Expression of p16INK4A and p14ARF in hematological ... - Nature

Leukemia (1999) 13, 1760–1769  1999 Stockton Press All rights reserved 0887-6924/99 $15.00 http://www.stockton-press.co.uk/leu

Expression of p16INK4A and p14ARF in hematological malignancies T Taniguchi1, N Chikatsu1, S Takahashi2, A Fujita3, K Uchimaru4, S Asano5, T Fujita1 and T Motokura1 1

Fourth Department of Internal Medicine, University of Tokyo, School of Medicine; 2Division of Clinical Oncology, Cancer Chemotherapy Center, Cancer Institute Hospital; 3Department of Hematology, Showa General Hospital; 4Third Department of Internal Medicine, Teikyo University, School of Medicine; and 5Department of Hematology/Oncology, Institute of Medical Science, University of Tokyo, Tokyo, Japan

The INK4A/ARF locus yields two tumor suppressors, p16INK4A and p14ARF, and is frequently deleted in human tumors. We studied their mRNA expressions in 41 hematopoietic cell lines and in 137 patients with hematological malignancies; we used a quantitative reverse transcription-PCR assay. Normal peripheral bloods, bone marrow and lymph nodes expressed little or undetectable p16INK4A and p14ARF mRNAs, which were readily detected in 12 and 17 of 41 cell lines, respectively. Patients with hematological malignancies frequently lacked p16INK4A expression (60/137) and lost p14ARF expression less frequently (19/137, 13.9%). Almost all patients without p14ARF expression lacked p16INK4A expression, which may correspond to deletions of the INK4A/ARF locus. Undetectable p16INK4A expression with p14ARF expression in 41 patients may correspond to p16INK4A promoter methylation or to normal expression status of the p16INK4A gene. All patients with follicular lymphoma (FL), myeloma or acute myeloid leukemia (AML) expressed p14ARF while nine of 23 patients with diffuse large B cell lymphoma (DLBCL) lost p14ARF expression. Patients with ALL, AML or blast crisis of chronic myelogenous leukemia expressed abundant p16INK4A mRNAs more frequently than patients with other diseases (12/33 vs 6/104, P ⬍ 0.01). Patients with FL and high p14ARF expression had a significantly shorter survival time while survival for patients with DLBCL and increased p14ARF expression tended to be longer. These observations indicate that p16INK4A and p14ARF expression is differentially affected among hematological malignancies and that not only inactivation but also increased expression may have clinical significance. Keywords: INK4A; ARF; leukemia; lymphoma; RT-PCR; prognosis

Introduction The human INK4A/ARF locus located on chromosome 9p21 attracts the attention of many oncologists, because it encodes two different candidate tumor suppressors, p16INK4A and p14ARF, which affect Rb and p53 pathways, respectively.1,2 p16INK4A is a cyclin-dependent kinase inhibitor (CKI) specific to CDK4 and CDK6 and can directly block cyclin Ddependent kinase activity.3 The cyclin D/CDK4 (or 6) complex facilitates G1-phase progression toward the S phase by phosphorylating and thus inactivating the Rb protein (pRb).4 Therefore, the upregulated expression of p16INK4A causes G1-phase arrest and function is dependent on normal Rb.5,6 In contrast, transcription of p16INK4A is repressed by Rb function.7 Once cells lack Rb, the levels of p16INK4A mRNA and protein were elevated without growth arrest.8,9 The aberrantly high expression of p16INK4A was evident in tumors without functional Rb.3,7 p16INK4A is a member of the INK4 family that has three other structurally related members, p15INK4B, p18INK4C and p19INK4D. Among them, INK4A and INK4B located on 9p21 just next to the INK4A/ARF locus are often inactivated in human malignancies and are considered to be candidate

Correspondence: T Motokura, Fourth Department of Internal Medicine, University of Tokyo, School of Medicine, 3-28-6 Mejirodai, Bunkyo-ku, Tokyo 112-8688, Japan; Fax: 81–3–3943–3102 Received 2 April 1999; accepted 13 July 1999

tumor suppressor genes.10,11 In human hematological malignancies, their inactivation occurs mainly by means of homozygous deletion or promoter region hypermethylation (reviewed in Ref. 12). In tumors such as pancreatic adenocarcinomas, esophageal squamous cell carcinomas and familial melanomas, p16INK4A is often inactivated by point mutation, which is not the case in hematological malignancies.12 On the other hand, genetic aberrations of p18INK4C or p19INK4D are rare in human tumors.12 The INK4A/ARF locus yields two transcripts derived from alternative first exons, exon 1␣ and exon 1␤, each of which is joined to sequences in exon 2.13,14 p16INK4A is translated from the ␣ form transcript derived from exon 1␣. The ␤ form transcript that has an initiation codon in exon 1␤ encodes an unrelated protein, the reading frame of which differs from that of p16INK4A; it is designated ARF, derived from an alternative reading frame protein.14 Ectopic expression of mouse p19ARF in the nucleus of rodent fibroblasts induces G1 and G2 phase arrest.14 Transfection of human ARF cDNA induces marked growth inhibition in head and neck squamous cell carcinoma cell lines and HeLa cells with nonfunctional Rb,15 while growth suppression by p16INK4A requires functional Rb.5,6 Thus, ARF is thought to function negatively on cell-cycle progression, in a manner different from p16INK4A. Human ARF protein, predicted to be 13 902 Da, is referred to as p14ARF.16 ARF is also a candidate tumor suppressor, because mice lacking p19ARF develop tumors and mouse embryo fibroblasts lacking p19ARF are transformed by oncogenic Ha-ras alone.17 Furthermore, in patients with T cell acute lymphoblastic leukemia (T-ALL) and rearranged alleles of this region, p14ARF encoding exons are always disrupted or deleted, whereas p16INK4A and p15INK4B encoding exons are spared in some patients.18 The human p14ARF binds directly to MDM2, resulting in stabilization of both p53 and MDM2, which induces p21Cip1 expression and cell-cycle arrest in both G1 and G2/M.16 In contrast, p14ARF is negatively regulated by wild-type p53 expression, resulting in a negative feedback loop.16,19 p14ARF can be inactivated by homozygous deletion,18 and promoter region hypermethylation,19 while mutations in exon 1␤ are not found in tumor-derived lung, bladder, glioma or melanoma cell lines or in primary T-ALL cells.18 DNA alterations and methylation status of the p16INK4A gene in hematological malignancies have been frequently examined.12 However, expression of p16INK4A and especially p14ARF in primary hematological malignancies has not been described in detail. We investigated the expression of p16INK4A and p14ARF in primary hematological malignancies and hematopoietic cell lines using a quantitative reverse transcriptionpolymerase chain reaction (RT-PCR) assay. We found that INK4A/ARF expression was often altered and differs among hematological malignancies. Patients with follicular lymphoma (FL) and increased p14ARF expression are likely to have a poor prognosis.

INK4A/ARF in hematological malignancies T Taniguchi et al

Materials and methods

Cell lines Cell lines used in this study are shown in Table 1. These cell lines, except for FLAM-76 and SP-49, were passaged in

Table 1

p16INK4A and p14ARF mRNA expressions in hematopoietic cell lines

Cell line

p16INK4A mRNA Northern

Lymphoid cell lines non-B non-T B cells

T cells

HTLV-1 infected

RT-PCR (unit)

p14ARF mRNA Northern

RT-PCR (unit)

pRB

p16 genome statusa

p53a

Source

Reh Nalm-6 SMS-SB LBW-2 BALL-1 Namalwa Ramos HS-sultan HA IM-9 SP-49c FLAM-76c P30/Ohkubo CEM RPMI-8402 HPB-ALL KOPT-K1 MOLT-3 Jurkat MOLT-4 MOLT-16 PEER SKW-3 A3/Kawakami MT-1 MT-2 HUT102

− − − − − − − − − − − + − − − − + − − − − − − − − − −

0 0 0 0 0 0 0 0 0 0.036 0 64 0 0 0 0 0.63 0 0 0 0 0.016 0 0 0.026 0.069 0

− − sm + − + + + − + − + − sm − − ab, sm − − sm − sm ab, sm + + + +

0 0 0 3.8 0 39 32 49 0 1.1 0 70 0 0 0 0 0 0 0 0 0 0 0 13 1.6 0.46 4

+ + + + + + + + + − + + + + + + + + + + + + + + + + +

del NR NR NR re met met met del pmet NR NR del del del NR met, mt del del re del re re w NR w w

NR w NR NR NR mt mt mtb NR NR NR NR NR mt NR NR w NR mt mt, w mt mt NR mt mt w w

A B B B C D B F E E G G F C H E H C C I E H E E C C C

KG-1 K562 KCL-22 HL-60 THP-1 U937 JK-1 HEL MEG-01s MEG-01 CMK CMK11–5 Meg-J MOLM-1

− − − − − − sm − + + − − − −

0.21 0 0.52 1.2 0 0 0 0 16 23 0 0 0 0.045

− − − + − + sm − + + − − − −

0.025 0 0.04 12 0 16 0 0 13 18 0 0 0 0.066

+ + + + + + + + − + + + + +

met del w mt del met del del NR deld w NR del w

mt mt mt del mt mt NR NR NR mt re NR NR NR

F F D F F D I F J J K K L E

Myeloid cell lines

a

RPMI1640 medium (GibcoBRL Life Technologies, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (Bio Whittaker, Walkersville, MD, USA) and 60 mg/l kanamycin (Meiji Seika Kaisha, Tokyo, Japan) at 37°C in a humidified atmosphere with 5% CO2. Cultures for FLAM-76 and SP-49 cells required additional interleukin-6 (10 ng/ml) and 5% fetal bovine serum, respectively.

Data from other studies (see Results and Discussion for references). HS-sultan is a derivative of Jijoye, which has mutated p53. c Cell lines with cyclin D1 overexpression. d In MEG-01 cells, homozygous deletion of p16 gene is reported, but another group reported p16 expression (see Discussion for references). ab, aberrant size transcript; del, deletion; met, methylated; mt, mutation; NR, not reported; pmet, partially methylated; re, gene rearrangement; w, wild type. Cell lines were obtained from: (A) Dr M Higashihara, First Department of Internal Medicine, University of Tokyo, (B) Dr T Nakamura, First Department of Internal Medicine, University of Tokyo, (C) Dr M Yoshida, Institute of Medical Science, University of Tokyo, (D) our lab, (E) Drs S Ogawa and H Hirai, Third Department of Internal Medicine, University of Tokyo, (F) Japanese Cancer Research Resource Bank, (G) Drs I Kubonishi and I Miyoshi, Department of Medicine, Kochi Medical School, (H) Dr Y Hayashi, Department of Pediatrics, University of Tokyo, (I) Dr K Tani, Institute of Medical Science, University of Tokyo, (J) Dr M Ogura, Aichi Cancer Hospital and Dr H Saito, First Department of Internal Medicine, Nagoya University, (K) Dr T Sato, Department of Pediatrics, Chiba University, and (L) Drs M Teramura and H Mizoguchi, Department of Hematology, Tokyo Women’s Medical School. b

1761


1762

Clinical specimens of hematological malignancies We examined a total of 137 tissue specimens from 137 patients in Japan with hematological malignancies. The specimens had been frozen with dimethyl sulfoxide in liquid nitrogen or in deep freezers at −80°C, until RNA preparation but some were freshly prepared. These patients were selected from the files of the Fourth Department of Internal Medicine, School of Medicine, University of Tokyo (Tokyo) between December 1985 and October 1998, Department of Hematology, Showa General Hospital (Tokyo) between March 1992 and February 1996, and Cancer Institute Hospital (Tokyo) between November 1993 and February 1996, on the basis of availability of frozen or fresh samples for molecular studies. The proportion of malignant cells in each sample was greater than 70%, judged with cytosmears. Eight patients had been referred from Nagano Red Cross Hospital (Nagano), Kurobe City Hospital (Toyama) and Teikyo University Ichihara Hospital (Chiba). Some of the specimens were obtained with written informed consent and others were archive specimens obtained at diagnostic procedures. The patients (87 men and 50 women) ranged in age from 14 to 89 years (median age: 59 years). Tissues examined included 69 lymph nodes, 34 bone marrow (BM) aspirates, 24 peripheral blood (PB) samples, seven pleural effusions, one peritoneal effusion and two extranodal tumors. The hematological malignancies in these patients are listed in Table 2. Among them, 93 patients (66 non-Hodgkin’s lymphomas (NHLs), 10 acute lymphoblastic leukemias (ALLs), four multiple myelomas (MMs), three chronic lymphocytic leukemias (CLLs), and two prolymphocytic leukemias (PLLs), two Waldenstro¨m’s macroglobulinemias (WMs), six adult T cell leukemia/lymphomas (ATLs)) were included in our previous studies on cyclin D1 overexpression.20,21 Survival time was measured from the time of pathologic diagnosis of the disease and from the time of sampling for molecular analysis. Table 2

Non-Hodgkin’s lymphoma B cell diffuse large small lymphocytic follicular mantle cell lymphoplasmacytoid Burkitt’s others T cell Unclassified Multiple myeloma/PCL/WM CLL/PLLa ALL/LBLb Acute myeloid leukemia CML-BC Adult T cell lymphoma/leukemia

a

RNA preparation RNAs of cell lines were extracted from exponentially growing cells by the acid guanidinium thiocyanate–phenol–chloroform (AGPC) method,22 boiled for 1 min, measured by optic density, and stored at −80°C. Frozen cells were thawed rapidly in a bath prewarmed at 37°C, and subsequently, RNAs of these cells were extracted by the AGPC method. From freshly prepared PB or BM, mononuclear cells (MNCs) were separated by standard Ficoll–Paque (Pharmacia Biotech, Uppsala, Sweden) density-gradient centrifugation, as outlined by the manufacturer. Fresh lymph node cells were prepared by sieving minced lymph nodes. RNAs of these cells were also extracted by the AGPC method.

cDNAs and Northern blot analysis The cDNA plasmids used in this study were as follows: a p16INK4A (␣ form) cDNA plasmid was kindly provided by Dr David Beach, Howard Hughes Medical Institute; a plasmid containing the insert of the ␤ form transcript (clone 13) was kindly provided by Dr Christian-Jacques Larsen, INSERM.13 Probes used for Northern blot analysis are as follows: a probe specific to exon 1␣ of p16INK4A was a 150 bp PCR product obtained by PCR using primers (P16AS206, P16S57; see Table 3) from the p16INK4A cDNA plasmid. A probe specific to exon 1␤ of p14ARF was a 197 bp PCR product obtained by PCR using primers (P16AS206, P16BS40; see Table 3) from clone 13. These two probes were 32P- radiolabeled with each anti-

Lack of p16INK4A and p14ARF expression in hematological malignancies

Disease

Total

Normal PB samples (n = 4) were obtained from healthy adult volunteers with informed consent. Normal BM aspirates (n = 3) from patients with NHL, without BM involvement, were obtained following acquisition of informed consent.

No. of patients

Lacking expression of p16INK4A

p14ARF

Both

p16INK4A only

p14ARF only

76

41

12

12

29

0

23 2 30 4 (4) 2 (2) 2 7 4 2 12 (6) 8 (3) 13 12 8 8

15 1 13 1 (1) 2 (2) 2 4 3 0 7 (3) 3 (3) 3 1 2 3

9 0 0 1 (1) 1 (1) 1 0 0 0 0 (0) 1 (1) 3 0 2 1

9 0 0 1 (1) 1 (1) 1 0 0 0 0 (0) 1 (1) 2 0 2 1

6 1 13 0 (0) 1 (1) 1 4 3 0 7 (3) 2 (2) 1 1 0 2

0 0 0 0 (0) 0 (0) 0 0 0 0 0 (0) 0 (0) 1 0 0 0

137 (15)

60 (9)

19 (3)

18 (3)

42 (6)

1 (0)

One patient with T-CLL is included and expressed both genes. Four patients with T-ALL are included and expressed both genes. PCL, plasma cell leukemia; WM, Waldenstro¨m’s macroglobulinemia; CLL, chronic lymphocytic leukemia; PLL, prolymphocytic leukemia; CML-BC, blast crisis of chronic myelogenous leukemia; ALL, acute lymphoblastic leukemia; LBL, lymphoblastic lymphoma. Numbers of cyclin D1-overexpressing patients are given in parentheses.

b


Table 3

1763

Primers used for PCR amplification

Name P16AS206 P16S57 P16BS40 BA299S BA538ASP16AS

Sequence

Primer length (bp)

5⬘-TGCCCATCATCATGACCTGG-3⬘ 5⬘-GGAGCAGCATGGAGCCTT-3⬘ 5⬘-TTCTTGGTGACCCTCCGGATT-3⬘ 5⬘-GCACCACACCTTCTACAATG-3⬘ 5⬘-TGCCCATCATCATGACCTGGTAGATGGGCACAGTGTGGGT-3⬘

20 18 21 20 40

sense PCR primer and Klenow fragment (Takara, Kyoto, Japan). A ␤-actin probe was a human ␤-actin 0.63-kb cDNA fragment, which was 32P-radiolabeled using a random primer labeling kit (Takara). An aliquot (10 ␮g/lane) of each total RNA was separated on a formaldehyde-agarose gel and blotted on to nitrocellulose membrane. The membrane was hybridized with each 32P-labeled probe as described,20 washed in high stringency of 0.1–0.2 × SSC and 0.1% sodium dodecyl sulfate (SDS) at 65°C and autoradiographed at −80°C with an intensifying screen.

RT-PCR cDNA was synthesized with oligo(dT)15 from total RNA, as described.20 An aliquot (1 ␮l) of cDNA (20 ng RNA equivalent) was placed in 20 ␮l of 1 × PCR buffer (10 mm TrisHCl, 50 mm KCl, 1.5 mm MgCl2, pH 8.3) with 200 ␮m each deoxyribonucleoside triphosphate (dNTP), 0.2 ␮m each primer (P16AS206, P16S57, P16BS40, BA299S) except for BA538ASP16AS which were added at 0.5 nm, 2 ␮Ci of ␣-32P dCTP, 0.5 U recombinant Taq DNA polymerase (Takara), and 1 ␮l dimethyl sulfoxide (Wako Pure Chemical Industries, Osaka, Japan). Reaction mixtures were overlaid with mineral oil (Sigma, St Louis, MO, USA) and PCR amplification was started by placing the capped tubes on the block in the DNA thermal cycler (Perkin Elmer, Norwalk, CT, USA) which had already been heated to over 90°C. Each cycle constituted denaturation (1 min at 94°C, first cycle 5 min), annealing (2 min at 64°C), and extension (3 min at 72°C, last cycle 10 min). PCR was run for 21 cycles unless otherwise stated. Five microliters of each PCR reaction were separated on a 4.5% polyacrylamide gel followed by autoradiography. An optical scanner was used and densitometrical analysis was made using NIH image 1.59 software (NIH, Bethesda, MD, USA). As negative controls, distilled water instead of cDNA or RT reactions without reverse transcriptase was subjected to PCR and we confirmed no false positive reaction. PCR primers were synthesized by Greiner Japan (Tokyo, Japan) or GIBCO BRL (Tokyo, Japan) and are listed in Table 3. Schematic presentation of design of the PCR primers is depicted in Figure 1a. Expected sizes of the PCR products are as follows: p16INK4A, 150 bp; p14ARF, 197 bp; ␤-actin, 260 bp. Three specific upstream primers, a shared downstream primer, and a composite downstream primer were used. The shared downstream primer, P16AS206, corresponded to 187–206 nucleotide (nt) in the sequence of p16INK4A (␣ form)3 and 217– 236 nt in the sequence of p14ARF (␤ form).13 The upstream primer specific to p16INK4A, P16S57, corresponded to 57– 74 nt in the sequence.3 The upstream primer specific to p14ARF, P16BS40, corresponded to 40–60 nt in the sequence and was the same as P1 primer described by Duro et al.13 The upstream primer specific to ␤-actin, BA299S, corresponded to 299–318 nt in the sequence of ␤-actin cDNA.23 The 3⬘ portion

Figure 1 Quantitative RT-PCR for relative expression levels of p16INK4A and p14ARF with endogenously expressed ␤-actin used as an internal control. (a) Schematic presentation of primer setting on p16INK4A, p14ARF and ␤-actin sequences. Thick lines indicate coding regions and thin lines represent truncated non-coding regions. Thick arrows indicate primers used in the PCR. The common primer, P16AS206, is derived from the identical region between the p16INK4A and p14ARF sequence. The composite primer, BA538ASP16AS, consisted of P16AS206 in the 5⬘ part and ␤-actin sequence in the 3⬘ part. The common primer, P16AS206, is shared in amplification of the three PCR products, while P16S57, P16BS40, and BA299S are specific to p16INK4A, p14ARF, and ␤-actin sequences, respectively. (b) PCR of p16INK4A, p14ARF and ␤-actin cDNA templates. The indicated template, ␤-actin, p14ARF or p16INK4A cDNA plasmid, was added at the indicated concentration while the other two were kept at 2 × 104 molecules/␮l for p14ARF and p16INK4A, and at 2 × 105 molecules/␮l for ␤-actin and PCR was done, as described in ‘Materials and methods’. After electrophoresis, the gels were dried and exposed to X-ray films at −80°C with intensifying screens. Arrows indicate PCR products corresponding to p16INK4A, p14ARF and ␤-actin.


1764

of the composite downstream primer, BA538ASP16AS, corresponded to 519–538 nt in the ␤-actin sequence23 and the 5⬘ portion was the same as P16AS206. The level of p16INK4A or p14ARF expression was determined as the ratio of the signal of the p16INK4A or p14ARF PCR product to those of the internal control (␤-actin). RT-PCR for cyclin D1 overexpression was done, as described.21

Western blot analysis for pRb Cells were lyzed with 1 × sample buffer (Tris-HCl 60 mm, SDS 2%, dithiothreitol 0.1 m, pH 6.8) and boiled for 5 min. Protein concentration was determined by spectrophotometry using BCA Protein Assay Reagent (Pierce, Rockford, IL, USA).24 A total of 30 ␮g protein per lane was loaded on an 8% SDSpolyacrylamide gel and subjected to Western blot analysis, as described.25 Anti-pRb monoclonal antibody (MAb1; TRITON Diagnostics, Alameda, CA, USA), at 0.4 ␮g/ml, was used.

Statistical analyses Univariate (Kaplan–Meier) analysis was used to assess survival differences. Survivals between different subgroups were compared using the Wilcoxon test.

Results

Quantitative RT-PCR to compare expression levels of p16INK4A, p14ARF and ␤-actin Co-amplification of targets (p16INK4A and p14ARF) and control (␤-actin) in a tube was designed to circumvent difficulties in conventional competitive RT-PCR assays (Figure 1a). Simple co-amplification of control and targets reveals the ratio of target to control templates by comparing target and control products generated during the exponential phase of the PCR reaction.26 Since ␤-actin is expressed at a level much higher than those of the targets, comparable co-amplification can be difficult. Therefore, we used a composite primer at a low concentration to reduce ␤-actin signals at a constant ratio to the level comparable to the target signals. As shown in Figure 1a, we designed a common primer, P16AS206, derived from 16 bases in 5⬘ end of exon 2 and four identical bases in 3⬘ ends of exons 1␣ and 1␤, and a composite primer, BA538ASP16AS, the 3⬘ part of which is derived from the sequence of ␤-actin and the 5⬘ part of which is the same as the common primer, P16AS206. The three specific primers for p16INK4A, p14ARF and ␤-actin are P16S57, P16BS40 and BA299S, respectively. Amplification of ␤-actin cDNA must be initiated with BA299S and BA538ASP16AS, then, P16AS206 takes the place of BA538ASP16AS because we use BA538ASP16AS at a lower concentration. Final products for ␤-actin were made by virtue of P16AS206 and BA299S. We confirmed identity of the PCR products by direct sequencing (data not shown). We found exponential amplification phases of three genes overlapped up to 21 cycles with a cDNA from MEG-01s cell line used as a template (data not shown). Therefore, PCR was next run for 21 cycles. In addition, using cDNA plasmids for the genes as templates, we confirmed the ratio of PCR products reflecting that of added templates (Figure 1b).

p16INK4A and p14ARF expression in hematopoietic cell lines We studied p16INK4A and p14ARF mRNA expression in hematopoietic cell lines, by RT-PCR described above and Northern blot analysis. As shown in Figures 2a and b, the RT-PCR and Northern blot analysis gave similar results and thus, the quantitative nature of the PCR was re-confirmed. The RT-PCR had a more sensitive detectability. In 22 (54%) of 41 cell lines, neither p16INK4A nor p14ARF mRNA was detected by the RT-PCR, which is consistent with frequent deletions of the INK4A/ARF locus in cell lines reported previously.12,27 In seven (17%) of 41 cell lines, only p14ARF was detected, which might be related to p16INK4A promoter hypermethylation (Table 1). p14ARF expression without p16INK4A expression was frequently (4/11) observed in B cell lines, while it was infrequent (1/12) in non-HTLV-1-infected T cell lines (Figure 2c). In contrast, p16INK4A expression without p14ARF expression was observed only in two non-HTLV-1infected T cell lines, KOPT-K1 and PEER (Figure 2b). Li et al7 reported that p16INK4A mRNA accumulates to a high level in cells lacking Rb function and that transcription of p16INK4A is repressed by Rb. Therefore, we studied pRb expression in these cell lines by Western blot analysis (Table 1). However, reciprocal expression of p16INK4A and pRb is not necessarily expected. ARF is reported to be expressed highly in cell lines lacking p53.16 p53 mutation, deletion or rearrangement was noted in at least 18 of the cell lines listed in Table 1.28–37 However, mutation of p53 and lack of p14ARF expression were not mutually exclusive. Seven cell lines showed smears in Northern blot analysis of p14ARF mRNA and JK-1 showed a smear of p16INK4A mRNAs, although the RT-PCR revealed no corresponding signals (Table 1 and Figure 2b). Among them, two cell lines (KOPT-K1 and SKW-3) also showed p14ARF bands of aberrant sizes.

p16INK4A and p14ARF expression in patients with hematological malignancies As shown in Figure 3a, in normal PB and BM MNCs and lymph nodes of reactive lymphadenitis, p14ARF mRNAs were at low levels and p16INK4A mRNAs were barely detectable. In clinical specimens of hematological malignancies, abnormally high expression levels of p16INK4A and/or p14ARF compared to normal tissues were frequently observed (Figure 3). On the other hand, p16INK4A mRNA expression was lacking in 60 of 137 patients and p14ARF mRNAs were undetectable in 19 of 137. Almost all patients without p14ARF mRNAs lacked p16INK4A expression (Table 2) except for one patient with ALL who had B cell phenotype and TCR rearrangement. p16INK4A and p14ARF expression differed among types of diseases. All patients with FL expressed p14ARF while nine (39%) of 23 patients with diffuse large B cell lymphoma (DLBCL) lacked p14ARF expression, thus corresponding to the relatively high incidence of p16INK4A deletion in diffuse lymphoma (13%).12 On the other hand, p16INK4A expression was frequently undetectable in both DLBCL (15/23, 65%) and FL (13/30, 43%). MM was similar in expression to FL. These findings are consistent with the previous reports that homozygous deletion of the INK4A/ARF locus in FL and MM is rare and that p16INK4A is hypermethylated in 40% of FL and in 75% of MM.12,38,39 In a study of p16INK4A protein expression in NHLs by Western blot analysis, loss of expression was rare (6%) in typical FL but frequent (28%) in large cell lymphoma.40 Two


1765

Figure 2 Analysis of hematopoietic cell lines. (a) Comparison between the RT-PCR and Northern blot analysis. RNAs extracted from the indicated cells on each lane were subjected to the RT-PCR (upper part) and to Northern blot analysis (middle part) as described in ‘Materials and methods’. RNA loading in each lane is shown in the photograph of the ethidium bromide-stained gel at the bottom. (b) Comparison between the RT-PCR and Northern blot analysis of a part of T cell lines. Note smears shown in lanes for CEM, KOPT-K1, MOLT-4, PEER and SKW-3, and bands of aberrant sizes in lanes for SKW-3 and KOPT-K1 in the lower panel. (c) The RT-PCR analysis of remaining hematopoietic cell lines for expression of p16INK4A and p14ARF. Arrows indicate PCR products corresponding to p16INK4A, p14ARF and ␤-actin or each transcript signal of Northern blot analysis.

other immunohistochemical studies revealed that the incidence of p16INK4A protein loss in FL is zero whereas in diffuse large lymphoma it is as high as 44–65%.41,42 Taken together, in a portion of FL, p16INK4A may be hypermethylated and its mRNA levels may be low, but not low enough to silence protein expression. In almost all patients with acute myeloid leukemia (AML), both expressions were detectable, such being consistent with the rarity of homozygous deletion of INK4A/ARF locus in this disease.12 High levels of p16INK4A expression (defined as INK4A/␤-actin ⬎0.2 arbitrarily) were significantly more frequent in those with acute leukemia, including ALL, AML and blast crisis of chronic myelogenous leukemia (CML-BC) than in other hematological malignancies (12/33 vs 6/104, P ⬍ 0.01, chi square test). We also examined cyclin D1 mRNA expression in these patients using the RT-PCR assay that we had devised;20,21 we identified 15 patients (six NHLs, one WM, five MMs, one CLL and two PLLs) with cyclin D1 overexpression. There was no apparent relationship between p16INK4A and p14ARF expression and cyclin D1 overexpression (Table 2).

or FL. We found no significant difference in OS between the presence and absence of p16INK4A expression for patients with DLBCL or FL. Regarding p14ARF expression, when survival time for patients with DLBCL was measured from the time of sampling for molecular analysis, patients with increased p14ARF expression (p14ARF/␤-actin ⬎0.19) compared to normal controls tended to have a longer OS than did other patients (P = 0.096, Figure 4b). Since all patients with FL expressed p14ARF, we determined if the high levels of p14ARF expression affected survival time. When high levels of p14ARF expression were defined arbitrarily as p14ARF/␤-actin ⬎0.6, patients with high p14ARF expression had a significantly shorter OS from the time of diagnosis than did other patients (P = 0.030, Figure 5a). OS measured from the time of molecular analysis showed no significant difference (P = 0.21) (Figure 5b). We found no significant correlations of p14ARF overexpression and other clinical parameters including age, LDH levels and clinical stage in FL. Discussion

Clinical correlation We then asked whether p16INK4A or p14ARF expression status affected overall survival time (OS) for patients with DLBCL

We used a quantitative RT-PCR assay to examine relative expression levels of p16INK4A and p14ARF in primary tumors of various hematological malignancies and hematopoietic cell lines. Major mechanisms of gene inactivation of this locus are


1766

Figure 4 Overall survival time for diffuse large B cell lymphoma patients, with or without increased p14ARF mRNA expression. (a) There is no significant difference in the overall survival time from the time of diagnosis between patients with increased p14ARF expression (solid line, n = 11) and those without (dotted line, n = 12). (b) Patients without increased p14ARF expression (dotted line) tended to have a shorter overall survival time from the time of molecular analysis than did other patients (solid line).

Figure 3 p16INK4A and p14ARF mRNA expression in normal tissues and clinical specimens of various hematological malignancies. (a) Representative results of the RT-PCR analysis are shown. Arrows indicate PCR products corresponding to p16INK4A, p14ARF and ␤-actin. (b) Histogram of relative p16INK4A and p14ARF mRNA expression levels determined by RT-PCR. Expression levels were standardized using the ␤-actin signal of each sample as an internal control and are plotted over a ratio of p16INK4A or p14ARF to ␤-actin signal. LN, lymph node of reactive lymphadenitis; PB MNC, mononuclear cells from normal peripheral blood; BM MNC, mononuclear cells from normal bone marrow; NHL, non-Hodgkin’s lymphoma; DLBCL, diffuse large B cell lymphoma; FL, follicular lymphoma; WM, Waldenstro¨m’s macroglobulinemia; MM, multiple myeloma; CLL, chronic lymphocytic leukemia; PLL, B cell prolymphocytic leukemia; ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; CML-BC, blast crisis of chronic myelogenous leukemia; ATL, adult T cell leukemia/lymphoma. Dotted lines indicate the maximum levels of expression observed in controls. *, Specimens were obtained during relapse or progression of diseases.

homozygous deletion and hypermethylation, both leading to lack of expression. This assay cannot detect some modes of gene inactivation such as point mutation. However, p16INK4A inactivation by point mutation is rare in hematological malignancies12 and mutations in p14ARF exon 1␤ are not found in tumor-derived lung, bladder, glioma or melanoma cell lines or in primary T-ALL cells.18 To maintain the quantitative nature of our RT-PCR assay, we limited the PCR cycles to 21 and the sensitivity of PCR is somewhat compromised. In reactive lymphoid tissues used as controls, p16INK4A mRNAs were barely detectable with this assay, although p16INK4A nuclear immunoreactivity has been reported.41 p16INK4A protein production from levels of p16INK4A mRNA undetectable in our RT-PCR assay is possible and it should be noted that the ‘lack of p16INK4A expression’ does not necessarily lead to complete silencing of the p16INK4A gene. Therefore, our assay may overestimate the inactivation of p16INK4A. However, it has as good a quantitative nature as Northern blot analysis and is useful when comparing the expression levels of p16INK4A and p14ARF mRNAs among a number of samples. To the best of our knowledge, this is the first report that characterizes simultaneously mRNA expression levels in various hematological malignancies. Our results were fairly consistent with the genomic status of the p16INK4A gene reported in hematopoietic cell lines (Table 1).12,27,37,43,44 In the cell lines in which p16INK4A genes were


Figure 5 Overall survival time for follicular lymphoma patients, subdivided according to the expression levels of the p14ARF mRNAs. (a) p14ARF highly-expressing patients (solid line, n = 6) had a significantly shorter overall survival time from the time of diagnosis than other patients (dotted line, n = 24). (b) There is no significant difference in the overall survival time from the time of molecular analysis between patients with high p14ARF expression (solid line) and those with low p14ARF expression (dotted line).

deleted, neither p16INK4A nor p14ARF mRNAs were detected, except for MEG-01 cells, contradictory results were reported.44,45 In cell lines reported to have methylated p16INK4A gene, p16INK4A mRNAs were either not detected or were at low to intermediate levels, and p14ARF mRNA levels were almost always elevated. This means that methylation of the p16INK4A gene does not necessarily lead to complete gene silencing and that even if the p16INK4A gene is silenced by methylation, p14ARF mRNA expression can be elevated. This is consistent with reported data that human ARF promoter region hypermethylation is a low-frequency event in cell lines.19 Smears and bands of aberrant sizes observed in seven cell lines by Northern blot analysis of p14ARF could be transcripts derived from exon 1␤ without exon 2. Since we used primers for RT-PCR on exon 1␤ and exon 2, breakpoints of rearrangement may be located between exon 1␤ and exon 2 in these cell lines. In fact, a breakpoint cluster region exists within 3kb 3⬘ to exon 1␤ and MOLT-4 has a breakpoint in this region.46 We confirmed the expression of exon 1␤ in these cell lines (except for KOPT-K1) using an RT-PCR assay and primer pairs within exon 1␤ (data not shown). These aberrant transcripts derived from exon 1␤ may produce functional p14ARF protein, because the aminoterminal domain encoded by exon 1␤ is both necessary and sufficient for inducing G1 arrest.1 Whether functional proteins are actually made remains to be elucidated. JK-1 cells also showed a smear on Northern blot analysis for exon 1␣ (data not shown) and are

negative for p16INK4A by the RT-PCR. This smear could be transcripts derived from exon 1␣ without an efficient termination signal, because in these cells, exon 2 is deleted.44 These transcripts would not produce functional p16INK4A protein, because exon 2 is essential for p16INK4A function.1 In primary tumors, p16INK4A expression more frequently than p14ARF expression was lacking. In only one ALL patient was p16INK4A expressed without p14ARF. Therefore, in many patients with hematological malignancies, the target of inactivation is p16INK4A alone, or both p16INK4A and p14ARF, but not p14ARF alone, although there remains the possibility that p16INK4A was expressed below detectable levels. However, in ALL, p14ARF may be a primary target of inactivation as suggested by findings in one ALL patient with TCR rearrangement and two T cell lines, KOPT-K1 and PEER, which expressed p16INK4A without p14ARF, probably through an illegitimate recombination mechanism reported by other investigators.18,46 Garcia-Sanz et al47 reported that deletions and rearrangements of p16INK4A are associated with a poor prognosis in B cell NHLs. Loss of p16INK4A expression and deletions of p16INK4A gene are associated with aggressive variants of mantle cell lymphomas48 and homozygous deletions of p16INK4A are associated with histologic progression in FL.49 Interestingly, when survival time for patients with DLBCL was measured from the time of molecular analysis (not from the time of diagnosis), the difference in OS between the absence and presence of increased p14ARF expression became bigger, but not significant (Figure 4a and b). Inactivation of the INK4A/ARF locus may occur by chance and change the course of the disease. On the other hand, in FL without homozygous deletion of INK4A/ARF locus, p14ARF overexpression seemed to be a poor prognostic factor. Several mitogenic stimuli such as E1A, myc, oncogenic ras, V-abl and E2F-1 upregulate ARF leading to p53 stabilization (reviewed in Ref. 2). p14ARF is negatively regulated by wild-type p53 expression.16,19 Therefore, the high expression of p14ARF observed in this study may reflect oncogenic stimuli in tumor cells and/or inactivation of other tumor suppressors such as p53 and Rb. Since we analyzed only a small number of patients, the relation to other prognostic factors remains to be clarified. p16INK4A mRNA accumulates to a high level in cells lacking Rb function7 and in senescent cells caused by an increase in the number of population doublings.9 High expression levels of p16INK4A mRNAs especially in acute leukemias might be related to a rapid increase in population doublings of leukemic cells. Recently, Mekki et al50 reported that childhood ALL patients with high p16INK4A expression have shorter disease-free survival than other such patients. In conclusion, p16INK4A and p14ARF expression is differentially affected and is more frequently deregulated in hematological malignancies than speculated from documented genetic data. In addition to inactivation, overexpression may have clinical significance and further study on the relation between expression and prognosis is warranted.

Acknowledgements This work was supported in part by grants from the Ministry of Education, Science, Sports and Culture of Japan. We thank M Yoshikawa for technical assistance and M Ohara for language assistance.

1767


1768

References 1 Sherr CJ. Tumor surveillance via the ARF-p53 pathway. Gene Dev 1998; 12: 2984–2991. 2 Sharpless NE, DePinho RA. The INK4A/ARF locus and its two gene products. Curr Opin Genet Dev 1999; 9: 22–30. 3 Serrano M, Hannon GJ, Beach D. A new regulatory motif in cellcycle control causing specific inhibition of cyclin D/CDK4. Nature 1993; 366: 704–707. 4 Sherr CJ. G1 phase progression: cycling on cue. Cell 1994; 79: 551–555. 5 Lukas J, Parry D, Aagaard L, Mann DJ, Bartkova J, Strauss M, Peters G, Bartek J. Retinoblastoma-protein-dependent cell-cycle inhibition by the tumour suppressor p16. Nature 1995; 375: 503–506. 6 Medema RH, Herrera RE, Lam F, Weinberg RA. Growth suppression by p16ink4 requires functional retinoblastoma protein. Proc Natl Acad Sci USA 1995; 92: 6289–6293. 7 Li Y, Nichols MA, Shay JW, Xiong Y. Transcriptional repression of the D-type cyclin-dependent kinase inhibitor p16 by the retinoblastoma susceptibility gene product pRb. Cancer Res 1994; 54: 6078–6082. 8 Parry D, Bates S, Mann DJ, Peters G. Lack of cyclin D-Cdk complexes in Rb-negative cells correlates with high levels of p16INK4/MTS1 tumour suppressor gene product. EMBO J 1995; 14: 503–511. 9 Hara E, Smith R, Parry D, Tahara H, Stone S, Peters G. Regulation of p16CDKN2 expression and its implications for cell immortalization and senescence. Mol Cell Biol 1996; 16: 859–867. 10 Kamb A, Gruis NA, Weaver-Feldhaus J, Liu Q, Harshman K, Tavtigian SV, Stockert E, Day RR, Johnson BE, Skolnick MH. A cell cycle regulator potentially involved in genesis of many tumor types. Science 1994; 264: 436–440. 11 Nobori T, Miura K, Wu DJ, Lois A, Takabayashi K, Carson DA. Deletions of the cyclin-dependent kinase-4 inhibitor gene in multiple human cancers. Nature 1994; 368: 753–756. 12 Drexler HG. Review of alterations of the cyclin-dependent kinase inhibitor INK4 family genes p15, p16, p18 and p19 in human leukemia–lymphoma cells. Leukemia 1998; 12: 845–859. 13 Duro D, Bernard O, Della Valle V, Berger R, Larsen CJ. A new type of p16INK4/MTS1 gene transcript expressed in B-cell malignancies. Oncogene 1995; 11: 21–29. 14 Quelle DE, Zindy F, Ashmun RA, Sherr CJ. Alternative reading frames of the INK4a tumor suppressor gene encode two unrelated proteins capable of inducing cell cycle arrest. Cell 1995; 83: 993–1000. 15 Liggett W Jr, Sewell DA, Rocco J, Ahrendt SA, Koch W, Sidransky D. p16 and p16␤ are potent growth suppressors of head and neck squamous carcinoma cells in vitro. Cancer Res 1996; 56: 4119– 4123. 16 Stott FJ, Bates S, James MC, McConnell BB, Starborg M, Brookes S, Palmero I, Ryan K, Hara E, Vousden KH, Peters G. The alternative product from the human CDKN2A locus, p14ARF, participates in a regulatory feedback loop with p53 and MDM2. EMBO J 1998; 17: 5001–5014. 17 Kamijo T, Zindy F, Roussel MF, Quelle DE, Downing JR, Ashmun RA, Grosveld G, Sherr CJ. Tumor suppression at the mouse INK4a locus mediated by the alternative reading frame product p19ARF. Cell 1997; 91: 649–659. 18 Gardie B, Cayuela JM, Martini S, Sigaux F. Genomic alterations of the p19ARF encoding exons in T-cell acute lymphoblastic leukemia. Blood 1998; 91: 1016–1020. 19 Robertson KD, Jones PA. The human ARF cell cycle regulatory gene promoter is a CpG island which can be silenced by DNA methylation and down-regulated by wild-type p53. Mol Cell Biol 1998; 18: 6457–6473. 20 Uchimaru K, Taniguchi T, Yoshikawa M, Asano S, Arnold A, Fujita T, Motokura T. Detection of cyclin D1 (bcl-1, PRAD1) overexpression by a simple competitive reverse transcriptionpolymerase chain reaction assay in t(11;14)(q13;q32)-bearing Bcell malignancies and/or mantle cell lymphoma. Blood 1997; 89: 965–974. 21 Taniguchi T, Fujita A, Takahashi S, Uchimaru K, Yoshikawa M, Asano S, Fujita T, Motokura T. Cyclin D1 overexpression detected by a simple competitive reverse transcription-polymerase chain reaction assay for lymphoid malignancies. Jpn J Cancer Res 1998; 89: 159–166.

22 Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate–phenol–chloroform extraction. Anal Biochem 1987; 162: 156–159. 23 Ponte P, Ng SY, Engel J, Gunning P, Kedes L. Evolutionary conservation in the untranslated regions of actin mRNAs: DNA sequence of a human beta-actin cDNA. Nucleic Acids Res 1984; 12: 1687–1696. 24 Hill HD, Straka JG. Protein determination using bicinchoninic acid in the presence of sulfhydryl reagents. Anal Biochem 1988; 170: 203–208. 25 Uchimaru K, Taniguchi T, Yoshikawa M, Fujinuma H, Fujita T, Motokura T. Growth arrest associated with 12-o-tetradecanoylphorbol-13-acetate-induced hematopoietic differentiation with a defective retinoblastoma tumor suppressor-mediated pathway. Leukemia Res 1998; 22: 413–420. 26 Raeymaekers L. A commentary on the practical applications of competitive PCR. Genome Res 1995; 5: 91–94. 27 Ogawa S, Hirano N, Sato N, Takahashi T, Hangaishi A, Tanaka K, Kurokawa M, Tanaka T, Mitani K, Yazaki Y, Hirai H. Homozygous loss of the cyclin-dependent kinase 4-inhibitor (p16) gene in human leukemias. Blood 1994; 84: 2431–2435. 28 Filippini G, Griffin S, Uhr M, Eppenberger H, Bonilla J, Cavalli F, Soldati G. A novel flow cytometric method for the quantification of p53 gene expression. Cytometry 1998; 31: 180–186. 29 Farrell PJ, Allan GJ, Shanahan F, Vousden KH, Crook T. p53 is frequently mutated in Burkitt’s lymphoma cell lines. EMBO J 1991; 10: 2879–2887. 30 Cheng J, Haas M. Frequent mutations in the p53 tumor suppressor gene in human leukemia T-cell lines. Mol Cell Biol 1990; 10: 5502–5509. 31 Chow VT, Quek HH, Tock EP. Alternative splicing of the p53 tumor suppressor gene in the Molt-4 T-lymphoblastic leukemia cell line. Cancer Lett 1993; 73: 141–148. 32 Yamato K, Oka T, Hiroi M, Iwahara Y, Sugito S, Tsuchida N, Miyoshi I. Aberrant expression of the p53 tumor suppressor gene in adult T-cell leukemia and HTLV-I-infected cells. Jpn J Cancer Res 1993; 84: 4–8. 33 Sugimoto K, Toyoshima H, Sakai R, Miyagawa K, Hagiwara K, Ishikawa F, Takaku F, Yazaki Y, Hirai H. Frequent mutations in the p53 gene in human myeloid leukemia cell lines. Blood 1992; 79: 2378–2383. 34 Bi S, Hughes T, Bungey J, Chase A, de Fabritiis P, Goldman JM. p53 in chronic myeloid leukemia cell lines. Leukemia 1992; 6: 839–842. 35 Wolf D, Rotter V. Major deletions in the gene encoding the p53 tumor antigen cause lack of p53 expression in HL-60 cells. Proc Natl Acad Sci USA 1985; 82: 790–794. 36 Kogawa K, Mogi Y, Morii K, Takimoto R, Niitsu Y. TGF-␤ and platelet. Jpn J Clin Hematol 1994; 35: 370–375. 37 Kawamura M, Ohnishi H, Guo SX, Sheng XM, Minegishi M, Hanada R, Horibe K, Hongo T, Kaneko Y, Bessho F, Yanagisawa M, Sekiya T, Hayashi Y. Alterations of the p53, p21, p16, p15 and RAS genes in childhood T-cell acute lymphoblastic leukemia. Leukemia Res 1999; 23: 115–126. 38 Martinez-Delgado B, Robledo M, Arranz E, Osorio A, Garciá MJ, Echezarreta G, Rivas C, Benitez J. Hypermethylation of p15/ink4b/MTS2 gene is differentially implicated among nonHodgkin’s lymphomas. Leukemia 1998; 12: 937–941. 39 Ng MH, Chung YF, Lo KW, Wickham NW, Lee JC, Huang DP. Frequent hypermethylation of p16 and p15 genes in multiple myeloma. Blood 1997; 89: 2500–2506. 40 Pinyol M, Cobo F, Bea S, Jares P, Nayach I, Fernandez PL, Montserrat E, Cardesa A, Campo E. p16INK4a gene inactivation by deletions, mutations, and hypermethylation is associated with transformed and aggressive variants of non-Hodgkin’s lymphomas. Blood 1998; 91: 2977–2984. 41 Villuendas R, Sańchez-Beato M, Martıńez JC, Saez AI, MartinezDelgado B, Garcıá JF, Mateo MS, Sanchez-Verde L, Benı´tez J, Martińez P, Piris MA. Loss of p16/INK4A protein expression in nonHodgkin’s lymphomas is a frequent finding associated with tumor progression. Am J Pathol 1998; 153: 887–897. 42 Geradts J, Andriko JW, Abbondanzo SL. Loss of tumor suppressor gene expression in high-grade but not low-grade non-Hodgkin’s lymphomas. Am J Clin Pathol 1998; 109: 669–674. 43 Klangby U, Okan I, Magnusson KP, Wendland M, Lind P, Wiman


44

45

46

47

KG. p16/INK4a and p15/INK4b gene methylation and absence of p16/INK4a mRNA and protein expression in Burkitt’s lymphoma. Blood 1998; 91: 1680–1687. Nakamaki T, Kawamata N, Schwaller J, Tobler A, Fey M, Pakkala S, Lee YY, Kim BK, Fukuchi K, Tsuruoka N, Kahan J, Miller CW, Koeffler HP. Structural integrity of the cyclin-dependent kinase inhibitor genes, p15, p16 and p18 in myeloid leukaemias. Br J Haematol 1995; 91: 139–149. Della Valle V, Duro D, Bernard O, Larsen CJ. The human protein p19ARF is not detected in hemopoietic human cell lines that abundantly express the alternative ␤ transcript of the p16INK4a MTS1 gene. Oncogene 1997; 15: 2475–2481. Cayuela JM, Gardie B, Sigaux F. Disruption of the multiple tumor suppressor gene MTS1/p16INK4a/CDKN2 by illegitimate V(D)J recombinase activity in T-cell acute lymphoblastic leukemias. Blood 1997; 90: 3720–3726. Garcia-Sanz R, Gonzalez M, Vargas M, Chillon MC, Balanzategui

A, Barbon M, Flores MT, San Miguel JF. Deletions and rearrangements of cyclin-dependent kinase 4 inhibitor gene p16 are associated with poor prognosis in B cell non-Hodgkin’s lymphomas. Leukemia 1997; 11: 1915–1920. 48 Pinyol M, Hernandez L, Cazorla M, Balbin M, Jares P, Fernandez PL, Montserrat E, Cardesa A, Lopez-Otin C, Campo E. Deletions and loss of expression of p16INK4a and p21Waf1 genes are associated with aggressive variants of mantle cell lymphomas. Blood 1997; 89: 272–280. 49 Elenitoba-Johnson KS, Gascoyne RD, Lim MS, Chhanabai M, Jaffe ES, Raffeld M. Homozygous deletions at chromosome 9p21 involving p16 and p15 are associated with histologic progression in follicle center lymphoma. Blood 1998; 91: 4677–4685. 50 Mekki Y, Catallo R, Bertrand Y, Manel AM, Ffrench P, Baghdassarian N, Duhaut P, Bryon PA, French M. Enhanced expression of p16ink4a is associated with a poor prognosis in childhood acute lymphoblastic leukemia. Leukemia 1999; 13: 181–189.

1769