BackgroundCurrently, treatment of autoimmune diseases is based on manipulation of general control mechanisms, including those mediated by immunoregulatory cytokines. This approach is non-curative and may cause unwanted side effects due to numerous beneficial and non-redundant functions of a particular cytokine.MethodsTechniques of reverse genetics, such as conditional gene targeting, were employed to uncover the contributions of two proinflammatory and immunomodulatory cytokines, tumour necrosis factor (TNF) and interleukin 6 (IL-6), in various disease states.ResultsSeveral non-redundant functions of TNF from distinct cellular sources were identified. TNF from myeloid cells is pathogenic in several autoimmune diseases, whereas TNF produced by T cells showed non-redundant protective functions in experimental arthritis and in a Mycobacterium tuberculosis infection model. To test the idea of selective pharmacological inhibition of “bad” TNF produced by myeloid cells while sparing “good” TNF produced by T lymphocytes, a myeloid-specific TNF inhibitor (MYSTI) was designed—a recombinant mini-antibody with dual specificity that can bind to the surface molecule F4/80 on myeloid cells and to TNF. In vitro experiments confirmed retention of TNF on the surface of TNF-producing cells and in vivo experiments indicated that MYSTI can protect mice from lethal TNF-mediated hepatotoxicity. MYSTI is also effective in experimental arthritis.ConclusionThe proposed therapeutic strategy may be more effective than systemic anti-cytokine therapy in several human autoimmune diseases, as it would preserve the potentially beneficial effects of the same cytokine produced by other cell types. Such bispecific biological agents may become interesting tools for experimental studies and, eventually, drug development.