Patient's Derived Organoids for Drug Screening in Glioblastoma

Development and Characterization of Patient's Derived Organoids as a Platform for the Screening of Novel Therapeutic Treatments for Glioblastoma Multiforme

The study will enroll patients suffering from glioblastoma, a malignant brain tumor. Intervention is intended as a laboratory intervention and not as a clinical intervention. In fact, tumor removed from patients' brains will be sent to a dedicated laboratory to obtain an "avatar" of the tumor, named patient-derived organoid (PDO). A number of experimental antitumor approaches will be studied on PDOs. Results of these experiments will be correlated to the prognosis of patients.

Patient's Derived Organoids (PDOs) from tumour surgical biopsies are an innovative tool to test the response of individual patients to specific therapeutic strategies. Organoids are three dimensional (3D) structures made of organ-specific and self-organizing cells which can be maintained and propagated in culture. To date, PDOs have been established from a great number of cancer types, including prostate, ovarian and breast cancers. Glioblastoma PDOs have also been produced, and they were shown to maintain the characteristics of their parent tumours, both at mutational level and in terms of gene expression profiles and cellular heterogeneity. Since they recapitulate the characteristics of the original tumour better than GBM cell lines, they represent an important advancement for personalized medicine approaches. Thus, glioblastoma PDOs represent useful pre-clinical models for drug screening, CAR-T cell testing and for the generation of brain orthotopic xenografts in model models . In this perspective, the PDOs offer an opportunity to better characterize the molecular heterogeneity of glioblastoma patients and to test new therapeutic strategies in a context that mimic parent tumour genetic properties.

Immunotherapy is emerging as a powerful anticancer approach in some cancer types. Immunotherapy exploits the ability of the immune system to recognise non self-antigens to target and destroy cancer cells. Immune checkpoints inhibitors (e.g. anti-PD-1 and anti-CTLA-4 monoclonal antibodies) were shown effective in tumours exhibiting a high mutational burden, such as melanoma. Unfortunately, glioblastoma has a low mutational burden, resulting in a small amount of neoantigens. Moreover, glioblastoma is highly heterogeneous, meaning that not all the patients produce the same antigens. Thus, higher benefits could be achieved by developing immunotherapies that target multiple neoantigens and by combining neoantigen recognition strategies with immune checkpoint blockade inhibitors. In this perspective, epigenetic regulation to activate the transcription of normally silent transposable elements (TEs) in glioblastoma by DNA demethylating agents can enhance the production of neoantigens and trigger a specific immune response.

Transcription of TEs is low or absent in most adult cells, while it is more active during embryonic development, in stem cells and, intriguingly, in tumors. TEs de-repression in tumors occurs through multiple epigenetic changes to TE loci, including DNA demethylation and histone deacetylation. Both epigenetic changes can be associated with oncogenesis, resulting in different levels of epigenetic de-regulation. TEs overexpression in tumors compared with healthy tissue has prompted the search for anti-TE T cell responses in cancer. Proteogenomic approaches have identified tumor-specific, non-canonical open reading frames (ORFs) that encode peptides presented by human leukocyte antigen (HLA)-I molecules on tumour cells. Most of the identified peptides derived from non-coding genomic regions. Interestingly some of these potential tumor-specific antigens are found in multiple patients and can induce immune responses in vitro or in mouse models. The investigators recently characterized a long non-coding RNA (lncRNA) in the antisense direction of SOD1 gene locus (SOD1-DT), that includes several transposable elements. Some of these (LTR and Alu) contain ORFs and could potentially encode different epitopes. By in silico translation of these elements, the investigators identified peptides corresponding to epitopes already tested as GBM-specific targets for cancer immunotherapies. However, the DNA sequence of these transposable elements is highly methylated in the nervous tissue and in the U87 GBM cell line (data from Genome Browser). The investigators will focus on the study of the TEs belonging to the LTR12C family, because they have been shown to act as enhancer-like and promoter-like elements, shaping the transcriptomics landscape in a tissue-specific manner. It has been already demonstrated that treatments with DNMTi and HDACi do not alter the expression of canonical genes but induce de novo transcription of LTRs, which in turn drive the expression of specific genes. In addition to producing the epitope, by activating specific LTRs, it is therefore possible to activate the genes connected to them. Notably, LTR12C was identified as regulator of proapoptotic genes, such as TP63 and TNFRSF10B. Thus, the proposed strategy could represent a generally applicable means to produce proapoptotic genes and immunogenic epitopes in a controlled manner, ensuring a very specific outcome.

Another potential source of neoepitopes is defective splicing. Splicing is a fundamental step in pre-messenger RNA (mRNA) maturation operated by a large macromolecular machinery named the spliceosome. The spliceosome removes the introns and ligates the flanking exons of the pre-mRNAs, yielding the mature mRNAs. Regulated alternative splicing (AS) of many exons is exploited by cells to generate multiple protein isoforms from a single gene. However, the altered splicing program is often deregulated in cancer cells, generating an actionable vulnerability for tumours, including brain tumours. Profiling of primary and recurrent GBM and non-malignant brain tissues datasets has identified AS events that are differently regulated between in GBM and that could be translated into neoepitopes. These results suggest that splicing modulation could represent a valid therapeutic strategy for glioblastoma. Indeed, inhibition of the arginine methyl transferase PRMT5 in GBM cells dysregulates splicing and leads to incremented intron retention and cell senescence both in vitro and in vivo. Furthermore, PRMT5 has a role in the preservation of GSCs, which are necessary for tumour self-renewal. Recently, it was shown that pharmacologic inhibition of splicing generates splicing-derived immunogenic neoepitopes, which are presented by MHC-I on tumour cells and induce a T cell immune response in vivo. Another potential therapeutic target is the Splicing Factor 3b Subunit 1 (SF3B1), a core component of the splicing machinery that is overexpressed in GBM. Taken together, these results support the rationale of studying the effects of DNA demethylating agents and splicing inhibitors in glioblastoma PDOs and GSCs to identify suitable candidates to develop new therapeutic strategies for this disease.

The above-described approaches will be applied to prospectively enrolled patients undergoing neurosurgery for glioblastoma. Neurosphere cultures and PDO will be established from primary tumor tissue. Drug screening and cell manipulation to induce TE expression and to modulate splicing will be applied. The results of in vitro tests will be correlated with tumor molecular profile, response to treatments and overall patients outcome.

Inclusion Criteria:

• adult patients undergoing resective neurosurgery for glioblastoma, IDH-wildtype

Exclusion Criteria:

• Needle biopsies
• Age <18 years
• Inability to give informed consent
• Brain surgery for tumor disease other than GBM
• Previous neoadjuvant chemotherapy or radiotherapy