Modeling Biological Networks

IV.1 Coordinators
IV.2 Participants
IV.3 Introduction
IV.4 Background and Significance
IV.5 Research Plan
IV.6 Specific Subprojects

IV.7 Connection to Specific Projects 2 (cytoskeleton) and 3 (organogenesis)
IV.8 Timeline

< Previous | Page 33 of 35 | Next >

IV.6.vii.e Research Plan:
IV.6.vii.e.1 Aim 1:

Dr. B. Latinkic presented, at the Ninth International Xenopus Conference in Cambridge, UK, results from injection of synthetic mRNA encoding GATA-4 into Xenopus laevis embryos at the single cell stage. These embryos were then grown to the late blastula stage of development and then animal caps were harvested and cultured in groups of three to four caps each. These animal cap cells form beating myocardiocytes with the same timing as stage-matched embryos and the same expression of every myocardial and endocardial heart marker examined to date.

This observation provides a powerful new tool for studying heart development. Cultured animal caps contain a homogeneous population of unspecified embryonic cells, much akin to embryonic stem cells. The fact that misexpression of GATA-4 can drive these cells down a cardiogenic pathway with normal timing suggests that while the specification of these cells may be somewhat artificial, the transcriptional regulation necessary to direct heart development occurs just as it does within normal heart fields. Additionally, when performing microarray experiments, it is preferable to obtain RNA from the least complex tissue possible (with a homogeneous cell population such as animal caps being the ideal). This reduces the transcriptional "noise" from other tissues, and increases the chance that a localized expression change from a ubiquitously expressed gene will be detected.

We will generate a large number of GATA-4-expressing animal caps and culture them along with embryos stage-matched to different stages of development (stages 10, 10.5, 11, 14, 18, 22, 25, 26, 27, 28, 29, 30, and 32). We will then harvest total RNA from the caps. Using the planned KEH2 microarray, we will hybridize the gene expression for each stage against a pool of RNA from all stages to generate a temporal profile of gene expression in the presumptive heart fields. We will identify the developmental stages at which each gene represented on KEH2 expresses and when the expression changes, to create a "domain of knowledge" about gene expression changes in vivo. To validate this temporal profile, we will examine known heart field and differentiation markers, and ensure that the temporal profile matches the known expression of these genes.

IV.6.vii.e.2 Aim 2:

We will individually coinject synthetic mRNA encoding wild-type XNkx2-3, XNkx2-5, XSRF, Myocardin, and GATAs 4, 5, and 6 with a lineage tracer (GFP mRNA) into 8-cell Xenopus laevis blastulas and select embryos exhibiting GFP within the heart field(s) for further assay. We will then expression profile the RNA from these embryos using the KEH2 to compare the experimental RNA with RNA collected from GFP-injected, stage-matched control embryos. Competitive hybridization and data analysis will use standard procedures. We will observe the transcriptional effects of these overexpressions at a variety of developmental stages. To identify genes regulated by complexes rather than individual transcription factors, we will also profile embryos that have been co-injected with multiple wild-type mRNAs.

To study the effects of inhibiting the entire tinman-related family, we will compare RNA from embryos injected with our XNkx2-3LP and XNkx2-5LP dominant inhibitory mutants to that of control embryos. To inhibit Myocardin, MADS box and the GATA factors, we will employ the injection of antisense morpholino oligonucleotides designed for each factor (see Organogenesis, Section VI)

We will create a theoretical cardiogenic transcriptional network from the temporal profile from aim 1 (IV.6.vii.e.1) and the expression profiles from these experiments. For each "treatment," we will select features that exhibit the most reproducibly significant transcriptional changes and sequence the clones expressing them if no sequence data is available. The resulting expression profiles from each "treatment" can reveal which genes lie downstream from the treatment gene and the type of regulation.

Because we cannot determine or regulate the timing of injected mRNA or morpholino effects, the cardiogenic temporal profile from specific aim 1 is crucial.

For this specific aim, we will rely heavily upon Consortium members and statisticians associated with the CMG for assistance and advice on the construction of the network model. The current conception is a probabilistic network that can be refined as data from additional expression-profiling experiments become available.

IV.6.vii.e.3 Aim 3:

We expect that specific aims 1 and 2 will identify a number of new genes as significant to early heart development. We will look at the putative functional class of each newly discovered gene (based on sequence homology and clustering with known genes), assigning highest priorities to transcription factors and signaling proteins. Where possible, we will use the clustering methods of Subproject 1 and the word identification of Subproject 3 to identify genes that exhibit transcriptional changes prior to cardiac differentiation.

We will examine the temporal and spatial expression for each gene of interest using whole mount in situ hybridization and will note restricted patterns of interest. We will confirm microarray results using other methods such as RT-PCR or RNase protection assays. We will create mRNA or morpholinos for genes that we choose to further characterize to perform transcriptional profiling as in Aim 2.

We will compare the resulting profile to the theoretical cardiogenic transcriptional network model, and refine the model in a step-wise fashion, beginning with genes known to change expression at the earliest stages of development.

In overexpression profiling experiments, we expect a significant upregulation of gene targets for Nkx2-5 and GATA-4 and dramatic differences in cardiomyocyte marker expression in embryos injected with tinman-related dominant inhibitory mutants. Such results and internal expression controls will ensure that other observed transcriptional effects are true effects of our manipulations.

Profiling experiments with whole embryos are difficult. The number and complexity of different cell types increase exponentially with time in the developing embryo. Slight transcriptional changes, or those localized to a small subset of cells, may be difficult to detect in whole embryos. We will investigate the possibility of profiling dissected tissue or cultured animal caps.