Differential cytostatic and cytotoxic action of metallocorroles against human cancer cells: Potential platforms for anticancer drug development

Article abstract of DOI:10.1021/tx200452w

A gallium(III)-substituted amphiphilic corrole noncovalently associated with a targeting protein was previously found by us to confer promising cytotoxic and antitumor activities against a breast cancer cell line and a mouse xenograft breast cancer model.

Full text of DOI:10.1021/tx200452w

Article
pubs.acs.org/crt
Differential Cytostatic and Cytotoxic Action of Metallocorroles
against Human Cancer Cells: Potential Platforms for Anticancer Drug
Development
Punnajit Lim,† Atif Mahammed,‡ Zoya Okun,‡ Irena Saltsman,‡ Zeev Gross,*,‡ Harry B. Gray,*,§
and John Termini*,†
†Department of Molecular Medicine, Beckman Research Institute of the City of Hope, 1450 E. Duarte Road, Duarte, California
91010, United States
‡Schulich Faculty of Chemistry, TechnionIsrael Institute of Technology, Haifa 32000, Israel
§Beckman Institute, California Institute of Technology, Pasadena, California 91125, United States
S
* Supporting Information
ABSTRACT: A gallium(III)-substituted amphiphilic corrole
noncovalently associated with a targeting protein was
previously found by us to confer promising cytotoxic and
antitumor activities against a breast cancer cell line and a
mouse xenograft breast cancer model. To further explore
potential anticancer applications, the cytostatic and cytotoxic
properties of six nontargeted metallocorroles were evaluated
against seven human cancer cell lines. Results indicated that toxicity toward human cancer cells depended on the metal ion as
well as corrole functional group substitution. Ga(III)-substituted metallocorrole 1-Ga inhibited proliferation of breast (MDA-
MB-231), melanoma (SK-MEL-28), and ovarian (OVCAR-3) cancer cells primarily by arrest of DNA replication, whereas 2-Mn
displayed both cytostatic and cytotoxic properties. Confocal microscopy revealed extensive uptake of 1-Ga into the cytoplasm of
melanoma and ovarian cancer cells, while prostate cancer cells (DU-145) displayed extensive nuclear localization. The
localization of 1-Ga to the nucleus in DU-145 cells was exploited to achieve a 3-fold enhancement in the IC50 of doxorubicin
upon coadministration. Time−course studies showed that over 90% of melanoma cells incubated with 30 μM 1-Ga internalized
metallocorrole after 15 min. Cellular uptake of 1-Ga and 1-Al was fastest and most efficient in melanoma, followed by prostate
and ovarian cancer cells. Cell cycle analyses revealed that bis-sulfonated corroles containing Al(III), Ga(III), and Mn(III) induced
late M phase arrest in several different cancer cell lines, a feature that could be developed for potential therapeutic benefit.
adenovirus serotype 5 (Ad5) capsid penton base protein.13 This
feature can be exploited to facilitate their transport into various
cell types and allow for specific target design. Moreover, some
metallocorroles, such as 1-Ga and 1-Al, are highly fluorescent, a
INTRODUCTION
Although cancer incidence is on the rise, mortality has slowly
declined over the past few years due to improved under-
■
standing of tumor biology and more sophisticated diagnostic
methods and therapeutic approaches.1,2 However, standard
property that can be used to monitor uptake and intracellular
localization.14
treatment regimens have not changed radically, and there is still
an urgent need to develop better molecular platforms for
Previous work demonstrated that 1-Ga noncovalently bound
to a heregulin-modified protein directed at the human
improved cancer-targeting efficacy. Corroles, aryl-substituted
epidermal growth factor receptor (HER) was cytotoxic to
corrin derivatives, are amenable to complex functionalization
breast cancer cells and induced tumor regression in a mouse
and can form stable chelates with a wide variety of metal ions,
xenograft model.13,14 To date, tumor-bearing animals treated
allowing for great flexibility in the design of molecules with
unique physical and chemical properties.3,4
with corroles have not shown any marked toxic side effects,
suggesting their potential for anticancer drug develop-
Although corroles share limited structural features with the
ment.11,14,15 These observations, along with renewed interest
cobalt-chelating corrins present in vitamin B12, they display
in exploring the potential therapeutic applications of porphyrins
physical and chemical properties similar to porphyrins and
and related macromolecules,16 prompted our investigation of
corroles as potential platforms for the development of
chemotherapeutic agents. Because the efficiency of anticancer
drugs can vary substantially among different cell types, we
chelate a diverse array of metal ions.5,6 Sulfonic acid
substitution on β-pyrrole carbon atoms 2 and 17 or di- and
tri-meso pyridinium substitution confers amphiphilic properties
(Figure 1),7−9 which facilitates the formation of tight
noncovalent complexes between metallocorroles and a variety
of proteins such as human serum albumin (HSA),10 lip-
oproteins (LDL and HDL),11 transferrin,12 and a recombinant
Received: October 18, 2011
Published: December 20, 2011
© 2011 American Chemical Society
400
dx.doi.org/10.1021/tx200452w | Chem. Res. Toxicol. 2012, 25, 400−409

Chemical Research in Toxicology
Article
Figure 1. Structures and UV−vis spectroscopic data for metallocorroles used in this study.
NMR (400 MHz, C6D6): δ = 8.98 (d, J = 4.8 Hz, 4 H), 8.80 (d, J = 3.6
Hz, 2 H), 8.69 (m, 4 H), 8.47 ppm (br. s, 4 H), 8.15 (d, J = 8 Hz, 2
H), δ = 8.02 (d, J = 4.8 Hz, 4 H), 7.26 (d, J = 8 Hz, 2 H), 3.60 (s, 3
H). Manganese insertion and alkylation: 2-Mn was prepared by
refluxing corrole 2 in pyridine with 15 equiv of Mn(OAc)2 4H2O
followed by chromatographic separation (silica, starting with CH2Cl2
and gradually adding methanol), affording 88% yield. The product was
dissolved in hot THF, and an excess of methyl iodide was added to the
solution, which was then left at 40 °C until complete precipitation.
The solid material was collected by centrifugation and washed with
THF and diethyl ether until the solvent was colorless affording pure 2-
Mn in 94% yield. Compound 2-Mn: UV/vis (H2O): λmax (ε 10−3) =
354 nm (15.9), 492 nm (32.1), 601 nm (8.54), 684 nm (16.3). MS
(MALDI-TOF LD+) m/z (%): 640 (20) [M+], 625 (100) [M − 15].
These were dissolved in DMSO, and concentrations were determined
by UV spectroscopy using the extinction coefficients listed in Figure 1.
Stock solutions were further diluted in culture medium (RPMI 1640,
Mediatech, Inc., Manassas, VA). The final amount of DMSO in drug
solutions applied to tissue culture was <0.2%.
decided to investigate whether the antiproliferative effects of
corroles observed in breast cancer could be extended to other
tumor-derived cell lines.
We examined the effects of a series of metallocorroles against
seven NCI-60 cell lines17 derived from cancers of the breast,
prostate, ovary, skin, lung, and colon (Table 1) and observed
Table 1. Human Cancer Cell Line Panel
cell line
tumor type
histology
carcinoma
HCT-116
HCT-15
colon
colon
adenocarcinoma
carcinoma
DU-145
prostate
lung
NCI-H23
MDA-MB-231
SK-MEL-28
OVCAR-3
adenocarcinoma
adenocarcinoma
malignant
breast
melanoma
ovarian
adenocarcinoma
Reagents were obtained from the indicated suppliers: Hank's
balanced salt solution (HBSS) without phenol red, Invitrogen Corp./
Gibco, Carlsbad, CA; Triton X-100 and DAPI, Sigma-Aldrich, Inc., St.
Louis, MO; Paraformaldehyde, Electron Microscopy Sciences, Hat-
field, PA; Paclitaxel, Calbiochem, Gibbstown, NJ; Anti-Phospho-
Histone H3 (Ser10) Antibody Alexa Fluor 647 Conjugate, Cell
Signaling Technology, Danvers, MA; Hoechst 33342, Wheat Germ
Agglutinin, Alexa Fluor 488 Conjugate, Invitrogen Corp./Molecular
Probes, Carlsbad, CA; and Vectashield Hard-Set Mounting Medium
with DAPI, Vector Laboratories, Burlingame, CA.
differential cytotoxic and cytostatic responses with marked
selectivity against melanoma, breast, and ovarian cancers.
Kinetic studies of the intracellular uptake and accumulation
of metallocorroles revealed cell-specific properties, which
indicated their potential for tumor-selective targeting. Results
are presented that show that the observed cytotoxic and
cytostatic activities are dependent upon metal ion and
functional group substitution and that corroles may also exert
anticancer action by disrupting normal cell cycle progression.
Human Cancer Cell Lines. Seven cell lines from the NCI-60 cell
line panel representing six distinct tumor types (colon, prostate, lung,
breast, melanoma, and ovarian) were used in this study (Table 1).
Cells were grown in RPMI 1640 medium (Mediatech, Inc.)
supplemented with 10% fetal bovine serum (Omega Scientific, Inc.,
Tarzana, CA), 100 U/mL penicillin, and 100 μg/mL streptomycin
(Life Technologies, Inc., Rockville, MD) and maintained at 37 °C
under 5% CO2 in a humidified incubator.
MTS and BrdU Assay. Cytotoxicity was determined by cell
viability measurements using the CellTiter 96 Aqueous One Cell
Proliferation Assay (MTS) from Promega (Madison, WI). Cytostatic
activity was measured by the BrdU incorporation assay using the Cell
Proliferation ELISA, BrdU (colorimetric assay) from Roche Applied
Science (Indianapolis, IN). Cells were plated in seven 96-well dishes
(5 × 103 cells per well; 0.1 mL per well) 24 h prior to treatment. Drug
treatment was initiated by replacement with fresh media containing
EXPERIMENTAL PROCEDURES
■
Chemicals and Reagents. Bis-sulfonated and tris-pyridinium
metallocorroles (Figure 1) were synthesized as previously de-
scribed.5,6,18 The synthesis and characterization of 2-Mn was as
follows: p-Anisaldehyde (49 μL, 0.4 mmol) was added to a 10 mL
solution of pyridine-substituted dipyrromethane (178 mg, 0.8 mmol)
in propionic acid, and the mixture was heated to reflux for 1 h. The
residue obtained after solvent evaporation was washed with hot water,
neutralized with ammonium hydroxide (25%), and washed again with
hot water. Purification of corrole 2 was achieved by column
chromatography (silica, CH2Cl2 followed by 1% methanol addition)
affording pure 2 in 9% yield. Compound 2: Rf = 0.34 (CH2Cl2/EtOH
10:1). UV/vis (CH2Cl2): λmax (ε 10−3) = 420 (48.9), 574 (6.6), 623
(4.8), 649 (4.7). MS (MALDI-TOF): m/z (%): 558 (100) [M+]. H
1
401
dx.doi.org/10.1021/tx200452w | Chem. Res. Toxicol. 2012, 25, 400−409

Chemical Research in Toxicology
Article
Figure 2. Cytotoxic and cytostatic activities of 1-Ga toward (A) MDA-MB-231 (breast), (B) SK-MEL-28 (melanoma), (C) OVCAR-3 (ovarian),
(D) DU-145 (prostate), (E) NCI-H23 (lung), and (F) HCT-15 (colon) human cancer cells. Treated cells were incubated with 0.3, 3, or 30 μM 1-
Ga, followed by determination of viability and DNA replication competency using the MTS and BrdU incorporation assays, respectively, as described
in the Experimental Procedures.
0.3, 3, or 30 μM metallocorrole in 0.1 mL per well. All drug treatments
were performed in the dark to minimize the effects of metallocorrole
photochemistry. The MTS and BrdU assays were performed according
to the manufacturers' directions once a day for six consecutive days.
Absorbances were measured using a microplate reader (Synergy 4,
Biotek Instruments, Inc., Winooski, VT) at 490 nm for MTS and 370
nm for BrdU. Experiments were performed in triplicate with standard
deviation ranging from ∼0.5 to 15% of the measured absorbances. The
means from triplicate spectrophotometric data over a 7 day period
were analyzed by nonlinear regression analysis using GraphPad Prism
5 (GraphPad Software, Inc., La Jolla, CA) to obtain a growth equation
for each treatment. The first derivatives of nonlinear fits were used to
obtain the rate of change, which was normalized to the control. The
rate of cell proliferation (BrdU) or growth (MTS) of corrole-treated
cells relative to control cells was reported as percent of untreated cells.
Confocal Microscopy. Confocal images of 1-Ga-treated cells were
obtained using an upright LSM510 2-Photon microscope (Carl Zeiss
MicroImaging, Inc., Thornwood, NY). Cells were seeded on glass
coverslips at 1.5 × 105 cells per well, 2 mL per well in 24-well plates,
and allowed to grow overnight. Compound 1-Ga was added directly to
the cell media to 3 μM final concentration, followed by vortexing for 2
min at 500 rpm, and incubation for 3 h at 37 °C, 5% CO2. The culture
medium was aspirated, and cells were washed 2× with HBSS and then
labeled with 5 μg/mL WGA-Alexa Fluor 488 conjugate in HBSS for 10
min at 37 °C. The labeling solution was removed; then, cells were
washed 2× with PBS, fixed in 2% PFA/PBS for 20 min at room
temperature (RT), followed by a final PBS wash. For mounting, the
coverslips containing cells were removed from 24-well dishes, dipped
briefly in water, and directly mounted on slides containing 5 μL of
Vectashield Hard-Set Mounting Medium with DAPI. Slides were
placed at 4 °C in the dark for 12−24 h prior to imaging. Images were
obtained at 40× magnification with constant fluorescence, exposure,
and gain settings.
Point-Scanning Confocal Imaging of Uptake of Metallo-
corroles. The intracellular uptake of metallocorroles was quantified
using the ImageXpressultra laser point-scanning confocal microscope
(Molecular Devices, Inc., Sunnyvale, CA). A wavelength of 405 nm
was used for excitation of DAPI and 1-Ga or 1-Al. DAPI fluorescence
emission (blue) was measured with a 450 nm bandpass filter. Corrole
fluorescence (red) was monitored using a 675 nm bandpass filter.
Melanoma, prostate, and ovarian cancer cells were seeded in 96-well
dishes (8 × 103 cells per well in 0.1 mL) and allowed to grow
overnight. Compound 1-Ga or 1-Al was added directly to the media at
0, 0.3, 3.0, or 30 μM final concentration, vortexed for 2 min at 500
rpm, and incubated for 15 min or 1, 3, or 24 h at 37 °C under 5% CO2.
Cells were then fixed and labeled in situ with WGA-Alexa Fluor 488
conjugate and DAPI to delineate the outer cell membrane and nuclear
DNA, respectively. The quantification of DAPI fluorescence signals in
each sample yielded the total number of cells in that population. The
corresponding determination of red (1-Ga or 1-Al) fluorescence
within each cells, corrected for background fluorescence from
untreated cells, established the number of cells containing metallo-
corroles. For each corrole concentration at various time points,
fluorescence data were obtained for 5500−7000 individual cells. Data
402
dx.doi.org/10.1021/tx200452w | Chem. Res. Toxicol. 2012, 25, 400−409

Chemical Research in Toxicology
Article
Figure 3. Cytotoxic and cytostatic activities of Al(III)- and Mn(III)-substituted corroles against MDA-MB-231 (breast), SK-MEL-28 (melanoma),
and OVCAR-3 (ovarian) cancer cells. Cells were treated with 0.3, 3, and 30 μM 1-Al (A−C), 2-Mn (D−F), or 3-Mn (G−I).
were acquired on the ImageXpressultra at 81 sites in each well using a
20x Plan Fluor objective equipped with DAPI and Cy5 filter cubes.
Image and statistical analyses were performed using the MetaXpress
and AcuityXpress image analysis software, respectively. The number of
cells containing metallocorroles relative to total number of cells was
reported as percent of corrole-positive cells; the median fluorescence
intensity was calculated for each treatment and represented in a 3-axis
graph using Origin 7.0 (OriginLab Corp., Northampton, MA).
Cell Cycle Analysis. Metallocorrole-induced perturbations of the
cell cycle were examined using ImageXpressultra in conjunction with
the cell cycle application module for the MetaXpress image analysis
software. Statistical analyses were performed using AcuityXpress
cellular informatics software. Cells were plated in 96-well dishes (8 ×
103 cells per well in 0.1 mL) and allowed to grow overnight.
Metallocorroles (30 μM final concentration of 1-Ga, 1-Fe, 1-Mn, 1-Al,
2-Mn, or 3-Mn) or paclitaxel (0.1 μM) was added to cell-containing
media. After 24 h at 37 °C under 5% CO2, cells were washed 2× with
PBS and fixed in 2% PFA/PBS for 20 min at RT, followed by three
additional PBS washes of 5 min each. Immunostaining was carried out
according to the manufacturer's protocol. Briefly, cells were incubated
in blocking buffer (0.3% Triton X-100/PBS) for 1 h at RT. Anti-
Phospho-Histone H3 (Ser10) Antibody (Alexa Fluor 647 Conjugate)
was diluted at 1:400 ratio in antibody dilution buffer (1% BSA/0.3%
Triton X-100/PBS) and applied to cells after the removal of blocking
buffer. Cells were incubated at 4 °C overnight in the dark. After they
were washed with PBS (3×) to remove free antibody, cells were
incubated for 30 min with Hoechst Dye (3 μg/mL in PBS), then
rinsed once, and stored in PBS at 4 °C for imaging. Images were
403
dx.doi.org/10.1021/tx200452w | Chem. Res. Toxicol. 2012, 25, 400−409

Chemical Research in Toxicology
Article
acquired on the ImageXpressultra at 16 sites in each well as described
above. For each treatment, fluorescence data and cell cycle
determination were obtained for ∼2000 individual cells. The number
of cells in various phases of the cell cycle relative to the total number
of cells was reported as percent cells in respective phases of the cell
cycle.
1-Ga as a Chemotherapy Adjuvant. The potential application of
metallocorroles as a carrier molecule for chemotherapeutic agents was
investigated by a coadministration of 1-Ga with doxorubicin. Cells
were plated in 96-well dishes (5 × 103 cells per well in 0.1 mL) and
allowed to grow overnight. Doxorubicin alone (final concentrations of
0.01, 0.1, 0.5, 1.0, or 1.75 μM) or in combination with 3 μM 1-Ga was
added to cell-containing media. Following 48 h of exposure at 37 °C,
cell viability was determined using the MTS assay according to the
manufacturer's instructions. Absorbances were measured using a
microplate reader (Synergy 4, Biotek Instruments, Inc.) at 490 nm.
Experiments were performed in triplicate with standard deviations
ranging from 0.02 to 0.12. Spectrophotometric data were analyzed
by sigmoid dose−response, nonlinear regression analysis, and the IC50
values were calculated using GraphPad Prism 5 (GraphPad Software,
Inc.).
cancer cell lines. The bis-pyridinium 2-Mn demonstrated
cytostatic and cytotoxic activities against all three cell lines
(Figure 3D−F). The activity was most pronounced against the
ovarian cancer cell line (Figure 3F). Approximately 60% of
OVCAR-3 cells were killed at a dose of 30 μM, whereas 80% of
the surviving fraction exhibited DNA replication arrest. In
contrast, tris-pyridinium 3-Mn showed neither cytostatic or
cytotoxic activity at all concentrations tested (Figure 3G−I).
Ga(III) and Al(III) derivatives of 2 and 3 were also prepared
but did not display any unusual biological or physicochemical
properties to warrant further testing. Compounds 2-Fe and 3-
Fe compounds were synthesized but were found to be unstable
and were not considered further.
Metallocorrole metal ion and functional group substitution
will influence their association with potential carrier proteins,
cell membranes, DNA, and other potential molecular targets as
well as their redox properties. For example, anionic bis-
sulfonated metallocorroles with structure 1 have been shown to
interact strongly with lipoproteins, while cationic dipyridyl
analogues 2 were predominantly unbound or only weakly
associated with serum proteins.19 Thus, cellular uptake of 2-Mn
may not be dependent on protein carriers. The para-pyridinium
substitution of 2-Mn, coupled with its hydrophobic tetra-
pyrrole core, stabilizes interactions with DNA via electrostatic
binding to the phosphate backbone and intercalation,
respectively, in a manner analogous to ethidium bromide,
whereas ortho-pyridinium-substituted 3-Mn cannot intercalate
into DNA.7 This may explain the significant inhibition of DNA
synthesis and subsequent cytotoxicity observed for 2-Mn across
three different cancer cell lines, whereas 3-Mn was completely
inactive against all tested lines (Figure 3).
The redox chemistry of Mn(III)-substituted corroles toward
reactive oxygen and nitrogen species may also contribute to
their biological activity. Mn(III)-substituted corroles are pro-
oxidants and have been shown to initiate lipid peroxidation as
well as catalyze the oxidation of aromatic thiols to sulfoxides via
the intermediacy of high-valent Mn(V)oxo complexes.6,11 The
absence of redox chemistry for Ga(III) and Al(III) would
suggest that the toxicity against cancer cell lines observed for 1-
Ga and 1-Al must be independent of free-radical mechanisms.
However, 1-Ga and 1-Al may induce oxidative damage
indirectly. For example, it has been shown that gallium
compounds can enhance the cellular uptake of iron from
transferrin,20 which could increase the potential for oxidative
damage via Fenton type reactions.21 Gallium complexes are also
known inhibitors of ribonucleotide reductase, presumably
because they can compete with iron binding at the active
site.22,23 This would reduce dNTP levels in the nucleotide
precursor pool and restrict de novo DNA synthesis, which
could account in part for the cytostatic activities observed for 1-
Ga and 1-Al.
There has been longstanding interest in the use of Ga(III)
compounds for cancer treatment;24 however, drug development
has been hampered by the availability of suitable ligands that
can stabilize Ga(III) compounds against hydrolysis and
facilitate cell permeability.23 Nitrate and chloride gallium salts
have been extensively investigated for therapeutic purposes and
have been evaluated in phase I/II clinical trials against
hepatoma, lymphoma, and bladder cancer.24−26 The cytostatic
effects of GaNO3 against lymphoma cells in culture have been
reported to occur around 120 μM, while cytotoxicity becomes
apparent at ∼500 μM.25,27 Chelated Ga(III) derivatives appear
to have improved biological activity relative to the free salts,
RESULTS AND DISCUSSION
■
Differential Toxicity of 1-Ga toward Human Cancer
Cells. In light of promising results obtained against breast
cancer models, the activity of 1-Ga (Figure 1) toward seven
human cancer cell lines representing six distinct tumor types
(Table 1) was initially examined. Each cell line was incubated in
complete (i.e., serum-containing) medium with 1-Ga at 37 °C
for up to 7 days. The effects of 1-Ga against cancer cell lines
were evaluated every 24 h for up to 6 days using the MTS and
BrdU assays, which measure cytotoxic and cytostatic activities,
respectively. The integrated values for cytoxicity (MTS assay)
and proliferative capacity (BrdU incorporation) as a function of
1-Ga concentration are shown in Figure 2. The two colon
cancer cell lines, HCT-116 and HCT-15, exhibited nearly
identical responses to 1-Ga; thus, data for only HCT-15 are
shown (Figure 2F). In all lines, cell killing (cytotoxicity) by 1-
Ga was only evident at relatively high concentrations (30 μM).
In contrast, inhibition of DNA replication by 1-Ga (cytostatic
activity) was observed at concentrations as low as 300 nM in
breast and ovarian cancer cells (Figure 2A,C) and 3 μM in
melanoma cells (Figure 2B). Approximately 40−50% inhibition
of cell proliferation was observed in all three cell lines at 30 μM
1-Ga. In contrast, the cytostatic and cytotoxic effects of 1-Ga
against prostate, lung, and colon cancer cells were found to be
negligible, suggesting defined rather than broad toxicological
activity (Figure 2D−F).
Cytotoxic and Cytostatic Activities of Differentially
Substituted Metallocorroles. The toxicological properties of
metallocorroles toward cancer cell lines were influenced by
both the chelated metal ion and the corrole functional group
substitution pattern. Metal ion effects on the cytotoxic and
cytostatic activities within a corrole series were examined by
substituting Al(III) for Ga(III) in bis-sulfonated corrole 1.
Compound 1-Al showed significantly greater cytotoxicity in the
MTS assay as compared to 1-Ga against breast and ovarian
carcinoma lines at the 30 μM dose (Figure 3A,C vs Figure
2A,C), although there was little change in cytotoxicity toward
melanoma cells as a result of Al(III) substitution. Significant
dose-dependent cytostatic activity for 1-Al was only observed
against melanoma cells (Figure 3B). Cationic Mn(III)
substituted meso-bis para-pyridinium and tris ortho-pyridinium
corroles (2-Mn and 3-Mn, respectively, Figure 1) were also
studied for activity against melanoma, breast, and ovarian
404
dx.doi.org/10.1021/tx200452w | Chem. Res. Toxicol. 2012, 25, 400−409

Chemical Research in Toxicology
Article
Figure 4. Intracellular accumulation of 1-Ga (red fluorescence). (A) DU-145 (prostate), (B) SK-MEL-28 (melanoma), and (C) OVCAR-3
(ovarian) cells were treated with 1-Ga (3 μM for 3 h) or untreated (medium only). Cells were stained with Alexa Fluor 488 conjugated to WGA and
DAPI. Images were obtained at 40× magnification using an upright confocal microscope and displayed in three-color channels with a merged image.
Fluorescence settings were kept constant between imaging, and images were captured at fixed exposure and gain settings. The scale bar represents 20
μm.
and gallium maltolate was described as exerting cytotoxicity
against the CCRF-CEM leukemic cells at ∼65 μM27 and
between 25 and 35 μM against several hepatoma lines.28 These
latter values are similar to what we observed for Ga(III)
corroles (Figure 2). In our hands, 1-Ga was most effective as a
cytostatic agent, inhibiting DNA replication in 40% of OVCAR-
3 cells at a dose of 0.3 μM. To the best of our knowledge, this is
the first report of a Ga(III) compound displaying appreciable
405
dx.doi.org/10.1021/tx200452w | Chem. Res. Toxicol. 2012, 25, 400−409

Chemical Research in Toxicology
Article
submicromolar cytostatic activity. The enhanced chelate
stability relative to the maltolate derivative, and particularly in
view of the efficient cellular uptake revealed in these studies
(see below), suggest that corroles may represent an improved
platform for therapeutic metal ion delivery to cancer cells.
Uptake and Localization of 1-Ga. To assess whether the
variability in cytostatic activity of 1-Ga against melanoma and
ovarian relative to prostate cancer cells resulted from
differences in uptake and/or intracellular localization, confocal
image analyses were performed. Cells were incubated with 3
μM 1-Ga in complete media for 3 h or in media alone (Figure
4A−C). Cells were labeled with DAPI and WGA-Alexa Fluor
488 conjugate to image nuclear DNA (blue fluorescence) and
N-acetylglucosamine of the plasma membrane (green fluo-
rescence), respectively. Compound 1-Ga was readily observable
due to its intense red fluorescence a feature absent in untreated
cells (Figure 4). In the prostate DU-145 cells (Figure 4A), the
colocalization of blue and red fluorescence observed in the
merged images indicated nuclear localization of 1-Ga. In
contrast, melanoma and ovarian cancer cells retained strong
blue nuclear fluorescence from DAPI in the presence of 1-Ga
and exhibited mainly cytoplasmic accumulation of metallo-
corrole in the merged images (Figure 4B,C).
The cell-specific mechanisms of metallocorrole uptake in
these cancer lines are presently under investigation; however, it
is likely that internalization of bis-sulfonated corroles of
structure 1 was facilitated through noncovalent association
with serum carrier proteins.13 The binding of metallocorroles 1-
Ga, 1-Fe, and 1-Mn to lipoproteins, LDL, HDL, and transferrin
has been previously described.11,19 Structurally related
porphyrins have been shown to bind to the LDL receptor as
well as the mitochondrial outer membrane-associated benzo-
diazepine receptor.29,30 More recently, porphyrins have been
shown to enter the mitochondria via the ABCB6 transporter of
the ATP binding cassette superfamily.31 Analogous mechanisms
may be involved in the uptake and subcellular accumulation of
metallocorroles. The molecular basis for the nuclear localization
of 1-Ga in the DU-145 prostate line is unclear; however, it is
possible that this metallocorrole is not retained strongly within
lysosomes once internalized in these cells and can more readily
enter the nucleus in a manner that may be dependent on
protein carriers. It is interesting that the nuclear accumulation
of 1-Ga in DU-145 cells (Figure 4A) did not adversely affect
growth or replication within the sensitivity limits of our assays.
Time-Dependent Uptake of 1-Ga and 1-Al. The
ImageXpressultra system, a laser point-scanning confocal micro-
scope that integrates digital microscopy and flow cytometry,
was used to measure the time course of uptake of metallo-
corroles and the degree of intracellular accumulation. Uptake
was studied only for 1-Ga and 1-Al since Mn(III)-substituted
corroles are not fluorescent. Time-dependent accumulation at a
3.0 μM dose was measured in melanoma, ovarian, and prostate
cancer cells, and results are presented in Figure 5A,B for 1-Ga
and 1-Al, respectively. The y-axis indicates the percentage of
cells displaying detectable corrole fluorescence, whereas the z-
axis shows the median fluorescence intensity, which is directly
proportional to the extent of corrole uptake. In general, the
internalization of 1-Al was more efficient than 1-Ga. For
example, after 15 min, ∼80% of melanoma cells displayed 1-Al
fluorescence (Figure 5B), whereas <20% of cells were labeled
using 1-Ga at the same time point (Figure 5A). Approximately
85% of the DU-145 prostate cells display 1-Al fluorescence after
3 h, whereas <40% are labeled with 1-Ga at this time interval.
Figure 5. Time-dependent uptake and accumulation of 1-Ga and 1-Al
by SK-MEL-28 (red), DU-145 (green), and OVCAR-3 (blue) cancer
cells determined using the ImageXpressultra system. Cells were exposed
to 3.0 μM 1-Ga (A) or 1-Al (B) for the time points indicated on the x-
axis. The presence of metallocorrole within cells yielded red
fluorescence, presented as a percentage of corrole positive cells (y-
axis). The extent of corrole uptake was directly proportional to the
median fluorescence intensity (z-axis). Cell images were obtained at
20× magnification using filters for blue (DAPI) and red fluorescence.
After 24 h, 100% of cells from all three cancer lines displayed 1-
Al fluorescence, whereas quantitative labeling was only
observed for melanoma and prostate lines with 1-Ga. The
uptake of 1-Ga was very slow in ovarian cancer cells; after 24 h
of incubation, only ∼40% of cells exhibited fluorescence.
Melanoma cells displayed the most rapid and efficient uptake
with either metallocorrole, followed by prostate and ovarian
cancer cells. The efficient uptake into SK-MEL-28 melanoma
cells is consistent with transferrin-mediated transport of
metallocorroles,12 since these cells are well-known to avidly
accumulate iron via this mechanism.32 The efficiency of 1-Ga
uptake did not necessarily correlate with biological activity. For
example, prostate cancer cells demonstrated efficient uptake as
well as nuclear localization; yet, DNA replication was only
inhibited in ∼10% of the cell population at 30.0 μM (Figure
2D). Toxicity at this dose was only ∼20%. Although the uptake
of 1-Ga was least efficient in the ovarian carcinoma line, a
therapeutically relevant dose of 300 nM was sufficient to elicit
significant (∼40%) arrest of DNA synthesis (Figure 2C).
Metallocorroles Induce Mitotic Arrest. To further
investigate potential mechanisms of metallocorrole-induced
cytostatic and cytotoxic action, cell cycle analyses were
performed using the ImageXpressUltra. Melanoma, ovarian,
and breast cancer cells were treated with 30 μM 1-Ga, 1-Al,
1-Fe, 1-Mn, 2-Mn, and 3-Mn. As a control, cells were also
406
dx.doi.org/10.1021/tx200452w | Chem. Res. Toxicol. 2012, 25, 400−409

Chemical Research in Toxicology
Article
treated with 0.1 μM paclitaxel, which blocks cell division at the
G2/M checkpoint. Classification of cell cycle phases was based
on the fluorescence intensity of DNA dye (Hoechst 33342) and
mitotic-specific stains anti-Phospho-Histone H3 (Ser10) Alexa
fluor 647 conjugate. Cells progressing from G0/G1 to G2/M
double their DNA content and can be identified by the
corresponding increase in Hoechst 33342 fluorescence
intensity. Changes in fluorescence intensity induced by
phosphorylation at Ser10 of histone H3, a modification tightly
correlated with chromosome condensation during mitosis, were
used to detect cells in M phase. A representative cell cycle data
set is shown in Figure 6 for melanoma cells following exposure
distributions similar to untreated cells. In contrast, treatment
with 1-Ga, 1-Mn, and 1-Al resulted in stalling at the late mitotic
(M) phase. Similar profiles of cell cycle distributions also were
observed in breast and ovarian cancer cells incubated with these
metallocorroles (see the Supporting Information).
Compound 1-Al appeared to induce the strongest mitotic
arrest at a 30 μM dose, while varying degrees of late M arrest
were also observed for Mn(III)- and Ga(III)-substituted bis-
sulfonated corroles with structure 1. This late mitotic arrest,
distinct from what was observed for paclitaxel (G2 arrest),
suggested a mechanism other than the well-known tubulin
stabilization associated with taxol derivatives. Metallocorrole-
induced late M arrest is consistent with inhibition of
cytokinesis, an event that would ultimately lead to cellular
apoptosis.
Metallocorroles as Potential Adjuvants in Chemo-
therapy. The localization of 1-Ga to the nucleus of DU-145
prostate cancer cells, as well as its ability to form noncovalent
complexes with a variety of molecules due to its amphiphilic
nature, suggested a potential application as a carrier molecule
for DNA targeting chemotherapeutic agents. The ability of 1-
Ga to enhance drug cytotoxicity in DU-145 cells was examined
with the DNA-intercalating anthracycline drug doxorubicin
(Figure 7). IC50 values were measured for doxorubicin against
prostate, ovarian, and skin cancer lines in the presence of 3 μM
1-Ga. No enhancement was found for ovarian and skin cancer
cells, which only showed cytoplasmic accumulation of 1-Ga;
however, an ∼3-fold decrease in IC50 was observed for prostate
carcinoma cells. In contrast, the IC50 for the nonaromatic
topoisomerase inhibitor etoposide in prostate cancer cells was
unaffected by coadministration of 1-Ga (data not shown). The
putative association of metallocorroles and doxorubicin is
plausible in light of recent studies showing sequestration of
doxorubicin by porphyrin-modified micelles.33
Figure 6. Influence of metallocorroles on cell cycle progression of SK-
MEL-28 cells. Cells were incubated with various metallocorroles (30
μM 1-Ga, 1-Fe, 1-Mn, 1-Al, 2-Mn, and 3-Mn), paclitaxel (0.1 μM), or
complete medium alone. Quantitation of the distribution of cell cycle
phases was based on the fluorescent intensity of Hoechst 33342 and
Anti-Phospho-Histone H3 (Ser10) Antibody (Alexa Fluor 647
Conjugate) at 10× magnification using the ImageXpressultra system.
CONCLUSIONS
to various metallocorroles. For untreated cells, the majority
(∼80%) were observed to be in G0/G1 followed by G2, S, and
M. Inhibition of tubulin assembly by paclitaxel inhibited mitosis
and blocked cells at the G2/M checkpoint. Metallocorroles that
did not exhibit cytostatic or cytotoxic activity such as 3-Mn
(Figure 3G−I) and 1-Fe (data not shown) displayed cell cycle
■
Metallocorroles exhibited variable cytostatic and cytotoxic
action against several human cancer cell lines. Metal ion
substitution within a specific corrole nucleus influenced the
biological activity. For example, substitution of Al(III) for
Ga(III) significantly improved the cytotoxicity of bis-sulfonated
Figure 7. Effect of coadministration of 1-Ga on doxorubicin IC50 values in DU-145 (prostate), OVCAR-3 (ovarian), and SK-MEL-28 (melanoma)
cells. Dose−response curves from treatment with doxorubicin alone (blue) or doxorubicin +3 μM 1-Ga (red) for 48 h.
407
dx.doi.org/10.1021/tx200452w | Chem. Res. Toxicol. 2012, 25, 400−409

Chemical Research in Toxicology
Article
corrole 1 and also increased the rate of cellular uptake in three
unrelated cancer cell lines. However, additional structure−
function studies will be required to further elucidate the
influence of metal ion and functional group substitution on the
biological activity of metallocorroles. Ga(III)-substituted
corroles with structure 1 demonstrated dose-dependent
cytostatic activity against breast, skin, and ovarian cancer cell
lines. Although 1-Ga did not exert strong cytotoxic effects in
these cell lines under our assay conditions, the effective
cytotoxic dose (∼30 μM) was comparable to other Ga(III)
chelates that have been proposed as therapeutic agents27,28 and
was much improved over Ga(III) salts that have already been
tested in phase I/II clinical trials.24−26 Compound 1-Ga was
most active as a cytostatic agent against the SK-MEL-28
melanoma, MDA-MB-231 breast cancer, and OVCAR-3
ovarian carcinoma lines and inhibited replication of ∼40−
50% of the cell population at a dose of 30 μM (Figure 2B).
Compound 2-Mn displayed dose-dependent cytotoxic and
cytostatic effects, with the best activity observed against the
OVCAR-3 ovarian carcinoma cell line. In this case, ∼60% of
OVCAR-3 cells were killed at 30 μM 2-Mn (Figure 3F). These
effective cytotoxic doses are relatively high, but they are close to
the 34 μM IC50 value reported for a 96 h of treatment of MDA-
MB-231 breast cancer cells with cisplatin.34
It has been previously demonstrated that many cytostatic
agents that cause cell cycle arrest ultimately induce apoptosis or
cell death.35 Whether cytotoxicity follows cytostasis is largely a
function of the cellular environment and depends upon the
relative contributions of the different cell death mechanisms
induced in response to drug treatment. Further studies with
metallocorroles are required to determine whether the observed
cytostasis and cell cycle arrest ultimately results in cytotoxicity.
This will likely require in vivo testing using mouse xenograft
models and ultimately clinical trials. Many kinase inhibitors,
such as sorafenib, that were originally described as cytostatic
agents prolonging progression free survival were later
demonstrated in some clinical studies to ultimately induce
tumor regression.36
AUTHOR INFORMATION
■
Corresponding Author
*Tel: 626-301-8169. Fax: 626-930-5463. E-mail: chr10zg@tx.
technion.ac.il (Z.G.), hbgray@caltech.edu (H.B.G.), or
jtermini@coh.org (J.T.).
Funding
This work was supported by Caltech-City of Hope Biomedical
Research Initiative and the U.S. Army RDECOM Acquisition
Center, Natick Contracting Division, Natick, MA, under
Contract no. W911QY-10-C-0176.
ACKNOWLEDGMENTS
■
The technical assistance of Dr. Richard Yip and Charlie Hsin of
the High Throughput Screening and Dr. Brian Armstrong of
the Light Microscopy Digital Imaging Core Facilities of the
City of Hope Comprehensive Cancer Center is gratefully
acknowledged.
REFERENCES
■
(1) American Cancer Society (2010) Cancer Facts & Figures 2010,
American Cancer Society, Atlanta.
(2) Jemal, A., Siegel, R., Xu, J., and Ward, E. (2010) Cancer statistics,
2010. CA: Cancer J. Clin. 60, 277.
(3) Aviv-Harel, I., and Gross, Z. (2011) Coordination chemistry of
corroles; with focus on main group elements. Coord. Chem. Rev. 255,
717−736.
(4) Aviv-Harel, I., and Gross, Z. (2009) Aura of corroles. Chem. Eur.
J. 15, 8382−8394.
(5) Mahammed, A., Goldberg, I., and Gross, Z. (2001) Highly
selective chlorosulfonation of tris(pentafluorophenyl)corrole as a
synthetic tool for the preparation of amphiphilic corroles and metal
complexes of planar chirality. Org. Lett. 3, 3443−3446.
(6) Mahammed, A., and Gross, Z. (2005) Albumin-conjugated
corrole metal complexes: Extremely simple yet very efficient
biomimetic oxidation systems. J. Am. Chem. Soc. 127, 2883−2887.
(7) Gershman, Z., Goldberg, I., and Gross, Z. (2007) DNA binding
and catalytic properties of positively charged corroles. Angew. Chem.
Int. Ed. 46, 4320−4324.
(8) Ma, H., Zhang, M., Zhang, D., Huang, R., Zhao, Y., Yang, H., Liu,
Y., Weng, X., Zhou, Y., Deng, M., Xu, L., and Zhou, X. (2010) Pyridyl-
substituted corrole isomers: Synthesis and their regulation to G-
quadruplex structures. Chem.Asian J. 5, 114−122.
Metallocorroles used in conjunction with other drugs may
also provide new approaches in cancer chemotherapy. A 3-fold
improvement in the IC50 of doxorucibin (to 0.8 μM) against a
prostate cancer cell line was achieved upon coadminstration
with 1-Ga (Figure 7). It is interesting to note in this regard that
Ga(III) salts have been previously used in conjunction with
vinblastine and ifosfamide (VIG therapy) for the treatment of
cisplatin resistant urothelial cancers.37 Future development of
metallocorroles into useful chemotherapeutic agents would
involve optimization of functional group substitution, expanded
screening to uncover additional susceptible cancer cell lines,
and improved intracellular targeting efficacy via covalent or
noncovalent association with carrier molecules and/or proteins.
(9) Saltsman, I., Botoshansky, M., and Gross, Z. (2008) Facile
synthesis of ortho-pyridyl-substituted corroles and molecular struc-
tures of analogous porphyrins. Tetrahedron Lett. 49, 4163−4166.
(10) Mahammed, A., Gray, H. B., Weaver, J. J., Sorasaenee, K., and
Gross, Z. (2004) Amphiphilic corroles bind tightly to human serum
albumin. Bioconjugate Chem. 15, 738−746.
(11) Haber, A., Mahammed, A., Fuhrman, B., Volkova, N., Coleman,
R., Hayek, T., Aviram, M., and Gross, Z. (2008) Amphiphilic/Bipolar
metallocorroles that catalyze the decomposition of reactive oxygen and
nitrogen species, rescue lipoproteins from oxidative damage, and
attenuate atherosclerosis in mice. Angew. Chem. Int. Ed. 47, 7896−
7900.
(12) Haber, A., Agadjanian, H., Medina-Kauwe, L. K., and Gross, Z.
(2008) Corroles that bind with high affinity to both apo and holo
transferrin. J. Inorg. Biochem. 102, 446−457.
(13) Agadjanian, H., Weaver, J. J., Mahammed, A., Rentsendorj, A.,
Bass, S., Kim, J., Dmochowski, I. J., Margalit, R., Gray, H. B., Gross, Z.,
and Medina-Kauwe, L. K. (2006) Specific delivery of corroles to cells
via noncovalent conjugates with viral proteins. Pharm. Res. 23, 367−
377.
(14) Agadjanian, H., Ma, J., Rentsendorj, A., Valluripalli, V., Hwang, J.
Y., Mahammed, A., Farkas, D. L., Gray, H. B., Gross, Z., and Medina-
Kauwe, L. K. (2009) Tumor detection and elimination by a targeted
gallium corrole. Proc. Natl. Acad. Sci. U.S.A. 106, 6105−6110.
ASSOCIATED CONTENT
■
S
* Supporting Information
Profile of cell cycle distributions for breast and ovarian cancer
cells obtained following treatment with various metallocorroles.
This material is available free of charge via the Internet at
http://pubs.acs.org.
408
dx.doi.org/10.1021/tx200452w | Chem. Res. Toxicol. 2012, 25, 400−409

Chemical Research in Toxicology
Article
(15) Aviezer, D., Cotton, S., David, M., Segev, A., Khaselev, N., Galili,
N., Gross, Z., and Yayon, A. (2000) Porphyrin analogues as novel
antagonists of fibroblast growth factor and vascular endothelial growth
factor receptor binding that inhibit endothelial cell proliferation, tumor
progression, and metastasis. Cancer Res. 60, 2973−2980.
(16) Berg, K., Selbo, P. K., Weyergang, A., Dietze, A., Prasmickaite,
L., Bonsted, A., Engesaeter, B. O., Angell-Petersen, E., Warloe, T.,
Frandsen, N., and Hogset, A. (2005) Porphyrin-related photo-
sensitizers for cancer imaging and therapeutic applications. J. Microsc.
218, 133−147.
(17) Shoemaker, R. H. (2006) The NCI60 human tumour cell line
anticancer drug screen. Nature Rev. 6, 813−823.
(18) Saltsman, I., Mahammed, A., Goldberg, I., Tkachenko, E.,
Botoshansky, M., and Gross, Z. (2002) Selective substitution of
corroles: Nitration, hydroformylation, and chlorosulfonation. J. Am.
Chem. Soc. 124, 7411−7420.
(19) Haber, A., Aviram, M., and Gross, Z. (2011) Protecting the
beneficial functionality of lipoproteins by 1-Fe, a corrole-based
catalytic antioxidant. Chem. Sci. 2, 295−302.
Porphysome nanovesicles generated by porphyrin bilayers for use as
multimodal biophotonic contrast agents. Nat. Mater. 10, 324−332.
(34) Pope, A. J., Bruce, C., Kysela, B., and Hannon, M. J. (2010)
Issues surrounding standard cytotoxicity testing for assessing activity of
non-covalent DNA-binding metallo-drugs. Dalton Trans. 39, 2772−
2774.
(35) Zaffaroni, N., Silvestrini, R., Orlandi, L., Bearzatto, A., Gornati,
D., and Villa, R. (1998) Induction of apoptosis by taxol and cisplatin
and effect on cell cycle-related proteins in cisplatin-sensitive and
-resistant human ovarian cells. Br. J. Cancer 77, 1378−1385.
(36) Escudier, B., Eisen, T., Stadler, W. M., Szczylik, C., Oudard, S.,
Siebels, M., Negrier, S., Chevreau, C., Solska, E., Desai, A. A., Rolland,
F., Demkow, T., Hutson, T. E., Gore, M., Freeman, S., Schwartz, B.,
Shan, M., Simantov, R., and Bukowski, R. M. (2007) Sorafenib in
advanced clear-cell renal-cell carcinoma. N. Engl. J. Med. 356, 125−134.
(37) Einhorn, L. H., Roth, B. J., Ansari, R., Dreicer, R., Gonin, R., and
Loehrer, P. J. (1994) Phase II trial of vinblastine, ifosfamide, and
gallium combination chemotherapy in metastatic urothelial carcinoma.
J. Clin. Oncol. 12, 2271−2276.
(20) Richardson, D. R. (2001) Iron and gallium increase iron uptake
from transferrin by human melanoma cells: Further examination of the
ferric ammonium citrate-activated iron uptake process. Biochim.
Biophys. Acta 1536, 43−54.
(21) Tampo, Y., Kotamraju, S., Chitambar, C. R., Kalivendi, S. V.,
Keszler, A., Joseph, J., and Kalyanaraman, B. (2003) Oxidative stress-
induced iron signaling is responsible for peroxide-dependent oxidation
of dichlorodihydrofluorescein in endothelial cells: Role of transferrin
receptor-dependent iron uptake in apoptosis. Circ. Res. 92, 56−63.
(22) Chitambar, C. R., Narasimhan, J., Guy, J., Sem, D. S., and
O'Brien, W. J. (1991) Inhibition of ribonucleotide reductase by gallium
in murine leukemic L1210 cells. Cancer Res. 51, 6199−6201.
(23) Jakupec, M. A., and Keppler, B. K. (2004) Gallium in cancer
treatment. Curr. Top. Med. Chem. 4, 1575−1583.
(24) Chitambar, C. R. (2010) Medical applications and toxicities of
gallium compounds. Int. J. Environ. Res. Public Health 7, 2337−2361.
(25) Hedley, D. W., Tripp, E. H., Slowiaczek, P., and Mann, G. J.
(1988) Effect of gallium on DNA synthesis by human T-cell
lymphoblasts. Cancer Res. 48, 3014−3018.
(26) Seidman, A. D., Scher, H. I., Heinemann, M. H., Bajorin, D. F.,
Sternberg, C. N., Dershaw, D. D., Silverberg, M., and Bosl, G. J. (1991)
Continuous infusion gallium nitrate for patients with advanced
refractory urothelial tract tumors. Cancer 68, 2561−2565.
(27) Chitambar, C. R., Purpi, D. P., Woodliff, J., Yang, M., and
Wereley, J. P. (2007) Development of gallium compounds for
treatment of lymphoma: gallium maltolate, a novel hydroxypyrone
gallium compound, induces apoptosis and circumvents lymphoma cell
resistance to gallium nitrate. J. Pharmacol. Exp. Ther. 322, 1228−1236.
(28) Chua, M. S., Bernstein, L. R., Li, R., and So, S. K. (2006)
Gallium maltolate is a promising chemotherapeutic agent for the
treatment of hepatocellular carcinoma. Anticancer Res. 26, 1739−1743.
(29) Candide, C., Morliere, P., Maziere, J. C., Goldstein, S., Santus,
R., Dubertret, L., Reyftmann, J. P., and Polonovski, J. (1986) In vitro
interaction of the photoactive anticancer porphyrin derivative
photofrin II with low density lipoprotein, and its delivery to cultured
human fibroblasts. FEBS Lett. 207, 133−138.
(30) Maziere, J. C., Morliere, P., and Santus, R. (1991) The role of
the low density lipoprotein receptor pathway in the delivery of
lipophilic photosensitizers in the photodynamic therapy of tumours. J.
Photochem. Photobiol. B 8, 351−360.
(31) Krishnamurthy, P. C., Du, G., Fukuda, Y., Sun, D., Sampath, J.,
Mercer, K. E., Wang, J., Sosa-Pineda, B., Murti, K. G., and Schuetz, J.
D. (2006) Identification of a mammalian mitochondrial porphyrin
transporter. Nature 443, 586−589.
(32) Richardson, D., and Baker, E. (1992) Two mechanisms of iron
uptake from transferrin by melanoma cells. The effect of desferriox-
amine and ferric ammonium citrate. J. Biol. Chem. 267, 13972−13979.
(33) Lovell, J. F., Jin, C. S., Huynh, E., Jin, H., Kim, C., Rubinstein, J.
L., Chan, W. C., Cao, W., Wang, L. V., and Zheng, G. (2011)
409
dx.doi.org/10.1021/tx200452w | Chem. Res. Toxicol. 2012, 25, 400−409