| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Medicine, The Louis Stokes Cleveland Department of Veterans Affairs Medical Center, and
Department of Pathology, Case Western Reserve University School of Medicine, Cleveland, OH 44106; and
Department of Neurology, University of California, San Francisco, CA 94143
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. IgG1, but not IgG2a, autoantibodies were produced. The engaged
T cells were not classic Th2 cells in that IL-4 and IL-5 were produced
by different cells. The IFA-induced violation of self tolerance,
including the deposition of specific autoantibodies in the respective
target organs, occurred in the absence of detectable pathology.
Exhaustion of the pool of naive precursor cells was shown to be one
mechanism of the IFA-induced tolerance. In addition, while the
IFA-primed T cells acted as suppressor cells, in that they adoptively
transferred disease protection, they did not interfere with the
emergence of a type 1 T cell response in the adoptive host. Both active
and passive tolerance mechanisms, therefore, contribute to
autoantigen:IFA-induced protection from autoimmune
disease. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
When the propounded suppressor cell models came under scrutiny (7), IFA-induced tolerance was reexamined. It was then shown that mice preinjected with Ag in IFA failed to generate proliferative (8, 9) and IL-2 recall responses (10) in their draining LNs when they were reimmunized with the same Ag in CFA. This lack of responsiveness seemed to occur in the absence of detectable regulatory cells in the draining LNs (8, 9). Direct visualization studies using adoptively transferred TCR transgenic cells showed that the Ag-specific cells disappeared from the draining LNs under these conditions (11). Moreover, the generation of T cell hybridomas from draining LNs of wild-type mice suggested that the residual cells lacked high affinity Ag-specific T cell clones following IFA preinjection (12). These types of data have led to the view that IFA injections induce tolerance via clonal deletion. Additionally, in favor of this interpretation is the prevalent infectious nonself/noninfectious self model of immunogenicity (13, 14), according to which the induction of a T cell response requires immunogenic Ags to be encountered in association with microbial products (13, 14). Unlike Ag injections in CFA (mycobacterial proteins in mineral oil), injections of Ag in IFA (mineral oil only) were thought to be prone to induce tolerance because such water:oil emulsions, by serving as excellent Ag depots that last for months (15, 16), provide the first signal without the bacterial products present to engage the second signal.
Despite these apparently convincing arguments for Ag:IFA injections
triggering clonal deletion, the issue remains controversial in that
there are recent publications implicating anergy (17) and
induction of an active type 2 immune response (18, 19) as
being operative under these conditions. The controversy derives in part
because it has been challenging to define the impact of Ag:IFA
injections on the specific T cells in normal (nontransgenic) mice, in
which the naturally occurring low frequencies of precursor cells have
stymied direct measurements. Recently, a new generation of cytokine
ELISPOT assays has emerged (20) that permit one to study
Ag-specific T cells directly ex vivo at single cell resolution and that
facilitate the direct visualization of Ag-induced cytokine produced by
even a single Ag-specific T cell in a million bystander cells
(20). Enabled by these assays, we set out to revisit the
effects of Ag:IFA injections. We previously showed that IFA injections
of foreign Ag are prone to induce CD4 cells that produce IL-4 and IL-5,
but not IFN-, and IgG1 and IgE, but not IgG2a Ab responses, as is
consistent with a polarized type 2 immune response (19).
These experiments also showed that the induced CD4 memory cells were
L-selectin negative, and that this lack of LN-homing receptor
expression accounted for the aforementioned deficient LN responses
(8, 9, 11); after the same Ag was reinjected, the memory
cells were not recruited to the LN, but were present in the spleen
(18). These data explained why clonal deletion was
implicated in the IFA-induced tolerance when LNs were the only organs
examined. With all foreign protein Ag, in all mouse strains studied, we
found IFA injections to induce this polarized, type 2 immunity when
spleen cells were tested (19).
T cell responses to autoantigens, however, might differ fundamentally from those generated against foreign Ags. Autoantigens frequently delete the high affinity end of the autoreactive T cell repertoire (21, 22). This also holds for MBP (23) and possibly for other neuroantigens that until recently were thought to be sequestered behind the blood brain barrier (BBB). Unlike the response to foreign Ags, therefore, the T cell response to autoantigens might comprise primarily low affinity clones (24). Because T cell affinity has been implicated in defining cytokine differentiation along the type 1/type 2 pathways (25, 26), the outcome of autoantigen:IFA injections might differ fundamentally from inoculations of foreign Ag in IFA. Along the same lines, there is emerging evidence that the autoreactive T cell repertoire is not necessarily naive (as is the case for foreign Ags), but instead can consist at least partially of T cells preactivated by exposure to endogenous Ag (27). Unlike naive cells, preactivated/memory cells are committed to a particular cytokine-secreting lineage (28, 29) and, thus, the polarizing effects of adjuvants may not easily redirect their type 1/type 2 differentiation status. It therefore remains to be determined whether autoantigen:IFA injections have the same immunizing/type 2-polarizing effects as seem to be the rule for injections of foreign Ags.
It was initially accepted that Th2 cells could have a regulatory function that suppresses Th1 cell-mediated effector functions, a hypothesis with the potential to explain the previously observed suppressor cell phenomena (30, 31, 32). This view has more recently come under significant scrutiny. It was shown, for example, that the cotransfer of autoreactive, Th2-polarized, putative regulatory T cells with Th1-polarized effector T cells did not prevent or ameliorate autoimmune pathology, even when the Th2 cells were in marked excess over the Th1 cells (33, 34, 35). To the contrary, such adoptive transfers indicated that the autoreactive Th2 cells could themselves induce inflammatory responses in the target organ, thus contributing to tissue destruction. Although this pathology was seen after injecting large numbers of Th2 cells preactivated in vitro and primarily in immune-deficient recipients, it remains controversial whether a type 2 response endogenously developing in an immunocompetent host can be similarly pathogenic (particularly as it pertains to violations of self tolerance following infections or vaccinations/active immunotherapy). Similarly, it is unclear whether and under which conditions this type 2 violation of self tolerance can protect from autoimmune diseases and how to reconcile the regulatory suppressor qualities of Th2 cells in such models (36, 37, 38).
The classic experiments on IFA-induced tolerance were done with MBP, an intracellular protein buried within the myelin sheath, where it is inaccessible to autoantibodies. Additionally, in this EAE model, the BBB may limit the access of the induced autoantibodies to the CNS in the absence of preexisting, activated T cells (39, 40). This model, therefore, is prone to underestimation of the potential hazards of autoantibodies induced along with Th2 cells. In contrast, MOG is a cell surface protein in the CNS, and the development of aggressive, demyelinating EAE in rats has been tied to the presence of anti-MOG Abs in addition to autoreactive T cells (41, 42, 43). Using MOG-induced EAE in marmoset monkeys, it was noted that the secondary injection of soluble MOG into MOG:CFA-preimmunized animals caused a Th2 shift in the T cell response, but eventually resulted in a delayed and more severe form of EAE (44). These and other experiments (45), including induction of EAE in some rat strains using neuroantigen/IFA injections (46), have raised significant concern regarding the potential pathogenicity of therapies aimed at exploiting active/regulatory immune mechanisms for treatment of autoimmune diseases. These concerns might be even more substantial when the extracellular target Ags used are not protected from autoantibodies by anatomic barriers such as the BBB.
In an effort to address these issues, we revisited three models of EAE
(mediated by injections of MBP, PLP, and MOG, respectively) and a
related model of murine interstitial nephritis, antitubular basement
membrane (
TBM) disease (47, 48, 49). Like EAE, TBM
disease is elicited by immunization with a tissue-specific autoantigen
(renal tubular Ag, RTA) in CFA, and injection of RTA with IFA has been
reported to prevent the disease (50). In this manner, we
studied autoimmunity directed toward an immune-privileged organ, the
CNS, and toward a nonimmune-privileged target organ, the kidney.
Although the BBB might make the CNS unique with respect to access by T
cells and Abs, the kidney is an organ whose anatomy makes it highly
accessible to Abs and immune cells. Moreover, while the
immune-privileged state of the CNS tends to cause entering T cells to
undergo apoptosis (51, 52, 53), this does not apply to the
kidney. Therefore, although EAE is the best studied and most commonly
used model of experimentally induced autoimmune disease, it is also a
unique autoimmune model. The inclusion of TBM disease was thus used
in an effort to understand more broadly pathogenic and protective
qualities of the different autoimmune response types.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Male and female SJL (H-2s), B10.PL
(H-2u), and C57.BL/6
(H-2b) mice, age 6–8 wk, were purchased
from The Jackson Laboratory (Bar Harbor, ME), and maintained under
specific pathogen-free conditions in the animal facilities of Case
Western Reserve University and the Louis Stokes Cleveland Department of
Veterans Affairs Medical Center (Cleveland, OH). RTA and MBP were
prepared in our laboratory, as described (47, 54). The
nonglycosylated recombinant protein corresponding to the extracellular
domain of rat MOG (aa 1–125) was expressed in the Escherichia
coli strain DH5- and purified to homogeneity, as previously
described (55). The PLP peptide (PLPp) 139–151
(H-HCLGKWLGHPDKF-OH) was synthesized and purchased from Princeton
Biomolecules (Columbus, OH). IFA was purchased from Life Technologies
(Grand Island, NY), and CFA was made by mixing Mycobacterium
tuberculosis H37RA (Difco Laboratories, Detroit, MI) at 2.5 mg/ml
into IFA. For inducing autoimmune diseases, either autoantigens or
peptides were mixed with CFA to yield emulsions at previously
established maximum pathogenic concentrations, 1 mg/ml for MBP, MOG,
and PLP(139–151), and 4 mg/ml for RTA. Of this emulsion, 50 µl was
injected once, s.c. at four sites on the flank. Pertussis toxin (200
ng; List Biological, Campbell, CA) was injected i.p. in 500 µl
saline, at the time of the CFA immunization and repeated 24 h
later. Injections of autoantigens in IFA emulsions involved the same
dose of Ags as specified above for their inoculation in CFA,
administered with a single 200 µl i.p. injection. Starting from day 5
after injections of neuroantigens, the mice were assessed daily for the
development of paralytic symptoms, and the severity of disease was
recorded according to the standard scale: grade 1, floppy tail; grade
2, hind leg weakness; grade 3, full hind leg paralysis; grade 4,
quadriplegia; grade 5, death.
Organ harvest, histologic analysis, and detection of in situ bound Ab
Kidneys and the CNS were harvested when the animals were sacrificed. Portions of each organ were fixed in 10% Formalin, and paraffin sections were made and stained with hematoxylin and eosin (portions of each harvested organ were also frozen in OCT compound for immunofluorescence studies of Ab staining; see below). The presence of infiltrating leukocytes was determined in a semiquantitative manner by a trained investigator, blinded to the experimental group. Disease scores for interstitial nephritis were defined as follows. Grade 0, normal kidney; grade 0.5, rare focal mononuclear cell infiltration; grade 1, mononuclear cell infiltration of the interstitium of <10% of the renal cortex; grade 2, mononuclear cell infiltration of the interstitium of >10% <50% of the renal cortex; grade 3, mononuclear cell infiltration of the interstitium of >50% of the renal cortex with tubular drop-out and interstitial fibrosis (47, 48). Standard histologic scoring of brain inflammation in EAE was used (23, 34). For detection of in situ bound Ab, frozen sections of kidney or brain tissue from naive or immunized mice were washed three times in PBS, blocked for 30 min with 1% BSA, stained with FITC-conjugated goat anti-mouse Ig or control Ig (1/50 dilution in PBS/0.1% BSA; PharMingen, San Diego, CA) for 30 min, and washed three times more in PBS. The dried sections were examined under an immunofluorescent microscope at 450 nm and photographed.
ELISA assays for detecting autoantigen-specific serum Ab
ELISA plates (Nunc Immunoplate; Fisher Scientific, Pittsburgh, PA) were coated overnight with 100 µl of 10 mg/ml RTA, MBP, or MOG at 4°C. The plates were washed three times with PBST, blocked for 1 h with 0.1% gelatin, both in PBS with 0.025% Tween (PBST) at room temperature, and washed again. Serial 3-fold dilutions of serum, starting at 1/500 in RPMI, were added in duplicate to the plate and placed at 4°C overnight. After washing with PBST, three times, the detection Abs were added. Affinity-purified goat anti-mouse IgG (H+L) from Southern Biotechnology Associates (Birmingham, AL) was used to detect total Ig; the isotype-specific Abs used to detect IgG1 and IgG2a were also from Southern Biotechnology Associates. p-Nitrophenylphosphate (PNPP; Research Organics, Cleveland, OH) was used for the development of the colorimetric reaction. After incubating for 4 h at room temperature, the plates were again washed three times in PBST, and streptavidin-alkaline phosphatase (1/2000 dilution; Dako, Carpinteria, CA) was added for 1 h. The plates were developed with a PNPP substrate, and the OD was read at 405 nm.
Cell purification and adoptive transfers
Single cell suspensions were prepared from spleens, and these were tested either as bulk populations or as purified cell fractions, as specified. Subpopulations of T cells were isolated using commercially available murine T cell isolation columns (R&D Systems, Minneapolis, MN), following the instructions supplied by the manufacturer. Resultant cells were washed in HBSS, counted by trypan blue exclusion, and resuspended at appropriate concentrations for use in the various assays. Purity was confirmed by FACS analysis using FITC-conjugated Abs (PharMingen). The columns routinely yielded >96% purity for CD4+ T cells. For preparation of APC, spleen cells from naive mice were incubated with mitomycin C (Boehringer Mannheim Biochemicals, Indianapolis, IN) at 50 µg/ml in PBS for 20 min at 37°C, followed by three washes in HBSS. For adoptive transfers, 80 x 106 T cells from spleens of immunized animals were injected, i.p., per recipient. Where indicated, these T cells have been preactivated by the respective Ag for 3 days before injection.
ELISPOT assays and ELISPOT image analysis
ImmunoSpot M200 plates (Cellular Technology, Cleveland, OH) were
coated overnight with the capture Abs in sterile PBS. R46A4, at 4
µg/ml (isolated and purified from hybridoma), was used for IFN-;
JES6-1A12, 3 µg/ml (PharMingen), for IL-2; 11B11, at 2 µg/ml
(isolated and purified from hybridoma), was used for IL-4; and TRFK5,
at 5 µg/ml (isolated and purified from hybridoma), was used for IL-5.
The plates were blocked for 1 h with sterile 1% BSA in PBS and
washed three times with sterile PBS. Spleen cells
(106 per well) or purified T cells (5 x
105/well, with the same number of mitomycin
C-treated syngeneic splenic APC) were plated in HL-1 medium
(BioWhittaker, Walkersville, MD) and Ag or peptides at 100 µg/ml, in
triplicate wells. In select experiments, cells and Ags were titrated.
The plates were incubated at 37°C, 7% CO2 for
24 h (IFN- and IL-2) or for 48 h (IL-4, IL-5, and
two-color assays). After washing three times with PBS, followed by
three times with PBS-0.025% Tween, detection Abs were added and plates
were incubated overnight at 4°C. XMG1.2-HRP (produced in our
laboratory) was used for IFN-, rat anti-mouse IL-4-biotin
(BVD6-24G2; PharMingen) was used for IL-4, rat anti-mouse
IL-2-biotin (JES6-5H4; PharMingen) was used for IL-2, and biotinylated
TRFK4 (PharMingen) was used for IL-5. The plates were then washed three
times in PBST. Streptavidin-HRP (Dako) was added at 1/2000 dilution in
PBST as a third reagent for IL-2, IL-4, and IL-5 for 2 h, followed
by three washes in PBS. The solution used to develop the plates
consisted of 800 µl AEC (Pierce, Rockford, IL; 1 g dissolved in
100 ml dimethyl formamide) mixed in 24 ml 0.1 M sodium acetate, pH 5,
plus 12 ml H2O2; 200 µl
was added per well.
For two-color ELISPOT assays (20), the plates were coated simultaneously with two capture Abs, and the two detection Abs were added simultaneously as well. The following coating mAbs were used for two-color IL-4 and IL-5 detection: BVD4-1D11 (2 µg/ml), TRFK5 (5 µg/ml). The combinations of detection Abs for the IL-4:IL-5 assays were: BVD4-24G2-biotin:TRFK4-HRP. HRP labeling of Abs was performed according to the standard method. The detection Ab concentrations were as follows: BVD4-24G2-biotin (2.5 µg/ml), TRFK4-HRP (2 µg/ml). For the biotinylated detection mAbs, the streptavidin-AP conjugate (Dako) was added (at 1/2000 dilution), incubated for 2 h at room temperature, and removed by washing twice with PBST and twice with PBS. The NBT/BCIP substrate (Kirkegaard & Perry Laboratories, Gaithersburg, MD) was added first, then, after washing twice with PBS, the AEC substrate was added and left for 15–30 min for NBT/BCIP and 20–40 min for AEC. The resulting spots were counted on an ImmunoSpot Series 1 Analyzer (Cellular Technology). For single-color ELISPOT, digitized images were analyzed for the presence of areas in which color density exceeds the background by a factor set on the basis of the comparison of control (containing T cells and APC without Ag) and experimental wells (containing Ag). After separating spots that touch or partially overlap, additional criteria of spot size and circularity are applied to gate out speckles and noise caused by spontaneous substrate precipitation and nonspecific Ab binding. Objects that do not meet these criteria are ignored, and areas that meet them are recognized as spots, counted, and highlighted. Two-color ELISPOT image analysis follows the same principles, except that the image analyzer detects red, blue, and double-colored spots separately by using three different color factors. When the blue is active, single-color blue and all double-positive spots (which have blue as a part of their color composition) are detected. Similarly, when red is active, only red single-color and double-color spots are detected. Each color factor is set in RGB (red, green, blue) mode and consists of three numbers reflecting the color intensity in red, blue, and green channels. The red and blue factors are set by using spots from single-color assays. The double-color factor is a mathematical intersection of the two single-color factors.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
As differences in mouse colonies and housing conditions might
influence Th1/Th2 T cell differentiation and its subsequent autoimmune
consequences, we first sought to establish the outcomes of autoantigen
injections on the phenotypic expression of autoimmune disease.
Specifically, we sought to determine whether injection with the CNS
autoantigens MBP, PLPp, and MOG or the kidney autoantigen RTA mixed in
IFA would induce clinical or histological evidence of target-organ
pathology as compared with the classic CFA injections that have been
the gold standard for inducing autoimmune disease. Table I
summarizes the results. When
susceptible mice were immunized with autoantigen mixed in CFA,
88–100% of the animals developed clinical and/or histologic
manifestations of autoimmune disease. The effects were target organ
specific in that mice injected with either MBP or PLP did not develop
detectable renal pathology, and RTA-injected mice did not show clinical
or histologic evidence of CNS pathology (not shown).
|
TBM disease: immunization in CFA
induces autoimmune disease, and injection in IFA does not. Induction of type 1 and type 2 cytokine signatures following injection of autoantigen in CFA and in IFA
The detection of low frequency cytokine-producing T cells has been
challenging, particularly for type 2 cytokines in which the addition of
receptor-blocking Abs may be required to detect them using standard
ELISAs (56, 57). To avoid loss of signal in supernatants
due to dilution or receptor uptake, we measured cytokine production at
single cell resolution with ELISPOT assays (20). As shown
in Fig. 1A, spleen cells from
RTA:IFA-injected mice produced 20–100/million RTA-specific IL-2, IL-4,
and IL-5 spots in the presence of RTA over medium control. These
RTA:IFA-injected mice did not produce any IFN- in the recall
response. The cytokine responses bore features of immunologic memory
because naive mice did not produce cytokines when challenged with RTA,
and RTA-injected mice did not respond to any of the neuroantigens (data
not shown). Overall, similar type 2 polarization was seen when MBP,
PLPp, and MOG were injected with IFA (Fig. 1, B,
C, and D). Recall responses characterized by the
production of IL-2, IL-4, and IL-5 in the absence of IFN- could be
specifically detected to the injected Ags. There were considerable
variations noted, however, in the production of the individual
cytokines, depending on which autoantigen was tested. Injections of
MOG:IFA, for example, reproducibly led to IL-4 but not IL-5 production,
while both cytokines were prevalent after injection of MBP, PLPp, and
RTA. Moreover, we found that all of the cytokines induced by IFA
immunization with these autoantigens were produced by CD4 cells (not
shown), analogous to our previously reported findings for IFA
immunizations using foreign Ags (19, 23, 47).
|
and IL-2 recall
responses in comparable frequencies to the IL-4 and IL-5 response in
IFA-injected mice (Fig. 1
, A–D), and cell separation showed
that this response was also derived from CD4 T cells (not shown). In
contrast to the IFA-injected animals, however, these mice did not
produce IL-5, although IL-4 recall responses were detectable after
immunization with some of the autoantigens. Overall, the induction of
IFN--producing memory cells was a constant feature of all CFA
immunizations, while the absence of such IFN--producing cells
characterized all injections in IFA. IL-5, in contrast, was seen with
most of the IFA injections (MOG being the only exception that we
noted), yet was completely absent after CFA immunizations. Injections
in IFA, therefore, were found to induce CD4 T cell immunity, albeit of
different cytokine signatures than immunity induced by CFA injection.
The frequencies of memory cells induced by IFA injection were similar
to those following the disease-inducing CFA immunizations. Unlike the
CFA immunizations, however, they occurred without signs of autoimmune
pathology. Type 2 cytokine signature in the absence of Th2 cells
The dissociated expression of IL-4 and IL-5 after MOG:IFA
injection, and the often discordant frequencies of memory cells
producing IL-4 and IL-5 induced by the other autoantigens prompted us
to study the coexpression of these cytokines by individual cells. Using
single cell resolution, two-color ELISPOT assays (20), we
found that IL-4 and IL-5 were, in fact, produced by different CD4 cells
and that the number of cells coexpressing both cytokines was <5% of
the total (Fig. 2). Moreover, IL-2 was
produced by the IFA-induced memory cells (which occurred in the absence
of IFN- production; Fig. 1), and this IL-2 was found to be secreted
by different CD4 memory cells than IL-5 (data not shown). These
measurements suggested that cytokine coexpression in the IFA-induced
memory cells is much more tightly regulated than predicted by the
Th0/Th1/Th2 model. Th0 cells should coexpress IL-2, IL-4, IL-5, and
IFN-, while Th2 cells should coexpress IL-4 and IL-5 without
producing IL-2. Overall, type 2, cytokine-producing, cellular immunity
was induced by autoantigen:IFA immunization at a population level, but
the Th2 model does not suffice to describe the effector cell class
induced at the level of the single cell.
|
Although autoantigen-specific IgG2a Abs were virtually absent in
autoantigen:IFA-injected mice, these animals developed high titers of
specific IgG1 Abs (Fig. 3). In contrast,
specific IgG2a was detectable at high titers in the serum of animals
injected with autoantigens in CFA (Fig. 3). As murine IgG2a Ab
production depends upon the presence of IFN--producing memory cells
to provide T cell help for Ig class switching (58), the
detection of this Ab isotype after CFA (but not IFA) injection matches
well with the detected cytokine profiles of the memory cells themselves
(Fig. 1). Similarly, the presence of IgG1 Abs after both CFA and IFA
immunization is consistent with the cytokine data, in that IgG1 Ig
class switching depends upon IL-2 and IL-4 (59, 60).
Moreover, this isotype may be less discriminatory for different immune
response types because the IgG1 locus lies upstream from the more
discriminatory IgG2a or IgE loci.
|
, kidneys harvested from animals 3 wk
after immunization with RTA:IFA showed staining of tubular basement
membranes. Similarly, the brains of MOG:IFA-immunized mice stained
positively for deposited IgG1. Notably, this autoantibody deposition
was detectable in the absence of inflammatory cellular infiltration of
the target organs (as assessed by hematoxylin and eosin) or clinical
manifestations of disease for up to 6 mo following the immunization.
Brains and kidneys from naive or control immunized mice did not bind
the anti-IgG1 (or the isotype control) Ab. The data demonstrate
that IFA injections did induce an actual humoral autoimmune response
that entails specific autoantibody deposition in target organs. Yet,
this violation of self tolerance occurred in the complete absence of
detectable pathologic consequences.
|
To test whether the previously reported IFA-induced protection
also applies to our studies, in which autoimmune responses and not
immunologic tolerance were clearly induced, we first injected mice with
autoantigen:IFA, then with the disease-inducing autoantigen:CFA
regimen. As shown in Table II, mice that
were preinjected with IFA plus saline (or control Ags) developed
organ-specific autoimmune disease of a severity corresponding to the
score of mice that were not preinjected. In contrast, mice preinjected
with RTA or MBP in IFA were largely protected from developing TBM
disease or EAE, respectively. Pretreatment of SJL mice with PLPp:IFA
reduced the severity of subsequent EAE in all animals, and fully
protected 20% of the animals compared with controls (Table II). In all
three models tested, the type 2 immunity induced with IFA injection was
therefore associated with disease protection. Although the data seem to
be consistent with the classic interpretation of Th2 cells inhibiting
Th1 cell-mediated pathology, we used the high resolution ELISPOT assay
to characterize directly the autoantigen-specific T cell repertoire in
mice with the apparent tolerant phenotype.
|
). These mice therefore retained the
type 2 component induced by the primary IFA treatment; the type 2
cytokine-producing cells were apparently not switched to type 1
producers. The mice secondarily injected with RTA:CFA, however,
additionally developed an IFN- component that was considerably
reduced in clonal size compared with a primary RTA:CFA immunization (or
with control mice primarily injected with PBS:IFA). Additionally, the
RTA:IFA-preinjected (protected) mice showed markedly decreased
frequencies of IL-2-producing cells compared with those that received
primary RTA:CFA injections. Moreover, unlike the control mice, such
RTA:IFA-preinjected mice did not engage an IgG2a autoantibody response;
the detected serum autoantibodies were of the IgG1 isotype. These data
were reproduced essentially identically in the EAE models (MBP in
B10.PL and PLPp in SJL mice; not shown). The data are consistent with
the primary autoantigen:IFA immunization exhausting the
naive/uncommitted T and B cell repertoire. As a consequence, the type
1-biasing secondary immunization with autoantigen:CFA apparently cannot
engage the full clonal sizes of the IFN--producing type 1 cells or
IgG2a Abs. These data are consistent with a passive tolerance mechanism
in which the mice cannot generate or maintain a pathogenic response
(thus contributing to their protected state). What they do not explain,
however, are the previous reports that showed that IFA-primed cells
convey active tolerance when adoptively transferred to naive
recipients.
|
We next tested whether transferring spleen cells from
autoantigen:IFA-immunized mice into naive recipients could prevent the
development of disease upon subsequent immunization with the
autoantigen in CFA. Reproducing previously published observations in
EAE and TBM disease (4, 5, 6, 50), we found that
IFA-induced immunity was in fact transferable to naive recipients
(Table III). As the adoptive recipients
have a naive T and B cell repertoire, this model allowed us to examine
how the injected regulatory/suppressor cells interfered with the
engagement of this naive repertoire by an immunization regimen that
would otherwise induce disease. The frequency of IFN- and IL-2
producers in the protected recipients of RTA:IFA-primed cells was
comparable with those induced following RTA:CFA immunization of naive
mice or mice adoptively transferred with control (MBP:IFA-primed) cells
(Fig. 6A). In contrast to the
s.c. primary immunized controls, RTA-specific IL-4 and IL-5 producers
were detectable in the recipients of RTA:IFA-primed cells. The adoptive
transfer of IFA-primed autoantigen-specific cells did not interfere
with the emergence of an IFN--producing, autoantigen-specific memory
cell population or with the induction of an IgG2a response (Fig. 6B). The induction of type 1 immunity seemed, therefore,
to be unaffected by the preexistence of type 2 memory cells specific
for the same autoantigen.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
), consistent with induction of a highly polarized type 2
response (Fig. 1). These immunizations additionally resulted in the
induction of specific IgG1, but not IgG2a Abs (Fig. 3), further
supporting the notion that an immune response, rather than immunologic
tolerance, was achieved.
While the overall Ab and cytokine profile of the IFA-induced autoimmune
responses corresponded to the classic description of type 2 immunity,
it was of interest that with MOG, IL-4 production occurred in the
absence of IL-5, while for other autoantigens both IL-4- and
IL-5-producing memory cells were induced. Moreover, the frequencies of
the individual cells producing either IL-4 or IL-5 did not exactly
match. This was true despite the fact that these assays have been
optimized so as to measure cytokines produced by individual cells, thus
providing accurate frequency measurements (20). These data
are consistent with the newly emerging view that type 1 and type 2
responses are not mediated by classic Th1 or Th2 cells that coexpress
the different type 1 or type 2 cytokines in predefined sets
(61). Instead, the data suggest that the production of
each cytokine within the type 1 or type 2 response is independently
regulated, as has been first noted with in vitro polarized
TCR-transgenic cells (62), and mitogen-stimulated cells
(63, 64, 65). The data are also consistent with the finding
that differentiation of IL-4-producing memory cells depends on the
presence of IL-4/13, but that of IL-5-producing memory cells is
independent of these cytokines (66). Also, our recent
studies of cytokine coexpression in individual in vivo differentiated T
cells showed that single cytokine expression seems to be the rule when
the T cells were tested directly ex vivo, activated by their
physiologic signal, the nominal Ag (20). In this study, we
report the same finding for the IFA-induced T cell responses to
autoantigens: IL-4 and IL-5 were produced by different T cells (Fig. 2). If our observation that regulation of IL-4/5 is under different
control mechanisms applies to other type 2 cytokines (i.e., IL-10,
IL-6, and TGF-ß), type 2 immunity may not be a uniform response type,
but one that occurs in many manifestations. Accordingly, IFA-induced
type 2 immunity might differ fundamentally from other type 2 immune
responses, e.g., those induced after immunization with Alum
(19) or after polarizing T cells in vitro. Some of the
controversy on the protective vs pathogenic role of type 2 immunity in
autoimmune disease models might result from such diversity within the
type 2 immunity studied. This notion about dissociated cytokine
expression might be important for understanding type 2 cell-mediated
effector and regulatory functions.
CFA is thought to be the strongest adjuvant with which to promote a T
cell response in mice (67). Strikingly, we found that the
injections of autoantigens in IFA induced comparable frequencies of
cytokine-producing memory T cells (Fig. 1). The magnitude of the type 2
immunity induced by IFA injections might, therefore, be as strong as
can be induced by active immunization (with or without using an
adjuvant) or during infection. Despite the strength of the induced
autoimmune responses, no pathology was detectable in the target organ,
and specifically there was no eosinophilic infiltration as might have
been predicted by the high frequency of IL-5-producing autoreactive T
cells. As opposed to the injection of large numbers of preactivated Th2
cells, therefore, the active engagement of a cellular type 2 response,
even of a most vigorous response, seems to be compatible with
health.
The pathogenicity of the autoantibodies induced in association with
type 2 immunity has also been a matter of debate. Since the initial
studies with IFA-induced tolerance were done with MBP, an intracellular
Ag, the lack of Ab-mediated immunopathology might have been due to the
lack of access the Abs had to the autoantigen. In contrast, as MOG is a
cell surface Ag, evidence arose that autoantibodies associated with
type 2 responses might be pathogenic for this Ag (44, 68).
Our experiments using MOG:IFA-injected mice do not point in this
direction. Such mice did not exhibit any evidence of clinical EAE (for
observations up to several months), and the brains were entirely
unaffected upon histologic examination. Nonetheless, the deposition of
IgG1 autoantibodies was readily detectable in the brains of such
MOG:IFA-immunized mice; the autoantibodies were able to penetrate the
BBB and bind to their target cells (Fig. 4). In the absence of
additional T cell-mediated inflammation, the exclusion of complement
from the CNS and the poor complement-fixing properties of the murine
IgG1 Abs (69) may explain the lack of pathogenicity on the
CNS exhibited by such autoantibodies. The lack of pathology associated
with the massive deposition of Ab in the kidney may further reinforce
the notion that the induction of autoantibodies in the mouse can occur
without consequence: in the kidney, complement and phagocytes are
abundant. It is important to note that our IFA injections in mice
induced an autoantibody response void of complement-fixing and
Ab-dependent cellular cytotoxicity-mediating IgG2a. It will be of
interest to determine whether such polarized (pure) Ab isotype
responses are induced by IFA injections in other species (i.e., the
rat) in which the pathogenicity of autoantibodies has been described,
and in which the observation was made that neuroantigen/IFA injections
can cause EAE (46). Strikingly, in the rat, IgG1 Abs are
effective mediators of Ab-dependent cellular cytotoxicity, while they
are not good complement fixers (70). Overall, our data
clearly show that, at least in the mouse, the induction of even the
most vigorous humoral type 2 autoimmunity is fully compatible with
health. The horror autotoxicus (71) may not be
so dreadful if a nonpathogenic class (or classes) of autoimmune
response(s) is engaged.
The classic observation that preinjection with MBP:IFA protects the
host from subsequent induction of MBP:CFA-induced EAE gave rise to the
notion that this pretreatment induces tolerance. Because we actually
found that vigorous type 2 autoimmunity (not immune tolerance) was
induced, we revisited at single cell resolution the
autoantigen-specific T cell response in both the originally protected
host and in hosts protected by adoptive transfer. We tested whether the
preexisting autoantigen-specific type 2 memory cells induced by the IFA
immunization would interfere with the induction of a type 1 T cell
response when the mice were reinjected with the same autoantigen in
CFA. Unlike with naive and control mice, the reinjections of
autoantigen in CFA led to a decreased frequency of IFN- and IL-2
producers and did not engage an IgG2a Ab response. These data are
consistent with exhaustion of naive precursors, representing a passive
tolerance mechanism. Accordingly, the strong primary Ag challenge with
IFA might have partially depleted the pool of naive Ag-specific T cells
and B cells available for the subsequent induction of a type 1 effector
response. In addition, the decreased frequency of IL-2-producing cells
in these mice (Fig. 5), that is, the decreased helper cell function,
may deprive the type 1 effector cells from sustaining their clonal
sizes and effector function.
Our findings of IFN- production in the absence of IL-2 could
alternatively be interpreted as being consistent with T cell anergy in
the reinjected mice (72, 73). Importantly, however, anergy
was originally defined in the context of a Th1/Th2 model in which
IFN- and IL-2 are produced by the same cell. Because we and others
have now demonstrated that IL-2 and IFN- are produced by different
CD4 cells (20, 62), we propose that these results signify
exhaustion of IL-2-producing helper cells.
The exhaustion of naive precursor cells and of IL-2-producing helper
cells both represent passive tolerance mechanisms that should not be
adoptively transferable to a naive recipient. Nevertheless, IFA-induced
tolerance is transferable, which demonstrates that there are additional
active tolerance mechanisms involved. The recipients of adoptively
transferred cells had a naive repertoire available to generate cells
producing IL-2 and IFN-; yet they were still protected from disease.
Although the type 2 component persisted after the reimmunization, it
apparently did not interfere with the induction of a type 1 response;
thus, the presence of the preprimed type 2 memory cells did not
suppress the generation of type 1 cytokine-secreting memory cells, nor
did they cause a type 2 shift of the newly engaged response as would be
predicted by the guided differentiation model (74, 75).
Guided differentiation predicts that the IL-4 produced by the
preexisting Th2 cells would create a type 2 microenvironment when the
Ag is reencountered by naive cells, causing them to differentiate along
the type 2 pathway as well. The CFA immunizations of mice that received
IFA-primed memory cells resulted in frequencies of IFN--secreting T
cells that were comparable with immunizations of naive mice, i.e., we
found no evidence for type 2 shifting of the naive cells that were
induced in the adoptive host (Fig. 6).
While the fact that the IFA-induced autoreactive T cells adoptively
transfer protection fits the classic definition of suppression, we
therefore found no evidence that the IFA-induced T cells would actually
suppress or otherwise inactivate/deviate de novo T cell responses in
the recipient. As an alternative explanation, it is conceivable that
the suppressive/regulatory effect of the IFA-induced T cells was not on
the effector arm itself, but instead directly on the target organ. For
example, subpathogenic immune responses have been shown to cause
up-regulation of the antioxidant stress-associated protein
hemeoxygenase (HO)-1 in cells of the target organ, rendering these
cells resistant to oxidative damage (76), a
prerequisite to development of T cell-mediated autoimmune disease
(77). This mechanism has been associated with protection
from T cell-mediated inflammatory damage of transplanted organs
(76). Immunohistochemical staining experiments did not,
however, reveal up-regulation of HO-1 expression in the target organs
of Ag:IFA-immunized mice when compared with naive mice or mice injected
with Ag:CFA (both in the MOG:EAE and the TBM disease models; data
not shown). Alternatively, classic suppressor cell activity and type 2
cytokine-mediated suppression (e.g., by IL-4, IL-10, TGF-ß) have been
implicated in the inactivation of APC function through incapacitating
effective Ag presentation as a mechanism of tolerance
(30). When we injected RTA:CFA-primed cells directly under
the kidney capsule of RTA:IFA-preinjected mice, however, the cells
produced inflammatory lesions comparable with control recipients (data
not shown). This finding also argues against significant end-organ
resistance.
In summary, while the issue of whether or not the injection of autoantigens in IFA induces clonal inactivation (passive tolerance) or a regulatory response (active tolerance) has been controversial, our utilization of high resolution ELISPOT analysis has permitted us to settle this issue: the injection of autoantigen:IFA induces vigorous immune responses with features of type 2 immunity. The induced tolerant phenotype has features of clonal inactivation because the precursor cells that could give rise to a pathogenic response seem to be converted into memory cells of a nonpathogenic class. Our data also show, however, that the term type 2 immunity does not sufficiently describe the induced immune response. We did not detect classic Th2 cells that coexpress type 2 cytokines as a set, but instead found that the individual Ag-reactive T cells produced these cytokines in a mutually exclusive manner. Moreover, while the IFA-induced cells actively transferred tolerance, no existing model seems sufficient to explain the underlying mechanism. While the exhaustion of naive precursors provides a novel mechanism for passive tolerance, adding to other well-defined pathways, the understanding of active IFA-induced infectious tolerance remains incomplete. An improved comprehension of type 2 immunity per se might be required before we can decipher the mechanisms of active protection afforded by this class of response.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 P.S.H and T.F. contributed equally to this work.
3 Address correspondence and reprint requests to Dr. Paul V. Lehmann, Department of Pathology, Case Western Reserve University, 10900 Euclid Avenue, BRB 929, Cleveland, OH 44106-4943.
4 Abbreviations used in this paper: EAE, experimental allergic encephalomyelitis; TBM, antitubular basement membrane; AEC, 3-amino-9-ethylcarbazole; BBB, blood brain barrier; ELISPOT, enzyme-linked immunospot; HO, hemeoxygenase; LN, lymph node; MBP, myelin basic protein; MOG, myelin oligodendrocyte glycoprotein; PLP, proteolipid protein; PLPp, PLP peptide; RTA, renal tubular Ag.
Received for publication December 30, 1999. Accepted for publication March 22, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ß-transgenic model. J. Exp. Med. 182:1579. and IL-2- producing cells in the CD8+ T cell subset. Eur. J. Immunol. 28:3630.[Medline]
1, 2a, 2b, 2c, 
, and . J. Immunol. 142:3145.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  N. Babel, G. Brestrich, L. P. Gondek, A. Sattler, M. W. Wlodarski, N. Poliak, N. Bethke, A. Thiel, M. H. Hammer, P. Reinke, et al. Clonotype Analysis of Cytomegalovirus-Specific Cytotoxic T Lymphocytes J. Am. Soc. Nephrol., February 1, 2009; 20(2): 344 - 352. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Korn, M. Mitsdoerffer, A. L. Croxford, A. Awasthi, V. A. Dardalhon, G. Galileos, P. Vollmar, G. L. Stritesky, M. H. Kaplan, A. Waisman, et al. IL-6 controls Th17 immunity in vivo by inhibiting the conversion of conventional T cells into Foxp3+ regulatory T cells PNAS, November 25, 2008; 105(47): 18460 - 18465. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. S. Lichtenegger, S. Kuerten, S. Faas, B. O. Boehm, M. Tary-Lehmann, and P. V. Lehmann Dissociation of Experimental Allergic Encephalomyelitis Protective Effect and Allergic Side Reactions in Tolerization with Neuroantigen J. Immunol., April 15, 2007; 178(8): 4749 - 4756. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Mahesh, H.-S. Leong, A. McCormack, P. Sarathchandra, A. Holder, and M. L. Rose Autoantibodies to Vimentin Cause Accelerated Rejection of Cardiac Allografts Am. J. Pathol., April 1, 2007; 170(4): 1415 - 1427. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Valujskikh, Q. Zhang, and P. S. Heeger CD8 T Cells Specific for a Donor-Derived, Self-Restricted Transplant Antigen Are Nonpathogenic Bystanders after Vascularized Heart Transplantation in Mice J. Immunol., February 15, 2006; 176(4): 2190 - 2196. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Visser, H. Jan de Heer, L. A. Boven, D. van Riel, M. van Meurs, M.-J. Melief, U. Zahringer, J. van Strijp, B. N. Lambrecht, E. E. Nieuwenhuis, et al. Proinflammatory Bacterial Peptidoglycan as a Cofactor for the Development of Central Nervous System Autoimmune Disease J. Immunol., January 15, 2005; 174(2): 808 - 816. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. B. Stephensen, X. Jiang, and T. Freytag Vitamin A Deficiency Increases the In Vivo Development of IL-10-Positive Th2 Cells and Decreases Development of Th1 Cells in Mice J. Nutr., October 1, 2004; 134(10): 2660 - 2666. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. C. Bartholomae, F. H. Rininsland, J. C. Eisenberg, B. O. Boehm, P. V. Lehmann, and M. Tary-Lehmann T Cell Immunity Induced by Live, Necrotic, and Apoptotic Tumor Cells J. Immunol., July 15, 2004; 173(2): 1012 - 1022. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Le Hir Histopathology of humorally mediated anti-glomerular basement membrane (GBM) glomerulonephritis in mice Nephrol. Dial. Transplant., July 1, 2004; 19(7): 1875 - 1880. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Darabi, A. Y. Karulin, B. O. Boehm, H. H. Hofstetter, Z. Fabry, J. C. LaManna, J. C. Chavez, M. Tary-Lehmann, and P. V. Lehmann The Third Signal in T Cell-Mediated Autoimmune Disease? J. Immunol., July 1, 2004; 173(1): 92 - 99. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Mana, M. Goodyear, C. Bernard, R. Tomioka, M. Freire-Garabal, and D. Linares Tolerance induction by molecular mimicry: prevention and suppression of experimental autoimmune encephalomyelitis with the milk protein butyrophilin Int. Immunol., March 1, 2004; 16(3): 489 - 499. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Tian, A. P. Olcott, and D. L. Kaufman Antigen-Based Immunotherapy Drives the Precocious Development of Autoimmunity J. Immunol., December 1, 2002; 169(11): 6564 - 6569. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Pantenburg, F. Heinzel, L. Das, P. S. Heeger, and A. Valujskikh T Cells Primed by Leishmania major Infection Cross-React with Alloantigens and Alter the Course of Allograft Rejection J. Immunol., October 1, 2002; 169(7): 3686 - 3693. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. H. Hofstetter, C. L. Shive, and T. G. Forsthuber Pertussis Toxin Modulates the Immune Response to Neuroantigens Injected in Incomplete Freund's Adjuvant: Induction of Th1 Cells and Experimental Autoimmune Encephalomyelitis in the Presence of High Frequencies of Th2 Cells J. Immunol., July 1, 2002; 169(1): 117 - 125. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Saarelainen, T. Zeiler, J. Rautiainen, A. Narvanen, M. Rytkonen-Nissinen, R. Mantyjarvi, P. Vilja, and T. Virtanen Lipocalin allergen Bos d 2 is a weak immunogen Int. Immunol., April 1, 2002; 14(4): 401 - 409. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Billiau and P. Matthys Modes of action of Freund's adjuvants in experimental models of autoimmune diseases J. Leukoc. Biol., December 1, 2001; 70(6): 849 - 860. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
). A,
RTA-specific responses in SJL mice. B, MBP-specific
responses in B10.PL mice. C, PLPp-specific responses in
SJL mice. D, MOG-specific responses in C57BL/6 mice.
Each data point represents a single mouse. Shown is the mean number of
spots counted by computer-assisted image analysis in two to three
replicate wells (<15% variability between wells) after subtracting
spontaneous production in wells with medium alone (<5 spots per well
for IFN-
