The Effect of Polyamine Homologation on the Transport and Cytotoxicity Properties of Polyamine-(DNA-Intercalator) Conjugates

Article abstract of DOI:10.1021/jo0002792

An efficient five-step synthetic method was developed to access a homologous series of spermidine-acridine and spermidine-anthracene conjugates. The derivatives were comprised of a spermidine fragment covalently tethered at its N4 position to either an ac

Full text of DOI:10.1021/jo0002792

5590
J . Org. Chem. 2000, 65, 5590-5599
Th e Effect of P olya m in e Hom ologa tion on th e Tr a n sp or t a n d
Cytotoxicity P r op er ties of P olya m in e-(DNA-In ter ca la tor )
Con ju ga tes
Otto Phanstiel, IV,* Harry L. Price, Lu Wang, J ane J uusola, Martin Kline, and
Sapna Majmundar Shah
Center for Discovery of Drugs and Diagnostics, Department of Chemistry, University of Central Florida,
Orlando, Florida 32816-2366
ophansti@mail.ucf.edu
Received February 29, 2000
An efficient five-step synthetic method was developed to access a homologous series of spermidine-
acridine and spermidine-anthracene conjugates. The derivatives were comprised of a spermidine
fragment covalently tethered at its N4 position to either an acridine or anthracene nucleus via an
aliphatic chain (e.g., spermidine-[aliphatic tether]-acridine). The distance separating the sper-
midine and aromatic nucleus was altered by using different tethers comprised of four or five
methylene units, respectively. These ligands (2-5) were shown to inhibit human DNA topo-
isomerase-II (TOPO-II) activity at 10 µM. Enzymatic activity was assessed as the ability to unknot
(decatenate) and cleave kinetoplast DNA (kDNA). Polyamine conjugation did not disrupt the ability
of the acridine-spermidine conjugates 2 and 3 to inhibit TOPO-II activity as compared with the
9-aminoacridine and 9-(N-butyl)aminoacridine controls (at 10 µM). In general, the acridine
derivatives (2 and 3) showed higher TOPO-II inhibitory activity than their anthracene counterparts
(4 and 5). However, this trend was reversed in a whole cell assay with L1210 (murine leukemia)
cells, wherein the anthracene analogues were more potent than their acridine counterparts. In
this regard the qualitative enzyme-based assay did not predict the trends in the corresponding
IC50 values. Within either series insertion of an additional methylene unit did not significantly
alter activity. While the appended spermidine unit did not disrupt TOPO II inhibition by the tethered
DNA intercalator, it did provide an alternative mode of entry into the cell as demonstrated by
spermidine protection assays. These results were compared with a spermine-intercalator analogue.
Of all the conjugates tested the N4-(4-(9-aminoacridinyl)butyl)spermine hexahydrochloride (conjugate
16)resulted in the highest degree of L1210 cell rescue upon cotreatment of the cells with exogenous
spermidine. It was concluded that the monoalkylated spermine motif present in 16 holds promise
as a better vector than its N4 monoalkylated spermidine counterpart.
In tr od u ction
mentary polynucleotide strands and ultimately the gen-
eration of relaxed supercoils (both of which are important
events for DNA replication and RNA transcription).
In eukaryotic cells, DNA is segmentally unfolded
within the nucleus by enzymes called topoisomerases.1
Topoisomerase type I (TOPO-I) and type II (TOPO-II)
represent two classes of the known mammalian DNA
topoisomerases. To untwist densely packed DNA, these
enzymes generate transient breaks within the DNA
strands, allow for topological changes to occur and then
reseal the break.1 TOPO-I and TOPO-II act by creating
temporary single-strand and double-strand breaks in
DNA, respectively. Presently, two equilibrating com-
plexes (“cleavable” and “noncleavable”) have been shown
to exist between topoisomerase II and DNA.2 The equi-
librium is normally shifted toward the “noncleavable”
complex as the “cleavable” form results in permanent
DNA strand breaks.3,4 The “noncleavable” complex allows
for the cleaved strands to untwist and to be reannealed,
which in turn allows for the local separation of comple-
An anti-cancer strategy suggested by Nelson et al.5
involves perturbing this equilibrium toward the “cleav-
able” complex, an event that results in permanent DNA
strand breaks and cell death. This strategy was realized
with amsacrine (AMSA, 1), a drug currently used in the
treatment of acute nonlymphocytic leukemia (see Figure
1).2 In particular, TOPO-II has been identified as the
primary target of 1 in vitro,4 and AMSA has been shown
to give similar DNA strand breaks in both cultured
mammalian cells and isolated nuclei.3,4 Nelson and others
have suggested that AMSA shifts the aforementioned
equilibrium toward the “cleavable” complex by forming
a ternary complex of AMSA/topoisomerase II/ and DNA.5
This AMSA-stabilized complex is the putative cause of
cytotoxicity of this weak DNA intercalator.5,6 Other
molecules, which stimulate the formation of the “cleav-
able” complex, would be of clear value in terms of
(1) Morgan, D. M. Biochem. Soc. Transactions 1990, 18, 1080-1084.
(2) Zwelling, L. A.; Michaels, S.; Erickson, L. C.; Ungerleider, R. S.;
Nichols, M.; Kohn, K. W. Biochemistry, 1981, 20, 6553-6563.
(3) Zwelling, L. A.; Kerrigan, D.; Micheals, S. Cancer Res. 1982, 42,
2687-2691.
(4) Pommier, Y.; Kerrigan, D.; Schwartz, R.; Zwelling, L. A. Biochem.
Biophys. Res. Commun. 1982, 77, 1150-1157.
(5) Nelson, E. M.; Tewey, K. M.; Liu, L. F. Proc. Natl. Acad. Sci.
U.S.A. 1984, 81, 1361-1365.
(6) Denstman, S. C.; Ervin, S. J .; Casero, R. A. Biochem. Biophys.
Res. Commun. 1987, 149, No. 1, 194-202.
10.1021/jo0002792 CCC: $19.00 © 2000 American Chemical Society
Published on Web 08/16/2000

Effects of Polyamine Homologation
J . Org. Chem., Vol. 65, No. 18, 2000 5591
F igu r e 1. Amsacrine (AMSA) and 9-(amino)acridine.
developing new chemotherapeutics and further elucidat-
ing the specific mechanisms involved in the DNA cleaving
process.5
It seemed possible that structural modifications of 1
would further enhance the TOPO-II inhibitory activity
and the anti-proliferative properties of AMSA. In design-
ing new ligands, we recognized that the intercalative
association of 1 with DNA was likely attributed to the
9-aminoacridine component of AMSA (Figure 1). Indeed,
many linear tricyclic systems are widely accepted as
efficient DNA intercalators.7,8 Mere intercalation of these
ligand types, however, is not sufficient for anti-neoplastic
activity as shown in a series of substituted acridines.8
After an extensive investigation, Denny and co-workers
found that both DNA intercalation and an appropriately
placed side chain appeared to be an absolute requirement
for antitumor activity with the acridine systems studied.8
We were interested in whether a DNA intercalator
with an appended polyamine side chain such as spermi-
dine (SPD, H2N(CH2)3NH(CH2)4NH2) or spermine (SPM,
H2N(CH2)3NH(CH2)4NH(CH2)3NH2) would show altered
transport behavior as the conjugate may interact with
the polyamine transporter (PAT) present in cancer cells
lines.9 Such an interaction was expected to provide an
alternative mode of entry into the cell and possibly
provide a more selective chemotherapeutic. Indeed, such
polyamine-DNA intercalator conjugates may act as dual
vector systems, which target the DNA of rapidly dividing
cell types.9 The design of conjugates 2-5 (structures
shown in Figure 2) is predicated upon the well-known
affinity of polyamines for DNA and the established DNA
binding modes of acridines5,9 and anthracenes.10 Fur-
thermore, DNA intercalation models predict that a
multitude of anionic site charges are available for binding
of additional basic functions beyond those present in the
intercalator framework.11 Since polyamines exist as poly-
cations in vivo, the conjugates (e.g., 2) which have an
appended polycationic tail may bind to these available
sites thereby increasing the binding affinity and potency
of the intercalator. Therefore, the combination of these
agents (polyamines and DNA intercalators) were ex-
pected to provide new DNA targeting vectors, which
significantly disrupt TOPO-II activity. This report out-
lines the synthesis and biological evaluation of several
conjugates of this type.
F igu r e 2. Acridine- and anthracene-based conjugates 2-5,
Sjo¨berg’s conjugate A, and Cohen’s conjugate B.
active agents to augment their activity and specificity.9,12-19
Earlier investigations have shown that one can utilize
polyamine conjugation to convey cytotoxic drugs (e.g.,
Figure 2: A and B) into rapidly growing cells.9,12,13 During
our biological evaluation of conjugates 2-5 Sjo¨berg and
co-workers published their synthesis of conjugate A, a
potential candidate for boron neutron capture therapy
(BNCT).9 In another example, Cohen and others found
that a chloroambucil-spermidine conjugate B was 10 000-
fold more effective in forming intrastrand cross-links with
naked DNA. In vivo studies with plasmacytoma cells in
mice showed that the polyamine conjugate was 4-fold
more potent in inhibiting tumor growth than the uncon-
jugated chloroambucil. They concluded that the conjugate
B utilized the polyamine uptake apparatus as evidenced
by its low Ki (vs spermidine) and the fact that exogenous
(12) Cohen, G. M.; Cullis, P.; Hartley, J . A.; Mather, A.; Symons,
M. C. R.; Wheelhouse, R. T. J . Chem. Soc., Chem. Commun. 1992, 298-
300.
(13) Holley J . L.; Mather, A.; Wheelhouse, R. T.; Cullis, P. M.;
Hartley, J . A.; Bingham J . P.; Cohen, G. M. Cancer Res. 1992, 52,
4190-4195.
In this regard, there has been recent interest in
tethering polyamines to existing cancer drugs and bio-
(14) Blagbrough, I. S.; Geall, A. J . Tetrahedron Lett. 1998, 439-
442.
(15) Blagbrough, I. S.; Taylor, S.; Carpenter, M. L.; Novoselskiy, V.;
Shamma, T.; Haworth, I. S. Chem. Commun. 1998, 929-930.
(16) Fenniri, H.; Hosseini, M. W.; Lehn, J .-M. Helv. Chim. Acta 1997,
80, 786-803.
(7) Albert, A. The Acridines, 2nd ed.; Edward Arnold: London, 1966.
(8) Atwell, G. J .; Cain, B. F.; Baguley, B. C.; Finlay, G. J .; Denny,
W. A. J . Med. Chem. 1984, 27, 1481-1485.
(9) Ghaneolhosseini, H.; Tjarks, W.; Sjoberg, S. Tetrahedron 1998,
54, 3877-3884.
(17) Haensler, J .; Szoka, F. C., J r. Bioconjugate Chem. 1993, 4, 85-
93.
(10) Kumar, C. V.; Asuncion, E. H. J . Am. Chem. Soc. 1993, 115,
8547-8553.
(18) Rodger, A.; Taylor, S.; Adlam, G.; Blagbrough, I. S.; Haworth,
I. S. Bioorg. Med. Chem. 1995, 3(6), 861-872.
(19) Geall, A. J .; Blagbrough, I. S. Tetrahedron Lett. 1998, 39, 443-
446.
(11) Atwell, G. J .; Cain, B. F.; Denny, W. A. J . Med. Chem. 1977,
20, 1128-1134.

5592 J . Org. Chem., Vol. 65, No. 18, 2000
Phanstiel et al.
Sch em e 1a
Sch em e 2a
a
Key: (a) BOC-ON, THF, 0 °C; (b) Br-CH2(CH2)nCH2CN,
refluxing acetonitrile, solid K2CO3; (c) H2 gas (75 psi), NH4OH,
Raney Ni (50% slurry), absolute EtOH saturated with NH3 gas.
spermidine reduced the toxicity of the conjugate.12,13 The
present study also demonstrated that polyamine conju-
gates can also function as vectors,19 which assist co-
valently attached components into cells.
a
Key: (a) 9-chloroacridine, molten phenol, 50 °C; (b) 4 N HCl;
(c) 9-anthraldehyde, 4 Å molecular sieves followed by NaBH4.
nitriles 7 and 8 in 92% and 86% yield, respectively. Raney
nickel reduction of 7 and 8 in the presence of ammonia
under a high pressure of H2 gave the respective amines
9 (40%) and 10 (74%).
Resu lts a n d Discu ssion
Syn th esis. The synthesis of each conjugate involves
a series of protection and deprotection steps necessary
for the synthesis of substituted polyamines and acri-
dines.11,20-25 In each case, spermidine was conjugated at
the N4 position via an aliphatic carbon tether to either
an acridine or anthracene nucleus. This central attach-
ment was predicated upon the findings by Porter in 1982,
wherein spermidine could be extensively derivatized at
the central N4 position and still be taken up by the
polyamine transporter.23 The N4-alkylation step was
designed to maintain the basicity of the N-4 nitrogen,
which was also shown to be critical to uptake.12,13,23
Each conjugate (2-5) required the generation of a
functionalized spermidine fragment with selectively pro-
tected amino groups. As shown in Scheme 1, the synthe-
sis began with the selective acylation of the primary
amino groups of spermidine with 2-(tert-butoxycarbony-
loxyimino)-2-phenylacetonitrile (BOC-ON) to give the
known N1,N8-bis(tert-butoxycarbonyl)spermidine 6 in
79% yield.9,24,25 N4-Alkylation of 6 with either 4-bromobu-
tyronitrile or 5-bromovaleronitrile gave the corresponding
The syntheses of the acridine conjugates 2 and 3 are
shown in Scheme 2. The respective amines 9 and 10 were
coupled to 9-chloroacridine using an excess of molten
phenol to give 11 and 12 in 81% and 78% yield, respec-
tively. The mechanism of the coupling reaction was
shown to go through a 9-phenoxyacridine intermediate
13.9,26a,27 This finding was confirmed by isolating 13 and
demonstrating its facile reaction with amine 9 to give
the acridine derivative 11.26b The BOC protecting groups
on the spermidine-acridine conjugates 11 and 12 were
then deprotected using 4 N HCl to give the respective
HCl salts (2 and 3) in 79% yield for each conjugate.
In addition to the acridine derivatives, two anthracene-
spermidine conjugates (4 and 5) were synthesized. As
shown in Scheme 2, the respective amines 9 and 10 were
condensed with 9-anthraldehyde and after imine reduc-
tion with NaBH4 afforded the tethered amine adducts 14
(66%) and 15 (68%), respectively. The BOC protecting
groups of 14 and 15 were then deprotected using 4 N HCl
to give the HCl salt of the desired conjugates 4 (91%)
and 5 (93%), respectively.
(20) Cain, B. F.; Seelye, R. N.; Atwell, G. J . J . Med. Chem. 1974,
17, 922-930.
For comparison purposes, the monoalkylated spermine
(SPM) derivative 16 (Scheme 3) and 9-(N-butylamino)-
acridine 17 (Figure 3) were also synthesized. Using
(21) Atwell, G. J .; Cain, B. F. J . Med. Chem. 1974, 17, 930-934.
(22) Cain, B. F.; Baguley, B. C.; Denny, W. A. J . Med. Chem. 1978,
21, 658-668.
(23) Porter, C. W.; Bergeron, R. J .; Stolowich, N. J . Cancer Res. 1982,
42, 2, 4072-4078.
(26) (a) Dupre, D. J .; Robinson, F. A. J . Chem. Soc. 1945, 549-551.
(b) For an alternative method using 9-(1,2,4-triazole) intermediates,
see: Carlson, C. B.; Beal, P. A. Org. Lett. 2000, 2(10), 1465-1468.
(27) Kohn, K. W.; Orr, A.; O’Connor, P. M.; Guziec, L. J .; Guziec, F.
S., J r. J . Med. Chem. 1994, 37, 67-72 and references therein.
(24) Lurdes, M.; Almeida, S.; Grehn, L.; Ragnarsson, U. J . Chem.
Soc., Chem. Commun. 1987, 1250-1251.
(25) Lurdes, M.; Almeida, S.; Grehn, L.; Ragnarsson, U. J . Chem.
Soc., Perkin Trans. 1 1988, 1905-1911.

Effects of Polyamine Homologation
J . Org. Chem., Vol. 65, No. 18, 2000 5593
Sch em e 3a
F igu r e 4. Topoisomerase II inhibition studies. Decatenation
assay results are displayed as a negative image of the
electrophoresis gel. All lanes contain 0.5 µg of kDNA and 4.0
units of human TOPO-II, with the exception of Lane 1. The
respective drugs and conjugates were evaluated at 10 µM
(panel A) and 20 µM (panel B). Lane 1: negative control
(kDNA only). Lane 2: positive control (TOPO-II). Lane 3:
acridine conjugate 2. Lane 4: acridine conjugate 3. Lane 5:
anthracene conjugate 4. Lane 6: anthracene conjugate 5. Lane
7: berenil. Lane 8: 9-aminoacridine 23. Lane 9: 9-(N-
butylamino)acridine 17. Lane 10: spermidine. Forms of DNA
are denoted as Roman numerals: (intact kDNA (I, top); nicked
circular DNA II, middle); and relaxed circular DNA (III,
bottom).
DNA (kDNA) has limited migratory abilities and remains
near the top well (form I), but the lower molecular weight
decatenated forms [i.e., nicked-circular (form II) and
relaxed DNA (form III)] migrate further down the gel. A
potent TOPO-II inhibitor is expected to completely
abrogate the decatenation process and to leave (after
electrophoresis) significant amounts of kDNA remaining
in the well.
a
Key: (a) 1 equiv of 4-bromobutyronitrile, K2CO3, refluxing
Decatenation reactions were carried out in the presence
of the conjugates 2-5 (at 10 and 20 µM, respectively).
For comparison, reactions were also carried out with
equivalent concentrations of berenil 22 (a.k.a. diminazene
aceturate, a known inhibitor of TOPO-II),33,34 9-ami-
noacridine 23, 9-(N-butylamino)acridine 17, and spermi-
dine (SPD), respectively (see structures in Figures 1 and
3). The results of these TOPO-II experiments are shown
in Figure 4 (10 µM, panel A; 20 µM, panel B). As shown
in Figure 4, conjugates 2-5 (lanes 3-6) clearly interfere
with the (TOPO II-mediated) decatentation of kDNA in
a concentration dependent manner. Independent of con-
centration (i.e., 10 or 20 µM), the acridine conjugates 2
and 3 exhibit significantly greater inhibitory activity than
the respective anthracene derivatives 4 and 5. This
difference is most readily observed by the significant
amount of kDNA remaining in the well in Figure 4 (panel
A, lanes 3 and 4 versus lanes 5 and 6). Another apparent
difference is the ladder of bands (i.e., partial kDNA
networks) observed between forms I and II. These bands
are indicative of diminished TOPO-II activity, wherein
only partial decatenation activity takes place. Complete
inhibition of TOPO-II activity should result in the
absence of these ladder bands (as all the kDNA remains
intact). Notably, the amount of these partial kDNA
networks is significantly reduced upon increasing the
concentration of the anthracene conjugates to 20 µM (see
panel B, lanes 5 and 6). The enhanced activity of the
acridine conjugates (versus the anthracene derivatives)
is consistent with the higher binding constants which
have been observed with acridines and DNA.10,18,35,36
CH3CN; (b) Raney Ni, concentrated NH4OH, H2 (75 psi); (c)
9-chloroacridine, molten phenol, 100 °C; (d) 4 N HCl, 1,4-dioxane.
F igu r e 3. Structures of 9-(N-butyl)aminoacridine 17 and
berenil 22.
similar methods, the bis BOC-protected spermine deriva-
tive 1828 was monoalkylated to its corresponding nitrile
19 (73%), which was then reduced to amine 20 (95%) and
coupled with 9-phenoxyacridine to give acridine 21 (70%),
which was then deprotected to give conjugate 16 as its
HCl salt (72%). Compound 17 was synthesized from
9-chloroacridine, phenol, and N-butylamine in 72% yield.
Biologica l Eva lu a tion . The spermidine conjugates
2-5 were evaluated for their ability to inhibit human
DNA topoisomerase II activity. Enzymatic activity was
assessed as the ability to unknot (decatenate) and cleave
Crithidia fasciculata kinetoplast DNA (kDNA).29-31
A
simple electrophoretic assay32 allowed for the visualiza-
tion of the reaction products (see the Experimental
Section). Due to its high molecular weight, catenated
(28) Adamczyk, M.; Fishpaugh, J . R.; Heuser, K. J . Org. Prep. Proc.
Int. 1998, 30, 339-348.
(29) Shapiro, T. A.; Klein, V. A.; Englund, P. T. J . Biol. Chem. 1989,
264, 4173-4178.
(30) kDNA and TOPO-II assay kits can be purchased from Topo-
GEN, Inc.
(31) Muller, M. T.; Helal, K.; Soisson, S.; Spitzner, J . R. Nucl. Acid
Res. 1989, 17, 9499.
(32) Zwelling, L. A.; Hinds, M.; Chan, D.; Mayes, J .; Sie, K. L.;
Parker, E.; Silberman, L.; Radcliffe, A.; Beran, M.; Blick, M. J . Biol.
Chem. 1989, 264, 16411-16420.
(33) Pearl, L. H.; Skelly, J . V.; Hudson, B. D.; Neidle, S. Nucl. Acids
Res. 1987, 15, 3469-3477.
(34) Yoshida, M.; Banville, D. L.; Shafer, R. H. Biochemistry, 1990,
29, 6585-6592.
(35) Denny, W. Anti-Cancer Drug Des. 1989, 4, 241-263.

5594 J . Org. Chem., Vol. 65, No. 18, 2000
Ta ble 1. IC50 Va lu es (µM)
Phanstiel et al.
relate potency to uptake by the polyamine transporter,
we conducted several spermidine protection experiments.
An a lysis of P olya m in e Con ju ga te Tr a n sp or t. Are
the polyamine-DNA intercalator conjugates recognized
and transported by the polyamine uptake apparatus of
L1210 cells? As mentioned previously, the observation
by Cohen et al.13 that a chloroambucil-spermidine
conjugate B (Figure 2) has a low Ki value in competition
experiments involving radiolabeled spermidine suggested
the possibility that some of the conjugates may also be
recognized by the polyamine transporter (PAT).
compd
IC50 (47 h)
IC50 (94 h)
2 (SPD C4 acridine)
3 (SPD C5 acridine)
4 (SPD C4 anthracene)
5 (SPD C5 anthracene)
16 (SPM mono C4 acridine)
17 9-(N-butylamino)acridine
75
>100
13
55
>100
6
5
17
2
7
23
2
Equally important results were obtained by carrying
out the decatenation reactions in the presence of the
minor-groove binder berenil 22 (lane 7) and the known
DNA intercalator, 9-aminoacridine 23 (lane 8).35 Other
controls included derivative 17 (lane 9), which represents
a molecular fragment of conjugate 2 without the polyamine
moiety, and spermidine alone (lane 10).
Several polyamine derivatives were examined. Two
controls, dimethylspermidine (DMSPD) and diethylsper-
midine (DESPD) have been shown to gain entry into the
cell via the PAT.38 Both of these derivatives are toxic to
L1210 cells, and kinetic assays have revealed that these
SPD analogues inhibited PAT-mediated transport of SPD
via a competitive mechanism.38 In terms of their respec-
tive affinity for PAT, the Ki for inhibition of SPD uptake
for DMSPD and DESPD has been reported to be 5 and
19 µM, respectively.38 Based on these Ki values, the
approximate Km for DMSPD and DESPD transport via
the PAT is 2.5 and 9.5 µM, respectively.38 The cellular
response to DMSPD and DESPD is death, which results
from polyamine depletion.38 Since the mode of SPD
inhibition is competitive, and an excess of SPD is present
in the system, it is expected that transport of both
DMSPD and DESPD will be reduced by more than 90%,
which in turn will significantly reduce cell death. There-
fore, inhibition of PAT-mediated transport of DMSPD and
DESPD will be observed as a decrease in cell death.
In general, the spermidine (SPD) protection assays
were performed to determine whether uptake of selected
polyamine analogues is mediated, in part or in whole,
by the polyamine transport apparatus (PAT). To answer
this question, competition assays were performed in the
absence and presence of SPD. Figure 6 below illustrates
the competitive uptake scenario probed by these experi-
ments. To maximize the protective effects of spermidine
(SPD), an excess of SPD (25 µM) was used during these
experiments. Using a 25Km excess of SPD ensures that
SPD is transported into the cell at nearly Vmax, and hence,
provides a high level of competition with the selected
polyamine derivatives for the PAT protein.
The results of experiments using DMSPD and DESPD
are shown in Figure 5, panels 1 and 2, respectively. The
graphs clearly indicate that over a 48 h period, the
selected assay format provides sufficient sensitivity to
reveal changes in cell viability. The low concentration
range of DESPD and DMSPD used in these experiments
was intended to produce a low-level of toxicity and, hence,
amplify any observed SPD-mediated protection.
Examination of data obtained with DMSPD reveals a
significant reduction of cell death in the presence of SPD.
Specifically, at the lowest concentration of DMSPD (0.6
µM), the presence of SPD decreases the observed re-
sponse by 97%. A 77% decrease in response was observed
at the highest concentration of DMSPD (11 µM). This
level of protection is consistent with a competitive mech-
anism. More significantly, the response can be predicted
using eq 1. When 0.6 and 2.5 µM are substituted for
[DMSPD] and Km, respectively, and 25 and 2.2 µM are
Evaluation of these compounds provided both a veri-
fication of inhibition, as well as, insight into the contribu-
tions that each molecular fragment (i.e., intercalator and
polyamine) had to the observed inhibitory activity.
Examination of both gels reveals that the acridine
derivatives (2, 3, 17, and 23) possess similar potency as
the berenil control 22. These results are consistent with
the avid binding characteristics observed with these
altered ligand types with model oligonucleotides in
vitro10,36,37a and the ability of other acridines to inhibit
the DNA decatenation activity of TOPO-II.37b Incremental
changes in tether length (2 vs 3 and 4 vs 5) did not seem
to influence the efficacy of either system. Since the
conjugates 2 and 3 were virtually equivalent to the
acridine controls (lanes 8 and 9) and spermidine (lane
10) had little effect on decatentation, one can conclude
that the appended polyamine did not disrupt the ability
of the intercalator component to inhibit TOPO-II activity.
L1210 Stu d ies. L1210 (murine leukemia) cells were
challenged with various concentrations of the conjugates
(2-5 and 16) and their respective IC50 values at 47 and
94 h are listed in Table 1. Even though the nonpoly-
amine-containing 9-(N-butyl)aminoacridine control was
the most potent derivative tested, conjugation of a
polyamine component onto the 9-aminoacridine archi-
tecture clearly modulates its bioactivity (Table 1: 2 vs
9-(N-butyl)aminoacridine). Surprisingly, even though 2
and 3 were more potent inhibitors of TOPO-II in the
above gel assay, 4 and 5 were more efficacious in the
whole cell-based assay. This reverse trend implies that
either (a) TOPO-II inhibition is not the mechanism of
action for these conjugates or (b) that gaining access to
the cell (and eventually the nucleus) is more important
than TOPO-II inhibition alone or (c) that a balance is
needed between transport into the cell and TOPO-II
inhibitory ability. Interestingly, conversion of the ap-
pended polyamine component into the longer spermine
backbone increased the potency of 2 as shown by the
lower IC50 of the mono-alkylated spermine derivative 16.
The reported mean IC50 value for AMSA 1 is 0.26 µM for
six human cancer lines with varying levels of TOPO-II R
and â isoforms.37b As both 137b and the acridines tested
(2 and 3) are potent inhibitors of TOPO-II (by gel assay),
the cytotoxicity difference is not likely due to TOPO-II
inhibition, but the drug’s ability to enter the cell. To
(36) Zegar, I. S.; Prakash, A. S.; Harvey, R. G.; LeBreton, P. R. J .
Am. Chem. Soc. 1985, 107, 7990-7995.
(37) (a) Price, H. L.; Phanstiel, O., IV. Unpublished results. (b)
Stanslas, J .; Hagan, D. J .; Ellis, M. J .; Turner, C.; Carmichael, J .;
Ward, W.; Hammonds, T. R.; Stevens, F. G. J . Med. Chem. 2000, 43,
1563-1572.
(38) Bergeron, R. J .; Feng, Y.; Weimar, W. R.; McManis, J . S.;
Dimova, H.; Porter, C.; Raisler, B.; Phanstiel, O. J . Med. Chem. 1997,
40, 1475-1494.

Effects of Polyamine Homologation
J . Org. Chem., Vol. 65, No. 18, 2000 5595
decrease was observed). These values are in excellent
agreement with the observed level of SPD-mediated
protection. These results also support the assumption
that PAT-mediated uptake is coupled to the toxicity of
these substrates.
Vo
[DMSPD]
[SPD]
)
(1)
Vmax
Km 1 +
+ [DMSPD]
(
)
Ki
Equation 1 was also used to predict the level of
inhibition that would be expected when DESPD was
present in the culture medium. According to the calcula-
tions, at both 0.6 and 11 µM DESPD, the relative rate of
transport would have been inhibited by more than 90%.
The results of this set of SPD protection experiments
reveal a significantly lower degree of protection (Figure
5, panel 2). The decreased sensitivity to SPD can be
better understood by inclusion of an alternative transport
process, and perhaps even an additional transport pro-
tein. Analogous behavior has been observed using in vitro
model systems designed to study the effect of uptake
blockers on cellular responses to agonists.39-41
A reduced sensitivity to added spermidine was also
observed with 2 (not shown) and 4 (panel 3, Figure 5).
There are several possible interpretations of these re-
sults. The first is that 4 exhibits an even greater affinity
for the transport apparatus than either DMSPD, DESPD,
or SPD. The second possibility, and the most reasonable,
is that 4 displays a low affinity for PAT and a second
process is responsible for transport of this analogue. In
this regard, the spermidine-based ligands (e.g., 2 and 4)
may not represent the optimum vector architecture.
However, experiments with the spermine (SPM) deriva-
tive 16 and L1210 cells revealed that significant cell
rescue (panel 4, Figure 5) was observed upon addition of
exogenous spermidine (25 µM). These observations are
compelling as they suggest that conjugates such as 16
(which has an IC50 comparable to 4) may be using the
polyamine transport apparatus to gain entry into L1210
cells. Moreover, these preliminary observations support
spermidine’s antagonistic character in these studies and
support the premise that certain conjugates can compete
with spermidine for the transporter while others are
excluded due to their structure.42 In summary, the
selectivity of the PAT protein can be used to deliver
specific architectures into cells.
F igu r e 5. Spermidine protection experiments. See Methods
section for description of experimental conditions. No spermi-
dine (triangles), 25 µM spermidine (circles). A decreased
response to added drug in the presence of spermidine is
consistent with reduced drug uptake. Data shown represent
the average of two independent trials, each performed in
triplicate.
Con clu sion s
In terms of vector design, the ability to modulate the
delivery characteristics of drug conjugates and to ulti-
mately direct them to specific cell types are mutually
inclusive goals. Our goal was not to generate a highly
toxic agent with little inherent selectivity (e.g., 17).
Instead, we sought to better understand the role of
(39) Kenakin, T. P. Pharmacologic Analysis of Drug-Receptor In-
teraction, 2nd ed.; Raven Press: New York, 1993; and references
therein.
(40) Kenakin, T. P.; Leighton, H. J . In Methods in Pharmacology.
Methods Used in Adenosine Research; Paton, D. M., Ed.; Plenum
Press: New York, 1984; Vol. 6, pp 213-237.
(41) Kenakin, T. P. J . Pharmacol. Exp. Ther. 1982, 222, 752-758.
(42) It should be noted that other protection experiments with a bis-
alkylated spermine derivative (Scheme 3, 16a : compound 16 where
R1 ) -(CH2)4NH-acridine and R2 dH) showed essentially no spermi-
dine protection, unpublished work.
F igu r e 6. Competitive scenario for polyamine uptake.
substituted for [SPD] and Ki, respectively, the relative
rate of transport in the presence of SPD, Vo/Vmax, is
calculated to be 0.02 or 98% inhibited (e.g., a 97%
decreased response was observed). The predicted Vo/Vmax
for 11 µM DMSPD is 0.25 or 75% inhibited (e.g.,77%

5596 J . Org. Chem., Vol. 65, No. 18, 2000
Phanstiel et al.
the absence and presence of increasing concentrations of drug.
A positive control containing methotrexate was included in all
trials. Cell viability was determined using a colorimetric assay
based on the NAD(P)-dependent production of Formazan.43,44
Trypan blue staining was used to determine cell viability
before the initiation of the cytotoxicity and transport experi-
ments. Typically, samples contained less than 5% trypan blue
positive cells (dead). In these experiments, L1210 cells in early
to mid log-phase were used.
Sp er m id in e P r otection Assa ys. L1210 cells were seeded
into 96-well plates, grown in medium containing the polyamine
oxidase inhibitor, aminoguanidine (250 µM), and increasing
concentrations of drug, in the presence and absence of a
saturating concentration of spermidine (25 µM, approximate
Km ) 1.0 µM). Cells were grown for 48 h. Relative cell viability
was determined using the colorimetric MTS/PMS assay.43,44
After a 2 h incubation, the absorbance at 490 nm was recorded
using a Wallac Victor2 plate reader operating in absorbance
mode. The relative viability was determined using the follow-
ing ratio:
polyamine conjugation in terms of drug delivery. This
report demonstrated not only the sensitivity of the
polyamine transporter to pendant ligands (like acridine
and anthracene) but also suggests a possible strategy for
using this pathway to gain entry into cells.
Lengthening of the polyamine tail of 2 by an additional
aminopropyl group directly enhanced the potency of the
altered conjugate (see 16). Although further work is
needed, these results suggest that of the series studied
thus far the monoalkylated tetraamine motif (16, R )
H) may be the best vector to access the transporter.
Intuitively, conjugates which more closely resemble the
parent polyamines are more likely to interact with and
be recognized by the transporter. These findings suggest
that superior vectors (with even higher affinities for PAT)
may be accessed by either extending the appended
polyamine further from the polyaromatic core or by
further aminoalkylation of the parent polyamine chain.
Our future studies will not only focus on this insight, but
will also round off the systematic study of tether length
we have begun to establish. In summary, these results
suggest that the appended polyamine is a “value-added”
fragment, which can be used to tailor the delivery and
potency of polyamine-containing conjugates.
100 × [(A490 cells + drug - A490 drug blank )/(A490 cells only
-
A490 blank)]
N1,N8-Dim et h ylsp er m id in e Tr ih yd r och lor id e (DM-
SP D) a n d N1,N8-Diet h ylsp er m id in e Tr ih yd r och lor id e
(DE SP D). The dimethyl and diethyl SPD derivatives were
synthesized by the methods of Bergeron38 in 65% and 63% yield
from spermidine, respectively. DMSP D: 1H NMR (D2O) δ
3.12-2.90 (m, 8H), 2.66 (s, 3H), 2.65 (s, 3H), 2.12-1.95 (m,
2H), 1.75-1.68 (m, 4H); 13C NMR (D2O): δ 51.1, 49.9, 48.6,
47.3, 35.6, 35.5, 25.6, 25.5, 25.4. Anal. Calcd for C9H26N3Cl3:
C, 38.24; H, 9.27; N, 14.86. Found: C, 38.29; H, 9.36; N, 14.90.
DESP D: 1H NMR (D2O) δ 3.19-3.04 (m, 12H), 2.18-2.06 (m,
2H), 1.81-1.73 (m, 4H), 1.34-1.24 (m, 6H). Anal. Calcd for
Exp er im en ta l Section
Ma t er ia ls. Silica gel (36-63) mesh 60 Å was purchased
from Selecto Scientific (Norcross, GA). Chemical reagents were
purchased either from the ACROS Chemical Co. or the Sigma
Chemical Co. and used without further purification. Agarose
was purchased from the Bio-Rad Co. 1H NMR spectra were
recorded at 200 MHz (unless otherwise indicated). TLC solvent
systems are based on volume % and NH4OH refers to concen-
trated aqueous NH4OH. The human type II topoisomerase
(p170 form) and kinetoplast DNA were purchased from Topo-
gen, Inc. (Columbus, OH). All biologically evaluated conjugates
gave satisfactory elemental analyses prior to testing. Elemen-
tal analyses were performed by Atlantic Microlabs (Norcross,
GA).
TOP O-II Assa y. The protocol for assaying topoisomerase
II activity is well established.29-31 Briefly, decatenation assays
were performed in buffer A containing 50 mM Tris-HCl (pH
8.0), 120 mM KCl, 10 mM MgCl2, 30 µg/mL bovine serum
albumin, 0.5 mM dithiothreitol, and 0.5 mM ATP. Reaction
mixtures contained 0.5 µg of catenated kinetoplast DNA, four
units of topoiosomerase II, and the appropriate polyamine
conjugate 2-5 at 10 µM (Figure 4, panel A) (or at 20 µM for
panel B results). Controls were run in parallel and contained
no inhibitor and no enzyme. Berenil, 9-aminoacridine, and
9-(N-butyl)aminoacridine were also evaluated at 10 µM (and
20 µM). Reaction mixtures were incubated at 37 °C for 30 min.
Decatenation reactions were quenched with 20 mM Na2EDTA,
and 100 µg/mL proteinase K. Reaction products were separated
by electrophoresis at 2.5 V/cm through a 1.5% agarose gel. The
gel contained 0.5 µg/mL ethidium bromide and was submerged
in buffer B [1X TAE buffer: 40 mM Tris-HCl, 25 mM sodium
acetate, and 1 mM EDTA (pH 8.5)]. The reaction products were
visualized under ultraviolet light.
C
11H30N3Cl3: C, 42.52; H, 9.73; N, 13.52. Found: C, 42.37; H,
9.73; N, 13.42.
N4-(4-(9-Am in oa cr id in yl)b u t yl)sp er m id in e Tet r a h y-
d r och lor id e (2). The butyl adduct (11) (0.27 g, 0.42 mmol)
was dissolved in 1,4-dioxane (5 mL) and stirred at 0 °C. HCl
(4 N, 14 mL) was added. The reaction mixture was stirred
under a nitrogen atmosphere for 4 h at room temperature. The
disappearance of 11 was monitored by TLC (0.5% NH4OH/CH3-
OH), and the reaction was complete after 4 h. The volatiles
were removed under reduced pressure to give 2 as a yellow
oil (0.213 g, 79%). An analytically pure sample was obtained
by recrystallization from (25% absolute ethyl alcohol/diethyl
ether) to give a yellow semisolid, which was protected against
light (with aluminum foil) and stored under nitrogen at 0 °C.
1
2: Rf ) 0.3, 12% NH4OH/CH3OH; H NMR (CD3OD) δ 8.59
(d, 2H), 7.90 (m, 4H), 7.55 (t, 2H), 4.25 (t, 2H), 3.35 (m, 6H),
3.06 (m, 4H), 2.39-1.70 (m, 10H); high-resolution FAB MS
m/z found 394.2965 (MH+), C24H35N5O4 requires 394.2970.
Anal. Calcd for C24H35N5O4‚4HCl/4.5H2O: C, 46.46; H, 7.80;
N, 11.28. Found: C, 46.34; H, 7.42; N, 11.31.
N4-(5-(9-Am in oa cr id in yl)p en tyl)-sp er m id in e Tetr a h y-
d r och lor id e (3). The pentylamine adduct 12 (0.373 g, 0.56
mmol) was dissolved in 1,4-dioxane (6 mL) at 0 °C. HCl (4 N,
18 mL) was added. The reaction mixture was stirred under a
nitrogen atmosphere for 4 h at room temperature. The disap-
pearance of 12 was monitored by TLC (1% NH4OH/CH3OH)
and was complete after 4 h at room temperature. The volatiles
were removed under reduced pressure to give 3 as a yellow
oil (0.29 g, 79%). An analytically pure sample of 3 was obtained
by recrystallization (from 25% absolute ethyl alcohol/diethyl
ether) to give a yellow semisolid, which was protected against
light (with aluminum foil) and stored under nitrogen at 0 °C.
3: Rf ) 0.4, 12% NH3/CH3OH; 1H NMR (CD3OD) δ 8.60 (d,
2H), 7.90 (m, 4H), 7.60 (t, 2H), 4.21 (t, 2H), 3.30 (m, 6H), 3.05
IC50 Det er m in a t ion s a n d Met h od s for Tr a n sp or t -
Rela ted Stu d ies. Murine leukemia cells (L1210) were ob-
tained from the American Type culture collection (ATCC). All
reagents and materials used specifically for the growth and
maintenance of this cell line were obtained from Sigma
Chemical Co. and Fisher Scientific, Inc. Cells were grown and
maintained in RPMI-1640 medium, supplemented with 10%
horse serum and 1% antibiotic/antimycotic cocktail (100 units
of penicillin, 100 µg of streptomycin, and 0.25 µg of amphot-
ericin B). Cells were grown at 37 °C in a 5% CO2 atmosphere.
For these studies, cells were plated out into sterile 96-well
microtiter plates and grown for the specified period of time in
(43) Malich, G.; Markovic, B.; Winder, C. Toxicology 1997, 124, 179-
192.
(44) Buttke, T. M.; McCubrey, J . A.; Owen, T. C. J . Immunol.
Methods 1993, 157, 233-240.

Effects of Polyamine Homologation
J . Org. Chem., Vol. 65, No. 18, 2000 5597
(m, 4H), 2.30-1.50 (m, 12H); Anal. Calcd for C25H37N5‚4HCl/
4H2O: C, 48.01; H, 7.89; N, 11.20. Found: C, 48.33; H, 7.64;
N, 11.31.
N4-(4-Cya n obu tyl)-N1,N8-bis(ter t-bu toxyca r bon yl)sp er -
m id in e (8). Protected spermidine 6 (1.5 g, 4.3 mmol) and
5-bromovaleronitrile (1.05 g, 6.5 mmol) were dissolved in 15
mL of acetonitrile, and solid K2CO3 (0.51 g, 5.2 mmol) was
added. The vigorously stirred mixture was heated to reflux.
The disappearance of 6 was monitored by TLC (Rf ) 0.35, 1%
NH4OH/CH3OH). The reaction was complete in 4 h. The
volatiles were removed under reduced pressure to give a
residue, which was redissolved in CH2Cl2 and washed with
aqueous Na2CO3 (pH ) 10). The organic phase was separated,
dried over anhydrous Na2SO4, filtered, concentrated and
subjected to flash chromatography (100% EtOAc) to remove
the impurities followed by 20% CH3OH/EtOAc to elute the
desired nitrile 8 (1.58 g: 86%). 8: Rf ) 0.39, 20% CH3OH/
EtOAc; 1H NMR (CDCl3) δ 5.25 (br s, 1H), 4.75 (br s, 1H), 3.12
(m, 4H), 2.49 (m, 6H), 2.40 (m, 2H), 1.65 (m, 4H), 1.50 (m,
6H), 1.43 (s, 18H). Anal. Calcd for C22H42N4O4: C, 61.94; H,
9.92; N, 13.13. Found: C, 62.29; H, 9.61; N, 13.48.
N4-(4-(9-Am in om et h yla n t h r a cen yl)b u t yl)sp er m id in e
Tetr a h yd r och lor id e (4). The anthracene derivative 14 (0.30
g, 0.49 mmol) was dissolved in 1,4-dioxane (5 mL) at 0 °C, and
4 N HCl (14 mL) was added. The solution was stirred under a
nitrogen atmosphere for 4 h at room temperature. The
complete consumption of 14 was monitored by TLC (Rf ) 0.3,
10% CH3OH/EtOAc). The volatiles were removed under vacuum
to give a pale yellow oil (0.28 g, 91%). Upon standing, a pale
yellow semisolid 4 was obtained which was protected against
light (with aluminum foil) and stored under
a nitrogen
atmosphere at 0 °C. 4: Rf ) 0.3, 8% NH4OH/CH3OH; 1H NMR
(CD3OD) δ 8.60 (s, 1H), 8.42 (d, 2H), 8.09 (d, 2H), 7.62 (m,
4H), 5.28 (s, 2H), 3.52-3.00 (m, 12H), 2.30 (m, 4H), 2.05-1.79
(m, 6H). Anal. Calcd for C26H38N4‚4HCl/4.4H2O: C, 49.43; H,
8.10; N, 8.87. Found: C, 49.10; H, 7.70; N, 8.71.
N 4-(5-(9-Am in om e t h yla n t h r a ce n yl)p e n t yl)sp e r m i-
d in e Tetr a h yd r och lor id e (5). The anthracene adduct 15
(0.620 g, 1 mmol) was dissolved in 1,4-dioxane (5 mL) at 0 °C,
and 4 N HCl (14 mL) was added. The reaction mixture was
stirred under nitrogen atmosphere for 4 h at room tempera-
ture. The complete consumption of 15 was monitored by TLC
(Rf ) 0.3, 10% CH3OH/EtOAc). The volatiles were removed
under reduced pressure to give a pale yellow oil (0.60 g, 93%).
Upon standing, a pale yellow semisolid 5 was obtained which
was protected against light (with aluminum foil) and stored
under a nitrogen atmosphere at 0 °C. 5: Rf ) 0.36, 8% NH4-
OH/CH3OH;1H NMR (CD3OD) δ 8.70 (s, 1H), 8.46 (d, 2H), 8.16
(d, 2H), 7.63 (m, 4H), 5.35 (s, 2H), 3.46-2.99 (m, 12H), 2.22
(m, 4H), 2.05-1.50 (m, 8H); High-resolution FAB MS m/z
found 421.3312 (MH+), C27H40N4 requires 421.3331. Anal.
Calcd for C27H40N4‚4HCl/4.2H2O: C, 50.50; H, 8.23; N, 8.73.
Found: C, 50.35; H, 7.87; N, 8.64.
N4-(4-Am in obu tyl)-N1,N8-bis(ter t-bu toxyca r bon yl)sp er -
m id in e (9). Nitrile 7 (2.0 g, 4.80 mmol) was dissolved in
absolute EtOH and transferred to a Parr shaker bottle. Raney
nickel (50% slurry: 1.6 gm) and 3.4 mL of concentrated NH4-
OH were added to the solution. NH3 gas was then passed
through the solution at 0° C for 15 min. The reaction vessel
was quickly attached to the Parr shaker and was partially
evacuated. H2 gas was introduced at a pressure of 75 psi. The
mixture was shaken and the disappearance of 7 was monitored
by TLC (10% CH3OH/EtOAc). After the reaction was complete
(78 h) the catalyst was filtered off and washed using hot
absolute EtOH. The filtrate was concentrated under reduced
pressure and the residue was dissolved in CH2Cl2 and washed
with aqueous Na2CO3 (pH ) 10). The organic layer was
separated, dried over anhydrous Na2SO4, filtered, concentrated
and subjected to flash chromatography using 3% NH4OH/ CH3-
OH to give the desired amine 9 (0.80 g: 40%). Note: when
this reaction was repeated on a smaller scale the yield was
80% (126 mg of 9). 9: Rf ) 0.4, 3% NH4OH/CH3OH; 1H NMR
(CDCl3) δ 5.56 (br s, 1H), 5.00 (br s, 1H), 3.13 (m, 4H), 2.84
(br s, 2H), 2.74 (br s, 2H), 2.42 (m, 6H), 1.62 (t, 2H), 1.55-
1.40 (m, 26H, includes a signal at 1.48 (s, 18H)); 13C NMR
(CDCl3) δ 156.34, 79.15, 79.02, 54.03, 53.75, 52.66, 41.88, 40.61,
40.05, 31.10, 28.68, 28.65, 28.16, 26.91, 24.59, 24.42. Anal.
Calcd for C21H44N4O4: C, 60.54; H, 10.64; N, 13.45. Found:
C, 60.36; H, 10.67; N, 13.45.
N1,N8-Bis(ter t-bu toxyca r bon yl)sp er m id in e (6). Spermi-
dine (2.22 g, 15.3 mmol) was dissolved in 45 mL of THF and
stirred for 20 min at 0 °C. A solution of 2-(tert-butoxycarbo-
nyloxyimino)-2-phenylacetonitrile) (BOC-ON, 7.53 g, 30.5
mmol) dissolved in 85 mL of THF was added dropwise with
constant stirring. The reaction was carried out under
a
nitrogen atmosphere. After the addition was complete, the
reaction was further stirred for 1 h at 0 °C. The disappearance
of spermidine was monitored by TLC (Rf ) 0.35, 3% NH4OH/
CH3OH). The volatiles were removed under reduced pressure
to give a residue, which was redissolved in CH2Cl2 and washed
with aqueous Na2CO3. The organic layer was separated and
dried over anhydrous Na2SO4, filtered, concentrated and
subjected to flash chromatography (1% NH4OH/CH3OH) to
give derivative 6 (4.41 g, 79%). An analytically pure sample
was obtained by recrystallization (from 25% diethyl ether, 1:1
N4-(5-Am in op en t yl)-N1,N8-b is(ter t-b u t oxyca r b on yl)-
sp er m id in e (10). Nitrile 8 (1.2 g, 2.80 mmol) was dissolved
in absolute EtOH and transferred to a Parr shaker bottle.
Raney nickel (50% slurry: 1.6 gm) and 3.4 mL of concentrated
NH4OH were added to the solution. NH3 gas was then passed
through the solution at 0 °C for 15 min. The reaction vessel
was attached to the Parr shaker and was partially evacuated.
H2 gas was introduced at a pressure of 75 psi. The disappear-
ance of 8 was monitored by TLC (20% CH3OH/EtOAc). After
the reaction was complete (78 h) the catalyst was filtered off
and washed with hot absolute EtOH. The filtrate was concen-
trated under reduced pressure and the residue was dissolved
in CH2Cl2 and washed with aqueous Na2CO3 (pH ) 10). The
organic phase was separated, dried over anhydrous Na2SO4,
filtered, concentrated and subjected to flash chromatography
using 1% NH4OH/ CH3OH to elute the impurities followed by
2.5% NH4OH/ CH3OH to elute the desired amine 10 (0.90 g:
1
EtOAc/ hexane). 6: Rf ) 0.35, 1% NH4OH/CH3OH; H NMR
(600 MHz, CDCl3) δ 5.20 (br s, 1H, NH), 4.82 (br s, 1H, NH),
3.19 (m, 4H), 2.61 (m, 4H), 1.62 (m, 2H), 1.54 (m, 4H), 1.42 (s,
18H). Anal. Calcd for C17H35N3O4 : C, 59.10; H, 10.21; N, 12.16.
Found: C, 59.18; H, 10.25; N, 12.18.
N4-(3-Cya n op r op yl)-N1,N8-b is(ter t-b u t oxyca r b on yl)-
sp er m id in e (7). Protected spermidine 6 (2.5 g, 7.2 mmol) and
4-bromobutyronitrile (1.4 g, 9.0 mmol) were dissolved in 15
mL of acetonitrile, and solid K2CO3 (0.87 g, 9.0 mmol) was
added. The vigorously stirred mixture was heated to reflux.
The disappearance of 6 was monitored by TLC (Rf ) 0.35, 1%
NH4OH/CH3OH). The reaction was complete in 4 h. The
volatiles were removed under reduced pressure to give a
residue, which was redissolved in CH2Cl2 and washed with
aqueous Na2CO3 (pH ) 10). The organic phase was separated,
dried over anhydrous Na2SO4, filtered, concentrated, and
subjected to flash chromatography (20% hexane/EtOAc) to
remove the impurities followed by 10% CH3OH/EtOAc to elute
the desired nitrile 7 (1.7 g, 92%). 7: Rf ) 0.39, 10% CH3OH/
EtOAc; 1H NMR (CDCl3) δ 5.19 (br s, 1H), 4.82 (br s, 1H), 3.14
(m, 4H), 2.49 (m, 2H), 2.41 (m, 6H), 1.77 (m, 2H), 1.62 (m,
2H), 1.56-1.38 (m, 22H, includes signal at 1.44 (s, 18H)); high-
resolution FAB MS m/z found 413.3130 (MH+), C21H41N4O4
requires 413.3128.
1
74%). 10: Rf ) 0.39, 2.5% NH4OH/CH3OH; H NMR (CDCl3)
δ 5.56 (br s, 1H), 5.01 (br s, 1H), 3.13 (m, 4H), 2.70 (m, 2H),
2.40 (m, 6H), 1.60 (m, 12H), 1.45 (s, 18H). Anal. Calcd for
C
22H46N4O4: C, 60.85; H, 10.77; N, 12.90. Found: C, 60.81;
H, 10.64; N, 12.78.
N4-(4-(9-Am in oa cr id in yl)bu tyl)-N1,N8-bis(ter t-bu toxy-
ca r bon yl)sp er m id in e (11). 9-Chloroacridine (0.175 g, 0.80
mmol) and phenol (0.771 g, 8.0 mmol) were heated to 50 °C
with constant stirring under a nitrogen atmosphere. The
amine 9 (0.340 g, 0.80 mmol) was predissolved in molten
phenol. The temperature of the phenol/amine 9 mixture was
maintained at 45 °C to keep phenol in the molten state. The
molten mixture was then added to the acridine mixture
dropwise and the reaction heated to 110 °C and stirred for 15

5598 J . Org. Chem., Vol. 65, No. 18, 2000
Phanstiel et al.
min. The reaction was monitored by TLC (Rf ) 0.3, 20%
EtOAc/hexane) and (3% NH4OH/CH3OH) for the complete
consumption of 9-chloroacridine and amine 9, respectively.
After the reaction was complete the mixture was subjected to
flash chromatography using (100% CH3OH) to elute impurities
followed by 0.5% NH4OH/CH3OH to give the acridine deriva-
tive 11 (0.39 g: 80%). 11: Rf ) 0.4, 0.5% NH4OH/CH3OH; 1H
NMR (CD3OD) δ 8.10 (m, 4H), 7.65 (t, 2H), 7.35 (t, 2H), 3.80
(t, 2H), 3.10 (m, 4H), 2.35 (m, 6H), 1.58 (m, 10H), 1.40 (s, 18H);
high-resolution FAB MS m/z found 594.3998 (MH+), C34H51N5O4
requires 594.3998. Anal. Calcd for C34H51N5O4‚1.5 H2O: C,
65.78; H, 8.76; N, 11.28; Found: C, 65.54; H, 8.64; N, 11.00.
N4-(5-(9-Am in oa cr id in yl)p en tyl)-N1,N8-bis(ter t-bu toxy-
ca r bon yl)sp er m id in e (12). 9-Chloroacridine (0.175 g, 0.82
mmol) and phenol (0.770 g, 8.2 mmol) were heated to 50 °C
with constant stirring under a nitrogen atmosphere. The
pentylamine derivative 10 (0.350 g, 0.82 mmol) was predis-
solved in molten phenol. The temperature of the phenol/amine
10 mixture was maintained at 45 °C to keep phenol in the
molten state. The molten mixture was then added to the
acridine mixture dropwise and the reaction was heated to 110
°C and stirred for 15 min. The reaction was monitored by TLC
(Rf ) 0.3, 20% EtOAc/hexane) and (2.5% NH4OH/CH3OH) for
the complete consumption of 9-chloroacridine and amine 10,
respectively. After the reaction was complete, the mixture was
subjected to flash chromatography using 100% CH3OH to elute
impurities followed by 1% NH4OH /CH3OH to give the acridine
derivative 12 (0.37 g: 78%). 12: (Rf ) 0.4, 1% NH4OH/CH3-
OH); 1H NMR (CDCl3) δ 8.10 (m, 4H), 7.65 (t, 2H), 7.32 (t,
2H), 5.31 (br s, 1H), 4.85 (br s, 1H), 3.81 (t, 2H), 3.10 (m, 4H),
2.35 (m, 6H), 1.80 (m, 2H), 1.55 (m, 10H), 1.41 (s, 18H). Anal.
Calcd for C35H53N5O4‚1.5 H2O: C, 66.22; H, 8.89; N, 11.03.
Found: C, 65.89; H, 9.12; N, 11.34.
same solvent mixture, 2 mL) was added dropwise. After the
mixture was stirred for 6 h at room temperature under a
nitrogen atmosphere, activated 4 Å molecular sieves were
added and the mixture was stirred for an additional 24 h. The
consumption of 9-anthraldehyde was monitored by TLC (Rf )
0.3, 14% EtOAc/hexane). After 30 h the reaction mixture was
filtered and the volatiles were removed under reduced pressure
to give a bright yellow powder that was used without further
purification. The powder was dissolved in 50% CH2Cl2/MeOH
(3 mL), and the mixture was cooled to 0 °C. NaBH4 (0.165 g,
0.44 mmol) was added, and the solution was stirred at 0 °C
for 3 h and at room temperature for 1 h. The volatiles were
removed under reduced pressure to give a pale yellow residue.
The residue was then dissolved in CH2Cl2 and washed with
water. The organic layer was then separated, dried over
anhydrous Na2SO4, filtered, and concentrated to give a pale
yellow crude mixture. The residue was then subjected to flash
chromatography using 100% EtOAc to elute the impurities
followed by 10% CH3OH/EtOAc to give the anthracene adduct
1
15 (0.62 g: 68%). 15: Rf ) 0.3, 10% CH3OH/EtOAc; H NMR
(CDCl3) δ 8.35 (m, 3H), 8.01 (d, 2H), 7.50 (m, 4H), 5.60 (s, 1H),
5.2 (s, 1H), 4.71 (s, 2H), 3.10 (m, 4H), 2.85 (t, 2H), 2.38 (m,
6H), 1.70-1.20 (m, 30H, this chemical shift range includes the
signal at 1.42 (s, 18H)); high-resolution FAB MS m/z found
621.4364 (MH+), C37H56N4O4 requires 621.4379. Anal. Calcd
for C37H56N4O4‚1.2 H2O: C, 69.17; H, 9.16; N, 8.72. Found:
C, 69.22; H, 8.80; N, 8.33.
N4-(4-(9-Am in oa cr id in yl)bu tyl)sp er m in e Hexa h yd r o-
ch lor id e 16. The acridine derivative 21 (0.12 g, 0.18 mmol)
was dissolved in dioxane (1.5 mL) at 0 °C followed by the
addition of 4N HCl (1.5 mL). The mixture was stirred at room
temperature. The disappearance of the starting material was
monitored by TLC (Rf ) 0.24, 2% NH4OH/CH3OH). After 4 h,
the mixture was concentrated under reduced pressure and the
product was recrystallized with CH3OH and Et2O to give the
hygroscopic salt 16 (84 mg, 72%). 16: 1H NMR (CD3OD) δ 8.50
(d, 2H), 7.90 (t, 2H), 7.75 (d, 2H), 7.50 (t, 2H), 4.20 (broad m,
2H), 3.38 (broad m, 6H), 3.20 (broad m, 8H), 2.40-1.80 (broad
m, 12H). Anal. Calcd for C27H47N6Cl5‚ 3.5 H2O: C, 46.59; H,
7.82; N, 12.08. Found: C, 46.22; H, 7.45; N, 11.95.
9-P h en oxya cr id in e (13). 9-Chloroacridine (0.5 g, 2.3
mmol) and phenol (2.25 g, 23 mmol) were heated to 100 °C
with constant stirring for 15 min under a nitrogen atmosphere.
The crude mixture was recrystallized using 25% EtOAc/hexane
to give 13 (4.62 g: 72%):9,26 1H NMR (CDCl3) δ 9.10 (d, 2H),
8.25 (d, 2H), 8.10 (t, 2H), 7.69 (m, 2H), 7.40 (m, 2H), 7.20 (m,
1H), 6.89 (d, 2H); high-resolution FAB MS m/z found 272.1073
(MH+), C19H13N1O1 requires 272.1075.
9-N-Bu tyla m in oa cr id in e 17. 9-Chloroacridine (0.5 g, 2.3
mmol) and phenol (2.26 g, 10 mmol) were heated to 50 °C with
constant stirring under a nitrogen atmosphere. Butylamine
(0.168 g, 2.3 mmol) was predissolved in molten phenol. The
molten mixture was then added to the acridine mixture
dropwise and the reaction heated to 110 °C and stirred for 15
min. The reaction was monitored by TLC (Rf ) 0.3, 20%
EtOAc/hexane), which revealed complete consumption of
9-chloroacridine. After the reaction was complete, the mixture
was subjected to flash chromatography eluting with 20%
MeOH/EtOAc (Rf ) 0.4) to give the N-butylamino derivative
17 (0.41 g, 72%). 17: 1H NMR (CD3OD) δ 8.50 (d, 2H), 7.98
(m, 2H), 7.80 (d, 2H), 7.57 (m, 2H), 4.15 (t, 2H), 1.98 (m, 2H),
1.51 (m, 2H), 1.01 (t, 3H); HRMS m/z calcd for C17H19N2 (MH+)
251.1545, found 251.1549. TLC (with both UV and I2 detection)
N4-(4-(9-Meth ylam in oan th r acen yl)bu tyl)-N1,N8-bis(ter t-
bu toxyca r bon yl)sp er m id in e (14). 9-Anthraldehyde (0.204
g, 0.99 mmol) was dissolved in 20% dry MeOH/CH2Cl2 (1.5
mL), and the amine 9 (0.456 g, 1.09 mmol) (dissolved in 1 mL
of the same solvent mixture) was added dropwise. After the
mixture was stirred for 6 h at room temperature under a
nitrogen atmosphere, activated 4 Å molecular sieves were
added and the mixture was stirred for an additional 24 h. The
consumption of 9-anthraldehyde was monitored by TLC (Rf )
0.3, 14% EtOAc/hexane). After 30 h the reaction mixture was
filtered and the volatiles were removed under reduced pressure
to give a bright yellow powder (0.65 g) that was used without
further purification. The powder was dissolved in 50% CH2-
Cl2/MeOH (2 mL), and the solution was cooled to 0 °C. NaBH4
(0.165 g, 0.44 mmol) was added and the solution was stirred
at 0 °C for 3 h and then at room temperature for 1 h. The
volatiles were removed under vacuum to give a pale yellow
residue. The residue was then dissolved in CH2Cl2 and washed
with water. The organic layer was separated, dried over
anhydrous Na2SO4, filtered and concentrated to give a pale
yellow crude mixture. The mixture was then subjected to flash
chromatography using 100% EtOAc to elute the impurities
followed by 10% CH3OH/EtOAc to give the desired adduct 14
(0.4 g: 66%). 14: Rf ) 0.3, 10% CH3OH/EtOAc; 1H NMR
(CDCl3) δ 8.35 (m, 3H), 8.00 (d, 2H), 7.50 (m, 4H), 5.42 (br s,
1H), 4.91 (br s, 1H), 4.72 (s, 2H), 3.06 (m, 4H), 2.86 (t, 2H),
2.35 (m, 6H), 1.60-1.20 (m, 28H, this chemical shift range
includes the signal at 1.40 (s, 18H)). Anal. Calcd for C36H54N4O4‚
0.5 H2O: C, 70.21; H, 9.00; N, 9.10. Found: C, 70.21; H, 9.00;
N, 9.10.
1
revealed one spot. The H NMR spectrum of 17 (in CD3OD) is
provided as additional proof of purity for 9-(N-butylamino)-
acridine 17 (see the Supporting Information).
N1,N12-Bis(ter t-bu toxyca r bon yl)sp er m in e 18. Spermine
(9.87 g, 49 mmol) was dissolved in THF (150 mL) and cooled
to 0 °C. BOC-ON (24.04 g, 98 mmol) in THF (300 mL) was
added dropwise, and the reaction mixture was stirred for 2 h
at 0 °C, resulting in the complete conversion of spermine as
mentioned by TLC (Rf ) 0.4, 40% NH4OH/CH3OH). The
reaction mixture was then warmed to room temperature and
partitioned between a saturated aqueous Na2CO3 solution and
dichloromethane. The organic layer was separated, dried over
anhydrous Na2SO4, filtered, and concentrated under reduced
pressure to afford the crude product. The crude product was
subjected to flash chromatography on silica gel and eluted with
5% NH4OH/CH3OH to give di-BOC amine 18 (14.67 g, 75%).28
18: Rf ) 0.30 in 5% NH4OH/CH3OH; 1H NMR (CDCl3) δ 5.19
(broad s, 2H), 3.20 (q, 4H), 2.62 (m, 8H), 1.62-1.25 (m, 26H,
this chemical shift range includes the signal at 1.40 (s, 18H)).
N5-(5-(9-Met h yla m in oa n t h r a cen yl)p en t yl)-N1,N8-b is-
(ter t-bu toxyca r bon yl)sp er m id in e (15). 9-Anthraldehyde
(0.305 g, 1.47 mmol) was dissolved in 20% dry MeOH/CH2Cl2
(1.5 mL), and amine 10 (0.7 g, 1.63 mmol, dissolved in the

Effects of Polyamine Homologation
J . Org. Chem., Vol. 65, No. 18, 2000 5599
Anal. Calcd for C20H42N4O4: C, 59.67; H, 10.52; N, 13.92.
Found: C, 59.37; H, 10.31; N, 13.74.
Calcd for C24H51N5O4‚ 0.7 H2O: C, 59.28; H, 10.86; N, 14.40.
Found: C, 58.99; H, 10.43; N, 14.32.
N4-(3-Cya n op r op yl)-N1,N12-b is(ter t-b u t oxyca r b on yl)-
sp er m in e 19. Amine 18 (4.23 g, 10.5 mmol) and 4-bromobu-
tyronitrile (1.55 g, 10.5 mmol) were dissolved in acetonitrile
(10 mL). Potassium carbonate (1.74 g, 12.6 mmol) was then
added. The mixture was refluxed with constant stirring. TLC
(15% NH4OH/CH3OH, Rf ) 0.45) was used to monitor the
consumption of the starting material. After 24 h the reaction
mixture was concentrated, redissolved in dichloromethane and
washed with aqueous Na2CO3. The organic layer was sepa-
rated, dried over anhydrous Na2SO4, filtered, and concen-
trated. The crude residue was subjected to flash chromatog-
raphy and eluted with 1% NH4OH/CH3OH to give 19 (2.54 g,
73% after the recovery of unreacted starting material). 19: Rf
) 0.20 in 1% NH4OH/CH3OH; 1H NMR (CDCl3) δ 5.40 (broad
s, 2H), 3.15 (m, 4H), 2.85-2.30 (broad m, 12H), 1.82-1.15 (m,
28H, this chemical shift range includes the signal at 1.40 (s,
18H)). Anal. Calcd for C24H47N5O4‚0.5 H2O: C, 60.22; H, 10.11;
N, 14.63. Found: C, 60.35; H, 9.92; N, 14.41.
N4-(4-(9-Am in oa cr id in yl)bu tyl)-N1,N12-bis(ter t-bu toxy-
ca r bon yl)sp er m in e 21. 9-Chloroacridine (0.06 g, 0.26 mmol)
and phenol (0.20 g, 2.09 mmol) were heated to 50 °C for 20
min. Amine 20 (0.15 g, 0.32 mmol) dissolved in hot phenol was
added to the stirred mixture. The reaction mixture was heated
to 100 °C. The disappearance of starting amine was monitored
by TLC (10% NH4OH/CH3OH, Rf ) 0.20). After 20 min, the
reaction mixture was concentrated, redissolved in CH2Cl2 and
washed with aqueous Na2CO3 (pH ) 10). The organic layer
was separated, dried over anhydrous Na2SO4, filtered, con-
centrated, and purified by flash chromatography (eluting with
2% NH4OH/CH3OH) to give the acridine derivative 21 (0.12
g, 70%). 21: Rf ) 0.24 in 2% NH4OH/CH3OH; 1H NMR (CDCl3
+ 25% CD3OD) δ 8.40 (d, 2H), 7.92 (d, 2H), 7.80 (t, 2H), 7.45
(t, 2H), 4.05 (t, 2H), 3.12 (m, 4H), 2.78 (m, 4H), 2.49 (broad m,
6H), 2.01-1.30 (broad m, 30H, this chemical shift range
includes a signal at δ 1.40 (s, 18H)); HRMS m/z calcd for
C
37H59N6O4 (M + H) 651.4598, found M + H 651.4561.
N4-(4-Am in obu tyl)-N1,N12-bis(ter t-bu toxycar bon yl)sper -
m in e 20. Nitrile 19 (2.54 g, 5.4 mmol) was dissolved in
absolute EtOH in a Parr shaker bottle. Concentrated NH4OH
(5 mL) was added to the solution followed by the addition of
Raney nickel catalyst (50% slurry, 3.50 g). NH3 gas was then
passed through the solution at 0 °C for 20 min. The reaction
mixture was placed in a Parr shaker, briefly evacuated, and
H2 gas introduced at a pressure of 75 psi. The disappearance
of starting material was monitored by TLC (1% NH4OH/CH3-
OH, Rf ) 0.20). The reaction took 78 h to complete. The
suspension was filtered and the recovered catalyst was washed
with absolute EtOH. The filtrate was concentrated and redis-
solved in CH2Cl2. The CH2Cl2 solution was washed with
aqueous Na2CO3 (pH ) 10), separated, dried over anhydrous
Na2SO4, filtered and concentrated. The product was purified
by flash chromatography and eluted with 16% NH4OH/CH3-
OH to give the desired amine 20 (2.4 g, 95%). 20: Rf ) 0.25 in
Ack n ow led gm en t. This research was supported by
an award from Research Corporation. We gratefully
acknowledge the additional financial support by the
University of Central Florida Sponsored Research In-
House Award, the Elsa U. Pardee Foundation, and the
Florida Hospital Gala Endowed Program for Oncologic
Research. The 200 MHz NMR spectrometer was pur-
chased by a grant from the National Science Foundation
(CHE-8608881). All mass spectral data were generously
provided by Dr. David H. Powell at the University of
Florida. We also acknowledge the assistance of Pooja
Varshneya in the preparation of this manuscript.
Su p p or tin g In for m a tion Ava ila ble: An 1H NMR spec-
trum of 17 in CD3OD is provided as additional proof of purity.
This material is available free of charge via the Internet at
http://pubs.acs.org.
1
10% NH4OH/CH3OH; H NMR (CDCl3) δ 5.60 (broad s, 1H),
5.20 (broad s, 1H), 3.20 (m, 4H), 2.69 (broad m, 4H), 2.40 (broad
m, 8H), 1.80-1.35 (broad m, 30H, this chemical shift range
includes the signal at 1.43 (s, 18H)); HRMS m/z calcd for
C
24H51N5O4 (M + H) 474.4019, found M + H 474.4001. Anal.
J O0002792