Independent induction of senescence by p16INK4a and p21CIP1 in spontaneously immortalized human fibroblasts.

In this work, we address the question of whether replicative senescence can be induced in immortal nontumorigenic human fibroblasts. The immortal fibroblasts used in this study were derived from two Li-Fraumeni (LF) patients who carry in their germ line one wild-type and one mutant p53 allele. Both immortal lines have lost the wtp53 allele and express no detectable p16INK4a protein, although they carry the p16INK4a gene. In contrast to immortal human fibroblasts, senescent human fibroblasts have a low content of 5-methyl-cytosine in their DNA. This observation suggested the possibility that a demethylating agent could revert the immortal phenotype and induce replicative senescence in the immortal cell lines. Cells of the two LF lines were exposed to the demethylating agent 5-aza-2'-deoxycytidine. Within 6 days, all cells were growth arrested and showed the enlarged and flat morphology characteristic of senescent cells, an accumulation of lipofuscin granules and senescence-associated beta-galactosidase activity at pH6, both biomarkers for senescence. Immunoblots of 5-aza-2'-deoxycytidine-treated cells showed a greatly increased expression of p16INK4a protein but no detectable change in the expression of p21CIP1, a gene known to be strongly expressed in senescent normal human fibroblasts. In two other experimental series, cells of the two LF lines were infected with retroviral constructs encoding either p16INK4a or p21CIP1. Each of the transduced genes induced senescence without affecting the expression of the other endogenous gene. The results show that induction of senescence in immortal LF fibroblasts can occur by different pathways: (a) by demethylation-dependent pathways that induce the expression of p16INK4a; and (b) by demethylation-independent pathways involving the expression of p21CIP1. The induction of senescence by p16INK4a and p21CIP1 occurred equally in the two human immortal fibroblast lines, which differed in the length of their telomeres and the activity of their telomerase.

Dissemination and tissue invasiveness in murine acute leukemia associated with acquisition of p53 mutation and loss of wild-type p53.

Approximately 60% of mice treated with split-dose radiation develop leukemias that disseminate widely through the body, whereas 40% of the treated mice incur leukemias that are contained entirely within the thymus. We studied the status of p53 in non-cultured samples of thymic leukemias and in cell lines established from these leukemias. In those mice with disseminated disease, primary samples were also obtained from visceral leukemic organs, and cell lines were established from these leukemic organs for further study. Using single-strand conformation polymorphism (SSCP), nucleic acid sequencing, and immunochemical analysis, we found that mutation of both p53 alleles occurred in leukemic cell lines developed from nine of 10 disseminated leukemias; mutation of one p53 allele with the other remaining wild-type occurred in one disseminated leukemia. A p53 mutation unique for each mouse was found in all cell lines established from the different leukemic organs of each mouse. The same mutation was also found in the non-cultured leukemic tissues of each mouse, indicating that the mutations originated in vivo and were clonal. Seven of seven non-disseminating thymomas possessed wild-type p53 only. Hence, in vivo dissemination and tissue invasiveness were associated with the loss of wild-type p53 by mutation of both alleles or by mutation and loss of heterozygosity, as revealed by studies of cell lines established from them. The selective in vivo dissemination of leukemia cells possessing p53 mutations had a parallel in vitro. Leukemia cell lines from mice harboring disseminating leukemia were established more readily (success rate greater than 80%) than lines from mice harboring thymic nondisseminating leukemia (success rate less than 10%). Additionally, while mice with disseminating leukemia harbored a mixture of wild-type and mutant p53-encoding thymoma cells, only cell lines possessing mutant p53 became established in culture. Mutations found in thymoma cell lines were always detectable by SSCP and sequencing of DNA extracted from non-cultured thymoma tissue. However, in non-cultured leukemic tissue of visceral organs, the clonal p53 mutations found in cell lines established from them were often not detectable by SSCP or sequencing but were detectable by immunochemical analysis or polymerase chain reaction amplification. This indicates an unexpected degree of masking of mutant genes by wild-type genes present in the leukemic tissue. Masking was evident even in leukemic organs that were grossly larger than normal organs. Hence, routine screening of leukemic tissue by SSCP and sequencing may result in a highly significant underestimation of the incidence of p53 mutations.(ABSTRACT TRUNCATED AT 400 WORDS)

Elevated levels of wild-type p53 induced by radiolabeling of cells leads to apoptosis or sustained growth arrest.

BACKGROUND: The tumor suppressor protein p53 regulates progression through the checkpoint between the G1 and S phases of the cell cycle in response to radiation- or drug-induced DNA damage. We have examined potential p53-mediated effects of metabolically labeling cultured mammalian cells with [35S]methionine and [3H]thymidine, methods that are commonly used to study the biochemical properties, synthesis, processing and degradation of proteins and the replication of DNA in proliferating cells. RESULTS: Wild-type p53 protein concentrations rapidly increased to high levels following metabolic radiolabeling of cells, as determined by four distinct assays. The increased concentration of wild-type p53 resulted in apoptosis of normal human peripheral blood lymphocytes and of murine T-cell acute lymphoblastic leukemia cells. In leukemia cells containing no p53, or only mutant p53 alleles, p53 protein levels were not increased and the cells did not undergo apoptosis in response to radioactive labeling. Radiolabeling of human diploid fibroblasts resulted in a prolonged growth arrest that was maintained for nearly three weeks. CONCLUSIONS: The results of experiments employing radiolabeling techniques to characterize various aspects of cellular physiology may be seriously influenced by the induction of aberrant cell-cycle arrest and/or apoptosis mediated by wild-type p53. Furthermore, our observations suggest that stabilization of wild-type p53 in response to irradiation may not act primarily to facilitate the repair of DNA damage by inducing a transient G1-phase arrest, but rather to ensure genetic stability through sustained cell-cycle arrest or apoptotic death of the damaged cells.

Gain-of-function mutations of the p53 gene induce lymphohematopoietic metastatic potential and tissue invasiveness.

Leukemia cell infiltration and the induction of lethal hematopoietic disease in immune-deficient SCID mice transplanted with human T cell acute lymphoblastic T leukemia (T-ALL) cells occurred only when the cells possessed mutant p53 genes and lacked a wild-type allele or when T-ALL cells lacking p53 protein were infected with specific mutant p53 genes. A series of six mutant p53 genes were cloned from relapse T-ALL-derived cell lines and were constructed into defective retroviral expression vectors. Viruses encoding mutant p53 proteins were used to infect relapse T-ALL cells in a study designed to compare their pathogenic potency. The mutant p53 genes possessed a distinct hierarchy in vivo and in vitro: mutants inducing the greatest increase in proliferation of different T-ALL lines in vitro and colony formation in methylcellulose cultures also induced tissue invasiveness of infected T-ALL cells in vivo. Mutant p53 gene transfer to a cell line lacking p53 protein showed that the more potent p53 mutants possessed a distinctive dominant oncogenic activity in vitro and in vivo. The dominant oncogenic activity of these mutant p53 proteins was not dependent on the presence of and on complex formation with wild-type p53 protein. These "hot" p53 mutations thus represent bona fide gain-of-function mutations. Infection of p53-negative T-ALL cells with viruses encoding gain-of-function mutant p53 genes resulted in the acquisition of metastatic potential and tissue invasiveness. Taken together, our results suggest that specific mutant p53 genes play a role in the generation of lymphohematopoietic metastatic potential and tissue invasiveness as assayed in SCID mice, whereas the expression of wild-type p53 is capable of keeping this metastatic potential in check.

Nonhereditary p53 mutations in T-cell acute lymphoblastic leukemia are associated with the relapse phase.

We have previously reported that greater than 60% of human leukemic T-cell lines possess mutations in the p53 tumor suppressor gene. To determine whether T-cell acute lymphoblastic leukemia (T-ALL) patient samples possess p53 mutations, we screened peripheral blood-and bone marrow-derived leukemia samples, taken at diagnosis and at relapse, for p53 mutations. Exons 4 through 9 and selected intron regions of the p53 gene were analyzed using polymerase chain reaction-single-strand conformation polymorphism and direct sequencing. p53 mutations were found in 0 of 15 T-ALL diagnosis samples, as compared with 10 of 36 (28%) T-ALL relapse samples. To determine whether p53 mutations play a role in the recurrence (relapse) of T-ALL, two special groups of T-ALL patients were studied: (1) a group of 8 relapse patients whose disease was refractory to chemotherapeutic treatment, and (2) a group of 6 "paired" T-ALL cell samples from patients for whom we possess both diagnosis and relapse samples. Three of 8 relapsed patients (37.5%) whose disease was refractory to the reinduction of remission by chemotherapy possessed missense mutations of the p53 gene. All 3 cases had mutations in exon 5. Among the paired samples, 3 of 6 patients harbored p53 mutations at disease recurrence, but possessed only wild-type p53 alleles at diagnosis. One case had mutation on exon 4, 1 case in exon 5, and 1 case in exon 8 with loss of heterozygosity. These data clearly indicate that recurrence of T-ALL is associated with missense mutations in p53. Our results indicate that (1) mutations of p53 do occur in T-ALL in vivo, and such mutations are associated with the relapse phase of the disease; and (2) p53 mutation is involved in the progression of T-ALL. This conclusion is supported by our observation that the introduction of T-ALL-derived mutant p53 expression constructs into T-ALL cell lines further increases their growth rate in culture, enhances cell cloning in methylcellulose, and increases tumor formation in nude mice.

P53 mutation in acute T cell lymphoblastic leukemia is of somatic origin and is stable during establishment of T cell acute lymphoblastic leukemia cell lines.

Samples donated by patients with T cell acute lymphoblastic leukemia (T-ALL) were screened for mutations of the p53 tumor suppressor gene. Peripheral blood cells of T-ALL relapse patient H.A. were found to possess a heterozygous point mutation at codon 175 of the p53 gene. To determine whether this was an inherited mutation, a B cell line (HABL) was established. Leukemic T cell lines (HATL) were concurrently established by growing peripheral blood leukemic T cells at low oxygen tension in medium supplemented with IGF-I. Previously we had shown that > 60% of leukemic T cell lines possessed mutations in the p53 gene (Cheng, J., and M. Hass. 1990. Mol. Cell. Biol. 10:5502), mutations that might have originated with the donor's leukemic cells, or might have been induced during establishment of the cell lines. To answer whether establishment of the HATL lines was associated with the induction of p53 mutations, cDNAs of the HATL and HABL lines were sequenced. The HATL lines retained the same heterozygous p53 mutation that was present in the patient's leukemic cells. The HABL line lacked p53 mutations. Immunoprecipitation with specific anti-p53 antibodies showed that HATL cells produced p53 proteins of mutant and wild type immunophenotype, while the HABL line synthesized only wild-type p53 protein. The HATL cells had an abnormal karyotype, while the HABL cells possessed a normal diploid karyotype. These experiments suggest that (a) p53 mutation occurred in the leukemic cells of relapse T-ALL patient HA; (b) the mutation was of somatic rather than hereditary origin; (c) the mutation was leukemia associated; and (d) establishment of human leukemia cell lines needs not be associated with in vitro induction of p53 mutations. It may be significant that patient HA belonged to a category of relapse T-ALL patients in whom a second remission could not be induced.

Role of the p53 tumor suppressor gene in the pathogenesis and in the suppression of acute lymphoblastic T-cell leukemia.

The tumor suppressor gene p53 has been shown to be mutated in 50% of acute lymphoblastic T-cell-leukemia (T-ALL) cell lines, all of which were established from T-ALL cases in relapse. In these lines both alleles of the p53 gene were independently affected by point mutation. In contrast, in human solid tumors possessing a mutated p53 allele, the second wild-type p53 suppressor allele is often lost by deletion rather than altered by mutation. This suggests that in T-ALL cell lines, the product encoded by the second mutated allele provides the cells with an additional oncogenic stimulus, beyond the loss of suppressive activity. While different p53 mutations have been shown to possess differential oncogenic potential in the p53 plus ras cotransformation assay, in T-ALL cells different mutations may in addition possess distinct functions, further contributing to the tumorigenic phenotype.

Suppression of acute lymphoblastic leukemia by the human wild-type p53 gene.

Independent mutations in both alleles of the p53 tumor suppressor gene are a frequent finding in human T-cell acute lymphoblastic leukemia (T-ALL) cell lines and in the cells of some T-ALL patients in relapse. One major goal of studying the status of p53 (and other tumor suppressor genes) in human cancer is to facilitate the suppression of the tumorigenic phenotype through the restoration of the expression of the wild-type allele. While the efficient insertion of a suppressor into all cells of solid/metastatic human tumors may at present be impossible, insertion into leukemia cells may be feasible due to the accessibility of the leukemia cells in the body. To examine the feasibility of suppressing the tumorigenicity of human T-leukemia cells, the human T-ALL cell line Be-13, which lacks endogenous p53 protein, was infected with a recombinant retrovirus encoding the wild-type allele of human p53 (hwtp53). Expression of p53 reduced the growth rate of infected Be-13 cells in vitro, suppressed colony formation in methylcellulose cultures, and abrogated their tumorigenic phenotype in nude mice in vivo. These results suggest that suppression of the leukemic phenotype of relapse T-ALL-derived Be-13 cells is feasible. Acute leukemia cell suppression via high-efficiency infection with retroviruses encoding wtp53 may be feasible and beneficial in T-ALL cases as part of a bone marrow transplantation regimen in an effort to reduce the frequency of posttransplantation relapse.

Insulin-like growth factor-I supports proliferation of autocrine thymic lymphoma cells with a pre-T cell phenotype.

We have studied the phenotypic characteristics and growth properties of murine T lymphoma cell lines derived from primary x-ray-induced thymic lymphomas at the earliest stage at which they can be detected, and well before spreading to other organs has occurred. These cell lines serve as model systems for the earliest events in T cell lymphoma induction, before tumor cell progression and spreading to other organs. We find that primary x-ray-induced T cell lymphoma lines have phenotypic characteristics of thymic pre-T cells and show no proliferative response to any of the IL tested nor to other hematopoietic growth factors. However, they do proliferate in response to insulin-like growth factor I (IGF-I) and to a small autocrine peptide distinct from IGF-I, which we term lymphoma growth factor. One of the earliest lesions in T cell lymphoma induction may therefore be an inhibition of differentiation at one of several specific points. In its early stages, T lymphoma cell growth may be restricted to an environment where local concentrations of specific growth factors such as IGF-I or lymphoma growth factor are sufficiently high.