A Rapid Screening Workflow to Identify Potential Combination Therapy for GBM using Patient-Derived Glioma Stem Cells

Summary

The glioma stem cells (GSCs) are a small fraction of cancer cells which play essential roles in tumor initiation, angiogenesis, and drug resistance in glioblastoma (GBM), the most prevalent and devastating primary brain tumor. The presence of GSCs makes the GBM very refractory to most of individual targeted agents, so high-throughput screening methods are required to identify potential effective combination therapeutics. The protocol describes a simple workflow to enable rapid screening for potential combination therapy with synergistic interaction. The general steps of this workflow consist of establishing luciferase-tagged GSCs, preparing matrigel coated plates, combination drug screening, analyzing, and validating the results.

Abstract

The glioma stem cells (GSCs) are a small fraction of cancer cells which play essential roles in tumor initiation, angiogenesis, and drug resistance in glioblastoma (GBM), the most prevalent and devastating primary brain tumor. The presence of GSCs makes the GBM very refractory to most of individual targeted agents, so high-throughput screening methods are required to identify potential effective combination therapeutics. The protocol describes a simple workflow to enable rapid screening for potential combination therapy with synergistic interaction. The general steps of this workflow consist of establishing luciferase-tagged GSCs, preparing matrigel coated plates, combination drug screening, analyzing, and validating the results.

Introduction

Glioblastoma (GBM) is the most common and aggressive type of primary brain tumor. Currently, the overall survival of GBM patients who received maximal treatment (a combination of surgery, chemotherapy, and radiotherapy) is still shorter than 15 months; so novel and effective therapies for GBM are urgently required.

The presence of glioma stem cells (GSCs) in GBM constitutes a considerable challenge for the conventional treatment as these stem-like cells play pivot roles in the maintenance of tumor microenvironment, drug resistance, and tumor recurrence¹. Therefore, targeting GSCs could be a promising strategy for GBM treatment². Nevertheless, a major drawback for the drug efficacy in GBM is its heterogenetic nature, including but not limited to the difference in genetic mutations, mixed subtypes, epigenetic regulation, and tumor microenvironment which makes them very refractory for treatment. After many failed clinical trials, scientists and clinical researchers realized that single-agent targeted therapy is probably incapable of fully controlling the progression of highly heterogeneous cancers such as GBM. Whereas, carefully selected drug combinations have been approved for their effectiveness by synergistically enhancing the effect of each other, thus providing a promising solution for GBM treatment.

Although there are many ways to evaluate the drug-drug interactions of a drug combination, such as the CI (Combination Index), HSA (Highest Single Agent), and Bliss values, etc.^3,4, these calculation methods are usually based on multiple concentration combinations. Indeed, these methods can provide affirmative assessment of drug-drug interaction but can be very laborious if they are applied in high-throughput screening. To simplify the process, a screening workflow for rapidly identifying the potential drug combinations that inhibit the growth of GSCs originated from surgical biopsies of patient GBM was developed. A sensitivity Index (SI) that reflects the difference of the expected combined effect and the observed combined effect was introduced into this method to quantify the synergizing effect of each drug, so the potential candidates can be easily identified by the SI ranking. Meanwhile, this protocol demonstrates an example screen to identify the potential candidate(s) that can synergize the anti-glioma effect with temozolomide, the first-line chemotherapy for GBM treatment, among 20 small molecular inhibitors.

Protocol

GBM specimen was acquired from a patient during a routine operation after obtaining fully informed consent by human research ethics committee of The First Affiliated Hospital of Nanjing Medical University.

1. Isolation and culture of patient-derived GSCs

1. Place fresh surgically resected glioblastoma tissue in a 15 mL centrifuge tube filled with sterile PBS and store the tissue on ice until further operation.
2. Mince the GBM tissue into approximately 0.5 to 1 mm diameter pieces using dissection scissors and wash the tissue specimens with neuronal basal medium to remove cellular debris in a biosafety cabinet.
3. Digest the tissue fragments with 1 mg/mL collagenase A at 37 °C for 30 min and centrifuge at 400 x g for 5 min at 4 °C.
4. Remove the supernatant and suspend the pellet with blank neuronal basal medium and dissociate the pellet mechanically by repetitive pipetting on ice.
5. Culture the mixture in ultra-low attachment 6-well culture plates filled with GSC culture medium (see Table 1 for the recipe) in a sterile cell incubator with 5% CO₂ and 90% humidity at 37 °C until neurosphere formation.
6. On sufficient neurosphere formation, collect them using a pipette in a 1.5 mL microtube and centrifuge at 800 x g for 5 min at room temperature.
7. Resuspend the pellet and split it into several flasks filled with the above culture medium for maintaining the primary GSCs.
   NOTE: Patient-derived GSCs used in the example were derived from surgical biopsies of a 34-year-old male patient with WHO grade IV recurrent GBM. The GSCs were named as XG387 for the future experiments. PCR-based mycoplasma tests were performed for the above GSCs to confirm no mycoplasma contamination is present. All the experiments involving GSCs used in this protocol were carried out <15 passages.

2. Preparing luciferase-tagged GSCs

1. Collect the GSCs from the medium culture and centrifuge them at 70 x g for 3 min at room temperature.
2. Remove the supernatant, digest the cells with accutase for 4 min at 37 °C. Use a 200 µL tip and pipette repeatedly to dissociate and resuspend the cell pellet.
3. Dilute the cells to 2 x 10⁵ cells per well in a 12-well culture plate and culture the cells overnight.
4. Add 30 µL luciferase-EGFP virus supernatant (titer >10⁸ TU /mL) into each well in the plate and then centrifuge the cells at 1,000 x g for 2 h at 25 °C. Culture the cells overnight.
5. Refresh the medium the next day and culture the cells for another 48 h.
6. Observe the cells under a florescent microscope to confirm the appearance of the GFP positive cells.
7. Use a flow cell sorter to sort and select the GSCs with high GFP fluorescence to culture the cells further.

3. Bio-luminescence based measurement of cell viability

1. Coating plates with the extracellular matrix (ECM) mixture (e.g., Matrigel): Add 40 µL of 0.15 mg/mL ECM mixture to each well and incubate the plate for 1 h at 37 °C. Remove the excess ECM mixture and gently rinse once with PBS.
2. Add 100 µL culture medium containing 15,000, 10,000, 8,000, 6,000, 4,000, 2,000, 1,000, and 500 XG387-Luc cells together with 100 µL blank medium as control into each well for 6 replicates in a 96-well optical bottom plate and culture the cells overnight at 37 °C.
3. Remove the supernatant, add 50 µL culture medium containing 150 ng/µL D-luciferin into each well and incubate the cells for 5 min at 37 °C.
4. Take images of the cellular bio-luminescence in the plate using the IVIS spectrum imaging system. Use the built-in software to create multiple circular areas of the region of interest (ROI) and quantify the cellular bio-luminescence.

4. Temozolomide treatment and combination screening

1. Precoat four 96-well plates as described above, prior to the treatment.
2. Seed XG387-Luc cells at a density of 1,000 cells in 100 µL culture medium into each well of a 96-well optical bottom plate and culture the cells overnight.
3. Prepare temozolomide and the targeted agents from the stock solution in advance. Prepare a concentration series composed of 800 µM, 600 µM, 400 µM, 300 µM, 200 µM, 100 µM, and 50 µM temozolomide in culture medium for the single-agent treatment. Dilute temozolomide and the targeted agents in stock solution in the culture medium, respectively, to obtain final concentrations of 200 µM and 2 µM for combination drug screening (Table 2).
4. Remove the culture medium when most of the GSCs adhere to the bottom of the plates; add the above-prepared medium containing temozolomide into each well for three technical replicates per treatment.
5. To treat Temozolomide and to screen the drug combinations remove the blank medium and add the above-prepared medium containing either 200 µM temozolomide, or 2 µM targeted agent, or a combination of both into each well for three technical replicates per treatment.
6. Incubate all plates at 37 °C, 5% CO2 for 3 days.
7. Remove the drug-containing medium, add 50 µL blank medium containing 150 ng/µL D-luciferin into each well and incubate the cells for 5 min at 37 °C.
8. Take images of the cellular bio-luminescence in the plate using the IVIS spectrum imaging system. Use the built-in software to create multiple circular ROIs and quantify the cellular bio-luminescence.

5. Combination treatment of temozolomide and UMI-77 in XG387-Luc and XG328-Luc cell lines

1. Precoat three 96-well plates as described above, prior to the treatment.
2. Seed XG387-Luc and XG328-Luc cells at a density of 1,000 cells respectively in 100 µL of culture medium into each well of a 96-well optical bottom plate and culture the cells overnight.
3. Prepare a concentration series composed of 600 µM, 400 µM, 300 µM, 200 µM, 100 µM, 50 µM, and 0 µM temozolomide and a concentration series composed of 6 µM, 4 µM, 3 µM, 2 µM, 1 µM, 0.5 µM, and 0 µM UMI-77 in the culture medium for six-by-six dose titration matrix treatments.
4. Remove the blank medium when most of the GSCs adhere to the bottom of the plate; add the above-prepared medium into each well for three technical replicates per treatment.
5. Incubate these plates for 3 days at 37 °C, 5% CO₂.
6. Remove the drug-containing medium; add 50 µL of blank medium containing 150 ng/µL D-luciferin into each well and incubate the cells for 5 min at 37 °C for bioluminescence measurement.

6. Data analysis

1. Calculate the sensitivity Index (SI) score of temozolomide and targeted agent according to the formula in Figure 2A.
   NOTE: The SI score is to quantify the influence of the addition of another drug. It ranged from -1 to +1, with positive values indicating temozolomide synergistic effects.
2. Calculate the combination index (CI) values between temozolomide and UMI-77 using CompuSyn software to analyze their combined interactions. CI value <1 indicates synergy; CI value >1 indicates antagonism.
3. Calculate the high single agent (HSA) values between temozolomide and UMI-77 using Combenefit software. HSA value indicates the combined inhibitory effect. HSA value >0 indicates synergy and the HSA value <0 indicates antagonism.

Representative Results

The XG387 cells formed neurospheres in the culture medium described in the Table 1 in an ultra-low attachment 6-well culture plate or a non-coated plate⁵ (Figure 1A). First, a test was performed to check whether the bio-luminescence intensity from XG387-Luc cells was proportional to the cell number. As shown in Figure 1B, the bio-luminescence intensity increased proportionally to the cell density and resulted in a linear correction between them (Pearson r = 0.9872; p < 0.0001; Figure 1C). Since the bio-luminescence of luciferase tagged cells is easy and quick to measure, this provides a simple method to measure the density of viable GSCs. Next, the anti-proliferative activity of temozolomide was assessed. As shown in Figure 1D, 400 µM temozolomide caused approximately 20% proliferation inhibition of XG387-Luc cells, suggesting it is useful but its anti-GBM effect can be further improved. The concentration of 200 µM was selected for the combination screening.

To give an example, 20 target-selective small molecule inhibitors was utilized for the drug combination screening to identify the potential candidate(s) that enhances the anti-GBM effect of temozolomide. As a result, the sensitive index (SI) values of 13 targeted agents were above 0, and 5 of them were above 0.1 (Figure 2B,C). Especially, the SI of the top two candidate drugs UMI-77 and A 83-01 were higher than 0.25, suggesting their potential to synergize with temozolomide.

To validate the above finding, the classical synergy models of HSA and Bliss^3,4,6 were applied to determine the combined effect of temozolomide and UMI-77 in GSCs. In addition, XG328-another patient-derived GSC model established early-was used to perform the same evaluation. Anti-proliferative assay of the combined treatment of temozolomide and UMI-77 was performed in a six-by-six dose titration matrix. The results were analyzed to acquire the HSA and Bliss values which are readouts for synergistic inhibition and depict the difference between the expected inhibition and the observed inhibition. As shown in Figure 3B-D, the combination index (CI) values <1 and the high single agent (HSA) values >0 for most of the combinations of temozolomide and UMI-77 at different concentrations, suggests an overall synergistic interaction of temozolomide and UMI-77 in both XG387 and XG328 GSCs.

Figure 1: GBM patient-derived GSCs XG387. (A) Neurosphere formation. (B) Bio-luminescence generation of luciferase tagged GSCs. (C) Bio-luminescence generated by XG387-Luc cells was proportional to the cell density. (D) Temozolomide treatment of XG387. Each treatment was performed in triplicate with two independent experiments. The data are expressed as the mean ± SD. Please click here to view a larger version of this figure.

Figure 2: Drug combination screening using GSCs. (A) The formula to calculate the SI (sensitivity index) of 20 targeted agents with temozolomide (TMZ) in the combination screen. (B) The distribution of SI values of 20 targeted agents. Red dots: the top five candidate drugs. (C) Information of the top five candidate targeted agents. Please click here to view a larger version of this figure.

Figure 3: Combination treatment of temozolomide (TMZ) and UMI-77 in XG387-Luc and XG328-Luc cell lines. (A) Single and combinatorial titration of temozolomide and UMI-77 in a proliferation assay in XG387-Luc and XG328-Luc cell lines. (B) Isobologram and combination index analysis of the proliferation inhibition in XG387-Luc and XG328-Luc cells treated with temozolomide and UMI-77. CI <1 indicates a synergistic effect. (C,D) Synergy plots generated by Combenefit showing the interaction between temozolomide and UMI-77. Analysis of interaction resulted in HSA (high single agent) values and Bliss values, indicating synergistic efficacy as calculated from the expected. HSA and Bliss values >0 indicate synergistic effects. Each treatment was performed in triplicate with two independent experiments. Please click here to view a larger version of this figure.

Discussion

In the present study, a protocol that can be applied to identify potential combination therapy for GBM using patient-derived GSCs was described. Unlike the standard synergy/additivity metric model such as Loewe, BLISS, or HSA methods, a simple and quick workflow was used that does not require a drug pair to be combined at multiple concentrations in a full factorial manner as the traditional methods. In this workflow, SI (sensitivity index) which is originated from a study to evaluate the sensitizing effect of siRNAs in combination with small molecular inhibitor was introduced to quantify the synergistic drug effect of two small molecular inhibitors⁷. The range of SI values is from -1 to 1, and the positive SI value indicates a sensitizing effect between each of the drugs. The higher the SI value achieved, the stronger the synergy was. Although the SI value alone is incapable to provide an affirmative answer about the type (synergistic, additive, or antagonistic) of drug-drug interaction, those top-ranked candidates have high probability to synergize with the drug of interest and therefore are worthy of further validation. In comparison, most of the current high-throughput drug combination screening methodologies are still laborious and involve difficult algorithms^8,9.

To exemplify the feasibility of this method, a small-scale test screen was performed. As a result, it was possible to identify UMI-77, a selective MCL1 inhibitor, as the top candidate among 20 targeted agents to synergize with temozolomide in GSCs growth suppression. In fact, in a previous study, UMI-77 was also found to synergistically enhance the anti-glioma activity of temozolomide in established GBM cells¹⁰. In the current study, the synergistic interaction between UMI-77 and temozolomide was approved again in GSCs using the classical Chou-Talalay combination index, BLISS or HSA methods. Another advantage of this protocol is the usage of luciferase-tagged GSCs for measuring the viable proportion of cells. The luciferase activity of cells can be easily measured by the addition of the luciferin, the substrate of luciferase, and capture the luminescence by any instrument with the function of luminometric measurement. Because the luciferase-luciferin reaction is quick, herein it provides a cheap and quick solution in comparison with traditional MTT (3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide), MTS(3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H tetrazolium), or CCK-8 (cell counting kit-8) assays, all of which require long incubation times. Together, the protocol presents a high-throughput screening of potential drug combination for GBM. The protocol also provides optional quick and simple solution for drug combination screen in addition to the standard synergy evaluation methods.

Materials

Name	Company	Catalog Number	Comments
B-27	Gibco	17504-044	50X
EGF	Gibco	PHG0313	20 ng/ml
FGF	Gibco	PHG0263	20 ng/ml
Gluta Max	Gibco	35050061	100X
Neurobasal	Gibco	21103049	1X
Penicillin-Streptomycin	HyClone	SV30010	P: 10,000 units/ml     S:  10,000 ug/ml
Sodium Pyruvate	Gibco	2088876	100 mM
Table 1. The formulation of GSC complete culture medium.
ABT-737	MCE		Selective and BH3 mimetic Bcl-2, Bcl-xL and Bcl-w inhibitor
Adavosertib (MK-1775)	MCE		Wee1 inhibitor
Axitinib	MCE		Multi-targeted tyrosine kinase inhibitor
AZD5991	MCE		Mcl-1 inhibitor
A 83-01	MCE		Potent inhibitor of TGF-β type I receptor ALK5 kinase
CGP57380	Selleck		Potent MNK1 inhibitor
Dactolisib (BEZ235)	Selleck		Dual ATP-competitive PI3K and mTOR inhibitor
Dasatinib	MCE		Dual Bcr-Abl and Src family tyrosine kinase inhibitor
Erlotinib	MCE		EGFR tyrosine kinase inhibitor
Gefitinib	MCE		EGFR tyrosine kinase inhibitor
Linifanib	MCE		Multi-target inhibitor of VEGFR and PDGFR family
Masitinib	MCE		Inhibitor of c-Kit
ML141	Selleck		Non-competitive inhibitor of Cdc42 GTPase
OSI-930	MCE		Multi-target inhibitor of Kit, KDR and CSF-1R
Palbociclib	MCE		Selective CDK4 and CDK6 inhibitor
SB 202190	MCE		Selective p38 MAP kinase inhibitor
Sepantronium bromide (YM-155)	MCE		Survivin inhibitor
TCS 359	Selleck		Potent FLT3 inhibitor
UMI-77	MCE		Selective Mcl-1 inhibitor
4-Hydroxytamoxifen(Afimoxifene)	Selleck		Selective estrogen receptor (ER) modulator
Table 2. The information of 20 targeted agents used in the test screen. All of these are target selective small molecular inhibitors. The provider, name, and targets were given in the table.