A novel class of self-complementary AAV vectors with multiple advantages based on cceAAV lacking mutant ITR

Abstract

Self-complementary AAV vectors (scAAV) use a mutant inverted terminal repeat (mITR) for efficient packaging of complementary stranded DNA, enabling rapid transgene expression. However, inefficient resolution at the mITR leads to the packaging of monomeric or subgenomic AAV genomes. These noncanonical particles reduce transgene expression and may affect the safety of gene transfer. To address these issues, we have developed a novel class of scAAV vectors called covalently closed-end double-stranded AAV (cceAAV) that eliminate the mITR resolution step during production. Instead of using a mutant ITR, we used a 56-bp recognition sequence of protelomerase (TelN) to covalently join the top and bottom strands, allowing the vector to be generated with just a single ITR. To produce cceAAV vectors, the vector plasmid is initially digested with TelN, purified, and then subjected to a standard triple-plasmid transfection protocol followed by traditional AAV vector purification procedures. Such cceAAV vectors demonstrate yields comparable to scAAV vectors. Notably, we observed enhanced transgene expression as compared to traditional scAAV vectors. The treatment of mice with hemophilia B with cceAAV-FIX resulted in significantly enhanced long-term FIX expression. The cceAAV vectors hold several advantages over scAAV vectors, potentially leading to the development of improved human gene therapy drugs.

Keywords: Adeno-associated virus; covalently closed-end double-stranded AAV; genome integrity; self-complementary AAV vectors; transduction.

Graphical abstract

Figure 1

cceAAV vector design, replication, and packaging model In the cceAAV vector plasmid, a 56-bp sequence (tel) replaced the mITR in the scAAV vector plasmid (1). Before the triple plasmid transfections, cceAAV vector plasmid is treated with TelN, leaving 2 cce at the site of cleavage (2). After transfection (3), a similar process including Rep-mediated resolution, replication, and packaging as scAAV occurs is carried out in producer cells (4). Finally, an ssAAV DNA with 2 unwound complementary stands is encapsidated into the AAV virion (5). Once released, it will self-anneal the dsDNA form with 1 cce end. Solid line, old DNA strand; dash-dot line, new synthesis DNA strand. iso-dimer, dimer with a closed end; NE, no end. tel, TelN recognition site (TAT CAG CAC AAT TGC CCA TTA TAC GCG CGT ATA ATG GAC TAT TGT GCT GAT A).

Figure 2

cceAAV vectors yield comparable titers with scAAV vectors The vector genome copy number was measured by quantitative real-time PCR. The specific primers targeting the transgene GFP, Gluc, and hFIX, respectively, were applied for the quantitative real-time PCR assay. The mean ± SD represents the average titer from 3 independent batches. NS, no statistically significant difference (p > 0.05, Mann-Whitney test, n = 2).

Figure 3

Analysis of cceAAV genome replicative intermediates by Southern blot The triple plasmids of pHelper (pFΔ6), pAAV ITR-transgene, and pAAVrep-cap were, at the equal mole ratio, cotransfected into 293 cells. At 48 h after transfection, Hirt DNA was extracted and further subjected to DpnI treatment. DNA samples were run by native (A) and alkaline (B) gel electrophoresis, respectively. The ITR probe was applied for detecting replicated AAV DNA. The relative intensity ratio of the dimeric replicative form (D) to the monomeric replicative form (M) are depicted by the graphics at right. cceAAV DNA replicates more efficiently compared to scAAV. p < 0.01: statistically significant (Mann-Whitney test, n = 2).

Figure 4

Analysis of encapsidated cceAAV genomes by Southern blot (A and B) The vector scAAV2-CB-EGFP (1), cceAAV2-EGFP (2), scAAV8-TTR-hFIX (3), and cceAAV8-TTR-hFIX (4) were purified by iodixanol gradient. The vector DNA was extracted and subsequently subjected to separate by native (A) and alkaline (B) gel electrophoresis, respectively. The ITR probe was applied for detecting AAV DNA. (C) The relative intensity ratio of subgenomes versus full genomes by alkaline gel electrophoresis. Triangles indicate subgenomes. cceAAV genome population contains fewer subgenomes than that of scAAV. ∗∗p < 0.01: statistically significant (Mann-Whitney test, n = 2).

Figure 5

In vitro transduction efficacy The scAAV2-CB-EGFP or Gluc and cceAAV2-EGFP or Gluc vectors were transduced into GM16095 or HepG2, respectively, at an MOI of 1,000, and then incubated for 48 h. (A) EGFP⁺ cells were analyzed by flow cytometry, and (B) Gluc activity was measured by Spectra Max iD3. ∗∗p < 0.01: significant difference.

Figure 6

Comparison of transduction efficacy between the 2 vectors in vivo scAAV-TTR-FIX or cceAAV-TTR-FIX were injected into hemophilia B mice, at 1 × 10¹² vg/mouse (10 mice/group). FIX levels in plasma were analyzed by ELISA assay at the set time points.