Ketamine-Induced Apoptosis in Normal Human Urothelial Cells: A Direct, N-Methyl-d-Aspartate Receptor-Independent Pathway Characterized by Mitochondrial Stress

Abstract

Recreational abuse of ketamine has been associated with the emergence of a new bladder pain syndrome, ketamine-induced cystitis, characterized by chronic inflammation and urothelial ulceration. We investigated the direct effects of ketamine on normal human urothelium maintained in organ culture or as finite cell lines in vitro. Exposure of urothelium to ketamine resulted in apoptosis, with cytochrome c release from mitochondria and significant subsequent caspase 9 and 3/7 activation. The anesthetic mode-of-action for ketamine is mediated primarily through N-methyl d-aspartate receptor (NMDAR) antagonism; however, normal (nonimmortalized) human urothelial cells were unresponsive to NMDAR agonists or antagonists, and no expression of NMDAR transcript was detected. Exposure to noncytotoxic concentrations of ketamine (≤1 mmol/L) induced rapid release of ATP, which activated purinergic P2Y receptors and stimulated the inositol trisphosphate receptor to provoke transient release of calcium from the endoplasmic reticulum into the cytosol. Ketamine concentrations >1 mmol/L were cytotoxic and provoked a larger-amplitude increase in cytosolic Ca(2+) concentration that was unresolved. The sustained elevation in cytosolic Ca(2+) concentration was associated with pathological mitochondrial oxygen consumption and ATP deficiency. Damage to the urinary barrier initiates bladder pain and, in ketamine-induced cystitis, loss of urothelium from large areas of the bladder wall is a reported feature. This study offers first evidence for a mechanism of direct toxicity of ketamine to urothelial cells by activating the intrinsic apoptotic pathway.

Figure 1

Histological assessment of human ureteric organ cultures after 72 hours' exposure to 3 mmol/L ketamine showed thinning of the epithelium (shown in the images of donor x) and clear signs of apoptosis, including pyknotic nuclei and karyorrhexis (shown in the images of donor y). Immunoreactivity with a cleaved cytokeratin 18 antibody (M30 Cytodeath) was observed in cells being lost from the epithelium (arrows). n = 6 donors with two representative examples shown. Scale bar = 100 μm. H&E, hematoxylin and eosin.

Figure 2

Effects of ketamine on population growth in cultures of normal (nonimmortalized) human urothelial (NHU) cells. A: Concentrations of ketamine between 0.3 and 6 mmol/L were assessed during a 6-day time course by Alamar Blue reduction assay. This graph is representative of repeats in three independent donor cell lines. B: Counts from two donor cell lines after 96 hours of exposure to ketamine indicated a half maximal inhibitory concentration (IC₅₀) for NHU cells of 0.93 mmol/L. The IC₅₀ was calculated using a sigmoidal fit and is illustrated as a solid black line; the 95% CI is shown as dashed lines. Error bars represent SD (A). n = 6 (A); n = 4 replicates per donor (B).

Figure 3

Induction of apoptosis in normal (nonimmortalized) human urothelial (NHU) cell cultures by ketamine. A: Western blot analysis of NHU cells separated into cytoplasmic and mitochondrial fractions showed elevated cytochrome c in the cytosolic fractions of cultures treated with 3 mmol/L ketamine for 24 hours. Loading controls were Bcl2 for the mitochondrial fraction and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) for the cytoplasmic fraction. B: Densitometry analysis showed a significant mean 2.4-fold change in cytoplasmic cytochrome c content in three independent NHU cell lines after 24 hours' exposure to 3 mmol/L ketamine. C: Western blot analysis of phospho-Akt (active form), phospho-ERK1/2 (active form), and S9 phospho-glycogen synthase kinase (GSK) 3β (inactive form) showed early depletion of these forms of the kinases in response to 3 mmol/L ketamine. Abundance quantified by densitometry is shown as a percentage of control cells for each time point, normalized to β-actin (combined data). D: Inhibition of GSK3β by SB415286 in ketamine-exposed NHU cells was capable of a slight, but significant, inhibition of toxicity, as assessed by Alamar Blue reduction. E: Inhibition of the mitochondrial permeability transition pore with cyclosporin A (CsA) in ketamine-exposed NHU cells was capable of a small, but significant, inhibition of toxicity, as assessed by Alamar Blue reduction. F: Western blot analysis of caspase 9, caspase 3, and cleaved poly (ADP-ribose) polymerase (PARP) in NHU cells after 72 hours' exposure to 3 mmol/L ketamine. G: The three markers of apoptosis were all significantly increased by more than twofold in densitometry, which was normalized to β-actin. H: Caspase 3/7 activity was assessed in lysates from NHU cell cultures exposed to 0.1 to 6 mmol/L ketamine and normalized to baseline caspase activity in untreated cells. Significant increases in caspase activity were observed after exposure to 3 and 6 mmol/L ketamine. Error bars represent SD (B, D, E, G, and H). n = 4 donors (C); n = 6 (D, E, and H); n = 3 donors (B and G). ^∗P < 0.05, ^∗∗P < 0.01, and ^∗∗∗P < 0.001.

Figure 4

RT-PCR of NMDA receptor transcripts (GRIN isoforms). No expression of GRIN isoforms was detected in normal (nonimmortalized) human urothelial (NHU) cells under either proliferating or differentiated cell culture conditions. Furthermore, no GRIN expression was detected in freshly isolated urothelium after its separation from the underlying stroma. Representative images of one donor (proliferating and differentiated NHU) and one donor P0 urothelium are shown. Uroplakin 2 (UPK2) transcript expression was used to confirm urothelial differentiation, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was included as an internal housekeeping control to confirm RNA integrity. In addition, no products were detected in reverse transcriptase–negative cDNA controls that were generated using each RNA preparation.

Figure 5

Treatment of normal (nonimmortalized) human urothelial (NHU) cells with ketamine (Ket) resulted in elevation of cytosolic [Ca²⁺]. A: The addition of 1 mmol/L ketamine (noncytotoxic) to a culture of NHU cells produced a transient increase in cytosolic [Ca²⁺], whereas exposure to 3 mmol/L ketamine (cytotoxic) induced a higher [Ca²⁺] in the cytoplasm, which did not return to a normal baseline. B: The role of purinergic receptors in mediating the urothelial response to 1 mmol/L ketamine was investigated by preincubating cultures for 10 minutes in 100 μmol/L pyridoxal-phosphate-6-azophenyl-2′,4′-disulfonate (PPADS; purinergic receptor inhibitor) or ARL-67156 (ecto-ATPase inhibitor). PPADS (dotted line) almost completely blocked the calcium transient observed after addition of 1 mmol/L ketamine to NHU cells, whereas ARL-67156 (dashed line) increased both the amplitude and duration of the transient. C: Preincubating cultures in PPADS was not sufficient to inhibit the [Ca²⁺] increase triggered by 3 mmol/L ketamine. All images are traces representative of triplicate experiments.

Figure 6

Normal (nonimmortalized) human urothelial (NHU) cell cystolic [Ca²⁺] was elevated by the release of stores from the endoplasmic reticulum in response to ketamine exposure. A: The repeated administration of 1 mmol/L ketamine to NHU cells in the presence of 2 mmol/L extracellular [Ca²⁺] resulted in repeated transient increases in cytosolic [Ca²⁺], with no sign of reduced intensity in subsequent transients. B: The source of the calcium transient was elucidated by performing repeated challenges of urothelial cells with 1 mmol/L ketamine (solid line) and 20 μmol/L ATP (dashed line) in the absence of extracellular calcium. An initial treatment with either ketamine or ATP generated a calcium transient showing that, in both cases, calcium was released from internal stores. A second stimulation triggered a significantly attenuated transient in response to either ketamine or ATP, indicating incomplete refilling of the internal stores. C: Pretreatment of urothelial cell cultures with 100 μmol/L 2-APB produced complete inhibition (dashed line) of the 1 mmol/L ketamine-induced calcium release (solid line). D: The source of the sustained [Ca²⁺] elevation was confirmed as internal by stimulating the cells with 3 mmol/L ketamine in the absence of exogenous calcium. E: Thapsigargin was used to inhibit SERCA pumps and effectively empty the ER by allowing Ca²⁺ ions to leak out. The 3 mmol/L ketamine-induced elevation in cytosolic [Ca²⁺] persisted. Images are traces representative of triplicate experiments.

Figure 7

Ketamine exposure causes mitochondrial stress in normal (nonimmortalized) human urothelial (NHU) cells. A: NHU cells exposed to 3 mmol/L ketamine for 48 hours had a significantly lower resting mitochondrial oxygen consumption rate (OCR) than controls (54.4%; Mann-Whitney test P < 0.001). The proportion of OCR devoted to physiology other than ATP generation remained the same, whereas the spare mitochondrial capacity in control cells was far greater (4.6-fold; Mann-Whitney test). Data are the means of experiments performed in duplicate on three independent cell lines. B: NHU cells labeled with tetramethylrhodamine (TMRM) showed that 3 mmol/L ketamine exposure led to elevated mitochondrial membrane potential (images are representative of experiments performed in cells from three donors). C: Results in B were confirmed quantitatively by flow cytometry (mean of three independent donor cell lines given). D: Cellular ATP was significantly reduced to 76.8% of control after 3 mmol/L ketamine exposure for 48 hours. Data are the means of duplicate measurements in three independent donor cell lines. Error bars represent SD (A, C, and D). ^∗P < 0.05, ^∗∗∗P < 0.001. Scale bar = 50 μm (B). FCCP, carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone.

Figure 8

Schematic diagram summarizing the proposed mechanism of ketamine-induced cytotoxicity downstream of initial receptor activation/inhibition and illustrating how the experiments performed in this study support the proposed mode-of-action. Ketamine exposure leads to ATP release from the cells (the ATP signal could be enhanced by preventing its breakdown with ARL-67156), which binds to P₂Y receptors [that were effectively inhibited by pyridoxal-phosphate-6-azophenyl-2′,4′-disulfonate (PPADS)]. Activation of P₂Y receptors releases inositol trisphosphate (IP₃), which binds IP₃ receptor on the endoplasmic reticulum [inhibited using 2-aminoethoxydiphenyl borate (2-APB)] and causes release of stored Ca²⁺ into the cytoplasm. Prolonged elevation of cytoplasmic [Ca²⁺] triggers further Ca²⁺ release from mitochondria, probably via the activated mitochondrial permeability transition pore (MPTP). Ketamine exposure triggers a reduction in the phosphorylated/active forms of Akt and ERK, promoting an increase in glycogen synthase kinase (GSK) 3β activity (which can be inhibited by SB415286). Acting in concert, GSK3β activity and sustained elevation of cytosolic [Ca²⁺] activate the MPTP (which can be inhibited by cyclosporin A), leading to mitochondrial depolarization, release of cytochrome c, and caspase-mediated apoptosis.