A drug combination targeting hypoxia induced chemoresistance and stemness in glioma cells

Abstract

Hypoxia is a characteristic of solid tumors especially Glioblastoma and is critical to chemoresistance. Cancer stem cells present in hypoxic niches are known to be a major cause of the progression, metastasis and relapse. We tried to identify synergistic combinations of drugs effective in both hypoxia and normoxia in tumor cells as well as in cancer stem cells. Since COX-2 is over-expressed in subset of glioblastoma and is also induced in hypoxia, we studied combinations of a prototype Cyclooxygenase (COX-2) inhibitor, NS-398 with various drugs (BCNU, Temozolomide, 2-Deoxy-D-glucose and Cisplatin) for their ability to abrogate chemoresistance under both severe hypoxia (0.2% O₂) and normoxia (20% O₂) in glioma cells. The only effective combination was of NS-398 and BCNU which showed a synergistic effect in both hypoxia and normoxia. This synergism was evident at sub-lethal doses for either of the single agent. The effectiveness of the combination resulted from increased pro- apoptotic and decreased anti-apoptotic molecules and increased caspase activity. PGE₂ levels, a manifestation of COX-2 activity were increased during hypoxia, but were reduced by the combination during both hypoxia and normoxia. The combination reduced the levels of epithelial-mesenchymal transition (EMT) markers. It also resulted in a greater reduction of cell migration. While single drugs could reduce the number of gliomaspheres, the combination successfully abrogated their formation. The combination also resulted in a greater reduction of the cancer stem cell marker CD133. This combination could be a prototype of possible therapy in a tumor with a high degree of hypoxia like glioma.

Keywords: BCNU; COX-2 inhibitor; PGE2; chemosensitization; hypoxia.

Figure 1. COX-2 expression in glioma cell lines

(A) mRNA level expression. COX-2 expression was found in LN229 and U87MG cell lines and it was up-regulated under hypoxia in LN229 (9 fold change, p < 0.001) and U87MG cell line (2.2 fold change, p < 0.01). (B) Protein level expression. COX-2 protein expression was found in LN229 and U87MG cell lines and it was up-regulated under hypoxia in LN229 (1.9 fold change) and U87MG (1.3 fold change) cell lines. Lane N denotes Normoxia control, Lane H denotes Hypoxia control. β-actin was used as a control. Overall there is increase in both COX-2 mRNA and protein expression during hypoxia.

Figure 2

(i) Cell viability analysis at different doses of BCNU and NS-398. Both the cell lines were treated with increasing concentrations of BCNU and NS-398 and cell viability was assessed by MTT assay. Dose dependent cytotoxicity of BCNU and NS-398 was observed in (A, C) LN229 and (B, D) U87MG cell lines under hypoxia and normoxia. Results were expressed as mean ± SD of three experimental replicates. Under all conditions, drug effect of individual drugs was reduced under hypoxia. (ii) Cell viability analysis after treatment with the combination (BCNU + NS-398). Cells were treated with BCNU, NS-398 and BCNU + NS-398 and cell viability was assessed after 72 hours of treatment by MTT assay. The combination of BCNU + NS-398 (75 μM of BCNU and 200 μM of NS-398) showed synergism (CI < 0.9) under both hypoxia and normoxia in both cell lines: (A, B) LN229 and (C, D) U87MG cell lines. Results were expressed as mean ± SD of three experimental replicates. The combination had more effect than either single agent. This was true for normoxia (p < 0.01 with BCNU and p < 0.01 with NS-398) as well as hypoxia (p < 0.01 for BCNU and p < 0.01 for NS-398) for LN229 with reference to this synergistic combination. Also, for U87MG the levels of significance for normoxia were p < 0.01 for BCNU and p < 0.05 for NS-398. For hypoxia the levels of significance were p < 0.05 for BCNU and p < 0.05 for NS-398.

Figure 3. PI and Annexin V staining of glioma cells after drug treatment

(i) LN229 and (ii) U87MG cells were treated with drugs (BCNU, NS-398 and combination) after 24 hours of normoxia and hypoxia and the same state continued throughout the experiment. PI and Annexin V staining was done after 48 hours of drug treatment. It was observed that there was a distinct increase in early apoptosis induced by the combination (BCNU + NS-398) as compared to single agent (BCNU or NS-398). Early apoptosis is depicted by Annexin V positive and PI negative cells. The percentage of cells with staining are depicted below each plot.

Figure 4

(i) Drug induced Caspase 3/7 activity and effect of caspase inhibitor. Cells were treated with BCNU, NS-398 and BCNU + NS-398 and caspase 3/7 activity was assessed after 48 hours. Caspase 3/7 activity was upregulated in the combination of BCNU and NS-398 as compared to single agents under both hypoxia and normoxia in (A) LN229 and (B) U87MG cell lines. Addition of caspase inhibitor reduced drug induced caspase activity. Results were expressed as mean ± SD of three experimental replicates. (ii) Cell viability analysis after treatment with caspase inhibitor. Both cells lines were treated with BCNU, NS-398 and BCNU + NS-398 in presence and absence of caspase inhibitor and MTT assay was done after 72 hours. The caspase inhibitor was able to reverse the decrease in cell viability by both single agent and combination in (A) LN229 and (B) U87MG cell lines. Overall the results indicate the role of caspase 3/7 activity in determining reduced cell viability, induced by the drug combination.

Figure 5

(i, ii) Expression of pro-apoptotic and anti-apoptotic markers under various conditions. This figure shows the quantitative mRNA expression for both pro-apoptotic markers (Bax, cytochrome c and caspase 3) and anti-apoptotic marker (Bcl2) by Real time PCR. There was more up-regulation of the pro-apoptotic markers and more down-regulation of the anti-apoptotic marker by the combination as compared to single agent under both hypoxia and normoxia in (i) LN229 and (ii) U87MG cell lines. Results were expressed as mean ± SD of three experimental replicates. (iii) Western blot analysis of Bcl2 and Bax in the LN229 cell line. The protein levels of Bcl2 and Bax were also studied in the LN229 cell line. Bax was up-regulated and Bcl2 was down-regulated by the combination as compared to single agent under normoxia and hypoxia. Lane C denotes control, Lane B denotes BCNU, Lane N denotes NS-398 and Lane (B + N) denotes BCNU + NS-398.

Figure 6. Relative PGE₂ concentrations indicating COX-2 activity under various conditions

Culture supernatants of drug treated (BCNU, NS-398 and BCNU + NS-398) and control samples of LN229 and U87MG cell lines under hypoxia and normoxia were assayed for PGE₂ concentration by EIA kit. Relative PGE2 concentration was significantly reduced in the cells treated with the combination (BCNU + NS-398) under both hypoxia and normoxia in (A) LN229 and (B) U87MG cell lines. Results were expressed as mean ± SD of three experimental replicates. The combination had more effect than either single agent. This was true for normoxia (p < 0.01 with BCNU and p < 0.05 with NS-398) as well as hypoxia (p < 0.01 for BCNU and p < 0.01 for NS-398) for LN229. Also, for U87MG the levels of significance for normoxia were p < 0.05 for BCNU and p < 0.05 for NS-398. For hypoxia the levels of significance were p < 0.01 for BCNU and p < 0.05 for NS-398.

Figure 7

(i, ii) Real time PCR analysis of mRNA expression of EMT markers vimentin and N-cadherin. mRNA levels of the markers of epithelial to mesenchymal transition (Vimentin and N-cadherin) were more down-regulated by the combination of BCNU and NS-398 as compared to single agent in (i) LN229 and (ii) U87MG cell lines under hypoxia and normoxia. Results were expressed as mean ± SD of three experimental replicates. (iii) Western blot analysis of N-cadherin and Vimentin in the LN229 cell line. The protein levels of Vimentin and N-cadherin were also studied in the LN229 cell line. Here again, the decrease by the combination was more than that for single agent in both hypoxia and normoxia. Lane C denotes control, Lane B denotes BCNU, Lane N denotes NS-398 and Lane (B + N) denotes BCNU + NS-398.

Figure 8. Cell migration as assessed by transwell membrane assay

There was greater reduction in the migration of both cell lines after treatment with the combination as compared to the single agent under both hypoxia and normoxia. A pictorial representation of the cells is given in (i) while the percentage of the migrated cells are depicted in (ii). The combination had more effect than either single agent. This was true for normoxia (p < 0.01 with BCNU and p < 0.05 with NS-398) as well as hypoxia (p < 0.001 for BCNU and p < 0.01 for NS-398) for LN229. Also, for U87MG the levels of significance for normoxia were p < 0.01 for BCNU and p < 0.05 for NS-398. For hypoxia the levels of significance were p < 0.001 for BCNU and p < 0.01 for NS-398.

Figure 9

Photomicrographs of (i) gliomaspheres of U87MG cells. There was reduction in the size of the gliomasphere of the U87MG cells after treatment with the single agent under both hypoxia and normoxia. After treatment with the combination, the formation of gliomaspheres was abrogated in both the conditions, though some small cell clusters of less than 100 μm were observed. (ii) Analysis of CD133 expression in U87MG gliomaspheres. After 48 hours of drug treatment of gliomaspheres under hypoxia and normoxia, expression of CD133 was checked through flow cytometry. There was reduction in the expression of CD133 in gliomaspheres treated with the combination (BCNU + NS-398) as compared to the single agent under both hypoxia and normoxia. The number on each plot depicts the percentage of CD133 positive cells.