Mast Cell and Myeloid Marker Expression During Early In Vitro Differentiation from Human Peripheral Blood Mononuclear Cells

Pia Welker, JuÈrgen Grabbe,* Torsten Zuberbier, Sven Guhl, and Beate M. Henz Departments of Dermatology, Humboldt-University, Berlin, Germany; *Medical University, LuÈbeck, Germany

In order to characterize the phenotype of human after 2 wk of culture showed that FceRIa-positive mast cell precursors in the peripheral blood mono- cells were mostly CD14+ (90%),CD64+ (82%),and nuclear fraction and its alterations during in vivo mast CD68+ (52%) on ¯ow cytometry. Intracellular tryp- cell differentiation,cells were studied before and tase activity was ®rst detectable after 1 wk of culture, during culture with stem cell factor or stem cell fac- increased FceRIa expression was only detectable by tor-containing cell supernatants. Prior to culture, week 2. Cultured cells acquired the ability to release 86% of cells were immunoreactive for the monocytic during IgE-dependent stimulation,and marker CD14,slightly fewer for CD11b and CD64, culture with the c-Kit antibody YB5.B8 resulted in a <10% expressed FceRIa,rare cells were CD34+ downregulation of tryptase and FceRIa,but not of (<0,1%), and none stained for CD1, CD33, c-Kit, c-Kit. These data show that human mast cells and tryptase. After 2 wk of culture,there was de novo develop from c-Kit- and tryptase-negative precursors expression of c-Kit (14%±43% positive cells),tryptase in the myelomonocytic fraction of peripheral blood (26%±79%),CD33 (57%),and CD64 (64%),an upre- and that they upregulate,maintain,and share many gulation of FceRIa (23%±52%),CD11b (93%),and phenotypic characteristics of cells from the mono- CD68 (95%),but no expression of CD34. Levels of cyte/macrophage lineage during early phases of in mRNA for FceRIa and c-Kit were detectable prior vitro differentiation. Keywords: c-/FceRI/SCF/tryp- to culture and increased during culture,together tase. J Invest Dermatol 114:44±50, 2000 with de novo expression of tryptase. Double staining

ast cells are bone marrow-derived, tissue SCF is produced by a number of tissue resident cells including resident cells whose numbers are known to ®broblasts, keratinocytes, endothelial, and bone marrow stroma increase in a wide variety of in¯ammatory and cells, supporting the concept that mast cell precursors can neoplastic conditions (Weber et al, 1995; differentiate in situ in diverse organs including the skin. A MMetcalfe et al, 1997). Mechanisms underlying potential role of SCF in the pathogenesis of has these processes, particularly in the human system, are however also been discussed by different groups (Longley et al, 1993; largely unclear (Denburg, 1995; Czarnetzki et al, 1996). One Castells et al, 1996; Henz, 1998). As suggested by its name, possible way for mast cells to increase at tissue sites might SCF is however not mast cell speci®c, as it also affects a involve the in¯ux of precursors from the peripheral blood into number of other cells expressing its c-Kit receptor, including the tissue, with subsequent differentiation of the cells under the hematopoietic stem cells, melanocytes, and germ cells (Grabbe in¯uence of locally secreted growth factors. Stem cell factor et al, 1994a). (SCF), also known as mast cell growth factor, Steel factor, or c- We and others have shown in the past that cells expressing all Kit ligand, has been identi®ed as a potent mast cell growth major features of mast cells can be induced in cultures of the factor (Zsebo et al, 1990; Williams et al, 1990; Takagi et al, monocytic fraction from human peripheral blood (PBMC) under 1990) and appears to be the main mast cell differentiation factor the in¯uence of SCF or SCF-containing ®broblast- and keratino- identi®ed so far in humans (reviewed in Grabbe et al, 1994a). cyte-derived mast cell conditioning medium (Czarnetzki et al, 1983, 1984; Valent et al, 1992; Agis et al, 1993; Grabbe et al, 1994b; Welker et al, 1995). In the bone marrow and in cord blood, mast Manuscript received September 19, 1997; revised September 14, 1999; accepted for publication September 20, 1999. cells have been shown to developfrom CD34+, SCF-responsive, Reprint requests to: Dr. Pia Welker, ChariteÂ, Campus Virchow, multipotent hematopoietic progenitor cells (Kirshenbaum et al, Department of Dermatology, Augustenburger-Platz 1, D 13344 Berlin, 1992; Agis et al, 1993; Rottem et al, 1994; Huang and Terstappen, Germany. Email: [email protected] 1995). The phenotype of the committed, speci®c mast cell Abbreviations: APAAP, alkaline phosphatase anti-alkaline phosphatase; precursor remains elusive however. In peripheral blood that serves FceRI, high af®nity IgE receptor; HPLC, high performance liquid as a direct source of mast cell precursors for diverse organs, chromatography; HFS, human ®broblast supernatants; HKS, human keratinocyte supernatants; Ig, immunoglobulin; LCS, L-cell ®broblast including the skin, CD34+ cells are extremely rare and yield only a supernatant; NGF, nerve growth factor; PBMC, peripheral blood minor fraction of mast cells during in vitro culture, whereas CD14± monocytic cells; SCF, stem cell factor. cells have been described to provide the main proportion of these

0022-202X/00/$15.00 ´ Copyright # 2000 by The Society for Investigative Dermatology, Inc. 44 VOL. 114, NO. 1 JANUARY 2000 MAST CELL PRECURSOR 45 cells (Valent, 1994). In nasal mucosa, a c-Kit+, tryptase±,FceRI± Flow cytometric analysis For ¯ow cytometric analysis, 5 3 105 cells cell has recently been identi®ed and proposed to serve as a were incubated for 60 min at 4°C with the monoclonal receptor antibodies committed mast cell precursor (Kawabori et al, 1997). in 50 ml of phosphate-buffered saline (PBS), containing human AB serum (Seromed) (0.1%). After two washes with PBS, cells were incubated for In order to further delineate the phenotype of the mast cell 45 min at 4°C with a 1:20 dilution of ¯uorescein isocyanate (FITC)- precursor in peripheral blood, which would serve as source of mast conjugated anti-mouse IgG (DAKO, Denmark) or phycoerythrine (PE)- cells in diverse organs including the skin, this study was designed to conjugated anti-mouse IgG (Dianova, Hamburg, Germany). Following focus on the identi®cation of several typical mast cell markers and two washes, cells were suspended in 400 ml PBS containing 0.1% NaN3, of monocytic markers on PBMC prior to and during culture in the ®xed with 50 ml of a 37% formaldehyde solution, and analysed on an presence of SCF or mast cell growth factor-containing condition- EPICES Pro®le ¯ow cytometer (Coulter, Krefeld, Germany). Negative ing media. The data con®rm the previously suggested close controls were done in the absence of antibody, with IgG and isotype- relationship between mast cells and peripheral blood myelomono- matched desmin antibody D33 (DAKO). For ¯ow cytometric analysis, the cytic cells (Czarnetzki et al, 1992, 1983, 1984; Valent et al, 1989). elite workstation software (Coulter) was used. They furthermore underline the central role of SCF compared with Tryptase activity The enzyme activity of mast cell tryptase was detected possible additional mast cell growth factors in this process. by cleavage of the peptide Z-Gly-Pro-Arg-pNA (4 mM) in the presence of heparin (5 mg per ml) and a-1-antitrypsin (2 mg per ml) (all from Sigma, St MATERIALS AND METHODS Louis, MO), as described before (Harvima et al, 1988). Cells were lysed by three cycles of freezing and thawing for determination of intracellular Cells PBMC were prepared from peripheral blood of different healthy tryptase. Activity was expressed as mU per 106 cells. donors by a sequence of differential centrifugation on Ficoll-Hypaque, In order to analyse tryptase activities of FceRI-positive and negative adherence to polystyrene culture ¯asks for 2 h, and subsequent washing cells, PBMC cultured for 14 d were separated using the FceRI antibody with RPMI containing 10% fetal calf serum (both from Seromed, Berlin, 29C6 and magnetic cell sorting with microbeads (MACS) (Miltenyi Biotec, Germany) to remove nonadherent cells (Grabbe et al, 1994b). Human Bergisch Gladbach, Germany), as described before (Zuberbier et al, 1999), leukemic mast cells (HMC-1) were kindly provided by J.H. Butter®eld, yielding >95% purity of either cell population. Minneapolis, U.S.A. (Butter®eld et al, 1988) and human basophilic leukemic cells (KU 812 cells) (Kishi, 1985) were from the Research Assessment of intracellular and released histamine Histamine was Institute in Borstel, Germany. quanti®ed in perchloric acid lysed cells and in cell supernatants, using a modi®ed automated ¯uorometric method as described (Zuberbier et al, Cell culture Cells were kept routinely in mast cell growth medium 1995). For studies of histamine release, cells were preincubated with IgE consisting of Iscove's medium (GIBCO, Eggenstein, Germany) and (1 mg per ml, Calbiochem, Bad Soden, Germany) for 30 min at 37°C, 30% horse serum (Seromed) because this basic medium has been followed by addition of anti-IgE (4000 U per ml, Behring, Marburg, proven in the past to be optimal for supporting the activity of mast cell Germany) and another 30 min incubation. Samples were kept at ±20°C differentiation factors (Czarnetzki et al, 1983). For mast cell until analysis. differentiation, L-cell ®broblasts (LCS), human ®broblast (HFS), or HaCaT keratinocyte supernatants (HKS), obtained from con¯uent culture of cells after a 5 d culture in Dulbecco's minimal essential Isolation of RNA Cells from days 0 and 14 were lysed with 3 M lithium medium (DMEM, GIBCO) with 5% newborn calf serum or 10% fetal chloride and 6 M urea, centrifuged at 20 000 rpm for 60 min and extracted calf serum (both GIBCO), were added to the basic mast cell growth with phenol-chloroform, according to Sambrook et al (1989a). medium. In addition to SCF, LCS contained 250 pg NGF per ml and 500 pg TGFb per ml, HKS 125 pg NGF per ml and 1250 pg TGFb Reverse transcription polymerase chain reaction cDNA was per ml, as measured by commercial ELISA (Biermann, Bad Nauheim, synthesized by reverse transcription of 5 ml total RNA, using a cDNA Germany). These conditioning media were added at a 30% synthesis kit (InVitrogen, Stade). The following sets of oligonucleotide concentration throughout, as lower concentrations had proven in the primers were used to amplify cDNA (expected fragment lengths are given past to be less effective for the induction of mast cell differentiation in parentheses): tryptase, 5¢ GGA GCT GGA GGA GCC CGT GA and 5¢ from PBMC (Czarnetzki et al, 1983; Welker et al, 1995). rh SCF ACC TGG GTA AGG AAG CAG TGG TG (531 bp) (Miller et al, 1989); (Peptro Tech, London, U.K.) was added to the basic culture medium FceRIa,5¢ CTG TTC TTC GCT CCA GAT GGC GT and 5¢ TAC at 100 ng per ml because this concentration has been shown to be equal AGT AAT GTT GAG GGG CTG AG (536 bp); FceRIb,5¢ GGA CAC or superior to other concentrations during mast cell differentiation AGA AAG TAA TAG GAG AG and 5¢ GAT CAG GAT GGT AAT studies (Grabbe et al, 1994a), and because this concentration TCC CGT T (446 bp) (Robertson et al, 1991; SchoÈneich et al, 1992); approximated the SCF concentration in HKS (Grabbe et al, 1996). GAPDH, 5¢ GAT GAC ATC AAG AAG GTG GTG and 5¢ GCT GTA Cells were seeded into the different culture media at 1 3 105 cells GCC AAA TTC GTT GTC (197 bp) (Togumaga et al, 1987); c-Kit, 5¢ per ml and kept under standard conditions, as described (Czarnetzki et al, CGT TGA CTA TCA GTT CAG CG AG and 5¢ CTA GGA ATG TGT 1983; Grabbe et al, 1994b). AAG TGC CTC C (369 bp) (Ratajczak et al, 1992). Ampli®cation was In another set of experiments, cells were cultured with LCS in the done using taq polymerase (GIBCO) over 35 cycles with an automated presence of the SCF-receptor antibody YB5.B8 or an irrelevant antibody of thermal cycler (Perkin Elmer, Germany). Each cycle consisted of the the same isotype (anti-desmin, see below). Both antibodies had been following steps: denaturation at 94°C, annealing at 55°C, and extension at puri®ed by HPLC using ion exchange columns (Pharmacia, Freiburg, 72°C for 1 min each. PCR products were analysed by agarose gel Germany), with subsequent dialysis. The antibodies were added together electrophoresis and enzymatic digestion, using standard techniques with LCS to culture media from days 0±14 at 2 mg per ml. (Sambrook et al, 1989b).

Monoclonal antibodies The antibody against mast cell tryptase (AA1) Statistical analysis Data are expressed as means 6 1 SD, and signi®cant (Walls et al, 1990) was kindly provided by A. Walls, Southampton, U.K., a differences between values were calculated, using the unpaired two-tailed competitively binding antibody against the a-chain of the FceRI (29C6) Student t-test. (Riske et al, 1991) by J. Hakimi, Nutley, U.S.A., a partially competitively binding antibody against c-Kit (YB5.B8) (Lerner et al, 1991; Ashman et al, RESULTS 1994) by L. Ashman, Adelaide, Australia, and an HLA-DR antibody (TuÈ36) by A. Ziegler, Berlin Germany (Ziegler et al, 1986). Antibodies Characterization of monocytic cells prior to culture by against CD1, CD3, CD11b, CD14, CD26, CD33, CD34, CD64, and immunocytochemistry,¯ow cytometry,and PCR Prior to CD68 as well as isotype matched control antibodies against desmin and culture, <10% of adherent PBMC reacted with the FceRIa laminin were all purchased from Dianova (Hamburg, Germany). antibody 29C6 on immunocytochemistry (Fig 1), and none were positive for mast cell speci®c tryptase and for c-Kit. The majority of Immunocytochemical staining Immunocytochemistry using the APAAP technique was performed on cytocentrifuge preparations of cells cells stained with markers of the /macrophage lineage, at days 0 and 14 of in vitro culture. Antibody reactivity was evaluated by two namely CD11b, CD14, CD68, and HLA-DR (Table I). No or independent observers, counting at least 100 cells, recording distinctly hardly any CD1, CD3, CD16, CD33, or CD26-positive cells were stained cells only, as described before (Schadendorf et al, 1991; Hamann detected and CD34+ cells were found at a frequency of <1:10,000 et al, 1994). (100,000 cells counted). 46 WELKER ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Double staining of the FceRIa-positive cells with markers for Quantitative assessment of mRNA of selected molecules and/or macrophages showed that 87% were also typically expressed on mast cells was done by serial dilution, with CD14+, CD68+, 45% stained weakly for CD64. None were cDNA samples of cells being adjusted to equal quantities, reactive for c-Kit or CD34 (Table II). controlled by RT-PCR ampli®cation using primers of the house-keeping gene GAPDH. Density of bands of PCR products attained with these primers was analysed by agarose gel electro- phoresis (Fig 2). With the respective dilution of PBMC cDNA, RT-PCR ampli®cation with FceRIa and c-Kit primers was faintly detectable at day 0, and that of mast cell speci®c tryptase and FceRIb mRNA not at all. Mast cell mRNA and protein expression and associated activity in cultured cells On analysis of mRNA expression of the same mast cell markers in cells cultured for 2 wk, bands for FceRIa and c-Kit were increased and those for tryptase and FceRIb de novo detectable (Fig 2). This mRNA expression was paralleled by increased immunoreactivity of PBMC cultured in the presence of SCF, LCS, HKS, and HFS at day 14 for FceRIa and tryptase (Fig 1). Increase of tryptase was con®rmed by functional analysis of the enzyme (Table III). C-Kit protein expression was always lower than that of the other two mast cell markers, and tryptase expression was signi®cantly increased in HKS and HSF, compared with SCF or LCS (Fig 1). Further studies on the kinetics of induction of tryptase and FceRIa expression in PBMC cultured in LCS showed that neither molecule was detectable at days 2 and 4 of culture. Low levels of tryptase activity were ®rst noted on day 7, and FceRIa was ®rst detected on analysis of cells cultured for 2 wk. Both tryptase and FceRIa were further upregulated during culture of PBMC for up to 52 d (three different experiments), resulting in an approximately 2-fold further increase of FceRIa and tryptase positive cells, Figure 1. Percentage of immunoreactive PBMC to antibodies against FceRIa (29C6),tryptase (AA1),or c-Kit (YB5.B8) at day 0 compared with values on day 14, whereas immunoreactivity for c- and after 14 d of culture with SCF or different conditioning media. Kit remained unchanged or was even slightly decreased (data not Conditioning media contained 30% supernatants from either cultured shown). mouse ®broblasts (L-cells) (LCS), the human HaCaT keratinocyte cell line The functional signi®cance of FceRIa expression was demon- (HKS), or normal human skin ®broblasts (HFS) (means 6 SD, N = 3). strated by stimulating the cells with IgE/anti-IgE. On day 0, there *p<0.05, compared with cells cultured with SCF or LCS. was no histamine release at all, whereas this was clearly so with cells

Table I. Immunoreactivity of adherent PBMC prior to and after 14 d of culture with LCS,as determined by the APAAP methoda

% stained cells

Antibody (speci®city)b Known expression onc PBMC day 0 PBMC day 14 HMC-1 KU-812

CD1 Langerhans cells, B cells, 0 6 006 006 006 0 dendritic cells CD3 T cells, dendritic cells 0 6 0336 15* 0 6 006 0 CD11b (C3b) Monocytes, , 77 6 15 93 6 5* 15 6 5656 15 macrophages, basophils, mast cells CD14 Monocytes, granulocytes, 86 6 12 80 6 806 006 0 macrophages CD16 (FcgRIII) Monocytes, granulocytes, 0 6 0376 15* 0 6 006 0 macrophages CD33 Monocytes, myeloic cells, 0 6 0576 9* n.d. n.d.d mast cells CD34 Hematopoietic stem cells <0.1 0 6 006 006 0 CD64(FcgRI) Monocytes, macrophages 0 6 0646 11* 0 6 006 0 CD68 Macrophages 77 6 21 95 6 5 100 6 0 100 6 0 TuÈ36 (HLA-DR) Langerhans cells, mono- 65 6 18 95 6 5* 0 6 006 0 cytes, macrophages, mast cells CD26 Activated T cells, B cells, <1 <1 0 6 006 0 macrophages Desmin (IgG1control) 0 6 006 006 006 0 Laminin IgG2control) 0 6 006 006 006 0

aReactivity with human leukemic mast cells (HMC-1) and basophils (KU812) is shown for comparison. Data are expressed as means 6 SD of the percentage of positively staining cells (N=3), *p<0.05. bMarkers coexpressed on mast cells are underlined. cListing of relevant cell types only. dPreviously described to be positive (Hiroshi et al, 1995). VOL. 114, NO. 1 JANUARY 2000 MAST CELL PRECURSOR 47

Table II. Double staining of adherent PBMC prior to and after 14 d of culture with LCS,as determined by ¯ow cytometry a

% positive cells Day 0 Day 14

FceRIa (single staining) 7 6 3346 12 FceRIa+ cells (100%) also positive for: CD14 87 6 5906 4 CD68 16 6 6526 10 CD64 45 6 12b 82 6 15 c-Kit negative negativec CD34 negative negative

aData are expressed as means 6 SD of the percentage of positively staining cells (N=3). bFaint staining only. cNote that intracellular staining was detected with the APAAP technique (Fig 1). cultured for 2 wk. At this time point, the cells also contained increased intracellular histamine stores (Table IV). Spontaneous histamine release from cultured cells incubated in medium alone was close to zero (not shown). In order to further assess whether there was a direct correlation between FceRIa and tryptase expression, cells cultured for 2 wk in LCS were separated into FceRIa positive (40%) and negative (60%) cells by immunobeads and then analysed for tryptase activity. Protease activity was detected in both cell populations, with surprisingly higher values in the FceRI-negative (48 6 10 mU per 106 cells) than in the FceRI- Figure 2. Expression of mRNA for mast cell markers FceRIa, 6 FceRIb,tryptase,and c-Kit compared with GAPDH in PBMC at positive cell fraction (16 6 8 mU per 10 cells). day 0 and after 14 d of culture with LCS,using a semiquantitative Alterations of myelomonocytic markers during culture Of RT-PCR methodology. Representative results from one of three the other markers studied after 2 wk of culture in LCS by experiments. Lane 1, DNA-ladder; lane 2, negative control; lane 3, day 0; immunocytochemistry, CD1 and CD26 remained undetectable, lane 4, day 14. and no CD34+ cells were noted at all among >10 000 cells (Table I). The number of CD14+ cells showed a tendency to decrease but Table III. Total cell numbers and tryptase activity in PBMC stayed at levels above 70%, whereas CD11b, CD68, and HLA-DR before and at different days of culture in SCF-containing positive cells increased to almost 100%. CD3, CD16, CD33, and mast cell conditioning mediaa CD64 were newly expressed in cultured cells (Table I). On comparison of these cells with immature leukemic mast cells and Time of Conditioning Total cell number Tryptase activity basophils, all of the latter expressed CD68, and variable fractions culture medium added (3106) (mU per 106 cells) were also CD11b and CD64 positive. In order to further characterize the potential mast cell precursor, Day 0 ± 12.0 6 1.5 0.0 6 0.0 ¯ow cytometric analysis of double staining cells was done after a Day 7 LCS 5.2 6 1.2 1.5 6 0.5 2 wk culture, using antibodies against the a-chain of the high- Day 14 LCS 1.7 6 0.2 24.0 6 5.1 af®nity IgE receptor and antibodies against selected myelomono- Day 14 HKS 1.2 6 0.1 36.0 6 8.2 cytic marker molecules. Ninety per cent of FceRIa-positive cells Day 14 HFS 1.5 6 0.2 27.0 6 6.9 were also CD14+, 52% CD68+, 82% CD64+, and none CD34+ (Table II). C-Kit protein membrane staining was not detected at aNo tryptase activity was detected at days 2 and 4 of culture. Differences all on ¯ow cytometry (Table II), although cytoplasmatic staining between cells cultured with different media and analysed at day 14 did not reach was seen on immunocytochemistry (Fig 1). Double staining statistical signi®cance (means 6 SD of N=3). experiments with c-Kit were therefore unsuccessful.

Role of SCF and its c-Kit receptor during mast cell Kit, tryptase, CD1, CD3, CD16, CD26, CD33, and the a-chain of differentiation In order to elucidate the role of SCF contained the high-af®nity IgE receptor. This excludes dendritic cells and in LCS with regard to the induction of FceRIa and tryptase, lymphocytes as mast cell precursors. The low level of c-Kit mRNA PBMC were cultured in the presence of LCS and the SCF receptor identi®ed in the PBMC is possibly a feature of the precursors, as antibody YB5.B8 or an irrelevant, isotype matched antibody as with FceRIa, although this might also be due to the presence of control. Compared with cells cultured in the presence of LCS FceRI-positive monocytes previously described in allergic (Maurer alone, both antibodies had no effect on the number and viability of et al, 1994) and at low levels also in non-allergic donors (Reischl cells (not shown). Addition of the SCF receptor antibody to the et al, 1996). The same might apply to histamine, as low quantities of cultured cells, however, markedly reduced the staining of cells with this molecule were detected in the PBMC prior to culture (Table FceRI and tryptase antibodies (Fig 3), whereas cells cultured with III), and monocytes have been described before to express the control antibody developed levels of immunoreactivity at day histamine and histamine decarboxylase (Mirossay et al, 1994). 14 for both FceRIa and tryptase, similar to those of cells cultured The extremely rare detection of CD34+ cells in PBMC prior to in LCS alone (Fig 1). The percentage of YB5.B8 positive cells, culture makes it unlikely that these cells are the sole source of mast however, was not signi®cantly reduced during culture with the cells among PBMC, as we have shown in the past that cell SCF receptor antibody (Fig 3). proliferation, as measured by thymidine incorporation, occurs only at a low (<0.05%) level during the ®rst days of culture, with DISCUSSION maximally 4% positive cells by day 7 and a sharp decline thereafter In this investigation, we have attempted to further characterize the (Czarnetzki et al, 1983). Proliferation of the rare CD34+ cells mast cell precursor among PBMC, showing that these cells lack c- among cells seeded into culture can thus hardly account for the 48 WELKER ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Table IV. Intracellular histamine levels and percentage of histamine release from PBMC before and after 2 wk of culture with LCSa

Histamine Day 0 Day 14

Intracellular contents (ng per 105 cells)b 1.0 6 0.2 3.0 6 0.9 % of IgE-independent release 0.0 6 0.0 19.6 6 7.5

aFor details, see Materials and Methods (means 6 SD of N=5) increased histamine levels after day 4 of culture and of tryptase and other mast cell markers in upto 40% of cells already after 1 wk of culture, as shown in Table III or reported before by our group (Czarnetzki et al, 1983, 1984). The lack of suf®cient cells expressing the hematopoietic stem cell marker CD34+ to serve as precursors for mast cells in peripheral blood might be explained by a downregulation of this molecule on mast cell precursors, as ®ts with the fact that it is no longer expressed on cultured and mature mast cells (Table I; Escribano et al, 1998). Low numbers of CD34+ cells among PBMC, however, were reported by Agis et al (1993) and, after isolation, served as precursors for mast cell colonies. Unfortunately, these cells were not further characterized with regard to other characteristics of interest, such as c-Kit or FceRI expression. Figure 3. Immunoreactivity (APAAP) of PBMC to antibodies Interestingly, the majority of cells at day 14 of culture in LCS still against FceRIa (29C6),tryptase (AA1),and c-Kit (YB5.B8) after express a number of molecules that are also present on monocytes/ 14 d of culture with LCS in the presence of an irrelevant control macrophages (Table I). This holds particularly for CD11b, CD14, antibody (hatched bars) or the anti-SCF receptor antibody YB5.B8 (speckled bars) (means 6 SD,N = 3). *Signi®cant differences (p<0.01) CD33, CD64, CD68, and HLA-DR. CD33 has so far been between cells cultured with the anti-SCF receptor antibody, compared detected on all tissue mast cells (Escribano et al, 1998; Guo et al, with the control antibody. 1992; Saito et al, 1995; Valent et al, 1989; Ghannadan et al, 1998), in agreement with its de novo expression on the cultured cells (Table I). CD11b has furthermore been described before on HMC-1 cells lineages among the PBMC might account for the high percentage (Weber et al, 1996), and expression of MHC-II has been reported of non-mast cells at week 2 of culture and would contribute to the by a number of authors for HMC-1 and other human mast cells non-mast cell marker expression among the cultured cells. (reviewed in Grabbe et al, 1997). Because SCF is the most potent human mast cell growth factor On double staining, the majority of cultured FceRIa+ cells still (reviewed in Grabbe et al, 1994a), a key to the understanding of express several monocyte/macrophage markers, albeit to a variable mast cell development from PBMC is most likely the elucidation of extent (Table II). The high percentages of CD14+ cells are mechanisms regulating c-Kit expression on mast cell precursors. dif®cult to reconcile with ®ndings by Agis et al (1993), who were Although this receptor is restricted to only a few cell types like mast unable to grow mast cell colonies from the CD14+ human cells and melanocytes in the skin, immature and mature cells do not peripheral blood monocyte subpopulation. They were, however, invariably express it, as shown with melanocyte cell lines and with able to induce 5% pure mast cell colonies and 2.9% mixed mast the mast cell line HMC-1 where in each case, 10%±20% of cells are cell-containing colonies after 20 d of culture from sorted CD14± c-Kit negative (Natali et al, 1992; Hamann et al, 1994), as are monocytes in the presence of SCF. The overall yield of mast cells avidin-positive mast cells in normal human skin (Hermes et al, obtained by this groupwith the colony assay was thus very low, 2000). From data published so far, IL-4, TNFa, and vitamin D3 compared with our methodology where 30%±60% of cultured cells downregulate c-Kit (Sillaber et al, 1991; Jacobson et al, 1995; exhibit mast cell characteristics already by week 2 of culture (Fig 1; Toyota et al, 1996), as does SCF through an internalization of its Czarnetzki et al, 1983, 1984), even if one takes into account that receptor-ligand complex (Shimizu et al, 1996), a ®nding that may total cell numbers decrease markedly during culture (Table III). explain the relatively low c-Kit expression in cultured cells, The discrepancy in ®ndings between Agis et al (1993) and our compared with other mast cell markers (Fig 1). Receptor groupis most likely due to methodologic differences regarding internalization might also account for our inability to detect both cell culture and cell analysis. In mature tissue mast cells, CD14 membrane c-Kit in cells cultured with SCF-containing media on is nevertheless no longer expressed (Escribano et al, 1998). A ¯ow cytometry; however, on immunocytochemistry where decrease of macrophage marker expression on more mature intracellular c-Kit is detected as well, the receptor is detectable, cultures, however, seems to be not unusual and has, for example, even during coculture with YB5.B8 (Fig 3). A functional also been described by us in the past for the macrophage marker Ki- interference with the initiation of c-Kit-dependent signalling by M1P (Hamann et al, 1995). the YB5.B8 antibody, however, must still have been operative to Taken together, the expression of monocyte/macrophage explain the low numbers of FceRIa- and tryptase-positive cells markers on cultured cells should be viewed with caution, as it during culture with this antibody (Fig 3). may be subject to dynamic changes in dependence of culture In contrast to the data on c-Kit downregulation, no data are conditions or stages of cellular maturation. The growth of other available in the literature on the mechanisms involved in de novo clones or the presence of other types of committed precursors in expression or upregulation of c-Kit. In these experiments, the cultures might also contribute to the persistence or de novo induction of c-Kit expression may have occurred via SCF itself, expression of some of these markers. As shown by us before for possibly even via other receptors on the c-Kit-negative precursors, thioglycollate-induced rat peritoneal macrophages that had lost the as recently described for the erythropoietin receptor (Wu et al, ability to assume mast cell features (Czarnetzki et al, 1982), 1995) or via only a few receptors on the precursors that would be terminally committed or differentiated cells from non-mast cell undetectable by immunocytochemistry. Because SCF is as active as VOL. 114, NO. 1 JANUARY 2000 MAST CELL PRECURSOR 49 the SCF-containing cell supernatants with regard to c-Kit Grabbe J, Welker P, MoÈller A, Dippel E, Ashman LF, Czarnetzki BM: Comparative induction (Fig 1), other human mast cell growth factors in these cytokine release from human monocytes, monocyte-derived immature mast cells and a human mast cell line (HMC1) J Invest Dermatol 103:504±508, 1994b media, like NGF and TGFb (see Materials and Methods) are unlikely Grabbe J, Welker P, Rosenbach T, et al: Release of stem cell factor from a human to account for this effect. The cell supernatants are also not more keratinocyte line, HaCaT, is increased in differentiating versus proliferating active than SCF alone regarding FceRIa expression in this model, cells. J Invest Dermatol 107:1±6, 1996 possibly due to mast cell growth inhibitory factors also present in Grabbe J, Karau L, Welker P, Ziegler A, Henz BM: Induction of MHC class II antigen expression on human mast cells. J Derm Sci 16:67±73, 1997 the mast cell conditioning media like GM-CSF (Welker et al, 1995, Guo CB, Kagey-Sobotka A, Lichtenstein LM, Bochner BS: Immunophenotypic and 1997).1 The quantitative differences in expression of mast cell functional analysis of puri®ed human uterine mast cells. Blood 79:708±712, markers within the same culture suggest that each of them is subject 1992 to different molecular regulatory mechanisms during mast cell Hamann K, Grabbe J, Welker P, Haas N, Algermissen B, Czarnetzki BM: Phenotypic evaluation of cultured human mast and basophilic cells and of differentiation. normal human skin mast cells. Arch Dermatol Res 286:380±385, 1994 Taken together, these observations further characterize the Hamann K, Haas N, Grabbe J, Welker P, Czarnetzkl BM: Two novel mast cell peripheral blood mast cell precursors and processes involved in their phenotypic markers, monoclonal antibodies KiMC1 and Ki-M1P, identify differentiation, con®rming that the precursors circulate in an distinct mast cell subtypes. Br J Dermatol 133:547±552, 1995 Hartmann K, Henz BM, KruÈger-Krasagakes S, et al: C3a and C5a stimulate undetectable form among myelomonocytic cells. From there, they chemotaxis of human mast cells. Blood 89:2863±2870, 1997 are recruited into the tissue, most likely under the in¯uence of mast Harvima IT, Schechter NM, Harvima RJ, Fraeki JE: Human skin tryptase: cell chemotactic factors like SCF, C3a, and C5a, and through puri®cation, partial characterisation and comparison with human lung interaction with matrix molecules (Nilsson et al, 1994; KruÈger- tryptase. Biochem Biophys Acta 957:71±80, 1988 Henz BM: SCF and c-Kit in mastocytosis ± a pandora's box holding more theories Krasagakes et al, 1996; Hartmann et al, 1997). In the tissue, they can than proven facts. J Invest Dermatol 110:186, 1998 further differentiate under the in¯uence of locally secreted growth Hermes B, Feldmann-BoÈddeker I, Welker P, Algermisseu B, Steckelings MU, factors like endothelial cell-, ®broblast-, and keratinocyte-derived Grabbe J, Henz BM: Altered expression of mast cell chymase and tryptase and SCF. During this process, the cells apparently maintain but also lose of c-Kit in human cutaneous scar tissue. J Invest Dermatol 114:51±55, 2000 Hiroshi S, Katsui M, Takahashi G, et al: Development of tryptase-positive KU812 some characteristics of their precursors and of the closely related cells cultured in the presence of Steel factor. Int Arch Allergy Immunol 107:330± monocyte-macrophage lineage. Details regarding the exact nature 332, 1995 of the mast cell precursor during its circulation in the blood and Huang S, Terstappen LW: Lymphoid and myeloid differentiation of single human after its migration into the tissue, its interplay with diverse CD34+, HLA-DR±, CD38± hematopoietic stem cells. Blood 83:1515±1526, 1995 differentiation inducing and modulating factors, and the associated Jacobson FW, Dubois CM, Rusten, LS, Veiby OP, Jacobson SE: Inhibition of stem molecular events will need to be further clari®ed. cell factor induced proliferation of primitive murine hematopoietic progenitor cells signaled through the 74-kilodalton tumor necrosis factor receptor. J Immunol 151:3932±3941, 1995 Kawabori S, Kanai N, Tosho T, Adachi T: Existence of c-kit receptor-positive, This work was supported by a grant from the German Research Foundation, DFG. tryptase-negative, IgE-negative cells in human allergic mucosa: A candidate for The authors wish to thank Mrs Regina Nordheim and Mrs Elke Bauer for excellent mast cell progenitor. Int Arch Allergy Immunol 112:36±43, 1997 technical assistance. Kirshenbaum AS, Goff JP, Kessler SW, Mican JM, Zsebo KM, Metcalfe DD: Effect of IL3 and stem cell factor on the appearance of human basophils and mast cells from CD34+ and pluripotent progenitor cells. J Immunol 148:772±777, 1992 Kishi K: A new leukemic cell line with characterized as REFERENCES basophil precursor. Res 9:381±390, 1985 Agis H, Willheim M, Sperr WR, et al: Monocytes do not make mast cells when KruÈger-Krasagakes S, GruÈtzkau A, Baghramian R, Czarnetzki BM: Interaction of cultured in the presence of SCF. Characterization of the circulating mast cell immature human mast cells with extracellular matrix: Expression of speci®c progenitors as a c-kit+, LY±, CD14±, CD17±, colony-forming cell. J Immunol integrin receptors and their role in cell adhesion to extracellular proteins. J 151:4221±4227, 1993 Invest Dermatol 106:538±543, 1996 Ashman LK, Buhring RJ, Aylett GW, Broudy VC, Muller C: Epitope mapping and Lerner NB, Nocka KH, Cole SR, Qui F, Strife A, Ashman LK, Besmer P: functional studies with three monoclonal antibodies to the c-kit receptor Monoclonal antibody YB5.B8 identi®es the human c-kit protein product. tyrosine kinase. J Cell Physiol 158:545±554, 1994 Blood 77:1876±1883, 1991 Butter®eld JH, Weiler D, Dewald G, Gleich GJ: Establishment of an immature mast Longley BJ, Morganroth GS, Tyrrell L, Ding TG, Anderson DM, Williams DE, cell line from a patient with mast cell leukemia. Leukemia Res 12:345±355, 1988 Halaban R: Altered metabolism of mast-cell growth factor (c-kit ligand) in Castells MC, Friend DS, Bunnell CA, Hu X, Fraus M, Osteen RT, Austen KF: The cutaneous mastocytosis. N Engl J Med 328:1302±1307, 1993 presence of membrane-bound stem cell factor on highly immature Maurer D, Fiebiger E, Reininger B, et al: Expression of functional high af®nity nonmetachromatic mast cells in the peripheral blood of a patient with immunoglobulin E receptors (FceRI) on monocytes of atopic individuals. J aggressive systemic mastocytosis. J Allergy Clin Immunol 98:831±840, 1996 Exp Med 179:745±750, 1994 Czarnetzki BM, Sterry W, Bazin H, Kalveram KJ: Evidence that tissue mast cells Metcalfe DD, Baram D, Mekori YA: Mast cells. Physiol Rev 77:1033±1079, 1997 derive from mononuclear phagocytes. Int Arch Allergy Appl Immunol 67:44±48, Miller JS, Westin EH, Schwartz LB: Cloning and characterisation of complementary 1982 DNA for human tryptase. J Clin Invest 84:1188±1195, 1989 Czarnetzki BM, KruÈger G, Sterry W: In vitro generation of mast cell-like cells from Mirrossay L, Chastre E, Callebert J, et al: Histamine H2 receptors and histidine human peripheral mononuclear phagocytes. Int Arch Allergy Appl Immunol decarboxylase in normal and leucemic human monocytes and macrophages. 71:161±167, 1983 Am J Physiol 267:602±611, 1994 Czarnetzki BM, Figdor CG, Kolde G, Vroom T, Aalberse R, de Vries J: Natali PG, Nicotra MR, Sures I, Santoro E, Bigotti A, Ullrich A: Expression of c-kit Development of human connective tissue mast cells from puri®ed blood receptor in normal and transformed human nonlymphoid tissue. Cancer Res monocytes. Clin Exp Immunol 51:549±554, 1984 52:6139±6144, 1992 Czarnetzki BM, Grabbe J, Kolde G, KruÈger-Krasagakes S, Zuberbier T: Mast cells in Nilsson G, Butter®eld JH, Nilsson K, Siegbahn A: Stem cell factor is a chemotactic the cytokine network ± the what, where from and what for. Exp Dermatol factor for human mast cells. J Immunol 153:3717±3725, 1994 4:221±226, 1996 Ratajczak MZ, Luger SM, De Riel K, Abraham J, Calabretta B, Gewirtz AM: Role Denburg JA: Differentiation of human basophils and mast cells. In: Marone G (ed.). of the KIT protooncogene in normal and malignant human hematopoiesis. Proc Human Basophils and Mast Cells: Biological Aspects, Vol. 61: Chem Immunol. Natl Acad Sci USA 89:1710±1718, 1992 Basel: Karger, 1995, pp. 49±71 Reischl IG, Corvaia N, Effenberger F, Wolff-Winiski B, KroÈmer E, Mudde GC: Escribano L, Ocqueteau M, Almeida J, Orfao A, San Miguel JF: Expression of the Function and regulation of FceRI expression on monocytes from non-atopic c-kit (CD117) molecule in normal and malignant hematopoiesis. Leukemia donors. Clin Exp Allergy 26:630±641, 1996 Lymphoma 30:459±466, 1998 Riske F, Hakimi J, Mallamaci M: High af®nity human IgE receptor (FceRI). Analysis Ghannadan M, Baghestamian M, Wimazal F, et al: Phenotypic characterization of of functional domains of the a-subunit with monoclonal antibodies. J Biol human skin mast cells by combined staining with toluidine blue and CD Chem 266:11245±11251, 1991 antibodies. J Invest Dermatol 111:689±695, 1998 Robertson MW, Mehl VS, Richards ML, Fu-Tong L: m-RNA variants encoding Grabbe J, Welker P, Dippel E, Czarnetzki BM: Stem cell factor, a novel cutaneous multiple forms of the high-af®nity IgE receptor alpha subunit in transformed growth factor for mast cells and melanocytes. Arch Dermatol Res 287:78±84, and nontransformed mast cells. Int Arch Allergy Appl Immunol 96:289±295, 1991 1994a Rottem M, Okada T, Goff JP, Metcalfe DD: Mast cells cultured from peripheral blood of normal donors and patients with mastocytosis originate from a CD34+/FceRI± cell population. Blood 84:2489±2496, 1994 1Welker P, Grabbe J, GruÈtzkau A, Henz BM: In¯uence of the growth Saito H, Miura K, Takahashi G, et al: Development of tryptase-positive KU812 cells factors NGF and GM-CSF on the expression of mast cell characteristics. J cultured in the presence of Steel factor. Int Arch Allergy Immunol 107:330±332, Allergy Clin Immunol 97:266, 1996 (abstr.) 1995 50 WELKER ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Sambrock J, Fritsch EF, Maniatis T: Analysis of genomic DNA. In: Ford N, Nolan C, Valent P, SpanbloÈchl E, Sparr WR, et al: Induction of differentiation of human mast Ferguson M (eds). Molecular Cloning. New York: Cold Spring Harbor cells from bone marrow and peripheral blood mononuclear cells by rhSCF/c- Laboratory Press, 1989a, pp. 7.37±7.79 kit-ligand (KL) in long time culture. Blood 80:2237±2245, 1992 Sambrook J, Fritsch EF, Maniatis T: Analysis of RNA. In: Ford N, Nolan C, Walls AF, Jones DB, Williams JH, Church MK, Holgate ST: Immunohistochemical Ferguson M (eds). Molecular Cloning. New York: Cold Spring Harbor identi®cation of mast cells in formaldehyde-®xed tissue using monoclonal Laboratory Press, 1989b, pp. 9.31±9.57 antibodies speci®c for tryptase. J Pathol 162:119±126, 1990 Schadendorf D, Tiedemann K-H, Haas N, Czarnetzki BM: Detection of human Weber S, KruÈger-Krasagakes S, Grabbe J, Zuberbier T, Czarnetzki BM: Mast cells. papilloma viruses in paraf®n-embedded condylomata accuminata ± comparison Int J Dermatol 34:1±10, 1995 of immunohistochemistry, in situ hybridization, and polymerase chain reaction. Weber S, Babina M, KruÈger-Krasagakes S, GruÈtzkau A, Henz BM: A subclone (5C6) J Invest Dermatol 97:549±554, 1991 of the human mast cell line HMC-1 represents a more mature phenotype than SchoÈneich JT, Victoria VL, Kado-Fong H, Presky DH, Kochan JP: Association of the original cell line. Arch Dermatol Res 288:778±782, 1996 the human FceRI subunit with novel cell surface polypeptides. J Immunol Welker P, Grabbe J, Czarnetzki BM: Human keratinocytes release mast cell 148:2181±2185, 1992 differentiation factors other than stem cell factor. Int Arch Allergy Immunol Shimizu Y, Ashman LK, Du Z, Schwartz LB: Internalization of Kit together with 107:139±141, 1995 stem cell factor on human fetal liver-derived mast cells. J Immunol 156:3443± Welker P, Grabbe J, Zuberbier T, Henz BM: GM-CSF downregulates mast cell 3449, 1996 speci®c tryptase, FceRI and histamine. Int Arch Allergy Appl Immunol 113:284± Sillaber CH, Bevec D, Ashman LK, et al: Il-4 regulates c-kit gene product expression 286, 1997 in human mast and myeloid cell progenitors. J Immunol 147:4224±4228, 1991 Williams DE, Eisenmann J, Baird A, et al: Identi®cation of a ligand for the c-kit Takagi M, Nokahata T, Kubo T, et al: Stimulation of mouse connective tissue-type protooncogen. Cell 63:167±174, 1990 mast cells by hematopoietic stem cell factor, a ligand for c-kit receptor. J Wu H, KlingmuÈller U, Besmer P, Lodish HF: Interaction of the erythropoietin and Immunol 148:3446±3453, 1990 stem-cell-factor receptors. Nature 377:242±246, 1995 Togumaga K, Nakamura Y, Sakata K, Fujimori F, Ohkubo M, Sawada K, Sakiyama Zsebo KM, Williams DA, Geissler EN: Stem cell factor is encoded at the S1 locus of S: Enhanced expression of glyceraldehyde-3-phosphate. Cancer Res 47:5616± the mouse and is the ligand for the c-kit tyrosine kinase receptor. Cell 63:213± 5619, 1987 224, 1990 Toyota N, Sakai H, Takahashi H, Hashimoto Y, Iizuka H: Inhibitor effect of 1a,25- Ziegler A, Heinig J, MuÈller C, GoÈtz H, Thinnes FP, Uchanska-Ziegler B, Wernet P: dihydroxyvitamin D3 on mast cell proliferation and A23187-induced histamine Analysis by sequential immunoprecipitations of the speci®cities of monoclonal release, also accompanied by a decreased c-kit receptor. Arch Dermatol Res antibodies TUÈ 22, 34, 35, 36, 37, 39, 43, 58 and YD1/63.HLK directed against 288:709±715, 1996 human HLA Class II antigens. Immunobiol 171:77±90, 1986 Valent P: The riddle of the mast cell: kit (CD117)-ligand as the missing link? Im Zuberbier T, Pfrommer C, BeinhoÈlzl J, Hartmann K, Czarnetzki BM: Gangliosides Today 15:111±114, 1994 enhance IgE-receptor dependent histamine and LTC4 release from human mast Valent P, Ashman LK, Hinterberger W, Eckersberger F, Majdic O, Lechner K, cells. Biochem Biophys Acta 1269:79±84, 1995 Bettelheim P: Mast cell typing: demonstration of a distinct hematopoietic cell Zuberbier T, Guhl S, Handtke T, Handtke C, Welker P, Grabbe J, Henz BM: Role type and evidence for immunophenotypic relationship to mononuclear of gangliosides during the in vitro human mast cell differentiation. Exp Dermatol phagocytes. Blood 73:1178±1784, 1989 8:380±385, 1999