Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cells in human blood constitute a rare population of mast cell progenitors

Mast cells are infamous for their role in allergic disease, and their activation can lead to a severe life-threatening condition, an anaphylactic reaction.¹  Most recognized is the powerful mast cell activation caused by allergen cross-linking of immunoglobulin E–loaded high-affinity immunoglobulin E receptors (FcεRIs), which leads to the release of an array of different mediators.²  In allergic asthmatics, mast cell mediators such as histamine and prostaglandin D₂ are secreted rapidly after allergen provocation.^3-5  These mediators are devastating to the asthmatic lung causing, for example, bronchoconstriction.^6,7  In comparison with healthy individuals, the mast cell numbers are increased in the airway smooth muscle⁸  and alveolar parenchyma⁹  of asthmatics. As a consequence, a high number of mast cells can be activated during allergen exposure, and the symptoms can be severe.

Mast cells originate from the bone marrow but are absent in the blood in their fully granulated mature state. In mice, mast cell progenitors are present in the circulation and mature on arrival in the peripheral tissues.¹⁰  Progenitors committed to the mast cell lineage can be found in the bone marrow^11,12  and circulate in the blood of naïve mice at very low frequencies as lineage-negative (Lin⁻) c-kit^hi (CD117^hi) ST2⁺ integrin β7^hi CD16/32^hi FcεRI⁺ or FcεRI⁻ cells.¹³  Virtually all mouse mast cell progenitors express FcεRI once entering peripheral tissues, such as the lungs and the peritoneal cavity.¹⁴  In mice with experimental allergic asthma, mast cell progenitors are recruited to the lung¹⁵  and give rise to increased numbers of lung mast cells.^16-18  In humans, mast cells can be derived from CD34^+19,20  and CD34⁺ CD117⁺²¹  progenitor cells in peripheral blood by in vitro culture. However, whether human mast cells originate from a distinct population of progenitors has not previously been determined. Identification of the ancestor of mast cells is important for understanding the underlying mechanisms of allergic disorders and hematologic diseases such as systemic mastocytosis. Possibly, such progenitors would be a novel drug target in mast cell–related diseases.

Because FcεRI is involved in allergen-induced mast cell activation in asthma, the goal of the present investigation was to identify novel human blood mononuclear cell populations that could give rise to CD117⁺ FcεRI⁺ mast cells. In vitro culture of prospectively isolated CD34⁺ blood progenitors showed that the CD117⁺ FcεRI⁺ mast cell–forming potential was mainly found in the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cell fraction. This population of blood cells contained high levels of mast cell–associated genes in comparison with human blood basophils and had detectable levels of messenger RNA (mRNA) transcripts of, for example, tryptase. Collectively, the data suggest that this rare population of blood cells constitutes precursors to human tissue mast cells.

Blood samples were obtained from 13 patients with allergic asthma (median age 24 years, range 16-36 years; 9 women; median asthma control test²²  21, range 17-25), 1 patient with nonallergic asthma (age 14 years, male, asthma control test 21), and 10 healthy, nonallergic controls (median age 20 years, range 16-35 years; 7 women) who participated in the follow-up of the MIDAS (Minimally-Invasive Diagnostic Procedures in Allergy, Asthma or Food Hypersensitivity) study.^23-25  The MIDAS and the follow-up MIDAS II study were performed in 2010-2012 and 2013-2015, respectively. Two of the subjects with allergic asthma did not have anti-inflammatory treatment (inhaled corticosteroids or antileukotrienes) the year previous to this study, and 4 of the subjects with allergic asthma received only short treatment periods with anti-inflammatory treatment during the past 3 months. The rest of the asthmatic patients reported daily anti-inflammatory treatment during the past 3 months previous to the MIDAS II study. The prebronchodilator forced expiratory volume in 1 second (FEV1) was recorded in all subjects in connection to blood sampling and subsequent mast cell progenitor analysis. The results were expressed as percent of the predicted value.²⁵ 

Lung tissues were obtained from 2 patients undergoing lung resection because of lung cancer. The tissue samples were taken from a tumor-free part of the resected lobe.

Mononuclear cells from blood were enriched from EDTA-treated whole blood using Ficoll-Paque Premium (ρ = 1.077 g/mL; GE Healthcare, Little Chalfont, UK). Platelets were removed by centrifugation. Human lung tissues were processed using the Human Tumor Dissociation Kit, gentleMACS C tubes, and the gentleMACS Octo Dissociator with Heaters (all from Miltenyi Biotec, Bergisch Gladbach, Germany). The 37C_m_LDK_1 program was used. Red blood cells were lysed using an ammonium chloride potassium buffer (0.15 M NH₄Cl, 10 mM KHCO₃, and 1 mM EDTA, pH 7.4). The cells were counted using a hemocytometer.

The following fluorochrome-conjugated antibodies were used: CD4 (RPA-T4), CD8 (RPA-T8), CD13 (WM15), CD14 (M5E2), CD19 (HIB19), CD34 (581), CD45RA (HI100), CD117 (104D2), CRTH2 (BM16), FcεRI (AER-37), and integrin β7 (FIB504) (all from BD Biosciences, Franklin Lakes, NJ, and eBiosciences, San Diego, CA). Staining of cells was performed in phosphate-buffered saline pH 7.4 with 2% heat-inactivated fetal calf serum. Flow cytometry and cell sorting were performed on an LSR II or a FACS Aria III (BD Biosciences). Internal controls were used to set the gates. Cells were sorted into 96-well plates or 60-well Terasaki plates.

Cells were cultured in serum-free Stempro-34 SFM medium (ThermoFisher, Waltham, MA) with 100 U/mL penicillin, 100 μg/mL streptomycin, and 2 mM l-glutamine (Sigma-Aldrich). The medium was supplemented with 30 ng/mL interleukin (IL) 3, 100 ng/mL IL-6, 20 ng/mL IL-9, and 100 ng/mL stem cell factor alone or with addition of 20 ng/mL IL-5, 50 ng/mL granulocyte macrophage–colony-stimulating factor, 20 ng/mL thrombopoietin, 10 ng/mL IL-11, 10 ng/mL FMS-like tyrosine kinase 3 ligand, and 4 U/mL erythropoietin. Cytokines were from PeproTech (Rocky Hill, NJ) and R&D Systems (Minneapolis, MN). HMC-1 cells²⁶  were cultured as described.²⁷ 

Tryptase activity in cytocentrifuged cells was detected with Z-Gly-Pro-Arg-4-methoxy-2-naphthylamide peptide substrate (MP Biomedicals, Santa Ana, CA).²⁸  Standard protocols were followed for May-Grünwald Giemsa staining. Shandon Octospot was used to cytocentrifuge small number of cells (ThermoFisher). Images were captured as described.¹⁴ 

Ninety-one to 100 cells of each population were sorted into separate Terasaki plate wells containing 12 μL direct lysis buffer, provided with the Ovation One-Direct System (Nugen Technologies Inc., San Carlos, CA). Whole-transcriptome analyses were performed as described previously,¹⁴  with the exception that GeneChip Human Gene 2.0 ST Arrays were used. The raw data were normalized using the robust multiarray method.^29,30  The gene expression data have been deposited in National Center for Biotechnology Information’s Gene Expression Omnibus database (accession number GSE69030; http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE69030). A gene set enrichment analysis (GSEA) was performed as described previously.^31,32  Gene sets <15 and >200 were filtered out. To establish the rank lists, default settings were used. Processes with an enrichment score >0.5 with a false discovery rate q value <0.05 were considered active. For quantitative reverse transcription–polymerase chain reaction (qRT-PCR), RNA was isolated from lysates of Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cells and HMC-1 cells using NucleoSpin RNA II (Macherey-Nagel, Dueren, Germany), and first-strand complementary DNA (cDNA) was synthesized using total RNA as template by iScript cDNA synthesis kit (Bio-Rad, Hercules, CA) according to the manufacturer’s instructions. Gene expression levels were determined using the primer pairs in Table 1. The primers for the proteases were previously published.²⁰  qRT-PCR was performed using SYBR GreenER SuperMix (ThermoFisher), 200 nM primers, following recommended PCR cycling conditions. Melting curve analysis was performed to ensure product uniformity.

Informed written consent was obtained from all participants, and the study was approved by Uppsala Regional Ethical Review Board.

The statistical tests used are indicated in the respective figure legends. A P value <.05 was considered significant.

Progenitors with mast cell–forming potential have been found in human peripheral blood. These cells were enriched in the CD34⁺ CD117⁺ CD13⁺ fraction.²¹  In the present investigation, subpopulations of the CD34⁺ CD117⁺ CD13⁺ cells were sorted to identify committed mast cell progenitors. To isolate rare progenitor cell populations, the Lin markers CD4, CD8, CD19, and CD14 were included in the analysis. Lin⁻ CD34^hi CD117^int/hi CD13⁺ cells were divided and sorted into FcεRI⁻ and FcεRI⁺ cells. After 1 week in culture with IL-3, IL-6, IL-9, and stem cell factor, the Lin⁻ CD34^hi CD117^int/hi CD13⁺ FcεRI⁻ cells had formed only 4.4% CD117⁺ FcεRI⁺ mast cells (supplemental Figure 1, available on the Blood Web site). In contrast, Lin⁻ CD34^hi CD117^int/hi CD13⁺ FcεRI⁺ cells had formed 75.3% CD117⁺ FcεRI⁺ mast cells (supplemental Figure 1). FcεRI⁺ cells were also found in the Lin⁻ CD34^hi CD117^int/hi CD13^−/lo fraction (supplemental Figure 2). Therefore, the mast cell potential was evaluated in the total Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ fraction irrespective of CD13 expression (Figure 1A). Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cells had formed 72.9% and 78.8% CD117⁺ FcεRI⁺ mast cells after 2 and 7 days, respectively (supplemental Figure 3). Lin⁻ CD34^hi CD117^int/hi FcεRI⁻ cells were used as control. These cells had only formed 2.7% and 5.7% CD117⁺ FcεRI⁺ mast cells, respectively (supplemental Figure 3). As the mast cell–forming capacity was high among the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cells, their level of commitment was evaluated. Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cells were cultured with a myeloerythroid cytokine cocktail (Figure 1A). Lin⁻ CD34^hi CD117^int/hi FcεRI⁻ cells were cultured as control. The sorted cells were analyzed already after 7 days to enable analysis of short-lived cells that could be present in the culture. The Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cells mainly formed CD117⁺ cells, which contained granules and had a mast cell–like morphology (Figure 1C). Approximately half of the cells were CD117⁺ FcεRI⁺ mast cells (Figure 1C,K-L). In contrast, the Lin⁻ CD34^hi CD117^int/hi FcεRI⁻ cell fraction formed heterogeneous cells, a lower frequency of CD117⁺ cells, and almost no CD117⁺ FcεRI⁺ mast cells after 1 week (Figure 1B,K-L). A distinct population of CD117⁻ FcεRI⁺ cells was also developed. Thus, the blood progenitors within the Lin⁻ CD34^hi CD117^int/hi FcεRI⁻ population likely constitute multipotent progenitors. Some dead cells without stainable nucleus and structures in the cytoplasm were observed in cells cultured for 7 days, both from the Lin⁻ CD34^hi CD117^int/hi FcεRI⁻ and the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cell fraction (supplemental Figure 4).

Depending on culture conditions mast cells derived from peripheral blood cells contain tryptase to a high degree after several weeks in culture.^20,33  To confirm that the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ progenitors could mature in vitro, cell-associated trypsin-like activity was used as a measure of mast cell tryptase content. Already after 17 to 19 days of culture in the myeloerythroid cytokine cocktail, a fraction of the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cells had tryptase activity, indicating that these cells had matured (Figure 1E). As expected, the Lin⁻ CD34^hi CD117^int/hi FcεRI⁻ cells were overall negative for tryptase activity after 17 to 19 days of culture (Figure 1D). However, occasional cells stained positive for tryptase activity (supplemental Figure 5).

Three additional markers were added to the antibody panel in an attempt to enrich the mast cell potential within the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ fraction. CD45RA, which was absent on virtually all of the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cells, was used to exclude granulocyte/monocyte as well as lymphoid progenitors (Figure 1F). Integrin β7, which is expressed by mouse mast cell progenitors, was found on the majority of the Lin⁻ CD34^hi CD117^int/hi CD45RA⁻ FcεRI⁺ cells (Figure 1I). CRTH2, which is expressed on mature basophils and mast cells, was absent on essentially all of the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ progenitors (Figure 1I). Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ CD45RA⁻ CRTH2⁻ integrin β7⁺ cells were sorted and cultured with the myeloerythroid cytokine cocktail (Figure 1J). Lin⁻ CD34^hi CD117^int/hi FcεRI⁻ CD45RA⁻ CRTH2⁻ integrin β7⁻ cells were used as control (Figure 1H). The frequencies of CD117⁺ cells and CD117⁺ FcεRI⁺ mast cells were higher in the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ CD45RA⁻ CRTH2⁻ integrin β7⁺ fraction than in the Lin⁻ CD34^hi CD117^int/hi FcεRI⁻ CD45RA⁻ CRTH2⁻ integrin β7⁻ fraction after 1 week in culture (Figure 1H,J-L). In 1 experiment, the CD117⁺ FcεRI⁺ mast cells even reached 90.4% after 1 week (Figure 1J,L). However, in a second experiment, 16% of the progeny constituted CD117⁺ FcεRI⁺ mast cells (Figure 1L), which revealed a large variation in frequency of CD117⁺ FcεRI⁺ mast cells after in vitro culture between individuals. Even though only 16% of the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ CD45RA⁻ CRTH2⁻ integrin β7⁺ progeny constituted CD117⁺ FcεRI⁺ mast cells in this second subject, 92% were still CD117⁺ cells (Figure 1K). We concluded that even though the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ mast cell progenitors could be classified as CD45RA⁻ integrin β7⁺ CRTH2⁻ cells, these 3 markers did not enhance the frequency of CD117⁺ cells retaining FcεRI expression 7 days postculture of the mast cell progenitors (Figure 1L). In conclusion, the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ characteristics alone are sufficient to identify this rare blood cell population.

To check if single Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ or Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ CD45RA⁻ CRTH2⁻ integrin β7⁺ progenitors could give rise to different cell lineages, single cells were sorted. After 7 days, the number of cells that each single progenitor gave rise to was evaluated (Figure 2A). The median was 3 cells after 1 week in culture, indicating that the progenitors divide slowly. May-Grünwald staining of the cytocentrifuged cells indicated that the vast majority of the cells had developed granules, some of which stained metachromatically. In fact, 24 out of 25 individual Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ or Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ CD45RA⁻ CRTH2⁻ integrin β7⁺ progenitors had exclusively given rise to granulated cells (Figure 2B). From a single well, 1 cell that lacked granules was recovered. No other cells were recovered from the same well, and no well containing both nongranulated and granulated cells was identified. However, because of the slow division of the single cells, it is not possible to draw definite conclusions whether one progenitor could give rise to some other cell type besides mast cells.

Next, Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ and Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ CD45RA⁻ CRTH2⁻ integrin β7⁺ progenitors were sorted and stained with May-Grünwald Giemsa without culturing the cells. These progenitors were small. Some of the cells lacked granules, whereas some had few metachromatic granules (Figure 3A). To provide a morphologic comparison with the blood mast cell progenitors, Lin⁻ CD117⁺ FcεRI⁺ human lung mast cells and Lin⁻ CRTH2⁺ FcεRI⁺ blood basophils were sorted and stained with May-Grünwald Giemsa. In contrast to the mast cell progenitors, Lin⁻ CD117⁺ FcεRI⁺ mature lung mast cells and Lin⁻ CRTH2⁺ FcεRI⁺ mature blood basophils contained numerous metachromatic granules (Figure 3B-C). The lung mast cells and the blood basophils had no or low expression of CD34, which indicates that they were mature (Figure 3D-E). In summary, Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ and Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ CD45RA⁻ CRTH2⁻ integrin β7⁺ progenitors have an immature morphology, but they are in the process of forming granules.

Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ mast cell progenitors and Lin⁻ CRTH2⁺ FcεRI⁺ blood basophils cells were sorted in parallel from 3 blood donors. A whole-transcriptome microarray analysis was performed on these 2 cell populations to investigate the differential gene expression profile. Genes that were expressed highly specifically in human mast cells in the FANTOM (Functional Annotation of the Mammalian Genome) study³⁴  were selected for evaluation. Production of prostaglandin D₂ is catalyzed by hematopoietic prostaglandin D synthase. Transcripts of hematopoietic prostaglandin D synthase (Hpgds) were exceptionally abundant in mast cell progenitors, and Hpgds was the most differentially expressed gene of the whole genome (Figure 4A). Transcripts for the gene coding for CD117 (Kit) and histidine decarboxylase (Hdc), which codes for the enzyme that catalyzes the production of histamine, were higher in mast cell progenitors than in basophils. There was a strong trend toward mast cell progenitors having more transcripts of the β chain of FcεRI (Ms4a2), although not significantly. There was also a tendency that the transcripts for the mast cell α- and β-tryptases (Tpsab1 and Tpsb2) and carboxypeptidase A (Cpa3) were all expressed to a higher degree in the mast cell progenitors than in the basophils, even though there were no significant differences in these transcripts (Figure 4A). To provide further evidence for the commitment to the mast cell lineage, qRT-PCR analysis of proteases that are expressed in mast cells derived from human blood²⁰  and serglycin (Srgn), the core protein for heparin/chondroitin sulfate E proteoglycans, was performed on the primary mast cell progenitors. Tryptase (Tpsab1/Tpsb2), carboxypeptidase (Cpa3), and serglycin (Srgn) were detected (Table 2). In conclusion, the mast cell progenitors express the mast cell–related genes Hpgds, Kit, Hdc, Tpsab1/Tpsb2, Cpa3, and Srgn.

A GSEA was performed to explore what biological processes were active in mast cell progenitors. Many of the processes were involved in protein synthesis and localization (Figure 4B).

As asthmatics have an increased number of lung mast cells, we examined if this phenomenon was reflected by an increased frequency of Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ blood cells. No difference in the frequency of Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ blood mast cell progenitors was found when comparing healthy nonatopic individuals with patients with well-controlled asthma recruited in the MIDAS study^23-25  (Figure 5A). Interestingly, the asthmatics with reduced lung function (FEV1 ≤90% of predicted) had a higher frequency of circulating mast cell progenitors than those with normal lung function (FEV1 >90% of predicted) (Figure 5B). The same pattern was observed among all individuals including healthy controls (Figure 5C).

This study describes a rare population of Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ mast cell progenitors in human blood. As shown earlier, blood progenitors within the CD34⁺ CD117⁺ CD13⁺ population can give rise to mast cells.²¹  In agreement with these results, we found mast cell–forming potential among the Lin⁻ CD34^hi CD117^int/hi CD13⁺ cells, especially in the FcεRI⁺ fraction. However, as we observed that the CD13 gate divided the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cell population, the CD13 marker was not further used to discriminate mast cell progenitors. In the study by Kirshenbaum et al, the CD34⁺ CD117⁺ CD13⁺ cells gave rise to mast cells, but also to CD117⁻ FcεRI⁻ CD14⁺ monocytes, when cultured in vitro.²¹  In our study, the Lin⁻ (CD14⁻ CD4⁻ CD8⁻ CD19⁻) CD34^hi CD117^int/hi FcεRI⁺ progenitors gave rise to few (if any) CD117⁻ FcεRI⁻ cells. In contrast, Lin⁻ CD34^hi CD117^int/hi FcεRI⁻ cells formed a clear population of CD117⁻ FcεRI⁻ cells upon culture. Therefore, it is likely that the gating of FcεRI⁺ cells from the Lin⁻ CD34^hi CD117^int/hi progenitors separates the mast cell lineage from the monocyte lineage.

Tryptase⁺ (mature) mast cells without FcεRI expression have been reported in vitro,³⁵  and mast cells without FcεRI expression can be found in vivo.³⁶  In our investigation, the focus was to identify progenitors that give rise to CD117⁺ FcεRI⁺ mast cells. As Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ did not form a 100% pure CD117⁺ FcεRI⁺ progeny after culture, attempts to further enrich the mast cell purity were made. The Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cells were gated based on expressions of CD45RA, CRTH2, and integrin β7. However, the progeny of the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ CD45RA⁻ CRTH2⁻ integrin β7⁺ cells still showed a variable phenotype with 16% to 90% CD117⁺ FcεRI⁺ cells, although nearly all were CD117⁺ after 7 days in vitro. Yet, apart from some dead cells, they all had an immature mast cell–like morphology. Because the sorted cells were cultured in a cytokine cocktail, it is highly unlikely that all CD117⁺ FcεRI⁻ cells at day 7, formed from either Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ CD45RA⁻ CRTH2⁻ integrin β7⁺ cells or Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cells, were multipotent progenitors. Rather, the CD117⁺ FcεRI⁻ cells may constitute immature mast cells that lost FcεRI expression during the culture.

The whole-transcriptome microarray analyses verified that the Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ progenitors were expressing mast cell–related genes and showed that they were actively differentiating. The genes Hpgds, Kit, and Hdc, which are highly expressed by mast cells according to the FANTOM consortium study,³⁴  were significantly more expressed in Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cells than in basophils. Furthermore, mRNA for Tpsab1/Tpsb2, Cpa3, and Srgn, but not Cma1, could be detected in the mast cell progenitors using qRT-PCR. Human mast cells are divided into 2 subtypes (ie, tryptase- and chymase-positive mast cells and tryptase only–positive mast cells). Both subtypes also contain transcripts of carboxypeptidase A (Cpa3).³⁷  Because the mast cell progenitors had undetectable levels of chymase transcripts, this expression is likely turned on later during mast cell development, although the possibility that tryptase- and chymase-positive mast cells arise from another population cannot be excluded. Altogether, the whole-genome expression microarray analyses and the qRT-PCR data support the conclusion that Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ blood cells are predominantly mast cell progenitors.

There are several similarities between mast cell progenitors in mice and humans such as expression of CD117 and integrin β7, and the majority of murine BALB/c blood mast cell progenitors express FcεRI.¹³  Approximately 0.0045% of the blood mononuclear cells in these mice are committed mast cell progenitors,¹³  which is comparable to the median value of 0.0053% Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ cells observed in healthy humans. A major difference between mice and humans is the proliferative capacity of the mast cell progenitors. Mouse mast cell progenitors undergo extensive proliferation when cultured in vitro. In fact, 1 single mast cell progenitor gives rise to roughly 10³ mast cells after 7 days in culture.³⁸  Here, human blood mast cell progenitors divided poorly. Sorted single cells became in median only 3 cells after 1 week of culture. We speculate that this may be because either an essential unknown growth factor is to be found for optimal growth in vitro or the human blood mast cell progenitors have lost a great part of their proliferative ability during their differentiation.

The ontogeny of mast cells is still to be traced. In humans, the lymphoid-primed multipotent progenitor branch is characterized by expression of CD45RA.³⁹  Given that Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ mast cell progenitors were CD45RA⁻, and bipotent mast cell/monocyte progenitors have been described,²¹  it is reasonable to assume that the mast cell/monocyte axis branches off directly from multipotent progenitors. Studies delineating the origin of human mast cell progenitors are awaited.

As asthmatics have an increased number of mast cells in the airways,^8,9,40  we determined whether this was reflected by a higher frequency of blood mast cell progenitors in asthmatics than in healthy controls. We conclude that there was no major difference in the frequency of human blood mast cell progenitors between the 10 healthy subjects and 14 asthmatics. This may reflect that the asthmatics had good asthma control. Nevertheless, when the individuals with normal lung function and those with reduced lung function were grouped among asthmatics or all individuals, those with reduced lung function had a higher frequency of circulating mast cell progenitors than those with a normal lung function. Notably, the differences were significant even for a relatively small number of individuals analyzed. High alveolar mast cell numbers are found in uncontrolled asthmatics with reduced lung function.⁹  Therefore, a high blood mast cell progenitor frequency may reflect ongoing recruitment to the lung resulting in increased mast cell density, which could contribute to the reduced lung function. Quantification of Lin⁻ CD34^hi CD117^int/hi FcεRI⁺ mast cell progenitors may also be clinically important in patients with, for example, mastocytosis and parasitic infections. CD34 is lost during mast cell maturation.⁴¹  Therefore, the mast cell progenitors identified in the present study are likely more immature than CD34⁻ CD117⁺ blood cells with mast cell–forming capacity quantified in mastocytosis patients.⁴² 

To summarize, we have identified a rare population of human mast cell progenitors in blood, which gives rise to predominantly mast cells, although the limited cell division capacity and partial loss of FcεRI after culture in vitro prevent us from concluding their full commitment to the mast cell lineage. However, none of our data suggest that another lineage develops from this cell population. Our finding provides a novel opportunity to phenotype immature mast cells in diseases by a blood test. It also enables the mast cell progenitors to serve as a therapeutic target in diseases characterized by an increased mast cell density in peripheral organs and as a diagnostic tool to identify aberrancies in the mast cell lineage development.

The authors thank Pia Kalm-Stephens for providing the blood samples and Birgitta Heyman for critical reading of the manuscript. The gene expression microarray analyses were performed by the Array and Analysis Facility at the Science for Life Laboratory (Uppsala Biomedical Center, Uppsala, Sweden). We performed the fluorescent-activated cell sorting on equipment provided by the BioVis Facility at the Science for Life Laboratory, Uppsala, Sweden.

This work was supported by grants from Agnes and Mac Rudberg Foundation, Lennander’s Foundation, Ellen, Walter, and Lennart Hesselman Foundation, and the Royal Swedish Academy of Sciences (J.S.D.); and from the Swedish Research Council, Malin and Lennart Philipson Foundation, the Swedish Heart-Lung Foundation, Konsul Th C Bergh Foundation, Bror Hjerpstedt Foundation, and Mats Kleberg Foundation (J.H.). The build-up of the MIDAS cohort was supported by the Swedish Governmental Agency for Innovation Systems.

Contribution: J.S.D. and J.H. designed the experiments and wrote the manuscript; J.S.D. and H.Ö. performed the experiments; J.S.D., H.Ö., and J.H. analyzed the experiments; A.M., C.J., and K.A. were responsible for recruitment of patients and control subjects; and M.S. coordinated the collection of the lung samples.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

The current affiliation for J.S.D. is Clinical Immunology and Allergy Unit, Department of Medicine, Karolinska Institutet and Karolinska University Hospital, Stockholm, Sweden.

Correspondence: Jenny Hallgren, Department of Medical Biochemistry and Microbiology, Uppsala University, Biomedical Center, Husargatan 3, 75123 Uppsala, Sweden; e-mail: jenny.hallgren@imbim.uu.se.