SETBP1 variants outside the degron disrupt DNA-binding, transcription and neuronal differentiation capacity to cause a heterogeneous neurodevelopmental disorder

Abstract

Different types of germline de novo SETBP1 variants cause clinically distinct and heterogeneous neurodevelopmental disorders: Schinzel-Giedion syndrome (SGS, via missense variants at a critical degron region) and SETBP1-haploinsufficiency disorder. However, due to the lack of systematic investigation of genotype-phenotype associations of different types of SETBP1 variants, and limited understanding of its roles in neurodevelopment, the extent of clinical heterogeneity and how this relates to underlying pathophysiological mechanisms remains elusive. This imposes challenges for diagnosis. Here, we present a comprehensive investigation of the largest cohort to date of individuals carrying SETBP1 missense variants outside the degron region (n = 18). We performed thorough clinical and speech phenotyping with functional follow-up using cellular assays and transcriptomics. Our findings suggest that such variants cause a clinically and functionally variable developmental syndrome, showing only partial overlaps with classical SGS and SETBP1-haploinsufficiency disorder. We provide evidence of loss-of-function pathophysiological mechanisms impairing ubiquitination, DNA-binding, transcription, and neuronal differentiation capacity and morphologies. In contrast to SGS and SETBP1 haploinsufficiency, these effects are independent of protein abundance. Overall, our study provides important novel insights into diagnosis, patient care, and aetiology of SETBP1-related disorders.

Fig. 1. SETBP1 variants outside the degron cluster in the SKI domain, facial photographs and speech phenotyping of individuals with SETBP1 variants.

a Schematic representation of the SETBP1 protein (UniProt: Q9Y6X0) indicating the locations of variants included in this study. These comprise eleven novel germline variants (circles), including ten missense variants (blue) and one in-frame deletion (de novo, green) outside the canonical degron. Three missense variants outside (blue) and nine within (orange) the degron that were previously reported (diamond) are also annotated in the schematic. ^§ represent variants for which speech phenotyping data were available. ^# represent variants included in functional assays. Five exons (black bars) encode isoform A of the protein (NP_056374.2, 1596 amino acids). The known SETBP1 protein domains are indicated^,,. An overview with variant details per subject is provided in Supplementary Data 1. b For individuals with an amino acid change in the 854–858 region, no visually recognisable facial features could be delineated. Shared facial features present in at least three out of five individuals with an amino acid change in 862–874 include prominent ears, shallow orbits, midface retraction, and microcephaly. Individuals with a 962 and 957 variant show similar facial features, including a round face, blepharophimosis, ptosis, hypertelorism, and a short nose with a bulbous tip, features that are also often noted in individuals with SETBP1 haploinsufficiency disorder. c PhenoScore is able to distinguish individuals with SGS and SETBP1-haploinsufficiency disorder. Subgroup analysis of (left) facial photographs and (right) phenotypic data (HPO terms) of five individuals with SGS and five individuals with SETBP1-haploinsufficiency disorder. Score: 0 = SETBP1-haploinsufficiency disorder; 1 = SGS. d Phenotypic similarity of individuals with missense SETBP1 variants outside the degron predicted using PhenoScore (purple circle), HPO terms only (green) or facial features only (blue). Score: 0 = SETBP1-haploinsufficiency disorder; 1 = SGS. Each datapoint represents one individual (n = 18). Details in Supplementary Data 2. e Performance on Vineland-3 subtests. Lines denote median scores; X denotes mean scores; ABC, Adaptive Behaviour Composite. Standard scores between 85 and 115 are considered within the average range, with a mean of 100 and a standard deviation of 15. The performance of five individuals was measured. All details of statistical tests and p-values are provided in Source Data 2. Source data are provided as a Source Data file.

Fig. 2. SETBP1 variants outside the degron show increased protein abundance but do not affect SET binding or pPP2A/PP2A ratio in patient fibroblasts.

a Immunoblot of whole cell lysates of control and patient human dermal fibroblasts (HDF) probed with anti-SETBP1, anti-SET, anti-pPP2A (T307) and anti-PP2A antibodies. β-actin was used as a loading control. b Quantification of protein levels of SETBP1 variants normalised to β-actin (right). Bars represent the mean ± SEM of three independent experiments (vs controls, one-way ANOVA and a post-hoc Dunnett’s test). c Immunoblot of whole cell lysates of HEK293T/17 cells expressing FLAG-tagged SETBP1 variants probed with anti-SETBP1 and anti-FLAG antibodies. β-actin was used as a loading control (left). Representative blots of three independent experiments are shown. d Quantification of protein levels of FLAG-tagged SETBP1 variants normalised to β-actin (right). Values are expressed relative to wild type (WT) and represent the mean ± SEM of three independent experiments (*p < 0.05, **p < 0.01, ***p < 0.001, using one-way ANOVA and a post-hoc Dunnett’s test). e Normalised SET transcript expression (bottom) in control and patient HDFs. Bars represent the mean ± SEM of three independent experiments (vs controls, one-way ANOVA and a post-hoc Dunnett’s test). f Quantification of pPP2A/PP2A ratio. Bars represent the mean ± SEM of three independent experiments (vs controls, one-way ANOVA and a post-hoc Dunnett’s test). Source data are provided as a Source Data file. All details of statistical tests and p-values are provided in Source Data 4.

Fig. 3. Variable impairment of SETBP1 degradation via proteasome and autophagy pathways is associated with partial reduction of ubiquitination.

a Fibroblasts derived from healthy controls and patients carrying SETBP1 variants outside the degron were treated with a translation inhibitor, cycloheximide (CHX; 100 μg/ml) or vehicle control DMSO for 4 h. Immunoblots of whole cell lysates probed with an anti-SETBP1 antibody were shown. β-actin was used as a loading control. Results are representative of three independent experiments. b Relative fluorescence intensity of YFP-tagged SETBP1 variants overexpressed in HEK293T/17 cells treated with translation inhibitor cycloheximide (CHX; 50 µg/ml; top), proteasomal degradation inhibitor MG132 (5 µg/ml; middle), or autophagy inhibitor Bafilomycin A1 (BafA1; 100 nM; bottom). An equal volume of DMSO was used as a vehicle control. Fluorescence intensity was measured for 24 h with 3-h intervals and normalised to an mCherry transfection control. Values are expressed relative to t = 0 h and represent the mean ± SEM of three independent experiments, each performed in triplicate (^*p < 0.05, ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001; two-way ANOVA and a post-hoc Dunnett’s test). Details of statistical tests and p-values are provided in Source Data 4. c Immunoprecipitation of FLAG-tagged SETBP1 wild type and variants using FLAG-conjugated magnetic agarose and blotted with an anti-FLAG, anti-SETBP1 or anti-ubiquitin antibody. β-actin was used as a loading control. d Quantification of SETBP1 (top), ubiquitin (middle) in FLAG-IP fractions for inherited (outside degron, grey) and SGS variants (orange, left), and de novo variants outside the degron (right). Ratio of ubiquitin/SETBP1 in the FLAG-IP fractions was plotted (bottom). Bars represent the mean ± SEM of three independent experiments (^*p < 0.05, vs WT; one-way ANOVA and a post-hoc Dunnett’s test). Details of statistical tests and p-values are provided in Source Data 4. Source data are provided as a Source Data file.

Fig. 4. Genotype-specific reduction of binding capacity to AT-rich DNA sequences and transcriptional activation for SETBP1 variants outside the degron.

a Results of luciferase assays with constructs containing WT and SETBP1 variants, and the reporter constructs with the consensus SETBP1 binding sequences. Values are expressed relative to the control condition, which used a pCMV-YFP construct without SETBP1. b Results of the M1H assay for SETBP1 transcriptional regulatory activity with WT and SETBP1 variants fused with an N-terminal GAL4 in combination with a reporter construct with or without the GAL4-binding site. Values are expressed relative to the control condition, which used a pBIND2-GAL4 construct without SETBP1. c) Results of luciferase assays with constructs containing WT and SETBP1 variants, and reporter constructs with FOXP2 promoters: TSS1 (left) and TSS2 (right). Values are expressed relative to the control condition, which used a pCMV-YFP construct without SETBP1. All graphs for luciferase assays show the mean ± SEM of three independent experiments, each performed in triplicate (^*p < 0.05, ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001 vs WT; one-way ANOVA and a post-hoc Dunnett’s test). All graphs for the mammalian-one-hybrid assay show the mean ± SEM of three independent experiments, each performed in triplicate (^*p < 0.05, ^***p < 0.001, ^****p < 0.0001 vs reporter without GAL4-binding site; ^#p < 0.05, ^###p < 0.001, ^####p < 0.0001 vs WT; two-way ANOVA and a post-hoc Tukey’s test). Details of statistical tests and p-values are provided in Source Data 5. Source data are provided as a Source Data file.

Fig. 5. Fibroblasts carrying SETBP1 variants outside the degron showed distinct transcriptomic profiles from healthy controls, SGS and SETBP1-haploinsufficiency disorder patients.

a Schematic representation of the SETBP1 protein (UniProt: Q9Y6X0) indicating the locations of variants included in the RNA-seq experiment. b Fibroblasts derived from patients harbouring a SETBP1 variant (missense or in-frame deletion) outside the degron formed separate clusters from healthy controls, and partially overlapped with those from SGS and SETBP1-haploinsufficiency disorder patients. Principal component analysis (PCA) plot of variance distribution of eight control fibroblast lines (black), two lines carrying a variant within the degron (orange), six lines carrying a variant outside the degron (blue), and three lines carrying a truncating variant (purple); three technical replicates were included for all patient lines and control lines 5–8; six replicates were included for control lines 1–4. A list of cell lines included in the RNA-seq experiment can be found in Supplementary Data 5. Principal components (PC) 1 and PC2 accounted for 30% and 13% of total variance, respectively. c Volcano plots showing the gene expression data of control vs patient fibroblast lines (left: within degron; middle: outside degron; right: truncating) as a function of log ratios and mean average gene counts. The number of significant DEGs that are up- and down-regulated are shown (log₂ fold change ≥ 1 or ≤−1, p < 0.05, multiple testing correction with FDR). d Venn diagram showing the overlap of genes that demonstrated significant differential expression (p < 0.05, multiple testing correction with FDR) in control vs SGS variants within degron (orange), variants outside degron (blue) and truncating variants (purple). e Top four dysregulated GO biological processes revealed by over-representation analysis of the up- (top) and down-regulated (bottom) DEGs in patient fibroblasts carrying different SETBP1 variants (p < 0.05, multiple testing correction with FDR; orange: within degron; blue: outside degron; purple: truncating). f RRHO heatmaps showed transcriptomic overlaps between two comparisons. Bottom left corner of the heatmap indicates overlap signal from up-regulated genes in both conditions. The top right corner of the heatmap indicates an overlap signal from down-regulated genes in both conditions. Source data are provided as a Source Data file (Source Data 6–8).

Fig. 6. Induced neurons carrying different SETBP1 variant groups showed distinct differentiation capacity and morphology, and transcriptomic profiles.

a Schematic representation of the SETBP1 protein (UniProt: Q9Y6X0) indicating the locations of variants included. b Schematic representation of the generation of induced neurons from fibroblasts. Asterisks indicate the time of harvest. CA: compound set A; CB: compound set B. c Confocal microscopy images of immunostained neuronal markers Tuj1 (green), MAP2 (orange), and DCX (magenta) in control and patient-derived induced neurons. Nuclei were stained with Hoechst 33342 (white). Merged images are shown. Results are representative of three independent experiments. Scale bar = 100 μm. d Quantification of the proportion of Tuj1-positive cells (%Tuj1/Hoechst) at days 10 and 12. Bars represent the mean ± SEM of three independent experiments (vs controls; one-way ANOVA and a post-hoc Dunnett’s test). Details of statistical tests and p-values are provided in Source Data 9. e Zoomed in confocal microscopy images of immunostained neuronal markers Tuj1 (green) in control and patient-derived induced neurons. Nuclei were stained with Hoechst 33342 (white). Results are representative of three independent experiments. Scale bar = 50 μm. f Quantification of neuronal morphology using Sholl analysis. Violin plots of three independent experiments (vs controls; one-way ANOVA and a post-hoc Dunnett’s test). Brown crossbars indicate the median. At least 40 neurons per cell line from three independent differentiation experiments were measured. Details of violin plot statistics, statistical tests and p-values are provided in Source Data 9. g Sholl profile of neuronal morphology. Data are from three independent experiments. Individual data points represent the number of dendritic intersections per radial distance from the soma. Solid lines show the predicted mean per condition, fitted using cubic polynomial regression (3rd-degree) via linear modelling; shaded bands represent 95% confidence intervals around the model fit. Statistical analysis was performed using one-way ANOVA and a post-hoc Tukey’s test. At least 40 neurons were analysed per cell line. Details of statistical tests and adjusted p-values are provided in Source Data 9. Source data are provided as a Source Data file.

Fig. 7. Induced neurons carrying different SETBP1 variant groups showed distinct transcriptomic profiles.

a PCA plot of induced neurons derived from control (red), variant within the degron (green), variant outside the degron (blue), and truncating variant (purple); two lines each from patient variant groups and control in each differentiation; three independent differentiation experiments were included for all patient lines and control lines. A list of cell lines included in differentiation and RNA-seq is included in Supplementary Data 7. b Volcano plots showing the gene expression data of control vs patient induced neurons (left: within degron; middle: outside degron; right: truncating) as a function of log ratios and mean average gene counts. The number of up- and down-regulated DEGs that had a log₂ fold change ≥1 or ≤−1 are shown (p < 0.05, multiple testing correction with FDR). c Venn diagram showing the overlap of genes that demonstrated significant differential expression (p < 0.05, multiple testing correction with FDR) in control vs SGS variants within degron (orange), variants outside degron (blue) and truncating variants (purple). d Top 10 dysregulated GO biological processes revealed by over-representation analysis of the up- (top) and down-regulated (bottom) DEGs in patient-induced neurons (p < 0.05, multiple testing correction with FDR; orange: within degron; blue: outside degron; purple: truncating). Source data are provided as a Source Data file. Details of statistical tests and p-values are provided in Source Data 10−11.

Fig. 8. Summary of functional assays: variants outside the degron disrupt a spectrum of SETBP1 functions by dysregulating transcription, reducing DNA binding capacity, variable protein degradation impairment and ubiquitination independent of SETBP1 level or the SET/PP2A pathway.

A heat map summarising the functional characterisation of SETBP1 variants within (classical SGS) and outside the degron. Classical SGS variants showed increased protein stability and higher SETBP1 protein levels. While the binding capacity of SGS variants to AT-rich DNA sequences was not affected, a subset showed higher transcriptional activation, suggesting a gain of function. In contrast, SETBP1 variants outside the degron demonstrated a broad spectrum of functional disruptions, which could be largely categorised into two groups. Although the majority of these variants resulted in more abundant SETBP1 protein, there were variable degrees of disruption in SETBP1 degradation via the proteasome and autophagy machinery. Variants furthest from the degron (grey) showed mostly reduced SETBP1 degradation by both proteasome and autophagy due to disrupted ubiquitination, while only a subset of those in the proximity of the degron demonstrated impaired degradation and mildly reduced ubiquitination. Although increased proliferation was seen for patient fibroblasts carrying two variants outside the degron, suggesting a partially overlapping mechanism with classical SGS, SET expression levels or pPP2A/PP2A ratio (indicative of PP2A activity) were not affected. Two variants outside the degron led to lower affinity to AT-rich DNA sequences, suggesting loss of function. Transcriptional activation was affected in the majority of variants outside the degron to various degrees. Patient fibroblasts carrying variants outside the degron showed different transcriptomic profiles compared to healthy controls, and to those from SGS and SETBP1-haploinsufficiency disorder patients, with differentially expressed genes involved in different gene ontology biological processes. Patient-induced neurons with different types of SETBP1 variants showed different differentiation capacity, morphology, and transcriptomic profiles. HD: SETBP1-haploinsufficiency disorder.