Gene Therapy (2001) 8, 966–968 2001 Nature Publishing Group All rights reserved 0969-7128/01 $15.00 www.nature.com/gt BRIEF COMMUNICATION Standard heparin, low molecular weight heparin, low molecular weight heparinoid, and recombinant hirudin differ in their ability to inhibit transduction by recombinant adeno-associated virus type 2 vectors UT Hacker1,2,*, FM Gerner1,3,*, H Bu¨ ning1, M Hutter1, H Reichenspurner3, M Stangl4 and M Hallek1,2,5 1Gene Center of the Ludwig-Maximilians-University Munich; 2Medical Department III, 3Department of Cardiac Surgery, 4Department of Surgery, University Hospital Munich/Grosshadern, Ludwig-Maximilians-University, Munich; and 5GSF, National Research Center for Environment and Health, Munich, Germany Recombinant adeno-associated virus type 2 (rAAV) is a and 37.1 ± 1.8% at 1.0 U/ml. The inhibitory effect was sig- promising vector for in vivo gene therapy. Transduction by nificantly weaker compared with heparin at 1.0 U/ml, rAAV requires binding to heparan sulfate proteoglycan on (P Ͻ 0.01). The LMW heparinoid danaparoid inhibited rAAV the cell surface, and heparin can block this binding. Because transduction by 8.8 ± 3.5% at 0.1 U/ml (P Ͻ 0.01 compared heparin is administered to most patients undergoing cardio- with heparin). In contrast, recombinant hirudin did not inter- vascular gene transfer in order to prevent thrombotic events, fere at all with rAAV transduction. In summary, the results it is important to identify anticoagulants which do not inter- demonstrate that inhibition of rAAV transduction by heparin fere with rAAV transduction. Therefore, we examined the occurs rapidly and at therapeutically used concentrations. influence of different anticoagulants on rAAV transduction in LMW heparinoids and above all recombinant hirudin might vitro. rAAV transduction was inhibited by 40.5 ± 7.9% at hep- be alternatives for heparin when vascular gene transfer with arin concentrations of 0.1 U/ml, and by 81.7 ± 3.6% at rAAV requires transient anticoagulation. Gene Therapy 1.0 U/ml. The low molecular weight (LMW) heparin tinzapa- (2001) 8, 966–968. rin inhibited rAAV transduction by 20.2 ± 3.8% at 0.1 U/ml Keywords: adeno-associated virus type 2; anticoagulants; heparin; gene transfer; inhibition Recombinant adeno-associated virus type 2 (rAAV) is a the influence of four anticoagulants, heparin, tinzaparin, viral vector system which allows long-term transgene danaparoid and recombinant hirudin, on the transduc- expression after in vivo application in various tissues.1–5 tion efficiency of rAAV in vitro. Heparin exerts its rAAV has the potential to be used for the treatment of antithrombotic effect through anti-thrombin III (AT III) genetic diseases,4,6,7 cancer gene therapy8,9 and cardio- mediated inhibition of coagulation factor Xa.18 Tinzaparin vascular diseases.10–12 Results demonstrating that rAAV sodium is a low molecular weight (LMW) heparin, pro- can mediate efficient gene transfer into rat arteries10 and duced by enzymatic depolymerization of unfractionated into mouse hearts12 in vivo underline that rAAV might heparin.19 It is a heterogeneous mixture of sulfated poly- be a promising vector for cardiovascular gene therapy. saccharide glycosaminoglycan chains with a mass rang- Recently, heparan sulfate proteoglycan has been pro- ing between 5500 and 7500 daltons. Danaparoid is a posed to be an important attachment molecule13 or recep- LMW heparinoid consisting of heparan sulfate, dermatan tor14 for rAAV. Transduction efficiency of rAAV is sulfate and chondroitin sulfate. Binding to AT III leads decreased in a dose-dependent manner in the presence to an accelerated inhibition of factor Xa resulting in the of heparin by abolishing rAAV binding to, and transduc- antithrombotic effect of danaparoid.20 Hirudin is a potent tion of, target cells.14 Because heparin is administered to thrombin inhibitor originally derived from the medicinal most patients undergoing cardiovascular gene transfer in leech which can now be produced using recombinant order to prevent the incidence of thrombotic events,15–17 technology. Unlike heparin, hirudin acts directly on it is important to identify anticoagulants which do not thrombin, rather than through other clotting factors.21 interfere with rAAV transduction. Therefore, we tested It has recently been demonstrated that heparin can inhibit rAAV-mediated gene transfer in a dose-depen- dent manner by binding to rAAV and thus blocking viral Correspondence: M Hallek, Medical Department III, University Hospital attachment to target cells.14 During vascular interventions Munich/Grosshadern, LMU, Marchioninistr. 15, 81377 Munich, Germany and cardiovascular surgery heparin is routinely used to 15,18 *UTH and FMG contributed equally to this work prevent the incidence of thrombotic events. In order Received 16 January 2001; accepted 19 March 2001 to identify the effects of different clinically used antico- Inhibition of transduction by rAAV vectors UT Hacker et al 967 agulants on rAAV-mediated HeLa cell transduction, we inhibition after a 10 min incubation time. Prolongation of examined the inhibitory effects of heparin, LMW heparin, the incubation time from 10 min to 2 h did not signifi- LMW heparinoid and recombinant hirudin at equipotent cantly enhance the inhibition of transduction (Figure 1a). concentrations. When using heparin at concentrations When using a 10-fold concentrated virus a significantly from 0.1 to 1000 anti-factor-Xa U/ml, rAAV-mediated lower inhibition of transduction of 18.1 ± 1.2% was target cell transduction was blocked up to 100% in a detected at low heparin concentrations of 0.1 U/ml, com- dose-dependent manner (Figure 1a). At concentrations of pared with a one-fold concentrated virus preparation 0.1 U/ml, heparin inhibited viral transduction by (P = 0.029). However, at higher heparin concentrations 40.5 ± 7.9%, and 1.0 U/ml heparin caused 81.7 ± 3.6% (1.0–1000 U/ml) no significant differences could by ident- ified (Figure 2). Taken together, these data confirm the results of Summerford and Samulski,14 and demonstrated that heparin decreased rAAV-mediated target cell trans- duction. Furthermore, they indicate that the inhibition of rAAV transduction efficiency by heparin may be a rel- evant problem in the clinical setting when rAAV is used as a vector system for cardiovascular gene transfer because (1) it occurs at therapeutical heparin concen- trations of 0.1–1 U/ml, (2) the binding of heparin to rAAV is a rapid event, and (3) increasing the viral titer by a factor of 10 can reduce the inhibitory effect of heparin only at a low concentration of 0.1 U/ml. In the presence of the LMW heparin tinzaparin, inhi- bition of target cell transduction was 20.2 ± 3.8% at 0.1 anti-factor-Xa U/ml and 37.1 ± 1.8% at 1.0 U/ml. Nearly complete inhibition occurred only at a high concentration of 1000 U/ml, a dose which is not used clinically. A pro- longation of the incubation time (2 h versus 10 min) had Figure 1 For the generation of rAAV virions, 293 cells (a gift from M no influence on the results (Figure 1b). Compared with Lose, Max-Planck-Institute of Biochemistry, Martinsried, Germany) were heparin, the inhibitory properties of tinzaparin were sig- cotransfected by calcium phosphate precipitation with the plasmids pRC, nificantly weaker at a concentration of 1.0 U/ml containing the AAV rep and cap genes, pGFP, containing the EGFP (P Ͻ 0.01). cDNA (Clontech, Palo Alto, CA, USA) within the adeno-associated virus Danaparoid consists of a mixture of the glycosamino- ITR-sequences (both kindly provided by Anne Girod, Gene Center, Mun- ich, Germany), and pXX6, containing the adenovirus helper genes (kindly glycans heparan sulfate, dermatan sulfate and chondro- provided by RJ Samulski22 University of North Carolina at Chapel Hill, itin sulfate. It is a low molecular weight heparinoid. NC, USA). After 48 h, the cells were lysed and fractioned ammonium Therapeutic concentrations of 0.1 U/ml inhibited HeLa sulfate precipitation was performed. rAAV virions were further purified cell transduction by 8.8 ± 3.5%, (P Ͻ 0.01) compared with and concentrated by ultracentrifugation on a CsCl gradient = heparin (Figure 1c). Full inhibition was obtained at a con- (P 1.37 g/ml) and column chromatography. Genomic titers were determ- centration of 1000 U/ml which is not utilized in in vivo ined by dot blot as described.23 For determination of the infectious titers, HeLa cells were plated in 12-well plates at 7 × 104 cells per well. Twenty- conditions. The incubation time did not significantly alter four hours later the cells were irradiated with 100 Gy and serial dilutions the inhibitory effect of danaparoid (Figure 1c). The differ- of the recombinant virus were added to the cells. After 48 h, the number ences in the inhibitory properties between the com- of GFP-expressing cells was quantified by flow cytometry. Using these pounds can be explained by differences in their chemical × 11 methods, titers up to 8 10 genomic particles/ml and infectious titers composition. Danaparoid is a mixture of heparan sulfate, × 8 of up to 8 10 infectious particles/ml were detected. For the inhibition dermatan sulfate and chondroitin sulfate.20 It has been experiments HeLa cells were pretreated as described above. rAAV at a concentration of 3.2 × 104 infectious particles/ml was incubated with serial demonstrated previously that dermatan sulfate and chon- dilutions of anticoagulants for 10 min or 2 h at room temperature in droitin sulfate exert only weak inhibitory effects on serum-free DMEM (Biochrom, Berlin, Germany). 250 ␮l of these sol- rAAV-mediated target cell transduction.14 utions were then added to the HeLa cells, corresponding to an MOI of 0.1. After a 2 h incubation time, the cells were washed twice with PBS, and DMEM/10% FCS was added.