Complement-dependent Cytotoxicity Assay Research Articles

No Expression of a Rous Sarcoma Virus-Induced Tumor Antigen in Mammalian Cells Infected With Retroviruses Transducing Other Oncogenes of the <italic>src</italic> Gene Family<xref ref-type="fn" rid="FN1">2</xref>

Immunization with mouse and rat cells transformed by Rous sarcoma virus (RSV) or by B77 avian sarcoma virus (ASV) induced complete transplantation resistance against an RSV-induced mouse tumor (CSA1M) in syngeneic hosts. In contrast, most of the mice immunized with a Fujinami sarcoma virus-transformed rat fibroblast line (FSV-3Y1), a feline sarcoma virus-transformed cat fibroblast line (FeSV-FEF), an Abelson leukemia virus-infected Balb/3T3 cell line (AbLV-3T3), or an uninfected 3Y1 cell line could not reject the CSA1M. Serologic analysis with the use of a complement-dependent cytotoxicity assay supported the results of transplantation studies. The mouse and rat cells transformed by RSV or B77 ASV expressed a common tumor-specific cell surface antigen (TSSA) detected by syngeneic antiserum against the CSA1M, whereas none of the FSV-3Y1, FeSV-FEF, and AbLV-3T3 cells expressed the TSSA. These results suggest that the common TSSA in the mouse and rat cells transformed by RSV or B77 ASV containing src gene is not shared with mammalian cells infected with retroviruses transducing other oncogenes of the src gene family (i.e., fps, fes, and abl).

Immune rejection mechanisms in murine leukemia. I. Timing of tumor cell rejection process relative to the development of humoral and cell-mediated cytotoxic immune responses.

Studies were undertaken to investigate the relationship between cell-mediated and humoral immune responses in the rejection of L1210/MTX-Rev (LR) leukemia cells in CD2F1 mice. The anti-LR antibody (humoral) response was defined by both its cytotoxic antibody titer and isotype composition, assessed by a complement-dependent cytotoxicity assay and radioimmune assay, respectively. The cytotoxic thymus (T)-derived lymphocyte response in the spleen was quantitated by 125I release by 125I-IUdR-labelled target cells. Analysis of the sera of tumor-bearing mice indicated that the LR leukemia cells elicited a wide spectrum of anti-tumor antibody isotypes. In mice that mounted a rejection response, the IgM anti-LR response was transient, while in those that did not, the IgM response persisted. Antibody titers for all isotypes remained low until the onset of LR rejection. At that time, high titers of IgG anti-LR antibodies, predominantly of the IgG2a subclass, were detected in sera of tumor-free mice. Thus, the LR rejection response coincided best with the appearance of high titers of IgG2a anti-LR antibodies. A single i.p. injection of viable LR cells elicited a potent cell-mediated immune response; neither the appearance nor the magnitude of the cell-mediated immune response as measured in the spleen correlated well with the onset or the strength of the LR rejection response in the peritoneal cavity. The LR peritoneal cell population grew unabated in the presence of an intense spleen cell-mediated immune response, and the rejection process began at a time when little cellular immunity could be detected. These results suggest that in the rejection response IgM antibodies contribute little, if any, to the process, and the role of cytotoxic T lymphocytes is questionable (uncertain). The rejection of LR cells appears to be primarily mediated by IgG2a antibodies.

Monoclonal Antibodies Detecting Antigenic Determinants With Restricted Expression on Erythroid Cells: From the Erythroid Committed Progenitor Level to the Mature Erythroblast

Two new cell surface antigens specific for the erythroid lineage were defined with cytotoxic IgM monoclonal antibodies (McAb) (EP-1; EP-2) that were produced using BFU-E-derived colonies as immunogens. These two antigens are expressed on in vivo and in vitro derived adult and fetal erythroblasts, but not on erythrocytes. They are not detectable on resting lymphocytes, concanavalin-A (Con-A) activated lymphoblasts, granulocytes, and monocytes or granulocytic cells or macrophages present in peripheral blood or harvested from CFU-GM cultures. Cell line and tissue distributions distinguish McAb EP-1 and EP-2 from all previously described monoclonal antibodies. McAb EP-1 (for erythropoietic antigen-1) inhibits the formation of BFU-E and CFU-E, but not CFU-GM, colonies in complement-dependent cytotoxicity assays. By cell sorting analysis, about 90% of erythroid progenitors (CFU-E, BFU-E) were recovered in the antigen-positive fraction. Seven percent of the cells in this fraction were progenitors (versus 0.1% in the negative fraction). The expression of EP-1 antigen is greatly enhanced in K562 cells, using inducers of hemoglobin synthesis. McAb EP-2 fails to inhibit BFU-E and CFU-E colony formation in complement-dependent cytotoxicity assays. EP-2 antigen is predominantly expressed on in vitro derived immature erythroblasts, and it is weakly expressed on mature erythroblasts. The findings with McAb EP-1 provide evidence that erythroid progenitors (BFU-E and CFU-E) express determinants that fail to be expressed on other progenitor cells and hence appear to be unique to the erythroid lineage. McAb EP-1 and EP-2 are potentially useful for studies of erythroid differentiation and progenitor cell isolation.

Monoclonal antibodies detecting antigenic determinants with restricted expression on erythroid cells: from the erythroid committed progenitor level to the mature erythroblast

Two new cell surface antigens specific for the erythroid lineage were defined with cytotoxic IgM monoclonal antibodies (McAb) (EP-1; EP-2) that were produced using BFU-E-derived colonies as immunogens. These two antigens are expressed on in vivo and in vitro derived adult and fetal erythroblasts, but not on erythrocytes. They are not detectable on resting lymphocytes, concanavalin-A (Con-A) activated lymphoblasts, granulocytes, and monocytes or granulocytic cells or macrophages present in peripheral blood or harvested from CFU-GM cultures. Cell line and tissue distributions distinguish McAb EP-1 and EP-2 from all previously described monoclonal antibodies. McAb EP-1 (for erythropoietic antigen-1) inhibits the formation of BFU-E and CFU-E, but not CFU-GM, colonies in complement-dependent cytotoxicity assays. By cell sorting analysis, about 90% of erythroid progenitors (CFU-E, BFU-E) were recovered in the antigen-positive fraction. Seven percent of the cells in this fraction were progenitors (versus 0.1% in the negative fraction). The expression of EP-1 antigen is greatly enhanced in K562 cells, using inducers of hemoglobin synthesis. McAb EP-2 fails to inhibit BFU-E and CFU-E colony formation in complement-dependent cytotoxicity assays. EP-2 antigen is predominantly expressed on in vitro derived immature erythroblasts, and it is weakly expressed on mature erythroblasts. The findings with McAb EP-1 provide evidence that erythroid progenitors (BFU-E and CFU-E) express determinants that fail to be expressed on other progenitor cells and hence appear to be unique to the erythroid lineage. McAb EP-1 and EP-2 are potentially useful for studies of erythroid differentiation and progenitor cell isolation.

Polymorphic and Monomorphic HLA-DR Determinants on Human Hematopoietic Progenitor Cells

The expression of monomorphic Ia-like antigens and polymorphic (allotypic) HLA-DR determinants on CFU-GM, BFU-E, CFU-E, and CFU-GEMM was studied in bone marrow and peripheral blood cells from normal healthy individuals. Using various polyclonal and monoclonal anti-Ia-like antibodies, the presence of HLA-DR backbone antigens was shown on all hematopoietic progenitor cells (HPC) studied, both in complement-dependent cytotoxicity assays and in fluorescence-activated cell sorting (FACS). The expression of allotypic determinants was demonstrated on all HPCs, using the HLA-DR typing sera anti-HLA-DR1, 2, 3, 4, 5, and 7. The Class II antigen MT-2 was also shown on all HPCs, using both monoclonal and alloantisera, whereas the MB-1 (DC-1) determinant could not be demonstrated on HPCs. This might open the possibility of removing MB-1-positive malignant cells from the graft in autologous bone marrow transplantation.

Monoclonal anti-Ia murine alloantibodies crossreactive with the Ia-homologues of other mammalian species including humans.

Two murine monoclonal anti-Ia alloantibodies (ISCR2 and ISCR3) crossreactive with lymphocytes of other mammalian species including humans have been established. These hybridomas were derived from A.TH mice immunized with A.TL lymphocytes, and they reacted with murine Ia antigens, as judged by complement-dependent cytotoxicity assays and chemical analysis. Genetic mapping analysis suggested that ISCR2 recognizes the allotypic determinants shared between the I-A and the I-E molecules. On the other hand, the specificity of ISCR3 was shown to correspond to the serological specificity Ia.7 and directed to the I-E subregion gene product (E alpha). The genetic specificity of ISCR3 was also confirmed by sequential precipitation analysis performed on the internally labeled B10.BR spleen cell antigen. These antibodies showed interspecies crossreactions with non-murine lymphocytes. ISCR2 showed reactivity on human and dog lymphocytes, but not with rat lymphocytes. ISCR3 reacted with human, dog, and rat lymphocytes but not with chicken or frog lymphocytes. Furthermore, ISCR3 was shown to be capable of eliminating a population of cells necessary for induction of human allogeneic killer cells in vitro. Treatment of stimulator cells with ISCR3 and complement drastically reduced the induction of the killer activity.

Comparison of complement-dependent cytotoxicity and indirect immunofluorescence for enumeration of T-cell subpopulations in human peripheral blood

The monitoring of T-lymphocyte subsets of recipients of organ grafts enables studies on immune reconstitution (after bone-marrow transplantation) and may predict graft rejection (after kidney transplantation). Quantitation of human peripheral T-lymphocyte subsets from healthy volunteers and from recipients of a bone-marrow graft by a complement dependent cytotoxicity (CDC) assay, based on the use of propidium iodide, and by an indirect immunofluorescence (IIF) technique has been compared using the monoclonal antibodies OKT3, OKT4 and OKT8. Except for OKT3 in healthy individuals — for which no significant difference was found between CDC and IIF — CDC detected significantly more cells of each subset than IIF. Furthermore, the CDC results indicated the presence of low numbers of OKT4 +8 + cells in the peripheral blood of healthy individuals and — with higher numbers — following marrow transplantation. Results of depletion experiments, obtained by fluorescence activated cell sorting (FACS) for either OKT4 or OKT8, supported this conclusion. OKT4/OKT8 ratios were calculated from enumerations by the CDC assay and by the IIF assay and found to be linearly related, both in healthy persons and in marrow-graft recipients. Thus, the CDC assay is a reliable method for monitoring T-cell subsets, allowing detection of lymphocytes carrying low densities of membrane determinants.

Comparison of approaches for augmenting the serologic response to the individually specific methylcholanthrene-induced sarcoma-Meth A: pretreatment with cyclophosphamide is most effective.

We have previously reported the production of antisera against a highly restricted antigen expressed on the BALB/c sarcoma Meth A. Because induction of the antibody required many vaccinations with irradiated and unirradiated Meth A cells over a prolonged period, we have investigated methods of improving the efficiency of producing Meth A antibody in BALB/c mice. Three general approaches were used: immunizing with irradiated Meth A cells mixed with adjuvants, immunizing with irradiated Meth A cells modified by treatment with chemicals, or administration of tumor cell vaccines in conjunction with low dose cyclophosphamide. These vaccine preparations were injected subcutaneously into groups of BALB/c mice five to six times at 2-wk intervals. Sera from all mice were screened for reactivity to Meth A by complement-dependent cytotoxicity, protein A, and mixed hemadsorption assays. Twenty-three vaccine preparations containing adjuvants or modified cells were tested in individual groups of mice. Mice in some groups receiving adjuvants and in all groups receiving modified cells produced antibody, but only after four to six vaccinations. In contrast, nine of the 14 mice immunized with irradiated Meth A cells (unmodified and without adjuvant) in conjunction with 10 or 100 mg/kg body weight of cyclophosphamide, and all nine of the mice receiving 25 mg/kg cyclophosphamide made Meth A antibodies after only one or two vaccinations. The titer of antibodies produced after one treatment with cyclophosphamide and irradiated Meth A cells was at least as high as that achieved after five or six vaccinations in our other trials. The specificity of these antibodies for the Meth A antigen was established by absorption analysis. We conclude that treatment with cyclophosphamide before vaccination is highly effective in augmenting the humoral immune response to the Meth A antigen.

Humoral response of melanoma patients to two different tumor-associated antigens.

Sera of 8 stage II melanoma patients undergoing surgical adjunctive immunotherapy with bacille Calmette Guérin (BCG) plus melanoma cell vaccine (MCV) were assayed for humoral response by the complement-dependent antibody cytotoxicity (CDAC) assay and by the microcomplement fixation (MCF) test. The patients developed high levels of cytotoxic (CTX) and complement-fixing (CF) antibodies to the UCLA-SO-M14 (M14) cells, one of the three melanoma cell lines in the MCV. Significant rises in CTX and CF antibodies occurred one month post-immunotherapy. While the level of CTX antibodies was maintained for 11 months thereafter, the titer of CF antibodies was sustained for seven months, then gradually declined. When the sera were absorbed with lymphoblastoid (ML14) cells which are autologous to the M14 cells, two residual peaks of CTX antibodies, one and four months postimmunotherapy and two peaks of CF antibodies, one and seven months postimmunotherapy, emerged. Two sera that exhibited high levels of CTX and CF antibodies one month postimmunotherapy were absorbed with ML14 cells and human fetal brain tissue. The reactivity of one serum in both the CDAC and MCF assays were abolished, whereas the reactivity of the other serum was not significantly diminished in either assay. These data indicate that the stage II post-surgical melanoma patients developed a humoral immune response to at least two distinct tumor antigens on the membrane of the M14 cells. One of these antigens appeared to be of fetal origin (OFA), the other M14-associated (TAA). Both antigens, OFA and TAA, were involved in complement-dependent antibody cytotoxic and complement fixation reactions in vitro.

Serotherapy of L1210 murine leukaemia--reasons for ineffectiveness of in vivo treatment by L.1 monoclonal antibody.

A monoclonal antibody (L.1), reacting in vitro specifically with L1210 leukaemia cells in a complement-dependent cytotoxicity assay (CDC), has been exploited for serotherapy studies. Different regiments of L.1 treatment of CD2F1 mice bearing the semi-syngeneic L1210 leukaemia did not prolong the life span of tumor-bearing animals. Moreover, the administration of L.1 did not enhance the antitumour effects of cyclophosphamide. Studies of in vivo localization showed that L.1 was able to bind specifically to L1210 leukaemic cells, although 30-40% of the cells remained negative. The presence of L.1 in mouse blood was demonstrated up to 15 days after the inoculation. On the other hand, in vivo administration of L.1 was probably accompanied by loss of the cytotoxic activity, perhaps through a mechanism of complement inactivation, since the presence of undiluted normal mouse serum in a CDC assay inhibited the cytotoxic activity of L.1. Moreover, serum from L.1-treated mice did not display any cytotoxic activity, although the presence of the antibody could be demonstrated by indirect immunofluorescence. Shedding of the antigen defined by L.1 was probably not responsible for the failure of the serotherapy, since the L.1 neutralizing antigen could be found in body fluids only long after the start of therapy.

Immunological studies in patients with aplastic anemia who rejected marrow grafts from HLA-identical sibling donors.

Marrow graft rejection after transplantation from an HLA-identical sibling for the treatment of aplastic anemia has an immunological basis in most cases. We previously studied the post-transplantation sera of 20 consecutive patients using the chromium release assays of complement-dependent cytotoxicity (CDC) and/or antibody-dependent cell-mediated cytotoxicity (ADCC). Reactivity against peripheral blood mononuclear cells of the marrow donors was found which was associated with graft rejection. Sera of 17 consecutive patients who rejected their grafts have now been studied. Donor cells susceptible to cytotoxicity in both the CDC and ADCC assays were composed of both adherent and nonadherent cells not forming spontaneous rosettes with sheep red blood cells. Absorption studies demonstrated that the reactivities were mediated by antibody, and inhibition studies with 2-mercaptoethanol indicated that the active component was IgM.

Serologically detectable MHC and tumor-associated antigens on B16 melanoma variants and humoral immunity in mice bearing these tumors.

We compared the expression of serologically detectable MHC and tumor-associated antigens on low and high metastasis variants of B16 melanoma tumor. Complement-dependent cytotoxicity, radioimmunobinding, and quantitative absorption assays showed that with respect to both types of antigens the low metastasis variant B16-F1 expressed a higher serologically detectable antigenicity than B16-F10, its high metastasis counterpart. A reverse situation existed in terms of in vivo immunogenicity of these metastasis variants. Sera from C57BL/6 mice bearing locally growing B16-F10 tumors had a higher binding activity to B16 cells than sera from B16-F1 bearers. Accordingly, in vivo propagating B16-F10 tumors had a higher content of tumor-associated Ig than B16-F1 tumors.

Lack of reactivity of sera from [formula omitted][formula omitted]-infected and recovered cattle against cell membrane antigens of [formula omitted] transformed cell lines

Sera from Theileria parva infected, recovered and rechallenged cattle were tested in complement-dependent cytotoxicity, membrane immunofluorescence and antibody-dependent cellular cytotoxicity assays for the presence of antibodies against cell membrane antigens of T . parva transformed cell lines. In the complement-dependent antibody-mediated cytotoxicity assay, sera from lethally infected animals were negative. Some recovered cattle showed a positive reaction, but such reactions were also observed when an eland cell line infected with T . taurotragi , and bovine lymphoblastoid cells were used as targets. Reaction was less against Ig-negative peripheral blood lymphocytes. Evidence is presented that these reactions could be evoked by attachment of immune complexes to Fc-receptors. It is concluded that cattle exposed to T . parva infection do not develop antibodies against specific T . parva (or T . parva -induced) cell surface antigens.

Detection of H-2 antigens on 8-cell mouse embryos.

H-2 antigens were detected on 80-cell mouse embryos of the b, k, and d haplotypes, but not on 8-cell embryos of the a haplotype. The technique used for detection of H-2 antigens on the embryos was a highly sensitive, automated, complement-dependent cytotoxicity assay, based on the principle that live embryos will incorporate 3H-thymidine into DNA, but embryos killed with antiserum and complement will not. The amount of H-2 present on 8-cell embryos is apparently less than that present on blastocysts. A possible function of H-2 antigens in development is proposed.

Heterogeneity of the natural humoral anti-tumor immune response in mice as shown by monoclonal antibodies.

Four anti-tumor cytotoxic antibodies, namely, A6 of the IgG2 class and B3, C2, and D1 of the IgM class, were obtained in monoclonal form, from hybridization of mouse myeloma cells with spleen cells of normal untreated mice selected for high natural anti-tumor immune responses. The specificity of the four monoclonals was tested by complement-dependent cytotoxicity assay on the reference EL4 lymphoma at different days of in vivo transplant, on normal adult and fetal tissues, on the SC-1 fibroblastic cell line uninfected or infected with a murine ecotropic type-C virus, and on a panel of murine lymphomas of different origin. The four antibodies had different specificities: the A6 and B3 recognized virus-related structures, the C2 a structure expressed on fetal cells, and the D1 a normal component of fibroblasts. The different classes and specificities of the four monoclonals, as well as their in vitro-demonstrated synergistic cooperation, give support to the hypothesis that the natural humoral anti-tumor cytotoxic immune response could be the result of the cooperative activity of an array of heterogeneous antibody molecules.

Serologic and cellular characterization of products of a new major histocompatibility gene region, RT1.C, of the rat; possible homology to mouse H-2 Qa.

The A region of the major histocompatibility system of the rat, RT1, codes for ubiquitously expressed class I antigens and the RT1.B region determines class II (Ia) molecules. A third region, RT1.C, adjacent to B, has been described recently by histogenetic methods. Data are now presented on the analysis of RT1.C region products by the use of alloantisera and cytotoxic T lymphocytes, both prepared in two, C region congenic strain combinations. C region antigens have been detected serologically by complement-dependent cytotoxicity assays on the majority of B and T lymphocytes, but not on erythrocytes or liver and brain tissue by absorption tests. With the typing sera available only three "alleles" could be identified among 13 different rat strains, so that the RT1.C polymorphism appears to be restricted. Cytotoxic T lymphocytes could be induced against C region products after in vivo immunization and in vitro restimulation; killing was not restricted by the A and B regions. The concept that RT1.C is homologous to H-2 Qa is discussed.

Characterization of a monoclonal antibody to L1210 leukaemia.

A mouse of monoclonal cell line (L1) was produced by fusing the mouse myeloma P3X63/Ag8 with CD2F1 spleen cells immunized with a highly immunogenic subline of L1210 leukaemia (L1210/DTIC). A very few positive clones (1%) were isolated and one of these was chosen for detailed study. The monoclonal antibody L1 is an IgM immunoglobulin strongly reacting in a complement-dependent cytotoxicity assay against L1210/Cr leukaemia and its more or less immunogenic sublines. The specificity of the L1 antibody against L1210 leukaemia was studied by extensive screening with normal adult and foetal tissues, lymphoid tissues from several independent strains and a panel of the most common experimental tumours, to all of which it was unreactive. Attempts at immunotherapy were carried out in DBA/2 mice challenged with L1210 leukaemia and treated with L1 (ascites) and complement. Although the in vitro cytotoxic titre of ascites fluid from mice bearing hybridoma was very high (10(-7)), no therapeutic effect was obtained in vivo.

B lymphoblast antigen (BB-1) expressed on Epstein-Barr virus-activated B cell blasts, B lymphoblastoid cell lines, and Burkitt's lymphomas.

Two new cell surface antigens expressed on B lymphoblastoid cell lines (B-LCL) were defined with cytotoxic mouse monoclonal antibodies. One marker, BB-1 (for B lymphoblast antigen-1), was detected on human and nonhuman primate B-LCL, Epstein-Barr virus (EBV)-activated B cell blasts, most Burkitt's lymphomas, and Ia+ B lymphoblast-like myelomas. Polyclonal B cell activators such as pokeweed mitogen (PWM) and lipopolysaccharide (LPS) also induced the expression of BB-1 on immunoglobulin (Ig)-positive cells. In contrast, BB-1 could not be detected on normal lymphoid tissues by complement-dependent cytotoxicity and immunofluorescence (IF) assays or by analysis with a fluorescence-activated cell sorter (FACS). T cell blasts, T cell leukemias, and pre-B cell or erythroblastic leukemia cell lines were also BB-1 negative. Of particular interest was the finding that BB-1 was expressed on the Jijoye lymphoma but only marginally on a subline of Jijoye, P3HR-1, that lacks receptors for EBV and produces a defective virus incapable of transforming lymphocytes. A second lymphoblast antigen (LB-1) unlike BB-1, was present on both T and B cell blasts and virus-transformed T- and B-LCL but not on normal lymphoid tissues.

Expression of Ia-antigens on normal and chronic myeloid leukemic human granulocyte-macrophage colony-forming cells (CFU-GM) is associated with the regulation of cell proliferation by prostaglandin E

The presence of la-antigens and their relationship to the inhibitory effect of prostaglandin E on the proliferation of human CFU-GM was studied in animals and patients with chronic myeloid leukemia. Consistent reduction of normal colony formation to approximately 50% of baseline levels was observed using a monoclonal anti-human la antibody in a complement-dependent cytotoxicity assay titrated over serial dilutions. ELimination of the la-antigen-bearing CFU-GM population was associated with virtually a complete loss of responsiveness to the inhibitory effects of prostaglandin E. Maintenance of bone marrow cells in short-term suspension culture at 37 degrees C prior to agar culture resulted in the loss of detectable la-antigen on the CFU-GM and, similarly, loss of response to prostaglandin. In contrast, most patients with chronic myeloid leukemia showed greatly reduced levels of la-antigens on their CFU-GM in fresh marrow together with lack of prostaglandin sensitivity, suggesting a correlation with the abnormal growth regulation observed in these patients. In two chronic myeloid leukemia patients, levels of la-antigen higher than that observed in the majority of patients could be detected and correlated with a residual response to prostaglandin E. These results suggest a relationship in normals between the expression of la-antigens on CFU-GM and the physiologic response to regulation by prostaglandin E, and a possible mechanism for the aberrant regulatory response in patients with chronic myeloid leukemia.

Expression of Ia-antigens on normal and chronic myeloid leukemic human granulocyte-macrophage colony-forming cells (CFU-GM) is associated with the regulation of cell proliferation by prostaglandin E

The presence of la-antigens and their relationship to the inhibitory effect of prostaglandin E on the proliferation of human CFU-GM was studied in animals and patients with chronic myeloid leukemia. Consistent reduction of normal colony formation to approximately 50% of baseline levels was observed using a monoclonal anti-human la antibody in a complement-dependent cytotoxicity assay titrated over serial dilutions. ELimination of the la-antigen-bearing CFU-GM population was associated with virtually a complete loss of responsiveness to the inhibitory effects of prostaglandin E. Maintenance of bone marrow cells in short-term suspension culture at 37 degrees C prior to agar culture resulted in the loss of detectable la-antigen on the CFU-GM and, similarly, loss of response to prostaglandin. In contrast, most patients with chronic myeloid leukemia showed greatly reduced levels of la-antigens on their CFU-GM in fresh marrow together with lack of prostaglandin sensitivity, suggesting a correlation with the abnormal growth regulation observed in these patients. In two chronic myeloid leukemia patients, levels of la-antigen higher than that observed in the majority of patients could be detected and correlated with a residual response to prostaglandin E. These results suggest a relationship in normals between the expression of la-antigens on CFU-GM and the physiologic response to regulation by prostaglandin E, and a possible mechanism for the aberrant regulatory response in patients with chronic myeloid leukemia.