Heparan sulfate proteoglycan binding properties of adeno-associated virus retargeting mutants and consequences for their in vivo tropism

Abstract

Adeno-associated virus type 2 (AAV-2) targeting vectors have been generated by insertion of ligand peptides into the viral capsid at amino acid position 587. This procedure ablates binding of heparan sulfate proteoglycan (HSPG), AAV-2's primary receptor, in some but not all mutants. Using an AAV-2 display library, we investigated molecular mechanisms responsible for this phenotype, demonstrating that peptides containing a net negative charge are prone to confer an HSPG nonbinding phenotype. Interestingly, in vivo studies correlated the inability to bind to HSPG with liver and spleen detargeting in mice after systemic application, suggesting several strategies to improve efficiency of AAV-2 retargeting to alternative tissues.

FIG. 1.

Proposed model for the influence of several peptide classes on capsid stability and on binding to heparin. A) Three-dimensional atomic structure of the 587 region. R585 and R588 are depicted in blue. B) The two arginines are part of the binding motif. C) A bulky peptide disrupts the heparin binding motif, taking the arginines apart. D) A bulky peptide obstructs the HSPG binding motif. E) Small peptides could preserve the original structure of the loop and heparin binding motif. F) and G) The presence of one or more arginines in the inserted peptide restores the heparin binding ability. H) Due to its proximity to R588, an arginine in the last amino acid position of the insertion is prone to restore heparin binding. In all panels, a functional heparin binding site is indicated by a red pattern. A loop conformation that confers capsid stability is indicated by the blue arrow. wt, wild type.

FIG. 2.

Transduction efficiency of rAAV-RC, the B-rAAV-pool, the NB-rAAV-pool, and rAAV-A3 on HeLa and HepG2 cells. A) HeLa cells were transduced with rAAV-RC (RC; 500 genomic particles per cell [GOI]), the B-rAAV-pool (B; 1,000 GOI), and the NB-rAAV-pool (NB; 5,000 GOI), coding for the green fluorescent protein (GFP) in the absence (gray bars) or presence (black bars) of 425 IU of heparin/ml medium. Transduction levels were determined by fluorescence-activated cell sorter analysis. Each experiment was repeated independently twice, and the mean values and standard deviations (indicated by error bars) are shown. B) HepG2 cells were transduced in the presence of adenovirus (1 PFU/cell) with a 1,000 GOI of rAAV-RC, the B-rAAV-pool, the NB-rAAV-pool, or rAAV-A3 (coding for beta-galactosidase) and were analyzed by Galactolight Plus beta galactosidase assay according to manufacturer's instructions (Tropix). Gene expression was normalized for total protein using bicinchoninic acid (Perbio, United Kingdom) and expressed as relative light units (RLU) per milligram of protein. Each experiment was repeated independently twice, and the mean values and standard deviations (indicated by error bars) are shown. rAAV-RC, dotted bars; B-rAAV-pool, black bars; NB-rAAV-pool, striped bars; rAAV-A3, gray bars.

FIG. 3.

Bioistribution of rAAV-RC, the B-rAAV-pool, the NB-rAAV-pool, and rAAV-A3 in C57/B6 mice. A total of 4 × 10⁹ genomic particles of the different vector preparations (coding for beta-galactosidase) were injected into the tail vein of 12-week-old C57/B6 mice. At 24 h postinfection mice were sacrificed. DNA was extracted from blood and tissues. Vector genomes per tissue were quantified by PCR (TaqMan). rAAV-RC, dotted bars; B-rAAV-pool, black bars; NB-rAAV-pool, striped bars; rAAV-A3, gray bars.