| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Divisions of
*
Immunology and Allergology and
Analytical Chemistry, Medical Institute of Environmental Hygiene, and
Dermatology Clinic, Heinrich Heine University, Düsseldorf, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The situation in humans is comparable with that seen in mouse models. Whereas administration of nickel ions alone failed to sensitize naive mice, sensitization was achieved by the combined administration of NiSO4 and CFA (8), NiCl2 plus irritant, or NiCl2 and H2O2 (9); the latter is produced at high levels by inflammatory phagocytes and can act as an endogenous adjuvant (10). In mice already sensitized to nickel, however, nickel ions alone sufficed to elicit hypersensitivity reactions (8, 9, 11). Likewise, when T cells from nickel-sensitized mice (9) or men (12, 13, 14, 15, 16) were exposed to nickel ions on APCs in vitro, this sufficed to restimulate them. The collective evidence prompted us to conclude (9) that the combined administration of nickel ions and H2O2 induces both signal 1 and signal 2 required for T cell priming and induction of nickel allergy, whereas nickel ions alone can generate the neoantigens recognized by nickel-specific T cells and thus provide signal 1 but are unable to induce the costimulation, or signal 2.
Two groups of investigators have shown that tolerance to nickel can be induced in naive animals by several weeks of oral exposure to nickel ions, thus preventing subsequent sensitization of the animals (8, 11, 17). In orally tolerized guinea pigs, nickel tolerance lasted up to 24 mo (17); in mice, the maximal duration of the tolerant state has not yet been determined. The mechanism underlying tolerance to nickel has not been elucidated. Based on findings obtained with nickel-specific human T cell clones, new concepts implicate IL-10, immature dendritic cells, and chemokines in nickel tolerance (15, 18, 19); however, it is unclear at present how far these findings may be generalized, because only a relatively small number of T cell clones was analyzed in detail and it is not known to what extent these clones were representative. Both nickel-allergic and nonallergic persons harbor a variety of different subsets of nickel-reactive T cells that are partially identical in both groups (15, 16, 20).
The aim of the present study was first to define the optimal conditions for tolerance induction to nickel in a mouse model (9). Applying these conditions, we determined the minimal number of T cells capable of adoptively transferring tolerance and studied their capacity to proliferate and produce IL-2 after in vivo sensitization with NiCl2 in H2O2 and in vitro restimulation with NiCl2. With respect to a possible therapeutic usage of orally applied nickel ions in men, we determined the duration of tolerance induced in naive mice and of desensitization of sensitized mice, respectively, and the concentration of nickel ions in different organs after administration of NiCl2.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
NiCl2·6H2O (in the following denoted as NiCl2), and 2,4-dinitrofluorobenzene (DNFB)4 were purchased from Sigma-Aldrich Chemie (Steinheim, Germany), 2,4-dinitrobenzenesulfonic acid (DNBS) was bought from ICN Biomedicals Inc. (Aurora, OH), and H2O2 was from E. Merck (Darmstadt, Germany). Recombinant mouse IL-2 and streptavidin-FITC were obtained from BD PharMingen (Heidelberg, Germany).
Antibodies
PE- or FITC-labeled anti-CD4 (clone RM4-5) and
anti-CD8
.2 (clone 53-5.8) mAbs, biotin-labeled anti-CD8.2
(clone 53-5.8), APC-labeled anti-CD3
(clone 145-2C11),
FITC-labeled anti-CD19 (clone 1D3), FITC-labeled
anti-I-Ab (clone AF6-120.1), biotin-labeled
anti-CD11b (clone M1/70), biotin-labeled anti-CD11c (clone
HL3), purified anti-mouse IL-2 (capture Ab: clone JES6-1A12), and
biotin-labeled anti-mouse IL-2 (detecting Ab: JES6-5H4) were
purchased from BD PharMingen (Heidelberg, Germany). CD4 Microbeads and
Streptavidin Microbeads were from Miltenyi Biotec (Bergisch Gladbach,
Germany).
Mice
Specific pathogen-free female C57BL/6J (H-2b) mice obtained from Janvier (Le Genest St. Isle, France) were used throughout. Animals were 7–10 wk of age at the onset of experiments. They had free access to drinking water (tap water) and standard rodent laboratory chow (no. 1324; Altromin, Lage/Lippe, Germany). Other than in the study by van Hoogstraten et al. (8), no measures were taken to protect the animals from exposure to nickel; cages (made from plastic) were covered by stainless steel covers, and drinking water was provided in glass bottles covered with stainless steel water outlets.
Sensitization of mice
Mice were sensitized, as described previously (9). In the case of nickel, mice were injected intradermally in both flanks (50 µl each) with either 10 mM NiCl2 in sterile, pyrogen-free saline (negative control) or 10 mM NiCl2 in saline containing 1% H2O2. In the case of DNFB, mice were primed by painting 0.5% (w/v) DNFB on the shaved flanks (25 µl each); DNFB was resolved in a 4:1 (v/v) mixture of acetone and olive oil.
Challenge for recall and ear swelling test
Ten days after priming, mice were challenged for recall by injecting 50 µl 10 mM NiCl2 in sterile, pyrogen-free saline into the pinna of each ear or by applying 50 µl 0.2% DNFB onto each ear. Forty-eight hours after challenge with NiCl2 and 24 h after challenge with DNFB, respectively, delayed-type hypersensitivity reactions were determined by measuring the increment in ear thickness compared with prechallenge value. For determination of prechallenge values, mice were anesthetized with ether; for measurement after challenge, the mice were killed by asphyxiation with CO2. Measurements were performed using a micrometer (Oditest D 1000 gauge; Dyer, Lancaster, PA) and in a blinded manner. Data shown represent the mean ear-swelling response of groups comprising five to six mice, expressed in units of millimeters x 10-2 + SEM.
Tolerance induction
For oral tolerization, mice were treated with
NiCl2 in the drinking water at the concentrations
and for the periods of time specified in Table I
. Control mice
received tap water not enriched with nickel ions. For i.p.
tolerization, mice received three weekly i.p. injections (50 µl each)
sterile, pyrogen-free saline containing NiCl2 at
the concentrations and for the periods of time specified in Table I.
Control mice received saline only. Based on the data shown in Table I,
standard treatment protocols for oral and i.p. tolerization were
selected, as described under Results.
|
Different NiCl2 concentrations, duration
of treatment, and routes of administration effective in desensitizing
sensitized mice were investigated, as specified in Table II. Mice
were first sensitized with NiCl2 in
H2O2 and, starting 5 wk
later, subjected to the desensitization treatment with oral or i.p.
administration of NiCl2 indicated. After a
treatment-free interval of 1 wk, the mice were challenged for recall
with NiCl2. Two days later, the increase in ear
thickness was determined. In one experiment, mice were first sensitized
with 10 mM NiCl2 in saline containing 1%
H2O2, left untreated for 5
wk, and then treated with 10 mM NiCl2 in the
drinking water for 4 wk, followed by a 20-wk oral maintenance treatment
with 0.1 and 1 mM NiCl2, respectively. After a
treatment-free interval of 1 wk, mice were challenged and their
ear-swelling responses were measured.
|
Single-cell suspensions of erythrocyte-free spleen cells were prepared in RPMI 1640 containing 10% FCS and passed through nylon wool columns once or twice, until the T cell fraction contained 80–85% T cells. In one experiment, T cells enriched by nylon wool were divided into CD4+ and CD8+ cells, respectively, by depleting the opposite subset using magnetic cell sorting (MACS; Miltenyi Biotec), as described (21). The enriched CD4+ and CD8+ T cell fractions were contaminated with <3% CD8+ and <2% CD4+ T cells, respectively. In two other experiments, T cells were further purified by depletion of CD11b+, CD11c+, CD19+, and MHCII+ cells using the sorting unit of the FACSCalibur (BD Biosciences, San Jose, CA). The sorted T cell fractions were contaminated with <0.5% of CD11b+, CD11c+, CD19+, and MHCII+ cells, respectively. Furthermore, in one experiment, enriched splenic T cells (104/ml) from naive donors were incubated in vitro with 100 µM NiCl2 for 30 min at 37°C, washed with sterile, pyrogen-free PBS, and then used for adoptive transfer. In yet another experiment, enriched splenic T cells (102/150 µl) from orally tolerized donors were killed before transfer by three cycles of freezing (-196°C) and thawing (37°C). Finally, in one experiment pooled popliteal, inguinal, and auricular lymph node cells (LNC), instead of enriched or purified T cells, were transferred.
Adoptive transfer
After nylon wool enrichment or MACS sorting, cell suspensions were diluted 1/10 with trypan blue, whereas with FACS sorting dilution was only 1/2. Then either 0.4 or 0.8 µl cell suspensions were counted in a Neubauer counting chamber. T cell fractions to be transferred were serially diluted to the desired concentration in sterile, pyrogen-free PBS with a maximal dilution factor of 10. Cell suspensions (150 µl), containing the type and the number of cells indicated, were injected i.v. into the tail vein of recipient mice. One day later, mice were sensitized intradermally, as described above. Ten days thereafter, they were challenged for recall at the ears, and 48 h later (24 h in the case of DNFB), their ear-swelling response was measured.
In vitro restimulation of LNC
Groups of tolerant and naive mice were injected s.c. into both hindfoot pads (50 µl each) with saline containing 100 µM NiCl2, saline containing 100 µM NiCl2 plus 1% H2O2, or saline alone. For priming to DNFB, mice were painted at both hindfoot pads with 25 µl 0.5% DNFB. Ten days later, mice were sacrificed, draining popliteal and inguinal lymph nodes from each group (2 mice/group) were isolated and pooled, and single-cell suspensions prepared. The cells were pipetted into 96-well round-bottom plates (105 cells/well) in triplicates or quadruplicates and cultured (37°C, 6.5% CO2) either in complete medium alone (200 µl/well) or in complete medium containing 75 µM NiCl2 or 100 µM DNBS. Complete medium contained RPMI 1640 supplemented with 10% FCS, 2 mM glutamine, 1 mM sodium pyruvate, essential and nonessential amino acids, 10 U/ml penicillin/streptomycin, and 5 nl/ml 2-ME. After 3 days, cultures were pulsed with 0.5 µCi/well [3H]thymidine for 16 h. Cells were then harvested onto Ready Filters with Xtalscint (Beckman Coulter, Fullerton, CA) using a PHD Cell Harvester (Cambridge Technology, Cambridge, MA), filters were dried, and [3H]thymidine incorporation was measured with a LS 6000 IC series Liquid Scintillation System (Beckman Coulter). The results are expressed as stimulation index (SI) + SD (SI = mean cpm of restimulated cells/mean cpm of cells cultured in medium only).
Sandwich ELISA for IL-2
IL-2 in the supernatants of cultured LNC (5 x 105/well) was measured using an ELISA with a detection limit of 3–5 pg/ml.
Immune flow cytometry
As control for the separation procedure, enriched T cells and
their subsets, respectively, were stained with anti-CD3,
anti-CD4, anti-CD8.2, anti-CD11b, anti-CD11c,
anti-CD19, and anti-I-Ab mAbs. Flow
cytometry analysis were performed on FACSCalibur (BD Biosciences) and
analyzed with CellQuest software.
Determination of nickel ion concentration
After the indicated treatment, the mice were bled by heart
puncture, and the indicated organs (Table III) were isolated with
Teflon-coated instruments to avoid nickel ion contamination. Nickel
analysis was performed by sector-field inductively coupled plasma mass
spectrometry in the medium resolution mode (22). Sample
preparation was done by high pressure ashing in the case of tissue
samples and by UV photolysis for blood.
|
This is depicted in Fig. 1.
|
Statistical significance of results was determined by ANOVA followed by the Newman-Keuls test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To determine optimal experimental conditions for induction of
tolerance to nickel, we compared a number of different oral and i.p.
treatment regimens. The general experimental design is shown in Fig. 1
A. As can be seen in Table I, both oral and i.p. administration of
NiCl2 to naive mice decreased the ear-swelling
response to NiCl2 in a dose-dependent manner. If
the mice were treated with 10 mM NiCl2 in the
drinking water for periods of 10, 5, and 4 wk, respectively, complete
tolerance was induced. If the animals received 2 mM
NiCl2 in the drinking water for 5 wk, partial
tolerance was induced. Administration of 10 mM
NiCl2 for 4 wk was therefore selected as the
standard treatment for oral induction of tolerance. Similarly, three
weekly i.p. injections of 10 mM NiCl2 for a
period of 4 wk resulted in complete tolerance, whereas i.p. injections
of 1 or 0.3 mM NiCl2 for 4 wk only induced
partial tolerance. The former treatment protocol was selected as the
standard treatment for i.p. induction of tolerance. All other i.p.
administration protocols studied failed to cause a statistically
significant decrease in ear-swelling response.
The background concentration of nickel ions that mice received via
their drinking water was negligible; whereas the concentration of
nickel ions in tap water was below the detection limit of 1 nM, it was
8.5 nM after release from the drinking bottles covered with stainless
steel outlets. In the case of oral administration of 10 mM nickel ions
(m.w. 58.7) for 4 wk, a total dose of 3.3 mg nickel/g body weight (bw)
was taken up via the drinking water, assuming a daily intake of 4 ml
drinking water and a bw of 20 g/mouse. Since 
27 ± 17% of
nickel ions are absorbed from the human intestine (23),
the estimated cumulative dose for oral induction of complete tolerance
in mice was 900 µg/g bw. In the case of repeated i.p. injections,
a cumulative dose of 17.6 µg nickel/g bw was needed to induce
complete tolerance.
Specificity and duration of orally induced tolerance to nickel
Based on the results shown in Table I, a 4-wk course of 10 mM
NiCl2 in the drinking water was chosen as the
standard treatment regimen for induction of oral tolerance. When mice
thus treated were sensitized 1 wk after the termination of oral
treatment and challenged 10 days later, they only showed a background
ear-swelling response to nickel, but a completely normal response to
DNFB (Fig. 2A, groups 3 and
6). Virtually identical results were obtained when the mice were
sensitized and challenged after a treatment-free interval of 20 wk
(Fig. 2B), indicating that long term tolerance had been
induced.
|
The results of experiments aiming at desensitization of mice
already sensitized to nickel are shown in Table II. As schematically depicted in Fig. 1B, mice were first sensitized and left untreated for 5 wk
before the different oral and i.p. treatment regimens were started. One
week after termination of the desensitization treatment, the mice were
challenged at the ears. At that time, treatment with 10 mM
NiCl2 in the drinking water for a period of 5 wk
had led to complete desensitization (Table II). However, when mice
thus treated were not challenged 1 wk after oral desensitization but
rested for 20 wk and then challenged with NiCl2
at the ears, they proved to be no longer desensitized (Fig. 3, group 3). Only if mice received an
additional, continuous treatment with 1 mM NiCl2
in the drinking water after the 4-wk oral treatment with 10 mM
NiCl2, their desensitization was maintained for
20 wk (Fig. 3, group 5). Partial desensitization was achieved if mice
were orally treated with 2 mM NiCl2 for 5 wk or
10 mM and 2 mM, respectively, for 2 wk (Table II). All of the i.p.
treatments studied induced only partial desensitization (Table II).
|
Adoptive transfers of enriched splenic T cells, or LNC, to naive
syngeneic recipients were performed according to the general
experimental design shown in Fig. 1C. We found that
107, 105 (data not shown),
104, 103, and even as few
as 102 bulk T cells from nickel-tolerant donors
were able to transfer tolerance (Fig. 4A, groups 4–6). In contrast,
when naive mice were used as donors, transfer of
107 (data not shown) or 104
enriched T cells failed to render the recipients resistant to
subsequent sensitization with NiCl2 in
H2O2 (Fig. 4A,
group 7). Tolerance could also be transferred by
102 FACS-sorted (instead of nylon wool-enriched)
tolerant T cells (Fig. 4B), by 104
peripheral LNC of tolerant mice (Fig. 4C, group 6), and by
104 sorted T cells of tolerized mice left
untreated for 20 wk before transfer (Fig. 5). In contrast,
101 tolerant T cells (Fig. 4A, group
3), 102 naive T cells that had been loaded with
nickel ions in vitro (Fig. 4C, group 3), or
102 killed tolerant T cells (Fig. 4C,
group 4) failed to transfer tolerance to the recipients. The results
obtained with the last two groups show that tolerance to nickel could
be transferred only by living cells, and they exclude the possibility
that the observed transfer of tolerance was merely due to
inadvertently transferred tolerogen rather than specific T cell
activity. The tolerance induced by transfer of nickel-tolerant T cells
was specific for nickel ions, because recipients of such T cells showed
a normal immune response to DNFB (Fig. 4C, group 7).
|
|
Next, we asked which T cell subset was able to adoptively transfer
tolerance to nickel. Whereas 102 enriched splenic
T cells from nontolerized donors failed to induce tolerance in the
recipient as expected, 102 cells from orally
tolerized donors again proved able to do so (Fig. 6, groups 1 and 2). In contrast, neither
102 CD4+- nor
102 CD8+-sorted T cells of
tolerized donors were sufficient to transfer the tolerance (Fig. 6, groups 4 and 6). However, when 50 CD4+ and 50
CD8+ T cells were recombined after sorting, the
mixed cells were able to transfer tolerance, just like
102 unseparated T cells (Fig. 6, group
8).
|
Tolerance to nickel could also be induced by three weekly i.p.
injections of 10 mM NiCl2 (Table I). The
tolerance thus induced was specific for nickel ions (Fig. 7A, cf. groups 4 and 7),
comparable with orally induced tolerance (Fig. 2). A difference between
the oral and i.p. tolerance is that 20 wk after termination of the
tolerization treatment orally induced tolerance was still complete
(Fig. 2B), whereas i.p. induced tolerance was only partial
(Fig. 7A, group 6). As with oral tolerance, the i.p. induced
tolerance could specifically be transferred by enriched splenic T cells
(Fig. 7B). This result was again comparable with orally
induced tolerance to nickel except that 104
enriched T cells from i.p. tolerized donors were needed for the
transfer of tolerance (Fig. 7B, group 3), whereas only
102 cells were needed in the case of orally
induced tolerance.
|
The capacity of T cells from orally tolerized donors to actively
transfer specific tolerance (
Figs. 4–7) indicated that the tolerized
cells, at least some of them, had not been deleted but were able to
exert specific suppressor function. We then asked whether the T cells
of tolerized mice were anergic in the presence of
NiCl2. To this end, we tested their capacity for
proliferation and IL-2 production after in vivo sensitization with
NiCl2 plus
H2O2 and restimulation with
nickel ions in vitro (Fig. 8). Tolerized
and nontolerized groups of mice were sensitized with either saline,
NiCl2 alone, NiCl2 in
H2O2, or DNFB. Ten days
later, LNC were restimulated in vitro with NiCl2
and DNBS, respectively. As expected, LNC of nontolerized mice
sensitized with NiCl2 in
H2O2 showed an enhanced
cell proliferation and IL-2 production (group 4 in both Fig. 8A and Fig. 8B). In contrast, LNC of tolerized
animals completely failed to do so (group 5 in both Fig. 8A
and Fig. 8B). Their anergic state was specific for
NiCl2, because they did respond to DNFB
sensitization and restimulation with DNBS (group 10 in both Fig. 8A and Fig. 8B). These results obtained in vitro
parallel those obtained in vivo, demonstrating nickel-specific
unresponsiveness (cf. groups 3 and 6 in both Fig. 2A and
Fig. 2B, and groups 2 and 7 in Fig. 4C).
When the LNC of nickel-tolerant mice were tested by adoptive
transfer, they prevented sensitization to nickel (Fig. 4C,
group 6). Taken together, these data indicate that the LNC from
tolerant donors were both anergic and suppressive.
|
For toxicological assessment of treatment risk, nickel ion
concentrations in six different organs and blood were determined after
4 wk of oral treatment with 10 mM NiCl2. Except
for liver, in the other organs and in blood, the nickel ion content of
treated mice was increased by a factor of 2–223 when compared with
that of untreated mice (Table III).
The highest increase in nickel ion content was found in the small
intestine and blood (factor of 124–223).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It is easier to induce tolerance in nonsensitized than in sensitized
individuals. Confirming this rule, oral administration of
NiSO4 induced long-lasting tolerance in naive
guinea pigs, but only a transient desensitization of guinea pigs
already sensitized to nickel (17). The present
investigation confirmed this in mice. Moreover, we found that
desensitization persisted if the mice sensitized to nickel continued to
receive 1 mM NiCl2 in the drinking water. The
concentration of 1 mM NiCl2 used for maintenance
treatment was 105–106
times higher than the background concentration of nickel ions routinely
received via the drinking water by mice housed in our animal
facility.
Continuous delivery of signal 1 in the absence of signal 2 induces T
cell tolerance and anergy (24, 25). This concept helps to
understand our observation that prolonged administration of nickel
ions, which lack intrinsic adjuvanticity (9), succeeded to
induce tolerance and anergy among the T cells reacting to this
allergen. This view is further supported by our finding that tolerance
induction to nickel was not dependent on the oral route of
administration, which is known to favor tolerance (26),
but could also be achieved by i.p. administration. Nonetheless, oral
administration of nickel ions apparently induced a more profound state
of tolerance than i.p. administration, because after a treatment-free
period of 20 wk the i.p. induced nickel tolerance remained only
partially, whereas it was still complete after oral administration (cf.
Figs. 2B and 7A). Moreover, the minimal number of
T cells capable of transferring tolerance was only
102 after oral administration, but
104 after i.p. administration (cf. Figs. 4A and 7B). In addition to the two routes of
administration used, however, there also was a difference in the
cumulative nickel ion doses received after i.p. and after oral
administration. The estimated total dose of nickel ions taken up after
oral administration (900 µg/g bw) was 50 times higher than that
received via the i.p. injections (17.6 µg/g bw). Because the level of
peripheral T cell tolerance that can be achieved is known to be dose
dependent (27, 28, 29), the more complete and persistent state
of tolerance seen after oral NiCl2 administration
might also be explained by the higher dose of tolerogen received via
this route.
Several mechanisms that may underlie oral Ag-induced tolerance have
been proposed: clonal deletion (30); anergy
(31); active suppression (32); altered
trafficking (33); and immune deviation due to a Th1-Th2
cytokine shift (34). Clonal deletion cannot be the
mechanism responsible for the oral tolerance to nickel reported here,
because the existence of nickel-specific T suppressor cells was
demonstrated by adoptive transfer of splenic T cells and LNC,
respectively, obtained from tolerized donors. When the LNC of orally
tolerized and subsequently sensitized mice were restimulated in vitro
with NiCl2, they failed to show the enhanced
proliferation and IL-2 production seen with the LNC of mice that were
not tolerized, but only sensitized before restimulation (Fig. 8).
Apparently, the nickel-specific T cells of the tolerized mice were
anergic. A congruence between a state of T cell anergy and suppressive
activity has been also observed in other tolerance models (25, 35, 36, 37). In their in vitro studies of autoreactive T cell
clones, in which anergy was induced with increasing doses of peptide,
Taams and Wauben (27, 28) observed three distinct anergic
phenotypes, ranging from simple unresponsiveness, to an unresponsive
suppressive phenotype, to an unresponsive suppressive phenotype that
was persistently present. In our in vivo model, the nickel-specific,
primed suppressor T cells transferred after a treatment-free interval
of 20 wk would seem to belong to the latter phenotype accounting for
long term tolerance.
T cell anergy toward nickel has not been reported in nonallergic, patch
test-negative humans (15, 16, 38, 39). This might be
explained by the assumption that the average level of human exposure to
tolerogenic nickel ions is considerably lower than that used in our
mouse experiments resulting in less profound T cell tolerance
(27, 28). This assumption is indirectly supported by the
different concentrations of nickel ions detected in the tissues of men
and of the mice studied by us. At the time of autopsy, the mice treated
for tolerization by exogenous NiCl2 showed nickel
ion concentrations in whole blood that were 110–230 times higher
than those found in untreated mice or humans not occupationally exposed
to nickel (40), and were still 10 times higher than
those found in workers in a nickel refinery (41). However,
humans showing transient intoxication after hemodialysis against
nickel-contaminated dialysate had a 25- to 30-fold higher level of
nickel ions in the blood (42). Apart from blood and the
small intestine, the nickel ion content in different tissues of mice
treated with NiCl2 was not considerably different
from that in naive mice, and for unknown reason, the nickel ion content
in the liver of NiCl2-treated mice was even lower
than that of untreated mice.
A different question concerns the phenotype of the cells preventing nickel hypersensitivity. In our experiments, we used T cells either enriched by nylon wool passage (purity of 80–85%) or sorted by FACSCalibur (>98% pure). Numerous reports have described T cell-mediated suppression of cellular immune responses (43, 44, 45, 46), but a uniform phenotype of the pertinent T suppressor cells has not emerged. The various T cell subsets include CD4+CD25+CTLA-4+ cells (47, 48) and CD8+ cells (32, 49). Apparently, several different T cell subsets may be involved, and conceivably this varies with both the level of tolerance achieved (27, 28) and the type of Ag studied. We found that among the 102 T cells transferred, both CD4+ and CD8+ T cells were needed for adoptive transfer of nickel tolerance. This differs from the results of van Hoogstraten et al. (8), who reported successful transfer of nickel tolerance with CD8+ cells from spleen and lymph nodes; however, they transferred 107 instead of 102 cells. In other mouse models, such as the anterior chamber-associated immune deviation model, it was shown that CD4+ T cells alone as well as CD8+ T cells alone were capable of transferring tolerance: afferent-regulatory CD4+ T cells suppressing the induction; and efferent-regulatory CD8+ T cells suppressing the elicitation of delayed-type hypersensitivity (50). Similarly, in murine EAE both CD4+ T cells alone and CD8+ T cells alone were capable of transferring the suppression induced by oral tolerance (32).
An unexpected finding was that the orally induced nickel tolerance could be adoptively transferred by as few as 102 T cells. This was not due to inadvertent carryover of nickel ions, the tolerizing agent, but required an active role of live T cells. To our knowledge, the successful transfer of tolerance with as few as 102 T cells has no precedence in the literature. In other mouse models, the number of T cells used for adoptive transfer of tolerance was several orders of magnitude higher. For instance, in the anterior chamber-associated immune deviation model, 5 x 107 spleen cells from intraocularly treated donors were used (50), in the UV light-induced tolerance to the hapten DNFB 108 unsorted or 5 x 105 CTLA-4+ spleen and regional lymph node cells were used (43), in a model of transplantation tolerance toward heart allografts 3.5 x 105 CD4+ cells from tolerized donors were used (51), and even in a model using TCR-transgenic mice orally tolerized to OVA peptide 5 x 106 splenic T cells were used (52). Although this has not been explicitly reported, for the sake of the argument we suppose that transfers of smaller cell inocula were unsuccessful in those models. What then is special about nickel ions as a tolerogen and the manner it was used here when compared with foreign proteins, cells, or tissue grafts acting as tolerogens?
Our finding that as few as 50 CD4+ plus 50
CD8+ T cells were able to transfer the tolerance
indicates that at least 2% of both subsets were able to react to
nickel ions. Such an extraordinarily high frequency of Ag-reactive,
primed T cells can hardly be explained in terms of conventional
concepts of T cell specificity, even though it is realized now that the
-TCR does not possess the exquisitely narrow specificity as
previously assumed but shows considerable cross-reactivity to different
peptides (53, 54) and
haptens5 tested.
Instead, it should be pointed out that nickel ions differ from the
above-mentioned classical Ags, or tolerogens, by the high tissue
concentrations they can achieve and by their promiscuous behavior with
regard to protein and peptide binding (55, 56). Moreover,
unlike foreign proteins, cells, and tissues, nickel ions need not be
processed to generate neoantigens, but can directly attach to
MHC-embedded self peptides on the cell surface (9, 14).
The basis of neoantigen formation by this metal appears to be its
ability to form reversible metal-protein and metal-peptide coordination
complexes (9, 55, 56, 57, 58). In doing so, nickel ions can engage
a variety of different protein side-chains as ligands, and these may
vary both within each complex and from complex to complex (55, 57). The number of ligands engaged in nickel-protein and
nickel-peptide complexes, respectively, also can vary from complex to
complex (58, 59). In all likelihood, complexes formed by
nickel ions may comprise more than one protein or peptide molecule. If
so, nickel ions might act like a bioinorganic glue sticking together
two different molecules. This could affect, for instance, appropriate
MHC-embedded self peptides and certain TCRs (56, 59). In
this context, it should be realized that the T cells rescued from
thymic death are those with TCR that possesses a certain degree of low
affinity for self peptides-MHC complexes. According to the
affinity-threshold concept of T cell activation (61), the
TCR of peripheral T cells must not be engaged by unmodified self
peptides but will be engaged by even slightly modified ones, if these
increase the binding affinity of the TCR and the overall avidity of a
given T cell for those peptides. The ability of nickel ions for complex
formation, their poorly selective chemical behavior in this process,
and their potentially high tissue concentrations should enable them to
generate a great diversity of changes, which enhance the avidity of a
large variety of different T cell clones for the self peptides
displayed in the periphery and prompt them to react.
In fact, some authors (62, 63) previously hypothesized
that nickel ions would act as mitogens, since bulk T cells from the
vast majority of human subjects tested, including those who did not
suffer from contact hypersensitivity to nickel and were patch test
negative, did proliferate in response to nickel ions (38, 39, 64). For the following reasons the mitogen hypothesis is
unlikely: 1) T cell clones reacting to nickel ions do so by using their
-TCR (58); 2) they must be able to see MHC molecules
on the APCs pulsed with nickel ions (9, 14, 65); and 3)
the majority of T cell clones in a given individual fail to react to
nickel ions (9, 14, 16, 66). Nonetheless, the original
observation remains valid that there is an unusually broad T cell
reactivity toward this versatile, complex-building metal (16, 39, 64).
A high frequency of nickel-reactive T cells cannot, of course, be the only explanation for the ability of only 102 bulk T cells to transfer the tolerant state. In addition, there must be a powerful amplification mechanism enabling so small a number of nickel-specific, primed T suppressor cells of the donor to become dominant in the recipient, inactivating the numerous nickel-specific T cells of the latter. A candidate mechanism here is infectious tolerance (67), and, indeed, there is evidence that infectious tolerance operates in our model, involving both the tolerant T cells reported here and tolerogenic APCs (K. Haarhuis et al., manuscript in preparation).
To conclude, we found that oral NiCl2 administration to naive mice induced a profound state of immunological tolerance due to the persistent suppressor activity exerted by an unusually high number of specific T cells.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 S.A. and K.H. contributed equally to this paper.
3 Address correspondence and reprint requests to Dr. Ernst Gleichmann, Division of Immunology and Allergology, Medical Institute of Environmental Hygiene, Auf’m Hennekamp 50, D-40225 Düsseldorf, Germany. E-mail address: Ernst.Gleichmann{at}uni-duesseldorf.de
4 Abbreviations used in this paper: DNFB, 2,4-dinitrofluorobenzene; DNBS, 2,4-dinitrobenzenesulfonic acid; bw, body weight; LNC, lymph node cells.
5 M. Wulferink, S. Dierkes, and E. Gleichmann. Cross-sensitization to haptens can be due to different mechanisms: formation of common metabolites, T cell recognition of cryptic peptides, and true cross-reactivity. Submitted for publication.
Received for publication July 19, 2001. Accepted for publication October 9, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
repertoire of nickel-specific T cells. Arch. Dermatol. Res. 290:397.[Medline]
0.01; ***, p 
