The Effect of the Classical TSPO Ligand PK 11195 On in Vitro Cobalt Chloride Model of Hypoxia in Non-Neuronal and Brain Cell Lines

The mitochondrial translocator protein (TSPO) is a modulator of the apoptotic pathway involving reactive oxygen species (ROS) generation, mitochondrial membrane potential (Δψm) collapse, activation of caspases and eventually initiation of the apoptotic process. In this in vitro study, H1299 lung cells and BV-2 microglial cells were exposed to the hypoxic effect of CoCl 2 with or without PK 11195. Exposing the H1299 cells to 0.5 mM CoCl 2 for 24 hours resulted in decreases in cell viability (63%, p<0.05), elevation of cardiolipin peroxidation levels (38%, p<0.05), mitochondrial membrane potential depolarization (13%, p<0.001), and apoptotic cell death (117%, p<0.05). Pretreatment with PK 11195 (25 µM) exhibited signi�cant protective capacity on CoCl 2 -induced alterations in the mentioned processes. Exposure of BV-2 cells to increasing concentrations of CoCl 2 (0.3, 0.5, 0.7 mM) for 4 hours resulted in alterations in the same cellular processes. These alterations were obtained in a dose-dependent manner, except the changes in caspases 3 and 9. The novel ligands as well as PK 1195 attenuated the in vitro hypoxic effects of CoCl 2 . It appears that the TSPO ligand PK 11195 can prevent CoCl 2 -induced cellular damage in both non-neuronal and brain cell lines, and they may offer new therapeutic options in hypoxia-related lung and brain diseases which fail to respond to conventional therapies.


Introduction
Hypoxia refers to an abnormally exposure of cells or tissues to low level of oxygen. The cellular response to hypoxia has been of great interest to researchers since hypoxia is relevant to essential biological processes including angiogenesis, cellular survival/proliferation, energy metabolism, erythropoiesis, extracellular matrix function, invasion/metastasis, iron metabolism, pH regulation, multi-drug resistance and stem cell survival 1,2 .
Neurodegenerative diseases are de ned by the progressive loss of neurons, synapses, and protein misfolding, and aggregation of proteins 3 . Reduced oxygen supply has been suggested as an important contributor to pathogenesis of neurodegenerative diseases. Hypoxia was found to induce oxidative stress, in ammation and apoptosis, among other cellular processes, contributing to the pathophysiology of neurodegeneration 3 .
CoCl 2 is a hypoxia-mimicking agent 4 . It inhibits prolyl hydroxylase-domain enzymes (the oxygen sensors) through replacement of Fe 2+ with Co 2+ making these enzymes unable to label hypoxia-inducible factor (HIF)-alpha for degradation 5 . It should be noted that CoCl 2 mimics HIF-1 alpha accumulation, but no other effects of hypoxia. In addition to its effects on HIF-1 alpha accumulation, exposure to CoCl 2 can also modulate processes and pathways, such as apoptosis and reactive oxygen species (ROS) generation 6,7 .
The mitochondrial translocator protein (TSPO) was shown to regulate ROS generation, proliferation, angiogenesis and apoptosis, cellular processes that are relevant to the toxic effect of CoCl 2 8, 9 .
TSPO is highly expressed during in ammation and in various tumor types, such as glioblastomas 10 . In the U118MG glioblastoma cell line, exposed to CoCl 2 at various concentrations was shown to increase ROS formation, cause the collapse of the mitochondrial membrane potential and cause cell death [11][12][13][14] .
TSPO knockdown using siRNA, or its blockade using the TSPO antagonist PK11195, signi cantly counteracted the CoCl 2 -induced effects 9 . Several studies investigated the role of TSPO in CoCl 2 -induced hypoxia 7,9,15 . Our research group evaluated the effects of TSPO ligands on CoCl 2 -induced cytotoxicity, on the human H1299 lung cancer and glial cell line. Further investigations on this topic may produce implications on the role of TSPO as a therapeutic target in hypoxic conditions.
In the present study, we evaluated the protective effects of the TSPO ligand PK 11195 in non-neuronal (lung cancer cell line) and brain (microglial cell line) cells exposed to the hypoxic agent CoCl 2 7 in an attempt to identify the cellular mechanisms that are involved in such putative bene cial effects. Eagle's medium high glucose containing 4.5 g/l glucose, 4 mM L-glutamine and supplemented with 5% fetal bovine serum, penicillin (100 U/ml), and streptomycin (100 µg/mL) 16 . These two cell types were cultured at 37 °C in 5% CO2 and 90% relative humidity.

TSPO ligands pretreatment
The in vitro experiments included the following groups: vehicle control group pretreated with 1% ethanolcontaining 0.5% Fetal Calf Serum (FCS, biological industries, Beit Ha'Emek, Israel); a group pretreated with 1% ethanol-containing serum starvation medium.
The two cell lines were seeded in 96-well plates (5x10 3 cells per well) or 6-well plates (2.5x10 5 cells per well) (depending upon the type of experiment) and grown in complete medium for 48 hours at 37ºC and 5% CO 2 until the desired con uency 80% was reached. Then, pretreatment with the TSPO ligand PK 11195 (25 µM) in serum-starvation medium was applied for another 24 hours.

Cobalt Chloride (CoCl 2 ) exposure
CoCl 2 (Sigma-Aldrich, Rehovot, Israel) was prepared at the concentrations required for each speci c experiment and applied to the CoCl 2 -treated groups for 24 hours 7,9 .

Cell counting
Cells were grown until 80% con uency was reached, the medium was discarded, cells were washed with phosphate buffer saline (PBS) and collected following trypsinization. For cell counting, 100 µl of the cells was placed in an Eppendorf tube then mixed with 100 µl of trypan blue (ratio 1:1). Under the light microscope, the cells were counted by hemocytometer (Neubauer slide, Sigma Aldrich, Rehovot).

XTT assay
The two cell lines (H1299 and BV-2) were seeded in 96-well plates (5,000 cells/well) and grown for 48 hours in complete medium. Then pretreated with the required TSPO ligand for another 24 hours, followed by exposure of the cells to the desired concentrations of CoCl 2 for 30 minutes, 4 hours, or 24 hours.
Assessment of cellular viability was performed using XTT cell viability kit (Biological Industries, Bait Ha'Emek, Isreal), following the manufacturer's protocol:150 µl medium from each well was removed followed by adding 50 µl from the XTT mixture to the remaining 50 µl medium within the plate, then the plates were incubated in dark for one hour and a half. Reduction of 2,3-bis-(2-methoxy-4-nitro-5sulfophenyl)-2H-tetrazolium-5-carboxanilide (XTT) depends on mitochondrial dehydrogenases and reductases, which results in orange formazan dye production, a process that occurs only in viable cells. The amount of the orange dye indicates the cellular viability and the optic density (O.D.) was measured using Infinite M200 Pro plate reader (Tecan, Männedorf, Switzerland) with absorbance with endpoint photometric at 492 nm wavelength and reference wavelength of 620 nm.

JC-1 assay
The mitochondrial membrane potential (Δψm) depolarization was assessed using the JC-1 assay, which based on cationic, lipophilic tetra-ethyl-benzimidazolyl-carbocyanine iodide JC-1 dye. After seeding the H1299 cells in 6-well plates (250,000 cells/well) for 48 hours and pretreated with 25 µM PK 11195 for another 24 hours, cells were exposed to 0.5 mM CoCl 2 for 24 hours, then cells were trypsinized (600 µl trypsin), collected and centrifuged (660 g for 5 minutes at room temperature) followed by removing the supernatant and resuspended in 600 µl of PBS and cells were centrifuged again. Dilution of JC-1 with PBS (1:500), 600 µl were applied and incubated for 30 minutes in dark. Cells were centrifuged again and 400 µl of PBS were added followed by cells ltration and transfer to FACS tubes. In case of intact cells with high Δψm, JC-1 enters the mitochondria and forms J-aggregates emitting red uorescence at 590 nm. In contrast, cells exposed to CoCl 2 , with low Δψm, the JC-1 dye remains in cytosol compartment in a monomer form emitting green uorescence at 527 nm indicating Δψm depolarization (Zeno et al., 2009). The median uorescence intensity (MFI) indicates Δψm depolarization that was calculated by red/green ratio using FACS, and the results were analyzed using FlowJo (FlowJo LLC, Ashland, Oregon)

Nonyl Acridine Orange (NAO) assay-cardiolipin peroxidation indicator
H1299 Cells were seeded in 6-well plates (250,000 cells/well), after 48 hours, 25 µM of PK 11195 was applied for 24 hours, then cells were exposed to 0.5 mM CoCl 2 for another 24 hours. Afterwards, cells were trypsinized and centrifuged (660 g for 5 minutes at room temperature) followed by neutralizing trypsin with complete medium. The cells were washed with PBS and centrifuged again the supernatant was aspirated and 400 µl of Nonyl Acridine Orange (NAO) stain (diluted with PBS at a ratio of 1:1000) was added. The cells were incubated in dark for 30 minutes, and centrifuged again, then 400 µl of PBS were added and transferred to FACS tubes.
NAO stain was used to assess the cardiolipin peroxidation level. Cardiolipin, a polyunsaturated acidic phospholipid, biosynthesized and localized in the inner mitochondrial membrane.

Results
The protective ability of PK 11195 to counteract the effect of CoCl 2 -induced mitochondrial damage via TSPO-related processes in non-neuronal (lung cancer) and central (microglial) cell lines included the assessment of: cell viability, oxidative stress, mitochondrial membrane potential depolarization, TSPO protein expression levels, and apoptotic markers.  1-1 mM). Signi cant decreases in cell viability were detected at 0.3 to 1 mM of CoCl 2 . Since a su cient, but not complete suppressive effect of CoCl 2 on cellular viability was observed at 0.5 mM, this concentration was chosen for all the further experiments (p<0.001; Figure 1).

Discussion
In the present study the capacity of the TSPO ligand PK 11195 to counteract the CoCl 2 -induced hypoxic cellular damage was evaluated in both non-neuronal (lung) and central (microglia) cell lines.

Lung cancer cell line (H1299)
We used an established in vitro model of pulmonary hypoxia by exposing H1299 lung cancer cell line to cobalt chloride and investigated the protective effects of TSPO ligands in this cellular model 7,9 . According to our data, lung derived cells exposed to various concentrations of CoCl 2 , ranging from 0.1 mM to 1 mM, for 24 hours led to a dose-dependent reduction in cell viability with maximal toxicity at 1 mM as compared to a control group. Based on the current dose-dependent results, and similar to previous studies 8, 9 , which showed su ciently effective toxic damage that mimics hypoxia at a concentration of 0.5 mM of CoCl 2 , we assessed in our experiments the impact of this concentration on a variety of TSPOrelated cellular processes, including: cell death, mitochondrial membrane potential depolarization, cardiolipin peroxidation and ROS generation. We evaluated the protective effects of the TSPO ligands PK 11195 at a concentration of 25 µM in the CoCl 2 hypoxic cellular model. PK 11195 exhibited signi cant inhibitory effect on CoCl 2 -induced cell viability reduction.
TSPO has been reported to be involved in several mitochondrial processes affected by CoCl 2 exposure, including apoptosis, ROS generation and collapse of mitochondrial membrane potential. One putative pathway for the CoCl 2 -induced cytotoxicity is through an harmful impact on mitochondrial functions mediated by TSPO, including CoCl 2 -induced apoptosis mediated by ROS generation, cardiolipin peroxidation, mitochondrial membrane potential depolarization, and decreased cellular metabolism and viability 8, 9 .
In our present study, using ROS generation as an indicator of oxidation, PK 11195 exhibited a potent inhibitory effect on CoCl 2 -induced cardiolipin peroxidation. These ndings further strengthen the previously published data on U118MG cells as a model for glioma 9 .
It is likely that the CoCl 2 -induced accumulation of ROS interfered with the mitochondrial homeostasis of the mitochondrial membrane potential. Indeed, a signi cant increase was seen in depolarization of the mitochondrial membrane potential in CoCl 2 -treated cells as compared to unexposed control group. It appears that CoCl 2 led to ROS generation and disruption of mitochondrial potential which eventually led to cytochrome c release and subsequent initiation of apoptotic pathway 8 . A previous microscopic study revealed morphological, nuclear, and cytological changes as features of apoptosis including condensed chromatin, DNA fragmentation, cell shrinkage and cell surface blabbing 17 . In this study, necrosis/apoptosis assay was used to assess the harmful effects of CoCl 2 on cell viability (apoptotic or necrotic cell death). Following exposure to CoCl 2, H1299 cells showed signi cant elevation in apoptosis levels, but not necrosis, as was reported previously 8, 9 . Notably, another study performed by our group using cigarette smoke as a hypoxia-causing agent leading damage to cellular hypoxia, also resulted in apoptotic cell death, rather than necrotic cell death 8 .
In the current study, PK 11195 exhibited signi cant inhibitory effect on CoCl 2 -induced cell viability reduction. A previous study demonstrated the e cacy of the classical TSPO ligand PK 11195 in counteracting the effect of the CoCl 2 -induced damages in astrocytic cell line (U118MG) 9 . In another study, in the same H1299 lung derived cells, the classical high a nity TSPO ligand PK 11195 exhibited a signi cant protective activity 8 . However, the relationship between the a nity of the ligand to TSPO and the pharmacological activity in the various cell lines and the various cellular functions/pathways and models for speci c pathological damage is yet unclear. Moreover, the anti-hypoxic effects of PK 11195 are relevant to long-term hypoxia (24 hours), but the relevance to shorter period of hypoxia is yet unclear.

Microglial BV-2 cells
BV-2 microglial cells were chosen to investigate the effect of CoCl 2 in a cell line from the central nervous system which will enable the differentiation of the effects of CoCl 2 in cell lines from non-neuronal (lung) and central (brain) origin. Additionally, the nding of apoptotic cell death occurrence in H1299 cells and BV2 cells supports further our previous observations 9,18 . Apoptosis is a form of programmed cell death, in which caspases are strongly involved. Caspases divided into initiator caspases and executioner caspases. In the present study, caspase 3 (initiator) and 9 (executioner) were assessed as apoptotic markers after exposure to the required duration and concentration of CoCl 2 . The elevation in apoptotic markers was inhibited by PK11195. Interestingly no signi cant effect was shown regarding caspase 9 levels.
In conclusion, CoCl 2 as mimicking agent of hypoxia leads to alterations in several apoptosis-associated processes which occur in parallel to a reduction in the levels of TSPO protein levels. Such processes involve essential mitochondrial functions that after a certain time point and at speci c CoCl 2 concentration (0.7 mM) may reach an irreversible damage. The high a nity classical TSPO ligand PK 11195 ligand at a concentration of 25 µM, can prevent some cellular damages caused by exposure to CoCl 2 , however, there are cytotoxic cellular pathways that are insensitive to TSPO ligands and the bene cial effects at present are relevant to only long-term hypoxia (24 hours). Such pathways, mainly relevant to generation of oxidative stress, might occur in a non TSPO-related fashion, and thus no inhibitory/protective impact of TSPO ligands can be obtained. Figure 1 Dose-response analysis of H1299 cell viability (using XTT assay kit) after 24 hours' exposure to CoCl2 (0.1 -1 mM). Results are expressed as mean ± SEM (8 replicates for each group). ***p<0.001 compared to control. Evaluation of the protective ability of the TSPO ligand PK 11195 to coun-teract the effect of CoCl2induced cell death. The protective ability of PK 11195 (n=5 replicates) to counteract the effect of 0.5 mM of CoCl2 was measured using XTT analysis. Results are expressed as means ± SEM ***p<0.001 vs.

Figure 5
Evaluation of apoptotic cell death levels assessed by apopxin dye following 24 hours' exposure of H1299 to 0.5 mM CoCl2 with or without pretreatment with 25 µM PK 111195. Results are expressed as mean ± SEM (n=4 repli-cates in each group). # and *p<0.05.