Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
[Preprint]. 2024 Jun 15:2024.06.13.597128.
doi: 10.1101/2024.06.13.597128.

Reverse engineering neuron type-specific and type-orthogonal splicing-regulatory networks using single-cell transcriptomes

Daniel F Moakley  1   2   3 , Melissa Campbell  1   2   3   4 , Miquel Anglada-Girotto  1   5 , Huijuan Feng  1   2   3   6 , Andrea Califano  1 , Edmund Au  7   8 , Chaolin Zhang  1   2   3
Affiliations

Reverse engineering neuron type-specific and type-orthogonal splicing-regulatory networks using single-cell transcriptomes

Daniel F Moakley et al. bioRxiv. .

Abstract

Cell type-specific alternative splicing (AS) enables differential gene isoform expression between diverse neuron types with distinct identities and functions. Current studies linking individual RNA-binding proteins (RBPs) to AS in a few neuron types underscore the need for holistic modeling. Here, we use network reverse engineering to derive a map of the neuron type-specific AS regulatory landscape from 133 mouse neocortical cell types defined by single-cell transcriptomes. This approach reliably inferred the regulons of 350 RBPs and their cell type-specific activities. Our analysis revealed driving factors delineating neuronal identities, among which we validated Elavl2 as a key RBP for MGE-specific splicing in GABAergic interneurons using an in vitro ESC differentiation system. We also identified a module of exons and candidate regulators specific for long- and short-projection neurons across multiple neuronal classes. This study provides a resource for elucidating splicing regulatory programs that drive neuronal molecular diversity, including those that do not align with gene expression-based classifications.

Keywords: Elavl2; RNA splicing regulation; RNA-binding proteins; network inference; neuronal subtypes.

PubMed Disclaimer

Conflict of interest statement

Competing Interest Statement: The authors declare no competing interests.

Figures

Figure 1.
Figure 1.. Reconstruction of splicing-regulatory networks using scRNA-seq data from adult mouse neocortical cell types.
A, Adult mouse neocortical cell scRNA-seq data are pooled by cell types for the simultaneous quantification of RBP expression and cassette exon inclusion. B, Schematic of splicing regulation inference workflow. The ARACNe algorithm is applied to the cell-type-level RBP expression and exon inclusion data to predict the regulatory targets (regulon) of each RBP, which is then used to estimate neuron type-specific RBP activity using the VIPER algorithm. C, Visualization of the inferred splicing-regulatory network. D, Number of inferred targets per RBP. RBPs are grouped into 50 bins based on the regulon size. The inferred regulon size per RBP approximately follows a power-law distribution. E, Heatmap of glutamatergic and GABAergic neuron type-specific exons with inferred positive and negative regulatory RBPs indicated on the right as red or blue tick marks, respectively. F, Similar to (E) but for CGE and MGE-specific exons.
Figure 2.
Figure 2.. Validation of RBP regulons and activities inferred by ARACNe and VIPER algorithms.
A, ARACNe-inferred regulons show significant overlap (Venn diagrams) with high-confidence lists of RBP family targets from integrative Bayesian models based on analysis of multimodal splicing data–. Among the exons common to both lists, the predicted directions, or modes of regulation (MOR), by VIPER are also highly consistent with the integrative models (bar graphs). P-values shown are derived from Fisher’s exact tests. B,C, Similar to (A) but additional regulons inferred by ARACNe were compared to RBP target lists identified from exon inclusion changes after RBP perturbations, as measured by RNA-seq or exon-junction microarray data. D, Bar plot of RBP activity difference estimates in RBP-depleted samples using the predicted regulons as compared to wild-type (WT) control. For perturbations of multiple RBP family members, activity value sums of the multiple RBPs perturbed were used. The statistical significance of the directional change was tested using a Binomial test.
Figure 3.
Figure 3.. Differential RBP activity analysis provides candidate regulators of neuron type-specific splicing across different clades.
In each panel, RBP activity was compared between two groups of cell types. A, neurons and glia. B, glutamatergic and GABAergic neurons. C, MGE- and CGE-lineage interneurons. D, long- and short-projecting glutamatergic neurons. Q-values are derived from empirical Bayes-moderated t-tests followed by multiple test correction.
Figure 4.
Figure 4.. Validation of Elavl2 as a key MGE-lineage-specific splicing factor.
A, Elavl2 expression RPKM (reads per kb per million) in MGE- and CGE-lineage interneurons at various time points during cortical development. The gene shows a consistent preferential expression in MGE-lineage neurons as early as embryonic day 12.5 (E12.5) and persists into adulthood. B, Construct design for the MGE/CGE dual-reporter mouse ESC line. eGFP is positioned downstream of the MGE-specific Lhx6 promoter, and tdTomato is contained in an Ai9 reporter with Cre under the control of the CGE-specific marker 5ht3a. Ascl1/Mash1 overexpression is driven by the neural progenitor marker Nestin to promote interneuron differentiation. C, Experimental schema for testing the role of Elavl2 in interneuron type-specific splicing regulation. Elavl2 was knocked out in the dual-reporter mouse ESC line using CRISPR-Cas9. WT and KO ESCs were differentiated into interneurons using an embryoid body-based protocol. On day 16 of differentiation, cells were isolated based on reporter fluorescence by FACS and RNA was isolated for RNA-seq. D, Centered exon inclusion (percent spliced in or PSI) values of inferred Elavl2 targets ordered by predicted MOR. Inclusion differences of these exons in WT GFP+ (ESC-MGE cells) vs. tdTomato+ (ESC-CGE cells) samples and adult MGE- vs. CGE-lineage interneurons are shown at right. E-F Scatter plots (top) and quantification of directionality agreement (bottom) for predicted Elavl2 target inclusion value differences in WT versus Elavl2 KO eGFP+ samples (x-axis) compared to WT versus Elavl2 KO tdTomato+ samples (y-axis, e) or WT eGFP versus WT tdTomato+ samples (y-axis, F). Positive and negative Elavl2 targets predicted by ARACNe/VIPER are indicated in the scatter plots by red and blue, respectively. P-values above the barplots are calculated by Fisher’s exact test. G,H, Genome browser views of Slit2 exon 31 (G) and Alcam exon 13 (H) inclusion in WT or KO ESC-interneurons and adult MGE- or CGE-lineage neurons. Both exons are inferred targets of Elavl2.
Figure 5.
Figure 5.. Characterization and validation of an exon module specific to long- or short-projecting neurons.
A, Top left: Venn diagram showing the overlap of exons differentially spliced between glutamatergic versus GABAergic neurons, long- versus short-projecting glutamatergic neurons, and long- versus short-projecting Sst interneurons. Bottom right: Comparisons in the directionality of differentially spliced exons in the identified common exon between each of the neuron type comparisons, including a comparison of ES-derived globus pallidus neurons (ES-GPNs) and ES-derived interneurons (ESC-INs). P-values are calculated by Fisher’s exact test. B, A heatmap showing the differential inclusion of the 61 exons identified in all three comparisons. Inferred positive (red) ore negative (blue) regulators are indicated on the right. C, Gene ontology (GO) analysis of genes containing the 61 overlap exons shows an enrichment of genes related to axonogenesis and their maintenance. D, A schematic depicting Ptk2 exon 13, which is specifically included in short axon neurons. The cartoon depicts the possible biological function of the exon, which contains a reverse calmodulin binding domain and may link the gene’s function to calcium signaling.
Figure 6.
Figure 6.. Inferred networks identify candidate drivers of projection length-associated splicing.
A, PCA plot of the first two principal components of RBP activity with cell type clusters and neuronal classes indicated by color. B, Similar to (A) but with neuron projection types indicated by color. C-E, Comparisons of differential RBP activity in long- versus short-axon neuron types correlate across different neuronal classes (Pearson). F-G, Conchordance of candidate RBP regulons with the module of projection-length associated exons identified in Figure 5. Each scatter plot shows the predicted mode of regulation by the RBP on the x-axis and change in exon inclusion between projection types on the y-axis. Positive and negative target exons overlapping with the projection length-associated module and are differentially spliced between the groups are colored and counted in red or blue, respectively, and all other exons in the regulon are shown with reduced opacity. P-values are from Fisher’s exact test.
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Zeng H. & Sanes J. R. Neuronal cell-type classification: challenges, opportunities and the path forward. Nat Rev Neurosci 18, 530–546, doi:10.1038/nrn.2017.85 (2017). - DOI - PubMed
    1. Berg J. et al. Human neocortical expansion involves glutamatergic neuron diversification. Nature 598, 151–158, doi:10.1038/s41586-021-03813-8 (2021). - DOI - PMC - PubMed
    1. Christodoulou O., Maragkos I., Antonakou V. & Denaxa M. The development of MGE-derived cortical interneurons: An Lhx6 tale. The International Journal of Developmental Biology, doi:10.1387/ijdb.210185md (2022). - DOI - PubMed
    1. Jiang X. et al. Principles of connectivity among morphologically defined cell types in adult neocortex. Science 350, aac9462, doi:10.1126/science.aac9462 (2015). - DOI - PMC - PubMed
    1. Han X. et al. Whole human-brain mapping of single cortical neurons for profiling morphological diversity and stereotypy. Sci Adv 9, eadf3771, doi:10.1126/sciadv.adf3771 (2023). - DOI - PMC - PubMed

Publication types