Effect of Structural Modifications to Glyoxal-bis(thiosemicarbazonato

J. Med. Chem. , 2018, 61 (3), pp 711–723. DOI: 10.1021/acs.jmedchem.7b01158. Publication Date (Web): December 12, 2017. Copyright © 2017 American C...
0 downloads 4 Views 6MB Size
Article Cite This: J. Med. Chem. 2018, 61, 711−723

pubs.acs.org/jmc

Effect of Structural Modifications to Glyoxalbis(thiosemicarbazonato)copper(II) Complexes on Cellular Copper Uptake, Copper-Mediated ATP7A Trafficking, and P‑Glycoprotein Mediated Efflux Karla M. Acevedo,†,‡ David J. Hayne,‡,§ Lachlan E. McInnes,‡,§ Asif Noor,‡,§ Clare Duncan,† Diane Moujalled,† Irene Volitakis,∥ Angela Rigopoulos,# Kevin J. Barnham,§,∥,⊥ Victor L. Villemagne,∇ Anthony R. White,*,†,○ and Paul S. Donnelly*,‡,§ †

Department of Pathology, ‡School of Chemistry, §Bio21 Institute, ∥Florey Institute of Neuroscience and Mental Health, and Department of Pharmacology and Therapeutics, University of Melbourne, Melbourne, Victoria 3010, Australia # Olivia Newton-John Cancer Research Institute, Heidelberg, Victoria 3084, Australia ∇ Centre for PET, Department of Molecular Imaging & Therapy, Austin Health, 145 Studley Road, Heidelberg, Victoria 3084, Australia ⊥

S Supporting Information *

ABSTRACT: Bis(thiosemicarbazonato)copper(II) complexes are of interest as potential therapeutics for cancer and neurodegenerative diseases as well as imaging agents for positron emission tomography (PET). The cellular uptake of six bis(thiosemcarbazonato)copper(II)complexes derived from glyoxal, with different functional groups Cu(gtsx) where x = different functional groups, was investigated in SKOV-3, HEK293, and HEK293 P-gp cell lines. Treatment of the cells with the copper complexes increased intracellular copper and increased levels of p-ERK due to activation of the Ras-RafMEK-ERK pathway. Treatment of SKOV-3 cells with low concentrations (μM) of two of the copper complexes led to trafficking of the endogenous copper transporter ATP7A from the Golgi network to the cell membrane. Experiments in HEK293 and HEK293-P-gp cells suggest that Cu(gtsm) and Cu(gtse) are substrates for the P-gp efflux protein but the complex with a pyrrolidine functional group, Cu(gtspyr), is not. A PET experiment in mice showed that [64Cu]Cu(gtspyr) has reasonable brain uptake but high liver uptake.



INTRODUCTION The biological activity of copper(II) complexes of bis(thiosemicarbazone) ligands (Cu(btsc)) has led to them being investigated as potential anticancer agents, antimicrobials, and neurotherapeutics. Treatment of a sarcoma 180 mouse model with the metal free bis(thiosemicarbazone), glyoxalbis(N-4-methyl-3-thiosemicarbazone) H2gtsm (Figure 1), led to a reduction in tumor weight when compared to untreated controls.1 In this early investigation, the authors recognized “a mechanism of action involving inactivation or translocation of metal ions was possible”. The analogous bis(thiosemicarbazones) derived from 2,3-butanedione, featuring two methyl substituents on the backbone, such as H2atsm (Figure 1) were inactive in this model. Bis(thiosemicarbazones) derived from 1,2-diones, such as glyoxal, form charge neutral complexes with CuII with the ligand acting as a dianionic tetradentate N2S2 donor. The CuII complexes are stable (log KA ≈ 18) and relatively lipophilic, and selected examples, such as Cu(gtsm) (Figure 1), have the ability to cross cell membranes and even the blood−brain barrier.2 The CuII complexes can be reduced to CuI complexes that are less stable than the CuII complexes. The CuII/I reduction potential is remarkably sensitive to the substituents © 2017 American Chemical Society

on the backbone of the ligand as illustrated by the dramatically different reduction potentials for Cu(gtsm) (E0′ = −0.52 V vs SCE) and Cu(atsm) (E0′ = −0.65 V vs SCE) when measured in dimethylformamide.2−5 The inactivity of the ligands derived from 2,3-butanedione, H2atsm, in the early studies against sarcoma 180 in Swiss mice was possibly due to the lower reduction potential of Cu(atsm) making it harder to release bioavailable copper.1 It has been suggested that the reducing environment encountered inside cells is sufficient to reduce Cu(gtsm) to less stable CuI species that are susceptible to dissociation from the ligand making the copper “bioavailable” to known copper binding proteins.5,6 Copper is an essential trace nutrient necessary for many enzymes and cellular processes, and its cellular biochemistry is carefully controlled by an array of transport and chaperone proteins.7,8 The Menkes coppertranslocating P-type ATPase (ATP7A) is a copper transport protein central to systemic copper absorption that uses the energy of ATP hydrolysis to transport copper from the cytosol into the lumen of the secretory pathway where copper is Received: August 9, 2017 Published: December 12, 2017 711

DOI: 10.1021/acs.jmedchem.7b01158 J. Med. Chem. 2018, 61, 711−723

Journal of Medicinal Chemistry

Article

Figure 1. Chemical structures and abbreviations for Cu(btsc) complexes.

incorporated into various copper requiring enzymes.7 In polarized cells, the ATP7A protein resides in the trans-Golgi network and responds to increases in cellular copper concentrations by relocating to the plasma membrane.9 Treatment of the human neuroblastoma cell line BE(2)-M17 with the preformed copper complex, Cu(gtsm), resulted in cell cycle arrest that was not associated with the onset of apoptosis. Analysis of the treated cells by protein microarray techniques revealed that Cu(gtsm) rapidly and potently reduced cyclin D1 expression, while increasing Kip2 expression decreased Cdk7 expression, activated CHK2, and resulted in a potent decrease of total and phosphorylated insulin-like growth factor receptor (IGF-IR).10 Treatment of the transgenic adenocarcinoma of mouse prostate (TRAMP) model with preformed Cu(gtsm) (2.5 mg/kg) significantly reduced prostate cancer burden (70%) and severity (grade) but caused mild kidney toxicity in the mice, associated primarily with interstitial nephritis and luminal distention.11 In a separate study, treatment of a HCT116 xenograft mouse model with another Cu(btsc) complex derived from glyoxal, glyoxal-bis(4-methyl-4-phenyl3-thiosemicarbazonato)copper(II), inhibited tumor growth by 95 ± 3.9% when compared to control mice.12 We have also investigated bis(thiosemicabazonato)copper(II) (Cu(btsc)) complexes as potential therapeutics for neurodegenerative diseases.13−16 Treatment of an amyloid mouse model (APP/PS1) of relevance to Alzheimer’s disease with Cu(gtsm) resulted in cognitive improvement associated with altered amyloid metabolism and microtubule tau protein phosphorylation mediated through modulation of kinase signaling including phosphatidylinositol 3-kinase (PI3K), glycogen synthase kinase 3β (GSK3β), and extracellular signal regulated kinase (ERK).6,17 There are several positron-emitting isotopes of copper that are of interest in the development of copper based positron emission tomography (PET) imaging agents. Complexes of bis(thiosemicarbazones) formed with radioactive copper isotopes have been investigated as myocardial perfusion and hypoxia tracers.2,18−22 Both [64Cu]Cu(gtsm) and [64Cu]Cu(atsm) cross the blood−brain barrier in mice, and there is a significant increase in the uptake of [64Cu]Cu(gtsm) in a transgenic mouse model of amyloid pathology (APP/PS1) when compared to control animals (3.0 ± 0.25% ID/g and 1.58 ± 0.14% ID/g, respectively), possibly due to altered copper metabolism in the animal model of the disease. In addition, in the APP/PS1 model, [64Cu]Cu(gtsm) showed 40% higher

brain retention than 64Cu(atsm).23 A similar difference in brain uptake of copper following administration of [64Cu]Cu(gtsm) was observed in an alternative transgenic amyloid model (TASTPM) when compared to wild-type control.24 The blood−brain barrier consists of a series of endothelial cells with tight junctions that control the passage of chemicals from the blood into the brain tissue. It is generally accepted that only certain small, MW < 600 Da, relatively lipophilic (log P of 0.9−3.0) molecules are capable of readily crossing the blood−brain barrier and it has been estimated that only 2% of central nervous system drug discovery compounds can cross the blood−brain barrier.25 The blood−brain barrier also contains several transporters that facilitate uptake or efflux and introduce yet further complications. Efflux transporters that are essential to the function of the blood−brain barrier include P-glycoprotein (P-gp) and breast cancer resistant protein (BCRP) and both can impede entry of imaging agents into the brain.26 P-gp is widely expressed and found in most barrier and excretory tissues including liver, kidney, intestine, testes and placenta. In the brain, P-gp is also found in astrocytes, microglia, neurons, and the luminal side of brain capillary endothelial cells.27 In mice, P-gp is partially responsible for clearance of amyloid-β from the brain, and patients with mild Alzheimer’s disease have compromised P-gp activity.28,29 Overexpression of P-gp also plays a role in the multidrug resistance (MDR) of some human cancers and correlates with poor prognosis. There is an inverse relationship between P-gp expression and the efficacy of chemotherapy drugs, with high expression being a predictor of poor prognosis in patients with advanced disease.30−33 Despite the importance of P-gp mediated efflux to chemotherapeutic drugs and imaging agents intending to target the brain, the precise molecular interactions that confer P-gp efflux remain poorly defined. There is a paucity of published examples where P-gp efflux had been purposefully circumvented. Functional groups that favor binding to P-gp are typically those adorned with hydrogen bond donors.34 The effect of P-gp expression on the cellular uptake and retention of the perfusion tracer 64Cu(ptsm) and the hypoxia tracer 64Cu(atsm) was investigated in MES-SA (P-gp negative) and MESSA/Dx5 (P-gp positive) cells under ambient oxygen conditions to investigate the influence of P-gp independent of hypoxia or perfusion. MES-SA/Dx5 (P-gp positive) cells showed lower retention of both tracers in a time-dependent fashion and more rapid efflux.35 A structurally related N-heterocyclic thiosemicarbazone, di-2-pyridyl ketone 4,4-dimethyl-3-thiosemicarba712

DOI: 10.1021/acs.jmedchem.7b01158 J. Med. Chem. 2018, 61, 711−723

Journal of Medicinal Chemistry

Article

zone, is a substrate for P-gp.36 To our knowledge the interactions between P-gp and Cu(btsc) derived from glyoxal remain to be investigated. In this manuscript we substantiate the claim that Cu(btsc) complexes derived from glyoxal, Cu(gtsx) where x = different functional groups, ethyl (e), phenyl (p) and pyrolidine (pyr), cross cell membranes and release bioavailable copper inside cells by monitoring copper-induced trafficking of ATP7A between the trans-Golgi network and the plasma membrane in SKOV3 cells. We also investigate the effects of P-gp expression on cellular uptake of Cu(gtsx) complexes in a HEK293 cell line that expresses the P-gp protein. A series of six complexes were synthesized; three compounds with different “N4” substituents but each retaining a NH-R functional group (potential hydrogen bond donor) were compared to three N4-dialkyl derivatives that lack a potential H-bond donor in this position (Figure 1). It was hoped that we could use this information to help us predict derivatives with potential to be more effective in MDR tumors and derivatives with better blood−brain barrier penetration to provide superior PET imaging agents to enable study of copper metabolism in the brain.



pellets by ICP-MS following a 2 h treatment with the complex (1 μM) (Figure 3). Measurements were based on the copper

RESULTS

Cu(gtsx) Complexes Have Quasi-Reversible CuII/I Reduction Potentials. The redox chemistry of each complex was investigated by cyclic voltammetry in dimethylformamide. Each of the six complexes displayed quasi-reversible reductions assigned to essentially metal based CuII/I processes. The CuII/I reduction potentials for the complexes with either one or two alkyl substituents on the N4-position were all similar ranging from Eo′ = 0.49 V for Cu(gtsm2) to Eo′ = 0.53 V for Cu(gtspyr) (vs SCE where ferrocene/ferricenium (FeII/III) Eo′ = 0.45 V in dimethylformamide) and consistent with previous reports (Figure 2).3−5 The aromatic functional group in Cu(gtsp) results in a significant shift in the CuII/CuI couple to Eo′ = 0.31 V (vs SCE). Cell Uptake and P-gp Recognition of Cu(gtsx) Complexes Is Dependent on Their Substituents. The cellular uptake of Cu(gtsx) complexes was investigated in SKOV3 cells by measuring the concentration of Cu in cell

Figure 3. Intracellular Cu levels following treatment with Cu(gtsx) compounds in SKOV3 ovarian carcinoma cells. SKOV3 cells were treated with Cu(gtsx) complexes (1 μM) or vehicle (DMSO) for 2 h before cell pellets were collected and metal levels measured by inductively coupled plasma mass spectrometry (ICPMS). The total Cu detected in pellets was compared to total protein in sample (Cu μmol L−1 μg−1). Values are the mean ± SEM. Values are compared to those of Cu(gtsm) as the most studied PET imaging agent in this class of compound.23,24 P values were calculated by one-way ANOVA followed by Tukey’s multiple comparison test: (∗) P < 0.05; (∗∗∗∗) P < 0.0001; n = 2.

concentration (μmol/L) and the total protein in the cell pellet (Cu μmol L−1 μg−1). Cell viability was not compromised following Cu(gtsx) treatment under these conditions as indicated by both 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays and lactate dehydrogenase (LDH) assays (see Supporting Information, Figure S1). Treatment with Cu(gtsm) led to the greatest increase in intracellular Cu when compared to Cu(gtsm2) [P < 0.0001], Cu(gtse) [P < 0.0001], and Cu(gtse2) [P < 0.0001] (Figure 3). From this series of compounds treatment with Cu(gtse2) resulted in the lowest intracellular Cu concentrations, with no statistically significant increase in Cu levels in comparison to vehicle (DMSO) control. Treatment with Cu(gtsm) and Cu(gtspyr) led to relatively high intracellular copper concentrations, significantly (P < 0.0001) higher than treatment with much higher concentrations of CuCl2 (300 μM) (Figure 3). Treatment of Cells with Cu(gtsm) and Cu(gtspyr) Increased p-ERK Levels and Induced Cellular Trafficking of ATP7A. Treatment with Cu(gtsx) complexes results in activation of the PI3K-Akt-GSK3, ERK, and JNK cell signaling cascades, and measuring increases to cellular levels of p-ERK has proved to be a reasonable indirect prediction of increased bioavailable copper.6,17,37 In SKOV-3 cells there is a general trend for all Cu(gtsx) complexes to increase p-ERK and p-Akt following a 2 h treatment (1 μM) (Figure 4A). However, only Cu(gtsm) showed a statistically significant increase in both pERK and p-Akt levels as determined by densitometry analysis (P < 0.01; n = 3) (Figure 4Aii and iii). Treatment with

Figure 2. Representative cyclic voltammograms showing CuII/I process for Cu(gtsm) and Cu(gtspyr) in dimethylformamide, with 0.1 M tetrabutylammonium hexafluorophosphate electrolyte, scan rate 0.1 V s−1, glassy carbon working electrode, potentials quoted versus SCE (where ferrocene/ferricenium E0′ = 0.45 V in dimethylformamide). 713

DOI: 10.1021/acs.jmedchem.7b01158 J. Med. Chem. 2018, 61, 711−723

Journal of Medicinal Chemistry

Article

Figure 4. Copper bioavailability following treatment with Cu(gtsx) complexes. (A) Western blot analysis to investigate the levels of p-ERK, p-Akt and P-gp following 2 h treatment with Cu(gtsx) complexes. Treatment with both Cu(gtsm) and Cu(gtse) led to a decrease in endogenous P-gp expression; P < 0.01 and P < 0.05, respectively (Aii). Treatment with all Cu(gtsx) complexes increased the levels of p-ERK and p-Akt in comparison to the DMSO control. Graphs show densitometric analysis of Western blot images, and the phosphoprotein levels were normalized to total protein 714

DOI: 10.1021/acs.jmedchem.7b01158 J. Med. Chem. 2018, 61, 711−723

Journal of Medicinal Chemistry

Article

Figure 4. continued and GAPDH loading controls (n = 3). (B) To test whether Cu(gtsm) or Cu(gtspyr) induced the copper-responsive trafficking of ATP7A, SKOV3 cells were treated with 1 μM complex for 2 h followed by co-staining with ATP7A (red channel), golgin-97 (green), F-actin (gray), and the nuclear stain DAPI (blue). Results show that in vehicle-only conditions ATP7A colocalizes mainly with golgin-97. CuCl2, Cu(gtsm), and Cu(gtspyr) promote the trafficking of ATP7A toward the cell periphery. Scale = 10 μm. Inset highlights the Golgi region.

Figure 5. Intracellular Cu levels in HEK293 vector and P-gp expressing cells following treatment with Cu(gtsx) over a period of time. (A) Cu levels in HEK293 cell lines were measured by ICPMS at 0.5, 1, 2, and 4 h following treatment with 1 μM Cu(gtsx) complexes. Linear graphs indicate the level of Cu (μmol L−1 μg−1) at each time point for both vector alone HEK293 and P-gp expressing cells. (B) Column graphs comparing the levels of Cu following 2 h treatment between the six Cu(gtsx) complexes for vector cells and P-gp expressing cells. Values are the mean ± SEM. P values were calculated by two-way ANOVA followed by Sidak’s multiple comparison’s test: (∗) P < 0.05; (∗∗) P < 0.01; (∗∗∗) P < 0.001; (∗∗∗∗) P < 0.0001; n = 3.

715

DOI: 10.1021/acs.jmedchem.7b01158 J. Med. Chem. 2018, 61, 711−723

Journal of Medicinal Chemistry

Article

Figure 6. Inhibiting P-gp increased uptake of Cu(gtsm) in HEK293 stably expressing P-gp. HEK293 expressing P-gp were incubated with Cu(gtsm) (1 μM) for (A) 2 h in the presence or absence of cyclosporin A (cyA) (5 μM) or (B) 24 h in the presence of elacridar (0.2 μM). Cu levels were tested by ICPMS. Values are the mean ± SEM: (∗) P < 0.05; n = 3.

Figure 7. p-ERK and p-Akt levels following treatment with Cu(gtsx) in HEK293 P-gp cells. HEK293 expressing P-gp cells were treated with 1 μM Cu(gtsx) complexes as well as Cu:histidine as a control. (A) Protein levels (P-pg, p-ERK, total ERK, p-Akt, total Akt) were analyzed by Western blot analysis. (B) Column graph representing the fold-increase of p-ERK levels relative to DMSO control, as calculated by densitometry analysis. The pERK levels were normalized to total ERK and GAPDH loading control (n = 3). Values are the mean ± SEM: (∗∗) P < 0.01; (∗∗∗) P < 0.001; (∗∗∗∗) P < 0.0001; n = 3.

Cu(gtspyr) also led to an increase in p-Akt levels (P < 0.01, n = 3). Treatment with both Cu(gtsm) and Cu(gtspyr) resulted in the highest increases in intracellular Cu following 2 h treatment. Interestingly, treatment with both Cu(gtsm) and Cu(gtse) treatment led to a decrease in endogenous P-gp expression; P < 0.01 and P < 0.05, respectively (Figure 4Ai). In addition, Western blot analysis showed a shift in the mobility/ molecular weight of P-gp when treated with Cu(gtsm) and Cu(gtse), indicative of post-translational modifications such as phosphorylation and glycosylation. Increased intracellular Cu promotes trafficking of the Cu transporter ATP7A from the Golgi network to the cell membrane for efflux. To further assess whether Cu is released from complexes once they are inside cells, the trafficking of endogenous ATP7A following treatment with Cu(gtsm) or Cu(gtspyr) was investigated in SKOV3 cells. The trafficking of ATP7A was assessed using confocal immunofluorescence microscopy with antibodies raised against the amino-terminal

590 amino acids of ATP7A. Treatment with both Cu(gtsm) or Cu(gtspyr) (1 μM) induced significant trafficking of ATP7A from the Golgi to a wider diffuse cellular distribution (Figure 4B) consistent with a dramatic increase of intracellular and bioavailable copper. It was necessary to treat with much higher concentrations of CuCl2 (300 μM) to achieve similar degree of ATP7A trafficking.38−40 The Golgi was co-stained with golgin97 (a trans-Golgi network resident protein; green channel), and the cell periphery was identified by staining filamentous actin (F-actin) with rhodamine phalloidon (gray channel). Cross sections (X−Z) show that some of the ATP7A reaches the membrane following Cu(gtsm) and Cu(gtspyr) (1 μM), and this is comparable to treatment with 300-fold higher concentrations of CuCl2 (300 μM). The Golgi appears green rather than yellow in the cells treated with Cu(gtsm) due to the lack of colocalization of ATP7A with Golgi marker in these cells when compared to DMSO control (Figure 4B; refer to insets and cross-section panels). 716

DOI: 10.1021/acs.jmedchem.7b01158 J. Med. Chem. 2018, 61, 711−723

Journal of Medicinal Chemistry

Article

Figure 8. (a) Representative microPET/MRI images of [64Cu]Cu(gtsm) and (b) [64Cu]Cu(gtspyr) at 9 min postinjection. (c) Liver and brain uptake of [64Cu]Cu(gtspyr).

Cu(gtsm) and Cu(gtse) Are Substrates of P-Glycoprotein. To assess the effect of P-gp on the uptake of Cu(gtsx) complexes, HEK293 expressing human P-gp and vector control cells (lower P-gp levels) were used. The P-gp transfected cell line expresses approximately 12-fold higher levels of P-gp compared to vector cells (see Supporting Information, Figure S2). Importantly, the P-gp expressed by HEK293 cells was functionally active as shown by rhodamine uptake/efflux assays (see Supporting Information, Figure S3). To investigate the uptake of Cu(gtsx) complexes (1 μM), Cu levels in cell pellets were measured following 0.5, 1, 2, and 4 h of treatment. At each time point treatment with Cu(gtsm) resulted in the highest intracellular copper concentrations when compared to Cu(gtse2), Cu(gtsm2), Cu(gtsp), and Cu(gtspyr) (Figure 5A). The cellular copper concentrations following treatment with Cu(gtse) were comparable to Cu(gtsm) particularly after 2 h (0.18 μmol L−1 μg−1 vs 0.17 μmol L−1 μg−1) and 4 h (0.2 μmol L−1 μg−1 vs 0.16 μmol L−1 μg−1) treatment. However, the uptake of both Cu(gtsm) and Cu(gtse) is significantly reduced at all times points in HEK293 P-gp cells when compared to vector control cells (n = 3; P < 0.001), suggesting these complexes are substrates of P-gp (Figure 5A and Figure 5B). In contrast, treatment of Cu(gtsp), Cu(gtsm2), Cu(gtse2) led to significantly lower levels of Cu uptake in comparison to Cu(gtsm) and Cu(gtse) in HEK293 vector cells. Cu levels did not vary greatly between P-gp expressing and vector cells following Cu(gtsp) and Cu(gtsm2) treatment. In the instance of Cu(gtse2), there was no statistical difference in Cu levels between cell lines. Interestingly, treatment with Cu(gtspyr) resulted in moderate uptake, which was not largely reduced by P-gp expression. In addition, Cu(gtsm) and Cu(gtspyr) did not affect P-gp activity as measured by P-gp functional assays using rhodamine uptake and flow cytometry (see Figure S4). Treatment with the P-gp Inhibitors Cyclosporin A and Elacridar Resulted in an Increase in Cellular Accumulation of Cu after Treatment with Cu(gtsm). To investigate the role of P-gp in the reduced uptake of Cu(gtsm) in HEK293 P-gp, the levels of intracellular copper were measured following treatment with of Cu(gtsm) (1 μM) for 2 h in the presence and absence of the P-gp inhibitors, cyclosporin A (Figure 6A), and elacridar (Figure 6 B).36 Both P-gp inhibitors led to increased

uptake of Cu(gtsm) in HEK293-P-gp cells to levels comparable to vector cells (Figure 6). Treatment with Cu(gtsx) Complexes Lead to an Increase in p-ERK but Not p-Akt in HEK293-P-gp Cells. An increase in p-ERK, as measured by densitometry analysis, was detected for five complexes following treatment of HEK293 P-gp cells with Cu(gtsx) (1 μM) (Figure 7A). There was no statistical significant increase in p-ERK when cells were treated with Cu(gtsm2) consistent with relatively low levels of intracellular copper, as measured by ICPMS. In contrast, treatment with Cu(gtsm) led to the greatest increase in p-ERK (∼6-fold increase; Figure 7B) in comparison to vehicle treated cells. Treatment with all complexes did not lead to a change in p-Akt levels, suggesting this signaling pathway is not stimulated in HEK293-P-gp cells in contrast to SKOV3 cells, highlighting subtle cell-line dependent differences. [64Cu]Cu(gtspyr) Crosses the Blood−Brain Barrier but Also Has a High Liver Uptake. Radioactive [64Cu]Cu(gtsm) has been used to probe copper metabolism in the brain in animal models of relevance to the pathology of Alzheimer’s disease.23,24 From this series of compounds Cu(gtspyr) was selected for further investigations as a potential PET imaging agent as it displayed good membrane permeability but, unlike Cu(gtsm), is not a substrate for P-gp. The biodistribution of [64Cu]Cu(gtspyr) was investigated in wild type mice using small animal PET/MRI imaging and compared to a representative scan of [64Cu]Cu(gtsm) (Figure 8). Wild-type balb-c mice were administered with [64Cu]Cu(gtspyr) (n = 3) via intravenous tail vein injection, and dynamic PET imaging over 30 min was used to assess biodistribution. The initial brain uptake of [64Cu]Cu(gtspyr) was 1.02 ± 0.04% IA/g at 4 min postinjection (mpi), and this activity did not clear over the image acquisition time (0.9 ± 0.01% IA/g at 29 mpi). In our previous studies with [64Cu]Cu(gtsm), the brain uptake in wild-type mice was 1.58 ± 0.14% IA/g.23 The uptake in the liver of [64Cu]Cu(gtspyr), 10.34 ± 2.47% IA/g at 4 mpi, and qualitative inspection of the PET/MRI images (Figure 8) suggest a higher liver uptake for [64Cu]Cu(gtspyr) than for [64Cu]Cu(gtsm). 717

DOI: 10.1021/acs.jmedchem.7b01158 J. Med. Chem. 2018, 61, 711−723

Journal of Medicinal Chemistry



Article

DISCUSSION A small library of glyoxal(bisthiosemicarbazonato)copper(II) complexes with different substituents in the N4 position was prepared to investigate cell permeability and P-gp mediated efflux. Functional groups featuring hydrogen-bond donors tend to favor binding to P-gp, so we altered the N4 functional group by removing one N−H hydrogen bond donor to give complexes that contain two alkyl groups at the N4 position (Cu(gtsm2), Cu(gtse2), and Cu(gtspyr)).34 Substantiation of the assertion that once glyoxal(bisthiosemicarbazone)copper(II) complexes enter cells that they are reduced to Cu(I) complexes in the intracellular reducing environment leading to release of the metal ion from the ligand leading to an increase in bioavailable copper was also sort. Altering the substituents at the N4 position had little effect on the CuII/I reduction potential, for example, for Cu(gtsm2) Eo′ = −0.49 V was marginally easier to reduce than Cu(gtsm) E0′ = −0.52 V. Treatment of SKOV-3 cells with the different complexes did result in different levels of intracellular copper, with Cu(gtsm), Cu(gtse), and Cu(gtspyr) being the most effective at increasing intracellular copper levels whereas Cu(gtse2) was the least effective. There has been increased interest in the role of copper in the activating Ras-Raf-MEKERK signal transduction pathway that is central to cell survival, proliferation, and migration. Aberrant signaling through this pathway can lead to either cell death or uncontrolled growth and cancer. In this instance, the interest in measuring activation of the RAS/MAPK pathway was an indirect measure of intracellular, bioavailable copper. The copper transporter Ctr1 is responsible for delivering copper into cells, and in cultured cells reduction in intracellular copper due to loss of function of Ctr1 leads to impaired activation of Ras/MAPK signaling and reduced phosphorylation of ERK (extracellular signal-regulated kinases). Copper is important to the Mek1/2 dependent phosphorylation of ERK1/2, and Mek1 binds CuII with appreciable affinity.41,42 Treatment of SKOV-3 cells with the Cu(gtsx) complexes led to increased levels of p-ERK consistent with increased intracellular and bioavailable copper. Copper ions can also inhibit the activity of protein-tyrosine phosphatases such as PTP1B that also play an important role in modulating phosphorylation based signal transduction pathways.37 Tracking intracellular labile intracellular copper with new copper responsive fluorescent imaging probes is providing increased insight into copper biology.43,44 Tracking copper that becomes bound to high affinity copper proteins following uptake by cells remains a challenge.45,46 An increase in “bioavailable” copper in SH-SY5Y cells following treatment with Cu(gtsm) (500 nM) could be detected using a transcription-based copper biosensor that takes advantage of the capacity of cells to activate metallothionein genes in response to elevated cytosolic copper availability. In this system, increased copper concentration in the cytosol leads to displacement of zinc from metallothioneins. The released zinc binds to and activates MTF-1 (MTF = metal responsive transcription factor), which subsequently induces gene expression by binding to MREs in the promoter region of specific genes cloned upstream of the firefly luciferase leading to the formation of the MRE-luciferase reporter.47,48 Here we demonstrate that treatment of SKOV-3 cells with relatively low concentrations of Cu(gtsm) and Cu(gtspyr) (1 μM) leads to significant trafficking of the endogenous copper transporter

ATP7A from the Golgi network to give a diffuse cytosolic distribution of the copper trafficking protein and some distinct localization in the cell membrane. Much higher concentrations of CuCl2 (300 μM) were required to achieve a comparable amount of ATP7A trafficking. Initiation of this intrinsic cellular response of trafficking an endogenous protein provides strong evidence for the release of copper from the bis(thiosemicarbazone) ligand. Previously, we investigated the mechanisms of cellular accumulation of Cu(gtsm), Cu(gtse), and Cu(gtsp) in human neuronal (M17) and glial (U87MG) cell lines, and these studies indicated that the complexes were taken into these cells by combination of passive and facilitated (protein-carriermediated) processes.45 A complicating factor to these studies was the rapid efflux of the copper delivered by the complexes through active mechanisms, but it was not known if the copper was released in ionic form through ATP7A or as an intact metal complex. In this study, we investigated the potential of the important P-gp efflux protein in modulating the uptake of Cu(gtsx) complexes. The effect of P-gp on the uptake of this series of Cu(gtsx) complexes was assessed by comparing cellular levels of copper in HEK293 and HEK293 P-gp cells that have 12-fold increase in P-gp expression. The two complexes that result in the highest intracellular concentrations of copper in HEK293 cells are Cu(gtsm) and Cu(gtse), 0.18 μmol L−1 μg−1 and 0.17 μmol L−1 μg−1, at 2 h after a treatment with 1 μM. The intracellular concentrations of copper are dramatically lower in HEK293 P-gp cells following the same treatment (Figure 5A). These data strongly suggest that Cu(gtsm) and Cu(gtse) are substrates for P-gp, and this was further substantiated by the fact that repeating the treatment of the cells in the presence of a P-gp inhibitor, cyclosporin A, led to an increase in the intracellular copper concentrations to levels that were comparable to the control HEK293 cells. The complex with an aromatic phenyl functional group, Cu(gtsp), was less effective at increasing intracellular copper than Cu(gtsm) and Cu(gtse), and it is notable that there is little difference in uptake between the HEK293 and HEK293 P-gp cells. It is possible that the addition of the lipophilic phenyl functional group leads to increased retention in the cell membrane and less accumulation in the cytosol.45 It has been suggested that functional groups that contain potential hydrogen-bonding donors, such as −NH- substituents, tend to favor recognition by P-gp recognition. To investigate the potential role of the −NH- groups present in Cu(gtsm) and Cu(gtse) in recognition by P-gp, analogues were prepared where this functional group had been replaced by another alkyl group, Cu(gtsm2) and Cu(gtse2), or by a pyrrolidine functional group, Cu(gtspyr). Each of the complexes that lacked the −NH- functional group displayed similar degrees of cellular uptake of copper in both the HEK293 and HEK293 P-gp cells suggesting that, unlike Cu(gtsm) and Cu(gtse), they are not substrates for P-gp. Although the cellular levels of copper following treatment with these complexes were not reduced in the P-gp overexpressing cells, the actual levels of intracellular copper were lower when compared to treatments with Cu(gtsm) and Cu(gtse). The complex with a pyrrolidine functional group, Cu(gtspyr), was a standout in that it was capable of delivering modest amounts of intracellular copper, but the uptake was not affected by the presence of P-gp, so it was selected for further investigations as a PET imaging agent. The radiolabeled complex, [64Cu]Cu(gtsm), has been suggested as an imaging agent to probe altered copper 718

DOI: 10.1021/acs.jmedchem.7b01158 J. Med. Chem. 2018, 61, 711−723

Journal of Medicinal Chemistry



metabolism in neurodegenerative diseases as two transgenic mouse models of amyloid pathology exhibit significantly higher brain uptake when compared to control animals. 23,24 Considering the significantly lower uptake of Cu(gtsm) in cells overexpressing P-gp, it is possible that [64Cu]Cu(gtsm) could be a useful PET tracer for assessing P-gp activity.49−51 Considering the importance of P-gp in the preventing transport of molecules across the blood−brain barrier and the identification of Cu(gtsm) as a likely P-gp substrate, it was hoped that derivatives that had similar degrees of high membrane permeability but were not P-gp substrates could result in even higher levels of brain uptake. An original intention of this study was to investigate whether measuring uptake in the HEK293 P-gp cell line would be useful in identifying complexes that would readily cross the blood−brain barrier.52 Structure−activity studies on a small series of compounds, which focused on removing −NH- hydrogen bonding functional groups, identified Cu(gtspyr) as a compound with good membrane permeability with cell uptake that was not affected by P-gp expression. While [64Cu]Cu(gtspyr) did display a different biodistribution to [64Cu]Cu(gtsm) in wild-type mice, the brain uptake was less for Cu(gtspyr) when compared to [64Cu]Cu(gtsm). One reason for the lower levels of uptake in the brain following administration of [64Cu]Cu(gtspyr) could be the significantly high liver uptake for [64Cu]Cu(gtspyr), 10.34 ± 2.47% IA/g. The high liver uptake of [64Cu]Cu(gtspyr) could reflect the lipophilicity of the tracer as at pH 7.4 [64Cu]Cu(gtspyr) log D = 1.53 compared to log D = 1.37 for [64Cu]Cu(gtsm) and the lack of P-gp-mediated efflux of [64Cu]Cu(gtspyr) from the liver when compared to [64Cu]Cu(gtsm). P-gp is expressed in hepatocytes and plays a role in the excretion of lipophilic drugs.53,54



Article

EXPERIMENTAL SECTION

Chemical Synthesis. General Conditions. All reagents and solvents were obtained from commercial sources and used as received unless otherwise stated. Nuclear magnetic spectra were acquired on an Agilent 400-MR (1H NMR at 400 MHz and 13C{1H} NMR at 101 MHz) or Varian FT-NMR 500 spectrometer (1H NMR at 500 MHz and 13C{1H} NMR at 126 MHz) at 298 K. All chemical shifts were referenced to the internal solvent residue and quoted in ppm relative to TMS. Microanalyses for C, H, and N were carried out by the Campbell Microanalytical Laboratory (Dunedin, New Zealand). ESIMS spectra were recorded on an Agilent 6220 ESI-TOF LC/MS Mass Spectrometer. The ligands used to make the copper(II) complexes were analyzed by NMR spectroscopy and ESI-MS. The paramagnetic copper(II) complexes were characterized by ESI-MS, microanalysis, and cyclic voltammetry, confirming ≥95% purity. Cyclic voltammograms were recorded on an AUTOLAB PGSTAT100 electrochemical workstation using GPES V4.9 software and employing a glassy carbon working electrode, a platinum counter electrode, and a Ag/Ag+ reference electrode. All measurements were carried out in dimethylformamide solutions that were 5 mM analyte with an electrolyte of 0.1 M tetrabutylammonium hexafluorophosphate. Dimethylformamide was obtained from standard commercial sources and dried over 3 Å molecular sieves before use. Each solution was purged with argon prior to analysis and measured at ambient temperatures under an argon atmosphere. Peak potentials, E0′ = [Epa + Epc]/2, were quoted versus SCE (where ferrocene/ferricenium E0′ = 0.45 V in dimethylformamide). Copper-64 was produced with the IBA Nirta target by the 64Ni(p,n)64Cu reaction at the Austin Hospital, Melbourne, Australia.55 4-Pyrrolidine-3-thiosemicarbazide,56 Cu(gtsm),57 Cu(gtse),57 Cu(gtsp),6 and Cu(gtsm2)57 were prepared following literature procedures. 4,4′-Diethyl-3-thiosemicarbazide. To a cooled (∼4 °C) aqueous solution of NaOH (2.52 g, 63 mmol, 25 mL H2O) were added diethlyamine (15 mL, 146 mmol) and carbon disulfide (3.6 mL, 60 mmol). The mixture was allowed to warm to ambient temperature. After 2 h the mixture was heated at reflux for 2 h, and then sodium chloroacetate (7.5 g, 64 mmol) was added and the mixture was heated at reflux for a further 2 h. To the cooled mixture hydrazine hydrate (78−82%, 18 mL) was added, and the mixture was heated at reflux for 2 h. The mixture was then cooled in ice and a colorless crystalline solid precipitated that was isolated by filtration, washed with H2O and pentane to afford colorless crystalline solid (2.46 g, 16.7 mmol, 28%). 1 H NMR (400 MHz; DMSO-d6): δ 8.70 (s, 1H, NH), 4.62 (s, 2H, NH2), 3.56 (q, 3JHH = 7.0 Hz, 4H), 1.06 (t, 3JHH = 7.0 Hz, 6H). 13 C{1H} NMR (101 MHz; DMSO-d6): δ 180.7, 44.2, 12.6. ESI-MS (+ve ion) m/z [M + H]+ 148.103 (experimental), 148.09 (calculated for [C5H14N3S]+). H2gtse2. A mixture of 4,4-diethyl-3-thiosemicarbazide (0.93 g, 6.3 mmol) and glyoxal (40% w/w, 0.39 g, 2.7 mmol) in ethanol (50 mL) was stirred at ambient temperature for 16 h. The tan-colored product was isolated by filtration and washed with cold ethanol (0.54 g, 1.7 mmol, 63%). Calcd for C12H24N6S2: C, 45.54; H, 7.64; N, 26.55. Found: C, 45.65; H, 7.58; N, 26.46. 1H NMR (400 MHz; DMSO-d6): δ 11.00 (br, 2H, 2NH), 7.92 (s, 2H, 2N = CH), 3.69 (q, 3JHH = 6.8 Hz, 8H, 4CH2), 1.18 (t, 3JHH = 6.9 Hz, 12H, 3CH3). {1H}13C NMR (101 MHz; DMSO-d6): δ 178.5, 142.5, 45.9, 12.8.ESI-MS (+ve ion) m/z [M + H] + 317.146 (experimental), 317.16 (calculated for [C12H25N6S2]+). Cu(gtse2). A mixture of H2gtse2 (0.16 g, 0.42 mmol) and copper(II) acetate hydrate (0.094 g, 0.47 mmol) in acetonitrile (15 mL) was heated at reflux for 5 h. The mixture was allowed to cool and a brown precipitate was collected by filtration, washed with acetonitrile and diethyl ether to give Cu(gtse2) (0.070 g, 0.47 mmol, 43%). Calcd for C12H22CuN6S2: C, 38.13; H, 5.87; N, 22.23. Found: C, 37.38; H, 5.74; N, 21.50. ESI-MS (+ve ion) m/z [M + H]+ 378.102 (experimental), 378.07 (calculated for [C12H23CuN6S2]+). H2gtspyr. A mixture of 4-pyrrolidine-3-thiosemicarbazide (0.54 g, 3.70 mmol) and glyoxal (40% w/w, 0.26 g, 1.8 mmol) in ethanol (50 mL) was stirred at ambient temperature for 16 h. The product was

CONCLUDING REMARKS

Of the six complexes investigated, treatment with Cu(gtsm), Cu(gtse), and Cu(gtspyr) results in the highest levels of intracellular copper in SKOV-3, HEK293, and HEK293 P-gp cell lines. The increased intracellular copper leads to increased levels of p-ERK in all three cell lines, presumably due to activation of the Ras-Raf-MEK-ERK signal transduction pathway. Treatment of SKOV-3 cells with relatively low concentrations of Cu(gtsm) and Cu(gtspyr) (1 μM) leads to significant trafficking of the endogenous copper transporter ATP7A from the Golgi network to the cell membrane. The trafficking of endogenous ATP7A from the Golgi to the cell membrane is an intrinsic cellular response to remove excess copper from the cell and suggests the copper has been released from the ligand to become bioavailable. Comparisons of cellular uptake of the complexes in HEK293 and HEK293-P-gp cells suggest that Cu(gtsm) and Cu(gtse) are substrates for the P-gp efflux protein. The cellular uptake of Cu(gtsm2) and Cu(gtse2), which have two alkyl substituents and lack a −NH- hydrogen bond donor, was the same in both the control and P-gp overexpressing cells but less than the cellular uptake of Cu(gtsm) and Cu(gtse). The complex with a pyrrolidine functional group, Cu(gtspyr), delivered modest amounts of intracellular copper, and the uptake was not affected by the presence of P-gp. A preliminary microPET experiment in wildtype mice showed that intravenous tail vein injection of [64Cu]Cu(gtspyr) resulted in reasonable brain uptake but also high liver uptake. 719

DOI: 10.1021/acs.jmedchem.7b01158 J. Med. Chem. 2018, 61, 711−723

Journal of Medicinal Chemistry

Article

isolated as a tan colored powder (0.40 g, 1.3 mmol, 72%). ESI-MS (+ve ion) m/z [M + H]+ 313.113 (experimental), 313.46 (calculated for [C12H21N6S2]+). 1H NMR: (500 MHz; DMSO-d6, acquired at 283 K): δ 7.88 (s, 2H), 5.26 (s, 2H), 3.67 (m, 8H), 1.89−1.88 (m, 8H). Cu(gtspyr). A mixture of H2gtspyr (0.088 g, 0.28 mmol) and copper(II) acetate hydrate (0.062 g, 0.31 mmol) in acetonitrile (15 mL) was heated at reflux for 5 h. The mixture was allowed to cool to room temperature. A brown precipitate was collected by filtration and washed with acetonitrile and ditheyl ether (0.085 g, 0.23 mmol, 82%). Calcd for C12H18CuN6S2.H2O: C, 36.77; H, 5.14; N, 21.44. Found: C, 37.14; H, 4.69; N, 21.34. ESI-MS (+ve ion) m/z [M + H]+ 374.070 (experimental), 374.04 (calculated for [C12H19CuN6S2]+). [64Cu]Cu(gtsm). An aliquot of H2gtsm (1 mg/mL, 2 μL) was added to 64Cu(CH3CO2)2 (17 MBq, 1 M acetate buffer, pH 5.5, 40 μL), and the mixture was incubated at room temperature for 30 min. Radiochemical purity was determined by radio-TLC (EtOH; Rf > 0.6, 97% yield). The reaction mixture was then diluted in PBS (pH 7.4, 60 μL) for injection. [64Cu]Cu(gtspyr). An aliquot of H2gtspyr (1 mg/mL, 6 μL) was added to 64Cu(CH3CO2)2 (66 MBq, 1 M acetate buffer, pH 5.5, 160 μL) and incubated at room temperature for 30 min. Radiochemical purity was determined by radio-TLC (EtOH; Rf > 0.6, 97% yield). The reaction mixture was then diluted in PBS (pH 7.4, 180 μL) for injection. Cell-Based Experiments. General. Rhodamine 123, cyclosporin A, and elacridar were purchased from Sigma-Aldrich (St. Louis, MO). Cell culture media and supplements were obtained from Gibco BRL (Grand Island, NY, USA). The following antibodies were used in Western blot analysis and were obtained from Cell Signaling Technologies (Danvers, MA, USA): anti-p-ERK, anti-ERK, anti-pAkt, anti-Akt, and anti-GAPDH. Anti-P-glycoprotein was purchased from Calbiochem (EMD Millipore Corp., Billerica, MA, USA). The antibody CT77 was used to detect the copper transporter ATP7A and was a kind gift from Prof. B. Eipper (Neuroscience and Molecular, Microbial, and Structural Biology Division Institute of Connecticut). Golgin-97 was used to visualize the trans-Golgi network (Invitrogen). Cell Lines. Human embryonic kidney cells (HEK) 293 expressing Pglycoprotein or vector only were a kind gift from Des R. Richardson (University of Sydney) and have been previously described. Cells were cultured in DMEM supplemented with 10% FCS, 1% NEEA-1, 1% sodium pyruvate, 100 U/mL penicillin, 100 μg/mL streptomycin, 2 mM L-glutamine and selected with 2 mg/mL Geneticin. SKOV3 ovarian carcinoma cells (kind gift from Dr. Carleen Cullinane, Peter MacCallum Cancer Centre) were used to investigate copper availability by cell imaging and were maintained in RPMI media supplemented with 10% FCS and penicillin/streptomycin. Metal Analysis via Inductively Coupled Plasma Mass Spectrometry (ICP-MS). HEK293 or SKOV3 cell lines were seeded in 6-well plates and grown until 80% confluency prior to treating with 1 μM Cu(gtsx) compounds for 0.5, 1, 2, or 4 h. Cells were then washed ×3 with ice-cold 1× PBS prior to harvesting cell pellet by scraping and centrifugation. Each treatment was carried out in triplicate, and experiments were repeated at least twice. Cell pellet was digested in 65% nitric acid, and copper levels were measured using an Agilent 7700 ICP-MS instrument. Results were expressed as micromole per liter concentrations of metal (μmol/L). The concentration of Cu was calculated as μM metal per μg of protein in the cell sample. MTT Reduction and LDH Assay. To assess the general health of SKOV3 cells following treatment with 1 μM Cu(gtsx) for 2 h, the 3(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay and the lactate dehydrogenases assay were carried out, as previously described.58 SDS−PAGE and Immunoblotting. Protein levels were analyzed by SDS−PAGE following 2 h treatment with Cu(gtsx) (1 μM) compounds as previously described. Briefly, cell lysates were prepared by dissolving cell pellets in Phosphosafe extraction reagent (Merck, Kilsyth, VIC, Australia) supplemented with phenylmethylsulfonyl fluoride (PMSF) and deoxyribonuclease I (DNase I, Roche Diagnostics, Castle Hill, NSW, Australia). Following centrifugation,

supernatant was collected and protein concentration measured using Pierce BCA protein determination assay as per manufacturer’s instructions (Thermo Scientific, Rockford, IL, USA). Samples (40 μg) were loaded on a NuPAGE 4−12% Bis-Tris gel (Invitrogen) and separated at 125 V for 1 h and then transferred onto a PVDF membrane (Roche Diagnostics) at 25 V; 90 min. Membranes were blocked in 4% skim milk in PBST (PBS and 0.05% v/v Tween-20) and incubated with primary antibodies overnight at 4 °C. Unbound antibodies were washed in PBST (×3), and the membrane was then probed with anti-mouse/rabbit HRP-conjugated secondary antibodies (Cell Signaling Technologies). Protein bands were visualized using ECL Advanced reagent (GE Healthcare Biosciences, Castle Hill, NSW Australia) and a Fujifilm LAS-3000 imager. Changes in protein levels were quantified by densitometry analysis using ImageJ software and standardized relative to GAPDH levels (n = 3). P-Glycoprotein Functional Flow Cytometry Assay. P-Glycoprotein (P-gp) function in HEK293 cell lines was measured by rhodamine-123 accumulation assay. Rhodamine-123 is a fluorescent substrate for efflux transporters including P-gp and has been used as a marker to study Pgp activity. Cells in 6-well plates were treated with 0.5 μg/mL of rhodamine-123 ± 5 μM cyclosporin A for 30 min at 37C in 5% CO2. Alternatively, cells were treated with elacridar (0.2 μM), a potent and selective inhibitor of P-gp, for 24 h prior to addition of Cu(gtsx) complexes (2 h), as previously described.36 The medium was then removed and incubated with rhodamine-free medium for 30 min. Cells were collected by trypsinization, and rhodamine fluorescence within cells was measured by flow cytometry using either LSR Fortessa (BD Biosciences) or CytoFlex S analyzer (Beckman Coulter) operated with a 15 mW argon ion laser tuned to λ = 488 nm excitation wavelength. Flow cytometry data was analyzed using FlowJo, LLC software. Immunocytochemistry. SKOV3 cells were seeded on poly-D-lysine (0.1 mg/mL) coated glass coverslips at a density of 0.025 × 106/ well. Cells were treated with 1 μM copper complex or vehicle (DMSO) for 2 h. Cells were then fixed in 4% paraformaldehyde in PBS, permeabilized with 0.1% Triton-X100 in PBS, and nonspecific sites blocked with 1% bovine serum albumin (Sigma) for at least 2 h prior to incubating with primary antibodies CT77 (1:500) and golgin-97 (1:200) overnight. Following 20 min washes with 1× PBS, primary antibody was detected with secondary IgG antibodies conjugated to Alexa Fluor 488, 568, or 647 fluorophores (1:400). Rhodamine phalloidin (Invitrogen) was used at 1:50 dilution. The nucleic acid stain DAPI was used at 100 ng/mL (Invitrogen). Images were taken using a confocal Leica SP8 microscope. Small Animal PET Imaging (MicroPET/MRI). Animal ethics approval from the Austin Health Animal Ethics Committee was granted for these studies. All mice were housed at the Florey Animal facility, in conditions of controlled temperature (22 ± 2 °C) and lighting (14:10 h light/dark cycle), with free access to food and water. Mice were injected (tail vein) with [64Cu]Cu(gtsx) (3.7 MBq), and isofluorane anesthesia was induced and maintained using a vaporizer (O2 set to 2 (L/min)/2% isoflurane). A series of 6 × 5 min emission scans were acquired on a Mediso MicroPET/MR (