Enhancement of therapeutic transgene insertion for treatment of murine phenylketonuria

Abstract

Low in vivo transgene integration frequency limits the therapeutic efficacy of homology-directed repair (HDR) as a treatment for genetic disorders. This study demonstrates improved efficacy of HDR-mediated gene insertion for the treatment of murine phenylalanine hydroxylase (PAH) deficiency, a model of human phenylketonuria (PKU), through pharmacologic inhibition of competing DNA repair pathways. The targeted integration of a Pah-expressing transgene was enhanced with vanillin, a potent inhibitor of non-homologous end joining (NHEJ), reducing mean serum phenylalanine concentrations by 56.8% in treated mice. This was further improved following co-inhibition of NHEJ and microhomology-mediated end joining (MMEJ), yielding transgene insertions in approximately 10% of genomes with an associated 70.6% decrease in serum phenylalanine. Phenylalanine concentrations were further reduced to 392 μM after oral administration of sapropterin dihydrochloride, a pharmacologic cofactor of PAH. Separately, we demonstrate that rare hepatocytes harboring transgene insertions were successfully expanded to a therapeutically relevant population through selection based upon resistance to acetaminophen toxicity, but this method was hampered by adverse effects upon AAV vector production and the neurologic function of treated neonatal mice related to the presence of shRNA sequences in the AAV vector. These results demonstrate that pharmacologic inhibition of alternative DNA repair pathways can significantly enhance HDR-mediated transgene insertion in vivo.

Keywords: CRISPR/Cas9; MT: RNA/DNA Editing; adeno-associated virus vectors; gene editing; gene therapy; homologous recombination; novobiocin; phenylalanine hydroxylase; phenylketonuria; vanillin.

Graphical abstract

Figure 1

Experimental design for HR-mediated gene editing treatment of murine PAH deficiency (A) Canonical model of DSB repair by non-homologous end joining (NHEJ), homologous recombination (HR), and microhomology-mediated end joining (MMEJ). Inhibition of DNA-PKcs and polymerase theta, vital factors for NHEJ and MMEJ, is hypothesized to increase frequency of repair through HR. (B) Neonatal mice at neonatal day 3 (P3) received dual AAV8 vectors through facial vein injection and daily i.p. injections of NHEJ or MMEJ inhibiting small molecules. Dual AAV8 treatment is composed of a streptococcus pyogenes Cas9 (spCas9) vector expressed from a liver specific promoter (LSP) and a repair template with 5′ and 3' homology arms (HA) for integration into target genomic locus. Single guide RNA (sgRNA) is delivered on the repair template, therefore requiring transduction with both vectors to induce a double-strand break (DSB). Episomal expression of murine PAH (mPAH) dissipates during hepatocyte turnover, while expression of the repair template after integration into the genome remains.

Figure 2

Vanillin-assisted insertion of Pah transgene in a murine model of PKU (A) Schematic of dual-AAV vectors including 5′ and 3' homology arms (HA), liver-specific promoter (LSP) composed of a bikunin enhancer and human thyroxine binding globulin (TBG) promoter, murine Pah cDNA, HGH poly(A) signal (PolyA), U6 promoter (U6), and sgRNA guide for targeted Cas9 cleavage (Guide 1). Red arrows indicate forward and reverse primers for qPCR. (B) Serum Phe concentrations in Dexon1 mice (n = 14) treated with dual AAV8 Cas9 and Pah repair template (mean = 917 ± 342 μM) versus control mice that received no treatment (mean = 2,132 ± 172 μM). Treated animals received facial vein injections of viral vectors (8.2 × 10¹³ vg/kg, 1:1 ratio) at day P3 with subsequent intraperitoneal injections of 100 mg/kg vanillin for 5 days following viral transduction. (C) PAH enzyme activity in treated (mean = 7.66% ± 3.17%) versus untreated (mean = 0.72% ± 0.45%) animals measured as percentage of wild-type PAH activity. (D) Anti-PAH (red, cytoplasmic), anti-lectin (green, cell membrane), and DAPI (blue, nuclear) of livers of treated (bottom) and untreated (top right) Pah^{Δexon1/Δexon1} animals as well as untreated C57BL/6 mice (top left). Scale bars, 100 μm. (E) Percentage of hepatocytes expressing PAH protein based on immunostaining. (F) Percentage of hepatocytes with Pah transgene insertions (mean = 3.76% ± 2.07%). Data are shown as means ± SD. ∗∗∗∗p < 0.0001 by unpaired t test.

Figure 3

Insertion of the transgene into Pah Exon 7 (A) Schematic of the repair template used to insert murine Pah cDNA into the Pah Exon 7 locus. Expression of sgRNA guide to this locus (Guide 7) under the expression of a U6 promoter (U6) and the adaptation of the 5′ and 3' homology arms to the target locus are the major differences between Figure 2A. Animals were injected with a viral dose of 1.0 × 10¹⁴ vg/kg per vector and the same vanillin regime as in Figure 1. (B) Serum Phe concentrations in animals treated with both repair template (RT) and Cas9 carrying viral vectors and vanillin (Van) (mean = 898 ± 324 μM, n = 10) versus those treated with RT and Cas9 without vanillin (mean = 1702 ± 149 μM, n = 5) and those only treated with the RT (mean = 2026 ± 232 μM, n = 4). (C) PAH enzyme activity in RT + Cas9 + Van (mean = 6.57% ± 2.11%), RT + Cas9 (mean = 0.21% ± 0.11%), and RT (mean = 0.17% ± 0.12%) treated animals. (D) Anti-Pah (red, cytoplasmic), anti-lectin (green, cell membrane), and DAPI (blue, nuclear) staining in liver section from RT + Cas9 treated animal. Scale bars, 100 μm. (E) Insertion frequency of the Pah cDNA in RT + Cas9 + Van (mean = 7.44% ± 2.49%), RT + Cas9 (mean = 0.25% ± 0.09%), and RT (mean = 0.01% ± 0.00%)-treated animals, as determined by qPCR. ∗∗∗∗p < 0.0001 by unpaired t-test. Data are shown as means ± SD.

Figure 4

Inhibition of NHEJ and MMEJ increases efficacy of homology-directed recombination in vivo (A) Serum Phe concentrations in untreated mice (mean = 2,134 ± 197 μM, n = 6) and animals treated with Cas9 and RT AAV vectors with Vanbiocin (vanillin 100 mg/kg and novobiocin 50mg/kg) (mean = 631 ± 211 μM, n = 6). Animals received Vanbiocin treatment at 5 days of age for 3 days. Black dashed line indicates a therapeutic threshold of 360 μM. Data are shown as means ± SD. ∗∗∗∗p < 0.0001 by paired t test. (B) Insertion frequency of the mPah cDNA (mean = 9.71% ± 3.62%) in treated animals, as determined by qPCR. (C) PAH enzyme activity in Vanbiocin-treated mice (mean = 6.14% ± 1.15%) measured as percentage of wild-type PAH activity. (D) Anti-PAH (Red), anti-lectin (green), and nuclear stain (blue) in treated animals. Scale bars, 100 μm. (E) Coat color rescue of wild-type C57BL/6, untreated Dexon1 mice, and Vanbiocin-treated mice. Baseline serum Phe concentrations without sapropterin provided. (F) Serum Phe concentrations in Vanbiocin-treated animals were measured over a 5-day course of daily treatment with 100 mg/kg sapropterin dihydrochloride, administered via oral gavage 6 h prior to retro-orbital serum collection. Black dashed line indicates a therapeutic threshold of 360 μM. p < 0.1188, ∗p < 0.0242, and ∗∗p < 0.0037 by one-way ANOVA. (G) Serum Phe concentrations in the progeny (mean = 2,124, n = 4) of Vanbiocin-treated female post-weaning. Dunnett’s multiple comparison used with all ANOVAs. Data are shown as means ± SD.

Figure 5

Providing Cypor-shRNA-induced selective advantage to PAH transgene under APAP exposure (A) Schematic of dual-AAV vectors including 5′ and 3' homology arms (HA), liver-specific promoter (LSP) including a bikunin enhancer and human TBG promoter, murine Pah cDNA, HGH poly(A) signal (PolyA), U6 promoter (U6), Cypor-targeting shRNA (shRNA), and sgRNA guide for targeted Cas9 cleavage (Guide 7). (B) Serum Phe concentrations over 18 weeks of APAP diet in male and female Dexon1 mice initiated after weaning. Dashed line indicates a therapeutic threshold of 360 μM. (C) Transgene insertion frequency in treated males that received APAP diet (9.55 ± 1.16%, n = 3) versus normal chow (0.04 ± 0.02%, n = 3), and females that received APAP diet (3.45 ± 2.48%, n = 4) versus normal chow (0.17 ± 0.20%, n = 4) (∗p < 0.0193 and ∗∗∗∗p < 0.0001). (D) Terminal serum Phe concentrations between APAP-treated males (425 ± 205 μM) and females (1309 ± 603 μM), as well as males (2165 ± 388 μM) and females (2524 ± 122 μM) that did not receive APAP selection but did receive viral vectors (∗∗∗p < 0.0005, and ∗∗p < 0.0026). (E) PAH enzyme activity in treated male (6.24% ± 1.64%), treated female (2.36% ± 1.93%), and untreated male (0.46% ± 0.32%) and female (0.68% ± 0.23%) animals after 18 weeks of APAP-induced hepatocyte selection (p < 0.1944 and ∗∗∗p < 0.0006). (F) Anti-CYPOR (red, cytoplasmic), anti-lectin (green, cell membrane), and DAPI (blue, nuclear) in treated animals. (G) Anti-PAH (red), anti-lectin (green), and nuclear stain (blue) in respective treated animals. Scale bars, 100 μm. Bonferroni’s multiple comparison test was used to calculate the significance of group differences. Significance calculated by two-way ANOVA with sex and treatment as variables. Data are shown as means ± SD.