Elevated Expression of the Proto-Oncogene C-Kit in Patients with Mastocytosis H Nagata, AS Worobec, T Semere and DD Metcalfe

Home , Oncogene

Elevated expression of the proto-oncogene c-kit in patients with mastocytosis H Nagata, AS Worobec, T Semere and DD Metcalfe

Laboratory of Allergic Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD, USA

The stem cell factor (SCF)c-kit receptor interaction plays a criti- and germ cell lineages.12,14 The c-kit receptor is known to cal role in the development and survival of mast cells. Several mediate critical signals for the proliferation and maturation studies have also associated c-kit receptor mutations with the 15–17 human diseases, mastocytosis and piebaldism. Over- of these cells, and recent studies have demonstrated that expression of c-kit has been reported to be associated with abnormalities of c-kit are associated with human diseases myeloproliferative disorders and myelodysplastic syndromes. which are characterized by impaired or enhanced growth of Using peripheral blood mononuclear cells (PBMCs) from 11 cells expressing this receptor. Mutations of c-kit are involved patients with indolent mastocytosis (category I), mastocytosis in the hereditary disease piebaldism, which is characterized with an associated hematologic disorder (category II), or by loss of pigmentation,18–21 while overexpression of c-kit aggressive mastocytosis (category III); a patient with CMML mRNA is reported to be associated with myeloproliferative unassociated with mastocytosis, and PBMCs from 13 normal 22 subjects, we examined the level of expression of c-kit mRNA disorders and myelodysplastic syndromes (MDS). along with other c-kit isoforms to determine if overexpression Mastocytosis is a disease of unknown pathogenesis charac- of the c-kit receptor was associated with mastocytosis. Using terized by mast cell hyperplasia in the bone marrow, liver, quantitative competitive PCR, c-kit mRNA levels on average spleen, lymph nodes, gastrointestinal tract and skin.23 It is were found to be statistically elevated in the five patients with classified into four categories: (I) indolent mastocytosis, (II) mastocytosis with an associated hematologic disorder and in the patient with aggressive mastocytosis as compared with mastocytosis with an associated hematologic disorder, (III) controls, but not elevated in patients with indolent masto- lymphadenopathic mastocytosis with eosinophilia (also cytosis. The relative mRNA expression for the two c-kit iso- known as aggressive mastocytosis), and (IV) mast cell leuke- forms was not significantly different in the mastocytosis mia. Based on the information that the SCF/c-kit receptor sys- patients compared with controls. This demonstration of the tem plays a role in the development of mast cells, several stud- overexpression of c-kit mRNA in mastocytosis, and particularly ies investigating the role of SCF or the c-kit receptor in the those patients with clinical evidence of myelodysplastic syn- development of mastocytosis have been performed in patients drome, adds evidence to support the conclusion that masto- 24 cytosis, at least in some patients, is a feature of myelodyspla- with mastocytosis. Longley et al reported that there is sia; and suggests that determination of c-kit mRNA expression increased soluble SCF in the skin of indolent mastocytosis in PBMCs may provide an additional approach to assessing patients. We recently identified an activating point mutation prognosis. of c-kit in the peripheral blood mononuclear cells (PBMCs) Keywords: c-kit; mastocytosis; stem cell factor; mast cells of patients with mastocytosis with an associated hematologic disorder which was a substitution of Val for Asp-816, the same mutation as one of the two mutations found in the human Introduction mast cell leukemia cell line HMC-1.25–27 This mutation has now also been reported in the skin of a patient with indolent The proto-oncogene c-kit encodes a transmembrane tyrosine mastocytosis.28 Nevertheless, it still remains to be revealed kinase receptor for the growth factor, SCF, and belongs to the whether there are any abnormalities of c-kit receptor same family of receptors as those which bind platelet-derived expression in mastocytosis patients. Thus, in the present study, growth factor and colony-stimulating factor-1.1,2 The common we decided to evaluate the level of expression of c-kit mRNA structural feature of this family of receptors is a five immuno- in patients with mastocytosis to determine whether or not globulin-like loop structure in the extracellular domain and a there is overexpression of c-kit. Furthermore, we examined the kinase insert which splits the catalytic domain between the relative expression of mRNA for Kit and Kit A as well as mRNA ATP binding site and the phosphotransferase active site. In for the other isoforms. human c-kit, there are two kinds of isoforms which are generated by alternative splicing.3 The alternative splice site for the first is located at the 5′ splice junction involving exon 9, thus Materials and methods generating two isoforms of mRNA differing in size by 12 bases; the alternative splice site for the second is located at Subjects the 3′ acceptor site involving exon 15 in the kinase insert domain and generates two mRNA isoforms differing in size by All procedures used to obtain human specimens were three bases. The protein products of the former isoforms differ approved by the National Institute of Allergy and Infectious by four amino acids, Gly-Asn-Asn-Lys (+)or(−), and they are Diseases (NIAID) Institutional Review Board at the National referred to as Kit A and Kit, respectively. Previous murine stud- Institutes of Health and followed informed consent. Blood was ies of homologous isoforms have demonstrated that Kit but obtained from 11 adult patients with mastocytosis, and one not Kit A has constitutive weak autophosphorylation activity.4 adult patient with chronic myelomonocytoic leukemia Transcripts and protein products of c-kit are expressed on (CMML) unassociated with mastocytosis. The diagnosis for mast cells,5,6 hematopoietic stem cells,6–10 melanocytes11–13 each patient was based on clinical symptomatology, physical findings, bone marrow biopsy and aspirate, and other appro- priate biopsy specimens (Table 1). Subjects with mastocytosis Correspondence: DD Metcalfe, NIH/NIAID/LAD, Building 10, Room 11C205, 10 Center Drive, MSC 1881, Bethesda, MD 20892-1881, were divided into those with indolent mastocytosis (five USA; Fax: 301 480-8384 patients total, patients I–1, 2, 3, 4, 5), mastocytosis with an Received 7 January 1997; accepted 16 October 1997 associated hematologic disorder (five patients total, patients Elevated expression of c-kit H Nagata et al 176 Table 1 Patient data PCR amplification

Patient No. Age/Sex Diagnosis PCR of reverse transcribed cDNA (RT-PCR) was performed with a reagent mixture composed of 200 nmol/l primers, I-1 43/F Indolent mastocytosis 150 ␮mol/l dNTPs, 20 mmol/l Tris HCl, pH 8.4, 50 mmol/l I-2 59/M Indolent mastocytosis KCl, 1.5 mmol/l MgCl and 1.5 units Taq DNA polymerase in I-3 16/F Indolent mastocytosis 2 a50␮l reaction solution. The PCR program was as follows: I-4 44/F Indolent mastocytosis ° ° ° I-5 32/F Indolent mastocytosis 94 C for 1 min, 60 C for 1 min and 72 C for 1 min. II-1 64/F Mastocytosis with an AHD (myelofibrosis) II-2 39/M Mastocytosis with an AHD Quantitative competitive PCR for c-kit (MDS: RA) II-3 46/M Mastocytosis with an AHD To determine the level of c-kit mRNA expressed in PBMCs, we (MDS: CMML) performed quantitative competitive PCR (qc-PCR) with reverse II-4 64/M Mastocytosis with an AHD transcribed cDNA. The sequence of primers were: primer 1 (MDS: RA) ′ ′ II-5 56/F Mastocytosis with an AHD (sense), 5 GTAAGGCTTACAACGATGTGGGCA 3 (nucleotide 1 (MDS: unclassified) 1493–1516 of the published c-kit cDNA sequence ) and primer 2 (anti-sense), 5′ TTGAGCATCTTTACAGCGA- III-1 20/M Aggressive mastocytosis CAGTCA 3′ (1898–1874). To make an internal standard for IV-1 69/F MDS (CMML), no the qcPCR, PCR products from cDNA were cleaved with the mastocytosis restriction endonuclease Mnl1. Because of the presence of two isoforms, predicted sizes of the PCR products from the AHD, associated hematologic disorder; MDS, myelodysplastic syndrome; CMML, chronic myelomonocytic leukemia; RA, refractory two isoforms were 406 and 394 bp, respectively, and the pre- anemia. dicted sizes of DNA fragments after Mnl1 cleavage were 202, 132 and 72 bp for Kit A cDNA, and 190, 132 and 72 bp for Kit cDNA (Figure 1). After separation on a 20% polyacrylamide gel, the 190 and 72-bp fragments were excised, purified by the crush and soak method,30 and ligated to make a 262- II-1, 2, 3, 4, 5), and aggressive mastocytosis (one patient, bp internal standard. This internal standard, thus, had the patient III–1). Eight of these respective mastocytosis patients, same primer sequences as the c-kit cDNA, and competed with (I-1, 2, 3, II–1, 2, 3, 4, III-1), and the patient with CMML (IV- the cDNA for primers and other reagents in one tube. The 1) are the patients whose PBMCs were screened for c-kit ligated 262-bp DNA fragment was then amplified by PCR with mutations in our previous study, with their clinical features the same primers, separated on a 20% polyacrylamide gel, and hematological findings described in that report.25 Four of these mastocytosis patients (II-1, 2, 3, 4) had the point mutation, Asp816Val, in the catalytic domain of c-kit.An additional fifth patient with mastocytosis with an associated hematologic disorder (II-5) whose PBMCs we recently analyzed for this c-kit mutation was also found to have the same mutation, and her associated hematologic disorder was unclassifiable myelodysplastic syndrome (MDS). Normal controls consisted of blood obtained from 13 heal- thy male and female adult volunteers whose ages ranged between 25 and 40 years of age.

RNA extraction and cDNA synthesis

PBMCs were isolated by Ficoll–Hypaque density gradient cen- trifugation (density 1.077, Sigma Chemical, St Louis, MO, USA). Total RNA was extracted from PBMCs by the acid guanidium phenol-chloroform method,29 precipitated and kept at −20°C until use. After measurement of the amount of RNA by spectrophotometry, 5 ␮g of each sample was reverse transcribed to cDNA with Moloney murine leukemia virus reverse transcriptase without RNase H activity and oligo dT primer (Life Technologies, Gaithersburg, MD, USA) in 20 ␮l of reac- Figure 1 Creation of a 262-bp internal competitive standard for tion solution. After reverse transcription, the reaction solution qcPCR for c-kit. After PCR of cDNA with primers 1 and 2, PCR pro- was diluted with deionized water to 100 ␮l, then aliquoted to ducts were cleaved with Mnl1 endonuclease. The 190-bp fragments 10 ␮l, and kept frozen at −80°C until use. To minimize varia- cleaved from isoform 1 and 72-bp fragments from both isoforms were then separated on a polyacrylamide gel, purified, and ligated to make bility of the efficiency of the reverse transcription reaction, a 262-bp internal standard. Isoform 1 and isoform 2 are the PCR pro- all 25 RNA samples from patients and controls were reverse ducts amplified from mRNA for Kit A and Kit, respectively. EC, extra- transcribed at the same time. cellular domain; TM, transmembrane; CP, cytoplasmic domain. Elevated expression of c-kit H Nagata et al 177 and purified. The concentration of the purified PCR product was determined by comparison of the intensity of electrophor- etic bands on a 1.5% agarose gel against those generated by known quantities of ␾X174/HaeIII digest. The concentration of the PCR product was adjusted to 100 attomole/␮l(1 attomole = 10−18 mole) in 50 ␮g/ml glycogen (Boehringer Mannheim Biochemica, Indianapolis, IN, USA) and stored as a stock solution of the internal standard. We performed two sets of qcPCR to quantitate c-kit in each sample as follows: 2 ␮l of cDNA sample, ie cDNA made from 100 ng of total RNA, was amplified for 40 cycles in the presence of 1 ␮lof a 10-fold dilution series of internal standards whose concen- trations were 10−1 to 100−5 attomole/␮l (Figure 2a). After approximating the concentration of standard which would give a 1:1 ratio of the intensity of bands of PCR products amplified from cDNA and standard, cDNA was further analyzed with a two-fold dilution series of standards. After the second PCR, the intensity of bands was assessed with a densitometer (Molecular Dynamics, Sunnyvale, CA, USA), and the concentration of the standard which would give a 1:1 ratio of band intensity was obtained by a statistical analysis with a computer (Figure 2b). The result was then converted to the amount of the internal standard which would equilibrate with cDNA reverse transcribed from 1 ␮g of total RNA. This value was used to express the amount of c-kit mRNA in 1 ␮gof total RNA.

Figure 2 An example of qcPCR for c-kit. (a) First, cDNA reverse transcribed from 100 ng of total RNA from the PBMCs was amplified with 1 ␮l of a 10-fold dilution series (10−1 to 10−5 attomole/␮l) of Studies on isoforms competitive standard. Because of the presence of isoforms of c-kit mRNA, PCR products from cDNA containing 406 and 394-bp fragments, which ran in one band, were separated from the 262-bp PCR To visualize the isoforms of c-kit mRNA, radiolabeled PCR products derived from competitive standard. Knowing the approxi- products containing the 12 base and three base differing iso- mate concentration of standard which would give a 1:1 ratio of the intensity of bands of PCR products from cDNA and standard, cDNA forms were separated on a sequencing gel. The primers used was further analyzed with a two-fold dilution series. (b) The intensity for amplification of 12-base differing isoforms were primer 1 of each band of the second PCR was measured by densitometry. A and primer 3 (anti-sense), 5′ CCACTGTACTTCATACAT- straight line which represents a relationship of the ratio of the intensity GGGTTTC 3′ (1692–1668). The predicted sizes of the PCR of bands vs the concentration of internal standard was statistically products were 200 and 188 bp, respectively. The primer drawn by a computer, and the concentration of the internal standard which would give a 1:1 ratio of bands was obtained. The amount of sequences for the three base differing isoforms were: primer 4 ′ ′ mRNA in the PBMCs was expressed by the amount of internal stan- (sense), 5 GAATTTTTTGAGAAGAAAACGTGATTC 3 (2058– dard which would equilibrate with cDNA reverse transcribed from 2084) and primer 5 (anti-sense), 5′ TTCATGTCCATG- 1 ␮g of total RNA. In this case, the amount of mRNA in 1 ␮g of total TACTCCATTAGTAC 3′ (2195–2171); hence, the predicted RNA was expressed as 1.31 × 10−3 times 10, ie 1.31 × 10−2 attomole. sizes of PCR products were 138 and 135 bp, respectively. One microliter of cDNA was amplified in the presence of 30 nmol/l Results of ␣-32P dCTP. The radiolabeled PCR products were then elec- trophoresed on a 1.5% agarose gel, and a band including both Expression of c-kit mRNA in the PBMCs of patients isoforms was cut out and purified through a Suprec-01 column and controls (Takara, Kyoto, Japan). After radioactivity was measured with a liquid scintillation counter, 1000 c.p.m. of each of the pur- We performed qcPCR to determine the level of c-kit mRNA ified products were loaded on to a 6% sequencing gel after expressed in the PBMCs of patients with mastocytosis and ° heat denaturing at 90 C for 3 min. As a size marker for 12 CMML, along with control subjects, and demonstrated that base differing isoforms, a sequencing product was run in par- this technique was sensitive enough to be carried out with allel. For this purpose, PCR products amplified with primer 6, cDNA synthesized from 100 ng of total RNA. The intensity of 5′ GATCCCATCGCAGCTACCGCG 3′ (1–21) and primer 2 bands amplified from c-kit cDNA which contained 406 and were directly sequenced with radiolabeled primer 3 as 394-bp fragments was stronger when amplified in the pres- described.25 For the three base differing isoforms, PCR pro- ence of a smaller amount of internal standard and weaker ducts amplified with primer 7, 5′ GTGCTGGAGCTT- when amplified with a larger amount of standard, indicating TCGGGAAGG 3′ (1808–1828) and primer 5 were directly that the cDNA and internal standard were competitively sequenced with primer 5, and run in parallel with samples. amplified in the same tube (Figure 2). Two sets of qcPCR were The gel was dried on Whatman 3MM paper, autoradio- performed for each sample with a 10-fold and a two-fold graphed at −70°C, and the density of bands was measured dilution series of standard, and the results of these two sets of by densitometry. qcPCR did not show any obvious discrepancy (Figure 2a). As Elevated expression of c-kit H Nagata et al 178 may be seen in Figure 3, the level of c-kit mRNA in the PBMCs to 100% (94.5 ± 4.9%), and was 97.0 and 94.0%, respect- of 13 controls ranged from 0.21 to 13 × 10−3 ively, in two patients with indolent mastocytosis, 88.9 and (mean ± s.d. = 3.9 ± 4.3 × 10−3) attomole/␮g total RNA, and 91.4% in the two patients with mastocytosis with an associa- that of five patients with indolent mastocytosis ranged from ted hematologic disorder, 90.3% in the patient with aggressive 1.6 to 23 × 10−3 (9.0 ± 8.4 × 10−3) attomole/␮g total RNA. Stu- mastocytosis, and 88.3% in the patient with CMML unassocia- dent’s unpaired t-test showed no statistically significant differ- ted with mastocytosis. ence between the mean for these two groups. However, the level of c-kit transcripts in the PBMCs of five patients with mastocytosis with an associated hematologic disorder ranged Expression of three base differing isoforms in the from 7.7 to 34 × 10−3 (21 ± 9.3 × 10−3) attomole/␮g total RNA, PBMCs of patients and controls and the mean level of c-kit transcripts was statistically significant compared to the mean c-kit transcript level in 13 control After RT-PCR using ␣-32P dCTP with a primer pair flanking subjects (P Ͻ 0.01). The levels of c-kit mRNA in the PBMCs the alternative splice site for three base differing isoforms, the of a patient with aggressive mastocytosis and a patient with radiolabeled PCR products of two isoforms were separated on CMML unassociated with mastocytosis were 21 × 10−3 and a 6% sequencing gel (Figure 5). Two bands were observed in 24 × 10−3 attomole/␮g total RNA, respectively. each isoform, possibly because Taq DNA polymerase randomly added dATP to the 3′ end of PCR products.31 The intensity of two bands in each isoform was measured together by Expression of mRNA for Kit A and Kit in the PBMCs densitometry, and the percentage of three base smaller iso- of patients and controls forms was calculated. There was no obvious difference in the ratio of three base smaller isoforms between patients and con- To visualize the 12 base differing isoforms of c-kit mRNA, RT- trols; the ratio of the density of three base smaller bands in PCR products containing both isoforms (Kit A and Kit) were the five control subjects was 17.1–33% (26.5 ± 6.1%), and separated on a 6% sequencing gel. Run with a sequencing 24.5 and 28.3% in the two indolent mastocytosis patients, reaction product, the PCR products derived from the two iso- 25.4 and 25.7% in the two patients with mastocytosis with an forms were demonstrated to differ in size by 12 bases associated hematologic disorder, 27.9% in the patient with (Figure 4). After the density of autoradiographic bands as mea- aggressive mastocytosis, and 28.0% in the patient with CMML sured by densitometry, the percentage of the 12 base smaller unassociated with mastocytosis. bands was calculated. The ratio of the 12 base smaller bands which was derived from mRNA for Kit was markedly greater than that of the larger ones which were derived from Kit A Discussion mRNA in all subjects studied. Furthermore, there was no significant difference in this ratio in the six mastocytosis patients In the present study, we demonstrate that the mean level of as compared with the five control subjects. The ratio of the c-kit mRNA per 1 ␮g of total RNA extracted from PBMCs was density of smaller bands in the five controls ranged from 89.1 statistically significantly elevated in patients with mastocytosis with an associated hematologic disorder compared with controls, but not elevated in patients with indolent mastocytosis. Elevated c-kit mRNA levels are not, however, specific for mastocytosis with an associated hematologic disorder, as the levels of c-kit mRNA in a patient with aggressive mastocytosis and a patient with CMML unassociated with mastocytosis were as high as the mean c-kit mRNA level of the patients with mastocytosis with an associated hematologic disorder (Figure 3). Moreover, there is a report that the levels of c-kit mRNA in the PBMCs of patients with myeloproliferative disorders and MDS were statistically significantly higher than that of normal controls.22 Therefore, it appears that such an increase in c-kit mRNA is correlated with the presence of a hematologic disorder and may reflect an expansion of c-kit- positive cells in the peripheral blood. In our experience, the patient with aggressive mastocytosis was also noted to have unique hematologic abnormalities characterized by a dysplastic megakaryocyte lineage, clumped giant hypogranular plate- lets in the peripheral blood, an increased leukocyte alkaline Figure 3 The level of c-kit mRNA in the PBMCs of patients and phosphatase (LAP) score, and diagnosed to have a reactive controls. The level of c-kit mRNA in the PBMCs of 13 controls ranged 25 × −3 ± = ± × −3 ␮ marrow. from 0.21 to 13 10 (mean s.d. 3.9 4.3 10 ) attomole/ g + total RNA, and that of five patients with indolent mastocytosis ranged In the human hematopoietic system, a population of CD34 from 1.6 to 23 × 10−3 (9.0 ± 8.4 × 10−3) attomole/␮g total RNA, while stem cells expresses c-kit receptors. Mast cells, in contrast to the level of c-kit transcripts in the PBMCs of the five patients with neutrophils and other myeloid cells, express c-kit as mature mastocytosis with an associated hematologic disorder ranged from 7.7 cells.6,8–10 Because there are a small number of CD34+ cells × −3 ± × −3 ␮ to 34 10 (21 9.3 10 ) attomole/ g total RNA. Student’s within PBMCs, and there are no circulating mast cells in nor- unpaired t-test showed no statistically significant difference between mal circumstances, CD34+ cells are considered to represent controls and the indolent mastocytosis patients. Patients with mastocytosis with an associated hematologic disorder were found to have the majority of c-kit expressing cells in the PBMCs in normal 32–36 a statistically significantly greater amount of c-kit mRNA in the PBMCs subjects. As mastocytosis progresses, however, this situ- than controls (P Ͻ 0.01). ation changes as immature mast cells are identified in periph- Elevated expression of c-kit H Nagata et al 179

Figure 4 Expression of the 12 base differing isoforms of c-kit in patients and controls. PCR products ampliﬁed with primers 1 and 3, which are ﬂanking the alternative splice site for the 12 base differing isoforms, were labeled with ␣-32P dCTP by direct incorporation. PCR products derived from the two isoforms were then separated on a 6% sequencing gel. A sequence on the right side is of a PCR product sequenced with primer 3. Because smaller isoforms comprise the consensus sequence, the sequence of primer 1 appears at the level of 12 base smaller isoforms. The intensity of bands was measured by densitometry, and the percentage of smaller isoforms are shown below. Samples are: lane 1, patient I-1; lane 2, patient I-2; lane 3, patient III-1; lane 4, patient IV-1; lane 5, patient II-3; and lane 6, patient II-1.

Figure 5 Expression of three base differing isoforms of c-kit in patients and controls. PCR products amplified with primers 4 and 5, which are flanking the alternative splice site for the three base differing isoform, were labeled with ␣-32P dCTP and separated on a 6% sequencing gel. A sequence on the right side is of a PCR product sequenced with primer 5. Because both isoforms comprise templates for sequencing, the same sequence repeatedly appears three bases apart. Two bands are seen in each isoform, possibly because Taq DNA polymerase randomly added dATP to the 3′ end of PCR strands.28 The intensity of bands was measured with a densitometer, and the percentage of shorter isoforms was calculated. Samples are: lane 1, patient I-1; lane 2, patient I-2; lane 3, patient III-1; lane 4, patient IV-1; lane 5, patient II-3; and lane 6, patient II-1. eral blood, and because mast cells as they differentiate rapidly transduction pathway of the mast cell. In this study, the rela- become CD34−.22,35 Hence, overexpression of c-kit in masto- tive expression level of mRNA for these two isoforms was not cytosis may be due to both an increase in the number of obviously different in patients as compared with controls. CD34+ cells and an increase in mast cells and mast cell pre- Likewise, there was no change in the relative level of mRNA cursors as detected on bone marrow aspirates and evaluation for three base differing isoforms among patients and controls. of peripheral blood smears.25,28,32 These data suggest that the relative levels of Kit and Kit A are We asked the additional question of whether or not there unlikely to contribute to an increase in mast cell number in is altered expression of isoforms of c-kit mRNA, as isoforms mastocytosis patients. There are two previous reports which of c-kit receptors are known to have different baseline activi- discuss the relative expression of isoforms of c-kit mRNA. ties. Kit but not Kit A is known to have weak constitutive auto- Crosier et al3 studied the relative levels of both 12 base and phosphorylation activity and is known to associate with PI- three base differing isoforms in normal tissue, transformed cell 3′ kinase and PLC␥14 – important intermediaries in the signal lines and acute myelogenous leukemia (AML) blasts, and Elevated expression of c-kit H Nagata et al 180 reported that 12 base larger transcripts were 10- to 15-fold 7 Ogawa M, Matsuzaki Y, Nishikawa S, Hayashi S, Kunisada T, more abundant than 12 base smaller ones and three base dif- Sudo T, Kina T, Nakauchi H, Nishikawa S. Expression and function fering transcripts were coexpressed with dominance of three of c-kit in hemopoietic progenitor cells. J Exp Med 1991; 174: 63–71. base larger transcripts in all subjects studied; these data are 8 Papayannopoulou T, Brice M, Broudy VC, Zsebo KM. Isolation of consistent with our results. Another study showed that the c-kit receptor expressing cells from bone marrow, peripheral relative levels of 12 base smaller transcripts compared to blood, and fetal liver: functional properties and composite anti- larger isoforms in AML blasts are variable with values ranging genic profile. Blood 1991; 78: 1403–1417. 9 Reisbach G, Bartke I, Kempkes B, Kostka G, Ellwart J, Birner A, from as low as 1.3 to as high as 11.8, while the ratio of these ¨ isoforms in normal bone marrow was tightly clustered Thalmeier K, Mailhammer R, Bornkamm GW, Ullrich A, Dormer 37 P. Characterization of hemopoietic cell populations from human between 4.4 and 5.5. Both studies concluded that the 12 cord blood expressing c-kit. Exp Hematol 1993; 21: 74–79. base and/or three base differing isoforms have no significance 10 Yamaguchi Y, Gunji Y, Nakamura M, Hayakawa K, Maeda M, in AML blasts and transformed cell lines. Thus, the production Osawa H, Nagayoshi K, Kasahara T, Suda T. Expression of c-kit of these isoforms may occur along with physiological pro- mRNA and protein during the differentiation of human hematopo- cesses in bone marrow development. ietic progenitor cells. Exp Hematol 1993; 21: 1233–1238. In summary, patients with mastocytosis showed no clear 11 Orr-Urtreger A, Avivi A, Zimmer Y, Givol D, Yarden Y, Lonai P. abnormalities in the relative expression level of isoforms of c- Developmental expression of c-kit, a proto-oncogene encoded by the W locus. Development 1990; 109: 911–923. kit in the PBMCs. On average, however, patients with masto- 12 Manova K, Bachvarova RF. Expression of c-kit encoded at the W cytosis with an associated hematologic disorder and aggres- locus of mice in developing embryonic germ cells and presump- sive mastocytosis had elevated levels of c-kit mRNA. Further- tive melanoblasts. Dev Biol 1991; 146: 312–324. more, clinical features of indolent mastocytosis and other cat- 13 Funasaka Y, Boulton T, Cobb M, Yarden Y, Fan B, Lyman SD, egories of mastocytosis are known to be distinct, although Williams DE, Anderson DM, Zakut R, Mishima Y, Halaban R. c- patients with indolent disease are sometimes reclassified.23 kit Kinase induces a cascade of protein tyrosine phosphorylation in normal human melanocytes in response to mast cell growth Mastocytosis with an associated hematologic disorder and factor and stimulates mitogen-activated protein kinase but is aggressive mastocytosis have a more rapid course of disease, down-regulated in melanomas. Mol Biol Cell 1992; 3: 197–209. poorer prognosis and are not always associated with skin 14 Manova K, Nocka K, Besmer P, Bachvarova RF. Gonadal involvement. Conversely, most patients with indolent masto- expression of c-kit encoded at the W locus of the mouse. Develop- cytosis have a more optimistic prognosis and show urticaria ment 1990; 110: 1057–1069. pigmentosa, telangiectasia macularis eruptiva perstans (TMEP) 15 Morrison-Graham K, Takahashi Y. Steel factor and c-kit receptor: or diffuse cutaneous mastocytosis at some time in their clinical from mutants to a growth factor system. Bioessays 1993; 15: 77–83. course. These findings suggest that varied pathological pro- 16 Hashimoto K, Tsujimura T, Moriyama Y, Yamatondani A, Kimura cesses may underlie the different categories of mastocytosis, M, Tohya K, Morimoto M, Kitayama H, Kanakura Y, Kitamura Y. but that in some patients, mastocytosis is a concomitant of a Transforming and differentiation-inducing potential of constitut- dysplastic process, one manifestation of which is the over- ively activated c-kit mutant genes in the IC-2 murine interleukin- expression of c-kit. 3-dependent mast cell line. Am J Pathol 1996; 148: 189–200. 17 Tsujimura T, Morimoto M, Hashimoto K, Moriyama Y, Kitayama H, Matsuzawa Y, Kitamura Y, Kanakura Y. Constitutive activation of c-kit in FMA3 murine mastocytoma cells caused by deletion of Acknowledgements seven amino acids at the juxamembrane domain. Blood 1996; 87: 273–283. We thank Janet Ludwig and Cindy Pagonis for their help in 18 Giebel LB, Spritz RA. Mutation of the KIT (mast/stem cell growth the preparation of the manuscript. factor receptor) protooncogene in human piebaldism. Proc Natl Acad Sci USA 1991; 88: 8696–8699. 19 Fleishman RA, Saltman DL, Stastny V, Zneimer S. Deletion of the References c-kit protooncogene in the human developmental defect piebald trait. Proc Natl Acad Sci USA 1991; 88: 10885–10889. 1 Yarden Y, Kuang WJ, Yang-Feng T, Coussens L, Munemitsu S, Dull 20 Fleischman RA. Human piebald trait resulting from a dominant TJ, Chen E, Schlessinger J, Francke U, Ullrich A. Human proto- negative mutant allele of the c-kit membrane receptor gene. J Clin oncogene c-kit: a new cell surface receptor tyrosine kinase for an Invest 1992; 89: 1713–1717. unidentified ligand. EMBO J 1987; 6: 3341–3351. 21 Spritz RA, Holmes SA, Ramesar R, Greenberg J, Curtis D, Beighton 2 Qiu F, Ray P, Brown K, Baker PE, Jhanwar S, Ruddle FH, Besmer P. Mutations of the KIT (mast/stem cell growth factor receptor) P. Primary structure of c-kit: relationship with the CSF-1/PDGF proto-oncogene account for a continuous range of phenotypes in receptor kinase family-oncogenic activation of v-kit involves human piebaldism. Am J Hum Genet 1992; 51: 1058–1065. deletion of extracellular domain and C terminus. EMBO J 1988; 22 Siitonen T, Savolainen ER, Koistinen P. Expression of the c-kit 7: 1003–1011. proto-oncogene in myeloproliferative disorders and myelodysplas- 3 Crosier PS, Ricciardi ST, Hall LR, Vitas MR, Clark SC, Crosier KE. tic syndromes. Leukemia 1994; 8: 631–637. Expression of isoforms of the human receptor tyrosine kinase c-kit 23 Metcalfe DD. Classification and diagnosis of mastocytosis: current in leukemic cell lines and acute myeloid leukemia. Blood 1993; status. J Invest Dermatol 1991; 96 (Suppl): 2S–4S. 82: 1151–1158. 24 Longley J, Morganroth GS, Tyrrell L, Ding TG, Anderson DM, Wil- 4 Reith AD, Ellis C, Lyman SD, Anderson DM, Williams DE, Bern- liams DE, Halaban R. Altered metabolism of mast-cell growth fac- stein A, Pawson T. Signal transduction by normal isoforms and W tor (c-kit ligand) in cutaneous mastocytosis. New Engl J Med 1993; mutant variants of the Kit receptor tyrosine kinase. EMBO J 1991; 328: 1302–1307. 10: 2451–2459. 25 Nagata H, Worobec AS, Oh CK, Chowdhury BA, Tannenbaum S, 5 Nocka K, Majumder S, Chabot B, Prabir R, Cervone M, Bernstein Suzuki Y, Metcalfe DD. Identification of a point mutation in the A, Besmer P. Expression of c-kit gene products in known cellular catalytic domain of the proto-oncogene c-kit in peripheral blood targets of W mutations in normal and W mutant mice. Evidence mononuclear cells of patients with mastocytosis. Proc Natl Acad for an impaired c-kit kinase in mutant mice. Genes Devel 1989; Sci USA 1995; 92: 10560–10564. 3: 816–826. 26 Furitsu T, Tsujimura T, Tono T, Ikeda H, Kitayama H, Koshimizu 6 Ashman LK, Cambareri AC, To LB, Levinsky RJ, Juttner A. U, Sugahara H, Butterfield JH, Ashman LK, Kanayama Y, Matsu- Expression of the YB5.B8 antigen (c-kit proto-oncogene product) zawa Y, Kitamura Y, Kanakura Y. Identification of mutations in in normal human bone marrow. Blood 1989; 78: 30–37. the coding sequence of the proto-oncogene c-kit in a human mast Elevated expression of c-kit H Nagata et al 181 cell leukemia cell line causing ligand-independent activation of 33 Travis WD, Li CY, Hoagland HC, Travis LB, Banks PM. Mast cell c-kit product. J Clin Invest 1993; 92: 1736–1744. leukemia, report of a case and review of the literature. Mayo Clin 27 Butterfield JH, Weiler D, Dewald G, Gleich GJ. Establishment of Proc 1986; 61: 957–966. an immature mast cell line from a patient with mast cell leukemia. 34 Kirshenbaum AS, Kessler SW, Goff JP, Metcalfe DD. Demon- Leukemia Res 1988; 12: 345–355. stration of the origin of human mast cells from CD34+ bone mar- 28 Longley BJ, Tyrrell L, Lu S-Z, Ma Y-S, Langley K, Ding T-Q, Duffy row progenitor cells. J Immunol 1991; 146: 1410–1415. T, Jacobs P, Tang LH, Modlin T. Somatic c-kit activating mutation 35 Kirshenbaum AS, Goff JP, Kessler SW, Mican JM, Zsebo KM, Met- in urticaria pigmentosa and aggressive mastocytosis: establishment calfe DD. Effect of IL-3 and stem cell factor on the appearance of of clonality in a human mast cell neoplasms. Nat Genet 1996; 12: human basophil and mast cells from CD34+ pluripotent progeni- 312–315. tor cells. J Immunol 1992; 148: 772–777. ¨ 29 Chomczynski P, Sacchi N. Single-step method of RNA isolation 36 Agis H, Willheim M, Sperr WR, Wilfing A, Kromer E, Kabrna E, ¨ by acid guanidium thiocyanate-phenol-chloroform extraction. Spanblochl E, Strobl H, Geissler K, Spittler A, Boltz-Nitulescu G, Anal Biochem 1987; 162: 156–159. Majdic O, Lechner K, Valent P. Monocytes do not make mast cells 30 Maxam AM, Gilbert W. A new method for sequencing DNA. Proc when cultured in the presence of CF. Characterization of the circu- Natal Acad Sci USA 1977; 74: 560–564. lating mast cell progenitor as a c-kit+, CD34+,Ly−, CD14−, 31 Gelfand DH, White TJ. Thermostable DNA polymerases. In: Innis CD17−, colony-forming cell. J Immunol 1993; 151: 4221–4227. MA, Gelfand DH, Sninsky JJ, White TJ (eds). PCR Protocols. Aca- 37 Piao X, Curtis JE, Minkin S, Minden MD, Bernstein A. Expression demic Press: San Diego, 1990, pp 129–136. 32 Rottem M, Okada T, Goff JP, Metcalfe DD. Mast cells cultured of the Kit and Kit A receptor isoforms in human acute myelogen- from the peripheral blood of normal donors and patients with mas- ous leukemia. Blood 1994; 83: 476–481. tocytosis originate from a CD34+/Fc⑀RI-cell population. Blood 1994; 84: 2489–2496.