Synthesis and Bioevaluation of Macrocycle-Polyamine Conjugates as Cell Migration Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.7b01222

The motuporamines are natural products isolated from the New Guinea sea sponge Xestospongia exigua. Dihydromotuporamine C contains a large macrocycle and an appended polyamine component and was shown to be both antimetastatic and cytotoxic to human L3.6pl pancreatic cancer cells. A series of macrocycle-polyamine conjugates were prepared, and the sequence of the polyamine component was varied to optimize the antimigration properties (as measured in L3.6pl cells) of this molecular class. A one-carbon spacer between the 15-membered carbocycle and the appended polyamine showed improved antimigration properties. A survey of different polyamine sequences containing two, three, or four carbon spacers revealed that the natural polyamine sequence (norspermidine, a 3,3-triamine) was superior in terms of inhibiting the migration of L3.6pl cells in vitro. An investigation of the respective ceramide and sphingomyelin populations in L3.6pl cells revealed that these molecules can modulate both ceramide and sphingomyelin pools in cells and inhibit cell migration.

Full text of DOI:10.1021/acs.jmedchem.7b01222

Article
pubs.acs.org/jmc
Synthesis and Bioevaluation of Macrocycle−Polyamine Conjugates
as Cell Migration Inhibitors
Kristen Skruber, Kelvin J. Chaplin, and Otto Phanstiel, IV*
University of Central Florida, College of Medicine, 12722 Research Parkway, Orlando, Florida 32826, United States
*
S Supporting Information
ABSTRACT: The motuporamines are natural products isolated from the
New Guinea sea sponge Xestospongia exigua. Dihydromotuporamine C
contains a large macrocycle and an appended polyamine component and
was shown to be both antimetastatic and cytotoxic to human L3.6pl pancreatic
cancer cells. A series of macrocycle−polyamine conjugates were prepared, and
the sequence of the polyamine component was varied to optimize the
antimigration properties (as measured in L3.6pl cells) of this molecular class. A
one-carbon spacer between the 15-membered carbocycle and the appended polyamine showed improved antimigration
properties. A survey of different polyamine sequences containing two, three, or four carbon spacers revealed that the natural
polyamine sequence (norspermidine, a 3,3-triamine) was superior in terms of inhibiting the migration of L3.6pl cells in vitro. An
investigation of the respective ceramide and sphingomyelin populations in L3.6pl cells revealed that these molecules can
modulate both ceramide and sphingomyelin pools in cells and inhibit cell migration.
INTRODUCTION
upregulating import of exogenous polyamines to escape this
blockade.
■
(
10
In 1998, the Andersen group isolated a series of compounds
motuporamines) from the sea sponge Xestospongia exigua and
showed that they inhibit the invasiveness of MDA-MB-231
breast carcinoma cells in vitro. The motuporamines have two
1
The PTS is important in the development and progression of
metastatic cancers, as it provides a means to import polyamine
metabolites from the tumor microenvironment to maintain the
high levels needed for rapidly dividing cells. The PTS has a
2
key structural features, a large macrocycle and a polyamine
11
wide tolerance for ligands, which can be leveraged to deliver
1
component. In an attempt to optimize these properties,
polyamine vectors tethered to antimetastatic and cytotoxic
agents. Therefore, it is possible to selectively target cancer cells
via their increased requirement for polyamines and active
Andersen et al. synthesized over 40 derivatives of the original
isolates and found dihydromotuporamine C (1, Figure 1) to be
2
the most potent. Derivatization of these isolates included
12
import processes.
changes to the degree of saturation and size of the macrocyclic
ring as well as changes to the polyamine component itself.
Previously, the Anderson group probed the nature and size of
2
the macrocyclic ring as well as the distance between a 13-
The fact that the motuporamines contain polyamines within
their structure may provide a way to target them specifically to
cancer cells via the high polyamine import activity often present
2
membered heterocycle and the tethered polyamine. This prior
work demonstrated that the 15-membered macrocycle provided
an improved balance between cytotoxicity and inhibition of
invasion of MDA-231 breast cancer cells and the presence of
3
,4
in human cancers. The native polyamines (putrescine 4h,
spermidine 4i, spermine 4j; Figure 2) are low molecular weight
aliphatic amines that are positively charged at physiological pH.
They are crucial for chromatin condensation, DNA replication,
RNA synthesis, and in the translation of mRNA into
2
aminopropyl tethers led to improved properties. The authors
did not, however, assess the ability of these materials to target
the polyamine transport system. Indeed, alteration of the
polyamine motif is a strategy that should improve targeting but
only if the polyamine conjugate is recognized by the PTS.
Surprisingly, compound 2a showed a significant increase in
antimetastatic efficacy but demonstrated no improvement in
5
−7
protein.
Polyamines are particularly of interest to the field
of oncology, as an increase in intracellular polyamine content is
concomitant with the initiation of cancer and is maintained
throughout oncogenesis. The intracellular content of poly-
8
12
PTS targeting.
amines in cells is under tight control. Polyamines can be
biosynthesized from amino acid precursors, as well as imported
Here we describe the synthesis of additional analogues of the
motuporamines. For example by synthesizing longer tether
lengths, such as compound 2b, we can further define the
relationship between tether length and the antimigration
properties of this class of compounds. Other analogues were
from the environment via the polyamine transport system
8
(
PTS). Several compounds have been developed to inhibit
9
these key biochemical targets. For example, difluoromethyl-
ornithine (DFMO) is an inhibitor of ornithine decarboxylase
(ODC), the rate limiting enzyme in putrescine biosynthesis.
Received: August 18, 2017
However, cancer cells often respond to DFMO therapy by
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.jmedchem.7b01222
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 1. Dihydromotuporamine C (1), synthesized motuporamine derivatives (2a, 2b, and 3), 2-hydroxyoleic acid (a sphingomyelin synthase
agonist), imipramine and desipramine (acid sphingomyelinase inhibitors), and MotuN33 (2c).
Figure 2. Structures of polyamines. Norspermidine (4a), tert-butyl (2-aminoethyl)carbamate (4b), N-(2-aminoethyl)ethane-1,2-diamine (4c), tert-
butyl (2-aminoethyl)(2-((tert-butoxycarbonyl)amino)ethyl)carbamate (4d), N1,N1′-(ethane-1,2-diyl)bis(ethane-1,2-diamine) (4e), tert-butyl (2-
aminoethyl)(2-((tert-butoxycarbonyl)(2-((tert-butoxycarbonyl)amino)ethyl)amino)ethyl)carbamate (4f), tetraethylene pentaamine (4g), and the
native polyamines putrescine (4h), spermidine (4i), and spermine (4j).
a
Scheme 1
a
Reagents and conditions: (a) CH PPh I, BuLi, THF, 0 °C; (b) BH −THF, 0 °C, then, H O , 3 M NaOH, rt; (c) CH SO Cl, TEA, 0 °C, then rt;
3
3
3
2
2
3
2
(
d) KCN, 18-crown-6, dry CH CN, reflux; (e) LiAlH , THF, 0 °C, then reflux; (f) PCC, CH Cl , rt; (g) PBr , rt for 1.5 h, then reflux for 1.5 h.
3 4 2 2 3
synthesized utilizing ethylene amine motifs (compounds 3a−c)
for comparisons. These short polyamine motifs are charge-
deficient analogues of the native polyamines and have been
shown to interact with established oncogenic targets such as
polyamine motif influence PTS targeting, cell migration, and
sphingolipid metabolism.
RESULTS AND DISCUSSION
13,14
■
eIF-5a and telomerase.
In summary, since the parent 1 and
Synthesis. A number of synthetic approaches were
investigated using the commercially available ketone cyclo-
pentadecanone 5a (Sigma-Aldrich). The synthesis of the 15-
compound 2a were also shown to enhance intracellular
12
sphingomyelin pools, the new analogues described herein
were utilized to probe how structural alterations in the
B
DOI: 10.1021/acs.jmedchem.7b01222
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 2
a
Reagents and conditions: (a) TEA, MsCl, 0 °C, then rt overnight; (b) amine 8, Na CO , rt, 48 h, 20% yield; (c) EtOH, 4 M HCl, 0 °C, then rt.
2
3
a
Scheme 3
a
Reagents and conditions: (a) CH CN, K CO , 50 °C, 6 days, 12% yield; (b) 4 M HCl, EtOH, rt.
3
2
3
membered macrocycle by Furstner was considered, but it was
that the performance of 2a was predicated upon the increased
15
reported to suffer from low yield. As such, we adopted a
strategy that utilized the preformed macrocycle 5a. Our group
also developed a multistep pathway for synthesis of the parent
availability of the polyamine message for its putative receptor;
12
however, 2a had poor PTS targeting. The synthesis of 2b was,
therefore, an attempt to understand how further extending this
message by increasing the number of methylene spacers (from
two to three) would affect PTS targeting and the antimigration
properties.
1
6,17
compound 1
and used ketone 5a and Boc-protected
12
polyamines to make carbocyclic derivatives. This latter
approach was facilitated by initial conversion of the ketone to
its corresponding alkene followed by anti-Markovnikov
addition of borane to give after workup the primary alcohol
We envisioned synthesis of 2b via the N-alkylation of
macrocyclic amine 8 with the electrophilic polyamine
component 10. Therefore, a selectively Boc-protected diamine
9 was constructed for this purpose. As shown in Scheme 2, the
synthesis of 2b began with the conversion of the Boc-protected
polyamine 9 to its corresponding mesylate 10 in 80% yield. The
N-alkylation of mesylate 10 with amine 8 was performed in the
presence of Na CO in DCM and resulted in poor yield of 11
1
2
5
b. From alcohol 5b a number of synthetic routes were
attempted to use the alcohol as a platform to create an
electrophilic or a nucleophilic reactant. For example, alcohol 5b
was previously converted to its corresponding aldehyde 5c in a
12
reaction that unfortunately resulted in low yields. While
extended compound 2a could be made from the aldehyde,
other chemistries were investigated here in an attempt to
improve the yields for extended motuporamine structures with
longer tether lengths. In this study, alcohol 5b was converted to
extended primary amine 8 which could behave as a nucleophile
to attack an electrophilic polyamine scaffold, as shown in
Schemes 1 and 2. The alcohol was also readily converted to its
bromide 5d through phosphorus tribromide (Scheme 1) for
use in Scheme 4.
2
3
(20%). The low yield in the synthesis of 11 was attributed to
the facile formation of a self-cyclized byproduct 12 where (due
to the low reactivity of the macrocyclic amine 8) the terminal
carbamate group of 10 was observed to react with and displace
the appended mesylate group to form byproduct 12. This
lowered the amount of 10 available to fully convert 8 to 11
(step b in Scheme 2). Efforts to facilitate this reaction by
heating to reflux further increased byproduct formation at the
expense of the product 11. A similar conversion of 3-((tert-
butoxycarbonyl)amino)propyl methanesulfonate to 2-(tert-
butoxy)-5,6-dihydro-4H-1,3-oxazine has been observed by
As shown in Scheme 1, alcohol 5b was converted to its
mesylate 6 using methanesulfonyl chloride/TEA in DCM in
good yields. Mesylate 6 was then converted to nitrile 7 using
18
18-crown-6 ether and KCN with a yield of 87%. The resultant
Brown. Nevertheless, the isolated 11 was converted to 2b
nitrile 7 was then reduced to primary amine 8 using lithium
aluminum hydride in THF in 49% yield.
in 74% yield.
To create the ethylene amine based systems (3a−d), several
synthetic strategies were tried. Due to the low yields of 2b and
the self-cyclization of the polyamine component, alternative
chemistries were investigated that reversed the polarity of the
coupling partners so that a nucleophilic polyamine component
could be joined to an electrophilic macrocycle. The bulky
ketone was very unreactive, and numerous attempts to convert
it into other useful functionalities gave very poor yields. For
The synthesis of 2b was an attempt to further probe the
effects of extending the norspermidine “message” away from
the macrocycle core by increasing the number of methylene
spacers (Scheme 2). Previously, we demonstrated that
compound 2a had the best in vivo performance in terms of
antimetastatic efficacy in a mouse model of metastatic
12
pancreatic cancer using L3.6pl xenografts. It was hypothesized
C
DOI: 10.1021/acs.jmedchem.7b01222
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
these reasons, mesylate 6 was selected for coupling to a series of
Boc-protected ethylene polyamines (e.g., 4b, 4d, and 4f).
Attempts to join 6 with either the Boc- protected triamine 4d
or the Boc-protected tetramine 4f failed due the formation of a
self-cyclized byproduct. This was again due to the poor
reactivity of the bulky macrocycle, where the polyamine reagent
cyclized upon itself likely via urea formation due to the high
temperatures and prolonged reaction times needed to drive this
reaction forward. Nevertheless, the mono-N-Boc diamine 4b
was available commercially (Sigma-Aldrich) and was the only
Boc-protected ethylene amine motif to be used successfully in
forming the desired structure.
separated by first installing Boc groups at every secondary
and primary amine center via excess di-tert-butyl dicarbonate.
Since the tertiary amine was not N-tert-butylcarbonylated, the
byproduct retained its amine functional group, whereas the
desired product was converted to a polycarbamate motif. This
change in functional group allowed for easy separation by
column chromatography of the undesired tertiary amine
byproduct. By use of this approach, a 25% yield was achieved
for each of the desired products.
19
The use of the pentamine 4g to create 3d with this method
gave a different result. The free base of pentamine 4g was
prepared, and mesylate 6 was added in CH CN, refluxed for 72
3
1
As shown in Scheme 3, the Boc-protected polyamine 4b was
heated with 6 at 50 °C overnight and then for 5 days at reflux
h, and monitored by H NMR. The mixture was then treated
with di-tert-butyl dicarbonate, worked up and the crude
(
80 °C). The product 13 (53% yield) was separated by column
separated by column chromatography (1% NH OH/4.5%
4
chromatography (7% MeOH/DCM R of 13: 0.4) from a
MeOH/DCM, R suspected product 0.4 and an additional
f
f
cyclized urea byproduct that was present in large quantities.
Indeed, facile urea formation has been observed in related
byproduct at R = 0.47). This approach was challenging as not
f
all the amine centers were tert-butylcarbonylated likely due to
steric constraints. Attempts to use Boc-group introduction to
drive the reaction to completion resulted in undesired self-
cyclization where the secondary amine cyclized back on an
existing carbamate group to form a urea linkage. Multiple
attempts to separate the product failed, and ultimately the
major isolated product appeared to be a cyclized byproduct by
NMR and was unresolvable from its upper spot. The crude
weight also suggested it was a significantly less productive
reaction due to the low conversion of starting material at high
temperatures and the long reaction period. Therefore, there
were clear limitations to using this unprotected approach for
the longer polyamine systems like 4g. Nevertheless, the method
did provide the desired systems 3b and 3c (Scheme 4).
19
systems in the presence of base and heat. Subsequent removal
of the Boc group with 4 M HCl provided the desired adduct 3a
in high yield but in a disappointing 10% overall yield from 6. In
short, the longer Boc-protected ethylene amine motifs appeared
to have poor reactivity and readily displaced their own tert-butyl
groups to form cyclic ureas in favor of reacting with the
mesylate 6.
To avoid these steric constraints, we devised a “naked”
polyamine approach using the free bases of 4c, 4e, and 4g
(Scheme 4). This approach was initially not pursued in order to
a
Scheme 4
1
N -Anthracen-9-ylmethylethane-1,2-diamine 17 was gener-
ated as a control to evaluate the role of the large N-alkyl
substituent by comparison to 3a. Compound 17 was based on
prior derivatives, which showed that anthrylpolyamines readily
20
entered cells through the PTS. As shown in Scheme 5, the
Boc-protected ethylene amine motif 4b was joined via reductive
amination to the commercially available 9-anthraldehyde
(
Sigma-Aldrich) and resulted in a 75% yield of 17 over three
steps via the intermediate imine.
Biological Evaluation. After their syntheses were com-
plete, the compounds were screened for cytotoxicity in Chinese
hamster ovary (CHO), CHO-MG (a CHO mutant defective in
polyamine transport), and L3.6pl human pancreatic cancer
cells. The wild type (wt) CHO cells have high polyamine
transport activity and were very sensitive to compounds that
exploit the PTS for cell entry. In contrast, the CHO-MG cell
line was previously generated by Flintoff et al. via random
DNA-alkylation of CHO cells and subsequent cell selection for
the ability to survive in the presence of a cytotoxic PTS
targeting compound, methylglyoxal bisguanyl hydrazine
a
Reagents: (a) CH CN, K CO , refluxed 48 h for 14a and 72 h for
4b; (b) di-tert-butyl dicarbonate, rt; (c) EtOH, 4 M HCl, rt.
3
2
3
1
avoid reaction of the internal secondary amine(s) with the
mesylate, which would form the undesired tertiary amine
branched structures instead of the desired linear polyamine
architectures. Nevertheless, this naked polyamine approach
avoided the possibility of the polyamine portion reacting with
an attached Boc group prior to N-alkylation. Indeed, the
reaction was much more facile without the steric contribution
of the N-Boc substituents and was successful with 4c and 4e,
while 4g was not productive. As shown in Scheme 4, the
production of 3b involved the reaction of triamine 4c with
bromide 5d. The related derivative 3c was generated from the
reaction of tetraamine 4e with mesylate 6. With both 4c and 4e,
21
(MGBG). The CHO-MG cells have defective polyamine
transport and are significantly less sensitive to cytotoxic
22
polyamine compounds. L3.6pl cells are hyper-metastatic
human pancreatic cancer cells, trained for their ability to
migrate from the pancreas to the liver and provide a model of
23
very aggressive pancreatic cancer cells primed for metastasis.
Polyamine Transport Selectivity Studies. The CHO
(CHO-K1, ATCC) and CHO-MG cell screen uses cytotoxicity
11
the respective N-alkylations were performed in CH CN and
showed complete disappearance of each starting material after
measurements to assess transport preference. Since the CHO-
3
22
MG mutant cell line is defective in polyamine transport,
72 h. After workup, the expected mixture of secondary and
compounds that selectively enter via the PTS should be less
tertiary N-alkylated products was obtained. These were
toxic to these cells and give a high CHO-MG IC50 value. In
D
DOI: 10.1021/acs.jmedchem.7b01222
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 5
Table 1. Biological Evaluation of Motuporamine Derivatives
on Cytotoxicity in CHO and CHO-MG Cells To Assess PTS
Targeting at 48 h
a
IC50 (μM)
compd
CHO-MG
CHO
IC50 ratio CHO-MG/CHO
b
1
2.96 ± 0.10
2.80 ± 0.11
5.95 ± 0.50
2.37 ± 0.12
2.55 ± 0.07
2.62 ± 0.13
>100
2.90 ± 0.20
2.51 ± 0.06
2.84 ± 0.20
2.66 ± 0.11
2.48 ± 0.10
2.37 ± 0.09
>100
1
2
2
3
3
3
5
1
b
1.1
2.1
0.9
1
b
c
a
b
c
1.1
ND
5.8
b
7
11.33 ± 0.34
1.96 ± 0.11
a
All compounds were dosed as aqueous solutions and were compared
to a media control with no compound added, except compound 5b.
Compound 5b was not soluble in water and was dosed in such a
manner that the final DMSO concentration was 1% DMSO and was
thus compared to a separate 1% DMSO in PBS control with no
compound added. For all replicates, n = 3. ND = not determined, due
to low toxicity of the compound at 100 μM. Note that an
anthracenylmethylhomospermidine control (Ant44, bottom in Scheme
14 b
5
) gave IC50 ratio CHO-MG/CHO of 148. Note that compound 2a
10
also gave CHO-MG/CHO IC50 ratio of 1.
L3.6pl Cytotoxicity Studies. Compounds 2b, 3a−c, 5b,
and 17 were tested for cell growth inhibition in human L3.6pl
pancreatic cancer cells. The known K-Ras mutation present in
this cell line has been shown to correlate with high polyamine
uptake due to its ability to affect caveolin-mediated
a
Reagents and conditions: (a) 25% MeOH/DCM, rt; (b) NaBH , 0
25
4
endocytosis, which leads to an increase in uptake of
°C, then rt; (c) EtOH, 4 M HCl, rt.
polyamine compounds. For this study, L3.6pl cells were treated
with a range of compound concentrations (0.1 μM to 100 μM)
for 48 h in the presence of aminoguanidine (AG, 250 μM).
Prior studies have shown that this amine oxidase inhibitor
(AG) was needed to maintain the polyamine compound’s
contrast, PTS-dependent compounds should be lethal to wt
CHO cells (which have high PTS activity) and give low CHO
IC values. A ratio of the CHO-MG/CHO IC values is then
5
0
50
12
used to assess PTS targeting. A compound that does not enter
cells via the PTS should give similar toxicity in both cell lines
and an IC50 ratio near 1. In contrast, a compound that targets
the PTS should give a high CHO-MG/CHO IC50 ratio.
Compounds 2b, 2c, 3a−c, 5b, and 17 were tested in the
CHO and CHO-MG cell lines at a range of concentrations
from 0.1 μM to 100 μM for 48 h in the presence of
aminoguanidine (1 mM). Aminoguanidine is an inhibitor of
amine oxidases present in calf serum, which readily oxidize
potency. Cell viability was evaluated via the tetrazolium
compound [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxy-
phenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt (MTS
reagent from Promega), and the IC and IC values were
5
50
calculated from the respective plots of % viability versus
concentration of compound. Each value represents the
concentration of the compound needed to produce the
designated level of toxicity (n = 3). The IC value is the
5
concentration of the compound that gives 95% viability and
represents the dose where minimum toxicity (5% toxicity) from
the compound is expected. The IC value is the concentration
12
amine-containing compounds. As shown in Table 1, the
motuporamine derivatives did not target the PTS in CHO cells
and gave ratios near 1. In contrast, the N -(anthracenylmethyl)
50
1
of the compound needed to reduce viability by 50% compared
to an untreated control with vehicle (phosphate buffered saline,
PBS). The results are shown in Table 2.
20
homospermidine derivative Ant44 and diamine 17 gave a
20
CHO-MG/CHO IC50 ratios of 148 and 5.8, respectively.
These findings are consistent with previous studies that showed
that the parent compound, dihydromotuporamine C, did not
The new motuporamine derivatives (2b, 3a−c) displayed
sharp cytotoxicity curves (see Supporting Information) with
their respective IC values very close to their IC concentration
24
use the PTS for cellular entry. The results of this study
suggest that the appended aliphatic macrocycle is not
conducive for PTS targeting. Indeed, modulation within the
polyamine message itself and the distance between the
polyamine message and the macrocycle did not improve PTS
targeting in this homologous series. For example, we noted that
the polyamine message 4b, when appended to the anthracene
substituent (17), was able to accomplish modest PTS targeting
5
50
and reflect trends observed with other members of this
12,16
series.
Interestingly, the toxicity of the new motuporamine
derivatives was not dramatically altered through modulation of
the polyamine message. However, the absence of the polyamine
component (such as in 5b) or the addition of an unsaturated
anthracene core in place of the aliphatic macrocycle (e.g., 17)
was sufficient to decrease the toxicity. In the case of 17, PTS
targeting was also improved (Table 1). These results suggest
that the specific combination of polyamine and macrocycle is
responsible for the cytotoxicity of these compounds. We noted
(CHO-MG/CHO IC ratio 5.8) while the related macrocycle-
50
containing compound 3a did not (CHO-MG/CHO IC50 ratio
.9).
0
E
DOI: 10.1021/acs.jmedchem.7b01222
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 2. Cytotoxicity Evaluation of Motuporamine
Table 3. Inhibition of L3.6pl Cell Migration by
Derivatives (2b, 3a−c, 5b) and Anthryl Derivative (17) in
Motuporamine Derivatives (1−3 and 5b), Ethylenediamine
,
ab
L3.6pl Cells for 48 h
and Unsubstituted Polyamines 4c and 4e, and Anthracene
, ,
a b c
Derivative (17)
compd
L3.6pl, 48 h, IC50 (μM)
L3.6pl, 48 h, IC (μM)
5
b
%
migration
1
0.99 ± 0.07
89.4 ± 5.4
3.34 ± 0.09
1.44 ± 0.09
1.25 ± 0.05
1.64 ± 0.05
42.0 ± 3.4
24.1 ± 1.2
0.60 ± 0.04
80.0 ± 4.0
1.64 ± 0.09
0.6 ± 0.09
0.6 ± 0.05
0.92 ± 0.05
8.4 ± 0.6
b
% cell
% migration
inhibition
2
2
3
3
3
5
1
a
b
a
b
c
migration at normalized to
compared to
a
a
a
compd (concn)
24 h
the control
control
1
% DMSO in PBS
control
44.3 ± 2.5
b
b
10% aqueous control
1 (0.5 μM)
72.3 ± 8.8
57.4 ± 5.0
43.6 ± 4.4
57.7 ± 1.6
50.7 ± 6.8
83.5 ± 3.2
71.2 ± 4.2
67.8 ± 5.6
53.3 ± 13.0
64.9 ± 7.5
52.9 ± 3.7
67.9 ± 8.2
60.2 ± 5.5
100
b
b
b
7
79.4 ± 6.9
60.2 ± 6.6
79.8 ± 6
20.6 ± 6.9
4.15 ± 0.2
1 (1 μM)
39.8 ± 6.6
20.2 ± 6
29.9 ± 7.8
−15.5 ± 6
1.5 ± 6.5
6.2 ± 7.2
26.2 ± 10
10.2 ± 8.2
26.8 ± 6.3
6.1 ± 8.5
16.8 ± 7.2
0
a
2a (0.5 μM)
All compounds were dosed as aqueous solutions and were compared
2
a (1 μM)
70.1 ± 7.8
115.2 ± 6
98.5 ± 6.5
93.8 ± 7.2
73.7 ± 10
89.8 ± 8.2
73.2 ± 6.3
93.9 ± 8.5
83.2 ± 7.2
to a media control with no treatment, except compound 5b. L3.6pl
cells were incubated with 250 μM aminoguanidine (AG) for 24 h prior
to addition of compound. Compound 5b was not soluble in water and
was dosed in such a manner that the final DMSO concentration was
% DMSO and was thus compared to a 1% DMSO in PBS control
with no compound. For all replicates, n = 3. Reference 12.
2b (0.5 μM)
2b (1 μM)
3a (0.5 μM)
1
3
3
3
3
3
a (1 μM)
b (0.5 μM)
b (1 μM)
c (0.5 μM)
c (1 μM)
b
that no cell lysis was observed at the IC50 concentrations of
these compounds, and direct cell lysis was discarded as a
potential mechanism of action for these amphiphilic molecules.
Antimigration Assay. A scratch assay was performed to
assess the antimigratory properties of compounds 2b, 3a−c, 5b,
and 17 in L3.6pl cells. The parent compound 1 and extended
compound 2a were tested as controls, and non-native
polyamine motifs 4c and 4e and ethylene diamine HCl were
also tested to understand the effect of unsubstituted polyamine
structures. It was critical that these cell migration experiments
be conducted at compound concentrations that were not toxic
to the L3.6pl cells to avoid the interpretation of cell death as
migration inhibition. Therefore, the compounds were dosed at
ethylenediamine·HCl
(1 μM)
4c (1 μM)
4e (1 μM)
5b (1 μM)
17 (0.5 μM)
100
0
100
0
60.7 ± 1.6
67.8 ± 4.0
137.2 ± 2.1
93.8 ± 6.4
−37.2 ± 2.1
6.2 ± 6.4
12
a
All compounds were dosed as aqueous solutions and were compared
to the 10% water control, except compound 5b. Compound 5b was
not soluble in water and was dosed in DMSO in such a manner that
the final DMSO concentration was 1% DMSO and was thus compared
to the 1% DMSO in PBS control. For all values n = 3 with the
exception of those marked otherwise. n = 4. L3.6pl cells were
incubated for 24 h with 250 μM AG prior to compound addition. For
additional details about the antimigration essay, see the Supporting
Information. Note that similarly compound 2c was previously reported
to inhibit 19.3% (±1.3) of L3.6pl cell migration at 0.6 μM after a 24 h
b
c
the IC value (0.5 μM) of the parent compound 1 to directly
5
compare their potencies at a nontoxic concentration. We also
tested them at 1 μM which was close to the MTD of 3b, and
some toxicity may be seen at this higher dose. All experiments
were run in triplicate and representative microscopy images are
shown in their respective figures in the Supporting Information
and the numeric results are shown in Table 3.
12
incubation.
μM concentration but did not surpass the antimigration efficacy
seen with the parent compound 1.
As mentioned earlier, the Andersen group did not probe the
effect of changing the distance between the ring system and the
The unsubstituted non-native ethylene amine motifs (4b, 4c,
and 4e) were also assessed for their antimigratory properties via
a scratch assay. With incubation of each polyamine for 24 h at 1
μM, however, these compounds showed an increase in
migration compared to control and complete wound closure
at 24 h. This experiment defines the importance of both the
macrocycle (e.g., 5b) and the polyamine (e.g., 4b) for efficacy
of these compounds because independently each component is
less effective than the conjugate containing both the macrocycle
and the amine motif (e.g., 3a).
12
polyamine message. The synthesis of 2a by Muth et al. and
b here allows for an understanding of the effect of increased
2
distance on cell migration inhibition. Interestingly, antimigra-
tion efficacy is retained in compounds 1 and 2a at 0.5 μM, but
appears to drop off drastically with the two methylene spacer
extension 2b (Table 3).
Compounds 5b and 17 were tested well below their
maximum tolerated concentration (IC ) in order to comment
5
on their efficacy in comparison to the parent compound.
Neither control showed any antimigration behavior (see Table
Sphingolipid Metabolism Modulation. Prior work in
yeast suggested that these materials may also affect sphingolipid
26
3
). Interestingly, the extended system 2b and the macrocyclic
metabolism. Indeed, a heterozygous diploid yeast strain
lacking one copy of the genes associated with sphingolipid
metabolism (LCB1 and TSC10) was found to be more sensitive
alcohol 5b appear to have no antimigration effect at 0.5 μM.
As both compound 2b and the 3a−c series exhibit sharp
cytotoxicity curves (Supporting Information), the window
between the IC value and IC concentration was very small
26
to 1 than a wild-type strain. Therefore, compounds 1, 2a, and
3c were studied for their ability to modulate sphingolipid and
ceramide pools in L3.6pl human pancreatic cancer cells. As
shown in Table 4, the parent compound 1 was unique in its
ability to increase specific ceramide pools (N16:0 and N24:1)
when dosed at 500 nM for 48 h at 37 °C, which may explain its
5
50
and a dose-limiting toxicity was observed. All compounds were
also tested at 1 μM, a concentration that doubles the potency of
the parent compound 1. In a similar outcome the efficacy for
the ethylene amine based compounds improved at the higher 1
F
DOI: 10.1021/acs.jmedchem.7b01222
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 4. L3.6pl Cell Ceramide Pool Modulation after 48 h
reported to affect cell migration (and because 1 and 2a were
shown to influence sphingomyelin pools). The SGMS assay
utilized a commercially available fluorescently tagged ceramide
12
Incubation with 500 nM of Each Compound at 37 °C
a
(Expressed in nmol/mg Protein)
(
NBD-C12-ceramide, Figure 3) as a substrate for the SGMS
ceramide
species
untreated
control
enzyme present in the whole cell lysates obtained from PanO2
murine pancreatic cancer cells. The results are shown in Figure
4
1
2a
3c
N16:0
N24:1
N24:0
sum
0.08 ± 0.01
0.37 ± 0.06
0.16 ± 0.04
0.62 ± 0.10
0.17 ± 0.03*
0.51 ± 0.03*
0.20 ± 0.00
0.89 ± 0.05*
0.11 ± 0.02
0.37 ± 0.03
0.14 ± 0.02
0.63 ± 0.06
0.09 ± 0.01
0.28 ± 0.02
0.11 ± 0.03
0.49 ± 0.06
and quantified in Table 6.
The data in Figure 4 and Table 6 clearly demonstrated that
these compounds were not agonists of SGMS, at least under
the conditions of our study. An agonist would be expected to
increase the SGMS enzyme’s ability to convert starting
ceramide to sphingomyelin. In addition, one would expect to
see an increase in sphingomyelin production in the presence of
a SGMS agonist in a dose dependent manner. We observed
neither the enhanced production of sphingomyelin nor the
expected dose dependence in sphingomyelin production for the
three motuporamine derivatives tested: parent 1, 2a, and 2c. In
sum, the results obtained with the NBD-C12 ceramide probe
a
Data with asterisk were statistically significant from the untreated
control with p < 0.05.
high toxicity compared to the other derivatives (see Table 2; 1,
1
2
IC = 600 nM). In contrast, this ceramide increase was not
5
observed with 2a and 3c at the 500 nM dose.
Compounds 1 and 2a, which were effective antimigration
agents at 500 nM (Table 3, ∼20% inhibition), significantly
increased specific sphingomyelin pools (Table 5). In contrast,
compound 3c, which had low antimigration properties at 500
nM (Table 3, ∼6% inhibition), resembled the untreated
control. As shown in Figure 3 (panel A), the increased
sphingomyelin pools can be rationalized by either increased
sphingomyelin synthase (SGMS) activity (which generates
sphingomyelin from its corresponding ceramide) or inhibition
of sphingomyelinase (an enzyme which degrades sphingomye-
lin to the corresponding ceramide).
(
i.e., a fluorescent model of N16:0 ceramide, Figure 3)
suggested that SGMS was not the target of motuporamines.
DISCUSSION
■
Both a norspermidine motif 4a and ethylene amine motifs (4b,
4c, and 4e) were used to synthesize derivatives of
dihydromotuporamine C (1). The new derivatives were used
to develop structure−activity relationships that relate molecular
structure to inhibition of cell migration and their ability to
modulate intracellular sphingolipid pools. The original strategy
for construction of these motuporamine targets was based on
using regioselective Boc-protection of polyamines to maximize
formation of linear conjugates between Boc-protected poly-
In an effort to define the target, we modified a sphingomyelin
27
synthase (SGMS) assay and screened the motuporamine
derivatives for their possible agonist activity. Since 3c had little
effect on sphingomyelin pools, we elected to screen the
homologous series 1, 2a, and 2c, which have been previously
Table 5. L3.6pl Cell Sphingomyelin Pool Modulation after 48 h Incubation with 500 nM of Each Compound at 37 °C
a
(Expressed in nmol/mg Protein)
sphingomyelin species
untreated control
1
2a
3c
N14:0
N15:0
N16:1
N16:0
N17:0
N18:1
N18:0
N19:0
N20:1
N20:0
N21:0
N22:2
N22:1
N22:0
N23:2
N23:1
N23:0
N24:2
N24:1
N24:0
N25:1
N25:0
N26:1
sum
0.13 ± 0.01
0.12 ± 0.00
0.28 ± 0.01
2.60 ± 0.06
0.05 ± 0.01
0.07 ± 0.00
0.19 ± 0.01
0.01 ± 0.00
0.02 ± 0.00
0.06 ± 0.01
0.03 ± 0.00
0.01 ± 0.00
0.17 ± 0.01
0.23 ± 0.01
0.01 ± 0.00
0.10 ± 0.00
0.07 ± 0.00
0.30 ± 0.01
1.07 ± 0.03
0.32 ± 0.01
0.02 ± 0.01
0.01 ± 0.00
0.02 ± 0.00
5.91 ± 0.11
0.14 ± 0.01
0.14 ± 0.03
0.13 ± 0.01
0.13 ± 0.02
0.30 ± 0.01
2.84 ± 0.19
0.06 ± 0.00
0.10 ± 0.00**
0.27 ± 0.03*
0.01 ± 0.00
0.03 ± 0.00*
0.09 ± 0.01**
0.04 ± 0.00*
0.02 ± 0.00
0.22 ± 0.01**
0.31 ± 0.02*
0.01 ± 0.00
0.12 ± 0.02*
0.09 ± 0.01*
0.36 ± 0.02*
1.22 ± 0.06*
0.36 ± 0.02*
0.03 ± 0.01
0.02 ± 0.01
0.03 ± 0.00
0.13 ± 0.01
0.12 ± 0.01
0.28 ± 0.02
2.66 ± 0.22
0.05 ± 0.01
0.09 ± 0.01*
0.24 ± 0.02
0.01 ± 0.00
0.02 ± 0.01
0.08 ± 0.01*
0.03 ± 0.00
0.02 ± 0.00
0.19 ± 0.01
0.28 ± 0.03
0.01 ± 0.00
0.11 ± 0.01
0.08 ± 0.00*
0.31 ± 0.03
1.15 ± 0.14
0.32 ± 0.03
0.02 ± 0.00
0.02 ± 0.01
0.03 ± 0.00
6.25 ± 0.53
0.33 ± 0.02*
3.26 ± 0.09**
0.06 ± 0.01*
0.11 ± 0.01**
0.32 ± 0.01**
0.01 ± 0.00
0.03 ± 0.00**
0.10 ± 0.00*
0.05 ± 0.00**
0.02 ± 0.00
0.27 ± 0.02**
0.35 ± 0.01**
0.01 ± 0.00
0.14 ± 0.01**
0.10 ± 0.00**
0.41 ± 0.02**
1.49 ± 0.04**
0.39 ± 0.01**
0.03 ± 0.00
0.03 ± 0.01
0.04 ± 0.00**
7.84 ± 0.21**
b
6.79 ± 0.39
a
All experiments were performed in triplicate. Data entries with one asterisk (p < 0.05) and double asterisk (p <0.01) were compared in a t test (two
b
samples assuming unequal variances, two-tailed) with the untreated control samples. p = 0.06.
G
DOI: 10.1021/acs.jmedchem.7b01222
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Table 6. SGMS Agonism Results for 1, 2a, and 2c
compd
concn (μM)
relative % NBD-sphingomyelin produced
untreated
NA
0.5
1
100
89
75
88
97
75
82
1
1
2
2
2
2
a
a
c
c
0.5
1
0.5
1
a
The reaction mixture containing PanO2 cell lysate, NBD-C -
1
2
ceramide, and the phosphocholine source were combined and stirred
at 37 °C for 3 days. The lipids were then extracted using a 9:1 CHCl3/
MeOH solution and resolved by TLC. The TLC plate was first eluted
with 15% MeOH/CHCl3 (to easily separate the ceramide and
sphingomyelin compounds). The plate was then dried under a stream
of N2 gas and re-eluted with 30% MeOH/CHCl3 to move the
sphingomyelin product cleanly off the baseline. The plate was then
irradiated with two UV lamps, and an image was captured and
analyzed by ImageJ. The relative % NBD-sphingomyelin produced
compared to the untreated control was calculated by first subtracting
out the background (set by the lysis buffer only negative control) to
generate corrected integrated density counts. By use of these corrected
values, the respective pixel counts obtained for the NBD-
sphingomyelin in each lane were divided by the corrected NBD-
sphingomyelin pixel counts obtained in the untreated control (lane 4
in Figure 4) × 100. See the Supporting Information for more detailed
information regarding the ImageJ analysis.
Figure 3. Ceramide and sphingomyelin structures and interconversion
pathways: (A) enzymes involved in sphingomyelin−ceramide
interconversion and the native and NBD-tagged substrates for these
enzymes; (B) structure of NBD-C12 acid, a potential hydrolysis
product of NBD-C -ceramide.
materials. In contrast, N-alkylations using naked polyamines
(without Boc protection) allowed access to the triamine and
tetramine motifs, 3b and 3c, albeit in low yield.
12
The steric hindrance of the bulky 15-membered ring system
seems to be a significant determining factor in synthesizing
motuporamine derivatives. The 15-membered ketone 5a was
relatively inert to a variety of attempts to convert it to other
functional groups. In a similar vein, the low reactivity of
polyamine chains containing nearby bulky Boc substituents also
severely limited the feasibility of the regioselective strategies
employed here. In this regard, the steric demands of the
polyamine component were also important determinants of
reaction success. Indeed, the “naked” polyamine strategy used
here improved the yields of this coupling step. The ethylene
amine motifs are shorter than norspermidine and homospermi-
dine, and the decrease in distance between nitrogen centers
coupled with the closeness in proximity of Boc groups to the
reactive site appears to make this nitrogen-to-carbon coupling
reaction especially challenging. Future investigations into
synthetic processes that lead to higher yields for these systems
are warranted especially for future two carbon polyamine
systems based upon 4e.
Figure 4. SGMS agonism studies. NBD-C12 ceramide was incubated
with PanO2 cell lysates to generate NBD-C12-sphingomyelin in situ in
the presence and absence of motuporamine derivatives. TLC was used
to separate the NBD-tagged starting ceramide and product
sphingomyelin for later analysis by ImageJ (in Table 6). Lane 1:
starting material NBD-C12-ceramide. Lane 2: a potential hydrolysis
byproduct NBD-C12 acid (12-(7-nitrobenzofurazan-4-ylamino)-
dodecanoic acid). Lane 3: product NBD-C12-sphingomyelin. Lane
4
: reaction system + deionized water (positive control). Lane 5:
reaction system + cell lysis buffer only. Lane 6: reaction system +
compound 1 (0.5 μM). Lane 7: reaction system + compound 1 (1
μM). Lane 8: reaction system + compound 2a (0.5 μM). Lane 9:
reaction system + compound 2a (1 μM). Lane 10: reaction system +
compound 2c (0.5 μM). Lane 11: reaction system + compound 2c (1
μM). Note that all compounds (1, 2a, and 2c) were introduced via
stock solutions in deionized water.
The IC50 and PTS targeting determinations for the
synthesized compounds showed that the two carbon motupor-
amine systems 3a−c were unable to target the PTS and had
12,16
sharp cytotoxicity curves.
The effect of the reduced distance
between nitrogen centers had no significant effect on improving
potency when appended to the large cyclopentadecane ring.
Studies with the macrocycle alcohol 5b and the unsubstituted
polyamines revealed that the individual components were
relatively nontoxic. In contrast, the conjugation of the
macrocycle and polyamine components was significantly
more toxic to cells.
amines and the 15-membered macrocycle. Surprisingly, we have
shown that due to steric issues, this strategy was nonproductive.
This is due, in part, to self-cyclization of the Boc-protected
polyamines during the prolonged reaction times and high heat
needed to drive this coupling step between two bulky starting
Studies with 3a and 17 showed that PTS targeting was
dependent upon the N-substituent and may require longer
H
DOI: 10.1021/acs.jmedchem.7b01222
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
1
tethers separating bulky N -substituents from the polyamine
novel starting points for developing antimetastatic medicines
for clinical use. We speculate that changes in sphingomyelin
pools may affect cell migration via downstream Rho activation
and stress fiber formation, outcomes that have both been
message to improve polyamine recognition by cell surface
receptors.
The macrocyclic alcohol 5b and the unsubstituted non-native
ethylene amine motifs (as their respective HCl salts) were all
found to be nontoxic and promigratory. While unable to
improve upon the potency of 1 at 500 nM to inhibit the
migration of L3.6pl cells, compounds 3a−c, nevertheless, were
antimigration compounds, (e.g., see Table 3 for performance of
compounds 3a−c at 1 μM).
The results obtained suggest that both the polyamine and
macrocycle components are required for efficacy as cell
migration inhibitors. Indeed, moving the linear triamine
message of 1 away from the ring (as performed here with 2a
and 2b) significantly affected the ability of the compound to
block cell migration. While 2a (containing a methylene spacer)
retained the antimigration properties of 1, this property was
lost in compound 2b (which contains an ethylene spacer). This
study shows that efficacy can be significantly altered by minor
structural changes in the appended polyamine architecture and
suggests that an optimal placement of the appended nitrogen
centers is necessary to fully interact with the putative biological
target.
30−32
observed with motuporamines.
It remains unclear how
these pathways are precisely connected, but we note that
released sphingosine 1-phosphate (S1P) can bind to one of five
S1P receptors, which in turn lead to differential activation of
distinct G-protein responsive pathways including Rho activa-
33
tion. Regardless of mechanism, compound 2a has been shown
12
to block L3.6pl cell metastasis in vivo.
The ability of motuporamine derivatives to increase intra-
cellular sphingomyelin levels and affect membranes can have
broad biomedical applications beyond cancer and metastasis. In
fact, we recently described the use of motuporamine derivatives
as antimicrobials. We observed that these derivatives can create
small leaks in bacterial membranes and can de-energize the
34
bacteria’s efflux pump without cell lysis. Therefore, future
work will also focus on the ability of these compounds to
increase the potency of existing antibiotics in drug-resistant
microbes as well as inhibit cancer cell migration and alter
sphingomyelin pools.
We speculated that the observed increased sphingomyelin
pools induced by these compounds could be rationalized by
either increased sphingomyelin synthase activity or inhibition of
sphingomyelinase. Indeed, as shown in Figure 1, the motupor-
amine derivatives resemble molecules known to interfere with
enzymes associated with sphingomyelin−ceramide conversion.
For example, 2-hydroxyoleic acid is a known sphingomyelin
synthase (SGMS) agonist, where the 2-hydroxyl group and 18-
EXPERIMENTAL SECTION
■
Materials. Silica gel (32−63 μm) and chemical reagents were
purchased from commercial sources and used without further
purification. All solvents were distilled prior to use or purchased as
analytical grade. All reactions were carried out under atmospheric
1
13
pressure unless an N atmosphere was specified. H and C spectra
2
were recorded at 500 or 125 MHz, respectively. TLC solvent systems
were listed as volume percentages. All tested compounds provided
satisfactory elemental analyses and were tested at ≥95% purity, with
the exception of compound 2b which was tested at 93% purity.
28
carbon chain length were shown to be critical for this activity.
We speculated that motuporamine derivatives could mimic this
agonist by using the distal polyamine headgroup as a bidentate
ligand similar to the α-hydroxycarboxylic acid group in 2-
hydroxyoleic acid. In addition, the conformational restriction
imparted by the cis alkene in 2-hydroxyoleic acid may be
mimicked by the cyclopentadecane ring.
22
Biological Studies. CHO (ATCC) and CHO-MG as well as
23
L3.6pl cells (RRID: CVCL_0384) were grown in RPMI 1640
medium with the addition of 10% fetal bovine serum and 1%
penicillin/streptomycin and grown at 37 °C under a humidified 5%
CO atmosphere. CHO and CHO-MG cells were seeded at 10 000
2
cells/mL, while L3.6pl cells were grown at 5000 cells/mL for
cytotoxicity studies and 55 500 cells/mL for antimigration experi-
ments. Cells were treated 24 h prior to compound addition with
aminoguanidine (AG, a known inhibitor of polyamine oxidase present
in bovine serum) through addition to the growth medium at a
concentration of 1 mM for CHO/CHO-MG cells and 250 μM for
L3.6pl cells. The presence of the amine oxidase inhibitor (AG) was
important as the compounds tested contained polyamines within their
structure and active amine oxidase could cause degradation during the
course of the assay.
IC50 Determinations. Cell viability was assessed in sterile 96-well
plates (Costar 3599, Corning). After an initial overnight incubation,
the respective stock solutions (10×) of each compound (10 μL/well)
were added to 90 μL/well of cell suspension in media with AG (see
above). This provided a total volume of 100 μL in each well. After
addition of the compounds, the plates were incubated for 48 h at 37
As shown in Figure 4 and Table 6, a SGMS activity assay
using a fluorescent C -ceramide substrate failed to show
1
2
significant increases in C -sphingomyelin formation in the
1
2
presence of these putative SGMS agonists (the motuporamine
derivatives) compared to the untreated control. There are
several potential interpretations of this result. The most simple
conclusion is that the motuporamines are not SGMS agonists.
One caveat to this conclusion is that since the motuporamine
compounds modulate specific sphingomyelin pools, it is
possible that SGMS agonism may be observed using ceramides
with other chain lengths (such as C ). Nevertheless, the
1
8
negative result for SGMS agonistic activity here suggests that
the motuporamines likely increase sphingomyelin pools by
affecting other pathways. For example, they may inhibit
sphingomyelinases like acid sphingomyelinase (ASM).
°
C (in a 5% CO atmosphere) and cell viability was assessed via
2
formazan formation (after a 4 h incubation at 37 °C) from the 3-(4,5-
dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfenyl)-2H-
tetrazolium (MTS) reagent (Promega) via absorbance (490 nm) with
a SynergyMx Biotek microplate reader.
Antimigratory Assay. Antimigratory properties were assessed in
sterile 96-well plates (Costar 3599, Corning). After an initial 24 h
incubation of L3.6pl cell suspension in medium (90 μL/well)
containing 250 μM AG, a channel was scratched laterally across the
plate using a 100 μL plastic pipet tip. The cells were then washed with
PBS (1×, 100 μL) and medium was replaced (90 μL/well cell
suspension in medium with 250 μM AG) and the respective solution
of compound was added (10 μL/well). Samples were run in
Indeed, the motuporamines resemble two known functional
29
ASM inhibitors, imipramine and desipramine (Figure 1), both
of which contain a large macrocycle with 15 atoms along its
circumference and an appended amine motif similar to 1.
Future work will investigate whether sphingomyelinases like
ASM are inhibited by motuporamines.
Developing molecules, which target cell migration and
metastasis, is important because many cancer patients die
from metastatic disease. In this regard identifying new scaffolds,
which interfere with cell migration and tumor spread, provides
I
DOI: 10.1021/acs.jmedchem.7b01222
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
quadruplicate. After 24 h, the cells were imaged (Nikon TE 200) with
a 10× objective. To image the same location at two time points,
images were taken next to a predrawn line bisecting each well
perpendicular to the channel scratched. To standardize each image, all
images were cropped in ImageJ software to 700 × 1280 pixels. Cell
migration was assessed by measuring the area devoid of cells over a 24
h period and calculated as the following:
nicely defined bands for quantification purposes. Separate controls
were run to confirm the elution profile of each analyte using pure
NBD-C -ceramide (Sigma) and NBD-C -sphingomyelin (Nova).
12
12
The NBD-C -acid (lane 2 in Figure 4, a potential hydrolysis product
12
of ceramide that would also contain a NBD tag) was run in parallel to
show that it did not interfere with the NBD-C -sphingomyelin band.
12
Synthetic Procedures and Characterization. Compounds 1,
a, and 2c were synthesized previously.
1
2
2
%
cell migration at 24 h
N-(3-Aminopropyl)-N′-(2-cyclopentadecylethyl)propane-
1
,3-diamine 2b. Compound 11 was dissolved in absolute EtOH (1
= (
area with no cells at 0 h) − (area with no cells at 24 h)
× 100
mL) and slowly added dropwise to 4 M HCl (1.56 mL) at 0 °C. After
the addition was complete, the solution was brought to rt and stirred
area with no cells at 0 h
for 24 h. It was then concentrated under reduced pressure to yield the
%
cell migration normalized to control
1
product 2b (20 mg, 0.042 mmol, 74% yield). H NMR (D O): δ 3.10
2
(m, 10H, J3
1
3
= 6.1 Hz), 2.09 (m, 4H), 1.58(m, 2H), 1.45 (m, 1H),
.29 (s, 28H). C NMR (CDCl ): δ 45.9, 43.9, 36.8, 33.7, 33.5, 31.3,
0.9, 30.3, 28.7, 28.1, 27.4, 27.4, 26.5, 25.8, 25.6, 24.0, 23.5. HRMS
%
migration with drug at 24 h
migration of control at 24 h
H−H
= %
× 100
13
3
calcd for C H N (M + H) 367.392, found 367.3926. Compound 2b
23
49
3
%
cell inhibition at 24 h
was 93% pure by HPLC analysis [UV detection at 210 nm showed a
major peak eluted (∼5 min) on a C18 column using 60% acetonitrile/
aqueous heptanesulfonate buffer at pH 3.8 with a flow rate of 1 mL/
min].
⎛
⎞
%
migration with drug at 24 h
=
⎜1 −
⎟ × 100
⎝
% migration of control at 24 h ⎠
1
N -Cyclopentadecylmethylethane-1,2-diamine 3a. Com-
SGMS Assay. The assay was used to test for potential agonism of
the ceramide to sphingomyelin conversion by the motuporamine
compounds 1, 2a, and 2c. The sphingomyelin synthase (SGMS)
activity assay was modified from Ding et al. Briefly, a cell pellet
derived from Pan02 murine pancreatic cancer cells (16 M cells) was
lysed using a lysis buffer containing 50 mM Tris-HCl, 1 mM EDTA,
% sucrose, and protease inhibitors (10 μL/mL of buffer). The cell
homogenate was then centrifuged at 5000 rpm for 10 min, and the
supernatant was collected and used fresh for the reaction assay. C -
pound 13 (61 mg, 0.16 mmol) was dissolved in EtOH (1 mL), and
4
M HCl/EtOH (1 mL) was added dropwise and stirred at rt
27
overnight. The solvent was then removed under reduced pressure to
1
give a white solid 3a (50 mg, 0.157 mmol, 98% yield). H NMR
(D
28H). C NMR (D
26.1. Anal. Calcd for C18
Found: C 61.12, H 11.44, N 7.72.
O): δ 3.39 (br s, 4H), 3.01 (br s, 2H), 1.82 (br s, 1H), 1.33 (br s,
2
13
5
2
O): δ 55.5, 47.3, 38.0, 36.7, 31.8, 29.3, 28.9, 28.6,
·0.05H O: C 60.83, H 11.34, N 7.88.
H
40Cl
N
2
2
2
12
NBD-ceramide (10 μL of 20 μM stock in CHCl ) and
phosphatidylcholine (PC; 10 μL of a 0.1 mg/mL solution comprising
N-(2-Aminoethyl)-N′-cyclopentadecylmethylethane-1,2-dia-
mine 3b. Compound 14a (96 mg, 0.153 mmol) was dissolved in
solution containing EtOH (3 mL) and 4 M HCl in EtOH (3 mL) and
stirred overnight at rt. The solvent was then removed under reduced
3
9
:1 CHCl /MeOH) were pipetted into a 0.6 mL centrifuge tube and
3
concentrated under reduced pressure to provide a tube containing
dried residues of both substrates. These were redissolved in a buffer
containing 50 mM Tris-HCl (5 μL, pH 7.4) and 25 mM KCl (5 μL).
Next, the motuporamine derivative was added (10 μL of either a 2 μM
or 4 μM stock in water), and as a final step the cell lysate (20 μL) was
added to begin the reaction in a total reaction volume of 40 μL. Note
that the 2 μM and 4 μM stock solutions of the motuporamines in
deionized water were used to provide final concentrations of 0.5 μM
and 1 μM, respectively. A positive control using deionized water (10
μL) in place of the motuporamine stock solution was run to confirm
that the cell lysate was responsible for the synthesis of sphingomyelin
and not the motuporamine compound. A negative control was run
concurrently to confirm that the sphingomyelin synthase contained in
the cell lysate was responsible for synthesis of sphingomyelin and not
the lysis buffer itself. In this negative control, the reaction was run
using lysis buffer only (containing no enzyme). The respective
mixtures were incubated at 37 °C for 72 h. Note that an earlier time
course study revealed that a 72 h incubation period provided a strong
sphingomyelin signal compared to a 24 h experiment (data not
shown). The lipids in each reaction mixture were then extracted in the
presence of a 2:1 mixture of chloroform/methanol (40 μL), and the
chloroform layer containing the unreacted ceramide and sphingomye-
lin product was carefully separated from the reaction mixture via glass
pipet and the organic layer (∼25−35 μL obtained) placed into a 0.6
mL microcentrifuge tube. Thin layer chromatography (TLC) used
normal phase TLC plates to separate the resulting lipids. Each
respective lipid extract (20 μL) was spotted via pipet onto a separate
lane on the TLC plate and dried under a stream of nitrogen. The TLC
1
pressure to give a white solid 3b (60 mg, 0.138 mmol, 90% yield). H
NMR (D
(br s, 28H); C NMR (D
28.94, 28.88, 28.55, 26.1. HRMS for C20
325.3439; found, 325.3457. Anal. Calcd for C20
O): δ 3.45 (m, 8H), 3.02 (br s, 2H), 1.83 (br s, 1H), 1.32
2
13
O): δ 55.6, 47.1, 46.0, 37.9, 36.7, 31.8, 29.3,
2
H
43
N
(M + H): theory,
3
H
46Cl N ·0.56H O: C
3 3 2
53.98, H 10.67, N 9.44. Found: C 54.38, H 10.68 N 9.05.
N-[2-(2-Aminoethylamino)ethyl]-N′-cyclopentadecylmethyl-
ethane-1,2-diamine 3c. Removal of the Boc groups of 14b was
performed in 4 M HCl/EtOH by first dissolving 14b in 200 proof
EtOH (5 mL), which required sonication. 4 M HCl in EtOH (5 mL)
was then added dropwise while stirring and then stirred overnight. The
solvent was then removed under reduced pressure to provide a white
1
solid, 3c (252 mg, 0.49 mmol, 98% yield). H NMR (D
O): δ 3.51 (m,
2
1
3
12H), 3.04 (br s, 2H), 1.86 (br s, 1H), 1.33 (s, 28H). C NMR
(D O): δ 55.4, 47.1, 46.4, 46.2, 38.0, 36.6, 31.8, 30.5, 29.4, 29.0, 28.8,
26.1. HRMS C22 (M + H): theory, 368.3878; found, 368.3879.
Anal. Calcd for C22 : C 51.36, H 10.19, N 10.89. Found: C
51.35, H 10.46, N 10.62.
2
H N
48 4
H
52Cl N
4 4
Synthesis of N-(2-Aminoethyl)-N′-[2-(cyclopentadecyl-
methylamino)ethyl]ethane-1,2-diamine 4f. The free base of
2,2-tetraamine 4c was generated using N,N′-bis-(2-aminoethyl)-
ethane-1,2-diamine (3.00 g, 10.27 mmol, Sigma-Aldrich) with 4
equiv of aq NaOH (1 M, 41 mmol, 41 mL). The water was removed
under vacuum, and then the residue was further dried by adding and
removing benzene under reduced pressure to remove any excess water.
Anhydrous Na
SO
(8 equiv, 82.3 mmol, 11.69 g) was added to a
2
4
dried vessel with 25% MeOH/CH
Cl (50 mL). One equivalent of
2
2
plate was eluted first with 15% MeOH/CHCl until the solvent front
salicyclaldehyde (10.27 mmol, 1.10 mL) in MeOH (10 mL) was then
added dropwise over 2 h at 0 °C while stirring. Upon addition of the
reagent, the reaction immediately turned yellow. Imine formation was
3
reached 90% up the plate (R of NBD-C -ceramide was 0.73). In this
f
12
solvent system, the sphingomyelin analyte remained at the origin.
Therefore, the plate was dried under a stream of nitrogen and re-
1
monitored by H NMR. After imine formation was complete, the
eluted using the more polar solvent system (30% MeOH/CHCl ) to
remaining reactive amine centers were reacted with di-tert-butyl
3
elute the sphingomyelin product off the baseline, and elution was
stopped before the new solvent front reached the originally eluted
NBD C -ceramide band. This double elution approach provided
dicarbonate (3 equiv., 30.87 mmol, 6.74 g) and stirred overnight. The
reaction progress was checked by H NMR for full N-bocylation, and
additional di-tert-butyl dicarbonate was then added (0.3 equiv, 0.67g)
1
12
J
DOI: 10.1021/acs.jmedchem.7b01222
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
while heating at 40 °C overnight to drive the reaction to completion.
The reaction was again checked for N-tert-butylcarbonylation and
2-Cyclopentadecylethanamine 8. Nitrile 7 (412 mg, 1.64
mmol) was dissolved in dry THF (5 mL, 55.5 mmol) and added
confirmed to be complete by NMR. Then MeONH ·HCl (10.3 mmol,
dropwise to a stirred solution of LiAlH (207 mg, 5.46 mmol) in THF
2
4
0
.86 g) was added in 1.5 mL of TEA. Note that MeONH ·HCl was
(5 mL) at 0 °C. The reaction was then warmed to rt and refluxed
overnight. The reaction was monitored for the disappearance of the
2
initially not soluble in the TEA, which was used to generate the free
base. Upon addition of 25% MeOH/CH Cl (5 mL) the methoxy-
nitrile with TLC (1% NH
(7) = 0.55). After disappearance of the starting material was
confirmed by TLC, the reaction mixture was concentrated and
redissolved in CH Cl . The organic phase was washed with a solution
of water (0.9 mL) and 5 M NaOH (0.15 mL), and the organic layer
was separated, dried over anhydrous Na SO , filtered, and
concentrated under reduced pressure to yield a colorless oil. Column
chromatography (85% CH Cl , 15% methanol, 1% NH OH) provided
amine 8 (203 mg, 0.80 mmol, 49% yield). H NMR: δ 2.72 (m, 2H),
OH/15% MeOH/CH Cl , R (8) = 0.28,
2
2
4
2 2 f
amine easily dissolved and the TEA/MeOH/CH Cl solution was
R
f
2
2
then added dropwise to the stirring solution. Imine cleavage was
monitored by NMR. Note that an oxime byproduct is formed via
methoxyamine exchange with the imine. The MeOH was removed
2
2
under reduced pressure, and the residue was redissolved in CH Cl2
2
4
2
and washed with saturated Na CO (10 mL). The aqueous layer was
2
3
extracted three times with CH Cl and the organic layers were
2
2
4
2
2
1
combined and concentrated under reduced pressure to yield a yellow
oil (3.65 g). The product was then isolated by column chromatog-
raphy (1% NH OH/10%MeOH/CH Cl , R (4f) = 0.65). The product
1
2
2
.41 (m, 4H), 1.34 (m, 28H). 13C NMR: δ 40.0, 38.9, 34.2, 32.4, 27.6,
6.9, 26.6. HRMS for C H N (M + H): theory, 253.2789; found,
17 35
4
2
2
f
53.2770. Anal. Calcd for C H N·0.2H O: C 79.43, H 13.88, N 5.45.
4
f eluted as a light yellow solid (1.48 g, 3.31 mmol, 32% yield). 4f:
17 35
2
1
Found: C 79.24, H 13.94, N 5.37.
C H N O . H NMR (CDCl ): δ 3.31 (br s, 10H), 2.84 (br s, 2H),
2
1
42
4
6
3
13
Methanesulfonic Acid 3-[tert-Butoxycarbonyl-(3-tert-
butoxycarbonylaminopropyl)amino]propyl Ester 10. (3-tert-
Butoxycarbonylaminopropyl)-(3-hydroxypropyl)carbamic acid tert-
1
4
4
5
.79 (br s, 2H), 1.46 (br s, 27H). C NMR (CDCl ): δ 155.5, 76.6,
6.2, 45.5, 40.2, 39.1, 28.1. HRMS for C H N O (M + H): theory,
46.3121; found, 446.3104. Anal. Calcd for C H N O ·0.2H O: C
6.03, H 9.49, N 12.45. Found: C 55.87, H 9.39, N 12.23.
Cyclopentadecylmethanol 5b and Cyclopentadecyl-
carbaldehyde 5c. Syntheses of 5b and 5c have been described by
3
21
42
4
6
21
42
4
6
2
16
butyl ester 9 (100 mg, 0.3 mmol) was added to TEA (127 μL, 0.9
mmol) and CH Cl (3 mL). Methanesulfonyl chloride (34.9 μL, 0.45
2
2
mmol) was added dropwise at 0 °C via syringe under a nitrogen
atmosphere. Once the addition was complete, the syringe was rinsed
with CH Cl (0.6 mL). The reaction progress was monitored by TLC
12
1
Muth et al. H NMR analysis matched the literature spectra for these
compounds.
12
2
2
(
2
5% MeOH/CH Cl , R (alcohol) = 0.25; R (mesylate) = 0.37). After
Bromomethylcyclopentadecane 5d. The alcohol 5b (250 mg,
.04 mmol) was placed under an inert atmosphere. Phosphorus
2 2 f f
4 h, 4 M NaOH (5 mL) was added with stirring. The organic layer
1
was separated, dried over anhydrous Na SO , filtered, and
tribromide (0.52 mmol, 49 μL) was added by syringe. The reaction
immediately turned yellow and started bubbling and was stirred at rt
for 1.5 h. Hexane (3 mL) was added, and the reaction was refluxed at
2
4
concentrated under reduced pressure to yield mesylate 10 (99 mg,
0
3
4
.24 mmol, 80% yield).
1
9
: H NMR (CD OD) δ 3.55 (t, 2H), 3.28 (t, 2H), 3.24 (t, 2H),
3
16
6
9 °C for another 1.5 h. The reaction turned a brownish yellow color.
.04 (q, 2H), 1.62−1.81 (m, 4H), 1.44 (s, 9H), 1.42 (s, 9H).
The vessel was rinsed, and the brown crude (0.48 g) was isolated.
Column chromatography (100% n-hexane) was performed (with a
1
3
1
0: H NMR (CDCl ) δ 4.25 (t, 2H, J
= 6.2 Hz), 3.28 (br s,
3
H−H
H), 3.11 (br s, 2H), 3.02 (s, 3H), 1.99 (m, 2H), 1.67 (m, 2H), 1.56
s, 3H), 1.47 (s, 9H), 1.44 (m, 9H)
3-tert-Butoxycarbonylaminopropyl)-[3-(2-cyclopentadecyl-
3
0:1 ratio of silica gel/crude) due to the large R difference (R (5b) =
f
f
(
0
, R (5d) = 0.8). Visualization of the TLC plate using phosphomo-
f
(
lybdic acid and heat provided a convenient monitoring tool. The
ethylamino)propyl]carbamic Acid tert-Butyl Ester 11. 2-
Cyclopentadecylethylamine 8 (0.07g, 0.28 mmol) was dissolved in
CH Cl and added to Na CO (75 mL, 1.79 mmol) while stirring at rt.
product 5d was isolated and concentrated under reduced pressure to
1
yield an oil (0.21g, 0.695 mmol, 67% yield). H NMR (CDCl ): δ
3
.38(d, 2H, J3
13
2
2
2
3
3
= 6.1 Hz), 1.72 (m, 1H), 1.38 (br s, 28H).
C
H−H
Di-Boc mesylate 10 (121 mg, 0.296 mmol) was dissolved in CH Cl
2
2
NMR (CDCl ): δ 40.4, 38.7, 31.2, 27.3, 26.9, 26.6, 26.5, 24.6.
3
(
1 mL) and added to the solution dropwise. The reaction was stirred
Methanesulfonic Acid Cyclopentadecylmethyl Ester 6.
Alcohol 5b (998 mg, 4.15 mmol) was added to TEA (643 μL, 4.58
mmol) in CH Cl . Methanesulfonyl chloride was then added by
for 48 h and was monitored by TLC (10% MeOH/CH Cl ; R =
2
2
f
0
.34). After the reaction was complete, CH Cl (2 mL) was added and
2 2
2
2
the solution was washed three times with aq Na CO (10% by w/v, 3
2
3
syringe (355 μL, 4.58 mmol) at 0 °C, and then the solution slowly
warmed to rt and stirred overnight. The reaction was monitored by
TLC (100% CH Cl , R (mesylate 6) = 0.63; R (alcohol 5b) = 0.37)
mL). The organic layer was separated, dried over anhydrous Na SO ,
2
4
filtered, and concentrated under reduced pressure to yield a crude
2
2
f
f
yellow oil (248 mg). Column chromatography (10% MeOH/CH Cl ,
2
2
and then quenched with 1 M NaOH (2 mL). The organic phase was
washed three times with 1 M NaOH (5 mL), then separated, dried
over anhydrous Na SO , filtered, and concentrated to give the
followed by 10% MeOH/1% NH OH, CH Cl ) provided a product
4
2
2
mixture as a clear oil (60 mg, 0.106 mmol, 38% yield) with 11 along
with a self-cyclized starting material 12. These were then separated by
running a second column (4.5% MeOH/CH Cl ) to yield the
2
4
1
mesylate 6 as a yellow oil (1.12 g, 3.52 mmol, 85% yield). H NMR
CDCl ): δ 4.15 (d, 2H, J
2
2
3
(
= 6.1 Hz), 3.0 (s, 3H), 1.34 (br s, 29H).
3
H−H
byproduct 12 (27.3 mg, 0.0572 mmol, 21% yield) and the desired
Cyclopentadecylacetonitrile 7. KCN (3.09 g, 47.6 mmol), 18-
product 11 (32 mg, 0.056 mmol, 20% yield) as a yellow oil.
crown-6 ether (145 mg, 0.48 mmol), and dry CH CN (48 mL) were
1
3
11: H NMR (CDCl ) δ 3.31(br s, 2H), 3.16 (br s, 2H), 3.03 (br s,
3
added to mesylate 6 (1.69 g, 5.29 mmol), and the reaction was
refluxed overnight. The reaction was monitored by TLC (70%
2
(
H), 2.84 (br s, 4H), 2.06 (m, 2H), 1.65 (m, 4H), 1.40 (s, 18H), 1.36
13
br s, 29H). C NMR: δ 46.9, 45.0, 37.8, 34.6, 32.4, 31.3, 29.7, 29.1,
28.3, 27.5, 26.8, 26.6, 26.5, 25.0, 24.3, 22.7. HRMS for C H Cl N O
4
hexanes/CH Cl , R = 0.28), and then volatiles were removed under
2
2
f
33
65
3
3
vacuum. The residue was then redissolved in CH Cl and washed with
2
2
(M + H): theory, 567.4973; found, 567.4975.
water. The layers were separated, and the organic layer was dried over
anhydrous Na SO , filtered, and concentrated to give a yellow crude
1
1
2: H NMR (CDCl ) δ 4.25 (m, 2H), 3.30 (br s, 4H), 3.11 (s,
3
2
4
2H), 2.0 (m, 2H), 1.66 (m, 2H), 1.46 (s, 18H).
2-(Cyclopentadecylmethylamino)ethyl]carbamic Acid tert-
oil (1.29 g). The crude was then purified by column chromatography
70% hexanes/CH Cl ) to yield a colorless oil (0.92g, 3.69 mmol, 82%
[
(
Butyl Ester 13. The commercially available ((2-aminoethyl)carbamic
acid tert-butyl ester 4b (286 mg, 1.78 mmol, Sigma) was added to
mesylate 6 (508 mg, 1.59 mmol) in CH CN (20 mL) with K CO
3
2
2
yield, 91% conversion). Unreacted mesylate 6 was also recovered as a
white crystalline solid (0.15 g, 0.47 mmol, 9% recovery) in (50%
3
2
1
3
hexanes/CH Cl , R = 0.3). 7: H NMR δ 2.30 (d, 2H, J = 6.6
(0.72 g) at 50 °C. After 5 days the reaction showed 50% conversion,
2
2
f
H−H
13
Hz), 1.80 (m, 1H), 1.34 (br s, 28H). C NMR: δ 119.2, 33.7, 31.8,
7.1, 26.8, 26.7, 26.6, 26.5, 24.3, 22.9. HRMS for C H N (M + H):
and most of the CH CN was stripped off (leaving 5 mL) and heated
3
2
overnight for complete conversion. The solvent was removed under
reduced pressure, redissolved in DCM, and washed with water. The
crude (620 mg) was then purified by column chromatography (7%
17
31
theory, 250.2535; found, 250.2527. Anal. Calcd for C H N: C 81.86,
17
31
H 12.53, N 5.62. Found: 82.04, 12.63, N 5.51.
K
DOI: 10.1021/acs.jmedchem.7b01222
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
1
MeOH/DCM R (13) = 0.4) to give 13 as a clear oil (75 mg, 0.196
complete (monitored by H NMR). The solvent was removed in
vacuo and the crude imine was redissolved in 50% methanol/DCM
f
mmol, 12% yield) as well as a large amount of what appeared to be a
1
cyclized urea byproduct (250 mg). 13: H NMR (CD O) δ 3.23
and cooled to 0 °C. NaBH (167 mg, 4.28 mmol) was added, and the
4
4
(
broad s, 2H), 2.78 (br s, 2H), 2.60 (br s, 2H), 1.63 (s, 1H), 1.45 (s,
mixture was stirred at rt overnight. The solvents were removed under
vacuum, and the residue was redissolved in DCM and washed with
9
2
3
7
H), 1.37 (br s, 28H). 13C NMR (CD O): δ 150.8, 68.0, 50.4, 31.9,
4
8.9, 28.2, 28.0, 27.9, 27.8. HRMS for C H N O (M + H): theory,
saturated Na
anhydrous Na
2
CO
3
. The organic layer was separated, dried over
, filtered, and concentrated to give a solid (521 mg
23
46
2
2
82.3558; found, 382.3559. Anal. Calcd for C H N O ·0.3H O: C
2
SO
4
23
46
2
2
2
1.19, H 12.10, N 7.22. Found: C 71.02, H 12.05, N 7.20.
crude) which was purified by 5% MeOH/CHCl to yield 412 mg (1.18
3
(2-tert-Butoxycarbonylaminoethyl)-[2-(tert-butoxycarbonyl-
mmol) of adduct 16 (80% yield). Debocylation with 2 mL of ethanol
cyclopentadecylmethylamino)ethyl]carbamic Acid tert-Butyl
Ester 14a. Triamine 4c (N -(2-aminoethyl)ethane-1,2-diamine, 255
mg, 2.47 mmol, Sigma) was added dropwise to bromide 5d (250 mg,
0
at rt, and then the mixture was refluxed for 48h. After 48 h, H NMR
showed complete disappearance of starting material and the reaction
was concentrated and resuspended in DCM and washed with 0.1 M
NaOH. The organic layer was separated, dried over anhydrous
Na SO , filtered, and concentrated to give a light yellow oil (272 mg).
Di-tert-butyl dicarbonate was added (540 mg, 2.47 mmol) in 25%
MeOH/DCM (10 mL) at 40 °C and the reaction stirred overnight.
and 2 mL of 4 M HCl gave 17 as its dihydrochloride salt (340 mg, 1.05
1
1
mmol, 89% yield) for an overall yield of 75%. H NMR (D O): δ 8.40
2
(s, 1H), 8.05 (m, 2H), 7.63 (m, 2H), 7.53 (m, 2H), 4.95 (s, 2H), 3.55
(m, 2H), 3.35 (m, 2H). 13C NMR (25% DMSO-d in D O): δ 133.4,
.82 mmol) in CH CN (3 mL) along with K CO (1.17 g, 8.47 mmol)
3
2
3
6
2
1
132.8, 131.9, 130.3, 128.3, 125.4, 124.3, 47.3, 46.0, 38.3. Anal. Calcd
C H N Cl ·0.3H O: C 62.13, H 6.32, N 8.52. Found: C 62.10, H
17
20
2
2
2
6.19, N 8.47.
2
4
ASSOCIATED CONTENT
Supporting Information
■
*
S
1
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jmed-
chem.7b01222.
Bocylation was assessed by H NMR, and when complete, the reaction
was concentrated, redissolved in DCM, and washed with aq Na CO .
The organic layer was separated, dried over anhydrous sodium sulfate,
2
3
filtered, and concentrated under reduced pressure to give a yellow oil
Sample cytotoxicity curves for 3a−c with L3.6pl cells,
additional scratch assay protocol details, representative
scratch assay images using a non-native polyamine as well
(
(
(
2
706 mg). The crude was then purified by column chromatography
2% MeOH/DCM R (14a) = 0.4) to give 14a as a viscous yellow oil
f
1
154 mg, 25% yield). H NMR (CDCl ): δ 3.30 (m, 8H), 3.06 (m,
3
1
13
13
as H and C NMR spectra for compounds 2b, 3a−c, 4f,
H), 3.04 (m, 1H), 1.68 (s, 1H), 1.45 (s, 27H), 1.31 (br s, 28H).
C
1
1
1
5
1
d, 6 ( H only), 7, 8, 10 ( H only), 11, 12 ( H only), 13,
4a, 14b, 16, and 17, and more information regarding
NMR (CDCl ): δ. 155.9, 79.9, 52.5, 46.7, 45.5, 39.8, 36.2, 30.2, 28.4,
3
2
6
6
6.8, 24.4. HRMS for C H N O (M + H): theory, 625.5043; found,
35 67 3 6
25.5030. Anal. Calcd for C H N O : theory, C 67.16, H 10.79, N
the ImageJ analysis (PDF)
Molecular formula strings and some data (CSV)
35
67
3
6
.71. Found: C 67.42, H 10.92, N 6.63.
2-(tert-Butoxycarbonyl-(2-[tert-butoxycarbonyl-(2-tert-
[
butoxycarbonylaminoethyl)amino]ethyl)amino)ethyl]-
cyclopentadecylmethylcarbamic Acid tert-Butyl Ester 14b.
Triethylene tetraamine tetrahydrochloride (N,N′-bis(2-aminoethyl)-
ethane-1,2-diamine, Sigma) (1.0 g, 3.42 mmol) was dissolved in 1 M
NaOH (13.8 mL) to give the free tetraamine base 4e. The water was
then removed under vacuum and benzene added to the residue and
removed under vacuum to facilitate trace water removal (via benzene−
water azeotrope) to give “dry” 4e. Mesylate 6 (628 mg, 2 mmol) was
dissolved in CH CN with K CO (391 mg, 2.83 mmol), and the
AUTHOR INFORMATION
Corresponding Author
*E-mail: otto.phanstiel@ucf.edu. Phone: 407-823-6545. Fax:
07-384-2062.
■
4
ORCID
Otto Phanstiel IV: 0000-0001-7101-1311
Notes
The authors declare no competing financial interest.
3
2
3
1
reaction showed complete conversion after 72 h at 50 °C by H NMR.
The reaction was concentrated to provide a residue, which was
dissolved in DCM and washed with 0.1 M NaOH (30 mL). A difficult
emulsion ensued. The organic layer was separated, dried over
anhydrous Na SO , filtered, and concentrated to give organic layer
ACKNOWLEDGMENTS
■
The authors thank the 2011 Department of Defense Congres-
sionally Directed Medical Research Program (CDMRP) Peer
Review Cancer Research Program (PRCRP) Discovery Award
CA110724 and the UCF College of Medicine for financial
support of this research and Dr. Isaiah Fidler at the University
of Texas M. D. Anderson Cancer Center and Dr. Wayne
Flintoff at the University of Western Ontario for their generous
gifts of L3.6pl and CHO-MG cells, respectively, for this
research. We also acknowledge the assistance of Dr. Julia
Soulakova and Bee Ben Khallouq at UCF-COM with the
statistical analyses and John Reyenga for his assistance in
developing the synthetic methods to access compound 10.
2
4
no. 1. The remaining emulsion/aqueous layer was concentrated and
redissolved in DCM at a lower temperature, and any remaining
precipitates were filtered off. The precipitates were washed with DCM
and the organic filtrate was pooled with organic layer no. 1 and they
were collectively concentrated to give a yellow viscous oil (806 mg).
The crude oil was redissolved in 25% MeOH/DCM and reacted with
di-tert-butyl dicarbonate (8.75 mmol, 1.91 g). The reaction was
1
monitored for bocylation by H NMR and then worked up using
saturated aq Na CO and DCM. The organic layer was separated,
2
3
dried over anhydrous Na SO , filtered, and concentrated to give a
2
4
yellow oil (1.57 g). Column chromatography (25% EtOAc/hexanes, Rf
(
14b) = 0.39) gave 14b as a viscous oil (383 mg, 0.5 mmol, 25%
1
yield). H NMR (CDCl ): δ 3.32 (m, 12H), 3.08 (br s, 1H), 3.03 (br
s, 2H), 1.46 (s, 36H), 1.33 (br s, 29H). C NMR (CDCl ): δ 155.4,
3
13
ABBREVIATIONS USED
3
■
8
0.2, 79.4, 51.7, 45.4, 36.1, 35.8, 36.1, 35.8, 30.2, 28.5, 26.8, 26.4, 24.4.
AG, aminoguanidine; ASM, acid sphingomyelinase; Boc, tert-
butylcarbonyl; CHO, Chinese hamster ovary; CHO-MG,
Chinese hamster ovary cell defective in polyamine transport;
DCM, dichloromethane; DFMO, difluoromethylornithine;
DMSO, dimethylsulfoxide; MGBG, methylglyoxal bisguanyl
hydrazine; MTS, 3-(4,5-dimethylthiazol-2-yl)-5-(3-
carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H- tetrazolium,
inner salt; NBD, nitrobenzoxadiazole; ODC, ornithine
HRMS C H N O (M + H): theory, 768.5987; found, 768.5976.
42
80
4
8
Anal. Calcd C H N O : C 65.59, H 10.48, N 7.28. Found: C 65.89,
42
80
4
8
H 10.64, N 7.26.
1
N -Anthracen-9-ylmethylethane-1,2-diamine 17. To a stirred
solution of mono-Boc-diamine 4b (268 mg, 1.67 mmol) in 25%
methanol/DCM (10 mL) was added a solution of 9-anthraldehyde
(
289 mg, 1.40 mmol) in 5 mL of 25% methanol/DCM under N . The
2
solution was allowed to stir at rt overnight until imine formation was
L
DOI: 10.1021/acs.jmedchem.7b01222
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
decarboxylase; PBS, phosphate buffered saline; PC, phospha-
tidylcholine; PCC, pyridinium chlorochromate; PTS, poly-
amine transport system; S1P, sphingosine 1-phosphate; SGMS,
sphingomyelin synthase; TEA, triethylamine
(18) Brown, M. J. Synthesis, Structure and Functionalisation of 1,2-
Diazetidines. Ph.D. Dissertation, Department of Chemistry, University
of Warwick, Coventry, U.K. CV47AL, 2011, (Order No. U568598).
Available from ProQuest Dissertations & Theses A&I; ProQuest
Dissertations & Theses Global (1033192588) (Order No. U568598).
Retrieved from https://login.ezproxy.net.ucf.edu/login?url=https://
search.proquest.com/docview/1033192588?accountid=10003.
REFERENCES
■
(19) Kutovaya, I. V.; Shmatova, O. I.; Tkachuk, V. M.; Melnichenko,
(
1) Williams, D. E.; Lassota, P.; Andersen, R. J. Motuporamines A-C,
cytotoxic alkaloids isolated from the marine sponge Xestospongia exigua
N. V.; Vovk, M. V.; Nenajdenko, V. G. Aza-Henry reaction with CF3-
Ketimines: An efficient approach to trifluoromethylated-nitroamines,
1,2-diamines, -aminooximes, and imidazolidinones. Eur. J. Org. Chem.
2015, 2015 (30), 6749−6761.
(
Kirkpatrick). J. Org. Chem. 1998, 63, 4838−4841.
(
2) Williams, D. E.; Craig, K. S.; Patrick, B.; McHardy, L. M.; van
Soest, R.; Roberge, M.; Andersen, R. J. Motuporamines, anti-invasion
and anti-angiogenic alkaloids from the marine sponge Xestospongia
exigua (Kirtpatrick): Isolation, structure elucidation, analogue syn-
thesis, and conformational analysis. J. Org. Chem. 2002, 67, 245−258.
(20) Phanstiel, O., IV; Kaur, N.; Delcros, J.-G. Structure-activity
investigations of polyamine-anthracene conjugates and their uptake via
the polyamine transporter. Amino Acids 2007, 33, 305−313.
(21) Heaton, M. A.; Flintoff, W. F. Methylglyoxal-bis-
(guanylhydrazone)-resistant Chinese hamster ovary cells: Genetic
evidence that more than a single locus controls uptake. J. Cell. Physiol.
1988, 136, 133−139.
(
3) Muth, A.; Kamel, J.; Kaur, N.; Shicora, A. C.; Ayene, I. S.;
Gilmour, S. K.; Phanstiel, O. Development of polyamine transport
ligands with improved metabolic stability and selectivity against
specific human cancers. J. Med. Chem. 2013, 56 (14), 5819−5828.
(
4) Madan, M.; Patel, A.; Skruber, K.; Geerts, D.; Altomare, D. A.;
(22) Mandel, J. L.; Flintoff, W. F. Isolation of mutant mammalian
cells altered in polyamine transport. J. Cell. Physiol. 1978, 97, 335−344.
Phanstiel, O. ATP13A3 and caveolin-1 as potential biomarkers for
difluoromethylornithine-based therapies in pancreatic cancers. Am. J.
Cancer Res. 2016, 6, 1231−1252.
(23) Bruns, C. J.; Harbison, M. T.; Kuniyasu, H.; Eue, I.; Fidler, I. J.
In vivo selection and characterization of metastatic variants from
human pancreatic adenocarcinoma by using orthotopic implantation in
nude mice. Neoplasia 1999, 1 (1), 50−62.
(
5) Saini, P.; Eyler, D. E.; Green, R.; Dever, T. E. Hypusine-
containing protein eIF5A promotes translation elongation. Nature
009, 459 (7243), 118−121.
6) Hoopes, B. C.; McClure, W. R. Studies on the selectivity of DNA
precipitation by spermine. Nucleic Acids Res. 1981, 9 (20), 5493−5504.
7) Park, M. H. The post-translational synthesis of a polyamine-
(24) Muth, A.; Madan, M.; Archer, J. J.; Ocampo, N.; Rodriguez, L.;
2
(
Phanstiel, O. Polyamine transport inhibitors: design, synthesis, and
combination therapies with difluoromethylornithine. J. Med. Chem.
2014, 57 (2), 348−363.
(
derived amino acid, hypusine, in the eukaryotic translation initiation
(25) Roy, U. K.; Rial, N. S.; Kachel, K. L.; Gerner, E. W. Activated K-
RAS increases polyamine uptake in human colon cancer cells through
modulation of caveolar endocytosis. Mol. Carcinog. 2008, 47 (7), 538−
553.
(26) Sturgeon, C. M.; Kemmer, D.; Anderson, H. J.; Roberge, M.
Yeast as a tool to uncover the cellular targets of drugs. Biotechnol. J.
2006, 1 (3), 289−298.
(27) Ding, T.; Li, Z.; Hailemariam, T.; Mukherjee, S.; Maxfield, F. R.;
Wu, M. P.; Jiang, X. C. SMS overexpression and knockdown: impact
on cellular sphingomyelin and diacylglycerol metabolism, and cell
apoptosis. J. Lipid Res. 2008, 49 (2), 376−385.
(28) Barcelo-Coblijn, G.; Martin, M. L.; de Almeida, R. F.; Noguera-
Salva, M. A.; Marcilla-Etxenike, A.; Guardiola-Serrano, F.; Luth, A.;
Kleuser, B.; Halver, J. E.; Escriba, P. V. Sphingomyelin and
sphingomyelin synthase (SMS) in the malignant transformation of
glioma cells and in 2-hydroxyoleic acid therapy. Proc. Natl. Acad. Sci. U.
S. A. 2011, 108 (49), 19569−19574.
(29) Kornhuber, J.; Tripal, P.; Reichel, M.; Terfloth, L.; Bleich, S.;
Wiltfang, J.; Gulbins, E. Identification of new functional inhibitors of
acid sphingomyelinase using a structure-property-activity relation
model. J. Med. Chem. 2008, 51, 219−237.
(30) Lin, M.; DiVito, M. M.; Merajver, S. D.; Boyanapalli, M.; van
Golen, K. L. Regulation of pancreatic cancer cell migration and
invasion by RhoC GTPase and caveolin-1. Mol. Cancer 2005, 4 (1), 21.
(31) McHardy, L. M.; Sinotte, R.; Troussard, A.; Sheldon, C.;
Church, J.; Williams, D. E.; Andersen, R. J.; Dedhar, S.; Roberge, M.;
Roskelley, C. D. The tumor invasion inhibitor dihydromotuporamine
C activates RHO, remodels stress fibers and focal adhesions, and
stimulates sodium-proton exchange. Cancer Res. 2004, 64 (4), 1468−
1474.
factor 5A (eIF5A). J. Biochem. 2006, 139 (2), 161−169.
(
8) Palmer, A. J.; Wallace, H. M. The polyamine transport system as
a target for anticancer drug development. Amino Acids 2010, 38 (2),
15−422.
9) Meyskens, F. L., Jr.; Gerner, E. W. Development of
4
(
difluoromethylornithine (DFMO) as a chemoprevention agent. Clin.
Cancer Res. 1999, 5 (5), 945−951.
(
10) Seiler, N. Thirty years of polyamine-related approaches to
cancer therapy. Retrospect and prospect. Part 1. Selective enzyme
inhibitors. Curr. Drug Targets 2003, 4 (7), 537−564.
(
11) Seiler, N.; Delcros, J. G.; Moulinoux, J. P. Polyamine transport
in mammalian cells. An update. Int. J. Biochem. Cell Biol. 1996, 28 (8),
43−861.
12) Muth, A.; Pandey, V.; Kaur, N.; Wason, M.; Baker, C.; Han, X.;
8
(
Johnson, T. R.; Altomare, D. A.; Phanstiel, O. Synthesis and biological
evaluation of antimetastatic agents predicated upon dihydromotupor-
amine C and its carbocyclic derivatives. J. Med. Chem. 2014, 57 (10),
4
(
023−4034.
13) Hyvonen, M. T.; Ucal, S.; Pasanen, M.; Peraniemi, S.; Weisell, J.;
Khomutov, M.; Khomutov, A. R.; Vepsalainen, J.; Alhonen, L.;
Keinanen, T. A. Triethylenetetramine modulates polyamine and
energy metabolism and inhibits cancer cell proliferation. Biochem. J.
2
(
016, 473 (10), 1433−1441.
14) Liu, J.; Guo, L.; Yin, F.; Zheng, X.; Chen, G.; Wang, Y.
Characterization and antitumor activity of triethylene tetramine, a
novel telomerase inhibitor. Biomed. Pharmacother. 2008, 62 (7), 480−
4
85.
(
15) Furstner, A.; Rumbo, A. Ring-closing alkyne metathesis.
Stereoselective synthesis of the cytotoxic marine alkaloid motupor-
amine C. J. Org. Chem. 2000, 65, 2608−2611.
(32) To, K. C.; Loh, K. T.; Roskelley, C. D.; Andersen, R. J.;
O’Connor, T. P. The anti-invasive compound motuporamine C is a
robust stimulator of neuronal growth cone collapse. Neuroscience 2006,
139 (4), 1263−1274.
(
16) Kaur, N.; Delcros, J.-G.; Martin, B.; Phanstiel, O. Synthesis and
biological evaluation of dihydromotuporamine derivatives in cells
containing active polyamine transporters. J. Med. Chem. 2005, 48,
3
(
832−3839.
(33) Hannun, Y. A.; Obeid, L. M. Principles of bioactive lipid
signalling: lessons from sphingolipids. Nat. Rev. Mol. Cell Biol. 2008, 9
(2), 139−150.
17) Breitbeil, F.; Kaur, N.; Delcros, J.-G.; Martin, B.; Abboud, K. A.;
Phanstiel, O. Modeling the preferred shapes of polyamine transporter
ligands and dihydromotuporamine-C mimics: Shovel versus hoe. J.
Med. Chem. 2006, 49, 2407−2416.
(34) Borselli, D.; Blanchet, M.; Bolla, J. M.; Muth, A.; Skruber, K.;
Phanstiel, O.; Brunel, J. M. Motuporamine derivatives as antimicrobial
M
DOI: 10.1021/acs.jmedchem.7b01222
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
agents and antibiotic enhancers against resistant Gram-negative
bacteria. ChemBioChem 2017, 18 (3), 276−283.
N
DOI: 10.1021/acs.jmedchem.7b01222
J. Med. Chem. XXXX, XXX, XXX−XXX