Discovery and therapeutic exploitation of Master Regulatory miRNAs in Glioblastoma

Abstract

Glioblastoma is a fatal primary malignant brain tumor, with an average survival of only 15 months despite surgical resection, chemotherapy, and radiation therapy. Due to the concurrent deregulation of numerous genes in glioblastoma, molecular monotherapies have not improved clinical outcomes. Evidence suggests that effectively targeting multiple deregulated molecules is essential for better therapies; however, this is limited by the lack of suitable drugs and the increased toxicity of combination therapies. To address this, we hypothesized that miRNAs, small gene-regulatory RNAs that suppress multiple target genes via sequence complementarity, could be developed to inhibit multiple deregulated genes simultaneously, leading to more effective treatments. We identified master regulatory miRNAs-those that target several deregulated genes in glioblastoma-using PAR-CLIP screenings in glioblastoma cells and analyzed TCGA tumor data to find which targets were deregulated. An algorithm ranked these targets based on their significance in glioblastoma malignancy. We selected two tumor suppressor master regulatory miRNAs, miR-340 and miR-382, and one oncogenic miRNA, miR-17. Validation showed that these miRNAs target critical glioblastoma pathways and significantly inhibit cell growth, survival, invasion, and tumor growth in vivo. We developed an innovative therapeutic delivery approach using Brain Penetrating Nanoparticles in combination with MRI-guided focused ultrasound and microbubbles, resulting in reduced tumor volume and extended survival in glioblastoma-bearing mice. This strategy offers a promising pathway for translating miRNA-based therapies into clinical trials for glioblastoma and other cancers.

Figure 1:. Identification and prioritization of master regulatory miRNAs in glioblastoma:

A) Schematic overview of PAR-CLIP. B) The phosphorimage of SDS-PAGE gel of RNA-Argonaute (Ago) complexes labeled with 5′-32P that were immunoprecipitated with a Flag-tag antibody. The complex is expected to appear near 100 kDa. C) Agarose gel separation of PCR products from AGO1, AGO2, AGO3 PAR-CLIP cDNA libraries. D) A Venn diagram illustrating the overlap in miRNA targets identified with AGO1, AGO2, AGO3. E) Bar graph showing the distribution of identified targets within coding and 3’-UTR regions. F) Diagram depicting the algorithm used for determining master regulatory miRNAs in glioblastoma. G) Volcano plot denoting the significant miRNAs based on Cox Propportional Hazard. H) Classification of miRNAs as oncogenic and tumor-suppressive based on their targets derived from AGO1, AGO2, AGO3 PAR-CLIP.

Figure 2:. Validation of miR-340 and miR-382 targets in glioblastoma cells and patient-derived stem cells:

A) Glioblastoma cell lines A172, U87, U251, and patient-derived stem cell lines GSC-28 and GSC-34 were transfected with either a scrambled negative control, miR-340, or miR-382. Immunoblots were probed with antibodies against TOP2A (A), RHOC (B), CD44 (C), HMGA2 (D), and MDM2 (E), CD44 (F), NUSAP1 (G), PLAU (H), and HMGA2 (I). GAPDH served as the internal loading control. The data show that miR-340 and miR-382 downregulated protein expression compared to negative controls. J,K) 3’UTR Luciferase activity assay in U87 cells were performed by co-transfecting cells with miR-340 (J) or miR-382 (K) and a psiCheck2 luciferase reporter plasmid containing the 3’-UTR regions of targets CD44, TOP2A, RHOC, HMGA2, MDM2, EGFR, PDGFRA (J), or PLAU, CD44, NUSAP1, HMGA2, and MDM2 (K). The data show that miR-340 and miR-382 decreased luciferase activity for all respective targets compared to controls. Statistical significance was determined using a two-tailed Student’s t-test, with * = P<0.05.

Figure 3:. Validation of miR-17 targets in glioblastoma cells and patient-derived stem cells:

A-D) Glioma cell lines A172, U251, GSC-28, and GSC-34 were transfected with a miRNA inhibitor targeting miR-17. Immunoblots were performed with antibodies against ZBTB4, ANKRD11, EHD3, and EPHA4. GAPDH served as the internal control for loading. The data show that the miR-17 inhibitor increased protein expression compared to negative controls. E,F) 293T cells were co-transfected with miR-17 or anti-miR-17 along with psiCheck2 luciferase reporter plasmid containing the 3’-UTR regions of targets ANKRD11, EHD3, EPHA4 and ZBTB4. The data show that the miR-17 inhibitor increased luciferase activity for all respective 3’UTR targets compared to negative controls 48 hours after the transfection, cells were lysed and luciferase signals were measured. * = P<0.05.

Figure 4:. miR-340 and miR-382 regulate cancer pathways:

A-F) Targets identified for miR-340 and miR-382 through PAR-CLIP and The Cancer Genome Atlas (TCGA), were utilized as input for pathway analysis involving Gene Ontology (GO) terms for Biological Processes (BP), Kyoto Encyclopedia of Genes and Genomes (KEGG), and gene set enrichment analysis. The significant pathways are highlighted by red dots.

Figure 5:. miR-340 and miR-382 inhibit cell proliferation, invasion, and neurosphere formation in glioblastoma:

A-D) Glioblastoma cell lines U87 and U251 were transfected with either a scrambled control, or miR-340, or miR-382. Cell counts were performed at various time points, and miR-340 and miR-382 showed decreased cell growth compared to negative controls. E-J) A172, U87, and U251 cells were transfected with mimic miR-340 and miR-382 and invasion assays were executed. Invaded Images from 5–10 random fields were captured and analyzed using ImageJ software for quantification. miR-340 and miR-382 decreased invasion compared to negative controls K-N) Six-well plates were pre-coated with poly-ornithine, and glioblastoma stem cell lines GSC-28 and GSC-34 were plated and transfected with either scrambled control miRNA, miR-340, or miR-382. Neurosphere images were taken from five distinct fields 72 hours post-transfection and categorized into large, medium, and small using ImageJ software, with quantifications presented in (M, N). miR-340 and miR-382 decreased neurosphere formation compared to negative controls. Data represent mean ± SEM from three independent experiments. * = P<0.05.

Figure 6:. Effects of miR-17 inhibition on cell proliferation, invasion, and neurosphere formation in glioblastoma:

A-C) A172, U87, and U251 were transfected with either scrambled control miRNA or a miR-17 inhibitor. Cells were counted 48 hours post-transfection using trypan blue exclusion at various intervals to assess viability. D-I) A172, U87, and U251 cells were transfected with control or miR-17 inhibitor. Invasion assay was carried out and the invaded cells were stained with crystal violet, and images were captured and analyzed using ImageJ software for quantification. J-M) Glioma stem cell lines GSC-28 and GSC-34, plated on poly-ornithine-coated 6-well plates, were transfected with either a scrambled control miRNA or an miR-17 inhibitor. Neurospheres were imaged 72 hours post-transfection in the neurobasal complete growth medium. Images from five different microscopic fields were taken, and neurosphere sizes were categorized into large, medium, and small for quantification using ImageJ software. Data are presented as mean ± SEM from three independent experiments. * = P<0.05.

Figure 7:. miR-340, miR-382 and miR-17 regulate in vivo tumor growth:

A) Schematic overview of the experimental design for tumor implantation and timeline for MRI imaging to assess tumor volume. Each group comprised seven mice for surgery. B-E) U87 cells were transfected with scrambled negative miRNA, miR-340, miR-382, or an inhibitor of miR-17. 48 hours post-transfection, cells were implanted intracranially into the striata of 5–6-week-old immunodeficient mice. Mice were monitored for 3–4 weeks, and MRI imaging was performed to evaluate tumor generation. Representative MRIs and quantification showing that miR-340, miR-382 and anti-miR-17 inhibitor showed reduction in tumor volume compared to Scramble controls. * = P<0.05.

Figure 8:. Inhibition of in vivo glioma growth through MRI-guided (MRIg) Focused Ultrasound, Microbubbles and Brain-Penetrating Nanoparticles (FUS-MB-BPN) delivery of miR-340:

A) A schematic representation of FUS-MB-BPN-mediated miRNA delivery into mice. B) Overall experimental plan and timeline for FUS-MB-BPN. C) Tumors in mice were sonicated pre and post FUS-MB-BPN, and images were captured to validate blood-brain barrier opening. MB were employed to facilitate blood-brain barrier opening. Arrows indicated the tumor location and show dispersion of the contrast demonstrating successful opening of the blood-brain barrier. D,E) Microbubbles were injected through the mice’s tail vein, and MRIgFUS was conducted. Upon confirmation of blood-brain barrier opening, BPN conjugated with either scrambled or miR-340 were injected through the mice’s tail vein. The mice were imaged using MRI, and tumor volume shows significant reduction upon delivery of miR-340 compared to Scramble controls at day 15. F) Kaplan Meir survival curve showing miR-340 significantly prolonged survival compared to scrambled control mice. n = 7 mice per group. *=P < 0.05 for miR-340-treated vs. scramble control.