Blinatumomab-Induced T Cell Activation at Single Cell Transcriptome Resolution

Introduction Blinatumomab, a bi-specific T cell engager (BiTE®) combining the VH and VL domains of two antibodies against human CD19 and CD3, has been approved by U.S. Food and Drug Administration (FDA) for the treatment of patients with relapsed or refractory B precursor ALL (r/r B-ALL) for its clinical benefit demonstrated in different clinical trials. Clinical trials have also shown that there are still patients refractory to blinatumomab. It is thus important to understand the resistance mechanisms. Blinatumomab connects patients' autologous T cells and target cells to form immunologic synapse which potently triggers the activation signaling cascades in T cells and guides T cells to recognize and induce perforin/granzyme-mediated lysis of CD19+ B-ALL cells. Previous studies showed blinatumomab-mediated cytotoxicity involves different T cell subpopulations. But response of each T cell subpopulation to blinatumomab treatment remained largely unknown.

Methods and Results In this study, we used 10X Genomics based single cell RNA sequencing to analyze the transcriptome of single T cells before and after blinatumomab treatment. First, ex vivo blinatumomab cytotoxicity model was established, in which healthy PBMCs were used as effector cells and cocultured with target cells (RS4;11 cells or SUP-B15 cells) at an effector-to-target cell ratio of 10:1 with 0.1 ng/mL blinatumomab. Then, PBMCs and BMMCs from 2 B-ALL patients were cultured with 10 ng/mL blinatumomab. Cells from both ex vivo model and patient samples were sequenced using 10X Genomics platform. In total, transcriptome of 17920 single T cells from the ex vivo model and 2271 single T cells from patient sample were analyzed.

Based on T cell trajectory analysis, we identified four distinct populations of blinatumomab-activated T cells, which were derived from CD8+ effector memory T (T_EM) cells, CD4+ central memory (T_CM) cells, naïve T cells and Tregs, respectively. The differentially expressed genes in activated clusters were analyzed to reflect T cell activation mechanisms. The result showed blinatumomab induced the upregulation of aerobic glycolysis pathway (PKM, PGAM1, ENO1, GAPDH and LDHA), cytoskeleton dynamics pathway (ACTD1, ACTB, NME1 and TUBA1B), IFN-responsive pathway (GBP1, PSME2, WARS, CXCL10 and STAT1), and the upregulation of well-known immune-related genes (TNFRSF4, TNFRSF18, LAG3, CD69, IL2RA, MIR155HG, BATF, SH2D2A, LTA, NFKBIA and NDFIP2). We found blinatumomab-activated CD8+ T_EM cells showed stronger cytotoxic capability than other activated populations with specific production of cytotoxic factors (PRF1, IFNG and FASLG) and cytokines (CCL2, CCL3, CCL3L1, CCL4, CCL4L2, CCL8, XCL1, XCL2, TNFSF9 and TNFSF14). Last, differential gene expression analysis revealed that co-stimulatory (TNFRSF4,TNFRSF9 and TNFRSF18) and co-inhibitory receptors (LAG3 and TIGIT) were similarly up-regulated in clusters activated from memory and naïve T cells, indicating ligand dependent T cell functional outcomes induced by blinatumomab.

Conclusion In summary, we used single cell sequencing to map the blinatumomab-mediated T cell activation state transition and reveal the molecular changes in different T cell subpopulations. Memory T cells, naïve T cells and Tregs were identified functional populations after blinatumomab treatment. CD8+ T_EM accounted for the majority of blinatumomab-induced cytotoxicity. Furthermore, T cell co-regulatory receptors were identified as potential targets accountable for blinatumomab sensitivity or resistance mechanisms. The study demonstrated that the collected cellular transcriptional profiles can serve as resource to explore novel strategies to enhance the efficacy of blinatumomab.