AUTHOR=Marques Rui Ferreira , Moreno Daniel Antunes , da Silva Luciane , Leal Leticia Ferro , de Paula Flávia Escremim , Santana Iara , Teixeira Gustavo , Saggioro Fabiano , Neder Luciano , Junior Carlos Almeida , Mançano Bruna , Reis Rui Manuel 

TITLE=Digital expression profile of immune checkpoint genes in medulloblastomas identifies CD24 and CD276 as putative immunotherapy targets

JOURNAL=Frontiers in Immunology

VOLUME=14

YEAR=2023

URL=https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2023.1062856

DOI=10.3389/fimmu.2023.1062856

ISSN=1664-3224

ABSTRACT=<sec><title>Introduction</title><p>Medulloblastoma is the most common and lethal pediatric malignant brain tumor. It comprises four main molecular subgroups: WNT-activated, SHH-activated, Group 3, and Group 4. Medulloblastoma treatment is surgical resection, craniospinal radiation, and chemotherapy. However, many patients do not respond to therapy, and most suffer severe side effects. Cancer immunotherapy targeting immune checkpoints (IC) (PD-1, PD-L1, and CTLA4) has been getting disappointing outcomes in brain tumors. Nevertheless, other less explored immune checkpoints may be promising candidates for medulloblastoma therapy.</p></sec><sec><title>Objectives</title><p>In the present study, we aimed to characterize the expression profile of 19 immune checkpoints in medulloblastoma.</p></sec><sec><title>Methods</title><p>We analyzed 88 formalin-fixed paraffin-embedded medulloblastomas previously classified for each molecular subgroup and three non-tumoral brain tissue. mRNA levels of 19 immune checkpoint-related genes were quantified using the nCounter (PanCancer Immune Profiling Panel) assay. Further <italic>in silico</italic> analysis was performed in two larger public microarray datasets, one of which enabled comparisons between tumoral and non-tumoral tissues. Immunohistochemistry of PD-L1 was performed in a subset of cases. Microsatellite instability was also molecularly analyzed.</p></sec><sec><title>Results</title><p>We observed an absence of expression of the canonic ICs, namely <italic>PDCD1</italic> (PD-1), <italic>CD274</italic> (PD-L1), and <italic>CTLA4</italic>, as well as <italic>CD80, CD86, BTLA, IDO1, CD48, TNFSF14, CD160, CEACAM1</italic>, and <italic>CD244.</italic> PD-L1 protein expression was also practically absent. We found higher mRNA levels of <italic>CD24, CD47</italic>, <italic>CD276</italic> (B7-H3), and <italic>PVR</italic>, and lower mRNA levels of <italic>HAVCR2</italic>, <italic>LAG3</italic>, and <italic>TIGIT</italic> genes, with significant differences across the four molecular subgroups. Compared to the non-tumor tissues, the expression levels of <italic>CD276</italic> in all subgroups and <italic>CD24</italic> in SHH, Group 3, and Group 4 subgroups are significantly higher. The <italic>in silico</italic> analysis confirmed the expression profile found in the Brazilian cohort, including the lower/absent expression of the canonic ICs. Moreover, it confirmed the overexpression of <italic>CD24</italic> and <italic>CD276</italic> in medulloblastomas compared with the non-tumor tissue. Additionally, <italic>CD276</italic> and CD24 high levels were associated with worse survival.</p></sec><sec><title>Conclusion</title><p>These results highlight the low or absence of mRNA levels of the canonic targetable ICs in medulloblastomas. Importantly, the analysis revealed overexpression of <italic>CD24</italic> and <italic>CD276</italic>, which can constitute prognostic biomarkers and attractive immunotherapy targets for medulloblastomas.</p></sec>