Gene Therapy (2000) 7, 1417–1420  2000 Macmillan Publishers Ltd All rights reserved 0969-7128/00 $15.00 www.nature.com/gt VIRAL TRANSFER TECHNOLOGY BRIEF COMMUNICATION Hyaluronidase enhances recombinant adeno- associated virus (rAAV)-mediated gene transfer in the rat skeletal muscle D Favre 1 , Y Cherel 2 , N Provost 1 , V Blouin 1 , N Ferry 1 , P Moullier 1 and A Salvetti 1 Laboratoire de The ´rapie Ge ´nique, CHU Ho ˆtel-Dieu, Nantes; and 2 UMR INRA-Ecole Ve ´te ´rinaire, Nantes, France Skeletal muscle is a privileged target for long-term rAAV- mediated gene transfer in mouse, rat, dog and non-human primates. Intramuscular injections of rAAV encoding human factor IX in hemophilia B patients have been initiated, based on promising results gathered in affected dogs. We found that intramuscular rAAV administration in rats resulted in restricted transduction essentially along the myofibers axis with poor lateral diffusion. This suggested that the transduc- tion rate might be limited by the ability of the virus to reach sites distant from the injection point. We tested whether hya- luronidase, an enzyme which dissociates the extracellular matrix, could enhance vector diffusion when injected in the rat muscle before administration of rAAV encoding either nuclear-localized -galactosidase (rAAVCMVnlsLacZ) or the human -1-antitrypsin (rAAVCMVhAAT) under the control of the cytomegalovirus immediate–early promoter (CMV). The Keywords: skeletal muscle; recombinant adeno-associated virus; hyaluronidase Recombinant adeno-associated viral vectors (rAAV) have emerged as an attractive alternative to retroviral and adenoviral vectors to transduce nondividing cells in vivo. 1 Recombinant AAV vectors are derived from the ‘wild- type’ virus by deleting the rep and cap genes and replac- ing them with the transgene and transcription control elements. Usually, rAAV stocks are purified from 293 cells which are cotransfected with the AAV vector and a rep-cap expressing plasmid and subsequently infected with helper adenovirus. Alternatively, the adenoviral helper functions required for the production of rAAV can be provided by a nonreplicative plasmid resulting in rAAV stocks free of detectable adenovirus. 2 Despite the broad host range of rAAV in vitro, three preferential targets have been identified in vivo: the ret- ina, skeletal muscle, and the central nervous system, where rAAV efficiently transduced post-mitotic and dif- ferentiated cells such as photoreceptors, myofibers and neurons. 3–5 Skeletal muscle is an easily accessible target for gene delivery and is highly vascularized. Efficient rAAV transduction of myofibers can be used to achieve the sustained in situ expression of reporter genes or the Correspondence: P Moullier, Laboratoire de The ´rapie Ge ´nique, CHU Hotel-Dieu, 44035 Nantes, France Received 18 February 2000; accepted 3 May 2000 results showed that pretreatment of the rat anterior tibialis muscle with hyaluronidase resulted in: (1) a larger diffusion of the virus indicated by an increase in the area containing LacZ-transduced fibers, and (2) a two- to three-fold increase of transduction efficiency measured by the number of LacZ- positive fibers or by the hAAT serum concentration. We also provide evidence that hyaluronidase was well tolerated and was not associated with short- or long-term toxicity evalu- ated by morphological studies. Finally, in our experimental conditions, hyaluronidase did not promote rAAV dissemi- nation to other organs as assessed by PCR to detect vector sequences. We conclude that pretreament of skeletal mus- cle by hyaluronidase, a clinically available reagent, was harmless and resulted in a consistent and significant increase in rAAV diffusion and transduction levels. Gene Therapy (2000) 7, 1417–1420. systemic delivery into the circulation of therapeutic pro- teins. The mouse has been used most commonly for rAAV evaluation in vivo and the anterior tibialis muscle remains the preferred site of vector administration. The small size of the target combined with large volumes of injected rAAV (30–100 l) allow diffuse transduction of the muscle reaching up to 40% of myofibers. 6,7 In the present study rAAV-mediated transduction of the rat anterior tibialis muscle was analyzed. Intramuscu- lar injection of rAAVCMVnlsLacZ into 250 g Wistar rats resulted in transduction essentially along the axis of myofibers with little lateral diffusion (data not shown). On average, 1% myofibers were routinely transduced using 7.8 × 10 9 rAAVCMVnlsLacZ particles (2.5 × 10 7 transducing particles as measured by LacZ staining of adenovirus co-infected 293 cells) in a total volume of 250 l saline, injected perpendicularly at three different sites at equal distances along the longitudinal axis. Increasing the total volume of injection to 500 l did not improve the level of transduction, nor did the longitudinal admin- istration of the rAAV vector (data not shown). Actually, increasing the volume above 250 l resulted in obvious leakage outside the muscle despite making a cutaneous incision and exposing the muscle before rAAV injection. Overall, these observations indicated that rAAV had a restricted capacity to diffuse from the injection site and suggested that the confined diffusion of the vector was limiting the transduction rate of muscle fibers.