Anaphylaxis --- a brief introduction

This web page assumes a small amount of knowledge of antibodies, cytokines, T and B cells, basophils, mast cells and genes. With this knowledge in mind, read on and enjoy! After an initial overview of anaphylaxis, we have three main sections, firstly, a discussion of IgE, secondly a look at the basophil and mast cell and what they produce, and lastly a note on modulation of the system. Here's an accompanying powerpoint presentation (650K).

An overview

Of the nine different immunoglobulins (Ig) in man --- IgM, IgG1 to IgG4, IgE, IgA1 and IgA2, and IgD --- only IgE is associated with atopy and anaphylaxis. It's a very special molecule.

Production of the immunoglobulin IgE is a potential cause of major illness (and even death) in man, but has persisted over the ages, probably because it provides a potent defence against parasitic infestation. Anaphylaxis is a term coined by Richet and Portier in 1902 as a direct challenge to the prevailing dogma that the immune system was always a good thing which guarded against attack (phylaxis). Asked by Prince Albert the First of Monaco to investigate the toxicity of jellyfish stings, they showed that repeated challenge with minuscule amounts of antigen, far from guarding against disease, sometimes induces life-threatening reactions.

IgE is central to human anaphylaxis by virtue of its binding to mast cells and basophils. When these cells are activated by binding of antigen to the IgE, they release of an alphabet soup of active substances which may cause hypotension, bronchospasm, urticaria and, in some unfortunate individuals, circulatory collapse and death. The main treatment of this catastrophe is administration of adrenaline (epinephrine) intramuscularly or intravenously. Suitable IM doses of adrenaline are 0.3 mg (or 0.15 mg in children); intravenous doses can be about a third of this dose (0.1 mg). Administration of single larger IV doses may well be harmful to the patient.

We cannot be too dogmatic about anaphylaxis, especially as we don't understand it too well. Much of the confusion has arisen because of arbitrary decisions about labelling. For example, although anaphylaxis is 'classically' said to be characterised by the triad of hypotension, bronchospasm or uticaria, we may encounter the patient who only develops hypotension, without either bronchospasm or urticaria. Only the 'clincher' of a massively raised serum tryptase may demonstrate that the cause of the hypotension was anaphylaxis! There's much confusion about alleged differences between anaphylactic and 'anaphylactoid' reactions (with recent recommendations that we abandon the latter term).

There's another problem, that of grading the severity of anaphylaxis, with multiple systems being prevalent. Brown's recent work may help to dispel a lot of the confusion. [Brown SGA, J Allergy Clin Immunol 2004 114 371-6]. He recognises just three grades --- mild, moderate, or severe, and provides justification for his grading by using logistic regression analysis.

Our tests for the presence of anaphylaxis are rather primitive and rudimentary, and are, as is so common in medicine, far from 100% reliable. Skin testing is still our somewhat debased gold standard, although some have claimed that newer tests are of substantial value. Newer tests generally involve exposing basophils to the suspected offending antigen, and looking at their response using flow cytometry to identify specific markers on the basophil cell surface. For example, Sudheer et al. note sensitivity of 64% for skin-prick testing, and 79% for flow cytometry of CD63, using tryptase as a marker for peri-anaesthetic anaphylaxis [Anaesthesia. 2005 Mar;60(3):251-60].

In the following sections, we will explore a little of what is known about IgE mediated disease. A lot of the available literature is highly technical, densely written, and made more confusing due to a proliferation of abbreviations, and the frequent use of multiple names for the same thing. Let's try to wade through the morass.

A. Anaphylaxis starts with IgE

First let's explore IgE a little!

In anaphylaxis, the mast cell seems to be the main player (The basophil is similar to the mast cell, and may play a role). We may learn a lot by carefully examining 'atopic' processes which also involve IgE and the mast cell, processes such as asthma, and allergic rhinitis. In fact, anaphylactic reactions seem to be more common in atopic individuals, with the exception of reactions to insect stings and possibly to penicillin.

The word 'atopy' was coined by Coca and Cooke in the 1920s --- it translates as 'without a place', that is, a strange disorder which doesn't fit in! Coca's problem was that despite the immune reaction that was obviously occurring, no antibodies could be identified by mechanisms available to him. The mysterious hidden factor which conferred hypersensitivity was called reagin, and only much later was this identified (by the husband-and-wife Ishizaka team in 1965) and called IgE. IgE is present in such small amounts in blood that it is indeed difficult to identify (normally about 150 nanograms/ml, as opposed to about 10mg/ml for IgG).

In anaphylaxis there seems to be the unfortunate collision of genetic predisposition and an environmental allergen to which the person is exposed. The individual develops an immune response to the allergen (antigen), and for reasons that are just becoming apparent, produces substantial amounts of IgE. IgE has an extraordinary affinity for mast cells, and by avidly binding even prolongs the lifespan of the cell!

Re-exposure to the offending antigen then causes cross-linking of pairs of IgE molecules, with devastating consequence. But we're getting ahead of ourselves. First, let's look at the circumstances surrounding IgE production. The process is complex, but here's a summary:

  1. A B lymphocyte 'tuned' to the specific antigen meets up with that antigen. The B lymphocyte acts as an antigen-presenting cell(APC). It processes the antigen, and presents a small fragment, suitably packaged, to a T lymphocyte;
  2. The T cell then helps the B lymphocyte to proliferate;
  3. A switch occurs in the B lymphocyte clone, resulting in production of IgE;
  4. Terminal differentiation occurs with formation of plasma cells, but some B lymphocytes remain as memory cells;

1. Antigen-presenting cells (APCs)

The simplest APC is simply a 'naive' B lymphocyte. There are over ten million (and perhaps a billion) different clones of B lymphocytes within the body, each with its own peculiar immunoglobulin receptors on the cell surface. When a member of such a clone encounters a matching, specific antigen, it won't proliferate unless there's a second signal, for example the complement component C3d binds a receptor on the surface of the lymphocyte. (This is a safety mechanism which helps to protect the body from turning on itself; if the second signal doesn't occur, the cell usually undergoes apoptosis, which can be seen as a sort of ritual cellular suicide)! T4 lymphocytes then promote proliferation of the B cell. A key component of the response of the B cell is processing of antigens into small peptide fragments which are presented on the surface of the B cell as 'peptide epitopes'.

(Some antigens, for example large carbohydrate and lipid molecules, do not require interaction with T cells for a vigorous B-cell response to occur. Such responses can occur via Toll-like receptors, or simultaneous cross-linking of multiple B-cell receptors).

There seems to be a degree of confusion as to which 'antigen presenting cell' does what, but despite the fact that APCs other than B cells largely present antigens to T cells and stimulate a T-cell response, they also seem to be involved in atopy! The main involvement of APCs other than the B cell in initiating atopy is possibly a negative one. If these cells are active, then they may promote suppression of allergy; if they are inactive, then allergy may be allowed to emerge. However, on repeat presentation of an antigen, APCs seem to be heavily involved in promoting the IgE response! It's all rather complex.

2. Lymphocyte proliferation / apoptosis

A B cell which encounters an antigen has two options: proliferate or die. If the second signal we mentioned above isn't forthcoming, then the B cell turns up its toes and dies. Otherwise...

Somatic hypermutation (SHM)

Once B cells have encountered an antigen they enter a process of somatic hypermutation. Here, point mutations occur in the 'V' region of DNA, resulting in production of a variety of similar antibodies of varying affinity for a particular antigen. Cells with antibodies which have a greater affinity for the antigen seem to survive, while less capable cells vanish due to apoptosis ("clonal selection"). These 'fine-tuned' cells often produce antibodies that have a tenfold increase in affinity for the desired antigen.

There are many different V genes, and some of these encode for the V region of the heavy (H) chain of IgE. These "VH-Cε" sequences have been found to differ in atopic and non-atopic individuals --- a particular subset of VH genes called "VH5" has been found to be over-represented in people with atopy, particularly in those with severe asthma.

Mechanisms of lympocyte survival and proliferation

Once it's started, complex mechanisms seem to keep IgE production on the go, even in the absence of antigen! These mechanisms may involve innate properties of the cells, as well as presence of factors that influence lymphocyte survival and differentiation. We discuss two of these factors below --- IL-4 and sCD23.

3. The switch to IgE (isotype switching)

Although B cells initially are programmed to produce IgM, later on they will produce either IgG, or IgE. A key question in allergy is therefore "Why is IgE preferentially produced in atopic individuals?" We still don't know the full answer to this question, but we have filled in some of the details along the way.

Initial production of IgE doesn't just happen. It requires an orchestrated series of events. Initially, B cells carry IgM molecules on their cell surface (are 'IgM positive'), and in order for IgE to be produced, the genome of the B cell needs to reorganised! This reorganisation involves movement of a whole 'cassette' of genes (VHDJH) next to the genes coding for the constant part of the IgE heavy chain. Two important events necessary for moving about of this cassette are:

  1. stimulation of the B cell by the cytokines IL-4 or IL-13;
  2. surface interaction with T cells bearing something called CD40L.

Each of the above is complex. Exposing the B cell to cytokines does half the job, by stimulating transcription of genes which code for the IgE heavy chain (its constant part, that is). The rest of the task only starts after a specific interaction with the T cell --- knowing that the term CD40L, the vital protein sticking out of the surface of the T cell, is shorthand for "CD40 ligand", we can guess that the specific receptor on the B cell is --- wait for it --- CD40.


This final step of lymphocyte interaction results in the gene cassette being moved around so that IgE can actually be formed. Not content with this complexity, experts have come up with further fancy names for the process --- for example, the gene movement is sometimes called "deletional switch recombination", but more commonly rejoices under the name class switch recombination or 'CSR'. The stimulation of IgE heavy chain formation is referred to as "epsilon germline gene transcription"! Central to the mechanism of CSR seems to be "activation-induced cytidine deaminase" (AID), normally seen only in B cells.

The Th2 phenotype

IL-4 is primarily produced by T helper cells. Let's look at this production..

Intimately associated with atopy and production of IgE are particular characteristics of helper T cells. Helper T cells are otherwise known as CD4 positive cells because they carry this antigen on their surface; they are the CD4 cells devastated by the AIDS virus. Some individuals have CD4+ lymphocytes with the 'normal' "Th1 phenotype", but allergic individuals have CD4+ lymphocytes with peculiar characteristics labelled "the Th2 phenotype". Let's explore this, because it's important.

From the above description of how IgE is formed, you might guess that Th2 cells are responsible for production of lots of IL-4 and IL-13. You might also speculate that these cells are laden with CD40L. You'd be right in both cases. (A few other cells can also produce some IL-4, including mast cells themselves, basophils, and NK1.1+ cells).

Before we blame the Th2 cell for production of IgE however, we must realise that the formation of Th1 or Th2 cells is itself determined by specific signals. Such signals orchestrating T cell behaviour come from antigen-presenting cells (APCs: macrophages, dendritic cells, and the odd B cell). The key issue seems to be whether the APC produces interleukin-12 or not. In the absence of IL-12, T cells default to a Th2 phenotype; if IL-12 is present, then Th cells are directed into the Th1 pathway. The Th1 phenotype is responsible for lots of production of interferon gamma, which in turn suppresses the Th2 phenotype! Conversely, a vigorous Th2 response inhibits production of Th1 cytokines (this suppression of the Th1 response by IL4 is possible because T cells themselves carry IL4 receptors).

Some antigens seem to be particularly good at eliciting a Th1 response (e.g. lipopolysaccharide and some oligonucleotides); others such as double-stranded viral RNA and schistosomal egg antigens seem to favour Th2. It's reasonable to expect the innate immune system (especially Toll-like receptors) to be involved in such 'decision making'!

Many common allergens have protease activity, which appears to modify receptors such as CD23 and the IL-2 receptor; these changes in turn apparently promote a Th2 phenotype and increase IgE production. (Details are complex).

There is even some evidence that the basophil (which contains tiny amounts of pre-formed IL-4 in its granules, at least in the mouse) may shift the immune system towards a Th2 phenotype.

4. Terminal differentiation to plasma cells

A plasma cell is a little immunoglobulin factory --- up to half of the protein synthesis in such committed cells is immunoglobulin. As with all immune processes, regulation of plasma cell formation is insanely complex (but now broadly understood). Antigen binding to a B cell results in destruction of a key inhibitor protein called BCL-6. In the absence of BCL-6 there's substantial production of a protein quaintly called "Blimp-1" which brings about formation of a plasma cell. It's like a molecular switch, because a lot of Blimp-1 floating around turns off further production of BCL-6. In fact, a remarkable number of genes get turned off in the process of making a plasma cell.

Memory cells

B cells can also turn into memory cells, with a little help from IL-10 in particular. Memory cells, as their name suggests, hang around for a loooong time, preserving the response to a particular antigen. Their formation is promoted by IL-10 (a Th2 cytokine), and they survive with a little help from CD40 and CD23.

B. Basophils and mast cells

We now understand a little more about IgE and its formation, but how does it do the damage? We need to examine the basophil, and its 'equivalent' in tissues other than blood, the mast cell, as these produce and release the mediators that actually do the damage. The main mediator is of course histamine.

In this section we will examine:

  1. Basophils in blood - a general note;
  2. The mast cell;
  3. The FcεRI receptor;
  4. The FcεRII receptor;
  5. IgE receptors on antigen-presenting cells;
  6. What the mast cell releases;
  7. The late phase of anaphylaxis which occurs in about 10% of cases.

1. Basophils in the blood

Paul Ehrlich first described the basophil in 1879, but there's still an enormous amount we don't know about this fascinating cell. Despite constituting under one percent of circulating leukocytes, the basophils may be important immune cells. On cross-linking of the IgE receptors bound to their surface by FcεRI, they release the contents of the basophilic granules from which they derive their name. There's been a lot of recent speculation that the basophil may be important in other processes as well.

It seems that normal basophils contain histamine but insignificant amounts of tryptase. From the fact that tryptase levels go screaming up during many, perhaps most cases of anaphylaxis, we can deduce that in such cases there is prominent degranulation of mast cells --- whether the blood basophil plays a role seems most uncertain.

We're not even sure where basophils come from, although they appear to arise from precursor cells found both in the blood and in bone marrow. (These precursors, which carry the antigen CD34, probably also give rise to monocytes, neutrophils and eosinophils, but some seem to become committed to form mast cells or basophils). The main growth factor for basophils appears to be IL-3. In contrast, mast cells seem to depend largely on 'stem cell factor'; basophils and mast cells, which can otherwise look very similar, can be distinguished by the presence of receptors for one or other of these factors on their surfaces.

2. The mast cell

Ehrlich described the mast cell at about the same time that he described the basophil. He called them 'mastzellen' or "well-fed cells", and unaware of their role in allergy, speculated that they might have a nutritive role. There seem to be two distinct populations of mast cells, those found in connective tissue, and those found in mucosa. We're more interested in the latter, as they seem to be the ones concerned with atopy and anaphylaxis. They are packed with metachromatic granules (up to a thousand per cell), which on release ('degranulation') mediate allergic reactions.

Each mast cell is covered in about 200 000 FcεRI receptors. When an antigen meets up with two IgE molecules, each bound to an FcεRI receptor, all hell breaks loose as the mast cell degranulates. Such 'cross-linking' of two bound IgE molecules is essential for mast cell activation.

It's interesting to note that maximal sensitisation of the mast cell occurs when under 10% of FcεRI receptors are occupied by IgE. Before we look at degranulation, let's examine this important receptor in a little more detail.

3. FcεRI

FcεRI binds the Fc fragment of the IgE molecule with great avidity --- there's also some binding of IgE to another receptor called FcεRII, but the latter has a lower affinity for IgE. FcεRI really is remarkable in its capacity to bind IgE --- the 'association constant' (a measure of how tightly IgE binds) is about 10-10M-1, about a thousand times more avid than many other receptor/ligand bindings.

FcεRI is mainly found on mast cells and basophils, but is also seen on the Langerhans cell, as well as on monocytes, platelets and eosinophils.

FcεRI is made up of several individual components --- necessary are an alpha and two gamma chains, but on mast cells and basophils the receptor has an added beta chain. The beta chain strengthens signal transduction, and helps move the receptor to the cell surface.

Lowered levels of IgE result in decreased expression of FcεRI. This isn't surprising, as IgE stabilises the receptor on the cell surface.

A few asides

The literature about FcεRI has been made more confusing by use of terms like ITAM (for "immunoreceptor tyrosine-based activation motif"), which refers to the structure of the receptor, and the immune activation which results from stimulating it. ITAM should not be confused with ITIM ("immunoreceptor tyrosine-based inhibition motif"), which refers to the ability of the IgG receptor (FcγRIIb) to inhibit allergic reactions.

Artificial modulation of T-cell responses

A monoclonal anti-IgE (omalizumab) has been created that appears to down-regulate FcεRI expression on basophils and dendritic cells. This may have clinical relevance in the management of allergic asthma and rhinitis, and even (perhaps) in peanut allergy. Cost may be a significant issue, and numbers reported so far are relatively small, with a lot of post-hoc analyses in the various studies. The antibody has been carefully engineered so that the complex is biologically inactive, non-immunogenic, and doesn't cross link mast-cell bound IgE. [Poole, Matangkasombut & Rosenwasser: J Allergy Clin Immunol 115(3) S375-385, March 2005]

Blockade of IgE/FcεRI interaction

Another possible way to modulate IgE mediated responses is to use humanised antibodies ("GE2") that cross-link FcεRI and FcγRIIb receptors on the surface of mast cells, decreasing mediator release on antigenic stimulation. [Saxon A et al, Current Opinion in Allergy and Clinical Immunology, 2004 4:563-568] This product also appears to work on basophils, mast cells and "Langerhans-like dendritic cells", inhibiting 'proallergic' responses. Similar responses may be a mechanism by which 'desensitisation' immunotherapy works.

4. FcεRII (CD23)

This receptor is commonly called CD23, and binds IgE somewhat less avidly than FcεRI. It's not a member of the immunoglobulin superfamily, and is related to adhesion molecules and the like from the "C-type lectin superfamily". Other things apart from IgE (such as some complement receptors and vitronectin, CD11b and CD11c) bind to CD23. Binding depends on calcium ions and is helped by a 'co-receptor' called CD21, which itself can bind to many different ligands (including some complement components, Ebstein Barr virus, and alpha interferon). Binding of IgE to CD23 may down-regulate IgE production.

CD23 also facilitates antigen presentation (see below). There are two forms of CD23, called CD23a (on B cells), and CD23b (on other cells expressing CD23).

Soluble CD23 (sCD23) has a variety of actions, including up-regulation of IgE synthesis!

5. What do IgE receptors do on antigen-presenting cells (APCs)?

APCs definitely express and use FcεRI. Such APCs include Langerhans cells, moncytes, and dendritic cells --- the presence of antigen-specific IgE increases the efficiency with which they present antigens to T cells by about a thousand fold! Dendritic cells in peripheral blood are apparently far more efficient at such antigen presentation than are monocytes.

The end result of such enhancement is that subsequent IgE responses will be massively amplified the second time an antigen is encountered. APCs can even produce cytokines and eicosanoids themselves!

CD23 seems to be important in both presentation of antigens and feedback regulation of IgE synthesis (Mice lacking CD23 produce large amounts of IgE). Soluble CD23 promotes differentiation of immune cells, notably of B cells into plasma cells, but also promyelocytes into basophils.

6. What the mast cell releases

The mast cell contains a lot of nasty substances in its granules. On activation, mast cells release:

Cellular events which occur between binding of antigen and mast cell degranulation are very complex. A common mediator has however been identified, called PI 3-kinase. Blocking of this enzyme appears to completely abrogate mast cell degranulation but we haven't yet identified clinically useful blockers. (The laboratory ones are wortmannin and LY-294003).

7. The late-phase reaction

In about ten percent of acute IgE mediated reactions, there is a second, later response. This occurs many hours after the initial response, and seems to be eosinophil-dependent. This reaction has been particularly well documented in asthmatics, but is also common following anaphylaxis. [J Allergy Clin Immunol. 2002 Sep;110(3):341-8] Biphasic (or refractory) anaphylaxis cannot be predicted from the severity of the initial phase, mandating cautious observation of subjects for 24 hours following the initial phase.

The chemokine eotaxin seems to lure the eosinophil into areas of allergic inflammation when it binds to the 'CCR3' receptor on their surface. Basophils also express CCR3, so it seems quite reasonable to implicate them in late phase allergic reactions as well! (Other eosinophil-recruiting chemokines also seem to elicit a response from basophils).

C. Control and modulation of atopy

The preceding sections may have lulled you into a sense of false security, that at least somebody understands what is going on (even if it's neither you nor me)! This is far from the truth, as most of the systems controlling atopy haven't yet been worked out. Several linked systems probably control the expression of atopy, and whether an individual's lymphocytes have a Th1 or Th2 phenotype. Two important players are the innate immune system, and dendritic cells. We've also begun to appreciate that there are several different controlling pathways, including an 'alternative pathway' to IgE (Which is unrelated to the 'alternative' complement pathway, despite the similar name).

1. The innate immune system

Although everyone gets really excited about antibodies and the like, recent evidence underlines the importance of the innate immune system in the body's response to, well, almost anything you care to mention! We now know that this response is specific to the invader attacking the body. We know very little else about the innate immune system, although details are starting to emerge.

Central to the functioning of the innate immune system is identifying 'strangers' despite never having encountered them before. The body does this by recognising 'motifs' that are not found in mammalian cells. These include lipopeptides, double-stranded RNA, lipopolysaccharide (LPS), flagellin, and CpG DNA. Such motifs are commonly present on viral and bacterial invaders. The motifs are otherwise known as PAMPS (pathogen associated molecular patterns).

The receptors that do the recognising are called Toll-like receptors, which are named from genes found in the fruit fly, and highly conserved throughout evolution (They're even found in plants)! The name 'Toll' comes from the German word for 'crazy'. TLRs are similar in structure to the interleukin-1 receptor.

Here's a list of mammalian TLRs:

Receptor What it recognises
TLR2 + TLR1 diacyl lipopeptides
TLR2 + TLR6 triacyl lipopeptides, zymosan
TLR3 double-stranded RNA
TLR4 Lipopolysaccharide (from bacterial cell wall)
TLR5 flagellin
TLR7 single-stranded RNA, imiquimod!
TLR8 single-stranded RNA
TLR9 CpG (bacterial) DNA, hemozoin (malaria pigment)
TLR10 unknown

See how some TLRs (for example, TLR2+TLR6) seem to combine to form dimers that recognize specific motifs. There is probably a host of other molecules which modulate the function of TLRs. Signals subsequent to TLR activation are extremely complex.

Unfortunately, the relationship between the innate immune system and IgE hasn't been well explored. There are some interesting preliminary findings..

The million dollar question is really whether agents like Mycoplasma pneumoniae (heavily implicated in exacerbations and even causation of asthma) affect toll-like receptors. They do --- TLR2 is critical in airway mucin expression in human lung epithelial cells, in response to M. pneumoniae infection! [J Immunol. 2005 May 1;174(9):5713-9]

2. The role of the dendritic cell

Dendritic cells vary. These are potent antigen-presenting cells, and there are several subsets. Traditionally it was said that dendritic cells with 'myeloid precursor' characteristics (CD-11c+) produce IL-12 and a Th-1 response, whereas plasmacytoid ones (CD123+) induce a Th2 response. The myeloid-type dendritic cells are often called pDC1 and the plasmacytoid ones, pDC2.

This simple binary classification seems to be far from the truth. Different sub-populations may do various things at different times. In addition, there is a lot of cross-modulation. For example, histamine may extensively modulate mature dendritic cell function! [J Clin Invest 2001 Dec 108 (12) 1865-73].

3. Revised mechanisms of IgE class switching

An 'alternative pathway' to IgE

Recently, an 'alternative' pathway has been defined that doesn't appear to require interaction of B cells with T cell CD40L for the IgE switch! In fact there may be multiple 'alternative' mechanisms involving:

The literature is made even more confusing by use of the term CD154 in place of CD40L.

Conversely, a host of factors inhibit IgE class switching. These include gamma interferon and IL-21, B cell surface receptors (B-cell receptor, CD45, CTLA4, CD23), and some transcription factors ("B cell lymphoma 6", "inhibitor of DNA binding 2").

4. A final note

We have interspersed References throughout the text; if you view the source in your web browser, you'll find a lot more!

Finally, a lot of our basic knowledge is derived from experiments on mice, which is unfortunate, as anaphylaxis in mice can be rather different from that in man. To give one example, anaphylaxis can be induced in mice with immunoglobulin G and monocytes, and nary an IgE molecule, basophil or mast cell in sight! As regards anaphylaxis, rats seem to provide a better model of anaphylaxis than do mice.