Ruthenium Complexes are pH-Activated Metallo Prodrugs (pHAMPs) with Light-Triggered Selective Toxicity Toward Cancer Cells

Article abstract of DOI:10.1021/acs.inorgchem.7b01065

Metallo prodrugs that take advantage of the inherent acidity surrounding cancer cells have yet to be developed. We report a new class of pH-activated metallo prodrugs (pHAMPs) that are activated by light- and pH-triggered ligand dissociation. These ruthenium complexes take advantage of a key characteristic of cancer cells and hypoxic solid tumors (acidity) that can be exploited to lessen the side effects of chemotherapy. Five ruthenium complexes of the type [(N,N)₂Ru(PL)]²⁺ were synthesized, fully characterized, and tested for cytotoxicity in cell culture (1_A: N,N = 2,2′-bipyridine (bipy) and PL, the photolabile ligand, = 6,6′-dihydroxybipyridine (6,6′-dhbp); 2_A: N,N = 1,10-phenanthroline (phen) and PL = 6,6′-dhbp; 3_A: N,N = 2,3-dihydro-[1,4]dioxino[2,3-f][1,10]phenanthroline (dop) and PL = 6,6′-dhbp; 4_A: N,N = bipy and PL = 4,4′-dimethyl-6,6′-dihydroxybipyridine (dmdhbp); 5_A: N,N = 1,10-phenanthroline (phen) and PL = 4,4′-dihydroxybipyridine (4,4′-dhbp). The thermodynamic acidity of these complexes was measured in terms of two pK_a values for conversion from the acidic form (X_A) to the basic form (X_B) by removal of two protons. Single-crystal X-ray diffraction data is discussed for 2_A, 2_B, 3_A, 4_B, and 5_A. All complexes except 5_A showed measurable photodissociation with blue light (λ = 450 nm). For complexes 1_A-4_A and their deprotonated analogues (1_B-4_B), the protonated form (at pH 5) consistently gave faster rates of photodissociation and larger quantum yields for the photoproduct, [(N,N)₂Ru(H₂O)₂]²⁺. This shows that low pH can lead to greater rates of photodissociation. Cytotoxicity studies with 1_A-5_A showed that complex 3_A is the most cytotoxic complex of this series with IC₅₀ values as low as 4 μM (with blue light) versus two breast cancer cell lines. Complex 3_A is also selectively cytotoxic, with sevenfold higher toxicity toward cancerous versus normal breast cells. Phototoxicity indices with 3_A were as high as 120, which shows that dark toxicity is avoided. The key difference between complex 3_A and the other complexes tested appears to be higher uptake of the complex as measured by inductively coupled plasma mass spectrometry, and a more hydrophobic complex as compared to 1_A, which may enhance uptake. These complexes demonstrate proof of concept for dual activation by both low pH and blue light, thus establishing that a pHAMP approach can be used for selective targeting of cancer cells.

Full text of DOI:10.1021/acs.inorgchem.7b01065

Article
pubs.acs.org/IC
Ruthenium Complexes are pH-Activated Metallo Prodrugs (pHAMPs)
with Light-Triggered Selective Toxicity Toward Cancer Cells
Fengrui Qu,† Seungjo Park,‡ Kristina Martinez,§ Jessica L. Gray,† Fathima Shazna Thowfeik,∥
John A. Lundeen,† Ashley E. Kuhn,⊥ David J. Charboneau,†,⊥ Deidra L. Gerlach,† Molly M. Lockart,†
James A. Law,† Katherine L. Jernigan,† Nicole Chambers,† Matthias Zeller,# Nicholas A. Piro,g
W. Scott Kassel,⊥ Russell H. Schmehl,§ Jared J. Paul,*,⊥ Edward J. Merino,*,∥ Yonghyun Kim,*,‡
and Elizabeth T. Papish*,†
†Department of Chemistry, The University of Alabama, Tuscaloosa, Alabama 35487-0336, United States
‡Department of Chemical and Biological Engineering, The University of Alabama, Tuscaloosa, Alabama 35487-0203, United States
§Department of Chemistry, Tulane University, New Orleans, Louisiana 70118, United States
∥Department of Chemistry, University of Cincinnati, Cincinnati, Ohio 45221, United States
⊥Department of Chemistry, Villanova University, Villanova, Pennsylvania 19085, United States
#Department of Chemistry, Purdue University, West Lafayette, Indiana 47907, United States
gDepartment of Chemistry, Albright College, Reading, Pennsylvania 19612, United States
S
* Supporting Information
ABSTRACT: Metallo prodrugs that take advantage of the
inherent acidity surrounding cancer cells have yet to be devel-
oped. We report a new class of pH-activated metallo prodrugs
(pHAMPs) that are activated by light- and pH-triggered ligand
dissociation. These ruthenium complexes take advantage of a key
characteristic of cancer cells and hypoxic solid tumors (acidity)
that can be exploited to lessen the side effects of chemotherapy.
Five ruthenium complexes of the type [(N,N)2Ru(PL)]2+ were
synthesized, fully characterized, and tested for cytotoxicity in cell
culture (1A: N,N = 2,2′-bipyridine (bipy) and PL, the photo-
labile ligand, = 6,6′-dihydroxybipyridine (6,6′-dhbp); 2A: N,N =
1,10-phenanthroline (phen) and PL = 6,6′-dhbp; 3A: N,N = 2,3-
dihydro-[1,4]dioxino[2,3-f][1,10]phenanthroline (dop) and PL = 6,6′-dhbp; 4A: N,N = bipy and PL = 4,4′-dimethyl-6,6′-
dihydroxybipyridine (dmdhbp); 5A: N,N = 1,10-phenanthroline (phen) and PL = 4,4′-dihydroxybipyridine (4,4′-dhbp). The
thermodynamic acidity of these complexes was measured in terms of two pKa values for conversion from the acidic form (XA) to
the basic form (XB) by removal of two protons. Single-crystal X-ray diffraction data is discussed for 2A, 2B, 3A, 4B, and 5A. All
complexes except 5A showed measurable photodissociation with blue light (λ = 450 nm). For complexes 1A−4A and their
deprotonated analogues (1B−4B), the protonated form (at pH 5) consistently gave faster rates of photodissociation and larger
quantum yields for the photoproduct, [(N,N)2Ru(H2O)2]2+. This shows that low pH can lead to greater rates of photo-
dissociation. Cytotoxicity studies with 1A−5A showed that complex 3A is the most cytotoxic complex of this series with IC50
values as low as 4 μM (with blue light) versus two breast cancer cell lines. Complex 3A is also selectively cytotoxic, with sevenfold
higher toxicity toward cancerous versus normal breast cells. Phototoxicity indices with 3A were as high as 120, which shows that
dark toxicity is avoided. The key difference between complex 3A and the other complexes tested appears to be higher uptake of
the complex as measured by inductively coupled plasma mass spectrometry, and a more hydrophobic complex as compared to
1A, which may enhance uptake. These complexes demonstrate proof of concept for dual activation by both low pH and blue light,
thus establishing that a pHAMP approach can be used for selective targeting of cancer cells.
target all quickly dividing cells, including healthy cells. There is a
need for new targeted therapies. Targeted therapies can use the
unique redox properties of the cell, the lowered pH of the cells,
INTRODUCTION
■
For certain types of cancer (e.g., testicular cancer), platinum-
based drugs are commonly prescribed as one of the most effective
treatments. However, the side effects of treatment can include
nerve damage, nausea, gastrointestinal distress, hair loss, and
other life-threatening side effects. Cisplatin and carboplatin
Received: April 27, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.7b01065
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
and focused light to minimize off target effects. Light is already
used to treat certain cancers using the Food and Drug
Administration-approved technique of photodynamic therapy
(PDT).1 Targeted therapies for cancer often rely on the inherent
differences between cancerous and normal tissues.2,3 Each cancer
has unique genetic mutations, but certain metabolic character-
istics are common to most cancers. These similarities include that
most solid tumors are hypoxic and exhibit decreased extracellular
pH (pHe) relative to normal tissue.4,5 These metabolic
differences are known as the Warburg effect, and they present
a unique opportunity in cancer research that has been
unexploited thus far in metal-based drugs. Our work aims to
target the lowered external pH of cancerous cells through pH-driven
differences in photodissociation with visible light. Ruthenium com-
plexes have shown promise for pH and light-activated cyto-
toxicity (Schemes 1 and 2).2,3
an IC50 of 20 μM versus a colon carcinoma cell line,7 and it is
activated by reduction of RuIII to RuII. Both NKP-1339 and
KP1019 (a close analogue) have performed well in clinical trials
with minimal side effects.8,9 While these drugs do bind DNA in
the cell nuclei, the evidence suggests that DNA binding is not
the primary mode of action.9 Rather, NKP-1339 and KP1019
are believed to kill cancer cells by disturbing the cellular redox
balance, which can lead to cell cycle arrest and apoptosis.9
NAMI-A also entered clinical trials and was less cytotoxic (IC50 ≈
400 μM) but had anti-metastasis activity.10 NAMI-A’s activity
appears to derive from its ability to scavenge nitric oxide and
thereby prevent endothelial cell migration and angiogenesis.11,12
The extensive studies of NAMI-A and NKP-1339 also illustrate
that similar complexes can have vastly different modes of action
and, furthermore, that in vitro toxicity (in cell culture) often does
not predict in vivo toxicity (in animal models) or success in
clinical trials.9,13 Most notably, a light-activated ruthenium com-
plex (TLD-1433) is about to enter clinical trials in Canada for
bladder cancer, showing the promise of drugs that are visible
light-activated via PDT.14,15 TLD-143316,17 and many other
PDT agents18−20 act as photosensitizers and generate singlet
oxygen to kill cancer cells. Our system is different, in that it does not
require oxygen and can work in hypoxic solid tumors. Our work falls
under photoactivated chemotherapy (PACT),21,22 but our
system is unique in using both pH and light for activation
(Scheme 1).
The use of prodrugs that are activated by the low pHe present
in cancerous tissue has been rare.4 Tumor-activated prodrugs
that use pH for drug activation have been limited to organic
drugs and polymers.23−25 Some of the hardest cancers to treat
(including “triple negative” breast cancer (TNBC) and cancer
stem cells (CSCs)) acidify the extracellular matrix (ECM) to a
significant extent26 and should be vulnerable to a strategy that
derives its selectivity from inherent pH differences. Cancerous
cells typically have a low pHe of 6.2−6.8 (some sources mention
pHe values as low as 5),26 whereas normal cells have a higher pHe
of 7.2−7.4.27 Cancer cells exhibit fast growth and high aerobic
glycolysis that leads to low pHe.27,28 Our hypothesis is that by
designing compounds with appropriate features we can ensure both
good uptake and light-triggered toxicity (Schemes 1 and 2).
Glazer et al. have made great progress in the use of ruthenium
complexes that readily photodissociate to bind DNA and cause
cytotoxicity.21,29,30 These studies have used steric bulk near the
metal center to destabilize the octahedral geometry and facilitate
photolabilization of a ligand. The N,N ligands 2,3-dihydro-
[1,4]dioxino[2,3-f][1,10]phenanthroline (dop), 2,9-dimethyl-
1,10-phenanthroline (dmphen), and 2,2′-biquinoline (biq) are
known to enhance photodissociation (with blue, red, or near-IR
light) by causing strain in octahedral ruthenium com-
plexes.21,29,30 Turro, Dunbar, and co-workers have used cytotoxic
ligands (e.g., 5-cyanouracil) on ruthenium for light-triggered
release.31−34 Meggers and Gasser have shown that attachment of
a drug to a metal center can alter its cytotoxicity and its
subcellular localization.19,35−37 McFarland, Sadler, and others
have shown that changing the charge of the metal complex (e.g.,
with cyclometalated C,N bound ligands) can greatly alter the
cytotoxicity.6,38−41 We plan to expand upon these studies by
investigating new strategies with pH-sensitive hydroxy-sub-
stituted ligands, using our synthetic experience to make new
ligands and metal complexes.42−46
Scheme 1. Conceptual Framework for pH-Activated Metallo
Prodrugs
a
Scheme 2. A Mechanism for Cancer Selectivity Based upon
pH with 1A = [(bipy)2Ru(6,6′-dhbp)]2+
a
A similar mechanism is proposed for derivatives that vary the
spectator and photolabile ligands.
Ruthenium anticancer agents have been studied less than
platinum agents, but they offer advantages in terms of more sites
for binding ligands (six for octahedral Ru(II) or Ru(III) versus
four for square planar Pt(II)) and thus more structural diversity.
The activity of ruthenium complexes in cells is often governed by
their three-dimensional structure, redox potential, and their
relative lipophilicity/hydrophilicity, which determines uptake.6
Metal-based prodrugs that are activated in a hypoxic, reducing
environment have been developed, and two of these redox-
activated prodrugs have entered clinical trials (Chart 1). For
example, NKP-1339, [Na][trans-RuIIICl4(benzimidazole)2] has
Chart 1. Ruthenium Complexes That Have Entered Clinical
Trials
In a prior publication and patent application we reported
a prototype complex, [(bipy)2Ru(6,6′-dhbp)]2+ (1A), where
6,6′-dhbp = 6,6′-dihydroxybipyridine, which was both light- and
B
DOI: 10.1021/acs.inorgchem.7b01065
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
pH-activated by ligand loss and showed modest toxicity (IC50
=
photodissociation rates). The new complexes 2A−4A were fully
characterized by spectroscopic and analytical techniques.
88 μM) toward HeLa (cervical cancer) cells.2,3 In this work, we
vary both the spectator and the photolabile ligand to make new
ruthenium complexes that are more cytotoxic and show selective
toxicity for cancerous versus normal cells. These new complexes
also offer improved phototoxicity indices.
Synthesis and Characterization of the Ligands and
Metal Complexes. The spectator ligands (N,N) were
purchased (bipy, phen) or synthesized (dop) using known
procedures.49−51 The new photolabile ligand dmdhbp was
synthesized by modification of known procedures.52 As shown in
Scheme 4, the commercially available 2,6-dichloro-4-methylpyridine
RESULTS
■
Research Design. Our prototype molecule, [(bipy)2Ru-
(6,6′-dhbp)]2+ (1A), is an active prodrug in vitro due to photo-
dissociation in blue light to give a complex that binds to bio-
logical targets (Scheme 2). The photolabile ligand 6,6′-dhbp and
closely related ligands (e.g., 4,4′-dimethyl-6,6′-dihydroxybipyr-
idine (dmdhbp) in Scheme 3) were specifically designed for this
Scheme 4. Synthesis of 4,4′-Dimethyl-6,6′-
dihydroxybipyridine
Scheme 3. New Complexes Designed as pHAMPs
was treated with a large excess of sodium metal in dry methanol
under air-free conditions. The resulting 2-chloro-6-methoxy-4-
methylpyridine was used without purification in the coupling
reaction with itself in the presence of Zn, Bu4NBr, and
NiCl2(PPh3)2 as the catalyst in dry dimethylformamide
(DMF) to afford the 6,6′-dimethoxy-4,4′-dimethyl-2,2′-bipyr-
idine. A deprotection procedure involving treatment with HBr/
AcOH yielded the final product, dmdhbp.
project for three reasons. (1) The ligand provides steric bulk near
the metal center, which distorts the metal geometry and
facilitates light-triggered ligand loss due to weakened M−N
bonds. Upon photoexcitation, an electron will be promoted into
a metal−ligand antibonding molecular orbital that disrupts the
metal−ligand interaction and causes ligand dissociation. Light-
triggered excitation leads to occupation of this metal-centered
excited state that weakens metal−ligand bonding (Figure 7).
(2) The ligand is pH-sensitive and will have a different pro-
tonation state in cancerous versus normal tissue. These proton-
ation states will differ with respect to how readily they undergo
[light-triggered] ligand loss. (3) The ligand is new in transition-
metal chemistry, with the Papish group reporting some of the
first metal complexes of this ligand in 2011.44,46−48 Thus, its
biochemistry and potential uses were unknown when we began
this project.
The synthesis of 1A has been reported previously.2 Using a
similar procedure, RuCl3·(H2O)x was treated with cycloocta-1,5-
diene (cod) to afford the intermediate (η4-cod)RuCl2. This
intermediate was then treated with 2 equiv of the spectator ligand
(N,N) in 1,2-dichlorobenzene to yield cis-(N,N)2RuCl2. After
purification, treatment of cis-(N,N)2RuCl2 with the photolabile
ligand (PL = 6,6′-dhbp or dmdhbp) in aqueous and/or alcoholic
solution led to complexes 1A−4A as the chloride salt,
[(N,N)2Ru(PL)]Cl2 (Scheme 5). The final products were
Scheme 5. Synthesis of the Metal Complexes
We varied the spectator bidentate (N,N) ligand from bipy (1A)
to phenanthroline (phen in 2A) to 2,3-dihydro-[1,4]dioxino-
[2,3-f][1,10]phenanthroline (dop in 3A; Scheme 3). The dop
ligand is described in the literature as enhancing photo-
dissociation by creating a twist via the nonplanar six-membered
ring containing sp3 atoms (O−CH2−CH2−O).21 Thus, we syn-
thesized a series of new ruthenium complexes (2A−4A) that are
both pH and light activated (Scheme 3). These new complexes
are only photoactive under conditions in which the dhbp ligand is
protonated. Throughout this paper, XA indicates the acidic form
of the complex (e.g., 2A = [(phen)2Ru(6,6′-dhbp)]2+, isolated
as the chloride salt), and XB is the deprotonated form (e.g., 2B).
In this work, we varied the photolabile ligand to look at the
impact of methyl electron donor groups in dmdhbp. Complex 4A
was initially designed for its higher pKa value (to give a larger
proportion of 4A versus 4B in solution and thereby better
recrystallized by slow diffusion of diethyl ether in a solution of
dry ethanol to afford the pure product. In a similar fashion,
[(phen)2Ru(4,4′-dhbp)]Cl2 (5A) was synthesized to test the
impact of moving the position of the hydroxyl groups (4,4′-dhbp =
4,4′-dihydroxybipyridine).
C
DOI: 10.1021/acs.inorgchem.7b01065
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 1. ORTEP diagrams (with ellipsoids at 50% probability) of the cation of 2A (left) and the neutral species 2B (right) with hydrogen atoms (except
OH) and counteranions (for 2A) omitted for clarity. Selected bond lengths (Å) and angles (deg): 2A: Ru1−N1, 2.0855(17); Ru1−N2, 2.0872(17);
Ru1−N3, 2.0752(17); Ru1−N4, 2.0627(17); Ru1−N5, 2.0549(17); Ru1−N6, 2.0520(17); 2B: Ru1−N1, 2.0940(14); Ru1−N2, 2.0516(15); Ru1−N3,
2.0613(15). A full list of structural parameters is included in the Supporting Information.
Table 1. Selected Bond Lengths and Angles of the Complexes 1A, 2A, 3A, 1B, 2B, and 4B
1A
1B
2A
2B
3A
4B
5A
Ru−N1
Ru−N2
Ru−N3
Ru−N4
Ru−N5
Ru−N6
avg Ru−N
avg_Ru−N (PL)
avg Ru−N (spectator)
N−Ru−N1
N−Ru−N2
N−Ru−N3
avg trans angle
2.091(2)
2.094(2)
2.046(2)
2.043(2)
2.066(2)
2.053(2)
2.066(1)
2.093(3)
2.052(1)
169.4(1)
173.9(1)
175.7(1)
173.0(1)
10.2(3)
−21.6(3)
15.9(4)
2.6(3)
2.097(2)
2.102(2)
2.044(1)
2.057(2)
2.051(1)
2.047(2)
2.066(1)
2.100(2)
2.050(1)
174.7(1)
173.4(1)
172.5(1)
173.5(1)
−16.6(2)
−3.1(2)
9.9(3)
2.086(2)
2.087(2)
2.075(2)
2.063(2)
2.055(2)
2.052(2)
2.070(1)
2.086(2)
2.061(1)
172.0(1)
172.8(1)
177.1(1)
174.0(1)
16.8(2)
−16.4(2)
16.6(3)
4.1(2)
2.094(1)
2.094(1)
2.052(2)
2.052(2)
2.061(2)
2.061(2)
2.069(1)
2.094(2)
2.056(1)
172.5(1)
171.9(1)
171.9(1)
172.1(1)
2.096(2)
2.096(2)
2.048(2)
2.048(2)
2.065(2)
2.066(2)
2.070(1)
2.096(2)
2.057(1)
171.0(1)
173.6(1)
173.6(1)
172.7(1)
2.083(1)
2.091(1)
2.055(1)
2.048(1)
2.047(1)
2.040(1)
2.061(1)
2.087(2)
2.048(1)
175.9(1)
170.6(1)
176.1(1)
174.2(1)
−12.9(2)
16.6(2)
14.8(3)
8.1(2)
2.062(2)
2.058(2)
2.058(2)
2.070(2)
2.061(2)
2.050(2)
2.060(1)
2.060(3)
2.060(1)
173.1(1)
173.4(1)
175.9(1)
174.1(1)
−2.9(3)
−5.2(3)
4.1(4)
a
b
b
L1 bend (PL)
−8.1
−7.4
b
b
−8.1
−7.4
b
b
avg |L1 bend|
L2 bend (spectator)
8.1
7.4
a
b
b
−3.1(2)
2.8(2)
0.7
4.1
4.5(3)
b
b
2.4(3)
−4.5(2)
−6.0(2)
5.0(2)
−4.3
−5.5
−0.3(2)
2.3(2)
−5.6(3)
−5.2(3)
5.1(3)
a
b
b
L3 bend (spectator)
6.4(3)
−1.2(2)
−1.4(2)
2.1(1)
0.7
4.1
b
b
−12.4(3)
6.0(2)
−4.3
−5.5
1.8(2)
b
b
avg |L2 or L3 bend|
4.9(1)
2.5
4.8
3.1(1)
5.1(2)
a
b
Bends are defined as the torsion angles of Ru−N−C-C′ as well as Ru−N′−C′−C. The estimated standard deviations (esds) for these are
estimated at zero by Mercury software, because C and C′ (or N and N′) are symmetry-related. In practice, the true esd should be 0.1 or 0.2.
Complexes 6−8, [(N,N)2Ru(H2O)2]SO4, where (N,N) =
bipy (6), phen (7), dop (8), represent the metal-containing
products of photodissociation with 1A−4A. We independently
synthesized complex 6 to test its cytotoxicity and to confirm that
it is the product of photodissociation with 1A and 4A. Complex 6
was synthesized by treating cis-(bipy)2RuCl2 with 1 equiv of
Ag2SO4 in water, followed by filtration to remove precipitated
AgCl. These complexes (1−6) were spectroscopically and
This suggests that the 6,6′-dhbp is more weakly bound to the
metal relative to the phen ligand. Significantly, the 6,6′-dhbp
ligand is distorted away from a planar geometry, as illustrated
by viewing the photolabile ligand side on in Figure 2 (red
wireframe = 2A) and Figure 5. For example, the oxygen atoms of
the dhbp ligand are 0.723 and 0.733 Å out of the plane defined by
Ru1, N1, and N2. This “bowing” type of distortion suggests that
the 6,6′-dhbp ligand introduces strain that may help promote
photodissociation. Similar bowing of the photolabile ligand was
seen for 1A (Figure 5).2
1
analytically characterized by H NMR, 13C NMR, IR, UV−vis
spectroscopy, and MS and/or elemental analysis.
X-ray Crystallography. Quality crystals of 2A for single-
crystal X-ray diffraction were obtained by slow diffusion
of diethyl ether into an ethanol solution of 2A. The structure
(Figure 1 and Table 1) shows a distorted octahedral geometry,
with the dhbp ligand showing longer Ru−N bond distances
(by 0.025(2) Å on average) than those for the phen ligand.
When 2A was deprotonated with Bu4NOH in ethanol solution,
a color change from bright red to black was observed. Crystals
suitable for X-ray diffraction were grown by slow diffusion of
diethyl ether into an ethanol solution of the complex. The crystal
structure confirmed that the acidic hydrogens were removed to
yield the neutral complex 2B, as shown in Scheme 3 and Figure 1.
D
DOI: 10.1021/acs.inorgchem.7b01065
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Similarly, crystals of 3A were grown by diffusion of diethyl
ether into an ethanol solution and these were used for X-ray
diffraction. Complex 3A features a distorted octahedral geometry,
and again, Ru−N distances were significantly longer (by 0.03−
0.05 Å) involving the dhbp ligand versus the dop ligand. Here,
the average Ru−N(dhbp) distance (2.096(2) Å) is 0.040(2) Å
longer than the average bond distance for Ru−N(dop)
(2.057(1) Å; Table 1). As a comparison, the bond distances
differ by 0.025 Å (2.086(2)Å − 2.061(1) Å) for 2A. The
other important feature of the structure of 3A is that the dop
ligands are not planar, as expected. The O−CH2−CH2−O con-
taining ring imparts a slight twist in the ligand (torsion angle,
defined by O−CH2−CH2−O chain = 61.3(4)°), as shown for
other Ru complexes of the dop ligand.21 The complex 3A also
shows a twist away from planarity in the photolabile 6,6′-dhbp
ligand (Figure 5). Unfortunately, we were unable to obtain X-ray
quality crystals for 3B, the deprotonated analogue.
Figure 2. Crystal structures of 2B (ellipsoids) and 2A (overlaid in red
wireframe) to show the changes upon protonation.
X-ray quality crystals of compound 4A were not obtained
despite several attempts. However, when the ethanol solution of
4A was treated with excess base (Bu4NOH), the bright red color
turned to black. Single crystals were grown by slow diffusion of
diethyl ether into the solution. A crystal structure of 4B is shown
in Figure 3. The central ruthenium ion features a mildly distorted
octahedral geometry, as seen with the other three ruthenium
compounds discussed above. The dmdhbp ring is mildly bowed/
puckered, similar to the structure in 2A, but with less strain. For
example, the oxygen atoms are 0.551 and 0.662 Å off the plane
defined by N5−Ru1−N6. As a comparison, in 2A the oxygen
atoms are 0.723 and 0.733 Å off the plane defined by Ru1, N1,
and N2. It is interesting to note that the dmdhbp ligand is bowed
in 4B in a similar way as for the photolabile ligand in 2A, as
opposed to the twisting of the photolabile ligand in 2B (Figure 5).
The dmdhbp ligand also shows a substantial downward tilt of the
hydroxyl groups below the plane of the Ru−N bonds (Figure 5).
In contrast to the other crystal structures above, the structure
of 5A (Figure 4) shows that the plane defined by Ru, N1, and N2
also contains most of the atoms in the 4,4′-dhbp ligand, including
the hydrogens at the 6 and 6′ positions. Clearly, upon moving
the hydroxy groups away from the metal center and to the
4,4′-positions, there is now no distortion at the metal center
(Figure 5) or in the bond lengths and angles involving the
4,4′-dhbp ligand (Ru−N distances are similar between the phen
The crystal structures for 2A and 2B are overlaid in Figure 2 to
illustrate the differences in geometry that occur with ligand
deprotonation. Similar to its acidic congener, the bond distances
of Ru−N in the deprotonated dhbp ligand are slightly longer
(by 0.038(2) Å on average, see Table 1) than for the phen ligand,
indicating weaker bonds, and deprotonated dhbp is predicted to
be the more labile ligand. Interestingly, the deprotonated dhbp,
[O2-bipy]2−, also features distortion away from a planar ligand
structure, but it is less severe than with the neutral 6,6′-dhbp
ligand in 2A (Figure 5). For example, the two oxygen atoms of
the dhbp ligands are 0.219 Å out of the plane defined by the
N1−Ru1−N2, and these two oxygen atoms are on the opposite
sides of this plane. This resulted in the ring twisting around the
C2−C2′ bond in the dhbp ligand (torsion angle 10.7(2)−
10.9(3)°), as compared to the ring “bowing” in the case of 2A.
The lesser distortion as well as the mild twisting (vs severe
bowing/puckering) suggest that 2A and 2B may have significant
differences in their tendencies to photodissociate (vide infra).
Strain that distorts the geometry of the ligand and the octahedral
geometry at the metal center is known to lower the energy of the
triplet metal-centered (3MC) excited state, from which photo-
dissociation typically occurs.32,53 Thereby, strain can increase the
quantum yields for photodissociation.
Figure 3. ORTEP diagrams (with ellipsoids at 50% probability) of the cations of 3A (left) and 4B (right) with hydrogen atoms (except OH) and
counteranions (for 3A) omitted for clarity. Selected bond lengths (Å) and angles (deg): 3A: Ru1−N1, 2.0962(17); Ru1−N2, 2.0477(16); Ru1−N3,
2.0655(16); 4B: Ru1−N1, 2.0548(13); Ru1−N2, 2.0483(13); Ru1−N3, 2.0403(13); Ru1−N4, 2.0473(13); Ru1−N5, 2.0906(13); Ru1−N6,
2.0832(13). A full list of structural parameters is included in the Supporting Information.
E
DOI: 10.1021/acs.inorgchem.7b01065
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 4. ORTEP diagram (with ellipsoids at 50% probability) of the
dication of 5A with hydrogen atoms (except OH) and counteranions
omitted for clarity. Selected bond lengths (Å) and angles (deg):
Ru1−N1, 2.062(2); Ru1−N2, 2.058(2); Ru1−N3, 2.058(3); Ru1−N4,
2.070(2); Ru1−N5, 2.061(2); Ru1−N6, 2.050(3). A full list of structural
parameters is included in the Supporting Information.
ligands and 4,4′-dhbp in Table 1, in contrast to significant differ-
ences in Ru−N for complexes 1A, 1B, 2A, 2B, 3A, and 4B). Not
surprisingly, complex 5A does not undergo photodissociation
with blue light. Figure 5 shows that complex 5A shows the least
distortion in the dhbp-derived ligand of all the complexes we
studied.
Thermodynamic Acidity. The pKa values for removing two
protons from 1A−4A are shown in Table 2, in which the acidity
data for 1A were previously reported.2 Complexes 2A−5A were
dissolved in aqueous solution, and these were treated with
4 equiv of HCl(aq) to ensure that the complexes were fully pro-
tonated. This solution was then titrated with NaOH to determine
the equivalence point. These experiments were done in low
lighting conditions to avoid photodissociation (with 2A−4A).
Although these complexes are all diprotic acids, after neutralizing
the excess HCl only one equivalence point was observed (with
2 equiv of base) showing that an averaged (avg) pKa value is
observed. Thus, pKa1 and pKa2 are close enough that we could
not observe individual protonation events using the titration
method. Similar results have been seen with other dhbp metal
complexes.47,48
However, spectrophotometric investigation of the pH-depen-
dent absorbance allowed determination of pKa1 values for several
complexes. We examined the spectra of each complex from pH 4
to pH 8 to determine a pH range in which the isosbestic points
are clearly defined. For example, for complex 2A, a clean iso-
sbestic point is observed at 351 nm (Figure 6) when the spectra
from pH 4 to pH 5.66 are plotted. By plotting the absorbance
values at 359 nm as a function of pH, we were able to estimate
pKa1 for 2A at 5.2(1) from two experiments. Similarly, we
estimated pKa1 for complexes 3A and 4A in the same manner
(Table 2). While clean isosbestic points were not observed in the
UV−vis spectra above pH 6 due to peaks shifting with pH,
Figure 5. A comparison of the photolabile ligand geometry in complexes
1A, 2A, 3A, 1B, 2B, and 4B.
Not surprisingly, 2A and 5A have very similar pKa values as the
hydroxy groups are simply moved to a different position on the
rings but are still in conjugation with the N donor atoms.
Complex 4A has the highest pKa values of the series (due to the
electron-donating methyl groups) and is expected to have the
highest concentration of protonated species at pH 7.2−7.4 in cell
culture.
Quantum Yields. The first step in photodissociation
involves exciting an electron into an metal-to-ligand charge
transfer (MLCT) excited state that is most likely localized on the
photolabile 6,6′-dhbp ligand (Figure 7). (While the spectator
we were able to estimate pKa2 from pKa avg and pKa1 (pKa2
=
2pKa avg − pKa1). See the Supporting Information for further
experimental details.
Thus, the average pKa value is known more precisely than the
values of pKa1 and pKa2. Focusing therefore on pKa avg, the trend
is that the pKa avg values increase in this order: 3A < (2A = 5A) < 1A
< 4A. This trend is expected because electron-withdrawing
groups (of phen and dop) tend to lower the pKa values.
F
DOI: 10.1021/acs.inorgchem.7b01065
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 2. A Comparison of Thermodynamic Acidity, Quantum Yields for Photodissociation upon Irradiation at 450 nm, and
a
log(Do/w) Values for Compounds 1A−5A
b
compound
structure
pKa1
pKa2
pKa avg
ϕ at pH = 5.0 (mostly XA) ϕ at pH = 7.5 (mostly XB) log(Do/w) at pH = 7.4
c
1A
[(bipy)2Ru(6,6′-dhbp)]2+
[(phen)2Ru(6,6′-dhbp)]2+
[(dop)2Ru(6,6′-dhbp)]2+
[(bipy)2Ru(dmdhbp)]2+
[(phen)2Ru(4,4′-dhbp)]2+
5.26
7.27
6.3
0.0058(5)
0.0020(2)
0.001(1)
0.0012(1)
0.76(3)
1.5(1)
1.2(1)
1.4(2)
2A
3A
4A
5A
5.2(2)
5.0(2)
5.6(2)
6.8(2)
6.8(2)
7.2(2)
6.0(1)
5.9(1)
6.4(1)
6.01(8)
0.000036(1)
0.00022(3)
0.00076(5)
0.0048(3)
a
b
c
Estimated standard deviations are in parentheses. The value of pKa2 was computed mathematically from pKa avg and pKa1. The pKa data for
complex 1A have been previously reported.
for reacting with DNA or other biomolecules.2 These results give
us qualitative measures of how pH influences photodissociation;
below we present quantum yields for a more quantitative
approach.
We measured the quantum yields for the photodissociation
reaction starting from complexes 1A−4A (Table 2). The quantum
yield is defined as moles of photoproduct ((N,N)2Ru(OH2)2)2+
(6, 7, or 8) per mole of photons. These experiments were done at
pH = 5.0, because this pH is below pKa1 in most cases, and
therefore the complexes should be mostly in the acidic form
(XA). Similarly, quantum yields were also measured at pH = 7.5
to represent the complexes in mostly the basic form (XB). Here
the trends are more important than the absolute numbers. The
quantum yield measurements are complicated by the fact that the
Figure 6. pH dependent UV−vis spectra of 2A as used to determine pKa1
from the change in absorbance at λ = 359 nm.
excited states can potentially undergo deprotonation/proto-
nation events. Our data support that XA is more photolabile than
XB, with the ratio of the quantum yields (ϕpH5/ϕpH7.5) ranging
from 4.5 (for 3A) to 4.8 (for 1A) to 6.3 (for 4A) to 56 (for 2A;
Table 2). The greater quantum yields for the acidic forms are
in line with the crystal structures of complexes 1A, 2A, and 3A,
which show significant distortion away from an ideal octahedral
geometry at the metal center and also distortions in the 6,6′-dhbp
ligands. In viewing the crystal structures of 2A and 2B (Figure 1),
it is clear that simply removing a proton changes the distortion at
Figure 7. Typical Jablonski diagram for photodissociation reaction
involving octahedral Ru(II) with imine ligands.
the metal center and thereby the efficiency of photodissociation,
as evident from ϕpH5/ϕpH7.5. Complex 4A was initially designed
for its relatively high pKa value (to give a larger proportion of
4A vs 4B in solution and thereby better photodissociation rates).
However, 4A has not led to an improvement in quantum yields.
The quantum yield for 4 at pH 7.5 is lower than that of 1 (mostly
4B and 1B at pH 7.5), and at pH 5.0 the quantum yields for
1 and 4 are similar (mostly 4A and 1A at pH 5.0). The low
quantum yield for 4B may be due to the electron-donating methyl
groups perturbing the relative energies of the excited-state
molecular orbitals (Figure 7, computational studies are ongoing).
Cytotoxicity toward Cancerous and Normal Cells. Cells
were treated with varied concentrations of 1A−5A in both the
dark and with irradiation by 450 nm blue light. The resulting IC50
values and phototoxicity indices (the ratio of dark IC50 to light
IC50) are shown in Table 2. These complexes are isolated and
administered in the acidic form (XA), but given that the pH
external is typically 7−7.5 in cell culture, these complexes would
exist as the neutral species (XB) in all but the most acidic
organelles. Treatment with 3A gives the best results with IC50
values of ∼4 μM in two breast cancer cell lines with light treat-
ment and phototoxicity indices up to 120 (red box, Table 2).
Importantly, complex 3A also displays selective toxicity toward
cancer cells (sevenfold more active toward breast cancer cells
vs normal cells). The toxicity of complex 3A is more moderate
toward MOLM-13 cells, but again some selectivity is seen for
cancerous versus normal umbilical cord blood (UCB) cells.
ligands also show MLCT bands, these excitations do not lead to
1
photodissociation due to stronger M−N bonds.) The MLCT
state is then typically converted (with near unity efficiency) to a
3MLCT state via intersystem crossing, and then conversion to
3
the MC (metal centered) state can lead to photodissociation
products (Figure 7). During much of this process, back reaction
to the ground-state starting material competes with product
formation.22
Our past work has shown that 1A has pKa values of 5.26 and
7.27, which indicate that the complex is predominantly (64%) in
the form of 1A at pH 5 and predominantly (63%) in the form of
1B at pH 7.5.2 At pH 7.5, UV−visible spectroscopy studies after
irradiating the sample with blue light (photoexcitation at
450 nm) show that ligand loss does not occur to a significant
extent. The resonance structures in Scheme 2 offer an expla-
nation in terms of a dianionic ligand binding strongly to the
ruthenium(II) center.2 Thus, at normal physiological pH values
the ruthenium center does not have free sites for DNA binding.
The data also suggest that the mono-deprotonated form,
[(bipy)2Ru(6,6′-O,OH-bipy)]+, which is the minor component
(37%) at pH 7.5, is not photolabile.2 In contrast, at pH 5, upon
irradiation with blue light complete 6,6′-dhbp ligand loss occurs
over 1 h. This indicates that, in an acidic environment, exposure
to light would cause complex (1A) to form 6, which has free sites
G
DOI: 10.1021/acs.inorgchem.7b01065
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Complex 4A appears to have low toxicity in most cases and is not
light-activated to any appreciable extent (PI values ∼1 for most
cell lines). In conclusion, it appears that the order of increasing
cytotoxicity, phototoxicity, and cancer selectivity is 4A ≤ 1A < 2A
≪ 3A. The reasons for this trend appear to be a combination of
drug uptake and light activation. Complex 5A has only been
tested in one cell line, but like 4A it also shows poor cytotoxicity
and is not light-activated to any significant extent. This under-
scores the need for both distortion at the metal center and effi-
cient photodissociation.
ruthenium complexes are localized in any particular organelles.
Furthermore, the uptake of 3A (dosed at the concentration that
gave 50% viability, incubated, and the cells were washed with
media before being lysed) was 7.2 and 3.9 μg/L with MDA-MB-
231 and MCF10A cell lines, respectively, as measured by
inductively coupled plasma mass spectrometry (ICP-MS). In a
separate experiment, cells were dosed with 10 μM (chosen to be
well below the dark IC50 values) of 1A−4A, and Ru uptake was
measured in the dark (Table 3, in green). Uptake is best for 3A
(Figure 9 and Table 3), but even in this case it is at most 0.32%
in MCF10A (normal cells) and ∼0.02% in MDA-MB-231 and
MCF7. Clearly, improving uptake is an appropriate goal for
future studies.59
Some pyridone derivatives are drugs,54−56,75 but 6,6′-dhbp is
nontoxic as the free ligand (IC50 > 100 μM) versus the breast cell
lines above. We cannot exclude the possibility that 6,6′-dhbp has
better uptake when bound to Ru versus as the free ligand.
We synthesized one product of photodissociation, [(bipy)2Ru-
(OH2)2]SO4 (6), and this complex was nontoxic (>100 μM), but
cellular uptake has not been measured yet. Ru complexes
with labile sites typically have poor uptake and may react with
molecules outside the cell.6,57,75
Octanol/Water Partition Coefficients. The hydrophilic
versus lipophilic (hydrophobic) nature of the complexes gives an
estimate as to how readily these complexes will penetrate the cell
membrane by passive diffusion. Ideally the complexes should be
more lipophilic while still maintaining sufficient water solubility
for drug delivery (log(Do/w) should ideally be 4−6).58 We
measured the octanol−water partition coefficient of complexes
1A−4A at pH 7.4, and we report log(Do/w) in Table 2. At
pH = 7.4, the major species in solution should be the basic
form XB, which is a neutral species, but there can still be a
significant fraction of the singly deprotonated monocation, and
therefore total ruthenium complex concentration was measured.
Lipophilicity increases in the order of 1 < 3 < 2, as expected
because of increased hydrophobic rings on the spectator ligands.
The methyl groups of the dmdhbp ligand serve to increase the
lipophilicity of 4 (cf. 1). Here, the species 1 (or 2−4) refers to all
protonation states that result when 1A is dissolved at pH 7.4.
Cellular Uptake. We measured cellular uptake of 3A and
other drugs both qualitatively and quantitatively. Most ruthe-
nium complexes are phosphorescent and can be visualized by
microscopy. Figure 8 shows that 3A is in the cytoplasm of the
cells. Further studies would be needed to determine if these
DISCUSSION
■
Initially, we anticipated that pKa would be a key factor in
determining which complex would be most effective at killing
cancer cells, and we expected that a higher pKa value would be
beneficial in giving a greater proportion of the protonated form
XA, which is more photolabile. Thereby, we expected that
complex 4A with higher pKa values (pKa avg = 6.4) would be the
most cytotoxic compound, but actually complex 4A is among the
least cytotoxic complexes we studied (cytotoxicity increases as
4A ≲ 1A < 2A ≪ 3A). Putting this factor in perspective, the pKa
values range from 5.9 to 6.4 for pKa avg and from 5.0 to 5.6 for
pKa1, and thus all of these values are pretty similar. In cell culture
at pH 7.0−7.5, these complexes will exist mostly as the mono-
deprotonated and doubly deprotonated (XB) species. Relatively
little of these complexes will be in the fully protonated and more
photolabile form XA. Building up the protonated species in a
cellular environment will require localization of the ruthenium
prodrug in an acidic organelle.
Complex 2 shows the largest change in quantum yields
between their acidic and basic forms (ϕpH5/ϕpH7.5 = 56 for
predominantly 2A vs 2B). The crystal structure for 2A shows
significant bowing of the photolabile ligand, whereas 2B shows
twisting but no bowing. For the complexes for which we
crystallized both the acidic and basic form, it is clear that the
acidic form (1A and 2A) always involves a greater twist
(as illustrated by the torsion angle, L1 bend, Table 1) than the
basic form (1B and 2B). The quantum yields for the acidic forms,
XA, were greatest for 1 and 4 at pH 5 (predominantly 1A or 4A in
solution). At pH 7.5, complex 1 is most photolabile. This may
suggest that greater light-triggered activity is expected for
complexes 1 and 4; however, all of the quantum yields are low
(ϕ = 1 × 10−3 or less),53 and the quantum yields are all relatively
similar (ranging from 5.8 × 10−3 for 1 to 1 × 10−3 for 3 at pH 5).
Thus, the uptake factor appears to be more important than either
the quantum yields or the pKa values. Complex 3 leads to the
highest ruthenium concentration in cells and is most cytotoxic
and shows the best phototoxicity indices. Only if these complexes
enter the cells does the photodissociation lead to drug activation
and binding to biological targets. The products of photo-
dissociation (e.g., 6) appear to have low cytotoxicity, perhaps due
to worse drug uptake in this form.
Thus, it appears that complex 3 is most cytotoxic due favorable
uptake in cancer cells and sufficient quantum yields. Selective
toxicity toward cancer cells is seen, but this effect cannot be
explained by uptake results, because uptake of 3 is actually greater
in normal cells versus cancer cells. It appears that the cancer cells
are more vulnerable to ruthenium treatment, perhaps due to
complex 3 localizing in an acidic environment within the cancer
cells (and thereby forming 3A rather than 3B). Further studies will
Figure 8. MDA-MB-231 cells treated with 3A at 3.65 μM and stained
with DAPI. (blue) The nucleus. (red) The Ru complex in the cytoplasm.
(top) 10×. (bottom) 40×.
H
DOI: 10.1021/acs.inorgchem.7b01065
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
a
Table 3. Cell Viability Data for Treatment with 1A−5A in the Dark and upon Irradiation for One Hour with Blue Light (450 nm)
vs MDA-MB-231 (breast CSC)
vs MCF7 (breast cancer)
vs MCF10A (normal)
com-pound
IC50 dark
IC50 light
PI
3.5
% uptake
IC50 dark
IC50 light
PI
% uptake
IC50 dark
IC50 light
PI
% uptake
1A
2A
3A
4A
5A
1010
280
290
83
0.0016
0.0071
0.022
>500
490
>500
180
∼1
2.8
0.0021
0.14
>500
110
210
13
>2.4
9
0.0086
0.022
0.32
3.3
52
190
3.7
490
4.1
120
∼5
1.5
0.025
0.0033
58
29
2
>1000
1140
∼1
0.0048
>500
>300
>100
194
>800
>460
∼1.7
0.011
vs MOLM-13 (leukemia)
vs UCB (normal)
vs HeLa (cervical cancer)
com-pound
IC50 dark
>100
IC50 light
PI
IC50 dark
IC50 light
PI
IC50 dark
IC50 light
PI
1A
2A
3A
4A
>100
27
∼1
2.2
1.8
14
114
120
122
132
0.93
0.91
0.96
∼1
148
1440
730
202
383
0.73
3.8
60
53
29
110
115
120
6.0
250
18
>100
>100
>100
>100
∼1
a
All IC50 values are in μM. Phototoxicity index (PI) is defined as IC50 dark/IC50 light. CSC: cancer stem cells, UCB: umbilical cord blood.
The 4,4′-dhbp was synthesized according to a previously published
procedure.63 Dry solvents were obtained by passing through a column of
activated alumina using a Glass Contour Solvent Purification System
built by Pure Process Technology, LLC.
General Specifications: Instruments and Analysis (Com-
1
plexes 1A−4A and 6−8, Papish Group). H, 13C, and 19F NMR
spectra were acquired at room temperature on a Bruker AV360
360 MHz or AV500 500 MHz spectrometer, as designated. Chemical
shifts are reported in parts per million and referenced to residual solvent
resonance peaks. Abbreviations for the multiplicity of NMR signals
are s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), br
(broad). UV−visible spectra were recorded on a PerkinElmer Lambda
35 UV−visible spectrometer. Mass spectrometric data were collected on
a Waters AutoSpec-Ultima NT spectrometer with electron ionization
method. Electrospray ionization mass spectrometry (ESI-MS) was
provided by the University of Alabama Mass Spectrometry Resource.
Elemental analyses were performed by Nu-Mega Resonance Labo-
ratories, Inc., San Diego, CA. Quantum yields were measured at Tulane
University; see the Supporting Information for further details.
General Specifications: Instruments and Analysis (Com-
plex 5A, Paul Group). UV−visible absorption spectra were collected
on a Scinco S-3100 diode-array spectrophotometer at a resolution of
1 nm. IR spectra were collected on a PerkinElmer Spectrum One FTIR
with ATR accessory. Cyclic voltammetry measurements were
performed on a Bioanalytical Systems CW-50 potentiostat. For studies
performed in 0.1 M tetrabutylammonium hexafluorophosphate
(TBAPF6) electrolyte in acetonitrile, a standard three-electrode setup
with platinum wire auxiliary electrode, glass carbon working electrode,
and Ag/Ag+ reference electrode filled with 0.1 M TBAPF6 was used with
a scan rate of 200 mV/s. The acetonitrile was dried over molecular sieves
and filtered prior to use to remove any particulates. Ferrocene was used
as an internal standard with E1/2 = +0.40 V versus saturated calomel
electrode (SCE).64
Figure 9. Ru uptake as measured by ICP-MS for cells treated with
1A−4A (10 μM) in the dark. Blue = MDA-MB-231. Red = MCF7,
Green = MCF10A.
be needed to clarify this matter and the pathway by which
3A causes cell death.
CONCLUSIONS
■
These experiments (with IC50 as low as 4 μM with 450 nm light)
show the potential of our prodrugs, which are already
comparable in cytotoxicity to cisplatin (IC50 < 2 μM)60,61 and
NKP-1339 (IC50 = 20 μM),7 which have entered clinical trials
(p 1).8,9,62 We have demonstrated that at lower pH values (pH 5)
these complexes exist in their acidic form (XA) and that the
quantum yields for photodissociation are higher in this form. The
change in quantum yields upon lowering the pH from 7.5 to 5 is
greatest for complexes 2 (56-fold), and changes in geometric
distortion are evident in the crystal structures of 2A and 2B. This
illustrates that pH can activate metallo prodrugs toward
photodissociation. Significantly, drug uptake is also enhanced
by the use of ligands with an extended ring structure for greater
hydrophobicity. These complexes are selectively cytotoxic
toward cancer cells (as compared with normal cells), and they
show efficient light activation with phototoxicity indices as high
as 120. These complexes are activated by two triggers (light and
pH) to minimize off target effects, thus showing the promise of a
pHAMP approach. Most metal-based drugs are prodrugs that are
activated by light or the reducing environment in a cancer cell,
but here we have used pH to activate these prodrugs, which
represents a new approach in metal-based chemotherapy.
Thermodynamic Acidity Measurements. The pH measurements
were performed using a VWR SympHony pH meter, utilizing a three-
point calibration at pH = 4, 7, and 10. Aqueous solutions were prepared
using a Millipore DirectQ UV water purification system. Buffer solutions
were made according to previous published methods.65
(cod)RuCl2. A patent procedure66 was followed. A 250 mL oven-
dried Schlenk flask equipped with a stir bar was evacuated and refilled
with nitrogen three times. RuCl3(H2O)x (2.074 g, 10.0 mmol) was
measured into the flask. Absolute EtOH (50 mL) was injected to the
flask. 1,5-Cyclooctadiene (3.70 mL, 30.0 mmol, 3.0 equiv) was injected.
A condenser was attached while flushing with N2. The final dark brown/
black solution was heated to reflux in an oil bath at 85 °C under N2 for
20 h. After it cooled, the product was collected by suction filtration. The
brown powder was washed with a small amount of EtOH, then 50 mL of
diethyl ether, then air-dried. Yield: 2.223 g, 7.935 mmol, 79.4%.
(phen)2RuCl2. A procedure67 reported for the synthesis of
cis-[Ru(bpy)2Cl2] was adapted. A 100 mL oven-dried Schlenk flask
equipped with a stir bar was evacuated and refilled with nitrogen three
EXPERIMENTAL SECTION
■
General Specifications: Materials. All experiments were
performed under a nitrogen atmosphere using glovebox or standard
Schlenk techniques if not indicated otherwise. All commercially available
reagents were purchased from Sigma-Aldrich, Acros, Strem, or Pressure
Chemical and were used as received without further purification.
I
DOI: 10.1021/acs.inorgchem.7b01065
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
times. (cod)RuCl2 (0.50 g, 1.785 mmol) and 1,10-phenanthroline
(0.643 g, 3.57 mmol, 2.0 equiv) were measured into the flask, followed
by 25 mL of anhydrous o-dichlorobenzene. The final dark brown
suspension was heated to reflux in a sand bath at 180 °C under N2 for
3 h. After it cooled, the product was collected by suction filtration. The
dark purple/black powder was washed with Et2O, then air-dried. Yield:
1.011 g, 1.778 mmol, 99.6%. To remove potential [(phen)3Ru]Cl2 the
crude product was stirred in 200 mL of deionized (DI) water (200 mL/g
of crude product) for a few hours. The product was then collected and
washed with ether and dried. 1H NMR (360 MHz, DMSO) δ 10.28 (dd, J =
5.2, 1.4 Hz, 2H), 8.71 (m, 2H), 8.29 (d, J = 8.9 Hz, 2H), 8.23 (m, 4H), 8.14
(d, J = 8.9 Hz, 2H), 7.75 (m, 2H), 7.33 (dd, J = 8.1, 5.4 Hz, 2H).
9.06; and H, 4.40%. Calc. for C38H28Cl2N6O6Ru·EtOH·H2O: C, 53.34;
N, 9.33; and H, 4.03%. ESI-MS (m/z): 383.1, Calcd for [(dop)2Ru-
(dhbp)]2+: 383.1. pKa1: 4.9; pKa2: 6.9; average: 5.9.
2-Chloro-6-methoxy-4-methylpyridine. A patent procedure68
was followed with adaptation. A 250 mL oven-dried Schlenk flask
equipped with a stir bar was evacuated and refilled with nitrogen three
times. Anhydrous MeOH (150 mL) was injected to the flask. Na (0.92 g,
40 mmol, 4.0 equiv) was added portionwise while flushing with N2.
To the resultant pale white emulsion the 2,6-dichloro-4-methyl-pyridine
(1.62 g, 10 mmol) was added. The final clear faint yellow solution was
heated to reflux in an oil bath at 80 °C under N2 for 5 d. After it cooled,
solvent was removed by rotary evaporation. The residue was diluted in
dichloromethane and washed with water. The organic layer was isolated,
washed with brine, and dried over MgSO4. Solvent was removed to give
the final product as a pale, yellow oil. Yield: 1.427 g, 9.05 mmol, 90.5%.
1H NMR (360 MHz, CDCl3) δ 6.74 (s, 1H), 6.46 (s, 1H), 3.91 (s, 3H),
2.28 (s, 3H).
[(phen)2Ru(6,6′-dhbp)]Cl2 (2A). A procedure2 reported for the
synthesis of [(bipy)2Ru(6,6′dhbp)](PF6)2 was adapted. The whole
process was protected from light, as the final product is sensitive to light.
A 100 mL oven-dried Schlenk flask equipped with a stir bar was charged
with DI water (30 mL) and was purged with nitrogen for 20 min.
(phen)2RuCl2 (0.200 g, 0.375 mmol) and 6,6′-dihydroxy-2,2′-
bipyridine (0.776 g, 0.412 mmol, 1.1 equiv) were measured into the
flask. The final dark brown suspension was heated to reflux in an oil bath
at 110 °C under N2 for 22 h. After it cooled, the solution was filtered to
remove any excess dhbp ligand. To the filtrate, a few drops of concen-
trated HCl were added. Water was removed by rotary evaporation. The
product was collected by suction filtration with the help of diethyl ether.
The red powder was washed with Et2O, then air-dried. Yield: 0.220 g,
0.3053 mmol, 81.4%. 1H NMR (360 MHz, dimethyl sulfoxide
(DMSO)) δ 11.88 (s, 2H), 8.82 (dd, J = 8.4, 1.3 Hz, 2H), 8.54 (dd,
J = 8.3, 1.3 Hz, 2H), 8.44 (dd, J = 5.3, 1.2 Hz, 2H), 8.34 (d, J = 8.9 Hz,
2H), 8.27 (d, J = 8.9 Hz, 2H), 8.19 (dd, J = 7.9, 1.1 Hz, 2H), 8.00 (dd, J =
8.3, 5.2 Hz, 2H), 7.86 (t, J = 8.0 Hz, 2H), 7.70 (dd, J = 5.4, 1.2 Hz, 2H),
7.49 (dd, J = 8.2, 5.3 Hz, 2H), 6.77 (d, J = 8.2 Hz, 2H). 13C NMR
(126 MHz, DMSO) δ 168.64, 156.13, 152.80, 152.06, 148.25, 147.90,
139.90, 136.07, 135.10, 129.94, 129.64, 127.73, 127.34, 125.97, 125.03,
115.34, 111.70. Elem. Anal.: Found: C, 55.04; N, 10.64; and H, 4.28%.
Calc. for C36H32Cl2N6O4Ru·EtOH·H2O: C, 55.11; N, 10.71; and H,
4.11%. MALDI-MS (m/z): 650.2, Calcd for [(phen)2Ru(dhbp)]+,
[C34H24N6O2Ru]+: 650.1. pKa1: 5.2; pKa2: 6.8; average: 6.0.
6,6′-Dimethoxy-4,4′-dimethyl-2,2′-bipyridine. A patent proce-
dure52 was followed with adaptation. A 250 mL oven-dried Schlenk flask
equipped with a stir bar was evacuated and refilled with nitrogen three
times. Dry DMF (150 mL) was injected to the flask. 2-Chloro-6-
methoxy-4-methylpyridine (2.0 g, 12.69 mmol), Zn (0.83 g, 12.69
mmol, 1 equiv), NiCl2(PPh3)2 (2.49 g, 3.807 mmol, 0.3 equiv), and
nBu4NBr (4.091 g, 12.69 mmol, 1 equiv) were added while flushing
with N2. The final black solution was heated to reflux in an oil bath at
55 °C under N2 for 5 d. After it cooled, the solvent was removed by
rotary evaporation. The residue was diluted in dichloromethane and
washed with water. The organic layer was isolated, washed with brine,
and dried over MgSO4. The solvent was removed to give the final
product as a light yellow residue. Yield: 0.276 g, 1.13 mmol, 17.8%.
1H NMR (500 MHz, CDCl3) δ 7.83 (dd, J = 1.3, 0.7 Hz, 1H), 6.56 (dq,
J = 1.7, 0.8 Hz, 1H), 4.03 (s, 3H), 2.38 (t, J = 0.7 Hz, 3H). 13C NMR
(126 MHz, CDCl3) δ 164.06 (s), 153.40 (s), 150.67 (s), 115.53 (s),
110.93 (s), 53.44 (s), 21.44 (s).
4,4′-Dimethyl-[2,2′-bipyridine]-6,6′-diol (dmdhbp). A litera-
ture procedure69 reported for the synthesis of analogous 6,6′-
dihydroxyl-bipyridine was followed. To a 25 mL round-bottom flask
equipped with a stir bar 6,6′-dimethoxy-4,4′-dimethyl-bipyridine
(0.244 g, 1.0 mmol) was added, followed by 5 mL of HBr/AcOH
(33 wt %). The final suspension was heated to reflux in a sand bath for
3 d. After it cooled, the white solid was collected by filtration, which was
washed with small amounts of water and acetone. Then the solid was
dispersed in 10 mL of boiling water and brought to neutral pH using a
KOH solution. The resultant solid was collected by filtration and dried.
(dop)2RuCl2. A procedure67 reported for the synthesis of
cis-[Ru(bpy)2Cl2] was adapted. A 100 mL oven-dried Schlenk flask
equipped with a stir bar was evacuated and refilled with nitrogen
three times. CODRuCl2 (0.081 g, 0.2896 mmol) and dop (0.138 g,
0.579 mmol, 2.0 equiv) were measured into the flask, followed by 15 mL
of anhydrous o-dichlorobenzene. The final dark brown suspension was
heated to reflux in a sand bath at 180 °C under N2 for 3 h. After it cooled,
the product was collected by suction filtration. The black powder was
washed with Et2O and then air-dried. Yield: 0.170 g, 0.262 mmol, 90.5%.
To remove potential [(dop)3Ru]Cl2: the crude product was stirred in
10 mL of DI water (200 mL/g of crude product) for a few hours. Then
the product was collected, washed with ether, and air-dried.
1
Yield: 0.1646 g, 0.761 mmol, 76.1%. H NMR (360 MHz, DMSO) δ
10.86 (s, 2H), 7.10 (s, 2H), 6.37 (s, 2H), 2.24 (s, 6H).
[(bpy)2Ru(dmdhbp)]Cl2 (4A). A procedure2 reported for the
synthesis of [(bipy)2Ru(6,6′dhbp)](PF6)2 was adapted. The whole
process was protected from light, as the final product was sensitive to
light. A 100 mL oven-dried Schlenk flask equipped with a stir bar was
charged with 10 mL of DI water with 10 mL of EtOH and was degassed
for 20 min. (bpy)2RuCl2 (0.1275 g, 0.261 mmol) and 4,4′-dimethyl-6,6′-
dihydroxy-2,2′-bipyridine (0.062 g, 0.2867 mmol, 1.1 equiv) were
measured into the flask. The final dark brown suspension was heated to
reflux in an oil bath at 110 °C under N2 for 24 h. After it cooled, the
solution was filtered to remove any excess dmdhbp ligand. To the
filtrate, a few drops of concentrated aqueous HCl were added. Water was
removed by rotary evaporation. The product was collected by suction
filtration with the help of diethyl ether. The brown-red powder was
washed with Et2O and then air-dried. The crude product was purified by
ether diffusion into an ethanol solution to crystallize the product. Yield:
0.1497 g, 0.2137 mmol, 65.4%. 1H NMR (500 MHz, DMSO) δ 11.76
(s, 2H), 8.77 (d, J = 8.2 Hz, 2H), 8.65 (d, J = 8.1 Hz, 2H), 8.13 (t, J = 7.9,
2H), 8.04 (s, 2H), 7.96 (m, 4H), 7.58 (t, J = 6.3 Hz, 2H), 7.53 (d, J =
5.4 Hz, 2H), 7.31 (t, J = 6.7 Hz, 2H), 6.56 (s, 2H), 2.36 (s, 6H). 13C
NMR (126 MHz, DMSO) δ 167.91 (s), 157.83 (s), 157.05 (s), 155.14
(s), 151.48 (s), 151.34 (s), 150.89 (s), 136.88 (s), 136.13 (s), 127.16 (s),
126.18 (s), 123.58 (s), 122.98 (s), 116.93 (s), 111.67 (s), 20.52 (s).
[(dop)2Ru(6,6′-dhbp)]Cl2 (3A). A procedure2 reported for the
synthesis of [(bipy)2Ru(6,6′dhbp)](PF6)2 was adapted. The whole
process was protected from light, as the final product is sensitive to light.
A 100 mL oven-dried Schlenk flask equipped with a stir bar was charged
with 5 mL of DI water with 5 mL of EtOH and was purged with nitrogen
for 20 min. (dop)2RuCl2 (0.089 g, 0.137 mmol) and 6,6′-dihydroxy-2,2′-
bipyridine (0.028 g, 0.151 mmol, 1.1 equiv) were measured into the
flask. The final dark brown suspension was heated to reflux in an oil bath
at 110 °C under N2 for 30 h. After it cooled, the solution was filtered to
remove any excess dhbp ligand. To the filtrate, a few drops of con-
centrated HCl were added. Water was removed by rotary evaporation.
The product was collected by suction filtration with the help of diethyl
ether. The brown-red powder was washed with Et2O and then air-dried.
1
Yield: 0.075 g, 0.0896 mmol, 65.4%. H NMR (500 MHz, DMSO) δ
11.89 (s, 2H), 8.67 (dd, J = 8.5, 1.2 Hz, 2H), 8.42 (dd, J = 8.3, 1.2 Hz,
2H), 8.30 (d, J = 5.3, 2H), 8.18 (d, J = 7.8 Hz, 2H), 7.93 (dd, J = 8.4,
5.2 Hz, 2H), 7.87 (t, J = 8.0 Hz, 2H), 7.60 (dd, J = 5.3, 1.2 Hz, 2H), 7.46
(dd, J = 8.4, 5.3 Hz, 2H), 6.81 (d, J = 8.3 Hz, 2H), 4.66 (d, J = 5.0 Hz, 8H).
13C NMR (126 MHz, DMSO) δ 168.53, 156.05, 150.84, 150.31, 143.86,
143.57, 140.01, 134.05, 133.73, 128.74, 127.76, 125.67, 124.81, 123.98,
123.56, 115.51, 111.71, 64.96, 64.91. Elem. Anal.: Found: C, 53.65; N,
ESI-MS (m/z): 315.1, Calcd for [(bpy)2Ru(dmdhbp)]2+
,
[C32H28N6O2Ru]2+: 315.1. Elem. Anal.: Found: C, 48.81; N, 10.67;
J
DOI: 10.1021/acs.inorgchem.7b01065
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
and H, 4.86%. Calc. for C32H28Cl2N6O2Ru.5H2O: C, 48.61; N, 10.63;
and H, 4.84%. pKa1: 5.6; pKa2: 7.2; average: 6.4.
these regions were unstable, so the data were treated using a solvent
mask as implemented in OLEX2.72−74 The procedure found two voids
of 600 Å3, each containing the equivalent of 150 e−, for approximately
four molecules of methanol per asymmetric unit. More details regarding
disorder and refinement details for all structures are given in the
Supporting Information.
Cell Culture. MDA-MB-231, MCF7, and MCF10A cell lines were
purchased from American Type Culture Collection (ATCC). MDA-
MB-231 and MCF7 of less than 10 passages were grown using DMEM
(Gibco, 21063−029) supplemented with 10% FBS (Gibco, 26140079).
MCF10A was cultured with MEBM (Lonza) supplemented with
MEGM kit according to manufacturer’s protocol (Lonza). Human
umbilical cord blood (UCB) and MOLM-13 cells were cultured in
IMDM supplemented with 20% FBS. UCB required SCF, IL-3, IL-6,
Flt-3L, and TPO growth factors at 10 ng mL−1.
IC50 Measurement. MDA-MB-231, MCF7, and MCF10A were
seeded at a density of 5000 cells per well in 100 μL of media in 96-well
plates and incubated for 24 h to adhere to plate. Compounds were
dissolved in DMSO and diluted in media to avoid cytotoxic effect from
DMSO on cells. Final concentration of DMSO was set to less than 1%
(v/v). Cells were treated with 100 μL of serially diluted compounds and
incubated for 48 h in the dark. After 48 h, cells were washed with Hank’s
Balanced Salt Solution (HBSS; Gibco) and irradiated for an hour with
blue light (Philips, goLITE BLU) in the dark. Cells were then provided
with 100 μL of fresh media per well and incubated overnight. Cytotoxic
effects of the compounds were measured using a Cell Counting Kit-8
according to the manufacturer’s protocol (Enzo Life Sciences). The IC50
of each cell line was determined using a Minitab 17.
[Ru(phen)2(4,4′-dhbp)][PF6]2 (5A). A round-bottom flask contain-
ing 30 mL of a 1:1 ethanol and water solution was degassed for 30 min.
A 0.5532 g (0.001 001 mol) sample of Ru(phen)2(Cl)2 and a 0.2251 g
(0.001 196 mol) sample of 4,4′-dhbp were added to the reaction flask
and heated at 80 °C under argon for 5 h. After the reaction was
completed, the solution was removed from heat and allowed to cool to
room temperature, then filtered to remove insoluble impurities. The
filtered solution was added to 150 mL of H2O. A few drops of
concentrated HCl was added to ensure protonation of the complex, and
an excess of NH4PF6 in 20 mL of water was added to precipitate the
orange complex. The complex was collected on a fritted flask and
washed with water, followed by ether. Crystals of [Ru(phen)2(4,4′-
dhbp)][PF6]2 for X-ray diffraction were grown by the slow diffusion
1
of ether into a methanol solution. Yield: 0.8431 g, 88%. H NMR
(300 MHz, CD3CN): δ 8.6 (d, 2H), 8.5 (d, 2H), 8.3 (d, 2H), 8.2 (m,
4H), 7.8 (m, 6H), 7.5 (dd, 2H), 7.2 (d, 2H), 6.7 (d, 2H). FT-IR (ATR,
cm−1): 3634 (w), 3087 (w), 2161 (w), 1975 (w), 1619 (m), 1577 (m),
1490 (m), 1449 (m), 1427 (m), 1412 (m), 1311 (m), 1204 (m), 1146
(m), 1095 (m), 1057 (m), 1026 (m), 972 (m), 823 (s), 719 (s), 555 (s).
Anal. Calcd for RuC34N6O2H24P2F12·1H2O: C, 42.65%; N, 8.78%; H,
2.74%. Found: C, 42.73%; N, 8.74%; H, 2.83%. Once the character-
ization of [Ru(phen)2(4,4′-dhbp)][PF6]2 was complete, a salt meta-
thesis was performed to isolate the chloride salt, [Ru(phen)2(4,4′-
dhbp)]Cl2, for cell studies.
[(bpy)2Ru(H2O)2]SO4 (6). Ru(bpy)2Cl2 (132.1 mg, 0.273 mmol)
was added to a flask and transferred to the glovebox, where Ag2SO4
(85 mg, 0.274 mmol) was added. Milli-Q water (6 mL) was added, and
the solution was stirred at room temperature for 18 h. The solution was
filtered and washed with DI H2O (3 × 10 mL). The filtrate was collected
and dried under vacuum to yield a maroon solid (136.0 mg, 91% yield).
1H NMR 360 MHz (D2O) δ 9.33 (d, 2H), 8.54 (d, 2H), 8.33 (d, 2H),
Immunocytochemistry. After irradiation with blue light as
described above, cells were fixed by 4% paraformaldehyde (PFA) and
permeabilized by Triton X-100 (Fisher Scientific). Cells were stained
with DAPI (1:5,000; Invitrogen). Ruthenium compounds were detected
by mCherry channel.
ICP-MS. Cells were treated with 10 μM of compound concentration
to analyze the uptake rate of each cell line for each compound. After
irradiation, the HBSSs were aspirated and washed with PBS. Nitric acid
(70%) was added to lysate the cells. The cell lysates were diluted with 5%
nitric acid in distilled water. The concentration of ruthenium was
analyzed using ICP-MS (PerkinElmer).
8.20 (t, 2H), 7.85 (t, 2H), 7.72 (m, 4H), 7.06 (t, 2H). MALDI-TOF MS
(m/z): 547.0, Calcd for [(bpy)2Ru(H2O)2SO4 + H+]+ 547.0,
C20H21N4O6SRu. Elem. Anal.: Found: C, 41.26; N, 9.68; and H, 3.97%.
Calc. for C20H20N4O6SRu·2H2O: C, 41.30; N, 9.63; and H, 4.16%.
Single-Crystal X-ray Diffraction Structure Determinations of
2A, 2B, 3A, 4B, 5A. Crystals of appropriate dimension were mounted on a
Mitgen cryoloop or glass filament in a random orientation. Preliminary
examination and data collection were performed on a Bruker ApexII
CCD-based X-ray diffractometer equipped with an Oxford N-Helix
Cryosystem low-temperature device and a fine focus Mo-target X-ray
tube (λ = 0.710 73 Å) operated at 2000 W power (50 kV, 40 mA). The
X-ray intensities were measured at low temperature (100(2) K or
173(2) K). The collected frames were integrated with the Saint70
software using a narrow-frame algorithm. Data were corrected for
absorption effects using the multiscan method in SADABS.71 The space
groups were assigned using XPREP of the Bruker ShelXTL72 package,
solved with ShelXT72 and refined with ShelXL72 and the graphical
interface ShelXle73 (for 2A, 2B, 3A, 4B) or OLEX2 (5A).73,74 All non-
hydrogen atoms were refined anisotropically. H atoms attached to
carbon were positioned geometrically and constrained to ride on their
parent atoms.
The structure of 2A was found to have two occupationally disordered
ethanol molecules, with two-component disorder at each site. The
structure of 2B was also found to have two occupationally disordered
ethanol molecules, and both of them are located on symmetry elements.
The structure of 3A was found to contain several regions of residual
density. Part of them was modeled as one occupationally disordered
ethanol molecule. Attempts to model the remaining residual density
were not successful, so the residual density was corrected for using a
reverse Fourier transform approach (the SQUEEZE procedure) using
the PLATON program. The solvent-accessible volume was found to be
1193 Å3. The electrons found in the solvent-accessible void is 294 e−,
which corresponds to approximately three ethanol molecules.
The structure of 4B was found to contain two occupationally disordered
ethanol molecules, with one of the molecules being threefold dis-
ordered. For structure 5A, large regions of residual density were again
associated with disordered solvent molecules. Attempted refinements of
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the ACS
Publications website at DOI: 10.1021/acs.inorgchem.7b01065.
1H, 13C NMR, IR, UV−vis spectra; methods and details of
thermodynamic acidity (pKa) measurements; quantum
yield measurement details; methodology for partition
coefficient determination; structural parameters for X-ray
crystallography (PDF)
Accession Codes
CCDC 1545429−1545433 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge via www.ccdc.cam.ac.uk/data_request/cif, or by emailing
data_request@ccdc.cam.ac.uk, or by contacting The Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge
CB2 1EZ, UK; fax: +44 1223 336033.
AUTHOR INFORMATION
Corresponding Authors
■
*E-mail: jared.paul@villanova.edu. (J.J.P.)
*E-mail: merinoed@ucmail.uc.edu. (E.J.M.)
*E-mail: ykim@eng.ua.edu. (Y.K.)
*E-mail: etpapish@ua.edu. (E.T.P.)
ORCID
Fengrui Qu: 0000-0002-9975-2573
K
DOI: 10.1021/acs.inorgchem.7b01065
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(13) Antonarakis, E. S.; Emadi, A. Ruthenium-based chemo-
therapeutics: are they ready for prime time? Cancer Chemother.
Pharmacol. 2010, 66, 1−9.
Nicholas A. Piro: 0000-0003-4219-0909
W. Scott Kassel: 0000-0002-6764-9045
Jared J. Paul: 0000-0003-0641-1671
(14) http://theralase.com/pressrelease/health-canada-approves-
clinical-trial-application-anti-cancer-drug/ (accessed 2/16/17).
(15) Fong, J.; Kasimova, K.; Arenas, Y.; Kaspler, P.; Lazic, S.; Mandel,
A.; Lilge, L. A novel class of ruthenium-based photosensitizers effectively
kills in vitro cancer cells and in vivo tumors. Photochem. Photobiol. Sci.
2015, 14, 2014−23.
Elizabeth T. Papish: 0000-0002-7937-8019
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
(16) Stephenson, M.; Reichardt, C.; Pinto, M.; Wachtler, M.;
̈
■
Sainuddin, T.; Shi, G.; Yin, H.; Monro, S.; Sampson, E.; Dietzek, B.;
McFarland, S. A. Ru(II) Dyads Derived from 2-(1-Pyrenyl)-1 H-
imidazo[4,5- f][1,10]phenanthroline: Versatile Photosensitizers for
Photodynamic Applications. J. Phys. Chem. A 2014, 118, 10507−10521.
(17) Shi, G.; Monro, S.; Hennigar, R.; Colpitts, J.; Fong, J.; Kasimova,
K.; Yin, H.; DeCoste, R.; Spencer, C.; Chamberlain, L.; Mandel, A.;
Lilge, L.; McFarland, S. A. Ru(II) dyads derived from α-
oligothiophenes: A new class of potent and versatile photosensitizers
for PDT. Coord. Chem. Rev. 2015, 282−283, 127−138.
We thank NSF CAREER for past support (Grant Nos. CHE-
0846383 and CHE-1360802 to E.T.P. and her group), the
Undergraduate Creativity and Research Academy at Univ. of
Alabama (UA), the Research Grants Committee at UA to E.T.P.
and Y.K., and NSF EPSCoR Track 2 Grant to E.T.P. and R.H.S.
for support (PI N. Hammer, Grant No. OIA-1539035), the UA
(E.T.P. and Y.K.), and Villanova Univ. (J.J.P.) for generous
financial support. E.J.M. and his group are supported by an
NIH grant (R21A185370) and U.S. Army Medical Research
(CA150055) grants. We also thank Q. Liang (UA) for MS
analysis. Finally, we thank the members of the Papish group for
assistance and suggestions, including S. E. Brown, K. Hughes,
and S. Reed for preliminary experiments on this project.
(18) Naik, A.; Rubbiani, R.; Gasser, G.; Spingler, B. Visible-Light-
Induced Annihilation of Tumor Cells with Platinum-Porphyrin
Conjugates. Angew. Chem., Int. Ed. 2014, 53, 6938−6941.
(19) Mari, C.; Huang, H.; Rubbiani, R.; Schulze, M.; Wurthner, F.;
̈
Chao, H.; Gasser, G. Evaluation of Perylene Bisimide-Based Ru IIand Ir
IIIComplexes as Photosensitizers for Photodynamic Therapy. Eur. J.
Inorg. Chem. 2017, 2017, 1−9.
REFERENCES
(20) Pierroz, V.; Rubbiani, R.; Gentili, C.; Patra, M.; Mari, C.; Gasser,
G.; Ferrari, S. Dual mode of cell death upon the photo-irradiation of a
RuII polypyridyl complex in interphase or mitosis. Chemical Science
2016, 7, 6115−6124.
■
(1) Dolmans, D. E. J. G. J.; Fukumura, D.; Jain, R. K. Photodynamic
therapy for cancer. Nat. Rev. Cancer 2003, 3, 380−387.
(2) Hufziger, K. T.; Thowfeik, F. S.; Charboneau, D. J.; Nieto, I.;
Dougherty, W. G.; Kassel, W. S.; Dudley, T. J.; Merino, E. J.; Papish, E.
T.; Paul, J. J. Ruthenium dihydroxybipyridine complexes are tumor
activated prodrugs due to low pH and blue light induced ligand release. J.
Inorg. Biochem. 2014, 130, 103−111.
(3) Papish, E. T.; Paul, J. J.; Merino, E. J. Ruthenium Complexes as
Tumor Activated Metallo-Prodrugs. 2014, Patent Application filed with
US Patent Office.
(21) Hidayatullah, A. N.; Wachter, E.; Heidary, D. K.; Parkin, S.;
Glazer, E. C. Photoactive Ru(II) Complexes With Dioxinophenanthro-
line Ligands Are Potent Cytotoxic Agents. Inorg. Chem. 2014, 53,
10030−10032.
(22) White, J. K.; Schmehl, R. H.; Turro, C. An overview of
photosubstitution reactions of Ru(II) imine complexes and their
application in photobiology and photodynamic therapy. Inorg. Chim.
Acta 2017, 454, 7−20.
(4) Denny, W. A. Prodrug strategies in cancer therapy. Eur. J. Med.
(23) Chari, R. Targeted delivery of chemotherapeutics: tumor-
activated prodrug therapy. Adv. Drug Delivery Rev. 1998, 31, 89−104.
(24) Zhou, L.; Cheng, R.; Tao, H.; Ma, S.; Guo, W.; Meng, F.; Liu, H.;
Liu, Z.; Zhong, Z. Endosomal pH-Activatable Poly(ethylene oxide)-
graft-Doxorubicin Prodrugs: Synthesis, Drug Release, and Biodistribu-
tion in Tumor-Bearing Mice. Biomacromolecules 2011, 12, 1460−1467.
(25) Denny, W. A. Tumor-activated ProdrugsA New Approach to
Cancer Therapy. Cancer Invest. 2004, 22, 604−619.
(26) Montcourrier, P.; Silver, I.; Farnoud, R.; Bird, I.; Rochefort, H.
Breast cancer cells have a high capacity to acidify extracellular milieu by a
dual mechanism. Clin. Exp. Metastasis 1997, 15, 382−392.
(27) Cardone, R. A.; Casavola, V.; Reshkin, S. J. The role of disturbed
pH dynamics and the Na+/H+ exchanger in metastasis. Nat. Rev. Cancer
2005, 5, 786−795.
(28) Gatenby, R. A.; Gillies, R. J. Why do cancers have high aerobic
glycolysis? Nat. Rev. Cancer 2004, 4, 891−899.
(29) Howerton, B. S.; Heidary, D. K.; Glazer, E. C. Strained Ruthenium
Complexes Are Potent Light-Activated Anticancer Agents. J. Am. Chem.
Soc. 2012, 134, 8324−8327.
(30) Wachter, E.; Heidary, D. K.; Howerton, B. S.; Parkin, S.; Glazer, E.
C. Light-activated ruthenium complexes photobind DNA and are
cytotoxic in the photodynamic therapy window. Chem. Commun. 2012,
48, 9649.
(31) Garner, R. N.; Gallucci, J. C.; Dunbar, K. R.; Turro, C. [Ru(bpy)
2(5-cyanouracil) 2] 2+as a Potential Light-Activated Dual-Action
Therapeutic Agent. Inorg. Chem. 2011, 50, 9213−9215.
(32) Arora, K.; White, J. K.; Sharma, R.; Mazumder, S.; Martin, P. D.;
Schlegel, H. B.; Turro, C.; Kodanko, J. J. Effects of Methyl Substitution
in Ruthenium Tris(2-pyridylmethyl)amine Photocaging Groups for
Nitriles. Inorg. Chem. 2016, 55, 6968−6979.
Chem. 2001, 36, 577−595.
(5) Stubbs, M.; McSheehy, P. M.; Griffiths, J. R.; Bashford, C. L. Causes
and consequences of tumour acidity and implications for treatment. Mol.
Med. Today 2000, 6, 15−19.
(6) Gaiddon, C.; Pfeffer, M. The Fate of Cycloruthenated
Compounds: From C-H Activation to Innovative Anticancer Therapy.
Eur. J. Inorg. Chem. 2017, 2017, 1−17.
(7) Flocke, L. S.; Trondl, R.; Jakupec, M. A.; Keppler, B. K. Molecular
mode of action of NKP-1339 − a clinically investigated ruthenium-based
drug − involves ER- and ROS-related effects in colon carcinoma cell
lines. Invest. New Drugs 2016, 34, 261−268.
(8) Hartinger, C. G.; Zorbas-Seifried, S.; Jakupec, M. A.; Kynast, B.;
Zorbas, H.; Keppler, B. K. From bench to bedside − preclinical and early
clinical development of the anticancer agent indazolium trans-
[tetrachlorobis(1H-indazole)ruthenate(III)] (KP1019 or FFC14A). J.
Inorg. Biochem. 2006, 100, 891−904.
(9) Trondl, R.; Heffeter, P.; Kowol, C. R.; Jakupec, M. A.; Berger, W.;
Keppler, B. K. NKP-1339, the first ruthenium-based anticancer drug on
the edge to clinical application. Chemical Science 2014, 5, 2925−2932.
(10) Bergamo, A.; Sava, G. Ruthenium anticancer compounds: myths
and realities of the emerging metal-based drugs. Dalton Trans. 2011, 40,
7817.
(11) Morbidelli, L.; Donnini, S.; Filippi, S.; Messori, L.; Piccioli, F.;
Orioli, P.; Sava, G.; Ziche, M. Antiangiogenic properties of selected
ruthenium(III) complexes that are nitric oxide scavengers. Br. J. Cancer
2003, 88, 1484−1491.
(12) Bencini, A.; Failli, P.; Valtancoli, B.; Bani, D. Low molecular
weight compounds with transition metals as free radical scavengers and
novel therapeutic agents. Cardiovasc. Hematol. Agents Med. Chem. 2010,
8, 128−146.
L
DOI: 10.1021/acs.inorgchem.7b01065
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(33) Li, A.; White, J. K.; Arora, K.; Herroon, M. K.; Martin, P. D.;
Schlegel, H. B.; Podgorski, I.; Turro, C.; Kodanko, J. J. Selective Release
of Aromatic Heterocycles from Ruthenium Tris(2-pyridylmethyl)amine
with Visible Light. Inorg. Chem. 2016, 55, 10−12.
Hydroxy Groups During Basic Electrocatalysis. Inorg. Chem. 2014, 53,
12689−12698.
(49) Keegstra, E. M. D.; Zwikker, J. W.; Roest, M. R.; Jenneskens, L. W.
A highly selective synthesis of monodisperse oligo(ethylene glycols). J.
Org. Chem. 1992, 57, 6678−6680.
(34) Loftus, L. M.; White, J. K.; Albani, B. A.; Kohler, L.; Kodanko, J. J.;
Thummel, R. P.; Dunbar, K. R.; Turro, C. New Ru IIComplex for Dual
Activity: Photoinduced Ligand Release and 1O 2Production. Chem. -
Eur. J. 2016, 22, 3704−3708.
(50) Paw, W.; Eisenberg, R. Synthesis, Characterization, and
Spectroscopy of Dipyridocatecholate Complexes of Platinum. Inorg.
Chem. 1997, 36, 2287−2293.
(51) Rice, C. R.; Guerrero, A.; Bell, Z. R.; Paul, R. L.; Motson, G. R.;
Jeffery, J. C.; Ward, M. D. New hybrid ditopic ligands containing fused
phenanthroline and crown ether units. New J. Chem. 2001, 25, 185−187.
(52) Dubreuil, D. M.; Pipelier, M. G.; Pradere, J. P.; Bakkali, H.;
Lepape, P.; Delaunay, T.; Tabatchnik, A. Pyridazine and pyrrole
compounds, processes for obtaining them and uses. WO2008012440A2,
2008.
(53) Ross, H. B.; Boldaji, M.; Rillema, D. P.; Blanton, C. B.; White, R. P.
Photosubstitution in Tris Chelate Complexes of Ruthenium(II)
Containing the Ligands 2,2′Bipyrazine, 2,2′-Bipyrimidine, 2,2′-Bipyr-
idine, and 4,4′-Dimethyl-2,2′-bipyridine: Energy Gap Control. Inorg.
Chem. 1989, 28, 1013−1021.
́
(35) Debreczeni, J. E.; Bullock, A. N.; Atilla, G. E.; Williams, D. S.;
Bregman, H.; Knapp, S.; Meggers, E. Ruthenium Half-Sandwich
Complexes Bound to Protein Kinase Pim-1. Angew. Chem., Int. Ed.
2006, 45, 1580−1585.
(36) Kastl, A.; Wilbuer, A.; Merkel, A. L.; Feng, L.; Di Fazio, P.; Ocker,
M.; Meggers, E. Dual anticancer activity in a single compound: visible-
light-induced apoptosis by an antiangiogenic iridium complex. Chem.
Commun. 2012, 48, 1863−1865.
(37) Streib, M.; Kraling, K.; Richter, K.; Xie, X.; Steuber, H.; Meggers,
̈
E. An Organometallic Inhibitor for the Human Repair Enzyme 7,8-
Dihydro-8-oxoguanosine Triphosphatase. Angew. Chem., Int. Ed. 2014,
53, 305−309.
(38) Liu, Z.; Sadler, P. J. Organoiridium Complexes: Anticancer Agents
and Catalysts. Acc. Chem. Res. 2014, 47, 1174−1185.
(54) Groutas, W. C.; Stanga, M. A.; Brubaker, M. J.; Huang, T. L.; Moi,
M. K.; Carroll, R. T. Substituted 2-Pyrones, 2-Pyridones, and Other
Congeners of Elasnin as Potential Agents for the Treatment of Chronic
Obstructive Lung Diseases? J. Med. Chem. 1985, 28, 1106−1109.
(55) Jacinto Demuner, A.; Moreira Valente, V. M.; Almeida Barbosa, L.
C.; Rathi, A.; Donohoe, T. J.; Thompson, A. L. Synthesis and Phytotoxic
Activity of New Pyridones Derived from 4-Hydroxy-6-Methylpyridin-
2(1H)-one. Molecules 2009, 14, 4973−4986.
(39) Zeng, L.; Chen, Y.; Huang, H.; Wang, J.; Zhao, D.; Ji, L.; Chao, H.
Cyclometalated Ruthenium(II) Anthraquinone Complexes Exhibit
Strong Anticancer Activity in Hypoxic Tumor Cells. Chem. - Eur. J.
2015, 21, 15308−15319.
(40) Rico Bautista, H.; Saavedra Díaz, R. O.; Shen, L. Q.; Orvain, C.;
Gaiddon, C.; Le Lagadec, R.; Ryabov, A. D. Impact of cyclometalated
ruthenium(II) complexes on lactate dehydrogenase activity and
cytotoxicity in gastric and colon cancer cells. J. Inorg. Biochem. 2016,
163, 28−38.
(56) Dutta, U.; Deb, A.; Lupton, D. W.; Maiti, D. The regioselective
iodination of quinolines, quinolones, pyridones, pyridines and uracil.
Chem. Commun. 2015, 51, 17744−17747.
(41) Sainuddin, T.; McCain, J.; Pinto, M.; Yin, H.; Gibson, J.; Hetu, M.;
McFarland, S. A. Organometallic Ru(II) Photosensitizers Derived from
π-Expansive Cyclometalating Ligands: Surprising Theranostic PDT
Effects. Inorg. Chem. 2016, 55, 83−95.
(42) Papish, E. T.; Taylor, M. T.; Jernigan, F. E., III; Rodig, M. J.;
Shawhan, R. R.; Yap, G. P. A.; Jove, F. A. Synthesis of Zinc, Copper,
Nickel, Cobalt, and Iron Complexes Using Tris(pyrazolyl)methane
Sulfonate Ligands: A Structural Model for N,N,O Binding in
Metalloenzymes. Inorg. Chem. 2006, 45, 2242−2250.
(57) Wachter, E.; Zamora, A.; Heidary, D. K.; Ruiz, J. x.; Glazer, E. C.
Geometry matters: inverse cytotoxic relationship for cis/trans-Ru(ii)
polypyridyl complexes from cis/trans-[PtCl2(NH3)2]. Chem. Commun.
2016, 52, 1−4.
(58) Tardito, S.; Bassanetti, I.; Bignardi, C.; Elviri, L.; Tegoni, M.;
Mucchino, C.; Bussolati, O.; Franchi-Gazzola, R.; Marchio, L. Copper
̀
Binding Agents Acting as Copper Ionophores Lead to Caspase
Inhibition and Paraptotic Cell Death in Human Cancer Cells. J. Am.
Chem. Soc. 2011, 133, 6235−6242.
(43) Jernigan, F. E., III; Sieracki, N. A.; Taylor, M. T.; Jenkins, A. S.;
Engel, S. E.; Rowe, B. W.; Jove, F. A.; Yap, G. P. A.; Papish, E. T.;
Ferrence, G. M. Sterically Bulky Tris(triazolyl)borate Ligands as Water-
Soluble Analogues of Tris(pyrazolyl)borate. Inorg. Chem. 2007, 46,
360−362.
(44) Nieto, I.; Livings, M. S.; Sacci, J. B.; Reuther, L. E.; Zeller, M.;
Papish, E. T. Transfer Hydrogenation in Water via a Ruthenium Catalyst
with OH Groups near the Metal Center on a bipy Scaffold.
Organometallics 2011, 30, 6339−6342.
(45) Dixon, N. A.; McQuarters, A. B.; Kraus, J. S.; Soffer, J. B.; Lehnert,
N.; Schweitzer-Stenner, R.; Papish, E. T. Dramatic tuning of ligand
donor properties in (Ttz)CuCO through remote binding of H+ (Ttz =
hydrotris(triazolyl)borate). Chem. Commun. 2013, 49, 5571−5573.
(46) Marelius, D. C.; Bhagan, S.; Charboneau, D. J.; Schroeder, K. M.;
Kamdar, J. M.; McGettigan, A. R.; Freeman, B. J.; Moore, C. E.;
Rheingold, A. L.; Cooksy, A. L.; Smith, D. K.; Paul, J. J.; Papish, E. T.;
Grotjahn, D. B. How Do Proximal Hydroxy or Methoxy Groups on the
Bidentate Ligand Affect [(2,2′;6′,2″-Terpyridine)Ru(N,N)X] Water-
Oxidation Catalysts? Synthesis, Characterization, and Reactivity at
Acidic and Near-Neutral pH. Eur. J. Inorg. Chem. 2014, 2014, 676−689.
(47) DePasquale, J.; Nieto, I.; Reuther, L. E.; Herbst-Gervasoni, C. J.;
Paul, J. J.; Mochalin, V.; Zeller, M.; Thomas, C. M.; Addison, A. W.;
Papish, E. T. Iridium Dihydroxybipyridine Complexes Show That
Ligand Deprotonation Dramatically Speeds Rates of Catalytic Water
Oxidation. Inorg. Chem. 2013, 52, 9175−9183.
(59) Kalaivani, P.; Prabhakaran, R.; Poornima, P.; Dallemer, F.;
Vijayalakshmi, K.; Padma, V. V.; Natarajan, K. Versatile Coordination
Behavior of Salicylaldehydethiosemicarbazone in Ruthenium(II)
Carbonyl Complexes: Synthesis, Spectral, X-ray, Electrochemistry,
DNA Binding, Cytotoxicity, and Cellular Uptake Studies. Organo-
metallics 2012, 31, 8323−8332.
(60) Graf, N.; Mokhtari, T. E.; Papayannopoulos, I. A.; Lippard, S. J.
Platinum(IV)-chlorotoxin (CTX) conjugates for targeting cancer cells.
J. Inorg. Biochem. 2012, 110, 58−63.
(61) Lovejoy, K. S.; Lippard, S. J. Non-traditional platinum compounds
for improved accumulation, oral bioavailability, and tumor targeting.
Dalton Trans. 2009, 10651.
(62) Trondl, R.; Heffeter, P.; Jakupec, M. A.; Berger, W.; Keppler, B. K.
NKP-1339, a first-in-class anticancer drug showing mild side effects and
activity in patients suffering from advanced refractory cancer. BMC
Pharmacol. Toxicol. 2012, 13, A82.
(63) Hong, Y.-R.; Gorman, C. B. Synthetic Approaches to an
Isostructural Series of Redox-Active, Metal Tris(bipyridine) Core
Dendrimers. J. Org. Chem. 2003, 68, 9019−9025.
(64) Connelly, N. G.; Geiger, W. E. Chemical Redox Agents for
Organometallic Chemistry. Chem. Rev. 1996, 96, 877−910.
(65) Shugar, G. J.; Bauman, R. S.; Shugar, R.; Bauman, L. Chemical
Technicians’ Ready Reference Handbook, 2nd ed.; McGraw-Hill Book
Company: New York, NY, 1981.
(66) Saitou, R.; Chung, K.-G.; Nishimura, H.; Sakai, T.; Komiya, S.;
Noda, N.; Nishiguchi, M. Method for producing ruthenium compound.
US20120101290A1, 2012.
(48) Gerlach, D. L.; Bhagan, S.; Cruce, A. A.; Burks, D. B.; Nieto, I.;
Truong, H. T.; Kelley, S. P.; Herbst-Gervasoni, C. J.; Jernigan, K. L.;
Bowman, M. K.; Pan, S.; Zeller, M.; Papish, E. T. Studies of the Pathways
Open to Copper Water Oxidation Catalysts Containing Proximal
(67) Norris, M. R.; Concepcion, J. J.; Glasson, C. R. K.; Fang, Z.;
Lapides, A. M.; Ashford, D. L.; Templeton, J. L.; Meyer, T. J. Synthesis of
M
DOI: 10.1021/acs.inorgchem.7b01065
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Phosphonic Acid Derivatized Bipyridine Ligands and Their Ruthenium
Complexes. Inorg. Chem. 2013, 52, 12492−12501.
(68) Payne, L. J.; Downham, R.; Sibley, G. E. M.; Edwards, P.; Davies,
G. M. 2-Oxo-2-(2-phenyl-5,6,7,8-tetrahydroindolizin-3-yl)acetamide
derivatives as antifungal agents and their preparation.
WO2008145963A1, 2008.
(69) Zhang, T.; Wang, C.; Liu, S.; Wang, J.-L.; Lin, W. A Biomimetic
Copper Water Oxidation Catalyst with Low Overpotential. J. Am. Chem.
Soc. 2014, 136, 273−281.
(70) Apex2, v2008.2−4 (2008) or Apex2, v2009.7−0 (2009) Bruker
Advanced X-ray Solutions; Bruker AXS Inc.: Madison, WI, 2009.
(71) SADABS: An Empirical Absorption Correction Program, v. 2008/1;
Bruker AXS: Madison, WI, 2009.
(72) Sheldrick, G. M. A short history of SHELX. Acta Crystallogr., Sect.
A: Found. Crystallogr. 2008, A64, 112−122.
(73) Dolomanov, O. V.; Bourhis, L. J.; Gildea, R. J.; Howard, J. A. K.;
Puschmann, H. OLEX2: a complete structure solution, refinement and
analysis program. J. Appl. Crystallogr. 2009, 42, 339−341.
(74) Sheldrick, G. M. Crystal structure refinement with SHELXL. Acta
Crystallogr., Sect. C: Struct. Chem. 2015, A71, 3−8.
(75) Cuello-Garibo, J.-A.; Meijer, M. S.; Bonnet, S. To cage or to be
caged? The cytotoxic species in ruthenium-based photoactivated
chemotherapy is not always the metal. Chem. Commun. 2017,
DOI: 10.1039/c7cc03469e.
N
DOI: 10.1021/acs.inorgchem.7b01065
Inorg. Chem. XXXX, XXX, XXX−XXX