Removal of developmentally regulated microexons has a minimal impact on larval zebrafish brain morphology and function
- Department of Neurobiology, The University of Alabama at Birmingham Heersink School of Medicine, United States
Figures
Figure 1 with 4 supplements
Generation and analysis of zebrafish with mutations that remove conserved, developmentally regulated microexons.
(A) Pipeline of the screen. Mutant lines with alternatively spliced microexons removed were generated with CRISPR/Cas9, crossed together, and sibling larvae were assessed for changes to brain morphology, brain activity, and behavioral profiling. (B) The amino acid sequence identity of 95 zebrafish microexons compared to mouse. (C) The amino acid sequence identity of 95 zebrafish microexons compared to mouse and divided by the sequence identity of the entire protein. (D) Gene Ontology analysis of biological processes associated with the 95 mouse microexons that are conserved in zebrafish. The analysis was completed using the PANTHER classification system. (E) Quantification of reverse transcription PCR (RT-PCR) for microexon-containing regions over zebrafish development. These data were clustered using the default seaborn clustermap settings (method = ‘average’). Panels A and E were created in BioRender.
-
Figure 1—source data 1
Quantified reverse transcription PCR (RT-PCR) for microexon-containing regions over zebrafish development for panel E.
- https://cdn.elifesciences.org/articles/101790/elife-101790-fig1-data1-v1.csv
Figure 1—figure supplement 1
Microexons in relationship to cut sites, neighboring exons, putative regulatory elements, and gRNA sites.
Diagrams were generated with DNA Features Viewer (https://edinburgh-genome-foundry.github.io/DnaFeaturesViewer/). The first and third columns show the microexons and neighboring exons, with a unique scale bar included for each mutant. The second and fourth columns show the zoomed area around each microexon with the start position for putative regulatory elements and gRNA sites. When the microexon neighbored either the first or last exon, the UTR is included in the labeled exon (e.g., csnk1g1 and synj1-1). When determining possible UC-repeat elements, a distance of 100 base pairs from the microexon was considered, although <50 base pairs is canonical. The label for the UC repeat represents the beginning of possible repeat sequences and not the span; thus, there may be closer elements.
Figure 1—figure supplement 2
Transcriptomic outcomes of eliminating selected microexons.
(A) Reverse transcription PCR (RT-PCR) validations of microexon removal at 6 dpf for wild-type (+/+), heterozygous (+/−), and homozygous (−/−) siblings. The expected length of the product with and without the microexon is shown above each sample, and the ladder is shown on the left side. The upper band in the vav2 wild-type and heterozygous samples is a heterodimer of the two products, an outcome we often see when genotyping smaller deletions. (B) Quantification of the gels in panel A. (C) Nanopore sequencing (Plasmidsaurus) of the heterozygous nrxn1a sample with corresponding read counts (right). The second read (R2) represents the intermediate band on the nrxn1a gel above. This sequence corresponds to an extension of the upstream exon (gray), indicating that there are two isoforms of nrxn1a in addition to the developmentally regulated inclusion of the microexon. (D) Nanopore sequencing (Plasmidsaurus) of the heterozygous vav2 sample with corresponding read counts. The R2 sequencing missing the microexon was additionally confirmed with Sanger sequencing of the homozygous sample. (E) qRT-PCR for selected microexon lines. The expression of each gene was normalized to rpl13a. Statistical significance was calculated with the Brown–Forsythe and Welch ANOVA corrected for multiple comparisons.
-
Figure 1—figure supplement 2—source data 1
qRT-PCR for selected microexon lines normalized to rpl13a.
- https://cdn.elifesciences.org/articles/101790/elife-101790-fig1-figsupp2-data1-v1.xlsx
Figure 1—figure supplement 3
Phenotypes for mutants in microexon genes that remove protein sequence beyond the microexon.
(A) Developmental phenotypes observed in homozygous larvae for three genes with truncating mutations. Chi-square analysis p < 0.01 from 12/12 animals with and without a phenotype. Additionally, healthy homozygous animals were never recovered from multiple heterozygous incrosses assessed at 4–6 dpf. (B) Brain activity maps for heterozygous incrosses. The comparisons are shown above the sum-of-slices intensity projection with a 6 dpf brain outline, where the genotype before the | is being compared to the one after. All N are in Supplementary file 2. (C) Brain structure maps for heterozygous incrosses. All N are in Supplementary file 2. (D) Brain activity maps for heterozygous outcrosses for those mutants with lethal developmental phenotypes that prohibit homozygous imaging. The comparisons are shown above the sum-of-slices intensity projection with a 6 dpf brain outline, where the genotype before the | is being compared to the one after. All N are in Supplementary file 2. (E) Brain structure maps for heterozygous outcrosses for those mutants with lethal developmental phenotypes that prohibit homozygous imaging. The comparisons are shown above the sum-of-slices intensity projection with a 6 dpf brain outline, where the genotype before the | is being compared to the one after. All N are in Supplementary file 2.
Figure 1—figure supplement 4
Behavior summary data for mutants in microexon genes that remove protein sequence beyond the microexon.
(A) Baseline and stimulus-driven behavior dot plots for heterozygous incrosses. Dot size corresponds to the percent of significant assays in the category (e.g., Magnitude). Replicate experiments using different parental pairs are shown side-by-side. All N are in Supplementary file 2. (B) Baseline and stimulus-driven behavior dot plots for heterozygous outcrosses. Dot size corresponds to the percent of significant assays in the category (e.g., Magnitude). (C) The caska truncating mutant has repeatable phenotypes in motion frequency and dark flash response. Example frequency of motion plot for caska mutants. The nighttime movement frequency is increased. The wild-type siblings (black) are compared to the homozygous (red). (D) Graph demonstrating the reduced dark flash responsivity of caska mutants. The latency is increased in this line (see panel H), which can be visualized by this response plot. (E) Increased daytime movement frequency and increased dark flash response frequency in ckap5 heterozygous mutants (red) compared to wild-type siblings (black). This phenotype represents the strongest behavioral outcome for heterozygous compared to wild-type. (F) The ppp6r3sa16892 has an increased response to acoustic stimuli. These three plots are response traces (shown from left to right) induced by strong taps that occur after tap habituation block 2 (day5dpfhab2post), after tap habituation block 3 (day5dpfhab3post), and during the night (a0f1000d5pD300a1f1000d5p).
Figure 2 with 4 supplements
Larval behavioral phenotypes of zebrafish with microexons removed.
(A) Summary of behavioral pipeline. (B) Baseline behavioral phenotypes for microexon mutants. The labels ‘1’ and ‘2’ indicate biological replicates. The data shown is for homozygous mutant larvae compared to wild-type siblings. Comparisons to the heterozygous larvae are removed for clarity and available in the Supplementary Materials, as they often have even milder phenotypes than homozygous. The size of the bubble represents the percent of significant measurements in the summarized category, and the color represents the mean of the strictly standardized mean difference (SSMD) of the significant assays in that category. (C) Stimulus-driven behavioral phenotypes for microexon mutants. The labels ‘1’ and ‘2’ indicate biological replicates. The bubble size and color are calculated the same as in panel B. (D) Examples of behavioral phenotypes. The black boxes in panels B and C correspond to the selected plots. Wild-type siblings (black) are compared to the homozygous (red), and the plots show mean ± SEM. All N are in Supplementary file 2. Kruskal–Wallis ANOVA p-values for the selected plots are as follows: dop1a center preference during the first night (day0night_boutcenterfraction_3600) = 0.00006/0.0008, N +/+ = 15 (run 1) and 14 (run 2), N −/− = 17 (run 1) and 15 (run 2); eif4g3b p-values are not calculated for response traces (shown is all dark flashes in block 1), p-values for the latency for the first 10 dark flashes in block 1 (day6dpfdf1a_responselatency) = 0.026/0.0008, N +/+ = 16 (run 1) and 15 (run 2), N −/− = 15 (run 1) and 21 (run 2); ppp6r3 frequency of response to strong acoustic stimuli with a sound frequency of 1000 Hz that precede the habituation block (day5dpfhab1pre_responsefrequency_1_a1f1000d5p) = 0.002/0.016, N +/+ = 19 (run 1) and 20 (run 2), N −/− = 21 (run 1) and 29 (run 2); ptprd-1 pixels moved in each bout for the duration of the experiment (combo_boutcumulativemovement_3600) = 0.001/0.00002, N +/+ = 22 (run 1) and 21 (run 2), N −/− = 26 (run 1) and 18 (run 2); rapgef2 number of bouts for the duration of the experiment calculated using the delta pixel data in each frame (dpix_numberofbouts_3600) = 0.002/0.001, N +/+ = 21 (run 1) and 23 (run 2), N −/− = 21 (run 1) and 20 (run 2).
-
Figure 2—source data 1
Zip file of summarized behavior data and script to generate the heatmap in panel B.
Raw data is available in Zenodo.
- https://cdn.elifesciences.org/articles/101790/elife-101790-fig2-data1-v1.zip
-
Figure 2—source data 2
Zip file of summarized behavior data and script to generate the heatmap in panel B.
Raw data is available in Zenodo.
- https://cdn.elifesciences.org/articles/101790/elife-101790-fig2-data2-v1.zip
Figure 2—figure supplement 1
Baseline behavior summary data for microexon mutants.
Baseline behavior dot plots for all sibling comparisons. Dot size corresponds to the percent of significant assays in the category (e.g., Magnitude). Replicate experiments using different parental pairs are shown side-by-side. All N are in Supplementary file 2.
Figure 2—figure supplement 2
Frequency of motion example plots for microexon mutants.
Frequency of motion measured as number of bouts/hour over the entire duration of the experiment. The biological replicates are shown side-by-side. All N are in Supplementary file 2. The wild-type siblings (black) are compared to the homozygous (red).
Figure 2—figure supplement 3
Stimulus-driven behavior summary data for microexon mutants.
Stimulus-driven behavior dot plots for all sibling comparisons from a heterozygous parental in-cross.
Figure 2—figure supplement 4
Additional behavioral phenotypes for microexon mutants.
All N are in Supplementary file 2. (A) Frequency phenotype of dop1a mutants. Shown is the day3msdf_dpix_numberofbouts_60 graph for run 1/run 2. p-values: 0.033/0.038 (linear mixed model) and 0.036/0.11 (Kruskal–Wallis ANOVA, which performs poorly on data with a large range). (B) Reduced dark flash response frequency phenotype of dop1a mutants. Shown is the response frequency for dark flash block 3 (day6dpfdf3_responsefrequency). p-values: 0.047/0.036 (Kruskal–Wallis ANOVA). (C) Reduced dark flash response of nrg2b mutants. p-values are not calculated for response graphs. (D) Reduced dark flash response frequency phenotype of ptprd-1 mutants. Shown is the response frequency for dark flashes from all three blocks (day6dpfdfall_responsefrequency). p-values: 0.0056/0.0018 (Kruskal–Wallis ANOVA). (E) Reduced nighttime sleep bout frequency phenotype of ptprd-1 mutants. Shown is the plot for the entire 3-night/day experiment (combo plot). p-values: 0.001/0.013 (linear mixed model). The same measure from a subsection of the data, the entire second night (day2nightall_dpix_numberofboutsSLEEP, not shown), has Kruskal–Wallis ANOVA p-values of 0.001/0.003. (F) Altered number of bouts frequency phenotype of ptprd-1 mutants. Shown is the entire 3-night/day experiment (combo plot). p-values are not significant for the entire duration, but a subsection (day1morn_dpix_numberofbouts_60, not shown) has Kruskal–Wallis ANOVA p-values of 0.003/0.012. (G) Reduced daytime bout total pixels magnitude (dpix_boutcumulativemovement) phenotype of ptprd-2 mutants. Shown is the entire 3-night/day experiment (combo plot). p-values are not significant for this duration. The same measure from a subsection of the data, the evening of day 1 of the experiment (day1evening_dpix_boutcumulativemovement, not shown), has Kruskal–Wallis ANOVA p-values of 0.002/0.019. (H) Increased sleep bout (stronger in daytime) frequency phenotype of ptprd-2 mutants. Shown is the entire 3-night/day experiment (combo plot). p-values: 0.001/0.013 (Kruskal–Wallis ANOVA). The same measure from a subsection of the data, the entire second night (day2nightall_dpix_numberofboutsSLEEP_60, not shown), has Kruskal–Wallis ANOVA p-values of 0.001/0.01. (I) Reduced number of bouts (stronger in daytime) frequency phenotype of ptprd-2 mutants. Shown is the entire 3-night/day experiment (combo plot). p-values: 0.001/0.04 (linear mixed model) and 0.003/0.08 (Kruskal–Wallis ANOVA). (J) Reduced acoustic response frequency phenotype of ptprd-2 mutants. Shown are the escape responses for the strong stimuli occurring following acoustic habituation (habituation_day5dpfhab1post_responsefrequency). The responses are filtered for those that are true escapes, and the phenotype emerges for only the true, short-latency C-bends. p-value: 0.03/0.006 (Kruskal–Wallis ANOVA). (K) Reduced strong acoustic responses of ptprd-2 mutants. This response plot goes along with the frequency data, but this plot is not filtered for only the true C-bends. p-values are not calculated for response graphs. (L) Increased daytime location center preference for vav2 mutants. Shown is the evening of the second day (day2evening_boutcenterfraction), but multiple subsections show significant increase in center dwelling preference. p-value: 0.006/0.013 (Kruskal–Wallis ANOVA).
Figure 3 with 4 supplements
Whole-brain activity and morphology phenotypes of zebrafish with microexons removed.
(A) Clustered summary of pErk comparisons between homozygous mutants and wild-type control siblings, where magenta represents decreased activity and green represents increased. The signal in each region was summed and divided by the region size. The N for all experiments is available in Supplementary file 2. (B) Clustered summary of structure comparisons between homozygous mutants and wild-type control siblings, where magenta represents decreased size and green represents increased. (C) Location of major regions in the zebrafish brain based on masks from the Z-Brain atlas (Randlett et al., 2015). (D) Brain imaging for microexon mutants with repeatable brain activity phenotypes. The brain images represent the significant signal difference between homozygous and wild-type control siblings. They are shown as sum-of-slices projections (Z- and X-axes) with the white outline representing the zebrafish brain. (E) Structural phenotype of the ppp6r3 mutant, with replicates shown side-by-side. The magenta indicates decreased size. (F) Brain imaging for two microexon mutants with brain activity phenotypes that are similar for both the heterozygous and homozygous mutants. (G) Comparison of the total brain activity signal between homozygous microexon mutants from this work and mutants for autism risk genes from Capps et al., 2024. Both increased and decreased activity are considered a single comparison rather than the average of the repeats.
-
Figure 3—source code 1
Python script that sums the intensities from image stacks to generate the source data for panels A and B.
- https://cdn.elifesciences.org/articles/101790/elife-101790-fig3-code1-v1.zip
-
Figure 3—source code 2
Python script that generates the heatmap in panels A and B using the source data.
- https://cdn.elifesciences.org/articles/101790/elife-101790-fig3-code2-v1.zip
-
Figure 3—source data 1
Red channel signal for activity differences heatmap in panel A.
Raw image stacks are available in Zenodo.
- https://cdn.elifesciences.org/articles/101790/elife-101790-fig3-data1-v1.csv
-
Figure 3—source data 2
Green channel signal for structural differences heatmap in panel A.
Raw image stacks are available in Zenodo.
- https://cdn.elifesciences.org/articles/101790/elife-101790-fig3-data2-v1.csv
-
Figure 3—source data 3
Red channel signal for structural differences heatmap in panel B.
Raw image stacks are available in Zenodo.
- https://cdn.elifesciences.org/articles/101790/elife-101790-fig3-data3-v1.csv
-
Figure 3—source data 4
Green channel signal for structural differences heatmap in panel B.
Raw image stacks are available in Zenodo.
- https://cdn.elifesciences.org/articles/101790/elife-101790-fig3-data4-v1.csv
Figure 3—figure supplement 1
Brain activity maps for multiple genetic comparisons for microexon mutants.
The comparisons are shown above the sum-of-slices intensity projection with a 6 dpf brain outline, where the genotype before the | is being compared to the one after. All N are in Supplementary file 2.
Figure 3—figure supplement 2
Second set of brain activity maps for multiple genetic comparisons for microexon mutants.
The comparisons are shown above the sum-of-slices intensity projection with a 6 dpf brain outline, where the genotype before the | is being compared to the one after. All N are in Supplementary file 2.
Figure 3—figure supplement 3
Brain structural maps for multiple genetic comparisons for microexon mutants.
The comparisons are shown above the sum-of-slices intensity projection with a 6 dpf brain outline, where the genotype before the | is being compared to the one after. All N are in Supplementary file 2.
Figure 3—figure supplement 4
Second set of brain structural maps for multiple genetic comparisons for microexon mutants.
The comparisons are shown above the sum-of-slices intensity projection with a 6 dpf brain outline, where the genotype before the | is being compared to the one after. All N are in Supplementary file 2.
Tables
Key resources table
| Reagent type (species) or resource | Designation | Source or reference | Identifiers | Additional information |
|---|---|---|---|---|
| Genetic reagent (D. rerio) | Microexon mutants | This paper | Supplementary file 2 | |
| Antibody | Mouse monoclonal anti-Erk | Cell Signaling | #4696 | IF(3:1000) |
| Antibody | Rabbit monoclonal anti-phospho-Erk | Cell Signaling | #4370 | IF(1:500) |
| Commercial assay or kit | E.Z.N.A. MicroElute Total RNA Kit | Omega Bio-Tek | R6834-02 | |
| Commercial assay or kit | iScript Reverse Transcription Supermix | Bio-Rad | #1708840 | |
| Commercial assay or kit | GoTaq 2x master mix | Promega | M7123 | |
| Commercial assay or kit | SsoAdvanced Universal SYBR Green Supermix | Bio-Rad | #1725270EDU | |
| Sequence-based reagent | Oligonucleotide primers for genotyping and cDNA amplification | Life technologies, this paper | Supplementary files 1 and 2 | |
| Software, algorithm | Fiji/ImageJ | https://imagej.net/software/fiji/downloads; Schindelin et al., 2012 | ||
| Software, algorithm | Image registration with CMTK | https://www.nitrc.org/projects/cmtk; Jefferis et al., 2007 | ||
| Software, algorithm | Zebrafish brain mapping and Z-Brain atlas | Randlett et al., 2015 | ||
| Software, algorithm | Zebrafish behavior analysis | https://github.com/thymelab/ZebrafishBehavior; Joo et al., 2020 |
Additional files
-
Supplementary file 1
Sequences and locations of 95 microexons conserved between zebrafish and mouse.
- https://cdn.elifesciences.org/articles/101790/elife-101790-supp1-v1.xlsx
-
Supplementary file 2
Mutants generated and corresponding genotyping and experimental information.
- https://cdn.elifesciences.org/articles/101790/elife-101790-supp2-v1.xlsx
-
MDAR checklist
- https://cdn.elifesciences.org/articles/101790/elife-101790-mdarchecklist1-v1.pdf
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Removal of developmentally regulated microexons has a minimal impact on larval zebrafish brain morphology and function
eLife 13:RP101790.
https://doi.org/10.7554/eLife.101790.3
