A systems biology approach to hematopoietic stem-cell differentiation

Genomic-level expression profiling, facilitated by the development of DNA micro or macro-arrays, allows measurement of the transcriptional response of a large set of genes to a cellular state in a single assay. The expression profile constitutes a detailed molecular phenotype of the cellular state, whether it is a disease, a process such as cell division or differentiation, or a response to a chemical or environmental perturbation. Although still in its technological infancy, DNA arrays are a powerful new tool (the first of the high throughput genomic tools) that will change the way we do basic and applied biological research, IP protection and process validation. Data are presented on the temporal expression profile of stem-cell differentiation along the granulocytic (G) lineage using Clontech/BD Plastic 8K Human Arrays to survey transcript levels of over 8,300 human genes. Data from several time points of the 14-day cultures were normalized by the new mask method of ratio normalization developed in our laboratory. This mask method was used to conservatively filter out genes with low fold changes (at 95% confidence level). A total of 380 genes were observed to be differentially regulated over time point and they were clustered into 5 clusters using self-organizing maps (SOMs). Two clusters displayed sequential upregulation of differentiation and function related genes. Another two exhibited sequential downregulation of protein synthesis and folding related genes. The remaining clustered showed early upregulation followed by downregulation of proliferation and differentiation related genes. Detailed examination of the differentially regulated genes revealed many are categorized into the expected clusters and with functionally complementary genes.