Article Text

Download PDFPDF
Complement deficiency and disease
  1. D J Unsworth
  1. Dr D J Unsworth, Department of Immunology and Immunogenetics, Southmead Hospital, Bristol BS10 5ND, UK; joe.unsworth{at}nbt.nhs.uk

Abstract

There are approximately 30 serum complement proteins (15% of the globulin fraction), excluding cell surface receptors, and regulatory proteins. Many are manufactured in the liver, and reduced complement is a feature of severe liver failure. Complement proteins contribute to the acute phase response, and high levels are seen in chronic untreated inflammation (eg, rheumatoid arthritis). Once activated, complement is strongly pro-inflammatory. Indeed, almost half of the complement system proteins/receptors play regulatory roles, reflecting the importance of controlling inappropriate activation. This review focuses on disease states arising as a direct consequence of complement deficiency or dysfunction.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Take-home messages

  • Complement is a powerful pro-inflammatory system of host defence proteins.

  • Multiple regulatory factors help ensure that potential host damage is limited during complement activation.

  • Complement deficiency can lead to autoimmunity (lupus, for example) or infection (meningitis, for example).

  • Defects of complement regulatory factors cause diseases including glomerulonephritis, angioedema and haemolysis.

In the 1880s, heat labile, serum factors, capable of killing bacteria were described. The mechanism was organism non-specific. It was later observed that these factors “complemented” the action of antibody. We now understand how antibodies bound to target antigen (for example, microbial structures), alter their three-dimensional conformation, revealing complement binding sites. Complement bound to immunoglobulin heavy chain is activated, and this flags immune complex for phagocytic disposal. All vertebrates show some evidence of serum complement activity as part of their innate, antigen non-specific, host defence mechanisms. Evolutionary advantage is however not the only line of evidence demonstrating the importance of complement in host defence. Because complement helps removal of pathological antibody/antigen immune complexes, deficiency is also associated with poor immune complex clearance, leading to autoimmunity, including vasculitis.

Components were initially numbered (eg, C1, C2) but discovery of a then novel (“alternative”) activation pathway identified newer factors, which were each ascribed a letter (eg, factor B). Figure 1 is an overview (omitting some components for simplicity), divided into three important functional areas: the classical pathway, the alternative pathway, and the common or terminal pathway. Full details are available elsewhere.1 Many complement proteins are inactive proteases, themselves activated by proteolysis (so called zymogens). Serial amplification occurs akin to activation of the blood coagulation cascade.

Figure 1 Simplified diagram of the complement system.

Classical pathway (CP) activation

C1 is composed of three subunits, C1q, C1r and C1s. C1q binds to antibody Fc within an antigen–antibody complex, activating C1s (also called C1 esterase) initiating the CP, leading to proteolysis of C4 and C2. The CP can be activated by antibody independent mechanisms, including C1q binding to CRP (C-reactive protein) bound to bacteria. Mannose binding lectin (MBL) is structurally similar to C1q, and can also activate the CP. MBL shows affinity for bacteria expressing particular mannose moieties. MBL associated serine proteases 1 (SP1) or 2 (SP2) are analogous to C1r and C1s and can activate the CP by cleavage of either C4 or C2 (MBL–SP2), or C3 (MBL–SP1). Thus, MBL bypasses the need for C1q.

Activation via C1 or MBL leads to C4 proteolysis, generating C4b, which binds covalently to local activating surfaces (eg, bacteria). C2 is only susceptible to enzyme cleavage after it binds to C4b. C2 splits into C2a and C2b. A complex of C4b and C2a (C4b2a) capable of “converting” C3 to C3a and C3b is generated (see C3 convertases, below).

Alternative pathway (AP) activation

The AP is antibody independent. “Tick over” (hydrolytic) C3 activation is amplified by this route. Control mechanisms ensure that activation is restricted to the surface of pathogens. Comparing and contrasting with the CP, C3b is analogous to C4b, and factor B is analogous to C2. Factor B binds to C3b, and factor D (an active plasma protease not requiring activation, in contrast to the zymozans) cleaves factor B to Ba and Bb. The resulting fluid phase C3 convertase (C3bBb) is unstable, and although much of the C3b is destroyed, some binds covalently to local surfaces.

The terminal pathway

C3b initiates a cascade of events allowing C5 activation, generating C5a and C5b. Thereafter, C5b and other complement proteins (C6, C7, C8, C9) aggregate, producing a lipophilic “membrane attack complex” (MAC) capable of punching holes or pores in cell membranes, causing apoptosis or cell death. IgG autoantibody mediated haemolysis would typically, for example, involve MAC generation and cell death.

CONSEQUENCES OF COMPLEMENT ACTIVATION14

Waste disposal—“opsonisation”

C3b is a very efficient opsonin, facilitating removal of antibody–antigen immune complexes or complement coated pathogens, by phagocytes expressing C3b receptors. Erythrocytes express complement receptor 1 (CR1) and are key in transporting immune complexes to the liver and spleen for destruction. CR1 and other complement receptors (CR3 and CR4) are also expressed on professional phagocytes including macrophages, neutrophils and monocytes.

C3a and C5a (“anaphylotoxins”)

These signalling molecules attract phagocytic cells and promote a powerful inflammatory response. If experimentally injected, they cause a precipitous, anaphylaxis-like circulatory response, hence the term “anaphylotoxin”.

Targeted membrane lysis

As described above by generation of the membrane attack complex (MAC).

Positive effect on antibody responses

Co-presentation of antigen with activated complement (C3d), via complement receptor 2 (expressed by B lymphocytes) lowers the threshold for B lymphocyte activation, favoring antibody production. CR2 incidentally is the portal of entry for Epstein–Barr virus (EBV) explaining why EBV targets B lymphocytes.

DIAGNOSTIC TESTS

Biochemical quantification (fig 2)

Genetically determined complement deficiencies are rare (see below) and less likely explanations for abnormal complement levels than the pathologies quoted in fig 2. The biochemical level of particular serum complement proteins can be easily quantified (eg, by nephelometry). Routinely measuring serum C3 and C4 levels allows a range of pathogenic processes to be excluded or diagnosed, and can be useful in disease monitoring.

Figure 2 Quantitative measurements of C3 and C4 and their interpretation. HUVS, Hypocomplementaemic urticarial vasculitis syndrome; SLE, systemic lupus erythematosus.

Thus low levels of C4 but normal C3 (assuming reduced C4 is not due to an inherited deficiency state) would suggest an active CP process, for example cryoglobulinaemia, which consumes C4. Audits suggest that normal C4 levels effectively exclude the presence of a mixed cryoglobulin. Treatment of the underlying process (immunosuppression if autoimmune, chemotherapy if clonal B cell expansion) removes cryoglobulin, allowing C4 levels to rise.

By contrast, normal C4 but reduced/consumed C3 suggest AP activation, as seen, for example, with nephritic factor or bacterial activation without significant CP activation.

Functional assays (AP50 and CH50) (fig 3)

Figure 3 Functional complement assays.

Patient serum provides the complement source, and MAC mediated haemolysis of sheep erythrocytes can be used as readout. Presence or absence of anti-erythrocyte antibody and selective use of buffers allows the AP haemolytic potential (AP50), or the CP haemolytic potential (CH50) to be separately measured. Normal (or heat treated) human serum (complement is inactivated by incubation at 56°C for 30 min) acts as a control. The number 50 refers to the serum dilution at which 50% haemolysis is achieved. If any one single complement component (eg, C4, or factor H; or factor C8) is 100% deficient (or consumed), then complement activation is abruptly halted, and haemolysis is not possible. Complete deficiency (eg, C2) in the CP allows normal AP50 but a CH50 result of zero (no haemolysis). By contrast, factor B or H deficiency leads to a normal CH50 but zero AP50 activity. A terminal pathway deficiency (eg, C6 or C8) leads to no haemolysis in either assay, and zero haemolysis in both the AP50 and the CH50. Fresh samples (within a few hours of sampling) are required, and decay artefact (rather than true deficiency) is the commonest reason for abnormal results in both the CH50 and AP50.

Immunohistological detection of complement in biopsy material

Monoclonal antibodies, for example, can be used to detect C4d (a degradation product of C4 consistent with CP activation) in renal allografts post-transplant and is regarded as a useful predictor of humoral rejection.5 Figure 4 shows pathognomonic C3 and IgA deposition in peri-lesional dermatitis herpetiformis skin. The pathogenesis is unclear. Wheat gluten related hypersensitivity mechanisms are suspected.

Figure 4 C3 deposition at the dermo-epidermal junction in dermatitis herpetiformis skin (arrow).

COMPLEMENT REGULATORY FACTORS14

C1 inhibitor is a serine protease inhibitor (or serpin) which blocks activated C1s and C1r, preventing uncontrolled C4 and C2 activation. It also inhibits MBL associated proteases. However, it is not functionally restricted to the complement system. For example, it suppresses blood coagulation by interfering with activation of factor XI, kallikrein and plasmin.

C4 binding protein (C4BP) binds C4b, regulating CP activation in the fluid phase. It is a factor I cofactor, and these two factors together break down the CP C3 convertase.

Factor H (another factor I cofactor) acts both in the fluid phase and at the cell surface. It regulates the AP, analogous to C4BP in the CP. It competes with factor B for C3b binding. Factor H is preferentially expressed on host surfaces rather than a microbial surfaces, so complement activation is less likely to proceed on host cells.

Properdin stabilises the AP convertase, but achieves the same end result as factor H, because it binds preferentially to microbial rather than host cells.

Complement receptor 1 (CR1) is a factor I cofactor which prevents generation of both AP and CP C3 convertases.

Instability of C3 convertases. Both the CP (C4b2a) and AP (C3bBb) convertases (above) are powerful complement activators. Control arises in part because they are inherently unstable with a brief half-life.

Factor I (C3b inactivator) breaks down both the AP and CP convertases.

Decay accelerating factor (DAF) has a similar role to CR1, but is not a cofactor for factor I.

Membrane cofactor protein is a factor I cofactor which facilitates C4b and C3b breakdown.

CD59 (protectin) is a widely expressed host cell surface molecule which prevents assembly of the MAC at autologous cell surfaces, by blocking the C8 and C9 assembly steps.

COMPLEMENT DEFICIENCY AND DISEASE

Classical pathway deficiencies and systemic lupus erythematosus

Inherited (or acquired) deficiencies of CP components are associated with increased risk of developing systemic lupus erythematosus (SLE). Increased susceptibility to infection is also a risk.37 The incidence of SLE in C1q, C4 and C2 deficiency states is 90%, 75% and 15%, respectively. Conversely, in active SLE, immune complex generation with complement consumption may lead to reduced CP components (eg, C4) during disease flares (fig 2). Hence, low levels of CP factors can be “cause”, “effect” or both. Overall women are at increased risk of lupus, but in C1q deficiency, the gender bias is lost. MBL deficiency also increases the risk of lupus.

Acquired CP defects may be secondary to auto antibodies (eg, anti-C1q which are found in up to one third of SLE cases). Anti-C1q is also associated with a rare syndrome called hypocomplementaemic urticarial vasculitis syndrome (HUVS).6

C1 inhibitor deficiency

C1 inhibitor deficiency with associated angioedema is very rare. Idiopathic angioedema is at least 1000 times commoner. Association with urticaria, and/or normal C4 levels points to idiopathic angioedema. C1 inhibitor deficiency is easily and frequently misdiagnosed.8

Inherited deficiency

This shows a prevalence of around 1 in 50 000. In type 1 C1 inhibitor deficiency (80% of cases), inadequate C1 inhibitor is produced. In type II (20%), quantity may be normal but the C1 inhibitor is dysfunctional. Both are autosomal dominant, but tend to present in teenage or adult life. Patients experience a random pattern of angioedema (without associated urticaria), affecting lips, tongue, eyelids, genitals and sometimes bowel. Obstructive swelling around the airway or gut can be life threatening. Gut associated angioedema can be very painful, leading to unnecessary laparotomy. Low levels of C4 are secondary to CP consumption. CP trigger factors include infection and trauma. Infusion of purified C1 inhibitor can correct the problem, and restore C4 levels.9 Current experimental treatments undergoing clinical trial aim to block physiological mediators of angioedema.10 In type I deficiency, anabolic steroids help by increasing production of C1 inhibitor by the single functional gene.

Acquired deficiency

This is either due to excessive C1 consumption (for example, associated with a haematological malignancy, chronic immune complex activation in lupus or cryoglobulinaemia), or due to a C1q specific autoantibody. Autoantibodies directed against C1 inhibitor, compromising function, are another possibility. Late onset angioedema, with low C4, and no family history, should suggest this as a possible diagnosis. Treatment of the primary condition (associated malignancy or autoimmune disease) is the key.

Paroxysmal nocturnal haemoglobinuria

This is an acquired intravascular complement mediated haemolytic disorder, with haemoglobinuria and associated thrombosis. Somatic mutation of the Pig-A gene on the X chromosome causes abnormalities in an enzyme required to synthesise a lipid constituent of normal cell membranes. This lipid moiety normally anchors a range of cell membrane proteins, including the complement regulatory proteins DAF and CD59. Defective synthesis of the lipid anchor causes absent expression of these cell surface regulatory proteins, allowing MAC deposition with haemolysis. Treatment is discussed below.

Complement receptor 3 and leucocyte adhesion molecule deficiency (LAD-1)

CR3 is a heterodimer, composed of CD11b and CD18. Endothelial CD18 receptors are up-regulated at sites of infection. Defective or absent CD18 prevents neutrophil homing to sites of inflammation, allowing local invasive pyogenic bacterial infection.11 Paradoxically elevated blood neutrophil counts are seen because neutrophils cannot exit the bloodstream normally.

Inherited complement deficiency and bacterial infection

Clinically significant inherited deficiencies are rare.2 4 Occult deficiency was found in around 1/1000 healthy Japanese blood donors. The majority had C9 deficiency, which seems less common in Europeans. No deficiencies were found in 4000 Swiss army recruits.

Terminal pathway deficiency and Neisseria meningitidis

Absolute deficiency of membrane attack complex components (C5, C6, C7, C8 or C9) leads to susceptibility to neisserial infection. These deficiencies are inherited as autosomal recessive traits. Deficient patients may survive two or more invasive meningococcal episodes. Complement deficiency reduces inflammatory sequelae, explaining why infusion of fresh frozen plasma (as a replacement source of complement) can lead to deterioration.12 Confirmed terminal pathway deficiency associated with meningococcus in the UK approximates to one or two cases per million (personal observation).

Properdin deficiency and Neisseria meningitidis

This shows X-linked inheritance. Again, the link is with meningococcus.

Despite the association between meningococcal disease and complement deficiencies, routine screening is generally unrewarding. In a single supra-regional UK paediatric centre, 297 consecutive children who survived proven invasive meningococcal infection (mostly serogroups B and C) over a four year study period, were checked. Only one case of complement deficiency was identified.13 Complement deficiency is more likely to be detected in association with the meningococcal serogroups W135, X, Y and Z.14

MBL deficiency

Up to 5% of the population show some degree of MBL deficiency. Deficient adults appear healthy, but low levels in children can be associated with increased risk of bacterial infection, especially in the time period between loss of passively acquired maternal IgG (6 months of age) and generation of protective antibody. Following bone marrow transplantation, individuals with particular MBL polymorphisms are at greatest risk of infection.15 Trials using prophylactic infused MBL are under consideration.

Inherited complement deficiency and renal disease

Immune complex diseases and glomerulonephritis

Diseases such as SLE or cryoglobulinaemia can lead to the deposition of immune complexes with complement activation. Defective CP function (see above), or defects in phagocytosis, increase the risk.

Factor H deficiency and recurrent haemolytic uraemic syndrome

In the absence of factor H, increased production of AP convertase (C3bBb) leads to membranoproliferative glomerulonephritis. Further work has shown that abnormalities of factor I and/or C4BP can also lead to a similar familial predisposition to nephritis.16 Ideal management would be combined liver (as source of normal complement regulatory proteins) and kidney transplantation.

C3 nephritic factor

This is an IgG autoantibody, which binds to and stabilises the AP C3 convertase. It promotes C3 activation, causing C3 consumption. Mesangiocapillary glomerulonephritis type II is one classical association. Electron microscopy shows dense C3 and IgG deposits within the glomerular basement membrane. Partial lipodystrophy is also associated with nephritic factor; the lipodystrophy spares the lower limbs. Proximal adipose cells express less factor D, explaining this observation. The anti-diabetes mellitus drug rosiglitazone stimulates adipocyte regeneration (despite ongoing complement activation), and can be helpful cosmetically.17

POTENTIAL FOR ANTI-COMPLEMENT THERAPIES

Many ingenious strategies based on using bioengineered copies or derivatives of normal complement regulatory proteins or receptors have been considered in the treatment of inflammatory disease.18 Sadly, many of these strategies, while promising in experimental models, have not proved effective in patients.

Myocardial infarction and cerebrovascular ischaemic damage are both associated with local complement activation. Pexelizumab, a designer Ig fragment with C5 specificity was disappointing in trials. In patients with acute ST elevation, infusion in parallel with fibrinolysis, resulted in evidence of reduced complement activation, but no clinical benefit.19

One exception is the humanised monoclonal antibody eculizumab, specific for complement C5. The treatment prevents C5b generation, blocking formation of the membrane attack complex (C5b is a mandatory prerequisite). Eculizumab is well tolerated, and in patients with paroxysmal nocturnal haemoglobinuria, it reduces risk of haemolysis and lowers transfusion requirements.20

REFERENCES

Footnotes

  • Competing interests: None.