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