Florie Borel

Recombinant AAV as a Platform for Translating the Therapeutic Potential of RNA Interference

RNA interference has become a ubiquitous biological tool, and is being harnessed for therapeutic purposes as well. Therapeutic posttranscriptional gene silencing takes advantage of the endogenous RNAi pathway through delivery of either chemically synthesized siRNAs, or transgenes expressing hairpin-based inhibitory RNAs (e.g., shRNAs and artificial miRNAs). RNAi has expanded the field of viral gene therapy from gene replacement to gene knockdown. Here, we review various noncoding RNAs such as shRNAs, miRNAs, and miRNA decoys which can be utilized for therapeutic applications when expressed from recombinant adeno-associated vectors (AAV), and present examples of their basic design. In addition the basis of exploiting cellular miRNA profiles for detargeting AAV expression from specific cells is described. Finally, an overview of AAV-mediated RNAi preclinical studies is presented, and current RNAi-based clinical trials are reviewed.

Apolipoprotein B Knockdown by AAV-delivered shRNA Lowers Plasma Cholesterol in Mice

Serum low-density lipoprotein cholesterol (LDL-C) levels are proportionate to the risk of atherosclerotic cardiovascular disease. In order to reduce serum total cholesterol and LDL-C levels in mice, RNA interference (RNAi) was used to inhibit expression of the structural protein of LDL-C, apolipoprotein B100 (ApoB). We developed and screened 19 short hairpin RNAs (shRNAs) targeting conserved sequences in human, mouse, and macaque ApoB mRNAs (shApoB) and subsequently narrowed our focus to one candidate for in vivo testing. Self-complementary adeno-associated virus serotype 8 (scAAV8) was used for long-term transduction of murine liver with shApoB. A strong dose-dependent knockdown of ApoB mRNA and protein was observed, which correlated with a reduction in total cholesterol levels, without obvious signs of toxicity. Furthermore, shApoB was found to specifically reduce LDL-C in diet-induced dyslipidemic mice, whereas high-density lipoprotein cholesterol (HDL-C) remained unaffected. Finally, elevated lipid accumulation was shown in murine liver transduced with shApoB, a known phenotypic side effect of lowering ApoB levels. These results demonstrate a robust dose-dependent knockdown of ApoB by AAV-delivered shRNA in murine liver, thus providing an excellent candidate for development of RNAi-based gene therapy for the treatment of hypercholesterolemia.

Therapeutic rAAVrh10 Mediated SOD1 Silencing in Adult SOD1(G93A) Mice and Nonhuman Primates.

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease; survival in ALS is typically 3–5 years. No treatment extends patient survival by more than three months. Approximately 20% of familial ALS and 1–3% of sporadic ALS patients carry a mutation in the gene encoding superoxide dismutase 1 (SOD1). In a transgenic ALS mouse model expressing the mutant SOD1G93A protein, silencing the SOD1 gene prolongs survival. One study reports a therapeutic effect of silencing the SOD1 gene in systemically treated adult ALS mice; this was achieved with a short hairpin RNA, a silencing molecule that has raised multiple safety concerns, and recombinant adeno-associated virus (rAAV) 9. We report here a silencing method based on an artificial microRNA termed miR-SOD1 systemically delivered using adeno-associated virus rAAVrh10, a serotype with a demonstrated safety profile in CNS clinical trials. Silencing of SOD1 in adult SOD1G93A transgenic mice with this construct profoundly delayed both disease onset and death in the SOD1G93A mice, and significantly preserved muscle strength and motor and respiratory functions. We also document that intrathecal delivery of the same rAAVrh10-miR-SOD1 in nonhuman primates significantly and safely silences SOD1 in lower motor neurons. This study supports the view that rAAVrh10-miR-SOD1 merits further development for the treatment of SOD1-linked ALS in humans.

Embedding siRNA sequences targeting Apolipoprotein B100 in shRNA and miRNA scaffolds results in differential processing and in vivo efficacy

Overexpression of short hairpin RNA (shRNA) often causes cytotoxicity and using microRNA (miRNA) scaffolds can circumvent this problem. In this study, identically predicted small interfering RNA (siRNA) sequences targeting apolipoprotein B100 (siApoB) were embedded in shRNA (shApoB) or miRNA (miApoB) scaffolds and a direct comparison of the processing and long-term in vivo efficacy was performed. Next generation sequencing of small RNAs originating from shApoB- or miApoB-transfected cells revealed substantial differences in processing, resulting in different siApoB length, 5′ and 3′ cleavage sites and abundance of the guide or passenger strands. Murine liver transduction with adeno-associated virus (AAV) vectors expressing shApoB or miApoB resulted in high levels of siApoB expression associated with strong decrease of plasma ApoB protein and cholesterol. Expression of miApoB from the liver-specific LP1 promoter was restricted to the liver, while the H1 promoter-expressed shApoB was ectopically present. Delivery of 1 × 1011 genome copies AAV-shApoB or AAV-miApoB led to a gradual loss of ApoB and plasma cholesterol inhibition, which was circumvented by delivering a 20-fold lower vector dose. In conclusion, incorporating identical siRNA sequences in shRNA or miRNA scaffolds results in differential processing patterns and in vivo efficacy that may have serious consequences for future RNAi-based therapeutics.

Optimization and comparison of knockdown efficacy between polymerase II expressed shRNA and artificial miRNA targeting luciferase and Apolipoprotein B100

BackgroundControlling and limiting the expression of short hairpin RNA (shRNA) by using constitutive or tissue-specific polymerase II (pol II) expression can be a promising strategy to avoid RNAi toxicity. However, to date detailed studies on requirements for effective pol II shRNA expression and processing are not available. We investigated the optimal structural configuration of shRNA molecules, namely: hairpin location, stem length and termination signal required for effective pol II expression and compared it with an alternative strategy of avoiding toxicity by using artificial microRNA (miRNA) scaffolds.ResultsHighly effective shRNAs targeting luciferase (shLuc) or Apolipoprotein B100 (shApoB1 and shApoB2) were placed under the control of the pol II CMV promoter and expressed at +5 or +6 nucleotides (nt) with reference to the transcription start site (TSS). Different transcription termination signals (TTS), namely minimal polyadenylation (pA), poly T (T5) and U1 were also used. All pol II- expressed shRNA variants induced mild inhibition of Luciferase reporters carrying specific targets and none of them showed comparable efficacy to their polymerase III-expressed H1-shRNA controls, regardless of hairpin position and termination signal used. Extending hairpin stem length from 20 basepairs (bp) to 21, 25 or 29 bp yielded only slight improvement in the overall efficacy. When shLuc, shApoB1 and shApoB2 were placed in an artificial miRNA scaffold, two out of three were as potent as the H1-shRNA controls. Quantification of small interfering RNA (siRNA) molecules showed that the artificial miRNA constructs expressed less molecules than H1-shRNAs and that CMV-shRNA expressed the lowest amount of siRNA molecules suggesting that RNAi processing in this case is least effective. Furthermore, CMV-miApoB1 and CMV-miApoB2 were as effective as the corresponding H1-shApoB1 and H1-shApoB2 in inhibiting endogenous ApoB mRNA.ConclusionOur results demonstrate that artificial miRNA have a better efficacy profile than shRNA expressed either from H1 or CMV promoter and will be used in the future for RNAi therapeutic development.

5 Year Expression and Neutrophil Defect Repair after Gene Therapy in Alpha-1 Antitrypsin Deficiency

Alpha-1 antitrypsin deficiency is a monogenic disorder resulting in emphysema due principally to the unopposed effects of neutrophil elastase. We previously reported achieving plasma wild-type alpha-1 antitrypsin concentrations at 2.5%–3.8% of the purported therapeutic level at 1 year after a single intramuscular administration of recombinant adeno-associated virus serotype 1 alpha-1 antitrypsin vector in alpha-1 antitrypsin deficient patients. We analyzed blood and muscle for alpha-1 antitrypsin expression and immune cell response. We also assayed previously reported markers of neutrophil function known to be altered in alpha-1 antitrypsin deficient patients. Here, we report sustained expression at 2.0%–2.5% of the target level from years 1–5 in these same patients without any additional recombinant adeno-associated virus serotype-1 alpha-1 antitrypsin vector administration. In addition, we observed partial correction of disease-associated neutrophil defects, including neutrophil elastase inhibition, markers of degranulation, and membrane-bound anti-neutrophil antibodies. There was also evidence of an active T regulatory cell response (similar to the 1 year data) and an exhausted cytotoxic T cell response to adeno-associated virus serotype-1 capsid. These findings suggest that muscle-based alpha-1 antitrypsin gene replacement is tolerogenic and that stable levels of M-AAT may exert beneficial neutrophil effects at lower concentrations than previously anticipated.

Survival Advantage of Both Human Hepatocyte Xenografts and Genome-Edited Hepatocytes for Treatment of α-1 Antitrypsin Deficiency

Hepatocytes represent an important target for gene therapy and editing of single-gene disorders. In α-1 antitrypsin (AAT) deficiency, one missense mutation results in impaired secretion of AAT. In most patients, lung damage occurs due to a lack of AAT-mediated protection of lung elastin from neutrophil elastase. In some patients, accumulation of misfolded PiZ mutant AAT protein triggers hepatocyte injury, leading to inflammation and cirrhosis. We hypothesized that correcting the Z mutant defect in hepatocytes would confer a selective advantage for repopulation of hepatocytes within an intact liver. A human PiZ allele was crossed onto an immune-deficient (NSG) strain to create a recipient strain (NSG-PiZ) for human hepatocyte xenotransplantation. Results indicate that NSG-PiZ recipients support heightened engraftment of normal human primary hepatocytes as compared with NSG recipients. This model can therefore be used to test hepatocyte cell therapies for AATD, but more broadly it serves as a simple, highly reproducible liver xenograft model. Finally, a promoterless adeno-associated virus (AAV) vector, expressing a wild-type AAT and a synthetic miRNA to silence the endogenous allele, was integrated into the albumin locus. This gene-editing approach leads to a selective advantage of edited hepatocytes, by silencing the mutant protein and augmenting normal AAT production, and improvement of the liver pathology.

Artificial miRNAs reduce human mutant Huntingtin throughout the striatum in a transgenic sheep model of Huntington’s disease

Huntingtons disease (HD) is a fatal neurodegenerative disease caused by a genetic expansion of the CAG repeat region in the huntingtin (HTT) gene. Studies in HD mouse models have shown that artificial miRNAs can reduce mutant HTT, but evidence for their effectiveness and safety in larger animals is lacking. HD transgenic sheep express the full-length human HTT with 73 CAG repeats. AAV9 was used to deliver unilaterally to HD sheep striatum an artificial miRNA targeting exon 48 of the human HTT mRNA under control of two alternative promoters: U6 or CβA. The treatment reduced human mutant (m) HTT mRNA and protein 50-80% in the striatum at 1 and 6 months post injection. Silencing was detectable in both the caudate and putamen. Levels of endogenous sheep HTT protein were not affected. There was no significant loss of neurons labeled by DARPP32 or NeuN at 6 months after treatment, and Iba1-positive microglia were detected at control levels. It is concluded that safe and effective silencing of human mHTT protein can be achieved and sustained in a large-animal brain by direct delivery of an AAV carrying an artificial miRNA.

Safe and Efficient Silencing with a Pol II, but Not a Pol lII, Promoter Expressing an Artificial miRNA Targeting Human Huntingtin

Huntington’s disease is a devastating, incurable neurodegenerative disease affecting up to 12 per 100,000 patients worldwide. The disease is caused by a mutation in the Huntingtin (Htt) gene. There is interest in reducing mutant Huntingtin by targeting it at the mRNA level, but the maximum tolerable dose and long-term effects of such a treatment are unknown. Using a self-complementary AAV9 vector, we delivered a mir-155-based artificial miRNA under the control of the chicken β-actin or human U6 promoter. In mouse brain, the artificial miRNA reduced the human huntingtin mRNA by 50%. The U6, but not the CβA promoter, produced the artificial miRNA at supraphysiologic levels. Embedding the antisense strand in a U6-mir-30 scaffold reduced expression of the antisense strand but increased the sense strand. In mice treated with scAAV9-U6-mir-155-HTT or scAAV9-CβA-mir-155-HTT, activated microglia were present around the injection site 1 month post-injection. Six months post-injection, mice treated with scAAV9-CβA-mir-155-HTT were indistinguishable from controls. Those that received scAAV9-U6-mir-155-HTT showed behavioral abnormalities and striatal damage. In conclusion, miRNA backbone and promoter can be used together to modulate expression levels and strand selection of artificial miRNAs, and in brain, the CβA promoter can provide an effective and safe dose of a human huntingtin miRNA.

Editing out five Serpina1 paralogs to create a mouse model of genetic emphysema

Significance Chronic obstructive pulmonary disease affects 10% of the worldwide population, and the leading genetic cause is a genetic disease, α-1 antitrypsin (AAT) deficiency. Humans have only one gene that codes for the AAT protein, but mice have up to six, which made it impossible for decades to create a mouse model of the disease. Here we succeeded in creating this mouse model using CRISPR technology to target all of the mouse genes at once. Importantly, this mouse model spontaneously develops lung disease and recapitulates many aspects of the human disease. We anticipate that this model will be highly relevant not only to the preclinical development of therapeutics for AAT deficiency, but also to emphysema and smoking research. Chronic obstructive pulmonary disease affects 10% of the worldwide population, and the leading genetic cause is α-1 antitrypsin (AAT) deficiency. Due to the complexity of the murine locus, which includes up to six Serpina1 paralogs, no genetic animal model of the disease has been successfully generated until now. Here we create a quintuple Serpina1a–e knockout using CRISPR/Cas9-mediated genome editing. The phenotype recapitulates the human disease phenotype, i.e., absence of hepatic and circulating AAT translates functionally to a reduced capacity to inhibit neutrophil elastase. With age, Serpina1 null mice develop emphysema spontaneously, which can be induced in younger mice by a lipopolysaccharide challenge. This mouse models not only AAT deficiency but also emphysema and is a relevant genetic model and not one based on developmental impairment of alveolarization or elastase administration. We anticipate that this unique model will be highly relevant not only to the preclinical development of therapeutics for AAT deficiency, but also to emphysema and smoking research.