Notch-1 and Notch-2 Exhibit Unique Patterns of Expression in Human B-Lineage Cells FE Bertrand1, CE Eckfeldt1, AS Lysholm1 and TW Lebien1,2

Leukemia (2000) 14, 2095–2102  2000 Macmillan Publishers Ltd All rights reserved 0887-6924/00 $15.00 www.nature.com/leu Notch-1 and Notch-2 exhibit unique patterns of expression in human B-lineage cells FE Bertrand1, CE Eckfeldt1, AS Lysholm1 and TW LeBien1,2

1University of Minnesota Cancer Center, 2Department of Laboratory Medicine/Pathology, University of Minnesota, Minneapolis, MN, USA

The Notch genes encode a conserved family of receptors that bearing the extracellular domain of Jagged-1, exhibited an influence developmental fate in many species. Prior studies increase in primitive hematopoietic precursors.12 Recent stud- have indicated that Notch-1 and Notch-2 signaling influence the development of hematopoietic stems cells and thymocytes, but ies with human hematopoietic stem cells suggest that mainte- little is known regarding Notch expression and function in B- nance of precursors in the presence of a Notch signal may lineage cells. We analyzed the expression of Notch receptors reflect enhanced progression through the G0/G1 phase of the and Notch ligands in human B-lineage cells and bone marrow cell cycle.13 (BM) stromal cells. Notch-1 mRNA and protein is expressed The role of Notch signaling in thymocyte development has throughout normal B cell development and in leukemic B-lin- been investigated (reviewed in Refs 14 and 15). Constitutively eage cells. In contrast, Notch-2 expression is limited to pre-B cells expressing low levels of surface ␮. The Notch ligand Delta active Notch was originally reported to favor development of + + 16 ␣␤+ ␥␦+ is expressed in BM B-lineage cells. The Notch ligand Jagged- CD8 over CD4 thymocytes and TCR over TCR thy- 1 is not expressed in B-lineage cells, but is expressed in BM mocytes.17 More recent studies suggest that Notch-1 signaling stromal cells. These results suggest a model wherein lateral may render thymocytes more resistant to apoptotic sig- signaling between Notch and Delta on B-lineage cells and/or nals.18,19 A physiologic role for Notch signaling in normal Notch/Jagged-1 interactions between B-lineage cells and BM murine thymocyte development is reinforced by the finding stromal cells may regulate human B cell development. Leuke- that Jagged-1 and Jagged-2 are expressed on thymic mia (2000) 14, 2095–2102. 20 Keywords: Notch; B cell precursors epithelial cells. In contrast to the extensive analysis of Notch function in thymocyte development, much less is known regarding the Introduction role of Notch in mammalian B cell development. A recent study demonstrated that retroviral infection of constitutively active Notch-1 into murine bone marrow (BM) hematopoietic The Notch family of receptors and their DSL (Delta, Serrate, + + Lag-2) family of ligands mediate cell fate decisions in myogen- cells led to a precocious development of CD4 /CD8 T-lineage cells in BM, with a concomitant suppression of B cell esis, neurogenesis, adipogenesis, boundary formation, wing 21 development, eye development and lymphohematopoiesis in development. However, because B cell development was organisms ranging from drosophila to humans (reviewed in suppressed in this model of constitutive Notch signaling, it Refs 1–3). A potential common denominator of Notch func- was not possible to determine the expression and function of tion in these diverse cell fate decisions appears to be that Notch in cells developing and expanding into the B-lineage. Notch signaling maintains cells in an undifferentiated state; In the current study, we examined the expression of Notch most likely by preventing cells from responding to other devel- and Notch ligands in human fetal BM B-lineage cells and nor- opmental cues. Notch signaling initiates with the ligand- mal human BM stromal cells using RT-PCR and flow cytome- induced generation of an intracytoplasmic Notch (ICN) cleav- try. We show that Notch is expressed in normal and leukemic age product, followed by nuclear translocation of ICN and B-lineage cells and Notch ligands are expressed by B-lineage transcriptional activation or repression of a large number of cells and BM stromal cells. These results suggest a potentially target genes.1 complex role for Notch signaling in regulating the prolifer- The human genome contains four Notch receptor genes ation and/or differentiation of human B-lineage cells. (Notch-1, -2, -3 and -4) and at least three DSL ligand genes (Jagged-1, Jagged-2 and Delta).4–8 Human Notch-1 was originally identified at the breakpoint of the t(7;9)(q34;q34.3) Materials and methods chromosomal translocation in a subset of T cell acute lymphoblastic leukemias (ALL).9 Expression of Notch-1 mRNA Isolation of human B-lineage cells and FACS sorting was subsequently found in human CD34+ hematopoietic 10 Human fetal BM B-lineage cells and stromal cells were iso- stem cells. 22 Notch signaling exerts complex effects on mammalian hem- lated as described previously. Briefly, single cell suspensions atopoietic stem cells (reviewed in Ref. 3). For example, were layered on to Ficoll–Hypaque cushions and mono- truncated/constitutively active forms of Notch-1 and Notch-2 nuclear interface cells were collected following room tem- inhibit murine myeloid cell development in response to G- perature centrifugation at 1000 r.p.m. for 30 min. The mono- CSF and GM-CSF, respectively.11 Co-culture of the murine nuclear cells were washed, plated on plastic tissue culture 32D myeloid cell line with Jagged-1+ stromal cells inhibits G- dishes in RPMI-1640 supplemented with 10% heat-inacti- 6 − vated FBS and penicillin/streptomycin (Gibco-BRL, Gaithers- CSF-induced differentiation. Furthermore, murine Lin Sca- ° 1+c-kit+ cells (expressing Notch-1 and Notch-2) cultured on burg, MD, USA), and incubated for 2 h at 37 Cin5%CO2. 3T3 cells transfected with human Jagged-1, or with beads The adherent cells were expanded to confluence, passaged three times, and maintained in X-VIVO 10 serum-free medium (Bio-Whittaker, Walkersville, MD, USA) supplemented with 2 mml-glutamine prior to use.22 Correspondence: FE Bertrand, University of Minnesota Cancer Center, Mayo Mail Code 806 420 Delaware St SE, Minneapolis, MN 55455, The adherence-depleted mononuclear cells were stained USA; Fax: 612-624-2400 with anti-CD34, anti-CD19, and anti-␮ mAbs as previously Received 14 June 2000; accepted 11 August 2000 described.23 In most experiments the sample was split into two Notch expression in human B-lineage cells FE Bertrand et al 2096 halves. One half was stained with anti-CD34-FITC (HPCA-2; PCR primers, annealing temperature, and cycle number for Becton Dickinson Immunocytometry Systems (BDIS), San each gene analyzed. Jose, CA, USA) and anti-CD19-biotin revealed with streptavi- Ten microliters of each PCR product were separated on a din-PE (Molecular Probes, Eugene, OR, USA). The other half 1.5% agarose gel and transferred to a Nytran membrane was stained with anti-CD19-biotin (revealed with streptavidin- (Schleicher and Schuell, Keene, NH, USA). Following stan- PE) and anti-␮-FITC. The anti-CD19 and anti-␮ conjugates dard pre-hybridization the blots were hybridized at 42°C with have been previously described.24 In some experiments the the oligonucleotide probes listed in Table 1. Probes were lab- entire sample was stained with anti-CD34-FITC, anti-CD19- eled with digoxygenin (DIG) using the DIG Oligonucleotide allophycocyanin (25C1, BDIS) and anti-␮-biotin revealed with 3Ј-End Labeling Kit (Boehringer Mannheim) per the manufac- streptavidin-PE. The two staining/sorting strategies yielded turer’s instruction. Following washing at 42°C, hybridization equivalent results. Cells with lymphoid light scatter character- signals were revealed using the DIG Luminescent Detection istics were sorted into the following B-lineage compartments Kit for Nucleic Acids (Boehringer Mannheim) per the using a FACSVantage (BDIS): CD34+CD19+␮− pro-B cells, manufacturer’s instruction. CD34−CD19+␮− early pre-B cells, CD34−CD19+␮lo late pre-B cells and CD34−CD19+␮hi immature B cells.23 In experiments involving Western blot analysis, total fetal Western analysis of Notch-1 protein expression BM B cell precursors were isolated by magnetic bead depletion of non-B-lineage cells and surface ␮hi immature B Cells were lysed in RIPA buffer containing 10 mM Tris HCl + cells, as described previously.22 CD19 cells were thawed and pH 8.0, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 0.5% sodium sorted from cryopreserved tonsilar mononuclear cells. B-lin- deoxycholate, and 0.1% SDS supplemented with the follow- eage ALL cells were thawed from cryopreserved BM samples. ing protease inhibitors: aprotinin (22 ␮g/ml), PMSF (1 ␮g/ml), The use of all human tissue was approved by the Insti- DTT (1 ␮M), iodoacetamide (0.01 ␮M), pepstatin A (1 ␮g/ml) tutional Review Board: Human Subjects Committee of the and leupeptin (1 ␮g/ml). The protease inhibitors were University of Minnesota. obtained from Sigma Chemical Co (St Louis, MO, USA). Between 60 and 80 ␮g of total protein was electrophoresed per lane on a 7.5% SDS-PAGE gel with a 5% stacking gel. Isolation of murine B-lineage cells Following electrophoresis the proteins were transferred to nitrocellulose membranes and incubated overnight at 4°Cin BM from C57BL/6 mice was prepared by flushing femurs with 5% milk dissolved in 1 × TBST. Blots were incubated with 15 a 23-gauge needle to generate a single cell suspension. Spleen ␮g/ml of rat anti-human Notch-1 mAb 18G25 (produced in the and thymus were disaggregated in cold PBS using a mesh laboratory of S Artavanis-Tsakonas and kindly provided by I screen. Red cells were removed from the single cell suspen- Bernstein, Fred Hutchinson Cancer Research Center, Seattle), sions by incubation with hemolytic buffer comprised of 0.15 washed with 1 × TBST for 50 min at room temperature, and

M NH4Cl, 1.0 M NaHCO3 and 0.1 M Na2EDTA. Spleen and probed with a 1:5000 dilution of sheep anti-rat Ig conjugated BM cells were incubated with antibodies against murine to horseradish peroxidase (Amersham Life Science, Arlington CD19 and CD3 or CD19 alone, respectively, and FACS Heights, IL, USA) in 1 × TBST for 30 min at room temperature. purified using a FACSVantage. PE-conjugated anti-CD19 and Following a 90 min wash in 1 × TBST at room temperature FITC-conjugated anti-CD3 were obtained from BD- the blots were developed using enhanced chemiluminescence Pharmingen (San Diego, CA, USA). (Amersham Life Science). Equal protein loading between lanes was confirmed using mouse anti-human ␤-tubulin detected with sheep anti-mouse Ig conjugated to horseradish peroxi- RT-PCR and Southern analysis dase (Amersham). For experiments using murine BM, the blots were incubated Total RNA was extracted from 5 × 104 cells of each sorted with rabbit anti-human Notch-1 (kindly provided by J Aster, fraction using Tri-Reagent (Molecular Research Center, Cin- Brigham and Women’s Hospital, Boston) diluted 1:1000 in cinnati, OH, USA). Oligo-dT primed cDNA was synthesized 5% milk dissolved in 1 × TBST. Blots were counter-stained as follows: 10 ␮l of RNA was incubated with 1.6 ␮lof5× RT with donkey anti-rabbit Ig conjugated to horseradish peroxi- buffer (Gibco-BRL), 0.5 ␮l of DNAse I (Boehringer Mannheim, dase (Amersham) and developed by enhanced chemilumi- Indianapolis, IN, USA) and 0.5 ␮l of Rnasin (Promega, Madi- nescence as described above. The rabbit anti-human Notch- son, WI, USA) at 37°C for 1 h, followed by heat inactivation 1 (designated anti-TC) was made against a peptide comprising at 75°C. To this reaction was added 24 ␮l5× RT buffer, 2 ␮l amino acids 2278–2470 of human Notch-1.26 This antibody 0.1 M DTT, 1 ␮l of Oligo-dT, 1 ␮lof10mM dNTPs, and 1 ␮l cross-reacts with murine Notch-1.26 of RT (Gibco-BRL). Following 1 h incubation at 42°C and 10 min heat inactivation at 65°C, 1 ␮l of this preparation was used for each PCR reaction. Flow cytometric analysis of Notch-1 surface PCR was performed essentially as described previously.23 expression ␮ ␮ × Each PCR reaction mixture consisted of 39 lH2O, 5 l10 ␮ ␮ buffer (Gibco-BRL), 1.5 l50mM MgCl2,1 lof10mM BLIN-1 cells, a pre-B ALL cell line, or SUP T1 cells, a T-ALL dNTPs, 1 ␮l each primer at 20 ␮M, 2.5 U Taq polymerase cell line, were incubated with a 1:20 dilution of rabbit anti- (Gibco-BRL), and 1 ␮l cDNA. Cycle conditions were 95°C for rat Notch-1 extracellular domain (Notch-1EC) IgG (Upstate 1 min, annealing at various temperatures for 1 min (see Table Biotechnology, Lake Placid, NY, USA) for 20 min on ice. 1), and extension at 72°C for 1 min. A final extension at 72°C This Ab cross-reacts with mouse and human Notch-1EC. was performed for 10 min. Annealing temperatures and cycle Cells were washed in PBS + 1% calf serum, then incubated conditions for each primer pair were optimized such that each with a 1:50 dilution of human adsorbed goat anti-rabbit IgG PCR was in the linear range of amplification. Table 1 lists the conjugated to biotin (Southern Biotechnology Associates,

Leukemia Notch expression in human B-lineage cells FE Bertrand et al 2097 Table 1 Primers, oligo probes and cycle conditions used in RT-PCR analysis

Gene 5Ј Primer 3Ј Primer Annealing Cycle Probe Temp °C Number

GAPDH ATGGCCTTCCGTGTCCCCACTG TGAGTGTGGCAGGGACTCCCCA 60 25 TTCGTTGTCATACCAGGA VCAM-1 ATGACATGCTTGAGCCAGG GTGTCTCCTTCTTTGACACT 51 30 TTGGAGTAACTTGGATAATT Notch-1 GCCGCCTTTGTGCTTCTGTTC CCGGTGGTCTGTCTGGTCGTC 62 40 GCTGCTGTCCCGCAAGCGCC Notch-2 CATAGAATGATTAGCAGAGAG CAACATCAGAGCTAGCAAGAG 51 35 AGGGTGTCAACTGTGAGTATG Notch-3 TAGGAGGGAGAAGCCAAGTCG TAGCAGGAGGTACGTGCATG 40 35 AGACTGGGGCCTATTGCCTT Jagged-1 CAAGGAGATGTCACCAG CAGCTGCTCCCGCACGTTGT 59 30 GCTGCCGTTGCAGAAGTAAG Delta TGCCGCTGTGGTGGTCT AGTCCGCCTTCTTGTTGGTGT 62 40 ACGGAGACCATGAACAACCT

Birmingham, AL, USA). The biotinylated antibody was revealed with streptavidin-PE (Molecular Probes). Fetal BM mononuclear cells isolated by Ficoll–Hypaque centrifugation and adherence depletion were stained for Notch-1EC as above, except that the cells were incubated with 50 ␮g of human IgG for 10 min on ice prior to the addition of the anti-Notch-1EC. The fetal BM cells were also co-stained with anti-CD19 (25C1) that had been conjugated to FITC in our laboratory. Cells were analyzed on a FACSCalibur (BDIS) using CellQuest software.

Jagged-1 ﬂow cytometric and Western analysis

Jagged-1 surface staining was conducted using goat anti-rat Jagged-1 kindly provided by Frank Mortari (R&D Systems, Minneapolis, MN, USA). Binding was revealed with rabbit anti-goat FITC (Southern Biotechnology Associates). Cells were analyzed on a FACSCalibur (BDIS) using CellQuest software. For Western blotting, membranes were incubated with 20 ␮g/ml of the goat anti-rat Jagged-1 antibody, washed and then probed with a 1:10000 dilution of rabbit anti-goat Ig conjugated to horseradish peroxidase (Southern Biotechnol- ogy Associates). All other steps were as described above for Notch-1 Western analysis.

Results

Notch receptors are expressed in normal human Figure 1 B-lineage compartments isolated on the basis of CD34, B-lineage cells CD19 and surface ␮ expression. (a) Sorting gates on fetal BM cells stained for CD34 vs CD19 and CD19 vs ␮. Early pre-B cells were distinguished from pro-B cells on the basis of high CD19 surface Notch gene expression was analyzed in specific stages of expression on the former. (b) Post-sort analysis of cells gated in (a). human B cell development by sorting fetal BM B-lineage cells on the basis of CD34, CD19 and ␮ surface expression.23 The sorted compartments are shown in Figure 1a. CD34+CD19+␮− pro-B cells harbor DJH and some VDJH rearrangements and the vast majority express TdT.23,27 CD34−CD19+␮− early pre- B cells express cytoplasmic ␮ H chains, but have yet to assemble ␮ H chains on the cell surface with VpreB and ␭5.27 CD34−CD19+␮lo late pre-B cells are enriched for cells expressing surface ␮ H chains associated with VpreB and ␭5 to form the ␮-surrogate light chain (␺LC).28,29 The CD34−CD19+␮hi compartment is comprised of immature B cells that express surface ␮ H chains associated with ␬ or ␭ L chains. Post-sort analysis revealed that the purity of each compartment was greater than 95% (Figure 1b). Each sorted compartment was also analyzed by RT-PCR for the expression of B-lineage asso- Figure 2 Notch-1 and Notch-2 are differentially expressed in ciated genes as further confirmation that discrete stages of B human fetal BM B-lineage cells. Fetal BM cells FACS-purified as shown in Figure 1 were analyzed for Notch gene expression by RT- cell development were isolated (data not shown). PCR followed by Southern blotting. (a) and (b) were separate samples. − Figure 2 shows the results obtained when specific stages of RT , no reverse transcriptase; H2O, no cDNA template; Fetal BM, fetal B cell development were analyzed for Notch expression by BM mononuclear cells.

Leukemia Notch expression in human B-lineage cells FE Bertrand et al 2098 RT-PCR. Notch-1 transcripts were detected at all stages of of cell surface Notch-1, further confirming that human B- human fetal BM B cell development (Figure 2a). Interestingly, lineage cells express the Notch-1 receptor. Notch-2 transcripts were only detected in the late pre-B compartment of B cell development (Figure 2b). Notch-3 transcripts were not detected in BM B-lineage cells. This pattern Murine BM B-lineage cells express Notch-1 protein of expression was observed in five independently sorted and analyzed fetal BM samples. Notch-1 and Notch-2 transcripts The high degree of inter-species conservation of the Notch were also detected within the heterogeneous CD34+CD19− pathway, and the finding that Notch-1 is expressed throughout hematopoietic stem cell compartment (data not shown), human BM B cell development (Figures 2, 3 and 4), prompted consistent with previous studies.3,10 us to examine Notch-1 protein expression in murine B-lineage cells. Murine BM was FACS-purified into CD19+ and CD19− compartments and murine spleen was FACS-purified into Human B-lineage cells express Notch-1 protein CD19+ B-lineage cells and CD3+ T-lineage cells. Post-sort analysis revealed that each fraction was Ͼ90% pure (data not The expression of Notch-1 protein in human B-lineage cells shown). Murine thymocytes were used as positive control for was assessed by Western blotting using a rat mAb against the Notch-1 expression. As shown in Figure 5, Notch-1 protein cytoplasmic domain of human Notch-1. Results in Figure 3 was readily detected in murine thymocytes. In addition, show that the pre-B ALL cell line BLIN-1, normal human B CD19+ BM B-lineage cells also expressed Notch-1 protein. cell precursors, tonsil B cells, and leukemic cells from three Notch-1 protein was detected in CD3+ splenic T cells, but individual cases of B-lineage ALL all expressed p120 Notch- expression in CD19+ splenic B cells was below the level of 1. This is the expected size of the portion of the Notch-1 heter- detection. odimer containing the cytoplasmic domain, as reported by others.30 The Notch ligand Delta is expressed by normal human B-lineage cells Human B-lineage cells express surface Notch-1 receptors The sorted fetal BM B-lineage compartments were analyzed by RT-PCR for the expression of the Notch ligands Jagged-1 In order to determine if B-lineage cells express cell surface and Delta. Jagged-1 transcripts were not detected in BM B- Notch-1, we analyzed B-lineage cells for expression of the lineage cells (Figure 6). However, Delta was expressed in pro- Notch-1 extracellular region (Notch-1EC) by flow cytometry. B (CD34+CD19+), early pre-B (CD19+␮−) and late pre-B Analysis of the pre-B ALL cell line BLIN-1 revealed a high (CD19+␮lo) cells. level of cell surface Notch-1 expression (Figure 4a), consistent with the readily detectable Notch-1 expression observed by Western analysis (Figure 3). SUP T1 is a T-lineage ALL cell line BM stromal cells that support B-cell differentiation that bears two copies of a t(7;9) chromosomal translocation express Jagged-1 between Notch-1 and the TCR-␤ locus, resulting in expression of a constitutively active ICN. SUP T1 cells have no normal Our laboratory has developed an in vitro culture system that Notch-1 alleles and do not express cell surface Notch-1.30 supports development of IgM+ immature B cells from Consistent with this, the anti-Notch-1EC antibody did not react CD34+CD19− progenitors.24,31 This culture system utilizes with the cell surface of SUP T1 cells (Figure 4a). Figure 4b freshly isolated human BM stromal cells and does not require shows that CD19+ normal fetal BM cells expressed a low level the addition of exogenous cytokines. These BM stromal cell preparations were analyzed by RT-PCR for Notch ligand expression (Figure 7a). BM stromal cells expressed Jagged-1 but not Delta transcripts, although the latter was expressed in skin fibroblasts and HUVEC. Third passage fetal BM stromal cells were tested for Jagged-1 protein using a goat anti-Jagged- 1 antibody raised against the extracellular domain of rat Jag- ged-1. As shown in Figure 7b, fetal BM stromal cells expressed readily detectable cell surface Jagged-1. Rat fibroblasts were also Jagged 1+, whereas the pre-B ALL cell line BLIN-1 was Jagged-1−. Western blotting using this goat anti-rat Jagged-1 reagent revealed a protein of 150–180 kDa (Figure 7c), consistent with the size of human Jagged-1 previously detected by others.6,12

Discussion

Figure 3 Notch-1 protein is expressed in normal and leukemic B- The function of the Notch pathway in lymphopoiesis is begin- ␮ lineage cells. Whole cell lysates (80 g total protein/lane) were separ- ning to emerge and a role for Notch signaling in thymocyte ated on a 7% SDS-PAGE gel under reducing conditions and trans- development has been established (reviewed in Ref. 14). ferred to nitrocellulose for Western blotting. Membranes were probed with a rat anti-human Notch-1 mAb that recognizes the cytoplasmic Enforced expression of constitutively active Notch-1 in thymo- domain of Notch-1. Membranes were re-probed with mouse anti- cytes favors the development of TCR␣/␤ cells vs TCR␥/␦ human ␤-tubulin to conﬁrm equal protein loading between lanes. cells17 and the development of CD8+ cells vs CD4+ cells.16

Leukemia Notch expression in human B-lineage cells FE Bertrand et al 2099

Figure 4 CD19+ human fetal BM B-lineage cells express cell surface Notch-1. (a) BLIN-1 and SUP T1 cells were stained with rabbit anti- rat Notch-1EC (solid dark line) or normal rabbit IgG (dotted line) and analyzed by flow cytometry. (b) Adherence-depleted fetal BM mononuclear cells were co-stained with rabbit anti-rat Notch-1EC and mouse anti-human CD19 and analyzed by flow cytometry. CD19+ lymphoid cells defined by light scatter (gate R3) and CD19 expression (gate R4) express low levels of cell surface Notch-1 (far right histogram). The solid dark line is Notch-1EC staining. The dotted line is normal rabbit IgG staining.

Figure 5 Murine B-lineage cells express Notch-1 protein. Murine lymphoid populations from BM, spleen, and thymus were lysed, and ෂ80 ␮g protein was separated on a 7.5% SDS-PAGE gel under reducing conditions and transferred to nitrocellulose for Western blotting. Membranes were probed with rabbit anti-human Notch-1 that cross- reacts the murine Notch-1.26 Membranes were also probed with mouse anti-human ␤-tubulin (that cross-reacts with murine ␤-tubulin) to assess protein loading. Murine thymocytes were on a separate blot.

One interpretation of these results is that Notch-1 signaling alters the developmental fate of thymocytes mechanistically akin to what has been described in other developmental systems.1 More recent experiments suggest that Notch-1 signaling confers apoptotic resistance to developing thymocytes, which Figure 6 B-cell precursors express Delta. Fetal BM cells FACS-pur- may influence the pool size of thymocyte compartments.18,19 ified as shown in Figure 1 were analyzed for Notch ligand expression Moreover, retrovirus-mediated expression of constitutively by RT-PCR followed by Southern blotting. VCAM-1 expression was analyzed to ensure there were no stromal cells contaminating the active Notch-1 in early lymphoid progenitors promotes thym- − purified B-lineage compartments. RT , no reverse transcriptase; H2O, ocyte development at the expense of B cell development in no cDNA template; Fetal BM, fetal BM mononuclear cells. The 21 murine BM. This suggests that Notch signaling may influ- expression of VCAM-1 in fetal BM mononuclear cells probably ence the T- vs B-lineage fate decision, possibly via a dual out- represents rare stromal cells present in this population.

Leukemia Notch expression in human B-lineage cells FE Bertrand et al 2100

Figure 7 BM stromal cells express cell surface Jagged-1. (a) RT-PCR analysis and Southern blotting for Jagged-1 and Delta. RT−, no reverse

transcriptase; H2O, no cDNA template; Fetal BM, fetal BM mononuclear cells. (b) Flow cytometric analysis of Jagged-1 protein. Shaded histogram, control staining; open histogram, staining with goat anti-rat Jagged-1. (c) Western blot analysis for Jagged 1 expression. BM stromal cells express the ෂ180 kDa Jagged 1 protein (arrow), while BLIN-1 and JURKAT cells do not.

come of CBF-1-mediated induction of HES-1 and/or deltex- site in the Notch extracellular domain proximal to the trans- mediated suppression of E47 homodimers.14 However, studies membrane domain;44 an event that precedes generation of in fibroblasts and EBV-transformed B cell lines have suggested ICN by ␥-secretase activity. The role of Notch modifying that Notch signaling in B-lineage cells may be CBF-1 inde- genes and various enzymes in regulating Notch signaling in pendent.32,33 Pui and colleagues21 reported that Notch-1 was B-lineage cells will require further investigation. not detected in normal murine BM B-lineage cells en- It is interesting that Notch-2 gene expression is restricted to compassing Hardy fractions A-F when analyzed by RT-PCR. a compartment of pre-B cells (Figure 2b). The human In contrast, we have detected Notch-1 protein in murine CD19+␮lo compartment is enriched for large and small pre-B CD19+ cells from BM and spleen (Figure 5) using a rabbit anti- cells which express surface ␮ H chains in association with human Notch-1 antibody26 that is known to cross-react with VpreB and ␭5.45,46 Cells expressing the pre-BCR undergo murine Notch-1. The reason for this discrepancy is unclear. selection based on the ability of the ␮-H chains to pair with The current study provides direct evidence that Notch the ␺LC.47,48 B-lineage cells expressing ␮-H chains that fail to mRNA and protein are expressed in normal and leukemic pair with the ␺LC undergo programmed cell death.49 Human human B-lineage cells. Notch-1 is expressed at multiple stages fetal BM surface VpreB+ cells enriched by FACS express of B cell development, while Notch-2 expression is restricted Notch-2 by RT-PCR analysis (FE Bertrand et al, unpublished to the late pre-B (CD19+␮lo) compartment (Figures 2 and 3). observation). It is therefore conceivable that Notch-2 signaling We have also shown that CD19+ normal human BM B-lineage augments or modulates the cellular response to pre-BCR cells express cell surface Notch-1 receptors (Figure 4). This stimulation. For example, Notch-2 may confer an anti-apop- flow cytometric analysis of cell surface Notch-1 at an individ- totic signal to pre-BCR+ cells at a point when these cells are ual cell level rules out the unlikely possibility that rare non- poised to undergo L chain rearrangement. The anti-apoptotic B-lineage contaminating cells were the source of the Notch effects of Notch-1 signaling in thymocytes is consistent with signals observed in the RT-PCR and Western blotting this hypothesis.18,19 experiments. Human BM stromal cells used in the current study are a The ubiquitous expression of Notch-1 in human B cell population of fibroblast-like adventicial reticular cells that development suggests that Notch-1 signaling could modify support the differentiation of CD34+CD19− progenitors into proliferation and differentiation through several developmen- surface IgM+ immature B cells.24,31 These BM stromal cells tal checkpoints. However, it seems unlikely that all stages of express cell surface Jagged-1 but do not express Delta (Figure B cell development constantly undergo Notch activation by 7). This contrasts with the expression of Delta but not Jagged- virtue of simple ligand binding. Signaling through Notch-1 in 1 in pro-B and pre-B cells (Figure 6). These results raise the B-lineage cells could be regulated by mechanisms other than possibility that lateral signaling between B cell precursors may constitutive cell surface receptor expression. Genetic studies occur via Notch–Delta interactions. Notch signaling between in Drosophila have indicated the existence of Notch mod- equivalent cells of equal developmental potential, termed lat- ifying genes such as numb and fringe.34,35 For example, eral signaling, is a common paradigm in the Notch field that association of Fringe with Notch-1 impairs Serrate (Jagged)- explains how some cells may be signaled to commit to a parti- mediated signaling, but is permissive for Delta-mediated sig- cular fate, while others are maintained in the original precur- naling.36 Mammalian homologues of these genes have been sor state (see Ref. 1 for review). Such mechanism could poss- identified.37,38 In addition, current models of Notch function ibly play a role in maintaining normal homeostasis of B- indicate that generation of ICN must occur in order for a lineage developmental compartments. A report showing that Notch signal to be transduced (reviewed in Ref. 1). A wealth Rel/NF-␬B can induce Jagged-1 expression on murine splenic of recent data implicates presenilin-1 and/or its associated ␥- B cells also supports a lateral signaling model.50 Signaling secretase activity in the generation of ICN,39–42 although pre- between B cell precursors via Notch and Delta may also con- senilin-1 may not be a universal requirement for Notch sig- tribute to other cell–cell interactions described in B-lympho- naling.43 Furthermore, TNF-␣ converting enzyme cleaves a poiesis.51

Leukemia Notch expression in human B-lineage cells FE Bertrand et al 2101 In conclusion, we have provided direct evidence that Notch 11 Bigas A, Martin DI, Milner LA. Notch1 and Notch2 inhibit myeloid and Notch ligands are expressed in human B-lineage cells and differentiation in response to different cytokines. Mol Cell Biol BM stromal cells. Notch signaling in human B-lineage cells 1998; 18: 2324–2333. 12 Varnum-Finney B, Purton LE, Yu M, Brashem-Stein C, Flowers D, could be activated by the Delta ligand on B-lineage cells, or Staats S, Moore KA, Le Roux I, Mann R, Gray G, Artavanis- the Jagged-1 ligand on BM stromal cells. These signals may Tsakonas S, Bernstein ID. The notch ligand, jagged-1, influences act independently and/or cooperatively to modulate sub- the development of primitive hematopoietic cells. Blood 1998; 91: sequent gene expression, differentiation and apoptotic fate 4084–4091. within the B-lineage. 13 Carlesso N, Aster JC, Sklar J, Scadden DT. Notch1-induced delay of human hematopoietic progenitor cell differentiation is associated with altered cell cycle kinetics. Blood 1999; 93: 838–848. 14 Osborne B, Miele L. Notch and the immune system. Immunity Acknowledgements 1999; 11: 653–663. 15 Robey E. Regulation of T cell fate by Notch. Ann Rev Immunol We thank Barbara Ercole for technical assistance, Janet Peller 1999; 17: 283–295. of the University of Minnesota Cancer Center Flow Cytometry 16 Robey E, Chang D, Itano A, Cado D, Alexander H, Lans D, Core for cell sorting support, and Diane Hasz and David Lar- Weinmaster G, Salmon P. An activated form of notch influences the choice between CD4 and CD8 T cell lineages. Cell 1996; 87: gaespada for provision of murine lymphoid tissues/reagents. 483–492. We deeply appreciate provision of reagents from and dis- 17 Washburn T, Schweighoffer E, Gridley T, Chang D, Fowlkes BJ, cussion with Irv Bernstein (FHCRC, Seattle) and Jon Aster Cado D, Robey E. Notch activity influences the ␣␤ versus ␥␦ T (Brigham and Womens Hospital, Boston). We also thank cell lineage decision. Cell 1997; 88: 833–843. Sandi Sherman for assistance with the manuscript and mem- 18 Deftos ML, He YW, Ojata EW, Bevan MJ. Correlating Notch sig- bers of the LeBien laboratory for helpful discussion. This work naling with thymocyte maturation. Immunity 1998; 9: 777–786. 19 Jehn BM, Bielke W, Pear WS, Osborne BA. Protective effects of was supported by grants R01 CA31685 and R01 CA76055 notch-1 on TCR-induced apoptosis. J Immunol 1999; 162: 635– from the National Institutes of Health, the Graduate School of 638. the University of Minnesota, the Minnesota Medical Foun- 20 Felli MP, Maroder M, Mitsiadis TA, Campese AF, Bellavia D, dation, and the Apogee Enterprises Professorship. ASL was the Vacca A, Mann RS, Frati L, Lendahl U, Gulino A, Screpanti I. recipient of the Margaret Mitchell Research Scholarship from Expression pattern of notch-1, 2, and 3 and Jagged1 and 2 in the American Cancer Society (Minnesota Council). FEB is sup- lymphoid and stromal thymus components: distinct ligand-recep- ported by The Chris P Tkalcevic Foundation for Leukemia tor interactions in intrathymic T cell development. Int Immunol 1999; 11: 1017–1025. Research as a Special Fellow of the Leukemia and Lym- 21 Pui JC, Allman D, Xu L, DeRocco S, Karnell FG, Bakkour S, Lee phoma Society. JY, Kadesch T, Hardy RR, Aster JC, Pear WS. Notch 1 expression in early lymphopoiesis influences B versus T lineage determination. Immunity 1999; 11: 299–308. References 22 Pribyl JAR, Shah N, Dittel BN, LeBien TW. Methods for purifi- cation and growth of human B cell precursors in bone marrow 1 Artavanis-Tsakonas S, Rand MD, Lake RJ. Notch signaling: cell stromal cell-dependent cultures. In: Lefkovits I (ed). Immunology fate control and signal integration in development. Science 1999; Methods Manual. Academic Press: London, 1997, pp 902–923. 284: 770–776. 23 Bertrand FE, Billips LG, Burrows PD, Kubagawa H, Schroeder HW 2 Robey E. Notch in vertebrates. Curr Opin Genet Dev 1997; 7: Jr. IgH gene segment transcription and rearrangement prior to sur- 551–557. face expression of the pan B cell marker CD19 in normal human 3 Milner LA, Bigas A. Notch as a mediator of cell fate determination bone marrow. Blood 1997; 90: 736–744. in hematopoiesis: evidence and speculation. Blood 1999; 93: 24 Pribyl JAR, LeBien TW. IL-7 independent development of human 2431–2448. B cells. Proc Natl Acad Sci USA 1996; 93: 10348–10353. 4 Larson C, Lardelli M, White I, Lendahl U. The human NOTCH1, 25 Zagouras P, Stifani S, Blaumueller CM, Carcangiu ML, Artavanis- 2 and 3 genes are located at chromosome positions 9q34, 1p13- Tsakonas S. Alterations in Notch signaling in neoplastic lesions of p11, and 19p13.2-p13.1 in regions of neoplasia-associated trans- the human cervix. Proc Natl Acad Sci USA 1995; 92: 6414–6418. location. Genomics 1994; 15: 253–258. 26 Hasserjian RP, Aster JC, Davi F, Weinberg DS, Sklar J. Modulated 5 Li L, Huang GM, Banta AB, Deng Y, Smith T, Dong P, Friedman expression of Notch 1 during thymocyte development. Blood C, Chen L, Trask BJ, Spies T, Rowen L, Hood L. Cloning, charac- 1996; 88: 970–976. terization, and the complete 56.8-kilobase DNA sequence of the 27 Loken M, Shah VO, Dattilio KL, Civin CI. Flow cytometric analysis human NOTCH4 gene. Genomics 1998; 51: 45–58. of human bone marrow. II. Normal B lymphocyte development. 6 Li L, Milner LA, Deng Y, Iwata M, Banta A, Graf L, Marcovina S, Blood 1987; 70: 1316–1324. Friedman C, Trask BJ, Hood L, Torok-Storb B. The human homolog 28 Burrows PD, Cooper MD. B cell development and differentiation. of rat Jagged-1 expressed by marrow stroma inhibits differentiation Curr Opin Immunol 1997; 9: 239–244. of 32D cells through interaction with Notch1. Immunity 1998; 8: 29 LeBien TW. B-cell lymphopoiesis in mouse and man. Curr Opin 43–55. Immunol 1998; 10: 188–195. 7 Luo B, Aster JC, Hasserjian RP, Kuo F, Sklar J. Isolation and func- 30 Aster J, Pear W, Hasserjian R, Erba H, Davi F, Luo B, Scott M, tional analysis of cDNA for human Jagged2, a gene encoding a Baltimore D, Sklar J. Functional analysis of the TAN-1 gene, a ligand for the Notch1 receptor. Mol Cell Biol 1997; 17: 6057– human homolog of Drosophila Notch. Cold Spring Harbor Symp 6067. Quant Biol 1994; LIX: 125–136. 8 Gray GE, Mann RS, Mitsiadis E, Henrique D, Carcangiu M-L, 31 Kurosaka D, LeBien T, Pribyl JAR. Comparative studies of different Banks A, Leiman J, Ward D, Ish-Horowitz D, Artavanis-Tsakonas stromal cell microenvironments in support of human B-cell devel- S. Human ligands of the Notch receptor. Am J Pathol 1999; 154: opment. Exp Hematol 1999; 27: 1271–1281. 785–794. 32 Callahan J, Aster J, Sklar J, Robertson ES. Intracellular forms of 9 Ellisen LW. Bird J, West DC, Soreng AL, Reynolds TC, Smith SD, human NOTCH1 interact at distinctly different levels with RBP- Sklar J. TAN-1, the human homolog of the Drosophila notch gene, Jkappa in human B and T cells. Leukemia 2000; 14: 84–92. is broken by chromosomal translocation in T lymphoblastic neo- 33 Ordentilich P, Lin A, Shen CP, Blaumueller C, Matsuno K, Artav- plasms. Cell 1991; 66: 649–661. anis-Tsakonas S, Kadesch T. Notch inhibition of E47 supports the 10 Milner LA, Kopan R, Martin DI, Bernstein ID. A human homologue existence of a novel signaling pathway. Mol Cell Biol 1998; 18: of the Drosophila developmental gene, notch, is expressed in 2230–2239. CD34+ hematopoietic precursors. Blood 1994; 83: 2057–2062. 34 Guo M, Jan LY, Jan YN. Control of daughter cell fates during asym-

Leukemia Notch expression in human B-lineage cells FE Bertrand et al 2102 metric division: interaction of Numb and Notch. Neuron 1996; 44 Brou C, Logeat F, Gupta N, Bessia C, LeBail O, Doedens JR, Cum- 17: 27–41. ano A, Roux P, Black RA, Israel A. A novel proteolytic cleavage 35 Panin VM, Papayannopoulos V, Wilson R, Irvine KD. Fringe involved in notch signaling: the role of the disintegrin-metallopro- modulates Notch-ligand interactions. Nature 1997; 387: 908–912. tease TACE. Mol Cell 2000; 5: 207–216. 36 Ju BG, Jeong S, Bae E, Hyun S, Carroll SB, Yim J, Kim J. Fringe 45 Wang Y-H, Nomura J, Faye-Petersen OM, Cooper MD. Surrogate forms a complex with Notch. Nature 2000; 405: 191–195. light chain production during B cell differentiation: differential 37 Zhong W, Feder JN, Jiang MM, Jan LY, Jan YN. Asymmetric local- intracellular versus cell surface expression. J Immunol 1998; 161: ization of a mammalian numb homolog during mouse cortical 1132–1139. neurogenesis. Neuron 1996; 17: 43–53. 46 Lassoued, K., Nun˜ez CA, Billips L, Kubagawa H, Monteiro RC, 38 Moran JL, Johnston SH, Rauskolb C, Bhalerao J, Bowcock AM, LeBien TW, Cooper MD. Expression of surrogate light chain recep- Vogt TF. Genomic structure, mapping, and expression analysis of tors is restricted in a late stage in pre-B cell differentiation. Cell the mammalian Lunatic, Manic, and Radical fringe genes. Mamm 1993; 73: 73–86. Genome 1999; 10: 535–541. 47 ten Boekel E, Melchers F, Rolink AG. Changes in the V(H) gene 39 Ye Y, Lukinova N, Fortini ME. Neurogenic phenotypes and altered repertoire of developing precursor B lymphocytes in mouse bone Notch processing in Drosophila Presenilin mutants. Nature 1999; marrow mediated by the pre-B cell receptor. Immunity 1997; 7: 398: 525–529. 357–368. 40 Struhl G, Greenwald I. Presenilin is required for activity and nuclear 48 Wasserman R, Li YS, Shinton SA, Carmack CE, Manser T, Wiest access of Notch in Drosophila. Nature 1999; 398: 522–525. DL, Hayakawa K, Hardy RR. A novel mechanism for B cell reper- 41 De Strooper B, Annaert W, Cupers P, Saftig P, Craessaerts K, toire maturation based on response by B cell precursors to pre-B Mumm JS, Schroeter EH, Schrijvers V, Wolfe MS, Ray WJ, Goate receptor assembly. J Exp Med 1998; 187: 259–264. A, Kopan R. A presenilin-1-dependent gamma secretase-like pro- 49 Rajewsky K. Clonal selection and learning in the antibody system. tease mediates release of Notch intracellular domain. Nature Nature 1996; 381: 751–758. 1999; 398: 518–522. 50 Bash J, Zong WX, Banga S, Rivera A, Ballard DW, Ron Y, Gelinas 42 Mumm JS, Schroeter EH, Saxena MT, Griesemer A, Tian X, Pan C. Rel/NF-kappaB can trigger the Notch signaling pathway by DJ, Ray WJ, Kopan R. A ligand-induced extracellular cleavage inducing the expression of Jagged1, a ligand for Notch receptors. regulates ␥-secretase-like proteolytic activation of Notch-1. Mol EMBO J 1999; 18: 2803–2811. Cell 2000; 5: 197–206. 51 Ray RJ, Stoddart A, Pennycook JL, Huner HO, Furlonger C, 43 Berechid BE, Thinakaran G, Wong PC, Sisodia SS, Nye JS. Lack Wu GE, Paige CJ. Stromal cell-independent maturation of of requirement for presenilin1 in notch1 signaling. Curr Biol 1999; IL-7-responsive pro-B cells. J Immunol 1998; 160: 5886– 9: 1493–1496. 5897.

Leukemia