• No results found

(1)1 Engineering herpes simplex viruses by infection-transfection methods including 1 recombination site targeting by CRISPR/Cas9 nucleases 2 3 Tiffany A

N/A
N/A
Protected

Academic year: 2023

Share "(1)1 Engineering herpes simplex viruses by infection-transfection methods including 1 recombination site targeting by CRISPR/Cas9 nucleases 2 3 Tiffany A"

Copied!
25
0
0

Loading.... (view fulltext now)

Full text

(1)

1

Engineering herpes simplex viruses by infection-transfection methods including 1

recombination site targeting by CRISPR/Cas9 nucleases 2

3

Tiffany A. Russella, Tijana Stefanovica and David C. Tscharkea, # 4

5

aResearch School of Biology, Bldg #134 Linnaeus Way, The Australian National University, 6

Canberra, ACT, 0200, Australia ([email protected]; [email protected]; 7

[email protected]).

8 9

#Address for correspondence: David Tscharke, Research School of Biology, Bldg #134 10

Linnaeus Way, The Australian National University, Canberra ACT 11

0200, [email protected], T: +61 2 6125 3020, F: +61 2 6125 0313 12

(2)

2 Summary

13

Herpes simplex viruses (HSV) are frequent human pathogens and the ability to engineer 14

these viruses underpins much research into their biology and pathogenesis. Often the 15

ultimate aim is to produce a virus that has the desired phenotypic change and no additional 16

alterations in characteristics. This requires methods that minimally disrupt the genome and, 17

for insertions of foreign DNA, sites must be found that can be engineered without disrupting 18

HSV gene function or expression. This study advances both of these requirements. Firstly, 19

the use of homologous recombination between the virus genome and plasmids in 20

mammalian cells is a reliable way to engineer HSV such that minimal genome changes are 21

made. This has most frequently been achieved by cotransfection of plasmid and isolated 22

viral genomic DNA, but an alternative is to supply the virus genome by infection in a 23

transfection-infection method. Such approaches can also incorporate CRISPR/Cas9 genome 24

engineering methods. Current descriptions of infection-transfection methods, either with or 25

without the addition of CRISPR/Cas9 targeting, are limited in detail and the extent of 26

optimisation. In this study it was found that transfection efficiency and the length of 27

homologous sequences improve the efficiency of recombination in these methods, but the 28

targeting of the locus to be engineered by CRISPR/Cas9 nucleases has an overriding 29

positive impact. Secondly, the intergenic space between UL26 and UL27 was reexamined as 30

a site for the addition of foreign DNA and a position identified that allows insertions without 31

compromising HSV growth in vitro or in vivo.

32

Keywords 33

Herpes simplex virus, genome engineering, recombinant virus, CRISPR, Cas9 34

(3)

3 1. Introduction

35

Herpes simplex virus (HSV) types 1 and 2 are highly prevalent human pathogens, with HSV- 36

1 infecting approximately 60% of people worldwide (Cunningham et al., 2006; Bradley et al., 37

2014). HSV is also extensively studied as the prototypical alphaherpesvirus due to the 38

relative ease with which it can be grown and the wide variety of in vitro and in vivo models 39

available (Simmons Nash, 1984; Sawtell Thompson, 1992; Shimeld et al., 1996; Leland 40

Ginocchio, 2007; Hogk et al., 2013). Recombinant HSV expressing foreign genes have 41

proven invaluable for studying viral pathogenesis and growth, as well as for screening for 42

potential antiviral agents (Tanaka et al., 2004; Balliet et al., 2007; Ramachandran et al., 43

2008; Ding et al., 2012). Further, HSV has shown some promise as a recombinant vaccine 44

vector, especially against cancer (Markert et al., 2000; Rampling et al., 2000; Goins et al., 45

2008). Ideally methods for making recombinant HSV should a) leave minimal changes other 46

than those desired in the genome and b) where foreign genes are added, these should be 47

inserted at sites that do not impact the growth and pathogenesis of the virus.

48

The original method for making such viruses relies upon homologous recombination 49

between a transfer plasmid that has copies of the viral sequences flanking the desired 50

insertion site and the virus genome in cultured mammalian cells. The relatively low rate at 51

which this occurs means that efficient methods are required to select or screen the few 52

recombinant viruses that are produced (Tanaka et al., 2004; Ramachandran et al., 2008).

53

More recently, recombineering of HSV genomes propagated as Bacterial Artificial 54

Chromosomes, or BACs, has been used. However, viruses recovered from these usually 55

contain residual BAC sequences and/or are attenuated in vivo due to other unwanted 56

changes (Horsburgh et al., 1999; Tanaka et al., 2003; Gierasch et al., 2006). Therefore, the 57

original methods remain essential tools that continue to be used.

58

In non-BAC homologous recombination-based methods, cotransfection of viral and transfer 59

plasmid DNA is the most common way of generating recombinant HSV. While detailed 60

(4)

4

reports in the literature are sparce, anecdotally this relies heavily on obtaining very high 61

quality HSV genomic DNA. A simpler alternative is to provide the HSV genome by infection 62

of cells transfected with a transfer plasmid (transfection/infection) and at least one report of 63

the use of such as method can be found, but few details were included (Orr et al., 2005).

64

Transfection/infection is also a common way to engineer poxviruses, which have large 65

dsDNA genomes that unlike HSV are non-infectious (Mackett et al., 1984; Wong et al., 66

2011). In addition, such methods can be combined with CRISPR/Cas9 genome editing tools 67

(Bi et al., 2014; Suenaga et al., 2014). However, thus far the improvement in recombination 68

frequency associated with the application of CRISPR/Cas9 targeting has not been made 69

against optimised transfection/infection methods.

70

A variety of different locations have been identified in the HSV-1 genome which allow the 71

insertion of foreign DNA with minimal disruption of genes. These include intergenic regions 72

between UL3 and UL4, UL50 and UL51 and US1 and US2, but only the first of these has 73

been well characterized (Tanaka et al., 2004; Morimoto et al., 2009). In each case, the 74

genes either side of the insertion site are convergently transcribed and each has its own 75

polyA signal between which there is enough sequence for an insertion to be made without 76

disrupting either transcription unit. Most other common sites of insertion, such as the 77

US5/US6 location and UL23 (thymidine kinase) lead to disruption of some ORFs, generally 78

leading to attenuation in vivo (Rinaldi et al., 1999; Proenca et al., 2008). The space between 79

UL26 (glycoprotein B, gB) and UL27 genes has the ideal structure described above, but 80

previous attempts to use this insertion site have led to some loss of virulence (Halford et al., 81

2004; Orr et al., 2005). It remains possible that this site can accept insertions without 82

compromising virulence if these are targeted to ensure there is no disruption of the 83

transcription units, including polyA sites.

84

The aims of this study were to explore transfection-infections approaches for generating 85

recombinant HSV, including CRISPR/Cas9 targeting and to identify a precise position 86

between UL26 and UL27 where foreign genes can be inserted without loss of virulence.

87

(5)

5 88

2. Materials and Methods 89

2.1. Viruses and cell lines 90

The unmodified HSV-1 strain KOS was provided by Francis Carbone (University of 91

Melbourne, Australia). HSV-1 pCmC contains the fluorescent reporter mCherry under the 92

control of the cytomegalovirus immediate early (CMV IE) promoter located in the intergenic 93

region between UL3 and UL4 of HSV-1 KOS (HSV-1 KOS 11649). This virus was 94

constructed by standard homologous recombination based methods following four rounds of 95

plaque purification.

96

All viruses were grown and titrated on Vero cells (ATCC CCL-81). The immortalized Vero 97

cell line was maintained in Minimal Essential Medium (MEM; Gibco/Life Technologies, 98

Carlsbad, USA) supplemented with 2 or 10% heat-inactivated fetal calf serum, 5 mM 4-(2- 99

hydroxyethyl)-1-piperazineethanesulfonic acid, 4 mM L-glutamine and 50 mM 2- 100

mercaptoethanol. All transfections were carried out on 293A cells with Lipofectamine 2000 101

(Life Technologies, Carlsbad, USA).

102

2.2. Plasmid construction 103

All sequence references below are to the HSV-1 genome, accession JQ673480. To 104

construct the generic transfer vector pT UL3/4, the UL3/UL4 region (HSV-1 10534-12682) 105

was cloned into pTracer CMV/bsd (Life Technologies, Carlsbad, USA) by In-Fusion cloning 106

(Clontech Laboratories, Mountain View, USA). These HSV-1 sequences were generated in 107

two polymerase chain reactions (PCR) to enable the addition of EcoRV, PstI and SpeI sites 108

between the polyA signals of UL3 and UL4 (HSV-1 11649) by the use of extended primers to 109

make pT UL3/4.

110

The cytomegalovirus immediate early (CMV IE) promoter and bovine growth hormone (BGH) 111

poly A termination sequence were amplified from pTracer CMV/bsd and the eGFP Cre 112

(6)

6

cassette was amplified from pIGCN21 (Lee et al., 2001). These fragments were then cloned 113

into the SpeI site of pT UL3/4 by In-Fusion cloning to construct pT pC_eGC (Fig 1A).

114

To construct plasmids with different lengths of homology sequence, sequences flanking the 115

intergenic UL3/UL4 region were amplified and cloned into the pCR bluntII vector (Life 116

Technologies, Carlsbad, USA). Four plasmids were made in this way, namely pU3.0.5kbF 117

(HSV-1 11200-12179), pU3.1kbF (HSV-1 10700-12722), pU3.2kbF (HSV-1 9803-13698) 118

and pU3.3kbF (HSV-1 8689-14663), such that a MCS containing KpnI and NotI sites are 119

inserted in the middle of a fragment of the UL3/UL4 intergenic region (HSV-1 11649). The 120

following three synthetically constructed elements were inserted into the MCS of each of 121

these plasmids (Genscript, Piscataway, USA): A) The ICP47 promoter lacking the origin of 122

replication (OriS) sequence (Summers Leib, 2002). The sequence encoding the OriS was 123

removed as it has been shown that this plays no role in regulating the transcription of ICP47 124

(Summers Leib, 2002). B) A Venus reporter gene containing a SV40 nuclear localization 125

sequence. C) A BGH polyA terminator sequence. The resulting plasmids were named 126

pU3.0.5kbF-Venus, pU3.1kbF-Venus, pU3.2kbF-Venus and pU3.3kbF-Venus (Fig 2A).

127

To construct pU26/7, the UL26/UL27 region (HSV-1 51431-54154) with EcoRV, NotI and 128

SpeI sites added between the two polyA signals (at HSV-1 52809) was inserted into pUC19 129

(Clontech Laboratories, Mountain View, USA) to make pU26/7. Into the NotI site of this 130

generic vector was inserted the ICP47 promoter (described above) upstream of a Tdtomato 131

gene with a BGH polyA termination sequence (from pCIGH3) to make pU26/7 132

pICP47/TdTom (Fig 3C).

133

The plasmid pX330 (Addgene plasmid 42230) has been previously (Cong et al., 2013). The 134

plasmid pX330-mC was constructed by annealing two complimentary oligodeoxynucleotides 135

(CACCGGATAACATGGCCATCATCA and AAACTGATGATGGCCATGTTATCC) and 136

ligating the resulting dsDNA fragment into the BbsI site of pX330.

137

2.3. Generation of recombinant HSV-1 by transfection/infection 138

(7)

7

Recombinant HSV-1 were produced by transfection of 293A cells with the required amount 139

of plasmid DNA. After 5 hours incubation (37°C, 5% CO2), cells were infected with HSV-1 140

KOS at an appropriate MOI. All cell-associated and supernatant virus was harvested from 141

the transfection with the aid of a cell lifter. This was then subjected to three cycles of 142

freezing and thawing to lyse the cells and release the virus. The virus was then serially 143

diluted and used to infect fresh cultures of Vero cells overlaid with phenol red-free semisolid 144

MEM-2 with 0.4% (w/v) carboxy-methyl cellulose (M2-CMC). This allowed the development 145

of individual plaques after 48 hours which were then able to be identified and selected by 146

fluorescence microscopy. Multiple rounds of plaque purification were carried out as 147

appropriate. PCR screening and sequencing was used to confirm the correct modification 148

occurred and to identify plaque isolates free of parental virus where appropriate. Two 149

independent recombinant viruses were isolated from parallel transfection/infection 150

experiments.

151

2.4. Replication in vitro 152

Confluent Vero cell monolayers in six well plates were infected with 1 × 104 PFU (MOI 0.01) 153

virus in 1 mL M0. After 1 h at 37°C, virus inocula were removed, the cell monolayer was 154

washed once and 2 mL fresh M2 added. The first samples (zero hour) were harvested 155

immediately after the addition of fresh media and virus from further wells was collected at the 156

times indicated. To harvest virus, cells were scraped into the media so that both were 157

collected in a single sample. These were subjected to three freeze/thaw cycles and virus 158

titres in each determined by plaque assay on Vero cells.

159

2.5. Measurement of Plaque Size 160

Confluent Vero cell monolayers in six well plates were infected with 50 PFU virus. After 161

incubation for 90 min at 37°C, 5% CO2, the inoculum was replaced with M2-CMC. 48 hours 162

later, cells were crystal violet stained and 30 representative photographs per virus were 163

(8)

8

taken at 100x magnification using an Olympus CKX41 microscope and DP20 camera.

164

Plaque area was calculated using ImageJ (Rasband, 1997-2012).

165

2.6. Mice and infections 166

This study was carried out in accordance with the Australian NHMRC guidelines contained 167

within the Australian Code of Practice for the Care and Use of Animals for Scientific 168

Purposes. Female specific pathogen free C57Bl/6 mice greater than 8 weeks of age were 169

obtained from the APF (Canberra, Australia). Mice were housed and experiments carried 170

out according to ethical requirements and under approval of the Animal Ethics Committee of 171

the Australian National University (Protocol Number: A2011.001).

172

To assess the virulence of HSV, a mouse flank infection model was used where virus was 173

introduced onto the flanks of shaved mice using a tattoo machine (Figure S1). This is a 174

variation of the flank scarification or abrasion technique sometimes referred to as the 175

zosteriform model (Blyth et al., 1984; Van Lint et al., 2004). The advantage of tattooing over 176

scarification is that the skin remains unbroken by the inoculation, so on the first day after 177

infection there is no sign of damage to the skin allowing the development of the primary 178

lesion to be clearly observed from two days later (Fig S1A). After five days, secondary (or 179

zosteriform) spread is seen, usually peaking on day seven and typically all lesions resolve by 180

14 days after infection (Fig. S1B).

181

Female C57Bl/6 mice eight weeks of age or greater were used. Mice were anaesthetized by 182

i.p. injection of avertin (1,1,1 Tribromoethanol in 2-methyl-2-butanol) given at 250 mg/kg and 183

kept warm when not being handled using an infrared lamp. The left flank of each mouse was 184

clipped and depilated with Veet cream (Reckitt Benckiser; Sydney, Australia). For tattooing, 185

a 10 round shader needle (a cluster containing 10 needles in a round pattern) was mounted 186

on a Swiss rotary tattoo machine (Pullman Tools; Widnau, Switzerland) and charged with 187

virus by dipping for 10 seconds in a suspension containing 1 × 108 PFU/mL HSV. The site 188

for infection was determined by identifying the tip of the spleen (seen through the skin) and a 189

(9)

9

5 × 5 mm area was tattooed for 10 seconds with gentle pressure and even coverage of the 190

area. Mice were monitored daily following infection for lesion development. Where mice have 191

been weighed they generally lose around 5% of body weight in the days after the infection 192

procedure and then recover; there is no evidence of generalized illness as a result of lesion 193

formation.

194

2.7. Titration of virus from skin and dorsal root ganglia (DRG) 195

A 1 cm2 portion of skin located over the inoculation site and the 10 DRG on the ipsilateral 196

side corresponding to spinal levels L1 – T5 were collected from each mouse 5 days after 197

infection. Samples were homogenized in M2, subjected to three cycles of freeze/thawing and 198

infectious virus quantified by plaque assay on Vero cells.

199

2.8. Statistical analysis 200

Statistical comparisons were performed using an unpaired t-test with Welch’s correction with 201

the aid of Prism software (version 5.01; GraphPad, La Jolla, USA).

202 203

3. Results 204

3.1. Transfection/infection methods for generating recombinant HSV-1 205

To establish the transfection/infection method a recombinant HSV was designed that would 206

express a fusion protein of enhanced green fluorescent protein and Cre recombinase 207

(eGFP/Cre) using the cytomegalovirus immediate early (CMV IE) promoter from the 208

intergenic space between HSV UL3 and UL4 genes. A fluorescent reporter was chosen to 209

enable the easy identification of recombinant viruses and the UL3/UL4 intergenic region was 210

selected because insertions at this site do not compromise growth or virulence (Tanaka et 211

al., 2004; Morimoto et al., 2009). The point of insertion was between the two native polyA 212

sequences which are necessary for proper termination of UL3, UL4 and UL5 transcription 213

(10)

10

(Morimoto et al., 2009). The plasmid used (pT pC_eGC) and the eGFP/Cre expression 214

cassette are shown in Figure 1A.

215

Three parameters associated with infection/transfection were tested to determine which 216

were important determinants of the frequency of recombinant virus generation: 1) the 217

amount of virus, or multiplicity of infection (MOI); 2) the efficiency of transfection; 3) the 218

length of flanking region sequence.

219

To determine if the amount of virus used to infect the cells influenced the frequency of 220

recombination, 293A cells were transfected with linearized pT pC_eGC DNA five hours prior 221

to infection with HSV-1 strain KOS at MOIs of 0.01, 0.001 or 0.0001. Virus was harvested 222

after three days and serial dilutions used to infect new cultures. This allowed quantification of 223

eGFP+ and eGFP- progeny. As expected, as MOI increased, total virus yields were 224

correspondingly higher but proportions of eGFP+ and eGFP- plaques remained similar (Fig 225

1B).

226

Next, to examine transfection efficiency, varied amounts of linearized or circular plasmids 227

were transfected into 293A cells to achieve differing transfection efficiencies as measured by 228

flow cytometry. These cells were then infected with HSV-1 KOS at an MOI of 0.01 and after 229

three days, virus was harvested. Serial dilutions of this virus were used to infect new cultures 230

and the proportion of total plaques that were eGFP+ was determined (Fig 1C, D). Higher 231

transfection efficiency improved the proportion of eGFP+ plaques in a roughly linear manner 232

and notably, efficiencies below 20% did not reliably produce any recombinants.

233

The third parameter tested was the length of viral sequences flanking the insertion site used 234

in the transfer plasmid. Plasmids were generated that contained left and right flanks either 235

side of the UL3/UL4 intergenic region of approximately 0.5, 1, 2, or 3 kb (Fig 2A). Venus was 236

chosen as a marker so that we could continue to use fluorescence to identify recombinant 237

viruses while widening the range of foreign genes shown to be inserted using the 238

(11)

11

transfection/infection method. In two independent experiments, these Venus transfer 239

plasmids were transfected into 293A cells with conditions that ensured transfection efficiency 240

was similar (~70 - 80% by flow cytometry, not shown) and then infected with HSV-1 at an 241

MOI of 0.01. As in previous experiments, virus was harvested after 3 days. The proportion of 242

Venus+ plaques of total virus was determined by fluorescence microscopy of cell monolayers 243

infected with serial dilutions of the progeny from these transfection/infections (Fig 2B). In 244

both experiments the frequency of Venus+ plaques was directly proportional to the length of 245

the flanking sequence in the transfer plasmids with the range of efficiency across the 246

plasmids being in the order of 10-fold.

247

3.2. CRISPR/Cas9 targeting of the recombination site has an overriding influence on 248

recombination frequency of transfection-infection methods 249

The methods detailed above gave recombination frequencies high enough to allow visual 250

selection of viruses engineered to express a fluorescent marker, but even with the 251

optimizations made thus far it would remain challenging to identify recombinants without this 252

visual aid. The recently developed use of CRISPR/Cas9 genome engineering approaches 253

offers an avenue to improve the efficiency of homologous recombination in a variety of 254

settings (Cong et al., 2013). These methods use an RNA guided nuclease (Cas9) to cleave 255

dsDNA at a desired position and these double-stranded breaks can be repaired either by 256

non-homologous end joining or, if a suitable template is available, homologous 257

recombination (Cong et al., 2013). There have been two reported applications that used 258

CRISPR/Cas9 to aid the generation of recombinant HSV-1, but little optimisation was 259

reported (Bi et al., 2014; Suenaga et al., 2014).

260

First, a preliminary experiment was done that found co-transfection of the transfer plasmid 261

with a CRISPR/Cas9 construct designed to cleave the HSV genome at the site of 262

recombination greatly improved the frequency of recombinant HSV that can be obtained by 263

transfection-infection (not shown). Next, the impact of two parameters associated with the 264

(12)

12

incorporation of CRISPR/Cas9 plasmids into the strategy were examined 1) the length of the 265

flanking region sequence in the transfer plasmid and 2) the ratio of the CRISPR-Cas9 266

targeting plasmid to the repair plasmid used.

267

To test the first of these, Venus transfer plasmids (as described in Fig. 2A) were transfected 268

into 293A cells such that transfection efficiency was similar along with either pX330-mC (that 269

will cleave mCherry coding sequence) or pX330 (a control with no targeting sequence) in a 270

1:1 ratio. Five hours later, these cells were infected with HSV-1 pCmC at an MOI of 0.01.

271

Virus was harvested after 3 days and used to infect new cultures and the numbers of 272

Venus+, mCherry+ and non-fluorescent plaques were determined by microscopy (Fig 3A).

273

The use of the mCherry-targeting pX330-mC had a dramatic effect, improving the frequency 274

of Venus+ plaques by >100-fold and up to almost a third of all plaques in one case. In the 275

presence of the mCherry targeting plasmid, increasing the length of flanking region 276

sequence made only a marginal difference in two independent experiments.

277

In the previous experiment a substantial proportion of plaques were non-fluorescent, 278

indicating that the genome had been cleaved by CRISPR-Cas9, but was repaired without 279

recombination with the repair plasmid. Therefore, it was reasoned that altering the ratio of 280

the repair plasmid DNA to pX330-mC may increase the frequency of the desired 281

recombinant virus. To test this 293A cells were transfected with 2 µg of the repair plasmid 282

pU3.1kbF-Venus and various amounts of pX330 or pX330-mC to generate molar ratios of 283

4:1, 2:1, 1:1 or 1:2, and then infected with HSV-1 pCmC at an MOI of 0.01. Virus was 284

harvested after 3 days and the proportion of Venus+, mCherry+ and fluorescence negative 285

plaques determined by microscopy of cell monolayers infected with serial dilutions of the 286

progeny from these transfection/infections (Fig. 3B). This experiment further confirmed the 287

large improvement in efficiency associated with CRISPR/Cas9 targeting. Altering the ratio of 288

the CRISPR-Cas9 plasmid to the repair plasmid only had a modest impact on the frequency 289

of fluorescent virus generated and this was repeated in a second experiment.

290

(13)

13

3.3. Foreign genes can be inserted between UL26 and UL27 of HSV-1 without loss of 291

virulence 292

To develop the UL26-UL27 intergenic region as a site that can accept foreign genes 293

available annotations of this region with predicted transcription termination sites were 294

inspected. An insertion position between base pairs 52809 and 52810 (based on the KOS 295

sequence, accession JQ673480) was chosen being roughly equidistant between the full 296

polyA sites for these transcription units (Fig 3A, B). This information was used to design 297

transfer plasmid pUC26/7 into which a cassette containing the ICP47 promoter, TdTomato 298

coding sequence and a BGH polyA signal was inserted (Fig 4C). The transfection/infection 299

method detailed above, without the aid of CRISPR/Cas9 was used to generate recombinant 300

virus. Two TdTomato+ plaques were selected from the progeny of two independent 301

transfection/infections and pure stocks of both were obtained after three rounds of plaque 302

purification. One of these (named HSV-1 pICP47/TdTom) was chosen for further 303

examination and restriction digests of the genome and PCR and DNA sequencing done to 304

confirm its integrity (not shown). This virus was found to have identical replication kinetics 305

compared with the parent KOS in Vero cells (Fig 3D). In addition, HSV-1 pICP47/TdTom 306

also exhibited a normal plaque phenotype (by microscopy) and size (Fig 3E&7; ImageJ, 307

Rasband, 1997-2012). Finally this virus was compared with its parent HSV-1 KOS in a flank 308

model of infection in which virus is introduced by tattoo (Supplemental Fig. S1). The 309

virulence of the HSV-1 pICP47/TdTom was similar to KOS based on observation of lesions 310

(not shown) and virus loads in DRG and skin (Fig 3G).

311 312

4. Discussion 313

This study shows that transfection/infection methods are sufficiently efficient to reliably 314

generate recombinant HSVs where a strong marker for screening, for example a fluorescent 315

protein, is available. In total this method has been used to generate ten viruses using either 316

(14)

14

the UL3-UL4 or the UL26-UL27 sites and expressing a range of fluorescent proteins under 317

the control of several promoters, some of which are published elsewhere (Mackay et al., 318

2013; Macleod et al., 2014). For this approach, transfection efficiency is of key importance, 319

with efficiencies of >20% being required and higher efficiencies being preferable. In addition, 320

increasing flank sequence lengths in transfer plasmids improved the frequency of 321

recombination in a roughly linear manner. However, these improvements need to be 322

weighed against the lower transfection efficiencies typically achieved with larger plasmids.

323

Despite influencing efficiency by up to 10-fold, none of these optimizations improved 324

efficiency to the point that recombinant viruses could be identified by PCR screening in the 325

absence of an additional selectable marker to enrich the desired viruses. By contrast the use 326

of CRISPR/Cas9 targeting dramatically improved the frequency of initial recombination. The 327

data above show in some cases a third of all progeny are recombinant using this method.

328

Several more viruses have been generated using this method and frequencies have been as 329

high as 70% and viruses with small deletions and no markers have been made (not shown).

330

Further, the importance of using transfer plasmids with long homology sequences flanking 331

the insertion site is greatly reduced when CRISPR/Cas9 is used. The availability of 332

CRISPR/Cas9 plasmids in repositories and the relative insensitivity of the methods to 333

changes in protocol such as ratio of plasmid suggest that adoption of this technology will 334

greatly expand the accessibility of recombinant virus generation for HSV-1.

335 336

In terms of developing insertion sites, a position between the polyA signals associated with 337

the UL26 and UL27 transcription units was chosen and a plasmid designed so that no HSV 338

sequence was deleted. It remains unclear why previous attempts to use this region to add 339

genes as led to attenuation (Halford et al., 2004; Orr et al., 2005). However, in the best 340

described case, the insertion disrupted the native polyA signal of UL26, which was then 341

replaced with one from SV40 (Orr et al., 2005). This suggests that all elements associated 342

with transcription in this region cannot be easily replaced or predicted. The design detailed in 343

Figure 4 avoids these problems as shown by the generation of HSV-1 pICP47/TdTom, which 344

(15)

15

had wild type virulence. This establishes a new site that can be used for future recombinant 345

viruses.

346 347

Acknowledgements 348

We wish to thank RSB animal services for husbandry of mice. We thank Francis Carbone 349

(University of Melbourne) for HSV-1 KOS, the National Cancer Institute (NIH) Biological 350

Resources Branch for pIGCN21 and Andrew Lew (Water and Eliza Hall Institute, Melbourne) 351

for pCIGH3. This work was funded by NHMRC Project grant APP1005846 and ARC Future 352

Fellowship FT110100310.

353

References 354

Balliet, J.W., Kushnir, A.S., Schaffer, P.A., 2007. Construction and characterization of a 355

herpes simplex virus type I recombinant expressing green fluorescent protein: Acute 356

phase replication and reactivation in mice. Virol. 361, 372-383.

357

Bi, Y., Sun, L., Gao, D., Ding, C., Li, Z., Li, Y., Cun, W., Li, Q., 2014. High-efficiency targeted 358

editing of large viral genomes by RNA-guided nucleases. PLoS Pathog. 10, 359

e1004090.

360

Blyth, W.A., Harbour, D.A., Hill, T.J., 1984. Pathogenesis of zosteriform spread of Herpes 361

Simplex virus in the mouse. J. Gen. Virol. 65, 1477-1486.

362

Bradley, H., Markowitz, L.E., Gibson, T., McQuillan, G.M., 2014. Seroprevalence of Herpes 363

Simplex Virus Types 1 and 2—United States, 1999–2010. J. Infect. Dis. 209, 325- 364

333.

365

Cong, L., Ran, F.A., Cox, D., Lin, S., Barretto, R., Habib, N., Hsu, P.D., Wu, X., Jiang, W., 366

Marraffini, L.A., Zhang, F., 2013. Multiplex genome engineering using CRISPR/Cas 367

systems. Science 339, 819-823.

368

Cunningham, A.L., Taylor, R., Taylor, J., Marks, C., Shaw, J., Mindel, A., 2006. Prevalence 369

of infection with herpes simplex virus types 1 and 2 in Australia: A nationwide 370

population based survey. Sex Transm. Infect. 82, 164-168.

371

Ding, X., Sanchez, D.J., Shahangian, A., Al-Shyoukh, I., Cheng, G., Ho, C.M., 2012.

372

Cascade search for HSV-1 combinatorial drugs with high antiviral efficacy and low 373

toxicity. Int. J. Nanomed. 7, 2281-92.

374

(16)

16

Gierasch, W.W., Zimmerman, D.L., Ward, S.L., VanHeyningen, T.K., Romine, J.D., Leib, 375

D.A., 2006. Construction and characterization of bacterial artificial chromosomes 376

containing HSV-1 strains 17 and KOS. J. Virol. Met. 135, 197-206.

377

Goins, W.F., Krisky, D.M., Wechuck, J.B., Huang, S., Glorioso, J.C. 2008. Construction and 378

production of recombinant herpes simplex virus vectors, Methods Mol. Biol., Vol. 433, 379

pp. 97-113.

380

Halford, W.P., Balliet, J.W., Gebhardt, B.M., 2004. Re-evaluating natural resistance to 381

herpes simplex virus type 1. J. Virol. 78, 10086-10095.

382

Hogk, I., Kaufmann, M., Finkelmeier, D., Rupp, S., Burger-Kentischer, A., 2013. An in vitro 383

HSV-1 reactivation model containing quiescently infected PC12 cells. Biores Open 384

Access 2, 250-7.

385

Horsburgh, B.C., Hubinette, M.M., Qiang, D., MacDonald, M.L., Tufaro, F., 1999. Allele 386

replacement: an application that permits rapid manipulation of herpes simplex virus 387

type 1 genomes. Gene Ther. 6, 922-30.

388

Lee, E.C., Yu, D., Martinez de Velasco, J., Tessarollo, L., Swing, D.A., Court, D.L., Jenkins, 389

N.A., Copeland, N.G., 2001. A highly efficient Escherichia coli-based chromosome 390

engineering system adapted for recombinogenic targeting and subcloning of BAC 391

DNA. Genomics 73, 56-65.

392

Leland, D.S., Ginocchio, C.C., 2007. Role of cell culture for virus detection in the age of 393

technology. Clin. Microbiol. Rev. 20, 49-78.

394

Mackay, L.K., Rahimpour, A., Ma, J.Z., Collins, N., Stock, A.T., Hafon, M.-L., Vega-Ramos, 395

J., Lauzurica, P., Mueller, S.N., Stefanovic, T., Tscharke, D.C., Heath, W.R., Inouye, 396

M., Carbone, F.R., Gebhardt, T., 2013. The developmental pathway for CD103+CD8+ 397

tissue-resident memory T cells of skin. Nature Immunol. 14, 1294-1301.

398

Mackett, M., Smith, G.L., Moss, B., 1984. General method for production and selection of 399

infectious vaccinia virus recombinants expressing foreign genes. J. Virol. 49, 857-64.

400

Macleod, B.L., Bedoui, S., Hor, J.L., Mueller, S.N., Russell, T.A., Hollett, N.A., Heath, W.R., 401

Tscharke, D.C., Brooks, A.G., Gebhardt, T., 2014. Distinct APC subtypes drive 402

spatially segregated CD4+ and CD8+ T-cell effector activity during skin infection with 403

HSV-1. PLoS Pathog. 10, e1004303.

404

Markert, J.M., Medlock, M.D., Rabkin, S.D., Gillespie, G.Y., Todo, T., Hunter, W.D., Palmer, 405

C.A., Feigenbaum, F., Tornatore, C., Tufaro, F., Martuza, R.L., 2000. Conditionally 406

replicating herpes simplex virus mutant, G207 for the treatment of malignant glioma:

407

results of a phase I trial. Gene Ther. 7, 867-74.

408

Morimoto, T., Arii, J., Akashi, H., Kawaguchi, Y., 2009. Identification of multiple sites suitable 409

for insertion of foreign genes in herpes simplex virus genomes. Microbiol. Immunol.

410

53, 155-161.

411

(17)

17

Orr, M.T., Edelmann, K.H., Vieira, J., Corey, L., Raulet, D.H., Wilson, C.B., 2005. Inhibition 412

of MHC class I is a virulence factor in herpes simplex virus infection of mice. PLoS 413

Pathog. 1, 0062-0071.

414

Proenca, J.T., Coleman, H.M., Connor, V., Winton, D.J., Efstathiou, S., 2008. A historical 415

analysis of herpes simplex virus promoter activation in vivo reveals distinct 416

populations of latently infected neurones. J. Gen. Virol. 89, 2965-2974.

417

Ramachandran, S., Knickelbein, J.E., Ferko, C., Hendricks, R.L., Kinchington, P.R., 2008.

418

Development and pathogenic evaluation of recombinant herpes simplex virus type 1 419

expressing two fluorescent reporter genes from different lytic promoters. Virol. 378, 420

254-264.

421

Rampling, R., Cruickshank, G., Papanastassiou, V., Nicoll, J., Hadley, D., Brennan, D., 422

Petty, R., MacLean, A., Harland, J., McKie, E., Mabbs, R., Brown, M., 2000. Toxicity 423

evaluation of replication-competent herpes simplex virus (ICP 34.5 null mutant 1716) 424

in patients with recurrent malignant glioma. Gene Ther. 7, 859-66.

425

Rasband, W.S. 1997-2012. Image J, U.S. National Institutes of Health, Bethesda, Maryland, 426

USA.

427

Rinaldi, A., Marshall, K.R., Preston, C.M., 1999. A non-cytotoxic herpes simplex virus vector 428

which expresses Cre recombinase directs efficient site specific recombination. Virus 429

Res. 65, 11-20.

430

Sawtell, N.M., Thompson, R.L., 1992. Rapid in vivo reactivation of herpes simplex virus in 431

latently infected murine ganglionic neurons after transient hyperthermia. J. Virol. 66, 432

2150-2156.

433

Shimeld, C., Whiteland, J.L., Williams, N.A., Easty, D.L., Hill, T.J., 1996. Reactivation of 434

herpes simplex virus type 1 in the mouse trigeminal ganglion: An in vivo study of 435

virus antigen and immune cell infiltration. J. Gen. Virol. 77, 2583-2590.

436

Simmons, A., Nash, A.A., 1984. Zosteriform spread of herpes simplex virus as a model of 437

recrudescence and its use to investigate the role of immune cells in prevention of 438

recurrent disease. J. Virol. 52, 816-821.

439

Suenaga, T., Kohyama, M., Hirayasu, K., Arase, H., 2014. Engineering large viral DNA 440

genomes using the CRISPR-Cas9 system. Microbiol. Immunol. 58, 513-522.

441

Summers, B.C., Leib, D.A., 2002. Herpes simplex virus type 1 origins of DNA replication play 442

no role in the regulation of flanking promoters. J. Virol. 76, 7020-7029.

443

Tanaka, M., Kagawa, H., Yamanashi, Y., Sata, T., Kawaguchi, Y., 2003. Construction of an 444

excisable bacterial artificial chromosome containing a full-length infectious clone of 445

herpes simplex virus type 1: viruses reconstituted from the clone exhibit wild-type 446

properties in vitro and in vivo. J. Virol. 77, 1382-1391.

447

(18)

18

Tanaka, M., Kodaira, H., Nishiyama, Y., Sata, T., Kawaguchi, Y., 2004. Construction of 448

recombinant herpes simplex virus type I expressing green fluorescent protein without 449

loss of any viral genes. Microbes Infect. 6, 485-493.

450

Van Lint, A., Ayers, M., Brooks, A.G., Coles, R.M., Heath, W.R., Carbone, F.R., 2004.

451

Herpes simplex virus-specific CD8+ T cells can clear established lytic infections from 452

skin and nerves and can partially limit the early spread of virus after cutaneous 453

inoculation. J. Immunol. 172, 392-397.

454

Wong, Y.C., Lin, L.C.W., Melo-Silva, C.R., Smith, S.A., Tscharke, D.C., 2011. Engineering 455

recombinant poxviruses using a compact GFP-blasticidin resistance fusion gene for 456

selection. J. Virol. Met. 171, 295-298.

457 458

Figure Legends 459

Figure 1. Role of virus multiplicity and transfection efficiency on recombinant HSV 460

generation by transfection/infection. (A) Map of pT pC_eGC indicating the base pair 461

positions of the two flanking regions (in grey), using numbers from HSV-1 KOS (accession 462

JQ673480), other features are as marked. (B) Effect of MOI on virus output of 463

transfection/infections. 293A monolayers were transfected with pT pC_eGC and infected at 464

the MOIs shown 5 hours later. Progeny of these transfection/infections at 72 hrs were used 465

to infect monolayers of Vero cells, and the number of total (open bars) and eGFP+ (black 466

bars) plaques counted. Results are representative of two experiments. (C and D) The effect 467

of transfection efficiency was tested for linearized (C) and intact (D) plasmids. 293A were 468

transfected with pT pC_eGC to achieve a range of efficiencies and infected at an MOI of 469

0.01 5 hours later. Progeny of these transfection/infections were collected at 72 hours to 470

determine the rate of recombinant virus generation. The proportion of eGFP+ plaques is 471

plotted against the transfection efficiency as determined by flow cytometry.

472

Figure 2. Influence of flank sequence length on recombinant HSV generation by 473

transfection/infection. (A) Representative map of plasmids with different lengths of 474

UL3/UL4 flanking sequences. Four different lengths were used as depicted by the concentric 475

(19)

19

grey boxes to generate plasmids pU3.0.5kbF (HSV-1 KOS 11200-12179), pU3.1kbF (HSV-1 476

KOS 10700-12722), pU3.2kbF (HSV-1 KOS 9803-13698) and pU3.3kbF (HSV-1 KOS 8689- 477

14663). Other features are as marked. (B) 293A monolayers were transfected with the each 478

of the plasmids shown in (A) and infected at an MOI of 0.01 5 hours later. Progeny of these 479

transfection/infections was harvested at 72 hours and used to infect monolayers of Vero 480

cells. The percentage of Venus+ plaques of all HSV plaques is shown. Two independent 481

experiments are indicated with markers in grey and black.

482

Figure 3. Targeting the site of insertion using CRISPR-Cas9 has an overriding effect 483

on recombination frequency. (A) 293A monolayers were cotransfected with 2 µg of one of 484

the plasmids shown in Fig. 2A and either pX330 or pX330-mC in a 1:1 ratio, and infected 485

with HSV-1 pCmC at an MOI of 0.01 5 hours later. Progeny of these transfection/infections 486

was harvested at 72 hours and used to infect monolayers of Vero cells. Pie charts show the 487

percentage of Venus+, mCherry+ and non-fluorescent plaques where mCherry was targeted 488

(with pX330-mC) and boxes below are the approximate percent of Venus+ plaques found 489

when the control (pX330) plasmid was used. (B) 293A monolayers were cotransfected with 490

2 µg pU3.1kbF-Venus and the appropriate mass of either pX330 or pX330-mC so the ratio of 491

these plasmids was 4:1, 2:1, 1:1 or 1:2, and infected with HSV-1 pCmC at an MOI of 0.01 5 492

hours later. Progeny of these transfection/infections was harvested at 72 hours and used to 493

infect monolayers of Vero cells. The pie charts and boxes show data as for panel A, nd = not 494

determined. Experiments in A and B were repeated with similar results.

495

Figure 4. Use of UL26-UL27 intergenic region for insertion of foreign DNA into HSV. (A) 496

Schematic representation of the HSV-1 genome with the location of UL26 and UL27 497

indicated. (B) Detail of the insertion of the TdTomato expression cassette in the intergenic 498

space between UL26 and UL27. (C) Map of pU26/7 pICP47/TdTom indicating the base pair 499

positions of the UL26/UL27 flanking regions (in grey), using numbers from HSV-1 KOS 500

(accession JQ673480), other features are as marked. (D)Multiple step growth analysis (MOI 501

(20)

20

0.01) in Vero cells comparing parent HSV-1 KOS (shown in black) and HSV-1 502

pICP47/TdTom (shown in grey). Data are mean±SEM of three replicates. (E and F) Plaques 503

of HSV-1 KOS and pICP47/TdTom on Vero cells under semi-solid media were similar.

504

Morphology (E) is shown by phase contrast microscopy at 100× magnification (scale bar = 505

150μm) and size (F) was measured for 30 plaques of each virus (mean size indicated by the 506

black bar). (G) Amounts of infectious virus in the skin and innervating DRG of C57Bl/6 mice 507

5 days after flank infection with HSV-1 KOS (black) and HSV-1 pICP47/TdTom (grey).

508

Circles show results for each mouse (n=4) and bars represent mean±SEM. (ns = not 509

significant).

510

Figure S1. Pathogenesis of HSV in mice following flank infection by tattoo. C57Bl/6 511

mice were infected with 1 × 108 PFU/mL WT HSV-1 KOS by tattoo. (A) Photographs of a 512

representative mouse were taken at 1, 4, and 7 days after infection. (B) Estimation of total 513

lesion size over time. Circles and bars represent mean±SEM (n=3).

514 515

(21)

0 20 40 60 80 100 0.000

0.001 0.002 0.003

R2=0.883

Transfection efficiency (%)

% eGFP+ Plaques (of total)

0 20 40 60 80 100

0.000 0.001 0.002 0.003

R2=0.825

Transfection efficiency (%)

% eGFP+ Plaques (of total)

pT pC_eGC

UL4 homology arm HSV-1 KOS 11650-12682

CMV IE promoter UL3 homology arm HSV-1 KOS 10534-11649 pUC

origin

AmpR

BGH PolyA eGFP/Cre

A B

C D

Figure 1. Role of virus multiplicity and transfection efficiency on recombinant HSV generation by transfection/infection. (A) Map of pT pC_eGC indicating the base pair positions of the two flanking regions (in grey), using numbers from HSV-1 KOS (accession JQ673480), other features are as marked.

(B) Effect of MOI on virus output of transfection/infections. 293A monolayers were transfected with pT pC_eGC and infected at the MOIs shown 5 hours later. Progeny of these transfection/infections at 72 hrs were used to infect monolayers of Vero cells, and the number of total (open bars) and eGFP+ (black bars) plaques counted. Results are representative of two experiments. (C and D) The effect of transfection efficiency was tested for linearized (C) and intact (D) plasmids. 293A were transfected with pT pC_eGC to achieve a range of efficiencies and infected at an MOI of 0.01 5 hours later. Progeny of these transfection/infections were collected at 72 hours to determine the rate of recombinant virus generation. The proportion of eGFP+ plaques is plotted against the transfection efficiency as determined by flow cytometry.

0.0001 0.001 0.01 0

2 4 6 8

GFP+ PFU

MOI Titre (log10 PFU/mL)

Total PFU

(22)

2 4 6 0.00

0.05 0.10 0.15 0.20

R2=0.93

R2=0.99

Flanking sequence length (kb)

% Venus+ plaques (of total)

pU3.3kbF Venus

UL4 homology

arm

ICP47 promoter UL3 homology

arm KanR

pUC origin

BGH polyA Venus

A

B

11200 10700 9803 8689

12179 12722 13698 14663

Figure 2. Influence of flank sequence length on recombinant HSV generation by transfection/infection. (A) Representative map of plasmids with different lengths of UL3/UL4 flanking sequences. Four different lengths were used as depicted by the concentric grey boxes to generate plasmids pU3.0.5kbF (HSV-1 KOS 11200-12179), pU3.1kbF (HSV-1 KOS 10700-12722), pU3.2kbF (HSV-1 KOS 9803-13698) and pU3.3kbF (HSV-1 KOS 8689-14663). Other features are as marked. (B) 293A monolayers were transfected with the each of the plasmids shown in (A) and infected at an MOI of 0.01 5 hours later. Progeny of these transfection/infections was harvested at 72 hours and used to infect monolayers of Vero cells. The percentage of Venus+ plaques of all HSV plaques is shown. Two independent experiments are indicated with markers in grey and black.

(23)

19

39

42 27

30 43

23

33 44

18

38 44

A

B

Figure 3. Targeting the site of insertion using CRISPR-Cas9 has an overriding effect on recombination frequency. (A) 293A monolayers were cotransfected with 2 µg of one of the plasmids shown in Fig. 2A and either pX330 or pX330-mC in a 1:1 ratio, and infected with HSV-1 pCmC at an MOI of 0.01 5 hours later. Progeny of these

transfection/infections was harvested at 72 hours and used to infect monolayers of Vero cells. Pie charts show the percentage of Venus+, mCherry+ and non-fluorescent plaques where mCherry was targeted (with pX330-mC) and boxes below are the approximate percent of Venus+ plaques found when the control (pX330) plasmid was used. (B) 293A monolayers were cotransfected with 2 µg pU3.1kbF-Venus and the appropriate mass of either pX330 or pX330-mC so the ratio of these plasmids was 4:1, 2:1, 1:1 or 1:2, and infected with HSV-1 pCmC at an MOI of 0.01 5 hours later. Progeny of these

transfection/infections was harvested at 72 hours and used to infect monolayers of Vero cells. The pie charts and boxes show data as for panel A, nd = not determined.

Experiments in A and B were repeated with similar results.

Non fluorescent mCherry+ Venus+

4:1 2:1 1:1 1:2

Ratio of repair:CRISPR plasmid Combined length of flanking sequence (kb)

pX330-mC CRISPR plasmid

pX330

22

34 44

23

33 44

23

35

42 30

39 41

1 2 4 5.6

0.08 0.1 0.1

<0.05

nd nd 0.6%

0.2%

pX330-mC CRISPR plasmid

pX330

(24)

0 24 48 72 2

4 6 8

Time (hpi) Viral Titre (log10 PFU/mL)

0.0 0.1 0.2 0.3

Virus Plaque size (mm2 )

DRG Skin

1 2 3 4 5 6

Titre (log10PFU/mL)

pU26/7 ICP47/TdTom

UL27 homology arm HSV-1 KOS 52810-54154

ICP47 promoter

UL26 homology arm HSV-1 KOS 51431-52809 pUC

origin AmpR

UL26 PolyA

UL27 PolyA

BGH PolyA

Tdtomato

UL26 UL27

ICP47 promoter

UL26.5 UL28

SV40 Terminator

UL27 PolyA UL26 PolyA Tdtomato

UL US

TRL IRL IRS TRS

a’ a’ a’

A

B

D

E C

HSV-1 KOS

HSV-1 pICP47/TdTom

F

ns

Figure 4. Use of UL26-UL27 intergenic region for insertion of foreign DNA into HSV. (A) Schematic representation of the HSV-1 genome with the location of UL26 and UL27 indicated. (B) Detail of the insertion of the TdTomato expression cassette in the intergenic space between UL26 and UL27. (C) Map of pU26/7 pICP47/TdTom indicating the base pair positions of the UL26/UL27 flanking regions (in grey), using numbers from HSV-1 KOS (accession JQ673480), other features are as marked. (D)Multiple step growth analysis (MOI 0.01) in Vero cells comparing parent HSV-1 KOS (shown in black) and HSV-1 pICP47/TdTom (shown in grey). Data are mean±SEM of three replicates. (E and F) Plaques of HSV-1 KOS and pICP47/TdTom on Vero cells under semi-solid media were similar. Morphology (E) is shown by phase contrast microscopy at 100× magnification (scale bar = 150μm) and size (F) was measured for 30 plaques of each virus (mean size indicated by the black bar). (G) Amounts of infectious virus in the skin and innervating DRG of C57Bl/6 mice 5 days after flank infection with HSV-1 KOS (black) and HSV-1 pICP47/TdTom (grey). Circles show results for each mouse (n=4) and bars represent mean±SEM. (ns = not significant).

KOS pICP47Td/Tom

G

ns

ns

(25)

0 2 4 6 8 10 12 0.0

0.5 1.0 1.5

Time (dpi) Lesion area (cm2 )

B

Figure S1. Pathogenesis of HSV in mice following flank infection by tattoo. C57Bl/6 mice were infected with 1 x 108 PFU/mL WT HSV-1 KOS by tattoo. (A) Photographs of a representative mouse were taken at 1, 4, and 7 days after infection. (B) Estimation of total lesion size over time. Circles and bars represent mean±SEM (n=3).

A

4 dpi 7 dpi 1 dpi

References

Related documents

Our investigation considered the accessibility of ACT Policing’s complaints management framework for Aboriginal and Torres Strait Islander peoples, the policies and procedures in place