A new human T-cell differentiation antigen: Unexpected expression on chronic lymphocytic leukemia cells

Monoclonal antibody 10.2 reacts with a monomorphic antigen expressed on the surface of virtually all thymocytes, as well as thymus-dependent lymphocytes in the peripheral blood and bone marrow. In contrast, antibody 10.2 did not react with normal peripheral blood B cells, monocytes, or the non-T-cell fraction of bone marrow. This complement fixing IgG2a antibody also reacted with extablished leukemic T-cell lines, but not with cell lines of either normal or malignant B-cell origin. Similarly, when tested against acute leukemia blasts, the 10.2 antibody reacted with those from patients with T-cell acute leukemia, but not with those from patients with acute null cell or non-lymphocytic leukemia. An unexpected exception to this pattern was the reaction of 10.2 antibody with leukemic cells from patients with B-cell type chronic lymphocytic leukemia. Immune precipitates formed with 10.2 antibody and detergent lysates of radiolabeled T-cells contained three polypeptides with molecular weights of 65 000, 55 000, and 50000 daltons. It has not been determined whether all three of these polypeptides contain the 10.2 antigenic determinant, or whether these proteins represent a multimeric antigen complex.PJM is a Junior Faculty Clinical Fellow of the American Cancer Society.     

A novel human T cell antigen preferentially expressed on mature T cells and shared by both well and poorly differentiated B cell leukemias and lymphomas

A new lymphocyte differentiation antigen shared by all normal T cells and some malignant B cells was defined by a monoclonal antibody designated 12.1. This antibody reacted with all peripheral blood T cells but not with normal B cells and B cell lines. Analysis with a fluorescence activated cell sorter showed that the expression of 12.1 antigen changes during T cell maturation. Most thymocytes, blasts of acute T cell leukemia, and cells from established leukemic T cell lines bear a small amount of 12.1 antigen. In contrast the majority of peripheral blood T cells, activated T cells, and leukemic T cells of the Sezary syndrome bear a large amount of 12.1 antigen. Unexpectedly, antibody 12.1 reacted with leukemic cells from most patients with B-type chronic lymphocytic leukemia (CLL) and some patients with lymphosarcoma cell leukemia (LSCL). Among these leukemias, expression of the 12.1 antigen was not correlated with the stage of B cell maturation, with the amount of surface immunoglobulin on the cells, or with the presence or absence of monoclonal gammapathy. In a comparative serologic analysis the antigen defined by antibody 12.1 was distinct from the p67 T cell antigen (defined by monoclonal antibody 10.2) that is also known to be expressed by B-type CLL cells.     

Monoclonal antibodies identifying a novel T-Cell antigen and Ia antigens of human lymphocytes

We describe here two new monoclonal antibodies that react with surface antigens of human lymphocytes. Antibody 7.2 identified a determinant on the framework region of the human Ia antigen. It was cytotoxic for all cultured B-cell lines, normal B cells, and monocytes. The antibody was not cytotoxic for normal T cells or for established T leukemic cell lines. In immune precipitation assays, the 7.2 antibody reacted with a bimolecular complex of two chains that resolved in polyacrylamide gels as polypeptides with molecular weights of 29000 and 34000 daltons. These precipitation results were analogous to those achieved with a rabbit antiserum prepared against human Ia antigens. Antibody 9.3 identified a determinant on the framework region of a T-cell antigen. It was cytotoxic for 50–80% of peripheral T cells and for 20–50% of thymocytes. It was not cytotoxic for cultured B-cell lines, normal B cells, or monocytes. In immune precipitation assays, the 9.3 antibody reacted with a single polypeptide with a molecular weight of 44000 daltons. Due to the expression of this antigen on a limited subpopulation of human T cells, we have designated the antigen HuLyt-1.     

Acute lymphoblastic leukemia (ALL) antigens detected with antisera to E rosette-froming and non-E rosette-forming ALL blasts.

Based on the presence or absence of erythrocyte receptors(E) a T cell marker, acute lymphocytic leukemia (ALL), can be divided into E+ALL and E-ALL. We studied cell surface antigens on blasts from 12 children with untreated ALL: eight with E-ALL and four with E+ALL. Heterologous antisera were raised against thymus cells, E+ and E-ALL blasts, appropriately absorbed and tested by immunofluorescence and a radiolabeled antibody assay with normal and leukemic lymphoid cells. By both methods, anti-thymus and anti-E+ALL sera reacted with human thymocytes. Specific binding of anti-E+ALL serum to T antigens was indicated by the fact that a single absorption with thymocytes abolished its binding to allogenic thymocytes, and the reactivity of anti-E+ALL serum with thymus, blood and bone marrow lymphocytes was similar to that of anti-thymus serum. After exhaustive absorption with blood leukocytes, anti-E+ALL and E-ALL sera were negative against normal lymphocytes and bone marrow cells from children with ALL in remission. Anti-thymus and anti-E+ALL sera reacted with blasts from patients with E+ALL, but not with E-ALL. In contrast, anti-E+ALL serum reacted with 40 to 96% of blasts from all children with E-ALL, whereas of the four patients with E+ALL, two were negative and two had the lowest percentage of immunofluorescent cells (10 to 22%). These results were confirmed with the radiolabeled antibody assay. Patients with active E-ALL had cells bearing E-ALL antigen(s) in the peripheral blood and bone marrow, but the number of immunofluorescent cells was lower in blood. Cells reactive with anti-E-ALL serum did not react with thymus cells, blood lymphocytes, remission bone marrow cells, Raji cells, PWM and PHA-induced blasts and CLL cells bearing mIg (uk). These data suggest that the antigen detected on E-ALL blasts by anti-E-ALL serum is neither a HLA-related nor a cell differentiation antigen. Thus, by using antiserum to E+ALL blasts, we have confirmed the presence of a T cell-specific antigen(s) on E+ALL cells. This antiserum did not recognize other leukemia-associated antigens common to E+ and E-ALL. We have also demonstrated an antigen(s) which is regularly expressed on E-ALL blasts and is either not detectable or is present in a lower proportion of E+ALL blasts.     

Effect of thymic humoral factor in vitro on human bone marrow and blood cells from B-, T- and null-cell acute lymphatic leukemia.

The nature of null-cell acute lymphatic leukemia (ALL) was investigated with the aid of a thymic humoral factor (THF), bone marrow cells, and a local xenogeneic graft-versus-host reaction (GVHR). Lymphocytes obtained from the blood and bone marrow of six children with T-cell ALL, five with null-cell ALL, one with perinatal B-cell ALL, one with acute myelocytic leukemia, and one with erythroleukemia were tested for membrane surface markers (E, EAC, and SM Ig); functional activity of T cells was tested by a local GVHR. All of the specimens obtained at the initial presentation showed a lack of functional activity of the lymphocytes. Incubation of null cell and acute myelocytic leukemia (AML) bone marrow with THF led to the acquisition of the characteristics of functional, immunocompetent T cells. No such effect was seen when the bone marrow of T-cell ALL and peripheral blood lymphocytes of B-cell perinatal ALL were incubated with THF. This study demonstrates that the null cell in ALL bone marrow can be differentiated into a T cell whereas the stem cell in AML bone marrow constitutes a pluripotential undifferentiated cell which also can mature into a T cell.     

Antibody-producing human-human hybridomas. III. Derivation and characterization of two antibodies with specificity for human myeloid cells

Peripheral blood mononuclear cells from a patient with acute myeloid leukemia (AML) and spleen cells from a patient with chronic myeloid leukemia (CML) were fused with HAT-sensitive human B lymphoma cells (RH-L4) in attempts to generate human monoclonal antibodies (Mab) against antigens with high specificity for myeloid leukemia cells. Forty-seven of 246 hybridomas secreted Ig that bound to AML cell surface constituents, as determined by FACS analysis of viable cells that were FITC-stained with the human Mab as the first-step reagent and FITC-conjugated rabbit anti-human Ig as second-step. Two of the 47 human Mab (one from each patient and designated AML-19 and CML-20, respectively) bound to both autologous and allogeneic myeloid leukemia cells. No significant binding was observed to cell surface constituents on human bone marrow cells, granulocytes, lymphocytes, erythrocytes, thymocytes, monocytes, lymphoblastic leukemia cells, fibroblasts, malignant B and T lymphocytic cell lines, and murine bone marrow cells. Both human Mab were IgG and were cytotoxic to myeloid leukemia cells in the presence of complement. About 70% of peripheral blood cell samples from 46 AML patients contained AML-19- and CML-20-positive cells, but the reactivity pattern had no correlation to the morphologic FAB classification of the samples. The promyelocytic HL60 cell line and the K562 cell line reacted with the two antibodies. Dot blot analysis of binding of AML-19 and CML-20 to cellular extracts immobilized on nitrocellulose paper showed that both human Mab in this assay also reacted with normal bone marrow cells. This was supported by microscopic immunofluorescence because both human Mab stained intracytoplasmatic structures in normal bone marrow cells, but both intracytoplasmatic and cell surface components stained in myeloid leukemia cells. Moreover, immunoblotting demonstrated that both human Mab in leukemia cells reacted with two cellular proteins with Mr approximately 14,500 and 18,000, and in normal bone marrow cells with a molecule with Mr approximately 20,000. Immunoprecipitation of cell membrane molecules with both the AML-19 and CML-20 antibody precipitated from leukemic cells only the molecule with Mr approximately 18,000 and no components from normal bone marrow cells. It is concluded that myeloid leukemogenesis may result in generation of cell surface expression of either new or abnormally processed molecules that are immunogenic in the autochthonous host. These molecules may also be useful as markers in diagnosis of myeloid leukemia.     

A Monoclonal Antibody Defining Human B Cell Differentiation Antigen (HLB-1 Antigen)

A new monoclonal antibody specific for human B cell differentiation antigen (HLB-1) is produced by a hybridoma established by fusion of splenocytes of mice immunized with the Epstein-Barr virus (EBV)-transformed peripheral B cell line, RPMI-8057. This monoclonal, antibody designated anti-HLB-1 monoclonal antibody (anti-HLB-1), reacted with surface immunoglobulin (sIg)-positive B cells of normal peripheral blood and lymphoid tissues and sIg-positive leukemic cells. The cells of T cell leukemia, non-T non-B acute lymphoblastic leukemia (ALL) and nonlymphoid leukemia were HLB-1 negative. These data were further confirmed by studying a panel of cultured human hematopoietic cell lines. Anti-HLB-1 reacted with B cell lines derived from pre-B, Burkitt's lymphoma, B cell type ALL and EBV-transformed peripheral B cells. Anti-HLB-1 was reactive with only one of three human myeloma cell lines, and with none of the T cell, myeloid and non-T non-B ALL cell lines. This newly defined HLB-1 antigen is different from other conventional human B cell markers such as sIg, Ia antigens, and receptors for the Fc portion of Ig and complement C3.     

Cytostatic drug testing in human leukemias by means of multiparametric flow cytometry

Human bone marrow cells from 20 patients as well as the permanent human B-cell lines RPMI 1788, Raji, Daudi, T-cell lines Molt, CEM, Jurkat and the promyelocytic line HL 60 were assayed by means of a newly developed in vitro flow cytometric cytostatic drug assay. The cells were exposed to cytosine-arabinoside, L-asparaginase, daunorubicin, prednisone or vincristine. Surviving cells were stained after an incubation period of 2 to 7 days with esterase and pH-indicator dye ADB (1,4-diacetoxy-2,3-dicyanobenzene), dead cells with DNA-dye PI (propidium iodide). Dose-response curves were established using percent surviving cells. It was possible to evaluate bone marrow samples from 16 out of 20 patients. Seven samples were leukemic (acute myeloid leukemia (AML) n = 6, Non-Hodgkin's Lymphoma (NHL) n = 1). Nine samples were from patients either in complete remission or with benign diseases. Daunorubicin and cytosine-arabinoside were cytotoxic in both groups, whereas vincristine was effective mainly in the leukemic group (p less than 0.05). There was significant heterogeneity in the reactivity of AML-marrow cells from different patients to different drugs. The cell lines exhibited different patterns of sensitivity. Vincristine arrested cells in G2/M-phase, cytosine-arabinoside caused an increase of cells in the S-phase.     

A 125I-protein A-binding assay detecting antibodies to cell surface antigens

A 125I-protein A-binding assay detecting antibodies to cell surface antigens on human blood cells was developed and evaluated using sera from multitransfused nonleukemic patients sensitized against HLA antigens. The binding assay was found to be reproducible and more sensitive than conventional HLA testing. Seven patients with acute myelogenous leukemia and two patients with acute lymphoblastic leukemia successfully treated by chemotherapy were then investigated. Sera from seven of the patients studied in partial or complete remission demonstrated significant binding to autochthonous leukemic cells obtained from bone marrow or peripheral blood. In two cases sera taken during the leukemic stage demonstrated the most pronounced binding to the patients' own leukemic cells. Sera from four patients with demonstrable significant binding to autochthonous leukemic cells failed to bind to autochthonous remission cells when both types of target cells were tested in parallel. Differences in serum concentrations of IgG, IgA, and IgM were not the cause of the demonstrated increased binding of leukemic sera to autochthonous target cells. We propose that the 125I-protein A-binding assay presented in this paper detects antibodies reacting selectively with acute leukemia cells.     

Evidence for a common leukemia-associated antigen in acute myelogenous leukemia

Summary An antiserum was raised in rabbits to extracts of a pool of acute myelogenous leukemia cells. The immunization protocol used (antibody feedback) gave rise to antisera with marked specificity for AML extracts. After absorption, the antiserum demonstrated essentially no reactivity with cell extracts from 12 individual normal peripheral blood samples, while it reacted positively with 16 individual extracts from AML cells. Reactivity was assayed by the enzyme-linked immunosorbent assay (ELISA). The antiserum was not reactive with extracts from normal PHA-induced blast cells, with extracts of bone marrow cells from six individuals, or with three individual extracts of chronic lymphocytic leukemia (CLL) blast cells. These data indicate that this antiserum is detecting an antigen that is common to AML cells but may not be common to other blast cells.     

Increased proliferation of activated T-lymphocytes in response to a monoclonal antibody (anti-p68).

A series of monoclonal antibodies was produced by immunization of mice with cells of the human promonocytic cell line CM-S; one of these recognized a membrane antigen (MW 68,000) constitutively expressed by these cells. Antigen p68 was also found to be expressed on all granulocytic cells and most mononuclear leukocytes from normal human peripheral blood, but not on hemopoietic precursor cells from bone marrow. Various types of leukemic cells also expressed antigen p68 as did various transformed human cell lines whether derived from hemopoietic cells or from other tissues. Antigen p68 is involved in T-lymphocyte regulation. In fact, the antibody anti-p68 has a strong synergistic effect increasing the proliferative response of peripheral blood T-lymphocytes both in the mixed lymphocyte reaction and when the lymphocytes are stimulated by suboptimal doses of lectin (phytohemagglutinin), tumor promoter phorbol esters, or tetanus toxoid. The anti-p68 antibody synergizes with the active metabolite of vitamin D3, 1,25-dehydroxyvitamin D3, to induce monocyte to macrophage maturation and enhances the function of mature granulocytes stimulated with the granulocyte-macrophage colony-stimulating factor in vitro.     

Novel neutral glycolipid associated with adult T-cell leukemia and its pathological biochemistry

During attempts to make antibodies cross-reactive with human lymphocytes, we established a monoclonal antibody (VJ-41) from the alloimmunization of mice, that is, B10. A (3R) anti-B10. A (5R). VJ-41 reacted with all cases of freshly isolated adult T-cell leukemia cells (36 cases) but not with cells from other hematological disorders (more than 50 cases). Human T-cell leukemia virus type I healthy carriers also seemed to possess these VJ-41 antigen positive cells. However, in vitro established adult T-cell leukemia cell lines did not show the reactivity with VJ-41. Normal lymphocytes from humans or mice apparently did not carry this antigen, but mitogen activated lymphocytes or some in vitro maintained malignant cell lines of both human and mouse origins showed positive reaction. Having established solid phase radioimmunoassay to detect the VJ-41 antigen in plasma, it was found that healthy human T-cell leukemia virus type I carriers, but not the majority of adult T-cell leukemia patients, predominantly possessed this antigen. Even though immunochemical characterizations of cellular materials were unsuccessful, a certain neutral glycolipid was detected from healthy human T-cell leukemia virus type I carrier plasma by using thin-layer chromatography and immunostaining with the VJ-41 antibody.     

Distribution and modulation of a human leukemia-associated antigen (CALLA)

CALLA is a 100,000-dalton surface glycoprotein expressed by malignant cells of patients with clinically important subtypes of acute leukemia. Incubation of human leukemic cells expressing CALLA with specific monoclonal antibody (J5) at 37 degrees C causes rapid and selective internalization and degradation of this antigen (antigenic modulation). In these studies we show that CALLA-specific monoclonal antibodies also identify a cell surface glycoprotein having a m. w. of approximately 100,000 on 2 to 6% of nonmyeloid nucleated cells of normal adult bone marrow, on normal fibroblasts in tissue culture, and on cells of several nonhematopoietic human tumor cell lines. J5 antibody similarly modulates the surface expression of CALLA on nonleukemic cell populations, although the extent of modulation at a given concentration of antibody varied considerably. Modulation was almost complete for CALLA on cells of normal bone marrow, but was highly variable for cells of nonhematopoietic cell lines, possibly reflecting variability in antibody access to surface antigen. Using fluoresceinated or iodinated J5 antibody to modulate expression of CALLA on cells of leukemic cell lines, we show that antibody-antigen complexes undergo a temperature-dependent redistribution on the cell surface during modulation to form microaggregates. Antibody as well as antigen is then internalized. Studies of [35S]methionine-labeled cells indicate that synthesis of CALLA continues despite modulation of its surface expression by specific antibody, implying that the presence of CALLA on the cell surface reflects a dynamic equilibrium between the processes of surface expression of newly synthesized glycoprotein and its spontaneous and antibody-mediated clearance. The implications of these observations for immunotherapy are discussed.     

Spontaneous cytotoxicity of human lymphoblast cell lines mediated by normal peripheral blood lymphocytes. I. Differential susceptibility of T-versus B-cell lines.

Human lymphoblast cell lines of B- and T-cell origin have been tested for their ability to serve as targets in a 4-hr ⁵¹Cr release microcytotoxicity assay using normal human peripheral blood lymphocytes as effector cells. Cell lines of T-cell origin were susceptible to lysis in this assay by effector lymphocytes from all normal donors tested. Cell lines of B-cell origin were repeatably lysed by normal lymphocytes from some, but not all donors. Spontaneous cytotoxicity of B-cell lines, when observed, was also quantitatively less than was obtained using T-cell lines as targets. One cell line (RPMI-7666), of B-cell origin, was not susceptible to spontaneous cytotoxicity by almost all of the normal lymphocyte effectors tested. Lymphocytes from patients with acute lymphoblastic leukemia in remission were less capable of effecting lysis in this assay.     

Ex vivo elimination of neoplastic T-cells from human marrow using an anti-Mr 41,000 protein immunotoxin: potentiation by ASTA Z 7557

ASTA Z 7557 potentiated the ex vivo efficiency of a T-cell directed immunotoxin containing pokeweed antiviral protein (PAP). We used an immunotoxin of pan-T monoclonal antibody 3-A1 directed against p41 antigen expressed both on normal and leukemic T-cells. Treatment with 3A1-PAP in combination with ASTA Z 7557 produced 7-8 logs elimination of target lymphoblastic leukemia cells. Our data suggest that this new strategy shows potential for more effective ex vivo marrow purging in autologous marrow transplantation for acute lymphoblastic leukemia.     

A novel cell surface antigen (T305) found in increased frequency on acute leukemia cells and in autoimmune disease states

Monoclonal antibody T305, prepared by immunizing mice with the T-ALL derived cell line RPMI-8402, immunoprecipitates a single chain glycoprotein with m.w. 160,000 daltons (under reducing conditions) or 180,000 daltons (under nonreducing conditions). In immunofluorescence assays, antibody T305 reacted with a subpopulation of T cells in normal blood (22 +/- 6%), thymus (28 +/- 11%), and lymph node (24 +/- 6%). Increased frequency of T cells reactive with antibody T305 was found in peripheral blood of patients with infectious mononucleosis (greater than 80%), graft-vs-host disease after bone marrow transplantation (65 +/- 11%), acquired immunodeficiency syndrome (53 +/- 12%). The T cells in synovial fluid of patients with rheumatoid arthritis had increased frequency of antibody T305 reactive cells (59 +/- 8%) as compared to their peripheral blood (18 +/- 7%). Two color immunofluorescent studies demonstrated that the T305+ T cells predominantly co-stained with antibody Leu 2a (suppressor/cytotoxic subset) in both normals and disease state blood. After cell sorting to obtain T305+ and T305- subpopulations, we demonstrated that a) natural killer and antibody-dependent cellular cytotoxicity activity in normal blood was in the T305+ but not T305- T cells; b) cytotoxic T cells induced by mixed lymphocyte reaction were predominantly T305+; c) T305- T cells could be induced in vitro to express T305 antigen by mitogens or allogeneic B cells; d) the DNA content of T305+ and T305- T cells in normal blood was similar (greater 95% of cells with G0/G1 level); e) after mitogen stimulation, T305 antigen induction on previously T305- cells occurs before S-phase; and f) significantly more [3H]-thymidine after mitogen stimulation was incorporated by originally T305- cells than by originally T305+ cells (p less than 0.001). The T305 antigen was not restricted to T cells because it was also found on myeloid precursors in bone marrow but was not present on polymorphonuclear leukocytes, red blood cells, platelets, muscle, liver, skin, kidney, lung, or brain. Antibody T305 was found on 24/25 cases of acute leukemia (6 T-ALL, 10/11 cALL, 7 AML, and 1 AMOL) but not on 18 cases of chronic leukemia (B-CLL, T-CLL, null CLL, CML). The importance of the T305 antigen is that it is present on a high number of T cells in certain autoimmune diseases and on virtually all acute leukemia cells. Its distribution on immature and in vitro activated cells suggests that it may represent a receptor for signals related to cellular replication or differentiation.     

A newly established human acute lymphoblastic leukemia cell line with characteristics of the earliest B-cell maturation

A new human acute lymphoblastic leukemia (ALL) cell line, designated HBL-3, was established from the bone marrow of a patient with non-T-ALL. The HBL-3 cell line expressed B4 (CD 19), BA-1 (CD 24) and HLA-DR antigens, but not surface immunoglobulin (SIg) or cytoplasmic immunoglobulin (CIg). The cell line lacked the common acute lymphoblastic leukemia antigen (CALLA) and antigenic markers characteristic of T-cell and myeloid cell lineages. The HBL-3 cells had structural rearrangements of both the homologous chromosome 9s, including a translocation with chromosome 1 which has been reported in a patient with common ALL. The cell line had rearranged immunoglobulin heavy chain genes but retained germ-line kappa light chain genes and germ-line T-cell receptor beta- and gamma-chain genes. The HBL-3 cell line was strongly positive for terminal deoxynucleotidyl transferase (TdT). These findings indicate that the HBL-3 cell line is derived from the earliest B-cell committed to B-cell lineage.     

A comparison of the rate of DNA synthesis in myeloblasts from peripheral blood and bone marrows of patients with acute nonlymphocytic leukemia

Durations of S-phase (Ts) and total cell cycle times (Tc) were measured from the peripheral blood (PB) and bone marrow aspirates (BM) of five patients with acute nonlymphocytic leukemia (ANLL). Intravenous bromodeoxyuridine (BrdU) was used as the first label for S-phase cells and a monoclonal anti-BrdU antibody was used to detect the positive cells. Tritiated thymidine [( 3H]Tdr) was used as a second label in vitro, and the Ts was calculated by counting the number of cells labeled either by BrdU or by [3H]Tdr or by both. Our data demonstrate that the duration of S-phase in myeloblasts obtained from BM is quite similar to that of circulating leukemic cells. Finally, the most accurate assessment of percentage of myeloblasts actively engaged in DNA synthesis can be obtained only from bone marrow biopsies following in vivo labeling.     

Monoclonal antibodies to carbohydrate antigens in autologous bone marrow transplantation

Normal and malignant myeloid cells express a highly immunogenic oligosaccharide, lacto-n-fucopentaose-III (LNF-III), that has been identified by numerous monoclonal antibodies (MoAb). We have been interested in the use of a particular monoclonal antibody to LNF-III, PM-81, in the treatment of patients with acute myelogenous leukemia using the antibody to treat bone marrow in vitro. Following in vitro treatment of bone marrow with PM-81 and another MoAb, AML-2-23, the remaining cells are used as an autograft in a patient treated with high-dose chemotherapy and radiotherapy. In order to enhance the ability of the MoAb to lyse leukemic cells in the remission bone marrow, we have explored the effect of neuraminidase treatment on leukemia cells. In this paper we describe that myeloid leukemia cells expressing low levels of LNF-III by immunofluorescence can be shown to have high levels of LNF-III after neuraminidase treatment. In addition, we show that normal bone marrow progenitor cells do not have cryptic LNF-III antigen, thus allowing the application of this finding to the clinical setting. Moreover, we have shown that leukemia colony-forming cells from one patient with acute myelogenous leukemia express cryptic LNF-III and that after exposure to neuraminidase there was an increased ability of PM-81 in the presence of complement to eliminate these colony forming cells. These data indicate that the LNF-III moiety is almost universally expressed on myeloid leukemia cells and their progenitors but not expressed on normal progenitors. Thus, it may be possible to enhance leukemia cell kill in vitro by neuraminidase treatment of bone marrow.