Lipophilic phosphonium-lanthanide compounds with magnetic, luminescent, and tumor targeting properties

Article abstract of DOI:10.1016/j.jinorgbio.2011.09.035

Multifunctional phosphonium-lanthanide compounds that simultaneously possess paramagnetism, luminescence, and tumor mitochondrial targeting properties were prepared by use of a facile method. These compounds were fully characterized by use of ¹H, ¹³C, ³¹P NMR, FT-IR, and elemental analyses. The thermal properties of these compounds including melting points and decomposition temperatures were investigated using DSC and TGA analyses. In addition, the paramagnetism, luminescence, and tumor targeting properties of these multifunctional compounds were confirmed by respective use of SQUID, fluorescence, and cell cytotoxicity studies. All compounds exhibited paramagnetism at room temperature, which could provide target delivery of these compounds to parts of the body containing tumor cells using a strong external magnetic field. In addition, these compounds display two major characteristic emissions originating from Dy^{3 +}, which can be utilized for imaging tumor cells. The IC₅₀ values of these compounds measured against normal breast cell line (Hs578Bst) are significantly greater than those measured against the corresponding carcinoma breast cell line (Hs578T), clearly indicating the selective tumor targeting properties of these compounds. Confocal fluorescence microscopy studies were used to confirm the yellowish-green fluorescence corresponding to the emission of dysprosium thiocyanate anion within cancer cells upon exposure of cancer cell lines such as human pancreatic carcinoma cell line (MIAPaCa-2) and human breast carcinoma (MDA-MB-231) to a solution of these phosphonium-dysprosium compounds.

Full text of DOI:10.1016/j.jinorgbio.2011.09.035

Journal of Inorganic Biochemistry 107 (2012) 40–46
Contents lists available at SciVerse ScienceDirect
Journal of Inorganic Biochemistry
journal homepage: www.elsevier.com/locate/jinorgbio
Lipophilic phosphonium–lanthanide compounds with magnetic, luminescent, and
tumor targeting properties
Min Li a, Gabriela M. Ganea a, Chengfei Lu a, Sergio L. De Rooy a, Bilal El-Zahab a, Vivian E. Fernand a,
Rongying Jin b, Sita Aggarwal c, Isiah M. Warner a,
⁎
a
Department of Chemistry, Louisiana State University, Baton Rouge, LA 70803, USA
b
Department of Physics & Astronomy, Louisiana State University, Baton Rouge, LA 70803, USA
William Hansel Cancer Prevention, Pennington Biomedical Research Center, Baton Rouge, LA 70808, USA
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 22 July 2011
Received in revised form 22 September 2011
Accepted 29 September 2011
Available online 3 November 2011
Multifunctional phosphonium–lanthanide compounds that simultaneously possess paramagnetism, lumines-
cence, and tumor mitochondrial targeting properties were prepared by use of a facile method. These com-
pounds were fully characterized by use of 1H, 13C, 31P NMR, FT-IR, and elemental analyses. The thermal
properties of these compounds including melting points and decomposition temperatures were investigated
using DSC and TGA analyses. In addition, the paramagnetism, luminescence, and tumor targeting properties
of these multifunctional compounds were confirmed by respective use of SQUID, fluorescence, and cell cyto-
toxicity studies. All compounds exhibited paramagnetism at room temperature, which could provide target
delivery of these compounds to parts of the body containing tumor cells using a strong external magnetic
Keywords:
Lanthanides
Luminescence
Magnetic
field. In addition, these compounds display two major characteristic emissions originating from Dy3+
,
Phosphonium
Target delivery
which can be utilized for imaging tumor cells. The IC50 values of these compounds measured against normal
breast cell line (Hs578Bst) are significantly greater than those measured against the corresponding carcino-
ma breast cell line (Hs578T), clearly indicating the selective tumor targeting properties of these compounds.
Confocal fluorescence microscopy studies were used to confirm the yellowish-green fluorescence corre-
sponding to the emission of dysprosium thiocyanate anion within cancer cells upon exposure of cancer cell
lines such as human pancreatic carcinoma cell line (MIAPaCa-2) and human breast carcinoma (MDA-MB-
231) to a solution of these phosphonium–dysprosium compounds.
© 2011 Elsevier Inc. All rights reserved.
1. Introduction
inhibition of this process in tumor cells has provided an effective
strategy for treatment of cancer. Numerous in vitro and in vivo stud-
The aim of drug targeting is to deliver therapeutic compounds to a
particular tissue in the body, a strategy which has found tremendous
significance in the field of cancer therapy. A targeted delivery ap-
proach may overcome major delivery limitations such as low bio-
availability, significant toxicity, undesirable side effects, and
unrestricted biodistribution of anticancer drugs [1,2]. As one class of
major cellular targeting agents, delocalized lipophilic cations (DLCs)
accumulate more rapidly in the mitochondria of most carcinoma-
derived cells than in most untransformed cells in response to nega-
tive inside transmembrane potentials [3–5]. This difference has been
attributed to the abnormally high mitochondrial membrane poten-
tials (ΔΨm) characteristic of most tumor cells [6–10]. Accumulation
of DLCs in mitochondria may inhibit oxidative phosphorylation, an
energy-generating process, and thus disrupt the cellular ATP produc-
tion required to sustain normal cellular functions [3]. The selective
ies have demonstrated the selective cytotoxicity of DLCs against
tumor cells [11–14].
One class of lipophilic compounds based on phosphonium is called
cationic lipophilic phosphonium salts (CLPS). These compounds have
shown much higher carcinoma selectivity in vitro than clinically used
anti-tumor agents such as cisplatin and cytosine arabinoside (ara-C)
[15]. Tetraphenylphosphonium chloride (TPP) is the prototype com-
pound of CLPS. Radiolabeled TPP has been used as a molecular
probe for tumor imaging because TPP can selectively accumulate in
the mitochondrial membrane of tumor cells [16]. TPP-conjugated an-
tioxidants have also been developed for their targeted delivery to mi-
tochondria which could prevent mitochondrial oxidative damage far
more effectively than untagged antioxidants [17,18]. Interestingly,
the counteranions of all these lipophilic salts reported to date have
been exclusively limited to halides, e.g., chloride and bromide. These
halides seem to be simply used as counteranions in order to simply
balance the positive electric charge of DLCs, without serving any
other useful function. For example, anions were even omitted in the
illustrations of lipophilic salt molecular structures in a number of
⁎
Corresponding author. Tel.: +1 225 578 2829; fax: +1 225 578 3971.
E-mail address: iwarner@lsu.edu (I.M. Warner).
0162-0134/$ – see front matter © 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jinorgbio.2011.09.035

M. Li et al. / Journal of Inorganic Biochemistry 107 (2012) 40–46
41
articles [19–21]. In our view, however, a functional anion introduced
by a simple anion-exchange of the original halide anion may afford
additional useful properties. This strategy has been employed by us
to prepare magnetic chiral room temperature ionic liquids by repla-
cing the Cl− of an amino acid methyl ester hydrochloride with a para-
magnetic anionic species such as FeCl4−[22].
decoupling. FT-IR was measured using a Bruker Tensor 27 FT-IR spec-
trometer. Samples were analyzed in pure form by use of a DuraSamp
IR apparatus. All spectra were obtained using 32 scans for both sam-
ple and background with a resolution of 4 cm−1. Elemental analyses
were contracted to Atlantic Microlab (Atlanta, GA, USA).
A TA Q50 thermal gravimetric analyzer (TA Instruments, New
Castle, DE) was used to analyze the thermal stability of phospho-
nium–dysprosium salts and their decomposition behaviors. Sam-
ples (about 3–4 mg) were scanned from room temperature to
500 °C under nitrogen flow (50 mL/min) with a heating rate of
10 °C/min. Values for the onset degradation temperature and
peak temperature were determined from the derivative TGA
curves. Melting points of the products were investigated by use
of a Q100 differential scanning calorimeter (TA Instruments, New
Castle, DE). The phosphonium–dysprosium salts (about 5 mg)
were put in a sealed aluminum crucible and run under a nitrogen
flow (50 mL/min). Samples were cooled and heated with the fol-
lowing sequence: isothermal for 3 min at 20 °C; then heated to
120 °C with a rate of 5 °C/min, isothermal for 3 min at 120 °C;
then cooled to 20 °C with a rate of 5 °C/min.
Ionic liquids (ILs) have been studied extensively in a great va-
riety of fields owing to their unique and beneficial properties
[23–26]. One of the most desirable properties of ILs is their high
tunability. By simply replacing the halide anions of organic salts
with functional anionic species, one can generate numerous
novel multifunctional ionic liquid materials with many attractive
properties [27–32]. Consequently, within a single molecule both
the cation and anion may possess distinct and desirable properties
and functions, which in turn could be used to prepare multifunc-
tional materials more conveniently. Using a similar rationale pre-
viously employed for synthesizing functional ILs, we introduce
multifunctional anions based on a lanthanide (dysprosium) com-
plex to exchange the bromide anion of aryltriphenylphosphonium
salts. The dysprosium–thiocyanate anion is intrinsically paramag-
netic and luminescent, which provides great utility for applica-
tions such as targeted delivery in the presence of a magnetic
field and fluorescence imaging of tumor cells. Simultaneously,
the original lipophilic phosphonium cation can serve as a mito-
chondrial targeting agent, selective for tumor cells.
Magnetic properties were measured using approximately 80 to
100 mg of the sample in a Quantum Design Superconducting Quantum
Interference Device (SQUID) magnetometer (San Diego, CA, USA) at
temperatures between 2 and 300 K and fields between −70 000 and
70 000 Oe. Absorbance measurements were performed on
a
We herein report on the synthesis and characterization of
novel multifunctional phosphonium–lanthanide compounds that
simultaneously contain paramagnetism, luminescence, and tumor
mitochondrial targeting properties within a single compound. In
addition, we report the application of these compounds as
tumor-targeting, therapeutic, and imaging agents in human cell
cultures.
Shimadzu UV-3101PC UV–vis-near-IR scanning spectrometer
(Shimadzu, Columbia, MD). Steady-state fluorescence measurements
were recorded at room temperature by use of a Spex Fluorolog-3 spec-
trofluorimeter (model FL3-22TAU3; Jobin Yvon, Edison, NJ) equipped
with a 450-W xenon lamp and R928P photomultiplier tube (PMT)
emission detector. Fluorescence emission spectra were collected in a
4-mm quartz fluorescence cuvet with slit widths set for entrance exit
bandwidths of 4 nm on both excitation and emission monochromators.
2. Experimental
2.1. Chemicals and materials
2.3. Synthesis and characterization of the phosphonium–lanthanide
compounds
Phosphonium bromides including tetraphenylphosphonium
bromide (98.0%), benzyltriphenylphosphonium bromide (97.0%),
(4-ethoxybenzyl)triphenylphosphonium bromide (98.0%), (4-
A facile two-step synthesis procedure was used to prepare the
phosphonium–lanthanide compounds. In the first step, the four
intermediate compounds ([Ph4P][SCN], [Ph3PBnOEt][SCN], [Ph3-
PBnNO2][SCN], and [Ph3PBn][SCN]) were synthesized by reacting
the respective phosphonium bromide (1 equiv.) with KSCN
(2 equiv.) in acetonitrile at room temperature for 2 d through
anion-exchange reactions [33]. Dy(ClO4)3.6H2O was obtained by
dissolving Dy2O3 in 70% HClO4 aqueous solution followed by sub-
sequent removal of water by lyophilization. In the second step, the
four phosphonium–dysprosium products ([Ph4P]5[Dy(SCN)8],
[Ph3PBnOEt]5[Dy(SCN)8], [Ph3PBnNO2]5[Dy(SCN)8], and [Ph3PBn]5
[Dy(SCN)8]) were synthesized according to the procedure
reported previously with slight modifications [27,28]. A typical
nitrobenzyl)triphenylphosphonium
bromide
(98.0%)
were
obtained from TCI America (Portland, Oregon, USA). Potassium
thiocyanate, dysprosium(III) oxide, and perchloric acid (70%)
were obtained from Sigma-Aldrich (Milwaukee, Wisconsin, USA)
and used without further purification. Ethanol, methanol, acetoni-
trile, chloroform, and carbon tetrachloride were of anhydrous
grade (Sigma-Aldrich, Milwaukee, WI), and all other solvents
such as acetone, hexane, and water are of HPLC grade (J. T. Baker,
Phillipsburg, NJ).
Normal human breast fibroblast cells (Hs578Bst, ATCC no. HTB-
125), human breast carcinoma cells (Hs578T, ATCC no. HTB-126),
hormone-independent human breast carcinoma cells (MDA-MB-
231, ATCC no. HTB-26), human pancreatic carcinoma cells (PANC-1,
ATCC no. CRL-1469), and human colorectal adenocarcinoma cells
(HT-29, ATCC no. HTB-38) were obtained from the American Tissue
Culture Collection (ATCC, Manassas, VA) and grown to 90% conflu-
ence according to ATCC's instructions.
synthesis procedure is as follows.
A mixture of [Ph4P][SCN]
(5 equiv.), KSCN (3 equiv.), and Dy(ClO4)3.6H2O (1 equiv.) was
stirred in anhydrous ethanol at room temperature overnight. The
white suspension of KClO4 byproduct was removed by filtration
and ethanol was evaporated by use of a rotavapor. The residue
was redissolved in anhydrous dichloromethylene and the solution
was allowed to stand overnight in a refrigerator. After filtration
dichloromethylene was evaporated under vacuum to afford the
products which were further dried under vacuum for 48 h. Cau-
tion: Although we did not experience any problems in this work,
special care should be taken when perchlorates are handled due to
2.2. Instrumentation and methods
1H NMR (400 MHz) and 13C NMR (100 MHz) spectra were ac-
quired by use of a Bruker Avance 400 NMR spectrometer. 31P NMR
(101 MHz) spectra were taken with a 250 Brucker Avance spectrom-
eter. The 1H and 13C chemical shifts are given in parts per million (δ)
with TMS as an internal standard. The chemical shifts of 31P are
recorded relative to external 85% H3PO4 with broad-band 1H
a
potential hazard of explosion. Characterization of the four
phosphonium–dysprosium compounds by use of 1H, 13C, and 31P
NMR, FT-IR, and C, H, N, S elemental analysis is provided in
Supporting information.

42
M. Li et al. / Journal of Inorganic Biochemistry 107 (2012) 40–46
2.4. Cell viability assay
transmitted light images. Images were acquired by collecting emitted
light at wavelengths from 512 to 663 nm.
The cytotoxicities of the compounds against various cell lines were
determined by use of CellTiter 96® AQueous One Solution Cell Prolifera-
tion Assay (Promega Corporation, Madison, Wisconsin, USA) according
to the manufacturer's instructions. Specifically, the cells (5×103 cells/
well, 100 μL) were incubated with solutions of compounds dissolved in
appropriate culture media (containing 2% DMSO). The cells were incu-
bated with various concentrations of the four compounds at 37 °C,
in 5% CO2 atmosphere for 48 h. At the end of the incubation period,
the cells were treated with 3-(4,5-di-methylthiazol-2-yl)-5-(3-
carboxymethoxyphenyl)-2-(4-sulfo-phenyl)-2H-tetrazolium salt (MTS)
(20 μL). After 1 h, the absorbance was measured at 490 nm using a
microplate spectrophotometer (Benchmark Plus, Bio-Rad Laboratories,
Hercules, CA, USA). Percentage cell viability was calculated as 100
times the ratio of absorbance in the presence of investigated compounds
and the absorbance for treatment-free control samples. The concentra-
tion at which the growth of 50% of cells was inhibited (IC50) was calculat-
ed using the linear fit equation of the linear part of the cytotoxicity
graphs (cell survival versus concentration) for each compound. In the
case of compounds for which the linear portion of the graph did not in-
clude 50% cell growth inhibition, the IC50 was reported as higher than
the minimum concentration at which a constant cytotoxicity was
reached irrespective of the increase in concentration. The values of IC50
were used for comparison of the cytotoxicities among the synthesized
compounds.
3. Results and discussion
3.1. Synthesis and characterization
Four phosphonium–dysprosium salts were synthesized by follow-
ing a previously reported method [27,28] with slight modifications
(Scheme 1). These compounds were characterized by use of 1H, 13C,
31P NMR, FT-IR, and elemental analysis. Their thermal properties, in-
cluding melting points and thermal stabilities, were also investigated
by use of differential scanning calorimetry (DSC) and thermogravi-
metric analysis (TGA) (Supporting information). These phosphoni-
um–dysprosium compounds exhibit low solubility in water (Table
S1, Supporting information) due to their hydrophobic triphenylaryl-
phosphonium cations. In contrast, as a control compound, K5Dy
(SCN)8 containing a different potassium cation, but the same dyspro-
sium thiocyanate complex anion shows much higher water solubility
than the phosphonium-containing compounds. This comparison
highlights the significant influence of the cation on the solubility of
these compounds. Additional resistivity tests of the aqueous solution
containing [Ph4P]5[Dy(SCN)8] and the control compound K5[Dy
(SCN)8] demonstrated that the saturated aqueous solutions of
[Ph4P]5[Dy(SCN)8] had higher resistivity (14.58 versus 8.38 Ω). This
comparison of resistivity indicates that the compounds are stable
and primarily exist in a non-dissociated state in aqueous solution.
All experiments were performed in triplicate. Statistical analysis
was conducted using one-way analysis of variance (ANOVA). Tukey's
studentized range test was performed to ascertain the significant dif-
ference between treatments within the 95% confidence interval using
SAS 9.2 software (SAS, Cary, NC, USA).
3.2. Paramagnetic properties of the phosphonium lanthanide compounds
All the four compounds contain a magnetically active ion Dy3+
with a 4f9 electron configuration and exhibit paramagnetic behavior
at room temperature. Their magnetic susceptibilities were measured
using a superconducting quantum interference device (SQUID) mag-
netometer. The magnetization at 300 K was measured in a magnetic
field ranging from −70 000 to 70 000 Oe and exhibited an expected
linear dependence on the applied magnetic field according to Curie's
law (Fig. 1a) [29–32]. From the slopes of the fitted lines, the respec-
tive molar magnetic susceptibilities of the four phosphonium–dypro-
sium compounds at 300 K were obtained as χmol =0.0453, 0.0400,
0.0385, and 0.0353 cm3 mol−1, respectively, which fit well to the
2.5. Tumor cell uptake and imaging study by confocal fluorescence
microscopy
Human pancreatic carcinoma cell lines (MIAPaCa-2) (1×105 cells/
well) were incubated with solutions of the phosphonium–lanthanide
compounds (10 μM) for 1 h at 37 °C in 5% CO2 atmosphere. The cell
medium was subsequently removed. Treated cells were then washed
with prewarmed phosphate buffered saline (PBS) to remove excess
fluorescent probes and possible interferences of the medium. As con-
trol, non-treated cells containing no phosphonium–lanthanide com-
pound were incubated and imaged using the same protocol as for
the treated cells. Confocal fluorescence microscopy images of live
cells were acquired using a confocal laser microscope (Zeiss Confocal
LSM510, Carl Zeiss MicroImaging Inc.Thornwood, NY, USA) equipped
with an argon-krypton laser Blue Stain. The sample was illuminated
with a 458 nm Argon/Krypton laser for both the fluorescent and
expected value for Dy3+[27]. The effective magnetic moment (μeff
)
for Dy3+ has been calculated as 10.48 μB[27]. The measured μeff values
of the four compounds were close to the calculate→d value (Table 1).
All four compounds showed very weak anti-ferromagnetic interac-
tions as observed from their Weiss constants (Table 1). The increase
in χmol with decreasing temperature in the plot of magnetic suscepti-
bility against temperature also proves that [Ph3PBnNO2]5[Dy(SCN)8]
Scheme 1. Synthesis of multifunctional phosphonium–dysprosium compounds.

M. Li et al. / Journal of Inorganic Biochemistry 107 (2012) 40–46
43
b
a
4
3000
2000
1000
0
3
2
1
0
-1000
-2000
-3000
0
50 100 150 200 250 300
-6 -4 -2
0
2
4
6
H / 104 Oe
T / K
c
d
20
15
10
5
14
12
10
8
0
6
0
50 100 150 200 250 300
T / K
0
50 100 150 200 250 300
T / K
Fig. 1. Magnetic property of [Ph3PBnNO2]5[Dy(SCN)8]. Field dependence of a) the molar magnetization Mmol at 300 K. Temperature dependence of b) the static molar magnetic sus-
ceptibility χmol, c) the reciprocal molar magnetic susceptibility χm−o1l, and d) the product of temperature and static molar magnetic susceptibility χmolT at a field of 1000 Oe.
is paramagnetic (Fig. 1b). Due to the large anisotropic magnetic mo-
ment of Dy3+, all four compounds can be manipulated by a neodym-
ium magnet when the distance between the magnet and compound
is less than about 4 mm (Fig. 2). Therefore, the phosphonium–
dysprosium compounds could possibly be target delivered under
a strong external magnetic field and concentrated in parts of the
body containing tumor cells. The field and temperature depen-
dences of the other three phosphonium–dysprosium salts are
listed in the Supporting information.
yellowish-green luminescence. It is well known that lanthanide ions
have very low absorption coefficients because optical transitions
within the 4f subshells of lanthanide ions are partially forbidden
[35]. However, energy transfer from an organic antenna chromophore
may generate indirect excitation, which allows for excitation at a
wavelength where the lanthanide ion does not exhibit significant ab-
sorption [36]. In addition, lanthanide ions are usually good quenchers
of triplet states so that photobleaching is substantially reduced [36].
For our compounds, the aryltriphenylphosphonium cations may
have served as sensitizers (antenna) and are able to activate Dy3+
via energy transfer. This bright fluorescence originating from Dy3+
3.3. Luminescence properties of the phosphonium lanthanide compounds
Luminescence properties of the four compounds were studied as
well. Dy3+ with a 4f9 electron configuration is known for its intense
luminescent transition, 4F9/2 → 6H15/2 in the blue region of visible
light spectrum and 4F9/2 → 6H13/2 in the yellowish-green spectral re-
gion [34]. All four compounds showed the two major characteristic
emissions (Fig. 3). The intensity of the hypersensitive transition var-
ies with the environment. The transition 4F9/2 → 6H13/2 at 573 nm is
the most intense for all the compounds in agreement with their
Table 1
Summary of magnetic properties of the compounds.
a
b
Compounds
χmol
μeff
θc
(K)
(cm3 mol−1
)
(μB)
[Dy(SCN-)8
]
P
[Ph3PBnNO2]5[Dy(SCN)8]
[Ph3PBnOEt]5[Dy(SCN)8]
[Ph3PBn]5[Dy(SCN)8]
[Ph4P]5[Dy(SCN)8]
4.53×10−2
4.00×10−2
3.85×10−2
3.53×10−2
10.4
10.2
9.9
−4.3
−4.4
−3.7
−3.1
5
9.6
C
a
Molar magnetic susceptibility measured at 300 K
Effective magnetic moment.
b
c
Fig. 2. Response of the pale yellow-colored [Ph4P]5[Dy(SCN)8] to
a neodymium
Weiss constant.
magnet.

44
M. Li et al. / Journal of Inorganic Biochemistry 107 (2012) 40–46
than 200 μM (Fig. 5a). Our results also show that there was a signifi-
cant difference between the viability of Hs578T and Hs578Bst cells
when treated with [Ph4P]5[Dy(SCN)8] at concentrations between 25
and 200 μM. This comparison demonstrates that [Ph4P]5[Dy(SCN)8]
selectively induced cytotoxicity against tumor cells versus normal
cells under the same cell culture conditions. The same trend was
also observed when [Ph3PBnOEt]5[Dy(SCN)8] was incubated with
both the normal and cancer cell lines for 48 h. For example, IC50 of
[Ph3PBnOEt]5[Dy(SCN)8] against Hs578T was 19 μM, while a much
higher IC50 of 74 μM against the normal cell line Hs578Bst was ob-
served (Fig. 5b). Our data also indicate that there was at all concentra-
tions of [Ph3PBnOEt]5[Dy(SCN)8] a significant difference in viability
between tumor and normal cells, except at 15 μM. Similar results
were obtained when cancer and normal cells were treated for 48 h
with [Ph3PBnNO2]5[Dy(SCN)8] (Fig. 5c), and [Ph3PBn]5[Dy(SCN)8]
(Fig. 5d). A significant difference was found between cancer and nor-
mal cells for treatment with [Ph3PBnNO2]5[Dy(SCN)8] (5–200 μM)
and [Ph3PBn]5[Dy(SCN)8] (5–100 μM). The IC50 values of these com-
pounds for Hs578T and Hs578Bst are displayed in Table 2. Altogether,
these results demonstrate the selective cytotoxicity of the lipophilic
phosphonium–dysprosium salts.
4F9/2
6H13/2
4F9/2
6H15/2
400
450
500
λ / nm
550
600
Fig. 3. Emission spectra with transition assignments of [Ph4P]5[Dy(SCN)8] (red), [Ph3-
PBnOEt]5[Dy(SCN)8] (green), [Ph3PBnNO2]5[Dy(SCN)8] (black), and [Ph3PBn]5[Dy
(SCN)8] (blue) in acetonitrile (1 mM), λex =365 nm.
was subsequently tested for application as an imaging agent for
tumor cells. The excitation spectra obtained by monitoring the emis-
sions at 480 nm and absorption spectra of the four compounds are
shown in the Supporting information.
3.4. Selective cytotoxicities of the phosphonium lanthanide compounds
against tumor cell lines
In contrast, most clinical antineoplastic agents such as cisplatin
and ara-C do not show appreciable selectivity against tumor-derived
versus untransformed cells if the cells are actively proliferating in cul-
ture [12]. The anti-carcinoma selectivity of aryltriphenylphospho-
nium salts has been postulatively ascribed to their different
diffusion rates from the extracellular environment into the mitochon-
dria and cytoplasm of tumor cells [15]. Among the four salts investi-
gated, [Ph3PBnOEt]5[Dy(SCN)8] with an electron donating group
(−OCH2CH3) exhibited the highest anticancer activities. Rideout et
al. have also reported significantly lower IC50 value of a modified aryl-
triphenylphosphonium chloride that contains an electron donating
group (−N(CH3)2) [15]. IC50 values against other tumor cell lines in-
cluding hormone-independent human breast carcinoma cells (MDA-
MB-231), human pancreatic carcinoma cells (PANC-1), and human
colorectal adenocarcinoma cells (HT-29) were also investigated
(Table 2). The IC50 values of [Ph3PBnOEt]5[Dy(SCN)8] against HT-29
and MDA-MB-231 were 86 and 28 μM, respectively. The IC50 values
of all four compounds against PANC-1 were greater than the concen-
trations investigated, which suggests that the four compounds have
low anticancer activity against the PANC-1 cells after 48 h incubation.
These new phosphonium–dysprosium compounds were also in-
vestigated for their cytotoxicities against four cancer cell lines and
one normal breast cell line. The concentrations (IC50) at which the
growth of 50% of cells was inhibited were calculated and used for
comparison. In the control tests, the starting material [Ph4P][Br], a
prototype CLPS compound, displayed very low cytotoxicity against
all cancer cell lines studied. The IC50 values of [Ph4P][Br] were greater
than the highest concentration investigated (400 μM) (Fig. 4). In con-
trast, the anion-exchange of Br− with the dysprosium–thiocyanate
anion resulted in lower IC50 of the four products. It is possible that
the increased cytotoxicities of the new compounds are due to a
higher amount of active phosphonium per mole of phosphonium–
dysprosium compounds as compared with the bromide salt. In addi-
tion, therapeutic applications including anticancer agents of lantha-
nides have been reported [37–40]. Previous studies at the cellular
level and animal tests indicated that a high dose of Ln3+ can induce
apoptosis of cancer cells a result of the damaging effect of cytotoxic
behavior.[37] Therefore, it is also possible that the increased cytotox-
icities against cancer cell lines were obtained due to the replacement
of bromide by the dysprosium-containing anion.
3.5. Targeted delivery of the phosphonium lanthanide compounds to
tumor cells
To investigate the selective cytotoxicity of the four compounds
against tumor cells, a cell viability study was also performed with
the normal breast cell line Hs578Bst as control. The IC50 of [Ph4P]5
[Dy(SCN)8] against breast cancer cell line Hs578T was 83 μM. Con-
versely, IC50 against normal breast cell line Hs578Bst was greater
Confocal fluorescence microscopy has been shown to be effective
in observing cellular uptake, intracellular distribution, and localiza-
tion of fluorescent drugs [41–43]. This technique has also been used
in imaging tumor cells [44–48]. In order to demonstrate the potential
of these lipophilic phosphonium–dysprosium compounds as contrast
agents for imaging tumor cells, we examined the cellular uptake and
fluorescence images of [Ph4P]5[Dy(SCN)8] using the human pancreat-
ic carcinoma cell line (MIAPaCa-2). In this study, the cell line
MIAPaCa-2 was used as a model cell of tumor cells. The potential ap-
plication of these phosphonium–dysprosium compounds as tumor
cell imaging markers was confirmed using a confocal fluorescence
microscope. We observed that upon exposure of MIAPaCa-2 cells to
a solution of [Ph4P]5[Dy(SCN)8] for 1 h at 37 °C, the phosphonium–
dysprosium compound was internalized. The yellowish-green fluo-
rescence light in Fig. 4c corresponds to the emission of Dy3+ inside
MIAPaCa-2 cells. The observed fluorescence image is attributed to
the significant uptake of [Ph4P]5 [Dy(SCN)8] in the tumor cells
through highly selective and efficient interactions between phospho-
nium cations and the mitochondria of the tumor cell [3,4]. In the con-
trol experiment, however, no fluorescence emission was observed
using the same cell line MIAPaCa-2, but without addition of
150
120
*
90
*
*
*
*
*
*
60
30
0
Hs578T
HT29
MDA-MB-231
PANC1
0
50 100 150 200 250 300 350 400
Concentration of [Ph4P][Br] (μM)
Fig. 4. Cytotoxicity of [Ph4P][Br] against cancer cell lines, 48 h incubation. All points are
mean s.d. of triplicate wells of three independent experiments. Data is considered
significant when pb0.05.

M. Li et al. / Journal of Inorganic Biochemistry 107 (2012) 40–46
45
b
a
Hs578Bst (normal breast cell)
Hs578Bst (normal breast cell)
*
125
100
75
50
25
0
Hs578T (breast cancer cell)
Hs578T (breast cancer cell)
100
75
50
25
0
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
0
50
100
150
200
0
25
50
75
100
Concentration of [Ph4P]5[Dy(SCN)8] (μM)
Concentration of [Ph3PBnOEt]5[Dy(SCN)8] (μM)
c
d
Hs578Bst (normal breast cell)
Hs578T (breast cancer cell)
*
*
Hs578Bst (normal breast cell)
Hs578T (breast cancer cell)
125
125
100
75
50
25
0
*
*
*
*
*
100
*
*
*
75
50
25
0
*
*
*
*
*
*
*
*
0
50
100
150
200
0
25
50
75
100
Concentration of [Ph3PBnNO2]5[Dy(SCN)8] (μM)
Concentration of [Ph3PBn]5[Dy(SCN)8] (μM)
Fig. 5. Cytotoxicity of a) [Ph4P]5[Dy(SCN)8], b) [Ph3PBnOEt]5[Dy(SCN)8], c) [Ph3PBnNO2]5[Dy(SCN)8], and d) [Ph3PBn]5[Dy(SCN)8] against breast cancer (Hs578T, diamonds) and
normal (Hs578Bst, squares) cell lines, 48 h incubation. All points are mean s.d of triplicate wells of three independent experiments. Data is considered significant when pb0.05.
phosphonium–dysprosium complex (Fig. 6a). Note that the blue fluo-
rescence emission observed in Fig. 6a is due to the treatment of the
cell line with a nuclear marker. This comparison clearly demonstrates
that the green fluorescence emission observed in Fig. 6c should be as-
cribed to the fluorescence emission of the [Ph4P]5[Dy(SCN)8]. A con-
focal fluorescence micrograph of a single MIAPaCa-2 cell is also
provided (Fig. 6) which clearly demonstrates that [Ph4P]5[Dy(SCN)8]
was primarily associated with the cell membrane and can be imaged
by use of the greenish fluorescence. These cellular uptake and fluores-
cence imaging experiments confirm that [Ph4P]5[Dy(SCN)8] has a
unique binding (accumulating) property on or in the MIAPaCa-2 cell
membrane. In order to further verify the experimental results from
fluorescence imaging, we acquired an additional confocal fluores-
cence microscopic imaging of a different hormone-independent
human breast carcinoma cell line (MDA-MB-231) using another
phosphonium–dysprosium compound, [Ph3PBnNO2]5[Dy(SCN)8], as
the fluorescence imaging agent (Fig. S42, Supporting information).
This cell line has been previously subjected to extensive studies of
cellular uptake of different tumor images or therapeutic agents by
use of fluorescence microscopy [49–52]. Again, the bright greenish
yellow fluorescence emission confirms that [Ph3PBnNO2]5[Dy
(SCN)8] can be internalized by the cancer cell line MDA-MB-231
and visualized by use of confocal laser microscope. As a control test,
the image of the cancer cells without adding any phosphonium–
lanthanide complex proved that the cell line itself does not produce
greenish yellow fluorescence (Fig. S43, Supporting information).
Note that the nuclear marker was not used in this test.
Conclusions
We have developed four multifunctional lipophilic phosphonium–
dysprosium compounds. Three functionalities were incorporated into
a single molecule through a simple and straightforward synthesis.
These compounds possess paramagnetic and luminescent properties
which arise from the dysprosium-containing anions, as well as carci-
noma cell selectivity attributable to the aryltriphenylphosphonium
cations. In vitro studies demonstrate that the four compounds selec-
tively inhibit the growth of tumor cells, as compared to normal
cells, suggesting that they may be of potential use as cancer therapeu-
tic agents. In addition, these compounds could also be used as fluores-
cence imaging markers in vital cell cultures.
Acknowledgment
This work was supported in part by a grant from the National Sci-
ence Foundation (NSF, CHE-0911118). We acknowledge K.V. McDo-
nough for cell cultures, Dr. M. Brown for confocal fluorescence
microscopy experiments, and Dr. R. Cueto for DSC and TGA tests.
Table 2
Inhibitory concentrations (IC50) of the four compounds against cancer cell lines and normal cell line: 48 h assay.
Cell lines
IC50 (μM)a
[Ph4P]5[Dy(SCN)8]
[Ph3PBnOEt]5[Dy(SCN)8]
[Ph3PBnNO2]5[Dy(SCN)8]
[Ph3PBn]5[Dy(SCN)8]
Hs578Bst
Hs578T
>200
83
74
19
147
91
85
22
MDA-MB-231
HT29
121
251
28
86
141
303
194
85
PANC-1
>400
>450
>400
>400
a
The values for IC50 were calculated from the slopes of cell viability as a function of concentration graphs for each of the investigated compounds and cell lines.

46
M. Li et al. / Journal of Inorganic Biochemistry 107 (2012) 40–46
a
c
b
d
Fig. 6. Visualization of live cancer cells in the presence of [Ph4P]5[Dy(SCN)8] by Confocal Laser Microscopy. Pancreatic cancer cells (MIAPaCa-2) were incubated with [Ph4P]5[Dy(SCN)8] at
a concentration of 10 μM for 1 h and washed with PBS to remove excess of [Ph4P]5[Dy(SCN)8] and live cells were visualized under confocal laser microscope equipped with an argon–krypton
laser Blue Stain:(a) Localization of nucleus (control test showing cell nucleus without using phosphonium–dysprosium compound; Green Stain: (b) localization of [Ph4P]5[Dy(SCN)8] only)
(c) overlapped (a) and (d) showing [Ph4P]5[Dy(SCN)8] was up-taken by the cells. Excitation at 458 nm. Data clearly indicate that [Ph4P]5[Dy(SCN)8] localizes in the cancer cell membrane.
[26] T. Welton, Chem. Rev. 99 (1999) 2071–2084.
[27] B. Mallick, B. Balke, C. Felser, A.-V. Mudring, Angew. Chem. Int. Ed. 47 (2008)
Appendix A. Supplementary data
7635–7638.
Supplementary data to this article can be found online at doi:10.
1016/j.jinorgbio.2011.09.035.
[28] P. Nockemann, B. Thijs, N. Postelmans, K. Van Hecke, L. Van Meervelt, K. Binnemans,
J. Am. Chem. Soc. 128 (2006) 13658–13659.
[29] S. Tang, A. Babai, A.-V. Mudring, Angew. Chem. Int. Ed. 47 (2008) 7631–7634.
[30] S. Hayashi, H. Hamaguchi, Chem. Lett. 33 (2004) 1590–1591.
[31] Y. Zheng, Y. Lan, W. Wernsdorfer, C.E. Anson, A.E. Powell, Chem. Eur. J. 15 (2009)
12566–12570.
[32] Y. Yoshida, G. Saito, Phys. Chem. Chem. Phys. 12 (2010) 1675–1684.
[33] A. Kamal, G. Chouhan, Tetrahedron Lett. 46 (2005) 1489–1491.
[34] A. Getsis, B. Balke, C. Felser, A.-V. Mudring, Cryst. Growth. Des.
References
[1] K. Riehemann, S.W. Schneider, T.A. Luger, B. Godin, M. Ferrari, H. Fuchs, Angew.
Chem. Int. Ed. 48 (2009) 872–897.
[2] R. Haag, Angew. Chem. Int. Ed. 43 (2004) 278–282.
[3] L.B. Chen, Ann. Rev. Cell Biol. 4 (1988) 155–181.
[4] M.J. Weiss, J.R. Wong, C.S. Ha, R. Bleday, R.R. Salem, G.D. Steele Jr., L.B. Chen, Proc.
Natl. Acad. Sci. USA 84 (1987) 5444–5448.
9 (2009)
4429–4437.
[35] Z. Li, L. Zhou, J. Yu, H. Zhang, R. Deng, Z. Peng, Z. Guo, J. Phys. Chem. C 111 (2007)
2295–2300.
[36] J.-C.G. Bünzli, Chem. Rev. 110 (2010) 2729–2755.
[37] K. Wang, R. Li, Y. Cheng, B. Zhu, Coord. Chem. Rev. 190–192 (1999) 297–308.
[38] S.P. Fricker, Chem. Soc. Rev. 35 (2006) 524–533.
[39] Z.F. Chen, M.X. Tan, Y.C. Liu, Y. Peng, H.H. Wang, H.G. Liu, H. Liang, J. Inorg. Bio-
chem. 105 (2001) 426–434.
[40] B. Wang, Z.Y. Yang, P. Crewdson, D. Wang, J. Inorg. Biochem. 101 (2007)
1492–1504.
[41] D.J. Stephens, V.J. Allan, Science 300 (2003) 82–86.
[5] G. Kroemer, B. Dallaporta, M. Resche-Rigon, Annu. Rev. Physiol. 60 (1998)
619–642.
[6] J.S. Modica-Napolitanoa, J.R. Aprille, Adv. Drug Delivery Rev. 49 (2001) 63–70.
[7] V. Weissig, V.P. Torchilin, Adv Drug Deliver Rev. 49 (2001) 127–149.
[8] S.T. Richard, W. Horobin, Eur. Biophys. J. 34 (2005) 959–966.
[9] T.J. Preston, A. Abadi, L. Wilson, G. Singh, Adv Drug Deliver Rev. 49 (2001) 45–61.
[10] E.A. Liberman, V.P. Topall, L.M. Tsonna, A.A. Jasatis, V.P. Skulachev, Nature 222
(1969) 1076–1078.
[11] S.D. Bernal, T.J. Lampidis, R.M. McIsaac, L.B. Chen, Science 222 (1983) 169–172.
[12] S.D. Bernal, T.J. Lampidis, R.M. McIsaac, L.B. Chen, Science 218 (1982) 1117–1119.
[13] E.J. Delikatny, W.A. Cooper, S. Brammah, N. Sathasivam, D.C. Rideout, Cancer Res.
62 (2002) 1394–1400.
[14] X. Sun, J.R. Wong, K. Song, J. Hu, K.D. Garlid, L.B. Chen, Cancer Res. 54 (1994)
1465–1471.
[42] C.A. Puckett, J.K. Barton, J. Am. Chem. Soc. 129 (2007) 46–47.
[43] U. Neugebauer, Y. Pellegrin, M. Devocelle, R.J. Forster, W. Signac, N. Moran, T.E.
Keyes, Chem. Comm. (2008) 5307–5309.
[44] O. Alptürk, O. Rusin, S.O. Fakayode, W. Wang, G.O. Escobedo, I.M. Warner, W.E.
Crowe, V. kraal, J.M. Pruet, R.M. Strongin, Proc. Nat. Acad. Sci. USA (2006)
9756–9760.
[15] D.C. Rideout, T. Calogeropoulou, J.S. Jaworski, R. Dagnino Jr., M.R. McCarthy, Anti-
Cancer Drug Des. 4 (1989) 265–280.
[45] K. Yong, J. Qian, I. Roy, H. Lee, E.J. Bergey, K.M. Tramposch, S. He, M.T. Swihart, A.
Maitra, P.N. Prasad, Nano Lett. 7 (2007) 761–765.
[16] J. Min, S. Biswal, C. Deroose, S.S. Gambhir, J. Nucl. Med. 45 (2004) 636–643.
[17] J. Asin-Cayuela, A.B. Manas, A.M. James, R.A.J. Smith, M.P. Murphy, FEBS Lett. 571
(2004) 9–16.
[46] E.B. Voura, J.K. Jaiswal, H. Mattoussi, S.M. Simon, Nat. Med. 10 (2004) 993–998.
[47] N. Nasongkla, X. Shuai, H. Ai, B.D. Weinberg, J. Pink, D.A. Boothman, J. Gao, Angew.
Chem. Int. Ed. 43 (2004) 6323–6327.
[18] M.P. Murphy, R.A. Smith, Annu. Rev. Pharmacol. Toxicol. 47 (2007) 629–656.
[19] J.L. Hickey, R.A. Ruhayel, P.J. Barnard, M.V. Baker, S.J. Bernes-Price, A. Filipovska, J.
Am. Chem. Soc. 130 (2008) 12570–12571.
[20] C. Yang, Y. Kim, J. Wang, L. Wang, J. Shi, Z. Li, X. Chen, M. Fan, J. Li, S. Liu, Biocon-
jugate Chem. 19 (2008) 2008–2022.
[21] J. Wang, C.T. Yang, Y.S. Kim, S.G. Sreerama, Q. Cao, Z.B. Li, Z. He, X. Chen, S. Liu, J.
Med. Chem. 50 (2007) 5057–5069.
[22] M. Li, S.L. De Rooy, D.K. Bwambok, B. El-Zahab, J.F. Ditusa, I.M. Warner, Chem.
Commun. (2009) 6922–6924.
[48] J. Chen, S. Chen, X. Zhao, L.V. Kuznetsova, S.S. Wong, I. Ojima, J. Am. Chem. Soc.
130 (2008) 16778–16785.
[49] A.E. Karnoub, A.B. Dash, A.P. Vo, A. Sullivan, M.W. Brooks, G.W. Bell, A.L. Richardson,
K. Polyak, R. Tubo, R.A. Weinberg, Nature 449 (2007) 557–563.
[50] A. Müller, B. Homey, H. Soto, N. Ge, D. Catron, M.E. Buchanan, T. McClanahan, E.
Murphy, W. Yuan, S.N. Wagner, J.L. Barrera, A. Mohar, E. Verástegui, A. Zlotnik,
Nature 410 (2001) 50–56.
[51] W.J. Parak, R. Boudreau, M. Le Gros, D. Gerion, D. Zanchet, C.M. Micheel, S.C. Williams,
A.P. Alivisatos, C. Larabell, Adv. Mater. 14 (2002) 882–885.
[23] N.V. Plechkovaa, K.R. Seddon, Chem. Soc. Rev. 37 (2008) 123–150.
[24] F. van Rantwijk, R.A. Sheldon, Chem. Rev. 107 (2007) 2757–2785.
[25] V.I. Prvulescu, C. Hardacre, Chem. Rev. 107 (2007) 2615–2665.
[52] O.P. Mgbonyebi, J. Russo, I. Russo, Cancer Res. 59 (1999) 1903–1910.