| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
School of Pharmacy, University of Southern California, Los Angeles, CA 90033
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Abs inhibited destruction of
infiltrated islets. Single islets were effective stimulators of IFN-
production by cloned CD8+ T cells, which varied >50-fold
depending on the degree of islet infiltration. This effect of the islet
mononuclear infiltrate could be mimicked by adding spleen cells to
NIT-1 cells, which augmented IFN- production above the level
stimulated by NIT-1 cells alone. The enhancing effect of spleen cells
could be attributed to their macrophage subpopulation and was not MHC
restricted, although recognition of islet Ag by cloned CD8+
T cells and subsequent islet destruction was restricted to islets
expressing H-2Db molecules. An inhibitor of inducible NO
synthase inhibited destruction of inflamed islets by cloned
CD8+ T cells. We propose that macrophages in inflamed
islets provide a form of bystander costimulation of ß cell-specific
CD8+ T cells. CD8+ T cells respond to Ag and
costimulation by producing IFN- that activates macrophages.
Activated macrophages facilitate islet destruction by CD8+
T cells through a NO synthesis-dependent
pathway. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although in special experimental systems individual cell types appear
to be sufficient for disease induction, most studies have shown that
the interaction of several cell types is essential for natural,
spontaneous disease development. Thus, while at later stages of disease
development CD4+ T cells appear to be sufficient
for disease transfer in NOD mice, blocking CD8+ T
cell function with anti-CD8 Abs at an early stage prevented
spontaneous disease development, and T cells from prediabetic mice no
longer transferred disease when depletion of CD8+
T cells was combined with anti-CD8 Ab treatment (11).
Likewise, monoclonal CD8+ T cells grown on islets
that express a B7 transgene could transfer disease on their own
(12), but disease transfer by naturally developing
polyclonal T cells requires both the CD4+ and
CD8+ subsets (6). In addition to T
cells, macrophages have been found to be essential for both the
development and the effector phases of anti-islet immunity.
Depletion of resident macrophages prevented disease development, and
inhibiting the entry of host macrophages into the islet environment
prevented transfer of disease by donor T cells (5).
Recruitment of macrophages into the islet environment by islet-specific
expression of a TNF-
transgene accelerated disease development in
neonatal mice (13). These studies suggest that all of
these cell types are essential for natural disease development and
appear to cooperate at various stages during disease development.
CD8+ T cells can, at least in principle, recognize and destroy ß cells directly, and primed CD8+ T cells would be expected to be independent from APCs in the local environment of islets. However, without costimulatory signals derived from APCs, the effect of CD8+ T cells is short lived, and APCs are thought to be essential not only for priming of naive CD8+ T cells but also for the maintenance of CD8+ T cell memory and CTL effector function in response to exogenous Ags (14, 15, 16, 17). Priming as well as maintenance stimulation are thought to occur in lymphoid organs such as lymph nodes or spleen. How CD8+ T cells become and remain activated during the development of autoimmune diabetes is still unresolved.
Ag presentation is the best known but probably not the only mechanism by which professional APCs may contribute to priming of T cells and maintenance of their activated state. Zinkernagel et al. have shown that fibroblasts can prime and activate CD8+ T cells provided that the fibroblasts are localized in lymphoid tissue but not outside it (18). In that study, the Ag was clearly presented by fibroblast MHC class I molecules, not by local professional APCs. The authors concluded that localization in lymphoid tissue is the key determinant of immunogenicity, not Ag presentation in the context of costimulatory activity by the presenting cell (19). Presumably some form of bystander costimulation by cells in close proximity or the cytokine milieu constitutes this environment. It is not clear whether cells in an inflammatory infiltrate cooperate in similar ways.
During the development of some organ-specific autoimmune diseases, the chronic inflammatory infiltrate begins to resemble organized lymphoid tissue. Locally assembled organized lymphoid tissue may obviate the need to transport autoantigen to distant lymphoid organs. In infiltrated islets, the lymphoid tissue would be located close to ß cells rather than ß cells or Ags released by them being transported to lymphoid tissue. Analogous to fibroblasts in lymphoid tissue, ß cells in the vicinity of the locally assembled infiltrate may activate CD8+ T cells without a requirement to shed Ag for presentation by phagocytic APCs. However, no direct evidence is available for a role of local islet-infiltrating accessory cells in the activation of CD8+ T cells.
Cells of the monocyte-macrophage-dendritic cell lineage not only
generate signals that control the function of T cells, but also respond
to T cell-derived signals. For example, T cell-derived IFN- may
activate local macrophages and up-regulate expression of costimulatory
activity and contribute to a local environment in which priming of T
cells can occur and an autoimmune response can be initiated and
maintained (20, 21). This mutual exchange of signals may
lead to a process of self-enhancing activation and tissue organization.
After an initial trigger, disease progression may be the result of an
ongoing process of self-organization of the lymphocytic infiltrate,
which would be determined by local as well as systemic factors. During
the destructive phase of insulitis, activated macrophages can produce
amounts of NO and other products that are toxic to ß cells
(22) or up-regulate expression of Fas and increase the
sensitivity of ß cells to killing by T cells (23).
However, it is not clear how the release of these macrophage products
is linked to the specific recognition of islet Ags by T cells.
In this work, we have investigated the role of islet-infiltrating
mononuclear cells in the activation and effector function of
CD8+ T cells. The inflammatory infiltrate greatly
enhanced both islet destruction and IFN- production by cloned
CD8+ T cells. In vitro reconstitution
experiments, in which spleen cells were added to NIT-1 cells, showed
that at least some of the enhancing effect of non-ß cells could be
attributed to bystander costimulation by macrophages. Macrophages were
also implicated in enhancing islet destruction by cloned
CD8+ T cells because inhibitors of the inducible
form of NO synthase protected inflamed islets from destruction by the
CD8+ T cell clones. We propose a novel two-way
signal exchange between macrophages and CD8+ T
cells that enhances and complements Ag-specific interaction of ß
cells and CD8+ T cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
NOD, NOD/Lt-scid/scid (NOD-scid), BALB.B, BALB/c, B10.BR, and C57BL/6 mice were obtained from The Jackson Laboratory (Bar Harbor, ME) and were bred and maintained in the University of Southern California Animal Facility under pathogen-free conditions. The spontaneous incidence of diabetes in our colony reaches 65–70% in female NOD mice by 20 wk of age, and diabetes usually commences by 13 wk of age. For experiments, 8- to 12-wk-old mice were used.
Reagents, Abs, and cell culture
Murine rIL-2 and rIFN-, Con A, and bacterial LPS
were obtained from Sigma (St. Louis, MO). Ionomycin,
PMA, and
L-N6-(1-iminoethyl)lysine · HCl
(L-NIL) were purchased from Calbiochem (La Jolla,
CA). Anti-mouse IFN- mAb (XMG1.2, rat IgG1), and murine rTNF-
were obtained from PharMingen (San Diego, CA). Murine rIL-7 was
obtained from Life Technologies (Gaithersburg, MD), and collagenase P
was obtained from Boehringer Mannheim (Indianapolis, IN). Hybridoma
YCD3–1 (anti-CD3
), A20 lymphoma cells, NXA cells, and CTLL-2
cells were kind gifts from Dr. C. A. Janeway, Jr. (Yale
University, New Haven, CT). J774A.1 cells (H-2d),
hybridomas 10-2.16 (anti-I-Ak), 30-H12
(anti-Thy1.2), H57-597 (anti-ßTCR), 3.155 (anti-CD8),
GK1.5 (anti-CD4), M1/70.15.11.5.HL (anti-Mac-1), N418
(anti-CD11c), and RA3-3A1/6.1 (anti-B220) were obtained from
American Type Culture Collection (Manassas, VA). Anti-CD28 mAb HM3500
and FITC-goat F(ab')2 anti-hamster IgG were
purchased from Caltag Laboratories (South San Francisco, CA), and goat
F(ab')2 anti-rat IgMµ and goat
F(ab')2 anti-rat IgGFc
were obtained from Accurate Chemicals (Westbury, NY). Pooled rabbit
complement was obtained from ICN Pharmaceuticals (Aurora, OH). RMA
(H-2d) cells were kindly provided by Dr. M.
McMillan (University of Southern California, Los Angeles, CA), and EL-4
(H-2b), P815 (H-2d), and
YAC-1 cells were kindly provided by Dr. G. Dennert (University of
Southern California). The ß cell line NIT-1
(I-Ag7; Kd,
Db) was kindly provided by Dr. E. H. Leiter
(The Jackson Laboratory), and TCX6310 cells were kindly provided by Dr.
F. Melchers (Basel Institute for Immunology, Basel, Switzerland).
Abs were used in the form of diluted cell culture supernatant or were purified from culture supernatant using GammaBind Plus Sepharose (Pharmacia Biotech, Piscataway, NJ) columns. The tissue culture medium (TCM) used for cell culture and all experiments was based on Click’s medium (Irvine Scientific, Santa Ana, CA), which was supplemented with 4 mM L-glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin (Life Technologies), 40 µM 2-ME, and 10% FBS (BioWhittaker, Walkersville, MD). For routine T cell culture, we used supernatant of TCX6310 cells (24) as a source of IL-2.
Islet isolation
Islets were prepared by collagenase digestion as described
previously (25) with minor modifications. Islets were
manually picked using a dissection microscope and cultured for 1–2
days in TCM. Unless indicated otherwise, only islets that had
intact borders after this culture period were used in experiments. For
some experiments, islets were treated with IFN- for 72
h.
Generation of an islet-specific CTL line
To obtain islet-destructive CTLs, spleen cells (5 x
105) from newly diabetic female NOD mice were
cultured with five islets in the presence of IFN- (10 U/ml), IL-2
(10 U/ml), and IL-7 (10 ng/ml) in 24-well plates. After 5 days, T cell
blasts surrounding disintegrating islets were picked up using a
pipette, pooled, and restimulated with islets 1 wk later. Four days
after restimulation, T cells were placed in cultures of adherent NIT-1
cells in the presence of cytokines as above. Cultures were checked
daily for NIT-1 cell destruction. Cells from positive wells were pooled
and stimulated again with islets and NIT-1 cells. We applied two cycles
of expansion alternating between islets and NIT-1 cells. Restimulated
cells, referred to as "polyclonal cytotoxic T cells (polyclonal
CTLs)," were used for cloning.
T cell cloning and maintenance of clones
For cloning and expansion, we stimulated T cells with
anti-CD3 mAb. Polyclonal cytotoxic T cells were seeded at a density
of 0.3–1 cells/well into 96-well round-bottom plates in TCM containing
anti-CD3 mAb (culture supernatant of hybridoma YCD3–1 diluted
1:20) and mitomycin C-treated NOD spleen cells (3 x
104 per well). The anti-CD3 mAb was removed
after 48 h by washing cells two to three times, and cell
proliferation was stimulated by resuspending cells in TCM supplemented
with IL-2 (40 U/ml) and IL-7 (10 ng/ml). After 48 h, cells were
fed once more with the same medium, and after that every 3 to 4 days
with TCM alone or, during every other feeding cycle, with TCM
supplemented with IL-2 and IL-7 (10 U/ml and 10 ng/ml, respectively). T
cells were stimulated in this way every 10–14 days. After the second
and the third stimulation, plates were checked for apparent growth of T
cells, and cells from positive wells were expanded further.
For routine maintenance, clones were restimulated with anti-CD3 mAb every 2–3 wk. T cells (1–5 x 105) were cultured with mitomycin C-treated spleen cells (5 x 106) in 5 ml of TCM in the presence of YCD3–1 cell culture supernatant diluted 1:20. After 48 h, cells were washed and treated with cytokines as described above. In all experiments T cell clones were used after removal of dead cells using lymphocyte separation medium (Organon Teknika, Durham, NC).
Disease transfer experiments
Newly diabetic 12- to 20-wk-old female NOD mice were used as donors of diabetogenic spleen cells, and 8-wk-old female NOD-scid mice were used as recipients. T cell clones (107), spleen cells (107), or T cell clones (107) mixed with spleen cells from diabetic mice (107) were suspended in 200 µl of HBSS and injected into the tail vein of recipients. Onset of disease was monitored by testing the urinary glucose level three times a week with Chemstrip uG (Boehringer Mannheim). Mice were considered diabetic from the first day of twice consecutive detection of glucosurea (100 mg/dL or higher).
Morphological detection of islet destruction
Islets were placed in 96-well flat-bottom assay plates together
with 1 x 105 clonal T cells. Anti-IFN-
mAb XMG1.2, the NO synthase inhibitor L-NIL, and the
cytokines IL-2, IFN-, and IL-7 were added as indicated. T cells were
used 5–6 d after stimulation. Islet morphology was assessed at regular
time intervals by phase contrast microscopy. Residual islet mass at
intermediary stages before complete destruction could be assessed after
removing the large cluster of T cells around them. To achieve this, the
T cells surrounding the islet were dispersed by pipetting the islet T
cell cluster up and down several times through a 100-µl pipette tip.
This had no effect on islet integrity in the absence of T cells. In
this assay, 
8 ± 3% (average ± SEM, n =
84) of untreated islets disintegrated spontaneously. In the presence of
IL-7 (10 ng/ml) and IL-2 (10 U/ml), islet-specific clones increased the
rate of disintegration to 88 ± 3% (n = 192),
whereas in the presence of irrelevant clones 5 ± 2%
(n = 35) of the islets disintegrated.
Stimulation of cytokine release from T cells
To determine their cytokine release profile, T cell clones were
stimulated simultaneously with anti-CD3 and anti-CD28 mAbs
immobilized in tissue culture plates (96-well flat-bottom plates
(Falcon, Becton Dickinson, San Diego, CA) coated with 10 µg/ml of
each Ab in PBS at 4°C overnight). Cloned T cells (5 x
104 in 100 µl TCM) were added to Ab-coated
plates, and, 24–48 h later, cell culture supernatant was collected for
assay of cytokines. In experiments where islets or cell lines were used
to stimulate IFN- release, clones were used 7–8 days after
stimulation. Cloned T cells (5 x 104) were
cultured with a single islet in 96-well round-bottom plates in the
presence of 10 U/ml IL-2. After 24–48 h, supernatant was collected for
assay of IFN-. In controls, only islets or only T cells were added
to wells. In experiments where cell lines or spleen cells were used as
stimulators, T cells (5 x 104) were added
to either 1 x 105 spleen cells or to 5
x 104 RMA, J774A.1, YAC-1, P815, or EL-4 cells
alone or in combination with 1 x 105
adherent NIT-1 cells. IL-2 was added at 10 U/ml as indicated in the
figure legends. The incubation time was 24–48 h. For maximal
stimulation of IFN- release from islet-infiltrating or cloned T
cells, single islets or cloned T cells (105/well)
were treated with Con A (4 µg/ml) or a combination of PMA (10 ng/ml)
and ionomycin (0.5 µM).
Cytokine assays
IL-2 activity in cell culture supernatant was measured using the
IL-2-dependent cell line CTLL-2. The assay was calibrated using rIL-2.
The sensitivity of the assay was 0.1 U/ml (0.02 ng/ml). The
concentrations of IFN- and TNF- in culture supernatants were
measured by sandwich ELISA, using paired anti-cytokine Abs
(PharMingen), following protocols recommended by the manufacturer. The
sensitivities of the assays were 1 U/ml (67 pg/ml) for IFN- and 0.4
U/ml (40 pg/ml) for TNF-.
Depletion of splenocyte subpopulations
Spleen cells were incubated with anti-MHC class II Ab or anti-Thy-1.2 Ab (supernatant from hybridomas 10-2.16 or 30-H12, respectively, diluted 1:2) for 30 min on ice, washed, and treated with diluted (1:10) pooled rabbit complement for 45 min at 37°C. To deplete adherent cells (macrophages and dendritic cells), 15 x 106 spleen cells were cultured in 60 x 15 mm tissue culture dishes in TCM for 2 h at 37°C. Nonadherent cells were collected after gentle pipetting.
Preparation of macrophages and B cell blasts
For preparation of macrophages, NOD or BALB/c mice were injected
with 1.5 ml 3% thioglycollate, and, 4 days later, peritoneal exudate
cells (PEC) were obtained by peritoneal lavage with PBS supplemented
with 20 U/ml heparin. PEC were cultured in TCM for 2 h at 37°C
to allow macrophages to adhere to the tissue culture plastic.
Nonadherent cells were removed by washing with culture medium. Adherent
cells contained 
90% of Mac-1+ cells as
determined by FACS analysis. B cell blasts were prepared by incubating
spleen cells with 10 µg/ml LPS for 48 h. After stimulation with
LPS, dead cells were removed by centrifugation through lymphocyte
separation medium. This preparation of B cell blasts contained 87%
B220+ cells, 9% Thy1.2+
cells, <1% Mac-1+, and <1%
CD11c+ cells.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To investigate how islet-reactive CD8+ T
cells interact with both ß cells and accessory cells that infiltrate
the islets of Langerhans during the development of IDDM, we isolated
CD8+ T cell clones that recognize and destroy
islets and an insulinoma cell line in vitro and contribute to disease
development in vivo. The insulinoma cell line NIT-1 (26)
was to be used for in vitro reconstitution experiments instead of
islets because it would provide a homogeneous and readily available
model for ß cells devoid of APCs. To first select islet-specific T
cells and then a subset of those T cells that also recognize NIT-1
cells, we expanded CD8+ T cells using an islet
preparation as antigenic stimulus, followed by expansion using the NOD
syngeneic insulinoma cell line NIT-1 as a source of Ag. Because we had
found that a combination of IL-2, IFN-, and IL-7 was optimal in
promoting islet destruction (data not shown), we used this combination
of cytokines to expand a polyclonal population of T cells. Using spleen
cells from a 12-wk-old newly diabetic NOD female mouse, we obtained a
polyclonal population of T cells, termed polyclonal CTLs, that potently
destroyed both islets and NIT-1 cells. Because islets and NIT-1 cells
did not support the growth of highly diluted T cells, we used an
anti-CD3 Ab together with splenocytes and IL-2 for cloning by
limiting dilution. The cloning efficiency of this protocol was as high
as 50%. We also used anti-CD3 mAb stimulation followed by
cytokine treatment to clone directly from islet infiltrate without
prior expansion on islets and NIT-1 cells.
To characterize the clones, we determined their expression of surface markers, stimulation-induced cytokine release profile, cytotoxic activity toward islets and NIT-1 cells, and ability to contribute to disease development in vivo. We also investigated whether they specifically recognized islet Ags in an MHC-restricted manner.
All clones expressed TCR and were single positive for either CD8 or CD4
(Table I
). The panel of clones that was
obtained after the expansion step on NIT-1 cells contained only
CD8+ T cells, whereas cloning without such prior
expansion yielded both CD4+ and
CD8+ T cells. Because NIT-1 cells express MHC
class I but not MHC class II molecules, this suggests that the
expansion step before cloning was driven by Ags presented by MHC class
I molecules.
|
, only two produced
IL-2, and none of them produced significant levels of TNF- when
stimulated with immobilized anti-ßTCR Ab or a combination of
anti-CD3 and anti-CD28 mAbs.
In vitro destruction of islets and NIT-1 cells was monitored using a
morphological assay (Fig. 1A).
T cells first accumulated around islets and enlarged while the border
of the islets remained intact. After an initial attack at the islet
perimeter, the islet was destroyed rapidly. Different islets in the
same well had different lag times for destruction. In most cases,
islets were destroyed over a time period of 48–72 h. A similarly slow
process of destruction was observed for NIT-1 cells.
|
.
The results show that clones can be classified into four groups, those
that destroyed 1) islets only, 2) NIT-1 cells only, 3) both, or 4)
neither. The third group was by far the largest, suggesting that NIT-1
cells present at least some Ags in common with islets. Interestingly,
14 of 18 clones that were obtained after expansion on both islets and
NIT-1 cells were cytotoxic to both, whereas this was true for only 1 of
11 clones that were derived from the islet lymphocytic infiltrate
without the prior expansion step. This striking difference together
with the shift toward CD8-positive T cells suggests again that
expansion on islets and NIT-1 cells is driven by Ag presented by MHC
class I molecules.
We next tested some of the clones from group 3, which are islet and
NIT-1 cell destructive, for the ability to influence disease
development in vivo. Clone 9B7 accelerated disease (Fig. 1B), as did clone 8F7 (not shown). Four other clones tested
had no effect on disease development.
NOD mice express the MHC class I molecules H-2Kd
and H-2Db. To determine the MHC restriction of
the clones, we used islets from different strains of mice sharing with
NOD mice either Kd, Db, or
neither of the two as targets. As shown in Fig. 1C, islets
from C57BL/6 mice (H-2b) were destroyed, whereas
islets from BALB/c mice (H-2d) were not,
indicating that the clones examined here were Db
restricted. Surprisingly, all of the other five clones examined showed
the same MHC restriction. To confirm that it is indeed the MHC locus
rather than some other polymorphism that distinguishes NOD and C57BL/6
mice on the one hand and BALB/c mice on the other, we also tested
islets prepared from the BALB/c congenic strain BALB.B. In
contrast to BALB/c islets, BALB.B islets were susceptible to killing by
clones 9B7 and 8F7. A summary of the results of many more experiments
is shown in Fig. 1D. Although these data clearly confirm
that islet destruction is Db restricted, our data
also show that, on average, NOD islets are destroyed more effectively
than islets from other strains that express Db, a
finding that cannot be explained solely by allelic differences in the
MHC locus.
In separate experiments, we confirmed that activation of killing activity by clone 8F7 is not only MHC restricted but also islet specific. Thus, neither NOD fibroblasts nor NOD Con A blasts were destroyed. Furthermore, several non-ß cell lines expressing H-2Db, such as A20 cells fused with NOD splenocytes (NXA cells) and EL-4 cells, were not destroyed (data not shown). Islet destruction was contact dependent, as supernatant from cloned CD8+ T cells taken during an islet destruction assay or collected from cells activated by anti-TCR Abs had no destructive effect (not shown).
We selected clones 9B7 and 8F7 for experiments designed to investigate the role of both ß cells and accessory cells within the islet inflammatory infiltrate in the activation and islet-destructive function of CD8+ T cells.
The activation and islet-destructive function of cloned CD8+ T cells are influenced by the degree of mononuclear cell infiltration in islets
Does the mononuclear cell infiltrate influence the
islet-destructive function of CD8+ T cells? The
large variation of the severity of lymphocytic infiltration, observed
between islets even within a single animal before the onset of disease,
can be exploited to address this. To asses the severity of
inflammation, we placed islets at low density into tissue culture
dishes, cultured them for 16 h, and monitored the number of
mononuclear cells emanating from them. We separated the islets into
three groups: those that did not show any sign of infiltration, those
that showed an intermediate degree of infiltration, and those that were
severely infiltrated as indicated by having none, between 1 and 10, or
>10 cells around them, respectively. We also used islets from NOD-scid
mice as these should be free from any inflammatory infiltrate. Cloned T
cells were added to islets from different groups, and, after a further
culture period, islet destruction was assayed. As shown in Fig. 2A, infiltrated islets were
destroyed faster than islets that did not show signs of infiltration.
The latter group may still contain some level of infiltration that is
not detected by our scoring method. Indeed, destruction of islets
prepared from NOD-scid mice required an even longer time. In all cases,
destruction was clearly an effect of the added clones, because even
heavily infiltrated islets only rarely underwent spontaneous
destruction. Without added clones, infiltrated islets lost infiltrating
cells over time and damaged islets recovered in culture.
|
that fully
up-regulated MHC class I expression and compared the speed of islet
destruction with that of untreated NOD-scid and that of NOD islets with
or without signs of infiltration. The results (Fig. 2A) show
that pretreatment with IFN- increased the speed of islet destruction
to that of noninfiltrated NOD islets, but not to that of infiltrated
islets. These results suggested that an increase of Ag presentation by
ß cells is not likely be the only effect of the infiltrate.
For the experiments shown in Fig. 2A, both IL-2 (10 U/ml)
and IL-7 (10 ng/ml) were included in the islet-destruction assay
because they had been found to optimize killing during cloning. Without
exogenously added IL-7 and at a reduced level of IL-2 (2.5 U/ml), the
effect of the infiltrate was essential, not just accelerating (Fig. 2B). NOD-scid islets were no longer destroyed, whereas the
majority of heavily infiltrated islets were. As in Fig. 2A,
NOD islets that did not show obvious signs of infiltration were
destroyed to a lesser degree than heavily infiltrated islets.
Islet destruction is a complex consequence of T cell activation.
Although clearly an effect of cloned CD8+ T
cells, islet destruction could also involve accessory cells in the
islet infiltrate. To have a second and more direct assay for activation
of CD8+ T cells, we used IFN- release as a
marker. Fig. 3 shows that it is possible
to measure IFN- release from cloned CD8+ T
cells triggered by single islets, an observation that facilitated the
use of the large heterogeneity of the magnitude of the mononuclear
infiltrate for studies of its influence on IFN- release from
CD8+ T cells. The signal-to-background ratio of
IFN- release varied from 1:1 to close to 50:1 relative to controls
containing T cells but no islets. The IFN- release data confirmed
that the infiltrate plays an important role in the activation of T
cell clones. Thus, the most severely infiltrated islets stimulated more
IFN- release from clone 8F7 than moderately infiltrated islets
(p < 0.02), which in turn were more effective
than NOD islets without signs of infiltration
(p < 10-5). In accord
with the islet-destruction experiments, NOD islets with no signs of
infiltration still stimulated significantly more IFN- release than
NOD-scid islets (p <
10-6). NOD-scid islets stimulated very little
IFN- release from cloned CD8+ T cells,
suggesting that ß cells alone may not be sufficient to stimulate high
levels of IFN- release.
|
release during coculture of infiltrated islets with cloned
CD8+ T cells came from the clones and not from
cells in the islet infiltrate because even heavily infiltrated islets
released only low levels of IFN- on their own (Fig. 3, 
), and
infiltrated islets did not have the capacity to produce sufficient
amounts of IFN-, even when stimulated with potent T cell activators
(10 ng/ml PMA + 0.5 µM ionomycin, 10 ± 2 U/ml; 4 µg/ml Con A, 2 ±
1 U/ml). In contrast, in the same experiment, these agonists potently
triggered IFN- release from cloned CD8+ T
cells (PMA + ionomycin, 24,081 ± 2,825 U/ml; Con A, 2,143 ± 551 U/ml;
anti-TCR mAb, 13,075 ± 684 U/ml), and, during a coculture of
clones with islets, run in parallel, 194 ± 20 U/ml were released. The enhancing effect of the islet infiltrate was a general property of islet-reactive CTLs, not just a characteristic of the clones studied here, because it could be reproduced with the polyclonal islet-reactive CD8+ T cells (data not shown).
Although the severity of mononuclear infiltration was a highly
significant determinant of an islet’s ability to stimulate IFN-
release from exogenously added cloned T cells, it was not the only
factor. Even within one group, the magnitude of IFN- release varied
greatly between individual islets, suggesting that not only the number
of infiltrating cells but also their composition and/or organization
may be important.
As above for islet destruction, we tested whether the effect of the
islet infiltrate on IFN- release from cloned
CD8+ T cells was due to up-regulation of MHC
class I molecules on ß cells. We pretreated NOD-scid islets and
MHC-mismatched BALB/c islets with IFN- and used both for stimulation
of cloned CD8+ T cells. IFN- pretreatment of
NOD-scid islets clearly increased their effectiveness
(p < 10-9) to stimulate
IFN- release from cloned CD8+ T cells (Fig. 3). This increase was not observed with MHC-mismatched BALB/c islets,
suggesting that it is indeed mediated by NOD MHC molecules and not due
to some other, nonspecific effect of IFN-. Although clearly
detectable, the effect of IFN- on NOD-scid islets does not fully
account for the much stronger effect of the islet infiltrate. Some or
all of the cells that constitute the mononuclear islet infiltrate
appear to play an important role other than merely up-regulating of MHC
class I molecules on ß cells.
The effect of the mononuclear infiltrate was not due to some soluble factor acting on cloned T cells because supernatant collected from inflamed islets had no effect (data not shown).
We asked whether the ability of CD8+ T cell
clones to destroy islets and to produce IFN- are linked. Fig. 2B shows that an anti-IFN- Ab reduced the ability of
cloned CD8+ T cells to destroy islets in vitro.
An isotype-matched control Ab had no effect. This result indicates that
IFN- plays an important role in islet destruction in our system.
Accessory cells from the islet infiltrate stimulate cloned CD8+ T cells directly
The results described above suggest that accessory cells in the
islet infiltrate play a role in enhancing the activation of cloned
CD8+ T cells added to islets. To demonstrate a
direct effect of accessory cells, we collected cells that had emanated
from infiltrated islets during overnight culture and assayed their
ability to stimulate cloned CD8+ T cells. The
data in Fig. 4 demonstrate the efficacy
of infiltrate cells. However, the data also show that infiltrate cells
alone are not as potent as infiltrated islets. As many as 25,000 cells
collected from 10 infiltrated islets stimulated only 30 U/ml,
whereas many of the single islets with infiltrate stimulated between
100–200 U/ml. This observation suggests that accessory cells within
the infiltrate are either more numerous or more effective activators of
cloned CD8+ T cells, or that infiltrating cell in
combination with ß cells are more effective than ß cells alone or
infiltrating cells alone.
|
release from CD8+ T cell clones
We have demonstrated that the severity of lymphocytic infiltration
in islets correlates with their ability to stimulate IFN- release
from cloned CD8+ T cells. In the following
experiments, we attempted to simulate the effect of accessory cells in
the islet infiltrate by adding mononuclear cells to a ß cell model in
vitro. Although NOD-scid islets should be completely free from a
lymphocytic infiltrate, they may still contain nonlymphoid accessory
cells such as dendritic cells and macrophages, whereas NIT-1 cells are
a homogeneous ß cell preparation devoid of any APCs. We used spleen
cells as a model for mononuclear cells. To test a cooperative effect of
NIT-1 cells and mononuclear cells, clones 9B7 and 8F7 were stimulated
with NIT-1 cells in the presence or absence of splenocytes. The effect
of splenocytes in combination with NIT-1 cells was not only more than
that of NIT-1 cells or splenocytes alone, but also more than the sum of
both (Fig. 5). Although both clones
showed similar qualitative responses to the combination of NIT-1 cells
and splenocytes, 9B7 cells responded better to splenocytes alone than
to NIT-1 cells alone and the reverse was true for 8F7 cells. Despite
these quantitative differences, there was clear cooperative effect
between NIT-1 cells and splenocytes for both clones.
|
The following experiments were conducted to investigate the mechanism
by which splenocytes cooperated with NIT-1 cells to stimulate IFN-
release from cloned CD8+ T cells and how
splenocytes may stimulate CD8+ T cells on their
own. For both situations, we considered Ag presentation, costimulation,
and a contribution of splenic T cells.
The effect of spleen cells on IFN- release is not MHC restricted
Could the cooperative effect between NIT-1 and spleen cells be due
to presentation of NIT-1 cell Ags by spleen cells? There is precedence
that phagocytic cells such as macrophages and dendritic cells, both of
which are constituents of splenocytes, can present exogenous cell-bound
Ags to CD8+ T cells (15). Even
stimulation of IFN- release by NOD spleen cells alone is, at least
in principle, compatible with this notion as it may reflect uptake and
presentation of ß cell Ag in vivo before the experiment and retention
of specific peptide-MHC complexes for the duration of the experiment.
If Ag presentation is the mechanism of enhancement, it should be MHC
restricted. To test this, we assayed the ability of splenocytes from
various strains of mice to enhance NIT-1 cell-triggered IFN- release
from clones 9B7 and 8F7. Surprisingly, splenocytes from strains NOD
(H-2g7), BALB/c (H-2d),
B10.BR (H-2k), and BALB.B
(H-2b) all had an enhancing effect (Fig. 6A), arguing against Ag
presentation as the sole mechanism for the enhancing effect of
splenocytes.
|
release on
their own did not require expression of NOD-specific MHC alleles (Fig. 6B). For all strains of mice used, the NIT-1
cell-independent stimulatory effect of spleen cells was greatly
enhanced by exogenous IL-2 and to a large extent depended on
it.
Taken together, Ag presentation can be ruled out as the sole or even
major mechanism for the in vitro ability of spleen cells to enhance
IFN- release from CD8+ T cell clones. It
appears that spleen cells provide a stimulatory signal other than Ag
that can trigger IFN- release on its own but at the same time
cooperates with, and enhances, Ag-specific stimulation of cloned
CD8+ T cells by NIT-1 cells. In the absence of
Ag, prior activation or IL-2 appear to be required in lieu of
simultaneous stimulation through the TCR.
Depletion of MHC class II-expressing and adherent cells from spleen
cells reduces their effectiveness to enhance and stimulate IFN-
release from CD8+ T cell clones
We next asked which subpopulation within spleen cells might
contribute to their cooperative effect with NIT-1 cells or their
independent stimulatory effect. As shown in Fig. 7A, removing T cells from
splenocytes, before adding them to cloned T cells together with NIT-1
cells, reduced the cooperative effect, but did not eliminate it. This
reduction is probably due to some T cell-derived cytokine, such as
IL-2, that enhances IFN- production. In contrast, removing adherent
cells abolished the enhancing effect. This result suggests that either
macrophages and/or dendritic cells account for the stimulatory effect
of spleen cells because removal of adherent cells led to a drastic
decrease, as determined by flow cytometry, of the number of dendritic
cells and macrophages (CD11c+ from 4.0 ±
0.2% to 1 ± 0.4%, Mac-1+ from 4.1 ±
0.6% to 1 ± 0.4%, respectively), but did not change the number
of B cells (B220+ from 33 ± 5% to 31
± 2%) or T cells (Thy-1.2+ from 50 ± 3%
to 51 ± 3%).
|
B). As an alternative approach to depletion by adherence,
we removed APCs, as defined by their expression of MHC class II
molecules, from spleen cells before adding them to cloned
CD8+ T cells. In controls, we removed T cells
instead of APCs. The data (Fig. 7C) show that MHC class
II-expressing cells but not T cells account for a large part of the
enhancing effect of the total spleen cell population. Conversely, B
cell blasts had no effect (Fig. 7D), not only confirming the
specificity of the clones but again suggesting that MHC class
II-expressing cells remaining after excluding B cells, i.e.,
macrophages and/or dendritic cells, account for the independent
stimulatory effect of splenocytes.
The experiments demonstrating that the effect of spleen cells on
IFN- release from CD8+ T cells is not MHC
restricted have ruled out Ag presentation but not alloreactivity as the
underlying mechanism. However, in contrast to B10.BR spleen cells,
B10.BR B cell blasts had no effect (not shown). We also compared the
effect of the macrophage cell line J774A.1 and T cell line RMA.
Although J774A.1 and RMA cells express the same MHC haplotype
(H-2d) as BALB/c spleen cells, RMA cells were
ineffective (Fig. 8A).
Therefore, alloreactivity is unlikely to account for the effect of
H-2k- or H-2d-expressing
spleen cells, or of J774A.1 cells. Several other nonmacrophage cell
lines were equally ineffective, such as the T cell lines EL-4
(H-2b) and YAC-1, as well as the mastocytoma line
P815 (H-2d) (not shown).
|
release. This rules out alloreactivity to non-NOD MHC molecules
expressed by J774A.1 cells as the mechanism of stimulation.
Because J774A.1 is a clonal cell line of macrophage origin, these
experiments also suggest again that macrophages may contribute to the
stimulation of IFN- release from cloned CD8+ T
cells. To test this more directly, we confirmed these results with a
bona fide macrophage preparation, adherent PEC. PEC macrophages were
effective stimulators of IFN- release from both clone 8F7 and 9B7
(Fig. 8B). Interestingly, this was also true for BALB/c
macrophages, confirming that this effect is not MHC restricted.
Analogous to the enhancing and the independent stimulatory effects of
spleen cells, clone 9B7 responded better to these macrophages than
clone 8F7. To ensure that the response to macrophages was
representative of NOD CD8+ T cells and not
restricted to the clones used in this study, we repeated the
experiments with polyclonal islet-reactive CTLs. Again NOD macrophages
as well as BALB/c macrophages effectively stimulated IFN- release
from polyclonal CD8+ T cells (Fig. 8B).
CD8+ T cells recruit an inducible NO synthase-dependent pathway in islet destruction
The in vitro reconstitution experiments above have identified
macrophages to be important contributors to the activation of IFN-
release from cloned CD8+ T cells. Islet
destruction by CD8+ T cells was blocked by an Ab
to IFN- and was dependent on a pre-existing inflammatory infiltrate.
Because IFN- is a potent activator of macrophages, a component in
the inflammatory infiltrate, it is possible that macrophages play a
role not only in costimulating CD8+ T cells but
may also cooperate with CD8+ T cells in islet
destruction. Indeed, if macrophages costimulate IFN- production by
CD8+ T cells in situ, they would be in close
proximity of IFN--producing CD8+ T cells.
Activated macrophages produce several factors that may contribute to
ß cell destruction. These include NO, IL-1, and TNF-
(22). In rodents, IL-1 alone (27) or in
combination with IFN- or TNF- (22, 27), stimulates
ß cells to synthesize NO. Because macrophages are the only source of
IL-1 in islets (22, 28, 29), both NO synthesis by
macrophages or by ß cells is indicative of macrophage activation. To
implicate macrophages as partners in islet destruction by cloned
CD8+ T cells, we returned to the islet
destruction assays and used an inhibitor of the inducible form of NO
synthase, L-NIL (30). Fig. 9A shows that this inhibitor
protected inflamed islets from being destroyed by
CD8+ T cells. This inhibition was observed at a
dose range where selective inhibition of the inducible form of NO
synthase occurs. The same observations were made using polyclonal CTLs,
indicating that the participation of a NO synthesis in islet
destruction is not restricted to the clones. In the presence of IL-7,
the clones destroyed islets without inflammation also, such as NOD-scid
islets (Fig. 9B). This type of killing was much less
sensitive to inhibition by the NO synthesis inhibitor, suggesting that
there are two pathways of islet destruction by cloned
CD8+ T cells. In the presence of IL-7, the clones
can kill directly, without the participation of macrophages. Without
IL-7, the clones recruit a NO-dependent pathway. Most likely, IFN-
released from activated clones activates local macrophages, which
contribute to islet destruction either by producing NO themselves or by
releasing IL-1, which up-regulates the inducible form of NO synthase in
ß cells.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
release. A second major finding is that
accessory cells costimulate CD8+ T cells, and we
have identified macrophages to account for this effect. A third finding
is that CD8+ T cells, when destroying inflamed
islets, are able to recruit a NO synthesis-dependent pathway.
The enhancement of islet destruction by the islet inflammatory
infiltrate may in part be due to an increase in activation of cloned
CD8+ T cells. The increase of IFN- production
in response to infiltrated islets is clearly an indicator of such
increased activation. While information about the role of
islet-infiltrating accessory cells in the activation of
CD4+ T cells is accumulating (13, 31), evidence for a role of local accessory cells in inflamed
islets in the activation of CD8+ T cells has not
been reported before.
What could be the mechanism for the influence of inflammatory cells on
activation of CD8+ T cells? Such mechanisms could
be classified as direct or indirect, depending on whether infiltrating
cells interact directly with cloned T cells or modify the interaction
of ß cells with cloned T cells. A third possibility is that IFN-
release from cloned T cells is maximal only in case of a tripartite
interaction where CD8+ T cells interact with both
ß cells and infiltrating cells either simultaneously or
consecutively.
Several of the possible indirect effects, such as up-regulation of MHC
class I molecules or adhesion molecules on ß cells would be mediated
by IFN- released from infiltrating mononuclear cells. However,
pretreatment of islets with IFN- did not reproduce the effect of the
inflammatory infiltrate. Induction of the release of soluble factors
from ß cells, another potential indirect effect, is equally unlikely
to be the key mechanism of the infiltrate because supernatant from
inflamed islets had no effect on cloned CD8+ T
cells. Other indirect effects relating to islet tissue integrity are
also possible. For example, ß cell death may create gaps in the islet
tissue that facilitate access of cloned T cells to ß cells. However,
dissociated ß cells (data not shown) and NIT-1 were very ineffective
stimulators of CD8+ T cells, suggesting that
access to ß cells was not a severe limitation in the activation of
cloned CD8+ T cells.
A contribution of direct effects, in which the infiltrating mononuclear
cells themselves contribute to activation of CD8+
T cell clones, was suggested by our experiments with spleen cells and
cells collected from the infiltrate. Cells of the
monocyte-macrophage-dendritic cell lineage are most likely to
account for these effects because acute depletion of MHC class
II-expressing cells and strongly adherent cells from splenocytes
drastically reduced stimulation of IFN- production by cloned T
cells. Based on these data we cannot differentiate between macrophages
and dendritic cells. In this regard, it is noteworthy that macrophages
and dendritic cells have recently been shown to belong to a common
lineage of phagocytic cells that originates from circulating monocytes
and branches upon tissue entry and contact with Ag (32).
In principle, these cells could present islet Ag locally. Regardless of
the contributing mechanisms, inflamed single islets were far more
effective stimulators of IFN- release than mononuclear cells
collected from the infiltrate of many islets, suggesting that Ag
presentation or other direct effects do not fully account for the
enhancing effect of the infiltrate.
Taken together, neither indirect nor direct effects alone fully account
for the enhancing effect of the infiltrate to activate cloned
CD8+ T cells. In quantitative terms, even the sum
of the effect of ß cells (or NIT-1 cells) alone and that of
infiltrating mononuclear cells alone did not account for the ability of
inflamed islets to stimulate IFN- release from cloned
CD8+ T cells. The data strongly suggest that ß
cells and accessory cells in the infiltrate cooperate in a more than
additive way in stimulating CD8+ T cells.
The in vitro reconstitution experiments with spleen cells, PEC, and a
macrophage cell line were designed to identify candidate mechanisms
that may underlie the cooperative effect of ß cells and infiltrating
cells. These experiments have revealed an unexpected function of
accessory cells that is, at least in part, most likely due to some form
of bystander costimulation by macrophages. Spleen cells, when added to
NIT-1 cells, appear to costimulate IFN- production by cloned
CD8+ T cells. This effect did not depend on
expression of NOD MHC class I molecules although the
CD8+ T cell clones could recognize ß cell Ag in
an MHC-restricted way nor was it due to alloreactivity of
MHC-mismatched spleen cells. These findings suggests that mechanisms
other than Ag presentation may underlie the enhancing effect of spleen
cells. Bystander costimulation is a remaining possibility. A similar
conclusion was reached by Shimizu et al. (33) based on the
fact that spleen cells expressing mismatched MHC alleles enhanced
proliferation of NIT-1 cell-specific CD8+ T cells
as well as NOD spleen cells did. The cell type and the nature of the
costimulatory factor was not resolved in the work by Shimizu et al.
(33). In this paper, we have provided evidence that
macrophages can provide bystander costimulation. Among the known
costimulators, B7-2 is unlikely to be a candidate molecule mediating
this effect because B cell blasts expressed B7-2 but had no effect.
Similar statements cannot be made for B7-1 because the expression
levels were too low.
Spleen cells and PEC can, at least in the presence of IL-2, stimulate cloned CD8+ T cells without a source of ß cell Ags. However, this effect required prior activation of CD8+ T cells through their TCR, suggesting that accessory cells provide costimulatory signals but these do not need to be delivered simultaneously with signal one.
In this study, we have used IFN- as a marker of
CD8+ T cell activation. Our observation that
anti-IFN- Abs inhibited islet destruction suggests that IFN-
may play a role also in their islet-destructive effector function.
Although IFN- is necessary for islet destruction in our system, it
is not sufficient because recombinant IFN- did not destroy cultured
islets without cloned CD8+ T cells during the
time scale of our experiments (<72 h, data not shown). This was true
even if the islets were heavily infiltrated. Others have found that
IFN- is toxic to ß cells, but only in combination with other
cytokines or at very high concentrations (34). Those and
our present data suggest that ß cells are not likely to be the only
target of IFN- action.
IFN- is a potent coactivator of macrophages. Activation of
macrophages is thought to be a major mechanism by which
IFN--producing CD4+ T cells contribute to ß
cell damage during the development of diabetes. By releasing IFN-,
Th1 cells are thought to activate macrophages to produce NO and other
oxidants that are toxic to ß cells. Our results suggest that a
similar interaction may occur between macrophages and
CD8+ T cells, because a blocker of inducible NO
synthase inhibited destruction of inflamed islets by
CD8+ T cells. Although ß cells can also express
the inducible form of NO synthase, this again requires products of
activated macrophages, such as IL-1. Therefore, our data suggest a
second pathway of signal exchange between CD8+ T
cells and macrophages. Thus, macrophages may receive signals from
CD8+ T cells and participate in their islet
destructive function. The model of a tripartite interaction between ß
cells, CD8+ T cells, and macrophages described
above can account not only for costimulation of IFN- release from
CD8+ T cells, but also in their islet-destructive
function. NO, produced either by activated macrophages themselves or by
ß cells in response to macrophage products, may increase the
sensitivity of ß cells to killing by CD8+ T
cell, for example by up-regulating Fas on ß cells (23). In this
model, CD8+ T cells still need to recognize Ag
presented by ß cells, but are fully effective only if they are also
exchanging signals with macrophages. Our CD8+ T
cell clones can clearly recognize and kill ß cells directly as
indicated by their ability to destroy NOD-scid islets in the presence
of IL-7. IL-7 appears to have an enhancing effect that can substitute
the enhancing effect of the infiltrate.
Our data on the role of IFN- in islet destruction and
IFN--mediated signal exchange between CD8+ T
cells and macrophages agree well with studies by others who found that
IFN- is an important cytokine in the development of type I diabetes.
Although earlier studies had suggested that knockout of the IFN-
gene does not protect against disease (35), recent
evidence from IFN- receptor knockout mice strongly suggests that
IFN- receptor signaling is essential for disease development and
impacts at several stages of disease development (36).
Expression of receptors for IFN- on ß cells may not be required
(37), but these receptors are essential on non-ß cells
for disease development (36). Several studies have shown
that blocking IFN-, either by using anti-IFN- Abs or soluble
IFN- receptor postnatally, prevented natural disease development and
adoptive transfer of disease in NOD mice (38, 39), and
regulation of IFN- synthesis by IL-18 was found to be abnormal in
NOD mice (40). By activating local macrophages, IFN-
may not only recruit their destructive potential, but also up-regulate
their expression of costimulatory activity and contribute to a local
environment in which activation of T cells can occur and an autoimmune
response can be maintained (20, 21).
While it is well established that IFN- plays an important role in
disease development, it is not clear which cells produce IFN- in the
local environment of islets. The data in this paper strongly suggest
that CD8+ T cells are, alongside Th1 cells,
candidates for IFN- production. A similar suggestion as to the role
of CD8+ cells as local producers of IFN- has
been made by Rabinovitch et al. (41). In biobreeding rats,
depletion of CD8+ T cells prevented up-regulation
of IFN- during disease development, and disease incidence was
reduced and its onset was delayed (42).
Although our data showing that infiltrated islets from NOD mice are much better stimulators of CD8+ T cells than those that do not contain an infiltrate may in part be explained by the increase in the number of APCs, it is possible that the mere presence of different types of APCs found in islets is not sufficient. Indeed, even heavily infiltrated islets still showed a large variation of stimulatory capacity. We propose that the exact composition or the organization of the infiltrate may also be important. The architecture of the mononuclear infiltrate may bring macrophages in close proximity to ß cells, which may allow macrophages to provide costimulatory signals to CD8+ cells within the islet environment and receive activating signals from them and contribute to ß cell destruction.
The ability to measure the efficacy of single islets to stimulate
IFN- release from cloned T cells will facilitate the identification
of tissue architectural elements that determine functional
immunogenicity of the islet environment. In future experiments, cloned
CD8+ T cells will serve as probes for the
detection of changes in the islet environment that precede the
development of destructive insulitis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Hermann von Grafenstein, School of Pharmacy, University of Southern California, 1985 Zonal Avenue, Los Angeles, CA 90033. E-mail address:
3 Abbreviations used in this paper: IDDM, insulin-dependent diabetes mellitus; NOD, nonobese diabetic; NOD-scid, NOD/Lt-scid/scid; TCM, tissue culture medium; L-NIL, L-N6-(1-iminoethyl)lysine · HCl; PEC, peritoneal exudate cells.
Received for publication May 20, 1998. Accepted for publication September 9, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
in neonatal NOD mice promotes diabetes by enhancing presentation of islet antigens. Immunity 9:733.[Medline]
. Nature 346:844.[Medline]
. J. Clin. Invest. 95:486.
and tumor necrosis factor-: cytotoxicity to murine islets of Langerhans. J. Immunol. 141:2325.[Abstract]
-interferon delays but does not prevent diabetes in NOD mice. Diabetes 45:812.[Abstract]
impacts at multiple points during the progression of autoimmune diabetes. Proc. Nat. Acad. Sci. USA 94:13844.. J. Exp. Med. 189:1053. gene expression in pancreatic islet-infiltrating mononuclear cells correlates with autoimmune diabetes in nonobese diabetic mice. J. Immunol. 154:4874.[Abstract]
gene expression in pancreatic islets and autoimmune diabetes development in biobreeding rats. J. Autoimmun. 12:109.[Medline]
This article has been cited by other articles:
|
|
 |
|
  M. K. Mohammad, M. Morran, B. Slotterbeck, D. W. Leaman, Y. Sun, H. v. Grafenstein, S.-C. Hong, and M. F. McInerney Dysregulated Toll-like receptor expression and signaling in bone marrow-derived macrophages at the onset of diabetes in the non-obese diabetic mouse Int. Immunol., July 1, 2006; 18(7): 1101 - 1113. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Lieberman, T. Takaki, B. Han, P. Santamaria, D. V. Serreze, and T. P. DiLorenzo Individual Nonobese Diabetic Mice Exhibit Unique Patterns of CD8+ T Cell Reactivity to Three Islet Antigens, Including the Newly Identified Widely Expressed Dystrophia Myotonica Kinase J. Immunol., December 1, 2004; 173(11): 6727 - 6734. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Fallarino, R. Bianchi, C. Orabona, C. Vacca, M. L. Belladonna, M. C. Fioretti, D. V. Serreze, U. Grohmann, and P. Puccetti CTLA-4-Ig Activates Forkhead Transcription Factors and Protects Dendritic Cells from Oxidative Stress in Nonobese Diabetic Mice J. Exp. Med., October 18, 2004; 200(8): 1051 - 1062. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Gurlo and H. v. Grafenstein Antigen-independent cross-talk between macrophages and CD8+ T cells facilitates their cooperation during target destruction Int. Immunol., September 1, 2003; 15(9): 1063 - 1071. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. Grohmann, F. Fallarino, R. Bianchi, C. Orabona, C. Vacca, M. C. Fioretti, and P. Puccetti A Defect in Tryptophan Catabolism Impairs Tolerance in Nonobese Diabetic Mice J. Exp. Med., July 7, 2003; 198(1): 153 - 160. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Trembleau, G. Penna, S. Gregori, N. Giarratana, and L. Adorini IL-12 Administration Accelerates Autoimmune Diabetes in Both Wild-Type and IFN-{gamma}-Deficient Nonobese Diabetic Mice, Revealing Pathogenic and Protective Effects of IL-12-Induced IFN-{gamma} J. Immunol., June 1, 2003; 170(11): 5491 - 5501. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Williams, J. Trambley, J. Ha, A. B. Adams, M. M. Durham, P. Rees, S. R. Cowan, T. C. Pearson, and C. P. Larsen Genetic Characterization of Strain Differences in the Ability to Mediate CD40/CD28-Independent Rejection of Skin Allografts J. Immunol., December 15, 2000; 165(12): 6849 - 6857. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Ganapathy, T. Gurlo, H. O. Jarstadmarken, and H. von Grafenstein Regulation of TCR-induced IFN-{gamma} release from islet-reactive non-obese diabetic CD8+ T cells by prostaglandin E2 receptor signaling Int. Immunol., June 1, 2000; 12(6): 851 - 860. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, The mean of IFN-
) and dendritic cells (DC) abolishes the
cooperative effect of spleen cells and NIT-1 cells (A)
as well as the IL-2-dependent stimulatory effect of spleen cells
(B) on IFN-
), M
) or B cell blasts (
). IL-2 was added to the
culture at 10 U/ml (B and D, closed
symbols; C, as indicated). All experiments were done
using clone 8F7.
