Microarray gene expression profiling during the segmentation phase of zebrafish development

https://doi.org/10.1016/j.cca.2004.08.008Get rights and content

Abstract

We analyzed 15,512 unique transcripts from wild-type Danio rerio using a long oligonucleotide microarray containing >16,000 65-mers probes. Total RNA was isolated from staged embryos at 2 h intervals over a 24-h period. On average, at any given time point, 27% of the probe set detected corresponding transcripts in embryonic RNA. There were two predominant patterns in the nearly 4000 genes that changed expression in at least one time point during the first 24 hpf. At 12 hpf, we detected 420 up-regulated and 386 down-regulated genes. By 24 hpf, the number of up- and down-regulated genes had increased to 954 and 766, respectively. While the majority of these genes maintained their new level of expression for the duration of the time course, we identified five genes with phasic regulation over the 24-h time course. Two of these genes, germ cell nuclear factor and mesogenin, have been identified as being expressed during gastrulation (5 1/4 to 10 h postfertilization) and subsequently repressed. A cluster containing 36 distinct ribosomal proteins was up-regulated at 12 h, indicating a capability for de novo protein synthesis during and after this stage. Twenty-three muscle-specific genes were up-regulated late during the initial 24 hpf, corresponding to the development and differentiation of the somites.

Introduction

Microarray analysis provides information regarding the expression of thousands of genes in a single experiment. The power of this technique, however, is not in the analysis of a single experiment, but rather in the analysis of multiple experiments over a period of biological transitions. Microarray analysis can provide information on temporal or spatial patterns of gene expression. Correlations between expression patterns in multiple samples can help to identify differentially expressed genes that play a role in embryogenesis, tissue patterning, organ development, and other physiological processes. Whole genome microarray studies have been used to characterize early developmental gene expression patterns in several model organisms. Multiple time points examined during embryogenesis in Drosophila and Caenorhabditis elegans demonstrate patterns of coordinately regulated genes (Arbeitman et al., 2002, Jiang et al., 2001) Novel genes were identified by spatial differences in gene expression in Xenopus gastrula embryos (Altmann et al., 2001). Both somite formation and liver development have been studied in the mouse embryo and microdissected tissue (Buttitta et al., 2003, Jochheim et al., 2003). Developmental pathways have been dissected in sea urchin using a combination of array analysis coupled with gene knockdown techniques (Davidson et al., 2002).

The zebrafish serves as a model for understanding normal vertebrate development as well as dissecting the mechanisms underlying human disease. Zebrafish are ideal vertebrate models because of their small size, ease of culture, and transparent embryos. The embryos are amenable to manipulation and numerous mutant phenotypes have been identified in large-scale mutagenesis screens (Driever et al., 1996, Mullins and Nusslein-Volhard, 1993, Amsterdam et al., 1999). Pathways of gene regulation can be discerned when microarray analysis is combined with the injection of gene-specific, antisense morpholinos (Nasevicius and Ekker, 2000). Many aspects of vertebrate development, including those of humans, are mirrored in the zebrafish embryo. Therefore, transcript profiling of early zebrafish development and how it can be affected by environmental influences should facilitate our understanding vertebrate embryogenesis.

Several developmental processes are evident during the first 24 h. For example, the segmentation phase is characterized by the development of somites from which skeletal muscle, dermis, ribs, and vertebrae form. New somites form approximately every 20 min between 10 and 12 hpf and every 30 min between 12 and 24 hpf (Kimmel et al., 1995). Neurulation occurs coincident with segmentation and inductive signals from adjacent tissues play a role (Yoda et al., 2003). Myogenic differentiation of cardiac tissue and somites is coordinated by expression of cell-specific and cell-restricted transcription factors (Lassar and Münsterberg, 1994). In particular, the transcription factors MEF2, myoD, and hsp90α regulate skeletal and cardiac muscle-specific gene expression in zebrafish development (Ticho et al., 1996, Sass et al., 1996).

In this study, we used an available library of zebrafish long oligonucleotide probes representing 15,512 unique genes to evaluate the sensitivity and utility of an oligonucleotide-based microarray resource for analyzing specific stages of zebrafish development. The probes were 65 nucleotides in length, as compared to more traditional cDNA array probes that are typically thousands of base pairs in length. The long oligonucleotide probes were designed to generate minimal cross-hybridization between closely related gene family members and to maximize hybridization across splice variants of the same gene. In contrast, cDNA probes, with their longer sequence, tend to have more incidences of cross-hybridization between closely related genes. Such differences in sequence lengths, which can affect cross-hybridization efficiencies, may explain potential expression differences between the two probe types. We corroborated our long oligonucleotide microarray results using two complementary methods, quantitative real-time PCR and in situ hybridization. We identified several genes known to be involved in zebrafish development, particularly myogenesis, as well as novel genes with unique temporal expression patterns.

Section snippets

Embryo husbandry and RNA isolation

Two separate collections of staged embryos were provided by Scientific Hatcheries. Procedures for feeding, spawning, and hatching are available at http://www.scientifichatcheries.com/zebraFishFolder/Zebrafish_page.html. Breeders were maintained at 28±0.5 °C. The embryos were staged by hours postfertilization (hpf) (Kimmel et al., 1995). To study gene expression during early zebrafish development, we isolated total RNA from two pools of staged embryos (Scientific Hatcheries) independently

Quality control of hybridizations

To assess the reproducibility of microarray hybridizations and embryo variability, we determined the correlation between technical (dye-swap hybridizations) and biological (two separate collections of staged embryos) replicates. For each of the nine time points, two independent embryo collections (pools) were obtained, from which RNA was isolated (Pool 1, nine time points; Pool 2, nine time points). We performed dye-swap hybridizations, removed poor quality elements, and LOWESS normalized the

Discussion

Few reports on global gene expression during zebrafish development have been described thus far. The first published study utilized a cDNA microarray consisting of ∼4500 unique genes (Ton et al., 2002). A second study utilized a cDNA microarray representing ∼3100 unique clusters (Lo et al., 2003). Both of these studies examined gene expression at multiple time points during development but few time points during the initial 24 h when the nervous system is forming and most major organ systems

Acknowledgements

This work was supported by NIH grants U19 ES11375 (E.L. and R.L.M.), R03 AG022703 (R.L.M.), and PO1 HD39948 (E.L.). We would like to thank Jennifer Tsai, of the TIGR Gene Indices Group, for invaluable bioinformatic support and Liqing Li for some of the RNA isolations.

References (35)

  • C. Ton et al.

    Construction of a zebrafish cDNA microarray: gene expression profiling of the zebrafish during development

    Biochem. Biophys. Res. Commun.

    (2002)
  • T. Toramoto et al.

    Multiple gene organization of pufferfish Fugu rubripes tropomyosin isoforms and tissue distribution of their transcripts

    Gene

    (2004)
  • J.A. White et al.

    Identification of the retinoic acid-inducible all-trans-retinoic acid 4-hydroxylase

    J. Biol. Chem.

    (1996)
  • H. Yoda et al.

    An expression pattern screen for genes involved in the induction of the posterior nervous system of zebrafish

    Differentiation

    (2003)
  • A. Amsterdam et al.

    A large-scale insertional mutagenesis screen in zebrafish

    Genes Dev.

    (1999)
  • M.N. Arbeitman et al.

    Gene expression during the life cycle of Drosophila melanogaster

    Science

    (2002)
  • P.L. Bardet et al.

    Characterization of oestrogen receptors in zebrafish (Danio rerio)

    J. Mol. Endocrinol.

    (2002)
  • Cited by (0)

    This paper is based on a presentation given at the conference: Aquatic Animal Models of Human Disease hosted by the American Type Culture Collection and the University of Miami in Manassas, Virginia, USA, September 29–October 2, 2003.

    View full text