Abstract
Gene expression profiling relies on mRNA extraction from defined cell systems, which in the case of pathological processes necessarily results in the use of small quantities of tissues, sometimes as little as a few cells. This obviates the use of many systems of gene expression profiling and is best carried out using cDNA amplified by poly(A) reverse transcription polymerase chain reaction, which is capable of generating material representative of all the expressed genes in samples as small as one cell. Analysis of this material using subtractive hybridization compares the genes expressed at different stages of a biological/pathological process allowing identification of the all the genes upregulated during the process. The identification of the genes present is not dependent on their prior description or on the choice of genes used in a screen and as such the method is ideal for identifying novel genes or unsuspected genes. We have used the method to identify genes involved in normal osteoblastic differentiation and in Paget's disease of bone and it has been widely used to study normal differentiation and pathological processes in a number of systems. The method, its applications and its relationship with the other methods of gene expression profiling are reviewed.