The rate-limiting step for any gene therapy application is the efficiency of gene transfer. Two approaches were discussed: direct, in vivo gene delivery and indirect, ex vivo gene transfer using synovial fibroblasts, chondrocytes, antigen-presenting cells (APCs), lymphocytes or mesenchymal stem cells (MSCs). Although a number of presentations discussed improvements in viral and nonviral vectors for direct gene transfer to joints or to bone defects, no vector yet appears ready to be used in in vivo clinical trials.
W van den Berg (Nijmegen, The Netherlands), P Robbins (University of Pittsburgh, USA) and T Oligino (University of Pittsburgh, USA) all presented evidence that adenoviral vectors are able to infect murine and rabbit synovial tissue as well as infecting infiltrating monocytes following injection into the joint space. Expression was transient, however, and the induction of neutralizing antibodies prevents repeat dosing. T Huizinga (Leiden University, The Netherlands) also demonstrated that synovial fluid from many rheumatoid arthritis (RA) patients contains neutralizing antibodies to adenovirus serotype 5 (Ad5), but that there are only low levels of neutralizing antibodies to certain serotypes of adenoviruses in the synovial fluids of RA patients. In particular, Ad35 is able to infect human synovial cells more efficiently than Ad5, but was not neutralized by pre-existing antibodies in synovial fluids. P Yeh (Aventis-Gencell, France) discussed strategies for targeting adenoviruses to particular cell types.
The properties of adeno-associated virus (AAV) were described by AM Douar (Genthon, France). AAV also appears to be able to infect cells in the joint space efficiently, resulting in prolonged gene expression (R Hirsch, University of Cincinnati, USA). Hirsch suggested, however, that AAV fails to transduce murine synovium but gives the appearance of doing so by transducing the adjacent muscle very effectively. Nevertheless, his data suggested that human synovial fibroblasts support transduction by AAV much more effectively than their murine counterparts. Given the initial success of clinical trials using AAV for gene transfer to muscle, there was optimism that AAV could be useful for intra-articular gene delivery as well as for systemic delivery of soluble proteins following intra-muscular injection. Both Oligino and Robbins reported that transgene expression following the intra-articular injection of an AAV vector was progressively lost during the course of 2–3 weeks. Expression could not be restored by readministration of the same vector, possibly because of the induction of neutralizing antibodies. It is not yet known whether such antibodies would be directed against the transgene or the virus. Immunotherapy for arthritis may be achieved through intramuscular injection of recombinant AAV that contains genes encoding anti-inflammatory cytokines. F Apparailly (Montpellier, France) reported the efficacy of viral IL-10 expression under the control of a tetOn inducible promoter using an AAV construct in collagen-induced arthritis, and M-C Boissier (Bobigny, France) also reported efficacy in this model using AAV-IL-4.
P Corbeau (Montpellier, France) discussed the development of lentiviral vectors and their ability to transduce cultures of human synovial fibroblasts, a finding in agreement with data showing that an HIV-based lentiviral vector was able to infect rat synovium following intra-articular IL-1 receptor antagonist (IL-1Ra) injection. Surprisingly high levels of intra-articular transgene expression were obtained (E Gouze, Harvard Medical School, USA).
Two laboratories (C Pitzalis, London, UK and P Robbins) have initiated studies to identify peptides that can target cells within the joint following systemic delivery or those that are able to transduce synovial cells following intra-articular delivery. These peptides could be used to deliver therapeutic proteins, drugs and possible plasmids or viral vectors to cells within the joint. Finally, the use of nonviral vectors was also presented. Although a number of nonviral vectors were able to transfect synovium in vivo following intra-articular injection, the duration of gene expression (of over 100 different nonviral formulations tested) was less than 1 week, with some of the formulations inducing inflammation (P Robbins). The ability to express soluble proteins in the long term from muscle by electroporation of the plasmid DNA encoding them was demonstrated by D Scherman (Aventis Gencell, France). The muscle thus represents an attractive target for regulated systemic expression of IL-1Ra, soluble tumor necrosis factor receptors (TNF-R) or IL-10, especially if naked plasmid DNA can be used as the vector. The same electroporation technology was used with remarkable success for in vivo gene delivery to chondrocytes, with up to 40% of chondrocytes from the patella expressing the transgene for 3 months (P Gillet, Vandoeuvre, France).