This protocol presents an approach for whole transcriptome analysis from zebrafish embryos, larvae, or sorted cells. We include isolation of RNA, pathway analysis of RNASeq data, and qRT-PCR-based validation of gene expression changes.
The analysis of global gene expression changes is a valuable tool for identifying novel pathways underlying observed phenotypes. The zebrafish is an excellent model for rapid assessment of whole transcriptome from whole animal or individual cell populations due to the ease of isolation of RNA from large numbers of animals. Here a protocol for global gene expression analysis in zebrafish embryos using RNA sequencing (RNASeq) is presented. We describe preparation of RNA from whole embryos or from cell populations obtained using cell sorting in transgenic animals. We also describe an approach for analysis of RNASeq data to identify enriched pathways and Gene Ontology (GO) terms in global gene expression data sets. Finally, we provide a protocol for validation of gene expression changes using quantitative reverse transcriptase PCR (qRT-PCR). These protocols can be used for comparative analysis of control and experimental sets of zebrafish to identify novel gene expression changes, and provide molecular insight into phenotypes of interest.
Comparative analysis of global gene expression is a valuable tool to identify novel genes contributing to observed phenotypes. Such analyses typically rely on quantitative assessment of transcript abundance compared between experimental and control samples. Targeted approaches, such as qRT-PCR are relatively rapid and accurate for investigation of single gene expression changes. RNA sequencing (RNASeq) offers a broad, hypothesis-free approach to identify significant changes in gene expression between samples, making it now the standard for such investigations across experimental systems.
Zebrafish have emerged as a prominent model across many disease areas. Originally developed for their utility in developmental biology studies, due to their high fecundity and relatively low cost of maintenance, experimental use of zebrafish has evolved to include a broad range of phenotypes from embryonic to adult stages as well as a wide array of molecular assays1,2,3. Indeed, these advantages make molecular mechanistic studies rapid and cost-effective because of the ease of acquiring large amounts of material combined with the ease of both genetic and environmental manipulation at all stages of life. Moreover, the transparent nature of zebrafish embryos and larvae make it ideal for generating cell- and tissue-specific transgenic reporter lines allowing in vivo visualization of discrete cell populations4. Exploitation of such lines permits global gene expression analysis in specific isolated cell types based on reporter gene expression.
Here we present a comprehensive protocol for global gene expression analysis using RNASeq after culture of zebrafish embryos. Genetic experimental manipulations, including morpholino (MO)-based transient gene knockdown or CRISPR-mediated genome editing, have been presented elsewhere5,6,7. We therefore focus on a detailed protocol for isolation of RNA from whole embryos or sorted transgenic reporter-expressing cells followed by simple computational analysis of RNASeq results using pathway tools and gene ontology (GO) terms. Finally, we have included a strategy for validation of gene expression changes by quantitative reverse transcriptase PCR (qRT-PCR). These protocols are applicable to zebrafish embryos subjected to a wide range of experimental conditions, including comparison of genetic mutants or environmental conditions.
All animal protocols outlined below are in accordance with and approved by the University of Maryland Institutional Animal Care and Use Committee (IACUC).
1. Embryo Preparation
2. Single-cell Dissociation: Whole Embryo and Sorted Cell Populations
3. RNA Preparation
4. Pathway and GO Term Analysis
NOTE: See Figure 1 for representative output of the gene expression analysis after RNASeq provided by the core or vendor.
5. Verification by qRT-PCR
NOTE: Individual genes identified with significant gene expression changes in RNASeq should be verified by targeted qRT-PCR in replicate experiments.
Sorting of Differentially Expressed Genes:
To identify differentially expressed genes in the larval stage of zebrafish models of Alström Syndrome and Bardet-Biedl Syndrome (BBS), we targeted either alms1 or bbs1 transcripts by injecting previously validated splice-blocking MOs into wild-type zebrafish embryos16,17. At 5 days post fertilization (dpf), two replicates of RNA were extracted from each condition, as well as embryos injected with control MO, as described in Section 3 above. RNA from each sample was sequenced to a depth of ~ 120 million reads with ~ 107 million reads aligned to the zebrafish genome. Between 85-90% of the aligned reads mapped to exonic regions of the genome.
1,348 total genes were significantly changed in the Alström model and 471 total genes were significantly changed in the BBS model. Shared changes in gene expression between the models or changes unique to one of the models were identified as described in Section 4.2. 1,084 total genes were uniquely changed in the Alström model, 612 up-regulated and 472 down-regulated, and 207 genes were uniquely changed in the BBS model, 176 up-regulated and 31 down-regulated (Figure 2, Table S1, and Table S2). 264 genes were found to be significantly changed in both models, 73 were up-regulated and 182 were down-regulated, and 9 showed opposite changes in expression between the two models (Figure 2, Table S1, and Table S2 (See Supplemental Tables.xlsx)). The discrepant numbers of differentially expressed genes between the two models suggest that bbs1 and alms1 may impact different stages in development.
Interestingly, the large number of differentially expressed genes in the Alström model suggest that alms1 plays an important role in development at the 5 dpf stage, while the smaller number of gene changes in the BBS model suggest that the role of bbs1 may not be as prominent at this time point. In contrast, previous transcriptomic analyses at 2 dpf suggested the reverse18.
Determination of Enriched Pathways and Gene Ontologies in Zebrafish Model of Alström Syndrome:
To more clearly elucidate the molecular profile of the Alström model, the pathways and gene ontologies enriched in the differentially expressed genes were identified. Enriched pathways were determined by querying the ConsensusPathDB and enriched Gene Ontologies were determined by querying the Gene Ontology Consortium as described in Sections 4.3-4.5. The most highly enriched among the genes up-regulated in the Alström model, by number of genes or fold enrichment, were analyzed. 31 total pathways were up-regulated in the Alström model. Aside from the broad grouping of "metabolism," the most highly affected pathways were the innate and adaptive immune system with 32 and 20 genes, respectively (Figure 3A). Downstream β-cell receptor signaling events were also enriched suggesting immune system up-regulation in this model. Several signaling pathways were also enriched among the up-regulated genes in the Alström model, consistent with the association of Alström Syndrome with primary cilium dysfunction, a major signaling center of the cell (Figure 3A). In addition, three pathways associated with insulin secretion were up-regulated: fatty acids bound to GPR40, free fatty acids, and acetylcholine (Figure 3A). This is consistent with the highly penetrant insulin resistance and type 2 diabetes phenotypes in Alström patients. Six GO terms were enriched among the up-regulated genes in the Alström model, including erythrocyte differentiation, erythrocyte homeostasis, myeloid cell homeostasis, and homeostatic processes (Figure 3B). To assess the interconnectedness of the up-regulated pathways, a pathway network of the Alström model up-regulated pathways was generated (Figure 3C). The major node of interconnected pathways revealed a high degree of connectivity among the pathways associated with the immune system and signaling, while one of the minor nodes revealed connectivity between the three insulin regulatory pathways. Finally, we verified the gene expression changes identified by RNASeq by assessment of 5 gene targets, which were either up- or down-regulated in the Alström model. The directionality of gene expression changes by qRT-PCR in aste, gck, bmb, btr26, and ifi35, relative to control, recapitulated that observed by RNASeq (Figure 4A–B).
Figure 1. Representative output of gene expression analysis after RNASeq provided by the core or vendor. Feature ID indicates the ENSEMBL gene ID number. Read counts indicates the number of reads in either control or experimental samples that aligned to that gene in the genome. Fold change and the log of the fold change indicate the degree of change in expression of the gene between the experimental and control sample in either the positive (increase) in expression in the experimental sample direction, or negative (decrease) in expression in the experimental sample direction. The p-value and FDR are statistical tests to determine the significance of results; for RNASeq experiments, a more stringent FDR value of 0.05 is used to determine significance. Gene_symbol indicates the NCBI gene ID, and gene_name lists the full name of the gene. Please click here to view a larger version of this figure.
Figure 2. Differentially expressed genes in the BBS and Alström models. Number of genes up-regulated (red) or down-regulated (yellow) in alms1 MO (green) injected embryos, bbs1 MO (orange) injected embryos, or both compared to standard control MO injected embryos. *Denotes number of genes changed in both but having opposite changes in expression. Please click here to view a larger version of this figure.
Figure 3. Upregulated pathways and enriched GO terms in the Alström model. (A) Upregulated pathways by number of genes. (B) Upregulated GO terms by fold enrichment. (C) Pathway connectivity of upregulated pathways in the Alström model. Pathway connections determined by a minimum of 20% shared genes between pathways and at least 2 genes overlap. Please click here to view a larger version of this figure.
Figure 4. qRT-PCR validation of gene expression changes. Changes in gene expression identified by RNASeq were confirmed in age- and treatment-matched 5 dpf embryos, injected with control, alms1, or bbs1 MO. (A) Subset of genes showing mRNA expression of genes using qRT-PCR. (B) Equivalent data showing read counts from the RNASeq dataset. All data are relative to control, and qRT-PCR were normalized to actin. aste, asteroid homolog 1; gck, glucokinase; bmb, brambleberry; btr26, bloodthirsty-related gene family, member 26; ifi35, interferon-induced protein 35. Please click here to view a larger version of this figure.
The approach described in this protocol offers a relatively rapid and cost-effective strategy for transcriptome-level analysis of whole animals or specific sorted cell populations. The zebrafish provides an advantageous model for this type of study because of the ease and rapidity in generating large amounts of starting material, the ease of implementing genetic or environmental experimental conditions, and the availability of a large spectrum of transgenic reporter lines allowing for isolation of cell-type specific and tissue-specific populations.
While not detailed here, this method could easily accommodate tissue sections collected from juvenile or adult fish as an alternative to FACs and collection. We also developed a basic follow-up analysis strategy that only relies on the use of basic spreadsheet commands and pre-existing pathway and GO databases. The major advantage of this approach is the ease of accessibility without requiring high-level expertise in computer programming or database management. As such, this strategy is applicable for investigators of all backgrounds for informative analysis of large gene expression data sets.
Our approach combines basic zebrafish embryo and larval culture with standard RNA extraction techniques followed by RNASeq. RNASeq requires a large quantity of high-quality starting material; some vendors require 2 µg of total RNA. As a result, rare/infrequent cell types or individual embryo analysis is out of reach. However, the last five years have demonstrated dramatic improvements in low-yield analyses, datasets generated from smaller samples, and lower RNA quantities, and we expect that these applications will be possible in the near future. To ensure accurate results, the addition of both technical replicates, i.e., generating two RNA libraries from a single sample, and biological replicates, i.e., collecting embryos from multiple matings, is ideal. These steps allow the researchers to control for variation in samples, and sequences identified in all samples are more likely to be true results, with the reduced likelihood of missing a lower read-count result.
The authors have nothing to disclose.
This work was supported by R01DK102001 (N.A.Z.), P30DK072488 (N.A.Z.), and T32DK098107 (T.L.H. and J.E.N.).
Commercial Reagents | |||
TriZol | Thermo Scientific | 15596026 | lysis reagent |
TrypLE | Gibco | 12604013 | dissociation buffer 1 |
FACSMax | Genlantis | T200100 | dissociation buffer 2 |
DEPC-treated water | Sigma | 95284 | |
FirstStrand cDNA conversion | Thermo Scientific | K1621 | cDNA conversion kit |
2X SYBR Green Master Mix | Roche | 4707516001 | qRT-PCR Master Mix |
FACS buffer | Fisher Scientific | 50-105-9042 | |
chloroform | Sigma Aldrich | 288306 | |
sodium acetate | Sigma Aldrich | S2889 | |
Name | Company | Catalog Number | Comments |
Zebrafish Strains | |||
Tuebingen | ZIRC | ZL57 | |
ins2a:mCherry | ZIRC | ZL1483 | |
Name | Company | Catalog Number | Comments |
Equipment | |||
40 micron cell strainer | Sigma | CLS431750 | |
FACS tube | BD Falcon | 352063 | |
hemocytometer | Sigma | Z359629 | |
Dissecting Microscope | Zeiss | ||
Inverted Microscope | Zeiss | ||
Nanodrop | Thermo Scientific | ||
Illumina HiSeq | Illumina | ||
LightCycler 480 | Roche | ||
Mating tanks 1.0L Crossing Tank Set | Aquaneering | ZHCT100 | |
FACS tube 5 mL polypropylene tube | BD Falcon | 352063 | |
Name | Company | Catalog Number | Comments |
Software | |||
Excel | Microsoft | ||
Consensus Path DB | http://cpdb.molgen.mpg.de/ | ||
GO Enrichment Analysis | http://geneontology.org/page/go-enrichment-analysis |