Development of a non-immunoglobulin protein scaffold based on human Kringle domain
인간 크링글 도메인 기반 비항체 단백질 골격 개발
초록/요약
The therapeutic monoclonal antibodies (mAbs) and fragments thereof, mostly based on the IgG1 immunoglobulin domain, have been a huge success in the pharmaceutical industry for more than a decade. Despite of the huge success, whole antibodies have several well-known limitations, such as poor tissue penetration, restricted target molecules, a requirement for expression in mammalian cells, intellectual property barriers, and etc. Here I report a novel non-antibody protein scaffold developed based on the fold of kringle domain (KD) derived from human plasminogen. In humans, 39 KDs composed of 78–80 amino acids are present as modular structural units of one to tens of copies in 31 functionally distinct proteins, most of which are present in blood plasma, such as coagulation factors, proteases, apolipoprotein, and growth factors. In particular, the KDs are present independently from the other proteins, or present with the other proteins (for example, endostatin, angiostatin, NK1, NK3, NK4). And they serve to inhibit angiogenesis (antiangiogenesis) or tumor invasion and metastasis, also are being developed as potential cancer treatment in preclinical trials or phase II trials. The plasminogen (Pgn) is second common protein in human blood (100-200 mg/L), and has five KDs. Plasminogen is able to be cleaved into angiostatin (PgnKD1-4) and PgnKD5, and both of them show antiangiogenesis activities1; 2. KDs share a rigid core structure composed of two short antiparallel β-sheets and three disulfide bonds connected in a characteristic 1–6 (Cys1-Cys80), 2–4 (Cys22-Cys63), and 3–5 (Cys51-Cys75) pattern, and these residues are highly conserved. The core folding of KD, which is constrained by the disulfide bonds, presents a structure with seven surface-exposed loops. And the residues in these loops can diverge substantially between individual KDs. A typical KD has good characteristics as a non-antibody protein scaffold, such as avimer which is small in size, individually folded, highly soluble and thermally stable. To exploit the KD scaffold as a target-specific binder, a synthetic KD library was generated on the yeast cell surface by introducing designed mutations into the flexible seven loops (~45 residues). From the synthetic KD library, KD variants isolated biologically functional agonists against TRAIL receptors, death receptor 4 (DR4) and DR5 as well as antagonist against human tumor necrosis factor-α (TNFα). Though incorporated by many mutations, the selected KD variants demonstrated good thermal stability and retained the typical secondary structure of KD, suggesting that the stability and conformation of KD scaffold is strongly sequence-tolerant. The selected KD variants function in vitro and in vivo as agonists to DR4 and/or DR5 or antagonist against TNFα, suggesting that KD-library could play a role of the post-therapeutic antibody. Although the functional KD variants were successfully isolated from KD-library, there is a need to improve the biological activities for development as therapeutics. The anti-cancer KD variants against DR4 and DR5 have low affinity and efficacy compared to the natural ligand, TRAIL. And anti-TNFα KDT26 has low affinity and IC50 compared to infliximab that are therapeutic Abs in rheumatoid arthritis. To optimize the selected KD variants, three different strategies were conducted. First, the bivalent Fc-fused form was constructed to increase the biological functionality by improving the antigen binding affinity through the avidity effect and by cross-linking of the target proteins into oligomeric form. Second, bivalent and/or bispecific KD variants were generated by grafting the high-affinity binding loop onto other loops of the same domain, suggesting that natural KDs can be engineered to have bivalency and/or bispecificity against targeted molecules, generating KD variants with more potent and/or dual function. Third, the affinity of KD548 was increased by spiked oligonucleotides methods of loop 5 and 6 followed by affinity selection. The engineered anti-cancer KD variants with each three methods showed enhanced cell-death-inducing activity of tumor cells compared with their parent KD variants. And anti- TNFα KDT26-Fc showed 7.7 folds higher affinity and 150-fold enhanced IC50, compare with the parent KDT26 In conclusion, a novel non-antibody scaffold based on KD fold can be developed as a specific binder to given targets to modulate the biological activity, and KD variants could be optimized by Fc-fusion, grafting of high-affinity binding loops into other loops within the single domain, and affinity maturation of the target binding loops, like antibodies. Therefore isolated KD variants have a potential to be developed as diagnostics and therapeutics for the target molecule-associated diseases.
more목차
TABLE OF CONTENTS
ABSTRACT OF DISSERTATION
LIST OF TABLES
LIST OF FIGURES
CHAPTER 1. General introduction
1.1 Alternative protein scaffold
1.2 Kringle domain
1.3 Cancer target molecules: death receptor 4 (DR4) and DR5
1.4 Rheumatoid arthritis target molecule: tumor necrosis factor α(TNFα)
CHAPTER 2. Development of non-immunoglobulin protein scaffold based on Kringle domain (KD)
2.1 Abstract
2.2 Introductions
2.3 Materials and Methods
2.3.1 Cell lines and reagents
2.3.2 Expression and purification of protein
2.3.3 Construction of KD library
2.3.4 Screening of KD library against DR4, DR5 and TNFα
2.3.5 Binding analysis by ELISA
2.3.6 Far-UV CD (Circular dichroism) spectroscopy
2.3.7 Size exclusion chromatography
2.3.8 Thermal stability assay
2.3.9 Disulfide bond mapping
2.3.10 Cell viability assay
2.3.11 Annexin V and PI staining
2.3.12 Cell surface binding assay
2.3.13 in vitro TNFα neutralization assay
2.3.14 in vivo antitumor assay
2.4 Results
2.4.1 Strategy and construction of a protein scaffold library based on PgnKD2
2.4.2 Screening and isolation of target-specific KD variants
2.4.3 Biological efficacy of the isolated KD variants
2.4.4 The biochemical characterization of the selected KD variants
2.4.5 Construction and characterizations of Fc-Fused KD variants
2.4.6 in vivo efficacy
2.5 Discussion
CHAPTER 3. Development of bivalent or/and bispecific anti-DR4 or/and DR5 KD variants by loop grafting
3.1 Abstract
3.2 Introductions
3.3 Materials and Methods
3.3.1 Cell lines and reagents
3.3.2 Binding loop mapping
3.3.3 Construction of KD-LCR and loop-grafted KD variants
3.3.4 Expression and purification of KD variants
3.3.5 ELISA, competitive ELISA and sandwich ELISA
3.3.6 Size exclusion chromatography
3.3.7 Surface Plasmon Resonance (SPR)
3.3.8 Mass spectrometer
3.3.9 Cell viability assay
3.4 Results
3.4.1 The LCR1 and LCR2 of KD variants have independent target binding abilities
3.4.2 KD variants can have bivalent target binding ability via LCR1 and LCR2 on the single domain
3.4.3 Binding loop mapping for target recognition in the LCR1 and LCR2 of KD variants
3.4.4 The high-affinity target-binding L56 of KD variants can be grafted onto L34 located on the opposite side
3.4.5 Generation of bivalent and/or bispecific KD mutants against DR4 and/or DR5 by loop grafting
3.4.6 Composition and stoichiometry of bivalent and/or bispecific KD complexes with DR4 and/or DR5
3.4.7 Biological activity of loop-grafted bivalent/bispecific KD mutants against DR4 and/or DR5
3.5 Discussion
CHAPTER 4. Affinity maturation of anti-DR5 KD548
4.1 Abstract
4.2 Introductions
4.3 Materials and Methods
4.3.1 Library construction of TRAIL-like KD548
4.3.2 Isolation of affinity improved clones from library
4.3.3 Expression and purification of KD variants
4.3.4 ELISA
4.3.5 SPR
4.3.6 Cell viability assays
4.4 Results
4.4.1 Strategy and construction of KD548 library
4.4.3 Binding analysis of affinity improved KD variants
4.4.4 Cell viability assay
4.5 Discussion
Conclusion
REFERENCE
ABSTRACT IN KOREAN
PUBLICATIONS AND PATENTS
