Concepts and approaches to in situ luminescence dating of Martian sediments.

In this paper we present the concept of a robotic instrument for in situ luminescence dating of near-surface sediments on Mars. The scientific objectives and advantages to be gained from the development of such an instrument are described, and the challenges presented by the Mars surface environment to the design and operation of the instrument are outlined.

Martian "microfossils" in lunar meteorites?

One of the five lines of evidence used by McKay et al. (1996) for relic life in the Martian meteorite Allan Hills (ALH) 84001 was the presence of objects thought to be microfossils. These ovoid and elongated forms are similar to structures found in terrestrial rocks and described as "nanobacteria" (Folk, 1993; McBride et al., 1994). Using the same procedures and apparatus as McKay et al. (1996), we have found structures on internal fracture surfaces of lunar meteorites that cannot be distinguished from the objects described on similar surfaces in ALH 84001. The lunar surface is currently a sterile environment and probably always has been. However, the lunar and Martian meteorites share a common terrestrial history, which includes many thousands of years of exposure to Antarctic weathering. Although we do not know the origin of these ovoid and elongated forms, we suggest that their presence on lunar meteorites indicates that the objects described by McKay et al. (1996) are not of Martian biological origin.

Rat class III Fc gamma receptor isoforms differ in IgG subclass-binding specificity and fail to associate productively with rat CD3 zeta.

Several new rat class III Fc gamma R isoforms are described here, extending the genetic complexity of this receptor family and further distinguishing rat CD16 from mouse CD16, represented by only one receptor isoform, and human CD16, represented by only two isoforms. RNase protection assays reveal that three rat tumor cell lines--RBL-1 basophilic leukemia cells, RM-SV1 macrophages, and CRNK-16 NK cells--all coordinately express multiple and probably identical rtFc gamma RIII-related transcripts in similar relative proportions but at significantly different levels. These results indicate that no single isoform predominates in these cell types but that the overall level of rtFc gamma RIII-related transcripts is differentially regulated. Two of the rtFc gamma RIII isoforms found to have extensive amino acid sequence differences in their second extracellular (EC2) domains are shown to bind rat and mouse IgG subclasses differently. This result suggests that the receptor isoform diversity in this species may function as a mechanism for extending the IgG-binding capacity of rat leukocytes. Cloned cDNA for the rat CD3 zeta protein was also isolated in this study and its ability to augment surface expression of class III Fc gamma R was tested by rosetting of cDNA-transfected COS cells. Like the structurally homologous mouse CD3 zeta, rat CD3 zeta fails to promote surface expression of Fc gamma RIII, sharply contrasting the efficient receptor expression produced by human CD3 zeta. Variations in the transmembrane amino acid sequences correlate with the divergent capacities of these CD3 zeta molecules to augment receptor expression. The high levels of CD3 zeta message expressed in rat NK cells may indicate that other unidentified hetero-subunits are required for assembly of rat CD3 zeta into functional CD16 receptors.

The breakup of a meteorite parent body and the delivery of meteorites to Earth.

Whether many of the 10,000 meteorites collected in the Antarctic are unlike those failing elsewhere is contentious. The Antarctic H chondrites, one of the major classes of stony meteorites, include a number of individuals with higher induced thermoluminescence peak temperatures than observed among non-Antarctic H chondrites. The proportion of such individuals decreases with the mean terrestrial age of the meteorites at the various ice fields. These H chondrites have cosmic-ray exposure ages of about 8 million years, experienced little cosmic-ray shielding, and suffered rapid postmetamorphic cooling. Breakup of the H chondrite parent body, 8 million years ago, may have produced two types of material with different size distributions and thermal histories. The smaller objects reached Earth more rapidly through more rapid orbital evolution.

Rat CD16 is defined by a family of class III Fc gamma receptors requiring co-expression of heteroprotein subunits.

Rat CD16 is defined here by a family of highly homologous class III Fc gamma receptor isoforms. Northern blot and polymerase chain reaction analysis indicate that multiple rtFc gamma RIII alpha isoforms are expressed by rat NK and macrophages contrasting the expression of a single class III receptor isoform by human and mouse NK and macrophages. Analysis of genomic Southern blots and splice variants identified by polymerase chain reaction suggests the existence of several homologous rat Fc gamma RIII alpha genes organized similar to human and mouse class III receptor genes. All rat Fc gamma RIII alpha isoform protein sequences have conventional transmembrane insertion sequences containing the unique LFAVDTGL motif conserved in all other class III Fc gamma and Fc epsilon RI alpha receptor sequences. A model is proposed for the protein-protein interactions between this sequence and the transmembrane sequences of two heteroprotein subunits, Fc epsilon RI gamma and CD3 zeta, known to interact with human and mouse class III receptors. Rat NK, monocytes, and neutrophils were all found to express transcripts for both of these subunits, whereas rat macrophages express only Fc epsilon RI gamma subunit transcripts. Furthermore, the Fc epsilon RI gamma subunit was found to promote the cell surface expression of rtFc gamma RIII alpha isoforms in transfected COS cells.

Rat and human natural killers exhibit contrasting immunoglobulin G subclass specificities in antibody-dependent cellular cytotoxicity reflecting differences in their Fc receptors (Fc gamma R).

Rat and human natural killers (rtNK and huNK, respectively) were compared in quantitative antibody-dependent cellular cytotoxicity (ADCC) assays for their capacity to recognize mouse and rat IgG monoclonal antibodies (MAb) of different subclasses. NK from these two species exhibit considerably different patterns of IgG subclass recognition as determined by the relative antibody concentrations required for comparable levels of target cells lysis. ADCC assays with a panel of 16 MAb revealed that the efficiency of rtNK-mediated target lysis diminished according to IgG subclass in the following order: molgG1 greater than rtlgG2a greater than molgG2b approximately molgG2a greater than rtlgG2b greater than molgG3. By comparison, huNK recognized the same antibodies with nearly the opposite order of efficiency: rtlgG2b much greater than molgG2a greater than molgG3 greater than molgG2b much greater than rtlgG2a approximately molgG1. Only molgG2a antibodies were equally potent with rtNK and huNK. The contrasting difference in IgG subclass recognition by rat and human NK reflects the comparatively low protein sequence homology between their respective IgG Fc receptors (Fc gamma R).

Characterization and expression of an Fc gamma receptor cDNA cloned from rat natural killer cells.

A cDNA clone for an IgG-binding Fc receptor, rtFc gamma R alpha, of the rat natural killer cell line CRNK-16 is characterized here. This clone encodes an Fc gamma receptor as shown by the ability of cDNA-transfected COS cells to rosette IgG-coated sheep erythrocytes. The rtFc gamma R alpha is exceptionally homologous to the mouse moFc gamma R alpha, with 77% protein sequence identity and 71% nucleic acid identity overall. The transmembrane region of the rtFc gamma R alpha contains the sequence Leu-Phe-Ala-Val-Asp-Thr-Gly-Leu, which is present in the membrane sequences of four other Fc receptors including mouse Fc gamma R alpha, human Fc gamma RIII-2, and the Fc epsilon R alpha subunits of the rat and human high-affinity IgE-binding receptors. Also, the rtFc gamma R alpha cytoplasmic domain exhibits specific homology to other receptors derived from natural killer cells, human Fc gamma RIII-2 and mouse Fc gamma R alpha. However, the rtFc gamma R alpha cDNA clone is complementary to at least two different-sized mRNAs expressed by CRNK-16 cells, contrasting the single Fc gamma R-related mRNA species expressed by human and mouse natural killer cells. These rat mRNAs are homologous to both the 5' and the 3' end of the cDNA clone, suggesting that they may be (i) splice variants of one transcript or (ii) products of different but highly related genes.

H-2 restriction and serotype crossreactivity of anti-reovirus cytotoxic T lymphocytes (CTL).

Murine anti-reovirus cytotoxic T lymphocytes (CTLs) were analyzed for H-2 restricted recognition of virus infected target cells and for potential cross-reactivity with cells infected by reovirus serotype 1 (T1; Lang strain) or by serotype 3 (T3; Dearing strain). Anti-reovirus CTL specifically lysed virus infected cells and lysis was shown to be H-2 restricted by the H-2Dd, H-2Ld, H-2Kd, H-2Kb, and H-2Kk antigens. No H-2 antigens were identified which failed to restrict virus recognition by anti-reovirus CTL. Anti-T1 and anti-T3 CTLs were also shown to crossreact completely with cells infected with the opposite virus serotype. Thus, anti-reovirus CTLs are restricted by a broad spectrum of H-2 antigens and they detect common rather than unique structural components of these two viral serotypes.

Molecular cloning and expression of the mouse high affinity Fc receptor for IgG.

Full length cDNA clones encoding the mouse Fc gamma RI were isolated by using redundant oligonucleotide probes based on previously determined amino acid sequence of protein bound to an IgG2a antibody column. Sequence analysis of cDNA clones indicates that mouse Fc gamma RI is a transmembrane glycoprotein that is composed of three disulfide bonded extracellular Ig binding domains unlike Fc gamma RII of man and mouse. These extracellular domains contain five potential sites of N-linked glycosylation; three sites in the first domain and one in each of the second and third domains. In addition a transmembrane region is present followed by a cytoplasmic tail of 84 amino acids. Analysis of the amino acid sequence of the first two extracellular domains of Fc gamma RI indicate that these are highly homologous to the extracellular domains of Fc gamma RII; the third domain is different and shows a lower level of homology to other FcR domains but is clearly related to the Ig super-family. Transfected cells expressing Fc gamma RI were shown to bind immune complexes of rabbit IgG; and monomeric IgG2a bound to transiently transfected cells with an affinity of approximately 5 x 10(7) M-1, i.e. the receptor was of high affinity and therefore was by definition Fc gamma RI. Northern analysis demonstrated that Fc gamma RI mRNA could be detected in the Fc gamma RI+ myeloid cell lines WEH1 3B and J774. Finally, Southern analysis indicated that Fc gamma RI is likely to be encoded by a single copy gene of approximately 9 kb.

Mouse macrophage beta subunit (CD11b) cDNA for the CR3 complement receptor/Mac-1 antigen.

The cDNA for the common beta Mac-1 subunit (CD11b) of the mouse LFA-1/Mac-1/p150,95 group of leukocyte cell adhesion receptors, formally designated integrin beta 2, has been cloned and sequenced. Clones were isolated from cDNA libraries made from J774 macrophage and WEHI-3B myelomonocytic tumor cells which express this subunit as a component of the macrophage activation antigen 1 (Mac-1), also known as complement receptor type 3 (CR3). This subunit is expressed as a single, abundant mRNA species approximately 2.7 kilobase (kb) in size. The 2422 base pair (bp) cDNA sequence obtained codes for a 771 amino acid protein organized with leader, extracellular, transmembrane, and cytoplamic domains of 23, 680, 23, and 46 amino acids, respectively, yielding an 82,700 mature protein of 747 amino acids. The mouse beta Mac-1 subunit is highly similar to its human counterpart with an overall sequence identity of 81% and identical positioning of 5 out of 6 potential N-linked glycosylation sites, as well as 56 Cys residues that are organized in repeating motifs characteristic of integrin beta subunits. The most highly conserved regions are the transmembrane and cytoplasmic domains where only 4 out of 69 amino acids differ, indicating that the functions associated with this domain in Mac-1-mediated processes, such as iC3b-triggered phagocytosis, have been evolutionarily conserved.

Differential expression of the ASGM1 antigen on anti-reovirus and alloreactive cytotoxic T lymphocytes (CTL).

Anti-reovirus cytotoxic effectors were found to be: (i) H-2 restricted; (ii) virus specific; (iii) non-lytic (in 4 h) for natural killer (NK)-sensitive YAC-1 cells; and (iv) positive for the Thy-1 and Lyt-2 lymphocyte markers. Thus, anti-reovirus cytotoxic effectors have the functional and phenotype characteristics of cytotoxic T lymphocyte (CTL). A significant fraction of anti-viral CTL, as well as alloreactive CTL, were also found to be positive for the asialo GM1 (ASGM1) cell surface antigen, generally considered to be a NK cell marker. ASGM1 expression on these CTL, as determined by sensitivity to antibody plus complement (C), appeared to be highly variable and dependent on two factors-the nature of the antigenic stimulus (viral vs. alloantigen), and the mouse strain from which the CTL originated. Thus, ASGM1 antigen expression on CTL appears to be regulated and may be under the control of lymphokines, development differentiation signals and/or other strain-dependent genetic factors.

Apparent sensitivity of human K lymphocytes to the spatial orientation and organization of target cell-bound antibodies as measured by the efficiency of antibody-dependent cellular cytotoxicity (ADCC).

Although monoclonal antibodies (mAb) can elicit potent ADCC by human K lymphocytes, different mAb, even of the same antibody subclass or even of the same target antigen specificity, vary considerably as to their efficiency in eliciting ADCC. The extensive variability in ADCC efficiencies of murine IgG2a mAb is analyzed here. In cold-target inhibition experiments it was found that only cells coated with "ADCC-efficient" IgG2a mAb, and not "ADCC-inefficient" IgG2a mAb, inhibit K effector cell lysis of radiolabeled target cells by ADCC. This result indicates that the spatial orientation of the antibodies on the target cell membrane influences the net efficiency of ADCC reactions by affecting the efficiency of interaction between antibody and the Fc receptors (FcR) of K cells. It is proposed that a "favorable" orientation of antibodies on the target cell membrane is required for efficient ADCC reactions. This proposal is directly supported by the observation that one IgG2a mAb (20.8.4), which cross-reacts with several different H-2 alloantigens, was found to elicit efficient ADCC only when bound to certain of its possible target cell antigens. It was also observed in these studies that the organization of antibodies on a target cell membrane influences the net efficiency of ADCC reactions. It is proposed that a "favorable" antibody organization on the target cell membrane is also required for efficient ADCC reactions. This proposal is supported by the observation that certain antihuman beta 2m (anti-Hu beta 2m) IgG2a mAb, which elicit efficient ADCC lysis of human target cells, fail to elicit the lysis of murine cells having Hu beta 2m molecules coupled randomly to their external membrane surfaces. The differences in the way the Hu beta 2m was organized on the surfaces of the human cells and the murine-Hu beta 2m cell conjugates presumably caused differences in the way the bound antibodies were organized on the cell surfaces, which in turn resulted in the ADCC efficiency differences observed for the same mAb with the different target cell types. Because ADCC reactions appear to be sensitive to both the orientation and the organization of cell surface-bound antibodies, certain types of structural alterations or variations in the membrane molecules (relative to other neighboring structures on the target cell membrane) are potentially detectable by quantitative differences or variations in ADCC reactions.

Lack of lymphocyte-induced DNA fragmentation in human targets during lysis represents a species-specific difference between human and murine cells.

Significant cytotoxic lymphocyte-induced DNA fragmentation does not occur in four human target cells lysed by human natural killer lymphocytes, killer lymphocytes, or cytotoxic T lymphocytes. These results contrast with the extensive DNA fragmentation that occurs in murine target cells lysed by the analogous effector lymphocytes. Thus, DNA fragmentation, or the lack thereof, represents a species-specific difference between human and murine cells responding to lysis by cytotoxic lymphocytes. The DNase activity observed in murine cells is probably internally activated rather than delivered by the cytotoxic effector cells, since human killer lymphocytes selectively caused DNA fragmentation in murine but not in human target cells lysed by antibody-dependent cellular cytotoxicity.

Polymorphism of murine major histocompatibility Class I antigen: assignment of putative allodeterminants to distinct positions of the amino acid sequence within the first external domain of the antigen.

Forty-five new monoclonal antibodies reacting with the mouse H-2Dd antigen have been established. The specificities of 34 of these antibodies were mapped into the first external domain (N) of the Dd antigen by testing reactivities with the products of mosaic H-2 genes in which the coding sequences of the first and/or the second external domains of the H-2Dd genes were recombined in vitro with the remaining portion of the H-2Ld gene. These antibodies reacted with at least 13 distinct allodeterminants located in the N domain, composed of 91 amino acids, as judged from panel tests carried out on various H-2 haplotypes. To assign possible positions of antigenic determinants of these and other anti-H-2Dd antibodies, we compared primary sequences of seven H-2 antigens and searched for correspondence between the pattern of amino acid substitutions in the N domain, allowing 15 positions to be assigned for the antigenic sites. These putative antigenic determinants were assessed for possible relationships with several parameters of protein secondary structure postulated according to predictive methods. Many of these sites appear to be associated with greatest local hydrophilicity, known to correlate with sites of antibody binding in various proteins. We therefore propose that some of the correspondences found in this work represent structural correlates of allodeterminants.

Evidence of only one H-2D/L-related glycoprotein in H-2dm1 mutant cells.

Analyses of the H-2D/L-related glycoproteins from dm1 mutant cell extracts by sequential immunoprecipitation, by SDS gel electrophoresis and by tryptic peptide mapping indicate that dm1 cells express only a single glycoprotein with H-2D/L-related determinants. In contrast to the four H-2D/L-related antigens identified for the parental d haplotype viz. H-2Dd, H-2Md, H-2Ld and H-2Rd, separate and distinguishable "H-2Ddm1", "H-2Mdm1", "H-2Ldm1" and "H-2Rdm1" glycoprotein counterparts are apparently lacking in the dm1 mutant haplotype. Only a single H-2D/L-related glycoprotein is identified in dm1 extracts by standard serological methods and this glycoprotein is designated H-2D/Ldm1 because of its H-2Dd/H-2Ld hybrid characteristics, as recently shown by Burnside and colleagues (1984). Thus, the seemingly complex phenotype of the dm1 mutant appears to originate primarily from one molecule having properties of two (or more) molecules of the parental haplotype.

Rapid covalent coupling of proteins to cell surfaces: immunological characterization of viable protein-cell conjugates.

A rapid and efficient 2-step procedure is described for covalently attaching proteins to cell surfaces by using a heterobifunctional cross-linking agent, succinimidyl-4-(N-maleimidomethylcyclohexane)-1-carboxylate (SMCC). In the first step, protein is derivatized for about 30 min with a 1:1 (mole:mole) stoichiometric ratio of SMCC which creates 4-(N-maleimidomethylcyclohexane)-1-amidyl-protein (MCA-protein) covalent linkages with the primary amino groups of proteins. In the second step, cell-MCA-protein conjugates are rapidly prepared (in less than 30 min) by reacting MCA-protein with 'reduced' cells which have been pre-incubated (for about 1 h) with dithiothreitol (DTT). Stable protein-cell conjugates result from the covalent thioether linkages formed between the maleimido groups of the derivatized protein and the cell surface thiol groups created by DTT reduction. Protein-cell conjugates have been made in this way with several different proteins using several different types of cells. Such conjugates are characterized in this study by complement plus antibody-mediated cytotoxicity (CAMC) and by antibody-dependent cellular cytotoxicity (ADCC) with human effector lymphocytes. Because these protein-cell conjugates are shown to remain viable and to retain significant levels of coupled protein for at least 24 h in culture, they are potentially useful for probing various membrane-related phenomena such as the recognition of cell surface antigens by immune effector cells.

Unusually efficient tumor cell lysis by human effectors of antibody-dependent cellular cytotoxicity mediated by monoclonal antibodies.

Concentrations ranging between 0.01 and 10 pg per cell of certain monoclonal antibodies (MAbs) are shown to constitute 4-hr 50% lethal doses for tumor cells mixed with human effectors of antibody-dependent cellular cytotoxicity (ADCC). This efficient and rapid tumor cell lysis is achieved at low effector cell levels (effector:target ratios, less than or equal to 25:1) at which the effectors are nonadherent peripheral blood leukocytes (PBL) enriched by density centrifugation. Comparable MAb-mediated ADCC efficiency has not been reported previously, probably because most MAbs (e.g., 10 of 13 tested in this study) are typically inefficient or completely inactive in mediating ADCC, even at 100-fold greater concentrations. By analyzing the ADCC efficiencies of several MAbs specific for murine cell surface alloantigens, it is shown that murine IgG2a and IgG3 MAbs and a rat IgG2b MAb are very efficient mediators of ADCC. However, ADCC efficiency was found not to correlate strictly with subclass, since 4 of 6 murine IgG2a MAbs tested were completely inactive, even though they all bound the target cells readily. It is shown that the relative differences in ADCC efficiencies are not accounted for directly by antibody affinity for antigen; one MAb was very efficient in ADCC but had demonstrably low antigen affinity, while a second MAb showed no ADCC activity in spite of its high affinity for the same target antigen. These results point to other experimentally testable properties of MAbs and of MAb-antigen complexes which may be critical for efficient ADCC reactions. This study underscores an important immunotherapeutic value which certain MAbs potentially have for mediating tumor cell lysis: in low concentrations (and without toxic drug modification), some MAbs efficiently mediate the lysis of tumors by ADCC, which itself is as effective as other immune lytic processes but which requires no prior immunological education of effector cells.

A molecular hybrid of the H-2Dd and H-2Ld genes expressed in the dm1 mutant.

Sequential immunoprecipitates show that H-2dm1 mutant cells express a hybrid "H-2D/L" antigen exhibiting determinants normally associated with two different gene products of the parental d haplotype-i.e., the H-2Dd and H-2Ld antigens. The hybrid H-2D/Ldm1 antigen appears to consist of a portion of the NH2-terminal extracellular half of the H-2Dd antigen "fused" to a portion of the COOH-terminal extracellular half of the H-2Ld antigen. This structure is inferred from the reactivity of dm1 antigens with cytotoxic T lymphocytes specific for H-2Ld determinants and with monoclonal antibodies specific for determinants in the structural domains of H-2Ld or H-2Dd. The H-2D/Ldm1 molecule apparently retains all of the third external domain (C2 or alpha 3) and part of the second external domain (C1 or alpha 2) of H-2Ld, but its first external domain (N or alpha 1) derives from H-2Dd. From these findings and from previous peptide mapping studies, we propose that the H-2D/Ldm1 antigen is the product of a hybrid gene that has resulted from an unequal crossover between the parental H-2Dd and H-2Ld genes, leaving the N exon and part of the C1 exon of the H-2Dd gene joined to the H-2Ld gene beginning somewhere within its C1 exon.