Radiol Oncol 2006; 40(2): 87-93. Apoptosis of human malignant glioma-derived cell cultures treated with clomipramine hydrochloride, as detected by Annexin-V assay Katharine A. Parker and Geoffrey J. Pilkington Cellular and Molecular Neuro-oncology Group, Institute of Biomedical and Biomolecular Sciences, School of Pharmacy and Biomedical Sciences, University of Portsmouth, UK Background. Previous research in our laboratories has shown that Clomipramine Hydrochloride (CLOM), a tricyclic antidepressant in use for over thirty years, selectively kills neoplastic glial cells in vitro whilst leaving normal brain cells unaffected. The purpose of this study was to evaluate whether a range of early passage cell cultures and established cell lines, derived from a number of patients with malignant glioma, would display different sensitivities when exposed to CLOM. The particular assay of interest, following our discovery that CLOM targets the mitochondria of tumour cells and triggers Caspase 3 mitochondrially-medi-ated apoptosis, was Annexin-V flow cytometry. This assay was used to determine the mechanism of cell de-ath, either necrosis or apoptosis, according to drug concentration and period of incubation. Method. Cells grown to 90% confluence in 25cm3 flasks were incubated with concentrations of CLOM from 20µM – 100µM , for up to 6 hours. Cells were harvested and resuspended in calcium binding buffer, which triggers translocation of calcium-regulated phosphatidylserine residues to the nuclear envelo-pe, before removing 500µl of the single cell suspension to a Facs tube. Controls used in the analysis we-re performed by omission of the drug incubation in one flask, and addition of 1µM staurosporine to one flask. These served as negative and positive controls respectively. Annexin-V FITC and propidium iodi-de were added to all tubes and incubated for 15 minutes at room temperature, in the dark. Subsequent to this, binding buffer was added to each tube and analysed using a BD FACScalibur. Results. Results show that, of the five malignant gliomas tested, the two established cell lines had the lower apoptotic threshold, with a significant percentage of apoptotic cells present at 60µM and above when compared to the control sample. The three early passage cultures, developed ‘in house’ from bi-opsy, had higher apoptotic thresholds, withstanding up to 100µM CLOM incubation for six hours. Normal human astrocytes were assayed in parallel, and show that CLOM does not cause cell death at the concentrations tested. Conclusions. It may be possible, in a larger study, to predict individual patient response to CLOM us-ing the Annexin-V assay, alongside Bcl-2 analysis and CYP gene testing, on the individual patient’s tu-mour cells. The difference in sensitivities between glioma, in this small study, indicates the importance of analysing early passage cultures, which retain original morphology and characteristics to a greater ex-tent, alongside established cell lines. Key words: annexin V; apoptosis; brain neoplasms - drug therapy; glioma; clomipramine; tumour cells, cultured 88 Parker K and Pilkington GJ / Clomipramine hydrochloride cytotoxicity Received 20 March 2006 Accepted 12 June 2006 Correspondence to: Professor G.J. Pilkington, Cellular and Molecular Neuro-oncology Group, Institute of Bi-omedical and Biomolecular Sciences, University of Portsmouth, White Swan Road, Portsmouth, Hants, PO1 2DT UK Tel: +44 (0) 23 9284 2116, Fax: +44 (0) 23 9284 2116, E-mail: Geoff.Pilkington@port.ac.uk, Website: www.port.ac.uk/brainlab Introduction Previous research in our laboratories has shown that Clomipramine Hydrochloride (CLOM), a tricyclic antidepressant in use for over thirty years, has the ability to induce apoptosis in malignant glioma cells in vitro.1 Thus, following the initiation of a clinical tri-al based at King’s College Hospital, London (LREC 01-235), it is important to be able to predict the outcome of a given drug treat-ment, especially in the case of glioma which are often characterised by a poor prognosis. Cell death is often defined as occurring by either apoptosis or necrosis. Whilst ne-crosis is a relatively passive process, invol-ving loss of membrane integrity and cell membrane rupture, which leads to the release of intracellular debris and eventual in-flammatory response, apoptosis is an active process resulting in cell shrinkage, plasma membrane blebbing and chromatin con-densation which produces apoptotic bodi-es, rapidly recognised and phagocytosed by macrophages. This ‘clean’ mechanism of cell death, avoiding an inflammatory res-ponse, ensures minimal tissue damage to the surrounding brain parenchyma. Apoptosis after CLOM treatment is associated with the intrinsic pathway of mito-chondrial cytochrome C release. 1,2 follo-wing a rapid increase in p-c-Jun2, and acti-vation of caspase-3.3 The two main path-ways of apoptosis can be identified in mammalian cells, both are controlled by caspases and eventually converge on ‘exe-cutioner’ caspase-34, which is responsible for the cleavage of structural cytoplasmic Radiol Oncol 2006; 40(2): 87-93. and nuclear proteins, with consequent cell death and collapse.5 Phosphatidylserine (PS), made by the two PS synthases PSS1 and PSS2, is nor-mally located on the inner leaflet of the plasma membrane, but undergoes transbila-yer movement during apoptosis and beco-mes exposed on the cell surface.6 Annexin-V FITC (BD Biosciences) binds to the exposed PS residues in a calcium dependent manner, after a rise in nuclear calcium con-centration that causes the translocation of the calcium regulated proteins to the nucle-ar envelope.7 This mechanism, usually a pi-votal step in the recognition and removal of apoptotic cells by phagocytes8, allows the attachment of the Annexin V-FITC anti-body and allows us to visualise apoptotic cells via flow cytometry. The purpose of this study was to evaluate whether a range of both early passage cultu-res and established cell lines, derived from patients with malignant glioma, would display different sensitivities, with regard to apoptotic cell death, when exposed to CLOM. This, combined with other assays previously carried out by the research group, could go some way towards defining in vitro markers for patient response to CLOM. Materials and methods Cells; the following cell cultures were used: SNB-19 – an established glioblastoma mu-etiforme GBM (grade IV) cell line p20-24, derived from a 47-year old male (DSMZ Cell Bank) DK-MG – an established GBM (grade IV) cell line p23-27, derived from 67-year old female (DSMZ Cell Bank) UPAB – a primary GBM (grade IV) cell culture set up ‘in house’, p11-14, derived from a 73 year-old male Parker K and Pilkington GJ / Clomipramine hydrochloride cytotoxicity 89 UPMC – a primary GBM (grade IV) cell culture set up ‘in house’, p9-12, derived from a 69 year-old female UPJM – a primary astrocytoma (grade II) cell culture set up ‘in-house’, p7-10, derived from 42 year-old male CC-2565 – a normal human astrocyte cell line, p4-6, derived from an 18 year-old male (Cambrex Biosci-ences). Annexin V analysis The apoptosis assay was used to determine the mechanism of cell death according to drug concentration and period of incubati-on. Cells grown to 90% confluence in 25cm3 flasks were incubated with 10X concentrati-ons of CLOM (20, 40, 60, 80 & 100µM), ad-ded to flasks at 1:10, for up to 6 hours. Cells were harvested by firstly removing the com-plete media to centrifuge tubes (to ensure that all cells are subject to analysis), before adding 1.0ml of clear Tryplexpress (a non-enzymatic rapid dissociation solution; Gib-co). During the two-minute dissociation period, flasks were placed in the incubator (37°C, 5% CO2) to maintain the optimum temperature. The Tryplexpress was remo-ved by centrifugation at 200gav after neutralisation with 10% complete media. Following staining the cell pellet was re-suspended in 1ml of 1X calcium binding buffer, which triggers translocation of calci-um-regulated phosphatidylserine residues to the nuclear envelope, before removing 500µl of the single cell suspension to a FACS tube. Controls used in the analysis were performed by omission of the drug in-cubation in one flask, and addition of 1µM staurosporine to one flask. These served as negative and positive controls respectively. Five microlitres of annexin V FITC and 5µl of propidium iodide were added to all tu- bes, by placing a drop of the fluorochrome on the side of the tube and inverting it. The tubes were incubated for 15 minutes at ro-om temperature, in the dark. Subsequent to this, 400µl of binding buffer was added to each tube and analysed by the BD FACSca-libur within 1 hour. Results After a six-hour incubation with CLOM the cell lines/cultures undergoing a marked de-gree of apoptosis, when compared against negative controls were DK-MG and SNB-19. Because of the slow-growing nature of the lower grade astrocytoma UPJM, some sam-ples were not achieved due to lack of cells (20,000 minimum required for analysis). Al-though the percentage of apoptosis in UPMC appears high, when compared to the control values it demonstrates that CLOM does not exert any effect at the concentrati-ons tested (see Table 1 and Figure 1). From the five malignant gliomas tested, the two established cell lines had the lower apoptotic threshold, with a significantly higher percentage of apoptotic cells pre-sent at 60µM CLOM and above. The three early passage cultures, developed ‘in-hou-se’ from biopsy, had higher apoptotic thre-sholds, withstanding up to 100µM CLOM incubation for six hours. The normal human astrocytes, tested in parallel, demonstrated that CLOM did not cause cell death at the concentrations te-sted. The cells at the highest passage number (DK-MG) are the most responsive to Clomi-pramine in this assay; this could be due, in part, to the homogeneity of the sample population. Also, it is of interest to note that the CLOM was less effective at causing apoptosis in the CC-2565 cell line than the staurosporine in the positive control sample (Table 1). Radiol Oncol 2006; 40(2): 87-93. 90 Parker K and Pilkington GJ / Clomipramine hydrochloride cytotoxicity SNB-19 DK-MG UPAB UPMC UPJM CC-2565 Apoptosis (%) Control Sample 1.25 2.05 2.27 10.77 3.03 0.36 Staurosporine 21.47 71.56 5.04 47.43 38.94 1.94 (6hr control) 20µM 6h 1.25 1.87 2.35 9.96 1.01 0.29 5h 0.92 1.50 3.21 10.81 3.20 4.68 4h 0.95 1.38 2.11 12.74 0.50 3.21 3h 1.29 1.18 3.00 11.72 0.88 2.51 2h 0.71 1.40 1.87 10.99 0.24 4.12 1h 1.45 1.03 1.46 10.13 1.31 2.98 40µM 6h 3.67 1.95 2.84 11.67 - 3.62 5h 2.69 2.64 3.45 13.32 - 3.28 4h 2.13 1.30 4.19 14.11 - 2.92 3h 2.11 5.39 2.93 12.41 - 1.55 2h 2.64 4.26 3.00 12.31 - 3.74 1h 2.72 3.14 4.01 14.03 - 3.32 60µM 6h 4.84 23.27 3.86 14.26 - 2.86 5h 5.14 21.44 3.40 12.16 - 0.03 4h 11.50 51.25 2.99 12.80 - 5.65 3h 10.98 18.75 3.71 13.62 - 0.32 2h 5.9 26.53 3.66 12.33 - 4.49 1h 5.31 19.75 3.08 13.47 - 3.52 80µM 6h 5.89 27.13 3.00 14.22 12.32 0.23 5h 15.91 20.86 4.36 14.02 10.00 1.65 4h 10.55 38.92 3.18 15.88 10.47 0.15 3h 4.98 42.25 3.31 15.82 - 0.05 2h 9.18 23.76 2.51 13.19 - 0.02 1h 10.27 23.41 3.71 14.59 - 0.00 100µM 6h 11.03 49.16 3.91 10.97 10.73 0.36 5h 2.57 23.18 2.77 12.91 8.64 0.01 4h 10.91 22.81 2.97 11.64 5.99 3.21 3h 12.58 17.74 4.98 11.94 10.28 0.04 2h 17.13 20.66 4.68 10.76 6.50 0.05 1h 11.00 30.84 2.79 9.41 - 0.38 Table 1. A summary of the apoptosis data obtained by Annexin V flow cytometry, highlighted are the samples at which apoptosis (defined as a sample with more apoptosis than the negative control) was achieved when compa-red to the negative and positive controls, which were cells with no drug and cells with staurosporine added res-pectively. This table illustrates the differences in apoptotic sensitivity of the cell lines; with DK-MG being the most responsive when compared to the control values. Radiol Oncol 2006; 40(2): 87-93. Parker K and Pilkington GJ / Clomipramine hydrochloride cytotoxicity 91 Figure 1. Annexin V Data for DK-MG (A), SNB-19 (B), UPAB (C), UPMC (D), UPJM (E) and CC-2565 (F). Plots sho-wing the side scatter and gating selected for this assay (left) demonstrate the difference in cell size and heteroge-neity of the samples. The plots of interest shown (right) are samples analysed after a 6-hour incubation with 100KM Clomipramine Hydrochloride. The FL-1 detector detected annexin-V binding; the FL-2 detector detected cells counterstained with isotonic propidium iodide. The percentage values (%) for apoptotic and dead cells, res-pectively, are as follows A = 49.16; 13.23, B= 11.03; 5.96, C=3.91; 4.76, D=10.97; 2.24, E= 10.73; 4.85, F=0.36; 11.28. Discussion CLOM has previously been reported to ex-ert an apoptotic effect by Xia et al.9, on human myeloid leukaemia HL-60 cells (50µM), and Levkovitz et al.2 on C6 glioma cells (25µM) and human neuroblastoma SH-SY5Y cells (20µM). Significant mor-phological changes following incubation with 12µM CLOM, represented by red (propidium iodide) fluorescence of fra-gmented apoptotic nuclei, were observed by Levkovitz et al.2 when compared to blue (hoechst) fluorescence of the intact nuclei treated with vehicle (saline). Similar mor-phological findings were presented by Da-ley, E10, when human malignant glioma cells were incubated with CLOM (maxi- mum incubation period of 4 hours) and subsequently stained with ethidium bromide and acridine orange. Internucleoso-mal DNA fragmentation measured by elec-trophoresis in glioma cell lines was also demonstrated by Daley, E10, confirming DNA laddering and hence condensation of chromatin, the ‘classic’ hallmark of apop-tosis. These findings, from the literature on CLOM, confirm the potent apoptotic ef-fect that CLOM has on tumour cells. They also observe the higher resistance of pri-mary cell cultures11 which can be accoun-ted for by the high proportion of non-neo-plastic cells maintained in these short-term, low passage, cultures. This was reflected in the results of this study, whereby the control normal human Radiol Oncol 2006; 40(2): 87-93. 92 Parker K and Pilkington GJ / Clomipramine hydrochloride cytotoxicity brain astrocytes (CC-2565) were unaffected by CLOM. It is tempting to postulate that a population of normal astrocytes remains in the primary cultures developed ‘in-house’; further subculturing (leading to increased homogeneity of cell populations) and analysis may reveal passage-dependent apoptotic sensitivity to CLOM. It may be possible, in a larger study, to predict individual patient response to CLOM using the Annexin-V assay, alongsi-de Bcl-2 analysis and CYP gene testing, on the patients own tumour cells. Bcl-2 analy-sis, performed previously in our laboratories by Daley, E10 demonstrated that Bcl-2 expression correlated with apoptotic rate. Bcl-2 prevents cytochrome C release, and hen-ce glioma cell lines expressing a high per-centage of endogenous Bcl-2 had the lowest apoptotic rate. A common clinical observation within a patient cohort, diagnosed with the same tu-mour type, is the appearance of a few ‘res-ponders’ who respond very well to the test agent, and a majority of non-responders who do not gain any advantage from the test agent.12 One explanation for this could be the multidrug-resistant phenotype of brain tumours; Andersson et al.13 found a large heterogeneity in the expression of dif-ferent resistance markers (P-glycoprotein, PgP; Multidrug resistance protein, MRP1; lung-resistance related protein, LRP and O(6)-methylguanine-DNA methyltransfera-se, MGMT). The next steps in this research are to combine CLOM with other potenti-ally synergistic agents to enhance to apop-totic effect. It may also be possible to isola-te cancer stem cells and/or other ‘clones’ from heterogeneous primary glioma to test the resistance of subpopulations. The signi-ficance of further studies on Bcl-2, which acts upstream of the mitochondria, is that the expression may enhance the survival of cancer cells.3 The difference in sensitivities between glioma, in this small study, indica-Radiol Oncol 2006; 40(2): 87-93. tes the importance of analysing early passage cultures, which retain original morpho-logy and characteristics to a greater extent, alongside established cell lines.14 Acknowledgement The authors wish to thank Dr Verena Amberger-Murphy, National Institute for Cellular Biotechnology, Dublin, for dona-ting the two established cell lines and the Samantha Dickson Research Trust (www.sdrt.co.uk) for their continuing financial support. References 1. Daley E, Wikie D, Loesch A, Hargreaves IP, Ken-dall DA, Pilkington GJ, Bates TE. Chlorimiprami-ne: a novel anticancer agent with a mitochondrial target. Biochem Biophys Res Commun 2005; 328(2): 623-32. 2. Levkovitz Y, GilAd I, Zeldich E, Dayag M, Wei-zman A. Differential induction of apoptosis by an-tidepressants in glioma and neuroblastoma cell lines: evidence for p-c-Jun, cytochrome c, and caspa-se-3 involvement. J Mol Neurosci 2005; 27(1): 29-42. 3. Ekert PG, Read SH, Silke J, Marsden VS, Kaufmann H, Hawkins CJ, Gerl R, Kumar S, Vaux DL. Apaf-1 and caspase-9 accelerate apoptosis, but do not determine whether factor-deprived or drug-treated cells die. J Cell Biol 2004; 165(6): 835-42. 4. Cohen Z, Wilson J, Ritter L, McDonagh P. Caspa-se inhibition decreases both platelet phosphatid-ylserine exposure and aggregation: caspase inhibition of platelets. Thromb Res 2004; 113(6): 387-93. 5. Ceruti SA, Mazzola A, Abbracchio MP. Resistance of human astrocytoma cells to apoptosis induced by mitochondria-damaging agents: possible impli-cations for anticancer therapy. J Pharmacol Exp Ther 2005; 314(2): 825-37. 6. Grandmaison PA, Nanowski TS, Vance JE. Exter-nalization of phosphatidylserine during apoptosis does not specifically require either isoform of phosphatidylserine synthase. Biochim Biophys Acta 2004; 1636(1): 1-11. 7. Raynal P, Kuipers G, Rojas E, Pollard HB. A rise in nuclear calcium translocates annexins IV and V to the nuclear envelope. FEBS Lett 1996; 392(3): 263-8. Parker K and Pilkington GJ / Clomipramine hydrochloride cytotoxicity 93 8. Zwaal RF, Comfurius P, Bevers EM. Surface exposure of phosphatidylserine in pathological cells. Cell Mol Life Sci 2005; 62(9): 971-88. 9. Xia Z, DePierre JW, Nassberger L. Modulation of apoptosis induced by tricyclic antidepressants in human peripheral lymphocytes. J Biochem Mol To-xicol 1998; 12(2): 115-23. 10. Daley E. The effect of mitochondrial membranoly-tic drugs on brain tumours in vitro. PhD Thesis; University of London, 2001. 1. 11. Gil-Ad I, Shtaif B, Levkovitz Y, Dayag M, Zeldich E, Weizmann A. Characterization of phenothiazi-ne-induced apoptosis in neuroblastoma and glio-ma cell lines: clinical relevance and possible application for brain-derived tumors. J Mol Neurosci 2004; 22(3): 189-98. 12. Behin A. Hoang_Xiuan K, Carpentier AF, Delattre JY. Primary brain tumours in adults. Lancet 2003; 361(9354): 323-31. 13. Andersson U. Malmer B, Bergenheim AT, Brannstrom T, Henrikkson R. Heterogeneity in the expression of markers for drug resistance in brain tumors. Clin Neuropathol 2004; 23(1): 21-7. 14. Baguley BC, Marshall ES. In vitro modelling of human tumour behaviour in drug discovery pro-grammes. Eur J Cancer 2004; 40(6): 794-801. Radiol Oncol 2006; 40(2): 87-93. Slovenian abstracts 135 Radiol Oncol 2005; 40(2): 87-93. Klomipramin hidroklorid in ugotavljanje apopotoze na celičnih kulturah humanih malignih gliomov s pomočjo pretočne citometrije ob uporabi Annexina-V Parker K, Pilkington GJ Izhodišča. Predhodne raziskave v našem laboratoriju so pokazale, da klomipramin hidroklorid (CLOM), triciklični antidepresiv, ki ga uporabljamo že 30 let, in vitro selektivno ubija neoplastične glialne celice in pri tem ne prizadene normalnih možganskih celic. Namen naše raziskave je bil oceniti celične kulture malignega glioma, ki smo jih odvzeli različnim bolnikom. Želeli smo ugotoviti, ali so različno občutljive na CLOM. Posebno nas je zanimala apoptoza, saj CLOM deluje na mitohondrije tumorskih celic in na ta način sproži apoptozo. Pri tem smo uporabljali pretočno citometrijo in Annexin-V. Glede na koncentracijo zdravila in čas inkubacije smo želeli ugotaviti mehanizem celične smrti, ali ta nastane predvsem zaradi nekroze ali zaradi apoptoze. Metode. Celice smo inkubirali do 6 ur z različno koncentracijo CLOM-a (20µM – 100µM). Sledila je priprava celic za pretočno citometrijo, kjer smo uporabili tudi Annexin-V FITC in propidium iodid. Rezultati. Preiskavo smo naredili s petimi malignimi gliomi. Pri dveh so imele celice manj apoptoze, koncentracija CLOM-a je bila 60µM ali več. Pri treh, kjer smo uporabili zgodnje celične linije, pa smo opazili zelo izrazito apoptozo, koncentacija CLOM-a je bila do 100µM, inkubacija pa 6 ur. Vzporedno smo preiskovali normalne humane astrocite in ugotovili, da CLOM v omenjenih koncentracijah ni povzročil njihove smrti. Zaključki. Preiskava z Annexinom-V bi lahko služila testiranju posamičnih bolnikov – ob analizi Bcl-2 in genskem CYP preiskovanju – ugotavljali bi lahko, ali so njihove tumorske celice občutljive na CLOM. Radiol Oncol 2006; 40(2): 133-8.