Development of a new affinity maturation protocol for the construction of an internalizing anti-nucleolin antibody library

Abstract

Over the last decades, monoclonal antibodies have substantially improved the treatment of several conditions. The continuous search for novel therapeutic targets and improvements in antibody's structure, demands for a constant optimization of their development. In this regard, modulation of an antibody's affinity to its target has been largely explored and culminated in the discovery and optimization of a variety of molecules. It involves the creation of antibody libraries and selection against the target of interest. In this work, we aimed at developing a novel protocol to be used for the affinity maturation of an antibody previously developed by our group. An antibody library was constructed using an in vivo random mutagenesis approach that, to our knowledge, has not been used before for antibody development. Then, a cell-based phage display selection protocol was designed to allow the fast and simple screening of antibody clones capable of being internalized by target cells. Next generation sequencing coupled with computer analysis provided an extensive characterization of the created library and post-selection pool, that can be used as a guide for future antibody development. With a single selection step, an enrichment in the mutated antibody library, given by a decrease in almost 50% in sequence diversity, was achieved, and structural information useful in the study of the antibody-target interaction in the future was obtained.

Figure 1

WebLogo representations of the merged CDR3 region nucleotide sequence (a) and amino acid sequence (b) from the pre-selection library. The height of each stack represents the conservation at that position, measured in bits, and the height of each symbol is proportional to its frequency at that particular position. C—Cysteine, G—Guanine, A—Adenine, T—Thymine, N—Unidentified nucleotide, P—Proline, Q—Glutamine, R—Arginine, S—Serine, A—Alanine, L—Leucine, X—Unidentified amino acid.

Figure 2

Fluorescence signal of phages labelled with pH-sensitive dye pHrodo as function of pH. Phages displaying the reference antibody were labeled with pH sensitive dye pHrodo Green and used as coating in ELISA plates. After being incubated with neutral (pH 7.4) or acidic (pH 4) solutions and washed twice with PBS, fluorescence was measured at 512 nm. A control experiment using non-labeled phages was performed in parallel. Data represent the mean of three independent experiments, performed in duplicate. Differences in emitted fluorescence were evaluated by two-way ANOVA followed by Tukey’s test (*p < 0.05, **p < 0.01, ***p < 0.001).

Figure 3

Interaction of nucleolin-overexpressing cells with pHrodo-labeled phages displaying the reference anti-nucleolin antibody under non-internalizing (4 °C) or internalizing (37 °C) conditions. Incubations at both temperatures were performed in parallel experiments, and further analyzed by flow cytometry Data represent the mean of three independent experiments. Differences in emitted fluorescence were evaluated by Unpaired t-test (*p < 0.05, **p < 0.01, ***p < 0.001).

Figure 4

Interaction of pHrodo labeled phages displaying either the constructed antibody library (blue) or the reference antibody (red) with nucleolin-overexpressing cells, under internalizing conditions (37 °C). After analysis by flow cytometry, green fluorescence emitted by internalized labeled phages was represented in the form of histogram (a) and dot plot (b). Data result from a representative experiment. Green—Untreated cells.

Figure 5

Representative scatter plots of either labeled reference antibody- (a) or library-displaying phages (b). Nucleolin-overexpressing cells were incubated with the referred labeled phages at 4 °C for 45 min, followed by two washing steps, and a final incubation at 37 °C for 1.5 h. In both scatter plots, the FITC + gate was considered positive for the highest extent of internalization, and the corresponding phages were collected for further analysis. Data represent one of two experiments performed in parallel.

Figure 6

Comparison between the percentage of clusters with different number of sequence copies in the pre- and post-selection libraries. NGS analysis of VHH fragments from the original antibody library and the post-selection antibody pool included the clustering of amino acid sequences, thus allowing the identification of antibody clusters. Each cluster corresponds to a different amino acid sequence, and may be composed of a single or multiple copies of that same sequence.

Figure 7

WebLogo representation of the merged CDR3 region nucleotide sequence (a) and amino acid sequence (b) from the post-selection pool. The height of each stack represents the conservation at that position, measured in bits, and the height of each symbol is proportional to its frequency at that particular position. C—Cysteine, G—Guanine, A—Adenine, T—Thymine, N—Unidentified nucleotide, P—Proline, Q—Glutamine, R—Arginine, S—Serine, A—Alanine, L—Leucine, X—Unidentified amino acid.

Figure 8

Selection and sequencing of VHH mutant antibodies. Five VHH mutant clones were selected by performing an expression and binding assay. Eighty-eight bacteria colonies from the plated post-sorter library were randomly picked, and used to express correspondent VHH mutant clones, by IPTG induction. Medium only, ER2738 bacteria without VHH coding plasmid, and bacteria bearing reference VHH coding fragment were used as controls. The five clones with the highest expression levels, given by increased values of Absorbance at 405 nm (a), were selected for a binding assay against MDA-MB-435S nucleolin-overexpressing cells. Flow cytometry assay was performed upon incubating the cells with 1 µg of each antibody for 45 min at 4 °C, followed by incubation with anti-HA-Alexa Fluor 647 secondary antibody, for 30 min at room temperature. Mean Fluorescence Intensity (MFI) normalized to cells incubated with secondary antibody only (b) and corresponding overlayed histograms of the green fluorescence signal (c) are presented. The two clones showing highest binding to nucleolin-overexpressing cells were sequenced and aligned against the reference VHH (d). Observed mutations are highlighted in red. Data from a representative experiment.