Immune responses to AAV vectors: overcoming barriers to successful gene therapy

Abstract

Gene therapy products for the treatment of genetic diseases are currently in clinical trials, and one of these, an adeno-associated viral (AAV) product, has recently been licensed. AAV vectors have achieved positive results in a number of clinical and preclinical settings, including hematologic disorders such as the hemophilias, Gaucher disease, hemochromatosis, and the porphyrias. Because AAV vectors are administered directly to the patient, the likelihood of a host immune response is high, as shown by human studies. Preexisting and/or recall responses to the wild-type virus from which the vector is engineered, or to the transgene product itself, can interfere with therapeutic efficacy if not identified and managed optimally. Small-scale clinical studies have enabled investigators to dissect the immune responses to the AAV vector capsid and to the transgene product, and to develop strategies to manage these responses to achieve long-term expression of the therapeutic gene. However, a comprehensive understanding of the determinants of immunogenicity of AAV vectors, and of potential associated toxicities, is still lacking. Careful immunosurveillance conducted as part of ongoing clinical studies will provide the basis for understanding the intricacies of the immune response in AAV-mediated gene transfer, facilitating safe and effective therapies for genetic diseases.

Figure 1

Structure and tropism of wild-type AAV and of recombinant AAV vectors. (A) Gene therapy vectors are complex therapeutics requiring proper assembly of both DNA and protein components to generate the final product. Wild-type AAVs are small nonenveloped viruses, 20 to 25 nm in diameter, with a single-stranded DNA genome of ∼4.7 kb encoding 2 sets of genes, the rep genes required for replication and virion assembly and the cap genes that encode the 3 proteins that assemble to form the 60-mer viral capsid (upper bar). AAV vectors are composed of an outer protein shell, an exact or close replica of the AAV viral capsid, carrying a therapeutic gene of interest under the control of an appropriate promoter (lower bar). The vector is 74% protein by molecular weight. Maximum packaging capacity is ∼5 kb DNA, a limitation of AAV as a gene delivery vehicle. (B) Dozens of different naturally occurring AAV capsids, as well as genetically engineered ones, have been isolated for study, from humans and from other species. The capsid sequences are highly conserved, from 60% to >99%, but studies with naturally occurring serotypes and purpose-engineered capsids^- have shown that even small differences in capsid sequence may affect tissue tropism of a vector and can be exploited to improve therapeutic outcomes. Figure 1A reprinted from Xie et al with permission. Copyright 2000 National Academy of Sciences, USA. Figure 1B reprinted from Arrunda and Xiao with permission. Copyright 2006 John Wiley and Sons.

Figure 2

Clinical results following hepatic artery infusion of AAV2-F.IX. (A) Time course of F.IX levels and of transaminases in first subject to receive high-dose AAV2-F.IX (2 × 10¹² vg/kg) in hemophilia B trial. An F.IX level (red) of ∼10% persists for 4 weeks and then slowly declines, concomitant with a self-limited rise in liver transaminases, which peaked at levels close to 10-fold over the baseline and then gradually fell, in tandem with a falling F.IX level, so that all laboratory parameters had returned to the patient’s baseline by ∼12 weeks after vector infusion (ALT, blue; AST, green). The patient remained asymptomatic throughout these events and, when tested subsequently, responded normally to infused recombinant F.IX. (B) A subsequent subject received a fivefold lower dose, 4 × 10¹¹ vg/kg. F.IX levels (red) did not rise above 1%, but liver enzymes (ALT, blue; AST, green) rose and fell in a time course similar to that seen in the previous subject. Note difference in scales of liver enzymes. ALT, alanine aminotransferase; AST, aspartate aminotransferase. Reprinted from Manno et al.

Figure 3

Expansion of a population of capsid-specific CD8+ T cells after vector infusion, and working model of capsid processing and presentation in hepatocyte. (A) Time course of serum transaminases and frequency of AAV2 capsid peptide–specific CD8⁺ T cells in PBMCs, in subject infused at 4 × 10¹¹ vg/kg dose. (B) Working hypothesis of CD8⁺ T-cell–mediated destruction of AAV-transduced hepatocytes. On transduction, vector enters hepatocytes, and following escape from the endosome and uncoating, vector DNA directs synthesis of F.IX. Some capsid remains behind in the cytosol and undergoes proteasomal processing. Capsid-derived peptides are then transported to the endoplasmic reticulum and loaded onto MHC I molecules, making the transduced hepatocyte a target for circulating capsid-specific CD8⁺ T cells. Note that activation of CD8⁺ T cells and presentation of capsid-derived peptides must occur in a similar time course to result in a clinically detectable event. Figure 3A reprinted from Mingozzi et al, and Figure 3B reprinted from Mingozzi and High.

Figure 4

Capsid antigen presentation is dose-dependent, and can be blocked with a protosomal inhibitor. (A) Generation of a genetically modified T-cell line used as a sensitive detector of peptide-MHC complexes on the cell surface. The T-cell line is stably transfected with a luciferase gene under the control of a promoter containing the NFAT sequence. The cell also carries a TCR cloned from a human subject that recognizes a specific peptide from the AAV capsid complexed with HLA B*0702. Upon engagement of the TCR luciferase is produced. (B) Human hepatocyte cell line HHL5-B7 is transduced with AAV at progressively higher MOIs (x-axis). When the TCR recognizes its cognate peptide-MHC complex, luciferase expression increases (y-axis) in proportion to the number of peptide-MHC complexes engaged. (C) Treatment of cells with the proteasome inhibitor bortezomib reduces AAV capsid antigen presentation in a dose-dependent manner. MOI, multiplicity of infection; NFAT, nuclear factor of activated T cells; RLU, relative light units. Reprinted from Finn et al.

Figure 5

T-cell–mediated immunity to the capsid in the AAV8-F.IX trial. (A-B) Subject 5 received a vector dose of 2 × 10¹² vg/kg. At week 8 postinfusion, his F.IX levels (red line) began to decline and his liver enzymes increased (ALT, green line). Concomitantly, capsid-specific T cells became detectable in peripheral blood. A course of prednisolone was associated with resolution of the transaminitis and partial rescue of F.IX transgene expression levels. (C-D) Subject 6, dosed at 2 × 10¹² vg/kg, experienced an increase in liver enzymes and was promptly treated with steroids. In this subject, increase in liver enzymes and decrease in F.IX transgene expression levels was also associated with detection of capsid-reactive T cells in PBMCs. Reprinted from Nathwani et al.

Figure 6

Model of the relationship between capsid dose and outcome of gene transfer following systemic vector delivery. Low capsid doses are more likely to be neutralized by anti-AAV antibodies, even low-titer NAb. This results in lack of efficacy. Higher capsid doses overcome this limitation, leading to therapeutic efficacy. Capsid-specific T-cell activation is detected as the total capsid dose administered increases. This does not affect efficacy until a critical threshold is reached, above which immune-mediated clearance of transduced target cells results in loss of efficacy.