Human mature erythrocytes have been considered as unable to undergo programmed cell death (PCD), due to their lack of mitochondria, nucleus and other organelles, and to the finding that they survive two conditions that induce PCD in vitro in all human nucleated cells, treatment with staurosporine and serum deprivation. Here we report that mature erythrocytes can undergo a rapid self-destruction process sharing several features with apoptosis, including cell shrinkage, plasma membrane microvesiculation, phosphatidylserine externalization, and leading to erythrocyte disintegration, or, in the presence of macrophages, to macrophage ingestion of dying erythrocytes. This regulated form of PCD was induced by Ca 2+ influx, and prevented by cysteine protease inhibitors that allowed erythrocyte survival in vitro and in vivo. The cysteine proteinases involved seem not to be caspases, since (i) proforms of caspase 3, while present in erythrocytes, were not activated during erythrocyte death; (ii) cytochrome c, a critical component of the apoptosome, was lacking; and (iii) cell-free assays did not detect activated effectors of nuclear apoptosis in dying erythrocytes. Our findings provide the first identification that a death program can operate in the absence of mitochondria. They indicate that mature erythrocytes share with all other mammalian cell types the capacity to selfdestruct in response to environmental signals, and imply that erythrocyte survival may be modulated by therapeutic intervention.

Many investigators have studied this problem with numerous, and often conflicting, hypotheses proposed. We have reviewed the field and discussed the pros and cons of each (1). We favor the hypothesis that "RBC senescence" and clearance from the circulation involve the unmasking of ␤-galactosyl residues due to desialylation of glycoconjugates on the RBC cell surface. This hypothesis was based on classical biochemical studies on young and old RBCs (2) and was validated at the cellular level by the use of flow cytometric procedures (3,4). Following these studies, we demonstrated with flow cytometric techniques that the densest (5) and oldest RBCs were the most

This study prospectively analysed the relationships between observed to vary according to the type and number of lymphimmunophenotypic and cytogenetic features of blast cells in oid lineage antigens expressed. 9 The expression of some of 432 acute non-lymphoblastic leukemias (ANLL) at presentation.these antigens also appeared to be related to recurrent cyto- The study was initiated in March 1992 by the Groupe d'Etude Immunologique des Leucémies (GEIL), a French multicentric Introduction group that collects data from over 40 centers in France and Belgium. The endpoint of the present analysis was 30 SepIn 1988, the morphologic, immunologic and cytogenetic tember 1994. Patients eligible for the study met the following working classification (MIC) of acute non-lymphoblastic leucriteria: established diagnosis of ANLL and collection of a sufkemia (ANLL) 1 underlined the correlations between some ficient number of blast cells for both immunophenotype and recurrent specific cytogenetic abnormalities and defined cytosuccessful cytogenetic analysis. M0-ANLL were defined as logic subgroups. At that time, the relationships between cytopreviously reported. 2 FAB data were collected from each genetic features identified in ANLL and the immunophenotype center's cytology laboratory. of blast cells remained poorly characterised. During the last 10 years, immunophenotypic analysis of ANLL blast cells has provided new information and become a powerful tool in Immunological phenotyping assigning undifferentiated acute leukaemia to the myeloid lineage. 2 Meanwhile, investigation of lymphoid lineage antigens Mononuclear cells were recovered from heparinized bone on blast cells from a large number of ANLL has demonstrated marrow (391 cases) or peripheral blood (41 cases) collected the high immunophenotypic heterogeneity of morphologically at diagnosis. The percentage of blast cells after separation on and cytogenetically well-defined diseases. [3][4][5][6][7][8] a Ficoll gradient was higher than 50% in M2 and M4 subtypes Molecular analyses attempting to correlate these immunoand higher than 80% in the remaining cases. The immunophenotypic features to their genotypic counterpart did not phenotype was performed by flow cytometry on blast cells demonstrate any strict correlation between the rearrangement gated on their abnormal light scatter characteristics, using of immunoglobulin and T cell receptor genes and the monoclonal antibodies for the following antigens: CD9 expression of lymphoid lineage antigens on ANLL blast (IOB2), CD11b (FD11), CD13 (MY7), CD14 (UCHM1), cells. 3,9-11 However, the prevalence of genotypic changes was CD15a (SMY15a), CD16 (Leu11b), CD18 (IOT18), CD33 (MY9), CD34 (BI-3C5), CD35 (44-D), CD36 (OKM5), CDw65 (VIM2), CD41 (P2), CD42 (SZ2), T lymphoid differentiation

We report two cases of secondary acute lymphoblastic leukemia (ALL) with t (4;11) (q21;q23) translocation occurring after chemotherapy and radiotherapy for a prior cancer. Seven previously published cases of secondary ALL with t (4;11) (q21;q23) are also reviewed. Most patients had received a combination of topoisomerase II inhibitors (anthracyclines, mitoxantrone, or the epipodophillotoxin derivatives VP16 or VM26) and cyclophosphamide, which have also been implicated in the pathogenesis of secondary acute myeloid leukemia (AML) with 11q23 rearrangements. These observations give further support to the existence of a subgroup of secondary acute leukemias with cytogenetic findings "specific" for de novo ALL and AML, especially those with translocations involving the 11q23 region.

We performed immunocytochemical detection of myeloperoxidase (MPO), using monoclonal antibody MPO-7, in 15 consecutive cases of adult acute leukemia (AL) unclassified by conventional cytological and cytochemical criteria and 7 AML-M, with less than 10% of cytochemically MPO-positive blasts. In AL with negative MPO cytochemistry the anti-MPO reaction was positive in 5 of the 15 patients with 3, 3, 7, 11 and 45% positive blasts respectively. In AML-M,, immunocytochemistry was positive in a larger percentage of blasts than cytochemistry in 2 cases. Immunological detection of myeloid surface markers was positive in all 15 cases of unclassified AL (including the 10 AL with negative anti-MPO reaction). Eleven of the 22 patients from this study had mixed lymphoid-myeloid phenotype. Discrepancy between immunological MPO detection and light cytochemistry was more frequent in patients with mixed imrnunophenotype than in patients without lymphoid markers. No relationship between MPO-antigen positivity and clinical or biological features was seen. These findings confirm immunological detection of MPO as useful for the diagnosis of poorly differentiated AL. The high incidence of inactive MPO detectable only by immunocytochemistry in mixed lineage AL needs to be confirmed.