Twenty years ago, researchers took fibroblasts from the skin of eight Alzheimer’s patients, engineered them to produce nerve-growth factor, and slid them into each volunteer’s basal forebrain. They hoped the neurotrophin would halt or slow the neurodegeneration that robbed them of their memories, indeed their lives. The gamble failed and since then, scientists have shown little zest for gene therapy in neurodegenerative disorders. That is changing. As evident at this year’s Society for Neuroscience conference, held October 19–23 in Chicago, gene therapy is back. Buoyed by success in treating spinal muscular atrophy in infants, scientists are flush with new ideas—and funding.

  • Gene therapy research is undergoing a revival.
  • Treatments for neurodegenerative disorders are in development.
  • Main concerns? Safety and cost.

What was once considered risky, expensive, and unlikely to succeed is now seen by many as risky, expensive—and quite likely to succeed. A growing number of scientists think gene-based therapies may have the best chance of slowing, or even preventing, neurodegeneration, especially for disorders caused by mutations in a single gene. SfN hosted a press briefing on gene therapy, plus many projects are active throughout the field beyond those showcased at the conference. There was no breaking clinical trial news at the annual meeting, but the scope and challenges of such therapies were outlined at the briefing moderated by Rush University’ s Jeff Kordower, Chicago, as well as a translational roundtable moderated by Asa Abeliovich, Columbia University, New York. Abeliovich recently co-founded Prevail Therapeutics, New York.

Going viral. Researchers are tweaking the capsid of adeno-associated viruses to optimize gene therapies for a multitude of disease. Shown here, AAV2.

From Zolgensma to Alzheimer’s?
If the failure of the nerve growth factor therapy tempered enthusiasm for gene therapy (Mar 2018 news), then the success of AVXS-101, aka Zolgensma, reignited it. Developed by scientists at Nationwide Children’s Hospital, Columbus, Ohio, and AveXis, Bannockburn, Illinois, AVXS-101 uses an adeno-associated virus to deliver billions of copies of the survival motor neuron 1 gene to the brain. A small pilot trial tested the therapy in babies with spinal muscular atrophy (SMA) Type 1, the severest form of this neurodevelopmental disease. Lacking functional SMN1, these infants face progressive muscle weakness. Most die before their second birthday; those who live need a ventilator to breathe.

In Phase 1, AVXS-101 dramatically improved motor function of 15 treated infants; all were living 20 months later when historical data predicted only one would survive. Twelve babies who received the highest dose grew stronger within months, most sitting independently and rolling over. They hit the highest score on a scale of motor function, whereas untreated babies deteriorated. By 20 months, two of the treated babies had begun to walk (Mendell et al., 2017). The Food and Drug Administration approved zolgensma in May 2019. At SfN in Chicago, Petra Kaufmann, AveXis, played videos of the first patients treated with AVXS-101. Some four years later, they are walking, running, and appear to be playing almost normally. A video of a little girl walking downstairs with nary a hint of having SMA Type I visibly moved the audience.

Scientists say it’s a game-changer. “It is really the tremendous success with SMA that has renewed interest in gene therapy,” said Clive Svendsen, Cedars-Sinai Regenerative Medicine Institute, Los Angeles. Speaking with Alzforum before SfN, Bart De Strooper, Dementia Research Institute, London, said the same. “The success in SMA patients of both gene therapy and antisense therapy has revived interest in the whole area,” De Strooper said. Nowadays, researchers tend to lump gene therapy and antisense therapy under one moniker, i.e., gene-based therapy. The SMA antisense therapy nusinersen also works in babies with SMA Type 1 and is FDA-approved (Nov 2016 news; May 2018 conference news). Unlike gene therapy, antisense therapy needs to be delivered indefinitely.

How About Neurodegenerative Disease?
At SfN, scientists outlined strategies for treating adults who face years of decline due to Alzheimer’s, amyotrophic lateral sclerosis, frontotemporal dementia, Huntington’s (HD) and Parkinson’s diseases (PD), or other synucleinopathies. Some are being tested in clinical trials, others are in preclinical development. Some target specific losses or gains of function, others aim to rescue dying neurons more broadly. Scientists also believe that working on rare childhood diseases of lysosomal storage may give them an opening to treat this common phenotype in age-related neurodegeneration, as well.

Just this October, an ApoE gene therapy trial started enrolling. Led by Ronald Crystal at Weill Cornell Medical College, New York, it will inject adeno-associated virus carrying the gene for ApoE2 into patients with early to late-stage AD who inherited two copies of ApoE4. The idea is to flood their brains with the protective allele of this apolipoprotein to try to counteract the effects of the risk allele. AAV-rh10-APOE2 will be injected directly into the subarachnoid cisternae of participants’ brains. The Phase 1 trial will recruit 15 patients with biomarker-confirmed AD. Beverly Davidson, Children’s Hospital of Philadelphia, has a similar ApoE2 gene therapy in preclinical development.

At SfN, Abeliovich detailed Prevail’s programs for forms of PD and for frontotemporal dementias that are caused by risk alleles. A trial has begun for a glucocerebrosidase-based gene therapy. The enzyme GCase is essential for lysosomes to function properly. People who have loss-of-function mutations in both copies of the GBA1 gene develop Gaucher’s, a lysosomal storage disease. The severest form starts in babies, most of whom die before age 2. Milder forms cause later-onset Gaucher’s, while heterozygous mutations in GBA1 increase risk for Parkinson’s, making restoration of GCase an obvious strategy for PD. Some researchers are trying to develop ways to boost activity of the mutated enzyme (e.g., Oct 2019 news), whereas Abeliovich and colleagues have constructed AAV-9 vectors to deliver normal GBA1 into the brain to restore GCase production.

In preclinical studies, the AAV9-GBA1 construct PR001 rescued both lysosomal and brain function in models of GCase deficiency and of Parkinson’s, Abeliovich said. In mice fed the GCase inhibitor conduritol β epoxide (CBE), PR001 injected into the brain ventricles beefed up GCase activity and reduced glycolipid accumulation, which is a sign that lysosomes are functional. A single dose worked for at least six months. Similar results were seen in a commonly used model of Gaucher’s that expresses the V394L GBA mutation and only weakly expresses prosaposin and saposins, lysosomal proteins that metabolize lipids. In these 4L/PS-NA mice, PR001 made increased levels of active GCase, fewer lipids accumulated, and the mice were more mobile on a balance beam. 4L/PS-NA mice also accumulate α-synuclein, the major component of Lewy bodies in PD and other synucleinopathies. In these mice, and also in A53T α-synuclein mice made worse with CBE, PR001 halved the amount of insoluble α-synuclein, Abeliovich reported at SfN.

In search of the right dose for humans, the scientists next turned to nonhuman primates. They injected PR001 into the cisterna magna in hopes AAV9 would broadly distribute throughout the brain. At the highest dose, 8 x 1010 capsids per gram of brain weight, exposure in the brain was similar to that seen in the mice. The virus permeated the spinal cord, frontal cortex, hippocampus, midbrain, and putamen.

Also in October, Prevail scientists began recruiting for a Phase 1/2 double-blind, sham-controlled trial to test this gene therapy in 16 people with moderate to severe PD, who have mutations in one or both copies of their GBA1 genes. Six patients each will receive a low or high dose of PR001A. Blood and CSF biomarkers to be analyzed at three and 12 months, and at follow-up, include GCase, lipids, α-synuclein, and neurofilament light chain. Participants will also undergo cognitive, executive, and motor-function tests and brain imaging. A Phase 1/2 trial of PR001 in neuronopathic Gaucher’s, which affects the brain and spinal cord, will start soon, Abeliovich said.

Other groups are boosting dopamine production in Parkinson’s by way of gene therapy. VY-AADC, developed by Voyager Therapeutics, Cambridge, Massachusetts, packages the gene for L-amino acid decarboxylase (AADC), which converts L-dopa into dopamine, in an AAV-2 vector that is delivered into the brain. Two Phase 1 open-label trials are testing safety and efficacy. Both the PD-1101 and PD-1102 trials use MRI to guide injections of the vector bilaterally into the putamina of 15 or 16 patients, respectively. According to preliminary results presented at the annual meeting of the American Academy of Neurology this past May, the virus penetrated half of the putamen and AADC activity, as judged by 18F-DOPA PET, increased by 85 percent in the latter study. Seven of eight treated patients reported improvement after a year, along with longer “on” time on L-DOPA, and shorter “off” time. Off time is the period when L-DOPA effects wear off and patients experience loss of motor control. RESTORE-1, a Phase 2 study of 42 patients, started in 2018 and will run to the end of 2020.

Long-Lived Gene Therapy. When a Parkinson’s disease patient died eight years after neurturin gene therapy, the trophin was still being expressed in their putamen (top left) and substantia nigra (bottom left), where it corresponded with tyrosine hydroxylase activity (right). [Courtesy of Jeff Kordower.]

Also in PD, Kordower and colleagues plan to re-evaluate neurturin-based gene therapy. Previously, the gene for this neurotrophin was delivered in an AAV2 vector into the brains of Parkinson patients in Phase 1 and 2 trials. This did not improve motor function. Even so, in Chicago Kordower showed that in two patients who died eight and 10 years later, the inserted gene was still expressing neurturin and that dopamine levels were higher on the injected than the contralateral side of the substantia nigra/putamen. “This shows us that long-term gene expression can be achieved in the human brain,” said Kordower (see image above). He believes that by focusing delivery with ultrasound, or tweaking the capsid itself, he may be able to generate enough gene expression to improve function.

Separately, AAV-GAD, a gene therapy for PD that showed promise in Phase 2 (Mar 2011 news) was acquired by MeiraGTx, New York, which will continue to develop it in the U.S. and Europe, according to founder Samuel Waksal (Nov 2018 news). 

For its part, Prevail has a gene transfer construct for frontotemporal dementia in the pipeline, as well. Called PR006, it carries GRN, the gene encoding progranulin, on an AAV9 vector. GRN mutations cause familial FTD and, much like GBA mutations, do their dirty work via lysosomal dysfunction. In Chicago, Abeliovich reported that PR006 boosted progranulin release from neurons derived from FTD-GRN patients, nearly doubling their levels of mature Cathepsin D, the lysosomal protease that chops progranulin into granulins and indicates healthy lysosomes. In progranulin knockout mice, PR006 restored brain GRN expression and progranulin secretion into the CSF. Abeliovich said he expects a Phase 1/2 clinical trial in FTD patients to start in early 2020.

The biotech company Passage Bio, Philadelphia, is planning for clinical trials early next year with its AAV-GRN vector. MeiraGTx, New York, is banking on a different approach for FTD. They have developed an AAV carrying UPF1, which encodes regulator of nonsense transcripts 1. This protein helps clear out aberrant RNAs through a process call nonsense-mediated decay. MeiraGTx hopes this will restore homeostasis to RNA processing. AAV-UPF1 will be trialed for FTD and all forms of ALS bar those caused by mutations in SOD1. For SOD ALS, Novartis, Basel, Switzerland, and REGENXBIO, Rockville, Maryland, have a vector in preclinical testing.

For his part, Svendsen is taking a different approach. His lab tackles ALS with ex vivo gene therapy. The idea is to engineer clinical-grade human stem cells to produce glial-derived growth factor, and inject them into the spinal cord, much like the early NGF studies did in AD. Svendsen hopes the cells will churn out enough of the neurotrophin to protect spinal cord motor neurons. In a Phase 1/2a trial, 18 ALS patients have received these cells into one side of their spinal cords, such that each person serves as his or her own control. If this works, they would regain mobility only on the injected side. The trial finished in October; Svendsen expects results to come out in a few months. In a follow-up study, the scientists are trying to do the same with induced pluripotent stem cells. This would allow them to transplant autologous cells into patients, avoiding immune rejection

Other groups are deploying gene therapy as a way to improve immunotherapy, shield neurons from stress, or even generate neurons from astrocytes to make up for those lost to neurodegeneration.—Tom Fagan


  1. Viral-vector-based gene therapy has had emerging success stories in a number of neurological applications, especially for monogenic rare diseases. The trick with Alzheimer's disease (a complex disorder) will be targeting the most high-impact genes at the right time in order to forestall or reverse the course of the disease.

    For the past 15 years, I have been a firm believer in gene therapy for Alzheimer's disease, and yet high-impact targets have been elusive. The targets chosen for previous human AD gene therapy did not have an impact. This is partly a feature of our limited understanding of mechanisms of disease, and complex genetic interactions. But based on recent insights into proteolysis and metabolic contributors to AD, as well as new endothelial targets, it appears that effective AD gene therapies finally may be a few years away.

    I predict that a single treatment will need to incorporate multiple genes and anatomical targets. The goal for AD gene therapy is a one-time treatment, "one and done," which will likely involve a neurosurgical procedure. While this might sound impractical, in the long run it is likely to be much less burdensome on patients (and less expensive) than frequent IV dosing of medications for life.

Make a Comment

To make a comment you must login or register.


News Citations

  1. NGF Gene Therapy Trial Data Published
  2. Positive Trials of Spinal Muscular Atrophy Bode Well for Antisense Approach
  3. As RNA Therapies Come of Age, Efficacy Remains Weak
  4. Small Molecules Liven Up Lethargic Lysosomes in Parkinson’s Neurons
  5. First Phase 2 Success for Gene Therapy in Parkinson’s
  6. Imaging Data Resurrects Abandoned Parkinson’s Gene Therapy

Therapeutics Citations

  1. VY-AADC

Paper Citations

  1. . Single-Dose Gene-Replacement Therapy for Spinal Muscular Atrophy. N Engl J Med. 2017 Nov 2;377(18):1713-1722. PubMed.

External Citations

  1. trial 
  2. trial
  3. PD-1101
  4. PD-1102
  5. Phase 2 study
  6. Phase 1/2a trial

Further Reading


  1. . Safety and tolerability of putaminal AADC gene therapy for Parkinson disease. Neurology. 2009 Nov 17;73(20):1662-9. PubMed.