- Email: [email protected]
The Use of Reporter Antigens in the Popliteal Lymph Node Assay to Assess Immunomodulation by Chemicals
TOXICOLOGY AND APPLIED PHARMACOLOGY ARTICLE NO.
143, 102–109 (1997)
TO968078
The Use of Reporter Antigens in the Popliteal Lymph Node Assay to Assess Immunomodulation by Chemicals1 RUUD ALBERS, ANGELIQUE BROEDERS, ANITA
VAN DER
PIJL, WILLEM SEINEN,
AND
RAYMOND PIETERS
Section of Immunotoxicology, Research Institute of Toxicology, Utrecht University, 3508 TD Utrecht, The Netherlands Received May 30, 1996; accepted November 27, 1996
The Use of Reporter Antigens in the Popliteal Lymph Node Assay To Assess Immunomodulation by Chemicals. ALBERS, R., BROEDERS, A., VAN DER PIJL, A., SEINEN, W., AND PIETERS, R. (1997). Toxicol. Appl. Pharmacol. 143, 102–109. Various drugs and other chemicals can induce T-cell-dependent B-cell activation which may lead to allergic or autoimmune-like diseases. Because the nature of the relevant (neo-) antigens is generally not known and probably depends on the chemical, we have explored the potential use of reporter antigens to determine T-cell-dependent B-cell activation by chemicals. TNP-Ficoll and TNP-OVA were used for this purpose because they are recognized by the same TNP-specific B cells, but these cells require distinct costimulation for specific antibody production. It was found that HgCl2 , phenytoin, nitrofurantoin, and D-penicillamine stimulated IgG1 production to both antigens, incomplete Freund’s adjuvant, silica, and dimethylsulfoxide to TNP-OVA only, and LPS and hydroxyl-amino procainamide to TNP-Ficoll alone. The diabetogene streptozotocin did not enhance IgG1 production, but may enhance a cellular response instead. Tolerogens and a T-cell antigen without intrinsic adjuvant activity did not influence the responses. The IgG1 production to TNP-Ficoll was local and transient, and did not always require T cells. In contrast, responses to TNP-OVA could be measured in serum, led to specific memory, and were strictly T-cell dependent. These results demonstrate that specific antibody production to reporter antigens indicates immunostimulatory effects of chemicals more sensitive than PLN cell count and provides important mechanistic information. Moreover, with TNP-OVA as reporter antigen the kinetics and regulation of chemically enhanced immune responses can be studied without the need to know the relevant neo-antigens for each individual compound. q 1997 Academic Press
It is well known that exposure to drugs and other lowmolecular-weight compounds (LMWC) can stimulate the immune system and may lead to allergic or autoimmunelike diseases in susceptible individuals (Anonymous, 1992; Kilburn and Warshaw, 1994). Activation of Th cells by 1 This work was supported by the Life Science Foundation, Dutch Organization for Scientific Research, Project SLW 811-423-431.
0041-008X/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
TOX 8078
/
6h16$$$281
chemically altered or induced neo-epitopes is considered pivotal, but the nature of these epitopes remains largely unknown and probably depends on the chemical and on the MHC haplotype involved (Griem et al., in press). Nevertheless, recognition of T-cell activation by LMWC is crucial to elucidate the underlying mechanism and to successfully predict immunostimulatory potential of chemicals. The popliteal lymph node assay (PLNA) in rodents is one of few models that can be used for this purpose. In this assay, PLN responses are assessed following footpad injection of chemicals without adjuvant. In general, only chemicals known to induce adverse immune effects in man, induce positive PLN reactions in mice as judged by an increase in PLN weight or cell count (Bloksma et al., 1995; Gleichmann et al., 1990; Kammu¨ller et al., 1989). However, these nonspecific parameters are rather insensitive and do not unequivocally prove T-cell involvement, because chemicals causing irritation at the injection site may induce nonspecific PLN reactions (Bloksma et al., 1995; Brouland et al., 1994). Moreover, as the specificity of these responses is unknown, their kinetics and regulation cannot easily be monitored and mechanistic studies are difficult to perform. We have recently found that the prototype autoimmunogenic chemicals HgCl2 and diphenylhydantoin [phenytoin, DPH] strongly stimulate specific B-cell responses to coinjected reporter antigens in the PLNA (Albers et al., 1996a). TNP-Ficoll and TNP-ovalbumin (TNP-OVA) were used for this purpose because they are both recognized by TNP-specific B cells, but these cells require distinct T-cell help for specific antibody production. The T-cell-independent type 2 antigen TNP-Ficoll can independently stimulate IgM production, whereas cytokines are required for isotype switch to IgG1 (Mond et al., 1995; Pike et al., 1987). TNP-OVA, on the other hand, is a T-cell-dependent antigen and B cells forming antibodies to this antigen require help in the form of T–B cell interaction and cytokines (Clark and Ledbetter, 1994). Thus, by varying the antigen, effects of LMWC on B-cell responses with different requirements for T-cell help can be evaluated, without the need to identify the relevant neo-antigens for each chemical individually.
102
02-12-97 20:20:43
toxas
AP: Tox
CHEMICAL IMMUNOMODULATION ASSESSED WITH REPORTER ANTIGENS
The aim of this study was to determine whether stimulation of specific antibody responses to the reporter antigens TNP-Ficoll and TNP-OVA is a more sensitive indication of immunostimulation than increased PLN cell count. In addition it was assessed whether these responses provide information on the mechanisms underlying stimulation by LMWC, and allow monitoring of chemically enhanced responses over prolonged periods. MATERIALS AND METHODS Mice. Female, specific-pathogen-free BALB/c mice of 8–12 weeks were obtained from the Utrecht University breeding facility, and similar BALB/c nu// and nu/nu mice were obtained from Harlan CPB (Austerlitz, Netherlands), and B10.s from Olac Ltd. (Blackthorn, Bicester, UK). Mice were housed under hygienic barrier conditions in filter-topped macrolon cages with a bedding of wood chips at a temperature of 23 { 27C and 50– 60% relative humidity, and a 12-hr light/dark cycle. They received standard lab chow and acidified tap water ad libitum. Chemicals and reagents. HgCl2 was obtained from BDH (Poole, UK), LPS (lipopolysaccharide W, Escherichia coli O111:B4) and IFA (see Table 1 for abbreviations) from Difco (Detroit, MI), PB from Brocasef bv. (Maarssen, The Netherlands), NF from ACROS Chimica (’s Hertogenbosch, The Netherlands), Immobilon-P membranes from Millipore (Etten-Leur, The Netherlands), and alkaline phosphatase-conjugated goat anti-mouse IgG1 and IgM antibodies from Southern Biotechnology Associates Inc. (Birmingham, AL). The antigens TNP-OVA and TNP-Ficoll were prepared as described previously (Albers et al., 1996a). AT was synthesized according to (Alkan et al., 1972), purified by chromatography, and was 51 recrystallized before use. Innolia-kits were obtained from Innogenetics (Zwijndrecht, Belgium). Other chemicals were obtained from Sigma Chemical Company (St. Louis, MO). PLN assays. PLN assays were performed as described (Albers et al., 1996a). In brief, LMWC were coinjected with 10 mg TNP-Ficoll or 10 mg TNP-OVA sc into the right hindpaw (heel r toe direction) of naive mice. LMWC were injected in quantities that were previously demonstrated stimulatory in the PLNA, or were equimolar to a related compound. Thus, 50 mg HgCl2 , 5.2 mmol PA, 5.2 mmol (1.5 mg) HAPA, 1 mg AT, 1 mg SI, 10 mg LPS, 1 mg PEN or 7.3 mmol PB in saline, 7.3 mmol (2 mg) DPH in (1:1) saline/carbonate buffer (0.025 M Na2CO3 and 0.12 M NaCl; pH 11), 1 mg STZ in citrate buffer (pH 6), or 1 mg NF in DMSO were mixed with the antigens and were injected sc into the right hindpaw. Alternatively, antigens in saline were emulsified (1:1) in IFA and a total volume of 50 ml containing 10 mg antigen was injected. Seven days later, draining PLN were isolated. For memory responses, mixtures were similarly injected into the left footpad of B10.s mice, which are susceptible to glomerulonephritis induced by HgCl2 . Serum samples were obtained weekly by orbita puncture under ether anesthesia, and after 4–5 weeks mice were challenged in the right paw with 10 mg antigen in saline. Six days after challenge, serum was collected and PLN and spleen were isolated. In all experiments, cells were isolated in PBS/1% BSA, washed once by centrifugation, resuspended in PBS/1% BSA, counted using a Coulter Counter (Coulter Electronics, Luton, UK), and then adjusted to 106 cells/ml. Viability was checked occasionally by trypan blue exclusion and always exceeded 95%. ELISPOT assay. These assays were essentially performed as described (Schielen et al., 1995). In brief, graded numbers of PLN cells in 500 ml PBS/1% BSA were incubated (377C, 4 hr) in wells with TNP-BSA-coated (10 mg/ml, 47C, 18 hr) Immobilon-P membranes as bottom. Thereafter, membranes were washed and incubated with predetermined optimal dilutions of alkaline phosphatase-conjugated anti-mouse IgG1 or IgM in PBS/
AID
TOX 8078
/
6h16$$$281
02-12-97 20:20:43
103
1% BSA at 47C for 18 hr. After washing, spots were developed by incubation with a substrate of 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium. Specific antibody-forming cells (AFC) per 106 cells were then calculated from spot numbers counted with the aid of a stereo microscope. Line immunoassay. TNP-specific antibodies in serum were determined using a commercially available line immunoassay, composed of strips coated with bands of isotype-specific anti-mouse Ig. This strip assay is developed to determine the isotype of mouse Ig in culture supernatants, but can also be used to simultaneously detect specific Ig of different isotypes. To this end, strips were incubated with 20-fold-diluted serum samples (16 hr, 47C), washed, incubated with TNP-conjugated alkaline phosphatase (1 hr, room temperature), and then developed according to the manufacturer’s instructions. Staining intensity of individual bands was determined using a densitometer, yielding a semiquantitative measure of TNP-specific Ig of different isotypes. Statistics. Preceding statistical analysis, the number of AFC was transformed to log 10 values to homogenize variance. Differences between group means were analyzed for each antigen separately, using one-way ANOVA with LSD post hoc test for contrasts. Pearson correlations were calculated with data of all animals, and the coefficients mentioned are all highly significant (p õ 0.001). Correlation lines were determined mathematically by minimizing the variance, using (Y 0 Yestimated)2/(1 / b2) as loss function and y Å a / bx as model in the nonlinear module of Systat (Version 5.2 for the Macintosh).
RESULTS
Effect of Different Chemicals in T-Cell-Competent Mice To determine the effect of LMWC on antibody responses to defined reporter antigens, selected chemicals (Table 1) were coinjected with TNP-Ficoll and TNP-OVA into the hindpaw of BALB/c /// mice. After 7 days, PLN cell count and TNP-specific IgM and IgG1 AFC were determined. Injection of the antigens in saline, carbonate buffer, or citrate buffer had no effect on these parameters (data not shown). However, footpad injection of most (auto)immunogenic chemicals increased PLN cell count and enhanced production of specific IgG1 to the coinjected reporter antigens, whereas neither tolerogens nor the T-cell antigen AT had these effects (Table 2). The Data in Table 2 were used to construct scatter plots that indicate correlations between the different parameters. Figure 1 clearly illustrates that PLN cellularity correlated with IgG1 responses to TNP-Ficoll, except following stimulation by the macrophage activators IFA and SI (R 2 Å 0.43; 0.72 with IFA and SI excluded). IgM and IgG1 AFC responses to this antigen were not very closely related (R 2 Å 0.47, Fig. 2). However, when TNP-OVA was used as reporter antigen, specific IgG1 AFC increased exponentially with cell count for all chemicals (R 2 Å 0.79). The maximum increase in PLN cell count was É10-fold (Table 2), whereas IgG1 production to TNP-OVA increased up to É400-fold per 106 cells, yielding a 4000-fold increase per PLN. IgM and IgG1 production to TNP-OVA increased in concert, unless LPS was used as the stimulator. This B-cell mitogen enhanced specific IgM production, but failed to induce
toxas
AP: Tox
104
ALBERS ET AL.
TABLE 1 Abbreviations and Characteristics of Compounds Used in This Study Abbreviation
Chemical
Characteristics
AT
ABA-Tyr
DMSO
Dimethyl sulfoxide
DPH HAPA
Diphenylhydantoin Hydroxyl-amino procainamide
HgCl2 IFA LPS
HgCl2 Incomplete Freund’s adjuvant Lipopoly saccharide
NF PA PB
Nitrofurantoin Procainamide Phenobarbital
PEN SI
D-Penicillamine Silica
STZ
Streptozotocin
Complete T-cell antigen of low molecular weight Organic solvent, occasionally used in PLN assay Autoimmunogen Autoimmunogenic metabolite of procainamide Autoimmunogen Macrophage activator B-cell antigen and B-cell mitogen Autoimmunogen Tolerogen Tolerogen, structurally related to diphenylhydantoin Autoimmunogen, sensitizer Macrophage activator, inducer of scleroderma Autoimmunogen
switch to IgG1 (R 2 Å 0.68; 0.77 without LPS). STZ behaved different from the other chemicals in that it increased PLN size and weight (not shown) but not cellularity, and enhanced IgG2a rather than IgG1 production to TNP-OVA (IgG2a data not shown). Responses in T-Cell-Competent Versus Incompetent Mice The role of T cells in stimulation by HgCl2 and DPH was determined by comparing antibody production to the reporter antigens in T-cell-competent (///, nu//) and incompetent (nu/nu) BALB/c mice. HgCl2 increased PLN cell count in /// and nu//, but not in nu/nu mice, and accordingly enhanced specific AFC responses in T-cell-competent mice only (Table 3). DPH on the other hand, increased cell count most effectively in /// and nu// mice, but to a lesser extent also in nu/nu mice, particularly when combined with TNPFicoll. Interestingly, DPH stimulated IgM and IgG1 responses to TNP-Ficoll similarly in ///, nu//, and nu/nu mice, whereas responses to TNP-OVA were only enhanced in /// but not in nu/nu mice. Induction of Memory To evaluate whether stimulation by LMWC leads to systemic responses to the reporter antigen, T-cell-competent mice received one injection of TNP-Ficoll or TNP-OVA with saline, HgCl2 , or DPH. TNP-specific IgM but not IgG1 was increased in serum of TNP-Ficoll-immunized mice, and this IgM was produced mainly in the spleen (not shown). Moreover, challenge with TNP-Ficoll did not elicit second-
AID
TOX 8078
/
6h16$$$281
References
02-12-97 20:20:43
Alkan et al. (1972) Kammu¨ller et al. (1989) Alarco´n-Segovia and Alarco´n-Riquelme (1989) Uetrecht (1992); Kubicka-Muranyi et al. (1993) Schrallhammer-Benkler et al. (1992) Janeway and Travers (1994) Janeway and Travers (1994) Alarco´n-Segovia and Alarco´n-Riquelme (1989) Kubicka-Muranyi et al. (1993) Gleichmann (1981) Emery and Panayi (1989) Cabral et al. (1994) Kantwerk-Funke et al. (1991)
ary responses (Fig. 3) and serum levels were not influenced by the stimulating LMWC. In contrast, mice immunized with TNP-OVA plus HgCl2 or DPH, but not with TNP-OVA alone, had TNP-specific IgM as well as IgG1 in their serum. Challenge with TNP-OVA largely increased these levels and induced numerous TNP-specific IgG1 AFC in PLN, spleen, and bone marrow of these mice (AFC data not shown). DISCUSSION
It has been well established that increased PLN cell count following footpad injection of LMWC in mice correlates with immunostimulating potential of such compounds in man. Here it is shown that such LMWC also stimulate specific antibody production to coinjected reporter antigens in the PLN. In particular IgG1 production to TNP-OVA is a far more sensitive indicator of immunostimulation by LMWC than cell count, and this response is not stimulated by tolerogens (PB, PA), nor by the T-cell antigen (AT) that lacks intrinsic adjuvant activity. Importantly, these responses to defined reporter antigens may also provide information on the mechanisms underlying immunostimulation by chemicals. Antibody production to TNP-Ficoll and TNP-OVA was differently affected by distinct (sets of) compounds (Fig. 4). HgCl2 , DPH, PEN, and NF were the most potent immunostimulators as they increased IgG1 production to TNP-Ficoll as well as to TNP-OVA. The polysaccharide TNP-Ficoll cannot be recognized by T cells, and LMWC most likely
toxas
AP: Tox
AID
TOX 8078
/
6h16$$8078 2 1 4 3 4 5
AP: Tox
49 169 91 71 88 102
{ { { { { {
13** 91** 12** 27** 12** 34**
25 { 7* 48 { 17** 55 { 8**
33 { 12* 35 { 6**
14 { 3 23 { 5* 6{ 4
IgM (AFC/106 cells)
4 12 7 5 6 7
2 3 4
3 3
2 õ1
SI
33 14 58 42 105 215
{ { { { { {
18** 5 5** 20** 19** 56**
10 { 3 12 { 3* 18 { 4**
9{ 2 19 { 7*
8{ 3 4{ 1 6{ 1
IgG1 (AFC/106 cells)
4 2 7 5 13 27
1 2 2
1 2
õ1 õ1
SI
6.4 3.2 12.0 11.6 15.8 13.3
{ { { { { {
1.1** 0.4 1.0** 1.3** 1.1** 2.2**
9.3 { 1.6** 7.6 { 1.3** 17.5 { 0.9**
3.5 { 0.7 5.9 { 1.3**
1.5 { 0.9 2.9 { 0.6 2.7 { 0.5
Cells (1106)
4 2 8 8 11 9
6 5 12
2 3
2 2
SI
18 14 207 397 118 39
{ { { { { {
11* 4* 36** 159** 25** 14**
58 { 19** 47 { 10** 97 { 14**
14 { 5* 83 { 16**
6{ 9 12 { 2* 21 { 11**
IgM (AFC/106 cells)
TNP-OVA
3 2 35 66 20 7
10 8 16
2 14
2 4
SI
5 0 7*
6 6 683 1104 1269 165
{ { { { { {
2 3 154** 341** 254** 65**
145 { 28** 55 { 16** 797 { 225**
5{ 3 20 { 18
3{ 1{ 16 {
IgG1 (AFC/106 cells)
2 2 228 368 423 55
48 18 266
2 7
õ1 5
SI
a TNP-Ficoll and TNP-OVA were injected alone or in combination with the indicated chemicals into the footpad of naive BALB/c mice. Seven days later, the total number of cells and the number of TNP-specific IgM and IgG1 AFC in the draining PLN were determined. Values indicated are arithmetic means { SEM of 6–9 mice (NF and PEN n Å 3) and as stimulation indices (SI) of treatment/control group. Significant differences from mice injected with antigen and saline are indicated (*p õ 0.05 and **p õ 0.01).
0.7* 0.5 1.9** 1.3* 0.8* 1.0**
{ { { { { {
toxas 4.4 2.4 9.2 5.6 9.3 9.9
HAPA (autoimmunogen) STZ (autoimmunogen) NF (autoimmunogen) PEN (autoimmunogen) DPH (autoimmunogen) HgCl2 (autoimmunogen)
2 2
2.7 { 0.1 4.5 { 0.7*
AT (T-cell antigen) LPS (B-cell mitogen) 2 3 12
1 õ1
2.1 { 0.2 2.0 { 0.4 1.3 { 0.4
Saline (control) PA (tolerogen) PB (tolerogen)
4.4 { 0.9 6.5 { 0.6** 24.9 { 2.7**
02-12-97 20:20:43
DMSO (organic solvent) SI (MPh activator) IFA (MPh activator)
SI
Cells (1106)
TNP-Ficoll
TABLE 2 Summary of PLN Cell Count and TNP-Specific IgM and IgG1 AFC, Found after Injection of Indicated Chemials with the Reporter Antigen TNP-Ficoll and TNP-OVAa
CHEMICAL IMMUNOMODULATION ASSESSED WITH REPORTER ANTIGENS
105
106
ALBERS ET AL.
FIG. 1. Scatter plot of TNP-specific IgG1 AFC versus PLN cellularity, after injection of various chemicals with the reporter antigens TNP-Ficoll and TNP-OVA. BALB/c /// mice were immunized sc into the right hindpaw with TNP-Ficoll or TNP-OVA combined with one of several chemicals. Seven days later, draining PLN were isolated and cell count and TNP-specific IgG1 AFC/106 cells were determined. The areas indicated in the plot define the mean { SEM of 6–9 mice (PEN and NF, n Å 3). The black area indicates values obtained with antigen in saline, and the other areas show results of injection of antigen with the indicated chemical. Correlations calculated with the individual data are R 2 Å 0.43 for TNP-Ficoll (dashed line), 0.72 when SI and IFA are excluded (solid line), and 0.79 for TNP-OVA (solid line).
stimulate this response by creating neo-epitopes that lead to activation of T cells. Cytokines produced by those T cells can then stimulate production of IgG1 by B cells that recognize TNP-Ficoll (Albers et al., 1996a). IgG1 production to TNP-OVA, on the other hand, requires contact help provided during cognate interaction of TNP-OVA-specific T and B cells (Fig. 4). This response can be stimulated by tissue damage or release of proinflammatory cytokines leading to local inflammation and activation of antigen presenting cells. LMWC that enhance both responses therefore have the in-
trinsic adjuvant activity that stimulates responses to nonself antigens, and in addition may create such antigens by modifying self antigens. Macrophage-activating LMWC like IFA and SI and the solvent DMSO cannot stimulate IgG1 production to TNPFicoll, but do stimulate the response to TNP-OVA. The response to TNP-OVA is probably increased because activated macrophages present TNP-OVA more efficiently, leading to activation of specific Th cells, T–B cooperation, and subsequent IgG1 production. Clearly the response to TNP-Ficoll
FIG. 2. Scatter plot of TNP-specific IgM and IgG1 AFC per 106 PLN cells. TNP-specific IgM and IgG1 AFC were determined in the same mice described in the legend to Fig. 1, and are similarly represented. Correlations are R 2 Å 0.47 for TNP-Ficoll (solid line) and 0.68 (solid line) for TNPOVA (0.77 with LPS excluded, dashed line).
AID
TOX 8078
/
6h16$$$281
02-12-97 20:20:43
toxas
AP: Tox
107
CHEMICAL IMMUNOMODULATION ASSESSED WITH REPORTER ANTIGENS
TABLE 3 PLN Cell Count and TNP-Specific IgG1 AFC, after Injection of HgCl2 and DPH with TNP-Ficoll or TNP-OVA into Normal and T-Cell-Deficient Micea Cells (1106)
Antigen
SI
TNP-specific IgG1/106 cells
SI
Strain
Stimulant
// / nu// nu/nu
None None None
None None None
// / nu// nu/nu // / nu// nu/nu
HgCl2 HgCl2 HgCl2 DPH DPH DPH
TNP-Ficoll TNP-Ficoll TNP-Ficoll TNP-Ficoll TNP-Ficoll TNP-Ficoll
10.3 11.7 1.0 11.3 8.7 6.3
{ { { { { {
2.8 3.7 0.6* 1.7 0.8 1.3*
5 7 õ1 6 5 5
152 70 2 77 85 128
{ 62 { 8 { 1** { 9 { 23 { 70
38 35 õ1 19 43 16
// / nu/nu // / nu/nu
HgCl2 HgCl2 DPH DPH
TNP-OVA TNP-OVA TNP-OVA TNP-OVA
17.3 2.1 14.8 4.5
{ { { {
2.8 0.9* 3.8 1.3
9 2 8 4
188 2 792 7
{ 144 { 0* { 83 { 6**
47 õ1 198 õ1
1.9 { 0.2 1.7 { 0.2 1.2 { 0.2
4{ 2{ 8{
1 1 3
a TNP-Ficoll and TNP-OVA were injected with HgCl2 and DPH into BALB/c ///, nu//, and nu/nu mice. Seven days later, PLN cell count and the number of TNP-specific IgG1 AFC/106 cells were determined. Compiled data from contralateral PLN are given as baseline values for each strain. Data are indicated as arithmetic means { SEM of 3–6 mice and stimulation index (SI) of treatment/baseline. Significant differences from /// mice treated with same chemical and reporter antigen are indicated (*p õ 0.05 and **p õ 0.01).
cannot be stimulated in this manner and these LMWC apparently do not create neo-antigens that activate other T cells. Selective stimulation of IgG1 production to TNP-OVA, but not to TNP-Ficoll, may therefore identify LMWC that have intrinsic adjuvant activity, but do not create neo-epitopes. Other compounds like LPS, HAPA, and in nu/nu mice also DPH, increased IgG1 production to TNP-Ficoll, but not the response to TNP-OVA. LPS is a B-cell mitogen and at the dose used directly stimulates IgM production to both antigens, and to some extent augments IgG1 production to TNP-Ficoll. It is currently unclear why HAPA also selectively stimulated IgM and IgG1 production to TNP-Ficoll, as results of others have indicated that this compound can elicit a T-cell response by itself, suggesting intrinsic adjuvant activity (Kubicka-Muranyi et al., 1993). Interestingly, DPH stimulates T-cell-dependent IgG1 production to TNP-OVA exclusively in normal mice, but increases IgG1 production to TNP-Ficoll in normal and T-cell-deficient mice alike. This effect and the previously observed low, but significant, increased PLN cell count in nu/nu mice (Gleichmann, 1981, Kammu¨ller et al., 1987) are not likely caused by the few T cells present in PLN of nu/nu mice (õ3% of T cells in PLN of /// and nu//; R. Albers, unpublished results). Instead, these results may indicate that DPH has a T-cell-independent mode of action, on top of its undisputed effect on T cells. It remains to be established whether this results from a direct LPS-like effect on B cells or whether DPH stimulates production of switch-inducing cytokines by non-T cells (i.e., NK or mast cells (Mond et al., 1995).
AID
TOX 8078
/
6h16$$$281
02-12-97 20:20:43
FIG. 3. TNP-specific antibodies in serum. B10.s mice received one injection of TNP-Ficoll or TNP-OVA with saline (m), HgCl2 (s), or DPH (l) and were challenged 4–5 weeks later in the other hindpaw with antigen alone. Serum was collected weekly and TNP-specific IgM and IgG1 were determined; 0, before immunization, 1, 2, 3, 1 week after, 2 weeks after, and 3 weeks after immunization; C, 6 days after challenge. Values are arithmetic means of five mice per group.
toxas
AP: Tox
108
ALBERS ET AL.
regulation of chemically enhanced responses to symptomatic adverse immune effects without the need to know the relevant neo- or auto-antigens for each LMWC individually. ACKNOWLEDGMENTS HAPA was kindly provided by Prof. Dr. J. Uetrecht, Toronto, Canada. The helpful discussions with Drs. M. Frens, C. de Heer, E. Urrestarazu, W. Vaes, and H. Verhaar are gratefully acknowledged.
REFERENCES
FIG. 4. Schematic representation of the effect of different LMWC on IgG1 production to the reporter antigens TNP-Ficoll and TNP-OVA. For explanation see text.
STZ was distinct from all other compounds as it increased PLN weight but not cell count and increased IgG2a rather than IgG1 production to TNP-OVA. Moreover, production of interferon-g was increased and CD8/ cells and macrophages accumulated in STZ-stimulated PLN (manuscript in preparation). These and other observations indicate that unlike the other LMWC that drive humoral type 2 immune responses, STZ stimulates cell-mediated type 1 responses. This is in line with the antibody-mediated adverse effects of the other compounds, versus the cellular mechanism underlying STZinduced autoimmune diabetes. It appears therefore that the isotype of the antibodies produced to TNP-OVA indicates the type of response enhanced by a LMWC. The distinct characteristics of TNP-Ficoll and TNP-OVA were also reflected in the systemic responses developing after chemical stimulation. With TNP-Ficoll only TNP-specific antibodies of IgM isotype were found in serum; these were produced mainly in the spleen, and were not influenced by stimulating LMWC. This indicates that LMWC only transiently stimulate local cytokine-dependent IgG1 production to TNP-Ficoll. However, chemical stimulation of responses to TNP-OVA yielded increased serum levels of TNP-specific IgG1 and induced specific T- and B-cell memory to the antigen. This indicates that it may be feasible to compare the type and magnitude of such responses in mouse strains with varying susceptibility to the adverse immune effects of the stimulating LMWC, and relate this to the observed pathology (Albers et al., 1996b, manuscripts in preparation). In summary, we have demonstrated that reporter antigens can be very instrumental in detecting the immunostimulating potential of LMWC. It is shown that reporter antigen-specific IgG1 production provides a very sensitive parameter that can also provide information on the mechanism underlying the stimulation. The isotype of the antibodies produced to TNPOVA can indicate the type of response stimulated, and antibodies to reporter antigens may eventually link induction and
AID
TOX 8078
/
6h16$$$281
02-12-97 20:20:43
Alarco´n-Segovia, D., and Alarco´n-Riquelme, M. (1989). Autoimmune reactions in humans induced by diphenylhydantoin and nitrofurantoin. In Autoimmunity and Toxicology. Immune Disregulation Induced by Drugs and Chemicals (M. E. Kammu¨ller, N. Bloksma, and W. Seinen, Eds.), pp. 151–166. Elsevier, Amsterdam/New York/Oxford. Albers, R., van der Pijl, A., Seinen, W., Pieters, R., and Bloksma, N. (1996a). The autoimmunogenic chemicals HgCl2 and diphenylhydantoin stimulate IgG production to TNP-ficoll and TNP-OVA, supporting and extending the graft-versus-host hypothesis for chemical induction of autoimmunity. Immunology 86, 468–473. Albers, R., Volckmann, R., van der Pijl, A., Bol, M., Seinen, W., and Pieters, R. (1996b). Breaking of tolerance and MHC-dependent immune modulation by autoimmunogenic chemicals. Fundam. Appl. Toxicol. 30(Suppl.), 62. [Abstract] Alkan, S. S., Williams, E. B., Nitecki, D. E., and Goodman, J. W. (1972). Antigen recognition and the immune response. Humoral and cellular immune responses to small mono- and bifunctional antigen molecules. J. Exp. Med. 135, 1228–1246. Anonymous (1992). Biologic Markers in Immunotoxicology, pp. 1–206. National Academic Press, Washington, DC. Bloksma, N., Kubicka-Muranyi, M., Schuppe, H.-C., Gleichmann, E., and Gleichmann, H. (1995). Predictive immunotoxicological test systems: Suitability of the popliteal lymph node assay in mice and rats. Crit. Rev. Toxicol. 25, 369–396. Brouland, J. P., Verdier, F., Patriarca, C., Vial, T., and Descotes, J. (1994). Morphology of popliteal lymph node responses in Brown-Norway rats. J. Toxicol. Environ. Health 41, 95–108. Cabral, A. R., Alcocer, V. J., Orozco, T. R., Reyes, E., Fernandez, D. L., and Alarcon, S. D. (1994). Clinical, histopathological, immunological and fibroblast studies in 30 patients with subcutaneous injections of modelants including silicone and mineral oils. Rev. Invest. Clin. 46, 257–266. Clark, E. A., and Ledbetter, J. A. (1994). How B and T cells talk to each other. Nature 367, 425–428. Emery, P., and Panayi, G. S. (1989). Autoimmune reactions to D-penicillamine. In Autoimmunity and Toxicology. Immune Disregulations Induced by Drugs and Chemicals (M. E. Kammu¨ller, N. Bloksma, and W. Seinen, Eds.) pp. 167–182. Elsevier Science, Amsterdam. Gleichmann, H. (1981). Studies on the mechanism of drug sensitization: Tcell-dependent popliteal lymph node reaction to dipenylhydantoin. Clin. Immunol. Immunopathol. 18, 203–211. Griem, P., Shaw, C. F., III, and Gleichmann, E. Chemically-induced allergy and autoimmunity: What do T cells react against? In Comprehensive Toxicology. Vol 5. Immune System Toxicology (Sipes, Eds.). Pergamon Press, Oxford, in press. Janeway, C. A., and Travers, P. (1994). Immunobiology. The Immune System in Health and Disease. Current Biology Limited/Blackwell Scientific/ Garland Publishing, Oxford.
toxas
AP: Tox
CHEMICAL IMMUNOMODULATION ASSESSED WITH REPORTER ANTIGENS Kammu¨ller, M., Penninks, A. H., Bakker, J. M. D., Thomas, C., Bloksma, N., and Seinen, W. (1987). An experimental approach to chemically induced systemic (auto) immune alterations: The Spanish toxic oil syndrome as an example. In Mechanisms of Cell Injury: Implications for Human Health. Dahlem Konferenzen (B. A. Fowler, Ed.), pp. 175–192. John Wiley & Sons, Chichester. Kammu¨ller, M. E., Thomas, C., Bakker, J. M. D., Bloksma, N., and Seinen, W. (1989). The popliteal lymph node assay in mice to screen for the immune disregulating potential of chemicals—A preliminary study. Int. J. Immunopharmacol. 11, 293–300. Kantwerk-Funke, G., Burkart, V., and Kolb, H. (1991). Low dose streptozotocin causes stimulation of the immune system and of anti-islet cytotoxicity in mice. Clin. Exp. Immunol. 86, 266–270. Kilburn, K. H., and Warshaw, R. H. (1994). Chemical-induced autoimmunity. In Immunotoxicology and Immunopharmacology (J. H. Dean, M. I. Luster, A. E. Munson, and I. Kimber, Eds.), pp. 523–538. Raven Press, New York. Kubicka-Muranyi, M., Goebels, R., Goebel, C., Uetrecht, J., and Gleichmann, E. (1993). T lymphocytes ignore procainamide, but respond to its
AID
TOX 8078
/
6h16$$$281
02-12-97 20:20:43
109
reactive metabolites in peritoneal cells: Demonstration by the adoptive transfer popliteal lymphnode assay. Toxicol. Appl. Pharmacol. 122, 88– 94. Mond, J. J., Lees, A., and Snapper, C. M. (1995). T cell-independent antigens type 2. Annu. Rev. Immunol. 13, 655–692. Pike, B. L., Alderson, M. R., and Nossal, G. J. V. (1987). T-independent activation of single B cells: An orderly analysis of overlapping stages in the activation pathway. Immunol. Rev. 99, 120–152. Schielen, P., van Rodijnen, W., Tekstra, J., Albers, R., and Seinen, W. (1995). Quantification of natural antibody producing B cells in rats by an improved elispot technique using the polyvylidene difluoride membrane as the solid support. J. Immunol. Methods 188, 33–41. Schrallhammer-Benkler, K., Ring, J., Przybilla, B., Meurer, M., and Landthaler, M. (1992). Acute mercury intoxication with lichenoid drug eruption followed by mercury contact allergy and development of antinuclear antibodies. Acta Dermatol. Venereol. 72, 294–296. Uetrecht, J. P. (1992). The role of leukocyte-generated reactive metabolites in the pathogenesis of idiosyncratic drug reactions. Drug. Metab. Rev. 24, 299–366.
toxas
AP: Tox
143, 102–109 (1997)
TO968078
The Use of Reporter Antigens in the Popliteal Lymph Node Assay to Assess Immunomodulation by Chemicals1 RUUD ALBERS, ANGELIQUE BROEDERS, ANITA
VAN DER
PIJL, WILLEM SEINEN,
AND
RAYMOND PIETERS
Section of Immunotoxicology, Research Institute of Toxicology, Utrecht University, 3508 TD Utrecht, The Netherlands Received May 30, 1996; accepted November 27, 1996
The Use of Reporter Antigens in the Popliteal Lymph Node Assay To Assess Immunomodulation by Chemicals. ALBERS, R., BROEDERS, A., VAN DER PIJL, A., SEINEN, W., AND PIETERS, R. (1997). Toxicol. Appl. Pharmacol. 143, 102–109. Various drugs and other chemicals can induce T-cell-dependent B-cell activation which may lead to allergic or autoimmune-like diseases. Because the nature of the relevant (neo-) antigens is generally not known and probably depends on the chemical, we have explored the potential use of reporter antigens to determine T-cell-dependent B-cell activation by chemicals. TNP-Ficoll and TNP-OVA were used for this purpose because they are recognized by the same TNP-specific B cells, but these cells require distinct costimulation for specific antibody production. It was found that HgCl2 , phenytoin, nitrofurantoin, and D-penicillamine stimulated IgG1 production to both antigens, incomplete Freund’s adjuvant, silica, and dimethylsulfoxide to TNP-OVA only, and LPS and hydroxyl-amino procainamide to TNP-Ficoll alone. The diabetogene streptozotocin did not enhance IgG1 production, but may enhance a cellular response instead. Tolerogens and a T-cell antigen without intrinsic adjuvant activity did not influence the responses. The IgG1 production to TNP-Ficoll was local and transient, and did not always require T cells. In contrast, responses to TNP-OVA could be measured in serum, led to specific memory, and were strictly T-cell dependent. These results demonstrate that specific antibody production to reporter antigens indicates immunostimulatory effects of chemicals more sensitive than PLN cell count and provides important mechanistic information. Moreover, with TNP-OVA as reporter antigen the kinetics and regulation of chemically enhanced immune responses can be studied without the need to know the relevant neo-antigens for each individual compound. q 1997 Academic Press
It is well known that exposure to drugs and other lowmolecular-weight compounds (LMWC) can stimulate the immune system and may lead to allergic or autoimmunelike diseases in susceptible individuals (Anonymous, 1992; Kilburn and Warshaw, 1994). Activation of Th cells by 1 This work was supported by the Life Science Foundation, Dutch Organization for Scientific Research, Project SLW 811-423-431.
0041-008X/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
TOX 8078
/
6h16$$$281
chemically altered or induced neo-epitopes is considered pivotal, but the nature of these epitopes remains largely unknown and probably depends on the chemical and on the MHC haplotype involved (Griem et al., in press). Nevertheless, recognition of T-cell activation by LMWC is crucial to elucidate the underlying mechanism and to successfully predict immunostimulatory potential of chemicals. The popliteal lymph node assay (PLNA) in rodents is one of few models that can be used for this purpose. In this assay, PLN responses are assessed following footpad injection of chemicals without adjuvant. In general, only chemicals known to induce adverse immune effects in man, induce positive PLN reactions in mice as judged by an increase in PLN weight or cell count (Bloksma et al., 1995; Gleichmann et al., 1990; Kammu¨ller et al., 1989). However, these nonspecific parameters are rather insensitive and do not unequivocally prove T-cell involvement, because chemicals causing irritation at the injection site may induce nonspecific PLN reactions (Bloksma et al., 1995; Brouland et al., 1994). Moreover, as the specificity of these responses is unknown, their kinetics and regulation cannot easily be monitored and mechanistic studies are difficult to perform. We have recently found that the prototype autoimmunogenic chemicals HgCl2 and diphenylhydantoin [phenytoin, DPH] strongly stimulate specific B-cell responses to coinjected reporter antigens in the PLNA (Albers et al., 1996a). TNP-Ficoll and TNP-ovalbumin (TNP-OVA) were used for this purpose because they are both recognized by TNP-specific B cells, but these cells require distinct T-cell help for specific antibody production. The T-cell-independent type 2 antigen TNP-Ficoll can independently stimulate IgM production, whereas cytokines are required for isotype switch to IgG1 (Mond et al., 1995; Pike et al., 1987). TNP-OVA, on the other hand, is a T-cell-dependent antigen and B cells forming antibodies to this antigen require help in the form of T–B cell interaction and cytokines (Clark and Ledbetter, 1994). Thus, by varying the antigen, effects of LMWC on B-cell responses with different requirements for T-cell help can be evaluated, without the need to identify the relevant neo-antigens for each chemical individually.
102
02-12-97 20:20:43
toxas
AP: Tox
CHEMICAL IMMUNOMODULATION ASSESSED WITH REPORTER ANTIGENS
The aim of this study was to determine whether stimulation of specific antibody responses to the reporter antigens TNP-Ficoll and TNP-OVA is a more sensitive indication of immunostimulation than increased PLN cell count. In addition it was assessed whether these responses provide information on the mechanisms underlying stimulation by LMWC, and allow monitoring of chemically enhanced responses over prolonged periods. MATERIALS AND METHODS Mice. Female, specific-pathogen-free BALB/c mice of 8–12 weeks were obtained from the Utrecht University breeding facility, and similar BALB/c nu// and nu/nu mice were obtained from Harlan CPB (Austerlitz, Netherlands), and B10.s from Olac Ltd. (Blackthorn, Bicester, UK). Mice were housed under hygienic barrier conditions in filter-topped macrolon cages with a bedding of wood chips at a temperature of 23 { 27C and 50– 60% relative humidity, and a 12-hr light/dark cycle. They received standard lab chow and acidified tap water ad libitum. Chemicals and reagents. HgCl2 was obtained from BDH (Poole, UK), LPS (lipopolysaccharide W, Escherichia coli O111:B4) and IFA (see Table 1 for abbreviations) from Difco (Detroit, MI), PB from Brocasef bv. (Maarssen, The Netherlands), NF from ACROS Chimica (’s Hertogenbosch, The Netherlands), Immobilon-P membranes from Millipore (Etten-Leur, The Netherlands), and alkaline phosphatase-conjugated goat anti-mouse IgG1 and IgM antibodies from Southern Biotechnology Associates Inc. (Birmingham, AL). The antigens TNP-OVA and TNP-Ficoll were prepared as described previously (Albers et al., 1996a). AT was synthesized according to (Alkan et al., 1972), purified by chromatography, and was 51 recrystallized before use. Innolia-kits were obtained from Innogenetics (Zwijndrecht, Belgium). Other chemicals were obtained from Sigma Chemical Company (St. Louis, MO). PLN assays. PLN assays were performed as described (Albers et al., 1996a). In brief, LMWC were coinjected with 10 mg TNP-Ficoll or 10 mg TNP-OVA sc into the right hindpaw (heel r toe direction) of naive mice. LMWC were injected in quantities that were previously demonstrated stimulatory in the PLNA, or were equimolar to a related compound. Thus, 50 mg HgCl2 , 5.2 mmol PA, 5.2 mmol (1.5 mg) HAPA, 1 mg AT, 1 mg SI, 10 mg LPS, 1 mg PEN or 7.3 mmol PB in saline, 7.3 mmol (2 mg) DPH in (1:1) saline/carbonate buffer (0.025 M Na2CO3 and 0.12 M NaCl; pH 11), 1 mg STZ in citrate buffer (pH 6), or 1 mg NF in DMSO were mixed with the antigens and were injected sc into the right hindpaw. Alternatively, antigens in saline were emulsified (1:1) in IFA and a total volume of 50 ml containing 10 mg antigen was injected. Seven days later, draining PLN were isolated. For memory responses, mixtures were similarly injected into the left footpad of B10.s mice, which are susceptible to glomerulonephritis induced by HgCl2 . Serum samples were obtained weekly by orbita puncture under ether anesthesia, and after 4–5 weeks mice were challenged in the right paw with 10 mg antigen in saline. Six days after challenge, serum was collected and PLN and spleen were isolated. In all experiments, cells were isolated in PBS/1% BSA, washed once by centrifugation, resuspended in PBS/1% BSA, counted using a Coulter Counter (Coulter Electronics, Luton, UK), and then adjusted to 106 cells/ml. Viability was checked occasionally by trypan blue exclusion and always exceeded 95%. ELISPOT assay. These assays were essentially performed as described (Schielen et al., 1995). In brief, graded numbers of PLN cells in 500 ml PBS/1% BSA were incubated (377C, 4 hr) in wells with TNP-BSA-coated (10 mg/ml, 47C, 18 hr) Immobilon-P membranes as bottom. Thereafter, membranes were washed and incubated with predetermined optimal dilutions of alkaline phosphatase-conjugated anti-mouse IgG1 or IgM in PBS/
AID
TOX 8078
/
6h16$$$281
02-12-97 20:20:43
103
1% BSA at 47C for 18 hr. After washing, spots were developed by incubation with a substrate of 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium. Specific antibody-forming cells (AFC) per 106 cells were then calculated from spot numbers counted with the aid of a stereo microscope. Line immunoassay. TNP-specific antibodies in serum were determined using a commercially available line immunoassay, composed of strips coated with bands of isotype-specific anti-mouse Ig. This strip assay is developed to determine the isotype of mouse Ig in culture supernatants, but can also be used to simultaneously detect specific Ig of different isotypes. To this end, strips were incubated with 20-fold-diluted serum samples (16 hr, 47C), washed, incubated with TNP-conjugated alkaline phosphatase (1 hr, room temperature), and then developed according to the manufacturer’s instructions. Staining intensity of individual bands was determined using a densitometer, yielding a semiquantitative measure of TNP-specific Ig of different isotypes. Statistics. Preceding statistical analysis, the number of AFC was transformed to log 10 values to homogenize variance. Differences between group means were analyzed for each antigen separately, using one-way ANOVA with LSD post hoc test for contrasts. Pearson correlations were calculated with data of all animals, and the coefficients mentioned are all highly significant (p õ 0.001). Correlation lines were determined mathematically by minimizing the variance, using (Y 0 Yestimated)2/(1 / b2) as loss function and y Å a / bx as model in the nonlinear module of Systat (Version 5.2 for the Macintosh).
RESULTS
Effect of Different Chemicals in T-Cell-Competent Mice To determine the effect of LMWC on antibody responses to defined reporter antigens, selected chemicals (Table 1) were coinjected with TNP-Ficoll and TNP-OVA into the hindpaw of BALB/c /// mice. After 7 days, PLN cell count and TNP-specific IgM and IgG1 AFC were determined. Injection of the antigens in saline, carbonate buffer, or citrate buffer had no effect on these parameters (data not shown). However, footpad injection of most (auto)immunogenic chemicals increased PLN cell count and enhanced production of specific IgG1 to the coinjected reporter antigens, whereas neither tolerogens nor the T-cell antigen AT had these effects (Table 2). The Data in Table 2 were used to construct scatter plots that indicate correlations between the different parameters. Figure 1 clearly illustrates that PLN cellularity correlated with IgG1 responses to TNP-Ficoll, except following stimulation by the macrophage activators IFA and SI (R 2 Å 0.43; 0.72 with IFA and SI excluded). IgM and IgG1 AFC responses to this antigen were not very closely related (R 2 Å 0.47, Fig. 2). However, when TNP-OVA was used as reporter antigen, specific IgG1 AFC increased exponentially with cell count for all chemicals (R 2 Å 0.79). The maximum increase in PLN cell count was É10-fold (Table 2), whereas IgG1 production to TNP-OVA increased up to É400-fold per 106 cells, yielding a 4000-fold increase per PLN. IgM and IgG1 production to TNP-OVA increased in concert, unless LPS was used as the stimulator. This B-cell mitogen enhanced specific IgM production, but failed to induce
toxas
AP: Tox
104
ALBERS ET AL.
TABLE 1 Abbreviations and Characteristics of Compounds Used in This Study Abbreviation
Chemical
Characteristics
AT
ABA-Tyr
DMSO
Dimethyl sulfoxide
DPH HAPA
Diphenylhydantoin Hydroxyl-amino procainamide
HgCl2 IFA LPS
HgCl2 Incomplete Freund’s adjuvant Lipopoly saccharide
NF PA PB
Nitrofurantoin Procainamide Phenobarbital
PEN SI
D-Penicillamine Silica
STZ
Streptozotocin
Complete T-cell antigen of low molecular weight Organic solvent, occasionally used in PLN assay Autoimmunogen Autoimmunogenic metabolite of procainamide Autoimmunogen Macrophage activator B-cell antigen and B-cell mitogen Autoimmunogen Tolerogen Tolerogen, structurally related to diphenylhydantoin Autoimmunogen, sensitizer Macrophage activator, inducer of scleroderma Autoimmunogen
switch to IgG1 (R 2 Å 0.68; 0.77 without LPS). STZ behaved different from the other chemicals in that it increased PLN size and weight (not shown) but not cellularity, and enhanced IgG2a rather than IgG1 production to TNP-OVA (IgG2a data not shown). Responses in T-Cell-Competent Versus Incompetent Mice The role of T cells in stimulation by HgCl2 and DPH was determined by comparing antibody production to the reporter antigens in T-cell-competent (///, nu//) and incompetent (nu/nu) BALB/c mice. HgCl2 increased PLN cell count in /// and nu//, but not in nu/nu mice, and accordingly enhanced specific AFC responses in T-cell-competent mice only (Table 3). DPH on the other hand, increased cell count most effectively in /// and nu// mice, but to a lesser extent also in nu/nu mice, particularly when combined with TNPFicoll. Interestingly, DPH stimulated IgM and IgG1 responses to TNP-Ficoll similarly in ///, nu//, and nu/nu mice, whereas responses to TNP-OVA were only enhanced in /// but not in nu/nu mice. Induction of Memory To evaluate whether stimulation by LMWC leads to systemic responses to the reporter antigen, T-cell-competent mice received one injection of TNP-Ficoll or TNP-OVA with saline, HgCl2 , or DPH. TNP-specific IgM but not IgG1 was increased in serum of TNP-Ficoll-immunized mice, and this IgM was produced mainly in the spleen (not shown). Moreover, challenge with TNP-Ficoll did not elicit second-
AID
TOX 8078
/
6h16$$$281
References
02-12-97 20:20:43
Alkan et al. (1972) Kammu¨ller et al. (1989) Alarco´n-Segovia and Alarco´n-Riquelme (1989) Uetrecht (1992); Kubicka-Muranyi et al. (1993) Schrallhammer-Benkler et al. (1992) Janeway and Travers (1994) Janeway and Travers (1994) Alarco´n-Segovia and Alarco´n-Riquelme (1989) Kubicka-Muranyi et al. (1993) Gleichmann (1981) Emery and Panayi (1989) Cabral et al. (1994) Kantwerk-Funke et al. (1991)
ary responses (Fig. 3) and serum levels were not influenced by the stimulating LMWC. In contrast, mice immunized with TNP-OVA plus HgCl2 or DPH, but not with TNP-OVA alone, had TNP-specific IgM as well as IgG1 in their serum. Challenge with TNP-OVA largely increased these levels and induced numerous TNP-specific IgG1 AFC in PLN, spleen, and bone marrow of these mice (AFC data not shown). DISCUSSION
It has been well established that increased PLN cell count following footpad injection of LMWC in mice correlates with immunostimulating potential of such compounds in man. Here it is shown that such LMWC also stimulate specific antibody production to coinjected reporter antigens in the PLN. In particular IgG1 production to TNP-OVA is a far more sensitive indicator of immunostimulation by LMWC than cell count, and this response is not stimulated by tolerogens (PB, PA), nor by the T-cell antigen (AT) that lacks intrinsic adjuvant activity. Importantly, these responses to defined reporter antigens may also provide information on the mechanisms underlying immunostimulation by chemicals. Antibody production to TNP-Ficoll and TNP-OVA was differently affected by distinct (sets of) compounds (Fig. 4). HgCl2 , DPH, PEN, and NF were the most potent immunostimulators as they increased IgG1 production to TNP-Ficoll as well as to TNP-OVA. The polysaccharide TNP-Ficoll cannot be recognized by T cells, and LMWC most likely
toxas
AP: Tox
AID
TOX 8078
/
6h16$$8078 2 1 4 3 4 5
AP: Tox
49 169 91 71 88 102
{ { { { { {
13** 91** 12** 27** 12** 34**
25 { 7* 48 { 17** 55 { 8**
33 { 12* 35 { 6**
14 { 3 23 { 5* 6{ 4
IgM (AFC/106 cells)
4 12 7 5 6 7
2 3 4
3 3
2 õ1
SI
33 14 58 42 105 215
{ { { { { {
18** 5 5** 20** 19** 56**
10 { 3 12 { 3* 18 { 4**
9{ 2 19 { 7*
8{ 3 4{ 1 6{ 1
IgG1 (AFC/106 cells)
4 2 7 5 13 27
1 2 2
1 2
õ1 õ1
SI
6.4 3.2 12.0 11.6 15.8 13.3
{ { { { { {
1.1** 0.4 1.0** 1.3** 1.1** 2.2**
9.3 { 1.6** 7.6 { 1.3** 17.5 { 0.9**
3.5 { 0.7 5.9 { 1.3**
1.5 { 0.9 2.9 { 0.6 2.7 { 0.5
Cells (1106)
4 2 8 8 11 9
6 5 12
2 3
2 2
SI
18 14 207 397 118 39
{ { { { { {
11* 4* 36** 159** 25** 14**
58 { 19** 47 { 10** 97 { 14**
14 { 5* 83 { 16**
6{ 9 12 { 2* 21 { 11**
IgM (AFC/106 cells)
TNP-OVA
3 2 35 66 20 7
10 8 16
2 14
2 4
SI
5 0 7*
6 6 683 1104 1269 165
{ { { { { {
2 3 154** 341** 254** 65**
145 { 28** 55 { 16** 797 { 225**
5{ 3 20 { 18
3{ 1{ 16 {
IgG1 (AFC/106 cells)
2 2 228 368 423 55
48 18 266
2 7
õ1 5
SI
a TNP-Ficoll and TNP-OVA were injected alone or in combination with the indicated chemicals into the footpad of naive BALB/c mice. Seven days later, the total number of cells and the number of TNP-specific IgM and IgG1 AFC in the draining PLN were determined. Values indicated are arithmetic means { SEM of 6–9 mice (NF and PEN n Å 3) and as stimulation indices (SI) of treatment/control group. Significant differences from mice injected with antigen and saline are indicated (*p õ 0.05 and **p õ 0.01).
0.7* 0.5 1.9** 1.3* 0.8* 1.0**
{ { { { { {
toxas 4.4 2.4 9.2 5.6 9.3 9.9
HAPA (autoimmunogen) STZ (autoimmunogen) NF (autoimmunogen) PEN (autoimmunogen) DPH (autoimmunogen) HgCl2 (autoimmunogen)
2 2
2.7 { 0.1 4.5 { 0.7*
AT (T-cell antigen) LPS (B-cell mitogen) 2 3 12
1 õ1
2.1 { 0.2 2.0 { 0.4 1.3 { 0.4
Saline (control) PA (tolerogen) PB (tolerogen)
4.4 { 0.9 6.5 { 0.6** 24.9 { 2.7**
02-12-97 20:20:43
DMSO (organic solvent) SI (MPh activator) IFA (MPh activator)
SI
Cells (1106)
TNP-Ficoll
TABLE 2 Summary of PLN Cell Count and TNP-Specific IgM and IgG1 AFC, Found after Injection of Indicated Chemials with the Reporter Antigen TNP-Ficoll and TNP-OVAa
CHEMICAL IMMUNOMODULATION ASSESSED WITH REPORTER ANTIGENS
105
106
ALBERS ET AL.
FIG. 1. Scatter plot of TNP-specific IgG1 AFC versus PLN cellularity, after injection of various chemicals with the reporter antigens TNP-Ficoll and TNP-OVA. BALB/c /// mice were immunized sc into the right hindpaw with TNP-Ficoll or TNP-OVA combined with one of several chemicals. Seven days later, draining PLN were isolated and cell count and TNP-specific IgG1 AFC/106 cells were determined. The areas indicated in the plot define the mean { SEM of 6–9 mice (PEN and NF, n Å 3). The black area indicates values obtained with antigen in saline, and the other areas show results of injection of antigen with the indicated chemical. Correlations calculated with the individual data are R 2 Å 0.43 for TNP-Ficoll (dashed line), 0.72 when SI and IFA are excluded (solid line), and 0.79 for TNP-OVA (solid line).
stimulate this response by creating neo-epitopes that lead to activation of T cells. Cytokines produced by those T cells can then stimulate production of IgG1 by B cells that recognize TNP-Ficoll (Albers et al., 1996a). IgG1 production to TNP-OVA, on the other hand, requires contact help provided during cognate interaction of TNP-OVA-specific T and B cells (Fig. 4). This response can be stimulated by tissue damage or release of proinflammatory cytokines leading to local inflammation and activation of antigen presenting cells. LMWC that enhance both responses therefore have the in-
trinsic adjuvant activity that stimulates responses to nonself antigens, and in addition may create such antigens by modifying self antigens. Macrophage-activating LMWC like IFA and SI and the solvent DMSO cannot stimulate IgG1 production to TNPFicoll, but do stimulate the response to TNP-OVA. The response to TNP-OVA is probably increased because activated macrophages present TNP-OVA more efficiently, leading to activation of specific Th cells, T–B cooperation, and subsequent IgG1 production. Clearly the response to TNP-Ficoll
FIG. 2. Scatter plot of TNP-specific IgM and IgG1 AFC per 106 PLN cells. TNP-specific IgM and IgG1 AFC were determined in the same mice described in the legend to Fig. 1, and are similarly represented. Correlations are R 2 Å 0.47 for TNP-Ficoll (solid line) and 0.68 (solid line) for TNPOVA (0.77 with LPS excluded, dashed line).
AID
TOX 8078
/
6h16$$$281
02-12-97 20:20:43
toxas
AP: Tox
107
CHEMICAL IMMUNOMODULATION ASSESSED WITH REPORTER ANTIGENS
TABLE 3 PLN Cell Count and TNP-Specific IgG1 AFC, after Injection of HgCl2 and DPH with TNP-Ficoll or TNP-OVA into Normal and T-Cell-Deficient Micea Cells (1106)
Antigen
SI
TNP-specific IgG1/106 cells
SI
Strain
Stimulant
// / nu// nu/nu
None None None
None None None
// / nu// nu/nu // / nu// nu/nu
HgCl2 HgCl2 HgCl2 DPH DPH DPH
TNP-Ficoll TNP-Ficoll TNP-Ficoll TNP-Ficoll TNP-Ficoll TNP-Ficoll
10.3 11.7 1.0 11.3 8.7 6.3
{ { { { { {
2.8 3.7 0.6* 1.7 0.8 1.3*
5 7 õ1 6 5 5
152 70 2 77 85 128
{ 62 { 8 { 1** { 9 { 23 { 70
38 35 õ1 19 43 16
// / nu/nu // / nu/nu
HgCl2 HgCl2 DPH DPH
TNP-OVA TNP-OVA TNP-OVA TNP-OVA
17.3 2.1 14.8 4.5
{ { { {
2.8 0.9* 3.8 1.3
9 2 8 4
188 2 792 7
{ 144 { 0* { 83 { 6**
47 õ1 198 õ1
1.9 { 0.2 1.7 { 0.2 1.2 { 0.2
4{ 2{ 8{
1 1 3
a TNP-Ficoll and TNP-OVA were injected with HgCl2 and DPH into BALB/c ///, nu//, and nu/nu mice. Seven days later, PLN cell count and the number of TNP-specific IgG1 AFC/106 cells were determined. Compiled data from contralateral PLN are given as baseline values for each strain. Data are indicated as arithmetic means { SEM of 3–6 mice and stimulation index (SI) of treatment/baseline. Significant differences from /// mice treated with same chemical and reporter antigen are indicated (*p õ 0.05 and **p õ 0.01).
cannot be stimulated in this manner and these LMWC apparently do not create neo-antigens that activate other T cells. Selective stimulation of IgG1 production to TNP-OVA, but not to TNP-Ficoll, may therefore identify LMWC that have intrinsic adjuvant activity, but do not create neo-epitopes. Other compounds like LPS, HAPA, and in nu/nu mice also DPH, increased IgG1 production to TNP-Ficoll, but not the response to TNP-OVA. LPS is a B-cell mitogen and at the dose used directly stimulates IgM production to both antigens, and to some extent augments IgG1 production to TNP-Ficoll. It is currently unclear why HAPA also selectively stimulated IgM and IgG1 production to TNP-Ficoll, as results of others have indicated that this compound can elicit a T-cell response by itself, suggesting intrinsic adjuvant activity (Kubicka-Muranyi et al., 1993). Interestingly, DPH stimulates T-cell-dependent IgG1 production to TNP-OVA exclusively in normal mice, but increases IgG1 production to TNP-Ficoll in normal and T-cell-deficient mice alike. This effect and the previously observed low, but significant, increased PLN cell count in nu/nu mice (Gleichmann, 1981, Kammu¨ller et al., 1987) are not likely caused by the few T cells present in PLN of nu/nu mice (õ3% of T cells in PLN of /// and nu//; R. Albers, unpublished results). Instead, these results may indicate that DPH has a T-cell-independent mode of action, on top of its undisputed effect on T cells. It remains to be established whether this results from a direct LPS-like effect on B cells or whether DPH stimulates production of switch-inducing cytokines by non-T cells (i.e., NK or mast cells (Mond et al., 1995).
AID
TOX 8078
/
6h16$$$281
02-12-97 20:20:43
FIG. 3. TNP-specific antibodies in serum. B10.s mice received one injection of TNP-Ficoll or TNP-OVA with saline (m), HgCl2 (s), or DPH (l) and were challenged 4–5 weeks later in the other hindpaw with antigen alone. Serum was collected weekly and TNP-specific IgM and IgG1 were determined; 0, before immunization, 1, 2, 3, 1 week after, 2 weeks after, and 3 weeks after immunization; C, 6 days after challenge. Values are arithmetic means of five mice per group.
toxas
AP: Tox
108
ALBERS ET AL.
regulation of chemically enhanced responses to symptomatic adverse immune effects without the need to know the relevant neo- or auto-antigens for each LMWC individually. ACKNOWLEDGMENTS HAPA was kindly provided by Prof. Dr. J. Uetrecht, Toronto, Canada. The helpful discussions with Drs. M. Frens, C. de Heer, E. Urrestarazu, W. Vaes, and H. Verhaar are gratefully acknowledged.
REFERENCES
FIG. 4. Schematic representation of the effect of different LMWC on IgG1 production to the reporter antigens TNP-Ficoll and TNP-OVA. For explanation see text.
STZ was distinct from all other compounds as it increased PLN weight but not cell count and increased IgG2a rather than IgG1 production to TNP-OVA. Moreover, production of interferon-g was increased and CD8/ cells and macrophages accumulated in STZ-stimulated PLN (manuscript in preparation). These and other observations indicate that unlike the other LMWC that drive humoral type 2 immune responses, STZ stimulates cell-mediated type 1 responses. This is in line with the antibody-mediated adverse effects of the other compounds, versus the cellular mechanism underlying STZinduced autoimmune diabetes. It appears therefore that the isotype of the antibodies produced to TNP-OVA indicates the type of response enhanced by a LMWC. The distinct characteristics of TNP-Ficoll and TNP-OVA were also reflected in the systemic responses developing after chemical stimulation. With TNP-Ficoll only TNP-specific antibodies of IgM isotype were found in serum; these were produced mainly in the spleen, and were not influenced by stimulating LMWC. This indicates that LMWC only transiently stimulate local cytokine-dependent IgG1 production to TNP-Ficoll. However, chemical stimulation of responses to TNP-OVA yielded increased serum levels of TNP-specific IgG1 and induced specific T- and B-cell memory to the antigen. This indicates that it may be feasible to compare the type and magnitude of such responses in mouse strains with varying susceptibility to the adverse immune effects of the stimulating LMWC, and relate this to the observed pathology (Albers et al., 1996b, manuscripts in preparation). In summary, we have demonstrated that reporter antigens can be very instrumental in detecting the immunostimulating potential of LMWC. It is shown that reporter antigen-specific IgG1 production provides a very sensitive parameter that can also provide information on the mechanism underlying the stimulation. The isotype of the antibodies produced to TNPOVA can indicate the type of response stimulated, and antibodies to reporter antigens may eventually link induction and
AID
TOX 8078
/
6h16$$$281
02-12-97 20:20:43
Alarco´n-Segovia, D., and Alarco´n-Riquelme, M. (1989). Autoimmune reactions in humans induced by diphenylhydantoin and nitrofurantoin. In Autoimmunity and Toxicology. Immune Disregulation Induced by Drugs and Chemicals (M. E. Kammu¨ller, N. Bloksma, and W. Seinen, Eds.), pp. 151–166. Elsevier, Amsterdam/New York/Oxford. Albers, R., van der Pijl, A., Seinen, W., Pieters, R., and Bloksma, N. (1996a). The autoimmunogenic chemicals HgCl2 and diphenylhydantoin stimulate IgG production to TNP-ficoll and TNP-OVA, supporting and extending the graft-versus-host hypothesis for chemical induction of autoimmunity. Immunology 86, 468–473. Albers, R., Volckmann, R., van der Pijl, A., Bol, M., Seinen, W., and Pieters, R. (1996b). Breaking of tolerance and MHC-dependent immune modulation by autoimmunogenic chemicals. Fundam. Appl. Toxicol. 30(Suppl.), 62. [Abstract] Alkan, S. S., Williams, E. B., Nitecki, D. E., and Goodman, J. W. (1972). Antigen recognition and the immune response. Humoral and cellular immune responses to small mono- and bifunctional antigen molecules. J. Exp. Med. 135, 1228–1246. Anonymous (1992). Biologic Markers in Immunotoxicology, pp. 1–206. National Academic Press, Washington, DC. Bloksma, N., Kubicka-Muranyi, M., Schuppe, H.-C., Gleichmann, E., and Gleichmann, H. (1995). Predictive immunotoxicological test systems: Suitability of the popliteal lymph node assay in mice and rats. Crit. Rev. Toxicol. 25, 369–396. Brouland, J. P., Verdier, F., Patriarca, C., Vial, T., and Descotes, J. (1994). Morphology of popliteal lymph node responses in Brown-Norway rats. J. Toxicol. Environ. Health 41, 95–108. Cabral, A. R., Alcocer, V. J., Orozco, T. R., Reyes, E., Fernandez, D. L., and Alarcon, S. D. (1994). Clinical, histopathological, immunological and fibroblast studies in 30 patients with subcutaneous injections of modelants including silicone and mineral oils. Rev. Invest. Clin. 46, 257–266. Clark, E. A., and Ledbetter, J. A. (1994). How B and T cells talk to each other. Nature 367, 425–428. Emery, P., and Panayi, G. S. (1989). Autoimmune reactions to D-penicillamine. In Autoimmunity and Toxicology. Immune Disregulations Induced by Drugs and Chemicals (M. E. Kammu¨ller, N. Bloksma, and W. Seinen, Eds.) pp. 167–182. Elsevier Science, Amsterdam. Gleichmann, H. (1981). Studies on the mechanism of drug sensitization: Tcell-dependent popliteal lymph node reaction to dipenylhydantoin. Clin. Immunol. Immunopathol. 18, 203–211. Griem, P., Shaw, C. F., III, and Gleichmann, E. Chemically-induced allergy and autoimmunity: What do T cells react against? In Comprehensive Toxicology. Vol 5. Immune System Toxicology (Sipes, Eds.). Pergamon Press, Oxford, in press. Janeway, C. A., and Travers, P. (1994). Immunobiology. The Immune System in Health and Disease. Current Biology Limited/Blackwell Scientific/ Garland Publishing, Oxford.
toxas
AP: Tox
CHEMICAL IMMUNOMODULATION ASSESSED WITH REPORTER ANTIGENS Kammu¨ller, M., Penninks, A. H., Bakker, J. M. D., Thomas, C., Bloksma, N., and Seinen, W. (1987). An experimental approach to chemically induced systemic (auto) immune alterations: The Spanish toxic oil syndrome as an example. In Mechanisms of Cell Injury: Implications for Human Health. Dahlem Konferenzen (B. A. Fowler, Ed.), pp. 175–192. John Wiley & Sons, Chichester. Kammu¨ller, M. E., Thomas, C., Bakker, J. M. D., Bloksma, N., and Seinen, W. (1989). The popliteal lymph node assay in mice to screen for the immune disregulating potential of chemicals—A preliminary study. Int. J. Immunopharmacol. 11, 293–300. Kantwerk-Funke, G., Burkart, V., and Kolb, H. (1991). Low dose streptozotocin causes stimulation of the immune system and of anti-islet cytotoxicity in mice. Clin. Exp. Immunol. 86, 266–270. Kilburn, K. H., and Warshaw, R. H. (1994). Chemical-induced autoimmunity. In Immunotoxicology and Immunopharmacology (J. H. Dean, M. I. Luster, A. E. Munson, and I. Kimber, Eds.), pp. 523–538. Raven Press, New York. Kubicka-Muranyi, M., Goebels, R., Goebel, C., Uetrecht, J., and Gleichmann, E. (1993). T lymphocytes ignore procainamide, but respond to its
AID
TOX 8078
/
6h16$$$281
02-12-97 20:20:43
109
reactive metabolites in peritoneal cells: Demonstration by the adoptive transfer popliteal lymphnode assay. Toxicol. Appl. Pharmacol. 122, 88– 94. Mond, J. J., Lees, A., and Snapper, C. M. (1995). T cell-independent antigens type 2. Annu. Rev. Immunol. 13, 655–692. Pike, B. L., Alderson, M. R., and Nossal, G. J. V. (1987). T-independent activation of single B cells: An orderly analysis of overlapping stages in the activation pathway. Immunol. Rev. 99, 120–152. Schielen, P., van Rodijnen, W., Tekstra, J., Albers, R., and Seinen, W. (1995). Quantification of natural antibody producing B cells in rats by an improved elispot technique using the polyvylidene difluoride membrane as the solid support. J. Immunol. Methods 188, 33–41. Schrallhammer-Benkler, K., Ring, J., Przybilla, B., Meurer, M., and Landthaler, M. (1992). Acute mercury intoxication with lichenoid drug eruption followed by mercury contact allergy and development of antinuclear antibodies. Acta Dermatol. Venereol. 72, 294–296. Uetrecht, J. P. (1992). The role of leukocyte-generated reactive metabolites in the pathogenesis of idiosyncratic drug reactions. Drug. Metab. Rev. 24, 299–366.
toxas
AP: Tox