| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,
* Laboratory of Molecular Biology, Clinical Research Institute of Montreal, Montréal, Québec, Canada;
Department of Medicine, Université de Montréal, Montréal, Québec, Canada;
Department of Microbiology and Immunology, Université de Montréal, Montréal, Québec, Canada; and
Department of Experimental Medicine, McGill University, Montréal, Québec, Canada
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
CT). We expressed these mutants in Tg mice using the CD4 promoter. Both Notch1IC and Notch1CT, but not Notch1EC, Tg mice developed double-positive (DP) thymomas. These disseminated more frequently in Notch1CT Tg mice. Double (Notch1IC x myc) or (Notch1CT x myc) Tg mice developed thymoma with a much shorter latency than single Tg mice, providing genetic evidence of a collaboration between these two oncogenes. FACS analysis of preleukemic thymocytes did not reveal major T cell differentiation anomalies, except for a higher number of DP cells and an accumulation of TCRhighCD2highCD25high DP cells in Notch1IC, and less so in Notch1CT Tg mice. This was associated with enhanced in vivo thymocyte proliferation. However, Notch1IC, but not Notch1CT, DP thymocytes were protected against apoptosis induced in vivo by dexamethasone and anti-CD3 and in vitro by anti-CD3/CD28 Abs. This indicates that the C terminus of Notch1 and/or the conserved regulation by its ligands have a significant impact on the induced T cell phenotype. Therefore, Notch1IC and Notch1CT behave as oncogenes for T cells. Because these two Notch1 mutations are very similar to those described in some forms of human T cell leukemia, these Tg mice may represent relevant models of these human leukemias. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
-associated molecule (RAM) domain, the six cdc10/SW16/ankyrin motifs, a transcriptional-activated domain (TAD), and polyglutamine stretch (OPA), proline-glutamate-serine-threonine (PEST) sequences at its C terminus, is responsible for the signal transduction to the nucleus. Truncated forms of Notch1 representing various forms of the IC domain (Notch1IC) have been found to function as gain-of-function mutants in several biological systems.
However, studies on the effects of overexpression of Notch1IC in T cells have been controversial. Overexpression of a constitutively activated form of Notch1IC (deleted of its TAD and C-terminal PEST sequences) in transgenic (Tg) mice under the control of lck proximal promoter was initially found to promote the maturation of CD8 single-positive (SP) thymocytes (3) and to favor the TCR
cell fate (4). In contrast, expression of another Notch1IC mutant (deleted of only its PEST sequences) was reported to favor the development of both CD4+ and CD8+ SP thymocytes in Tg mice (5). Reassessment of these two Tg strains in the same laboratory led to the conclusion that in fact both strains promote CD8+ and inhibit CD4+ SP development (6). To add to the controversy, two additional studies of Notch1IC mutants expressed in T cells of lethally irradiated mice transplanted with retrovirus-transduced bone marrow cells reported phenotypes distinct of each other and distinct from those observed in the Tg mice described above (7, 8). In one study, Notch1IC was found to prevent development of both CD4+ and CD8+ SP thymocytes (7), while in the other, it had very little impact on thymocyte development (8).
In addition to their effect on T cell development, gain-of-function mutants of Notch1IC have also been implicated in the development of T-lymphoid malignancies (9, 10, 11, 12). TAN-1, the human homologue of Notch, was first identified as a locus involved in the t(7;9)(q34;q34.3) chromosomal translocation present in a subset of T cell acute lymphoblastic leukemias (T-ALL) (10). In these human leukemic cells, the translocation led to the aberrant expression of truncated Notch1 IC transcripts. Our group also found that Notch1 was mutated in a high proportion (52%) of thymic T cell lymphomas arising in Moloney murine leukemia virus-infected mouse mammary tumor virus (MMTVD) /c-myc Tg mice (12, 13). In these tumors, the inserted proviruses were integrated in two regions of the Notch1 gene, inducing two distinct truncated mutations: type I and II. Type I integration occurred around genomic regions coding for the 34th epidermal growth factor repeat and the TM of Notch1. Typically, type 1 tumors produced elevated levels of 3- to 4-kb truncated RNA transcripts, initiating at the integration site and terminating at the 3' end of the gene, and thus encoding truncated Notch1 IC proteins having conserved all IC motifs. These Notch1IC mutant proteins contain a small portion of extracellular domain, the TM domain, and the complete cytoplasmic domain (12, 13, 14). In addition, in tumor cells harboring type I mutants, truncated 5' end RNA coding for N-terminal truncated Notch1 ectodomain (Notch1EC) was overexpressed. In type II mutants, integration of proviruses occurred within the last coding exon of the gene, thus generating a Notch1 protein deleted of its C-terminal OPA and/or PEST domains (Notch1CT). Notch1CT protein appears to be cleaved physiologically, giving rise to an intact ectodomain and a truncated Notch1 IC domain (13). Interestingly, a recent study (15) of human T-ALL cells has shown that Notch1 mutations occurred at the same two type I and type II locations as the ones generated by provirus insertions, and thus most likely generate truncated molecules very similar to those observed in murine cells harboring type I and II mutants.
Together, these results strongly suggested that truncated Notch1IC might be involved in T cell transformation and may cooperate with c-myc in T cell oncogenesis. In fact, previous attempts to document the oncogenic potential of some truncated Notch1IC experimentally were positive. Transduction of donor bone marrow cells expressing truncated Notch1IC mutants led to the development of double-positive (DP) T cell malignancies in recipient mice (16). In this assay, the ankyrin repeat and C-terminal trans activation domains were required for T cell leukemogenesis (17). Similarly, expression of two distinct Notch1IC mutants under the regulation of the lck proximal promoter in Tg mice was also found to induce DP T cell leukemia (3, 5, 18).
Our finding of frequent truncated Notch1 mutations in myc Tg mice led us to assess the oncogenic potential of these Notch1 mutants to establish a relevant animal model of the human T-ALL harboring Notch1 mutations. We chose to express the very same type I and type II Notch1 mutants initially found in murine T cell tumors in Tg mice, because they were the only ones generated and selected for during the oncogenic process. We report in this study a comparative analysis of the oncogenicity of these two mutants expressed under the same regulatory sequences of the human CD4 gene and on their different impact on T cell development.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
The 14.9-kbp CD4C chimeric regulatory sequences containing the 1.9-kbp enhancer of the mouse CD4 gene and a 13-kbp fragment from the human CD4 gene have previously been described (19). The 3.0-kbp truncated Notch1 IC fragment was cleaved from mouse Notch1 cDNA (provided by H. Weintraub, the Fred Hutchinson Cancer Research Center, Seattle, WA) (20) at the NcoI site (nt 5051) and at the EcoRI site (nt 8064) at the end of C terminus and inserted downstream of the CD4C sequences to generate CD4C/Notch1IC. Similarly, the near-to-full-length 6.9-kbp Notch1CT DNA, deleted at the HindIII site from its C terminus, was inserted downstream of the CD4C regulatory sequences to generate CD4C/Notch1CT. The RI-H3 4.4-kbp Notch1 ectodomain (Notch1EC) DNA fragment and the 1.1-kbp HindIII/Eco/RI C-terminal PEST-containing DNA fragment were inserted downstream of the CD4C regulatory sequences to generate CD4C/Notch1EC and CD4C/NotchCT constructs, respectively. The SV40 polyadenylation sequences (0.88 kbp from plasmid NL4/SV40) (21) were ligated at the 3' end of each transgene DNA. The 18.8-kbp CD4C/Notch1IC, 23-kbp CD4C/Notch1CT, 20-kbp CD4C/Notch1EC, and 17-kbp CD4C/Notch1CT DNAs used for microinjection were cleaved with EcoRI, then isolated by preparative agarose gel electrophoresis and further purified on CsCl gradients, as described previously (22, 23).
One-cell (C57BL/6 x C3H)F2 embryos were collected, microinjected, and transferred into pseudopregnant CD1 females essentially as described previously (22, 23). The presence of the transgene was confirmed by Southern blot hybridization on tail DNA, using Notch1 cDNA as a probe. Six CD4C/Notch1IC Tg founders (F30856, F35748, F35735, F35753, and two other unnumbered founders), three CD4C/Notch1CT (F92743, F92744, F92745), four CD4C/Notch1EC (F60787, 60788, 98513, 98514), and two CD4C/Notch1CT (F87827, F87828) were produced from the pups born. All of the Tg founders were bred on the CD1 background as heterozygous for the transgene for at least 8 generations and for 10–15 generations for most experiments presented. The phenotypes described in this work have been observed over a period of 
4 years without noticeable changes. All founders transmitted the transgene in a Mendelian fashion and appeared phenotypically normal. The Tg mice were observed for spontaneous thymic lymphoma development, and the thymic lymphomas were collected for analysis.
Analysis of transgene RNA expression
Total RNA was extracted from thymus or thymic lymphoma samples with 1 ml of TRIzol Reagent (Invitrogen Life Technologies), according to the manufacturer’s protocol. To study transgene expression, RNA was separated on 1% formaldehyde-agarose gels, transferred to Hybond-N membranes (Amersham), and hybridized with Notch1 probe K for CD4C/Notch1IC Tg mice, as previously described (12). The Notch1 probe K is a 2.7-kbp BamHI-EcoRI cDNA fragment covering the entire Notch1 IC domain. Probe M was used to detect Notch1EC RNA and corresponds to a 1.9-kbp BamHI extracellular fragment of the MN7 full-length Notch1 cDNA, as previously described (12). These probes were prepared as previously described (13). For quantitation, the membrane was exposed on PhosphorImager screen and scanned with the Storm Imaging unit (Amersham Biosciences). Semiquantitation of the RNA bands was estimated with the ImageQuant software (Amersham Biosciences). A ratio of the Notch1-specific signal relative to the actin-specific signal was obtained.
For analysis of c-myc Tg RNA expression, RT-PCR was performed on the RNA of thymic lymphomas arising in Notch1IC/c-myc double-Tg mice. Total RNA (5 µg) was used for cDNA synthesis with reverse transcriptase (200 U, Moloney murine leukemia virus; BRL, 200 U/µl). The cDNA obtained was used as template for PCR using primers specific for the MMTVD long terminal repeat (LTR) (sense, 5'-GCAACAGTCCTAACATTCACCT-3') and the c-myc exon 3 (antisense, 3'-CGGAATGGAGATGAGCCCGAC-5'). PCR were performed in 2.5 mM MgCl2, 1 mM dNTP, 1 U of TaqDNA polymerase, and 0.5 pM of each primer. DNA was amplified for 35 cycles at an annealing temperature of 60°C. PCR products were separated on 1% agarose gels and stained with ethidium bromide.
Protein extraction and Western blotting
Total proteins were extracted in radioimmunoprecipitation assay buffer (12 mM Na2HPO4, 4 mM NaH2PO4, 1% Nonidet P-40, 1% sodium desoxycholate, 0.1% SDS, 150 mM NaCl, 2 mM EDTA, 10.5 mM EGTA) (24) containing 2 µg/ml aprotinin, 2 µg/ml leupeptin, 1 µg/ml pepstatin, 50 µg/ml N-p-tosyl-L-lysine chloro-methylketone, and 100 µg/ml PMSF. Extracts were cleared by centrifugation at 150,000 x g for 30 min. Protein extracts were mixed with sample buffer (62.5 mM Tris-HCl (pH 6.8), 20% glycerol, 2% SDS, and 5% 2-ME), boiled for 5 min, subjected to 6% SDS-PAGE, and transferred to nylon membranes. Protein molecular mass standards myosin (205 kDa) was purchased from Bio-Rad. Filters were blocked with 5% milk powder in TBST (10 mM Tris (pH 7.5), 150 mM NaCl, 0.5% Tween 20) overnight at 4°C and then probed with primary antiserum in 0.5% milk powder in TBST. Immunodetection was performed essentially as described previously, using rabbit anti-Notch1-intra1 or extra1 Abs, and secondary HRP-conjugated anti-rabbit antiserum (Sigma-Aldrich A0545), followed by chemiluminescent detection (NEN). The anti-Notch1-intra1 or extra1 Abs are specific, respectively, for the Notch1 IC or extracellular domain, and their generation and characterization have been described previously (13). The resulting autoradiograms were scanned with HP Deskscan II and reproduced for publication using PowerPoint (Microsoft) software. For quantitation, proteins were visualized by incubating the membranes with secondary Abs coupled to Alexa 680 fluorochrome, followed by scanning with Odyssey infrared imaging system (Licor). In each experiment, an image of the Notch and actin proteins was scanned on the Odyssey system and quantitated with ImageQuant software (Amersham Biosciences). A ratio of the Notch band over the actin band was first established. The ratio of the lower expressor founder was then set to 1 in each experiment, while the ratio of the higher expressor founder was corrected relative to 1 for each independent experiment.
Thymocyte apoptosis in vivo and in vitro
The CD4C/Notch1IC and CD4C/Notch1CT Tg mice and their littermates used in this experiment were 6–12 wk old. To induce thymocyte apoptosis in vivo, dexamethasone (0.3 mg) or purified anti-CD3 (clone 145-2C11) (50 µg) was injected i.p. to Tg and non-Tg littermates, while control treatment, respectively, RPMI 1640, or purified armenian hamster IgG1 of the same isotype as anti-CD3 was injected to the control group. The mice were sacrificed after 48 or 4 h. For in vitro apoptosis, thymocytes (2 x 106 cells) were incubated without or with dexamethasone (1 µg/ml) or on anti-CD3 (25 µg/well)-coated 24-well tissue culture plates supplemented with anti-CD28 (5 µg/ml) in the medium. After 3 h, cells were harvested. The levels of in vivo or in vitro apoptosis were evaluated by labeling cells with 7-aminoactinomycin D (7-AAD) and/or annexin V/propidium iodide (PI) and anti-CD4 and anti-CD8 mAb and analysis by FACS. The number of viable cells in control mice determined the 100% viability. The relative viability in Tg or non-Tg mice was calculated by number of viable cells in treated mice compared with the number of viable cells in control mice.
Transplantation of tumor cells into nude mice
Thymic lymphomas from CD4C/Notch1IC or CD4C/Notch1CT Tg mice were inoculated into nude mice. The tumors were cut into small pieces and passed through 18G needle (D = 0.3 mm) in 1x PBS buffer. This cell suspension (2 x 107 cells in 0.2 ml) was injected s.c. to 40- to 50-day-old CD1 nude mice. The tumor cells were injected on the right side and PBS buffer on the left side as a control.
Flow cytometry (FACS) analysis
Thymus, spleen, and lymph nodes (LN) from young Tg mice and their littermates, or thymic lymphomas from Tg mice were dissociated and passed through a 70-µm nylon mesh filter to obtain single-cell suspensions. To remove RBC from spleen cells, cells were subjected to lysis buffer (NH4Cl, KHCO3, EDTA (pH 7.2)) for 5 min on ice.
Abs conjugated to either FITC, PE, allophycocyanin, PE-Cy5, or biotin were obtained from BD Pharmingen or Cedarlane Laboratories. Biotinylated Abs were revealed using streptavidin-TRICOLOR (Caltag Laboratories) or FITC. Thymic lymphoma cells, splenocytes, and LN cells at 1 x 106 cells/sample were stained for surface expression of Thy-1.2, CD2, CD3, CD5, CD44, CD25, TCR, TCR
, CD4, CD8, CD24 (heat shock Ag; HSA), B220, Mac-1, CD11c, F4/80, CD69, and anti-NK. Cells were stained in FACS buffer containing Abs and analyzed on a FACSCalibur apparatus (BD Biosciences). Live cells were gated according to their forward light scatter (FSC) and side light scatter profiles after staining with PI. Data were analyzed using CellQuest software.
Gene rearrangement analysis
DNA was extracted from thymic lymphomas. The DNA from the kidney of the same mouse was used as control. For studying the TCR gene and Ig gene rearrangement, DNAs were digested with EcoRI and HindIII, respectively, and Southern blot analysis was performed with 32P-labeled probes, as described previously (25). The probe used for analyzing TCR gene was a 700-bp RBL-5 DNA fragment containing most of the murine C region and 3' untranslated sequences of C1. The JH region of Ig gene was analyzed with a 6.2-kbp EcoRI germline JH DNA probe.
Histological study
Tissues used for histology were fixed in 3.7% formaldehyde, sectioned, and stained with H&E, as described previously (26).
Detection of cell proliferation in vivo
This procedure has been described previously (27). Briefly, it involves the detection by FACS analysis of BrdU incorporated into cellular DNA of animals previously fed with BrdU in their drinking water (0.8 mg/ml) for 24 h or previously inoculated with BrdU (100 µg/g body weight) 4 h before sacrifice. Thymuses were collected and cell surface staining was first performed with anti-CD4 and anti-CD8 mAb, followed by an IC staining with FITC-labeled anti-BrdU Abs for detection of DNA synthesis and with 7-AAD staining for measuring total DNA content (cell cycle).
Fetal liver transplantation
Donors. Fetal livers from non-Tg and Tg 14.5-day-old embryos (E14.5) were harvested. Single-cell suspension was made in HBSS supplemented with 10% FBS, under sterile conditions, with a syringe plunger. Cell suspension was filtered through a Nytex mesh (BSH Thompson). The remaining fetal tissue was typed for Tg expression by PCR. When typing was known, cells coming from Tg or non-Tg embryos were pooled. After one wash, cells were counted (RBC cell lysis was performed in the counting aliquot). Cells were then resuspended in HBSS solution supplemented with 2% FBS at a concentration of 20 x 106 cells/ml.
Hosts. CD1 hosts (8–12 wk old) were lethally irradiated (950 rad) using Mark I-68A1 Irradiator (Cs-137; J.L. Shepherds & Associates). Hosts were injected, via the tail vein, with 4 x 106 fetal liver cells (FLCs) in 0.2 ml of HBSS solution supplemented with 2% FBS 4–6 h after irradiation.
PCR technique for transgene detection
A piece of fetal body was placed in 200 µl of lysis buffer (0.1 M NaCl, 0.01 M EDTA, 0.05 M Tris (pH 7.5), 0.5% Nonidet P-40, and 0.05% of Tween 20) supplemented with 10 µl of proteinase K (10 mg/ml) and digested at 55°C for 20 min to 2 h. After digestion, lysates were centrifuged to pellet nondigested tissue, and 100 µl of supernatant was taken. This aliquot was heated at 100°C for 8 min, further diluted (10 µl in 200 µl of water), and further heated at 55°C for 30 min. PCR was performed with Tg-specific primers (sense primer, CTGCTCCTACTCATTCCTTCC; antisense primer, CTCTGGAAGCACTGCGAGG; product size, 300 bp). Detection of the mouse Myb gene using sense primer, CCAGTCACGTTCCCTATCCT, and an antisense primer, GCCTGCTGTCCCTTCAGCTC, was done as a control (product size, 525 bp). The PCR was performed under the following conditions (3 min at 94°C, followed by 30 cycles of 30 s at 94°C, 1 min at 60°C, and 1 min at 72°C, followed by 5 min at 72°C).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
To assess the oncogenic potential of type 1 mutated truncated Notch1IC for T cells and myeloid cells, Tg mice expressing these sequences under the regulatory sequences of the human CD4 gene (CD4C) were generated (CD4C/Notch1IC) (Fig. 1A). The CD4C regulatory sequences have previously been shown to allow expression of surrogate genes in immature CD4+CD8+ and mature CD4+CD8– T cells as well as in cells of the macrophage/myeloid lineage, notably in macrophages and in dendritic cells (19, 28). The Notch1IC gene contained the complete IC domain of Notch1. Of the four Tg founders harboring the CD4C/Notch1IC transgene produced and exhibiting transgene RNA expression, two died with thymic lymphoma at the age of 1.5 mo without giving progeny. Founder lines F30856 and F35748 were established by mating the other two Tg founder mice with outbred CD1 mice. Expression of the transgene was confirmed by Northern (Fig. 1B) and by Western (Fig. 1C) blot analysis and found to be high in mice from both founder lines. As expected, expression was higher in the thymus than in the spleen (data not shown).
|
|
To determine the nature of cells being transformed, TCR and Ig gene rearrangements were studied on DNA from thymic lymphomas arising in CD4C/Notch1IC Tg mice. The TCR -chain gene was found to be rearranged or deleted in both alleles in all nine thymic lymphomas from Tg mice tested (data not shown). The Ig H chain gene was found to be rearranged in three of nine thymic lymphomas tested (data not shown). These results indicated that these thymic lymphomas are clonal or monoclonal in origin and that they belong to the T cell lineage.
These results were further confirmed by the characterization of the transformed thymic cell population by FACS analysis. Cells isolated from thymic lymphomas were analyzed for expression of various cell surface markers of T cells (Thy-1, CD2, CD3, CD5, CD44, CD25, CD4, CD8, TCR, TCR, and HSA (CD24)), B cells (B220), macrophages (F4/80, Mac-1), dendritic cells (Mac-1/CD11c), and NK cells (DX-5). The lymphoma cells were distinguished from normal thymocytes by their larger size (FSC, mean channel = 142) (Fig. 2Ba). All thymic lymphomas tested (n = 22; 15 from F35748 and 7 from F30856) were of the CD4+CD8+ DP phenotype and were Thy-1+, CD2+. Only a few tumors among all those tested from both founders show the presence of some SP CD4+ (2 of 22 tumors) or SP CD8+ (1 of 22 tumors) T cells. Two phenotypes could be observed according to the TCR and HSA expression: type A (TCRlowHSA–) and type B (TCRhighHSA+) (Fig. 2B). Type A represented 8 of 15 and 3 of 7 tumors among those analyzed, respectively, from F35748 and F30856, the remaining belonging to type B.
The CD25 and CD44 cell surface molecules are normally expressed in CD4–CD8– double-negative thymocytes or by activated T cells. In thymic lymphomas, DP lymphoma cells expressed high levels of CD25+ and CD44+ (12 of 15 (F35748) and 2 of 2 (F30856) tumors): F35748, SP CD44+ (2 of 15), mixed SP CD25+/SP CD44+ (4 of 15), mixed SP CD25+/SP CD44+/DP CD25+CD44+ (9 of 15); F30856, mixed SP CD25+/SP CD44+/DP CD25+CD44+ (2 of 7) and DP CD25+CD44+ (5 of 7) (Fig. 2C). However, these tumors did not express another T cell activation marker, CD69 (data not shown). The expression of CD25 and CD44, but not of CD69, in these thymic lymphoma cells suggests that these are abnormal DP cells. These DP thymic lymphomas also showed absence of B220, F4/80, Mac-1, DX-5, and CD11c staining (data not shown), ruling out an origin from B cells, macrophages, NK cells, or dendritic cells. However, in some thymic lymphomas, an increase of B220+ or Mac-1+ cells was found. This increase was not associated with any of the type A or B, nor with age.
Tg expression of Notch1EC is not oncogenic for CD4+CD8+ thymocytes
Type 1 Notch1 mutation generated by provirus insertion leads to the truncation of the gene and the production of not only Notch1IC, but also of a truncated Notch1 extracellular domain (Notch1EC). To assess the oncogenic potential of Notch1EC, Tg mice expressing Notch1EC under the regulation of the CD4C sequences were generated (CD4C/Notch1EC) (Fig. 3A). Although these Tg mice expressed the transgene at high levels (Fig. 3, B–D), none of them (0 of 21) developed thymic lymphoma during the 10- to 15-mo observation period (Fig. 3E). These results indicated that Notch1EC is not oncogenic by itself for T cells.
|
Tg expression of Notch1CT induces CD4+CD8+ metastatic thymic lymphomas
Among the Notch1 mutations induced by provirus insertions, we previously described type II mutations giving rise to Notch1 molecules deleted of their C terminus (Notch1CT). To assess the oncogenic potential of Notch1CT, Tg mice expressing Notch1CT under the CD4C regulatory sequences were generated (CD4C/Notch1CT) (Fig. 4A). Three founder lines (F92743, F92744, F92745) were established by breeding with CD1 mice. Expression of the transgene was highest in the F92745 Tg mice, as determined by Northern (Fig. 4B) or Western (Fig. 4C) blot analysis. Direct comparison of expression of Notch1CT and Notch1IC Tg RNA on the same gel with a probe common to both constructs showed that Notch1CT from both founders (F92744, F92745) were expressed at lower levels than Notch1IC (Fig. 4B).
|
CT Tg mice developed thymic lymphoma at various frequency in different founder lines (30%, F92743; 70%, F92744; 94%, F92745 at 10 mo of age) (Fig. 4D). In contrast to the thymic tumors of CD4C/Notch1IC Tg mice, which rarely metastasize, the CD4C/NotchCT thymomas were more frequently (14 of 14 total; 6 of 6, F92744; 8 of 8, F92745) accompanied by tumors in the peripheral lymphoid organs (spleen and/or LN), as observed by macroscopic pathology (14 of 14) and as confirmed by FACS analysis (8 of 10 analyzed) (see below), indicating a disseminating metastatic phenotype. When inoculated s.c. into nude mice, most of these thymomas (5 of 5, F92744; 5 of 6, F92745) grew as local tumors of medium size (1- to 2-cm diameter) within 3–5 (F92744) or 3.5–7 (F92745) wk, indicating their malignancy.
The phenotypic characterization of the transformed thymic and peripheral cell populations done by FACS confirmed that they were DP CD4+CD8+, TCRhigh/intermediate, CD2high, CD5high, and CD24+ or CD24– (Fig. 4, E and F), as well as CD3high (data not shown) T cells. These thymic lymphomas were negative for F4/80, Mac-1, DX-5, and CD11c staining. However, as in the CD4C/Notch1IC thymoma cells, the CD4/Notch1CT thymoma contained some B220+ and/or Mac-1+ cells. Among those analyzed (n = 6, F92744; n = 8, F92745), these tumors were positive for CD25 or for CD44 or for both (Fig. 4F): F92745, SP CD25+ (2 of 8), SP CD44+ (4 of 8), mixed SP CD25+/SP CD44+/DP CD25+CD44+ (2 of 8); F92744, SP CD44+ (3 of 6), mixed SP CD25+/SP CD44+/DP CD25+ CD44+ (3 of 6). Together, these results indicated that the transformed cells are abnormal DP-positive thymocytes. We found only 1 tumor (of 14) with a small percentage of SP CD8+ T cells (Fig. 4F).
At the same time as the type II mutated Notch1CT is generated by provirus insertions, the Notch1 C terminus domain is found to be located downstream of the viral LTR promoter and in the same transcriptional orientation. This suggested that expression of this C terminus domain (Notch1CT) may somehow be involved in the transformation process. A similar truncated Drosophila NotchCT domain expressed in Tg flies was found to abrogate disheveled-dependent pathways (29). To study the contribution of Notch1CT in the oncogenic process, Tg mice expressing Notch1CT under the regulation of CD4C sequences were constructed (CD4C/Notch1CT) (Fig. 4A). Two founder lines (F87827, F87828) were established by breeding with CD1 mice. Although the expression of the transgene was high, as assessed by RT-PCR analysis (data not shown), these Tg mice did not develop any thymic or peripheral tumors (0 of 32) during the 12-mo observation period, suggesting that Notch1CT is not oncogenic by itself.
Notch1IC and Notch1CT collaborate with c-myc for tumor formation
Our initial screen leading to the discovery of a high frequency of provirus insertional Notch1 mutants was conducted in Tg mice overexpressing c-myc in DP thymocytes (MMTVD/c-myc) (12). These results not only suggested that these Notch1IC truncated forms might be involved in T cell tumor development, but also that these truncated molecules may be cooperating with c-myc for oncogenesis. To test directly whether Notch1IC is one of the c-myc cooperators in lymphomagenesis, CD4C/Notch1IC Tg mice (F30856) were crossed with MMTVD/c-myc Tg mice to generate double-Tg mice. Both Tg lines were on a CD1 background. All (100%) the double-Tg mice analyzed (n = 17) developed thymic lymphomas with a latency of 29–44 days (mean = 37 ± 4 days) (Fig. 5A). This latent period for tumor development was much shorter than that in single Notch1IC or single c-myc Tg littermates (102 ± 40 days (56.2%) and 133 ± 43 days (88.8%), respectively), within 210 days for observation. No significant difference was found in the dissemination and/or infiltration of spleen and liver between the double-Tg mice and Notch1IC or c-myc single-Tg mice bearing thymic lymphomas. Northern blot analysis or RT-PCR showed that Tg Notch1IC and c-myc were expressed in these thymic lymphomas (data no shown). Interestingly, a novel phenotype of weight loss and wasting, not present in single Notch1IC or c-myc Tg mice, was observed in some Notch1IC/myc double-Tg mice bearing thymic lymphomas. Together, these results indicated a clear collaboration of Notch1IC and c-myc in accelerating the oncogenic process in thymic T cells.
|
gene (data not shown). All the tumors analyzed by FACS (n = 9) were CD4+CD8+ and expressed T cell-specific markers (data not shown). However, in contrast to CD4C/Notch1IC or MMTVD/c-myc single-Tg thymic lymphomas, Notch1IC/c-myc double-Tg thymic lymphomas could not be clearly divided into two phenotypes, and expressed different levels of TCR or HSA. In contrast to lymphomas arising from CD4C/Notch1IC single-Tg mice, a large fraction (6 of 9) of these lymphomas expressed CD44, and only a few were positive for both CD25 and CD44. None were positive for CD69 (data not shown).
To determine whether Notch1CT could also collaborate with c-myc for tumor formation, we generated double (Notch1CT x c-myc)-Tg mice. All (100%) these double-Tg mice analyzed (n = 14) developed thymic lymphomas with a latency of 48–80 days (mean 56 days) (Fig. 5B). The latency is much shorter than that of single Notch1CT or single c-myc Tg littermates (mean 99 and 90 days) for the 6-mo observation period (Kaplan-Meier test, p < 0.0001). Six of these double Tg thymomas were further analyzed by FACS and confirmed as DP. These thymomas could be characterized as TCRhigh/CD2high/CD24–/CD44+/CD25– (3 of 6) or TCRintermediate/CD2high/CD24–/CD44+/CD25– (2 of 6) or TCRhigh/CD2high/CD24+/CD44+/CD25+ (1 of 6). Most thymomas from double-Tg mice showed dissemination/infiltration of spleen and LN (11 of 14). The immunophenotype of spleen and LN lymphomas was confirmed by FACS analysis as abnormal DP cells similar to the DP thymomas.
Preleukemic thymocytes expressing Notch1IC, but not those expressing Notch1CT, show an increased number of abnormal (CD2high, TCRhigh) DP CD4+CD8+ T cells
To determine whether the expression of these gain-of-function Notch1 mutants affects T cell development, the lymphoid organs (thymus, spleen, and peripheral LNs) of young (6-wk-old) Tg mice and their control non-Tg littermates from two founder lines (F35748, F30856) were studied. A reproducible increase in the total thymocyte number was observed in the higher expressor founder (F35748) CD4C/NotchIC Tg mice (Fig. 6, A and B), but not in mice of the lower expressor F30856 line (data not shown). FACS analysis showed that there was no significant difference in the FSC profiles (non-Tg = 13 ± 1.5 vs Tg = 15 ± 1.8) nor in the percentage of thymocytes staining with Thy-1, CD5, CD44, HSA, TCR, and B220 or with double staining with Mac-1/CD11c CD44/CD25 between Tg mice and their non-Tg littermates (data not shown). This analysis showed a modest increase in the percentage of CD4+CD8+ thymocytes, whereas the percentage of CD4+ or CD8+ SP thymocytes was slightly decreased (data not shown). However, the increase in total cell number in mice of the F35748 line was mainly due to an increase in the DP cell population, whereas no difference could be documented for SP thymocyte subsets (Fig. 6B). Three-color analysis revealed that, in comparison with non-Tg thymocytes, a significantly higher number of these DP CD4+CD8+ Tg thymocytes from both founders were CD2high, TCRhigh, and CD25+ (Fig. 6C), although changes in expression of these markers seem to occur independently from each other. These results suggest an expansion of an abnormal DP subpopulation. A similar analysis conducted on very young mice (18 days old) revealed no significant change in thymocyte numbers, nor in the percentage of the DP CD4+CD8+, SP CD4+, SP CD8+, and T cell subsets in Tg as compared with non-Tg mice (data not shown). In the peripheral organs (spleen and LN) of CD4C/Notch1IC Tg young mice, no significant difference was observed in the numbers and proportions of T cells (CD4, CD8) and B cells, as compared with normal non-Tg controls (data not shown).
|
CT Tg mice (6 wk old) shows a small decrease of total thymocytes in mice of F92743 and F92744 Tg line (data not shown), but a very little difference in total thymocyte number in Tg mice from the high expressor F92745 line (Fig. 7B) relative to non-Tg mice. The FSC profiles of thymocytes (non-Tg = 13.5 ± 1.5 vs Tg = 14 ± 2) were comparable in Tg and non-Tg mice. The proportion and absolute cell number of DP cells were slightly decreased in Tg mice from F92743 and F92744 Tg lines (data not shown), but were also not statistically different in Tg mice from the F92745 founder line as compared with non-Tg littermates (Fig. 7, A and B). However, SP CD4 and SP CD8 thymocytes were slightly decreased only in mice from F92743 and F92744 lines (data not shown), but not in those from the F92745 line (Fig. 7B). In addition, specific changes found in CD4C/Notch1IC Tg mice were not observed. For example, we could not document a significant difference in the number of DP cells expressing TCRhigh (Fig. 7, A and C). However, as in CD4C/Notch1IC Tg mice, we could document a significant increase of the CD2high DP population, both in percentage (15.3 ± 8 Tg vs 7 ± 5 non-Tg) and in absolute cell number (Fig. 7C) in the three founder lines of CD4C/Notch1CT Tg mice. CD25+ DP cells were found to be significantly increased in percentage in all three founders (p < 0.01) (data not shown) and in absolute cell number in two (F92743, F92745) of three of the founders (Fig. 7C and data not shown). In peripheral LNs, we observed a modest increase in percentage of CD8+, and a small decrease in CD4+ T cells for both founders analyzed (F92743, F92745) (data not shown). In the spleen, no significant difference could be seen.
|
CT mutant, highlighting the importance of the C terminus Notch1 domain.
Notch1IC or Notch1CT expression leads to increased proliferation of DP CD4+CD8+ thymocytes in preleukemic mice
To determine whether the increased number of DP thymocytes of CD4C/Notch1IC Tg mice reflects an enhanced proliferation, cell cycle analysis was performed with 7-AAD on freshly isolated DP thymocytes from 6-wk-old Tg (n = 12) (F30856) and non-Tg (n = 5) mice. This analysis revealed that a comparable percentage of CD4C/Notch1IC Tg (28.4 ± 2.5%) and non-Tg (21.3 ± 4.6%) DP cells was in S phase. Similarly, no difference of total DNA content in DP thymocytes was observed between CD4C/Notch CT Tg mice (n = 12) from both founders and non-Tg littermates (n = 8) (data not shown). DNA synthesis was then assessed more directly with anti-BrdU FACS analysis 4 h following in vivo BrdU inoculation or after 24 h of BrdU provided in drinking water. No difference of BrdU incorporated in DP, SP CD4, and SP CD8 thymocytes of CD4C/Notch1IC Tg mice (F30856) (n = 9) relative to their non-Tg littermates (n = 9) could be documented in the 24-h analysis (data not shown). However, when the analysis was performed at a shorter interval after BrdU inoculation (4 h), an increase of BrdU incorporation in DP, SP CD4+, and SP CD8+ thymocytes from both CD4C/Notch1IC and CD4C/Notch1CT Tg mice was observed (Fig. 8).
|
Notch1IC, but not Notch1CT, protects thymocytes from dexamethasone or anti-CD3-induced apoptosis in vivo and in vitro
Because Notch1 insertional mutation was initially observed by our group in tumors of Tg mice overexpressing c-myc in the thymus, we speculated that these gain-of-function insertional Notch1 mutants could be involved in apoptosis. To test this hypothesis, Tg and non-Tg control mice were inoculated i.p. with dexamethasone or anti-CD3 Abs. Forty-eight hours after the treatment, Tg mice (F30856) exhibited a higher number of their total thymocytes than non-Tg littermates (Fig. 9A). FACS analysis of thymocytes of these mice revealed a significantly (p < 0.001) higher proportion of remaining CD4+CD8+ DP thymocytes in Tg than in control non-Tg mice treated with anti-CD3 (75.98 ± 6.67% vs 18.2 ± 11%) or with dexamethasone (72 ± 8% vs 10 ± 5%) (Fig. 9B). This protection of DP cells by Notch1IC was also evident by their higher (10-fold) numbers remaining after both treatments (Fig. 9C). Labeling with 7-AAD for detection of apoptotic/dead cells confirmed a lower proportion of dead cells among Tg DP thymocytes compared with that among non-Tg DP thymocytes, after dexamethasone or anti-CD3 treatment (Fig. 9D).
|
CT Tg mice (F92745) did not show significant protection against dexamethasone- nor anti-CD3-induced apoptosis of DP thymocytes (Fig. 9, C and D).
Because assessment of apoptosis at 48 h following the apoptotic stimuli may reflect not only the rate of apoptosis, but also the rate regenerative capacity of DP thymocytes within this interval, we performed the same experiment at 4 h following inoculation of dexamethasone and anti-CD3 mAb. The results of these experiments were very similar to those obtained after 48 h. A protection of DP thymocytes against apoptosis induced by dexamethasone and anti-CD3 mAb could be documented in CD4C/Notch1IC Tg mice (Fig. 9E). Such a protection was not evident in CD4C/Notch1CT Tg relative to non-Tg mice, except for a modest protection in Tg mice of one founder (F92744). However, when compared with sham-treated animals, this group of Tg mice still had significant loss of DP cells.
To determine whether this phenotype could be reproduced in vitro, total thymocytes from control and Tg mice were incubated in vitro for 3 h in the presence or absence of the same (dexamethasone) or related (anti-CD3 + CD28 mAb) apoptotic stimuli as those used in vivo. FACS analysis was then performed by gating on DP CD4+CD8+ cells after labeling with 7-AAD to detect apoptotic/dead cells. DP thymocytes from CD4C/N1IC Tg mice showed a significantly lower number of 7-AAD-positive cells relative to those from non-Tg mice (Fig. 9F), suggesting that Notch1IC had a protective effect. This was not the case for DP cells from CD4C/Notch1CT Tg mice, which showed comparable percentage of 7-AAD-positive cells as non-Tg ones (Fig. 9F). Similar results were obtained with annexin V/PI labeling (data not shown).
Therefore, overexpression of Notch1IC, but not Notch1CT mutant protected thymocytes, especially the CD4+CD8+ DP subpopulation, from dexamethasone or TCR-induced apoptosis in vivo and in vitro. These results suggest that the C terminus domain of Notch1 may be required for this protection.
Development of thymomas after transplantation of CD4C/Notch1IC FLCs into lethally irradiated mice
To assess whether Tg thymic stromal epithelial cells were required for tumor development, lethally irradiated recipient mice were reconstituted with FLCs from CD4C/Notch1IC Tg mice or from their normal non-Tg littermates and observed for up to 6 mo. None of the recipient mice receiving non-Tg FLC developed tumors. In contrast, four of six mice reconstituted with Tg FLC developed tumors (two thymomas, one peripheral LN tumor, and one with grossly enlarged spleen) after 3- to 4-mo latency. FACS analysis showed that one thymoma was constituted of a mix of TCRhigh DP and TCRhigh CD8+ SP T cells, while the other was composed of TCR+ DP cells. The peripheral LN tumor was TCRlow/negative DP cells.
These data indicate that Tg stromal epithelial cells are not absolutely required for the development of Notch1IC-mediated tumors.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
CT, in T cells of Tg mice. Notch1 mutants and T cell development
Notch1IC, but not Notch1CT, was found to induce an increase of total number of thymocytes, especi
, Human CD4C promoter with mouse enhancer (En) (
); 
, Notch1 IC domain (aa 1659–2533), including the TM domain, ankyrin (ANK) repeats, OPA, and PEST domains; 
, SV40 poly(A) signal; Bs, BssHII; E, EcoRI; S, SalI. B, Northern blot analysis of Notch1IC transgene RNA expression. Thymus (T) or spleen (S) RNA from animals of each Tg founder line (F30856 and F35748) and from non-Tg littermates were hybridized with IC Notch1 probe K. Note the higher detection of the 3.5-kb Notch1IC RNA in Tg thymuses. The gel was stripped and hybridized with the actin probe to monitor gel loading. E, endogenous Notch1 RNA. The numbers below the gel represent the intensity of the signals relative to that in the non-Tg mouse (= 1) after correction for actin intensity, as measured on the Storm PhosphorImager. C, Western blot analysis of Notch1IC transgene protein expression in thymocytes. Whole-thymocyte lysates from Tg mice of two founder lines (F30856 and F35748) were reacted on two gels with anti-Notch1-intra Abs with Alexa 680-coupled secondary Abs, as described in Materials and Methods. The mutated truncated Notch1IC protein was previously found to comigrate with the 110-kDa processed wild-type IC Notch1 protein in this gel system (
) and the F35748 (•) lines as well as for their non-Tg littermates (
). Results are indicated as the percentage of mice that were found dead and had thymic lymphoma or were sacrificed because of dyspnea caused by thymic lymphomas. B, FACS analysis of thymic lymphomas. FSC analysis (a). Notice that thymic lymphoma cells are larger than most normal thymocytes. No difference is observed between type A and type B tumors. The numbers over the markers show the percentage of FSC. Cell surface markers on thymic lymphoma cells included CD2 (b), TCR
) Tg mice and in each cross their control littermates single CD4C/Notch1IC or CD4C/Notch
) Tg mice, as well as non-Tg mice (
