Biased Generation and In Situ Activation of Lung Tissue–Resident Memory CD4 T Cells in the Pathogenesis of Allergic Asthma

Asthma is a chronic inflammatory lung disease characterized by airway hyperresponsiveness, for which there is no cure. Asthma is triggered by the immune response to inhaled allergens, which induces infiltration of effector T cells into the lung (1, 2). Type 2 helper T lymphocytes (Th2) and specific Th2 cytokines, including IL-4 and IL-5, are the major drivers of allergic asthma and promote airway inflammation, recruitment, and activation of effector cells such as eosinophils (3–6) and mast cells (7), mucous production (8, 9), and increased airway hyperresponsiveness. Fibrosis and lung remodeling, observed in chronic disease (10), are also linked to Th2-mediated effects (11, 12). The mechanisms for the induction and chronicity of asthma are not known, and providing insight into this process will allow the development of more targeted therapies to specifically inhibit the lung inflammatory response in this debilitating disease.

It has become increasingly clear that immune responses resident in the lung and other mucosal sites are critical to immune-mediated protection (13–16) and influence tissue inflammation and repair (17, 18). In mouse models of respiratory virus infection, we previously found that noncirculating tissue-resident memory CD4⁺ and CD8⁺ T cells were generated in the lung (designated lung tissue–resident memory T cells [TRMs]), potentially providing optimal protective responses to virus challenge, with minimal morbidity (19–21). Lung TRMs can also be generated to intranasally (i.n.) administered vaccines and to other respiratory pathogens (22–26), suggesting localized generation of lung TRMs. The generation, persistence, and functional role of memory T cells in asthma and chronic inflammatory lung disease are less clear. In mouse models of allergen sensitization, a previous study demonstrated generation of long-lived memory Th2 cells in response to OVA sensitization (27), and a more recent study demonstrated the development of allergen-specific lung TRMs in the more physiological model of house dust mite (HDM) allergen exposure (28). However, the role of lung TRM in perpetuating asthma chronicity is not well understood. Moreover, mechanisms by which lung TRMs may promote an inflammatory response in the lung either through direct in situ activation and/or rapid recruitment and mobilization of immune effectors to the lung are not known.

In this study, we report the biased generation and retention of lung CD4⁺ TRMs in allergic asthma from lung effector responses, whereas infiltrating effector CD8⁺ T cells are not retained as resident lung populations. HDM-primed lung CD4⁺ TRMs persist localized around airways following cessation of allergen exposure and exhibit rapid in situ reactivation upon secondary exposure to inhaled allergen, leading to airway hyperresponsiveness as a hallmark of chronic disease. This early, local reactivation of CD4⁺ TRM is independent of circulating T cell responses, is characterized by increased production of IL-4, IL-5, and IL-17, and increased activation and recruitment of dendritic cells (DCs), as a mechanism by which potent inflammatory responses and airway infiltration can derive from early in situ triggers. Together, our results demonstrate that persistence of lung CD4⁺ TRMs can potentiate long-term airway disease, suggesting that targeting this subset could be of therapeutic benefit in the treatment of chronic asthma.

Female C57BL/6 mice (6–7 wk of age) were purchased from Jackson Laboratory (Bar Harbor, ME) and maintained under specific pathogen-free conditions at Columbia University Medical Center. All animal procedures were conducted according to the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals and were approved by the Columbia University Institutional Animal Care and Use Committee.

After sedation with isoflurane (5% induction, 2–3% maintenance dose), mice were administered HDM extract (Dermatophagoides pteronyssinus; Greer Laboratories, Lenoir, NC) i.n. in PBS (40 μg/25 μl PBS) five times per week for 3 wk as previously described (29), and control mice received PBS i.n. at the same time points. For memory generation and secondary challenge, mice exposed to HDM for 3 wk were “rested” (no further manipulation) for 4–8 wk, then challenged i.n. with two doses of HDM (40 μg/25 μl PBS) 24 h apart. Airway hyperresponsiveness was measured 24 h after the final HDM challenge.

In vivo labeling of T cells with fluorescent Abs was performed as previously described (19). In brief, mice were administered, PE- or Alexa 647–conjugated anti-CD90.2 (anti-Thy1.2; clone 53-2.1; BioLegend), or PE-conjugated anti-CD45.2 Ab (clone 104; BioLegend) i.v., and 7–10 min later lungs were perfused, dissected, and digested in RPMI 1640 medium with collagenase D, DNAse, and Trypsin inhibitor for 1 h at 37°C. Mediastinal lymph nodes and spleen were dissected and manually disrupted to generate single-cell suspensions. Cell suspensions were stained with fluorescent-conjugated Abs for CD4 (clones RM4-5 [BD Biosciences] and GK1.5 [eBioscience]), CD8 (clone 53-6.7; BioLegend), CD11a (clone M17/4; BioLegend), CD11b (M1/70; eBioscience), CD11c (N418; eBioscience), CD25 (PC61.5; eBioscience), CD45 (30-F11; BioLegend), CD69 (clone H1.2F3; eBioscience), CD86 (GL-1; BioLegend), and CD103 (clone 2E7; eBioscience), and Anti-Mouse MHC Class II (I-A/I-E) (clone M5/114.15.2; eBioscience) according to the manufacturer’s protocol. Stained cells were analyzed using the BD LSRII flow cytometer and FlowJo software (Tree Star, Ashland, OR). Absolute cell numbers were determined by flow cytometry using CountBright Absolute Counting Beads (Invitrogen) according to the manufacturer’s protocol.

CD4⁺CD44⁺CD62L⁻ memory T cells protected from i.v. anti-CD90.2 staining as described earlier were sorted from lungs of HDM memory mice using BD Influx Cell Sorter. The resultant purified lung CD4⁺ TRMs were cocultured with an equal number of CD3⁻ splenocytes in the presence of either BSA or HDM (100 μg/ml) for 24 h in complete click’s media in a 96-well U-bottom plate. Supernatants were analyzed for cytokine content using the Legendplex Mouse Th Cytokine kit and a BD LSRII Flow Cytometer. Data were further analyzed using Legendplex v8.0 software (BioLegend).

After sedation with pentobarbital 75 mg/kg, the neck was dissected, and the trachea was cannulated with an 18-gauge beveled tracheal tube. The mouse was then placed on a mouse ventilator (flexiVent; SCIREQ, Montreal, QC, Canada) with a tidal volume of 8 ml/kg and frequency of 150 breaths/min. After 3–5 min equilibration on the ventilator and maintaining the mice at 37°C with a homeothermic blanket (Homeothermic Blanket System; Harvard Apparatus, Holliston, MA), mice were given succinylcholine 0.5 mg every 14 min by i.p. injection; a graded methacholine i.n. challenge was initiated, and airway resistance was measured using the flexiVent system.

Lungs were lavaged twice with 1 ml of sterile PBS, and the cells were centrifuged and resuspended in PBS. Total cell concentrations were counted with a hemocytometer. Cytospin preparations were stained with Quick-Diff (Imeb), and cells were analyzed for differential counts using morphological criteria.

Mice were sacrificed and the lungs were perfused, inflated with agarose, and sectioned using a Leica VT 1000s Vibratome as previously described (19). Tissue slices were first treated with endogenous streptavidin/biotin-blocking reagents (Life Technologies), then stained with biotin-conjugated anti-collagen I mAb (Rockland Immunochemicals, Gilbertsville, PA) followed by streptavidin-conjugated Dylight 488 (Abcam, Cambridge, MA). Lung sections were then stained with fluorescein-labeled lectin from the Cry-Baby Tree, Erythrina cristagalli (Vector Labs), AF647-conjugated or Pacific Blue–conjugated anti-CD4 mAb (clone RM4-4), and PE-conjugated anti-B220. Images were acquired and processed using a Zeiss LSM 700 Laser Scanning Microscope and software (Thornwood, NY).

For peripheral T cell depletion, mice were treated with anti-Thy1 mAb, clone T24/31 (Bio X Cell, West Lebanon, NH) or isotype control IgG2b (50 μg per mouse) daily for 3 d before and 1 d following HDM challenge.

To define the role of lung TRMs in the pathogenesis of allergic asthma, we employed a physiological mouse model of chronic HDM exposure (30), involving daily i.n. challenge of HDM over 3 wk. This protocol of repeated HDM exposure results in severe symptoms that mimic chronic asthma in humans including increased airway hyperresponsiveness to methacholine, severe eosinophilic inflammation, and increased expression of Th2-associated cytokines (30).

To assess whether HDM exposure generated lung TRMs, we used an in vivo, Ab-labeling assay involving i.v. injection of mice with fluorescently labeled anti-Thy1.2 mAb (i.v. Ab) before tissue harvest. With this approach, T cells resident within lung tissue are protected from i.v. Ab labeling, whereas circulating cells in the vasculature bind fluorescent Ab (19, 31). In PBS-treated mice, the majority of lung CD4 and CD8 T cells are labeled with i.v. Ab (>80% CD4⁺ and >90% CD8⁺; Fig. 1A, left) confirming that lungs of naive mice contain predominantly circulating T cells that are not retained in the tissue (19). By contrast, following HDM exposure, the majority of CD4⁺ and CD8⁺ T cells in the lung are protected from i.v. Ab labeling, as early as 4 d of HDM administration (given each day) with increasing frequencies after 3 wk of daily HDM exposure (Fig. 1A, 1B). These results indicate that acute and continuous exposure to HDM alters the composition of lung T cells from mostly circulating to predominantly in the tissue.

The persistence of these lung tissue T cells as TRMs was further assessed by stopping further HDM exposure in mice “rested” for 4–8 wk following the last HDM administration (HDM memory mice). Interestingly, whereas CD4⁺ T cells maintained their distribution within the protected fraction after i.v. Ab labeling, lung CD8⁺ T cells were predominantly labeled in HDM memory mice (Fig. 1A, 1B). We assessed absolute lung T cell counts to determine whether these distinct proportions of protected lung T cells were due to differential generation of lung CD4⁺ and CD8⁺ TRMs following HDM exposure. Although HDM exposure triggered increased numbers of CD4⁺ and CD8⁺ T cells in the lungs, which were predominantly distributed in protected fraction, CD4⁺ T cells were recruited in greater numbers and persisted in HDM memory mice, whereas numbers of protected lung CD8⁺ T cells were reduced significantly (Fig. 1C). Moreover, lung CD4⁺ T cells in HDM memory mice exhibited upregulated CD69 expression, a phenotypic hallmark of TRMs (21, 32), whereas CD8⁺ T cells were largely CD69⁻, even after acute HDM exposure (Fig. 1D). Taken together, these results show that in contrast with respiratory virus infection, which generates both CD4⁺ and CD8⁺ TRMs in the lung (19, 33), HDM exposure specifically primes for establishment of lung CD4⁺ TRMs.

We previously showed that lung TRMs generated from influenza infection were localized to specific niches around airways (19). Here, we assessed whether the TRM phenotype CD4⁺ T cells generated following HDM exposure were localized in similar niches. Overall, there was extensive mononuclear cell infiltration in the lungs of mice exposed to HDM compared with lungs of naive mice, which was concentrated around airways by H&E staining (Fig. 2A, top row). By confocal microscopy, lung CD4⁺ T cells were found to accumulate in clusters around the major airways and within bronchovascular bundles in lungs after 3 wk of acute HDM exposure, and were similarly localized in these regions following cessation of HDM exposure in HDM memory mice (Fig. 2A, lower panel). The numbers of CD4⁺ T cells clustering about the bronchovascular bundles were consistently increased in HDM-sensitized mice after both 3 wk and longer times after cessation of sensitization (Fig. 2B). The cells that localize around the blood vessels appear to be within the tunica media of the walls of the blood vessels, which appear to be enlarged (Fig. 2, lower). By contrast, circulating CD4⁺ T cells (in vivo labeled) were located within the capillaries of alveolar walls and appeared as punctate dots distributed across the lung (Supplemental Fig. 1). Thus, CD4⁺ TRMs generated from HDM exposure persist around airways, suggesting they are poised for early responses to inhaled allergens.

We assessed whether the persistence of lung CD4⁺ TRMs was associated with a secondary response to HDM challenge, manifested by rapid induction of airway hyperresponsiveness following short-term HDM exposure. For assessing secondary responses, we used a 2-d HDM challenge that does not trigger symptoms in naive mice (30). We observed an increase in airway hyperresponsiveness in HDM memory mice after 2-d challenge with HDM compared with PBS challenge and HDM challenge of naive mice (Fig. 3A), indicating that pathological symptoms of asthma occur with the kinetics of a secondary response in HDM memory mice. This asthmatic response was associated with an increase in immune cell infiltrates into the airways characteristic of allergic asthma. Overall, there was a substantial increase in the total immune cell infiltrate in the bronchoalveolar lavage (BAL) during the secondary HDM challenge with a significant increase in lymphocytes, eosinophils, and neutrophils (Fig. 3B, 3C), characteristic of an acute asthmatic response (34). Although the T cells present in the BAL of HDM memory mice challenged with PBS were predominantly CD4⁺ T cells with negligible CD8⁺ T cells (consistent with biased persistence of CD4⁺ TRMs), there was an increase in both CD4⁺ and CD8⁺ T cells in the airways of HDM memory mice following 2-d HDM challenge (Fig. 3B). Together, these results show a rapid infiltration of innate and adaptive lymphocytes into the airways in mice with persisting lung CD4⁺ TRMs generated from previous HDM exposure.

We examined whether lung CD4⁺ TRMs mediated secondary responses to allergen rechallenge in vivo and their functional responses to HDM stimulation in vitro. In vivo, we examined whether circulating or resident lung T cells were responding to HDM challenge. In naive mice, few lung TRMs are present [Fig. 1A (19)] and a 2-d HDM challenge did not result in significant lung TRM generation or lung T cell activation as assessed by upregulation of the activation marker CD25 (Supplemental Fig. 2). By contrast, 2-d HDM challenge of HDM memory mice triggered measurable lung T cell responses, including a significant increase in the frequency and absolute numbers of memory (CD11a^hi) CD4⁺ and CD8⁺ T cells protected from i.v. Ab (Fig. 4A). Moreover, in HDM-challenged HDM memory mice, there was early upregulation of CD25 expression exclusively by CD4⁺ TRMs (protected), but not circulating (labeled) lung CD4⁺ T cells, or by protected or labeled CD8⁺ T cells (Fig. 4B). These results provide evidence for rapid in situ activation of lung CD4⁺ TRMs by HDMs.

The ability of lung CD4⁺ TRMs generated from HDM sensitization to mediate functional HDM-specific secondary responses was assessed by ex vivo stimulation of sorted i.v. Ab-protected lung CD4⁺ T cells from HDM memory mice with HDM or control BSA. We measured the production of multiple proinflammatory and Th2-like cytokines in supernatants after 24 h of stimulation. Notably, lung CD4⁺ TRMs produced multiple cytokines, including Th2-like cytokines IL-5 and IL-4, the Th17-cytokine IL-17, and Th1-like cytokines IFN-γ and TNF-α, in response to HDM stimulation, but not in response to stimulation with an irrelevant Ag (Fig. 4C). Production of these cytokines exhibited an HDM dose-dependent response (data not shown), further establishing Ag specificity. Together, these results indicate that HDM-specific lung CD4⁺ TRMs comprise a multifunctional population, producing Th2-like and proinflammatory cytokines in response to allergen rechallenge.

To determine whether lung TRMs were sufficient to trigger secondary HDM responses and airway hyporesponsiveness, we treated HDM memory mice prior and following 2-d HDM challenge with a pan anti–T cell Ab shown previously to primarily deplete circulating T cells (35, 36). Treatment of HDM memory mice with anti-Thy1 Abs depleted 75% of lung CD4⁺ T cells and >95% of lung CD8⁺ T cells (Fig. 5A). The majority of residual CD4⁺ T cells in the lung following treatment were TRMs, as indicated by their protection from i.v. Ab labeling (Fig. 5B, left), whereas nearly all lung CD8⁺ T cells were depleted (Fig. 5B, right). Importantly, increased airway hyperresponsiveness to HDM challenge in HDM memory compared with naive mice was observed similarly in both control IgG-treated and anti-Thy1–treated mice as shown by Fig. 5C. Together, these results indicate that the asthma response mediated by HDM memory mice was not inhibited by peripheral T cell depletion, suggesting that lung CD4⁺ TRMs are sufficient to trigger lung-localized secondary responses and asthmatic symptoms.

Given the ability of lung-localized CD4⁺ T cells to respond to HDM challenge and trigger asthma, we hypothesized that lung CD4⁺ TRMs may mobilize a lung response through local activation of innate cells. To determine the effect of local reactivation of lung CD4⁺ TRMs on APCs, we analyzed the recruitment and activation of CD11c⁺ DCs in the lungs following 2-d HDM challenge of HDM memory mice. There was an increased frequency and absolute number of CD11c⁺MHC^hi DCs in the lung in HDM memory mice challenged with HDM compared with PBS, which had similar DC frequencies as PBS- or HDM-treated naive mice (Fig. 6A, right; Supplemental Fig. 3). Lung DCs in HDM-challenged compared with PBS-treated HDM memory mice exhibited increased upregulation of CD86 (Fig. 6B). The DCs present in naive mice treated with HDMs also expressed CD86, although many fewer DCs were present in the lung compared with HDM memory mice (Fig. 6A).

It was previously shown in an infection model that the lung contains two major types of DC subsets exhibiting CD103^hi and CD11b^hi phenotypes, respectively, which both have the capacity to migrate to tissue-draining lymph nodes and present Ags to T cells (37). We asked whether alterations in lung DC subsets were associated with HDM challenge of naive or HDM memory mice. The frequency of CD103^hi DCs in the lungs of naive mice treated with PBS or HDM and HDM memory mice treated with PBS was consistently 20–30% with the CD11b^hi subset predominating (Fig. 6C). In HDM memory mice challenged with HDM, the frequency of CD103^hi DCs was significantly decreased (Fig. 6C), resulting in a higher ratio of CD11b^hi/CD103^hi DCs in the lung. We observed similar changes in the composition of these DC subsets in the lungs of HDM memory mice treated with anti-Thy1 mAb (Fig. 6D). Interestingly, the numbers of CD103^hi DCs did not change significantly despite their reduced frequencies (Fig. 6D), indicating an increase in the CD11b^hi DC population perhaps because of recruitment from the blood. These results suggest that lung CD4⁺ TRMs coordinate rapid lung inflammation through DC activation and recruitment.

Tissue-resident memory T cells have been studied for their role in protecting from site-specific pathogens, although their presence and function in the context of chronic inflammation and Ag exposure are not well understood. In this study, we provide evidence for biased generation and long-term maintenance of lung CD4⁺ TRMs in response to chronic exposure to HDM, a major allergen. These lung CD4⁺ TRMs undergo site-specific reactivation at rapid times after allergen challenge and are sufficient to trigger lung inflammation and the asthma response, without significant participation of circulating T cell populations. We further demonstrate that CD4⁺ TRM reactivation triggers rapid mobilization and activation of lung DC populations, with loss of migratory DCs. Our results indicate that long-term allergen sensitization in the lung can be maintained through CD4⁺ TRMs and suggest that targeting TRMs in situ has potential to reduce the chronicity of lung allergic responses.

Our results show that during the acute and ongoing response to HDM, both CD4⁺ and CD8⁺ T cells are extensively recruited to the lung tissues; however, only CD4⁺ T cells are maintained as TRMs. In contrast, the increased numbers and frequencies of CD8 T cells in the lung are greatly reduced after cessation of HDM treatment, suggesting that they either die and/or migrate out of the lung in the absence of chronic HDM exposure. Although there are known MHC-I–restricted epitopes within HDM (38, 39), it is not clear whether CD8⁺ T cells present in the lung following HDM exposure are specifically primed or nonspecifically recruited due to inflammatory signals in the lung. By contrast, a proportion of lung CD4⁺ TRMs are HDM specific based on functional recall shown in this study, consistent with HDM-tetramer–specific CD4⁺ TRMs demonstrated in a previous study (28). Polyclonal HDM-specific responses by CD4⁺ TRMs identified in this article produce both Th2-like, as typically associated with asthma, as well as IL-17, which has more recently been associated with severe asthma responses in children (40, 41). CD4⁺ TRM-mediated asthma in this model may therefore recapitulate more severe disease as observed in humans.

Local reactivation of TRM populations is important for protection against tissue-tropic infections, including influenza virus infection in the lungs, and herpes virus infection in the skin and genital mucosa, and listeria infection in the intestines (35, 36, 42, 43). TRM-derived cytokines can induce the mobilization of immune cells into tissues, enhance the innate immune response by macrophages and NK cells, and also promote infiltration of CD8⁺ T and B cells (35, 43, 44). In the context of allergen challenge, we found that local reactivation of CD4⁺ TRMs was associated with increased recruitment of CD4⁺ and CD8⁺ T cells, granulocytes (eosinophils), and CD11c⁺ DCs into the lung, as well as DC activation, which is essential for induction of asthma in mouse models (45). There was also an increase in CD11b^hi DCs in the lung associated with HDM-specific CD4⁺ TRM recall. Because these recruitment/activation events occurred without the participation of circulating T cells, we propose that TRMs direct lung inflammation directly from their resident niche in the tissues through promoting rapid DC activation and recruitment in the lung.

Our results suggest that lung CD4⁺ TRMs contribute to the chronic symptoms of allergic asthma through recurrent activation and effector functions upon exposure to inhaled allergens. HDM sensitization in this model required repeated exposure to inhaled HDM over 3 wk, and contrasts earlier models of repeated inhalation of OVA protein leading to Ag nonresponsiveness (46, 47). In HDM-sensitized mice, allergen re-exposure leads to the rapid appearance of the characteristic asthma symptoms including airway hyperresponsiveness and airway eosinophilic infiltration. This secondary asthmatic response following cessation of allergen exposure closely mimics the hallmarks of human asthma that persist for decades of life. There is evidence in human asthma studies of local reactivation of T cells in the lung following allergen re-exposure (48), indicating that this model of HDM TRM generation and reactivation can be relevant to investigate mechanisms for allergen recall that occur clinically. Importantly, targeting memory CD4 TRMs in the lung for depletion or inhibition could potentially remove the critical source of initial inflammatory signals that recruit other effector cells to the lung and reduce disease persistence.

We thank Tina Zelonina for the excellent technical help and Rebecca Guyer for help with the mouse colony.

The authors have no financial conflicts of interest.