Cell-type-specific alternative splicing in the cerebral cortex of a Schinzel-Giedion Syndrome patient variant mouse model
2024-06-25
bioRxiv Preprint:
Emma F. Jones, Timothy C. Howton, Tabea M. Soelter, Anthony B. Crumley, Brittany N. Lasseigne
Schinzel-Giedion Syndrome (SGS) is an ultra-rare Mendelian disorder caused by gain-of-function mutations in the SETBP1 gene. While previous studies determined multiple roles for how SETBP1 and associated pathways may cause disease manifestation, they have not assessed whether cell-type-specific alternative splicing (AS) plays a role in SGS. We used STARsolo to quantify gene and splice junction (SJ) expression for 51,465 nuclei previously generated from the cerebral cortex of atypical Setbp1S858R SGS patient variant mice (n = 3) and wild-type control mice (n = 3). After cell type annotation, we performed pseudobulk differential gene expression and SJ usage (SJU) analyses across cell types and conditions. We identified 34 genes with statistically significant alterations in SJU. Oligodendrocytes had the most genes with changes in SJU, followed by astrocytes, excitatory, and inhibitory neurons. One gene, Son, a splicing cofactor known to cause the neurodevelopmental disorder ZTTK Syndrome, had SJU changes in all six non-vascular cell types we measured in Setbp1S858R compared to controls. This is the first research to report AS changes in the cerebral cortex of an SGS model and the first study to link SGS to perturbations in Son.
(A) Schematic overview of our processing and analysis pipeline. (B) We analyzed pseudobulk gene expression and calculated SJU for each cell type and condition. (C) We compared SJU values for each cell type using a permutation test (Methods) to identify cell-type-specific differences in AS between Setbp1S858R and wild-type mouse brain tissue. (D) Next, we visualized all annotated transcripts and splice junction locations for each significant SJU gene. (E) Finally, we compared the genes and pathways identified through functional enrichment analysis that overlap between cell types and predict their biological relevance.
The following bash scripts are for running STARsolo on the raw fastq files. These scripts are not required to run if you download processed data from zenodo.
## src/starsolo_conda
## ├── 01_build_STAR_genome.sh
## └── 02_run_STARsolo.sh
- 01_build_STAR_genome.sh - The purpose of this script is to build a genome to run STAR.
- 02_run_STARsolo.sh - The purpose of this script is to run STARsolo with custom filtering parameters to quantify gene and splice junction expression.
The following scripts are for processing and filtering the gene expression data.
## src/seurat
## ├── 01_import_filter_data.Rmd
## ├── 02_annotate_cell_types.Rmd
## ├── functions.R
## └── generate_cellcycle_lists.R
- 01_import_filter_data.Rmd - The purpose of this script is to import and filter data using seurat. It is dependent on running the scripts in STARsolo_conda or downloading the data to start. Please use docker image setbp1_alternative_splicing:1.0.4.
- 02_annotate_cell_types.Rmd - The purpose of this script is to annotate cell types using seurat. It is dependent on seurat script 01. Please use docker image setbp1_alternative_splicing:1.0.4.
- functions.R - The purpose of this script is to provide function scripts necessary for analysis.
- generate_cellcycle_lists.R
The following scripts are for formatting, importing, and analyzing splice junction and expression data using the MARVEL package.
## src/marvel
## ├── 01_format_MARVEL_data.Rmd
## ├── 02_MARVEL_differential_analysis.Rmd
## ├── 03_analyze_de_genes.Rmd
## ├── 04_calc_sj_usage.Rmd
## ├── 05_cell_specific_sj_expr.Rmd
## └── functions.R
- 01_format_MARVEL_data.Rmd - The purpose of this script is to format expression data from seurat and SJ outputs from STARsolo for MARVEL. It is dependent on running the seurat scripts 01 and 02. Please use docker image setbp1_alternative_splicing:1.0.5.
- 02_MARVEL_differential_analysis.Rmd - The purpose of this script is to run differential analysis with MARVEL. It is dependent on running all seurat scripts, and marvel script 01. Please use docker image setbp1_alternative_splicing:1.0.5.
- 03_analyze_de_genes.Rmd - The purpose of this script is to analyze differential analysis results MARVEL. It is dependent on running all seurat scripts, and marvel scripts 01 and 02. Please use docker image setbp1_alternative_splicing:1.0.6, since you need ComplexHeatmap so make the UpSet plots.
- 04_calc_sj_usage.Rmd - The purpose of this script is to calculate splice junction usage for each cell type for each splice junction. The data is too sparse to calculate splice junction usage for single cells. It is dependent on running all seurat scripts, and marvel scripts 01 through 03. Please use docker image setbp1_alternative_splicing:1.0.6.
- 05_cell_specific_sj_expr.Rmd - The purpose of this script is to quantify the splice junction expression across cell types, both per cell and aggregated for each cell type. It creates a new supplemental table that is referred to in the manuscript. It is dependent on running all seurat scripts, and marvel scripts 01 through 04. Please use docker image setbp1_alternative_splicing:1.1.0.
- functions.R - The purpose of this script is to provide function scripts necessary for analysis.
The following scripts are for running a pseudobulk differential gene expression analysis
## src/deseq2
## ├── 01_pseudobulk_analysis.Rmd
## └── functions.R
- 01_pseudobulk_analysis.Rmd - The purpose of this script is to run a pseudobulk gene expression analysis using the cell type data from seurat. It is dependent on seurat scripts 01 and 02. It should be run in docker 1.0.8.
- functions.R - The purpose of this script is to provide function scripts necessary for analysis.
The following script are for creating finalized figures for the manuscript
## src/figures
## ├── figure_2.Rmd
## ├── figure_3-4.Rmd
## ├── figure_5.Rmd
## ├── functions.R
## ├── geom_split_violin.R
## ├── mean_expression_celltype.R
## └── supp_figure_2.Rmd
- figure_2.Rmd - The purpose of this script is to create a finalized version of figure 2. It is dependent on Marvel scripts 01 through 06. Run in docker 1.0.9.
- figure_3-4.Rmd - The purpose of this script is to make and complete figures 3 and 4. Figure 3 has an upset plot of all alternatively spliced genes, and Figure 4 has functional enrichment analysis of cell-type specific AS genes. Run in docker 1.1.0.
- figure_5.Rmd - The point of this figure is to convey the importance of the Son gene and also add an additional panel to show where the splice junctions lit on transcript structures. Run in docker 1.1.0.
- functions.R - The purpose of this script is to provide function scripts necessary for analysis.
- geom_split_violin.R - This is a split violin ggplot function borrowed from jan-glx on stackoverflow
- mean_expression_celltype.R - This is a short R script to get average expression for each cell type of a Seurat object
- supp_figure_2.Rmd - The purpose of this script is to create a finalized version of supporting information figure 2. It is dependent on Seurat scripts 01 and 02. Run in docker 1.0.6.
All intermediate data generated by the following analyses can be found at the following links:
Data:
Docker Images:
GitHub Repository:
We performed all analyses in docker with R version 4.3.1. The repository with all the Docker images used for this project can be found on Docker Hub at emmafjones/setbp1_alternative_splicing. Docker images have backward compatibility, so all scripts can be run using tag 1.1.0. While individual docker image tags used are noted in every script, we provide an overview of analyses and their associated tags below:
| Tag | Associated Analyses |
|---|---|
| 1.0.4 | Seurat scripts |
| 1.0.5 | MARVEL scripts 01 and 02 |
| 1.0.6 | Seurat scripts 03 and 04, supporting information Figure 2 |
| 1.0.8 | DESeq2 scripts |
| 1.0.9 | Figure 2 script |
| 1.1.0 | MARVEL script 05, Figures 3-5 |
What is Happening in the Lasseigne Lab?
This work was funded by the UAB Lasseigne Lab funds and the UAB Pilot Center for Precision Animal Modeling (C-PAM) (1U54OD030167).
We would like to acknowledge all current and former members of the Lasseigne Lab for their thoughtful feedback, especially Elizabeth J. Wilk, Amanda D. Clark, and Vishal H. Oza. The graphical abstract was created using BioRender.
This repository is licensed under the MIT License, see LICENSE documentation within this repository for more details.