Cholesterol flux is required for endosomal progression of African ...

4 downloads 144 Views 8MB Size Report
Nov 25, 2015 - Present address for RM-M: c Department of Microbiology, Icahn School of Medicine at Mount Sinai, New. 19. York, USA and d Global Health ...
JVI Accepted Manuscript Posted Online 25 November 2015 J. Virol. doi:10.1128/JVI.02694-15 Copyright © 2015, American Society for Microbiology. All Rights Reserved.

1 2

Cholesterol flux is required for endosomal progression of African swine fever virions during the initial establishment of infection

3 4 5 6 7 8

Cuesta-Geijo, Miguel Ángela, Michele Chiappib, Inmaculada Galindoa, Lucía BarradoGila, Raquel Muñoz-Morenoa,c, d, José L. Carrascosab and Covadonga Alonsoa*.

9 a

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Department of Biotechnology, Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria, INIA, Ctra. de la Coruña Km 7.5, 28040 Madrid, Spain. b Department of Structure of Macromolecules, Centro Nacional de Biotecnología (CNB-CSIC), Universidad Autónoma de Madrid, 28049 Madrid, Spain *[email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected]; [email protected];

25 26

ABSTRACT

27

spreading in Eastern Europe since its first appearance in the Caucasus in 2007. ASFV enters

28

the cell by endocytosis and gains access to the cytosol to start replication from late endosomes

29

and multivesicular bodies. Cholesterol associated with low-density lipoproteins entering the

30

cell by endocytosis also follows a similar trafficking pathway as ASFV. Here we show that

31

cholesterol plays an essential role in the establishment of infection as the virus traffics through

32

the endocytic pathway. In contrast to other DNA viruses, such as Vaccinia or Adenovirus 5,

33

cholesterol efflux from endosomes is required for ASFV release/entry to the cytosol.

34

Accumulation of cholesterol in endosomes impairs fusion resulting in retention of virions

35

inside endosomes. ASFV also remodels intracellular cholesterol by increasing its cellular uptake

36

and redistributes free cholesterol to viral replication sites. Our analysis reveals that ASFV

*Corresponding author: [email protected] Present address for RM-M: c Department of Microbiology, Icahn School of Medicine at Mount Sinai, New York, USA and d Global Health and Emerging Pathogens Institute, Icahn School of Medicine at Mount Sinai, New York, USA.

Running title: Cholesterol flux in early ASFV infection

African swine fever virus (ASFV) is a major threat for porcine production that is slowly

1

37

manipulates cholesterol dynamics to ensure an appropriate lipid flux to establish productive

38

infection.

39 40

IMPORTANCE

41 42

Since its appearance in the Caucasus in 2007, African swine fever (ASF) has been

43

spreading westwards to neighboring European countries threatening porcine

44

production. Due to the lack of an effective vaccine, ASF control relies on early

45

diagnosis and widespread culling of infected animals. We investigated early stages of

46

ASFV infection to identify potential cellular targets for therapeutic intervention against

47

ASF. The virus enters the cell by endocytosis and soon thereafter, viral decapsidation

48

occurs in the acid pH of late endosomes. We found that ASFV infection requires and

49

reorganizes the cellular lipid cholesterol. ASFV requires cholesterol to exit the

50

endosome to gain access to the cytoplasm to establish a productive replication. Our

51

results indicate that there is a differential requirement for cholesterol efflux for

52

Vaccinia or Adenovirus 5 compared to ASFV.

53 54 55

INTRODUCTION

56 57

Endocytosis is a common uptake pathway for nutrients, lipids and receptors that is

58

frequently hijacked by viruses to gain entry into cells (49). Cargoes taken up by

59

endocytosis initially converge on early endosomes (EE) from where they can be

60

recycled back to the plasma membrane (28). However, more usually, internalized

61

cargo is transported to late endosomes (LE), and from where it can be directed to

62

lysosomes (LY) for degradation. Endosomal maturation from EE to lysosomes is a

63

dynamic process that involves a progressive reduction in intralumenal pH (28). At the 2

64

start of their maturation, invaginations of the limiting membrane into the lumen of the

65

EE give rise to the intraluminal vesicles (ILV), leading to the formation of multivesicular

66

bodies (MVB), in which the pH is proportional to the number of ILV (28). MVB mature

67

to late endosomes ending in the lysosomes (22). Both MVB and LE are late endosomal

68

compartments. The ordered maturation of early endosomes to lysosomes depends on

69

endosomal membrane signalling that is regulated by both proteins and lipids including

70

cholesterol (6, 28).

71 72

Cholesterol enters the cell by endocytosis in the form of low-density lipoproteins (LDL)

73

and accumulates in the intraluminal vesicles (ILV) in the multivesicular bodies and late

74

endosomes (60). Those ILV that accumulate cholesterol are also enriched in the long-

75

lived lysobisphosphatidic acid (LBPA) (14). Cholesterol must be redistributed to the

76

endosome limiting membrane so that it becomes available for export to other cellular

77

destinations (29). LBPA regulates cholesterol efflux by a mechanism known as

78

“backfusion”, in which the ILV fuses with the endosomal limiting membrane (14, 41).

79

Backfusion ensures that cholesterol in ILV escapes lysosomal degradation by exiting

80

the LE. This cholesterol efflux is controlled by a number of lipid transporters, including

81

NPC1 and NPC2 proteins (31). Mutations in these proteins leads to the hereditary

82

Niemann-Pick C (NPC) disease, characterized by a severe defect in the export of

83

cholesterol from late endosomes (55). The abnormal accumulation of cholesterol in LE

84

leads to altered protein and lipid trafficking that results in a progressive

85

neurodegenerative disorder (34). This disease phenotype can be mimicked by treating

86

cells with U18666A, an amphipatic steroid that blocks the exit of free cholesterol from

3

87

the late endosomal compartment resulting in aggregation of endosomes filled with

88

cholesterol and other lipids (43).

89 90

The endocytic pathway ensures a highly dynamic and ordered sorting of cargoes to

91

their correct cellular locations. Not surprisingly, many viruses exploit endocytosis to

92

facilitate their uptake and ensure they reach the correct site for replication or gain

93

access to the cytoplasm (21, 46, 49). For example, African swine fever virus (ASFV) is a

94

large double stranded DNA virus that enters the cell by dynamin, clathrin and

95

cholesterol-dependent endocytosis (19, 26). Actin-dependent macropinocytosis also

96

assists ASFV entry into cells (58). Once inside the endosome, viral decapsidation

97

occurs in the acid pH of LE within few minutes after entry (13). After uncoating, the

98

virus needs to exit the endosome and gain access to the cytoplasm to start replication

99

in a specialized structure called the viral factory (53). The viral factory is devoid of a

100

limiting membrane and is organized in a single perinuclear site near the microtubule-

101

organizing center (3, 51). After completing morphogenesis in the viral factory, newly

102

assembled virions are moved to the plasma membrane on microtubules by kinesin-1,

103

where they acquire an additional lipid membrane when they exit the cell by budding

104

(32, 57). In this study, we set out to investigate the relevance of cholesterol

105

homeostasis in ASFV entry and replication. Our results clearly demonstrate that

106

cholesterol efflux and endosomal trafficking play essential roles during the

107

establishment of ASFV replication.

108 109

MATERIAL AND METHODS

110

Cells, virus and infections 4

111

Vero cells were mock-infected or infected with ASFV Ba71V isolate as previously

112

described (18, 63) at a multiplicity of infection (moi) of 1, 10 (early time point

113

experiments) or 100 (for TEM) pfu/ml was used. When synchronization of infection

114

was required, virus adsorption was performed for 90 min at 4°C and after a cold PBS

115

wash to remove unbound virus, cells were rapidly shifted to 37°C with fresh pre-

116

warmed media. Virus stocks and/or infective ASFV production yields from infected

117

samples were titrated by plaque assay in Vero cells as previously described (18).

118

Recombinant B54GFP is a recombinant ASFV expressing green fluorescent protein as a

119

fusion protein of viral p54 (25). Infections with B54GFP were performed at a

120

multiplicity of infection (moi) of 1 pfu/cell. Vaccinia virus (VV) recombinant vtag2GFP

121

contained the tag2GFP under the control of a strong synthetic VV early/late promoter

122

(7) was kindly provided by Dr. R. Blasco (INIA, Spain). VV infections were performed at

123

a moi of 1 pfu/cell for 16 hpi. Adenovirus 5/attB (Adv) was provided by Dr. Carmen San

124

Martín (CNB, CSIC). This recombinant virus carries a packaging modification that

125

lengthens and viral cycle and GFP to facilitate detection (1, 2) (Alba et al, 2007; Alba et

126

al, 2011). Vero cells were infected with Adv at a moi of 1 pfu/cell for 16 hpi.

127 128

Detection and quantitation of the ASFV genome

129

The quantitation of the number of copies of ASFV genome was achieved by

130

quantitative real-time PCR (qPCR) as described (19). The qPCR assay used fluorescent

131

hybridization probes to amplify a region of the p72 viral gene as described (36). Each

132

sample was included in triplicates and values were normalized to the standard positive

133

controls. Reactions were performed using the ABI 7500 Fast Real-Time PCR System

134

(Applied Biosystems). 5

135 136

Antibodies

137

Monoclonal antibodies against the virus major capsid protein p72 (Ingenasa) and early

138

protein p30 were used at 1:1000 and 1:100 dilutions respectively (24). ASFV p72 is a

139

very abundant late protein in infection and is accumulated in the viral factory (11).

140

EE were labeled with anti-mouse EEA1 antibody (BD Biosciences Pharmingen), and LE

141

with anti-rabbit Rab7 (Cell Signalling) both at 1:50 dilution. MVB were labeled with

142

anti-CD63 (Developmental Studies Hybridoma Bank, University of Iowa, clone H5C6), a

143

tetraspannin characteristic of this compartment, at 1:200 dilution. LY were labeled

144

with anti-Lamp1 (Abcam) at 1:50 dilution. Antibody 6C4 against lysobisphosphatidic

145

acid (LBPA), enriched in intraluminal vesicles of multivesicular bodies was used at a

146

concentration of 50 μg/mL (kindly provided by Jean Gruenberg). Anti-mouse

147

immunoglobulin G (IgG) antibody conjugated to Alexa Fluor 594 and anti-rabbit IgG

148

antibody conjugated to Alexa Fluor 488 (Molecular Probes) were used as secondary

149

antibodies diluted 1:200.

150 151

Protein detection by Western Blot

152

Protein extracts were electrophoresed in 12% acrylamide–bisacrylamide gels.

153

Separated proteins were transferred to nitrocellulose membranes and detected with

154

corresponding antibodies. As secondary antibody, anti-mouse IgG (GE Healthcare) or

155

anti-rabbit IgG (Bio-Rad) conjugated to horseradish peroxidase was used at a 1:5000

156

dilution. β-tubulin (Sigma) was used as a load control in WB analysis. Finally, bands

157

obtained after development with ECL reagent were detected on Molecular Imager

6

158

Chemidoc XRSplus Imaging System. Bands were quantified by densitometry and data

159

normalized to control values using Image lab software (Bio-Rad).

160 161

Indirect immunofluorescence, conventional and confocal microscopy

162

Immunofluorescence experiments were performed as previously described (13). Vero

163

cells were fixed in PBS 4% paraformaldehyde (PFA) for 15 min and permeabilized with

164

PBS–0.1% Triton X-100 or Saponin (Sigma) for 10 min. Then, cells were incubated with

165

50 mM NH4Cl in PBS for 10 min. After blocking with Albumin bovine serum (BSA)

166

(Sigma) or normal goat serum (Sigma), cells were incubated with corresponding

167

antibodies and nucleus were stained with Topro3 (Molecular Probes) before mounting.

168

Confocal microscopy was carried out in a Leica TCS SPE confocal microscope.

169

Conventional fluorescence microscopy to analyze cholesterol staining with Filipin was

170

carried out using a Leica DM RB microscope. Image analyses were performed with

171

Leica Application Suite advanced fluorescence software (LAS AF) and ImageJ software.

172 173

Transmission Electron Microscopy

174

Vero cells were pretreated with U18666A for 16 h or with DMSO. Then, cells were

175

infected with Ba71V at a moi of 100 pfu/cell. Virus adsorption was carried out by

176

spinoculation as previously described (10) and then placed at 37ºC for 30 or 45 mpi.

177

Then, cells were washed with PBS and fixed with a mixture of 2% PFA and 2.5%

178

glutaraldehyde in PBS for 1 h. The cell monolayer on coverslips was post-fixed with 1%

179

osmium tetroxide in PBS (45 min), treated with 1% aqueous uranyl acetate (45 min),

180

dehydrated with ethanol and embedded in epoxy resin 812 (TAAB) for 2 days at RT as

181

described (8). Ultrathin, 70-nm-thick sections were obtained with the Ultracut UCT 7

182

ultramicrotome (Leica Microsystems), transferred to Nickel EM grids (Gilder) and

183

stained with 3% aqueous uranyl acetate for 20 min and lead citrate for 2 min. Sections

184

were examined on a JEOL JEM 1200 EXII electron microscope (operating at 100Kv).

185 186

Flow Cytometry

187

At 6 hpi cells were washed with PBS and harvested by trypsinization. Flow cytometry

188

experiments were performed by detection of infected cells with an anti-p30

189

monoclonal antibody as described elsewhere (13) or by GFP expression at 16 hpi when

190

using recombinant viruses. 104 cells per tube in triplicates were scored and analyzed in

191

a FACSCalibur flow cytometer (BD Sciences) to determine the percentage of infected

192

cells under these conditions. The obtained infection rates were normalized to the

193

corresponding controls.

194 195

Cell treatment with U18666A

196

U18666A (U; Sigma) is an amphipathic steroid which is a widely used chemical to block

197

the intracellular trafficking of cholesterol and mimic Niemann-Pick type C disease

198

phenotype (37, 43). Cytotoxicity was tested using CellTiter 96 non-radioactive Cell

199

Proliferation assay (Promega) following the manufacturer´s instructions. Vero cells

200

were pretreated for 16 h with DMSO or 5 or 10 µM U18666A and then infected with

201

the different viruses tested at a moi of 1 pfu/cell without washing.

202 203

Cholesterol quantitation and determination

204

Cholesterol was determined and quantified enzymatically using Amplex Red

205

Cholesterol Assay Kit, following manufacturer's instructions (Invitrogen). Fluorescence 8

206

values at an excitation 532 nm and emission 590 nm were counted in a Genios

207

Spectrafluor microplate reader (Tecan) normalized to control sample values and

208

expressed in percentages.

209 210

Cholesterol depleted medium, cholesterol enriched medium and cell treatment

211

Vero cells were grown in cholesterol depleted medium (CHDM) containing DMEM

212

(P/S+G), 5% Lipoprotein deficient serum (LPDS; Sigma), 10 µM mevastatin (Santa Cruz

213

Biotechnology), 50 µM mevalonate (Sigma) and 5 mM Methyl-β-cyclodextrin (MβCDX;

214

Sigma). Chemicals were previously tested for cytotoxicity at the concentrations used.

215

Cells were incubated in CHDM for 2 h before infection, at the time of infection (0 hpi)

216

or 2 and 4 h after infection (2 and 4 hpi). To avoid cytotoxicity of MβCDX (23), after 2 h,

217

CHDM was shifted to CHDM lacking MβCDX (CHDMminus-) and maintained up to 6 or

218

16 hpi. Cell extracts were analyzed for early p30 expression at 6 hpi and for viral

219

replication using qPCR at 16 hpi. Viral infectivity was evaluated by GFP expression of

220

recombinant ASFV, VV and Adv at 16 hpi in cells incubated in CHDM 2 h before

221

infection.

222

Cholesterol enriched medium (CHEM) was prepared as described (15, 47). Briefly, 10

223

mM Cholesterol (CH; Santa Cruz Biotechnology) stock concentration was prepared in

224

stirred and warmed DMSO at 80ºC on a heating block. When the solution clarified was

225

filtered and added in increasing concentrations (0, 100 and 500 µM) to DMEM (P/S+G)

226

containing 5 mM MβCDX to form CH/MβCDX complexes. Vero cells were incubated

227

with CHEM for 2 h, medium was removed and cells washed for 6 times to clear

228

remnant MβCDX. After washing, cells were infected at a moi of 1 pfu/cell and

229

maintained in DMEM (PS+G) containing 2% LPDS. At 2 hpi, medium was replaced with 9

230

DMEM 2% FCS (normal infection medium) and infection progressed for 16 hpi. Finally,

231

cell extracts were processed for qPCR.

232 233

Cholesterol staining

234

To detect free intracellular cholesterol we used fluorescent Filipin (Sigma) as

235

previously described (4, 33, 38, 39). Filipin signal was recorded using a 390–415 nm

236

wavelength excitation filter and with a 450–470-nm wavelength emission filter. For

237

living cells, we used fluorescently labeled Bodipy-cholesterol (TopFluor-Cholesterol,

238

Avanti Polar Lipids), a cholesterol live cell mimic. Vero cells were loaded with 5 µg/ml

239

of bodipy-cholesterol (Bodipy-CH) in DMEM 2% FBS for 30 min at 37°C. Then, cells

240

were infected at a moi of 5 pfu/cell for 1 or 3 h in order to analyze its uptake by

241

endocytosis. Then, cells were quickly rinsed with PBS and fixed with 4% PFA before

242

confocal imaging. Intracellular fluorescence measurement of Bodipy-CH from mock- or

243

ASFV infected cells was performed using ImageJ software.

244 245

Statistical analysis

246

Differences between groups were analyzed by Bonferroni test with GraphPad Prism

247

software and Instat software. All experiments were performed more than two times,

248

and data are presented as mean values and standard deviations (mean±SD) of

249

independent experiments. Metrics were normalized to control values. Asterisks denote

250

statistically significant differences (***P