Fig 1: HNRNPU+/- hCOs show disruption in ribosomal proteins and neurodevelopmental processes(A) Upregulated genes in HNRNPU+/- hCOs enriched for neurodevelopmental ontologies, including neurogenesis, morphogenesis and differentiation.(B) Downregulated genes enriched for functional annotations including DNA-binding and RNA-binding.(C) UMAP visualization of annotated clusters across all five samples for D11, M20 and PGP1. ‘npc’ = neuronal progenitor. ‘nsc’ = neural stem cell. ‘hmi’ = high metabolic intermediate progenitor. ‘scpn’ = sub-cortical projection neuron. ’opc’ = oligodendrocyte precursor.(D) Compositional analysis shows significant reduction (* = p< 0.05) in NPC and HMI populations for both D11 and M20 hCOs. Significant or trending significant (* = p< 0.06) increases in SCPN and GABAergic populations for both D11 and M20 hCOs.(E and F) show burden of dysregulated genes increased in precursor populations versus more mature neuronal population. Number of DEGs generated using 47 cells per cluster per sample to control for cell number and sample bias. FDR-correct pvalues and a log2 fold change threshold of 0.25 were used. Final DEGs averaged across 10 simulations to minimize impact of selection bias. Precursor populations were the four most (M20) or four of the five most (D11) impacted populations. Data are represented as mean ±SEM.
Fig 2: Consistent dysregulation between DIV45 HNRNPU+/- hCOs and E13 Hnrnpufl/- mouse cortices(A) Overview of differentially expressed genes (DEGs) in E13 cortices and hCOs.(B) Log2 fold change of all hCO DEGs were plotted against log2 fold change values of correlate genes in E13 Hnrnpufl/- mouse cortices, without consideration of whether or not those changes were identified as differentially expressed in E13 samples (left). A best fit linear regression line was plotted and resulted in an R2 coefficient of 0.13. Same analysis was performed considering only genes that were both differentially expressed in hCOs and FDR adjusted significance thresholds in mice of either p<0.5 (middle) or p<0.05 (right). When considering genes with FDR<0.05 in both models, R2 coefficient increases to 0.29. Quanitification of geometric mean enrichment further shows significant enrichment of both co-upregulated (C) and co-downregulated (D) genes.
Fig 3: Transcriptomic signature of HNRNPU-related disorder diverges in perinatal mice(A) Evidence of reduced transcriptomic dysregulation in perinatal Hnrnpufl/- cortices. Both heterozygous and homozygous mice segregate in PC space in RNA-sequencing of embryonic cortices, but heterozygous samples do not segregate at a perinatal time point. “F” = female. “M” = male.(B) Both Hnrnpufl/- and Hnrnpufl/fl cortices show most significant geometric mean enrichment is in genes differentially expressed in opposing directions at embryonic and perinatal time points.(C) Downregulated genes in DIV45 HNRNPU+/- hCOs are significantly enriched in gene sets that were identified as upregulated in both Hnrnpufl/- and Hnrnpufl/fl P1 cortices.
Fig 4: HNRNPU+/- stem cell lines successfully differentiate into human cortical organoids with significant size reduction(A) Two isogenic HNRNPU+/-mutant stem cell lines were generated by SynthegoTM and representative phase contrast images of stem cell populations are shown.(B) Both HNRNPU+/- stem cell lines show approximately 25% reduction in hnRNPU mRNA (left) and protein (right) expression compared to internal controls. * = p< 0.05. NS = not statistically significant.(C) HNRNPU+/- hCOs show significant size reduction emerging by DIV10 and continuing through 45 days in vitro. * = p< 0.05.(D) D11, M20 and PGP1 exhibit VZ-like proliferative zones at DIV25 (left) and widespread neuronal expression (right) at DIV45 with evidence of subpopulations of GABAergic neurons. Data are represented as mean ±SEM.
Supplier Page from Abcam for Anti-hnRNP U/p120 antibody [EPR12278(2)(B)]