Background
The type I interferon system is the first line of defense against virus infections [
1‐
5]. During viral replication, RNA intermediates are generated and recognized by distinct host pattern recognition receptors (PRRs). Two classes of PRRs, endosomal toll-like receptors (TLR) and cytoplasmic retinoid-inducible gene I (RIG-I)-like receptors (RLRs), are essential for the detection and protection against viral infection. However, the specific PRRs involved in mediating antiviral response are likely to be virus and cell type-specific [
6‐
8]. Endosomal TLR3 and TLR7 recognize viral double-stranded RNA (dsRNA) and single-stranded DNA (ssRNA) and signal through their adaptor proteins TRIF (TLR3) and Myd88 (TLR7) to induce type I interferons (IFNs) via transcription factors IRF3 and IRF7, respectively [
9]. RLRs such as RIG-I and melanoma differentiation antigen 5 (MDA5) are double stranded RNA binding DExD/H box RNA helicases [
10,
11]. After the recognition of viral RNA, they interact with mitochondrial associated adaptor protein interferon -β promoter stimulator 1 (IPS-1; also known as MAVS, VISA, or CARDIF) [
12‐
15]. This interaction leads to downstream signaling and activation of transcription factors IRF3 and NF-κB, which translocate to the nucleus and activate type I IFN and proinflammatory genes [
16,
17].
Type I IFNs are locally produced in the central nervous system (CNS) upon viral infection. The producing cell types differ from parenchymal cells upon influenza and La Crosse virus to astrocytes upon vesicular stomatitis virus (VSV) infection [
18‐
20]. However, very little is known about the importance and the relative contribution of the different PRRs in the different regions of the CNS for protection against neurotropic viruses.
Tick-borne encephalitis virus (TBEV) is one of the emerging arthropod-borne viral (arboviral) diseases in Europe and Asia. It is a neurotropic-enveloped RNA virus of the
Flaviviridae family and closely related to other human pathogens such as yellow fever virus, dengue virus, West Nile virus (WNV), Japanese encephalitis virus (JEV), and Murray Valley encephalitis virus [
21]. The virus is transmitted via tick bites or consumption of infected milk [
22,
23]. The infection is characterized by a biphasic course in the primary phase, patients show signs of fatigue, headache, pain, and fever, followed by a second phase of neurological involvement with symptoms of encephalitis, meningitis, and paralysis. The mortality rate ranges from 0.5 to 30 %, with 30 to 60 % of survivors developing neurological sequelae [
24‐
27]. Although an effective vaccine is available, treatment options are limited to supportive care [
28]. Given such high medical importance along with the increasing number of cases, and expansion into new unaffected areas, many aspects of TBEV pathogenesis and immunity are still unclear. Langat virus (LGTV) is a naturally attenuated member of TBEV serogroup [
29]. LGTV shares close molecular relationship (82–88 % amino acid identity) with TBEV and thus make it an ideal surrogate model to study TBEV pathogenesis [
24].
We have recently shown that type I IFNs protect and control TBEV- and LGTV-induced inflammation and encephalitis by limiting systemic LGTV replication, spreading to CNS and the associated immunopathology [
30]. However, the cell types and specific PRRs involved in IFN induction and clearance of LGTV are not known.
In the current study, we have determined the relative contribution of IPS-1 signaling in LGTV disease development and protection. We show that the tropism of the virus in the CNS is shaped by the IFN response, and that IPS-1 signaling is very important for IFN-β upregulation in the olfactory bulb. The absence of IPS-1 leads to uncontrolled viral replication in CNS but plays only a minor role in shaping the humoral immune response in periphery.
Methods
Ethics statement
All animal experiments were performed in compliance with the German animal welfare law (TierSchG BGBl. S. 1105; 25.05.1998). The mice were housed and handled in accordance with good animal practice as defined by FELASA. All animal experiments were approved by the responsible state office (Lower Saxony State Office of Consumer Protection and Food Safety) under permit number AZ 33.9-42502-04-11/0528.
Mice and viral infections
C57BL/6 wild-type (WT) mice were purchased from Harlan.
IPS-1
−/−
mice on the C57BL/6 background were bred under specific pathogen free conditions at the Helmholtz Centre for Infection Research. LGTV strain TP21 and TBEV strain Hypr71 (G. Dobler) were propagated in Vero B4 cells. Titers were determined by focus forming assays on Vero B4 cells [
31]. Six- to ten-week-old mice were intraperitoneally infected with the indicated focus forming units (FFU) of LGTV or TBEV in 100-μl phosphate buffered saline (PBS). For intracranial infections, mice were anesthetized by intraperitoneal injection with a mixture of ketamine (100 μg/g body weight) and xylazine (5 μg/g body weight). Mice were injected with indicated FFU of LGTV in 20 μl PBS. Mice that lost more than 20 % of their body weight were sacrificed and perfused with 10 ml of PBS. Experiments with TBEV strain Hypr were performed in the biosafety level 3 (BSL3) facility at the Helmholtz Center for Infection Research.
Viral titers were determined by using a focus forming assay as described previously [
31]. Briefly, serial dilution of virus samples were added on monolayer of Vero B4 cells in 96-well plates. After 1 h of incubation at 37 °C in a 5 % CO
2, inoculum was removed and overlaid with 1.2 % Avicel RC-591 NF (FMC Biopolymer), 1× DMEM. Cell monolayers were fixed with 6 % paraformaldehyde dissolved in phosphate-buffered saline (PBS) 48 h post infection and permeabilized with 0.5 % triton X 100 and 20 mM glycine in PBS. LGTV foci were stained with monoclonal TBEV E antibody (monoclonal antiserum 1786 [
32]) diluted 1:1000 in PBS supplemented with 10 % fetal calf serum (FCS) and 0.05 % tween 80 and horse reddish peroxidase-conjugated anti-mouse immunoglobulin (Thermo Scientific) diluted 1:2000 in PBS, 10 % FCS and 0.05 % tween 80. Antibody-bound infection foci were visualized with True Blue
TM Peroxidase substrate (KPL Inc., Maryland, USA), counted under microscope, and virus titer was expressed as focus forming unit per milliliter (FFU/ml).
RNA extraction and real-time RT-PCR
RNA extraction from serum was done with QIAamp Viral RNA mini kit (QIAGEN), and LGTV levels were determine by comparing with a standard curve. The standard curve for viral RNA in serum was generated by RNA extraction, complementary DNA (cDNA) synthesis, and quantification with real-time reverse transcription PCR (RT-PCR) of control serum spiked with 1 – 10
5 FFU of LGTV. Mice organs were homogenized in Trizol reagent (Invitrogen) using Lysis Matrix (Nordic Biolabs) and the tissue homogenizer Fast Prep-24 (MP). The RNA was extracted using the Nucleo-Spin RNA II kit (Macherey Nagel). Five hundred nanograms of total RNA was used to synthesize cDNA with the QuantiTect Reverse Transcription Kit (Qiagen). Expression levels of mouse GAPDH, IFN-β, IFN-α4, IFN-λ, Mx1, viperin, IL6, CCL5, CXCL10, IRF1, and TRIM79α were detected by validated QuantiTect primer assays (Qiagen) and the KAPA SYBR FAST quantitative polymerase chain reaction (qPCR) Kit using the 7900HT Fast-Real-time PCR System (Applied Biosystems). Viral RNAs were detected by two different TaqMan-based assays; the TaqMan probe for TBEV [
33] detecting the 3′ NCR with a sensitivity of 10
4 copies for LGTV or the newly developed LGTV NS3-based TaqMan assay, forward primer 5′-AACGGAGCCATAGCCAGTGA-3′, reverse primer 5′-AACCCGTCCCGCCACTC-3′ and probe FAM-AGAGACAGATCCCTGATGG-MGB, with a sensitivity of 10 copies. For both assays, the KAPA probe FAST qPCR kit was used. Signals of indicated messenger RNA (mRNA) or viral RNAs were normalized to the GAPDH mRNA signal.
Brain immune cell quantification
Mice brains were harvested from uninfected or LGTV-infected mice. Following perfusion, brains were homogenized through a 70-μm cell strainer, digested with a collagenase solution (500 μg/ml collagenase D, 0.1 μg/ml trypsin inhibitor TLCK, 10 μg/ml DNase I, 10 mM HEPES in HBSS) for 1 h at room temperature. Cells were separated by centrifugation on a discontinuous 70-to-30 % Percoll gradient. For the detection of resident microglia, infiltrating macrophages and dendritic cells were incubated with anti-CD45, anti-CD11b, and anti-CD11c antibodies (BD Biosciences). Infiltrating T cells were detected using anti-CD3, anti-CD4, and anti-CD8 antibodies (BD Biosciences). Brain leukocyte numbers were quantitated using TruCount beads (BD Biosciences). Analysis was performed on BD LSRII using BD FACSDiva and FlowJo software.
Immunhistology
Immunohistological analysis were performed from WT and
IPS-1
−/−
mice infected with LGTV at different time points (
n = 3 per time point). Brains were removed after cardiac perfusion with PBS followed by 4 % paraformaldehyde (PFA) and incubated 24 h in 4 % PFA followed by an incubation in 30 % sucrose in 0.1 M phosphate buffer for an additional 24 h. Subsequently, the brains were frozen in Tissue Tek® Compound at -80 °C. 30 μm sagittal slices of the whole brain were cut using a freezing microtome (Frigomobil, Leica, Germany). All staining procedures were performed on free floating sections. Following a 1-h blocking step at room temperature in PBS containing 1 % BSA, 0.2 % Triton, and 10 % goat serum, the slices were incubated overnight at 4 °C in primary antibody solution consisting of 10 % goat serum and 0.2 % Triton in PBS. The following antibodies were used: monoclonal TBEV E antibody (monoclonal antiserum 1786 [
32]) monoclonal mouse anti-GFAP (Sigma, 1:500), monoclonal guinea pig anti-NeuN (SySY, 1:500), polyclonal rabbit anti-IBA1 (Synaptic System, 1:500), and polyclonal rabbit anti- cleaved caspase 3 [Asp175] (Cell Signaling, 1:400). Secondary anti-mouse, anti-guinea pig, or anti-rabbit antibodies conjugated with FITC, Cy3, or Cy5 (Jackson ImmunoResearch) were incubated 1:500 in PBS for 2 h at room temperature. All analyzed samples were comparable and support the conclusion. Representative pictures are shown.
Primary cell isolation and infection
Primary mouse hippocampal neurons were generated from WT and
IPS-1
−/−
mouse embryos E18 as described previously [
34]. Briefly, 70,000 dissociated hippocampal cells were seeded on poly-
l-lysine coated cover slips and incubated in Neurobasal medium (Invitrogen) supplemented with 2 % B27 (Invitrogen), 1× N-2 supplement (Invitrogen) and 0.5 mM Glutamax at 37 °C, 5 % CO
2, and 95 % humidity. After 2 weeks, cells were infected with LGTV (MOI-0.001) and kinetics of viral replication was recorded by focus forming assay.
Western blot analysis
For preparation of extracts, the brains or brain parts were lysed in buffer containing 250 mmol/L Tris, 0.5 % Triton X-100, and Halt protease inhibitor cocktail (Thermo Scientific, Schwerte, Germany). Western blot analysis was performed according to standard procedures. The following primary antibodies were used: Anti-STAT1-P (Tyr701, 58D6; Cell Signaling, Frankfurt, Germany), anti-STAT1 (STAT91/84; Transduction Laboratories Lexington, USA), and anti-actin (MAB 1501R, Chemicon, Limburg, Germany). Horseradish-peroxidase-conjugated anti-rabbit and anti-mouse antibodies (Amersham, Freiburg, Germany) were used as secondary antibodies using enhanced chemoluminescence detection (Bio-Rad, München, Germany). The chemiluminescence signal was recorded digitally by a ChemiDoc DRS imaging system (Bio-Rad, München, Germany). Digital signal acquisition and analysis were performed using the Quantity One Program (version 4.6; Bio-Rad, München, Germany).
Discussion
Protection against neurotropic LGTV infection requires coordinated action of the type I IFN system in both peripheral and CNS to prevent LGTV-induced inflammation and development of encephalitis [
30]. Here, we have characterized the function of IPS-1, a key adaptor molecule that act downstream of RLRs (RIG-I and MDA5) to activate the IFN and NF-κB pathways in LGTV infection. We have shown that IPS-1 signaling is important for controlling lethal LGTV infection. IPS-1-deficient mice show high viral replication in the peripheral and CNS, increased BBB permeability, massive infiltration of immune cells, and uncontrolled inflammation. After close analysis of different brain sections, we found that IPS-1 signaling plays a role in determining the base line levels of some antiviral ISGs but also controlling LGTV viral replication in all brain sections. Upregulation of IFN-β is specifically dependent on IPS-1 in olfactory bulb. Thus, we propose a specific role of pattern recognition receptors in the different sections of the brain in neurotropic LGTV infection.
Neuroinvasiveness and neurovirulence are key steps in the pathogenesis of neurotropic viruses. The type I IFN system is an important component of innate immunity and limits viral load of many flavivirus infections [
4]. Previously, we showed that
IFNAR
−/−
mice succumb to LGTV infection within 5 days with uncontrolled systemic viremia [
30]. LGTV-infected
IPS-1
−/−
mice showed declined systemic type I IFN responses. This finding agrees with our previous studies indicating that recognition of LGTV is IPS-1-dependent [
31] and that induction of type I IFN production by other RNA viruses is mediated by IPS-1-dependent mechanisms [
45‐
47]. The lower systemic IFN response probably contributed to higher systemic viral levels, thus contributing to higher viral titers and earlier time point of neuroinvasiveness (Additional file
5: Figure S4 ). However, no difference in neutralizing antibody titers or spleen immune cell composition was detected. These later findings are markedly different compared to the effect seen in WNV infected
IPS-1
−/−
mice [
47]. Virus entry into the brain by LGTV, TBEV, and Japanese encephalitis is independent of the integrity of the BBB because virus is detectable even when the BBB is intact [
30,
48,
49], and this is also true in IPS-1 deficient-mice.
The brain has distinct immune responses to pathogens and injury where resident brain cells including microglia and astrocytes, which are unique innate immune cells without direct counterparts in the periphery, can produce IFN and proinflammatory cytokines and can crosstalk with infiltrating immune cells [
50‐
53]. Local type I IFN response is very important for controlling lethal LGTV infection in the CNS [
30]. Indeed, high levels of LGTV replication were detected in the brain of IPS-1-deficient mice compared to WT. This in turn could lead to elevated IFN-β response in the brains due to increased activation of IPS-1-independent pathways. However, increased IFN response neither led to the induction of antiviral effector genes early in infection nor was it able to control the viral infection. One explanation might be that LGTV and TBEV non-structural protein NS5 can inhibit JAK STAT pathway downstream of IFNAR to block effector functions of IFN, e.g., upregulation of ISGs [
43,
44]. Accordingly, in spite of the presence of high levels of viral RNA and IFN-β in brains of
IPS-1
−/−
mice, day 4 post infection, STAT1 phosphorylation was not detectable and neither was ISG expression. Other studies have shown that even in the presence of type I IFN, viral replication could not be controlled and that other mechanisms like IRF-1 are essential to control viral replication [
34]. IPS-1 located on peroxisomes has been shown to induce an IRF-1-dependent IFN-independent antiviral mechanism [
54], which has been shown to be important for both VSV and HCV [
34,
55]. This pathway might be one of the reasons why IPS-1-deficient mice cannot control the LGTV infection in CNS.
Higher viral burden observed in
IPS-1
−/−
mice brains was a result of higher viral replication in neurons, similar to LGTV-infected
TLR7
−/−
mice brains [
56]. Although not highly infected, astrocytes and microglia showed an activated phenotype as evident from increased expression of GFAP and IBA-1, respectively. Astrocytes are the most abundant glial cell population in the brain and when activated could act as potential source of proinflammatory cytokines. These cytokines might contribute to BBB breakdown and TBEV induced neurotoxicity and thus play a major role in TBEV pathogenesis [
57,
58]. We also detected a strong inflammatory response in the
IPS-1
−/−
mice leading to the opening of BBB and activation of the cell-mediated immune response with high quantity of infiltrating CD4 and CD8 T cells in later stages of infection. Although a strong T cell response was observed, it was not protective; similar findings were also seen in case of WNV [
47]. Notably rather than protection, this higher T cell response in the brain contributed to immunopathology [
59]. Apoptosis in the brains of LGTV-infected
IPS-1
−/−
mice were apparent as a result of high viral infection and inflammatory response, which combined ultimately lead to the death of mice.
Local immune response within the CNS plays an important role in combating viral infections [
60]. In the absence of IFNAR signaling in the CNS, LGTV replicated to higher levels and mice succumbed early to the infection [
30]. Studies with Sindbis, JEV, and WNV reports that the IPS-1 pathway is important in controlling viral replication in the brain [
47,
61,
62]. Comparison of the different brain regions displays specific antiviral mechanisms [
39,
63,
64]. Two neuronal subtypes, granule cell neurons of the cerebellum and cortical neurons from the cerebral cortex, have unique innate immune programs and showed differential permissiveness to replication of several positive-stranded RNA viruses [
39]. Cerebellum has also been shown to be more resistant to WNV viral infection compared to cerebrum [
39], and this holds true for LGTV as well. One explanation for this might be an increased basal expression of ISGs such as Ifi27, Irg1, viperin, and Stat1 in granule neurons [
39]. IPS-1 signaling seems to be very important to restrict LGTV growth throughout the brain. Although, high IFN-β levels were detected in
IPS-1
−/−
, it could not restrict replication in any of the brain regions. Comparison of the basal expression levels of some ISGs known to be active against TBEV, LGTV, or flaviviruses in general [
41,
42,
65,
66] revealed that the olfactory bulb generally shows higher levels of viperin, TRIM79α, Mx1, and IRF1 compared to cerebrum. This might play a role for defense against pathogens which enter the brain preferentially via the olfactory route. IPS-1-deficient mice have lower expression of some specific ISG but not all. Since TBEV and LGTV are very sensitive to the antiviral action of viperin [
42] (unpublished data), this might be one of the contributing reasons why IPS-1-deficient mice are more susceptible to LGTV infection. This initial advantage might also contribute to higher levels of viral NS5 and inhibition of STAT1 phosphorylation in IPS-1 deficient mice. This is in contrast to wild-type mice that have higher basal levels of antiviral ISGs, respond faster by production of IFN-β, and have a more robust STAT1-mediated expression of ISGs required to contain the infection. Notably, LGTV infection was restricted to olfactory bulb in WT mice. Similar observations were made in VSV and cytomegalovirus (CMV) infections. VSV- and CMV-induced long-distance interferon signaling within the brain that blocks virus spread [
67] and astrocytes were found to be the main producer of IFN-β in the olfactory bulb in response to VSV infection [
18]. In LGTV infection, the olfactory bulb seems to be more dependent on IPS-1 signaling for IFN-β production compared to the other brain parts, where RLR independent pathways (MyD88 and TRIF) compensated with higher IFN-β production. Thus, different cell types might exert specific immune response in different sections of the brain.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
CK, AK, and AKÖ planned experiment, analyzed the data, and wrote the manuscript. CK, LZ, EW, SN, and KMP performed the experiments. JS provided the reagents. NOG helped with the writing of the manuscript. AKÖ and AK provided the facilities and funding. All authors read and approved the final manuscript.