Controlled release of small interfering RNA targeting midkine attenuates intimal hyperplasia in vein grafts

Objective: Intimal hyperplasia is a major obstacle to patency after vein grafting. Despite of a diverse array of trials to prevent it, a satisfactory therapeutic strategy for clinical use has not been established. However, sufficient inhibition of early stages of intimal hyperplasia may prevent this long-term progressive disease. Midkine (MK) is a heparin-binding growth factor that was originally discovered as the product of a retinoic acid-responsive gene. We previously demonstrated that MK-deficient mice exhibit a striking reduction of neointima formation in a restenosis model, which is reversed on systemic MK administration. In this study, we evaluated a strategy of using small interfering RNA (siRNA) targeting MK as a therapy for vein graft failure. Methods: We first made a highly effective siRNA to rabbit MK. Jugular vein-to-carotid artery interposition vein grafts, which are applied to a low flow condition, were made in Japanese white rabbits. Small interfering RNA mixed with atelocollagen was administrated to the external wall of grafted veins. Cy3-conjugated stabilized siRNA was used to confirm its stability and successful transfer into the vein graft wall. Neointimal hyperplasia was evaluated 4 weeks after the operation. The proliferation index and leukocyte infiltration were determined. Results: MK expression was induced and reached the maximum level 7 days after operation. Fluorescence of Cy3-labeled siRNA could be detected in the graft wall even 7 days after operation. Knockdown of the gradually increasing expression was achieved by perivascular application of siRNA using atelocollagen. The intima-media ratio and the intima thickness at 28 days after grafting were both reduced > 90% by this treatment compared with controls. This phenomenon was preceded by significant reductions of inflammatory cell recruitment to the vessel walls and subsequent cell proliferation in MK siRNA-treated grafts. Conclusions: These results suggest that midkine is a candidate molecular target for preventing vein graft failure. Furthermore, for clinical applications of siRNA, a single intraoperative atelocollagen-based nonviral delivery method could be a reliable approach to achieve maximal function of siRNA in vivo. This strategy may be a useful and practical form of gene therapy against human vein graft failure.