Search

Analysis of proliferation and cytokine production at the single-cell level indicated that proliferation of pre-existing type 2 cytokine(+) human peripheral naive T cells (CD4(+) and CD8(+)) accounts for the accumulation of type 2 T cells in lymphocytes cultured with IL-2, with or without IL-4, and independently from TCR-mediated stimulation. This is because: firstly,the number of cells progenitor to the type 2 cytokine(+) T cells accumulated in culture is lower than that of the original cytokine(+) cells; secondly, percentages and numbers of the accumulated type 2 cytokine(+) T cells depend on those in the original lymphocyte population; thirdly, no accumulation occurs in cultures of lymphocytes experimentally depleted of type 2 cells; and fourthly, naive T cells do not require proliferation before producing type 2 cytokines. In contrast, accumulation of IFN-gamma(+) T cells in cultures with IL-12 can not be explained with induced proliferation of pre-existing IFN-gamma(+) cells, but depends on differentiation from more-immature cells. These novel insights into the regulation of type 2 and type 1 cytokine(+) T cells provide a new understanding of the cellular bases for the regulation of immune responses and for manipulating the immune system in clinical settings.

Detection of cytokines and identification of the producer cells are essential to define the interplay and the role of distinct leukocyte subsets in the development of immune and inflammatory responses. Several methods used to study cytokine expression are based on detection of the encoding mRNA (Northern blot, RNase protection assay, RT-PCR) or of protein in the supernatant from stimulated cells (ELISA, RIA, ELISPOT). These are simple and useful, but have limitations related to the need of using purified cell populations to precisely define the effector cells, and exception made for RT-PCR and ELISPOT assays, the requirement for relatively large numbers of cells for sufficient resolution. Here we present a method based on immunofluorescence (flow cytofluorimetry) to detect intracellular accumulation of cytokines in mixed leukocyte populations. It has the distinct advantage of: The detailed description/discussion of the method uses natural killer (NK) cells as an example, but this method can be applied to the study of other cell types, and is of special interest/use when analyzing subsets present in very low proportion.

A maturation-dependent change in phenotype and cytokine production from relatively immature CD161(-) or CD161(+), IL-13(+)IL-4(+), IFN-gamma(-), to mature CD161(+)CD56(+) IFN-gamma(+) cells occurs in primary human alpha-galactosyl ceramide-reactive CD1d-restricted natural killer T (NKT) cells under the control of IL-12 and other monokines. Modulation of this process upon alpha-galactosyl ceramide stimulation explains the opposite roles of NKT cells to drive type 1 and type 2 immune responses. Because the same developmental changes occurred and were similarly regulated in T cells, the data establish that NKT cells should no longer be considered a functionally unique regulatory subset. However, the results of their analysis can be taken as a model for immunotherapeutic approaches with T cells for which a nominal or surrogate antigen is defined.

Immature myeloid and NK cells exist, and undergo cytokine-induced differentiation, in the periphery. In this study, we show that also immature CD2−/low T cells exist in peripheral blood. These cells produce the type 2 cytokines IL-13, IL-4, and IL-5, but not IFN-γ or IL-10, and, upon culture with IL-12- and TCR-mediated stimuli, differentiate to IL-13+IFN-γ+ cells producing high IL-2 levels, and finally IL-13−IFN-γ+ cells. The monokine combination IL-12, IL-18, and IFN-α substitutes for TCR-mediated stimulation to induce the same differentiation process in both immature CD2−/low and primary mature CD2+ IL-13+ Τ cells. IFN-α is needed to maintain high level IL-2 production, which is confined to type 2 cytokine-producing cells and lost in the IFN-γ+ ones. Upon TCR-mediated stimulation, IFN-γ+ cells are then induced to produce IL-10 as they undergo apoptosis. These data indicate that peripheral type 2 cytokine+ T cells are immature cells that can differentiate to effector IFN-γ+ cells following a linear monokine-regulated pathway identical with that previously described for NK cells. They define the cellular bases to support that cell-mediated immune responses are regulated not only via Ag-induced activation of mature effector cells, but also via bystander monokine-induced maturation of immature T cells.

The hypothesis that distinct subsets of NK cells produce type 2 and type 1 cytokines in resting naive lymphocytes was tested analyzing cytokine production at the single-cell level. Two non-overlapping IL-13+ and IFN-gamma+ subsets were identified in adult and neonatal NK cells. IL-2 maintained their relative proportion. Accumulation of the former was induced by IL-4, but not IL-13, and inhibited by IL-12; that of the latter was induced by IL-12 and inhibited by IL-4 and IL-13. IL-4 induced preferential proliferation of the pre-existing peripheral IL-13+ cells, whereas IL-12 had minimal effect on proliferation of the IFN-gamma+ NK cells. The IL-13+ cells (CD161+ only) are phenotypically distinct from the IFN-gamma+ ones (CD56+) before and after culture under any condition analyzed, and produce IL-13 in response to NK-sensitive target cells and PMA+Ca(2+) ionophore, whereas also FcgammaRIIIA and IL-2+IL-12 stimulate IFN-gamma production. These data define the existence and regulation of two distinct resting peripheral NK cell subsets producing type 1 and type 2 cytokines, and suggest possible roles for IL-13+ NK cells in allergy.

α-Galactosylceramide (αGalCer) stimulates NKT cells and has antitumor activity in mice. Murine NKT cells may directly kill tumor cells and induce NK cell cytotoxicity, but the mechanisms are not well defined. Newly developed human CD1d/αGalCer tetrameric complexes were used to obtain highly purified human αGalCer-reactive NKT cell lines (>99%), and the mechanisms of NKT cell cytotoxicity and activation of NK cells were investigated. Human NKT cells were cytotoxic against CD1d− neuroblastoma cells only when they were rendered CD1d+ by transfection and pulsed with αGalCer. Four other CD1d− tumor cell lines of diverse origin were resistant to NKT cells, whereas Jurkat and U937 leukemia cell lines, which are constitutively CD1d+, were killed. Killing of the latter was greatly augmented in the presence of αGalCer. Upon human CD1d/αGalCer recognition, NKT cells induced potent cytotoxicity of NK cells against CD1d− neuroblastoma cell lines that were not killed directly by NKT cells. NK cell activation depended upon NKT cell production of IL-2, and was enhanced by secretion of IFN-γ. These data demonstrate that cytotoxicity of human NKT cells can be CD1d and ligand dependent, and that TCR-stimulated NKT cells produce IL-2 that is required to induce NK cell cytotoxicity. Thus, NKT cells can mediate potent antitumor activity both directly by targeting CD1d and indirectly by activating NK cells.

Natural killer (NK) and T cells appear to have a common progenitor cell. Interleukin 15 (IL-15) plays an important role in directing such progenitors to commit to NK-cell differentiation. Several other cytokines such as Flt-3 ligand and IL-7 appear to modulate or enhance the effects of IL-15.

Cytokine-mediated enhancement of spontaneous cytotoxicity depends, at least in part, on modulation of the expression of surface molecules responsible for recognition of target cell structures and triggering or inhibition of the cytotoxic machinery. We previously demonstrated that expression of transcription factors (e.g., Egr-1, JunB, and c-Fos) is differentially regulated by IL-2 and IL-12. Here we show that expression of CD161/NKR-P1A, a molecule involved in triggering cytotoxicity, is specifically up-regulated by IL-12. CD161 transcription, mRNA accumulation, and surface expression are increased by IL-12. Other cytokines sharing the IL-2R β- and/or common γ-chains (i.e., IL-15, IL-4, and IL-7) do not mediate these effects. In an effort to analyze the mechanisms by which IL-2, IL-12, and IL-15 differentially regulate gene transcription, we have isolated a novel gene, 197/15a, the expression of which in NK and T cells is down-regulated by IL-2 and IL-15, up-regulated by IL-12, and not affected by IL-4 and IL-7. IL-2 and IL-15 act, at least in part, repressing 197/15a transcription; their effect on 197/15a mRNA accumulation is partially independent of novel protein synthesis, likely not mediated by JunB, Bcl-2, or Bax, and requires the activity of rapamycin-sensitive molecule(s). The observation that IL-2 and IL-12 differentially modulate CD161 expression suggests the existence of cytokine-specific mechanisms of modulation of spontaneous cytotoxicity based on the regulation of expression of surface molecules involved in target cell recognition and/or triggering of the cytolytic machinery.

G-protein-coupled receptor kinases (GRKs) mediate agonist-specific phosphorylation and desensitization of G-protein-coupled receptors. Previous studies have shown that several of these kinases are expressed in hematopoietic cells and that GRK expression is modulated during T-lymphocyte activation. Here, we analyzed the regulation of beta-adrenergic receptor kinase (betaARK) and GRK6 expression and activity in myelomonocytic and lymphoid cells. In the promyelocytic cell line HL- 60, GRK6 protein levels and activity rose twofold to fourfold over the course of treatment with agents that induce differentiation toward either the myeloid (dimethyl sulfoxide and retinoic acid) or monocytic [1alpha,25(OH)2-vitamin D3] lineage, whereas betaARK protein levels and activity were only slightly altered. In contrast, treatment with phorbol 12,13-myristic acetate (PMA) led to a reduction in steady-state levels and activity of both betaARK and GRK6. Treatment of human lymphocytes with several polyclonal activators (phytohemagglutinin, anti-CD3 antibody and interleukin-2) resulted in enhanced betaARK and GRK6 mRNA and protein levels and increased activity of both kinases. In contrast, PMA and calcium ionophore treatment significantly elevated GRK6 protein levels, while decreasing betaARK expression. These data demonstrate that betaARK and GRK6 expression are differentially regulated during myelomonocytic cell development and lymphocyte activation.

Human natural killer (NK) cell differentiation from immature lineage negative (Lin-) umbilical cord blood cells was examined in vitro. Cells expressing differentiation antigens of mature NK cells (CD56, CD16, CD2, CD8, NKR-P1A) were generated from Lin- cells cultured with interleukin (IL)-2 and a murine bone marrow stromal cell line expressing the human membrane-bound form of stem cell factor. Two subsets of NK cells were identified in these cultures: one expressed both NKR-P1A and CD56 and, in variable proportions, all other NK cell differentiation antigens; the second subset expressed only NKR-P1A and, unlike the former, was not cytotoxic. Neither subset expressed interferon (IFN)-gamma mRNA even after stimulation with phorbol di-ester and Ca2+ ionophore, but both expressed tumor necrosis factor alpha mRNA and the cytotoxic granule-associated proteins TIA-1, perforin, and serine esterase-1. After 10-d culture with IL-2, IL-12, and irradiated B lymphoblastoid cells, approximately 45% of the NKR-P1A+/ CD56- cells became CD56+, and the same cultures contained cells capable of cytotoxicity and of IFN-gamma production. These results indicate that NKR-P1A expression in the absence of other NK cell markers defines an intermediate, functionally immature stage of NK cell differentiation, and that effector functions develop in these cells, concomitantly with CD56 expression, in the presence of IL-12. These cells likely represent the counterpart of a CD3-/NKR-P1A+/ CD56-/CD16- cell subset that, as shown here, is present both in adult and neonatal circulating lymphocytes.

Cross-linking the receptors for the Fc domain of IgG (Fc gamma R) on leukocytes induces activation of protein tyrosine kinases. The intermediary molecules that transduce to the nucleus the signals leading to induction of the diverse biological responses mediated by these receptors are not clearly identified. We have investigated whether mitogen-activated protein kinases (MAPK) are involved in transmembrane signaling via the three Fc gamma R present on monocytic, polymorphonuclear, and natural killer (NK) cells. Our results indicate that occupancy of Fc gamma RI and Fc gamma RII on the monocytic cell line THP-I and on polymorphonuclear leukocytes (PMN) induces, transiently and with fast kinetics, MAPK phosphorylation, as indicated by decreased electrophoretic mobility in sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and increased amounts of the proteins in antiphosphotyrosine antibody immunoprecipitates. This, associated with increased enzymatic activity, also occurs upon stimulation of the transmembrane isoform of CD16 (Fc gamma RIIIA) in NK cells and in a T cell line expressing transfected Fc gamma RIIIA alpha ligand-binding chain in association with zeta, but not upon stimulation of the glycosil-phosphatidylinositol-anchored Fc gamma RIIIB on PMN. Using the specific MAP kinase kinase inhibitor-PD 098059, we show that activation of MAPK is necessary for the Fc gamma R-dependent induction of c-fos and tumor necrosis factor alpha mRNA expression in monocytes and NK cells. These results underscore the role of MAPK as signal-transducing molecules controlling the expression of different genes relevant to leukocyte biology upon Fc gamma R stimulation.

The BCR/ABL oncogenic tyrosine kinase is responsible for initiating and maintaining the leukemic phenotype of Philadelphia chromosome (Ph1)- positive cells. Phosphatidylinositol-3 (PI-3) kinase is known to interact with and be activated by receptor and nonreceptor tyrosine kinases. We investigated whether PI-3 kinase associates with and/or is regulated by BCR/ABL, whether this interaction is functionally significant for Ph1 cell proliferation, and, if so, whether inhibition of PI-3 kinase activity can be exploited to eliminate Ph1-positive cells from bone marrow. We show that the p85 alpha subunit of PI-3 kinase associates with BCR/ABL and that transient expression of BCR/ABL in fibroblasts and down-regulation of BCR/ABL expression using antisense oligodeoxynucleotides (ODNs) in Ph1 cells activates and inhibits, respectively, PI-3 kinase enzymatic activity. The use of specific ODNs or antisense constructs to downregulate p85 alpha expression showed a requirement for p85 alpha subunit in the proliferation of BCR/ABL-dependent cell lines and chronic myelogenous leukemia (CML) primary cells. Similarly, wortmannin, a specific inhibitor of the enzymatic activity of the p110 subunit of PI-3 kinase, inhibited growth of these cells. The growth of normal bone marrow and erythromyeloid, but not megakaryocyte, progenitors was inhibited by p85 alpha antisense [S]ODNs, but wortmannin, at the concentrations tested, did not affect normal hematopoiesis. The proliferation of two BCR/ABL- and growth factor-independent cell lines was not affected by downregulation of the expression of the p85 alpha subunit or inhibition of p110 enzymatic activity, confirming the specificity of the observed effects on Ph1 cells. Thus, PI-3 kinase is one of the downstream effectors of BCR/ABL tyrosine kinase in CML cells. Moreover, reverse transcriptase-polymerase chain reaction performed on single colonies to detect BCR-ABL transcripts showed that wortmannin was able to eliminate selectively CML-blast crisis cells from a mixture of normal bone marrow and Ph1 cells.

PROBLEM: Giemsa stained cell isolates prepared from chorionic villus samples (CVS) contain granulated cells morphologically similar to large granular lymphocytes. METHOD: Phenotypic characterization of these cellular isolates by two-color immunofluorescence and subsequent in vitro culture in the presence of recombinant interleukin-2 (rIL-2) were done in order to determine whether CVS could serve as a source of decidual lymphocytes. RESULTS: A major fraction of the CVS-derived lymphocytes were characterized as decidual NK cells, exhibiting high levels of CD56 expression (CD56+bright), without concomitant expression of CD16. The T cell population present in CVS-derived lymphocytes contained both CD4+ and CD8+ cells in a ratio somewhat reduced compared to that found in peripheral blood. While both T cells and CD56+bright cells from CVS proliferate in vitro in response to rIL-2 alone, preferential growth of CD56+bnghtcells was accomplished using a selective culture technique wherein co-culture with an irradiated, B lymphoblastoid cell line promoted the growth of CD56+ cells. CONCLUSION: CVS contains decidual NK cells and T cells that proliferate in response to rIL-2 and/or third party stimulator cells. These culture techniques will allow investigations into the maturation and/or activation of decidual NK cells and T cells.

The p210bcr/abl tyrosine kinase appears to be responsible for initiating and maintaining the leukemic phenotype in chronic myelogenous leukemia (CML) patients. p21ras-p120GAP interactions play a central role in transducing mitogenic signals. Therefore, we investigated whether p21ras and p120GAP are regulated by p210bcr/abl, and whether this activation is functionally significant for CML cell proliferation. We report that transient expression of p210bcr/abl in fibroblast-like cells induces simultaneous activation of p21ras and inhibition of GTPase-promoting activity of p120GAP, and confirm these data showing that downregulation of p210bcr/abl expression in CML cells with bcr/abl antisense oligodeoxynucleotides induces both inhibition of p21ras activation and stimulation of GTPase-promoting activity of p120GAP. Tyrosine phosphorylation of two p120GAP-associated proteins, p190 and p62, which may affect p120GAP activity, also depends on p210bcr/abl tyrosine kinase expression. Direct dependence of these effects on the kinase activity is proven in experiments in which expression of c-MYB protein in fibroblast-like cells or downregulation of c-MYB expression resulting in analogous inhibition of CML cell proliferation does not result in the same changes. Use of specific antisense oligodeoxynucleotides to downregulate p21ras expression revealed a requirement for functional p21ras in the proliferation of Philadelphia chromosome-positive CML primary cells. Thus, the p210bcr/abl-dependent regulation of p120GAP activity is responsible, in part, for the maintenance of p21ras in the active GTP-bound form, a crucial requirement for CML cell proliferation.

Induced activation of protein tyrosine kinase(s) is a central event in signal transduction mediated via the low affinity receptor for IgG (Fc gamma RIIIA, CD16) in natural killer (NK) cells. Tyrosine phosphorylation may affect the function of several protein directly, or indirectly by inducing their association with other tyrosine phosphorylated proteins. Here, we report that Fc gamma RIII stimulation induces activation of phosphatidylinositol (PI)-3 kinase in NK cells. Phosphotyrosine immunoprecipitates from Fc gamma RIII-stimulated NK cells contain PI-kinase activity and PI-3 kinase can be directly precipitated from them. Conversely, a series of tyrosine-phosphorylated proteins is coprecipitated with PI-3 kinase from the stimulated, but not from control cells. Analogous results obtained using Jurkat T cells expressing transfected Fc gamma RIIIA alpha ligand binding chain in association with gamma 2 or zeta 2 homodimers indicate that both complexes transduce this effect, although the Fc gamma RIIIA-zeta 2 complexes do so with greater efficiency. Accumulation of phosphoinositide D3 phosphorylated products in stimulated cells confirms PI-3 kinase activation, indicating the participation of this enzyme in Fc gamma RIIIA-mediated signal transduction.

There is evidence to suggest that the p120 GAP (GAP), originally described as an inhibitor of p21ras, may also serve as a downstream effector of ras-regulated signal transduction. To determine whether GAP expression is required for the growth of human normal and leukemic hematopoietic cells, we used GAP antisense oligodeoxynucleotides to inhibit it and analyzed the effects of this inhibition on the colony-forming ability of nonadherent, T lymphocyte-depleted mononuclear cells and of highly purified progenitors (CD34+ MNC) obtained from the bone marrow and peripheral blood of healthy volunteers or chronic myeloid leukemia (CML, bcr-abl-positive) patients. The acute myelogenous leukemia cell line MO7, the Philadelphia BV173 cell line, and the acute promyelocytic leukemia NB4 and HL-60 cell lines were similarly examined. GAP antisense treatment inhibited colony formation from normal myelo-, erythro-, and megakaryopoietic progenitor cells as well as from CML progenitor cells. Proliferation of MO7 (growth factor-dependent) and BV173 (bcr-abl-dependent) cells, but not that of NB4 and HL-60 (growth factor-independent) cells, was also inhibited, even though a specific downregulation of GAP was observed in each cell line, as analyzed by either or both mRNA and protein expression. Stimulation of MO7 cells with hematopoietic growth factors increased the expression of GAP as well as the levels of active GTP-bound p21ras. Stimulation of GAP expression was inhibited upon GAP antisense treatment. These data indicate that p120 GAP is involved in human normal and leukemic hemopoiesis and strongly suggest that GAP is not only a p21ras inhibitor (signal terminator), but also a positive signal transducer.

In this study we identify and characterize a subset of human peripheral blood T cells, present in all individuals, that has features previously described for T cells either separately or in special circumstances. These cells are found in purified suspensions of resting peripheral blood lymphocytes within the CD8+ T lymphocytes, express alpha beta T cell receptor (TCR), and can be identified and isolated because of high-density expression of surface CD11b (TCR alpha beta +/CD3+/CD8+/CD11b+ cells). They coexpress constitutively the IL-2 receptor beta chain, Fc gamma RIIIA, and CD56. Although they do not mediate spontaneous cytotoxicity, CD3+/CD8+/CD11b+ cells have cytotoxic potential, demonstrated in redirected cytotoxicity assays with P815 target cells in the presence of anti-Fc gamma RIII (CD16) or anti-CD3 monoclonal antibodies. Stimulation of CD3+/CD8+/CD11b+ cells with rIL-2 induces proliferation, cytotoxicity against NK-sensitive and NK-resistant target cells, and expression of surface activation antigens, including IL-2 receptor alpha chain (CD25). CD3+/CD8+/CD16+/CD56+ cell clones with cytotoxic functions including those mediated by engagement of surface CD16 were obtained by limiting-dilution cloning of purified CD3+/CD8+/CD11b+ cells in the presence of rIL-2 and autologous feeder cells. Our data support the hypothesis that the CD3+/CD8+/CD11b+/CD16+ cells represent a discrete peripheral blood lymphocyte subset that could be the physiological counterpart of that expanded in several pathological conditions and in large granular lymphocyte lymphocytosis.

The transmembrane receptor for immunoglobulin G immune complexes on natural killer (NK) cells and macrophages, Fc gamma RIIIA (CD16), mediates cellular activation through a tyrosine kinase-dependent pathway. We show that Fc gamma RIII crosslinking results in activation of the src-related kinase p56lck in NK cells and demonstrate a physical association of p56lck with Fc gamma RIIIA in immunoprecipitates from NK cells obtained using anti-Fc gamma RIII antibodies or immune complexes. Our studies show that the zeta chain, the signal transducing subunit of Fc gamma RIIIA and of T cell receptor, associates with p56lck and, in NK cells, is a substrate for this kinase. Such direct association of p56lck with the zeta subunit as confirmed by demonstrating the interaction in heterologous cells transfected with cDNA expressing p56lck and zeta. Our findings demonstrate both functional and physical association of p56lck with Fc gamma RIIIA, through direct interaction of the kinase with the zeta and/or the gamma signal transducer subunits of the receptor. These data suggest a possible mechanism by which activation via Fc gamma RIIIA occurs.

Binding of ligand to the alpha subunit of Fc gamma RIIIA(CD16), expressed at the natural killer (NK) cell membrane in association with homo or heterodimers of proteins of the zeta family, results in phosphorylation of several proteins on tyrosine residues. We have analyzed the role of protein tyrosine phosphorylation in the regulation of molecular events induced upon stimulation of Fc gamma RIIIA in NK cells and in T cells expressing the Fc gamma RIII alpha chain in association with endogenous zeta 2 homodimers and devoid of other (CD3, CD2) transducing molecules. Our data indicate that treatment of these cells with protein tyrosine kinase inhibitors prevents not only Fc gamma RIIIA-induced protein tyrosine phosphorylation but also phosphatidylinositol 4,5 diphosphate hydrolysis and increased intracellular Ca2+ concentration, indicating a primary role of tyrosine kinase(s) in the induction of these early activation events. Occupancy of Fc gamma RIIIA by ligand results in phospholipase C (PLC)-gamma 1 tyrosine phosphorylation in NK cells and in Fc gamma RIIIA-transfected CD3-/CD2- T cells, and induces functional activation of p56lck in Fc gamma RIIIA alpha/zeta 2-transfected T cells, suggesting the possibility that the receptor-induced PLC-gamma 1 activation occurs upon phosphorylation of its tyrosine residues mediated by this kinase and is, at least in part, responsible for the signal transduction mediated via CD16 upon ligand binding.

Once natural killer cells identify their targets they engage their lysis machinery. Spontaneous, unlike antibody-dependent, cytotoxicity predominantly uses a Ras-independent pathway to accomplish this activation.

Natural killer (NK) cells are a discrete subset of leukocytes, distinct from T and B lymphocytes. NK cells mediate spontaneous non-MHC-restricted killing of a wide variety of target cells without prior sensitization and appear to be involved in initial protection against certain viral infections. Depressed NK cell-mediated cytotoxicity, one of the many immunological defects observed in AIDS patients, may contribute to secondary virus infections. Here we report that clonal and purified polyclonal populations of NK cells, which expressed neither surface CD4 nor CD4 mRNA, were susceptible to infection with various isolates of human immunodeficiency virus type 1 (HIV-1). Viral replication was demonstrated by detection of p24 antigen intracellularly and in culture supernatants, by the presence of HIV DNA within infected cells, and by the ability of supernatants derived from HIV-infected NK cells to infect peripheral blood mononuclear cells or CD4+ cell lines. Infection of NK cells was not blocked by anti-CD4 or anti-Fc gamma RIII monoclonal antibodies. NK cells from HIV-infected and uninfected cultures were similar in their ability to lyse three different target cells. Considerable numbers of cells died in HIV-infected NK cell cultures. These results suggest that loss of NK cells in AIDS patients is a direct effect of HIV infection but that reduced NK cell function involves another mechanism. The possibility that NK cells serve as a potential reservoir for HIV-1 must be considered.

We previously reported that natural killer cell stimulatory factor (NKSF), a heterodimeric lymphokine purified from the conditioned medium of human B lymphoblastoid cell lines, induces interferon gamma (IFN-gamma) production from resting peripheral blood lymphocytes (PBL) and synergizes with interleukin 2 in this activity. In this study, we show that human NKSF induces IFN-gamma production from both resting and activated human PBL and from freshly isolated murine splenocytes. Human T and NK cells produce IFN-gamma in response to NKSF, but resting PBL require the presence of nonadherent human histocompatibility leukocyte antigens DR+ (HLA-DR+) accessory cells to respond to NKSF. The mechanism(s) by which NKSF induces IFN-gamma production results in accumulation of IFN-gamma mRNA, is insensitive to cyclosporin A, and synergizes with those mediated by phytohemagglutinin, phorbol diesters, anti-CD3 antibodies, and allogeneic antigens, but not by Ca2+ ionophores. The ability of NKSF to directly induce IFN-gamma production and to synergize with other physiological IFN-gamma inducers, joined with the previously described ability to enhance lymphocyte cytotoxicity and proliferation, indicates that this lymphokine is a powerful immunopotentiating agent.

The low-affinity Fc receptor for immune-complexed IgG (FcγRIII; CD16) present on in vitro cultured human monocytes are encoded by an FcγRIII-2 gene that, by cDNA sequence analysis, is identical to that expressed on tissue macrophages and on natural killer cells. In macrophages, FcγRIII-2 encodes a glycoprotein of 52-62 kDa, with a peptide backbone of 33 kDa identical to that of the homologous receptor on natural killer cells. Like this and unlike in polymorphonuclear neutrophils, FcγRIII (CD16) on cultured monocytes is insensitive to phosphatidylinositol-specific phospholipase C, is not allelic for the neutrophil NA alloantigens NA-1/NA-2, is not recognized by a monoclonal antibody (1D3) detecting an epitope present only on neutrophil FcγRIII (CD16) and functions to trigger cytotoxicity upon ligand binding.

In the past year, natural killer cells have been the subject of much active investigation. The analysis of the effect of cytokines on the generation, proliferation and function of natural killer cells, and the definition of the lymphokines that they produce, have been particularly important areas of research in view of their possible application in adaptive immunotherapy, combined with biological response modifiers.

NK cells play critical roles in immune responses against tumors or virus infections by generating type 1 cytokine and cytotoxicity responses. In contrast, during type 2 dominant immune responses, such as allergic diseases, activities of NK cells are often impaired. These type 2 immune-mediated diseases have been reported to be closely associated with local production of PGD2. PGD2 is an eicosanoid primarily synthesized by mast cells and alveolar macrophages, and it functions through two major receptors, D prostanoid receptor (DP) and chemoattractant receptor-like molecule on the Th2 cell. Within the immune system, PGD2 binding to DP generally leads to suppression of cellular functions. In the current study, we show that: 1) DP is expressed in human NK cells as detected by mRNA analysis and Western blot; 2) PGD2 inhibits cytotoxicity, chemotaxis, and type 1 cytokine production of human NK cells via signaling through DP; 3) PGD2 signaling via DP elevates intracellular cAMP levels and the inhibitory effects on NK cells are cAMP dependent; 4) PGD2 binding to DP suppresses Ca2+ mobilization triggered by the cross-linking of the activating receptor, CD16. Together, these data uncover a novel mechanism by which PGD2 functions through DP to suppress type 1 and cytolytic functions of human NK cells, thus contributing to the promotion of a type 2 immune response.

We previously reported that a subset of human peripheral blood CD3+ T cells expresses low-to-null CD2 levels (CD2-/lo), produces type 2 cytokines and is inducible to differentiate to functionally mature IFN-gamma+ cells. Multiple-color immunofluorescence analysis indicated that this population, representing

We have defined the existence of developmental relationships among human peripheral NK cells with distinct phenotypic and functional characteristics. These findings closely parallel the changes that occur in vivo during NK cell development, and in vitro in experimental culture systems supporting NK cell generation from hematopoietic progenitors. These new insights provide a simplified framework to understand NK cell immunobiology and the cellular bases for their roles in innate immunity, initiation and maintenance of immune responses via regulation of adaptive and accessory cell functions, and immune pathologies.

The ability to perform biological studies on Natural Killer (NK) cells requires effective methods for their isolation from hematopoietic cells of other lineages. NK cells are a discrete lymphocyte subset distinguishable from B- and T-lymphocytes on the basis of both physical and phenotypic characteristics that can be exploited for then purification. Techniques based on differential cell buoyancy (centrifugation on discontinuous density gradients, such as Percoll [1]) have been used to enrich NK cells from mixed lymphocyte populations, but do not allow purification of these cells to homogeneity. The mononuclear cell suspensions obtained, although enriched in NK cells, also contain variable proportions of other cell types (notably monocytes and/or activated T- and B-lymphocytes) (2) and subsets of NK cells of higher density are lost in these preparations.

We report that human peripheral NK cells expressing high CD56 levels (CD56+high) are terminally differentiated cells indistinguishable from mature NK cells recently activated in the presence of IL-12, and not a functionally distinct NK-cell subset or progenitors to mature CD56+low NK cells. CD56+high NK cells coexpress all differentiation Ags constitutive or inducible in mature (CD56+) NK cells, except CD16, present at lower level than on most mature NK cells. Also, activation markers, activating receptors and adhesion molecules, and most inducible receptors are expressed exclusively and constitutively and are inducible at higher levels on CD56+high than on CD56+low NK cells. Consistent with their activated phenotype, many CD56+high NK cells are cycling and mediate heightened effector functions (proliferation, IFN-γ and IL-10 but not IL-13 production) in response to IL-12 and other NK cell-specific stimuli. Conversely, IL-12 induces on CD56+low NK cells all markers constitutively expressed on the CD56+high NK cells, concomitantly preventing the IL-2 (and IL-15)-inducible expression of NKp44 and CD16 re-expression after immune complex-induced down-modulation, and CD56−/+low NK cells acquire a CD56+high NK cell phenotype in short term in vitro culture with IL-12. The significance of these findings to the NK cell-mediated regulation of immune responses and NK cell development is discussed.

IFN, produced during viral infections by accessory (type I IFN) or NK cells (type II IFN), play a primary role in the regulation of immune and anti-viral NK cell effector functions. Because IFN have anti-proliferative effects on several cell types, including hematopoietic cells, we asked whether they modulate proliferation of human NK cells, and whether IFN-alpha and IFN-gamma mediate distinct effects on NK cells at different developmental stages. Analysis of proliferation at the single-cell level in human NK cells indicated that both IFN types inhibit IL-4-induced accumulation of immature CD56(-) IL-13(+) NK cells in freshly separated peripheral blood lymphocytes and in cells derived from them after short-term cultures. However, IFN-gamma inhibited specifically the IL-4-dependent proliferation of these cells without affecting the IL-2-dependent one or that of the IL-13(-) cells, whereas IFN-alpha attenuated proliferation of NK cells at any developmental stage (both immature CD56(-)IL-13(+) and mature CD56(+)IL-13(-) IFN-gamma(+) NK cells) and contributed to their monokine-induced differentiation to IFN-gamma-producing cells. Adding to our previous report that IL-13 inhibits accumulation of mature IFN-gamma(+) NK cells, the present data unravel a mechanism by which peripheral immature IL-13(+) and mature IFN-gamma(+) NK cells can negatively regulate each other's accumulation.

It has been proposed that immune responses to intra- (type 1) and extra-cellular (type 2) pathogens are regulated by two T-cell subsets branching from type 0 cells on T-cell receptor (TCR) engagement and terminally differentiating in distinct cytokine environments. However, analysis at the single-cell level of human T cells and natural killer (NK) cells has revealed that peripheral immature cells of both lineages exist, produce only type 2 cytokines and either proliferate or differentiate to interferon-γ producing cells in distinct cytokine environments, even without TCR engagement. These data support the hypothesis that a modified balance between proliferation and/or survival and differentiation of immature type 2 cytokine-producing cells regulates productive, type 1, immune responses. This new concept provides a simpler and testable framework to understand and manipulate the immune system and immune pathologies.

A maturation-dependent change in phenotype and cytokine production from relatively immature CD161(-) or CD161(+), IL-13(+)IL-4(+), IFN-gamma(-), to mature CD161(+)CD56(+) IFN-gamma(+) cells occurs in primary human alpha-galactosyl ceramide-reactive CD1d-restricted natural killer T (NKT) cells under the control of IL-12 and other monokines. Modulation of this process upon alpha-galactosyl ceramide stimulation explains the opposite roles of NKT cells to drive type 1 and type 2 immune responses. Because the same developmental changes occurred and were similarly regulated in T cells, the data establish that NKT cells should no longer be considered a functionally unique regulatory subset. However, the results of their analysis can be taken as a model for immunotherapeutic approaches with T cells for which a nominal or surrogate antigen is defined.

The impairment of NK cell functions in the course of HIV infection contributes to a decreased resistance against HIV and other pathogens. We analyzed the proportion of mature and immature NK cell subsets, and measured subsets of IFN-γ and TNF-α-producing NK and T cells in viremic or therapy-suppressed HIV-infected subjects, and noninfected control donors. Viremic HIV+ individuals had significantly lower proportions of mature CD3−/CD161+/CD56+ NK cells and of IFN-γ-producing NK cells compared with noninfected donors, independent of CD4+ T cell counts. HIV-infected subjects with undetectable viral load recovered mature CD3−/CD161+/CD56+ NK cells and cytotoxicity against tumor (K562) and HSV-infected target cells to percentages comparable with those of uninfected individuals, but their NK cells remained impaired in their ability to produce IFN-γ. In parallel to these ex vivo findings, in vitro NK cell differentiation of CD34-positive cord blood precursors in the presence of R5 or X4 HIV-1 resulted in the production of NK cells with a normal mature phenotype, but lacking the ability to produce IFN-γ, whereas coculture of uninfected PBMC with HIV failed to affect mature NK cell properties or IFN-γ secretion. Altogether, our findings support the hypothesis that mature NK cell phenotype may be uncoupled from some mature functions following highly active antiretroviral therapy-mediated suppression of HIV-1, and indicate that relevant innate immune functions of NK cell subsets may remain altered despite effective viral suppression following antiretroviral treatment.

To determine whether production of type 1 and type 2 cytokines defines discrete stages of natural killer (NK) cell differentiation, cytokine expression was analyzed in human NK cells generated in vitro in the presence of interleukin-15 (IL-15) and/or IL-2 from umbilical cord blood hematopoietic progenitors. Like peripheral NK cells, the CD161(+)/CD56(+) NK cells from these cultures contained a tumor necrosis factor alpha (TNF-alpha)(+)/granulocyte macrophage-colony-stimulating factor (GM-CSF)(+) subset, an interferon gamma (IFN-gamma)(+) subset, mostly included within the former, and very few IFN-gamma(-)/IL-13(+) cells. Instead, most immature CD161(+)/CD56(-) NK cells, detectable only in the cultures with IL-2, produced IL-13, TNF-alpha, and GM-CSF, but not IFN-gamma, and contained an IL-5(+) subset. In short-term cultures with IL-12 and feeder cells, a proportion of the immature cells acquired the ability to produce IFN-gamma. Part of these produced both IFN-gamma and IL-13, irrespective of induced CD56 expression. These in vitro data indicate that ability to produce the type 2 cytokines IL-13 and IL-5 defines CD161(+) NK cells at intermediate stages of differentiation, and is lost upon terminal functional differentiation, concomitant with acquired ability to produce IFN-gamma.

The hypothesis that distinct subsets of NK cells produce type 2 and type 1 cytokines in resting naive lymphocytes was tested analyzing cytokine production at the single-cell level. Two non-overlapping IL-13+ and IFN-gamma+ subsets were identified in adult and neonatal NK cells. IL-2 maintained their relative proportion. Accumulation of the former was induced by IL-4, but not IL-13, and inhibited by IL-12; that of the latter was induced by IL-12 and inhibited by IL-4 and IL-13. IL-4 induced preferential proliferation of the pre-existing peripheral IL-13+ cells, whereas IL-12 had minimal effect on proliferation of the IFN-gamma+ NK cells. The IL-13+ cells (CD161+ only) are phenotypically distinct from the IFN-gamma+ ones (CD56+) before and after culture under any condition analyzed, and produce IL-13 in response to NK-sensitive target cells and PMA+Ca(2+) ionophore, whereas also FcgammaRIIIA and IL-2+IL-12 stimulate IFN-gamma production. These data define the existence and regulation of two distinct resting peripheral NK cell subsets producing type 1 and type 2 cytokines, and suggest possible roles for IL-13+ NK cells in allergy.

The human B-myb gene encodes a transcriptional regulator that plays an important role in cell cycle progression, differentiation, and survival. To assess the in vivo role of B-myb, we investigated the phenotype of mouse transgenic lines in which B-Myb expression in lymphoid tissues was driven by the LCK proximal promoter. Overexpression of B-Myb had no measurable effect on the subsets of splenic and thymic lymphocytes, but was associated with increased expression of Fas ligand in NK and T cells. B-Myb-overexpressing splenocytes expressed higher IFN-γ levels and contained higher percentages of cytokine-producing cells than wild-type (wt) splenocytes, as detected by Western blot analysis and ELISPOT assays, respectively. Ex vivo-cultured transgenic thymocytes and splenocytes had decreased survival compared with the corresponding cells from wt mice, possibly dependent on increased expression of Fas ligand. In addition, Fas ligand-dependent cytotoxicity of transgenic T and NK cells was significantly higher than that mediated by their wt counterparts. Together, these results indicate that B-Myb overexpression results in T and NK cell activation and increased cytotoxicity. Therefore, in addition to its well-established role in proliferation and differentiation, B-myb also appears to be involved in activation of NK and T cells and in their regulation of Fas/Fas ligand-mediated cytotoxicity

The serine-threonine mitogen-activated protein kinase (MAPK) family includes extracellular signal-regulated kinases (ERK), c-Jun N-terminal kinases (JNK), and p38 kinases. In NK cells, spontaneous or Ab-mediated recognition of target cells leads to activation of an ERK-2 MAPK-dependent biochemical pathway(s) involved in the regulation of NK cell effector functions. Here we assessed the roles of p38 and JNK MAPK in NK cell-mediated cytotoxicity. Our data indicate that p38 is activated in primary human NK cells upon stimulation with immune complexes and interaction with NK-sensitive target cells. FcγRIIIA-induced granule exocytosis and both spontaneous and Ab-dependent cytotoxicity were reduced in a dose-dependent manner in cells pretreated with either of two specific inhibitors of this kinase. Target cell-induced IFN-γ and FcγRIIIA-induced TNF-α mRNA accumulation was similarly affected under the same conditions. Lack of inhibition of NK cell cytotoxicity in cells overexpressing an inactive form of JNK1 indicates that this kinase, activated only upon FcγRIIIA ligation, does not play a significant role in cytotoxicity. These data underscore the involvement of p38, but not JNK1, in the molecular mechanisms regulating NK cell cytotoxicity.

We used a monoclonal antibody (mAb) (B203.13, IgM) generated from a mouse immunized with the human B/myeloid bi-phenotypic B1b cell line, to analyse haemopoietic cells. The antigen recognized by this mAb is expressed on most adult and umbilical cord blood CD21+ B cells, at minimal density on mature monocytes, and is undetectable on granulocytes, T, natural killer (NK) cells, and erythrocytes. Within umbilical cord blood and adult bone marrow haemopoietic progenitor cells, the B203.13 mAb recognized a surface marker, present on progenitor cells of several haemopoietic lineages, that was transiently expressed on early erythroid and T/NK progenitors, and was preferentially maintained on cells of the B and myeloid lineages. Within the CD34+ cells, B203.13 was expressed on early committed myeloid (CD33+) and erythroid (CD71dim) progenitor cells, as confirmed in colony formation assays. The mAb also reacted with cells of B and myeloid chronic leukaemias and cell lines. These data define B203.13 mAb as a novel reagent useful for the characterization of haemopoietic progenitors and leukaemias.

Mature natural killer (NK) cells use Ca2+-dependent granule exocytosis and release of cytotoxic proteins, Fas ligand (FasL), and membrane-bound or secreted cytokines (tumor necrosis factor [TNF]-α) to induce target cell death. Fas belongs to the TNF receptor family of molecules, containing a conserved intracytoplasmic “death domain” that indirectly activates the caspase enzymatic cascade and ultimately apoptotic mechanisms in numerous cell types. Two additional members of this family, DR4 and DR5, transduce apoptotic signals upon binding soluble TNF-related apoptosis-inducing ligand (TRAIL) that, like FasL, belongs to the growing TNF family of molecules. Here, we report that TRAIL produced or expressed by different populations of primary human NK cells is functional, and represents a marker of differentiation or activation of these, and possibly other, cytotoxic leukocytes. During differentiation NK cells, sequentially and differentially, use distinct members of the TNF family or granule exocytosis to mediate target cell death. Phenotypically immature CD161+/CD56− NK cells mediate TRAIL-dependent but not FasL- or granule release–dependent cytotoxicity, whereas mature CD56+ NK cells mediate the latter two.

To determine the possible role of the BCR/ABL oncoprotein SH3 domain in BCR/ABL-dependent leukemogenesis, we studied the biologic properties of a BCR/ABL SH3 deletion mutant (▵SH3 BCR/ABL) constitutively expressed in murine hematopoietic cells. ▵SH3 BCR/ABL was able to activate known BCR/ABL-dependent downstream effector molecules such as RAS, PI-3kinase, MAPK, JNK, MYC, JUN, STATs, and BCL-2. Moreover, expression of ▵SH3 BCR/ABL protected 32Dcl3 murine myeloid precursor cells from apoptosis, induced their growth factor-independent proliferation, and resulted in transformation of primary bone marrow cells in vitro. Unexpectedly, leukemic growth from cells expressing ▵SH3 BCR/ABL was significantly retarded in SCID mice compared with that of cells expressing the wild-type protein. In vitro and in vivo studies to determine the adhesive and invasive properties of ▵SH3 BCR/ABL-expressing cells showed their decreased interaction to collagen IV- and laminin-coated plates and their reduced capacity to invade the stroma and to seed the bone marrow and spleen. The decreased interaction with collagen type IV and laminin was consistent with a reduced expression of α2 integrin by ▵SH3 BCR/ABL-transfected 32Dcl3 cells. Moreover, as compared with wild-type BCR/ABL, which localizes primarily in the cytoskeletal/ membrane fraction, ▵SH3 BCR/ABL was more evenly distributed between the cytoskeleton/membrane and the cytosol compartments. Together, the data indicate that the SH3 domain of BCR/ABL is dispensable for in vitro transformation of hematopoietic cells but is essential for full leukemogenic potential in vivo.

The evidence accumulated in the past few years has indicated that natural killer (NK) cells, originally identified on the basis of their ability to mediate spontaneous cytotoxicity, play a primary role in regulating immune responses and homeostasis via release of cytokines. Polyclonal populations of NK cells can be induced to produce an overlapping set of soluble factors. Some of these, represented primarily by IFN-γ, TNF-α, and several growth factors for hematopoietic cells, have been clearly identified; others may represent novel cytokines. Production of specific cytokine subsets by distinct NK-cell subpopulations has not been documented, and the same cytokines are produced, in response to distinct stimuli, by both resting and activated NK cells, with the differences being for the most part quantitative, rather than qualitative. No cytokines capable of regulating NK-cell proliferation in an autocrine fashion have been described, and, with the exception of IL-5, cytokine production is not correlated with the proliferative capability of the cells. In this review, the stimuli leading to cytokine production are discussed in light of the knowledge of the molecular mechanisms by which distinct ligands (target cells, ligands for specific receptors, antibodies to triggering molecules, cytokines) regulate expression of cytokine-encoding genes and variably modulate each others’ effects. The biological consequences of the production of specific cytokines are related to their effects on the NK cells themselves and on other hematopoietic and nonhematopoietic cell types. These are discussed, in an attempt to provide a picture that helps in understanding how NK cells may participate in regulating inflammatory and immune responses independently, at least in part, of their cytotoxic functions.

The putative factors that couple the signal transduction from surface receptors to the activation of cytokine synthesis in natural killer (NK) cells have not been elucidated. We report here that the nuclear factor of activated T cells (NFATp), a cyclosporin A (CsA)-sensitive factor that regulates the transcription of several cytokines, mediates CD16-induced activation of cytokine genes in human NK cells. CD16 (Fc gamma RIIIA)-induced expression of cytokine mRNA in NK cells occurs via a CsA-sensitive and Ca(2+)-dependent mechanism. Stimulation of NK cells with CD16 ligands induces NFAT-like DNA binding activity in the nuclear extracts from these cells, as detected in electrophoretic mobility shift assays. This occurs with fast kinetics after stimulation, via a CsA-sensitive and Ca(2+)-dependent mechanism that does not require de novo protein synthesis. NK cell NFAT is present in the cytosol of nonstimulated cells, migrates to the nucleus upon stimulation, and can associate with AP-1. Two distinct molecules, NFATp and NFATc, have been reported to mediate NFAT activity. The results of supershift assays using NFATp- and NFATc- specific antibodies indicate that NK cell activation early after CD16 ligand binding involves primarily, if not exclusively, NFATp, and Western blot analysis shows that this has the same electrophoretic mobility (approximately 120 kD) as that of T lymphocytes. NK cells do not express NFATc constitutively, but NFATc mRNA accumulation is induced in these cells within 2 h of stimulation with CD16 ligands. However, supershift assays using the available mAb recognizing the T cell NFATc revealed no detectable NFATc protein in nuclear and cytoplasmic extracts from CD16- or phorbol ester-stimulated cells at any time tested, up to 4 h. These results provide the first direct evidence that both CsA-sensitive transcription factors, NFATp and NFATc, are expressed in human NK cells, and that their activation and/or expression can be regulated in primary cells by a single stimulus, that, in the case of CD16 in NK cells, results in early activation of NFATp and subsequently induced expression of NFATc mRNA.

Natural killer cell stimulatory factor (NKSF) or interleukin-12 (IL-12) is a heterodimeric cytokine with pleiomorphic effects on T and NK cells, including induction of lymphokine production, mitogenesis, and enhancement of spontaneous cytotoxic activity. Similarly to IL-2, NKSF/IL-12 enhances NK cell-mediated cytotoxicity within a few hours and independently from induced proliferation. This effect is independent from other induced cytokines, because it is not prevented by antibodies neutralizing interferon (IFN)-alpha, IFN-beta, IFN-gamma, IL-2 or tumor necrosis factor (TNF)-alpha and, unlike the induction of IFN-gamma production by peripheral blood lymphocytes, it does not require HLA class II-positive accessory cells. Enhanced cytotoxicity is accompanied by morphologic changes in NK cells, including a significant increase in the number of cytoplasmic granules. In addition to the previously described ability to enhance the cytotoxic activity of NK cells against tumor-derived target cells, NKSF/IL-12 is also a potent stimulator of cytotoxicity against virus-infected cells, either fibroblasts acutely infected with herpes viruses or T cell lines chronically infected with human immunodeficiency virus-1. NK cell-mediated antibody-dependent cytotoxicity or anti-CD16 antibody-redirected lysis is not significantly enhanced by NKSF/IL-12. However, the ability of resting peripheral blood T cells to mediate anti-CD3 antibody-redirected lysis is enhanced by 18-h incubation with NKSF/IL-12, indicating that this lymphokine can modulate the cytotoxic capability of both NK and T cells.

We analyzed the phenotypic changes associated with monocyte activation and differentiation using a newly developed monoclonal antibody (B148.4). Among peripheral blood leukocytes, the antigen recognized by this antibody is expressed on monocytes and granulocytes at high and low density, respectively. Antigen expression is lost during in vitro differentiation of monocytes and is absent on tissue macrophages, indicating that expression of this antigen is related to monocyte differentiation. Only la, 25-dihydroxyvitamin D3 and phorbol diesters, of several inducers tested, up-regulate B48.4 antigen expression on cells (monocytes and certain myeloid cell lines) that constitutively bear the antigen, without, however, allowing its maintenance during monocytic differentiation or inducing it on negative cells. By im- munoprecipitation from B148.4* U937 cells, the antigen is a complex of a major 116-kd and two minor 38- and 46-kd molecules. Analysis of eight different tissues reveals that the antigen is shared with endothelial cells. Biochemical characteristics, cellular distribution, and expression pattern on monocytes, myeloid cell lines, and AML cells upon culture with different stimuli indicate that B148.4 is a novel monocyte differentiation antigen. J. Leukoc. Biol. 53: 390–398; 1993.