A p53 Defect Sensitizes Various Stages of B Cell Development to Lymphomagenesis in Mice Carrying an IgH 3′ Regulatory Region-Driven c-myc Transgene

Tumorigenesis is a multistep process in which several genetic lesions have to accumulate to produce a fully malignant phenotype. Ongoing recombination and mutation during B cell development make the IgH locus a hot spot for translocation. Thus, many mature lymphomas are marked by protooncogene translocation into the IgH locus. Cyclin D1 translocations found in mantle cell lymphoma (MCL) take place during V(D)J recombination (1). c-myc translocation, a typical hallmark of Burkitt’s lymphoma (BL), is related to either somatic hypermutation (SHM) and chromosomal breakage within the variable exon or class switch recombination (CSR) (2). The cyclin D1 or c-myc translocations often observed in plasma cell lymphoma (PCL) rather correspond to illegitimate CSR events (3). As a consequence of these translocations, oncogenes are constitutively expressed under the control of active IgH enhancers (4, 5). Among IgH cis-regulatory elements, Eμ was expected to be the critical oncogene deregulator in B lymphomagenesis. Thus, Eμ-c-myc transgenic mice express c-myc in B cell progenitors but develop lymphomas that harbor an immature B cell signature (6) far from the mature phenotype of human BL (2). Among IgH enhancers, those (hs3a, hs1,2, hs3b, and hs4) located in the IgH 3′ regulatory region (3′RR) are reported as the master control elements of CSR and of the transcriptional burst associated with plasma cell differentiation (4, 5, 7). Convincing demonstration of the essential contribution of the 3′RR in mature lymphomagenesis has been produced by murine models of B cell lymphomas with various types of IgH-c-myc translocations in the context of either a normal IgH locus or a locus with truncated and functionally deficient 3′RR (8). The 3′RR thus proved dispensable for pro-B cell lymphomas with V(D)J recombination-initiated translocations but is required for mature B cell lymphomas with CSR-associated translocations. It is thus tempting to speculate that the 3′RR is the master control element of c-myc deregulation after c-myc translocation in mature B cell lymphomas. Whereas transgenic models of oncogene deregulation have given insights into the molecular mechanisms of lymphomagenesis, they rarely provided true counterparts of human pathologies. Transgenics indeed often feature overexpression of an oncogene from the beginning of B cell development and without any regulation, although the natural history of lymphomagenesis rather reflects a sequence of oncogenic events in a defined order. In this context, a model yielding a broad spectrum of lymphoproliferative pathologies could be of interest and behave as a sensor for detecting which additional genetic or environmental influence may favor the development of one specific type of tumor. Thus, c-myc-3′RR transgenic mice are prone to develop nonactivated BL-like proliferations (IgM⁺IgD⁺CD43⁻ cells) (9) with a marked overproliferation signature but without downregulation of the apoptotic signature (10). p53 is a transcription factor that triggers growth inhibitory and apoptotic responses (11). Inactivation of p53 function is the more common feature of human tumors. The frequent loss of p53 function in human lymphomas underscores its critical role in suppressing the emergence of incipient tumors (12). A genetic anomaly initially restricted to c-myc deregulation clearly selects tumors with a homogeneous high proliferating signature and no need for additional alteration of apoptotic pathways. We tested the hypothesis that an early alteration of apoptosis in a model of deregulated proliferation (i.e., 3′RR-driven c-myc overexpression in B cells) would allow the onset of tumors in a greater number of animals and yield not only highly proliferative BL but also less-proliferative types of lymphomas as observed in human patients where c-myc deregulation frequently occurs in pathologies such as MCL and PCL. We thus crossed c-myc-3′RR transgenic mice to mice heterozygous for germline deletion in p53 (p53^+/−) (13). The onset of c-myc-3′RR/p53^+/− lymphomas was greatly accelerated compared with c-myc-3′RR control mice. Although c-myc-3′RR mice were prone to develop nonactivated CD43⁻ BL [CD43 is a sialylated single-chain membrane glycoprotein expressed on activated B cells (14)], a wider pattern of lymphomas occurred in c-myc-3′RR/p53^+/− mice including CD43⁻ BL, CD43⁺ BL, MCL-like lymphoma (IgM⁺CD5⁺CD43⁺CD23⁻ cells), and PCL (IgM^−/lowIgD^−/lowCD138⁺ cells). The observed spectrum of tumors provides strong evidence that in the natural history of lymphomagenesis, an initial defect of apoptosis can endow a 3′RR-mediated deregulation of c-myc to mimic c-myc translocation seen in multiple human lymphomas and promote the occurrence of various types of mature B cell lymphomas.

The study was approved by the Centre National de la Recherche Scientifique and the university review committee. p53-deficient mice were in a C57BL/6 background (13). p53-deficient mice were crossed with mice carrying a c-myc transgene driven by the IgH locus 3′ regulatory region (c-myc-3′RR in a C57BL/6 background) (9, 10) to obtain c-myc-3′RR/p53^+/− mice.

Genomic DNA was extracted from mouse tails. Ten micrograms of DNA was digested with EcoRI, loaded on 0.7% agarose gels, transferred on nylon sheets (Amersham, Buckinghamshire, U.K.), hybridized with a ³²P-labeled hs4 probe (a 1.3-kb PstI fragment), and autoradiographed to ensure the presence of the c-myc-3′RR transgene as previously reported (9). Genomic DNA prepared from splenic cells was digested with EcoRI and analyzed with a ³²P-labeled J_H4 probe to test lymphoma clonality as previously reported (9).

PCR experiments for wild-type (wt) p53 alleles were carried out with specific forward 5′-ACACGCTGGTGGTACCTTAT-3′ and reverse 5′-CACATGTACTTGTAGTGGATGG-3′ primers. PCR experiments for mutated p53 alleles were carried out with specific forward 5′-CTATCAGGACATAGCGTTGG-3′ and reverse 5′-CACATGTACTTGTAGTGGATGG-3′ primers. DNA was denatured for 120 s at 94°C and then submitted to 35 cycles consisting of 94°C for 30 s, 55°C for 30 s, and 72°C for 40 s. Amplification products were analyzed on an 0.7% agarose gel. Expected sizes of amplified products were 600 bp and 400 bp for mutated and wt alleles, respectively. PCR experiments for P1/P2 c-myc promoters were carried out with forward 5′-GGCTGCGCTGCTCTCAGCTG-3′ and 5′-CTGACTCGCTGTAGTAAT TCC-3′ primers (for P1 and P2 promoters, respectively) and reverse 5′-GACCACCAGATCTGTGCTTA-3′ primer. cDNA was denatured for 120 s at 94°C and then submitted to 40 cycles consisting of 94°C for 30 s, 58°C (for P1) or 62°C (for P2) for 30 s, and 72°C for 60 s. Amplification products were analyzed on an 0.7% agarose gel. Expected sizes of amplified products were 580 bp and 425 bp for P1 and P2 transcripts, respectively. RT-PCR analysis of Iα-Cα germinal transcripts and Iμ-Cα–switched transcripts were done as previously described (7) with the following PCR primers: Iμ forward, 5′-CTCTGGCCCTGCTTATGTTTG-3′; Cα reverse, 5′-GAGCTGGTGGGAGTGTCAGTG-3′; Iα forward, 5′-CCTGGCTGTTCCCC TATGAA-3′; β-actin forward, 5′-TACCTCATGAAGATCCTCA-3′; β-actin reverse, 5′-TTCGTGGATGCCACAGGAC-3′.

Total RNA was extracted using Trireagent (Ambion, Austin, TX). Real-time PCR was performed in duplicate by using TaqMan assay reagents and analyzed on an ABI Prism 7000 system (Applied Biosystems, Foster City, CA). Assay references were Mm00494449-m1 (INK4a); Mm00483241-m1 (INK4b); Mm00483243-m1 (INK4c); Mm00432448-m1 (p21/Cip); Mm00438168-m1 (p27/kip1); Mm00500563-m1 (MSH2); Mm00487769-g1 (MSH6); Mm00449156-m1 (UNG); Mm00453168-m1 (polymerase η [POLη]); Mm00450983-m1 (Rev1); Mm01184115-m1 (activation-induced cytidine deaminase); Mm00441964-g1 (Trp53); Mm0043259-m1 (cyclin D1); Mm01612362-m1 (cyclin D3); and Mm99999915-g1 (GAPDH for normalization of gene expression levels) (Applied Biosystems).

Single-cell suspensions from spleen and bone marrow of healthy mice and lymph node-derived tumors were labeled with various Abs (Southern Biotechnologies, Birmingham, AL) conjugated with PE and allophycocyanin. Control experiments included irrelevant isotype-matched Abs conjugated with PE and allophycocyanin. Cells were analyzed on an LSRII cytometer (BD Biosciences). For intracellular labeling experiments, cells were fixed and permeabilized with the Intraprep permeabilization reagent (Beckman Coulter) according to the manufacturer’s recommendations prior to incubation with FITC–anti-Ki67 (Becton Dickinson) or irrelevant Abs (Cell Signaling Technology)

Splenic B cells (2 × 10⁵ cells/well) were cultured (in sextuplicate) in 96-well plates, either alone or in the presence of LPS from Salmonella typhimurium (Sigma, L’Isle d’Abeau Chenes, France) or anti-CD40 (R&D Systems, Lille, France) for 72 h. The number of viable cells in proliferation was assessed using the CellTiter 96 One Solution Cell Proliferation assay (Promega) according to the manufacturer’s recommendations. The assay is a homogeneous, colorimetric method for determining the number of viable cells in proliferation assays. The assay is composed of solutions of a tetrazolium compound [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium; MTS] and an electron coupling reagent phenazine methosulfate. MTS is bioreduced by cells into a formazan product that is soluble in tissue culture medium. The absorbance of the formazan product at 490 nm can be measured directly from 96-well assay plates without additional processing. The conversion of MTS into the aqueous soluble formazan product is accomplished by dehydrogenase enzymes found in metabolically active cells. The quantity of formazan product as measured by the amount of 490-nm absorbance is directly proportional to the number of living cells in culture.

Freshly isolated splenocytes (1 × 10⁶ cells/ml) were cultured in growth medium in 24-well plates with or without 10 μg/ml LPS. After 24 and 72 h of growth, cells were fixed, permeabilized, and incubated with FITC-labeled anti-caspase 3 Abs or irrelevant Abs (Cell Signaling Technology).

Single-cell suspensions of spleen cells were cultured 5 d at 1 × 10⁶ cells/ml in RPMI 1640 with 10% FCS and 40 μg/ml LPS, with or without addition of cytokines: 20 ng/ml IL-4, 2 ng/ml TGF-β, or 2 ng/ml IFN-γ (PeproTech, Rocky Hill, NJ). Supernatants were then harvested and stored at −20°C until Ig evaluations. In paralleled experiments, 4-d in vitro-stimulated splenocytes were washed in PBS and stained with various Abs: anti-B220 (CD45R)–PC5, anti-CD138–PE, and anti-Ig subclasses conjugated with FITC. Cells were analyzed on an LSRII cytometer.

Blood samples were recovered from transgenic mice and wt controls with heparinized needles.

Culture supernatants and sera from c-myc-3′RR, p53^+/−, and c-myc-3′RR/ p53^+/− mice were analyzed for the presence of the various Ig classes (IgM, IgG₁, IgG_2a, IgG2b, IgG₃, IgE, and IgA) by ELISA as previously described (15).

Gene expression profiles of B lymphomas from c-myc-3′RR/p53^+/− mice were compared with B lymphomas from c-myc-3′RR mice. mRNA was extracted from B cells purified with CD19-coupled beads (>98% of purity) from Miltenyi Biotec (Bergisch Gladbach, Germany) according to the manufacturer’s recommendations. Experiments were done in quadruplicate. cDNA labeling, dual color microarray hybridization, and microarray data analysis were done in the Nice Sophia-Antipolis Microarray Facility (Nice Sophia Antipolis, France) essentially as previously described (16, 17). Statistical analysis was performed with the Bioconductor open source software, particularly its Limma package (18). The microarray data presented in this article have been submitted to the Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/geo/) under the accession number GSE31814.

Genomic DNA from c-myc-3′RR/p53^+/−–derived lymphoma was amplified by PCR using the following primers and multistep programs: the forward primer VHJ558 5′-GCGAAGCTTARGCCTGGGRCTTCAGTGAAG-3′ complementary to the VHJ558 segment and the backward primer 5′-AGGCTCTGAGATCCCTAGACAG-3′ corresponding to a sequence located 150 bp downstream of the J_H4 segment using 1 cycle at 94°C for 5 min, 35 cycles (94°C for 45 s, 58°C for 45 s, and 72°C for 5 min), and 1 cycle at 72°C for 10 min. PCR was carried out in 50 μl Taq 1× PCR buffer containing 300 μl of each deoxynucleoside triphosphate, 100 ng of each primer, and 1 U Taq polymerase.

The PCR products were separated on a 1.2% agarose gel and cloned into the pGEM-T easy vector system (Promega, Madison, WI) according to the manufacturer’s instructions. Plasmids were isolated using the NucleoSpin kit (Macherey-Nagel Eurl, Hoerdt, France), and inserts were sequenced using an automated laser fluorescent ANA ABI-PRISM sequencer (Perkin-Elmer, Branchburg, NJ).

The genetically engineered knockout p53 mice were crossed with mice carrying a c-myc transgene driven by the IgH 3′RR (c-myc-3′RR) (9) prone to develop CD43⁻ BL (CD43⁻IgM⁺IgD⁺ cells) to obtain c-myc-3′RR/p53^+/− mice. The presence of the c-myc transgene was assessed by Southern blot using a c-myc ³²P-labeled probe, and the p53 mutation was followed by using a specific PCR.

As shown in Fig. 1A, the life span of c-myc-3′RR/p53^+/− mice was markedly decreased compared with that of c-myc-3′RR mice and of p53^+/− mice. The overall tumor incidence reached 100% in c-myc-3′RR/p53^+/− mice compared with 75% in c-myc-3′RR mice. No lymphoma was found in p53^+/− mice over a period of 16 mo. The mean ages of death were 13 wk and 30 wk for c-myc-3′RR/p53^+/− mice and c-myc-3′RR mice, respectively. Beginning at the age of 10 wk, c-myc-3′RR/p53^+/− mice progressively developed profound enlargement of lymph nodes and spleen (Fig. 1B). Mice exhibiting obvious tumors or presenting signs of illness were sacrificed.

We analyzed cells in 4-wk-old c-myc-3′RR/p53^+/− mice before any manifestation of disease. Spleens were of normal sizes with germinal centers of normal morphologies. Similar numbers (p > 0.05, Mann–Whitney U test) of WBCs were found in blood of young c-myc-3′RR/p53^+/− mice (10.2 ± 3.9 × 10³ cells/μl, mean ± SEM of 4 mice), c-myc-3′RR mice (11.0 ± 2.0 × 10³ cells/μl, 4 mice), and p53^+/− mice (10.5 ± 1.2 × 10³ cells/μl, 4 mice). Similar numbers (p > 0.05) of WBCs were found in spleen of young c-myc-3′RR/p53^+/− mice (43.0 ± 1.9 × 10⁶ cells, 4 mice), c-myc-3′RR mice (58.8 ± 5.8 × 10⁶ cells, 4 mice), and p53^+/− mice (59.5 ± 11.4 × 10⁶ cells, 4 mice). Similar numbers (p > 0.05) of WBCs were found in bone marrow of young c-myc-3′RR/p53^+/− mice (18.3 ± 4.2 × 10⁶ cells, 4 mice), c-myc-3′RR mice (25.7 ± 2.6 × 10⁶ cells, 4 mice), and p53^+/− mice (19.6 ± 5.7 × 10⁶ cells, 4 mice). B splenocytes from young c-myc-3′RR/p53^+/− mice showed slightly higher proliferative responses to in vitro stimulations with high amounts of LPS (Fig. 2A) or anti-CD40 (Fig. 2B) compared with those of B splenocytes from 4-wk-old c-myc-3′RR mice. As shown in Fig. 2C, the apoptosis rate (assessed by the percentage of caspase 3-positive cells after 3-d in vitro LPS stimulation) was similar in B splenocytes from 4-wk-old c-myc-3′RR/p53^+/−, p53^+/−, and c-myc-3′RR mice. Plasma cell maturation was not affected by the p53 mutation as judged by similar amounts of CD138⁺ plasmablasts (Fig. 2D) after 3-d in vitro LPS activation in c-myc-3′RR/p53^+/− mice (28.9 ± 3.6%; mean ± SEM of 3 mice) and c-myc-3′RR mice (24.4 ± 0.5%, 3 mice). Finally, their in vivo and in vitro ability to secrete all the various Ig isotypes in response to LPS and cytokines was normal (data not shown). As shown in Fig. 2E, circulating WBC counts markedly increased with age in c-myc-3′RR/p53^+/− mice. In some mice, WBC counts became elevated as early as 8 wk. Beginning at the age of 10 wk, some c-myc-3′RR/p53^+/− mice exhibited obvious lymph node tumors. Taken together, these results strongly suggested B cell lymphomagenesis and not simply a proliferative B cell defect in c-myc-3′RR/p53^+/− mice.

Lymphomas from 37 animals were investigated (Table I). All animal had obvious lymph node involvement (Fig. 1B; Table I). In some cases, lymphomas were also located in GALT (Fig. 1C; Table I). Flow cytometry was used to monitor the immunophenotypic profile of these lymphomas. A B220⁺ (89%, 33 of 37) and CD19⁺ (100%, 37 of 37) population was characterized (Fig. 3), whereas staining for T lineage and monocyte/macrophage Ags (CD4, CD8, Cd11b) was negative (data not shown). Analysis of 37 animals revealed four different lymphoma types that appeared with similar latency. Of importance, animals exhibited only one type of lymphoma. The first type of lymphoma (8%, 3 of 37) was characterized as CD43⁻ BL (Fig. 3). Cells were IgM⁺IgD⁺ but negative for CD5, CD43, and CD138 surface markers. These tumors have a phenotype similar to those developed by c-myc-3′RR mice (9). The second type of lymphoma (38%, 14 of 37) was diagnosed as activated BL. Cells were IgM⁺IgD⁺ but differed from the previous ones by the presence of the CD43 Ag. The third type of lymphoma (24%, 9 of 37) was characterized as MCL-like lymphoma. Cells were IgM⁺IgD⁺, expressed CD43 and CD5 surface markers, but were negative for the CD138 surface marker (Fig. 3). Among them, three did not express the B220 Ag. c-myc-3′RR mice with a defect of the INK4-Cdk4 checkpoint (the Cdk4^R24C mutation) have been recently reported to develop MCL-like lymphoma (C. Vincent-Fabert, R. Fiancette, P. Rouaud, C. Baudet, V. Truffinet, V. Magnone, M. Cogné, P. Dubus, and Y. Denizot, submitted for publication). The fourth type of lymphoma (30%, 11 of 37) was characterized as PCL. Cells were IgM^low/−IgD^low/−, expressed CD43 and CD138 surface markers, but were negative for the CD5 Ag. One of them was negative for the B220 Ag. The expression of CD138 varied with degree of differentiation, and no evidence of secreted monoclonal Ig was documented. PCL arose predominantly (10 of 11, 91%) with GALT involvement (Table I). RT-PCR analysis of PCL transcripts revealed the presence of Iα-Cα germinal transcripts and Iμ-Cα–switched transcripts showing B cell commitment toward IgA isotype (Fig. 1D).

Southern analysis of V(D)J recombinations showed that all B cell lymphoma (i.e., BL, MCL-like lymphoma, and PCL) had undergone clonotypic Ig rearrangements. Thus, the use of a J_H4 probe revealed rearranged bands in addition to the germline band indicating the clonal origin of lymphomas from c-myc-3′RR/p53^+/− mice (Fig. 1E).

Analysis of blood of these animals revealed leukemic invasion with lymphoma cells (Table I). The number of circulating WBCs was markedly elevated (p < 0.001, Mann–Whitney U test) in lymphoma mice (80.8 ± 21.4 × 10³ cells/μl, mean ± SEM of 29 mice) compared with that in wt (i.e., c-myc-3′RR negative and p53^+/+) mice (11.1 ± 0.5 × 10³ cells/μl, 7 mice). However, no significant differences (p > 0.15) were found for number of circulating cells between CD43⁻ BL (30.5 ± 10.0 × 10³ cells/μl, 3 mice), CD43⁺ BL (73.2 ± 15.1 × 10³ cells/μl, 11 mice), MCL-like lymphoma (63.5 ± 23.4 × 10³ cells/μl, 6 mice), and PCL (123.8 ± 64.2 × 10³ cells/μl, 9 mice). The high (p < 0.05, Mann–Whitney U test) proliferative activity of these lymphoma cells was highlighted by the high expression of the nuclear proliferation-associated Ag Ki67 (assessed by flow cytometry), a nuclear protein present during G₁, S, G₂, and M phases of the cell cycle (67 ± 3% versus 20 ± 2% of Ki67⁺ cells for 22 lymphomas and WBCs from 4 wt mice, respectively). The Ki67 labeling index was 39.6 ± 14.2 (3 mice) in CD43⁻ BL, 70.3 ± 2.8 (7 mice) in CD43⁺ BL, 70.2 ± 4.8 (7 mice) in MCL-like lymphoma, and 76.3 ± 5.6 (6 mice) in PCL. Real-time PCR analysis did not highlight differences between c-myc transcript levels in CD43⁻ BL (relative expression of 2.33 ± 0.95, n = 3), CD43⁺ BL (relative expression of 2.98 ± 1.28, n = 9), MCL-like lymphoma (relative expression of 1.88 ± 0.12, n = 7), and PCL (relative expression of 2.81 ± 0.25, n = 6). Western blot analysis did not reveal significant differences in the amounts of c-Myc protein in these lymphomas (data not shown). Regression analysis showed a significant (p = 0.0002; r = 0.77) link between the number of circulating WBCs and c-myc transcript levels. Expression of c-myc transcripts in B cell lymphomas of c-myc-3′RR/p53^+/− mice involved not only the conventional P2 but also the P1 promoter. P1 c-myc transcripts were detected in 50 (4 of 8), 88 (8 of 9) and 100% (13 of 13) of MCL-like lymphoma, PCL, and BL, respectively.

Several human B cell lymphomas carrying c-myc chromosomal translocations harbor mutations in the rearranged c-myc gene. In human BL, mutations are reported upstream and downstream of P1/P2 (19–21). These mutations might be selected during tumor development and might affect c-myc mRNA transcription. Specific PCR primers were used to distinguish between the endogenous and the transgenic c-myc in c-myc-3′RR/p53^+/− tumors. No mutations were found in transgenic c-myc of BL, MCL-like lymphoma, and PCL of c-myc-3′RR/p53^+/− mice (Supplemental Fig. 1).

We analyzed the mutation status of V_H rearranged genes in B cell lymphoma of c-myc-3′RR/p53^+/− mice. This study was made possible because of the clonal nature of c-myc-3′RR/p53^+/− lymphomas (Fig. 1E). The 150 bp located downstream of the J_H4 exon were investigated as a hot spot of SHM (22, 23). Results from seven lymphomas (three BL and four MCL-like lymphomas) and 64 sequences are summarized in Supplemental Fig. 2 (upper panel). All lymphomas revealed unmutated sequences and thus featured a pregerminal center origin.

SHM introduces nucleotide substitution into V_H region genes at all four bases. Mutations at C/G pairs are a direct consequence of C→U deamination catalyzed by activation-induced cytidine deaminase (AID). The resulting U/G mispair is recognized by the UNG uracil-DNA glycosylase. UNG-mediated recognition followed by replication by error-prone polymerases such as REV1 results in the generation of both transition and transversion substitutions around the initially mutated C/G pairs. MSH2/MSH6-mediated recognition of the U/G mispair triggers a POLη-dependent patch repair process that results in mutations extended to A/T pairs (24). As shown in Supplemental Fig. 2 (lower panel), transcript levels for AID, MSH2, MSH6, and POLη were markedly lower in BL and MCL-like lymphoma compared with those of CD19⁺ B splenocytes of healthy c-myc-3′RR/p53^+/− mice. Transcripts levels were not significantly different for PCL (except elevated for REV1).

D-type cyclins (D1, D2, and D3) are G₁-specific cyclins that associate with Cdk4 or Cdk6 and promote restriction point progression during G₁ phase (25). Cdk4 and Cdk6 may have different functions (26). Cdk6, but not Cdk4, is reported to play a key role in cytokine-mediated B cell proliferation (27) and B cell differentiation (28). Deregulated expression of Cdk6 predisposes cells to malignant transformation (29), and Cdk6 is overexpressed in numerous B cell lymphomas (30). Western blot analysis of Cdk2, Cdk4, Cdk6, cyclin D1, cyclin D2, and cyclin D3 proteins was performed in BL, MCL-like lymphoma, and PCL of c-myc-3′RR/p53^+/− mice. No significant levels of Cdk6 were detected in nonmalignant B splenocytes of c-myc-3′RR/p53^+/− mice demonstrating that Cdk4 is the primary catalytic partner for D-type cyclin in these cells (Fig. 4). In contrast, even if Cdk4 levels were markedly elevated in c-myc-3′RR/p^53+/− lymphomas, only Cdk6–cyclin complexes were documented demonstrating that Cdk6 is the primary catalytic partner for D-type cyclins in c-myc-3′RR/Cdk4^R24C–derived lymphomas. In agreement with their high Ki67 index, high amounts of free cyclin D1, D2, and D3 as well as Cdk–cyclin D1 and Cdk–cyclin D2 complexes were evidenced in c-myc-3′RR/p^53+/− lymphomas whatever their B cell phenotype (BL, MCL-like phenotype, or PCL). Cyclin E–Cdk2 complexes have important roles in cell cycle progression, although selective Cdk2 inhibition is readily compensated (31). The vast majority of Cdk2 was found in Cdk2 complexes in nonmalignant B splenocytes of c-myc-3′RR/p^53+/− mice. In contrast, all types of c-myc-3′RR/p^53+/−–derived lymphomas expressed predominantly free Cdk2 protein, and only low amounts of Cdk2 complexes (Fig. 4). D-cyclins levels (especially D1 and D2) were markedly elevated in c-myc-3′RR/p^53+/−–derived lymphomas compared with those in nonmalignant B splenocytes.

As a step toward characterizing relevant differences in transcriptional profiles between MCL-like lymphoma, BL, and PCL in c-myc-3′RR/p53^+/− mice, we performed mRNA expression analysis using an array of 43,000 genes. The 220 most different genes are reported in Supplemental Table I. MCL-like lymphoma evidenced a clearly different signature compared with BL and PCL. Although many of these genes still have unknown functions, some of them seem relevant to the tumor phenotype (Table II). MCL-like lymphoma expressed high levels of genes implicated in cell signaling such as Irak1bp1, Mapkap1, Lst1, Pr2G4, Cdk1, and Lyar. Antiapoptotic genes (Pr2g4, Nos2, Uchl3) were also documented. BL and PCL exhibited a rather similar signature (Supplemental Table I). However specific differences could be evidenced between BL and PCL. BL expressed a marked antiapoptotic signature compared with PCL by overexpression of genes such as Bid, Tgfβ1, and Tada (Table II). BL also overexpressed genes implicated in cancer such as Prkcb and Mzf1. In turn, PCL overexpressed several anti-inflammatory and proliferation genes such as Ptgs1, Maged1, and Slpi.

Many mature lymphomas are marked by protooncogene translocation into the IgH locus (1–3). As a consequence of these translocations, oncogenes are constitutively expressed under the control of active IgH enhancers (4, 5). Among IgH enhancers, those located in the IgH 3′RR are reported as the master control elements of CSR and of the transcriptional burst associated with plasma cell differentiation (4, 5, 7). Expression of c-myc is tightly linked to the early G₁ phase of the cell cycle and plays a critical role in cell proliferation, differentiation, metabolism, and apoptosis (32). Although classically described as the driving oncogene in BL, abnormalities of c-myc have been reported in other types of lymphoma (33). Beyond BL, c-myc is also overexpressed in a subset of human diffuse large B cell lymphomas, MCL, and PCL (1–3). Despite the early recognition of the importance of c-myc during lymphomagenesis, efficient and robust in vivo animal models of mature B cell lymphomas are not available. c-myc transgenes linked to the Eμ element were shown to promote B cell malignancies, although with a pro-B/pre-B phenotype (6). A c-myc insertion downstream of J_H (mimicking the translocation observed in endemic BL) induced in aged animals (overall incidence of 68% by 21 mo) the development of B cell and plasma cell neoplasms resulting from c-myc activation (34, 35). A truncated 3′RR cassette (lacking hs3a) integrated upstream of the c-myc coding region on mouse chromosome 15 induced B cell lymphomas in aged animals (mean age of death of 314 and 379 d for homozygous and heterozygous animals, respectively) (36). Finally c-myc-3′RR mice are prone to BL-like lymphomas with incomplete penetrance (75% tumor incidence) and long latencies (10–12 mo) (9). To our knowledge, no animal model has been reported to develop c-myc–induced mature B cell malignancies with a short latency (3–4 mo), with full penetrance (100% tumor incidence), and with a lymphoma panel mimicking the wide diversity of human c-myc–induced mature B cell lymphomas.

p53 is a transcription factor that triggers growth inhibitory and apoptotic responses (11). Inactivation of p53 function is a common feature of human tumors. Because 3′IgH enhancers are the master control elements of CSR and of the transcriptional burst associated with plasma cell differentiation (4, 7), we tested the hypothesis that an intercross of the c-myc-3′RR mice on a genetic background of strain deficient for p53 would yield a robust mouse model of c-myc–induced B cell lymphomagenesis. The Eμ-c-myc B cell lymphoma model has been previously used to study alterations in the p53 pathway. Eμ-c-myc transgenic mice succumb to immature (pro-B, pre-B) B cell lymphomas with a mean survival of 4–6 mo (35). Crossing with p53^+/− mice that had a mean survival of >15 mo, Eμ-c-myc/p53^+/− mice exhibited a mean survival of 35 d (37). Similarly to Eμ-c-myc and Eμ-c-myc/p53^+/− mice, c-myc-3′RR/p53^+/− mice survived for significantly less time than c-myc-3′RR mice (mean of 13 wk and 30 wk for c-myc-3′RR/p53^+/− mice and c-myc-3′RR mice, respectively). c-myc-3′RR/p53^+/− mice developed B cell lymphoma with a full penetrance (100% tumor incidence) compared with that of c-myc-3′RR mice (75% tumor incidence). Whether or not the p53^+/− mutation increases the rate of B cell lymphomagenesis in Eμ-c-myc mice but without affecting its phenotype (already immature forms of B cell lymphoma), c-myc-3′RR/p53^+/− mice developed a wide pattern of B cell monoclonal malignancies such as BL (CD43⁺ or CD43⁻ forms), MCL-like lymphoma, and PCL compared with c-myc-3′RR mice. BL, activated BL, and MCL-like lymphoma were IgM⁺IgD⁺ thus in agreement with their lack of class-switched phenotype. Regarding PCL, they were IgA, potentially predicted on their formation in GALT. p53 heterozygosty has, thus, a discernible effect on disease onset and mortality in c-myc–driven lymphomagenesis. The main effect of p53 loss in c-myc-3′RR mice is not only an altered course of disease but also a change in B cell lymphoma phenotype. p53 regulates a plethora of processes that modulate cellular proliferation, survival, and apoptosis. The progression of lymphoma elicited by c-myc overexpression in Eμ-myc/p53 mice resulted in increased proliferation (37). Similarly, lymphomas found in c-myc-3′RR/p53^+/− mice exhibited an increased proliferation signature. Even when phenotypically related to indolent forms of human tumors such as MCL, the tumors observed in mice thus rather corresponded to the more aggressive and highly proliferative forms of these diseases. Thus, marked alterations in D-type cyclins and Cdks were evidenced not only on BL but also in MCL-like lymphoma and PCL. Such alterations have been previously evidenced in their human counterpart (1–3, 33). Whether or not Cdk4 is the primary catalytic partner for D-type cyclins in normal B cells of c-myc-3′RR/p53^+/− mice, Cdk6 is the one in c-myc-3′RR/p53^+/− lymphomas. Confirming previous studies (29), our current work further indicates a potential role for Cdk6 in malignant B cell transformation. Myc acts as a transcription factor that regulates >15% of all cellular genes by controlling the expression of some genes and repression of others(33). The Eμ intronic enhancer is the master control element for V(D)J recombination revealing its action during early lymphopoiesis. Confirming this, Eμ-GFP transgenic mice expressed GFP at the pro-B/pre-B cell stages (38). Thus, its action during lymphomagenesis takes place during immature stages of B cell maturation. The 3′RR acts from the pre-B to the mature B cell stage (39). Thus, its action during lymphomagenesis takes place during mature stages of B cell maturation. It is thus conceivable all types of lymphoma overexpressing c-myc share a common mechanism for immortalization whatever the B cell stage of maturity or differentiation of the initial B cell target.

In conclusion, the observed spectrum of tumors observed in c-myc-3′RR/p53^+/− mice provides strong evidence that the 3′RR, depending on a context including the mutational status of p53, can be considered as the master control element of oncogene deregulation after oncogene translocation into the IgH locus in mature B cell lymphomas. c-myc-3′RR/p53^+/− mice closely reproduce many features of human BL, MCL, and PCL and provide a novel and efficient model to dissect the molecular events leading to c-myc–induced lymphomagenesis and an important tool to test potential therapeutic agents or environmental conditions modulating the initial occurrence of lymphomas.

We thank S. Desforges and B. Remerand for help with animal care. We acknowledge the technological expertise of the Nice Sophia-Antipolis Functional Genomics Platform. We thank Pascal Barbry for helpful discussions and support during this work.

The authors have no financial conflicts of interest.