Chemotherapy Sensitizing Effects of Methadone in Glioblastoma Cells Are Drug- and Cell Line-dependent

D,L-methadone (MET), an analgesic drug used for pain treatment and opiate addiction has achieved attention from oncologist and social media as possible chemosensitizing agent in glioblastoma multiforme (GBM) treatment. MET has been reported to enhance doxorubicin-induced cytotoxicity in GBM cells via activation of the µ-opioid receptor (MOR) and subsequent apoptosis induction. Here, we further aimed at quantifying MET effects in comparison to other opioids alone and in combination with doxorubicin and clinically more relevant temozolomide (TMZ) in a set of GBM cell lines and primary GBM cells. and TMZ-induced cytotoxicity in established GBM cell lines and primary GBM cells. We recorded concentration-response curves in 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays of MET alone and doxorubicin or TMZ in combination with MET in a panel of cells to quantify effects and determine IC 50 values. Furthermore, we applied the MOR inhibitor naloxone and opioids used in oncology, morphine and oxycodone, to proof if opioid-induced cytotoxicity is mediated via MOR. Although MET was cytotoxic itself at high, clinically not relevant concentrations in all GBM cells, we only observed weak MET-induced sensitization of one cell line and primary cells to doxorubicin, while there was no effect of MET on TMZ-induced cell death.


Abstract
Background D,L-methadone (MET), an analgesic drug used for pain treatment and opiate addiction has achieved attention from oncologist and social media as possible chemosensitizing agent in glioblastoma multiforme (GBM) treatment. MET has been reported to enhance doxorubicin-induced cytotoxicity in GBM cells via activation of the µ-opioid receptor (MOR) and subsequent apoptosis induction. Here, we further aimed at quantifying MET effects in comparison to other opioids alone and in combination with doxorubicin and clinically more relevant temozolomide (TMZ) in a set of GBM cell lines and primary GBM cells.
Methods MOR expression in GBM cells was investigated by immuno uorescence and immunoblotting. Resistance to drugs alone or in combination was assessed by MTT assays. Concentration effect curves were tted to data points by nonlinear regression analysis and IC 50 values were calculated. Apoptotic rates were determined by Annexin V staining.

Results
We found that MET alone was cytotoxic to GBM cells at high micromolar concentrations in MTT assays by induction of apoptosis and necrosis while morphine and oxycodone were hardly cytotoxic. Naloxone was not able to block MET-induced cytotoxicity, indicating that cell death inducing effects of MET are not MOR dependent. We recorded doxorubicin and TMZ concentration response curves by MTT assays in combination with xed MET concentrations. MET only enhanced doxorubicin cytotoxicity in one cell line and in part in primary cells at certain MET concentrations. MET was not effective in sensitizing cells towards TMZ. Contrarily, in two cell lines MET even decreased sensitivity towards TMZ.

Conclusions
MET can be considered cytotoxic to GBM cells only at clinically not relevant concentrations by induction of apoptosis and necrosis. Sensitizing effects are only observed in combination with doxorubicin but not with TMZ and are highly dependent on cell line and applied drug concentrations.

Background
Grade IV glioblastoma multiforme (GBM) is the most aggressive form of glioma with highest incidence [1,2]. Median survival is 14.6 months under therapy, the 5-year survival rate only is at 6% [3], indicating that there is high medical need for new therapeutic options. Therapeutic standard of care in GBM treatment is radiotherapy with concomitant temozolomide (TMZ) treatment [4,5]. Recent studies indicate that D,Lmethadone (MET), an analgesic drug used for pain treatment and opiate addiction [6,7], increases apoptosis of leukemia cells and the cytotoxic effects of the topoisomerase II-inhibitor doxorubicin. In a follow up study, the same group investigated MET on GBM cell lines and showed that the opioid has the potential to enhance apoptosis induced by doxorubicin [8][9][10]. The proposed mechanism of action is activation of the µ-opioid receptor (MOR) and subsequent suppression of cAMP/protein kinase A (PKA) signaling via inhibitory G-proteins (Gi) which nally activates caspases and induces apoptosis. However, cAMP displays pro-and anti-apoptotic effects depending on cell type [11] raising the question if all GBM cells respond equally to MET treatment. In addition, it has been shown that MET increases intracellular doxorubicin levels probably by inhibiting P-glycoproteins (P-gps) in GBM cells. In a tumor model of nude mice implanted with U87 GBM cells, MET treatment reduced tumor growth and volume [9]. On the other hand, more recent data indicate no sensitizing effect of MET on various doxorubicin-treated canine tumor cells [12].
The rst-line therapy in GBM treatment is TMZ in combination with radiotherapy, while doxorubicin in rst place has no indication for GBM and second is rarely used off-label due to its poor blood-brain barrier penetration and neurologic side effects [13,14]. Despite lack of preclinical and clinical data on any bene cial effects of MET on TMZ therapy, it has been promoted as promising therapeutic option for GBM treatment. Current in vitro studies did not report a potentiation of TMZ effects in GBM cells [15,16], questioning the use of MET in GBM (reviewed in [17]). Therefore, we aimed at investigating the quantitative contribution of MET to doxorubicin-and TMZ- containing 10% fetal calf serum (FCS; Biochrom, Berlin, Germany), 100 IU/ml penicillin, and 100 µg/ml streptomycin (Biowest, Nuaillé, France). All other cells and cell lines were cultivated in Dulbecco's Modi ed Eagle Medium (DMEM) low glucose (Biowest, Nuaillé, France) containing 10% FCS (Biochrom, Berlin, Germany), 100 IU/ml penicillin, and 100 µg/ml streptomycin (Biowest, Nuaillé, France). Primary GBM cells derived from a primary GBM tumor biopsy were obtained from the University Hospital Cologne, genetically characterized and cultured as previously described [18].

Western Blot Analysis
In order to prepare whole cell lysates, cells were washed with ice cold phosphate-buffered saline (PBS; Biowest, Nuaillé, France) and lysed with ice cold Denaturing Cell Extraction Buffer (FNN00091, Thermo Fisher Scienti c, Waltham, MA, USA), incubated on ice for 30 min, and centrifuged for 15 min at 4 °C. The supernatant was used for protein content determination and subsequent immunoblotting.For immunoblotting standard procedures using the following antibodies were used as previously described [19]. Anti-MOR-1 (D-12; 1:1000, Santa Cruz Biotechnology, TX, USA) combined with goat anti-mouse IgG-HRP (1:5000, Santa Cruz Biotechnology, TX, USA) andβ-Actin-HRP (C-4; Santa Cruz Biotechnology, TX, USA). Immunoblots were developed with the enhanced chemoluminescence system (Amersham Biosciences, Little Chalfont, United Kingdom).

MTT Assay
For MTT assays we followed a published protocol [20]. Brie y, 5000 (U251, U87, U373) or 15,000 (primary) cells were plated on 96 wells and grown at 37 °C and 5% CO 2 overnight. Cell survival after exposure to either opioids alone or doxorubicin/TMZ in the presence of MET as indicated was determined by MTT assays after 72 h. In combination experiments MET or naloxone were added to culture medium 1 h prior to addition of compounds for 72 h. Final DMSO concentrations in media did not exceed 1%.

Annexin V Apoptosis Assay
For apoptosis measurements the BD Pharmingen™ FITC Annexin V Apoptosis Detection Kit (BD Biosciences, Franklin Lakes, NJ, USA) was used according to the manufacturer´s protocol. Brie y, 2.5 × 10 5 cells were seeded into 6-well plates and incubated at 37° C and 5% CO 2 overnight. After compound treatment for 72 h, cells were trypsinized and centrifuged for 4 min at 1500 × g. Supernatant was removed and cells were resuspended in 500 µL binding buffer. 5 µL propidium iodide and 5 µL Annexin V-FITC were mixed with 100 µL of cells in binding buffer. After 15 min of incubation on ice, samples were analyzed by ow cytometry (FACSCaliburTM, BD Bioscience, Franklin Lakes, NJ, USA). 75,000 cells/mL were seeded on cover-glasses and incubated at 37° C and 5% CO 2 for 48 h before staining. Thereafter, cells were washed 2 times for 1 min with PBS and xed with 4% paraformaldehyde/PBS for 10 min. Fixed cells were washed for 3 times with PBS and blocked in 2% bovine serum albumin (BSA; Sigma-Aldrich, St. Louis, MO, USA) including 0.2% Triton X-100 for 30 min. Thereafter cells were incubated with MOR-1-antibody (D12, Santa Cruz Biotechnology, TX, USA) in 2% BSA (1:50) at 4° C overnight. The next day cells were washed incubated with Anti-mouse IgG (H + L) DyLight™ 680 Conjugate (Cell Signalling Technology, Danvers, MA, USA) in 2% BSA (1:500) for 1 h. Finally, the stained cells were washed for 3 times with PBS and covered with a coverslip by using 1 drop of mounting medium. Only with secondary antibody stained cells, were used as negative control. MOR expression was visualized by using a Zeiss LSM 780 microscope (Carl Zeiss AG, Oberkochen, Germany) and 40 × magni cation.

Data Analysis and Statistical Methods
Concentration effect curves were tted to data points by nonlinear regression analysis using the fourparameter logistic equation (GraphPad™ Prism). Statistical differences between two groups were determined by unpaired 2-tailed Student's t-test. Comparisons among several groups were performed by ANOVA followed by Turkey post-hoc test. Data are presented as means ± SEM

Results
It has previously been shown that the MOR is expressed in GBM cell lines and primary cells [9,15]. Here, we demonstrate expression of MOR in commonly used GBM cell lines (U87, U251, U373) and primary GBM cells isolated from a patient biopsy by Western blotting (Fig. 1a) and immunocytochemistry (Fig. 1b) con rming that MOR is expressed in a wide range of GBM cells to a similar extend.
We next asked whether MET is cytotoxic in these cells in MTT experiments including the A172 cell line in which bene cial effects of MET on doxorubicin-induced cytotoxicity has been reported [9]. We obtained very similar IC 50 values in all GBM cells in the range from 62 to 130 µM after 72-hour treatment. MET displayed a steep concentration response in all cells (Hill slopes ranging from − 3.3 to -4.4). The lowest IC 50 of 62 µM was obtained for A172 cells (Fig. 2a). Annexin V/propidium iodide staining of cell lines treated with the previously determined IC 50 MET concentrations revealed that approximately 50% of cell deaths account for apoptosis and necrosis, respectively (Fig. 2b). In order to test if MET-induced cytotoxicity is mediated via MOR, we co-treated A172, U373 and primary cells with the MOR antagonist naloxone and MET. Naloxone was not capable of abolishing MET-induced cytotoxicity (Fig. 2c-e).
Contrarily, naloxone even signi cantly increased MET toxicity in A172 and primary cells (Fig. 2f), indicating that MET does not require the MOR for its cytotoxic action.
To further proof these ndings, we treated cells with other MOR agonists used in clinical practice for pain management such as morphine and oxycodone. Strikingly, both compounds showed very weak cytotoxicity in supra-therapeutic concentrations in a subset of GBM cell lines and primary cells (Fig. 3a-d). Determined IC 50 values were in the millimolar range or could not be established due to to low cytotoxicity (Fig. 3e).
Friesen et al have previously shown that various MET concentrations (e.g. 10, 3, 1 µg/mL corresponding to ~ 30, 10, 3 µM, respectively) were capable of sensitizing A172 cells towards a xed doxorubicin concentration (0.3 µg/mL = 0.5 µM) in apoptosis assays. In order to determine a concentration relationship, we treated cells with ascending doxorubicin concentrations combined with 10 µM MET, a concentration which itself is not cytotoxic but close to what might be reached in plasma of patients. MTT concentration-response curves overlapped in both tested GBM cell lines and in primary GBM cells ( Fig. 4a-c); respective IC 50 values showing no signi cant differences are displayed in Fig. 4d. However, it has to be noted that in A172 cells at 0.3 µM doxorubicin, combination treatment with 10 µM MET reduced cell viability as compared to doxorubicin treated cells alone (red circle in Fig. 4a). Interestingly, this is Next, we repeated this experiment with previously determined IC 50 concentrations of MET (Fig. 2a). High dose MET treatment (60 µM) resulted in a slight left shift of the doxorubicin concentration-response curve in A172 cells (Fig. 5a), while curves overlapped in U87 cells at 100 µM MET (Fig. 5b). Also in primary GBM cells curves largely overlapped independently of treatment (Fig. 5c). Statistical analysis revealed that MET treatment signi cantly reduced doxorubicin IC 50 values in A172 and primary GBM cells indicating a sensitization (Fig. 5d). Looking closer at the curves of primary GBM cells reveals that both curves overlap at most data points except at 1 µM (~ 0.5 µg/mL) doxorubicin (red circle in Fig. 5c), which explains the difference in IC 50 values and indicates that a potential synergism of MET and doxorubicin highly depends on cell line and applied drug concentrations.
Doxorubicin is not used in the rst-line treatment for GBM. Therefore, we studied the effect of MET on the toxicity of TMZ, which is the clinically relevant treatment option. We treated four GBM cell lines and primary GBM cells with ascending TMZ concentrations in combination with 10 µM MET and performed MTT assays. In A172 cells MET even reduced sensitivity towards TMZ (Fig. 6a) while in all other cells no effect was observed (Fig. 6b-d). The only statistical difference in IC 50 values was obtained for A172 cells (Fig. 6e) con rming the negative impact of MET on TMZ treatment in this cell line. Similar results were obtained when we treated cells with an IC 50 MET concentration in combination with TMZ (Fig. 7). At this high MET concentrations both, A172 and U373 cell lines, respectively, were less sensitive to TMZ.

Discussion
MET is a widely used therapeutic opioid in narcotic addiction and neuropathic pain syndromes. In oncological settings it is regularly used as a long-lasting analgesic. Recently it has also been proposed as a chemosensitzing agent in leukemia and GBM therapy based on results of in vitro studies and xenograft mouse models [8][9][10]. In these studies, xed doxorubicin concentrations combined with MET were applied to a subset of GBM cell lines and primary GBM cells and at tested concentrations MET was capable of sensitizing these cells to doxorubicin in terms of apoptosis induction. The proposed mechanism of action is activation of MOR and subsequent induction of apoptosis in addition to P-gp inhibition by MET leading to increased intracellular doxorubicin levels [10]. In order to quantify whether MET alone is cytotoxic in GBM cells, we performed MTT assays with various MET concentrations. MET was indeed cytotoxic in a similar concertation range in various cell lines and primary cells. Yet, recorded IC 50 values were far beyond concentrations, which might be reached in plasma of patients [21]. We also found that high dose MET induced apoptosis and necrosis to a similar extend in most cell lines except for U373 cells where about two thirds of recorded cell deaths accounted for necrosis. This is in line with what has been shown by others that MET can induce apoptosis [10] or necrotic-like cell death in cells of neuronal origin [22,23] depending on cell type. Interestingly, naloxone was not able to block MET-induced cytotoxicity and other MOR agonists, morphine and oxycodone, were hardly cytotoxic at clinically relevant concentrations. This implies that METs cytotoxic action on GBM cells is not MOR dependent. MET might interfere with other cellular targets which are responsible for its cytotoxicity. For instance, MET, in contrast to oxycodone and morphine, also inhibits members of the voltage-gated potassium channel family at low micromolar concentrations [24][25][26]. Pharmacological inhibition of these channels in turn leads to cell death in various tumor cells including GBM cell lines [27] and is proposed as promising target for cancer treatment [28]. However, whether MET indeed blocks voltage-gated potassium channel at the applied concentrations in GBM cells leading to cytotoxicity needs further investigation. When we tested MET in combination with doxorubicin in a panel of GBM cell lines and primary cells we found a sensitizing effect only for the A172 cell line and to some extend in primary GBM cells. We have previously shown that P-gp is not expressed in U87, U251 and U373 cell lines [18] which might explain why these cells could not be sensitized to doxorubicin by MET, while inhibition of doxorubicin e ux has previously been reported for the A172 cell line [10]. Our ndings are in line with a study which also reported no sensitizing effect of MET on various doxorubicintreated canine tumor cells [12].
Another important nding of our study is, that the sensitizing effect of MET in A172 and primary cells highly depends on applied drug concentrations. While a complete left shift of the doxorubicin concentration-response curve upon MET treatment only was observed at high, clinically not relevant MET concentrations (60 µM) in A172 cells, lower MET concentrations (10 µM) only further increased doxorubicin cytotoxicity at one single doxorubicin/MET concentration combination but not over the whole concentration range. This might have clinical implications as it is challenging to exactly achieve the needed doxorubicin and MET levels in vivo where synergistic cytotoxic effects might occur. In addition, tolerable MET plasma levels in addicts are maximally 0.6 µg/mL (corresponding to ~ 2 µM) after a dose of 60-120 mg/day [21,29] which is much lower compared to effective MET concentrations reported by Friesen [10] and in our present study. When a blood-brain barrier penetration of only 42% for MET is assumed [30], su cient brain levels can hardly be reached in GBM patients at feasible MET doses. It could be argued that MET accumulates in tissue stores and also in brain [31], however, if su cient MET levels are indeed are reached in brain tumor tissue at clinically achievable doses warrants further investigation. Apart from that, doxorubicin is not the indicated treatment option for GBM due to its low blood-brain barrier penetration and neurological side effects [13,14]. Although pegylated and liposomal-encapsulated formulations of doxorubicin (Caelyx®) with increased brain uptake and less side effects are available, GBM is rarely treated with doxorubicin in clinical practice. Due to these shortcomings, we also tested the clinically more relevant drug TMZ in combination with MET. One study has already reported no effect of MET on TMZ-induced cytotoxicity in various GBM cell lines [15]. Similar observations were made in another study with GBM cells and MET applied together with TMZ or irradiation (4 Gy) [16]. We could con rm these ndings and most importantly in U373 and A172 cells, MET even reduced sensitivity towards TMZ.

Conclusion
We conclude that cytostatic effects of MET alone on GBM cells are not mediated via the opioid receptor MOR implying that other cellular targets are involved what warrants further investigation. Furthermore, our ndings do not support the use of MET in treatment of GBM in combination with TMZ as no sensitizing effects of MET were observed which is also in line what has been reported by others. The use of doxorubicin in GBM therapy is questionable and we doubt that feasible MET concentrations might be reached in brain of patients to achieve potential synergistic effects with doxorubicin, which we observed in vitro only in two distinct cell types at dedicated drug concentrations. Availability of data and material The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.