Preclinical humanized genO‑hCD28 mouse model

Baseline validation

• hCD28 expression pattern recapitulates mCD28

  hCD28 and mCD28 expression on freshly isolated splenocytes on (A) CD8⁺ (viable, CD3⁺CD8⁺) T cells, (B) conventional CD4⁺ (viable, CD3⁺CD4⁺Foxp3^-), and (C) Tregs (viable, CD3⁺CD4⁺CD25⁺Foxp3⁺).

  

• hCD28 is functional: α-human CD28 co-stimulation induces T cell activation and proliferation

  Isolated T cells were labeled with CTV and activated with αCD3 and soluble mouse or human αCD28 for 4 days. Representative FACS bar graphs indicating (A) CD25 expression as a marker of activation, and (B) CTV dilution (as a measure of proliferation) gated on live T cells. Results are expressed as mean±SD. Unpaired T test analyses (**p<0.01, ***p<0.001).

  

• Extended validation

• Treatment of genO-hCD28 mice with aCD28/PD-L1 bispecific antibody (NI-3201) leads to inhibition of tumor growth and enrichment of the tumor microenvironment with T cells

  genO-hCD28 mice were implanted with OVA-MC38 cells and then treated twice with NI-3201 (A). NI-3201 treatment leads to a reduction in tumor size (B) as well as an increase in the number of T cells in the TME (C). Adapted from Majocchi et al., 2025.

  

• Extended validation

• Antibody safety assessment in genO-hCD28 mice

  genO-hCD28 mice were treated with TGN1412 compound as a positive control, a human-specific CD28 agonist, described to induce CRS in humans (Suntharalingam et al. 2006) or an anti-mouse CD3, and observed an induction of cytokine release following treatment with both antibodies (A). The associated body weight decrease was also observed (B), while the body temperature remained unchanged (C).

  

• Comparison of PK profiles of 3 similar bsAbs in genO-hCD28 mice

  genO-hCD28 mice were IV treated with 3 bsAbs containing a CD28 binding domain with binding affinities of two-fold difference. The 3 antibodies show relatively slow clearance rates with differences that do not correlate with CD28 binding affinity. Adapted from Thisted et al, 2025