Microarrays are powerful tools that can impact new diagnostic procedures and expedite the drug development process. A significant but largely unrealized application is gene expression analysis from RNAlimited samples, such as blood, needle biopsies, laser capture microdissection (LCM) samples, and even single cells. The most widely used T7 RNA polymerase-based amplification method currenly lacks the necessary sophistication to meet these emerging needs. To overcome these limitations, we propose to increase the efficiency of T7 RNA amplification using an innovative protein engineering approach. Through combined rational design and random mutagenesis, mutant library selection and screening, we aim to identify a hyperactive T7 RNA polymerase that is 4-5 times more kinetically proficient than the current enzyme. We will expand this novel approach in phase II to include the discovery of mutant T7 polymerase enzymes that are both more thermostable (and thus extend high efficiency transcription for longer times, creating more product) and can incorporate biotinylated and Cy-modified nucleotides 5- to 10-fold more efficiently. Taken together, these improvements will lower the demand for input RNA by approximately 20-fold, and reduce the expense of modified nucleotides by as much as an order of magnitude