Search

Following a cryopreservation step, short-term cultures of circulating leukemic blasts from a patient with acute myeloid leukemia (AML) were performed. Because cultured tumor cells became susceptible to natural killer (NK) activity, in vitro alteration of the blasts was studied. Immediately after thawing, cell suspensions consisted of a relatively homogeneous population of undifferentiated blasts. In culture, tritiated thymidine uptake by the leukemic cells was low during the first 24 hours and then increased (X20) to a peak on day 7. The cell concentration started to increase on day 4. On day 8, less than 10% of the cultured cells still appeared as undifferentiated blasts, whereas up to 60% were granular and 30% to 40% had a monocytoid morphology. Prior to being cultured, the blasts were resistant to resting and IL2- activated natural killing. When the kinetics of in vitro acquired susceptibility were studied, it was found that maximum cytotoxicity against these leukemic cells was reached within 24 hours. Thus, the blasts had become NK-sensitive prior to increase in DNA synthesis, proliferation, and differentiation based on morphological and cytochemical criteria. In contrast, there was a positive correlation between acquired susceptibility and surface expression of an activation antigen, termed TNKtar. To dissect further the mechanisms of acquired susceptibility, a series of six NK clones representing four distinct phenotypes of NK active lymphocytes were tested against the leukemic cells. Immediately after thawing, blasts were essentially resistant to all clones, whereas they were strongly killed by 5 of 6 clones when cultured for 24 hours. Cold target inhibition assays indicated that resistance of fresh blasts was likely to be due to a binding defect. These results suggested that tumor cells became susceptible because they surface-expressed NK target structure(s) in the early phase of an activation process leading to their proliferation and/or differentiation. This hypothesis was substantiated for one clone, termed JT9, because the anti-TNKtar antibody blocked cytotoxicity of JT9 cells against the cultured blasts.

To identify cells with potential antileukemia activity following bone marrow transplantation, we have monitored immunologic reconstitution in a patient with acute lymphocytic leukemia in second remission who received intensive chemotherapy and total body irradiation followed by infusion of allogeneic histocompatible marrow. Prior to transplantation, donor bone marrow cells were depleted of T lymphocytes by in vitro treatment with anti-T12 monoclonal antibody and rabbit complement. In the first 3 weeks following bone marrow transplantation, the predominant regenerating mononuclear cell population in peripheral blood exhibited a phenotype characteristic of natural killer (NK) cells. After 4 weeks, T lymphocytes became predominant, but NK cells persisted. Cultured peripheral blood lymphocytes obtained 12 weeks posttransplant were able to display significant cytotoxicity against leukemic blasts that had been cryopreserved at the time of relapse 5 months prior to bone marrow transplantation. To further characterize those cells with antileukemia activity, we used in vitro cloning techniques to identify four monoclonal populations, termed TC12, -48, - 50, and -59, with strong antitumor activity. Cytogenetic analysis demonstrated that each clone was of donor origin. Phenotypic characterization showed that the four clones expressed NKH1A but did not express T3, T4, or T8 antigens. Three of the four clones expressed T11/E rosette antigen. Each clone exhibited strong cytotoxicity against genetically unrelated hematopoietic tumor cell lines such as K562, Molt- 4, JM, and U937. In addition, we found that these patient clones were similar to cloned NK cells previously derived from normal individuals. Taken together, these results suggest that at least some clones with antileukemia activity following bone marrow transplantation are cells with NK-like function and phenotype. Functional analysis of these cytolytic cells in larger numbers of patients will be necessary to determine the clinical significance of this finding.

A rapid multisite radioimmunoassay for measurement of human alpha-fetoprotein (AFP) that uses two high-affinity monoclonal antibodies directed against distinct and separate determinants on the protein was developed and designated M-RIA. The sensitivity of the "simultaneous-sandwich" M-RIA is approximately equal to 0.5 ng/ml of serum after a 1-hr incubation period. Serum AFP levels have been measured in 1747 individuals with hepatocellular carcinoma (HCC), acute and chronic hepatitis B virus infection, chronic hepatitis B surface antigen (HBsAg)-carrier states, cirrhosis, other malignant tumors, and normal and disease controls to determine the specificity of the assay. Eighty percent (68/85) of patients with HBsAg-positive HCC had AFP levels of greater than 200 ng/ml (range, 260 to greater than 200,000 ng/ml). In contrast, all 450 normal subjects and 477 chronic HBsAg-positive carriers had levels of less than 20 ng/ml. More importantly, in acute and chronic hepatitis B, cirrhosis, and other malignant tumors and in the remaining disease controls, AFP levels were less than 20 ng/ml in 99.3% of the subjects, the great majority (greater than 96%) being less than 5 ng/ml. Indeed only two of 1635 individuals, one with acute hepatitis and the other with carcinoma of the esophagus had AFP levels of greater than 100 ng/ml. These observations are at variance with previous studies with conventional polyvalent RIAs of AFP levels of greater than 20 ng/ml in approximately equal to 40% of acute and chronic hepatitis and in 30% of cirrhosis. This striking specificity of the M-RIA is probably due in part to recognition of epitopes unique to AFP and suggest that such an assay may be used in the detection, early identification, and monitoring of AFP-producing tumors in high-risk populations.

In order to study antigenic site(s) present in the carboxyl-terminal part of the alpha-subunit of human choriogonadotropin (hCG-alpha), we attempted to produce site-specific antibodies directed against a 34-residue synthetic peptide analogous to region 59-92 of hCG-alpha. From a fusion experiment performed with a mouse injected with hCG-alpha-(59-92)-peptide conjugated to tetanus toxoid as immunogen, we selected a monoclonal antipeptide antibody (designated FA36) which has high binding activity for 125I-hCG-alpha but not for 125I-hCG in a radioimmunoassay. This antibody is of the IgG1 subclass and displays an affinity constant for 125I-hCG-alpha of 3.1 x 10(8) M-1. Hapten inhibition experiments performed by either radioimmunoassay or enzyme-linked immunosorbent assay with synthetic peptides spanning different portions of the region (59-92) demonstrated that the binding site of FA36 resides on (minimally) the six COOH-terminal amino acids of hCG-alpha, namely Cys-Tyr-Tyr-His-Lys-Ser, and that FA36 binds preferentially to peptides containing a carboxyl group on the COOH-terminal residue. Monoclonal immunoradiometric assays were established to determine the location of antigenic regions recognized by FA36, by antibody AHT20 (which binds only to hCG-alpha), and by antibody HT13 (which binds to both hCG and hCG-alpha). FA36 has the capacity to bind to hCG-alpha bound to either AHT20 or HT13, demonstrating that both AHT20 and HT13 antibodies are directed against antigenic regions distinct from the epitope of FA36. Monoclonal immunoradiometric assays were also carried out to study the binding of FA36 to hCG, the ovine and equine lutropin alpha-subunit, or hCG-alpha minus the 5 COOH-terminal residues (hCG-alpha core). Whereas significant binding of 125I-FA36 was observed with the ovine lutropin alpha-subunit, no binding was found with the equine lutropin alpha-subunit. As expected, FA36 did not bind to hCG-alpha core. Binding was also not detected with hCG, confirming that FA36 is specific for free hCG-alpha and that the COOH-terminal part of hCG-alpha is either weakly or (more likely) not at all accessible in the alpha/beta-dimer for antibody binding. Finally, immunoblots performed on hCG-alpha-(59-62)-peptide and various denatured alpha-subunits indicated that, with the exception of the equine lutropin alpha-subunit, FA36 detected various denatured alpha-subunits and particularly the alpha-subunit of carp gonadotropin-thyrotropin. This latter observation suggests a high degree of homology between the COOH-terminal regions of the alpha-subunits of fish gonadotropin and analogous mammalian hormones.(ABSTRACT TRUNCATED AT 400 WORDS)

We have conducted a phenotypic and functional analysis of 19 cloned cell lines generated after allogeneic stimulation of circulating lymphocytes from a normal human fetus aged 25 wk. Using a limited series of mAbs (Anti-T3, WT31, T4, T8, and NKH1A), cloned cells were found to fall in three groups. Three clones have a conventional "inducer" phenotype. Three clones have a phenotype (T3+, WT31+, T8+, and NKH1A+) similar to that of certain NK active mature T lymphocytes present in adult peripheral blood. In contrast, 13 cell lines display surface characteristics that have not been described previously. Indeed, they express T3 proteins but not the WT31 determinant. In light of previous studies, these results show that WT31 mAb is a unique reagent directed at an invariant epitope of the human T cell receptor that is not present on all circulating T3+ fetal lymphocytes. Functionally the T3+, WT31+, and NKH1A+ clones were found to kill immunizing LAZ 388 cells, as well as K562, while T3+, WT31- and NKH1A+ clones display NK-like function exclusively. Moreover, only WT31+ lymphocytes present in the cell line used for cloning experiments have the capacity to recognize alloantigen-bearing cells. Together, these data suggest that expression of WT31 may be necessary for recognition of alloantigens, while NK reactions mediated by T3+ lymphocytes are WT31-independent.

In an attempt to localize topographic domains specific to native human chorionic gonadotropin (hCG), we studied the discontinuous antigenic regions recognized by a monoclonal anti-hCG antibody designated as C8 which binds only to hCG and does not cross-react with either the free hCG-alpha and hCG-beta subunits or other glycoprotein hormones. Using two-site monoclonal immunoradiometric assays (M-IRMAs), we found that C8 antibody and an anti-peptide antibody (FB12) directed to residues 110-116 of hCG-beta did not bind simultaneously to hCG. This observation suggested that C8 binds to residues of hCG-beta included either in the antibody-binding region of FB12 or in close proximity to amino acids 110-116. To further delineate the regions of hCG-beta recognized by C8, we carried out hapten inhibition experiments with synthetic peptides corresponding to various regions of hCG-beta. The peptide corresponding to residues 109-122 and subpeptides (111-122 or 112-122) inhibited the binding of 125I-hCG to C8, whereas weak inhibition was observed with subpeptide 113-122. By studying the binding of C8 to the 1-112 disulfide-bonded part of hCG-beta (hCG-beta core) recombined with hCG-alpha, we were able to confirm that C8 binds to a region including or near to Asp112. M-IRMAs showed that C8 does not bind to the recombinant molecule lacking residues 113-145 of hCG. Taken together, these results indicate that a limited number of residues located on hCG-beta near to Asp112, and most likely the sequence Asp111-Asp112-Pro113, are included in the discontinuous antigenic region recognized by C8. We then attempted to localize residues of hCG-alpha that constitute another part of the determinant which bound to C8. Six synthetic peptides corresponding to various regions of hCG-alpha did not inhibit binding of 125I-hCG to C8. In contrast, M-IRMAs demonstrated that C8 is capable of binding recombinant products composed of the hCG-beta subunit and the alpha subunits from human, equine, and porcine species. These results indicate that C8 recognizes a region of the alpha subunit highly conserved in these three species. Finally, we determined that the discontinuous regions recognized by C8 are partially accessible on the CG/LH-receptor complex.

In the zinc metallopeptidases produced by the genus Bacillus, an active site histidine has been proposed to either stabilize the transition state in catalysis by donating a hydrogen bond to the hydrated peptide (Matthews, B. W. (1988) Acc. Chem. Res. 21, 333-340) or to polarize a water molecule, which subsequently attacks the peptidyl bond (Mock, W. L., and Aksamawati, M. (1994) Biochem. J. 302, 57-68). Site-directed mutagenesis techniques have been used to change this residue in the zinc endopeptidase from Bacillus stearothermophillus to either phenylalanine or alanine. At pH 7.0, the kcat/Km values of the substrate leucine enkephalin for the phenylalanine and alanine mutants were reduced by factors of 430- and 500-fold, respectively, as compared with the wild-type enzyme, mostly due to changes in kcat. In addition, the enzymatic activities of the mutant enzymes showed little pH dependence in the alkaline range, unlike the wild-type enzyme. The mutations did not greatly alter the binding affinities of inhibitors containing sulfydryl groups to chelate the active site zinc, while those of inhibitors containing hydroxamate or carboxylate zinc-chelating groups were increased between 80- and 250-fold. The largest change in the binding affinity of an inhibitor (> 5 orders of magnitude) was found with the proposed transition state mimic, phosphoramidon. The results are generally in agreement with x-ray crystallography studies and favor the involvement of the active site histidine in transition state binding.

We describe a first attempt to study the antibody-combining sites recognized by monoclonal antibodies raised against the beta-subunit of human choriogonadotropin (hCG). Two groups of antibodies were first defined by their ability to recognize only the free beta-subunit or the free and combined subunit. Antibodies FBT-11 and FBT-11-L bind only to hCG beta-subunit but not to hCG, whereas antibodies FBT-10 and D1E8 bind to both the beta-subunit and the hormone. In both cases, the antigenic determinants were localized to the core of the protein (residues 1-112), indicating the weak immunogenicity of the specific carboxyl-terminal extension of hCG-beta. Nine synthetic peptides spanning different regions of hCG-beta and lutropin-beta were assessed for their capacity to inhibit antibody binding. A synthetic peptide inclusive of the NH2-terminal region (residues 1-7) of the hCG beta-subunit was found to inhibit binding to the radiolabeled subunit of a monoclonal antibody specific for free hCG-beta (FBT-11). Further delineation of the antigenic site recognized by this antibody provided evidence for the involvement of fragment 82-92. Moreover, monoclonal antibody FBT-11 inhibited the recombination of hCG-beta to hCG-alpha, indicating that its antigenic determinant might be located nearby or in the hCG-beta portion interacting with the alpha-subunit. Binding of monoclonal antibody FBT-10, corresponding to the second antigenic determinant, was weakly inhibited by fragment 82-105 and did not impair the recombination of the hCG beta-subunit to the hCG alpha-subunit. Its combining site appeared to be located in a region of the intact native choriogonadotropin present at the surface of the hormone-receptor complex.