Pre-arrayed Pan-AAV Peptide Display Libraries for Rapid Single-Round Screening

Abstract

Display of short peptides on the surface of adeno-associated viruses (AAVs) is a powerful technology for the generation of gene therapy vectors with altered cell specificities and/or transduction efficiencies. Following its extensive prior use in the best characterized AAV serotype 2 (AAV2), recent reports also indicate the potential of other AAV isolates as scaffolds for peptide display. In this study, we systematically explored the respective capacities of 13 different AAV capsid variants to tolerate 27 peptides inserted on the surface followed by production of reporter-encoding vectors. Single-round screening in pre-arrayed 96-well plates permitted rapid and simple identification of superior vectors in >90 cell types, including T cells and primary cells. Notably, vector performance depended not only on the combination of capsid, peptide, and cell type, but also on the position of the inserted peptide and the nature of flanking residues. For optimal data availability and accessibility, all results were assembled in a searchable online database offering multiple output styles. Finally, we established a reverse-transduction pipeline based on vector pre-spotting in 96- or 384-well plates that facilitates high-throughput library panning. Our comprehensive illustration of the vast potential of alternative AAV capsids for peptide display should accelerate their in vivo screening and application as unique gene therapy vectors.

Figure 1

Establishment of Peptide Display in 12 Different AAV Serotypes (A and B) Cloning strategy to introduce two SfiI sites into the different AAV cap genes (shown as an example is wild-type AAV2 [wt2]). For comparison, the conventional strategy requiring a series of site-directed mutagenesis steps is depicted in (A), versus the streamlined version established and used here (B). Numbers denote amino acid positions within VP1. Amino acids are shown in single-letter code. Nucleotides or amino acids that are changed at each step are highlighted in red. Also indicated are the SfiI binding and cleavage sites. The additional G nucleotide that causes a disrupting frameshift, which is corrected upon oligonucleotide insertion, is circled. In (B), P is the PCR step and L the ligation reaction. Highlighted in green are the NsiI sites that are used to join the two PCR products. (C) Two alternative designs of peptide-encoding oligonucleotides, resulting in either a serine (left) or arginine (right) upstream of the peptide (PEP). The overhangs highlighted in orange match those in the SfiI-restricted plasmid backbone. (D) Overview of all combinations of AAV serotypes and peptide insertion sites studied here (see also Table S1). (E) Confirmation of successful insertion or correction, respectively, of a frameshift in the cap gene of all 12 serotypes (see also A–C) by western blotting. Antibody 303.9 was used to detect all four Rep proteins in the left column and B1 for detection of VP1–VP3 in the right column. Note that VP1–VP3 expression is ablated in the clones containing the two Sfii restriction sites (2× SfiI) but no insert, and reconstituted upon insertion of peptide-encoding oligonucleotides (+ peptide). Also note that B1 does not detect VP1–VP3 of AAV4 or AAV12. Unmodified wild-type cap genes were used as positive controls at the top.

Figure 2

Validation of the Feasibility to Express Peptides in 12 Different AAV Serotypes (A) Schematic overview of the panel of capsid-peptide combinations that was tested in this figure. See Table S2 for sequences of peptides P1–P6. (B) Confirmation of comparable VP expression and particle production. VP1–VP3 were detected using the B1 antibody in transfected cells (left) or cell supernatants after freeze-thaw cycles (right). Note that capsid proteins comprising peptide P6 were only detected in cell pellets but not in supernatants. (C–G) Functional validation through transduction of HeLa cells in 96-well format. Results are depicted as microscopic images of YFP expression (C), as heatmaps representing percentages of YFP-positive cells (D, transduction rate) or mean YFP expression measured via FACS (E), as a heatmap representing firefly luciferase expression (F), or as a bubble chart where the size of the inner circles represents the transduction rate and the intensity of the green color indicates the mean YFP expression (G). E, expression level; T, transduction rate.

Figure 3

Modeling of Selected Peptide Insertions in AAV1 or AAV2 (A and B) Peptide-modified VP3 monomers of AAV1 (A) or AAV2 (B) comprising the seven peptides shown. Shown underneath are superimposed VP3 models. (C and D) Two views of a trimer of AAV1 (C) or AAV2 (D) VP3 containing peptide P2. Left: side view. Right: view down the three-fold symmetry axis. (E and F) Depth cue surface representation of the wild-type AAV1 (E) or AAV2 (F) capsid (left in each panel) or the derivative structure comprising the P2 peptide (right in each panel; P2 is shown in magenta). The color keys show the depth cue in Å. (A)–(D) and the right structures in (E) and (F) were generated in PyMOL. The wild-type capsid structures shown on the left in panels (E) and (F) were generated in Chimera.

Figure 4

Modeling of Peptide Insertion Sites in 12 Different AAV Serotypes Shown are the VR-VIII and the sites of peptide insertions. The latter are indicated by arrows, with site (i) shown in blue and site (ii) in orange. The six balls represent the amino acids flanking the insertion site (site [i], in case two were tested). The most protruded residue in each loop is colored in turquoise, and the corresponding residue is shown above. Values in brackets indicate the numbers of amino acids left of, at, or right of the peak of each loop (e.g., 1-1-4 denotes one residue left of the peak, one at the peak, and four right of the peak). Based on these characteristics, the modified capsids carrying insertion site (i) were segregated into the shown three groups. Note that insertion site (i) is slightly shifted in groups 1 and 3 as compared to group 2, while insertion site (ii) matches the position in group 2.

Figure 5

Examples for Robust Transduction with Selected Capsid-Peptide Combinations Shown are representative fluorescence microscopy images of cells from different origins transduced with the indicated YFP-expressing AAV variant. Images were typically taken 2–3 days post-transduction. Letters in white circles denote (A) actin staining, (B) brightfield, (H) Hoechst staining of nuclei, (M) merge, or (Y) YFP expression. (A) Primary cortical neurons isolated from a rat on day 1 after birth. (B) Primary murine 3D hepatocyte culture transduced with one of the AAVDJ-NXXR peptide variants from Figure 6. The actin staining of canaliculi confirms the intact morphology of the hepatocytes. (C) Primary human monocyte-derived macrophages. (D) Human T cell line SupT1. The arrow indicates decreasing vector doses used for transduction (upper row, 1 × 10⁶ vector genomes [vg]/cell; lower row, 1 × 10⁵ vg/cell). (E) Primary human 3D hepatocytes transduced with wild-type AAV7 or its shown peptide-modified derivatives. Scale bars, (A), (C), and (E), 100 μm; (B) and (D), 50 μm.

Figure 6

Dissection of the Role of Individual Amino Acids in Peptides with an NXXRXXX Motif (A) Results of an alanine walk in peptide NDVRSAN (P4) in AAV1. Alanines that replace original residues are highlighted in red. All vectors were produced in parallel in small scale and used to transduce HeLa cells in a 96-well plate in 10-fold dilutions. Colors represent transduction rates. The two most critical residues whose change to alanine caused the largest drops in infectivity, N in position 1 and R in position 4, are boxed. The square brackets highlight the fact that in this peptide, residue 6 was alanine to begin with. (B) Results of screening of 85 vectors based on the shown four capsid scaffolds and displaying the indicated peptides in four cell types, i.e., HeLa (H), RAW (R), JAWS (J) and SupT1 (S). Transduction rates are depicted as heatmaps. In version (a), results in all four cell lines are shown at the same scale, whereas in (b), they are colored individually for each cell line, with white indicating the highest transduction rate that was measured within each line. Numbers 1 and 4 indicate residues N and R, respectively, that were kept constant in all peptides. The nature of individual residues is color-coded according to the legends at the bottom.

Figure 7

Analysis of the role of residues in the AAV backbone flanking the peptide insertions (A and B) Effect of the arginines at position 585 or 588 in AAV2 or of their replacement with glutamine (R585Q) or serine (R588S), respectively, on transduction of cells (A; human liver organoids are shown as example in B). The inserted peptide in this example was K2. (C) Analysis of 64 peptide-capsid combinations based on AAV2 or AAV-DJ in which R585 or Q585 was juxtaposed with R588 or S588 in all possible permutations, and then combined with the shown eight peptides. The heatmaps illustrate transduction efficiencies measured in the indicated cell lines. (D) Comparison of insertion of a published keratinocyte-specific peptide (boxed) in AAV2 using our own strategy (site [ii]) or the reported design with multiple flanking alanines (AAA/AA). The heatmap shows the results of transduction of human keratinocytes. For comparison, the same peptide was inserted in 11 other serotypes as well, always using site (ii) when two were available. (E) Comparison of various combinations of amino acids flanking peptide insertions in AAV2. Shown in this example is peptide A2 (boxed). (F) Effect of the juxtaposition of the shown three peptides with the indicated flanking residues in AAV2 on transduction of HeLa cells.

Figure 8

Reverse Spotting/Transduction for Rapid and Simple AAV Capsid Screening (A) Workflow for AAV spotting and drying in 96- or 384-well plates, permitting long-term storage and shipping. (B and C) Correlation plots showing the transduction rates (B) and the mean YFP signals (C) of liquid versus reverse transduction in HeLaP4 cells in a 96-well format. Values per AAV mutant were averaged. Shown are linear regression analyses and R2 values, indicating the reproducibility between the two conditions. (D) Bubble charts corresponding to the data in (B) and (C) and depicting transduction rates and YFP expression levels. wt, wild-type.