Converting cell types in vivo to alleviate Huntington’s Disease

gene therapy for Huntington Disease

Date: 28th February 2020

Huntington’s disease (HD) is a progressive brain disorder that is estimated to affect between five and seven people per 100,000 in western countries.  Now scientists have used gene therapy to convert and reprogramme striatal astrocytes into neurons that can replace ‘faulty’ neurons, partially rescuing motor functional deficits and increasing the survival of mouse HD models.

HD is caused by an unbalanced cytosine-adenine-guanine (CAG) repeat expansion in the huntingtin gene (HTT).  This leads to the production of an abnormally long version of the huntingtin protein, which is then cut into smaller, toxic fragments.  These toxic fragments bind together and accumulate in the particularly susceptible GABAergic medium spiny neurons (MSNs).  As MSNs represent approximately 95% of the total neurons within the human striatum the consequences are devastating.  The loss of these cells causes early neurodegeneration, and many of the symptoms of HD can be attributed to their loss.

There have been many efforts to alleviate the symptoms of HD, however treatments remain limited. Here a team from Pennsylvania State University, US, led by Gong Chen have used AAV-based gene therapy to co-express two neural transcription factors, NeuroD1 (ND1) and Dlx2, in a specific type of glial cell –  converting these cells in vivo into MSNs.

Glial cells are the most abundant cell types in the central nervous system and one type – astrocytes – make up ~30% of the cells in the mammalian CNS.  They surround every neuron in the brain and provide support and insulation, and thus make them an attractive choice of cell for conversion.

  • To track the astrocyte-converted neurons in the mouse brain, the team developed a Cre-FLEx (flip-excision) system.
  • One AAV contained a vector expressing Cre under the control of an astrocyte-specific promoter.  The other AAVs contained FLEx vectors containing inverted coding sequences of either fluorescent controls (mCherry), or ND1-mCherry, and/or Dlx2-mCherry. Only in the presence of Cre were the coding sequences in the FLEx vectors re-inverted allowing expression.
  • To test whether the system could drive the conversion of astrocytes into neurons in the striatum the team injected the AAV-Cre together with ND1 and Dlx2 AAVs into adult wild type mice brains.
  • 7 days post injection (dpi) viral infected cells (mCherry+) were identified as astrocytes and 81.5% contained both ND1 and Dlx2.  Neither transcription factors (TFs) were found in the surrounding neurons as expected at this stage.
  • In contrast, by 30 dpi most of the ND1 and Dlx2 signals were being co-expressed in neurons, with only a small percentage expressed in astrocytes, indicating that co-expression of ND1 and Dlx2 had converted striatal astrocytes into neurons.
  • The team wished to see whether the system could regenerate the GABAergic neurons in a mouse model of HD. The results demonstrated that the striatal astrocytes in the HD mouse brains could be converted into neurons with high efficiency. These neurons were functional and integrated into local synaptic circuits. Furthermore, axonal projection from MSNs which are severely disrupted in HD patients, were able to form in the converted MSNs.
  • Importantly, at 38 dpi, >93% of HD-mice that were injected with NeuroD1 plus Dlx2 were still alive whilst, only ~55% of the control mice were alive as expected. They also displayed significant improvements in motor function, and other characteristic symptoms of HD.

astrocyte cell conversion

Conclusion and future applications

The work here has demonstrated that striatal astrocytes in HD-mouse models can be converted into GABAergic neurons, with similar electrophysical properties to neighbouring pre-existing neurons, and subsequently alleviated motor functional deficits. This data showed promising results, especially with respect to the significant increase in life span, indicating that NeuroD1 + Dlx2 gene therapy treatment could offer potential disease-modifying therapy for HD sufferers.

However, whilst this therapy may alleviate symptoms it does not treat the root cause of the disease.  There are other therapies currently being explored that do address this issue such as UniQure’s gene therapy AMT-130 which inhibits production of the mutated form of the huntingtin protein.  In preclinical trials (mouse models) it has shown promising results,  and clinical trials have now been approved by the FDA and currently are in the recruitment phase.  However, there are still improvements to be made as not all parameters of the disease were improved.

Other attempts such as CRISPR have also been used to edit the mutations that occur in HD, however, off-target effects and the requirement to target the brain both from an ethical and practical perspective have not as yet yielded any promising translatable therapies.

There are of course limitations to in vivo cell conversion technology. It requires the presence of glial cells which may be limited in number in a disease environment, and as the mutations causing the disease are still present, newly reprogrammed neurons may suffer the same fate as the old ones (although this was shown to occur at a lower rate in this paper).

One potential powerful approach would be to combine this in vivo cell conversion therapy together with a CRISPR-based treatment.  CRISPR-gene editing to correct the mutation so that the newly converted neurons are able to survive.  This may provide a more comprehensive defence against the disease in a halt, repair and regenerate tactic.

Wherever, these next generation treatments take us, either in combination or alone, there is now an increased hope on the horizon for HD sufferers and indeed for those with other neurodegenerative disorders.


Wu, Z., M. Parry, X.-Y. Hou, M.-H. Liu, H. Wang, R. Cain, Z.-F. Pei, Y.-C. Chen, Z.-Y. Guo, S. Abhijeet and G. Chen (2020). “Gene therapy conversion of striatal astrocytes into GABAergic neurons in mouse models of Huntington’s disease.” Nature Communications 11(1): 1105.