An immunoglobulin E-dependent recombinant histamine-releasing factor induces interleukin-4 secretion from human basophils.

A novel recombinant histamine-releasing factor (rHFR), which stimulates secretion from a subpopulation of human basophils that express a particular type of immunoglobulin E (IgE) or IgE+, was found to induce interleukin-4 (IL-4) production from cells isolated from these same donors. The secretion of IL-4 protein induced by rHRF significantly correlated with histamine release and the amount of protein generated, and the kinetics were identical to those caused by anti-IgE activation. Furthermore, the ability of rHRF to induce IL-4 protein production from cells not normally responsive to this protein was transferred by passive sensitization with plasma containing IgE+ antibody. That this novel protein stimulates both mediator release and the secretion of IL-4 protein from human basophils suggests a prominent role for this molecule in allergic disease.

Interactions of the complement system with endotoxic lipopolysaccharides in immunoglobulin-deficient sera.

Bacterial lipopolysaccharides (LPS) derived from a variety of organisms effectively induced C consumption in humans, bovines, and porcines with developmental agammaglobulinemia; birds with experimental agammaglobulinemia; and humans with agammaglobulinemia syndromes. This interaction proceeded even in precolostral piglet sera which contained less than 2.5 x 10(-6) mg/ml gamma globulin, and led to generation of neutrophil chemotactic factor and anaphylatoxin in these sera. Hence, the LPS-C interaction can proceed in sera markedly deficient in immunoglobulin. The question of whether immunoglobulins can be bypassed in the LPS-C interaction, or whether they are regularly utilized in a way so efficient that their participation is masked, was considered.

Differences in the volume distributions of human lung mast cell granules and lipid bodies: evidence that the size of these organelles is regulated by distinct mechanisms.

We analyzed transmission electron micrographs of human lung mast cells by digitized planimetry and point counting to determine the cross-sectional areas of two distinct cytoplasmic organelles: specific granules and lipid bodies. Specific granules have a limiting membrane and often contain one or more cylindrical scroll-like inclusions. By contrast, lipid bodies are on average much larger than granules and lack both limiting membranes and inclusions. The measured cross-sectional areas of lipid bodies and scroll-containing granules were converted to equivalent volumes, and the noise in the frequency distribution of these volumes was smoothed using a moving bin technique. This analysis revealed (a) a periodic, multimodal distribution of granule equivalent volumes in which the modes fell at volumes that were integral multiples of the volume defined by the first mode (the "unit volume"), and (b) a modal granule equivalent volume frequency that occurred at a magnitude equal to four "unit volumes." Thus, specific granules appear to be composed of units of a narrowly fixed volume. Furthermore, the mean volume of intragranule inclusions was 0.0061 mu3, a value very similar to that calculated for the "unit volume" (0.0071 mu3). This result suggests that each "unit volume" comprising the individual scroll-type granules contains (or is capable of generating or accommodating) a single scroll-like inclusion. In contrast to the specific granules, mast cell lipid bodies lack a periodic, multimodal volume distribution. Taken together, these findings suggest that the volumes of human lung mast cell granules and lipid bodies are regulated by distinct mechanisms.

Differences in the behavior of cytoplasmic granules and lipid bodies during human lung mast cell degranulation.

We used a morphometric and autoradiographic approach to analyze changes in specific cytoplasmic granules and cytoplasmic lipid bodies associated with human lung mast cell degranulation. Mast cells were dissociated from lung tissue by enzymatic digestion and were then enriched to purities of up to 99% by countercurrent centrifugation elutriation and recovery from columns containing specific antigen bound to Sepharose 6 MB. Degranulation was induced by goat anti-IgE. At various intervals after stimulation, parallel aliquots of cells were recovered for determination of histamine release or were fixed for transmission electron microscopy. We found that lipid bodies, electron-dense structures that lack unit membranes, were present in both control and stimulated mast cells. Autoradiographic analysis showed that lipid bodies represented the major repository of 3H-label derived from [3H]arachidonic acid taken up from the external milieu. By contrast, the specific cytoplasmic granules contained no detectable 3H-label. In addition, lipid bodies occurred in intimate association with degranulation channels during mast cell activation, but the total volume of lipid bodies did not change during the 20 min after stimulation with anti-IgE. This result stands in striking contrast to the behavior of specific cytoplasmic granules, the great majority of which (77% according to aggregate volume) exhibited ultrastructural alterations during the first 20 min of mast cell activation. These observations establish that mast cell cytoplasmic granules and cytoplasmic lipid bodies are distinct organelles that differ in ultrastructure, biochemistry, and behavior during mast cell activation.

Histamine release in vitro: inhibition by catecholamines and methylxanthines.

Methylxanthines and catecholamines both inhibit antigenically induced histamine release from human leukocytes. They act synergistically to inhibit the reaction, but must be present when antigen is added; preincubation is not effective. Since both increase cellular levels of cyclic 3 ', 5'-adenosine monophosphate it is postulated that this compound plays a role in the regulation of allergic histamine release.

Selective display of histamine receptors on lymphocytes.

Histamine, acting on histamine type 2 receptors, increases intracellular cyclic adenosine monophosphate (AMP) and thus modulates the immunologic functions of lymphocytes. Lymphocyte cyclic AMP levels were used to follow the development of histamine receptors. The B lymphocytes have no functional histamine receptors. As T lymphocytes "mature" in immunologic function--from thymocytes to cortisone-resistant thymocytes to splenic T lymphocytes--their response to histamine increases. The response of these subpopulations of lymphocytes to isoproterenol is the inverse of the histamine response. It is suggested that the changing display of histamine receptors plays an important part in the control of immunologic responses.

Molecular identification of an IgE-dependent histamine-releasing factor.

An immunoglobulin E (IgE)-dependent histamine-releasing factor (HRF) produced by lymphocytes of atopic children and present in biological fluids of allergic patients has been identified and purified. Amino-terminal sequencing revealed extensive homology to a mouse protein, p21, and its human homolog, p23. Both recombinant proteins caused histamine release from the human basophils of a subpopulation of donors, and this release was dependent on IgE. Polyclonal antibodies recognized and removed the biological activity of recombinant and native HRF. HRF identifies a heterogeneity of IgE and is believed to play a prominent role in chronic allergic disease processes.

Histamine augments leukocyte adenosine 3',5'-monophosphate and blocks antigenic histamine release.

Extracellular histamine stimulates accumulation of adenosine 3',5'-monophosphate in human leukocytes and prevents antigenic release of histamine from cells of allergic donors. Both effects occur at histamine concentrations that can be achieved by antigenic release of the amine in vitro.

Histamine release from human leukocytes: studies with deuterium oxide, colchicine, and cytochalasin B.

Agents known to interact with either microtubules or microfilaments influenced the antigen-induced release of histamine from the leukocytes of allergic individuals. Deuterium oxide (D(2)O) which stabilizes microtubules and thereby favors their formation enhanced histamine release markedly. Concentrations as low as 5% increased antigen-induced release somewhat while concentrations as high as 75% had no effect on release in the absence of antigen. Enhancement occurred over a wide range of antigen concentrations and was also seen when release was initiated by antibody to IgE or IgG. When the release process was divided into two stages a D(2)O activity could be demonstrated only in the second stage. However, when D(2)O was present in the first stage together with agents which raise cyclic AMP levels and thereby inhibit release it partially reversed this inhibition. Colchicine, demecolcine, and vinblastine, compounds known to disaggregate microtubules, i.e., have an effect opposite to that of D(2)O, inhibited the release of histamine and counteracted the effects of D(2)O. The inhibitory action of colchicine was greater if cells were treated with colchicine before rather than after activation with antigen. Cytochalasin B, a compound which causes the disappearance of microfilaments, had variable effects on histamine release. The most frequently seen response was slight enhancement. Neither D(2)O nor cytochalasin B altered cyclic AMP levels in leukocytes. These observations support and strengthen the view that an intact and functioning microtubule system is directly important for the secretion of histamine from leukocytes and suggest that microfilaments might have multiple indirect effects.

Defective histamine release in chronic urticaria.

Histamine release from peripheral blood leukocytes challenged with anti-human IgE was studied in patients with chronic urticaria and nonatopic controls. 19 of 23 controls, but only 6 of 20 patients, released over 20% of the total available leukocyte histamine. The response to anti-IgE concentrations of 1.66, 0.33, 0.066, and 0.013 mug antibody N/ml was significantly lower in patients than in controls. Serum IgE levels were significantly higher in the patients but total histamine content of about 10(7) leukocytes was not. Deuterium oxide (D2O) greatly increased histamine release (in both groups), indicating that the anti-IgE interacted with the basophils of urticaria patients. Passive sensitization of leukocytes with biologically active IgE was achieved in both patients and control subjects whose cells responded to anti-IgE, but was not achieved in either patients or control subjects whose cells were nonresponsive to anti-IgE challenge. 125I-anti-IgE autoradiographic studies revealed no obvious quantitative abnormality in the amount of basophil-bound IgE in chronic urticaria patients. Ionophore stimulation of aliquots of the same leukocytes used for anti-IgE challenge demonstrated that the urticaria patients' basophils were capable of releasing normal amounts of histamine. Leukocyte cyclic AMP levels in the two groups were not significantly different either in base-line levels or in responsiveness to stimulation with isoproterenol. These data indicate that chronic urticaria patients have a (acquired?) defect in leukocyte histamine release that occurs after the anti-IgE-IgE interaction, but before the actual (second-stage) release process, and that is comparable to the phenomenon of desensitization.

Properties of a subpopulation of T cells bearing histamine receptors.

C57BL/6 mice immunized i.p. with alloantigen (P815 mastocytoma cells) develop cytolytically active thymus-derived (T) splenic lymphocytes. The definition of specific histamine receptor sites on effector T cells has been studied by measuring the in vitro effects of the hormone on cytolytic activity. Histamine was found to inhibit cytolysis reversibly and to increase lymphoid cell cyclic AMP levels. Both of these histamine activities were reversed by burimamide and metiamide; neither activity was affected by diphenhydramine or pyrilamine. These findings indicate that modulation of effector T cell activity by histamine is mediated only by one of the subtypes of tissue histamine receptors, designated a histamine-type 2 receptor. This receptor appears to be present on cytolytically active cells; there is no evidence for activation by histamine of auxiliary or "suppressor" cells. The estimated dissociation constant (KB) for the burimamide-receptor complex (9 times 10-minus 6 tm) and for the metiamide-receptor complex (8 times 10-minus 7 M) indicated that the histamine receptor on T cells is quite similar to histamine-type 2 receptors in other tissues. Cells bearing such receptors could not be isolated by passage through a column of histamine-coated tsepharose beads. The cytolytic activity of spleen cells taken from mice early (days 7-9) after immunization is virtually unaffected by histamine in vitro. In contrast, the activity of spleen cells taken from mice later in the immune response is progressively more susceptible to inhibition by histamine. After reaching a maximum at day 11, the spleen cell cytolytic activity falls in a pattern that parallels the increase in susceptibility to histamine. The susceptibility of effector T cells to histamine appears also to reflect their site of origin, for peritoneal exudate effector cells were found to be significantly less sensitive than spleen cells to inhibition by histamine. The progressive increase in inhibition by histamine apparently reflects the appearance of greater numbers of specific histamine-type 2 receptors, and is probably a general phenomenon, for spleen cells from A/J or C3H mice immunized with either P815 mastocytoma (H-2d) or EL-4 (H-2b) cells showed the same effect. However, the appearance of histamine receptors could be altered by prior immunization with an unrelated alloantigen: thus, when A/J mice are preimmunized with EL-4, a subsequent immunization with mastocytoma cells results in peak spleen anti-H-2d activity at day 9 instead of days 11-13, and the appearance of significant (greater than 40 percent) inhibition by histamine as early as day 8 instead of day 16. The physiological role of the histamine receptors is as yet undefined, though their unexpected rate of appearance on effector T cells, coincident with a decline in the number of lytically active cells in vivo, may be a significant hint that hormone receptors play a role in the control of T-cell proliferation.

Anaphylactic relase of a basophil kallikrein-like activity. II. A mediator of immediate hypersensitivity reactions.

This report describes the immune release of a new mediator from human peripheral leukocytes, a basophil kallikrein-like activity (BK-A). The release process is initiated by the interaction of antigen on anti-IgE with cell-bound IgE, and appears to be similar in mechanism to the relase of histamine and other mediators of the immediate hypersensitivity reaction. The dose-response relationships and kinetics of histamine and BK-A release from antigen-challenged peripheral leukocytes are similar. The relase of the BK-A is calcium and temperature dependent, requires metabolic energy, and is controlled by hormone-receptor interactions that influence the cellular level of cyclic AMP, as has been described for other mediators of immediate hypersensitivity reactions. The data indicate that the interaction of BK-A with human plasma kininogen, generates immunoreactive kinin. We conclude that the antigen-IgE interation leads to the release from human basophils of a new mediator, a basophil kallikrein-like activity which may well be a link between reactions of immediate hypersenstivitity and the plasma and/or tissue kinin-generating systems.

IgE receptors on human basophils. Relationship to serum IgE concentration.

As reported previously, and confirmed here in 26 donors, the serum IgE level (2.6-5,500 ng/ml) correlates well (rs = 0.95, P less than 0.001) with the in vivo number of IgE molecules/basophil (6,000-600,000). The total number of IgE receptors/basophil was monitored by incubating them with an IgE-rich serum (15 microgram/ml), quantitatively stripping IgE from the cells at pH 3.7, and measuring eluted IgE by a direct radioimmunosorbent test. Saturation of receptors for each donor was achieved with 15 nM IgE (3 microgram/ml). The proportion of receptors occupied in vivo correlated with the serum IgE (rs = 0.84, P less than 0.001) whereas the average association constant of the receptors was independent of serum IgE and ranged from 7.1 X 10(8)/M to 2.8 X 10(10)/M, averaging 7.7 X 10(9)/M. Unexpectedly, the total number of IgE receptors/basophil was closely related to the serum IgE level. (rs = 0.92, P less than 0.001). Thus, either there is genetic association between serum IgE and the number of basophil IgE receptors, or, more likely, the receptor number is modulated by the serum IgE concentration.

Effects of cholera toxin on in vitro models of immediate and delayed hypersensitivity. Further evidence for the role of cyclic adenosine 3',5'-monophosphate.

Cholera enterotoxin inhibits the antigen-induced. IgE-mediated release of histamine from human leukocytes and the lysis of allogeneic mastocytoma cells by splenic lymphocytes from specifically immunized mice. This effect requires a prolonged preincubation time of the toxin with the lymphocyte/leukocyte preparations: a demonstrable inhibition requires about 30 min of pre-incubation and the toxin activity is still increasing at 90-180 min. Cholera enterotoxin also stimulates adenyl cyclase and leads to increased levels of cyclic AMP in the lymphocyte/leukocyte preparations. The concentration of toxin required for both cyclic AMP accumulation and inhibition of the biologic responses is about the same (ca. 1 ng/ml), and the time course of cyclic AMP accumulation parallels the development of inhibitory activity. Both activities, inhibition of the in vitro hypersensitivity reactions and cyclic AMP accumulation, are blocked by cholera antitoxin and by a toxoid prepared from the toxin (choleragenoid). These are specific antagonists in that they do not block the inhibiting activity or rise in cyclic AMP levels caused by other adenyl cyclase stimulators. Because cholera enterotoxin has no known activity other than the stimulation of adenyl cyclase and because of its unusual time course and the availability of specific antagonists, this data considerably strengthens the hypothesis that the cyclic AMP system influences the expression of these two forms of hypersensitivity phenomena.

Effects of cholera enterotoxin on adenosine 3',5'-monophosphate and neutrophil function. Comparison with other compounds which stimulate leukocyte adenyl cyclase.

Cholera enterotoxin caused a delayed accumulation of adenosine 3',5'-monophosphate (cyclic AMP) in human leukocytes, associated with an increase in leukocyte adenyl cyclase activity. The action of cholera enterotoxin contrasted with that of other agents which stimulate adenyl cyclase: (a) the effects of the toxin were delayed in onset, while prostaglandin-E(1) (PGE(1)) and isoproterenol acted rapidly; (b) removal of the soluble toxin from the extracellular medium did not abolish its effects on cyclic AMP and inhibition of antigenic histamine release, while removal of PGE(1) did prevent its effects; (c) PGE(1), but not cholera enterotoxin, stimulated adenyl cyclase activity when added directly to broken cell preparations. Binding of the toxin to leukocytes was rapid and irreversible, and was followed by a gradual increase in cyclic AMP which was not prevented by cycloheximide. Cholera enterotoxin caused accumulation of cyclic AMP in purified human neutrophils as well as mono-nuclear cells, but did not prevent the extrusion of lysosomal hydrolases from phagocytic cells. The toxin only slightly inhibited the ability of human neutrophils to kill Candida albicans. Thus these results with the toxin cast doubt on previous proposals that cyclic AMP regulates these two functions of neutrophils. The unique action of cholera enterotoxin on cyclic AMP production provides a potentially useful pharmacologic tool, in addition to methylxanthines and dibutyryl cyclic AMP, for testing hypotheses relating cyclic AMP to altered function of leukocytes and, perhaps, of other mammalian cells.

In vivo suppression of the immune response to alloantigen by cholera enterotoxin.

The immune response of C57BL/6 mice to allogeneic (DBA/2) mastocytoma cell suspensions was profoundly suppressed by intraperitoneal administration of 1 mug cholera enterotoxin 4 days after antigenic stimulation. The immune response assayed 11 days after antigen showed decreased cytolytically active thymusderived (T) lymphocytes and markedly depressed serumagglutinating titers. A comparable suppression of the immune response to skin allografts (DBA/2-->C57BL/6) was also effected by cholera toxin administration, although there was no prolongation of allograft survival. The mechanism of the immune suppression is apparently related to the known adenylate cyclase stimulatory activities of choleragen.

Immunologic and cellular changes accompanying the therapy of pollen allergy.

We had found previously that children with ragweed hay fever were somewhat less symptomatic after preseasonal immunization with large doses of ragweed pollen extract than were placebo-treated children. To study further the immunologic changes which accompany immunotherapy, these children were treated again the following year. Each patient served as his own control. Serum blocking (IgG) antibody, measured by inhibition of antigen-induced leukocyte histamine release, was increased 20- to 40-fold after therapy. The anticipated postpollen season increase of serum reaginic (IgE) antibody, measured by passive sensitization of leukocytes from nonallergic donors, was suppressed. Instead, the mean titer was decreased after treatment. Total serum IgE levels, measured by radial radioimmunodiffusion assay, were higher than normal; were correlated with reaginic antibody titers; and also did not increase in the pollen season after treatment. The concentration of both IgE and reaginic antibody was lower in the older children, irrespective of treatment. Leukocyte response to ragweed antigen E and guinea pig anti-IgE antiserum was assessed by means of in vitro histamine release techniques. After treatment, the leukocytes of 21 patients were less sensitive (11 cases), or less reactive (10 cases), to antigen E. Response to anti-IgE antibody also was diminished after treatment. In four cases, neither anti-IgE nor antigen E induced histamine release, although both IgE protein and ragweed-specific IgE antibody were present in the patients' own sera. Clinical improvement was correlated best with decreased leukocyte sensitivity and leukocyte reactivity to ragweed antigen E. It appeared that decreased cell sensitivity was related to lower serum reaginic antibody levels. Decreased cell reactivity, in the presence of both IgE protein and IgE antibody in the serum, may indicate a change in cellular response mechanisms. These studies suggest that clinical improvement following specific immunotherapy must be the result of complex changes in the immunologic and cellular components of allergic disease.

Anaphylactic release of a basophil kallikrein-like activity. I. Purification and characterization.

These studies describe the IgE-mediated relase of a basophil kallikrein-like enzyme that is an arginine esterase and is inhibited by plasma, diisopropylphosphofluoridate, and Trasylol. The substrate specificity for the synthetic amino acid ester substrates p-toluenesulfonyl-L-arginien methyl ester, benzoyl-arginine methyl ester, and acetyl-tyrosine methyl ester is similar for the basophil enzyme and plasma kallikrein. The interaction of arginine esterase-active fractions from ion-exchange (DEAE-Sephacel) and gel filtration (Sepharose 6B) chromatography, with human plasma kininogen, generates immunoreactive kinin. The basophil arginine esterase and kinin-generating activities co-chromatograph on Sepharose 6B and the quantity of kinin generated is, in general, proportional to the arginine esterase activity of the column fractions, suggesting that these two activities are subserved by the same protease. The ability of this protease to generate kinin equally well from heat- and acid-treated plasma, as from fresh human plasma, suggests that this protease has kallikrein-like activity. These data suggest that kallikrein-like activity can be generated from human basophils as a direct result of a primary IgE-mediated immune reaction, thus providing a potential link between reactions of immediate hypersensitivity and the plasma and(or) tissue kinin-generating systems.

Hyperosmolar triggering of histamine release from human basophils.

Idiopathic reactions occurring during the infusion of hyperosmolar solutions, such as radiocontrast dyes, cause a significant number of deaths each year. These reactions are similar to those which follow mediator release during allergen-induced anaphylaxis. In attempting to explain these nonimmunologic reactions, we examined the direct effect of hyperosmolarity on normal human basophils with emphasis on release induced by mannitol. The cells of all donors released histamine in vitro in response to hyperosmolar (0.2-0.7 M) solutions of a number of solutes including mannitol. That this was not a toxic process was supported by a number of criteria, including inhibition of release by excess stimulus at 37 degrees C and a lack of release at 4 degrees C. Furthermore, electron microscopic studies revealed that hyperosmolar stimulation did not disrupt the cell membrane or lead to any signs of cytotoxicity. In contrast to antigen-stimulated release, where granules fuse only with the cell membrane, granules in mannitol-stimulated cells, in addition to fusing with the cell membrane, may also be extruded into a common intracellular sac before exteriorization. Characteristics similar to antigen-induced histamine release included the time-course for release, inhibition by drugs that modify phospholipid metabolism, p-bromophenacyl bromide, and eicosa-5,8,11,14-tetraynoic acid, and augmentation of release by deuterium oxide (D(2)O). The release process differed from antigen-induced release by a number of criteria, including independence from immunoglobulin (Ig)E-related mechanisms, insensitivity to agonists that elevate intracellular cyclic AMP, minimal dependence on extracellular calcium, lack of inhibition by 2-deoxyglucose and theophylline, and a temperature optimum of 32 degrees C. We conclude that this noncytotoxic hyperosmolar release process is different from IgE-mediated secretory events and may well play a role in the idiopathic reactions which occur secondary to the infusion of hyperosmolar solutions in man.

Anaphylatoxin-induced histamine release with human leukocytes: studies of C3a leukocyte binding and histamine release.

Purified human C3a and synthetic COOH-terminal peptides of C3a, i.e., a pentapeptide, Leu-Gly-Leu-Ala-Arg (5R), and an octapeptide, Ala-Ala-Ala-Leu-Gly-Leu-Ala-Arg (8R) induced histamine release from human basophil granulocytes. On a molar basis, 5R was one-tenth and 8R was one-fifth as active as C3a in causing histamine release. It was found that 125I-C3a binds to whole leukocytes, interacting with both mononuclear cells and neutrophils and the binding was inhibited by preincubation of cells with unlabeled C3a, but not by C5a. 5R and 8R also inhibited the binding of 125I-C3a to the cells. However, on a molar basis, 2,000 times more 8R or 6,000 times more 5R is required for 50% inhibition of 125I-C3a binding as compared with native C3a. Autoradiography of cells using 125I-C3a and 125I-C5a showed preferential binding of 125I-C3a to eosinophils and basophils, whereas 125I-C5a binds primarily to neutrophils and eosinophils and to a lesser extent to basophils. The preferential binding of C3a and C5a to different cell types may herald significance related to their physiological functions.