The ingenious mast cell: Contemporary insights into mast cell behavior and function

1.1 The origin of mast cells

Hematopoietic cells arise in several temporally distinct waves during prenatal development. The first MCs originate from primitive erythromyeloid progenitors in the extraembryonic yolk sac.^42-45 A second wave of MCs appears together with the first definitive hematopoietic progenitors. The first two waves mainly contribute to connective tissue MCs (CTMC) and serosal-type MCs.⁴³ The third hematopoietic wave originates from the aorta-gonado-mesonephros region. Cells formed during this wave produce hematopoietic stem cells that exhibit MC-forming potential.⁴³ It was for a long time assumed that MCs develop from hematopoietic stem cells in the bone marrow.⁴⁶ However, it is now clear that bone marrow hematopoietic stem cells produce a fraction, but not all, MCs. Experiments also show that bone marrow-derived MCs mainly replenish the mucosal MC population postnatally.⁴³ However, there is a redundancy in the potential of the waves as bone marrow-derived MCs can reconstitute depleted skin-resident (connective tissue-type) MCs and populate the skin following inflammation.⁴⁷

1.1.1 From hematopoietic trees to differentiation landscapes

The classic tree-like model that describes hematopoiesis assumes a stepwise loss of lineage potential with differentiation from hematopoietic stem cells to lineage-committed progenitors (Figure 4A). However, assignment of the MC differentiation trajectory to either the granulocyte-monocyte or the megakaryocyte-erythrocyte branch is a controversial topic. Studies supporting the idea that MCs belong to either branch have been presented.^48-53 On a different angle, other investigations propose that MCs branch off early—close to hematopoietic stem cells—and differentiate along a unique trajectory.^{54, 55}

Advances in high-throughput single-cell RNA sequencing and computational biology are now delineating the MC differentiation trajectory by suggesting a revision of the hematopoietic model.⁵⁶ The idea to reconstruct hematopoietic differentiation with single-cell transcriptomics is based on capturing a snapshot of single differentiating progenitors.⁵⁷ Such analysis reveals that differentiating cells traverse a continuous landscape of states, from multi- and bipotent progenitors to lineage-committed cells (Figure 4B). A single-cell transcriptional landscape of bone marrow hematopoiesis in mouse shows that the MC developmental trajectory is positioned adjacent to the erythrocyte and basophil trajectories.^{51, 58} The MC and basophil trajectories are particularly close to each other,^{58, 59} which indicates the presence of a bipotent basophil-MC progenitor population, a finding that is recapitulated in single-cell fate assay experiments.^{49, 58-60} Cell fate assays of mouse bone marrow progenitors also verified that basophil/MC differentiation is closely coupled with erythropoiesis.⁵¹ The combination of single-cell transcriptomics and lineage tracing provides even deeper insights into hematopoiesis with focus on MC differentiation. Tracing individually barcoded mouse hematopoietic progenitor cells during in vitro culture confirms a coupling between MC and basophil fates as well as between MC and erythrocyte fates.⁶¹

In human, single-cell transcriptomics analysis of bone marrow cells indicates that basophils and MCs arise from a common bipotent progenitor.⁵⁹ However, cell fate assays are yet to verify the existence of such progenitor. A link between basophil/MC progenitors and erythrocyte progenitors has also been proposed in human hematopoiesis.^62-65 In analogy to lineage tracing experiment in mouse, tracking a somatic mutation in human hematopoiesis could potentially decipher the relationship between basophil and MC differentiation. The cellular distribution of KIT D816V mutation, a mutation present in systemic mastocytosis patients causing constitutive KIT signaling, has been studied for this purpose. No evidence supporting a close developmental relationship between MCs and basophils has been observed using this approach.⁶⁶ In fact, the results based on charting the cellular distribution of the KIT D816V mutation seem to suggest that basophils and MCs are distantly related. However, such analysis rests on a disputable assumption; presence or absence of the somatic KIT D816V mutation in a given progenitor does not influence the cell's fate potential. Specifically, it is possible that the acquired D816 V mutation primes a multi- or bipotent progenitor toward the MC fate, given that KIT signaling promotes MC differentiation.⁶⁷ This latter idea supports a close association between basophil and MC differentiation in health and explains the seemingly contradictory results in disease.

In summary, single-cell RNA sequencing has revolutionized our understanding of MC differentiation. The data reveal an overall landscape of hematopoiesis and an association between the MC, erythrocyte, and basophil granulocyte differentiation trajectories. The landscape model of hematopoiesis unifies old and new theories of the MC differentiation field and constitutes a foundation for future research.

1.2 Mast cell heterogeneity—how different are they from each other?

MC heterogeneity was first described in the mid-60 s by differences in histochemical staining features, and the concept of connective tissue mast cells (CTMCs) and mucosal mast cells (MMCs) was born.^{68, 69} In humans, different MC subpopulations have been defined by their protease content; those that express tryptase (MC_T), tryptase and chymase (MC_TC), and chymase only (MC_C).^{70, 71} The latter was, for a long time, controversial, and the presence of chymase-only positive MCs has remained ignored and virtually unstudied. The two distinct MC subpopulations, MC_T, and MC_TC, as well as the MMC and CTMC, differ in their localization, mediator content, and responsiveness to secretagogues.⁷²

Besides dividing MCs into subpopulations based on their localization or protease content, there are also definitions of MCs being either constitutive or inducible, inflammatory (iMC), or pro- or anti-tumorigenic (MC1 or MC2).^{35, 73-75} While the division into constitutive and inducible subpopulations might be related to the prenatal origin of the MC (see above), the latter is probably dependent on changes in the microenvironment during an inflammation. It is clear that MCs in different organs differ in their receptor and mediator expression, but also within a single tissue there is a considerable heterogeneity, for example, among human lung MC_T.⁷⁶ Furthermore, the gene/protein expression within this MC_T subpopulation changes during inflammatory responses.^{75, 77} A question to address is if these changes relate to different subtypes or a plasticity within the MC population.

The origin and development of different MC subpopulations have been enigmatic, but one clue was described recently as mentioned above.⁴⁵ For hematopoiesis in the adult, the question has been if there are several different MC progenitor populations in circulation, or if there is one progenitor population that has the capacity to differentiate and mature into any of the MC subtypes. In other words, is the heterogeneity driven by locally produced factors or is it driven by the recruitment of different types of designated progenitors? In a study where this question was addressed, the results suggested the existence of a common MC progenitor that gives rise to all MC subpopulations.⁷⁸ However, single-cell RNA sequencing and single-cell cultures will likely provide further proof and insights into this process.

The separation of MCs into different subpopulations based on their protease content or histochemical properties is rather simplistic. MCs show a great plasticity, and detailed analysis of protease expression in human lung MCs demonstrate a gradient in the expression of the different proteases in these cells. Transcriptome comparisons of MCs from different mouse tissues with other immune cells revealed that MCs form a distinct population well separated from all other immune cells.⁷⁹ Within the MC population, there was a substantial heterogeneity across tissue. Similarly, proteome analysis of human skin and fat MCs confirmed the unique identity of MCs among immune lineages, including the granulocytes (neutrophils, eosinophils, and basophils).⁸⁰ In this study, there were few differences in the proteome between skin and fat MCs, raising questions on the impact of the microenvironment for MC heterogeneity.⁸⁰ Analysis of single-cell RNA sequencing data of MCs from human tissues will further decipher MC heterogeneity and plasticity and might provide a clearer view about differences among MCs in different tissues, and also within the same tissue.

1.3 Mechanisms behind directions that mast cells take during their life cycle

1.3.2 To migrate or to degranulate, that is the question

To avoid loss of their munitions before reaching their final destination at inflamed sites, mature MCs need to be regulated to migrate, but not yet degranulate⁹⁰ (Figure 5). This is a challenging requirement as some of the MC chemokines induce degranulation of basophils⁹¹ and synergize with secretagogues to potentiate MC degranulation.^{92, 93} Thus, it was proposed that chemokines may either elicit distinct signals in MCs, as opposed to basophils or that MC SGs are linked to the cell cytoskeleton differently from the basophil granules.⁹⁰ Marked differences observed between the actin skeleton in migrating versus secreting MCs support the latter possibility.⁹⁴ Analysis of actin rearrangements following chemokine stimulation of MCs revealed an accumulation of pericentral actin clusters that prevent cell flattening and converge the SGs in the cell center.⁹⁴ By contrast, reduction in the actin mesh density characterizes the secretory actin phenotype. Thus, the migratory actin phenotype immobilizes the secretory granules by trapping them in the cell center, whereas the secretory actin phenotype supports mobility and exocytosis. Diaphanous-related formin, mDia1, appears to be key player in these actin rearrangements⁹⁴ (Figure 5).

1.4 Mast cells in diseases

MCs are involved in the initiation and perpetuation of a number of inflammatory conditions. These conditions range from those associated with an intrinsic or primary defect in MCs due to inherited or acquired polymorphisms and mutations within the MC compartment; to diseases where MCs are activated and recruited through an extrinsic mechanism such as formation of antigen-dependent and IgE-mediated MC activation; to clinical conditions where MCs are implicated through the release of MC mediators but where the mechanism of activation and recruitment is not well understood, thus “idiopathic” (Table 1).¹⁰⁸

There are several well-characterized molecular aberrancies affecting the MC compartment. Some are associated with recognized clinical diseases due to effects on MC proliferation and survival, MC reactivity, or MC mediator production. Mastocytosis is perhaps the most widely known disease associated with a primary defect in the MC compartment. It is a clonal disease involving expansion of tissue MCs. Mastocytosis is most commonly associated with an acquired gain-of-function KIT p.D816V missense variant, resulting in ligand-independent MC activation. This leads to both unrestrained growth of MCs and a lowered threshold for activation. Individuals with mastocytosis may present with flushing, pruritus, gastrointestinal complaints, or systemic anaphylaxis that may occur following exposure to Hymenoptera venom or for unidentified reasons. Mastocytosis may present as cutaneous disease only, or as a systemic disease with or without cutaneous manifestations.¹⁰⁹

A group of patients with recurrent anaphylaxis has clonal MCs as demonstrated by evidence of one or two minor criteria for mastocytosis including aberrant MC morphology, CD25 expression, and/or presence of the KIT D816V point mutation. By consensus, such patients are currently said to have monoclonal MC activation syndrome (MMAS).¹¹⁰ If the marrow findings are observed in the absence of evidence of systemic MC activation, the term monoclonal MC disorder of uncertain significance has been suggested.

There are now a growing number of heritable genetic conditions that lead to increased MC reactivity. One such example is manifested as a physical urticaria. A missense substitution from cysteine to tyrosine (pC492Y) in the adhesion G protein-coupled receptor E2 (ADGRE2) is present in autosomal dominant vibratory urticaria (VU) characterized by localized hives and systemic manifestations in response to a local stimulus of frictional nature.¹¹¹ ADGRE2 (CD312) belongs to a large family of adhesion GPCRs, generally with an extracellular domain facilitating interactions with proteins from the extracellular matrix. In this case, activation of MCs after a vibratory stimulus is evidenced by MC degranulation and an increase in histamine in the venous blood from the affected areas. Thus, ADGRE2 functions as a mechanoreceptor and induces cutaneous MC degranulation. Although the physiological relevance of the limited MC responses to friction in normal individuals is not completely understood, possibilities are that ADGRE2 may alert both resident and immune cells to combat potential injury and wound healing. It may also play a role in pain modulation and perhaps help sense a parasite migrating through dermal tissues.

Hereditary alpha tryptasemia (HαT) is a term used to describe a genetic trait caused by an increased TPSAB1 copy number encoding alpha-tryptase that leads to elevated basal serum tryptase levels in 4–6% of Western populations.^{112, 113} Individuals with multiple duplications of alpha-tryptase are reported to have a higher risk for severe anaphylaxis. Further, the prevalence of HαT in clonal MC disease is twice that of the general population and may be a biomarker for severe mediator-related symptoms in those with mastocytosis.^{114, 115} Recent mechanistic studies have demonstrated that unique enzymatic properties of alpha-tryptase containing heterotetrameric tryptases may contribute to this association.¹¹⁶ At present, HαT is thus best thought of as a disease-modifying genetic finding.

Extrinsic or secondary MC activation occurs primarily in allergic diseases, diseases associated with complement activation, and in association with activation of MCs through MRGPRX2. Symptoms may be infrequent to frequent and resultant disease sporadic or chronic depending on the activating mechanism. The immediate effects of MC degranulation, if localized to skin, include a wheal and flare reaction or, in airways, contraction of airway smooth muscle, mucus secretion, and an increase in vascular permeability. If systemic, the results may include severe hypotension and extensive vascular leakage. The early responses often transition into a late phase reaction hours later associated with an influx of circulating cell types which promote further inflammation.

The idiopathic MC category includes urticaria, angioedema, and anaphylaxis where there is no identifiable etiology, but where MC activation is documented through MC mediator release or evidence of MC degranulation in tissues involved (Table 1). The term idiopathic MC activation syndrome (Idiopathic MCAS) has been applied as a diagnosis for individuals who present with such episodic allergic-like signs and symptoms such as flushing, urticaria, diarrhea, and wheezing involving two or more organ systems, where the etiology is unknown.¹¹⁷ Diagnostic criteria include response to anti-mediator therapy and an elevation in a validated urinary or serum marker of MC activation, such as serum tryptase with an episode. Primary and secondary MC disorders must be eliminated as possible causes of the clinical findings. However, the search should continue for the etiology of these idiopathic disorders including the possibility that MC activation may relate to a yet-to-be-identified endogenous or environmental stimulus or an intrinsic MC defect resulting in a hyperactive MC phenotype.

1.5 The development of mast cell-targeted treatments

1.5.2 Treatments that inhibit mast cell activation

More recent attempts to target MCs therapeutically aim at the inhibition of activating signals and receptors (Figure 6). These include, for example, the high-affinity receptor for IgE, FcεRI, as well as the intracellular signals involved in translating receptor activation into degranulation and mediator release. Examples for the latter include Bruton's tyrosine kinase (BTK) and spleen-associated tyrosine kinase (SYK) (Figure 7). Inhibitors of BTK or SYK inhibit the degranulation of human MCs induced via FcεRI. Two BTK inhibitors, Fenebrutinib and Remibrutinib, as well as the SYK inhibitor GSK2646264, are currently under development for the treatment of patients with CU (Figure 7).

In chronic spontaneous urticaria (CSU), the activation of skin MCs via FcεRI, either by IgE to autoallergens or by autoantibodies to its alpha chain, is held to drive the development of signs and symptoms, itchy wheals and angioedema, in most patients.^124-126 Treatment with omalizumab, an anti-IgE antibody, is effective in CSU¹²⁷ and prevents MC activation by reducing the levels of free IgE and FcεRI expression. Newer anti-IgEs, including ligelizumab and GI-301, with higher affinity to IgE than omalizumab, are in clinical development.¹²⁸

Several other activating signals and receptors are held to contribute to the activation and degranulation of MCs and, thereby, to the development of signs and symptoms of MC-driven diseases. These include the receptors for thymic stromal lymphopoietin (TSLP), IL-33, IL-4, IL-13, IL-5, and complement C5a as well as MRGPRX2. For all of these receptors, inhibitory compounds (or compounds that inhibit their ligands) are currently in clinical development for MC-driven diseases, primarily CU.¹²⁹

The benefit of targeting MC-activating signals, receptors, and pathways is that this approach protects from the release and, therefore, the effects of all MC mediators rather than only one. The limitation is that only one of many pathways of MC activation and degranulation is shut down. Other pathways are untouched, remain viable, and can lead to MC activation, as they often do in most MC-driven diseases.

1.6 Animal models to study mast cell function

Numerous types of animal models, primarily in the mouse, have been utilized throughout the years to outline the contribution of MCs in diverse pathological settings. In the first generation of such models, MC deficiency in the mouse and rat was due to various mutations in Kit, that is, the receptor for SCF. Since SCF is an essential growth factor for MCs, defects in Kit result in an essentially complete absence of MCs. However, KIT is also expressed by a number of other cell types, and it has, therefore, been challenging to ascertain that consequences of Kit defects are indeed explained by an impact on the MC niche as opposed to off-target effects on other populations (reviewed in^{135, 136} ). To account for these issues, new mouse models of MC-deficiency, independent of Kit, have been developed (Table 2). These include mice where MC deficiency is driven by Cre recombinase expression under the control of MC-specific promoters. In one strategy, Cre recombinase was driven by the promoter for Mcpt5, a gene specifically expressed by CTMCs. These mice can then be crossed with R26^DTA mice, leading to constitutive MC deficiency due to MC-specific expression of diphtheria toxin (DT).¹³⁷ Alternatively, the mice can be crossed with the iDTR line, leading to MC-specific expression of the DT receptor; treatment of these mice with DT will thus lead to conditional depletion of MCs.¹³⁷ In another approach, MC deficiency was accomplished by inserting Cre under the control of the promoter for Cpa3, a gene expressed predominantly by MCs.¹³⁸ This leads to constitutive MC depletion, apparently due to Cre-mediated genotoxicity, but also to a substantial reduction of basophils, the latter in agreement with studies showing that basophils express low levels of Cpa3. The Cpa3 promoter was also exploited to generate a mouse strain in which the Cre-LoxP recombination system was used for deletion of the gene coding for the anti-apoptotic factor Mcl-1. This led to an essentially complete absence of MCs but also to a major reduction in basophils.¹³⁹ MC depletion has also been accomplished in a model where the DT receptor gene was expressed under the control of an MC-specific IL-4 enhancer element.¹⁴⁰ Another strategy was to insert the DT receptor and bright red td-Tomato fluorescent protein genes into the gene coding for the β chain of FcεRI, which is expressed by basophils and MCs. This approach can be used for the depletion of MCs and basophils and also as an elegant tool to visualize MCs/basophils in vivo.¹⁴¹ More recently, a mouse line with reduced numbers of MMCs (CTMCs were not affected) was generated by expressing Cre under the control of the baboon-α-chymase gene and crossing these to mice with a floxed allele of Mcl-1.¹⁴²

By using these Kit-independent models of MC deficiency, important insight into the biological function of MCs has been obtained. As expected, the use of Kit-independent mouse models for MC deficiency has firmly confirmed the essential role of MCs in allergic responses.^{138, 143} Moreover, recent studies have shown that MCs can have a role in melanoma dissemination,¹⁴⁴ cutaneous lymphoma,¹⁴⁵ collagen-induced arthritis,¹⁴⁶ bone fracture-associated inflammation,¹⁴⁷ and bone healing.¹⁴⁸ It was also demonstrated that MCs aggravate osteoarthritis¹⁴⁹ and can mediate the detrimental impact of smoke components on asthmatic features.¹⁵⁰ Further, it has been demonstrated that MCs have a beneficial role in controlling bacterial clearance and promoting wound healing after Pseudomonas aeruginosa infection,¹⁵¹ whereas a detrimental impact of MCs was seen in skin infection by Sporothrix schenckii.¹⁵²

However, it is important to note that the use of these novel mouse models for MC deficiency has challenged some previous findings where a contribution of MCs in various pathologies has been implied. For example, recent findings based on Kit-independent mouse models for MC deficiency have questioned the role of MCs in certain models for autoimmune diseases¹³⁸ and obesity,¹⁵³ as well as the proposed adjuvant activities of MCs.¹⁴⁶ Hence, a more nuanced view of how MCs are involved in pathological settings is currently emerging.

In addition to the various mouse models described above, recent efforts have resulted in the generation of mice in which the MC niche is populated by human MCs. This was accomplished by transplanting human hematopoietic stem cells into NOD-scid IL2R-γ^−/− mice, and then promoting MC growth by administrating plasmids expressing human SCF, GM-CSF, and IL-3. In these mice (denoted “humice” or NSG-SGM3), mature human MCs (co-expressing KIT and FcεRI) were detected in multiple tissues. These “humice” have so far mainly been used to study the role of MCs in anaphylaxis and cutaneous drug reactions,^154-156 but also in the evaluation of new therapies, for example, using a BTK inhibitor.¹⁵⁷ Moreover, functional human MCs could be developed from the bone marrow of these mice. Clearly, this humanized model has the potential to be used as a highly valuable tool to study human MC function. Another approach for studying the function of MCs is to evaluate mice deficient in various MC mediators, including the MC-restricted proteases, serglycin, and histamine (reviewed in ^{16, 123, 158}).