A comprehensive overview of gene therapy, antisense oligonucleotide therapy, and small-molecule precision therapies for neurological genetic diseases — covering vector biology, approved treatments, ongoing clinical trials, and the practical and ethical considerations of delivering these transformative therapies.
Tags: Neurogenetics · Advanced
Learning Objectives
1.Describe the major categories of gene and molecular therapy and their mechanisms of action
2.Explain AAV vector biology and the factors limiting CNS gene delivery
3.Describe approved gene therapies and antisense oligonucleotide treatments for neurological conditions
4.Recognize the challenges of immune responses to AAV vectors and their clinical management
5.Apply knowledge of precision therapies to counsel patients and families about therapeutic expectations
01Categories of Gene and Molecular Therapy
Gene and molecular therapies for neurological disorders span a spectrum from gene replacement (delivering a functional copy of the deficient gene) to gene silencing (reducing production of a toxic gain-of-function protein) to gene editing (correcting a specific pathogenic variant in situ). Each approach has distinct mechanisms, delivery challenges, and appropriate applications depending on whether the disorder results from loss-of-function or gain-of-function pathogenic mechanisms.
Key Points
Gene addition/replacement: deliver functional gene copy via viral vector (AAV) or non-viral carrier; appropriate for recessive loss-of-function disorders; does not alter genomic DNA — the transgene remains episomal in neurons; examples: SMA (Zolgensma), Fabry (Phase 3), hemophilia
Gene silencing — antisense oligonucleotides (ASOs): short (18–25 nt) chemically modified single-stranded DNA oligomers that bind target mRNA via Watson-Crick base pairing → RNaseH-mediated mRNA degradation or steric blockade of translation/splicing; intrathecal or IV delivery; examples: nusinersen (SMA), tofersen (SOD1-ALS), inotersen (hATTR)
Gene silencing — RNA interference (RNAi): siRNA or shRNA; double-stranded RNA triggers RISC complex to cleave complementary mRNA; examples: inclisiran (PCSK9-targeting siRNA — subcutaneous), patisiran (TTR-siRNA for hATTR — IV lipid nanoparticle)
Gene editing — CRISPR-Cas9: guided by sgRNA; creates double-strand break at specific genomic locus; HDR corrects sequence (in dividing cells) or NHEJ creates indels (in post-mitotic neurons — limited HDR); base editors and prime editors avoid DSBs; Casgevy (sickle cell/beta-thal) — first approved CRISPR therapy (2023)
mRNA therapeutics: delivery of modified mRNA encoding the therapeutic protein; transient expression without genomic integration; lipid nanoparticle delivery; relevant to IEM (OTC deficiency trials), neurological applications emerging
02AAV Vector Biology and CNS Delivery
Adeno-associated virus (AAV) is the dominant vector for in vivo gene therapy in neurology. AAV is a small (25 nm), non-enveloped, single-stranded DNA virus that is naturally replication-deficient. Recombinant AAV (rAAV) retains only the inverted terminal repeats (ITRs) flanking the therapeutic transgene — all viral coding sequences are removed, reducing immunogenicity. Serotype selection (AAV1, 2, 5, 8, 9, AAVrh10, etc.) determines cell tropism, tissue distribution, and CNS penetration.
Key Points
AAV9 and AAVrh10: preferred serotypes for CNS gene therapy; cross the blood-brain barrier after intravenous administration (most efficiently in neonates/young children — BBB permeability decreases with age); transduce both neurons and astrocytes; used in SMA (Zolgensma), Batten disease, MLD trials
Cargo capacity: maximum ~4.7 kb insert — limits use for very large genes (DMD full-length cDNA = 14 kb, too large; micro-dystrophin constructs used instead); suitable for SMN1 (1.7 kb), MECP2 (1.5 kb), ARSA (MLD), CLN genes (Batten)
Episomal persistence: rAAV DNA remains largely episomal (non-integrating) in post-mitotic neurons — gene expression is stable long-term without insertional mutagenesis risk; in rapidly dividing cells, expression is lost with each division
Routes of administration for CNS: IV (crosses BBB in young patients, requires high dose); intrathecal/intracerebroventricular (reduces dose needed, bypasses BBB, spreads via CSF); direct intraparenchymal (focal delivery — Parkinson's, Alzheimer's trials); intravitreal (eye diseases — LCA/RPE65)
Immune responses: pre-existing anti-AAV neutralizing antibodies (from natural AAV infection) — prevalent in adults (40–70% positive for AAV9 NAb); NAb seropositivity may exclude patients from IV gene therapy trials; complement activation (CARPA — complement-mediated pseudo-allergic reaction) is a risk with high-dose IV AAV; immunosuppression protocols essential
03Approved Neurological Gene Therapies
The past decade has seen the approval of several transformative gene therapy products for neurological diseases, representing the first treatments that address the underlying genetic defect rather than managing symptoms. Each approval has provided important lessons about vector biology, immune management, patient selection, and long-term durability.
Key Points
Onasemnogene abeparvovec (Zolgensma, AveXis/Novartis): AAV9-SMN1; single IV infusion; approved FDA 2019 for SMA in patients <2 years; >30,000 USD/year → 2.1 million USD one-time dose; presymptomatic treatment near-normalizes motor outcomes; dorsal root ganglion toxicity seen in primate studies at high doses — clinical monitoring required; see the [[neuromuscular|Genetic Neuromuscular Disorders]] module for detailed SMA coverage including SMN2 copy number correlation and all approved therapies
Voretigene neparvovec (Luxturna, Spark): AAV2-RPE65; subretinal injection; approved FDA 2017 for RPE65-related Leber congenital amaurosis; bilateral treatment; sustained visual improvement; 425,000 USD per eye — first approved in vivo AAV gene therapy in the US
Delandistrogene moxeparvovec (Elevidys, Sarepta): AAVrh74-micro-dystrophin; IV infusion; approved FDA 2023 (accelerated) for DMD ages 4–5 years; uses engineered micro-dystrophin (functional mini-protein, 138 kDa); ongoing trials in older patients
Atidarsagene autotemcel (Libmeldy, Orchard): ex vivo autologous HSC transduction with lentiviral ARSA; approved EMA 2020 for metachromatic leukodystrophy in pre-symptomatic or early symptomatic patients; longest follow-up >10 years; no neurological progression vs. natural history
Betibeglogene autotemcel (Zynteglo) and elivaldogene autotemcel (Skysona): ex vivo HSC gene therapy for beta-thalassemia and cerebral adrenoleukodystrophy respectively — establish lentiviral HSC platform for CNS demyelinating leukodystrophies
04Antisense Oligonucleotide Therapies
Antisense oligonucleotides (ASOs) are synthetic oligonucleotides (18–25 nucleotides) designed to bind specific RNA sequences via complementary base pairing, modulating RNA fate by multiple mechanisms. Chemical modifications (phosphorothioate backbone, 2'-O-methyl, LNA, PMO) protect ASOs from nuclease degradation and improve target affinity. ASOs administered intrathecally distribute broadly throughout the CSF compartment and penetrate neurons and glia, providing effective CNS delivery without viral vectors.
Key Points
Nusinersen (Spinraza, Biogen): intrathecal ASO; targets ISS-N1 element in SMN2 intron 7, blocking hnRNP binding and forcing exon 7 inclusion → increased full-length SMN2 mRNA; approved 2016; loading doses then q4 months maintenance; transformative in infantile-onset SMA; ENDEAR and CHERISH trials
Tofersen (Qalsody, Biogen): intrathecal ASO; targets SOD1 mRNA; RNaseH-mediated cleavage reduces SOD1 protein; approved FDA 2023 for SOD1-ALS; first approved treatment specifically for a monogenic subtype of ALS; reduces CSF/plasma NFL (neurofilament light) biomarker
Inotersen and eplontersen: RNaseH ASO targeting TTR mRNA (both mutant and wild-type); reduce TTR production; subcutaneous injection; approved for hereditary transthyretin amyloid polyneuropathy (hATTR-PN); patisiran (siRNA) has similar mechanism and indication
UBE3A-ATS ASO for Angelman syndrome: targets the non-coding antisense transcript that silences paternal UBE3A in neurons; Phase 1/2 trials show paternal UBE3A protein restoration; toxicity in primates caused temporary pause — development ongoing with modified dosing regimen
05Practical Considerations, Challenges, and Future Directions
The clinical implementation of gene therapies requires addressing formidable practical challenges: patient selection (age, weight, immune status, antibody titers), cost and access, monitoring for immune adverse events, long-term durability of expression, and re-dosing limitations. The field is rapidly advancing with new targets, improved vectors, and evolving regulatory frameworks.
Key Points
Pre-existing AAV neutralizing antibodies: AAV9 NAb ≥1:50 typically excludes patients from IV gene therapy — antibodies block transduction; prevalence increases with age; patients can be screened and enrolled when NAb titers decline naturally; plasmapheresis to reduce NAb titers is experimental
Immunosuppression protocols: prednisolone 1 mg/kg/day initiated before and continued weeks after AAV administration to suppress T-cell response to viral capsid; anti-CD20 (rituximab) added in high-risk protocols; liver enzyme monitoring for AAV hepatotoxicity
Episomal dilution in growing livers: in infants, rapidly dividing hepatocytes dilute episomal rAAV → decreasing expression over years; this is a critical limitation for early-treated SMA patients approaching adolescence — combination with ASO (nusinersen) may address durability
Re-dosing challenges: once immune response to AAV capsid develops after first dose, subsequent doses of the same serotype are inefficient; cross-reactive immunity between serotypes complicates switching; next-generation capsid engineering (machine learning-designed capsids, deimmunized capsids) aims to address this
Emerging neurological gene therapy targets: Huntington disease (HTT lowering ASO/siRNA/CRISPR — clinical trials underway), Parkinson disease (AADC, GAD1/2 — symptom modification), Alzheimer disease (APOE4 gene editing), Batten disease/NCL (CLN2 AAV — cerliponase alfa approved intrathecal for CLN2), Rett syndrome (MECP2 AAV — dose-sensitive), GM1/GM2 gangliosidosis (multiple programs). Note: Lecanemab (Leqembi) is an anti-amyloid-beta monoclonal antibody (NOT an ASO or gene therapy) — FDA approved 2023 for early Alzheimer's disease. Mechanism: binds soluble Abeta protofibrils for clearance. ARIA (amyloid-related imaging abnormalities) is the key safety concern; APOE4 homozygotes have highest ARIA risk
Quiz Questions
1. Which statement correctly distinguishes gene addition therapy from gene editing for neurological disorders?
A.Gene addition integrates into the genome; gene editing uses episomal vectors
B.Gene addition delivers a functional gene copy that remains episomal in neurons; gene editing modifies the genomic DNA sequence at a specific locus✓
C.Gene editing is suitable for recessive disorders; gene addition is used only for dominant conditions
D.Both gene addition (AAV) and CRISPR gene editing permanently modify the patient's germline
In gene addition (e.g., AAV-based therapy), the recombinant AAV genome (including the therapeutic transgene) remains largely episomal (circular, non-integrated) in post-mitotic neurons — providing long-term expression without altering the patient's genomic DNA sequence. Gene editing (CRISPR-Cas9) creates targeted double-strand breaks to correct, delete, or modify a specific genomic sequence. Neither AAV-based gene therapy nor somatic CRISPR editing modifies the germline — only somatic cells are treated.
2. Why is the cargo capacity of AAV vectors (~4.7 kb) a significant challenge for DMD gene therapy?
A.DMD mRNA is too unstable to be packaged into AAV capsids
B.The full-length dystrophin cDNA is ~14 kb — far exceeding AAV capacity — requiring engineered micro-dystrophin constructs that may not fully replicate all dystrophin functions✓
C.AAV vectors cannot transduce muscle tissue, limiting their utility for DMD
D.The DMD gene is X-linked, and AAV cannot deliver X-linked transgenes
The DMD gene is the largest known human gene (~2.4 Mb), encoding dystrophin (427 kDa, 14 kb cDNA). AAV vectors can only package ~4.7 kb of insert, making full-length dystrophin delivery impossible in a single AAV vector. Researchers have developed engineered micro-dystrophin constructs (~3.8 kb) that retain essential functional domains (spectrin-repeats, C-terminal domain) but lack the central rod domain. While functional, micro-dystrophin may not fully recapitulate all dystrophin activities — this is a key limitation and active area of research.
3. An 8-month-old boy is diagnosed with SMA type 1 (SMN1 homozygous deletion, SMN2 copy number 2). He has progressive respiratory insufficiency. His parents ask whether gene therapy (onasemnogene abeparvovec/Zolgensma) is appropriate. Which factor most limits his eligibility?
A.His SMN2 copy number of 2 predicts a DMD-like phenotype that AAV9-SMN1 cannot correct
B.His age and weight may exceed FDA approval criteria; pre-existing AAV9 neutralizing antibody titer must be checked✓
C.AAV9 cannot cross the blood-brain barrier in children under 12 months
D.Only children with SMN2 copy number ≥4 qualify for gene therapy
Zolgensma is FDA-approved for SMA patients <2 years of age (and up to 21 kg intrathecally in some countries). Critically, pre-existing AAV9 neutralizing antibodies (NAb) are an exclusion criterion — NAb titers ≥1:50 typically exclude patients from IV administration. All patients must be screened for NAb before enrollment. Weight and ongoing respiratory status also affect eligibility. SMN2 copy number 2 is compatible with gene therapy — all SMA types benefit from presymptomatic or early treatment.
4. An infant with SMA type 1 (homozygous SMN1 deletion, SMN2 copy number = 2) is treated with nusinersen. The drug increases full-length SMN protein. Why does this approach target SMN2 rather than correcting SMN1?
A.SMN1 is located on mitochondrial DNA and cannot be targeted by ASOs
B.SMN2 is already present in all SMA patients; nusinersen modifies its splicing to produce full-length protein from the existing gene, avoiding the need for gene delivery✓
C.SMN1 and SMN2 produce identical proteins, so there is no functional difference between them
D.SMN2 has higher expression levels than SMN1 in motor neurons, making it the better therapeutic target
In SMA, SMN1 is deleted or non-functional, but SMN2 — a nearly identical paralog — is retained. SMN2 predominantly produces a truncated, unstable protein due to exon 7 skipping (caused by a C>T transition in exon 7). Nusinersen is an ASO that binds the ISS-N1 silencer element in SMN2 intron 7, blocking hnRNP A1/A2 binding and forcing exon 7 inclusion. This increases full-length SMN protein from the already-present SMN2 gene — no gene delivery required. SMN2 copy number modifies severity because more copies mean more baseline full-length SMN production.
5. A patient with SOD1-ALS is started on tofersen. Six months later, plasma neurofilament light (NfL) chain drops substantially. This biomarker change most directly indicates:
A.Successful reduction of SOD1 mRNA and protein in motor neurons
B.Reduced axonal injury and neurodegeneration, as NfL is released from damaged axons✓
C.Immune activation caused by the ASO backbone
D.The patient has developed compensatory upregulation of other SOD isoforms
Neurofilament light chain (NfL) is released from damaged neurons and axons into CSF and blood. Elevated NfL reflects ongoing neurodegeneration. A sustained decrease in plasma/CSF NfL after tofersen treatment indicates reduced axonal injury — providing biomarker evidence that the therapy is slowing neurodegeneration. NfL is increasingly used as a pharmacodynamic biomarker in neurodegenerative disease trials because it is sensitive, accessible (blood draw), and changes faster than clinical endpoints.