Molecular Immunology 46 (2009) 1416–1423
Contents lists available at ScienceDirect
Molecular Immunology journal homepage: www.elsevier.com/locate/molimm
HLA-association of serum levels of natural antibodies Éva Pozsonyi a , Bence György b , Timea Berki c , Zsófia Bánlaki d , Edit Buzás b , Katalin Rajczy a , Adrienn Hossó a , Zoltán Prohászka d,e , Ágnes Szilágyi d , László Cervenak e , George Füst d,∗ a
National Blood Transfusion Service, Budapest, Hungary Department of Genetics, Cell- and Immunobiology, Semmelweis University, Budapest, Hungary Department of Immunology and Biotechnology, Pécs University, Pécs, Hungary d 3rd Department of Internal Medicine and Szentágothai Knowledge Center, Semmelweis University, Budapest, Hungary e Research Group of Inflammation Biology and Immunogenomics, Semmelweis University and Hungarian Academy of Sciences, Budapest, Hungary b c
a r t i c l e
i n f o
Article history: Received 12 November 2008 Received in revised form 8 December 2008 Accepted 16 December 2008 Available online 23 January 2009 Keywords: HLA MHC DR15 DR16 Natural autoantibodies Heat shock proteins Citrate synthase Chondroitin sulphate C Regulation Genetic regulation
a b s t r a c t Natural antibodies of IgM or IgG types are present in sera of most healthy individuals and are important participants of the immune response. Little is known, however, about the genetic regulation of their plasma levels in humans. We determined the concentrations of three IgM type natural autoantibodies (NAAbs) reactive to certain conserved self-antigens (citrate synthase (A-CIT), chondroitin sulphate C (ACOS) and 60 kDa heat shock proteins (A-HSP) in the sera of 78 healthy individuals and in their 86 children. In case of all the 164 individuals alleles of several polymorphisms were determined in class II (HLA-DQ, -DR), class III (AGER-429T>C, HSP70-2 1267A>G, TNF-308G>A, CFB S/F, copy number of the C4A and C4B genes), and class I (HLA-A, -B) regions of the major histocompatibility complex (MHC). Since the samples originated from a family study, extended MHC haplotypes were also determined for each study participant. Our results show that children of parents with low NAAb concentration have significantly lower serum concentrations of all the three NAAbs, as compared to offsprings of parents without reduced serum concentration. This indicates that the serum levels of these NAAbs were partly regulated by factors which are inherited from the parents to offsprings. In further studies performed only in genetically independent parents, we found significant differences in the serum levels of the IgM type A-CIT and A-COS antibodies (Abs) between carriers and non-carriers of the HLA-DR2 (15 and 16) antigens. In both cases the Ab concentrations were higher in the HLA-DR15 carriers (p = 0.002 and p = 0.008, respectively) and lower in DR16 carriers (p = 0.029 and p = 0.049, respectively) than in the non-carriers. Even more significant differences were found when the levels of two Abs were evaluated together. Frequency of the DR15 carriers was significantly lower among subjects with one or two low (in the lowest quartile) titers of A-CIT/ACOS Abs (p = 0.014), A-CIT/A-HSP Abs (p = 0.016) and A-COS/A-HSP Abs (p = 0.013) as compared to those with normal Ab titers for both antigens. By contrast, frequency of the DR16 carriers was significantly higher among subjects with one or two low A-CIT/A-COS Abs (p = 0.001), A-CIT/A-HSP Abs (p = 0.002) and A-COS/A-HSP Abs (p = 0.021) as compared to those with normal Ab titers for both antigens. Similar differences were found for both IgM type antibodies when carriers and non-carriers of the HLA-DR15DQ6 and HLA-DR16-DQ5 haplotypes were considered. These novel observations indicate that not only adaptive immune response but also natural autoantibody pattern, as a part of innate immune response, is influenced by the MHC allele composition. © 2009 Elsevier Ltd. All rights reserved.
1. Introduction Antibodies that react with self or foreign molecules detectable in the absence of known immunization with the target antigen are termed natural antibodies (Quintana and Cohen, 2004). This topic
∗ Corresponding author at: 3rd Department of Internal Medicine, Faculty of Medicine, Semmelweis University, Kútvölgyi út 4, H-1125 Budapest, Hungary. Tel.: +36 1 212 9351; fax: +36 1 212 9351. E-mail address:
[email protected] (G. Füst). 0161-5890/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.molimm.2008.12.006
was recently reviewed by several authors in a special issue of the Journal of Autoimmunity (Avrameas et al., 2007; Cohen, 2007; Lutz, 2007; Zelenay et al., 2007). A significant part of the natural antibodies, the so-called natural autoantibodies (NAAbs) react with internal constituents of the organism, they recognize various intracellular and cell surface antigens as well as circulating macromolecules and haptens highly conserved during evolution (Avrameas et al., 2007). In humans and different animal species, these NAAbs are of IgG, IgM or IgA isotypes. NAAbs seem to be conserved during evolution suggesting that they are not simple side-products of exogenous immunization but they might have a physiological role
É. Pozsonyi et al. / Molecular Immunology 46 (2009) 1416–1423
in the homeostasis of the body (Cohen, 2007; Quintana and Cohen, 2004). Tissue homeostasis in vertebrates comprises the ability of the body to (a) clear proteins released by dead cells, (b) eliminate non-functional plasma proteins, (c) recognize and dispose of tumor, dead, senescent and apoptotic cells and (d) contribute to the elimination of bacteria, viruses and other pathogenic agents (Lutz, 2007). According to the “immune homunculus” theory of Irun Cohen, these antibodies have a crucial role in the control of autoimmune diseases as well (Cohen and Young, 1991). Production of NAAbs is genetically controlled and apparently independent of environmental antigen stimulation. Several indirect data support this assumption. Even if the repertoire of NAAbs differs from one individual to another, within one individual it is stable with age (Lacroix-Desmazes et al., 1999; Mirilas et al., 1999; Mouthon et al., 1995). IgM type NAAbs can already be detected in newborn humans (Merbl et al., 2007) and animals as well as in germ-free and antigen-free mice. It seems that their repertoires are largely independent of external antigenic contacts (Haury et al., 1997). In elegant experiments Lacroix-Desmazes et al. (1999) measured immunoreactivity against different (kidney, lung, stomach and thymus) human tissue extracts in the serum samples of the same five healthy men at 43 ± 2 and 69 ± 3 years of age by using a quantitative immunoblotting technique. They found that the densitometric profiles of self-reactivity of serum IgM and of purified serum IgG remained unchanged during this 25 years interval. Several data obtained in mice (reviewed by Zelenay et al., 2007) and recently in HIV-infected patients (Stahl et al., 2005) support the concept that similarly to antibodies against non-self-antigens, T cells contribute to the selection of natural self-antibody repertoires. Since it is well known that immune reactivity against non-selfantigens is regulated by the genes encoded in class II region of the human major histocompatibility complex (MHC), it seemed reasonable to study if the NAAbs are also controlled by these genes. Vasconcellos et al. (1998), by measuring auto-reactivity in four different MHC-congenic mice strains, demonstrated that NAAb repertoires are controlled among others by the MHC genes. Surprisingly, however, to our best knowledge, no data are available on the NAAb levels in healthy human beings carrying different class II antigens. Therefore we have decided to measure IgM type NAAbs in the sera of 78 healthy individuals with different HLA-DR and DQ antigens. Haplotypes were determined in a family study. Levels of three different NAAbs (anti-heat shock protein 60, anti-citrate synthase and anti-chondroitine sulphate C) were measured in the serum samples of the test subjects. In order to assess the heritability of the NAAb levels we also measured the same IgM antibodies in serum samples of 86 offsprings of the test subjects. We selected these NAAbs for the study since our laboratory published several studies on the regulation of these antibodies in health and disease. IgG type antibodies reactive to the important intracellular 60 kDa heat shock protein (highly conserved in phylogeny), anti-hsp60, are known to be present in serum samples of most healthy individuals (Burian et al., 2001). We demonstrated that their level is related to the polymorphism of the IL-6 gene in healthy Finnish blood donors (Veres et al., 2002) and healthy Hungarian individuals (Kiszel et al., 2006). We also reported that the genes of IL-6 and that encoding immunoglobulin GM show an epistatic effect on the serum concentration of anti-hsp60 (Pandey et al., 2004). In addition, in our unpublished studies (Várbíró Sz et al., submitted for publication) we found marked stability in serum levels of anti-hsp60 antibodies in healthy middle-aged persons over a 5-year-long period of time. Citrate synthase, which is present in all organisms as a key enzyme in aerobic energy metabolism, is one of the phylogenetically most conserved enzymes. This mitochondrial inner-membrane enzyme has a similar ancestry with the chloroplasts. Both have evolved by endosymbiosis from a prokaryotic cell. The plasma membrane of a prokaryotic cell is the site of
1417
oxidative phosphorylation and later on became the inner membrane of mitochondria (Nemeth et al., 1991). In vivo analysis of the immunological recognition of, and tolerance to this enzyme is a suitable model to investigate the mechanism of physiological and pathological autoimmune reaction. Mainly IgM isotype citrate synthase autoantibodies were found in healthy controls (Petrohai et al., 2005). Recently we summarized our studies which indicate that the anti-mitochondrial citrate synthase autoantibodies are components of the natural antibody network (Czompoly et al., 2006). Chondroitin sulphate C is a carbohydrate antigen, it is a glycosaminoglycan attached covalently to a protein core in proteoglycans. According to our data, glycosaminoglycan-reactive IgM natural autoantibodies are highly cross-reactive with other carbohydrate structures, and are abundant in the sera of healthy adult humans, but are absent in neonates. Recently we have identified the level of natural autoantibodies to this T cell independent (TI2) carbohydrate antigen (chondroitin sulphate C) as a disease state marker in rheumatoid arthritis (B. Gyorgy et al., 2008). Here we report on a strong association between carrier state of HLA-DR15 and -DR16 and the low levels of these antibodies. 2. Materials and methods 2.1. Family study Samples were collected from healthy members of families (father, mother and at least one child) who were investigated for donor search for one of the family members. Participants were informed about the purpose of the study and they gave their informed consent for sample collection and analysis. The study was performed on 39 families involving 39 mothers, 39 fathers and 86 offspring. The number of children investigated was 1, 2, 3, 4 and 5 in 3, 28, 6, 1 and 1 families, respectively. All families included in the study were of Hungarian ethnic origin. The study was approved by the Ethical Committee of the Semmelweis University. 2.2. Sample collection EDTA-anticoagulated blood was used for the preparation of genomic DNA (with commercial (Puregene) kit), while serum in which natural antibodies were determined later on was separated from native blood samples immediately after coagulation. DNA and serum samples were kept at −30 and −80 ◦ C, respectively until used. 2.3. Determination of the MHC class II, class I and class III alleles in the DNA samples Serologic typing for HLA-A and HLA-B was performed using the standard microlymphocytotoxicity method (Innotrain Diagnostik GmbH, Kronberg, Germany), which defined the 24 HLA-A and the 48 HLA-B antigens. HLA-DRB1 and HLA-DQB1 medium resolution genotyping was performed using polymerase chain reaction single-strand oligonucleotide reverse dot-blot kits (InnoLipa DRB key and InnoLipa DQB kits, respectively; Innogenetics, Zwijndrecht, Belgium). HLA-DRB1 and -DQB1 low-resolution typing and DRB1*16 genotyping were performed by polymerase chain reaction with sequence-specific primers (Olerup SSP AB QIAGEN Vertriebs GmbH, Vienna, Austria). In case of the class III polymorphisms tested, AGER-429T>C SNP was determined by the method of Hudson et al. (2001), complement factor B (CFB) polymorphism was detected as described by Jahn et al. (1994) copy number of the C4A and C4B genes were determined as described earlier (Szilagyi et al., 2006) and HSP70-2 1267A>G SNP was tested by the method of Vargas-Alarcon et al. (2002). Lymphotoxin-alpha
1418
É. Pozsonyi et al. / Molecular Immunology 46 (2009) 1416–1423
Fig. 1. Serum concentrations of anti-citrate synthase (A-CIT), anti-chondroitin sulphate C (A-COS) and anti-hsp60 (A-HSP) natural autoantibodies (NAAbs) in children with no, one or both parents with low levels of the same IgM type NAAbs. p-Values for Kruskall–Wallis test are indicated. If evaluated after the Bonferroni correction, p-values for A-COS and A-HSP were significant (G SNP was determined using the method of Seidemann et al. (2005) with a minor modification applying the StyI restriction enzyme. A new Taqman® assay was developed applying the software Primer Express v.2.0 for genotyping the −308G>A polymorphism in the TNF-alpha gene. PCR was performed in an ABI 7300 Real Time PCR System using 1× TaqMan® Universal PCR Master Mix, 0.3 M forward (308F: AAAAGAAATGGAGGCAATAGGTT) and reverse (308R: GGCCACTGACTGATTTGTGTGT) primer and 0.2 M of each probe (308A: FAM-AACCCCGTCCTCATG-MGB; 308G: VICAACCCCGTCCCCATG-MGB) in a final volume of 10 l. 2.4. Measurement of the serum levels of the three NAAbs tested IgM type antibodies to 60 kDa heat shock proteins (anti-HSP60) were measured by ELISA assay. In brief, plates were coated with 0.10 g/well recombinant human Hsp60 (Lionex GmbH, Braunschweig, Germany). After washing and blocking (PBS, 0.5% gelatine) the wells were incubated with 50 l of serum samples diluted 1:100 in PBS containing 0.5% gelatine and 0.05% Tween 20. Binding of anti-Hsp autoantibodies was determined using -chain-specific anti-human IgM peroxidase-labeled antibodies (DAKO, Glostrup, Danemark) and o-phenylenediamine (Sigma) detection system. The optical density was measured at 490 nm (reference at 620 nm) and the means of duplicate wells were calculated. All data points were from the same set of measurements. The inter-assay coefficient of the method is less than 10%. Serum concentration of the IgM type antibodies to citrate synthase was determined in an indirect ELISA test, described by Petrohai et al. (2005). Briefly ELISA plates (Nunc, France) were coated with 1 g/ml citrate synthase from porcine heart (Sigma, USA) in carbonate buffer (pH 8.2) with the conventional adsorption method for 12 h at 4 ◦ C followed by an incubation for 1 h at 37 ◦ C. Following saturation of the non-specific binding sites with 0.5% gelatine (Sigma, USA) in PBS, plates were incubated with 1:100 dilutions of sera in duplicates for 1 h. After three washing
steps with PBS/Tween20, plates were incubated with anti-human IgM-HRPO secondary antibody (Dako, Denmark) for 1 h. The color reaction was developed with o-phenylenediamine (Sigma) and the O.D. was measured at 492 nm on a Dynatech MR7000 ELISA reader. Concentrations were expressed in O.D. values. IgM antibody levels reactivity with citrate synthase was determined in the same set of measurement. A negative and a positive standard sample were used to control the reproducibility of the ELISA tests. Levels of the IgM type antibodies to chondroitin sulphate C were measured using CovaLink ELISA. Briefly, chondroitin sulphate C (Sigma–Aldrich, St. Louis, MO) was covalently bound to the surface of the CovaLink plates (Nunc, Wiesbaden, Germany) using 1% 1-(3-dimethylaminopropyl)-3-ethylcarbodiimid (1% EDC, Merck Whitehouse Station, NJ) at 1 g/well for 2 h at 37 ◦ C, and then overnight at room temperature. Blocking was carried out using 1% PBS-BSA-Na azide. Sera were applied onto the plates using 1:100 dilution (a concentration selected after preliminary experiments). HRP-conjugated anti-human IgM and anti-human IgG (both from Sigma–Aldrich) were used in 1:50,000 and 1:30,000 dilutions, respectively. Using ortho-phenylene-diamine (Sigma–Aldrich) as chromogenic substrate, the absorbance was detected at 492 nm. Concentrations were expressed in O.D. values. All data points were from the same set of measurements. 2.5. Statistical analysis Statistical analysis was performed with Prism for Windows 4.02 (GraphPad Software, San Diego, CA, USA) statistical software product. As many of the variables had non-Gaussian distributions we used non-parametric tests in the analysis. We used the Mann–Whitney’s U-test to compare two independent groups, the 2 -test to compare categorical variables and Spearman’s to calculate correlations. All statistical analyses were performed two-tailed and p < 0.05 was considered as significant. At multiple comparison Bonferroni correction was performed.
É. Pozsonyi et al. / Molecular Immunology 46 (2009) 1416–1423
1419
Fig. 2. Differences in the serum concentrations of the autoantibodies to citrate synthase (left panel), chondroitin sulphate C (middle panel) and 60 kDa heat shock protein (right panel) between carriers and non-carriers of the HLA-DR15 antigen (upper row) and the HLA-DR15-DQ6 haplotype (lower row). p-Values for Mann–Whitney test are indicated.
3. Results 3.1. Correlation between the serum concentrations of the three types of NAAbs tested In the sera of 164 members of the 39 families we found highly significant correlation between the levels of the IgM anti-citrate synthase (A-CIT) and IgM anti-chondroitin sulphate C (A-COS) antibodies (Spearman correlation coefficient R = 0.590 (p < 0.0001)), between A-COS and IgM anti-hsp60 (A-HSP) antibodies (R = 0.369,
p < 0.0001) and between A-CIT and A-HSP antibodies (R = 0.363, p < 0.0001). When the analysis was restricted to the 78 genetically independent parents, the A-CIT and A-COS exhibited highly significant correlation (R = 0.517, p < 0.0001), A-COS and A-HSP significantly correlated (R = 0.299, p = 0.0078) while the correlation between ACIT and A-HSP was found close to significant (R = 0.228, p = 0.053). Since here three different comparisons (three hypotheses) were made according to the Bonferroni correction p < 0.05/3 that is p < 0.016 could be considered as significant. Accordingly the first
Fig. 3. Differences in the serum concentrations of the autoantibodies to citrate synthase (left panel), chondroitin sulphate C (middle panel) and 60 kDa heat shock protein (right panel) between carriers and non-carriers of the HLA-DR16 antigen (upper row) and the HLA-DR16-DQ5 haplotype (lower row). p-Values for Mann–Whitney test are indicated.
1420
É. Pozsonyi et al. / Molecular Immunology 46 (2009) 1416–1423
Table 1 Association between the HLA-DR15 antigen and low serum concentration of three different IgM type natural autoantibodies. Antibody
Titer
HLA-DR15
Anti-citrate synthasec
High Low
43 (77%) 17 (100%)
13 (23%) 0 (0%)
0.030a
Anti-chondritin sulphate C
High Low High Low
45 (76%) 19 (100%) 46 (78%) 18 (95%)
14 (24%) 0 (0%) 13 (22%) 1 (5%)
0.017a
Anti-citrate synthase–anti-chondritin sulphate C
Neither low Either low Both low
36 (73%) 13 (100%) 11 (100%)
13 (27%) 0 (0%) 0 (0%)
0.011b
Anti-citrate synthase–anti-hsp60
Neither low Either low Both low
32 (73%) 21 (95%) 7 (100%)
12 (27%) 1 (5%) 0 (0%)
0.013b
Anti-chondritin sulphate C–anti-hsp60
Neither low Either low Both low
35 (73%) 21 (95%) 8 (100%)
13 (27%) 1 (5%) 0 (0%)
0.011b
Anti-citrate synthase–anti-chondritin sulphate C–anti-hsp60
Neither low One or two low All three low
27 (69%) 27 (96%) 6 (100%)
12 (31%) 1 (4%) 0 (0%)
0.004b
Non-carriers
Anti-hsp60
p-Value Carriers
0.167a
Percentage of the HLA-DR15 carriers and non-carriers in the group of individuals with different levels of NAAbs. Low: in the lowest quartile and high: rest of subjects. a Fisher’s exact test. b 2 -test for trend. c This NAAb was not measured in five subjects. All values for double or triple comparisons were significant even after Bonferroni correction.
Table 2 Association between the HLA-DR16 antigen and low serum concentration of three different IgM type natural autoantibodies. Antibody
Titer
HLA-DR16
p-Value
Non-carriers
Carriers
Anti-citrate synthasec
Low High
10 (59%) 51 (91%)
7 (41%) 5 (9%)
0.005a
Anti-chondritin sulphate C
Low High
13 (68%) 53 (90%)
6 (32%) 6 (10%)
0.061a
Anti-hsp60
Low High
14 (74%) 52 (88%)
5 (26%) 7 (12%)
0.152a
Anti-citrate synthase–anti-chondritine sulphate
Neither low Either low Both low
45 (92%) 10 (77%) 6 (55%)
4 (8%) 3 (23%) 5 (45%)
0.002b
Anti-citrate synthase–anti-hsp60
Neither low Either low Both low
40 (91%) 18 (82%) 3 (43%)
4 (9%) 4 (18%) 4 (57%)
0.004b
Anti-chondritin sulphate C–anti-hsp60
Neither low Either low Both low
43 (90%) 19 (86%) 4 (50%)
5 (10%) 3 (14%) 4 (50%)
0.017b
Anti-citrate synthase–anti-chondritin sulphate C–anti-hsp60
Neither low One or two low All three low
36 (92%) 23 (82%) 2 (33%)
3 (8%) 5 (18%) 4 (67%)
0.002b
Percentage of the HLA-DR16 carriers and non-carriers in the group of individuals with different levels of NAAbs. Low: in the lowest quartile and high: rest of subjects. a Fisher’s exact test. b 2 -test for trend. c This NAAb was not measured in five subjects. All values for double or triple comparisons were significant even after Bonferroni correction except A-COS-A-HSP where a p-value of marginal significance was obtained.
two correlation coefficients are significant while the third is not. 3.2. Inheritance of the low NAAb levels from parents to children First we tested if serum concentration of the NAAbs measured in the parents and their children correlated with each other. Therefore we divided the parents into two groups according to the NAAbs levels: those with or without low NAAbs levels as defined in the lowest quartile of all values. Children with no, one or both parents
with low NAAbs levels were compared. The analysis was separately performed for each NAAb (Fig. 1). Children of parents with low NAAb concentration had significantly (for A-CIT of marginal significance) lower serum concentrations of all the three NAAbs, as compared to offsprings of parents without low serum concentration. However, children who had one parent with low levels of A-CIT and A-COS, had lower levels of the corresponding antibodies than children without such parents. On the other hand, in the case of A-HSP, only children whose both parents were low NAAb carriers, had decreased A-HSP levels.
É. Pozsonyi et al. / Molecular Immunology 46 (2009) 1416–1423
1421
3.3. Relationship between the carrier state of different HLA-DR and HLA-DQ antigens and the serum concentration of the NAAbs tested Using the non-parametric Mann–Whitney test, the median serum concentration of the A-CIT, A-COS and A-HSP antibodies were compared in carriers and non-carriers of the different HLADR antigens. The following DR antigens were carried by at least five individuals: HLA-DR1, -DR3, -DR4, -DR7, -DR11, -DR13, -DR15, and DR16. Significant difference was found only with the HLA-DR15 and -DR16 antigens. Carrier state of the HLA-DR15 antigen was associated with higher A-CIT (p = 0.002) and A-COS levels (p = 0.008) while in the case of A-HSP no significant differences were found (p = 0.289). The first two p-values were significant even after Bonferroni corrections (p < 0.016) (Fig. 2, upper row). By contrast, carriers of the HLA-DR16 antigen had lower A-CIT (p = 0.029) and A-COS (p = 0.049) serum concentration, whereas no significant difference was found in the A-HSP levels with this antigen, either (p = 0.873). After Bonferroni correction these values could not be considered significant (Fig. 3, upper row). Out of the HLA-DQ antigens -DQ2, -DQ4, DQ5, -DQ6 and -DQ7 were carried by at least five test subjects. When carriers and noncarriers of these antigens were compared for the levels of NAAbs, significant (p = 0.035) association was found only with the HLA-DQ6 antigens for the A-COS antibodies: carriers had higher titres than non-carriers. Since strong linkage disequilibrium exists between the HLA-DR15 and -DQ6 antigens (p < 0.001) and -DR16 and -DQ5 antigens (p < 0.001), we also tested the possible association of the NAAb levels with the carrier state of the HLA-DR15-DQ6 and HLADR16-DQ5 haplotypes. Carriers of the HLA-DR15-DQ6 haplotype had significantly (even after Bonferroni correction) higher A-CIT (p = 0.001) and A-COS levels (p = 0.013) while no association was found with the A-HSP levels (Fig. 2, lower row). Carrier state of the DR16-DQ5 haplotype was associated with lower A-CIT (p = 0.062) and A-COS levels (p = 0.094) but these differences were not significant (Fig. 3, lower row). Next we examined whether the association of the HLA-DR15 and -DR16 with the level of A-CIT and A-COS NAAbs were due to some conserved extended haplotypes (CEHs) which contained these antigens. We determined the extended haplotypes carried by the 78 parents based on pedigree analysis (data to be published). We found that parents who carried the HLA-DR15 antigens had five different CEHs. Half of the HLA-DR15 alleles were constituents of the 7.1 CEH but there was no significant difference in the NAAb levels between the 7.1 CEH positive and negative subjects (data not shown). CEHs involving the HLA-DR16 antigen were even more heterogenous, seven different CEHs contained this HLA-DR antigen. Similarly we found no association between class I and class III antigens and the serum concentration of NAABS tested when these antigens were individually evaluated (data not shown). These findings indicate that the association of these HLA-DR and -DQ antigens with the levels of the A-CIT and A-COS NAAbs were due to this region itself. 3.4. Association between the carrier state of HLA-DR15 as well as -DR16 and low levels of NAAbs tested Comparing the NAAb titres of the HLA-DR15 and -DR16 carriers and non-carriers it became clear that the main difference between the two groups at both comparisons is due to the less and more frequent occurrence of the low NAAb levels, respectively. Thus, we continued the analysis in this direction. Low levels of all three NAAbs tested were defined as levels in the lowest quartile of the values obtained in the whole group. As it is expected from the previously summarized analysis, low values of both A-CIT and ACOS titres were found with significantly lower frequency in the
Fig. 4. Additive effect of the DR15 non-carrier and DR16 carrier state on the serum concentration of three (anti-citrate synthase, anti-chondroitin suphate C, antihsp60) IgM type natural autoantibodies in 78 unrelated individuals, p-value for 2 -test is indicated.
HLA-DR15 carriers (Table 1) and significantly higher frequency in the -DR16 carriers (Table 2), while no significant differences were obtained in the occurrence of A-HSP values (Tables 1 and 2). However, when we compared the paired occurrence of the three NAAbs tested, we found that the HLA-DR15 was associated with the lack of occurrence of the low Abs titers not only in the case of the A-CIT/ACOS combination, but also with the A-CIT/A-HSP and A-COS/A-HSP combinations, and the difference was even more significant when all the three NAAbs were considered (Table 1). Quite similar results were obtained with the DR16 antigen except that in that case the low NAAb levels were positively associated with the HLA-DR16 carrier state (Table 2). 3.5. Analysis of the combined effect of the HLA-DR15 and -DR16 carrier state on the occurrence of low NAAb levels None of the analyzed individuals carried both DR15 and DR16 antigens. The test subjects were divided into three groups. Group 1 consisted of 12 subjects who had a DR15, DR16 combination not associated with low NAAb levels (DR15 carriers–DR16 non-carriers), 11 members of Group 3 did not carry either DR15 or DR16, while the rest of the subjects (n = 50) (Group 2) carried both genetic traits associated with low NAAb titres (DR15 non-carriers–DR16 carriers) (Fig. 4). None of the NAAb levels was low in the sera of the vast majority (11/12) of the subjects of Group 1. By contrast, this situation occurred in only 26/50 and 2/11 subjects of Group 2 and Group 3, respectively (p < 0.001). 4. Discussion Here we report on novel observations which indicate that serum concentrations of three NAAbs are genetically regulated. First, we found that children of one or two parents with low serum concentration of the three NAAbs (A-CIT, A-COS and A-HSP) tested had lower levels of the corresponding NAAbs than children of parents without reduced low serum concentration of these NAAbs. This observation suggests that serum level of these NAAbs is partly regulated by factors which are inherited from the parents to offspring. Second, serum concentrations of these NAAbs were found to be influenced in healthy individuals by two antigens (HLA-DR15, -DR16) encoded in class II, major immunoregulatory region of the major histocompatibility complex (MHC). Carriers of the DR15 antigen and those of the DR15-DQ6 haplotype had higher A-CIT and A-COS NAAb levels, while carrier state of DR16 antigen and less strongly DR16-DQ5 haplotype was associated with lower levels of these two NAAbs. No such association was observed for the A-HSP
1422
É. Pozsonyi et al. / Molecular Immunology 46 (2009) 1416–1423
NAAb, however when we performed paired analysis, we found that carriers of DR15 and DR16, respectively, had significantly lower and higher probability for low levels of all the three NAAbs tested. Since we performed a family study in which several HLA class I and class II antigens as well as SNPs and copy number polymorphisms in MHC class III region were also tested, we could determine haplotypes extending through the whole MHC (conserved extended haplotypes) in all parents. Since both DR15 and DR16 were involved in several different CEHs and not even the most frequent CEH (7.1 in DR15 carrier) was associated with the serum levels of either of these NAAbs, it seems that the observed association can be restricted to the HLA-DR, -DQ genes that is the major immunoregulatory region of the MHC. It is well known that natural autoantibodies are subjects for negative and positive T cell regulation (Zelenay et al., 2007). Autoantibodies recognizing citrate synthase (Czompoly et al., 2006) and 60 kDa heat shock protein (Burian et al., 2001) are natural autoantibodies. Although carbohydrates, such as chondroitin sulphate, are generally considered to fall into the category of T cell independent antigens, recent data suggest that this is not necessarily the case in all instances. A special subset of T cells has been reported to expand in the bone marrow of glycosaminoglycan (GAG)-immunized mice (Wang and Roehrl, 2002). On the other hand a given MHC pattern may influence which epitopes of proteins or other macromolecules are important for the central immune system. This is in turn may determine the whole natural antibody composition including those ones that are reactive to antigens not directly presented on classical MHC molecules. NAAbs may be monoreactive or polyreactive (Avrameas et al., 2007), therefore our findings can be due either to HLA-association of one type of polyreactive antibodies, which reacts with CIT, COS and to a lesser extent HSP antigens as well. A-COS NAAb were recently found to be highly cross-reactive. This was reflected by the strong statistical correlation among levels of IgM type antibodies reactive with different glycosaminoglycans. We have also shown broad cross-reactivity among different types of GAGs in inhibition studies. Furthermore we have found that circulating A-COS antibodies also reacted with bacterial peptidoglycans and the fungal polysaccharide zymosan (Gyorgy et al., 2008). Observed association may be explained by a common MHC-related regulation of all the three NAAbs tested which, if further studies support the assumption, may be a common mechanism for most or all NAAbs in general. Clearly further studies are necessary in order to select between the two possibilities. These observations are in line with the fairly consistent reactivity repertoires of NAAbs during life (Lacroix-Desmazes et al., 1999; Mirilas et al., 1999; Mouthon et al., 1995). NAAbs of healthy individuals are essentially of germ-line type with very few somatic mutations, while disease-associated autoantibodies are determined by somatic mutations as well (Avrameas and Ternynck, 1993; Zelenay et al., 2007). New data, however, indicate that both types of autoantibodies are subjects to positive and negative T cell regulation. Thus, our present observations indicating that serum levels of some NAAbs are up- or down-regulated by two antigens (HLADR15 and -DR16) in the immunoregulatory region of MHC, does not seem to be surprising. On the other hand, if Cohen’s homunculus theory (Cohen and Young, 1991) is correct, the level of NAAbs may not be influenced by MHC directly, but it rather reflects to a HLAcontrolled T and B cell pattern as the antibodies and T cells interact with each other. As we already pointed out, no data have been reported on the association of NAAb levels in healthy individuals with either the DR15 or DR16 (or any other) class II MHC antigens. However, abundant amounts of information indicate that many pathological autoantibodies are related to one or the other of the two antigens. HLA-DR15 was found to be strongly associated with synthesis of the anti-SSA antibodies in patients with primary Sjögren syndrome
(Gottenberg et al., 2003). The HLA-DR15 antigen is a strong risk factor for multiple sclerosis as evidenced by recent whole genome scan studies (Lincoln et al., 2005; Prat et al., 2005; Sawcer et al., 2005). It is associated with cellular immune response against myelin antigens (Martin et al., 1992), but it is independent of the anti-EBNA antibody titers, a risk factor for multiple sclerosis (De Jager et al., 2008). HLA-DR15 was also found to up-regulate anti-DNA antibody titres in patients with SLE (Podrebarac et al., 1998), and it was found to be a risk factor of the immune-mediate aplastic anaemia (Marsh and Gordon-Smith, 2004; Sugimori et al., 2007) and idiopathic dilated cardiomyopathy (Liu et al., 2006) as well. The HLA-DR16 antigen seems to be associated mainly with susceptibility for immune-mediated pathological sequels of infections, such as rheumatoid heart disease after streptococcus infections (Hernandez-Pacheco et al., 2003) or heart damage in Chagas disease caused by the protozoon Trypanosoma cruzi (Cruz-Robles et al., 2004). In individuals naturally exposed to malaria, DR16 carriers develop antibodies to the circumsporozoite protein repeat antigen of Plasmodium vivax more frequently than non-carriers (OliveiraFerreira et al., 2004). The three NAAbs we tested in the present work may also have relevance to autoimmune diseases. A-COS level was recently shown to be increased in rheumatoid arthritis because of the increased efflux of GAGs from the inflamed joints. Higher A-COS levels were found to be associated with less severe diseases suggesting a protective role of the A-COS NAAb (Gyorgy et al., 2008). Earlier we found high levels of antibodies against hsp60 in patients with undifferentiated connective tissue disease (Horvath et al., 2001). Recently a protective role of the IgM type anti-hsp60 antibodies was suggested for the development of systemic necrotizing vasculitis (Sherer et al., 2008) and an important role of the hsp60 family was shown in the pathogenesis of murine model of SLE as well (Marengo et al., 2008). While A-CIT NAAb was found to be able to recognize a target antigen (nucleosome) in a systemic autoimmune disease (Czompoly et al., 2006). Thus, it seems that our present findings on HLA-DR15 and -DR16 related regulation of NAAb levels in healthy individuals are in line with some well known pathological associations. Therefore studying the mechanisms of the MHC-regulation of autoantibodies may be relevant for understanding the pathomechanisms of autoimmune diseases as well. Acknowledgements We are greatly indebted to the study participants. This study was supported by the National Office for Research and Technology, Hungary, the OTKA T049266 (G.F.), the OTKA NF72689 (Z.P.) and OTKA (K 73247) (E.B.) grants of the Hungarian Research Fund (Z.P.), the GVOP-2004-05-0537/3.0 grant of National Office for Research and Technology, Hungary (T.B.) and by the Marie Curie (MRTN-CT2005-019561) grant of E.B. References Avrameas, S., Ternynck, T., 1993. The natural autoantibodies system: between hypotheses and facts. Mol. Immunol. 30, 1133–1142. Avrameas, S., Ternynck, T., Tsonis, I.A., Lymberi, P., 2007. Naturally occurring B-cell autoreactivity: a critical overview. J. Autoimmun. 29, 213–218. Burian, K., Kis, Z., Virok, D., Endresz, V., Prohaszka, Z., Duba, J., Berencsi, K., Boda, K., Horvath, L., Romics, L., Fust, G., Gonczol, E., 2001. Independent and joint effects of antibodies to human heat-shock protein 60 and Chlamydia pneumoniae infection in the development of coronary atherosclerosis. Circulation 103, 1503–1508. Cohen, I.R., 2007. Biomarkers, self-antigens and the immunological homunculus. J. Autoimmun. 29, 246–249. Cohen, I.R., Young, D.B., 1991. Autoimmunity, microbial immunity and the immunological homunculus. Immunol. Today 12, 105–110. Cruz-Robles, D., Reyes, P.A., Monteon-Padilla, V.M., Ortiz-Muniz, A.R., VargasAlarcon, G., 2004. MHC class I and class II genes in Mexican patients with Chagas disease. Hum. Immunol. 65, 60–65. Czompoly, T., Olasz, K., Simon, D., Nyarady, Z., Palinkas, L., Czirjak, L., Berki, T., Nemeth, P., 2006. A possible new bridge between innate and adaptive immunity: are the
É. Pozsonyi et al. / Molecular Immunology 46 (2009) 1416–1423 anti-mitochondrial citrate synthase autoantibodies components of the natural antibody network? Mol. Immunol. 43, 1761–1768. De Jager, P.L., Simon, K.C., Munger, K.L., Rioux, J.D., Hafler, D.A., Ascherio, A., 2008. Integrating risk factors: HLA-DRB1*1501 and Epstein-Barr virus in multiple sclerosis. Neurology 70, 1113–1118. Gottenberg, J.E., Busson, M., Loiseau, P., Cohen-Solal, J., Lepage, V., Charron, D., Sibilia, J., Mariette, X., 2003. In primary Sjogren’s syndrome HLA class II is associated exclusively with autoantibody production and spreading of the autoimmune response. Arthritis Rheum. 48, 2240–2245. Gyorgy, B., Tothfalusi, L., Nagy, G., Pasztoi, M., Geher, P., Lorinc, Z., Polgar, A., Rojkovich, B., Ujfalussy, I., Poor, G., Pocza, P., Wiener, Z., Misjak, P., Koncz, A., Falus, A., Buzas, E.I., 2008. Natural autoantibodies reactive with glycosaminoglycans in rheumatoid arthritis. Arthritis Res. Ther. 10, R110. Haury, M., Sundblad, A., Grandien, A., Barreau, C., Coutinho, A., Nobrega, A., 1997. The repertoire of serum IgM in normal mice is largely independent of external antigenic contact. Eur. J. Immunol. 27, 1557–1563. Hernandez-Pacheco, G., Aguilar-Garcia, J., Flores-Dominguez, C., Rodriguez-Perez, J.M., Perez-Hernandez, N., Alvarez-Leon, E., Reyes, P.A., Vargas-Alarcon, G., 2003. MHC class II alleles in Mexican patients with rheumatic heart disease. Int. J. Cardiol. 92, 49–54. Horvath, L., Czirjak, L., Fekete, B., Jakab, L., Prohaszka, Z., Cervenak, L., Romics, L., Singh, M., Daha, M.R., Fust, G., 2001. Levels of antibodies against C1q and 60 kDa family of heat shock proteins in the sera of patients with various autoimmune diseases. Immunol. Lett. 75, 103–109. Hudson, B.I., Stickland, M.H., Futers, T.S., Grant, P.J., 2001. Effects of novel polymorphisms in the RAGE gene on transcriptional regulation and their association with diabetic retinopathy. Diabetes 50, 1505–1511. Jahn, I., Mejia, J.E., Thomas, M., Darke, C., Schroder, H., Geserick, G., Hauptmann, G., 1994. Genomic analysis of the F subtypes of human complement factor B. Eur. J. Immunogenet. 21, 415–423. Kiszel, P., Fust, G., Pessi, T., Hurme, M., Prohaszka, Z., 2006. Associations between interleukin-6 genetic polymorphisms and levels of autoantibodies to 60-kDa heat-shock proteins. Hum. Hered. 62, 77–83. Lacroix-Desmazes, S., Mouthon, L., Kaveri, S.V., Kazatchkine, M.D., Weksler, M.E., 1999. Stability of natural self-reactive antibody repertoires during aging. J. Clin. Immunol. 19, 26–34. Lincoln, M.R., Montpetit, A., Cader, M.Z., Saarela, J., Dyment, D.A., Tiislar, M., Ferretti, V., Tienari, P.J., Sadovnick, A.D., Peltonen, L., Ebers, G.C., Hudson, T.J., 2005. A predominant role for the HLA class II region in the association of the MHC region with multiple sclerosis. Nat. Genet. 37, 1108–1112. Liu, W., Li, W.M., Yang, S.S., Gao, C., Li, S.J., Li, Y., Kong, Y.H., Gan, R.T., 2006. Association of HLA class II DRB1 DPA1 and DPB1 polymorphism with genetic susceptibility to idiopathic dilated cardiomyopathy in Chinese Han nationality. Autoimmunity 39, 461–467. Lutz, H.U., 2007. Homeostatic roles of naturally occurring antibodies: an overview. J. Autoimmun. 29, 287–294. Marengo, E.B., de Moraes, L.V., Faria, M., Fernandes, B.L., Carvalho, L.V., Tambourgi, D.V., Rizzo, L.V., Portaro, F.C., Camargo, A.C., Sant’anna, O.A., 2008. Administration of M. leprae Hsp65 interferes with the murine lupus progression. PLoS ONE 3, e3025. Marsh, J.C., Gordon-Smith, E.C., 2004. Insights into the autoimmune nature of aplastic anaemia. Lancet 364, 308–309. Martin, R., McFarland, H.F., McFarlin, D.E., 1992. Immunological aspects of demyelinating diseases. Annu. Rev. Immunol. 10, 153–187. Merbl, Y., Zucker-Toledano, M., Quintana, F.J., Cohen, I.R., 2007. Newborn humans manifest autoantibodies to defined self molecules detected by antigen microarray informatics. J. Clin. Invest. 117, 712–718. Mirilas, P., Fesel, C., Guilbert, B., Beratis, N.G., Avrameas, S., 1999. Natural antibodies in childhood: development, individual stability, and injury effect indicate a contribution to immune memory. J. Clin. Immunol. 19, 109–115. Mouthon, L., Nobrega, A., Nicolas, N., Kaveri, S.V., Barreau, C., Coutinho, A., Kazatchkine, M.D., 1995. Invariance and restriction toward a limited set of self-antigens characterize neonatal IgM antibody repertoires and prevail in autoreactive repertoires of healthy adults. Proc. Natl. Acad. Sci. U.S.A. 92, 3839–3843.
1423
Nemeth, P., Small, W.C., Evans, C.T., Zhi, W., Persson, L.O., Srere, P.A., 1991. Immunological mapping of fine molecular surface structures of citrate synthase enzymes from different cell types. J. Mol. Recognit. 4, 77–83. Oliveira-Ferreira, J., Pratt-Riccio, L.R., Arruda, M., Santos, F., Ribeiro, C.T., Goldberg, A.C., Banic, D.M., 2004. HLA class II and antibody responses to circumsporozoite protein repeats of P. vivax (VK210 VK247 and P. vivax-like) in individuals naturally exposed to malaria. Acta Trop. 92, 63–69. Pandey, J.P., Prohaszka, Z., Veres, A., Fust, G., Hurme, M., 2004. Epistatic effects of genes encoding immunoglobulin GM allotypes and interleukin-6 on the production of autoantibodies to 60- and 65-kDa heat-shock proteins. Genes Immun. 5, 68–71. Petrohai, A., Nagy, G., Bosze, S., Hudecz, F., Zsiros, E., Paragh, G., Nyarady, Z., Nemeth, P., Berki, T., 2005. Detection of citrate synthase-reacting autoantibodies after heart transplantation: an epitope mapping study. Transpl. Int. 17, 834–840. Podrebarac, T.A., Boisert, D.M., Goldstein, R., 1998. Clinical correlates, serum autoantibodies and the role of the major histocompatibility complex in French Canadian and non-French Canadian Caucasians with SLE. Lupus 7, 183–191. Prat, E., Tomaru, U., Sabater, L., Park, D.M., Granger, R., Kruse, N., Ohayon, J.M., Bettinotti, M.P., Martin, R., 2005. HLA-DRB5*0101 and -DRB1*1501 expression in the multiple sclerosis-associated HLA-DR15 haplotype. J. Neuroimmunol. 167, 108–119. Quintana, F.J., Cohen, I.R., 2004. The natural autoantibody repertoire and autoimmune disease. Biomed. Pharmacother. 58, 276–281. Sawcer, S., Ban, M., Maranian, M., Yeo, T.W., Compston, A., Kirby, A., Daly, M.J., De Jager, P.L., Walsh, E., Lander, E.S., Rioux, J.D., Hafler, D.A., Ivinson, A., Rimmler, J., Gregory, S.G., Schmidt, S., Pericak-Vance, M.A., Akesson, E., Hillert, J., Datta, P., Oturai, A., Ryder, L.P., Harbo, H.F., Spurkland, A., Myhr, K.M., Laaksonen, M., Booth, D., Heard, R., Stewart, G., Lincoln, R., Barcellos, L.F., Hauser, S.L., Oksenberg, J.R., Kenealy, S.J., Haines, J.L., 2005. A high-density screen for linkage in multiple sclerosis. Am. J. Hum. Genet. 77, 454–467. Seidemann, K., Zimmermann, M., Book, M., Meyer, U., Burkhardt, B., Welte, K., Reiter, A., Stanulla, M., 2005. Tumor necrosis factor and lymphotoxin alfa genetic polymorphisms and outcome in pediatric patients with non-Hodgkin’s lymphoma: results from Berlin-Frankfurt-Munster Trial NHL-BFM 95. J. Clin. Oncol. 23, 8414–8421. Sherer, Y., Pagnoux, C., Chironi, G., Simon, A., Guillevin, L., Mouthon, L., Gilburd, B., Shoenfeld, Y., 2008. Carotid artery intima-media thickness, heat shock proteins and oxidized LDL autoantibodies in systemic necrotizing vasculitis. Rheumatol. Int. 28, 1099–1103. Stahl, D., Lacroix-Desmazes, S., Misra, N., Karmochkine, M., Kaveri, S.V., Costagliola, D., Sibrowski, W., Kazatchkine, M.D., 2005. Alterations of self-reactive antibody repertoires in HIV disease: an insight into the role of T cells in the selection of autoreactive B cells. Immunol. Lett. 99, 198–208. Sugimori, C., Yamazaki, H., Feng, X., Mochizuki, K., Kondo, Y., Takami, A., Chuhjo, T., Kimura, A., Teramura, M., Mizoguchi, H., Omine, M., Nakao, S., 2007. Roles of DRB1 *1501 and DRB1 *1502 in the pathogenesis of aplastic anemia. Exp. Hematol. 35, 13–20. Szilagyi, A., Blasko, B., Szilassy, D., Fust, G., Sasvari-Szekely, M., Ronai, Z., 2006. Realtime PCR quantification of human complement C4A and C4B genes. BMC Genet. 7, 1. Vargas-Alarcon, G., Londono, J.D., Hernandez-Pacheco, G., Gamboa, R., Castillo, E., Pacheco-Tena, C., Cardiel, M.H., Granados, J., Burgos-Vargas, R., 2002. Heat shock protein 70 gene polymorphisms in Mexican patients with spondyloarthropathies. Ann. Rheum. Dis. 61, 48–51. Vasconcellos, R., Nobrega, A., Haury, M., Viale, A.C., Coutinho, A., 1998. Genetic control of natural antibody repertoires. I. IgH MHC and TCR beta loci. Eur. J. Immunol. 28, 1104–1115. Veres, A., Prohaszka, Z., Kilpinen, S., Singh, M., Fust, G., Hurme, M., 2002. The promoter polymorphism of the IL-6 gene is associated with levels of antibodies to 60-kDa heat-shock proteins. Immunogenetics 53, 851–856. Wang, J.Y., Roehrl, M.H., 2002. Glycosaminoglycans are a potential cause of rheumatoid arthritis. Proc. Natl. Acad. Sci. U.S.A. 99, 14362–14367. Zelenay, S., Moraes Fontes, M.F., Fesel, C., Demengeot, J., Coutinho, A., 2007. Physiopathology of natural auto-antibodies: the case for regulation. J. Autoimmun. 29, 229–235.
