To explore the molecular genetic characteristics of primary and recurrent glioblastomas (GBMs) from the same patient in vivo, primary glioma stem cells cultured in vitro, and patient-derived xenograft (PDX).

(1) The primary and recurrent GBM specimens from one patient during surgical resection were collected; and the expressions of glial fibrillary acidic protein (GFAP), nestin and Ki-67 were detected by immunohistochemical staining; the methylation of O6-methylguanine DNA methyltransferase (MGMT) gene, mutation of isocitrate dehydrogenase (IDH) gene and amplification of epidermal growth factor receptor (EGFR) gene were analyzed. (2) The primary and recurrent GBM stem cells were cultured in vitro and named as SU5-1 and SU5-2 cells, respectively; the expressions of nestin and CD133 were detected by immunohistochemical staining; GFAP expression was detected by immunohistochemical staining after induced differentiation, and the growth curve was detected by CCK-8 assay; Transwell invasion assay was used to detect the invasion ability; cell resistance to temozolomide (TMZ), carboplatin (CBP), cisplatin (DDP) and adriamycin (ADM) was detected by CCK-8 assay; the protein expression of programmed death receptor-ligand 1(PD-L1) was detected by Western blotting. The rate of PD-L1 positive cells was detected by flow cytometry; genetic testing analysis was as above. (3) The primary and recurrent in situ PDX models in nude mice were established, and the expressions of nestin, GFAP and Ki-67 were detected by immunohistochemical staining.

(1) As compared with the primary GBM, the recurrent GBM had significantly higher percentages of Ki-67 and nestin positive cells, while statistically lower percentage of GFAP positive cells (P<0.05); genetic analysis showed that there was no mutation in IDH gene in the primary GBM tissues and recurrent GBM tissues; the MGMT gene in the primary GBM tissues was methylated and EGFR gene was not amplified, while the MGMT gene in recurrent GBM tissues was demethylated and EGFR gene amplification was positive. (2) Both SU5-1 and SU5-2 cells expressed nestin and CD133, and GFAP was expressed after induced differentiation; the growth curve showed that the proliferation of SU5-2 cells started earlier than that of SU5-1 cells, the two were equal on the 3^rd, 4^th, and 5^th d, and the proliferation of SU5-1 cells was faster than that of SU5-2 cells from the 6^th d; the invasion ability of SU5-2 cells was statistically stronger than that of SU5-1 cells (P<0.05); the inhibition rates of SU5-2 cells treated with 5, 10, and 15 mmol/L CBP, 0.3125, 1.25, and 5 mmol/L DDP, 0.5 and 2 mmol/L ADM, and 125 and 500 mmol/L TMZ were significantly lower than those of SU5-1 cells treated with the same concentrations and same drugs (P<0.05); the protein expression of PD-L1 in SU5-2 cells was higher than that in SU5-1 cells; the positive rate of PD-L1 in SU5-2 cells was statistically higher than that in SU5-1 cells (P<0.05); the results of genetic analysis were consistent with those of the primary and recurrent GBM samples. (3) As compared with those in the primary PDX model, the nestin and Ki-67 expressions were significantly higher and GFAP expression was significantly lower in the recurrent PDX model (P<0.05); the results of genetic analysis were consistent with those of the primary and recurrent GBM samples.

Genetic differences are detected between primary and recurrent GBMs; recurrent GBM has stronger invasive capacity and multi-drug resistance. The primary stem cells derived from surgical specimens and corresponding PDX models could replicate the molecular genetic characteristics of original tumors, which provide a reliable experimental platform for both tumor translation researches and screening of molecular therapeutic targets.

Glioblastoma; Molecular genetics; Patient-derived xenograft model; Tumor stem cell; Recurrence