Tick-Borne Encephalitis Virus Infects Rat Astrocytes but Does Not Affect Their Viability

Tick-borne encephalitis virus (TBEV) causes one of the most dangerous human neuroinfections in Europe and Asia. To infect neurons it must cross the blood-brain-barrier (BBB), and presumably also cells adjacent to the BBB, such as astrocytes, the most abundant glial cell type. However, the knowledge about the viral infection of glial cells is fragmental. Here we studied whether TBEV infects rat astrocytes. Rats belong to an animal group serving as a TBEV amplifying host. We employed high resolution quantitative fluorescence microscopy to investigate cell entry and cytoplasmic mobility of TBEV particles along with the effect on the cell cytoskeleton and cell survival. We report that infection of astrocytes with TBEV increases with time of exposure to TBEV and that with post-infection time TBEV particles gained higher mobility. After several days of infection actin cytoskeleton was affected, but cell survival was unchanged, indicating that rat astrocytes resist TBEV-mediated cell death, as reported for other mammalian cells. Therefore, astrocytes may present an important pool of dormant TBEV infections and a new target for therapeutic intervention.


Introduction
Tick-borne encephalitis virus (TBEV) is medically important human pathogen that causes one of the most dangerous neuroinfections in humans in Europe and in Asia [1][2][3]. TBEV is a member of the genus Flavivirus, family Flaviviridae [4]. Mature virus particles are enclosed within lipidic envelope which carry attachment molecules (E protein) for the host cell receptors (heparan sulfate) [1,[5][6][7]. The first and the most important host cells infected by TBEV are likely epidermal Langerhans cells (i.e. dendritic cells; [8]) which transport the virus to the lymph nodes and initiate the spread of infection to lymphoid compartments [3]. In some vertebrate species the virus is neurovirulent and crosses the blood-brain barrier (BBB), which isolates and protects the brain tissue from fluctuations in nutrients, hormones, metabolites, and from the direct influences of many endo-or exogenous compounds circulating in the blood [3,9]. The mechanism by which TBEV invades the central nervous system (CNS) is not clear, and may include cytokine-mediated entry, entry of TBEV-infected cells of the immune system and infection of endothelial cells [10,11]. After entering the CNS, neurons are primary targets, but other brain cells may also be infected [3]. Such cells are astrocytes which are positioned between synapses and endothelial cells, playing a role in neurovascular coupling [12,13]. The infection of astrocytes may have unfavourable consequences for brain functioning. In general, astrocytes have many leading roles in the brain, including integrating neuronal functions, neuronal support, and the maintenance of BBB inter-endothelial tight junctions in normal and pathologic conditions [14][15][16][17][18][19][20][21][22][23][24][25][26]. It has been shown recently that brain TBEV infection alters the permeability of the BBB in mice [11], and astrocytes may be implicated in this process, since these cells regulate blood flow in the brain [12,13]. In addition, the effects of neurovirulence can be observed relatively rapidly, compared to the clinical course of the disease [3,27], which likely indicates that a reservoir of TBEV may exist in the CNS.
The aim of this study was to test whether astrocytes, the most numerous glial brain cells [17], can be infected by TBEV and to measure cytoplasmic TBEV particle dynamics in the initial phases after the infection, along with the astrocyte viability. The susceptibility to TBEV infection would make astrocytes a potential TBEV reservoir. We used rat astrocytes as a model cell, since rodents of several species are known to be TBEV amplifying hosts and may maintain TBEV through latent persistent infections [3,28].
The results show that TBEV infects rat astrocytes and that the infection of a single cell, which progresses in time-dependent phases, is associated with changes in actin cytoskeleton, but astrocyte viability is unaffected. We propose that astrocytes represent an important reservoir of TBEV brain infection, which makes these cells a new target for therapeutic intervention. Given their tight morphological association with blood vessels, infected astrocytes may possibly affect the BBB and neurons.

Ethics Statement
The care for experimental animals and the euthanization of animals was carried out in strict accordance with the following ethical codes and directives:

Cell Cultures
Astrocyte cultures were prepared from cortices of 3 days old Wistar rats as described [29]. Cells were maintained in highglucose Dulbecco's modified Eagle's medium (Invitrogen, Life Technologies, Carlsbad, NM, USA) containing 10% fetal bovine serum (FBS), 1 mM pyruvate, 2 mM glutamine and 25 mg/ml penicillin/streptomycin at 37uC, 95% air/5% CO2. After reaching confluence, cells were manipulated as described [30]. Prior the experiments cells were sub-cultured onto Lab-Tek TM chambered coverglass (Thermo Scientific) or onto 22 mmdiameter poly-L-lysine-coated coverslips and used within 6 days after plating. Vero E6 cells were used for the preparation of virus stocks and as productively infected control cell type after infection with TBEV. Cells were maintained in Dulbecco's MEM with high glucose (DMEM GlutaMAX TM ; Invitrogen) supplemented with 10% FBS. Cultures were incubated at 37uC/5% CO2.) Unless stated otherwise all chemicals for maintaining cell cultures and experimental procedures were obtained from Sigma-Aldrich (Diesenhofen, Germany).

TBEV Labelling and Cytotoxicity Test
Viral strains and isolates of TBEV can be classified into three subtypes: the European, the Siberian and the Far Eastern [31]. In this study we used the European TBEV strain Ljubljana 1 [32]. TBEV was grown 7 days on Vero E6 cells. Supernatant was collected and centrifuged twice at 4uC (10 min at 3200 6 g and 5 min at 20800 6 g in Eppendorf 5804R centrifuge). Pellet was

TBEV RNA Concentration
The concentration of TBEV RNA was measured using one-step quantitative real time RT-PCR [34]. Vero E6 cells were infected with TBEV and the virus concentration was measured in the supernatant collected from Vero E6 cells 7 days post infection (p.i.) before and after labelling TBEV with fluorescent lipophilic VybrantH DiD labelling solution. Further on, in the experiments, we have used the TBEV in concentration of 10 8 copies/ml, labelled with 50 mM DiD.

Imaging
Imaging of fixed and live cells was performed with an inverted confocal microscope (Zeiss LSM 510 META, Carl Zeiss) using oilimmersion objective 636/NA 1.4. For excitation of DiD dye He/ Ne laser was used (633 nm), the emission light was filtered with long pass filter, with the cut off below 650 nm. The conjugate Alexa Fluor 488 was excited by argon laser (488 nm) and the emission light was collected through the band pass filter (505 to 530 nm). The conjugate Alexa Fluor 546 was excited by He/Ne laser and the emission light was filtered with long pass filter 560 nm. To eliminate possible bleed-through, the green and red emission fluorescence was acquired sequentially. In live cells the mobility of vesicles that expressed DiD fluorescence of labelled TBEV was recorded. Time series images were recorded in 2 s intervals for 2 min of total recording time. Experiments were conducted at 37uC (Heatable universal mounting frame, Carl Zeiss).

Immunocytochemistry
The cells were washed with the phosphate buffered saline (PBS), fixed in 2-4% formaldehyde (prepared from paraformaldehyde) in PBS for 5-15 min at room temperature (RT) and permeabilized with Triton X-100 for 10 min at RT. The non-specific back- ground staining was reduced by incubating cells in blocking buffer, containing 3% bovine serum albumin (BSA) and 10% goat serum in PBS, at 37uC for 1 h. The cells were then stained with primary antibodies, diluted into 3% BSA in PBS and incubated at 37uC for 2 h or at 4uC overnight. When two primary antibodies (raised in different species) were used, the staining was done sequentially. Afterwards, the cells were rinsed in PBS and stained with secondary antibodies at 37uC for 45 min. At the end of the staining protocol the cells were mounted onto glass slides using Slowfade Gold antifade reagent (Molecular Probes, Invitrogen).

Analysis
The mobility of fluorescently labelled vesicles was analysed by ParticleTR software (Celica, Slovenia). To describe vesicle mobility the parameters were calculated as described [36,37]: step length (displacement of a vesicle in the time interval of 2 s), track length (TL, the total length of the analysed vesicle pathway), velocity and maximal displacement (MD; [36]). Vesicle mobility was analysed in cells from three independent astrocyte cultures. The analysis of the vesicle mobility was performed for epochs of 30 s.
Vesicle size was analysed with ImageJ software (available at National Institute of Health, USA, http://rsbweb.nih.gov/ij/). Fluorescent particles above the threshold level 8 pixels 2 (0.1460.14 mm 2 ) were determined as vesicles, corresponding to the vesicle size area above 0.1568 mm 2 to cover a broad span of imaged vesicles with different fluorescence intensities at slightly different z- positions within optical slice, similarly as described [38]. The extent of the co-labelled vesicles by fluorescent probes was determined by manually counting the observed fluorescent probes in the green and red channels.
Statistical significance was determined with the Mann-Whitney Rank Sum test. Values presented on graphs are expressed as mean 6 s.e.m.

Identifying Initial Steps of TBEV Infection in Astrocytes
To image intracellular localization of TBEV particles, we prelabelled them with 50 mM DiD lipophylic dye and afterwards infected cultured astrocytes with conc. of 10 7 RNA copies/ml. Infection with TBEV was confirmed by immunolabelling with antibodies against the viral large envelope protein E [5] (Figure 1A, inset). TBEV labelling of cells increased with the time p.i. as assessed by counting TBEV particles per cell ( Figure 1A,B). On one hand, the screening of several time periods following the   Figure 1A). We analyzed up to 31 cells from three different animals. On the other hand, we also determined the fraction of cells in which at least 3 DiD-TBEV particles were observed in a coverslip with cultured astrocytes. Clearly, the percent of infected cells increased as a function of post-infection time: 6869% (4 h), 9262% (18 h; P,0.05; Figure 1B).

With Increasing p.i. time DiD-TBEV Particles Acquire Increased Directionality
Large vesicles with fluorescence area above 1.078 mm 2 were immobile. On the other hand, small vesicles were mobile and were subjected to mobility analysis. Speed, track length (TL) and maximal displacement (MD) were calculated as described previously [36]. In 30 s of recording time vesicle pathways appeared predominantly clumped, meaning that vesicles exhibited nondirectional mobility (NDM). However, some vesicles displayed directional mobility (DM), as evident from elongated trajectories ( Figure 4A). The percent of DM periods (Figure 4, MD.1 mm, (dashed line)) remained below 20% during all incubation times, although the percent of small vesicles increased from 41% (2 h) to 72% (18 h, Figure 3). MD and TL of DM periods and NDM periods were distinct ( Figure 5). Mean MD of DM periods significantly increased: from 1.6960.27 mm (2 h), to 2.6860.23 mm (4 h; P,0.05) and 2.4260.09 mm (18 h; P,0.01) ( Figure 5C). And mean TL of DM periods significantly increased from 3.7460.35 mm (2 h) to 5.0060.29 mm (4 h; P,0.05, Figure 5D) and to 4.

TBEV Induces Disintegration of Actin Filaments While Cell Viability Remains Unchanged
Cytoskeleton rearrangements of host cells are reported for several viral infections (Herpesvirus, [43]; TBEV, [27]). To examine whether TBEV infection triggered changes of the rat astrocyte cytoskeleton, molecular motor-associated microtubules and actin filaments were labelled in astrocytes at several p.i. times (4 h, 18 h, 48 h, 3 days and 6 days). No obvious alterations of cytoskeleton were noticed in rat astrocytes until day 2 p.i. Then, after day 3 p.i. significant reorganization of actin filaments was observed, while microtubules appeared unaffected ( Figure 6).
Finally, to asses whether TBEV triggers the cytopathic effect (CPE) in primary rat astrocytes, we recorded fields of view of TBEV infected primary rat astrocytes and control Vero E6 cells and measured their viability for 14 days (Figure 7). The CPE (morphologically altered cell shape, detached cells) was observed in Vero E6 cells at day 3 p.i, whereas in astrocytes no CPE was recorded ( Figure 7A). Morevover, the results of the trypan blue exclusion viability test show that the viability of primary rat astrocytes was not affected by TBEV infection (Figure 7 B). On the other hand, in the control Vero E6 cells, which were used to multiply TBEV, the viability was significantly affected already after 18 h (6662%, P,0.001) and was further reduced to 1662% (P,0.001) at day 14 p.i., compared to non-infected cells (9562%) ( Figure 7B). Successful replication of TBEV in both cell types was confirmed by measuring the virus load (TBEV copies/cell) at different times p.i. (Figure 8). Moreover, the supernatant collected from astrocytes at the end of the experiment (14 days) still reduced the viability of Vero E6 cells (39%) at day 7 p.i. and triggered CPE (Figure 8), confirming the presence of infective TBEV. From these data we conclude that TBEV infection does not significantly affect the viability of rat astrocytes, implying that these cells could act as a potential TBEV reservoir.

Discussion
Despite of the high clinical importance of TBEV neuroinfection (long-term neurological disabilities, over 1% mortality rate, growing incidence in Europe), nothing is known about the TBEV infection of the most abundant neuronal supportive glial cells, astrocytes, and their possible implication in active or dormant TBEV infections. Recent data indicate that TBEV can be detected in rodent organs for longer periods of time and that the brain tissue stands out by high virus load [34]. In this report experimental evidences show that astrocytes, the most abundant mammalian glial cells become infected with TBEV, which makes them a potential mediator of brain infection and a reservoir of brain TBEV in rodents.
TBEV internalization into astrocytes increased with longer postinfection time. Although the mechanism of internalization remains to be elucidated, a number of different proteins have been identified as potential flavivirus receptors on the cell surface, however there is no direct evidence for a role of any of these during TBEV entry into host cells, and some of the results are controversial [44]. Glycosaminoglicans, such as heparan sulphate (HS), are expressed on different tissues [45]. Frequently, HS appears to serve as a primary attachment molecule that concentrates viral particles on the cell surface and may facilitate the subsequent binding to more specific receptor molecules [7,46]. An involvement of HS during attachment and entry of TBEV has been demonstrated, although it appears that more than one single type of host-cell molecule is involved [3,5,7,47]. In astrocytes mimicking injury response, upregulation of HS and higher sulfation of heparan sulphate proteoglycans occurs [48]. Therefore, one possible explanation of increased time-dependent TBEV entry into astrocytes could be attributed to an increased expression of HS on astrocyte membrane of infected astrocytes. However, this was not monitored in our experiments. The internalization of TBEV particles into astrocytes was consistent with confirmed clathrin-dependent entry of several members of Flaviviridae family: West Nile virus (WNV) [2], Dengue virus [49,50], Hepatitis C virus [51] and Bovine Viral Diarrhoea virus (BVDV) [52]. And their localization in late endosomes/lysosomes was consistent with other flaviviruses: BVDV, [52]; WNV, [53]; DIL-labelled dengue virus, [50].
TBEV-loaded vesicles observed following 18 h post-infection time were predominantly smaller or comparable in size to early endosomes (300 nm, [2]; 300-400 nm, [54]); and much smaller from late endosomes which have on average 700 nm in diameter [55]. Taking into account the point-spread function related overestimation of vesicle diameter [38], the majority of TBEV vesicles exhibited diameters between 150-400 nm. This is in line with WNV vesicles, which have 100 nm in diameter until they start to fuse and become approximately the size of 500 nm [2]. A minor portion of DiD-TBEV particles (6% at 2 h p.i., 5% at 4 h p.i. and 20% at 18 h p.i.) was even comparable in apparent size to peptidergic and glutamatergic vesicles monitored in astrocytes (50-100 nm, [38]; [56]). The rest of TBEV-loaded vesicles were between 500 or 600 nm (diameter), which is in the range of late endosomes and lysosomes [54,55]. With longer p.i. time the average size of large vesicles decreased, which indicates changes in vesicle dynamics, such as attenuated formation or enhanced cleavage of very large vesicles.
Despite the increase in the percent of smaller highly mobile TBEV loaded vesicles recorded with longer time p.i. (up to 72%), the percent of directional mobility periods remained below 20% at all times p.i. This is in line with the mobility properties of other vesicles which travel along the cytoskeleton [37,57] and with the restricted availability of host cell trafficking apparatus to be used by internalized TBEV viruses (reviewed in Greber [58,59]). These vesicles apparently move along both molecular motor-associated filaments. Analysis of vesicle speed revealed that the speed of DiD-TBEV vesicles generally corresponded to speeds of processive human myosins (Va, VI, and X) which travel as fast as 0.3-0.9 mm/s in vitro and 0.1-0.4 mm/s in live cells (references within [60]). These values correlate with DENV particle trafficking towards endosomes [61]. On the other hand, maximal recorded speeds above 1 mm/s confirm also the involvement of kinesin motors along microtubules [62]. Increased mobility at longer periods p.i. may reflect the changes in local protein synthesis affecting vesicle trafficking, as it was observed in axons infected by pseudorabies virus [63].
It is possible that viruses are transported on the account of innate cell vesicles, therefore we assessed if this impairs cell viability. In general, little is known about how various cell types in the CNS react to viral infections causing encephalitis [64]. In TBEV infected rat astrocytes we observed alterations of actin cytoskeleton, and no apparent change in the arrangement of microtubules. This is in contrast from the observation in human glioblastoma cells, where TBEV caused a substantial microtubule rearrangement already after 48 h p.i. [27]. The cell viability of primary rat astrocytes remained unaltered during monitored time p.i. (14 days). We did not observe any TBEV-triggered necrosis in rat astrocytes in 14 days p.i. (95.060.9% viable cells in the noninfected control and 97.061.0% viable cells after 14 days p.i., P = 0. 267). It appears that astrocytes are much more resilient to TBEV infection than the control Vero E6 cells (monkey kidney epithelial cell lineage) that showed reduced viability already after 18 h (Figure 7) and than porcine kidney cells, where almost all cells were dead at 50 h p.i. [65]. The shape of astrocytes was unaltered in 14 days p.i., the cells were not rounded, which is also in contrast to glioblastoma cell lines where 20% of cells were apoptotic already after 48 h p.i.; determined by rounding of cells and TUNEL assay [27]. These data suggest that in contrast to human glioblastoma cells rat astrocytes are more resilient to TBEV infection. Thus, rat astrocytes may serve as a reservoir for spreading the viral infection or maintaining TBEV reservoir in wild rodents. It has been reported that the virus likely remains dormant in the brain tissue for long time in wild rodents, while they show no clinical symptoms of TBEV infection [28,34,66]. We propose that the cell reservoir of dormant virus may consist of astrocytes. Vesicle traffic in astrocytes can be modified by therapeutics, such as fingolimod [67], which has been recently introduced to treat multiple sclerosis. Therefore, a similar strategy may be used to prevent astrocyte TBEV infection and possibly other brain cells. Recently, it was reported [11] that TBEV entry is independent on the breakdown of the BBB, but that the BBB breakdown is a consequence of TBEV brain infection. Whether astrocytes play a role in this process remains to be investigated, especially in the light of astrocyte role in compromising the permeability of the BBB.