Skip to main content
NCBI home page
Clinical and Experimental Immunology logoLink to Clinical and Experimental Immunology
. 2005 Jun;140(3):572–579. doi: 10.1111/j.1365-2249.2005.02794.x

Defective prevention of immune precipitation in autoimmune diseases is independent of C4A*Q0

G J Arason *, R Kolka *, A B Hreidarsson *, H Gudjonsson *, P M Schneider †, L Fry ‡, A Arnason §
PMCID: PMC1809379  PMID: 15932521

Abstract

Increased prevalence of C4 null alleles is a common feature of autoimmune diseases. We have shown previously that complement-dependent prevention of immune precipitation (PIP) is defective in patients with systemic lupus erythematosus (SLE), and correlated this defect with C4A*Q0 and low levels of the C4A isotype. To further clarify the role of C4A in the aetiology of SLE, we now extend our studies to other diseases which have been associated with C4A*Q0. The frequency of C4A*Q0 was increased in Icelandic patients with coeliac disease (0·50; P < 0·001), Grave's disease (0·30; P = 0·002) and insulin-dependent diabetes mellitus (0·23; P = 0·04) and in British patients with dermatitis herpetiformis (0·42; P = 0·002) and this was reflected in low levels of C4A. In spite of this, PIP was normal in these patients, and in marked contrast to our previous observations on connective tissue diseases, PIP measurements in these patient groups correlated more strongly with levels of C4B (r = 0·51, P = 0·0000004) than C4A. Patients with increased levels of anti-C1q antibodies had significantly lower PIP than patients without such antibodies (P < 0·01) and a negative association of PIP with anti-C1q antibodies was also reflected in an increased prevalence (P = 0·006) and levels (P = 0·006) of anti-C1q antibodies in patients with subnormal PIP, as well as a negative correlation between PIP and anti-C1q antibodies (r = − 0·25, P = 0·02). These results show that the PIP defect cannot be explained by low levels of C4A alone and suggest that measurements of anti-C1q antibodies may be useful in future studies on the molecular cause of the PIP defect in autoimmune connective tissue disease.

Keywords: antigen–antibody complex, autoimmune disease, complement

Introduction

The classical pathway of complement is instrumental in the clearance of immune complexes to the liver, where they are safely eliminated [1]. The first two components, C1 and C4 are in a crucial position in this function. Unlike C1, which is non-polymorphic, C4 is a highly polymorphic protein, coded for by two tandem-duplicated genes located in the major histocompatibility complex (MHC) region on human chromosome 6. It exists as two isotypes, C4A and C4B, for which more than 40 allotypic variants are recognized [2]. Null alleles (C4A*Q0 and C4B*Q0) producing no identifiable product are common, and increased frequency of these alleles has been observed in the immune complex diseases (ICD), systemic lupus erythematosus (SLE), systemic sclerosis and Henoch–Schönlein purpura [3–8].

The high prevalence of C4A*Q0 in immune complex disease (ICD) has been linked to in vitro results indicating that C4A binds stronger to immune complexes than C4B [9–11], and used to argue the hypothesis that defective immune complex clearance could play a role in the aetiology or early pathogenesis of ICD [12–14]. This hypothesis owes its origin to the high prevalence of ICD observed in individuals with inherited absolute deficiencies of C1 or C4 [1,15], but to account for the majority of cases, who do not have any obvious classical pathway abnormalities, it is assumed that even subtotal deficiencies (e.g. resulting from partial deficiency of C4) may play a role [12–14]. Such deficiencies are considered to give rise to the autoimmune component of ICD through chronic release of autoantigens from inflamed tissues after immune complex deposition, and this is consistent with results indicating that SLE autoantibodies are driven by antigen [16,17]. Additional support is gained from the observation that the compounds most strongly implicated in drug-induced lupus erythematosus (DILE) are all strong inhibitors of C4A [18–21].

The main problem with the theory of complement involvement in ICD aetiology lies in the fact that almost all the evidence quoted so far has been circumstantial. However, we have recently confirmed that complement-dependent prevention of immune precipitation (PIP) is indeed defective in SLE patients, and that this defect is especially prominent in the early stages of the disease [5]. The defect was strongly correlated with low levels of C4, especially C4A, and a similar defect which we noted on a smaller scale in patients with systemic sclerosis was also correlated with levels of C4A [7]. At first sight these results might seem to favour the conclusion that C4A*Q0 and relative deficiency of C4A may predispose to connective tissue disease through defective immune complex clearance. However, one important problem with this argumentation is that C4A*Q0 is also a feature of several autoimmune diseases in which tissue deposition of immune complexes has not been established [22–34]. For further clarification of the relationship between C4A and prevention of immune precipitation we thus turned our attention to the C4A*Q0-associated diseases insulin-dependent diabetes mellitus (IDDM), autoimmune thyroid disease (Grave's and Hashimoto’s), and the autoimmune gluten-sensitive diseases (GSD) [35], dermatitis herpetiformis (DH) and coeliac disease (CD). Our results show that prevention of immune precipitation (PIP) may be normal even in the total absence of C4A, but was below normal in most patients who had elevated titres of IgG or IgA anti-C1q antibodies.

Materials and methods

Patients

The study group consisted of 24 patients with DH, 21 with CD, 25 with Grave's disease, 24 with IDDM (two of whom also had Grave's disease) and three with Hashimoto's disease. The diagnosis of DH was confirmed by the presence of IgA in the dermal papillae or in a linear granular band below the basement membrane of uninvolved skin. Criteria for the diagnosis of CD were (a) subjective and/or objective symptoms or signs of intestinal malabsorbtion; (b) total or subtotal villous atrophy on a small intestinal biopsy; and (c) unequivocal clinical improvement after gluten withdrawal and/or treatment with corticoid steroids. The diagnosis of IDDM and autoimmune thyroid disease was based on common criteria [36,37]. The DH patients were British, sampled consecutively at St Mary's Campus, but the remaining patients were Icelandic, sampled consecutively at Landspitali University Hospital. The age and male/female ratio was 21–69 (mean 48) in DH (13 males, 11 females), 16–84 (mean 43) in CD (seven males, 14 females), 17–56 (mean 35) in IDDM (18 males, four females) and 24–68 (mean 40) in Grave's (five males, 20 females). Two females (aged 42 and 61 years) had both IDDM and Grave’s, and two females (aged 49 and 57 years) and one male (aged 51 years) had Hashimoto's disease. Informed consent was obtained from all patients.

Materials

Microtitre plates were purchased from Nunc (Roskilde, Denmark), complement fixation diluent (CFD) from Flow (Irvine, Scotland), C1q, calf alkaline phosphatase (AP), AP substrate (p-nitrophenylphosphate) and AP-conjugated mouse α-human IgG from Sigma (St Louis, MO, USA), AP-conjugated mouse anti-IgA from BD Pharmingen (San Diego, CA, USA), goat α-human C4 from DiaSorin (Stillwater, MN, USA) and goat α-calf AP from ICN/Cappel (Aurora, OH, USA). Rabbit α-human C4c, rabbit α-human C3d and AP-conjugated rabbit α-mouse immunoglobulins were from Dako (Copenhagen, Denmark) and mouse monoclonal α-human C4d (A213) and α-human C4A (RgD1) were from Quidell (San Diego, CA, USA) and Biotest (Dreieich, Germany), respectively. In enzyme-linked immunosorbent assay (ELISA) measurements, phosphate buffered saline (PBS) containing 0·05% Tween was used for diluting sera and PBS with 0·005% Tween for washing.

C4 allotypes

All serum samples were stored at −70°C after collection and kept on ice during measurements except as indicated. C4 allotypes were determined by high-voltage agarose electrophoresis on carboxypeptidase- and neuraminidase-treated samples with subsequent immunofixation and staining [2]. Null alleles were determined by visual scoring of the relative intensity of C4A and C4B bands.

PIP

PIP was measured as previously described [38]. Briefly, AP (1/50, 5 µl) and goat α-AP (1/5, 5 µl) were added to serum (30 µl); after incubation (37°C, 1 h) and centrifugation (5500 g, 10 min), supernatants were diluted 1 : 10 in PBS and reacted with AP substrate. Absorbance was converted to arbitrary units (AU) by comparison to a serially diluted reference serum pool, defined as 100 AU. This assay is sensitive to minor variations within and below the normal range [38].

C3d-ELISA

Complement activation was assessed by monitoring serum levels of the C3d cleavage product by an ELISA [39], using rabbit α-C3d (1 : 1000) for coating and AP-conjugated α-C3d (1 : 2000) for developing. Sera were diluted 1 : 100 and results expressed in AU by comparison to a serially diluted zymosan-activated reference serum, defined as 100 AU. Specificity for C3d was ensured by removal of larger C3 fragments (C3, C3b, iC3b) by adding ethyline diamine tetra-acetic acid (EDTA) (10 m M) to the serum and mixing it with equal volumes of 22% polyethylene glycol, followed by incubation on ice (1 h) and centrifugation (1500 g, 30 min) at 4°C before adding it to the microtitre plate wells. Normal levels for C3d were determined by measuring sera from 100 blood donors.

CH50, complement components and autoantibodies

C4 and C3 were measured by rocket immunoelectrophoresis and total haemolytic complement (CH50) by standard titration methods. Results were expressed in g/l by comparison to a known standard, or as AU by comparison to a standard serum pool. Anti-C1q antibodies were measured by ELISA [40], using C1q (1 : 2500) for coating and AP-conjugated mouse α-human IgG (1 : 2000) or IgA (1 : 1000) antibodies for developing. Interference from immune complexes binding to the coated plate was avoided by increasing the molecular strength (1 M NaCl) of the serum dilution buffer. Results were expressed in AU by comparison to serially diluted reference sera with high levels of IgG or IgA α-C1q antibodies; after measuring 100 sera from blood donors, the upper limit of normal was calculated as the average + 1·96 s.d. and defined as 95 AU. The assay was calibrated against established methods by exchange of standards, sera and reagents between our laboratory and collaborators in Sweden; in our measurements, 100 AU correspond to 62 AU for IgG and 320 AU for IgA α-C1q in previous publications [41–44].

C4A and C4B levels

Levels of C4A and total C4 were measured by a modified ELISA [42] using goat α-human C4 for coating (1 : 200) and mouse monoclonal α-human C4A (RgD1; 1/1000) or C4d (A213; 1/2000) followed by AP-conjugated rabbit α-mouse immunoglobulin (1/1000) for developing; sera were diluted 1 : 800 and 1/1600. Results were expressed in g/l by comparison to serially diluted reference serum samples from phenotypically homozygous C4A deficient probands or individuals without null alleles; their C4 concentrations had previously been determined twice by rocket immunoelectrophoresis. C4B estimation was reached by subtracting C4A from total C4. This approach has been validated in previous studies in our laboratory by simultaneous measurements of C4B in ELISA using mouse monoclonal α-C4B (Mab 1228 from Biotest).

Statistical analysis

The means of test values of patients and controls were compared using the Mann–Whitney U-test. PIP was compared with other laboratory and clinical parameters using anova, χ2 and Pearson's correlation statistics. Significance was set at P < 0·05.

Results

C4 allotypes

The frequency of C4A*Q0 was greatly increased in the patients (Table 1), being 0·50 in CD (P < 0·001), 0·42 in DH (P = 0·002), 0·30 in Grave's disease (P = 0·002) and 0·23 in IDDM (P = 0·04) compared to 0·12 in Icelandic and 0·18 in British controls [4,46]. The carrier frequencies were 76% in CD (16/21 carriers, including five homozygotes), 79% in DH (19/24 carriers, one homozygote), 54% in Grave's disease (14/26 carriers, one homozygote) and 42% in IDDM (10/24, one homozygote) compared with control values of 25% in Iceland (49/194) and 36% (24/132) in the United Kingdom [4,46]. Two of the three patients with Hashimoto's disease had heterozygous C4A*Q0.

Table 1.

Distribution of C4 allotypes in Icelandic patients with gluten-sensitive diseases (GSD), insulin-dependent diabetes mellitus (IDDM), Grave's disease and Hashimoto's disease

GSD IDDM Grave’s Hashimoto’s Total
n freq. n freq. n freq. n freq. n freq.
A*Q0 41 0·46 11 0·23 15 0·30 2 0·33 69 0·36
A*2 3 0·03 1 3 0·06 0·00  7 0·04
A*3 40 0·44 33 0·69 26 0·52 4 0·67 103 0·53
A*4 5 0·06 2 0·04 5 0·10 0·00 12 0·06
A*6 1 0·01 1 1 0·02 0·00  3 0·02
Total 90 48 50 6 194
B*Q0 1 0·01 4 0·08 1 0·02  6 0·03
B*1 84 0·93 29 0·60 37 0·74 5 0·83 155 0·80
B*2 3 0·03 9 0·19 7 0·14 1 0·17 20 0·10
B*22 1 0·02  1 0·01
B*3 2 0·02 5 0·10 4 0·08 11 0·06
B*5 1 0·02  1 0·01
Total 90 48 50 6 194
Open in a new tab

CH50, C3d, C3, C4, C4A, C4B

Five patients had slightly raised C3d, indicating recent activation of complement (Fig. 1). However, measurements of C3 and CH50 were normal in the patients. Levels of C4 in the patients were suboptimal in 18 patients (Fig. 1) but were not decreased in the whole group of patients compared to controls (results not shown). The presence of C4A*Q0 was reflected in lower levels of C4A (P < 0·001) as well as total C4 (P = 0·002) than in the remaining patients without C4A*Q0 (Fig. 2); the effect of C4B null could not be analysed meaningfully as only two samples with C4B*Q0 were available for C4 measurements.

Fig. 1.

Fig. 1

Open in a new tab

CH50 and levels of C4, C3 and C3d in patients with gluten-sensitive diseases (GSD), insulin-dependent diabetes mellitus (IDDM) and autoimmune thyroid disease. Horizontal lines denote lower normal limits for CH50, C4 and C3, and upper normal limits for C3d.

Fig. 2.

Fig. 2

Open in a new tab

Levels of C4 and C4A in patients with and without C4A*Q0.

PIP

The PIP defect shown previously in patients with SLE and systemic sclerosis was not observed in patients with GSD, IDDM or autoimmune thyroid disease (Fig. 3) although a few patients were below the normal limit (< 68 AU). Low PIP could neither be correlated to C4A*Q0 (Fig. 3, open and dotted circles) nor complement activation as revealed by increased C3d. On the other hand, PIP did correlate with levels of C4 (r = 0·55, P = 0·00000004) and surprisingly, this was due mainly to C4B (r = 0·51, P = 0·0000004; Fig. 4, Table 2). A weaker correlation was found between PIP and C3 (Table 2), C3 and C4B (r = 0·24, P = 0·02), and C3 and C4 (r = 0·23, P = 0·03).

Fig. 3.

Fig. 3

Open in a new tab

Prevention of immune precipitation in patients with gluten-sensitive diseases (GSD), autoimmune thyroid disease and insulin-dependent diabetes mellitus (IDDM). Open circles denote heterozygous and dotted circles homozygous C4A*Q0 carriers. Filled circles represent C4A*Q0 non-carriers. Patients with Hashimoto's disease are marked with arrows. Arbitary units: AU.

Fig. 4.

Fig. 4

Open in a new tab

Correlation of PIP to C4B*Q0 in insulin-dependent diabetes mellitus (IDDM), gluten-sensitive diseases (GSD) and autoimmune thyroid disease.

Table 2.

Correlation between prevention of immune precipi tation (PIP) and other complement parameters in patients with gluten-sensitive diseases (GSD), insulin-dependent diabetes mellitus (IDDM) or thyroid disease

C3 C4 C4A C4B
PIP r = 0·24 r = 0·55 r = 0·25 r = 0·51
P = 0·02 P < 0·001 P = 0·02 P < 0·001
Open in a new tab

Anti-C1q antibodies

Levels of anti-C1q antibodies were elevated in 20 patients, and PIP values were markedly lower in these patients than in patients without these antibodies (P < 0·01, Fig. 5). This was reflected in an inverse correlation between PIP and anti-C1q antibodies (r = − 0·25, P = 0·02) as well as an increased prevalence and concentration of α-C1q antibodies in patients with PIP values close to or lower than 68 (the lower normal limit for PIP) (Table 3). Decreased PIP values of patients with elevated anti-C1q antibodies were not associated with levels of C3, C4 or its isotypes, or the presence of C4 null alleles.

Fig. 5.

Fig. 5

Open in a new tab

Prevention of immune precipitation in patients with and without elevated IgG or IgA anti-C1q antibodies. AU: arbitrary units; • = outlier value.

Table 3.

Prevalence and levels of α–C1q antibodies in patients with gluten-sensitive diseases (GSD), insulin-dependent diabetes mellitus (IDDM) or thyroid disease

High α–C1q n Normal α–C1q n P*(Fisher exact) Median α–C1q (AU) P* (M–W)
PIP < 68 AU 5  3 0·006 181·5 0·006
PIP 68–79 AU 6  6 0·012 76·6 0·11
PIP > 79 AU 9 53 29·0
Open in a new tab
*

Compared to the group of patients with PIP >79 AU. AU: arbitary units; M-W: Mann-Whitney U test.

Discussion

The distribution of C4 allotypes in Icelandic patients with CD, IDDM and Grave's disease is described here for the first time. The frequency of C4A*Q0 turned out to be very high, with carrier frequencies of 76%, 54% and 42% and gene frequencies of 0·50, 0·30 and 0·23 in CD, Grave's disease and IDDM, respectively. For the British DH patients the corresponding values were 79% and 0·42. An increased prevalence of C4A*Q0 has been recorded previously in Irish [31], American [32, 33] and Italian [34] patients with autoimmune gluten-sensitive disease (DH or CD) and the prevalence is especially high in Ireland. The west of Ireland is the geographical area with the highest rate of CD in the world [47], possibly reflecting selective pressure from diet and historic patterns of cereal digestion [48]; the recent decline in CD among Irish children is consistent with this notion [49]. The prevalence of C4A*Q0 in DH and CD in London and Iceland turned out to be even higher than in Ireland, and in fact constitute the highest C4A*Q0 prevalence values so far recorded for any disease. The prevalence of C4A*Q0 in our patients with IDDM and Grave's was somewhat lower but still raised compared to controls, and this is consistent with data from other populations [22–30].

The increased frequency of C4A*Q0 in our patient group was reflected in low levels of the C4A protein. Other conventional complement parameters, however, turned out to be normal, allowing for optimal conditions for studying the effect of C4A*Q0 and low C4A on prevention of immune precipitation. A popular theory [12–14] has suggested that this condition leads to impaired PIP and hence to immune complex disease. Prevention of immune precipitation, however, turned out to be entirely normal in our patient group, and even the eight patients who had homozygous C4A*Q0 and no serum C4A had PIP values of up to 149. Measurements of C4, C4A and C4B confirmed this, and showed that PIP was correlated much more strongly with C4B than C4A. This is in marked contrast to our previous results on patients with SLE and scleroderma, and completely inconsistent with the notion that defective immune complex handling in SLE may be caused by low levels of C4A due to the presence of C4A*Q0.

Our study contains two possible clues to the identity of the factor(s) responsible for defective PIP. On one hand, it seems possible that PIP may depend on levels of total C4, and the correlation of PIP to C4B in this study and C4A in patients with SLE and scleroderma may thus represent an epiphenomenon. In this respect it may be pertinent that we and others have found raised levels of C4B*Q0 in Henoch–Schoenlein purpura [8,50,51]. Perhaps it could be postulated that immune complex disease may result from subtle defects in complement activity caused, for example, by relatively low C4, and that other as yet unknown factors then determine the exact nature of the immune complex disease.

The other explanation could relate to our finding in this study of an inverse correlation between PIP and antibodies to C1q (r = − 0·26, P = 0·02). Although anti-C1q antibodies were not raised in the whole group of patients, both the prevalence and levels of these antibodies were raised in patients with low PIP, and PIP was markedly lower in the 20 patients who had raised titres of these antibodies than in the remaining patients. Levels of anti-C1q antibodies are increased in several vasculitic and nephritic disorders [52–57] and in 10% of healthy subjects, increasing with age [58]. Very high levels of these antibodies have been considered pathognomonic for SLE and hypocomplementaemic urticarial vasculitis syndrome (HUVS); however, the distinction of the latter disease from SLE has been questioned [59] and high titres of these antibodies may thus prove to be the distinctive hallmark of SLE. Our previous findings of decreased PIP in patients with SLE [5], together with our current findings of an inverse relationship between PIP and anti-C1q antibodies, suggest that these parameters may be correlated in SLE. It appears quite logical to assume that cross-binding of immune complexes by anti-C1q antibodies could interfere with the normal handling of such complexes and instead promote immune complex precipitation. Studies to test this hypothesis are forthcoming.

In conclusion, complement-dependent prevention of immune precipitation was normal in patients with GSD, IDDM and autoimmune thyroid disease in spite of a high prevalence of C4A*Q0 and low levels of C4A in these patients. Defective immune complex handling may be related to the presence of anti-C1q antibodies rather than low levels of individual complement components.

Acknowledgments

This work was supported by the Icelandic Research Council (grant no. 971310097) and the Science Fund of Landspitali University Hospital, and was approved by all relevant ethical committees.

References

  • 1.Schifferli JA, Ng YC, Peters DK. The role of complement and its receptor in the elimination of immune complexes. N Engl J Med. 1986;315:488–95. doi: 10.1056/NEJM198608213150805. [DOI] [PubMed] [Google Scholar]
  • 2.Schneider PM, Mauff G. Complement C4 protein and DNA typing methods. Methods Mol Biol. 2003;210:269–95. doi: 10.1385/1-59259-291-0:269. [DOI] [PubMed] [Google Scholar]
  • 3.Christiansen FT, Zhang WJ, Griffiths M, Mallal SA, Dawkins RL. Major histocompatibility complex (MHC) complement deficiency, ancestral haplotypes and systemic lupus erythematosus (SLE): C4 deficiency explains some but not all of the influence of the MHC. J Rheumatol. 1991;18:1350–8. [PubMed] [Google Scholar]
  • 4.Steinsson K, Jonsdottir S, Arason GJ, Fossdal R, Skaftadottir I, Arnason A. A study of the association of HLA DR, DQ, and complement C4 alleles with systemic lupus erythematosus in Iceland. Ann Rheum Dis. 1998;57:503–5. doi: 10.1136/ard.57.8.503. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Arason GJ, Steinsson K, Kolka R, Vikingsdottir Th, D’Ambrogio MS, Valdimarsson H. Patients with systemic lupus erythematosus are deficient in complement-dependent prevention of immune precipitation. Rheumatology. 2004;43:783–9. doi: 10.1093/rheumatology/keh183. [DOI] [PubMed] [Google Scholar]
  • 6.Briggs D, Stephens C, Vaughan R, Welsh K, Black C. A molecular and serologic analysis of the major histocompatibility complex and complement component C4 in systemic sclerosis. Arthritis Rheum. 1993;36:943–55. doi: 10.1002/art.1780360711. [DOI] [PubMed] [Google Scholar]
  • 7.Arason GJ, Geirsson ÁJ, Kolka R, Vikingsdottir Th, Valdimarsson H. Deficiency of complement-dependent prevention of immune precipitation in systemic sclerosis. Ann Rheum Dis. 2002;61:257–60. doi: 10.1136/ard.61.3.257. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Thors VS, Kolka R, Sigurdardottir S, Edvardsson VO, Arason GJ, Haraldsson A. Increased freqeuncy of C4B*Q0 alleles in patients with Henoch–Schönlein purpura. Scand J Immunol. pp. 274–8. [DOI] [PubMed]
  • 9.Law SK, Dodds AW, Porter RR. A comparison of the properties of two classes, C4A and C4B, of the human complement component C4. EMBO J. 1984;3:1819–23. doi: 10.1002/j.1460-2075.1984.tb02052.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Kishore N, Shah D, Skanes VM, Levine RP. The fluid-phase binding of human C4 and its genetic variants, C4A3 and C4B1, to immunoglobulins. Mol Immunol. 1988;25:811–19. doi: 10.1016/0161-5890(88)90117-4. [DOI] [PubMed] [Google Scholar]
  • 11.Schifferli JA, Steiger G, Paccaud JP, Sjoholm AG, Hauptmann G. Difference in the biological properties of the two forms of the fourth component of human complement (C4) Clin Exp Immunol. 1986;63:473–7. [PMC free article] [PubMed] [Google Scholar]
  • 12.Lachmann PJ, Walport MJ. Deficiency of the effector mechanisms of the immune response and autoimmunity. Ciba Found Symp. 1987;129:149–71. doi: 10.1002/9780470513484.ch11. [DOI] [PubMed] [Google Scholar]
  • 13.Walport MJ, Davies KA, Morley BJ, Botto M. Complement deficiency and autoimmunity. Ann NY Acad Sci. 1997;815:267–81. doi: 10.1111/j.1749-6632.1997.tb52069.x. [DOI] [PubMed] [Google Scholar]
  • 14.Sullivan KE. Complement deficiency and autoimmunity. Curr Opin Pediatr. 1998;10:600–6. doi: 10.1097/00008480-199810060-00011. [DOI] [PubMed] [Google Scholar]
  • 15.Agnello V. Lupus diseases associated with hereditary and acquired deficiencies of complement. Springer Sem Immunopathol. 1986;9:161–78. doi: 10.1007/BF02099020. [DOI] [PubMed] [Google Scholar]
  • 16.Tan EM. Pathophysiology of antinuclear antibodies in systemic lupus erythematosus and related diseases. Adv Dent Res. 1996;10:44–6. doi: 10.1177/08959374960100010801. [DOI] [PubMed] [Google Scholar]
  • 17.Casciola-Rosen L, Rosen A. Ultraviolet light-induced keratinocyte apoptosis. a potential mechanism for the induction of skin lesions and autoantibody production in LE. Lupus. 1997;6:175–80. doi: 10.1177/096120339700600213. [DOI] [PubMed] [Google Scholar]
  • 18.Sim E, Gill EW, Sim RB. Drugs that induce systemic lupus erythematosus inhibit complement component C4. Lancet. 1984;2:422–4. doi: 10.1016/s0140-6736(84)92905-2. [DOI] [PubMed] [Google Scholar]
  • 19.Sim E, Law SK. Hydralazine binds covalently to complement component C4. Different reactivity of C4A and C4B gene products. FEBS Lett. 1985;184:323–7. doi: 10.1016/0014-5793(85)80631-1. [DOI] [PubMed] [Google Scholar]
  • 20.Sim E, Dodds AW, Goldin A. Inhibition of the covalent binding reaction of complement component C4 by penicillamine, an anti-rheumatic agent. Biochem J. 1989;259:415–9. doi: 10.1042/bj2590415. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Sim E, Stanley L, Gill EW, Jones A. Metabolites of procainamide and practolol inhibit complement components C3 and C4. Biochem J. 1988;251:32. doi: 10.1042/bj2510323. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Bertrams J, Hintzen U, Schlicht V, Schoeps S, Gries FA, Louton TK, Baur MP. Gene and haplotype frequencies of the fourth component of complement (C4) in type 1 diabetics and normal controls. Immunobiology. 1984;166:335–44. doi: 10.1016/S0171-2985(84)80012-1. [DOI] [PubMed] [Google Scholar]
  • 23.Rich S, O'Neill G, Dalmasso AP, Nerl C, Barbosa J, Complement HLA. Further definition of high-risk haplotypes in insulin-dependent diabetes. Diabetes. 1985;34:504–9. doi: 10.2337/diab.34.5.504. [DOI] [PubMed] [Google Scholar]
  • 24.Skanes VM, Barnard J, Farid N, et al. Class III alleles and high-risk MHC haplotypes in type I diabetes mellitus, Graves’ disease and Hashimoto's thyroiditis. Mol Biol Med. 1986;3:143–57. [PubMed] [Google Scholar]
  • 25.Tucker WS, Jr, Niblack GD, McLean RH, et al. Serositis with autoimmune endocrinopathy: clinical and immunogenetic features. Medicine (Baltimore) 1987;66:138–47. doi: 10.1097/00005792-198703000-00005. [DOI] [PubMed] [Google Scholar]
  • 26.Wang C, Rivas ML, Burghen GA, Hudson EC, Wyatt RJ. C4 and Bf phenotypes in black and Caucasian patients with childhood onset insulin dependent diabetes mellitus. J Clin Lab Immunol. 1989;30:183–90. [PubMed] [Google Scholar]
  • 27.Ratanachaiyavong S, Lloyd L, McGregor AM. C4A gene deletion: association with Graves’ disease. J Mol Endocrinol. 1989;3:145–53. doi: 10.1677/jme.0.0030145. [DOI] [PubMed] [Google Scholar]
  • 28.Ratanachaiyavong S, Lloyd L, Darke C, McGregor AM. MHC-extended haplotypes in families of patients with Graves’ disease. Hum Immunol. 1993;36:99–111. doi: 10.1016/0198-8859(93)90112-e. [DOI] [PubMed] [Google Scholar]
  • 29.Tandon N, Yan SL, Morgan BP, Weetman AP. Expression and function of multiple regulators of complement activation in autoimmune thyroid disease. Immunology. 1994;81:643–7. [PMC free article] [PubMed] [Google Scholar]
  • 30.Lhotta K, Auinger M, Kronenberg F, Irsigler K, Konig P. Polymorphism of complement C4 and susceptibility to IDDM and microvascular complications. Diabetes Care. 1996;19:53–5. doi: 10.2337/diacare.19.1.53. [DOI] [PubMed] [Google Scholar]
  • 31.Mannion A, Stevens FM, McCarthy CF, Grimes-O’Cearbhaill H, Killeen AA. Extended major histocompatibility complex haplotypes in celiac patients in the west of Ireland. Am J Med Genet. 1993;45:373–7. doi: 10.1002/ajmg.1320450319. [DOI] [PubMed] [Google Scholar]
  • 32.Alper CA, Fleischnick E, Awdeh Z, Katz AJ, Yunis EJ. Extended major histocompatibility complex haplotypes in patients with gluten-sensitive enteropathy. J Clin Invest. 1987;79:251–6. doi: 10.1172/JCI112791. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Ahmed AR, Yunis JJ, Marcus-Bagley D, Yunis EJ, Salazar M, Katz AJ, Awdeh Z, Alper CA. Major histocompatibility complex susceptibility genes for dermatitis herpetiformis compared with those for gluten-sensitive enteropathy. J Exp Med. 1993;178:2067–75. doi: 10.1084/jem.178.6.2067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Amoroso A, Mazzola G, Canale L, et al. HLA in juvenile dermatitis herpetiformis: clinical heterogeneity correlated with DNA and serological polymorphism. J Immunogenet. 1990;17:195–206. doi: 10.1111/j.1744-313x.1990.tb00872.x. [DOI] [PubMed] [Google Scholar]
  • 35.Karpati S. Dermatitis herpetiformis: close to unravelling a disease. J Dermatol Sci. 2004;34:83–90. doi: 10.1016/j.jdermsci.2003.11.004. [DOI] [PubMed] [Google Scholar]
  • 36.Alberti KG, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1. Diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med. 1998;15:539–53. doi: 10.1002/(SICI)1096-9136(199807)15:7<539::AID-DIA668>3.0.CO;2-S. [DOI] [PubMed] [Google Scholar]
  • 37.Braverman LE, Utiger RD. Werner and Ingbars The thyroid: a fundamental and clinical text. 7. Philadelphia: Lippincott Williams & Wilkins; 1996. [Google Scholar]
  • 38.Arason GJ, D’Ambrogio MS, Vikingsdottir T, Sigfusson A, Valdi-marsson H. Enzyme immunoassays for measuring complement-dependent prevention of immune precipitation (PIP) and solubilization of preformed antigen–antibody complexes (SOL) J Immunol Meth. 1999;223:37–46. doi: 10.1016/s0022-1759(98)00199-9. [DOI] [PubMed] [Google Scholar]
  • 39.Mollnes TE. Quantification of the C3d split products of human complement by a sensitive enzyme-linked immunosorbent assay. Scand J Immunol. 1985;21:607–13. doi: 10.1111/j.1365-3083.1985.tb01851.x. [DOI] [PubMed] [Google Scholar]
  • 40.Siegert CE, Daha MR, van der Voort EA, Breedveld FC. IgG and IgA antibodies to the collagen-like region of C1q in rheumatoid vasculitis. Arthritis Rheum. 1990;33:1646–54. doi: 10.1002/art.1780331107. [DOI] [PubMed] [Google Scholar]
  • 41.Gunnarsson I, Ronnelid J, Lundberg I, Jacobson SH. Occurrence of anti-C1q antibodies in IgA nephropathy. Nephrol Dial Transplant. 1997;12:2263–8. doi: 10.1093/ndt/12.11.2263. [DOI] [PubMed] [Google Scholar]
  • 42.Ronnelid J, Gunnarsson I, Nilsson-Ekdahl K, Nilsson B. Correlation between anti-C1q and immune conglutinin levels, but not between levels of antibodies to the structurally related autoantigens C1q and type II collagen in SLE or RA. J Autoimmun. 1997;10:415–23. doi: 10.1006/jaut.1997.0147. [DOI] [PubMed] [Google Scholar]
  • 43.Gunnarsson I, Ronnelid J, Huang YH, et al. Association between ongoing anti-C1q antibody production in peripheral blood and proliferative nephritis in patients with active systemic lupus erythematosus. Br J Rheumatol. 1997;36:32–7. doi: 10.1093/rheumatology/36.1.32. [DOI] [PubMed] [Google Scholar]
  • 44.Ronnelid J, Huang YH, Norrlander T, et al. Short-term kinetics of the humoral anti-C1q response in SLE using the ELISPOT method: fast decline in production in response to steroids. Scand J Immunol. 1994;40:243–50. doi: 10.1111/j.1365-3083.1994.tb03457.x. [DOI] [PubMed] [Google Scholar]
  • 45.Rebmann V, Doxiadis I, Kubens BS, Grosse-Wilde H. Quantitation of the human component C4: definition of C4, Q0 alleles and C4A duplications. Vox Sang. 1992. pp. 117–23. [DOI] [PubMed]
  • 46.Davies EJ, Hillarby MC, Cooper RG, et al. HLA-DQ, DR and complement C4 variants in systemic lupus erythematosus. Br J Rheumatol. 1993;32:870–85. doi: 10.1093/rheumatology/32.10.870. [DOI] [PubMed] [Google Scholar]
  • 47.Michalski JP, McCombs CC, Arai T, et al. HLA-DR, DQ genotypes of celiac disease patients and healthy subjects from the West of Ireland. Tissue Antigens. 1996;47:127–33. doi: 10.1111/j.1399-0039.1996.tb02525.x. [DOI] [PubMed] [Google Scholar]
  • 48.Cronin CC, Shanahan F. Why is celiac disease so common in Ireland? Perspect Biol Med. 2001;44:342–52. doi: 10.1353/pbm.2001.0045. [DOI] [PubMed] [Google Scholar]
  • 49.Gumaa SN, McNicholl B, Egan-Mitchell B, Connolly K, Loftus BG. Coeliac disease in Galway, Ireland 1971–90. Ir Med J. 1997;90:60–1. [PubMed] [Google Scholar]
  • 50.Ault B, Stapleton FB, Rivas ML, et al. Association of Henoch–Schönlein purpura glomerulonephritis with C4B deficiency. J Pediatr. 1990;117:753–5. doi: 10.1016/s0022-3476(05)83336-0. [DOI] [PubMed] [Google Scholar]
  • 51.McLean RH, Wyatt RJ, Julian BA. Complement phenotypes in glomerulonephritis. Increased frequency of homozygous null C4 phenotypes in IgA nephropathy and Henoch–Schönlein purpura. Kidney Int. 1984;26:855–60. doi: 10.1038/ki.1984.228. [DOI] [PubMed] [Google Scholar]
  • 52.Uwatoko S, Aotsuka S, Okawa M, et al. Characterization of C1q-binding IgG complexes in systemic lupus erythematosus. Clin Immunol Immunopathol. 1984;30:104–16. doi: 10.1016/0090-1229(84)90011-4. [DOI] [PubMed] [Google Scholar]
  • 53.Strife CF, Leahy AE, West CD. Antibody to a cryptic, solid phase C1Q antigen in membranoproliferative nephritis. Kidney Int. 1989;35:836–42. doi: 10.1038/ki.1989.61. [DOI] [PubMed] [Google Scholar]
  • 54.Siegert CE, Daha MR, van der Voort EA, Breedveld FC. IgA antibodies to the collagen-like region of C1q in rheumatoid vascolitis. Arthritis Rheum. 1990;33:1646–54. doi: 10.1002/art.1780331107. [DOI] [PubMed] [Google Scholar]
  • 55.Wisnieski JJ, Baer AN, Christensen J, et al. Hypocomplementemic urticarial vasculitis syndrome. Clinical and serologic findings in 18 patients. Medicine (Baltimore) 1995;74:24–41. doi: 10.1097/00005792-199501000-00003. [DOI] [PubMed] [Google Scholar]
  • 56.Siegert CE, Kazatchkine MD, Sjoholm A, Wurzner R, Loos M, Daha MR. Autoantibodies against C1q: view on clinical relevance and pathogenic role. Clin Exp Immunol. 1999;116:4–8. doi: 10.1046/j.1365-2249.1999.00867.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Marto N, Bertolaccini ML, Calabuig E, Hughes GR, Khamashta MA. Anti-C1q antibodies in nephritis: correlation between titres and renal disease activity and positive predictive value in systemic lupus erythematosus. Ann Rheum Dis. 2004. in press. [DOI] [PMC free article] [PubMed]
  • 58.Siegert CE, Daha MR, Swaak AJ, van der Voort EA, Breedveld FC. The relationship between serum titers of autoantibodies to C1q and age in the general population and in patients with systemic lupus erythematosus. Clin Immunol Immunopathol. 1993;67:204–9. doi: 10.1006/clin.1993.1066. [DOI] [PubMed] [Google Scholar]
  • 59.Trendelenburg M, Courvoisier S, Spath PJ, et al. Hypocomplementemic urticarial vasculitis or systemic lupus erythematosus? Am J Kidney Dis. 1999;34:745–51. doi: 10.1016/S0272-6386(99)70402-6. [DOI] [PubMed] [Google Scholar]

Articles from Clinical and Experimental Immunology are provided here courtesy of British Society for Immunology

RESOURCES