Children that were administered a formalin-inactivated respiratory syncytial virus (FI-RSV) vaccine experienced enhanced respiratory disease, including pulmonary eosinophilia, after contracting a natural RSV infection. RSV vaccine-enhanced disease can be mimicked in BALB/c mice immunized with either FI-RSV or with a recombinant vaccinia virus (vacv) expressing the RSV attachment (G) protein. We have recently demonstrated that memory CD8 T cells directed against the RSV immunodominant M282–90 epitope inhibit the development of pulmonary eosinophilia in either vacvG- or FI-RSV-immunized mice by reducing the total number of Th2 cells in the lung after RSV challenge. In this study, we show that memory CD8 T cells specific to a subdominant epitope within the RSV fusion (F) protein fail to inhibit the development of pulmonary eosinophilia after RSV challenge of mice previously co-immunized with vacvF and with either vacvG or FI-RSV. We observed that the inability of RSV F85-specific memory CD8 T cells to inhibit the development of pulmonary eosinophilia was largely due to an inadequate total number of F85-specific memory CD8 T cells in the lung at early times after RSV challenge. Increasing the number of F85-specific memory CD8 T cells after immunization grants these cells the ability to inhibit RSV vaccine-enhanced pulmonary eosinophilia. Moreover, we demonstrate that RSV-specific memory CD8 T cells, when present in sufficient numbers, inhibit the production of the Th2-associated chemokines CCL17 and CCL22. Taken together, these results indicate that RSV-specific memory CD8 T cells may alter the trafficking of Th2 cells and eosinophils into the lung.

Respiratory syncytial virus (RSV)3 infection is the leading cause of hospitalization in children under the age of 5 and also causes significant morbidity in the elderly and immunocompromised (1, 2, 3). Despite the clear need for an RSV vaccine, efforts to create one have been unsuccessful. The potential pitfalls associated with RSV immunization were highlighted during a series of trials of a formalin-inactivated RSV (FI-RSV) vaccine that were conducted during the 1960s. During these trials, ∼80% of the FI-RSV-vaccinated children required hospitalization and two children died after contracting a natural RSV infection (4, 5, 6). Histological examination of the lungs from the deceased children revealed extensive mononuclear cell infiltration including the presence of eosinophils (4, 5, 6).

The successful establishment of a mouse model that mimics the RSV vaccine-enhanced disease observed in humans has greatly aided our understanding of the immune determinants that contribute to pathology. In this model, BALB/c mice are immunized with either FI-RSV or a recombinant vaccinia virus (vacv) that expresses the RSV attachment (G) protein (vacvG), both of which elicit an RSV-specific CD4 T cell response and do not prime a detectable RSV-specific CD8 T cell response (7, 8, 9, 10, 11, 12). After subsequent RSV challenge, both FI-RSV- and vacvG-immunized mice develop robust RSV-specific memory CD4 T cell responses, pulmonary eosinophilia, and vaccine-enhanced disease that is similar to the disease observed in the children from the FI-RSV vaccine trials (9, 10, 12, 13). Previous work has demonstrated that both FI-RSV- and vacvG-immunized mice that also undergo a concurrent RSV-specific memory CD8 T cell response specific to the immunodominant RSV transcription anti-termination factor (M2)82–90 CD8 T cell epitope have significantly reduced levels of pulmonary eosinophilia after RSV challenge as compared with controls (7, 14). These data suggest that RSV-specific memory CD8 T cells inhibit the Th2 response that is necessary for the development of pulmonary eosinophilia.

Besides the well-defined immunodominant M282-specific CD8 T cell response in BALB/c mice, several other subdominant CD8 T cell epitopes have been identified (15, 16). The RSV fusion (F) protein contains a subdominant CD8 T cell epitope at aa 85–93. CD8 T cells that recognize the F85 epitope comprise ∼2–5% of the CD8 T cells in the lung at the peak of the response after acute RSV infection (15). Previous work has demonstrated that vacvF-immunized mice develop a robust pulmonary CD4 and CD8 T cell response after RSV challenge (17, 18, 19). Interestingly, Ab-mediated depletion of either CD8 T cells or IFN-γ in vacvF-immunized mice results in the development of pulmonary eosinophilia after RSV challenge (19, 20). These data suggest that IFN-γ produced by F85-specific CD8 T cells prevents the development of pulmonary eosinophilia after RSV challenge of vacvF-immunized mice. However, it remains unclear whether F85-specific memory CD8 T cells can inhibit the development of pulmonary eosinophilia after RSV challenge of either vacvG- or FI-RSV-immunized mice.

We demonstrate in this study that only either vacvG- or FI-RSV-immunized mice that are also co-immunized with vacvM2, and not with vacvF, exhibit reduced levels of pulmonary eosinophilia after RSV challenge. Interestingly, there are a similar number of M282- and F85-specific CD8 T cells in the lungs of either vacvG plus vacvM2-immunized mice or vacvG plus vacvF-immunized mice at day 7 after RSV challenge. Although the total number of RSV-specific cells in the lung is similar in vacvG-immunized mice undergoing a concurrent M282- or F85-specific CD8 T cell response at day 7 after challenge, the total number of the F85-specific memory CD8 T cells is dramatically reduced compared with M282-specific CD8 T cells at early times (days 0–3) after RSV challenge. Importantly, decreasing the total number of M282-specific CD8 T cells early after RSV challenge leads to the development of pulmonary eosinophilia, and increasing the total number of F85-specific CD8 T cells early after RSV infection results in diminished eosinophilia. Furthermore, we go on to demonstrate that M282-specific memory CD8 T cells modulate the pulmonary protein levels of Th2-associated chemokines such as CCL17 and CCL22. Taken together, these data demonstrate that the total number of RSV-specific CD8 T cells in the lung early after infection dictates their ability to regulate the events that are required for the development of pulmonary eosinophilia.

The A2 strain of RSV was a gift from B. S. Graham (National Institutes of Health, Bethesda, MD) and was propagated in HEp-2 cells (American Type Culture Collections; ATCC). Recombinant vacv were a gift from T. J. Braciale (University of Virginia, Charlottesville, VA) and J. L. Beeler (U.S. Food and Drug Administration, Bethesda, MD) and were propagated in BSC-40 cells (ATCC). BALB/cAnNCr mice between 6 and 10 wk of age were purchased from the National Cancer Institute (Bethesda, MD). IFN-γ-deficient C.129S7(B6)-Ifngtm1Ts/J mice were originally purchased from The Jackson Laboratory. Each mouse was scarified with 3 × 106 PFU of recombinant vacv or a mixture of two different recombinant vacv and challenged with RSV 3 wk later as previously described (7). Alternatively, BALB/c mice were i.m. immunized with a 1/200 dilution of FI-RSV or formalin-inactivated mock preparation of HEp-2 cell supernatants as previously described (21). Four weeks after mock or FI-RSV immunization, mice were scarified with 3 × 106 PFU of vacvM2 or vacvF and 4 wk later were challenged intranasally with 3 × 106 PFU of RSV. All mouse experiments have been evaluated and approved by the University of Iowa Institutional Animal Care and Use Committee.

Lung mononuclear cells and bronchoalveolar lavage (BAL) cells were harvested and prepared as previously described (7). Cytospin (Cytospin 2; Thermo Shandon) preparations of BAL cells were stained with Diff-Quik (Baxter Healthcare) before analysis. Differential cell counts were performed on at least 200 cells based on standard morphology and staining characteristics. Lungs were subsequently perfused with 5 ml of PBS before removal. Lung mononuclear cells were isolated by pressing the lung tissue through a wire mesh screen (Cellector; Bellco Glass). Peptides corresponding to the CD4 T cell epitopes G183–195 and F51–66 and the CD8 T cell epitopes F85–93 and M282–90 were purchased from Biosynthesis. To enumerate the number of RSV-specific CD4 and CD8 T cells, lung mononuclear cells were stimulated in vitro in the presence of 1 μM peptide and 10 μg/ml brefeldin A (Sigma-Aldrich) for 5 h at 37°C. Cells were subsequently stained for cell surface CD4, CD8, Thy1.2 and intracellular IFN-γ as previously described (7). Peripheral blood was collected from isoflurane-anesthetized mice by eye bleed into 4% (w/v) sodium citrate. Blood was subsequently lysed with ammonium chloride and washed with RPMI 1640. Enumeration of F85- or M282-specific CD8 T cells in the blood was performed by in vitro stimulation in either the presence or the absence of an optimized target to effector ratio (1:3) of P815 cells (ATCC) coated with 1 μM F85–93 or M282–90 peptide followed by ICS for IFN-γ. Lamp1/2 and granzyme B staining were done in combination with IFN-γ ICS. Lamp1 (clone 1D4B; BD Biosciences) and Lamp2 (clone M3/84; Biolegend) Abs were added before incubation of lung mononuclear cells with P815 cells coated with 1 μM F85–93 peptide in the presence of 2 μM monensin (Sigma-Aldrich). Granzyme B (Caltag Laboratories) staining was done after the 5-h ICS incubation in the presence of 1 μM F85–93 peptide and 10 μg/ml brefeldin A.

Lungs were harvested from vacv-immunized mice on either day 4 or 7 after RSV challenge in 1 ml of serum-free RPMI 1640. Lung tissue was disrupted using a tissue homogenizer (Ultra-Turrax T25; IKA Works), and lung homogenates were subsequently centrifuged at 2000 rpm for 10 min. Cell-free supernatants were collected and frozen in liquid nitrogen. Frozen samples were thawed and then subjected to plaque assay on Vero cells as previously described (7).

Lungs and BAL were harvested at day 3 after RSV challenge of vacv-immunized mice. Lungs were disrupted using a tissue homogenizer (Ultra-Turrax T25; IKA Works) in 10% FCS containing RPMI 1640 containing a 1/200 dilution of a protease inhibitor cocktail (Sigma-Aldrich). BAL samples were collected via lavage of the lung with 1 ml of RPMI 1640 containing 10% FCS. Lung homogenates and BAL samples were centrifuged at 2000 rpm for 10 min and chemokine protein content for CCL11, CCL17, and CCL22 in the supernatants was subsequently determined by ELISA as previously described (19). For CXCL10 (IP-10) and CCL5 (RANTES) ELISA analysis, flat-bottom 96-well plates (Nunc) were coated overnight at 4°C with 2 μg/ml purified anti-CXCL10 and anti-CCL5 (all Abs from R&D Systems). After extensive washing, samples were added to coated plates in 50 μl per well followed by incubation overnight at 4°C. Biotinylated detection Abs for CXCL10 and CCL5 (R&D Systems) at 0.1 μg/ml were added to plates for 2 h at room temperature. Following incubation with the biotinylated Abs, plates were washed and developed as previously described (19). ELISA plates were analyzed by an Elx800 reader (Bio-Tek Instruments) and data were analyzed using KC Junior software (Bio-Tek Instruments). Detection limits for chemokine protein levels were: 624 pg/ml CXCL10; 62.5 pg/ml CCL5; 31.2 pg/ml CCL11; 62.5 pg/ml CCL17; and 15.6 pg/ml CCL22.

Flt3 ligand-secreting B16-melanoma cells were a gift from J. T. Harty (University of Iowa, Iowa City, IA) and were cultured in RPMI 1640 containing 10% FCS. A total of 1 × 106 Flt3 ligand-secreting B16 melanoma cells were i.p. injected into BALB/c mice in 200 μl of PBS. At 14 days post-injection, mice were i.p. injected with 2 μg of LPS (Sigma-Aldrich) in 200 μl of PBS. The following day, spleens from these mice were harvested in PBS and single cell suspensions were prepared as described. The DC content of these cells was determined by staining for CD11c (clone N418; eBioscience) and MHC class II (clone M5/114.15.2; eBioscience). Maturation of DC was monitored by expression of CD80 (clone 16–10A1; eBioscience) and CD86 (clone GL1; eBioscience). Over 90% of DC routinely expressed these costimulatory molecules (data not shown). The whole cell suspension was then incubated at 37°C for 1 h in either the presence or absence of 1 μM F85–93 peptide in PBS containing 2% normal rat serum (Jackson ImmunoResearch Laboratories). Cells were washed twice in PBS and 1 × 106 DC, as determined by flow cytometry, were i.v. injected into naive recipients.

Statistical analyses were performed using Prism software (GraphPad). Data were analyzed using a Student’s t test or, where noted, by ANOVA followed by a Tukey post-test. Differences were considered significant when p < 0.05.

Previous work has demonstrated that RSV M282-specific memory CD8 T cells inhibit the development of pulmonary eosinophilia after RSV challenge of either vacvG- or FI-RSV-immunized mice (7, 14). The F protein of RSV elicits a subdominant CD8 T cell response directed against aa 85–93 (15).

It is currently unclear whether F85-specific memory CD8 T cells can inhibit vacvG-induced pulmonary eosinophilia. Mice immunized with a 1:1 ratio of either vacvG and vacvF (vacvG plus vacvF) or vacvG and vacvM2 (vacvG plus vacvM2) induced a similar total number of Ag-specific CD8 T cells at day 7 after RSV challenge (Fig. 1, A and B). With a similar total number of F85- and M282-specific CD8 T cells, we predicted that F85-specific CD8 T cells would inhibit the development of pulmonary eosinophilia in vacvG-immunized mice. Surprisingly, vacvG plus vacv expressing β-galactosidase (β-gal)-immunized and vacvG plus vacvF-immunized mice developed a similar frequency (Fig. 1,C) and a total number (Fig. 1,D) of eosinophils in the BAL at day 7 after RSV challenge with RSV, whereas mice immunized with vacvG plus vacvM2 exhibited a significant reduction (p < 0.05) in both the frequency and the total number of eosinophils in the BAL at this time (Fig. 1, C and D). As previously described (7), only vacvG plus vacvM2- and not vacvG plus vacvF-immunized mice demonstrated a significant decrease (p < 0.05) in the total number of G183-specific CD4 T cells (Fig. 1,E). As expected only mice immunized with either vacvG plus vacvF or vacvF alone mounted an enhanced F51-specific CD4 and F85-specific CD8 T cell response (Fig. 1, A and F). Furthermore, only vacvG plus vacvM2-immunized mice exhibited an increased total number of M282-specific CD8 T cells over the acute (i.e., vacvβ-gal-immunized) RSV response (Fig. 1 B).

FIGURE 1.

F85-specific memory CD8 T cells do not inhibit the development of pulmonary eosinophilia after RSV challenge of mice previously immunized with vacvG. BALB/c mice were immunized with vacvβ-gal, vacvG plus vacvβ-gal, vacvF, vacvG plus vacvF, or vacvG plus vacvM2. At least 21 days after vacv immunization mice were intranasally challenged with RSV. Seven days post-infection, the total number of F85-specific (A) and M282-specific (B) CD8 T cells was determined by IFN-γ ICS. The frequency (C) and total number (D) of eosinophils in the BAL were enumerated by quantitative morphometry. The total number of G183-specific (E) and F51-specific (F) CD4 T cells were quantified by IFN-γ ICS. Detection limits at <1 × 103 cells were below limit of detection (B.L.D.). Error bar represents the SEM from n = 4 mice per group. These data represent three independent experiments. ∗, p < 0.05, significant difference compared with vacvG plus vacvβ-gal-immunized controls as determined by ANOVA.

FIGURE 1.

F85-specific memory CD8 T cells do not inhibit the development of pulmonary eosinophilia after RSV challenge of mice previously immunized with vacvG. BALB/c mice were immunized with vacvβ-gal, vacvG plus vacvβ-gal, vacvF, vacvG plus vacvF, or vacvG plus vacvM2. At least 21 days after vacv immunization mice were intranasally challenged with RSV. Seven days post-infection, the total number of F85-specific (A) and M282-specific (B) CD8 T cells was determined by IFN-γ ICS. The frequency (C) and total number (D) of eosinophils in the BAL were enumerated by quantitative morphometry. The total number of G183-specific (E) and F51-specific (F) CD4 T cells were quantified by IFN-γ ICS. Detection limits at <1 × 103 cells were below limit of detection (B.L.D.). Error bar represents the SEM from n = 4 mice per group. These data represent three independent experiments. ∗, p < 0.05, significant difference compared with vacvG plus vacvβ-gal-immunized controls as determined by ANOVA.

Close modal

Immunization of BALB/c mice with vacvG induces an oligoclonal Vβ14+ CD4 T cell population that is required for the development of pulmonary eosinophilia after RSV challenge (8, 22). Previously, we have demonstrated that RSV M282-specific CD8 T cells reduce the total number of CD4 as well as Vβ14+ CD4 T cells in the lung after RSV challenge of vacvG-immunized mice (7). Having shown that F85-specific CD8 T cells do not inhibit the development of pulmonary eosinophilia after RSV challenge of vacvG plus vacvF-immunized mice; we next asked whether these cells were also unable to inhibit the expansion of CD4 and Vβ14+ CD4 T cells in the lung after RSV challenge. Fig. 2 demonstrates that only mice immunized with vacvG plus vacvM2 and not vacvG plus vacvF exhibited a significantly reduced (p < 0.05) total number of CD4 (Fig. 2,A) and Vβ14+ CD4 (Fig. 2 B) T cells in the lung after RSV challenge.

FIGURE 2.

A concurrent M282-specific CD8 T cell response reduces the total number of Vβ14+ CD4 T cells in the lung after RSV challenge. BALB/c mice were immunized with recombinant vacv and subsequently intranasally challenged with RSV at least 21 days after vacv immunization as described in Fig. 1. The total number of CD4 (A) and Vβ14+ (B) CD4 T cells in the lung at day 7 after challenge were quantified by flow cytometry. Error bar represents the SEM from n = 3–4 mice per group. Data represent three independent experiments. ∗, p < 0.05. Levels at p > 0.05 were not significantly different (n.s.).

FIGURE 2.

A concurrent M282-specific CD8 T cell response reduces the total number of Vβ14+ CD4 T cells in the lung after RSV challenge. BALB/c mice were immunized with recombinant vacv and subsequently intranasally challenged with RSV at least 21 days after vacv immunization as described in Fig. 1. The total number of CD4 (A) and Vβ14+ (B) CD4 T cells in the lung at day 7 after challenge were quantified by flow cytometry. Error bar represents the SEM from n = 3–4 mice per group. Data represent three independent experiments. ∗, p < 0.05. Levels at p > 0.05 were not significantly different (n.s.).

Close modal

We analyzed virus clearance to determine its potential role in the differential development of pulmonary eosinophilia after RSV challenge of mice previously immunized with either vacvG plus vacvF or vacvG plus vacvM2. All immunized mice had significantly decreased (p < 0.05) viral titers at both day 4 and day 7 after RSV challenge as compared with vacvβ-gal-immunized mice that do not mount an RSV-specific memory T cell response (Fig. 3). Additionally, both vacvG plus vacvF- and vacvG plus vacvM2-immunized mice, which undergo memory CD4 and CD8 T cell responses, had significantly reduced (p < 0.05) viral titers in their lungs as compared with vacvG plus vacvβ-gal-immunized mice at day 4 after RSV challenge. However, there was no significant difference (p > 0.05) in the viral titers in the lungs of vacvG plus vacvF- and vacvG plus vacvM2-immunized mice at day 4 after RSV challenge, suggesting that differential clearance of the virus does not account for the difference in pulmonary eosinophilia exhibited by these mice (see Fig. 1).

FIGURE 3.

Viral clearance in vacv-immunized mice. BALB/c mice were immunized with recombinant vacv and challenged with RSV as described in Fig. 1. Lungs from vacv-immunized mice were harvested at 4 and 7 days post-challenge with RSV, and cell-free homogenates were assessed for viral titers by plaque assay on Vero cells. Error bar represents the SEM from n = 4 mice per group. These data represent three independent experiments. ∗, p < 0.05 as determined by ANOVA. Levels at p > 0.05 were not significantly different (n.s.). Detection limits at <1 × 102 PFU were below limit of detection (B.L.D.).

FIGURE 3.

Viral clearance in vacv-immunized mice. BALB/c mice were immunized with recombinant vacv and challenged with RSV as described in Fig. 1. Lungs from vacv-immunized mice were harvested at 4 and 7 days post-challenge with RSV, and cell-free homogenates were assessed for viral titers by plaque assay on Vero cells. Error bar represents the SEM from n = 4 mice per group. These data represent three independent experiments. ∗, p < 0.05 as determined by ANOVA. Levels at p > 0.05 were not significantly different (n.s.). Detection limits at <1 × 102 PFU were below limit of detection (B.L.D.).

Close modal

We have previously demonstrated that induction of an M282-specific memory CD8 T cell response in FI-RSV-immunized mice inhibits the development of pulmonary eosinophilia after RSV challenge (7). The pulmonary eosinophilia that occurs after RSV challenge of FI-RSV-immunized mice is independent of the oligoclonal Vβ14+ CD4 T cell expansion that is required for the development of pulmonary eosinophilia after RSV challenge of vacvG-immunized mice (11). We have shown (see Fig. 2) that only M282-specific and not F85-specific memory CD8 T cells reduce the total number of Vβ14+ CD4 T cells in the lung after RSV challenge of vacvG plus vacvM2- and vacvG plus vacvF-immunized mice, respectively. Therefore it remained possible that F85-specific memory CD8 T cells could inhibit the development of pulmonary eosinophilia after RSV challenge of FI-RSV- but not vacvG-immunized mice. To determine whether F85-specific memory CD8 T cells can inhibit pulmonary eosinophilia after RSV challenge of FI-RSV-immunized mice, mice were either mock treated or i.m. immunized with FI-RSV. Four weeks after initial priming, mice were subsequently immunized with vacvβ-gal, vacvF, or vacvM2 before intranasal challenge with RSV after an additional 4 wk. Fig. 4 demonstrates that only FI-RSV-immunized mice co-immunized with vacvM2 and not vacvF have significantly reduced (p < 0.05) total numbers (Fig. 4,A) of eosinophils in the BAL. Similar to what we have previously observed with vacvG plus vacvF- and vacvG plus vacvM2-immunized mice (see Fig. 1), there was not a significant difference (p > 0.05) in the total number of either F85- or M282-specific CD8 T cells in FI-RSV/vacvM2- and FI-RSV/vacvF-immunized mice (Fig. 4 B).

FIGURE 4.

F85-specific memory CD8 T cells do not inhibit FI-RSV-induced pulmonary eosinophilia. BALB/c mice were i.m. immunized with 100 μl of a 1/200 dilution of either FI-RSV or a mock control. Mice were subsequently challenged with RSV 4 wk later. Seven days post-infection, the total number of eosinophils (A) in the BAL was enumerated by quantitative morphometry. B, F85- or M282-specific CD8 T cells were quantified by IFN-γ ICS. Error bars represent the SEM from n = 3–4 mice per group. These data represent three independent experiments. ∗, p < 0.05, compared with vacvβ-gal-immunized controls. Levels at p > 0.05 were not significantly different (n.s.). Detection limits at <1 × 103 cells were below limit of detection (B.L.D.).

FIGURE 4.

F85-specific memory CD8 T cells do not inhibit FI-RSV-induced pulmonary eosinophilia. BALB/c mice were i.m. immunized with 100 μl of a 1/200 dilution of either FI-RSV or a mock control. Mice were subsequently challenged with RSV 4 wk later. Seven days post-infection, the total number of eosinophils (A) in the BAL was enumerated by quantitative morphometry. B, F85- or M282-specific CD8 T cells were quantified by IFN-γ ICS. Error bars represent the SEM from n = 3–4 mice per group. These data represent three independent experiments. ∗, p < 0.05, compared with vacvβ-gal-immunized controls. Levels at p > 0.05 were not significantly different (n.s.). Detection limits at <1 × 103 cells were below limit of detection (B.L.D.).

Close modal

One possible reason F85-specific memory CD8 T cells are unable to inhibit the development of pulmonary eosinophilia after RSV challenge of either vacvG- or FI-RSV-immunized mice may be due to a defect in their effector function. To address this possibility, we compared the ability of F85- and M282-specific CD8 T cells to produce common CD8 T cell effector cytokines (i.e., IFN-γ, TNF-α, or IL-2) by ICS. Interestingly, F85-specific CD8 T cells contained a higher frequency of cells capable of co-producing the cytokines IFN-γ and either TNF-α or IL-2 as compared with M282-specific CD8 T cells (Fig. 5, A and B). Furthermore, the relative amount of IFN-γ, TNF-α, and IL-2 produced, as measured by mean fluorescence intensity (MFI), was greater in F85-specific CD8 T cells as compared with M282-specific CD8 T cells (Fig. 5 C). Taken together, these data suggest that F85-specific CD8 T cells exhibit at least similar if not greater cytokine production capacity as compared with their M282-specific counterparts.

FIGURE 5.

Functional differences of F85- and M282-specific memory CD8 T cells. Lung mononuclear cells were harvested from vacvG plus vacvF-immunized and vacvG plus vacvM2-immunized mice 7 days post-challenge with RSV. ICS was performed by mixing lung mononuclear cells from vacv-immunized mice at a 3:1 ratio with P815 cells pulsed with 1 μM F85–93 or M282–90 peptide for 5 h in the presence of brefeldin A. A, CD8 T cells were identified by gating on CD8 and Thy1.2. Representative dot plots of costaining with IFN-γ and either IL-2 or TNF-α from gated lung mononuclear cells isolated from vacvG plus vacvF- and vacvG plus vacvM2-immunized mice. Numbers indicate frequency of cells in each quadrant. B, Frequency of either IFN-γ and TNF-α- or IFN-γ and IL-2-coproducing F85- and M282-specific CD8 T cells. C, MFI of IFN-γ, TNF-α, and IL-2 staining after ICS of F85- and M282-specific CD8 T cells. ∗, p < 0.05. Data represent two individual experiments with n = 4 mice per group. Error bar represents the SEM.

FIGURE 5.

Functional differences of F85- and M282-specific memory CD8 T cells. Lung mononuclear cells were harvested from vacvG plus vacvF-immunized and vacvG plus vacvM2-immunized mice 7 days post-challenge with RSV. ICS was performed by mixing lung mononuclear cells from vacv-immunized mice at a 3:1 ratio with P815 cells pulsed with 1 μM F85–93 or M282–90 peptide for 5 h in the presence of brefeldin A. A, CD8 T cells were identified by gating on CD8 and Thy1.2. Representative dot plots of costaining with IFN-γ and either IL-2 or TNF-α from gated lung mononuclear cells isolated from vacvG plus vacvF- and vacvG plus vacvM2-immunized mice. Numbers indicate frequency of cells in each quadrant. B, Frequency of either IFN-γ and TNF-α- or IFN-γ and IL-2-coproducing F85- and M282-specific CD8 T cells. C, MFI of IFN-γ, TNF-α, and IL-2 staining after ICS of F85- and M282-specific CD8 T cells. ∗, p < 0.05. Data represent two individual experiments with n = 4 mice per group. Error bar represents the SEM.

Close modal

Although we found an equivalent number of F85- and M282-specific CD8 T cells in the lung at day 7 after RSV challenge, it was unclear whether the early kinetics of the F85- and M282-specific memory CD8 T cell responses were similar. Therefore, we examined the total number of F85- and M282-specific CD8 T cells in the lung before and at various time points after RSV challenge. Fig. 6 demonstrates that before RSV challenge (21 days after vacv-immunization) there was a significantly greater (p < 0.05, 3–4 fold) total number of M282-specific CD8 T cells in the spleen, blood, inguinal lymph nodes, and lungs of vacvG plus vacvM2-immunized mice (Fig. 6,A) as compared with F85-specific CD8 T cells in vacvG plus vacvF-immunized mice. Furthermore, there are significantly (p < 0.05) more M282-specific CD8 T cells in the lungs of vacvG plus vacvM2-immunized mice as compared with the number of F85-specific CD8 T cells in the lungs of vacvG plus vacvF-immunized mice at day 3 after RSV challenge (Fig. 6,B). Consistent with the data presented in Fig. 1, the total number of F85- and M282-specific CD8 T cells at day 7 after RSV challenge is not significantly different (p > 0.05) (Fig. 6 B). These results suggest that the total number of RSV-specific memory CD8 T cells in the lung at early time points after RSV challenge may be a critical determinant to the inhibition of pulmonary eosinophilia.

FIGURE 6.

Kinetics of the F85- and M282-specific memory CD8 T cell responses in vacv-immunized mice. BALB/c mice were immunized with recombinant vacv and subsequently intranasally challenged with RSV at least 21 days after vacv immunization as described in Fig. 1. A, Spleens, lungs, inguinal lymph nodes (ILN), and peripheral blood were harvested 21 days post-immunization with vacv. The total number of F85-specific (▪) and M282-specific (□) were determined by IFN-γ ICS. Error bar represents the SEM from n = 3 mice per group, and inguinal lymph nodes were pooled from n = 3 mice. B, Lungs were harvested at 0, 3, or 7 days postchallenge with RSV and the total number of RSV-specific CD8 T cells was quantified by IFN-γ ICS. Error bar represents the SEM from n = 3–4 mice per group. Data represent three independent experiments. ∗, p < 0.05, significantly different as compared with vacvG plus vacvF-immunized controls.

FIGURE 6.

Kinetics of the F85- and M282-specific memory CD8 T cell responses in vacv-immunized mice. BALB/c mice were immunized with recombinant vacv and subsequently intranasally challenged with RSV at least 21 days after vacv immunization as described in Fig. 1. A, Spleens, lungs, inguinal lymph nodes (ILN), and peripheral blood were harvested 21 days post-immunization with vacv. The total number of F85-specific (▪) and M282-specific (□) were determined by IFN-γ ICS. Error bar represents the SEM from n = 3 mice per group, and inguinal lymph nodes were pooled from n = 3 mice. B, Lungs were harvested at 0, 3, or 7 days postchallenge with RSV and the total number of RSV-specific CD8 T cells was quantified by IFN-γ ICS. Error bar represents the SEM from n = 3–4 mice per group. Data represent three independent experiments. ∗, p < 0.05, significantly different as compared with vacvG plus vacvF-immunized controls.

Close modal

We next asked whether this discrepancy between the early kinetics of the F85- and M282-specific memory CD8 T cell responses could account for their differential ability to inhibit the development of pulmonary eosinophilia. We therefore altered the magnitude of the M282-specific memory response by immunizing mice with a ratio of 200:1 vacvG to vacvM2 (200 vacvG:vacvM2) in an attempt to match the early kinetics of the F85-specific CD8 T cell response. Fig. 7 demonstrates that the total number of M282-specific CD8 T cells in vacvG plus vacvM2-immunized mice at day 3 after RSV challenge is 3- to 4-fold greater than the total number of F85-specific CD8 T cells in vacvG plus vacvF-immunized mice (Fig. 7,A). Mice immunized with 200 vacvG:vacvM2 had an equivalent total number of M282-specific cells in the lung as compared with the total number of F85-specific cells in vacvG plus vacvF-immunized mice at day 3 after RSV challenge (Fig. 7,A). As discussed, at day 7 after RSV challenge, there was a similar total number of F85- and M282-specific CD8 T cells in the lungs of vacvG plus vacvF- and vacvG plus vacvM2-immunized mice, respectively (Fig. 7,B). Importantly, there were substantially more M282-specific CD8 T cells in 200 vacvG:vacvM2-immunized mice as compared with vacvβ-gal-immunized mice, indicating that we successfully primed a M282-specific memory CD8 T cell response (Fig. 7,B). However, the M282-specific memory CD8 T cell response in 200 vacvG:vacvM2-immunized mice is significantly reduced (2- to 3-fold, p < 0.05) as compared with vacvG plus vacvM2-immunized mice at day 7 after RSV challenge (Fig. 7,B). Mice immunized with vacvG plus vacvβ-gal have a robust RSV G183-specific CD4 T cell response at day 7 after RSV challenge (Fig. 7C). This G183-specific CD4 T cell response is significantly reduced (p < 0.05) in mice undergoing a concurrent M282-specific, but not a concurrent F85-specific, memory CD8 T cell response (Fig. 7,C). Interestingly, when we equalized the M282- and F85-specific response at day 3 after RSV challenge by immunizing mice with 200 vacvG:vacvM2 (Fig. 7,A) we observed a significant increase (p < 0.05) in the total number of G183-specific CD4 T cells at day 7 after RSV challenge. As demonstrated, mice immunized with vacvG plus vacvβ-gal and vacvG plus vacvF have a similar total number (Fig. 7,D) of eosinophils and this is significantly reduced (p < 0.05) in vacvG plus vacvM2-immunized mice after RSV challenge (Fig. 7,D). Interestingly, 200 vacvG:vacvM2-immunized mice have similar levels of pulmonary eosinophilia as compared with vacvG plus vacvβ-gal- and vacvG plus vacvF-immunized mice (Fig. 7,D). To ensure that M282-specific CD8 T cells induced by vacvG plus vacvM2 and 200 vacvG:vacvM2 immunization behave similarly after RSV challenge we compared the frequency of IFN-γ-producing CD8 T cells (Fig. 7,E) and the MFI of IFN-γ produced by these cells by IFN-γ ICS (Fig. 7,F). Fig. 7,E demonstrates that there is a slight decrease in the frequency of IFN-γ-producing cells that is consistent with the 2- to 3-fold decrease in the total number of M282-specific CD8 T cells noted earlier (Fig. 7 B). Additionally, M282-specific CD8 T cells in vacvG plus vacvM2- and 200 vacvG:vacvM2-immunized mice produce IFN-γ at a similar per cell capacity as measured by the MFI of IFN-γ, which is significantly greater than the nonstimulated controls (p < 0.05). These data indicate that the early kinetics of entry and accumulation in the lung by RSV-specific memory CD8 T cells is a major determinant of their ability to inhibit the development of pulmonary eosinophilia after RSV challenge of vacvG-immunized mice.

FIGURE 7.

Reducing the early M282-specific CD8 T cell response results in pulmonary eosinophilia. BALB/c mice were immunized with recombinant vacv and challenged with RSV as described in Figs. 1 and 6. An additional group of mice were immunized with 200 vacvG:vacvM2, a 1/200 dilution of vacvM2 mixed 1:1 with vacvG. A, Lungs were harvested 3 days after RSV challenge and the total number of F85-specific (▦) and M282-specific (▪) CD8 T cells were quantified by IFN-γ ICS. B, The total number of F85-specific (▦) and M282-specific (▪) CD8 T cells 7 days after RSV challenge as quantified by IFN-γ ICS. C, The total number of G183-specific CD4 T cells 7 days after RSV challenge as quantified by IFN-γ ICS. D, The total number of eosinophils in the BAL of vacv-immunized mice at day 7 after RSV challenge was quantified by quantitative morphometry. Detection limits at <1 × 103 cells were below limit of detection (B.L.D.). E, Representative dot plots of IFN-γ production by CD8 T cells in vacvβ-gal-, vacvG plus vacvM2-, or 200 vacvG:vacvM2-immunized mice isolated from the lungs at day 7 after RSV challenge. Frequency of IFN-γ-producing CD8 T cells is shown for each gate. F, Quantification of IFN-γ MFI from mice represented in E. Error bar represents the SEM from n = 3–4 mice per group. Data represent three independent experiments. ∗, p < 0.05.

FIGURE 7.

Reducing the early M282-specific CD8 T cell response results in pulmonary eosinophilia. BALB/c mice were immunized with recombinant vacv and challenged with RSV as described in Figs. 1 and 6. An additional group of mice were immunized with 200 vacvG:vacvM2, a 1/200 dilution of vacvM2 mixed 1:1 with vacvG. A, Lungs were harvested 3 days after RSV challenge and the total number of F85-specific (▦) and M282-specific (▪) CD8 T cells were quantified by IFN-γ ICS. B, The total number of F85-specific (▦) and M282-specific (▪) CD8 T cells 7 days after RSV challenge as quantified by IFN-γ ICS. C, The total number of G183-specific CD4 T cells 7 days after RSV challenge as quantified by IFN-γ ICS. D, The total number of eosinophils in the BAL of vacv-immunized mice at day 7 after RSV challenge was quantified by quantitative morphometry. Detection limits at <1 × 103 cells were below limit of detection (B.L.D.). E, Representative dot plots of IFN-γ production by CD8 T cells in vacvβ-gal-, vacvG plus vacvM2-, or 200 vacvG:vacvM2-immunized mice isolated from the lungs at day 7 after RSV challenge. Frequency of IFN-γ-producing CD8 T cells is shown for each gate. F, Quantification of IFN-γ MFI from mice represented in E. Error bar represents the SEM from n = 3–4 mice per group. Data represent three independent experiments. ∗, p < 0.05.

Close modal

Th2-associated chemokines are important in the development of RSV vaccine-enhanced pulmonary eosinophilia (19, 23, 24, 25). The production of several Th2-associated chemokines peaks early after RSV challenge of vacvG-immunized mice (26). Therefore, we asked whether either a concurrent F85- or M282-specific memory CD8 T cell response would reduce the early Th2-associated chemokine production after RSV challenge of vacvG-immunized mice. Lungs and BAL were harvested at day 3 after RSV challenge and subsequently analyzed for protein levels of either Th1-associated (CXCL10, CCL5) or Th2-associated (CCL11, CCL17, CCL22) chemokines by ELISA. Mice immunized with vacvG, vacvG plus vacvF, vacvG plus vacvM2, or 200 vacvG:vacvM2 showed similar enhanced protein levels of all Th1-associated chemokines and interestingly the Th2-associated chemokine CCL11 in both the lung and BAL (Fig. 8). Our data also demonstrates that only M282-specific, but not F85-specific, memory CD8 T cells are able to significantly decrease (p < 0.05) protein levels of CCL17 and CCL22 in both the lung and BAL (Fig. 8). Importantly, mice immunized with 200 vacvG:vacvM2 had increased levels (p < 0.05) of CCL17 and CCL22 as compared with mice immunized with vacvG plus vacvM2, suggesting that a reduced number of M282-specific memory CD8 T cells are unable to alter the chemokine environment (Fig. 8). These data indicate that M282-specific memory CD8 T cells prevent the trafficking of Th2 cells, and possibly eosinophils, into the lung by inhibiting the early production of the Th2-associated chemokines CCL17 and CCL22.

FIGURE 8.

Chemokine protein levels in vacv-immunized mice after RSV challenge. BALB/c mice were immunized with recombinant vacv and subsequently challenged with RSV as described in Figs. 1 and 6. At 3 days after RSV infection BAL (A) and lungs (B) were harvested. Lungs were homogenized, and chemokine protein levels from cell-free BAL and lung supernatants were determined by ELISA. Error bar represents the SEM from n = 3–4 mice per group. Data represent three independent experiments. ∗, p < 0.05, compared with vacvG plus vacvβ-gal-immunized controls as determined by ANOVA.

FIGURE 8.

Chemokine protein levels in vacv-immunized mice after RSV challenge. BALB/c mice were immunized with recombinant vacv and subsequently challenged with RSV as described in Figs. 1 and 6. At 3 days after RSV infection BAL (A) and lungs (B) were harvested. Lungs were homogenized, and chemokine protein levels from cell-free BAL and lung supernatants were determined by ELISA. Error bar represents the SEM from n = 3–4 mice per group. Data represent three independent experiments. ∗, p < 0.05, compared with vacvG plus vacvβ-gal-immunized controls as determined by ANOVA.

Close modal

Having established that reducing the total number of M282-specific CD8 T cells that enter the lung early after RSV infection results in the development of pulmonary eosinophilia (Fig. 7), we next asked whether increasing the number of memory F85-specific CD8 T cells early after RSV challenge would result in the abrogation of pulmonary eosinophilia. Mice were immunized with 1 × 106 DC that were either left untreated or coated with F85–93-peptide. After 2–3 wk, DC-immunized mice were subsequently immunized with vacvβ-gal, vacvG plus vacvβ-gal, or vacvG plus vacvF and challenged with RSV after an additional 3 wk of rest. We were unable to detect F85-specific CD8 T cells in either the peripheral blood or lungs of mice immunized with DC-F85–93 and vacvβ-gal, whereas we were able to detect a low number of F85-specific CD8 T cells after DC-no peptide and vacvG plus vacvF immunization (Fig. 9, A and B). However, mice immunized with DC-F85–93 and vacvG plus vacvF exhibited a significantly greater (p < 0.05) number of F85-specific CD8 T cells in both the peripheral blood and lung 21 days after vacv immunization as compared with mice immunized with DC-no peptide and vacvG plus vacvF (Fig. 9, A and B), demonstrating that we were able to boost the number of F85-specific memory CD8 T cells in the peripheral blood and lung 5- to 10-fold. At day 3 after RSV challenge of DC- and vacv-immunized mice there was a significantly (p < 0.05) greater total number of F85-specific CD8 T cells in the lungs of DC-F85–93 and vacvG plus vacvF-immunized mice as compared with their DC-no peptide-immunized counterparts (Fig. 9,B). After RSV challenge only DC-F85–93 and vacvG plus vacvF-immunized mice had a significantly decreased (p < 0.05) total number of eosinophils in the lungs as compared with vacvG plus vacvβ-gal-immunized mice (Fig. 9,C). Interestingly, there was only a 2- to 3-fold increase (p < 0.05) in the total number of F85-specific CD8 T cells at day 7 after RSV challenge in the lungs of DC-F85–93- and vacvG plus vacvF-immunized mice as compared with DC-no peptide- and vacvG plus vacvF-immunized mice (Fig. 9 D). These data demonstrate that F85-specific memory CD8 T cells are capable of inhibiting the development of pulmonary eosinophilia if present in high enough numbers at the time of or early after RSV challenge. Furthermore, these data provide strong evidence that both the total number and early kinetics of entry into the lung of RSV-specific memory CD8 T cells are critical for the control of RSV vaccine-enhanced pulmonary eosinophilia.

FIGURE 9.

Boosting the F85-specific memory CD8 T cell response results in inhibition of vacvG-induced pulmonary eosinophilia. BALB/c mice were immunized with 1 × 106 DC that were either left untreated or coated with F85–93 peptide. At 2–3 wk after DC immunization, mice were immunized with vacvβ-gal, vacvG plus vacvβ-gal, or vacvG plus vacvF. After an additional 3 wk, mice were bled and analyzed by IFN-γ ICS for the number of F85-specific CD8 T cells (A). Detection limits of <0.05% of PBL were below limit of detection (B.L.D.). Mice were subsequently challenged with RSV and at days 0, 3, and 7 after RSV challenge lungs and the total number of F85-specific CD8 T cells was determined by IFN-γ ICS (B). The total number of eosinophils in the BAL of DC-immunized and vacv-immunized mice was enumerated by quantitative morphometry at day 7 after RSV challenge (C). Levels at p > 0.05 were not significantly different (n.s.). The total number of F85-specific CD8 T cells in the lungs of all groups at day 7 after RSV challenge was enumerated by IFN-γ ICS (D). Error bar represents the SEM from n = 3–4 mice per group. These data represent two independent experiments. ∗, p < 0.05 as determined by ANOVA.

FIGURE 9.

Boosting the F85-specific memory CD8 T cell response results in inhibition of vacvG-induced pulmonary eosinophilia. BALB/c mice were immunized with 1 × 106 DC that were either left untreated or coated with F85–93 peptide. At 2–3 wk after DC immunization, mice were immunized with vacvβ-gal, vacvG plus vacvβ-gal, or vacvG plus vacvF. After an additional 3 wk, mice were bled and analyzed by IFN-γ ICS for the number of F85-specific CD8 T cells (A). Detection limits of <0.05% of PBL were below limit of detection (B.L.D.). Mice were subsequently challenged with RSV and at days 0, 3, and 7 after RSV challenge lungs and the total number of F85-specific CD8 T cells was determined by IFN-γ ICS (B). The total number of eosinophils in the BAL of DC-immunized and vacv-immunized mice was enumerated by quantitative morphometry at day 7 after RSV challenge (C). Levels at p > 0.05 were not significantly different (n.s.). The total number of F85-specific CD8 T cells in the lungs of all groups at day 7 after RSV challenge was enumerated by IFN-γ ICS (D). Error bar represents the SEM from n = 3–4 mice per group. These data represent two independent experiments. ∗, p < 0.05 as determined by ANOVA.

Close modal

One drawback to using the DC prime and boost methodology described in this study to increase the total number of F85-specific memory CD8 T cells in the lung is that the resultant cells are now secondary memory cells, as compared with primary memory cells generated after vacvG plus vacvF immunization. Consequently, the F85-specific secondary memory CD8 T cells may differ in their phenotypic or functional characteristics. To address this issue, we examined the functional and phenotypic characteristics of F85-specific memory CD8 T cells at day 21 after vacvG plus vacvF immunization of mice previously immunized with DC that were either left untreated or coated with F85–93 peptide. DC-no peptide-immunized mice exhibited a greater frequency of cells expressing CD62L but not TNF-α or IL-2 as compared with mice immunized with DC-F85–93 (Table I). However, after RSV challenge of mice previously immunized with either DC-no peptide and vacvG plus vacvF- or DC-F85–93 and vacvG plus vacvF, F85-specific primary and secondary memory CD8 T cells exhibited similar functional avidity (Fig. 10,A), degranulation, as determined by Lamp1/2 staining (Fig. 10,B) and similar expression of intracellular granzyme B (Fig. 10,C). However, there was a difference in the ability of these cells to produce the cytokines TNF-α and IL-2 (Fig. 10, D and E). These data suggest that although there are some phenotypical and functional differences in F85-specific cells in DC-no peptide- and DC-F85–93-immunized mice, these differences are relatively small and are unlikely to greatly alter their ability to inhibit the development of pulmonary eosinophilia.

Table I.

Phenotype of F85-specific CD8 T cells in DC-no peptide- and DC-F85–93-immunized mice after subsequent vacvG plus vacvF immunizationa

ImmunizationMarkerPercentage of Positive Cells of F85-Specificb CD8 T Cellsc
DC-no peptide → vacvG plus vacvF   
 IL-2 8.0 ± 4.9 
 TNF-α 59.8 ± 13.2 
 CD62L 32 ± 5.3d 
DC-F85–93 → vacvG plus vacvF   
 IL-2 2.8 ± 0.7 
 TNF-α 64.5 ± 1.9 
 CD62L 15.5 ± 1.3 
ImmunizationMarkerPercentage of Positive Cells of F85-Specificb CD8 T Cellsc
DC-no peptide → vacvG plus vacvF   
 IL-2 8.0 ± 4.9 
 TNF-α 59.8 ± 13.2 
 CD62L 32 ± 5.3d 
DC-F85–93 → vacvG plus vacvF   
 IL-2 2.8 ± 0.7 
 TNF-α 64.5 ± 1.9 
 CD62L 15.5 ± 1.3 
a

Mice were immunized with untreated DC (DC-no peptide) or DC coated with F85–93 peptide (DC-F85–93). Two weeks after DC immunization, mice were subsequently immunized with vacvG plus vacvF.

b

F85-specific CD8 T cells were identified in the peripheral blood 21 days after vacv immunization by IFN-γ ICS.

c

These data represent two individual experiments with n = 4 mice per group. Frequency of marker positive cells is expressed as the mean ± SE of the mean.

d

Significantly different (p < 0.05) than DC-F85–93 → vacvG plus vacvF-immunized mice.

FIGURE 10.

Function and phenotype of F85-specific memory CD8 T cells after RSV challenge of mice previously immunized with either DC-no peptide or DC-F85–93 followed by vacvG plus vacvF immunization. BALB/c mice were immunized with DC and subsequently immunized with vacvG plus vacvF as described in Fig. 9. At 3 wk after vacv immunization mice were challenged intranasally with RSV and at day 5 after RSV challenge the lungs were harvested. F85-specific CD8 T cells were identified by ICS for IFN-γ using P815 cells pulsed with 1 μM F85–93 peptide (at a ratio of 3:1 lung mononuclear cells to P815 cells) and assessed for their functional avidity (A), ability to degranulate as determined by Lamp1/2 staining (B), and expression of intracellular granzyme B (C). Levels at p > 0.05 were not significantly different (n.s.). The ability of F85-specific CD8 T cells to produce TNF-α (D) and IL-2 (E) was assessed by ICS. Data represent two individual experiments with n = 4 animals per group. Error bar represents the SEM. ∗, p < 0.05.

FIGURE 10.

Function and phenotype of F85-specific memory CD8 T cells after RSV challenge of mice previously immunized with either DC-no peptide or DC-F85–93 followed by vacvG plus vacvF immunization. BALB/c mice were immunized with DC and subsequently immunized with vacvG plus vacvF as described in Fig. 9. At 3 wk after vacv immunization mice were challenged intranasally with RSV and at day 5 after RSV challenge the lungs were harvested. F85-specific CD8 T cells were identified by ICS for IFN-γ using P815 cells pulsed with 1 μM F85–93 peptide (at a ratio of 3:1 lung mononuclear cells to P815 cells) and assessed for their functional avidity (A), ability to degranulate as determined by Lamp1/2 staining (B), and expression of intracellular granzyme B (C). Levels at p > 0.05 were not significantly different (n.s.). The ability of F85-specific CD8 T cells to produce TNF-α (D) and IL-2 (E) was assessed by ICS. Data represent two individual experiments with n = 4 animals per group. Error bar represents the SEM. ∗, p < 0.05.

Close modal

As demonstrated, M282-specific memory CD8 T cells in vacvG plus vacvM2-immunized mice are present at higher numbers early after RSV challenge as compared with F85-specific CD8 T cells in vacvG plus vacvF-immunized mice (Figs. 6 and 7). Additionally, these M282-specific CD8 T cells are able to alter the pulmonary chemokine environment, whereas the F85-specific CD8 T cells, that are present at a lower number, are not (Fig. 8). To determine whether increasing the total number of F85-specific CD8 T cells by DC-F85–93 and vacvG plus vacvF immunization (see Fig. 9) would also alter the pulmonary chemokine environment, DC- and vacv-immunized mice were harvested at day 3 after RSV challenge and subsequently analyzed for the protein levels of the chemokines CXCL10, CCL11, CCL17, and CCL22 by ELISA. Fig. 11 shows that, like before, a low number of F85-specific CD8 T cells in DC-no peptide- and vacvG plus vacvF-immunized mice were unable to decrease the protein levels of CCL17 and CCL22 in the lung as compared with DC-no peptide- and vacvG plus vacvβ-gal-immunized mice. However, in DC-F85–93- and vacvG plus vacvF-immunized mice, where there are a higher total number of F85-specific CD8 T cells, there are significant decreases (p < 0.05) in the levels of CCL17 and CCL22 protein in the lung (Fig. 11). Interestingly, as observed with M282-specific CD8 T cells in vacvG plus vacvM2-immunized mice (see Fig. 8), there are no significant decreases (p > 0.05) in protein levels of CXCL10 or CCL11 in DC-F85–93 and vacvG plus vacvF-immunized mice (Fig. 11). These data strongly suggest that a critical number of RSV-specific memory CD8 T cells are required to alter the pulmonary chemokine environment and prevent the development of pulmonary eosinophilia.

FIGURE 11.

F85-specific memory CD8 T cells alter the pulmonary chemokine environment if present at sufficient numbers early after RSV challenge. BALB/c mice were immunized with 1 × 106 DC that were either left untreated or coated with F85–93 peptide as described in Fig. 9. Two weeks after DC immunization, mice were subsequently immunized with vacvβ-gal, vacvG plus vacvβ-gal, and vacvG plus vacvF and were challenged with RSV 3 wk after vacv immunization. Protein levels of CXCL10, CCL11, CCL17, and CCL22 were measured by ELISA from lung homogenates at day 3 after RSV challenge. These data represent pooled data from three individual experiments for n = 9–12 mice per group. Error bar represents the SEM. ∗, p < 0.05 comparing vacvG plus vacvβ-gal-immunized and vacvG plus vacvF-immunized mice as determined by ANOVA.

FIGURE 11.

F85-specific memory CD8 T cells alter the pulmonary chemokine environment if present at sufficient numbers early after RSV challenge. BALB/c mice were immunized with 1 × 106 DC that were either left untreated or coated with F85–93 peptide as described in Fig. 9. Two weeks after DC immunization, mice were subsequently immunized with vacvβ-gal, vacvG plus vacvβ-gal, and vacvG plus vacvF and were challenged with RSV 3 wk after vacv immunization. Protein levels of CXCL10, CCL11, CCL17, and CCL22 were measured by ELISA from lung homogenates at day 3 after RSV challenge. These data represent pooled data from three individual experiments for n = 9–12 mice per group. Error bar represents the SEM. ∗, p < 0.05 comparing vacvG plus vacvβ-gal-immunized and vacvG plus vacvF-immunized mice as determined by ANOVA.

Close modal

BALB/c mice immunized with vacvF do not develop pulmonary eosinophilia after RSV challenge (19, 20). However, IFN-γ-deficient mice immunized with vacvF develop pulmonary eosinophilia after RSV challenge (19, 20). These data suggest that F85-specific CD8 T cells inhibit vacvF-induced pulmonary eosinophilia by an IFN-γ-dependent mechanism. We have previously demonstrated that M282-specific memory CD8 T cells inhibit vacvG-induced pulmonary eosinophilia via an IFN-γ-independent mechanism (7). Thus, it appears that there are differences in the requirements for IFN-γ based on the CD8 T cell population responding to the challenge as well as the immunization (i.e., either vacvG or vacvF). Therefore, we asked whether M282-specific memory CD8 T cells could inhibit vacvF-induced pulmonary eosinophilia in the absence of IFN-γ. BALB/c IFN-γ-deficient mice were immunized with vacvβ-gal, vacvF plus vacvβ-gal, vacvM2 plus vacvβ-gal, or vacvF plus vacvM2 and 3 wk later intranasally challenged with RSV. Fig. 12 demonstrates that M282-specific memory CD8 T cells decrease the total number of eosinophils in the BAL at day 7 after RSV challenge (Fig. 12,A). Again, as depicted in Fig. 12 B, there are a similar total number of RSV-specific CD8 T cells in mice immunized with vacvF plus vacvβ-gal, vacvM2 plus vacvβ-gal, or vacvF plus vacvM2 at day 7 after RSV challenge. These data suggest that M282-specific memory CD8 T cells can inhibit pulmonary eosinophilia independent of IFN-γ and the manner of immunization.

FIGURE 12.

M282-specific memory CD8 T cells inhibit vacvF-induced pulmonary eosinophilia in the absence of IFN-γ. IFN-γ-deficient BALB/c mice were immunized with vacvβ-gal, vacvF plus vacvβ-gal, vacvM2 plus vacvβ-gal, and vacvF plus vacvM2. At least 21 days later mice were intranasally challenged with RSV. The total number of eosinophils (A) in the BAL at day 7 after RSV challenge was enumerated by quantitative morphometry. B, The total number of F85- and M282-specific CD8 T cells were quantified by ICS for TNF-α. Error bar represents the SEM from n = 3–4 mice per group. These data represent three independent experiments. Detection limits at <1 × 104 cells were below limit of detection (B.L.D.). ∗, p < 0.05.

FIGURE 12.

M282-specific memory CD8 T cells inhibit vacvF-induced pulmonary eosinophilia in the absence of IFN-γ. IFN-γ-deficient BALB/c mice were immunized with vacvβ-gal, vacvF plus vacvβ-gal, vacvM2 plus vacvβ-gal, and vacvF plus vacvM2. At least 21 days later mice were intranasally challenged with RSV. The total number of eosinophils (A) in the BAL at day 7 after RSV challenge was enumerated by quantitative morphometry. B, The total number of F85- and M282-specific CD8 T cells were quantified by ICS for TNF-α. Error bar represents the SEM from n = 3–4 mice per group. These data represent three independent experiments. Detection limits at <1 × 104 cells were below limit of detection (B.L.D.). ∗, p < 0.05.

Close modal

CD8 T cells play a prominent role in the clearance of intracellular pathogens. In addition, CD8 T cells also have been shown to inhibit pathological CD4 T cell responses (7, 14, 27, 28). Previous work has established that RSV-specific CD8 T cells can inhibit the development of pulmonary eosinophilia during RSV challenge of previously immunized mice (14, 20). We have recently demonstrated that induction of an RSV M282-specific memory CD8 T cell response inhibits both the total number of Th2 cells and eosinophils in the lung after RSV challenge of mice previously immunized with either vacvG- or FI-RSV (7). Because the presence of Th2 cells in the lung is required to induce pulmonary eosinophilia in this model (29), these results suggest that RSV-specific memory CD8 T cells inhibit the trafficking or expansion of Th2 cells. In this study, we show that the ability of RSV-specific memory CD8 T cells to inhibit RSV vaccine-enhanced pulmonary eosinophilia is critically dependent on the total number of virus-specific memory CD8 T cells in the lung early after infection. Simply decreasing the number of M282-specific CD8 T cells present in the lungs early after RSV challenge (day 3) results in the development of pulmonary eosinophilia (Fig. 7). Moreover, we demonstrate that memory CD8 T cells specific to the subdominant F85–93 epitope, if present in high enough numbers early after RSV challenge, also inhibit the development of pulmonary eosinophilia in vacvG plus vacvF-immunized mice (Fig. 9). However, the underlying mechanisms of how these early memory CD8 T cell responses prevent the Th2-driven eosinophilic response remain unclear.

Chemokines are critical for trafficking of immune cells to sites of inflammation. Murine eosinophils and CD4 T cells have been shown to express different chemokine receptors based on their localization or functional phenotype (30). Our data indicates that the presence of a high number of RSV-specific memory CD8 T cells in the lung early after RSV challenge alters the chemokine environment in the lung (Figs. 8 and 11). Interestingly, we observed a preferential decrease in the protein levels of CCL17 and CCL22 and no change in the levels of CCL11 (Figs. 8 and 11). CCL11, CCL17, and CCL22 have all been shown to play a role in the recruitment of Th2 cells and eosinophils that express their cognate receptors (CCR4 for CCL17 and CCL22, CCR3 for CCL11) (30, 31, 32, 33). Matthews et al. (23) demonstrated that depletion of CCL11 in vacvG-immunized mice reduces the total number of eosinophils in the lung after RSV challenge. Importantly, the reduction of eosinophils in CCL11-depleted mice is ∼2-fold as compared with mock-treated mice, suggesting that other Th2-associated chemokines may compensate and facilitate the trafficking of eosinophils into the lung. These data are similar to allergic models of airway disease in which CCL11-deficient mice display only a partially decreased number of eosinophils in the lung, again suggesting roles for other Th2-associated chemokines (i.e., CCL17, CCL22) in the onset of eosinophilic disease (34, 35).

Interestingly, whereas the absence of CCL11 in asthma models only partially reduces airway eosinophilia, depletion of either CCL17 or CCL22 completely abrogates pulmonary eosinophilia induced by repeated allergen stimulation (34, 35, 36, 37). Further studies concluded that depletion of CCL22 dramatically reduced the trafficking of Th2 cells into the lung of allergen challenged mice (32). These data are consistent with our previous findings that vacvG-immunized mice undergoing a concurrent M282-specific memory CD8 T cell response have a similar number of RSV-specific Th1 cells at day 5 after RSV challenge but have a significantly decreased total number of RSV-specific Th2 cells at the same time point (7). Taken together, these data suggest that the modulation of CCL17 and CCL22 chemokine levels by RSV-specific memory CD8 T cells may alter the trafficking of both Th2 cells and eosinophils into the lungs after RSV challenge of vacvG-immunized mice.

It is currently unclear how memory CD8 T cells inhibit the production of CCL17 and CCL22 in the lungs early after RSV challenge of vacvG-immunized mice. It is unlikely that accelerating viral clearance is responsible for this early decrease in chemokine production as the kinetics of viral clearance is similar in vacvG plus vacvF- and vacvG plus vacvM2-immunized mice (Fig. 3). Hussell et al. (20) demonstrated that vacvF-immunized mice develop pulmonary eosinophilia after Ab-mediated depletion of either CD8 T cells or IFN-γ, suggesting that IFN-γ may be required for CD8 T cell inhibition of RSV vaccine-enhanced pulmonary eosinophilia. This concept is attractive because of the known inhibitory effects of IFN-γ on Th2 cells (38, 39, 40). Furthermore, IFN-γ has been previously demonstrated to inhibit the production of CCL17 or CCL22 (41, 42, 43). However, we have previously shown that RSV M282-specific memory CD8 T cells inhibit the development of pulmonary eosinophilia after RSV challenge of vacvG plus vacvM2-immunized mice in the absence of IFN-γ (7). We have also demonstrated in this study that M282-specific memory CD8 T cells inhibit the development of pulmonary eosinophilia after RSV challenge of IFN-γ-deficient mice that have been previously immunized with vacvF (Fig. 12). One potential alternative is that M282-specific CD8 T cells may eliminate CCL17- or CCL22-producing cells via a perforin- or Fas ligand-dependent mechanism; however, this alternative has not been directly tested.

RSV vaccine-enhanced disease is a prime example of CD4 T cell-mediated immunopathology. Our results suggest that immunization that induces both RSV-specific CD4 and CD8 T cell responses provides maximal resistance to viral infection as well as protection from the development of Th2 cell-driven pulmonary eosinophilia. Data presented in this study strongly suggests that memory CD8 T cells modulate disease-causing Th2 responses by altering Th2-associated chemokines, specifically CCL17 and CCL22, and thus the trafficking of Th2 cells, and possibly eosinophils, into the lung. CCL17 and CCL22 are also induced after RSV infection of lung epithelial cells, suggesting that these CCR4 ligands may potentially play a role in disease after acute RSV infection (25, 44). Our results highlight the critical role chemokines play in shaping the inflammatory response that enters the lung following RSV infection and how modulation of this inflammation can alter the development of immunopathology.

We thank John Harty and Stanley Perlman for helpful discussions.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by Grant R01 AI-063520 from the National Institutes of Health (to S.M.V.).

3

Abbreviations used in this paper: RSV, respiratory syncytial virus; FI-RSV, formalin inactivated-RSV; vacv, vaccinia virus; G, RSV attachment protein; F, RSV fusion protein; ICS, intracellular cytokine staining; BAL, bronchoalveolar lavage; DC, dendritic cell; β-gal, β-galactosidase; MFI, mean fluorescence intensity.

1
Heilman, C. A..
1990
. National Institute of Allergy and Infectious Diseases and the World Health Organization: respiratory syncytial and parainfluenza viruses.
J. Infect. Dis.
161
:
402
-406.
2
Falsey, A. R., P. A. Hennessey, M. A. Formica, C. Cox, E. E. Walsh.
2005
. Respiratory syncytial virus infection in elderly and high-risk adults.
N. Engl. J. Med.
352
:
1749
-1759.
3
Falsey, A. R., E. E. Walsh.
2005
. Respiratory syncytial virus infection in elderly adults.
Drugs Aging
22
:
577
-587.
4
Fulginiti, V. A., J. J. Eller, O. F. Sieber, J. W. Joyner, M. Minamitani, G. Meiklejohn.
1969
. Respiratory virus immunization. I. A field trial of two inactivated respiratory virus vaccines; an aqueous trivalent parainfluenza virus vaccine and an alum-precipitated respiratory syncytial virus vaccine.
Am. J. Epidemiol.
89
:
435
-448.
5
Kapikian, A. Z., R. H. Mitchell, R. M. Chanock, R. A. Shvedoff, C. E. Stewart.
1969
. An epidemiologic study of altered clinical reactivity to respiratory syncytial (RS) virus infection in children previously vaccinated with an inactivated RS virus vaccine.
Am. J. Epidemiol.
89
:
405
-421.
6
Kim, H. W., J. G. Canchola, C. D. Brandt, G. Pyles, R. M. Chanock, K. Jensen, R. H. Parrott.
1969
. Respiratory syncytial virus disease in infants despite prior administration of antigenic inactivated vaccine.
Am. J. Epidemiol.
89
:
422
-434.
7
Olson, M. R., S. M. Varga.
2007
. CD8 T cells inhibit respiratory syncytial virus (RSV) vaccine-enhanced disease.
J. Immunol.
179
:
5415
-5424.
8
Varga, S. M., E. L. Wissinger, T. J. Braciale.
2000
. The attachment (G) glycoprotein of respiratory syncytial virus contains a single immunodominant epitope that elicits both Th1 and Th2 CD4+ T cell responses.
J. Immunol.
165
:
6487
-6495.
9
Johnson, T. R., J. E. Johnson, S. R. Roberts, G. W. Wertz, R. A. Parker, B. S. Graham.
1998
. Priming with secreted glycoprotein G of respiratory syncytial virus (RSV) augments interleukin-5 production and tissue eosinophilia after RSV challenge.
J. Virol.
72
:
2871
-2880.
10
Johnson, T. R., M. N. Teng, P. L. Collins, B. S. Graham.
2004
. Respiratory syncytial virus (RSV) G glycoprotein is not necessary for vaccine-enhanced disease induced by immunization with formalin-inactivated RSV.
J. Virol.
78
:
6024
-6032.
11
Johnson, T. R., S. M. Varga, T. J. Braciale, B. S. Graham.
2004
. Vβ14+ T cells mediate the vaccine-enhanced disease induced by immunization with respiratory syncytial virus (RSV) G glycoprotein but not with formalin-inactivated RSV.
J. Virol.
78
:
8753
-8760.
12
Openshaw, P. J., S. L. Clarke, F. M. Record.
1992
. Pulmonary eosinophilic response to respiratory syncytial virus infection in mice sensitized to the major surface glycoprotein G.
Int. Immunol.
4
:
493
-500.
13
Waris, M. E., C. Tsou, D. D. Erdman, S. R. Zaki, L. J. Anderson.
1996
. Respiratory synctial virus infection in BALB/c mice previously immunized with formalin-inactivated virus induces enhanced pulmonary inflammatory response with a predominant Th2-like cytokine pattern.
J. Virol.
70
:
2852
-2860.
14
Srikiatkhachorn, A., T. J. Braciale.
1997
. Virus-specific CD8+ T lymphocytes downregulate T helper cell type 2 cytokine secretion and pulmonary eosinophilia during experimental murine respiratory syncytial virus infection.
J. Exp. Med.
186
:
421
-432.
15
Chang, J., A. Srikiatkhachorn, T. J. Braciale.
2001
. Visualization and characterization of respiratory syncytial virus F-specific CD8+ T cells during experimental virus infection.
J. Immunol.
167
:
4254
-4260.
16
Lee, S., S. A. Miller, D. W. Wright, M. T. Rock, J. E. Crowe, Jr.
2007
. Tissue-specific regulation of CD8+ T-lymphocyte immunodominance in respiratory syncytial virus infection.
J. Virol.
81
:
2349
-2358.
17
Openshaw, P. J., R. M. Pemberton, L. A. Ball, G. W. Wertz, B. A. Askonas.
1988
. Helper T cell recognition of respiratory syncytial virus in mice.
J. Gen. Virol.
69
: (Pt. 2):
305
-312.
18
Pemberton, R. M., M. J. Cannon, P. J. Openshaw, L. A. Ball, G. W. Wertz, B. A. Askonas.
1987
. Cytotoxic T cell specificity for respiratory syncytial virus proteins: fusion protein is an important target antigen.
J. Gen. Virol.
68
: (Pt. 8):
2177
-2182.
19
Castilow, E. M., M. R. Olson, D. K. Meyerholz, S. M. Varga.
2008
. Differential role of gamma interferon in inhibiting pulmonary eosinophilia and exacerbating systemic disease in fusion protein-immunized mice undergoing challenge infection with respiratory syncytial virus.
J. Virol.
82
:
2196
-2207.
20
Hussell, T., C. J. Baldwin, A. O'Garra, P. J. Openshaw.
1997
. CD8+ T cells control Th2-driven pathology during pulmonary respiratory syncytial virus infection.
Eur. J. Immunol.
27
:
3341
-3349.
21
Peebles, R. S., Jr, J. R. Sheller, R. D. Collins, K. Jarzecka, D. B. Mitchell, B. S. Graham.
2000
. Respiratory syncytial virus (RSV)-induced airway hyperresponsiveness in allergically sensitized mice is inhibited by live RSV and exacerbated by formalin-inactivated RSV.
J. Infect. Dis.
182
:
671
-677.
22
Varga, S. M., X. Wang, R. M. Welsh, T. J. Braciale.
2001
. Immunopathology in RSV infection is mediated by a discrete oligoclonal subset of antigen-specific CD4+ T cells.
Immunity
15
:
637
-646.
23
Matthews, S. P., J. S. Tregoning, A. J. Coyle, T. Hussell, P. J. Openshaw.
2005
. Role of CCL11 in eosinophilic lung disease during respiratory syncytial virus infection.
J. Virol.
79
:
2050
-2057.
24
Castilow, E. M., D. K. Meyerholz, S. M. Varga.
2008
. IL-13 is required for eosinophil entry into the lung during respiratory syncytial virus vaccine-enhanced disease.
J. Immunol.
180
:
2376
-2384.
25
Monick, M. M., L. S. Powers, I. Hassan, D. Groskreutz, T. O. Yarovinsky, C. W. Barrett, E. M. Castilow, D. Tifrea, S. M. Varga, G. W. Hunninghake.
2007
. Respiratory syncytial virus synergizes with Th2 cytokines to induce optimal levels of TARC/CCL17.
J. Immunol.
179
:
1648
-1658.
26
Culley, F. J., A. M. Pennycook, J. S. Tregoning, T. Hussell, P. J. Openshaw.
2006
. Differential chemokine expression following respiratory virus infection reflects Th1- or Th2-biased immunopathology.
J. Virol.
80
:
4521
-4527.
27
Muller, D., B. H. Koller, J. L. Whitton, K. E. LaPan, K. K. Brigman, J. A. Frelinger.
1992
. LCMV-specific, class II-restricted cytotoxic T cells in β2-microglobulin-deficient mice.
Science
255
:
1576
-1578.
28
Quinn, D. G., A. J. Zajac, J. A. Frelinger, D. Muller.
1993
. Transfer of lymphocytic choriomeningitis disease in β2-microglobulin-deficient mice by CD4+ T cells.
Int. Immunol.
5
:
1193
-1198.
29
Walzl, G., S. Matthews, S. Kendall, J. C. Gutierrez-Ramos, A. J. Coyle, P. J. Openshaw, T. Hussell.
2001
. Inhibition of T1/ST2 during respiratory syncytial virus infection prevents T helper cell type 2 (Th2)- but not Th1-driven immunopathology.
J. Exp. Med.
193
:
785
-792.
30
Lukacs, N. W., S. H. Oliveira, C. M. Hogaboam.
1999
. Chemokines and asthma: redundancy of function or a coordinated effort?.
J. Clin. Invest.
104
:
995
-999.
31
Sallusto, F., A. Lanzavecchia, C. R. Mackay.
1998
. Chemokines and chemokine receptors in T-cell priming and Th1/Th2-mediated responses.
Immunol. Today
19
:
568
-574.
32
Lloyd, C. M., T. Delaney, T. Nguyen, J. Tian, A. C. Martinez, A. J. Coyle, J. C. Gutierrez-Ramos.
2000
. CC chemokine receptor (CCR)3/eotaxin is followed by CCR4/monocyte-derived chemokine in mediating pulmonary T helper lymphocyte type 2 recruitment after serial antigen challenge in vivo.
J. Exp. Med.
191
:
265
-274.
33
Liu, L. Y., N. N. Jarjour, W. W. Busse, E. A. Kelly.
2003
. Chemokine receptor expression on human eosinophils from peripheral blood and bronchoalveolar lavage fluid after segmental antigen challenge.
J. Allergy Clin. Immunol.
112
:
556
-562.
34
Rothenberg, M. E., J. A. MacLean, E. Pearlman, A. D. Luster, P. Leder.
1997
. Targeted disruption of the chemokine eotaxin partially reduces antigen-induced tissue eosinophilia.
J. Exp. Med.
185
:
785
-790.
35
Yang, Y., J. Loy, R. P. Ryseck, D. Carrasco, R. Bravo.
1998
. Antigen-induced eosinophilic lung inflammation develops in mice deficient in chemokine eotaxin.
Blood
92
:
3912
-3923.
36
Gonzalo, J. A., Y. Pan, C. M. Lloyd, G. Q. Jia, G. Yu, B. Dussault, C. A. Powers, A. E. Proudfoot, A. J. Coyle, D. Gearing, J. C. Gutierrez-Ramos.
1999
. Mouse monocyte-derived chemokine is involved in airway hyperreactivity and lung inflammation.
J. Immunol.
163
:
403
-411.
37
Kawasaki, S., H. Takizawa, H. Yoneyama, T. Nakayama, R. Fujisawa, M. Izumizaki, T. Imai, O. Yoshie, I. Homma, K. Yamamoto, K. Matsushima.
2001
. Intervention of thymus and activation-regulated chemokine attenuates the development of allergic airway inflammation and hyperresponsiveness in mice.
J. Immunol.
166
:
2055
-2062.
38
Iwamoto, I., H. Nakajima, H. Endo, S. Yoshida.
1993
. Interferon gamma regulates antigen-induced eosinophil recruitment into the mouse airways by inhibiting the infiltration of CD4+ T cells.
J. Exp. Med.
177
:
573
-576.
39
Cohn, L., R. J. Homer, N. Niu, K. Bottomly.
1999
. T helper 1 cells and interferon γ regulate allergic airway inflammation and mucus production.
J. Exp. Med.
190
:
1309
-1318.
40
Suzuki, M., K. Maghni, S. Molet, A. Shimbara, Q. A. Hamid, J. G. Martin.
2002
. IFN-gamma secretion by CD8 T cells inhibits allergen-induced airway eosinophilia but not late airway responses.
J. Allergy Clin. Immunol.
109
:
803
-809.
41
Su, S. B., R. S. Grajewski, D. Luger, R. K. Agarwal, P. B. Silver, J. Tang, J. Tuo, C. C. Chan, R. R. Caspi.
2007
. Altered chemokine profile associated with exacerbated autoimmune pathology under conditions of genetic interferon-γ deficiency.
Invest. Ophthalmol. Vis. Sci.
48
:
4616
-4625.
42
Hino, R., T. Shimauchi, Y. Tokura.
2005
. Treatment with IFN-γ increases serum levels of Th1 chemokines and decreases those of Th2 chemokines in patients with mycosis fungoides.
J. Dermatol. Sci.
38
:
189
-195.
43
Jakubzick, C., H. Wen, A. Matsukawa, M. Keller, S. L. Kunkel, C. M. Hogaboam.
2004
. Role of CCR4 ligands, CCL17 and CCL22, during Schistosoma mansoni egg-induced pulmonary granuloma formation in mice.
Am. J. Pathol.
165
:
1211
-1221.
44
Zhang, Y., B. A. Luxon, A. Casola, R. P. Garofalo, M. Jamaluddin, A. R. Brasier.
2001
. Expression of respiratory syncytial virus-induced chemokine gene networks in lower airway epithelial cells revealed by cDNA microarrays.
J. Virol.
75
:
9044
-9058.