Discovery and Characterization of Monoclonal Antibodies for the Treatment of Metastatic Colorectal Cancer

Abstract

Monoclonal antibodies (mAbs) represent one of the most important classes of pharmaceuticals that have revolutionized the treatment of cancer and other serious conditions. There are over 100 antibody-based products that been approved by the U.S. FDA, generating estimated yearly revenues of $220 billion as of 2023. Antibody phage display technology is broadly practiced as it offers a time and cost-effective solution for isolating novel fully human mAbs against virtually any target of interest. The work described in this thesis focused on the isolation and characterization of novel fully human mAbs specific to Carcinoembryonic antigen (CEA) and Programmed cell death protein 1 (PD-1). Metastatic colorectal cancer (mCRC) remains a major cause of cancer-related deaths. The 5-year survival rate of patients with mCRC is 15%, highlighting the urgent need for novel and more effective pharmaceutical options. The standard of care for newly diagnosed mCRC is based on chemotherapy regimens, typically encompassing a cocktail of leucovorin, 5-fluorouracil or capecitabine and either oxaliplatin (FOLFOX protocol) or irinotecan (FOLFIRI protocol), often in combination with mAbs against either vascular endothelial growth factor-A (VEGFA; e.g., Bevacizumab) or epidermal growth factor receptor (EGFR; e.g., Cetuximab). At first recurrence, treatment aims to counteract resistance, switching oxaliplatin-based to irinotecan-based treatment as well as anti-EGFR to antiVEGF-A biological or vice versa. At second recurrence, patients typically receive multikinase inhibitors, such as regorafenib. The response rates to treatment go from 35-60% in the first line to 15% and 1% in the second and third line, respectively. Thus, there is a big unmet medical need to develop more efficacious and selective treatment strategies for mCRC. Carcinoembryonic antigen (CEA) represents the most specific and extensively validated antigen for antibody-based mCRC targeting. The availability of a fully human anti-CEA antibody should facilitate the implementation of pharmacodelivery strategies aimed at concentrating bioactive payloads at the site of disease, helping spare normal tissues. The first part of this thesis reports the isolation and affinity maturation of the antiCEA human F4 antibody by phage display technology. F4 in single-chain variable fragment (scFv) format exhibited a dissociation constant of 7.7 nM to the cognate antigen, as measured by surface plasmon resonance (SPR). The binding specificity was confirmed by flow cytometry and immunofluorescence analyses on human cancer samples. In contrast to previously generated CEA-specific antibodies, F4 did not cross-react with cells expressing other members of the carcinoembryonic antigen-related cell adhesion molecule (CEACAM) family, namely CEACAM1 and CEACAM6, which are abundant in healthy tissues. F4 showed preferential accumulation in CEA-positive neoplastic lesions following intravenous administration in tumor-bearing mice, as evidenced by two orthogonal in vivo biodistribution studies. Encouraged by these results, I genetically fused murine interleukin 12 to F4 in the single-chain diabody (scDb) format. F4-IL12 demonstrated potent antitumor efficacy in two murine models of colon cancer. Treatment of mice with F4-IL12 significantly elevated the density of tumorinfiltrating lymphocytes, which were stimulated to produce interferon-γ (IFNγ). Collectively, these results suggest that the F4 antibody may represent an attractive delivery vehicle for targeted cancer therapy. In the second part of the thesis, I used the F4 antibody and the 2C11 murine antimouse CD3 antibody to generate bispecific antibodies (TCBs) that simultaneously recognize tumor cells and T cells. I cloned and expressed four different formats based on the F4-2C11 fusions: (i) a bispecific immune T cell engager (BiTE) based on a tandem arrangement of scFv fragments, (ii) a 2+1 format based on scDb(F4) fused to scFv(2C11), (iii) a 2+1 format based on IgG(F4), wherein the scFv(2C11) was fused to the C-terminus of one heavy chain via knobs-into-holes, and (iv) a 2+2 format based on IgG(F4), with two scFv(2C11) fused to the C-termini of both heavy chains. The binding of the four TCBs to CEA-expressing cancer cells and murine T cells was confirmed by flow cytometry. In an in vitro killing assay, using activated mouse T cells incubated with CEA-expressing target cells, the potency of the formats was compared and the 2+1 scDb-scFv format emerged as the most potent variant. Subsequently, I evaluated the antitumor activity of the best TCB format, both as a single agent and in combination with the L19-IL2 antibodycytokine fusion, in murine models of CEA-positive cancer. Moderate antitumor activity was observed, but mice were not cured. These findings are in keeping with the experience of other research groups, who have previously reported that it is difficult to eradicate solid tumors using bispecific antibodies. The third and last part of the thesis focuses on the isolation and in vitro assessment of a novel fully human mAb specific to Programmed cell death protein-1 (PD-1). PD-1 is an immunoregulatory target expressed on activated T lymphocytes. The antibody-mediated blockade of PD-1 has yielded objective responses in patients with various types of cancer. I isolated a fully human antibody (termed D12) specific to the extracellular domain of human PD-1 from a novel synthetic antibody library, which comprised 750 million clones and featured combinatorially mutated residues in the CDR3 loops of a scFv fragment. The affinity of the D12 antibody was measured by SPR and binding to PD-1 expressed on primary human T cells was confirmed by flow cytometry. Blockade of the interaction between PD-1 and its cognate ligands programmed death-ligand 1 and 2 (PD-L1 and PD-L2) was demonstrated in a competition ELISA, with EC50 values of 4.4 nM and 12.5 nM, respectively. The detailed interaction between the D12 antibody and PD-1 was characterized by X-ray crystallography in collaboration with the group of Prof. Dr. Federico Forneris (Pavia, Italy), uncovering an unprecedented conformational shift at the N-terminus of PD-1 after D12 binding. A comparative analysis with other PD1 blocking antibodies, which are currently used in clinical practice, confirmed the distinctive attributes of the D12 antibody in recognizing a novel epitope and blocking PD-1 function. Collectively, these findings provide a rationale for a possible industrial development of D12 as a novel PD-1 blocker for cancer immunotherapy.