Mutations in the spliceosomal gene SNW1 cause neurodevelopment disorders with microcephaly

Abstract

The spliceosome is a critical cellular machinery responsible for pre-mRNA splicing that is essential for the proper expression of genes. Mutations in its core components are increasingly linked to neurodevelopmental disorders, such as primary microcephaly. Here, we investigated the role of SNW domain-containing protein 1 (SNW1), a spliceosomal protein, in splicing integrity and neurodevelopment. We identified 9 heterozygous mutations in the SNW1 gene in patients presenting with primary microcephaly. These mutations impaired SNW1's interactions with core spliceosomal proteins, leading to defective RNA splicing and reduced protein functionality. Using Drosophila melanogaster and human embryonic stem cell-derived cerebral organoids models, we demonstrated that SNW1 depletion resulted in significant reductions in neural stem cell proliferation and increased apoptosis. RNA-Seq revealed disrupted alternative splicing, especially skipping exons, and altered expression of neurodevelopment-associated genes (CENPE, MEF2C, and NRXN2). Our findings provide crucial insights into the molecular mechanisms by which SNW1 dysfunction contributes to neurodevelopmental disorders and underscore the importance of proper spliceosome function in brain development.

Keywords: Embryonic stem cells; Genetic diseases; Genetics; Neurodevelopment; Neuroscience.

Figure 1. Mutations in SNW1 lead to microcephaly and impair SNW1 functions in human.

(A) Schematic diagram of the SNW1 transcript (NM_012245.3, intron not to scale) (top panel) and schematic outline of the SNW1 protein domains (lower panel) with the locations of 9 loss-of-function variants identified in our study. The splice donor variants are shown in purple. The 2 variants reported in the literature are shown in green. (B) The c.330+2T>C construct showed complete skipping of exon 3 and a partial 63-bp skip. The c.426+1G>A and c.426+1G>T constructs showed complete skipping of exon 4. (C) Fluorescence images of the HEK293T cells after transfection with SNW1_WT or F412Lfs*17 vectors. pmCherry-C1 was used as an internal control and cotransfected with the WT and F412Lfs*17 vectors at the same ratio. Scale bars: 100 μm. (D) Expression analysis of SNW1 by Western blotting was performed in lysates from HEK293T cells transfected with either SNW1 WT or F412Lfs*17 vectors. (E) qPCR analysis for SNW1 in HEK293T cells transfected with SNW1 WT or F412Lfs*17 vectors. (F) qPCR analysis for SNW1 in SNW1 WT (left) or F412Lfs*17 (right) HEK293T cells after being stimulated by the NMD inhibitor cycloheximide (CHX; 100 μg/mL). (G) Overexpression of C-terminal FLAG-tagged WT and SNW1 variants in HEK293T cells. GAPDH served as a loading control. Quantification of overexpressed FLAG-tagged SNW1 proteins. (H) Effects of mutations on the localization of SNW1 in HEK293T cells. Fluorescence images were captured using a laser scanning confocal microscope (Leica TCS SP8) with ×63 oil glass. SNW1 (green) and DAPI (blue) are displayed. Scale bars: 2.5 μm. (I and J) Cryo-electron microscopy structure of the human spliceosome ILS complex (Protein Data Bank ID 6id0) highlighting SNW1 (surface in pink) and its interacting proteins in the spliceosome, including PPIL1 (sky blue), PLRG1 (olive drab), and PRPF8 (brown). Residues in patients were observed to be located at the interface where SNW1 interacts with these proteins, suggesting changes in molecular interactions. Data are presented as mean ± SEM. For comparisons among multiple groups, ordinary 1-way ANOVA was performed followed by Tukey’s multiple-comparison test for post hoc pairwise analysis. For comparisons between 2 independent groups, 2-tailed unpaired Student’s t tests were used (without post hoc correction). **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 2. SNW1 interacted with PPIL1, PLRG1, and PRPF8.

(A) Structure-based protein interaction interface analysis between SNW1 (pink) and PPIL1 (sky blue), where interaction hotspot residues are labeled. (B) CoIP of SNW1 and PPIL1 was performed. HEK293T cells were transfected with plasmids encoding SNW1-FLAG and PPIL1-HA. The cell lysates were subjected to anti-FLAG and anti-HA IP, followed by analysis via Western blotting. (C) Subcellular localization analysis of SNW1 and PPIL1. HEK293T cells were transfected with the plasmid encoding pEGFP-SNW1 and pmCherry-PPIL1 for 24 hours and then fixed and stained with DAPI. SNW1 (green), PPIL1 (red), and DAPI (blue) are displayed. Scale bars: 2.5 μm. (D) Structure-based protein interaction interface analysis between SNW1 (pink) and PLRG1 (olive drab), where interaction hotspot residues are labeled. (E) CoIP of SNW1 and PLRG1 was performed. HEK293T cells were transfected with plasmids encoding SNW1-FLAG and PLRG1-HA. The cell lysates were subjected to anti-FLAG and anti-HA IP, followed by analysis via Western blotting. (F) Structure-based protein interaction interface analysis between SNW1 (pink) and PRPF8 (brown), with interaction hotspot residues labeled. (G) CoIP of SNW1 and PRPF8 was performed. HEK293T cells were transfected with plasmids encoding SNW1-FLAG and PRPF8-HA. The cell lysates were subjected to anti-FLAG and anti-HA IP, followed by analysis via Western blotting.

Figure 3. Bx42 knockdown in NSCs leads to reduced brain lobe volume, stem cell number, and percent proliferating stem cells.

(A) Images of third-instar larval brains with knockdown of EGFP or Bx42 in NSCs (insc-GAL4). (B) The brain lobe volume of EGFP (control) and Bx42 knockdown in NSCs. (C) Bx42 transcript expression (normalized to RpL32) in control and Bx42 RNAi third-instar brains. (D) Confocal images of a single brain lobe from third-instar larvae with knockdown of EGFP or Bx42 in NSCs. Brains were stained for Deadpan (Dpn), a nuclear marker of NSCs, and p-HH3, a marker for proliferating cells. The central brain region is outlined in white. (E) The number of Dpn⁺ NSCs in the central brain region in Drosophila with EGFP (n = 10) and Bx42 (n = 8) knocked down in NSCs. (F) Proliferating NSCs (yellow arrowheads in D) and non-proliferating (white arrowheads in D) in EGFP- (n =10) and Bx42-knockdown (n = 8) brains were quantified. Dpn⁺ cells without p-HH3 puncta are noted with white arrowheads in D. There was complete loss of proliferating NSCs in third-instar larvae with Bx42 knockdown using insc-GAL4. (G) SNW1-HA was expressed in NSCs (insc-GAL4) and stained with Dpn (magenta), HA (green), or DAPI (white) to confirm presence of SNW1 protein in the nucleus of NSCs. (H) Images of third-instar larval brains with knockdown of EGFP, Luciferase + Bx42 RNAi, or SNW1_WT + Bx42 RNAi in NSCs (insc-GAL4). (I) Brain lobe volume of genotypes from (H); each dot represents 1 brain. (J) The brain lobe volume of Drosophila expressing SNW1 alone in NSCs was analyzed. (K) Confocal images of a single brain lobe from third-instar larvae with knockdown of EGFP, Luciferase + Bx42 RNAi, or SNW1_WT + Bx42 RNAi in NSCs. The central brain region is outlined in white. (L) Proliferating cells were quantified in EGFP (n = 8), Luciferase + Bx42 RNAi (n = 6), and SNW1_WT + Bx42 RNAi (n = 6) knockdown brains. (M) Proliferating NSCs (yellow arrowheads in K) and non-proliferating (white arrowheads in K) were quantified in EGFP (n = 8), Luciferase + Bx42 RNAi (n = 6), and SNW1_WT + Bx42 RNAi (n = 6) knockdown brains. Data are presented as mean ± SEM. For comparisons among 3 groups, ordinary 1-way ANOVA was performed followed by Tukey’s multiple-comparison test for post hoc pairwise analysis. For comparisons between 2 independent groups, 2-tailed unpaired Student’s t tests were used (without post hoc correction). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Scale bars: 50 μm (A and H), 40 μm (D and K), and 20 μm (G).

Figure 4. Developmental progression and phenotypic analysis of SNW1^+/– cerebral organoids reveals altered size and NSC properties.

(A) Schematic of making cerebral organoids based on Lancaster methods. (B) Representation images of H9 and SNW1^+/– brain organoids cultured for 10, 20, 30, and 40 days. Scale bars: 500 μm. (C) The size of brain organoids from different groups was quantified at multiple time points across 3 independent experiments. (D) The ventricular zone–like (VZ-like) PAX6⁺ rosette area was quantified, with each data point representing an individual rosette. Data were collected from 3 independent experiments. (E–H) Immunofluorescent staining was performed on sections of WT and SNW1^+/– brain organoids cultured for 45 days. PAX6 was used to label the VZ-like rosette area, which was further stained with p-HH3 (E), Ki67 (F), Caspase3 (G), and MAP2 (H). The proliferating rosettes and apical surface adjusted to ventricle-like regions are highlighted by white dash lines. Scale bars: 50 μm. (I) Quantification of the ratio of p-HH3⁺ cells per 100,000 μm² PAX6⁺ area of each rosette. Each plot represents an individual rosette. (J) Quantification of the ratio of Ki67 and PAX6 double-positive cells versus the total number of PAX6⁺ cells of each rosette. Each plot represents an individual rosette. (K) Quantification of the ratio of Caspase3⁺ cells per 100,000 μm² PAX6⁺ area of each rosette. Each plot represents an individual rosette. (L) Quantification of the ratio of MAP2 and PAX6 double-positive cells versus the total number of PAX6⁺ cells of each rosette. Each plot represents an individual rosette. Small gray symbols represent size measurements of single hCOs (technical replicates), different symbol shapes denote 3 independent biological replicate experiments (n = 3 per group), and large colored symbols indicate means of technical replicates within each biological replicate. Data are presented as mean ± SEM. Statistical significance was tested by 1-way ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 5. Developmental and transcriptomic analysis of human cerebral organoids derived from SNW1^+/– hESCs.

(A) Schematic overview of the experimental design. hESCs were differentiated into NPCs over 12 days and into cerebral organoids over 45 days. NPCs cultured for 18 days, cerebral organoids cultured for 18 days, and cerebral organoids cultured for 45 days were collected for bulk RNA-seq analysis. (B) GO biological process analysis of DEGs is shown, highlighting common enriched terms across the 3 groups (NPCs, day 18 organoids, and day 45 organoids). (C) KEGG pathway enrichment analysis of DEGs is presented, displaying shared enriched pathways among the 3 groups (NPCs, day 18 organoids, and day 45 organoids). (D) Reactome pathway enrichment circos plot for 45-day cerebral organoids. For each Reactome term around the circle (outermost ring), the adjacent ring shows upregulated (red) versus downregulated (teal) gene counts; the next inner ring shows –log₁₀(P value) as a heat map; and the innermost ring displays gene ratio bars. (E) DisGeNET disease association circos plot for 45-day cerebral organoids, plotted in the same concentric-ring format as in D.

Figure 6. Splicing integrity defects in SNW1^+/– hCOs.

(A) Impact of SNW1 depletion on 5 major types of AS events detected with rMATS in hCOs (45 days old). SEs were most affected, followed by mutually exclusive exons (MXE), alternative 5′ splice site (A5SS), alternative 3′ splice site (A3SS), and RI. (B) Columns showing numbers of significant events with higher inclusion level in SNW1^+/– (sky blue) or control (pink). (C) Distribution of differential splicing identified by rMATS in SNW1^+/– or control on the basis of SE length. Exons with long length show significantly higher SEs in SNW1^+/– hCOs. (D) Violin plot of ESGs with significant transcript lengths. (E) Violin plot of isoform numbers of ESGs. (F) Venn diagram of overlapping DEGs and ESGs. (G and H) Violin plot of transcript lengths of DEGs compared with nonsignificant genes. NonSig, nonsignificant genes (gray); UP, upregulated (light green); Down, downregulated (sky blue). (I) Schematic representation of human CENPE transcript isoforms. The purple arrow indicates the Matched Annotation from NCBI and EMBL-EBI (MANE)–selected canonical transcript. The identified exon in the rMATS analysis is marked in pink. (J) Sashimi plots of read density of CENPE transcript in 4 SNW1^+/– (no. 1-2) and 4 control brain organoids revealed that SNW1^+/– hCOs exhibited retention of exon 38, while the WT tended to skip this exon. (K) RT-qPCR validations of SE events of CENPE in WT and SNW1^+/– brain organoids (no. 1-2). Top: Schematic diagrams of CENPE transcript. The pink box represents the SEs. Bottom: Validation of significant SE events by semiquantitative RT-qPCR using GAPDH as reference gene. Relative level of overall mRNA (sky blue); ratio of exon-included mRNA to total mRNA (pink). Statistical significance was determined by Wilcoxon’s test. ***P < 0.001.