Skip to main content
Neurology: Genetics logoLink to Neurology: Genetics
. 2026 Nov 9;12(6):e200447. doi: 10.1212/NXG.0000000000200447

International Multicenter Retrospective Study on UBTF-Related Neurodevelopmental Disorder

Giulia Balletto 1,2, Sara Carli 3, Antonietta Coppola 4, Annalaura Torella 5, Hsiu-Fen Lee 6,7, Christoph Korenke 8, Sebastien Lebon 9, Lucie Sedlackova 10, Amanda Nagy 11, Florian S Eichler 11, Chaya N Murali 12,13, Vincenzo Nigro 5, Farah Alnoor Ladha 12,13, Mark S Ledoux 14,15, Silvia Pelle 3, Giulia Prato 2, Lino Nobili 1,2, Luigi Chiriatti 16, Cecilia Mancini 16, Giuseppe Zampino 17, Claudio Graziano 18, Kristin Baranano 19, Pauline Planté-Bordeneuve 20, Roseline Caumes 20, Barbora Lauerová 21, Jana Haberlová 21, Duccio Maria Cordelli 3,22, Chiara Fiorillo 1,2, Marco Tartaglia 16, Rory James Tinker 23, Caterina Garone 3,22,✉
PMCID: PMC13759263  PMID: 43031596

Abstract

Background and Objectives

UBTF encodes the upstream binding transcription factor (UBTF), a protein essential for rRNA synthesis. Dominantly acting pathogenic variants, either de novo or inherited, have been associated with a broad clinical childhood-onset neurodegenerative/neurodevelopmental spectrum whose genotype-phenotype correlation has not been accurately defined yet. This multicenter study aims to conduct a retrospective analysis of a large cohort of patients with pathogenic variants in the UBTF gene to refine diagnostic criteria and identify clinically relevant outcome measures for future trials.

Methods

Clinical centers provided clinical, molecular, and neuroimaging data on patients carrying pathogenic or predicted pathogenic UBTF variants, consistent with the clinical phenotype. Data were collected using an human phenotype ontology–based standardized electronic case report form and included unreported cases or updated information on patients reported in the literature from January 2017 to December 2024.

Results

A cohort of 49 patients (23 previously unreported) was included in the study. The disease typically presented at ≤3 years (89.6%) with neuroregression in 71.4% of cases, either in a child with preexisting neurodevelopmental delay (42.9%) or after previously normal psychomotor development (28.6%). Less frequently, neurodevelopmental delay occurred in isolation (26.5%). During disease progression, patients developed behavioral abnormalities (81.6%), movement disorders (78.3%), autonomic manifestations (50.0%), and epilepsy (34.7%). At the last follow-up (mean age 13 years and 6 months), only 2 patients were deceased (4.1%). The main neuroimaging findings included white matter hyperintensity (68.1%), cerebral (66.0%) and cerebellar (51.1%) atrophy, and thinning of the corpus callosum (38.3%). However, brain MRI findings were reported as normal in 21.3% of cases. We identified 7 novel UBTF variants (c.271G>C; c.349A>T; c.495C>A; c.517G>A; c.761A>G; c.1048-2A>G; and c.1625_1626+15del). Genotype-phenotype analysis led to the distinction of 2 allelic, clinically distinct disorders: a childhood-onset neurodegenerative disease due to gain-of-function variants, characterized by the triad of neuroregression, movement, and behavioral disorders, and a neurodevelopmental disorder with a milder clinical course associated with loss-of-function/hypomorphic variants.

Discussion

The results from our deep-phenotyping analysis suggest the classification of UBTF-related disease into 2 main clinical syndromes impacting the natural history of the disease with a different degree of severity and complexity. This novel observation should be considered in the management and prognosis of patients and will be fundamental in the design of targeted therapies.

Introduction

Neurodevelopmental disorders (NDDs) affect more than 3.0% of children worldwide1 and comprise a heterogeneous group of conditions with diverse etiologies, including neurodegenerative and neurometabolic diseases2; approximately 1,400 genes have been implicated in their pathogenesis.3 Recently, both dominantly acting and biallelic pathogenic variants in the UBTF gene (MIM: 600673) have been identified as one of the molecular events underlying NDDs.4 UBTF encodes the upstream binding transcription factor (UBTF), a nucleolar protein containing nonspecific high-mobility group (HMG) boxes that contributes to DNA bending. Main functions of UBTF include epigenetic remodeling at the rDNA loci and recruitment of the Selectivity Factor 1 (SL1), a transcription factor required for RNA polymerase I function and rDNA transcription.5

In 2017, a recurrent de novo dominantly acting variant in UBTF (c.628G>A, p. Glu210Lys [E210K]; NM_014233.4) was first associated with UBTF neuroregression syndrome or childhood-onset neurodegeneration with brain atrophy (CONDBA; OMIM 617672) by Edvardson et al.4 in 7 unrelated affected individuals, sharing a similar clinical presentation of developmental regression. Studies conducted on patient-derived primary fibroblasts showed increased accumulation of pre-rRNA and 18S rRNA associated with the E210K variant, consistent with a gain-of-function (GoF) mechanism.4,6 Since then, the UBTF E210K variant has been reported in an additional 11 patients,2,6-11 profiling the CONDBA clinical phenotype as characterized by neurodevelopmental regression typically manifesting around 2.5–7 years, and having pyramidal and extrapyramidal signs, seizures, deterioration of motor and language functions, dysphagia, and behavioral problems, as additional cardinal features. Neuroimaging findings revealed occurrence of progressive cortical atrophy, white matter hyperintensity, corpus callosum thinning, and ex vacuo dilation of ventricles.2,4,6-11 The neurodevelopmental regression was highly variable, ranging from milder cases that retain ambulation and some degree of autonomy to more severe presentations characterized by early-onset dystonic tetraparesis, profound intellectual disability, and the need for percutaneous endoscopic gastrostomy feeding and nocturnal ventilatory support.2

More recently, the clinical and molecular spectrum has further been expanded with additional functionally relevant UBTF substitutions, such as c.608A>G (p.Gln203Arg),12 which has been predicted to behave as an activating change, as well as suspected loss-of-function (LoF) variants reported to be linked to a milder phenotype characterized by neurodevelopmental delay without neuroregression, and brain MRI atrophy.13 However, an accurate genotype-phenotype correlation has not yet been defined, challenging the definition of patients' prognosis. Currently, no curative treatment is available for this condition; management is exclusively symptomatic and is tailored according to the patient's clinical course.

Our multicenter international study aims to conduct a retrospective analysis of a large cohort of patients with pathogenic variants in UBTF to refine diagnostic criteria and identify clinically relevant outcome measures for future trials.

Methods

The multicenter study involved a cohort of 49 participants carrying predicted functionally relevant or bona fide pathogenic UBTF variants, fitting either CONDBA syndrome or the NDD associated with LoF changes.

Data from 22 previously reported patients were collected through a PubMed review from January 2017 to December 2024, using the keywords “UBTF neurodegeneration,” “UBTF neurodevelopmental disorder,” “UBTF neuroregression syndrome,” and “UBTF E210K.” Follow-up information was available for 9 of these patients. Four patients who were unreported patients at the end of our data collection and analysis were described in a separate manuscript published in January 2026.14 Therefore, we have updated the final number of published cases to 26 after revisions, whereas 23 remain unpublished.

Studies published after the predefined search cutoff were not included in the formal analysis but were considered in the discussion where relevant to contextualize emerging findings. Clinicians were invited to provide updated information on previously published cases or to contribute data on unreported patients using an electronic case report form (eCRF). The eCRF collected demographic information (sex, country/ethnicity, survival, age at onset, age at last follow-up, and age at death), molecular findings, clinical suspicion, and observed features coded using human phenotype ontology terms (full list available in eTable 1). Secondary microcephaly was defined as head circumference within the normal range at birth, followed by a decline below −2 SD or below the 3rd percentile during follow-up. Data were marked as N/A when unavailable or not retrievable for a specific query.

Participating centers from China (1), Czech Republic (1), Germany (1), Italy (3), Switzerland (1), and the United States (4) contributed to the study and shared data in a pseudonymized manner.

UBTF variants were annotated according to HGVS nomenclature, using the MANE Select transcript NM_014233.4 (cDNA) and the reference protein NP_055048.1. For UBTF variants whose pathogenic mechanism had not been established, the predicted functional impact was assessed by integrating diverse orthogonal lines of evidence, including in silico predictions tools, the functional relevance of the affected residue, the inheritance pattern, and, where applicable, clinical correlations. Variants affecting the N-terminal region were considered to potentially affect the protein's dimerization, variants in the HMG-boxes the rDNA binding activity, and variants in the C-terminus the transcription activation and RNA polymerase recruitment. Truncating variants (e.g., nonsense and canonical splice-site changes) likely to disrupt transcript processing and/or to introduce a premature termination codon were considered putative LoF. The functional relevance of missense changes was inferred using LoGoFunc to predict a potential GoF or LoF mechanism, whose predictions were integrated with those of multiple in silico tools such as SIFT, PolyPhen-2, CADD-PHRED, and REVEL to evaluate variants deleteriousness. All UBTF variants were classified according to the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP) guidelines.15 The following criteria were applied when appropriate: PS2 (de novo occurrence), PM1 (location in mutational hotspot and/or critical domain), PM2 (rarity in population databases), PP2 (gene constraint supporting missense pathogenicity), PP3 (computational evidence), and PP4 (phenotype specificity). Each variant was assigned a final classification as pathogenic (P), likely pathogenic (LP), or variant of uncertain significance (VUS). A hypomorphic effect was assumed for the variants associated with mild presentation of the condition.

Standard Protocol Approvals, Registrations, and Patient Consents

This retrospective, observational study was conducted in accordance with the principles of the Declaration of Helsinki and Oviedo Convention. Approval from the institutional review boards or ethics committees was obtained from the coordinating center and all participating sites. Informed consent for the anonymous publication of clinical, neuroimaging, and genetic data was obtained from patients or their legal guardians, in accordance with local regulations when required. All patient data were pseudoanonymized before transfer to the coordinating center.

Statistical Analysis

The qualitative variables were summarized by using absolute and relative frequency distributions, and the continuous quantitative variables were expressed by reporting median, min, and max. Statistical analysis was conducted with GraphPad version 10.2. The results were expressed as a ratio and percentage of available data. Given the exploratory design and the limited sample size, genotype-phenotype comparisons were conducted descriptively, and no statistical tests were applied to assess differences between genotype groups.

Data Availability

The anonymized data can be accessed on reasonable request addressed to the corresponding author.

Results

Disease Onset

The cohort comprised 49 individuals (24 men, 25 women) carrying a putative clinically relevant UBTF variant, 23 of whom had not been previously reported in the literature. Follow-up data were obtained for 9 of the 26 previously published cases.

The perinatal history was unremarkable in most patients (89.4%), whereas prenatal or neonatal abnormalities were sporadically reported (Figure 1A; eTable 2).

Figure 1. The Disease Course.

Bar graphs compare perinatal history, disease onset, and clinical course in patients with different genetic variants.

Graphs show the clinical course of the overall cohort patients and in patients carrying GoF variants in the hotspot exon 7 (c.628G>A; c.608A>G) vs patients carrying variants outside exon 7. Panel A shows the perinatal course (available data: 47/49), panel B shows the disease onset (available data: 49/49) and neurodevelopment progression (available data: 48/49), and panel C shows the main clinical manifestations throughout the disease course (available data: 49/49).

Disease onset occurred at <2 years in 26/48 patients (54.2%, 1 N/A), between 2 and 3 years in 17/48 (35.4%), and >3 years in 5/48 (10.4%), and it was primarily characterized by signs and/or symptoms of abnormal neurodevelopment. Developmental progression could be categorized as follows: neurodevelopmental regression, which was preceded by neurodevelopmental delay in 42.9% of the cases, regression without prior neurodevelopmental delay in 28.6%, or isolated neurodevelopmental delay in 26.5% (Figure 1B; eTable 2). Overall, neurodevelopmental delay, with or without regression, was observed in 34/49 patients, and detailed developmental domain information was available for 28 of them; language delay was present in all cases, while motor milestone delay was observed in 60.7% (17/28). Neurodevelopmental regression was observed in 35/49 patients. Domain-specific onset data were available for 34/35 patients, with motor skills as the first affected domain in 30/34 (88.2%), followed by language in 22/34 (64.7%) and cognitive functions in 5/34 (14.7%). The mean age at regression was 3 years and 8 months, ranging from 2 to 7 years, and, in 7 participants, was preceded by febrile illness.

At disease onset, secondary microcephaly was also observed in 40.8% of cases, while macrocephaly was reported in a single case (Figure 1B; Table 1; eTable 2). Craniofacial features were reported in 9 individuals and included sparse eyebrows, hypertelorism, single eyelid folds, blepharoptosis, flat nasal bridge, anteverted nares, long and shallow philtrum, thin upper lip, abnormal ear morphology, and ptotic eyelids. Other dysmorphic features included pectus carinatum, cryptorchidism, hypospadias, toe syndactyly, congenital clubfoot, and livedo reticularis. Dysmorphisms were reported but not further specified in 5 additional patients (Table 1).

Table 1.

Characterization of Dysmorphic Features/Congenital Abnormalities Associated With UBTF Variants

UBTF variant (No. of patients) Phenotype Macrocephaly Primary microcephaly Facial dysmorphism Pectus carinatum Cryptorchidism Hypospadias Toe syndactyly Congenital clubfoot Livedo reticularis
c.608A>G (1) Known CONDBA — — — — — — — — —
c.628G>A (33) Known CONDBA — — 2 1 1 1 3 1 2
c.271G>C (2) Novel NDD — — — — — — — — —
c.349 A>T (1) Novel NDD — — — — — — — — —
c.495C>A (1) Novel NDD — — — — — — — — —
c.517G>A (1) Novel NDD — — — — — — — — —
c.761A>G (1) Novel NDD — — — — — — — — —
c.905+1G>A (4) Known NDD — — 4 — — — — — —
c.1048-2A>G (1) Novel NDD — 1 — — — — — — —
c.1327C>T (1) Known NDD 1 — 1 — — — — — —
1625_1626+15del (1) Novel NDD — — — — — — — — —
UBTF deletion (2) Known NDD — 1 2 — — — — — —
All patients (49) 1/49 2/49 9/49 1/49 1/49 1/49 2/49 1/49 2/49

Abbreviations: CONDBA = childhood-onset neurodegeneration with brain atrophy; NDD = neurodevelopmental disorder; UBTF = upstream binding transcription factor.

Disease Course

During the disease course, patients showed signs of neurodegeneration by losing motor, cognitive, and language skills. In detail, at the last follow-up (mean age: 13 years and 6 months; range: 1 year and 11 months–34 years; 1 N/A), independent ambulation was lost in 21 of 47 patients (44.7%, 2 N/A) at the mean age of 11 years and 6 months (data available for 14/21). One patient never acquired the ability to walk. Regarding language, 25 of 47 individuals (53.2%) (2 N/A) had lost verbal abilities at the last follow-up, while 4 never developed speech. The mean age when the loss of verbal functions occurred, available in 16 cases, was 8 years and 11 months.

During the clinical course, patients also developed behavioral abnormalities in 81.6%, movement disorders in 78.3%, autonomic manifestations in 50%, and epilepsy in 34.7%. The most frequent behavioral manifestations included attention deficit hyperactivity disorder (ADHD) (52.6%), autistic behavior (42.1%), and motor stereotypies (31.6%); among movement disorders, ataxia (65.2%), spasticity (54.3%), dystonia (45.7%), dysarthria (41.3%), and hyperreflexia (39.1%) were the most frequently reported findings. Additional details on behavioral abnormalities and movement disorders are presented in Figure 1C; eTable 2).

Autonomic manifestations were peripheral cyanosis (29.4%), cold extremities (26.5%), and altered pain perception, including pain insensitivity (20.6%) and abnormal pain sensation (2.9%).

Epilepsy had no clear predominance of focal or generalized seizure onset. EEG abnormalities in the absence of clinical seizures were reported in 7/49 individuals. The mean age at seizure onset was 8 years and 11 months.

Additional clinical features included dysphagia in 60.5%, sleep abnormalities in 52.6%, scoliosis in 42.1%, hypopnea in 34.2%, apnea in 26.3%, and pseudobulbar affect in 10.5% (Figure 2; eTable 2). Furthermore, during the disease progression, 5 individuals required assisted ventilation, and 8 required enteral nutrition.

Figure 2. The Clinical Features of UBTF-Related Disease.

Bar graph comparing clinical features in different patient cohorts.

Clinical signs and symptoms described in the overall cohort and in patients carrying GoF variants in the hotspot exon 7 (c.628G>A; c.608A>G) vs patients carrying variants outside exon 7. Data are presented as percentages of the available data (38/49 for behavioral abnormalities, 46/49 for movement disorders, 34/49 for autonomic manifestations, 49/49 for epilepsy, and 38/49 for other clinical features). UBTF = upstream binding transcription factor.

Only 2 patients in the cohort were deceased.

In 11/49 cases, the UBTF-related disorder was initially misdiagnosed most frequently as hypoxic-ischemic encephalopathy (4 cases) and, in isolated cases, as Rett syndrome, Angelman disease, subacute sclerosing panencephalitis, Landau-Kleffner syndrome, lysosomal storage disease, leukodystrophy, and CHARGE syndrome.

Neuroimaging Findings

Neuroradiologic alterations were found in 78.7% patients, as presented in Figure 3 and eTable 2, with a predominant pattern of cerebellar and cerebral atrophy with white matter T2 hyperintensity at the brain MRI. Chiari malformation type I was reported in 3 patients. Abnormal magnetic resonance spectroscopy findings were reported in 10/47 patients (21.3%). Specifically, a decrease in the N-acetyl aspartate peak was observed in 7 of these 10 patients (70.0%), an increased myo-inositol peak in 4 (40.0%), an increased choline peak in 3 (30.0%), a decreased choline/creatine ratio in 3 (30%), and a mild lactate peak in 2 patients (20.0%). In a single patient, N-isopropyl-p-[123I] iodoamphetamine single-photon emission computed tomography (SPECT) was performed and showed diffuse regional hypoperfusion at the cerebral cortex.

Figure 3. MRI Findings.

Bar graph showing MRI findings in different patient groups.

Percentage of imaging abnormalities found in the overall cohort, patients, and in patients carrying GoF variants in the hotspot exon 7 (c.628G>A; c.608A>G) vs patients carrying variants outside exon 7. Data are represented as percentages of the available data (47/49).

Molecular Genetics

Thirty-three individuals (67.3%) were molecularly diagnosed with the de novo recurrent activating c.628G>A (p.Glu210Lys) variant. Five novel de novo UBTF missense variants were identified: c.349A>T, p.(Thr117Ser); c.495C>A, p.(Phe165Leu), c.517G>A, p.(Glu173Lys), and c.608A>G, p.(Gln203Arg). Two missense changes were predicted to exert a GoF effect by LoGoFunc: p.(Thr117Ser), involving a residue located at the N-terminal region of the protein within the homodimerization domain, and p.(Gln203Arg), affecting a residue close to the recurrently affected Glu210 within the HMG-box 2. In contrast, for other 2 missense changes, c.495C>A (p.Phe165Leu) and c.517G>A (p.Glu173Lys), a LoF pathomechanism was suspected. Pathogenicity of c.761A>G, p.(Lys254Arg) had been functionally demonstrated by Hwang et al.16 All the identified missense variants were ultra-rare/private in population databases and predicted to be disruptive according to multiple in silico tools. With the exception of the c.495C>A and c.517G>A substitutions, all variants were classified as LP/P according to the ACMG criteria (details are listed in eTable 3).

In addition, 3 truncating variants were identified, 1 nonsense (c.1327C>T, p.Arg443Ter) and 2 splice-site (c.1048-2A>G and c.1625_1626+15del) changes, which were predicted to introduce a premature termination codon, and impair transcript processing, respectively, consistent with a loss-of-function pathogenic mechanism.

Two unrelated patients carried de novo microdeletions encompassing the gene: 17q21.31(42248053_42294092) × 1 and 17q21.31(42239218_42345522)×1, both resulting in a whole-gene deletion and therefore consistent with a haploinsufficiency mechanism.

Although most of UBTF variants were identified as de novo events, 2 additional changes were found to segregate with disease in 2 unrelated familial cases: the missense substitution c.271G>C (p.Asp91His) located at the N-terminus of the protein and the splice-site variant c.905+1G > A, which affected a canonical splice donor site and was identified in 3 affected siblings and their affected mother. The effect of this variant on transcript processing had been functionally demonstrated14 (Figure 4). The detailed ACMG/AMP criteria applied to each variant, together with the resulting classification, are presented in eTable 3.

Figure 4. UBTF Pathogenic Variants.

A diagram of the UBTF gene structure showing exons, protein domains, and variants.

Graphical representation of the 21 exons of the UBTF gene (NM_014233.4), the corresponding protein domains, the variants detected in our cohort, and their corresponding amino acid changes. Variants are color coded in red for gain-of-function and in green for loss-of-function/hypomorphic mechanisms. The 2 chromosomal coordinate ranges refer to patients carrying larger chromosomal deletions encompassing the entire UBTF locus. Exon 7 is highlighted as a mutational hotspot associated with the more frequent and severe CONDBA phenotype (red block). UBTF = upstream binding transcription factor.

Genotype-Phenotype Correlation

The recurrent GoF c.628G>A variant resulted in a clinical phenotype characterized by the triad of neurodevelopmental regression, movement disorders, and behavioral abnormalities. All of our 33 patients carrying recurrent variants had neurodevelopmental regression and movement disorders, and most of them developed during the clinical course behavioral abnormalities (83.3%). When the clinical course progresses, they present loss of verbal (25/30, 83.3%; 3 N/A) and motor (21/31, 67.7%; 2 N/A) functions. On brain MRI, cerebral (90.6%) and cerebellar (68.8%) atrophy and white matter hyperintensity (87.5%) were the main findings. Autonomic features were reported in 65.0% and epilepsy in 39.4%. Among the other clinical findings, dysphagia (88.0%), sleep disturbances (64.0%), scoliosis (52.0%), and apnea/hypopnea (40.0%) were reported as the most common signs. Dysmorphic features were described in 6 patients. At the last follow-up (mean age 14.8 years, range: 3.8–33), 2 patients had died. Similarly, the patient carrying the predicted GoF c.608A>G change, who was last evaluated at 4 years, presented with neurodevelopmental regression, dystonia, and behavioral abnormalities, confirming the presence of the diagnostic triad. On the contrary, the patient carrying the c.349A>T had a milder clinical course with no evidence of regression or movement disorders up to the last follow-up at 10 years. This observation suggests that exon 7 is not only a hot spot for UBTF-related disease but can also be associated with a more severe phenotype with neurodevelopmental regression. Accordingly, differences between patients carrying exon 7 variants and those carrying variants outside exon 7 are presented in Figures 1–3.

A LoF behavior was predicted for 11 UBTF truncating variants, including 3 splice-site substitutions (i.e., c.1048–2A>G, c.1625_1626+15del, and the c.905+1G>A, identified in 4 family cases), 1 nonsense variant (i.e., c.1327C>T), 1 missense change (i.e., c.271G>C, p.Asp91His), and 2 microdeletions [i.e., 17q21.31(42248053_42294092) ×1, and 17q21.31(42239218–42345522) ×1]. Overall, these patients showed a milder clinical course predominantly characterized by developmental delay without regression (90.0%), with only one individual showing developmental regression at the age 5 years. All patients retained independent ambulation and expressive language abilities. No patient had passed away at the last follow-up (mean age 12.6 years, range: 3–34). Clinically, behavioral abnormalities were the predominant findings (77.8%), whereas movement disorders were sporadically observed. Nystagmus was reported in a single case. Autonomic manifestations were present in 33.3% of cases and epilepsy in 18.2% of cases. Dysmorphic features were described in 7 patients. Among other clinical findings, sleep disturbances, scoliosis, and apnea/hypopnea were reported in 25.0% of cases.

From a neuroradiologic perspective, abnormalities on brain MRI were observed in 40.0% of cases. In particular, white matter hyperintensity and atrophy of the brain and cerebellum were observed only in the patient who carried the c.1048-2A>G variant. Other MRI findings included brainstem atrophy (30.0%), reduced pons volume (30.0%), and cervical spinal cord atrophy (20.0%). No patient had passed away at the last follow-up.

Similarly, patients carrying putative hypomorphic variants (c.517G>A, c.495C>A, c.761A>G), whose disease mechanism has not clearly been defined yet, had a milder clinical course with no evidence of movement abnormalities and neuroregression at their last evaluation (mean age 6.3 years, range: 1.9–14).

Discussion

UBTF-related disorders comprise an ultra-rare, dominant, and clinically heterogeneous spectrum caused by pathogenic variants in UBTF, mostly occurring as de novo events. Disease was initially described in association with GoF variants; more recently, however, LoF variants have been recognized to cause a clinically distinct condition. Deep-phenotyping studies are essential to delineate the natural history of these conditions and to define clinically relevant genotype-phenotype correlations, guiding the prognosis and management of patients. These insights are particularly timely considering the increasing therapeutic relevance of applying antisense oligonucleotides to address autosomal dominant variants.17

In our study, we analyzed a cohort of 49 patients with genetically confirmed UBTF-related presentations, including 23 previously unreported cases. Our findings confirm that the clinical spectrum includes an early-onset neurodevelopmental condition, typically presenting before age 3 years (89.6%) with developmental delay and/or regression. Language impairment emerged as the most consistently affected developmental domain, highlighting the importance of early speech and communication interventions in patient management. Details on specific motor milestones and language skills before regression or at the peak of development from standardized scales were not available, and although this represents a limitation ofthis study, it paves the way for a prospective natural history study.

The high frequency of behavioral abnormalities and movement disorders during disease progression supports their role as hallmark features of UBTF-related phenotype that may facilitate differential diagnosis, early recognition, and timely pharmacologic and behavioral interventions. Epilepsy occurred predominantly in individuals with developmental regression, and both deceased patients had a history of seizures, suggesting a possible association between epilepsy and a more severe disease course. However, given the retrospective design, the limited number of events, and the absence of detailed electroclinical characterization, this observation should be confirmed in prospective studies.

Our data further indicate that the UBTF-related presentation is associated with substantial morbidity despite relatively low mortality. Neuroimaging findings suggest that cerebral and cerebellar atrophy, corpus callosum thinning, and involvement of the thalami and basal ganglia are mainly associated with the neuroregressive phenotype and may represent markers of more severe disease progression. In contrast, white matter hyperintensities and, less commonly, brainstem and cervical spinal cord abnormalities may also occur in patients with milder phenotypes, suggesting that structural neuroimaging alterations are not exclusively linked to severe clinical presentations. In patients requiring ventilatory support, MRI of the brain did not reveal major abnormalities supporting a central structural cause. MRS findings, although nonspecific, further supported the presence of an underlying neurodegenerative process, most commonly showing reduced N-acetylaspartate peaks, altered choline metabolism, increased myo-inositol, and, less frequently, lactate elevation. These observations are consistent with evidence from ubtf knockout and knockdown mouse models.18 In detail, the CNS conditional homozygous ubtf−/− knockout was embryonically lethal, whereas the heterozygous ubtf+/− knockout presented an adult onset of movement disorders and behavioral abnormalities associated with cerebellum 53BP1 immunoreactivity.18

A substantial proportion of patients were initially misdiagnosed based on clinical and neuroimaging findings, highlighting the phenotypic overlap of UBTF-related disorders with other neurodevelopmental and neurodegenerative conditions, and the pivotal role of next-generation sequencing in establishing the correct diagnosis.

From a molecular perspective, besides the recurrent c.628G>A variant and the other previously described,12,13 we identified 7 novel variants. The c.628G>A UBTF variant has been demonstrated to exert a GoF effect on UBTF using in vitro models with increased rDNA expression and consequent chromatin rearrangement and enlarged nuclei.4 In vivo mouse embryonic stem cells and mouse embryonic fibroblasts carrying the E210K mutation showed that the substitution within the HMGbox2 domain disrupts the cooperative interaction between UBTF and the SL1 complex, leading to reduced rDNA transcription, decreased RNA synthesis, and impaired cell proliferation.19 Instead, the pathogenic mechanisms of the other predicted-damaging variants remain to be clarified.

Our genotype-phenotype analysis demonstrated a consistent clinical phenotype in patients carrying the bona fide GoF c.628G>A and the predicted activating c.608A>G variants, clustering in exon 7, characterized by the triad of neurodevelopmental regression, movement disorders, and behavioral abnormalities with brain atrophy on MRI in 90.9% of our cohort. In this subgroup of patients, the disease was fatal in 2 cases. Notably, exon 7 encodes part of the highly conserved HMG-box 2 domain, which plays a critical role in DNA binding, chromatin remodeling, and regulation of ribosomal DNA transcription. The clustering of recurrent activating variants in this domain further supports its functional relevance in CONDBA syndrome, suggesting that disruption of HMG-box 2 may contribute to neurodegeneration through altered rRNA transcription and nucleolar dysfunction. A different clinical phenotype was instead identified in patients carrying pLoF variants [c.1625_1626+15del, c.1327C>T, c.1048-2A>G, c.905+1G>A, or c.271G>C1, 7q21.31(42248053_42294092) 17q21.31(42239218-42345522)], presenting with developmental delay without neuroregression in almost all cases (90.0%) and with normal brain MRI in 60.0% of the cohort. During the clinical course, patients in this subgroup preserved ambulation and expressive language and had no pyramidal or extrapyramidal movement disorders. Behavioral abnormalities (77.8%) and dysmorphic features (77.8%) were instead recurrent. A similar clinical course was presented by patients carrying putative hypomorphic variants with a disease mechanism not clearly defined. No deaths were reported in these subgroups of patients. Recently, Yammine et al.20 and Yi et al.21 reported additional LoF variants, not included in our formal analysis, associated with neurodevelopmental phenotypes without neuroregression, further supporting our observations. All together, these findings bring evidence of the molecular and clinical heterogeneity within UBTF-related disorders and expand the spectrum of LoF and hypomorphic variants.

Our results suggest a classification of the UBTF-related clinical spectrum into 2 distinct disorders: (1) a childhood-onset neurodegenerative disorder characterized by the triad of neuroregression, behavioral abnormalities, and movement disorders, previously known as CONDBA, and due to GoF variants in the hotspot HMG-box2 domain; (2) a childhood-onset neurodevelopmental disorder without neurodegenerative features and milder clinical course (NDD-UBTF), due to LoF and/or hypomorphic variants in UBTF. Although functional studies were not feasible for the novel variants and the pathogenicity was inferred through prediction tools and ACMG criteria, representing a limitation of our study, the proposed classification is supported by recently published studies reporting the association of LoF variants with neurodevelopmental delay without neuroregression.20,21

Our observations are also limited by the retrospective nature of the data collection, which, although standardized by eCRF, lacks some information on the phenotype and follow-up of the patients including the absence of systematically administered standardized developmental scales, thus limiting the characterization of developmental delay and regression severity.

Nevertheless, our findings help identify key clinical domains that could be systematically assessed in prospective natural history studies using standardized and objective measures in a large cohort of patients. Moreover, our results have a translation implication for diagnosis, prognosis, management, and the future development of precision therapies.

Acknowledgment

The authors thank the family association Il Sorriso negli occhi di Richi for their support in patient engagement and advocacy.

Glossary

ADHD

attention-deficit hyperactivity disorder

CONDBA

childhood-onset neurodegeneration with brain atrophy

CRF

case report form

eCRF

electronic case report form

HMG

high-mobility group

MRS

magnetic resonance spectroscopy

N/A

not available

NAA

N-acetylaspartate

NDD

neurodevelopmental disorder

SL1

selectivity factor 1

SPECT

single-photon emission computed tomography

UBTF

upstream binding transcription factor

Author Contributions

G. Balletto: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. S. Carli: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data. A. Coppola: major role in the acquisition of data; analysis or interpretation of data. A.a Torella: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data. H.-F. Lee: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data. C. Korenke: major role in the acquisition of data. S. Lebon: major role in the acquisition of data. L. Sedlackova: major role in the acquisition of data; analysis or interpretation of data. A. Nagy: major role in the acquisition of data; analysis or interpretation of data. F.S. Eichler: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. C.N. Murali: major role in the acquisition of data. V. Nigro: major role in the acquisition of data. F.A. Ladha: major role in the acquisition of data. M.S. Ledoux: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data. S. Pelle: major role in the acquisition of data; analysis or interpretation of data. G. Prato: major role in the acquisition of data. L. Nobili: major role in the acquisition of data. L. Chiriatti: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data. C. Mancini: major role in the acquisition of data. G. Zampino: major role in the acquisition of data. C. Graziano: major role in the acquisition of data. K. Baranano: major role in the acquisition of data. P. Planté-Bordeneuve: major role in the acquisition of data. R. Caumes: major role in the acquisition of data. B. Lauerová: major role in the acquisition of data. J. Haberlová: major role in the acquisition of data. D.M. Cordelli: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data. C. Fiorillo: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data. M. Tartaglia: major role in the acquisition of data; analysis or interpretation of data. R.J. Tinker: major role in the acquisition of data; study concept or design. C. Garone: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data.

Study Funding

This work was partially supported by grants from the NIH (grant number R56NS123059 to M.S.LD. and T32GM082773-16 to RJ. T), Ministero della Salute (grants PNRR-MR1-2022-12376811and RF-2021-12374963, to M.T.) and Telethon Undiagnosed Diseases Program (TUDP, GSP1500124).

Disclosure

The authors report no relevant disclosures. Go to Neurology.org/NG for full disclosures.

References

  • 1.Gilissen C, Hehir-Kwa JY, Thung DT, et al. Genome sequencing identifies major causes of severe intellectual disability. Nature. 2014;511(7509):344-347. doi: 10.1038/nature13394 [DOI] [PubMed] [Google Scholar]
  • 2.Pietra A, Palombo F, Giannotta M, et al. Expanding the clinical spectrum of UBTF-Related neurodevelopmental disorder. Neurol Genet. 2023;9(6):e200098. doi: 10.1212/NXG.0000000000200098 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Anazi S, Maddirevula S, Salpietro V, et al. Expanding the genetic heterogeneity of intellectual disability. Hum Genet. 2017;136(11-12):1419-1429. doi: 10.1007/s00439-017-1843-2 [DOI] [PubMed] [Google Scholar]
  • 4.Edvardson S, Nicolae CM, Agrawal PB, et al. Heterozygous de novo UBTF gain-of-function variant is associated with neurodegeneration in childhood. Am J Hum Genet. 2017;101(2):267-273. doi: 10.1016/j.ajhg.2017.07.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Moss T, LeDoux MS, Crane-Robinson C. HMG-boxes, ribosomopathies and neurodegenerative disease. Front Genet. 2023;14:1225832. doi: 10.3389/fgene.2023.1225832 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Toro C, Hori RT, Malicdan MCV, et al. A recurrent de novo missense mutation in UBTF causes developmental neuroregression. Hum Mol Genet. 2018;27(7):1310. doi: 10.1093/hmg/ddy049 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Sedláčková L, Laššuthová P, Štěrbová K, et al. UBTF mutation causes complex phenotype of neurodegeneration and severe epilepsy in childhood. Neuropediatrics. 2019;50(1):57-60. doi: 10.1055/s-0038-1676288 [DOI] [PubMed] [Google Scholar]
  • 8.Bastos F, Quinodoz M, Addor MC, et al. Childhood neurodegeneration associated with a specific UBTF variant: a new case report and review of the literature. BMC Neurol. 2020;20(1):17. doi: 10.1186/s12883-019-1586-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Ikeda C, Kawarai T, Setoyama C, Orlacchio A, Imamura H. Recurrent de novo missense variant E210K in UBTF causes juvenile dystonia-parkinsonism. Neurol Sci. 2021;42(3):1217-1219. doi: 10.1007/s10072-020-04758-y [DOI] [PubMed] [Google Scholar]
  • 10.Gunasekaran PK, Laxmi V, Manjunathan S, Kumar A, Tiwari S, Saini L. Childhood-onset neurodegeneration with brain atrophy in association with c.628G>A in UBTF gene. Indian J Pediatr. 2023;90(8):835. doi: 10.1007/s12098-023-04642-5 [DOI] [PubMed] [Google Scholar]
  • 11.Chi CS, Lee HF, Tsai CR. Clinico-radiological phenotype of UBTF c.628G>A pathogenic variant-related neurodegeneration in childhood: a case report and literature review. Brain Sci. 2022;12(9):1262. doi: 10.3390/brainsci12091262 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Tinker RJ, Guess T, Rinker DC, et al. A novel, likely pathogenic variant in UBTF-related neurodegeneration with brain atrophy is associated with a severe divergent neurodevelopmental phenotype. Mol Genet Genomic Med. 2022;10(12):e2054. doi: 10.1002/mgg3.2054 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Wang X, Yang B, Wu S, et al. UBTF haploinsufficiency is associated with UBTF-related global developmental delay and distinctive facial features without neuroregression. J Med Genet. 2024;61(12):1089-1095. doi: 10.1136/jmg-2024-110061 [DOI] [PubMed] [Google Scholar]
  • 14.Chiriatti L, Priolo M, Leoni C, et al. UBTF haploinsufficiency-related disorder: report of a new case series and definition of the facial gestalt. Clin Genet. 2026;109(5):964-971. doi: 10.1111/cge.70126 [DOI] [PubMed] [Google Scholar]
  • 15.Richards S, Aziz N, Bale S, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17(5):405-424. doi: 10.1038/gim.2015.30 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Hwang YJ, Han D, Kim KY, et al. ESET methylates UBF at K232/254 and regulates nucleolar heterochromatin plasticity and rDNA transcription. Nucleic Acids Res. 2014;42(3):1628-1643. doi: 10.1093/nar/gkt1041 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Crooke ST. Addressing the challenges of treating patients with heterozygous gain of function mutations. Nucleic Acid Ther. 2024;34(6):273-275. doi: 10.1089/nat.2024.0060 [DOI] [PubMed] [Google Scholar]
  • 18.Hori RT, Moshahid Khan M, Xiao J, Hargrove PW, Moss T, LeDoux MS. Behavioral and molecular effects of Ubtf knockout and knockdown in mice. Brain Res. 2022;1793:148053. doi: 10.1016/j.brainres.2022.148053 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Tremblay MG, Sibai DS, Valère M, et al. Ribosomal DNA promoter recognition is determined in vivo by cooperation between UBTF1 and SL1 and is compromised in the UBTF-E210K neuroregression syndrome. PLoS Genet. 2022;18(2):e1009644. doi: 10.1371/journal.pgen.1009644 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Yammine T, Mercier S, Poirsier C, et al. Commentary on UBTF haploinsufficiency associated with UBTF-related global developmental delay and distinctive facial features without neuroregression. J Med Genet. 2025;62(10):653-655. doi: 10.1136/jmg-2025-110686 [DOI] [PubMed] [Google Scholar]
  • 21.Yi S, Fan L, Zhang Q, et al. Frameshift variants in the UBTF gene are associated with neurodevelopmental disorders. Clin Chim Acta. 2026;583:120841. doi: 10.1016/j.cca.2026.120841 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The anonymized data can be accessed on reasonable request addressed to the corresponding author.


Articles from Neurology: Genetics are provided here courtesy of American Academy of Neurology

RESOURCES