Description

MECHANISM: SMA is caused by loss-of-function mutations in SMN1, leaving SMN2 as the sole SMN-producing locus. A C-to-T transition at position +6 of SMN2 exon 7 disrupts an exonic splicing enhancer (ESE) recognized by SRSF1 and simultaneously creates a binding site for the splicing repressor hnRNP A1/A2, resulting in ~90% skipping of exon 7 and production of a truncated, rapidly degraded SMDELTA7 protein (PMID: 37194521). The ~10% of transcripts that retain exon 7 produce functional full-length SMN, which partially compensates for SMN1 loss; SMN2 copy number therefore acts as the primary disease severity modifier. Motor neurons are selectively vulnerable because they have exceptionally high demands for SMN in snRNP biogenesis and axonal mRNA transport, and because SMN-depleted cells accumulate splicing errors in transcripts critical for neuromuscular junction maintenance. The therapeutic objective is therefore to shift the SMN2 exon-7 inclusion ratio from ~10% toward ~50-100%, raising total SMN protein above the threshold required for motor neuron survival. EVIDENCE CONVERGENCE: Multiple mechanistically orthogonal interventions all converge on exon-7 inclusion or SMN2 sequence correction as the critical node. Nusinersen, an intrathecally delivered ASO targeting ISS-N1 in SMN2 intron 7, blocks hnRNP A1/A2-mediated repression to increase exon-7 inclusion, and has demonstrated favorable long-term benefit-risk in both infantile- and later-onset SMA across broad populations (PMID: 33501671, PMID: 38192577, PMID: 29689734). Risdiplam, an orally bioavailable small molecule, acts on two distinct ESEs in SMN2 exon 7 and exon 8 to stabilize the spliceosomal complex and achieve systemic SMN upregulation; it demonstrated efficacy across SMA Types 1, 2, and 3 in pivotal trials and became the first approved oral disease-modifying therapy (PMID: 35316106, PMID: 30044619, PMID: 37194521). Adenosine base editors can directly revert the C-to-T SMN2 mutation via an A-to-G edit, achieving up to 99% correction efficiency in SMA patient-derived fibroblasts and offering the prospect of a durable, single-intervention genomic fix (PMID: 38057426). Onasemnogene abeparvovec delivers a functional SMN1 transgene via AAV9, bypassing SMN2 splicing entirely, and is feasible even in presymptomatic patients with elevated initial AAV9 antibody titers provided titers decline on retesting (PMID: 38306058). The convergence of ASO, small-molecule, base-editing, and gene-replacement modalities on a common phenotypic endpoint—restored SMN protein and motor neuron function—constitutes exceptionally strong mechanistic validation of this target. CONTRADICTIONS AND LIMITATIONS: Despite strong clinical validation, several gaps remain. Nusinersen requires repeated intrathecal dosing for life, and ~9% of patients do not report subjective improvement at one year (PMID: 38192577); poor responders may benefit from switching to onasemnogene abeparvovec (PMID: 35606491), suggesting that SMN restoration alone may be insufficient in some disease contexts, possibly implicating SMN-independent pathological cascades. Risdiplam's off-target splicing effects on FOXM1 and other pre-mRNAs bearing similar ESE motifs raise long-term safety concerns not yet fully resolved by current trial durations. Base editing efficiency data come from fibroblasts (PMID: 38057426) and in vivo delivery to post-mitotic motor neurons at therapeutic scale remains undemonstrated. AAV9 antibody titer thresholds for onasemnogene abeparvovec eligibility are still being refined, and immune-mediated hepatotoxicity remains a monitored safety signal. No head-to-head RCT comparing modalities exists, limiting direct efficacy ranking. THERAPEUTIC ANGLE: The mechanistic clarity of the SMN2 splicing defect makes it an ideal multi-modality target. For patients requiring immediate systemic coverage, risdiplam offers oral dosing with CNS and peripheral tissue penetration. For CNS-predominant disease or patients with swallowing difficulties, nusinersen's intrathecal ASO delivery concentrates drug at the site of greatest pathology. For presymptomatic neonates identified by newborn screening, onasemnogene abeparvovec or base editing offers the potential for a one-time intervention before irreversible motor neuron loss. Combination strategies pairing risdiplam or nusinersen with neuroprotective agents targeting downstream SMN-independent pathways (e.g., plastin-3, NCALD) represent a logical next frontier. Base editing with adenosine base editors is the most mechanistically precise approach—permanently converting the disease-causing nucleotide to wild-type—but requires demonstration of durable in vivo delivery to motor neurons before clinical translation.

Key questions

• Does combined risdiplam treatment plus adenosine base editing of SMN2 exon 7 achieve additive or synergistic SMN protein restoration compared to either intervention alone in iPSC-derived motor neurons from Type 1 SMA patients?
• What is the minimum percentage of SMN2 exon-7 inclusion required to cross the survival threshold in human motor neurons, and does this threshold differ between spinal and bulbar motor neuron subtypes?
• Does intrathecal nusinersen rescue splicing defects in SMN-sensitive transcripts (e.g., STASIMON/Tmem41b, NCALD) in post-mortem spinal cord tissue from treated SMA patients compared to untreated controls?
• Can AAV-delivered adenosine base editors achieve >50% SMN2 C6T reversion efficiency in lumbar motor neurons of the Taiwanese SMA mouse model following a single intrathecal injection, and does this correlate with restored NMJ morphology at 90 days post-treatment?
• In nusinersen poor-responders, do transcriptomic profiles from CSF-derived exosomes reveal a distinct splicing dysregulation signature that predicts response to subsequent onasemnogene abeparvovec rescue therapy?

Supporting evidence (828)

• ee31a919-464c-8059-f7d7-9c4a94473130
• 6dc88c07-a7b9-6846-5a5c-c5b8ec470565
• c12e97fa-f10d-dde8-25ac-77f9b5f9a0fe
• fb121997-92c1-3fc0-d6b5-20d695f5adc4
• 1ed86f0a-2776-20a9-4129-c4cf74027e1b
• 89b6d8b3-669d-e287-a942-e8b601737210
• c7eb697a-37ec-8a4a-e138-6e0b6c52b664
• 82e24b2e-8481-23d2-30b2-205423fb4434
• 9d782c09-6bc8-fe63-f87d-4dbb38ede2a2
• 5cee4738-7b55-416f-b732-7b7b30916b2d
• b4037d39-4c9a-65d7-59dd-675515adfd10
• 21e798e2-0f95-bb51-add8-36b6b3aaa72b
• 109db88b-1b9a-c44e-ff44-f3e1d84f324f
• 7b44dfa3-ebbb-28d3-fc40-a79aedf5c426
• 1509eba5-757c-4bcd-01ca-f6ca541f4875
• be50d756-98f9-a11d-ac2b-f771be5e50c9
• c1369d5a-7edf-e0bb-4024-7e81f17f0bd7
• f306afcc-24cf-b6fb-10c8-21e6ef98a1a7
• 24a7a7ce-5574-4a01-cd12-f40e6d22f56e
• 0f724528-50a3-656f-8ef7-566a58289e86
• … and 808 more

Related claims (20)