Gene Editing for Huntington’s Disease Shows Promise in Pigs

Researchers have known since 1993 that Huntington’s disease is caused by a repeat expansion in the huntingtin gene, yet figuring out how to correct this defect has not been easy. One approach, suppressing expression of the mutant protein with antisense oligonucleotides, has so far been unsuccessful in trials. In the February 16 Nature Biomedical Engineering, researchers led by Shihua Li and Xiao-Jiang at Jinan University, and Liangxue Lai at the Chinese Academy of Sciences, both in Guangzhou, China, make a case for correcting the gene via CRISPR editing. In a pig model of HD, a single brain injection of a viral vector carrying the gene-editing molecules cut the amount of mutant protein in half, and improved the animal’s gait months later. The strategy deserves further study, the authors suggest.

  • Mini-pigs with mutant huntingtin knocked in provide a large-mammal model of HD.
  • Correcting the mutant gene with CRISPR improved their motor abilities.
  • But CRISPR also made off-target genetic changes.

Claire Clelland at the University of California, San Francisco, agreed. “One major question is whether CRISPR editing, which looks promising in mice, will scale to the human brain, which is bigger and anatomically more complex. These data demonstrating a functional motor benefit in adult pigs after gene therapy are very promising,” she wrote to Alzforum (full comment below).

Olivier Danos at biotech REGENXBIO in New York City also called the approach promising, and considers the pig a good model for the human brain and its pathology. However, he said more studies are needed to evaluate the efficacy and safety of gene repair and to make sure gene editing did not introduce translocations or other chromosomal rearrangements.

A Better Model? Newborn piglets carrying a knocked-in mutant huntingtin gene. [Courtesy of Yan et al., Cell, 2018.]

The authors previously generated this animal model by replacing endogenous huntingtin in Chinese miniature pigs with a version containing 150 CAG repeats. The knock-in pigs lost striatal neurons, developed movement problems, and died young, around 1 year of age (Yan et al., 2018). The normal lifespan of these mini-pigs is 15 years.

To see if the mutant gene could be corrected, first author Sen Yan at Jinan University injected an adenoviral vector carrying CRISPR/Cas9 into the striatum of six knock-in pigs that were 3 months old, the age when they first show symptoms. The gene editing tools snipped out the repeat expansion, replacing it with a 20-CAG sequence. This is different than the ASO approach, which suppresses mutant huntingtin expression but does not correct the mutation to restore normal expression.

The CRISPR injection lowered the concentration of mutant huntingtin protein in the striatum by half, as seen by Western blots performed postmortem. At 7 months of age, treated pigs walked and ran better than untreated controls. They also lived a bit longer, with one surviving to two years before dying from HD complications.

Straight and Narrow. Pigs treated with gene therapy (left) walked more efficiently than did untreated controls (right). [Courtesy of Yan et al., Nat Biomed Eng, 2023.]

The authors achieved similar results when they delivered the adenoviral vector into newborn pigs intravenously through a vein in their ear. The vector was broadly distributed through the brain. Again, the treatment cut mutant huntingtin by more than half, and improved gait.

Brain Entry. Coronal brain section from a 4-month-old pig that had been injected as a newborn with AAV guide RNA viruses that also express red fluorescent protein. Viral expression was widespread. [Courtesy of Yan et al., Nat Biomed Eng, 2023.]

Alas, the technology did not work exactly as intended. The scientists sequenced DNA in striatal samples from each pig. This revealed that, by either administration route, gene editing had created an insertion or deletion in the huntingtin gene that disrupted protein expression, rather than replacing the gene with a healthy version, in about two-thirds of cells. In essence, CRISPR had behaved functionally like an ASO. The desired targeted replacement occurred in about 10 percent of cells.

The researchers also found rare off-target genetic changes, with about a 0.5 percent increase in single nucleotide and structural variants compared to the wild-type genome. Other studies have found that CRISPR can introduce serious errors (Jul 2018 news).

Clelland said the off-target effects could be a concern. “This [low percentage] could still be thousands of genetic changes,” she noted. “The field needs better methods to find and quantify off-targets, and metrics to determine an off-target profile that will be safe and acceptable to patients.”

CRISPR technology has entered clinical trials, such as Regeneron’s Phase 1 study in peripheral transthyretin amyloidosis.—Madolyn Bowman Rogers

Comments

  1. This study is important because it studied the effects of CRISPR machinery delivery and editing in the pig brain. One major question is whether CRISPR editing, which looks promising in mice, will scale to the human brain, which is bigger and anatomically more complex. These data, demonstrating a functional motor benefit in adult pigs after gene therapy, are very promising.

    Limitations and open questions are:

    Is there a survival benefit? A sample size of six with approximately 25 percent improved survival translates to 1.5 surviving animals (compared to zero in the untreated group). It is hard to know if this is meaningful, especially without statistical analysis.

    The treatment in adult pigs led to inflammation, and it is impossible to tell if this is due to expression of the AAV vector, expression of nonhuman protein (Cas9 or GFP), or a combination of vector and nonhuman protein. I would not conclude that continual expression of Cas protein is safe based on these data, but rather that the question needs to be directly studied. The long-term effect of the inflammation observed here warrants further study.

    Detecting and quantifying off-target genome editing is a major challenge for the field. The data here suggest there is less than 1 percent difference in off-target analysis, but this could still be thousands of genetic changes. The concern with AAV is that off-targets will accumulate over time given the Cas9 protein is continually expressed. AAVs also may have unique off-target integration of AAV DNA. This study attempted to address these concerns. Whether pig-genome off-targets of human-specific gRNAs are applicable to human patients or clinical trials is questionable. The field needs better methods to find and quantify off-targets, and metrics to determine an off-target profile that will be safe and acceptable to patients.

    Lastly, templated repair, in its current form, in adult CNS is unlikely to have an impact, given that most cells are post-mitotic and do not use templated repair to fix DNA breaks induced by Cas cutting. The gRNA strategy and data reported here suggest the results are more likely to be a result of knock-down of HTT expression by cutting within and thus disrupting a coding region of the HTT gene.

    View all comments by Claire Clelland
    • User Profile Image Nancy Ip
      Hong Kong University of Science & Technology
    • Posted:

    Huntington’s disease is an autosomal-dominant genetic disorder caused by excessive CAG repeats in the HTT gene. The misfolding and subsequent aggregation of mutant HTT protein is toxic to neurons, leading to motor and cognitive deficits. While the pathogenesis underlying HD is relatively clear, there are no effective disease-modifying treatments. In this study, the authors used an adeno-associated virus (AAV)-mediated CRISPR/Cas9 system to replace the expanded CAG repeats and lower the expression of mHTT protein in a huntingtin knock-in pig model. The paper demonstrates that both intracranial delivery (in the adult stage) and intravenous delivery (in the neonatal stage) of the AAV–Cas9 system can reduce HD-associated pathologies and alleviate motor function deficits. The authors also demonstrate that this AAV–Cas9 system is safe, as it has no obvious off-target effects and no significant immune response, indicating its promising potential to be developed for HD therapeutics.

    This work represents a significant step forward from the studies conducted previously by the same group, which demonstrated that the AAV–Cas9 system ameliorates neurotoxicity in an HD mouse model (Yang et al., 2017) and that a huntingtin knock-in pig model recapitulates the overt and selective neurodegeneration observed in the brains of HD patients (Yan et al., 2018). Given that large animals, like pigs, are more similar to humans in brain size and structure when compared with mice, using CRISPR/Cas9-mediated gene editing on large animals is an important contribution to the development of CRISPR/Cas9-based treatment strategies for brain disorders. It provides a strong foundation and intriguing insights for future clinical translation.

    Accumulating evidence suggests that lowering mHTT protein expression reduces HD-related pathologies and could be a feasible therapeutic for HD (Yang et al., 2017; Southwell et al., 2018). Besides CRISPR/Cas9-based gene-editing technology, other gene therapy approaches can also reduce the mHTT protein. For instance, antisense oligonucleotides (ASOs) that are complementary to HTT mRNA can trigger RNase-mediated degradation of these transcripts and reduce toxic HTT levels. Compared to ASO, gene editing has several advantages. First, the effect of gene editing is “once for all”—that is, theoretically, it can last a lifetime. In contrast, ASO-based therapies require repeated injections. Second, this study demonstrated targeted replacement of excessive CAG repeats with a normal repeat number, which can restore the normal function of HTT protein, whereas ASOs only reduce mHTT protein levels. Therefore, gene editing may be a more feasible approach for HD treatment.

    While this study demonstrated the possibility of replacing expanded CAG repeats, the efficiency of the CAG repeat replacement in generating the small HTT DNA product was low (4 percent–14 percent of the total analyzed DNA sequences in each pig). Most of the editing events generated random insertion or deletion (60 percent–69 percent), which resulted in reduced mHTT expression. Therefore, it is important to increase the replacement rate. Also, notably, the authors found that AAV infection (rather than expression of Cas9 protein) can cause inflammation in pigs. Therefore, less immunogenic delivery vehicles are needed.

    Overall, this elegant study demonstrates the efficacy and safety of gene editing in a large animal model of HD, which is a remarkable progress in the development of gene editing as a therapeutic approach for HD and other neurodegenerative diseases.

    References:

    Southwell AL, Kordasiewicz HB, Langbehn D, Skotte NH, Parsons MP, Villanueva EB, Caron NS, Østergaard ME, Anderson LM, Xie Y, Cengio LD, Findlay-Black H, Doty CN, Fitsimmons B, Swayze EE, Seth PP, Raymond LA, Frank Bennett C, Hayden MR. Huntingtin suppression restores cognitive function in a mouse model of Huntington's disease. Sci Transl Med. 2018 Oct 3;10(461) PubMed.

    Yan S, Tu Z, Liu Z, Fan N, Yang H, Yang S, Yang W, Zhao Y, Ouyang Z, Lai C, Yang H, Li L, Liu Q, Shi H, Xu G, Zhao H, Wei H, Pei Z, Li S, Lai L, Li XJ. A Huntingtin Knockin Pig Model Recapitulates Features of Selective Neurodegeneration in Huntington's Disease. Cell. 2018 May 3;173(4):989-1002.e13. Epub 2018 Mar 29 PubMed.

    Yang S, Chang R, Yang H, Zhao T, Hong Y, Kong HE, Sun X, Qin Z, Jin P, Li S, Li XJ. CRISPR/Cas9-mediated gene editing ameliorates neurotoxicity in mouse model of Huntington's disease. J Clin Invest. 2017 Jun 30;127(7):2719-2724. Epub 2017 Jun 19 PubMed.

    View all comments by Nancy Ip

  3. This work represents a step forward toward a gene therapy for Huntington’s disease. This disorder is caused by a trinucleotide repeat expansion of a “CAG” codon in HTT, leading to a dominantly inherited disease with 100 percent penetrance. In the study, the authors assessed CRISPR-mediated gene editing as a strategy to target mutant HTT in a knock-in pig model. By taking a homology-directed repair (HDR)-based approach, they aimed at replacing the poly CAG expansion region of HTT exon 1 with its shorter, non-mutated version. They packaged CRISPR components along with a repair template in adeno-associated viral vectors (AAVs) and injected the pigs directly into their brain, or intravenously. With both of these strategies they observed a robust reduction of the mutant huntingtin protein in the brain.

    However, most of the effects were instead due to a non-homologous end joining (NHEJ)-based disruption of the mutated alleles, causing decreased overall levels of huntingtin. Intriguingly, motor-related features of this pig model were partially normalized with both modes of administering CRISPR/Cas9. Also, no apparent off-target effects were observed in the treated pigs, and the immunogenic changes observed seemed to have been caused by the virus, not the CRISPR components.

    A precise editing effect by HDR is hard to achieve in neuronal cells as they are not readily undergoing mitosis. Accordingly, the limited frequency of precisely edited cells in this study could instead be ascribed to the targeting of glial cells, which are more prone to divide. However, in terms of treatment effects, it may not matter so much whether the major effect is related to NHEJ or HDR, as a loss of huntingtin in the adult brain probably will not have detrimental consequences.

    For the peripherally administered treatment, the pigs received intravenous injections at a neonatal stage, before their blood-brain barrier was fully developed. This allowed for a robust CNS effect with a relatively low number of AAV9 capsids--1.8x1013 vg per animal for direct injections, or 2x1013 vg per kg for intravenous injection. In a previous study, a much higher dose, 1x1014 vg per animal, of the highly blood-brain-barrier-penetrable AAV 9 PHP.eB subtype, was used for efficient CRISPR/CAS 9 treatment effects on much smaller transgenic APP mice (Duan et al., 2022). A recently approved AAV vector for intravenous administration to treat spinal muscular atrophy also uses a considerably higher dose of 1.1x1014 vg per kg.  

    Whereas CRISPR/Cas9 seems to offer an excellent mode of action to either disrupt or repair huntingtin in affected patients, there are several limiting factors related to drug delivery. If the intravenous administration were to be translated into a functioning therapy on humans, the dose would have to be greatly upscaled. Moreover, editing will occur in several organs—which is especially concerning since this study shows that the reproductive organs also are affected. Local delivery strategies offer alternative options, but direct intracranial injections may be considered too invasive and are unlikely to be sufficient for delivery to all brain areas in need. Instead, vector delivery to the cerebrospinal fluid via intrathecal injections, an approach that has already been validated for nucleic acid therapies (i.e., nusinersen in SMA), may represent the most feasible strategy and should be explored in this context.

    Taken together, this work presents compelling evidence that gene editing can translate to a motor benefit in a large animal model of Huntington’s disease, but further studies are needed to translate this promising approach to humans.

    References:

    Duan Y, Ye T, Qu Z, Chen Y, Miranda A, Zhou X, Lok KC, Chen Y, Fu AK, Gradinaru V, Ip NY. Brain-wide Cas9-mediated cleavage of a gene causing familial Alzheimer's disease alleviates amyloid-related pathologies in mice. Nat Biomed Eng. 2021 Jul 26; PubMed.

    View all comments by Martin Ingelsson View all comments by Bence György

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. CRISPR Gene Editing May Not Be so Precise After All

Paper Citations

  1. Yan S, Tu Z, Liu Z, Fan N, Yang H, Yang S, Yang W, Zhao Y, Ouyang Z, Lai C, Yang H, Li L, Liu Q, Shi H, Xu G, Zhao H, Wei H, Pei Z, Li S, Lai L, Li XJ. A Huntingtin Knockin Pig Model Recapitulates Features of Selective Neurodegeneration in Huntington's Disease. Cell. 2018 May 3;173(4):989-1002.e13. Epub 2018 Mar 29 PubMed.

External Citations

  1. Phase 1

Further Reading

Primary Papers

  1. Yan S, Zheng X, Lin Y, Li C, Liu Z, Li J, Tu Z, Zhao Y, Huang C, Chen Y, Li J, Song X, Han B, Wang W, Liang W, Lai L, Li XJ, Li S. Cas9-mediated replacement of expanded CAG repeats in a pig model of Huntington's disease. Nat Biomed Eng. 2023 May;7(5):629-646. Epub 2023 Feb 16 PubMed.

Annotate

To make an annotation you must Login or Register.