Dual CAR Stimulation in CD19xCD22 Bicistronic CAR T Cells Results in Increased c-Jun Expression and Superior In Vivo Expansion

Chimeric antigen receptor (CAR) T cell therapy is effective in treating B-cell malignancies with a high rate of complete responses; however due to relapse, only 40-50% of patients achieve long term remissions. Relapses with CD19 antigen loss are a major mechanism of leukemic escape after CD19-directed CAR T cell therapy and downregulation of the CD19 antigen has been associated with disease progression after CAR therapy in non-Hodgkin lymphoma (NHL) patients. Treatment with CD22-directed CAR T cells has been an effective salvage regimen for acute lymphoblastic leukemia (ALL) patients with CD19-negative disease, however downregulation of the antigen allows for leukemic escape in the majority of patients. Strategies to simultaneously target CD19 and CD22 are hypothesized to prevent antigen-modulated relapses after CAR therapy, therefore we constructed a CD19xCD22 bicistronic CAR which is currently in clinical trial at the University of Colorado (NCT05098613). This construct incorporates a CD19 CAR with a CD28 costimulatory domain along with a CD22 CAR with a 4-1BB costimulatory domain. In preclinical models, the combination of CD28 and 4-1BB containing CARs was found to be superior to the use of either costimulatory molecule alone, suggesting a yet to be understood beneficial interaction when this combination is used. We hypothesized that integration of the signals from the two different costimulatory molecules underlies the enhanced efficacy of CD19xCD22 CAR T cells and that understanding the signaling downstream of our bicistronic CAR construct will allow for the generation of more effective CAR T cell therapies in the future.

Signaling analysis was carried out in Jurkats stably transduced with the 19x22 CAR and co-cultured with Nalm6 cell lines expressing variable combinations of the CD19 and CD22 antigens (19+/22+, 19-/22-, 19+/22-, 19-/22+). Signaling was assessed by whole cell or nuclear flow cytometry for phosphorylation of signaling molecules or translocation of transcription factors, respectively. In vivo expansion, persistence, memory formation, and exhaustion were assessed utilizing NSG mice engrafted with Nalm6 clones expressing various antigen combinations and treated with 19x22 bicistronic CAR T cells. Mice were monitored for leukemia progression using bioluminescent imaging and CAR T cells were evaluated at peak expansion (Day 7) and at a persistent time point (Day 28) for cell number and phenotype by flow cytometry.

19x22 CAR T cells displayed superior in vivo expansion in mice engrafted with 19+/22+ Nalm6 (Figure 1) with a subsequent trend toward better CAR T cell persistence, higher expression of IL7Ra, and lower expression of both PD1 and TIM3 when compared to mice engrafted with Nalm6 expressing either target antigen alone. This correlated to higher levels of ERK phosphorylation and nuclear translocation of the transcription factor c-Jun after stimulation through both CD19 and CD22 CARs relative to either alone (Figure 2). In contrast, nuclear translocation of NFAT was primarily driven by stimulation through the CD19 CAR alone. Signaling through other pathways, such as NF-κB, p38 MAPK, and PLCγ, were not significantly different upon stimulation through CD19 and CD22 CARs versus the CD19 CAR alone.

Expression of CD19-28z and CD22-BBz CARs on the same T cell in a bicistronic construct led to signal integration reflected in altered expression of intracellular phosphoproteins. Interestingly, the transcription factor c-Jun, whose overexpression has been previously associated with decreased exhaustion and improved persistence (Lynn RC et al., Nature 2019), was upregulated after stimulation of two CARs with different costimulatory domains. In vivo, CAR T cells activated through both CD19 and CD22 CARs demonstrated superior expansion and persistence with increased levels of IL7Ra, suggesting an enhanced capacity for self-renewal. They also displayed a lower expression of the exhaustion markers PD1 and TIM3. Better understanding of how signal integration in a bicistronic construct relates to CAR T cell function may have future implications in helping to predict clinical outcomes after CAR manufacturing, and the ability to harness knowledge of signal integration has the potential to aid in the design of future CAR T cell products.

No relevant conflicts of interest to declare.