AEM Accepted Manuscript Posted Online 20 February 2015 Appl. Environ. Microbiol. doi:10.1128/AEM.03847-14 Copyright © 2015, American Society for Microbiology. All Rights Reserved.
1
Oxidative stress correlates with Wolbachia-mediated antiviral protection
2
in Wolbachia-Drosophila associations
3 4
Zhee Sheen Wonga, Jeremy C. Brownlieb, Karyn N. Johnsona#
5 6
School of Biological Sciences, The University of Queensland, Brisbane
7
Queensland Australiaa, School of Natural Sciences, Griffith University,
8
Brisbane Queensland, Australiab
9 10
Running Title: Oxidative stress and Wolbachia antiviral protection
11 12
#
Address correspondence to Karyn N. Johnson,
[email protected]
1
13
Abstract
14
Wolbachia mediates antiviral protection in insect hosts and is being developed
15
as a potential biocontrol agent to reduce the spread of insect-vectored
16
viruses. Definition of the molecular mechanism that generates protection is
17
important for understanding the tripartite interaction between host insect,
18
Wolbachia and virus. Elevated oxidative stress was previously reported for a
19
mosquito line experimentally infected with Wolbachia, indicating that oxidative
20
stress may be important for Wolbachia-mediated antiviral protection.
21
However, Wolbachia experimentally introduced into mosquitoes impacts a
22
range of host fitness traits some of which are unrelated to antiviral protection.
23
To explore whether elevated oxidative stress is associated with antiviral
24
protection in Wolbachia-infected insects we analysed oxidative stress of five
25
Wolbachia-infected Drosophila lines. In flies infected with protective
26
Wolbachia strains hydrogen peroxide concentrations were 1.25- to 2-fold
27
higher than in paired fly lines cured of Wolbachia infection. In contrast, there
28
was no difference in the hydrogen peroxide concentrations in flies infected
29
with non-protective Wolbachia strains compared to flies cured of Wolbachia
30
infection. Using a Drosophila mutant that produces increased levels of
31
hydrogen peroxide we investigated whether flies with high endogenous
32
reactive oxygen species had altered response to virus infection and found
33
flies with high endogenous levels of hydrogen peroxide were less susceptible
34
to virus-induced mortality. Taken together, these results suggest that elevated
35
oxidative stress correlates with Wolbachia-mediated antiviral protection in
36
natural Drosophila hosts.
2
37
Introduction
38
The
39
Alphaproteobacterium predicted to infect at least 40% of insect species (1-3).
40
Best known for its ability to invade invertebrate populations via modification of
41
host reproductive systems (4), some Wolbachia infected insects are protected
42
from viruses and other pathogens
43
infection (13-17). Due to this ability to disrupt virus infection, there is an
44
increased interest in employing Wolbachia as a means of biological control of
45
arthropod transmitted infectious diseases, such as Dengue virus (18). Despite
46
this interest, the mechanisms of Wolbachia antiviral protection remain to be
47
fully elucidated.
48
Several studies have provided insight into Wolbachia-mediated antiviral
49
protection. Wolbachia can mediate broad protection against a range of
50
different RNA viruses in both Drosophila and mosquitoes (5-12). Wolbachia-
51
infected Drosophila are concomitantly protected against two diverse viruses
52
Drosophila C virus (DCV, Dicistroviridae) and Flock House virus (FHV,
53
Nodaviridae) and the protection is strongly genetically correlated (6, 11, 12,
54
19). Taken together, this suggests that Wolbachia mediates antiviral
55
protection through a mechanism with broad specificity. Interestingly,
56
Wolbachia infection often interferes with virus accumulation but examples
57
have been noted where insects accumulate high titres of virus but are
58
protected from virus-induced mortality, indicating that Wolbachia may impact
59
both viral resistance and tolerance (11, 12).
60
Many different strains of Wolbachia infect Drosophila, but not all mediate
61
antiviral protection. In Drosophila simulans a strong correlation between
maternally
inherited
endosymbiont
Wolbachia
pipientis
is
an
(5-12), while others have enhanced
3
62
Wolbachia density and protection was noted. Using natural host-Wolbachia
63
pairings antiviral protection was mediated by wAu and wRi, which occurred at
64
high density in CO and DSR fly lines respectively (11). In contrast, Wolbachia
65
strains wHa and wNo are found in low density in their natural host lines DSH
66
and N7NO and no protection was observed. Furthermore, treatment of CO
67
flies with antibiotics to decrease wAu density results in loss of protection (20).
68
This role of Wolbachia density in protection is supported in Drosophila where
69
a correlation has been shown between the strength of protection and
70
Wolbachia density in consistent host backgrounds (19, 21). Thus it is clear
71
that high density is important for Wolbachia mediated antiviral protection.
72
Recently a model linking induction of oxidative stress in experimentally
73
infected Ae. aegypti mosquitoes with the protection against Dengue virus was
74
described (22). Stimulation of excess reactive oxygen species (ROS) has
75
previously been reported in Drosophila infected by Wolbachia (23).
76
Oxidative stress is the imbalance between the production of ROS and
77
antioxidant defenses. ROS are derived from Superoxide (O2 -.) generated by
78
one electron reduction of oxygen; and Superoxide is converted to H2O2 by
79
superoxide dismutase (SODs). Under normal conditions, ROS are primarily
80
produced by the mitochondrial respiratory chain during the intermediate state
81
of reducing molecular oxygen to water (24). In Drosophila during microbial
82
infection, ROS are produced by dual oxidases (Duox) in the mid gut (25, 26).
83
Albeit important in combating microbial infections, high level of ROS is
84
detrimental to the host as it creates a state of oxidative stress (27). ROS
85
cause damage to lipids, nucleic acids, proteins and a reduction in the insect
86
life span (27). To combat this damage, Drosophila has developed a complex
4
87
antioxidant system including superoxide dismutases (SOD) and catalases (28)
88
to balance the damaging effects of ROS.
89
Although the specific molecular target of ROS is still unknown, mitochondrial
90
ROS (mROS) is known to be involved in a range of physiological systems,
91
including immunity regulation. mROS are induced in response to cellular
92
stress to alter signaling pathways as an adaptation to cellular stress (29).
93
Elevation of oxidative stress in response to bacterial infections in insects and
94
mammals is well established (30, 31). In response to viral infections, oxidative
95
stress is induced in lepidoptera and is correlated with cell death (32). In
96
humans, increased concentration of ROS as a result of virus infections is also
97
observed (33, 34).
98
As oxidative stress is implicated with viral infections in several models, and in
99
mosquitoes correlates with Wolbachia-mediated antiviral protection, we
100
hypothesise that elevated ROS leading to oxidative stress/cell signalling is
101
involved in Wolbachia-mediated antiviral responses in natural Wolbachia-
102
Drosophila associations. To investigate this, we analysed whether protective
103
and non-protective Wolbachia strains induce ROS relative to Wolbachia-free
104
controls and the impact of high endogenous ROS on resistance of flies
105
against virus infection and on viral replication. Taken together our results we
106
demonstrate that increased ROS and oxidative stress correlates with
107
Wolbachia-mediated antiviral protection.
5
108
Materials and Methods
109 110
Flies, Virus and Wolbachia
111 112
All fly lines were maintained on standard cornmeal diet at 25°C with 12-hours
113
light/dark cycle. The D. melanogaster Oregon RC (ORC), D. simulans DSR,
114
CO, N7No and DSH were naturally infected with Wolbachia strain wMelCS
115
(35), wRi (36), wAu (37), wNo (38) and wHa (39) respectively. Paired fly lines
116
cured of Wolbachia infection were generated by treating the flies with 0.03%
117
of tetracycline (40) and flies were maintained for 12 months before use. Gut
118
flora was reconstituted and normalized across fly-lines described below, using
119
standardized methods (21); all experiments were conducted a minimum of
120
seven generations post tetracycline treatment (41). The D. melanogaster line
121
24492 (Bloomington Drosophila Stock Centre) is a Cu/Zn SOD-null mutant
122
and has a heterozygous missense mutation at the n108 position which
123
disrupts the hydrogen bonds formed across the dimer interface of the Cu/ZN
124
SOD. Consequently, 24492 flies have lower SOD activity compared to wild
125
type (42) and hence higher endogenous oxidative stress compared to wild
126
type flies. The 24492 mutant line was created from D. melanogaster Canton S
127
(CS) background and thus CS flies were used as the wild type control for
128
24492 flies. CS and 24492 flies were confirmed to be free of Wolbachia
129
infection.
130
The Drosophila C virus isolate EB was purified as previously described (6,
131
43), suspended in Tris pH 7.4 and maintained at -20 °C in aliquots. The virus
132
titre was determined using a tissue culture infectious dose (TCID50) assay as
6
133
previously described (44). DCV aliquots were used once and then discarded
134
to prevent loss of virus infectivity through repeated freeze-thawing.
135 136
Analysis of hydrogen peroxide concentration
137 138
Male 4 to 7 day old flies were collected in chilled PBS and homogenised with
139
pestles. Follow homogenisation, fly debris was pelleted by centrifugation for 1
140
min at 12,000 x g and supernatants were collected. H2O2 concentration in the
141
supernatants
142
Peroxide/Peroxidase Assay Kit (Invitrogen) as per the manufacturer’s
143
protocol. Absorbance of the oxidised Amplex® Red reagent was detected at
144
560 nm using an absorbance microplate reader (Epoch). Total protein was
145
measured in each fly homogenate using the Bio-Rad Protein Assay (Bio-Rad)
146
as per manufacturer’s protocol and the concentration of H2O2 was normalised
147
to the concentration of total proteins in the sample. The difference between
148
the two groups of flies was determined using student’s t test and F test was
149
used to determine the variance (GraphPad Prism).
was
analysed
with
the
Amplex®
Red
Hydrogen
150 151 152
Survival bioassays
153 154
To analyse the susceptibility of flies to viral induced mortality, 4-7 day old
155
male flies were challenged with DCV. Flies were anaesthetised with CO2 and
156
virus was injected into the upper lateral part of the fly abdomen using needles
157
pulled from borosilicate glass capillaries and a pulse pressure mico-injector
7
158
(Drummond). For each experiment a fresh aliquot of fresh DCV was
159
defrosted, diluted to a concentration of 1x108 IU/ml in PBS and 52.6 nl was
160
injected into each fly.
161 162
For each fly line assayed, three vials of 15 flies were injected with DCV and
163
one vial of 10 flies were mock infected with PBS. Following challenge, flies
164
were maintained in a 25°C incubator and survival of the flies was scored
165
every day until mortality in the virus-infected flies reached 100%. Mortality
166
within the first 24 hours was deemed to be due to needle injury and these flies
167
were removed from the survival analysis. At least three independent survival
168
bioassays were done for each fly line. Survival curves of the flies in each
169
experiment were compared using Kaplan-Meier analysis and log-rank
170
statistics with GraphPad Prism.
171 172
Virus accumulation assays
173 174
Virus accumulation in the CS and 24492 flies were analysed using a TCID50
175
assay. For each fly line, groups of flies were injected with PBS or DCV as for
176
survival bioassays. At day 0 and 2 days post infection, 3 flies were collected
177
and frozen at -20°C. Each pool of 3 flies was homogenised in 100 μl of PBS
178
with two 3 mm beads (Sigma-Aldrich) using TissueLyser II (Qiagen) for 90
179
seconds with the frequency of 30 shakes/s. The homogenate was centrifuged
180
at 14,000 rpm for 8 minutes to pellet the fly debris. Virus titre was analysed
181
using the TCID50 assay as previously described (6). The geometric means of
8
182
the duplicate sample between the CS and 24492 flies was analysed using
183
student’s t tests (GraphPad Prism).
9
184
Results
185 186
Drosophila infected with protective Wolbachia have elevated hydrogen
187
peroxide concentrations
188 189
To investigate whether the presence of protective, but not the non-protective
190
Wolbachia strains influenced the regulation of ROS in Drosophila, the
191
concentration of H2O2 in D. melanogaster ORC, D. simulans DSR, CO, N7No
192
and DSH infected by wMelCS, wRi, wAu, wNo and wHa respectively was
193
analysed (Figure 1). The concentration of H2O2 in flies that harbored a
194
protective strain (wMelCS, wAu or wRi) was increased 1.25-2 fold relative to
195
Wolbachia-free controls (Figure 1 A, B and C; student’s t test: A, *p
