Rheumatoid arthritis (RA) is a chronic autoimmune inflammatory disease characterised by cartilage and bone erosion. In order to target the joint and the inflammation process specifically, avoiding systemic secondary effects of biological compounds, we have: engineered T cells to recognise collagen type II (CII), a main constituent of cartilage, with a chimeric receptor composed of a scFv extracellular domain which binds to CII and an intracellular domain which signals inducing cytokine production and T cell proliferation; and modified cytokines with a biodegradable 'shell' that is removed in sites of inflammation by MMP activity.

Primary T cells grafted with the chimeric scFv receptor with a ζ chain intracellular domain via retroviral transduction, form homodimeric receptors in the cell surface recognising CII in vitro both in solution and attached to plastic and respond producing IL-2 and IFN-γ. KLH-specific T cells produce also IL-4 when challenged with KLH but upon grafting with the chimeric receptor lose the ability to produce this Th2 cytokine. This is in agreement with preliminary findings that engineered T cells are arthritogenic in vivo following CII challenge.

We have produced a fusion protein between a cytokine and a protective protein subunit linked via an MMP cleavage site. We show in vitro that the engineered fusion protein is inactive (incapable of binding to its cellular receptors) unless cleaved by MMP1, MMP3 or by synovial fluid of RA patients. Intramuscular injection of the plasmid expressing a 'mutated' active versus the 'inactive' form of the protein showed that the latter was more effective as expected in the treatment of collagen-induced arthritis in DBA/1 mice.

Engineering cells and molecules to achieve site-specific activation has advantages for gene therapy of RA.