The immunosuppressant rapamycin has been shown to inhibit G 1 /S transition of the cell cycle. This inhibition is thought to be mediated by maintenance of the threshold levels of cyclin-dependent kinase (CDK) inhibitor p27 Kip1 (p27) and inhibition of p70 s6 kinase (p70 s6k ). However, recent evidence suggests that cells still remain sensitive to rapamycin in the absence of functional p27 or p70 s6k . Here, we show that rapamycin represses cyclin D3 levels in activated human T lymphocytes with no inhibitory effects on cyclin D2. Furthermore, rapamycin elicits similar cyclin D3 modulatory effects in B lymphocytes. The overall effect of rapamycin on cyclin D3 leads to impaired formation of active complexes with Cdk4 or Cdk6 and subsequent inhibition of cyclin D3/ CDK kinase activity. Decrease in cyclin D3 protein levels is due to translational repression and not due to attenuated transcription of the cyclin D3 gene. Importantly, stable overexpression of cyclin D3 (2-2.5 fold) in Jurkat T cell transfectants renders them resistant to lower doses (1-10 ng/ml) of rapamycin. These results point to a critical role of cyclin D3 in rapamycin-mediated immunosuppressive effects in T cells and cell cycle regulation in lymphocytes in general.

e12513 Background: Current immunotherapy involves redirection of T cells against tumor either by chimeric antigen receptors (CAR-T) or bispecific T-cell engagers (BiTEs). These T cell-engaging technologies have been enabled by engineering single-chain variable fragment (scFv). Most BiTEs are designed as heterodimers, with monovalent binding to tumor cell surface antigens and to CD3+ T cells. Here, we describe a novel tetravalent bispecific (TetraBi) platform: a tumor associated antigen specific, T-cell engaging antibody that provides significant advantages over traditional BiTEs. Specifically, we showcase three unique TetraBi antibodies targeted for various cancers. Methods: ABP-100 was engineered by fusing CD3-binding scFv to the c-terminus of anti-HER2 IgG1 light chain, resulting in two binding sites each for HER2 and CD3. In preclinical studies, ABP-100 was compared with a clinical-stage heterodimeric antibody that is monovalent for HER2 and CD3. We also built TetraBi for anti-Glypican 3 (GPC3, ABP-110) and Claudin 18.2 (CLDN18.2, ABP-150) for gastric cancer therapy. Results: We observed bivalent binding to the tumor antigens (HER2, GPC3, CLDN18.2) on tumor cells like the monoclonal antibodies do. Interestingly, the CD3-binding domains in our TetraBi format showed functionally monovalent binding to CD3. ABP-100 displayed comparable cytokine release to the heterodimer molecule both in vitro and in vivo. In HER2+ cancer models, ABP-100 showed highly potent antitumor activity, resulting in complete responses in tumor bearing mice at doses as low as 0.1 mg/kg with no evidence of tumor relapse. ABP110 and ABP-150 also demonstrated potent cancer cell killing in vitro. Conclusions: This novel TetraBi antibody format enables concurrent bivalent and monovalent co-engagement of distinct target antigens in tumor microenvironment with promising therapeutic implications. Furthermore, it provides a potentially larger therapeutic index than more traditional bispecific formats that feature monovalent recognition of tumor antigens.

No abstract