RareResearch.AI
← Back to atlas

S443R

Category 3/4 — Most DruggablePathogenic/Likely pathogenicTransmembrane · predictedEditorial
SerineArginine at position 443 · TM4 (427-447), helical transmembrane · WFS1 (Wolframin)

Serine → Arginine at position 443 inside wolframin's fourth transmembrane helix (TM4). ClinVar Pathogenic/Likely pathogenic. AlphaMissense 0.999 (near-maximum pathogenicity score), DynaMut2 ΔΔG -0.31 kcal/mol (destabilising). A pathogenic variant whose mechanism is charge-into-membrane disruption.

Interactive 3D Structure

Wild-type reference
Wild-type S443 — hydrogen bond to L447
Fullscreen ↗
DynaMut2 mutant · S443R
Mutant R443 — hydrogen bond to F365 lost (2 contacts lost)
Fullscreen ↗

Bond changes · DynaMut2 interaction analysis

2 lost11 gained8 preserved
Interaction typeWild-type partnerMutant partnerStatus
Hydrogen bondF365Lost
Hydrogen bondF439F439Preserved
Hydrogen bondS446S446Preserved
Hydrogen bondL447L447Preserved
Polar contactT361Gained
Polar contactF365Lost
Polar contactF439F439Preserved
Polar contactL445L445Preserved
Polar contactS446S446Preserved
Polar contactL447L447Preserved
Polar contactP533Gained
Van der WaalsT361Gained
Van der WaalsV364Gained
Van der WaalsF365Gained
Van der WaalsL445L445Preserved
Van der WaalsS446Gained
Van der WaalsC537Gained
HydrophobicF365Gained
HydrophobicL447Gained
HydrophobicV536Gained
HydrophobicC537Gained

Lost / gained / preserved interatomic contacts at the variant residue, from the DynaMut2 (Arpeggio) interaction analysis of the wild-type and energy-minimized mutant structures.

Computational Predictions

DynaMut2 ΔΔG
-0.31kcal/mol
Destabilising — mild
AlphaMissense
0.999
LPath
AlphaFold pLDDT
88
model confidence
Schema
Cat 3/4
Category 3/4 — Most Druggable

Clinical Evidence

ClinVar classificationPathogenic/Likely pathogenic
Review statuscriteria provided, multiple submitters, no conflicts
Associated conditions(no specific conditions catalogued for S443R — ClinVar Pathogenic/Likely pathogenic by review evidence)
InheritanceInheritance not specified in this ClinVar entry. Given the multiple submitters with consistent Pathogenic/Likely pathogenic classification, the variant likely contributes to the WFS1 spectrum across both AR and AD presentations.
Population frequency (gnomAD v4)Absent from gnomAD v4
cDNA changec.1329C>G
ClinVar accessionVCV001315675
Last evaluated2025/07/13 00:00

Not observed in ~730k individuals — consistent with a rare allele (ACMG PM2_supporting).

Structural Context

Position 443 sits inside TM4, one of wolframin's eleven transmembrane helices. The AlphaFold model places S443 within 5 Å of THR442 (2.5 Å) and TYR444 (2.5 Å), and into a packed environment with PHE439 (3.5 Å), SER446 (3.6 Å), PHE365 (3.7 Å, from TM3), and THR440 (3.9 Å). The wild-type serine's small polar hydroxyl fits well in this membrane-embedded context, possibly forming a hydrogen bond with SER446 or with the backbone carbonyl of a nearby residue.

Replacing serine with arginine here introduces three layered structural costs. First, the volume difference: arginine is one of the larger amino acids by side-chain volume, while serine is among the smaller. The local packing has to accommodate roughly four-fold more side-chain mass. Second, the charge: arginine's guanidinium group carries a positive charge that is thermodynamically penalized in the bilayer hydrophobic core. Third, the H-bonding capacity changes — the lost serine hydroxyl is replaced by the guanidinium's strong H-bond donor character, which would prefer to engage water or polar partners outside the membrane.

DynaMut2's |ΔΔG| of 0.31 kcal/mol underestimates the structural disruption. The variant probably forces local rearrangement: the arginine side chain extends toward the membrane-water interface (where its charge can be partially satisfied), pulling nearby residues out of their wild-type positions. The PHE365 contact from TM3 (3.7 Å) is particularly important — that's a helix-helix interaction, and disrupting it perturbs the relative geometry of TM3 and TM4.

AlphaMissense's score of 0.999 reflects the severity of this mechanism even with a small structural ΔΔG. The variant is pathogenic because it disrupts helix-helix packing in the membrane, not because it unfolds the protein.

Amino-acid chemistry
Serine (S) → Arginine (R) — a small polar hydroxyl-bearing residue replaced by a large, positively-charged residue with a long alkyl chain and a guanidinium group (the strongest hydrogen-bond donor in the amino acid alphabet).
Position in the protein
TM4 (residues 427–447) · position 443 is bilayer-embedded near the lumenal end of the helix. The lipid environment penalizes charged side chains heavily.

Druggability Assessment

Category 3/4 — Most Druggable. |ΔΔG| = 0.31 kcal/mol — fold survives. AlphaMissense 0.999 (near-maximum) confirms severe functional consequence despite modest structural cost.

The mechanism is charge-into-membrane plus TM3-TM4 helix-helix interface disruption. S443R is one of several Atlas variants where a charged residue is introduced into a bilayer-embedded position (compare T641K, V536E). Across these cases, the therapeutic target is the helix-helix interface that the wild-type residue stabilized, not the helix itself.

The therapeutic strategy is site-specific: a small molecule that occupies the TM3-TM4 interface near the PHE365 contact would compensate for the disrupted packing. The atlas-derived structural framework makes this target geometry visible — pre-atlas, the small ΔΔG would have deprioritized this variant for structure-based drug design.

Why this matters

S443R is the cleanest example in the Atlas of the "charge-into-membrane" mechanism class. The protein folds, the variant is pathogenic, and the mechanism is helix-helix interface disruption inside the bilayer. Drug discovery for this class of variant targets the TM-TM interface, not the individual TM helix. The Atlas surfaces this whole class by surfacing the cross-helix neighbor contacts in the PDB analysis.
Therapeutic Strategy Handoff · prediction

Feed this card to Wolfram Intelligence

Download the S443R PDF below and upload it to Wolfram Intelligence to generate therapeutic-strategy proposals — guanidinium mimetics, sigma-1 agonist docking, NAC thiol-capping. NAC is already on the bench-testing list.

Download S443R PDF card ↓Strategies are AI-generated predictions, not validated therapeutics.

UniProt Domain Annotation

Chain1890 · Wolframin
Transmembrane427447 · Helical
Natural variant443443 · in WFS1