ONCOGENIC ABERRATIONS IN THE p53 PATHWAY ARE ASSOCIATED WITH A HIGH S PHASE FRACTION AND POOR PATIENT SURVIVAL IN B-CELL NON-HODGKIN’S LYMPHOMA Trond STOKKE 1 * , Eivind GALTELAND 1 , Harald HOLTE 2 , Lars SMEDSHAMMER 1 , Zhenhe SUO 3 , Erlend B. SMELAND 4 , Anne-Lise BØRRESEN-DALE 5 , Paula DEANGELIS 6 and Harald B. STEEN 1 1 Department of Biophysics, The Norwegian Radium Hospital, Montebello, Oslo, Norway 2 Department of Oncology, The Norwegian Radium Hospital, Montebello, Oslo, Norway 3 Department of Pathology, The Norwegian Radium Hospital, Montebello, Oslo, Norway 4 Department of Immunology, The Norwegian Radium Hospital, Montebello, Oslo, Norway 5 Department of Genetics, The Norwegian Radium Hospital, Montebello, Oslo, Norway 6 Department of Pathology, The Norwegian National Hospital, Oslo, Norway The implications of aberrations in the p53 pathway for induction of apoptosis and regulation of S phase entry, and for patient survival, were investigated in 83 B-cell Non- Hodgkin’s lymphomas. Eight cases had missense mutations in exons 5, 7, 8 and 9 as revealed by constant denaturant gel electrophoresis and sequencing. Fifteen cases had lost 1 TP53 allele as revealed by fluorescent in situ hybridization and comparative genomic hybridization. Ten cases expressed high levels of p53 as assessed by immunoblotting and immu- nohistochemistry. S phase fractions were higher, apoptotic fractions were the same and survival times were shorter in all aberration groups compared with the cases with no TP53/p53 aberrations. Since many tumors had more than one TP53/p53 aberration, the tumors were divided into groups with the following characteristics: no TP53/p53 aberrations; loss of one TP53 allele only (9 cases), TP53 point mutation (8 cases), high-level p53 expression and no TP53 mutation (3 cases). Tumors from the 3 latter groups had higher median S phase fractions (5%, 7.6%, and 5%, respectively, p<0.02) than the cases without any aberrations (1.1%), and survival time for these patients was much shorter (relative risks of 5.9, 8.9, and 6.6, respectively, p<0.003). Apoptotic fractions were similar in all these groups (p0.09). Multivariate analysis showed that the presence of TP53/p53 aberrations is a strong and independent prognostic parameter in B-cell Non-Hodgkin’s lymphoma. Int. J. Cancer (Pred. Oncol.) 89:313–324, 2000. © 2000 Wiley-Liss, Inc. TP53 point mutations and deletions, and abnormal p53 expres- sion, are the most common aberrations detected in human tumors (Hollstein et al., 1994). Hence, the p53 pathway is involved in controlling cell growth, which is deregulated during tumorigene- sis. Tumors are often characterized by a high rate of cell division, but it has also been suggested that a reduced apoptotic rate in tumor cells is another way to increase the growth rate of a tumor. Cell proliferation, as measured by S phase fraction, has been shown to be a strong prognostic parameter in Non-Hodgkin’s lymphoma (NHL) (Stokke et al., 1998b; Holte et al., 1999, and references therein); wt p53 inhibits cell proliferation by arresting cells in the G 1 phase of the cell cycle (Kuerbitz et al., 1992; Guillouf et al., 1995; Paules et al., 1995; for reviews see Sherr, 1996; Levine, 1997), and this effect is, at least in some cases, mediated through upregulation of p21/waf1, a cyclin-dependent kinase inhibitor. A role in the G 2 arrest imposed when DNA is damaged (Paules et al., 1995), or incompletely replicated (Taylor et al., 1999), has also been suggested. In vitro studies have also suggested that p53 is involved in triggering apoptosis (Friedlander et al., 1996; Levine, 1997; Ryan and Vousden, 1998); bax and bcl-2, which promote and inhibit apoptosis, respectively, have been implicated in performing the downstream functions of p53 in this scenario because wt p53 increases and decreases, respectively, transcription from the pro- moters of these genes (Miyashita et al., 1994). However, the same group has reported that apoptosis and bax/bcl-2 expression is independent of TP53 mutation in B-cell chronic lymphocytic leu- kemia (Thomas et al., 1996). Constitutive expression of p53 has often been assumed to reflect mutations in TP53. However, p53 may be expressed at high levels in normal cells after different types of damage (Maltzman and Czyzyk, 1984; Merritt et al., 1994), showing that expression is not necessarily an effect of genetic errors acquired during carcinogen- esis. It has been suggested that p53 “senses” particular types of DNA damage, resulting in either arrest of proliferation until the damage is repaired, or, if repair is not successful, apoptosis (Sherr, 1996). In this scenario, the cell must have ways of monitoring the damage (and thereby its repair), such that p53 is turned off and proliferation resumed when damage is repaired, or apoptosis in- duced if the damage is not corrected and p53 levels remain high. However, if the downstream phenotypic and/or genotypic effect(s) of p53 do not occur in tumor cells, e.g., because the p53 protein is inactive (mutant), or if the downstream molecular activators or effectors of (wt) p53 are non-functional (tumor-suppressor genes) or over-ridden by an overexpressed oncogene-product, no feed- back is obtained, and p53 remains stabilized. The “damage” lead- ing to induction of p53 may be some of the genetic aberrations that have occurred in the tumor cells, including mutations in and/or loss of TP53 itself. It is not known if loss of one TP53 allele (no mutation on the remaining allele) gives a reduced potential for p53 induction (gene dosage effect), which could in principle contribute to malignancy in the same way as mutations and clonal p53 expression. In vitro studies are of great importance for establishing the exact biochemical interactions that take place in the pathways control- ling entry into S phase and apoptosis, and for suggesting interac- tions that could be important in vivo. However, answers obtained from experiments employing cell lines will not necessarily reflect the in vivo situation because the interactions with normal cells are lost and because cell lines typically acquire additional aberrations as they grow in vitro. The in vivo effects of specific gene products can in some cases be determined in transgenic or knockout mice. However, at present, we must rely on correlative studies to deter- mine the in vivo role(s) of, e.g., p53 in humans. We have investigated retrospectively if NHL’s of B cell type with different p53/TP53 aberrations are characterized by particular cellular phenotypes or clinical characteristics. These p53/TP53 aberrations were all associated with a high S phase fraction, poor Grant sponsor: The Norwegian Cancer Society. *Correspondence to: Department of Biophysics, The Norwegian Radium Hospital, Montebello, 0310 Oslo. Fax number: 47 22934270. E-mail: tstokke@radium.uio.no Received 26 November 1999; Revised 11 April 2000; Accepted 17 April 2000 Int. J. Cancer (Pred. Oncol.): 89, 313–324 (2000) © 2000 Wiley-Liss, Inc. Publication of the International Union Against Cancer