Metalloprobes: Fluorescence imaging of multidrug resistance (MDR1) P-Glycoprotein (Pgp)-mediated functional transport activity in cellulo

Journal of Inorganic Biochemistry 159 (2016) 159–164

Contents lists available at ScienceDirect

Journal of Inorganic Biochemistry journal homepage: www.elsevier.com/locate/jinorgbio

Metalloprobes: Fluorescence imaging of multidrug resistance (MDR1) P-Glycoprotein (Pgp)-mediated functional transport activity in cellulo G.S.M. Sundaram a,⁎, Monica Sharma b,c, Daniel Kaganov a,c, Junsang Cho a,c, Scott E. Harpstrite a, Vijay Sharma a,b,c,d,⁎ a

ICCE Institute, Molecular Imaging Center, Mallinckrodt Institute of Radiology, United States Department of Neurology, Washington University School of Medicine, St. Louis, MO 63110, United States Students and Teachers As Research Scientists, Washington University School of Medicine, St. Louis, MO 63110, United States d Department of Biomedical Engineering, School of Engineering & Applied Science, Washington University, St. Louis 63105, United States b c

a r t i c l e

i n f o

Article history: Received 18 September 2015 Received in revised form 15 February 2016 Accepted 23 February 2016 Available online 26 February 2016

a b s t r a c t Radiolabeled metalloprobes offer sensitive tools for evaluating quantitative accumulation of chemical entities within pooled cell populations. Although beneficial in translational nuclear imaging, this method precludes interrogation of effects resulting from variations at a single cell level, within the same segment of cell population. Compared with radiotracer bioassays, fluorescence imaging offers a cost-efficient technique to assess accumulation of metalloprobes at a single cell level, and determine their intracellular localization under live cell conditions. To evaluate, whether or not radiotracer assay and fluorescence imaging provide complementary information on utility of metalloprobes to assess functional expression of P-glycoprotein (Pgp) on plasma membrane of tumor cells, imaging studies of fluorescent cationic Ga(III)-ENBDMPI (bis(3-ethoxy-2-hydroxy-benzylidene)-N,N′bis(2,2-dimethyl-3-amino-propyl)ethylenediamine) and its neutral counterpart Zn(II)-ENBDMPI are performed. While the uptake profiles of the cationic metalloprobe are inversely proportional to expression of Pgp in tumor cells, the accumulation profiles of the neutral Zn(II)-ENBDMPI in non-MDR and MDR cells are not significantly impacted. The cationic Ga(III)-ENBDMPI maps with Mito-Tracker Red, thereby confirming localization within mitochondria of non-MDR (Pgp−) cells. Depolarization of both plasmalemmal and mitochondrial potentials decreased retention of the cationic Ga(III)-ENBDMPI within the mitochondria. Additionally, LY335979, an antagonist-induced accumulation of the cationic Ga(III) metalloprobe in MDR (Pgp+) cells indicated specificity of the agent. Compared with traits of Ga(III)-ENBDMPI as a Pgp recognized substrate, Zn(II)-ENBDMPI demonstrated uptake in both MDR and non-MDR cells thus indicating the significance of overall molecular charge in mediating Pgp recognition profiles. Combined data indicate that live cell imaging can offer a cost-effective methodology for monitoring functional Pgp expression. © 2016 Published by Elsevier Inc.

1. Introduction Multidrug resistance (MDR) is a major obstacle to successful chemotherapeutic treatment of many solid and hematological tumors. Among numerous biochemical pathways, MDR1 P-glycoprotein (Pgp, ABCB1; a family member of ABC transporters) is among the best-characterized barriers of chemotherapeutic resistance. Pgp, a 170 kD plasma membrane protein, is predicted by sequence analysis to comprise two symmetrical halves that share both homology with a family of ATPbinding cassette (ABC) membrane transport proteins and a common ancestral origin with bacterial transport systems [1,2]. Characterized by 12 transmembrane domains and two nucleotide-binding folds ⁎ Corresponding authors at: Mallinckrodt Institute of Radiology, Washington University School of Medicine, Box 8225, 510 S. Kingshighway Blvd., St. Louis, MO 63110, United States. E-mail address: [email protected] (V. Sharma).

http://dx.doi.org/10.1016/j.jinorgbio.2016.02.022 0162-0134/© 2016 Published by Elsevier Inc.

[1,3], the protein is thought to hydrolyze ATP to affect outward transport of substrates across or off the cell surface membrane [1,4]. Although the specific protein domains and amino acids involved in substrate recognition continue to be characterized, genetic and biochemical evidence is conventionally interpreted to show putative membraneassociated domains interacting directly with selected cytotoxic agents to affect transport [5–8]. Regardless of mode of interaction, the net effect is reduction of the intracellular concentration of recognized substrate drugs in Pgp-expressing multidrug resistant cells compared with nonPgp-expressing counterparts. Additionally, Pgp has also been postulated to play an important role in the development of Aβ (β-amyloid) pathophysiology within the brain [9–11] as well as other neurodegenerative disorders [12]. Physiologically, Pgp is located in several tissues responsible for excretory functions, such as the brush border of proximal tubule cells in the kidney, the biliary surface of hepatocytes, and the apical surface of mucosal cells in the small intestine and the colon [13,14]. Finally, Pgp expression has also been demonstrated on the luminal surface of

160

G.S.M. Sundaram et al. / Journal of Inorganic Biochemistry 159 (2016) 159–164

endothelial cell lining capillaries in the brain and in the testis [15,16], as well as on the apical surface of choroid plexus epithelial cells [17]. With the emergence of system biology and critical biomarkers of various disease states, a priori analysis of tumor markers to assess the presence or absence of a particular molecular pathway or target (such as a key receptor or enzyme activity) for a given therapeutic agent could offer a great deal of insight into the potential outcome of a particular treatment. Thus identification of tumor markers with diagnostic agents could assist in stratification of patient populations likely to benefit from a given targeted therapy. To accomplish this objective, measurement of MDR is one potentially important marker in planning systemic therapy. However, expression of MDR1 Pgp, as detected at the level of messenger RNA or protein, does not always correlate with a functional assessment of Pgp-mediated transport activity. Because Pgp transport activity is also affected by specific mutations as well as the phosphorylation state of the protein [18–20], altered or less active forms of Pgp may be detected by polymerase chain reaction (PCR) or immunohistochemistry which do not accurately reflect the status of tumor cell resistance. Therefore, strategies for interrogating Pgp-mediated functionally transport activity have been sought [12,21]. Imaging with a radiopharmaceutical that is transported by Pgp could be beneficial in identifying noninvasively those tumors in which the transporter is not only expressed, but also functional. Thus, significant effort has been directed worldwide toward synthesis and preclinical validation of SPECT (Single Photon Emission Computed Tomography) and PET (Positron Emission Tomography) radiopharmaceuticals to allow noninvasive detection of transportermediated resistance. Typically, Pgp-recognized substrates are moderately hydrophobic cationic molecules under physiological conditions. Among metalloprobes, 99mTc-Sestamibi and 99mTc-Tetrofosmin are well-validated Pgp substrates [12,22,23]. Several 67/68Ga-metalloprobes have also been validated for monitoring Pgp-mediated activity in vitro and in vivo. Earlier, we identified a cationic and moderately hydrophobic Ga(III) radiopharmaceutical (67/68Ga-ENBDMPI; gallium(III)-(bis(3ethoxy-2-hydroxy-benzylidene)-N,N′-bis(2,2-dimethyl-3-aminopropyl)ethylenediamine) as a robust probe of Pgp-mediated functional transport activity in cellulo and in vivo [22]. The agent showed cellular uptake profiles inversely proportional to the expression of Pgp in human epidermal carcinoma cells. Additionally, the agent demonstrated pharmacokinetic profiles in vivo consistent with its sensitivity to detect Pgp-mediated functional transport [22]. Importantly, the agent has been hypothesized to penetrate non-MDR cells in response to favorable electrochemical gradients prevalent across plasma- and mitochondrialmembranes [24]. While its neutral Zn(II) counterpart (Zn-ENBDMPI; zinc(II)-(bis(3-ethoxy-2-hydroxy-benzylidene)-N,N′-bis(2,2-dimethyl3-amino-propyl)ethylenediamine)) did not show cytotoxicity profiles modified by either presence or absence of Pgp in human epidermal

carcinoma cells, thus implying indirectly that the neutral hydrophobic Zn(II) metalloprobe may not be recognized by Pgp as its transport substrate. These data indicate the importance of overall charge on the surface of the molecule for Pgp recognition profiles within this class of compounds. However, the lack of an appropriate commercially available Zn(II) radionuclide presented hurdles on tracer assays. Among the available radionuclides, 65Zn (t1/2 = 244 d) was not considered to be a good alternative due to significant concerns on radioactive waste management and its disposal, thus precluding direct studies of cellular uptake profiles under identical conditions. Furthermore, the intracellular localizations of cationic Ga(III)-ENBDMPI and its neutral counterpart Zn(II)ENBDMPI have not been directly ascertained. Herein, we evaluate the ability of both cationic Ga(III)-ENBDMPI and its neutral counterpart Zn(II)-ENBDMPI to serve as molecular imaging probes for ascertaining: a) Pgp-mediated functional transport activity in human epidermal carcinoma (MDR and non-MDR cells) either in presence or absence of LY335979, a potent and specific Pgp antagonist; b) intracellular localization of Ga(III) and Zn(II) metalloprobes within non-MDR cells under identical conditions; and c) influence of membrane potential on cellular accumulation of Ga(III) and Zn(II) metalloprobes within non-MDR cells, employing live cell fluorescence imaging. 2. Results and discussion Literature precedents indicate that multidentate ligands incorporated with an N4O2 donor core (within the coordination sphere) of the organic scaffold have the ability to form stable monomeric, monocationic, and moderately hydrophobic metalloprobes with a variety of main group and transition metals [22]. Within this class of agents, Ga(III)-ENBDMPI and Zn(II)-ENBDMPI (Fig. 1) are known to possess octahedral geometry [24,25], wherein the central metal cores are hexacoordinated, with two phenoxy oxygen atoms (O1 and O2), two secondary amine nitrogen atoms (N2 and N3), and two imine nitrogen atoms (N1 and N4). Recently, we have discovered unprecedented fluorescent characteristics of these metalloprobes. Following excitation at 365 nm, Ga(III)-ENBDMPI and Zn(II)-ENBDMPI show emission at 480 nm (Fig. 2). Although the emission peak is independent of the overall charge of the molecule under physiological conditions, the intensity of Zn(II)-ENBDMPI (at same concentration) peaks is lower than its cationic gallium counterpart. Additionally, the fluorescence output could also be pH dependent. Within the intracellular compartments, the lysosomal pH is acidic (4.5–4.8) compared with that of cytosol (7.2). The emission spectra of metalloprobes recorded in citric acid-sodium phosphate buffer (pH 4.0–5.0) do not show any significant impact on fluorescence intensity thus indicating retention of stable fluorescent traits following their penetration of intracellular compartments in cellulo. Therefore, we

Fig. 1. Chemical structures of a cationic Ga(III)-ENBDMPI (A) and its electrochemically neutral Zn(II)-ENBDMPI counterpart (B).

G.S.M. Sundaram et al. / Journal of Inorganic Biochemistry 159 (2016) 159–164

Fig. 2. Excitation and emission spectra of metalloprobes (100 μM) in PBS containing 1% ethanol. A. [Ga(III)-ENBDMPI]+; B. [Zn(II)-ENBDMPI]. Excitationmax: 365 nm; Emissionmax: 480 nm.

postulate that the modest fluorescent characteristics of these molecules could be exploited to track their intracellular localization via live cell imaging, and also correlate cellular uptake data reported earlier employing toxicity and radiotracer assays. While human epidermal non-MDR (KB-3-1) cells lack Pgp, their colchicine-derived MDR (KB-8-5) counterparts show immunodetectable levels of Pgp in human epidermal carcinoma cells, using monoclonal antibody C219 [26]. Multiple biochemical pathways, such as ion channels, receptor-mediated endocytosis, and diffusion across the bilayer could potentially mediate the cellular accumulation of Ga(III) and Zn(II) metalloprobes. However, the net cell content of many hydrophobic cationic tracers transported by Pgp has also been shown to be a function of both passive membrane potential dependent influx and transporter-mediated efflux. Accordingly, cationic Ga(III) metalloprobes exhibiting favorable Pgp recognition profiles would be expected to penetrate KB-3-1 (non-MDR, Pgp−) cells in response to inwardly directed electrochemical driving forces at the plasma membrane and

161

mitochondrial membrane and excluded from Pgp-expressing KB-8-5 (MDR, Pgp +) cells [23]. Conversely, neutral Zn(II) metalloprobes would be expected to permeate both non-MDR and MDR cells, and thus unlikely to differentiate between drug resistant (MDR) and drug sensitive cells (non-MDR). For the live cell imaging studies, human epidermal carcinoma KB-31 (non-MDR) and KB-8-5 (MDR) cells were incubated with Ga(III)ENBDMPI at increasing concentrations (Fig. 3) for 30 min in media either in the absence or presence of LY335979, a potent and specific Pgp inhibitor [23,27]. Following incubation with Ga(III)-ENBDMPI, KB3-1 cells were visualized using a Nikon high resolution microscope. Based upon initial results of concentration dependent output of the fluorescence signal (data not shown), 20 μM of Ga(III)-ENBDMPI was selected as an optimal concentration for live cell fluorescence imaging (Fig. 3, Right Panel). Following permeation, the cationic metalloprobes have been postulated to localize within mitochondria of non-MDR cells [28]. To directly evaluate intracellular localization, Ga(III)ENBDMPI was also incubated in non-MDR cells in the presence or absence of Mito-Tracker® (Mito-Tracker® Red CMH2XRos), a wellknown fluorescent marker for staining mitochondria. Both Ga(III)ENBDMPI and Mito-Tracker® show localization within the mitochondria of non-MDR cells (Fig. 4). Conversely, Mito-Tracker® penetrated both non-MDR and MDR cells to localize within mitochondria under these conditions. To evaluate whether or not the cationic metalloprobe mimics membrane permeant monovalent cationic fluorophores, such as rhodamine 123, which accumulates electrophoretically within nonMDR cells in response to favorable electrochemical gradients across plasma membrane and mitochondria, Ga(III) metalloprobe was analyzed for its ability to monitor depolarization of the membrane potential. Literature precedents indicate that plasmalemmal potential ranges from − 30 to − 100 mV and mitochondrial potential ranges from − 120 to − 160 mV. Since these potentials are additive, the mitochondria are −150 to −260 mV more negative than extracellular space. Employing a membrane potential at equilibrium for uptake of cationic molecule using the Nernst equation, this could mean several thousand fold higher concentrations of a metalloprobe within mitochondria compared to the extracellular space. Therefore, uptake profiles of Ga(III)-ENBDMPI in non-MDR cells (Fig. 4) using live cell imaging are also consistent with data of other lipophilic cations exemplified by rhodamine dyes [28]. Of note, uptake of the metalloprobe in KB-3-1 cells is a net function of plasma- and mitochondrial-potentials at the equilibrium. For assessing contribution of mitochondrial potential to uptake of the Ga(III) metalloprobe, it was incubated in non-MDR KB-3-1 cells in the presence or absence of carbonyl cyanide 3-chlorophenylhydrazone (CCCP, 5 μM), a potent uncoupler of electron transport chain, and known to readily dissipate mitochondrial potential. Following depolarization of the mitochondrial potential with CCCP, substantial release of Ga(III) metalloprobe from mitochondria was observed (Fig. 5, Middle Panel). Additionally, incubation of the agent in the presence of a higher concentration of CCCP (10 μM) did not show any significant further release of the agent from mitochondria via decrease in fluorescence. The residual fluorescence in mitochondria under these conditions indicates uptake

Fig. 3. Cellular accumulation of Ga(III)-ENBDMPI. Fluorescence imaging in live human epidermal carcinoma non-MDR (KB-3-1, Pgp−) cells following incubation of the metalloprobe for 30 min. KB3-1 cells without probe (Left), KB3-1 cells with Ga(III)-ENBDMPI (10 μM; Middle), and KB3-1 cells with Ga(III)-ENBDMPI (20 μM; Right).

162

G.S.M. Sundaram et al. / Journal of Inorganic Biochemistry 159 (2016) 159–164

Fig. 4. Cellular accumulation of Ga(III)-ENBDMPI and detection of its location via correlation with Mito-Tracker. Fluorescence imaging in live human epidermal carcinoma non-MDR (KB-31, Pgp−) cells following 30 min treatment with Ga(III)-ENBDMPI (20 μM) (Left) and Mito-Tracker® (25 nM) (Middle). Ga(III)-ENBDMPI localizes to mitochondria as shown through fairly close co-registration (Right).

of the agent within the mitochondria of non-MDR cells, which is independent of negative transmembrane electrochemical gradients present at mitochondrial membrane (Fig. 5). Literature precedents also indicate that incubation of MCF-7 cells in the presence of valinomycin (25 nM), a potassium ionophore, and high concentration potassium buffer (120 mM K+/20 mM Cl−) have been known to depolarize both plasma- and mitochondrial-potentials to nearly zero [28]. Under these conditions, a further decreased retention of Ga(III)-ENBDMPI in the intracellular compartments of non-MDR (KB 3-1) cells was observed (Fig. 5, Right Panel). Overall, the residual retention in non-MDR cells under plasma- and mitochondrial-potential depolarization conditions could be attributed to a slightly higher hydrophobicity of Ga(III)ENBDMPI compared with that of 99mTc-Sestamibi [22,29], an extremely versatile probe of membrane potential. Finally, combined live cell imaging data correlate exceptionally well with indirect radiotracer bioassays reported earlier [24,29]. Following incubation of the Ga(III)-ENBDMPI in MDR KB-8-5 cells, a strikingly different profile was observed. While Ga(III)-ENBDMPI demonstrated rapid accumulation within non-MDR (KB-3-1) Pgplacking cells (Figs. 3 and 4), the metalloprobe was rapidly excluded from MDR (KB-8-5) cells (Fig. 6, Top Right Panel), thus indicating cellular uptake profiles of Ga(III)-ENBDMPI are inversely proportional to Pgp expression, under these conditions. Additionally, incubation of Ga(III)-ENBDMPI in MDR cells in the presence of LY335979, a potent, a highly specific MDR1 Pgp antagonist, also demonstrated rapid accumulation of the metalloprobe within MDR cells (Fig. 6, Bottom Right Panel). Importantly, neither media nor LY335979 (controls) indicated any background fluorescence under these conditions (Fig. 6, Top and Bottom Left Panels). Overall, the presence of LY335979-induced accumulation in MDR cells indicates specificity of the metalloprobe to monitor Pgp-mediated functional transport activity in cellulo. Zn(II)-ENBDMPI is slightly less fluorescent compared with its cationic Ga(III) counterpart, thus a higher concentration (30 μM) has been used for live cell fluorescence imaging. To assess the ability of neutral Zn(II)-ENBDMPI to differentiate between non-MDR and MDR cells, the cells were treated with the agent under identical conditions, employed for cationic Ga(III)-ENBDMPI. The agent showed

accumulation in both non-MDR (Fig. 7A) and MDR cells (Fig. 7B). While the Ga(III) metalloprobe localizes in the mitochondria (Fig. 4), the neutral Zn(II) counterpart shows localization in mitochondria and cytoplasm. Overall, the cellular uptake profiles of neutral Zn(II)ENBDMPI in MDR and non-MDR cells are not mediated by expression of Pgp (Fig. 7). Earlier, we have shown that Ga(III) and Zn(II) metalloprobes are stable under physiological conditions [25]. Therefore, uptake profiles of both cationic metalloprobe and its surrogate neutral counterpart are mediated by parental compounds. Although mechanism(s) of Zn(II) metalloprobe uptake need to be ascertained, and are independent of Pgp expression, the uptake profiles of cationic Ga(III) metalloprobe are inversely proportional to Pgp expression in human epidermal carcinoma cells. Combined data indicate the importance of overall charge in mediating Pgp recognition within this class of metalloprobes. 3. Conclusions Both Ga(III) and Zn(II) metalloprobes have nearly identical chemical structures with octahedral geometry yet differing only in the overall electrochemical charge. While Ga(III)-ENBDMPI is recognized by Pgp as its efficient transport substrate, its surrogate neutral Zn(II)-ENBDMPI is not a Pgp substrate. Like other lipophilic cationic dyes, Ga(III) metalloprobe penetrated non-MDR cells to localize within mitochondria and demonstrated good correlation with Mito-Tracker®, thereby confirming localization of the probe within the mitochondria of KB-31 cells. Furthermore, the release of Ga(III)-ENBDMPI from the mitochondria following depolarization of the electrochemical gradients across plasma- and mitochondrial-membranes indicated its ability to serve as a probe of the membrane potential. Additionally, LY-inducible uptake of Ga(III)-ENBDMPI in MDR (Pgp+) cells indicates target specificity of the probe. These data also confirm earlier reports of their Pgp recognition profiles using cytotoxicity and radiotracer bioassays. Combined data indicate the fluorescent characteristics of selected analogs within this class of metalloprobes can be employed as a versatile, time-efficient, and cost-effective tool to probe Pgp-mediated functional transport in cellulo. Further improvization of the template scaffold

Fig. 5. Impact of depolarization of membrane potential on cellular accumulation of Ga(III)-ENBDMPI. Fluorescence imaging demonstrating depolarization of electrochemical gradients across plasma- and mitochondrial membranes in live human epidermal carcinoma non-MDR (KB-3-1, Pgp−) cells following 30 min treatment with Ga(III)-ENBDMPI (20 μM) in the absence (Left) or presence of CCCP (5 μM) (Middle) or CCCP (5 μM) and valinomycin (25 nM) in high potassium buffer (120 mM K+/20 mM Cl−) (Right).

G.S.M. Sundaram et al. / Journal of Inorganic Biochemistry 159 (2016) 159–164

163

Fig. 6. Inhibitor (LY335979)-induced cellular accumulation of Ga(III)-ENBDMPI in MDR (Pgp+) cells. Fluorescence imaging showing cellular uptake of Ga(III)-ENBDMPI in live human epidermal carcinoma MDR (KB-8-5, Pgp+) cells following 30 min treatment with Ga(III)-ENBDMPI (20 μM) in the absence (Top Right) or presence (Bottom Right) of a Pgp-inhibitor LY335979 (1 μM). Compared to controls (media, Top Left) or LY335979 alone (media, Bottom Left), the inhibitor (LY335979)-induced accumulation of the metalloprobe in Pgp expressing MDR cells indicate specificity of the agent.

can also enable development of highly fluorescent molecular imaging agents for interrogating the role of Pgp on the surface of tumor cells and pathophysiology studies of mitochondrial myopathies. 4. Materials and methods 4.1. General methods 4.1.1. Synthesis of the gallium(III) and zinc(II) metalloprobes Ga(III)-ENBDMPI and Zn(II)-ENBDMPI were synthesized using ligand exchange reactions involving a Schiff base ligand (2-(2-hydroxy-

3-ethoxyphenyl)-1,3-bis[4-aza-5-(2′-hydroxy-3′-ethoxyphenyl)2″,2‴dimethyl-but-4′-ene-1′-yl]-1,3-imidazolidine (H33-Eabi) and either Ga(III) acetylacetonate or Zn(II) acetylacetonate in ethanol using methods described earlier and characterized using routine analytical methods [24,25]. 4.1.2. Chemical characterization of Ga(III)-ENBDMPI 13 N NMR (75.4 MHz, DMSO-d6) δ: 0.90 (s, 6H), 0.95 (s, 6H), 1.40 (t, 6H), 2.60 (m, 2H), 2.70–2.98 (m, 6H), 3.60 (dd, 2H), 3.78 (t, 2H), 4.02 (q, 4H), 4.90 (bs, 2H), 6.60 (t, 2H), 6.88 (dd, 2H), 7.02 (d, 2H), 8.19 (s, 2H); 13C NMR (75.4 MHz, CDCl3) δ: 14.8, 21.9, 26.1, 35.5, 47.6,

Fig. 7. Cellular accumulation of Zn(II)-ENBDMPI in non-MDR (Pgp−) and MDR (Pgp+) cells. A: Fluorescence imaging in live human epidermal carcinoma non-MDR (KB-3-1, Pgp−) cells following 30 min treatment with Zn(II)-ENBDMPI] (30 μM) (Left) and Mito-Tracker® (25 nM) (Middle). Zn(II)-ENBDMPI does not localize to mitochondria as shown through lack of co-registration (Right). B: Fluorescence imaging in live human epidermal carcinoma MDR (KB-8-5, Pgp+) cells following 30 min treatment with Zn(II)-ENBDMPI] (30 μM). Note Zn(II)-ENBDMPI penetration in both non-MDR (Pgp−) and MDR (Pgp+) cells.

164

G.S.M. Sundaram et al. / Journal of Inorganic Biochemistry 159 (2016) 159–164

59.0, 63.2, 68.9, 115.9, 116.2, 118.8, 125.3, 150.4, 157.2, 170.5; MS(FAB) for [C30H44N4O4Ga]+ m/z = 594.1.

4.5. Assessment of effects of plasma- and mitochondrial-potentials on uptake and retention of Ga(III) metalloprobe in non-MDR human epidermal carcinoma cells

4.1.3. Chemical characterization of Zn(II)-ENBDMPI 13 N NMR (75.4 MHz, DMSO-d6) δ: 0.73(s, 6H), 0.94 (s, 6H), 1.34 (t, 6H), 2.41 (m, 4H), 2.58–2.80 (m, 6H), 3.62–3.81 (m, 2H), 3.82–3.95 (m, 4H), 4.05–4.15 (m, 2H), 6.09 (t, 2H), 6.53 (dd, 2H), 6.63 (d, 2H), 7.68 (s, 2H); 13C NMR (75.4 MHz, CDCl3) δ: 15.2, 22.3, 27.3, 35.7, 48.8, 59.8, 62.8, 70.6, 109.0, 113.8, 120.1, 125.9, 151.5, 162.8, 165.4; MS(FAB) for [C30H45N4O4Zn] m/z = 588.2; Found: (M + H): 589.2; (M + Li): 595.2.

50,000 cells (KB-3-1) were seeded onto borosilicate 8-well chambered slides 1 d prior to the assay. Cells were then incubated with Ga(III)-ENBDMPI (20 μM), either with variable concentrations of CCCP (5 μM, 10 μM, 15 μM) in MEBSS buffer or carbonyl cyanide 3chlorophenylhydrazone (CCCP, 5 μM) and valinomycin (25 nM) in high potassium buffer (120 mM K+/20 mM Cl−), into each well in an eight-chamber slide and incubated for 30 min at 37 °C. Following incubation, images were acquired using the filter sets (DAPI excitation; FITC and Texas red emission) as indicated.

4.2. Tissue culture Monolayers of cells were grown at 37 °C under a 5% CO2 atmosphere. Human epidermal carcinoma non-MDR KB-3-1 (Pgp −) cells were grown in media (DMEM (GIBCO, 11965) supplemented with heatinactivated fetal bovine serum (10%) and L-glutamine (2 mM)). Human epidermal carcinoma colchicine-derived MDR KB-8-5 cells (Pgp +) were grown in media supplemented with colchicine (25 nM).

Acknowledgments Authors thank Prof. David Piwnica-Worms (MD Anderson, University of Texas, Houston, TX) for helpful discussions. Financial assistance to this work was provided by grants from the National Institutes of Health (NIH) in part by RO1 HL111163 (VS), R33 AG033328 (VS), and American Health Assistance Foundation (A2007-383; VS). References

4.3. Accumulation of gallium(III) and zinc(II) metalloprobes in live human epidermal carcinoma cells using fluorescence imaging For evaluation of cellular transport assays via microscopy, human epidermal carcinoma KB-3-1 (non-MDR, Pgp −) and KB-8-5 (MDR, Pgp +) cells were plated onto borosilicate 8-well chambered (25,000 cells/chamber) slides (Nalge Nunc International, Rochester, NY) and allowed to grow to approximately 70% confluence at 37 °C under 5% CO2 atmosphere. Prior to imaging, media were replaced with imaging-media (phenol-red free DMEM (Gibco 31053) supplemented with heat-inactivated fetal bovine serum (10%) and L-glutamine (2 mM)). KB-3-1 (non-MDR, Pgp −) and KB-8-5 (MDR, Pgp +) cells were incubated with either Ga(III) or Zn(II) metalloprobes (20 μM in imaging-media containing 0.1% DMSO) for 30 min at 37 °C under a continuous influx of 5% CO2 atmosphere either in absence or presence of LY335979 (Zosuquidar trihydrochloride, 1 μM), a potent and specific Pgp inhibitor. Cellular accumulation of the metalloprobe was assessed using a Nikon Ti-E PFS inverted microscope equipped with a Nikon 60 × 0.3 NA Plan APO objective (oil), Prior H117 ProScan flat top linear encoded stage, and Prior Lumen 200PRO illumination system with standard DAPI and FITC filter sets. Images were acquired using a Photometrics CoolSNAP HQ2 digital camera and MetaMorph Microscopy and Imaging Analysis Software (version 7.7.0.0, Molecular Devices). Images were processed and analyzed using the ImageJ software package (NIH).

4.4. Assessment of intracellular localization of gallium(III) and zinc(II) metalloprobes in live human epidermal carcinoma cells A stock solution of Mito-Tracker® (Mito-Tracker® Red CMH2XRos, Life technologies, M7512) (1 μM) in DMSO was diluted to a final concentration of 25 nM in imaging-media and kept at 37 °C for 30 min, prior to experiments. Literature precedents indicate that higher concentrations (50 nM) of Mito-Tracker® can potentially impair mitochondrial function. Therefore, all experiments were conducted using 25 nM MitoTracker®. Cells were incubated with either Ga(III)-ENBDMPI (20 μM) or Zn(II)-ENBDMPI (30 μM), and Mito-Tracker® (25 nM) for 30 min and then visualized using appropriate filter sets (metalloprobes: DAPI excitation and FITC emission; Mito-Tracker®: Texas red).

[1] [2] [3] [4] [5] [6] [7] [8] [9]

[10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29]

M. Gottessman, I. Pastan, Annu. Rev. Biochem. 62 (1993) 385–427. G. Ames, Annu. Rev. Biochem. 55 (1986) 397–425. P. Gros, Y. Ben Neriah, J.M. Croop, D.E. Housman, Nature 323 (1986) 728–731. P. Gros, R. Dhir, J. Croop, F. Talbot, Proc. Natl. Acad. Sci. 88 (1991) 7289–7293. S. Ambudkar, S. Dey, C. Hrycyna, M. Ramachandra, I. Pastan, M. Gottesman, Annu. Rev. Pharmacol. Toxicol. 31 (1999) 361–398. E.P. Bruggemann, U.A. Germann, M.M. Gottesman, I. Pastan, J. Biol. Chem. 264 (1989) 15483–15488. A. Yoshimura, Y. Kuwazuru, T. Sumizawa, S. Ikeda, M. Ichikawa, T. Usagawa, S. Akiyama, Biochim. Biophys. Acta 992 (1989) 307–314. Y. Raviv, H. Pollard, E. Bruggemann, I. Pastan, M. Gottesman, J. Biol. Chem. 265 (1990) 3975–3980. J.R. Cirrito, R. Deane, A.M. Fagan, M.L. Spinner, M. Parsadanian, M.B. Finn, H. Jiang, J.L. Prior, A. Sagare, K.R. Bales, S.M. Paul, B.V. Zlokovic, D. Piwnica-Worms, D.H. Holtzman, J. Clin. Invest. 115 (2005) 3285–3290. B. Zlokovic, Trends Neurosci. 28 (2005) 202–208. R. Deane, Neuron 43 (2004) 333–344. J. Sivapackiam, S.T. Gammon, S.E. Harpstrite, V. Sharma, Methods Mol. Biol. 596 (2010) 141–181. F. Thiebaut, T. Tsuruo, H. Hamada, M.M. Gottesman, I. Pastan, M.C. Willingham, Proc. Natl. Acad. Sci. U. S. A. 84 (1987) 7735–7738. R.N. Hitchins, D.H. Harman, R.A. Davey, D.R. Bell, Eur. J. Cancer Clin. Oncol. 24 (1988) 449–454. C. Cordon-Cardo, J.P. OBrien, D. Casals, G.L. Rittman, J.L. Biedler, M.R. Melamed, J.R. Bertino, Proc. Natl. Acad. Sci. U. S. A. 86 (1989) 695–698. J.M. Croop, M. Raymond, D. Haber, A. Devault, R.J. Arceci, P. Gros, D.E. Housman, Mol. Cell. Biol. 9 (1989) 1346–1350. V. Rao, J. Dahlheimer, M. Bardgett, A. Snyder, R. Finch, A. Sartorelli, D. PiwnicaWorms, Proc. Natl. Acad. Sci. U. S. A. 96 (1999) 3900–3905. M.M. Gottesman, I. Pastan, Annu. Rev. Biochem. 62 (1993) 385–427. U.A. Germann, M.C. Willingham, I. Pastan, M.M. Gottesman, Biochemistry (Mosc) 29 (1990) 2295–2303. S. Ahmad, A.R. Safa, R.I. Glazer, Biochemistry (Mosc) 33 (1994) 10313–10318. L. Homolya, Z. Hollo, U.A. Germann, I. Pastan, M.M. Gottesman, B. Sarkadi, J. Biol. Chem. 268 (1993) 21493–21496. V. Sharma, Bioconjug. Chem. 15 (2004) 1464–1474. J. Sivapackiam, S.E. Harpstrite, J.L. Prior, H. Gu, N.P. Rath, V. Sharma, Dalton Trans. 39 (2010) 5842–5850. V. Sharma, J. Prior, M. Belinsky, G. Kruh, D. Piwnica-Worms, J. Nucl. Med. 46 (2005) 354–364. S.E. Harpstrite, J.L. Prior, J. Sivapackiam, S.D. Collins, N.P. Rath, V. Sharma, Med. Chem. 6 (2010) 191–199. S.E. Harpstrite, H. Gu, R. Natarajan, V. Sharma, Nucl. Med. Commun. 35 (2014) 1067–1070. A. Dantzig, R. Shepard, J. Cao, K. Law, W. Ehlhardt, T. Baughman, T. Bumol, J. Starling, Cancer Res. 56 (1996) 4171–4179. S. Davis, M. Weiss, J. Wong, T. Lampidis, L. Chen, J. Biol. Chem. 260 (1985) 13844–13850. V. Sharma, A. Beatty, S.-P. Wey, J. Dahlheimer, C. Pica, C. Crankshaw, L. Bass, M. Green, M. Welch, D. Piwnica-Worms, Chem. Biol. 7 (2000) 335–343.