N⁸-coumaroyl spermidines (Hu and Hesse, 1996).

5.4. Reagents and solvents

Reagent grade solvents were used for all extraction

and work-up procedures without further puriﬁcation.

THF was dried over sodium metal in the presence of

benzophenone. Dichloromethane was distilled from

calcium hydride. Methanol was distilled from magne-

sium turnings and a catalytic amount of iodine.

ꢂ

perature (21 C).

5.2. Recombinant mammalian glutamate receptors

5.5. Generalmethods

Mammalian glutamate receptors were studied using

voltage-clamp recording of recombinant receptors

expressed in Xenopus oocytes. Plasmids with clones

encoding glutamate receptor subunits were linearized at

a site 3⁰to the insert using an appropriate restriction

endonuclease. cRNAs were synthesized from cDNAs

using the mMessage mMachine kit (Ambion, Austin,

TX), and oocytes were injected with the appropriate

cRNA or mixture of cRNAs (0.2–10 ng per oocyte).

The preparation and maintenance of oocytes was car-

ried out as described previously (Williams, 1993b; Wil-

liams et al., 1993). Macroscopic currents were recorded

with a two-electrode voltage-clamp using glass micro-

electrodes ﬁlled with 3 M KCl and having resistances of

0.5–3 Mꢀ (Williams, 1993b). Oocytes were con-

tinuously superfused with a saline solution (96 mM

NaCl, 2 mM KCl, 1.8 mM BaCl₂, 10 mM HEPES, pH

7.5) that contained BaCl₂rather than CaCl₂to minimize

Ca²⁺-activated Cl^ꢁcurrents, and in experiments with

NMDA receptors, oocytes were injected with

K⁺-BAPTA (100 nl of 40 mM, pH 7.0–7.4) on the day

of recording. Oocytes were voltage-clamped at ꢁ70 mV.

NMDA receptors were activated by 10 mM glutamate

plus 10 mM glycine. AMPA receptors were activated by

100 mM kainate. To study currents through d2(A654T)

channels, we determined the diﬀerence in the holding

current measured in an extracellular solution containing

Na⁺(composition as above) and one in which the Na⁺

was replaced by an equimolar concentration of the large

organic cation N-methyl-d-glucamine (NMDG). The

constitutively open d2(A654T) channels are permeable

to Na⁺but not to NMDG (Wollmuth et al., 2000; Zuo

et al., 1997). The d2(A654T) channels produced macro-

scopic currents (deﬁned as the diﬀerence in holding

current between Na⁺and NMDG) of several hundred

nA. In control experiments, uninjected oocytes, or

oocytes injected with the wild-type (inactive) d2 clone

had a diﬀerence of 5–15 nA in the Na⁺versus NMDG

holding current.

Unless otherwise stated, all starting materials were

obtained from commercial suppliers and were used with-

out further puriﬁcation. Chromatography was carried out

on Aldrich silica gel (230–400 mesh) with the indicated

solvent systems. Analytical TLC was performed on 0.25

mm, precoated silica gel plates (EM Science, Silica Gel 60

F-254). Visualization was achieved using a UV lamp at

254 nm or exposure to iodine vapor, or by immersion in a

solution of ninhydrin in ethanol (0.2 g in 100 ml) followed

by heating. Melting points were determined on a Koﬂer

hot stage apparatus and are uncorrected. Low resolution

mass spectra were obtained by electron impact (EI) at 70

eV or fast atom bombardment (FAB) using m-nitrobenzyl

alcohol (NBA) as the matrix. ¹H NMR were recorded at

300 MHz and ¹³C NMR at 75.6 MHz with deuterated-

chloroform as the solvent unless otherwise noted. The

chemical shifts, ꢀ, for the ¹H were recorded in ppm rela-

tive to CDCl₃(ꢀ=7.26 ppm) and that for ¹³C was relative

to CDCl₃(ꢀ=77.0 ppm) or DMSO (ꢀ=39.5 ppm).

5.6. Generalprocedure for triﬂuoracetyal tion of di- and

polyamines

Di- and polyamines were mono-triﬂuoracetylated fol-

lowing the published method (Osullivan and Dalrymple,

1995). Brieﬂy, the amine is dissolved in MeOH or THF,

ꢂ

cooled to ꢁ78 C and to this is added 1.0 equivalent of

ethyltriﬂuoroacetate. After 1 h the solution is allowed to

warm to room temperature and the product is recovered

by simple evaporation of the solvent. Further puriﬁ-

cation is dependent on the compound and is described

in the relevant sections below.

5.7. Generalprocedure for mono- t-butoxy carbonylation

of di- and polyamines

Mono-protection of di- and polyamines with (Boc)₂O

was performed using the method of Krapcho and Kuell

(1990).

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

325

5.8. Generalprocedure for acyal tion with (E)-3-(4-

acetoxyphenyl) prop-2-enoyl chloride

42.36, 40.04, 33.92, 33.62, 29.15, 29.04; FAB-MS

(CDCl₃) m/z 213 ([M+1]⁺, 100), 117 (10).

Under an argon atmosphere (E)-3-(4-acetoxyphenyl)

prop-2-enoic acid (1.0 eq., 2.15 g, 10.4 mmol) was sus-

pended in 20 ml of CH₂Cl₂and a solution of oxalyl

chloride (1.2 eq., 1.76 g, 1.21 ml) in CH₂Cl₂(20 ml) was

added drop wise. Approximately 70 mL of DMF was

added as a catalyst. The reaction mixture was reﬂuxed for

1 h, cooled and partially evaporated to remove HCl. The

(E)-3-(4-acetoxyphenyl) prop-2-enoyl chloride so pro-

duced was added drop wise to a solution of the appro-

priate amine (1.0 eq.) and 3 eq. of Et₃N at RT over 10 min.

After the reaction was judged complete the solvent was

removed, and the residue dissolved in distilled water. The

aqueous phase was then extracted with CH₂Cl₂and the

combined organic fractions were washed with 0.25 M

Na₂CO₃to remove the any coumaric acid in solution.

The organic phase was dried with anhydrous Na₂SO₄and

evaporated to dryness to aﬀord the various amides

described below.

5.10.2. Synthesis of tert-butyl6-[(triﬂuoroacetyl)amino]-

hexylcarbamate (4b)

A solution of (Boc)₂O (1 eq., 16.5 g, 75.5 mmol) in

THF was added dropwise to 16.0 g (75.5 mmol) of 4a in

35 ml of THF stirred overnight at room temp. The sol-

vent was removed under reduced pressure to give 4b as a

clear yellow viscous oil (23.3 g, 99%) which solidiﬁed on

standing. TLC R_f=0.85 (CH₂Cl₂/MeOH/NH₄OH

30%=10:4:1); ¹H NMR spectral data (CDCl₃) ꢀ 6.5 (br

s, 1H), 4.52 (br s, 1H), 3.39 (t, 2H), 3.11 (br s, 2H), 1.45

(s, 9H, t-Bu), 1.39–1.28 (m, 8H); ¹³C NMR spectral data

(CDCl₃) 160.0, 39.87, 30.43, 29.17, 29.07, 28.78, 26.74,

26.16, 25.90; FAB-MS m/z 313 ([M+1]⁺, 8), 256 (46),

239 (29), 161 (65).

5.10.3. Synthesis of tert-butyl6-aminohexyclarbamate

(4c)

4b (23.0 g) was suspended in MeOH (50 ml) and con-

centrated NH₄OH (1:1). After stirring for 48 h the

reaction mixture was evaporated to dryness, water (60

ml) was added and the aqueous phase extracted with

CH₂Cl₂. Compound 4c remained in the aqueous phase

and was concentrated under reduced pressure to give a

colorless oil (12.0 g, 75%). R_f=0.57 (CH₂Cl₂/MeOH/

NH₄OH 30%=10:4:1); ¹H NMR spectral data

(CDCl₃) ꢀ 4.63 (br s, 1H), 3.09 (t, 2H), 2.70 (m, 2H),

2.25 (2H, -–NH₂), 1.42 (s, 9H, t-Bu), 1.40 (m, 4H),

1.29–1.20 (m, 4H); ¹³C NMR spectral data (CDCl₃)

156.4, 79.4, 42.2, 40.8, 33.4, 30.4, 28.8, 26.8; FAB-MS

m/z 433 ([2M+1]⁺, 7), 217 ([M+1]⁺, 100), 161 (56),

117 (19).

Acylation of 2.50 g (11.6 mmol) of 4c with 29 aﬀorded

1.74 g of a yellow spongy voluminous product. The

crude product was puriﬁed on silica gel with CH₂Cl₂/

MeOH=20:1, followed by CH₂Cl₂/MeOH=10:1 as

eluant to give 1.63 g of 29a. The yield was 40%. TLC

R_f=0.50 (CH₂Cl₂/MeOH=10:1); ¹H NMR spectral

data (CDCl₃) ꢀ 7.64 (d, J=16 Hz), 7.54 (d, 2H, J=8

Hz), 7.12 (d, 2H, J=8 Hz), 6.4 (d, 1H, J=16 Hz) 5.90

(br s, 1H), 4.55 (br s, 1H) 3.42 (m, 2H), 3.1 (m, 2H), 2.34

(s, 3H, OCH₃), 1.61 (m, 2H), 1.46 (s, 9H, t-Bu), 1.39 (m,

4H); ¹³C NMR spectral data (CDCl₃) 169.4, 162.7,

153.3, 147.9, 131.8, 130.2, 129.2, 122.4, 121.5, 117.2,

42.2, 40.8, 33.4, 30.4, 28.8, 21.5; FAB-MS m/z 809

([2M+1]⁺, 6), 405 ([M+1]⁺, 22) 349 (24), 305 (63), 189

(49), 147 (100).

5.9. Estimation of purities

The impurities of chief concern are residual amounts

of free amines, which could confound the glutamate

receptor assays, and E/Z isomerism of the coumaric

acid moiety. TLC easily shows whether any free amines

remain since for these samples the origin stains purple

with ninhydrin. No sample destined for bioassay con-

tained any free amines as determined by this method.

The extent to which the coumaroyl group had under-

gone E/Z isomerization was best determined by NMR,

and judging by the larger J-values recorded all samples

contained >95% of the E-isomer.

5.10. 2(E)-N-(6-aminohexyl)-3-(4-hydroxyphenyl)prop-

2-enamide (29b)

5.10.1. Synthesis of N-(6-aminohexyl)-2,2,2-

triﬂuoroacetamide (4a)

Following the general method, 9.75 g (83.9 mmol) of

1,6-hexanediamine (1), was dissolved in MeOH (50 ml)

and cooled to ꢁ78 ^ꢂC. A solution of ethyl tri-

ﬂuoroacetate (CF₃COEt), (1 eq., 11.9 g, 9.98 ml) was

added drop wise to the diamine with a gas tight syringe

and the reaction temperature maintained for 1.5 hr.

After warming to 0 ^ꢂC over 1 h, the MeOH solvent was

evaporated to aﬀord 16.9 g (95%) of crude 4a as a clear,

viscous pale yellow oil containing only trace of hexane-

diamine. TLC R_f=0.47 (CH₂Cl₂/MeOH/NH₄OH

Deprotection of 1.63 g (4.50 mmol) of 29a in 40 ml

ꢂ

dry MeOH/HCl (2.15 g) at 0 C provided 29b as yellow

crystals after solvent evaporation and recrystallization

from absolute EtOH. The yield was 800 mg (68%).

TLC R_f=0.30 (CH₂Cl₂/MeOH=10:1); ¹H NMR

(CDCl₃) ꢀ 8.02 (br s, 1H), 7.37 (d, 2H, J=8 Hz), 7.31 (d,

1H, J=16 Hz), 3.94 (br s, 1H), 3.14–3.13 (m, 2H), 2.74

1

30%=10:4:1); H NMR spectral data (CDCl₃) ꢀ 8.27

(br s, 1H), 3.28 (t, 2H), 2.65–2.61 (t, 2H), 1.53 (2H),

1.40–1.38 (m, 4H), 1.30–1.28 (m, 4H); ¹³C NMR spec-

tral data (CDCl₃) 157.79 (q, 36 Hz), 116.43 (q, 286 Hz),

326

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

(m, 2H), 1.53–1.08 (m, 8H); ¹³C NMR (CDCl₃) 165.15,

159.65, 139.27, 129.93, 126.81, 119.74, 116.59, 31.51,

29.88, 27.76, 26.78, 26.35; FAB-MS m/z 525 ([2M+1]⁺,

40), 263 ([M+1]⁺, 100), 147 (78) HRMS (FAB, NBA).

Calculated mass: 263.1760 for C₁₅H₂₃N₂O₂, found:

263.1718.

(CDCl₃) 175.4, 50.6, 40.9, 40.3, 30.2, 29.0, 28.8, 27.6,

26.7, 26.4. FAB-MS m/z 489 ([2M+1]⁺, 7), 245

([M+1]⁺, 100), 189 (93), 145 (18), 57 (83).

5.11.4. Synthesis of 4-[(1E)-3-({8-[tert-butoxy-

carbonyl)amino]octyl}amino)-3-oxoprop-1-enyl]phenyl

acetate

5.11. 2(E)-(8-aminooctyl)-3-(4-hydroxyphenyl)prop-2-

enamide (30b)

Acylation of 4.00 g of 5c (16.4 mmol) with 28 follow-

ing the general method aﬀorded 30a (4.28 g, 60%) as a

yellow spongy solid. TLC R_f=0.37 (CH₂Cl₂/

1

5.11.1. Synthesis of N-(8-aminooctyl)-2,2,2-triﬂuoro-

acetamide (5a)

MeOH=10:1); H NMR spectral data (CDCl₃) ꢀ 7.63

(d, 1H, J=16 Hz), 7.53 (d, 2H, J=8 Hz), 7.11 (d, 2H,

J=8 Hz), 6.39 (d, 1H, J=16 Hz), 5.78 (s, br, 1H), 4.54

(br s, 1H) 3.41–3.32 (m, 2H), 3.14 (m, 2H), 2.34 (s, 3H,

OCH₃), 1.58 (m, 2H), 1.45 (s, 9H), 1.32 (10H); ¹³C

NMR spectral data (CDCl₃) 166.3 159.8, 139.7, 130.0,

129.9, 126.9, 119.5, 116.6, 70.1, 69.9, 30.1, 29.0, 28.8, 27.6,

26.9; FAB-MS m/z 865 ([2M+1]⁺, 5), 433 ([M+1]⁺, 17),

377 (14), 359 (5) 333 (31), 189 (20), 147 (40).

Following the general procedure1,8-octanediamine 2

(7.60 g, 52.7 mmol) was mono-triﬂuoracetylated to

aﬀord 5a. Puriﬁcation of the crude product on silica gel

(MeOH/CH₂Cl₂/AcOH=60:40:2) gave 11.9 g (94%) of

a pale yellow oil. TLC R_f=0.48 (CH₂Cl₂/MeOH/

NH₄OH 30%=10:4:1); ¹H NMR spectral data (CDCl₃)

ꢀ 7.68 (br s, 1H) 3.29 (t, 2H), 2.65 (t, 2H), 1.73 (2H),

1.55 (2H), 1.42 (2H), 1.28 (m, 8H); ¹³C NMR spectral

data (CDCl₃) 157.79 (q, J=36 Hz), 116.38 (q, J=286

Hz), 50.38, 42.45, 40.2, 33.96, 29.78, 29.43, 27.17, 27.06;

FAB-MS m/z 241 ([M+1]⁺, 100), 185 (7), 145 (34), 126

(10), 69 (22), 56 (21), 30 (52).

Deprotection of 1.10 g of 30a (2.54 mmol) was

ac_ꢂcomplished using dry HCl (g) in MeOH (30 ml) at

0 C. Fine yellow crystals of 30b (650 mg, 88%) were

isolated by ﬁltration. TLC R_f=0.36 (CH₂Cl₂/

1

MeOH=10:1); H NMR (DMSO) ꢀ 7.37 (d, 2H, J=8

Hz), 7.31 (d, 1H, J=16 Hz), 6.81 (d, 2H, J=8 Hz), 6.45

(d, 1H, J=16 Hz), 3.16 (2H) 2.88 (2H), 1.54–1.43 (4H),

1.26 (8H); ¹³C NMR (DMSO) 166.1, 159.7, 139.2, 129.9,

119.7, 116.6, 30.0, 29.3, 29.1, 27.7, 27.2, 26.5; FAB-MS

m/z 581 ([2M+1]⁺, 5), 291 ([M+1]⁺, 100), 176 (26), 147

(89) 145 (41). HRMS (FAB, NBA). Calculated mass:

291.2073 for C₁₇H₂₇N₂O₂, found: 291.1998.

5.11.2. Synthesis of tert-butyl8-[(triﬂuoroacety)l -

amino]octylcarbamate (5b)

A solution of 1 eq. (Boc)₂O (10.85 g) in THF (30 ml)

was added to a solution of 5a (11.94 g, 49.75 mmol) in

THF (30 ml) via syringe at room temp. After stirring

overnight, the solvent was evaporated and the product

puriﬁed on silica gel (CH₂Cl₂/MeOH/ammonia water

30%=10:4:1) to aﬀord 5b (16.71 g, 99%) as a yellow

5.12. 2(E)-N-(12-aminododecyl)-3-(4-hydroxyphenyl)-

prop-2-enamide (31b)

viscous

oil.

TLC

R_f=0.75

(CH₂Cl₂/MeOH/

NH₄OH=10:4:1); ¹H NMR spectral data (CDCl₃) ꢀ

6.48 (br s, 1 H), 4.53 (br s, 1 H), 3.38–3.33 (m, 2H),

3.13–3.07 (m, 2H), 1.66–1.57 (m, 2 H), 1.45 (s, 9H, t-

Bu), 1.43–1.32 (10H); ¹³C NMR spectral data (CDCl₃)

175.4, 157.64 (q, J=36 Hz), 116.32 (q, J=287 Hz), 40.9,

40.3, 30.4, 29.6, 29.3, 29.2, 28.8, 27.0, 26.9, 26.8; FAB-

MS m/z 363 ([M+Na]⁺, 82), 341 ([M+1]⁺, 17), 285

(74), 241 (100), 57 (96).

5.12.1. Synthesis of tert-butyl12-aminododecylcarbamate

(6)

Mono-tert-butoxycarbonylation of 10.0 g of 1,12-

dodecanediamine 3 (50.0 mmol) with (Boc)₂O (1.36 g,

6.25 mmol) following the general procedure aﬀorded 6

(1.36 g, 74%) as yellow oil. TLC R_f=0.40 (CH₂Cl₂/

MeOH/NH₄OH=10:4:1); ¹H NMR spectral data

(CDCl₃) ꢀ 4.52 (br s, 1H), 2.71 (t, 2H), 1.45 (s, 9H, t-

Bu), 1.28 (22H); ¹³C NMR spectral data (CDCl₃) 176.5,

79.1, 42.9, 40.4, 34.3, 30.1, 29.9, 29.8, 29.7, 29.3, 29.0,

26.9, 26.5, 26.1; FAB-MS m/z 601 ([2M+1]+, 2), 301

([M+1]⁺, 100), 245 (79), 201 (36).

5.11.3. Synthesis of tert-butyl6-aminooctylcarbamate ( 5c)

16.6 g (2.15 mmol) of 5b was suspended in 50 ml of

MeOH and concentrated NH₄OH (1:1). After stirring at

reﬂux for 48 h a white precipitate was ﬁltered oﬀ, the

solvent was evaporated, water (60 ml) was added and

the aqueous phase extracted with CH₂Cl₂. Compound

5c remained in the aqueous phase and was concentrated

under reduced pressure to give 11.78 g (99%) of a

colorless oil which solidiﬁed at room temp. TLC

R_f=0.50 (CH₂Cl₂/MeOH/NH₄OH 30%=10:4:1); ¹H

NMR spectral data (CDCl₃) ꢀ 7.79 (br s, 2H), 4.6 (br s,

1H) 3.08 (t, 2H), 2.93 (t, 2H), 1.67 (t, 2H), 1.45 (s, 9H,

t-Bu), 1.40–1.30 (m, 10H); ¹³C NMR spectral data

5.12.2. Synthesis of 4-[(1E)-3-({12-[(tert-butoxy-

carbonyl)amino]dodecyl} amino)-3-oxoprop-1-enyl]-

phenylacetate

Acylation of 6 with 28 (1.38 g, 4.60 mmol) following

the general procedure gave 540 mg (22%) of 31a as a

yellow spongy solid. TLC R_f=0.46 (CH₂Cl₂/

1

MeOH=10:1); H NMR spectral data (CDCl₃) ꢀ 7.65

(d, 1H, J=16 Hz), 7.54 (d, 2H, J=9 Hz), 7.13 (d, 2H,

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

327

J=9 Hz), 6.37 (d, 1H, J=16 Hz), 5.62 (br s, 1H), 4.51

5.13.3. Synthesis of tert-butyl2-[2-(2-aminoethoxy)-

ethoxy]ethylcarbamate (9c)

(br s, 1H), 3.41 (t, 2H), 3.11 (m, 2H), 2.33 (s, 3H,

OCH₃), 1.58 (m, 2H), 1.46 (s, 9 H, t-Bu). ¹³C NMR

spectral data (CDCl₃) 169.8, 166.5, 156.7, 152.1, 139.7,

133.3, 129.7, 129.1, 122.6, 121.9, 79.3, 71.4, 42.9, 40.6,

34.5, 30.2, 29.8, 29.7, 29.6, 29.4, 29.1, 28.9, 27.0, 26.8;

FAB-MS m/z 489 ([M+1]⁺, 7), 433 (5), 389 (31), 245

(48), 189 (20), 147 (60), 57 (100). 31a (520 mg, 0.107

mmol) was treated with dry HCl (2.15 g) in MeOH and

aﬀorded 31b as a yellow solid (120 mg, 32%) after

recrystallization from absolute EtOH; TLC R_f=0.66

(CH₂Cl₂/MeOH/NH₄OH=10:4:1); ¹H NMR (d₆-

DMSO) ꢀ 7.60 (d, 1H, J=15 Hz), 7.41 (d, 2H, J=9 Hz),

6.87 (d, 2H, J=9 Hz), 6.29 (d, 1H, J=16 Hz), 5.69 (br s,

1H), 3.43 (m, 2H), 3.11 (m, 2H), 1.60–1.57 (m, 4H), 1.27

(16H). ¹³C NMR (d₆-DMSO) 172.2, 165.5, 144.2, 134.6,

131.0, 124.2, 121.4, 42.7, 40.4, 34.3, 30.0, 29.9, 29.8, 29.7,

29.5, 29.4, 29.3, 27.3, 27.0; FAB-MS m/z 347 ([M+1]⁺,

100), 330 ([M+1-NH₃, 2), 371 (2), 201 (6), 147 (83), 86

(4), 72 (3). HRMS (FAB, NBA). Calculated mass:

347.2699 for C₂₁H₃₅N₂O₂, found: 347.2679.

9b (33.0 g) was suspended in MeOH (100 ml) and 40 ml

conc. NH₄OH was added to the solution. The reaction

mixture was stirred and heated to reﬂux for 24 h. After

evaporation to dryness water (100 ml) was added and

the aqueous phase extracted with CH₂Cl₂. Evaporation of

the aqueous phase gave 9c as yellow oil (23.1 g, 97%).

TLC R_f=0.53 (CH₂Cl₂/MeOH/NH₄OH= 10:4:1); ¹H

NMR spectral data (CDCl₃) ꢀ 5.38 (br s, 1H), 3.72 (t, 2H),

3.69 (m, 4H), 3.54 (t, 2H), 3.27 (t, 2H), 3.14 (2H), 1.42 (s,

9H, t-Bu); ¹³C NMR spectral data (CDCl₃) 158.5, 114.1,

65.7, 65.6, 65.4, 62.4, 62.1, 61.9, 35.9, 35.1, 23.9, 23.6;

FAB-MS m/z 497 [2M +1]⁺, 5), 249 ([M+1]⁺, 84), 193

(30), 149 (100), 106 (30), 57 (40), 44 (96).

5.13.4. Synthesis of 4-[(1E)-16,16-dimethyl-3,14-dioxo-

7,10,15-trioxa-4,13-diazaheptadec-1-en-1-yl]phenyl

acetate(32a)

Acylation of 9c with 28 (10.0 g, 40.3 mmol) following

the general procedure gave 12.9 g of a crude product

that was puriﬁed by on silica gel (CH₂Cl₂/MeOH=20:1,

10:1) to give 32a as a yellow viscous oil (7.90 g, 45%).

5.13. 2(E)-N-{2-[2-(2-aminoethoxy)ethoxy]ethyl}-3-

(4-hydroxyphenyl)prop-2-enamide (32b)

1

TLC R_f=0.52 (CH₂Cl₂/MeOH=10:1); H NMR spec-

tral data (CDCl₃) ꢀ 7.65 (d, 1H, J=15 Hz), 7.54 (d, 2H,

J=8 Hz), 7.12 (d, 2H, J=8 Hz), 6.46 (d, 1 H, J=15

Hz), 3.64 (m, 6H), 3.59 (m, 4H), 3.34 (2H), 2.34 (s, 3H,

OCH₃), 1.46 (s, 9H, t-Bu); ¹³C NMR spectral data

(CDCl₃) 166.32, 159.70, 139.57, 130.02, 126.71, 119.42,

116.57, 70.33, 70.03, 29.07; FAB-MS m/z 459

([M+Na]⁺, 19), 437 [M+1]⁺, 13), 339 (6), 337 (62),

245 (22), 189(46), 147 (100).

5.13.1. Synthesis of N-{2-[2[(2-aminoethoxy)ethoxy]-

ethyl}-2,2,2-triﬂuorocaetamide (9a)

Monotriﬂuoroacetylation of 30.0 g (29.5 ml, 202

mmol) of 2,2⁰-(ethylenedioxy) bis (ethylamine) 7 fol-

lowing the general procedure gave 26.5 g (54%) of 9a,

following evaporation of the THF, dissolution in water

(100 ml) and extraction with CH₂Cl₂. The organic phase

was concentrated and puriﬁed on silica gel (CH₂Cl₂/

MeOH=20:1 followed by 10:1) TLC R_f=0.58 (CH₂Cl₂/

MeOH/NH₄OH=10:4:1); ¹H NMR spectral data

(CDCl₃) ꢀ 8.75 (br s, 1H), 3.72 (t, 2H), 3.64 (m, 6H),

3.54 (t, 2H), 3.12 (t, 2H); ¹³C NMR spectral data

(CDCl₃) 158.16 (q, J=36 Hz), 116.37 (q, J=286 Hz),

70.5, 69.5, 67.4, 40.1, 39.7, 22.3; FAB-MS m/z 245

([M+1]⁺, 100), 149 (22), 140 (49), 44 (50).

Deprotection of 32a (2.00 g, 4.60 mmol) in 40 ml

MeOH using dry HCl (2.15 g) gave, after evaporation of

the MeOH, a crude product, that was washed with Et₂O

to aﬀord 880 mg (65%) of 32b as a yellow viscous oil.

1

TLC R_f=0.20 (CH₂Cl₂/MeOH/NH₄OH=10:4:0.5); H

NMR (CDCl₃) ꢀ 7.37 (d, 2 arom H, J=8.36), 7.31 (d, 1

olef H, J=16), 6.81 (d, 2 arom H, J=8.40), 6.52 (d, 1

olef H, J=16), 3.66 (t, 2H), 3.53 (m, 4 H), 3.45 (t, 2H),

3.30 (t, 2H), 2.93 (t, 2H); ¹³C NMR (CDCl₃) 171.2,

164.5, 144.3, 134.7, 131.4, 124.2, 121.4, 75.3, 75.1, 74.8,

73.4, 72.2; FAB-MS m/z 589 ([2M+1]⁺, 5), 295

[M+1]⁺, 96), 178 (14), 149 (48), 147 (7), 89 (16), 44

(53). HRMS (FAB, NBA). Calculated mass: 295.1658

for C₁₅H₂₃N₂O₄, found: 295.1612.

5.13.2. Synthesis of N-tert butyl2-(2- {2-[(triﬂuoro-

acetyl)amino] ethoxy}ethoxy) ethylcarbamate) (9b)

A solution of (Boc)₂O (1 eq., 21.97 g) in THF (60 ml)

was added to 9a (24.56 g, 100.7 mmol) in THF (100 ml)

at room temp. and stirred for 3 h. The reaction mixture

was evaporated to dryness to aﬀord 9b as a pale yellow

oil (33.25 g, 97%) which was used for the next step of

the reaction without further puriﬁcation. TLC R_f=0.73

(CH₂Cl₂/MeOH/NH₄OH=10:4:1); ¹H NMR spectral

data (CDCl₃) ꢀ 3.69 (m, 6H), 3.54 (m, 4H), 3.31 (t, 2H),

1.43 (s, 9H, t-Bu); ¹³C NMR spectral data (CDCl₃)

174.98, 157.29 (q, 36 Hz), 116.37 (q, J=286 Hz), 70.5,

70.4, 69.2, 67.2, 40.0, 39.6, 22.3; FAB-MS m/z 367

([M+Na]⁺, 100), 345 ([M+1]⁺, 5), 289 (13), 245 (42), 57

(33).

5.14. 2(E)-N{3-[4-(3-aminopropoxy)butoxy]propyl}-3-

(4-hydroxyphenyl)prop-2-enamide (33b)

5.14.1. Synthesis of N-{3-[4-(3-aminopropoxy)-

butoxy]propyl})-2,2,2-triﬂuoroacetamide (10a)

Following the general procedure12.0 g (58.8 mmol,

12.5 ml) of 4,9-dioxa-1,12-dodecanediamine, 8, was

mono-triﬂuoroacetylated to give a crude product

that was puriﬁed on silica gel (MeOH/CH₂Cl₂/

328

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

AcOH=10:20:1, followed by 20:20:1) to aﬀord 10a as a

yellow viscous oil (14.6 g, 82%). TLC R_f=0.60

(CH₂Cl₂/MeOH/NH₄OH=10:4:1); ¹H NMR spectral

data (CDCl₃) ꢀ 8.20 (br s, NHCOCF₃), 3.61–3.51 (m, 4H),

3.40 (t, 4H), 3.06 (m, 2H), 1.94 (m, 6H), 1.85 (m, 2H), 1.61

(4H); ¹³C NMR spectral data (CDCl₃) 157.5 (q, J=36

Hz), 116.4 (q, J=286 Hz), 71.5, 70.2, 69.1, 39.3, 38.6, 28.5,

27.7, 26.7, 23.8; FAB-MS m/z 301 ([M+1]⁺, 100), 205

(18), 154 (68), 126 (13), 58 (26), 30 (18).

139.7,133.1, 129.6, 129.1, 122.5, 121.7, 79.3, 71.1, 71.0,

69.9, 69.5, 39.1, 38.6, 30.0, 29.5, 28.8, 26.9, 21.5; FAB-

MS m/z 493 ([M+1]⁺, 14), 393 (64), 305 (15), 249 (35),

189 (20), 147 (55), 102 (68), 74 (25), 57 (100).

5.14.5. Synthesis of (2E)-N-{3-[4-(3-aminopropoxy)-

butoxy]propyl} -3-(4-hydroxyphenyl)prop-2-enamide (33b)

Deprotection of 33a (8.30 g, 16.9 mmol) using dry

HCl (2.15 g) in MeOH at 0 ^ꢂC gave a crude product after

solvent removal that was dissolved in 1N HCl (80 ml)

and extracted with CH₂Cl₂. Evaporation of the aqueous

phase gave 33b as a bright yellow oil (5.38 g, 91%). TLC

R_f=0.57 (CH₂Cl₂/MeOH/NH₄OH=10:4:1); HPLC

Rt=10.8 min; ¹H NMR spectral data (d₆-DMSO) ꢀ

7.38 (d, 2H, J=8 Hz), 7.31 (d, 1H, J=16 Hz), 6.81 (d,

2H, J=8 Hz), 6.46 (d, 1H, J=16 Hz), 3.42 (m, 10H),

3.18 (t, 2H), 2.83–2.78 (m, 2H), 1.83 (m, 2H), 1.70 (q,

2H), 1.51 (m, 4H); ¹³C NMR spectral data (d₆-DMSO)

166.3, 159.7, 139.4, 129.9, 126.7, 119.6, 70.7, 68.5, 67.7,

37.4, 36.8, 30.3, 28.1, 26.8, 23.4; FAB-MS m/z 351

([M+1]⁺, 100), 247 (41), 205 (51), 147 (64), 130 (7), 74

(12), 58 (79), 30 (60). HRMS (FAB, NBA). Calculated

mass: 351.2284 for C₁₉H₃₁N₂O₄, found: 351.2278.

5.14.2. Synthesis of tert-butyl3-(4- {3-[(triﬂuoro-

acetyl)amino]propoxy} butoxy)propylcarbamate (10b)

A solution of (Boc)₂O (1 eq., 9.13 g) in THF (35 ml)

was added to 10a (12.6 g, 41.8 mmol) in THF (45 ml) at

RT and the reaction mixture stirred overnight. Evap-

oration of the solvent aﬀorded 10b as a pale yellow oil

(16.3 g, 97%) which was used for the next step of the

synthesis without puriﬁcation. TLC R_f=0.80 (CH₂Cl₂/

MeOH/NH₄OH=10:4:1); ¹H NMR spectral data

(CDCl₃) ꢀ 7.54 (br s, 1H), 4.48 (br s, 1H), 3.61–3.57 (m,

2H), 3.49–3.45 (m, 10H), 3.2 (t, 2H), 1.88–1.81 (m, 4H),

1.72 (2H), 1.43 (s, 9H, t-Bu); ¹³C NMR spectral data

(CDCl₃) 176.0, 157.3 (q, 36 Hz), 116.4 (q, J=286 Hz),

71.6, 70.7, 69.8, 68.7, 39.8, 31.5, 30.1, 28.8, 28.3, 26.8,

25.9, 21.1; FAB-MS m/z 423 ([M+Na]⁺, 17), 401

([M+1]⁺, 10), 345 (12), 301 (100), 263 (6), 249 (21), 126

(11), 102 (64), 74 (18), 57 (83).

5.15. Synthesis of (2E)-N-{2-[(2-aminoethyl)amino]-

ethyl}-3-(4-hydroxyphenyl)prop-2-enamide (34b)

5.15.1. Synthesis of tert-butyl2-[(2-aminoethy)l -

amino]ethylcarbamate (14)

5.14.3. Synthesis of tert-butyl3-[4-(3-aminopropoxy)-

butoxy]propylcarbamate ( 10c)

Mono-protection of 10.0 g (97.1 mmol, 10.5 ml) of

diethylenetriamine 11 with (Boc)O₂(0.13 eq., 2.83 g,

12.9 mmol) following the general procedure gave 14 as a

colorless viscous oil (2.30 g, 89%). TLC R_f=0.83

(CH₂Cl₂/MeOH/NH₄OH=10:4:1); ¹H NMR spectral

data (CDCl₃) ꢀ 4.48 (2H), 3.35 (2H), 3.28 (2H), 2.18 (2H),

1.43 (9H); ¹³C-NMR spectral data (CDCl₃) 156.6, 79.5,

70.4, 53.2, 49.8, 39.7, 28.8; FAB-MS m/z 204 ([M+1]⁺,

76), 148 (54), 131 (16), 104 (40), 87 (34), 73 (35), 57 (100).

To a solution of 16.2 g (40.5 mmol) of 10b in MeOH

(100 ml) was added 40 ml of concentrated NH₄OH.

After heating to reﬂux for 20 h and evaporation to dry-

ness, the residue was puriﬁed on silica gel (CH₂Cl₂/

MeOH=20:1, then 10:1) to give 10c as a yellow oil (12.2 g,

99%). TLC R_f=0.70 (CH₂Cl₂/MeOH/NH₄OH=

1

10:4:1); H NMR spectral data (CDCl₃) ꢀ 5.01 (br s,

1H), 3.59 (t, 2H), 3.49–3.42 (m, 6H), 3.20 (t, 2H), 3.12

(t, 2H), 1.96 (m, 2H), 1.92 (m, 2H), 1.76 (t, 2H), 1.62 (m,

4H), 1.43 (s, 9H, t-Bu); ¹³C NMR spectral data (CDCl₃)

176.6, 71.5, 71.1, 70.9, 69.7, 69.3, 39.4, 38.9, 30.1 28.8,

27.4, 26.8, 26.7; FAB-MS m/z 327 ([M+Na]⁺, 14), 305

([M+1]⁺, 100), 249 (23), 231 (6), 205 (16), 146 (10), 130

(6), 102 (30), 74 (20), 58 (84), 57 (67).

5.15.2. Synthesis of 4-[(1E)-13,13-dimethyl-3,11-dioxo-

12-oxa-4,7,10-triazatetradec-1-en-1-yl]-3-(4-hydroxy-

phenylacetate ( 34a)

Acylation of 14 (1.00 g, 4.93 mmol) following the

general procedure gave 1.31 g of a ﬂuﬀy yellow crude

product which was suspended in water (50 ml) and

extracted with CH₂Cl₂. The organic phase was then

dried with Na₂SO₄. Puriﬁcation on silica gel (CH₂Cl₂/

MeOH=20:1) gave 34a (0.32 g, 0.82 mmol, 16%) and a

diacylated product 34c (0.86 g, 1.48 mmol, 30%).

5.14.4. Synthesis of 4-[(1E)-20,20-dimethyl-3,18-dioxo-

8,13,19-trioxa-4,17-diazahenicos-1-en-yl]phenyl acetate (33a)

Acylation of 10c (10,0 g, 32.9 mmol) with 28 follow-

ing the general procedure gave 33a as a yellow oil (10.0

g, 62%). TLC R_f=0.52 (CH₂Cl₂/MeOH=10:1); ¹H

NMR spectral data (CDCl₃) ꢀ 7.56 (d, 1H, J=16 Hz),

7.46 (d, 2H, J=8 Hz), 7.05 (d, 2H, J=16 Hz), 6.38 (d,

1H, J=8 Hz), 5.03 br, 1H), 3.51–3.47 (t, 2H,), 3.45–3.38

(m, 8H), 3.17 (m, 2H), 2.26 (s, 3H, OCH₃), 1.81 (m, 2H),

1.78 (m, 2H), 1.64 (m, 4H), 1.40 (s, 9H, t-Bu); ¹³C NMR

spectral data (CDCl₃) 169.6, 166.3, 156.5, 151.8,

Compound

34a

TLC

R_f=0.45

(CH₂Cl₂/

1

MeOH=10:1); H NMR spectral data (CDCl₃) ꢀ 7.30

(d, 2H, J=8 Hz), 7.10 (d, 1H, J=16 Hz), 6.80 (d, 2H,

J=8 Hz), 6.48 (d, 1 olef H, J=15.6), 3.16 (2H), 2.97

(2H), 2.73 (4H), 2.31 (s, 3H, OCH₃), 1.43 (s, 9H, t-Bu);

¹³C NMR spectral data (CDCl₃) 169.5, 166.8, 158.3,

151.7, 139.8, 133.4, 129.7, 129.0, 127.1, 121.9, 79.3, 70.4,

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

329

56.8, 49.9, 39.5, 36.8; FAB-MS m/z 392 ([M+1]⁺, 45),

335 (20), 318 (83), 189 (15), 147 (45), 57 (100).

Compound 34c TLC R_f=0.66 (CH₂Cl₂/MeOH=10:1);

FAB-MS m/z 580 ([M+1]⁺, 2), 480 (10), 438 (6), 189

(33), 147 (100), 57 (57).

Deprotection of 34a (300 mg, 0.77 mmol) in 30 ml

MeOH was accomplished by adding 20 drops of con-

centrated HCl and stirring at 35 ^ꢂC for 1.5 h. After

evaporation of solvent followed by chromatography

(CH₂Cl₂/MeOH/NH₄OH=10:4:0.5), 34b was isolated

as a yellow solid (140 mg, 73%).

gave 480 mg of a ﬂuﬀy yellow crude product which was

suspended in water (50 ml) and extracted with CH₂Cl₂.

The organic phase was then dried with Na₂SO₄and eva-

porated.

Puriﬁcation on silica gel gave the desired mono-

acylated product 35a (0.22 g, 0.52 mmol, 11%) and a

diacylated product 35c (0.23 g, 0.38 mmol, 8%).

Compound 35a TLC R_f=0.60 (CH₂Cl₂/MeOH=10:1);

¹H NMR spectral data (CDCl₃) ꢀ 7.37 (d, 2H, J=8 Hz),

7.30 (d, 1H, J=16 Hz), 6.98 (d, 2H, J=8 Hz), 6.38 (d, 1H,

J=116 Hz), 3.16 (m, 2H), 2.99–2.95 (m, 4H), 2.59 (m, 2H),

1.89 (s, 3H, OCH₃), 1.56 (2H), 1.50 (2H), 1.43 (s, 9H, t-Bu);

¹³C NMR spectral data (CDCl₃) 169.5, 164.8, 153.3, 147.9,

130.9, 130.2, 129.2, 122.4, 121.5, 117.2, 79.1, 53,9, 48.6,

41.7, 36.7, 34.0, 30.9, 28.6; FAB-MS m/z 420 ([M+1]⁺, 8),

320 (15), 216 (24), 147 (100), 129 (18), 100 (33), 57 (96).

Deprotection of 35a (200 mg, 0.48 mmol) was

Similarly, 0.59 g of 34c (1.02 mmol) suspended in

MeOH (20 ml) was reacted with 20 drops of con-

centrated HCl. The reaction mixture was stirred at 35 ^ꢂC

for 1.5 h. After evaporation of solvent followed by column

chromatography (CH₂Cl₂/MeOH/NH₄OH=10:4:0.5),

34d was isolated as a bright yellow solid (200 mg, 50%).

Compound 34b TLC R_f=0.44 (CH₂Cl₂/MeOH/

ꢂ

accomplished with dry HCl (2.15 g) in MeOH at 0 C

1

NH₄OH=10:4:1); H NMR (d₆-DMSO) ꢀ 7.30 (d, 2H,

and gave 35b (105 mg, 79%) after recrystallization from

absolute EtOH. TLC R_f=0.39 (CH₂Cl₂/MeOH/

NH₄OH=10:4:1); ¹H NMR (DMSO) ꢀ 7.37 (d, 2H,

J=8 hz), 7.30 (d, 1H, J=16 Hz), 6.98 (d, 2H, J=8 Hz),

6.38 (d, 1H, J=16 Hz), 3.10 (m, 2H), 2.99–2.85 (m, 4H),

2.52 (m, 2H), 1.56 (2H), 1.49 (2H); ¹³C NMR (d₆-

DMSO) 165.2, 159.7, 139.3, 129.9, 126.8, 119.7, 116.6,

15.6, 47.2, 40.8, 56.0, 33.8, 24.9; FAB-MS m/z 278

[M+1]⁺, 65), 260 (15), 190 (21), 162 (45), 176 (25), 147

(80), 57 (100). HRMS (FAB, NBA). Calculated mass:

278.1869 for C₁₅H₂₄N₃O₂, found: 278.1829.

J=8 Hz), 7.10 (d, 1H, J=16 Hz), 6.80 (d, 2H, J=8 Hz),

6.48 (d, 1H, J=16 Hz), 3.81 (br s, 1H), 3.60 (1H), 3.41–

3.39 (m, 4H), 3.07 (2H); ¹³C NMR (d₆-DMSO) 166.7,

159.9, 139.8, 130.0, 126.5, 119.5, 116.4, 55.8, 47.6, 45.3,

36.9; FAB-MS m/z 499 ([2M+1]⁺, 5%), 250 ([M+1]⁺,

100), 233 (13), 219 (5), 190 (22), 176 (10), 147 (94).

HRMS (FAB, NBA). Calculated mass: 250.1556 for

C₁₃H₂₀N₃O₄, found: 250.1517.

Compound 34d TLC R_f=0.76 (CH₂Cl₂/MeOH/

NH₄OH=10:4:1); FAB-MS m/z 396 ([M+1]⁺, 58), 379

(6), 250 (40), 190 (21), 147 (100).

5.17. 2(E)-N-{3-[4-aminocyclohexyl)amino]propyl}-3-

(4-hydroxyphenyl)prop-2-enamide (36b)

5.16. 2(E)-N-{3-[(3-aminopropyl)amino]propyl}-3-(4-

hydroxyphenyl)prop-2-enamide (35b)

5.17.1. Synthesis of tert-butyl 4-aminocyclohexyl-

carbamate (16)

5.16.1. Synthesis of tert-butyl3-[(3-aminopropy)l -

amino]propylcarbamate ( 15)

Monoprotection of 30.0 g (263 mmol) of 1,4-diami-

nocyclohexane 13 with 0.125 eq. of (Boc)₂O (32.8 mmol,

7.17 g) following the general procedure gave, after

removal of the reaction solvent, a residue that was par-

titioned between water and dichloromethane. The mono-

Boc and di-Boc derivatives were extracted into the organic

phase leaving the unreacted starting material in the aqu-

eous phase. Separation of the mixture on silica gel with

CH₂Cl₂/MeOH/NH₄OH =10:4:1 gave 5.18 g (74%) of

16. TLC R_f=0.58 (CH₂Cl₂/MeOH/NH₄OH=10:4:1); ¹H

NMR spectral data (CDCl₃) ꢀ 4.39 (br s, 1 H), 3.44–3.36

(m, 1H), 2.62–2.59 (m, 1H), 2.17 (2H), 1.99 (2H), 1.47

(2H), 1.43 (9H), 1.25–1.06 (4H); ¹³C NMR spectral data

(CDCl₃) 155.6 (C¼O), 79.5 (C–O), 50.3, 49.6, 35.7, 32.5,

28.8; FAB-MS m/z 429 ([2M+1]+, 11), 215([M+1]⁺,

61), 159 (45), 142 (100), 96 (26), 81 (36), 57 (65).

Mono-protection of 10.0 g (76.3 mmol, 10.7 ml) of 3,3⁰-

iminobispropylamine 12 with (Boc)₂O (0.13 eq., 2.22 g,

10.2 mmol) following the general procedure gave a crude

product which, after removal of the dioxane, was dis-

solved in water (60 ml) and extracted with CH₂Cl₂. The

organic phase was dried with Na₂SO₄and concentrated.

The crude product was puriﬁed on silica gel (CH₂Cl₂/

MeOH/NH₄OH =10:4:1). Compound 15 was obtained as

a clear yellow oil (1.48 g, 63%). TLC R_f=0.26 (CH₂Cl₂/

MeOH/NH₄OH=10:4:1); ¹H NMR spectral data (CDCl₃)

ꢀ 5.20 (br s, 1H), 3.21 (m, 2H), 2.78 (t, 2H), 2.67–2.62 (m,

4H), 1.69–1.59 (m, 4H), 1.43 (s, 9H, t-Bu); ¹³C NMR spec-

tral data (CDCl₃) 156.54, 79.86, 53.81, 48.10, 40.72, 39.52,

33.99, 30.26, 28.81; FAB-MS m/z 232 ([M+1]⁺, 100), 186

(5), 132 (36), 102 (11), 87 (8), 74 (9), 57(80), 44 (48).

5.16.2. Synthesis of 4-[(1E)-15,15-dimethyl-3,13-dioxo-

14-oxa-4,8,12-triazahexadec-1-en-1yl]phenyl acetate (35a)

Acylation of 15 with (E)-3-(4-acetoxyphenyl) chloride

28 (1.10 g, 4.72 mmol) following the general procedure

5.17.2. Synthesis of tert-butyl4-[(2-cyanoethy)l -

amino]cyclohexylcarbamate (17)

16 5.00 g, 23.4 mmol) was dissolved in MeOH (60 ml)

and 1 eq. of acrylonitrile (1.24 g, 1.54 ml) in MeOH

330

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

(40 ml) was added slowly over 1 h and stirred for 6 h.

Evaporation of the solvent aﬀorded 5.84 g (94%) of

of the solvent, followed by recrystallization from abso-

lute EtOH gave 0.75 g (77%) of 36b. TLC R_f=0.10

(CH₂Cl₂/MeOH=10:1); ¹H NMR (d₆-DMSO) ꢀ 7.51 (d,

1H, J=16 Hz), 7.42 (d, 2H, J=9 Hz), 6.78 (d, 2H, J=9

Hz), 6.45 (d, 1H, J=16 Hz), 3.62 (t, 2H), 2.96 (m, 4H),

2.70–2.64 (2H), 1.60 (m, 4H), 1.30–1.26 (m, 4H); ¹³C NMR

(d₆-DMSO) 166.7, 159.8, 139.7, 130.0, 126.5, 119.8, 116.6,

49.1, 49.0, 36.7, 28.8, 27.1 FAB-MS m/z 318 ([M+1]⁺,

48), 214 (59), 176 (10), 160 (22), 147 (17), 142 (9), 136 (100),

113 (6), 81 (22). HRMS (FAB, NBA). Calculated mass:

318.2182 for C₁₈H₂₈N₃O₂, found: 318.2188.

crude

17.

TLC

R_f=0.63

(CH₂Cl₂/MeOH/

NH₄OH=10:4:1); ¹H NMR spectral data (CDCl₃) ꢀ

4.38 (br s, 1H), 3.48 (1H), 3.42 (1H), 2.96 (t, 2H), 2.50

(t, 2H), 2.04 (2H), 1.95–1.92 (2H), 1.45 (s, 9H, t-Bu),

1.2–1.07 (m, 5H); ¹³C NMR spectral data (CDCl₃) 155.6

(C¼O), 119.1 (CN), 79.6 (C–O), 55.8, 49.8, 42.7, 32.4,

32.3, 28.8, 19.6; FAB-MS m/z 535 ([2M+1]⁺, 5), 268

([M+1]⁺, 48), 212 (35), 171 (17), 142 (82), 96 (22), 81

(44), 57 (100).

5.17.3. Synthesis of tert-butyl4-[(3-aminopropy)l

amino]cyclohexy l carbamate (18)

5.18. 2(E)-N-{3-[4-(3-aminopropyl)piperazin-1-yl}-

propyl}-3-(4-hydroxyphenyl)prop-2-enamide (37b)

17 (5.00 g, 18.7 mmol) was dissolved in 150 ml of 95%

EtOH/1N NaOH. One gram of Raney-Ni catalyst was

added and the mixture subjected to hydrogenation at 60

psi for 24 h. After ﬁltering through Celite¹521, the

reaction mixture was concentrated to a volume of about

40 ml. After a few hours an oil separated on top which

was removed. Puriﬁcation on silica gel (5:2=CH₂Cl₂/

MeOH) gave 18 (3.90 g, 77%) as a white ﬂuﬀy solid.

TLC R_f=0.39 (CH₂Cl₂/MeOH/NH₄OH=10:4:1); ¹H

NMR spectral data (CDCl₃) ꢀ 4.44 (br s, 1H), 3.45 (1H),

2.78 (t, 2H), 2.68 (t, 2H), 2.30 (1H), 2.15 (2H), 2.10–1.96

(5H), 1.65–1.60 (m, 2H), 1.43 (s, 9H, t-Bu), 1.19–1.13

(m, 4H); ¹³C NMR spectral data (CDCl₃) 155.7 (C¼O),

79.5 (C–O), 56.5, 49.9, 45.4, 40.8, 34.1, 32.4, 32.3, 28.8;

FAB-MS m/z 272 ([M+1]⁺, 100), 216 (30), 171(7), 142

(25), 96 (14), 81 (30), 57 (58).

5.18.1. Synthesis of tert-butyl3-[4-(3-aminopropy)l

piperazin-1-yl]propyl carbamate (22)

Mono-protection of 12.0 g of 1,4-bis (3-aminopropyl)

piperazine 19 (60.0 mmol, 12.3 ml) with (Boc)₂O (1.75 g,

8 mmol) following the general conditions provided a

residue after evaporation that was suspended in water

(80 ml) and extracted with CH₂Cl₂. Removal of the

solvent provided 21 as a colorless oil (2.13 g, 89%)

which formed needle-like crystals on standing. TLC

R_f=0.33 (CH₂Cl₂/MeOH/NH₄OH 30%=10:4:1); ¹H

NMR spectral data (CDCl₃) ꢀ 5.52 (br s, 1H, NHBoc),

3.17 (q, 2H), 2.76 (q, 2H), 2.44–2.35 (m, 10H), 1.67–1.57

(m, 6H), 1.42 (s, 9H, t-Bu); ¹³C NMR spectral data

(CDCl₃) 156.5, 67.5, 57.2, 56.9, 53.7, 53.6, 41.3, 40.4,

30.8, 28.8, 26.7; FAB-MS m/z 301 ([M+1]⁺, 92), 270

(10), 256 (14), 242 (11), 201 (68), 127 (20), 97 (39), 70

(42), 57 (100).

5.17.4. Synthesis of 4-((1E)-3-{[3-([4-[(tert-butoxy-

carbonyl)amino] cyclohexyl}amino)propyl]amino}-3-

oxoprop-1-enyl)phenyl acetate (36a)

5.18.2. Synthesis of 4-(1E)-3-{[3-(4-{3-[tert-butoxy-

carbonyl)amino]propyl} piperazin-1-yl)propyl]amino}-

3-oxoprop-1-enyl)phenyl acetate (37a)

(E)-3-4-(4-Acetoxyphenyl) prop-2-enoic acid (1.14 g,

5.54 mmol) of was suspended in CH₂Cl₂(50 ml) to

which was added 1.10 eq. (1.26 g) of DCC in CH₂Cl₂

and 0.25 eq. of 1-hydroxybenzotriazole. The solution

was then added to a mixture of 1.5 g (5.53 mmol) of 18

in CH₂Cl₂(80 ml) and the reaction mixture reﬂuxed for

36 h. Evaporation of the solvent gave a yellow ﬂuﬀy

crude product which was puriﬁed on silica gel (CH₂Cl₂/

MeOH =10:1) to give 1.60 g (63%) of 36a. TLC

R_f=0.1 (CH₂Cl₂/MeOH=10:1); ¹H NMR spectral data

(MeOH) ꢀ 7.51 (d, 1H, J=16 Hz), 7.43 (d, 2H, J=8

Hz), 6.82 (d, 2H, J=8 Hz), 6.45 (d, 1H, J=16 Hz), 3.72

(t, 1H), 2.92–2.85 (m, 4H), 2.07 (2H), 2.03 (s, 3H,

OCH₃), 1.96 (1H), 1.92–1.87 (m, 4H), 1.43 (s, 9H, t-Bu),

1.35–1.24 (m, 6H); ¹³C NMR spectral data (d-MeOH)

168.9, 159.9, 156.8, 141.3, 129.7, 126.4, 124.2, 117.7,

116.8, 115.9, 111.5, 79.0, 67.9, 56.3, 51.1, 43.1, 36.5,

31.0, 29.3, 27.9, 25.5; FAB-MS m/z 460 ([M+1]⁺, 100),

404 (18), 356 (12), 147 (49), 96 (17), 57 (65).

Acylation of 21 (1.80 g, 6 mmol) with 28 following the

general procedure gave a crude product that was puri-

ﬁed by chromatography on silica gel (CH₂Cl₂/MeOH/

NH₄OH=10:4:0.5) and yielded 37a as a yellow spongy

solid (1.36 g, 52%). TLC R_f=0.25 (CH₂Cl₂/

1

MeOH=10:1); H NMR spectral data (CDCl₃) ꢀ 7.57

(d, 1H, J=16 Hz), 7.49 (d, 2H, J=8 Hz), 7.32 (d, 2H,

J=8 Hz), 6.35 (d, 1H, J=16 Hz), 5.43 (br s, 1H), 3.48

(m, 2H), 3.16 (m, 2H), 2.51–2.47 (m, 8H), 2.43–2.38 (m,

4H), 2.29 (s, 3H, OCH₃), 1.74 (t, 2H), 1.68 (m, 2H), 1.43

(s, 9H, t-Bu); ¹³C NMR spectral data (CDCl₃) 169.7,

166.0, 156.5, 151.5, 139.5, 133.2, 129.7, 129.1, 122.4,

121.9, 116.5, 79.2, 57.9, 57.2, 53.8, 46.3, 40.2, 28.8, 26.7,

25.4, 21.5; FAB-MS m/z 489([M+1]⁺, 28), 389(10), 256

(11), 201 (17), 147 (42), 102 (18), 74 (14), 57 (100).

Deprotection of 37a (1.36 g, 2.79 mmol) was accom-

plished in MeOH (30 ml) by treating with 20 drops of

concentrated HCl. After 1.5 h the reaction was evapo-

rated to dryness, the residue was dissolved in absolute eth-

anol and ﬁltered. The ethanolic solution was evaporated

Deprotection of 36a (1.50 g, 3.27 mmol) was accom-

plished in MeOH (50 ml) containing 20 drops of conc.

HCl. After stirring at room temp. for 5 h, evaporation

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

331

giving, 37b (687 mg, 63%) as a pale yellow solid. TLC

1

that was puriﬁed on silica gel (CH₂Cl₂/MeOH/

AcOH=10:20:1). The yield of 24 was 3.25 g (46%).

TLC R_f=0.48 (CH₂Cl₂/MeOH/AcOH=60:40:2); ¹H

NMR spectral data (CDCl₃) ꢀ 3.56–3.52 (m, 1H), 3.29–

3.27 (m, 1H), 2.82–2.68 (m, 5H), 2.47–2.44 (q, 1H),

1.71–1.62 (t, 2H), 1.19 (2H), 0.93 (2H); ¹³C NMR spec-

tral data (CDCl₃) 157.4 (q, J=36 Hz), 116.5 (q, J=285

Hz), 62.3, 61.7, 45.8, 44.9, 40.7, 40.3, 34.1, 31.8, 31.7,

28.3, 25.5, 25.1; FAB-MS m/z 325 ([M+1]⁺, 100), 251

(27), 155 (37), 81 (33), 56 (35).

R_f=0.20 (CH₂Cl₂/MeOH/NH₄OH=10:4:1); H NMR

(d₆-DMSO) ꢀ 7.40 (d, 2H, J=8 Hz), 7.36 (d, 1H, J=16

Hz), 6.81 (d, 2H, J=8 Hz), 6.44 (d, 1H, J=16 Hz), 3.73

(m, 8H), 3.45 (m, 2H), 3.23–3.09 (4H), 2.91 (2H), 2.02

(m, 4H); ¹³C NMR (d₆-DMSO) 166.1, 159.3, 139.2,

129.5, 126.1, 118.7, 116.1, 55.0, 41.4, 36.5, 26.5, 24.1

FAB-MS m/z 369 ([M+Na]^+,5), 347 ([M+1]⁺, 100),

204 (17), 176 (8), 147 (49), 144 (6), 127 (14), 113 (16), 86

(13). HRMS (FAB, NBA). Calculated mass: 347.2447

for C₁₉H₂₃₁N₄O₂, found: 347.2409.

5.19.4. Synthesis of triBoc-N-[3-({2-[(3-aminopropyl)-

amino]cyclohexyl}amino)propyl]-2,2,2-triﬂuoroacetamide

38) and triBoc-N,N⁰-bis(3-aminopropyl)cyclohexane-

1,2-diamine (39)

5.19. 2(E)-N-[3-({2-[(3-aminopropyl)amino]-

cyclohexyl}amino)propyl]-3-(4-hydroxyphenyl)prop-2-

enamide (40b)

24 (3.10 g, 9.57 mmol) was suspended in THF (60 ml)

and 3.0 eq. of (Boc)₂O (28.7 mmol, 6.26 g) in THF (40

ml) was added at room temp.. When the formation of

38 was complete as judged by TLC, 15 ml of con-

centrated. NH₄OH was added followed by reﬂuxing at

5.19.1. Synthesis of N,N⁰-bis (2-cyanoethyl)cyclohexane-

1,2-diamine (22)

trans-1,2-Diaminocyclohexane 20 (10.0 g (87.6 mmol,

6.72 ml) was dissolved in MeOH (65 ml) and 2.50 eq. of

acrylonitrile (14.4 ml, 11.6 g) in 50 ml of MeOH was

added drop wise over 30 min. After stirring overnight at

room temp. and evaporation of the solvent 22 was

recovered as a yellow oil (19.1 g, 99%). TLC R_f=0.88

(CH₂Cl₂/MeOH/NH₄OH=10:4:1); ¹H NMR spectral

data (CDCl₃) ꢀ 3.04–2.98 (m, 2H), 2.79–2.75 (m, 2H),

2.53–2.48 (t, 4H), 2.17–2.15 (t, 2H), 2.14 (2H), 1.96 (br

s, 2H), 1.75–1.72 (m, 2H); ¹³C NMR spectral data

(CDCl₃) 119.3, 61.4, 42.8, 32.0, 25.2, 19.7; FAB-MS m/z

221 ([M+1]⁺, 100), 194 (4), 180 (12), 151 (33), 109 (14),

81 (13)

ꢂ

60 C for 24 h to remove the TFA protecting group.

After evaporation of the solvent the crude material was

puriﬁed by cc to give 3.28 g (60%) of the triBoc amine

39. TLC R_f=0.70 (CH₂Cl₂/MeOH=10:1); ¹H NMR

spectral data (CDCl₃) ꢀ 5.87 (br s, 1H), 4.90 (4H), 4.59

(2H), 4.03 (4H), 3.82 (1H), 3.03 (6H), 1.75–1.66 (7H),

1.45 (s, 27H); ESI-MS m/z 1057.8 ([2M+1]⁺, 8), 529.4

([M+1]⁺, 100), 429.4 (80), 329.4 (35), 229.2 (15).

5.19.5. Synthesis of compound 40a

Acylation of 39 (3.00 g, 5.68 mmol) following the

general procedure aﬀorded, after evaporation of the

reaction solvent, a residue which was dissolved in dis-

tilled water and extracted with CH₂Cl₂. Evaporation of

the organic phases gave 40a as a yellow spongy volumi-

nous solid (2.54 g, 69%). TLC R_f=0.53 (CH₂Cl₂/

5.19.2. Synthesis of N,N⁰-bis(3-aminopropyl)cyclohexane-

1,2-diamine (23)

Compound 23 was obtained by reducing 19.0 g (86.4

mmol) of 22 with hydrogen at 50 psi using 95% EtOH/

1N NaOH as solvent and Raney-Ni (3 g) as catalyst.

The reduction was complete after 24 h. The mixture was

ﬁltered using Celite¹521 and concentrated by rotary

evaporation to ꢀ45 ml. A pale yellow viscous oil sepa-

rated on top and this was removed. The oil was then

puriﬁed by cc on silica gel (CH₂Cl₂/MeOH/

NH₄OH=10:4:0.5) to aﬀord 14.7 g (75%) of 23. TLC

1

MeOH=10:1); H NMR spectral data (CDCl₃) ꢀ 7.70

(d, 1H, J=16 Hz), 7.63 (d, 2H, J=8 Hz), 7.13 (d, 2H,

J=8 Hz), 6.43 (d, 1H, J=16 Hz), 3.33–3.17 (m, 6H),

2.31 (s, 3H, OCH₃), 1.77 (m, 6H), 1.45–1.43 (27H);

FAB-MS m/z 717 ([M+1]⁺, 2), 617 (5), 529 (65), 429

(18), 254 (13), 147 (13), 81 (11), 57 (100).

Deprotection of 40a (3.0 g, 4.2 mmol) was accom-

plished with HCl (2.15 g) in 50 ml of dry MeOH at 0 ^ꢂC.

Evaporation of the solvent, followed by washing with

diethyl ether and recrystallization from absolute EtOH

gave 830 mg (53%) of 40b. TLC R_f=0.10 (CH₂Cl₂/

MeOH=10:1); ¹H NMR (d₆-DMSO) ꢀ 7.40 (d, 2H,

J=8 Hz), 7.37 (d, 1H, J=16 Hz), 6.82 (d, 2H, J=8 Hz),

6.48 (d, 1H, J=16 Hz), 3.26 (3H), 2.92 (4H), 2.11 (6H),

1.82–1.72 (4H), 1.21 (2H); ¹³C NMR (d₆-DMSO) 165.9,

159.0, 139.0, 129.2, 125.8, 118.4, 115.8, 56.2, 48.6, 41.8,

36.2, 25.4, 23.7, 22.1 FAB-MS m/z 375 ([M+1]⁺,40),

325 (92), 316 (42), 229 (100), 155 (68), 111 (25), 98 (42),

81 (60), 57 (62). HRMS (FAB, NBA). Calculated mass:

375.2760 for C₂₁H₃₅N₄O₂, found: 375.2780.

1

R_f=0.25 (CH₂Cl₂/MeOH/NH₄OH=10:4:1); H NMR

spectral data (CDCl₃) ꢀ 4.76 (br s, 1H), 2.99 (br s, 1H),

2.75–2.71 (m, 6H), 2.48–2.44 (m, 2H), 2.09–2.05 (m,

4H), 1.70 (2H), 1.67–1.59 (4H), 1.57 (1H), 1.20 (3H),

1.17 (2H); ¹³C NMR spectral data (CDCl₃) 62.0, 44.9,

40.5, 34.5,31.9, 25.5; FAB-MS m/z 229 ([M+1]⁺, 100),

155 (25), 110 (10), 81 (13), 58 (11).

5.19.3. Synthesis of N-[3-({2-[(3-aminopropyl)amino]-

cyclohexyl}amino) propyl]-2,2,2- triﬂuoroacetamide

(24)

Mono-triﬂuoracetulation of 23 (5.00 g, 21.9 mmol)

following the general procedure gave a crude product

332

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

5.20. 2(E)-N-[3-({4-[(3-aminopropyl)amino]cyclohexyl}

amino) propyl]-3-(4-hydroxyphenyl)prop-2-enamide

(43b)

5.20.4. Synthesis of N-[3-({4-[(3-aminopropyl)amino]-

cyclohexyl}amino)propyl]-2,2,2-triﬂuoroacetamide (27)

Mono-triﬂuoracetylation of 26 (3.70 g, 16.2 mmol)

following the general procedure gave a crude product

which was puriﬁed on silica gel (CH₂Cl₂/MeOH/

AcOH=10:20:1) and provided 5.16 g, (98%) of 27.

TLC R_f=0.48 (CH₂Cl₂/MeOH/AcOH=60:40:2); ¹H

NMR spectral data (CDCl₃) ꢀ 3.1–6–3.12 (m, 12H), 2.52

(m, 2H), 2.23 (2H), 1.79–(4H), 1.58 (1H), 1.45–1.39

(2H), 0.92 (4H); ¹³C NMR spectral data (CDCl₃) 57.4,

57.2, 49.4, 45.4, 44.9, 34.7, 32.5, 29.4; FAB-MS m/z 325

([M+1]⁺, 8), 251 (22), 191 (13), 155 (21), 107 (29), 96

(21), 81 (50), 65 (100), 56 (42).

5.20.1. Synthesis of N,N⁰-bis(2-cyanoethyl)cyclohexane-

1,4-diamine (25)

trans-1,4-Diaminocyclohexane 13 (10.0 g, 87.6 mmol,

10.6 ml) was dissolved in MeOH (65 ml) and 2.5 eq. of

acrylonitrile (14.4 ml, 11.6 g) in MeOH (50 ml) was

added drop wise over 30 min with continuous stirring.

After stirring overnight at room temperature, evapora-

tion of the solvent followed by puriﬁcation by cc (silica,

2:1 CH₂Cl₂: MeOH) gave 9.1 g (99%) of 25 as oﬀ-white

crystals. TLC R_f=0.87 (CH₂Cl₂/MeOH/NH₄OH=

10:4:1); ¹H NMR spectral data (CDCl₃) ꢀ 2.98–2.93

(m, 4H), 2.53–2.49 (m, 6H), 1.97–1.95 (4H), 1.20–1.07

(m, 6H); ¹³C NMR spectral data (CDCl₃) 119.1, 56.3,

42.8, 32.3, 19.6; FAB-MS m/z 441 ([2M+1]⁺, 13),

221 ([M+1]⁺, 84), 180 (12), 151 (100), 109 (21), 81

(34).

5.20.5. Syntheses of compounds protected amines 41 and

42

27 (5.00 g, 15.4 mmol) was dissolved in MeOH (80

ml) and 3 eq. of (Boc)₂O (46.3 mmol, 10.1 g) in THF (40

ml) was added at room temp. The reaction was com-

plete as judged by TLC after 3 h and provided crude 41

that was used without any further puriﬁcation. In the

same ﬂask, 30 ml concentrated NH₄OH was added and

the solution reﬂuxed for 24 h to aﬀord 6.20 g (76%) of

triBoc protected amine, 42. TLC R_f=0.70 (CH₂Cl₂/

MeOH/NH₄OH=10:4:1); ¹H NMR spectral data

(CDCl₃) ꢀ 4.93 (br s, 1H), 3.35 (8H), 1.76 (4H), 1.65

(8H), 1.48 (13H), 1.45 (27H); FAB-MS m/z 529

([M+1]⁺, 61), 429 (100), 329 (3), 254 (11), 229 (8), 199

(15), 57 (68).

5.20.2. Synthesis of N,N⁰-bis(3-aminopropyl)cyclohexane-

1,4-diamine (26)

Compound 26 was obtained by reducing 6.00 g (27.3

mmol) of 26 with hydrogen at 50 psi for 24 h using 95%

EtOH/1 N NaOH as solvent and 2 g of Raney-Ni as

catalyst. When complete the mixture was ﬁltered

through Celite¹521 and concentrated to a volume of

about 35 ml. A pale, yellow viscous oil separated on

top of the mixture and this was removed. The oil was

then puriﬁed by cc on silica gel to aﬀord 5.60 g

(90%) of 27. TLC R_f=0.20 (CH₂Cl₂/MeOH/

NH₄OH=10:4:1); ¹H NMR spectral data (CDCl₃) ꢀ

3.16 (br s, 1H), 2.64–2.50 (m, 8H), 2.30 (m, 2H), 1.84–

1.82 (m, 4H), 1.54–1.46 (q, 4H), 1.09–0.96 (m, 6H); ¹³C

NMR spectral data (CDCl₃) 56.2, 43.7, 39.1, 32.3,30.8,

17.2; FAB-MS m/z 229 ([M+1]⁺, 100), 176 (19), 155

(22), 57 (28).

5.20.6. Synthesis of 43a

Acylation of 42 (3.00 g, 5.68 mmol) in CH₂Cl₂(40 ml)

following the general procedure provided, after evap-

oration of the reaction solvent, a residue which was

dissolved in distilled water and extracted with CH₂Cl₂.

The organic phase was washed with 0.25 M Na₂CO₃

and then dried with Na₂SO₄. Evaporation of the solvent

gave 43a as a yellow spongy solid (3.20 g, 87%). TLC

R_f=0.53 (CH₂Cl₂/MeOH=10:1); ¹H NMR spectral

data (CDCl₃) ꢀ 7.60 (d, 1H, J=16 Hz), 7.46 (d, 2H, J=8

Hz), 7.15 (d, 2H, J=8 Hz), 6.45 (d, 1H, J=16 Hz),

3.40–3.20 (m, 6H), 2.35 (s, 3H, OCH₃), 1.75 (m, 6H),

1.45–1.43 (27H); FAB-MS m/z 717 ([M+1]⁺, 2), 617

(6), 529 (29), 471 (8), 375 (8), 271 (12), 133 (23), 73 (73),

57 (100).

5.20.3. Synthesis of 26a primary amide nitrile of 26

During reduction of 25 at larger scale (i.e. 19.0 g, 86.4

mmol) using 3.6 g Raney-Ni, TLC (CH₂Cl₂/MeOH/

NH₄OH=10:4:1) showed a spot that migrated with an

R_f=0.36, between the dinitrile starting material 25

(R_f=0.60) and fully reduced amine product 26

(R_f=0.2). Recovery of this material by silica chromato-

graphy eluting by with the same solvent as for TLC

provided 1.50 g of a viscous oil that solidiﬁed on

Deprotection of 43a (3.0 g, 4.2 mmol) in dry MeOH

(40 ml) was accomplished by bubbling HCl (2.15 g)

ꢂ

through the solution at 0 C and stirring for 2 h. After

1

standing. H NMR spectral data (CDCl₃) ꢀ 7.02 (bs,

evaporation and recrystallization from EtOH, 43b was

obtained as a pale yellow solid (550 mg, 35%). TLC

R_f=0.15 (CH₂Cl₂/MeOH=10:1); ¹H NMR (d₆-DMSO)

ꢀ 7.40 (d, 2H, J=9 Hz), 7.36 (d, 1H, J=16 Hz), 6.82 (d,

2H, J=9 Hz), 6.47 (d, 1H, J=16 Hz), 4.10 (2H), 2.89 (m,

10H), 2.17 (4H), 1.99 (3H), 1.81 (1H), 1.46 (4H); ¹³C NMR

(d₆-DMSO) 165.3, 158.8, 138.4, 129.1, 125.9, 118.8, 115.7,

55.5, 55.1, 52.8, 52.7, 37.1, 26.5, 25.7 FAB-MS m/z 375

2H), 5.47 (bs, 2H) 2.95 (m, 2H), 2.53 (m, 4H), 2.25

(broad multiplet, 6H), 2.13 (m, 6H), 1.96 (m, 3H), 1.71

(m, 1H), 1.55 (bs, 4H), 0.75 (m, 4H). ¹³C NMR spectral

data (d₆-DMSO) 174.2, 120.5, 56.4, 55.9, 43.1, 42.4,

36.1, 31.8, 18.6, 18.2 FAB-MS m/z 239 ([M+1, 98],

225 (100), 180 (10), 169 (20), 167 (14), 151 (72), 149

(45), 110 (42).

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

333

([M+1]+, 57), 302 (13), 271 (100), 229 (54), 155 (67),

147 (35), 57 (72). HRMS (FAB, NBA). Calculated

mass: 375.2760 for C₂₁H₃₅N₄O₂, found: 375.2767.

Converso, D.A., Fernandez, M.E., 1995. Peroxidase isozymes from

wheat germ: puriﬁcation and properties. Phytochemistry 40, 1341–

1345.

Cu, C.H., Bahring, R., Mayer, M.L., 1998. The role of hydrophobic

interactions in binding of polyamines to non NMDA receptor ion

channels. Neuropharmacology 37, 1381–1391.

Dingledine, R., Borges, K., Bowie, D., Traynelis, S.F., 1999. The glu-

tamate receptor ion channels. Pharmacol. Rev. 51, 7–61.

Duﬀey, S.S., Stout, M.J., 1996. Antinutritive and toxic components of

plant defense against insects. Arch. Insect Biochem. Physiol. 32, 3–

37.

Acknowledgements

This work was funded by an NSERC Strategic Grant

(to JKA and AJM), NSERC Operating Grants to JKA

and AJM and National Institutes of Health grant

NS35047 from NINDS to KW. Tannis Mayert per-

formed the antifeedant trials with Sitophilus oryzae at

the Cereal Research Centre, Agriculture and Agri-Food

Canada, Winnipeg, Manitoba.

Hollmann, M., Heinemann, S., 1994. Cloned glutamate receptors.

Annu. Rev. Neurosci. 17, 31–108.

Hu, W.Q., Hesse, M., 1996. The synthesis of p-coumaroylspermidines.

Helv. Chim. Acta 79, 548–559.

Hume, R.I., Dingledine, R., Heinemann, S.F., 1991. Identiﬁcation of a

site in glutamate receptor subunits that controls calcium perme-

ability. Science 253, 1028–1031.

Igarashi, K., Koga, K., He, Y., Shimogori, T., Ekimoto, H., Kashi-

wagi, K., Shirahata, A., 1995. Inhibition of the growth of various

human and mouse tumor cells by 1,15-bis(ethylamino)-4,8,12-tri-

azapentadecane. Cancer Research 55, 2615–2619.

References

Abe, T., Kawai, N., Miwa, A., 1983. Eﬀects of a spider toxin on the

glutaminergic synapse of lobster muscle. J. Physiol. 339, 243–252.

Anis, N., Sherby, S., Goodnow Jr., R., Niwa, M., Konno, K.,

Kallimopoulos, T., Bukownik, R., Nakanishi, K., Usherwood, P.,

Eldefrawi, A., 1990. Structure–activity relationships of philantho-

toxin analogs and polyamines on N-methyl-d-aspartate and nico-

tinic acetylcholine receptors. J. Pharmacol. Exp. Ther. 254, 764–773.

Aonuma, H., Nagayama, T., Takahata, M., 1998. L-Glutamate as an

excitatory transmitter of motor giant neurons in the crayﬁsh Pro-

cambarus clarkii. J. Crustacean Biol. 18, 243–252.

Igarashi, K., Shirahata, A., Pahk, A.J., Kashiwagi, K., Williams, K.,

1997. Benzyl-polyamines: novel, potent N-methyl-d-aspartate

receptor antagonists. J. Pharmacol. Exp. Ther. 283, 533–540.

Karigiannis, G., Papaioannou, D., 2000. Structure, biological activity

and synthesis of polyamine analogues and conjugates. Eur. J. Org.

Chem. 1841–1863.

Kashiwagi, K., Masuko, T., Nguyen, C.D., Kuno, T., Tanaka, I.,

Igarashi, K., Williams, K., 2002. Channel blockers acting at N-

methyl-d-aspartate receptors: diﬀerential eﬀects of mutations in the

vestibule and ion channel pore. Mol. Pharmacol. 61, 533–545.

Klose, M.K., Atkinson, J.K., Mercier, A.J., 2002. Eﬀects of a

hydroxy-cinnamoyl conjugate of spermidine on arthropod neuro-

muscular junctions. J. Comp. Physiol. A. 187, 945–952.

Krapcho, A.P., Kuell, C.S., 1990. Mono-protected diamines. N-tert-

butoxycarbonyl-a,o-alkanediamines from a,o-alkanediamines.

Synth. Commun. 20, 2559–2564.

Atwood, H.L., 1982. Synapses and neurotransmitters. In: Atwood,

H.L., Sandeman, D.C. (Eds.), The Biology of Crustacea. Academic

Press, New York, pp. 105–150.

Benson, J.A., Kaufmann, L., Hue, B., Pelhate, M., Schurmann, F.,

Gsell, L., Piek, T., 1993. The physiological action of analogues of

philanthotoxin-4.3.3 at insect nicotinic acetylcholine receptors.

Comp. Biochem. Physiol. C 105, 303–310.

Li, Y., Popaj, K., Lochner, M., Geneste, H., Budriesi, R., Chiarini, A.,

Melchiorre, C., Hesse, M., 2001. Analogues of polyamine alkaloids

and their synthetic advantages. Farmaco 56, 127–131.

Benson, J.A., Schurmann, F., Kaufmann, L., Gsell, L., Piek, T., 1992.

Inhibition of dipteran larval neuromuscular synaptic transmission

by analogues of philanthotoxin-4.3.3: a structure–activity study.

Comp. Biochem. Physiol. C 102, 267–272.

Makinen, K.K., Tenovuo, J., 1982. Observations on the use of guaia-

col and 2,2⁰-azino-di(3-ethylbenzthiazoline-6-sulfonic acid) as per-

oxidase substrates. Anal. Biochem. 126, 100–108.

Blagbrough, I.S., Brackley, P.T.H., Bruce, M., Bycroft, B.W., Mather,

A.J., Millington, S., Sudan, H.L., Usherwood, P.N.R., 1992. Arthro-

pod toxins as leads for novel insecticides. Toxicon 30, 303–322.

Blagbrough, I.S., Geall, A.J., 1998. Practical synthesis of unsymmet-

rical polyamine amides. Tetrahedron Lett. 39, 439–442.

McCormick, K.D., Meinwald, J., 1993. Neurotoxic acylpolyamines

from spider venoms. J Chem. Ecol. 19, 2411–2451.

Mercier, A.J., Atwood, H.L., 1989. Long-term adaptation of a phasic

extensor motoneurone in crayﬁsh. J. Exp. Biol. 145, 9–22.

Mercier, A.J., Farragher, S., Schmor, B., Kamau, M., Atkinson, J.,

1998. Eﬀect of a plant-derived spider toxin analogue on the crayﬁsh

neuromuscular junctions. Can. J. Zool. 76, 2103–2107.

Meurer-Grimes, B., 1995. New evidence for the systematic signiﬁcance

of acylated spermidines and ﬂavonoids in pollen of higher hama-

melidae. Brittonia 47, 130–142.

Blagbrough, I.S., Moya, E., Taylor, S., 1994. Polyamines and polyamine

amides from wasps and spiders. Biochem. Soc. Trans. 22, 888–893.

Blagbrough, I.S., Usherwood, P.N.R., 1990. Mode of action of poly-

amine-derived glutamate antagonists. Pest. Sci. 30, 451–453.

Blagbrough, I.S., Usherwood, P.N.R., 1992. Polyamine amide toxins

as pharmacological tools and pharmaceutical agents. Proc. R. Soc.

Edinburgh 99B, 67–81.

Moya, E., Blagbrough, I.S., 1995. Synthesis and neuropharmaco-

logical properties of arthropod polyamine amide toxins. In: Carter,

C. (Ed.), Neuropharmacology of Polyamines, pp. 167–183.

Osullivan, M.C., Dalrymple, D.M., 1995. A one-step procedure for

the selective triﬂuoroacetylation of primary amino groups of poly-

amines. Tetrahedron Lett. 36, 3451–3452.

Bokern, M., Witte, L., Wray, V., Nimtz, M., Meurer-Grimes, B.,

1995. Trisubstituted hydroxycinnamic acid spermidines from Quer-

cus dentat pollen. Phytochemistry 39, 1371–1375.

Bruce, M., Bukownik, R., Eldefrawi, A.T., Eldefrawi, M.E., Good-

now Jr., R., Kallimopoulos, T., Konno, K., Nakanishi, K., Niwa,

M., Usherwood, P.N., 1990. Structure–activity relationships of

analogues of the wasp toxin philanthotoxin: non-competitive

antagonists of quisqualate receptors. Toxicon 28, 1333–3346.

Chao, J., Seiler, N., Renault, J., Kashiwagi, K., Masuko, T., Igarashi, K.,

Williams, K., 1997. N1-dansyl-spermine and N1-(n-octanesulfonyl)-

spermine, novel glutamate receptor antagonists: block and permeation

of N-methyl-d-aspartate receptors. Mol. Pharmacol. 51, 861–871.

Panagabko, C., Chenier, D., Fixon-Owoo, S., Atkinson, J.K., 2000.

Ion-pair HPLC determination of hydroxycinnamic acid mono-

conjugates of putrescine, spermidine and spermine. Phytochem.

Anal. 11, 11–17.

Panchenko, V.A., Glasser, C.R., Mayer, M.L., 2001. Structural simi-

larities between glutamate receptor channels and K(+) channels

334

S. Fixon-Owoo et al. / Phytochemistry 63 (2003) 315–334

examined by scanning mutagenesis. J. Gen. Physiol. 117, 345–

360.

Tikhonov, D.B., Magazanik, L.G., Mellor, I.R., Usherwood, P.N.,

2000. Possible inﬂuence of intramolecular hydrogen bonds on the

three-dimensional structure of polyamine amides and their interac-

tion with ionotropic glutamate receptors. Receptors Channels 7,

227–236.

Parnas, I., Atwood, H.L., 1966. Phasic and tonic neuromuscular sys-

tem in the abdominal extensor muscle of the crayﬁsh and rock lob-

ster. Comp. Biochem. Physiol. 18, 701–723.

Quicke, D.L.J., Usherwood, P.N.R., 1990. Spider toxins as lead

structures for novel pesticides. In: Hodgson, E., Kuhr, R.J. (Eds.),

Safer Insecticides: Development and Use. Marcel Dekker, New

York, pp. 385–452.

Torrigiani, P., Altamura, M.M., Pasqua, G., Monacelli, B., Seraﬁni-

Fracassini, D., Bagni, N., 1987. Free and conjugated polyamines

during de novo ﬂoral and vegetative bud formation on thin cell lay-

ers of tobacco. Physiol. Plant 70, 453–460.

Raditsch, M., Geyer, M., Kalbitzer, H.R., Jahn, W., Ruppersberg,

J.P., Witzemann, V., 1996. Polyamine spider toxins and mammalian

N-methyl-d-apsartate receptors: structural basis for channel block-

ing and binding of argiotoxin(636). Eur. J. Biochem. 240, 416–426.

Schafer, A., Benz, H., Fiedler, W., Guggisberg, A., Bienz, S., Hesse,

M., 1994. Polyamine toxins from spiders and wasps. In: Cordell,

G.A., Brossi, A. (Eds.), Alkaloids, Vol. 45. Academic Press, San

Diego, pp. 1–124.

Usherwood, P.N.R., Blagbrough, I.S., 1991. Spider toxins aﬀecting

glutamate receptors: polyamines in therapeutic neurochemistry.

Pharmacol. Ther. 52, 245–268.

van Harreveld, A., 1936. A physiological solution for freshwater

crustaceans. Proc. Soc. Exp. Biol. Med. 34, 428–432.

Werner, C., Hu, W., Lorenzi-Riatsch, A., Hesse, M., 1995. Dicou-

maroylspermidines in anthers of diﬀerent species of the genus

Aphelandra. Phytochemistry 40, 461–465.

Sharma, P., Rajam, M.V., 1995. Spatial and temporal changes in

endogenous polyamine levels associated with somatic embryogen-

esis from diﬀerent hypocotyl segments of eggplant (Solanum mel-

ongena). J. Plant Physiol. 146, 658–664.

Williams, K., 1993a. Eﬀects of Agelenopsis aperta toxins on the

N-methyl-d-aspartate receptor- polyamine-like and high-aﬃnity

antagonist actions. J. Pharmacol. Exp. Therapeut. 266, 231–236.

Williams, K., 1993b. Ifenprodil discriminates subtypes of the

N-methyl-d-aspartate receptor: selectivity and mechanisms at

recombinant heteromeric receptors. Mol. Pharmacol. 44, 851–859.

Williams, K., Dattilo, M., Sabado, T., Kashiwagi, K., Igarashi, K.,

2003. Pharmacology of delta2 glutamate receptors: eﬀects of penta-

midine ad protons. J. Pharmacol. Exp. Ther. 305, in press.

Williams, K., Russell, S.L., Shen, Y.M., Molinoﬀ, P.B., 1993. Devel-

opmental switch in the expression of NMDA receptors occurs in

vivo and in vitro. Neuron 10, 267–278.

Shinozaki, H., Shibuya, I., 1974. A new potent excitant, quisqualic

acid: eﬀects on crayﬁsh neuromuscular junction. Neuropharmacol.

13, 665–672.

Shudo, K., Endo, Y., Hashimoto, Y., Aramaki, Y., Nakajima, T.,

Kawai, N., 1987. Newly synthesized analogues of the spider toxin

block the crustacean glutamate receptor. Neurosci. Res. 5, 82–85.

Smith, T.A., Negrel, J., Bird, C.R., 1983. The cinnamic acid amides of

the di- and polyamines. In: Bachrach, U. (Ed.), Advances in Poly-

amine Research. Raven Press, New York, pp. 347–370.

Strømgaard, K., Brier, T.J., Andersen, K., Mellor, I.R., Saghyan, A.,

Tikhonov, D., Usherwood, P.N., Krogsgaard-Larsen, P., Jar-

oszewski, J.W., 2000. Solid-phase synthesis and biological eval-

uation of a combinatorial library of philanthotoxin analogues. J.

Med. Chem. 43, 4526–4533.

Wollmuth, L.P., Kuner, T., Jatzke, C., Seeburg, P.H., Heintz, N.,

Zuo, J., 2000. The Lurcher mutation identiﬁes delta 2 as an AMPA/

kainate receptor-like channel that is potentiated by Ca(2+). J.

Neurosci. 20, 5973–5980.

Xie, Y.S., Bodnaryk, R.P., Fields, P.G., 1996. A rapid and simple

ﬂour-disk bioassay for testing substances active against stored-pro-

duct insects. Can. Ent. 128, 865–875.

Subramaniam, S., Donevan, S.D., Rogawski, M.A., 1994. Hydro-

phobic interactions of n-alkyl diamines with the N-methyl-d-aspar-

tate receptor: voltage-dependent and-independent blocking sites.

Mol. Pharmacol. 45, 117–124.

Zuo, J., De Jager, P.L., Takahashi, K.A., Jiang, W., Linden, D.J.,

Heintz, N., 1997. Neurodegeneration in Lurcher mice caused by

mutation in delta2 glutamate receptor gene. Nature 388, 769–773.

Article Doi

Preparation and biological assessment of hydroxycinnamic acid amides of polyamines

DOI: 10.1016/S0031-9422(03)00133-X

Source and publish data:

Authors:

Article abstract of DOI:10.1016/S0031-9422(03)00133-X

Full text of DOI:10.1016/S0031-9422(03)00133-X

Products guided by the article

R&D Labs maybe for 660426-60-8

Relevant to this article

Hot Product