Leukemia (1997) 11, 984–989 1997 Stockton Press All rights reserved 0887-6924/97 $12.00 Extensive analysis of the retinoblastoma gene in adult T cell leukemia/lymphoma (ATL) Y Hatta1, Y Yamada2, M Tomonaga2 and HP Koeffler1 1Division of Hematology/Oncology, Department of Medicine, Cedars-Sinai Research Institute, UCLA School of Medicine, Los Angeles, CA, USA; and 2Department of Hematology, Atomic Disease Institute, Nagasaki University, School of Medicine, Nagasaki, Japan The retinoblastoma susceptibility gene (Rb) plays a key role in Tax;38 a long period of clinical latency (20–30 years) precedes regulating the cell cycle in association with cyclins and cyclin- the development of ATL;39,40 only a small percentage of dependent kinases (CDKs). Alteration of the Rb gene as well 40 as CDK inhibitors (CDKIs) leads to deregulated cellular growth HTLV-I-infected individuals develop this malignancy; and which promotes cancer formation. We examined the genomic ATL cells are monoclonal. configuration of the entire Rb gene in 40 primary adult T cell The missense inactivating mutations of the p53 gene have leukemias/lymphomas (ATL) and two ATL cell lines by South- been observed in ATL.41–44 We tested for the inactivation of ern blotting and also by polymerase chain reaction-single several CDK inhibitors (CDKIs) including p15INK4B, p16INK4A, strand conformation polymorphism (PCR-SSCP) analyses. p18INK4C, p19INK4D and p27KIP1, and have found deletions of Homozygous loss of exon 1 was identified in one of 21 acute the p16INK4A gene in 10 of 37 cases (27%); deletions of the ATL, one of 15 chronic ATL, and none of four lymphomatous INK4B 45 ATL samples. No point mutations were identified. Previously, p15 gene in seven of 37 cases (19%); alterations of the we found that 10 of these same ATL samples had alterations of p27KIP1 gene in two of 42 ATL samples (5%);46 and no abnor- either p16INK4A (homozygous deletion) or p27kip1 (homozygous malities were found in the p18INK4C,78 and p19INK4D genes47 deletion or point mutation). Although the numbers are very low, in 43 samples. none of the samples with an aberrant Rb gene had an altered Altered expression and structural abnormalities of Rb have CDKI and vice versa, suggesting that both genes probably 29,30,48–58 operate in a common pathway and alteration of either can pro- been identified in several hematologic malignancies, vide these cells with a growth advantage. but no information exists regarding inactivation of the Rb gene Keywords: cyclin dependent kinase inhibitor; Rb; ATL; tumor sup- in ATL. The Rb gene is localized to human chromosomal band pressor gene; cell cycle 13q14.8,59 Chromosomal abnormalities of 13q14 have been detected in a variety of tumors including osteosarcomas.60 In contrast, ATL rarely has chromosomal abnormalities at 13q.61 Introduction Nevertheless, studies have shown that mutations and deletions of the Rb gene occur in other types of tumors without a detect- The retinoblastoma susceptibility gene (Rb) encoding a able cytogenetic abnormality at chromosome 13q.48,53,62 nuclear phosphoprotein is a prototype of tumor suppressor Therefore, we performed analysis for alterations of the Rb genes.1,2 The phosphorylation of the Rb gene is regulated in gene using Southern blotting and polymerase chain reaction- a cell cycle-specific manner in association with the cyclin– single strand conformation polymorphism (PCR-SSCP) for 40 cyclin dependent kinase (CDK) complex.3–7 Inactivation of the primary ATL samples and two ATL cell lines. Rb gene has been implicated in the pathogenesis of a wide variety of human solid tumors, including osteosarcomas and soft tissue sarcomas,1,8–15 carcinomas of the breast,16–18 Materials and methods lung,19–23 prostate24 and bladder,25 and of course retinoblas- tomas.26–28 High level of expression of the Rb protein or Rb Samples gene amplification has also been reported in some tumors.15,29,30 Although development of cancer is a multi-step Cells were collected from 40 Japanese ATL patients before process, these findings suggest that Rb abnormalities com- their treatment and from two ATL cell lines (TSUKA and monly occur in many different tumor types and may help to SO4).63 Of the samples from primary ATL patients, 21 cases explain the deregulated growth control associated with malig- were acute leukemias, four were lymphomas, and 15 were nancy. chronic ATL. The clinical subtypes of ATL were based on the Adult T cell leukemia/lymphoma (ATL) is an aggressive, diagnostic criteria proposed by the Lymphoma Study Group of fatal malignancy of mature CD4+ T lymphocytes.31–33 The Japan.64 One sample was obtained from an abdominal mass, human T cell lymphotropic virus type I (HTLV-I) has been another from the tonsil, two from lymph nodes, one from recognized as the etiologic agent of ATL;34–36 and the viral pleural effusion, and the remaining samples were from the factor responsible for the leukemia induction was proposed to peripheral blood. In each of the samples, at least 70% of the be Tax, a regulatory protein encoded by the HTLV-I gen- cells were morphologically ATL cells. As a control, normal ome.37 However, the mechanism by which HTLV-I induces human bone marrow was obtained after informed consent. ATL is not well understood. The following suggests that additional cellular events are required for the development of the malignant phenotype: ATL cells usually do not express Southern blot analysis After five or 10 mg of DNA was digested with HindIII, it was separated on 0.7% agarose gels and transferred to Biotrans Correspondence: Y Hatta, Division of Hematology/Oncology, Depart- ment of Medicine, Cedars-Sinai Research Institute, UCLA School of Nylon Membranes (ICN, Aurora, OH, USA). The Rb probes Medicine, 8700 Beverly Blvd, B-208, Los Angeles, CA 90048, USA were isolated by KpnI and EcoRI digestion of a plasmid con- Received 28 August 1996; accepted 26 February 1997 taining a 4.15 kb cDNA of the Rb gene which was kindly pro- Alterations of Rb in ATL Y Hatta et al 985 vided by Dr T Dryja (Harvard Medical School, Boston, MA, and one of 15 chronic ATL samples) using the 59 Rb cDNA USA). The resulting Rb cDNA probes were 1.15 kb (59 KpnI– probe (Figure 1). Four lymphomatous ATL samples showed no EcoRI fragment) and 3.8 kb (39 EcoRI fragment). Each of the alterations of the gene. No DNA rearrangements or deletions DNA samples were sequentially hybridized with the 39 and were observed in any of the cases when the blots were rehy- 59 Rb probes. A probe for myeloperoxidase (MPO) was used bridized with the 39 cDNA probe (Table 2). as an internal control.65 Southern blot hybridization, 32P-labe- The entire Rb coding sequence and 573 bp of the promoter ling of the probes, and autoradiography were accomplished in region starting at 692 bp upstream of the coding sequence essentially the same manner as previously described.66 Probes were analyzed for point mutations by using PCR-SSCP. were labeled with 32P by the random primer method67 using Aberrant shifts on SSCP were identified in three samples; how- the Random Primers DNA Labeling System (BRL, Life Techno- ever, DNA sequence analysis revealed that all these samples logies, Gaithersburg, MD, USA). Allele losses were ascer- had the base changes in the introns far from the exon–intron tained by visual inspection. When visible reduction of radio- boundaries and the base changes did not code for new intron graphic signal was not obvious, densitometry was performed splice sites. (Ultrascan XL laser densitometer; Pharmacia/LKB, Freiburg, Germany) to confirm our interpretation. A ratio of MPO/Rb hybridizing bands of at least four-fold and two-fold were Discussion required for a sample to be classified as having either a homo- zygous or hemizygous Rb deletion, respectively. To our knowledge, this is the first study to test for alterations of the entire Rb gene in ATL. Southern blot and SSCP analyses were performed on the entire coding region of Rb in 40 pri- PCR-SSCP analysis mary ATL and two ATL cell lines. Homozygous deletions of exon 1 of Rb were found in samples from two primary cases Oligonucleotide primer pairs were designed to amplify frag- for an incidence of 5%. Samples with homozygous deletions ments containing each of exons 1–27 as well as the promoter of the Rb gene had a normal SSCP pattern. This is not surpris- region of the Rb gene, according to recently published ing because PCR will amplify DNA from contaminated normal sequence data.68 Most of the PCR fragments included at least tissue. All of our samples were obtained before treatment; 24 bp of the flanking intron regions surrounding each exon. therefore, therapy did not cause the genetic changes observed. To increase the sensitivity of the SSCP analysis, the PCR frag- Although this frequency of alteration of Rb is almost compara- ments were further digested by appropriate restriction endonu- ble to those of other hematologic malignancies,49,52,53,56,58 it cleases in order to reduce their size to less than 250 bp; sensi- is lower than those of several other tumors (eg 29% in tivity of mutation detection by SSCP is inversely proportional osteosarcoma15). Because loss of expression of Rb protein was to the size of the amplified products.69 The restriction recently identified in five of 10 ATL samples,70 we may have enzymes were added directly to the PCR at 37°C, (30°C for underestimated the frequency of samples with altered SmaI and 65°C for Taq) for 2 h. The primer sequences, Mg expression of Rb. Some investigators have reported ‘dimini- concentrations, annealing temperature, restriction enzymes, shed’ expression of the Rb protein with no apparent abnor- and fragment lengths are listed in Table 1.