Gene Expression Differences in Lungs of Mice during Secondary Immune Responses to Respiratory Syncytial Virus Infection

Virus and mock.Human RSV type A2 was obtained from the American Type Culture Collection (ATCC). The virus was cultured on HEp-2 cells (ATCC) and stored in 10% fetal calf serum (FCS), which was needed to stabilize the virus concentration as described elsewhere (3, 21). The infectivity of the virus stock (PFU RSV/ml) was assessed by quantitative plaque-forming assay (3). Mock lysates were made by scraping HEp-2 cells that were not infected with RSV in the same way as described above.

FI-RSV vaccine, FI-mock, and medium.Human RSV stock A2 was grown on HEp-2 cells in RPMI 1640 medium (Gibco BRL, Life Technologies) containing 10% heat-inactivated FCS, 2 mM glutamine, penicillin, and streptomycin. Cells and media were harvested when ca. 90 to 100% cytopathic effect was observed and were centrifuged for 10 min at 1000 × g at 4°C. The supernatant was divided into aliquots and snap frozen at −80°C. The suspension contained 3 × 10⁷ PFU/ml as assessed by plaque-forming assay (3). Supernatant containing virus was thawed, filtered through a 5-μm filter (Millipore polyvinylidene difluoride [PVDF] membrane; Millipore Corporation), and then diluted in serum-free RPMI medium to a concentration of 2 × 10⁶ PFU/ml and 0.7% FCS. Afterwards, it was incubated for 3 days with formalin (1:4,000, Merck) at 37°C. The suspension was centrifuged at 50,000 × g at 4°C for 1 h. Subsequently, the pellet was resuspended in serum-free RPMI medium before binding to Al(OH)₃. The vaccine was stored at −80°C. Supernatant of mock-infected HEp-2 cells was treated in the same way and was designated the control vaccine (FI-mock). Next, the FI-RSV vaccine and FI-mock preparations were tested in a vaccination and challenge experiment, showing that the FI-RSV preparation resulted in more immunopathology than FI-mock (unpublished data). In the pilot study, culture RPMI medium was used as control vaccine (medium).

ΔG-RSV vaccine.Recombinant RSV virus lacking the G gene (ΔG-RSV) was generated by reverse genetics. The construction and characterization of this mutant have been described previously (43). The recombinant virus was cultured on monkey kidney Vero cells (ECACC lot 10-87/CCL-81; ATCC). This candidate vaccine is comparable to the mutant recombinant RSV described by Teng et al. (40).

Animals.Female specific-pathogen-free (SPF) BALB/c mice were obtained from Harlan (Horst, Netherlands) and were used at 6 to 10 weeks of age. Mice were kept in a 12-h light/dark cycle and received water and food ad libitum. A week before the experiments, the mice were housed in groups according to the experimental setup under SPF and temperature-controlled conditions. The study was approved by the Institute's committee on animal welfare.

Experimental design.First, a pilot experiment was performed to investigate vaccine-enhanced disease. Mice were vaccinated with FI-RSV or medium at 7 and 3 weeks prior to RSV infection. Transcription profiles of lungs were determined after RSV infection at the peak of viral replication to determine which early responses were involved in this enhanced pathological response (n = 6 per group). Expression profiles of unchallenged but FI-RSV- or medium-vaccinated mice also were determined. One mouse receiving FI-RSV vaccination was excluded from further analysis because of hepatosplenomegaly suggestive of underlying illness.

RSV challenge upon FI-RSV vaccination caused a different response from challenge after vaccination with medium only; predominantly Th2 genes were differentially regulated (see Table S1 in the supplemental material). These results showed that this method is informative to use in an extended study.

Subsequently, an extensive study was done in which mice (n = 6 per group) were immunized with different treatments (Fig. 1). Mice were infected intranasally (i.n.) with 10⁶ PFU RSV (in 50 μl) or mock infected at 7 weeks before challenge, immunized with 10⁵ PFU ΔG-RSV (in 20 μl) intranasally, or intramuscularly (i.m.) injected with 10⁷ PFU (in 50 μl) FI-RSV or FI-mock at 7 and 3 weeks before challenge. Before inoculation or intramuscular injection, mice were anesthetized with enflurane. At day 0, mice were intranasally challenged with 10⁶ PFU RSV in 50 μl. Mice were intraperitoneally anesthetized with ketamine, xylazine, and atropine (KRA) and sacrificed unchallenged or at 1, 2, or 5 days after challenge. Lungs were removed after perfusion. The left lung was fixed intratracheally using 4% formalin for histological examination and, after storage overnight, transferred to 70% ethanol. Fixed lungs were embedded in Paraplast (Monoject). Sections of 4 to 5 μm were stained with hematoxylin and eosin for routine histological examination and with sirius red to confirm eosinophilic infiltration (31). The spread and severity of the pneumonia as well as the composition of the perivascular infiltrates were evaluated. The right lung was immersed in RNAlater RNA stabilization reagent (Applied Biosystems), and after storage overnight at 4°C, RNA was extracted using the RNeasy kit (Qiagen). RNA concentrations were measured using a NanoDrop spectrophotometer (Thermo Scientific), and RNA quality was determined using the Bioanalyzer (Agilent Technologies).

• Open in new tab
• Download powerpoint

FIG. 1.

Time schedule of pretreatment and challenge. RSV, RSV i.n.; ΔG-RSV, ΔG vaccine i.n.; FI-RSV, FI-RSV vaccine i.m.; mock, control for RSV/ΔG vaccine i.n.; FI-mock, control for FI-RSV i.m.

Viral load.RSV A concentrations in the lungs of mice were analyzed by real-time PCR as described before (21), using the 7500 Fast real-time PCR system with primers 5′-TGAACAACCCAAAAGCATCAT-3′ and 5′-CCTAGGCCAGCAGCATTG-3′ and with probe 5′-FAM-AATTTCCTCACTTCTCCAGTGTAGTATTAGG-BHQ1-3′ (Biolegio). As an endogenous control, hypoxanthine phosphoribosyltransferase 1 (HPRT1) was used with primers 5′-GCCGAGGATTTGGAAAAAGTGTTTA-3′ and 5′-TTCATGACATCTCGAGCAAGTCTTT-3′ and with probe 5′-FAM-CAGTCCTGTCCATAATCA-MBG-3′ (Applied Biosystems). Threshold cycles (C_T values) were used for quantification of RNA copies of RSV A in the lungs of challenged mice. Low C_T values indicate high RSV concentrations, while high C_T values (with a maximum of 40 cycles) represent low viral loads.

Microarray analysis.RNA amplification and labeling were carried out with the Amino Allyl MessageAmp II aRNA kit (Applied Biosystems) according to the manufacturer's instructions, using 1.5 μg of total RNA as starting material. RNA samples from individual mice were labeled and hybridized against a common reference containing a labeled amino allyl-modified aRNA pool of all samples isolated. Microarray slides containing 22,680 oligonucleotides from the Sigma-Compugen mouse oligonucleotide library (and appropriate controls) were spotted at the Microarray Department of the University of Amsterdam. Arrays were scanned at two wavelengths using a ScanArray 4000XL microarray scanner (Perkin-Elmer). Following microarray scanning, median Cy3 and Cy5 signal intensities per spot were determined using Array Vision software (Imaging Research). Quality control was performed on raw data by means of visual inspection of the scanned images, as well as a check on the scatter and MA plots. All slides (n = 80) met our quality control criteria; i.e., fewer than 10% of the spots could be flagged as missing data, and the dye ratio did not show a signal-dependent trend exceeding a factor 10.

Statistical analysis.Raw signal data from oligonucleotide-containing spots were normalized in R using a three-step approach of natural log transformation, quantile normalization of all scans, and correction of the sample spot signal for differences in the corresponding reference spot signal between arrays (34).

Normalized data for individual genes were compared among all groups using a single one-way ANOVA. Genes with a P value of <0.001 and a maximum fold ratio of >2.0 (defined as maximum/minimum between groups) were considered sufficiently relevant for further analysis. This P value reflects the statistical likelihood that the gene is differentially expressed between groups. The false discovery rate (FDR), i.e., the fraction of false positives in the lists of regulated genes, was <0.01. The resulting gene lists were further refined using additional criteria and stringencies.

Gene expression patterns were visualized by hierarchical clustering (Euclidean distance clustering and Ward linkage) using the GeneMaths program (Applied Maths). Gene ontology term enrichment was assessed using the DAVID functional annotation tool (http://david.abcc.ncifcrf.gov/ ) (17), and additional pathway analysis was performed using MetaCore software (Genego, St. Joseph, MI).

Microarray data accession number.The raw and normalized microarray data were deposited at the ArrayExpress website (http://www.ebi.ac.uk/microarray-as/ae ) under accession number E-TABM-826.

Induction of RSV immunity.Previously we have used microarray analysis to study the early host response of BALB/c mice to primary RSV infection in the lung (21). We showed that this response occurs within 1 day and is characterized by the induction of a range of innate immune mechanisms, including chemokines, interferon (IFN) responses, and antigen processing. In addition, a pilot experiment showed that mice vaccinated with FI-RSV showed a response to RSV challenge that was distinct from that of medium-vaccinated mice (see Table S1 in the supplemental material). To explore the details and kinetics of the immune responses, a large study was performed with three different immunization regimens and two control groups. Since all groups received the same RSV challenge, the groups were designated based on their pretreatment regimen, i.e., RSV, FI-RSV, ΔG-RSV, mock, and FI-mock.

Pulmonary histopathology and viral load.To confirm infection in our model and to investigate the potential effects of cell influx on transcription profiles, lungs were analyzed by histopathology (see Fig. S1 in the supplemental material). All data are presented as the means for five mice per group (randomly selected out of the six mice per group). In unchallenged mice, only initial treatment with RSV resulted in minimal residual symptoms of perivascular inflammatory infiltrates; all other unchallenged mice did not show signs of inflammation. On day 1 after RSV challenge, mild perivasculitis with little edema was seen in the RSV- and ΔG-RSV-pretreated group; the FI-RSV-pretreated mice had only minimal inflammatory infiltrates. Minimal to moderate perivascular cuffs were seen on day 2 after challenge in all groups, with vasculitis and occasionally necrosis. At 5 days after challenge, extended perivascular reactions and interstitial pneumonia were determined in all groups. Infiltrating lymphocytes were seen in the RSV-pretreated group, in contrast to the other groups, where eosinophils had infiltrated the lung tissue as well and severe vasculitis with degeneration was observed. The lungs of the mock- and FI-mock-pretreated mice showed characteristics of primary infection. As expected, the most advanced pathology with the most severe eosinophil infiltration was seen at 5 days after challenge in the FI-RSV-pretreated group.

RSV load was estimated by the amount of viral RNA present in the lung. The viral load peaked at day 1 in all groups; this was the first time point at which viral load was measured after the inoculum was given (Fig. 2), and the load probably represents the viral RNA present in the inoculum. Threshold cycles increased (i.e., viral load diminished) in the RSV-, ΔG-RSV-, and FI-RSV-pretreated groups on day 5, while the RSV load remained invariably elevated in the mock- and FI-mock-pretreated mice up to 5 days after challenge. The severity of pathology and viral load in the lungs of RSV-infected mice associated well (data not shown), except for the FI-RSV-immunized mice, which had low RSV loads but severe histopathological changes.

• Open in new tab
• Download powerpoint

FIG. 2.

RSV load after challenge in the lungs of BALB/c mice. The concentrations of HPRT1 RNA and RSV RNA in the lungs of mice are expressed as threshold cycles (C_T values) of the PCR amplification. Low C_T values indicate high RSV concentrations, while high C_T values (with a maximum of 40 cycles) represent low viral loads. Error bars represent 95% confidence intervals.

PCA.Transcription profiles were analyzed before and 1, 2, and 5 days after RSV challenge. The transcription profile of the primary RSV response in this study (mock- and FI-mock-pretreated groups) was similar to the response measured during a primary RSV infection in our previous study (99.2% of 19,754 genes had expression levels that differed by less than a factor 2.0), indicating the robustness of this system (21). The individual gene expression differences among the mice per group were small (see Fig. S2 in the supplemental material for individual gene expression); therefore, all data are presented as the means for four mice per group (randomly selected out of the six mice per group). Up- or downregulation in gene expression was determined as >2.0-fold regulation by one-way analysis of variance (ANOVA) between the groups (P < 0.001). Figure 3 shows the principal-component analysis (PCA) of the microarray analysis of the mice per group. PCA is a mathematical algorithm that describes the data on the basis of their (dis)similarity, so that a greater distance corresponds to a greater dissimilarity. In this analysis, a principal component is defined as a mathematically derived combination of genes and their expression characteristics that can be used to describe part of the process observed. Clusters of unchallenged mice and mice on different days after RSV challenge can be clearly distinguished. From this PCA analysis it is also apparent that at early time points after challenge (days 1 and 2) all initial treatment groups cluster together, suggesting that at these early time points a strong response is induced irrespective of pretreatment of mice. This is probably related to the overwhelming innate response that occurs upon challenge. However, we cannot exclude that due to this strong innate response, other differences are no longer detectable. In contrast, on day 5 after challenge the different pretreated groups show a distinct PCA profile, and (as expected) only the two mock pretreatment groups cluster together.

• Open in new tab
• Download powerpoint

FIG. 3.

Principal-component analysis of the microarray analysis of the lungs of BALB/c mice. Every letter/number combination corresponds to the mean of one group (n = 4). The distance between the groups in the figure corresponds to the combined differences in gene expression in the groups. R, initial treatment with RSV; G, ΔG-RSV; F, FI-RSV; M, mock; N, FI-mock; 0, unchallenged; 1, 1 day after RSV challenge; 2, 2 days after RSV challenge; 5, 5 days after RSV challenge; PC, principal component.

Gene expression in unchallenged mice.First, host transcription profiles before challenge were determined, i.e., at 7 weeks after infection with RSV or 3 weeks after vaccination with ΔG-RSV or FI-RSV or after mock or FI-mock inoculation. In these last three groups no lung response was expected, and these data serve as a reference for the time points after challenge. ANOVA and hierarchical cluster analysis of microarrays of the lungs of unchallenged mice showed 90 genes which were differentially regulated, of which 82 were immunoglobulin genes (listed in Table S2 in the supplemental material). These differences were seen in the RSV- and ΔG-RSV-pretreated groups and were interpreted as indicating the presence of activated B cells, corresponding to the expected host immune response in the lungs of mice that had received prior intranasal RSV or ΔG-RSV treatment (Fig. 4). As expected, the other three groups showed no lung response after intramuscular vaccination or intranasal mock inoculation.

• Open in new tab
• Download powerpoint

FIG. 4.

Cluster analysis of regulated genes in the lungs of BALB/c mice in response to challenge with RSV. One single ANOVA was performed to study differences among all groups. All genes with a change of >2.0-fold (P < 0.001) are depicted. Gene expression changes are given relative to the median for the unchallenged groups. Each row represents the lungs of a group of mice (n = 4; treatment with RSV, ΔG-RSV, mock, FI-mock, or FI-RSV).

Gene expression on days 1 and 2 after RSV challenge.In the first 2 days after RSV challenge, there was strong up- and downregulation of gene expression compared to that in unchallenged mice; 1,137 and 959 genes were differentially regulated on days 1 and 2 after challenge compared to in unchallenged mice, respectively. This response was almost identical to the strong innate immune response observed in our previous study. However, as was already observed in the PCA analysis (Fig. 3), there were no differentially regulated genes among the different treatment groups compared to their respective controls (RSV and ΔG-RSV with mock pretreatment and FI-RSV with FI-mock pretreatment) (Fig. 4). Apparently, the first response to RSV challenge is so strong that differences between treatment groups are not visible. Based on the amount of viral RNA, one could argue that this response is a reflection of the quantity of viral RNA in the inoculum, which is similar in all treatment groups. Immunoglobulin genes were left out of consideration in this analysis, as these genes already differed in unchallenged mice.

Gene expression 5 days after RSV challenge. (i) Differences between primary and secondary immune responses.Next, host transcription profiles of mice with an initial RSV infection or mock administration were compared on day 5 after RSV challenge (Fig. 4); i.e., primary (N5 and M5 in Fig. 3) and secondary (R5 in Fig. 3) RSV infections were compared. The PCA analysis already shown that these groups were clearly different (Fig. 3), and indeed a large number of differentially expressed genes could be detected. Table 1 shows that 132 genes were differentially regulated when primary and secondary RSV infections were compared (see Table S3 in the supplemental material for a complete list of genes). These genes were categorized based on gene function (21). All categories were upregulated more strongly in response to primary RSV infection than in response to secondary RSV infection. Immunoglobulin genes were left out of consideration because these genes were already different in unchallenged mice. Only 2 genes (Ccl20 and Chi3l1) were downregulated during primary infection; all other genes were significantly more upregulated during primary infection than during secondary infection, where the response on day 5 was already diminished. The difference was most pronounced for the categories chemokine (Ccl2, Ccl7, Ccl8, Cxcl10, and Cxcl11), inflammation (Tgtp, Timp1, Gzmb, and Il18bp), and interferon response (Isg15, Ifi204, Irf7, Igtp, Ifi44, Ifi1, Ifit1, Ifi202b, Ifi27, and Iigp2). These data show that the kinetics of the primary and secondary responses to RSV infection were different (Table 1); the response to primary infection was more prolonged than the response to secondary infection. This was consistent with the amount of viral RNA present in the groups, i.e., low in the RSV group and high in the mock treatment group. This would fit with the assumption that gene expression is no longer switched on once the virus is cleared. The genes that were differentially regulated in the FI-RSV and FI-mock groups on day 5 after RSV challenge are shown in Table S4 in the supplemental material. In total, 46 genes were differentially regulated, especially interferon-regulated genes. These interferon-regulated genes were expressed at higher levels in the FI-mock group than in the FI-RSV group. No differences were seen between the primary RSV infection groups (mock and FI-mock [M5 and N5, respectively, in Fig. 3]).

(ii) Differences between FI-RSV and ΔG-RSV.On day 5, the secondary-infection group was clearly distinct from the primary-infection groups, with the latter showing gene expression corresponding to that of early time points after challenge, when high levels of viral RNA were found. The group with vaccine-induced immunity, which received the candidate RSV vaccine, was closer to the secondary-infection group, whereas the vaccine-enhanced disease group displayed remarkable resemblance to the findings for early time points after challenge even though viral clearance was apparent (Fig. 3). The distances among all three secondary-response groups (R5, G5, and F5) were almost equal. Thus, although mice with vaccine-enhanced disease had a reduction of viral replication resulting in low levels of viral RNA, the host response at the transcription level showed similarities with that seen in mice with high levels of viral RNA, i.e., with the primary immune response. The cluster analysis in Fig. 4 shows that the vaccine-enhanced disease group also showed results resembling aspects of the secondary immune responses on day 5 after challenge, indicating that this group displayed a mixed response with expression profiles of both the primary and secondary responses. One single analysis showed 88 differentially expressed genes among the RSV, ΔG-RSV, and FI-RSV groups on day 5 after challenge, and these genes were analyzed in more detail. Table 2 shows 75 genes which were differentially upregulated >2-fold between the groups with secondary RSV infection and FI-RSV vaccination, excluding immunoglobulin genes. In mice with vaccine-enhanced disease, 70 out of these 75 genes which were differentially regulated between these two groups were expressed like genes in mice with primary RSV infection (shown in Table 2 as mock and FI-mock), highlighting the similarities in expression profiles between these groups. In conclusion, at 5 days after challenge, mice with vaccine-enhanced disease have all genes expressed comparably to mice with secondary infection, except for these 75 genes. In 70 of the 75 genes which are differentially regulated, the expression in the FI-RSV mice is like that in mice with primary infection.

Next, we focused on the difference in expression levels between the groups with a reduced level of RSV replication caused either by previous RSV infection or by the candidate ΔG-RSV vaccine (Table 3). The gene expression in the ΔG-RSV group was almost similar to the expression in the secondary-infection group; only 17 genes were differentially expressed when these groups were compared. Ccl5 was expressed at a high level in the ΔG-RSV group only, whereas Ccl2, Gbp2, Gzma, Gzmb, Igtp, Il18bp, Saa2, Saa3, and Tgtp were expressed at higher levels in all groups compared to the low gene expression in the RSV-pretreated group. Both Tables 2 and 3 show that five genes (Chi3l3, Clca3, Fut2, Rdh11, and Timp1) were differentially expressed among all groups with a previously induced immune response.

Difference in Th1-Th2 polarization of the immune response.To investigate whether different pretreatment regimens induced a differentially polarized immune response, the genes in Tables 2 and 3 were designated Th1 or Th2 genes, corresponding to the classification we used before (21), when possible. A gene was designated Th1 if the gene was regulated by gamma interferon (IFN-γ) or involved in IFN-γ-mediated responses and as Th2 if that gene was regulated by interleukin-4 (IL-4) or IL-13 or was shown to be upregulated in two murine models for allergic asthma (27). The exact functions of some of these genes are not known, and therefore the phenotypic consequences of their expression cannot always be predicted. This analysis showed that on day 5 after RSV challenge, differentially expressed Th2 genes were all highly upregulated in FI-RSV-vaccinated mice, whereas their expression was not induced in the other two groups protected against RSV replication (see Fig. 5). Differentially regulated Th1 genes, on the other hand, which were highly expressed during primary RSV infection, displayed attenuated expression in the ΔG-RSV-vaccinated group and in the RSV-reinfected group. However, the FI-RSV group displayed a profile similar to that found in primary infection.

• Open in new tab
• Download powerpoint

FIG. 5.

Gene expression is upregulated in Th1 genes during primary infection and in Th2 genes in FI-RSV-pretreated mice at 5 days after challenge with RSV. Typical Th1 and Th2 genes with a change in expression of >2.0-fold (P < 0.001) on day 5 after RSV challenge are depicted. Genes regulated by IFN-γ or involved in IFN-γ responses are designated Th1 genes. Th2 genes are defined by regulation of IL-4 or IL-13 or are genes upregulated in two murine models of asthma/allergy (27). Data are presented relative to the median for the unchallenged groups. Each row represents the lungs of a group of mice (n = 4; initial treatment with RSV, ΔG-RSV, FI-RSV, mock, or FI-mock). Full gene names are described in Tables 2 and 3.