1

2

3

4

5

6

7

8

9

10

111213141516171819202122232425

2627

28

Journal of Immunological Methods xx (2006) xxx–xxx

+ MODEL

www.elsevier.com/locate/jim

ARTICLE IN PRESS

OOF

Research paper

Effective gene suppression using small interfering RNAin hard-to-transfect human T cells

Jiyi Yin a, Zhengyu Ma a,b, Nithianandan Selliah a, Debra K. Shivers a,Randy Q. Cron a,b,c, Terri H. Finkel a,b,c,⁎

a Division of Rheumatology, The Children's Hospital of Philadelphia, Philadelphia, PA 19104, United Statesb Immunology Graduate Program, University of Pennsylvania School of Medicine, Philadelphia, PA 19014, United States

c Department of Pediatrics, University of Pennsylvania School of Medicine, Philadelphia, PA 19014, United States

Received 3 June 2005; received in revised form 18 January 2006; accepted 20 January 2006

R

RECT

EDP

Abstract

RNA interference (RNAi) is an evolutionarily conserved cellular defense mechanism that protects cells from hostile genes andregulates the function of normal genes during growth and development. In this study, we established proof of principle of smallinterfering RNA (siRNA) silencing in hard-to-transfect human T cell lines and primary human CD4 T cells. We used public and in-house programs to design four siRNAs each for GFP, for our novel cellular gene HALP, and for their corresponding scrambledsiRNA controls. We generated siRNA expression cassettes (SECs) by PCR and directly transfected the PCR products into T cellsusing amaxa® Nucleofector technology. The most effective SECs were selected and cloned into a TA cloning vector and titeredwith their respective controls to increase transfection efficiency. Flow cytometry and fluorescence microscopy analyses wereperformed for GFP siRNAs, and Northern blot analysis was done to assess the HALP silencing effect. These experimentsdemonstrate that SECs are an excellent screening tool to identify siRNA sequences effective in silencing expression of genes ofinterest. The vector expressing the most effective siRNA robustly inhibited GFP expression (up to 92%) in the context of co-transfection in human T cell lines and primary CD4 T cells. The optimized siRNA for our endogenous cellular gene HALP alsosilenced its target RNA expression by more than 90%. These studies demonstrate that the combination of SEC, siRNA expressionvectors and Nucleofector technology can be successfully applied to hard-to-transfect human T cell lines and primary T cells toeffectively silence genes.© 2006 Published by Elsevier B.V.

R OKeywords: Human; CD4-positive T-lymphocytes; Cell line; Small interfering RNA; Transfection

UNC

JIM-10293; No of Pages 11

Abbreviations: RNAi, RNA interference; siRNA, small interfering RNA; SECs, siRNA expression cassettes; dsRNA, double stranded RNA;GFP, green fluorescent protein; PCR, polymerase chain reaction; FCS, fetal calf serum; HALP, “HIV-1 Associated Life Preserver”; PHA,phytohemagglutinin; BLAST, basic local alignment search tool; MFI, mean fluorescence intensity; GAPDH, glyceraldehyde-3-phosphatedehydrogenase; h, hour.⁎ Corresponding author. Abramson Research Center, Rm. 1102, Division of Rheumatology, The Children's Hospital of Philadelphia, 34th and

Civic Center Blvd., Philadelphia, PA 19104, United States. Tel.: +1 215 590 7180; fax: +1 215 590 1258.E-mail address: finkelt@email.chop.edu (T.H. Finkel).

0022-1759/$ - see front matter © 2006 Published by Elsevier B.V.doi:10.1016/j.jim.2006.01.023

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

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

2 J. Yin et al. / Journal of Immunological Methods xx (2006) xxx–xxx

ARTICLE IN PRESS

UNCO

RREC

1. Introduction

The phenomenon of gene silencing by RNAinterference (RNAi) is central to a number of naturalRNA-based silencing processes, and is becoming avaluable tool in functional genomics, target validation,and gene-specific therapeutic research. Introduction ofRNAi into different organisms induces both strong andspecific post-transcriptional gene silencing by directlymediating degradation of homologous target RNAs(Fire, 1999; Sharp, 2001; Tuschl, 2002; Scherer andRossi, 2003). The mediators of sequence-specificmRNA degradation are 21 and 22-nucleotide smallinterfering RNAs cleaved by ribonuclease III fromlonger dsRNAs (Hamilton and Baulcombe, 1999;Hammond et al., 2000; Zamore et al., 2000; Bernsteinet al., 2001; Elbashir et al., 2001b). In the mammaliansystem, long dsRNAs (usually longer than 30 bp) trig-ger the interferon pathway and lead to non-specificRNA degradation, while shorter siRNAs (21 bp) exo-genously introduced into mammalian cells specificallysuppress expression of endogenous and heterologousgenes (Elbashir et al., 2001a). Successful experimentsusing RNAi are largely dependent upon three factors:siRNA design, efficient delivery and cell type. To date,the majority of published studies have used adherentcells as the targeted cell population for siRNA expres-sion (Elbashir et al., 2001a; Lee et al., 2002; Miyagishiand Taira, 2002; Paul et al., 2002).

Methods for gene delivery into eukaryotic cells fallinto two main categories: viral gene transfer (by, forexample, retroviral or adenoviral vectors), and non-viral gene transfer. Viral gene transfer increases trans-duction efficiency, but the processes of virus packagingand transduction are time-consuming and safety con-cerns exist. Most non-viral gene transfer methods havelow transfection efficiencies in both adherent and sus-pension cells (Cron et al., 1997). We have optimized anovel transfection technology called Nucleofection,which has recently become commercially available(Cron, 2003). We transfected freshly isolated primaryhuman CD4 T cells with a GFP expression vector(pEGFP-N1). Flow cytometric analyses of GFP ex-pression showed minimal cell death and high transfec-tion frequencies of 60–70% at 6 h post-transfection.Furthermore, more than 25% of cells still expressedGFP after 7 days in culture with IL-2 alone (Cron,2003). This technique made feasible the transfection ofhard-to-transfect primary and suspension cells forfunctional analyses.

In this report, utilizing Nucleofector technologywith siRNA expression cassettes or siRNA expression

TEDPR

OOF

vectors, we optimized transfection programs for ourspecific cell lines, and achieved high efficiencies oftransfection and gene silencing in T cell lines andprimary human resting and activated T cells.

2. Materials and methods

2.1. Cell culture

The CD4 T cell line CEM-SS was obtained from therepository of the NIH AIDS Research and ReferenceReagent Program (from Dr. Peter Nara) and wasmaintained in RPMI 1640 medium (Gibco-BRL,Gaithersburg, MD) supplemented with 10% fetal calfserum (FCS), 2 mM L-glutamine, 100 U/ml penicillinand 100 μg/ml streptomycin.

2.2. Primary human CD4 T cell isolation and activation

CD4 T cells were isolated from blood of healthyadult human donors using the RosetteSep™ CD4+ TCell Enrichment Kit (StemCell Technologies, Vancou-ver, BC, Canada), as previously described (Cron,2001). Purified cells were phenotyped as 90–95%CD4+ T cells by flow cytometric analysis (data notshown). The CD4 T cells (1.5×106 cells/ml) were activated by phytohemagglutinin (PHA, M Form, Invi-trogen, Carlsbad, CA) for 72 h in RPMI 1640 medium(Gibco-BRL, Gaithersburg, MD) supplemented with10% FCS, 2 mM L-glutamine, 100 U/ml penicillin and100 μg/ml streptomycin at 37 °C in 5% CO2. The finalconcentration of PHA was 1 ml per 100 ml of culturemedium.

2.3. Cloning the human U6 promoter

CEM-SS cells were collected and genomic DNAwasisolated using the DNeasy tissue kit (Qiagen, Valencia,CA). The human U6 promoter sequence was amplifiedusing genomic DNA as template. Primers used in thePCR reaction flanked the U6 sequence. The 5′ primer (5′U6 universal primer) was: CGGAATTCCCCCAGTGGAAAGACGCG CAG; the 3′ primer was: CGGTGTTTCGTCCTTTCCACAAG. PCR primers weredesigned based upon the human small nuclear RNAgene sequence (GenBank Accession No: M14486). PCRreactions were carried out as follows: 30 s at 94 °C,followed by 35 cycles of 30 s at 94 °C, 30 s at 58 °C and1 min at 72 °C, then extended at 72 °C for 5 min. PCRproducts were cloned into a TA cloning vector (Invitro-gen) and the U6 sequence was confirmed by DNAsequencing.

TEDPR

OOF

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

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

Table 1 t1:1

GFP and HALP siRNA sequences and their random scrambledcounterparts t1:2

t1:3GFP siRNA sequencet1:4GFP1: GCAAGCTGACCCTGAAGTTCATa

t1:5GFP2: AAGGACGACGGCAACTACAAGACt1:6GFP3: AAGCTGGAGTACAACTACAACAGt1:7GFP4: AAGATCCGCCACAACATCGAGGAt1:8

t1:9GFP synthetic siRNA sequence (sense)t1:10GFP4: AAGAUCCGCCACAACAUCGAGGAt1:11

t1:12GFP scrambled controlt1:13GFP-SC1: GACCTATCGTGCAAATAGCCTGt1:14GFP-SC2: AAGCGACAGAGAGCTAAGCACACt1:15GFP-SC3: AAGAATATAAAGCACCGTCAGGCt1:16GFP-SC4: AAGCCGACGAAAGTACTCCAGACt1:17

t1:18GFP synthetic scrambled RNA sequence (sense)t1:19GFP4: AAGCCGACGAAAGUACUCCAGACt1:20

t1:21HALP siRNA sequencet1:22HALP1: GGA AGA CAC GGC TTA CCT GGAt1:23HALP2: GAT GCC TAG CCA GTT GGT AAGt1:24HALP3: GCC TAG CCA GTT GGT AAG Ct1:25HALP4: AAG ACA CGG CTT ACC TGG ATGt1:26

t1:27HALP scrambled controlt1:28HALP-SC1: GCG ATT GCG CCA GAG ACATAGt1:29HALP-SC2: GAT TCG GTG ATA CGATCG CGAt1:30HALP-SC3: GGC CGG TTT GAA GCA CTA Ct1:31HALP-SC4: AAG CTC TCG ATATGC AGA GCG

a This GFP siRNA sequence was taken from Caplen et al. (2001).Specific inhibition of gene expression by small double-stranded RNAsin invertebrate and vertebrate systems. Proc. Natl. Acad. Sci. U.S.A.98, 9742. t1:32

3J. Yin et al. / Journal of Immunological Methods xx (2006) xxx–xxx

ARTICLE IN PRESS

UNCO

RREC

2.4. Design of siRNA and construction of siRNAexpression cassettes (SECs)

The selection of siRNA sequences was based uponweb-based programs (web pages: http://www.dharmacon.com/ and www.ambion.com/techlib/misc/siRNA_finder.html). In brief, we selected siRNA sequences that startedwith AA and then analyzed every sequence by BLAST(http://www.ncbi.nlm.nih.gov/BLAST) to ensure thatthere were no significant sequence homologies withother genes. Four siRNA sequences each for GFP and thecellular gene HALP (“HIV-1 Associated Life Preserver”)(Yin et al., 2004) were selected for testing. Meanwhile,scrambled siRNA sequences for each correspondingsiRNA were designed, and these sequences also wentthrough BLAST analysis. GFP and HALP siRNA se-quences and their counterparts are listed in Table 1. Oncethe siRNA sequence was selected, we converted thesiRNA sequence into a primer with five A's (as ter-mination sites for polymerase III), a sense siRNAsequence, a 9 base pair spacer, an anti-sense siRNA se-quence and a 3′ U6 primer sequence (Fig. 1). PCRreactions were the same as those used in the human U6promoter cloning above. The PCR products were con-firmed using agarose gel electrophoresis and purified withthe QIAquick PCR purification kit (Qiagen). The mosteffective SECs were cloned into a TA cloning vector(Invitrogen) and confirmed by DNA sequencing. Syn-thetic siRNA forGFPwas synthesized by IDT (Coralville,IA) and annealed according to the manufacturer'sprotocol.

2.5. Transfection

All transfections of the CEM-SS T cell line andprimary human CD4 T cells were done with aNucleofector device and corresponding kits (amaxa,Inc., Cologne Germany) in 12-well plates. Transfectionprotocols were performed following the manufacturer'sinstructions. First, we optimized the transfection programfor CEM-SS T cells. In brief, CEM-SS cells weretransfected with Nucleofector solutions R, T or V, usingeight different Nucleofector programs; the Nucleofectorsolution and program that resulted in the highesttransfection efficiency with the lowest mortality wereselected. For CEM-SS cells, program O-17 and Cell LineNucleofector Kit R were identified as the best combina-tion. Human CD4 T cells were transfected according tothe manufacturer's recommendations; resting CD4 Tcellswere transfected with program U-14 and activated CD4 Tcells were transfected with program T-23. A GFPexpression vector (pEGFP-N1, Clontech, Mountain

View, CA) was co-transfected with the respective GFPsiRNA sequences, cells were collected at different timespost-transfection, and GFP silencing effects were quan-tified by FACS analysis.

PBMCs transfected with scrambled control or GFPsiRNAwere analyzed by real-time PCR for 18S, OAS1and MX1 (Assay-on-Demand kit, Applied Biosystems,Foster city, CA.). RNA was isolated 48 h aftertransfection (RNeasy Mini kit, Qiagen) and cDNA wassynthesized (RNA PCR Core kit, Applied Biosystems).Real-time PCR was performed using the SDS 7000(Applied Biosystems). Data were calculated by therelative quantitation method (ΔΔCt), compared to 18Sas internal control.

2.6. Fluorescence microscopy

CEM-SS cells transfected with GFP and GFP siRNAexpression vectors were collected 22 h post-transfec-tion, washed once and resuspended in PBS. Cells were

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

225

226

227

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

4 J. Yin et al. / Journal of Immunological Methods xx (2006) xxx–xxx

ARTICLE IN PRESS

EC

settled on microscope glass slides, mounted withcoverslips and analyzed immediately. Fluorescenceand Nomarski images were recorded by a digitalfluorescence microscopy system (Intelligent ImagingInnovations, Denver, CO) consisting of a ZeissAxioplan microscope fitted with a Xenon light sourceand a Sensicam CCD camera (Cooke, Auburn Hills,MI). Slidebook software (Intelligent Imaging Innova-tions) was used for image analysis. All images wererenormalized to the same range of intensities.

2.7. Northern blotting

HALP siRNA transfected CEM-SS cells werecollected 72 h post-transfection. Total cellular RNAwas isolated using the RNeasy Mini Kit (Qiagen),according to the manufacturer's instructions. Tenmicrograms total RNA was denatured and fractionatedin 1% formaldehyde-agarose gels and then transferred toHybond N+ membrane using a Vacuum blotter (Model785, Bio-Rad, Hercules, CA) for 90 min. The membranewas illuminated by a UV-crosslinker (UV Stratalinker2400, Stratagene, La Jolla, CA) and stored at roomtemperature until use. The blots were hybridized with aHALP coding region probe (Yin et al., 2004) labeledwith [α-32P]-dCTP (RadPrime DNA Labeling System,Life Technology, Gaithersburg, MD). Hybridization wasperformed at 68 °C for 3 h in the ExpressHybHybridization Solution (Clontech). After washing,membranes were exposed to X-ray film (Fuji, Tokyo,Japan) for varying lengths of time.Membranes were thenstripped and re-probed for GAPDH, and relative banddensities measured using a Bio-Rad Quantity Oneprogram.

UNCO

RR

B

A

5’ RNA Pol. III Promoter

5’ U6 primer

3’ U6 prim

5’

3’

GAUCCGCCACAACA

CUAGGCGGUGUUG

Fig. 1. Schema of a typical hairpin siRNA produced by an siRNA expressiosiRNA expression cassette; B. Hairpin siRNA structure.

TEDPR

OOF

3. Results and discussion

3.1. Design of siRNA, siRNA expression cassettes(SECs) and some practical considerations

The selected siRNA sequence has a strong influenceon the level and specificity of gene silencing. There are,however, no clear rules governing siRNA target site se-lection for specific mRNA sequences. In brief, we usedweb- and laboratory-based programs to select candidatesiRNA sequences comprised of 21 bp and an AA start, aspreviously described (website: http://www.rockefeller.edu/labheads/tuschl/sirna.html). If only a few candidatesiRNA sequences were identified using these criteria,other siRNA sequences were selected based on a GC ratioof about 50%. All sequences were then entered into aBLAST search to ensure that there was no significantsequence homology with other genes.

There are, in practice, three formats for siRNA:synthetic siRNAs, vector-expressed siRNAs andsiRNA expression cassettes (SECs). Synthetic siRNAsare usually comprised of 21-mer duplexes, whichsilence effectively in most adherent cells. SyntheticsiRNAs have a limited half-life and are diluted afterseveral cell divisions. In contrast, retroviral orlentiviral vectors are able to efficiently transduce avariety of target cells and confer sustained long-termexpression of siRNA (Lee et al., 2002; Miyagishi andTaira, 2002; Paul et al., 2002; Sui et al., 2002; Qin etal., 2003), although cloning of each siRNA into theexpression vector and testing of each vector sequencecan be extremely time-consuming. This is particularlyproblematic in the absence of defined rules guidingsiRNA design, since the choice of siRNA sequence for

Spacer

3’

er si-sense spacer si-antisense AAAAA

UCGAGGA

UAGCUCCU

CC

C

U AA

AAA

n vector or an siRNA expression cassette. A. A typical PCR-generated

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

305

306

307

308

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

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

5J. Yin et al. / Journal of Immunological Methods xx (2006) xxx–xxx

ARTICLE IN PRESS

UNCO

RREC

efficient silencing is still largely by trial and error.Thus, many siRNA sequences may need to be testedfor gene silencing.

The SEC is a PCR product consisting of a promoter(e.g., U6 promoter) and terminator sequence flanking aDNA insert encoding a hairpin siRNA (Castanotto etal., 2002). After transfection into cells, the DNA insertencoding the hairpin siRNA is expressed from the PCRproduct and induces gene specific silencing. SECs have anumber of advantages for gene suppression. First, SECpreparation is inexpensive and there is no need to syn-thesize siRNA in vitro. Second, SEC preparation by PCRis fast, particularly using our modified protocol (detailedinMaterials andmethods), in which a one-step rather thana two-step PCR is used. The entire process of SEC pre-paration can be completed within 2 h and does not requirecloning, plasmid preparation or sequencing. Third, testingof SECs is an efficient way to assay silencing by manydifferent siRNA sequences, an important feature given theuncertainties surrounding siRNA design. Once identified,the optimal SEC can then be PCR amplified and clonedinto the vector system of choice.

SECs are constructed by cloning siRNA templatesinto small, highly active RNA polymerase III (Pol III)U6 transcriptional units (Castanotto et al., 2002) (seeFig. 1). The transcriptional initiation site of the humanU6 promoter is always guanosine and its terminationsignal is a run of four or five thymidines. In our system,four candidate siRNA sequences were chosen for eachgene product and converted into 3′ primers for use inPCR of SECs. We discarded the first two nucleotides(AA) in the sequence and, if a G followed AA, used thesequence directly in the conversion. Otherwise, an extraGwas added before the siRNA sequence and used in theconversion. We integrated five adenines and the senseand antisense siRNA sequences directly into our 3′ U6primers, with a spacer between the sense and antisensesiRNA sequences. We then used a 5′ universal primer,the 3′ integrated primers and human U6 promoter as thetemplate for PCR. After transfection of SECs into cells,sense and antisense strands were transcribed under thecontrol of the human U6 promoter and the siRNAs wereexpressed as hairpin structures (Brummelkamp et al.,2002; Lee et al., 2002; Miyagishi and Taira, 2002; Paulet al., 2002). Fig. 1 shows a typical PCR generated RNAexpression cassette and hairpin siRNA structure.

We used Nucleofector technology to test candidateSECs in a human suspension T cell line and in primaryhuman resting and activated CD4 T cells. Unlike stan-dard electroporation systems, the Nucleofector techni-que uses a combination of specialized solutions andelectrical pulses to directly transfer DNA into the cell

TEDPR

OOF

nucleus. Optimized programs and proprietary solutionsare provided for maximizing viability and transfectionefficiency of specific cell types, or optimization may betailored to the cell type of interest. For example, testingof CEM-SS T cells with the cell line optimizing kitshowed that program O-17 yielded 60% cell viabilityand 60–70% transfection efficiency. In contrast, routinetransfection techniques generally yield transfectionefficiencies in T cells of only 1–2% (Cron et al.,1997). Thus, the Nucleofector technique can achievehigh transfection efficiencies in hard-to-transfecthuman T cells.

3.2. Suppression of GFP expression using SECs in ahuman T cell line

CEM-SS cells were transfected with pEGFP-N1(0.5 μg) or co-transfected with pEGFP-N1 and GFPSECs (2.5 μg). As a control, pEGFP-N1 was co-trans-fected with an SEC encoding the U6 promoter fragmentand a randomly scrambled GFP siRNA sequence. Equalamounts of DNA (3 μg total) were used for eachtransfection, except for the ‘no vector’ or pEGFP-N1controls. Cells were collected for flow cytometric ana-lysis 24 h post-transfection.

As shown in Fig. 2, the GFP SECs induced a decreasein mean fluorescence intensity (MFI), although nosignificant change was noted in the percent of cells ex-pressing GFP. Among the four GFP SECs tested(including SEC1, designed from published sequenceGFP1 siRNA, Table 1; (Caplen et al., 2001)), all inducedpartial silencing of GFP expression compared to trans-fection of pEGFP-N1 alone, while the random controlsequences had no reproducible silencing effect on GFPexpression. GFP SEC4 induced the greatest decrease(almost 70%) in GFP expression. These data suggest thatthe SEC is a useful tool for screening the effectiveness ofsiRNA design. To further increase transfection efficiency,the most effective GFP SEC, SEC4, was cloned into a TAcloning vector (GFP4 psiRNA). We also cloned GFPSEC1 into the cloning vector (GFP1 psiRNA), for pur-poses of comparison. These expression vectors were usedin subsequent analyses.

The mechanism of siRNA action is poorly under-stood; data show that siRNA works at both post-trans-criptional and translational levels, and recent studiessuggest that siRNA may directly inhibit transcription(Morris et al., 2004). Prior studies have shown thatsiRNA should be titered to the lowest functional dose toavoid non-specific effects. Thus, CEM-SS cells were co-transfected with pEGFP-N1 (0.5 μg) and differentamounts of siRNA expression vector. Cells were

PROO

F358

359

360

361

362

363

364

365

Co-transfected with pEGFP plus:

pEGFP + U6 vector

SEC1

SEC4

21%

No vectorTransfected with:

Silencing EffectGFP SEC Random scrambled

control SEC

EGFP

EGFP

65%MFI = 389

64%MFI = 246

47%MFI = 129

58%MFI = 357

3%MFI = 15

62%MFI = 310

64%

Co

un

ts

104100 104100

104100 104100

Co

un

tsC

ou

nts

200

0

200

0

200

0

Fig. 2. Screening the inhibitory effect of GFP siRNAs using siRNA expression cassettes (short PCR products). CEM-SS T cells were co-transfectedwith pEGFP-N1 (0.5 μg) and U6 or GFP siRNA cassettes or corresponding GFP siRNA random scrambled controls. Cells were collected for flowcytometric analysis 24 h post-transfection. 2.5 μg of PCR product was used for the U6 and pEGFP-N1 transfection controls. A total of 3 μg DNAwasused for each transfection. The silencing effect was calculated as (Random control MFI−Test sample MFI) /Random control MFI.

6 J. Yin et al. / Journal of Immunological Methods xx (2006) xxx–xxx

ARTICLE IN PRESS

collected for flow cytometric analysis 24 h post-trans-fection. Compared to their corresponding randomscrambled controls, transfection of 1 μg GFP4 psiRNAinduced a 49% decrease in GFP expression, while

UNCO

RRECCo-transfected with

pEGFP plus :

1 µµg

2.5 µg

5 µg

EGFP

41%MFI = 235

66%MFI = 968

42%MFI = 257

Random scrambled control psiRNA

200

0

200

0

200

0

Co

un

tsC

ou

nts

Co

un

ts

104100

Fig. 3. Titration of GFP siRNA (using the most effective siRNA, GFP4, fromscrambled control siRNA. CEM-SS cells were co-transfected with pEGFP-N1were collected for flow cytometric analysis 24 h later. A dose of 2.5 μg appGFP4 psiRNA.

EDtransfection of 2.5 μg GFP4 psiRNA decreased ex-pression by 90% (Fig. 3). In contrast, transfection oflarger amounts of GFP4 psiRNA or the random sc-rambled control, 3.5 μg (not shown) and 5 μg (Fig. 3),

T

Silencing Effect

49%

59%

90%

49%MFI = 120

47%MFI = 98

57%MFI = 105

GFP4 psiRNA

104100

experiments such as that shown in Fig. 2) and corresponding random(N1, 0.5 μg) and increasing amounts of siRNA expression vector. Cellseared to be the lowest effective dose and induced optimal silencing by

ECTEDPR

OOF

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

Untransfected

pEGFP +

Empty U6 vector

pEGFP +

Control psiRNA

pEGFP +

GFP4 psiRNA

Transfected w/:

Fig. 4. Fluorescence microscopic analysis of GFP silencing by siRNA. CEM-SS cells were co-transfected with pEGFP-N1 (0.5 μg) and the GFPsiRNA expression vector or corresponding random scrambled control vector (2.5 μg). Cells were collected for analysis by flow cytometry (notshown) and fluorescence microscopy 22 h later.

7J. Yin et al. / Journal of Immunological Methods xx (2006) xxx–xxx

ARTICLE IN PRESS

UNCO

RRinduced non-specific cell death and decreased the silencingeffect. We therefore considered 2.5 μg to be the optimizedconcentration for GFP4 psiRNA under these experimentalconditions. Fig. 4 confirmed the silencing effect of ouroptimized GFP4 psiRNA by fluorescence microscopy.

In order to assess the kinetics of silencing, CEM-SScells were co-transfected with pEGFP-N1 (0.5 μg) andGFP4 psiRNA expression vector or its control vector(2.5 μg) and collected for flow cytometric analysis atdifferent time points. The level of GFP expression wasmaximal between 24 and 48 h post-transfection and thendecreased gradually over 7 days in culture. Maximalsilencing of 85–90%, compared with the correspondingrandom controls, was also seen by 24 h post-trans-fection and remained stable over the next 7 days, dec-reasing in proportion to GFP expression (Fig. 5 and datanot shown).

In order to assess the effectiveness of SECs com-pared to synthetic siRNAs, CEM-SS cells were co-transfected with pEGFP-N1 (0.5 μg) and GFP siRNA orits scrambled control and collected for flow cytometricanalysis 48 h after transfection. As shown in Fig. 6,psiRNA (SEC) and synthetic siRNA inhibited GFPexpression to similar levels. These data show that theSEC is a good alternative to synthetic siRNA to silencespecific genes.

3.3. Inhibition of GFP expression using a vector-basedsystem in resting and activated primary human CD4 Tcells

Although CEM-SS is a human CD4 T lymphoblastoidcell line, our goal was to apply the siRNA technique toprimary human CD4 Tcells. Resting and activated primary

ROOF

398

399

400

401

402

403

404

405

406

407

408

409

410

411

Co-transfected with pEGFP plus :

GFP4 psiRNA

Silencing Effect

86%86%24 hrs

48 hrs

72 hrs

EGFP

70%MFI: 1220

31%MFI: 42

55%MFI: 351

69%MFI: 657

56%MFI: 78

61%MFI: 181

Random scrambledcontrol psiRNA

100

0

100

0

100

0104100 104100

Co

un

tsC

ou

nts

Co

un

ts

88%88%

88%88%

Fig. 5. Time course analysis of GFP silencing by siRNA. CEM-SS cells were co-transfected with pEGFP-N1 (0.5 μg) and the GFP siRNA expressionvector or its random control vector (2.5 μg). Cells were collected for flow cytometric analysis at different time points.

8 J. Yin et al. / Journal of Immunological Methods xx (2006) xxx–xxx

ARTICLE IN PRESS

humanCD4Tcells were transfected using theNucleofectorHuman T Cell Kit, using programs specific to each celltype, according to the manufacturer's instructions (seeMaterials and methods). Co-transfection of randomscrambled siRNA control vector (2.5 μg) with pEGFP-N1 (0.5 μg) resulted in approximately equal expression ofGFP in primary human resting and activated CD4 Tcells at

UNCO

RREC

Co-transfected with pEGFP plus :

100 pMoles

200 pMoles

EGFP

Scrambled controlsynthetic siRNA

200

0

41%MFI: 95

43%MFI: 103

200

0104100

Co

un

tsC

ou

nts

Co-transfected withpEGFP plus :

1µg

EGFP

Scrambled controlpsiRNA

47%MFI: 115

200

0104100

Co

un

ts

Fig. 6. GFP silencing by synthetic siRNA. CEM-SS cells were co-transfectsynthetic siRNA (top) or GFP siRNA expression vector (SEC, bottom). Cellswere similar 72 h after transfection (not shown).

EDP24 h post-transfection. Co-transfection of siRNA expres-

sion vector with pEGFP-N1 successfully silenced GFPexpression in both resting and activated T cell populations,although to different extents. Compared to the randomcontrol, the inhibitory effects of GFP1 psiRNAwere 16%for resting T cells and 21% for activated T cells (data notshown); the inhibitory effects of GFP4 psiRNAwere 56%

T

Silencing Effect

84%84%

GFP4synthetic siRNA

85%85%

20%MFI: 15

20%MFI: 15

104100

83%83%

GFP4psiRNA

26%MFI: 20

104100

ed with pEGFP-N1 (0.5 μg) and the indicated concentrations of GFPwere collected for flow cytometric analysis 48 h after transfection. Data

OF

412

413

414

415

416

417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

Fresh CD4 T cells

Activated CD4 T cells

56%

80%

EGFP

Co-transfected withpEGFP plus :

GFP4psiRNA Silencing Effect

Random scrambled control psiRNA

100

0

104100 104100

50

0C

ou

nts

Co

un

ts

62%MFI: 1041

50%MFI: 466

57%MFI: 880

30%MFI: 184

Fig. 7. GFP siRNA inhibits GFP expression in primary human CD4 T cells. Resting or PHA-activated CD4 T cells were transfected with 0.5 μgpEGFP-N1 and 2.5 μg GFP siRNA expression vector or its random control. Cells were collected for flow cytometric analysis 24 h later.

9J. Yin et al. / Journal of Immunological Methods xx (2006) xxx–xxx

ARTICLE IN PRESS

for resting T cells and 80% for activated T cells (Fig. 7).These data show that resting and activated primary humanT cells can be transfected with equal efficiencies byNucleofection, but that siRNA silencing is more pro-nounced in activated T cells.

3.4. Efficient suppression of the endogenous gene,HALP, using a vector-based system

Having established proof of principle for siRNAsilencing of exogenous gene expression in human T celllines and primary human CD4 T cells, we asked whether

UNCO

RREC

HALP

GAPDH

Control HALP1 psiR

0.5 1.0

A. Screening Control GFP4 HALP1

psiRNA psiRNA

GAPDH

HALP

B. Titration

HALP/GAP/GAPDH ratio 3.00 ratio 3.00 1.81.89 0.96 0.96

HALP/GAP/GAPDH ratio 0.94 ratio 0.94 0.73 0.73 0.45 0.45

Fig. 8. HALP siRNAs inhibit endogenous HALP expression in CEM-SS celcells were transfected with 2.5 μg GFP siRNA as control or with HALP siRNpost-transfection. B. Titration of the most effective HALP siRNA. Transfectioused for each transfection is as shown. Total RNA isolation and Northern bl

EDPR

Othis approach could efficiently inhibit endogenous geneexpression.HALP (an acronym for “HIV-1AssociatedLifePreserver”) is a novel gene isolated and cloned in a screenfor apoptosis regulators induced byHIV-1 infection of CD4T cells (Yin et al., 2004). As in our analyses of GFPsilencing, we designed four HALP siRNAs and fourcorresponding random scrambled controls; these constructswere then cloned into the vector expression systemdescribed above (HALP1–4 psiRNAs). First, we screenedthe effectiveness of the four HALP siRNAs in CEM-SScells. Northern blot analysis showed that HALP1 and 2psiRNAs induced efficient inhibition of endogenousHALP

T

NA (µµg) Control1 psiRNA (µg)

2.5 0.5 1.0

HALP2 HALP3 HALP4 Control1/2 Control3/4psiRNA psiRNA psiRNApsiRNA psiRNA

1. 1.0202 2.28 2.28 7.00 7.00 3.06 3.06 1.88 1.88

0.02 0.02 0.96 0.96 0.93 0.93 1.08 1.08

ls. A. Screening the effectiveness of four siRNAs for HALP. CEM-SSA 1–4 or their random scrambled controls. Cells were collected 3 daysns were performed as above. The amount of siRNA expression vectorot procedures were as detailed in Materials and methods.

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

489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541

10 J. Yin et al. / Journal of Immunological Methods xx (2006) xxx–xxx

ARTICLE IN PRESS

UNCO

RREC

expression (Fig. 8A). We then titered the most effectivepsiRNA, HALP1. As shown in Fig. 8B, the silencing effectof HALP1 psiRNA gradually increased from 0.5 to 2.5 μg.At 2.5 μg, the HALP1 psiRNA inhibited endogenousHALP expression by more than 90%.

In this study, we present a combined strategy toidentify effective siRNA target sequences and knockdown exogenous or endogenous genes in hard-to-transfect human suspension T cell lines and primary Tcells. As discussed, there are three forms of siRNA.Synthetic siRNAoligonucleotides can be used directly intransfection, and robustly silence genes transiently insome cell types, especially in adherent cells. However,synthetic oligonucleotides are very expensive, not verystable and in limited supply. In contrast, siRNA expres-sion cassettes are cost-effective, stable and have essen-tially unlimited availability. Once an effective siRNAtarget sequence is identified, this SEC can be cloned intoany vector, e.g., TA cloning vectors for transient trans-fection, or other vectors (including adeno-associatedvirus, retroviral or lentiviral vectors) to generate stable“knockdown” cell lines or to silence genes in vivo.

Silencing of specific genes with siRNA has beenreported in some systems to have off-target effects,including induction of interferon response genes (Bridgeet al., 2003). We analyzed interferon dependent andindependent effects by real-time PCR. In PBMCstransfected with the scrambled control or GFP psiRNA,the interferon response gene, OAS1, increased 1.5- or0.9-fold, respectively, when compared to mock trans-fected controls. We observed a similar lack of effect onanother interferon response gene,MX1 (data not shown).In contrast, human IFN-β (500 IU) induced an increasein OAS1 and MX1 gene expression of 44- and 56-fold,respectively. Furthermore, the HALP psiRNA did notinduce increases in GAPDH, an interferon independentgene, when compared to controls (data not shown).These data suggest that our SECs suppress target geneswithout nonspecific effects on representative interferondependent and independent off-target genes.

In summary, we provide strong evidence that thecombination of SEC, siRNA expression vectors andNucleofector technology can be successfully applied tohard-to-transfect human suspension T cell lines andprimary T cells to effectively silence genes. This processwill greatly facilitate the use of siRNA technology insuspension cells for immunologic research.

Acknowledgements

This study was supported by NIH R01 AI40003, R01AI35513 and R21 AI054233, the Penn CFAR and Cancer

Center, the Bender Foundation, the NIH AIDS Researchand Reference Reagent Program, the Joseph Lee Hollan-der Chair, and by the W.W. Smith Charitable Trust(THF). RQC was supported by NIH R01 AR48257.

TEDPR

OOF

References

Bernstein, E., Caudy, A.A., Hammond, S.M., Hannon, G.J., 2001.Role for a bidentate ribonuclease in the initiation step of RNAinterference. Nature 409, 363.

Bridge, A.J., Pebernard, S., Ducraux, A., Nicoulaz, A.L., Iggo, R.,2003. Induction of an interferon response by RNAi vectors inmammalian cells. Nat. Genet. 34, 263.

Brummelkamp, T.R., Bernards, R., Agami, R., 2002. A system forstable expression of short interfering RNAs in mammalian cells.Science 296, 550.

Caplen, N.J., Parrish, S., Imani, F., Fire, A., Morgan, R.A., 2001.Specific inhibition of gene expression by small double-strandedRNAs in invertebrate and vertebrate systems. Proc. Natl. Acad.Sci. U. S. A. 98, 9742.

Castanotto, D., Li, H., Rossi, J.J., 2002. Functional siRNA expressionfrom transfected PCR products. RNA 8, 1454.

Cron, R.Q., 2001. HIV-1, NFAT, and cyclosporin: immunosuppressionfor the immunosuppressed? DNA Cell Biol. 20, 761.

Cron, R.Q., 2003. CD154 transcriptional regulation in primary humanCD4 T cells. Immunol. Res. 27, 185.

Cron, R.Q., Schubert, L.A., Lewis, D.B., Hughes, C.C., 1997.Consistent transient transfection of DNA into non-transformedhuman and murine T-lymphocytes. J. Immunol. Methods 205,145.

Elbashir, S.M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K.,Tuschl, T., 2001a. Duplexes of 21-nucleotide RNAs mediate RNAinterference in cultured mammalian cells. Nature 411, 494–498.

Elbashir, S.M., Lendeckel, W., Tuschl, T., 2001b. RNA interfer-ence is mediated by 21- and 22-nucleotide RNAs. Genes Dev.15, 188.

Fire, A., 1999. RNA-triggered gene silencing. Trends Genet. 15, 358.Hamilton, A.J., Baulcombe, D.C., 1999. A species of small antisense

RNA in posttranscriptional gene silencing in plants. Science 286,950.

Hammond, S.M., Bernstein, E., Beach, D., Hannon, G.J., 2000. AnRNA-directed nuclease mediates post-transcriptional gene silenc-ing in Drosophila cells. Nature 404, 293.

Lee, N.S., Dohjima, T., Bauer, G., Li, H., Li, M.J., Ehsani, A.,Salvaterra, P., Rossi, J., 2002. Expression of small interferingRNAs targeted against HIV-1 rev transcripts in human cells. Nat.Biotechnol. 20, 500.

Miyagishi, M., Taira, K., 2002. U6 promoter-driven siRNAs with foururidine 3′ overhangs efficiently suppress targeted gene expressionin mammalian cells. Nat. Biotechnol. 20, 497.

Morris, K.V., Chan, S.W., Jacobsen, S.E., Looney, D.J., 2004. Smallinterfering RNA-induced transcriptional gene silencing in humancells. Science 305, 1289.

Paul, C.P., Good, P.D., Winer, I., Engelke, D.R., 2002. Effectiveexpression of small interfering RNA in human cells. Nat.Biotechnol. 20, 505.

Qin, X.F., An, D.S., Chen, I.S., Baltimore, D., 2003. Inhibiting HIV-1infection in human T cells by lentiviral-mediated delivery of smallinterfering RNA against CCR5. Proc. Natl. Acad. Sci. U. S. A.100, 183.

542543544545546547548549550

551552553554555556

557

558

11J. Yin et al. / Journal of Immunological Methods xx (2006) xxx–xxx

ARTICLE IN PRESS

Scherer, L.J., Rossi, J.J., 2003. Approaches for the sequence-specificknockdown of mRNA. Nat. Biotechnol. 21, 1457.

Sharp, P.A., 2001. RNA interference—2001. Genes Dev. 15, 485.Sui, G., Soohoo, C., Affar el, B., Gay, F., Shi, Y., Forrester, W.C.,

2002. A DNA vector-based RNAi technology to suppress geneexpression in mammalian cells. Proc. Natl. Acad. Sci. U. S. A. 99,5515.

Tuschl, T., 2002. Expanding small RNA interference. Nat. Biotechnol.20, 446.

UNCO

RREC

Yin, J., Chen, M.F., Finkel, T.H., 2004. Differential gene expressionduring HIV-1 infection analyzed by suppression subtractivehybridization. Aids 18, 587.

Zamore, P.D., Tuschl, T., Sharp, P.A., Bartel, D.P., 2000. RNAi:double-stranded RNA directs the ATP-dependent cleavage ofmRNA at 21 to 23 nucleotide intervals. Cell 101, 25.

TEDPR

OOF