Biomarkers specific to allergic lung inflammation are augmented in allergic-OVA mice infected with RSV. Female BALB/c mice were sensitized and challenged with ovalbumin (OVA) followed by intranasal infection with RSV-2A (Fig. 1A). At day 26 of the experimental protocol, allergy-associated gene expression was determined in whole lung tissue by RT-PCR array (Fig. 1B). We hypothesized that the combination of OVA treatment and RSV infection would induce a unique allergic-inflammatory gene profile in comparison to RSV and OVA treatment alone. RSV infection increased (P < 0.05) a number of genes as a percentage of our housekeeping gene (Gapdh) in comparison to saline-treated lungs; those genes include, Adam33, Areg, Bcl6, Ccl22, Ccr4, Cysltr1, Crlf2 (TSLP-R), Ifng, Ifngr2, Il12a, Il1rl1 (IL-33R, or ST2), Il33, Kit, Kitl, Ltb4r1, Mmp9, Postn, Rorc, Satb1, Stat6, Tnfrsf4 and Tslp. As expected, several genes were upregulated in response to OVA sensitization and challenge in comparison to saline-treated animals including Ccl11, Ccl12, Ccl17, Ccl5, Ccr3, Cd40l, Cysltr1, Il13, Il13ra, Il33, Il4, Itga4, and Tslp. The combination of OVA challenge and RSV infection (OVA-RSV), however, induced Ccl22, Ccr3, Ccr4, Ccl17, Clca1, Crlf2, Rnase2a, Foxp3, Gata3, Ptgdr1 (CRTH2), Il33, Il3ra, Il5, Kit, Mmp9, Muc5ac, Stat5a, Stat6, and Tslp in comparison to saline treated. Several genes were uniquely upregulated in the lungs of OVA-RSV mice in comparison to the OVA mice. Those genes that had significantly (P < 0.05) higher expression in the OVA-RSV mice compared to the OVA mice included, Ccl17, Ccl22, Ccr4, Clca1, Crlf2 (TSLP receptor), Rnase2a, Gata3, Ptgdr2, Il33, Kit, Mmp9, Muc5ac, Rorc, Stat5a and Stat6 (Fig. 1C). Tslp (P = 0.08) and the IL-5Ra (P = 0.12) were slightly elevated in OVA-RSV compared to OVA, but this did not reach statistical significance. Ccl11, Ccl12, Ccl5, Cd40lg, IL-4, and Itga4, interestingly, were suppressed in OVA-RSV in comparison to OVA. These data highlight an important gene expression signature unique to each treatment group.
IL-33 is rapidly upregulated following allergen challenge and RSV infection.
Previous studies have examined IL-33, IL-25 and TSLP in respiratory viral infections, lung injury models, and allergen challenge models (6, 7, 35, 36). Because of these studies, in conjunction with the data presented in Fig. 1, we chose to look at IL-33 and TSLP proteins 6 hours after RSV infection in naïve and OVA-challenged mice, comparatively. IL-33 and TSLP were elevated at this time point to higher levels in the OVA and OVA-RSV treated animals in comparison to other groups (Fig. 2A, 2B). Furthermore, because neutrophils and Th17 cells are reported as a means of exacerbating asthma following respiratory infections, we did measure IL-23 protein as well. Surprisingly, only RSV infection induced IL-23 significantly in comparison to all other groups at 6 hours after RSV infection (Fig. 2C).
Total leukocytes (Fig. 2D), neutrophils (Fig. 2E), macrophages (Fig. 2F), eosinophils (Fig. 2G), and lymphocytes (Fig. 2H) were quantified in BALF cytospins. As a means of understanding immune cells types, specifically neutrophils and eosinophils, that exacerbate allergic inflammation in humans (37, 38), we show that both populations were significantly elevated with the addition of RSV infection after the allergic phenotype is established with OVA. To confirm an effect of RSV infection, we used pulse oximetry to show a reduced blood O2 saturation in animals infected with RSV (Fig. 2I). This RSV effect was statistically significant in the OVA-RSV treated animals as well.
Innate lymphoid cells are increased as early as 6 hours after RSV infection in allergic and non-allergic mice
Because lymphocytes were elevated in the BALF cytospins of OVA-RSV treated animals, we completed a lymphocyte differential by flow cytometry that examined multiple lymphoid cell subsets comprehensively in the OVA-RSV infection model. 6 hours after RSV infection we detected increased CD19 + B cells, CD3 + T cells, ILC3 and ILC2 in RSV-infected animals compared to saline controls, and both of these populations were elevated in OVA-alone treated animals as well. Nkp46 + ILC3 were also elevated in all groups including RSV, OVA and OVA-RSV treated animals (Fig. 3B). Taken together these data further confirm that each treatment induces a unique host response dependent on the presence or lack of pre-existing allergic inflammation.
Eosinophils and neutrophils are present at increased numbers in pre-existing allergic airway inflammation that is amplified by RSV infection
Group 3 innate lymphoid cells (ILC3) have not been previously examined in OVA-RSV model, but studies show that IL-23 and TH17 responses are common under similar treatments. Furthermore, neutrophil recruitment to lungs is a well-established complication in respiratory viral infection; neutrophilic responses are supported by IL-23 and IL-17 in lungs. Coincidently, ILC2 activation with IL-33 and TSLP has been shown to support eosinophils and allergic inflammation by subsequent production of IL-5 and IL-13 (7, 29, 39, 40). Taken together, we hypothesized that group 2 innate lymphoid cells (ILC2) and eosinophils would be increased in the lungs, along with ILC3 and neutrophils at 2 days and 4 days after RSV infection. Indeed, OVA and OVA-RSV-treated mice had significantly more eosinophils at day 24 and day 26 in comparison to saline mice (Fig. 4D and 4E). However, only OVA-RSV-treated animals had increased eosinophils at day 26 in comparison to OVA-treated animals. The numbers of ILC2 increased from day 24 to 26 in the OVA-RSV mice, and contrastingly, the numbers of ILC2 decreased from day 24 to 26 in OVA mice. Similarly, NKp46 + ILC3 were increased in OVA and OVA-RSV treated mice, however, at day 24, RSV infection induced a marked increase of ILC3 and neutrophils compared to saline treated animals. The numbers of ILC3 and neutrophils were close to baseline levels by day 26 in these studies. These studies show temporal changes in innate immune populations with the most striking data showing that the numbers of ILC2 in the lungs remained higher in the OVA-RSV treated animals in comparison to the OVA-treated alone. The next set of experiments examine whether this was a product of IL-33 expression and secretion.
IL-33 is a potent stimulator of type 2 cytokine and chemokine release from group 2 innate lymphoid cells.
Because previous studies have identified TSLP, IL-25 and IL-33 as potent activators of innate lymphoid cells in allergic and non-allergic disease states (6, 7, 41), we chose to assess potential activation of lung ILC2 by these epithelium-derived cytokines. Lung ILC2 were isolated from naïve BALB/c animals and cultured with IL-2 in combination with TSLP, IL-25 or IL-33 for 5 days (Fig. 5A). We determined that IL-2 and IL-33 induced a higher IL-13 response as compared to treatment with IL-25 or TSLP, as was similarly shown by Mjosberg et al (42, 43). This confirmed that IL-33 is a more potent activator of type 2 inflammatory cytokine by ILC2.
In the next studies we compared the IL-33-induced responsiveness of innate lymphoid cells in separate cultures with the hypothesis that ILC2 from OVA-RSV mice produce more IL-5 and IL-13 than OVA treated animals, and ILC3 produce IL-17 in OVA-RSV mice more readily than ILC3 from OVA-treated animals. Lung ILC2 and pan ILC were enriched from Saline-, RSV-, OVA- and OVA-RSV-treated animals. We found significant increases in IL-33-stimulated IL-13 and IL-5 protein expression in OVA and OVA-RSV mice as compared to saline and RSV alone (P < 0.05, Fig. 5B and 5C). OVA-treated and OVA-RSV treated ILC2 were not different when comparing IL-5 and IL-13 production, however ILC2 from OVA-RSV treated animals produced higher levels of CCL22 on a pg/cell basis in comparison to OVA treated. CCL17 was released from IL-33-stimulated ILC2 from OVA mice to an extent greater than OVA-RSV. There was no difference in IL-5, IL-13 and CCL22 levels between the saline and RSV treatment groups. CCL17 production, however, was significantly increased in IL-33-stimulated ILC2 isolated from RSV-infected mice as compared to saline control. Similar trends were observed with isolated, IL-33-stimulated lung T CD3 + cells (data not shown); however, the magnitude of this response was reduced in comparison to lung ILC2. ILC2 demonstrated an approximate 10- and 15-fold increase in cytokine and chemokine production on a per cell basis when compared to T cells. Total innate lymphoid cells (LIN- cells; or pan ILC) were enriched and stimulated with IL-23 (Fig. 2C) and, importantly, approximately 20,000 of these cells produced IL-22 and IL-17 in RSV and OVA-RSV treated animals only. This was a unique feature of RSV infection that was not seen in OVA or saline treated animals. The IL-23 stimulated pan ILC cultures did not produce detectable levels of IL-5 and IL-13. As we were establishing a role for IL-33 in RSV, OVA and the combination of RSV and OVA challenge we also completed pan ILC culture experiments with both IL-23 and IL-33 co-stimulation. IL-33 had no effect on IL-22 or IL-17 production in the pan ILC experiments.
Neutralization of systemic IL-33 significantly reduces airway mucus production, and cytokine and chemokine following OVA-RSV treatment.
We found a significant increase in lung IL-33 mRNA and IL-33 protein in the OVA-RSV group compared to the other groups (Fig. 1C and 2A). We hypothesized that IL-33 was critical for the increased mucus, eosinophilia, and TH2 cytokine production during RSV infection in OVA-allergen challenged mice. To test this hypothesis, we neutralized IL-33 in our model using an anti-IL-33 mAb approach beginning one day prior to RSV infection (Day 21) and again at 3 days after RSV infection (Day 25) (Fig. 6A). Excess mucus production in the airways is a hallmark feature of allergic inflammation and is specifically induced in airway epithelial cells following IL-4 and IL-13 (44, 45). Here we show that anti-IL-33 treatment significantly decreased mucus production in airway epithelial cells in both OVA and OVA-RSV treated animals (Fig. 6B and 6C).
In the next studies, the impact of anti-IL-33 mAb on inflammatory cytokines/chemokines was investigated. Anti-IL-33 mAb treatment significantly reduced IL-33 (Fig. 6D), CCL22 (Fig. 6I), CCL17 (Fig. 6J) in the BALF of OVA and OVA-RSV treated animals, and only TSLP was reduced by anti-IL-33 in RSV treated animals (# indicates a significant anti-IL-33 effect; P < 0.05). Anti-IL-33 mAb had no effect on IL-23, IL-17 and IL-22 in any of the comparisons, although each one of these cytokines was significantly elevated by RSV infection alone. The anti-IL-33 mAb had no effect on these neutrophil-promoting cytokines at the early, 6 hour, time point either, indicating that anti-IL-33 mitigates the type 2 or allergic inflammation, but has no effect on the IL-17 or IL-22 in RSV or OVA-RSV treated animals.
Unique immune populations arise in RSV, OVA and OVA-RSV treated animals at day 4 after RSV infection.
In the last studies we show that neutralizing IL-33 significantly reduces total BALF and lung ILC2 (# P < 0.05 for an anti-IL-33 mAb effect) in only OVA-RSV-treated animals (Fig. 7C and 7G). Total cellularity was not altered by anti-IL-33 treatments and therefore not the reason for the reduced numbers of ILC2. Along with the BALF and lung ILC2 numbers, eosinophils were decreased in BALF and lung tissue following anti-IL-33 treatment. Interestingly, anti-IL-33 reduced both ILC3 and neutrophils in the BALF and total lung tissue of RSV and OVA-RSV treated animals, but not OVA treated. These results demonstrate a surprising pleiotropic role for IL-33 on both eosinophilic and neutrophilic responses generated following ovalbumin treatment with RSV infection.