Natural killer T cell sensitization during neonatal respiratory syncytial virus infection induces eosinophilic lung disease in re-infected adult mice

Respiratory syncytial virus (RSV) is a major viral pathogen that causes severe lower respiratory tract infections in infants and the elderly worldwide. Infants with severe RSV bronchiolitis tend to experience more wheezing and asthma in later childhood. Because invariant natural killer T (iNKT) cells are associated with the asthma pathology, we investigated whether neonatal iNKT cells are involved in the aggravation of pulmonary diseases following RSV infection in mice. Intranasal exposure to the iNKT cell ligand α-galactosylceramide (α-GC) with RSV primary infection in neonatal mice elicited neither cytokine production (except for a slight increase of IL-5) nor pulmonary eosinophilia, despite the presence of both CD1d+ cells and NKT cells. Interestingly, in adult mice re-infected with RSV, neonatal iNKT cell sensitization by α-GC during RSV primary infection resulted in much higher levels of pulmonary Th2 cytokines and elevated eosinophilia with airway hyperresponsiveness, whereas this was not observed in cd1d knockout mice. In contrast, α-GC priming of adults during RSV re-infection did not induce more severe airway symptoms than RSV re-infection in the absence of α-GC. α-GC co-administration during RSV primary infection facilitated RSV clearance regardless of age, but viral clearance following re-infection was not iNKT cell-dependent. This study clearly demonstrates that RSV-induced immune responses can be altered by iNKT cells, suggesting that neonatal iNKT cell sensitization during RSV primary infection is associated with exacerbation of pulmonary diseases following RSV re-infection in adulthood.


Introduction
Respiratory syncytial virus (RSV) is a negative-sense single-stranded RNA virus. Most people are infected with RSV at least once by age 2 and then infected again later in life [1]. In healthy adults, RSV infection usually induces mild symptoms. However, in infants, elderly over the age of 65 years, and immunocompromised persons, RSV is a cause of morbidity and mortality a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 [Chungbuk], Samtako Bio Korea, [Osan]). C.129S2-Cd1 tm1Gru /J (CD1d knockout; CD1d KO) mice were purchased from Jackson Laboratory (Orient Bio). Mice were kept under specific pathogen-free conditions at Animal Research Facility in the International Vaccine Institute. Mice were provided with water and food ad libitum. All efforts were made to minimize suffering and the number of mice for research. The treated mice were monitored daily as part of the approved protocol. Until the end of the experiments, no mice died as a result of RSV infection. However, two mice died of unknown causes during i.n. administration of solutions under anesthesia. No mice met IVI's specific humane endpoint criteria for euthanasia (weight loss, decrease in appetite, weakness/inability to obtain feed or water, moribund state, unrelieved pain/distress, and organ dysfunction/failure etc.) during the experiments. At the end of each experimental time point, animals were sacrificed by cervical dislocation or CO 2 inhalation.

Analysis of cells and cytokines in bronchoalveolar lavage (BAL) fluid
Four days after challenge with RSV A2, mice were sacrificed, and adult and neonate tracheas were cannulated and washed with 700 and 300 μL of PBS, respectively. After BAL fluid was centrifuged, supernatant was stored at -80˚C until analysis. Cells were incubated with violet fluorescent live-dead discriminator (Invitrogen, Eugene, OR) for 10 min at room temperature and then washed with 1 mL of PBS and blocked for 5 min with purified CD16/CD32 Fc (clone 2.4G2; BD Pharmingen, San Jose, CA). After blocking, 50 μL of antibody cocktail containing anti-CD45-APC (clone 30-F11), CD11c-FITC (clone HL3), Ly-6G (Gr-1)-PE-Cy7 (clone: 1A8), and Siglec F-PE (clone E50-2440; all from BD Pharmingen) were added to cells and incubated at 4˚C for 30 min. Cells were subsequently washed two times with PBS (2% FBS) and fixed with 200 μL of paraformaldehyde. Cells were analyzed by using a BD FACS LSR II flow cytometer and data were analyzed with FlowJo software (version 10; Tree Star, Ashland, OR). The cytokines in BAL supernatant was measured using the mouse Th1/Th2/Th17 BD Cytometric Bead Assay Kit (BD Biosciences, San Jose, CA) according to the manufacturer's recommendations. The cytokine levels were also analyzed using the Mouse Magnetic Luminex Screening Assay (R&D, Minneapolis, MN) [26].

Analysis of cells in lung
Four days after RSV A2 challenge, neonatal and adult mice were sacrificed and their lungs harvested. Lungs were chopped and incubated with lung enzyme cocktail including 10% RPMI (Sigma, St. Louis, MO), collagenase D (Roche, Mannheim, Germany), and DNase1 (Roche) for 30 min at 37˚C with stirring. After incubation, lung homogenates were transferred to a 70μm strainer on a 50-mL conical tube and cells were analyzed by flow cytometry as described above. Anti-CD1d-APC (clone 1B1; eBioscience, San Diego, CA), CD3ε-PerCP-Cy5.5 (clone 145-2C11; BD Pharmingen), and CD49b-PE (clone DX5; BD Pharmingen) were used for lung cell staining.

Lung RSV detection
Lung tissues were washed by vascular perfusion with PBS containing heparin (10 U/mL) and then homogenized by passing through a 70-μm cell strainer (BD Labware, Franklin Lakes, NJ) with 2 mL of MEM (10% FBS). Lung homogenates were centrifuged at 300 ×g for 5 min and 100 μL of supernatants was inoculated into 90% confluent HEp-2 cells in 6-well plates. After incubation with an agar overlay for 5 days, each well was stained with 0.1% crystal violet before plaques were counted to determine the PFU/mL.

Measurement of airway responsiveness
The methacholine challenge test was used to evaluate AHR, a hallmark of asthma. Four days after RSV challenge, mice received nebulized methacholine (0 and 10 mg/mL) for 3 min and enhanced pause (Penh) was recorded for 3 min using whole-body plethysmography (OCP 3000, Allmedicus, Korea). Penh at 10 mg/mL was expressed as the value obtained from each mouse subtracted from the value for the nebulized PBS inhalation control (0 mg/mL).

Statistical analysis
Data were analyzed using Prism software (version 5; GraphPad, La Jolla, CA) and expressed as mean ± SEM. Statistical significance was determined by using an unpaired, two-tailed Student t test. P values less than 0.05 were considered statistically significant.

α-GC does not cause eosinophil recruitment or induce cytokine secretion in neonatal mice but does in adult mice
We investigated the effect of iNKT cells on RSV-induced immune responses by age. First, neonatal (7-day-old) or adult (6-week-old) BALB/c wild-type mice were i.n. administered with vehicle (dimethyl sulfoxide and phosphate buffer saline), α-GC, RSV, or α-GC+RSV. Four days later, we examined eosinophil infiltration in BAL fluid. In adult mice, eosinophils were found in BAL fluid in the α-GC and α-GC+RSV treatment groups ( Fig 1B, S1B Fig); however, α-GC and RSV had no effect on the recruitment of eosinophils in BAL fluid of neonatal mice ( Fig 1A, S1A Fig). Eosinophil recruitment corresponded with H&E staining of lungs ( Fig 1A  and 1B). In adults, cell infiltration and mucus production were observed in both the α-GC and α-GC+RSV treatment groups ( Fig 1B), but no neonate treatment group showed additional cell infiltration or mucus production in lungs ( Fig 1A). Administration with neither α-GC nor RSV affected production of pulmonary cytokines (e.g., IFN-γ, IL-4, and IL-13) in neonates. There was a slight increase of IL-5, but the difference was not significant between the α-GC, RSV, and α-GC+RSV groups. In contrast, in adults, both α-GC and α-GC+RSV increased IFNγ, IL-4, IL-5, and IL-13 levels (Fig 2). In adult mice, α-GC+RSV administration elevated IL-4 production more than administration of α-GC alone (p < 0.05). IL-17 was below the detection level of the assay in all conditions. Collectively, these results show that in neonatal mice neither α-GC nor RSV affects eosinophil recruitment, mucus production, or cytokine production, with the exception of IL-5.
Neonatal NKT cells are unresponsive to α-GC administration despite the presence of CD1d-positive cells We assumed that unresponsiveness of neonatal iNKT cells to i.n. administration of α-GC might result from the lack of CD1d-expressing cells, which present α-GC to iNKT cells; however, both neonatal ( Fig 3A, S2A Fig) and adult (Fig 3B, S2B Fig) mice had similar ratios and absolute numbers of CD1d-expressing cell populations. In addition, CD3 + CD49b + NKT cells exist in neonatal mice, although in our study, the ratio of neonatal NKT cells remained consistent, regardless of α-GC or RSV administration (Fig 3A, S2A Fig). In contrast, in adult mice, the ratio and absolute numbers of NKT cells increased after α-GC administration (p < 0.001 for vehicle vs. α-GC; p < 0.05 for vehicle vs. α-GC+RSV) (Fig 3B). Collectively, these results suggest that neonatal mice have sufficient numbers of CD1d-expressing cells; however, unlike adult mice, they lack the ability to increase their NKT cell ratios, recruit eosinophils, and produce cytokines in response to α-GC.
iNKT cell ligand co-exposure during RSV primary infection in neonatal mice elevates adult immune responses after RSV re-infection We further examined whether neonatal NKT cell sensitization by α-GC affects immune responses in adults following RSV re-infection. Neonatal mice were administered vehicle, α-GC, RSV, or α-GC+RSV, and challenged with RSV at 8 weeks of age. There were increased numbers of eosinophils in the BAL fluid of groups exposed to RSV or α-GC+RSV in wild-type neonates. The number of infiltrated eosinophils was highest in the α-GC+RSV group. However, this pulmonary eosinophilia was abrogated in CD1d KO mice, suggesting that eosinophil infiltration following RSV re-infection is CD1d-dependent (Fig 5). In addition, lung histology showed similar CD1dmediated cell recruitment ( Fig 6A) and mucus production (Fig 6B) in both RSV and α-GC+RSV treatment groups. Th2 cytokines such as IL-4, IL-5, and IL-13 were increased most when mice were administered α-GC+RSV as neonates; IFN-γ levels were comparable in the RSV and α-GC +RSV groups (Fig 6C). CD1d KO mice produced much lower levels of cytokines than wild-type mice, and the differences between the RSV and α-GC+RSV groups were not statistically significant (Fig 6C).
We next examined whether elevated lung eosinophil infiltration and Th2 cytokine production were correlated with severe AHR. Methacholine testing 4 days after RSV re-infection showed that α-GC priming during RSV primary infection led to more severe AHR in neonatal mice than priming with RSV alone (Fig 7A). Severe AHR was not detected in the CD1d KO mice (Fig 7A). In contrast, when wild-type and CD1d KO mice were re-infected with RSV in the absence of α-GC, the virus was cleared (Fig 7B). These results suggest that RSV itself elicited CD1d-independent acquired immune protection against RSV re-infection.
Collectively, these results suggest that neonatal sensitization of iNKT cells by α-GC does not immediately induce an immune response, but that sensitized NKT cells elevate eosinophil recruitment, Th2 cytokine production, and AHR following RSV re-infection in adulthood. However, NKT cell sensitization in neonatal mice does not affect lung viral clearance following RSV re-infection in adulthood.

Adult NKT cell stimulation during RSV re-infection does not aggravate pulmonary eosinophilia
To examine whether adult iNKT cell stimulation during RSV re-infection also plays a role in the exacerbation of lung disease, wild-type mice were re-infected with RSV in the presence of α-GC following neonatal primary infection. The RSV re-infection groups showed significantly Role of neonatal NKT cell sensitization in RSV disease

Fig 6. Effects of neonatal α-GC sensitization with RSV primary infection on pulmonary immune responses following re-infection in adulthood.
Wild-type or CD1d KO neonatal mice were i.n. administered PBS, α-GC, RSV, or α-GC+ RSV and higher levels of eosinophils in BAL fluid than the group of neonates primed with RSV but administered α-GC only in adulthood (p < 0.01, RSV vs. α-GC). There was no significant difference in pulmonary eosinophilia between the two groups of mice re-infected with RSV, regardless of whether α-GC was co-administered ( Fig 8A, S3 Fig). Histology data also showed challenged at 8 weeks of age with RSV. Lung tissue was formalin-fixed, paraffin-embedded, and stained with hematoxylineosin (H&E) (A) or periodic acid-Schiff (PAS) (B) on day4 after RSVA2 challenge. (C) Cytokine levels in BAL fluid were determined by luminex assay. Data are expressed as mean ± SEM of 5 or 6 mice per group. *p<0.05.
https://doi.org/10.1371/journal.pone.0176940.g006 similar levels of cell infiltration and mucus production after RSV re-infection, regardless of α-GC priming in adulthood (Fig 8B). Similarly, the higher IFN-γ levels in BAL fluid were comparable in the two groups, whether or not α-GC was co-administered, but IL-4 and IL-13 levels were much higher in the group stimulated with α-GC during RSV re-infection. Of interest, α-GC priming during RSV re-infection led to decreased levels of IL-5, relative to levels in mice re-infected without α-GC (Fig 8C). AHR as a direct marker of asthma was not significantly different between the RSV and α-GC+RSV groups (Fig 8D). This suggests that adult iNKT cell stimulation during RSV re-infection does not exacerbate eosinophilic lung disease.

Discussion
Our results show that neonatal iNKT cell sensitization during RSV primary infection can promote immune responses in lungs upon RSV re-infection. This may be associated with the development of asthma following RSV re-infection, contrary to adult iNKT cell stimulation during RSV re-infection.
One study reported that the age of first RSV infection is critical for determining cytokine production and disease patterns during adult RSV re-infection [15]. Neonates tend to display Th2-biased responses with prolonged memory, whereas Th1 memory is unstable [27][28][29]. Although neonatal exposure to α-GC with RSV does not appear to induce an immediate response, our results suggest there may be prolonged Th2-biased memory. Therefore, during RSV re-infection in adult mice, iNKT cells sensitized in neonates promoted the recruitment of eosinophils to lungs and secretion of Th2-type cytokines (Fig 5). We found no difference in viral clearance after re-infection, perhaps because adaptive immunity is sufficient for clearance during RSV re-infection.
Eosinophilic inflammation of airways, which includes an increase in activated and degranulated eosinophils, is the key feature of both allergic and non-allergic asthma [30,31]. There are significant correlations between the activation of eosinophils and the severity of asthma as reflected in bronchial hyperresponsiveness and asthma symptom scores [32]. Th2 cytokines such as IL-4, -5, and -13 are considered markers of asthma, and IL-4 is the main cytokine involved in the pathogenesis of allergic disorders, including stimulation of mucus-producing cells and fibroblasts [33,34]. Large amounts of IL-4 can lead to lymphocytic and eosinophilic inflammation, but without airway hyperreactivity [35,36]. IL-13 is closely related to IL-4, which binds to IL-4R receptors and is also expressed by Th2 cells from asthma patients [37]. Increased amounts of IL-13 are observed in the airways of patients with atopic and non-atopic asthma [38,39]. IL-5 is highly specific for eosinophilic inflammation and may play an important role in eosinophil survival, maturation, and activation in asthma [40,41]. The increased percentage of eosinophils in sputum and elevated AHR in asthma are correlated with IL-5 secretion [42,43]. In addition, inhibition of IL-5 was shown to be effective in reducing eosinophilic inflammation and AHR in various species [40].
NKT cells can be immediately activated and rapidly produce large amounts of cytokines such as IL-4, IL-5, IL-13, IL-17, and IFN-γ in response to glycolipid antigens such as α-GC [21,[44][45][46]. During this process, NKT cells also interact with other cells of the immune system, including eosinophils, and lead cell activation, recruitment, and differentiation [47]. Activated NKT cells secrete IL-5, which recruits eosinophils directly to the lung, or secrete IL-4 and IL-13, which causes lung epithelial and endothelial cells and lung fibroblasts to secrete eotaxin. In response, eosinophils are recruited to the lung [48].
In neonatal mouse iNKT cells stimulated by α-GC, enhanced Th2 cytokine production and eosinophil infiltration in lungs after RSV re-infection seem to exacerbate AHR compared to primary infection with RSV alone. However, stimulation of iNKT cells with α-GC alone did not lead to airway inflammation in response to sequential RSV infection. These results suggest that RSVspecific adaptive immune responses are indispensable for the development of this pathology.
Priming of adult iNKT cells with α-GC during RSV re-infection elicits immune responses that differ from those associated with neonatal iNKT cell priming. iNKT cells stimulated with α-GC during RSV re-infection resulted in the secretion of more IL-4 and IL-13 in lungs than that found in mice re-infected with RSV alone. In our study, iNKT cell priming with RSV re-infection of adult mice did not promote recruitment of airway eosinophils or AHR with decreased IL-5. These results show that adult iNKT cell stimulation during RSV re-infection does not aggravate eosinophilic lung disease. The number of infiltrated neutrophils (CD11b + CD11c -Gr-1 + ) was increased in RSV re-infected adult mice, regardless of α-GC exposure during neonatal primary infection or re-infection (S4 Fig). In addition, IL-17 was below the detection level of the assay in our experimental conditions.
Collectively, our results show that neonatal iNKT cell sensitization is involved in exacerbation of lung inflammation and asthma pathology following RSV re-infection. RSV re-infection without α-GC co-administration also led to CD1d-mediated airway inflammation, suggesting that natural RSV infection results in CD1d-dependent NKT cell stimulation. If so, RSV itself may have NKT cell ligands or may promote presentation of natural ligands to NKT cells via CD1d-expressing cells in vivo.
Our results suggest that stimulation of iNKT cells during neonatal RSV infection may be a cause of asthma aggravation following later RSV re-infection.