Bystander killing by anti-PTPRZ1 CAR-T cells using additional tumor cell lines and scFv

Cytotoxic T cell killing is executed at the immunological synapse, whose nanoscale organization underlies function but remains difficult to resolve in native states. Here, we apply cryo-expansion microscopy (cryo-ExM) to visualize the near-native three-dimensional architecture of human T cell synapses and cytotoxic organelles. Cryo-ExM preserves actin, microtubules, membranes, and fine membrane protrusions with high fidelity, enabling volumetric quantification of synapse morphogenesis. We identify an adhesion-dependent, dome-like membrane architecture beneath activated T cells that collapses upon ICAM-1 engagement, linking synapse topology to adhesive cues. Cryo-ExM further resolves intact lytic granules in primary human CD4 and CD8 T cells, revealing single-core and multi-core ultrastructures, spatial organization, and perforin and granzyme loading. Using tissue-adapted expansion microscopy, we map cytotoxic granule content in tumor-infiltrating T cells in FFPE human brain tumors. Together, these data establish a near-native structural framework for human T cell cytotoxicity and an imaging workflow bridging cell models and clinical tissues.

PI3K inhibitors (PI3Ki) have shown promise in some hematological cancers, but further development has been hampered by reports of serious immune‐related adverse effects. Thus, identification of effective PI3Ki lacking these adverse effects is desirable. Here, we evaluated the in vitro effects of the investigational PI3Ki roginolisib (IOA‐244) and the approved PI3Ki idelalisib on immune cells and leukemic cells. Roginolisib inhibited chronic lymphocytic leukemia cell signaling and viability in a manner comparable to idelalisib. Both drugs specifically inhibited PI3K‐signaling in T cells, validating their on‐target effects. Both idelalisib and roginolisib reduced regulatory T‐cell frequency in a concentration‐dependent manner, with idelalisib demonstrating greater potency. Both inhibitors also reduced T‐cell activation and proliferation, but to differing extents. However, only idelalisib induced a pronounced impairment of CD8+ T‐cell cytotoxic function. Furthermore, idelalisib treatment promoted differentiation of conventional CD4+ T cells into Th1, Th2, and Th17 subsets—a response not observed with roginolisib. In summary, roginolisib functions as an effective PI3K inhibitor on leukemic cells while preserving T‐cell functions, posing it as an alternative to current PI3K inhibitors.

Abstract Phosphoinositide 3-kinase (PI3K) is a key target in cancer therapy, but first-generation PI3K inhibitors were associated with toxicities limiting their clinical use. Roginolisib, a next-generation PI3Kdelta inhibitor, is currently being investigated across multiple cancer types. In contrast to previous PI3K inhibitors, roginolisib has shown a well-tolerated profile. We hypothesize that this improved safety is in part due to high selectivity, driven by its structural properties. Structural and biophysical analyses—including X-ray crystallography, molecular dynamics simulations, and hydrogen-deuterium exchange mass spectrometry—demonstrate that roginolisib, unlike the first-generation inhibitor idelalisib, stabilizes the catalytic C-terminal helix (k-alpha12) of PI3Kdelta. This stabilization is specific for PI3Kdelta and locks the enzyme in an inactive conformation, resulting in potent and sustained inhibition of PI3Kdelta activity. This inhibition profile is observed in tumor samples from patients with chronic lymphocytic leukemia (CLL) and in lymphoma cell lines. Although roginolisib and idelalisib exhibit similar potency in CLL cells, their effects on immune cells differ. Notably, high concentrations of idelalisib impair the cytotoxic activity of CD8+ T cells and promote the differentiation of CD4+ T cells towards Th17 and Th2 subsets. By contrast, roginolisib preserves CD8+ T-cell function and does not alter CD4+ T-cell differentiation. This mechanism of inhibition, likely based on conformational stabilization of the inactive kinase, is novel for PI3K small-molecule inhibitors. Hence, our findings may enable the development of more effective and better-tolerated PI3K inhibitors. Citation Format: Giusy Di Conza, Oscar Vadas, Simon Tiede, Elise Solli, Maria Chaouki, Laura Tesmer, Shanlin Rao, Remy Visentin, Mathias Wenes, Denis Migliorini, Sigrid S. Skånland, Anne Quillet-Mary, Loïc Ysebaert, Julie Guillermet-Guibert, Alessio Bevilacqua, Lars van der Veen, Michael Lahn, Kjetil Tasken, Gerhard Hummer. Roginolisib stabilizes inactive PI3Kdelta to deliver specific target blockade with reduced immune toxicity [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 5763.

Primer pairs for CAR-T cell phenotyping by qPCR

CAR-T-cell-associated on-target off-tumor (OTOT) toxicity represents a major safety concern, as recognition of target antigens on healthy tissues can trigger severe and potentially life-threatening complications. Predicting OTOT toxicity remains a challenge because current preclinical models fail to capture the complexity of native human tissues. Here, we developed a human organotypic tissue platform that enables functional assessment of CAR-T-cell activity in intact human tissues across organ-specific and inflammatory contexts. Using a panel of clinically relevant CAR-T-cell products with known OTOT toxicities, we demonstrate that the platform faithfully recapitulates clinically observed tissue-specific toxicity profiles. CAR-T cells targeting EGFR, HER2, and mesothelin induced inflammatory and cytotoxic responses in healthy human lung tissue, whereas CD19 CAR-T cells remained inactive. We further show that OTOT toxicity cannot be reliably predicted from antigen abundance alone but instead results from the integration of multiple target-dependent determinants, including CAR affinity, inflammatory context, antigen accessibility, and effector-cell dose. The platform also enables quantitative assessment of inflammatory and cytotoxic responses and supports evaluation of pharmacological and CAR design-based strategies to mitigate toxicity. Together, this work establishes the first human organotypic platform for functional modeling of CAR-T-cell-associated OTOT toxicity, providing a clinically relevant framework for preclinical safety evaluation and the rational development of safer engineered cell therapies.

Glioblastoma (GBM) is characterized by a profoundly immunosuppressive tumor microenvironment (TME) that constrains the efficacy of chimeric antigen receptor (CAR)-T cell therapy. Here, we show that surgical resection in both male mice and human GBM ex vivo induces a rapid and sustained remodeling of the TME, marked by upregulation of TREM2 in myeloid cells followed by emergence of T cell exhaustion-like phenotypes. In male mice, targeting TREM2 reshapes the perioperative TME and potentiates tumor antigen-specific CAR-T cell responses, improving intratumoral persistence, proliferation, and effector differentiation, and resulting in enhanced survival. In parallel, we identify the timing of CAR-T cell administration as a critical determinant of therapeutic outcome, with neoadjuvant outperforming adjuvant treatment by preserving CAR-T cell effector function in mice. These findings establish perioperative myeloid cell remodeling and treatment timing as key determinants of CAR-T cell efficacy in GBM. Glioblastoma (GBM) is characterized by an immunosuppressive tumor microenvironment (TME). Here the authors report that surgical resection induces TREM2 associated immunosuppressive remodeling of the TME and test CAR-T based strategies to reshape the perioperative GBM TME.

Anti-PTPRZ1 CAR-T cell effector function after incubation with the Ge1302_PTPRZ1-KI cells

In vivo anti-tumor activity of anti-PTPRZ1 CAR-T cells

Recurrent glioblastoma has no established standard of care, and survival remains limited to a few months. Recent clinical evidence suggests that the intracranial delivery of CAR-T cells is feasible and shows signs of antitumor activity. In the phase I trial NCT02208362, anti-IL13Rα2 CAR-T cells were tested in 65 patients, yielding a 50% disease-control rate and 23% one-year survival. Two phase I studies of multi-target CAR-T products (NCT05168423, the ongoing NCT05660369) similarly showed bioactivity and early but transient radiological responses. Importantly, the intracranial route enabled rich translational research: serial biopsies and cerebrospinal fluid sampling allowed assessment of CAR-T cells function at the tumor site and the identification of candidate predictive biomarkers. In this review, we compare the designs and findings of these three studies and synthesize key lessons. We also outline future perspectives to improve the efficacy of CAR-T cell therapy in solid tumors, with a focus on glioblastoma.

Use of Jurkat triple reporter cells to assess CAR tonic signaling and specificity

Bystander killing by 471_28z CAR-T cells depends on soluble mediators and does not affect macrophages

Expression of effector/memory and activation/exhaustion markers by CAR-T cells in pre- vs. post-thawing samples

PTPRZ1 expression at the mRNA level according to different GBM clinicopathologic features

BACKGROUND:Glioblastoma (GBM) is an aggressive brain tumor associated with poor outcome and limited treatment options. Chimeric antigen receptor (CAR) T cells targeting cell surface antigens were shown to induce tumor regression in patients with GBM, although efficacy was transient. To broaden the range of tumor-restricted antigens, we developed CAR T cells targeting Tenascin-C (TNC), a secreted extracellular matrix protein that is overexpressed in GBM and plays a critical role in tumor progression. METHODS:Second-generation CAR T cells were engineered to target the alternatively spliced fibronectin type III (FNIII)-D domain of TNC using a single-chain variable fragment isolated from the R6N antibody and coupled to a CD28 costimulatory domain. TNC-CAR T cells were evaluated in vitro for antigen specificity, activation, and cell proliferation using TNC-expressing patient-derived GBM cell lines cultured as adherent cells or as neurospheres. Reactivity toward purified TNC protein, tumor supernatant, and ex vivo patient tumor samples was also assessed. Cytotoxic CAR T-cell activity was tested against TNC-positive and TNC-negative GBM cell lines, including bystander effects mediated by secreted TNC. In vivo efficacy and safety were determined in NOD scid gamma mice bearing patient-derived GBM tumors. RESULTS:TNC-CAR T cells demonstrated activation when exposed to TNC-positive GBM cells, cell-derived supernatants, or purified TNC protein. They exhibited potent cytotoxicity against TNC-expressing, GBM-derived adherent cells and neurospheres, and induced bystander killing of TNC-negative cells in the presence of either TNC-secreting cells or purified TNC. In vivo, TNC-CAR T cells efficiently infiltrated tumors, triggered cancer cell apoptosis, and significantly extended survival of mice bearing patient-derived GBM, with no evidence of off-tumor toxicity. Notably, TNC-CAR T cells were activated exclusively in the presence of tumor samples and showed no reactivity toward patient-derived non-tumor tissues. CONCLUSIONS:Targeting the alternatively spliced FNIII-D domain of TNC with CAR T cells offers a promising therapeutic approach for GBM. TNC-CAR T cells demonstrated specific tumor recognition, robust antitumor activity and the ability to induce bystander effects mediated by secreted TNC. Their efficacy in preclinical models, combined with a favorable safety profile, underscores their potential for clinical translation.

Immunotherapy is facing a revolution with the advent of immune cell engineering. Chimeric antigen receptor (CAR)-T cell therapy has shown unprecedented efficacy in B cell malignancies and is now being evaluated in other disease areas. Viral transduction is the most common method for immune cell genetic engineering, but presents important limitations, such as high reagent costs and regulatory concerns due to mutagenesis risk. One prevailing non-viral gene delivery strategy relies on the electroporation of non-integrating RNA. However, most modern electroporation technologies also require high reagent costs and rely on the use of proprietary software and transfection buffers. Nanoparticle-sensitized optoporation represents an alternative method for transient permeabilization of cells. Here, we introduce magnetic bead-sensitized optoporation, in which commercially available superparamagnetic beads coupled with anti-human CD3 and CD28 antibodies are used as photosensitizers for efficient genetic cargo delivery into human primary T cells and other immune cells. We show that magnetic bead-sensitized optoporation of human T cells generates functional mRNA-based CAR-T cells without affecting T cell product memory phenotype or activation potential. Importantly, optoporated T cells exhibited a greater proliferation capacity relative to electroporated T cells. In conclusion, our findings suggest that magnetic bead-sensitized optoporation holds promise as mRNA delivery strategy for immune cell therapy.

Immunotherapy offers a promising alternative approach for patients with glioma. In this study, we conducted a phase I/II clinical trial to evaluate the safety and tolerability of a multipeptide therapeutic vaccine (IMA950) comprising nine GBM-associated HLA-A2-restricted and two tumor-associated MHC class II-restricted peptides administered in combination with the adjuvant Poly-ICLC with or without pembrolizumab in patients with relapsing GBM or grade IV astrocytoma. Patients received up to nine subcutaneous injections of IMA950/Poly-ICLC without (Arm 1) or with (Aim 2) pembrolizumab (200 mg IV every 3 weeks). Primary endpoints were safety and tolerability; secondary endpoints included overall survival (OS), progression-free survival (PFS) as well as analysis of vaccine-induced T cell responses in the peripheral blood and tumor and treatment-induced changes in the tumor microenvironment (TME). This clinical trial showed that the IMA950/Poly-ICLC vaccine administered with or without pembrolizumab was safe and well tolerated. No differences in OS and PFS were observed between the two arms. Vaccine-induced CD8 and CD4 T cell responses were detected in peripheral blood samples collected before, during and after vaccination in some individuals but were of low magnitude. Tumor samples collected before treatment and after the 4th vaccine were submitted to single-nuclei (n=11) and bulk (n=19) RNA sequencing for transcriptomic analysis. Expression of the vaccine antigen in tumors was heterogeneous across patients, and no significant loss of antigen expression was observed following treatment in either arm. Single-nuclei RNA sequencing revealed that tumor cells predominantly exhibited astrocyte-like, oligodendrocyte progenitor-like and neural progenitor-like states. Exploratory analysis of the TME revealed the presence of SPP1⁺ and M2-like macrophages. Additionally, a distinct cluster of CD8⁺ IL7Rhi T cells was identified in both pre- and post-treatment samples across both treatment arms. Altogether, the IMA950/Poly-ICLC vaccine administered with or without pembrolizumab is safe but did not significantly impact patient survival.

Glioblastoma (GBM) is an aggressive primary brain tumor with dismal clinical prognosis and resistance to current therapies. GBM progression is facilitated by the tumor microenvironment (TME), with an immune infiltrate dominated by tumor-associated microglia/macrophages (TAMs) and regulatory T cells (Tregs). The TME is also characterized by hypoxia and the expression of hypoxia-inducible factors (HIFs), with HIF-2α emerging as a potential regulator of tumor progression. However, its role in GBM immunosuppression remains unknown. Here, we investigate HIF-2α and the use of the HIF-2α inhibitor PT2385 to modulate the TME in the immunocompetent GL261 mouse GBM model. PT2385 administration in vivo decreased tumor volume and prolonged survival of tumor-bearing mice, without affecting GL261 viability in vitro. Notably, HIF-2α inhibition alleviated the immunosuppressive TME and synergized with immune checkpoint blockade (ICB) using αPD-1 and αTIM-3 antibodies to promote long-term survival. Comprehensive analysis of the immune infiltrate through single-cell RNA sequencing and flow cytometry revealed that combining PT2385 with ICB reduced numbers of pro-tumoral macrophages and Tregs while increasing numbers of microglia, with a corresponding transcriptional modulation towards an anti-tumoral profile of these TAMs. In vitro, deletion of HIF-2α in microglia impeded their polarization towards a pro-tumoral M2-like profile, and its inhibition impaired Treg migration. Our results show that targeting HIF-2α can switch an immunosuppressive TME towards one that favors a robust and sustained response to ICB based immunotherapy. These findings establish that clinically relevant HIF-2α inhibitors should be explored not only in malignancies with defects in the HIF-2α axis, but also in those exhibiting an immunosuppressive TME that limits immunotherapy responsiveness.