Non-Hodgkin's Lymphomas' - Cancer Research · ICANCER RESEARCH 58. 1762-1767. April 15. 19981 ABSTRACT cdc2SA, cdc2SB, and cdc2SC are a family of human phosphatases that activate

ICANCER RESEARCH 58. 1762-1767. April 15. 19981

ABSTRACT

cdc2SA, cdc2SB, and cdc2SC are a family of human phosphatases thatactivate the cyclin-dependent kinases at different points of the cell cycle.cdc2SA and cdc2SB have been shown to have oncogenic potential, and theyhave been identified as transcriptional targets of c-myc. To determine therole of cdc25 genes in the pathogenesis of human lymphomas and theirpossible correlation with c-myc deregulation, we have analyzed the expression of cdc2SA, cdc2SB, and cdc25C and c-myc genes in a series of 63non-Hodgkin's Lymphomas and 8 nonneoplastic lymphoid tissues. The

mRNA levels of the three phosphatases in the nonneoplastic tissues werenegative or negligible. cdc2SB overexpression was detected in 35 tumors(56%). This overexpression was more frequently found in aggressive(81%) than In indolent lymphomas (36%; P < 0.01). cdc2SB overexpres

sion was also significantly associated with a higher proliferative activity ofthe tumors. No cdc25B gene amplification or rearrangements were detected by Southern blot analysis. A bialleic EcoRI polymorphism ofcdc25B gene was identified with a similar distribution in patients withlymphoma and in a normal population. cdc2SA was overexpressed in threeaggressive lymphomas. No detectable cdc25C mRNA levels were seen inany of the tumors. c-myc was overexpressed in 43% of tumors, and itcorrelated significantly with the presence ofcdc25B up-regulation. Twenty-six of 35 (74%) lymphomas with high levels of cdc2SB mRNA alsoshowed c-myc overexpression, whereas 27 of 28 (%%) tumors without

detectable or with very low cdc2SB expression also had undetectable c-myclevels (P < 0.0001). In addition, a significant linear correlation was found

between the cdc2SB and c-myc mRNA levels (r = 0.575, P < 0.001). Thesefindings suggest that cdc25B overexpression in non-Hodkin's lymphomamay participate in the pathogenesis of aggressive variants, and it maycooperate with c-myc oncogene in the development of these tumors.

INTRODUCTION

cdc25A, cdc25B, and cdc25C are three CDK4-activating phosphatases (1 , 2) with dephosphorylating activity of Thr- 14 and Tyr-l5 CDK

residues, which seem to act in different points of the cell cycle (3, 4).Particularly, cdc25A and cdc25B may play a role in early G1 and G1-Stransition, whereas cdc25C seems to be mainly active in G2 (1).cdc25A and cdc25B have been shown to have transforming potentialin cooperation with other oncogenic factors (5). In addition, cdc25A isstrongly induced by the adenovirus E1A oncoprotein (6), indicatingthat cdc25 may participate in the transforming pathway of this onco

Received 1/5/98; accepted 3/4/98.Thecostsof publicationof thisarticleweredefrayedin partby thepaymentof page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

I This work was supported by Comision Internainisterial de Ciencia y Tecnologia

(CICYT), Ministerio de Educación y Cultura Grant SAP 96/61; Marato-TV3 Cancer Grant95/3010D, FundaciónCientifica de Ia AsociaciónEspanola contra el Cancer, and Generalitat de Catalunya Grant SGR1W96. S. H. was a fellow supported by ComissióInterdepartamental de Recerca i InnovacióTecnologica (CIRIT), Generalitat de Catalunya; S. B. and M. C. were fellows supported by the Spanish Ministerio de EducaciónyCultura; and L. H. was a fellow supported by MaratO-TV3Cancer and Fundacid Rius iVirgili.

2The first two authors contributed equally to this study.3 To whom requests for reprints should be addressed, at Laboratory of Pathology,

Hospital Clinic, Villarroel 170, 08036 Barcelona, Spain. Phone: 343-2275450; Fax:343-2275717; E-mail: [email protected].

4 The abbreviations used are: CDK, cyclin-dependent kinase; CLL, chronic lympho

cytic leukemia.

protein. cdc25A and cdc25B are also overexpressed in some cancercell lines (7) and human breast (5) and head and neck (8) carcinomas,suggesting that these phosphatases may be implicated in the development of human tumors. The expression patterns of these cell cycleactivating phosphatases in human lymphoproliferative disorders andtheir possible contribution to the pathogenesis of these tumors is notknown. In addition, the mechanisms leading to cdc25 deregulation inhuman tumors are not clear.

The c-myc proto-oncogene is a transcription factor (9—11) that actsin conjunction with its partner Max in the regulation of genes implicated in the normal cell proliferation and apoptosis (12, 13). Mutations and overexpression of this gene are involved in the pathogenesisof several neoplasms, including human non-Hodgkin's lymphomas(14, 15). However, few transcriptional targets of c-myc with a potential oncogenic activity have been identified (14, 16). cdc25A andcdc25B have been recently described as direct transcriptional targetsof c-myc in murine fibroblast cell lines (17). The transcriptionalregulation of cdc25A and cdc25B by c-myc would link the oncogenicpotential of this gene with the activation of the cell cycle progression.However, there is no information about the expression of cdc25 genesin relation to c-myc expression in other in vivo cellular systems,especially in human tumors. Thus, in this study, we have investigatedthe hypothesis that cdc25 phosphatases may be involved in the pathogenesis of human non-Hodgkin's lymphomas and their possible association with c-myc activation in these tumors, where c-myc isfrequently deregulated (18, 19). Our results show that cdc25B is themain CDK-activating phosphatase overexpressed in these tumors andthat its expression was not due to amplification or rearrangement ofthe gene but correlated significantly with the c-myc mRNA levels.Overexpression of both genes was also significantly associated withthe more proliferative and clinically aggressive variants of theselymphomas. These findings suggest that cdc25B deregulation mayparticipate in the development of aggressive non-Hodgkin's lymphomas and that it may cooperate with c-myc in the pathogenesis of thesetumors.

MATERIALS AND METHODS

Patients and Tissues. Tumor specimens from 63 non-Hodgkin's lymphomas were selected from the files of the Hematopathology Section of theHospital Clinic of Barcelona, based on the availability of frozen tissue for

molecular analysis. These cases were diagnosed following the Revised Euro

pean-American Classification of Lymphoid Neoplasms (20). The tumors werefurther classified as indolent or low-grade (36 cases) and aggressive or highgrade (27 cases) lymphomas. The 36 indolent lymphomas consisted of 13CLLs, 3 hairy cell leukemias, 12 follicular cell lymphomas (grade I and H), 7typical mantle cell lymphomas, and 1 lymphoplasmacytoid lymphoma. The

mantle cell lymphomas were included in the indolent group because all of thecases examined had a low proliferative activity (less than 1 mitosis per

high-power field). The 27 aggressive lymphomas included 6 large cell lymphomas evolved from a CLL (Richter's syndrome), 4 large cell lymphomasprogressed from follicular lymphomas, 13 de novo diffuse B-cell large celllymphomas, 2 Burkitt's lymphomas, 1 B-cell lymphoblastic lymphoma, and 1anaplastic large cell lymphoma of null phenotype. Eight samples of nonneoplastic lymphoid tissues, including three reactive lymph nodes, three tonsils,

1762

cdc25 Cell Cycle-activating Phosphatases and c-myc Expression in Human

Non-Hodgkin's Lymphomas'

Silvia Hernández,2 Luis Hernández,2 Sflvia Beà, Maite Cazorla, Pedro L. Fernández, Alfons Nadal, Jaume Muntané,

Carme Mallofré, Emilio Montserrat, Antonio Cardesa, and Elias Campo3

Departments of Pathology [S. H., L H., S. B., M. C.. P. L F.. J. M., C. M.. A. C., E. C.) and Hematology [E. MI, Hematopathology Section, Hospital Cl(nic, and Casa deMaternitat (A. N.). Institut de Investigacions Biomèdiques Agust( P1 i Sunyer, University of Barcelona, 08036 Barcelona. Spain

on June 3, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Celllinescdc2SAcdc2SBcdc25Cc-mycU937+++++K562-———/+HSB2+++++Raji-I.+-+Daudi++++—++Molt-4++++++++

cdc25 IN NON-HODGKIN'S LYMPHOMAS

Northern blot in nonneoplastic lymphoid tissues, including three reactive lymph nodes, three tonsils, and two spleens and in six hematological cancer cell lines. The mRNA levels of the three phosphatasesin all of the nonneoplastics tissues were negative or negligible,whereas the expression in the cell lines was variable (Table 1).Molt-4, HSB2, and U937 cells showed detectable mRNA levels of alltypes of cdc25, although the levels in Molt-4 were higher than inHSB2 and U937. Raji and Daudi showed detectable expression ofcdc2SA and cdc25B but not cdc25C, and the mRNA levels in Daudiwere higher than those in Raji. No expression of the cdc25 genes weredetected in K562 cell line. Total RNA from Molt-4 and K562 wereused as positive and negative controls in the study of human lymphoidtumors (Fig. 1).

cdc25A, cdc25B, and cdc2SC Expression in Human NonHodgldn's Lymphomas. To know the possible implication of thesecell cycle phosphatases in human non-Hodgkin's lymphomas,cdc25A, cdc25B, and cdc25C expression was examined in a series of63 lymphomas, including tumors with different histology and biological behavior. Overexpression of cdc25B was detected in 35 tumors(56%; Fig. 1). This overexpressionwas more frequentlyfound inaggressive lymphomas (81%) than it was in low-grade tumors (36%;P < 0.01, two-tailed exact Fisher test; Table 2). Nevertheless, theexpression level within aggressive and indolent lymphomas was vanable. In the aggressive tumors, cdc25B overexpression was found in70% of de novo diffuse large cell lymphomas and in all samples oflymphoblastic lymphoma, Burkitt's lymphoma, large cell lymphomasprogressed from CLL (Richter's syndrome), and large cell lymphomasprogressed from follicular center cell lymphomas. In the low-gradelymphomas, cdc25B overexpression was found in 69% of the CLLs,25% of follicular center cell lymphomas, and 14% of mantle celllymphomas. No expression was detected in the hairy cell leukemiasand lymphoplasmacytoid lymphoma. The relatively high number oftypical CLLs overexpressing cdc25B gene may be related to theadvanced stage of the tumors included in this study because cdc25overexpression was only detected in one of the (20%) stage A CLLSbut it was detected in the eight (100%) stage B and C tumors (Table2).

The proliferative activity could be analyzed by flow cytometry in33 cases in which additional tissue was available. cdc25B overexpression was significantly associated with higher cell proliferation of thetumors because 74% lymphomas overexpressing cdc25B showed ahigh S-phase fraction, whereas it was present only in 20% tumorswithout cdc25B overexpression (P < 0.01 two-tailed exact Fischertest). No detectable expression of cdc2SC was seen in any of thecases, and significant levels of cdc25A mRNA were observed in threelymphomas (Fig. 1). These tumors were three aggressive lymphomascorresponding to a large cell lymphoma transformed from a CLL(Richter's syndrome), a large cell lymphoma evolved from a follicularlymphoma, and a Burkitt's lymphoma. In these three cases, highlevels of cdc25B and c-myc mRNA were also present.

Southern Blot Analysis. To determine whether cdc25B mRNAoverexpression in human lymphomas was associated with amplifications or rearrangements of the gene, genomic DNA from 27 lympho

Table I cdc25 and c-myc mRNAoverexpression in some hematological cancercell lines

and two spleens were also examined. Six hematological cancer cell lines werealso examined (Molt-4, K562, U937, Daudi, Rail, and HSB2). These cell lineswere used as controls in the study of human tumors and were purchased fromthe American Type Culture Collection. Cells were grown in RPMI (LifeTechnologies, Inc.) containing 15%FCS in a 5% CO2 humidified atmosphere.

RNA Extraction and Northern Blot Analysis. Total RNA was isolatedfrom frozen tissues by guanidine isothiocyanate extraction and cesium chloridegradient centrifugation, as described previously (21). Fifteen @gof total RNAwere electrophoresed on a denaturing 1.2% agarose formaldehyde gel andtransferred to Hybond-N membranes (Amersham, Buckinghamshire, UnitedKingdom). The membranes were prehybridized, hybridized with the differentprobes, and washed as described previously (22). Probes were radiolabeledusing a random primer DNA labeling kit (Promega Corp., Madison, WI) with[cr-32P]dCTP.The c-myc probe was a 1.4-kb Clal/EcoRI fragment containingthe third coding exon, and it was provided by R. Dalla Favera (ColumbiaUniversity, New York, NY). The cdc25 probes were a 1.6-kb NotI/NotIfragment containing the complete open reading frame of cdc25A, two 1.5-kbEcoRUEcoRI fragments of cdc25B, and the 2.2-kb BamHl/XhoI fragment,including the complete open reading frame of cdc25C. The cdc25A, cdc2SB,and cdc25C probes were provided by Drs. D. Beach and K. Galaktionov(Howard Hughes Medical Institute, Cold Spring Harbor, NY). Intensities ofc-myc and cdc25 bands were normalized to 28S RNA band. The signals werequantified using a UVP-5000 video densitometer (UVP, San Gabriel, CA).Overexpression was considered when the normalized signal of the tumor wasabove 0.5 relative units. In each blot, total RNA from Molt-4 and K562 celllines was also included as a positive and negative control, respectively. Thepresence of these cell lines in all of the blots allowed the interblot comparisonof the signals.

DNA Extraction and Southern Blot Analysis. Genomic DNA was cxtmcted from 27 frozen lymphomas and nonneoplastic tissues. In addition,genomic DNA was also obtained from blood cells in 56 cases of normal blooddonors. DNA was extracted using proteinase K/RNase treatment and phenolchloroform extraction, and 15 @tgwere digested with EcoRI and Hindffl,separated on 0.8% agarose gels, and transferred to Hybond-N membranes(Amersham). The membranes were prehybridized and hybridized with cdc25Band c-myc, as described previously (22). To normalize the loading, the blotswere hybridized with @-actinprobes, after the cdc25B and c-myc probes werestripped off. Intensities of cdc25B bands were normalized to (3-actin control.The signals of the Southern blot analysis were quantified with the same systemused in the Northern blot analysis.

Flow Cytometry. Proliferative parameters of the tumors were studied byflow cytometryin 33 cases (21 low-gradeand 12 high-gradelymphomas).Theanalysis was performed on 50-mm-thick sections from formalin-fixed, paraffin-embedded tissues using a technique described previously (23). The nuclearsamples were stained with propidium iodide and analyzed with an EpicsProfile II flow cytometer (Coulter Co., Hialeah, FL). Nonneoplastic cells in thesection under study were used as the internal standard of the diploid channel,as recommended by the DNA Cytometry Consensus Conference (24). Disciimination of doublets and aggregates was performed on the basis of the pulsepeak versus pulse area analysis with the Power Pack system (Coulter). Singleparameter histograms analyzed with the Multicycle software (Phoenix FlowSystems, San Diego, CA) were classified according to the guidelines onnomenclature of the Society for Analytical Cytology (25). S phase of thetumors was analyzed using the Power Pack doublet discrimination system(Coulter). The cutoff for a high S phase was considered at 5% of total cellpopulation.

Statistical Analysis. To determine the correlation between the expressionlevels of cdc25B and c-myc and the association of those with the aggressivelymphomas, a linear regression analysis was performed by means of thePearson product-moment correlation. Data were analyzed with the BMDPstatistical software package (BMDP Statistical Software, Los Angeles, CA).Flow cytometry data were analyzed with the same statistical software packagethat was used for the expression analysis.

RESULTS

cdc2SA, cdc25B, and cdc2SC Expression in NonneoplasticLympholdTissuesandHematologicalCancerCellLines. cdc25A,cdc25B, and cdc2SC mRNA expression was initially analyzed by

1763

on June 3, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

. @[ I .1

Table 2 cdc25 and c-myc mRNAoverexpression in non-Hodgkin‘slymphomasmRNA

overexpression

Diagnosis n cdc25A cdc2SB―c-myc―Indolent

36 0 13 (36%) 8(22%)SLLC/CLL 13 0 9 (69%) 4(30%)

StageA 5 0 1(20%) 0StagesBandC 8 0 8(100%) 4(50%)

Lymphoplasmacytoid lymphoma 1 0 0 0Hairy cell leukemia 3 0 0 0Follicular lymphoma 12 0 3 (25%) 3 (25%)Mantle cell lymphoma 7 0 1 (14%) 1 (14%)

Aggressive 27 3 (12%) 22 (81%) 18(70%)Large cell lymphoma 13 0 9 (70%) 7 (54%)Follicular lymphoma —@large cell lymphoma 4 1 4 (100%) 4(100%)Richter's syndrome 6 1 6 (100%) 5(83.3%)Burkitt'slymphoma 2 1 2(100%) 2(100%)

Lymphoblastic lymphoma 1 0 1 (100%) 1 (100%)Anaplastic large cell lymphoma I 0 0 0

Total 63 3 (5%) 35 (56%) 27 (43%)

cdc25 IN NON-HODGIUN'SLYMPHOMAS

levels were seen in nonneoplastic lymphoid tissues. In the cell lines,high levels of c-myc mRNA were found in U937, Molt-4, HSB2, Raji,and Daudi cells. Only K562 showed no c-myc detectable expression.c-myc overexpression was detected in 43% of the non-Hodgkin'slymphomas (27 of 63), and it was more frequently found in aggressive(70%) than in indolent (22%) types of tumors (P < 0.001, two-tailedexact Fisher test). The overexpression incidence in each type oflymphoma was similar to cdc25B up-regulation (Table 2). c-mycoverexpression was not due to gene amplifications or rearrangementsin most of the tumors because no gene amplification was observed inany case, and gene rearrangement was only detected in the twoBurkiu's lymphomas included in the study.

The finding of c-myc overexpression correlated significanfly withthe presence of cdc25B up-regulation (Table 3; Figs. 1 and 3). Twenty-six of 35 (74.3%) lymphomas with high levels of cdc25B mRNAalso showed c-myc overexpression, whereas 27 of 28 (96.4%) tumorswithout detectable or very low cdc25B expression also had undetectable c-myc levels (P < 0.0001, two-tailed exact Fisher test). Discordance between c-myc and cdc25B expression was found in 10 cases.Of note, six of these discordant cases were CLLs (five cases) and onewas a large cell lymphoma progressed from a CLL. In these cases,cdc2SB overexpression was found without high levels of c-myc,suggesting a different regulation of cdc25B expression in this subgroup of lymphomas. The coincidental overexpression of both geneswas significantly more frequent in aggressive than they were inlow-grade lymphomas (Table 2). To further analyze the possiblerelationship between c-myc and cdc25B overexpression, the mRNAlevels of these two genes were compared in all of the cases by a linearregression analysis. This study showed a significant linear correlationbetween these two parameters (r = 0.575, P < 0.001 Pearson productmoment correlation; Fig. 3).

DISCUSSION

cdc25 genes are a family of cell cycle-activating phosphatases thatfunction to remove the inhibitory phosphates of threonine and tyrosineresidues in the AlP-binding sites of the CDKs (26). These phosphatases act at different points of the cell cycle, including G@and the Ga-Stransition (3, 4). Deregulation of these early phases of the cell cycleseems to be an important step in the oncogenic transformation of cells,and several cell cycle-related oncogenes and tumor suppressor geneshave been shown to act in G@or G1-S transition (27, 28). In vitroexperiments using murine fibroblasts have demonstrated the onco

Fig. 1. Northern blot analysis of representative nonneoplastic lymphoid tissue (spleen)and human non-Hodgkin's lymphomas. Molt-4 and K562 cell lines were used as positiveand negative controls, respectively. Fifteen g.tgof total RNA were electrophoresed on a1.2% agarose-fonnaldehyde gel and transferred to nylon membranes. The filters werehybridized to the c-myc, cdc2SA, cdc2SB, and cdc25C probes. 28S RNA is shown asloading control. Tumors with c-myc overexpression also showed high levels of cdc25B.Only one MCL (l885T) with weak overexpression of c-myc did not show cdc25Bup-regulation. Two cases (14760/93 and 1765/95) with strong levels of c-myc showedcdc25A overexpression. No up-regulation of cdc25C was observed in any of the tumors.FCL, follicular cell lymphoma; DLCL, diffuse large B-cell lymphoma; MCL, mantle celllymphoma; BL. Burkitt's lymphoma; FCL)'-DLCL: diffuse large cell lymphoma progressed from follicular cell lymphoma.

mas and 4 normal tissues was examined by Southern blot (Fig. 2). Nogene alterations were observed in any of the tumors, including 22cases with mRNA overexpression, indicating that other mechanismscould be implicated in the up-regulation of this gene. Additionally, theabsence of gene rearrangement in the specific regions defined by thisSouthern blot analysis does not exclude the possibility of rearrangements in other regions that may influence transcriptional regulation.EcoRI detected a two-allele polymorphism of the cdc25B gene withpolymorphic bands at 7.25 (Al allele) and 5.6 kb (A2 allele) and twoadditional constant bands at 8.15 and 4.6 kb. The frequencies of thesetwo alleles in this series of lymphomas were 0.15 and 0.85, respectively. To know the relative frequency of this polymorphism in thenormal population, genomic DNA from 56 unrelated normal Caucasian individuals was also examined. The frequency of the two allelesin the normal population was similar to that observed in the lymphoma patients (Al, 0. 13; A2, 0.87).

Correlation between cdc25B and c-myc Expression. To determine whether the cdc25B overexpression observed in our series ofnon-Hodgkin's lymphomas was related to c-myc deregulation, weanalyzed c-myc mRNA expression in the same series of tumors,normal tissues, and cell lines. Very low or negative c-myc mRNA

a Indolent vs. aggressive tumors, P < 0.001.

b Indolent vs. aggressive, P < 0.001.

â€S̃LL, small lymphocytic lymphoma.

1764

I.... I,C-MYC

CDC2SB

CDC2SA

CDC25C

28S

on June 3, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

._ —,@,v'c'@@@ .,ø@@

...@ I@ • @a.@

@ ..@@ -—@ —@@

cdc2SB overexpressionnc-myc

overexpression+—+3526

(74.3%)9(25.7%)—281(3.6%)27(96.4%)P

< 0.0001.

cdc25 IN NON-HODOKINS LYMPHOMAS

Fig. 2. Southernblot analysisof cdc25Bgene in nonHodgkin's lymphomas. Genomic DNA was digested withEcoRl and hybridized with the cdc25B and f3-actin probes.No amplificationor rearrangementsweredetectedin anytumor,includingthecaseswithoverexpressionof thegene(5372/95,17246/95,15136/94,11313/92,and 2502195).Two polymorphicbands(Al, 7.25 kb; A2, 5.6 kb) wereidentified.Case5752189is homozygousfor the Al allele,cases 3230Tand 14796/95are heterozygous,and the remahsing cases arc homozygous for the A2 allele.

I. .@ Al

@A2

eight (100%) stage B and C tumors. Although the number of cases islimited, these findings suggest that cdc25B overexpression in CLLmay be implicated in the progression of these tumors. Mantle celllymphomas are neoplasms that show variable proliferative activity.Typical variants have a relatively low mitotic index, whereas the moreaggressive blastic variants show a very high proliferative activity

(29—31). Different studies have indicated that the proliferative activityis an important prognostic factor in these tumors (32, 33). The mantlecell lymphomas analyzed in our study were typical variants with arelatively low mitotic index, and concordantly, cdc25B mRNA levelswere undetectable or very low in most of these cases.

cdc25A and cdc25C genes were less expressed than were cdc2SBgenes in this series of human non-Hodgkin's lymphomas. cdc25Aoverexpression was seen in three aggressive tumors, and no detectable mRNA levels of cdc25C were found in any of the lymphomas. However, variable levels of cdc25A, cdc25B, and cdc25Cexpression were observed in hematological cancer cell lines. Thereasons for these variations are not clear. The different cell originof the cell lines included in the study may suggest a possible tumorspecificity of these phosphatases. Alternatively, cell lines mayshow different patterns of expression than tumors. In functionalstudies, cdc25A has been shown to be up to 10 times more activethan is cdc25B (4). Therefore, it is still possible that alterations incdc25A expression levels below the detection threshold of our

4.

Table3 Correlationbetweenc-mycand cdc25Bexpressionin non-Hodgkin'slymphomas

genic potential of cdc25A and cdc25B but not cdc25C cooperatingwith either mutated Ha-ras or loss of RBJ (5). In addition, cdc25A and

cdc25B overexpression has been observed in some cancer cell linesand human tumors, including breast and head and neck carcinomas (5,7, 8). Here, we have examined the cdc2SA, cdc2SB, and cdc2SCexpression in human reactive lymphoid tissue and in a large series ofindolent and aggressive NHLs, including a number of histologicallyprogressed variants. We have demonstrated that cdc25B is the maincell cycle-activating phosphatase overexpressed in these tumors, andhigh mRNA levels are associated with the more aggressive andtransformed variants of the tumors. Concordantly, cdc25B overexpression was significanfly associated with a high proliferative activityof the tumors because 74% lymphomas overexpressing cdc25Bshowed a high S-phase fraction, whereas it was present only in 20%of tumors without cdc2SB overexpression (P < 0.01). These findingsare the first evidence that cdc25B cell cycle-activating phosphatase isoverexpressed in lymphoproliferative disorders and that it may beimplicated in the pathogenesis of more aggressive variants and in thetransformation of indolent lymphomas.

cdc25 expression in nonneoplastic tissues was negative or negligible, indicating that the basal expression of these genes is very low inreactive lymphoid tissues. In the aggressive tumors, cdc25B overexpression was found in 70% (9 of 13) primary diffuse large celllymphomas and in all samples of large cell lymphomas transformedfrom follicular lymphomas (4 cases) and from CLLs-Richter's syndrome (6 cases). cdc25B overexpression was found in 36% of the totalgroup of indolent lymphomas. However, it was detected in 69% ofCLLs but only in 25% of follicular lymphomas and 14% of mantlecell lymphomas. The relatively high number of typical CLLs overexpressing cdc25B may be related to the advanced stage of the tumorsincluded in this study because cdc25B overexpression was only detected in one of five (20%) stage A CLLs but it was detected in the

CDC25B r= 0.575

p<0.0015.

C-MVC

Fig. 3. Correlation between c-myc and cdc25B mRNA expression. The levels ofcdc25B nsRNA expression are significantly related to the degree of c-myc mRNAup-regulation (r 0.575, P < 0.001).

1765

CDC25B

f3-ACTIN

on June 3, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

cdc25 IN NON-HODGKIN'SLYMPHOMAS

technique may be relevant in the pathogenesis of these lymphomas.In vitro transforming experiments have shown the oncogenic potential of cdc25A and cdc25B but not of cdc25C (5). Similarly,cdc25B but not cdc25C overexpression has been observed in somehuman cancer cell lines (7) by Northern blot and in 32% of breastcarcinomas by in situ hybridization (5). Moreover, cdc25A andcdc25B were found to be overexpressed by quantitative reversetranscription-PCR in head and neck squamous cell carcinomas, incomparison to their respective normal mucosa, whereas cdc25Clevels were similar in normal and neoplastic tissues (8). Thesefindings are consistent with the lack of expression of cdc25C inhuman lymphomas and suggest that cdc25C may not be relevant inhuman tumor development or progression.

The mechanisms leading to cdc25 overexpression in human tumorsare not clear. No cdc25B gene amplifications or rearrangements wereobserved in our study in any of the tumors, including 22 cases withmRNA overexpression, indicating that other mechanisms must beimplicated in the up-regulation of this gene. In this study, we detectedan EcoRI polymorphism in the cdc25B gene that does not seem to beassociated with these disorders because the frequency of both alleleswas the same in negative and overexpressed tumors and in the normaland neoplastic population.

Recent studies have identified cdc25A and cdc25B, but notcdc25C, as physiologically relevant transcriptional targets of theoncogenic and apoptotic properties of c-myc (17). In fact, cdc25Aand cdc25B genes were found to contain three and one Myc/Maxbinding sites, respectively, whereas no c-myc response elementswere found in cdc25C gene. The c-myc oncogene has been implicated in the pathogenesis of different human neoplasms, includinglymphoid tumors (14, 15, 34—36). Particularly, c-myc gene translocations and mutations are a characteristic finding in Burkitt'slymphoma (10, 37), and c-myc overexpression, independent ofthese genetic alterations, is also a relatively frequent phenomenonin other aggressive lymphomas, which further correlates with theproliferative activity of the tumors (15). In addition, c-myc coopcrates with other oncogenes, such as bcl-2 (38, 39) and cyclin Dl(40, 41), in the development of lymphoid neoplasms in transgenicanimal models. Therefore, it would be possible that cdc25 genescould be targets of c-myc induction in human lymphomas. In ourstudy, we have found a significant correlation between cdc25B andc-myc overexpression in these tumors because 74% cases with highlevels of cdc2SB also showed c-myc overexpression whereas 96%lymphomas without detectable or very low cdc25B expression alsohad undetectable c-myc levels (P < 0.0001). In addition, a significant linear correlation between cdc25B and c-myc mRNA levelswas also found. These findings are consistent with a potentialregulatory role of cdc25B expression by c-myc in human lymphomas. However, the presence of cdc25B overexpression in ninetumors with undetectable c-myc levels (Table 3) would indicatethat other mechanisms may also participate in the up-regulation ofcdc25B gene in these tumors. Three different alternative splicevariants of the cdc25B mRNA have been recently described, andthe resulting proteins seem to have different activities both in vivoand in vitro (42). The cdc25B probe used in our study was thefull-length cDNA, and it recognized the three splice variants.Nevertheless, the small differences in the size of the differenttranscripts precludes the possibility of their identification byNorthern blot analysis. It would be interesting to analyze the roleof cdc25B splice variants in different tumors.

In conclusion, our findings indicate that the overexpression of thecdc25B cell cycle-activating phosphatase may participate in the development of aggressive non-Hodgkin's lymphomas and in the progression of indolent variants. The concomitant up-regulation of c-myc

and cdc25B genes observed in our study suggest a potential cooperative role of both genes in the pathogenesis of these tumors.

ACKNOWLEDGMENTS

We thank Drs. Manuel Serrano and Carlos Lopez Otmnfor their helpfulcomments on this project, Drs. Konstantin Galaktionov and David Beach forthe gift ofthe cdc25 probes, Dr. R. Dalla Favera for the gift ofthe c-mycprobe,and Drs. J. Yague and J. Vives for the DNA from normal blood donors for the

study of cdc2SB polymorphism and the cell lines used in this study.

REFERENCES

1. Sadhu, K., Reed, S. I., Richardson, H., and Russell, P. Human homolog of fissionyeast cdc25 mitotic inducer is predominantly expressed in G2. Proc. Nati. Aced. Sci.USA,87: 5139—5143,1990.

2. Galaktionov, K., and Beach, D. Specific activation ofcdc25 tyrosine phosphatases byB-type cyclins: evidence for multiple roles of mitotic cyclins. Cell, 67: 1181—1194,1991.

3. Barth, H., Hoffmann, I., Klein, S., Kaszkin, M., Richards, J., and Kinzel, V. Role ofcdc25-C phosphatase in the immediate G2 delay induced by the exogenous factorsepidermal growth factor and phorbolester. I. Cell Physiol., 168: 589—599,1996.

4. Jinno, S., Suto, K., Nagala, A., Igarashi, M., Kanaoka, Y., Nojima, H., and Okayama,H. Cdc25Ais a novelphosphatasefunctioningearlyin thecellcycle.EMBOJ., 13:1549—1556, 1994.

5. Galaktionov, K., Lee, A. K., Eckstein, J., Draetta, G., Meckler, J., Lode, M., andBeach, D. CDC25 phosphatases as potential human oncogenes. Science (WashingtonDC), 269: 1575—1577,1995.

6. Spitkovsky, D., Jansen-Durr, P.. Karsenti, E., and Hoffman, I. S-phase induction byadenovirus EIA requires activation of cdc25A tyrosine phosphatase. Oncogene, 12:2549—2554, 1996.

7. Nagata, A., Igarashi, M., Jinno, S., Suto, K., and Okayama, H. An additiOnalhomologof the fission yeast cdc25+ gene occurs in humans and is highly expressed in somecancer cells. New Biol., 3: 959—968, 1991.

8. Gasparotto, D., Maestro, R., Piccinin, S., Vulkosavljevic, T., Barzan, L., Sulfaro, S.,and Boiocchi, M. Overexpression of CDC25Aand CDC25Bin head and neck cancers.Cancer Res., 57: 2366—2368,1997.

9. Blackwell,T. K.,Kretzner,L, Blackwood,E. M.,Eisenman,R.N.,andWeintraub,H. Sequence-specific DNA binding by the c-myc protein. Science (Washington DC),250: 1149—1151,1990.

10. Marcu. K. B., Bossone, S. A., and Patel, A. 3. Myc function and regulation. Annu.Rev. Biochem., 61: 809-860, 1992.

II. Stone, J., de Lange, T., Ramsay, G., Jakobovits, E., Bishop, J. M., Varmus, H., andLee, W. Definition of regions in human c-mycthat are involved in transformation andnuclear localization. Mol. Cell. Biol., 7: 1697—1709,1987.

12. Amati, B., Dalton, S., Brooks, M. W., Littlewood, T. D., Evan, G. I., and Land, H.Transcriptional activation by the human c-myc oncoprotein in yeast requires interaction with max. Nature (Lond.), 359: 423—426,1992.

13. Hanson, K. D., Shichiri, M., Follansbee, M. R., and Sedivy. J. M. Effects of c-mycexpression on cell cycle progression. Mol. Cell. Biol., 14: 5748—5755, 1994.

14. Henriksson, M., and Luscher, B. Proteins of the Myc network: essential regulators ofcellgrowthanddifferentiation.Adv.CancerRes.,68: 109-182,1996.

15. Spencer, C. A., and Groudine, M. Control of c-myc regulation in normal andneoplastic cells. Adv. Cancer Res., 68: 1—48,1991.

16. Packham, G., and Cleveland, J. L. The role of ornithine decarboxylase in c-mycinduced apoptosis. Curr. Top. Microbiol. Immunol., 194: 283—290,1995.

17. Galaktionov, K., Chen, X., and Beach, D. cdc25 cell-cycle phosphatase as a target ofc-myc.Nature(Lond.),382:511—517,1996.

18. Bhatia, K., Huppi, K., Spangler. G., Siwarski, D., Iyer. R., and Magrath, 1. Pointmutations in the c-myc transactivation domain are common in Burkitt's lymphomaandmouseplasmacytomas.Nat.Genet.,5: 56—61,1993.

19. Fellbaum, C., Radaszkiewicz, T., Ruhri, C., Puts, B., Lehmacher, W., and Hofler. H.c-myc mRNA expression in non-Hodgkin's lymphomas. Virchows Arch. B CellPathol.,62: 61—68,1992.

20. Harris, N. L., Jaffe, E. S., Stein, H., Banks, P. M., Chan, 3. K., Cleary, M. L, Delsol,G., Dc Wolf-Peeters,C., Falini,B., Garter,K. C., Grogan,T. M., Isaacson,P.,Knowles, D. M., Mason, D. Y., Muller-Hermeink, H. K., Pileri, S., Pins, M. A.,Ralfkiaer, E., and Warnke, R. A. A revised European-American classification oflymphoid neoplasms: a proposal from the International Lymphoma Study Group.Blood,84: 1361—1392,1994.

21. Glissin, V., Crkvenjakov, R., and Byus, C. Ribonucleic acid isolated by cesiumchloride centrifugation. Biochemistry, 13: 2633—2637, 1974.

22. Bosch,F., Jares,P., Campo,E., Lopez-Guillenno,A., Pins, M. A., Villamor,N.,Tassies, D., Jaffe, E. S., Montserrat, E., Rozman, C., and Cardesa, A. Prad-l/cyclinDl gene overexpression in chronic lymphoproliferative disorders: a highly specificmarker of mantle cell lymphoma. Blood, 84: 2726—2732, 1994.

23. Hedley, D. W., Friedlander, H. L, Taylor, J. W., Rugg, C. A., and Musgrove, E. A.Method for analysis of cellular DNA content of paraffin-embedded pathologicalmaterial using flow-cytometry. J. Histochem. Cytochem., 31: 1333—1335,1983.

24. Shanicey. T. V., Rabinovitch, P. S., Baqwell, B., Bauer, K. D., Duque, R. E., Hedley,D. W.,Mayall,B. H., Wheeless,L., andCox,C. Guidelinesfor implementationofclinical DNA cytometry. Cytometry, 14: 472—477, 1993.

1766

on June 3, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

cdc25 IN NON-HODGKIN'S LYMPHOMAS

25. Hiddemann. W.. Schumann. J.. Andreef. M. B. B.. Herman. C. J.. Leif. R. C., Magali.B. H.. Hurphy. R. F.. and Sandberg. A. A. Convention on nomenclature foi DNAcytometry. Cytometry, 5: 445-446, 1986.

26. Gu. Y.. Rosenblatt. J.. and Morgan. D. O. Cell cycle regulation of cdk2 activity byphosphorylation of thrl60 and tyrlS. EMBO J.. //: 3995-4005. 1992.

27. Hall, M.. and Peters, G. Genetic alterations of cyclins. cyclin-dependent kinases, andcdk inhibitors in human cancer. Adv. Cancer Res., 68: 67-108. 1996.

28. Sherr, C. J., and Roberts, J. M. Inhibitors of mammalian G, cyclin-dependent kinases.Genes Dev., 9: 1149-1163, 1995.

29. Ott. M. M., Ott. G.. Kuse. R.. Porowski, P., Gunzer, U., Feller, A. C., and Muller-

Hermelink. H. K. The anaplastic variant of centrocytic lymphoma is marked byfrequent rearrangements of the bcl-l gene and high proliferation indices. Histopa-thology (Oxford). 24: 329-334, 1994.

30. Jares. P.. Campo. E.. Pinyol, M.. Bosch, F.. Miquel, R.. Fernandez. P. L.. Sánchez-Beato. M.. Soler, F.. Pérez-Losada, A., Nayach. I., Mallofré.C.. Piris, M. A.,Montserrat, E., and Cardesa, A. Expression of retinoblastoma gene produci (;irb) inmantle cell lymphomas. Correlation with cyclin DI (pradl/ccndl) mRNA levels andproliferative activity. Am. J. Pathol., 148: 1591-1600. 1996.

31. Lardelli, P., Bookman, M. A.. Sundeen. J.. Longo, D. L., and Jaffe, E. S. Lympliocyticlymphoma of intermediate differentiation. Morphologic and immunophenotypic spectrum and clinical correlations. Am. J. Surg. Pathol.. 14: 752-763, 1990.

32. Velders, G. A., Kluin-Nelemans, J. C.. De Boer, C. J., Hermans, J., Noordijk. E. M.,

Schuuring, E., Kramer. M. H.. Van Deijk. W. A.. Rahder. J. B.. Kluin, P. M.. and VanKrieken. J. H. Mantle-cell lymphoma: a population-based clinical study. J. Clin.Oncol.. 14: 1269-1274. 1996.

33. Bosch. F., López-Guüermo,A., Campo. E.. Ribera. J. M., Conde. E.. Piris, M. A.,Vallespí. T.. Woessner. S.. and Montserrat, E. Mantle-cell lymphoma: presenting

features, response to therapy, and prognostic factors. Cancer (Phila.). 82: 567-575,

1998.34. Inghirami. G.. Macri, L., Cesarman. E.. Chadburn, A., Zhong. J., and Knowles. D. M.

Molecular characterization of cd30+ anaplastic large-cell lymphoma: high frequencyof c-mvr proto-oncogene activation. Blood. f<J: 3581-3590. 1994.

35. Yano, T.. Jaffe, E. S., Longo. D. L., and Raffeid, M. Myc rearrangements inhistologically progressed follicular lymphomas. Blood. Hü:758-767, 1992.

36. Ballerini. P.. Gaidano. G.. Gong. J. Z.. Tassi. V.. Saglio. G., Knowles. D. M., andDalla-Favera, R. Multiple genetic lesions in acquired immunodeficiency syndrome-related non-Hodgkin's lymphoma. Blood. HI: 166-176, 1993.

37. Clark. H. M., Yano, T., Otsuki, T., Jaffe, E. S.. Shibata, D.. and Rattcld. M. Mutationsin the coding region of c-mvr in AIDS-associated and other aggressive lymphomasCancer Res., 54: 3383-3386. 1994.

38. Marin. M. C.. Hsu. B., Stephens. L. C.. Brisbay. S.. and McDonnell. T. J. Thefunctional basis of c-mvr and bcl-2 complementation during multistep lymphomagen-esis in vivo. Exp. Cell Res.. 217: 240-247, 1995.

39. McDonnell. T. J., and Korsmeyer. S. J. Progression from lymphoid hyperplasia tohigh-grade malignant lymphoma in mice transgenic for the U14;18). Nature (Lond.).349: 254-256, 1991.

40. Lovec, H., Grzeschiczek. A.. Kowalski, M. B.. and Moray. T. Cyclin Dl/bcl-lcooperates with mvr genes in the generation of B-cell lymphoma in transgenic mice.EMBO J.. 13: 3487-3495, 1994.

41. Bodrug. S. E., Warner. B. J.. Bath. M. L., Lindeman, G. J.. Harris. A. W., and Adams.J. M. Cyclin DI transgene impedes lymphocyte maturation and collaborates inlymphomagenesis with the myr gene. EMBO J., 13: 2124-2130, 1994.

42. Baldin, V., Cans, C.. Superti-Furga. G.. and Ducommun. B. Alternative splicing ofthe human CDC25B tyrosine phosphatase. Possible implications for growth control?Oncogene. 14: 2485-2495. 1997.

1767

on June 3, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1998;58:1762-1767. Cancer Res   Silvia Hernández, Luis Hernández, Sílvia Beà, et al.   Expression in Human Non-Hodgkin's Lymphomas

myc Cell Cycle-activating Phosphatases and c-cdc25

  Updated version

  http://cancerres.aacrjournals.org/content/58/8/1762

Access the most recent version of this article at:

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  [email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/58/8/1762To request permission to re-use all or part of this article, use this link

on June 3, 2020. © 1998 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from