Search for a command to run...
Pathogenic variants in the gene STUB1 cause both autosomal dominant ataxia (STUB1/SCA48)1 and autosomal recessive spinocerebellar ataxia (STUB1/SCAR16)2; SCA48 is mainly characterized by late-onset cerebellar ataxia and cognitive involvement, but the presence of choreic movements is also described.3 An abnormal CAG repeat expansion in the TBP gene is responsible for a complex form of autosomal dominant ataxia (SCA17), characterized by cerebellar ataxia, movement disorders, and cognitive involvement, mimicking Huntington's disease4, 5: whereas TBP46–66 expanded alleles are considered fully penetrant, TBP41–45 expanded alleles are considered as intermediate alleles with reduced penetrance.6 The recurrent association between STUB1 dominant variants and TBP intermediate alleles is an intriguing and quite recent finding. This was initially described as digenic inheritance, because in a series of subjects with intermediate TBP alleles the presence of a concomitant STUB1 variant was a required condition to develop symptoms, explaining the reduced penetrance.7 However, in more recent series of ataxic patients with STUB1 pathogenic variants (SCA48), intermediate TBP41–45 alleles are reported only in about 19% to 58% of the patients, according to the series.8-10 This suggest that SCA48 (ATX/STUB1) is primarily a monogenic disorder, although the co-occurrence of intermediate TBP41–45 alleles is a significantly frequent finding and an important modificatory factor, with co-occurrence giving a more severe and complex phenotype.11 We agreed and have personal experience of ataxic families with STUB1 pathogenic variant and “normal” (<41 repeats) TBP alleles, as already reported.12 However, the high percentage of co-occurrence is still a striking and an unusual finding in monogenic disorders. Moreover, many aspects of this association should still be clarified in particular: (1) the boundaries between normal, intermediate, and fully penetrant TBP repeat expansions, and especially the lower threshold used to define an intermediate TBP allele, is not clearly established; and (2) the implication for diagnostic purposes is not clearly defined. Of course, another important issue is genetic counseling, but this will not be the aim of this viewpoint. We introduce this viewpoint with two exemplary cases from our clinical practice. Case 1.A 58-year-old woman was referred to our clinic for progressive choreic movements with anosognosia, gait instability, and cognitive impairment since the age of 53 years. No family history was reported. Huntington's disease was suspected, but genetic testing was negative. A complete search for Huntington's disease-like (HD-like) disorders was performed, including dentatorubro-pallidoluysian atrophy, SCA17, and HD-like disease 2 (JPH3) and turned out to be negative. Because cerebral magnetic resonance imaging showed a global cerebellar atrophy, we also performed the analysis for CAG expansion dominant spinocerebellar ataxia (SCA1, SCA2, SCA3, SCA6, and SCA7) and mitochondrial DNA analysis, which were normal. Because the phenotype really evoked Huntington's disease, genetic analysis was performed twice. Finally, a trio genome sequencing was performed in the index case and her parents, that were reported as asymptomatic. A pathogenic NM_005861.4 c.433_435delAAG p.(Lys145del) STUB1 variant was found in both the index case and her father. This variant was already described as pathogenic for SCA48 in 5 patients from the literature8 and therefore considered pathogenic. However, it is present in 9 individuals in the general population in gnomAD, version 4.1 (https://gnomad.broadinstitute.org/), corresponding to 1 of 65,000 heterozygous carriers. Unfortunately, the 87-year-old father died in the meantime, and a detailed neurological examination was not performed; however, after further investigation, it turned out that for several years he had had an unsteady gait requiring a stroller and cognitive impairment, confirming the segregation in the family. The analysis of genome data using Expansion Hunter found TPB39/37 alleles in the index case and TBP39/30 alleles in the possibly affected father. Case 2.A 45-year-old subject presented with cerebellar ataxia, dysarthria, involuntary choreoathetoid movements, anosognosia, and executive and behavioral abnormalities. The sister of the index case also had similar symptoms, but genetic analysis was not possible. Genome sequencing was performed and found a missense variant c.780C>A; p.(His260Gln) in STUB1 in the index case but not in her 2 unaffected siblings. The variant was considered pathogenic based on in silico prediction, as it involved a very conserved amino acid of the U-box domain and was absent from gnomAD, version 4.1. The analysis of TBP found TBP39/37 alleles in the proband and also in the 2 unaffected siblings (who did not have the STUB1 pathogenic variant). According to the current criteria and threshold for TBP expansions,6 these 2 patients had STUB1 pathogenic variants and a normal repeat number in the TBP gene (fewer than the 41-repeat threshold for intermediate alleles and fewer than the 46-repeat threshold for pathogenic alleles). For both cases the phenotype associated ataxia and choreoathetoid movements, mimicking Huntington's disease. However, we observed that in the literature the lower threshold used to define an intermediate TBP allele is not clearly established: the thresholds of both TBP>39 and TBP>40 were used9; a possible impact of TBP3910 and TBP>387 was sometimes suggested. Indeed, considering different thresholds, the number of STUB1 ataxic patients with concomitant TBP intermediate expanded allele was considerably greater in most of the published papers (Table 1). In Barbier et al.,8 62 of 63 patients (98%) had a TBP>36 allele, 61 of 63 (97%) had a TBP>37 allele, and 34 of 63 (54%) had a TPB>38 allele; in van Prooije et al.,9 20 of 21 STUB1 (95%) patients had a TBP>36 allele, 17 of 21 (81%) had TBP>37, and 12 of 21 (57%) had TBP>38; in De Winter et al.,10 the 5 patients with the so-called “normal TBP allele” harbored in fact TBP38 (n = 3) or TBP39 (n = 2) alleles; in Palombo et al.,13 all 4 (100%) had TBP>37 and 50% had TBP>38 (Table 1). To better assess whether TBP alleles with shorter expansions (36–40 repeats) might have an impact, we compared their frequency in the STUB1 group (117 ataxic patients reported in the literature)7-10, 13 and in 2 control cohorts: (1) AURAGEN, comprising 6554 individuals randomly selected from the Auragen whole-genome sequencing database (French Genomic Medicine Initiative); and (2) gnomAD, comprising 16,351 individuals from the tandem repeat dataset of gnomAD, version 4.1 (Genome Aggregation Database). Because the methods used to assess TBP repeat size differ among datasets, we manually corrected the repeat number reported by ExpansionHunter (https://github.com/Illumina/ExpansionHunter) for the AURAGEN and gnomAD cohorts. ExpansionHunter systematically underestimates the length of the complex expansion motif (CAG)3–(CAA)3–(CAG)N–CAA–CAG–CAA–(CAG)N–CAA–CAG by one repeat. This occurs because it counts GCA repeats from GRCh38 positions 170,561,907 to 170,562,017 instead of counting CAG repeats from positions 170,561,908 to 170,562,021. The results (Table 1; Fig. 1) indicate a distinct distribution of TBP alleles (considering the longer allele) in the STUB1 group compared with controls. In the STUB1 group, alleles with ≥40 and 39 TBP repeats were enriched by ~5- to 15-fold and 2- to 6-fold, respectively, relative to controls. TBP38 alleles are also enriched relative to the AURAGEN cohort, though this enrichment is not observed when compared with gnomAD. The reason for this discrepancy between the two control groups remains unclear. In addition, the distribution of the shorter TBP allele also differs between the STUB1 and control groups (Fig. 1). In patients with STUB1 mutations, TBP38 and TBP39 shorter alleles were approximately 3- and 20-fold more frequent, respectively, than in controls. The significance of this observation remains unclear and warrants further investigation. Overall, these observations suggest that the threshold at which TBP expansions act as modifiers in patients carrying STUB1 mutations may be lower than previously reported.6 From a practical standpoint, we propose that in a symptomatic patient, the presence of a pathogenic or likely pathogenic STUB1 variant may be sufficient to establish the diagnosis of SCA48. The search for a concurrent intermediate TBP expansion could then be reserved for cases with atypical features (eg, unusually early onset, significant cognitive impairment, choreic movements, such as in the 2 clinical cases reported here) or for research purposes. Conversely, in a symptomatic patient carrying an intermediate TBP expanded allele (up to 45 CAG repeats), we recommend systematically assessing the presence of a co-occurring STUB1 mutation, as this may be crucial for explaining the phenotype and for accurate genetic counseling. In the presence of a STUB1 pathogenic or likely pathogenic variant, the lower threshold defining an intermediate TBP expansion should likely be reconsidered: 39 and ≥40 repeats appear likely pathogenic, 38 repeats should be regarded as of uncertain significance, and alleles with 37 repeats are probably benign. In the absence of STUB1 pathogenic variant, the threshold to define intermediate and full penetrant TBP expansion is defined in accordance with the published criteria, that is, reduced penetrant allele for TBP41–45 and fully penetrant alleles for TBP>45 expansions.6 From a pathophysiological point of view, STUB1 encodes CHIP, a co-chaperone of the ubiquitin-proteasome system, which could be involved in the degradation of TBP protein. In the presence of STUB1 variants, the haploinsufficiency or dysfunction of CHIP could decrease degradation and favor the accumulation of intermediate expanded TBP7; moreover, the accumulation of TBP or its interaction with CHIP can be modulated by the CAG length. In contrast, fully expanded TBP alleles are pathogenic per se, independently of the presence of STUB1 mutation. The size of the repeat in the blood may not strictly correlate with its size in the tissues involved in the pathology (ie, central nervous system), and somatic variations could add further complexity to the comprehension of the interaction between TBP and STUB1 and be responsible for the phenotypic variability. (1) Research project: A. Conception, B. Organization, C. Execution; (2) Statistical analysis: A. Design, B. Execution, C. Review and critique; (3) Manuscript preparation: A. Writing of the first draft, B. Review and critique. C.M.: 1C, 2B, 3A Q.C.: 1C, 2B C.G.: 1C, 3B J.T.: 1C, 3B V.B.: 1C, 2B F.O.M.: 1B, 2C P.F.: 1C, 2B M.R.: 1C, 2B M.D.;1B, 2C M.K.: 1B, 1C, 3A F.R.: 1C, 2B, 3A This research was made possible through access to data generated by the 2025 French Genomic Medicine Initiative (Plan France Médecine Génomique 2025 [PFMG2025]). Open access publication funding provided by COUPERIN CY26. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.