Abstract

CD28 is a costimulatory receptor for T cell receptor activation, and CD28 binding antibodies are in clinical development to block or promote T cell activation[1,2]. CD28-specific binding domains are also used in bi- or tri-specific antibody formats to generate T cell engaging antibodies[3]. We set out to generate CD28-specific mAbs from rabbits using a single B cell cloning workflow on the Beacon system. The Beacon system enables rapid antibody discovery using a function-first approach. Antibody-secreting cells (ASCs) are isolated from immunized animals and individually placed into nanowells (pens) of a Beacon chip. Due to the small size of the pens, the secreted antibody concentration reaches µg/ml levels within minutes, and the functional properties of the antibodies can be tested in sequential or multiplexed bead-based assays with a fluorescent readout. Typical assay duration is 30-60 minutes.

Rabbits were immunized with CHOK1 cells stably transfected to carry human CD28 (hCD28) on their surface. At the end of the immunization schedule, memory B cells were isolated from blood using Ficoll separation and magnetic sorting. B cells were cultured in an activation media, and cells were analyzed after 4 and 5 days of activation. After activation, the average cell size increased and cells proliferated roughly 15-fold, with cell numbers unchanged after that. Flow cytometry showed a reduction of surface IgG and a clear increase of intracellular IgG during activation, consistent with a shift from B cells to antibody secreting cells. On day 4 and 5, a total of 32,716 activated B cells were screened on Beacon, first for overall IgG secretion and then for IgG specific for hCD28. 14,889 cells secreted IgG (76%), and 108 cells secreted IgG specific to hCD28 (0.43% of IgG secretors). cDNA from identified hits was individually exported from the chip and submitted for VH/VL sequencing.

Antibody discovery on the Beacon enabled rapid screening of a large number of ASCs and identification of antigen-specific antibodies within 4 days of obtaining PBMCs from immunized rabbits. Of 23 recombinantly-expressed antibodies, most confirmed good binding to CD28 while several showed superior binding and receptor activation compared to a reference antibody from a large pharmaceutical company.

Request the full poster

Abstract

KRAS is one of the most frequently mutated oncogenes in cancers, especially in lung cancer. Clinical research has shown that about 14% of non–small-cell lung cancers (NSCLCs) carry the KRASG12C mutation, and brain metastases are a relatively common occurrence in KRASG12C-mutated NSCLC. Approximately 30% of patients with KRASG12C mutations develop brain metastases, and those with KRAS-mutated NSCLC and untreated brain metastases have limited treatment options and poor clinical outcomes. AMG510 (Sotorasib) is a small molecule inhibitor of KRASG12C that specifically binds to inactive GDP-bound KRASG12C, preventing its oncogenic signaling. It has been approved for NSCLC by the FDA. Recent clinical studies have shown that Sotorasib exhibits some activity against KRASG12C NSCLC brain metastasis.

To evaluate the in vivo activity of AMG510 for NSCLC brain metastasis treatment and provide insights for the preclinical development of new KRAS targeted inhibitors, we established advanced orthotopic models, including intracranial, intracarotid, and intracardiac metastatic models, using the NCI-H358-luc cell line, which harbors the KRASG12C mutation. We validated these models with AMG510 as a single agent or in combination with radiation, which is commonly used in brain tumor therapeutic strategies, to reveal the efficacy and brain concentration of AMG510 in KRASG12C brain metastasis models. Tumor growth and metastasis incidences were assessed using bioluminescence imaging (BLI). All animal studies were performed in an AAALAC accredited facility and aligned with animal welfare regulations. Our preliminary results reveal that AMG510 could penetrate the blood-brain barrier (BBB) and show therapeutic activity for KRASG12C brain metastatic tumors. These models represent powerful tools for clinically relevant assessment of the efficacy of novel KRAS G12C inhibitors and can be expanded for other KRAS mutations-associated brain metastasis or inhibitors.

Request the full poster

Abstract

Upon immune activation, in addition to the first signal driven by the TCR‐peptide/MHC interactions, the costimulatory signals are indispensable to optimal T-cell activation, the absence of which will trigger naive T cells to an anergic state. The CD28‐CD80/CD86 pathway is widely known as the most classical and critical costimulatory signal for the initial activation of T cells. Bispecific T-cell engagers targeting CD3e can activate cytotoxic T-cells and have shown remarkable tumor regressions in clinical trials. However, the absence of a costimulatory signal through CD28 results in insufficient T-cell activation and early exhaustion. Combination of both CD3e and CD28 targeting antibodies offers attractive approach to optimally boost T-cell activity. Development of superagonistic CD28-targeting antibody has been paused after TeGenero‘s TGN1412 Phase 1 trial in 2006 due to life-threatening cytokine release syndrome. Unlike superagonist TGN1412, which activates T cells and induce cytokine storm, tumor targeting antigen and cross-linking dependent CD28 bispecific antibody is supposed to display little or no toxicity. In the present study, we explored Alpaca immunized phage library approach to generate cross-linking dependent agonist CD28 single domain antibody for bispecific antibody development. Alpacas were immunized subsequently with recombinant human CD28 extracellular domain and 293F cells stably expressing human CD28. Serum samples after immunizations showed decent immune response against CD28. At the end of Boost 2 and 3, peripheral blood was collected and phage display libraries were constructed using VHH specific primers. Phage display libraries were panned against recombinant CD28 ECD protein and CHOK1 cells overexpressing CD28.After one round of panning, both protein and cell panning tracks showed robust enrichment in output titer and repeating sequences. Next, 600 clones were randomly picked for screening by the binding to U266B1 cells endogenously expressing CD28. Of the 326 clones showed positive binding to U266B1, 79 unique sequences were identified. Secondary screening by ELISA and FACS using periplasmic extract confirmed 69 unique sequences that bound to both CD28 protein and cells expressing CD28. Seventeen unique sequences were selected for VHH-Fc production. All purified VHH-Fc showed sub-nanomolar affinity by ELISA and bind to primary T cells by flow cytometry. Three VHH-Fc induced IL-2 production in a primary T cells activation assay only when cross-linked and in the presence of CD3 co-stimulation and are suitable for further bispecific T cell engager development.

Author Block

Yunyun Chen, Yannan Ding, Xiaoyan Liu, Qianqian Zhu, Wenfang Xin, Hu Liu, Teddy Yang, Yikai Qiu

Biologics Discovery, ChemPartner Biologics Co., Ltd., Shanghai, China

Request the full poster

Abstract

Antibody therapeutics has been proven clinically against various cancer types and have achieved great success. Majority of the therapeutic antibodies target cell surface or extracellular proteins. Most of the tumor-specific antigens that control cell growth, proliferation, and death are intracellular, and traditional antibody therapies fail to recognize intracellular antigen targets. Such tumor specific intracellular targets consist of overexpressed self-antigens, oncofetal protein, cancer testis antigens, viral protein and neo-antigens of mutant proteins. Recently, an emerging approach to target these neo-epitopes has been developed called T-cell receptor-like/TCR-mimic antibodies as they recognize similar epitopes to those of T cell receptors which target specific peptide-MHC complexes (pMHC). TCR mimic antibodies open a new avenue for targeting intracellular antigens yet there are lots of challenges for generating highly selective and potent TCR mimic antibodies. MAGE-A4, melanoma-associated antigen A4, is a member of the MAGE protein family of cancer-testis antigen (CTA). In healthy adult, MAGE-A4 expression is restricted to immune-privileged sites. But MAGE-A4 is widely expressed on many cancers such as lung cancer, head and neck squamous cell cancer, synovial sarcoma (SS), ovarian cancer, urothelial cancer and melanoma. The decapeptide GVYDGREHTV (amino acids 230-239) derived from MAGE-A4, can be presented by HLA-A*02:01 molecule, elicits specific cytotoxicity T cell response. Afami-cel, a genetically modified autologous T cell therapy against this pMHC complex, showed promising potency for patients with metastatic SS. So MAGE-A4/ HLA-A*02:01 may be a good target for TCR mimic antibodies and cancer immunotherapy. In the present study, we utilized a newly established single cell cloning platform LyTARS to generate TCR mimic antibodies against HLA restricted MAGE-A4 peptide. Recombinant HLA-A*02:01MAGE-A4 protein was used to immunize mice and plasma B cells from immunized mice were sorted and loaded onto LyTARS system. Clones that showed reactivity to recombinant HLA-A*02:01 MAGE-A4 but not to irrelevant pMHC (WT-1) were exported and sequenced. Purified antibodies were extensively characterized by the specificity, affinity, and potency by ADCC. Results demonstrated that the TCR mimic antibodies identified have sub-nano molar affinity by protein based ELISA, bind to peptide pulsed T2 cells that endogenously expressing HLA with nano-Molar apparent affinity. TCR mimic antibodies also demonstrated antibody dependent cell killing with the potency of sub-nanomolar. In summary, we have generated and extensively characterized a panel of high affinity and selective TCR mimic antibodies against HLA-A*02:01 MAGE-A4 using a newly established single cell cloning platform.

Author Block

Yunyun Chen, Yannan Ding, Xiaoyan Liu, Qianqian Zhu, Wenfang Xin, Hu Liu, Teddy Yang, Yikai Qiu

Biologics Discovery, ChemPartner Biologics Co., Ltd., Shanghai, China

Request the full poster

Abstract

Proteolysis-targeting chimeras (PROTACs) are a novel class of small molecules that induce the degradation of target proteins. Unlike conventional inhibitors that disable target proteins by direct binding, PROTACs are event-driven, meaning that a single PROTAC molecule can induce the degradation of multiple target protein molecules. This catalytic mechanism allows PROTACs to achieve complete inhibition of target proteins at lower concentrations compared to conventional inhibitors, making them a promising therapeutic approach for previously considered “undruggable” targets.

Despite their promise, the development of PROTACs faces several challenges. The large molecular weight of PROTACs can lead to poor cell permeability and bioavailability. Additionally, the lack of high-throughput, quantitative assays for evaluating PROTAC efficacy hinders structure-activity relationship (SAR) studies. The formation of the Protein-PROTAC-E3 ternary complex is crucial for PROTAC function, and reliable methods for assessing this interaction are essential for PROTAC development.
In this poster, ChemPartner showcases a comprehensive suite of preclinical models for PROTAC evaluation, encompassing both in vitro and in vivo approaches. Using low-cost, high-throughput biochemical assays, we can provide valuable insights into PROTAC efficacy, facilitating SAR studies and elucidating the relationship between molecular structure and activity. Followed by in-cell and in vivo function assays, we can test PROTAC PK/PD, toxicity, and efficacy. This suite of assays provides a powerful tool to speed up PROTAC development and reduce the fail rate.

Request the full poster