In Sandhoff disease (HEXB, OMIM 268800), the master switch sits on the ↓ DOWN branch ([F] forced below); the healthy state is the ON well. The forced corrective direction is to raise s toward ON branch (polarity: supply). Classification: ○ Novel direction-only lead. Direction only — no dose, no efficacy magnitude.
The disease maps to an R19 double-well whose corrective lever is read straight off the cusp (barrier 0.4827, γ 1.3895, spinodal 0.6304). The corrective direction is forced [F]; the agent names below are direction-concordant labels at grade [O] (no dose, no efficacy or safety magnitude). This is a research hypothesis offered to experts, not medical advice.
The emergence switch
The switch for Sandhoff disease is an R19 double-well emerged from the real proximal-promoter DNA of HEXB. Its geometry is fixed by measured stiffness γ (never fitted); the numbers below are read directly off that cusp.
emergent axis
↓ DOWN [F]
healthy branch
ON
lesion
LOF_null
γ (stiffness)
1.3895
barrier
0.4827
spinodal
0.6304
s_on / s_off
1.1788 / -1.1788
fragility
0.54
corrective polarity
supply
forced direction
raise s toward ON branch
Chemistry feasibility (DIRECTION only [O]). small-molecule recrossing of the fold is geometrically plausible (barrier 0.4827 at neutral drive; lower barrier = smaller drive to supply). [O] -- no affinity/potency/dose/efficacy asserted.
The derived corrective lever
The cusp forces a corrective direction: raise s toward ON branch. This is the load-bearing output of the framework — it is [F] forced, and it is falsifiable. 2 lever families are derived; the lead is h-restore (drive modulator).
Mechanism. supply / restore the positive drive (replace-or-agonise the missing competence)
Applies to. allele-agnostic (downstream/environmental tilt)
favoured when the switch is fragile -- a small drive shift recrosses the fold
Falsifier. If supplying/agonising HEXB (raising the switch drive) does NOT move the DOWN [F] switch toward the healthy branch in a HEXB-deficient model, the h-restore(supply) lever is refuted for this axis.
Falsifier. If a folding-stabiliser/chaperone does NOT re-deepen the healthy well for a residual HEXB allele (no rescue that a null allele lacks), the gamma-restore lever is refuted; null alleles are excluded by construction.
Explicitly not levers.
{"agent_class": "substrate-reduction miglustat -- no established human efficacy in GM2 gangliosidosis and NOT approved for Sandhoff -> a downstream NON-lever, not the corrective", "axis_effect": "DOWN", "lever": "restrain", "source": "Maegawa 2009 Pediatrics; Bley 2011 (miglustat in GM2, limited benefit)", "status": "investigational", "target": "HEXB", "verdict": "WORSENS"}
Agents mapped onto the lever
Existing agents whose known action is direction-concordant with the derived lever (status as pinned in the corpus; no dose, no efficacy magnitude).
agent (class)
dir.
status
phase
map
AAV-HEXA+HEXB in vivo gene therapy (dual-vector or single bidirectional/bicistronic AAV: AAVrh8 thalamic+intrathecal; AAV9-Bic_HexA/HexB systemic/CSF) -- DIRECT on the causal gene; Phase 1/2 (TSD+SD), NOT approved
Flotte 2022 Nat Med 28:251; Gray-Edwards 2018 (AAV Hex in sheep); UMass/Axovant GM2 dual-vector Phase 1/2
increase
investigational
2
◇ in trials
⚠ Mechanism-direction only — do not self-administer; no dose (unvalidated; set by a physician / national authority); not an approved use.
Direction-only candidate leads (corpus join)
Each lead is surfaced only because its known mechanism points the same way as the derived lever, and each comes with a source and a falsifier. None is a treatment.
Class. bicistronic AAV9 HEXA+HEXB (beta-hexosaminidase A) gene therapy
Mechanism. An investigational AAV9 vector co-delivering the HEXA (alpha) and HEXB (beta) subunits to reconstitute functional beta-hexosaminidase A and clear GM2 ganglioside. Direction on the disease axis: increase -- supply the deficient enzyme. Engaged nodes: HEXA and HEXB. Allele scope: agnostic. Because functional Hex A requires both subunits, the single bicistronic agent is medically relevant to BOTH GM2 gangliosidoses, so it is scoped to applies_to_genes=[HEXA,HEXB] and surfaces on both tay_sachs_disease and sandhoff_disease (the safe analogue of the mexiletine-on-both-myotonias pattern). It does NOT supply the GM2-activator protein, so it is correctly not scoped to the GM2A-activator-variant node, which remains a miss. Addresses the honest-miss reason (no approved CNS-crossing enzyme replacement). INVESTIGATIONAL direction-only hypothesis (no efficacy/approval/availability claim); status per training knowledge to Jan 2026 -- confirm current development.
Prior-art. Direction-only hypothesis: the geometry derives an increase (supply the deficient enzyme) for HEXB; the investigational bicistronic AAV9 HEXA+HEXB gene therapy TSHA-101 reconstitutes beta-hexosaminidase A, a matched increase, addressing the honest-miss reason (no approved CNS-crossing enzyme replacement). The same bicistronic agent serves both GM2 gangliosidoses (Tay-Sachs and Sandhoff); it does not supply the GM2-activator protein, so the GM2A-variant node remains a miss. Investigational; confirm current status. No efficacy/availability claim.
Safety (qualitative; no magnitude). investigational CNS-directed AAV gene therapy; vector-immune, delivery and durability considerations (qualitative; no magnitude); confirm current development status
Falsifier. If TSHA-101 (a gene-specific supply/increase agent) does NOT move the DOWN [F] switch toward the healthy branch in a HEXB-deficient model, the h-restore(supply) direction is refuted for TSHA-101 here.
Source: TSHA-101 investigational bicistronic HEXA/HEXB gene-replacement programme in GM2 gangliosidosis
⚠ Mechanism-direction only — do not self-administer; no dose (unvalidated; set by a physician / national authority); not an approved use.
Evidence & provenance
What is reproduced vs. cited for this page.
element
grade
basis
R19 switch & cusp geometry (this page)
[V] verified
emerged from measured promoter γ of HEXB; deterministic, 2×sha256 identical
corrective direction
[F] forced
forced by the cusp sign; falsifiable (see lever falsifiers)
mapped agents / leads
[O] open
direction-concordance only; corpus-pinned (2026-06-21); no dose, no efficacy/safety magnitude
Reading the agent column. VALIDATION SIGNAL -- the agent is already approved / established standard for this disease, so the direction logic recovered known clinical practice. Not a new claim by this kit. GENUINE HYPOTHESIS -- no approved indication for this disease; the direction match is a direction-only, untested [O] lead. Basis field grades the evidence (investigational here / in a related condition / off-label / speculative).