| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Institute of Experimental Immunology, Department of Pathology, University of Zürich, Zürich, Switzerland; and
Department of Immunohematology and Blood Bank, University of Leiden, Leiden, The Netherlands
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cellular debris, such as that resulting from cytopathic viral infections or traumatic tissue damage, may be taken up by DC and processed via the exogenous pathway to be presented as peptides associated with MHC class II molecules (7). Indeed, DC are highly efficient in acquisition of exogenous Ags and presentation of high levels of MHC class II-restricted peptides in vitro (8). DC also present pancreatic self-Ags via MHC class II during the development of autoimmune diabetes in nonobese diabetic mice (9). Furthermore, it has been shown in vitro that DC may present MHC class I-associated peptides after uptake of concentrated protein (10) or after uptake of virus-infected, apoptotic cells (11). MHC class I-restricted presentation of exogenous soluble, particulate, or cell-associated Ags has been termed cross-priming (12) and can be demonstrated in vivo in certain transgenic model situations where self-reactive CTL are peripherally deleted (13, 14). However, if truly ignored self-Ags could easily gain access to the exogenous MHC class I cross-presentation pathway, e.g., in the course of a cytopathic virus infection, self-reactive CTL could be activated and initiate a self-perpetuating circle of tissue destruction and subsequent increased release of self-Ags. Therefore, it is conceivable that a number of mechanisms exist that limit the induction of anti-self reactions, particularly in the CD8 compartment. These may be signals such as negative feedback stimulation via CTLA-4 during DC-T cell interaction (15). However, it is unclear and unexplained to what extent presentation of self-peptides via MHC class I, particularly after exogenous uptake of either peptides or whole proteinacious Ag by DC, influences the induction of autoimmunity.
In the present report, we investigated mechanisms that might contribute to limit the autoimmune sequelae after release of self-Ags and subsequent presentation by DC. We found that the duration of peptide presentation locally during DC-CTL interaction is critical for the induction of autoreactive CTL and that exogenous, cell-associated Ags are poorly immunogenic for CTL, even when presented by DC.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
C57BL/6 (B6) mice were obtained from the Institut für
Labortierkunde (University of Zurich, Zurich, Switzerland). Mice
expressing lymphocytic choriomeningitis virus (LCMV)-glycoprotein (gp)
under the control of the rat insulin promotor (RIP)-gp
(2), mice expressing the LCMV gp33 epitope ubiquitously
(H8 mice; Ref. 16), and mice transgenic for a
V
2/V
8.1 TCR specific for H2-Db and the
major LCMV-gp epitope gp33–41 (gp33) (17), or for a
V2/V8.3 TCR specific for I-Ab and the
LCMV-gp peptide p13 (18) have been described previously.
Experiments were conducted with age (8–16 wk)- and sex-matched
animals.
Cell lines, Abs, and peptides
Methyl-collanthrene-induced fibrosarcoma cells expressing LCMV-gp (MC-gp) have been described previously (19). MC-gp were made apoptotic by cultivation at 43°C for 24 h; necrotic MC-gp were produced by three rounds of freeze-thawing. Apoptosis of MC-gp was tested by FACS analysis, and routinely 50–75% of the apoptotic MC-gp were propidium iodide negative and annexin V positive; necrotic MC-gp were always 100% trypan blue positive. EL-4 (H-2b), a thymoma cell line, was used as target cell. LCMV-gp peptides KAVYNFATM (gp33), FQPQNGQFI (nucleoprotein (NP) 396), and GLNGPDIYKGVYQFKSVEFD (p13) were produced from Neosystem Laboratoire (Strasbourg, France).
Supernatants or purified Ig from the following mAb-producing hybridomas were used: rat anti-mouse CD4 (YTS191.1; Ref. 20), rat anti-mouse CD8 (YTS169.4.2; Ref. 20), rat anti-mouse CD45R (RA3-3A1/6.1; American Type Culture Collection (ATCC), Manassas, VA), rat anti-mouse I-Ab (B21-2; ATCC), anti-CD40 (21), anti-CTLA-4 (22), and anti-H2-Db (141-51; Ref. 23).
Preparation and Ag pulse of DC
Generation of DC from B6 and H8 bone marrow cultures has been described previously (24). Both DC populations express high levels of the costimulatory molecules CD80, CD86, and CD40, and >50% of the cells express high levels of MHC class II Ags. Incubation of the DC with LPS leads to further up-regulation of these surface markers (not shown). For Ag pulsing, DC were resuspended in RPMI 1640/5% FCS at 106/ml and incubated with the indicated peptides for 60 min. Alternatively, DC were cocultivated with apoptotic or necrotic MC-gp at a ratio of 1:5 (DC:MC-gp) for 6 h at 37°C followed by metrizamide gradient centrifugation to separate the fibroblasts from the DC. DC were washed two times with balanced salt solution and i.v. injected in a volume of 0.5 ml.
Cytotoxicity assays
Spleen cells (4 x 106/well) from
primed mice were restimulated for 5 days in 24-well tissue culture
plates with 2 x 106 gp33-labeled,
irradiated (1000 rad) spleen cells in IMDM supplemented with 10% FCS,
penicillin/streptomycin, and 0.001 M 2-ME. Restimulated spleen effector
cells from one well were resuspended in 1 ml MEM/2% FCS, and 3-fold
serial dilutions were made (indicated as dilution of culture). For
detection of primary ex vivo cytotoxicity, effector cell suspensions
were prepared from spleens of immunized mice at the indicated time
point after priming. EL-4 cells were pulsed with LCMV-gp33 or
LCMV-NP394 (10-5 M, 1.5 h at 37°C) and
used in a standard 5-h 51Cr release assay or in
an overnight (15 h) assay. Unlabeled EL-4 cells served as controls. The
supernatant of the cytotoxicity cultures was counted in a Cobra II 
Counter (Canberra Packard, Downers Grove, IL). Spontaneous release was
always below 20% for 5 h assays and below 30% for overnight
assays.
Proliferation assay
DC were preincubated with gp33 at the indicated concentration or left untreated and washed twice with balanced salt solution. Spleen cells from 318 TCR-transgenic mice recognizing the gp33 epitope were incubated in round-bottom 96-well plates with 3-fold serial dilutions of DC, starting with 2 x 104 DC per well. In blocking experiments, DC were incubated with the indicated concentration of anti-H2-Db mAb 141-51 for 30 min before the addition of 318 responder cells. After incubation for 48 h at 37°C/5% CO2, [3H]thymidine (1 µCi/well) was added for a further incubation period of 12 h. Proliferation was determined as [3H]thymidine incorporation by using a Microbeta scintillation counter (Wallac, Turku, Finland).
Cytofluorometry
Naive B6 mice were transfused i.v. with 5 x
105 syngeneic spleen cells from 318 mice
(containing 
5 x 104 TCR-transgenic CTL)
on day -1 (B6/318). To detect expansion of transgenic TCR-expressing T
cells (25) after immunization with H8-DC, peripheral blood
cells were stained for CD8 and transgenic V2 and V8.1 using
FITC-conjugated rat anti-mouse CD8, PE-conjugated rat
anti-mouse V2 and biotinylated rat anti-mouse V8.1,
respectively (all obtained from BD PharMingen, San Diego, CA) followed
by streptavidin-Tricolor (Caltag, South San Francisco, CA). Activation
of gp33-specific CTL was detected by staining with biotinylated
anti-CD44 (BD PharMingen) followed by incubation with
streptavidin-Tricolor (Caltag). Tetrameric class I complexes of
H-2Db and gp33–41 were produced as described
previously (26) and used to identify gp33-specific
CD8+ T cells in spleen cell suspensions.
Erythrocytes were lysed with FACS lysis solution (Becton Dickinson),
and the cell suspensions were analyzed on a FACScan flow cytometer
(Becton Dickinson) after gating on viable lymphocytes.
Immunohistology
Freshly removed organs were immersed in HBSS and snap-frozen in liquid nitrogen. Tissue sections of 5 µm thickness were cut in a cryostat and fixed in acetone for 10 min. Sections were incubated with anti-mouse mAb against CD8+ cells (YTS169.4.2), and polyclonal guinea pig Abs against insulin (Dako, Glostrup, Denmark). Alkaline phosphatase-labeled, species-specific goat Abs (Tago Scientific, Burlingame, CA) were used as secondary reagents. The substrate for the red color reaction was AS-BI phosphate/New Fuchsin. Sections were counterstained with hemalum. Fifteen to 20 islets from two or three histological sections of each mouse were evaluated. The fluorescent dye CFSE (Molecular Probes, Eugene, OR) was used to label DC (24). A total of 2 x 105 CFSE-labeled DC were injected s.c. into the footpad of naive recipients. Draining popliteal lymph nodes were removed at different time points, immersed in HBSS, and snap-frozen in liquid nitrogen. Histological procedures were performed as above using rabbit anti-fluorescein Ab (Dako).
Measurement of blood glucose
The glucose concentration in blood obtained from a tail vein was measured by using an ELITE Haemoglucometer (Bayer, Wuppertal, Germany). Mice were considered diabetic with values >14 mM at two consecutive measurements.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have shown previously that DC derived from H8 mice (H8-DC) and
constitutively expressing a viral neoself-peptide (the
H-2Db-restricted peptide gp33 derived from
LCMV-gp) are able to induce autoimmune diabetes after adoptive transfer
into RIP-gp mice (27). In RIP-gp mice, the viral transgene
(LCMV-gp) is immunologically ignored such that both strong initial
expansion and sustained activation of self-reactive CTL by repetitive
immunization with DC (27) or by infection with LCMV
(2) are required to mediate autoimmune disease. Here, we
assessed whether DC exogenously loaded with gp33 are able to induce
diabetes in RIP-gp mice. We chose a protocol where repetitive i.v.
injection of 2 x 105 H8-DC led to manifest
diabetes in > 90% of the RIP-gp mice (Fig. 1
A). In contrast to the rapid
hyperglycemia that developed after injection of H8-DC, RIP-gp mice
remained normoglycemic when the same protocol was applied with
nontransgenic DC derived from B6 mice (B6-DC) exogenously pulsed with
gp33 (Fig. 1B). H8-DC exogenously loaded with gp33 also
induced diabetes (Fig. 1C), indicating the endogenous
production of self-peptide is responsible for the high autoimmune
potential of the constitutively gp33-expressing H8-DC. Increasing the
dose and frequency of application of exogenously peptide-loaded B6-DC
partially compensated for the limited ability of peptide-pulsed B6-DC
to elicit strong anti-self CTL responses in 6 of 14 mice (Fig. 1D and Table I), suggesting
that the amount of Ag available for a certain time during interaction
of DC with self-reactive CTL is probably a critical parameter for
CD8+-T cell-mediated autoimmune disease.
|
|
, A and E),
gp33-pulsed B6-DC/gp33 (Fig. 2, B, C,
F, and G) or unpulsed control B6-DC (Fig. 2, D and H) revealed that activation of
self-reactive CTL by H8-DC led to strong and destructive infiltration
of insulin-positive pancreatic islets (Fig. 2A) with
CD8-positive lymphocytes (Fig. 2E). Similarly, repetitive
priming with a high dose of B6-DC/gp33 (six injections of 5 x
105) elicited a destructive insulitis (Fig. 2, C and G), whereas less frequent immunization with
a lower dose of B6-DC/gp33 (four injections of 2 x
105) induced only a mild inflammation in the
endocrine pancreas with peri-insular infiltration of
CD8+ T cells (Fig. 2, B and
F). RIP-gp mice immunized repetitively with a high dose of
control B6-DC did not show inflammatory responses in the pancreas (Fig. 2, D and G). Taken together, it appears that
there are significant qualitative differences in the autoimmune
potential between DC constitutively expressing a self-peptide and DC
that have acquired the peptide exogenously: H8-DC induced destructive
insulitis and manifest hyperglycemia, whereas exogenously pulsed DC
induced mild insulitis, and only when administered in high numbers and
with high frequency did a hyperglycemia develop.
|
The above findings suggested that the duration of peptide
presentation by DC in the microenvironment of lymphoid tissues might
critically influence the activation of self-reactive CTL. In principle,
two steps limiting peptide presentation by DC in these experiments may
be envisaged: 1) short persistence of the exogenous peptide on DC,
and/or 2) shorter life-span of peptide-pulsed DC. To test the latter
possibility, B6-DC and H8-DC were labeled with CFSE and injected s.c.
into the footpad of naive C57BL/6 recipients. One and 2 days after
injection of 2 x 105 cells, similar numbers
of both H8-DC or B6-DC/gp33 reached the draining popliteal lymph node
(Fig. 3, A, B, and
D). From day 2 on, the number of DC decreased continuously
and both H8-DC and B6-DC/gp33 were hardly detectable by day 5 after
injection (Fig. 3, A, C, and E). As
the persistence of H8-DC and B6-DC/gp33 appeared to be comparable, it
appears more likely that differences in the persistence of peptide may
account for the profound differences in the autoimmune potential of
exogenously peptide-pulsed vs constitutively peptide-expressing
DC.
|
A) or first cultured in the
absence of peptide for 1 (Fig. 4B) or 2 days (Fig. 4C) before they were used as stimulators. H8-DC and unpulsed
B6-DC were processed in parallel and served as controls. B6-DC pulsed
with gp33 at high concentrations, i.e., at 10-6
M or 10-8 M, maintained the capacity to
stimulate naive CTL in vitro for >48 h (Fig. 4C). After
pulsing of the DC at the low concentration of
10-10 M, the stimulatory capacity of the DC was
strongly reduced after 1 day (Fig. 4B) and lost after 2 days
(Fig. 4C), indicating that B6-DC were efficiently pulsed
with gp33 and presented MHC class I-associated peptides only for a
limited period of time when pulsed with low doses of gp33. Next, we
evaluated the availability of gp33 on H8-DC and B6-DC by blocking with
an Ab specific for H2-Db (141-51; Ref.
23). H8-DC or B6-DC/gp33 were incubated with increasing
concentrations of the blocking Ab and subsequently cocultured with 318
spleen cells. The anti-MHC class I Ab inhibited the proliferative
response induced by H8-DC and B6-DC pulsed with
10-6 and 10-8 M gp33 in a
similar extent, whereas proliferation induced by B6-DC pulsed with
10-10 M gp33 was completely inhibited (Fig. 4D). These data suggest that the relative density of the
gp33 Ag on H8-DC and B6-DC pulsed with gp33 at high concentrations is,
at least initially, equivalent.
|
, A–D) or
B6-DC/gp33 (Fig. 5, E–H) were adoptively
transferred into naive recipients, and TCR-transgenic 318 spleen cells
were cotransferred either on the same day (Fig. 5, A,
B, E, and F) or 2 days later (Fig. 5, C, D, G, and H). Four days
after the transfer of TCR-transgenic T cells, the expansion of
gp33-specific CTL was determined by staining for the transgenic TCRs
V2 and V8.1. In addition, the activation status of the cells was
assessed by staining for the activation marker CD44. H8-DC transferred
and "parked" in vivo for 48 h were still able to significantly
expand and activate naive CTL (Fig. 5, C and D),
compared with H8-DC cotransferred with 318 cells on the same day (Fig. 5, A and B). Peptide-pulsed B6-DC showed a
similar efficiency to activate naive CTL (Fig. 5, E and
F) when cotransferred with 318 cells on the same day.
However, B6-DC pulsed with gp33 at 10-6 M had
lost their ability to expand and activate naive 318 cells after 48
h (Fig. 5, G and H). Fig. 5, I and
J summarize these experiments and show that B6-DC pulsed
with gp33 at 10-6 M or
10-8 M, but not with
10-10 M, efficiently expanded and activated
naive CTL after transfer on day 0. Importantly, after only 1 day,
gp33-pulsed B6-DC had lost most of their in vivo-stimulatory capacity
for gp33-specific CTL (Fig. 5, I and J, day 1),
indicating that the turnover of gp33 on B6-DC in vivo is rapid and
might be a limiting factor for the activation of naive CTL.
|
A) and B6-DC/gp33
(Fig. 6D) induced comparable polyclonal CTL responses as
determined in a 51Cr release assay after
secondary restimulation in vitro or by staining with tetrameric MHC
class I complexes (Fig. 6, B and E). In contrast,
primary ex vivo cytotoxicity was significantly stronger after priming
with H8-DC (Fig. 6C) compared with priming with B6-DC/gp33
(Fig. 6F), indicating that the short persistence of gp33 on
B6-DC limits the differentiation into effector CTL. Collectively, these
data suggest that DC presenting exogenously acquired self-peptides can
efficiently prime self-reactive CTL. However, the autoimmune potential
of these self-reactive CTL is reduced because the differentiation into
CTL with effector function is limited by the rather short half-life of
the peptides on DC in vivo.
|
We next examined whether activation of cognate
CD4+ Th cells or nonspecific activation signals
might influence the response of gp33-specific CTL. This was achieved by
either cotransferring TCR-transgenic Th cells specific for the LCMV-gp
epitope p13 (Smarta cells; Ref. 18), by nonspecific
activation of APC via CD40 (28), or by blocking potential
negative regulatory signals, such as inhibition via CTLA-4
(15). After injection of Smarta splenocytes and activation
of both Smarta Th cells and gp33-specific CTL by peptide-pulsed DC,
RIP-gp mice rapidly developed diabetes, comparable to the kinetics
after transfer of constitutively gp33-expressing H8-DC (Table I
). Only
1 x 105 DC-activated Smarta splenocytes
containing 1–2 x 104 TCR-transgenic Th
cells were sufficient to compensate for the limited activation of
gp33-specific CTL by peptide-pulsed DC (Table I). In contrast, in the
absence of additional TCR-transgenic Th cells, B6-DC pulsed with gp33
and p13 did not induce diabetes (Table I). Similarly, DC pulsed with
p13 alone and coinjected with 1 x 107
Smarta splenocytes failed to elicit a diabetogenic response (Table I),
indicating that both islet-specific CTL and relatively large numbers of
specific Th cells had to be activated by peptide-pulsed DC to induce a
strong autoimmune response. A similar effect was achieved when
unspecific "Th-like" signals were provided by repeated injection of
a cross-linking anti-CD40 Ab; 33% of the RIP-gp mice treated with
gp33-pulsed DC and anti-CD40 developed diabetes (Table I).
Interestingly, blocking of CTLA-4 did not lower the threshold for
activation of diabetogenic CTL after stimulation with gp33-pulsed B6-DC
(Table I). Thus, rather strong signals, provided either by
concomitantly activated transgenic Th cells or via stimulation of APC
via CD40 were required to convert a weak anti-self response induced
by DC presenting exogenously acquired self-peptides into clinically
manifest autoimmune disease.
Inefficient presentation of MHC class I-associated peptides by DC after acquisition of exogenous Ags
It has been suggested that acquisition of exogenous Ags, e.g.,
released from either necrotic or apoptotic cells, by bone
marrow-derived APC might be a mechanism for induction of CTL responses
(29). To evaluate the role of exogenously acquired
self-Ags by DC for the induction of autoimmune responses, we tested the
activation of gp33-specific CTL with either apoptotic or necrotic
LCMV-gp-expressing fibroblasts (MC-gp). The specific CTL response after
adoptive transfer of B6-DC pulsed with necrotic MC-gp was negative
(Fig. 7A). Similarly, B6-DC
exposed to apoptotic MC-gp for 6 h did not elicit a measurable CTL
response after adoptive transfer into naive recipients (Fig. 7C). However, when the DC were pulsed with gp33 for the last
60 min of the incubation period, a strong CTL response was generated
(Fig. 7, B and D), indicating that the DC were
not damaged by the cellular debris and were still capable of activating
gp33-specific CTL. The minimal amount of peptide that DC had to be
exposed to to induce a minimal CTL response was
10-10 M (Fig. 7E).
|
, F and
G) or apoptotic MC-gp (Fig. 7, H and
I) into the spleen, where splenic DC and other professional
host APC could take up and process the cellular proteins and induce
gp33-specific CTL. Repeated intrasplenic injection of necrotic MC-gp
induced a minimal CTL response (Fig. 7G), whereas after
single injection of 5 x 105 MC-gp, CTL
activation was barely detectable (Fig. 7F). Neither single
(Fig. 7H) nor repeated (Fig. 7I) injection of
106 apoptotic MC-gp elicited a significant CTL
response. The number of necrotic cells had to be reduced, because the
cellular debris formed clumps that caused death of the animal when
>5 x 105 cells were injected directly into
the spleen. As mentioned above, MC-gp express considerable amounts of
LCMV-gp; therefore, the failure of apoptotic or necrotic MC-gp to
elicit a significant CTL response is not likely to be a result of
insufficient Ag amount, because only 1 x
104 irradiated, i.e., nonreplicating, MC-gp
induced a CTL response after intrasplenic injection (Fig. 7J).
We next determined whether repeated delivery of apoptotic or necrotic
MC-gp via the s.c. route, where immature dermal and epidermal DC might
efficiently pick up and transport the cellular Ag to secondary lymphoid
organs, elicits a significant anti-self response in RIP-gp mice. As
a control, 2 x 105 H8-DC were repetitively
administered s.c. Similar to the i.v. route, s.c. application of H8-DC
induced severe insulitis and long-lasting hyperglycemia (Table II). Peripheral s.c. injection of high
doses (107) of irradiated MC-gp that migrate to
secondary lymphoid organs and directly induce CTL responses (19, 30) elicited a weak anti-islet response with mild insulitis
and diabetes in one of six mice. Importantly, repetitive application of
high doses of necrotic or apoptotic MC-gp did not suffice to induce
diabetes or insulitis (Table II). To test whether concomitant delivery
of LCMV-gp-specific Th might convert the weak anti-self response
after priming with MC-gp into an autoimmune response with pathological
consequences, 107 Smarta spleen cells were
adoptively transferred 1 day before the start of the treatment with
MC-gp cells. In the presence of high numbers of islet-specific Th
cells, repetitive priming with irradiated MC-gp induced diabetes in
50% of the RIP-gp mice and exacerbated the insulitis in these mice
(Table II). Similarly, in RIP-gp mice treated with necrotic MC-gp, the
islet-specific Th cells contributed to an aggravation of the islet
infiltration leading to mild insulitis (Table II).
|
A) or B6
(Fig. 8B) mice were repetitively immunized s.c. with
107 necrotic MC-gp and subsequently challenged
with LCMV. Similar to untreated controls, RIP-gp mice primed with
necrotic MC-gp developed diabetes between days 8 and 12 (Fig. 8A). In addition, B6 mice primed with necrotic MC-gp (Fig. 8B) showed a strong CTL reactivity against gp33 as well as
the nucleoprotein epitope 396 (NP396) on day 8 after infection with
LCMV, comparable to untreated controls (Fig. 8D). In
contrast, repetitive i.p. administration of 100 µg of gp33 in the
mild adjuvant IFA tolerized gp33-reactive CTL, but not NP394-reactive
CTL (Fig. 8C) and therefore prevented LCMV-induced diabetes
in RIP-gp mice (Fig. 8A). Taken together, these data clearly
show that, in our model system, MHC class I presentation of
cell-associated Ags derived from cellular debris is rather inefficient
and did not lead to tolerization of self-reactive CTL. Importantly,
even the presence of high numbers of islet-specific Th cells did not
convert the weak anti-self CTL activity after priming with necrotic
MC-gp into a self-perpetuating and clinically manifest autoimmune
response.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Duration of self-Ag presentation by DC is a critical parameter for autoimmune responses
During their ontogeny, DC undergo characteristic functional changes that allow Ag acquisition in peripheral nonlymphoid organs during the immature stage and, after maturation, efficient Ag presentation in secondary lymphoid tissues. The turnover of MHC class II-associated peptides in vitro is high on immature DC, whereas inflammatory stimuli leading to DC maturation significantly enhance the half-life of MHC class II molecules and prolong Ag presentation (8). Similarly, the turnover of MHC class I-associated peptides on human DC in vitro appears to be regulated by maturation stimuli such as virus infections (31). Our comparison of exogenously peptide-pulsed DC with DC constitutively expressing the respective peptides allowed evaluation of the significance of peptide persistence on DC in vitro and in vivo. We found that the half-life of gp33 on nontransgenic DC is short and thereby limits the activation of gp33-specific CTL. This is most likely attributable to the spatial distribution and the initially rather rare cognate interaction between adoptively transferred DC and the responding CTL. In vivo, DC and responding CTL firstly have to home to secondary lymphoid organs and, secondly, most probably have to go through a number of noncognate interactions before meeting the correct partner.
Our data clearly show that the amount of self-peptides on DC and the duration of peptide presentation by DC in secondary lymphoid tissues is of prime importance for the induction of anti-self CTL responses and the degree of tissue destruction. Constitutively gp33-presenting DC efficiently expanded the gp33-specific CTL pool and induced differentiation of the CTL into effectors. In contrast, although the initial expansion of gp33-specific CTL precursors was comparable, exogenously peptide-pulsed DC could not mediate complete differentiation of gp33-specific CTL. Therefore, it appears that the prolonged presentation of peptide by DC for 2–4 days is important to mediate both expansion and differentiation of CTL. Extended presence and/or high doses of Ag in local proliferation/differentiation clusters of APC with CTL might be a prerequisite for optimal induction of T cell responses as suggested by Mitchinson and O’Malley (32). Accordingly, gp33-pulsed DC had to be adoptively transferred repetitively over several days and in higher doses to achieve diabetogenic CTL responses in our model situation.
The failure to induce autoimmunity because of the limited persistence
of exogenous peptide on DC could be compensated for by simultaneous
induction of strong T help provided by simultaneously activated
TCR-transgenic Th cells. The observed effect of Th may be a result of
various mechanisms: bystander help to limited CTL responses, activation
of IFN- or TNF effector pathways, or enhanced maturation or
activation of DC, perhaps by stimulation via CD40 (28, 33). Indeed, injection of an activating anti-CD40 mAb was
able to partially compensate for the reduced immunogenicity of
peptide-pulsed DC. However, the addition and activation of specific Th
cells was more efficient than CD40 ligation in lowering the threshold
for obtaining autoimmunity in response to DC presenting exogenous
self-peptides. It has been shown that DC also mature rapidly during
cognate interaction with CTL, even in the absence of CD40
(34), indicating that a number of Th cell signals are
required to mediate complete CTL activation in the three-cell
interaction between DC, Th cells, and CTL. CTLA-4 has been suggested to
be important for the maintenance of peripheral tolerance by inhibiting
T cells that received only low-level costimulation for brief periods
(35). However, inhibition of T cell activation via CTLA-4
does not appear to be involved in the present system as blockade of
CTLA-4 ligation was not able to increase the ability of DC presenting
exogenous self-Ag to mediate autoimmunity. The presented system of
DC-induced autoimmune disease appears to be particularly well suited to
further investigate the Ag-specific interactions of the three cell
types and their importance of CTL and Th cell activation and subsequent
induction of an autoimmune disease.
A role for cross-presentation by DC in the induction of autoimmunity?
The genetically regulated cognate interaction between target cells and activated CTL resulting in target cell lysis has been termed "MHC restriction" (36). However, it has also been shown that exogenous, cell-associated proteins may gain access to the MHC class I presentation pathway via "cross presentation" (12). In addition, cross-presentation has been demonstrated in transgenic situations, where model Ags such as OVA (37) or influenza hemagglutinin (HA; Ref. 14) are expressed in peripheral nonlymphoid tissues. In OVA-transgenic mice, the peripheral Ag is presented via MHC class I in draining secondary lymphoid organs by bone marrow-derived APC (37). However, it appears that whenever the expression of a self-Ag is high and widespread, central (thymic) tolerance eliminates self-reactive, high-avidity T cells (37, 38). Local and intermediate expression of self-Ags, such as HA in RIP-HA mice, may also lead to elimination of high-avidity anti-self CTL by deletional mechanisms possibly involving cross-presentation via MHC class I (14, 39). However, complete deletion of self-reactive CTL via cross-priming in the "intermediate expressor" situation of RIP-HA mice appears to be rather inefficient, because >100 days were needed for the elimination of only 104 self-reactive, high-avidity CTL (14).
Spontaneous autoimmunity that is mediated by cross-presentation can only be detected in "high expresser" situations with widespread expression of transgenic OVA, when central tolerance mechanisms are bypassed by transfer of TCR-transgenic CTL (13). Similarly, peripheral tolerance mechanisms that prevent spontaneous autoimmunity, e.g., in RIP-HA mice (39), can be circumvented by crossing with TCR-transgenic animals, thereby providing great numbers of self-reactive T cells that can either not be tolerized by the peripheral tolerizing mechanism (40, 41), or perhaps may be accidentally activated too frequently when compared with a "normal" low-frequency CTL precursor situation. In contrast, in transgenic situations with lower and more restricted expression of the model self-Ag, e.g., OVA (5) or LCMV-gp (2) in pancreatic islet cells, the presence of high numbers of specific TCR-transgenic T cells does not induce autoimmunity, despite potential release of self-Ag from pancreatic islets and subsequent cross-presentation in local lymph nodes (5). Therefore, it appears that self-perpetuating autoimmunity attributable to cross-presentation is a phenomenon that appears in certain transgenic constellations where T cell frequencies are extremely high and/or central tolerance mechanisms are bypassed. The data presented in this report strongly support this interpretation because not even a minimal autoimmune response was detected in RIP-gp when large amounts of self-Ag in the form of apoptotic or necrotic cells were repetitively injected. Only in the presence of high numbers of specific TCR-transgenic Th cells did a mild insulitis develop after injection of necrotic LCMV-gp expressing fibroblasts. However, in this situation as well, a self-perpetuating and progressive autoimmune response did not develop and the mice stayed normoglycemic. Taken together, it appears that high threshold levels for presentation of exogenous Ags via MHC class I prevent self-perpetuating autoimmune disease.
Use of exogenously peptide- or tumor protein-pulsed DC in antitumor vaccination
The migratory function and strong Ag-presenting properties make DC excellent vehicles for transport of tumor Ags into secondary lymphoid organs to facilitate induction of antitumor immunity. Indeed, DC are used currently in many tumor treatment programs as a means to generate primary antitumor immune responses or to enhance the existing tumor immunity (42, 43). However, Ag-specific tumor immunotherapy can be only successful if the tumor Ag has been immunologically ignored previously, i.e., specific, high-avidity T cells have not been centrally or peripherally deleted, and antitumor T cells can be induced by appropriate immunization. Therefore, it is possible to apply the conclusions from the present report also for the induction of tumor immunity with DC. First, tumor-specific CTL are probably most efficiently induced when a previously ignored tumor Ag is constitutively expressed by the DC. A number of approaches have been described to achieve in vitro transfection of DC with tumor Ags (44, 45). Second, exogenous pulsing of DC with tumor peptides is less efficient, because the rapid in vivo turnover of MHC class I-associated peptides limits the induction of specific CTL or requires additional exogenous stimuli, such as activation of APC via CD40. Finally, exogenous pulsing of DC with whole tumor proteins is probably very inefficient. It has been shown that either large amounts of pure protein or, alternatively, special delivery vehicles such as microbeads are required to overcome this inefficiency of exogenous loading of MHC class I molecules by DC (46).
In conclusion, we have demonstrated that the rapid turnover of MHC class I-associated peptides and the rather inefficient presentation of exogenous cellular self-Ags are likely to be basic and important mechanisms to limit the autoimmune consequences of self-Ag presentation by DC.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Burkhard Ludewig, Institute of Experimental Immunology, Department of Pathology, University of Zürich, Schmelzbergstrasse 12, CH-8091 Zürich, Switzerland.
3 Abbreviations used in this paper: DC, dendritic cells; LCMV, lymphocytic choriomeningitis virus; gp, glycoprotein; RIP, rat insulin promotor; MC-gp, methyl-collanthrene-induced fibrosarcoma cells expressing LCMV-gp; H8-DC, DC derived from H8 mice; HA, influenza hemagglutinin; NP, nucleoprotein; B6, C57BL/6.
Received for publication April 13, 2000. Accepted for publication January 4, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
in neonatal NOD mice promotes diabetes by enhancing presentation of islet antigens. Immunity 9:733.[Medline]
heavy chain class switching. Immunity 5:319.[Medline]
This article has been cited by other articles:
|
|
 |
|
  M. Basler, C. Lauer, U. Beck, and M. Groettrup The Proteasome Inhibitor Bortezomib Enhances the Susceptibility to Viral Infection J. Immunol., November 15, 2009; 183(10): 6145 - 6150. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Li, D. K. Krishnadas, R. Kumar, D. L. J. Tyrrell, and B. Agrawal Priming and stimulation of hepatitis C virus-specific CD4+ and CD8+ T cells against HCV antigens NS4, NS5a or NS5b from HCV-naive individuals: implications for prophylactic vaccine Int. Immunol., January 1, 2008; 20(1): 89 - 104. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Yoshino, H. Yamazaki, L. D. Shultz, and S.-I. Hayashi Constant rate of steady-state self-antigen trafficking from skin to regional lymph nodes Int. Immunol., November 1, 2006; 18(11): 1541 - 1548. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Delamarre, R. Couture, I. Mellman, and E. S. Trombetta Enhancing immunogenicity by limiting susceptibility to lysosomal proteolysis J. Exp. Med., September 4, 2006; 203(9): 2049 - 2055. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Li, D. K. Krishnadas, J. Li, D. L. J. Tyrrell, and B. Agrawal Induction of Primary Human T Cell Responses against Hepatitis C Virus-Derived Antigens NS3 or Core by Autologous Dendritic Cells Expressing Hepatitis C Virus Antigens: Potential for Vaccine and Immunotherapy J. Immunol., May 15, 2006; 176(10): 6065 - 6075. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Saito, M. Hirokawa, K. Inaba, H. Fukaya, Y. Kawabata, A. Komatsuda, J. Yamashita, and K. Sawada Phagocytosis of codeveloping megakaryocytic progenitors by dendritic cells in culture with thrombopoietin and tumor necrosis factor-{alpha} and its possible role in hemophagocytic syndrome Blood, February 15, 2006; 107(4): 1366 - 1374. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Dieckmann, E. S. Schultz, B. Ring, P. Chames, G. Held, H. R. Hoogenboom, and G. Schuler Optimizing the exogenous antigen loading of monocyte-derived dendritic cells Int. Immunol., May 1, 2005; 17(5): 621 - 635. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X.-L. Huang, Z. Fan, B. A. Colleton, R. Buchli, H. Li, W. H. Hildebrand, and C. R. Rinaldo Jr. Processing and Presentation of Exogenous HLA Class I Peptides by Dendritic Cells from Human Immunodeficiency Virus Type 1-Infected Persons J. Virol., March 1, 2005; 79(5): 3052 - 3062. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Fassnacht, A. Ackermann, P. Staeheli, and J. Hausmann Immunization with dendritic cells can break immunological ignorance toward a persisting virus in the central nervous system and induce partial protection against intracerebral viral challenge J. Gen. Virol., August 1, 2004; 85(8): 2379 - 2387. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Wiethe, K. Dittmar, T. Doan, W. Lindenmaier, and R. Tindle Enhanced Effector and Memory CTL Responses Generated by Incorporation of Receptor Activator of NF-{kappa}B (RANK)/RANK Ligand Costimulatory Molecules into Dendritic Cell Immunogens Expressing a Human Tumor-Specific Antigen J. Immunol., October 15, 2003; 171(8): 4121 - 4130. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Thiel, N. Karl, B. Schelle, P. Disterer, I. Klagge, and S. G. Siddell Multigene RNA Vector Based on Coronavirus Transcription J. Virol., September 15, 2003; 77(18): 9790 - 9798. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. L. Ackerman and P. Cresswell Regulation of MHC Class I Transport in Human Dendritic Cells and the Dendritic-Like Cell Line KG-1 J. Immunol., April 15, 2003; 170(8): 4178 - 4188. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Bondanza, V. S. Zimmermann, G. Dell'Antonio, E. Dal Cin, A. Capobianco, M. G. Sabbadini, A. A. Manfredi, and P. Rovere-Querini Cutting Edge: Dissociation Between Autoimmune Response and Clinical Disease After Vaccination with Dendritic Cells J. Immunol., January 1, 2003; 170(1): 24 - 27. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. de Jersey, D. Carmignac, T. Barthlott, I. Robinson, and B. Stockinger Activation of CD8 T Cells by Antigen Expressed in the Pituitary Gland J. Immunol., December 15, 2002; 169(12): 6753 - 6759. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or on EL-4 cells without peptide (
). Spontaneous release
was <12% in the 5-h assay and < 28% after 15 h.
Representative data from one of three experiments.
