Radiol Oncol 2006; 40(2): 73-85.
review
The role of tricyclic drugs in selective triggering of
mitochondrially-mediated apoptosis in neoplastic glia:
a therapeutic option in malignant glioma?
Geoffrey J. Pilkington1, James Akinwunmi2 and Sabrina Amar1
1Cellular & Molecular Neuro-Oncology Group, School of Pharmacy & Biomedical Sciences,
Institute of Biomedical & Biomolecular Sciences, University of Portsmouth, Portsmouth and
2Hurstwood Park Neurological Centre, Haywards Heath West Sussex, UK
We have previously demonstrated that the tricyclic antidepressant, Clomipramine, exerts a concentration-dependant, tumour cell specific, pro-apoptotic effect on human glioma cells in vitro and that this effect is not mirrored in non-neoplastic human astrocytes. Moreover, the drug acts by triggering mitochondrially-media-ted apoptosis by way of complex 3 of the respiratory chain. Here, through reduced reactive oxygen species and neoplastic cell specific, altered membrane potential, cytochrome c is released, thereby activating a cas-pase pathway to apoptosis. In addition, while we and others have shown that further antidepressants, in-cluding those of the selective serotonin reuptake inhibitor (SSRI) group, also mediate cancer cell apoptosis in both glioma and lymphoma, clomipramine appears to be most effective in this context. More recently, oth-er groups have reported that clomipramine causes apoptosis, preceded by a rapid increase in p-c-Jun levels, cytochrome c release from mitochondria and increased caspase-3-like activity. In addition to clomipramine we have investigated the possible pro-apoptotic activity of a range of further tricyclic drugs. Only two such agents (amitriptyline and doxepin) showed a similar, or better, effect when compared with clomipramine. Since both orally administered clomipramine and amitriptyline are metabolised to desmethyl clomipramine (norclomipramine) and nortriptyline respectively it is necessary for testing at a tumour cell level to be car-ried out with both the parent tricyclic and the metabolic product. In addition, reversal of multidrug resistance in a number of solid cancers following treatment with both clomipramine and amitriptyline has been re-ported. This additional role for tricyclics may be of some significance in the treatment of primary and secon-dary brain tumours. Since a substantial number of patients with malignant glioma have already received and are receiving clomipramine, both anecdotally and within a clinical trial, we have carried out CYP (P450) gene expression studies and determined blood plasma levels of clomipramine and norclomipramine, in order to ascertain whether differences in individual patient metabolism influence clinical outcome. While the pro-apoptotic effect of norclomipramine appears to be inferior to that of the parent tricyclic, amitriptyline and nortriptyline share a similar propensity for eliciting apoptosis in neoplastic but not non-neoplastic astroc-ytes. The potential value of these agents as adjuvants in the management of patients with malignant glioma is apparent.
Key words: brain neoplasms – drug therapy; glioma; clomipramine; antidepressive agents, tricyclic; apop-tosis
74
Pilkington GJ et al. / Tricyclic drugs in treatment of malignant glioma
Received 24 May 2006 Accepted 30 May 2006
Correspondence to: Professor G.J. Pilkington, Cellular and Molecular Neuro-oncology Group, Institute of Bi-omedical and Biomolecular Sciences, School of Pharmacy and Biomedical Sciences, University of Portsmouth, White Swan Road, Portsmouth, Hants, PO1 2DT, UK Tel: +44 (0) 23 9284 2123, Fax: +44 (0) 23 9284 2118, e-mail: Geoff.Pilkington@port.ac.uk, Website: www.port.ac.uk/brainlab
Introduction
Tricyclic drugs and neoplastic cells
Tricyclic drugs, whose name is derived from their characteristic three ring nucleus (Table 1), were first thought to be useful as antihista-mines with sedative properties and later as anti-psychotics. They include an important group of tricyclic antidepressants (TCAs) which have been in clinical use over 40 years. In the 1970’s, it was found that TCAs showed selective inhibition of mitochondrial activity in yeast cells.1 It was surmised that the wide range of actions shown by the TCAs in vivo was due to interactions with membranes and membrane bound enzymes, in particular the mitochondrial membrane1, resulting in inhibition of cellular respiration and limitation of adenosine triphosphate (ATP) production. Further experiments showed that cancer cells were much more susceptible to the inhibitory effects of TCAs than non-transformed cells.2 After treatment with TCAs, it was observed that the respiration rate of transformed cells was significantly less than their normal coun-terparts in oxygen electrode studies.2 It was concluded that anti-mitochondrial drugs, such as TCAs, depress mitochondrial activity in cancer cells, thereby leading to cell death, whereas non-transformed cells were able to recover after treatment.2 This mode of action of the TCAs was found to be a common feature amongst members of the group but there appears to be no clear relationship between chemical   structure   and   pharmacological
Radiol Oncol 2006; 40(2): 73-85.
action.3 However, the chlorine containing drugs are said to be more toxic than others to the functions of the mitochondrial membrane.4
Impairments of mitochondrial function may lead to ATP depletion and necrotic cell death.5 More recently, however, mitochon-dria have been implicated in both the regulation of apoptotic cell death and cancer for-mation.3 It has been reported that mito-chondrial respiration is decreased in neo-plastic tissue, along with a lowering of the cellular content of mitochondria. These fin-dings indicate that tumour cells rely upon glycolysis as an energy source and this ena-bles them to survive under hypoxic conditi-ons.6 There are at least three established mechanisms through which mitochondria can trigger apoptosis although these events may be inter-related.7 Apoptosis may be triggered by disruption of electron transport, oxidative phosphorylation and ATP transport, release of proteins that trigger ac-tivation of caspases and alteration of cellu-lar redox potential.7 A number of agents ap-pear to target the mitochondria and promo-te the release of cytochrome c and other pro-apoptotic proteins, which can trigger caspase activation resulting in cell death.5 Caspases are cysteine proteases and exist in a latent state in ‘healthy’ cells.8 In response to damage or a malfunction of vital metabo-lic processes, cells generate signals that lead to activation of caspases, which result in apoptotic cell death.8 Figure 1 shows some of the signalling pathways involved in tric-yclic-initiated, mitochondrially-mediated neoplastic cell apoptosis.
Defects in apoptosis signaling pathways are, however, common in cancer cells. Mo-reover, tumour development, progression and resistance to radiotherapy and chemo-therapy are all the direct result of defects in the regulation of apoptosis in glioma9, due to raised apoptotic thresholds. Human mi-tochondrial DNA (mtDNA) consists of a small circular genome of 165kb that enco-
Pilkington GJ et al. / Tricyclic drugs in treatment of malignant glioma
75
Table 1. Chemical structure of the tricyclic antidepressants used in laboratory studies
Name          Chemical
structure
Chemical structure of clomipramine hydrochloride
Chemical structure of norclomipramine hydrochloride (N-desmethylclomipramine)
Chemical structure of amitriptyline hydrochloride
Chemical structure of doxepin hydrochloride
des a complex array of proteins including 13 respiratory chain sub-units. Expression of the entire genome is required to mainta-in proper function of the mitochondria. The identification of the specific proteins responsible for the regulation of apoptosis may be expected to lead to the development of cancer therapies directed at altering the levels of expression of pro-apopto-tic proteins and enhancing the effects of current radiotherapy and chemotherapy. The bcl-2 proto-oncogene represses a num-ber of cellular apoptotic pathways and is
known to be expressed in increasing amo-unts in glial tumours with increasing de-gree of malignancy.10 Transfection of glio-ma cells with antisense bcl-2 has been re-ported to result in an increase in apoptotic cell death. This indicates that bcl-2 plays a role in tumour progression of gliomas by acting as an oncogene and inhibition of the bcl-2 gene could have a therapeutic effect.10 It has been determined that chemothera-peutic drug-induced apoptosis of human malignant glioma cells involves the death receptor-independent activation of caspa-Radiol Oncol 2006; 40(2): 73-85.
76
Pilkington GJ et al. / Tricyclic drugs in treatment of malignant glioma
Figure 1. Flow pathway of clomipramine pro-apoptotic effect.
ses other than 3 and 8.11 Caspases 1, 2, 3, 7, 8 and 9 are constitutively expressed in most human malignant glioma cell lines and drug-induced apoptosis involves dela-yed activation of caspases 2, 7 and 9 and is blocked by a broad spectrum caspase inhi-bitor.11 It has also been established that the cytotoxic effects of many chemotherapeu-tic agents are mediated via apoptotic path-ways; therefore developing drugs that target the mitochondria may provide a new strategy to induce apoptosis in tumour cells. 12
It has been shown that the TCAs imipra-mine and clomipramine, and the selective serotonin reuptake inhibitor (SSRI) citalo-pram, induce apoptosis in cancer cells and that this process is associated with an ear-Radiol Oncol 2006; 40(2): 73-85.
ly increase in the production of reactive oxygen species (ROS) and subsequent loss of mitochondrial membrane potential.13 The literature suggests that TCAs can in-duce apoptosis in acute myeloid leukemic cells14 and lymphomas15 as well as glio-mas.3,16,17 The mechanism of action of clo-mipramine involves the inhibition of com-plex III of the respiratory chain, resulting in elevated levels of ROS, cytochrome c release and caspase-activated apoptosis.16 Indeed the data presented in a study carri-ed out by Daley et al. indicated that clomi-pramine might be useful in the treatment of patients with primary brain tumours.16 In fact it is estimated that there are now over 300 ‘anecdotal’ cases of patients with a range of different primary brain tumours
Pilkington GJ et al. / Tricyclic drugs in treatment of malignant glioma
77
who are taking, or have taken, clomiprami-ne in the UK. With respect to these cases, there have been numerous reports of survi-val benefit and increased quality of life. Currently, there is a clinical study in pro-gress in which patients newly diagnosed with either an anaplastic astrocytoma or glioblastoma multiforme receive an initial dose of 25mg clomipramine, escalating to 150mg in steps of 25mg at 3-day intervals.3
In addition, reversal of multidrug resistance in a number of solid cancers follo-wing treatment with both clomiprami-ne18,19 and amitriptyline20 has been repor-ted. This additional role for tricyclics may, albeit at differing concentrations, provide an additional novel approach to the trea-tment of primary and secondary brain tu-mours.
In order to address some of the issues re-lated to a possible further role of TCAs in the therapy of glioma we are carrying out the following studies:
Clinical
•  Determination of blood plasma levels of clomipramine and its metabolite, norclo-mipramine, in patients with brain tumour taking the drug.
•  Assessment of CYP (P450) gene expression in glioma patients taking clomipramine.
•  Monitoring of outcome of »anecdotal« gli-oma patients treated with clomipramine through the Samantha Dickson Research Trust (www.sdrt.co.uk).
•  Clinical trial at King’s College Hospital, London in newly diagnosed patients with histologically verified anaplastic astroc-ytoma and glioblastoma multiforme.
Laboratory
•  Assessment of viability in a dose response series of in vitro experiments in low passage, biopsy derived glioma cultures, high passage, established glioma cell lines and non-ne-oplastic human astrocytes to clomipramine.
•  Assessment of oxygen utilisation of the abo-ve cells after treatment with clomipramine.
•  Assessment of apoptosis of the above cells after treatment with clomipramine.
•  Repeating the above studies with norclomi-pramine, amitriptyline, nortriptyline and va-rious combinations of amitriptyline and clo-mipramine.
•  Establishing the possible influence of diffe-rent concentrations of dexamethasone on clomipramine-induced apoptosis.
Methods used in ongoing laboratory studies
Blood samples/clomipramine distribution
Blood plasma samples taken at regular in-tervals, from both anecdotal and trial pati-ents taking clomipramine, are analysed us-ing standard high-performance liquid chro-matography (HPLC), to detect both clomi-pramine and its metabolite norclomiprami-ne. A methodology is currently being deve-loped for the measurement of dexametha-sone via HPLC, and amitriptyline/nortrip-tyline can potentially be added to the range of tricyclics that we are able to offer testing for, should it be required. The data taken from the analysis of blood plasma will be used to track the metabolic progress of in-dividual patients, and over a period of months could be used to ‘tailor’ the indivi-dual dose according to side effects. The preliminary studies that have been carried out are based upon the therapeutic window for patients using clomipramine as an anti-depressant, however when enough data is gathered it will be possible to determine the target range for use in malignant glio-ma. In combination with the plasma testing of patients, it will be possible for a series of basic liver function tests (aspartate amino-transferase (AST), alanine aminotransfera-se (ALT) and gammaglutamyltransferase (GGT)) to be performed ‘in-house’.
Radiol Oncol 2006; 40(2): 91-103.
78
Pilkington GJ et al. / Tricyclic drugs in treatment of malignant glioma
Blood samples taken from patients in-cluded in the above studies are also analy-sed for the presence/absence of markers related to the metabolism of clomipramine. DNA extracted from Whatman FTA cards is analysed by PCR using primers for the CYP genes 2D6 and 2C19. By determining the gene expression of the individual patient it will be possible to classify them as ‘good’ or ‘poor’ metabolisers of clomipra-mine and this information will be of use when clinical decisions are taken concer-ning the optimal daily dose.
Treatment of cells
Neoplastic and non-neoplastic glial cells are used for treatment with the following agents: amitriptyline, nortriptyline, clomi-pramine, norclomipramine, dexamethaso-ne & sodium valproate (valproic acid). These experiments will show if there is any synergy, additive effect or antagonism bet-ween agents in combination.
Cell viability
Cell viability is used, in conjunction with clonogenicity assays to determine the effi-cacy of the drugs in vitro. Studies are also performed using normal human astrocytes (Cambrex Biosciences) to demonstrate that the tricyclic drugs affected only neoplastic cells in the brain. The MTT, Neutral Red and Alamar Blue cytotoxicity assays are used to initially determine the IC50 for each of the agents, and then subsequent studies are performed using pertinent concentrati-ons of the tricyclics. Using a Beckman Co-ulter Vi-Cell XR trypan blue analyzer cells exposed to test agents for varying lengths of time can be analysed to determine per-centage cell death, via uptake of trypan blue. The instrument also provides information about viability of different sub-po-pulations based upon cell size after drug
Radiol Oncol 2006; 40(2): 73-85.
exposure and is used to prepare growth curves and population doubling times via its »Bioprocess« software programme.
Oxygen electrode assay
Oxygen electrode studies using Hansatech multiple Oxytherm/Oxygraph O2 electrode apparatus are performed to establish any decrease in oxygen uptake on tumour cell exposure to the test agents. Reduction in oxygen utilisation gives an indication of the affects of test agents on mitochondrial function and is a useful indicator of events culminating in mitochondrially-mediated apoptotic cell death.
Apoptosis assays
Annexin V/Propidium iodide flow cytome-try: is used to determine the mechanism of cell death subsequent to exposure to the test agent by way of a BD FACScalibur flow cyto-meter. The annexin V fluorochrome binds to the ‘flipped’ phosphotidyl serine residues of the inner leaflet of the cell membrane, after the apoptotic signalling cascade has been ac-tivated. The assay can differentiate between early and late apoptotic cells as well as ne-crotic cells. The protocol had to be optimi-sed when using it to study the tricyclics in combination with dexamethasone as the propidium iodide, which is taken up by the ‘leaky’ cell membrane of necrotic cells, is also taken up due to the effect that dexame-thasone exerts on the pores of the cell membrane via the glucocorticoid receptors.
Live cell imaging: with monolayers of tu-mour cells is carried out over periods of up to 72 hours of drug exposure using a Zeiss Axiovert 200M incorporating a temperatu-re/humidity/CO2 controlled chamber toge-ther with Improvision Openlab, Velocity and Acquisition software. Cell proliferation and apoptotic events are recorded as time-lapse DVDs.
Pilkington GJ et al. / Tricyclic drugs in treatment of malignant glioma
79
Change in Oxygen Utilisation of Cell Line IFSB-18 Following Addition of Various Concentrations
of Amitriptyline
O.lmM
0.25mM
0.5mM
0.75mM
l.OmM
Control
Time Points
Figure 2. Graph representing the oxygen consumption of an anaplastic astrocytoma after treatment with different concentrations of Clomipramine.
Caspase 3 Activity: Caspase 3 activity is measured by its ability to cleave Ac-DEVD-AMC, whereby it produces a fluorescent AMC subunit. Cytosol extracted from cells exposed to the test agents, are read on a Mithras LB950 plate reader (Berthold Technologies) and compared to controls in which a pan-caspase inhibitor is added.
Results
Additional tricyclics and neoplastic cell apoptosis
In addition to clomipramine we have, in pilot experiments, investigated the possible pro-apoptotic activity of a range of further tricyclic drugs. Only two such agents (ami-triptyline and doxepin) showed a similar (doxepin), or better (amitriptyline), effect when compared with Clomipramine. Ami-triptyline has previously been reported to reduce proliferation in cancer cell lines.21 Preliminary studies carried out using ami-triptyline and nortriptyline show that it ex-erts a cytotoxic effect on the established
anaplastic astrocytoma line (IPSB-18 p39) and the glioblastoma-derived culture (CLOM 15 p23). We have also found that Amitriptyline induces a dose-dependent re-duction in oxygen utilisation in human gli-oma cells (Figure 2) as well as apoptosis as seen in Annexin V/PI flow cytometry. Mo-reover, when early passage cultures of human glioma were treated sequentially with clomipramine & amitriptyline apoptosis was initiated. Only a small proportion of cells recovered from this treatment
The possible role of dexamethasone in modulation of tricyclic drug-mediated brain tumour cell apoptosis
In the UK the great majority of patients suf-fering from malignant glioma receive the glucocorticoid steroid, dexamethasone, to reduce raised intracranial pressure.22 This steroid has been reported to be both anti-and pro-apoptotic, in its own right, accor-ding to concentration, in various cancer cells.23 In addition, it has been shown to pro-tect established glioblastoma-derived cultu-Radiol Oncol 2006; 40(2): 73-85.
80
Pilkington GJ et al. / Tricyclic drugs in treatment of malignant glioma
res from temozolomide-induced apoptosis by influence on caspase-3 activity and Bax: Bcl-2 ratio.24,25 In our laboratories when studying concomitant dexamethasone/clo-mipramine treatment of glioma cells and de-xamethasone pre-treatment prior to clomi-pramine treatment, we were able to demon-strate both inhibition and potentiation of clomipramine-mediated apoptosis.26 These studies, however, merit greater investigation using different combinations and doses pri-mary and early passage glioma-derived cul-tures as well as established cell lines.
Clinical studies with TCAs in brain tumour
Since a substantial number of patients with malignant glioma have already recei-ved and are receiving clomipramine, both anecdotally and within a clinical trial at King’s College Hospital, London we are ca-rrying out two experiments. One to determine the CYP (P450) genetic profile of in-dividuals and the other to determine blood plasma levels of clomipramine and norclo-mipramine, in order to determine whether differences in individual patient metabo-lism influences clinical outcome. CYP (450) are hydroxylases situated on the P450 loci and are responsible for the bre-akdown of antidepressant in particular CYP2C19 and CYP2D6, which are highly polymorphic.
The first of these was to determine the CYP (P450) (27, 28) genotypic profile of in-dividuals, in particular the CYP2D6 and CYP2C19 and the other was to test blood plasma levels of clomipramine and norclo-mipramine in order to determine whether differences in individual patient metabo-lism, measured by HPLC analysis, influen-ces clinical outcome. We now wish, in col-laboration with our clinical colleagues, to extend these studies in order to obtain sta-tistically meaningful data with which to inform clinical practice.
Radiol Oncol 2006; 40(2): 73-85.
Discussion and future studies
The pro-apoptotic role of clomipramine in neo-plastic cells
Clomipramine acts by triggering mitochon-drially-mediated apoptosis by way of com-plex 3 of the respiratory chain. Here, thro-ugh reduced reactive oxygen species and neoplastic cell specific, altered membrane potential, cytochrome c is realeased thereby activating a caspase pathway to apoptosis (Figure 1).16 Indeed, Xia et al.13,14 previously reported that clomipramine induced increa-ses in reactive oxygen species, lead to mito-chondrial membrane potential alterations and increased caspase-3 activity in human acute leukaemia HL-60 cells which prece-ded apoptosis. Similarly, the tricyclic analog, desipramine, has also been shown to induce mitochondrially-mediated apoptosis in C6 glioma cells via increased caspase-3 gene expression and intracellular calcium homeostasis changes.29 In addition, while we and others have shown that further an-tidepressants, including those of the selecti-ve serotonin reuptake inhibitor (SSRI) group, also mediate cancer cell apoptosis in both glioma and lymphoma, clomipramine appears to be most effective in this con-text.15 Very recently, Levkovitz et al.30 inde-pendently reported that clomipramine, in a comparative study between SSRIs and clo-mipramine in C6 rat glioma and human ne-uroblastoma cells, caused apoptosis prece-ded by a rapid increase in p-c-Jun levels, cytochrome c release from mitochondria and caspase-3-like activity. Significantly lo-wer sensitivity to the drug’s pro-apoptotic activity was demonstrated in primary mou-se brain and neuronal cultures. The authors therefore concluded – as we had previously - that the high sensitivity of cancer cells to the drug suggested that clomipramine may have potential in the treatment of brain tu-mours. We have also demonstrated the role
Pilkington GJ et al. / Tricyclic drugs in treatment of malignant glioma
81
of cathepsin L in interfering with activity of pro-apoptotic agents such as clomipramine by use of cathepsin inhibitors and anti-sense technology.31
Amitriptyline has previously been repor-ted to reduce proliferation in cancer cell li-nes21 and to decrease glioma cell viability.32 In our preliminary experiments we have fo-und that amitriptyline induces a dose-de-pendent reduction in oxygen utilisation in human glioma cells as well as apoptosis as seen in Annexin V/PI flow cytometry. Mo-reover, when early passage cultures of human glioma were treated sequentially with clomipramine & amitriptyline apoptosis was initiated. Only a small proportion of cells recovered from this treatment. In addition, reversal of multi drug resistance in a number of solid cancers following trea-tment with both clomipramine19,33 and amitriptyline20 has been reported. This ad-ditional role for tricyclics may, albeit at dif-fering concentrations, be of some signifi-cance in the treatment of primary and se-condary brain tumours.
Cancer stem cells
CD133 is a 120kDa five-transmembrane cell surface protein, originally described as a haematopoietic stem cell marker.34,35 More recently, however, it was shown to mark normal human neural stem cells.36 Subse-quently, Singh et al.37 demonstrated CD133 positivity, by both immunohistochemistry and flow cytometry, on two common forms of paediatric brain tumour; the high grade malignancy medulloblastoma and the low grade pilocytic astrocytoma. Moreover, bra-in tumour stem cells can be magnetic im-muno-bead and fluorescence activated cell sorted by use of dissociated cell suspensi-ons using CD133 antibodies. The subsequ-ent CD133 positive selected sub-populati-on of tumour cells also express nestin but fail to express markers associated with dif-
ferentiated cells of neural lineage.38 Altho-ugh these CD133/nestin positive stem cells represent a minority fraction of the overall tumour cell complement, they are able to generate clonal tumour neurospheres in suspension culture. They also show increa-sed self-renewal capacity and can be indu-ced to differentiate into cells phenotypi-cally similar to those seen in the original patient histology. The same group then de-veloped an in vivo, serially-transplantable, xenograft model in NOD-SCID (non-obese diabetic, severe combined immunodefici-ent) mouse brains by injecting as few as 100 CD133-positive brain tumour stem cells. The histological appearance of the re-sulting tumours resembled that of the original resected tumour. Conversely, injection of as many as 105 CD133-negative cells fai-led to produce tumours.39 We have noted that while human glioma biopsies normally grow well in standard DMEM growth con-ditions, cells from four clomipramine trea-ted patients cells taken at second biopsy grow poorly in DMEM culture media. We hypothesise that cancer stem cells, as deno-ted by CD133 (plus CD44/CD24/ne-stin/Musashi-1) expression may increase in number & are resistant to clomipramine.40 We, therefore, propose to culture these se-cond (recurrent case) biopsies, as well as new cases of glioma in stem cell defined medium to see if yield of CD133 +ve cells has increased. Primary/ex-vivo cultures will be prepared from human glioma obtained from King’s College Hospital (KCH) London (LREC 00-173) and Hurstwood Park Neurological Centre, Haywards Heath, Sussex (LREC applied for). Epilepsy surgi-cal brain resection tissue from KCH will be used to provide additional non-neoplastc astrocyte cultures (LREC 02-056). Biopsied glioblastoma primary cultures taken from both newly diagnosed patients and those previously treated with clomipramine and isolated in vitro41 in stem cell defined fee-Radiol Oncol 2006; 40(2): 73-85.
82
Pilkington GJ et al. / Tricyclic drugs in treatment of malignant glioma
der cell conditions.42 The monoclonal AC133 antibody (Miltenyi Biotech) will be used to identify CD133 positive stem cells and early progenitor cells. Immunocytoche-mistry, using fluorescence/TIRF micro-scopy, and flow cytometry will be used to identify and quantify CD133 antigen expression. CD133 positive cells will then be separated either by MACS/CD133 immu-nobeads (Miltenyi Biotech) or by sterile FACS and grown in bulk culture for subse-quent testing with various drug combinati-ons. Although we expect a low yield of CD133-positive cells we feel this would be a timely study.
Valproic acid
Many brain tumour patients also suffer from seizures and are, consequently, pre-scribed anti-convulsants. One particular an-ti-convulsant, the histone deacetylase inhibitor, sodium valproate (valproic acid), has recently attracted attention for its potential anti-cancer properties. Histone deacetylati-on is critical for regulation of gene expression which may affect chromatin structure and chromatin interaction with regulatory factors. In this context valproic acid has be-en shown to rapidly hyperacetylate histo-nes H3 and H4 in breast cancer cells and depleted the structural maintenance of chromatin proteins, DNA methyltransfera-se and heterochromatin proteins with a consequent enhanced sensitivity of DNA to DNA-damaging agents, both in vitro and in xenograft models.43 In addition, valproic acid has been reported to enhance radiosen-sitivity of human brain tumour cell lines and xenografts.44 Combination therapy of histone deacetylase inhibitors and radiothe-rapy has also resulted in increased neuro-blastoma cell necrosis and apoptosis com-pared with either single modality trea-tment. Interestingly, Beecken et al.45 have shown that it also positively modulates ne-Radiol Oncol 2006; 40(2): 73-85.
ural cell adhesion molecule (NCAM) polysi-alylation, thereby blocking adhesion of se-veral neuroectodermal tumour-derived cell lines to HUVEC (human umbilical vein en-dothelial cells) while downregulation of CD44 expression has been reported on human and rat glioma cells in vitro.46 These findings may be suggestive of reduced invasion but increased tumour cell differentiation and apoptosis have also been reported in human brain tumour xenograft models.46 Indeed, enhanced differentiated gene expression, growth inhibition, cell cycle arrest, induction of apoptosis and down-regu-lation of the pro-survival genes bcl-2 and bcl-xl has also been reported in thyroid can-cer cells.47,48 We are, therefore, eager to ex-plore the potential of valproic acid in com-bination with tricyclics.
Current literature available on dexame-thasone and its actions on glioma cells is conflicting. It has been reported that gluco-corticoids have a functional role at the level of the mitochondria.49 It has also been shown that glucocorticoids are neurotoxic and appear to play a role in neuronal cell loss following neuropathological insults.50 Dexamethasone has been shown to enhan-ce necrotic cell death of glioma cells indu-ced by serum deprivation.50 The steroid also reversibly and significantly inhibits growth of C6 glioma cells both at early and late passage.51 Despite evidence suggesting dexamethasone exerts a necrotic type of cell death, some studies have indicated that its mechanism of action is via apoptosis and interference with apoptotic pathways. In leukaemia cells dexamethasone-induced apoptosis has been demonstrated through the mitochondria-dependent pathway.52 Glucocorticoids are known to influence the ability of cells to undergo apoptosis, di-rectly inducing apoptosis in thymocytes while inhibiting it in hepatoma and carci-noma cells.23 It has been suggested that de-xamethasone inhibits  the induction of
Pilkington GJ et al. / Tricyclic drugs in treatment of malignant glioma
83
apoptosis in astrocytoma cells, probably via up-regulation of Bcl-xL, which could pre-vent cytochrome c release from mitochon-dria and subsequent caspase activation.23 Dexamethasone was also shown to confer protection against the induction of apopto-sis by anti-cancer agents.23 This indicates that dexamethasone could potentially in-terfere with the efficacy of chemotherapeu-tic agents. The laboratory and clinical stu-dies described are aimed at identifying a possible role for tricyclics in combination with standard therapies for glioma pati-ents. It is hoped that such a combinatorial, and possibly customised, approach may enhance both quality of life and survival time for patients suffering from malignant brain tumour.
Acknowledgements
Research on clomipramine in the author’s laboratories was generously supported by grants from the Samantha Dickson Research Trust.
References
1.   Linstead D, Wilkie D. A comparative study of in vivo inhibition of mitochondrial function in Sac-charomyces cerevisiae by tricyclic and other cen-trally-acting drugs. Biochem Pharmacol 1971; 20: 839-46.
2.   Wilkie D. Antimitochondrial drugs in cancer che-motherapy: preliminary communication. J Roy Soc Med 1979; 72: 599-601.
3.   Rooprai HK, Christidou M, Pilkington GJ. The potential for strategies using micronutrients and he-terocyclic drugs to treat invasive gliomas. Acta Ne-urochir (Wien) 2003; 145: 683-90.
4.   Eto K, Fukuda T, Araki Y, Inoue B, Ogata M. Effect of tricyclic drugs on mitochondrial membrane. Ac-ta Med Okayama 1985; 39: 289-95.
5.   Orrenius S. Mitochondrial regulation of apoptotic cell death. Toxicol Lett 2004; 149: 19-23.
6.   Warburg O. On the origin of cancer cells. Science 1956; 123(3191): 309-14.
7.   Green DR, Reed JC. Mitochondria and apoptosis. Science 1998; 281(5381): 1309-12.
8.   Vaux DL. Apoptosis and toxicology—what relevan-ce? Toxicology 2002; 181-182: 3-7.
9.   Amirlak B, Couldwell WT. Apoptosis in glioma cells: review and analysis of techniques used for study with focus on the laser scanning cytometer. J Neurooncol 2003; 63: 129-45.
10. Julien T, Frankel B, Longo S, Kyle M, Gibson S, Shillitoe E, et al. Antisense-mediated inhibition of the bcl-2 gene induces apoptosis in human mali-gnant glioma. Surg Neurol 2000; 53: 360-9.
11. Glaser T, Weller M. Caspase-dependent chemo-therapy-induced death of glioma cells requires mi-tochondrial cytochrome c release. Biochem Biophys Res Commun 2001; 281: 322-7.
12. Morisaki T, Katano M. Mitochondria-targeting the-rapeutic strategies for overcoming chemoresistan-ce and progression of cancer. Curr Med Chem 2003; 10: 2517-21.
13. Xia Z, Bergstrand A, DePierre JW, Nassberger L. The antidepressants imipramine, clomipramine, and citalopram induce apoptosis in human acute myeloid leukemia HL-60 cells via caspase-3 activa-tion. J Biochem Mol Toxicol 1999; 13: 338-47.
14. Xia Z, Lundgren B, Bergstrand A, DePierre JW, Nassberger L. Changes in the generation of reac-tive oxygen species and in mitochondrial membrane potential during apoptosis induced by the antidepressants imipramine, clomipramine, and citalopram and the effects on these changes by Bcl-2 and Bcl-X(L). Biochem Pharmacol 1999; 57: 1199-208.
15. Meredith EJ, Holder MJ, Chamba A, Challa A, Dra-ke-Lee A, Bunce CM, et al. The serotonin transporter (SLC6A4) is present in B-cell clones of diverse malignant origin: probing a potential anti-tumor target for psychotropics. Faseb J 2005; 19: 1187-9.
16. Daley E, Wilkie D, Loesch A, Hargreaves IP, Ken-dall DA, Pilkington GJ, et al. Chlorimipramine: a novel anticancer agent with a mitochondrial target. Biochem Biophys Res Commun 2005; 328: 623-32.
17. Spanova A, Kovaru H, Lisa V, Lukasova E, Rittich B. Estimation of apoptosis in C6 glioma cells treated with antidepressants. Physiol Res 1997; 46: 161-4.
18. Merry S, Hamilton TG, Flanigan P, Freshney RI, Ka-ye SB. Circumvention of pleiotropic drug resistance in subcutaneous tumours in vivo with verapamil and clomipramine. Eur J Cancer 1991; 65: 31-4.
Radiol Oncol 2006; 40(2): 73-85.
84
Pilkington GJ et al. / Tricyclic drugs in treatment of malignant glioma
19. Pommerenke EW, Volm M. Reversal of doxorubi-cin-resistance in solid tumors by clomipramine. In Vivo 1995; 9: 99-101.
20. Varga A, Nugel H, Baehr R, Marx U, Hever A, Nac-sa J, et al. Reversal of multidrug resistance by ami-triptyline in vitro. Anticancer Res 1996; 16: 209-11.
21. Volpe DA, Ellison CD, Parchment RE, Grieshaber CK, Faustino PJ. Effects of amitriptyline and fluo-xetine upon the in vitro proliferation of tumor cell lines. J Exp Ther Oncol 2003; 3: 169-84.
22. Swaroops GR, Kelly PA, Holmes MC, Shinoda J, Whittle IR. The effects of dexamethasone therapy on permeability, blood flow and iNOS expression in experimental glioma. J Clin Neurosci 2001; 8: 35-9.
23. Gorman AM, Hirt UA, Orrenius S, Ceccatelli S. Dexamethasone pre-treatment interferes with apoptotic death in glioma cells. Neuroscience 2000; 96: 417-25.
24. Das A, Banik NL, Patel SJ, Ray SK. Dexamethaso-ne protected human glioblastoma U87MG cells from temozolomide induced apoptosis by mainta-ining Bax:Bcl-2 ratio and preventing proteolytic ac-tivities. Molecular Cancer 2004; 3: 36.
25. Sur P, Sribnick EA, Patel SJ, Ray SK, Banik NL. Dexamethasone decreases temozolomide-induced apoptosis in human gliobastoma T98G cells. Glia 2005; 50: 160-7.
26. Amar S PK, R Lisle, G J Pilkington. Effect ofdexa-methasone on the cytotoxic effect of clomiprami-ne in human astrocytic cells in vitro. Journal of Ne-uro-oncology 2005; 7: 3.
27. Allabi AC, Gala JL, Desager JP, Heusterspreute M, Horsmans Y. Genetic polymorphisms of CYP2C9 and CYP2C19 in the Beninese and Belgian popula-tions. Br J Clin Pharmacol 2003; 56: 653-7.
28. Yokono A, Morita S, Someya T, Hirokane G, Oka-wa M, Shimoda K. The effect of CYP2C19 and CYP2D6 genotypes on the metabolism of clomi-pramine in Japanese psychiatric patients. J Clin Psychopharmacol 2001; 21: 549-55.
29. Qi H, Chen HZ, Jin ZJ. Caspase 3 gene expression and [Ca2+]i homeostasis underlying desipramine-induced C6 glioma cell apoptosis. Acta Pharmacol Sin 2002; 23: 803-7.
30. Levkovitz Y, Gil-Ad I, Zeldich E, Dayag M, Wei-zman A. Differential induction of apoptosis by anti-depressants in glioma and neuroblastoma cell lines: evidence for p-c-Jun, cytochrome c, and caspase-3 involvement. J Mol Neurosci 2005; 27: 29-42.
31. Levicar N, Dewey RA, Daley E, Bates TE, Davies D, Kos J, Pilkington GJ, Lah TT. Selective suppres-sion of cathepsin L by antisense cDNA impairs human brain tumor cell invasion in vitro and pro-motes apoptosis. Cancer Gene Therapy 2002; 10: 141-51.
32. Pilkington G, Amar S, Parker K. Induction of apoptosis in glioma by tricyclics and selective se-rotonin reuptake inhibitors: laboratory and clini-cal studies. In: AACR. Proceedings, 96th Annual Meeting; Anaheim, USA, April 2005.
33. Merry S, Hamilton TG, Flanigan P, Freshney RI, Ka-ye SB. Circumvention of pleiotropic drug resistance in subcutaneous tumours in vivo with verapamil and clomipramine. Eur J Cancer 1991; 27: 31-4.
34. Yin AH, Miraglia S, Zanjani ED, Almeida-Porada G, Ogawa M, Leary AG, et al. AC133, a novel marker for human hematopoietic stem and progenitor cells. Blood 1997; 90: 5002-12.
35. Miraglia S, Godfrey W, Yin AH, Atkins K, Warnke R, Holden JT, et al. A novel five-transmembrane hematopoietic stem cell antigen: isolation, charac-terization, and molecular cloning. Blood 1997; 90: 5013-21.
36. Uchida K, Mukai M, Okano H, Kawase T. Possible oncogenicity of subventricular zone neural stem cells: case report. Neurosurgery 2004; 55: 977-8.
37. Singh SK, Clarke ID, Terasaki M, Bonn VE, Haw-kins C, Squire J, et al. Identification of a cancer stem cell in human brain tumors. Cancer Res 2003; 63: 5821-8.
38. Singh SK, Clarke ID, Hide T, Dirks PB. Cancer stem cells in nervous system tumors. Oncogene 2004; 23: 7267-73.
39. Singh SK, Hawkins C, Clarke ID, Squire JA, Bayani J, Hide T, et al. Identification of human brain tumo-ur initiating cells. Nature 2004; 432(7015): 396-401.
40. Pilkington GJ. Cancer stem cells in the mammali-an central nervous system. Cell Proliferation 2005; 38: 423-33.
41. Lee A, Kessler JD, Read TA, Kaiser C, Corbeil D, Huttner WB, et al. Isolation of neural stem cells from the postnatal cerebellum. Nat Neurosci 2005; 8: 723-9.
42. Stojkovic P, Lako M, Stewart R, Przyborski S, Armstrong L, Evans J, et al. An autogeneic feeder cell system that efficiently supports growth of undiffe-rentiated human embryonic stem cells. Stem Cells 2005; 23: 306-14.
Radiol Oncol 2006; 40(2): 73-85.
Pilkington GJ et al. / Tricyclic drugs in treatment of malignant glioma
85
43. Marchion DC, Bicaku E, Daud AI, Sullivan DM, Munster PN. Valproic acid alters chromatin struc-ture by regulation of chromatin modulation proteins. Cancer Res 2005; 65: 3815-22.
44. Dote H, Cerna D, Burgan WE, Carter DJ, Cerra MA, Hollingshead MG, et al. Enhancement of in vitro and in vivo tumor cell radiosensitivity by the DNA methylation inhibitor zebularine. Clin Cancer Res 2005; 11: 4571-9.
45. Beecken WD, Engl T, Ogbomo H, Relja B, Cinatl J, Bereiter-Hahn J, et al. Valproic acid modulates NCAM polysialylation and polysialyltransferase mRNA expression in human tumor cells. Int Immu-nopharmacol 2005; 5: 757-69.
46. Cinatl J, Jr., Cinatl J, Driever PH, Kotchetkov R, Pouckova P, Kornhuber B, et al. Sodium valproate inhibits in vivo growth of human neuroblastoma cells. Anticancer Drugs 1997; 8: 958-63.
47. Catalano MG, Fortunati N, Pugliese M, Costantino L, Poli R, Bosco O, et al. Valproic acid induces apoptosis and cell cycle arrest in poorly differenti-ated thyroid cancer cells. J Clin Endocrinol Metab 2005; 90: 1383-9.
48. Shen WT, Wong TS, Chung WY, Wong MG, Kebe-bew E, Duh QY, et al. Valproic acid inhibits growth, induces apoptosis, and modulates apopto-sis-regulatory and differentiation gene expression in human thyroid cancer cells. Surgery 2005; 138: 979-85.
49. Koufali MM, Moutsatsou P, Sekeris CE, Breen KC. The dynamic localization of the glucocorticoid re-ceptor in rat C6 glioma cell mitochondria. Mol Cell Endocrinol 2003; 209: 51-60.
50. Morita K, Ishimura K, Tsuruo Y, Wong DL. Dexa-methasone enhances serum deprivation-induced necrotic death of rat C6 glioma cells through acti-vation of glucocorticoid receptors. Brain Res 1999; 816: 309-16.
51. Goya L, Feng PT, Aliabadi S, Timiras PS. Effect of growth factors on the in vitro growth and differentiation of early and late passage C6 glioma cells. Int J Dev Neurosci 1996; 14: 409-17.
52. Renner K, Amberger A, Konwalinka G, Kofler R, Gnaiger E. Changes of mitochondrial respiration, mitochondrial content and cell size after induction of apoptosis in leukemia cells. Biochim Biophys Ac-ta 2003; 1642: 115-23.
Radiol Oncol 2006; 40(2): 73-85.
134
Slovenian abstracts
Radiol Oncol 2006; 40(2): 73-85.
Pomen tricikličnih zdravil pri selektivnem proženju
mitohondrijske apoptoze pri neoplastični gliji.
Nova možnost zdravljenja malignih gliomov?
Pilkington G J, Akinwunmi J, Amar S
Izhodišča. Naše raziskave so že pokazale, da ima triciklični antidepresiv klomipramin in vitro specifičen pro-apoptotičen učinek na humanih malignih gliomskih celicah. Učinek je odvisen od koncentracije klomipramina in ga ni opaziti na normalnih humanih astrocitih. Zdravilo sproža apoptozo tako, da deluje na mitohondrije, kjer učinkuje na dihalno verigo. Čeprav so ob naših tudi druge raziskave pokazale, da imajo različni antidepresivi (vključno s selektivnimi zaviralci ponovnega prevzema serotonina – SSRI) vpliv na apoptozo pri lim-fomih in gliomih, je klomipramin najbolj učinkovit. Ugotavljali smo tudi pro-apoptotično aktivnost drugih tricikličnih zdravil in odkrili, da imata le dve takšni zdravili (amitriptilin in doksepin) enako ali boljšo učinkovitost kot klomipramin. Per os zaužiti zdravili klomipramin in amitriptilin se metabolizirata v desmetil klomipramin (norklomipramin) in nortriptilin. Menimo, da bi bilo potrebno testirati tumorske celice na omenjeni zdravili in na njuna metabolita. Učinkovitost obeh zdravil je namreč lahko močno zmanjšana zaradi odpornosti na zdravila (multidrug resistance), ki pa je pri obeh zdravilih različna. Ugotovili smo tudi, da ima metabolit klomipramina norklomipramin slabši pro-apoptotični učinek, medtem ko ima metabolit amitriptina nortriptilin enak učinek kot amitriptin. Zaključki. Menimo, da bo potrebno v kliničnih raziskavah ugotoviti učinkovitost tricik-ličnih antidepresivov pri malignem gliomu, najprej kot dopolnilno zdravljenje.
Radiol Oncol 2006; 40(2): 133-8.