Induction of an activation antigen on human endothelial cells in vitro

This study describes the expression characteristics of a cell membrane protein recognized by a monoclonal antibody ENA1, which was obtained by immunizing mice with human umbilical vein endothelial (HUVE) cells cultured with a mixture of interleukin 1 and tumor necrosis factor-alpha. The expression of this ENA1 antigen could also be induced by lipopolysaccharide and phorbol esters. Expression was only demonstrated on HUVE cells and human umbilical arterial endothelial cells, pretreated with one or with a mixture of these reagents. No expression was detected on human fibroblasts, renal epithelial cells or on mesothelial cells derived from omental tissue, either pretreated or not with the aforementioned inducers of the antigen. Furthermore, no reactivity was observed with either polymorphonuclear cells, peripheral blood lymphocytes or the monocytic cell line U937. Time course experiments revealed that the expression of the ENA1 antigen was time dependent. Maximal expression on HUVE cells was observed after 5 h of incubation with activator, after which a decline in expression occurred. Induction of expression could be completely blocked by the mRNA synthesis inhibitor actinomycin D and the protein synthesis inhibitor cycloheximide, indicating that de novo synthesis occurs. Other pharmacological reagents tested had no effect on the induction of ENA1 expression. The putative role of the newly described antigen is discussed in relation to the current knowledge of molecules involved in adhesion of immune cells in inflammatory processes.     

Endotoxin activation of endothelium for polymorphonuclear leucocyte transendothelial migration and modulation by interferon-gamma.

Endotoxin [lipopolysaccharide (LPS)] is a potent inflammatory stimulus and can activate human umbilical vein endothelium (HUVE) for leucocyte adhesiveness and transendothelial migration. Here we investigated the role of HUVE-secreted cytokines in this process. When HUVE monolayers were grown on filters and preincubated for 3 hr with LPS, 51Cr-labelled polymorphonuclear leucocytes (PMNL) migrated across the HUVE in a dose- and time-dependent manner. Maximal PMNL transmigration with LPS (1 ng/ml) was 26 +/- 3% of added PMNL in 75 min. Neutralizing antibodies to interleukin-1 alpha (IL-1 alpha) and IL-1 beta, tumour necrosis factor-alpha (TNF-alpha), IL-8 or recombinant IL-1 receptor antagonist had no effect on the activation by LPS of the HUVE for supporting migration of PMNL. The HUVE 'activated state' declined with prolonged (22 hr) exposure to LPS, as reflected by a decrease in PMNL transendothelial migration to 5.5 +/- 1% and in the expression of the endothelial cell adhesion molecule, E-selectin, as compared to stimulation with LPS for 3 hr. However, simultaneous exposure to interferon-gamma (IFN-gamma) (200 IU/ml) and LPS maintained maximal PMNL transendothelial migration (28 +/- 4%) for at least 24 hr, prolonged E-selectin expression by HUVE and superinduced intracellular adhesion molecule-1 (ICAM-1) expression. The PMNL transendothelial migration was blocked by > 90% by monoclonal antibody (mAb) to CD18 with either 3 hr of LPS or 22 hr LPS + IFN-gamma stimulation. Migration was partially inhibited by mAb to E-selectin (30-40%) or to ICAM-1 (35-45%) and by a combination of both reagents (50-60%) under both stimulation conditions. Thus, LPS activation of HUVE for PMNL transendothelial migration: (a) does not require secretion of IL-1, TNF-alpha or IL-8 by the endothelium, (b) IFN-gamma enhances and prolongs endothelial activation by LPS and may increase leucocyte infiltration in LPS or bacterial inflammatory reactions, and (c) CD18-dependent mechanisms are equally important for PMNL transendothelial migration under both acute (3 hr) and prolonged (22 hr LPS + IFN-gamma) activation of endothelium.     

Adhesion of polymorphonuclear cells to human endothelial cells. Adhesion-molecule-dependent, and Fc receptor-mediated adhesion-molecule-independent mechanisms.

Activation of human umbilical vein endothelial (HUVE) cells with the inflammatory mediators tumour necrosis factor-alpha (TNF), interleukin-1 (IL-1), lipopolysaccharide (LPS) and phorbol esters enhanced their adhesiveness for leucocytes. The appearance of an activation antigen ELAM-1, recognized by a monoclonal antibody (MoAb) ENA1, parallels the kinetics of the enhanced adherence of leucocytes to endothelial cells. Adhesion of polymorphonuclear cells (PMN) to activated HUVE cells could be blocked by F(ab')2 fragments of MoAb ENA1 up to 60%. An additive inhibition of the adhesion was established by pre-incubation of the PMN with anti-CD18 MoAb and/or leucocyte adhesion inhibitor (LAI), produced by endothelial cells. An opposite reaction, however, was observed when HUVE cells were pre-incubated with intact MoAb ENA1, resulting in an enhancement of the adhesion up to 200%. Apparently, the blocking effect of MoAb ENA1 could be bypassed by the strong interaction of the Fc part of the MoAb with the Fc receptor (FcR) on the PMN. Similarly, anti-CD18 MoAb and/or LAI reduced the adhesion observed if intact ENA1 were used, and Fc-FcR interaction took place. The results presented in this study indicate that adhesion via ELAM-1, the CD18 antigen and via the receptor for LAI are different mechanisms. These mechanisms may act in concert to strengthen the binding of PMN to HUVE cells. Moreover, a strong adhesion could be established via the Fc part of MoAbs directed against HUVE cells with the FcR on the PMN. The phenomenon described may play a role in graft rejection and in diseases where antibodies directed against endothelium are involved.     

Differences in E-selectin expression and leucocyte infiltration induced by inflammatory agents in a novel subcutaneous sponge matrix model.

We have developed a novel subcutaneous sponge matrix model in major histocompatibility complex (MHC) homozygous SLAb/b inbred pigs to study lymphocyte-endothelial cell interactions during inflammation. Polyether sponges were implanted subcutaneously and left for 12 days before injection of proinflammatory agonists. Implanted sponges became highly vascularized and showed markedly increased uptake of i.v.-injected 51Cr-labelled lymphocytes 5 hr after injection of tumour necrosis factor-alpha (TNF-alpha) (3000 U) or phytohaemagglutinin (PHA) (37 micrograms). Lower levels of traffic were seen in sponges 5 hr after injection with interleukin-1 alpha (IL-1 alpha) (3000 U) and no significant traffic occurred in sponges injected with phorbol 12-myristate 13-acetate (PMA) (15 ng) at 5 hr or PHA at 24 hr (compared to sponges injected with medium alone). Electron microscopy of control sponges revealed low numbers of infiltrating leucocytes and relatively 'flat' endothelium. Many more infiltrating leucocytes were present in PHA-injected sponges. However, no ultrastructural evidence was seen of any significant difference between control and activated endothelium. Immunocytochemistry of frozen sections from sponges showed that E-selectin expression was up-regulated markedly by TNF-alpha and PHA at 5 hr, only moderately by IL-1 alpha at 5 hr, and not at all by PMA at 5 hr. By 24 hr in PHA-injected sponges E-selectin expression had fallen markedly from the level seen at 5 hr. Flow cytometric analysis of cellular infiltrates dispersed from sponges injected with TNF-alpha, PHA, IL-1 alpha or medium alone, revealed differences in lymphocyte subset populations. The infiltrate in sponges injected with TNF-alpha 5 hr before removal was dominated by high numbers of CD2+ lymphocytes, whereas the infiltrate induced by PHA showed relatively higher levels of CD2- CD4-CD8- gamma delta T-cell receptor+ (TCR+) T cells revealed by population-specific monoclonal antibodies (mAb). This model, which permits harvesting of leucocytes and endothelial cells for manipulation in vitro, will be useful for the study of leucocyte-endothelial cell interactions in subacute and chronic inflammation.     

Identification and characterization of a novel membrane activation antigen with wide cellular distribution

A monoclonal antibody, 7F7, raised against the Raji B lymphoblastoid cell line reacted with about 35% of peripheral blood B cells, germinal center B cells, follicular dendritic cells and some vascular endothelial cells. Although not found on resting T cells this antigen was strongly expressed on CD4+ and CD8+ T cells activated with phytohemagglutinin, pokeweed mitogen and anti-T3, and its expression on B cells was enhanced after stimulation with pokeweed mitogen. It is strongly expressed 24 h after stimulation with phytohemagglutinin and its expression is reduced but not eliminated by cyclosporin A treatment. The molecule defined by antibody 7F7 is found on some, but not all, B cell lines, on an HTLV-I-transformed T cell line and on the promyelocytic cell line U937. Immunoprecipitation from externally and metabolically labeled Raji cells revealed a single-chain molecule of 85 kDa.     

Induction of neovascularization and nonlymphoid mesenchymal cell proliferation by macrophage cell lines

The mature murine macrophage-like cells NCTC-3749 and J-774, the immature human macrophage-like cells U-937-1, and their conditioned media exhibited potent angiogenic activity in rat corneas and stimulated proliferation of bovine aortic endothelial cells (BAEC) and DNA synthesis in BALB/c-3T3 cells in culture. In contrast, the immature human macrophage-like cells HL-60 and their conditioned media either failed to produce or release detectable quantities of these activities. Exposure of HL-60 cells to phorbol-myristate-acetate (PMA) did not enhance expression of angiogenic and growth stimulating activities by these cells. Both the angiogenic and growth stimulating activities appear to be mediated by a factor(s) that has biochemical properties in common with macrophage-derived growth factor (MDGF) produced by normal rat peritoneal macrophages. These results suggest that NCTC-3749, J-774, and U-937-1 macrophage-like cell lines may be a useful source for the large scale production and characterization of MDGF and macrophage-derived angiogenic activity.     

Characterization and structural analysis of Fc gamma receptors of human monocytes, a monoblast cell line (U937) and a myeloblast cell line (HL-60) by a monoclonal antibody

A monoclonal antibody, FR51, raised against the IgG Fc receptor (Fc gamma R) of the human monoblast cell line U937 was used to analyze the distribution of this antigen on various human cells. This antibody inhibited the binding of human IgG to the Fc gamma R on U937 cells, HL-60 cells and human peripheral blood monocytes. In contrast, the Fc gamma R on human granulocytes (neutrophil cells) and on an Epstein-Barr virus-transformed human lymphoblastoid cell line (Raji) were not recognized, indicated by the failure of blocking the binding of human IgG ligand to the Fc gamma R on these cells. By affinity chromatography of detergent-containing cell free lysates of surface-iodinated U937 cells, HL-60 cells and monocytes, a protein of 70-kDa was isolated. This protein was identified as the Fc gamma R by rebinding the isolated protein to immobilized human IgG. Removal of the carbohydrate moiety with endo-beta-N-acetylglucosaminidase F demonstrated that the receptors consist of a 40-kDa polypeptide. Analysis of the polypeptide patterns obtained by proteolytic digestion of either mature (70-kDa) or deglycosylated (40-kDa) receptors isolated from monocytes, U937 cells and HL-60 cells strongly suggests that the Fc gamma R are identical. The monoclonal antibody FR51 specifically reacts with Fc gamma R on human monocytes, a myeloblast and a monoblast cell line but not with the receptors on a B cell line and neutrophil cells.     

Monoclonal antibodies that bind to the My23 human myeloid cell surface molecule: epitope analysis and antigen modulation studies

It has previously been shown that the AML-2-23 monoclonal antibody (MoAb) reacts with a glycoprotein on differentiated myeloid cells. The antigen, My23, is released from these cells in culture and is also detectable in normal human plasma. We have now raised a panel of MoAbs against the soluble form of the antigen which reacts with monocytes and calcitriol-treated U937 and HL-60 myeloid cell lines, but not with lymphocytes or undifferentiated U937 and HL-60 cells. As with the AML-2-23 MoAb, the anti-My23 MoAbs immunoprecipitated a 50,000-55,000 protein from calcitriol treated HL-60 cells. Besides binding to cell surface My23, all eight MoAbs as well as the 63D3 MoAb reacted with crude and purified forms of soluble My23. A novel ELISA epitope analysis assay was developed to identify four distinct antigenic determinants on My23. Thus, the soluble and cell surface forms of My23 share several antigenic determinants and are biochemically very similar. Pooled anti-My23 caused selective patching, capping and clearing of My23 from the cell surface. My23 was not associated with cell surface, HLA-DR, HLA-A,B,C, beta 2-microglobulin or the Fc receptors for IgG or IgA. These anti-My23 MoAbs should be of importance in antigen functional studies and in clinical treatment protocols for patients with acute myelogenous leukemia. Furthermore, we have shown that a soluble myeloid differentiation antigen can serve as an effective immunogen for the preparation of MoAbs against important cell surface molecules thus obviating many problems inherent in the use of whole cells or detergent lysate-derived immunoprecipitates as immunogens.     

Comparative susceptibility of peripheral blood leucocytes and related cell lines to killing by T-cell perforin.

The comparative susceptibility of lymphocyte subsets, monocytes and polymorphonuclear leucocytes (PMN) to killing by murine perforin was measured using physical separation techniques, cell-surface phenotyping and scatter characteristics to isolate cell types, together with propidium iodide (PI) uptake as a measure of cell death. In the majority of individuals, PMN were more resistant to perforin than other peripheral blood cells including natural killer (NK) cells and CD8+ lymphocytes. Among the lymphocytes, CD4+ cells were the most susceptible subset, followed by CD19+, CD8+ and CD56+ lymphocytes respectively. The human promyelocytic leukaemia cell line, HL-60, and the human histiocytic lymphoma cell line, U937, were readily killed by perforin. When HL-60 were differentiated to either macrophage- or neutrophil-like end cells, and U937 differentiated to macrophage-like end cells, there was no difference between differentiated and undifferentiated cells in their relative susceptibility to perforin. The relative resistance of PMN to perforin may be important in protecting them from damage in in vivo situations where both NK cells and neutrophils are localized in the same inflammatory areas.     

Engagement of Na,K-ATPase beta3 subunit by a specific mAb suppresses T and B lymphocyte activation

In order to identify new molecules involved in regulation of T cell proliferation, we generated various mAb by immunization of mice with the T cell line Molt4. We found one mAb (termed P-3E10) that down-regulated the in vitro T cell proliferation induced by CD3-specific OKT3 mAb. The P-3E10 mAb was also able to inhibit IFN-gamma, IL-2, IL-4 and IL-10 production of OKT3-activated T cells. The antigen recognized by P-3E10 mAb is broadly expressed on all hematopoietic as well as on all non-hematopoietic cell lines tested so far. Within peripheral blood leukocytes, the P-3E10 antigen was detected on lymphocytes, monocytes and granulocytes. Human umbilical vein endothelial cells (HUVEC) also scored positively. By evaluating the effect of P-3E10 mAb on these cell types we found that it also inhibited anti-IgM-induced B cell proliferation. However, it did not block growth factor-mediated proliferation of HUVEC, and spontaneous proliferation of SupT-1, Jurkat, Molt4 and U937 cell lines. Moreover, it did not influence phagocytosis of human blood monocytes and granulocytes. Biochemical analysis revealed that the P-3E10 antigen is a protein with a mol. wt of 45-50 kDa under non-reducing and 50-55 kDa under reducing conditions. By using a retroviral cloning system, the P-3E10 antigen was cloned. Sequence analysis revealed the P-3E10 antigen to be identical to the beta3 subunit of the Na,K-ATPase.     

Regulation of c-fgr messenger RNA levels in U937 cells treated with different modulating agents.

The U937 cell line was used to investigate the induction of messenger RNA (mRNA) for the c-fgr mRNA tyrosine kinase proto-oncogene in cells of the monocyte-macrophage lineage. U937 cells were exposed to tumour necrosis factor-alpha (TNF-alpha), TNF-beta and transforming growth factor-alpha (TGF-alpha), alone and in combination with PMA and 1,25-dihydroxycholecalciferol (1,25-DHCC). TNF-alpha and TNF-beta, but not TGF-alpha, decreased the proliferation of U937 cells in a time- and dose-dependent manner, and both TNF-alpha and TNF-beta enhanced the response of U937 cells to PMA during the first 24 hr of treatment and to 1,25-DHCC over 72 hr. TNF-alpha induced a rapid increase in c-fgr mRNA levels within 4 hr, in contrast to slower induction by PMA and 1,25-DHCC. TNF-alpha and 1,25-DHCC together had an additive effect on c-fgr mRNA levels. In U937 cells exposed to PMA, c-fgr mRNA levels continued to increase over 72 hr. Levels of c-fgr mRNA induced by the various modulating agents did not correlate clearly with the changes in proliferation. Therefore, we suggest that although the c-fgr gene product may have a role in differentiation, the more significant role is likely to be in the fully differentiated macrophage.     

Characterization of a novel myeloid antigen regulated during differentiation of monocytic cells

The monoclonal antibody HC1/6 generated against phorbol 12-myristate 13-acetate-treated U-937 cells recognizes a new cell surface antigen with a broad relative molecular mass ranging from 100 to 150 kDa. This antigen is also present on monocytes, platelets and endothelial cells and is weakly expressed by granulocytes. In contrast, it is absent from T, B and erythroblastoid cells. The antigen HC1/6 is also expressed by normal tissue macrophages in tonsil, lung and kidney, as well as in skin biopsies from pathologies such as sarcoidosis and lepromatous leprosy. The expression of the HC1/6 antigen is increased up to 5-fold when U-937 (promonocytic) and HL-60 (myelomonocytic) cell lines are stimulated with phorbol 12-myristate 13-acetate. Conversely, the expression of the HC1/6 antigen is down-regulated in monocytes upon treatment with interferon-gamma. These findings are discussed in relation with other myeloid cell surface markers.     

sLex is not responsible for the interaction of sLex-positive memory T lymphocytes with E-selectin.

E-selectin in an adhesion molecule that is transiently and exclusively expressed on endothelial cells in response to inflammatory cytokines. In addition, E-selectin participates in the initial interaction of leucocytes with activated endothelial cells. This role of E-selectin in cell adhesion has made it a potential target for modulation of inflammatory processes that, for example, are occurring in autoimmune diseases such as rheumatoid arthritis. Although on granulocytes the ligand for E-selectin has been identified as the tetrasaccharide sialyl Lewis x (sLex), the molecular nature of this ligand on T lymphocytes has not yet been identified. In the present study, it was shown by fluorescence-activated cell sorter (FACS) analysis with the anti-sLex antibody CSLEX1 that T lymphocytes stimulated with phytohaemagglutanin (PHA), interleukin-2 (IL-2) and transforming growth factor-beta 1 (TGF-beta 1) expressed sLex. Furthermore, in a cell adhesion assay these activated T cells of the memory phenotype bound specifically to E-selectin-transfected Chinese hamster ovary (E-CHO) cells. This adhesion could be blocked with an anti-E-selectin antibody but not with CSLEX1. In the same assay, the interaction of sLex-positive U937 cells with the E-CHO cells could be inhibited both with anti-E-selectin and CSLEX1 antibodies. From these results it can be inferred that sLex on activated T lymphocytes is not responsible for the interaction with E-selectin. Rather, these results suggest that stimulated T lymphocytes express an additional E-selectin ligand(s) with much higher avidity for E-selectin than sLex.     

Regulation of Fc epsilon receptor expression on a human monoblast cell line U937.

Fc epsilon receptor (Fc epsilon R) expression on several human cell lines (U937, RPMI 8866, HL 60, THP-1, and Molt 4) and its regulation were examined by immunofluorescent analysis using a monoclonal anti-human Fc epsilon R antibody, H107. Phorbol ester (PMA), recombinant gamma interferon (IFN-gamma) and H107 itself enhanced Fc epsilon R expression on a FC epsilon R positive cell line U937, whereas these reagents did not induce FC epsilon R expression on the Fc epsilon R negative cell lines, Molt 4, HL 60 and THP-1. Dexamethasone not only suppressed by 50% the spontaneous Fc epsilon R expression on U937 cells but also completely inhibited the enhancement of their Fc epsilon R expression on U937 cells induced by PMA, IFN-gamma or H107. Dexamethasone caused a little suppression of Fc epsilon R expression by RPMI 8866 cells. The results showed that Fc epsilon R expression on a human monoblast cell line U937 was up- or down-regulated by a variety of physiological or pharmacological agents. These experimental systems provide a good model for the investigation of the regulatory mechanisms of Fc epsilon R expression.     

Selective induction of apoptosis by ar-turmerone isolated from turmeric (Curcuma longa L) in two human leukemia cell lines, but not in human stomach cancer cell line

We have investigated the effects of ar-turmerone isolated from turmeric (Curcuma longa L) on DNA of human leukemia cell lines, Molt 4B, HL-60 and stomach cancer KATO III cells. It was found that selective induction of apoptosis by ar-turmerone was observed in human leukemia Molt 4B and HL-60 cells, but not in human stomach cancer KATO III cells. Morphological changes showing apoptotic bodies were observed in the human HL-60 and Molt 4B cells treated with ar-turmerone. The fragmentation of DNA by ar-turmerone to oligonucleosomal-sized fragments that is a characteristic of apoptosis was observed to be concentration- and time-dependent in Molt 4B and HL-60 cells, but not in KATO III cells. The data of the present study show that the suppression by ar-turmerone of growth of these leukemia cell lines results from the induction of apoptosis by this compound.     

Apoptosis as a Mechanism of Lectin-Dependent Monocyte-Mediated Cytotoxicity

In this study we investigated the mechanisms of cytotoxicity mediated by pokeweed mitogen (PWM)-activated human peripheral blood monocytes. By using DNA electrophoresis and propidium iodide (PI)-DNA staining flow cytometry, we demonstrated that apoptotic cell death of target U937 cells and Raji cells was induced in lectin (PWM)-dependent monocyte-mediated cytotoxicity (LDMC). The LDMC-mediated DNA fragmentation in U937 cells and Raji cells was induced in lectin (PWM)-dependent monocyte mediated cytotoxicity(LDMC). The LDMC-mediated DNA fragmentation in U937 cells was completely inhibited by anti-TNFα monoclonal antibody (mAb), but not by the addition of monosaccharide (N-acetylglucosamine, GlcNAc, a sugar specifically recognized by PWM and a lectin-like receptor on monocytes). In contrast, GlcNAc inhibited the DNA fragmentation in Raji cells induced by LDMC which the anti-TNFα mAb had no effect. PWM was found to stimulate the production of nitric oxide (NO) from monocytes. The NO-production was enhanced in the presence of target Raji cells, while the enhancement was abolished by the treatment with GlcNAc. By flow cytometry, we found that PWM bound to tumour cells as well as monocytes, and inhibited the expression of HLA-DR antigen on tumour cells. These results suggest that the presence of lectin molecules on the surface of monocytes and tumour cells may bring the two cells together, thus facilitating the induction of apoptosis in target cells by triggering the production of cytolytic factors (TNF and NO) and the modification of target cell surface antigen (HLA-DR).     

Mucosal vaccination increases endothelial expression of mucosal addressin cell adhesion molecule 1 in the human gastrointestinal tract

Homing of leukocytes to various tissues is dependent on the interaction between homing receptors on leukocytes and their ligands, addressins, on endothelial cells. Mucosal immunization results in homing of antigen-specific lymphocytes back to the mucosa where they first encountered the antigen. However, it is unknown whether this homing of antigen-specific cells is mediated by an altered endothelial addressin expression after vaccination. Using different immunization routes with an oral cholera vaccine, we show that the endothelial expression of mucosal addressin cell adhesion molecule 1 (MAdCAM-1) is increased in the gastric and upper small intestinal mucosae after immunization through various local routes in the upper gastrointestinal tract. In contrast, rectal immunization did not influence the levels of MAdCAM-1 in the gastric or duodenal mucosa. Furthermore, we show that MAdCAM-1 can be induced on human endothelial cells by tumor necrosis factor alpha (TNF-alpha) and gamma interferon. The vaccine component cholera toxin B subunit (CTB) increased MAdCAM-1 expression on endothelial cells in cultured human gastric explants, an effect that seemed to be mediated by TNF-alpha. In conclusion, MAdCAM-1 expression is increased in the upper gastrointestinal tract after local immunizations with a vaccine containing CTB. This strongly suggests the involvement of MAdCAM-1 in the preferential homing of mucosal lymphocytes to their original site of activation.     

Poliovirus infection of established human blood cell lines: relationship between the differentiation stage and susceptibility of cell killing

The replication of type 1 poliovirus in 13 established human blood cell lines differing in the differentiation stage and cell lineage was investigated. Three T (CCRF-CEM, CCRF-HSB-2, and Molt-3) and three B (Raji, CCRF-SB, and RPMI 8226) cell lines showed no cytopathic effects (CPE) or virus production. CPE associated with virus production were detected in the other seven cell lines: HL-60, ML-1, and KG-1 (granulocytic lineage), U-937 and THP-1 (monocytic lineage), K-562 (erythroid lineage), and Molt-4 (T cell lineage). These susceptible cell lines greatly differed in the speed at which the CPE progressed. The progression of CPE was faster in relatively well-differentiated cell lines such as HL-60 and U-937, independently of the multiplicity of infection, than in less differentiated cell lines such as K-562, KG-1, and THP-1. Thus, for the same lineage, the speed at which CPE progressed became proportionally higher with subsequent differentiation stages. In the K-562 cell culture, CPE were not observed until at least 5 days postinfection (p.i.), while more than 80% of HL-60 cells were killed within 3 days p.i. There were no significant differences between infected HL-60 and K-562 cells in the efficiency of infection determined at 8 hr p.i. by the indirect immunofluorescent technique, the rate of virus growth, or the amount of viral capsid protein synthesized. This indicated that there were similar viral replication cycles in the two cell lines. These observations suggest that the killing function of the virus is expressed more slowly in K-562 cells than in HL-60 cells.