Lethal Encephalitis in Myeloid Differentiation Factor 88-Deficient Mice Infected with Herpes Simplex...

Immunopathology and Infectious Disease

Lethal Encephalitis in Myeloid Differentiation Factor88-Deficient Mice Infected with Herpes SimplexVirus 1

Daniel S. Mansur,* Erna G. Kroon,*Maurıcio L. Nogueira,† Rosa M.E. Arantes,‡

Soraia C.O. Rodrigues,§¶ Shizuo Akira,�

Ricardo T. Gazzinelli,§¶ and Marco A. Campos§

From the Departamentoes de Microbiologia,* Patologia,‡ and

Bioquımica e Imunologia,¶ Universidade Federal de Minas

Gerais, Belo Horizonte; the Centro de Pesquisas Rene Rachou,§

Fundacao Oswaldo Cruz, Belo Horizonte, Brasil; the Laboratory

of Viral Diseases,† National Institute of Allergy and Infectious

Disease, National Institutes of Health, Bethesda, Maryland; and

the Department of Host Defense,� Research Institute for Microbial

Diseases, Osaka University, Osaka, Japan

Herpes simplex virus 1 (HSV-1), a large DNA virusfrom the Herpesviridae family, is the major cause ofsporadic lethal encephalitis and blindness in hu-mans. Recent studies have shown the importance ofToll-like receptors (TLRs) in the immune response toHSV-1 infection. Myeloid differentiation factor 88(MyD88) is a critical adaptor protein that is down-stream to mediated TLR activation and is essential forthe production of inflammatory cytokines. Here, westudied the relationship between MyD88 and HSV-1using a purified HSV-1 isolated from a natural oralrecurrent human infection. We observed the activa-tion of TLR-2 by HSV-1 in vitro using Chinese hamsterovary cells stably transfected with a reporter gene.Interestingly, we found that only peritoneal macro-phages from MyD88�/� mice, but not macrophagesfrom TRL2�/� or from wild-type mice, were unableto produce tumor necrosis factor-� in response toHSV-1 exposure. Additionally, although TLR2�/�

mice showed no enhanced susceptibility to intranasalinfection with HSV-1, MyD88�/� mice were highlysusceptible to infection and displayed viral migrationto the brain, severe neuropathological signs of en-cephalitis, and 100% mortality by day 10 after infec-tion. Together, our results suggest that innate resis-tance to HSV-1 is mediated by MyD88 and may rely onactivation of multiple TLRs. (Am J Pathol 2005,166:1419–1426)

Herpes simplex virus 1 (HSV-1), from the Herpesviridaefamily, is a complex virus containing a large 140-kb DNA,which encodes 84 proteins and is the ubiquitous neuro-tropic human pathogen most commonly associated withoro-labial and ocular infections.1 The most serious infec-tion caused by HSV-1 is sporadic encephalitis,2 whichhas a mortality rate of �70%, when not treated.3 HSV-1 istransmitted primarily by contact with oral secretions. Onoral entry into skin and mucosal sites, HSV-1 replicateslocally in epithelial cells, resulting in cell lysis and localinflammatory response. After primary infection, HSV-1can travel along sensory nerve pathways and may be-come latent in the sensory ganglia, where it can eventu-ally be reactivated.3 Animal models of human HSV en-cephalitis in mice using intranasal inoculation have beendescribed.2 This inoculation pathway leads to an inflam-matory response that can be dangerous to the host.However, the precise mechanisms by which HSV-1causes death are not clear.

Toll-like receptors (TLRs) are innate immunity recep-tors linked with the response to pathogen-associatedmolecular patterns. Since the first description of TLRs inmammals, many TLR agonists have been described:peptidoglycans4 and Trypanasoma cruzi GPI anchor forTLR2,5 lipopolysaccharide (LPS) for TLR4,6–8 dsRNA forTLR3,9 flagellin for TLR5,10 and CpG DNA for TLR9.11

TLRs activate inflammatory responses and modulate im-munity by several different signal transduction pathways.The most well known pathway involves myeloid differen-tiation factor 88 (MyD88), an adapter molecule com-posed of a Toll-interleukin-1 receptor domain and a death

Supported by the Fundacao de Amparo a Pesquisa do Estado de MinasGerais (grant CBB 311/02 and CDS 185/02) and the Conselho Nacionalde Desenvolvimento Cientıfico e Tecnologico (Programa de Nucleo deExcelencia award 2400/03).

Accepted for publication January 24, 2005.

E.G.K., M.A.C., and R.T.G. are research fellows from Conselho Nacio-nal de Desenvolvimento Cientıfico e Tecnologico.

Address reprint requests to Marco A. Campos, Laboratory of Immuno-pathology, Centro de Pesquisas Rene Rachou, FIOCRUZ, Av. Augusto deLima 1715, Barro Preto, 30190-002, Belo Horizonte, MG, Brazil. E-mail:[email protected].

American Journal of Pathology, Vol. 166, No. 5, May 2005

Copyright © American Society for Investigative Pathology

1419

domain.12 MyD88 recruits the serine threonine kinase,interleukin receptor associated kinase-4, that activatestumor necrosis factor-� receptor-associated factor-6(TRAF-6) which in turn phosphorylates I�B, causing it todissociate from and leave nuclear factor (NF)-�B free inthe cytoplasm. NF-�B then translocates to the nucleusand acts as a transcription factor of innate immunity-associated genes.12,13 In addition, TLR3 appears to ac-tivate the inflammatory response through another adaptermolecule, named Toll-interleukin-1 receptor domain-con-taining adaptor-inducing interferon-�.13 This pathway isMyD88-independent, and culminates with the transloca-tion of interferon regulatory factor 3 (IRF-3) to the nucleus,leading to production of interferon (IFN)-� and IFN-induc-ible genes.13

A role for the TLR2, TLR3, TLR4, and TLR9 in theresponse to viruses has been previously estab-lished.9,12,14–18 Lund and colleagues17 showed thatgenomic HSV-2 DNA, which is closely related to HSV-1,was recognized by TRL9 and mediated activationthrough an MyD88-dependent endocytic pathway lead-ing to type I IFN response. Using a recombinant HSV-1KOS strain, Krug and colleagues14 confirmed the involve-ment of TLR9 in type I IFN response. Lundberg andcolleagues18 also showed that HSV-1 DNA is stimulatoryboth in vitro and in vivo. Recently, Kurt-Jones and col-leagues16 demonstrated that TLR2 mediates the induc-tion of inflammatory cytokines in response to intravenousinoculation with the HSV-1 KOS strain, whereas in micelacking functional TLR2, they detected a reduction inencephalitis symptoms.

Here we used a HSV-1 isolated from a natural oralrecurrent human infection, expanded in Vero cells, andpurified in sucrose gradient.19 We demonstrate the acti-vation of TLR2 by HSV-1 in vitro using Chinese hamsterovary (CHO) cells stably transfected with human TLR2and a reporter gene. We also show for the first time, usingan in vivo mouse model of intranasal inoculation,3 which isa natural route of infection, that HSV-1 leads to lethalencephalitis in 100% of the mice lacking the functionalMyD88 protein. These results further suggest the impor-tance of TLRs and innate immunity in host resistance toHSV-1.

Materials and Methods

Viruses, Staphylococcus aureus, and LPS

HSV-1 strain EK,20 isolated from a human case of recur-rent oral herpes with blisters and Vaccinia virus WesternReserve (VV) were allowed to multiply in Vero cells, main-tained with minimal essential medium (GIBCO, GrandIsland, NY) containing 5% fetal bovine serum (FBS)(GIBCO) and 25 �g/�l of ciprofloxacin (Fesenius, Pune,India) at 37°C in a 5% CO2 atmosphere. HSV-1 and VVwere purified in sucrose gradients,19 and the titers deter-mined in Vero cells as previously described.21 The virustiters obtained were: 1.1 � 108 PFU/ml for HSV-1 and 2 �1010 PFU/ml for VV. LPS from Escherichia coli O55:B5 was

obtained from Sigma (St. Louis, MO) and UV-inactivatedS. aureus was described before.5

Vero Cells

Vero cells were maintained in minimal essential mediumsupplemented with 5% heat-inactivated FBS and antibi-otics in 5% CO2 at 37°C. These cells were used formultiplication and titration of virus and in neutralizationtests.

CHO Cell Lines

The CHO reporter cell lines,22,23 a kind gift from DouglasT. Golenbock (University of Massachusetts MedicalSchool, Worcester, MA), were maintained as adherentmonolayers in Ham’s F-12/Dulbecco’s modified Eagle’smedium supplemented with 5% FBS and antibiotics at37°C, 5% CO2. All of the cell lines were derived fromclone 3E10, a CHO/CD14 cell line that has been stablytransfected with a reporter construct containing the struc-tural gene for CD25 under the control of the humanE-selectin promoter. This promoter contains a NF-�Bbinding site; CD25 expression is completely dependenton NF-�B translocation to the cell nucleus.23 Cells ex-pressing TLRs were constructed by stable transfection ofthe CHO/CD14 reporter cell line with the cDNA for humanTLR2 or expressing endogenous TLR4 as described.22 Inaddition to the LPS-responsive cell lines describedabove, we also tested a LPS nonrespondent cell linederived from 3E1022 designated clone 7.19, as well as aclonal line derived from this mutant that was transfectedwith CD14 and TLR2 (7.19/CD14/TLR2). The LPS nonre-sponsive phenotype of the 7.19 cell lines is due to amutation in the MD-2 gene, and thus is defective in sig-naling via TLR4.22 These cell lines report NF-�B activa-tion via surface expression of CD25, similarly to the otherCHO cell lines described.

Flow Cytometry Analysis

CHO reporter cells were plated at a density of 1 � 105

cells/well in a 24-well tissue culture dish. After 20 hours,UV-inactivated bacteria, HSV-1 or VV were added in atotal volume of 250 �l of medium/well for 18 hours. Thecells were then harvested with trypsin-ethylenediamine-tetraacetic acid (Sigma, St. Louis, MO) and washed oncewith medium containing 5% FBS and then with phos-phate-buffered saline (PBS). Cells harvested physicallywithout trypsin displayed similar results. Subsequently,the cells were counted and 1 � 105 cells stained withphycoerythrin-labeled anti-CD25 (mouse monoclonal an-tibody to human CD25, R-PE conjugate; Caltag Labora-tories, Burlingame, CA) 1:200 in PBS, on ice in the dark,for 30 minutes. After labeling, the cells were washedtwice with PBS containing 1 mmol/L sodium azide(Sigma), and 10,000 cells were examined by flow cytom-etry (BD Biosciences, San Jose, CA) for the expression ofsurface CD25 as described.5,8,22 After excluding deadcells by gating with forward and side scatter parameters,

1420 Mansur et alAJP May 2005, Vol. 166, No. 5

an average of 8750 � 312 live cells, were analyzed forthe expression of CD25. Analysis was performed usingCellQuest software (BD Biosciences).

Animals

TLR2�/� and MyD88�/� mice were generated at OsakaUniversity (Osaka, Japan) and backcrossed in theC57BL/6 background for eight generations. IFN��/�

mice in the C57BL/6 background were obtained from TheJackson Laboratory (Bar Harbor, ME). The knockout micewere transferred to the Federal University of MinasGerais, Institute of Biological Sciences (Belo Horizonte,Minas Gerais, Brazil) and maintained in a pathogenic-free, barrier environment. C57BL/6 mice, used as wild-type (WT) control, were obtained from the Centro dePesquisas Rene Rachou, Oswaldo Cruz Foundation(Belo Horizonte, Minas Gerais, Brazil). Four-week-oldmale mice were anesthetized with ketamine (Agribrandsdo Brasil Ltda, Paulinia, Brazil), and 104 PFU of the puri-fied HSV-1 contained in 10 �l were inhaled by mice asdescribed previously.5 Control mice inhaled PBS. Ninemice from each knockout or WT group were used in thesurvival experiments shown in Figure 2. Eight days afterinfection, brain, lung, liver, and spleen were removedfrom three animals per group and either frozen or fixed informalin (Sciavicco Comercio e Industria Ltda, Belo Hori-zonte, Brazil). Each experiment was repeated threetimes. Mice presenting symptoms such as total paralysisand/or seizures were sacrificed. The mouse colonies andall experimental procedures were performed accordingto the institutional animal care and use guidelines fromthe Centro de Pesquisas Rene Rachou, Oswaldo CruzFoundation.

Murine Macrophage Preparation and TumorNecrosis Factor (TNF)-� Measurement

Thioglycollate-elicited peritoneal macrophages were ob-tained from either C57BL/6, TLR2�/�, or MyD88�/� miceby peritoneal washing. Adherent peritoneal macro-phages were cultured in 96-well plates (2 � 105 cells/well) at 37°C/5% CO2 in Dulbecco’s modified Eagle’smedium (Life Technologies, Paisley, UK) supplementedwith 5% heat-inactivated FBS (Life Technologies), 2mmol/L L-glutamine (Sigma) and 40 �g/ml of gentamicin(Schering do Brasil, Rio de Janeiro, Brazil). Cells werethen stimulated with HSV-1 (multiplicity of infection, 40),LPS, or S. aureus for 24 hours to evaluate TNF-� produc-tion. TNF-� was quantified using a DuoSet ELISA kit fromR&D Systems (Minneapolis, MN).

Virus Detection by Nested Polymerase ChainReaction (PCR) and Determination of VirusConcentration in Mice Tissues

Frozen mice tissues were ground with sterile sand and200 �l of minimal essential medium, centrifuged, and thesupernatant was used for titration in a standard tissue

culture infectious dose (TCID50) assay24 and for nestedPCR. The primers and conditions used for the first reac-tion of PCR were described previously by Nogueira andcolleagues.20 The nested PCR was developed using theprimers specific for HSV-1 thymidine kinase: TKI3 CCAGCA TAG CCA GGT CAA GC and TKI5 GCG AAC ATCTAC ACC ACA CAA CA. The reaction was performed at95°C for 1 minute, 50°C for 1 minute, and 72°C for 1minute for 40 cycles.

Immunohistochemistry

Brain samples were fixed with 10% formaldehyde inphosphate buffer and then embedded in paraffin. Sec-tions were mounted on glass slides, deparaffinized, andthen treated with 3.5% H2O2 in PBS. Tissue sections wereblocked with 7% normal goat serum in PBS for 30 minutesat room temperature and incubated overnight with poly-clonal rabbit anti-herpes simplex I antibody p0175(DAKO, NH) diluted 1:50 or monoclonal anti-gC HSV-1diluted 1:500 in PBS containing 0.4% Triton X-100. Afterincubation with primary antibody, tissue sections werewashed three times with PBS and incubated for 90 min-utes at room temperature with the secondary biotinylatedantibody solution (DAKO, NH), washed with PBS, andtreated with the tertiary solution, containing peroxidase-conjugated streptavidin (DAKO, NH), for 60 minutes atroom temperature. Sections were then rinsed in PBS andwith 3,3�-diaminobenzidine tetrahydrochloride (Sigma)(0.05%) and hydrogen peroxide (0.03%). The sectionswere then rinsed in PBS and stained with hematoxylinand eosin (H&E) (Reagen, Rio de Janeiro, Brazil).

Neutralization Test

Sera of mice were serially diluted from 1:10 to 1:1280 inminimal essential medium in a total volume of 100 �l. Onehundred TCID50 of HSV-1 was added to each dilution andincubated for 1 hour at 37°C in a 5% CO2 atmosphere.The mixture was added to a 96-well plate containing Verocells. The plates were incubated and observed during 5days. All of the samples were titrated in duplicate. Thetiter was determined as the inverse of the highest dilutionof serum that protected Vero cells from cytopathic effectof HSV-1.

Statistical Analysis

Statistical analyses were performed with Student’s t-testusing the software program Minitab (Minitab Inc., StateCollege, PA).

Results

In Vitro Activation of TLR2 by HSV-1

We investigated if a recently isolated strain of HSV-1,purified by sucrose gradient, induces activation of TLR2in vitro. Flow cytometry analysis of the expression of a

Lethal HSV-1 Infection in MyD88�/� Mice 1421AJP May 2005, Vol. 166, No. 5

CD25 reporter gene in CHO cells, stably transfected withthe TLR constructions described previously22,23 is shownin Figure 1A. We used cells stably transfected with CD14alone (CHO/CD14) or with CD14 and TLR2 (CHO/CD14/TLR2), both expressing endogenous TLR4, as well as theclone 7.19, which does not have a functional TLR4 sig-naling pathway, but are stably transfected with CD14(7.19) or CD14 and TLR2 (7.19/CD14/TLR2). These cellswere exposed for 18 hours to 104 PFU of HSV-1 or Vac-cinia virus (VV). VV served as a negative control,25 be-cause it was produced in cell cultures and purified by thesame process19 used to purify HSV-1. We observed thatthe cells stimulated with HSV-1 were activated throughTLR2, but not through TLR4. Figure 1B shows an in-creased percentage of CD25-positive cells in TLR2/CHOor TLR2/TLR4/CHO cells stimulated with HSV-1. Thesedata indicate that purified HSV-1 triggers NF-�B throughTLR2/CD14, but not through TLR4/CD14.

To compare the host innate immune response afterHSV-1 challenge ex vivo, we measured the levels ofTNF-� in culture supernatants of macrophages from WT,TLR2�/�, and MyD88�/� mice (Figure 1C). Both WT andTLR2�/� produced significant amounts of this cytokinewhen challenged by HSV-1, whereas the production inMyD88�/� was totally abrogated. LPS and S. aureus wereused as controls. LPS,6,7 differently from S. aureus,22 stillactivated macrophages from TLR2�/� mice, whereas

none of the microbial stimuli were effective on macro-phages from MyD88�/� mice.13

MyD88�/� Mice Are Highly Susceptible to HSV-1 Infection

Our next step was to evaluate the importance of TLR2and MyD88 during infection with HSV-1 in an in vivomodel. We used the intranasal model2 because it is anatural route of infection with HSV-1. Four-week-oldC57BL/6, TLR2�/�, and MyD88�/� mice were inoculatedwith 104 PFU of HSV-1 intranasally. Because it has beenpreviously demonstrated that IFN-� receptor-deficientmice are more susceptible to infection with HSV-126–29

and that mice with herpetic stromal keratitis produce highlevels of IFN-�,30 IFN��/� mice were also tested.

In our initial studies we used different isogenic mousestrains, such as the C57BL/6, BALB/c, and 129 strains.All of these strains were found to be resistant to intranasalinfection and showed 100% survival at 4 weeks of age,after infection with 104 PFU of HSV-1 (data not shown).The TLR2�/� mice did not show any observable clinicalsymptoms, and were as resistant as C57BL/6 mice toinfection with HSV-1. In contrast, 100% of MyD88�/�

mice died between 6 to 10 days after infection (Figure2A). The IFN��/� mice were also more susceptible to

Figure 1. TLR2 mediates cellular activation after exposure to HSV-1. A: CHO/CD14 (expressing endogenous TLR4), 7.19/CD14/TLR2 (expressing TLR2),CHO/CD14/TLR2 (expressing TLR4/TLR2), or 7.19 (LPS nonresponder control) cells were left untreated (black area) or exposed (gray lines) to 104 PFU ofHSV-1 (top) or to 104 PFU of VV (bottom), and the expression of the reporter gene (CD25) was measured 18 hours later by flow cytometry. The data arerepresentative of three experiments. B: The cell lines were exposed to 104 PFU of HSV-1 (black bars) or of VV (gray bars) and the expression of the reportertransgene CD25 was measured by flow cytometry. The percentage of CD25-positive cells was obtained by subtracting the percentage of stimulated cells expressingCD25 from the percentage of nonstimulated cells expressing CD25. An average of 8750 � 312 cells were analyzed in each experiment. This experiment isrepresentative of three performed. Asterisks indicate that differences in reporter gene expression on TLR2 or TLR4/TLR2 cells is statistically significant (P � 0,01)when compared to TLR4 or control cell lines. C: Macrophages derived from WT (black columns), TLR2�/� (gray columns), or MyD88�/� (white columns)mice were exposed to HSV-1 (multiplicity of infection, 40), LPS, and S. aureus and the levels of TNF-� were measured in the culture supernatants at 24 hoursafter macrophage stimulation. Asterisks indicate that differences are statistically significant (P � 0.05), when comparing cytokine levels produced bymacrophages from WT or TLR2�/� mice to macrophages from MyD88�/� mice. The experiment was performed in triplicates and the results shown are onerepresentative of two experiments that yielded the same results.


HSV-1 infection (Figure 2A). After brain tissues from micesacrificed on the 8th day after infection were inoculatedinto Vero cell cultures, the samples from MyD88�/� andfrom IFN��/� mice with symptoms of infection were de-termined to have high TCID50 (Figure 2B), as comparedto brain tissues of TLR2�/�, C57BL/6, or IFN��/� micethat did not display clinical symptoms of infection. At-tempts to recover virus from lung, spleen, and liver, fromthese mice, either by nested PCR or isolation in Vero cellswere unsuccessful (data not shown).

The brains of mice sacrificed at 8 days after infectionwere processed for nested PCR reactions, as previouslydescribed,20 with specific primers for HSV-1 thymidinekinase (TK) gene, and the results are shown in Table 1.Only the brains from MyD88�/� and from IFN��/� micewith symptoms were positive for HSV-1 TK. To confirmthat all mice were infected, a neutralization test was per-formed (Table 1) using sera from C57BL/6 and TLR2�/�

mice obtained at 30 days after infection. Our results showthat all mice were seropositive (Table 1). Of note, theneutralization test in MyD88�/� and symptomaticIFN��/� was performed after 8 days of infection, becauseof their early death (Figure 2). No mice (ie, WT, TLR2�/�,MyD88�/�, or IFN��/�) presented seropositive results(data not shown) at this time. Together, these resultsindicate that the absence of IFN-� or MyD88 enhancesthe entry of HSV-1 into the brain and results in 50% or100% mortality, respectively.

Lethal Encephalitis in MyD88�/� and IFN��/�

Mice Infected with HSV-1

Macroscopic observation of brains from MyD88�/� andIFN��/� mice with clinical symptoms revealed hemor-rhagic and necrotic areas, differing substantially fromTLR2�/�, C57BL/6, or IFN��/� without clinical symptomsthat failed to show these gross changes. To further con-firm the effects of the infection in vivo, we used immuno-histochemical and histopathological methods on sectionsof brain and trigeminal ganglia of mice.

Microscopic examination of the brains stained withH&E revealed focal encephalitis characterized by mono-nuclear cell infiltrates and activated glial cells associatedwith necrosis and vascular congestion in some areas ofcortex tissue of MyD88�/� and of IFN��/� mice present-ing clinical symptoms (Figure 3A and Table 2), whileTLR2�/� mice showed only mild vascular congestion(Figure 3A and Table 2). In contrast, brains of IFN��/�

without clinical symptoms (data not shown) or C57BL/6mice did not show morphological alterations. Immunore-activity to mouse polyclonal anti-HSV-1 was observed inMyD88�/� and in IFN��/� mice with clinical symptoms,but not in TLR2�/� or IFN��/� mice without clinical symp-toms, or C57BL/6 mice (Figure 3C). Viral infection wasconfirmed in trigeminal ganglia in all experimentalgroups, including C57BL/6, TLR2�/�, IFN��/�, andMyD88�/� mice, at the 8th day after infection, usingimmunohistochemistry with monoclonal antibody anti-HSV-1 (Figure 3B).

Discussion

Immune response against infection with HSV-1 is verycomplex. Using the murine experimental model, it hasbeen reported that type I and type II IFNs as well asTNF-� are the main elements activated in the innate

Figure 2. MyD88�/� mice are highly susceptible to infection with HSV-1. A:Nine MyD88�/� (diamonds), IFN��/� (triangles), TLR2�/� (squares),and C57BL/6 (circles) 4-week-old mice were intranasally inoculated with104 PFU of HSV-1 or PBS, and survival was assessed daily. B: Eight days afterintranasal infection with 104 PFU of HSV-1 the brains from nine mice of eachgroup, or from animals that had died from infection, were collected aftersacrifice, macerated, and inoculated into Vero cell cultures to perform thetitration procedure in triplicate. sIFN�/�, IFN�/� mice with clinical symp-toms; nsIFN�/�, IFN�/� mice with no clinical symptoms. This experiment isrepresentative of three performed. *, Virus brain concentration in MyD88�/�

was statistically higher (P � 0,05) as compared to virus brain concentrationin sIFN��/�, nsIFN��/�, TLR2�/�, or in C57BL/6 mice. **, Virus brainconcentrations in MyD88�/�, or in sIFN�/�, mice were statistically higher(P � 0001) as compared to virus brain concentration in nsIFN��/�, TLR2�/�,or in C57BL/6 mice.

Table 1. HSV-1-Specific PCR and Serum Neutralization to Confirm Mice Infection with HSV-1

MyD88�/� TLR2�/� C57BL/6 nsIFN��/� sIFN��/�

Brain PCR using HSV-1TK gene primers � � � � �Abs neutralization* ND 20 60 160 ND

The brains from mice sacrificed 8 days after infection, or from animals that had died from infection, were processed to PCR reactions with specificprimers for TK HSV-1 gene. PCR reactions run on brain from uninfected mice were negative. The sera from surviving animals were obtained 30 daysafter infection and used in the neutralization test.

*The titers are correspondent of the inverse values of the sera dilution that protects Vero cells from cytopathic effect of HSV-1, calculated as themedian titer from sera from four animals. The neutralization test was performed in duplicate. sIFN�/�, IFN�/� mice with clinical symptoms; nsIFN�/�,IFN�/� mice with no clinical symptoms; ND, not done.


immune response against infection with HSV-1.27–32 Ithas also been shown that HSV-1 activates both TLR2 andTLR9 in a MyD88-dependent manner, suggesting theimportance of TLRs in encephalitis development and hostresistance to this viral infection.14,16,17 Although we con-firmed activation of TLR2 by HSV-1 in transfected CHOcells, the lack of functional MyD88, but not functionalTLR2, resulted in severely impaired cytokine synthesis byinflammatory macrophages exposed to HSV-1. Consis-tently, MyD88 knockout, but not TLR2 knockout mice,displayed enhanced susceptibility to experimental infec-tion with HSV-1. We favor the hypothesis that a combinedeffort of different TLRs is implicated in the activation of theinnate immune system and host resistance to infectionwith HSV-1. HSV-1 is a complex enveloped virus, whichhas 140-kb DNA and expresses 84 proteins.1,31 There-fore, we speculate that HSV-1 is recognized by multipleTLRs, such as TLR2 and TLR9, that may have additiveeffects in activation of MyD88 on infection with HSV-1.

Thus, simultaneous blocking of the function of multipleTLRs may be required to yield the same phenotype asseen in MyD88�/� mice on infection with HSV-1. Never-theless, our findings provide new information that corrob-orates the hypothesis that MyD88 and possibly TLRshave an important role in host resistance to viral infectionand pathogenesis observed during infection with HSV-1.

In a recent report, Boivin and colleagues2 describedthe enhanced expression of TLR2 in the hindbrain of miceinfected with HSV2. More importantly, Kurt-Jones andcolleagues16 demonstrate that HSV-1 activates TLR2 invitro in CHO-transfected cells. In this study, infection ofadult mice with 109 PFU of HSV-1 KOS strain, showedthat WT mice were more susceptible to the virus infectionas compared to TLR2�/� mice. When Kurt-Jones andcolleagues16 infected neonates (4-day-old mice) with 104

PFU of HSV-1 KOS strain, they also observed thatTLR2�/� mice, with a mortality of 30%, were more resis-tant than TLR4�/� or WT mice, which presented more

Figure 3. HSV-1 is able to enter in MyD88�/� and IFN��/� brains and cause brain degenerative changes and necrosis. Three TLR2�/�, MyD88�/�, or IFN��/�

mice were sacrificed 8 days after intranasal infection with 104 PFU HSV-1, and representative sections of cerebral cortex (A and C) or trigeminal ganglia (B) weremade. A: H&E-stained sections (see semiquantitative analysis of encephalitis in Table 2). B: Immunohistochemical staining for anti-gC protein from HSV-1 showingreactions against the virus (arrows). C: Anti-HSV-1 (DAKO)-immunoreactive cells (arrows) in affected areas of Myd88�/� and sIFN��/� mice. The inset in B(left) is a negative control from the trigeminal ganglion of a noninfected WT control animal. Original magnifications: �200 (A); �400 (B); �1000 (C).

Table 2. Semiquantitative Histopathological Analysis of Encephalitis in HSV-1-Infected Animals

Parameters*Degenerative changes

and necrosisPerivascular cuffing and

congestive changesMononuclear cell

infiltrates

GroupsWT �(3/3) �(3/3) �(3/3)TLR2�/� �(3/3) �(3/3) �(3/3)MyD88�/� ��(3/3) ��(3/3) ��(3/3)sIFN��/� ��(3/3) ��(3/3) ��(3/3)

*The histopathological evaluation criteria for quantification of changes was based on the alterations observed in the brain tissues collected from thedifferent groups of mice as compared to WT; see Figure 3A. �, Absence of significant alterations; �, mild alterations; ��, intense alterations; (3/3),three of three examined animals presented the degree of changes indicated. This experiment is representative from three experiments performed.sIFN�/�, IFN�/� mice with clinical symptoms.


than 90% mortality. They demonstrated in their model thatHSV-1-induced encephalitis and lethality was primarilymediated by TLR2.

Our in vitro data further confirmed that an earlier inter-action from HSV-1 with the innate immunity could happenthrough TLR2. However, when measuring TNF-�, a criti-cal cytokine for host resistance against HSV-132 we foundthat induction of TNF-� production by inflammatory mac-rophages exposed to HSV-1 was abolished in cells fromMyD88�/�, but not in cells from TLR2�/� mice. Further,we found that all MyD88�/� mice died after HSV-1 intra-nasal inoculation of 104 PFU. Conversely, the mice lack-ing the TLR2 functional gene have the same survival rateof WT mice, suggesting that other TLRs could be involvedin the response against HSV-1. In summary, our studyindicates a critical role of MyD88 in anti-viral defense,whereas Kurt-Jones and colleagues16 demonstrate thatactivation of TLR2 by HSV-1 will lead to detrimental in-flammatory response and lethal encephalitis. Thus, oneimportant goal of our future studies will be the identifica-tion of the TLR involved in host anti-viral defenses againstHSV-1. In any case TLRs are not evenly distributed in thedifferent organ tissues and cells and the difference in theresults obtained in these studies, could be explained bythe different HSV-1 strains, the size of HSV-1 inoculum,and/or the route of infection. The Kurt-Jones group16

used the intraperitoneal route and gave 109 PFU of HSV-1KOS in adults mice or 104 PFU in neonates, whereas weused a clinically isolated strain, with a concentration of104 PFU in 4-week-old mice. Further, the genetic back-ground of mice used in these studies may also haveinfluenced the outcome of infection. We used MyD88�/�

and TLR2�/� mice, which have been backcrossed eighttimes into the C57BL/6 background, and C57BL/6 ascontrol. Kurt-Jones and colleagues16 used a F2 TLR2�/�

mice, and the interbred 129 � C57BL/6 as control.Consistent with the mortality results, we found HSV-1

replication in the brain of mice lacking the MyD88, but notin brain of TLR2�/� mice. As previously shown,26–30 wealso observed an enhanced susceptibility of IFN��/�

mice infected with HSV-1 (50% of mortality). Further, weshowed that MyD88�/� and symptomatic IFN��/� pre-sented severe neuropathological signs of encephalitis,whereas TLR2�/� presented only mild neuropathologicalsigns and the WT showed no signals in the histopathol-ogy analysis of the brain. Because HSV-1 remains intrigeminal ganglia after infection, we performed immuno-histochemistry against gC protein of HSV-1 in trigeminalganglia, and showed that all mice (WT, TLR2�/�,MyD88�/�, and IFN��/�) were efficiently infected. Addi-tionally, we demonstrated that after 30 days of infection,all mice that survived infection produced neutralizingantibodies against HSV-1. The neutralization test wasalso performed at day 8 day after infection but no mice(WT, TLR2�/�, MyD88�/�, or IFN��/�) presented sero-positive results. The early death from MyD88�/� andIFN��/�, when the acquired defense was not yet estab-lished, further indicates that in our model innate immuneresponse has a critical role in host defense against HSV-1infection.

Finally, Lundberg and his colleagues18 described theimmunostimulatory role of HSV-1 genome, which is un-methylated and rich in G�C. They showed that mousesplenocytes treated with HSV-1-derived oligonucleotidesproduced IFN-�, TNF-�, and interleukin-6, and pos-sessed a potent adjuvant activity in vivo, leading to TH1response after immunization and restimulation withovalbumin. Krug and colleagues14 also demonstrated inplasmacytoid dendritic cells, that HSV-1 activates murinecells through TLR9. They showed that these highly spe-cialized IFN producer cells responded in vitro to stimuluswith HSV-1 through TLR9 and MyD88. Further, Lund andcolleagues17 described that activation of plasmacytoiddendritic cells by HSV-2 also occurs via TLR9. However,in vivo experiments14 showed that mice deficient in eitherMyD88 or in TLR9, although presenting impaired theresponse from plasmacytoid dendritic cells, could stillcontrol corneal infection with HSV-1. The 100% lethalityobserved in infected MyD88�/� mice in this study, incomparison with the controlled infection in the mice in-fected by scarring of cornea, further suggest that theinoculation route and/or the strain of the virus could playan important role in the outcome of the experimentalinfection. Therefore, additional studies will be necessaryto define what is (are) the critical TLR(s) in controlling viralreplication in the brain and host resistance to infectionwith HSV-1.

Acknowledgments

We thank Douglas T. Golenbock (University of Massa-chusetts Medical School, Worcester, MA) for providing uswith the CHO cell lines, Susanne Facchin for neutraliza-tion tests, and Gregory T. Kitten for critical reading of themanuscript.

References

1. Roizman B, Whitley RJ: The nine ages of herpes simplex virus. Her-pes 2001, 8:23–27

2. Boivin G, Coulombe Z, Rivest S: Intranasal herpes simplex virus type2 inoculation causes a profound thymidine kinase dependent cere-bral inflammatory response in the mouse hindbrain. Eur J Neurosci2002, 16:29–43

3. Hirsh HH, Bossart W: Two-centre study comparing DNA preparationand PCR amplification protocols for herpes simplex virus detection incerebrospinal fluids of patients with suspected herpes simplex en-cephalitis. J Med Virol 1999, 57:31–35

4. Takeuchi O, Hoshino K, Kawai T, Sanjo H, Takada H, Ogawa T,Takeda K, Akira S: Differential roles of TLR2 and TLR4 in recognitionof Gram negative and Gram positive bacterial cell wall components.Immunity 1999, 11:443–451

5. Campos MA, Almeida IC, Takeuchi O, Akira S, Paganini E, ProcopioDO, Travassos LR, Smith JA, Golenbock DT, Gazzinelli RT: Activationof Toll-like receptor-2 by glycosylphosphatidylinositol anchors from aprotozoan parasite. J Immunol 2001, 167:416–423

6. Poltorak A, He X, Smirnova I, Liu MY, Van Huffel C, McNally O,Birdwell D, Alejos E, Silva M, Galanos C, Freudenberg M, Ricciardi-Castagnoli P, Layton B, Beutler B: Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: mutations in TLR4 gene. Science 1998,282:2085–2088

7. Lien E, Means TK, Heine H, Yoshimura A, Kusumoto S, Fukase K,Fenton MJ, Oikawa M, Qureshi N, Monks B, Finberg RW, Ingalls RR,


Golenbock DT: Toll-like receptor 4 imparts ligand-specific recognitionof bacterial lipopolysaccharide. J Clin Invest 2000, 105:497–504

8. Campos MA, Rosinha GMS, Almeida IC, Salgueiro XS, Jarvis BW,Splitter GA, Bruna-Romero O, Gazzinelli RT, Oliveira SC: The role ofToll-like receptor 4 in induction of cell-mediated immunity and resis-tance to Brucella abortus infection in mice. Infect Immun 2004,72:176–186

9. Alexopoulou L, Holt AC, Medzhitov R, Flavell RA: Recognition ofdouble-stranded RNA and activation of NFKB by Toll-like receptor 3.Nature 2001, 413:432–438

10. Hayashi F, Smith KD, Ozinsky A, Hawn TR, Yi EC, Goodlett DR, EngJK, Akira S, Underhill DM, Aderem A: The innate immune response tobacterial flagelin is mediated by Toll-like receptor 5. Nature 2001,410:1099–1103

11. Hemmi H, Takeuchi O, Kawai T, Kaisho T, Sato S, Sanjo H, Matsu-moto M, Hoshino K, Wagner H, Takeda K, Akira S: A Toll-like receptorrecognizes bacterial DNA. Nature 2000, 408:740–745

12. Takeda K, Kaisho T, Akira S: Toll-like receptors. Annu Rev Immunol2003, 21:335–376

13. Yamamoto M, Takeda K, Akira S: TIR domain-containing adaptorsdefine the specificity of TLR signaling. Mol Immunol 2004,40:861–868

14. Krug A, Luker GD, Barchet W, Leib DA, Akira S, Colonna M: Herpessimplex virus type 1 (HSV-1) activates murine natural interferon-producing cells (IPC) through Toll-like receptor 9. Blood 2004,103:1433–1437

15. Kurt-Jones EA, Popova L, Kwinn L, Haynes LM, Jones LP, Tripp RA,Walsh EE, Freeman MW, Golenbock DT, Anderson LJ, Finberg RW:Pattern recognition receptors TLR4 and CD14 mediate response torespiratory syncytial virus. Nat Immunol 2000, 1:398–401

16. Kurt-Jones EA, Chan M, Zhou S, Wang J, Reed G, Bronson R, ArnoldMM, Knipe DM, Finberg RW: Herpes simplex virus 1 interaction withToll-like receptor 2 contributes to lethal encephalitis. Proc Natl AcadSci USA 2004, 101:1315–1320

17. Lund J, Sato A, Akira S, Medzhitov R, Iwasaki A: Toll-like receptor9-mediated recognition of Herpes simplex virus-2 by plasmacytoiddendritic cells. J Exp Med 2003, 198:513–520

18. Lundberg P, Welander P, Han X, Cantin E: Herpes simplex virus type1 DNA is immunostimulatory in vitro and in vivo. J Virol 2003,77:11158–11169

19. Joklik WK: The purification of four strains of poxvirus. Virology 1962,18:9–18

20. Nogueira ML, Siqueira RC, Freitas N, Amorim JB, Bonjardim CA,

Ferreira PC, Orefice F, Kroon EG: Detection of herpesvirus DNA bythe polymerase chain reaction (PCR) in vitreous samples from pa-tients with necrotising retinitis. J Clin Pathol 2001, 54:103–106

21. Campos MA, Kroon EG: Critical period of irreversible block of vac-cinia virus replication. Rev Bras Microbiol 1993, 24:104–110

22. Lien E, Sellati TJ, Yoshimura A, Flo TH, Rawadi G, Finberg RW, CarrollJD, Espevik T, Ingalls RR, Radolf JD, Golenbock DT: Toll-like receptor2 functions as a pattern recognition receptor for diverse bacterialproducts. J Biol Chem 1999, 274:33419–33425

23. Delude RL, Yoshimura A, Ingalls RR, Golenbock DT: Construction ofa lipopolysaccharide reporter cell line and its use in identifying mu-tants defective in endotoxin, but not TNF-alpha, signal transduction.J Immunol 1998, 161:3001–3009

24. Schmidt, N J: Cell culture techniques for diagnostic virology. Diag-nostic Procedures for Viral, Rickettsial and Chlamydial Infections.Edited by Lennette EH, Schmidt, NJ. Washington, American PublicHealth Association, Inc., 1979, p 100

25. Bowie A, Kiss-Toth E, Symons JA, Smith GL, Dower SK, O’Neil LAJ:A46R and A52R from vaccinia virus are antagonists of host IL-1 andToll-like receptor signaling. Proc Natl Acad Sci USA 2000,97:10162–10175

26. Smith PS, Wolcott RM, Chervenak R, Jennings SR: Control of acuteHerpes simples virus infection: T-cell-mediated viral clearance isdependent upon interferon-� (IFN-�). Virology 1994, 202:76–88

27. Liu T, Khanna KM, Carriere BN, Hendricks RL: Gamma interferon canprevent Herpes simplex virus type 1 reactivation from latency insensory neurons. J Virol 2001, 75:11178–11184

28. Sainz Jr B, Halford WP: Alpha/beta interferon and gamma interferonsynergize to inhibit the replication of Herpes simplex virus type 1.J Virol 2002, 76:11541–11550

29. Vollstedt S, Arnold S, Schwerdel C, Franchini M, Alber Gottfried, DiSanto JP, Ackermann M, Suter M: Interplay between alpha/beta andgamma interferons with B, T, and natural killer cells in the defenseagainst Herpes simplex virus type 1. J Virol 2004, 78:3846–3850

30. Keadle TL, Usui N, Laycock KA, Kumano Y, Pepose JS, Stuart PM:Cytokine expression in murine corneas during recurrent herpeticstromal keratitis. Ocul Immunol Inflamm 2001, 9:193–205

31. Whitley RJ: Herpes simplex viruses. Fields Virology. Edited by KnipeDM, Howley PM. Philadelphia, Lippincott Williams and Wilkins, 2001,pp 2461–2510

32. Minagawa H, Hashimoto K, Yanagi Y: Absence of tumour necrosisfactor facilitates primary and recurrent herpes simplex virus-1 infec-tions. J Gen Virol 2004, 85:343–347


Immunopathology and Infectious Diseases

Toll-Like Receptor (TLR) 2 and TLR9 Expressed inTrigeminal Ganglia are Critical to Viral ControlDuring Herpes Simplex Virus 1 Infection

Graciela Kunrath Lima,* Guilherme Pimenta Zolini,†

Daniel Santos Mansur,* Braulio Henrique Freire Lima,†

Uschi Wischhoff,† Ruiz Gerhardt Astigarraga,†

Marcela Franca Dias,† Mariana das GracasAlmeida Silva,† Samantha Ribeiro Bela,†

Lis Ribeiro do Valle Antonelli,† Rosa Maria Arantes,‡

Ricardo Tostes Gazzinelli,†§¶ Andre Bafica,�

Erna Geessien Kroon,* and Marco Antonio Campos†

From the Departamentos de Microbiologia,* Patologia Geral,‡ and

Bioquimica e Imunologia,§ Instituto de Ciencias Biologicas,

Universidade Federal de Minas Gerais, Belo Horizonte, Brazil; the

Laboratorio de Imunopatologia,† Centro de Pesquisas Rene Rachou,

Fiocruz, Belo Horizonte, Brazil; the University of Massachusetts

Medical School,¶ Worcester, Massachusettes; and the Laboratorio de

Imunologia e Doencas Infecciosas,� Departamento de Microbiologia

e Parasitologia, Universidade Federal de Santa Catarina,

Florianopolis, Brazil

Herpes simplex virus 1 (HSV-1) is a neurotropic DNAvirus that is responsible for several clinical manifesta-tions in humans, including encephalitis. HSV-1 triggerstoll-like receptors (TLRs), which elicit cytokine produc-tion. Viral multiplication and cytokine expression inC57BL/6 wild-type (WT) mice infected with HSV-1 wereevaluated. Virus was found in the trigeminal ganglia(TG), but not in the brains of animals without signs ofencephalitis, between 2 and 6 days postinfection(d.p.i.). Cytokine expression in the TG peaked at 5 d.p.i.TLR9�/� and TLR2/9�/� mice were more susceptible tothe virus, with 60% and 100% mortality, respectively, asopposed to 10% in the WT and TLR2�/� mice. Increasedlevels of both CXCL10/IP-10 and CCL2/MCP-1, as well asreduced levels of interferon-� and interleukin 1-� tran-scripts, measured in both the TG and brains at 5 d.p.i.,and the presence of virus in the brain were correlatedwith total mortality in TLR2/9�/� mice. Cytokine alter-ations in TLR2/9�/� mice coincided with histopatholog-ical changes in their brains, which did not occur in WTand TLR2�/� mice and occurred only slightly inTLR9�/� mouse brain. Increased cellularity, macro-phages, CD8 T cells producing interferon-�, and ex-

pression levels of TLR2 and TLR9 were detected in theTG of WT-infected mice. We hypothesize that HSV-1 in-fection is controlled by TLR-dependent immune responsesin the TG, which prevent HSV-1 encephalitis. (Am J

Pathol 2010, 177:2433–2445; DOI: 10.2353/ajpath.2010.100121)

Herpes simplex virus 1 (HSV-1) infections are widespread,and seropositivity may exceed 70% of the world popula-tion.1,2 The virus is transmitted primarily by contact betweenskin or mucosa with contaminated oral secretions.3 Primaryinfections are usually acquired during childhood and oftenpresent as mild self-limiting pharyngitis or are asymptomat-ic.3 After HSV-1 replicates in the skin and mucosa, itreaches the dorsal root ganglia termini, from which it isintraxonally transported to the trigeminal ganglia (TG),where it becomes latent.4,5 HSV-1 reactivation may be stim-ulated by hormonal alterations, UV exposure, and immuno-suppression, but the mechanisms that underlie reactivationare not well understood.3 The virus causes a wide range ofmanifestations, from the most common herpes labialis toherpes keratitis, which is a major cause of blindness indeveloped countries.2 HSV-1 is also the leading cause ofsporadic encephalitis in immunocompromised as well as inimmunocompetent individuals and without early manage-

Supported by the Fundacao de Amparo a Pesquisa do Estado de MinasGerais (Brazil), the Conselho Nacional de Desenvolvimento Científico eTecnologico (Brazil), Instituto Nacional de Ciência e Tecnologia de Vaci-nas/Conselho Nacional de Desenvolvimento Cientifico e Tecnologico(INCTV/CNPq; Brazil), Coordenacao de Aperfeicoamento de Pessoal deNível Superior (CAPES, Brazil), and the Programa Estrategico de Pes-quisa em Saude IV (PAPES) IV/FIOCRUZ/CNPq (Brazil). E.G.K., M.A.C.,R.T.G., A.B., L.R.A., and R.M.A. have fellowships from CNPq.

G.K.L. and G.P.Z. contributed equally to this work.

Accepted for publication July 15, 2010.

None of the authors disclosed any relevant financial relationships.

Supplemental material for this article can be found on http://ajp.amjpathol.org.

Address reprint requests to Marco Antonio Campos, Ph.D., Labora-torio de Imunopatologia, CPqRR/FIOCRUZ. Av. Augusto de Lima,1715. CEP: 30.190-002 Belo Horizonte, MG, Brazil. E-mail: [email protected].

The American Journal of Pathology, Vol. 177, No. 5, November 2010

Copyright © American Society for Investigative Pathology

DOI: 10.2353/ajpath.2010.100121

2433

ment is usually fatal.6 Although drug treatment has im-proved the outcome of these patients, morbidity remainshigh, and many individuals suffer from relapses or do notrespond well to treatment.6 The mechanisms underlyingHSV-1 manifestations, especially encephalitis, have notbeen well defined but involve the immune system.

Toll-like receptors (TLRs) are membrane-bound patternrecognition receptors that recognize pathogen-associatedmolecular patterns in endosomes (TLR3, 7 to 9) and theextracellular space (all remaining TLRs).7,8 There are 10human (TLRs 1 to 10) and 12 murine (TLR1 to 9 and TLR11to 13) TLR family members.7,8 Since the first description ofTLRs in mammals, many TLR agonists have been de-scribed: peptidoglycans9 and the Trypanosoma cruzi glyco-sil phosphatidylinositol (GPI) anchor for TLR2,10 lipopoly-saccharide (LPS) for TLR4,11–14 double-stranded RNA(dsRNA) for TLR3,15 flagellin for TLR5,16 and CpG DNA forTLR9.17 TLRs activate inflammatory responses and modu-late immunity by several different signal transduction path-ways. The most well-characterized pathway involves my-eloid differentiation factor 88 (MyD88), an adapter moleculecomposed of a Toll-interleukin-1 receptor domain and adeath domain.18 MyD88 recruits the serine/threonine kinaseinterleukin (IL) receptor associated kinase-4, which acti-vates tumor necrosis factor (TNF)-� receptor-associatedfactor-6 that, in turn, phosphorylates inhibitor NF kappa Bkinase (IkB) and causes it to dissociate from and releasenuclear factor �B in the cytoplasm. Nuclear factor �B thentranslocates to the nucleus and acts as a transcription fac-tor of innate immunity-associated genes.18,19 In addition,TLR3 appears to activate the inflammatory responsethrough another adapter molecule, named Toll-interleu-kin-1 receptor domain-containing adaptor-inducing inter-feron (IFN)-�.19 This pathway is MyD88 independent andculminates with the translocation of interferon regulatoryfactor 3 to the nucleus, leading to the production of IFN �and IFN-inducible genes.19

Many studies have examined the participation of innateimmunity in HSV-1-related diseases and in the control ofinfection. In particular, TLRs have been intensively investi-gated. In 2004 Kurt-Jones et al20 demonstrated that HSV-1activated TLR2 in vitro and that TLR2 null (�/�) mice inocu-lated intraperitoneally with HSV-1 KOS showed increasedresistance to infection. In the same year, Hochrein et al21

and Krug et al22 showed that TLR9 was important for thedendritic cell response to HSV-1. However, Krug et al22

could not find differences in viral replication or in suscepti-bility in TLR9�/� and MyD88�/� mice infected in the foot-pad or in the corneas with HSV-1. Nevertheless, our group23

demonstrated that TLRs (and/or IL-1�) are essential to con-trol the virus in an intranasal model of HSV-1 infection be-cause 100% of MyD88�/� mice developed lethal enceph-alitis after viral inoculation. We also showed that 50% of theinoculated IFN-� knockout (KO) mice died from encephali-tis.23 Additionally, cooperation between TLR2 and TLR9 inHSV control has been demonstrated in HSV-1 infected den-dritic cells24 and, more recently, in an HSV-2 mouse modelof vaginal and intraperitoneal infection.25 TLRs have alsobeen proposed to be important in Herpes simplex enceph-alitis in humans.26,27

How innate immunity and which TLRs contribute to thecontrol of HSV-1 and related diseases are still unknown. Inthis study, we investigated how HSV-1 infection is controlledin a murine model of intranasal infection by using the HSV-1EK strain, which was isolated from a human case of recur-rent oral herpes with blisters. Our experiments indicatedthat in C57BL/6 wild-type (WT) mice, control of virus infec-tion seemed to be highly regulated at the level of the TG.The levels of cytokine transcripts were directly related to theviral load in TG, and once the virus was controlled, thecytokine levels were reduced. Additionally, we found thatTLR2 and, more importantly, TLR9 play a role in immuneresponses and immune control in the TG and mouse brain.Thus, it seems that HSV-1 infection control in the intranasalmurine model occurs in the TG and brain, and TLR defi-ciencies may cause deregulated inflammation in these or-gans, which consequently allows virus entry into the brainand raises the susceptibility of mice to infection.

Materials and Methods

Virus

HSV-1 strain EK,28 isolated from a human case of recur-rent oral herpes with blisters, was multiplied in Vero cellsas previously described23 and purified as previously de-scribed.29 The virus titers obtained were 3.0 � 109

plaque forming units (PFU)/ml.

Vero Cells

Vero cells (American Type Culture Collection, Manassas,VA) were maintained in minimal essential medium sup-plemented with 5% heat-inactivated fetal bovine serumand antibiotics in 5% CO2 at 37°C. These cells were usedfor multiplication and titration of the virus.

Human Embryo Kidney Cells

Human embryo kidney (HEK) 293 cells stably transfectedwith the pcDNA3 plasmid (Invitrogen, Carlsbad, CA) con-taining the human TLR2, TLR4MD.2, or TLR9 sequencesor the empty vector, each fused with yellow fluorescentprotein, were a kind gift from Dr. Douglas T. Golenbock(Division of Infectious Diseases and Immunology, Depart-ment of Medicine, University of Massachusetts MedicalSchool, Worcester, MA). Cells were stimulated withHSV-1 (105 PFU/ml or at multiplicities of infection [MOIs]of 2 or 10), LPS (100 ng/ml; from Escherichia coli, 055:B5;Sigma, St. Louis, MO), Malp-2 (10 ng/ml; Alexis Bio-chemicals, San Diego, CA), CpG 2006 and 1826 (5�mol/L; Alexis Biochemicals), or E. coli (100 units/cell) for6 or 24 hours. Human IL-8 was measured by enzyme-linked immunosorbent assay (ELISA; BD, Franklin Lakes,NJ) in the supernatants of HEK293, HEK TLR4, and HEKTLR2 cells 24 hours after stimulation, and the relativeincrease in luciferase activity was measured in HEK TLR9cells 6 hours after stimulation.

2434 Lima et alAJP November 2010, Vol. 177, No. 5

Intraperitoneal Macrophages

Thioglycollate-elicited peritoneal macrophages were ob-tained from either C57BL/6, TLR2�/�, TLR9�/�, or TLR2/9�/� mice by peritoneal washing, activated with murineIFN-� as previously described,23 and then stimulated withHSV-1 (MOI of 10) for 24 hours. Murine TNF � and IL-12 p40were measured in the supernatants by ELISA (BD).

Mice

TLR2�/� and TLR9�/� mice were generated at Osaka Uni-versity (Osaka, Japan) and were kind gifts from ShizuoAkira, and the TLR2/9�/� mice were obtained by crossingTLR2�/� and TLR9�/� mice at the National Institutes ofHealth (Bethesda, MD) and were kind gifts from Alan Sher.The mice were backcrossing to the C57BL/6 backgroundfor eight generations. The C57BL/6 (wild-type, control) andthe knockout mice were maintained in a pathogen-free,barrier environment in the Centro de Pesquisas Rene Ra-chou, Oswaldo Cruz Foundation (CPqRR/FIOCRUZ; BeloHorizonte, Minas Gerais, Brazil). Six- to ten-week-old malemice were anesthetized with ketamine (Agribrands do Bra-sil Ltda, Brazil), and 106 PFU of purified HSV-1 in 10 �l wasinhaled by the mice as described previously.30 The controlmice inhaled PBS. The mouse colonies and all experimentalprocedures were performed according to the institutionalanimal care and use guidelines from the CPqRR/FIOCRUZ.The project was previously approved by the Ethics Com-mittee on Animal Experimentation (Comitê de Etica em Ex-perimentação Animal (CETEA) from Universidade Federalde Minas Gerais (UFMG) and Comitê de Etica em Utiliza-çào de Animais (CEUA) from CPqRR/FIOCRUZ).

Tissue Culture Infectious Dose Titration

Frozen mouse tissues were ground with sterile sand and300 �l (trigeminal ganglia) or 500 �l (brains) of Dulbecco’smodified Eagle’s medium containing 1% fetal bovine serumand antibiotics. Then the samples were centrifuged at6700 g for 10 minutes at 4°C, and the supernatants wereused for titration in a standard tissue culture infectious dose(TCID50) assay.31

RNA Extraction

Trigeminal ganglia and brains were aseptically removedand stored at �70°C until processing. RNA extraction wasperformed by using the TRIzol reagent (Invitrogen) accord-ing to the manufacturer’s procedures. One microliter of theextracted RNA was quantified with a Nanodrop ND-1000spectrophotometer at wavelengths of 260 and 280 nm.

Reverse Transcription

Reverse transcription was performed according to theprocedures provided by the manufacturer of the M-MLVRT enzyme (Promega, Madison, WI).

Real-Time PCR

Real-time quantitative PCR (Applied Biosystems, Carls-bad, CA) was performed to measure mRNA expression inthe trigeminal ganglia and brains of mice infected withHSV-1. The reactions were performed by using the SYBRGreen PCR Master Mix (Applied Biosystems) in an Ap-plied Biosystems 7000 Sequence Detection System andat 50°C for 2 minutes, 95°C for 10 minutes, and 40 cyclesof 95°C for 15 minutes and 60°C for 1 minute, followed bya final dissociation stage. The following oligonucleotideswere used in the reactions: Hypoxanthine-guanine phos-phoribosyltransferase (forward: 5�-GTTGGATACAGGCCAGA-CTTTGTTG-3�; reverse: 5�-GATTCAACTTGCGCTCATCTT-AGGC-3�); IFN � (forward: 5�-CTGGAGCAGCTGAATGG-AAA-3�; reverse: 5�-TGTCTGCTGGTGGAGTTCAT-3�); IP-10(CXCL10; forward: 5�-GCCGTCATTTTCTGCCTCAT-3�;reverse: 5�-GCTTCCCTATGGCCCTCATT-3�); MCP-1 (CCL2;forward: 5�-CTTCTGGGCCTGCTGTTCA-3�; reverse: 5�-CCAGCCTACTCATTGGGATCA-3�); MIP-1� (CCL3; for-ward: 5�-ACTGCCTGCTGCTTCTCCTA-3�; reverse: 5�-TTGGAGTCAGCGCAGATCTG-3�32); IL-1� (forward: 5�-CGCAGCAGCACATCAACAAGAGC-3�; reverse: 5�-TGT-CCTCATCCTGGAAGGTCCACG-3�); � trans-inducing fac-tor (forward: 5�-TTTGACCCGCGAGATCCTAT-3�; reverse:5�-GCTCCGTTGACGAACATGAA-3�33); TLR2 (forward: 5�-TTGCTCCTGCGAACTCCTAT-3�; reverse: 5�-AGCCTGGT-GACATTCCAAGA-3�); and TLR9 (forward: 5�-ACCTCAGC-CACAACATTCTC-3�; reverse: 5�-TGCACCTCCAACAGT-AAGTC-3�.34 The comparative Ct method with the formula2��Ct was used to analyze the data. Gene expression wasnormalized to the expression of the constitutively expressedgene Hypoxanthine-guanine phosphoribosyltransferase. Allreactions were replicated.

ELISA Assays

Supernatants from HEK293 cells (empty vector and TLR2and TLR4MD.2 transfected cells) were tested for thepresence of human IL-8 (BD) according to the manufac-turer’s protocols. Supernatants from macrophages stim-ulated with HSV-1 were tested for the presence of murineTNF � and murine IL-12 p40 (BD) according to the man-ufacturer’s protocols. Mice sera were tested for the pres-ence of murine IL-1� and CXCL10 (IP10) by using ELISAkits (R&D Systems, Minneapolis, MN) and for murineIFN-� and CCL2 (MCP1) by using the cytometric beadarray (CBA) mouse inflammation kit (BD) according to themanufacturer’s protocols.

Luciferase Activity Measurement

HEK TLR9 cells were cultured in 96-well plates (2 � 104

cells/well). After incubation for 1 day, cells were transientlytransfected (using Genejuice [Novagen, Darmstadt, Ger-many] according to manufacturer’s instructions) with a plas-mid containing an artificial promoter preceding the fireflyluciferase gene with five binding sites for nuclear factor �B.Cells were also co-transfected with a plasmid containing aconstitutively expressed Renilla-luciferase reporter gene

Trigeminal Cytokines in Response to HSV1 2435AJP November 2010, Vol. 177, No. 5

(Promega). After 24 hours, the cells were stimulated asdescribed in Human Embryo Kidney Cells. After 6 hours, thecells were lysed, and 20 �l of each protein extract was usedfor the activity measurement. Extracts were distributed in96-well plates, and 100 �l of luciferin was added to eachwell at room temperature immediately before the reading.Luciferase activity levels were detected in a Lumat LB 9501over 10 seconds of luminosity.

Histopathology

Brain samples were fixed with 10% formaldehyde in phos-phate buffer, routinely processed, and embedded in paraf-fin as previously described.23 For each group, topograph-ically matched temporal lobe and periventricular whitematter consecutive sections were scored in 14 microscopicfields (20� objective; n � 3 per group). The histopatholog-ical aspects of parenchyma and meninges were evaluatedby vascular reactivity (endothelial reactivity and proliferationand level of perivascular cell infiltration) and scored as mild(�), moderate (��), or intense (��). Vascular changes,edema, and cell infiltration of leptomeninges were scoredas mild (�), moderate (��), or intense (��).

Histopathology and Immunostaining

For trigeminal ganglia immunostaining, samples were fro-zen in Tissue-Tek O.C.T. compound (Sakura, Finetek, Tor-rance, CA), and 5-�m slices were cut with a HM505Nmicrotome cryostat (Mikron, Vista, CA). Tissues werestained as previously described35 with modifications (seeSupplemental Figure 1, A–E, at http://ajp.amjpathol.org).Briefly, the tissue sections were incubated with primaryantibodies for 2 hours, washed, and incubated with labeledsecondary antibody. The sections were counterstained withHoechst and mounted in Hydromount aqueous medium(National Diagnostics, Atlanta, GA). The primary antibodiesused were CD3 (1:100) and CD8 (1:100; Serotec, Raleigh,NC). The secondary antibody was Alexa Fluor 488 goatanti-rat IgG (1:500; Molecular Probes, Carlsbad, CA). Nu-clei counterstaining was performed by using Hoechst (0.2�g/ml; Molecular Probes). The stained sections were ob-served and photographed on an Olympus BX51 micro-scope (Olympus, Tokyo, Japan) by using a Megacybernet-ics color digital camera and the Image Pro-Expresssoftware.

Cell Preparation and Flow Cytometry

Single-cell suspensions were prepared from the TG of WTmice that were divided into control and infected groups.Tissues were digested with collagenase for 30 minutes at37°C. After incubation, the TGs were disrupted by grindingwith a syringe plunger. The cells were washed, suspendedin RPMI with 10% fetal bovine serum, and counted bytrypan blue exclusion with high viability. A total of 106 cellswere plated for macrophage and dendritic cell measure-ments, and 5 � 105 for CD4, CD8, and natural killer (NK)cell measurements. Cells were incubated for 1 hour in 5%CO2 at 37°C. Then the BD GolgiPlug Protein Transport

Inhibitor (BD Biosciences, San Jose, CA) was added to thewells where IFN-� production was evaluated, and the plateswere incubated for 8 hours. Plates were centrifuged at532 � g for 5 minutes at 4°C and suspended in Fluores-cence activated cell sorting (FACS) buffer solution withanti-mouse CD11c (FITC; BD Pharmingen, San Diego, CA),anti-mouse CD11b (PECy7; eBioscience, San Diego, CA),and anti-mouse F4/80 (APC; eBioscience) antibodies toidentify macrophages and dendritic cells; with anti-mouseNK1.1 (FITC; BD Pharmingen), anti-mouse CD4 (APC; eBio-science), anti-mouse CD8 (PECy5; BD Pharmingen), andanti-mouse I-Ab (PE; BD Pharmingen) antibodies to identifyCD4, CD8, and NK cells; and with FC Block (BD Pharmin-gen) for all of the markers. The cells were then incubated for20 minutes at room temperature, washed two times, andsuspended in 200 �l of FACS buffer. For T and NK cellsmarked with IFN-�, cells were incubated with BD cytofix/cytoperm solution (15 minutes at 4°C in the dark), followedby washing and incubation with BD Perm/Wash solution(150 ml/well). After washing and centrifugation at 532 � gfor 5 minutes at 4°C, the cells were suspended in BDPerm/Wash solution with IFN-� (PE; BD Pharmingen) andincubated for 30 minutes at 4°C in the dark. The cells werewashed with BD Perm/Wash and centrifuged twice (532 �g for 5 minutes, 4°C). The cells were finally suspended in200 �l of FACS buffer solution. For each sample, between100,000 and 200,000 events were captured for analysis.The data were analyzed using the FlowJo 7.2.5 software(Tree Star, Inc., Ashland, OR) and are presented as thepercentage of positive cells within the gated population.

Statistical Analysis

The real-time PCR results were statistically analyzed byusing Mann-Whitney nonparametric t-tests. The HEKcells results were analyzed with analysis of variancetests. The macrophage and flow cytometry results wereanalyzed with unpaired t-tests. The analyses were per-formed by using the GraphPad Prism 5 software forWindows (GraphPad Software, Inc., La Jolla, CA).

Results

Infectious HSV-1 Reaches the TG of C57BL/6Mice on Day 2 Postinfection and Peaks on Day5 Postinfection

C57BL/6 WT mice were intranasally infected with 106 PFU ofHSV-1. The animals were euthanized over the course of 8days postinfection (d.p.i.), and the brains and TG wereaseptically removed and snap frozen to verify the presenceof infectious virus particles. Mouse TG supernatants in-duced an HSV-1 cytopathic effect in Vero cells from 2 to6 d.p.i. (Table 1). For a few animals, it was possible tocalculate infectious particles in the TG (3 to 5 d.p.i.), with anaverage of 102,99 TCID50/ml. The mouse brain supernatantsdid not present any detectable infectious particles.


Viral, Cytokine, and TLR Transcripts AreIncreased in C57BL/6 Mice TG on Day 5Postinfection

Another group of C57BL/6 WT mice was intranasally in-fected with 106 PFU of HSV-1. This experiment was per-formed as described above, except that only TGs wereanalyzed to verify the expression of viral and cytokinetranscripts by real-time quantitative PCR. HSV-1 VirionProtein 16 (VP-16; also known as � trans-inducing factor),a late viral gene expressed only during the replicationcycle of the virus,33 was detected in mouse TG from 2 to8 d.p.i. (Figure 1A). The level of the VP-16 transcriptincreased until 5 d.p.i. (when it had a more pronouncedexpression) and decreased until 8 d.p.i. (Figure 1A). Thecytokines/chemokines IFN-�, CXCL10/IP-10, CCL2/MCP-1, and CCL3/MIP 1 � measured in mouse TGshowed a similar profile, with increased expression until

5 d.p.i., followed by a decrease until 8 d.p.i., when thelevels returned to baseline (Figure 1, B–E). The increasein the transcripts was only statistically significant for thechemokines CXCL10/IP10 (4, 5, and 6 d.p.i.; Figure 1C),CCL2/MCP-1 (5 and 6 d.p.i.; Figure 1D), and CCL3/MIP 1� (5 d.p.i.; Figure 1E). As for IFN-�, the level of the IFN �transcript was higher on day 5 postinfection but wasnot statistically significant (data not shown). TLRs 2and 9 expression levels were also significantly in-creased in WT infected mice on day 5 postinfection(Figure 1F). Thus, the presence of infectious virus par-ticles and its replication seemed to correlate with in-creases in cytokine and TLR expression in the TG ofC57BL/6 mice, with a peak of viral and cytokine ex-pression on day 5 postinfection.

HSV-1 Activates TLR2 and TLR9 in TransfectedHEK293 Cells

HEK293 cells stably transfected with plasmids ex-pressing TLR2, TLR4, or TLR9, or an empty vector werestimulated with HSV-1. Control cells had only basallevels of IL-8 expression for all stimuli tested (Figure2A). In TLR4-expressing cells (Figure 2B), HSV-1 didnot stimulate IL-8 production, indicating that our puri-fied virus was not contaminated with LPS and that thevirus does not activate this receptor. Only cells ex-pressing TLR2 (Figure 2C) or TLR9 (Figure 2D) wereactivated after viral stimulation, as measured by theproduction of IL-8 in the supernatants or by luciferaseactivity, respectively. These results indicate that TLR2and TLR9 recognize HSV-1 and stimulate an immuneresponse against the virus.

Table 1. Kinetics of HSV-1 Multiplication in C57BL/6 MouseOrgans

d.p.i.CPE of TGs in

Vero cells* TCID50/ml in TGs†CPE of brainsin Vero cells*

1 ND‡ ND ND2 �(3/4) ND ND3 �(3/4) 103,3 (1/4) ND4 �(3/4) 102,8 (1/4) ND5 �(3/4) 102,7 and 102,9(2/4) ND6 �(2/4) ND ND7 ND ND ND8 ND ND ND

*Cytopathic effect (CPE) detected in Vero cells infected with TG orbrain supernatants; n/4 indicates the number of positive mice in a totalof four mice per day.

†TCID50/ml, median tissue infective dose.‡ND, virus was not detected.

Figure 1. Patterns of viral, cytokine, and che-mokine transcript expression and TLR expres-sion in C57BL/6 mouse trigeminal ganglia.C57BL/6 mice were intranasally infected with106 PFU HSV-1, and four pools of trigeminalganglia of three animals were collected each dayfrom 1 to 8 d.p.i. Uninfected mice (NI) aspiratedonly PBS. After RNA extraction and reverse tran-scription, real-time PCR was performed. (A)VP-16 HSV-1 transcript; (B) IFN-�; (C) CXCL10(IP-10); (D) CCL2 (MCP-1); and (E) CCL3 (MIP 1�). In F, TLR2 and 9 expression levels weremeasured in WT mice on day 5 postinfection.*P � 0.05. Statistical analyses were performedwith Kruskal-Wallis nonparametric tests andDunn’s multiple comparison tests. Bars repre-sent the SEM. The results shown are representa-tive of two experiments that yielded similarresults.


TLR2 and TLR9 Contribute to Production ofIL-12p40 and TNF � in Macrophages

C57BL/6 WT, TLR2�/�, TLR9�/�, and TLR2/9�/� mouse-derived macrophages were stimulated with HSV-1, andthe levels of IL12p40 and TNF � were measured in thecell supernatants. All infected macrophages derived fromTLR null mice showed reduced production of cytokineswhen compared with infected macrophages derived fromWT infected mice (Figure 3, A and B). TLR2 and TLR9seemed to contribute synergistically to the production ofIL12p40 in macrophages because the double KO micehad a more pronounced reduction in this cytokine com-pared with the single KOs (Figure 3A).

TLR9�/� and TLR2/9�/� Mice Have HigherSusceptibility to HSV-1 Intranasal Infection

C57BL/6 WT, TLR2�/�, TLR9�/�, and TLR2/9�/� micewere intranasally infected with 106 PFU of HSV-1. Micewere observed daily for clinical signs of encephalitis(prostration, ruffled fur, hunched posture, and posteriorpaw paralysis). After the symptoms were observed, mice

were euthanatized, and their brains were collected andsnap frozen to verify the presence of the virus. Most of themice died on day 6 postinfection, but the TLR2/9�/� micebegan to die earlier than the other groups (Figure 4).C57BL/6 WT and TLR2�/� infected mice had low mortal-ity rates of around 10% (Figure 4). TLR9�/� infectedanimals had a more pronounced mortality of approxi-mately 60% (Figure 4). The mortality was even higher forTLR2/9�/� infected mice, with 100% of the mice dyingfrom infection (Figure 4). Brain TCID50 titrations demon-strated that all euthanized animals with encephalitis hadHSV-1 in their brains (Table 2), and the virus titers weretwo logs higher in TLR9�/� and TLR2/9�/� mice and onelog lower in TLR2�/� mice compared with the titer of WTmice (Table 2). Infectious virus particles were not found inthe brains of mice without signs of encephalitis (Table 2).

TLR2/9�/� Mice Have Major HistopathologicalChanges in the Brain on Day 5 Postinfection

For all experiments described from this point on, C57BL/6WT, TLR2�/�, TLR9�/�, and TLR2/9�/� mice were intra-nasally infected with 106 PFU of HSV-1. Because the

Figure 2. TLR2 and TLR9 activation in HEK cellsafter HSV-1 stimulation. HEK cells were stablytransfected with plasmids expressing TLR2,TLR4, or TLR9 and were then stimulated withHSV-1 (105 PFU/ml in A, B, and C; or at MOIs of2 and 10 in D), LPS (10 ng/ml), Malp-2 (10ng/ml), CpG 2006, and 1826 (5 �mol/L), or E.coli (100 units/cell), as indicated in the figure.After 24 hours of stimulation, IL-8 in the super-natants (for A, B, and C) or the relative increasein luciferase expression (for D) was measured.(A) HEK 293; (B) HEK TLR4; (C) HEK TLR2; and(D) HEK TLR9. The results shown are represen-tative of two experiments that yielded similarresults. Statistical analyses were performed withanalysis of variance tests, and the bars representthe SEM. *P � 0.05; statistical difference betweenthe bar and the respective negative control (me-dium stimulated). **P � 0.05; statistical differ-ence between the indicated bars.

Figure 3. TNF � and IL-12 p40 production inperitoneal macrophages stimulated with HSV-1.A and B: Macrophages derived from C57BL/6WT mice (white columns), TLR2�/� mice(striped columns), TLR9�/� mice (pointed col-umns), or TLR2/9�/� mice (black columns)were exposed to HSV-1 (MOI of 10), and thelevels of IL12 p40 (A) and TNF � (B) weremeasured in the culture supernatants 24 hoursafter stimulation. This experiment is representa-tive of two experiments. Statistical analyses wereperformed with unpaired t-tests, and the barsrepresent the SEM. *P � 0.05; statistical differ-ence between the bar and the respective nega-tive control (medium stimulated). **P � 0.05;statistical difference between the indicated bars.


previous experiments indicated that 5 d.p.i. was probablyan important point for virus control in WT mouse TG andbecause many mice died after day 5, we euthanized theanimals on day 5 postinfection. Half of each brain wasformalin fixed for histopathological analysis. Microscopicexamination of the brains stained with H&E on day 5postinfection revealed slight endothelial cell reactivityand perivascular edema of parenchyma and meninges ofWT and TLR2�/� infected mice (Figure 5, A–I; Table 3).TLR9�/� mice presented an intermediate intensity ofchanges, as shown in Figure 5, A–I, and Table 3. Themost severe neuropathological signs of encephalitis wereseen in TLR2/9�/� infected mice and were characterizedby intense leptomeningitis associated with edema, whitematter vacuolization, endothelial reactivity, and perivas-cular edema (Figure 5, J–L; Table 3). Focal encephalitischaracterized by mononuclear cell infiltrates and acti-vated glial cells associated with perivascular cuffing werefound exclusively in these animals (Figure 5; Table 3).

Cytokine and Chemokine Transcript Profiles AreAltered in the TG of TLR2/9�/� Mice on Day 5Postinfection

To compare viral and cytokine expression in the TG andbrains of WT and knockout mice, sera, TG and half of eachbrain were aseptically removed from the animals, snapfrozen, and processed for analysis. Real-time quantitativePCR indicated that the HSV-1 VP-16 mRNA was detected insimilar amounts in TG of mice from all groups (Figure 6A).All cytokine transcripts measured were significantly up-reg-ulated in the TG of all groups of infected mice but in differentamounts (Figure 6, B–E). IFN-� (Figure 6B) was down-regulated in all infected knockout mice compared with in-fected WT mice. A lower level of expression of IL-1� (Figure6C) was found in TLR2/9�/� infected mice compared with

the other groups of infected animals. CXCL10/IP-10 (Figure6D) was overexpressed in infected TLR2/9�/� animalscompared with the other groups. CCL2/MCP-1 (Figure 6E)had a similar profile, but there was not a difference betweeninfected TLR2/9�/� and TLR9�/� mice. IFN � and CCL3/MIP 1 � had similar levels of transcript up-regulation in allinfected groups (data not shown).

Cytokine Expression Profile Alterations Are NotEvident in TLR2/9�/� Mouse Brains and Seraon Day 5 Postinfection

Brains of the same animals from the experiment describedabove were analyzed for cytokine transcripts by real-timequantitative PCR. The cytokine transcripts were significantlyup-regulated in brains of almost all groups of infected mice(in different amounts; Figure 7, A–D) compared with unin-fected mice, except for IFN-� and IL1 � transcripts in TLR2/9�/� mice (Figure 7, A and B) and for IFN-� in WT mice(Figure 7A). CXCL10/IP-10 (Figure 7C) and CCL2/MCP-1(Figure 7D) were over expressed in infected TLR2/9�/�

animals compared with the other groups. Unlike the resultsfor TG, IFN-� transcripts (Figure 7A) were not up-regulatedin infected C57BL/6 WT mouse brains compared with in-fected knockouts; however, the transcript levels in brains ofinfected TLR2/9�/� mice were not increased comparedwith the respective control mice. A lower level of IL-1�(Figure 7B) was observed in TLR2/9�/� mice comparedwith the other groups of infected animals, but this differencewas significant only when compared with infected WT mice.IFN � and CCL3/MIP 1 � showed similar patterns of up-regulation in all infected groups (data not shown). To quan-tify cytokine production in mice sera, ELISA and CBA wereperformed. IP-10/CXCL10 and IL 1� were not detected inthe sera (data not shown), although there were detectableamounts of IFN-� and MCP-1/CCL2 in infected mice com-pared with controls (Figure 7, E and F). However, onlyMCP-1/CCL2 exhibited a significant difference between thegroups, which was shown by an increase in production inTLR2/9�/� when compared with WT mice (Figure 7F).

TG of HSV-1 Infected C57BL/6 Mice HaveIncreased Macrophage Population inComparison to TG of Noninfected Mice

To study the cell populations in TG of infected and non-infected WT mice, we performed Giemsa staining of dif-ferent areas of WT control (see Supplemental Figure 1A athttp://ajp.amjpathol.org) and WT infected (see Supple-

Figure 4. Survival rates of mice intranasally inoculated with HSV-1. C57BL/6(crosses), TLR2�/� (triangles), TLR9�/� (squares), and TLR2/9�/� (circles)mice were intranasally inoculated with 106 PFU HSV-1, and mouse mortalitywas quantified daily (n � 10 for each group). This experiment is represen-tative of three analyses.

Table 2. HSV-1 Presence in Mouse Brains

Mouse group

Mice without signs of encephalitis Mice with signs of encephalitis

% Mean brain TCID50/ml % Mean brain TCID50/ml

C57BL/6 92 0 8 102

TLR2�/� 90 0 10 10TLR9�/� 40 0 60 104

TLR2/9�/� 0 0 100 104


mental Figure 1B at http://ajp.amjpathol.org) ganglia.The data showed increased cellularity of the gangliaroot (double asterisks) and close to neurons (singleasterisk), where many small mononuclear cells were

detected in infected mice (arrows). Furthermore, flowcytometry showed an increase in macrophages in theTG of WT infected mice compared with noninfectedmice (Figure 8A).

Figure 5. Histopathological analysis of knockout mouse brains on day 5 postinfection. C57BL/6, TLR2�/�, TLR9�/�, and TLR2/9�/� mice were intranasallyinfected with 106 PFU HSV-1, and the brains of three animals per group were collected on day 5 postinfection and processed for histopathological analysis. A–I:Changes in the brain parenchyma and meninges of C57BL/6, TLR2�/�, and TLR9�/� mice were mild. J and K: Intense leptomeninges cell infiltration and edemawere apparent in TLR2/9�/� mice (asterisks). L: Endothelial vascular reactivity in a small vessel of the brain parenchyma vessels was more intense in TLR2/9�/�

mice (arrows). In C, note the regular aspect of the parenchymal vessels (arrows). Original magnification of the mouse brain, �100 (left). Original magnificationof the mouse brain meninges, �400 (middle). Original magnification of the mouse brain blood vessels, �400 (right).

Table 3. Histopathological Changes in HSV-1-Infected Mice at 5 d.p.i.

Parameters*Endothelial reactivity and

proliferation

Perivascular cuffing andperivascular cell

infiltrationEdema and cell infiltration

of leptomeninges

GroupsC57BL/6 � (3/3) � (3/3) � (3/3)TLR2�/� � (3/3) � (3/3) � (3/3)TLR9�/� �� (3/3) �� (3/3) � (3/3)TLR2/9�/� �� (3/3) �� (3/3) �� (3/3)

*Histopathological changes were scored as mild �, moderate ��, or intense ��; (3/3), three of three examined animals presented the degree ofindicated changes.


CD8 Cells Produce a Higher Amount of IFN-� inTG after Infection with HSV-1

We detected CD3 (see Supplemental Figure 1D at http://ajp.amjpathol.org) and CD8 (see Supplemental Figure 1E

at http://ajp.amjpathol.org) positive cells in the TG of in-fected mice but not in noninfected mice (see Supplemen-tal Figure 1C at http://ajp.amjpathol.org). Additionally, weperformed flow cytometry analysis of TG after treatmentwith collagenase. Although the differences in CD8 T cells

Figure 6. Viral, cytokine, and chemokine transcripts in knockout mouse trigeminal ganglia on day 5 postinfection. C57BL/6, TLR2�/�, TLR9�/�, and TLR2/9�/�

mice were intranasally infected with 106 PFU HSV-1, and trigeminal ganglia of six to nine animals per group were collected on day 5 postinfection. Control miceaspirated only PBS (n � 4 for each group). After RNA extraction and reverse transcription, real-time PCR was performed. (A) VP-16 HSV-1 transcript; (B) IFN-�;(C) IL-1�; (D) CXCL10 (IP-10); and (E) CCL2 (MCP-1). This experiment is representative of two replicates. *P � 0.05; statistical difference between the bar andthe respective negative control. **P � 0.05; statistical difference between the indicated bars. Statistical analyses were performed with Mann-Whitney t-tests, andthe bars represent the SEM.

Figure 7. Cytokine and chemokine transcriptexpression levels in knockout mouse brains andcytokine production in mice sera on 5 d.p.i.C57BL/6, TLR2�/�, TLR9�/�, and TLR2/9�/�

mice were intranasally infected with 106 PFUHSV-1, and brains and sera of six to nine animalsper group were collected on day 5 postinfection.Control mice aspirated only PBS (n � 4 for eachgroup). After RNA extraction and reverse tran-scription of the brains, real-time PCR was per-formed. (A) IFN-�; (B) IL-1�; (C) CXCL10 (IP-10); and (D) CCL2 (MCP-1). In the sera, levels ofIFN-� (E) and CCL2 (MCP-1; F) were measuredby CBA. The results are representative of twoexperiments that yielded similar results. *P �0.05; statistical difference between the bar andthe respective negative control. **P � 0.05; sta-tistical difference between the indicated bars.Statistical analyses were performed with Mann-Whitney t-tests, and the bars represent the SEM.


were not statistically significant between WT infected andnoninfected TG mice (Figure 8A), the CD8 T cells ofinfected mice produced a higher level of IFN-� than theCD8 T cells of noninfected mice (Figure 8B).

Discussion

Innate immunity to HSV-1 has been studied by manydifferent research groups in recent years.20–24,26,27,35,36

Although these studies show the importance of TLRs andother immune receptors and cytokines/chemokines inHSV-1 recognition and control, some discrepancies re-main concerning the benefits of these responses. Thedifferences in results are probably due to the high diver-sity in the animal models used in these works.37 Viralstrain, mouse strain, and inoculation route variations mayproduce different immune responses,38 making compar-isons between the results of each experiment difficult.However, each different model may have indications forone specific clinical manifestation of the broad spectrumof herpetic manifestations. Our research group uses astrain of HSV-1 that was isolated from a naturally occur-ring human lip recurrence. Additionally, the intranasalinoculation route that we used in our experiments30 issimilar to the natural route of infection because the lips(natural infection) and nose have common innervations.

In this work, we demonstrated that most C57BL/6 WTmice, which have a low susceptibility to an intranasalinoculation of 106 PFU of HSV-1, are capable of control-ling viral replication in TG, which prevents the infectiousvirus from reaching the brain, because the infectiousvirus was found in TG but not in the brains of WT animals.HSV-1 infection generally leads to a localized infection;the virus initiates replication in mucosa and then reachesthe dorsal root ganglia termini, from which it is intrax-onally transported to the TG, where it replicates or be-comes latent.3 However, sometimes, for reasons that arenot well understood, the virus reaches the brain, where itcauses encephalitis.6 Using an intranasal inoculum of104 PFU, a previous study from our group demonstratedby immunohistochemistry that only mice with signs ofencephalitis expressed viral proteins in brain cells, but allinfected mice expressed viral proteins in TG cells8 d.p.i.23 With the higher PFU inoculation used in thisstudy, the virus reached the TG in WT mice around2 d.p.i. (Table 1; Figure 1A), replicated until 6 d.p.i.(Table 1) and then was controlled by the mice, although

the viral VP-16 transcripts remained, in lower amounts,until day 8 postinfection (Figure 1A), which was the lastday analyzed. In most of the WT mice, once the virusreached the TG, it stimulated the expression of cytokinesand chemokines that control virus replication, preventingmost of the infectious virus from reaching the brain. Thisresponse seems to be very well regulated because theexpression of these molecules in WT mouse TG was moreelevated on day 5 postinfection (Figure 1, B–E)—whenthe virus is replicating in higher amounts and TLR expres-sion is increased (Figure 1F)—and returns to basal levelsas soon as the infectious virus particles are controlled.The TLR-dependent immune responses in TG seem to befundamental for controlling HSV-1 infection and prevent-ing the virus from reaching the brain, where it couldcause lethal encephalitis.

Many groups have already shown that TLR2 and TLR9recognize HSV-1 or HSV-2, which activates an immuneresponse to the virus.20–25 Because different viral strainsmay elicit different responses,31 we decided to confirmthat our isolate was able to activate those receptors.Using HEK293 cells stably expressing TLRs, we haveshown that HSV-1 activates TLR2 and TLR9 but not TLR4or cells that do not express these receptors (Figure 2).Using similar cell constructs, Kurt-Jones et al20 could notdetect TLR9 activation on HSV-1 KOS stimulation (butthey detected TLR2 activation). Although Kurt-Jones etal20 did not determine whether this was due to a technicallimitation, Sato et al24 found similar results when compar-ing TLR stimulation of HSV-1 KOS and other HSV strains;however, Sato et al24 also demonstrated that, althoughnot recognized by TLR9 in HEK cells, the analyzed virusseemed to use this receptor to stimulate dendritic cellresponses. Because different systems seemed to elicitdifferent responses, we investigated whether our HSV-1strain elicited TLR2 and TLR9 responses in HEK cells andin macrophages. We discovered that both receptorswere important for cytokine production in these cells, butthe level of importance varied with the type of cytokine. Inmacrophages (Figure 3), TLR2 and TLR9 contribute toIL12p40 and TNF � production. Although, there seemedto be some cooperation between these receptors forIL12p40 (because the double KO had a lower productionwhen compared with the single KOs), an effect that wasnot observed for TNF �. Thus, the importance of the TLRsvaries by cytokine type.

Figure 8. TG of HSV-1-infected C57BL/6 micehave increased macrophage population andCD8 cells that produce IFN-�. Mice were intra-nasally infected with 106 PFU of HSV-1, and theTG of five animals per group were collected onday 5 postinfection. Control mice aspirated onlyPBS (n � 3 for each group). TGs were treatedwith collagenase, and the population of cellswas analyzed by flow cytometry by using anti-bodies against the indicated cells (A). An anti-body against IFN-� was used in CD8 T cells (B).All data presented are the mean � SEM; *P �0.05 for comparison by t-test.


After we demonstrated that our isolate stimulated TLR2and TLR9 in vitro and ex vivo, we analyzed the roles ofthese receptors in vivo. C57BL/6 WT, TLR2�/�, TLR9�/�,and TLR2/9�/� mice were infected with HSV-1, and themortality rates indicated that TLR2 plays a minor role inmouse survival because TLR2�/� mice had a low mortal-ity rate that was very similar to that of WT mice (Figure 4).TLR9 seemed to play a major role in infection controlbecause null animals showed a higher mortality rate com-pared with WT mice (Figure 4). In a previous study,23 itwas shown that MyD88 is fundamental for immune de-fense against HSV-1, with 100% mortality observed inMyD88�/� mice and 50% mortality in IFN-� �/� mice. Ourdata strongly suggested, for the first time, that TLR2 andTLR9 cooperate in vivo to induce this MyD88-dependentresponse in HSV-1 infected mice; we showed that infec-tion of TLR2/9�/� mice results in 100% mortality (Figure4), as previously shown for MyD88�/� mice. All animalsthat died with signs of encephalitis (prostration, ruffledfur, hunched posture, and posterior paw paralysis) hadinfectious virus particles in their brains, which were notobserved for animals without signs of encephalitis (Table2). Therefore, the presence of infectious HSV-1 in mousebrains correlated with the signs of encephalitis anddeath, as previously shown.23,39 The virus titers werehigher in infected TLR9�/� and TLR2/9�/� animals com-pared with infected WT or TLR2�/� animals (Table 2).

The analysis of cytokine and chemokine expressionindicated that high expression of CXCL10/IP-10 andCCL2/MCP-1 was related to deaths because infectedTLR2/9�/� mice had an increased level of these chemo-kines in the TG and brain (Figures 6 and 7). Kurt-Jones etal20 have already shown that mice that succumbed toHSV-1 infection have high levels of the CCL2/MCP-1transcript in the brain. Although it may be important tocontrol the virus, CXCL10/IP-10 could also have somedeleterious role in HSV-1 infection because mice treatedwith an anti-CXCL10 antibody showed prolonged sur-vival, even though viral yields were significantly greater inthe infected tissue.40 Importantly, we show a correlationbetween low levels of IFN-� and IL-1� transcripts in TGand poor survival; all infected knockout mice showedreduced expression of IFN-� in TG (Figure 6). In WT mice,the level of IFN-� in TG would already be sufficient tocontrol the virus, whereas the TLR2�/� mice still pro-duced IFN-� in the brain because production in the TGwas lower than in WT mice (Figures 6 and 7).

Unlike TLR2�/� mice, TLR2/9�/� mice had highamounts of virus in the brain (Table 2), and the absenceof IFN-� in brains of the latter (Figure 7) seems to beimportant for their immune response. IFN-� possesses animportant antiviral potential and is capable of controllingviral replication during the early phases of infection.41

Additionally, He et al42 also found that IFN �-deficientmice had a higher mortality rate and delayed virus clear-ance after an ocular inoculation of the HSV-1 Amakatastrain. However, as we have reported previously,23 50%of IFN-� �/� mice died when infected with HSV-1,whereas all infected MyD88�/� mice died. Therefore,although expression of IFN-�, either in the TG or brain, isnecessary to control the virus, there must be another

MyD88-dependent process that is important for total im-mune defense against HSV-1. Interestingly, IL-1� signal-ing is MyD88 dependent.43 It has been shown that IL-1�plays a critical role in Herpes Simplex virus encephalitis;in an intranasal murine model of infection with HSV-1strain H25, IL-1 knockout mice had lethal encephalitis,higher brain viral loads, and inadequate immune re-sponses in the brain.44 IL-1� also has a role in promotingthe repair of the central nervous system.45 We observedthat IL-1� expression levels were significantly reduced ininfected TLR2/9�/� mice TG and brains (Figures 6 and7), and it was not reduced in the TG of infected TLR9�/�

or TLR2�/� mice (Figure 6). This cytokine was also notreduced in the brain of TLR2�/� mice (Figure 7), and itwas reduced in TLR9�/� mice, although it was not sta-tistically lower compared with expression in WT mice.This reduction in IL-1� could explain why TLR2�/� andTLR9�/� mice have a insignificant or only 60% mortality,respectively, compared with 100% mortality in TLR2/9�/�

mice, even though both TLR2�/� and TLR9�/� mice, aswell as TLR2/9�/� mice, produced lower levels of IFN-� inthe TG compared with WT mice. Additionally, TLR2�/�

mice showed a higher level of IFN-� expression in thebrain (Figure 7) and a higher expression level of IL-1� inthe TG and brain (Figures 6 and 7) compared with TLR2/9�/� mice; this increased cytokine expression could beresponsible for the effective immune response againstHSV-1 in this knockout mouse.

Our results showed that IFN � and MIP1 � were up-regulated in the TG of all groups of infected mice comparedwith uninfected mice (data not shown). Although importantin HSV control,46–48 expression of these molecules proba-bly does not depend on TLR2 or TLR9 in the TG.

Furthermore, TLR2/9�/� infected mice have major his-topathological alterations in the brain compared withthose observed in WT brains or in the brains of the otherknockout mice evaluated (Figure 5; Table 3). The brainparenchyma of TLR2/9�/� mice had focal lesions, andthe infiltration of meninges was visibly more intense thanthe degenerative changes in the brain parenchyma.Brain meninges were compromised in infected TLR2/9�/� mice, as shown by the massive presence of inflam-matory cells inside and outside the meningeal vessels.Similar brain alterations were found in susceptible mice inanother study.20 Marques et al49 suggested that neuro-logical damage might be related to CXCL10/IP-10 andCCL2/MCP-1 production and that these chemokines areprobably microglia derived, a hypothesis that demandsfurther in vivo and in vitro investigation. Our data indicatedthat TLR2/9�/� mice had significantly higher histopathol-ogy scores (brain and meninges) than WT and TLR2�/�

mice, but the differences were less apparent when com-pared with TLR9�/� mice.

Together, the results indicate that although TLR9seems to have a more critical role in vivo than TLR2, bothreceptors are not only important receptors in HSV-1 rec-ognition and control but also may cooperate in the acti-vation of the immune response because the doubleknockout mice had higher mortality rates and more prom-inent immune alterations and brain pathology on day 5postinfection. Immunological and pathological alterations


were not as evident in TLR9�/� mice, probably due todifferences in kinetics compared with TLR2/9�/� micebecause the latter began to die earlier than the former.Additionally, IFN � and CCL3/MIP 1 � transcripts hadsimilar expression levels in the TG for all groups of ani-mals, suggesting that other receptors could be involvedand that other pattern recognition receptors may alsocontribute to HSV-1 control. Retinoic acid inducible gene-like receptors and DNA-dependent activator of IFN reg-ulatory factors are some of the pattern recognition recep-tors that have also been shown to recognize HSV.50,51

Finally, we have shown that infection control dependson an immune response in the TG and brain in the herpesintranasal model. HSV-1 control seems to be mainly localbecause the only cytokine profile altered in knockoutsusceptible mice sera was that of CCL2/MCP1 (Figure7F). The response in the TG seems to limit virus replica-tion to prevent too much virus (if any) from reaching thebrain—an essential process to prevent encephalitis. Theresponse in the brain may not only be an attempt tocontain the virus but also may be designed to preventbrain damage. These responses must be well regulatedbecause not only the virus but also the immune responseitself may cause tissue damage and encephalitis.52 Thisindicates that, in addition to activating cytokine and che-mokine production, TLRs may have a role in the regula-tion of the immune response. This idea is corroborated byrecent findings that indicate that TLRs may suppress theimmune response of T cells and NK cells,53 control Bcells responses,54 transiently silence pro-inflammatorygenes,55 modulate chemokine receptors, and redirectleukocyte migration.56 Because the brain histopathology5 d.p.i. showed mononuclear infiltration only in the me-ninges of TLR2/9�/� mice, we must further investigateTLR participation in HSV-1 control by identifying all of thecells that are responsible for cytokine/chemokine produc-tion in the TG and brains of resistant and susceptiblemice. However, on day 5 postinfection the TG of WTinfected mice show increased cellularity of the gangliaroot and in areas close to neurons, where many smallmononuclear cells were detected by Giemsa stainingand macrophages were detected by flow cytometry.These mononuclear cells could be the TLR-producingcells. Additionally, cells of the TG from infected miceshowed positive immunostaining for CD8 T cells andproduced higher levels of IFN-�, which we showed byflow cytometry. Our next goal is to study in detail eachcell of the TG that participates in the immune responseand its respective cytokine. The data from this workshowed that the immune response in the TG is crucial toprevent against encephalitis, contributing to the under-standing of innate immune responses to HSV-1 and may,in the future, provide some clues to aid in the fight againstviral brain pathogens.

Acknowledgments

We thank Dr. Douglas T. Golenbock (University of Mas-sachusetts Medical School) for providing us with the HEKcell lines; Dr. Shizuo Akira (Osaka, Japan) for providing

the TLR2�/� and TLR9�/� mice; and Dr. Alan Sher (NIH,Bethesda, MD) for providing us with the TLR2/9�/� mice.

References

1. Cowan FM, French RS Mayaud P, Gopal R, Robinson NJ, Artimos deOliveira S, Faillace T, Uuskula A, Nygard-Kibur M, Ramalingam S,Sridharan G, El Aouad R, Alami K, Rbai M, Sunil-Chandra NP, BrownDW: Seroepidemiological study of herpes simplex virus types 1 and 2in Brazil, Estonia, India, Morocco and Sri Lanka. Sex Transm Infect2003, 79:286–290

2. Fatahzadeh M, Schwarz RA: Human herpes simplex virus infections:Epidemiology, pathogenesis, symptomatology, diagnosis, and man-agement. J Am Acad Dermatol 2007, 57:737–763

3. Roizman B, Knipe DM, Whitley RJ: Herpes Simplex Viruses. Fieldsvirology ed 5. Edited by DM Knipe, RM Howley. Philadelphia, Lippin-cott Williams & Wilkins, 2007, pp 2501–2576

4. Garner JA: Herpes simplex virion entry into and intracellular transportwithin mammalian cells. Advanced Drug Delivery Reviews 2003,55:1497–1513

5. Theil D, Derfuss T, Paripovic I, Herberger S, Meinl E, Schueler O,Strupp M, Arbusow V, Brandt T: Latent Herpes virus infection inhuman trigeminal ganglia causes chronic immune response. Am JPathol 2003, 163:2179–2184

6. Whitley RJ: Herpes simplex encephalitis: adolescents and adults.Antiviral Res 2006, 71:141–148

7. Ishii KJ, Koyama S, Nakagawa A, Coban C, Akira S: Host innateimmune receptors and beyond: making sense of microbial infections.Cell Host Microbe 2008, 3:352–363

8. Rasmussen SB, Reinert LS, Paludan SR: Innate recognition of intra-cellular pathogens: detection and activation of the first line of de-fense. APMIS 2009, 117:323–337

9. Takeuchi O, Hoshino K, Kawai T, Sanjo H, Takada H, Ogawa T,Takeda K, Akira S: Differential roles of TLR2 and TLR4 in recognitionof Gram negative and Gram positive bacterial cell wall components.Immunity 1999, 11:443–451

10. Campos MA, Almeida IC, Takeuchi O, Akira S, Paganini E, ProcopioDO, Travassos LR, Smith JA, Golenbock DT, Gazzinelli RT: Activationof Toll-like receptor-2 by glycosylphosphatidylinositol anchors from aprotozoan parasite. J Immunol 2001, 167:416–423

11. Campos MA, Rosinha GMS, Almeida IC, Salgueiro XS, Jarvis BW,Splitter GA, Bruna-Romero O, Gazzinelli RT, Oliveira SC: The role ofToll-like receptor 4 in induction of cell-mediated immunity and resis-tance to Brucella abortus infection in mice. Infect Immun 2004,72:176–186

12. Latz E, Visintin A, Lien E, Fitzgerald KA, Monks BG, Kurt-Jones EA,Golenbock DT, Espevik T: Lipopolysaccharide rapidly traffics to andfrom the Golgi apparatus with the toll-like receptor 4-MD-2-CD14complex in a process that is distinct from the initiation of signaltransduction. J Biol Chem 2002, 277:47834–47843

13. Lien E, Means TK, Heine H, Yoshimura A, Kusumoto S, Fukase K,Fenton MJ, Oikawa M, Qureshi N, Monks B, Finberg RW, Ingalls RR,Golenbock DT: Toll-like receptor 4 imparts ligand-specific recognitionof bacterial lipopolysaccharide. J Clin Invest 2000, 105:497–504

14. Poltorak A, He X, Smirnova I, Liu MY, Van Huffel C, McNally O,Birdwell D, Alejos E, Silva M, Galanos C, Freudenberg M, Ricciardi-Castagnoli P, Layton B, Beutler B: Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: mutations in TLR4 gene. Science1998, 282:2085–2088

15. Alexopoulou L, Holt AC, Medzhitov R, Flavell RA: Recognition ofdouble-stranded RNA and activation of NFkB by Toll-like receptor 3.Nature 2001, 413:432–438

16. Hayashi F, Smith KD, Ozinsky A, Hawn TR, Yi EC, Goodlett DR, EngJK, Akira S, Underhill DM, Aderem A: The innate immune response tobacterial flagelin is mediated by Toll-like receptor 5. Nature 2001,410:1099–1103

17. Hemmi H, Takeuchi O, Kawai T, Kaisho T, Sato S, Sanjo H, Matsu-moto M, Hoshino K, Wagner H, Takeda K, Akira S: A Toll-like receptorrecognizes bacterial DNA. Nature 2000, 408:740–745

18. Takeda K, Kaisho T, Akira S: Toll-like receptors. Annu Rev Immunol2003, 21:335–376

19. Yamamoto M, Takeda K, Akira S: TIR domain-containing adaptors


define the specificity of TLR signaling. Mol Immunol 2004,40:861–868

20. Kurt-Jones EA, Chan M, Zhou S, Wang J, Reed G, Bronson R, ArnoldMM, Knipe DM, Finberg R W: Herpes simplex virus 1 interaction withtoll-like receptor 2 contributes to lethal encephalitis. Proc Natl AcadSci USA 2004, 101:1315–1320

21. Hochrein H, Schlatter B, O’Keeffe M, Wagner C, Schmitz F, Schi-emann M, Bauer S, Suter M, Wagner H: Herpes simplex virus type-1induces IFN-alpha production via Toll-like receptor 9-dependent and-independent pathways. Proc Natl Acad Sci USA 2004, 101:11416–11421

22. Krug A, Luker GD, Barchet W, Leib DA, Akira S, Colonna M: Herpessimplex virus type 1 (HSV-1) activates murine natural interferon-producing cells (IPC) through Toll-like receptor 9. Blood 2004,103:1433–1437

23. Mansur DS, Kroon EG, Nogueira ML, Arantes RME, Rodrigues SCO,Akira S, Gazzinelli RT, Campos MA: Lethal encephalitis in myeloiddifferentiation factor 88: deficient mice infected with herpes simplexvirus 1. Am J Pathol 2005, 166:1419–1426

24. Sato A, Linehan MM, Iwasaki A: Dual recognition of herpes simplexviruses by TLR2 and TLR9 in dendritic cells. Proc Natl Acad Sci USA2006, 103:17343–17348

25. Sorensen LN, Reinert LS, Malmgaard L, Bartholdy C, Thomsen AR,Paludan SR: TLR2 and TLR9 synergistically control herpes simplexvirus infection in the brain. J Immunol 2008, 181:8604–8612

26. Casrouge A, Zhang S, Eidenschenk C, Jouanguy E, Puel A, Yang K,Alcais A, Picard C, Mahfoufi N, Nicolas N, Lorenzo L, Plancoulaine S,Senechal B, Geissmann F, Tabeta K, Hoebe K, Du X, Miller RL, HeronB, Mignot C, de Villemeur TB, Lebon P, Dulac O, Rozenberg F,Beutler B, Tardieu M, Abel L, Casanova JL: Herpes simplex virusencephalitis in human UNC-93B deficiency. Science 2006, 314:308–312

27. Zhang S, Jouanguy E, Ugolini S, Smahi A, Elain G, Romero P, SegalD, Sancho-Shimizu V, Lorenzo L, Puel A, Picard C, Chapgier A,Plancoulaine S, Titeux M, Cognet C, von Bernuth H, Ku CL, CasrougeA, Zhang XX, Barreiro L, Leonard J, Hamilton C, Lebon P, Heron B,Vallee L, Quintana-Murci L, Hovnanian A, Rozenberg F, Vivier E,Geissmann F, Tardieu M, Abel L, Casanova JL: TLR3 deficiency inpatients with herpes simplex encephalitis. Science 2007, 317:1522–1527

28. Nogueira ML, Siqueira RC, Freitas N, Amorim JB, Bonjardim CA,Ferreira PC, Orefice F, Kroon EG: Detection of herpesvirus DNA bythe polymerase chain reaction (PCR) in vitreous samples from pa-tients with necrotising retinitis. J Clin Pathol 2001, 54:103–106

29. Campos MA, Kroon EG: Critical period of irreversible block of Vac-cinia virus replication. Rev Bras Microbiol 1993, 24:104–110

30. Boivin G, Coulombe Z, Rivest S: Intranasal herpes simplex virus type2 inoculation causes a profound thymidine kinase dependent cere-bral inflammatory response in the mouse hindbrain. Eur J Neurosci2002, 16:29–43

31. Schmidt NJ: Cell culture techniques for diagnostic virology. Diagnos-tic Procedures for Viral, Rickettsial, and Chlamydial Infections. Editedby EH Lennette, NJ Schmidt. Washington DC, American PublicHealth Association Inc, 1979, p 100

32. Giulietti A, Overbergh L, Valckx D, Decallonne B, Bouillon R, MathieuC: An overview of real-time quantitative PCR: applications to quantifycytokine gene expression. Methods 2001, 25:386–401

33. Broberg EK, Nygardas M, Salmi AA, Hukkanen V: Low copy numberdetection of herpes simplex virus type 1 mRNA and mouse Th1 typecytokine mRNAs by light cycler quantitative real time PCR. J VirolMethods 2003, 112:53–65

34. Kuhlicke J, Frick JS, Morote-Garcia JC, Rosenberger P, Eltzschig HK:Hypoxia inducible factor (HIF)-1 coordinates induction of toll-likereceptors TLR2 and TLR6 during hypoxia. PLoS One 2007, 2:e1364

35. Prehaud C, Megret F, Lafage M, Lafon M: Virus infection switchesTLR3 positive human neurons to become strong producers of betainterferon. J Virol 2005, 79:12893–12904

36. Sarangi PP, Kim B, Kurt-Jones E, Rouse BT: Innate recognition net-

work driving the HSV induced corneal immunopathology: role of Tollpathway in the early inflammatory events in stromal keratitis. J Virol2007, 81:11128–11138

37. Sancho-Shimizu V, Shen-Ying Z, Laurent A, Tardieu M, Rozemberg F,Jouanguy E, Casanova J: Genetic susceptibility to herpes simplexvirus 1 encephalitis in mice and humans. Curr Opin Allergy ClinImmunol 2007, 7:495–505

38. Kern ER: Use of viral infections in animal models to assess changesin the immune system. EHP 1982, 43:71–79

39. Lopez C: Genetics of natural resistance to herpesvirus infections inmice. Nature 1975, 258:152–153

40. Carr DJJ, Chodosh J, Ash J, Lane TE: Effect of anti-CXCL10 mono-clonal antibody on Herpes Simplex Virus Type 1 keratitis and retinalinfection. J Virol 2003, 77:10037–10046

41. Ellerman-Eriksen S: Macrophages and cytokines in the early defenseagainst herpes simplex virus. Virol J 2002, 2:59–89

42. He J, Ichimura H, Iida T, Minami M, Kobayashi K, Kita M, Sotozono C,Tagawa Y, Iwakura Y, Imanishi J: Kinetics of cytokine production inthe cornea and trigeminal ganglion of C57BL/6 mice after cornealHSV-1 infection. J Interferon Cytokine Res 1999, 19:609–615

43. Sun D, Ding A: MyD88-mediated stabilization of interferon-gamma-induced cytokine and chemokine mRNA. Nature Immunol 2006,7:375–381

44. Sergerie Y, Rivest S, Boivin G: Tumor necrosis factor-alpha andinterleukin-1beta play a critical role in the resistance against lethalherpes simplex virus encephalitis. J Infect Dis 2007, 196:853–860

45. Mason JL, Suzuki K, Chaplin DD, Matsushima GK: Interleukin 1-betapromotes repair of the CNS. J Neurosci 2001, 21:7046–7052

46. Carr DJJ, Campbell IL: Herpes simplex virus type 1 induction ofchemokine production is unrelated to viral load in the cornea but notin the nervous system. Viral Immunol 2006, 19:741–746

47. Sainz Jr., B, Halford WP: Alpha/Beta Interferon and gamma interferonsynergize to inhibit the replication of herpes simplex virus type 1.J Virol 2002, 76:11541–11550

48. Vollstedt S, Arnold S, Schwerdel C, Franchini M, Alber G, Di Santo JP,Ackermann M, Suter M: Interplay between alpha/beta and gammainterferons with B, T, and natural killer cells in the defense againstherpes simplex virus type 1. J Virol 2004, 78:3846–3850

49. Marques CP, Hu S, Sheng W, Lokensgard JR: Microglial cells initiatevigourous yet non-protective immune responses during HSV-1 braininfection. Virus Res 2006, 121:1–10

50. Rasmussen SB, Jensen SB, Nielsen C, Quartin E, Kato H, Chen ZJ,Silverman RH, Akira S, Paludan SR: Herpes simplex virus infection issensed by both Toll-like receptors and retinoic acid-inducible gene-like receptors, which synergize to induce type I interferon production.J Gen Virol 2009, 90:74–78

51. Takaoka A, Wang Z, Choi MK, Yanai H, Negishi H, Ban T, Lu Y,Miyagishi M, Kodama T, Honda K, Ohba Y, Taniguchi T: DAI (DLM-1/ZBP1) is a cytosolic DNA sensor and an activator of innate immuneresponse. Nature 2007, 448:501–505

52. Lundberg P, Ramakrishna C, Brown J, Tyska JM, Hamamura M,Hinton DR, Kovats S, Nalcioglu O, Weinberg K, Openshaw H, CantinEM: The immune response to herpes simples virus type 1 infection insusceptible mice is a major cause of central nervous system pathol-ogy resulting in fatal encephalitis. J Virol 2008, 82:7078–7088

53. Vaknin V, Blinder L, Wang L, Gazit R, Shapira E, Genina O, Pines M,Pikarsky E, Baniyash M: A common pathway mediated through Toll-like receptors leads to T- and natural killer–cell immunosuppression.Blood 2008, 111:1437–1447

54. Pasare C, Medzhitov R: Control of B-cell responses by Toll-like re-ceptors. Nature 2005, 438:364–368

55. Foster SL, Hargreaves DV, Medzhitov R: Gene-specific control ofinflammation by TLR-induced chromatin modifications. Nature 2007,447:972–978

56. McKimmie CS, Moore M, Fraser AR, Jamieson T, Xu D, Burt C, PitmanNI, Nibbs RJ, McInnes IB, Liew FY, Graham GJ: A TLR2 ligandsuppresses inflammation by modulation of chemokine receptors andredirection of leukocyte migration. Blood 2009, 113:4224–4231


Lethal Encephalitis in Myeloid Differentiation Factor 88-Deficient Mice Infected with Herpes Simplex...

Documents

Transcript of Lethal Encephalitis in Myeloid Differentiation Factor 88-Deficient Mice Infected with Herpes Simplex...