- Email: [email protected]
Methyl Mercury-Induced Autoimmunity in Mice
Toxicology and Applied Pharmacology 154, 203–211 (1999) Article ID taap.1998.8576, available online at http://www.idealibrary.com on
Methyl Mercury-Induced Autoimmunity in Mice Per Hultman and Hele´n Hansson-Georgiadis Division of Molecular and Immunological Pathology, Department of Health and Environment, Linko¨ping University, S-581 85 Linko¨ping, Sweden Received April 7, 1998; accepted September 30, 1998
Methyl Mercury-Induced Autoimmunity in Mice. Hultman, P., and Hansson-Georgiadis, H. (1999). Toxicol. Appl. Pharmacol. 154, 203–211. Female SJL/N, A.SW, B10.S (H-2s), BALB/C, DBA/2 (H-2d), A.TL and B10.TL (H-2t1) mice were treated with sc injections of 1.0 mg CH3HgCl/kg body weight every third day for 4 weeks. Controls were given sterile, isotonic NaCl. CH3HgCl (MeHg) induced in SJL, A.SW and B10.S mice antinucleolar antibodies (ANoA) targeting the nucleolar 34-kDa protein fibrillarin. The susceptibility to develop ANoA in response to MeHg was linked to the mouse major histocompatibility complex (H-2), since H-2s but not H-2t1 mice sharing background (non-H-2) genes developed ANoA. However, the background genes decided the strength of the ANoA response in the susceptible H-2s mice, and the ANoA titer was in the order: A.SW > SJL > B10.S. Although MeHg as well as inorganic mercury induced ANoA, the two forms of mercury differed both quantitatively and qualitatively in their effect on the immune system. MeHg induced in H-2s mice a weaker general (polyclonal) and specific (ANoA) B-cell response than HgCl2, probably due to weaker activation of Th2 cells with lower IL-4 production, as indicated by the minimal increase in serum IgE. The A.TL strain with a susceptible genetic background, but a H-2 haplotype resistant to HgCl2, responded to MeHg with a modest polyclonal B-cell response dominated by Th1-associated Ig isotypes. H-2s mice treated with MeHg showed in contrast to HgCl2treated mice no systemic immune-complex (IC) deposits, which may be due to the weaker immune activation after MeHg treatment. The increase in serum IgE concentration and ANoA titer 2– 6 weeks after stopping treatment with MeHg is identical to reactions during the first 2–3 weeks of HgCl2 treatment. Therefore, demethylation of MeHg probably increased the concentration of inorganic mercury in the body sufficiently to reactivate the immune system. This reactivation indicated that genetically susceptible mice are not resistant to challenge with mercury, making them distinctly different from rats. © 1999 Academic Press
Inorganic mercury is, due to dental amalgam fillings, the principal source of mercury exposure in the general population. The average mean daily retention of methyl mercury (MeHg) in man has been estimated to 2.3 mg Hg/day, 99% of which is due to fish consumption (Clarkson, 1992). MeHg retention comprises 86% of the total mercury retention derived from nondental sources (WHO, 1991). However, the effect of MeHg
on the immune system has only been sparsely investigated. Nakatsuru et al. found that MeHg in vitro inhibits T- and B-cell activation (Nakatsuru et al., 1985), and MeHg has been shown to suppress humoral immunity in vivo (Blakely et al., 1980; Koller, 1973). Natural killer cell activity is severely reduced in adult BALB/C mice exposed to MeHg (Ilba¨ck, 1991). Indirect (perinatal) exposure to MeHg in mice (Thuvander et al., 1996) and rats (Wild et al., 1997) leads to an enhanced response to B-cell mitogens. A slightly increased response to T-cell mitogen is found in adult mice (Ilba¨ck, 1991) and rats (Ortega et al., 1994) exposed to MeHg. Genetically susceptible Brown Norway rats given MeHg orally or by the respiratory tract develop a systemic autoimmune condition similar to what is seen after mercuric chloride (HgCl2) treatment (Bernaudin et al., 1981). The interaction of mercury with the immune system is complex. Inorganic mercury induces, within a narrow dose range, proliferation of lymphocytes derived from humans (Caron et al., 1970) and a number of animals species (Pauly et al., 1969). The proliferating cells have been identified as T-cells in humans (Nordlind, 1984), rats (Pelletier et al., 1988), and mice (Jiang and Mo¨ller, 1995; Reardon and Lucas, 1987). Similarly, certain inbred mouse strains develop a strong lymphoproliferation, including both CD41 and CD81 T-cells, while other strains respond with only a slight proliferation of CD81 T-cells (Jiang and Mo¨ller, 1995). The genetic susceptibility to inorganic mercury is even more pronounced in vivo. Parenteral and oral treatment with HgCl2 and exposure to elemental mercury in the form of vapor in mouse strains of the H-2s haplotype induces systemic autoimmunity including antinucleolar antibodies (ANoA) targeting fibrillarin (AFA) and immune-complex (IC)-deposits (Hultman et al., 1992, 1989a, 1996; Hultman and Enestro¨m, 1992; Reuter et al., 1989; Warfvinge et al., 1995). Treatment of other inbred mouse strains, such as BALB/C (H-2d), with inorganic mercury leads to a T-celldependent immune response with systemic IC deposits but no ANoA/AFA (Hultman and Enestro¨m, 1987). Inorganic mercury induces a strong and broad (polyclonal) activation of the immune system in mice of the H-2s haplotype (Al-Balaghi et al., 1996; Hultman and Enestro¨m, 1987; Hultman et al., 1996; Johansson et al., 1997b; Pietsch et al., 1989; van Vliet et al., 1993), but the antibodies in this polyclonal response differ with respect to duration and isotype pattern from the simultaneously appearing AFA (Johansson et al.,
203
0041-008X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
204
HULTMAN AND HANSSON-GEORGIADIS
1998). Treatment with silver in H-2s mice causes only a slight activation of the immune system and no polyclonal B-cell response. However, the maximum ANoA titer does not differ significantly between mice given AgNO3 and HgCl2 (Johansson et al., 1997a, 1998). This clearly illustrates that a strong and polyclonal activation of the immune system, as seen after treatment with inorganic mercury, is not necessary for induction of ANoA/AFA. Instead, recent studies show that mercury causes alterations in the structure and processing of fibrillarin (Griem and Gleichmann, 1995; Pollard et al., 1997), which gives rise to fibrillarin fragments with cryptic epitopes. T-cell clones with specificity for modified fibrillarin antigens have recently been identified (Kubicka-Muranyi et al., 1996). This strongly indicates that AFA arise by antigen-specific mechanism and not as part of a polyclonal activation. The ability of organic mercury, especially MeHg, to induce systemic autoimmunity in genetically susceptible mouse strains (Hultman et al., 1992; Mirtcheva et al., 1989) has so far not been studied. Using the popliteal lymph node test, StillerWinkler et al. (1988) found that a strain resistant to the lymphoproliferative effect of inorganic mercury could be susceptible if MeHg was used. We have therefore examined to what extent organic mercury, in the form of MeHg, activates the immune system and causes systemic autoimmunity in a number of strains genetically susceptible or resistant to inorganic mercury (Hultman et al., 1992, 1993, 1996; Mirtcheva et al., 1989). The same dose of MeHg and HgCl2 was used in the present and these previous studies, respectively. The MeHg exposure in our mice was ca. 350 mg Hg/kg/day, which is much higher than the exposure recorded in highly exposed populations such as Polar Eskimos (ca. 5 mg Hg/kg/day) (Hansen and Danscher, 1997). MATERIALS AND METHODS Mice. Female SJL/N and A.SW (H-2s), and BALB/C and DBA/2 (H-2d) mice were obtained from Bommice Breeding and Research Centre (Ry, Denmark). Female B10.S (H-2s) and A.TL (H-2t1) mice were from Harlan Ltd. (Oxon, England). B10.TL mice, originally obtained from Department of Immunogenetics, University of Tu¨bingen, Germany, were maintained by brothersister mating at our animal facility department, and female progeny were used in this study. All mice were 9 –13 weeks old at onset of the experiments. The mice were housed under 12-h dark-12-h light cycles in a high-barrier unit, kept in steel-wire cages with beechwood bedding, and given sterilized pellets (Type R 36, Lactamin, Vadstena, Sweden) and tap water ad libitum. Treatment. Treatment consisted of giving 1.0 mg CH3HgCl/kg body wt diluted in sterile 0.84% NaCl as a 0.1 ml sc injection on the dorsum every third day for 27 days. Controls received an equal volume of sterile 0.84% NaCl. In the first set of experiments groups of 8 –11 mice were injected with CH3HgCl or sterile, isotonic NaCl as above for 4 weeks, which were followed with blood sampling 2, 3, 4, 6, and 10 weeks after onset of injections. One B10.TL mouse treated with CH3HgCl (MeHg) died after blood sampling 3 weeks after onset of injections. One control A.SW mouse died at blood sampling after 6 weeks. Among the SJL mice, 2 given MeHg died after blood sampling at 4 weeks, and 2 controls died at 4 and 6 weeks, respectively. One MeHg-treated and two control A.TL mice died after blood sampling at 3 weeks. Among the BALB/C, DBA/2, B10.S, and B10.TL mice, only a single MeHg-treated BALB/C mouse
died at 2 weeks. In the second set of experiments, groups of 7 A.SW, A.TL, and BALB/C mice were treated with CH3HgCl or NaCl as described above for 4 weeks, and blood samples were obtained before treatment as well as after 2, 3, 4, and 6 weeks. Serum Ig concentrations assessed by ELISA. Analysis of serum Ig concentrations was performed as previously described (Johansson et al., 1997a). Briefly, microtiter plates (Nunc, Copenhagen, Denmark) were coated with rat anti-mouse IgG1 and rat anti-mouse IgM MAb (LO-IMEX, Brussels, Belgium), respectively. Following blocking, the wells were incubated with diluted serum. Bound IgG1 and IgM were detected using diluted horseradish peroxidase (HRP)-conjugated rat anti-mouse and rat anti-mouse IgM MAb (LOIMEX), respectively. The plates were read at OD450 and the background values in wells coated with PBS instead of serum were subtracted. Actual concentration was obtained from a standard curve using mouse myeloma proteins of the IgG1 and IgM isotype (LO-IMEX), respectively. For serum IgG2a analysis, microtiter wells (Nunc) were coated with affinity-purified goat anti-mouse IgG2a isotype (Southern Biotechnology). The wells were incubated with diluted sera, and bound IgG2a was detected with alkaline phosphatase (ALP) conjugated goat anti-mouse IgG (Southern Biotechnology). The IgG2a concentration in the samples was obtained from a standard curve using mouse myeloma proteins of the IgG2a isotype (LO-IMEX). Serum IgE was determined as described before (Warfvinge et al., 1995). Briefly, microtiter plates were coated with rat anti-mouse IgE (Southern Biotechnology), followed by blocking and incubation with diluted serum. Bound IgE was detected by HRP-conjugated goat anti-mouse IgE (Nordic Immunological Lab, Tilburg, Netherlands), and the IgE concentration in the samples was derived from a standard curve using mouse myeloma protein of the IgE isotype (Sigma). Serum anti-ssDNA antibodies assessed by ELISA. The method used has been described before (Johansson et al., 1997a). Microtiter plates were coated overnight with ssDNA, washed with PBS-Tween 20, blocked with BSATween 20-PBS, and repeatedly washed first with PBS-Tween and then with PBS. Sera diluted in 1% BSA-PBS were incubated in the wells, and the plates washed six times with BSA-Tween-PBS and incubated with ALP-conjugated rabbit anti-mouse Ig (reacting with IgG, IgM and IgA) (Sigma) diluted in BSA-Tween-PBS. The plates were repeatedly washed, substrate was added, reaction was stopped with 3 M NaOH, optical density was read at OD405, and the background was subtracted. A pool of sera from MRL-mice was used as positive control. Using a monoclonal antibody (clone HB2) reacting with dsDNA (SeraLab), we detected no contamination with dsDNA in the coating (data not shown). For detection of anti-ssDNA ab of the IgG and IgM type, ALP-conjugated goat anti-mouse IgG and IgM antibodies (Caltag Laboratories, San Francisco, CA) were used. Serum anti-DNP antibodies assessed by ELISA. The method used has been described before (Johansson et al., 1997a). Microtiter plates (Nunc) were coated over night with human serum albumin conjugated with 30 – 40 mol DNP per mol albumin (Sigma). Following repeated washes with BSA-PBS, the wells were incubated with diluted sera and washed, and ALP-conjugated rabbit anti-mouse Ig (reacting with IgG, IgM, and IgA) (Sigma) was added. After repeated washes with BSA-PBS, substrate was added, and the reaction was stopped with 3 M NaOH. The optical density was read at OD405, and background values were subtracted. For detection of anti-DNP abs of the IgG and IgM type, ALP-conjugated goat anti-mouse IgG and IgM antibodies (Caltag Laboratories) were used. Mouse anti-DNP antibodies (LO-IMEX) were used as positive controls. For detection of anti-DNP abs of the IgG1 and IgG2a isotype, we used ALP-conjugated rat anti-mouse IgG1 and IgG2a antibodies (Pharmingen Inc., CA). Serum antinuclear antibodies assessed by indirect immunofluorescence. The presence, pattern, and titer of serum antinuclear antibodies of the IgG class was determined by indirect immunofluorescence using HEp-2 cells as a substrate (Hultman et al., 1989b). Serum antinuclear antibodies assessed by immunoblotting. The specificity of the antinuclear antibodies in the serum was assessed by immunoblotting as described before with minor modifications (Warfvinge et al., 1995). Briefly,
METHYL MERCURY-INDUCED AUTOIMMUNITY
205
FIG. 1. Serum Ig concentration in mice treated with sc injection of 1.0 mg MeHg/kg body wt. (F) or 0.1 ml NaCl (E) every third day for 4 weeks. The results from identically MeHg-treated mice and controls in a second set of experiments are also shown: and ƒ, respectively. The IgE concentration after sc injections of an equimolar amount of HgCl2 to A.SW and BALB/C mice is included for comparison (r). *p , 0.05; **p , 0.01; ***p , 0.001 comparing MeHgand NaCl-treated animals using ANOVA followed by Bonferroni post test. Bars are 6 SD.
rat liver nuclei were isolated (Chan and Pollard, 1992), aliquots of the boiled nuclei were SDS-PAGE separated using a gradient gel, and electrophoretic transfer to nitrocellulose membranes was performed for 1.5 h at 0.8 mA/cm2 under water cooling (Semiphor Transfer Unit, Hoefer Scientific Instruments, San Francisco, CA). Nitrocellulose strips were blocked in a solution of PBS-5% casein-0.05% Tween 20 for 60 min before being incubated with sera diluted 100-fold in PBS-Tween. Bound antibody was detected with rabbit anti-mouse IgG (Southern Biotechnology, Birmingham, AL) diluted 1:5000, followed by enhanced chemiluminescence (ECL Western blotting detection reagents, Amersham, Stockholm, Sweden). Tissue immune deposits. Pieces of the left kidney were examined as described before (Hultman et al., 1995b), using FITC-conjugated goat antimouse IgG antibodies (Southern Biotechnology) and anti-C3c antibodies (Organon-Technica, West Chester, PA). The titer was determined by serial dilution of the antibodies to 1:5120. Statistics. Differences between mercury-treated animals and controls were assessed by ANOVA and the Bonferroni post test.
RESULTS
Serum Ig Concentrations In the first set of experiments A.SW mice showed slightly varying IgE values which were not significantly different between MeHg-treated and control mice. In the second set of
experiments, A.SW mice treated with MeHg showed at 2 weeks a significant (p , 0.05) increase of IgE compared with the controls (Fig. 1). However, this increase was very marginal compared with the 5-fold increase of IgE 10 weeks after onset of MeHg treatment (6 weeks after stopping treatment) and the increase after 2 weeks HgCl2 treatment (Fig. 1). The mean IgG1 concentration in A.SW mice showed a tendency to increase 2–10 weeks after onset of MeHg treatment, but the difference was significant only after 10 weeks (Fig. 1). The IgG2a concentration was significantly increased in A.SW mice 2– 6 weeks after starting MeHg treatment, whereas IgM was significantly increased only after 3 weeks (Fig. 1). MeHg-treated A.TL mice showed no significant increase in IgE concentration compared with controls in either the first or the second experiment (Fig. 1). The IgG1 and IgG2a concentrations were significantly increased in A.TL mice 3– 6 and 2–3 weeks after onset of MeHg treatment, respectively. In SJL mice the IgE concentration fluctuated in both MeHg-treated and control mice (Fig. 1), and the IgG1, IgG2a, and IgM concentrations were not significantly increased at any time. MeHg-treated BALB/C mice showed a significant increase of IgE after 10 weeks, which was however marginal compared
206
HULTMAN AND HANSSON-GEORGIADIS
controls after 4 –10 weeks. The second experiment showed the same difference at 6 weeks (Fig. 1). The only major difference between MeHg-treated and control DBA mice was an increase of serum IgM in the former after 2 weeks (Fig. 1). Serum ANA/ANoA/AFA
FIG. 2. Reciprocal serum ANoA titer of the IgG class in individual A.SW mice treated with sc injections of 1.0 mg MeHg/kg body wt. every third day for 4 weeks. Assessment was by indirect immunofluorescence using HEp-2 cells as a substrate, serial dilution of the serum, and FITC-conjugated goat antimouse IgG antibodies as detecting reagent. Horizontal lines indicate median titer.
with the large increase after 3– 4 weeks HgCl2-treatment (Fig. 1). In the first experiment, the IgG2a concentration in MeHgtreated BALB/C mice was substantially lower than in the
ANoA first developed in A.SW mice after 2 weeks MeHg treatment, and the titer increased during the following 2 weeks (Fig. 2). The pattern of ANoA was clumpy with bright grains outlining the nucleoli, multiple nuclear dots corresponding to coiled bodies, lack of nucleoplasmic staining in interphase cells, and strong staining of the condensed chromosomes in metaphase cells (Fig. 3). This pattern is characteristic for AFA (Hultman and Pollard, 1996). Nine of 17 SJL mice showed a homogeneous ANA pattern already before onset of treatment, in accordance with previous observations (Hultman and Enestro¨m, 1988), but none showed ANoA. After 4 weeks treatment with MeHg, 6 of 9 SJL mice had developed ANoA with a titer of 247 6 86 (mean 6 SD), which increased to 328 6 129 at 10 weeks when no treatment had been given for 6 weeks (data not shown). SJL controls showed no ANoA. Sera from MeHg-treated A.SW and SJL mice with an ANoA titer of $1:640 consistently targeted in immunoblotting a nuclear protein with an apparent molecular weight of 34 kDa, corresponding to fibrillarin, and some of these sera with a high ANoA titer
FIG. 3. Pattern of serum antinucleolar antibodies from MeHg-treated A.SW mice. Assessment was by indirect immunofluorescence using HEp-2 cells and FITC-conjugated goat anti-mouse IgG antibodies. Bright grains decorate nucleoli (arrow), staining of multiple (4 – 6) nuclear dots corresponding to coiled bodies (arrowhead), and lack of nucleoplasmic staining. Magnification, 1200 3.
207
METHYL MERCURY-INDUCED AUTOIMMUNITY
FIG. 4. Immunoblotting of sera from A.SW and SJL mice. Lane 1 and 2: serum from an A.SW and an SJL mouse, respectively, treated with 1.0 mg HgCl2/kg body wt. every third day for 4 weeks. Serum ANoA titer was 1:2560 and 1:1280, respectively. The ANoA positive sera distinctly blot a protein with an apparent molecular weight corresponding to fibrillarin (34 kDa). In addition, the high-titered A.SW serum blots one or more proteins of 65–75 kDa. Lane 3: serum from an A.SW control mouse with no ANoA and no blotting of the 34-kDa protein. Lane 4: serum from an SJL mouse treated with 1.0 mg CH3HgCl/kg body wt. every third day for 4 weeks with an ANoA titer of 1:320. Weak blotting of the 34 kDa protein is shown. Lane 5: serum from an A.SW mouse treated with 1.0 mg CH3HgCl/kg body wt. every third day for 4 weeks with an ANoA titer of 1:640. Strong blotting of the 34-kDa protein and one or more proteins of 65–75 kDa is shown. In addition, blotting of a protein around 24 kDa is shown. All sera also blotted a protein of around 50 kDa. Rat liver nuclei were fractionated by SDS-PAGE and transferred to nitrocellulose. After incubation with mouse serum, detection with rabbit anti-mouse IgG antibody followed by enhanced chemiluminescence. Molecular weight markers are indicated on the right.
blotted in addition proteins in the range of 65–70 kDa (Fig. 4) as previously observed (Hultman and Pollard, 1996). Three out of 10 B10.S mice showed after 4 weeks MeHgtreatment ANoA with a low titer (1:40, 1:80, and 1:160). At 10 weeks these titers were 0, 1:80, and 1:640. Two out of 10 B10.S mice treated with MeHg developed a homogeneous ANA pattern (titer 1:160 and 1:80), but this occurred also in 2 control B10.S mice. Neither MeHg- nor NaCl-treated B10.TL mice showed ANA/ANoA. Two out of 9 MeHg-treated and 2 of 9 control A.TL mice had developed a low titer of ANA with a homogeneous pattern but no ANoA after 4 weeks treatment. Two out of 8 MeHgtreated and 1 of 9 control BALB/C mouse showed the same type of ANA but no ANoA. The DBA strain never showed ANA/ANoA. Serum Anti-ssDNA and Anti-DNP Antibodies In the first set of experiments MeHg-treated A.SW mice showed a significant increase of anti-ssDNA antibodies once, 2 weeks after starting treatment, whereas no significant increase was recorded in the second set of experiments (Fig. 5). The anti-DNP antibody level was substantially and significantly increased after 3–10 weeks in MeHg-treated A.SW mice in the first set of experiments. An significant increase in anti-DNP
antibodies was also seen in the second set of experiments, although the increase was much smaller and appeared only after 2–3 weeks treatment (Fig. 5). A.TL mice given MeHg showed a transient, significant increase of anti-ssDNA antibodies 2 weeks after onset of treatment in both the first and second set of experiments. An early, transient increase of anti-DNP antibodies was found in MeHg-treated A.TL mice after 2–3 weeks in both set of experiments, although the magnitude was much larger in the second experiment (Fig. 5). Further isotype analysis in the first experiment showed that IgG anti-DNP antibodies dominated in A.SW as well as A.TL mice with IgM contributing during the first 2–3 weeks, while the anti-DNP antibodies consisted of the IgG2a but not the IgG1 isotype (data not shown). MeHg-treated SJL and BALB/C mice showed only occasional increase in anti-ssDNA antibodies. A significant increase of anti-DNP antibodies was seen after 3– 4 weeks in BALB/C mice (Fig. 5). B10.S and B10.TL mice given MeHg showed a slight increase of anti-ssDNA and anti-DNP antibodies 6 weeks after onset of treatment (data not shown). Systemic IC Deposits At sacrifice, none of the strains showed a significant increase in renal IC-deposits (Table 1). Occasional mice in all strains after MeHg, as well as NaCl treatment, showed IgG or C3 deposits in the glomerular mesangial structures. Vessel wall deposits were, however, never observed in the kidney or spleen. DISCUSSION
This study showed that MeHg in mice of the H-2s haplotype induces ANoA targeting the 34-kDa nucleolar protein fibrillarin, whereas mice with the same background genes but another H-2 haplotype (t1) are resistant, linking the susceptibility to H-2. These observations are in agreement with findings in HgCl2-treated mice (Hultman et al., 1992, 1989a; Reuter et al., 1989). However, the ANoA response was weaker after treatment with MeHg. For example, after 4 weeks MeHg treatment the ANoA titer in A.SW mice was 370 6 237 (mean 6 SD), which is substantially lower (p , 0.01; Student’s t test) than the ANoA titer (8160 6 2163) in A.SW mice given the same amount of Hg in the form of HgCl2 (Johansson et al., 1998). The weaker ANoA response in MeHg-treated B10.S mice compared with MeHg-treated A.SW and SJL mice is in agreement with findings after HgCl2 treatment and has been attributed to a dampening effect of the B10 background genes (Hultman et al., 1992). Treatment with HgCl2 has been described as inducing a predominant Th2 reaction (Goldman et al., 1991), and blocking the action of IL-4 has been shown to drastically reduce the titer of IgG1 ANoA (Ochel et al., 1991). We found that a dose of 1.0 mg HgCl2/kg body wt. given every third day for 4 weeks
208
HULTMAN AND HANSSON-GEORGIADIS
FIG. 5. Titer of serum anti-ssDNA and anti-DNP antibodies of the IgG 1 IgM 1 IgA class after sc injections of 1.0 mg MeHg/kg body wt (F) or 0.1 ml NaCl (E) every third day for 4 weeks in the first experiment. Data from a second set of experiment with identical treatment is shown as and ƒ. Assessment by ELISA. *p , 0.05; **p , 0.01; ***p , 0.001 comparing MeHg- and NaCl-treated animals using ANOVA followed by Bonferroni post test. Significance signs to the right of the bar refers to the first experiment, signs to the left of the bars refer to the second experiment. Bars are 6 SD.
caused ANoA of the IgG1 isotype in 6 of 6 A.SW mice with a titer of 891 6 360 (mean 6 SD), while the same dose of MeHg caused IgG1 ANoA in only 3 of 9 A.SW mice and with a significantly (p , 0.05) lower titer (133 6 46). The observations in the present study thus indicate that the form of mercury to which the immune system is exposed affects the balance between Th1 and Th2 cells: inorganic mercury causes a strong Th2 response including ANoA of the IgG1 isotype, while MeHg favors a Th1 response. The titer of ANoA in HgCl2-treated A.SW mice is maximal after 4 weeks, being followed by a decrease in ANoA titer to
a plateau level around 60% of the maximal (Johansson et al., 1998). In contrast, MeHg-treated A.SW mice showed a higher ANoA titer 6 weeks after stopping MeHg treatment than after 4 weeks treatment. This late ANoA response in MeHg-treated A.SW mice and the increase in serum IgE is identical to what is seen 2–3 weeks after starting treatment with HgCl2 (Johansson et al., 1998). We interpret these late reactions as being caused by increased concentration of inorganic mercury in the body, derived from demethylation of accumulated MeHg (Norseth, 1971). This observation indicates that mercury treatment in the
209
METHYL MERCURY-INDUCED AUTOIMMUNITY
TABLE 1 Immune-Complex Deposits in Mice Treated with Methyl Mercury Chloride and Controls Kidney
Spleen
Glomeruli Strain
Number
Treatment
IgGa
A.SW A.SW A.TL A.TL SJL SJL BALB/C BALB/C DBA DBA
9 9 5 6 7 8 8 9 9 9
MeHg NaCl MeHg NaCl MeHg NaCl MeHg NaCl MeHg NaCl
13 6 28 107 6 226 176 6 267 187 6 256 897 6 1155 225 6 260 25 6 30 27 6 40 40 6 106 06 0
a
Vessels C3a
84 6 110 18 6 26 16 6 36 40 6 67 269 6 460 45 6 26 30 6 59 31 6 39 49 6 105 4 6 13
Vessels
IgG
C3
IgG
C3
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
Reciprocal titer 6 SD.
form of MeHg does not make the murine immune system resistant to reexposure with HgCl2. Indeed, recent data show that rechallenge with HgCl2 following primary exposure to the same mercury species reactivates the autoimmune response (Hultman and Nielsen, 1997). Such a reaction pattern is distinctly different from the situation in genetically mercurysusceptible rats, where even a very low dose of mercury makes the animal resistant to subsequent rechallenge with mercury (Mathieson et al., 1991). This indicates fundamental differences in immunoregulation between mice and rats treated with mercury. The lack of significantly increased IC deposits in glomeruli and systemically in vessel walls following MeHg treatment is distinctly different from the abundant IC deposits in HgCl2treated H-2s mice (Hultman et al., 1992, 1989b). However, development of ANoA but not IC deposits is seen after treating H-2s mice with silver (Hultman et al., 1994, 1995a). This dissociation between induction of ANoA and IC may be due to the weaker activation of the immune system after silver treatment compared with HgCl2 treatment (Johansson et al., 1997a). What about the activation of the immune system by MeHg compared with HgCl2? First, a major difference was the marginal tendency of serum IgE to increase in MeHg-treated A.SW mice compared with the 40 –160-fold increase of IgE after treatment with HgCl2 (Johansson et al., 1998; Pietsch et al., 1989). Since IgE is the most heavily IL-4-dependent Ig isotype (Kopf et al., 1993; Kuhn et al., 1991), the lack of increase in IgE may indicate a low IL-4 production and a low activation of Th2 cells, which is distinctly different from the strong Th2- and IL-4 response induced by HgCl2 in A.SW mice (Johansson et al., 1997b; Ochel et al., 1991). However, we cannot formally exclude the possibility that MeHg may instead have inhibited produced IL-4. A second difference was the strong immune activation in
MeHg-treated A.TL mice, including increase in Ig and signs of a polyclonal B-cell response, compared with the minimal increase of serum IgG1 and no polyclonal B-cell response after HgCl2 treatment (Johansson et al., 1998). Furthermore, the dominance of the IgG2a isotype in the A.TL strain after MeHg treatment indicates a Th1-dominated response. These observations support earlier notions that MeHg may induce cellular immune activation in strains which are resistant to HgCl2 (Stiller-Winkler et al., 1988). A third difference was the very transient increase of antissDNA antibodies in MeHg-treated A.SW mice and a moderate increase of anti-DNP antibodies, compared with the stronger and more persistent anti-ssDNA and anti-DNP antibody response after HgCl2 treatment (Johansson et al., 1998). This indicates a weaker and more restricted B-cell activation after MeHg treatment. We interpret the IgE response seen 6 weeks after stopping MeHg treatment as an effect of increasing body concentrations of inorganic mercury due to demethylation. A fourth difference was observed in BALB/C mice. While MeHg-treated BALB/C mice showed no increase in serum IgE and only a minimal increase of IgG1, HgCl2 treatment leads to an increase in both parameters (Hultman and Enestro¨m, 1987; Johansson et al., 1998). This again indicates a lower activation of the Th2-IL-4 pathway in MeHg-treated animals. In both sets of experiments, we found significantly lower levels of IgG2a in MeHg-treated as compared with control BALB/C mice after 4 – 6 weeks. We can only speculate that MeHg treatment in some way suppressed the age-related increase in IgG2a seen in the controls. In conclusion, although mercury in the form of MeHg, as well as inorganic mercury, induces ANoA/AFA, the effect of the two forms of mercury on the immune system differs quantitatively as well as qualitatively. Generally, a less Th2- and more Th1-dominated immune response was seen after MeHg treatment. By reducing T-helper cell activity, this may lead to
210
HULTMAN AND HANSSON-GEORGIADIS
the weaker general and specific immune response seen in MeHg-treated H-2s mice. Furthermore, we obtained indications that demethylation of MeHg may cause reactivation of the immune response with a response pattern characteristic for inorganic mercury.
Hultman, P., Ganowiak, K., Turely, S. J., and Pollard, K. M. (1995a). Genetic susceptibility to silver-induced antifibrillarin autoantibodies in mice. Clin. Immunol. Immunopathol. 77, 291–197. Hultman, P., Johansson, U., and Dagnaes-Hansen, F. (1995b). Murine mercuryinduced autoimmunity: the importance of T-cells. J. Autoimmun. 8, 809 – 824.
REFERENCES
Hultman, P., and Pollard, K. M. 1996. Fibrillarin autoantibodies. In Autoantibodies. (J. B. Peters and Y. Shoenfeld, Eds.), p. 253–265. Elsevier, Amsterdam.
Al-Balaghi, S., Mo¨ller, E., Mo¨ller, G., and Abed-Valugerdi, M. (1996). Mercury induces polyclonal B-cell activation, autoantibody production and renal immune complex deposits in young (NZBxNZW)F1 hybrids. Eur. J. Immunol. 26, 1519 –1526.
Hultman, P., Turley, S. J., Enestro¨m, S., Lindh, U., and Pollard, K. M. (1996). Murine genotype influences the specificity, magnitude and persistence of murine mercury-induced autoimmunity. J. Autoimm. 9, 139 –149.
Bernaudin, J. F., Druet, E., Druet, P., and Masse, R. (1981). Inhalation or ingestion of inorganic mercurials produce autoimmune disease in rats. Clin. Immunol. Immunopathol. 20, 129 –135.
Hultman, P., and Nielsen, J. B. (1997). Effect of repeated treatment with HgCl2 on development of autoimmune disease in genetically susceptible mice. In Diversification in Toxicology: Man and Environment. XXXVI European Congress of Toxicology, June 25–28, 1997, Aarhus, Denmark (Abstract).
Blakely, B. R., Sisodia, C. S., and Mukkur, T. K. (1980). The effect of merhylmercury, tetraethyl lead, and sodium arsenite on the humoral immune response in mice. Toxicol. Appl. Pharmacol. 52, 245–254.
Ilba¨ck, G. (1991). Effects of methyl mercury exposure on spleen and blood natural killer (NK) cell activity in the mouse. Toxicology 67, 117–124.
Caron, G. A., Poutala, S., and Provost, T. T. (1970). Lymphocyte transformation induced by inorganic and organic mercury. Int. Arch. Allergy Appl. Immunol. 37, 76 – 87.
Jiang, Y., and Mo¨ller, G. (1995). In vitro effects of HgCl2 on murine lymphocytes. I. Preferable activation of CD41 T cells in a responder strain. J. Immunol. 154, 3138 –3146.
Chan, E. K. L., and Pollard, K. M. 1992. Autoantibodies to ribonucleoprotein particles by immunoblotting. In Manual of Clinical Laboratory Immunology. (N. R. Rose, H. Friedman, and J. L. Fahey, Eds.), p. 755. American Society for Microbiology, Washington, DC.
Johansson, U., Hansson-Georgiadis, H., and Hultman, P. (1997a). Murine silver-induced autoimmunity: silver shares induction of antinucleolar antibodies with mercury, but causes less activation of the immune system. Int. Arch. Allergy Immunol. 113, 432– 443.
Clarkson, T. W. (1992). Principles of risk assessment. Adv. Dent. Res. 6, 22–27.
Johansson, U., Sander, B., and Hultman, P. (1997b). Effects on the murine genotype on T cell activation and cytokine production in murine mercuryinduced autoimmunity. J. Autoimm. 10, 347–355.
Goldman, M., Druet, P., and Gleichmann, E. (1991). A role for Th2 cells in systemic autoimmunity:insights from allogeneic diseases and chemicallyinduced autoimmunity. Immunol. Today. 12, 223–227. Griem, P., and Gleichmann, E. (1995). Metal-ion induced autoimmunity. Curr. Opin. Immunol. 7, 831– 838. Hansen, J. C., and Danscher, G. (1997). Organic mercury: An environmental threat to the health of the dietary-exposed societies? Rev. Env. Health. 12, 107–116. Hultman, P., Bell, L. J., Enestro¨m, S., and Pollard, K. M. (1992). Murine susceptibility to mercury. I. Autoantibody profiles and systemic immune deposits in inbred, congenic, and intra-H-2-recombinant strains. Clin. Immunol. Immunopathol. 65, 98 –109. Hultman, P., Bell, L. J., Enestro¨m, S., and Pollard, K. M. (1993). Murine susceptibility to mercury. II. Autoantibody profiles and renal immune deposits in hybrid, backcross, and H-2d congenic mice. Clin. Immunol. Immunopathol. 68, 9 –20. Hultman, P., and Enestro¨m, S. (1987). The induction of immune-complex deposits in mice by peroral and parenteral administration of mercuric chloride. Clin. Exp. Immunol. 67, 283–292. Hultman, P., and Enestro¨m, S. (1988). Mercury induced antinuclear antibodies in mice: characterization and correlation with renal immune complex deposits. Clin. Exp. Immunol. 71, 269 –274. Hultman, P., and Enestro¨m, S. (1992). Dose-response studies in murine mercury-induced autoimmunity and immune-complex disease. Toxicol. Appl. Pharmacol. 113, 199 –208.
Johansson, U., Hansson-Georgiadis, H., and Hultman, P. (1998). The genotype determines the B cell response in mercury-treated mice. Int. Arch. Allergy Immunol. 116, 295–305. Koller, L. D. (1973). Immunosuppression produced by lead, cadmium and mercury. Am. J. Vet. Res. 34, 1457–1458. Kopf, M., Le Gros, G., Bachmann, M., Lamers, M. C., Bluethmann, H., and Ko¨hler, G. (1993). Disruption of the IL-4 gene blocks Th2 cytokine response. Nature 362, 245–247. Kubicka-Muranyi, M., Kremer, J., Rottmann, N., Lu¨bben, B., Albers, R., Bloksma, N., Lu¨hrmann, R., and Gleichmann, E. (1996). Murine systemic autoimmune disease induced by mercuric chloride: T helper cells reacting to self proteins. Int. Arch. Allergy Immunol. 109, 11–20. Kuhn, R., Rajewski, K., and Mueller, W. (1991). Generation and analysis of interleukin-4 deficient mice. Science 254, 707–710. Mathieson, P., Stapleton, K., Oliveira, D., and Lockwood, C. (1991). Immunoregulation of mercuric chloride-induced autoimmunity in Brown Norway rats: A role for CD81 T cells revealed by in vivo depletion studies. Eur. J. Immunol. 21, 2105–2109. Mirtcheva, J., Pfeiffer, C., De Bruijn, J. A., Jaquesmart, F., and Gleichmann, E. (1989). Immunological alterations inducible by mercury compounds. III. H-2A acts as an immune response and H-2E as an immune “suppression” locus for HgCl2-induced antinucleolar antibodies. Eur. J. Immunol. 12, 2257–2261.
Hultman, P., Enestro¨m, S., Pollard, K. M., and Tan, E. M. (1989a). Antifibrillarin antibodies in mercury-treated mice. Clin. Exp. Immunol. 78, 470 – 477.
Nakatsuru, S., Oohashi, J., Nozaki, H., Nakada, S., and Imura, K. (1985). Effects of mercurials on lymphocyte functions in vitro. Toxicology 36, 297–305.
Hultman, P., Enestro¨m, S., and Skogh, T. (1989b). Circulating and tissue immune complexes in mercury-treated mice. J. Clin. Lab. Immunol. 29, 175–183.
Nordlind, K. (1984). Fractionation of human thymocytes and peripheral blood lymphocytes on percoll density gradients and DNA synthesis stimulating effect of mercuric chloride. Int. Arch. Allergy Appl. Immunol. 75, 16 –19.
Hultman, P., Enestro¨m, S., Turley, S. J., and Pollard, K. M. (1994). Selective induction of anti-fibrillarin autoantibodies by silver nitrate. Clin. Exp. Immunol. 96, 285–291.
Norseth, T. (1971). Biotransformation of methylmercuric salts in the mouse studied by specific determination of inorganic mercury. Acta Pharmacol. Toxicol. 29, 375–384.
METHYL MERCURY-INDUCED AUTOIMMUNITY Ochel, M., Vohr, H. W., Pfeiffer, C., and Gleichmann, E. (1991). Il-4 is required for the IgE and IgG1 increase and IgG1 autoantibody formation in mice treated with mercuric chloride. J. Immunol. 146, 3006 –3011. Ortega, H., Lopez, M., and Salvaggio, J. (1994). Effects of interferon gamma production after in vitro exposure to methyl mercury chloride. J. Invest. Med. 43, 67–74. Pauly, J. L., Caron, G. A., and Suskind, R. R. (1969). Blast transformation of lymphocytes from guinea pigs, rats and rabbits induced by mercuric chloride in vitro. J. Cell Biol. 40, 847– 856. Pelletier, L., Pasquier, R., Rossert, J., Vial, M. C., and Druet, P. (1988). Autoreactive T cells in mercury-induced autoimmune disease. Ability to induce the autoimmune disease. J. Immunol. 140, 750 –754. Pietsch, P., Vohr, H.-W., Degitz, K., and Gleichmann, E. (1989). Immunological alterations inducible by mercury compounds. II. HgCl2 and gold sodium thiomalate enhance serum IgE and IgG concentrations in susceptible mouse strains. Int. Arch. Allergy Appl. Immunol. 90, 47–53. Pollard, K., Lee, D., Casiano, C., Blu¨thner, M., Johnston, M., and Tan, E. (1997). The autoimmunity-inducing xenobiotic mercury interacts with the autoantigen fibrillarin and modifies its molecular and antigenic properties. J. Immunol. 158, 3521–3528. Reardon, C., and Lucas, D. O. (1987). Heavy metal mitogenesis: Zn21, Hg21 induce cellular cytotoxicity and interferon production in murine T lymphocytes. Immunobiology 175, 455– 469.
211
Reuter, R., Tessars, G., Vohr, H. W., Gleichmann, E., and Lu¨hrmann, R. (1989). Mercuric chloride induces autoantibodies against U3 small nuclear ribonucleoprotein in susceptible mice. Proc. Natl. Acad. Sci. USA 86, 237–245. Stiller-Winkler, R., Radaszkiewicz, T., and Gleichmann, E. (1988). Immunopathological signs in mice treated with mercury compounds - I. Identification by the popliteal lymp node assay of responder and non-responder. Int. J. Immunopharmacol. 10, 475– 484. Thuvander, A., Sundberg, J., and Oskarsson, A. (1996). Immunomodulation effects after perinatal exposure to methylmercury in mice. Toxicology 114, 163–175. van Vliet, E., Uhrberg, M., Stein, C., and Gleichmann, E. (1993). MHC control of Il-4-dependent enhancement of B cell Ia expression and Ig class switching in mice treated with mercuric chloride. Int. Arch. Allergy Appl. Immunol. 101, 392– 401. Warfvinge, K., Hansson, H., and Hultman, P. (1995). Systemic autoimmunity due to mercury vapor in genetically susceptible mice: dose-response studies. Toxicol. Appl. Pharmacol. 132, 299 –309. WHO, Environmental Health Criteria 118. Inorganic mercury. 1991, Geneva: World Health Organization. Wild, L. G., Ortega, H. G., Lopez, M., and Salvaggio, J. E. (1997). Immune system alteration in the rat after indirect exposure to methyl mercury chloride or methyl mercury sulfide. Env. Res. 74, 34 – 42.
Methyl Mercury-Induced Autoimmunity in Mice Per Hultman and Hele´n Hansson-Georgiadis Division of Molecular and Immunological Pathology, Department of Health and Environment, Linko¨ping University, S-581 85 Linko¨ping, Sweden Received April 7, 1998; accepted September 30, 1998
Methyl Mercury-Induced Autoimmunity in Mice. Hultman, P., and Hansson-Georgiadis, H. (1999). Toxicol. Appl. Pharmacol. 154, 203–211. Female SJL/N, A.SW, B10.S (H-2s), BALB/C, DBA/2 (H-2d), A.TL and B10.TL (H-2t1) mice were treated with sc injections of 1.0 mg CH3HgCl/kg body weight every third day for 4 weeks. Controls were given sterile, isotonic NaCl. CH3HgCl (MeHg) induced in SJL, A.SW and B10.S mice antinucleolar antibodies (ANoA) targeting the nucleolar 34-kDa protein fibrillarin. The susceptibility to develop ANoA in response to MeHg was linked to the mouse major histocompatibility complex (H-2), since H-2s but not H-2t1 mice sharing background (non-H-2) genes developed ANoA. However, the background genes decided the strength of the ANoA response in the susceptible H-2s mice, and the ANoA titer was in the order: A.SW > SJL > B10.S. Although MeHg as well as inorganic mercury induced ANoA, the two forms of mercury differed both quantitatively and qualitatively in their effect on the immune system. MeHg induced in H-2s mice a weaker general (polyclonal) and specific (ANoA) B-cell response than HgCl2, probably due to weaker activation of Th2 cells with lower IL-4 production, as indicated by the minimal increase in serum IgE. The A.TL strain with a susceptible genetic background, but a H-2 haplotype resistant to HgCl2, responded to MeHg with a modest polyclonal B-cell response dominated by Th1-associated Ig isotypes. H-2s mice treated with MeHg showed in contrast to HgCl2treated mice no systemic immune-complex (IC) deposits, which may be due to the weaker immune activation after MeHg treatment. The increase in serum IgE concentration and ANoA titer 2– 6 weeks after stopping treatment with MeHg is identical to reactions during the first 2–3 weeks of HgCl2 treatment. Therefore, demethylation of MeHg probably increased the concentration of inorganic mercury in the body sufficiently to reactivate the immune system. This reactivation indicated that genetically susceptible mice are not resistant to challenge with mercury, making them distinctly different from rats. © 1999 Academic Press
Inorganic mercury is, due to dental amalgam fillings, the principal source of mercury exposure in the general population. The average mean daily retention of methyl mercury (MeHg) in man has been estimated to 2.3 mg Hg/day, 99% of which is due to fish consumption (Clarkson, 1992). MeHg retention comprises 86% of the total mercury retention derived from nondental sources (WHO, 1991). However, the effect of MeHg
on the immune system has only been sparsely investigated. Nakatsuru et al. found that MeHg in vitro inhibits T- and B-cell activation (Nakatsuru et al., 1985), and MeHg has been shown to suppress humoral immunity in vivo (Blakely et al., 1980; Koller, 1973). Natural killer cell activity is severely reduced in adult BALB/C mice exposed to MeHg (Ilba¨ck, 1991). Indirect (perinatal) exposure to MeHg in mice (Thuvander et al., 1996) and rats (Wild et al., 1997) leads to an enhanced response to B-cell mitogens. A slightly increased response to T-cell mitogen is found in adult mice (Ilba¨ck, 1991) and rats (Ortega et al., 1994) exposed to MeHg. Genetically susceptible Brown Norway rats given MeHg orally or by the respiratory tract develop a systemic autoimmune condition similar to what is seen after mercuric chloride (HgCl2) treatment (Bernaudin et al., 1981). The interaction of mercury with the immune system is complex. Inorganic mercury induces, within a narrow dose range, proliferation of lymphocytes derived from humans (Caron et al., 1970) and a number of animals species (Pauly et al., 1969). The proliferating cells have been identified as T-cells in humans (Nordlind, 1984), rats (Pelletier et al., 1988), and mice (Jiang and Mo¨ller, 1995; Reardon and Lucas, 1987). Similarly, certain inbred mouse strains develop a strong lymphoproliferation, including both CD41 and CD81 T-cells, while other strains respond with only a slight proliferation of CD81 T-cells (Jiang and Mo¨ller, 1995). The genetic susceptibility to inorganic mercury is even more pronounced in vivo. Parenteral and oral treatment with HgCl2 and exposure to elemental mercury in the form of vapor in mouse strains of the H-2s haplotype induces systemic autoimmunity including antinucleolar antibodies (ANoA) targeting fibrillarin (AFA) and immune-complex (IC)-deposits (Hultman et al., 1992, 1989a, 1996; Hultman and Enestro¨m, 1992; Reuter et al., 1989; Warfvinge et al., 1995). Treatment of other inbred mouse strains, such as BALB/C (H-2d), with inorganic mercury leads to a T-celldependent immune response with systemic IC deposits but no ANoA/AFA (Hultman and Enestro¨m, 1987). Inorganic mercury induces a strong and broad (polyclonal) activation of the immune system in mice of the H-2s haplotype (Al-Balaghi et al., 1996; Hultman and Enestro¨m, 1987; Hultman et al., 1996; Johansson et al., 1997b; Pietsch et al., 1989; van Vliet et al., 1993), but the antibodies in this polyclonal response differ with respect to duration and isotype pattern from the simultaneously appearing AFA (Johansson et al.,
203
0041-008X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
204
HULTMAN AND HANSSON-GEORGIADIS
1998). Treatment with silver in H-2s mice causes only a slight activation of the immune system and no polyclonal B-cell response. However, the maximum ANoA titer does not differ significantly between mice given AgNO3 and HgCl2 (Johansson et al., 1997a, 1998). This clearly illustrates that a strong and polyclonal activation of the immune system, as seen after treatment with inorganic mercury, is not necessary for induction of ANoA/AFA. Instead, recent studies show that mercury causes alterations in the structure and processing of fibrillarin (Griem and Gleichmann, 1995; Pollard et al., 1997), which gives rise to fibrillarin fragments with cryptic epitopes. T-cell clones with specificity for modified fibrillarin antigens have recently been identified (Kubicka-Muranyi et al., 1996). This strongly indicates that AFA arise by antigen-specific mechanism and not as part of a polyclonal activation. The ability of organic mercury, especially MeHg, to induce systemic autoimmunity in genetically susceptible mouse strains (Hultman et al., 1992; Mirtcheva et al., 1989) has so far not been studied. Using the popliteal lymph node test, StillerWinkler et al. (1988) found that a strain resistant to the lymphoproliferative effect of inorganic mercury could be susceptible if MeHg was used. We have therefore examined to what extent organic mercury, in the form of MeHg, activates the immune system and causes systemic autoimmunity in a number of strains genetically susceptible or resistant to inorganic mercury (Hultman et al., 1992, 1993, 1996; Mirtcheva et al., 1989). The same dose of MeHg and HgCl2 was used in the present and these previous studies, respectively. The MeHg exposure in our mice was ca. 350 mg Hg/kg/day, which is much higher than the exposure recorded in highly exposed populations such as Polar Eskimos (ca. 5 mg Hg/kg/day) (Hansen and Danscher, 1997). MATERIALS AND METHODS Mice. Female SJL/N and A.SW (H-2s), and BALB/C and DBA/2 (H-2d) mice were obtained from Bommice Breeding and Research Centre (Ry, Denmark). Female B10.S (H-2s) and A.TL (H-2t1) mice were from Harlan Ltd. (Oxon, England). B10.TL mice, originally obtained from Department of Immunogenetics, University of Tu¨bingen, Germany, were maintained by brothersister mating at our animal facility department, and female progeny were used in this study. All mice were 9 –13 weeks old at onset of the experiments. The mice were housed under 12-h dark-12-h light cycles in a high-barrier unit, kept in steel-wire cages with beechwood bedding, and given sterilized pellets (Type R 36, Lactamin, Vadstena, Sweden) and tap water ad libitum. Treatment. Treatment consisted of giving 1.0 mg CH3HgCl/kg body wt diluted in sterile 0.84% NaCl as a 0.1 ml sc injection on the dorsum every third day for 27 days. Controls received an equal volume of sterile 0.84% NaCl. In the first set of experiments groups of 8 –11 mice were injected with CH3HgCl or sterile, isotonic NaCl as above for 4 weeks, which were followed with blood sampling 2, 3, 4, 6, and 10 weeks after onset of injections. One B10.TL mouse treated with CH3HgCl (MeHg) died after blood sampling 3 weeks after onset of injections. One control A.SW mouse died at blood sampling after 6 weeks. Among the SJL mice, 2 given MeHg died after blood sampling at 4 weeks, and 2 controls died at 4 and 6 weeks, respectively. One MeHg-treated and two control A.TL mice died after blood sampling at 3 weeks. Among the BALB/C, DBA/2, B10.S, and B10.TL mice, only a single MeHg-treated BALB/C mouse
died at 2 weeks. In the second set of experiments, groups of 7 A.SW, A.TL, and BALB/C mice were treated with CH3HgCl or NaCl as described above for 4 weeks, and blood samples were obtained before treatment as well as after 2, 3, 4, and 6 weeks. Serum Ig concentrations assessed by ELISA. Analysis of serum Ig concentrations was performed as previously described (Johansson et al., 1997a). Briefly, microtiter plates (Nunc, Copenhagen, Denmark) were coated with rat anti-mouse IgG1 and rat anti-mouse IgM MAb (LO-IMEX, Brussels, Belgium), respectively. Following blocking, the wells were incubated with diluted serum. Bound IgG1 and IgM were detected using diluted horseradish peroxidase (HRP)-conjugated rat anti-mouse and rat anti-mouse IgM MAb (LOIMEX), respectively. The plates were read at OD450 and the background values in wells coated with PBS instead of serum were subtracted. Actual concentration was obtained from a standard curve using mouse myeloma proteins of the IgG1 and IgM isotype (LO-IMEX), respectively. For serum IgG2a analysis, microtiter wells (Nunc) were coated with affinity-purified goat anti-mouse IgG2a isotype (Southern Biotechnology). The wells were incubated with diluted sera, and bound IgG2a was detected with alkaline phosphatase (ALP) conjugated goat anti-mouse IgG (Southern Biotechnology). The IgG2a concentration in the samples was obtained from a standard curve using mouse myeloma proteins of the IgG2a isotype (LO-IMEX). Serum IgE was determined as described before (Warfvinge et al., 1995). Briefly, microtiter plates were coated with rat anti-mouse IgE (Southern Biotechnology), followed by blocking and incubation with diluted serum. Bound IgE was detected by HRP-conjugated goat anti-mouse IgE (Nordic Immunological Lab, Tilburg, Netherlands), and the IgE concentration in the samples was derived from a standard curve using mouse myeloma protein of the IgE isotype (Sigma). Serum anti-ssDNA antibodies assessed by ELISA. The method used has been described before (Johansson et al., 1997a). Microtiter plates were coated overnight with ssDNA, washed with PBS-Tween 20, blocked with BSATween 20-PBS, and repeatedly washed first with PBS-Tween and then with PBS. Sera diluted in 1% BSA-PBS were incubated in the wells, and the plates washed six times with BSA-Tween-PBS and incubated with ALP-conjugated rabbit anti-mouse Ig (reacting with IgG, IgM and IgA) (Sigma) diluted in BSA-Tween-PBS. The plates were repeatedly washed, substrate was added, reaction was stopped with 3 M NaOH, optical density was read at OD405, and the background was subtracted. A pool of sera from MRL-mice was used as positive control. Using a monoclonal antibody (clone HB2) reacting with dsDNA (SeraLab), we detected no contamination with dsDNA in the coating (data not shown). For detection of anti-ssDNA ab of the IgG and IgM type, ALP-conjugated goat anti-mouse IgG and IgM antibodies (Caltag Laboratories, San Francisco, CA) were used. Serum anti-DNP antibodies assessed by ELISA. The method used has been described before (Johansson et al., 1997a). Microtiter plates (Nunc) were coated over night with human serum albumin conjugated with 30 – 40 mol DNP per mol albumin (Sigma). Following repeated washes with BSA-PBS, the wells were incubated with diluted sera and washed, and ALP-conjugated rabbit anti-mouse Ig (reacting with IgG, IgM, and IgA) (Sigma) was added. After repeated washes with BSA-PBS, substrate was added, and the reaction was stopped with 3 M NaOH. The optical density was read at OD405, and background values were subtracted. For detection of anti-DNP abs of the IgG and IgM type, ALP-conjugated goat anti-mouse IgG and IgM antibodies (Caltag Laboratories) were used. Mouse anti-DNP antibodies (LO-IMEX) were used as positive controls. For detection of anti-DNP abs of the IgG1 and IgG2a isotype, we used ALP-conjugated rat anti-mouse IgG1 and IgG2a antibodies (Pharmingen Inc., CA). Serum antinuclear antibodies assessed by indirect immunofluorescence. The presence, pattern, and titer of serum antinuclear antibodies of the IgG class was determined by indirect immunofluorescence using HEp-2 cells as a substrate (Hultman et al., 1989b). Serum antinuclear antibodies assessed by immunoblotting. The specificity of the antinuclear antibodies in the serum was assessed by immunoblotting as described before with minor modifications (Warfvinge et al., 1995). Briefly,
METHYL MERCURY-INDUCED AUTOIMMUNITY
205
FIG. 1. Serum Ig concentration in mice treated with sc injection of 1.0 mg MeHg/kg body wt. (F) or 0.1 ml NaCl (E) every third day for 4 weeks. The results from identically MeHg-treated mice and controls in a second set of experiments are also shown: and ƒ, respectively. The IgE concentration after sc injections of an equimolar amount of HgCl2 to A.SW and BALB/C mice is included for comparison (r). *p , 0.05; **p , 0.01; ***p , 0.001 comparing MeHgand NaCl-treated animals using ANOVA followed by Bonferroni post test. Bars are 6 SD.
rat liver nuclei were isolated (Chan and Pollard, 1992), aliquots of the boiled nuclei were SDS-PAGE separated using a gradient gel, and electrophoretic transfer to nitrocellulose membranes was performed for 1.5 h at 0.8 mA/cm2 under water cooling (Semiphor Transfer Unit, Hoefer Scientific Instruments, San Francisco, CA). Nitrocellulose strips were blocked in a solution of PBS-5% casein-0.05% Tween 20 for 60 min before being incubated with sera diluted 100-fold in PBS-Tween. Bound antibody was detected with rabbit anti-mouse IgG (Southern Biotechnology, Birmingham, AL) diluted 1:5000, followed by enhanced chemiluminescence (ECL Western blotting detection reagents, Amersham, Stockholm, Sweden). Tissue immune deposits. Pieces of the left kidney were examined as described before (Hultman et al., 1995b), using FITC-conjugated goat antimouse IgG antibodies (Southern Biotechnology) and anti-C3c antibodies (Organon-Technica, West Chester, PA). The titer was determined by serial dilution of the antibodies to 1:5120. Statistics. Differences between mercury-treated animals and controls were assessed by ANOVA and the Bonferroni post test.
RESULTS
Serum Ig Concentrations In the first set of experiments A.SW mice showed slightly varying IgE values which were not significantly different between MeHg-treated and control mice. In the second set of
experiments, A.SW mice treated with MeHg showed at 2 weeks a significant (p , 0.05) increase of IgE compared with the controls (Fig. 1). However, this increase was very marginal compared with the 5-fold increase of IgE 10 weeks after onset of MeHg treatment (6 weeks after stopping treatment) and the increase after 2 weeks HgCl2 treatment (Fig. 1). The mean IgG1 concentration in A.SW mice showed a tendency to increase 2–10 weeks after onset of MeHg treatment, but the difference was significant only after 10 weeks (Fig. 1). The IgG2a concentration was significantly increased in A.SW mice 2– 6 weeks after starting MeHg treatment, whereas IgM was significantly increased only after 3 weeks (Fig. 1). MeHg-treated A.TL mice showed no significant increase in IgE concentration compared with controls in either the first or the second experiment (Fig. 1). The IgG1 and IgG2a concentrations were significantly increased in A.TL mice 3– 6 and 2–3 weeks after onset of MeHg treatment, respectively. In SJL mice the IgE concentration fluctuated in both MeHg-treated and control mice (Fig. 1), and the IgG1, IgG2a, and IgM concentrations were not significantly increased at any time. MeHg-treated BALB/C mice showed a significant increase of IgE after 10 weeks, which was however marginal compared
206
HULTMAN AND HANSSON-GEORGIADIS
controls after 4 –10 weeks. The second experiment showed the same difference at 6 weeks (Fig. 1). The only major difference between MeHg-treated and control DBA mice was an increase of serum IgM in the former after 2 weeks (Fig. 1). Serum ANA/ANoA/AFA
FIG. 2. Reciprocal serum ANoA titer of the IgG class in individual A.SW mice treated with sc injections of 1.0 mg MeHg/kg body wt. every third day for 4 weeks. Assessment was by indirect immunofluorescence using HEp-2 cells as a substrate, serial dilution of the serum, and FITC-conjugated goat antimouse IgG antibodies as detecting reagent. Horizontal lines indicate median titer.
with the large increase after 3– 4 weeks HgCl2-treatment (Fig. 1). In the first experiment, the IgG2a concentration in MeHgtreated BALB/C mice was substantially lower than in the
ANoA first developed in A.SW mice after 2 weeks MeHg treatment, and the titer increased during the following 2 weeks (Fig. 2). The pattern of ANoA was clumpy with bright grains outlining the nucleoli, multiple nuclear dots corresponding to coiled bodies, lack of nucleoplasmic staining in interphase cells, and strong staining of the condensed chromosomes in metaphase cells (Fig. 3). This pattern is characteristic for AFA (Hultman and Pollard, 1996). Nine of 17 SJL mice showed a homogeneous ANA pattern already before onset of treatment, in accordance with previous observations (Hultman and Enestro¨m, 1988), but none showed ANoA. After 4 weeks treatment with MeHg, 6 of 9 SJL mice had developed ANoA with a titer of 247 6 86 (mean 6 SD), which increased to 328 6 129 at 10 weeks when no treatment had been given for 6 weeks (data not shown). SJL controls showed no ANoA. Sera from MeHg-treated A.SW and SJL mice with an ANoA titer of $1:640 consistently targeted in immunoblotting a nuclear protein with an apparent molecular weight of 34 kDa, corresponding to fibrillarin, and some of these sera with a high ANoA titer
FIG. 3. Pattern of serum antinucleolar antibodies from MeHg-treated A.SW mice. Assessment was by indirect immunofluorescence using HEp-2 cells and FITC-conjugated goat anti-mouse IgG antibodies. Bright grains decorate nucleoli (arrow), staining of multiple (4 – 6) nuclear dots corresponding to coiled bodies (arrowhead), and lack of nucleoplasmic staining. Magnification, 1200 3.
207
METHYL MERCURY-INDUCED AUTOIMMUNITY
FIG. 4. Immunoblotting of sera from A.SW and SJL mice. Lane 1 and 2: serum from an A.SW and an SJL mouse, respectively, treated with 1.0 mg HgCl2/kg body wt. every third day for 4 weeks. Serum ANoA titer was 1:2560 and 1:1280, respectively. The ANoA positive sera distinctly blot a protein with an apparent molecular weight corresponding to fibrillarin (34 kDa). In addition, the high-titered A.SW serum blots one or more proteins of 65–75 kDa. Lane 3: serum from an A.SW control mouse with no ANoA and no blotting of the 34-kDa protein. Lane 4: serum from an SJL mouse treated with 1.0 mg CH3HgCl/kg body wt. every third day for 4 weeks with an ANoA titer of 1:320. Weak blotting of the 34 kDa protein is shown. Lane 5: serum from an A.SW mouse treated with 1.0 mg CH3HgCl/kg body wt. every third day for 4 weeks with an ANoA titer of 1:640. Strong blotting of the 34-kDa protein and one or more proteins of 65–75 kDa is shown. In addition, blotting of a protein around 24 kDa is shown. All sera also blotted a protein of around 50 kDa. Rat liver nuclei were fractionated by SDS-PAGE and transferred to nitrocellulose. After incubation with mouse serum, detection with rabbit anti-mouse IgG antibody followed by enhanced chemiluminescence. Molecular weight markers are indicated on the right.
blotted in addition proteins in the range of 65–70 kDa (Fig. 4) as previously observed (Hultman and Pollard, 1996). Three out of 10 B10.S mice showed after 4 weeks MeHgtreatment ANoA with a low titer (1:40, 1:80, and 1:160). At 10 weeks these titers were 0, 1:80, and 1:640. Two out of 10 B10.S mice treated with MeHg developed a homogeneous ANA pattern (titer 1:160 and 1:80), but this occurred also in 2 control B10.S mice. Neither MeHg- nor NaCl-treated B10.TL mice showed ANA/ANoA. Two out of 9 MeHg-treated and 2 of 9 control A.TL mice had developed a low titer of ANA with a homogeneous pattern but no ANoA after 4 weeks treatment. Two out of 8 MeHgtreated and 1 of 9 control BALB/C mouse showed the same type of ANA but no ANoA. The DBA strain never showed ANA/ANoA. Serum Anti-ssDNA and Anti-DNP Antibodies In the first set of experiments MeHg-treated A.SW mice showed a significant increase of anti-ssDNA antibodies once, 2 weeks after starting treatment, whereas no significant increase was recorded in the second set of experiments (Fig. 5). The anti-DNP antibody level was substantially and significantly increased after 3–10 weeks in MeHg-treated A.SW mice in the first set of experiments. An significant increase in anti-DNP
antibodies was also seen in the second set of experiments, although the increase was much smaller and appeared only after 2–3 weeks treatment (Fig. 5). A.TL mice given MeHg showed a transient, significant increase of anti-ssDNA antibodies 2 weeks after onset of treatment in both the first and second set of experiments. An early, transient increase of anti-DNP antibodies was found in MeHg-treated A.TL mice after 2–3 weeks in both set of experiments, although the magnitude was much larger in the second experiment (Fig. 5). Further isotype analysis in the first experiment showed that IgG anti-DNP antibodies dominated in A.SW as well as A.TL mice with IgM contributing during the first 2–3 weeks, while the anti-DNP antibodies consisted of the IgG2a but not the IgG1 isotype (data not shown). MeHg-treated SJL and BALB/C mice showed only occasional increase in anti-ssDNA antibodies. A significant increase of anti-DNP antibodies was seen after 3– 4 weeks in BALB/C mice (Fig. 5). B10.S and B10.TL mice given MeHg showed a slight increase of anti-ssDNA and anti-DNP antibodies 6 weeks after onset of treatment (data not shown). Systemic IC Deposits At sacrifice, none of the strains showed a significant increase in renal IC-deposits (Table 1). Occasional mice in all strains after MeHg, as well as NaCl treatment, showed IgG or C3 deposits in the glomerular mesangial structures. Vessel wall deposits were, however, never observed in the kidney or spleen. DISCUSSION
This study showed that MeHg in mice of the H-2s haplotype induces ANoA targeting the 34-kDa nucleolar protein fibrillarin, whereas mice with the same background genes but another H-2 haplotype (t1) are resistant, linking the susceptibility to H-2. These observations are in agreement with findings in HgCl2-treated mice (Hultman et al., 1992, 1989a; Reuter et al., 1989). However, the ANoA response was weaker after treatment with MeHg. For example, after 4 weeks MeHg treatment the ANoA titer in A.SW mice was 370 6 237 (mean 6 SD), which is substantially lower (p , 0.01; Student’s t test) than the ANoA titer (8160 6 2163) in A.SW mice given the same amount of Hg in the form of HgCl2 (Johansson et al., 1998). The weaker ANoA response in MeHg-treated B10.S mice compared with MeHg-treated A.SW and SJL mice is in agreement with findings after HgCl2 treatment and has been attributed to a dampening effect of the B10 background genes (Hultman et al., 1992). Treatment with HgCl2 has been described as inducing a predominant Th2 reaction (Goldman et al., 1991), and blocking the action of IL-4 has been shown to drastically reduce the titer of IgG1 ANoA (Ochel et al., 1991). We found that a dose of 1.0 mg HgCl2/kg body wt. given every third day for 4 weeks
208
HULTMAN AND HANSSON-GEORGIADIS
FIG. 5. Titer of serum anti-ssDNA and anti-DNP antibodies of the IgG 1 IgM 1 IgA class after sc injections of 1.0 mg MeHg/kg body wt (F) or 0.1 ml NaCl (E) every third day for 4 weeks in the first experiment. Data from a second set of experiment with identical treatment is shown as and ƒ. Assessment by ELISA. *p , 0.05; **p , 0.01; ***p , 0.001 comparing MeHg- and NaCl-treated animals using ANOVA followed by Bonferroni post test. Significance signs to the right of the bar refers to the first experiment, signs to the left of the bars refer to the second experiment. Bars are 6 SD.
caused ANoA of the IgG1 isotype in 6 of 6 A.SW mice with a titer of 891 6 360 (mean 6 SD), while the same dose of MeHg caused IgG1 ANoA in only 3 of 9 A.SW mice and with a significantly (p , 0.05) lower titer (133 6 46). The observations in the present study thus indicate that the form of mercury to which the immune system is exposed affects the balance between Th1 and Th2 cells: inorganic mercury causes a strong Th2 response including ANoA of the IgG1 isotype, while MeHg favors a Th1 response. The titer of ANoA in HgCl2-treated A.SW mice is maximal after 4 weeks, being followed by a decrease in ANoA titer to
a plateau level around 60% of the maximal (Johansson et al., 1998). In contrast, MeHg-treated A.SW mice showed a higher ANoA titer 6 weeks after stopping MeHg treatment than after 4 weeks treatment. This late ANoA response in MeHg-treated A.SW mice and the increase in serum IgE is identical to what is seen 2–3 weeks after starting treatment with HgCl2 (Johansson et al., 1998). We interpret these late reactions as being caused by increased concentration of inorganic mercury in the body, derived from demethylation of accumulated MeHg (Norseth, 1971). This observation indicates that mercury treatment in the
209
METHYL MERCURY-INDUCED AUTOIMMUNITY
TABLE 1 Immune-Complex Deposits in Mice Treated with Methyl Mercury Chloride and Controls Kidney
Spleen
Glomeruli Strain
Number
Treatment
IgGa
A.SW A.SW A.TL A.TL SJL SJL BALB/C BALB/C DBA DBA
9 9 5 6 7 8 8 9 9 9
MeHg NaCl MeHg NaCl MeHg NaCl MeHg NaCl MeHg NaCl
13 6 28 107 6 226 176 6 267 187 6 256 897 6 1155 225 6 260 25 6 30 27 6 40 40 6 106 06 0
a
Vessels C3a
84 6 110 18 6 26 16 6 36 40 6 67 269 6 460 45 6 26 30 6 59 31 6 39 49 6 105 4 6 13
Vessels
IgG
C3
IgG
C3
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
Reciprocal titer 6 SD.
form of MeHg does not make the murine immune system resistant to reexposure with HgCl2. Indeed, recent data show that rechallenge with HgCl2 following primary exposure to the same mercury species reactivates the autoimmune response (Hultman and Nielsen, 1997). Such a reaction pattern is distinctly different from the situation in genetically mercurysusceptible rats, where even a very low dose of mercury makes the animal resistant to subsequent rechallenge with mercury (Mathieson et al., 1991). This indicates fundamental differences in immunoregulation between mice and rats treated with mercury. The lack of significantly increased IC deposits in glomeruli and systemically in vessel walls following MeHg treatment is distinctly different from the abundant IC deposits in HgCl2treated H-2s mice (Hultman et al., 1992, 1989b). However, development of ANoA but not IC deposits is seen after treating H-2s mice with silver (Hultman et al., 1994, 1995a). This dissociation between induction of ANoA and IC may be due to the weaker activation of the immune system after silver treatment compared with HgCl2 treatment (Johansson et al., 1997a). What about the activation of the immune system by MeHg compared with HgCl2? First, a major difference was the marginal tendency of serum IgE to increase in MeHg-treated A.SW mice compared with the 40 –160-fold increase of IgE after treatment with HgCl2 (Johansson et al., 1998; Pietsch et al., 1989). Since IgE is the most heavily IL-4-dependent Ig isotype (Kopf et al., 1993; Kuhn et al., 1991), the lack of increase in IgE may indicate a low IL-4 production and a low activation of Th2 cells, which is distinctly different from the strong Th2- and IL-4 response induced by HgCl2 in A.SW mice (Johansson et al., 1997b; Ochel et al., 1991). However, we cannot formally exclude the possibility that MeHg may instead have inhibited produced IL-4. A second difference was the strong immune activation in
MeHg-treated A.TL mice, including increase in Ig and signs of a polyclonal B-cell response, compared with the minimal increase of serum IgG1 and no polyclonal B-cell response after HgCl2 treatment (Johansson et al., 1998). Furthermore, the dominance of the IgG2a isotype in the A.TL strain after MeHg treatment indicates a Th1-dominated response. These observations support earlier notions that MeHg may induce cellular immune activation in strains which are resistant to HgCl2 (Stiller-Winkler et al., 1988). A third difference was the very transient increase of antissDNA antibodies in MeHg-treated A.SW mice and a moderate increase of anti-DNP antibodies, compared with the stronger and more persistent anti-ssDNA and anti-DNP antibody response after HgCl2 treatment (Johansson et al., 1998). This indicates a weaker and more restricted B-cell activation after MeHg treatment. We interpret the IgE response seen 6 weeks after stopping MeHg treatment as an effect of increasing body concentrations of inorganic mercury due to demethylation. A fourth difference was observed in BALB/C mice. While MeHg-treated BALB/C mice showed no increase in serum IgE and only a minimal increase of IgG1, HgCl2 treatment leads to an increase in both parameters (Hultman and Enestro¨m, 1987; Johansson et al., 1998). This again indicates a lower activation of the Th2-IL-4 pathway in MeHg-treated animals. In both sets of experiments, we found significantly lower levels of IgG2a in MeHg-treated as compared with control BALB/C mice after 4 – 6 weeks. We can only speculate that MeHg treatment in some way suppressed the age-related increase in IgG2a seen in the controls. In conclusion, although mercury in the form of MeHg, as well as inorganic mercury, induces ANoA/AFA, the effect of the two forms of mercury on the immune system differs quantitatively as well as qualitatively. Generally, a less Th2- and more Th1-dominated immune response was seen after MeHg treatment. By reducing T-helper cell activity, this may lead to
210
HULTMAN AND HANSSON-GEORGIADIS
the weaker general and specific immune response seen in MeHg-treated H-2s mice. Furthermore, we obtained indications that demethylation of MeHg may cause reactivation of the immune response with a response pattern characteristic for inorganic mercury.
Hultman, P., Ganowiak, K., Turely, S. J., and Pollard, K. M. (1995a). Genetic susceptibility to silver-induced antifibrillarin autoantibodies in mice. Clin. Immunol. Immunopathol. 77, 291–197. Hultman, P., Johansson, U., and Dagnaes-Hansen, F. (1995b). Murine mercuryinduced autoimmunity: the importance of T-cells. J. Autoimmun. 8, 809 – 824.
REFERENCES
Hultman, P., and Pollard, K. M. 1996. Fibrillarin autoantibodies. In Autoantibodies. (J. B. Peters and Y. Shoenfeld, Eds.), p. 253–265. Elsevier, Amsterdam.
Al-Balaghi, S., Mo¨ller, E., Mo¨ller, G., and Abed-Valugerdi, M. (1996). Mercury induces polyclonal B-cell activation, autoantibody production and renal immune complex deposits in young (NZBxNZW)F1 hybrids. Eur. J. Immunol. 26, 1519 –1526.
Hultman, P., Turley, S. J., Enestro¨m, S., Lindh, U., and Pollard, K. M. (1996). Murine genotype influences the specificity, magnitude and persistence of murine mercury-induced autoimmunity. J. Autoimm. 9, 139 –149.
Bernaudin, J. F., Druet, E., Druet, P., and Masse, R. (1981). Inhalation or ingestion of inorganic mercurials produce autoimmune disease in rats. Clin. Immunol. Immunopathol. 20, 129 –135.
Hultman, P., and Nielsen, J. B. (1997). Effect of repeated treatment with HgCl2 on development of autoimmune disease in genetically susceptible mice. In Diversification in Toxicology: Man and Environment. XXXVI European Congress of Toxicology, June 25–28, 1997, Aarhus, Denmark (Abstract).
Blakely, B. R., Sisodia, C. S., and Mukkur, T. K. (1980). The effect of merhylmercury, tetraethyl lead, and sodium arsenite on the humoral immune response in mice. Toxicol. Appl. Pharmacol. 52, 245–254.
Ilba¨ck, G. (1991). Effects of methyl mercury exposure on spleen and blood natural killer (NK) cell activity in the mouse. Toxicology 67, 117–124.
Caron, G. A., Poutala, S., and Provost, T. T. (1970). Lymphocyte transformation induced by inorganic and organic mercury. Int. Arch. Allergy Appl. Immunol. 37, 76 – 87.
Jiang, Y., and Mo¨ller, G. (1995). In vitro effects of HgCl2 on murine lymphocytes. I. Preferable activation of CD41 T cells in a responder strain. J. Immunol. 154, 3138 –3146.
Chan, E. K. L., and Pollard, K. M. 1992. Autoantibodies to ribonucleoprotein particles by immunoblotting. In Manual of Clinical Laboratory Immunology. (N. R. Rose, H. Friedman, and J. L. Fahey, Eds.), p. 755. American Society for Microbiology, Washington, DC.
Johansson, U., Hansson-Georgiadis, H., and Hultman, P. (1997a). Murine silver-induced autoimmunity: silver shares induction of antinucleolar antibodies with mercury, but causes less activation of the immune system. Int. Arch. Allergy Immunol. 113, 432– 443.
Clarkson, T. W. (1992). Principles of risk assessment. Adv. Dent. Res. 6, 22–27.
Johansson, U., Sander, B., and Hultman, P. (1997b). Effects on the murine genotype on T cell activation and cytokine production in murine mercuryinduced autoimmunity. J. Autoimm. 10, 347–355.
Goldman, M., Druet, P., and Gleichmann, E. (1991). A role for Th2 cells in systemic autoimmunity:insights from allogeneic diseases and chemicallyinduced autoimmunity. Immunol. Today. 12, 223–227. Griem, P., and Gleichmann, E. (1995). Metal-ion induced autoimmunity. Curr. Opin. Immunol. 7, 831– 838. Hansen, J. C., and Danscher, G. (1997). Organic mercury: An environmental threat to the health of the dietary-exposed societies? Rev. Env. Health. 12, 107–116. Hultman, P., Bell, L. J., Enestro¨m, S., and Pollard, K. M. (1992). Murine susceptibility to mercury. I. Autoantibody profiles and systemic immune deposits in inbred, congenic, and intra-H-2-recombinant strains. Clin. Immunol. Immunopathol. 65, 98 –109. Hultman, P., Bell, L. J., Enestro¨m, S., and Pollard, K. M. (1993). Murine susceptibility to mercury. II. Autoantibody profiles and renal immune deposits in hybrid, backcross, and H-2d congenic mice. Clin. Immunol. Immunopathol. 68, 9 –20. Hultman, P., and Enestro¨m, S. (1987). The induction of immune-complex deposits in mice by peroral and parenteral administration of mercuric chloride. Clin. Exp. Immunol. 67, 283–292. Hultman, P., and Enestro¨m, S. (1988). Mercury induced antinuclear antibodies in mice: characterization and correlation with renal immune complex deposits. Clin. Exp. Immunol. 71, 269 –274. Hultman, P., and Enestro¨m, S. (1992). Dose-response studies in murine mercury-induced autoimmunity and immune-complex disease. Toxicol. Appl. Pharmacol. 113, 199 –208.
Johansson, U., Hansson-Georgiadis, H., and Hultman, P. (1998). The genotype determines the B cell response in mercury-treated mice. Int. Arch. Allergy Immunol. 116, 295–305. Koller, L. D. (1973). Immunosuppression produced by lead, cadmium and mercury. Am. J. Vet. Res. 34, 1457–1458. Kopf, M., Le Gros, G., Bachmann, M., Lamers, M. C., Bluethmann, H., and Ko¨hler, G. (1993). Disruption of the IL-4 gene blocks Th2 cytokine response. Nature 362, 245–247. Kubicka-Muranyi, M., Kremer, J., Rottmann, N., Lu¨bben, B., Albers, R., Bloksma, N., Lu¨hrmann, R., and Gleichmann, E. (1996). Murine systemic autoimmune disease induced by mercuric chloride: T helper cells reacting to self proteins. Int. Arch. Allergy Immunol. 109, 11–20. Kuhn, R., Rajewski, K., and Mueller, W. (1991). Generation and analysis of interleukin-4 deficient mice. Science 254, 707–710. Mathieson, P., Stapleton, K., Oliveira, D., and Lockwood, C. (1991). Immunoregulation of mercuric chloride-induced autoimmunity in Brown Norway rats: A role for CD81 T cells revealed by in vivo depletion studies. Eur. J. Immunol. 21, 2105–2109. Mirtcheva, J., Pfeiffer, C., De Bruijn, J. A., Jaquesmart, F., and Gleichmann, E. (1989). Immunological alterations inducible by mercury compounds. III. H-2A acts as an immune response and H-2E as an immune “suppression” locus for HgCl2-induced antinucleolar antibodies. Eur. J. Immunol. 12, 2257–2261.
Hultman, P., Enestro¨m, S., Pollard, K. M., and Tan, E. M. (1989a). Antifibrillarin antibodies in mercury-treated mice. Clin. Exp. Immunol. 78, 470 – 477.
Nakatsuru, S., Oohashi, J., Nozaki, H., Nakada, S., and Imura, K. (1985). Effects of mercurials on lymphocyte functions in vitro. Toxicology 36, 297–305.
Hultman, P., Enestro¨m, S., and Skogh, T. (1989b). Circulating and tissue immune complexes in mercury-treated mice. J. Clin. Lab. Immunol. 29, 175–183.
Nordlind, K. (1984). Fractionation of human thymocytes and peripheral blood lymphocytes on percoll density gradients and DNA synthesis stimulating effect of mercuric chloride. Int. Arch. Allergy Appl. Immunol. 75, 16 –19.
Hultman, P., Enestro¨m, S., Turley, S. J., and Pollard, K. M. (1994). Selective induction of anti-fibrillarin autoantibodies by silver nitrate. Clin. Exp. Immunol. 96, 285–291.
Norseth, T. (1971). Biotransformation of methylmercuric salts in the mouse studied by specific determination of inorganic mercury. Acta Pharmacol. Toxicol. 29, 375–384.
METHYL MERCURY-INDUCED AUTOIMMUNITY Ochel, M., Vohr, H. W., Pfeiffer, C., and Gleichmann, E. (1991). Il-4 is required for the IgE and IgG1 increase and IgG1 autoantibody formation in mice treated with mercuric chloride. J. Immunol. 146, 3006 –3011. Ortega, H., Lopez, M., and Salvaggio, J. (1994). Effects of interferon gamma production after in vitro exposure to methyl mercury chloride. J. Invest. Med. 43, 67–74. Pauly, J. L., Caron, G. A., and Suskind, R. R. (1969). Blast transformation of lymphocytes from guinea pigs, rats and rabbits induced by mercuric chloride in vitro. J. Cell Biol. 40, 847– 856. Pelletier, L., Pasquier, R., Rossert, J., Vial, M. C., and Druet, P. (1988). Autoreactive T cells in mercury-induced autoimmune disease. Ability to induce the autoimmune disease. J. Immunol. 140, 750 –754. Pietsch, P., Vohr, H.-W., Degitz, K., and Gleichmann, E. (1989). Immunological alterations inducible by mercury compounds. II. HgCl2 and gold sodium thiomalate enhance serum IgE and IgG concentrations in susceptible mouse strains. Int. Arch. Allergy Appl. Immunol. 90, 47–53. Pollard, K., Lee, D., Casiano, C., Blu¨thner, M., Johnston, M., and Tan, E. (1997). The autoimmunity-inducing xenobiotic mercury interacts with the autoantigen fibrillarin and modifies its molecular and antigenic properties. J. Immunol. 158, 3521–3528. Reardon, C., and Lucas, D. O. (1987). Heavy metal mitogenesis: Zn21, Hg21 induce cellular cytotoxicity and interferon production in murine T lymphocytes. Immunobiology 175, 455– 469.
211
Reuter, R., Tessars, G., Vohr, H. W., Gleichmann, E., and Lu¨hrmann, R. (1989). Mercuric chloride induces autoantibodies against U3 small nuclear ribonucleoprotein in susceptible mice. Proc. Natl. Acad. Sci. USA 86, 237–245. Stiller-Winkler, R., Radaszkiewicz, T., and Gleichmann, E. (1988). Immunopathological signs in mice treated with mercury compounds - I. Identification by the popliteal lymp node assay of responder and non-responder. Int. J. Immunopharmacol. 10, 475– 484. Thuvander, A., Sundberg, J., and Oskarsson, A. (1996). Immunomodulation effects after perinatal exposure to methylmercury in mice. Toxicology 114, 163–175. van Vliet, E., Uhrberg, M., Stein, C., and Gleichmann, E. (1993). MHC control of Il-4-dependent enhancement of B cell Ia expression and Ig class switching in mice treated with mercuric chloride. Int. Arch. Allergy Appl. Immunol. 101, 392– 401. Warfvinge, K., Hansson, H., and Hultman, P. (1995). Systemic autoimmunity due to mercury vapor in genetically susceptible mice: dose-response studies. Toxicol. Appl. Pharmacol. 132, 299 –309. WHO, Environmental Health Criteria 118. Inorganic mercury. 1991, Geneva: World Health Organization. Wild, L. G., Ortega, H. G., Lopez, M., and Salvaggio, J. E. (1997). Immune system alteration in the rat after indirect exposure to methyl mercury chloride or methyl mercury sulfide. Env. Res. 74, 34 – 42.