Insertion of c-Myc into Igh Induces B-Cell and Plasma-Cell Neoplasms in Mice

We used gene targeting in mice to insert a His(6)-tagged mouse c-Myc cDNA, Myc(His), head to head into the mouse immunoglobulin heavy-chain locus, Igh, just 5' of the intronic enhancer, Emu. The insertion of Myc(His) mimicked both the human t(8;14)(q24;q32) translocation that results in the activation of MYC in human endemic Burkitt lymphomas and the homologous mouse T(12;15) translocation that deregulates Myc in certain mouse plasmacytomas. Beginning at the age of 6 months, Myc(His) transgenic mice developed B-cell and plasma neoplasms, such as IgM(+) lymphoblastic B-cell lymphomas, Bcl-6(+) diffuse large B-cell lymphomas, and CD138(+) plasmacytomas, with an overall incidence of 68% by 21 months. Molecular studies of lymphoblastic B-cell lymphoma, the most prevalent neoplasm (50% of all tumors), showed that the lymphomas were clonal, overexpressed Myc(His), and exhibited the P2 to P1 promoter shift in Myc expression, a hallmark of MYC/Myc deregulation in human endemic Burkitt lymphoma and mouse plasmacytoma. Only 1 (6.3%) of 16 lymphoblastic B-cell lymphomas contained a BL-typical point mutation in the amino-terminal transactivation domain of Myc(His), suggesting that most of these tumors are derived from naive, pregerminal center B cells. Twelve (46%) of 26 lymphoblastic B-cell lymphomas exhibited changes in the p19(Arf)-Mdm2-p53 tumor suppressor axis, an important pathway for Myc-dependent apoptosis. We conclude that Myc(His) insertion into Igh predictably induces B-cell and plasma-cell tumors in mice, providing a valuable mouse model for understanding the transformation-inducing consequences of the MYC/Myc-activating endemic Burkitt lymphoma t(8;14)/plasmacytoma T(12;15) translocation.


Introduction
Reciprocal chromosomal translocations that activate the cellular oncogene MYC are widely accepted as the initiating events in the natural history of human endemic Burkitt lymphoma, an aggressive form of mature, postgerminal center, non-Hodgkin lymphoma.The most common translocation found in endemic Burkitt lymphoma is t(8;14)(q24;q32), which recombines MYC at 8q24 with the immunoglobulin heavy-chain gene locus, IGH, at 14q32 (1,2).The breakpoints of t (8;14) usually occur in the joining gene region of the IGH locus, J H , and in the 5V flank of MYC.The resulting exchanges allocate the intact MYC gene, including the noncoding regulatory first exon with the two major MYC promoters, P1 and P2, to the MYC-deregulated product of translocation, where MYC is apposed to the intronic heavy-chain enhancer, EA, in opposite transcriptional orientation (Fig. 1).Mouse plasmacytoma, a malignant tumor of terminally differentiated B cells, is thought to be initiated, just like human endemic Burkitt lymphoma, by Mycactivating chromosomal translocations (3).The most common translocation observed in plasmacytoma is the T (12;15), the direct counterpart of the human t(8;14) translocation.A recent molecular analysis of T (12;15) in IL-6 transgenic plasmacytomas and their precursors has shown that 8 (18%) of 45 translocation breakpoint regions exhibited the endemic Burkitt lymphoma t(8;14)-typical Myc-EA juxtaposition on the Myc-deregulated product of translocation (Fig. 1; ref. 4).Thus, homologous MYC/Myc-EA rearrangements lead to B-cell and plasma-cell neoplasms in humans and mice.
The mechanism by which human endemic Burkitt lymphoma t (8;14) and mouse plasmacytoma T (12;15) deregulate the expression of MYC/Myc and, thereby, promote the malignant transformation of B lymphocytes and plasma cells has not been shown (5).Further insights into this mechanism might be possible if a mouse model that accurately reproduces the key features of the endemic Burkitt lymphoma t(8;14)/plasmacytoma T(12;15) could be produced.Previous attempts to develop such a model used mice that harbor multiple copies of human MYC or mouse Myc under control of the human or mouse EA enhancer (6)(7)(8)(9).Clearly, the EA-driven MYC/Myc genes are incomplete models of the endemic Burkitt lymphoma t(8;14)/plasmacytoma T (12;15) because they reproduce just one aspect of this translocation, the juxtaposition of EA and MYC/Myc (5).Another approach used mice that carry a 240-kb single-copy yeast artificial chromosome (YAC), which consists of a human germ line VDJ-CA-Cy locus and the MYC gene from the BL cell line Raji (10,11).Serious shortcomings of the YAC-MYC approach include the absence of the Ea enhancer, which is believed to be the critical element for MYC up-regulation in tumors of late B and plasma cells (12), the unusual location of MYC in the IGH locus (3V of Cy) and the head-to-tail orientation of MYC relative to IGH, features that are not observed in human endemic Burkitt lymphoma or mouse plasmacytoma.Consistent with the apparent inability of the YAC-MYC and EA-MYC/Myc transgenes to faithfully recreate the endemic Burkitt lymphoma t(8;14)/plasmacytoma T (12;15), the transgenic mice develop predominantly lymphomas with features of pre-B, immature B, or transitional B cells, rather than mature B cells (human endemic Burkitt lymphoma) or plasma cells (mouse plasmacytoma).
To model endemic Burkitt lymphoma t(8;14)/plasmacytoma T(12;15) more precisely, we inserted a His 6 -tagged mouse Myc cDNA in the J H -EA intervening region of mouse Igh and developed a mouse strain, designated iMyc EA , in which Myc is juxtaposed head to head with EA, just as in human endemic Burkitt lymphoma and mouse plasmacytoma (Fig. 1).Between the age of 6 and 21 months, nearly 70% of the heterozygous iMyc EA mice developed lymphoblastic B-cell lymphomas with a BL-like morphology, diffuse large B-cell lymphomas and plasmacytomas.All these are tumors of mature B cells.The observed spectrum of tumors provided strong empirical evidence that the iMyc EA transgene mimicked the critical features of the endemic Burkitt lymphoma t(8;14)/plasmacytoma T(12;15) translocation.

Materials and Methods
Generation of iMyc EA Mice.Gene targeting (13) and cre-loxP recombination (14) were used to insert a mouse Myc gene into the mouse germ line Igh locus.The inserted Myc consisted of an intronless cDNA clone, the noncoding first exon with the natural P1/P2 promoter, 1.5 kb of genomic 5V flank containing the normal transcription-regulatory region, and a short stretch of 3V untranslated region harboring the Myc major polyadenylation site.The untranslated region downstream of the polyadenylate signal, which has been shown to be dispensable in vivo (15), was not present in the construct.The Myc also contained the coding sequence for an artificial histidine tag, His 6 , added in frame at the 3V end of the gene.The tag made it possible to distinguish Myc mRNA and Myc protein encoded by the inserted Myc His gene on chromosome 12 from the endogenous Myc gene (no tag) on chromosome 15.Myc His was inserted in the intervening region between J H 4 and EA, f0.5 kb 5V of EA, in opposite transcriptional orientation to Igh.The construction of the inserted gene, Myc His Neo loxP , the assembly of the targeting vector, and the generation of Neo-less iMyc EA mice are described in the Supplementary Materials and Methods section and illustrated in Supplementary Fig. 1.This study used heterozygous transgenic mice harboring one targeted Igh allele, iMyc EA , and one wild-type Igh allele.The targeted and wild-type Igh alleles were from strains 129SvJ (129) and C57BL/6 (B6), respectively.The genetic background of the iMyc EA colony was initially a variable, segregating mixture of 129 and B6 alleles.The mice used for the studies reported here were from generations three to six of an ongoing backcross of the iMyc EA transgene onto B6, resulting in an average content of B6 alleles from 93.8% (N 3 ) to 99.2% (N 6 ).
Turnover of ''Premalignant'' B Splenocytes.Cell surface marker expression was assessed using a FACSCalibur flow cytometer (BD Biosciences Pharmingen, San Diego, CA) and a panel of reagents useful for distinguishing subsets of hematopoietic cells.Total Myc mRNA levels, encoded by Myc and Myc His , were determined in fluorescence-activated cell-sorting (FACS)-purified subpopulations of bone marrow B cells using the quantitative PCR assay on demand kit from Applied Biosystems (Foster City, CA).The turnover of MACS-purified B220 + splenic B cells was measured by bromodeoxyuridine (BrdUrd) incorporation in vivo using the kit from BD Biosciences Pharmingen.For cell cycle analysis, B cells were stained with 50 Ag/mL propidium iodide in buffer containing 0.1% sodium citrate and 0.1% Triton X-100 followed by FACS analysis.For determination of apoptosis, the FITC-conjugated monoclonal active caspase-3 antibody kit I was used (BD Bioscience).Proliferation of B cells in vitro was measured with the assistance of the colorimetric 3-(4,5-dimethylthiazol-2-yl)-5-(3carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt (MTS) assay (CellTiter 96 Aqueous Non-Radioactive Cell Proliferation Assay) from Promega (Madison, WI).To that end, B220 + splenic cells were purified by MACS (Miltenyi, Auburn, CA) and grown for 48 hours in RPMI 1640 supplemented with 25 Ag/mL lipopolysaccharide (LPS) from Escherichia coli 055:B5 (Sigma, St. Louis, MO), 10% fetal calf serum, 200 mmol/L L-glutamine, 50 Amol/L 2-mercaptoethanol, and penicillin/streptomycin (Life Technologies, Inc., Gaithersburg, MD).The immunologic competence of iMyc EA mice was assessed by s.c.injection of 100 Ag of the synthetic peptide CLQLKEQKNAGTRTNEG that had been conjugated to keyhole limpet hemocyanin and emulsified in complete Freund's adjuvant.This was followed by two i.p. booster injections of keyhole limpet hemocyaninpeptide in incomplete Freund's adjuvant spaced 3 weeks apart.Total immunoglobulin levels and peptide-specific antibodies were measured by ELISA as previously described (16).
Histological and Immunohistochemical Examination of Tumors.Histological features of B-cell and plasma-cell tumors were assessed in 5-Am-thick sections of paraffin-embedded tissues stained with H&E, Giemsa according to Lennert, or periodic acid-Schiff.Apoptosis was determined by using the terminal deoxynucleotidyl transferase-mediated nick end labeling (TUNEL) assay (17).Briefly, tissue sections were digested with proteinase K and placed in a reaction buffer containing digoxigeninlabeled dUTP and deoxynucleotidyl transferase (15 units/AL).Sections were incubated with anti-digoxigenin antibody, developed with New Fuchsin substrate, and counterstained with hematoxylin.To measure proliferation of tumor cells in vivo, mice were pulsed with BrdUrd (200 mg/m 2 i.p., 8 hours), followed by immunostaining of tissue sections with antibody to BrdUrd.To visualize surface antigen expression in tumor cells, tissue sections were deparaffinized, treated with hot steam or microwaving to unmask antigen, incubated with 1% hydrogen peroxide to eliminate endogenous peroxidase activity, blocked with 5% goat serum, incubated with primary antibody overnight at 4jC, and detected by horseradish peroxidase-conjugated secondary antibody (1:200) and 3,3Vdiaminobenzidine. Bcl-6 expression in diffuse large B-cell lymphomas was detected with a rabbit polyclonal antibody (N-3, Santa Cruz Biotechnology, Santa Cruz, CA) as previously described (18).
Molecular and Cytogenetic Studies with Lymphoblastic B-Cell Lymphoma.For Southern hybridization of clonotypic V(D)J rearrangements, genomic DNA obtained from tumor tissues was digested with EcoRI ( for Igh) or EcoRI and BamHI ( for Igk), fractionated by electrophoresis on 0.7% agarose gels, transferred onto a nitrocellulose membrane, and hybridized to a [ 32 P]dCTP-labeled 1.5-kb HindIII/EcoRI Igh probe (pJ11) spanning J H 3 and EA or a 1.1-kb Cn probe.The latter probe was generated by PCR using a primer pair ( forward: 5V-GAT GCT GCA CCA ACT GTA TCC A; reverse: 5V-GGG GTG ATC AGC TCT CAG CTT) and method developed by Dr. Michael Kuehl (National Cancer Institute, NIH).For spectral karyotyping, metaphase chromosomes were prepared from primary tumor cells ex vivo or tumor cells cultured in vitro for 1.5 hours in 20 Ag/mL Colcemid (Life Technologies).Cells were lysed in hypotonic KCl solution; chromosomes were fixed in methanol-acetic acid (3:1); and spectral karyotyping analysis was done as previously described (19).For each tumor, at least 10 complete metaphase plates were analyzed.
Detection of Myc His by Immunoblotting and Total Myc mRNA by Northern Blotting.Tissue samples were homogenized in 50 mmol/L TRIS, 150 mmol/L NaCl, 0.5% Triton X-100, 0.5% SDS, 0.5% sodium deoxycholate, and 10% glycerol.Proteins from clarified lysates (40 Ag) were resolved electrophoretically in denaturing 10% SDS-PAGE gels and transferred by electroblotting to nitrocellulose membranes.Membranes were probed with rabbit anti-His 6 -tag antibody (1:5,000, Clontech, Palo Alto, CA).The position of the Myc His protein was visualized with horseradish peroxidase-conjugated goat anti-rabbit antibody (1:5,000; Amersham, Arlington Heights, IL) using the chemiluminescence detection kit from Amersham.To confirm equal loading, the membrane was stripped and reprobed using an antibody specific for -actin (Sigma).For Northern blot analysis of Myc mRNA, polyadenylated RNA was extracted from tumor nodules, fractionated by electrophoresis on formaldehyde-containing 1.0% agarose gels, transferred to Hybond N membranes, and hybridized to a Myc exon 1 probe (pMmyc54) that was nick translated to a specific activity of 2 Â 10 8 cpm/Ag (20).
Determination of Myc and Myc His mRNA Levels by Allele-Specific Reverse Transcription-PCR.Myc His and Myc expression was determined by semiquantitative reverse transcription-PCR (RT-PCR) using a common 5V primer (5V-TCT CCA CTC ACC AGC ACA AC-3V) but 3V primers specific either for Myc His (5V-CCT CGA GTTAGG TCA GTT TA-3V) or wild-type Myc (5V-(5V-ATG GTG ATG GTG ATG ATG AC-3V).cDNA was synthesized using the AMV reverse transcriptase kit (Roche, Indianapolis, IN) and amplified by PCR using the following thermal cycling conditions: 95jC for 5 minutes (initial template denaturation) followed by 20 cycles of amplification at 62jC (primer annealing), 72jC (extension) and 95jC (melting), each for 1 minute.PCR amplification of the housekeeping gene Gapd (glyceraldehyde-3-phosphate dehydrogenase) was done for each sample as a control using the following primer pair: 5V-GGT GGA GCC AAA CGG GTC ATC ATC C-3V and 5V-CAC ATT GGG GGT AGG AAC ACG GAA GG-3V.
Analysis of Myc Promoter Shift by RNase Protection Analysis.RNA was extracted from B220 + MACS-purified splenic B cells and tumor tissues of iMyc EA mice with the TRIZOL reagent (Invitrogen, San Diego, CA) and then treated with DNase I. RNA samples from mouse plasmacytomas TEPC 1165 and ABPC 20, pooled thymus glands from twenty 4-to 6-weeks-old BALB/c mice and splenic B cells from C57BL/6 mice were included as controls.A 1,144-base antisense RNA probe was synthesized by T7 RNA polymerase using a MAXIscript in vitro transcription kit from Ambion (Austin, TX).The probe was synthesized in the presence of [a- 32 P]UTP at 800 Ci/mmol, 40 mCi/mL (PerkinElmer Life Sciences, New England Nuclear, Boston, MA) and purified on a 6% acrylamide-urea gel.The Myc probe covered sequences from an Eco RI star site 626 bp upstream of the P1 promoter to the Sst I site near the 3V end of exon 1. Ribonuclease protection was done with the Ambion RPA III kit following the instructions of the manufacturer.Briefly, 10 Ag of total RNA was hybridized with about 1 ng of Myc probe at 58jC for 16 hours, followed by digestion with 0.4 units of RNase A and 15 units of RNase T1 at 37jC for 30 minutes.The protected products were resolved on a 6% acrylamide-urea gel.Transcripts derived from the Myc P1 promoter protected 518 bases of the probe and transcripts derived from the P2 promoter protected 355 bases of the probe.Signals were quantitated with a Storm phosphor imager (Molecular Dynamics, Sunnyvale, CA) and normalized to correct signal strength for the number of uridine residues in the protected probe fragments.RNAs with known P1/P2 ratios of about 2 (TEPC 1165) and 0.4 (ABPC 20), respectively, were used as controls (21).
Detection of Point Mutations in Myc Amino-Terminal Transactivation Domain.Genomic DNA was purified from BLL using the Trizol reagent (Life Technologies), removing contaminating mRNA by digestion with Rnase A. Myc His was amplified by PCR using the following thermal cycling conditions: 95jC for 5 minutes followed by 30 cycles of amplification at 62jC for 20 seconds, 72jC for 60 seconds, and 95jC for 20 seconds.PCR primer sequences were 5V-ACA ATC TGC GAG CCA GGA CAG GAC T-3V ( forward) and 5V-TCC TCA TCT TCT TGC TCT TCT TCA GAG T-3V (reverse).The resultant amplicon was 942 bp long and contained the protein-encoding exons 2 and 3 of Myc His together with the regulatory exon 1, which harbors the P1/P2 promoter.Normal Myc sequence was not amplified because the primers were separated by f2.5 kb of intron 1 and 2 sequence, which was not present in the Myc His cDNA.PCR products were sequenced using equipment and reagents from Applied Biosystems.

Results
Myc His Is Expressed throughout B-Cell Development.Because onset and level of Myc His expression during B-cell development are likely to play an important role in determining incidence and types of lymphomas that might arise in iMyc EA mice, we assessed the expression of Myc His protein in lysates of FACS-sorted B cells at different maturation stages using immunoblotting with antibody to the His 6 tag (Fig. 2A).B220 + CD43 + pro-B cells ( fraction ABC according to Hardy;refs. 22,23) and B220 + CD43 À pre-B cells (DEF) collected from the bone marrow exhibited a 1.7-fold difference in Myc His levels after normalization to the actin loading controls (experiment 1).Further dissection of the pre-B cell compartment DEF into fractions D (early pre-B; B220 + CD43 À IgM À ) and EF (late pre-B; B220 + CD43 À IgM + ) showed that the Myc His levels varied again in a similar range (f2-fold) in these fractions (experiment 2).Myc His expression was also comparable in transitional and mature B splenocytes, including T1 B cells (IgM bright IgD dull ), T2 B cells (IgM bright IgD bright ), mature (M) B cells (IgM dull IgD bright ), and sIg + n + B cells further separated based on CD138 expression (experiment 3).These results showed that Myc His is overexpressed throughout B-cell development, presumably driven by the EA enhancer, which is thought to be active at relatively constant levels throughout B-cell development (24).
To determine the elevation of Myc mRNA caused by insertion of Myc His in the vicinity of EA, we did quantitative PCR in three major B-cell fractions collected by FACS based on expression of B220 and IgM (Fig. 2B).Total Myc mRNA was f10-fold higher in B220 low IgM À pro-B/pre-B cells (columns 1 and 4), B220 low IgM + transitional B cells (columns 2 and 5), and B220 hi IgM + mature B cells (columns 3 and 6) obtained from the bone marrow of iMyc EA mice, compared with the same cell populations in normal littermates.These findings were consistent with EA controlling Myc His expression and suggested that the constitutive overexpression of Myc His leads to chronic changes in the B-cell compartment of iMyc EA mice before malignant transformation occurs.
Increased Turnover of Premalignant Myc His Transgenic B cells In vivo.To evaluate whether constitutive expression of Myc His leads to chronic activation and expansion of B cells before their malignant transformation, we studied proliferation and apoptosis in B cells from 4-to 6-months-old, tumor-free iMyc EA mice.FACS analysis of MACS-sorted B220 + splenocytes stained with propidium iodide showed that, on average, 4.73% Myc His B cells were in S phase of the cell cycle, a 3.5-fold increase compared with B cells from nontransgenic littermates (1.36%; Fig. 3A).To better quantify the apparent increase in Myc His -dependent proliferation, we labeled B cells with BrdUrd in vivo followed by FACS analysis.The Myc His -transgenic B cells (8.79% BrdUrd + ) proliferated, on average, 3.8 times faster than controls (2.34% BrdUrd + ; Fig. 3B).Increased proliferation of Myc His B cells was accompanied by elevated apoptosis in these cells (7.46 F 2.18%) compared with normal B cells (0.14 F 0.02%), using reactivity for activated caspase 3 as the FACS label (Fig. 3C).The increased proliferation of Myc His B cells in vivo noted above was also observed in cell culture upon examination of the LPS response using the MTS assay.MACSpurified B splenocytes from iMyc EA mice proliferated more vigorously than B cells from wild-type littermates after stimulation with three different doses of LPS (Fig. 3D).
We next evaluated whether Myc-transgenic B cells exhibit changes in surface marker expression compared with wild-type controls.FACS analysis using a panel of antibodies to B cell-associated differentiation and activation antigens showed that B cells from 4-to 6-months-old iMyc EA mice expressed heightened levels of costimulatory receptor CD40 (Fig. 4A, top) and activation markers CD38, CD86, and MHC class II (results not shown).In contrast, CD48 and CD95 (Fas) were down regulated (Fig. 4A, center and bottom).The down-regulation of Fas may be an adaptive change to counterbalance Myc-induced apoptosis (25), but this has not been shown here.
To assess whether the changes in the B-cell compartment of iMyc EA mice compromise immune responsiveness, we immunized the mice with the T cell-dependent antigen, CLQLKEQKNAGTRT-NEG-keyhole limpet hemocyanin.The response of transgenics and wild-type controls was indistinguishable, based on antigen-specific titers of IgM and the four IgG isotypes (Fig. 4B).Transgenic mice had also normal overall serum IgG and IgM levels compared with wild-type littermates, which served as controls (Fig. 4C).Stained sections of lymphoid tissues of 4-to 6-months-old mice revealed the long-term consequences of Myc His -driven B-cell proliferation at the morphologic level.Spleen (Fig. 4D) and lymph nodes (not shown) of transgenic mice were enlarged relative to nontransgenic controls because of progressive, follicular hyperplasia indicative of B-cell accumulation in vivo.This finding indicated that in B cells of tumor-free iMyc EA mice, the balance of Myc-dependent proliferation and apoptosis is tipped in favor of proliferation, leading to characteristic prelymphomatous changes in these mice.
Development of Mature B-Cell and Plasma-Cell Neoplasms in iMyc EA Mice.To examine the possibility that Myc His -dependent lymphomas might arise in iMyc EA mice, we monitored 109 mice for tumor development.The overall tumor incidence was 68% at 21 months of age (Fig. 5A).Histological examination of 76 tumorbearing mice revealed three major tumor types (Fig. 5B), which appeared with similar latency.The most prevalent tumors were lymphoblastic B-cell lymphomas with a Burkitt-like ''starry sky'' morphology due to tingible body macrophages (ref.26; Fig. 5C, top).Immunostaining showed that the tingible body macrophages engulfed TUNEL-positive cells (not shown).Lymphoblastic B-cell lymphomas were IgM + IgD + B220 + CD19 + CD5 À using flow cytometry and had a high proliferative index (>70%) after BrdUrd labeling in vivo (not shown).Nearly one quarter of the lymphomas were diagnosed as diffuse large B-cell lymphomas (26) strongly positive for Bcl-6 by immunohistochemistry (Fig. 5C, center).Another fifth of the lymphomas were CD19 + CD138 + plasmacytomas (Fig. 5C, bottom) that exhibited varying degrees of differentiation, including plasmablastic, anaplastic, and plasmacytic variants.Plasmacytoma produced monoclonal immunoglobulin of g, a (shown in Fig. 5C, bottom), and A isotypes.Although the morphologic characteristics of lymphoblastic B-cell lymphoma, diffuse large B-cell lymphoma, and plasmacytoma will be described in depth in a subsequent article, the findings presented here clearly indicated that Myc His induced a spectrum of mature B-cell and plasma-cell tumors.Because lymphoblastic B-cell lymphoma was the most common lymphoma iMyc EA mice, it was chosen for further study described below.
Lymphoblastic B-Cell Lymphomas Are Monoclonal Tumors and Express Myc His Predominantly from the P1 Promoter.Southern analysis of V(D)J recombination showed that lymphoblastic B-cell lymphomas had undergone clonotypic Ign rearrangements Samples with Myc His /actin ratios set at 1.0 were used as internal standards in the three independent experiments.Myc His expression is not strictly limited to B cells and is, therefore, leaky.This is shown in Supplementary Fig. 2, which shows transgene expression in splenic T cells, but not in splenic monocytes.B, elevated Myc mRNA expression in pro-B/ pre-B cells (B220 low IgM À ), transitional B cells (B220 low- IgM + ), and mature B cells (B220 high IgM + ) obtained from bone marrow of iMyc EA mice and nontransgenic littermates (Control ).Average levels and SDs based on three independent measurements.
(Fig. 6A, left).Rearrangements of Igh were also readily detected, but these were limited to the B6-derived locus (Fig. 6A, right).Western blotting with anti-His 6 antibody showed that lymphoblastic B-cell lymphoma invariably expressed Myc His , although the expression levels varied (Fig. 6B, left).Myc His protein was clearly elevated in lymphoblastic B-cell lymphoma compared with premalignant B splenocytes from iMyc EA mice and normal B splenocytes (negative controls) using immunocytochemistry (Figure 6B, right).Northern blotting showed that total Myc mRNA (Myc His plus Myc) was overexpressed compared with controls (Fig. 6C).Estimation of the abundance of Myc His and Myc mRNA by allele-specific RT-PCR showed that the bulk of the Myc message was encoded by the inserted transgene (Fig. 6D).A characteristic feature of Myc deregulation in human BL and mouse plasmacytoma is a shift in promoter usage from the second Myc promoter, P2, which initiates f80% to 90% of Myc mRNA in normal B cells, to the first promoter, P1, which initiates f10% to 20% of normal Myc mRNA.To evaluate whether mouse lymphoblastic B-cell lymphoma exhibited a shift in the primacy of promoter usage, RNase protection assays were done with primary lymphoblastic B-cell lymphoma.Two mouse plasmacytoma cell lines that displayed a pronounced promoter shift (TEPC 1165) or did not (ABPC 20) were included as positive and negative controls (21).The analysis (summarized in Fig. 7A and presented in detail in Supplementary Fig. 2) showed that lymphoblastic B-cell lymphoma exhibited the full P2 to P1 promoter shift in Myc expression.In contrast, B220 + B cells from iMyc EA mice did not show a promoter shift relative to B220 + B cells from normal B6 mice, BALB/c thymus, and ABPC 20.These findings suggested that the juxtaposition of Myc to EA is, on its own, not sufficient to cause the promoter shift.The shift seems to require additional changes associated with acquisition of the malignant phenotype of lymphoblastic B-cell lymphoma. .Splenic B cells were sorted using magnetic beads followed by labeling with antibody to activated caspase 3 and FACS analysis.The average FACS reactivity to activated caspase 3 was markedly increased in transgenic B cells (7.46%) compared with controls (0.14%) in three independent determinations.D, enhanced LPS response in iMyc EA B cells.MACS-purified B220 + splenocytes from transgenic mice and nontransgenic littermates (Control ) were cultured for 24 hours in the absence of LPS (background control ) and presence of 25, 50, or 100 Ag/mL LPS.Proliferation was measured using the MTS assay.Average results of three independent measurements.The clustering of the great majority of these mutations in a highly conserved region of MYC amino-terminal transactivation domain (MYC homology box 1, MbI) suggests that these mutations were selected during tumor development.The underlying reason for this selection and the mechanism of mutagenesis in MbI are not known, but some circumstantial evidence points to VDJ hypermutation in germinal center B cells as the underlying mutagenic principle (27).To evaluate whether mouse lymphoblastic B-cell lymphomas contain point mutations in Myc His , we sequenced the entire Myc His gene (2.2 kb for both DNA strands) in 34 tumors obtained from 16 lymphoblastic B-cell lymphoma-bearing mice.74.8 kb of DNA sequence analysis resulted in the detection of just one mutation, a tandem missense mutation, L56F/P57S, which was found in two tumors harvested from different anatomic sites in the same mouse (Fig. 7B).A recent tabulation of data from the literature (28) indicates that the P57S mutation is the second most common base substitution mutation in human BL, comprising 8 (14.5%) of 55 mutations in MbI.Interestingly, biological studies with the P57S allele have showed that this mutation results in lower oncogenic activity than wild-type MYC when assayed for neoplastic transformation of rat embryo cells in cooperation with the H-rasG12V oncogene (28).The significance of the L56F mutation, which has not been observed thus far in human BL, is not known.It may be an innocent ''bystander'' mutation in the present case of mouse lymphoblastic B-cell lymphoma.
Interruption of the Arf-Mdm-p53 Tumor Suppressor Axis in Lymphoblastic B-Cell Lymphoma.Myc-induced B-cell transformation and lymphomagenesis in EA-Myc transgenic mice ( 29) is limited by Myc-dependent apoptosis (30).One important mechanism of curtailing Myc-dependent apoptosis in EA-Myc mice is inactivation of the p19 Arf -Mdm2-p53 tumor suppressor pathway (31,32).When activated, this pathway involves the up-regulation of p19 Arf , which results, in turn, in sequestration of Mdm2 and activation of the p53 checkpoint (33).To determine whether interruption of this pathway also plays a role in lymphoma development in the iMyc EA mice, we studied 26 lymphoblastic B-cell lymphomas using immunoblotting.Applying stringent assay conditions aimed at detecting pronounced alterations, we found changes in 12 (46.2%) of 26 tumors.Four (33%) of the 12 tumors with alterations overexpressed p19 Arf (Fig. 7C, LBL 1-4).Loss of p19 Arf protein expression, apparently due to biallelic deletions in the Ink4a locus, occurred in 6 (50%) of 12 tumors, as detected by Western blot and genomic PCR (LBL 5-10).Overexpression of Mdm2 was found in 2 tumors (17%, .Elevation of (presumably dysfunctional) p53 was seen in 3 tumors (25%; LBL 2, 4, and 9, indicated by red squares), however, only in combination with overexpression of p19 Arf (LBL 2 and 4) or loss of p19 Arf (LBL 9).Irrespective of the underlying reasons, which may be complex, all of the above alterations are likely to result in the crippling of the p53 response (cell cycle arrest, apoptosis).Considering that loss of p53 (32,34,35) or p19 Arf (31,32) accelerates EA-Myc-induced lymphoma development, our data suggest that the interruption of the p19 Arf -Mdm2-p53 pathway also plays a significant role in lymphomagenesis in iMyc EA mice (31).
Lymphoblastic B-Cell Lymphomas Contain Clonal Cytogenetic Alterations.Because Myc His -dependent genomic instability (36) may be one mechanism by which the p53 pathway is interrupted during tumor progression, we used spectral karyotyping to analyze three randomly chosen lymphoblastic B-cell lymphomas.One of them exhibited a translocation/deletion at chromosome 11B (Fig. 8, top) and another one showed a large deletion in chromosome 13 (Fig. 8, center).The third tumor had undergone tetraploidization and contained an extra copy of chromosome 18 (Fig. 8, bottom).Thus, three of three lymphoblastic B-cell lymphomas contained cytogenetic aberrations that were readily detectable by spectral karyotyping.Because these markers or extra chromosomes were found in 10 metaphase plates of each tumor, this provides additional evidence for the clonality of these tumors.

Discussion
Gene insertion was used to recreate as a germ line mutation in mice the MYC/Myc-activating Igh translocations seen in human endemic Burkitt lymphomas and certain mouse plasmacytomas.The newly developed mouse model, designated iMyc EA , has several noteworthy features.First is the precise reconstruction of the translocation breakpoint region on human der( 14) and mouse der (12), that is, the head-to-head juxtaposition of Myc His and EA.Second is the potential for Myc His to interact with the complete set of correctly spaced Igh enhancers, designed to create the same complex interplay of promoter and enhancer interactions that govern the expression of translocated MYC/Myc in human endemic Burkitt lymphoma/mouse plasmacytoma.Third is the insertion of Myc His in the Igh chromatin domain, which presumably subjects the gene to the same higher-order regulatory influences (e.g., those imposed by chromatin remodeling and positional effects in the interphase nucleus) that affect MYC/Myc involved in the t(8;14)/T(12;15) exchange.Our finding that many of the heterozygous iMyc EA mice spontaneously developed lymphoblastic B-cell lymphoma, diffuse large B-cell lymphomas, and plasmacytomas suggests that the iMyc EA transgene reproduced the key features of the t(8;14)/T(12;15) translocation that initiate the development of endemic Burkitt lymphoma in humans and plasmacytoma in mice.
Our Myc His transgene model of the endemic Burkitt lymphoma t(8;14)/plasmacytoma T(12;15) effectively induces B-cell neoplasia even though it differs from the ''natural'' chromosomal translocations found in the human and mouse tumors in at least three potentially important aspects.First, the natural translocations are spontaneously occurring somatic mutations in tumor precursors, whereas Myc His is a germ line mutation present in all cells.Second, Myc His insertion reproduced only the MYC/Mycactivating product of translocation, not the reciprocal product of translocation.Third, in endemic Burkitt lymphoma t(8;14), the breakpoints on chromosome 8 are scattered over a considerable genomic distance in the 5V flank of MYC (37,38), resulting in separation of MYC from EA by more than 340 kb (37) in some cases (class III translocations according to Cory 39).As shown here, these aspects of endemic Burkitt lymphoma t(8;14)/plasmacytoma T (12;15) are not critical to the development of B-cell and plasmacell neoplasms following MYC/Myc-activating translocation.It follows that the juxtaposition of Myc and EA in the appropriate chromosomal context of the Igh chromatin domain defines the minimal, critical feature of the endemic Burkitt lymphoma t(8;14)/ plasmacytoma T(12;15) translocation.The long latency of B-cell and plasma-cell tumors developing in iMyc EA mice suggests that secondary oncogenic events complement Myc His during lymphomagenesis.Alternatively, the oncogenic potential of Myc His may be offset in vivo by factors that inhibit Myc-induced transformation (e.g., Myc -induced apoptosis).The apparent occurrence of t(8;14)(q24;q32) in healthy blood donors (40), the repeated detection of T (12;15) in normal mice (41)(42)(43), and recent insights from mouse models of nonhematopoietic cancer driven by inducible Myc transgenes (44,45) support this view.
The occurrence of plasmacytoma in f20% of the tumor-bearing iMyc EA mice suggests that B cells bearing our model of Mycactivating translocation sometimes acquire the same subsequent genetic alterations that lead to the development of plasmacytoma following somatic acquisition of T(12;15).Plasmacytoma have not been observed in the EA-MYC/Myc and YAC-MYC mice.What mechanisms might be responsible for the unique property of the iMyc EA transgene to induce plasma cell tumors?One feature that distinguishes the iMyc EA mice from the EA-MYC/Myc and YAC-MYC mice is the genomic integration site of the Myc transgene.Whereas this site is ectopic in the EA-Myc and YAC-MYC, the insertion site of the Myc His transgene in iMyc EA corresponds precisely to the t(8;14)/T(12;15) translocation break site in endemic Burkitt lymphoma/T(12;15).The strategic location of Myc His in Igh may link timing and level of Myc activation to the developmentally controlled accessibility (46) and activity (47) of the Igh chromatin domain (48), generating thereby a temporal and quantitative pattern of Myc His activation that is conducive to plasma cell tumors.A second distinguishing feature of the iMyc EA is the potential for Myc His to interact with Ea and Ey in addition to EA.The crucial importance of the Myc-Ea interaction for Myc deregulation in plasma cell tumors has been clearly shown (12).The in vivo relevance of the Myc-Ey interaction has been recently noted by the surprising observation that EA is dispensable for MYC expression and lymphomagenesis in the YAC-MYC mouse (11), in which EA can apparently be substituted for by the Ey enhancer (49).Thus, the ability of the iMyc EA to promote plasma cell transformations may be created by the complex interplay of Myc His with all three Igh enhancers taking place in the appropriate genomic context of the Igh domain.
The iMyc EA mice were generated as part of a program to recreate and study in transgenic mice the molecular and oncogenic consequences of MYC/Myc-activating t(8;14)/T(12;15) translocations that characterize human and mouse B-cell and plasma-cell tumors, such as human endemic Burkitt lymphoma and mouse plasmacytoma.Among the new opportunities afforded by the iMyc EA is the identification of the secondary oncogenic events that complement Myc during neoplastic transformation, leading to lymphoblastic B-cell lymphoma in most cases, but diffuse large B-cell lymphomas and plasmacytoma in the others.Furthermore, although strain iMyc EA was not developed to model human endemic Burkitt lymphoma or any other specific type of human NHL in mice, it may provide guidance for possible future attempts to model human BL in mice more accurately.Human BL exhibits the phenotype of a germinal center (GC) B cell and contains somatic mutations in expressed immunoglobulin variable and MYC genes, indicating that the tumors have germinal center experience (50).Human BL acquire MYC translocations as spontaneous somatic mutations that are thought to be accidents of VDJ hypermutation (endemic Burkitt lymphoma) and isotype switching (sBL/iBL) in the germinal center (51).Human BL contains clonal EBV (52) and expresses the EBV-encoded EBNA-1 protein (53), which is consistent with malignant transformation occurring in an EBV-infected centroblast differentiating into a memory B cell (54).These features of human BL suggest that in order to model this neoplasm in mice more precisely, Myc activation should be delayed to the germinal center stage, for example, by placing Myc under control of a germinal center-specific promoter.Myc transgenes of this sort should then be combined with EBNA-1 transgenes (55) that could perhaps also be improved by targeting expression to germinal center B cells.Although these approaches may help to settle the long-standing thorny question of whether EBV plays an active role in the development of BL or is just a passive passenger in the tumor cells (56), the present study has shown that the iMyc EA transgene is a valuable model of endemic Burkitt lymphoma t(8;14) that predictably induces B-cell and plasma-cell tumors in mice.

Figure 1 .
Figure 1.Targeted transgenic insertion of Myc His in iMyc EA mice.Shown is a scheme of the normal mouse Igh locus (top ) and the targeted Igh locus (bottom ) with the inserted Myc His in iMyc EA mice (red ).The EA and Ea enhancers are depicted by blue diamonds .The transcriptional orientations of Igh and Myc are indicated by black and red arrows , respectively.The gene insertion site is also the preferred recombination site of human t(8;14)(q24;q32) IGH-MYC exchanges in human endemic Burkitt lymphoma and mouse T(12;15) Igh-Myc exchanges in certain mouse plasmacytomas.

Figure 2 .
Figure2.Overexpression of Myc His throughout B-cell development in iMyc EA mice.A, expression of Myc His at different stages of B-cell development.Samples with Myc His /actin ratios set at 1.0 were used as internal standards in the three independent experiments.Myc His expression is not strictly limited to B cells and is, therefore, leaky.This is shown in Supplementary Fig.2, which shows transgene expression in splenic T cells, but not in splenic monocytes.B, elevated Myc mRNA expression in pro-B/ pre-B cells (B220 low IgM À ), transitional B cells (B220 low- IgM + ), and mature B cells (B220 high IgM + ) obtained from bone marrow of iMyc EA mice and nontransgenic littermates (Control ).Average levels and SDs based on three independent measurements.

Figure 3 .
Figure 3. Increased turnover of premalignant B cells in tumor-free iMyc EA mice.A, FACS analysis indicates increased cell cycle progression in Myc His -transgenic B cells (bottom ) compared with normal B cells (top ).Freshly isolated B220 + splenocytes were stained with propidium iodide (PI ) and cell cycle profiles determined by FACS.Means and SDs (%) of cells in three different cell cycle stages.The results are based on five independent determinations.B, increased proliferation in Myc His -transgenic B cells (right ) compared with normal B cells (left ).B220 + splenocytes were labeled with antibody to BrdUrd and analyzed by FACS to determine the fraction of dividing cells that had incorporated BrdUrd in vivo (B ).Means and SDs (%) of BrdUrd + cells are shown.The results are based on three independent determinations.C, elevated apoptosis in Myc His -transgenic B cells (bottom ) relative to normal B cells (top ).Splenic B cells were sorted using magnetic beads followed by labeling with antibody to activated caspase 3 and FACS analysis.The average FACS reactivity to activated caspase 3 was markedly increased in transgenic B cells (7.46%) compared with controls (0.14%) in three independent determinations.D, enhanced LPS response in iMyc EA B cells.MACS-purified B220 + splenocytes from transgenic mice and nontransgenic littermates (Control ) were cultured for 24 hours in the absence of LPS (background control ) and presence of 25, 50, or 100 Ag/mL LPS.Proliferation was measured using the MTS assay.Average results of three independent measurements.

Figure 4 .
Figure 4. Lymphoid hyperplasia and normal immune response in tumor-free iMyc EA mice.A, up-regulation of CD40 and down-regulation of CD48 and CD95(Fas) on the cell surface of premalignant B220 + splenocytes obtained from iMyc EA transgenic mice (pink ) and nontransgenic littermates (green ).Shown are representative FACS histograms with isotype-specific antibody controls depicted in purple .B and C, normal immune response of iMyc EA mice after immunization with a thymus-dependent peptide antigen.Antigen-specific (F ) and overall IgG and IgM titers (G ) were determined by ELISA in transgenic mice and wild-type littermate controls.Means and SDs based on four mice in each group.D, lymphoid hyperplasia in the spleen of a 6-month-old tumor-free iMyc EA mouse.The enlarged white pulp (blue ) occupies more than 90% of the tissue section stained with H&E.

Figure 5 .
Figure 5. B-cell and plasma-cell tumors in iMyc EA mice.A, tumor incidence and survival of mice.B, graphical representation of the observed tumor spectrum: LBL, lymphoblastic B-cell lymphoma; DLBCL, diffuse large B-cell lymphoma; PCT, plasmacytoma.Tumor classification was difficult in 5% of cases, designated as Other .C, representative histologic tissue sections of the three main tumor types observed in the iMyc EA mice, either stained with H&E (top ) or immunostained with antibody to Bcl-6 (middle ) or IgA (bottom ).

Figure 6 .
Figure 6.Lymphoblastic B-cell lymphomas are monoclonal tumors overexpressing Myc His .A, clonotypic VDJ rearrangements.Shown are Southern blots that contain DNA from six tumors after Eco RI restriction digestion and hybridization to an Ign (left ) or J H probe (right ).Lymphoblastic B-cell lymphoma arose in mice that harbored one normal C57BL/6 (B6)-derived Igh allele and one mutant 129SvJ(129)-derived Igh allele with the Myc His .The B6 allele is indicated by one fragment (6.3 kb), whereas the 129 allele is indicated by two fragments (5.3 and 5.1 kb) that are poorly separated on the gel.VDJ rearrangements occurred only on the B6 allele, indicated by a reduction of the fragment relative to the germ line fragment in iMyc EA liver (Tg liver ) or liver from normal B6 mice.B, lymphoblastic B-cell lymphomas express Myc His protein using Western blotting (left) and immunocytochemistry (right ).Left, two immunoblots that contain lysates from 13 randomly chosen, primary lymphoblastic B-cell lymphomas.Myc His was detected with an anti-His antibody.Anti-actin was used as loading control.Right, cytofuge specimens of lymphoblastic B-cell lymphoma cells (bottom ) next to FACS-purified premalignant B cells from iMyc EA mice (middle ), and normal splenic B cells from C57BL/6 mice (control, top ) after immunostaining with an antibody to the histidine tag.C, lymphoblastic B-cell lymphomas overexpress Myc mRNA.Shown is a Northern blot that contains RNA samples from six primary lymphoblastic B-cell lymphoma.RNA samples obtained from two lymph nodes and the BALB/c plasmacytoma line, TEPC 1165, were included as negative and positive controls, respectively.The ethidium bromide-stained agarose gel showing the 28S rRNA band (bottom ) is included as loading control.D, lymphoblastic B-cell lymphomas express Myc mRNA mainly from the inserted Myc His gene.Shown are ethidium bromide-stained agarose gels with RT-PCR products that reflect the abundance of Myc mRNA (top ), Myc His mRNA (middle ), and Gapd mRNA (bottom ) in 16 primary lymphoblastic B-cell lymphomas and B220 + C57BL/6 splenocytes (control sample).White asterisks in the top, presence of faint PCR indicator fragment.Size markers are in the left-hand lane .

Figure 7 .
Figure 7. Lymphoblastic B-cell lymphomas exhibit the P2 to P1 promoter shift of Myc expression and distortions of the p19 Arf -Mdm2-p53 tumor suppressor axis.A, lymphoblastic B-cell lymphomas exhibit the P2 to P1 promoter shift of Myc expression.Summary of the experimental results presented in Supplementary Fig. 2. Mean values of normalized P1/P2 ratios and SDs of the mean are plotted.The normalization was based on the P1/P2 ratio in TEPC 1165, which was arbitrarily set at 1.The P1/P2 ratio was significantly elevated in lymphoblastic B-cell lymphoma relative to splenic B cells from iMyc mice (two-sided Student's t test; P = 0.0007).B, tandem point mutation in Myc homology box I, a primary lymphoblastic B-cell lymphoma.Top, scheme of the Myc His protein that contains the two Myc homology boxes MbI and MbII, the basic region (B), the helix-loop-helix domain (HLH ), the leucine zipper (LZ ), and the histidine tag (His ).The MbI domain spans residues 47 to 62. Bottom, histograms of DNA sequencing reactions of MbI nucleotides encoding residues 53 to 59 of normal Myc and mutated Myc found in two independent specimens in 1 of 16 primary lymphoblastic B-cell lymphomas.The mutated Myc gene harbored point mutations in two neighboring codons, resulting in L56F and P57S substitutions.C, p19 ARF , Mdm2, and p53 expression and biallelic Ink4a deletions in primary lymphoblastic B-cell lymphomas.Protein levels of p19 ARF , Mdm2 and p53 were compared by immunoblotting in 12 lymphomas (LBL 1-12 ) and B splenocytes from nontransgenic littermates (Normal ), using h-actin as loading control.Two of three major isoforms of Mdm2 are discernable in some lanes (e.g., LBL 3 and 4).Potential deletions of the genes encoding p19 ARF and p53 were assessed by genomic PCR for exon 1 of Ink4a and exon 2 of Trp53 , respectively.Apparent biallelic deletions of Ink4a were found in LBL 5-10.Deletions of Trp53 were not found.

Figure 8 .
Figure 8. Chromosomal aberrations in lymphoblastic B-cell lymphoma detected by spectral karyotyping.Shown are a nonreciprocal T(11;3) translocation and a large internal deletion of chromosome 13B found in two primary tumors.Aberrant chromosomes are depicted in spectral karyotyping display colors (middle ) and classification colors (right ) next to inverted 4V,6-diamidino-2-phenylindole images (left ).Normal chromosomes (Chr.)11 and 13 are included for comparison.The T(11;3) was highly variable in different metaphases, presumably reflecting ongoing genomic instability of the chromosome.Three variants of T(11;3) from different metaphase spreads of LBL 1 are depicted to the right of the complete spectral karyotyping karyotype of this tumor.An image of Del 13B found in LBL 2 is also shown.