Rational Engineering and Preclinical Evaluation of Neddylation and SUMOylation Site Modified Adeno-Associated Virus Vectors in Murine Models of Hemophilia B and Leber Congenital Amaurosis

Abstract

Synthetic engineering of viral vectors such as adeno-associated virus (AAV) is crucial to overcome host transduction barriers observed during clinical gene therapy. We reasoned that exploring the role of cellular ubiquitin-like modifiers (UBLs) such as Neddylation or SUMOylation during AAV transduction could be beneficial. Using a combination of in silico biochemical and molecular engineering strategies, we have studied the impact of these UBLs during AAV2 infection and further developed Neddylation or SUMOylation site-modified AAV vectors and validated them in multiple disease models in vitro and in vivo. Hepatic gene transfer of two novel vectors developed, K105Q (SUMOylation-site mutant) and K665Q (Neddylation-site mutant), demonstrated a significantly improved human coagulation factor (F) IX expression (up to two-fold) in a murine model of hemophilia B. Furthermore, subretinal gene transfer of AAV2-K105Q vector expressing RPE65 gene demonstrated visual correction in a murine model of a retinal degenerative disease (rd12 mice). These vectors did not have any adverse immunogenic events in vivo. Taken together, we demonstrate that gene delivery vectors specifically engineered at UBLs can improve the therapeutic outcome during AAV-mediated ocular or hepatic gene therapy.

Keywords: AAV; Neddylation; SUMOylation; capsid; hemophilia; rational engineering; retinal degeneration.

Figure 1.

Transduction efficiency of AAV2 mutant vectors in vitro. About 3 × 10⁴ Huh7, ARPE19, or HeLa cells were mock-infected or infected with scAAV2-EGFP and scAAV2-EGFP mutant vectors for 3 h. Forty-eight hours later, the transgene (GFP) expression was measured by flow cytometry (CyFlow, Sysmex-Partec, Kobe, HP, Japan). Quantitative data for Neddylation or SUMOylation target-site mutants (i) and representative histograms (ii, iii) in Huh7 cells (a), ARPE19 cells (b), or HeLa cells (c) are shown. scAAV2-EGFP control was shared between Neddylation and SUMOylation mutants. Dunnett's multiple comparisons test was used to determine the statistical significance. Data are expressed as mean ± SD, n = 4, ***p ≤ 0.001. Data presented are a representative set from two independent biological replicate analysis. Values on histograms (ii, iii) have been manually shown for one replicate sample from entire analysis. The MFI (mean fluorescence intensity) for the same samples are shown in Supplementary Fig. S3. AAV, adeno-associated virus; ARPE, adult retinal pigmental epithelium; EGFP, enhanced green fluorescent protein; sc, self-complementary. Color images are available online.

Figure 1.

Transduction efficiency of AAV2 mutant vectors in vitro. About 3 × 10⁴ Huh7, ARPE19, or HeLa cells were mock-infected or infected with scAAV2-EGFP and scAAV2-EGFP mutant vectors for 3 h. Forty-eight hours later, the transgene (GFP) expression was measured by flow cytometry (CyFlow, Sysmex-Partec, Kobe, HP, Japan). Quantitative data for Neddylation or SUMOylation target-site mutants (i) and representative histograms (ii, iii) in Huh7 cells (a), ARPE19 cells (b), or HeLa cells (c) are shown. scAAV2-EGFP control was shared between Neddylation and SUMOylation mutants. Dunnett's multiple comparisons test was used to determine the statistical significance. Data are expressed as mean ± SD, n = 4, ***p ≤ 0.001. Data presented are a representative set from two independent biological replicate analysis. Values on histograms (ii, iii) have been manually shown for one replicate sample from entire analysis. The MFI (mean fluorescence intensity) for the same samples are shown in Supplementary Fig. S3. AAV, adeno-associated virus; ARPE, adult retinal pigmental epithelium; EGFP, enhanced green fluorescent protein; sc, self-complementary. Color images are available online.

Figure 1.

Transduction efficiency of AAV2 mutant vectors in vitro. About 3 × 10⁴ Huh7, ARPE19, or HeLa cells were mock-infected or infected with scAAV2-EGFP and scAAV2-EGFP mutant vectors for 3 h. Forty-eight hours later, the transgene (GFP) expression was measured by flow cytometry (CyFlow, Sysmex-Partec, Kobe, HP, Japan). Quantitative data for Neddylation or SUMOylation target-site mutants (i) and representative histograms (ii, iii) in Huh7 cells (a), ARPE19 cells (b), or HeLa cells (c) are shown. scAAV2-EGFP control was shared between Neddylation and SUMOylation mutants. Dunnett's multiple comparisons test was used to determine the statistical significance. Data are expressed as mean ± SD, n = 4, ***p ≤ 0.001. Data presented are a representative set from two independent biological replicate analysis. Values on histograms (ii, iii) have been manually shown for one replicate sample from entire analysis. The MFI (mean fluorescence intensity) for the same samples are shown in Supplementary Fig. S3. AAV, adeno-associated virus; ARPE, adult retinal pigmental epithelium; EGFP, enhanced green fluorescent protein; sc, self-complementary. Color images are available online.

Figure 2.

Western blot analysis of AAV2 vectors. About 1.42 × 10¹⁰ vgs of AAV2 and AAV2 K105Q vectors were resolved by denaturing SDS-PAGE. The level of SUMO-1 protein on vector capsids (a) were quantified (c) as described in the Materials and Methods section. Anti-AAV capsid B1 antibody was used as a loading control (b). The field of view pertaining to loaded samples from the entire gel is shown in the image. Exposure time for SUMO-1 and B1 antibody immune-reactive blots was 2 min 49 s and 36 s, respectively. Data presented are representative set from three independent biological replicates. Paired t-test was performed to determine the statistical significance. Data are expressed as mean ± SD, n = 3, ***p ≤ 0.001. SDS-PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; SUMO, small ubiquitin-like modifier; vgs, vector genomes. Color images are available online.

Figure 3.

Efficiency of Neddylation or SUMOylation site–modified AAV2 vectors for gene transfer into hepatic cells in vitro or in a murine model of hemophilia B. (a) Human factor IX (h.FIX) transcript levels assessed by quantitative PCR from Huh7 cells infected with scAAV2-h.FIX wild type or mutant vectors are shown. (b) Levels of h.FIX in plasma were determined 5 and 8 weeks after injection of 5 × 10¹⁰ vgs of AAV2 vectors per animal. Dunnett's multiple comparisons test was used to determine the statistical significance. Data are expressed as mean ± SD (n = 5 animals per experimental group). ***p ≤ 0.001, *p ≤ 0.05. PCR, polymerase chain reaction. Color images are available online.

Figure 4.

Immunohistochemistry for human factor IX after hepatic gene transfer in vivo. Human factor IX (h.FIX) expression was detected by fluorescence microscopy 9 weeks postinjection of 5 × 10¹⁰ vgs/animal of scAAV2 LP1 h.FIX (a), scAAV2 K105Q LP1 h.FIX (b), or scAAV2 K665Q LP1 h.FIX (c). Representative images are shown. Multiple bright nonspecific fluorescent spots were detected (marked by arrow) surrounding specific signals during imaging. Original magnification 400 × . Color images are available online.

Figure 5.

Ocular gene transfer in C57BL6/J mice with SUMOylation site–modified K105Q vector. A fundus imaging of murine eyes was performed 4 weeks after administration of AAV2 wild-type and AAV2 K105Q vectors. For these experiments, we used a Micron IV imaging system (Phoenix Research Laboratories, Pleasanton, CA) that employs a standard mouse objective and a field of view at 50° (1.8 mm diameter). Intensity was set at maximum and gain was set at 15 db; the frame rate was set at 6 fps for imaging of all the groups. Image analysis was performed by using Concentric Circle Plugin in the ImageJ software, and AAV2 K105Q group had a 1.57-fold higher EGFP expression in comparison to AAV2 wild-type eyes (n = 4 to 7 eyes). Representative images are shown. Color images are available online.

Figure 6.

SUMOylation site–modified vectors expressing RPE65 demonstrate phenotypic rescue in rd12 mice model. Eyes of rd12 mice were mock-injected or injected with ssAAV2-RPE65 and ssAAV2K105Q-RPE65 vectors at a dose of 7 × 10⁸ vectors via subretinal route. Scotopic ERG recordings were performed 10 weeks postvector administration and their representative image (a) and quantification data (b) are shown. Completely opaque eyes caused by injury were eliminated from the recording data set. Dunnett's multiple comparisons test was used to determine the statistical significance. Data are expressed as mean ± SD (n = 11–13 eyes per experimental group). ***p ≤ 0.001. ERG, electroretinogram. Color images are available online.