The Ikaros gene has been shown to be essential for the development of B lymphocytes and to function as a tumour suppressor in T lymphocytes (Georgopoulos et al, 1994; Winandy et al, 1995; Wang et al, 1996). The finding that Ikaros functions as a tumour suppressor in mice prompted us (Nakayama et al, 1999; Nakase et al, 2000) and others (Sun et al, 1999a,b,c) to examine whether alterations in Ikaros activity were important in the pathology of human haematological malignancies. We have demonstrated that the dominant-negative isoform of Ikaros, Ik-6, was overexpressed in patients with blast crisis of chronic myelogenous leukaemia (CML), suggesting that Ikaros may be involved in the disease progression of CML (Nakayama et al, 1999). Moreover, we found Ik-6 over-expression in patients with B-cell acute lymphoblastic leukaemia (ALL) (Nakase et al, 2000). Ikaros is the founding member of a family of homologous genes, including Aiolos (Morgan et al, 1997; Hosokawa et al, 1999a; Wang et al, 1999) and Helios (Hahm et al, 1998; Kelley et al, 1998; Hosokawa et al, 1999b), which are co-expressed with Ikaros in several haematopoietic cell types. We, therefore, hypothesized that human haematological malignancies involve mutations not only of the Ikaros gene, but also in one or more of the other Ikaros family genes. As a result, we reported over-expression of short isoforms of Helios in the HD-Mar cell line and in a patient with T-cell ALL (Nakase et al, 2002). In the current study, we examined a panel of samples obtained from patients with a range of haematological malignancies, including chronic and acute forms of adult T-cell leukaemia/lymphoma (ATLL), and determined the expression levels of the three Ikaros family genes: Ikaros, Aiolos and Helios. Human T lymphotropic virus-I (HTLV-I) has a prolonged latency period of decades before clinical symptoms appear. Mutations in p53, p15, p16 and the Fas gene, and also increased methylation of the cyclin-dependent kinase inhibitor 2A (CDKN2A) gene have been shown to be associated with the progression of ATLL (Cesarman et al, 1992; Hatta et al, 1995; Maeda et al, 1999; Nosaka et al, 2000). Here, we show the over-expression of short isoforms of Helios in patients with chronic and acute forms of ATLL.
Summary. In previous studies, we demonstrated an over-expression of the dominant-negative isoform of the transcription factor Ikaros in patients with blast crisis of both chronic myelogenous leukaemia and B-cell acute lymphoblastic leukaemia (ALL). Recently, we reported an over-expression of the short isoforms of Helios, which is one of the members of the Ikaros gene family, in a patient with T-cell ALL. In the present study, we found over-expression of short isoforms of Helios in human T lymphotropic virus-I (HTLV1)-infected patients who had developed chronic and acute forms of adult T-cell leukaemia/lymphoma. In contrast, we could not detect any over-expression of short isoforms of Helios in healthy HTLV1 carriers. By Southern blotting, we detected a small deletion in the Helios gene locus of adult T-cell leukaemia/lymphoma patients. The present results suggest that Helios gene abnormalities might be one of the important mechanisms in the disease progression of HTLV1 infection.
Materials and methods
Reverse transcription polymerase chain reaction (RT-PCR) and sequencing analysis. Peripheral blood samples were obtained from normal volunteers (n = 10), healthy carriers of HTLV-I (n = 14) and patients with ATLL (n = 19; 15 acute, one lymphoma, two chronic and one smoldering) after informed consent was obtained. RT-PCR and sequencing analysis were performed as previously described (Nakayama et al, 1999; Nakase et al, 2000, 2002).
Southern blotting. Southern blotting was performed as previously described (Nakayama et al, 1999; Nakase et al, 2000, 2002). The human Helios cDNA, used as a hybridization probe, was amplified by RT-PCR, confirmed by sequencing analysis, and labelled using a high prime DNA labelling kit (Boehringer Mannheim, Indianapolis, IN, USA) with [α-32P]-dCTP.
As previously reported, in normal volunteers, the predominant isoforms of Ikaros family expressed were Ik-1 and Ik-2,3 (Molnar et al, 1996), Aio-1 (Hosokawa et al, 1999a), and Hel-1 and Hel-2 (Hosokawa et al, 1999b). Surprisingly we found that over-expression of the short isoforms of Helios in ATLL patients occurred at a high frequency (Fig 1). We examined 19 ATLL patients (15 acute, one lymphoma, two chronic and one smoldering) and five of them (four acute and one chronic) over-expressed short isoforms of Helios (Fig 1). In addition, there appeared to be a decreased expression of Hel-1 and Hel-2 in nine of these patients (Fig 1). Patient 8, diagnosed with acute ATLL, expressed two isoforms of Helios, Hel-5 and Hel-6, which were identical to the short isoforms over-expressed in HD-Mar cells (Nakase et al, 2002). Hel-5 and Hel-6 lack exons 2–4, and exons 2–4 and 6, respectively, leading to loss of three N-terminal zinc fingers. In the case of Ikaros, isoforms that have fewer than two N-terminal zinc fingers have a dominant-negative effect on the DNA binding activity of larger Ikaros isoforms. We, therefore, presume that, in a similar manner, these short Helios isoforms that have fewer than two N-terminal zinc fingers may have a dominant-negative effect on the DNA binding properties of Hel-1 and Hel-2.
Patients 8, 12, 15 and 19 (Fig 1) with acute ATLL, and also patient 2 (Fig 1) with chronic ATLL, showed an over-expression of the short isoform of Helios (Hel-5). In contrast, we could not detect any over-expression of the short isoforms of Helios in normal volunteers (n = 10) and healthy carriers of HTLV-I (n = 14) (Fig 2 and data not shown), although we found isoforms of Aiolos (Aio-2, -3, -4 and -5) and Helios (Hel-3 and -4) as minor populations (Nakase et al, 2002). We interpreted these results to indicate that a deregulation of Helios expression, which causes the short isoforms to become overexpressed, is an event that can occur during the chronic phase of ATLL and can either contribute to or directly cause the onset of the acute-phase ATLL.
We performed Southern blotting analysis to assess whether a genetic mutation could be correlated with the over-expression of the short isoforms of Helios in patients with ATLL (Fig 3). We could not detect any differences in the restriction fragments of genomic DNA from normal volunteers, Jurkat cells (a human T-cell leukaemia line), HD-Mar cells, and patients with short isoforms of Helios (Fig 1, lanes 2, 8, and 19) after EcoRI, HindIII, and PstI restriction enzyme digestion (data not shown). However, one of the bands observed by EcoRV and ScaI digestion of the DNA from normal volunteers and Jurkat cells, was at a markedly decreased intensity from similarly digested DNA from HD-Mar cells and patients with the short isoforms of Helios (Fig 3) (Nakase et al, 2002). Densitometrical analysis confirmed that there was a decreased intensity of this band in both HD-Mar cells and patients with the short isoforms of Helios compared with both Jurkat cells and normal volunteers. In the EcoRV digest, the ratios of the upper band with a markedly reduced intensity a to lower band b were as follows: Jurkat, 1·50; HD-March, 0·72; patient 2, 0·97; patient 19, 0·77; patient 8, 0·81 (Fig 3). In the ScaI digest, the ratios of upper band c to the lower band with a markedly reduced intensity d were as follows: normal volunteer, 1·72; patient 8, 2·37; patient 19, 3·42; patient 2, 2·73; HD-March, 3·88 (Fig 3). Contamination by normal cells may explain the variable decrease in the intensity of the markedly reduced band intensity observed in patients with the short isoforms of Helios. These results suggest that there may be a small deletion in the Helios gene locus of patients with the short isoforms of Helios.
The fairly high frequency with which over-expression of the short isoforms of Helios was observed (4/15 of acute samples) suggests that altered expressions of the Helios isoforms may be one of the important mechanisms driving monoclonal T-cell proliferation in HTLV-I infection. Recently, we found an over-expression of the short isoforms of Helios, Hel-7 and Hel-8, in a patient with T-cell ALL (Nakase et al, 2002). However, the frequency was not nearly as striking as in this study of ATLL patients. This suggests either that the Helios locus is particularly susceptible to mutation or that over-expression of short isoforms of Helios confer a particular selective advantage in HTLV-I infection. While Helios expression was found to differ, Ikaros expression appeared to be mostly normal in the present study. These results demonstrate that mutations in the Helios locus affect T-cell homeostasis in the absence of mutations in the Ikaros locus. Thus, over-expression of short isoforms of either Ikaros or Helios is sufficient to perturb lymphocyte homeostasis. Further characterization of the short isoforms of Helios and also of the differential function of Ikaros family proteins in human haematological malignancies is warranted.
We thank Dr Makoto Takeuchi (National Sanatorium Minami Okayama Hospital), Dr Akira Miyata (Chugoku Central Hospital of the Mutual Aid Association of Public School Teachers), Drs Kensuke Kojima and Masamichi Hara (Ehime Prefectural Central Hospital), Dr Masaki Yasukawa (Department of Medicine, University of Ehime), and Drs Hidetaka Takimoto and Seiichi Yorimitsu (Kochi Municipal Central Hospital) for support.