A role for p53 in the regulation of lysosomal permeability by Δ9-tetrahydrocannabinol in rat cortical neurones: implications for neurodegeneration


Address correspondence and reprint requests to Professor Veronica A. Campbell, Department of Physiology & Trinity College Institute of Neuroscience, Trinity College Dublin, Dublin 2, Ireland. E-mail: vacmpbll@tcd.ie


The psychoactive ingredient of marijuana, Δ9-tetrahydrocannabinol (Δ9-THC), can evoke apoptosis in cultured cortical neurones. Whilst the intracellular mechanisms responsible for this apoptotic pathway remain to be fully elucidated, we have recently identified a role for the CB1 type of cannabinoid (CB) receptor and the tumour suppressor protein, p53. In the current study, we demonstrate the Δ9-THC promotes a significant increase in lysosomal permeability in a dose- and time-dependent manner. The increase in lysosomal permeability was blocked by the CB1 receptor antagonist, AM251. Δ9-THC increased the localization of phospho-p53Ser15 at the lysosome and stimulated the release of the lysosomal cathepsin enzyme, cathepsin-D, into the cytosol. The p53 inhibitor, pifithrin-α and small interfering RNA-mediated knockdown of p53 prevented the Δ9-THC-mediated increase in lysosomal permeability. Furthermore, the Δ9-THC -mediated induction of apoptosis was abrogated by a cell-permeable cathepsin-D inhibitor (10 μM). Thus, the study demonstrates that Δ9-THC impacts on the lysosomal system, via p53, to evoke lysosomal instability as an early event in the apoptotic cascade. This provides evidence for a novel link between the CB1 receptor and the lysosomal branch of the apoptotic pathway which is crucial in regulating neuronal viability following exposure to Δ9-THC.

Abbreviations used





acridine orange






phosphate-buffered saline


small interfering RNA


terminal deoxynucleotidyl transferase

Δ9-Tetrahydrocannabinol (Δ9-THC) is the psychoactive ingredient of the cannabis plant, Cannabis sativa (marijuana). Δ9-THC exerts its effect on the CNS by activating the G protein-coupled cannabinoid (CB1) receptor, which is widely distributed in the brain (Herkenham et al. 1991), and is linked to activation of a number of signal transduction pathways including extracellular-regulated protein kinase (Derkinderen et al. 2003), stress-activated protein kinases (Downer et al. 2003) and sphingomyelinase (Blázquez et al. 2003). CB1 receptor activation is involved in the physiological control of synaptic activity (Carlson et al. 2002), motor function (Kishimoto and Kano 2006), as well as feeding, appetite and pain perception (Fride 2005).

Marijuana is a commonly used drug of abuse that induces euphoria (Ameri 1999), although the detrimental effects of the drug include memory impairments (Ranganathan and De Souza 2006) and an increased risk of psychosis (Moore et al. 2007). Although the potential neurotoxicity of marijuana is poorly defined, chronic recreational use has been linked with morphological changes in brain structures that are indicative of toxicity (Scallet 1991; Lawston et al. 2000). Functional magnetic resonance imaging studies have shown a reduction in frontal white-matter volume in cannabis abusers (Schlaepfer et al. 2006) and heavy marijuana users are found to have reduced grey matter in the parahippocampal gyrus and reduced white matter in the left parietal lobe (Matochik et al. 2005). Some of the deleterious effects of marijuana may be related to alterations in pathways associated with neurogenesis, synaptogenesis and wiring (Harkany et al. 2007), impaired myelination (Schlaepfer et al. 2006) or possibly aberrant neuronal death. The impact of CBs on neuronal viability is controversial with both neuroprotective and neurotoxic effects having been reported. Thus, in cultured neurones (Chan et al. 1998; Campbell 2001; Downer et al. 2003) and glioma cells (Galve-Roperh et al. 2000), Δ9-THC evokes apoptosis via activation of c-jun N terminal kinase and formation of ceramide respectively. However, Δ9-THC also protects neurones against excitotoxicity both in vitro (Gilbert et al. 2007) and in vivo (Raman et al. 2004) and has antioxidant (Hampson et al. 1998) and anti-inflammatory (Lyman et al. 1989) actions in the brain which may mediate neuroprotection (Campbell and Gowran 2007). The impact of CBs on neuronal fate is possibly determined by the stage of neuronal development as the neonatal brain is more vulnerable than the adult brain to the neurotoxic actions of Δ9-THC (Downer et al. 2007a) and such toxicity may contribute to the deficits in neuronal function that are observed following prenatal exposure to cannabis (Fried and Smith 2001).

Apoptosis is a programmed form of cell death which is essential for brain development, although excessive apoptosis is a feature of neurodegenerative disease (Blomgren et al. 2007). The apoptotic pathway involves the translocation of mitochondrial cytochrome c into the cytosol with a subsequent activation of a caspase protease cascade (Slee et al. 1999). Indeed, we have previously shown that exposure of cultured cortical neurones to Δ9-THC causes cytochrome c release and caspase 3 activation with a subsequent demise of the cell (Campbell 2001). Modulation of intracellular organelles is a common phenomenum during apoptosis but, whilst most studies focus on the mitochondrial regulation of apoptosis, several reports have indicated a role for lysosomes in early apoptotic events (Kågedal et al. 2001;Mathiasen and Jäättelä 2002; Stoka et al. 2007). Destabilization of the lysosomal membrane and translocation of lysosomal cathepsin enzymes from the lysosomal compartment to the cytosol have been reported as upstream apoptotic events induced by various stimuli such as synthetic retinoids (Zang et al. 2001), oxidative stress (Roberg and Ollinger 1998; Antunes et al. 2001), staurosporine (Kågedal et al. 2001), and over-expression of the tumour suppressor protein, p53 (Yuan et al. 2002). The mechanisms that control lysosomal permeability during apoptosis are ill-defined but over-expression of Bcl-2 can inhibit lysosomal-dependent apoptosis by stabilizing the lysosome (Zhao et al. 2001) and the pro-apoptotic member of the Bcl-family protein member, Bax, can insert into the lysosomal membrane to promote the release of lysosomal enzymes during staurosporine-induced apoptosis (Kågedal et al. 2001). Thus, members of the Bcl-family of proteins, classically associated with regulating the mitochondrial branch of apoptosis, also appear to influence lysosomal events associated with apoptosis. The reports that p53 induces lysosomal destabilization (Yuan et al. 2002) is of particular interest given that we have recently reported that p53 is instrumental in inducing Δ9-THC-mediated apoptosis in cultured neurones (Downer et al. 2007b). In the current study, we examined the influence of Δ9-THC on lysosomal membrane permeability and the association of p53 with lysosomes, as well as the role of lysosomal cathepsin proteases in mediating Δ9-THC-induced apoptosis. The findings demonstrate that the neuronal lysosomal system is a novel target for modulation by CBs and that this interaction is pertinent in the CB-mediated induction of neuronal apoptosis.

Materials and methods

Culture of cortical neurones

Primary cortical neurones were established as we have previously described (MacManus et al. 2000). Rats were decapitated in accordance with Institutional and National Ethical Guidelines and cerebral cortices were removed. The dissected cortices were incubated in phosphate-buffered saline (PBS) containing trypsin (0.3%) for 25 min at 37°C. The tissue was then triturated (5×) in PBS containing soyabean trypsin inhibitor (0.1%) and Dnase (0.2 mg/mL) and gently filtered through a sterile mesh filter. Following centrifugation, 2000 g for 3 min at 20°C, the pellet was resuspended in neurobasal medium, supplemented with heat-inactivated horse serum (10%), penicillin (100 U/mL), streptomycin (100 U/mL) and glutamax (2 mM). Suspended cells were plated out at a density of 0.25 × 106 cells on circular 10 mm diameter coverslips, coated with poly-l-lysine (60 μg/mL) and incubated in a humidified atmosphere containing 5% CO2 : 95% air at 37°C. After 48 h, 5 ng/mL cytosine-arabino-furanoside was included in the culture medium to prevent proliferation of non-neuronal cells. Culture media were exchanged every 3 days and cells were grown in culture for 7 days prior to Δ9-THC treatment.

Drug treatment

Δ9-Tetrahydrocannabinol was obtained under license from Sigma-Aldrich Company Ltd. (St. Louis, MO, USA) and diluted to the required concentration with warmed culture media. Absolute alcohol was used as vehicle control. In some experiments, cells were incubated with the CB1 receptor antagonist, AM251 (1-(2,4-dichlorophenyl)-5-(4-iodophenyl]])-4-methyl-N-(1-piperidyl)pyrazole-3-carboxamide; 10 μM) for 30 min prior to Δ9-THC treatment as previously described (Downer et al. 2003). The p53 inhibitor, pifithrin-α (Calbiochem, Darmstadt, Germany) was made up as a 1 mM stock in dimethylsulphoxide and diluted to a final concentration of 100 nM in culture medium. Cells were exposed to pifithrin-α for 60 min prior to Δ9-THC treatment. Pifithrin-α is a cell permeable highly lipophilic molecule that efficiently inhibits p53 phosphorylation (Chua et al. 2006) and p53-dependent transactivation of p53-responsive genes (Culmsee et al. 2001). The cell-permeable cathepsin-D inhibitor octapeptide, H-Gly-Glu-Gly-Phe-Leu-Gly-d-Phe-Leu-OH (Bachem, St. Helens, UK) was stored at −20°C as a 5 mM stock solution in dimethylsulphoxide, and was used at a final concentration of 10 μM. Cells were pre-treated with the cathepsin-D inhibitor for 30 min before exposure to Δ9-THC.

RNA interference

Custom ON-TARGET plus smart pool small interfering RNA (siRNA) containing a mixture of four SMART selection designed siRNAs targeting rat p53 (GenBank™ accession number NM_030989; p53 siRNA) was purchased from Dharmacon (Chicago, IL, USA). Primary cortical neurones were transfected with p53 siRNA (100 nM) using Dharmacon transfection lipid number 3. After 48 h of transfection, cells were treated with THC (5 μM) or vehicle (0.006% ethanol). A control siRNA duplex containing at least four mismatches to any rat gene (ON-TARGET Plus siControl Non-Targeting siRNA; Con siRNA) was used in parallel experiments. Optimal transfection efficiency and conditions were determined by using carboxyfluorescein-labelled non-specific siRNA (SiGlo green; Dharmacon). Effective p53 knockdown was analysed using immunocytochemistry and western immunoblot as we have previously described (Downer et al. 2007b).

Lysosomal localization of phospho-p53

The fluorescent probe, LysoTracker Red (Molecular Probes, Leiden, The Netherlands) was used to visualize lysosomes in intact cells. Cells were exposed to pre-warmed neurobasal medium containing LysoTracker Red (700 nM) for 25 min prior to exposure to Δ9-THC. Cells were then fixed with p-formaldehyde (4%) for 30 min at 37°C, permeabilized with Triton X-100 (0.2%) and re-fixed with 4%p-formaldehyde for 10 min. Cells were incubated overnight at 4°C with a rabbit polyclonal antibody which recognizes p53 phosphorylated on serine 15 (p-p53Ser15; Cell Signalling Technologies, Beverly, MA, USA). Immunoreactivity was detected using a biotinylated goat anti-rabbit IgG. Cells were then incubated with Alexa Fluor 488 avidin-conjugate (1 : 600 dilution in 2.5% serum; Molecular Probes) and viewed under 63× magnification using a confocal microscope (Zeiss LSM 510 META; Carl Zeiss Jena GmgH, Jena, Germany). The multitrack FITC/Rhodamine channel configuration was selected, emission spectra for Alexa 488 (excitation 488 nm and emission 520 nm) and for LysoTracker probe (excitation 543 nm and emission 599 nm).

Lysosomal integrity assay: acridine orange relocation

The lysosomal integrity assay was carried out as described by Yuan et al. (2002). Cells were exposed to pre-warmed supplemented neurobasal medium containing acridine orange (AO; 5 μg/mL; Molecular Probes) for 15 min at 37°C. Cells were rinsed in neurobasal medium and exposed to THC (5 μM) for 5–60 min and viewed by confocal microscopy. Visualization of the fluorophore was achieved using the 488 nm argon laser in the lambda mode where emission over the 499–670 nm range was collected. The configuration parameters were as follows: (i) Filters: Ch3-BP 585–615, Ch2-BP 505–530, ChS1 499.3–670.7 nm; (ii) Beam Splitters: HFT 488; (iii) Scan zoom 1. For each digital image, 512 × 512 pixels were used. The leakage of AO from the lysosome produces a decrease in the 633 nm emission and this parameter was used as an index of lysosomal integrity, as previously reported (Antunes et al. 2001). The average cellular fluorescence at 633 nm was counted from at least 200 cells for each treatment, from at least four independent experiments.

Cathepsin-D localization

To assess the intracellular distribution of cathepsin-D, neurones were incubated with LysoTracker Red (700 nM; Invitrogen, Paisley, UK) for 25 min at 37°C to label lysosomes and BIODIPY FL-pepstatin A (1 μM; Molecular Probes) for 1 h at 37°C to label cathepsin-D. BODIPY FL-pepstatin A is Pepstatin A (isovaleryl-l-valyl-4-amino-3-hydroxy-6-methylheptanoyl-l-alanyl-4-amino-3-hydroxy-6-methlheptanoic acid), covalently conjugated with the Boron dipyrromethene difluoride fluorophore (BODIPY; Chen et al. 2000). Following treatment with Δ9-THC, the cells were fixed with 4%p-formaldehyde and incorporated fluorophores were examined with a confocal microscope (Zeiss, LSM 510 META), as previously described (Chen et al. 2000).

Cathepsin-D activity

Cathepsin-D activity was measured using a fluorogenic assay. Cells were harvested in ice-cold buffer [25 mM HEPES, 5 mM MgCl2, 5 mM EDTA, 8 mM dithiothreitol (DTT) and 2 μg/mL leupeptin, pH 7.5], centrifuged at 10 000 g for 10 min at 4°C to yield the cytosolic fraction. Cathepsin-D was purified from cytosolic fractions using a 96-well plate coated with monoclonal anti-cathepsin-D antibody. The cathepsin-D activity was then detected using an internally quenched fluorescent cathepsin-D substrate peptide, Mca-Gly-Lys-Pro-Ile-Leu-Phe-Arg-Leu-Lys-(Dnp)-d-Arg-NH2. Release of the fluorescent product, Mca-Gly-Lys-Pro-Ile-Leu-Phe was determined fluorometrically at excitation 328 nm and emission 393 nm. Cathepsin-D activity was read from a standard curve of affinity purified cathepsin-D enzyme.

Caspase 3 analysis

Cleavage of the fluorogenic caspase 3 substrate (DEVD-aminofluorocoumarin; Alexis Corporation, San Diego, USA) to its fluorescent product was used to measure caspase 3 activity. Following treatment the cultured neurones were lysed in buffer (25 mM HEPES, 5 mM MgCl2, 5 mM DTT, 5 mM EDTA, 2 mM phenylmethylsulphonyl fluoride, 10 μg/mL leupeptin and 10 μg/mL pepstatin, pH 7.4), sonicated for 2 s and centrifuged at 10 000 g for 10 min at 4°C. Samples of supernatant (90 μL) were incubated with the DEVD peptide (500 μM; 10 μL) for 1 h at 30°C. Incubation buffer (900 μL; 50 mM HEPES, pH 7.4, containing 2 mM EDTA, 20% glycerol and 10 mM DTT) was added and fluorescence was assessed (excitation 400 nm and emission 505 nm). Results are expressed as the fold-change in caspase 3 activity induced by Δ9-THC.

Caspase 3 activity was also assessed by immunocytochemistry using an anti-active caspase 3 antibody (Promega Corporation, Madison, WI, USA). Cells were fixed with 4%p-formaldehyde, blocked with 30% goat serum in PBS overnight at 4°C. Following blocking the cells were incubated in rabbit anti-active caspase 3 (1 : 1000 in 30% blocking buffer; Promega Corporation) for 1 h at 20°C. The primary antibody was detected using a biotinylated goat anti-rabbit secondary antibody (1 : 1500 in 30% blocking buffer). The biotinylated secondary antibody was detected by the avidin conjugated fluorophore Alexa Fluor 488 (1 : 2000 in 30% blocking buffer). Cells were viewed under 40× magnification using a confocal microscope (Zeiss LSM 510 META; Carl Zeiss). The flurophore was visualized using the following scan configurations: excitation 488 nm, 5% argon laser transmission, beam splitters HFT 488, channel filters LP 505, detector gain 652, pinhole 66 μm and 16 scan averages.

Terminal deoxynucleotidyl transferase-mediated UTP-end nick labelling

Apoptotic cell death was assessed using the Dead End™ Fluorometric apoptosis detection system (Promega Corporation). Cells were fixed with p-formaldehyde (4%), permeabilized with Triton X-100 (0.1%) and fluorescein nucleotide was incorporated at 3′-OH DNA ends using the enzyme terminal deoxynucleotidyl transferase (TdT). Fluorescein was visualized by fluorescent confocal microscopy (Zeiss LSM 510-META) using the 488 nm Argon/2 laser with the following scan configurations; laser output 50%, % transmission 5%, band pass filter 505–530, beam splitter 488/543, detector gain 719, amplifier gain 1, amplifier offset −0.14, pinhole 96 μm (1 Airy unit) and 16 scan averages. All cells were counterstained with propidium iodide.


Data are reported as the mean ± SEM of the number of experiments indicated in each case. Statistical analysis was carried out using one-way anova followed by the post hoc Student–Newman–Keuls test when significance (at the < 0.05 level) was indicated. When comparisons were being made between two treatments, an un-paired Student’s t-test was performed and p < 0.05, p < 0.01, or p < 0.001 were considered significant.


Δ9-THC evokes a transient decrease in lysosomal membrane stability

Cortical neurones were loaded with AO and the mean fluorescence intensity at 633 nm emission was observed (Fig. 1). Δ9-THC evoked a dose- and time-dependent reduction in 633 nm emission, reflective of leakage of AO from the lysosomal compartment. Thus, exposure of cells to Δ9-THC at a concentration of 0.5, 5 and 50 μM evoked a significant 70% decrease in the fluorescence signal within 15 min (p < 0.01, anova, n = 6, Fig. 1a) indicating a loss in lysosomal membrane integrity. The inset image shows the pattern of AO staining in control and Δ9-THC-treated cells where the orange punctate staining reflects AO accumulation in the lysosome which emits at 633 nm, while the green diffuse staining represents cytosolic AO. The reduction in 633 nm emission occurred within 15 min of exposure to Δ9-THC (5 μM) and was retained up until 30 min (p < 0.01, anova, n = 6, Fig. 1b). However, by 60 min the fluorescence signal was approaching that of control cells. These data demonstrate that Δ9-THC evokes a rapid, but transient, decrease in lysosomal membrane integrity, which is reflected by an inability of AO to accummulate in the acidic organelles and produce a fluorescence signal at 633 nm.

Figure 1.

 Δ9-THC causes lysosomal rupture in a CB1-dependent manner. (a) Cultured neurones were loaded with acridine orange (AO), exposed to Δ9-THC (0.05–50 μM) for 15 min and viewed by confocal microscopy. The fluorescence emission at 633 nm, representing intact lysosomes, was significantly decreased by 0.5, 5 and 50 μM Δ9-THC. The lower concentration of Δ9-THC, 0.05 μM, had no effect on AO distribution. Results are expressed as mean ± SEM for four independent experiments, *p < 0.05 and **p < 0.01. Inset: AO staining in (i) control cells and (ii) cells exposed to Δ9-THC (5 μM, 15 min). The punctate distribution of orange fluorescence at 633 nm emission indicates a high proportion of cells with intact lysosomes. (b) cultured neurones were loaded with AO and exposed to Δ9-THC (5 μM) for 15–60 min. The fluorescence emission at 633 nm, representing intact lysosomes, was significantly decreased following exposure to Δ9-THC (5 μM) for 15 and 30 min. Results are expressed as mean ± SEM for four independent experiments, *p < 0.05 and **p < 0.01. (c) The Δ9-THC-induced decrease in fluorescence emission at 633 nm (5 μM, 15 min) was reversed by the CB1-receptor antagonist, AM251 (10 μM). Results are expressed as mean ± SEM for four independent experiments, **p < 0.01. (d) sample confocal images of (i) control cells, (ii) cells exposed to Δ9-THC (5 μM, 15 min), (iii) cells exposed to AM251 (10 μM) and (iv) cells exposed to Δ9-THC in the presence of AM251. The punctate distribution of orange fluorescence (633 nm emission) indicates a high percentage of cells with intact lysosomes (i, iii and iv) that are able to retain AO, whilst the diffuse green fluorescence (ii) represents AO within the cytosol as a consequence of lysosomal rupture.

Figure 1c demonstrates that the impact of Δ9-THC on lysosomal membrane integrity is mediated through the CB1 receptor as AM251 (10 μM) abrogated the Δ9-THC-mediated reduction in mean fluorescence intensity at 633 nm. Thus in control cells, mean fluorescence intensity was 122 ± 19 (mean arbitrary units ± SEM) and was significantly decreased to 39 ± 3 in cells exposed to Δ9-THC (5 μM) for 15 min (p < 0.01, anova compared with control cells, n = 6). AM251 alone had no effect on mean fluorescence intensity at 633 nm, but it significantly reversed the Δ9-THC-induced reduction in mean fluorescence intensity (p < 0.01, anova compared with cells exposed to THC, n = 6). Figure 1d illustrates sample AO staining demonstrating the CB1-dependent relocalization of AO to the cytosol following exposure to Δ9-THC (5 μM) for 15 min.

THC increases the association of phospho-p53 at the lysosome

Figure 2 demonstrates that Δ9-THC increases the association of phospho-p53Ser15 with lysosomes. Thus, in control cells phospho-p53Ser15 immunoreactivity was undetectable, and the staining of florescence produced by LysoTracker Red was punctate, indicative of intact lysosomes. However, when cells were exposed to Δ9-THC (5 μM, 15 min) phospho-p53Ser15 immunoreactivity was observed in punctate regions of the cell, some of which was found to co-localize with the lysosomal marker, LysoTracker red. Although we found that by 15 min Δ9-THC had caused the release of AO (Fig. 1b), indicative of increased lysosomal permeability, the LysoTracker red staining was still apparent in Δ9-THC-treated neurons. However, under those conditions the lysosomes were larger [Fig. 2b(ii)] reflecting a change to the lysosomal system (He et al. 2005). The AO protocol is considered to be more sensitive to changes in lysosomal permeability and can monitor very fast granular alkalinization events in live cells (Yuan et al. 2002). In contrast, LysoTracker Red staining was performed in fixed cells and the probe is retained in lysosomal membranes through an as yet uncharacterized mechanism. Thus, LysoTracker Red is considered to be less sensitive to rapid changes in lysosomal permeability and often the loss of LysoTracker staining is not apparent until the very final stages of the apoptotic cascade (Kaasik et al. 2005). In Δ9-THC-treated cells some p53 immunoreactivity was also detected within the nucleus, consistent with a role for p53 in governing transcriptional events (Green and Chipuk 2006). The observation that Δ9-THC promotes the association of phospho-p53 with lysosomes prompted us to examine whether p53 was pertinent in the Δ9-THC-mediated destabilization of the lysosomal membrane.

Figure 2.

 Δ9-THC promotes the association of phospho-p53 with lysosomes. Confocal microscopy was used to visualize the distribution of phospho-p53Ser15 within cortical neurones following treatment with Δ9-THC (5 μM, 15 min). Cells were double labelled with (i) an Alexa-488-conjugated phospho-p53Ser15 antibody and (ii) the lysosomal-specific marker, LysoTracker Red; panel (iii) represents the overlay of phospho-p53Ser15 immunoreactivity with LysoTracker Red. In control cells (a) there was no evidence of phospho-p53Ser15 at the lysosomes. However, 15 min following exposure to Δ9-THC (b) phospho-p53Ser15 immunoreactivity was increased and, in part, co-localized with the lysosomal marker, as indicated by purple staining (iii). In image (iv), phospho-p53Ser15 immunoreactivity at the lysosomes is indicated by purple. Panel (iv) in a zoomed image of lysosomes using 63× objective and scan zoom 6 and panel (v) is a zoomed image of a lysosome, with 63× objective and scan zoom 9, illustrating LysoTracker Red (red) co-localizing with phospho-p53Ser15 immunoreactivity (purple). Images are single 1 μm thick z-sections taken midway through the cell. Arrows indicate regions of co-localization. Scale bar: 10 μm.

p53 plays a role in the THC-mediated destabilization of the lysosomes

Figure 3a demonstrates that the impact of THC on lysosomal membrane integrity is mediated through p53 as the p53 inhibitor, pifithrin-α (100 nM) abrogated the THC-mediated reduction in mean fluorescence intensity at 633 nm. Thus in control cells, mean fluorescence intensity was 118 ± 21 and this significantly decreased to 39 ± 3 in cells exposed to THC (5 μM) for 15 min (p < 0.01, anova compared with control cells, n = 6). In the presence of pifithrin-α, the Δ9-THC-induced reduction in mean fluorescence intensity was significantly attenuated (p < 0.05, anova, n = 6). Sample AO staining demonstrating the p53-dependent relocalization of AO to the cytosol, and concomitant decrease in 633 nm emission, following exposure to THC (5 μM) for 15 min is shown in Fig. 3b. This result was confirmed with knockdown of p53 with siRNA (Fig. 3c) where the Δ9-THC-induced increase in lysosomal membrane permeability, as reflected by a significant reduction in fluorescence emission at 633 nm, was not observed in cells pre-treated with p53 siRNA.

Figure 3.

 Δ9-THC evokes lysosomal destabilization in a p53-dependent manner. (a) Cells were loaded with acridine orange (AO), exposed to Δ9-THC (5 μM, 15 min) and viewed by confocal microscopy. Δ9-THC evoked a significant reduction in fluorescence at 633 nm emission which was abrogated by the p53 inhibitor, pifithrin-α (100 nM). Results are expressed as mean ± SEM for six independent experiments, *p < 0.05 and **p < 0.01. (b) Representative AO staining in (i) control cells, (ii) cells exposed to Δ9-THC (5 μM, 15 min), (iii) cells exposed to pifithrin-α (100 nM) and (iv) cells exposed to Δ9-THC in the presence of pifithrin-α. The punctate distribution of orange fluorescence (633 nm emission) indicates a high percentage of cells with intact lysosomes. In Δ9-THC-treated cells (ii) the punctate distribution of orange fluorescence was reduced and increased diffuse green fluorescence was observed, which was abrogated by pifithrin-α (iv). Arrows indicate cells with reduced punctate orange staining and increased diffuse green fluorescence, indicative of lysosomal rupture and redistribution of AO to the cytosol. C, siRNA knockdown of p53 prevents the Δ9-THC-induced decrease in lysosomal integrity. Exposure of neurones to Δ9-THC (5 μM) for 15 min in the presence of control siRNA significantly decreased the fluorescence emission at 633 nm, indicative of lysosomal rupture. Treatment with p53 siRNA (100 nM; 48 h) prior to Δ9-THC treatment prevented the THC-induced decrease in fluorescence emission at 633 nm. Results are expressed as mean ± SEM for six observations, ***p < 0.001.

Δ9-THC evokes a redistribution of cathepsin protease

A destabilization of the lysosomal membrane would be expected to promote the release of lysosomal constituents into the cytosol. Cultured neurones were incubated with BODIPY®-FL pepstatin-A to label cathepsin-D, and LysoTracker Red to observe lysosomes. Figure 4a demonstrates that the BODIPY®-FL pepstatin-A co-localizes with lysosomes in untreated cells. When neurones were incubated with Δ9-THC (5 μM, 15 min) the punctate distribution of BODIPY®-FL pepstatin-A, reflective of cathepsin-D localization within lysosomes, changed to a diffuse pattern of staining, representative of the presence of cathepsin-D in the cytosol (Fig. 4b). This result demonstrates that the increase in lysosomal permeability is associated with a redistribution of cathepsin-D from the lysosomal compartment into the cytosol. Furthermore, exposure of neurones to Δ9-THC (5 μM, 15 min) evoked a significant increase in the activity of cathepsin-D (p < 0.01, anova compared with vehicle-treated cells, n = 6) and this was significantly reduced by pifithrin-α (100 nM, p < 0.01, anova compared with THC-treated cells, n = 6; Fig. 4c).

Figure 4.

 Δ9-THC increases the cytosolic expression and activity of cathepsin-D. BIODIPY FL-pepstatin-A fluorescence was used to label intracellular cathepsin-D localization. (a) BIODIPY FL-pepstatin-A co-localized with LysoTracker Red in untreated cells. (i) BIODIPY FL-pepstatin-A fluorescence, (ii) LysoTracker Red fluorescence and (iii) co-localization of BIODIPY FL-pepstatin-A and LysoTracker Red is indicated by yellow fluorescence. Scale bar: 10 μm. (b) Cells were exposed to (i) vehicle control or (ii) Δ9-THC (5 μM) for 15 min and BIODIPY FL-pepstatin-A fluorescence was used to monitor intracellular cathepsin-D localization. In control cells, BIODIPY staining was punctate, indicative of compartmentalization within lysosomes. In Δ9-THC-treated cells, BODIPY staining was present throughout the cytosol, reflecting the redistribution of cathepsin-D from the lysosomes to the cytosol. Scale bar: 5 μm. (c) Δ9-THC (5 μM; 15 min) evoked a significant increase in cathepsin-D activity and this was abolished by pifithrin-α (100 nM). Results are expressed as mean ± SEM for six independent observations, **p < 0.01.

Δ9-THC-induced apoptosis involves cathepsin-D

We have previously reported that Δ9-THC induces the activation of pro-apoptotic caspase 3 (Campbell 2001). The Δ9-THC-induced activation of caspase 3, as determined by expression of anti-active caspase 3 immunoreactivity (Fig. 5a) and fluorogenic assay (Fig. 5b) was prevented by the cathepsin-D inhibitor peptide (10 μM). Thus, exposure to Δ9-THC (5 μM, 1 h) caused an increase in expression of active-caspase 3 immunoreactivity and evoked a 2.28 ± 0.27-fold increase in caspase 3 activity (p < 0.05, anova, n = 6). In the presence of the cathepsin-D inhibitor, the Δ9-THC-induced increase in active-caspase 3 immunoreactivity (Fig. 5a) and caspase 3 activity (Fig. 5b; p < 0.01, anova, n = 6) was abolished. This result suggests that the impact of Δ9-THC on the lysosomal system promotes a redistribution of cathepsin-D which in turn contributes to the activation of caspase 3. The Δ9-THC-induced increase in apoptotic cell death, as assessed by TdT-mediated UTP-end nick labelling staining (Fig. 5c) was also prevented by the cathepsin-D inhibitor, further demonstrating that this lysosomal protease is involved in the apoptotic cell death evoked by Δ9-THC. In neurones exposed to the cathepsin-D inhibitor alone, a significant increase in DNA fragmentation was observed and this indicates that cathepsin-D may have a pro-survival role in this cell type. However, the lack of any further increase in DNA fragmentation in cells exposed to Δ9-THC in the presence of cathepsin-D inhibitor, would lend support of a role for cathepsin-D in Δ9-THC-iduced neuronal apoptosis.

Figure 5.

 The Δ9-THC increase in caspase 3 activity and DNA fragmentation is dependant upon cathepsin-D. (a) Representative active-caspase 3 immunoreactivity in (i) control cells, (ii) cells exposed to Δ9-THC (5 μM, 1 h), (iii) cells exposed to cathepsin-D inhibitor (10 μM) and (iv) cells exposed to Δ9-THC in the presence of cathepsin-D inhibitor. THC evoked an increase in active-caspase 3 immunoreactivity (ii; indicated by arrows), and this was prevented in cells exposed to Δ9-THC in the presence of the cathepsin-D inhibitor (iv). The left side of each image is active-caspase 3 immunofluorescence and the corresponding phase contrast image is shown on the right. Scale bar: 10 μm. (b) Δ9-THC (5 μM, 1 h) evoked a twofold increase in caspase 3 activity and this was significantly reduced by the cathepsin-D inhibitor (10 μM). Results are expressed as mean fold-increase in caspase 3 activity evoked by Δ9-THC ± SEM for six independent observations, ***p < 0.001. (c) Δ9-THC (5 μM, 3 h) evoked a significant increase in DNA fragmentation and this was prevented by the cathepsin-D inhibitor (10 μM). Results are expressed as mean ± SEM for six independent observations, ***p < 0.001 and +++p < 0.001 (compared with control cells, anova, n = 6).


The aim of this study was to examine the impact of the phytocannabinoid, Δ9-THC, on lysosomal membrane permeability and to assess whether lysosomes play a role in the Δ9-THC-induced activation of the apoptotic pathway previously reported (Campbell 2001; Downer et al. 2003, 2007a). The results demonstrate that Δ9-THC causes a dose- and time-dependent destabilization of the lysosomal membrane via activation of the CB1 receptor. Δ9-THC evoked co-localization of p53, phosphorylated at Ser15, with the lysosome and this association is likely to contribute to the destabilization of the lysosomal membrane as both the p53 inhibitor, pifithrin-α and siRNA knockdown of p53 reversed the Δ9-THC-induced loss in lysosomal integrity. The transient nature of this event suggests that lysosomal constituents may play a necessary role in orchestrating downstream cellular events associated with apoptosis. In support of this, Δ9-THC caused an increase in expression of the aspartic protease, cathepsin-D, in the cytosol, and the cathepsin-D inhibitor abrogated that Δ9-THC-induced activation of caspase 3 and DNA fragmentation. These data provide evidence for a lysosomal branch of the apoptotic programme induced by Δ9-THC in neurones.

Lysosomes are emerging as important regulators of the cell death cascade. Originally thought to be stable organelles, only becoming destabilized during the end stages of cell death, there is a large body of evidence demonstrating that lysosomes occupy an upstream regulatory role in apoptosis (Li et al. 2000; Brunk et al. 2001). Early in apoptosis lysosomes become permeable causing the translocation of lysosomal cathepsin proteases into the cytosol (Werneburg et al. 2002; Kagedal et al. 2005). The redistribution of cathepsins is suspected to be an important initiating event in apoptosis (Reiners et al. 2002) and the release of lysosomal enzymes may cause changes in mitochondrial permeability directly (Zhao et al. 2001) or indirectly (Stoka et al. 2001), followed by cytochrome c release, apoptosome formation with Apaf-1 and caspase activation. There may also be a direct activation of caspases by lysosomal cathepsins (Vancompernolle et al. 1998). Furthermore, cathepsin-D, both in its mature or inactive form, can impact on an as yet unidentified substrate to induce apoptosis (Schestkowa et al. 2007). In this study, the lysosomal destabilization occurred rapidly following exposure to Δ9-THC. Furthermore, the destabilization of lysosomes observed at 15 min occurs prior to the onset of activation of caspase 3 and preceeds induction of morphological features of apoptosis, which we have previously found to occur at a later time point when neurones are exposed to the same concentration of Δ9-THC to that used in the current study (Campbell 2001; Downer et al. 2001). The early lysosomal destabilization, coupled with the transient nature of this event, would suggest that the lysosomes have a role in orchestrating CB-mediated neuronal death. In neurones, the lysosomal permeabilization evoked by Δ9-THC is likely mediated through the CB1 receptor, although a role for CB2 receptors cannot be completely excluded as in macrophages Matveyeva et al. (2000) have demonstrated that Δ9-THC increases cathepsin-D activity via the CB2 receptor which may contribute to deficits in antigen-dependent processing. Early reports in the literature describing the lytic effect of Δ9-THC on isolated liver lysosomes and subsequent release of acid hydrolases (Britton and Mellors 1973) would indicate that Δ9-THC, which is highly lipophilic, may directly influence lysosomal permeability in some cell types.

Following exposure to Δ9-THC, an increase in expression of the phosphorylated form of the apoptotic effector, p53, was observed, which may reflect stabilization of the protein by c-jun N-terminal kinase (Downer et al. 2007b). Our previous studies have demonstrated that Δ9-THC evokes a rapid increase in phospho-p53 expression in cell lysates and that p53 knockdown with siRNA protects neurones from Δ9-THC-induced apoptosis (Downer et al. 2007b). In the current study, we found that Δ9-THC promotes association of the p53 protein with the nucleus where it is likely to participate in the transcription of apoptotic genes (Green and Chipuk 2006). However, a substantial proportion of p53 immunoreactivity co-localized with the lysosomal marker, LysoTracker Red, indicating the association of p53 with the lysosomal compartment. To assess whether p53 regulates lysosomal integrity we used the reversible p53 inhibitor, pifithrin-α, which prevents p53 transactivation (Komarov et al. 1999; Culmsee et al. 2001) and p53 phosphorylation (Chua et al. 2006), as well as siRNA-mediated depletion of p53. Using both experimental approaches the THC-induced permeabilization of lysosomes was prevented, providing evidence of a role for p53 in controlling lysosomal permeability. To the authors’ knowledge, this is the first demonstration that p53 directly associates with lysosomes during an apoptotic cascade, although the lysosomal targets that may be regulated by p53 to control lysosomal integrity remain to be elucidated. Indeed, there is currently no consensus on the mechanisms that are responsible for control of lysosomal membrane destabilization during apoptosis (Stoka et al. 2007). Cytoplasmic p53 can directly activate pro-apoptotic members of the Bcl family of proteins to cause mitochondrial permeabilization and apoptosis (Chipuk et al. 2005), and given that Bax can directly insert into the lysosomal membrane during staurosporine-induced apoptosis (Kagedal et al. 2005), it is interesting to speculate that functional crosstalk exists between p53 and Bax in relation to the control of lysosomal permeability. Other members of the Bcl family, such as Bcl-2, block oxidative stress-induced apoptosis by stabilizing lysosomes (Zhao et al. 2001). A number of other signalling molecules are pertinent in the regulation of lysosomal dynamics, such as intracellular sphingosine (Kagedal et al. 2005), and this is significant given that CBs couple to ceramide production to evoke apoptosis in glioma cells (Galve-Roperh et al. 2000). Another explanation for the triggering of lysosomal rupture implicates reactive oxygen species as oxidative stress can induce lysosomal destabilization very rapidly, resulting in the release of cathepsins, in vitro (Kalra et al. 1989) and in vivo (Ollinger and Brunk 1995). Lysosomes are particularly vulnerable to oxidative stress as they contain the most important pool of reactive iron in the cell (Antunes et al. 2001), and again it is significant that CBs are coupled to the generation of reactive oxygen species in neurones (Chan et al. 1998) and glia (Massi et al. 2006).

Using a BODIPY-conjugated cathepsin-D inhibitor, we demonstrate that Δ9-THC causes a redistribution of cathepsin-D from the lysosome to the cytosol, and this correlated with an increase in activity of cathepsin-D. The impact of Δ9-THC on cathepsin-D was blocked by the CB1 antagonist, AM251, as well as pifithrin-α. Lysosomal rupture and subsequent release of cathepsin-D have been reported to occur prior to changes in mitochondrial membrane potential (Turk et al. 2002) in a number of apoptotic pathways (Stoka et al. 2007). It is likely that the release of cathepsin-D is a regulated process and not a consequence of general lysosomal rupture because the increase in cytosolic cathepsin-D that we observe in Δ9-THC-treated cells does not coincide with an increase in cytosolic activity of other cathepsins, such as cathepsin-L (data not shown). Our observation that the Δ9-THC-induced increase in caspase 3 activity and DNA fragmentation was blocked by the cathepsin-D inhibitor peptide provides further support for a lysosomal involvement in this CB-induced apoptotic cascade. A link between cathepsin-D and caspase 3 has been reported in staurosporine-induced apoptosis in fibroblasts via a Bid-signalling pathway (Johansson et al. 2003), although pathways other than Bid are also likely to signal between cathepsin-D and caspase 3 to evoke apoptosis (Houseweart et al. 2003). Our in vivo studies have recently demonstrated that administration of Δ9-THC increases the cytosolic expression and activity of cathepsin-D in the cerebral cortex as part of an apoptotic cascade (Downer et al. 2007a) and the results presented herein would indicate that p53-mediated regulation of lysosomal integrity is a key factor in this neurotoxic event. Given that Δ9-THC has been reported to evoke apoptosis in other cell types such as gliomas (Velasco et al. 2004), leukaemic cells (Jia et al. 2006) and airway epithelia (Sarafian et al. 2005), it is interesting to speculate that those CB-induced apoptotic pathways also involve a lysosomal component.

While the potential neurotoxicity associated with recreational use of cannabis is controversial, some imaging studies of chronic cannabis users have identified reductions in grey and white matter volume (Scallet 1991; Lawston et al. 2000; Tzilos et al. 2005; Schlaepfer et al. 2006). Δ9-THC and other CBs have been shown to evoke apoptosis in neuronal (Chan et al. 1998; Downer et al. 2003) and glial (Galve-Roperh et al. 2000) cells. Potential mechanisms of apoptosis include formation of ceramide (Galve-Roperh et al. 2000), stress-activated protein kinase activity (Downer et al. 2003), calpain activation (Movsesyan et al. 2004), and engagement of the TRPV1 channel (Ligresti et al. 2006). The pro-apoptotic effect of the CB system is particularly robust in transformed cells and some studies have demonstrated that CBs do not induce apoptosis in healthy tissue (Carracedo et al. 2006; Massi et al. 2006); a property that may be harnessed in the development of a CB-based therapy for the selective eradication of tumour cells. It remains to be established whether the p53/lysosomal pathway is pertinent in the cannabinoid-induced apoptosis of transformed cells.

The current study demonstrates that in neurones Δ9-THC causes lysosomal destabilization early in the apoptotic cascade in a manner that is dependent upon p53. This novel pathway may reflect a role of the endocannabinoid system in brain development (Harkany et al. 2007), in which physiological apoptosis is a feature. In this regard, it is notable that cathepsin-D gene expression outlines the regions of physiological cell death during embryonic development (Zuzarte-Luis et al. 2007). Also, the engagement of this apoptotic pathway by exogenous phytocannabinoids during cannabis abuse may contribute to the profound volumetric changes in the brain that have been reported (Schlaepfer et al. 2006). Neuronal apoptosis evoked by Δ9-THC would also be expected to interfere with the physiological role of the pre-synaptic CB1 receptor in controlling neurotransmitter release and this is likely to be of importance during brain development (Galve-Roperh et al. 2006).

In summary, this study identifies a novel lysosomal branch to the CB-mediated induction of apoptosis involving the tumour suppressor protein, p53. Given the interest in the ability of CBs to regulate cell fate (Guzman 2005), this pathway may be important for the anti-tumoural properties of CBs, as well as being involved in the control of neural cell viability.


This work was funded by Science Foundation Ireland.