Please use this identifier to cite or link to this item:
Authors: Lui, Bonny Gaby
Title: Targeting the Tumor Vasculature with an Engineered Cystine-Knot Miniprotein Applying an Optimized Phage Display Discovery Platform
Online publication date: 20-May-2019
Language: english
Abstract: Engineered cystine-knot miniproteins represent an auspicious alternative protein scaffold for therapeutic and diagnostic applications, allowing specific tailoring of several molecular properties in addition, to target affinity and specificity, in particular stability solubility and pharmacokinetic behavior of the binding protein. Furthermore, the simple architecture of those small peptidic molecules facilitates conventional chemical manufacturing and the construction of multi-functional fusion molecules. Cystine-knot miniproteins are characterized by a remarkable stability and an extraordinary tolerance to sequence variation enabling the construction of combinatorial libraries. The starting basis for this thesis was a previously established cystine-knot miniprotein technology platform that allows the development of new drug candidates for diagnostic and therapeutic applications. The platform comprised a randomized cystine-knot miniprotein phage library (MCopt 1.0) that was constructed on the basis of an open chain variant of the squash trypsin inhibitor MCoTI-II (Momordica cochinchinensis) and a phage display screening system for ligand selection against soluble target proteins. The first objective of this thesis was the establishment of an efficient cell-based phage panning protocol and the development of a robust procedure to identify cystine-knot miniproteins that bind to plasma membrane proteins in their native conformation. In a proof of concept experiment a previously identified model ligand (MC-FA-010), which binds to FAP-α was utilized. MC-FA-010 phages were mixed at different ratios into the established MCopt 1.0 phage library and the mix was screened against FAP-α expressing eukaryotic cell line. Using an adapted screening protocol, MC-FA-010 phages could be specifically enriched after one, two or three successive screening rounds, depending on the initially applied mixing ratio. Additionally, a novel cell-based hit identification procedure consisting of high-throughput compatible protein expression and purification of individual clones combined with target binding analysis based on flow cytometry was established. The applicability of the whole cell-based panning and downstream hit identification procedure was confirmed by successful identification of high-affinity binders upon screening of the naïve MCopt 1.0 phage library against a FAP-α expressing eukaryotic cell line. The detection sensitivity was further enhanced with the development of a second expression system enabling biotin/streptavidin based tetramerization of proteins in order to gain higher binding strength as a result of the avidity effect. The second objective of this thesis was the design and construction of novel cystine-knot miniprotein phage libraries to increase the repertoire and size of the library. Three rationally designed sub-libraries with scattered randomized positions in loop 1 (MCopt 2.0), loop 5 (MCopt 2.1) or both loops (MCopt 3.0) in the open chain sequence of MCoTI-II were successfully generated. Phage displaying of cystine-knot miniproteins through the minor coat protein pIII was obtained using a pJuFo phagemid system, comprising Fos-cystine-knot miniprotein and Jun-pIII´ gene fusions for independent protein folding. Library quality controls included the expression analysis on phage level via ELISA and western blotting indicating the functional surface presentation of cystine-knot miniproteins. MCopt 2.0 library clones showed good protein expression rates and overall proper folding capabilities, while MCopt 2.1 and MCopt 3.0 library clones performed in a marginal inferior manner. This suggested a moderate tolerance of MCoTI-II scaffold towards sequence randomization in loop 5 or in both adjacent loops 1 and 5. Moreover, also the phagemid backbone was optimized to circumvent an enrichment of deletion mutants (clones lacking the M13 coat protein and the cystine-knot miniprotein encoding sequence part). In the novel phagemid vector, pPDIII-1, repetitive sequence parts were completely eliminated and different genetic elements for tighter gene expression control were introduced. A model selection experiment showed a clearly improved genetic stability of the pPDIII-1-MCopt 2.0 sub-library as compared to the original pJuFo vector. The third objective of this thesis focused on the development of engineered cystine-knot miniproteins for in vivo targeting of the tumor vasculature. To this end, MC-FN-010, a cystine-knot miniprotein identified by screening of MCopt 1.0 und MCopt 2.0 libraries that specifically binds to fibronectin extra domain B (EDB) was used. Detailed characterization of MC-FN-010 via alanine scanning mutagenesis revealed its target binding relevant amino acids and resulted in defining a second derivative candidate (MC-FN-016). Both cystine-knot miniproteins featured high EDB specificity but relatively low affinities. However, chemical oligomerization of the ligands and site-directed fluorescence dye conjugation increased the binding strength enormously, while retaining its high specificity. The resulting trimeric constructs were analyzed in vivo in a U-87 MG based xenograft glioblastoma mouse model. For both variants, a strong accumulation in the tumor and overall low background signals could be detected via in vivo and ex vivo fluorescence measurement. These findings together emphasize the high potential of cystine-knot miniproteins as molecular scaffolds for tumor imaging technologies.
DDC: 570 Biowissenschaften
570 Life sciences
Institution: Johannes Gutenberg-Universität Mainz
Department: FB 10 Biologie
Place: Mainz
URN: urn:nbn:de:hebis:77-diss-1000027889
Version: Original work
Publication type: Dissertation
License: In Copyright
Information on rights of use:
Extent: XIII, 115 Blätter
Appears in collections:JGU-Publikationen

Files in This Item:
  File Description SizeFormat
100002788.pdf5.07 MBAdobe PDFView/Open