Nickel(II), copper(II) and zinc(II) binding properties and cytotoxicity of tripodal, hexadentate tris(ethylenediamine)--analogue chelators.

Article abstract of DOI:10.1039/b403476g

Three tripodal hexamine chelators based on cis,cis-1,3,5-triaminocyclohexane (tach) have been synthesized and their aqueous coordination chemistry with Ni(II), Cu(II) and Zn(II) is reported. The chelators have a 2-aminoethyl pendant arm attached to each nitrogen of tach, specifically 'tachen'(N,N',N''-tris(2-aminoethyl)cyclohexane-cis,cis-1,3,5-triamine), and two with S,S,S-chiral pendant arms, 'tachpn'(N,N',N''-tris(2-aminopropyl)cyclohexane-cis,cis-1,3,5-triamine) and 'tachbn'(N,N',N''-tris(2-amino-3-phenylpropyl)cyclohexane-cis,cis-1,3,5-triamine. These chelators complex Ni(II), Cu(II) and Zn(II) in aqueous or aqueous/methanolic medium. The crystalline products [M(II)L](X)2 are isolated, where M = Ni(II), Cu(II) or Zn(II), L = tachen, tachpn or tachbn, and X = ClO4-. Crystallographic study of selected tachpn and tachbn complexes shows the chelate arms are constrained in a Lambda(deltadeltadelta) configuration about M(II), which is attributed to their chirality. Solution UV-vis spectroscopy of the Ni(II) and Cu(II) complexes indicates six-coordination and little effect of the pendant arm substitution on ligand-field strength. The single exception is [Cu(tachbn)]2+, whose spectrum is consistent with five-coordination in solution. The cytotoxicities of tachen, tachpn and tachbn toward cultured cancer cells is in the order tachen < tachpn < tachbn < tachpyr, where tachpyr is the aminopyridyl chelator N,N',N''-tris(2-pyridylmethyl)cyclohexane-cis,cis-1,3,5-triamine. The cytotoxicity difference is attributed to an order of increasing lipophilicity, tachen < tachpn < tachbn.

Full text of DOI:10.1039/b403476g

View Article Online / Journal Homepage / Table of Contents for this issue
Nickel(II), copper(II) and zinc(II) binding properties and
cytotoxicity of tripodal, hexadentate tris(ethylenediamine) –
analogue chelators
Neng Ye,a Gyungse Park,a Ann M. Przyborowska,a Paula E. Sloan,a Thomas Clifford,b
Cary B. Bauer,d Grant A. Broker,e Robin D. Rogers,e Rong Ma,c Suzy V. Torti,c
Martin W. Brechbielb and Roy P. Planalp*a
a Department of Chemistry, University of New Hampshire, Durham, NH 03824-3598, USA
b Radiation Oncology Branch, National Institutes of Health, Bethesda, MD 20892, USA
c Department of Biochemistry, Wake Forest University Health Sciences, Winston-Salem,
NC 27157, USA
d Bruker AXS, Inc, 5465 East Cheryl Parkway, Madison, WI 53711, USA
e Department of Chemistry, The University of Alabama, Tuscaloosa, AL 35487, USA
Received 5th March 2004, Accepted 10th March 2004
First published as an Advance Article on the web 31st March 2004
Three tripodal hexamine chelators based on cis,cis-1,3,5-triaminocyclohexane (tach) have been synthesized and their
aqueous coordination chemistry with Ni(), Cu() and Zn() is reported. The chelators have a 2-aminoethyl pendant
arm attached to each nitrogen of tach, specifically ‘tachen’ (N,NЈ,NЉ-tris(2-aminoethyl)cyclohexane-cis,cis-1,3,5-
triamine), and two with S,S,S-chiral pendant arms, ‘tachpn’ (N,NЈ,NЉ-tris(2-aminopropyl)cyclohexane-cis,cis-1,3,5-
triamine) and ‘tachbn’ (N,NЈ,NЉ-tris(2-amino-3-phenylpropyl)cyclohexane-cis,cis-1,3,5-triamine. These chelators
complex Ni(), Cu() and Zn() in aqueous or aqueous/methanolic medium. The crystalline products [MIIL](X)2 are
isolated, where M = Ni(), Cu() or Zn(), L = tachen, tachpn or tachbn, and X = ClO4Ϫ. Crystallographic study of
selected tachpn and tachbn complexes shows the chelate arms are constrained in a Λ(δδδ) configuration about M(),
which is attributed to their chirality. Solution UV-vis spectroscopy of the Ni() and Cu() complexes indicates six-
coordination and little effect of the pendant arm substitution on ligand-field strength. The single exception is
[Cu(tachbn)]2ϩ, whose spectrum is consistent with five-coordination in solution. The cytotoxicities of tachen, tachpn
and tachbn toward cultured cancer cells is in the order tachen < tachpn < tachbn < tachpyr, where tachpyr is the
aminopyridyl chelator N,NЈ,NЉ-tris(2-pyridylmethyl)cyclohexane-cis,cis-1,3,5-triamine. The cytotoxicity difference
is attributed to an order of increasing lipophilicity, tachen < tachpn < tachbn.
Introduction
Externally-introduced chelating agents may have either
beneficial or detrimental effects on organisms. As examples,
the elimination of toxic metals or the correction of metal
imbalances are beneficial, while the deprivation of essential
ions or the promotion of toxic metal catalysis are detrimental
effects.1,2 For these reasons we have studied aqueous co-
ordination chemistry and structural chemistry of the first-row
transition metals Mn, Fe, Co, Ni, Cu and Zn with families of
tripodal chelators. We investigated coordination chemistry of
tachpyr, a hexadentate chelator based on cis,cis-1,3,5-triamino-
cyclohexane (tach) with N-pendant 2-pyridylmethyl groups.3–6
Tachpyr has significant cytotoxic effects on tumor cells,
for which chelation of Fe, Cu or Zn were suggested causes.7
Tachpyr-induced cell death occurs by apoptosis.8 Recently,
Scheme
1
Syntheses of tach-2-aminoethyl derivatives, indicating
chirality of tachpn and tachbn.
it has been shown to be independent of p53 activation and to
involve caspase activation.9
We recently expanded this work with the preparation of
2-aminoethyl-pendant arm complexes of tach wherein S-chiral
substituents have been introduced on the ethylene chain. The
metal-binding preferences of these hexamine chelators should
differ from those of the aminopyridyl family, due to changes in
donor properties. Moreover, cellular targeting and distribution
is expected to differ, due to diverse biological effects. This
article details the synthesis and metal complexation reactions
of the en-framework derivatives tachen, tachpn and tachbn
(Scheme 1). The structure and bonding of [ML]2ϩ (Scheme 2;
M = Ni(), Cu(), Zn(); L = tachen, tachpn, tachbn) have been
characterized by solution spectroscopic and X-ray crystallo-
graphic methods. The cytotoxicities of the chelators have been
determined and related to a lipophilicity effect.
Experimental
Materials and methods
All the materials listed below were of a research grade or a
spectro-quality grade in the highest purity available and were
generally used without further purification except Et2O. Et2O
was distilled from Na and used immediately. Anhydrous grade
DMF, MeOH, CH3CN and tert-butanol and the salts
Zn(ClO4)2ؒ6H2O, Cu(ClO4)2ؒ6H2O and Ni(ClO4)2ؒ6H2O were
obtained from Aldrich. Anhydrous grade EtOH was obtained
from Pharmco. Yields of metal complexes were not optimized.
DMSO-d6, and CH3CN-d3 were obtained from Cambridge
Isotope Laboratories. MOPS buffer was prepared by dissolving
D a l t o n T r a n s . , 2 0 0 4 , 1 3 0 4 – 1 3 1 1
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4
1304

View Article Online
near dryness (vacuum rotary evaporation), the solid residue was
suspended in EtOAc (200 cm3) and vigorously stirred. In the
case of the phenylalanine derivative, the product precipitated
directly from the DMF solution. The respective triamides were
collected on a Buchner funnel, washed with EtOAc followed by
hexanes and dried under vacuum.
Glycine triamide (4). Tachؒ3HBr (6.52 g, 17.5 mmol) was
treated as described above with NaOH (2.102 g, 52.6 mmol)
and BOC-glycine p-nitrophenyl ester (15.57 g, 52.6 mmol) to
1
M
tachen (R=H)
tachpn (R=Me)
tachbn (R=Bn)
produce a white solid (54%). H NMR (d6-DMSO) δ 7.742 (d,
1H, J = 7.8), 6.839 (t, 1H, J = 6.0), 3.597 (m, 1H), 3.456 (d, 2H,
J = 5.7), 1.797 (br d, 1H, J = 11.4), 1.359 (s, 9H), 1.068 (q, 1H,
J = 11.7); 13C NMR (d6-DMSO) δ 168.58, 155.95, 77.99, 44.78,
42.14, 37.74, 28.15; MS (CI, NH3) 601 (Mϩ ϩ 1). Anal. Calc.
for C27H48N6O9: C, 53.97; H, 8.07; N, 13.99. Found: C, 53.70;
H, 8.00; N, 13.85%.
Ni()
Cu()
Zn()
7a
8a
9a
7b
8b
9b
7c
8c
9c,9cЈa
aThe anions X2 = (CIO4
) with exception of 9cЈ, where X2 =
2
Ϫ
(ClϪ)(ClO4Ϫ).
Scheme 2 Structures and numbering of the metal complexes of
Alanine triamide (5). Tachؒ3HBr (7.0 g, 18.8 mmol) was
treated as described above with NaOH (2.26 g, 56.45 mmol)
and BOC--alanine p-nitrophenyl ester (17.5 g, 56.45 mmol) to
tachpn and tachbn.a
solid 3-(N-morpholino)propanesulfonic acid (Sigma-Aldrich)
in water, then adding NaOH to reach pH 7.4, and finally
appropriate dilution to 0.1 or 0.2 M. The cis,cis-1,3,5-triamino-
cyclohexane (tach) trihydrobromide was prepared as reported
in the literature.4 The BOC-amino acid nitrophenyl esters
were obtained from Sigma. All other anhydrous solvents and
reagents were obtained from Aldrich.
23
produce a white solid (86.5%). [α]D ϩ19.39Њ (c 0.98, MeOH);
1H NMR (d6-DMSO) δ 7.742 (d, 1H, J = 7.8), 6.773 (d, 1H,
J = 7.5), 3.874 (m, 1H), 3.580 (m, 1H), 1.794 (br d, 1H,
J = 11.7), 1.363 (s, 9H), 1.128 (d, 4H, J = 7.0, one proton
obscured); 13C NMR (d6-DMSO) δ 171.78, 154.93, 77.88, 49.60,
44.71, 37.65, 28.17, 18.35; MS (CI, NH3) 643 (Mϩ ϩ 1). Anal.
Calc. for C30H54N6O9: C, 56.04; H, 8.84; N, 13.08. Found: C,
55.85; H, 8.39; N, 12.96%.
1H and 13C NMR were obtained using Varian Gemini 300,
Bruker AM360, or Varian Mercury 400 instruments and chem-
ical shifts are reported in ppm on the δ scale, referenced against
internal TMS, TSP, or solvent. Proton chemical shifts are anno-
tated as follows: ppm (multiplicity, spin system if determined,
integral, coupling constant (Hz)). UV-Vis electronic spectra
were measured using a Varian Cary 5 or Varian Cary 50-Bio
UV-vis spectrometer with 1 cm3 or 3 cm3 quartz cuvettes (1 cm
path-length) and in MOPS buffer (0.1 M, pH 7.3) unless
otherwise indicated. Chemical ionization mass spectra (CI-MS)
were obtained on a Finnegan 3000 instrument. Fast-atom
bombardment mass spectra (FAB-MS) were taken on an Extrel
400 instrument. Samples were desorbed from mixtures of
glycerol, thioglycerol (“TG”), DMSO, nitrobenzyl alcohol
(“NBA”) or Magic Bullet (“MB”, a 3 : 1 mixture of dithio-
threitol and dithioerythritol) as indicated. Electrospray ioniz-
ation (ESI) mass spectra in the positive ion detection mode
were obtained on a Finnegan LCQ Classic instrument with dual
optical Paul traps and Picoview nanospray source. Assignments
were based on calculations by the program developed by
Scientific Instrument Services (http://www.sisweb.com/cgi-bin/
mass10.pl, accessed December 2003). Lipophilicity calculations
were made with the CLogP program (Biobyte Corporation)
as provided on the website of the Virtual Computational
Chemistry Laboratory (http://vcclab.org/lab/alogps/start.html,
accessed December 2003). Optical rotations were obtained
with a Rudolf Autopol III polarimeter. Elemental analysis was
performed by Atlantic Microlabs (Atlanta, Georgia).
Phenylalanine triamide (6). Tachؒ3HBr (3.2 g, 8.6 mmol) was
treated with NaOH (1.03 g, 25.8 mmol) and BOC--phenyl-
alanine p-nitrophenyl ester (9.96 g, 25.8 mmol) as described
above to produce a white solid (70.5%). This material was
insoluble in DMF or DMSO, thus the optical rotation and
NMR spectra are of the derivative obtained from the treatment
of a small portion of the product with triflouroacetic acid for
4 h, followed by rotary evaporation and vacuum drying to leave
23
1
a white foam. [α]D ϩ45.0Њ (c 0.8, MeOH); H NMR (D2O)
δ 7.40–7.10 (m, 5H), 4.025 (t, 1H, J = 7.8), 3.570 (br t, 1H),
3.155 (dd, 1H, J = 13.5, 6.6), 3.041 (dd, 1H, J = 13.8, 8.7), 1.613
(br d, 1H, J = 11.7), 0.721 (q, 1H, J = 12.0); 13C NMR (D2O)
δ 171.03, 137.03, 132.66, 132.30, 57.56, 48.33, 40.01, 38.55; MS
(CI, NH3) 871 (Mϩ ϩ 1). Anal. Calc. for C48H66N6O9: C, 66.17;
H, 7.65; N, 9.65. Found: C, 65.98; H, 7.60; N, 9.56%.
General procedures for preparation of hexamines (1–3)
Deprotection of triamides 4, 5. The triamide was added to
anhydrous 1,4-dioxane saturated with HCl(g) (200 cm3) and
stirred for 18 h. The precipitate was collected, washed with
Et2O, and dried under vacuum.
Deprotection of triamide 6. The phenylalanine triamide was
treated with trifluoroacetic acid (50 cm3) for 6 h which, after
solvent removal by rotary evaporation and vacuum drying, left
a white foam.
Ligand syntheses
Borane reduction. The deprotected material was washed into
a three-necked round bottom flask with THF (150 cm3) and
1 M BH3ؒTHF (6 eq. for HCl deprotection, 28 eq. for tri-
fluoroacetic acid deprotection) was added while cooling the
suspension with an ice-bath. The reaction mixture was then
vigorously refluxed for 6 d, after which any excess borane was
decomposed with MeOH. The solution was rotary evaporated
to a gummy solid and taken up in 100% EtOH (150 cm3). The
EtOH solution was saturated with HCl(g) while cooling with an
ice bath and then the acidic solution was refluxed for 6 h during
which a white precipitate formed. The flask was held at 4 ЊC for
8 h and the precipitate was collected on a Buchner funnel,
washed with Et2O, and dried in vacuo.
General procedure for preparation of tach BOC-amino acid
triamides. Tachؒ3HBr (1 eq.) was treated with NaOH (3 eq.) in
H2O (10 cm3) in a two-necked flask at room temperature fitted
with a Dean–Stark trap and condenser. After formation of a
clear homogeneous solution, benzene (200 cm3) was added, and
the H2O was removed via azeotropic distillation. The benzene
solution was then reduced to ca. 25 cm3, cooled to room tem-
perature, diluted with DMF (150 cm3), and then the BOC-
amino acid nitrophenyl ester (3 eq.) in DMF (30 cm3) was
added. The temperature was raised to ca. 45 ЊC and the bright
yellow solution was stirred for 18 h. In the case of the glycyl and
alaninyl derivatives, the DMF solution was concentrated to
D a l t o n T r a n s . , 2 0 0 4 , 1 3 0 4 – 1 3 1 1
1305

View Article Online
Tachenؒ6HCl (1ؒ6HCl). BOC-triamide 4 (8.05 g, 13.4 mmol)
was deprotected as described above and isolated as the hydro-
chloride salt (5.13 g, 93%). The hydrochloride salt was
converted to the free base triamide by ion exchange resin (AG
1-X8 Biorad anion exchange resin, hydroxide form). After
removal of water at reduced pressure the oily residue was dried
azeotropically with benzene, and then used without further
purification (4.02 g, 86%). Free-base triamide (4.02 g, 10.78
mmol) was reduced with 1 M BH3ؒTHF (90 cm3, 90 mmol)
refluxing overnight. A further portion of 1 M BH3ؒTHF
(50 cm3, 50 mmol) was added and the mixture refluxed for
another 12 h. After the acid work-up, the product was obtained
as a white solid by recrystallization from aqueous ethanol.
Yield 3.57 g (7.48 mmol, 69.3%). 1H NMR (D2O. 25 ЊC) δ 3.58
(m, 4H, NCH2CH2N), 3.45 (br s, 1H, cyclohexyl methine H’s),
2.76 (br d, 1H, equatorial cyclohexyl methylene H’s), 1.83
(d of t, 1H, axial cyclohexyl methylene H’s); 13C NMR (D2O)
δ 51.42, 40.74, 35.16, 30.65; MS (CI, NH3) 259 (Mϩ ϩ 1). Anal.
Calc. for C12H30N6(HCl)6: C, 30.20; H, 7.62; N, 17.61. Found:
C, 29.94; H, 7.64; N, 17.69%.
(M – 2ClO4Ϫ). λmax/cmϪ1 × 10Ϫ3 (MOPS) 29.7 (ε/cmϪ1 MϪ1 12.6),
19.0 (9.1), 12.5 (11.3), 11.5 sh. Anal. Calc. for C12.5H32Cl2-
N6NiO8.5 (7aؒ0.5MeOH): C, 28.22; H, 6.06; N, 15.80. Found:
C, 28.44; H, 6.02; N, 16.25%.
[Ni(tachpn)](ClO4)2ؒ(7b) and 7bؒMeOH. Solid tachpnؒ6HCl
(0.0400 g, 7.79 × 10Ϫ5 mol) was dissolved in aqueous NaOH
(4.67 cm3, 0.1 M). The reaction solution was filtered and dried
under reduced pressure for 12 h. The residue was extracted with
anhydrous EtOH (3 cm3) and the filtrate was reacted with a
solution of Ni(ClO4)2ؒ6H2O (0.0282 g, 7.71 × 10Ϫ5 mol) in
EtOH (2 cm3). The cloudy solution became clear pink after
stirring and warming to ca. 70 ЊC for 1 min. Pink needles
formed while cooling. After standing for 1 h, the crystals were
isolated by decanting the supernatant, washing with ether, and
drying at 50 ЊC in the air. Yield 0.024 g (4.30 × 10Ϫ5 mol,
55.1%). Slow diffusion of Et2O into a MeOH solution
provided pink prisms of a solvate (7bؒMeOH) suitable for
X-ray crystallography. MS (FAB/MB): 457 (M Ϫ ClO4Ϫ).
λmax/cmϪ1 × 10Ϫ3 (MeCN) 29.7 (ε/cmϪ1 MϪ1 15.3), 19.6 (9.4),
12.4 (12.0), 11.5 sh. Anal. Calc. for C15H36Cl2N6NiO8: C, 32.50;
H, 6.55; N, 15.16. Found: C, 32.51; H, 6.71; N, 14.98%.
Tachpnؒ6HCl (2ؒ6HCl). BOC-triamide 5 (1.59 g, 2.45 mmol)
was deprotected as described above and after isolation and
drying, reduced with 1 M BH3ؒTHF (14.9 cm3). After acidic
work-up, the product was obtained as a white solid (86.5%).
[Ni(tachbn)](ClO4)2ؒH2O (7cؒH2O). Tachbnؒ6HCl (0.0467 g,
6.25 × 10Ϫ5 mol) was dissolved in aqueous NaOH (3.75 cm3,
0.1 M). The reaction solution was filtered and dried under
reduced pressure for 12 h. The residue was extracted with
anhydrous EtOH (6 cm3) and the filtrate was reacted with
Ni(ClO4)2ؒ6H2O (0.023 g, 6.25 × 10Ϫ5 mol) in EtOH (6 cm3).
The cloudy solution became clear pink after stirring and warm-
ing to ca. 70 ЊC for 1 min. The volume was reduced to 8 cm3
under reduced pressure to afford pink needles. These were
recrystallized by warming to 70 ЊC followed by cooling to
ambient temperature over 1 h. Decantation of supernatant,
washing with ether, and drying at 50 ЊC in air provided 0.034 g
(4.23 × 10Ϫ5 mol, 67.7% yield) of pink needles. MS (FAB/MB)
685 (M Ϫ ClO4Ϫ). λmax/cmϪ1 × 10Ϫ3 (MeOH–H2O, 1 : 1 v/v) 19.1
(ε/cmϪ1 MϪ1 14.4), 12.4 (11.5), and 11.5 sh. Anal. Calc. for
C33H50N6Cl2NiO9 (7cؒH2O): C, 49.22; H, 6.36; N, 10.44. Found:
C, 49.26; H, 6.08; N, 10.35%.
23
1
[α]D 7.84Њ (c 1.02, H2O); H NMR (D2O, 25 ЊC) δ 3.82 (m,
AMMЈXXЈ, 1H, cyclohexyl methine H’s), 3.60 (m, 1H,
NHC(Ha)(Hb)C(Hc)(CH3)NH2), 3.51, 3.44 (each d of d, each
1H, NHC(Ha)(Hb)C(Hc)(CH3)NH2, Ha, Hb are a diastereotopic
pair), 2.73 (br d, 1H, equatorial cyclohexyl methylene H’s), 1.82
(d of t, 1H, axial cyclohexyl methylene H’s); 13C NMR (D2O)
δ 52.48, 47.08, 44.67, 30.86, 15.31; MS (CI, NH3) 301 (Mϩ ϩ 1).
Anal. Calc. for C15H36N6(HCl)6(H2O)4: C, 30.46; H, 7.51; N,
14.21. Found: C, 30.54; H, 7.59; N, 14.31%.
Tachbnؒ6HCl (3ؒ6HCl). BOC-triamide 6 (1.26 g, 1.45 mmol)
was deprotected as described above and after isolation and
vacuum drying, reduced with 1 M BH3ؒTHF (40.3 cm3). After
acidic work-up, the product was obtained as a white solid
23
1
(74%). [α]D ϩ31.63Њ (c 0.98, H2O); H NMR (D2O) δ 7.32,
7.22, 7.19 (t, t, d, 5H, CH2–C6H5)), 3.84 (m, 1H, N(H)C(Ha)-
(Hb)C(Hc)(C(Hd)(He)–Ph)NH2), 3.26 (m, 3H, cyclohexyl
methine H’s and N(H)C(Ha)(Hb)C(Hc)(C(Hd)(He)–Ph)NH2),
3.05 (m, 2H, N(H)C(Ha)(Hb)C(Hc)(C(Hd)(He)–Ph)NH2), 2.61
(br d, 1H, equatorial cyclohexyl methylene H’s), 1.69 (br q, 1H,
axial cyclohexyl methylene H’s); 13C NMR (D2O) δ 133.04,
128.35, 128.26, 126.93, 118.02, 51.92, 49.45, 45.16, 35.18, 30.18;
MS (CI, NH3) 529 (Mϩ ϩ 1). Anal. Calc. for C33H48N6(HCl)6-
(H2O)2: C, 50.57; H, 7.48; N, 10.73. Found: C, 50.46; H, 7.38;
N, 10.72%.
[Cu(tachen)](ClO4)2 (8a). Solid Na2CO3 (0.0482 g, 4.22 ×
10Ϫ4 mol) was added to aqueous tachenؒ6HCl (0.0671 g, 1.41 ×
10Ϫ4 mol, in 2 cm3 H2O). The solution was heated to 70 ЊC until
5 min after gas evolution ceased. After drying under reduced
pressure (6 h), the solid residue was extracted with anhydrous
MeOH (3.5 cm3) and sonicated for 10 min. The extract of
tachen was then decanted from the white solid. Methanolic
Cu(ClO4)2ؒ6H2O (0.304 cm3 of 0.103 M solution) was added
dropwise to 1 cm3 of methanolic tachen while heating and
stirring in a warm water bath (5 min at 60 ЊC), affording a blue
solution. Vapor-phase diffusion of Et2O yielded a mixture of
dark blue prisms and lighter blue feathery crystals. The super-
natant was decanted and the crystals were dried under reduced
pressure. The solid was recrystallized from MeCN by vapor-
phase diffusion of Et2O to obtain shiny, feathery blue crystals
(0.0054 g, 9.60 × 10Ϫ6 mol, 33% yield). MS (ESI): m/1, 420 amu
(M Ϫ ClO4Ϫ); m/2, 160.5 amu (M Ϫ 2ClO4Ϫ). λmax/cmϪ1 × 10Ϫ3
(MOPS) 18.0 (ε/cmϪ1 MϪ1 87). Anal. Calc. for C12H30Cl2-
CuN6O8: C, 27.67; H, 5.81; N, 16.14. Found: C, 28.17; H, 5.99;
N, 16.21%.
Metal complexes
CAUTION: Metal perchlorate salts may be explosive and
should be handled with due care!
[Ni(tachen)](ClO4)2ؒ0.5MeOH (7aؒ0.5MeOH). The tachen
free base was prepared by adjusting an aqueous solution of
1ؒ6HCl to pH > 11 followed by evaporation to dryness under
reduced pressure. The colorless oil was used without further
purification. Tachen free base (0.50 g, 1.9 mmol) was refluxed
in aqueous ethanol (10 cm3, 1 : 1) and Ni(ClO4)2ؒ6H2O (0.50 g,
1.4 mmol) in aqueous ethanol (5 cm3, 1 : 1) added dropwise.
Initially a blue solid formed, which on further heating, gave
a pink solution. The solution was filtered hot and allowed to
cool. Over a few hours pink plates formed. These were filtered
off and washed with ethanol (2 cm3 × 2) and diethyl ether
(2 cm3 × 2) giving a crude yield of 0.72 g (1.4 mmol, 100%).
Analytically pure material was recrystallized from aqueous
methanol. MS (ESI): m/1, 415 amu (M Ϫ ClO4Ϫ); m/2, 158 amu
[Cu(tachpn)](ClO4)2 (8b). Tachpnؒ6HCl (0.0400 g, 7.71 ×
10Ϫ5 mol) was dissolved in aqueous NaOH (4.63 cm3 0.1 M).
After standing 1 h, the reaction solution was filtered and dried
under reduced pressure for 12 h. The residue was extracted into
anhydrous EtOH (3 cm3), filtered and added to Cu(ClO4)2ؒ
6H2O (0.0286 g, 7.72 × 10Ϫ5 mol) in EtOH (2 cm3). The solution
became cloudy blue and the solid product precipitated out
immediately. After standing for 1 h, the product was isolated by
D a l t o n T r a n s . , 2 0 0 4 , 1 3 0 4 – 1 3 1 1
1306

View Article Online
decanting the supernatant, washing with ether, and drying
under reduced pressure. An analytically-pure blue powder was
obtained in 28% yield (0.0126 g, 2.16 × 10Ϫ5 mol). MS (FAB/
MB): 462 (M Ϫ ClO4Ϫ). λmax/cmϪ1 × 10Ϫ3 (MOPS) 18.0 (ε/cmϪ1
MB): 463 (M Ϫ ClO4Ϫ). Anal. Calc. for C15H38Cl2N6O9Zn
(9bؒH2O): C, 30.91; H, 6.57; N, 14.42. Found: C, 31.13; H, 6.56;
N, 13.94%.
M
Ϫ1 97). Anal. Calc. for C15H38N6Cl2CuO9 (8bؒH2O): C, 31.01;
[Zn(tachbn)](ClO4)2ؒH2O (9cؒH2O). Tachbnؒ6HCl (0.0450 g,
6.03 × 10Ϫ5 mol) was dissolved in aqueous NaOH (3.62 cm3,
0.1 M). The reaction solution was filtered and dried under
reduced pressure for 12 h. The residue was extracted with
anhydrous EtOH (5 cm3) and the filtrate was reacted with
Zn(ClO4)2ؒ6H2O (0.0224 g, 6.03 × 10Ϫ5 mol) in EtOH (3 cm3).
The colorless solution was refluxed for 1.5 h, and then cooled to
ambient temperature over 8 h to provide colorless prisms in
H, 6.26; N, 14.47. Found: C, 31.05; H, 6.26; N, 14.18%.
[Cu(tachbn)](ClO4)2 (8c) and 8cؒH2O. Aqueous tachbnؒ6HCl
(0.0400 g in 3 cm3, 5.36 × 10Ϫ5 mol) was neutralized with
aqueous NaOH (3.21 cm3, 0.1 M). After standing for 1 h, the
solution was filtered and dried under reduced pressure for 14 h.
The residue was extracted into anhydrous EtOH (5 cm3) and
filtered. The filtrate was reacted with a blue solution of
Cu(ClO4)2ؒ6H2O (0.0199 g, 5.36 × 10Ϫ5 mol) in EtOH (5 cm3)
affording a clear purple solution. This was dried under reduced
pressure giving a blue solid that was recrystallized from CH3CN
by vapor-phase diffusion of Et2O. The blue prisms and plates,
suitable for X-ray crystallography, were isolated and dried
under reduced pressure (45% yield, 0.0271 g, 1.32 × 104 mol).
MS (FAB/glycerol): 690 (M Ϫ ClO4Ϫ). λmax/nm (MeCN–MOPS,
1 : 5 v/v) 15.8 with high-energy shoulder (ε/cmϪ1 MϪ1 60). Anal.
Calc. for C33H50Cl2CuN6O9 (8cؒH2O): C, 48.98; H, 6.23; N,
10.38. Found: C, 49.03; H, 6.04; N, 10.29%.
1
49% yield (0.0246 g, 3.11 × 10Ϫ5 mol). H NMR (D2O, 25 ЊC):
δ 7.46, 7.38, 7.34 (t, t, d, AAЈMXXЈ, 2H, 1H, 2H, CH2–C6H5),
3.27 (br s, AMMЈXXЈ, 1H, cyclohexyl methine H’s), 2.89 (br m,
2H, N(H)C(Ha)(Hb)C(Hc)(C(Hd)(He)–Ph)NH2) and N(H)-
C(Ha)(Hb)C(Hc)(C(Hd)(He)–Ph)NH2), 2.77 (br d of d, 2H,
NH2C(H)(C(Ha)(Hb)–Ph)C(Hc)(Hd)N(H)), 2.60 (apparent t,
2H, N(H)C(Ha)(Hb)C(Hc)(C(Hd)(He)–Ph)NH2), 2.24 (br d,
AMMЈXXЈ, 1H, equatorial cyclohexyl methylene H’s); 1.90 (br
d, AMMЈXXЈ, 1H, axial cyclohexyl methylene H’s). 13C NMR
(DMSO-d6, 25 ЊC): δ 138.02, 129.04, 128.59, 126.45 (1C, 2C,
2C, 1C, CH2–C6H5); 52.04 (1C, cyclohexyl methine C’s); 50.01
(1C, NH2–(CH)(CH2)–CH2–Ph); 49.34 (1C, NH2–(CH)(CH2)–
CH2–Ph); 41.83 (1C, NH2–(CH)(CH2)–CH2–Ph); 30.48 (1C,
cyclohexyl methylene C’s). MS (FAB/MB): 691 (M Ϫ ClO4Ϫ).
Anal. Calc. for C33H50Cl2N6O9Zn (9cؒH2O): C, 48.87; H, 6.21;
N, 10.36. Found: C, 49.39; H, 6.13; N, 10.28%.
[Zn(tachen)](ClO4)2 (9a). Tachenؒ6HCl (0.0601 g, 1.26 × 10Ϫ4
mol) in 3 cm3 H2O was treated with aqueous NaOH (0.756 cm3,
1.0 M), and the resulting solution was dried under reduced
pressure for 18 h. The residue was extracted with anhydrous
MeOH (3.4 cm3) and the extract of tachen was decanted away
from the white NaCl(s). Zn(ClO4)2ؒ6H2O (0.0240 g, 6.45 × 10Ϫ5
mol) in MeOH (1 cm3) was added to 1.7 cm3 of the methanolic
tachen giving a slightly cloudy solution. Vapor-phase diffusion
of Et2O yielded a white solid that was isolated by decantation
of the supernatant and dried in air at 40 ЊC. The solid was taken
up in MeCN (2 cm3) leaving a white residue; the clear liquid was
decanted away and subjected to vapor-phase diffusion of Et2O
yielding a white solid that was fully soluble in MeCN (0.020 g,
3.8 × 10Ϫ5 mol, 61%). 1H NMR (D2O, 25 ЊC): 3.29 (br s,
AMMЈXXЈ, 1H, cyclohexyl methine H’s); 3.11, 2.76, 2.53
(3 multiplets, 1H, 2H, 1H, NHC(Ha)(Hb)C(Hc)(Hd)NH2, 4
diastereotopic H’s); 2.23 (br d, AMMЈXXЈ, 1H, equatorial
cyclohexyl methylene H’s); 1.88 (br d, AMMЈXXЈ, 1H, axial
cyclohexylmethyleneH’s).MS(ESI):m/1,421amu(MϪClO4Ϫ);
m/2, 161 amu (M Ϫ 2ClO4Ϫ). Anal. Calc. for C12H30Cl2N6O8Zn:
C, 27.57; H, 5.79; N, 16.08. Found: C, 27.48; H, 5.82; N,
15.75%.
[Zn(tachbn)]Cl(ClO4)ؒ0.5H2O (9cЈؒ0.5H2O). Tachbnؒ6HCl
(0.0262 g, 3.50 × 10Ϫ5 mol) was dissolved in aqueous NaOH
(2.10 cm3, 0.1 M). The reaction solution was filtered and dried
under reduced pressure for 12 h. The residue was extracted with
MeOH (5 cm3) and the filtrate was reacted with Zn(ClO4)2ؒ
6H2O (0.0048 g, 3.52 × 10Ϫ5 mol) in MeOH (3 cm3). The
colorless solution was refluxed for 1.5 h, and then cooled to
25 ЊC over 18 h by standing in a Dewar flask of water initially
at 70 ЊC. The colorless prisms obtained were suitable for
X-ray study (0.0052 g, 3.5 × 10Ϫ6 mol, 20%). Anal. Calc. for
C33H49Cl2N6O4.5Zn (9cЈؒ0.5H2O): C, 53.70; H, 6.69; N, 11.39.
Found: C, 53.98; H, 6.79; N, 11.22%.
Solution-phase complexation
Solution complexation studies for Ni(), Cu() and Zn() were
carried out as described below and the solutions obtained were
1
studied by UV-vis (Table 1) or H NMR spectroscopy. The
shapes and positions of peaks, or chemical shifts and coupling
constants, from solution reaction spectra agreed well with those
of the isolated metal complexes in solution. The yields of solu-
tion complexation were not determined, although the measured
absorbances suggested >75% yield of solution complexation.
[Zn(tachpn)](ClO4)2ؒH2O (9bؒH2O). Tachpnؒ6HCl (0.0316 g,
6.09 × 10Ϫ5 mol) was dissolved in aqueous NaOH (3.58 cm3,
0.1 M). The reaction solution was filtered and dried under
reduced pressure for 12 h. The residue was extracted with
anhydrous EtOH (5 cm3) and the filtrate was reacted with
Zn(ClO4)2ؒ6H2O (0.0227 g, 6.09 × 10Ϫ5 mol) in EtOH (3 cm3),
forming a white solid. The supernatant was decanted and the
solid was washed with minimal Et2O and dried under reduced
pressure. The solid was dissolved in CH3CN (3 cm3); slow diffu-
sion of Et2O into the solution at 5 ЊC provided small colorless
prisms. These were isolated and dried under reduced pressure
Ni(II) complexes. Aqueous Lؒ6HCl (L = tachen, tachpn or
tachbn, 100 µL, 0.1 M) was neutralized with Na2CO3 (100 µL,
0.3 M), and NiCl2 (100 µL, 0.1 M) was added producing a clear
pink solution in the case of tachen and a pink-white precipitate
with tachpn and tachbn. MOPS (0.1 M, pH 7.3) was added to
the tachen reaction providing a pink solution. MeCN (200 µL)
was added to the others and the suspension was warmed (60 ЊC,
4 min) to afford a pink solution and a trace of white precipitate.
A further 100 µL of H2O and 400 µL of MOPS (0.1 M) were
added to give final reactant concentrations of 1 × 10Ϫ2 M.
Similar results were obtained using MeOH or DMSO in place
of MeCN for tachpn and tachbn.
1
affording a 54% yield (0.0186 g, 3.29 × 10Ϫ5 mol). H NMR
(D2O, 25 ЊC): δ 3.26 (br s, AMMЈXXЈ, 1H, cyclohexyl methine
H’s); 2.74 (m, 1H, NHC(Ha)(Hb)C(Hc)(CH3)NH2); 2.68
2
(br d, 1H, NH–C(Ha)(Hb)C(Hc)(CH3)NH2, JHa–Hb = 12 Hz);
2.47 (apparent t, 1H, NH–C(Ha)(Hb)C(Hc)(CH3)NH2, 2JHa–Hb
=
3JHb–Hc = 12 Hz); 2.21 (br d, AMMЈXXЈ, 1H, equatorial
cyclohexyl methylene H’s), 1.87 (br d, AMMЈXXЈ, 1H, axial
cyclohexyl methylene H’s); 1.17 (d, 3H, NH–C(Ha)(Hb)-
C(Hc)(CH3)NH2, 3JH–Hc = 6 Hz). 13C NMR (CH3CN-d3, 25 ЊC):
δ 70.53 (1C, cyclohexyl methine C’s); 69.92 (1C, NH2(CH)–
CH3); 66.09 (1C, NH–CH2–(CH)NH2CH3); 50.38 (1C, cyclo-
hexyl methylene C’s); 41.83 (1C, NH2(CH)–CH3). MS (FAB/
Cu(II) complexes. Aqueous Lؒ6HCl (L = tachen, tachpn
or tachbn, 20 µL, 0.1 M) was neutralized with Na2CO3 (20 µL,
0.3 M), and Cu(ClO4)2 (20 µL, 0.1 M) was added to provide a
clear periwinkle-blue solution in the case of tachen, and blue–
D a l t o n T r a n s . , 2 0 0 4 , 1 3 0 4 – 1 3 1 1
1307

View Article Online
white precipitate with tachpn and tachbn. MOPS buffer was
added to the tachen reaction (940 µL, 0.1 M, pH 7.3). MeCN
(200 µL) was added to the others and the suspension was
warmed (60 ЊC, 4 min) to afford a blue-periwinkle solution.
Further 400 µL of H2O and 400 µL of MOPS (0.1 M) were
added to give final reactant concentrations of 2 × 10Ϫ3 M.
Similar results were obtained when the reaction of tachbn was
repeated with DMSO or MeOH in place of MeCN.
< 19.93Њ, of which 1501 reflections had I > 2σ(I ). Final results:
R(F) = 0.052, Rw(F) = 0.101, GOF = 1.107, maximum residual
electronic density = 0.284 e ÅϪ3
.
Crystal data for 9cЈؒ0.5H2O
Colorless parallelpiped, crystal dimensions 0.10 × 0.12 × 0.32
mm, C33H49Cl2N6O4.50Zn, M = 738.05, orthorhombic, space
group P212121, a = 14.260(2), b = 14.790(2), c = 17.453(2) Å, V =
3681.1(8) Å, Z = 4, Dc = 1.332 g cmϪ3, µ(Mo-Kα) = 0.857 mmϪ1
.
Zn(II) complexes. All work was conducted using D2O in place
of H2O. Aqueous Na2CO3 (150 µL, 0.3 M) was added to
aqueous Lؒ6HCl (150 µL, 0.1 M, L = tachen, tachpn and
tachbn) followed by 1.050 cm3 D2O and lastly Zn(O3SCF3)2
(150 µL, 0.1 M) to give [ZnL]2ϩ(aq) of nominal concentration
1 × 10Ϫ2 M. In the case of tachbn, a large amount of white
precipitate formed, which was dissolved by addition of 0.5 cm3
A total of 5293 independent reflections were used with 1.80 < θ
< 23.25Њ of which 2500 reflections had I > 2σ(I ). Final results:
R(F) = 0.084, Rw(F) = 0.133, GOF = 1.912, maximum residual
electronic density = 0.417 e ÅϪ3
.
CCDC reference numbers 221223–221225.
See http://www.rsc.org/suppdata/dt/b4/b403476g/ for crystal-
lographic data in CIF or other electronic format.
1
of d6-DMSO. H NMR spectra confirmed the formation of
[ZnL]2ϩ in all cases.
Cell culture and cytotoxicity assays
X-Ray data collection, structure solution and refinement
MBT2 is a bladder transitional cell tumor line originally
induced in C3H/HeN mice by oral ingestion of FANFT.14
It was kindly provided by Dr. Hideyuki Akaza, Tsukuba
University School of Medicine, Ibaraki, Japan. Cells were
grown in a humidified 5% CO2 atmosphere at 37 ЊC in RPMI
medium (Gibco BRL) supplemented with 10% fetal bovine
serum and penicillin/streptomycin. Hela cells were obtained
from the American Type Culture Collection and grown in a
humidified 5% CO2 atmosphere at 37 ЊC in DME medium
(Gibco BRL) supplemented with 10% fetal bovine serum and
penicillin/streptomycin. 5 × 103 MBT2 cells or 2 × 103 Hela cells
were plated in 96 well tissue culture dishes and allowed to attach
overnight before test compounds were added. Six replicate
cultures were used for each point. After 72 h, viability was
assessed using an MTT assay, in which (3-(4,5-dimethylthiazol-
2-yl)-2,5-diphenyltetrazolium bromide is added to the medium
and the formation of a reduced product is assayed by measur-
ing the optical density at 560/650 nm after 3 h. Color formation
is proportional to viable cell number.15 IC50 was calculated
using the Multiple Drug-Effect Analysis Program (Biosoft).
Single crystals of Ni(tachpn)(ClO4)2ؒMeOH (7bؒMeOH),
[Cu(tachbn)](ClO4)2 (8c) and [Zn(tachbn)]Cl(ClO4)ؒ0.5H2O
(9cЈؒ0.5H2O) were grown as described, mounted on fibers
and transferred to the goniometer. The crystals were cooled to
Ϫ100 ЊC with a stream of nitrogen gas and data were collected
on a Siemens SMART diffractometer with a CCD area detector
using graphite monochromated Mo-Kα radiation. The
SHELXTL software, version 5, was used for solutions and
refinement.10 Absorption corrections were made with
SADABS.11 Each structure was refined by full-matrix least-
squares on F 2. The twist angles α were calculated by first
determining the centroid in each of the triangles of the three
cyclohexyl-amino nitrogens ((N(tach); designated as X1) and of
the three ethyl-amino nitrogens (N(en); designated as X2).
Then, the torsion angles N(tach)–X1–X2–N(en), each involving
the two nitrogens of the same chelate arm, were calculated and
averaged. ORTEP drawings were made with SHELXTL10 and
ORTEP-3.12,13 The perchlorate anions for all three structures
were found to be disordered, however, the disorder could only
be resolved for [Ni(tachpn)](ClO4)2. In this structure alternate
positions were found for O2, O3, O4 which were refined
isotropically at an occupancy of 75/25%. The disorder in the
other two structures could not be resolved from the data sets
acquired and these anions were refined with full weighted
oxygen atoms in each position. Samples of [Cu(tachbn)](ClO4)2
diffracted weakly, making the determination of absolute
configuration less practical. This complex was refined as the
S,S,S-enantiomer based on the known configuration of tachbn
and unlikelihood of conversion to the R,R,R-form under
complexation conditions.
Results and discussion
Ligand syntheses
The hexamine chelators 1–3 were prepared by acylation of the
readily available base triamine tach4 with the p-nitrophenyl
active esters of amino acids, followed by BH3 reduction
(Scheme 1). Amino acids were chosen to exploit the exquisite
potential of substituents readily available, both in natural and
synthetic amino acids, for the stereospecific introduction of the
methyl and benzyl groups of compounds 2 and 3. The choice
of type of active ester was arbitrary and other active esters, i.e.,
N-hydroxysuccinnimidyl esters have been employed with
success in this chemistry. The choice of tert-butyloxycarbonyl
(BOC) as carbamate protecting group was based solely on its
ease of removal.16
The acylation reaction proceeded quite smoothly after
generation of the free base of the triamine in a DMF solution
and treatment with the desired amino acid ester. Cleavage of
the BOC groups of tri-amides 4–6 was quite straightforward
and either HCl–dioxane or trifluoroacetic acid proved
acceptable as required. Direct BH3 reduction of the deprotected
amides was occasionally associated with problems with
incomplete reduction, particularly so with the preparation of
1ؒ6HCl. The problem was not resolved with greater excess of
BH3, nor did it appear influenced by choice of deprotecting
agent. It was found that satisfactory yield required reduction of
the azeotropically dried free-base triamide and addition of BH3
Crystal data for 7bؒMeOH
Light purple parallelpiped, crystal dimensions 0.04 × 0.12 ×
0.28 mm, C16H40Cl2N6NiO9, M = 590.15, orthorhombic, space
group P212121, a = 9.4038(6), b = 13.7359(9), c = 20.1269(13) Å,
V = 2599.8(3) Å, Z = 4, Dc = 1.508 g cmϪ3, µ(Mo-Kα) = 1.007
mmϪ1. A total of 6063 independent reflections were used with
2.02 < θ < 27.92Њ, of which 4006 reflections had I > 2σ(I ). Final
results: R(F) = 0.056, Rw(F) = 0.091, GOF = 1.017, maximum
residual electronic density = 0.421 e ÅϪ3
.
Crystal data for 8c
Blue plate, crystal dimensions 0.40 × 0.20 × 0.15 mm,
C33H48Cl2CuN6O8, M = 791.21, orthorhombic, space group
P212121, a = 13.3903(11), b = 16.0536(12), c = 17.8755(14) Å, V =
3842.6(5) Å, Z = 4, Dc = 1.368 g cmϪ3, µ(Mo-Kα) = 0.762 mmϪ1
.
A total of 2027 independent reflections were used with 1.98 < θ
D a l t o n T r a n s . , 2 0 0 4 , 1 3 0 4 – 1 3 1 1
1308

View Article Online
in two portions. The final products were isolated from the BH3
reduction milieu by routine acid workup17 and the hexamine
products were thus isolated as the hexa-protonated species.
M() complexes of tachpyr and derivatives.3,19 Comparison
across the chelate series 1–3 suggests no significant differences
between H, Me and Bn substituents as regards ligand-field
strength of Ni() and Cu() chelation. The cyclohexyl frame-
work does increase the field strength of the hexamine relative
to the tris(ethylenediamine) coordination sphere, because the
energies of the Ni() compounds 7a,b,c are slightly above those
of [Ni(en)3]2ϩ. The reported spectrum of [Ni(tame-en)]2ϩ is
comparable to those of the Ni–tach complexes, indicating that
the chelate effect of the 1,1,1-tris(methylene)ethyl framework is
similar to that of tach.
Complexation processes and products
The complexing properties of tachen, tachpn and tachbn (1, 2,
and 3 respectively) were evaluated by determination of reaction
conditions needed to form [ML]2ϩ (M = Ni(), Cu() and
Zn()) and physical characterization of the resulting compound
series 7x, 8x and 9x (Scheme 2). To obtain the coordination
geometry, and to investigate steric and electronic effects
in complexation of M(), the 7, 8, 9 series were isolated and
characterized by X-ray crystallography (for 7b, 8c and 9cЈ), and
UV-vis or 1H NMR spectroscopy. A trial reaction showed that
the perchlorate counterion facilitated the crystallization of
complexes from nonaqueous solvent, thus, ClϪ was removed
from the salts of 1, 2 and 3 by neutralization with aqueous
NaOH followed by drying under reduced pressure and extract-
ing base-free ligands into dry alcohol. The complexes were then
prepared from equimolar amounts of the respective metal ion
salts and the ligands in alcohol. In an initial study, halide
removal was not complete and the mixed-anion complex
[Zn(tachbn)]Cl(ClO4)ؒ0.5H2O (9cЈؒ0.5H2O) was isolated and
structurally characterized.
It is also notable that the tach framework enforces hexamino
coordination on Cu() in [Cu(tachen)]2ϩ and [Cu(tachpn)]2ϩ
,
but not in [Cu(tachbn)]2ϩ
[Cu(tachbn)]2ϩ manifests a lower-energy band at 15900 cmϪ1
. The UV-vis spectrum of
,
which is consistent with five-coordinate Cu().20,21 The corre-
sponding ethylenediamine analog [Cu(en)3]2ϩ has not been
isolated to our knowledge. In sum, the solution electronic
spectra demonstrate the chelate effect of tach, the absence
of apparent effects from substitutions on the ethylene-
diamine backbone in this ligand series, and a tendency to five-
coordination of Cu().
Solid-state structure
The M–N bond lengths and N–M–N bite angles in
[Ni(tachpn)]2ϩ and [Zn(tachbn)]2ϩ are in good agreement with
the tris(ethylenediamine) complexes [Ni(en)3]2ϩ and [Zn(en)3]2ϩ
(Table 2). Bond angles within bound tachpn and tachbn are
normal and unstrained. Both complexes possess trigonally
distorted coordination geometries, as indicated by their twist
angles, α. The value of α is defined as the angle between the two
triangles of amino N’s and tach-amino N’s, as viewed down the
pseudo-C3 axis, where the octahedral geometry possesses an
angle of 60Њ and the trigonal-prismatic geometry an angle of 0Њ.
The twist angles in 7b and 9cЈ are slightly decreased relative
to the ethylenediamine complexes. Comparison of 7b to [Ni-
(tame-en)]2ϩ 22 suggests that the geometry constraints imposed
by tame are similar to tach.
The most striking feature of tachpn and tachbn is their chiral
metal-binding pocket, arising from the influence of the chiral
centers on the configuration of the N–C–C–N chelate arms.
The sterics of the Me or Bn substituent at the S-chiral center
require the arm to assume the δ configuration in the five-
membered chelate ring, hence tachpn and tachbn complexes
should have the (δδδ) configuration. The observed conform-
ations are (δδδ) in all complexes (Figs. 1–3).
The ability of 1–3 to bind Ni(), Cu() and Zn() was studied
in aqueous or mixed aqueous/nonaqueous solvent as necessary.
The chelator HCl salts were neutralized with Na2CO3 (aq), then
M() (aq) was added. At concentrations needed for analysis,
(2 × 10Ϫ3 to 1 × 10Ϫ2 M), tachen formed soluble complexes
immediately, while the tachpn and tachbn complexes formed
but precipitated from water, tachbn more so than tachpn.
Addition of MeOH, MeCN or DMSO was needed, and final
solvent compositions were aqueous pH 7.3 in tachen cases and
20 : 80 (v/v) aqueous pH 7.3/nonaqueous in tachpn and tachbn
cases. The spectra from solution complexation closely
resembled those of isolated complexes 7–9 (Table 1 and
Experimental section). The spectral analysis of the solutions
was consistent with six-coordination in all cases except
[Cu(tachbn)]2ϩ, which is discussed below. All solutions were
unchanged after several days at 20 ЊC. In sum, 1–3 bind Ni(),
Cu() and Zn() effectively in aqueous medium although the
aqueous solubility of [ML]2ϩ indicates that the presence of
more organic residues on the chelator, as for tachbn, leads to
lower aqueous solubility of complexes.
The proton NMR spectra of complexes 9 were compared to
those of the free ligands 1–3. The methylene protons Ha, Hb of
the pendant arms (N(H)C(Ha)(Hb)C(Hc)(R)NH2) are chem-
ically inequivalent in both complexes and free ligands (Scheme
1; except for 1 where R = H). However, the coupling constant
JHa–Hc differs considerably from JHb–Hc in the complexes, but not
the free ligands, as would be expected for rigid pendant arms
that would maintain gauche or trans orientations of Hc relative
to Ha and Hb (Scheme 2). This suggests that the three pendant
arms of the Zn() complexes are static with respect to ∆–Λ
interconversion on the NMR time-scale at room temperature.
Biological properties
The cytotoxicities of tachen, tachpn and tachbn toward
cultured Hela and MBT2 cancer cells were determined, with
reference to tachpyr. As shown in Table 3, the toxicity order is
tachen < tachpn < tachbn < tachpyr. The aminopyridyl binding
groups of tachpyr are substantially different from the ethylene-
diamino donors of the present compounds, so differences
in biological properties could be expected. However, the
preliminary assessment of the coordination chemistry of
tachen, tachpn and tachbn does not explain why their cyto-
toxicities differ. One possible explanation of the difference
is chelator lipophilicity; increases in lipophilicity are known
to enhance drug activity by facilitating the ability to cross the
The tris(2-pyridylmethyl)-tach complexes, [M(tachpyr)]2ϩ
,
(M = Fe(), Ga(), In()) have also been found to be static
at 25 ЊC.18,7,5 However, due to the S,S,S-chirality of tachpn
and tachbn, their metal-bound chelate arms are likely further
constrained to the Λ(δδδ) configuration in solution (see below).
cell membrane.23 The lipophilicities (logpn-octanol/pH O) of these
2
Electronic structure
chelators were calculated by the CLogP program24 as 3.41,
Ϫ1.30 and Ϫ2.22 for tachbn, tachpn and tachen, respectively.
Further, the solubility properties of the metal complexes 7–9
indicate that [M(tachbn)]2ϩ is the most lipophilic. The probable
order of decreasing lipophilicity for both chelators and metal
complexes is then tachbn > tachpn > tachen. Their cytotoxic
activity is therefore consistent with this proposal.
To examine metal-complex structure in solution and assess
effects of the tripodal framework and ethylene backbone
substituents in the tach system, the UV-vis spectra of the Ni()
and Cu() complexes were measured, assigned, and compared
with other ethylenediamine-based metal compounds (Table 1).
The spectral assignments were made as previously reported for
D a l t o n T r a n s . , 2 0 0 4 , 1 3 0 4 – 1 3 1 1
1309

View Article Online
Fig. 1 Left: ORTEP view of the complex cation of Ni(tachpn)(ClO4)2ؒMeOH (7bؒMeOH) (50% probability ellipsoids). Right: Perspective view of
one ethylenediamine arm in 7b illustrating the Λ conformation of chelate arms and δ conformation within one arm.
Table 1 Visible-near IR spectra and colors of Ni() and Cu() complexes, plus selected reference compoundsa
ν/10Ϫ3cmϪ1
Assignment
Color
Pink
[Ni(tachen)]2ϩ b
[Ni(tachpn)]2ϩ c
[Ni(tachbn)]2ϩ c
11.6 sh, 12.4
19.2
29.5
11.3 sh, 12.5
20.1
29.7
11.4 sh, 12.4
19.2
3A2g
3A2g
3A2g
3A2g
3A2g
3A2g
3A2g
3A2g
3A2g
2Eg
3T2g(F), 3A2g
1Eg(D)
1Eg(D)
1Eg(D)
3T1g(F)
3T1g(P)
3T2g(F), 3A2g
3T1g(F)
Pink
Pink
3T1g(P)
3T2g(F), 3A2g
3T1g(F)
29.9
18.0
18.1
3T1g(P)
[Cu(tachen)]2ϩ b
[Cu(tachpn)]2ϩ c
[Cu(tachbn)]2ϩ c
[Ni(en)3]2ϩ b
2T2g
Blue
Blue
Blue
Pink
2Eg
2T2g
15.9 (with high-energy shoulder)
10.6, 12.1
17.5
See text
3A2g
3T2g(F), 3A2g
3T1g(F)
1Eg(D)
1Eg(D)
3A2g
28.1
3A2g
3T1g(P)
[Ni(tame-en)]2ϩ d, e 27
(Not reported)
(Not reported)
19.1
3A2g
3T2g(F), 3A2g
Pink
3A2g
3A2g
3T1g(F)
3T1g(P)
30.7
a From solution complexation studies except [Ni(tame-en)]2ϩ; solvent system as indicated. b MOPS buffer, pH 7.3, 0.07–0.1 M. c Mixture of MOPS
buffer, 0.04 M, pH 7.3 (80% v/v) and MeCN, MeOH or DMSO (20% v/v). d H2O. e tame-en = 1,1,1-tris(N-2-aminoethylaminomethyl)ethane.
Table 2 Structural parameters of Ni(), Cu() and Zn() complexes, plus selected reference compounds
Ionic
M–N(tach)
bond distances/Å
M–N(primary amine)
bond distances/Å
Bite angle
N–M–N/Њ
Twist angle,
radius25,26
αa/Њ
[Ni(tachpn)](ClO4)2ؒMeOH (7bؒMeOH)
0.83
2.098(4)
2.113(3)
2.117(3)
2.129(3)
2.129(3)
2.125(3)
2.130(2)
2.075(10)
2.078(8)
2.331(9)
2.187(8)
2.152(7)
2.154(8)
2.193(2)
2.109(4)
2.126(5)
2.124(4)
81.91(13)
82.00(13)
81.84(13)
81.9(1)
79.8(4)
79.3(4)
83.6(4)
79.8(3)
80.1(3)
79.6(3)
80.2(1)
83.5(2)
83.0(2)
82.9(2)
48.7(7)
28
[Ni(en)3](NO3)2
0.83
0.87
52.1
46(2)
[Cu(tachbn)](ClO4)2 (8c)
2.057(8)
2.107(9)
2.245(9)
2.192(8)
2.183(8)
2.205(9)
[Zn(tachbn)]Cl(ClO4)ؒ0.5H2O (9cЈؒ0.5H2O)
0.88
44(1)
29
[Zn(en)3](NO3)2
0.88
0.83
49.8
51.5
b 27
[Ni(tame-en)]Br2
2.112(4)
2.099(4)
2.113(4)
a See text; average of three values, where individual values differ from the average by less than 1.5Њ. b tame-en = 1,1,1-tris(N-2-amino-
ethylaminomethyl)ethane.
complexes formed are relatively strain-free but their conform-
ations are constrained by stereospecific substitution of the
Conclusion
This work demonstrates that tripodal chelators with pendant
arms based on ethylenediamine bind Ni(), Cu() and Zn() in
the aqueous or mixed aqueous/nonaqueous environment. The
arms. The aqueous coordination chemistry of these substances
with other biometal ions is under continuing study, as is their
cellular transition-metal ion speciation.
D a l t o n T r a n s . , 2 0 0 4 , 1 3 0 4 – 1 3 1 1
1310

View Article Online
Table 3 Growth inhibitory concentrations of tach chelators in Hela
Sciences, Office of Research, US Department of Energy (Grant
DE-FG02-96ER14673).
and MBT2 cellsa
Chelator
IC50 (Hela cells)/µM
IC50 (MBT2 cells)/µM
References
tachpyr
tachbn
tachpn
tachen
5
8
28
6
10
55
1 Uses of Inorganic Chemistry in Medicine; ed. N. P. Farrell, Royal
Society of Chemistry: Cambridge, UK, 1999.
2 Iron Chelators: New Development Strategies, ed. D. Badman,
R. J. Bergeron and G. M. Brittenham, Saratoga Group, Ponte Vedra
Beach, FL, 2000.
200
not measd
a Data were obtained from six replicate cultures.
3 G. Park, A. M. Przyborowska, N. Ye, N. Tsoupas, C. B. Bauer,
G. A. Broker, R. D. Rogers, M. W. Brechbiel and R. P. Planalp,
Dalton Trans., 2003, 318–324.
4 T. Bowen, R. P. Planalp and M. W. Brechbiel, Bioorg. Med. Chem.
Lett., 1996, 6, 807–810.
5 G. Park, N. Ye, R. Rogers, M. Brechbiel and R. Planalp, Polyhedron,
2000, 19, 1155–1161.
6 G. Park, E. Dadachova, A. Przyborowska, S.-J. Lai, D. Ma,
G. Broker, R. Rogers, R. Planalp and M. Brechbiel, Polyhedron,
2001, 20, 3155–3163.
7 G. Park, F. Lu, N. Ye, M. Brechbiel, S. Torti, F. Torti and R. Planalp,
J. JBIC, Biol. Inorg. Chem., 1998, 3, 449–457.
8 R. D. Abeysinghe, B. T. Greene, R. Haynes, M. C. Willingham,
J. Turner, R. P. Planalp, M. W. Brechbiel, F. M. Torti and S. V. Torti,
Carcinogenesis, 2001, 22, 1607–1614.
9 B. T. Greene, J. Thorburn, M. C. Willingham, A. Thorburn,
R. P. Planalp, M. W. Brechbiel, J. Jennings-Gee, J. Wilkinson IV,
F. M. Torti and S. V. Torti, J. Biol. Chem., 2002, 277, 25568–25575.
10 G. M. Sheldrick, SHELXTL, version 5.05, Siemens Analytical
X-Ray Instruments Inc., 1996.
11 G. M. Sheldrick, Program for Semiempirical Absorption Correction
of Area Detector Data, University of Göttingen, Germany, 1996.
12 M. Burkitt, Methods Enzymol., 1994, 234, 66–79.
13 L. J. Farrugia, J. Appl. Crystallogr., 1997, 30, 565.
14 M. S. Soloway, J. B. deKernion, D. Rose and L. Persky, Surg. Forum.,
1973, 24, 542–544.
Fig. 2 ORTEP view of the complex cation of [Cu(tachbn)](ClO4)2 (8c)
(50% probability ellipsoids).
15 T. Mosmann, J. Immunol. Methods, 1983, 65, 55–63.
16 T. Greene and P. Wuts, Protective Groups in Organic Synthesis,
Wiley, New York, 2nd edn., 1991, p. 327.
17 M. W. Brechbiel, O. A. Gansow, R. W. Atcher, J. Schlom, J. Esteban,
D. E. Simpson and D. Colcher, Inorg. Chem., 1986, 25, 2772–2781.
18 K. A. Hilfiker, R. D. Rogers, M. W. Brechbiel and R. P. Planalp,
Inorg. Chem., 1997, 36, 4600–4603.
19 R. D. Hancock and G. J. McDougall, J. Chem. Soc., Dalton Trans.,
1977, 67–70.
20 G. Park, J. Shao, F. H. Lu, R. D. Rogers, N. D. Chasteen,
M. W. Brechbiel and R. P. Planalp, Inorg. Chem., 2001, 40,
4167–4175.
21 E. D. McKenzie, J. Chem. Soc. A, 1970, 3095–3099.
22 A. McAuley, S. Subramanian and T. W. Whitcombe, J. Chem. Soc.,
Dalton Trans., 1993, 2209–2214.
23 A. Albert, Selective Toxicity: The Physicochemical Basis of Therapy,
Chapman and Hall, London, 7th edn., 1985.
24 A. Leo, C. Hansch and D. Elkins, Chem. Rev., 1971, 71, 525–616.
25 R. D. Shannon and C. T. Prewitt, Acta Crystallogr., Sect. B, 1969,
25, 925–946.
Fig. 3 ORTEP view of the complex cation of [Zn(tachbn)]Cl(ClO4)ؒ
0.5H2O (9cЈؒ0.5H2O) (50% probability ellipsoids).
26 R. Shannon, Acta Crystallogr., Sect. A, 1976, 32, 751–767.
27 A. McAuley, S. Subramanian and T. W. Whitcombe, J. Chem. Soc.,
Dalton Trans., 1993, 2209–2214.
28 J. D. Korp, I. Bernal, R. A. Palmer and J. C. Robinson, Acta
Crystallogr., Sect. B, 1980, 36, 560–564.
29 D. Neill, M. J. Riley and C. H. L. Kennard, Acta Crystallogr., Sect.
C, 1997, 53, 701–703.
Acknowledgements
The NIH is thanked for support (grant no. R01-DK57781 to
S. V. T.). Research at The University of Alabama is supported
by the Division of Chemical Sciences, Office of Basic Energy
D a l t o n T r a n s . , 2 0 0 4 , 1 3 0 4 – 1 3 1 1
1311