Journal of Autoimmunity (1997) 10, 441–446

Major Histocompatibility Complex Class II Antigens areRequired for both Cytokine Production and ProliferationInduced by Mercuric Chloride in Vitro

Hui Hu, Go ran Mo ller and Manuchehr Abedi-Valugerdi

Department of Immunology, ArrheniusLaboratories for Natural Sciences,Stockholm University, S-10691,Stockholm, Sweden

Received 30 July 1996Accepted 20 May 1997

Autoimmune diseases induced by mercuric chloride are genetically deter-mined, at least one gene being major histocompatibility complex (MHC)-linked. Previously, we showed that in vitro mercury stimulation induced ahigh proliferative response in lymphocytes from susceptible mice (high-responders) and that the proliferative response could be restored in lym-phocytes from low-responders by pretreating the cells with mercury. We alsofound that the continuous presence of mercury induced IL-2 and IFN-ãproduction, while pretreatment with mercury induced IL-4 production. In thisstudy, we showed that anti-MHC class II monoclonal antibodies blocked boththe mercury-induced proliferative responses in lymphocytes from high-responders and the restored proliferative responses in low-responders. Inaddition, anti-MHC class II antibodies also inhibited the mercury-inducedIL-2, IFN-ã and IL-4 cytokine production in vitro. The results demonstrate thatMHC class II antigens directly participate in mercury-induced cytokineproduction and cell activation, and are required at the onset of the initiation.

© 1997 Academic Press Limited

Key words: Majorhistocompatibility complex classII, mercury, proliferation,cytokine

Introduction

Autoimmune diseases induced by mercuric chlorideare genetically determined. Susceptibility is under thecontrol of three or four genes, one of which is withinthe major histocompatibility complex (MHC) class IIregion [1–4]. Mice possessing the H-2s haplotypeare particularly susceptible to mercury and respondwith T cell-dependent polyclonal B cell activation,increased levels of serum IgE and IgG1, formation ofanti-nuclear or anti-nucleolar autoantibodies, andrenal immune complex deposits which lead toglomerulonephritis [5–8]. In contrast, mouse strainswith other MHC class II haplotypes show one or someof the autoimmune manifestations, while DBA/2(H-2d) mice are completely resistant [5, 6, 9].In mercury-induced autoimmune disease, the pro-

duction of certain autoantibodies, such as anti-nucleolar/anti-fibrillarin autoantibodies, are mappedto the H-2A region and found only in mercury-treatedH-2s mouse strains, but not in susceptible mousestrains with other MHC class II haplotypes [5, 6].Some other forms of autoantibody production,such as that of anti-chromatin and anti-histone

440896-8411/97/050441+06 r25.00/0/au970150

autoantibodies, however, could not be linked to theH-2 complex [5, 6]. In BALB/c mice, lacking bothanti-nuclear and anti-nucleolar autoantibodies, theformation of renal mesangial and vessel wall immunecomplex deposits is found to be linked to one or moreH-2 loci to the right of the A region [6]. Thus, thedifferent patterns of autoantibody specificities in dif-ferent mouse strains raised the question of what roleMHC class II antigens may play in the activationprocess and the development of the autoimmunemanifestations.How mercury ions primarily activate the immuno-

competent lymphocytes is not known. In addition, therole of MHC class II antigens in the mercury-inducedimmune response has not yet been studied, though ithas been suggested that mercury might have directnon-specific effects on lymphocytes [10–13]. Pre-viously we found that in vitro mercuric chlorideinduced a strong lymphoproliferative response inlymphocytes from susceptible mice, but only a weakresponse in resistant mice (low-responders) [14, 15].By pretreating the cells with mercury, followed bywashing away the excess mercury, a high proliferativeresponse could be restored in lymphocytes from low-responders [15]. The reversal of low-responsiveness inmercury low-responders shed more importance onwhat role MHC class II antigens may play in mercury-induced response. Recently we have been able todetect the cytokine production induced by mercury

Correspondence to: Hui Hu, Department of Immunology, ArrheniusLaboratories for Natural Sciences, Stockholm University, S-10691Stockholm, Sweden. Fax: +46-8-154163.

1© 1997 Academic Press Limited

442 H. Hu et al .

in vitro at the protein level [15]. In this study, wemainly address the involvement of MHC class IIantigens in mercury-induced primary cytokineproduction and cell proliferation.

Materials and Methods

Animals

BALB/c (H-2d), DBA/2 (H-2d) and SJL (H-2s) micewere bred in our animal facilities at the Department ofImmunology, Stockholm University. All mice wereused at 6–10 weeks of age.

Stimuli and reagents

Stock solution of mercuric chloride (HgCl2) (Merck,Darmstadt, Germany) was prepared in physiologi-cal saline at a concentration of 1×10−3 M andfilter sterilized (0.22 ìM, Costar, Cambridge, MA).Lipopolysaccharide (LPS) was extracted from E. coli0.55:B5 (Dept. of Bacteriology, Karolinska Institute,Stockholm, Sweden). Concanavalin (Con A) waspurchased from Pharmacia (Uppsala, Sweden).

Monoclonal antibodies

Some monoclonal antibodies were produced from thefollowing hybridoma cell lines: anti-IgM (rat IgG)from 3704, anti-H-2Kd (mouse IgG) from HB159, anti-H-2Ad (mouse IgG) from HB3, anti-H-2As (mouseIgG) from TIB92, anti-interferon-ã (IFN-ã) fromRA-6A2 (capture) and AN18 (detection). The antibodyculture supernatant was purified by ammonium sul-phate precipitation and passage through protein Gcolumn after extensive dialysis against PBS. Theamount of Ig was determined by spectrophotometricquantitation. Anti-H-2Ed (mouse IgG), anti-IL-2 andanti-IL-4 antibodies (both capture and detection)were purchased from Pharmingen (San Diego, CA).Anti-Lyt-1 (rat IgG) antibodies were purchased fromBecton Dickinson (California, CA).

Cell preparation and culture conditions

Spleen cells from different mouse strains were washedtwice in Earle’s balanced salt solution (BSS) andcultured at a concentration of 3×106 cells/ml in RPMI1640 medium containing 10% fetal calf serum (FCS),NaHCO3 (0.075%), penicillin/streptomycin (50 IU/50 ìg), HEPES (10 mM) and L-glutamine (2 mM) inflat-bottom 96-well plates at a final volume of 0.2 ml/well. Mercuric chloride of appropriate concentration(10 ìM) was added to the medium at the start of thecultures. In mercury-washing-away experiments, after1 day’s cultivation with mercuric chloride (10 ìM), thecells were washed twice with BSS and the living cells

were recovered by lympholyte-M gradient (CedarlaneLaboratories Ltd, Canada) and were incubated untilday 3 (3×106 ml; 0.2 ml/well; 96-well plate) withoutfurther addition of mercury. In the blocking exper-iments, the different monoclonal antibodies wereadded at indicated concentrations. Cells were culturedat 37°C in a 5% CO2 incubator.

3H-thymidine incorporation assay

3H-thymidine (Amersham International, AmershamUK) was added (2 ìCi/ml) to the cultures (3×106 cells/ml; 0.2 ml/well; 96-well plate) at differenttimes. After 6 h the cells were harvested (semi-automatic cell harvester, SKATRON, Lier, Norway)and the incorporated radioactivity was analysedby liquid scintillation counting in a â-counter (1218Rackbeta Liquid Scintillation Counter; LKB Vallac,Bromma, Sweden).

Cytokine detection

In brief, splenic IL-2, IFN-ã and IL-4-producing cellswere enumerated by using ELISPOT assay [15].Spleen cells cultured with mercuric chloride(10 ìm) were added (3×106/ml; 0.1 ml/well) to tripli-cate wells of nitro-cellulose plates (Millipore AB,Stockholm, Sweden) coated with anti-IL-2 (10 ìg/ml),anti-IL-4 (15 ìg/ml) or anti-IFN-ã (15 ìg/ml) anti-bodies. After different times of incubation, the plateswere washed and the corresponding biotinylateddetection antibodies (1 ìg/ml) were added. After2–4 h incubation at RT, the plates were washed andincubated with streptavidine-alkaline-phosphatase(-ALP) for 90 min. BCIP/NBT substrate solution(Bio-Rad, Richmond, CA) was added and incubationperformed for 1–2 h until dark spots emerged. Spotswere counted in a dissection microscope (×40). In theblocking experiments, the monoclonal antibodieswere added at the beginning of the culture.

Statistical analysis

Results were subjected to statistical analysis by theStudent’s t-test.

Results

Anti-MHC class II antibodies block HgCl 2-inducedproliferative response in high-responders

Although MHC-linked susceptibility and the increaseof MHC class II antigen expression after mercuryinjection have indicated a role of MHC class II anti-gens in the mercury-induced autoimmune disorder[16, 17], direct involvement of MHC molecules in

MHC class II antigens in mercury stimulation 443

mercury-induced lymphocyte activation has not beenstudied. In order to analyse such direct engagement ofMHC molecules in lymphocyte activation by mercury,mercury-stimulated spleen cells from high-responderBALB/c mice were cultured in the presence orabsence of monoclonal antibodies against MHC anti-gens (both class I and II). As shown in Figure 1, theaddition of anti-MHC class II antibodies (either anti-H-2Ad or anti-H-2Ed) partially blocked the prolifera-tive response induced by mercuric chloride inBALB/c mice, whereas the addition of both of theanti-MHC class II antibodies to the cell cultureresulted in complete inhibition of the proliferativeresponse. Anti-MHC class I antibodies (anti-H-2Kd

antibodies) had no effect (Figure 1). The additionof anti-MHC class II antibodies did not block theproliferative response induced by the lymphocytemitogens Con A or LPS (data not shown). We alsoperformed analogous blocking experiments inanother mercury high-responder strain (SJL) with adifferent MHC haplotype (H-2s) and found thatthe addition of anti-H-2As antibodies completelyblocked the proliferative response induced bymercury (Figure 2).

Anti-MHC class II antibodies block the restoredproliferative response in mercury low-respondersat the initial stage

Recently we have shown that a high proliferativeresponse can be restored in lymphocytes from low-responder mice by pretreating the cells with mercuryand washing away the excess mercury [15]. In the nextseries of experiments, we further investigatedwhether and at which stage anti-MHC class II anti-bodies could interfere with the proliferative responseof lymphocytes from low-responders in mercury-washing-away experiments. Anti-MHC class II anti-bodies were added either at the beginning of cultureor to the lymphocytes which had been pretreatedwith mercury. As shown in Figure 3A & B, anti-MHCclass II antibodies added at the beginning of cultureinhibited the cell proliferation in both BALB/c (high-responder) and DBA/2 (low-responder) mice. How-ever, once the lymphocytes had been pretreated withmercuric chloride for 1 day, the addition of anti-MHCclass II antibodies had no inhibitory effect. Anti-MHCclass I antibodies had no effect at any stage.

15

0

Med

ium

SI

10

5

Hg

Hg

+ an

ti-H

-2A

d

Hg

+ an

ti-H

-2E

d

Hg

+ an

ti-I

gM

Hg

+ an

ti-H

-2K

d

Hg

+ an

ti-H

-2A

d an

d H

-2E

d

*

*

*

Figure 1. Blockade of anti-MHC class II antibodies onmercury-induced cell proliferation in BALB/c mice. Spleencells (3×106/ml) from BALB/c mice were incubated with10 ìM mercuric chloride as described in Materials andMethods. Different monoclonal antibodies were added atthe beginning of culture at a final concentration of 2.5 ìg/ml. On day 3 3H-thymidine was added (2 ìCi/ml) and thecells were pulsed for 6 h. Each column represents themean±SD SI of three separate experiments. SI: stimulationindex, calculated by dividing cpm from wells with mercuryor together with antibodies by cpm from medium wells.*P<0.01 compared with stimulation of mercuric chloridealone.

100

0

Sti

mu

lati

on (

%)

50

0 1.0 2.5 5.0Antibody concentration (µg/ml)

Figure 2. Blockade of anti-MHC class II antibodies onmercury-induced cell proliferation in SJL mice. Spleen cells(3×106/ml) from SJL mice were incubated with 10 ìMmercuric chloride as described in Materials and Methods.Different monoclonal antibodies were added at the begin-ning of culture at different concentrations. On day 33H-thymidine was added (2 ìCi/ml) and the cells werepulsed for 6 h. Results represent the mean±SD of threeseparate experiments. •, Anti-H-2As antibodies; C, anti-Lyt-1antibodies.

444 H. Hu et al .

Anti-MHC class II antibodies block HgCl 2-inducedcytokine production in vitro

We have found that in vitro the continuous presence ofmercury induced IL-2 and IFN-ã production and thatpretreatment with mercury induced IL-4 productionin mouse spleen cells [15]. In rats, because of a rapiddetection of cytokine mRNA after mercury stimula-tion in vitro [11], it has been suggested that themercury-induced cytokine production might be dueto direct, non-specific effects of the metal on cells[10–13]. In mice, our finding that MHC class II anti-gens were required at the onset of mercury-inducedlymphocyte activation led us to test whether MHCclass II antigens are required in mercury-inducedcytokine production. Spleen cells from SJL mice werestimulated by continuous presence of mercury orwere pretreated with mercury in the presence orabsence of anti-MHC class II antibodies and cytokineproduction was measured by ELISPOT assays. Asshown in Figure 4, the addition of anti-MHC class IIantibodies at the beginning of culture almost com-pletely blocked the IL-2 and IFN-ã productioninduced by the continuous presence of mercury. Fur-thermore, the simultaneous incubation of anti-MHCclass II antibodies during mercury pretreatment alsocompletely blocked the IL-4 production induced afterwashing (Figure 5). Control antibodies did not haveany effect (not shown).

Discussion

In this study, we have shown that anti-MHC class IIantibodies not only blocked the mercury-induced pro-liferative response in spleen cells from high-respondermice but also inhibited the restored proliferativeresponse in low-responder mice when added at thebeginning of cell culture. The results clearly demon-strated that MHC class II antigens are directlyinvolved in the mercury-induced cell activation oflymphocytes and are required at the onset of activa-tion. That anti-H-2A or anti-H-2E antibodies alonecould partially block the mercury-induced prolifera-tive response suggests that at the initial stage bothH-2A and H-2E molecules were participating in theactivation.The blocking effects of anti-MHC class II antibodies

on cytokines production both in the continuouspresence of mercury and in the washing-away exper-iments suggest that MHC class II antigens are essen-tial in mercury-induced primary cytokine productionas well. In rats, however, it has been shown thatmercuric chloride induced a quick, transient increaseof IL-4 mRNA in purified T cells from BN rats within2 h in vitro [11]. The swift turn-on of cytokine produc-tion indicated that mercury might activate lym-phocytes by direct effects. In mice, although theinduction of cytokine production in purified T cellsby mercury has not been studied, the fact that deple-tion of adherent spleen cells completely abolishedmercury-induced cytokine production made this

possibility unlikely (data not shown). The possibledifference in the requirement of MHC class II antigensin mercury-induced cytokine production betweenmouse and rat is not yet clearly understood.

15

0

cpm

(×

10–3

)

12

6

9

3

15

0

cpm

(×

10–3

)

12

6

9

3

A

B

Figure 3. Initial involvement of MHC class II antigens inmercury stimulation. Spleen cells (3×106/ml) from BALB/c(A) or DBA/2 (B) mice were incubated with 10 ìMmercuricchloride as described in Materials and Methods. After 1 dayculture, the cells were washed twice by BSS. The living cellswere recovered by lympholyte-M gradient and incubatedfor 2 days more. On day 3 3H-thymidine was added (2 ìCi/ml) and the cells were pulsed for 6 h. Each column repre-sents the mean±SD cpm of triplicate cultures. Results wererepresentative of two separate experiments. (1) Anti-H-2Ad

or anti-H-2Kd (2.5 ìg/ml) antibodies were added togetherwith mercuric chloride at the beginning of culture. Afterwashing, no more antibodies were added. (2) Anti-H-2Ad

or anti-H-2Kd (2.5 ìg/ml) antibodies were added to therecovered cells after washing. , medium; , Hg (wash); ,(1) Hg+anti-H-2Ad (wash); , (1) Hg+anti-H-2Kd (wash);, (2) Hg(wash)+anti-H-2Ad; , (2) Hg(wash)+anti-H-2Kd.

MHC class II antigens in mercury stimulation 445

In vivo, it has been shown that mercury inducedMHC class II antigen expression in renal tubulecells and caused infiltration of T cells [18]. Theinduction of renal MHC expression preceded thedevelopment of immune complexes in the glomeruli.When MHC induction was blocked by anti-IFN-ãantibodies, the amount of infiltrating T cells wasreduced to minimal [18]. Such findings suggest thatincreased MHC expression might present antigenswhich are recognized by T cells and cause theinfiltration [18]. It has been shown that the produc-tion of anti-nucleolar autoantibodies was CD4+ Tcell-dependent [19], which indirectly suggests theinvolvement of MHC class II molecules in the devel-opment of mercury-induced autoantibody formationand autoimmune disease. Our in vitro results thatMHC class II antigens are necessary for the initiallymphocyte activation provide the basis to supportthis idea.In mercury-induced cell activation and cytokine

production, though we showed the direct involve-ment of MHC class II antigens, what role the mercuryion itself may play is still not known. Several mecha-nisms have been suggested to explain how mercuryprimarily activates the lymphocytes:

(1) Redox-linked cross-linkage of cell surface recep-tors by bivalent mercury ions delivers integratedsignals and induces cell activation directly [10].(2) We hypothesized that mercury ions bind to mem-brane proteins, such as MHC class II molecules,modify and transform them into superantigens, whichthen induces a polyclonal activation.

(3) Mercury binding to peptide/protein inducespresentation of altered/cryptic self-peptides whichconsequently activate the immune system [13, 20–21].

Since both superantigen- and altered/cryptic-peptide-induced cell activation, but not redox-induced activation, require MHC class II antigens, ourresults to a very large extent exclude the redox mech-anism, though the lowered redox potential may helpthe cell activation.In vivo, it has been shown that MHC class II antigen

expression increased in lymph node and spleenic Bcells in both susceptible and resistant mice after mer-cury injection [16]. The expression was much moreincreased and maintained for a longer time on B cellsfrom susceptible mice than resistant mice. To a certaindegree, the swift shut-down of MHC class II antigenexpression in resistant mice could be responsible forthe non-responsiveness.

AcknowledgementsThe excellent technical assistance of Ms SirkkaHellman is gratefully acknowledged.

150

0

Pos

itiv

e ce

lls/

wel

l

50

IL-2

100

**

IFN-γ

*

IL-2

**

IFN-γ

*

Day 2Day 1

Figure 4. Blockade of anti-MHC class II antibodies oncytokine production induced by the continuous presence ofmercury. Spleen cells from SJL mice were incubated with10 ìM mercuric chloride in ELISPOT plate (3×106/ml;0.1 ml/well) coated with anti-IL-2 or anti-IFN-ã antibodies.Anti-H-2As antibodies were added at the beginning ofculture. After 1 or 2 days’ culture, the cytokine productionwas measured as described in Materials and Methods. Eachcolumn represents the mean of triplicate cultures. Resultswere representative of three separate experiments. *, P<0.05;**, P<0.01 compared mercury stimulation with eithermedium control or mercury stimulation plus anti-H-2As

antibodies. , Medium; , Hg; , Hg+anti-H-2As (2.5 ìg/ml); , Hg+anti-H-2As (5 ìg/ml).

100

0

Pos

itiv

e ce

lls/

wel

l

50

*

IL-4

Figure 5. Blockade of anti-MHC class II antibodies oncytokine production induced by pretreatment with mercury.In mercury-washing-away experiments, spleen cells fromSJL mice were preincubated with 10 ìM mercuric chloridetogether with anti-H-2As antibodies for 1 day. Then cellswere washed twice with BSS and incubated in ELISPOTplate (coated with anti-IL-4 antibodies) without furtheraddition of mercury or anti-H-2As antibodies for another2 days and cytokine production was measured as describedin Materials and Methods. Each column represents the meanof triplicate cultures. Results were representative of threeseparate experiments. *, P<0.05 compared mercury stimula-tion with either medium control or mercury stimulationplus anti-H-2As antibodies. , Medium; , Hg; , Hg+anti-H-2As (2.5 ìg/ml); , Hg+anti-H-2As (5 ìg/ml).

446 H. Hu et al .

This work was supported by grants from theSwedish Medical Research Council, the Swedish WorkEnvironmental Fund, an EU concerted action project(Biomed 2), and a grant from the Swedish Foundationfor Health Care Sciences and Allergy Research.

References1. Druet E., Sapin C., Gunther E., Feingold N., Druet P.

1977. Mercuric chloride-induced anti-glomerularbasement membrane antibodies in the rat. Geneticcontrol. Eur. J. Immunol. 7: 348–351

2. Sapin C., Mandet C., Druet E., Gunther G., Druet P.1981. Immune complex type disease induced by HgCl2:genetic control of susceptibility. Trans. Pro. XIII:1404–1406

3. Sapin C., Hirsch F., Delaporte J.P., Bazin H., Druet P.1984. Polyclonal IgE increase after HgCl2 injections inBN and LEW rats: a genetic analysis. Immunogenetics20: 227–236

4. Aten J., Veninga A., Heer E.D., Rozing J., NieuwenhuisP., Hoedemaeker P.J., Weening J.J. 1991. Susceptibilityto the induction of either autoimmunity orimmunosuppression by mercuric chloride is related tothe major histocompatibility complex class IIhaplotype. Eur. J. Immunol. 21: 611–616

5. Hultman P., Bell L.J., Enestrom S., Pollard K.M. 1992.Murine susceptibility to mercury I. Autoantibodyprofiles and systemic immune deposits in inbred,congenic, and intra-H-2 recombinant strains. Clin.Immunol. Immunopathol. 65: 98–109

6. Hultman P., Bell L.J., Enestrom S., Pollard K.M. 1992.Murine susceptibility to mercury II. Autoantibodyprofiles and renal immune deposits in hybrid,backcross, and H-2d congenic mice. Clin. Immunol.Immunopathol. 68: 9–20

7. Hultman P., Enestrom S. 1992. Dose-response studiesin murine mercury-induced autoimmunity andimmune-complex disease. Toxicol. Appl. Pharmacol. 113:199–208

8. Monestier M., Losman M.J., Novick K.E., Aris J.P. 1994.Molecular analysis of mercury-induced antinucleolarantibodies in H-2s mice. J. Immunol. 152: 667–675

9. Hultman P., Enestrom S. 1987. The induction ofimmune complex deposits in mice by peroral andparenteral administration of mercuric chloride: straindependent susceptibility. Clin. Exp. Immunol. 67:283–292

10. Nakashima I., Pu M.Y., Nishizaki A., Rosila I., Ma L.,Katano Y., Ohkusu K., Rahman S.M.J., Isobe K.-I.,

Hamaguchi M., Saga K. 1994. Redox mechanism asalternative to ligand binding for receptor activationdelivering disregulated cellular signals. J. Immunol.152: 1064–1071

11. Prigent P., Saoudi A., Pannetier C., Graber P., BonnefoyJ.-Y., Druet P., Hirsch F. 1995. Mercuric chloride, achemical responsible for T helper cell (Th)2-mediatedautoimmunity in Brown Norway rats, directly triggersT cells to produce interleukin-4. J. Clin. Invest. 96:1484–1489

12. Druet P. 1995. Metal-induced autoimmunity. Hum. Exp.Toxicol. 14: 120–121

13. Griem P., Gleichmann E. 1995. Metal ion inducedautoimmunity. Curr. Opin. Immunol. 7: 831–838

14. Jiang Y., Moller G. 1995. In vitro effects of HgCl2 onmurine lymphocytes. I. Preferable activation of CD4+

T cells in responder strain. J. Immunol. 154: 3138–314615. Hu H., Abedi-Valugerdi M., Moller G. 1997.

Pretreatment of lymphocytes with mercury in vitroinduces a response in T cells from geneticallydetermined low-responders and a shift of theinterleukin profile. Immunology 90: 198–204

16. van Vliet E., Uhrberg M., Stein C., Gleichmann E.1993. MHC control of IL-4-dependent enhancement ofB cell Ia expression and Ig class switching in micetreated with mercuric chloride. Int. Arch. Allergy.Immunol. 101: 392–401

17. Dubey C., Bellon B., Hirsch F., Kuhn J., Vial M.C.,Goldman M., Druet P. 1991. Increased expression ofclass II major histocompatibility complex molecules onB cells in rats susceptible or resistant to HgCl2-inducedautoimmunity. Clin. Exp. Immunol. 86: 118–123

18. Madrenas J., Parfrey N.A., Halloran P.F. 1991.Interferon gamma-mediated renal MHC expression inmercuric chloride-induced glomerulonephritis. Kidney.Int. 39: 273–281

19. Hultman P., Johansson U., Dagnaes-Hansen F. 1995.Murine mercury-induced autoimmunity: the role ofT-helper cells. J. Autoimmun. 8: 809–823

20. Kubicka-Muranyi M., Griem P., Lubben B., RottmannN., Luhrmann R., Gleichmann E. 1995.Mercuric-chloride-induced autoimmunity in miceinvolves up-regulated presentation by spleen cells ofaltered and unaltered nucleolar self antigen. Int. Arch.Allergy. Immunol. 108: 1–10

21. Kubicka-Muranyi M., Kremer J., Rottmann N., LubbenB., Albers R., Bloksma N., Luhrmann R., GleichmannE. 1996. Murine systemic autoimmune disease inducedby mercuric chloride: T helper cells reacting to selfproteins. Int. Arch. Allergy. Immunol. 109: 11–20