Contributions to drug resistance in glioblastoma derived from malignant cells in the sub-ependymal zone

Abstract

Glioblastoma, the most common and aggressive adult brain tumor, is characterized by extreme phenotypic diversity and treatment failure. Through fluorescence-guided resection, we identified fluorescent tissue in the sub-ependymal zone (SEZ) of patients with glioblastoma. Histologic analysis and genomic characterization revealed that the SEZ harbors malignant cells with tumor-initiating capacity, analogous to cells isolated from the fluorescent tumor mass (T). We observed resistance to supramaximal chemotherapy doses along with differential patterns of drug response between T and SEZ in the same tumor. Our results reveal novel insights into glioblastoma growth dynamics, with implications for understanding and limiting treatment resistance.

©2014 American Association for Cancer Research.

Figures

Figure 1. Identification of SEZ in human GB

(A) Top panel. Illustrative MRI scans of a GB in the right temporal lobe. Red dots indicate tissue sampling from T and SEZ. Middle panel. Peri-operative images of visible fluorescence from T and SEZ indicating the presence of tumor tissue. Bottom panel. Haematoxylin&Eosin staining of T and SEZ. Both show cardinal features of GB (high mitotic index, nuclear atypia, mitosis and microvascular proliferation). Scale bar, 100 μm. (B) T and SEZ tissues have been stained for Gfap (green), Nestin (red), CD31 (pink) and counterstained with DAPI (blue). SEZ is characterized by Gfap expression and high vascularization. Scale bar, 100 μm. (C) Quantitative analysis of Gfap-, Nestin- and CD31-positive cells in T and SEZ tissues of the same patient.

Figure 2. The SEZ harbors residual disease

(A) Summary of the most common putative drivers in GB. In 6/8 patients the SEZ contains an equal or smaller number of aberrations with respect to its corresponding T. Cultured cells from T and SEZ of patient sp14 have been used for comparison with the corresponding tissues. (B) FISH of 3 patients confirms gain of the region of chromosome 7 including EGFR and the centromere (centr 7) in sp49 and sp58 (T and SEZ in both cases) and gain of the region including MET and centr 7 in sp58 and cells from sp14 (T and SEZ in both cases); as expected, based on copy number data, gain of MET was not observed in sp49T. Orange denotes EGFR or MET and green the centr 7. The single nuclei are representative of what was generally observed in each sample. (C) Clustering of gene expression profiles from 9 patients reveals that 5 SEZ samples cluster tightly together instead of with their matched T. Samples are color-coded based on their origin. (D) Each compartment of 9 GBs has been assigned to a previously described classifier (6): T samples are either Proneural, Classical, Mesenchymal or Neural and 7/9 SEZ are Mesenchymal.

Figure 3. TICs can be isolated from SEZ of GB patients

(A) Cells from T and SEZ form spheres in serum-free medium. Magnification 100×; scale bar 50μm. (B) Left. Growth curve analysis shows that cells from T and SEZ are capable of long-term expansion. Right. Limiting-dilution assays reveal similar frequencies of sphere-formation in T and SEZ. Each data point represents the average of triplicates. Error bars, s.d. (C) Top. Cumulative Kaplan-Meier survival analysis of 30 animals reveals the tumorigenic potential of T and SEZ cells from 5 GBs. There is a statistically significant difference (p-value <0.05, Logrank test) between the survival of animals injected with T cells versus those injected with SEZ cells. Bottom. Table summarizing the number of tumors/injections and the type of cells used (primary or briefly cultured). (D) Haematoxylin&Eosin staining of T and SEZ cells-derived xenografts. Scale bar, 30 μm.

Figure 4. TICs from SEZ contribute to drug resistance

Cells from T and SEZ of 4 patients were used to evaluate response to chemotherapeutics. (A) Differential response patterns to temozolomide (TMZ) were observed in T and SEZ cells of the same patients. Dose escalation has little impact on drug-resistant cells. (B) Distinct patterns of response are observed when the assay is performed with Cisplatin and Cediranib. Data are shown as normalized to control. The mean normalized absorbance as well as the minimum (−) and maximum (+) absorbance are plotted. Asterisks indicate statistical significance.

Figure 5. Phylogenetic reconstruction reveals different patterns of tumor evolution involving SEZ malignant cells

(A) Cartoon illustrating the phylogeny based on molecular clock analysis. Non-dividing tissues, such as heart and brain, do not undergo mitotic events and do not accumulate methylation errors. Epithelial tissues, such as colon, show increased methylation with age due to the mitotic turnover of cells maintaining tissue homeostasis. Tumors undergo a large number of divisions and accumulate methylation changes in a manner proportional to the number of mitotic events. (B, C) Top. The IRX2 locus on chromosome 5 (in B) and the NETO1 locus on chromosome 18 (in C) have been validated as molecular clock loci by verifying that age-related methylation increases as a result of cell division in patient-derived tissues by comparing heart, brain and colon. Bottom. The graphs show tumor cellularity using IRX2 (in B) and NETO1 (in C) molecular clock data. The values of cellularity indicate no significant difference between T and SEZ. (D) Tumor phylogenies based on SEZ and multiple tumor mass samples (T, T1-T6) were reconstructed independently with 2 methods: copy number breakpoints (a) and two molecular clock loci (b, IRX2 and c, NETO1). As an example, in sp52 the SEZ (blue) harbors cells that arise earlier than tumor mass cells (red) (sample T1 failed hybridization). NA= not available.