AAV2 production with optimized N/P ratio and PEI-mediated transfection results in low toxicity and high titer for in vitro and in vivo applications

G

V

Ati

XQ1

Ja

b

ARRAA

KRiIGGOAABCHCPNQ2

1

s1vlJIbltles

0h

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

ARTICLE IN PRESS Model

IRMET 12170 1–8

Journal of Virological Methods xxx (2013) xxx– xxx

Contents lists available at SciVerse ScienceDirect

Journal of Virological Methods

jou rn al hom ep age: www.elsev ier .com/ locate / jv i romet

AV2 production with optimized N/P ratio and PEI-mediatedransfection results in low toxicity and high titer for in vitro andn vivo applications

inping Huanga,∗, Antja-Voy Hartleya, Yishi Yina,eremy H. Herskowitza, James J. Laha, Kerry J. Resslerb

CND, Department of Neurology, Emory University, Atlanta, GA 30322, United StatesDepartment of Psychiatry and Behavioral Sciences, Emory University, Atlanta, GA 30322, United States

rticle history:eceived 30 August 2012eceived in revised form 22 May 2013ccepted 3 June 2013vailable online xxx

eywords:ecombinant virus

n vivonfectioneneticene expressionptimizationAV

a b s t r a c t

The adeno-associated virus (AAV) is one of the most useful viral vectors for gene delivery for both in vivoand in vitro applications. A variety of methods have been established to produce and characterize recom-binant AAV (rAAV) vectors; however most methods are quite cumbersome and obtaining consistentlyhigh titer can be problematic. This protocol describes a triple-plasmid co-transfection approach with25 kDa linear polyethylenimine (PEI) in 293T cells for the production of AAV serotype 2. Seventy-twohours post-transfection, supernatant and cells were harvested and purified by a discontinuous iodixanoldensity gradient ultracentrifugation, then dialyzed and concentrated with an Amicon 15 100,000 MWCOconcentration unit. To optimize the protocol for AAV2 production using PEI, various N/P ratios and DNAamounts were compared. We found that an N/P ratio of 40 coupled with 1.05 �g DNA per ml of media(21 �g DNA/15 cm dish) was found to produce the highest yields for viral replication and assembly mea-sured multiple ways. The infectious units, as determined by serial dilution, were between 1 × 108 and2 × 109 IU/ml. The genomic titer of the viral stock was determined by qPCR and ranged from 2 × 1012 to6 × 1013 VG/ml. These viral vectors showed high expression both in vivo within the brain and in vitro in

AV2-GFPrainortexippocampuserebellum

cell culture. The use of linear 25 kDa polyethylenamine PEI as a transfection reagent is a simple, morecost-effective, and stable means of high-throughput production of high-titer AAV serotype 2. The use ofPEI also eliminates the need to change cell medium post-transfection, lowering cost and workload, whileproducing high-titer, efficacious AAV2 vectors for routine gene transfer.

45

46

47

48

49

50

51

52

53

54

55

56

olyethylenimine (PEI)/P ratio

. Introduction

Recombinant adeno-associated viruses (rAAV), a 22 nm naked,ingle-stranded DNA member of parvoviridae family (Hoggan et al.,966), has emerged as one of the preferred gene-delivery viralectors in the past decade. Use of rAAV has lead to promisingeaps in gene therapy (Bainbridge et al., 2008; Bowles et al., 2012;ayandharan et al., 2011; Kaplitt et al., 2007; Niemeyer et al., 2009).t has become such a powerful tool in the laboratory and clinicecause of its ability to infect a large range of tissues. It exhibits

ong-term gene transduction, offers a wide variety of cell andissue tropism due to availability of multiple serotypes, exhibits

Please cite this article in press as: Huang, X., et al., AAV2 production wittoxicity and high titer for in vitro and in vivo applications. J. Virol. Methods

ow immunogenicity, and has an overall good safety profile (Lockt al., 2010; Wright, 2009). High-throughput production of rAAVerotype 2 (rAAV2) is the goal of our viral vector core, and it requires

∗ Corresponding author.E-mail address: [email protected] (X. Huang).

166-0934/$ – see front matter © 2013 Published by Elsevier B.V.ttp://dx.doi.org/10.1016/j.jviromet.2013.06.008

57

58

59

© 2013 Published by Elsevier B.V.

the subsequent implementation and optimization of a productionprotocol that is adaptable to multiple inserted transgenes, in orderto obtain a pure and high titer virus without extensive workloadand cost.

In this protocol, rAAV2 was produced by a three-plasmid co-transfection system in HEK293 T cells, wherein (1) a helper plasmid(pHelper) contained the adenovirus genes E2a, E4 and viral asso-ciated RNAs necessary for AAV reproduction, (2) a vector plasmidcarries the transgene flanked by inverted terminal repeats (ITRs),and (3) the packaging plasmid provides the AAV serotype 2 replica-tion (rep2) and encapsidation (cap2) genes. The HEK293 T cell lineprovides trans-acting adenovirus E1 genes. This protocol reports anoptimization of rAAV2 production by reducing the amount of DNAused for transfection (to 1.05 �g/ml), which decreased the amountof work and cost in DNA purification while observing no decrease

h optimized N/P ratio and PEI-mediated transfection results in low (2013), http://dx.doi.org/10.1016/j.jviromet.2013.06.008

in viral titers.Established calcium-phosphate transfection (Ca–P) has been

found to be more laborious, expensive, and yield inconsistent andunstable results because of narrow optimal conditions, especially

60

61

62

63

dx.doi.org/10.1016/j.jviromet.2013.06.008


http://www.sciencedirect.com/science/journal/01660934

http://www.elsevier.com/locate/jviromet

mailto:[email protected]


ING Model

V

2 logica

t2ewa(

ataTbbBsalpocOftse

2uwtspts2vi1olali

2

2

cp(p

2

#o0aow(S

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

ARTICLEIRMET 12170 1–8

X. Huang et al. / Journal of Viro

emperature and pH (Wright, 2009). Comparatively, using linear5 kDa polyethylenimine (PEI) as a transfection reagent is cost-ffective, has improved stability, and has the ability to function at aider pH range (Lock et al., 2010; Reed et al., 2006). The use of PEI

lso eliminates the need to change cell medium post-transfectionDurocher et al., 2007), drastically lowering cost and workload.

PEI is a polymer in which a secondary amine occupies the thirdtom per monomer unit. It is the amino group of PEI that causeshe polymer to display a high cationic charge density potentialnd buffer capacity (Boussif et al., 1995b; Clamme et al., 2003a,b).his allows the PEI to form ionic interactions with the phosphateackbones of DNA to form a condensed PEI-DNA complex that cane transported into the cell via endocytosis (Godbey et al., 1999).ehr (Behr, 1997) observed that DNA-PEI complexes escape endo-omes via a proton sponge effect that promotes osmotic swellingnd disrupting of the endosomal membrane. However, how DNAater enters the nucleus is still unknown. Within the complex, thearticular ratio between molar units of PEI nitrogen atoms to unitsf DNA phosphate atoms, known as the N/P ratio, has been found toorrelate with enhanced transfection efficiencies (Guo et al., 2012;gris et al., 1998; Reed et al., 2006). However, the N/P ratio can dif-

er between protocols, and most ratios utilize an excess of PEI forransfection. One reason that these ratios sometimes include exces-ive PEI is centered on PEI’s ability to enhance the proton spongeffect and disrupt endosomes (Boeckle et al., 2004)

In the described experiments, it was observed that using a linear5 kDa PEI to transfect HEK293T cells with a three plasmid systemsing N/P ratio of 40, with an excess of PEI, yields high titer rAAV2,ithout observable cell death after 72 h transfection. In addition

o harvesting viral particles in lysed whole cell homogenate, theupernatant was also harvested using a 40% PEG8000/2.5 N NaClrecipitation. While some consider this to be trivial to final viraliters (Reed et al., 2006; Zolotukhin et al., 1999), others have foundupernatant to contain up to 50% of total viral particles (Ayuso et al.,010) and suggest that it is a pure (no cellular debris) source of rAAVectors (Lock et al., 2010). To further isolate virions, a discontinuousodixanol gradient centrifugation followed by dialysis in an Amicon5 100,000 MWCO concentration unit was employed. Optimizationf this PEI protocol resulted in the use of less DNA, decreased work-oad, and lower cost for reagents. A simple, effective, and adaptablepproach for producing high titer recombinant AAV2 using 25 kDainear PEI is described. This further demonstrates the utility of AAV2n vitro in cell culture and in vivo in the rodent brain.

. Materials and methods

.1. AAV vectors

Three vectors were used in each experiment: (1) An AAV vectorontaining GFP or GOI (gene of interest) flanked by the ITRs; (2) aackaging vector containing the AAV serotype rep2 and cap2 genes;3) a helper vector containing the adenovirus helper functions. (Alllasmids purchased from Cell biolabs Inc.).

.2. PEI preparation and calculation

PEI (Polyethylenimine, linear, MW 25,000 from Polysciences, cat 23966) was prepared to a final concentration of 7.5 mM (basedn monomer units). pH was adjusted to 8.0 and filtered through.22 �m filter. 20 ml-aliquots were frozen and thawed three times,nd stored at −20 ◦C. The amount (�l) of PEI was calculated based


n the following equation (Reed et al., 2006): PEI (�l) = 3 × D × R/S,here D = total amount of plasmid DNA used (�g), R = N/P ratio

ratio of nitrogen content in PEI to phosphorous content in DNA), = concentration of the PEI stock (mM, monomer unit).

PRESSl Methods xxx (2013) xxx– xxx

2.3. Preparation of AAV-293T cells for transfection

HEK 293 T cells (gift from Dr. Thomas Kukar; also available fromATTC) were maintained in complete medium (4.5 g/l-Glucose andl-Glutamine containing DMEM supplemented with 10% FBS and 1%Pen-Strep) and incubated at 37 ◦C, 5% CO2. One day before trans-fection, HEK 293 T cells were seeded onto 20 150 mm plates at adensity of 1 × 107 cells per plate in 18 ml of complete medium. Thecells were approximately 70% confluent on the day of transfection.

2.4. Co-transfection of three plasmids

All plasmid concentrations were adjusted to 1 �g/ul in TE bufferbefore mixing the following: 105 �g of pAAV vector, 105 �g ofpAAV-R2C2 and 210 �g of pHelper (total of 420 �g DNA for 2015 cmplates; equal to 1.05 �g per ml of media).

The transfection mixture, which included a total of 420 �g DNA,4 ml of 1.5 M NaCl, 6.72 ml of PEI and 29.28 ml of sterile dd H2O ina 50 ml sterile tube, was vortexed a few seconds, and incubated atroom temperature for 20 min. 2 ml of the mixture was then addedto each plate in a dropwise manner. The plates were returned tothe incubator for 72 h at 37 ◦C, 5%CO2.

2.5. Virus harvest

Cell culture media and transfected cells were harvested sepa-rately. 40% PEG (cat# P2139, Sigma) in 2.5 N NaCl was added tothe supernatant to a final concentration of 8%, and incubated onice for 2 h. The cell pellet was suspended in 14 ml of lysis buffer(50 mM Tris–Cl, 150 mM NaCl and 2 mM MgCl2) and stored at 4 ◦C.Following the 2 h incubation, the supernatant was centrifuged at2500 g for 30 min at 4 ◦C to pellet the PEG-precipitated virus. Thecell lysate and pelleted supernatant precipitate were combinedand then treated with 750 �l of 10% sodium deoxycholate (cat#BP349-100, Fisher) to final concentration 0.5%, and benzonase (cat#E8263-25KU, Sigma) to a final concentration of 50 �g/ml. After30 min incubation at 37 ◦C, 3 ml of 5 M NaCl were added to thesample to decrease aggregation of the virus followed by a 30 min50 ◦C incubation, and three freeze–thaw cycles between −80 ◦C and37 ◦C. Debris was pelleted by spinning at 12,000 × g for 30mins at4 ◦C, and separated from the lysate with the use of a popper pipetteneedle attached to a 20 ml syringe. The lysate was then pipettedinto a Beckman 25 × 89 quick seal tube and immediately underlaidwith 9 ml of 15%, 6 ml of 25%, 5 ml of 40% and 5 ml of 60% iodix-anol (cat# D1556-250, sigma) using a pipetting needle attachedto a 10 ml syringe. If necessary, 1× PBS was added to fill the tubecompletely.

2.6. Centrifugation and virus retrieval

The quick seal tube containing the prepared iodixanol gradientand lysate was centrifuged at 500,000 × g using rotor 70 Ti for 1.5 hat 18 ◦C. To retrieve the virus, the tube was then secured in a clampstand set to eye-level, after which an 18 G needle was insertedinto the top of the tube to allow air to enter. Another 18 G needleattached to a 10 ml syringe was inserted just below the interface ofthe 40% and 60% iodixanol layers with the bevel of the needle up.Approximately 4 ml of sample were extracted containing the AAVin the 40% iodixanol layer.

2.7. Dialysis and concentration


The extracted virus was then added to 7 ml of PBS previouslydispensed into the top portion of an Amicon 15 100,000 MWCO(cat# UFC910024, Fisher) concentration unit followed by a 10 min

177

178

179


ARTICLE IN PRESSG Model

VIRMET 12170 1–8

X. Huang et al. / Journal of Virological Methods xxx (2013) xxx– xxx 3

iated c

ccbtbf

2

fFcUTta1bBtGsfct

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

Fig. 1. Schematic of AAV2 production using PEI – med

entrifugation at 3000 × g. When most of the solution had suffi-iently passed through the filter, 10 ml of PBS were added followedy 4 rounds of repeated washing and spinning. After the final wash,he remnant solution was spun repeatedly until 150 �l remainedefore the virus was subsequently aliquoted and stored at −80 ◦Cor future use.

Fig. 1 shows the summarizing process of AAV2 production.

.8. Determination of virus titer

The infectious units (IU) of rAAV were determined by a 10-old serial dilution. GFP expressing HEK293T cells were counted byluorescence microscopy 72 h post-infection. The vector genomeopy number (VG) was determined by qPCR using Brilliant IIIltra Fast SYBRgreen qPCR Master Mix (cat# 600882, Stratagene).he viral DNA was extracted from 1 �l of purified virus and wasreated with 0.5 U DNase I (cat# 18068-015, Invitrogen) to digestny contaminating unpackaged DNA, followed by an additional0 �g proteinase K (25530-049, Invitrogen) treatment to assist inreaking capsids and releasing viral DNA. qPCR was run in Appliediosystems Mx3000P with primers for the ITRs common to AAVransfer vector plasmids: forward primer 5′-GGA ACC CCT AGT GATGA GTT-3′ and reverse primer 5′-CGG CCT CAG TGA GCG A-3′;


et with a program: 95 ◦C 10 min, then cycled 40 times at 95 ◦Cor 15 s, 60 ◦C for 30 s and 72 ◦C for 30 s. To generate a standardurve, a pAAV-GFP plasmid was used in serial dilutions from 1 × 107

o 1 × 103 genome copies, performed in triplicate. A no-template

o-transfection with three plasmids in HEK 293T cells.

negative control was also performed in triplicate. The positivecontrol was 5 × 105 genome copies of pAAV-GFP, performed induplicate.

2.9. Cell survival assay

To quantify cell death following transfection, a tetrozolium-based technique, the WST-1 (Water soluble Tetrazolium salts) assaywas performed. HEK 293T cells were seeded into each 24-well plateat a density of 1 × 105 in 500 �l of complete medium. The cells wereapproximately 70% confluent on the day of transfection. Three plas-mids were co-transfected as described previously. 0.35 �g of totalDNA per well was used. Cell survival was ascertained for N/P ratios20, 40, 60 and 80. A blank control of media only was used whilecells were used as a negative control. 72 h after transfection, 50 �lof WST1 (cat# 5015944001, Roche) were added to each well andincubated for 30 min at 37%, 5% CO2. The absorbance was mea-sured in a microplate reader (BioTek instrument Inc.) at 450 nmand the percent of cell survival was calculated using this formula:Survival rate (%) = Asample−Ab/Ac−Ab × 100; Ab = absorbance ofblank; Ac = absorbance of negative control.

2.10. P0 and stereotaxic injections of mice


For P0 injections, newborn (P0) C57BL/6 mice were injectedwith AAV2-GFP (2 �l of 1.5 × 1012 VG/ml) in both hemispheresapproximately 1–1.5 mm deep according to published protocols (Li

225

226

227


IN PRESSG Model

V

4 logical Methods xxx (2013) xxx– xxx

atkEso1p1b(ai

2i

whp2(asra

3

3p

otiafcNhAD

bbc1o

toF

3p

f11mDo

Fig. 2. Differential effects on AAV2 titer of various N/P ratios. (A) Effect of N/P ratios10, 20, 30 and 40 on supernatant, precipitate, crude, purified mean Infectious unittiters (IU/ml) and the mean vector genome titers (VG/ml) of purified virus. The sameDNA amount of 21 �g/15 cm dish (or 1.05 �g/ml of media) was employed through-out. (B) Effect of N/P ratios 20, 40, 60 and 80 on crude and purified mean Infectiousunit and mean vector genome titers of purified virus. The same DNA amount of

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304



nd Daly, 2002; Passini et al., 2003). For stererotaxic injections,wo month old C57BL/6 mice were anesthetized with 95 mg/kgetamine and 5 mg/kg xylazene and placed in stereotaxic frame.yes were covered in ophthalmic ointment, and the scalp wasterilized before the skull was exposed. A unilateral injectionf AAV2-GFP (2 �l of 1.5 × 109 VG/ml) was given at a rate of

�l/2.5 min into the hippocampus of the right hemisphere (antero-osterior (AP), −2.1 mm from the bregma; mediolateral (ML),.8 mm from the bregma; and dorsoventral (DV), −1.8 mm fromelow the surface of the dura) according to published protocolsKiyota et al., 2009). Four weeks after P0 or stereotaxic injection,nimals were sacrificed and tissue was processed for fluorescencemmunohistochemistry as described below.

.11. Perfusion and tissue processing for fluorescencemmunohistochemistry

Four weeks post-injection, animals were deeply anesthetizedith isofluorane and intracardially perfused with 4% paraformalde-yde in PBS (pH 7.4). Brains were fixed for 48 h in the samearaformaldehyde solution and cryoprotected in 10% sucrose for4 h followed by 30% sucrose for 24 h. Serial coronal sections40 �m thick) through the cortex and cerebellum were collectednd stored in anti-freeze solution for further analyses. Tissueections were washed and mounted using a standard immunofluo-escence protocol (Pranski et al., 2012). Images were captured using

Zeiss LSM 510 laser scanning confocal microscope.

. Results

.1. Optimization of N/P ratios for PEI-mediated AAV2-GFProduction

It has been indicated that the N/P ratio is of paramount influencen the formation of PEI complexes, which in turn affects optimumransfection efficiency (Thomas et al., 2005b). We initially exam-ned N/P ratios of 10–40, while maintaining the transfected DNAmount at 21 �g/15 cm plate (Fig. 2A). We found that the lowestunctional virus titers (IU/ml) were observed for the supernatant,rude and purified virus for N/P ratio of 10 (2.5 × 108 IU/ml). The/P ratio of 10 also had the lowest vector genome titers while theighest titers corresponded to an N/P ratio of 40 (7.7 × 109 IU/ml).ll experiments were conducted in triplicate while maintaining aNA amount of 21 �g/15 cm plate (1.05 �g/ml) for each trial.

Given these experiments were conducted with ratios 40 andelow, the effect of higher N/P ratios on titer was further testedy employing a broader range of ratios under similar experimentalonditions (N/P 20, 40, 60, and 80). The lowest averaged titer of.4 × 109 IU/ml was obtained for N/P ratio 80 while the highest wasbserved for N/P ratio 40 at 4.3 × 1010 IU/ml (Fig. 2B).

These findings demonstrated that optimal IU/ml and VG/mliters were obtained for N/P ratio 40 with very little cell deathbserved by fluorescence (Fig. 3) and WST-1 assays (see below,ig. 5) 72 h post-transfection.

.2. Optimization of DNA amount for PEI-mediated AAV2-GFProduction

After optimizing the N/P ratio, comparable or higher titers wereurther sought by comparing the DNA amount of 2.25 �g/ml to.05 �g/ml. For this experiment, DNA amounts of 0.525 �g/ml,


.05 �g/ml and 2.25 �g/ml were compared. A ratio of N/P 40 wasaintained for all trials. The highest titers were observed withNA 1.05 �g/ml of total plasmid DNA with functional titers justver 109 IU/ml which corresponds to a genome copy titer in excess

21 �g/15 cm dish (or 1.05 �g/ml of media) was employed throughout. Crude andpurified infectious unit titers were performed in triplicate, whereas genomic titers(VG/ml) was performed once per group. Bars show mean value ± sem.

of 1012 VG/ml as determined by qPCR. An N/P ratio of 40 cou-pled with DNA amount of 1.05 �g of total plasmid DNA per ml ofculture media was subsequently employed for all other AAV2 pro-ductions (5.2 �g of PEI per �g of DNA). These titers ranged from108–6 × 109 IU/ml and 4 × 1010–6 × 1013 VG/ml.

A total of 17 complete viral productions were thenexamined across numerous different packaging plasmids.The 1.05 �g/ml/40 N/P ratio (9 productions, mean titer2.0 × 1013 VG/ml) consistently produced superior titers tothe 2.25 �g/ml/27.5 N/P ratio (8 productions, mean titer5.8 × 1012 VG/ml) combination, (Fig. 4A, p < 0.05). The utilityand reproducibility of this particular N/P ratio and DNA amountfor the production of rAAV2 with diverse transgene sizes andstructures was thus further demonstrated.

3.3. Influence of the supernatant on final AAV2 vector titers

Current methodology for AAV production and harvestingfocuses on the heparin-binding mechanism of the virus to cells, andhence the commonly held view that most of the virus (80–90%) isstrongly cell-associated (Reed et al., 2006; Zolotukhin et al., 1999).


However, it has been reported that AAV heparin affinity is serotypespecific (Guo et al., 2012), and thus it was important to deter-mine whether harvesting the supernatant along with the lysate ofHEK293T cells affected overall titer for serotype 2. A 60 fold increase

305

306

307

308


ARTICLE IN PRESSG Model

VIRMET 12170 1–8

X. Huang et al. / Journal of Virological Methods xxx (2013) xxx– xxx 5

F ells. (Ap filter.A

iwwsnct

3

favci2rp4

Fsoastw

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

ig. 3. Efficacy and viability of AAV2 transfection and infection in vitro in 293T cAAV-R2C2, showing high efficiency. (B) Same as figure A in brightfield, without GFPAV2-GFP virus, MOI = 3.

n purified virus was observed when the supernatant was treatedith 40%PEG8000/2.5 N NaCl and added to the cell lysate (Titersere 1.4 × 107 IU/ml without supernatant and 8.3 × 108 IU/ml with

upernatant). The data also showed that virus released to the super-atant constitutes 43–62% of the total AAV2-containing clarifiedell homogenate (Fig. 4B). Thus, for all subsequent experiments,he supernatant was harvested and treated according to protocol.

.4. Effect of N/P ratio on WST-1 based cell survival

As a secondary means of testing post-transfection cell survivalor various N/P ratios, the WST-1 assay was employed. The WST-1ssay provided a sensitive and accurate method of measuring celliability and proliferation by the reduction of tetrazolium salts toolored formazan compounds by succinate-tetrazolium reductasen live cells. N/P ratios 20, 40, 60 and 80 were used in this assay. N/P


0 and N/P 40 had the highest percent cell survival (76% and 74%,espectively, which were not significantly different). The lowestercent survival was obtained for N/P ratios 60 and 80 at 65% and1%, respectively (Fig. 5).

ig. 4. Differential effects on AAV2 titer of DNA amounts and N/P ratios and inclu-ion of Cellular Supernatant in Virus Production. (A) Comparison of mean VG titersf DNA amount: 2.25 �g DNA/ml coupled with N/P ratio 27.5 and optimized DNAmount of 1.05 �g DNA/ml coupled with N/P ratio 40. The 1.05 �g/40 N/P ratio wasignificantly more effective at producing high-titer virus (p < 0.05). (B) Comparingiters of crude and purified virus harvested with and without supernatant (sup virusas precipitated by 40% PEG/2.5 N NaCl). Bars show mean value ± sem.

342

343

344

345

346

347

348

) 72-hours post-transfection using protocol above with pHelper, pAAV-GFP and Approximately 20 percent cell death observed. (C) 293T cells infected with purified

3.5. Variability of the RT-PCR method for quantification of AAV2vector genome particles

Infectious unit and genome copy viral titers were respectivelyascertained by serial dilution replication assays and real-time PCR.Quantification of rAAV2 particles by qPCR has been established asone of the most practical and time-efficient methods for obtainingvector genome copy titers (Rohr et al., 2002). However, the limita-tions of standardizing this procedure as a principal means of viralparticle quantification was also recognized as there yet remained asignificant degree of inter-experimental variability. Results showedratios of the genomic copies (VG) to infectious rAAV2 particles (IU)ranging from 333 (VG:IU) to 6000 (VG:IU), with a mean ratio of1138 (data not shown). Rohr and others quantified the mean ratioof the same sample as 253. Still others have reported ratios rangingfrom 60 to 120 (Atkinson et al., 1998) and mean ratios as low as 8.3(VG:IU) (Zhen et al., 2004). This discrepancy was attributed to thedifference in assay sensitivity, which may result in underestimatedinfectious unit titers and subsequently overestimated VG:IU ratios.Possible high multiplicity of infection of a single cell as well as thepresence of infectious-defective particles in the rAAV2 sample pop-ulation has been another school of thought (Rohr et al., 2002). Forthe experiments discussed herein, the high sensitivity of the qPCR


method and failure to achieve identical experimental conditionsconsistently may have further confounded titers. For this reason, itwas posited that the functional titers (IU) provided a more reliable

Fig. 5. WST1 based cell survival. Comparison of N/P ratios 20, 40, 60 and 80 post-transfection cell survival using WST1 assay. No significant differences observed forN/P 20 and 40. Significant loss in cell viability observed for N/P 60 and 80 bars showmean% survival ± sem.

349

350

351


ING Model

V

6 logica

rcc

3a

titflptfltsAerbfld

4

cAtpvppnp

Fqro

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415



epresentation of the presence of infective competent viral parti-les in a given rAAV sample albeit the number of genome copiesould not have been known by this method alone.

.6. P0 injection of neonatal mice and stereotaxic injection ofdult mouse brain

Injection of AAV2 in neonatal (P0) murine pups as well as stereo-axic injection of adult mice was employed to test virus efficacyn vivo. Brains of P0 pups were injected with AAV2 harboringhe chicken �-actin (CBA) promoter to drive expression of greenuorescent protein (GFP, 2 �l of 1.5 × 1012 VG/ml). Four weeksost-injection, brains were harvested and prepared for immunohis-ochemistry. Free-floating tissue sections revealed extensive GFPuorescence in nearly all brain regions, including but not limitedo, the cortex, hippocampus, and cerebellum (Fig. 6A). In parallel,tereotaxic injection of 8 week old mice was performed to targetAV2 to the hippocampus. A single stereotaxic injection of AAV2xpressing GFP (2 �l of 1.5 × 109 VG/ml) was delivered into theight hemisphere targeting the dentate gyrus. Four weeks later,rains were sectioned for analysis by confocal microscopy. Free-oating tissue samples revealed intense GFP fluorescence in theentate gyrus and other hippocampal areas in all mice (Fig. 6B).

. Discussion

In previously described protocols, adeno-associated virus washiefly produced via co-infection with another helper virus such asdenovirus or herpes-virus for effective viral replication. Unfor-

unately, with the use of a helper virus, there yet remained aossibility of contamination from the helper viruses in the finalector preparation. In the current experiments, a helper-free triple-


lasmid transfection protocol was used, conferring greater purity,otency and safety (Wright, 2009) in the production of recombi-ant AAV. Of the three plasmids used for transfection, pHelperlasmid contained the required E2A, E4, and VA RNA adenoviral

ig. 6. Efficacy of AAV2 in vivo across different brain areas. To test efficacy of AAV2 in vualitatively evaluated four weeks post-injection. (A) Representative images of P0 injectegions, including (left to right) cortex, hippocampus, and cerebellum. Scale bars = 100 �Mbserved in all hippocampal regions. Image on right is more posterior. Scale bars = 500 �M


genes, which eliminated the need for a helper adenovirus (Xiaoet al., 1998)(See Fig. 1).

A variety of approaches have been reported for the productionof rAAV using triple plasmid co-transfection. A co-precipitationmethod utilizing Calcium phosphate as the transfection reagenthas long been described and was established as early as 1973(Graham and Vandereb, 1973) as an efficient means of transferringexogenous DNA into mammalian cells. Under carefully stringentconditions, yields exceeding 105 VG/cell (Wright, 2009) have beenobtained and consistently reproduced. However, failure to achieveand maintain highly specific transfection cocktail conditions with avery narrow window for pH variability, results in appreciably lowerproductivity overall (Wright, 2009). With such exacting conditionsas pH (7.05 ± 0.01units), reagent concentrations, precipitate forma-tion size, temperature, and the requirement for post-transfectionmedia change, the calcium phosphate-mediated production of AAVproves time-consuming, laborious and ultimately more costly. Stillothers have employed cationic lipids such as lipofectamine (Reedet al., 2006), which can be extremely efficient but remain inhibitoryto recurrent, practical use due to high costs.

The advent of the polycation PEI has proven to be very promis-ing, practical and cost-efficient (127X less than CaCl2 for reagentcost only) for the production of rAAV. Titers up to ∼1014 VG per 3.5 lbioreactor have been reported using PEI-mediated transfection ofsuspension HEK-293 cells under serum-free conditions, and with-out the need for exchange of the cell culture medium (Durocheret al., 2007). While this approach appears quite reasonable for thelarge-scale, high-titer production required for clinical grade cGMPrAAV, it may not be as pragmatic for the pre-clinical research-grade, smaller-scale productions of only 400 ml starting volumeand concentrated volume of 130-140 �l as employed in this study.By optimizing current PEI-mediated protocols for AAV2 production,


it was demonstrated that high-titer, research-grade virus in excessof 1013 VG/ml was achieved with feasible starting volumes and lessDNA amounts. Additionally, this PEI protocol for adherent cells wasuninhibited by serum. AAV serotype 2 has efficiently transduced

ivo, P0 or stereotaxic injection of mice was performed and expression of GFP wasion of AAV2. Extensive spread of GFP fluorescence was observed in multiple brain. (B) Representative images of stereotaxic injection of AAV2. GFP fluorescence was. Data representative of three mice.

416

417

418

419


ING Model

V

logica

b(ata

cn2(wt8NnddtttaaRaao

gteeccmabPpm

obtpfyvam

sDcPQ3awPawse

1ntH

420

421

422

423

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

474

475

476

477

478

479

480

481

482

483

484

485

486

487

488

489

490

491

492

493

494

495

496

497

498

499

500

501

502

503

504

505

506

507

508

509

510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

525

526

527

528

529

530

531

532

533

534

535

536

537

538

539

540

541

542

543

544

545

546

547

548

549



rain, bone, kidney, lung, liver and, numerous other tissue typesGrimm et al., 2003). All reported viral stocks were near-purifiednd infected with high transduction efficiency, displaying broadropism both in vivo in rodent brain (Fig. 6) and, in vitro cell culturessays (see Fig. 3).

Consistent with prior reports, PEI-mediated transfection effi-iency and AAV vector assembly varied according to the ratio ofitrogen to phosphorus in the PEI-DNA polyplexes (Zhang et al.,004). While the lowest efficiency (5–20%) and crude virus titer2.3 × 106 IU/ml) were observed with N/P = 10, the titer increasedith higher ratios in a tenable manner, with N/P = 40 producing

he highest crude titers of up to 3 × 108 IU/ml (corresponding to.5 × 1012 VG/ml). To investigate whether further increasing the/P ratios would yield correspondingly higher titers, with little too exacerbated cell death, functional and vector genome titers wereetermined for N/P ratios 20, 40, 60, and 80 using the previouslyescribed protocol. An approximate ten to thirty-fold reduction initer was observed for N/P ratios 60 and 80, which could be par-ially attributed to the post-transfection cytotoxicity observed forhese ratios (Fig. 5). These findings suggested that there might be

threshold for the N/P ratio beyond which PEI becomes unfavor-bly toxic to the cell, resulting in inefficient viral vector assembly.educed viral assembly and thus reduced titers for N/P ratios 60nd 80 may have also been due to the exclusion of excessively largeggregates from endocytotic pathways as reported by Boussif andthers (Boussif et al., 1995a).

Although PEI has been proven to be one of the most efficientene delivery vectors, its efficiency and cytoxicity has been thoughto be dose dependent (Thomas et al., 2005a). In this study, wemployed a WST-1 based assay to measure cell viability and prolif-ration in live cells for a relatively broad range of N/P ratios. In thisase, as the N/P ratio increased, cell viability decreased (Fig. 5). Fis-her and others have reported that increasing cytotoxicity of higholecular weight PEI such as 25 kD could be caused by the high

ffinity binding of the PEI to the outer surface of the plasma mem-rane. This particular study has demonstrated that, as the overallEI concentration increased above N/P 40, a commensurate cyto-athic effect was observed morphologically by fluorescence and,etabolically, by WST-1 assays.For approximately 1/5th of AAV2 batches, a total DNA amount

f 2.25 �g per ml of media and a ratio of N/P = 27.5 were employedefore the attempt was first made to optimize. To further inves-igate the utility of an optimal N/P ratio of 40, experiments wereerformed to test whether approximately halving the DNA amountrom 2.25 �g/ml to 1.05 �g/ml while maintaining this ratio wouldield similar or disparate results. An almost 7-fold increase in crudeirus titers was observed when the DNA amount was halved, whilen approximate 75% reduction in the DNA amount (0.525 �g/ml ofedia) was analogous to a 10-fold decrease in productivity.These results were congruent with previous findings, demon-

trating that AAV2 triple plasmid transfection with the optimalNA amount and N/P ratio (1.05 �g/ml and 40 respectively), cru-ially affected the size, weight and dynamic geometry of theEI-DNA complex formation (Hou et al., 2011) and ultimately, over-ll efficiency. In this case, reasonably decreasing the DNA amounthile maintaining a high N/P ratio, increased the amount of excess

EI relative to DNA in solution (5.2 �g PEI: 1 �g DNA), which in turnccelerated both the settling of the complexes on the cell surface asell as the escape of the DNA from the endosomes via the proton

ponge effect (Behr, 1997) as soon as 4 h post-transfection (Reedt al., 2006).

Based on findings in the literature, it was anticipated that


0–20% of total purified virus would be recovered from the super-atant (Zolotukhin et al., 1999). Reed and colleagues also reportedhat an approximate 73% of total virus was found in cell lysates ofeLa and XDC293 cells and hence, did not deem the supernatant

PRESSl Methods xxx (2013) xxx– xxx 7

harvest a worthwhile endeavor (Reed et al., 2006). Surprisinglyhowever, this study has shown that an average of 43–62% of thetotal virus was contributed by the media harvest for AAV2, ayield well worth the effort of harvesting (Fig. 4B). While this phe-nomenon remains incompletely understood, it could be partiallyattributed to a possible active virion export pathway (Guo et al.,2012), cell lineage, as well as incubation time post-transfection.The latter is consistent with data revealing that culture media titerswere higher with increased incubation time post transfection, withas much as 86% of virus in the supernatant being reported for variedserotypes (Lock et al., 2010; Vandenberghe et al., 2010).

In conclusion, the importance of DNA amount and N/P ratio inoptimization of the PEI-mediated AAV2 vector production protocolwas elucidated. This optimized protocol for increasing overall titersby harvesting the supernatant along with crude cell lysate, was alsodemonstrated as being efficacious. With reasonable starting vol-umes, less plasmid DNA amounts, and by eliminating the need formedia change, overall production cost and workload were loweredsubstantially compared to Calcium phosphate-based protocols. Thedescribed method was simple, highly reproducible, and extremelyefficient in yielding high-titer functional AAV2 vector particles. Itis expected that the utility of this protocol proves applicable to allAAV serotypes with relevance for the production of other viral vec-tors such as lentivirus. Thus far, this protocol has been successfullyimplemented in this laboratory for serotypes 1, 5, and 9. Futureoptimization approaches should focus on the influence of serumon vector productivity as well as the role of differential effects ofmedia choice on PEI-mediated transfection efficiency.

Acknowledgements

We would like to thank Dr.Thomas Kukar for gifting theHEK293T cells and Dr. Nancy Ciliax for administrative support. Thisresearch project was supported by Emory Neuroscience NINDS CoreFacility grant for the Viral Vector Core, P30NS055077.

References

Atkinson, E.M., Debelak, D.J., Hart, L.A., Reynolds, T.C., 1998. A high-throughputhybridization method for titer determination of viruses and gene therapy vec-tors. Nucl. Acids Res. 26, 2821–2823.

Ayuso, E., Mingozzi, F., Montane, J., Leon, X., Anguela, X.M., Haurigot, V., Edmonson,S.A., Africa, L., Zhou, S., High, K.A., Bosch, F., Wright, J.F., 2010. High AAV vectorpurity results in serotype- and tissue-independent enhancement of transduc-tion efficiency. Gene Ther. 17, 503–510.

Bainbridge, J.W., Smith, A.J., Barker, S.S., Robbie, S., Henderson, R., Balaggan, K.,Viswanathan, A., Holder, G.E., Stockman, A., Tyler, N., Petersen-Jones, S., Bhat-tacharya, S.S., Thrasher, A.J., Fitzke, F.W., Carter, B.J., Rubin, G.S., Moore, A.T.,Ali, R.R., 2008. Effect of gene therapy on visual function in Leber’s congenitalamaurosis. N. Engl. J. Med. 358, 2231–2239.

Behr, J.P., 1997. The proton sponge: a trick to enter cells the viruses did not exploit.Chimia 51, 34–36.

Boeckle, S., von Gersdorff, K., van der Piepen, S., Culmsee, C., Wagner, E., Ogris, M.,2004. Purification of polyethylenimine polyplexes highlights the role of freepolycations in gene transfer. J. Gene Med. 6, 1102–1111.

Boussif, O., Lezoualc’h, F., Zanta, M.A., Mergny, M.D., Scherman, D., Demeneix, B.,Behr, J.P., 1995a. A versatile vector for gene and oligonucleotide transfer intocells in culture and in vivo: polyethylenimine. Proc. Natl. Acad. Sci. U.S.A. 92,7297–7301.

Boussif, O., Lezoualch, F., Zanta, M.A., Mergny, M.D., Scherman, D., Demeneix, B.,Behr, J.P., 1995b. A versatile vector for gene and oligonucleotide transfer intocells in culture and in vivo – polyethylenimine. Proc. Natl. Acad. Sci. U.S.A. 92,7297–7301.

Bowles, D.E., McPhee, S.W.J., Li, C.W., Gray, S.J., Samulski, J.J., Camp, A.S., Li, J., Wang,B., Monahan, P.E., Rabinowitz, J.E., Grieger, J.C., Govindasamy, L., Agbandje-McKenna, M., Xiao, X., Samulski, R.J., 2012. Phase 1 gene therapy for duchennemuscular dystrophy using a translational optimized AAV vector. Mol. Ther. 20,443–455.


Clamme, J.P., Azoulay, J., Mely, Y., 2003a. Monitoring of the formation and dis-sociation of polyethylenimine/DNA complexes by two photon fluorescencecorrelation spectroscopy. Biophys. J. 84, 1960–1968.

Clamme, J.P., Krishnamoorthy, G., Mely, Y., 2003b. Intracellular dynamics of thegene delivery vehicle polyethylenimine during transfection: investigation by

550

551

552

553

554


http://refhub.elsevier.com/S0166-0934(13)00221-8/sbref0005



















































































































































































































ING Model

V

8 logica

D

G

G

G

G

H

J

K

K

L

L

N

infectious events. Hum. Gene Ther. 15, 709–715.

555

556

557

558

559

560

561

562

563

564

565

566

567

568

569

570

571

572

573

574

575

576

577

578

579

580

581

582

583

584

585

586

587

588

589

590

591

592

593

594

595

596

597

598

599

600

601

602

603

604

605

606

607

608

609

610

611

612

613

614

615

616

617

618

619

620

621

622

623

624

625

626

627

628

629

630



two-photon fluorescence correlation spectroscopy. Biochim. Et Biophy. ActaBiomembr. 1617, 52–61.

urocher, Y., Pham, P.L., St-Laurent, G., Jacob, D., Cass, B., Chahal, P., Lau, C.J., Nal-bantoglu, J., Kamen, A., 2007. Scalable serum-free production of recombinantadeno-associated virus type 2 by transfection of 293 suspension cells. J. Virol.Methods 144, 32–40.

odbey, W.T., Wu, K.K., Mikos, A.G., 1999. Tracking the intracellular path ofpoly(ethylenimine)/DNA complexes for gene delivery. Proc. Natl. Acad. Sci. U.S.A.96, 5177–5181.

raham, F.L., Vandereb, A.J., 1973. New technique for assay of infectivity of humanadenovirus 5 DNA. Virology 52, 456–467.

rimm, D., Kay, M.A., Kleinschmidt, J.A., 2003. Helper virus-free, optically control-lable, and two-plasmid-based production of adeno-associated virus vectors ofserotypes 1–6. Mol. Ther. 7, 839–850.

uo, P., El-Gohary, Y., Prasadan, K., Shiota, C., Xiao, X.W., Wiersch, J., Paredes, J.,Tulachan, S., Gittes, G.K., 2012. Rapid and simplified purification of recombinantadeno-associated virus. J. Virol. Methods 183, 139–146.

oggan, M.D., Blacklow, N.R., Rowe, W.P., 1966. Studies of small DNA viruses foundin various adenovirus preparations – physical biological and immunologicalcharacteristics. Proc. Natl. Acad. Sci. U.S.A. 55, 1467.

ayandharan, G.R., Aslanidi, G., Martino, A.T., Jahn, S.C., Perrin, G.Q., Herzog, R.W.,Srivastava, A., 2011. Activation of the NF-kappaB pathway by adeno-associatedvirus (AAV) vectors and its implications in immune response and gene therapy.Proc. Natl. Acad. Sci. U.S.A 108, 3743–3748.

aplitt, M.G., Feigin, A., Tang, C., Fitzsimons, H.L., Mattis, P., Lawlor, P.A., Bland, R.J.,Young, D., Strybing, K., Eidelberg, D., During, M.J., 2007. Safety and tolerabil-ity of gene therapy with an adeno-associated virus (AAV) borne GAD gene forParkinson’s disease: an open label, phase I trial. Lancet 369, 2097–2105.

iyota, T., Yamamoto, M., Schroder, B., Jacobsen, M.T., Swan, R.J., Lambert, M.P.,Klein, W.L., Gendelman, H.E., Ransohoff, R.M., Ikezu, T., 2009. AAV1/2-mediatedCNS gene delivery of dominant-negative CCL2 mutant suppresses gliosis, beta-amyloidosis, and learning impairment of APP/PS1 mice. Mol. Ther. 17, 803–809.

i, J., Daly, T.M., 2002. Adeno-associated virus-mediated gene transfer to the neona-tal brain. Methods 28, 203–207.

ock, M., Alvira, M., Vandenberghe, L.H., Samanta, A., Toelen, J., Debyser, Z., Wilson,J.M., 2010. Rapid, simple, and versatile manufacturing of recombinant adeno-


associated viral vectors at scale. Hum. Gene Ther. 21, 1259–1271.iemeyer, G.P., Herzog, R.W., Mount, J., Arruda, V.R., Tillson, D.M., Hathcock, J., van

Ginkel, F.W., High, K.A., Lothrop Jr., C.D., 2009. Long-term correction of inhibitor-prone hemophilia B dogs treated with liver-directed AAV2-mediated factor IXgene therapy. Blood 113, 797–806.


Ogris, M., Steinlein, P., Kursa, M., Mechtler, K., Kircheis, R., Wagner, E., 1998. The sizeof DNA/transferrin-PEI complexes is an important factor for gene expression incultured cells. Gene Ther. 5, 1425–1433.

Passini, M.A., Watson, D.J., Vite, C.H., Landsburg, D.J., Feigenbaum, A.L., Wolfe, J.H.,2003. Intraventricular brain injection of adeno-associated virus type 1 (AAV1)in neonatal mice results in complementary patterns of neuronal transductionto AAV2 and total long-term correction of storage lesions in the brains of beta-glucuronidase-deficient mice. J. Virol. 77, 7034–7040.

Pranski, E.L., Van Sanford, C.D., Dalal, N.V., Orr, A.L., Karmali, D., Cooper, D.S., Costa,N., Heilman, C.J., Gearing, M., Lah, J.J., Levey, A.I., Betarbet, R.S., 2012. Compar-ative distribution of protein components of the A20 ubiquitin-editing complexin normal human brain. Neurosci. Lett. 520, 104–109.

Reed, S.E., Staley, E.M., Mayginnes, J.P., Pintel, D.J., Tullis, G.E., 2006. Transfection ofmammalian cells using linear polyethylenimine is a simple and effective meansof producing recombinant adeno-associated virus vectors. J. Virol. Methods 138,85–98.

Rohr, U.P., Wulf, M.A., Stahn, S., Steidl, U., Haas, R., Kronenwett, R., 2002. Fast and reli-able titration of recombinant adeno-associated virus type-2 using quantitativereal-time PCR. J. Virol. Methods 106, 81–88.

Thomas, M., Ge, Q., Lu, J.J., Chen, J., Klibanov, A.M., 2005a. Cross-linked smallpolyethylenimines: while still nontoxic, deliver DNA efficiently to mammaliancells in vitro and in vivo. Pharm. Res. 22, 373–380.

Thomas, M., Ge, Q., Lu, J.J., Chen, J.Z., Klibanov, A.M., 2005b. Cross-linked smallpolyethylenimines: while still nontoxic, deliver DNA efficiently to mammaliancells in vitro and in vivo. Pharm. Res. 22, 373–380.

Vandenberghe, L.H., Xiao, R., Lock, M., Lin, J.P., Korn, M., Wilson, J.M., 2010. Efficientserotype-dependent release of functional vector into the culture medium duringadeno-associated virus manufacturing. Hum. Gene Ther. 21, 1251–1257.

Wright, J.F., 2009. Transient transfection methods for clinical adeno-associated viralvector production. Hum. Gene Ther. 20, 698–706.

Xiao, X., Li, J., Samulski, R.J., 1998. Production of high-titer recombinant adeno-associated virus vectors in the absence of helper adenovirus. J. Virol. 72,2224–2232.

Zhen, Z., Espinoza, Y., Bleu, T., Sommer, J.M., Wright, J.F., 2004. Infectious titer assayfor adeno-associated virus vectors with sensitivity sufficient to detect single


Zolotukhin, S., Byrne, B.J., Mason, E., Zolotukhin, I., Potter, M., Chesnut, K., Summer-ford, C., Samulski, R.J., Muzyczka, N., 1999. Recombinant adeno-associated viruspurification using novel methods improves infectious titer and yield. Gene Ther.6, 973–985.

631

632

633

634















































































































































































































































































































































































































































































































































































Documents

AAV2 production with optimized N/P ratio and PEI-mediated transfection results in low toxicity and high titer for in vitro and in vivo applications