Rearrangement of CCNDl (BCLI/PRADI) 3’ Untranslated Region in Mantle- Cell Lymphomas and t(llql3)-Associated Leukemias

By Ruth Rimokh, FranGoise Berger, Christian Bastard, Bernard Klein, Martine French, Eric Archimbaud, Jean Pierre Rouault, Benedicte Santa Lucia, Laurent Duret, Michele Vuillaume, Bertrand Coiffier,

Paul-Andre Bryon, and Jean Pierre Magaud

Rearrangement and overexpression of CCNDl iBCLl/ PRADl), a member of the cyclin G, gene family, are consis- tent features of t(llql31-bearing B-lymphoid tumors (partic- ularly mantle-cell lymphoma [MCL]). lts deregulation is thought to perturb the G,-S transition of the cell cycle and thereby to contribute to tumor development. As suggested by previously published studies, rearrangement of the 3‘ untranslated region (3’ UTR) of CCNDl may contribute to its activation in some lymphoid tumors. To define further the prevalence of such rearrangements, we report here the result of the molecular study of 34 MCL and six t(llq13)- associated leukemias using a set of probes specific to the different parts of the CCNDl transcript. We also sequenced the entire cDNA of the overexpressed CCNDl transcripts in

MAJORITY OF HUMAN hematopoietic malignancies carry nonrandom chromosomal alterations; experi-

mental evidence indicates that genes located at recurring chromosomal breakpoints are directly involved in tumor pathogenesis.’ The t(l1; 14)(q13;q32) translocation and its molecular counterpart, bcl-l rearrangement, are consistent features of the subtype of non-Hodgkin’s lymphoma desig- nated centrocytic lymphoma or the similar intermediate lymphocytic lymphoma:” now referred to as mantle-cell lymphoma (MCL).’ As a result of this translocation, the putative BCLUPRADI proto-oncogene, on chromosome 11, is juxtaposed to an IgH-enhancer sequence located on chro- mosome 14.9 The BCLUPRADI gene is known under differ- ent names (PRADI, cyclin Dl , BCLI, DllS287E). We will refer to it as CCNDI, in accordance with the official nomen- clature.” Chromosomal breakpoints are widely scattered on chromosome 1 lq13, but three hot spots have been individu- alized: the major translocation cluster (MTC)9 lying approxi- mately 110 kb centromeric of the CCNDl gene, and two minor translocation clusters (mTCs), which are less fre- quently involved: mTCl and mTC2.I’ mTCl is localized approximately 22 kb telomeric of MTC,” and mTC2,”.’’ which has been recently identified, maps to the 5’ flanking region of the CCNDl gene (probe B as described by Wil- liams et all’). CCNDl is a member of the cyclin G, fam- ilyI4-l7; its deregulation is thought to perturb the G,-S transi- tion of the cell cycle and thereby to contribute to human tumor development. This hypothesis is in agreement with the finding that CCNDl is overexpressed in the vast majority of the MCL and t( 1 lql3)-associated leukemias analyzed

With regard to the mechanisms of activation of CCNDl in lymphoid tumors, one can consider several possibilities. It has been first postulated that activation of CCNDl in t( 11; 14)-carrying tumors is a consequence of its juxtaposi- tion to an immunoglobulin &-acting sequence. According to another hypothesis, CCNDI activation might occur at the posttranscriptional level through rearrangement of its 3’ un- translated region (3’ UTR). In fact, in primary tumors and in

A

so far.ll.18-20

a t(llql3)-associated leukemia. DNA from four of these 40 patients showed rearrangement of the 3’ UTR of CCNDl coexisting with major translocation cluster (MTC) re- arrangement. Southern blot and sequence analyses showed that, as a result of these rearrangements, the 3’ AU-rich region containing sequences involved in mRNA stabilii and in translational control is eliminated. Moreover, the finding that the CCNDl mRNA half-life was greater than 3 hours (normal tissues, 0.5 hours) in three t(llql3)-associated cell lines stresses the importance of posttranscriptional de- rangement in the activation of CCNDl. Finally, we did not observe any mutation in the coding frame of the CCNDl cDNA analyzed. 0 1994 by The American Society of Hematology.

cell lines with chromosome 1 lq13 structural abnormalities, a 1 S-kb mRNA can be detected in addition to the normal 4.5- kb mRNA. Sequencing of CCNDl cDNA in a few cell lines has shown that the smaller transcript corresponds to a short- ened form of the normal 4.5-kb transcript as a result of the use of different polyadenylation signals or of deletions of the 3’ end of the gene.’5,’6,’9 These data suggested that, in some cases, activation of CCNDl might result from the loss of 3‘ regulatory sequences, which are known to be implicated in mRNA stability and in translational control. Finally, se- quence analysis of the CCNDl coding region in primary human tumors has shown that amino acid sustitutions are not required for CCNDl’s oncogenic effects.”

The aim of the present study was to determine the inci- dence and the consequences on the transcription rate of re- arrangement of the 3’ UTR of CCNDl in primary human tumors. For this purpose, we first sequenced the coding re- gion and the 3’ UTR of the overexpressed CCNDl transcripts in one t(llql3)-associated leukemia. We next analyzed breakpoints that map to the CCNDI 3’ UTR in a large series of MCL and t(l lql3)-associated leukemias by using a set

From the Luboratoire de Cytoginitique Moliculaire and the Laboratoire d’Anatomie Pathologique, Hbpital Edouard Herriot, Lyon, France; Laboratoire de Cytoginitique, CRTS, Bois Guil- laume, France: and Dipartement d’Himatologie, Centre Hospitalier Lyon Sud, Lyon, France.

Submitted August 2, 1993; accepted February 23, 1994. Supported by grants from INSERM (CRE 92 0108, CJF 93-08),

CNRS, Hospices Civils de Lyon, Ligue Nationale Contre le Cancer (Comiti du Rhbne, de la Haute-Savoie et de la Sabne et Loire).

Address reprint requests to Ruth Rimokh, PhD, Laboratoire de Cytoginitique Moliculaire, Pavillon E, Hbpital Edouard Herriot, 69437 Lyon Cedex 03, France.

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “advertisement” in accordance with I 8 U.S.C. section 1734 solely to indicate this fact.

0 1994 by The American Society of Hematology. 0006-4971/94/8312-0028$3.00/0

Blood, Vol 83, No 12 (June 15). 1994: pp 3689-3696 3689

3690 RIMOKH ET AL

of probes specific to the different parts of the CCNDl tran- script. The results of MTC, mTC 1, and mTC2 rearrangement analysis of most of the samples presented here have been previously reported.” Finally, when material was available, CCNDl protein level was assessed in these tumors by West- em blot analysis.

MATERIALS AND METHODS

Tissue samples and cell lines. Thirty-three cases of MCL with available frozen material were analyzed (cases no. 1 to 33). Cytoge- netic data were not available for any of these MCL cases. Each case was classified by using histologic criteria previously

We also analyzed four cases of B-cell chronic lymphocytic leuke- mia (B-CLL), one case of plasma-cell leukemia carrying a t(11;14)(q13;q32) (cases no. 34 to 38), and one case of B-CLL carrying a variant t(l1; 19)(q13;q13) translocation (case no. 39). All cases (no. 1 to 39) were included in a previous report.”

Cell-suspension and frozen-section immunophenotypic studies were performed as previously reported.26 The following markers were used in this study: IgG, IgM, IgD, IgA, K and X light chains, B1 (CDZO), B4 (CD19), Leu-l4 (CD22), TlOl (CD5), OKT3 (CD3), Leu-9 (CD7), and JS(CDl0). All MCL cases were monoclonal B cells; 24 of the 30 tested showed CD5 expression and CD10 was positive in only seven of the 31 cases tested. All B-cell leukemias expressed monoclonal immunoglobulin and were positive for mark- ers of the B-cell lineage.

The Gob. cell line was derived from the leukemic cells of patient no. 39. Malignant cells continued to proliferate in RPM1 1640 with 20% fetal calf serum only in the presence of recombinant human interleukin-2 (rIL-2,20 to 30 IUlmL; Boehringer Mannheim, Mann- heim, Germany). After 3 months of culture, rIL-2 was no longer necessary for the cells to proliferate. It is likely that, as a consequence of its mitogenic effect, rIL-2 favored the occurrence of secondary genetic changes leading to growth factor-independence. Cells from the original culture have now been growing continuously for nearly 1 year. Immunophenotype and karyotype were identical to those seen during the initial examination.

Two cell lines carrying a t(11;14)(q13;q32) translocation: Rec-l (referred to as case 40) established from a MCL,” and XG5, a human myeloma cell line”; Ramos, a Burkitt’s lymphoma cell line; reactive lymph nodes, normal tonsils, and human fibroblasts were also used in this study.

DNA and RNA isolation and analysis. High-molecular weight DNA was extracted from fresh cells or from frozen material follow- ing standard procedures. After digestion with appropriate endonucle- ases as recommended by the suppliers (Boehringer Mannheim), DNA fragments were electrophoresed on 0.8% agarose gels and transferred onto nylon filters.

Total cellular RNA was isolated from cultured cell lines or from frozen samples by the acid guanidium thiocyanate-phenol-chloro- form method. For Northern blot analysis, poly(A+) or total RNA was size fractionated in formaldehyde-1.2% agarose gels and trans- ferred onto nylon filters.

DNA probes and hybridization procedures. A 2.3-kb Sac1 frag- ment representing the MTC on chromosome 1 lq13 was a gift from Y. Tsujimoto (Philadelphia, PA). Probes “A” and “E” arepolymer- ase chain reaction (PCR)-amplified DNA fragments encompassing nucleotides 148 to 310 and 2,856 to 3,091, respectively, on the human PRADl cDNA sequence.14 Probes “B,” “C,” and “D” correspond to CCNDl cDNA subclones (Fig 1). MYC exon 3 probe was a gift from D. Stehelin (Lille, France). Alpha-”P-labeling, pre- hybridization, hybridization, and washing were performed as pre- viously described.*’

Preparation and analysis of cDNA library. Poly(A+) RNA ex- tracted from case no. 39 was used to construct an oligo-dT-primed human cDNA library in the lambda gtl 1 vector as recommended by the supplier (Pharmacia, Uppsala, Sweden).

Sequencingprocedure and sequence analysis. Overlapping dele- tions of CCNDl cDNA cloned into Bluescript. SK(-) (Stratagene, La Jolla, CA) were obtained by the unidirectional exonuclease 111 digestion method (Erase-a-Base System; Promega, Madison, WI). Deletion clones were sized on agarose gels, and the inserts of the selected clones were sequenced using the double-stranded DNA se- quencing technique (dideoxy chain termination procedure) with Sequenase I1 (USB, Cleveland, OH) as described by the manufac- turer.

Study of CCNDI mRNA stability. Exponentially growing cells (Rec-l, XG5, and Gob, cell lines) were treated with 10 pg/mL of actinomycin D (Sigma Chemical, St Louis, MO). CCNDl mRNA level was then assessed by Northern blot analysis at various times after addition of actinomycin D. The Northern blot was sequentially hybridized to a CCNDl probe and to a MYC exon 3 probe as a control.

Western blot analysis. For Western blot analysis, proteins from the pathological samples were solubilized in lysis buffer ( 1 50 mmol/ L NaCL, 50 mmoVL Tris-HCL, pH 8, 0.5% sodium deoxycholate, 1% Triton X100). Proteins (30 pg) from each sample were then separated by sodium dodecyl sulfate 10% polyacrylamide slab gel electrophoresis and electrophoretically transferred onto nitrocellu- lose filters. For immunodetection, the filters were incubated with a U750 dilution of a polyclonal rabbit anti-cyclin D antibody (UBI, New York, NY) at room temperature. After washing, fixation of the antibody was shown by using a blotting detection kit for rabbit antibodies (Amersham, Uppsala, Sweden). The polyclonal antibody used in this study was raised against a peptide contained in the C- terminal domain of the protein (residues 208 to 295).

RESULTS

In a previous study, we showed that in MCL and t( 1 Iq 13)- associated leukemias, the steady-state level of CCNDl tran- script is dramatically elevated relative to any other lymphoid tissues” and that a 1.5-kb mRNA is predominantly ex- pressed.

One case of t( 11; 19)-associated leukemia (case no. 39) expressed, in addition to the 1.5-kb transcript, two aberrant transcripts of 2 and 3 kb, indicating that the CCNDl mRNA may be altered in this tumor (Fig 2) . To characterize further the CCNDl transcripts in this lymphoid malignancy, a cDNA library was made from fresh tumoral cells. Screening of this library with probe A allowed us to isolate 12 positive clones. Eleven clones covering less than 1,500 bp may correspond to the 1.5-kb transcript. The longest of them was entirely sequenced; it is almost identical to the region encompassing nucleotides 1 to 1346 on the PRADl cDNA sequence,14 except for two base changes: (1) a G is present at nucleotide 870 instead of an A in the coding sequence; (2) a c is present at nucleotide l100 instead of an A in the 3’ UTR. None of the base changes observed altered the amino acid sequence of the protein. There is no recognizable polyadenylation se- quence (AATAAA) anywhere within the sequence of the 1 i clones analyzed, and none of them contains a poly(A) tail. However, the corresponding transcripts are polyadenylated, since poly(A+) selection increased the signal obtained on Northern blot (Fig 2) . It is likely that, in this oligo-dT-

REARRANGEMENT OF CCNDl 3' UNTRANSLATED REGION 3691

5' 3' MTC W € H P € x

CEN. +/ B BP

I I I I I I - TEL. l10 kb

CCND1 cDNA U- U

A B C O E

p n P SPP L I . 1 . L .

&

39- cDNA 3 kb

39- cDNA 1.35 kb Fig 1. Comparison of the restriction maps of the 3' end of CCNDl with those of the normal 4.5-kb CCNDl cDNA and of two cDNA clones

isolated from case no. 39. Black boxes indicate the coding region; dotted line represents the MER11 sequence; double-headed arrow corresponds to the CCNDT last exon; slided-horizontal lines indicate the position of the probes used in this study. CEN, centromere; TEL. telomere, X, Xbd; E, EcoRI; H, Hindlll; P, P A ; B, BsmHI; Bg, Bgfll; S, Sad.

primed library, all the cDNA clones were initiated from the numerous poly(A) stretches present within the 3' end of the CCNDI mRNA.

We also isolated a 3,013-nucleotide long cDNA clone representative of the 3-kb aberrant transcript. The first 2,275 nucleotides of this clone are identical to nucleotides 2 to 2276 of the PRADI cDNA ~equence, '~ with the same base changes at positions 870 and 1 1 0 0 , and are followed by a sequence (nucleotides 2276 to 3013) unrelated to the CCNDI/PRADI gene (Fig 3). Comparison of this sequence with those in data bases (EMBL data base) showed that it corresponds to a new member of the MER1 1 sequences family. In fact, this region has greater than 90% similarity with two members (HUMP45C17, HUMSIGG3) of this fam- ily of medium reiteration frequency repetitive ~ e q u e n c e s . ~ ~ ~ * ~ It should be noted that in all the analyzed cDNA clones, the CCNDl coding frame is retained and that the mRNA-

4 34 40

39 A T

Fig 2. Northern blot analysis of CCNDl expression in MCL or leu- kemias with rearrangement of the CCNDT 3' UTR. (Left1 10 p g of total RNAfrom each sample was processed for Northern blot analysis as described in the Methods and the filter was hybridized to a CCNDT probe (probe Al. (Right) 5 p g of poly(A'1 RNA (A) and 5 p g of total RNA from case no. 39 were blotted and hybridized to probe A.

destabilizing signals AUUUA present in the normal tran- script are eliminated. In case no. 39, PCR amplification of genomic DNA with primers flanking the CCNDllMERll junction (nucleotides 2090 to 2465) yielded a 375-bp frag- ment whose sequence matched perfectly that of the 3-kb aberrant transcript (data not shown), demonstrating that the rearrangement occurred at the genomic level and was not the result of a cloning artifact. The fact that the same base changes was observed in the two groups of cDNA suggests that the 1.5-kb transcript represents an alternatively pro- cessed form of the 3-kb mRNA and that they both originate from the same allele.

On Southern blots containing BamHl and HindIII-cleaved DNA from case no. 39, probes D and E detected two differ- ent rearranged bands in addition to the same germline band (Fig 4). The first explanation of this result is that re- arrangement of the 3' end of CCNDI corresponds to an insertion of the MER1 1 sequence at this site. However, as case no. 39 is associated with a variant t(11;19)(q13;q13) translocation, we cannot rule out the possibility that the MER1 I sequence originates from chromosome 19 and that the rearranged band detected by probe E corresponds to the der( 19) chromosome. In Fig 4, case no. 39, the intensity of the rearranged bands differs considerably as compared with the residual germline band. This pattern was not observed when using probes A, B, and E (data not shown). This indi- cates that on the der( 11) chromosome, the 3' CCNDl UTR corresponding to probe D was amplified during the transloca- tion process.

This observation and a previously published study" show- ing that the 3' end of CCNDl was rearranged in one case of lymphoma suggested that, in some tumors, its deregula- tion might occur at the posttranscriptional level. To test this hypothesis, we have analyzed a large series of MCL and t( 1 Iql3)-associated leukemias using probes specific to the different parts of the CCNDl transcript (Fig 1). Three of the 40 tumors (cases no. 4, 34, and 40; case no. 39 not included) showed rearrangement of the 3' end of CCNDI. Southern blot analysis using probes B or C detected a rearranged fragment in BamHI- and Sad-cleaved DNA (Fig 5). None

3692 RIMOKH ET A t

GCGCAGTAGCAGCGAGCAGCAGAGTCCGCACGCTCCGGCGAGGGGCAGAAGAGCGCGAGGGAGCGCGGGGCAGCAGAAGCGAGAGCCGAGCGCGGACC

CAGCCAGGACCCACAGCCCTCCCCAGCTGCCCAGGAAGAGCCCCAGCCATGGAACACCAGCTCCTGTGCTGCGAAGTGGAAACCATCCGCCGCGCGTAC MetGluHisGlnLeuLeuCysCysGluValGluThrIleArgArgAlaTyr

CCCGATGCCAACCTCCTCAACGACCGGGTGCTGCGGGCCATGCTGAAGGCGGAGGAGACCTGCGCGCCCTCGGTGTCCTACTTCAAATGTGTGCAGAAG PrOASpA~aASnLeuLeuAsnAspArgValLeuArgAlaMetLeuLysAlaGluGluThrCysAlaProSerValSerTyrPheLysCysValGlnLys

G~uValLeuProSerMetArgLysIleValAlaThrTrpMetLeuGluValCysGluGluGlnLysCysGluG1uGluValPheProLeuAlaMetAsn GAGGTCCTGCCGTCCATGCGGAAGATCGTCGCCACCTGGATGCTGGAGGTCTGCGAGGAACAGAAGTGCGAGGAGGAGGTCTTCCCGCTGGCCATGAAC

TyrLwAspArgPheLeuSerLeuGluProValLysLysSerArgLeuGlnLeuLeuGlyAlaThrCysMetPheValAlaSerLysHetLysGluThr TACCTGGACCGCTTCCTGTCGCTGGAGCCCGTGAAAAAGAGCCGCCTGCAGCTGCTGGGGGCCACTTGCATGTTCGTGGCCTCTAAGATGAAGGAGACC

ATCCCCCTGACGGCCGAGAAGCTGTGCATCTACACCGACAACTCCATCCGGCCCGAGGAGCTGCTGCAAATGGAGCTGCTCCTGGTGAACMGCTCAAG

TGGAACCTGGCCGCAATGACCCCGCACGATTTCATTGAACACTTCCTCTCCAAAATGCCAGAGGCGGAGGAGAACAAACAGATCATCCGCAAACACGCG TrpAsnLeuAlaAlaMetThrProHisAspPheIleGluHisPheLeuSerLysMetProGluAlaGluGluAsnLysGlnIleIleArgLysHiSA~a

CAGACCTTCGTTGCCCTCTGTGCCACAGATGTGAAGTTCATTTCCAATCCGCCCTCCATGGTGGCAGCGGGGAGCGTGGTGGCCGCAGTGCAAGGCCTG G1nThrPheValAlaLeuCysAlaThrAspValLysPheIleSerAsnProProSerMetValAlaAlaGlySerValValAlaAlaValGlnGlyLeu

I1eProLeuThrAlaGluLysLeuCysIleTyrThrAspAsnSerIleArgProGluGluLeuLeuGlnHetGluLeuLeuLwValASnLysLeuLyS

AACCTGAGGAGCCCCAACAACTTCCTGTCCTACTACCGCCTCACACGCTTCCTCTCCAGAGTGATCAAGTGTGACCC~GACTGCCTCCGGGCCTGCCAG ASnLwArgSerProAsnAsnPheLeuSerTyrTyrArgLeuThrArgPheLeuSerArgValIleLysCysAspProAspCysLeuArgAlaCySGln

GluGlnIleGluAlaLeuLeuGluSerSerLeuArgGlnAlaGlnG1nAsnHetAspProLysAlaAlaGluGluGluGluGluGluGluGluGluVal GAGCAGATCGAAGCCCTGCTGGAGTCAAGCCTGCGCCAGGCCCAGCAGAACATGGACCCCAAGGCCGCCGAGGAGGAGGAAGAGGAGGAGGAGGAGGTG

AspLeuAlaCysThrProThrAspValArgAspValAspIle GACCTGGCTTGCACACCCACCGACGTGCGGGACGTGGACATCTGAGGGCGCCAGGCAGGCGGGCGCCACCGCCACCCGCAGCGAGGGCGGAGCCGGCCC

CAGGTGCTCC~CTGACAGTCCCTCCTCTCCGGAGCATTTTGATACCAGAAGGGAAAGCTTCATTCTCCTTGTTGTTGGTTGTTTTTTCCTTTGCTCTTT CCCCCTTCCATCTCTGACTTAAGCAAAAGAAAAAGATTACCCAAAAACTGTCTTTAAAAGAGAGAGAGAGAAAAAAAAAATAGTATTTGCATMCCCTG AGCGGTGGGGGAGGAGGGTTGTGCTACAGATGATAGAGGATTTTATACCCCAATAATCAACTCGTTTTTATATTAATGTACTTGTTTCTCTGTTGTAAG AATAGGCATTMCACAMGGAGGCGTCTCGGGAGAGGATTAGGTTCCATCCTTTACGTGTTTAAAAAAAAGCATAAAAACATTTTAAAAACATAGAAAA ATTCAGCAMCCATTTTTAAAGTAGAAGAGGGTTTTAGGTAGAAAAACATATTCTTGTGCTTTTCCTGATAAAGCACAGCTGTAGT~~TTCTAGGCA TCTCTGTACTTTGCTTGCTCATATGCATGTAGTCACTTTATAAGTCATTGTATGTTATTATATTCCGTAGGTAGATGTGTAACCTCTTCACCTTATTCA TGGCTGMGTCACCTCTTGGTTACAGTAGCGTAGCGTGGCCGTGTGCATGTCCTTTGCGCCTGTGACCACCACCCCAACAMCCATCCAGTGACMACC

AGTGACATMTATATTCTATTTTTATACTCTTCCTATTTTTGTAGTGACCTGTTTATGAGATGCTGGTTTTCTACCCAACGGCCCTGCAGCCAGCTCAC GTCCAGGTTCMCCCACAGCTACTTGGTTTGTGTTCTTCTTCATATTCTAAAACCATTCCATTTCCAAGCACTTTCAGTCCAATAGGTGTAGGAAATAG CGCTGTTTTTGTTGTGTGTGCAGGGAGGGCAGTTTTCTAATGGAATGGTTTGGGAATATCCATGTACTTGTTTGCAAGCAGGACTTTGAGGCMGTGTG GGCCACTGTGGTGGCAGTGGAGGTGGGGTGTTTGGGAGGCTGCGTGCCAGTCAAGAAGAAAAAGGTTTGCATTCTCACATTGCCAGGATGATAAGTTCA

ATCCAGTGGAGGTTTGTCGGGCACCAGCCAGCGTAGCAGGGTCGGGAAAGGCCACCTGTCCCACTCCTACGATACGCTACTATAAAGAGAAGACGAMT

ACTGAGCCTTAAACWIGCAAGTTTTTTATTAAGGGTTTCAAGAGGGGAGGGGGTGTGAGAACATGGAGTAGAGCTCATGCTTCAAAGG~AAAAAACAG AACAAAGA TCACTTGCTTCTGAGGGAACAGGAGAAAAGGCAACACAGAACTACTGA TAAGGGTCCATGTTCAGCGGTGCACGT4 TTATCTTGATAAACA TCTTGAACAAAAAA TAGGGAAACGTCTGACCACAAA TTCACCAGGGTGGAGTTTTTCCCCACCGTAGTAAGCCTGAGGGTACTGCAGGAGATCAGGGCG TATCTCAGTCCTTATCTCAACCACGTAAGACAGACATTCCCAGAGCAGCTGTTTATAGACCTCTCCCCAGGAATGCATTCCTTTCCCAGGGTATTAATA TTAA TATTGCTTGCTAAC;GAAAAGAATTTAGCGATATCCTCCCTACTTGCATGTCCTTTTATAW;CTCTCTGCAAGAAGAAAAATATGGCTTTTTTTGCC TGACCCTGCAGGCAGTCAGACCTTATGGTTGTCTTCCCTTGTTCCCTAAAATCGCTGTTATTCTGTTTTTTCTCAAGGTGCACTGA TTTCATATTGTTC AAACACACGTTTTACAATCAATTTGTACAGTTAACACAATTATCACGGTGGTCCTGAGGTGACATACATCCTCAGCTTACGAAGATAAWIGGATTAAGA GATGAGAGTAAGACAGGCGTAAGAAATTATAAAAGTATTAATTT

98

19 7 17

296 50

39 5 83

494 116

593 149

692 182

79 1 215

89 0 248

989 28 1

1088 295

1187 1286 1385 1484 1583 1682

1880 1781

1979 2078 2177 2276 2375 2474 2573 2672 2771 2870 2969 3013

Fig 3. Sequence of the 3-kb CCNDl aberrant transcript observed in case no. 39. The 295-amino acid coding is shown. The MER1 1 nucleotide sequence is in italics. *Position of the base changes (nucleotides 870 and 1100) compared with the normal CCNDl cDNA sequence." The sequence reported here has been deposited in the EMBL data base (accession no. 223022).

of the rearranged bands cohybridized to the immunoglobulin JH probe (data not shown). Probes D and E did not detect any rearranged fragment in these cases, although they de- tected the same germline bands, as did probes B and C in BamHI- and Sad-cleaved DNA. This led us to postulate that, in these cases, the genomic region corresponding to probes D and E was deleted, and that the break occurred between the stop codon and the 5' end of probe D.

Interestingly, all of these tumors (case no. 39 included) also demonstrated bcl-l rearrangement with a break in the MTC (Fig 6), each rearrangement being identified on two or more restriction digests. We are indeed unable to specify if the two breaks occurred on one or on both alleles.

Northern blot analysis showed that the steady-state level of the CCNDl transcripts is dramatically elevated in the samples showing complex rearrangements, and that the 1 S- kb transcript was predominantly represented, the normal 4.5- kb mRNA being undetectable (Fig 2 ) . When material was

available, Western blot analysis using a polyclonal rabbit anti-cyclin D antibody confirmed, at the protein level, that CCNDI is really expressed in these pathological samples (Fig 7). In fact, a protein with an expected size of 35-kD was detected in cases no. 39 and 40 and in normal human fibroblasts used as positive control. The CCNDI protein was also present in three other cases of MCL (cases no. 9, 19, and 22), but was undetectable in normal lymphoid tissues.

To determine if the high expression of CCNDI in lymphoid malignancies with break in the 3' UTR was a consequence of an increase in stability of its transcript, we assessed CCNDI mRNA half-life in Rec-l and Gob. cell lines, which exclusively express short transcripts, and in the XG5 cell line, where both the 1.5-kb and the 4.5-kb tran- scripts are detected, using the transcriptional inhibitor actino- mycin D. It is to be noted that the CCNDI 3' UTR is not rearranged in the XG5 cell line (data not shown). In normal tissue, the estimated CCNDI mRNA half-life is 0.5 hours.'"

REARRANGEMENT OF CCND1 3' UNTRANSLATED REGION

B H c 39 c 39 -23

W - 9.4

- 6.7

B H c 39 c 39 C

- 23

e - 9.4

- 6.7

I.

probe D probe E Fig 4. Southern blot analysis of CCNDI3' end rearrangement in

case no. 39. DNA from case no. 39 and from normal peripheral blood leukocytes (C) was digested with the indicated enzyme (B, BamHI; H, Hindlll), and processed for Southern blot analysis as described in the Methods. The filters were hybridized to probe D and E. The scale is in kilobases.

I n XGS. the half-lives of the I .S- and 4.5-kb transcripts are approxitnately 3 hours each: in Rcc-l and Gob.. the short transcripts appeared t o he even more stable. with a half- life greater than S hours (Fig X). Indeed. n o comparison can he macle with normal B-lymphoid tissues. which do n o t express CCNIII. This showed that increase in mRNA stability could constitute one of the Inechanisms by which CCNIII is deregulated i n MCL and t( I Iql3)-associated leukemias.

B S c 4 c 4 0 s

C 40 C 40 -23 -23

L S 0

W U

c 34 c 34 - 4, -9.4 * - '. - 9 4

-4.3 .4.3

probe C . . L - 23 - 2 z.3

z probe B

probe B

Fig 5. Southern blot analysis of rearrangement of the CCNDl 3' end in MCL and tlllql3)-associated leukemias. DNA, extracted from the pathological samples and from normal peripheral blood leuko- cytes (C), was digested with the indicated enzymes (S, Sad; B, BamHI) and processed for Southern blot analysis as described in the Methods. The filters were hybridized to probes C and B. The scale is in kilobases.

3693

4 39 3440

- 9.4

Fig 6. MTC rearrangement in MCL or leukemias with a break in the 3' end of CCNDI. DNA extracted from the pathological samples and from normal peripheral blood leukocytes (C) was digested with BarnHl endonuclease and processed for Southern blot analysis as described in the Methods. The filter was hybridized to a MTC probe. The scale is in kilobases.

DISCUSSION

The rcsults presented here constitute additional eletnents in favor of the role o f CCNIII i n lymphoid neoplasia antl provide so~ne clues with regard to its activation mcchanisms.

Cloning and sequence analysis o f CCNDI cDNAs in ;I

case of R-CLL showed that the coding franw \vas retained and that the different s ix s of CCNIII mRNA resulted from different 3' untranslated structures. I n this Ieuketnia. the CCNIII gene was expressed primarily a s ;I I.5-kb transcript. but two aberrant transcripts o f 2. antl 3 kb were also detected. The absence o f ;I recognizable polyxlcnylation site and of ;I poly(A) tail in the sequence o f I 1 cDNA clones covering less than I .SO0 hp did not allow 11s to specify the nature of the shortest transcript. I t may correspond. ;IS i t has been previously rcpor~cd.'~.' ' ' t o ;I truncntcd form of the normal 4.5-kh mRNA through the use of different polyaclcn!.latiol~ sign;ds. Sequence analysis showed that the 3-kb aberrant transcript results from the juxtaposition of a repetitive se- quence of the MER1 I family to the CCNDI coding sc- qucncc. This might first represent a n example of insertional mutagenesis lending to gcnc deregulation. ;IS it has been

- 35 kDa

Fig 7. Western blot analysis of the CCNDI protein in MCL and t(llql3)-associated leukemias. Proteins were extracted from patho- logical samples, normal tonsil IT), reactive lymph node (L), Ramos cell line (R), and normal human fibroblasts (F). lmmunodetection of the CCNDl protein was performed as described in the Methods.

3694

0 1 2 3 4 5

ccND1: G

MYC: G

CCND1: R

MYC: R

ccND1: x

MYC X

0 "

Hours

-3m

"2kb

-15kb

~ 1 5 k b

-4.5 kb

-3kb

Fig 8. Study of CCNDl mRNAstabilii in t(llql3l-bearing cell line: Rec. (RI, Gob. (G), and XG5 (X) cell lines. Total RNA was extracted at the indicated times after addition of actinomycin D and processed for Northern blot analysis as described in the Methods. The filters were sequentially hybridized to a CCNDl probe (probe A) and to a MYC exon 3 probe. The size of CCNDl transcripts is in kilobases.

described when proviral or human repetitive sequence are integrated in the vicinity of certain genes3' However, we have to consider the possibility that this aberrant transcript is the consequence of the t( l I ; 19)(q13;q13), which charac- terized this leukemia, and that the MER1 I sequence origi- nates from the chromosome 19.

The continuation of this work led us to show that re- arrangement of the 3' end of CCNDl was not a rare event in MCL and in t( 1 Iql3)-bearing leukemias. Using a set of probes specific to the different parts of CCNDI. such re- arrangements were observed in DNA from four of the 40 patients tested. Consistent with previously published works,'5.'" Southern blot and sequence analyses indicated that, as a result of these rearrangements, the greatest part of

RIMOKH ET AL

the CCNDl 3' UTR and. in particular. the AU-rich region containing the mRNA-destabilizing signals AUUUA. is eliminated. I t is thus possible that the CCNDI 3' UTR corre- sponds to a new breakpoint cluster (mTC3).

The finding that the half-life of CCNDI mRNA is in- creased in three cell lines carrying a t( I Iq13) translocation stresses the importance of posttranscriptional derangement in activation of CCNDI. The fact that in the XGS cell line the half-lives of the 4.5-kb and 1 .S-kb CCNDI mRNAs are identical allows us to rule out the hypothesis that the AU- rich region contained in the 3' UTR. and deleted in some tumors, is responsible for the greater stability of the truncated messages. However. it must be pointed out that the half-life of the CCNDl short mRNA is significantly longer in tumors with 3' UTR rearrangement and. thus. that these re- arrangements may alter KNA stability. There exist other situ- ations where certain genes implicated in the control of cell proliferation are deregulated through stabilization of their transcripts. So. the inappropriate expression of the M Y C proto-oncogene in Burkitt's lymphoma can result from an increase of the half-life of its transcript.""' However. in this case, the sequences implicated in the control of the mRNA stability are not only localized in the 3' UTR. but also in the S' region encompassing the first exon of MYC. I t has also been reported that the mRNA stability of different cyto- kines and cyclins G, is increased in some hematopoietic and solid tumors."'.2J without gross genetic abnormalities of the transcripts. From these data. it appears that the AU-rich se- quences present in the 3' UTR of certain genes are not always implicated in mRNA stability and that other mechanisms should exist. Their nature remains to be determined. Com- parison of the sequences of the rearranged 3' UTR with those of the short 1 .S-kb transcripts observed in most of the MCLs and of the normal 4.5-kb mRNA should provide data information with regard to the regions putatively involved in mRNA stability.

AUUUA motifs in the 3' UTR of RNAs can also help to modulate the efficient translation of these transcripts. In particular. cytokine-derived UA-rich sequence can be re- sponsible for a translational blockade in ~itro.~' .~ ' Consistent with this hypothesis is the finding that in two cases of tumors (cases no. 39 and 40) with deletion of the CCNDI mRNA AU-rich regions, the CCNDI protein is expressed at a high level. Finally, the fact that in most of the MCL and t( 1 lq13)- associated leukemias analyzed. CCNDI is primarily ex- pressed as a I .5-kb transcript".'" prompts us to pinpoint the importance of the translational control of CCNDl expression in the pathogenesis of these tumors. It remained to be deter- mined why a short transcript is also produced in some tumors without detectable rearrangement of the CCNDI 3' UTR.

All cases presented here demonstrated rearrangement of both the MTC region and the CCNDI 3' UTR. This observa- tion indicates that, in some cases. activation of CCNDI may be the consequence of two independent genetic events. namely, the juxtaposition of the CCNDI promoting region to IgH enhancer in the t( I 1 ; 14) translocation and the loss of 3' end regulatory sequences.

Nucleotide changes leading to amino acid substitutions in

REARRANGEMENT OF CCNDl 3‘ UNTRANSLATED REGION 3695

the sequence of the CCNDl protein have been described in one cell line.” The present report and the finding that in two primary human tumors the CCNDl coding sequence is normalz1 suggest that amino acid changes are not required for CCNDl to develop oncogenic properties in primary human tumors.

The results of this study also provide the first evidence that deregulation of CCNDI in MCLs and t(llql3)-associated leukemias leads to the accumulation of abnormally high lev- els of a normal 35-kD CCNDl protein, which we did not find expressed in normal lymphoid tissues and in three follicular lymphomas (Fig 7). The normal size of the CCNDI protein in two cases of tumor (cases no. 39 and 40) with a break in the 3’ UTR confirms that the rearrangement did not alter the CCNDl coding frame.

ACKNOWLEDGMENT

We gratefully acknowledge Jean-Marie Blanchard and Jacques Samarut for helpful discussion, Peter Griggs for reading the manu- script, and B6nidicte Santaluccia for excellent technical assistance.

REFERENCES I . Solomon E, Borrow J, Goddard AD: Chromosome aberrations

and cancer. Science 254:1153, 1991 2. Rimokh R, Berger F, Corniller P, Wahbi K, Rouault J-P,

Ffrench M, Bryon P-A, Gadoux M, Gentilhomme 0, Germain D, Magaud J-P: Break in the BCL-I locus is closely associated with intermediate lymphocytic lymphoma subtype. Genes Chrom Cancer 2:223, 1990

3. Medeiros LJ, van Krieken JHJM, Jaffe ES, Raffeld M: Associa- tion of BCL- 1 rearrangements with lymphocytic lymphoma of inter- mediate differentiation. Blood 76:2086, 1990

4. Williams ME, Westermann CD, Swerdlow SH: Genotypic characterization of centrocytic lymphoma: Frequent rearrangement of the chromosome 11 BCL-l locus lymphomas. Blood 76:1387, 1990

5. Williams ME, Meeker TC, Swerdlow SH: Rearrangement of the chromosome 11 BCL-1 locus in centrocytic lymphoma: Analysis with multiple breakpoint probes. Blood 78:493, 1991

6. Leroux D, Le Marc’hadour F, Gressin R, Jacob MC, Keddari E, Monteil M, Caillot P, Jalbert P, Sotto JJ: Non-Hodgkin’s lympho- mas with t(l I; 14)(q13;q32): A subset of mantle zonehntermediate lymphocytic lymphoma? Br J Haematol 77:346, 1991

7. Raffeld M, Jaffe ES: Bcl-l, t(ll;l4), and mantle cell-derived lymphomas. Blood 78:259, 1991

8. Banks PM, Chan J, Cleary ML, Delsol G, De Wolf-Peeters C, Gatter K, Grogan TM, Harris NL, Isaacson PG, Jaffe ES, Mason D, Pileri S, Ralfkiaer E, Stein H, Warnke RA: Mantle cell lymphoma. Am J Surg Pathol 16:637, 1992

9. Tsujimoto Y, Jaffe E, Cossman J, Gorham J, Nowell PC, Croce CM: Clustering of breakpoints on chromosome 11 in human B- cell neoplasms with the t(l1; 14) chromosome translocation. Nature 315:340, 1985

10. Inaba T, Matsushime H, Valentine M, Roussel MF, Sherr CJ, Look AT: Genomic organisation, chromosomal localization, and independent expression of human cyclin D genes. Genomics 13565, 1991

11. Rimokh R, Berger F, Delsol G, Charrin C, BerthLas MF, Ffrench M, Garoscio M, Felman P, Coiffier C, Bryon PA, Rochet M, Gentilhomme 0, Germain D, Magaud JP: Rearrangement and overexpression of the BCL-I/PRAD-I gene in intermediate lympho-

cytic lymphomas and in t(llql3)-bearing leukemias. Blood 81:3063, 1993

12. Meeker TC, Sellers W, Harvey R, Withers D, Carye K, Xiao H, Block A-M, Dadey B, Han T: Cloning of the t(I I ; 14)(q13;q32) translocation breakpoints from two human leukemia cell lines. Leu- kemia 5:733, 1991

13. Williams ME, Swerdlow SH, Rosenberg CL, Arnold A: Char- acterization of chromosome 11 translocation breakpoints at the BCLl and PRADl loci in centrocytic lymphoma. Cancer Res 52:5541s, 1992 (suppl)

14. Mokotura T, Bloom T, Kim HG, Juppner H, Ruderman JV, Kronenberg I”, Arnold A: A novel cyclin encoded by a BCLl linked candidate oncogene. Nature 350:512, 1991

15. Withers DA, Harvey RC, Faust JB, Melnyk 0, Carey K, Meeker TC: Characterization of a candidate BCLl gene. Mol Cell Biol 11:4846, 1991

16. Xiong Y, Conolly T, Futcher B, Beach D: Human D-type cyclin. Cell 65:691, 1991

17. Sherr CJ: Mammalian G1 cyclins. Cell 73:1059, 1993 18. Rosenberg CL, Wong E, Petty EM, Bale AE, Tsujimoto Y,

Hams NL, Arnold A: PRAD1, a candidate BCLl oncogene: Map- ping and expression in centrocytic lymphoma. Proc Natl Acad Sci USA 88:9638, 1991

19. Set0 M, Yamamoto K, Iida S, Aka0 Y, Utsumi R, Kubonishi I, Miyoshi I, Ohtsuki T, Yawata Y, Namba M, Motokura T, Arnold A, Takahashi T, Ueda R Gene rearrangement and overexpression of PRADl in lymphoid malignancy with t(l1; 14)(q13;q32) translo- cation. Oncogene 7:1401, 1992

20. Raynaud SD, Bekri S, Leroux D, Grosgeorge J, Klein B, Bastard C, Gaudray P, Simon MP: Expanded range of llq13 breakpoints with differing patterns of cyclin D1 expression in B- cell malignancies. Genes Chrom Cancer 8:80, 1993

21. Rosenberg CL, Motokura T, Kronenberg HM, Arnold A: Coding sequence of the overexpressed transcript of the putative oncogene PRADlkyclinDI in two primary human tumors. Onco- gene 8519, 1993

22. Weisenburger DD, Warren G, Sanger WG, Armitage C O , Purtilo DT: Intermediate lymphocytic lymphoma: immunopheno- typic and cytogenetic findings. Blood 69:1617, 1987

23. Jaffe ES, Bookman MA, Longo DL: Lymphocytic lymphoma of intermediate differentiation-mantle zone lymphoma: A distinct subtype of B-cell lymphoma. Hum Path01 18:877, 1987

24. Bookman MA, Lardelli P, Jaffe ES, Duffey PL, Longo DL: Lymphocytic lymphoma of intermediate differentiation: Morpho- logic, immunophenotypic, and prognostic factors. J Natl Cancer Inst 82:742, 1990

25. Perry DA, Bast MA, Armitage JO, Weinsenburger DD: Dif- fuse intermediate lymphocytic lymphoma: A clinicopathologic study and comparison with small lymphocytic lymphoma and diffuse small cleaved cell lymphoma. Cancer 66:1995, 1990

26. Brochier J, Magaud JP, Cordier G: Heterogeneity of human B lymphocytes as revealed by monoclonal antibodies. Ann Immunol 133D:283, 1984

27. Rimokh R, Magaud JP, Coiffier B, Samarut J, Germain D, Mason DY: Translocation involving a specific breakpoint (5q35) on chromosome 5 is characteristic of anaplastic large cell lymphomas. Br J Hematol 71:36, 1989

28. Kaplan DJ, Jurka J, Solus JF, Duncan CH: Medium reiteration frequency repetitive sequences in the human genome. Nucleic Acids Res 19:4731, 1991

29. Jurka J, Walichiewicz J, Milosavljevic: Prototypic sequences for human repetitive DNA. J Mol Evol 35:286, 1992

3696 RIMOKH ET AL

30. Keyomarsi K, Pardee AB: Redundant overexpression and gene amplification in breast cancer cells. Proc Natl Acad Sci USA 90: 1 I 12, 1993

3 I . Lammie GA, Smith R, Silver J, Brookes S, Dickson C, Peters G: Proviral insertions near cyclin D1 in mouse lymphomas: A paral- lel for BCLl translocations in human B-cell neoplasms. Oncogene 7:2381, 1992

32. Rabbits PH, Forster A, Stinson MA, Rabbits TH: Truncation of exon I from the c-myc gene results in prolonged c-myc mRNA stability. EMBO i 4:3727, 1985

33. Eick D, Piechaczyk M, Henglein B, Blanchard JM, Traub B, Kofler E, Weist S, Lenoir GM, Bornkamm GW: Aberrant c-myc RNAs of Burkitt's lymphoma cells have longer half-lifes. EMBO J 4:3717, 1985

34. Ross HJ, Sato N. Ueyama Y, Koeffler P: Cyto!cine messenger RNA stability is enhanced in tumor cells. Blood 77: 1787. 199 I

35. Shaw G, Kamen R: A conserved AU sequence from the 3' untranslated region of GM-CSF mRNA mediates selective mRNA degradation. Cell 46:659, 1986

36. Akashi M, Shaw G, Gross M, Saito M, Koeffler HP: Role of AUUU sequences in stabilization of granulocyte-macrophage col- ony-stimulating factor RNA in stimulated cells. Blood 78:2005, 1991

37. Kruys V, Marinx 0, Shaw G. Deschamps J , Huez G: Transla- tional blockade imposed by cytokine derived UA-rich sequences. Science 245:852, 1989

38. Han J, Brown T, Beutler B: Endotoxin-responsive sequences control cachectidtumor necrosis factor biosynthesis at translational level. J Exp Med 171 :465, 1990