Effector cells of anaphylaxis: mast cells and basophils

Effector cells of anaphylaxis: mast cells and basophils

Abstract. Systemic anaphylaxis arises when mast cells, possibly along with other cell types, are provoked to secrete mediators that evoke a systemic response. Mast cells in perivascular, respiratory, gastrointestinal and cutaneous tissues are likely involved, regardless of whether IgE or non-IgE-dependent pathways are invoked. affib tryptases are selectively and abundantly produced by mast cells. Tryptase levels in the circulation provide a precise indicator of mast cell involvement. Mature b tryptase is stored in secretory granules and is released when the cells are activated to degranulate, as occurs in anaphylaxis. affib proffipro’ tryptases are spontaneously secreted by mast cells. Consequently, mature tryptase levels in serum (normally 1 ngffiml) are elevated in systemic anaphylaxis. Total tryptase levels (mature plus precursor forms), normally 1^ 15 ngffiml in baseline serum samples, are elevated in patients with systemic mastocytosis (420 ngffiml), a disease that also predisposes one to anaphylactic reactions. The assessment of basophils in systemic anaphylactic reactions has been problematic, because an assay for a speci¢c releasable marker from this cell type has not been developed. Nevertheless, in cases of anaphylaxis in which elevations of histamine, but not tryptase, have been detected, it is enticing to speculate that basophil-dependent anaphylaxis may have occurred.

2004 Anaphylaxis. Wiley, Chichester (Novartis Foundation Symposium 257) p 65^79

Definition

Systemic anaphylaxis arises when mast cells are provoked to secrete mediators that evoke a systemic response. Although mast cells in any organ system may be involved, depending on the distribution of the instigating stimulus, the principal targets include the cardiovascular, cutaneous, respiratory and gastrointestinal systems, sites where mast cells are most abundant. The terms anaphylactic and anaphylactoid, respectively, attempt to distinguish between mast cell activation initiated by allergen and IgE through FceRI, classical immediate hypersensitivity, versus those that initiate mast cell activation by alternative pathways. Although the mediators elicited from mast cells will overlap extensively in anaphylaxis and anaphylactoid reactions, and thereby invoke similar acute therapies, understanding di¡erences in causation will likely impact on therapeutic interventions aimed at preventing future attacks. Cells other than mast cells also undoubtedly participate in systemic anaphylaxis, particularly those armed with antigen-speci¢c IgE. Basophils, like mast cells, constitutively express substantial amounts of the tetrameric (abg2), high a⁄nity receptor for IgE, FceRI, and when activated through this pathway, also release mediators within minutes. Eosinophils, monocytes, antigen-presenting cells and epithelial cells may be induced to express this receptor, primarily in its trimeric (ag2) form and thereby a¡ect the intensity, duration or character of anaphylactic reactions. Whether some cases of systemic anaphylaxis occur through one or more of these cell types without involving mast cells is theoretically possible, but remains controversial.

Precipitating factors

Most IgE-dependent mast cell activation events occur at local sites, and result in local disease. For example, allergic conjunctivitis, allergic rhinitis or allergic asthma typically occurs when allergen lands on the corresponding mucosal surface of a sensitive individual. Systemic anaphylaxis presumably requires the allergen (or non-allergen agonist) to distribute systemically before mast cells at remote sites will be activated. This is most likely to occur with parenteral administration, less likely with administration by the oral route, by inhalation or by direct cutaneous or ocular contact. Activation of mast cells in perivascular locations should have the greatest e¡ect on systemic vascular responses, even though large amounts of mediators released locally in theory could spill into the circulation and a¡ect remote sites. However, the precise distributions of mast cells that are activated during anaphylactic reactions are undetermined. Also, the numbers of mast cells available to respond, or factors a¡ecting mast cell releasability may have an impact on the severity of the response of a sensitive subject to an allergen challenge, and need to be better understood.

Allergens

The most common allergens causing systemic anaphylactic reactions include drugs, e.g. penicillin, insect venoms, foods, radiocontrast media, allergen immunotherapy injections, latex and autoantigens. Proteins or glycoproteins typically serve as complete antigens. In contrast, most drugs act as haptens after coupling to self proteins. Allergen multivalency is important, because cross- linking of IgE on the surface of cells brings together at least two FceRI molecules that then transmit an activating signal into the cell. A monovalent antigen would block cross-linking and thereby should block activation.

An allergen exposure must lead to sensitization before an immediate hypersensitivity reaction can occur, a process involving antigen processing, presentation to Th2 cells and class-switching of allergen-speci¢c B cells to IgE, a process that takes between one to two weeks. Consequently, anaphylaxis does not occur upon ¢rst exposure to an allergen, but may occur after subsequent exposures.

Non-IgE-dependent agonists

Most non-IgE-dependent foreign agents do not require antigen processing, and can elicit a mast cell-activation response upon ¢rst exposure. These would include radiocontrast dyes, narcotics such as codeine and morphine, and vancomycin. Endogenous mast cell activators include neuropeptides such as substance P, neurokinin A and calcitonin gene-related peptide, and the complement anaphylatoxin C5a. Whether a magnitude of mast cell activation su⁄cient to cause systemic anaphylaxis can result from endogenous secretion or generation of these peptides by themselves is unproven.

Aspirin hypersensitivity can manifest either at mucosal or cardiovascular sites.

Most cases of aspirin hypersensitivity appear to be pharmacologically (not IgE) mediated, and in sensitive subjects can occur to any of the cyclooxygenase (COX)-1 inhibitors. A mechanism to explain mast cell activation has not yet emerged. COX-2-selective inhibitors are relatively safe in aspirin-sensitive asthmatics, but have not been adequately tested in aspirin-sensitive anaphylaxis subjects. The recent identi¢cation of COX-3 and COX-related enzymes (Chandrasekharan et al 2002) that may have distinct pharmacological pro¢les adds to the complexity of aspirin-mediated hypersensitivity reactions.

Autoimmunity

Some patients present with spontaneous episodes of anaphylaxis. In certain cases this may be an extension of a physical urticaria. Because chronic urticaria may be associated with IgG and IgM antibodies against FceRI or IgE, an analogous autoimmune process might cause anaphylaxis.

Epidemiology

Assessing the annual incidence of systemic anaphylaxis and the prevalence of those at risk for systemic anaphylaxis are compromised by imprecise diagnostic measures. Approximately 1500 to 2000 deaths occur per year from systemic anaphylaxis in the USA (Neugut et al 2001). Non-fatal cases are much more common, estimated to occur in about 0.2% of the population per year. Further analyses suggest that between 3 and 43 million (1^15% of the population) may be at risk for such reactions. Drug reactions account for the majority of cases. Food, insect sting and latex reactions account for many of the other cases.

Pathophysiology

Mast cells participate in both acquired and innate forms of immunity (Wedemeyer et al 2000). They develop in peripheral tissues from bone marrow progenitors primarily under the in£uence of stem cell factor, the ligand for the tyrosine kinase receptor called Kit. Armed with allergen-speci¢c IgE, they are activated by multivalent allergens that cross-link IgE and associated FceRI molecules on the mast cell surface. This may be important in the defence against certain parasites that elicit a strong IgE response. Experiments performed in rodents suggest that mast cells also can be directly activated by certain bacterial products, leading to the secretion of mediators that recruit neutrophils, and thereby restrain bacterial dissemination until a more potent acquired immune response develops. Activation of mast cells by endogenous peptides such as substance P or calcitonin gene-related peptide may in£uence basic biological processes such a wound healing and angiogenesis. Whether mast cells have a critical, non-redundant role in these biologic and immunological processes remains controversial. However, their central role in immediate hypersensitivity is clear.

Mediators released by mast cells include preformed mediators stored in secretory granules, newly generated products of arachidonic acid, and an array of cytokines and chemokines (Schwartz 2002). Histamine, formed from histidine by histidine decarboxylase, is the sole biogenic amine stored in all granules of human mast cells and human basophils. Histamine released by mast cells or basophils di¡uses freely, and interacts with H1, H2 and H3 receptors. Histamine-mediated bronchial and gastrointestinal smooth muscle contraction, vascular smooth muscle relaxation and increased permeability of postcapillary venules account for many of the signs and symptoms of systemic anaphylaxis, primarily but not exclusively through H1 receptors. Excessive levels of histamine in the CNS may account for the sense of doom commonly experienced at the onset of anaphylaxis. Once secreted, histamine is rapidly metabolized to methyl histamine and indole acetic acid. Consequently, the plasma histamine level is not a practical test for anaphylaxis. Urinary histamine levels re£ect the small portions of histamine not metabolized in the circulation before renal clearance, and are a¡ected by ingested histamine-containing foods and histamine-producing mucosal bacteria, compromising the utility of this test.

Prostaglandin D2 (PGD2) is the principal COX-catalysed product of arachidonic acid secreted by activated mast cells, but is not made by basophils. It binds to the G protein-coupled receptors CRTH2 and DP (Hirai et al 2001). Both COX-1 and COX-2 are involved in PGD2 production by mast cells. Consequently, a COX inhibitor that is bipotent might be better than one which is selective at

blocking PGD2-mediated responses during anaphylaxis, which may include hypotension, bronchospasm and inhibition of platelet aggregation.

Leukotriene C4 (LTC4), is released by both mast cells and basophils after its formation from arachidonic acid and glutathione is sequentially catalysed ¢rst by 5-lipoxygenase and 5-lipoxygenase activating protein and then by LTC synthase. Conversion to LTD4 and LTE4, which also are bioactive, occurs in the extracellular space. These sul¢dopeptide leukotrienes bind to CysLT1 (smooth muscle, epithelial and endothelial cells, lung macrophages, eosinophils) (Evans 2002) and CysLT2 (endothelial and epithelial cells), both G protein-coupled receptors. Sul¢dopeptide leukotrienes cause bronchoconstriction, mucus secretion, eosinophil recruitment, increased vasopermeability, diminished cardiac contractility, vasoconstriction of coronary and peripheral arteries and vasodilation of venules.

Mast cells also are the principal source of heparin proteoglycan and the proteases a tryptase, b tryptase, chymase and mast cell carboxypeptidase. Like neutrophils and monocytes, they also contain cathepsin G. Basophils are relatively de¢cient in these enzymes. Mature tryptase is stored in the secretory granules of all mast cells, while the other proteases appear together in a subset of mast cells. Those with tryptase alone are called MCT cells, and are the predominant type of mast cell in the lung and small intestinal mucosa, while those with all proteases are called MCTC cells, and account for most of the mast cells in skin, intestinal submucosa, conjunctiva and blood vessel walls. The role(s) of these molecules in the pathophysiology of anaphylaxis are unde¢ned.

Tryptase (EC3.4.21.59) is the most abundant protein product produced by human mast cells, accounting for about 20% of the cell protein, and is derived principally from two genes on chromosome 16, a tryptase and b tryptase. The product of the b tryptase gene(s) is autoprocessed from b protryptase to b pro’tryptase at acidic pH, optimally in the presence of heparin proteoglycan, and then to b tryptase by a dipeptidase, thought to be dipeptidyl peptidase I in humans (Sakai et al 1996). Tryptase in murine mast cells utilizes a di¡erent dipeptidase (Wolters et al 2001). Mature b tryptase is stored in secretory granules as an enzymatically active tetramer in a complex with heparin proteoglycan until the cells are activated to degranulate and release the protease^proteoglycan complex. In contrast, a protryptase may not undergo autoprocessing from a protryptase to a pro’tryptase, because a tryptase-resistant Q-3 rather than a tryptase-sensitive R-3 is present in the -3 position of the propeptide (Sakai et al 1996). Indeed, when mature a tryptase is produced in vitro, though it forms a tetramer, the protein appears to be enzymatically inactive (Huang et al 1999, Marquardt et al 2002, Selwood et al 2002).

The major form of tryptase found in normal blood fails to bind to the G5 mAb, which recognizes mature forms of rha and rhb tryptases, but not the corresponding pro or pro’ forms of tryptase (Sakai et al 1996, Schwartz et al 1995, 2003). Thus, an immunoassay using this mAb measures mature tryptase. Although mature tryptase levels are undetectable in normal serum (51 ngffiml), they are elevated in the blood of most cases of systemic anaphylaxis with haemodynamic compromise, particularly when the precipitating agent is administered parenterally. In such cases, the magnitude of mast cell degranulation appears to be the primary determinant of clinical severity (Schwartz et al 1987, 1989, van der Linden et al 1992). In contrast, based on mAbs that recognize all forms of a and b tryptases, a total tryptase immunoassay was developed that detected levels of tryptase in baseline serum from essentially all individuals (mean ± SD, 4.9 ± 2.3 ngffiml) (Schwartz et al 1994). So-called total tryptase levels are elevated in subjects with systemic mastocytosis, and re£ect the total body burden of mast cells (Schwartz et al 1995). These observations can largely be explained by the in vitro observations that precursor forms of a and b tryptases are spontaneously secreted by unstimulated mast cells, while mature tryptase is preferentially stored in secretory granules and released by activated mast cells.

Approximately 25% of individuals lack a gene for a tryptase (Guida et al 2000, Soto et al 2002). Whether such a genetic di¡erence a¡ects the level of tryptase in the blood has been investigated. Preliminary data in healthy subjects indicates that a de¢ciency in the gene for a-tryptase does not in£uence the circulating levels of tryptase precursors in healthy subjects (Schwartz et al 2003). Finally, the hypothesis that mature forms of tryptase are preferentially stored in secretory granules, while immature forms of tryptase are preferentially selected for spontaneous secretion has not been directly examined.

Cytokines (TNFa, interleukins 4, 5, 6, 13 and 16, GM-CSF) and chemokines (interleukin 8,  monocyte chemotactic protein 1, monocyte in£ammatory protein 1a) represent another dimension of the mediators released by mast cells and basophils. Though not selectively produced by these cell types, the vasoactive and in£ammatory potential of such mediators could impact the severity and duration of anaphylaxis. Although cytokine secretion typically occurs in association with granulation, experiments in vitro indicate that cytokine secretion also can be induced in the absence of degranulation. As selective antagonists of the relevant cytokines and chemokines become available and are tested for therapeutic bene¢ts, their roles in the pathogenesis of anaphylaxis will be better understood.

Diagnosis and di¡erential diagnosis

Systemic anaphylaxis, with various combinations of hypotension, tachycardia, urticaria, bronchoconstriction, laryngeal oedema, colics and diarrhoea, often associated with a sense of doom, beginning within minutes of the provoking stimulus, can be precisely con¢rmed by demonstrating antigen-speci¢c IgE (sensitization) and an elevated b tryptase level in serum (mast cell activation). Skin testing or in vitro measurements of antigen-speci¢c IgE should be delayed for at least two weeks after the precipitating event to prevent false negative results. An increased level of mature tryptase in serum obtained within several hours following a hypotensive event, normal levels being undetectable, strongly suggests that mast cell activation occurred. During a study of experimental insect sting-induced anaphylaxis, the increased level of mature tryptase correlated closely with the drop in mean arterial pressure, indicating that the magnitude of mast cell activation is a primary determinant of clinical severity (van der Linden et al 1992). Objective criteria, such as an elevated level of mature tryptase, provide greater precision for the diagnosis of anaphylaxis than clinical signs and symptoms alone, and may be useful to distinguish anaphylaxis from other conditions. However, some cases of apparent IgE-mediated anaphylaxis, particularly some cases of food-induced anaphylaxis, are not associated with an elevated level of mature tryptase. This raises the question of whether there are anaphylactic pathways not involving mast cell activation, but instead perhaps basophil activation.

Anaphylaxis should be distinguished from a variety of disorders with overlapping presentations. Vasovagal syncope presents with diaphoresis, nausea, hypotension and bradycardia, but without urticaria. Flushing disorders may be benign and unrelated to anaphylaxis, or could be a manifestation of pathologic conditions such as the carcinoid syndrome, in which urticaria and profound hypotension are not typically associated, and phaeochromocytoma, which causes episodic hypertension. Precise detection of these latter conditions involves determining serum serotonin and urinary 5-hydroxyindole acetic acid, catecholamines and vanillylmandelic acid levels. Panic attacks and vocal cord dysfunction can be a challenge to distinguish from anaphylaxis, especially by history alone, but nevertheless must be considered. Acute attacks of hereditary and acquired angioedema due to C1 esterase inhibitor de¢ciency are not associated with pruritic urticaria, and persist longer than attacks of anaphylaxis. Shock due to complement activation by contaminated haemodialysis tubing, without involving mast cell activation, also has been reported. Scombroidosis occurs 5^90 min after ingestion of histamine in poorly-stored ¢sh, and presents with £ushing, palpitations, headache and gastrointestinal symptoms. The condition lasts several hours, both duration and severity depending on the amount of histamine ingested, and usually responds to H1 receptor and H2 receptor antihistamines, but occasionally requires epinephrine and intravenous £uids. Acute serum sickness, genetic cell activation syndromes, endotoxin- mediated septic shock and superantigen-mediated toxic shock syndromes present with fever, which is not characteristic of anaphylaxis by itself. Also, hypoglycaemia, seizure and primary pulmonary or cardiac events should be considered.

In some cases, systemic anaphylaxis may occur together with another disorder. For example, a 65 year-old male after being stung by a wasp complained of dizziness and shortness of breath, was hypotensive with urticaria, responded to treatment with subcutaneous epinephrine, then complained of chest pressure, and had an EKG indicating an inferior wall infarction. Both b tryptase level and cardiac enzymes were elevated, indicating both anaphylaxis and myocardial infarction had occurred.

Systemic mastocytosis is an important condition to consider in the di¡erential diagnosis of anaphylaxis (Schwartz 2001). In adults, somatic activating mutations in the gene for Kit in mast cell progenitors result in an excessive body burden of mast cells. In children with this disorder the disease may regress spontaneously, possibly due to the lack of this activating mutation. Patients with too many mast cells are at increased risk for anaphylaxis, and anaphylaxis may be a presenting manifestation of systemic mastocytosis. For example, anaphylaxis to an insect sting, particularly in the absence of venom-speci¢c IgE, should suggest the possibility of systemic mastocytosis. Diagnostic tests for systemic mastocytosis might include a biopsy of a skin lesion suspected to be urticaria pigmentosa, a bone marrow biopsy stained for mast cells (anti-tryptase immunohistochemistry being most sensitive), detection of bone marrow mast cells expressing surface CD2 and CD25, and an elevated serum level of total tryptase (mature plus immature forms of a and b tryptases) during a non-acute interval (Valent et al 2001).

Prevention and therapy

Acutely, treatment of systemic anaphylaxis ¢rst requires that airway patency, blood pressure and cardiac status be addressed. Epinephrine administration is critical, the earlier during the course of an anaphylactic event the better. Glucagon may be used in patients taking a b blocker. Parenteral administration of H1 and H2 receptor antihistamines may prevent progression of some of the signs and symptoms. Glucocorticosteroids may reduce the risk of protracted or recurrent anaphylaxis, but are unlikely to be of bene¢t acutely.

Patients who have experienced an anaphylactic reaction are at greatest risk of su¡ering another episode. Elevated baseline levels of total tryptase may be another indicator of anaphylactic risk (Fricker et al 1997, Ludolph-Hauser et al 2001, Schwartz et al 1994). Individuals at risk should wear a Medic-Alert bracelet, carry epinephrine, and avoid b blockers and ACE inhibitors as well as agents to which they are sensitive. In subjects with recurrent anaphylaxis, prophylactic use of H1 and H2 receptor antihistamines is bene¢cial. A leukotriene antagonist and cyclooxygenase inhibitor theoretically would provide additional bene¢t, but have not been systematically studied. Immunotherapy for venom-sensitive subjects, desensitization for certain cases of drug allergy, anti-IgE therapy for subjects at risk of food-induced anaphylaxis and cyclosporin A for recurrent anaphylaxis are considerations. Glucocorticosteroids, which do not inhibit experimental mast cell activation, are unlikely to provide a major bene¢t in most patients with recurrent anaphylaxis. More e¡ective and long-lasting therapies for IgE- and non-IgE-mediated anaphylaxis are still needed.

References

Chandrasekharan NV, Dai H, Roos KL et al 2002 COX-3, a cyclooxygenase-1 variant inhibited by acetaminophen and other analgesicffiantipyretic drugs: cloning, structure, and expression. Proc Natl Acad Sci USA 99:13926^13931

Evans JF 2002 Cysteinyl leukotriene receptors. Prostaglandins Other Lipid Mediat 68^69:587^  Fricker M, Helbling A, Schwartz L, Muller U 1997 Hymenoptera sting anaphylaxis and urticaria pigmentosa: clinical ¢ndings and results of venom immunotherapy in ten patients. J Allergy Clin Immunol 100:11^15

Guida M, Riedy M, Lee D, Hall J 2000 Characterization of two highly polymorphic human tryptase loci and comparison with a newly discovered monkey tryptase ortholog.

Pharmacogenetics 10:389^396

Hirai H, Tanaka K, Yoshie O et al 2001 Prostaglandin D2 selectively induces chemotaxis in T helper type 2 cells, eosinophils, and basophils via seven-transmembrane receptor CRTH2. J Exp Med 193:255^261

Huang C, Li L, Krilis SA et al 1999 Human tryptases alpha and betaffiII are functionally distinct due, in part, to a single amino acid di¡erence in one of the surface loops that forms the substrate-binding cleft. J Biol Chem 274:19670^19676

Ludolph-Hauser D, Rue¡ F, Fries C, Schopf P, Przybilla B 2001 Constitutively raised serum concentrations of mast-cell tryptase and severe anaphylactic reactions to Hymenoptera stings. Lancet 357:361^362

Marquardt U, Zettl F, Huber R, Bode W, Sommerho¡ C 2002 The crystal structure of human alpha1-tryptase reveals a blocked substrate-binding region. J Mol Biol 321:491^502

Neugut AI, Ghatak AT, Miller RL 2001 Anaphylaxis in the United States: an investigation into its epidemiology. Arch Intern Med 161:15^21

Sakai K, Ren S, Schwartz LB 1996 A novel heparin-dependent processing pathway for human tryptase. Autocatalysis followed by activation with dipeptidyl peptidase I. J Clin Invest 97:988^995

Schwartz LB 2001 Clinical utility of tryptase levels in systemic mastocytosis and associated hematologic disorders. Leuk Res 25:553^562

Schwartz LB 2002 Mast cells and basophils. Clin Allergy Immunol 16:3^42

Schwartz LB, Metcalfe DD, Miller JS, Earl H, Sullivan T 1987 Tryptase levels as an indicator of mast-cell activation in systemic anaphylaxis and mastocytosis. N Engl J Med 316:1622^1626 Schwartz LB, Yunginger JW, Miller JS, Bokhari R, Dull D 1989 The time course of appearance and disappearance of human mast cell tryptase in the circulation after anaphylaxis. J Clin Invest 83:1551^1555

Schwartz LB, Bradford TR, Rouse C et al 1994 Development of a new, more sensitive immunoassay for human tryptase: use in systemic anaphylaxis. J Clin Immunol 14:190^204

Schwartz LB, Sakai K, Bradford TR et al 1995 The a form of human tryptase is the predominant type present in blood at baseline in normal subjects and is elevated in those with systemic mastocytosis. J Clin Invest 96:2702^2710

Schwartz LB, Min H, Ren S et al 2003 Tryptase precursors are preferentially and spontaneously released, while mature tryptase is retained by HMC-1 cells, monomac-6 cells and human skin-

derived mast cells. J Immunol 170:5667^5673

Selwood T, Wang ZM, McCaslin DR, Schechter NM 2002 Diverse stability and catalytic properties of human tryptase alpha and beta isoforms are mediated by residue di¡erences at the S1 pocket. Biochemistry 41:3329^3340

Soto D, Malmsten C, Blount JL, Muilenburg DJ, Caughey GH 2002 Genetic de¢ciency of human mast cell a-tryptase. Clin Exp Allergy 32:1000^1006

Valent P, Horny H-P, Escribano L et al 2001 Diagnostic criteria and classi¢cation of mastocytosis: a consensus proposal. Leuk Res 25:603^625

van der Linden P-WG, Hack CE, Poortman J, Vivie¤ -Kipp YC, Struyvenberg A, van der Zwan JK 1992 Insect-sting challenge in 138 patients: relation between clinical severity of anaphylaxis and mast cell activation. J Allergy Clin Immunol 90:110^118

Wedemeyer J, Tsai M, Galli SJ 2000 Roles of mast cells and basophils in innate and acquired immunity. Curr Opin Immunol 12:624^631

Wolters PJ, Pham CTN, Muilenburg DJ, Ley TJ, Caughey GH 2001 Dipeptidyl peptidase I is essential for activation of mast cell chymases, but not tryptases, in mice. J Biol Chem 276:18551^18556

DISCUSSION

Galli: How many instances have you observed in which good specimens of blood had been obtained and yet there was evidence for histamine release without elevated tryptase? You showed details about one case.

Schwartz: We haven’t done a systematic study of this. You never know when you are going to get that. In some of the fatal or near-fatal food anaphylactic patients that Hugh Sampson has reported, assays were performed in which there was no elevation of total or mature tryptase. There are a number of potential explanations. Some of these patients had prolonged hypotension. It is not clear to me whether the same pathogenesis that is involved in the initial event is involved in the persistence of that event. Another possibility is that different groups of mast cells may get activated through di¡erent routes of allergen administration. Intestinal mast cells may not have the same quantity of tryptase per cell as the mast cells that are in the skin and around the vasculature. They conceivably could be more distant from blood vessels. Although I still favour the possibility that mast cells may not be involved in some of those reactions, perhaps other cell types have to be considered.

Vercelli: What do you mean when you say that a signi¢cant proportion of a¡ected individuals don’t have the a tryptase gene?

Schwartz: There is a tryptase locus. Each haploid chromosome 16 has two affib tryptase genes (one or two b tryptase genes and one or zero a tryptase genes). About 25% of individuals have all b tryptase genes and no a tryptase gene.

Vercelli: So it is not as if they have a truncated gene that may give rise to some functional dominant negative, or anything like that. Do we know that the gene is not there at all?

Schwartz: There is allelic variation. One can have one of the b tryptase subtypes or of the a tryptase subtypes at one site, and at the other site just the b tryptases. There are also polymorphisms among both a and b tryptases that have been documented.

Austen: Larry Schwartz, could you say more about the number of tryptase genes? How many do you think there are, and what are the alleles? What about the relevance of the g and d tryptase?

Schwartz: If we look at a peptide phylogram showing the relationships between tryptases, there are two principal a tryptase forms that vary by one or two amino acids, and three principal forms of b tryptase which vary by one to three amino acids. a and b tryptases are more closely related to one another than they are to any of the rodent or other animal tryptases. George Caughey has performed the best work on this. About 25% of people have no a tryptase gene in either of their chromosome 16s. There is also a tryptase d, a tryptase g and a tryptase e. We didn’t originally describe the d, but we published a paper saying that this was probably a pseudogene (Min et al 2001). It is truncated 40 amino acids earlier than are affib tryptases. We detected little to no mRNA for it in mast cells. Subsequent to this there has been a very nice paper from an Australian group which does detect the mRNA in mast cells and possibly other cell types and shows that the protein is also expressed in these cells (Wang et al 2002). Because of a stop codon it would be about 5^10 000 Da less than tryptase. In Western blots with our anti-tryptase mAbs, we don’t see a product of that size in mast cells. Either our antibodies aren’t detecting it, or the amount present compared to affib tryptases is very low. Tryptase g has been described by both Rick Stevens and George Caughey. It has a transmembrane portion to it. It is biochemically and immunologically quite distinct from affib tryptases. From what I recall, tryptase e also was a little bit less similar to the affib tryptases.

Austen: There are papers claiming that in systemic mastocytosis there are some patients who do not have elevated tryptase. George Caughey and colleagues have provided substantial evidence for a null humans (Soto et al 2002). Does this account for tryptase-negative mastocytosis patients?

Schwartz: At the recent International Mastocytosis meeting, this issue wasn’t raised. When I spoke to Peter Valent and Luis Escribano at the meeting in Vienna, they told me that they didn’t have anyone with systemic mastocytosis without an elevated tryptase level. At that meeting, a total tryptase level greater than 20 ngffiml was adopted as one of the diagnostic criteria. I think those patients without an elevated tryptase that are said to have systemic mastocytosis need to be looked at carefully and reported. One possibility would be that they are a tryptase de¢cient and for some reason don’t secrete enough tryptase. However, our results with healthy subjects, show that total tryptase levels are una¡ected by the presence or absence of the a tryptase gene. Whether there is an e¡ect in mastocytosis is being examined. Another possibility is that early in the course of systemic mastocytosis, when the mast cell burden is still low, the total tryptase level would be normal. The more distantly related tryptic-like enzymes that have been named tryptase, e.g. NK cell tryptase and clara cell tryptase are unrelated to the affib tryptases of mast cells.

Simons: With regard to the normal values for total tryptase of 1^15 ngffiml, were young children included in the population from which these norms were derived?

Schwartz: Young children haven’t been looked at systematically. With adults we’ve examined 100 to 200 subjects on at least two occasions. The mean and two standard deviations gives a range of about 2^12 ngffiml. I have taken three standard deviations to go from 1 to 15 ngffiml, because we have seen some individuals who appear healthy in that upper range, so I arbitrarily decided to extend it to three standard deviations.

Simons: My point is that for so many normal values there are age-related di¡erences between infants and prepubertal children versus teens and adults. This might be an interesting line of investigation to pursue.

Schwartz: I agree.

Finkelman: We have some mouse data which are possibly relevant to the mast cell versus basophil issue in anaphylaxis. We have looked at two phenomena: the development of anaphylaxis as assessed by changes in behaviour and hypothermia, versus increase in in vivo production of Th2 cytokines, following anti-IgE treatment of either wild-type or c-Kit de¢cient mice. There is a pretty strong dichotomy between the two phenomena. When c-Kit de¢cient mice are treated with anti-IgE they shown no change in behaviour and don’t develop hypothermia. Nor do they show an increase in MMCP1 or histamine levels. Yet they show as great an increase in IL4 production as do wild-type mice. In the mouse there is a relatively simple situation: the only two cell types that are thought to have FceRI are mast cells and basophils. So these responses are coming from one of these two cell types. If one were to ¢nd that mast cells require c-Kit for development and the basophils don’t, then one could say that the anaphylaxis is not coming from basophils. Again, I don’t know whether the mouse situation will be pertinent to the human, but it creates a bias in my mind.

Schwartz: Certainly within those experimental circumstances, I think your interpretation is reasonable. Mice normally seem to have very few basophils. If you did something to induce higher levels of basophils in your c-Kit-defective animals - such as using certain immunization techniques - and see what happens then, this may be revealing.

Finkelman: These experiments were done in mice that had been immunized in a way that causes a Th2 response. This results in the production of a large number of cells that are variously described as basophils or immature basophils.

Galli: About ¢ve years ago, Choi et al (1998) reported that active fatal anaphylaxis could be elicited to penicillin V in mast cell-de¢cient WffiWv mice and that this response was IgE-dependent. Based on a pharmacological approach, the authors attributed the reaction to PAF production, presumably by basophils. Although there was no formal proof that basophils were the origin of the PAF, the authors provided evidence that this form of active anaphylaxis is an IgE- dependent response (Choi et al 1998, Park et al 1997). So there has been at least that one reported example of an IgE-dependent anaphylactic reaction in the mast cell-de¢cient mouse.

Finkelman: I would argue with you about that paper. I don’t think there was good evidence that it was IgE mediated.

Galli: If it was not IgE-mediated, then that may be why the reaction could be elicited in WffiWv mice.

Fisher: In one of the studies of the mast cell tryptase postmortem, there were three patients, I think, where the diagnosis was myocardial infarction. Have you any explanation for that?

Schwartz: That’s a good point. There have been a couple of interesting studies where elevated b tryptase levels are found without any evidence for IgE-mediated anaphylaxis. Some of these are in postmortem samples in adults, associated with trauma. About 1 in 20^25 knife murders or shootings end up having markedly elevated tryptase. There have also been some hospital postmortem cases where elevated levels were thought by the medical examiner not to be associated with an IgE-mediated reaction. But many of these patients received a variety of drugs close to the terminal event, so it is hard to sort this out. Then there is a series of sudden infant death syndrome patients, where about 40% of these infants had postmortem tryptase levels that were elevated compared with explained death controls (Platt et al 1994). This has been more or less reproduced in two out of three subsequent studies. All groups agree that there is no antigen-speci¢c IgE detectable in blood. There may be some non-IgE-dependent ways of activating mast cells, and these need to be better understood.

Fisher: One of the things that interested us about the cardiac cases is that in the last eight cases of anaphylaxis we have seen have had elevated troponins when they were measured in four. This is supposed to be very speci¢c for myocardial ischaemia, yet there has been no evidence of myocardial ischaemia in these patients.

Schwartz: This sounds fascinating.

Sampson: Could you comment on the paper by Lin et al (2000) in which they looked at 92 patients coming into the emergency room with anaphylaxis? Only 21% had elevated tryptase.

Schwartz: Those elevations were all near the limit of detection. Almost none of those patients were hypotensive. Clinically these were not severe episodes. I am not sure that those results were clinically signi¢cant.

Golden: With respect to insect sting anaphylaxis and mastocytosis, you commented on the absence of IgE. That raises an interesting question. Is it anaphylaxis or mastocytosis? Can they coexist if there is no IgE? If I remember Ulrich Mˇller’s reports correctly, in virtually all cases of insect sting anaphylaxis with mastocytosis there is detectable IgE. My question on the case that you mentioned, is what laboratory did the assays? Could there have been a very low level of IgE that might have been detected in a sensitive assay?

Mˇller: In the paper you are mentioning (Fricker et al 1997) we describe 10 cases with urticaria pigmentosa and anaphylaxis to Hymenoptera stings. Eight had elevated basal serum tryptase levels. In two of the patients we had both negative i.c. skin tests at 1 gffiml and in ¢ve no venom speci¢c IgE was detectable by Phadezym RAST. Failure to detect speci¢c IgE in mastocytosis patients in spite of positive skin tests could be explained by absorption of most of the speci¢c IgE to the abundant mast cells.

Schwartz: With a higher sink for IgE, it is conceivable you could have mast cells that have become sensitized. In such a scenario, you could have a positive skin test and a negative assay in the serum.

Golden: Since the redescription in the past two years of sting anaphylaxis with negative skin tests the situation has gotten better. In 90^95% of those cases with negative skin tests we were able to detect at least a trace of venom-speci¢c IgE. It is really hard to ¢nd someone with insect allergy but without venom-speci¢c IgE.

Galli: Ulrich Mˇller, did you interpret your cases as probably being independent of IgE?

Mˇller: We thought of direct mediator release by venom components such as melittin, MCD-peptide or mastoparan.

Schwartz: This is a distinct possibility.

Ring: Perhaps 5^10% of our patients have negative skin tests. In these cases we then look at histamine or release from basophils. When we get a positive result we know they are sensitized, even though we have not detected venom-speci¢c IgE. It may be the basophil that is responsible. You showed us a lot about mast cells. Do we have a marker or basophil product similar to tryptase?

Schwartz: No. There have been two basophil-speci¢c antibodies. Both recognize a granule component, but in neither case has the antigen been identi¢ed.

Galli: So your point, Johannes Ring, is that in the skin test and radioallergosorbent test (RAST)-negative patients with anaphylaxis, a large proportion give you in vitro basophil histamine release in response to allergen.

Ring: Yes. We do this routinely. If they are negative in skin test and RAST, then we look at histamine release. Very few of them are still negative, yet they have a very severe reaction.

Schwartz: I take it that normal basophils won’t respond to venom.

Ring: That is correct, in the respective doses.

Ohtsu: You showed that tryptase levels are very high in systemic mastocytosis. Theoretically, if there is a lot of tryptase in the blood, tryptase receptors should be down-regulated. Are there any tryptase receptors?

Schwartz: It is not clear that there are. Because tryptase is a protease, it would have substrates that could be on a cell’s surface. What these biological substrates might be is not clearly resolved.

Ohtsu: Even so, I would expect that there would be some sort of diminished response with time to these elevated tryptase levels. Is there any evidence in these mastocytosis patients that the occurrence of systemic anaphylaxis reaction is di⁄cult.

Schwartz: What likely is being detected in the mastocytosis serum is the pro- and

pro’- forms of tryptase, which are enzymatically inactive. Those forms of tryptase would not be able to carry out the enzymatic function(s) of the mature enzyme.

References

Choi IH, Shin YM, Park JS et al 1998 Immunoglobulin E-dependent active fatal anaphylaxis in mast cell-de¢cient mice. J Exp Med 188:1587^1592

Fricker M, Helbling A, Schwartz L, Mˇller U 1997 Hymenoptera sting anaphylaxis and urticaria pigmentosa: clinical ¢ndings and results of venom immunotherapy in 10 patients. J Allergy Clin Immunol 100:11^15

Lin RY, Schwartz LB, Curry A et al 2000 Histamine and tryptase levels in patients with acute allergic reactions: an emergency department-based study. J Allergy Clin Immun 106:65^71

Min HK, Kambe N, Schwartz LB 2001 Human mouse mast cell protease 7-like tryptase genes are pseudogenes. J Allergy Clin Immunol 107:315^321

Park JS, Choi IH, Lee DG et al 1997 Anti-IL-4 monoclonal antibody prevents antibiotics-

induced active fatal anaphylaxis. J Immunol 158:5002^5006

Platt MS, Yunginger JW, Sekula-Perlman A et al 1994 Involvement of mast cells in sudden infant death syndrome. J Allergy Clin Immunol 94:250^256

Soto D, Malmsten C, Blount JL, Muilenburg DJ, Caughey GH 2002 Genetic de¢ciency of human mast cell alpha-tryptase. Clin Exp Allergy 32:1000^1006

Wang HW, McNeil HP, Husain A et al 2002 Delta tryptase is expressed in multiple human tissues, and a recombinant form has proteolytic activity. J Immunol 169:5145^5152