|Home | About | Journals | Submit | Contact Us | Français|
Toll-like receptor 7 (TLR7) recognizes guanidine-rich viral ssRNA and is an important mediator of peripheral immune responses to several ssRNA viruses. However, the role that TLR7 plays in regulating the innate immune response to ssRNA virus infections in specific organs such as the central nervous system (CNS) is not as clear. This study examined the influence of TLR7 on the neurovirulence of Langat virus (LGTV), a ssRNA tick-borne flavivirus. TLR7 deficiency did not substantially alter the onset or incidence of LGTV-induced clinical disease; however, it did significantly affect virus levels in the CNS with a log10 increase in virus titres in brain tissue from TLR7-deficient mice. This difference in virus load was also observed following intracranial inoculation, indicating a direct effect of TLR7 deficiency on regulating virus replication in the brain. LGTV-induced type I interferon responses in the CNS were not dependent on TLR7, being higher in TLR7-deficient mice compared with wild-type controls. In contrast, induction of pro-inflammatory cytokines including tumour necrosis factor, CCL3, CCL4 and CXCL13 were dependent on TLR7. Thus, although TLR7 is not essential in controlling LGTV pathogenesis, it is important in controlling virus infection in neurons in the CNS, possibly by regulating neuroinflammatory responses.
Members of the genus Flavivirus are enveloped, positive-sense ssRNA viruses. Several flaviviruses are known human and/or animal pathogens, including yellow fever virus, dengue virus, West Nile virus (WNV), Japanese encephalitis virus and tick-borne encephalitis virus (TBEV). These viruses share similar virion architecture, genomic organization and replication cycles (Kuno et al., 1998). Flaviviruses routinely emerge beyond their known geographical ranges, causing millions of infections annually. Flaviviruses belonging to the TBEV serocomplex are highly virulent to humans, causing severe neurological disease. Mortality ranges from 1 to 40% depending on virus strain (Gritsun et al., 2003; Mandl, 2005).
Langat virus (LGTV) is a naturally attenuated member of the TBEV serocomplex and is therefore an ideal model for studying host responses to flavivirus infection (Charlier et al., 2006). LGTV may induce neurological disease in humans and other mammals, similar to TBEV, although infections in humans are less severe than those caused by more virulent strains of TBEV (Maximova et al., 2008). Therefore, the close molecular relationship of LGTV with TBEV provides an opportunity to study the pathogenesis of tick-borne flavivirus infection, including virus-induced encephalomyelitis (Denk & Kovac, 1969; Pletnev & Men, 1998; Semenov et al., 1981).
The innate immune response is essential for the control of virus infection. In specific tissues such as the central nervous system (CNS), the innate immune response appears to both regulate virus infection and influence viral pathogenesis. Several studies have shown that the type I interferon (IFN) response is critical for control of flavivirus replication and that these viruses have developed mechanisms to inhibit the type I IFN pathway, which may be important for virus pathogenesis (Best et al., 2005; Laurent-Rolle et al., 2010; Liu et al., 2006; Samuel et al., 2006). Furthermore, pro-inflammatory cytokines including interleukin (IL)-1b, IL-12, IL-6, tumour necrosis factor (TNF), CCL2 and CCL5 are increased in the CNS of mice following infection with flaviviruses including TBEV (Růžek et al., 2011; Tigabu et al., 2010). However, how these responses are induced and which pattern recognition receptors (PRRs) are involved in detecting virus infection in the periphery as well as in the CNS is not completely understood. Studies with WNV suggest that numerous PRRs are involved in recognizing flavivirus infection, including the RNA helicases, retinoic acid-inducible gene 1 (RIG-I) and melanoma differentiation-associated protein 5 (MDA-5), as well as the Toll-like receptors TLR3 and TLR7 (Daffis et al., 2008; Fredericksen et al., 2008; Town et al., 2009; Wang et al., 2004; Welte et al., 2009). Interestingly, the role of these PRRs may differ depending on the route of inoculation. For example, TLR7 was found to influence the pathogenesis of WNV following intraperitoneal (i.p.) infection, but not following intradermal (i.d.) infection or transmission of the virus by mosquito bite (Town et al., 2009; Welte et al., 2009). TLR7 is present on numerous cell types in the CNS, and direct ligand inoculation into brain tissue induces a strong localized type I IFN response (Butchi et al., 2008, 2011), indicating that TLR7 may be a contributing factor in the development of local antiviral responses in the brain.
In the current study, we examined the innate immune response to LGTV infection in the mouse CNS and determined the role of TLR7 in regulating virus infection, neuroinflammation and disease. Surprisingly, there was a dichotomy in the effect of TLR7 during LGTV infection in the CNS. TLR7-deficient mice developed clinical neurological disease at a similar rate to that observed for wild-type mice. However, TLR7 deficiency led to increased virus load in the CNS. This increase was associated with heightened type I IFN responses but decreased expression of pro-inflammatory chemokines and TNF production. The innate immune response mediated by TLR7 appeared to be the primary influence on the heightened virus levels in the CNS, as LGTV replicated at similar levels in wild-type and TLR7-deficient neurons. These studies suggest that TLR7 can control flavivirus replication by altering the pro-inflammatory response to LGTV infection in the brain. Furthermore, they demonstrate that the type I IFN response is not dependent on TLR7 and is not sufficient to control virus infection in the absence of TLR7.
Several PRRs may influence the neuroinflammatory response to LGTV infection and affect viral pathogenesis. Studies with other flavivirus such as WNV have indicated that RIG-I, MDA-5 and TLR3 may all play a role in modulating the inflammatory response to flavivirus infection (Daffis et al., 2008; Fredericksen et al., 2008; Wang et al., 2004). The role of TLR7 in flavivirus infection is less clear, with different studies indicating that TLR7 may or may not affect virus-mediated disease depending on the route of infection (Town et al., 2009; Welte et al., 2009). To examine the function of TLR7 in LGTV pathogenesis and modulation of innate immunity, wild-type and TLR7-deficient mice were infected i.d. with 105 p.f.u. LGTV. Interestingly, no difference was observed in the development of clinical disease between the wild-type and TLR7-deficient mice (Fig. 1a), demonstrating that the deficiency in TLR7 did not influence virus pathogenesis. Similar results were also observed for mice infected with a low dose (103 p.f.u.) of LGTV (data not shown).
The lack of effect of TLR7 deficiency on virus-induced clinical disease suggested that TLR7 did not modulate LGTV infection of the CNS. However, virus levels in the CNS at the onset of disease [11 days post-infection (p.i.)] in mice infected with 105 p.f.u. virus were 1 log10 higher in TLR7-deficient mice than in wild type (Fig. 1b). To determine whether the increased virus levels in TLR7-deficient mice were due to increased neuroinvasion of the virus, we analysed a time course of virus infection in wild-type and TLR7-deficient mice. In wild-type mice, virus levels peaked at 6 days p.i. and did not increase at later time points (Fig. 1c). Virus levels were similar to wild-type in TLR7-deficient mice at 6 days p.i.; however, virus levels then continued to increase at 8 days p.i. and at the time of clinical disease (Fig. 1c). Analysis of virus RNA at early time points showed similar viral RNA levels in the CNS at 2 and 4 days p.i., confirming comparable early virus levels in wild-type and TLR7-deficient mice (Fig. 1d). This suggested that TLR7 deficiency did not affect virus entry in the CNS but rather influenced virus replication within the CNS after virus infection was established. To examine directly whether TLR7 affected virus levels within the CNS, mice were inoculated with LGTV by intracerebral (i.c.) inoculation and followed for disease pathogenesis and virus load. Similar to the results observed with i.d. inoculation, there was no difference in the onset of disease between wild-type and TLR7-deficient mice infected i.c. with LGTV (Fig. 2a). However, there was a significant difference in virus load, with a log10 higher titre of virus at the time of clinical disease (7 days p.i.) in the brain tissue of TLR7-deficient mice compared with wild-type controls (Fig. 2b). Thus, the influence of TLR7 on higher virus load appeared to be the result of virus replication within the CNS.
Immunohistochemical analysis for viral protein demonstrated that the increase in virus levels was associated with an increase in virus infection of neurons within the dentate gyrus and hippocampus (Fig. 3), as well as neurons in the cortex (data not shown). Neither microglia (Fig. 3b, c, red fluorescence) nor astrocytes (Fig. 3e, f, green fluorescence) were positive for virus in wild-type or TLR7-deficient mice, indicating that the tropism of the virus did not change in the absence of TLR7. Thus, TLR7 deficiency resulted in increased infection of neurons in the CNS by LGTV. However, this increase in infection did not appear to influence the onset of neurological disease directly.
To examine whether TLR7 deficiency directly affected virus replication in neurons, we infected primary cortical neurons with LGTV. No consistent difference was observed in virus titre from wild-type or TLR7-deficient neurons, indicating that a direct effect of TLR7 in neurons was not responsible for the increased virus titres in the CNS (Fig. 4a). As TLR7 can induce strong innate immune responses and neurons can produce cytokines such as beta IFN (IFN-β1) and TNF in response to stimuli, we analysed whether the innate immune response by neurons to LGTV was dependent on TLR7. Surprisingly, neurons deficient in TLR7 expressed significantly higher levels of Ifnb1 mRNA and slightly higher levels of IFN-regulatory factor 1 (Irf1) and Tnf mRNA in response to LGTV infection (Fig. 4b–d), indicating a suppressive effect of TLR7 on neuron-specific innate immune responses to LGTV infection.
Brain tissues from wild-type and TLR7-deficient mice infected i.d. with LGTV with clinical disease were examined in a blinded analysis to determine whether the higher virus titres were associated with significant histological changes in cellular infiltration or neuronal damage. Surprisingly, this difference in virus infection of neurons did not correlate with increased necrosis of neurons in either the hippocampus or the cerebrum (Fig. 5). Similar levels of meningitis and perivascular cuffing were observed in wild-type and TLR7-deficient mice, with a wide range of severity in clinical mice (Fig. 5). mRNA analysis revealed higher levels of the T-cell marker Cd3ϵ mRNA, as well as increased Cd8α mRNA, suggesting an increase in CD8+ T-cells in the brain in TLR7-deficient mice compared with wild-type controls (Fig. 6a, c). In contrast, Cd4 and Cd11c mRNA levels were comparable between wild-type and TLR7-deficient mice (Fig. 6b, f). Thus, TLR7 deficiency did not substantially alter LGTV-induced pathology in the brain, with the exception of a slight increase in the presence of CD8 T-cells.
Analysis of gliosis by immunohistochemistry revealed no striking difference between TLR7-deficient mice and wild-type controls (Fig. 3 and data not shown). mRNA expression of GFAP, an astrocyte activation marker, indicated a twofold higher increase in Gfap mRNA expression in TLR7-deficient mice compared with wild-type mice (Fig. 6d). No difference in F4/80, a marker for microglia/macrophages was detected (Fig. 6e). Thus, TLR7 is not required for the activation of glial cells or the recruitment of peripheral cells in response to LGTV infection in the CNS.
TBEV infection of the CNS can induce a strong pro-inflammatory cytokine response including IL-1b, IL-12, IL-6, TNF, CCL2 and CCL5 (Růžek et al., 2011; Tigabu et al., 2010). Analysis of mRNA expression of innate immune response genes in response to LGTV infection demonstrated a greater than tenfold upregulation of mRNA for a number of innate immune genes at the time of clinical disease (Table 1). Genes with increased mRNA included chemokines such as Ccl2 (MCP-1), Ccl3 (MIP-1a), Ccl4 (MIP-1b), Ccl5 (RANTES), Ccl7 (MCP-3), Ccl8 (MCP-2), Ccl12 (MCP-5), Cxcl1 (KC), Cxcl9 (MIG), Cxcl11 (Itac) and Cxcl13 (Blc); cytokines such as Il-1β, Il-10, Ifna2, Ifna4, Ifnb1, Ifng and Tnf, as well as IFN-stimulated genes (ISGs) such as Cxcl10, Ifit1, Ifit3, Irf1, Isg15 and Isg20. Comparison of the innate immune response between LGTV-infected wild-type and TLR7-deficient mice demonstrated a difference in only eight of these genes (Fig. 7). This included increased mRNA production of type I IFN response genes Ifna4 and Irf1 (Fig. 7), as well as Ifna2 and Spp1 (data not shown). In contrast, a significant decrease in mRNA expression was observed for the pro-inflammatory cytokine genes Tnf, Ccl3, Ccl4 and Cxcl13 (Fig. 7 and data not shown). Thus, the increased virus levels in TLR7-deficient mice correlated with an increase in the type I IFN response and a decrease in the pro-inflammatory cytokine/chemokine response of genes associated with macrophage/microglia activation in the CNS. Interestingly, type I IFN response genes, including Irf1, were not significantly different between wild-type and TLR7-deficient mice at early time points p.i. (Fig. 7e and data not shown), consistent with similar virus levels in the CNS at these time points (Fig. 1c). In contrast, cytokine mRNAs, including Cxcl13, were consistently lower at early time points p.i. (Fig. 7f and data not shown), despite similar levels of virus (Fig. 1c). This suggests that TLR7 has a critical role in generating the pro-inflammatory cytokine response to LGTV infection in the CNS, with the exception of the type I IFN response, which may be more closely associated with virus levels.
Our current studies demonstrate a role for TLR7 in controlling flavivirus replication in the CNS but not in influencing neuropathogenesis. The influence of TLR7 on LGTV levels in the CNS appeared to be primarily at the level of regulating neuronal infection. Increased virus levels were observed following both i.d. and i.c. infection, indicating that the difference in virus levels was not due to the ability of the virus to enter the CNS. Furthermore, immunohistochemical analysis showed increased neuronal infection in TLR7-deficient mice (Fig. 3) compared with wild-type controls, with no alteration in cellular tropism. TLR7-deficient neurons were not more susceptible to LGTV infection than wild-type neurons (Fig. 4), suggesting an extracellular mechanism of controlling neuronal infection in vivo. Analysis of TLR7-deficient mice indicated that the production of pro-inflammatory cytokines CCL3, CCL4, CXCL13 and TNF during LGTV infection were dependent on TLR7, whereas the type I IFN response was not (Fig. 7). Thus, TLR7 appeared to be involved primarily in the generation of a pro-inflammatory cytokine response, and this response may have a role in controlling virus spread within the CNS.
Studies with another flavivirus, WNV, demonstrated conflicting roles for TLR7 depending on the route of inoculation. When mice were infected with WNV by the i.d. route or by mosquito transmission, TLR7 deficiency did not affect disease induction or virus levels in the CNS (Welte et al., 2009). However, when WNV was administered i.p., TLR7 deficiency resulted in increased onset of disease and an eightfold increase in viral RNA in the CNS (Town et al., 2009). This was associated with a lack of infiltrating macrophages in TLR7-deficient mice. This does not appear to be the mechanism of TLR7-mediated suppression of virus levels, as TLR7 deficiency resulted in a slight increase, not decrease, in the recruitment of peripheral cells to the CNS and no detectable change in infiltration of macrophages was observed (Fig. 6 and data not shown).
It is interesting that TLR7 deficiency increased virus burdens following both LGTV and WNV infection but only resulted in altered kinetics of disease development with WNV infection (Town et al., 2009) (Figs 1 and and2).2). This suggests that virus titres are not directly correlative with neurological disease induction in LGTV infection. It is unclear why the increase in virus load in the CNS in the absence of TLR7 is not associated with increased disease incidence or onset. Possibly, a minimum virus titre is sufficient in the CNS to induce clinical neurological disease. In support of this theory, the p.f.u. of virus (g brain tissue)−1 did not appear to increase significantly in wild-type mice from 6 days p.i. to the onset of clinical signs (Fig. 1c). Furthermore, the onset of clinical disease was only a few days apart in i.c.-inoculated mice (7–8 days p.i.) compared with i.d.-inoculated mice (10–11 days p.i.), despite a 100-fold difference in virus titre in the CNS (108 p.f.u. in i.c.-inoculated mice compared with 106 p.f.u. in i.d.-inoculated mice). Thus, a log10 increase in virus titre in the CNS may not be a significant factor in altering the course of LGTV-induced neurological disease. The difference observed between WNV and LGTV may be due either to inherent differences in viral virulence unrelated to the ability of the virus to reach the CNS or to specific anti-flavivirus immune responses that contribute to pathogenesis.
Increased cytokine production in the CNS has been associated with a number of flavivirus infections including TBEV (Růžek et al., 2011; Tigabu et al., 2010). This increase in cytokines in the CNS was associated with the increased presence of inflammatory cells in the brain (Růžek et al., 2011). In the current study, decreased levels of CCL3, CCL4, CXCL13 and TNF mRNA expression were observed in TLR7-deficient mice, even though there was no decrease in the level of inflammatory cell markers. This suggests that these cytokines are being expressed by glial cells activated by LGTV infection, rather than by inflammatory cells. Microglia readily produce several cytokines and chemokines following TLR7 stimulation (Butchi et al., 2010). Furthermore, microglial production of CCL3, CCL4 and TNF has been associated with virus infections in the CNS (Cheeran et al., 2001, 2005; D’Aversa et al., 2004; Lokensgard et al., 2001; Mack et al., 2003; Sopper et al., 1996; Szretter et al., 2009; Zhou et al., 2008). Although we did not detect increased mRNA expression of the microglial/macrophage activation marker F4/80 in brain tissue from TLR7-deficient mice (Fig. 6), we did observe Iba1 staining in areas of LGTV infection. The presence of microglia/macrophages in these areas may indicate that these cells have a role in regulating virus infection of neurons, possibly through the production of the above cytokines.
One group of cytokines known to have an active role in virus suppression are the type I IFNs including IFN-α2, IFN-α4 and IFN-β1. Interestingly, induction of the type I IFN response was altered by TLR7 deficiency, with higher levels of Ifna2 and Ifna4 mRNA in brain tissue from TLR7-deficient mice compared with wild-type mice (Fig. 7). Similarly, Ifnb1 mRNA was upregulated in LGTV-infected TLR7-deficient neurons compared with wild-type neurons (Fig. 4). In correlation with increased type I IFN production, ISGs were also upregulated in TLR7-deficient neuronal cultures and in brain tissue from infected mice (Figs 4 and and77 and data not shown). Thus, TLR7 deficiency enhanced the type I IFN response both in vivo and in vitro. The increased IFN response in vivo could be explained at least in part by the higher virus levels in TLR7-deficient mice. However, the IFN response was also substantially higher in TLR7-deficient neuronal cultures where there was no significant increase in virus levels (Fig. 4). Possibly, TLR7 signalling may inhibit the signalling pathways of other virus-recognizing PRRs. We have shown previously that TLR7 inhibits TLR9-mediated neuroinflammatory responses (Butchi et al., 2010, 2011), although the mechanism of inhibition is unknown. As other PRRs, including TLR3, RIG-I and MDA-5, can recognize flavivirus infection (Daffis et al., 2008; Fredericksen et al., 2008), it is possible that TLR7 signalling limits the response of these PRRs.
The inverse correlation between virus load and type I IFN mRNA expression in both the CNS and in neuronal cultures suggests that high levels of type I IFN and induction of ISG expression may not be the only limiting factors in controlling flavivirus infection in neurons. The lack of impact of type I IFNs on LGTV infection may be due to the ability of flaviviruses to limit the IFN response. The non-structural protein NS5 of flaviviruses inhibits the JAK-STAT signalling pathway dependent on the IFN-α/β receptor (IFNAR) and limits virus recognition by the IFN-induced protein with tetratricopeptide repeats (IFIT) family members (Best et al., 2005; Daffis et al., 2010). Thus, high levels of type I IFNs may not limit LGTV replication in neurons after establishment of infection, either in vivo or in vitro.
In addition to the increased IFN response, a twofold increase in Cd3ϵ and Cd8α mRNA expression in brain tissue was observed in LGTV-infected TLR7-deficient mice compared with wild-type mice (Fig. 6). Although this difference was not discernible by immunohistochemical analysis, it does suggest an increase in the number of CD8+ T-cells in the brain in TLR7-deficient mice compared with wild-type mice. CD8+ T-cells can contribute to the disease state in the CNS through direct and/or indirect damage to virus-infected neurons (Rowell & Griffin, 2002; Sobottka et al., 2009). However, CD8+ T-cells are also important for the control of virus replication through non-lytic mechanisms that do not appear to damage neurons but do control neuronal virus replication (Binder & Griffin, 2001; Permar et al., 2003; Sitati et al., 2007). Thus, the increased numbers of CD8+ T-cells in TLR7-deficient mice could regulate virus levels in infected neurons without contributing to virus-mediated damage.
In summary, TLR7 deficiency did not substantially alter the onset or incidence of LGTV-induced clinical disease but did result in a log10 increase in virus titres in brain tissue. The latter effect was similarly observed after i.c. inoculation and thus was not due to altered neuroinvasion in TLR7-deficient mice. The heightened virus load did not correlate with a decrease in cellular inflammation or type I IFN production but was associated with a decrease in pro-inflammatory cytokines and chemokines including TNF, CCL3, CCL4 and CXCL13. From these observations, we conclude that TLR7 deficiency alters the neuroinflammatory response to LGTV infection with a corresponding increase in virus levels in the CNS.
TLR7-deficient inbred Rocky Mountain White (IRW) mice have been described previously (Lewis et al., 2008). IRW mice were used as wild-type controls. All of the animal procedures were approved by and conducted in accordance with the Louisiana State University Animal Care and Use Committee (ACUC) guidelines or the Rocky Mountain Laboratories ACUC guidelines under protocols LSU06-120 and RML2008-46.
LGTV (strain TP21) was a kind gift from Dr Alexander Pletnev (Laboratory of Infectious Disease, NIAID, NIH, USA) and was stored at −80 °C until use. After thawing, 5 µl original virus stock was injected into each lateral ventricle of anaesthetized 2-day-old C57BL/6 mice. At 4 days p.i., the mouse pups were euthanized. One half of each brain was collected and frozen at −80 °C. The remaining hemisphere was homogenized and centrifuged briefly and the supernatants collected. Aliquots of pooled virus suspension supernatant were inoculated onto confluent Vero cell (C76; ATCC CRL-1587) monolayers. At 3 days p.i., the supernatant was collected from infected Vero cell cultures, aliquotted and stored at −80 °C for virus quantification by immunofocus assay. This virus stock (108 p.f.u. ml−1) was used for all experiments. For mock-infected controls, supernatants from uninfected Vero cells that had been treated in the same manner as virus stocks were utilized.
For i.c. inoculations, 14-day-old mouse pups were injected with 102 p.f.u. LGTV by inoculating 5 µl diluted virus stock into each cerebral ventricle. Mock-infected controls were treated with diluted supernatants from uninfected Vero cells. For i.d. inoculations, 21-day-old mouse pups were injected with 50 µl LGTV containing 103 or 105 p.f.u. Mice were observed daily for signs of encephalitis. Mice showing clinical signs were euthanized and their brains removed for histological or molecular analysis.
Mice were followed daily for signs of neurological disease. Clinical disease was defined as obvious signs of distress, hunched posture, weight loss, and inability to move and/or maintain an upright position. Upon development of clinical signs or at pre-determined times p.i., mice were euthanized under isoflurane anaesthesia and their tissues removed. For histological analysis, brains were removed and fixed in 10% neutral-buffered formalin. For RNA analysis, brains were removed and transected into two hemi-brains. Each hemi-brain was placed in 2 ml RNAlater (Qiagen) in 24-well plates overnight. The brains were then snap frozen in N2, placed in 2 ml round-bottomed tubes, weighed and frozen at −80 °C until processing. For the virus immunofocus assay, whole or hemi-brains were weighed and homogenized in 0.5 ml minimal essential medium (MEM) with 10% FBS and 2 mM l-glutamine. Homogenized tissue was centrifuged at high speed (4500 g) for 15 min. The supernatant was aspirated into cryotubes, the volume recorded and the samples frozen at −80 °C for future processing.
Mouse brains were fixed in 10% neutral-buffered formalin, embedded in paraffin and cut into 4 µm sections that included cerebrum, cerebellum and brainstem. Tissue sections were adhered to coated microscope slides and stained with H&E using an automated histological stainer. Stained slides were examined in a blind fashion by a board-certified veterinary pathologist for histological evidence of neuropathology. Slides were scored on a scale of 0–3. For all scales, 0 was considered normal, with no histological changes. Meningitis was scored as: 1, scattered meningeal inflammatory cells infiltrates; 2, diffuse one- to two-cell-layer cell infiltrate; and 3, diffuse greater than two-cell-layer cell infiltrate. Perivascular cuffing was scored as: 1, plump endothelial cells with scattered one- to three-cell-layer perivascular cuffs in <20% of the vessels; 2, plump endothelial cells with frequent one- to three-cell-layer perivascular infiltrates in 20–50% of the vessels; and 3, plump endothelial cells with frequent one- to three-cell-layer perivascular infiltrates in >50% of the vessels. Cerebral neuronal necrosis was scored as: 1, a mean of one necrotic neuron per high-powered field (of five fields counted); 2, two to three necrotic neurons per high-power field; 3, more than three necrotic neurons per high-power field. Hippocampal neuronal necrosis was scored as: 1, scattered single-cell neuronal necrosis; 2, clusters of up to 20 necrotic neurons; and 3, clusters of >20 necrotic neurons.
Immunohistochemical analysis was competed using an antigen-retrieval protocol described previously (Du et al., 2010). LGTV infection was detected using Russian spring–summer virus (RSSV) ascites fluid (ATCC VR-1264AF) using Alexa Fluor 488- or Alexa Fluor 555-conjugated goat anti-mouse secondary antibodies (Invitrogen). Microglia/macrophages were detected using anti-Iba1 antibody (Wako) and Alexa Fluor 594-conjugated donkey anti-rabbit secondary antibodies. All sections were counterstained with DAPI at a concentration of 100 ng ml−1. Slides incubated without primary antibodies or with isotype controls were used to confirm specificity.
Primary cultures of cortical neurons were prepared from 16-day-gestation mouse embryos as described previously (Butchi et al., 2010). Cells were plated at a concentration of 8×105 cells ml−1. Following attachment (2 h), the medium was replaced with neurobasal medium with 2% B-27 and 0.5 mM glutamine. The purity of the neuronal population was confirmed as reported previously (Butchi et al., 2010) by positive staining for microtubule-associated protein 2 (MAP-2) as well as the absence of Iba1-positive cells and only a few GFAP-positive cells per well. At 1–2 days post-plating, neurons were infected at an m.o.i. of 0.01 with LGTV. Mock-infected controls were treated with supernatant from uninfected Vero cells diluted in medium to the same concentration as virus stocks. Supernatants and cells were harvested at the time points indicated and used to test for virus or for mRNA analysis.
RNA was isolated from tissues as described previously (Butchi et al., 2008). First-strand cDNA was synthesized using 100 ng cleaned-up RNA and analysed with PCR arrays (SABiosciences) on an ABI PRISM 7900 Sequence Detection System (Applied Biosystems). Data were calculated as the fold difference for the treatment groups compared with mock groups. For time-point analysis, RNA samples were converted to cDNA using an iScript Reverse Transcription kit (Bio-Rad Laboratories) following the manufacturer’s instructions. The primers to detect Gapdh, Cxcl13, Ccl3, Ccl4, Tnf, Ifnb1, Ifna4 and Irf1 cDNA were designed using Primer3 software with a melting temperature of 60 °C for all primers. SYBR Green dye with ROX (Bio-Rad) was used for the measurement of real-time PCR amplification. Viral cDNA was detected using forward primer 5′-GTCTCCGGTTGCAGGACT-3′, reverse primer 5′-CTCGGTCAGTAGGATGGTGTTG-3′ and the probe 5′-FAM-CCACAGGAAGATCAGTGAG-NFQ-3′. Universal PCR master mix (Applied Biosystems) was used for real-time PCR amplification. Data for each sample were calculated initially as the percentage difference in threshold cycle (CT) value (ΔCT=CT of Gapdh−CT of gene of interest). Data were plotted as the mean percentage of Gapdh values for each gene of interest for each sample.
For quantification of replication-competent virus, serial tenfold dilutions of each sample (supernatant or brain homogenate) were added to confluent Vero cell monolayers. Following a 1 h incubation, the cells were washed and overlaid with 0.8% methylcellulose in Dulbecco’s MEM with 2% FBS. After 5 days, methanol was added to each well. The wells were washed gently to remove the methylcellulose and blocked using OptiMEM I/2% bovine growth serum. Virus was detected using RSSV immune ascites fluid followed by a secondary antibody of HRP-conjugated goat anti-mouse IgG. Foci were detected using AEC substrate (Biomedia). The results were expressed as p.f.u. (g brain)−1.
The authors thank the Louisiana State University animal caretakers for providing excellent animal care throughout these experiments. The project was supported in part by Louisiana State University School of Veterinary Medicine Competitive Organized Research Program and in part by the Division of Intramural Research, National Institutes of Health, National Institute of Allergy and Infectious Diseases.