Scientific evidence points to the important role of IFN-gamma in AA development and it has been shown that IFN-gamma -/- mice do not develop AA and that children with AA can be treated with humanized IFN-gamma antibodies. However, “anti-gamma” therapies deprive the organism of the vitally important IFN-gamma and bear considerable side effects. Therefore, the aim of this project is to develop a new approach for AA treatment, which is based on the suppression of IFN-gamma activity. This is achieved by engineering of biologically inactive recombinant analogues of the human IFN-gamma. They are designed to be endowed with the same (or close) affinity as that of the native IFN-gamma to the IFN-gamma receptor but to be incapable of triggering the intracellular signal transduction pathway. This was done on the basis of the three-dimensional structures of the IFN-gamma homodimer and its receptor. Therefore, random mutations have been introduced in one (upstream) of the nuclear localization sequences (NLS) and/or the C-terminus (bearing the downstream NLS). The mutations were designed to cause minimal or no effect on the of hIFN-gamma tertiary structure and therefore on its ability to bind the hIFN-gamma cell receptor. As a result a combinatorial expression library carrying mutant IFN-gamma genes were created and more than 100 clones were well characterized. The mutant proteins were tested for their residual antiviral and antiproliferative activities by measuring the protective effect on WISH cells against the cytopathic action of the vesicular stomatitis virus, the antiproliferative activity was determined by a modified kynurenine bioassay and the less actives were further examined for their ability to compete with the wild type IFN-gamma and its receptor. On that basis several prospective hIFN-gamma mutants, such as RAN26, or RAN134, were selected and we will report why those are promising new molecules for treatment of AA.