Oligonucleotide array analysis of Toll-like receptors and associated signalling genes in Venezuelan equine encephalitis virus-infected mouse brain

Venezuelan equine encephalitis virus (VEEV) is a member of the genus Alphavirus in the family Togaviridae. VEEV is an emerging pathogen (Weaver et al., 2004), is highly infectious when delivered by aerosol (Steele et al., 1998) and has been developed for use as a bioweapon (Hawley & Eitzen, 2001). The virus may cause central nervous system (CNS) disease in horses and is transmitted to humans by mosquito. VEEV enters the CNS primarily through the olfactory tract (Charles et al., 2001; Ryzhikov et al., 1995) and causes encephalitis characterized by neuronal cell death, severe mononuclear cell cuffing of cerebral vessels, meningitis and demyelinating lesions. It is hypothesized that VEEV-induced reactive gliosis and inflammation also leads to the observed secondary neuronal damage in the brain independent of direct virus infection of neuronal cells (Charles et al., 1995, 2001; Davis et al., 1994; Grieder et al., 1995; Jackson et al., 1991; Jackson & Rossiter, 1997; Schoneboom et al., 1999, 2000a). Mechanisms underlying the inflammatory response to VEEV infection in the brain are poorly understood. Thus, the characterization of the innate immune response in the CNS following VEEV infection is important in understanding how VEEV mediates tissue damage in the brain.

Toll-like receptors (TLRs) have been implicated in brain pathogenesis of several neurotropic viruses such as cytomegalovirus, rabies virus, herpes simplex virus (HSV) and West Nile virus (WNV) and in influenza-associated encephalopathy (Tabeta et al., 2004; Wang et al., 2004; Ménager et al., 2009; Sørensen et al., 2008; Daffis et al., 2008; Town et al., 2009; Hidaka et al., 2006). In our previous microarray study, we reported activation of genes involved in response to viral infection, inflammation, antigen presentation and apoptosis following VEEV infection in the mouse brain (Sharma et al., 2008). Although direct upregulation of TLRs was not detected, upregulation of some of the TLR downstream signalling genes such as chemokines and interferon (IFN) regulatory factors (IRFs) were reported. Since TLRs are implicated in innate immune response and pathogenesis of other neurotropic viruses and their expression in the brain during VEEV infection is largely unknown, we further evaluated TLRs and TLR-associated signalling in VEEV-infected mouse brains. TLR-specific oligonucleotide arrays, more sensitive RT-PCR and immunohistochemistry techniques were used to evaluate TLRs and their downstream signalling genes in a VEEV-infected mouse brain. A mouse model of VEEV infection was used as it mimics the biphasic infection, i.e. initial replication in peripheral organs followed by entry and replication in the CNS, as is observed in humans.

In this study, we demonstrate the upregulation of multiple TLRs and other genes that are involved in TLR downstream signalling following VEEV infection of mouse brains. TLR 1, 2, 3, 7 and 9 were upregulated in VEEV-infected brains, accompanied by upregulation of TLR interacting adaptor (MyD88), kinases (IRAK1 and Tbk1), transcription factors (Nfkb1, IRF-1 and -7) and target genes (IFN-β, Mcp1, Cxcl10, Caspase8 and IL12 α/β). Collectively, these results show that multiple TLRs and a potential MyD88-dependent signalling may be implicated in the innate immune response and inflammation against VEEV infection in the brain.

Animals.
Three- to five-week-old male CD-1 mice were obtained from Charles River Laboratories. All experiments were carried out in a bio-safety level 3 facility at Uniformed Services University of the Health Sciences (USUHS), Bethesda, MD, and in accordance with the Guide for the Care and Use of Laboratory Animals (Committee on Care and Use of Laboratory Animals of the Institute of Laboratory Animal Resources, National Research Council, NIH Publication No. 86-23, revised 1996).

Virus, anaesthesia and infection procedure
Virus.
V3000, a virus generated from a full-length cDNA clone of VEEV subtype IA/B, that was used in this study was kindly provided by Dr Franziska B. Grieder, USUHS, Bethesda, MD. V3000 is infectious and virulence is similar to the VEEV subtype IA/B strain (Grieder et al., 1995).

Anaesthesia.
Mice were anaesthetized by inhalation anaesthesia using the isoflurane by drop jar method. Briefly, a transparent glass jar was used containing a dry sponge overlaid with a wire gauge at the bottom. Several layers of paper towel were then placed on the top of the wire gauge. Isoflurane (3 ml) was added directly onto the sponge, the lid was closed and the jar was allowed to saturate with the isoflurane vapours for about 5 min. Mice were individually anaesthetized by placing each animal in the jar and closing the lid. Mice were closely monitored for the stabilization of breathing and then taken out of the jar and injected in the foot pad with virus. After every three animals were anaesthetized, 1 ml isoflurane was added to the sponge.

Infection procedure.
Virus was diluted in 1x Dulbecco's PBS (DPBS) containing Ca²⁺ and Mg²⁺ (Gibco). One thousand p.f.u. V3000 in a final volume of 25 µl was injected in the left rear footpad of each mouse. Control animals were injected with 25 µl 1x DPBS containing Ca²⁺ and Mg²⁺.

Histopathology and immunohistochemistry (IHC).
Five mice (n=25) were sacrificed at each of several time points after infection (24, 48, 72, 96 and 120 h p.i.) and control animals (n=5) were sacrificed at 120 h p.i. Animals were perfused first with 5 ml ice-cold 1x PBS (Gibco) followed by 10 ml cold 10 % buffered neutral formalin (BNF) (VWR). The whole brain was removed and fixed in 10 % BNF for 3–4 weeks. Tissues were embedded in paraffin blocks and 5 µm-thick sections were prepared. Immunostaining was performed using primary antibodies against TLR 3 (1 : 250), TLR 7 (1 : 250), TLR 9 (1 : 100), IL12 (1 : 200), MyD88 (1 : 500) and Mcp1 (1 : 750) (United States Biological). For VEEV-specific staining, rabbit polyclonal antiserum (1 : 10000), raised against VEEV, eastern equine encephalitis virus, western equine encephalitis virus and sindbis virus (kindly provided by Dr Cindy Rossi and Dr George Ludwig, United States Army Medical Research Institute for Infectious Diseases, Frederick, MD), was used as described previously (Sharma et al., 2008). An indirect avidin-biotin-immunoperoxidase technique was used for IHC using Vectastain ABC Elite kit (Vector Laboratories) as specified by the manufacturer. As a negative control, separate sections from each test were incubated with species-matched normal serum. Tissues were counter-stained with Gill modified haematoxylin (EMD Chemicals) or haematoxylin QS (Vector Laboratories).

RNA isolation.
Three mice (n=15) were sacrificed at each of several time points after infection (24, 48, 72, 96 and 120 h p.i.). The right hemisphere of the mouse brain was removed and immediately frozen at –80 °C. Control mice (n=3) were sacrificed at 96 h p.i. Total RNA was extracted from frozen brain tissues using the TriZol kit (Invitrogen) according to the manufacturer's protocol. RNA was further purified using the RNeasy mini kit (Qiagen). RNA was quantified spectrophotometrically using a Beckman DU640 Spectrophotometer (Beckman Instruments). RNA was checked for degradation and genomic DNA contamination by electrophoresis on a 1 % agarose formaldehyde gel.

TLR signalling pathway-specific gene expression.
The pathway-focused oligo GEArray, mouse Toll-like receptor signaling microarray kit (Catalogue no. OMM-003) was used as per the manufacture's protocol. Each array had 112 genes spotted on the array membrane (full details in Supplementary Table S1, available in JGV Online). Three biological replicates were performed. Briefly, a cRNA probe was synthesized from 1 µg RNA using the TrueLabelling-AMP linear RNA amplification kit. The amplified cRNA was purified using a spin column (ArrayGrade cRNA Cleanup kit; all materials from SuperArray Biosciences) and quantified spectrophotometrically. Array membranes were prehybridized for 2 h at 60 °C with GEAHyb hybridization solution followed by overnight hybridization with 4 µg cRNA mixed in 750 µl hybridization buffer at 60 °C in a hybridization chamber (Techne hybridizer HB-1D). Array membranes were then washed for 15 min each at 60 °C with pre-warmed wash solution 1 (2x SSC, 1 % SDS) and wash solution 2 (0.1x SSC, 0.5 % SDS). Blocking was done for 2 h at room temperature with 2 ml blocking solution Q, followed by incubation with 2 ml binding solution (1 : 15000 AP-Strep in 1x buffer F). Array membranes were then washed with 1x buffer F and rinsed with buffer G. Detection was performed by incubating membranes with 1 ml of CDP star substrate supplied with the chemiluminescence detection kit (SuperArray Biosciences), for 3 min at room temperature. All hybridizations and washes were carried out with rotation at 10–20 r.p.m. X-ray films (Kodak Scientific Imaging Film) were exposed to the membrane for different time periods and films were developed using the Kodak Image developer. X-ray images were scanned and converted to digital images (Supplementary Fig. S1, available in JGV Online) for further analysis.

Analysis.
Images were analysed using the GEArray Expression Analysis Suite (SuperArray Biosciences). Gene densities were expressed as the average density (total density divided by number of pixels). Background detection was done locally, i.e. each expression value was individually subtracted from the value from the area outside the capture grid but within the spot cell area. The minimum positive value and common mean were adjusted to 10 and 100, respectively. Data normalization was done with a non-modulating housekeeping gene, glyceraldehyde-3-phosphate dehydrogenase (GAPDH). A gene was considered absent where the mean density of the spot was less than the mean value of the local backgrounds of the lower 75 percentile of all non-bleeding spots. All other spots were considered present.

Reverse transcription and PCR.
cDNA was synthesized using the Superscript first strand synthesis system for RT-PCR kit (Invitrogen). Two to four biological replicates were performed. Briefly, primer mix (1 µg RNA, dNTP and oligo DTs) was incubated at 65 °C for 5 min then mixed with the reaction mixture (10x PCR buffer, 25 mM MgCl₂, 0.1 M DTT, 40 U RNase inhibitor) and incubated at 42 °C for 2 min. cDNA synthesis was done using RT enzyme (SSII) at 42 °C for 50 min. The reaction was stopped by incubating at 70 °C for 15 min. Residual RNA was digested with Escherichia coli RNase H at 37 °C for 20 min and samples were stored at –20 °C.

PCR was performed to validate the genes that were up- or downregulated by twofold in the microarray. Primers and conditions used for the different genes are described in Supplementary Table S2 (available in JGV Online). For TLR PCR, one additional step of non-specific amplification was carried out for 5–10 cycles of 95 °C for 60 s, 45 °C for 90 s, 72 °C for 90 s, followed by specific amplification. PCR products were visualized by electrophoresis using a 1.2 % agarose gel and staining with ethidium bromide. Gene expression of PCR gel images was quantified by using the Scion Imaging software. Further PCR products were sequenced and checked for specific amplification by BLAST analysis and alignment against known gene sequences using Clone Manager Professional Suite Version 8 software (data not shown). Briefly, PCR products were pooled from 72 and 96 h for each sample and purified using the QIAprep Spin Miniprep kit (Qiagen). The sequencing reaction was performed using BD BigDye Version 3.1 (Applied Biosystems). The product was then purified using Performa DTR Gel Filtration Cartridges (Edge BioSystems). Sequencing was carried out on a DNA Sequencer 3100 (Applied Biosystems) at USUHS.

Statistical analysis.
Mean fold expression was determined from three replicates for microarray and two replicates for PCR. Student's t-test was done using individual gene expression values from VEEV-infected and control samples to evaluate statistical significance. P-values ≤0.05 were considered significant.

TLR signalling pathway-specific gene expression
TLR receptor and signalling pathway-specific oligonucleotide array analysis of RNA samples isolated from the brains of VEEV-infected mice at 96 h p.i. was done. Results showed the upregulation of 43 genes by ≥1.5-fold over the control. These genes have been grouped according to their function in TLR signalling (Table 1). These groups are as follows. (i) TLRs. TLR 3 was upregulated by 10-fold and TLR 7 expression was induced upon VEEV infection. (ii) Downstream pathways and target genes. NF-κB pathway: expression of Chuk, Hrb, Nfkb1 and Nfkbil1 was upregulated in VEEV-infected brains. Ccl2 (or Mcp1), IFN-β1, IFN-γ, Il1β, Il1r1, IL12β, Map3k14, Nfkb2, Nfkbie and Rela expression were induced upon VEEV infection. JNK/p38 pathway: expression of Fos, Jun and Map2k3 was upregulated. Map2k6, Mapk10 and Mapk12 were downregulated in VEEV-infected brains. NF/IL6 pathway: Ptges expression was upregulated. Cebpb, Il6 and Ptgs2 expression was induced upon VEEV infection. IRF pathway: Tbk1 expression was upregulated. Chemokine (C-X-C motif) ligand 10 (Cxcl10) or IP10, IRF-1 and IRF-7 expression was induced upon VEEV infection. (iii) Adaptors and TLR interacting proteins. Hmgb1 and MyD88 expression was upregulated. Expression of Hspd1, Mal and Tollip was downregulated. Cd14 and Pglyrp1 expression was induced in VEEV-infected brains. (iv) Effectors. IRAK-1 and Prkr expression was upregulated. Pkr expression was downregulated. Casp8 and Map3k7 expression was induced in VEEV-infected brains. (v) Regulation of adaptive immunity. Cd86 was upregulated in VEEV-infected brains.

Table 1. TLR signalling pathway-specific genes in VEEV-infected mice brains at 96 h p.i Microarray was performed using brain RNA samples from 96 h p.i., as extensive inflammation and VEEV antigen was observed in brains at this time. Total RNA was isolated and the array was performed as described in Methods. Three biological replicates were taken from two different experiments. Values are expressed as mean±SEM. Student's t-test was done to determine statistical significance. NE, No expression; NA, not applicable as expression was only in either infected or control brains.

RT-PCR and PCR
Genes involved in TLR signalling that were up- or downregulated, as seen by oligonucleotide array analysis, were validated and their expression kinetics were studied using semiquantitative RT-PCR. In addition, PCR analysis was carried out for all the TLRs included on the membrane, and also for Ticam1, an adaptor in TLR signalling, which was not present on the microarray blot. Genes that were sequenced from PCR products are given with their accession numbers in parentheses. TLR 1 (NM_030682.1[GenBank] ), 2, 3 (NM_126166.2), 7 (NM_133211.2) and 9 (NM_031178.1) expression was significantly upregulated in VEEV-infected brains (Fig. 1a and Supplementary Fig. S2). TLR 4, 5 and 6 (NM_011604.2) did not show any statistically significant difference from controls (data not shown). The TLR-interacting proteins genes MyD88 (NM_010851.2[GenBank] ) and Cd14 were significantly upregulated, whereas Ticam1 (NM_174989.2) upregulation was not statistically significant (Fig. 1b and Supplementary Fig. S2). Target genes such as IL12-α/β (NM_008351.1/NM_008352.1), Cxcl10 (NM_021274.1[GenBank] ), Mcp1 (NM_011333.1) and IFN-β were also significantly upregulated upon VEEV infection of the brain (Fig. 1c and Supplementary Fig. S2). IFN-α (NM_010502[GenBank] ) expression was also upregulated but this was not statistically significant. Expression of effectors such as Casp8 (NM_009812.1) and Cd86 (NM_019388.2) was significantly upregulated at 72 and 96 h p.i. Pkr expression was downregulated as the disease progressed (Fig. 1d and Supplementary Fig. S2). Transcription factors such as IRF-1 (NM_008390.1), IRF-7 (NM_016850.1) and Nfkb1 (NM_008689.1) were significantly upregulated in VEEV-infected brains (Fig. 1e and Supplementary Fig. S2). Fos (NM_010234.2[GenBank] ), Jun (NM_010591.1) and IRAK-1 (NM_008363.1) were significantly upregulated in VEEV-infected brain. Map2k6 (NM_011943.1) expression was significantly downregulated at 96 h p.i. Mapk12 was downregulated at 48 h p.i. but its expression level was comparable to the control at 72 and 96 h p.i. (Fig. 1e and Supplementary Fig. S2).

(42K): Fig. 1. Quantification of gene expression in VEEV-infected brains. Quantitative estimation was done by densitometry analysis of the PCR products shown in Supplementary Fig. S2. Values were normalized to the corresponding values of the housekeeping genes for individual samples. Values are expressed as mean±SEM. *P≤0.05. Proteins analysed are: (a) TLRs, (b) adaptors and TLR-interacting proteins, (c) target genes of TLR signalling, (d) effectors in TLR signalling and (e) TLR pathway downstream signalling genes.

Histopathology and IHC
Brain sections were evaluated for inflammation on the basis of perivascular cuffing, endothelial swelling, oedema and compromise of blood vessel endothelial lining. Inflammation started with perivascular cuffing and meningeal inflammation at 48 h p.i., and was established in the olfactory and frontal areas of the brain at 72 h p.i. By 96 h p.i., inflammation was observed throughout the brain. Extensive inflammation was observed at 120 h p.i. (Fig. 2a–d and Supplementary Fig. S3a, b). VEEV-specific antigen staining showed virus presence in the olfactory and prefrontal areas of the brain at 48 h p.i. The number of neuronal and mononuclear cells infected with VEEV increased as the disease progressed and the VEEV antigen was present throughout the brain by 96 and 120 h p.i. (Fig. 2e–h and Supplementary Fig. S3c, d). Staining for TLR 3 (Fig. 3a–d and Supplementary Fig. S4a, b), TLR 7 (Fig. 4a–d and Supplementary Fig. S5a, b), TLR 9 (Fig. 3e–h and Supplementary Fig. S4c, d), MyD88 (Fig. 4a–d and Supplementary Fig. S5c, d), Mcp1 (Fig. 5a–d and Supplementary Fig. S6a, b) and IL12 (Fig. 5e–h and Supplementary Fig. S6c, d) showed expression of these molecules localized in and around the endothelial lining of inflamed blood vessels. TLR 3, TLR 9, Mcp1 and IL12 expression could also be seen in mononuclear cells at 96 and 120 h p.i.

(100K): Fig. 2. Inflammation and evidence of VEEV infection in the brains of VEEV-infected mice. The images shown are of mid-brain sections and are representative images of samples from five animals. Haematoxylin and eosin-stained sections of uninfected brain (a) or VEEV-infected brain at 72 (b), 96 (c) and 120 (d) h p.i. were evaluated for inflammation. Thickening of the endothelial lining of microvessels and oedema around the microvessels was observed at 72 h p.i., shown by an arrow on (b). Perivascular cuffing and migrating mononuclear cells were observed at 96 h p.i., shown by an arrow in the insert on (c). Inflammation was spread throughout the brain at 120 h p.i. (d). Hyperchromatic neurons and increased numbers of migrating mononuclear cells were also observed around the inflamed microvessels. VEEV-specific-antigen staining in uninfected brain (e) or infected brain at 72 (f), 96 (g) and 120 (h) h p.i. is also shown. Several foci of VEEV infection were observed at 72 h p.i. and VEEV was localized in hyperchromatic neurons (insert), endothelial linings and mononuclear cells around the inflamed microvessels (f). Scattered foci of VEEV were observed throughout the brain at 96 h p.i. VEEV was localized in mononuclear cells around the microvessels [arrow and insert in (g)]. VEEV was localized throughout the brain and the number of VEEV-infected foci increased in the brain at 120 h p.i. Areas highlighted by squares are magnified in the inserts. Bars, 200 µm.

(96K): Fig. 3. Evidence of TLR 3 and TLR 9 expression in VEEV-infected mouse brains. Brain sections were stained for TLR 3- and TLR 9-specific antigens as described in Methods. The images shown are of mid-brain sections and are representative images of samples from five animals. TLR 3 from uninfected brain (a) and VEEV-infected brains at 72 (b), 96 (c) and 120 (d) h p.i. showed that TLR 3 was localized in the mononuclear cells around the blood vessels at 96 and 120 h p.i. [arrow and insert in (c) and (d)]. Diffused TLR staining was also observed adjacent to the microvessels at 96 h p.i., shown by the arrow in (c). The number of microvessels and mononuclear cells around the vessels that were positive for TLR 3 staining increased at 120 h p.i. TLR 9-specific antigen staining in uninfected brain (e) and in VEEV-infected brains at 72 (f), 96 (g) and 120 (h) h p.i. showed that TLR 9 was localized in the mononuclear cells and the endothelial linings of the microvessels at 120 h p.i. Areas highlighted by squares are magnified in the inserts. Bars, 200 µm.

(99K): Fig. 4. Evidence of TLR 7 and MyD88 expression in VEEV-infected mouse brains. Brain sections were stained for TLR 7 and MyD88-specific antigen as described in Methods. The images shown are of mid-brain sections and are representative images of samples from five animals. TLR 7 in uninfected brain (a) and VEEV-infected brains at 72 (b), 96 (c) and 120 (d) h p.i. showed that TLR 7 was localized in the mononuclear cells around the blood vessels at 96 h p.i. and also in the endothelial linings of the microvessels at 120 h p.i., shown by arrows and inserts in (c) and (d). MyD88 in uninfected brain (e) and VEEV-infected brains at 72 (f), 96 (g) and 120 (h) h p.i. showed that MyD88 was localized in the endothelial linings of the microvessels at 72 h p.i. [arrow and insert in (f)]. Its expression increased and was localized in the endothelial linings of the microvessels and mononuclear cells at 96 h p.i. (g). MyD88 expression further increased at 120 h p.i. (h). Areas highlighted by squares are magnified in the inserts. Bars, 200 µm.

(100K): Fig. 5. Evidence of Mcp1 and IL12 expression in VEEV-infected mouse brains. Brain sections were stained for Mcp1and IL12-specific antigen as described in Methods. The images shown are of mid-brain sections and are representative images of samples from five animals. Mcp1 in uninfected brain (a) and VEEV-infected brains at 72 (b), 96 (c) and 120 (d) h p.i. showed that Mcp1 was localized in the cells around and endothelial linings of the inflamed microvessels at 96 h p.i., shown by arrow and insert in (c). Mcp1 localization in endothelial linings of the microvessels and mononuclear cells increased at 120 h p.i. (d). IL12 in uninfected brain (e) and VEEV-infected brains at 72 (f), 96 (g) and 120 (h) h p.i. showed that IL12 was localized in the mononuclear cells in and around the inflamed microvessels at 72, 96 and 120 h p.i. (arrows and inserts). The expression increased as the disease progressed. Areas highlighted by squares are magnified in the inserts. Bars, 200 µm.

TLRs have been implicated in the inflammatory responses following infection or injury of the brain (Crack & Bray, 2007). TLR 2, 3, 4 and 7 participate in neuroinflammatory reactions resulting in the increased severity of disease, and are also implicated in inducing apoptosis in epithelial and microglial cells (Yoon et al., 2008; Babcock et al., 2006; Kurt-Jones et al., 2004; Park et al., 2006; Wang et al., 2004; Walter et al., 2007; Caso et al., 2007; Chakravarty & Herkenham 2005; Butchi et al., 2008; Aravalli et al., 2007; Srivastava et al., 2005; Kaiser & Offermann 2005; Haase et al., 2003). Inflammation following VEEV infection is important in conferring protection against VEEV, but neurodegeneration following VEEV infection in the brain is also widely believed to be due to excessive inflammatory response to virus presence in the brain (Charles et al., 2001; Grieder & Vogel, 1999; Jackson et al., 1991; Schoneboom et al., 1999, 2000a, b; White et al., 2001). In an earlier study by our laboratory, although direct upregulation of TLRs was not observed, several proinflammatory cytokines, chemokines and apoptotic genes that may also be induced in TLR signalling were upregulated (Sharma et al., 2008). We decided to evaluate TLR signalling, since the expression of TLRs in VEEV-infected brains is largely uncharacterised, despite not only our prior observations but also studies done by numerous others that demonstrate the association of TLRs with CNS inflammation.

The expression of TLRs 1–5 has been reported in the blood of VEEV-infected macaques, and the VEEV replicon particles have been shown to induce TLR and cytokine response in T cells and natural killer cells (Hammamieh et al., 2007; Saikh et al., 2003). In the present study, we demonstrate that multiple TLRs such as TLR 1, 2, 3, 7 and 9 were significantly overexpressed following VEEV appearance in the mouse brain. TLR expression increased with the increase in the viral load in the brain and was localized in and around the endothelial linings of the inflamed blood vessels and in mononuclear cells. These results are consistent with the other neurotropic viruses such as Semliki forest virus (SFV), rabies virus and HSV, where upregulation of TLRs has been established in the brain following viral infection (McKimmie et al., 2005; Prehaud et al., 2005). The expression of TLR 1, 2 and 4, which are thought to play a role in recognizing non-viral pathogens, was also upregulated following VEEV infection in this study. Similar observations have been reported in other neurovirulent viruses such as SFV and HSV (McKimmie et al., 2005; Bottcher et al., 2003; Mansur et al., 2005). The specific roles that these TLRs may play in recognition of or in the innate immune response against VEEV would need further exploration.

A functional IFN system is important in protection against VEEV, and mice lacking an active IFN system are highly sensitive to VEEV infection (Grieder & Vogel, 1999; Schoneboom et al., 2000a, b). IFN-dependent expression of TLR 3 and 9 has been shown in another alphavirus, SFV (McKimmie et al., 2005). TLR 3 and 9 and IFNβ were also upregulated in this study. However, whether the IFN system plays a role in the activation of TLR 3 and 9 in VEEV-infected brains remains to be evaluated.

MyD88-dependent antigen presentation and activation of CD8⁺ T-cell responses in alphaviruses and MyD88-dependent inflammatory response to brain injury have been reported previously (Chen et al. 2005; Koedel et al. 2007). In this study, MyD88 expression was upregulated and was localized in and around the endothelial linings of the inflamed blood vessels. The increase in Ticam-1 transcription was not statistically significant. This leads us to believe that VEEV infection depicts a bias towards a MyD88-dependent signalling pathway over a Ticam-1-dependent signalling pathway in the brain. Further, these results are consistent with other neurotropic virus infections such as HSV- and vesicular stomatitis virus (VSV)-induced encephalitis, where mice lacking MyD88 showed enhanced susceptibility to HSV-induced encephalitis (Mansur et al., 2005; Lang et al., 2007). Specific roles of MyD88 and Ticam-1 receptor and the significance of their presence or absence will require further investigation.

Genes that are involved in the TLR downstream signalling, such as transcription factors NF-κB and IRF and their activation-mediating kinase, IRAK1 (Honda et al., 2004; Moynagh, 2005; Takeda & Akira 2005), were also upregulated in the VEEV-infected brains. Several chemokines and cytokines that are under transcriptional regulation of NF-κB and IRF, such as the Mcp1, IL12-α, IL12-β, IFN-β1 and Cxcl10, were also significantly overexpressed in the VEEV-infected brains. These results are consistent with previous reports by us and others, in which upregulation of cytokines and chemokines such as IFN-β and Cxcl10 have been shown in VEEV-infected animals (Grieder & Vogel, 1999; Hammamieh et al., 2007; Koterski et al., 2007; Moynagh, 2005; Sharma et al., 2008).

The endothelial lining of brain microvessels plays a significant role in CNS inflammation during injury or infection. IHC analysis showed the localization of TLR, cytokine and chemokine protein expression in the endothelial linings of the inflamed blood vessels in VEEV-infected brains, thus indicating a dynamic role of brain microvessel endothelium in VEEV pathogenesis. Mcp1 is a chemokine that has been implicated in the direct alteration of the blood–brain barrier (BBB) and it also serves as a chemotactic signal for leukocytes and macrophages in CNS inflammation (Eugenin et al., 2006; Song & Pachter, 2004; Stamatovic et al., 2005). Mcp1 expression in this study was also associated with the inflamed and the BBB-compromised blood vessels, indicating a similar role for Mcp1 in VEEV-infected brains. Mononuclear cells were also positively stained for TLR 3, TLR 9, IL12 and Mcp1 expression, and their presence and association with the inflamed vessels indicates their potential participation in the inflammatory response in the brain. Several other TLR signalling-associated genes such as Cd14, a co-receptor for TLR 4 (Rolland et al., 2006), and Cd86 that encodes a co-stimulatory molecule (Ahmed et al., 2006; Takeda & Akira, 2005), were also overexpressed in VEEV-infected brain.

Collectively, for the first time, we have established the upregulation of multiple TLRs and their associated signalling gene activation in VEEV-infected mouse brains. These results may have a significant implication on our understanding of the VEEV-induced immune response and inflammation mechanisms in the brain.

This work was supported by grants from the US Army Medical Research and Material Command, Defense Threat Reduction Agency and the US–India Foreign Currency Fund from the US Department of State. The authors acknowledge the help received from Dr D. Khetawat in PCR product sequencing. The authors are thankful to Drs Clifford Snapper and Mathanraj Packiam and Ms. Sai P. Majji for critically reading the manuscript. The opinions or assertions contained herein are the scientific views of the authors and should not be construed as official or necessarily reflecting the views of the Uniformed Services University of the Health Sciences or the Department of Defense, USA.