CTLA4 p.Pro137Arg
c.410C>G · rs1553657429 · VCV827701
ClinGen expert panel — Antibody Deficiencies VCEP
Likely Pathogenic for autoimmune lymphoproliferative syndrome due to CTLA4 haploinsuffiency. Applied ACMG criteria (verbatim):
PM2_Supporting PS3_Supporting PS4_Moderate PM1
Curation rationale (ClinGen, verbatim)
NM_005214.5(CTLA4):c.410C>G (p.Pro137Arg) is a missense variant encoding the substitution of proline with arginine at amino acid 137. This variant is located within the MYPPPY functional domain (residues 134-139), which is required for interaction with CD80 and CD86 (PM1, PMID: 31396201). Another missense variant in the same codon, NM_005214.5(CTLA4):c.410C>T (p.Pro137Leu), has been reported in a patient with CTLA4 insufficiency (PMID: 27102614), but has not been counted for PM5 in order to avoid circularity. This variant is absent from gnomAD v4.1.0 (PM2_Supporting). This variant has been reported in 2 apparently unrelated probands meeting the VCEP standard for phenotypic criteria, both exceeding 10 phenotype points without genotyping that excluded causes in other loci (PMID: 27102614, PMID: 29729943, PS4_Moderate). A third proband harboring the p.Pro137Arg variant with apparently de novo origins (parental genotyping methods not described) has been diagnosed with inherited immunodeficiency and has a phenotype that includes recurrent bacterial infections (1 pt) and severe viral infections (2 pts), but lacks the additional phenotype points required for inclusion in PM6_Supporting (SCV001190136.1). At least one patient harboring this variant had a phenotype that included bronchiectasis (4 pts), type I diabetes mellitus (1 pt), autoimmune cytopenia (2 pts), and enteropathy (4 pts), but was genotyped by methods that were not sufficient to rule out an alternative basis for disease in LRBA, which together were not sufficiently specific for autoimmune lymphoproliferative syndrome due to CTLA4 haploinsufficiency to meet PP4 (11 total points, PMID: 27102614). The computational predictor REVEL gives a score of 0.364, which is below the ClinGen Antibody Deficiencies VCEP threshold of >0.75 and does not predict a damaging effect. The computational predictor CADD gives a PHRED score of 24.0, which is above the ClinGen Antibody Deficiencies VCEP threshold of >20 and predicts a damaging effect on CTLA4 function. Because the two predictors do not agree, this criterion is not met. The splicing impact predictor SpliceAI gives a score of 0.02 for donor loss, which is below the ClinGen Antibody Deficiencies VCEP recommended threshold of ≥0.2 and does not strongly predict an impact on splicing. Chinese hamster ovary cells expressing the variant showed defective uptake of CD80–Ig or CD86–Ig ligands relative to the wild-type control in a soluble ligand endocytosis assay (PMID: 27102614) and ablation of binding to both soluble ligands in a flow cytometry assay (PMID: 35999394), indicating a damaging effect on protein function (PS3_Supporting). In summary, this variant meets the criteria to be classified as likely pathogenic for autoimmune lymphoproliferative syndrome due to CTLA4 haploinsufficiency based on the ACMG/AMP criteria applied, as specified by the ClinGen Antibody Deficiencies VCEP: PM1, PM2_Supporting, PS4_Moderate, and PS3_Supporting. (VCEP specifications version 1.0.0; date of approval 09/18/2025).
ClinGen Evidence Repository → · reported verbatim, not re-tallied by Sugi Variant.
Computational evidence
| Protein change | ||
| AlphaMissense | 0.8728 | likely pathogenic ✓ concordant with ClinVar · top 10% in CTLA4 |
|---|---|---|
| REVEL | 0.364 | indeterminate |
| SaProt | LLR -11.21 | damaging |
| ⇒ AlphaMissense and SaProt both call this damaging; REVEL makes no call. | ||
| Splicing | ||
| SpliceAI | — | not assessed · no allele here is in the release |
| Constraint at this position | ||
| Conservation | phyloP 7.7 | GERP 1.5, phastCons 1 |
| Regional constraint | o/e 0.66 | unconstrained · MANE · 2,198 bp · p 3.2e-07 · regional, not allelic |
| Overall | ||
| CADD | 24.3 | PHRED · no single calibrated threshold — expert panels disagree |
AlphaMissense carries the ACMG weight for missense; REVEL and SaProt are independent agreement signals.
Population frequency
Absent from gnomAD v4.1. Not observed in any sequenced alleles — a rarity signal (ACMG PM2-supporting), not evidence of pathogenicity on its own.
Variant & gene context
- Gene
- CTLA4
- Protein
- p.Pro137Arg
- Coding
- c.410C>G
- MANE Select
- NM_005214
- dbSNP
- rs1553657429
- Type
- single nucleotide variant
- Location
- chr2:203870886 (GRCh38)
- Gene constraint
- LOEUF 0.211 · pLI 1.00 (gnomAD)
- Structure
- AlphaFold pLDDT 79.67 · 22 PDB
- Residue 137
- a β-strand; the ig-like v-type domain; the 'important for interaction with cd80 and cd86' region
Disease context
autoimmune lymphoproliferative syndrome due to CTLA4 haploinsufficiency — disease report on Sugi Atlas → Associated genes, phenotypes, clinical trials & treatmentsFor Autoimmune lymphoproliferative syndrome due to CTLA4 haploinsuffiency (Orphanet):
- Inheritance
- Autosomal dominant
- Onset
- Adolescent, Adult, Childhood
Similar variants
- CTLA4 p.Pro137Leu Likely pathogenic ★★★ same residue
- CTLA4 p.Pro137Gln Conflicting classifications of pathogenicity ★ ⚠ same residue
- CTLA4 p.Pro137= Likely benign ★ same residue
- CTLA4 p.Arg51Ter Pathogenic ★★★ same condition
- CTLA4 p.Arg75Trp Pathogenic ★★★ ⚠ same condition
- CTLA4 p.Asp153Asn Pathogenic ★★★ ⚠ same condition
- CTLA4 p.Cys35Ter Pathogenic ★★★ same condition
- CTLA4 p.Pro136Ala Likely pathogenic ★★★ same condition
Built on 18 open, versioned datasets — ClinVar, gnomAD, AlphaMissense, REVEL, ClinGen & more — plus Sugi Variant's own cross-source analysis. Source data 2026-01–2026-10, ClinVar 2026-10. Full sources & versions →
Cite Sugi Variant
Gür T. Sugi Variant: A Deterministic Per-Variant Germline Reference from a Knowledge Graph for Humans and AI Agents. 2026. doi:10.5281/zenodo.21509445. This variant: https://sugi.bio/variant/ctla4-p-pro137arg (accessed [today's date]).
Preprint & method: sugi.bio/variant/method Machine-readable: this page as Markdown