Freshly cultured vascular endothelial cells express the CD34 antigen in a diffuse cell surface pattern with some concentration on microvilli. Expression is downregulated with proliferation in continuous culture and undetectable after nine population doublings but can be maintained by restraining cell proliferation and promoting cell contact. Expression of CD34 at the antigen and mRNA levels on early passage cells is rapidly downregulated by interleukin-1 beta (IL-1 beta), interferon-gamma (INF-gamma), and tumor necrosis factor-alpha (TNF- alpha) under conditions in which these ligands upregulate the adhesion molecules: endothelial leukocyte adhesion molecule 1 (ELAM-1) and intracellular adhesion molecule 1 (ICAM-1). This reciprocal pattern of expression and the topographic distribution of CD34 molecules on the lumenal interdigitated microprocesses of adjacent endothelial cells in vivo suggest that CD34 might have a negative modulating role on adhesion functions of endothelia.

Indeterminate cell histiocytosis (ICH) is a proliferation of indeterminate CD1a+, CD68+, S100+ and CD207- dermal dendritic cells. We describe a 39-year-old man who developed diffuse ICH and, 6 years later, acute myeloblastic leukaemia (AML). He was treated with cyclophosphamide, etoposide and vinblastine until 2003. In August 2004, he presented dyspnoea, hyperpyrexia and infiltration of the lung parenchyma, compatible with an AML invasion, and died after a course of induction chemotherapy. Cytomorphology and immunophenotype analyses suggested an ICH clonal evolution. The leukaemogenic role of etoposide is discussed. ICH has previously been reported in association with B-cell malignancy, but only one case has shown systemic progression.

The present study provides immunobiochemical and molecular data on the differentiation-linked expression of the bcl-2 proto-oncogene in normal and neoplastic myeloid cells. Using a recently developed monoclonal antibody (MoAb) to the bcl-2 molecule, staining of normal bone marrow myeloblasts, promyelocytes, and myelocytes, but neither monocytes nor most polymorphonuclear cells, was demonstrated. By two-color flow cytometric analysis, bcl-2 was evidenced in CD33+ and CD33+/CD34+ myeloid cells as well as in the more primitive CD33-/CD34+ population. The leukemic cell lines HL-60, KG1, GM-1, and K562 were bcl-2 positive together with 11 of 14 acute myeloid leukemias (AML) and three of three chronic myeloid leukemias (CML) in blast crises; six of seven CML were negative. Among myelodysplastic cases, augmentation of the bcl-2 positive myeloblastic compartment was found in refractory anemia with excess of blasts (RAEB) and in transformation (RAEB-t). Western blots of myeloid leukemias and control lymphocytes extracts evidenced an anti- bcl-2 immunoreactive band of the expected size (26 Kd). Moreover, the HL-60 and KG1 cell lines, both positive for the bcl-2 protein, exhibited the appropriate size bcl-2 mRNA (7.5 Kb). These findings clearly indicate that the bcl-2 gene is operative in myeloid cells and that the anti-bcl-2 MoAb identifies its product and not a cross- reactive epitope. Induction of HL-60 differentiation toward the monocytic and granulocytic pathways was accompanied by a marked decrease in bcl-2 mRNA and protein levels; bivariate flow cytometric analysis showed that the fraction becoming bcl-2 negative was in the G1 phase of the cell cycle. These data establish that the bcl-2 proto- oncogene is expressed on myeloid cells and their progenitors and is regulated in a differentiation-linked manner.

Freshly cultured vascular endothelial cells express the CD34 antigen in a diffuse cell surface pattern with some concentration on microvilli. Expression is downregulated with proliferation in continuous culture and undetectable after nine population doublings but can be maintained by restraining cell proliferation and promoting cell contact. Expression of CD34 at the antigen and mRNA levels on early passage cells is rapidly downregulated by interleukin-1 beta (IL-1 beta), interferon-gamma (INF-gamma), and tumor necrosis factor-alpha (TNF- alpha) under conditions in which these ligands upregulate the adhesion molecules: endothelial leukocyte adhesion molecule 1 (ELAM-1) and intracellular adhesion molecule 1 (ICAM-1). This reciprocal pattern of expression and the topographic distribution of CD34 molecules on the lumenal interdigitated microprocesses of adjacent endothelial cells in vivo suggest that CD34 might have a negative modulating role on adhesion functions of endothelia.

The present study provides immunobiochemical and molecular data on the differentiation-linked expression of the bcl-2 proto-oncogene in normal and neoplastic myeloid cells. Using a recently developed monoclonal antibody (MoAb) to the bcl-2 molecule, staining of normal bone marrow myeloblasts, promyelocytes, and myelocytes, but neither monocytes nor most polymorphonuclear cells, was demonstrated. By two-color flow cytometric analysis, bcl-2 was evidenced in CD33+ and CD33+/CD34+ myeloid cells as well as in the more primitive CD33-/CD34+ population. The leukemic cell lines HL-60, KG1, GM-1, and K562 were bcl-2 positive together with 11 of 14 acute myeloid leukemias (AML) and three of three chronic myeloid leukemias (CML) in blast crises; six of seven CML were negative. Among myelodysplastic cases, augmentation of the bcl-2 positive myeloblastic compartment was found in refractory anemia with excess of blasts (RAEB) and in transformation (RAEB-t). Western blots of myeloid leukemias and control lymphocytes extracts evidenced an anti- bcl-2 immunoreactive band of the expected size (26 Kd). Moreover, the HL-60 and KG1 cell lines, both positive for the bcl-2 protein, exhibited the appropriate size bcl-2 mRNA (7.5 Kb). These findings clearly indicate that the bcl-2 gene is operative in myeloid cells and that the anti-bcl-2 MoAb identifies its product and not a cross- reactive epitope. Induction of HL-60 differentiation toward the monocytic and granulocytic pathways was accompanied by a marked decrease in bcl-2 mRNA and protein levels; bivariate flow cytometric analysis showed that the fraction becoming bcl-2 negative was in the G1 phase of the cell cycle. These data establish that the bcl-2 proto- oncogene is expressed on myeloid cells and their progenitors and is regulated in a differentiation-linked manner.

Two anti-nerve growth factor receptor (LNGFR or p75NGFR) antibodies, Me20.4 and Me8211, label stromal cells with dendritic features in fresh smears and in formalin-fixed, paraffin-embedded human bone marrow (BM). The LNGFR+ cells have an oval nucleus, a scanty cytoplasm with long dendrites that intermingle with the hematopoietic cells, line the abluminal side of sinus endothelial cells, and provide the scaffold for the hematopoietic marrow. At the electron microscopy level, the immunogold tag labels the body and the long branching dendrites of fibroblast-like cells with scanty cytoplasm containing mitochondria, endoplasmic reticulum, and dense bodies. The LNGFR+ cells are positive for alkaline phosphatase, reticulin, collagen III, vimentin, TE-7, and CD13 but negative for endothelial (vWF, CD34, Pal-E), neural (CD56, neurofilament) and leukocyte markers (CD45, CD68). The LNGFR+ stromal cells appear in the fetal BM before the hematopoietic activity begins, originate from the vessel adventitia, and radiate in the Bm cavity. Long-term BM culture (LTBMC) in vitro contain LNGFR+ stromal cells. We document the presence of RNA message for the low- (LNGFR) and the high- affinity NGF receptor (NTRK1) by using RT-PCR on fresh BM aspirate and on LTBMC. BM biopsies from patients with hematologic fibrogenic diseases and in cytokine-treated cancer patients are evaluated for LNGFR+ cells: the amount of stained cells is correlated with the traditional reticulin stain in cases of myelofibrosis, therapy-related myelodysplasia, leukemia, and detected an increase of stromal cells in cytokine-treated patients. The anti-LNGFR antibodies represent a specific membrane marker for the adventitial reticular cells (ARC) of the human marrow and allow precise evaluation and quantitation of this important BM microenvironment component in vivo and in vitro.

Bone marrow biopsy (BMB) is a routine investigation in the diagnosis and staging of Hodgkin's disease (HD) and non-Hodgkin's lymphoma (NHL), and there is evidence supporting its prognostic importance in some histological varieties. The histological characteristics of BMB in 433 NHL and 155 HD patients were reviewed for clinicopathological correlations; 36 of these cases were also studied by means of immunohistochemistry. BM infiltrates were discovered in 171 NHL patients. In 36 cases, the diagnosis of NHL was directly established by BMB; a discordance between lymph node and BM histology was observed in 38 of the other 135 cases. BM-positive centroblastic and immunoblastic NHL were significantly associated with larger infiltrates, BM fibrosis, and megakaryocytic hyperplasia. Leukemization at diagnosis was more frequent in low-malignancy NHL. No correlation was found between histology and prognosis, although immunohistochemistry revealed a B-cell phenotype in all but two cases. BMB was positive in 18 of the 155 HD patients and directly diagnostic in two; Reed-Sternberg and Hodgkin cells were CD-30 positive and surrounded by T-cell infiltration. The concordance between BM and lymph node histology was fairly satisfactory, although the relationships between BM infiltration and other histological parameters may reflect peculiar interactions with BM microenvironmental factors. The usefulness of BMB in the diagnosis of malignant lymphomas has been demonstrated, and further progress can be expected from the availability of reliable immunohistochemical markers of clonality reacting on paraffin-embedded BM sections.

The prognostic impact of bone marrow biopsy (BMB) histology and CD34 immunoreactivity was compared with that of the more conventional parameters (the FAB diagnosis, peripheral blood values, percentage of BM blasts and some common prognostic scores) in 100 MDS patients. Statistical correlations among the cytological, haematological, histological and immunohistochemical parameters and their relationship with clinical outcome were searched for. At univariate analysis, FAB classification (P < 0.001), pattern of blastic infiltration at BMB (P < 0.005), presence of CD34+ aggregates (P < 0.0005), percentage of blasts in BM aspirate (P < 0.0001) and percentage of CD34 positivity (P < 0.0001) proved to be linked to leukaemic transformation and, except for FAB classification, retained a high degree of prognostic significance in terms of survival. Leukaemic transformation occurred in 16/18 patients simultaneously presenting 'large' blastic infiltrates at BMB and CD34+ aggregates (P < 0.00001); 9/17 evaluable patients died within 12 months of diagnosis (P < 0.001)> Discriminant functions for leukaemic transformation and survival did not offer any advantage over univariate analysis in the prognostic work-up. The results indicate that the size of blastic aggregates and CD34 positivity allowed patients with a worse prognosis to be identified irrespective of their FAB subtype, but the prognostic impact is considerably greater when both parameters are simultaneously taken into account, as testified by the restricted and homogeneous subgroup of patients with both 'large' and CD34-positive aggregates.