Besides their well-known role in type I hypersensitivity reactions, mast cells (MCs) are multifunctional, long living sentinel cells with a key role in the innate immune response (1, 2). MCs derive from hematopoietic progenitor cells of the bone marrow (3), and circulate through the blood stream as immature precursors and lymphatics until they reach their final destination such as mucosal and epithelial tissues (4–6).

In the tissue, MC precursor differentiates under the influence of growth factors and cytokines that ultimately determine their mature phenotype (7). Two subtypes of MCs have been well characterized: mucosal MCs and connective tissue MCs (8). In vitro MCs differentiation is mediated by interleukin-3 (IL-3) leading phenotypically to the mucosal MC subtype (9). Murine bone marrow-derived MCs are commonly differentiated in vitro over 4–6 weeks by cultivating hematopoietic progenitor cells from murine bone-marrow in the presence of IL-3 (10). The differentiation status is generally confirmed by CD117/c-kit, a transmembrane protein with tyrosine kinase activity that regulates cell differentiation and proliferation (11). When 95% purity of mature MC suspension is reached, the cells are commonly used for functional studies (12–14). In the hematopoietic system, CD117/c-kit is expressed in most early progenitors including the stem-cell compartment (15). In most lineages, expression of CD117/c-kit is then lost during differentiation; only MCs maintain its expression throughout differentiation and maturation steps (16). For negative selection, Ly6G/C, small GPI-linked proteins on the surface of myeloid-derived cells, are used to identify neutrophils, monocytes, and granulocytes (17, 18).

In vitro experiments on the functionality of MCs are normally performed under normoxic oxygen levels with normoxic differentiated bone marrow-derived cells (oxygen conditions of in vitro cell culture systems: 150 mmHg or 19.95% oxygen, 20.3 kPa). Importantly, maturation of MCs occur in vivo under lower oxygen levels since the destination tissues of MCs, were the final differentiation takes place, vary in their physiological oxygen levels, depending on the local need of oxygen in the tissue (19): bone marrow (approx. 49 mmHg, 6.4%), blood (arterial blood: approx. 100 mmHg, 13.2%), intestinal mucosa and submucosa (approx. 57.6 mmHg 7.6%), alveolar wall and bronchi (gradient between 100 and 120 mmHg in the alveolus, 13.4–16%) (19).

Furthermore, during inflammatory processes, when MCs are involved, the physiological oxygen content of the tissue, termed physioxia, locally decreases to approx. 1% due to oxygen consumption by invading host immune cells and pathogens (20). This decreased physiological oxygen level in the tissue is called hypoxia (19). Interestingly, it has been shown that MCs functions significantly differ under acute hypoxia (1% oxygen) mimicking the oxygen content of acute inflamed tissue (21). MCs seem to rapidly adapt to low oxygen levels to better orchestrate the immune response under hypoxia by avoiding uncontrolled degranulation and tissue damage, e.g., by downregulating important proinflammatory cytokines like TNF-α and increasing the release of histamine (21). TNF-α and histamine are important proinflammatory effectors stored and released upon stimulation (22, 23). These mediators are immediately released by MCs in response to stimuli like bacteria including S. aureus to recruit other effector cells, i.e., neutrophils. Histamine increases the vascular permeability (24), while TNF-α acts as proinflammatory cytokine by recruiting and stimulating phagocytosis and degranulation of neutrophils (25). Since in those studies, the initial differentiation of MCs was performed under normoxic oxygen conditions, the question arises if MCs already display a distinct functional setup and gene expression when they differentiate in vitro in their physiological oxygen milieu.

To maintain homeostasis under low oxygen levels, the transcription factor HIF-1α is known to act as central regulator and key-player of the cellular adaptation to oxygen stress (26–28). HIF-1α activity is regulated by oxygen on the protein level controlling the transcription of numerous target genes: under normoxic conditions HIF-1α forms a complex with VHL for proteasomal degradation (29), whereas it is stabilized under low oxygen conditions and subsequently translocates into the nucleus to fulfill its regulatory functions (30). HIF-1α mediates inflammatory responses in various immune cells including MCs and also affects the bactericidal capacity (21, 30–34). Importantly, our own previous work revealed a minor role of HIF-1α in the short-term response of MCs to hypoxic 1% oxygen level (21). Therefore, it was of special interest to clarify the role of HIF-1α on MCs differentiation at physiological oxygen level.

We hypothesized that physiologic oxygen conditions affect the differentiation, cultivation, and functionality of MCs in vitro. Therefore, we compared the differentiation rate, transcript expression of hif-1α, selected target genes, foxo3 and vhl as well as reactive oxygen species (ROS) production and histamine production as functional markers of MCs differentiated at 7% oxygen (here termed as physioxia) compared to MCs differentiated at atmospheric oxygen level (21%, here termed normoxia).

It is well known that reduced oxygen levels, which predominantly occur during infection and inflammation, have an important impact on immune cell functions in vitro (21, 30–34). However, the importance of oxygen on differentiation of MCs in vitro is still not completely understood, although it is well known that the physiological oxygen level in resident tissues—the place where immature precursors differentiate to mature MCs—are much lower than atmospheric oxygen concentrations (19). Nowadays, the importance of this fact is increasing, especially regarding the reproducibility of in vitro culture systems when focusing on oxygen levels in immunological studies, as recently reviewed by Zenewicz (42). Our own previous studies with neutrophils or MCs revealed that there is an immune cell-specific response to hypoxic (1%) oxygen conditions that differs from the normoxic phenotype (21, 34). Here, we investigated the impact of reduced physioxic (7%) oxygen level on the differentiation of bone marrow-derived hematopoietic stem cells into mature MCs.

By cultivating MCs under physioxia (7% O₂), no difference in cell viability and cell number was observed over 39 days of differentiation compared to normoxic (21% O₂) differentiation (Figure 2A). Interestingly, for other cell types, it is known that physioxia and prolonged hypoxia influence the cellular survival: e.g., human fibroblasts increase their lifespan by 25% under 10% O₂ compared to normoxically cultivated cells (43). For primary neutrophils, it was also shown that culturing under hypoxia protects from apoptosis (33, 44). In stem cell biology, the culturing of cells under low oxygen already found its way into routine. Culturing embryonic stem cells at 5% oxygen retains their pluripotency (45). Importantly, Gornostaeva and colleagues conclude from their results that “the ‘hypoxic’ phenotype of ASCs (allogeneic adipose tissue-derived multipotent mesenchymal stem cells) may be more ‘desirable’ for the interaction with allogeneic immune cells” (46). In addition, oxygen is considered as an important signaling molecule, which has an integral role in the hematopoietic stem cell maintenance as well as in stem cell proliferation and differentiation (47).

Although no difference in cell viability was observed in this study (Figure 2A), the differentiation rate was found to be significantly slower on day 21 and 25 post-isolation under physioxic compared to normoxic differentiated MCs (Figures 2B,C). Interestingly, Köhler et al. (48) showed in 2012 that the differentiation of bone marrow-derived dendritic cells (BMDCs) under hypoxia (1% O₂) leads to enhanced phenotypic maturation into mature BMDCs but reduced cell growth and production of several cytokines. This was independent of HIF-1α, as shown by using mice with a conditional deletion of HIF-1α in dendritic cells. In good correlation to these data, we found that the transcript expression of hif-1α and selected target genes is downregulated in MCs cultured under physioxia (Figures 3A–D). The data support the assumption that under physioxia HIF-1α plays only a minor role during MC differentiation.

Possible candidates for an alternative response to stress stimuli including oxygen stress are Forkhead transcription factors (FOXO) (38). Mammalian cells express four FOXO factors: FOXO1, FOXO3a, FOXO4, and FOXO6. These factors are evolutionarily conserved and also identified in C. elegans, here named as DAF-16. DAF-16 promotes a developmental stage of C. elegans that ensures survival under nutrient-poor conditions (49). FOXO3 is the main factor detected in the thymus and bone marrow (38). Beside their role as tumor suppressors in several systems (50), they are identified to play a role in the adaptation of the cell to hypoxic stress. Thereby, especially FOXO3 antagonizes cMyc function, which results in downregulation of mitochondrial activity, thus preventing hypoxia-induced ROS formation which would consequently result in HIF-1α stabilization (51–53). We show here, that foxo3 expression is significantly increased in MCs cultivated under physioxia (Figure 3E), along with reduced hif-1α expression (Figure 3A), suggesting that FOXO3 plays also a role in differentiation of MCs under physioxia and may also play a role in the downregulation of hif-1α. Controversially to this suggestion, it was published that FOXO3 is activated downstream of HIF-1α in fibroblasts as well as in breast cancer cells (54). Consequently, MCs upregulation of foxo3 might occur in a HIF-1α-independent manner or even HIF-1α degradation occurs. Nevertheless, further studies are needed to unravel the role of FOXO3 and HIF-1α as well as their interplay during physioxic differentiation of innate immune cells.

Liu et al. recently showed in a zebrafish model that FOXO3 directly binds to VHL, an E3 ubiquitin ligase complex that is responsible for HIF-1α degradation. They stated that in zebrafish the disruption of foxo3b (the homolog of foxo3 in human) led to impaired hypoxic tolerance and discussed that this might also be due to the influence of foxo3b on HIF activity via regulation of vhl. The expression of vhl is downregulated in MCs differentiated under physioxia, though this effect is not significant (Figure 3F). Possibly, MCs differentiate under physioxia by expressing foxo3 and vhl, leading to degradation of HIF-1α.

Besides antimicrobial activities of immune cells, like phagocytosis, mediator-, and extracellular traps release, ROS play a central role in directly killing pathogens or mediating inflammation (55). ROS include, e.g., superoxide anions (O2⋅−), hydrogen peroxides (H₂O₂), and hydroxyl radicals (OH^⋅), and others that confer reactivity to different biological targets (56). One example is lysosomal digestion of pathogens after phagocytosis, by enzymes and H₂O₂ (57). Moreover, ROS stimulate the production of proinflammatory cytokines (58) and are discussed as second messenger in signal transduction (59, 60). Increased production results in hyperactivation of inflammatory responses, tissue damage and disease (41). MCs are well known to be strongly involved in acute lung injury and tissue damage due to MC degranulation and systemically circulating ROS (61). The additional production of ROS by activated MCs tremendously alters the risk for injury severity; therefore, ROS-generation needs to be strictly controlled. MCs differentiated under physioxia produce significantly less ROS without or with stimulus (PMA) (Figure 4A). Since it is published that acute hypoxia increases ROS production in MCs and various other cell types (62), ROS production might be otherwise regulated and strictly controlled in MCs during physioxic differentiation. A possible explanation might be the relation of mitochondrial ROS production and FOXO3. As already mentioned, FOXO3a could be a down-regulator of HIF-1α. Interestingly, FOXO3a activation blocks the hypoxia-dependent increase in ROS in colon cancer cells and prevents HIF-1α stabilization (53). In another study, it was demonstrated that FOXO3a protects quiescent cells from oxidative stress by inducing manganese superoxide dismutase (Mn-SOD), leading to less ROS. The fact that hypoxia increases cellular ROS production, could lead to the conclusion that the decreased ROS level under physioxia could prevent detrimental accumulation of ROS in MCs. One hind for this hypothesis is the increased expression level of foxo3 during physioxic differentiation, which could decrease ROS by inducing the expression of antioxidant enzymes, like Mn-SOD. By analyzing ROS production in response to short-term physioxia, we could also confirm that this effect is due to physioxic differentiation.

A key function of MCs is the release of mediators to mediate the recruitment of effector cells and orchestrate the immune response. One of the most important mediators of MCs is histamine, which was shown to be significantly and drastically decreased intracellularly under physioxia, suggesting an impact of physioxic differentiation on secretory granule composition. Histamine increases vascular permeability to enhance the blood flow and to allow the recruitment of other effector cells. Furthermore, histamine enhances epithelial cell mucus production to avoid bacterial residence (1). The observed decreased histamine storage after differentiation under physioxia (Figure 5A) may decrease the vasoactive capability of MCs, thus antimicrobial activity during infection is impaired. The intracellular storage after short-term physioxic incubation was not affected, thus the effect of physioxic differentiation on histamine storage was confirmed (Figure 5B). Nevertheless, histamine storage was reduced after physioxic differentiation. However, the histamine release in response to mastoparan was still observable, thus MCs are still responsive (Figure S1 in Supplementary Material). It was shown in addition that gene expressions of important pro-inflammatory cytokines, i.e., il-6 and tnf-α and also vegf, one of the most important vascular permeability factors, were downregulated after physioxic differentiation (Figures 3B–D). In good correlation to our data, the differentiation of dendritic cells under hypoxia strongly reduced cytokine levels, e.g., IL-6, TNF-α, and IL-1β under low oxygen concentrations (defined as 1% O₂), proving that hypoxia has a major influence on cytokine production (48). We suggest that the physioxically differentiated phenotype with less stored histamine, less ROS production in response to different stimuli and decreased gene expression of hif-1α and target genes, including important proinflammatory cytokines like il-6 and tnf-α reflects the tissue status of MCs in in vivo. Since the differentiation of blood-derived monocytes into tissue macrophages under normoxia increases HIF and its target gene VEGF levels during the differentiation (63), atmospheric oxygen concentrations, namely normoxia, which are normally used to differentiate MCs in vitro, might reflect a stress situation for the cell due to an excessive oxygen level. Normoxia might, thus, resemble physiologically a hyperoxic state that provoke oxygen stress in MCs and explain their significant role in hyperoxia-induced lung injury (64). Oxygen stress may cause an upregulation of intracellular histamine levels, ROS production and various genes with all its consequences for immune cell recruitment, antimicrobial defense upon infection, and impact on tissue injury. However, the detailed consequences of MC differentiation under physioxia on MC functionality in response to infection and inflammation still needs to be investigated more in detail in future studies. Furthermore, it is still unclear if there might be differences to oxygen stress response in mucosal MCs versus connective tissue MCs.

Our data show that physiological oxygen conditions substantially affect the maturation of MCs with regard to differentiation rate, gene expression of hif-1α and its target genes as well as foxo3 and vhl, ROS production and cellular histamine storage. We could shed light on the relevance of experiments focusing on the key role of oxygen in cell culture. Finally, the differentiation under physiological conditions produces a different MC phenotype in comparison to normoxic differentiation, which more reflects the in vivo phenotype preferred for further application-oriented studies.

This study was carried out in accordance with the recommendations of Nds. Landesamt für Verbraucherschutz und Lebensmittelsicherheit. The protocol was approved by the local ethical commission of Lower Saxony.

MK-B, KB-H, HM, and HN: conceived and designed the experiments; HM, KB-H, KS, KM, and HF: performed the experiments; HM, KM, MK-B, and KB-H: analyzed the data; HM, KB-H, and MK-B: wrote the paper. All authors proofread the paper.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.