Back

Rational redesign of antigen binding domain improves in vivo efficacy of CD22-CAR T cells

DeGolier, K. R.; Walsh, Z. H.; Leach, L.; Brzezinski, C.; Novak, A.; Dimitrov, D.; Kohler, M. E.; Fry, T. J.

2025-03-15 immunology
10.1101/2025.03.13.643183 bioRxiv
Show abstract

Chimeric antigen receptor (CAR) T cells targeted to CD19 are an effective therapy for B-lineage malignancies. However, about half of patients relapse and this therapeutic, often with antigen-negative disease, warranting the targeting of other antigens. CD22 represents another promising target, with highly restricted but ubiquitous expression across the B-lineage. However, despite promising preclinical work by several groups with CD22-targeted CAR T cells targeting of this antigen in the clinic has proven difficult, with many patients relapsing with CD22Lo leukemia, contrasting to complete loss of CD19 expression post CD19-CAR. While prior work has demonstrated that a CAR with so-called "tonic" antigen-independent signaling properties has proven to be highly efficacious, tonic signaling has been shown be detrimental to long-term T cell function. Here, we demonstrate a balance between binding affinity and antigen-independent tonic signaling (as determined by length of flexible linker) in determining CAR function. We show that maximal CAR function in the settings CD22Lo and WT leukemia is maintained by boosting binding affinity without shortening flexible linker to induce tonic signaling, establishing rational modification of antigen binding domain as an important approach for modulating the function of cellular therapeutics. GRAPHICAL ABSTRACT O_FIG O_LINKSMALLFIG WIDTH=179 HEIGHT=200 SRC="FIGDIR/small/643183v1_ufig1.gif" ALT="Figure 1"> View larger version (27K): org.highwire.dtl.DTLVardef@1d543e3org.highwire.dtl.DTLVardef@5bf226org.highwire.dtl.DTLVardef@76cd2forg.highwire.dtl.DTLVardef@198fd37_HPS_FORMAT_FIGEXP M_FIG C_FIG

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.