sign in sign up
<<
Home , Cyclin D1, Multiple myeloma

Leukemia (2002) 16, 1844–1851  2002 Nature Publishing Group All rights reserved 0887-6924/02 $25.00 www.nature.com/leu Alterations of the D1/pRb/p16INK4A pathway in A Kra¨mer1, B Schultheis2, J Bergmann2, A Willer2, U Hegenbart1,ADHo1, H Goldschmidt1 and R Hehlmann2

1Medizinische Klinik und Poliklinik V, Universita¨t Heidelberg, Heidelberg, Germany; and 2III.Medizinische Klinik, Klinikum Mannheim der Universita¨t Heidelberg, Mannheim, Germany

The retinoblastoma (pRb), p16INK4A, D-type , and istic feature of mantle cell .10,11 Also, their partners cyclin-dependent kinase (CDK) 4 and 6 constitute amplification and overexpression is a negative prognostic fac- aG1 regulatory pathway commonly targeted in tumorigenesis. 12,13 Several malignancies show a reciprocal correlation between tor in t(11;14)-negative lymphoproliferative disorders and 14,15 genetic alterations of single members of the pRb pathway. a variety of solid tumors. CDK4 overexpression or Therefore, we determined the frequency of Rb deletions and can be detected in various tumor types, including cyclin D1 alterations by fluorescence in situ hybridization as and glioblastoma.16–18 Rb deletions and Ј INK4A well as 5 CpG island hypermethylation of the at the chromosomal 13q14 are among the most com- using methylation-specific polymerase chain reaction in INK4A marrow mononuclear cells from 82 individuals with mon genetic aberrations found in human neoplasias. p16 disorders. Alterations in at least one of the components of the occurs in a broad spectrum of solid tumors and is pathway were found in 75%. Cyclin D1 translocations or ampli- the most frequently observed genetic alteration in adult T cell fications were detected in 14/82 (17.1%), Rb deletions at 13q14 lymphoblastic .9–22 In addition, recently, de novo 5Ј in 23/82 (28%) of the cases, including three (3.6%) homozygous INK4A CpG island hypermethylation has been found to be associated deletions. p16 was hypermethylated in 33/57 (57.9%) of the INK4A INK4B samples. Further analysis revealed a highly significant corre- with transcriptional silencing of the p16 and p15 23 lation between cyclin D1 alterations and extramedullar or leu- in human . kemic myeloma manifestations (P = 0.014; Fisher’s test). Interestingly, a reciprocal correlation between genetic alter- Whereas Rb deletions seemed to occur alternatively to cyclin ations of single members of the pRb pathway has been D1 alterations, no reciprocal correlation was found between described for several solid tumors.16,18,24–26 An alteration of p16INK4A hypermethylations and cyclin D1 or Rb locus aber- rations. Cyclin D1 locus alterations and Rb deletions were any one gene in the pathway seems to eliminate selective INK4A associated with a significantly worse prognosis whereas pressure for changes in the others, so that Rb or p16 inac- p16INK4A hypermethylation had no impact on survival. We con- tivation and cyclin D1 or CDK4 overexpression are mutually clude that cyclin D1 and Rb aberrations seem to occur as alter- exclusive events in the generation of tumor cells. In contrast, native events in plasma cell malignancies and contribute to INK4A some lymphoid neoplasias do not show this alternative pattern clinical course and prognosis. In contrast, although p16 22,27,28 hypermethylation is frequent, inactivation of p16INK4A seems of pRb pathway alterations. not to be involved in the pathogenesis of plasma cell disorders. The molecular pathogenesis of multiple myeloma (MM) is Leukemia (2002) 16, 1844–1851. doi:10.1038/sj.leu.2402609 still largely unknown and, unlike other types of neo- Keywords: multiple myeloma; ; cyclin D1; reti- plasms, no genetic specifically associated with this INK4A noblastoma; p15 ; fluorescence in situ hybridization neoplasm have yet been found.29–31 Nevertheless, as detected by flow cytometric analysis of nuclear DNA con-

tent of G0/1 cells or by fluorescence in situ hybridization (FISH) Introduction has been described in almost 90% of patients with MM.32,33 In contrast, due to difficulties in obtaining analyzable meta- Recent genetic and biochemical investigations of the molecu- phases, abnormal karyotypes have been observed in only 30 lar mechanisms governing the G1 to progression in to 45% of cases.34,35 Among those, only partial or complete mammalian cells demonstrated a central role for D-type deletions of 13 harboring the Rb gene or abnor- cyclins and their partner cyclin-dependent kinases (CDK) 4 malities involving 11q, where the cyclin D1 gene is located, and 6.1–5 The -accelerating function of - are associated with poor prognosis.35,36 In addition, monoal- associated kinases appears to be mediated by phosphorylation lelic deletions of the Rb gene locus at 13q14 have been and concomitant inactivation of the reported to be an independent adverse prognostic variable in (pRb) in G phase, a process negatively regulated by p16INK4A 1 MM.37 These findings suggest a potential important role of the and other CDK inhibitors. As a result, transcription factors like INK4A and CBP/cycA are released, thereby inducing gene cyclin D1/pRb/p16 pathway in the pathogenesis of MM. Therefore, we determined the frequency of Rb deletions and expression and metabolic changes leading to onset of DNA Ј replication.4,6–8 cyclin D1 alterations by FISH as well as 5 CpG INK4A The emerging critical role of the cyclin D1/pRb/p16INK4A island hypermethylation of the p16 gene using methyl- pathway in cell cycle regulation is further supported by fre- ation-specific polymerase chain reaction (PCR) in bone mar- quent alterations of individual components of this checkpoint row mononuclear cells from three individuals with mono- mechanism in human tumors.1–5,9 The chromosomal translo- clonal gammopathy of undetermined significance (MGUS), 73 cation t(11;14)(q13;q32), which leads to juxtaposition of the patients with MM, and six patients with plasma cell leukemia immunoglobulin heavy chain promoter and the cyclin D1 (PCL). To further estimate the biological effect of these alter- gene, resulting in overexpression of cyclin D1 is a character- ations, the aberration state was correlated to the clinical course of the plasma cell disorders.

Correspondence: A Kra¨mer, Medizinische Klinik und Poliklinik V, Universita¨t Heidelberg, Hospitalstr. 3, 69115 Heidelberg, Germany; Fax: +49–6221–56–5721 Received 3 January 2002; accepted 28 March 2002 Cyclin D1/Rb/p16INK4A alterations in multiple myeloma A Kra¨mer et al 1845 Materials and methods was performed to detect numerical chromosomal aberrations. A two-color FISH with probes for cyclin D1 and the 14q32 Patients IgH region was used for the detection of a chromosomal trans- location t(11;14)(q13;q32). Means and standard deviations Heparinized samples as well as bone marrow (s.d.) of the percentages of nuclei with none, one and two smears were obtained from three individuals with MGUS and domain signals were calculated for the control specimens. A 73 patients with MM. Peripheral blood samples and smears cut-off value for a decision of mono- or nullisomy and from six patients with PCL were also included. Peripheral deletions was based on the percentage of cells with zero and blood samples and smears from eight normal individuals one signal, respectively, being greater than the mean percent- served as controls for determination of cut-off values. Bone age + 3 s.d. of the controls. marrow aspirations and blood drawings were performed for diagnostic purposes, and patients gave informed consent according to institutional guidelines to allowing an aliquot of Methylation-specific PCR the specimen to be used for research purposes. Main charac- teristics of patients are summarized in Table 1. The conditions and primer sets used for methylation-specific PCR to detect methylated p16INK4A promoter DNA are described elsewhere.42 The PCR was performed in a final vol- Interphase FISH ume of 25 ␮l containing 2.5 ␮l buffer (10 mM Tris-HCl, 50

mM KCl, 0.1% Triton X-100), 1.5 mM MgCl2, 1.0 mM dNTPs, Heparinized bone marrow or peripheral blood samples were 10 pM primers, 2 units of Taq polymerase (GibcoBRL, subjected to Ficoll gradient centrifugation (density 1.077 g/ml; Karlsruhe, Germany) and 60 ng of bisulfate-treated DNA. The Pharmacia, Freiburg, Germany). Buffy coats were washed amplification was performed in a thermal cycler (Biometra, once with phosphate-buffered saline (PBS) and brought on to Go¨ttingen, Germany) as described.42 A control PCR without slides, fixed in methanol/acetic acid 3:1 and stored at −70°C. DNA was performed for every PCR set in order to exclude Dual-color FISH was performed as described previously.38 contamination. Five ␮l of each PCR product was loaded on Commercially available DNA probes for cyclin D1 (Vysis, to agarose gels (0.5% agarose and 4% Nu-Sieve), electrophor- Stuttgart, Germany) and the centromeric region of chromo- esed and visualized after ethidium bromide staining by UV some 11 (Vysis) as well as DNA probes for Rb/13q14.2 illumination. If an amplification product was detected in the (digoxigenin-labeled; Oncor, Heidelberg, Germany) were PCR reaction with primers specific for methylated DNA, the used according to the manufacturer’s recommendations. sample was classified as methylated. If a specific PCR product Cosmid contigs spanning the 14q32 IgH region and the 9p21 was detected exclusively after PCR with primers specific for p16INK4A/p15INK4B region (8 cosmids, 250 kb) were kindly pro- unmethylated DNA, the sample was classified as unmethyl- vided by Dr R Siebert (Department of Human Genetics, Uni- ated. versity of Kiel, Germany) and Dr S Bohlander (Institute of Human Genetics, University of Go¨ttingen, Germany), respect- ively.39,40 From those, FISH probes were generated using a Statistical analysis standard protocol.41 Evaluation was performed by analyzing at least 200 interphase nuclei per sample using conventional The chi square test and Fisher’s exact test were used to exam- fluorescence microscopy. For cyclin D1, two-color FISH with ine the relation between any two factors. Overall survival was the unique probe and the corresponding centromeric probe calculated from the date of diagnosis until the patients death or last visit. Survival probability curves were analyzed and plotted according to the method of Kaplan and Meier and Table 1 Characteristics of 82 individuals with plasma cell dis- compared using the log-rank test. orders

Patient characteristics n % Results Age Ͼ60 yr 29 35.4 Sex Interphase FISH analysis of cyclin D1 and Rb Male 46 56.1 Female 36 43.9 Fourteen of the 82 individuals with plasma cell disorders Ͻ6 months from initial diagnosis 30 36.6 (17.1%) exhibited aberrations of chromosomal region 11q13, Ig-subtype five cases (6.1%) had a translocation t(11;14)(q13;q32), and IgG 51 62.2 nine cases (11%) had a 11. Distribution of deletions IgA 13 15.9 IgD 1 1.2 according to the stage of the disease is shown in Table 2. Bence-Jones 15 18.3 Except for one patient with stage I disease, who had a Non-secretory 2 2.4 t(11;14)(q13;q32), all patients with cyclin D1 locus alterations Stage (Durie and Salmon) had stage II/III MM or PCL. The percentage of cells with a MGUS 3 3.6 t(11;14)(q13;q32) or a polysomy 11 varied between 10.9 and I 9 11.0 72.9% and between 9.5 and 97.5%, respectively. In all II 8 9.8 III 54 65.9 diploid cases with a chromosomal translocation EMP 2 2.4 t(11;14)(q13;q32), two copies of the control centromeric PCL 6 7.3 probe for were detected in 97.8 ± 2.2%, pro- ving the reliability of the method. In the total group of 82 MGUS, of undetermined significance; plasma cell disorders, cyclin D1 gene locus alterations were EMP, extramedullary plasmocytoma; PCL, plasma cell leukemia. closely related to extramedullar or leukemic disease manifes-

Leukemia Cyclin D1/Rb/p16INK4A alterations in multiple myeloma A Kra¨mer et al 1846 Table 2 Results of FISH and MSP analysis in 82 individuals with plasma cell disorders according to stage

Stage *11q13 del(13q14) p16INK4A 5ЈCpG methylation t(11;14) Polysomy 11 themizygous homozygous

MGUS 0/3 0/3 1/3 1/3 0/2 I 1/9 0/9 2/9 0/9 4/6 II 0/8 2/8 1/8 1/8 6/7 III 2/54 5/54 14/54 0/54 17/35 EMP/PCL 2/8 2/8 2/8 1/8 6/7 Total 5/82 (6.1%) 9/82 (11%) 20/82 (24.4%) 3/82 (3.6%) 33/57 (57.9%)

FISH, fluorescence in situ hybridization; MSP, methylation-specific PCR; MGUS, monoclonal gammopathy of undetermined significance; EMP, extramedullary plasmocytoma; PCL, plasma cell leukemia.

Table 3 Cyclin D1/pRb/p16INK4A pathway aberrations in MM vs EMP/PCL

Disease *11q13/Polysomy 11 del(13q14) p16INK4A hypermethylation

n%P n%P n%P

MM 10/74 13.5% 0.014 20/74 27% 0.984 27/50 54% 0.385 EMP/PCL 4/8 50% 3/8 37.5% 6/7 83%

MM, multiple myeloma; EMP, extramedullary plasmocytoma; PCL, plasma cell leukemia

tations (13.5% vs 50% cyclin D1 aberrations in MM and square test; Table 5). Seven out of eight patients with EMP or extramedullary plasmocytoma (EMP)/PCL, respectively; Table PCL harbored either a cyclin D1 alteration or a Rb deletion. 3). This correlation was highly significant (P = 0.014; Fish- None of these patients had a combination of both aberrations. er’s test). Both cyclin D1 locus alterations and Rb deletions were Rb deletions were found in 23 of the 83 individuals with associated with a worse prognosis (P = 0.058; Figure 1a, b). plasma cell disorders (28%), 20 cases (24.4%) had hemizyg- Patients with and without cyclin D1 aberrations or Rb ous deletions, and three cases (3.6%) had homozygous deletions had 5-year survival rates of 55% and 72%, respect- deletions (Table 2). The percentage of cells with Rb deletions ively (Figure 1a). Interestingly, no difference was found varied between 13.8 and 60.1%. Rb deletion showed no clear between the survival curves of patients with cyclin D1 locus relation to an extramedullar or leukemic course of the plasma alterations and Rb deletions (Figure 1b). Five-year survival cell disorder (27% vs 37.5% Rb deletions in MM and rates of both groups were identical. Of note, all three patients EMP/PCL, respectively; P = 0.984; Fisher’s test; Table 3). with concurrent cyclin D1 alterations and Rb deletions are Only three cases (3.6%) exhibited both cyclin D1 locus alive at a median follow-up of 51 months. Neither cyclin D1 alterations and Rb deletions (Table 4). All three patients had alterations nor Rb deletions were more frequent in patients in stage III disease. Two had a t(11;14)(q13;q32) and a hemizyg- whom the plasma cell disorder was diagnosed 6 months or ous Rb deletion and one had a polysomy 11 and a hemizyg- longer before FISH analysis as compared to patients with ous Rb deletion. In the subgroup of patients with EMP or PCL, newly diagnosed disease (Ͻ6 months from initial diagnosis; there was a highly signifcant negative correlation between Table 6). cyclin D1 locus alteraltions and Rb deletions (P = 0.001, chi

Interphase FISH analysis of p16INK4A

Table 4 Results of FISH analysis in 34 individuals with 11q13 In 22 of the 82 individuals with plasma cell disorders a FISH aberrations and/or 13q14 deletions with a cosmid spanning the chromosomal region 9p21 for the detection of p16INK4A gene deletions was performed. Only one n *11q13 del(13q14) case with stage II MM exhibited a hemizygous p16INK4A deletion. Bone marrow mononuclear cells from this patient t(11;14) polysomy 11 hemizygous homozygous also harbored a polysomy 11 and a hypermethylation of the remaining p16INK4A allele. 3 +− −− 8 −+ −− 2 +− +− INK4A Ј 1 −+ +−Hypermethylation of the p16 5 CpG island by 17 −− +−methylation-specific PCR 3 −− −+ 5ЈCpG island hypermethylation of the p16INK4A gene was FISH, fluorescence in situ hybridization. detected in 33 of 57 analyzed individuals with plasma cell

Leukemia Cyclin D1/Rb/p16INK4A alterations in multiple myeloma A Kra¨mer et al 1847 Table 5 Results of FISH analysis in eight patients with plasma cell leukemia or multiple myeloma with extramedullary manifestations

Patient No. Disease *11q13 del(13q14)

t(11;14) polysomy 11 hemizygous homozygous

1PCL+− −− 2PCL+− −− 3EMP−+ −− 4PCL−+ −− 5EMP−− +− 6PCL−− +− 7PCL−− −+ 8PCL−− −−

FISH, fluorescence in situ hybridization; EMP, extramedullary plasmocytoma; PCL, plasma cell leukemia.

Table 6 Results of FISH and MSP analysis in 82 individuals with plasma cell disorders according to time from diagnosis

Aberration Time from diagnosis

Ͻ6 months Ͼ6 months

*11q13 6/30 (20) 8/52 (15.4) del(13q14) 13/30 (43.3) 10/52 (19.2) p16INK4A methylation 14/19 (73.7) 19/38 (50)

FISH, fluorescence in situ hybridization; MSP, methylation-specific PCR; MGUS, monoclonal gammopathy of undetermined signifi- cance; EMP, extramedullary plasmocytoma; PCL, plasma cell leu- kemia.

locus alterations and hypermethylation of the p16INK4A pro- moter. p16INK4A 5ЈCpG island hypermethylation was present in 15/22 (68.2%) plasma cell disorders with Rb deletions or 11q13 aberrations vs 18/35 (51.4%) plasma cell disorders without Rb deletions or 11q13 aberrations (P = 0.5; chi square test). p16INK4A hypermethylation had no impact on survival (Figure 2). Patients with and without p16INK4A hypermethyl- ation had identical 5 year survival rates of 72%. As with cyclin D1 alterations and Rb deletions, no statistically significant dif- ference was found between the frequency of p16INK4A hyper- methylation in patients in whom the plasma cell disorder was

Figure 1 Survival of 82 individuals with plasma cell disorders with and without cyclin D1 or Rb gene locus aberrations as detected by FISH analysis. (a) Cyclin D1 and Rb gene locus aberrations are plotted as a single group. (b) Cyclin D1 and Rb gene locus aberrations are plotted as separate groups. disorders (57.9%; Table 2). Methylation of the p16INK4A pro- moter was independent of the disease stage. No correlation between extramedullar/leukemic disease and p16INK4A pro- moter methylation status was found (54% vs 83% p16INK4A promoter hypermethylations in MM and EMP/PCL, respect- = ively; P 0.385; Fisher’s test; Table 3). Also, in the total group Figure 2 Survival of 57 individuals with plasma cell disorders with INK4A of 33 patients with p16 promoter hypermethylation, no and without p16INK4A 5ЈCpG island hypermethylation as detected by correlation was found between Rb deletions or cyclin D1 methylation-specific PCR.

Leukemia Cyclin D1/Rb/p16INK4A alterations in multiple myeloma A Kra¨mer et al 1848 diagnosed 6 months or longer before MSP analysis as com- cytic leukemia (CLL). In this disease, it has been suggested pared to patients with newly diagnosed disease (Ͻ6 months that the gene affected by the deletion at chromosomal region from initial diagnosis; Table 6). 13q14 is not Rb but a novel tumor suppressor gene.57 How- ever, the lack of a deleterious effect of such 13q deletions in CLL suggests the involvement of distinctly different genes in Discussion CLL and MM, respectively.56 This interpretation is corrobor- ated by recent findings of Shaughnessy and coworkers,56 who In the present study, an extensive analysis of the three most have generated a deletion map of chromosome 13q in MM common mechanisms of cyclin D1/pRb/p16INK4A pathway by interphase FISH with 11 probes spanning the long arm of deregulation, namely cyclin D1 locus alteration, Rb deletion . In this study a marked deletion heterogeneity and p16INK4A 5ЈCpG island hypermethylation revealed the with deletion evolution in different subclones of individual

abrogation of this G1/S phase regulatory pathway in 75% of patients and Rb deletions in 60% of the cases have been those individuals with plasma cell disorders for whom every described. In contrast to our results, Shaughnessy et al report parameter was evaluated. Using interphase FISH and methyl- an increased frequency of Rb deletions in patients with long- ation-specific PCR, cyclin D1 locus aberrations, Rb deletions, standing MM as opposed to newly diagnosed patients and and p16INK4A 5ЈCpG island hypermethylations were identified therefore suggest a participation of Rb deletions in MM in 17.1, 28 and 57.9% of patients with plasma cell disorders, progression. respectively. Notably, our data indicate a correlation between Interestingly, for some solid tumors, one hit at a single pos- cyclin D1 gene locus alteration and extramedullar or leu- ition of the pRb pathway seems to be sufficient to destabilize kemic disease manifestation. In addition, at least in EMP/PCL, cell cycle regulation. In melanoma, there is a reciprocal corre- a reciprocal correlation between cyclin D1 alterations and Rb lation between pRb inactivation, p16INK4A deletion, and CDK4 deletions was found whereas p16INK4A hypermethylations mutation.16 Also, in glioblastoma, either Rb deletion, p16INK4A occurred in addition to cyclin D1 alterations or Rb deletions. inactivation or CDK4 amplification have been described.18,24 Of particular interest, cyclin D1 locus alterations and Rb For non-small and small cell lung , reciprocal inacti- deletions were associated with a worse prognosis whereas vations of either Rb and p16INK4A or Rb and cyclin D1 over- p16INK4A hypermethylation had no impact on survival. expression have been described.25,26 These observations are Previous studies on cell cycle deregulation in MM are ham- in agreement with our finding of a reciprocal correlation pered by selective examination of just one component rather between cyclin D1 alterations and Rb deletions in EMP/PCL. 37,43–54 than multiple genes involved in G1/S control. Neverthe- Seven out of eight patients with PCL or MM with extramedul- less, those studies have already emphasized a relatively high lary manifestations had strictly reciprocal alterations of the incidence of chromosomal abnormalities involving 11q, cyclin D1 or Rb locus. From 34 individuals with 11q13 aber- where the cyclin D1 gene is located. Chromosomal translo- rations or 13q14 deletions only three harbored both a cyclin cation of cyclin D1 by a t(11;14)(q13;q32) has been found in D1 locus aberration and a Rb deletion. However, due to the 5–40% of MM patients34,35,45,46 and is closely associated with relatively low frequencies of these alterations, no firm con- enhanced expression of cyclin D1.55 Furthermore, it has been clusions for the total group of patients with plasma cell dis- described that the t(11;14)(q13;q32) translocation in MM is orders can be drawn. Alterations in at least one of the compo- linked to a more aggressive disease and a poorer prog- nents of the cyclin D1/pRb/p16INK4A pathway were found in nosis.35,46 In addition to the (11;14)(q13;q32) translocation, 75% of the patients examined. Aberrations of pathway other 11q abnormalities have also been reported to contribute elements that have not been examined in this study may

to a poor prognosis in MM, with one possible mechanism account for G1/S phase deregulation in additional cases. being the amplification of the cyclin D1 gene as has been Examples like overexpression by a chromosomal described for breast cancers.14 Likewise, polysomy 11 translocation t(6;14)(;q32) and p15INK4B or p18INK4C resulting in overexpression of cyclin D1 in MM has been deletions in MM have recently been reported.49,58 reported recently.47 However, despite an association between Both cyclin D1 gene locus aberrations and Rb deletions t(11;14)(q13;q32) and an aggressive clinical course with occurred with similar incidences in newly diagnosed and extranodal disease and peripheral blood involvement being longstanding plasma cell disorders, a finding that argues in suggested,35 the impact of cyclin D1 overexpression on mye- favor of a role of these chromosomal abnormalities as an early loma cell circulation still remains controversial.45,46 In our event in MM development.29 Cyclin D1 alterations as well as study, cyclin D1 locus alterations including t(11;14)(q13;q32) Rb deletions were associated with a poor prognosis. Interest- and polysomy 11 were detected in 14/82 (17.1%) cases and ingly, when analyzed separately, both types of abnormalities closely correlated with EMP/PCL. Of note, mantle cell lym- lead to similar survival curves, a finding that supports the view phoma frequently presents with extranodal disease including that cyclin D1 and Rb aberrations are alternative modes of

peripheral blood involvement. Moreover, cyclin D1-positive G1/S phase deregulation with identical consequences. unclassified chronic lymphoproliferative disorders are charac- In transgenic mice, the overexpression of cyclin D1 alone terized by a predominantly leukemic presentation and an is not sufficient for the development of malignant lym- aggressive clinical course.13 phoma.59,60 Consequently, hemizygous as well as homo- Consistent with our observations, hemizygous deletions of zygous deletions of p16INK4A have been described to occur in chromosomal region 13q14, where the Rb gene is located, approximately 50% of mantle cell in addition to have been described at frequencies of 50% to 70% in MM by its characteristic translocation t(11;14)(q13;q32) and a highly FISH analysis and are associated with a lower rate of response significant correlation between p16INK4A deletions and the to and a shorter overall survival.35–37,43,44,56 proliferation index has been demonstrated.27 In addition, con- Also in agreement with our data, in rare cases, biallelic comitant aberrations of p16INK4A and cyclin D1 in several deletions of Rb with loss of pRb expression have been tumor cell lines of different origin,61 concurrent over- reported. Deletions affecting the chromosomal region 13q14 expression of cyclin D1 and CDK4 in familial adenomatous 62 are the most frequent genetic abnormality in chronic lympho- polyposis patients, and inactivation of multiple

Leukemia Cyclin D1/Rb/p16INK4A alterations in multiple myeloma A Kra¨mer et al 1849 regulating genes in primary lymphoid malignancies have been A transcription at the G1/S boundary consists of NF-Y and a novel described.22,28,63 However, in MM deletions or mutations of 115 kDa subunit. Cancer Res 1997; 57: 5117–5121. the p16INK4A gene have only been detected occasionally so 9 Hunter T, Pines J. Cyclins and cancer II: cyclin D and CDK inhibi- 48–50 tors come of age. Cell 1994; 79: 573–582. far. These reports are in agreement with our observation 10 Bosch F, Jares P, Campo E, Lopez-Guillermo A, Piris MA, Villamor INK4A of only one hemizygous p16 gene deletion among 22 N, Tassies D, Jaffe ES, Monserrat E, Rozman C. Prad-1/cyclin D1 samples analyzed. Also, data on inactivation of p16INK4A by gene overexpression in chronic lymphoproliferative disorders: a 5ЈCpG island hypermethylation in MM have been contro- highly specific marker for . Blood 1994; 84: versial with frequencies between 0 and 75%.51–53 Our finding 2726–2732. of a p16INK4A promoter hypermethylation frequency of 57.9% 11 de Boer CJ, van Krieken JHM, Kluin-Nelemans HC, Kluin PM, Schuuring E. Cyclin D1 messenger RNA overexpression as a confirms earlier studies by Ng et al51and Tasaka and cowork- 52 marker for mantle cell lymphoma. 1995; 10: 1833– ers, who have reported hypermethylation frequencies of 75 1840. and 50%, respectively. However, in contrast to Uchida et al54 12 Yatabe Y, Suzuki R, Tobinai K, Matsuno Y, Ichinohasama R, Oka- we did not find a significant correlation between p16INK4A pro- moto M, Yamaguchi M, Tamaru J, Uike N, Hashimoto Y, Morish- moter hypermethylation and aggressiveness of the disease. ima Y, Suchi T, Seto M, Nakamura S. Significance of cyclin D1 Also, no inverse correlation between p16INK4A hypermethyl- overexpression for the diagnosis of mantle cell lymphoma: a clin- ation and cyclin D1 or Rb gene locus aberrations was found. icopathologic comparison of cyclin D1-positive MCL and cyclin INK4A D1-negative MCL-like B-cell lymphoma. Blood 2000; 95: 2253– Instead, p16 hypermethylations appeared in addition to 2261. cyclin D1 alterations and Rb deletions, a finding that has not 13 Levy V, Ugo V, Delmer A, Tang R, Ramond S, Perrot JY, Vrhovac been reported earlier in plasma cell disorders. Together with R, Marie JP, Zittoun R, Ajchenbaum-Cymbalista F. Cyclin D1 over- a lack of a deleterious effect of p16INK4A hypermethylations expression allows identification of an aggressive subset of leu- on the prognosis of the patients analyzed by us, these findings kemic lymphoproliferative disorders. Leukemia 1999; 13: 1343– argue against an additional effect of p16INK4A hypermethyl- 1351. 14 Gillett C, Fantl V, Smith R, Fisher C, Bartek J, Dickson C, Barnes ation on cell cycle deregulation as it has been described for INK4A D, Peters G. Amplification and overexpression of cyclin D1 in p16 gene deletions in other lymphoid malig- detected by immunohistochemical staining. Cancer 21,27,28,63,64 nancies. However, since G1/S phase deregulating Res 1994; 54: 1812–1817. events seem to occur in the majority of cases in MM and CDK 15 Michalides R, van Veelen N, Hart A, Balm A. Overexpression of inhibitors like flavopiridol show in vitro activity against MM cyclin D1 correlates with recurrence in a group of forty-seven cell lines,65 clinical trials with these compounds might be operable squamous cell carcinomas of the head and neck. Cancer warranted. 1995; 55: 975–978. 16 Bartkova J, Lukas J, Guldberg P, Alsner J, Kirkin AF, Zeuthen J, Bartek J. The p16-cyclin D/cdk4-pRb pathway as a functional unit frequently altered in melanoma pathogenesis. Cancer Res 1996; Acknowledgements 56: 5475–5483. 17 Wo¨lfel T, Hauer M, Schneider J, Serrano M, Wolfel C, Klehmann- Hieb E, De Plaen E, Hankeln T, Meyer zum Buschenfelde KH, We are grateful to Dr Stefan Bohlander, Institute of Human Beach D. A p16INK4a-insensitive CDK4 mutant targeted by cyto- Genetics, University of Go¨ttingen, Germany and Dr Rainer lytic T lymphocytes in a human melanoma. Science 1995; 269: Siebert, Department of Human Genetics, University of Kiel, 1281–1284. Germany for kindly providing the cosmids spanning the 18 He J, Olson JJ, James CD. Lack of p16INK4 or retinoblastoma pro- 14q32 IgH region and the 9p21 p16INK4A/p15INK4B region, tein (pRb), or amplification-associated overexpression of cdk4 is respectively. We thank Dr Axel Benner, Department of Biom- observed in distinct subsets of malignant glial tumors and cell lines. Cancer Res 1995; 55: 4833–4836. edical Statistics, German Cancer Research Center, Heidelberg, 19 Kamb A, Gruis NA, Weaver-Feldhaus J, Liu Q, Harshman K, Tav- Germany for help with the statistical analysis. This work was tigian SV, Stockert E, Day RS 3rd, Johnson BE, Skolnick MH. A supported by the Deutsche Krebshilfe and the For- cell cycle regulator potentially involved in genesis of many tumor schungsfonds, Fakulta¨tfu¨r Klinische Medizin Mannheim, types. Science 1994; 264: 436–440. Universita¨t Heidelberg, Germany. 20 Nobori T, Miura K, Wu DJ, Lois A, Takabayashi K, Carson DA. Deletions of the cyclin-dependent kinase-4 inhibitor gene in mul- tiple human cancers. Nature 1994; 368: 753–756. 21 Hebert J, Cayuela JM, Berkeley J, Sigaux F. Candidate tumor-sup- References pressor gene MTS1 (p16/INK4A) and MTS2 (p15/INK4B) display frequent homozygous deletions in primary cells from T- but not 1 Kamb A. Cell-cycle regulators and cancer. Trends Genet 1995; from B-cell lineage acute lymphoblastic . Blood 1994; 11: 136–140. 84: 4038–4044.

2 Sherr CJ, Roberts JM. Inhibitors of mammalian G1 cyclin-depen- 22 Hangaishi A, Ogawa S, Imamura N, Miyawaki S, Miura Y, Uike dent kinases. Genes Dev 1995; 9: 1149–1163. N, Shimazaki C, Emi N, Takeyama K, Hirosawa S, Kamada N, 3 Strauss M, Lukas J, Bartek J. Unrestricted cell cycling and cancer. Kobayashi Y, Takemoto Y, Kitani T, Toyama K, Ohtake S, Yazaki Nat Med 1995; 1: 1245–1246. Y, Ueda R, Hirai H. Inactivation of multiple tumor-suppressor 4 Weinberg RA. The retinoblastoma protein and cell cycle control. genes involved in negative regulation of the cell cycle, Cell 1995; 81: 323–330. MTS1/p16INK4A/CDKN2, MTS2/p15INK4B, , and Rb genes in pri- 5 Hall M, Peters G. Genetic alterations of cyclins, cyclin-dependent mary lymphoid malignancies. Blood 1996; 87: 4949–4958. kinases, and Cdk inhibitors in human cancer. Adv Cancer Res 23 Merlo A, Herman JG, Mao L, Lee DJ, Gabrielson E, Burger PC, 1996; 68: 67–108. Baylin SB, Sidransky D. 5ЈCpG island methylation is associated 6 Nevins JR. E2F, a link between the Rb tumor suppressor protein with transcriptional silencing of the tumour suppressor and viral oncoproteins. Science 1992; 258: 424–429. p16/CDKN2/MTS1 in human cancers. Nat Med 1995; 1: 686–692. 7 Kra¨mer A, Carstens CP, Fahl WE. A novel CCAAT-binding protein 24 He J, Allen JR, Collins VP, Allalunis-Turner MJ, Godbout R, Day necessary for adhesion-dependent transcription at the RS 3rd, James CD. CDK4 amplification is an alternative mech-

G1/S boundary is sequestered by a retinoblastoma-like protein in anism to p16 homozygous deletion in glioma cell lines. Cancer G0. J Biol Chem 1996; 271: 6579–6582. Res 1994; 54: 5804–5807. 8 Kra¨mer A, Carstens CP, Wasserman WW, Fahl WE. CBP/cycA, a 25 Shapiro GI, Edwards CD, Kobzik L, Godleski J, Richards W, Sugar- CCAAT-binding protein necessary for adhesion-dependent cyclin baker DJ, Rollins BJ. Reciprocal Rb inactivation and the p16INK4A

Leukemia Cyclin D1/Rb/p16INK4A alterations in multiple myeloma A Kra¨mer et al 1850 expression in primary lung cancers and cell lines. Cancer Res Deletion of the retinoblastoma gene in multiple myeloma. Leuke- 1995; 55: 505–509. mia 1994; 8: 1280–1284. 26 Schauer IE, Siriwardana S, Langan TA, Sclafani RI. Cyclin D1 over- 44 Juge-Morineau N, Mellerin MP, Francois S, Rapp MJ, Harousseau expression vs. retinoblastoma inactivation: implications for growth JL, Amiot M, Bataille R. High incidence of deletions but infrequent control evasion in non-small and small cell lung cancer. Proc Natl inactivation of the retinoblastoma gene in human myeloma cells. Acad Sci USA 1994; 91: 7827–7831. Br J Haematol 1995; 91: 664–667. 27 Dreyling MH, Bullinger L, Ott G, Stilgenbauer S, Muller-Hermel- 45 Avet-Loiseau H, Li J-Y, Facon T, Brigaudeau C, Morineau N, Malo- ink HK, Bentz M, Hiddemann W, Do¨hner H. Alterations of the isel F, Rapp MJ, Talmant P, Trimoreau F, Jaccard A, Harousseau cyclin D1/p16-pRB pathway in mantle cell lymphoma. Cancer Res JL, Bataille R. High incidence of translocations t(11;14)(q13;q32) 1997; 57: 4608–4614. and t(4;14)(p16;q32) in patients with plasma cell malignancies. 28 Jadayel DM, Lukas J, Nacheva E, Bartkova J, Stranks G, De Cancer Res 1998; 58: 5640–5645. Schouwer PJ, Lens D, Bartek J, Dyer MJ, Kruger AR, Catovsky D. 46 Fonseca R, Witzig TE, Gertz MA, Kyle RA, Hoyer JD, Jalal SM, Potential role for concurrent abnormalities of the cyclin D1, Greipp PR. Multiple myeloma and the translocation p16CDKN2 and p15CDKN2B genes in certain B cell non-Hodgkin’s t(11;14)(q13;q32): a report on 13 cases. Br J Haematol 1998; 101: lymphomas. Functional studies in a cell line (Granta 519). Leuke- 296–301. mia 1997; 11: 64–72. 47 Troussard X, Avet-Loiseau H, Macro M, Mellerin MP, Malet M, 29 Hallek M, Bergsagel PL, Anderson KC. Multiple myeloma: evi- Roussel M, Sola B. Cyclin D1 expression in patients with multiple dence for a multistep transformation process. Blood 1998; 91: myeloma. J 2000; 1: 181–185. 3–21. 48 Gernone A, Iolascon A, Pietrafesa A, Dammacco F. p16 gene 30 Gould J, Alexanian R, Goodacre A, Pathak S, Hecht B, Barlogie analysis in multiple myeloma (MM). Leukemia 1996; 10: 1401. B. Plasma cell karyotype in multiple myeloma. Blood 1988; 71: 49 Tasaka T, Berenson J, Vescio R, Hirama T, Miller CW, Nagai M, INK4A INK4B 453–456. Takahara J, Koeffler HP. Analysis of the p16 , p15 , and INK4C 31 Barlogie B, Hoover R, Epstein J. Multiple myeloma – recent devel- p18 genes in multiple myeloma. Br J Haematol 1997; 96: opments in molecular and cellular biology. Curr Top Microbiol 98–102. Immunol 1995; 194: 37–41. 50 Dilworth D, Liu L, Stewart K, Berenson JR, Lassam N, Hogg D. 32 Bunn PA, Krasnow S, Makuch RW, Schlam M, Schechter G. Flow Germline CDKN2A mutation implicated in predisposition to mul- cytometric analysis of DNA content of bone marrow cells in tiple myeloma. Blood 2000; 95: 1869–1871. patients with plasma cell myeloma: clinical implications. Blood 51 Ng MHL, Chung YF, Lo KW, Wickham NWR, Lee JCK, Huang DP. 1982; 59: 528–535. Frequent hypermethylation of the p16 and p15 genes in multiple 33 Drach J, Schuster J, Nowotny H, Angerler J, Rosenthal F, Fiegl M, myeloma. Blood 1997; 89: 2500–2506. Rothermundt C, Gsur A, Jager U, Heinz R. Multiple myeloma: 52 Tasaka T, Asou H, Munker R, Said JW, Berenson J, Vescio RA, INK4A high incidence of chromosomal aneuploidy as detected by Nagai M, Takahara J, Koeffler HP. Methylation of the p16 interphase fluorescence in situ hybridization. Cancer Res 1995; gene in multiple myeloma. Br J Haematol 1998; 101: 558–564. 55: 3854–3859. 53 Urashima M, Teoh G, Ogata A, Chauhan D, Treon SP, Sugimoto Y, Kaihara C, Matsuzaki M, Hoshi Y, DeCaprio JA, Anderson KC. 34 Weh HJ, Gutensohn K, Selbach J, Kruse R, Wacker-Backhaus G, INK4A Seeger D, Fiedler W, Fett W, Hossfeld DK. Karyotype in multiple Characterization of p16 expression in multiple myeloma and myeloma and plasma cell leukaemia. Eur J Cancer 1993; 29A: plasma cell leukemia. Clin Cancer Res 1997; 3: 2173–2179. 54 Uchida T, Kinoshita T, Ohno T, Ohashi H, Nagai H, Saito H. Hyp- 1269–1273. INK4A 35 Tricot G, Barlogie B, Jagannath S, Bracy D, Mattox S, Vesole DH, ermethylation of the p16 gene promoter during the pro- Naucke S, Sawyer JR. Poor prognosis in multiple myeloma is asso- gression of plasma cell dyscrasia. Leukemia 2001; 15: 157–165. 55 Akiyama N, Tsuruta H, Sasaki H, Sakamoto H, Hamaguchi M, ciated only with partial or complete deletions of chromosome 13 Ohmura Y, Seto M, Ueda R, Hirai H, Yazaki Y. Messenger RNA or abnormalities involving 11q and not with other karyotype levels of five genes located at chromosome 11q13 in B-cell tumors abnormalities. Blood 1995; 86: 4250–4256. with chromosome translocation t(11;14)(q13;q32). Cancer Res 36 Pe´rez-Simo´n JA, Garci´a-Sanz R, Tabernero MD, Almeida J, Gon- 1994; 54: 377–379. zalez M, Fernandez-Calvo J, Moro MJ, Hernandez JM, San Miguel 56 Shaughnessy J, Tian E, Sawyer J, Bumm K, Landes R, Badros A, JF, Orfao A. Prognostic value of numerical chromosome aber- Morris C, Tricot G, Epstein J, Barlogie B. High incidence of chro- rations in multiple myeloma: a FISH analysis of 15 different chro- mosome 13 deletion in multiple myeloma detected by mutliprobe mosomes. Blood 1998; 91: 3366–3371. interphase FISH. Blood 2000; 96: 1505–1511. 37 Zojer N, Ko¨nigsberg R, Ackermann J, Fritz E, Dallinger S, Kromer 57 Corcoran MM, Rasool O, Liu Y, Iyengar A, Grander D, Ibbotson E, Kaufmann H, Riedl L, Gisslinger H, Schreiber S, Heinz R, Lud- RE, Merup M, Wu X, Brodyansky V, Gardiner AC, Juliusson G, wig H, Huber H, Drach J. Deletion of 13q14 remains an inde- Chapman RM, Ivanova G, Tiller M, Gahrton G, Yankovsky N, pendent adverse prognostic variable in multiple myeloma despite Zabarovsky E, Oscier DG, Einhorn S. Detailed molecular delin- its frequent detection by interphase fluorescence in situ hybridiz- eation of 13q14.3 loss in B-cell chronic lymphocytic leukemia. ation. Blood 2000; 95: 1925–1930. Blood 1998; 91: 1382–1390. 38 Schultheis B, Kra¨mer A, Willer A, Hegenbart U, Goldschmidt H, 58 Shaughnessy J Jr, Gabrea A, Qi Y, Brents L, Zhan F, Tian E, Sawyer Hehlmann R. Analysis of and p53 gene deletions in multiple J, Barlogie B, Bergsagel PL, Kuehl M. Cyclin D3 at 6p21 is dysreg- myeloma. Leukemia 1999; 13: 2099–2103. ulated by recurrent chromosomal translocations to immunoglob- 39 Siebert R, Matthiesen P, Harder S, Zhang Y, Borowski A, Zuhlke- ulin loci in multiple myeloma. Blood 2001; 98: 217–223. Jenisch R, Metzke S, Joos S, Weber-Matthiesen K, Grote W, Schle- 59 Lovec H, Grzeschiczek A, Kowalski MB, Mo¨ro¨y T. Cyclin D1/bcl- gelberger B. Application of interphase fluorescence in situ 1 cooperates with the genes in the generation of B-cell lym- hybridization for the detection of the Burkitt translocation phoma in transgenic mice. EMBO J 1994; 13: 3487–3495. t(8;14)(q24;q32) in B-cell lymphomas. Blood 1998; 91: 984–990. 60 Bodrug SE, Warner BJ, Bath ML, Lindemann GJ, Harris AW. Cyclin 40 Dreyling MH, Bohlander SK, LeBeau MM, Olopade OI. Refined D1 transgene impedes lymphocyte maturation and collaborates in mapping of genomic rearrangements involving the short arm of lymphogenesis with the myc gene. EMBO J 1994; 13: 2124–2130. chromosome 9 in acute lymphoblastic leukemia and other hema- 61 Lukas J, Aagaard L, Strauss M, Bartek J. Oncogenic aberrations of tological malignancies. Blood 1995; 86: 1931–1938. p16INK4/CDKN2 and cyclin D1 cooperate to deregulate G1 con- 41 Lichter P, Cremer T. Chromosome analysis by non-isotopic in situ trol. Cancer Res 1995; 55: 4818–4823. hybridization. In: Rooney DE, Czepulkowski BH (eds). Human 62 Zhang T, Nanney LB, Luongo C, Lamps L, Heppner KJ, DuBois . Oxford University Press: New York, 1992, pp RN, Beauchamp RD. Concurrent overexpression of cyclin D1 and 157–192. cyclin-dependent kinase 4 (Cdk4) in intestinal adenomas from 42 Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB. Methyl- multiple intestinal neoplasia (Min) mice and human familial aden- ation-specific PCR: a novel PCR assay for methylation status of omatous polyposis patients. Cancer Res 1997; 57: 169–175. CpG islands. Proc Natl Acad Sci USA 1996; 93: 9821–9826. 63 Stock W, Tsai T, Golden C, Rankin C, Sher D, Slovak ML, Pallavic- 43 Dao DD, Sawyer JR, Epstein J, Hoover RG, Barlogie B, Tricot G. ini MG, Radich JP, Boldt DH. Cell cycle regulatory gene abnor-

Leukemia Cyclin D1/Rb/p16INK4A alterations in multiple myeloma A Kra¨mer et al 1851 malities are important determinants of leukemogenesis and dis- ciated with unfavorable clinical outcome. Blood 1997; 89: ease biology in adult acute lymphoblastic leukemia. Blood 2000; 4161–4166. 95: 2364–2371. 65 Semenov, Akyuz C, Roginskaya V, Chanhan D, Corey SJ. Growth 64 Kees UR, Burton PR, Lu¨C, Baker DL. Homozygous deletion of the inhibition and of myeloma cells by the CDK inhibitor p16/MTS1 gene in pediatric acute lymphoblastic leukemia is asso- flavopiridol. Leuk Res 2002; 26: 271–280.

Leukemia