Development of a secreted cell-permeable NF-κB inhibitor to control inflammation

Abstract

Rheumatoid arthritis (RA) is an autoimmune chronic inflammatory disease, of unknown aetiology. Several disease-modulating approaches have been developed in the past years, however these are expensive, usually accompanied by unwanted side-effects and 30% of the patients fail to respond. The transcription factor NF-kB is a key factor in the development and perpetuation of the disease, as it regulates a number of inflammatory genes. The activity of certain signalling pathways can be modulated by delivering into cells inhibitors coupled to Protein Transduction Domains (PTDs). The aim of this study was to develop a secretable PTD-fusion NF-κΒ inhibitor that is produced and secreted by genetically engineered mammalian cells in sufficient amounts to subsequently transduce and regulate NF-κΒ activity in neighbouring cells. Such methodology could be useful for the management of RA by transplantation of engineered cells or directly using gene delivery into the synovial joints. In this study, PTD-fusion proteins were fused to the IL-2 secretion signal and their ability to be secreted from mammalian cells was explored. Secretable forms of TAT-IgG2A and TAT-eGFP were generated as control PTD-fusion proteins, and the TAT-srIκΒα (super repressor IκBα, a NF-κΒ inhibitor) was generated as an NF-κΒ PTD-fusion inhibitor. Western blotting analysis of supernatants from transiently transfected 293T cells revealed that TAT-IgG2A, TAT-eGFP and TAT-srIκΒα are secreted with variable efficiencies. When concentrated, PTD proteins were able to transduce mammalian cells, as demonstrated with Jurkat cells by confocal microscopy and western blotting analysis. The TAT PTD domain was replaced to a more stable, furin cleavage-resistant and less positively charged PTD domain, the TAT3 PTD domain, to ensure that PTDfusion proteins will be secreted more efficiently. This change of the PTD domain did not increase secretion levels of the srΙκBα. Subsequently, the Latent Associated Peptide (LAP) of TGFβ, was fused to the TAT3-srIκΒα inhibitor, via a matrix metalloproteinase (MMP) cleavage linker. This LAP-MMP-PTD-fusion NF-κΒ inhibitor was again poorly secreted. In turn, the srIκΒα inhibitor was replaced with a small synthetic NF-κΒ inhibitor, termed Nemo Binding Domain (NBD), in the form of LAP MMP-TAT3-NBD NF-κΒ inhibitor. Western blotting analysis of supernatants from transiently transfected 293T cells revealed that the LAP-MMP-TAT3-NBD was efficiently secreted. The ability of LAP-MMP-TAT3-NBD to inhibit NF-κΒ was tested in vitro with the use of a cell-assay based on HeLa cells that are permanently transfected with the luciferase gene driven by an NF-κB regulated promoter. In this assay, HeLa cells that were treated with the secreted LAP-MMP-TAT3-NBD, showed reduced levels of luciferase activity after IL-1β stimulation. Subsequently, using a replication-deficient lentiviral vector, genetically engeneered DBA/1 fibroblasts (DTF) able to produce the secreted LAP-MMP-TAT3-NBD were generated. The NF-κB inhibitory properties of the secreted LAP-MMP-TAT3-NBD were tested in vivo in the Carrageenan-induced paw oedema, Antigen Induced Arthritis and Air-Pouch acute inflammation models. Paws of mice that were treated with engineered cells or lentivirus encoding LAPMMP- TAT3-NBD demonstrated milder paw swelling, suggesting that LAP-MMPTAT3- NBD had a protective role in the induction of inflammation. However, the LAPMMP- TAT3-NBD did not demonstrate anti-inflammatory effects in the Air-Pouch model. In this study, I present a method to design PTD-fusion proteins that can be efficiently secreted from mammalian cells and I demonstrate a novel gene therapy approach for the local delivery of a therapeutic agent.