Introduction: : Despite success in hematologic malignancies, Chimeric antigen receptor (CAR) T-cell therapy faces significant challenges in solid tumor treatments, including tumor heterogeneity, on-target off-tumor toxicity, immunosuppressive tumor microenvironment, and limited T cell infiltration. Numerous engineering strategies have been proposed to address the challenges; however, most solutions target individual limiting factors, making it difficult to reach full efficacy when multiple inhibitory mechanisms are present concurrently. Recent advances in engineering approaches, such as CRISPR-Cas9-based gene editing, have expanded the potential for solid tumor treatment. However, how these technologies can be used to prime solid tumor and further boost immunotherapy outcomes remain underexplored. In this study, we developed an integrated platform for in vivo tumor gene editing to enhance CAR-T-mediated clearance in triple-negative breast cancer (TNBC; Figure 1A). Using a lentiviral-based particle engineered with receptor-ligand interactions (ENTER), we achieved simultaneous delivery of genetic payloads and CRISPR-Cas9/sgRNA ribonucleoprotein complexes (RNPs) to tumor cells. Following focused ultrasound (FUS) stimulation, the genetic payloads enabled heat shock promoter (HSP)-driven expression of the inducible antigen in a spatiotemporally controlled manner. These modified tumor cells functioned as training centers, activating synthetic Notch (synNotch) CAR-T cells to produce CARs targeting the universally expressed tumor antigen, thereby eliminating neighboring tumor cells. Delivery of Cas9/sgRNA RNPs targeting telomeres induced significant tumor cell death; RNA sequencing and bioinformatic analysis revealed activation of apoptosis, inflammatory pathways, and increased chemokine secretion. Engineering synNotch-CAR-T cells to express the corresponding chemokine receptor enhanced T cell trafficking to the gene-edited tumor site, further improving CAR-T therapeutic efficacy.
Materials and
Methods: : To enable delivery of both genetic payloads and CRISPR-Cas9 RNPs to MDA-MB-231 cells, we engineered the ENTER system to display an antibody single-chain variable fragment (scFv) on the particle membrane targeting endogenous EGFR on tumor cells. CRISPR-Cas9 RNPs were encapsulated within the particles, and the HSP-driven truncated CD19 (tCD19) construct was delivered as the genetic payload (Figure 1B). To verify FUS inducibility, MDA-MB-231 cells receiving the genetic payload were subjected to heat shock (HS), mimicking heat generated by FUS, and subsequently analyzed by flow cytometry to assess surface expression of tCD19. To evaluate the impact of telomere disruption, we assessed cell growth following CRISPR-mediated knockout and extracted total RNA for sequencing. Differential gene expression analysis was performed, and potential CAR-T targets were identified based on both high upregulation and expression abundance. The synNotch-CAR-T cells were activated upon binding of the synNotch receptor to the inducible antigen tCD19, triggering the expression of PD-1-based CARs that target endogenous PD-L1 on tumor cells. To enhance tumor infiltration, CAR-T cells were further armored with a chemokine receptor (Figure 1C). CAR-T engineering was confirmed by flow cytometry, chemokine receptor function was validated using a transwell migration assay, and cytotoxicity was assessed in a 2D co-culture system with MDA-MB-231 cells expressing HSP-driven tCD19, with or without heat shock stimulation. Finally, ENTER-mediated delivery and CAR-T cytotoxicity were evaluated in MDA-MB-231 spheroids, which more closely mimic the 3D physiological tumor environment.
Results, Conclusions, and Discussions:: To demonstrate the efficiency of ENTER delivery, we first validated the successful expression of mCherry as a model payload and the knockout of surface β2-microglobulin (B2M) by CRISPR-Cas9 RNPs (Figure 1D). We then replaced the gene of interest with HSP-driven tCD19 and used sgRNA targeting telomere. Following heat shock, tCD19 was robustly expressed, and telomere knockout significantly impaired cell growth compared to non-targeting (NT) sgRNA/Cas9 RNPs (Figure 1E). Bioinformatic analysis revealed significant upregulation of inflammatory and apoptotic pathways following telomere knockout, along with increased secretion of chemokines CXCL1, CXCL3, and CXCL8. While these chemokines attract innate immune cells, they do not recruit T cells due to the absence of their receptor, CXCR2, on T cells. To address this, we engineered T cells to express CXCR2, enabling migration toward recombinant chemokine gradients (Figure 1F). We further demonstrated that synNotch-CAR-T cells exhibited strong cytotoxicity against MDA-MB-231 cells expressing heat shock–induced tCD19, whereas unmodified T cells or synNotch-CAR-T cells exposed to non–heat-shocked cells showed no significant killing (Figure 1G). We next assessed cytotoxicity in MDA-MB-231 spheroids with HSP-tCD19 expression and either telomere-targeting or control CRISPR-Cas9 RNPs. Only CAR-T cells, not unmodified T cells, effectively killed spheroids. Moreover, spheroids with telomere knockout showed greater size reduction than controls (Figure 1H). We further compared spheroid killing using synNotch-CAR-T cells with or without CXCR2 in the presence or absence of telomere knockout. The killing effect of telomere knockout was more pronounced with CXCR2-expressing CAR-T cells, indicating a synergistic effect between tumor editing and T cell chemokine receptor engineering (Figure 1I). In summary, using a TNBC model, we achieved tumor gene editing via concurrent CRISPR-Cas9–mediated disruption and FUS-controlled antigen expression delivered by the ENTER system. Bioinformatic analysis guided the design of chemokine receptor–armored synNotch-CAR-T cells, resulting in improved tumor control. In vivo studies are ongoing.
Acknowledgements and/or References (Optional):: We thank all the co-authors for their contribution. The full author list will be released during the meeting.
