Research ArticleALLERGY

Inception of early-life allergen–induced airway hyperresponsiveness is reliant on IL-13+CD4+ T cells

See allHide authors and affiliations

Science Immunology  07 Sep 2018:
Vol. 3, Issue 27, eaan4128
DOI: 10.1126/sciimmunol.aan4128

Age matters in allergy

Development of allergy is driven by type 2 cytokines, IL-4, IL-5, and IL-13. With the discovery that type 2 innate lymphoid cells (ILC2s) and T cells can produce these cytokines, understanding the contributions of T cells and ILC2s in allergic responses has become important. Using mouse models of allergic airway inflammation, Saglani et al. report the contributions of T cells and ILC2s to be dependent on age. They found T cells to be the predominant source of IL-13 in neonatal mice as compared with ILC2s in adult mice. Given that neonates have fewer T cells as compared with adults, the results are contrary to expectations and bring to the fore an unappreciated role of neonatal T cells in this context.


Airway hyperresponsiveness (AHR) is a critical feature of wheezing and asthma in children, but the initiating immune mechanisms remain unconfirmed. We demonstrate that both recombinant interleukin-33 (rIL-33) and allergen [house dust mite (HDM) or Alternaria alternata] exposure from day 3 of life resulted in significantly increased pulmonary IL-13+CD4+ T cells, which were indispensable for the development of AHR. In contrast, adult mice had a predominance of pulmonary LinnegCD45+CD90+IL-13+ type 2 innate lymphoid cells (ILC2s) after administration of rIL-33. HDM exposure of neonatal IL-33 knockout (KO) mice still resulted in AHR. However, neonatal CD4creIL-13 KO mice (lacking IL-13+CD4+ T cells) exposed to allergen from day 3 of life were protected from AHR despite persistent pulmonary eosinophilia, elevated IL-33 levels, and IL-13+ ILCs. Moreover, neonatal mice were protected from AHR when inhaled Acinetobacter lwoffii (an environmental bacterial isolate found in cattle farms, which is known to protect from childhood asthma) was administered concurrent with HDM. A. lwoffii blocked the expansion of pulmonary IL-13+CD4+ T cells, whereas IL-13+ ILCs and IL-33 remained elevated. Administration of A. lwoffii mirrored the findings from the CD4creIL-13 KO mice, providing a translational approach for disease protection in early life. These data demonstrate that IL-13+CD4+ T cells, rather than IL-13+ ILCs or IL-33, are critical for inception of allergic AHR in early life.


The key pathophysiological abnormalities of allergic asthma include airway hyperresponsiveness (AHR), eosinophilic inflammation, and remodeling (1). Childhood onset disease is common, affecting about 10% of children, and is characterized by the key clinical symptom of recurrent wheeze (2). About one-third of all children develop wheezing in the first 5 years of life, but only one-third of those will develop asthma (3). However, the mechanisms by which allergic immune responses are initiated and the factors that mediate onset of preschool wheezing and progression to asthma are currently unidentified (4, 5). AHR is a central feature of recurrent wheezing in children who develop asthma, and impaired lung function (6, 7) and AHR shortly after birth (8, 9) are known to be associated with asthma in adolescence and adulthood (10).

The importance of innate immunity, specifically innate lymphoid cells (ILCs), in the inception of allergic asthma is increasingly proposed (11). However, during pregnancy, there is a change in the uterine environment toward a T helper type 2 (TH2) cytokine profile, and the thymic microenvironment is TH2-skewed in the early postnatal period and undergoes age-related suppression in favor of increasing TH1 maturation (12). Despite this, the current dogma is that pulmonary type 2 ILCs (ILC2s), not CD4+ T cells, are the primary cellular source of type 2 cytokines [interleukin-5 (IL-5) and IL-13] in early life (13). Although allergic asthma begins in childhood (14), mechanistic studies of allergic airways disease had predominantly used adult experimental models (1517), thus disregarding the specific developmental effects of postnatal immune maturation (18). A number of recent studies have used age-appropriate murine neonatal models and have demonstrated that, in C57BL/6 mice, perinatal type 2 immunity depends on IL-33 that is immediately up-regulated from the first day of life and drives accumulation and activation of IL-13–secreting ILC2s and pulmonary eosinophils after house dust mite (HDM) exposure (13, 19, 20). However, the predominant clinical manifestation in infants and preschool children is recurrent wheezing with associated AHR and reduced lung function (21), but neonatal studies to date have not investigated the mechanisms driving AHR.

Age-dependent maturation of the immune system occurs after birth once the neonate encounters the antigen-rich external environment (22). The composition of the airway bacterial profile in utero and in early life is also important because exposure to a diverse bacterial mix, such as that found on traditional cattle farms, has been shown to protect from the development of allergy and asthma (23, 24). The capacity of the adaptive immune system to induce memory responses is limited and is thought to gradually develop after early-life environmental exposure to microbes, pollutants, and allergens (25). A population of fetally derived CD4+ T cells with an effector memory phenotype are present in cord blood. These cells develop during fetal life but have a variety of effector inflammatory functions associated with CD4+ T helper cells at birth (26). However, little is known about the phenotype or function of tissue-specific (pulmonary) effector T cells in early life. Studies in infants have shown that allergen-induced immune responses in whole-blood mononuclear cells can be detected at birth with IL-13 predominating after stimulation with the egg protein ovalbumin (27). A differential developmental pattern of IL-13 versus IL-4, IL-5, and interferon-γ (IFN-γ) production was evident in infants in the first 3 months of life (28). Although these data implicate IL-13 in the inception of early-life allergic immune responses in children, there is little direct mechanistic evidence, particularly for identification of a cellular source for IL-13 during this crucial period.

We have previously demonstrated that exposure of neonatal mice from day 3 of life to inhaled HDM promotes robust eosinophilia, TH2 immune responses, and AHR (29). We show here that the cellular source of type 2 mediators in neonatal mice is not restricted to ILCs but that IL-13–secreting CD4+ T cells are crucial for the development of AHR in early life. In addition, we show that IL-33, which is elevated in school-age children with severe asthma and has been linked to airway remodeling, is not a requirement for the initiation of allergic airways disease. Moreover, protection from AHR was achieved in neonatal mice using inhaled farmyard bacteria administered concomitantly with HDM, with a selective reduction in IL-13+CD4+ T cells and IL-13, despite elevated IL-33 and IL-13+ ILCs. Our data demonstrate that the cellular source of IL-13 is essential in determining the development of early-life AHR and underpins the concept of a window of immune development in early life that has implications for the development of AHR.


IL-13+CD4+T cells are induced by IL-33 in neonatal mice

IL-33 is sufficient to generate allergic airway responses in adult mice via the induction of ILC2s and without the development of an adaptive immune response (30). Although ILCs have been shown to drive type 2 immunity in some models, their role in initiating AHR in early life has not been investigated. To determine whether ILCs are also the predominant cellular source of IL-13 in neonatal mice, we delivered intranasal recombinant IL-33 (rIL-33) for 2 weeks to adult mice or neonatal mice from day 3 of life (Fig. 1A) and enumerated IL-13+CD4+ T cells and ILCs by flow cytometry (fig. S1, A and B). There are numerous published strategies to define an ILC, using combinations of extracellular markers and intracellular cytokine or transcription factor expression (31). However, expression of cell surface markers by ILCs is variable and context-dependent (32). Given the importance of IL-13 in driving the pathological features of early-life airways disease, we focused on cytokine-secreting cells, initially gating on pulmonary LinnegCD45+IL-13+ cells, and then examined the expression of the extracellular ILC markers CD127, CD90, CD25, ST2, and inducible costimulator (ICOS) within this population (fig. S1C). We found that, in neonatal BALB/c mice, only CD90 reliably marked the IL-13+ ILC population at steady state and during IL-33–driven inflammation (fig. S1C); therefore, we used LinnegCD45+CD90+IL-13+ as our definition for ILC2. ST2, CD25, and ICOS were variable expressed, being present on only 30 to 40% of ILC2 at baseline [phosphate-buffered saline (PBS)–treated], although this increased to 75% on rIL-33–induced ILC2 (fig. S1, C and D). CD127 staining showed little separation from fluorescence minus one controls in BALB/c ILC2s under all conditions tested (fig. S1E), precluding its usefulness as a definitive marker of ILC2 populations in this strain of mice. Notably, more substantial CD127 staining was observed on ILC2 in C57BL/6 mice both at steady state and during allergic airway inflammation (fig. S1E), suggesting that strain differences exist in murine ILC2 surface phenotype. LinnegCD45+CRTH2+ ILC2 in children with severe therapy-resistant asthma (STRA) is also predominantly CD127neg (33). Collectively, these findings rationalize the use of LinnegCD45+CD90+IL-13+ as a robust definition of functional IL-13–producing ILC2 in neonatal BALB/c mice.

Fig. 1 T cells and ILCs are equally important sources of IL-13 in neonatal mice exposed to rIL-33.

(A) BALB/c mice ages 3 days and 6 to 8 weeks were exposed to intermittent intranasal rIL-33 (50 μg/kg) or PBS for 2 weeks (red arrows). Harvest was carried out 24 hours after the final dose. Numbers of CD3+CD4+IL-13+ T cells compared with ILCs (LinnegCD45+CD90+IL-13+) in lungs of (B) adult and (C) neonatal BALB/c mice. (D) Pulmonary eosinophils (SiglecF+CD11clow/neg) enumerated by flow cytometry. Airway resistance to methacholine (Mch) in (E) neonatal and (F) adult mice exposed to intranasal rIL-33. (G) IL-13 gene expression in the lung. (H) Levels of IL-13 in the lungs. (I) IL-5 gene expression in the lung. Expression of (J) Muc5ac and (K) Muc5b. n = 6 to 8 rIL-33, n = 4 to 5 PBS. Data are representative of at least two experiments. *P < 0.05, **P < 0.01.

There was a substantial difference in the balance of IL-13+ lymphoid cellular phenotypes observed in the lungs of adult and neonatal mice exposed to the innate cytokine rIL-33. In adult mice, as expected, the effect was predominantly on induction of IL-13+ ILCs, compared with IL-13+CD4+ T cells (CD3+CD4+; Fig. 1B). In contrast, neonatal mice responded to rIL-33 with comparable increases in levels of both IL-13+CD4+ T cells and IL-13+ ILCs (Fig. 1C). Both adult and neonatal mice developed significant eosinophilia after challenge with rIL-33, and this response was greatest in adult mice (Fig. 1D). In addition, rIL-33 increased airway resistance at both ages, concomitant with increased IL-13 levels (Fig. 1, E to H). Levels of IL-5 were also increased in mice of both ages, although the magnitude of the response was greatest in adult mice (Fig. 1I). Although 2 weeks of exposure to cytokine is too short to observe phenotypic changes in airway remodeling, increases in the mucin genes Muc5ac and Muc5b, the principle components of airway mucus, were observed in both neonatal and adult mice (Fig. 1, J and K). Muc5ac has been demonstrated to be necessary for the development of AHR, is increased in asthmatic patients including children, and is regulated by levels of IL-13 (34).

Both T cells and ILCs comprise the neonatal response to allergen

To determine the cellular source of IL-13 after exposure to an antigenically complex, clinically relevant allergen, we compared intranasal HDM exposure for 2 weeks in neonatal mice from day 3 of life to adult mice (Fig. 2). Duration of allergen exposure was limited to 2 weeks to focus on mechanisms underlying disease inception before established adaptive allergic immunity with elevated immunoglobulin E levels. Both neonatal and adult mice developed AHR (Fig. 2, A and B) and eosinophilia (Fig. 2C) in response to allergen. In contrast to the immune response to rIL-33, which in adult mice skewed the IL-13–secreting cells toward an ILC2-dominated response, HDM resulted in significantly elevated numbers of IL-13+CD4+ T cells (Fig. 2D). In contrast, neonatal mice exposed to HDM had a mixed response with induction of both TH2 cells and ILC2s (Fig. 2E). However, significantly more IL-13+CD4+ T cells were induced compared with IL-13+ ILCs (Fig. 2E). The increase in IL-13+ cells, irrespective of their source, resulted in elevated levels of pulmonary IL-13 in adult and neonatal mice (Fig. 2F). To extend our observations to another clinically relevant allergen, we enumerated ILCs and T cells after exposure to the fungal allergen Alternaria alternata (fig. S2A), which is associated with severe pediatric asthma (35) and with elevated pulmonary IL-33 (36). In both adult and neonatal mice, inhaled A. alternata induced a strong inflammatory response with significant elevations of both IL-13+CD4+ T cells and IL-13+ ILCs (Fig. 2, G and H) with associated increased levels of IL-13 (Fig. 2I). The elevated IL-13 concentration in the lung correlated with increased AHR (fig. S2, B and C). Congruent with a type 2 inflammatory response, both adult and neonatal mice exhibited significant pulmonary eosinophilia (fig. S2D).

Fig. 2 In response to allergen, T cells and ILCs are equally important sources of IL-13.

Airway responsiveness to methacholine in (A) neonatal and (B) adult mice. (C) Pulmonary eosinophils (SiglecF+CD11clow/neg). Numbers of LinnegCD45+CD90+IL-13+ILCs and CD3+CD4+IL-13+ T cells in lungs of (D) adult and (E) neonatal mice. (F) Levels of IL-13 in the lung of mice exposed to HDM. IL-13+ ILCs and T cells in lungs of (G) adult and (H) neonatal mice exposed to ALT. (I) Levels of IL-13 in the lung of mice exposed to A. alternata (ALT). n = 6 to 8 allergen-exposed, n = 4 to 6 PBS. Data are representative of two experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

We next phenotyped the allergen-induced T cells present in neonatal lungs and compared them with adults. In the CD4+IL-13neg population, the proportion of cells with a memory-like phenotype (CD44+ CD62Llow/neg) was equivalent in adults and neonates (Fig. 3, A and B), whereas the percentage expressing CD103, an integrin associated with epithelial lymphocyte localization, and the activation markers CD69 and ICOS were greater in adults compared with neonates (Fig. 3, C to F). Examination of the CD4+IL-13+ T cells revealed similar patterns of extracellular marker expression in cells recovered from both HDM and A. alternata–treated mice, although the proportions of cells expressing specific markers was greatest in mice administered A. alternata. In contrast to the IL-13neg population, the proportion of IL-13+CD4+ T cells with a memory-like phenotype was significantly higher in neonates than in adults (Fig. 3, A and B). Likewise, more neonatal IL-13+CD4+ T cells expressed CD103 than adults after HDM (Fig. 3, C and D). A very high proportion of cells expressed CD69 and ICOS in adults and neonatal mice exposed to either HDM or Alternaria (Fig. 3, E and F). Thus, the effector T cells in neonatal lungs, far from having an immature phenotype, appear analogous to adult T cells.

Fig. 3 IL-13 is a feature of allergen-activated neonatal T cells.

CD4+ T cells from HDM and A. alternata (ALT)–treated neonatal and adult mice were classified as either IL-13+ or IL-13neg and analyzed for surface markers. Cells were defined as either (A and B) memory-like (CD44+CD62Llow), (C and D) epithelial associated (CD103+), or (E and F) activated (CD69+/ICOS+). The proportion of neonatal and adult CD4+ T cells from (G and H) PBS control, (I and J) HDM-treated, or (K and L) ALT-exposed mice expressing cytokine were also enumerated. n = 7 to 9 allergen-exposed, n = 5 PBS. Data are representative of two experiments. Human cord blood was analyzed for (M) total CD4+ T cells and total ILC2 (LinnegCD161+CD127+CRTH2+CD56neg) and (N) IL-13+ CD4+ T cells and IL-13+ ILC2. (O) Neonatal T cells expressed TCR-β but lacked expression of CD1d dimer. Values are expressed as a percentage of live CD45+ lymphoid cells. n = 20. **P < 0.01, ***P < 0.001.

We next characterized the intracellular cytokine profile of the T cells. At baseline (PBS control mice), most neonatal cytokine-producing CD3+CD4+ cells had the capacity to secrete IFN-γ and IL-17, whereas this was more limited in cells from adult mice (Fig. 3, G and H). HDM elicited a mixed T cell response with an increase in the proportion of TH2 (IL-5+ or IL-13+), TH1 (IFN-γ+), and regulatory (IL-10+) T cells in both adults and neonates (Fig. 3, I and J). One notable difference in the T cell response to HDM between adults and neonates is that, in neonates, a high proportion of IL-13+CD4+ T cells coexpress IL-5; however, in adults, these dual-positive cells are relatively rare. Exposure to A. alternata stimulated the greatest proliferation of CD4+ T cells, which predominantly generated IL-13 or IL-17 in neonates (Fig. 3K). In adult mice, A. alternata polarized T cells to a type 2 phenotype with very few TH17 cells compared with neonates (Fig. 3, K and L).

CD4+ effector cells have recently been identified in human cord blood (26). We therefore investigated the relative proportions of CD4+IL-13+ T cells and IL-13+ ILCs in cord blood. Fifty percent of CD45+ cells were CD4+, whereas only 0.02% were LinnegCD161+ ILCs (Fig. 3M). In keeping with our observations regarding pulmonary IL-13+ cells in neonatal mice, ILC2s in cord blood (from healthy term infants) were not the predominant source of IL-13. Instead, CD4+ T cells contributed to the potential pool of neonatal IL-13 (Fig. 3N). Neonatal T cells were positive for the T cell receptor (TCR) and negative for the CD1d dimer expressed on natural killer (NK) T cells (Fig. 3O).

Allergen exposure in neonatal SCID mice protects from AHR, whereas Il33−/− mice are unaffected

To determine the functional significance of the pulmonary IL-13+CD4+ T cells in early life, we exposed neonatal severe combined immunodeficient (SCID) mice, which lack T, B, and NK cells, to HDM for 2 weeks (fig. S3A). While the wild-type (WT) mice showed the expected increase in AHR (fig. S3B), the neonatal SCID mice did not exhibit any AHR (fig. S3C). The observed AHR in WT mice correlated with increased levels of IL-13 and IL-33, which were not apparent in SCID mice (fig. S3, D and E). IL-13 is known to be critical for the generation of AHR. However, the dependence of AHR on the innate cytokine IL-33 in neonates has not been investigated. Therefore, we exposed neonatal WT and Il33−/− mice to HDM from day 3 of life. Neonatal Il33−/− mice developed equivalent AHR and an inflammatory response of similar composition and magnitude to WT mice (Fig. 4, A to D). IL-33 was absent in the Il33−/− mice as expected, but the type 2 cytokines IL-5 and IL-13 were induced to similar levels in WT and knockout (KO) mice (Fig. 4, E to G), demonstrating that, in neonatal mice, AHR is independent of IL-33 but totally reliant on IL-13. These data underscore the functional importance of perinatal IL-13+CD4+ T cells in inducing AHR in early life and suggest that pulmonary T cells are a critical source of IL-13 in the immediate postnatal period.

Fig. 4 IL-33 is not critical for initiation of allergic airways disease phenotype.

Three-day-old WT BALB/c and Il33−/−mice were exposed to intermittent doses of HDM or PBS. (A) Airway responsiveness to methacholine in neonatal WT and Il33−/− mice. (B) Pulmonary eosinophils (SiglecF+CD11clow/neg). Numbers of pulmonary (C) LinnegCD45+IL-13+ILCs and (D) CD3+CD4+IL-13+ T cells. Levels of (E) IL-33, (F) IL-5, and (G) IL-13 in the lung. n = 7 to 12 HDM, n = 6 to 8 PBS. Data are representative of two experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

Mice lacking IL-13 in CD4+ T cells do not develop AHR in response to allergen

To confirm the importance of T cell–derived IL-13 on the generation of allergen-induced AHR, we exposed either neonatal WT or Cd4-cre Il-4Il13fl/fl mice to inhaled HDM or A. alternata. T cells from Cd4-cre Il-4Il13fl/fl mice are unable to generate IL-13 or IL-4, thus permitting the specific role of T cell–derived IL-4/IL-13 to be investigated. Initial experiments confirmed the previously observed allergen-induced increase in IL-13+CD4+ T cells in WT mice and established that these cells were absent in the Cd4-cre Il-4Il13fl/fl mice (Fig. 5A). Confirming our hypothesis, mice lacking these cells did not develop AHR after exposure to either HDM or A. alternata (Fig. 5, B and C). We next determined the composition of the pulmonary cellular infiltrate in these mice. The number of eosinophils (Fig. 5, D and E) and IL-13+ ILCs (Fig. 5, F and G) was not significantly different between WT and Cd4-cre Il-4Il13fl/fl mice. Likewise, T cells expressing IL-5 were not affected by the specific loss of IL-13+CD4+ T cells (Fig. 5, H and I). Despite the presence of increased IL-13+ ILC2 with the capacity to secrete IL-13 (Fig. 5, F and G), there was only a minimal increase in pulmonary IL-13 levels after HDM or A. alternata exposure (Fig. 5, J and K) in the absence of IL-13+CD4+ T cells. Expression of Muc5ac was increased in neonates exposed to both HDM and A. alternata (Fig. 5, L and M). In mice lacking IL-13+CD4+ T cells, the HDM-induced increase in Muc5ac was ablated (Fig. 5L), indicating that expression is dependent on IL-13 levels. However, in mice administered A. alternata, allergen-induced Muc5ac gene expression was maintained (Fig. 5M). This suggests that IL-13+ ILC2s alone are not sufficient to generate levels of IL-13 necessary to drive AHR in the absence of IL-13+CD4+ T cells in early life. Thus, mice lacking IL-13 specifically in CD4+ T cells were completely protected from developing AHR after exposure to different allergens, underscoring the importance of these cells in the inception of early-life AHR.

Fig. 5 T cell–derived IL-13 is essential for the inception of AHR.

Three-day-old WT BALB/c and Cd4-cre Il-4Il13fl/fl (CD4Cre) mice were exposed to intermittent doses of HDM, A. alternata (ALT), or PBS for 2 weeks. (A) Representative flow cytometry plots of cytokine expressing (IL-5 and/or IL-13) CD3+CD4+ T cells. Airway responsiveness to methacholine in neonatal mice exposed to (B) HDM or (C) ALT. Lung eosinophils enumerated by flow cytometry in mice exposed to (D) HDM or (E) ALT. LinnegCD45+CD90+IL-13+ILCs in mice exposed to (F) HDM or (G) ALT. IL-5+ T cells to in mice exposed to (H) HDM or (I) ALT. IL-13 levels in mice exposed (J) HDM or (K) ALT. Muc5ac gene expression in mice exposed to (L) HDM or (M) ALT. n = 6 to 8 allergen-exposed, n = 4 to 6 PBS. Data are representative of at least two experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

Therapeutic manipulation of IL-13+CD4+ T cells protects neonatal mice from developing AHR

Although studies using KO mice show proof of principle, to determine whether it is possible to translate these findings and therapeutically manipulate neonatal IL-13+CD4+ T cells that drive allergen-induced changes to lung function, we investigated the mechanism of action of an immunomodulatory agent, the lyophilized bacteria Acinetobacter lwoffii, which is associated with protection from asthma in children (23) and AHR in adult mouse models (20). We administered A. lwoffii (a farmyard isolate with strong allergy-protective properties) intranasally concomitant with HDM to neonatal BALB/c mice from day 3 of life for 3 weeks (Fig. 6A). Neonatal mice that received both the inhaled bacteria and HDM were completely protected from the development of AHR (Fig. 6B). Moreover, eosinophils were significantly reduced in the mice that received A. lwoffii (Fig. 6C). Pulmonary IL-33 levels remained elevated in all mice that received HDM and even in those that received the bacterial isolate (Fig. 6D); however, in keeping with the AHR results, allergen-induced IL-13 levels were almost completely abrogated in the mice that received A. lwoffii (Fig. 6E). When we assessed the lymphoid cellular source of IL-13, numbers of IL-13+ ILCs were increased in mice that received bacteria irrespective of allergen exposure (Fig. 6F), but IL-13+CD4+ T cells were significantly reduced (Fig. 6G). Levels of allergen-induced IL-5 were unaffected by the A. lwoffii (Fig. 6H). To further investigate how the A. lwoffii prevented the recruitment of IL-13+CD4+ T cells, we enumerated pulmonary dendritic cells (DCs). In agreement with data in the literature, exposure to HDM alone resulted in a significant increase in the number of CD11b+ conventional DCs (cDCs; Fig. 6I). Plasmacytoid DC (pDC) and monocyte-derived DC (MoDC) numbers were also elevated. However, concomitant exposure to A. lwoffii completely blocked the expansion of CD11b+ cDCs and MoDCs (Fig. 6I) correlating with the reduction in IL-13+CD4+ T cells and improvement in AHR in these mice. These data confirm the critical role of IL-13 from IL-13+CD4+ T cells for the generation of neonatal AHR and show that IL-13+ ILCs cannot compensate for the absence of IL-13 from T cells in inducing neonatal AHR.

Fig. 6 Protection from AHR in neonatal mice can be achieved by reduction of IL-13+ T cells.

(A) BALB/c mice aged 3 days were exposed to intermittent intranasal A. lwoffii F78 (1.3 × 108 cfu for the first 2 weeks and then 2 × 108 cfu) or PBS (black arrows), followed by HDM (10 μg for the first 2 weeks and then 15 μg) or PBS for 3 weeks (red arrows). Analysis was carried out 24 hours after the final dose. (B) AHR to methacholine after 3 weeks of A. lwoffii F78 and HDM coexposure. (C) Numbers of eosinophils (SiglecF+CD11clow/neg) in the lung. (D) IL-33 and (E) IL-13 levels from lung homogenate. (F) Numbers of ILCs (LinnegCD45+IL-13+) and (G) CD3+CD4+IL-13+ T cells. (H) IL-5 levels in the lung tissue. (I) DC populations in the lung enumerated by flow cytometry. cDC (CD11b+CD11chigh), pDC (CD11cintLy6c+CD64neg), MoDC (CD11c+Ly6c+CD64+). n = 8 for HDM and bacteria exposed groups, n = 5 for PBS. *P < 0.05, **P < 0.01, ***P < 0.001. Data are representative of three experiments.


An essential clinical feature of asthma and wheezing in childhood, including infancy and the preschool years, is the presence of AHR (38). In addition, cohort studies have shown that the single most important factor that determines development of asthma in children is AHR in early life, which may be apparent even before the onset of the manifest symptom of wheezing (39). To identify mechanisms mediating allergen-induced disease inception, and targets for intervention allowing asthma prevention, achieving a reduction in AHR is essential. However, the underlying molecular mechanisms remained unclear. AHR was the focus of this study to optimally reflect pediatric symptoms. We determined that effector IL-13+ CD4+ cells are critical for the development of AHR after exposure to either the ILC2 promoting cytokine rIL-33 or clinically relevant allergens HDM and A. alternata in the first weeks of life. IL-13+ ILCs were insufficient to compensate for an absence of IL-13+CD4+ T cells in early life, demonstrated by the lack of AHR in neonatal SCID mice or in mice specifically lacking IL-13 in CD4+ T cells after exposure to inhaled allergen. Mice lacking functional T and B cells or IL-13+CD4+ T cells did not develop increased levels of IL-13 after allergen exposure from day 3 of life despite comparable numbers of IL-13+ ILCs. Moreover, administration of inhaled farm dust bacteria during exposure to HDM to neonatal mice resulted in complete protection from AHR with a significant reduction in pulmonary IL-13+CD4+ T cells and levels of IL-13 but sustained elevation of IL-33 and IL-13+ ILCs. Collectively, these data show that, in neonatal mice, T cells are an essential early source of IL-13 to drive AHR. Our results indicate that, although ILCs have the potential to generate IL-13 when stimulated with phorbol 12-myristate 13-acetate (PMA) and ionomycin in vitro, in the absence of IL-13+CD4+ T cells in vivo, functional levels of IL-13 are not generated. Thus, despite the assumption that ILC2s initiate pulmonary allergic immune responses, this is dependent on age, and in neonatal mice, IL-13+CD4+ T cells are critical for disease inception in early life. This has significant implications for therapies to prevent wheeze and asthma inception in childhood, because molecular targets that prevent the induction of ILC2s are unlikely to be effective in preventing early onset disease.

Key distinctions between adult and neonatal immune responses after HDM exposure were also shown in a neonatal model that used a single dose of allergen to achieve sensitization at day 3, day 14, or adult life, followed by allergen challenge a week later (13). Similarly, previous neonatal murine studies have shown the importance of the first 2 weeks of life in causing exaggerated immune responses (40) but did not examine associated AHR or interrogate the cellular source of the mediators that potentially generate AHR. We used a model of continuous, low-dose allergen exposure rather than sensitization, followed by allergen challenge several days later, because this reflects the type of exposure that young children likely experience with perennial allergens such as HDM and A. alternata. Although ILC2s have been shown to be important in generating TH2 immunity in experiments with papain or helminths in adult mice (41, 42), our data indicate that, in neonatal mice, IL-13+CD4+ T cells drive early-life allergen-induced AHR. We cannot rule out the role of T cell–derived IL-4 in neonatal allergen-driven responses and, given the potential importance of IL-4, future studies are needed to clarify the relative importance of IL-4 and IL-13 production in this process. In adult mice, an elegant series of experiments using ILC-deficient mice reconstituted with naïve CD4+ T cells has also shown that activation of primed TH2 cells is independent of ILC2s (43). However, in contrast to our findings in neonatal mice, the adult TH2 cell activation was dependent on pulmonary IL-33. Similarly, in response to HDM or papain, adult TH2 cells secreting IL-13 but not IL-4 have been shown to mediate TCR-independent, IL-33–dependent innate-like immune responses (44). Thus, interactions between TH2 cells and ILC2s are vital in developing pathophysiology but are likely to be contextual depending on environmental or temporal factors.

The in utero environment is biased toward TH2 immunity to support a successful pregnancy. CD4+ effector cells with a memory phenotype have previously been identified in human cord blood (26), and we have also shown the presence of IL-13+ CD4+ cells but very few IL-13+ ILCs in cord blood from healthy newborns. A specific subpopulation of IL-4+CD4+ T cells that are present in cord blood from naïve human neonates, but are lost during aging, has also been described (45). This distinct subpopulation of IL-4+CD4+ T cells was only found in neonates, but not in adults, and supports the hypothesis of an endogenously poised type 2 cytokine profile of T cells in neonates and a link between cytokine production and developmental stage (45). Using neonatal BALB/c mice, we have shown that CD4+ T cells make a vital contribution to the pool of IL-13 in the lung that drives AHR. Caution should therefore be exercised when interpreting data from C57BL/6 mice, where allergen appears to induce ILC2s as the major source of IL-13–secreting cells in an IL-33–dependent manner (13). An important factor that determined our use of BALB/c mice is the direct reflection of the disease phenotype of our patients with severe wheezing and asthma, incorporating a marked airway eosinophilia, AHR, and remodeling (46, 47), in this strain and protocol of allergen exposure. Previous murine studies, also in BALB/c mice, have shown that polyclonal stimulation of lung T cells results in a bias toward IL-4 and IL-5 and increased ratio of GATA3+ T cells compared with T-bet+ T cells (48). This would appear to be an intrinsic feature of neonatal T cells because Bacille Calmette-Guérin–primed lung DCs from either neonates or adults prime adult naïve T cells toward TH1, whereas a TH2 cytokine response is observed from naïve neonatal T cells. Cord blood cells, which are reflective of fetal blood, showed a substantial IL-13 response to allergen stimulation in vitro, and the newborns had a TH2 cytokine bias that was restricted to IL-13 (28). Although these results were in peripheral blood rather than in the lung, they do indicate that early life is associated with a skewed IL-13 response. We now show that T cell–derived IL-13 is critical for the inception of allergen-induced lung function changes.

Our data highlight the critical role of T cell–derived IL-13 in the neonatal period. Neonatal mice have a population of IL-13+CD4+ T cells that have the capacity to rapidly promote AHR when exposed to allergen. Modulating the allergen-induced increase in DCs in the lung and, consequently, in the IL-13+CD4+ T cells, via a farm dust bacterial isolate, specifically abrogated AHR, even while IL-33 and IL-13+ ILCs were maintained. The concept of the neonatal “window of opportunity” is gathering momentum with regard to the mucosal microbiota (40), and we now know that life-long immune homeostasis and susceptibility to immune-mediated diseases (asthma, allergies, and bronchiectasis) can be shaped during the postnatal period (22, 49, 50). The specialized neonatal adaptive immune response after birth also has a predisposition to higher expression of GATA3+, type 2 cytokine-producing pulmonary T cells (48) as a result of both normal development and in response to environmental exposures. Here, we have shown that interventional approaches to prevent AHR and asthma in early life need to focus on reducing IL-13+CD4+ T cells rather than IL-13+ ILCs, highlighting the need for age-specific therapeutic approaches in infants and young children compared with adults.


Study design

This study aimed to determine the immune mechanisms that drive the inception of AHR in neonatal mice.

Research samples

Immune cells were collected from the lung tissue of adult and neonatal mice at the times indicated depending on the experimental setup. Cord blood was collected at delivery of full-term pregnancies.

Experimental design


In all experiments, mice from the control and experimental groups came from the same cohorts, were reared under the same environmental conditions, and were age-matched. Adult female mice were randomly placed in either the control group or the experimental group. Neonatal mice were of either sex, and litters were randomly assigned to control or experimental groups.

Sample size

The number of mice analyzed for each different experimental approach is indicated on each figure. All experiments were repeated at least once with similar sample sizes and a minimum number of four mice per group.

Animals and reagents

WT female BALB/c and Beige SCID mice were initially obtained from Charles River (Saffron Walden, UK) and maintained by in-house breeding. Cd4-cre Il-4Il13fl/fl mice (51) on a BALB/c background were a gift from D. Voehringer. Il33−/− mice were a gift from MedImmune Inc. Each mother with its litter was housed separately. Mice were maintained in specific pathogen–free conditions and given food and water ad libitum. In individual experiments, all mice were matched exactly for age and background strain. All procedures were conducted in accordance with the Animals (Scientific Procedures) Act 1986. Recombinant mouse IL-33 (50 μg/kg) for intranasal administration was purchased from R&D Systems (UK).

Allergen challenge

In experiments to assess the effect of allergen challenge on allergic airways disease, 3-day-old neonatal mice and adults (6 to 8 weeks) received intranasal administration of either HDM or A. alternata (Greer, Lenior, NC, USA). From birth to 2 weeks of age, mice were administered 20 μg of HDM extract, 5 μg of A. alternata, or 10 μl of PBS, three times a week. Adult mice received 25 μg of HDM or 10 μg of A. alternata. All outputs were assessed at 24 hours after allergen challenge (52).

Bacteria and allergen coexposure

BALB/c mice were exposed to intermittent intranasal A. lwoffii F78 [1.3 × 108 colony-forming units (cfu) for the first 2 weeks and then 2 × 108 cfu] (53) (a gift from Johann Bauer) or PBS, followed by HDM (10 μg for the first 2 weeks and then 15μg) or PBS for 3 weeks starting on day 3 of life for 3 weeks. All outputs were assessed 24 hours after the final challenge as described below.

Measurement of AHR

Airway resistance was calculated using the flexiVent small animal ventilator (SciReq) using our established protocols (29). Mice were anesthetized with pentobarbital sodium (50 mg/kg, intraperitoneally) and ketamine (100 mg/kg, intramuscularly), tracheostomized, and connected to the flexiVent ventilator via a blunt-ended 21-gauge needle (neonate) or a 19-gauge needle (adult). The mice were ventilated with an average breathing frequency of 150 breaths/min, tidal volume of 10 ml/kg body weight, positive end-expiratory pressure of about 2 cm H2O. Changes in resistance to increasing concentrations of nebulized methacholine (3- 100 mg/ml) were calculated from the snapshot perturbation measurements. Resultant data were fitted using multiple linear regression to the single-compartment model in the following form: pressure = resistance × flow + elastance × volume + fitting constant.

Inflammation and cell recovery

Bronchoalveolar lavage (BAL) was performed with PBS via a tracheal cannula. The volume of BAL fluid instilled was 3 × 200 μl of aliquots for neonatal mice, 3 × 300 μl of aliquots for 3-week-old mice, and 3 × 400 μl of aliquots in adults (29). After lavage, the large left lobe of the lung was mechanically chopped and incubated at 37°C for 1 hour in complete media [RPMI + 10% fetal calf serum, 2 mM l-glutamine, and penicillin/streptomycin (100 U/ml)] containing collagenase (0.15 mg/ml; Type D, Roche Diagnostics) and deoxyribonuclease (25 mg/ml; Type 1, Roche Diagnostics). Cells were recovered by filtration through a 70-μm nylon sieve (Falcon, BD Biosciences, MA), washed twice, resuspended in 1-ml complete media, and counted in a hemocytometer (Immune Systems). The total cell yield was quantified by hemocytometer. All cell counts were performed blind by the same observer.

Flow cytometry

To reduce nonspecific binding, we incubated cells with rabbit serum (Sigma-Aldrich) for 15 min before staining. When staining for intracellular cytokines, single-cell suspensions were incubated at 37°C in complete RPMI for 4 hours, in the presence of PMA (20 ng/ml; Sigma-Aldrich), ionomycin-free acid (1.5 μg/ml) from Streptomyces conglobatus (Merck), and Brefeldin A (5 μg/ml; Sigma-Aldrich). Cells were subsequently washed in PBS and stained with LIVE/DEAD Fixable Blue Dead Cell Stain (Thermo Fisher Scientific/Life Technologies), as per the manufacturer’s directions, before washing twice in PBS. Cell suspensions were then stained with fluorochrome-conjugated monoclonal antibodies to surface markers (see table S1) in staining buffer (PBS containing 1% bovine serum albumin and 0.01% sodium azide) for 20 min at 4°C. Cells were then washed twice in staining buffer and fixed in IC Fixation Buffer (Thermo Fisher Scientific/eBioscience) for 15 min at room temperature. Where necessary, fixed cells were permeabilized using Permeabilization Buffer (Thermo Fisher Scientific/eBioscience) and stained with fluorochrome-conjugated antibodies to intracellular cytokines (see table S1) in Permeabilization Buffer for 20 min at 4°C. “Fluorescence minus one” controls for extracellular and intracellular antigens were used on matched tissue samples for quality control purposes and to assist with gating. Data were acquired on a BD LSR Fortessa using FACSDIVA software (both from BD Biosciences) and analyzed using FlowJo software (v10, Tree Star). For ILC identification in mouse samples, lineage exclusion gates consisting of the surface markers TCRβ, TCRγδ, CD3e, CD5, CD19, CD11b, CD11c, FCεR1, GR-1, F4/80, NKp46, and TER-119 were used (table S1). For ILC identification in human samples, a lineage exclusion gate consisting of CD14, CD16, CD19, CD20, CD3, and CD56 was used.

Quantification of cytokines

Lung tissue was homogenized at 50 mg/ml in Hanks’ balanced salt solution (Gibco) containing protease inhibitor tablets (Roche Diagnostics) and centrifuged at 800g for 20 min, and the supernatant was collected. Cytokines were analyzed in lung homogenate supernatants. Paired antibodies for mouse IL-33 (R&D Systems) and IL-5 (BD Biosciences) were used in standardized sandwich enzyme-linked immunosorbent assay according to the manufacturer’s protocols. IL-13 was measured using a Quantikine kit (R&D Systems) as per the manufacturer’s protocol.

Quantitative PCR

RNA was extracted from the lung using the Qiagen RNeasy Plus Mini Kit, following the manufacturer’s instructions. Reverse transcription was performed with 1 to 2 μg of RNA using the High Capacity complementary DNA (cDNA) Reverse Transcription Kit (Applied Biosystems), following the manufacturer’s instructions. Generated cDNA was used for quantitative real-time polymerase chain reaction (PCR) analysis using TaqMan Fast Advanced Master Mix (Applied Biosystems) and quantified on the ViiA 7 (Applied Biosystems). Relative gene expression was determined via normalization to the housekeeping gene Gapdh. All TaqMan primers were purchased from Thermo Fisher Scientific. Primers are as follows: Gapdh (Mm99999915_g1), dye: FAM-MGB; Il5 (Mm00439646_m1), dye: FAM-MGB; Il13 (Mm00434204_m1), dye: FAM-MGB; Muc5ac (Mm01276726_g1), dye: FAM-MGB; and Muc5b (Mm00466391_m1), dye: FAM-MGB.

Statistical analysis

All results were expressed as median and interquartile range, and data were analyzed using GraphPad Prism 7 software (GraphPad Software). Nonparametric tests (Mann-Whitney U) were used to detect differences between groups, and statistical significance was accepted when P < 0.05. *P <0.05, **P < 0.01, and ***P < 0.001.

Correction: Two reference citations were incorrect and one additional reference was mistakenly omitted from the published version. These have been corrected. There are no changes to data or conclusions.


Fig. S1. Defining IL-13+ CD4+ T cells and ILCs in neonatal mice.

Fig. S2. A. alternata induces AHR in neonatal mice.

Fig. S3. Allergen exposure in neonatal SCID mice does not result in AHR.

Table S1. Antibodies used for flow cytometry.

Table S2. Raw data sets.


Acknowledgments: We thank J. Rowley and J. Srivastava of the Imperial College Core Flow Cytometry facility for assistance. We also thank J. Bauer (Lehrstuhl für Tierhygiene, Technische Universität München) for his gift of A. lwoffii. We are grateful to M. Johnson (Chelsea and Westminster Hospital) for the cord blood samples. Funding: This work was supported by the Wellcome Trust (grants 083586/Z/07/Z, 087618/Z/08/Z, and 107059/Z/15/Z). E.v.M has been awarded the Leibniz Prize. C.M.L. is a Wellcome Senior Fellow in Basic Biomedical Sciences, A. Bush is an National Institute of Health Research (NIHR) Senior Investigator, and S.S. is an NIHR Career Development Fellow. R.G. received a Research Fellowship from the German Research Foundation (GR 4379/1-1). L.J.E. holds a visiting worker permit at the Francis Crick Institute. Author contributions: S.S. wrote the manuscript draft and conceived and designed the experiments. E.v.M. supplied the A. lwoffii and provided intellectual input on the farmyard dust. J.E.V., A.K.M., R.G., R.S., A. Byrne, S.L., S.A.W., J.B., V.F., L.D., F.P., F.U., L.J.E., W.J.B., R.A.O., and L.G.G. performed the experiments. D.V. provided the Cd4-cre Il-4Il13fl/fl mice. S.A.W., R.A.O., and L.G.G. carried out the statistical analyses. A.K.M., A. Bush, and L.G.G. revised the manuscript. C.M.L. conceived the study, designed the experiments, and edited the manuscript. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials.

Stay Connected to Science Immunology

Navigate This Article