Toll-like receptor 4-mediated respiratory syncytial virus disease and lung transcriptomics in differentially susceptible inbred mouse strains

Abstract

Respiratory syncytial virus (RSV) causes severe lower respiratory tract disease in infants, young children, and susceptible adults. The pathogenesis of RSV disease is not fully understood, although toll-like receptor 4 (TLR4)-related innate immune response is known to play a role. The present study was designed to determine TLR4-mediated disease phenotypes and lung transcriptomics and to elucidate transcriptional mechanisms underlying differential RSV susceptibility in inbred strains of mice. Dominant negative Tlr4 mutant (C3H/HeJ, HeJ, Tlr4^Lps-d) and its wild-type (C3H/HeOuJ, OuJ, Tlr4^Lps-n) mice and five genetically diverse, differentially responsive strains bearing the wild-type Tlr4^Lps-n allele were infected with RSV. Bronchoalveolar lavage, histopathology, and genome-wide transcriptomics were used to characterize the pulmonary response to RSV. RSV-induced lung neutrophilia [1 day postinfection (pi)], epithelial proliferation (1 day pi), and lymphocytic infiltration (5 days pi) were significantly lower in HeJ compared with OuJ mice. Pulmonary RSV expression was also significantly suppressed in HeJ than in OuJ. Upregulation of immune/inflammatory (Cxcl3, Saa1) and heat shock protein (Hspa1a, Hsph1) genes was characteristic of OuJ mice, while cell cycle and cell death/survival genes were modulated in HeJ mice following RSV infection. Strain-specific transcriptomics suggested virus-responsive (Oasl1, Irg1, Mx1) and epidermal differentiation complex (Krt4, Lce3a) genes may contribute to TLR4-independent defense against RSV in resistant strains including C57BL/6J. The data indicate that TLR4 contributes to pulmonary RSV pathogenesis and activation of cellular immunity, the inflammasome complex, and vascular damage underlies it. Distinct transcriptomics in differentially responsive Tlr4-wild-type strains provide new insights into the mechanism of RSV disease and potential therapeutic targets.

INTRODUCTION

Respiratory syncytial virus (RSV) is ubiquitous and is the primary cause of hospitalization for lower respiratory tract illness in infants and immune-compromised adults, infecting over 80% of children by their second year of life. RSV is extremely virulent, and infected patients exhibit symptoms ranging from mild cold-like illness to potentially fatal severe bronchiolitis and pneumonia. RSV disease is one of the major causes of morbidity and mortality worldwide, particularly in children less than 1 yr of age (21). Multiple factors including virus strain (i.e., serotypes) and host-specific factors (e.g., prematurity, pre-existing disease, genetic polymorphisms) can influence the severity of RSV symptoms (38, 40). Pathologic changes due to RSV infection include sloughing of the respiratory epithelium, inflammatory cell infiltration, and the production of intrabronchiolar plugs of fibrin and mucus, resulting in partial to complete obstruction of the airways. RSV does not induce an effective immunological memory, and repeated infections are frequent (6). Due to the complex nature of RSV pathogenesis, no clinically approved efficacious therapy is currently available.

Immune response to RSV begins in the upper airways where it primarily infects epithelial cells (29). Released proinflammatory mediators recruit innate immunity effector cells including neutrophils, dendritic cells, macrophages, and natural killer cells, that cause massive leukocyte infiltration and type 1 T helper (Th1) and Th2 cytokine storms (12). While inadequate amplification of the adaptive immune response causes severe RSV disease, production of type I viral interferons (e.g., IFN-β, IFN-α) and type II immune IFN-γ contribute to antiviral signaling and viral clearance (4, 29). This response is mediated in part by pattern recognition receptors including toll-like receptors (TLRs), a set of membrane signal transducers that detect conserved molecular motifs from pathogens during the innate immune response (18).

TLRs recognize viral nucleic acids and signal through the nuclear factor kappa-light-chain-enhanced of activated B cells (NF-κB) pathway, recruiting inflammatory and antiviral cytokines. TLR2/6, TLR3, and TLR4 are activated by RSV in the airways, and TLR4 interacts with the RSV surface fusion (F) protein to initiate a downstream signaling cascade in a CD14- and lymphocyte antigen 96-dependent manner (20). Spontaneous Tlr4 mutations cause deficient lipopolysaccharide (LPS) responsiveness in two strains of mice, Tlr4^Lps-d (C3H/HeJ, with a nonsynonymous functional point mutation +2342C/A resulting in P712H), and Tlr4^Lps-del (C57BL/10ScNJ, with Tlr4 deleted) (27, 31, 32). In genetic Tlr4-knockout (Tlr4^−/−) mice, RSV caused delayed lung viral clearance and natural killer (NK) cell cytotoxicity with suppression of NK cells, CD14⁺ macrophages and cytokine release compared with wild-type mice (15, 18). In human populations, two loss-of-function single nucleotide polymorphisms (SNPs) have been found that encode substitutions (Asp299Gly and Thr399Ile) in the ectodomain of TLR4 (1). The role of these TLR4 SNPs in RSV disease remains uncertain, as these mutations can be either protective or detrimental to severe RSV disease in infants, depending on previous environmental exposure to LPS (5). For example, Awomoyi et al. (2) demonstrated that the TLR4 SNPs are highly associated with symptomatic RSV disease in preterm infants. Mononuclear cells isolated from children with TLR4 mutations, however, exhibited decreased NF-κB activation and cytokine production in response to RSV (39).

In the current study, we first determined potential molecular mechanisms of TLR4-mediated RSV disease by comparative lung phenotype analysis and genome-wide transcriptomics between Tlr4-mutant HeJ (Tlr4^Lps-d, +2342C/A) and Tlr4-normal (or -sufficient, Tlr4^Lps-n) C3H/HeOuJ (OuJ) mice. We then used genome-wide transcriptomics for five inbred strains bearing the Tlr4-normal allele (Tlr4^Lps-n) to determine downstream transcriptional mechanisms underlying their differential RSV susceptibility.

MATERIALS AND METHODS

Animals and RSV Infection

Age matched (5–8 wk), male OuJ, HeJ, BALB/cJ (BALBc), C57BL/6J (B6), A/J (AJ), AKR/J (AKR), and DBA/2J (D2) mice were purchased from Jackson Laboratories (Bar Harbor, ME) and acclimated 1 wk before infection. All experimental protocols were in accordance with National Institutes of Health guidelines, and all animal use was approved by the National Institute of Environmental Health Sciences (NIEHS) Animal Care and Use Committee. Mice were intranasally instilled with 10⁶ plaque forming units of human RSV L19 (Virasource, Durham, NC) or vehicle (HBSS). Mice were euthanized 1 and 5 days post-RSV [or postinfection (pi)] or 1 and 5 days after vehicle treatment by sodium pentobarbital overdose (100 mg/kg).

Lung Injury Assessment

Bronchoalveolar lavage (BAL) fluid was collected from right lungs of HeJ and OuJ mice to assess total protein and inflammatory cell numbers in the lung (8). Left lungs were inflated and zinc formalin-fixed, and paraffin-embedded lung sections were stained with hematoxylin and eosin (H&E) for histological evaluation (EPL Inc., Research Triangle Park, NC), and with an antiheat shock protein 70 (HSP70) antibody (Santa Cruz Biotechnology, Santa Cruz, CA) for immunohistochemistry. Lung tissues were snap-frozen for protein and RNA isolation.

Lung RNA Isolation and cDNA Microarray

RNA was isolated from lung homogenates (n = 3/group; vehicle 1 day, RSV 1 day, RSV 5 day) with the RNeasy Mini-Plus kit (Qiagen, Valencia, CA). The Affymetrix Mouse Genome 430_2.0 microarray (Santa Clara, CA) was run by the NIEHS Molecular Genomics Core. Baseline strain differences were analyzed by moderated t test. Effects of TLR4 on RSV infection were analyzed by two-way ANOVA with multiple testing correction (factors: strain, treatment). Tlr4-normal strain data were analyzed by one-way ANOVA. All microarray data were analyzed with GeneSpring GX software (Agilent, Santa Clara, CA), and significantly altered genes (P < 0.01) were examined with Ingenuity Pathway Analysis software (Qiagen, Redwood City, CA). Microarray data are deposited in Gene Expression Omnibus [GSE111861 (HeJ, OuJ) and GSE112956 (AJ, AKR, B6, D2, and BALBc)].

RT-PCR

We reverse-transcribed 1 μg of total lung RNA into cDNA, and an aliquot was amplified with gene-specific primer sets by quantitative PCR using the StepOnePlus (Applied Biosystems, Foster City, CA) or by semiquantitative PCR on the GeneAmp PCR System 9700 (Applied Biosystems) following published procedures (8).

ELISA

Aliquots of BAL (50 μl) or serum (100 μl) from vehicle (1 day)- or RSV (1 and 5 day)-treated mice were assayed for protein levels of interleukin (IL)-6, C-X-C motif chemokine ligand 2 (CXCL2), and IL-1β using mouse-specific ELISA kits (R&D Systems, Minneapolis, MN).

Western Blot Analysis

Total lung proteins from vehicle (1 day)- or RSV (1 and 5 day)-treated mice were isolated from lung homogenates in radioimmunoprecipitation assay buffer (8). Nuclear proteins were extracted using a kit (Active Motif, Carlsbad, CA). Protein aliquots were denatured, fractionated on 10–20% Tris·HCl SDS-PAGE gels (Bio-Rad, Hercules, CA), and blotted with antibodies (Santa Cruz Biotechnology) for HSP70, IL-1β, IL-18, caspase 1 (CASP1) and nuclear NF-κB p65 using actin and laminin B as loading controls for total and nuclear proteins, respectively. Representative images from multiple blots were scanned using the FluorChem HD2 System (ProteinSimple, San Jose, CA).

Statistical Analysis

BAL, RT-PCR, and ELISA data were analyzed by two-way ANOVA (factors: strain, treatment) and comparisons of means followed by Student-Newman-Keuls a posteriori test. Lung injuries assessed by BAL in 5-day vehicle control mice did not differ from those in 1-day vehicle controls (Student’s t test) in each strain and pooled vehicle control data were used in BAL analyses. Analyses were conducted using SigmaStat (Jandel Scientific Software, San Rafael, CA) with statistical significance accepted at P < 0.05.

Supplemental Data

Supplemental tables and figure have been deposited to a generalist public access repository (https://doi.org/10.5281/zenodo.3532142).

RESULTS

TLR4-dependent Lung Injury by RSV in OuJ and HeJ Mice

Blunted lung cell injury and viral load in Tlr4-mutant HeJ mice.

Mean numbers of total BAL cells were significantly increased 1 day pi in both strains and remained elevated 5 days pi in OuJ mice (Fig. 1A). Neutrophils accounted for much of the increase in total cells at 1 day, and lymphocytes were the predominant cell type at 5 days pi in OuJ mice (Fig. 1A). The RSV-induced increases in total lung cells, neutrophils, and lymphocytes were significantly greater in OuJ than in HeJ mice (Fig. 1A). Protein levels in BAL were not significantly changed by RSV infection in either mouse strain (data not shown). Pulmonary viral load measured by the expression of RSV surface attachment protein (G) was significantly greater in OuJ mice relative to HeJ mice (Fig. 1B).

Fig. 1.

Toll like receptor 4 (TLR4)-dependent development of respiratory syncytial virus (RSV) disease. A: bronchoalveolar lavage (BAL) analysis identified differential cellular injury and inflammation in RSV-infected Tlr4-normal C3H/HeOuJ (OuJ, Tlr4^Lps-n) and Tlr4-mutant C3H/HeJ (HeJ, Tlr4^Lps-d) mice. Data are presented as means ± SE (n = 3–5/group). *Significantly different from strain-matched pooled (1- and 5-day) vehicle control (P < 0.05). †Significantly different from exposure-matched OuJ mice (P < 0.05). B: RSV G gene was amplified by quantitative RT-PCR to determine the differential viral load and amplification in the lung. Data are normalized to β-actin mRNA and presented as means ± SE (n = 3/group). *Significantly different from strain-matched vehicle controls (P < 0.05). †Significantly different from treatment-matched OuJ mice (P < 0.05). C: representative micrographic images of hematoxylin and eosin-stained lung cross sections from OuJ and HeJ mice at 1 day post-RSV or -vehicle. Arrows indicate inflammatory cell infiltration. Arrow heads indicate epithelial proliferation and hyperplasia. Bars = 100 μm. AV, alveoli; BR, bronchiole; BV, blood vessel.

Mild lung histopathologic changes in Tlr4-mutant HeJ mice.

No pulmonary histopathologic changes were noted in vehicle-treated HeJ or OuJ mice (Fig. 1C). In agreement with BAL phenotypes, there were severe bronchial epithelial cell exfoliation and inflammatory cell accumulation in the alveolar air space and perivascular-peribronchial regions in OuJ mice 1 day pi (Fig. 1C). Relative to OuJ mice, mild and sporadic pulmonary changes were observed in HeJ mice (Fig. 1C). Quantitative analysis of H&E-stained lung sections indicated that total lung inflammation score (ranging from 0 for no inflammation to 10) was twofold higher in OuJ (means ± SE; 2.33 ± 1.20, n = 3) than in HeJ (1.00 ± 0.58, n = 3) mice 1 day pi. Inflammatory cell infiltration was greatly diminished 5 days pi in either strain with persisting epithelial cell hyperplasia and alveolar vacuolization (emphysema) evident in OuJ mice (data not shown).

TLR4-dependent Lung Transcriptomics in OuJ and HeJ Mice

RSV infection.

RSV caused relatively greater transcriptome changes in OuJ (one-way ANOVA, P < 0.05, 1,435 genes; Supplemental Table S1) than in HeJ mice (one-way ANOVA, P < 0.05, 333 genes, Supplemental Table S2), which peaked 1 day pi (Fig. 2A). Approximately 95% of RSV-responsive genes in OuJ mice were altered 1 day pi (Supplemental Table S1). Overexpression of transcripts encoding HSP70 (Hspa1a, Hspa1b), chemokines (e.g., Cxcl3, Cxcl2, Ccl7, Ccl2), acute inflammatory response genes (e.g., Saa1, Timp1, Il6), and tudor domain containing 5 (Tdrd5) was unique to OuJ mice 1 day pi (Table 1, Supplemental Table S3). In RSV-altered genes specific to HeJ mice (244 out of 333 genes by Venn diagram analysis shown in Fig. 2A), cell cycle (e.g., Ccnl1, Plk2), DNA binding protein (e.g., Nr4a1, Nr4a2, Junb, Fosb, Nfil3), and dual-specificity phosphatase (Dusp1, Dusp2) genes were distinct (Supplemental Table S2). These transcripts were downregulated at 1 day pi but increased over baseline 5 days pi (Fig. 2B, left). Another subset of genes encoding gastrokines (Gkn1, Gkn2), trefoil factors (Tff1, Tff2), proteases (Ctse, Capn5), small proline-rich protein 2A1 (Sprr2a1), keratin (Krt42), and Ly6/PLAUR domain containing 8 (Lypd8) was markedly (7- to 47-fold) upregulated 1 day pi only in HeJ mice (Fig. 2B, right), and they may play roles in RSV resistance by regulating cellular growth, mucosal protection, epidermal development, and anti-inflammatory processes (24, 28).

Fig. 2.

Toll-like receptor 4 (TLR4)-dependent lung transcriptome changes induced by respiratory syncytial virus (RSV). A: heat map depicts differential gene expression profiles of RSV response (n = 855, two-way ANOVA with P < 0.01) between Tlr4-normal C3H/HeOuJ (OuJ, Tlr4^Lps-n) and Tlr4-mutant C3H/HeJ (HeJ, Tlr4^Lps-d) mice. Venn diagram analysis shows number of strain-specific RSV-responsive genes determined by one-way ANOVA for each strain. Color bar indicates average expression intensity (n = 3/group) normalized to OuJ-Vehicle. Graph colored by expression levels of OuJ at 1 day post-RSV (vs. OuJ-Vehicle). B: a distinct profile of HeJ-specific RSV-responsive genes includes genes encoding cell cycle and DNA binding proteins (e.g., Ccnl1, Plk2, Nr4a1, Nr4a2, Junb, Fosb, Nfil3) downregulated at 1 day post-RSV but increased over baseline 5 days post-RSV (gene profiles in left graph). Another set of genes involved in cellular growth, mucosal protection, epidermal development, and anti-inflammatory processes (e.g., gastrokines, trefoil factors, keratins) was markedly increased at 1 day post-RSV only in HeJ mice (gene profiles in right graph). Graph colored by expression levels of HeJ at 1 day post-RSV (vs. OuJ-vehicle). C: pathway analysis identified MYD88 and TLR4 as key upstream regulators of the suppressed immune response in HeJ mice infected with RSV. Tlr4 deficiency may inhibit immune cell recruitment and T cell activation and potentiate B cell proliferation against RSV in HeJ mice. Color intensity of the genes (green or red) indicates fold difference in HeJ relative to OuJ at 1 day post-RSV.

Table 1.

Profiles of representative Tlr4-dependent genes differentially expressed between OuJ and HeJ mice following RSV infection

		Fold Difference
Gene Symbol	Gene Title	Vehicle	RSV 1 day	RSV 5 days	P
Cell differentiation/development/stress response
Hspa1a	heat shock protein 1A	1.95	−19.46	−1.56	0.0000889994
Hspa1b	heat shock protein 1B	3.45	−15.52	−1.91	0.0009941262
Hsph1	heat shock 105 kDa/110kDa protein 1	1.45	−3.89	−1.71	0.0007323779
Dnajb1	DnaJ (Hsp40) homolog, subfamily B, member 1	1.04	−3.33	−1.10	0.0000002735
Nr4a1	nuclear receptor subfamily 4, group A, member 1	3.11	−1.22	4.90	0.0001070260
Csrnp1	cysteine-serine-rich nuclear protein 1	2.33	−1.69	3.04	0.0000094890
Fos	FBJ osteosarcoma oncogene	3.02	1.41	3.99	0.0098647230
Tp53	transformation related protein 53	−2.01	−2.16	−1.34	0.0025209690
Atf3	activating transcription factor 3	1.47	−1.84	2.74	0.0000073226
Slfn3,Slfn4	schlafen 3, schlafen 4	1.10	−2.08	1.89	0.0031019310
Ahsa2	AHA1, activator of heat shock protein ATPase 2	1.02	−2.40	−1.32	0.0003335664
Zfp36	zinc finger protein 36	3.18	−1.06	3.72	0.0000591925
Immune response
Cxcl3	chemokine (C-X-C) motif, ligand 3	−1.41	−19.68	−1.53	0.0082591700
Ccl3	chemokine (C-C) motif, ligand 3	−1.27	−5.73	−1.04	0.0004370719
Il1rn	interleukin 1 receptor antagonist	−1.94	−4.34	−1.44	0.0007002755
Il1b	interleukin 1 beta	−1.33	−3.00	1.20	0.0000094890
C2	complement component 2	−1.15	−2.86	−1.14	0.0031499850
Il18bp	interleukin 18 binding protein	−1.44	−2.43	−1.15	0.0022781708
Saa1	serum amyloid A1	−1.15	−14.41	−1.31	0.0000085549
Il6	interleukin 6	4.25	−9.80	4.74	0.0000094890
Cxcl2	chemokine (C-X-C) motif, ligand 2	1.93	−8.59	2.94	0.0003612355
Msr1	macrophage scavenger receptor 1	−1.27	−1.87	−1.13	0.0018328646
Orm2	orosomucoid 2	−1.20	−6.79	−1.14	0.0000102062
Timp1	tissue inhibitor of metalloproteinase 1	−1.14	−5.87	−1.40	0.0002498063
Clec4e	C-lectin domain 4e	−1.52	−5.30	−1.06	0.0015814515
Serpina3n	serine (or cysteine) peptidase inhibitor, clade A, member 3N	1.02	−2.70	−1.12	0.0002498063
Nfil3	nuclear factor, interleukin 3, regulated	2.73	−1.66	2.56	0.0003307226
Metabolism and transport
Ch25h	cholesterol 25-hydroxylase	−1.47	−2.38	−1.75	0.0092866210
Apold1	apolipoprotein L domain containing 1	2.54	−1.80	4.00	0.0000567671
Lcn2	lipocalin 2	−2.17	−2.76	−1.40	0.0043132780
Kcne4	potassium voltage-gated channel, Isk-related subfamily, gene 4	3.47	−1.22	3.20	0.0000412757
Cyp7b1	cytochrome P450, family 7, subfamily b, polypeptide 1	−1.28	−2.58	−1.24	0.0001415120
Noxo1	NADPH oxidase organizer 1	−1.14	−2.00	−1.02	0.0042059068
Ptgs2	prostaglandin-endoperoxide synthase 2	2.19	−1.19	2.02	0.0007379816

Full gene list in Supplemental Table S3. OuJ, C3H/HeOuJ; HeJ, C3H/HeJ; RSV, respiratory syncytial virus. n = 855, 2-way ANOVA P < 0.01. Negative (−) fold difference indicates lower in HeJ than in OuJ mice. Positive fold difference indicates higher in HeJ than in OuJ mice.

As anticipated, RSV-mediated Tlr4-dependent genes (two-way ANOVA for OuJ and HeJ, n = 855, P < 0.01; Supplemental Table S3) regulated by myeloid differentiation primary response gene 88 (MYD88) and other TLR signal transduction molecules (e.g., IL-1β, TNF, IL-4, IFN-γ) were predicted to exert immune cell trafficking and inflammatory responses in OuJ mice (Fig. 2C, top and middle). Lowered inflammatory cell recruitment, suppressed T lymphocyte activation, and reduced B cell proliferation was also predicted 1 day pi in HeJ (Fig. 2C, bottom). Specific genes associated with inflammatory cell trafficking and immune response (e.g., Cxcl1, Cxcl2, Il1b) were elevated in HeJ mice compared with OuJ (Table 1, Supplemental Table S3), suggesting a delayed immune response concomitant with delayed RSV replication 5 days pi (see Fig. 1B).

Vehicle controls.

Lung injury in 5-day vehicle controls did not differ from 1-day controls. For clarity only 1-day vehicle controls were included in further analyses. A total of 2,824 lung gene transcripts were differentially expressed between OuJ and HeJ vehicle-treated mice (moderated t test, P < 0.01, ≥2-fold genes in Supplemental Table S4). Among these, Tlr4-mutant HeJ lungs showed marked suppression of immune response and immunity genes (e.g., Ccl9, Tlr2, Mmp12, immunoglobulin heavy chains) and upstream regulators resistin-like beta (Retnlb) and histidine-rich glycoprotein (Hrg), which suggests their importance in the RSV-mediated inflammatory cascade (Fig. 3A, Supplemental Table S1). Many of these genes are postulated to be regulated directly through TLR4. Genes that encode transport (e.g., Copz2, Fabp1) and Tp53 were also suppressed in HeJ mice (Supplemental Table S4). In contrast, transcripts associated with vasculogenesis and organogenesis (e.g., Nr4a1, Egr1) were constitutively higher in HeJ mice (Fig. 3B), probably through activation of upstream kinases (e.g., mitogen-activated protein kinase 4, Map2k4) and growth factors (e.g., insulin-like growth factor 1 receptor, Igf1r). Solute carrier family 12, member 4 (Slc12a4), neurexin I (Nrxn1), and Il6 genes were higher in HeJ lungs than in OuJ lungs (Supplemental Table S4). p53 was suggested to be a central player of TLR4-mediated cell death and survival, cellular maintenance, and development in baseline gene networks (Fig. 3C) consistent with the previous observation (23).

Fig. 3.

Pathway analysis of toll like receptor 4 (TLR4)-dependent basal lung transcriptomics. Analysis of upstream and downstream regulator effects on Tlr4-dependent basal lung genes (moderated t test, P < 0.01) indicates a role for reduced immune response (A) and increased vasculogenesis and tissue development (B) networks in Tlr4-mutant C3H/HeJ (HeJ, Tlr4^Lps-d) compared with Tlr4-normal C3H/HeOuJ (OuJ, Tlr4^Lps-n) mice. TLR4-dependent genes were predicted to inhibit p53 (Tp53), which may play an essential role in cell death and survival, cellular maintenance, and development in HeJ lungs (C, Network score 49). Color intensity of the genes (green or red) indicates fold difference in HeJ relative to OuJ.

Validation of microarray analysis data.

RT-PCR, Western blot, and ELISA were used to verify TLR4-dependent differential expression of HSP70 (Hspa1b), inflammatory mediators (Cxcl2, Cxcl10, Saa1, Il1rn, Il6, Il1b), pattern recognition receptors (Clec4e), inflammasome mediators (Casp1, Nlrp3), and NF-κB signaling at the transcript or protein level. Significantly lower message levels of CASP1 and NLR Family Pyrin Domain Containing 3 (NLRP3) were accompanied by activated precursor IL-1β (Western blot) and lower mature IL-1β (ELISA) in HeJ compared with OuJ mice (Fig. 4, B and C). IL-18 transcripts were significantly increased in OuJ lungs after viral infection, with precursor (pre) and mature (mat) protein levels suppressed in HeJ mice after infection (Fig. 4, B and C). Overall, inflammasome-mediated proteolytic cleavage/maturation of IL-1β and IL-18 was reduced in HeJ mice compared with OuJ mice after RSV infection. RSV infection increased nuclear localization of NF-κB (detected as p65 subunit) only in OuJ mice (Fig. 4, A and C). Consistent with microarray analyses (Fig. 2C), HSP70 proteins were minimally (HeJ) or highly (OuJ) detected in alveolar airways (mostly macrophages) of vehicle-treated lungs (Fig. 4D). RSV infection increased the number of HSP70-positive cells in OuJ, but not in HeJ mice (Fig. 4D, B6 shown as an intermittent positive control). The HSP70-positive cells in RSV-infected OuJ lungs were mostly infiltrating leukocytes localized in severely injured alveoli or in blood vessels (Fig. 4D).

Fig. 4.

Microarray validation of toll like receptor 4 (TLR4)-dependent transcriptomics in respiratory syncytial virus (RSV) disease. Transcript levels of immune mediators (A) and inflammasome activators (B) from C3H/HeOuJ (OuJ) and C3H/HeJ (HeJ) mice at 1 and 5 days post-RSV determined by quantitative RT-PCR. β-Actin mRNA or 18s rRNA-normalized group means ± SE presented (n = 3/group). *Significantly different from strain-matched vehicle controls (P < 0.05). †Significantly different from treatment-matched OuJ mice (P < 0.05). C: TLR4-dependent lung protein levels determined by Western blot (internal controls: pan-actin for cytoplasmic protein, lamin B for nuclear NF-κB p65 subunit). Pre, precursor; mat, matured. Representative digitized images from duplicate assays are presented. Selected cytokine levels from bronchoalveolar lavage (BAL) or sera determined by ELISA. Group means ± SE (n = 3/group) presented. *Significantly different from strain-matched vehicle controls (P < 0.05). †Significantly different from treatment-matched OuJ mice (P < 0.05). D: representative lung micrographs (n = 2–4/group) display immunohistochemically stained heat shock protein 70 (HSP70) in alveoli of OuJ, HeJ, and C57BL/6J (B6) mice 1 day post-RSV (large images). Insets are lung sections from vehicle control (1 day) of each strain. Brown dots marked with arrows indicate HSP70 proteins. Bars = 100 μm.

Transcriptomics of Trl4-Normal Inbred Strains

RSV infection: strains overview.

We previously characterized experimental RSV disease in a panel of differentially responsive strains of inbred mice (17). Statistically significant inter-strain variation was found for each of the RSV disease phenotypes (selected strain data in Supplemental Figure S1). To further examine the mechanisms of differential RSV susceptibility with normal Tlr4 signaling, we compared RSV-induced lung transcriptome changes in five inbred strains bearing the wild-type Tlr4^Lps-n allele, B6, BALBc, AJ, AKR, D2, and OuJ mice (1-way ANOVA for each strain, P < 0.01). Compared with strain-matched vehicle controls, RSV infection caused more marked upregulation of inflammatory transcripts in the susceptible BALBc, AJ, and OuJ strains relative to the resistant D2, HeJ, and AKR strains (Table 2, Fig. 5A). The exception was the resistant B6 mice, which showed a marked induction of inflammatory genes (Table 2, Fig. 5A). Common pathways of genes altered by RSV in most strains included granulocyte and agranulocyte adhesion (e.g., Cxcl family), IL-10/IL-6/IL-17α/IL-12 signaling, dendritic cell maturation (e.g., Il6, histocompatibility 2 family), acute phase response signaling (e.g., Saa3), altered T cell and B cell signaling (Cxcl13, Csf1, Fcer1g), and TLR signaling and pattern recognition receptor signaling (Supplemental Table S5). In addition, channel genes (e.g., Scn3a) and proteasome subunits (e.g., Psmb8), serine (or cysteine) peptidase inhibitor family (Serpinb3b), and ubiquitin-related (e.g., Usp18) genes were altered, and transcription factor genes for signal transducer and activator of transcription (Stat1, Stat2) were markedly induced by RSV in many strains. Expression of 11 Tlr family genes was detected, with Tlr1 and Tlr2 transcripts highly induced by RSV, particularly in B6, BALBc, AJ, and AKR mice (Fig. 5B). Tlr genes in D2, HeJ, and OuJ mice were only marginally altered by RSV, while Tlr4 and/or Tlr6 transcripts were suppressed by RSV in HeJ mice (Fig. 5B). All Tlr transcript levels were relatively lower in HeJ mice than in OuJ mice basally and after RSV infection (Fig. 5B).

Table 2.

Inbred strain comparison of genes significantly altered by RSV infection

Symbol	Title	C57BL/6J	BALB/cJ	C3H/HeOuJ	A/J	AKR/J	DBA/2J
Cxcl3	chemokine (C-X-C motif) ligand 3	108.6/3.8	221.2/20.0	47.5/1.8	80.5/1.3	107.6/1.1
Saa3	serum amyloid A 3	99.7/25.9	146.5/26.9	30.1/2.4	23.2/1.2	84.2/3.4
Cxcl10	chemokine (C-X-C motif) ligand 10	155.6/7.5	216.7/34.1	7.4/1.8	107.1/3.7
Irg1	immunoresponsive gene 1	157.2/3.4	133.4/6.0	4.2/1.0	60.6/1.4
Slfn4	schlafen 4	57.1/2.3	100.0/6.0	1.6/-1.4	27.7/-1.4
Timp1	tissue inhibitor of metalloproteinase 1	31.4/2.2	52.7/3.6	6.5/1.1	16.4/1.7
Il1b	interleukin 1 beta	24.5/1.3	28.7/4.2	1.4/-1.5	9.9/-1.7
Serpina3n	serine (or cysteine) peptidase inhibitor, clade A, member 3N	10.7/-1.3	34.1/1.8	3.2/1.0	4.6/1.3
Ccl2	chemokine (C-C motif) ligand 2	26.7/2.5	71.2/4.5	4.4/1.1	24.2/1.7		1.8/1.2
Slc26a4	solute carrier family 26, member 4	11.1/4.8	20.7/9.4	2.3/-1.1			5.5/12.9
Chil1	chitinase-like 1	2.0/1.3	2.4/1.5	2.0/1.0			6.5/4.6
Clec4e	C-type lectin domain family 4, member e	34.6/1.5	77.5/3.9	3.0/-1.1
Il6	interleukin 6	41.1/1.2	71.0/2.0	8.0/-1.1
Orm1	orosomucoid 1	34.9/1.2	51.6/-2.2	10.1/1.4
Msr1	macrophage scavenger receptor 1	11.3/2.5	13.8/3.1	1.7/1.1
Tlr2	toll-like receptor 2	6.5/1.7	12.2/3.0	1.3/1.0
H2	histocompatibility 2	D1,T10 (3.4/<1.8)	D1,Q7,T10 (<6.3/<3.8)	Ob (−2.0/-1.1)
Mx1	myxovirus (influenza virus) resistance 1	38.6/1.6	70.0/3.3		30.8/1.6
Usp18	ubiquitin-specific peptidase 18	18.5/1.2	25.9/2.2		13.5/1.1.
Ms4a4c	membrane-spanning 4-domains, subfamily A, member 4C	33.1/4.5	45.6/3.5
Zbp1	Z-DNA binding protein 1	23.0/2.5	34.5/6.8
Scn3a	sodium channel, voltage-gated, type III, alpha	−24.2/-1.8	−15.7/-1.6
Cyp2a4	cytochrome P450, family 2, subfamily a, polypeptide 4	−7.9/-5.2	−21.0/-42.3
Krt4	keratin 4	152/233.5
Lce3a	late cornified envelope 3A	164.4/202.9
Ubd	ubiquitin D	34.7/3.2
Fos	FBJ osteosarcoma oncogene		3.2/1.3	−1.6/1.0		−6.8/-1.8
Nppa	natriuretic peptide type A		−26.5/-65.6	−1.4/36.4
Ifi202b	interferon activated gene 202B		65.2/11.5		23.3/3.3
Pyhin1	pyrin and HIN domain family, member 1		36.6/3.1
Adipoq	adiponectin, C1Q and collagen domain containing		−20.6/-24.6
Cdc20	cell division cycle 20		−1.6/2.0				1.2/4.1
Hspa1b	heat shock protein 1B			20.3/1.9
Nr4a1	nuclear receptor subfamily 4, group A, member 1				−4.0/-4.7	−9.3/-1.6

Selected significantly changed genes by RSV in each strain (one-way ANOVA, P < 0.01). Fold changes over the corresponding vehicle control (1 day) are shown as 1-day post-RSV/5-day post-RSV (negative values indicate decrease). Blank indicates no significant change. Details of gene list and pathway analysis reports in Supplemental Tables S1 (C3H/HeOuJ), S5 (Pathway analyses), S6 (BALB/cJ), and S7 (C57BL/6J).

Fig. 5.

Lung transcriptomes changed by respiratory syncytial virus (RSV) in differentially susceptible toll like receptor 4 (Tlr4)-normal inbred strains. A: heat maps depict gene expression profiles significantly changed by RSV in Tlr4-normal (Tlr4^Lps-n) inbred strains (one-way ANOVA with P < 0.01 for each strain). Marked transcriptome changes in C57BL/6J (B6) and BALBc/J (BALBc) and minimal alteration in DBA/2J (D2) strains are contrasted. Tlr4-mutated C3H/HeJ (HeJ) expression profile is shown for comparison. Color indicates average expression intensity (n = 3/group) normalized to strain-specific vehicle group (V, in yellow). Upregulation by RSV in red, downregulation by RSV in blue. Graph color by expression level of each gene at 1 day (1^d) post-RSV. B: TLR family gene expression profiles in inbred strains. Group means from microarray data normalized to OuJ Vehicle (n = 3/group). C: profiles of C57BL/6J (B6)-specific genes (i.e., remained elevated by 5 days post-RSV in B6) identified from Venn diagram overlap of RSV-responsive B6, BALBc and OuJ mice transcriptomes were compared between inbred strains. Graph color relative to 5-day (5^d) post-RSV expression in B6 mice. Pathway analysis indicated a role of these genes in suppression of viral replication in B6 mice. D: heat maps show expression profiles of RSV-upregulated A/J (AJ) mouse genes in inbred strains. Graph color relative to 1-day post-RSV expression in AJ mice. E: heat maps show inbred expression profiles of RSV-downregulated AKR/J (AKR) genes in inbred strains. Graph color relative to 1-day post-RSV expression in AKR mice. F: while minimal changes were observed in the transcriptome profile of DBA/2J (D2) mice, RSV-altered genes in these mice were mostly associated with suppressed cell cycle and mitosis pathways. RSV-upregulated genes in red, -downregulated genes in green.

BALBc.

Immune and inflammatory transcripts were significantly elevated in BALBc mice by RSV (n = 2,886; Supplemental Table S6). The BALBc-specific RSV response gene expression peaked 1 day pi (Fig. 5A), and these genes were also increased in AJ, AKR, B6, and OuJ mice (Table 2, Fig. 6). In contrast, genes associated with cardiac dysfunction and glycogen degradation (e.g., Pln, Ryr2, Csrp3, Actc1, Pygm) were downregulated exclusively in BALBc mice (Supplemental Table S6).

Fig. 6.

Respiratory syncytial virus (RSV)-altered lung gene expression profiles in inbred strains of mice and pathway analysis. *One-way ANOVA (vehicle, 1-day post-RSV; pi, 5-day post-RSV). †Expression profiles of the significantly changed genes in the strain of interest (boxed) and in the others. Gene expression levels were normalized to those of vehicle in each strain. Order of strains from left to right are A/J, AKR/J, C57BL/6J, BALB/cJ, DBA/2J, C3H/HeJ (HeJ), and C3H/HeOuJ (OuJ). Each profile shows in order of vehicle, 1-day post-RSV, and 5-day post-RSV. Color shown as 1-day post-RSV of the strain (blue, decreased vs. vehicle; red, increased vs. vehicle). §Details of gene list and pathway analysis reports in Supplemental Tables S1 and S5-S7.↑Genes increased over the vehicle control level. ↓Genes decreased relative to the vehicle control level. Analyzed by GeneSpring software and Ingenuity Pathway Analysis.

B6.

The largest RSV-induced transcriptome changes among examined strains were found in RSV-resistant B6 mice (n = 4,821; Supplemental Table S7). A distinct feature of their transcriptome profile was sustained upregulation of numerous genes 5 days pi (gene profiles in red in Figs. 5C and 6, Table 2). Interestingly, some of these genes were elevated 5 days pi in RSV-resistant strains D2, AKR, and HeJ, but not in BALBc or OuJ mice (Fig. 5C), suggesting their protective roles against RSV infection. These RSV-induced B6 signature genes (up to 234-fold increased by 5 days) are keratins (e.g., Krt4, Krt13) and many epidermal differentiation complex (EDC) genes including the late cornified envelope family [e.g., Lce3a, Lce1d), loricrin (Lor), repetin (Rptn), cornifelin (Cnfn), hornerin (Hrnr)], S100 calcium binding proteins (S100a14), and Sprr2a1. They encode proteins known to play critical roles in epidermal development, keratinization, and cellular morphogenesis. In addition, host defense genes such as myxovirus resistance genes (Mx1, Mx2), defensin beta 4 (Defb4), macrophage scavenger receptor 1 (Msr1), and interferon signaling genes (Ifi44l, Ifi44, Irf7) were uniquely upregulated in B6 mice, but not in susceptible BALBc mice. Concurrently, a key predicated function of RSV-responsive B6 genes was to inhibit viral replication 5 days pi (Fig. 5C).

other strains.

Interestingly, RSV-altered transcripts in AJ mice (n = 307) were mostly upregulated (gene profiles in red in Figs. 5D and 6, Table 2). In contrast, 85% of the genes significantly changed by RSV in AKR mice (n = 214) were downregulated (gene profiles in blue in Figs. 5E and 6, Table 2). These genes were also reduced in RSV-resistant HeJ mice (Fig. 5E). While the lung transcriptome of the resistant D2 mice marginally responded to RSV infection (n = 56, see Fig. 5A), RSV-modulated D2 genes are likely to play roles in cell cycle inhibition and DNA damage checkpoint regulation (Figs. 5F and 6).

Vehicle control: strains overview.

Basal lung transcriptome profiles were compared by normalizing expression values to those of OuJ mice. There were significant basal gene expression differences between AJ, AKR, B6, BALBc, and OuJ lungs, and different sets of genes varied constitutively between inbred mouse lungs (Supplemental Table S8, one-way ANOVA, fold differences compared with OuJ). For example, neurexophilin and PC-esterase domain family member 4 (Nxpe4), microRNA 1931 (Mir1931), angiogenin, ribonuclease RNase A, family 5 (Ang), zinc finger protein 874a (Zfp874a), and the tripartite motif family (Trim5, Trim21) were constitutively overexpressed in B6 lungs. In BALBc mice, a wide variety of immune mediators and mucosal immunity genes activated by viral infection (e.g., chemokines/cytokines, immune cell surface receptors, Fc receptors, histocompatibility complex 2, immunoglobulin kappa chain, lymphocyte antigen, TLR family) were basally suppressed compared with OuJ and other strains. In contrast, constitutively higher levels of natriuretic peptide precursor A (Nppa), mucin 5B (Muc5b), and lactotransferrin (Ltf) transcripts were suppressed by RSV in BALBc mice. Many genes, including BPI fold containing family A, member 1 (Bpifa1), carbonic anhydrase 3 (Car3), secretoglobin, family 3A, member 1 (Scgb3a1), cytochrome P450, family 2, subfamily a, polypeptide 4 (Cyp2a4), protocaherin beta 9 (Pcdhb9), and Cd200r4, were basally suppressed in AKR mice compared with other strains (Supplemental Table S8). However, melanoma antigen (Mela) and cathepsin E (Ctse) were constitutively higher in AKR as well as in D2 mice compared with other strains (Supplemental Table S8). In general, lower expression of Nppa, H2-D1, Serpina1b, solute carrier family 12, member 4 (Slc12a4), and elevated Krt79, Clec2d, kallikrein 1-related peptidase b22 (Klk1b22), and tumor necrosis factor receptor superfamily, member 14 (Tnfrsf14) expression was found in OuJ mice compared with other strains (Supplemental Table S8).

DISCUSSION

The current study demonstrated TLR4 signaling deficiency significantly reduced experimental RSV disease during the acute phase of infection. RSV replication and pulmonary inflammatory responses were significantly blunted in spontaneous Tlr4-mutant (Tlr4^Lps-d) HeJ mice compared with the corresponding Tlr4-normal (Tlr4^Lps-n) OuJ mice. We also identified Tlr4-dependent transcriptome changes in the lung following infection and elucidated potential molecular events modulated by TLR4 during RSV disease progression. Furthermore, comparative transcriptome analyses of Tlr4-normal inbred mouse strains with differential RSV susceptibility provide strain-specific gene expression profiles during RSV infection.

As predicted, TLR4 and TLR4-related signal transducer molecules, including IL-1β, MYD88, and NF-κB, were identified as key upstream regulators of the differentially expressed genes following RSV infection. Consistent with our findings, reduced levels of NF-κB in the lung and isolated macrophages were found in HeJ mice compared with the wild-type C3H/HeSnJ strain during RSV pathogenesis (14). Reduced NF-κB activity could explain the negligible induction of target chemo-attractants for monocytes, macrophages, and NK cells (e.g., CXCL3, CXCL10, CCL2, CCL4, etc.) as well as neutrophils (e.g., IL-6, CXCL2, CXCL5, etc.) in HeJ mice. Considering the antiapoptotic property of NF-κB (22), increased apoptosis and clearance of infected cells from suppressed NF-κB expression may also restrict viral spread in Tlr4-mutant HeJ mice. Platelet-derived growth factor B dimer (PDGF-BB) is another putative upstream regulator for HeJ-specific transcriptome changes. Because PDGF-BB-induced vascular smooth muscle cell proliferation and inflammatory vascular disease are mitigated in Tlr4^−/− mice (30), suppressed PDGF-BB activity may inhibit TLR4-mediated vascular damage in HeJ mice following RSV infection.

We found TLR4-dependent expression of the IL-1 axis (e.g., Il1β, Il1rn, Il-1r2, Il1a) and NLRP3 inflammasome genes after RSV infection. TLR4 is involved in dysregulation of inflammasome signaling by IL-1β overproduction, which causes NF-κB activation and prolonged and exacerbated inflammation leading to host tissue damage (19). Our transcriptome analysis also suggests novel TLR4 effectors of RSV disease. Relative to the large (10- to 100-fold) induction in OuJ and other Tlr4 wild-type strains, serum amyloid A1 (Saa1) was not activated in HeJ mice by RSV, indicating TLR4-dependent regulation as shown in recent studies (11). SAA1 is an apolipoprotein associated with chronic inflammatory disease and interacts with components of the extracellular matrix as a chemoattractant and regulator of immune cell trafficking (16). C-type lectin domain family 4, member e (CLC4E)-encoding gene (Clec4e), involved in cell adhesion, cell-cell signaling, and immune system surveillance, was only slightly activated in HeJ mice by RSV. Overall, in addition to reduced NF-κB signaling, the blunted transcriptional responses of potential TLR4 downstream effectors IL-1/inflammasome, SAA1, and CLC4E could be another means by which Tlr4-mutated HeJ mice are protected from RSV infection and disease.

Cellular defense systems involving the HSP family are often the target of viral pathogens that hijack the host cell’s transcriptional machinery for replication. Among the Tlr4-dependent lung transcripts, HSP70 genes (Hspa1a, Hspa1b) varied greatly between OuJ and HeJ strains but not among the other Tlr4-normal inbred strains during RSV infection. HSP70 is a known inhibitor of caspase that initiates apoptosis of virus infected cells and recruits neutrophils to produce cytokines via TLR4 (41). HSP70 localizes to mature viral particles or associates with viral polymerase complexes in lipid rafts of RSV-infected cells (3, 33). It also helps the polymerase remodel the RSV nucleocapsid for efficient viral RNA synthesis (25). Therefore, the upregulation of HSP70 in OuJ mice may increase RSV replication and susceptibility even though immunomodulatory genes were not as highly expressed as compared with other wild-type Tlr4, RSV-susceptible strains (BALBc, AJ). As there are no known genetic polymorphisms for Hspa1a and Hspa1b in OuJ mice, further investigation of upstream factors that modulate HSP70 expression (e.g., potential cis-binding proteins such as germ cell nuclear factor, early growth response protein-1, paired box gene 2) is warranted in this strain. RSV-induced upregulation of HSP90 genes (Hsp90aa1, Hsp90ab1) associated with RSV maturation, function, and stability (25, 33) was also observed in OuJ mice, suggesting an interaction of HSP70 and HSP90 in OuJ-specific RSV pathogenesis.

Our comparative transcriptome analyses between five differentially susceptible, inbred strains (all Tlr4-normal) added significant information on strain-specific transcriptomics of RSV pathogenesis. That is, while RSV-resistant (AJ, AKR, B6) and -susceptible (BALBc, OuJ) strains (17) shared some common transcriptomics in response to RSV (e.g., granulocyte and agranulocyte adhesion, IL-10/IL-6/IL-17α/IL-12 signaling, dendritic cell maturation, acute phase response), the magnitude of these transcript changes was comparable to RSV susceptibility, except for the B6 strain. More importantly, there were strain-specific gene subsets altered by RSV infection (See Figs. 6 and 7, Table 2). This reveals the significant influence of genetic factors on transcriptional and posttranscriptional mechanisms of RSV pathogenesis. It was noted that RSV mostly activated the lung transcriptome in susceptible AJ mice, but RSV-responsive lung genes in the resistant AKR mice were mostly downregulated. Predicted inhibition of cell cycle and proliferation networks in RSV-infected D2 mice suggested suppressed viral replication in the host as an underlying mechanism of RSV resistance. The most distinct expression profile of RSV-responsive genes was found in B6 mice 5 days pi. These B6 signature genes encode various keratin and EDC genes (e.g., Lce family, Rptn, S100a14, Sprr2a1) clustered in a region of mouse chromosome 3. While the EDC encodes proteins involved in terminal differentiation and cornification of basal keratinocytes, keratins such as Krt5 and Krt14 act as airway epithelium progenitor cells and are upregulated in response to lung injury (36). Distal airway keratin 5 (Krt5+/p63+) stem cells undergo proliferative expansion and assemble into nascent alveoli at sites of interstitial lung injury (13, 43). Krt5 has been implicated in repair of injured airways following influenza virus infection through repopulation of epithelial cell progenitors in the distal airways and could operate in a similar manner in RSV infection (34). Upregulation of these EDC genes was also evident in the other RSV-resistant strains, HeJ, D2, and AKR. Although the role of epidermal differentiation in RSV pathogenesis has not been well studied, our transcriptomic data suggest that host epidermal differentiation is closely linked to RSV resistance in mice, particularly in the B6 strain. While we found BALBc-like patterns of marked immune and inflammatory gene activation by RSV in B6 mice, paradoxically we found low RSV titer and lung inflammation in B6 mice (17, 26, 37). It is therefore postulated that RSV may induce a morphologic shift of lung cells to keratinizing epithelia or epidermal development that provides an “unfavorable” environment for the virus to replicate and spread, serving as an anti-viral mechanism in B6 and other RSV-resistant strains.

Fig. 7.

Proposed strain-specific mechanisms underlying differential susceptibility to respiratory syncytial virus (RSV). RSV enters airway epithelial cells (e.g., ciliated cells) by fusion of surface F protein or via endocytosis (micropinocytosis). Toll-like receptors (e.g., TLR4) also recognize RSV. Invading virus replicates in inclusion bodies, and viral entry causes overproduction of reactive oxygen species (ROS). Cell signaling (e.g., NF-κB, IRF3, STATs, p53) in airway cells and residential leukocytes via TLR4 or by ROS releases early inflammatory mediators (e.g., IL-1β, IL-6, TNF, IFN-γ), which recruit inflammatory cells for heightened immune response. The inflammasome, IL-1, C-type lectin domain family, and serum amyloid a1 are transcriptionally activated mainly through TLR4-NF-κB. Dendritic cells release type I interferons (IFN-α/IFN-β), opsonize virus, and migrate to the lymph nodes for antigen presentation. Neutrophils release cytokines and elastases, and matured Th2 lymphocytes illicit antibody production by B cells. Infiltrating natural killer cells release granzymes, perforins, IFN-γ, and TNF to kill the virus and infected cells. Macrophages kill the virus by phagocytosis. Lack of TLR4-mediated immunity in C3H/HeJ mice blunted these RSV responses. Heat shock proteins (e.g., HSP70, HSP90) may contribute to the enhanced susceptibility of C3H/HeOuJ mice by chaperoning viral entry, replication, and assembly. C57BL/6J strains have marked transcriptional activation of keratins and epidermal differentiation complex (EDC) components, which probably generate barriers for viral replication. Downregulated RNA transcription in the AKR/J strain and suppressed cell cycle genes in the DBA/2J strain are predicted to contribute to their RSV resistance.

The impact of genetic variation on predisposition to disease cannot be excluded in RSV susceptibility and transcriptomics. In addition to the Tlr4^Lps-d allele, we found additional SNPs (Supplemental Table S9) between OuJ and HeJ mice (https://phenome.jax.org/), which include a nonsynonymous mutation (rs4223249, GR:218) in Dusp19. While this SNP is not located in the Dusp19 functional domain, its role alone or in association with Tlr4^Lps-d in RSV resistance and nominal transcriptome changes in HeJ mice deserves further study. In addition to Tlr4, immune modulation was found in AJ, AKR, and D2 mice, all of which bear a mutant complement component 5 allele (Hc⁰) (35). RSV susceptibility in inbred strains of mice is polygenic and multifactorial, and alleles both common and specific to each strain [BALBc (e.g., Cdh23^ahl, Hld, Micrlⁿ), AJ (e.g., Cdh23^ahl, Hc⁰, Il3ra^m1, Hc⁰, Micrlⁿ), D2 (e.g., Fbrwt1^DBA/2J, Gpnmb^R150X, Ahr^d, Cdh23^ahl, Hc⁰, Klrd1^DBA/2J), AKR (e.g., Hc⁰, Soat1^ald), B6 (e.g., Cdh23^ahl, Ahr^b-1, Gluchos1^C57BL/6J, Gluchos2^C57BL/6J, Gluchos3^C57BL/6J), OuJ (e.g., Pde6b^rd1), and HeJ (e.g., Tlr4^Lps-d, Ahr^b-2, Gria4^spkw1, Pde6b^rd1)] could individually or collectively influence RSV susceptibility as well as the magnitude and diversity of gene expression profiles (Fig. 6, Table 2, Supplemental Table S6).

Although Tlr4 deficiency was protective in the acute-phase lung injury by RSV, we have not investigated the role of TLR4 signaling in adaptive immune responses during late-phase infection (i.e., at or after 5 days pi). At 5 days pi, several inflammatory genes (e.g., Il1b, Il6, Cxcl2) were upregulated only in HeJ or at higher levels in HeJ than in OuJ mice. Similarly, delayed clearance and persistent viral accumulation were reported in the lungs of C57BL/10ScCr (currently C57BL/10ScNJ) mice bearing Tlr4 deletion compared with the corresponding wild-type mice at 5 days pi (18). TLR signaling is known to directly activate B cells that generate antibodies against T cell-dependent antigens. In a long-term vaccination study (60 days) in mice, a formalin-inactivated or UV-inactivated RSV vaccine aggravated disease compared with live RSV, and it was suggested that poor TLR stimulation by the inactivated vaccine lead to insufficient antibody affinity maturation in mice upon viral rechallenge (10). Additional immunization studies by Cyr et al. (9) also demonstrated that Tlr4^−/− mice were defective in RSV-specific antibody and IFN-β production compared with wild-type mice. Further investigations will provide details on the role of TLR4 in the initial inflammatory response as well as the complex adaptive immune response during reinfection.

In addition to IFNs, our transcriptome analyses identified common antiviral defense genes induced by RSV in multiple inbred strains. They included the 2′-5′ oligoadenylate synthetase family (Oasl1, Oasl2) and other humoral immunity genes (Irg1, Mx1, Gzmb, Socs3). A marked decrease in sodium channel genes (e.g., Scn3a) suggests that RSV interferes with sodium transport across nonvoltage-gated epithelial sodium channels, leading to reduced alveolar fluid clearance (7). Nppa encoding atrial natriuretic peptide is a powerful vasodilator involved in the homeostatic control of body water, sodium, potassium, and fat and plays a protective role in acute lung injury associated with gram-positive infection (42). Significant RSV-induced activation of Nppa in Tlr4-sufficient OuJ, D2, B6, and AKR mice may prevent IFN-related microvascular endothelial leakage in these strains.

In conclusion, we report a detrimental role for TLR4 signaling in a murine model of acute RSV disease. Our transcriptome analyses identified TLR4-dependent as well as strain-specific RSV lung transcriptomics (Summary in Fig. 7). TLR4 signaling may affect the overall immune response including dendritic cell maturation, NK cell activation, inflammatory cell infiltration, T cell activation, and B cell proliferation as well as inflammasome activation and vascular damage during RSV infection. Constitutively lower immune cell transcripts and an increase in developmental and vasculogenic genes in HeJ mice indicate they counter RSV in a Tlr4 deficient environment. HSP70 may play a role in potentiation of RSV disease in OuJ mice, while EDC components may be essential for lung defense in B6 and other RSV-resistant strains. Results from this study provide mechanistic insight into the differential susceptibility to RSV disease and may help in the development of therapeutic intervention strategies.

GRANTS

This study was supported by funds from the Intramural Research Program of the NIEHS.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

H.-Y.C., F.P., and S.R.K. conceived and designed research; J.M., H.-Y.C., and M.H. performed experiments; J.M., H.-Y.C., and Z.R.M. analyzed data; H.-Y.C., F.P., and S.R.K. interpreted results of experiments; H.-Y.C. prepared figures; J.M. and H.-Y.C. drafted manuscript; J.M., H.-Y.C., M.H., Z.R.M., F.P., and S.R.K. edited and revised manuscript; S.R.K. approved final version of manuscript.