Catching our breath: updates on the role of dendritic cell subsets in asthma

Abstract

Dendritic cells (DCs), as potent antigen presenting cells, are known to play a central role in the pathophysiology of asthma. Our understanding of DC biology has evolved over the years to include multiple subsets of DCs with distinct functions in the initiation and maintenance of asthma. Furthermore, asthma is increasingly recognized as a heterogenous disease with potentially diverse underlying mechanisms. The goal of this review is to summarize the role of DCs and the various subsets therein in the pathophysiology of asthma and highlight some of the crucial animal models shaping the field today. Potential future avenues of investigation to address existing gaps in knowledge are discussed.

GRAPHICAL ABSTRACT

Dendritic cells (DCs) play a central role in the pathophysiology of asthma. There are multiple subsets of DCs with distinct functions. Some subsets are involved in the induction and maintenance of eosinophilic asthma while others promote tolerance. There is also emerging evidence for a role of certain DC subsets in neutrophilic asthma.

1. Introduction

Asthma is a multifaceted disease that has been increasing in incidence and prevalence, particularly in the pediatric population, over the past 50 years.^[1] Currently, asthma affects more than 300 million people worldwide and the impact to society has far reaching effects beyond the significant morbidity and mortality from the disease itself.^[2] The financial burden of asthma in the United States alone is estimated at >$80 billion.^[3]

Asthma is a chronic disease with airway hyperresponsiveness leading to reversible narrowing of the bronchial airways; if left unchecked, this can result in permanent airway remodeling. While the underlying pathology is heterogeneous and not fully understood, the most common and best studied cause is atopic or allergic asthma (~80% of all pediatric asthma and ~60% of asthma overall).^[4–6] In atopic asthma, antigen exposure leads to eosinophilic lung inflammation, increased mucus production secondary to goblet cell hyperplasia, airway edema, changes in airway smooth muscle (hyperplasia, hypertrophy, and acute constriction) which serve to narrow the airways and cause the classic symptoms of shortness of breath, chest tightness, wheezing, and cough.

The understanding of asthma as a heterogenous disease process involving complex interactions among multiple cell types continues to evolve. Mechanistic mouse studies have elucidated many of the underlying cytokines, chemokines, and cell types responsible for individual symptoms /phenotypes of asthma (reviewed^[7–9]). Our understanding of the influence of epigenetic changes, microbiome, and dietary modifications continues to expand.^[10–16] In the midst of this complexity, dendritic cells (DCs) have emerged as critical players in the communication between the innate and adaptive immune system in initiation and maintenance of asthma.

The goal of this review is to summarize the role of DCs and the various subsets therein in the pathophysiology of asthma. Findings from mouse models of experimental asthma as well as human studies are individually presented and compared to provide an overall context of our understanding of DC subsets. Potential future avenues of investigation to address existing gaps in knowledge will be discussed.

2. Mouse models

Our understanding of the pathomechanisms of many human diseases is informed extensively by animal models. Asthma is no exception; however, it is important to understand the evolution and limitations of these models. Asthma is a heterogenous disease with potentially multiple unique underlying pathomechanisms which have recently been categorized into endotypes (reviewed^[17]). Atopic asthma, which is predominantly Th2-mediated and characterized by lung eosinophilia and steroid-sensitivity, represents the bulk of the disease burden especially in the pediatric population. However, there is a group of non-atopic asthma endotypes characterized by neutrophilia, severe/refractory disease course, and absence of atopic disease which accounts for 5–10% of asthma.^[17] Clinically, these endotypes are associated with late onset, steroid-resistant asthma which tends to have a more ominous disease course, be mediated by Th17 inflammatory responses, and is more likely to be associated with obesity, smoking, and/or advanced age which define the individual endotypes within this group^[17] (Figure 1). There are other rare endotypes of asthma that fall on this continuum between Th2 and Th17 responses including aspirin-exacerbated respiratory disease (AERD) which has clinical features of a Th2 response including eosinophilia but is not steroid responsive and has a discrete underlying mechanism distinct from atopic asthma^[17] (Figure 1). In order to learn more about these disease processes and design appropriate interventions, models for the various endotypes are necessary. Most current mouse models focus on Th2-mediated atopic asthma which is the focus of this review; however, the field is shifting to develop and include models that are representative of additional endotypes which has the potential to take the field in exciting new directions.

Figure 1.

Human asthma endotypes compared to existing mouse models of experimental asthma on a spectrum of Th2 vs Th17, steroid responsiveness, eosinophils vs neutrophil infiltration, and roles of DC subsets.

AERD- aspirin-exacerbated respiratory disease; Tg- transgenic mouse; OVA- Ovalbumin; HDM- house dust mite; DEP- diesel exhaust particles; NO2- nitric dioxide; LPS- lipopolysaccharide; CFA- complete Freund’s adjuvant ; RORγt- retinoid orphan receptor gamma t

There are two primary mouse strains used to study asthma: BALB/c and C57BL/6 mice. BALB/c have been reported to exhibit an innate Th2-biased immune response whereas C57BL/6 are reported to be Th1 predominate responders.^[18–20] Despite this, C57BL/6 still develop experimental atopic asthma,^{[18, 21–23]} but exhibit differences in disease phenotype. For example, BALB/c mice have a greater increase in airway reactivity to methacholine and increased numbers of mast cells in lung tissue than C57BL/6 mice.^[18] Conversely, eosinophil and neutrophil counts in BAL fluid as well as peribronchial eosinophilia were greater in C57BL/6 mice.^{[18, 23]} Cytokine levels also vary between the models and tend to be higher from whole lung extracts in BALB/c mice, whereas select cytokines (CCL11 and CCL5) levels were higher in BAL fluid of C57BL/6 mice.^[18]

Regardless of which strain is used, inbred mice in specific-pathogen-free facilities do not develop asthma spontaneously and so models designed to mimic the pathophysiology have been introduced. Atopic asthma can be achieved by genetic modification, adoptive transfer of activated T cell or dendritic cells, or induction by exposure to allergen (reviewed^[24–26]; Figure 1). The allergen-based models usually include 2 phases: sensitization (intraperitoneal injection of antigen+/− adjuvant, subdermal injection, or intranasal administration depending on allergen) and challenge (often inhaled but could also be intranasal or intratracheal). Advantages and disadvantages of each protocol including physiologic relevance of allergens to human disease are reviewed by Aun et al.^[26] Protocols for inducing asthma in mouse models last up to 20 weeks and the resultant airway inflammation and airway hyperresponsiveness (AHR) resolve within weeks after the last allergen challenge. Animal models have also revealed a reproducible phenotype of allergen hyperresponsiveness in offspring of allergic mothers which has led to important investigations of contributions of the in utero environment to the development of the fetal/neonatal immune system.^{[21, 22, 27–30]}

There are several allergens used in mouse models of allergic asthma, including house dust mite extract (HDM) and purified chicken egg ovalbumin (OVA). It is important to examine mechanisms of allergy with more than one type of allergen, including purified allergen proteins and complex allergen extracts. As a crude extract of an allergen implicated in human asthma, HDM contains multiple antigens and other immune stimulants (i.e. toll-like receptor agonists) and has intrinsic sensitization properties. OVA, while not a common airway allergen in human disease, is well characterized as an experimental asthma model and allows for the use of specialized tools for quantifying antigen-specific T cell activation and response such as OTII mice with OVA-specific T cell receptor expression.^[31–33] Additionally, fungi (Aspergillus fumigatus and Alternatia alternata), cockroach extracts, Ascaris antigens, cotton dust ragweed and latex have also been used in murine models of asthma.^[31] In order to stimulate robust allergic response, allergens (particularly those such as OVA which lack intrinsic sensitization properties) are sometimes used in conjunction with an adjuvant. Aluminum hydroxide (alum) is one of the most common adjuvants, but nitrogen dioxide,^[34] diesel exhaust particles (DEP),^[35–37] lipopolysaccharide (LPS),^{[38, 39]} and complete Freund’s adjuvant (CFA)^{[40, 41]} have also been used. Importantly, the choice of adjuvant affects the resultant immune response. For example, alum induces a potent Th2-mediated inflammatory response, LPS and CFA induce a Th17 neutrophilic response, and low dose LPS, nitrogen dioxide, and diesel exhaust particles induce a mixed response.^{[36, 37, 42]} These various models can produce a spectrum of immune responses (similar to human asthma endotypes) which fall along a continuum of Th2- to Th17-mediated inflammation (Figure 1).

3. Pathophysiology of Dendritic cells in Asthma

DC anatomic location combined with their antigen presenting expertise and ability to influence Th-type responses poise DCs as an essential component of atopic asthma. DCs have been shown to be necessary^[43–48] and sufficient^{[49, 50]} for both induction and maintenance of Th2-mediated asthma as well as for establishing allergic predisposition in offspring of allergic mothers.^{[21, 22, 27, 29, 51]}

This has been extensively shown by deleting CD11c+ cells prior to sensitization or after sensitization but prior to challenge, both of which eliminate the clinical features of asthma.^{[49, 50]} In another approach, administering CD11c+ allergen-pulsed DCs induces asthma in naïve mice.^[49] Furthermore, transplantation of CD11c+ DCs isolated from pups of allergic mothers prior to antigen exposure into pups born from non-allergic mothers is sufficient to induce allergic predisposition.^[27] The mechanism underlying these findings has been of considerable interest over the past 25 years.

DCs present on the basolateral side of epithelial cells in lungs sample antigen from the lumen of airways by extending dendrites through the epithelial barrier. This is regulated by interactions between epithelial cells and DCs (reviewed^{[8, 9]}). Interestingly, both human and murine DCs express tight junction proteins which are hypothesized to functionally maintain the epithelial barrier during sampling. Once the antigen is acquired, the DC migrates to the draining lymph node. The DCs in the conducting airways have a rapid half-life of 2-days because of this migration.^[52] In the lymph nodes, DCs carry out their professional antigen presenting cell role and present processed antigen to naïve T cells. In addition to their role in lymph nodes, DCs interact locally (in the lung mucosa) with primed T cells.^[53–55] This then sets the adaptive immune system in motion to a Th2 pathway leading to production of IL-4, 5, and 13 as well as eosinophilic lung inflammation. Allergen exposure also stimulates generation of stem cell growth factors that induce bone marrow production of DCs and eosinophils which then circulate to the lungs and perpetuate the inflammatory process.^{[56, 57]}

Allergic responses during pregnancy alter the in utero environment for offspring development of DCs. This altered in utero environment leads to changes in fetal DCs which predispose offspring to allergen hyperresponsiveness postnatally.^{[21, 22, 27–29, 51, 58]} In particular, the immune-modulatory lipid beta glucosylceramide (βGlcCer) is elevated in allergic moms and crosses the placenta to alter fetal dendritic cell differentiation^{[29, 30]} and has recently been shown to be both necessary and sufficient to establish allergic predisposition in offspring.^[29]

In humans, levels of βGlcCer, in conjunction with other immune modulatory lipids present in cord blood, have been shown to correlate with incidence of childhood wheeze.^{[59, 60]} These findings combined with the observation that a child born to an allergic mother is more likely to suffer from allergic disease than one born to an allergic father^[61] implicate a role of the in utero environment in establishing allergic predisposition in offspring.

Given the relative inaccessibility of DC collection and study in humans, much less is definitively known about the specific mechanisms of DC involvement in humans. However, in humans, allergen exposure leads to increased DCs and eosinophils in the lungs and increases in Th2 cytokines and chemokines in conjunction with the clinical manifestations of airway hyperreactivity, goblet cell hypertrophy, and eosinophilic inflammation.^[62] Furthermore, when human monocyte-derived DCs are exposed to allergen (HDM^[63] or pollen^[64]) in vitro, they acquire Th2 cell polarizing capacity similar to what is seen in mouse models. However, not all DCs are the same, and, in recent years, we have learned more about the specific contributions of individual DC subsets.

4. Specific roles of DC subsets

Dendritic cells differentiate from bone marrow-derived common dendritic cell progenitors (CDPs) and subsequently differentiate into plasmacytoid DCs (pDCs) and conventional DCs (cDCs) (Figure 2). cDCs are further subdivided into cDC1s (homologous subset in humans is also referred to as myeloid DC (mDC) type 1) and cDC2s (homologous subset in humans is also referred to as mDC type 2) (Figure 2). Additionally, there are monocyte-derived dendritic cells (moDCs) which have a prominent function in the pathophysiology of multiple diseases including atopic asthma. However, there are challenges identifying and defining moDCs more so than other DC subsets due to their plasticity and more complex lineage.^[65–69] Each DC subset is defined based on their profiles of cytokine production, antigen processing and presentation efficiency, and capacity to induce different T cell responses—all of which help define the different physiologic roles DC subsets play in asthma. These subsets are differentiated via flow cytometry using a combination of markers (Table 1). The biggest caveat for all subsets is that the markers vary by location (including blood, lymphoid tissue, lung, or other primary organ tissue) and context (inflammation/contextual signals both locally and systemically as well as maturation and activation state) of the DCs.^{[8, 70–74]} Furthermore, there is no standardized gating scheme to study DC subsets but there is ongoing work to develop techniques incorporating unsupervised computational analysis of flow cytometry data to improve reproducibility and allow for comparisons across various tissues and species.^{[70, 75]} Characteristic markers for which there seems to be a consensus in the literature are bolded in Table 1. Localization within the lung is listed for murine subsets (Table 1) but there is little known about in situ localization in humans given the inaccessibility of tissue samples.

Figure 2.

Ontological relationship of DC subsets during differentiation

HSC- hematopoietic stem cell; MDP- monocyte-macrophage DC progenitor r; CDP- common DC progenitor; Mono- monocyte; pDC- plasmacytoid DC; cDC- conventional DC; moDC- monocyte-derived DC

Adapted from Vroman, 2017^[107]; Collin, 2018^[106]; See, 2017^[67])

Table 1.

DC subsets and their defining characteristics

DC Subset	Locationa)	Markersb)	Role in immune system
cDC1 (also mDC1 in humans)	Associated with epithelium of large conducting airways above basement membrane 0.1% of mononuclear cells in blood	Mouse: CD11b-, CD11c+, CD103+, Ly6c-, MHCII+, CLEC9A, XCR1e, IRF8+, CD24-, CD26+ Human: CD11b low/-, CD11c+/intermediate, MHCII+, BDCA3(CD141)+, CLEC9A, XCR1, IRF8+	Anti-viral immunity and pulmonary tolerance
cDC2 (also mDC2 in humans)	Submucosa/lamina propria of conducting airways beneath basement membrane and lung parenchyma 1% of mononuclear cells in blood	Mouse: CD11b+, CD11c+/high, Ly6c-, CD64-, MHCII+, CD103-, IRF4+, CD24+, CD26+, CD172a(SIRPα)+ Human: CD11c+/high, MHCII+, BDCA1(CD1c)+, CD172a+ (SIRPα), ESAM+, IRF4+	Highly implicated in allergic disease are major APCs in the uptake and processing of the antigens to prime and stimulate CD4+ T-lymphocytes
pDC	Large conducting airways beneath basement membrane <0.1% mononuclear cells in blood	Mouse: CD317(PDCA1)+,CD11b-, CD11c-/low, MHCII+, siglec-H, Gr-1(Ly6G/C) intermediate, B220+ subgroups with CD8α-β-, α+β-, or α+β+ Human: CD123+, MHCII+, CD11c-/low, ILT7(CD85g)+, CD303(BCDA2)+, CD304(BDCA4)+, CD33-	Anti-viral immunity and pulmonary tolerance; produce large amounts of IFN-α, may play a role in Th-17 mediated asthma
moDC (also mDC or inflammatory DC in mouse)	No consensus location as is a highly variable subset induced by inflammation	Mouse: CD11b+, CD11c+, Ly6c+, MHCII+, CD64+, FcεR1 (MAR1)+, CX3CR1, CD209 (DC-SIGN), CD206, CD26- Human: CD14+, CD1c+, CD1a+,CD11c+, CD206, Sirpα+, FcεR1+,CD16-	Thought to play a similar role to cDC2s in antigen presentation. Capable of inducing Th1, Th2, or Th17 responses. Role convoluted by plasticity and use of reliable markers

^a)Vroman, 2017^[107]; Lambrecht, 2014^[9]; Merad, 2013^[141]

^b)Gaurav, 2013^[72]; Lambrecht, 2009^[8]; Bosteels, 2020^[71]; Abdala-Valencia, 2014^[21]; Merad, 2013^[73]; Collin, 2013^[74]; Guilliams, 2016^[70]

^c)Abbreviations: CD (cluster of differentiation molecule); Ly6c (lymphocyte antigen 6 complex, locus C1); MHCII (major histocompatibility complex); CLEC (C-type lectin-like receptor); XCR1 (X-C Motif Chemokine Receptor 1); IRF (interferon regulatory factor); BDCA (blood dendritic cell antigen); PDCA1 (bone marrow stromal antigen 2); SIRPa (signal regulatory protein alpha); ESAM (endothelial cell-specific adhesion molecule); siglec (sialic acid-binding immunoglobulin-like lectins); Gr-1 (Ly6G/C); B220 (CD45 receptor); ILT7 (immunoglobulin-like transcript 7); FcεR1 (high-affinity receptor for the Fc region of immunoglobulin E); CX3CR1 (C-X3-C Motif Chemokine Receptor 1)

Much work has been done to show that CD11c+ DCs (including all subsets other than pDCs) are necessary and sufficient for initiation and maintenance of atopic asthma as well as transmission of allergic predisposition in neonatal offspring of allergic mothers.^{[49, 50]} This section will focus on the studies dedicated to understanding the roles of subsets of DCs as defined by the markers outlined in Table 1.

4.1. cDC1s

4.1.1. Mouse cDC1s

cDC1s are identified as the CD11b− CD103+ DCs in mice. More recently IRF8 has been used a marker for this subset as well; however, it is not specific for cDC1s as it is also expressed in pDCs, inflammatory cDC2s, and moDCs.^{[71, 76]} cDC1s are generally thought to be less important in atopic asthma as their primary function lies in cross-presentation of intracellular antigens to CD8+ T cells and skewing CD4+ T cells toward Th1 and Th17 responses.^[77] Furthermore, they do not increase in number in allergic mouse models^{[21, 22, 78]} and have diminished capacity compared to other DC subsets to take up antigen.^{[30, 79]} The bulk of the literature indicates that cDC1s play a role in pulmonary tolerance. Also, animal models demonstrate worse allergic airway inflammation in response to HDM or OVA in the absence of CD103+ DCs.^{[80, 81]} Additionally, absence of cDC1s in the CD103 null mouse (on C57BL/6 background) results in loss of allergen tolerance in a tolerized mouse model.^[82] While this seems to be the prevailing understanding in the field, there is some controversy as cDC1s have been shown to stimulate a Th2 response to HDM in vitro,^{[83, 84]} and, more recently, a role for cDC1s in mediating eosinophil recruitment in a model of chronic asthma has been described.^[85] Furthermore, some studies have shown decreased allergic airway inflammation in mouse models with decreased CD103+ DCs.^{[83, 84]} These data are complicated by the use of multiple different mouse models and treatment conditions; however, this complexity is consistent with the findings in human studies.

4.1.2. Human cDC1s

In humans with asthma, cDC1s do not increase in bronchoalveolar lavage (BAL) fluid or bronchial tissue (from biopsy) after exposure to inhaled allergen.^[86] There are mixed results when looking in the peripheral blood. Some studies show an increase in cDC1s^[87] while others report a decrease.^[88] Whether changes in circulating cDC1s are reflective of an absolute change in the number of cDC1s present or if it is more indicative of recruitment to the lung (or other tissue) remains to be seen. Functionally, there are reports of cDC1s inducing Th2 cell responses; however, these studies were either conducted with DCs expressing CD141 derived from peripheral blood monocytes^[89] or cDC1s exposed to live-attenuated viral particles which would be expected to stimulate cDC1s disproportionally given their role in antigen cross-presentation.^[90]

CD141 itself has a rich and convoluted history as a human DC marker. It has both been reported as a marker of cDC1s,^{[65, 70, 91–100]} and in other studies is used as the definitive marker for cDC2s.^{[87, 89, 101–104]} Most mechanistic studies conclude that CD141+ human DCs are functionally homologous to murine CD103+ DCs (cDC1s). Gene expression profile analysis confirm this conclusion.^{[65, 70, 92–97]} However, it is worth noting that although pDCs are very similar between mouse and human; there is significantly less homology between cDC subsets of these two species.^[92] CD141 can also be expressed on moDCs which can lead to further confusion.^{[74, 105, 106]} In recent years there has been a push to use IRF8 as the functional marker of cDC1s.^{[88, 107, 108]} However, this approach has been cautioned against as well because emerging data in mice indicate inducible IRF8 expression in cDC2s under inflammatory conditions.^[71] Consensus is still lacking and further studies are required to definitively delineate cDC1s, cDC2s, and moDCs under both steady state and inflammatory conditions.

4.2. cDC2s

4.2.1. Mouse cDC2s

cDC2s are defined as CD11b+ CD11c high DCs and are differentiated from moDCs by their lack of monocyte markers such as Ly6c and CD64.^{[8, 21, 109]} More recently, IRF4 and CD172a (SIRPα) have also become important markers, although these markers are also expressed on moDCs.^{[47, 48, 71, 76, 110–112]} cDC2s are often considered the primary DC subset to mediate atopic asthma given their antigen processing and presenting efficiency combined with their propensity to induce a Th2 response in multiple different models of experimental asthma.^{[21, 22, 48, 50, 77, 79, 112–117]} Adoptive transfers of cDC2s induce experimental asthma.^[112] cDC2s are increased in the lung and BAL fluid in models of asthma,^{[21, 22, 50, 54–57, 78]} in the fetal liver of fetuses gestating in allergic mothers, and in the lungs of offspring of allergic mothers after antigen exposure postnatally.^{[21, 22, 27]} Not only do cDC2s increase under allergic conditions, they have also been shown to migrate to lung-draining lymph nodes and induce allergen-specific Th2 cells. This results in allergic inflammation in the lung,^{[48, 50, 79]} thus fulfilling the primary role in the classic mechanism of induction of atopic asthma. They also have been hypothesized to activate local resident memory T cells leading to a robust Th2 response after re-exposure to allergen.^[7] Overall, the data are convincing that cDC2s play a predominate role in the pathogenesis of experimental asthma.

4.2.2. Human cDC2s

In humans with asthma, cDC2s are increased in peripheral blood, BAL fluid, and bronchial tissue both at baseline and after exposure to inhaled antigen.^{[86, 88, 108, 118, 119]} Functionally, cDC2s isolated from patients with asthma, when primed with HDM, skew CD4+ T cells isolated from non-asthmatic, non-allergic donors to a Th2 phenotype.^[120] Lung cDC2s are implicated in Th2 airway inflammation in humans as they express the high-affinity receptor FcεRI^{[121, 122]} and express increased CD86 and OX-40L.^[88] Interestingly, there is differential expression of markers on cDC2s of patients who experience few asthma exacerbations compared to those with multiple exacerbations per year indicating that there may be biomarkers of disease severity specific to cDC2s. Furthermore, the increase in number of cDC2s was more prominent in patients with increased exacerbations.^[88] Taken together these data further emphasize that greater understanding of DC subsets and their context may inform clinical practice either through targeted treatments or by triaging/prognosticating disease severity.

4.3. pDCs

4.3.1. Mouse pDCs

pDCs are identified among DC subsets by their lack of expression of CD11b and low expression of CD11c as well as the presence of CD317+, siglec-H, and B220 (Table 1). They can be further subdivided into 3 populations based on expression of CD8α and β. Adoptive transfers indicate that CD8α-β- pDCs enhance allergic lung inflammation and the other two (α+β− and α+β+) contribute to tolerance by inducing FOXP3+ regulatory T cells.^[123] pDCs are functionally distinct from cDCs in that they have a decreased capacity to uptake, process, and present antigens.^{[30, 124, 125]} The primary cytokine pDCs produce is interferon-alpha (IFN-α) and they play a role in antiviral immunity (reviewed^[126–129]) and allergen tolerance through the induction of regulatory T cells.^{[123, 129–132]} In the context of allergy and asthma, pDCs serve to inhibit airway hyperreactivity and allergic lung inflammation through IFN-α-mediated suppression of group 2 innate lymphoid cell (ILC2) survival and function^[133] and to restrain sensitization to otherwise harmless antigens.^[134] Furthermore, increasing the number of pDCs through Fms-like tyrosine kinase receptor-3 ligand (Flt3L) administration or administering antigen pulsed pDCs prior to sensitization decreases allergic lung inflammation whereas depletion of pDCs prior to OVA sensitization worsens allergic lung inflammation.^[135–140] Interestingly, pDC depletion via 120G8 (pDC specific antibody) after OVA sensitization reduces AHR and eosinophils in BAL.^[140] Therefore, pDCs seem to be protective during allergy/asthma initiation and may play a role in maintenance/worsening of disease after sensitization, but this requires further study. Interestingly, there may be age-specific effects as study of neonatal mice which naturally lack pDCs or depletion of pDCs in young mice predisposed to development of severe asthma whereas older mice were protected.^{[129, 141]} Furthermore, this predisposition in neonatal mice was rescued by adoptive transfer of pDCs or administration of IFN-α.^[141] Others have shown in models of neutrophilic asthma (more severe and refractory to steroid treatments) that pDCs are the primary DC subset involved as they are markedly increased in BAL fluid of mice with neutrophilic asthma in contrast to the eosinophilic asthma.^[142] Furthermore, neutrophils have been shown to activate pDCs in both humans and mouse models.^{[143, 144]} Taken together, these data indicate that the role of pDCs in disease may depend on the underlying pathophysiology of the disease (neutrophilic > eosinophilic asthma and chronic>acute) as well as potential functional differences among subsets.

4.3.2. Human pDCs

In humans, pDCs are classically identified by expression of CD123 or CD303+ (blood DC antigen 2 (BCDA2))(Table 1). While pDCs more classically promote a Th1 response,^[145] HDM exposure can elicit pDCs to promote Th2 responses.^{[63, 146, 147]} Some propose that this result is due to contamination of preDC cells which were recently discovered to share classical markers of pDCs and can be differentiated by their expression of CD33 which pDCs lack.^[67] Regardless, the physiologic significance of ability of pDCs to induce a Th2 response is unclear as the clinical data tends to support a protective role for pDCs in the initiation of Th2-mediated disease via induction of regulatory T cells (similar to that seen in mouse models).^{[60, 148–151]} For example, increased numbers of circulating pDCs in children are protective from a diagnosis of asthma in the first 5 years of life whereas decreased pDCs is a risk factor for wheezing, diagnosis of asthma, and increased number and severity of respiratory tract infections.^{[152, 153]} Furthermore, children with atopic asthma tend to have decreased pDCs in circulation.^{[154, 155]} Conversely, in adults with asthma, circulating pDCs are increased compared to healthy controls^{[156, 157]} as are pDCs in BAL fluid collected 24 hrs post allergen exposure.^{[152, 158]} This is mediated by an increase in bone marrow production of pDC and cDC precursors which are then recruited to the lungs.^[86] Even in the absence of allergen challenge, pDCs in the sputum of adults with atopic asthma are increased compared to controls and drastically increase from baseline (up to 4-fold increase) during an asthma exacerbation.^[140] Clinically, pDC lung infiltration has been associated with refractory, non-atopic asthma^[159] and may serve as a biomarker for severe disease. Given the growing focus on asthma endotypes (particularly separating atopic from non-atopic asthma), the mouse data showing increased role for pDCs in non-atopic asthma, and reports in humans of pDC numbers correlating with severity of inflammation and increased risk for asthma exacerbations in some populations, there is compelling reason for further investigation of pDCs in asthma. They may play a more active role particularly in the non-atopic endotype than previously recognized.

4.4. moDCs

4.4.1. Mouse moDCs

moDCs have a considerable heterogeneity and thus are the most difficult to definitively characterize as a distinct population independent of macrophages.^[65] There is considerable heterogeneity between moDCs even within a single organ/tissue under the same conditions and the markers used to define this subset has been subject to frequent revisions/updates.^{[7, 8, 68–73, 160]} There is, indeed, a population of monocyte-derived cells which are considered functionally similar to cDC2s: they express CD11b, CD11c, and MHCII (characteristic of cDC2s) while simultaneously expressing what are considered to be monocyte-specific markers (i.e Ly6c or CD64).^{[8, 71, 161–163]} Functionally, moDCs have been shown to process and present antigen to activate T cells in the context of allergic disease.^{[30, 44, 79, 164]} moDCs play important roles in Th1, Th17, and Th2-mediated inflammation.^{[112, 165–172]} In the context of asthma, the number of moDCs present in the lung and BAL fluid of allergic animals is increased.^{[21, 22]} They have been shown to be both necessary (by depletion with FcεRI antibody^[173]) and sufficient (adoptive transfer of moDCs) for the development of allergic lung inflammation.^{[112, 171, 172, 174]}

Recently, these results have been called into question by a report indicating that inflammatory cDC2s (which have now been shown to share many of the same markers as moDCs including FcεRI) may be present in previous studies of moDCs and that properties such as antigen presentation and migration may have been misattributed to moDCs.^{[7, 71]} Indeed, when CD26+ cells (indicative of cDC lineage) are removed from moDC cultures, moDCs are no longer able to process and present antigens to either CD4+ or CD8+ T cells.^[71] Further studies are needed to understand the differentiation, function and full impact that this finding will have on the field.

4.4.2. Human moDCs

Human moDCs are a versatile subset capable of stimulating CD4+T cells, cross-presenting antigen to CD8+ T cells and producing many cytokines including IL-1, 6, 12, 23, and TNF-α.^[175] Similar to their murine counterparts, human moDCs share many markers and functional roles with cDC2s. These similarities make differentiating the two subsets challenging, but CD14 is often used to identify cells of monocyte lineage and thus differentiate moDCs from cDC2s (Table 1).^{[67, 176, 177]} moDCs cultured from peripheral blood of allergic patients exhibit recruitment to epithelial cells exposed to HDM in in vitro migration assays^[178] as well as increased capacity to induce HDM-specific T cell proliferation.^[179] moDCs, similar to other DC subsets, also undergo complex interactions with the epithelium which differentially regulates moDC gene expression.^{[180, 181]} When exposed to allergen, the DCs also secrete more IL-10 and less IL-12 shifting the immune response to the Th2 response classically seen in allergic disease.^{[63, 146, 147, 182]} Although human moDCs have been studied in culture (derived from peripheral blood monocytes), few studies have looked at moDCs present in the lung and their biology. However, there appear to be links to disease severity because moDCs are increased in the outer wall of the large airways in patients who succumb to fatal asthma exacerbation.^[183] Given the similarities in markers expressed, it is likely that moDCs are included in studies that separate cDCs and pDCs without additional markers to exclude moDCs;^[71] however, there are few human studies dedicated to studying this subset in situ. Given the plasticity of moDCs and their context-dependence, in vitro studies may represent only a fraction of their behavior in situ. We do not yet fully understand the breadth of the roles this subset may play and more work dedicated to moDCs is needed.

5. Conclusion

With asthma being increasingly recognized as a heterogenous disease composed of multiple endotypes with distinct underlying pathomechanisms, it becomes increasingly important to have mouse models for each of the endotypes to facilitate design of targeted interventions (reviewed^{[25, 184]}). Clinicians strive to allow the asthma endotype of the patients to help inform treatment; however, there currently exists a dearth of targeted therapies to address the clinical heterogeneity present.

The basis of current asthma treatment is symptom control with inhaled beta-adrenergic agonists and steroids. There are disease modifying treatments focused at Th2 inflammation such as monoclonal antibodies against IL-4R, IL-5/5R, and IgE.^{[16, 185]} In the past few years, DC targeted therapies are being developed which show promise in mouse models and are a focus of ongoing investigation but none are yet approved for use in humans.^{[102, 185–190]} Understanding of DC biology, the role of DC subsets, and utilization of mouse models which phenocopy specific endotypes of human disease will allow for generation and testing of targeted therapeutics to improve the lives of patients.