Supplementary Figure S1 describes the establishment and scRNA-seq profiling of GCOs, related to Figure 1.

Supplementary Figure S5 describes extracellular vesicle mediated intercellular transfer in GCOs, related to Figure 5.

Supplementary Figure S4 describes widespread transfer of endogenous malignant transcripts in GCOs, related to Figure 4.

Supplementary Table S3 shows gene sets defining malignant programs in human GCOs.

Supplementary Figure S2 describes malignant cell states in GCOs, related to Figure 2.

Text file shows the GFP sequence from pLV-CMV-GFP used to align GFP reads from scRNA-seq data.

Supplementary Table S4 shows gene sets defining malignant (mouse) programs in mouse-human (m005) mixed-species GCOs.

Supplementary Table S2 shows gene sets defining non-malignant cell types in GCOs.

Supplementary Figure S3 describes GFP transfer process in GCOs, related to Figure 3.

Supplementary Table S1 shows the timeline of derivation, establishment, and sequencing of GCOs.

Glioblastoma (GBM) is characterized by heterogeneous malignant cells that are functionally integrated within the neuroglial microenvironment. In this study, we model this ecosystem by growing GBM into long-term cultured human cortical organoids that contain the major neuroglial cell types found in the cerebral cortex. Single-cell RNA sequencing analysis suggests that, compared with matched gliomasphere models, GBM cortical organoids more faithfully recapitulate the diversity and expression programs of malignant cell states found in patient tumors. Additionally, we observe widespread transfer of GBM transcripts and GFP to nonmalignant cells in the organoids. Mechanistically, this transfer involves extracellular vesicles and is biased toward defined GBM cell states and astroglia cell types. These results extend previous GBM organoid modeling efforts and suggest widespread intercellular transfer in the GBM neuroglial microenvironment.Significance: Models that recapitulate intercellular communications in GBM are limited. In this study, we leverage GBM cortical organoids to characterize widespread mRNA and GFP transfer from malignant to nonmalignant cells in the GBM neuroglial microenvironment. This transfer involves extracellular vesicles, may contribute to reprogramming the microenvironment, and may extend to other cancer types.See related commentary by Shakya et al., p. 261

Tuberous sclerosis complex (TSC) is a dominantly inherited disease in which most individuals are born with one defective allele encoding for either hamartin (TSC1) or tuberin (TSC2), with a somatic loss of the other allele leading to abnormal neurodevelopment and upregulation of cell growth in susceptible tissues. Ninety percent of affected individuals have brain involvement, including epilepsy, cognitive impairment, autism, and/or sleep disorders. In the stochastic, cerebral mouse model of Tsc1, loss of function of hamartin is induced in the CNS by injection of an adeno-associated virus (AAV) vector encoding Cre recombinase into the cerebral ventricles of homozygous Tsc1flox/flox mice at birth. In the brain, Tsc1 loss leads to increased proliferation of subventricular zone cells, disrupted neuronal migration and cortical cytoarchitecture, dysmyelination, and microglia-mediated inflammation, ultimately resulting in early mortality. Systemic administration of an AAV9 vector encoding human hamartin at postnatal day 21 significantly ameliorated these abnormalities at 3 and 6 weeks post-injection and markedly extended survival in this TSC1 mouse model. This work reveals the ability of hamartin replacement therapy to reverse some of the brain abnormalities caused by its loss in different cell types and provides support for the potential use of gene replacement therapy in the treatment of TSC1 patients.

Extracellular vesicles (EVs) are mediators of intercellular communication through the transfer of nucleic acids, lipids and proteins between cells. This property makes bioengineered EVs promising therapeutic vectors. However, it remains challenging to isolate EVs with a therapeutic payload due to the heterogeneous nature of cargo loading into EVs. In this study, enrichment of EVs with a desired cargo was possible through engineering of the hallmark CD63 transmembrane protein. E-NoMi refers to engineered CD63 with mCherry on the inside of the EV membrane and a tag (3xFLAG) exposed on the outside of the EV membrane. To facilitate EV loading during biogenesis, cargo proteins, such as EGFP, Cre recombinase and the CRISPR-Cas nuclease (SaCas9), were fused to a nanobody (Nb) protein with a high affinity for mCherry. FLAG-tag-based immunocapture from cell conditioned media allowed selection of cargo-loaded E-NoMi-EVs, and tobacco etch virus (TEV) protease cleavage sites were used to remove the 3xFLAG-tag from the surface of E-NoMi-EVs after capture. For functional payload delivery to recipient cells, the vesicular stomatitis virus G (VSV-G) fusogenic protein was incorporated into E-NoMi-EVs to form fusogenic EV-based vectors (EVVs) and proved to be 10-fold more effective at cargo delivery than EVs generated by size-exclusion chromatography. Functional delivery of cargo with E-NoMi-EVVs was validated in two mouse brain models in vivo.

Glioblastoma (GBM) remains one of the most aggressive brain tumors, with limited effective therapies including temozolomide (TMZ) and resistance to immunotherapy. The adenosine A2A receptor (A2AR) is a key mediator of immunosuppression in the tumor microenvironment (TME), where its activation dampens T cell function, inhibits pro-inflammatory cytokine production, and supports tumor immune evasion. While A2AR-driven immunosuppression has been well described in several cancer types, its expression pattern, cellular specificity, and therapeutic relevance in GBM remain largely unexplored. In this study, we evaluated the expression and therapeutic potential of A2AR in the GBM TME using a syngeneic mouse model. Immunofluorescence analysis revealed that A2AR was markedly upregulated in the glioma-bearing brain compared to peripheral tissues, while staining of the spleen showed low and predominantly intracellular A2AR expression. This suggests that its upregulation is localized and acquired within the GBM TME. Within the tumor, A2AR co-localized with THY1.2⁺ T cells, PD-1⁺ cells, and IFN-γ–secreting cells, indicating its association with activated and potentially exhausted T cell subsets. A2AR was undetected in TME-associated microglia, monocytes, and astrocytes. CT-2A glioma cells were implanted intracranially into mice, which were assigned to four treatment groups: sham (vehicle), FDA-approved, low-dose temozolomide (TMZ; 5 mg/kg/day, IP), brain-penetrant A2AR antagonist istradefylline (10 mg/kg/day, IP), or a combination of both. Survival analysis showed that istradefylline and TMZ monotherapies each significantly prolonged survival compared to sham (p = 0.0101), with no difference between monotherapies (p = 0.312). The combination therapy significantly improved survival over sham (p = 0.0042), TMZ (p = 0.0029), and istradefylline (p = 0.0027). Our findings support the therapeutic relevance of A2AR in GBM and suggest that combining its inhibition with standard-of-care TMZ may enhance efficacy while reducing systemic toxicity.