Evaluation of copper chelation agents as anti-angiogenic therapy

Article abstract of DOI:10.1016/S0968-0896(03)00401-2

The design, synthesis and evaluation of N,N′,N″-tris(2-pyridylmethyl)-cis,cis-1,3,5,-triaminocyclohexane (tachpyr, 1) derivatives as novel anti-angiogenic agents were performed in an in vitro endothelial cell proliferation assay to assess their cytotoxicity and selectivity. The selective nature of the anti-angiogenic agents for human umbilical vein endothelial cells (Huvec) was compared to a normal fibroblast cell line and a human Glioma cell line to evaluate these compounds. N,N′,N″-tris(2-mercaptoethyl)-cis,cis-1,3,5-triaminocyclohexane trihydrochloride (3b) was superior to tachpyr in terms of selectivity of its inhibitory activity toward the proliferation of Huvec compared to the fibroblast and human Glioma cell lines.

Full text of DOI:10.1016/S0968-0896(03)00401-2

Bioorganic & Medicinal Chemistry 11 (2003) 4287–4293
Evaluation of Copper Chelation Agents as Anti-Angiogenic
Therapy
Kevin Camphausen, Mary Sproull, Steve Tantama, Sandeep Sankineni, Tamalee Scott,
Cynthia Menard, C. Norman Coleman and Martin W. Brechbiel*
Radiation Oncology Branch, National Cancer Institute, National Institutes of Health, 10 Center Drive, Building 10,
Room B3B69, Bethesda, MD 20892-1002, USA
Received 20 January 2003; accepted 10 April 2003
Abstract—The design, synthesis and evaluation of N,N0,N00-tris(2-pyridylmethyl)-cis,cis-1,3,5,-triaminocyclohexane (tachpyr, 1)
derivatives as novel anti-angiogenic agents were performed in an in vitro endothelial cell proliferation assay to assess their cyto-
toxicity and selectivity. The selective nature of the anti-angiogenic agents for human umbilical vein endothelial cells (Huvec) was
compared to a normal fibroblast cell line and a human Glioma cell line to evaluate these compounds. N,N0,N00-tris(2-mercapto-
ethyl)-cis,cis-1,3,5-triaminocyclohexane trihydrochloride (3b) was superior to tachpyr in terms of selectivity of its inhibitory activity
toward the proliferation of Huvec compared to the fibroblast and human Glioma cell lines.
# 2003 Elsevier Ltd. All rights reserved.
Introduction
stimulate angiogenesis while the non-copper binding
form of ceruloplasmin cannot stimulate vessel growth in
the rabbit corneal assay.10,11 The first examination of
the effects of copper deprivation in animal tumor mod-
els utilized penicillamine, an early and routinely
employed chelation therapy agent, and/or a low copper
diet, and demonstrated a decreased level of angiogenesis
and hence, controlled tumor growth.12,13
Angiogenesis is a tightly regulated process that in a
normal adult is only found in the proliferating endome-
trium during menstruation and during wound healing.
However, angiogenesis is common in many pathologic
conditions including tumor formation, diabetic retino-
pathy and rheumatoid arthritis.1 Anti-angiogenic drug
therapy targets the proliferating endothelium in an
effort to arrest aberrant growth of vessels.2 Multiple
targets for anti-angiogenic drug therapy exist. An intri-
guing target is the proliferating endothelial cell’s
requirement for copper as a co-factor in its cellular
process.3 Therefore, drugs that could decrease the amount
of available copper at the level of the endothelium could
become anti-angiogenic agents.4,5
Certain chelating agents have been shown to bind cop-
per with a high affinity.12 Previous work on copper
chelation agents have focused on Wilson’s disease,
which is an inherited metabolic disease of copper toxi-
city that is fatal if left untreated. Early agents employed
to combat copper overload included penicillamine and
trientine (triethylenetetramine).14,15 However, some
patients developed severe toxicity to these agents. New
therapy for Wilson’s disease includes the use of zinc
acetate and tetrathiomolybdate (TM).16,17 Combining
the anti-tumor observations in animals and the clinical
utility of TM in patients with Wilson’s disease provided
impetus for an ongoing Phase I trial of TM in patients
with metastatic cancer. Early results have demonstrated
the therapy to be well tolerated and that copper levels
appeared to be reduced.18
The evidence that copper is important in tumor forma-
tion is inferential. First, biochemical studies have
demonstrated a higher copper concentration in tumor
tissue as compared to non-tumor tissue.6,7 Second,
multiple stimulators of angiogenesis including fibroblast
growth factor and vascular endothelial growth factor
bind copper with a high affinity.8,9 Third, both CuSO4
and ceruloplasmin (a copper containing protein) can
One family of chelating agents that have yet to be inves-
tigated for this application are those based upon cis,cis-
1,3,5,-triaminocyclohexane (TACH),19 specifically those
*Corresponding author. Tel.: +1-301-495-0591; fax: +1-301-402-
1923; e-mail: martinwb@mail.nih.gov
0968-0896/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0968-0896(03)00401-2

4288
K. Camphausen et al. / Bioorg. Med. Chem. 11 (2003) 4287–4293
agents prepared that employ this triamine as a platform
to append additional metal ion binding sites. Of this
class of compounds, the tris-pyridyl derivative known as
tachpyr (1) (Fig. 1) has been extensively investigated.
Recent publications have reported on the formation and
stability of radioactive Cu(II) complexes as well as the
synthesis of several closely related compounds.20 Based
upon the high stability and affinity for copper as
expressed in these radiopharmaceutical investigations,
the potential to exploit the properties of these com-
pounds as Cu(II) sequestration and/or deprivation
agents appeared attractive. The previously reported
compounds included permutations of methylation
wherein a methyl group had been added to the pyridyl
ring in all positions, N-methylation of the secondary
amines, and C-methylation of the ligand framework
between the secondary amine and pyridyl groups.
Additional analogues have also been prepared to initi-
ate investigation into donor character, exchanging imi-
dazole, hydroxyl, and mercapto donors for pyridyl, as
well as inclusion of some less configurationally
restrained tachpyr analogues.
Compound 1 and the direct analogues 1a–f thereof with
methyl groups (Fig. 2) on the various positions of the
pyridyl ring or elsewhere were prepared from TACH
and the appropriately substituted carboxaldehydes (1–
1d) or ketone (1e) as previously reported.19,20,22 In brief,
the triamine was reacted with minimal excess of the
ketone with elimination of water (Dean–Stark distilla-
tion) and the resultant tris-imine then reduced with
excess NaBH4. The synthesis of 1f was also performed
from the N,N0,N00-tris-methylated parent amine directly
alkylated with 2-(chloromethyl)pyridine.
Tris-imidazole 2 was also formed from the respective
tris-imine, however, reduction of the tris-imine was per-
formed by direct hydrogenation to yield the triamine as
hydride mediated reduction was either unsuccessful or
complicated by numerous side products.19
Novel chelators 3a and 3b (Fig. 3) were prepared from
cis,cis-1,3,5,-triaminocyclohexane by first tris-acylation
employing an active ester derivative to introduce the
two carbons bearing a protected hydroxyl or thiol.
Attempts at direct acylation by either acid chlorides
where possible or mediated by diimide coupling reac-
tions generated mixtures of products in poor yields.
Several protecting groups were investigated for this step
including benzyl and benzoyl, however, while some
provided a measure of desired product, all were ham-
pered by complications due to incomplete acylation or
requirements of chromatographic isolation. Addition-
ally, use of an acid labile protecting group in both cases
was desirable since this would then be removed during
the acidic cleavage of the polyamine boron adducts that
frequently complicate polyamine synthesis. Use of the
tert-butyl group for the synthesis of 3a was found to be
very convenient and was efficiently removed to provide
the final product in the last step. The EOE protected
One of the essential characteristics of anti-angiogenic
agents, that is used to screen prospective agents, is a
compound’s ability to inhibit endothelial cell prolifera-
tion while not inhibiting the proliferation of other cell
types, that is selective cytostasis.21 To evaluate the abil-
ity of all of the chelating agents in this report to func-
tion as anti-angiogenic agents, their ability to inhibit the
proliferation of human umbilical vein endothelial cells
(Huvec) in vitro while not inhibiting the proliferation of
non-endothelial cells was determined. A normal fibro-
blast cell line (NIH3T3) and a human glioma tumor cell
line (U251) were utilized as control non-endothelial cell
types. A compound that inhibits the proliferation of the
Huvec cell line at a concentration lower than the con-
centration needed to inhibit the other cell lines may be
an attractive anti-angiogenic agent. Herein, we report
the design, synthesis and biological evaluation of these
compounds.
Results and Discussion
The tachpyr (1) family of compounds had previously
been investigated for its formation of Cu(II) complexes
in regards to use as radiopharmaceuticals and also for
their application as therapeutics via iron depletion. For
this study these chelators were chosen as they had
demonstrated significant stability in forming Cu(II)
complexes and were thought then as viable Cu(II)
depletion agents as well.20,22
Figure 1.
Figure 2.

K. Camphausen et al. / Bioorg. Med. Chem. 11 (2003) 4287–4293
4289
thiol active ester had been prepared for the preparation
of bifunctional chelating agents for either 99mTc or
186,188Re and served to introduce the ethylene thiol
functionality after amide reduction and acid cleavage of
the protecting group.23
the NIH3T3 or U251 cell lines. The rest of the com-
pounds demonstrated either no activity or very little
differential in the inhibition of the cell lines.
Compound 4a was included to provide a much less
geometrically constrained coordination sphere as com-
pared to 1 yet maintain the same donor character and
number permitting some evaluation of this specific
variable. Similar activity and selectivity results were
obtained using 4a as compared to 1 which tend to indi-
cate that there may be no advantage gained by the pre-
organization of the donor components as achieved by
use of 1 in comparison to the much more unrestricted 4a.
Synthesis of 4a and 4b (Fig. 4) from the commercial
triamine, tris(2-aminoethyl)amine (tren), were both per-
formed in a very straightforward manner by modifi-
cation of previously reported synthesis of the tris-imine
to directly form the two respective tris-imines with sub-
sequent reduction with excess NaBH4.
There were numerous compounds within the 1–1f
family of chelators that inhibited Huvec proliferation at
a micromolar concentration (Table 1). There were also
many derivatives that also inhibited the proliferation of
either the NIH3T3 or U251 cell lines. Thus, while cyto-
toxic, these derivatives exhibited little differential
between any of the cells lines used for this purpose. Two
compounds, 1e and 1f, were also evaluated in part as a
control for 1 in that the previous investigations indi-
cated their unfavorable stability with radio-copper.
Surprisingly, 1f demonstrated at least a 2-fold greater
inhibition in Huvec proliferation as compared to either
In parallel, to that reported above, the analogous 4b
was also prepared and evaluated to determine whether
again the more open geometry could be more efficient
and whether the additional C-methylation would have
any effect. Similarly to 1e,4b was active, but both 1e and
4b demonstrated no greater selectivity towards the
Huvec cell line than their parental ligands. There are
several possible simple explanations for this result that
involve either modified coordination geometry, either 4-
coordination of metal ion or some form of multimeric
coordination, or in conjunction with this explanation,
C-methylation may be a source of significant change in
lipophilicity thereby permitting greater cell membrane
translation. The impact of increasing lipophilicity with
metal chelators to provide a significant enhancement of
activity has been reported wherein cyclam was found to
lack cytotoxicity while tetra-alkylated analogues were
very active while the intermediate amides were also
inactive.24 Additionally, there was no control of the
stereochemistry of these methyl groups through these
syntheses. The crystal structure of the Cu(II) complex of
1 clearly indicates that from the spiral nature that this
complex would only be stable if the methyl groups could
be aligned in a consistent configuration.20 This current
result supports a coordination geometry then that
would be either 4-, 5-, or some multimeric arrangement
for the 1e while in the case of the 4b the decreased geo-
metric constraints of the ligand may permit this chelator
to function well to sequester Cu(II).
One of other goals of this preliminary investigation was
to determine the proper donor character set of the
ligand. With this in mind, the pyridyl groups were
exchanged for imidazole giving 2 which was predicted in
part to have some selectivity for Cu(II) over iron and
therefore exhibit differential toxicity towards the Huvec
cell line. This indeed was the result, and 2 demonstrated
Figure 3.
Table 1. IC50 values (mM) of the compounds 1a–f for inhibition of
proliferation of Huvec, NIH3T3 and U251 cells
0
Compd
RR
X
Huvec
NIH3T3
U251
Tachpyr (1)
H
H
H
H
H
H
H
H
H
H
H
1
1.24
1.4
1.3
100
20
2.1
4.0
1.6
1.5
100
50
500
0.35
1.7
1.5
2.0
100
20
100
1a
1b
1c
1d
1e
1f
3-Me
4-Me
5-Me
6-Me
H
CH3
H
H
CH3
H
50
Figure 4.

4290
K. Camphausen et al. / Bioorg. Med. Chem. 11 (2003) 4287–4293
at least a 2-fold greater inhibition in Huvec proliferation
as compared to either the NIH3T3 or U251 cell lines.
The magnitude of this result was not large, and again
may stem from the donor atoms of 2 being excessively
constrained to less than optimal coordination geometry
for this select purpose. With that being stated, efforts to
prepare the tren-based analogue of 2 are ongoing to relieve
possible steric inhibitions to efficient complexation.
related to both some measure of lipophilicity and hence
cellular availability and then again donor character.
Further studies of novel compounds to optimize these
conditions of hexacoordination, soft donor character,
and moderate pre-organization are ongoing.
Conclusion
Lastly, recognizing that heterocyclic amines might not
be an optimal choice for chelation of intracellular
Cu(II), the pyridyl groups were replaced with either an
alkyl hydroxyl or thiol generating 3a and 3b, respec-
tively, with the latter compound being of primary inter-
est and the former acting as a control for coordination
number and geometry. Gratifyingly, 3b (Fig. 5) did
indeed demonstrate significant activity as well as the
best differentiation in activity towards the Huvec cell
line as compared to either NIH3T3 or U251 cells,
respectively, while 3a was completely inactive. Inhibi-
tion of the Huvec cell line was 10-fold greater than the
inhibition of the NIH3T3 cell line and 18-fold greater
than the U251 cell line. This differential should lead to
effective anti-angiogenic therapy with minimal toxicity.
Copper chelation is an attractive anti-angiogenic strat-
egy to prevent tumor endothelial cell proliferation and
tumor vascularization. Previous work with copper che-
lators has focused on Wilson’s disease and the neurolo-
gical manifestations of that disease.14,15,17 Recent
studies have examined the role of TM in the treatment
of cancer patients in a Phase I study.18 The goal of our
study was to use Huvec proliferation assays to evaluate
1, 4, and their derivatives for their ability to differen-
tially inhibit the proliferation of Huvec versus a fibro-
blast and tumor cell line. Compounds 1f, 2 and 3b
demonstrated a differential inhibition of Huvec with 3b
having the greatest effect. This lead compound and 3a, a
closely related compound, but of different donor char-
acter, that is hydroxyl versus thiol, that did not show
activity, will now be tested in vivo for their ability to
inhibit angiogenesis and prevent tumor formation.
From this preliminary evaluation compound 3b
emerged as a promising lead compound that will serve
as a starting point for novel variants to be derived to
obtain an agent with the greatest anti-angiogenic effect.
To help validate that the anti-proliferative effect
observed against Huvec was due to copper chelation,
studies to determine relative affinity constants of several
of the compounds reported in this work for Cu, Zn and
Fe have been initiated and will be reported in due
course. Anti-angiogenic compounds have also demon-
strated value added when combined with radiotherapy
and this strategy will also be employed with these com-
pounds.
Clearly, the result observed for 3b is related to the
donor character of this chelating agent. Thiols are
known to be good donors for binding Cu(II) being of a
‘soft’ nature and thiol donors have been extensively
investigated for myriad applications.25 As observed in
the results, that is zero activity, the contrary condition
of the ‘hard’ hydroxy being a poor donor for Cu(II) was
equally demonstrated with 3a. There were some con-
cerns that thiols would be too reactive in a biological
milieu and be subject to significant oxidation deleting
this component from being available for binding Cu(II)
and that perhaps thioethers might be a somewhat better
choice despite some loss of binding strength. In spite of
this concern, this compound appears not to suffer from
a compromising reactivity and displays not only reac-
tivity, but also promising selectivity.
Experimental
Taken in combination with all the information garnered
through this study, pre-organization on the order of
requiring the TACH scaffolding seems not a major pre-
requisite for activity as evidenced by the tren based
compounds. However, issues of selectivity appear to be
Materials and methods
All chemicals and solvents were purchased from Fluka,
Sigma, or Aldrich and were used as received. The cis,cis-
1,3,5-triaminocyclohexane trihydrobromide (TACH)
was prepared as reported in the literature.19 Com-
pounds 1, 1a–f, and 2 were prepared as previously
reported.19,20,22 The N-hydroxysuccinimidyl S-(1-ethoxy-
ethyl)mercaptoacetate was prepared as reported by
Kasina et al.23 The precursor tris-imines for 4a,b were
prepared by modification of a reported procedure.26
1H and 13C NMRwere obtained using a Varian Gemini
300 instrument and chemical shifts are reported in ppm
on the d scale relative to TMS, TSP, or solvent. Proton
chemical shifts are annotated as follows: ppm [multi-
plicity, integral, coupling constant (Hz)]. Chemical
ionization mass spectra (CI-MS) were obtained on a
Finnnegan 3000 instrument. Fast atom bombardment
mass spectra (FAB-MS) were taken on a Extrel 400.
Figure 5.

K. Camphausen et al. / Bioorg. Med. Chem. 11 (2003) 4287–4293
4291
Elemental analyses were obtained from Atlantic Micro-
labs (Norcross, GA, USA).
stirred for 18 h. The solvent was removed by vacuum
rotary evaporation and the residue was taken up in
EtOAc (300 mL). The solution was sequentially extrac-
ted with salt solution (100 mL), 5% NaHCO3 (2ꢀ100
mL), salt solution (100 mL), dried over Na2SO4, filtered,
and rotary evaporated to a white solid (2.18 g, 69%).
Analytical HPLC was performed using a Beckman gra-
dient system equipped with Model 114M pumps con-
trolled by System Gold software and a Model 165 dual
wavelength detector set at 254 and 280 nm. An Altex C-
18 reverse-phase column (5-mm particles, 4.6ꢀ250 mm)
and a binary gradient of 0–100% B/25 min (Solvent
A=0.05 M Et3N/HOAc, pH 5.5, Solvent B=MeOH) at
1.0 mL/min was used for all analyses.
1H NMR(DMSO- d6) d 7.314 (d, 1H, J=7.8), 3.743 (s,
2H), 3.620 (m, 1H), 1.791 (br.d, 1H, J=11.7), 1.377 (q,
1H, J=11.7), 1.159 (s, 9H); 13C NMR(DMSO- d6) d
169.489, 73.986, 62.096, 44.539, 37.193, 27.054; MS (CI/
NH3) 472,489 (M++1,18). Anal. calcd for
C24H45N3O6; C, 61.10; H, 9.63; N, 8.91. Found: C,
61.05; H, 9.59; N, 8.79.
Chemistry—ligand synthesis
N-Hydroxysuccinimidyl glycolate tert-butyl ether.
Methyl glycolate tert-butyl ether. Methyl glycolate (20
g, 222 mmol) was dissolved in dry cyclohexane (500 mL)
and tert-butyltrichloroacetimidate (50 g, 229 mmol) was
N,N0,N00-Tri(2-hydroxyethyl)-cis,cis-1,3,5-triaminocyclo-
hexane trihydrochloride. Amide 5 (1.00 g, 2.12 mmol)
was dissolved in THF (75 mL) in a two-necked flask,
cooled in an ice bath, and 1 M BH.3THF (13 mL) was
injected via syringe. The reaction was allowed to come
to room temperature over 1 h, after which the reaction
was vigorously refluxed for 24 h. After cooling to room
temperature, the excess hydride was decomposed with
MeOH, the solution concentrated to a gum by rotary
evaporation. The residue was taken up in 100% EtOH
(50 mL) and while cooling with an ice bath, saturated
with HCl(g). The acidic solution was then vigorously
refluxed for 18 h during which a fine white precipitate
formed. The suspension was cooled at 4 C for 8 h, col-
lected on a Buchner funnel, washed with Et2O, and
.
added. BF3 Et2O (1.75 mL) was added by syringe and a
white precipitate gradually formed. After 18 h, solid
Na2CO3 was added, and after stirring for 1 h, the sus-
pension was filtered through a silica pad on a coarse
glass frit. The filtrate was carefully rotary evaporated to
a light-yellow oil which was vacuum distilled (11 mm,
70 ꢁC) to isolate the product (21.4 g, 66%). H NMR
1
(CDCL3) d 4.046 (s, 2H), 3.753 (s, 2H), 1.240 (s, 9H);
MS (CI/NH3) 164 (M++18).
Glycolic acid tert-butyl ether. The above ester methyl
glycolate tert-butyl ether (12.0 g, 82.3 mmol) was added
to EtOH (200 mL) in which KOH (82.3 mmol) was been
previously dissolved and the solution was stirred for 18
h. The EtOH was removed by rotary evaporation and
the residue taken up in H2O. The aqueous solution was
acidified with 3 N HCl and immediately extracted with
CHCl3 (3ꢀ100 mL). The combined extracts were dried
dried in vacuo (0.50 g, 63.4%). 1H NMR(D O) d 3.874
2
(t, 2H, J=4.8), 3.503 (tt, 1H, J=11.5,3.0), 3.288 (t, 2H,
J=4.8), 2.681 (br.d, 1H, J=10.8), 1.709 (q, 1H,
J=11.7); 13C NMR(D 2O) d 59.803, 54.946, 49.846,
33.089; MS (FAB/glycerol) 262 (M++1). Anal. calcd
for C12H27N3O3(HCl)3; C, 38.87; H, 8.17; N, 11.33.
Found: C, 38.93; H, 8.10; N, 10.94.
over Na2SO4, filtered, and rotary evaporated to leave a
1
clear oil (10.5 g, 97%). H NMR(CDCL ) d 4.049 (s,
3
2H), 1.27 (s, 9H); Mass Spect. (CI/NH3) 150 (M++18).
The crude acid (8.4 g, 63.6 mmol), N-hydroxy-
succinimide (7.32 g, 6.37 mmol), and EDC (12.2 g, 63.7
mmol) were suspended in EtOAc (200 mL) to which
DMF (100 mL) was added to form a clear solution.
After stirring for 18 h, the reaction solution was diluted
with EtOAc (100 mL) and sequentially extracted with
H2O (100 mL), 5% NaHCO3 (100 mL), and salt solu-
tion (100 mL). After drying over Na2SO4, filtration, and
rotary evaporation, a white solid (11.8 g) was obtained
cis,cis-1,3,5-Tris(1-ethoxyethylmercaptoacetamido) cyclo-
hexane (6). TACH (HBr)3 (5.00 g, 13.44 mmol) and
NaOH (1.61 g, 40.25 mmol) were dissolved in H2O (10
mL). After a clear solution had formed, benzene (120
mL) was added and the H2O was distilled off by use of a
Dean–Stark trap. The benzene solution was then con-
centrated to ꢂ30 mL. After cooling to room tempera-
ture, DMF (120 mL) was added followed by addition of
N-hydroxysuccinimidyl S-(1-ethoxyethyl) mercaptoace-
tate23 (10.53 g, 40.3 mmol) in DMF (20 mL). The reac-
tion solution was warmed mildly and stirred for 18 h.
The solvent was removed by vacuum rotary evaporation
and the residue suspended in EtOAc (300 mL). The
precipitate was collected on a Buchner funnel and
washed with EtOAc to leave the product as a white
powder (5.64 g, 74%). 1H NMR(DMSO- d6) d 7.967 (d,
1H, J=7.8), 4.766 (q, 1H, J=6.3), 3.70–3.55 (m, 2H),
3.44–3.32 (m, 1H), 3.106 (s, 2H), 1.829 (br. d, 1H,
J=11.1), 1.425 (d, 3H, J=6.0), 1.099 (t, 4H, J=6.6,
signal for the ring proton obscured by the signal for the
methyl group); 13C NMR(DMSO- d6) d 168.214,
80.826, 62.212, 44.964, 37.474, 31.547, 22.684, 14.891;
MS (CI/NH3) 568 (M++1). Anal. calcd for
C24H45N3O6S3; C, 50.74; H, 8.00; N, 7.40. Found: C,
50.48; H, 7.90; N, 7.35.
1
which was further dried under high vacuum. H NMR
(DMSO-d6) d 4.46 (s, 4H), 2.822 (s, 2H), 1.185 (s, 9H);
MS (CI/NH3) 247 (M++18). The active ester was used
without further purification.
cis,cis-1,3,5-triaminocyclohexane-tris(O-tert-butylglycol-
amide) (5). TACH (HBr)3 (2.5 g, 6.75 mmol) and
NaOH (0.81 g, 20.3 mmol) were dissolved in H2O (10
mL). After a clear solution had formed, benzene (180
mL) was added and the H2O eliminated by a Dean-
Stark trap. The benzene solution was then concentrated
to ca. 30 mL by distillation. After cooling to room
temperature, DMF (150 mL) was added followed by
addition of the above prepared active ester (5.0 g, 20.3
mmol) in DMF (30 mL). The reaction solution was

4292
K. Camphausen et al. / Bioorg. Med. Chem. 11 (2003) 4287–4293
N,N0,N00 -Tri(2-mercaptoethyl)-cis,cis-1.3.5-triaminocy-
clohexane trihydrochloride (3b). Amide 6 (4.087 g, 7.199
mmol) was suspended in THF (100 mL) in a two-necked
flask, cooled in an ice bath, and 1 M BH.3THF (86 mL)
was injected via syringe. The reaction was allowed to
come to room temperature over 1 h, after which the
reaction was vigorously refluxed for 120 h. After cooling
to room temperature, the excess hydride was decom-
posed with MeOH. The solution was concentrated to a
gum by rotary evaporation and held under vacuum for
72 h. The residue was taken up in 100% EtOH (200 mL)
and while cooling with an ice bath, saturated with
HCl(g). The acidic solution was then vigorously refluxed
for 18 h during which a fine white precipitate formed.
The suspension was cooled at 4 ꢁC for 8 h, collected on
a Buchner funnel, washed with Et2O, and dried in vacuo
(1.93 g, 64%). 1H NMR(D 2O) d 3.485 (tt, 1H, J=11.7,
3.6), 3.346 (t, 2H, J=6.9), 2.866 (t, 2H, J=6.9), 2.644
(br.d, 1H, J=8.7), 1.703 (q, 1H, J=12.0); 13C NMR
(D2O) d 54.940, 50.562, 33.268, 23.076; MS (FAB/gly-
cerol) 310 (M++1). Anal. calcd for C12H27N3S3(HCl)3;
C, 34.39; H, 7.23; N, 10.03. Found: C, 34.26; H, 7.41; N,
10.07.
MgSO4, filtered, and then rotary evaporated to leave the
1
product as an oil (7.75 g, 82%). H NMR(CDCl ) d
3
8.55 (m, 1H), (td, 1H, J=6.6, 1.8), (dt, 1H, J=8.1, 0.9),
(ddd, 1H, J=7.2, 4.8, 1.2), 3.870 (q, 1H, J=6.6), 2.62–
2.43 (m, 4H), 1.368 (dd, 3H, J=6.6, 0.9); 13C NMR
(CDCl3) d 164.57, 149.04, 136.34, 121.68, 120.82, 59.44,
53.99, 45.23, 22.66; MS (CI/NH3) 462 (M++1). Anal.
calcd for C27H39N7; C, 70.23; H, 8.53; N, 21.24. Found:
C, 70.12; H, 8.23; N, 21.45.
Biological methods
In vitro cellular proliferation assay. HUVEC (Clo-
netics), U251 and NIH-3T3 (ATCC) were grown in T75
flasks in EBM-2 media supplemented with the EGM-2
bullet (Clonetics). At confluence the cells were washed
once with sterile PBS and released with 1.5 mL of tryp-
sin. The cells were then placed in a 50-mL conical flask
containing 50 mL of supplemented EBM-2 media. Cells
were then evenly seeded into four 24-well plates and
allowed to adhere for 24 h at 37 C in a 5% CO2 atmo-
sphere. The media was then removed and the study drug
was added to the wells, four wells per drug dilution, in
400 mL of un-supplemented media and allowed to incu-
bate for 30 min. At which time 400 mL of media con-
taining twice the supplementation was added and
incubated for 72 h. Positive controls had stimulated
media alone while negative controls had un-stimulated
media added. Positive and negative controls were added
to every plate to generate an IC50 for each drug. After
72 h the media was removed, the cells were trypsinized
with 400 mL volume, and the total cell number counted
in a Coulter Counter. IC50 values were generated by
comparison of the proliferation of the cells, done in
quadruplicate, with various concentrations of drug ver-
sus the stimulated positive controls and the un-stimu-
lated negative controls. An IC50 value was that number
where the cellular proliferation was halfway between the
stimulated positive control and the unstimulated
negative control.
Tris[N-(2-pyridylmethylene)-2-aminoethyl]amine
(4a).
Tris(2-aminoethyl)amine (3.00 g, 20.5 mmol) and 2-
pyridine-2-carboxaldehyde (6.585 g, 61.55 mmol) were
dissolved in benzene (150 mL) and the resultant H2O
was removed by Dean–Stark distillation for 18 h. After
cooling to room temperature, the solution was decanted
and rotary evaporated to isolate the crude tris-imine
that was then taken up in MeOH (150 mL). The imine
was then treated with NaBH4 (2.34 g, 61.6 mmol) and
stirred for 18 h. The reaction solution was then rotary
evaporated to a solid and CHCl3 (125 mL), saturated
salt solution (50 mL), and 5% NaHCO3 (50%) were
added and vigorously stirred for 2 h. The mixture was
then poured into a separatory funnel, the organic layer
retained, dried over MgSO4, filtered, and then rotary
evaporated to leave the product as an oil (7.39 g, 86%).
1H NMR(CDCl ) d 8.480 (d, 1H, J=4.8, 0.09), 7.577
3
(dt, 1H, J=7.5, 2.1), 7.306 (d, 1H, J=7.8), 7.110 (dd,
1H, J=6.9, 4.5), 3.918 (s, 2H), 3.10 (br.s, 1H), 2.769 (t,
2H, J=5.7), 2.669 (t, 2H, J=5.7); 13C NMR(CDCl ) d
3
References and Notes
159.24, 149.08, 136.28, 122.14, 121.77, 54.70, 54.03,
47.05; MS (CI/NH3) 420 (M++1). Anal. calcd for
C24H33N7; C, 68.69; H, 7.94; N, 23.37. Found: C, 68.45;
H, 7.48; N, 23.69.
1. Folkman, J. Nat. Med. 1995, 1, 27.
2. Folkman, J. N. Engl. J. Med. 1971, 285, 1182.
3. Hu, G. F. J. Cell. Biochem. 1998, 69, 326.
4. Pan, Q.; Kleer, C.; van Golen, K.; Irani, J.; Bottema, K.;
Bias, C.; De Carvalho, M.; Mesri, E.; Robins, D.; Dick, R.;
Brewer, G.; Merajver, S. Cancer Res. 2002, 62, 4854.
5. Yoshi, J.; Yoshiji, H.; Kuriyama, S.; Ikenaka, Y.; Noguchi,
R.; Okuda, H.; Tsujinoue, H.; Nakatani, T.; Kishida, H.;
Nakae, D.; Gomez, D.; De Lorenzo, M.; Tejera, A.; Fukui, H.
Int. J. Cancer 2001, 94, 768.
6. Arnold, M.; Sasse, D. Cancer Res. 1961, 21, 761.
7. Apelgot, S.; Coppey, J.; Fromentin, A.; Guille, E.; Poupon,
M.-F.; Roussel, A. Anticancer Res. 1986, 6, 159.
8. Connolly, D.; Olander, J.; Heuvelman, D.; Nelson, R.;
Monsell, R.; Siegel, N.; Haymore, B.; Leimgruber, R.; Feder,
J. J. Biol. Chem. 1989, 264, 20017.
Tris[N-(2-pyridyl(2-ethylene))-2-aminoethyl]amine (4b).
Tris(2-aminoethyl)amine (3.00 g, 20.5 mmol) and 2-
acetylpyridine (7.46 g, 61.65 mmol) were dissolved in
benzene (150 mL) and the resultant H2O was distilled
off by use of a Dean–Stark trap. After cooling to room
temperature, the solution was decanted and rotary eva-
porated to isolate the crude tris-imine that was taken up
in MeOH (200 mL). The imine was then treated with
NaBH4 (2.34 g, 61.6 mmol) and stirred for 18 h. The
reaction solution was then rotary evaporated to a solid
and CHCl3 (125 mL), saturated salt solution (50 mL),
and 5% NaHCO3 (50%) were added and vigorously
stirred for 2 h. The mixture was then poured into a
separatory funnel, the organic layer retained, dried over
9. Engleka, K.; Maciag, T. J. Biol. Chem. 1992, 267, 11307.
10. Parke, A.; Battacherjee, P.; Palmer, R.; Lazarus, N. Am.
J. Clin. Path 1988, 137, 1121.

K. Camphausen et al. / Bioorg. Med. Chem. 11 (2003) 4287–4293
4293
11. Raju, K.; Alesandrii, G.; Zinche, M.; Gullino, P. J. Natl.
Cancer Inst. 1982, 69, 1183.
12. Brem, S.; Zagzag, D.; Tsanaclis, A.; Gately, S.; Elkouby,
M.; Brien, S. Amer. J. Path 1990, 137, 1121.
13. Brem, H.; Tsanaclis, A.; Zagzag, D. Neurosurgery 1990,
26, 391.
20. Park, G.; Dadachova, E.; Przyboroeska, A. M.; Lai, S.;
Ma, D.; Broker, G. A.; Rogers, R. D.; Planalp, R. P.; Brech-
biel, M. W. Polyhedron 2001, 3155.
21. O’Reilly, M.; Holmgren, L.; Shing, Y.; Chen, C.; Rosen-
thal, R.; Moses, M.; Lane, W.; Cao, Y.; Sage, H.; Folkman, J.
Cell 1994, 79, 315.
14. Walshe, J. Amer. J. Med. 1982, 21, 487.
15. Walshe, J. Lancet 1982, 1, 643.
22. Ma, D.; Lu, F.; Overstreet, T.; Milenic, D.; Brechbiel,
M. W. Nucl. Med. Biol. 2002, 29, 91.
16. Brewer, G.; Dick, R.; Johnson, V.; Brunberg, J.; Kluin,
K.; Fink, J. J. Lab. Clin. Med. 1998, 132, 264.
17. Brewer, G.; Dick, R.; Yuzbasiyan-Gurkin, V.; Tanakow,
R.; Young, A.; Kluin, K. Arch. Neurol. 1991, 48, 42.
18. Brewer, G.; Dick, R.; Grover, D.; LeClaire, V.; Tseng, M.;
Wicha, M.; Pienta, K.; Redman, B.; Thierry, J.; Sondak, V.;
Strawderman, M.; LeCarpentier, G.; Merajver, S. Clin. Cancer
Res. 2000, 6, 1.
23. Kasina, S.; Rao, T.; Srinivasan, A.; Sanderson, J.; Fitzner,
J.; Reno, J.; Beaumier, P.; Fritzberg, A. J. Nucl. Med. 1991,
32, 1445.
24. Sibert, J.; Cory, A. H.; Cory, J. G. Chem. Commun. 2002,
154.
25. Sukal, S.; Bradshaw, J.; He, J.; Yap, G.; Rheingold, A.;
Kung, H.; Francesconi, L. Polyhedron 1998, 17, 1.
26. Morgenstern-Badarau, I.; Lanbert, F.; Deroche, A.;
Cesario, M.; Guilhem, J.; Keita, B.; Nadjo, L. Inorg. Chim.
Acta 1998, 275–276, 234.
19. Bowen, T.; Planapl, R.; Brechbiel, M. Bioorg. Med. Chem.
Lett. 1996, 6, 807.