Journal: bioRxiv

Article Title: Ex vivo stem-like cell families model evolution of glioblastoma therapeutic resistance

doi: 10.64898/2026.04.02.716158

Figure Lengend Snippet: ( A ) Schematic of ex vivo treatment of whole GBM cell populations and generation of res-GSC families with chemo- or radiotherapy. Whole GBM cell suspensions are cultured in stem cell medium (EFPH20) supplemented with EGF, FGF2, PDGFBB, and HGF (20 ng/mL each) and exposed in parallel to temozolomide (TMZ) or ionizing radiation (IR) as indicated. Each res-GSC family includes TMZ-GSC, IR-GSC, and an untreated CTRL-GSC member. Families were considered established after passage 10. ( B ) Genetic profiling of original GBMs (n = 25) grouped by ability to generate res-GSC families under therapeutic pressure (columns). Top rows: res-GSC family members derived from each GBM (black squares). Bottom rows: altered genes and frequencies in original GBMs assessed by the NGS GBM-target panel. p TERT mutation was evaluated by ddPCR. Purple: GBMs yielding res-GSC families; light orange: GBMs yielding only CTRL-GSCs. ( C ) Associations between genetic alterations in original GBMs and res-GSC family derivation. Bars indicate the number of GBMs that are wild-type (wt) or altered (alt) for each gene and that did (dark gray) or did not (light grey) yield res-GSC families. Thicker dotted outline: tendency to anti-correlation between EGFR alterations and res-GSC family generation (p=0.40). PTEN loss and PIK3CA alterations showed gene-level trends (p = 0.06 and p > 0.99, respectively), while combined AKT-pathway alterations ( PTEN and PIK3CA ) were significantly associated with res-GSC family derivation (p = 0.03). TP53 mutations were also significantly associated (p = 0.0005) (Fisher’s exact test). Red: oncogenes. Blue: tumor suppressor genes. ( D-E ) Derivation times of TMZ-GSCs (D, upper) and IR-GSCs (E) from GBM processing (as in B ) to p5. Bars show the percentage increase in derivation time relative to the matched CTRL-GSC. Grey dotted bars: not derived GSCs. Absolute derivation times (days) are reported in . In ( D ), dots indicate the MGMT status of the original GBM (or of CTRL-GSC when the GBM status was unavailable). Lower panel: schematic of the functional relationship between MGMT promoter methylation and MGMT transcription (arrow).

Article Snippet: To detect possibly rare gene alterations in original UA, ddPCR was performed using probe–based assays with ddPCR Supermix for Probes (no dUTP) (Biorad), according to manufacturer’s instructions.

Techniques: Ex Vivo, Cell Culture, Derivative Assay, Mutagenesis, Functional Assay, Methylation

Journal: bioRxiv

Article Title: Ex vivo stem-like cell families model evolution of glioblastoma therapeutic resistance

doi: 10.64898/2026.04.02.716158

Figure Lengend Snippet: ( A ) Pathogenic alterations (variant allele frequency, VAF, and gene copy number variation, CNV) detected by the NGS GBM-target panel or ddPCR (p TERT ) in MGMT-neg whole GBMs and matched CTRL-GSCs and TMZ-GSCs. MSS: microsatellite stable; MSI: microsatellite instable, assessed at passage 10. ( B ) MMR activity measured by FM-HCR in matched CTRL-GSC and TMZ-GSC from GBM17 and GBM151 families, 24 hours after transfection. Data are z-values (mean ± SEM). n = 2 independent experiments for GM17 and n = 3 independent experiments for GBM151, Student’s paired t test. GBM17: *p=0.018734; GBM151: *p= 0.006080. ( C ) Ploidy (n) in MGMT-pos (orange square) and MGMT-neg (green square) TMZ-GSCs and CTRL-GSCs assessed by chromosome G-banding. ( D ) Cumulative chromosomal aberration number in all TMZ-GSCs vs. matched CTRL-GSCs, assessed by G-banding and multicolor fluorescent in situ hybridization (M-FISH). Data are shown as delta versus the corresponding CTRL-GSC; median is indicated. One sample Wilcoxon test. *p=0.0156.

Article Snippet: To detect possibly rare gene alterations in original UA, ddPCR was performed using probe–based assays with ddPCR Supermix for Probes (no dUTP) (Biorad), according to manufacturer’s instructions.

Techniques: Variant Assay, Activity Assay, Transfection, In Situ Hybridization

Journal: bioRxiv

Article Title: Ex vivo stem-like cell families model evolution of glioblastoma therapeutic resistance

doi: 10.64898/2026.04.02.716158

Figure Lengend Snippet: ( A ) BRAF (top panel) and TP53 (bottom panel) mutations evaluated by ddPCR in GBM149 parental GBM, CTRL-GSC and TMZ-GSC. Data are shown as fractional abundance of the indicated mutant (black) and wild type (grey) alleles. ( B ) Representative metaphase images showing chromosome number and aberrations in all res-GSC families evaluated by G-banding and M-FISH.

Article Snippet: To detect possibly rare gene alterations in original UA, ddPCR was performed using probe–based assays with ddPCR Supermix for Probes (no dUTP) (Biorad), according to manufacturer’s instructions.

Techniques: Mutagenesis

Journal: bioRxiv

Article Title: Ex vivo stem-like cell families model evolution of glioblastoma therapeutic resistance

doi: 10.64898/2026.04.02.716158

Figure Lengend Snippet: ( A ) Genetic profiling of matched CTRL-GSCs and IR-GSCs using the NGS GBM-target panel; p TERT mutations were assessed by ddPCR. Samples are ordered by primary and secondary radioresistance. *Significant increase in MYCN copy number; see also A Table EV4. ( B ) Ploidy (n) of matched CTRL-GSCs e IR-GSCs assessed by chromosome G-banding. ( C ) Cumulative chromosomal aberration number in all IR-GSCs vs. matched CTRL-GSCs, assessed by G-banding and M-FISH. Median number is indicated. Data are shown as delta relative to the corresponding CTRL-GSC; median is indicated. One-sample Wilcoxon test. *p = 0.0313. ( D ) Western blots showing phospho-H2AX (γH2AX), phospho-CHK2 (pCHK2; Tyr68), total CHK2 and p21 in GBM95 CTRL-GSC and IR-GSC in a representative time-course experiment after a single 5 Gy dose. Calnexin: loading control. ( E-F ) Immunofluorescence for phospho-H2AX (γH2AX) in GBM95 CTRL-GSC and IR-GSC in a time-course experiment after a single 5 Gy dose. ( E ) Quantification of γH2AX-positive cell nuclei. n = 4 technical replicates, Student’s paired t test. * p<0.000001. ( F ) Representative images. Scale bar, 25 μm. ( G ) Fraction of GBM95 CTRL-GSC and IR-GSC cells in G2 and M phases, measured by flow cytometry using phospho-Histone 3 (pH3; Ser10) staining at 6 and 24 h after a single 5 Gy dose, shown as fold change vs. mock. Mock: no IR. n = 2 independent experiments, Student’s paired t test. ( H ) Schematic of adaptive mechanisms of secondary radioresistance in GBM95 GSC family. In (E, G), data are presented as mean ± SEM.

Article Snippet: To detect possibly rare gene alterations in original UA, ddPCR was performed using probe–based assays with ddPCR Supermix for Probes (no dUTP) (Biorad), according to manufacturer’s instructions.

Techniques: Western Blot, Control, Immunofluorescence, Flow Cytometry, Staining