Effect of LPS dose on T-cell priming in draining LNs
. We first investigated the impact of inhaled LPS dose on the initiation of antigen-specific T-cell priming, which occurs in regional LNs. C57BL/6 mice retroorbitally injected with OVA-specific T cells from OT-II mice (CD45.1) were exposed via inhalation to highly purified OVA together with various amounts of LPS (0–10 μg). After incubating cells from excised regional LNs with OVA, the culture supernatants were analyzed for signature cytokines of different T helper cell lineages. The concentration of the Th2 cytokines IL-4 and IL-5 in these supernatants increased in concert with the dose of LPS used during the sensitization, except for the highest dose of LPS (10 μg); with this dose, almost no IL-4 or IL-5 was detected (). IL-17 production in regional LNs also increased with the dose of LPS used, including 10 μg LPS. The concentrations of IFN-γ also increased with the amount of inhaled LPS, including the 10-μg dose, in agreement with previous reports that high doses of LPS prime Th1 responses (Eisenbarth et al. 2002
Figure 1 Effect of LPS on cytokine production in cultured LNs from C57BL/6J recipients of OVA-specific T cells. Mice were sensitized to OVA using LPS as adjuvant, and the cytokines IL-4, IL-5, IL-6, and IFN-γ were measured (more ...)
Impact of LPS dose on cytokines in the lung and on serum IgE. We next analyzed BALF to determine whether the production of cytokines in the lung after OVA sensitization and challenge would follow the same trend as those produced in regional LNs after sensitization. BALF was collected from mice sensitized to OVA using various doses of LPS and subsequently exposed to aerosolized OVA either on a single occasion to mimic intermittent allergen exposures or on 6 consecutive days to mimic more chronic exposures (). After a single challenge, mice previously administered OVA alone or OVA plus the very high LPS dose (10 μg) had very low concentrations of IL-5 in the airway (), consistent with the low concentrations of this cytokine we observed in regional LNs of similarly sensitized mice (). Mice sensitized with either 10–3 μg or 10–1 μg LPS had much higher concentrations of IL-5 in the airway after OVA challenge, again reflecting the increased amounts of this cytokine seen in regional LNs. After 6 challenges, the concentrations of IL-5 in BALF were lower than after a single challenge, and the trend generally correlated well with the concentrations we had observed in regional LNs. One exception to this trend was mice that had been sensitized to OVA by 10–1 μg LPS and challenged on six occasions. Despite having very high levels of IL-5 in regional LNs and in the airway after a single challenge, these animals had lower concentrations of IL-5 in the airway than mice sensitized using lower doses of LPS.
Figure 2 Effect of inhaled LPS dose on cytokine production in the lung following allergen challenge of C57BL/6 mice. (A) Time line of sensitization, challenge(s), and harvest. (B,C) Naïve (previously untreated) mice (N) and mice sensitized (more ...)
The serum concentration of IL-4, a Th2 cytokine that promotes immunoglobulin class switching to IgE, was lower after multiple challenges than after a single challenge, but it was not lower in mice sensitized with 10–1 μg LPS than in those sensitized with 10–3 μg LPS, indicating that IL-4 and IL-5 are differentially regulated.
We analyzed production of IL-17 in the airways of challenged mice because it can promote neutrophil recruitment and AHR in this murine model of asthma (Wilson et al. 2009
). As we expected, almost no IL-17 was observed in mice that inhaled OVA without LPS. However, the concentrations of this cytokine in BALF increased in proportion to the LPS dose used during sensitization. After six challenges, IL-17 levels were lower than those after a single challenge, but the trend of increased IL-17 with increasing LPS during sensitization was maintained. Thus, unlike IL-5, the concentation of IL-17 in the lung after challenge was closely associated with levels of that cytokine in draining LNs across the entire range of LPS doses tested. IFN-γ, the signature cytokine of Th1 cells, was not significantly different among the various LPS dose groups, suggesting that at these doses, increased Th1 responses do not account for the observed reduction in Th2 responses.
Atopic diseases are associated with elevated levels of IgE antibodies, which bind to Fc receptors on the surface of several cell types, including mast cells and basophils. We therefore studied total serum IgE in mice that had been sensitized to OVA by LPS. We found that titers of this isotype increased in concert with the LPS dose used during sensitization (). The exception to this pattern was mice that had been sensitized using 10 μg LPS and challenged on six occasions. These animals trended toward lower IgE levels than similarly challenged animals that had been sensitized with 10–1 μg LPS.
LPS dose during allergic sensitization determines the nature of allergen-induced pulmonary inflammation. When we examined the impact of LPS dose during sensitization on the types of leukocytes that accumulate in the airway following OVA challenge either on a single occasion or on 6 consecutive days, we observed that C57BL/6J mice that inhaled OVA alone () or LPS alone (data not shown) had very few inflammatory cells in the airway after a single OVA challenge. This suggests that levels of LPS in the home cages were not sufficient to act as a strong adjuvant for OVA sensitization. However, mice receiving OVA together with as little as 10–7 μg LPS developed low but measurable airway eosinophilia after a single challenge; this response was strongly increased after six daily challenges (). Surprisingly, we observed the strongest eosinophilic response in mice that had been sensitized with 10–5 μg LPS. Mice sensitized to OVA with 10–1 μg LPS had markedly fewer eosinophils than mice sensitized with lower doses of LPS, despite having the highest amounts of IL-4 and IL-5 in lung-draining LNs (). Similar results were obtained for BALB/c mice (see Supplemental Material, Figure S1), ruling out the possibility that our findings were unique to C57BL/6J mice. In animals sensitized to OVA by 10–1 μg LPS and challenged by OVA six times, the low number of eosinophils () was consistent with the low concentration of IL-5 in the airways (). Together, these observations suggest that although very low doses of LPS induce relatively weak Th2 responses in the draining LN, they lead to persistent airway eosinophilia, whereas the moderate dose of 10–1 μg LPS initially triggers stronger Th2 responses but results in shorter-lived pulmonary eosinophilia.
Figure 3 Effect of inhaled LPS dose on leukocyte subset recruitment to the lung following allergen challenge of C57BL/6 mice. Naïve (previously untreated) mice (N) and mice sensitized to OVA using LPS were challenged to aerosolized OVA on a single occasion (more ...)
In contrast to our findings for eosinophil accumulation, the number of neutrophils in the airway following a single OVA challenge was proportional to the amount of LPS administered during OVA sensitization, and were not markedly increased after six OVA challenges. Generally, the number of neutrophils correlated well with airway levels of IL-17. Interestingly, although LPS during sensitization was required for lymphocyte infiltration after OVA challenge, higher LPS doses did not lead to increased lymphocytes. By contrast higher amounts of LPS during sensitization generally led to more macrophages in the airway.
Effect of LPS dose on subsequent OVA challenge–induced AHR. AHR is a cardinal feature of asthma. Accordingly, we used invasive measurements of airway resistance to study the effect of LPS dose during allergic sensitization on the development and progression of AHR. As expected, mice that received OVA alone during the sensitization phase did not develop AHR after a single OVA challenge or after six challenges. By contrast, mice sensitized using doses of LPS ranging from 10–7 to 10–1 μg displayed increased airway resistance in response to methacholine after a single OVA challenge (; see also Supplemental Material, Figure S2A). This AHR was sustained in mice that had been sensitized by very low doses of LPS but not in mice sensitized by 10–1 μg LPS. Similar results were obtained when BALB/c mice were examined (see Supplemental Material, Figure S2B). Thus, like eosinophilic inflammation, AHR was also sustained after multiple OVA challenges in mice sensitized with very low doses of LPS but not in mice sensitized by the higher, but still moderate, dose of 10–1 μg LPS.
Figure 4 Effect of inhaled LPS dose on airway response to subsequent allergen challenge. Airway resistance (R; cmH20/mL/sec) was measured in anesthetized C57BL/6J mice that were sensitized to OVA using LPS, challenged with aerosolized OVA on a single occasion (more ...)
Environmental amounts of LPS are sufficient to prime both Th17 and regulatory responses to inhaled allergens. We examined whether the amounts of LPS found in natural environments are capable of priming Th2 and Th17 responses that lead to airway eosinophilia, neutrophilia, and AHR, reasoning that common house dust would provide a good representation of indoor environments. We therefore prepared extracts from dust samples and tested their abilities to promote allergic responses to co-instilled OVA. We evaluated two HDEs that contained dust mite allergens but had different endotoxin activity (see Supplemental Material, Figure S3A). Endolo had a relatively low endotoxin activity (approximately equal to 10–2 μg LPS/20 μL HDE), and Endomod had a higher endotoxin activity (approximately equal to 10–1 μg LPS/20 μL HDE). In a dose–response experiment in C57BL/6J mice, we found that after OVA challenge, neutrophilic and eosinophilic responses generally increased with increasing doses of HDE, and that at low doses, the Endomod HDE was more effective than Endolo (). However, mice that were sensitized with 20 μL of Endomod HDE had fewer airway eosinophils after the 6-day challenge than did mice sensitized using lower doses of this HDE. This observation was reminiscent of the reduced airway eosinophilia observed in mice exposed to 10–1 μg LPS and challenged six times compared with mice exposed to lower doses of LPS ().
Figure 5 Effect of LPS on induction of effector and regulatory responses by HDEs. Values are shown as the mean ± SE cell number of total leukocytes, eosinophils, and neutrophils in BALF. (A,B) Mice were sensitized with OVA together (more ...)
HDEs are complex mixtures and typically contain multiple adjuvants and allergens (see Supplemental Material, Figure S3B). To confirm that moderate amounts of LPS in house dust can reduce Th2 responses and eosinophilic inflammation, we carried out two additional experiments. First, we studied the Endomod HDE in Tlr4-deficient mice, which are unable to respond to LPS. These Tlr4-deficient mice displayed very little neutrophilic inflammation after OVA challenge, but they had increased numbers of eosinophils (). This result suggests that LPS residing in HDE promotes adaptive immune responses that drive neutrophil accumulation in the lung and suppress eosinophil accumulation. To confirm this, we added 10–1 μg of exogenous LPS to the Endolo HDE and tested its adjuvant activity. This addition of LPS led to significantly reduced eosinophils in the airway, particularly in mice that had been challenged on six occasions ().
Effect of LPS dose during sensitization on OVA-specific and nonspecific Tregs in the chronic asthma model
. We next investigated mechanisms that might be responsible for the reduction in eosinophils observed after multiple exposures to OVA in mice sensitized using moderate amounts of LPS. Because pulmonary levels of IFN-γ were not significantly increased in mice sensitized with 10–1
μg LPS compared with those sensitized using lower LPS doses, it seemed unlikely that Th1 responses were entirely responsible for suppressing eosinophilic inflammation. Moreover, analysis of draining LNs revealed that Th2 cytokines were highest when 10–1
μg LPS was used during sensitization. We considered the possibility that inhaled LPS induces regulatory responses that gain in strength after multiple challenges and suppress eosinophilic inflammation. In some murine models of asthma, IL-10 can suppress allergic inflammation in the lung (Joetham et al. 2007
; Kearley et al. 2005
; Oh et al. 2002
), but amounts of this cytokine were similar in lungs of mice sensitized using high or low doses of LPS (data not shown). We therefore studied Foxp3+
Tregs, which can effectively suppress established allergic responses (Lloyd and Hawrylowicz 2009
), including those induced by OVA/LPS (Whitehead et al. 2011
). Two main types of Tregs have been identified: natural (n) Tregs that develop in the thymus and recognize self-antigens, and induced (i) Tregs that develop in the periphery and recognize exogenous antigens. These two cell types have overlapping but distinct activities and can be distinguished by their transcriptional profile (Bilate and Lafaille 2012
). When we evaluated total Tregs in the lung using Foxp3gfp
mice, we found that total GFP+
Tregs were present at similar levels in the lung whether 10–3
μg LPS or 10–1
μg LPS was used in the sensitization phase (). These experiments ruled out total Treg cell number as being the sole factor responsible for suppression, but left open the possibility that allergen-specific iTregs might be particularly important in this regard. To investigate this possibility, we adoptively transferred OVA-specific OT-II (CD45.1) cells into recipient mice (C57BL/6J; CD45.2) prior to LPS sensitization and OVA challenge on 6 consecutive days. Recipient CD4+
T cells comprised the vast majority of total CD4+
T cells and their numbers were increased in OVA/LPS-sensitized mice () compared with mice treated with PBS or OVA alone (data not shown). The number and percentage of recipient Foxp3+
Tregs were similar in mice sensitized using 10–3
μg LPS or 10–1
μg LPS (), but OVA-specific CD45.1 CD4+
donor T cells, including Foxp3+
Tregs, were more abundant in mice sensitized using 10–1
μg LPS than in those sensitized using 10–3
μg LPS (). This suggested that inhalation of 10–1
μg LPS might, after multiple OVA challenges, lead to suppression of allergic responses by increasing the number of OVA-specific Foxp3+
iTregs in the lung.
Figure 6 Analysis of Tregs in Foxp3gfp mice. Abbreviations: FSC, forward scatter; GMFI, geometric mean of fluorescent intensity; SSC, side scatter. (A) Gating strategy used to identify CD4+ Foxp3+ Tregs, as well as total numbers of Tregs in mice (more ...)
Tregs are heterogeneous cells that might have functional differences. For example, Foxp3+
nTregs in humans are composed of populations with distinct cell surface display levels of ICOS (Ito et al. 2008
). Previous studies have shown that suppression of airway inflammation and AHR is associated with high levels of ICOS expression on CD4+
Tregs (Whitehead et al. 2011
) and that adoptive transfer of ICOS+
T cells, but not ICOS–
cells, suppresses established AHR in mice (Shalaby et al. 2012
). We therefore compared numbers of ICOS+
Tregs, as well as the amount of ICOS on the cell surface of Tregs, in mice sensitized using 10–3
μg LPS. The total numbers of recipient ICOS+
Tregs were similar in mice from these two groups (). However, the number of OVA-specific donor ICOS+
Tregs, as well as the intensity of ICOS staining on these cells, was higher in mice sensitized using the suppressive dose of LPS (10–1
μg) than in those sensitized with the nonsuppressive dose (10–3
μg). This suggests that moderate doses of inhaled LPS might lead to suppression of established allergic responses in the chronic challenge model by increasing the number of ICOS+
Tregs and the levels of ICOS on these cells.