| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Institute of Pathology, Division of Immunopathology, University of Bern, Bern, Switzerland
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ß+CD8+ IEL exert only minimal
virus-specific cytotoxicity, maximum specific killing mediated by
TCRß+CD8ß+ IEL on day 8 postinfection
exceeds maximum cytotoxic activity observed with CD8 splenocytes when
assessed in vitro. Maximum cytotoxic activity of IEL is preceded by
peak perforin and granzyme B mRNA expression in IEL around day 6
postinfection, suggesting a recent activation in situ. The antivirus
cytotoxicity of in vivo primed IEL is further demonstrated by the
protection from virus production in the spleen of mice infected with
LCMV 10 h before adoptive cell transfer. These data indicate a
potent priming of LCMV-specific IEL in situ after systemic LCMV
infection and suggest that cytotoxic IEL markedly contribute to the
elimination of virus-infected cells in the intestinal
mucosa. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ß or a TCR
and
show a strong bias toward cell surface expression of CD8, either as a
CD8 homodimer or a CD8ß heterodimer. Due to their potential
exposure to a vast array of Ags, including food Ags and commensal
bacteria, IEL of the gut are thought to be controlled by distinct
immunoregulatory pathways (reviewed in Refs. 2 and
3). IEL have been reported to respond poorly
to mitogens or IL-2, but to produce a vast array of regulatory
cytokines, such as IL-4, IL-5, and IFN- (1, 4). This
leads to the suggestion that IEL exert primarily immunomodulatory
functions. On the other hand, however, ex vivo isolated IEL have
repeatedly been reported to exert constitutive cytotoxic activity in
vitro in both mouse (5, 6, 7, 8) and man (9).
The generation of Ag-specific IEL that exert cytotoxic activity upon
isolation ex vivo has been reported in different systems, including
rotavirus (10, 11), Toxoplasma gondii
(12), and lymphocytic choriomeningitis virus (LCMV).
(13). LCMV is a noncytopathic single-stranded RNA virus
from the arena virus family (14). Several different
strains of LCMV have been isolated, some of which cause systemic
infection by interacting with a broad variety of cells, possibly, as
has been shown in vitro, via the highly conserved membrane protein
-dystroglycan (15). Upon infection of its natural
host, the mouse, the virus replicates rapidly to a maximal virus load
around day 4 postinfection. Thereafter, it is efficiently cleared
by virus-specific CD8+ CTL in an immune-competent
host (16, 17). Upon adoptive transfer into recently
LCMV-infected recipients, splenic T cells from acutely infected donor
mice are able to protect the recipients from further proliferative
expansion of the virus (18, 19), thus demonstrating the
central role of T cells in the control of LCMV infections in the
mouse.
In the present study we attempted to determine the relative potential
of IEL and splenic T cells in the control and cytotoxic elimination of
infected cells during a systemic, acute virus infection in the
intestinal mucosa and the spleen, respectively. To this end we directly
examined the relative virus-specific cytotoxic activity of ex vivo
isolated CD8ß and CD8 TCRß IEL and splenic CD8 T cells,
respectively, during a systemic infection with LCMV-WE. In parallel, we
followed by in situ hybridization the appearance and clearance of
virus-infected cells in the intestinal mucosa and the spleen.
These studies revealed the generation of potent cytotoxic IEL and
splenocytes alike upon appearance of LCMV-infected cells in vivo. When
assayed in vitro, ex vivo isolated CD8ß TCRß IEL exert
virus-specific cytotoxicity that equals or even exceeds the maximum
cytotoxicity of splenic CD8 T cells, whereas ex vivo isolated CD8
TCRß IEL are less potent in lysing Ag-primed target cells. Upon
adoptive transfer into freshly LCMV-infected recipient mice, in vivo
primed IEL and splenocytes alike protect from excessive LCMV
proliferation. Together with the observed disappearance of
LCMV-infected cells located in subepithelial regions of the intestinal
mucosa around day 8 post-LCMV infection, these results suggest an
important role of cytotoxic IEL in the control and elimination of
virus-infected host cells in situ.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
C57BL/6 mice were originally obtained from the Institut für Labortierkunde (Zurich, Switzerland). Mice are currently bred and reared in the animal facility of the Medical Faculty, University of Bern.
LCMV infections
A stock of LCMV strain WE (14) was provided by S. Oehen (Zurich, Switzerland). Unless otherwise indicated, 104 PFU of LCMV-WE/mouse in a volume of 100 µl of MEM and 2% FCS was injected i.p.
Antibodies
Anti-CD45/B220 (RA3-6B2), anti-CD4 (GK1.5), anti-CD8
(53-6.7), anti-TCR (GL3), anti-TCRß (H57-597),
anti-CD11c (N418), and F4/80 (specific for mature M
) were
protein G purified from hybridoma supernatants and either biotinylated
or FITC labeled according to standard protocols (20).
Anti-CD8-CyChrome (53-6.7) and anti-CD8ß-FITC or -PE (53-5.8)
were purchased from PharMingen (San Diego, CA). The hybridoma producing
the LCMV-WE-specific mAb (VL4) was provided by R. M. Zinkernagel
(Zurich, Switzerland).
Tissue preparation
Tissue samples were placed in Tissue-Tek O.C.T. compound (Sakura, Tokyo, Japan), frozen on dry ice, and stored at -80°C until they were used to prepare frozen sections. Five-micron cryostat sections were placed on poly-L-lysine-coated glass slides (Polysine, Menzel Gläser, Braunschweig, Germany). For in situ hybridization slides were fixed in freshly prepared PBS-buffered 4% paraformaldehyde for 20 min, washed sequentially with 3x PBS, 1x PBS, and H2O and dehydrated through graded ethanol. After air-drying, the slides were stored at -80°C for up to several months. For immunohistochemistry, slides were fixed for 30 s in acetone, air-dried, and stored at 4°C for up to several weeks.
35S-labeled RNA probe preparation and in situ hybridization
Preparation of 35S-labeled RNA probes. A 511-bp fragment of the mouse perforin cDNA (provided by H. Hengartner, Zurich, Switzerland) and the entire coding regions of the mouse granzyme B (provided by C. G. Lobe and R. C. Bleackley) and LCMV glycoprotein (GP; provided by H. Pircher, Freiburg, Germany) cDNAs, cloned into the expression vectors pGEM-2, pSPT 671, and pGEM-1, respectively, were used to prepare 35S-labeled RNA probes using T7 or SP6 RNA polymerase reactions, as previously described (21).
In situ hybridization. Cryostat sections and sorted cells were hybridized with appropriate probes at 2 x 105 cpm/µl hybridization solution as previously described (21). After hybridization and washing off nonhybridized probe, slides were dipped into prewarmed NTB-2 emulsion (Eastman Kodak, New Haven, CT) 1/2 diluted in 800 mM ammonium acetate, dried, and exposed in the dark at 4°C for 7 (GP) or 28 (perforin, granzyme B) days. Slides were developed with Kodak developer PL 12 for 5 min and fixed with Kodak fixer for 8 min at room temperature. Counterstaining was performed in Nuclear Fast Red (0.05% in 5% aluminum sulfate) for 20 min.
Immunohistochemistry
Slides were postfixed in acetone for 10 min and air-dried. After rehydration in Tris-buffered saline (pH 7.5) containing 10% horse serum, slides were incubated with the appropriate primary mAb for 40 min at room temperature. When peroxidase was used, endogenous peroxidase was quenched by incubation of slides for 25 min in methanol and 0.3% H2O2 preceding second-step incubation. For spleen sections, anti-rat Ig (DAKO, Glostrup, Denmark) for 30 min at room temperature was used as a second step, followed by rat alkaline phosphatase-anti-alkaline phosphatase (DAKO) according to the manufacturer’s instructions. On intestinal tissue sections biotinylated anti-rat Ig Abs (DAKO) were used for 30 min as a second step, followed by incubation with premixed avidin-biotin complex and avidin-peroxidase (DAKO), according to the manufacturer’s instructions. As an exception, detection of CD11c (N418, hamster IgG) was performed using biotinylated primary Ab after blocking with hamster IgG isotype standard (PharMingen) followed by premixed avidin-biotin complex and avidin-alkaline phosphatase or avidin-peroxidase (DAKO), according to the manufacturer’s instructions. Color reaction to detect alkaline phosphatase was performed with naphthol-AS-bisphosphate and New Fuchsin (Sigma, St. Louis, MO); to detect peroxidase, 3,3'-diaminobenzidine (Fluka, Buchs, Switzerland) was used.
Cell preparations
Spleens were placed in Ca2+- and Mg2+-free HBSS, 10 mM HEPES, and 5% horse serum (HBSS-5). Splenocytes were released by grinding the tissue through a 70-µm pore size cell strainer (Falcon, Becton Dickinson, San Jose, CA) with a plunger. Erythrocytes were lysed by osmotic shock treatment for 30 s. For in situ hybridization (ISH) on sorted cell subsets, spleens were cut into small pieces and digested three times for 25 min each time at 37°C under gentle agitation in HBSS-5 reconstituted with Ca2+ and Mg2+ containing 50 U/ml collagenase D and 30 Kunitz units/ml DNase I (Roche Diagnostics, Rotkreuz, Switzerland).
IEL and lamina propria lymphocyte were isolated from the small intestine as previously described (22, 23). Briefly, the entire small intestine was placed in Ca2+- and Mg2+-free HBSS, 10 mM HEPES, and 2% horse serum (HBSS-2). After removing Peyer’s patches and longitudinally opening the gut, the tissue was cut into pieces of 1–2 cm each. Enterocytes and IEL were detached from the basement membrane by incubating pieces in HBSS-2 containing 2 mM DTT and 0.5 mM EDTA at 37°C with stirring. Released cells were incubated for 30 min in HBSS-5 at 37°C in 5% CO2 to ameliorate subsequent separation of IEL from enterocytes by discontinuous (44/67%) Percoll (Pharmacia, Uppsala, Sweden) gradient centrifugation. With this procedure 0.5–1 x 107 and 2–3 x 107 IEL were reproducibly obtained from noninfected controls and from day 8 LCMV-infected mice, respectively. Tissue pieces were used for lamina propria lymphocyte isolation. To remove remaining epithelial cells, tissue pieces were incubated three times for 15 min each time at 37°C in HBSS-5 containing 5 mM EDTA with stirring and subsequent short vortex mixing. Thereafter, pieces were washed and incubated three times for 25 min each time at 37°C in HBSS-5 reconstituted with Ca2+ and Mg2+ containing 50 U/ml collagenase type D and 30 Kunitz units/ml DNase I with gentle agitation. Supernatants were pooled, serially passed through 70- and 40-µm pore size cell strainers, and purified by 44/67% Percoll gradient centrifugation.
Cell staining, cell sorting, and FACS analysis
Cells were stained with 0.25 µg of the appropriate mAb/106 cells for 10 min on ice at a density of 106-107 cells/100 µl of HBSS-5. Magnetic cell separation was performed to enrich for CD8+ splenocytes, following the instructions provided by the manufacturer (Miltenyi Biotec, Bergisch Gladbach, Germany) with the minor modifications that HBSS-5 and, for the incubation with avidin-coupled microbeads, Ca2+- and Mg2+-free HBSS and 10 mM HEPES without serum were used instead of the indicated buffers. For sorting with the FACSvantage (Becton Dickinson, San Jose, CA), stained cells were resuspended in HBSS-5 at 2 x 106 cells/ml. The FACSvantage cytometer and sort parameters were controlled using LYSYS II software (Becton Dickinson). For ISH of distinct cell subsets, cells were directly sorted onto poly-L-lysine-coated glass slides. Control stainings were acquired on a FACScan and analyzed on a Macintosh computer using CellQuest software (Becton Dickinson).
In vitro cytotoxicity assay
The H-2b-expressing thymoma cell line RMA (provided by D. Kägi, Zurich, Switzerland) was used as peptide-pulsed target cells; 106 target cells were labeled with 75 µCi of 51Cr-sodium (Amersham Life Sciences, Aylesbury, U.K.) in IMDM and 2% FCS (target medium) for 1 h at 37°C in 5% CO2. Labeled target cells were washed twice, counted, and resuspended at 3 x 104 cells/ml target medium. Shortly before the assay, labeled target cells were pulsed with 1 µg/ml of the H-2Db-binding, immunodominant, LCMV GP-derived peptide gp33–41 (gp33) or, as a control, H-2Db-binding, adenovirus-derived, nonapeptide adn5 (provided by H. P. Pircher, Freiburg, Germany). When LCMV-infected target cells were used, a 30% confluent monolayer of the fibroblast cell line MC57G (provided by R. M. Zinkernagel) was infected with LCMV-WE in MEM and 2% FCS at a multiplicity of infection of 0.01 for 1 h at room temperature on a rocking platform. Thereafter, infected cells were supplemented with fresh MEM and 5% FCS and were further grown for 48 h at 37°C in 5% CO2. Noninfected MC57G cells were used as control target cells. Shortly before the assay MC57G cells were detached with trypsin/EDTA and labeled with 51Cr-sodium as described above. Sorted effector cells were washed and resuspended at 8.1 x 105 cells/ml in IMDM and 5% FCS containing 4 mM L-glutamine (Seromed, Biochrom, Berlin, Germany), 1 µM 2-ME, and nonessential amino acid mix (Roche; effector medium). Effector cells were placed in a V-bottom 96-well plate (Costar, Cambridge, MA) and serially diluted in 1/3 steps. One hundred microliters of target cells (3000 cells) were added to all wells, including wells with effector medium only, for spontaneous and maximum release measurements. Effector and target cells were quickly mixed and shortly spun before Nonidet P-40 was added (final concentration, 0.5%) to wells designed for maximal release. Plates were incubated at 37°C in 5% CO2 for 5 h, and supernatant from each well was counted in a TopCount beta/gamma scintillation counter (Canberra Packard, Meriden, CT). Specific lysis was calculated as [(experimental counts - spontaneous counts)/(maximum counts - spontaneous counts)] x 100.
Virus protection assay
Unfractionated splenocytes (5 x 106) or IEL isolated from one day 8 LCMV-infected C57BL/6 mouse or 5 x 106 unfractionated splenocytes or IEL pooled from two noninfected C57BL/6 mice were injected i.v. in a volume of 100–200 µl MEM and 2% FCS into C57BL/6 recipients that had been infected with 104 PFU LCMV-WE i.v. 10 h previously. (Splenocytes and IEL were isolated from the same animals for adequate comparison.) Twenty or forty-four hours later (i.e., 30 or 54 h after infection) recipient spleens were harvested. Spleens were placed in 1 ml of MEM and 2% FCS and frozen at -80°C for later determination of the virus titer in a plaque assay.
LCMV plaque assay
The virus titer of infected spleens was determined in a virus plaque assay as previously described (24). Briefly, spleens frozen at -80°C in MEM and 2% FCS were thawed, homogenized, and mixed with MC57G cells in 10x dilution steps in a 24-well plate (TPP, Trasadingen, Switzerland). Virus was allowed to infect MC57G cells for 4 h. Thereafter, methylcellulose (Methocel, Sigma) diluted 1/1 in 2x DMEM was added to inhibit budding of virus. After 48 h at 37°C in 5% CO2, plaques of virus-infected cell clusters were detected immunohistochemically. For that purpose cells in the 24-well plate were fixed with PBS-buffered 4% paraformaldehyde and, after permeabilization in balanced salt solution containing 1% Triton X-100 (Fluka), were incubated with VL-4 Ab. Peroxidase-coupled anti-rat Ig Ab (DAKO) was used as second step, and color reaction was performed using o-phenylenediamine dihydrochloride (Sigma) according to the manufacturer’s instructions.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To monitor the extent of a systemic LCMV strain WE (subsequently
termed LCMV) infection, frozen tissue sections of
104 PFU of LCMV i.p. infected or, as a control,
noninfected C57BL/6 mice were used for ISH with a virus GP
mRNA-specific, 35S-labeled, antisense RNA probe.
Maximal signals for virus mRNA were found between day 4 (Fig. 1
) and day 6 (data not shown)
postinfection on both spleen and intestinal tissue sections. To further
substantiate the specificity of our detection method, semiserial tissue
sections from LCMV-infected animals on day 4 postinfection were either
used for ISH with GP mRNA-specific, 35S-labeled,
antisense RNA probe or immunohistochemistry (IHC) with the
virus-specific mAb VL-4 for direct comparison (Fig. 2, A and D, and
B and E, respectively). Virus protein-specific
signals perfectly match with signals for GP mRNA, although IHC is
slightly less sensitive than ISH, i.e., GP mRNA can also be detected at
sites where VL-4-specific signals seem to be absent.
|
|
, infected cells
and signals for CD11c were frequently observed in the same area of the
spleen (Fig. 2C) and the small bowel lamina propria (Fig. 2F). Most notably, CD11c+ DC were also
detected in close spatial association with the basement membrane and
the overlaying epithelium.
Detailed analysis of the spleen and small bowel of infected animals was
performed by ISH of isolated and sorted single cells with an LCMV
GP-specific RNA probe (Table I). The
results revealed that CD11c+ DC represent a
prominent subset of LCMV-infected cells 4 days after infection, with
2.11 ± 1.14 and 1.08 ± 0.32% of total DC being positive
for LCMV GP mRNA in the spleen and small bowel, respectively. In
addition, 1.81 ± 0.24 and 0.73 ± 0.16% of
F4/80+ macrophages (M) expressed LCMV GP mRNA
in the spleen and small bowel, respectively, at detectable levels. In
contrast, no signs of infection of small intestinal enterocytes with
LCMV were detected using this method. Because the LCMV strain WE
readily infects fibroblast cell lines in vitro, fibrocytes may also
constitute a considerable fraction of infected cells during an acute
infection in vivo. In contrast, however, LCMV-WE does not infect B and
T cells, as assessed by analysis of total lymphoid cells from acutely
infected mice (Table I) and from persistently infected carrier mice
(data not shown)
|
ß+CD8ß+, but not
TCRß+CD8+, IEL selectively expand
during acute LCMV infection
To determine changes in relative frequencies of potential CTL
subsets during a primary LCMV-specific immune response, splenocytes and
IEL isolated at different time points after LCMV infection were stained
for CD8 (CD8+), and
TCRß+CD8ß+
(CD8ß) or
TCRß+CD8+
(CD8), respectively, for subsequent FACS analysis. Relative
frequencies of CD8+ splenocytes and CD8ß IEL
increased about 3-fold until days 8–10 after LCMV infection, whereas
relative frequencies of CD8 IEL remained rather constant (Fig. 3A). After 8–10 days
postinfection, relative frequencies of CD8+
splenocytes and CD8ß IEL begin to decrease again to reach
frequencies close to initial values on day 20 postinfection. The
relative frequencies of CD8+ splenocytes and
CD8ß IEL correlated well with the absolute numbers of cells that
were isolated at discrete time points after LCMV infection (Fig. 3B), i.e., on day 8 post-LCMV infection >3-fold higher
numbers of CD8+ splenocytes and CD8ß IEL
were obtained compared with those for noninfected controls.
|
Expressions of perforin and granzyme B mRNA were followed to
monitor in situ the activation of cytotoxic cells in the spleen and
intestinal mucosa (Fig. 4). Although
perforin mRNA was virtually undetectable on control tissue sections
(Fig. 4, A and C), granzyme B mRNA was found to
be constitutively expressed, particularly in the small bowel (Fig. 4, E and G). Upon systemic LCMV infection, perforin
and large amounts of granzyme B mRNA were detected in both spleen and
intestinal mucosa (Fig. 4, B, D, F, and H).
Maximum numbers of positive cells and maximum levels of mRNA expression
per cell were observed around day 6 after LCMV infection in both
organs. Histologically, perforin and granzyme B mRNA-expressing cells
in the spleen were evenly distributed in the red pulp and
preferentially accumulated in the T cell zones of the white pulp and in
the marginal zones, where most of the virus-infected cells were located
(Fig. 4, B and F).
|
D and inset high power
photograph in D). Perforin mRNA-expressing IEL were found in
the CD8ß and, to a lesser extent, also in the CD8 subset by
ISH after isolation and sorting (data not shown).
The intestinal mucosa of noninfected control animals already contains
numerous granzyme B mRNA-expressing cells. Nevertheless, both the
number of granzyme B mRNA-positive cells and the level of expression
per cell markedly increased upon LCMV infection (Fig. 4, G
and H).
Kinetics of CTL activity of splenocytes and IEL in vitro during systemic LCMV infection
To determine the onset, peak, and decrease in LCMV-specific
cytotoxicity of IEL subsets compared with splenic CTL,
CD8+ splenocytes, and IEL subsets were isolated
from noninfected mice and from mice infected 4, 6, 8, 10, 12, and 20
days earlier with LCMV. Cytotoxic activity was assessed in vitro in a
51Cr release assay using syngeneic target cells
pulsed with the immunodominant peptide gp33 (Fig. 5). Detectable in vitro cytotoxicity
appeared simultaneously on day 6 postinfection in all subsets tested,
i.e., CD8+ splenic T cells, and CD8ß and
CD8 IEL. Maximum LCMV gp33-specific cytotoxicity was measured ex
vivo on day 8 postinfection with all cell subsets tested. In vitro
gp33-specific cytotoxic activity of CD8ß IEL was repeatedly found
to be even more potent than that of CD8+ splenic
T cells, although differences were not statistically significant
(p > 0.05) at a fixed E:T cell ratio of 5:1.
In contrast, as depicted in Fig. 5, CD8 IEL exerted about 5-fold
reduced in vitro gp33-specific cytotoxicity compared with CD8ß
IEL. After the peak on day 8, the cytotoxic activity of CD8ß IEL
dropped faster compared with that of splenic CD8+
T cells, leading to a statistically significant difference
(p = 0.047) on day 10 postinfection.
|
A). Equal in vitro
LCMV-specific cytotoxic activity of CD8+
splenocytes on day 8 postinfection is achieved only with a 2-fold
higher E:T cell ratio than with CD8ß IEL, and this difference
becomes statistically significant at higher E:T cell ratios
(p = 0.016 at an E:T cell ratio of 15:1). The
observed maximum value of CD8 IEL-mediated LCMV-specific
cytotoxicity is already obtained at 10 times lower E:T cell ratios
using CD8+ splenocytes or CD8ß IEL as
effector cells instead. Lysis of target cells pulsed with the MHC class
I H-2Db-binding, unspecific control peptide adn5
was <5% in all experiments and did not show a titration with E:T cell
ratios, indicating tight Ag specificity of all CTL subsets tested (Fig. 6A). During acute, systemic LCMV infection T cells
constitute an important, although nonexpanding, fraction of small bowel
IEL (data not shown). Therefore, LCMV-specific cytotoxic activity of ex
vivo isolated TCR IEL on day 8 postinfection was assessed using
LCMV-infected 51Cr-labeled target cells (Fig. 6B). LCMV-infected MC57G cells were used instead of
peptide-pulsed RMA cells because the majority of T cells do not
recognize processed peptides in the context of classical MHC class I
molecules (reviewed in Ref. 25). Fig. 6B shows
that TCR IEL do not exert LCMV-specific cytotoxicity in this
experimental set-up, whereas CD8ß IEL and
CD8+ splenocytes isolated from the same mouse
efficiently lyse LCMV-infected, but not uninfected, MC57G target
cells.
|
To test the functional CTL activity of LCMV-specifically primed
IEL and splenocytes in vivo, IEL or splenocytes were isolated from day
8 LCMV-infected donor animals and adoptively transferred i.v. into
syngeneic recipient animals that had been infected with
104 PFU of LCMV i.v. 10 h previously. The
capability of transferred cells to control virus production in
recipient animals was tested by assessing the number of infectious
virus particles in the spleens of recipient animals 20 h later.
Upon transfer of either 5 x 106 total IEL
or splenocytes from the same donors, infected with LCMV 8 days
previously, about 100 times less infectious virus particles could be
isolated from recipient spleens compared with nontransferred animals or
from recipients of IEL or splenocytes from noninfected donor animals
(Fig. 7). Notably, the virus titer in the
spleen of recipient animals remained virtually unaltered during the
next 20 h after adoptive transfer of primed CTL compared with that
at the time of adoptive transfer, i.e., 10 h after LCMV infection.
Although in this experimental set-up LCMV is generally able to increase
in number >30 h postinfection, about 100 times less infectious virus
particles are isolated from spleens of mice transferred with either
primed splenocytes or primed IEL compared with those of nontransferred
control animals at 54 h post-LCMV infection (data not shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ß IEL to 1)
lyse LCMV Ag-presenting target cells in vitro and 2) protect naive,
freshly LCMV-infected recipient mice from productive virus replication
in vivo after adoptive transfer.
In a direct comparison we could show that the main virus-specific
cytotoxic activity of IEL is mediated by the CD8ß T cells and that
CD8 IEL exert reduced, although virus peptide-specific, cytotoxic
activity in vitro (Figs. 5 and 6). Hence, these findings support the
concept that CD8 IEL are primarily responsible for the
maintenance of local immune homeostasis and surveillance of the
epithelial cell layer integrity. In this respect it is also important
to note that most CD8 IEL normally express autoreactive TCR
(26, 27, 28) (S. Müller, unpublished observations) and
that CD8 IEL may require a higher avidity of the TCR to fully
respond to Ag compared with CD8ß T cells. Data in support of this
idea have been recently reported using TCR transgenic animals
(28, 29).
In a previous report the virus-specific cytotoxicity of TCRß IEL
has been considered to be lower than the corresponding values measured
for splenic CD8+ T cells (13). In
contrast, our results show that on day 8 postinfection in vivo primed
CD8ß IEL exert comparable cytotoxic activity as splenic
CD8+ T cells at a 2-fold lower E:T cell ratio.
Differences in the isolation procedure for IEL, amount of virus, and
differences in the virus strains used for infection may have
contributed to these discrepant results. In addition, our data indicate
that the use of pooled TCRß+ IEL, which
include both CD8 and CD8ß IEL as effector cells, as in this
previous study, may result in a markedly lower specific cytotoxic
activity at comparable E:T cell ratios compared with CD8ß IEL
alone.
During systemic LCMV infection CTL from spleen and intestinal
epithelium display similar kinetics of expansion in vivo and similar
onset, peak, and decrease in specific cytotoxic activity in vitro.
Notably, on day 8 postinfection CD8ß IEL mediate significantly
stronger virus-specific cytotoxic activity than splenic
CD8+ T cells, whereas 2 days later inverse
results were obtained. These results may indicate a more rapid
elimination of LCMV-specific CTL in the IEL compartment compared with
the spleen after day 8 postinfection. Such an accelerated loss of
activated effector cells upon successful elimination of the triggering
Ags may contribute to the minimization of CTL-mediated tissue damage in
the intestinal mucosa.
With the appearance of a potent cell-mediated cytotoxicity among IEL the frequency of LCMV-infected cells in the underlying lamina propria greatly decreases. By day 10 postinfection, LCMV-infected cells are barely detectable on tissue sections by IHC or ISH (data not shown). However, while the appearance of potent cell-mediated cytotoxicity in the spleen is accompanied by a marked and transient disruption of the splenic architecture (Ref. 30 and our own observations), the potent LCMV-specific cytotoxic activity mediated by IEL did not result in obvious histopathological consequences (data not shown). This may indicate a tighter control of T cell activity in the intestinal epithelial compartment, such as the inhibition of bystander killing that might lead to tissue destruction and eventually to damage of the physical mucosal barrier.
When ex vivo isolated IEL are cultured, they poorly proliferate in
response to TCR-mediated triggers (31) and a considerable
fraction of freshly isolated IEL undergo spontaneous apoptosis after
overnight in vitro culture regardless of their Ag-specific or mitogenic
stimulation (S. Müller, unpublished observations; T. Brunner,
manuscript in preparation). On this background the assessment of
anti-virus protective activity upon re-exposure to the specific Ags
in vivo represents an important issue. To this end we have chosen an
adoptive transfer model established by Assmann-Wischer et al.
(18) and recently used by Ehl et al. (19) to
demonstrate potent inhibition of initial virus replication in freshly
LCMV-infected animals, mediated by the adoptively transferred, primed
splenocytes. Our results clearly show that in vivo LCMV-primed IEL
mediate identical protection from productive expansion of LCMV upon
adoptive transfer into freshly infected recipients compared with
splenocytes (Fig. 7). Hence, these results demonstrate that activated
IEL are able to mount a potent immune response after reexposure to
specific Ag in vivo. These results suggest an important contribution of
IEL to systemic antivirus immune response through their sustained
Ag-specific cytotoxic activity. Unfortunately, fractionation of
isolated IEL and splenocytes alike into distinct T cell subsets by
cytofluorometric or magnetic bead sorting and subsequent transfer into
LCMV-infected mice consistently resulted in a poor protection of
recipients from virus propagation. Hence, the respective capacity of
CD8ß vs CD8 or TCR IEL to control virus propagation in
vivo could not be directly assessed. However, because during an acute
LCMV infection 1) CD8ß T cells are the only IEL subset to expand
in both relative and absolute numbers, and 2) LCMV-specific cytotoxic
activity mediated by ex vivo isolated CD8 IEL is marginal or even
absent when TCR IEL are analyzed, protection of LCMV-infected
recipients from virus propagation may mainly be mediated by this cell
subset. This assumption is further supported by the finding that,
particularly at time points beyond 30 h after transfer, the level
of protection directly correlates with the frequency of CD8ß IEL
within the transferred cells and does not correlate or even inversely
correlate with the frequencies of CD8 or TCR IEL,
respectively (data not shown).
In a recent report Kim et al. (32) state that 3 days after
activation, peripheral T cells may also migrate to the IEL compartment.
However, several observations argue against an exclusive recruitment of
primed CD8+ T cells to the intestinal epithelium:
1) after systemic infection with LCMV functionally cytotoxic T cells
are detected simultaneously in the IEL compartment and in the spleen
(Fig. 5); 2) virus-infected cells are in close spatial association with
IEL as early as virus is manifest in the spleen (Figs. 1 and 2); 3)
perforin and granzyme B mRNA expressions, as an indication of recently
activated CTL, are simultaneously induced in splenocytes and IEL (Fig. 4); and 4) an important fraction of the virus-specific, cytotoxic
CD8ß IEL expresses the IEL-specific integrin
IELß7 (data not
shown); in contrast,
IELß7 is not expressed
on obviously recruited alloreactive, donor-derived splenic CTL isolated
from the small intestinal epithelium after adoptive transfer in a mouse
model of graft-vs-host
disease.4 Hence, our
data are consistent with the idea that LCMV-infected dendritic cells in
the intestinal lamina propria present virus-derived Ag across the
basement membrane to IEL, leading to the activation of virus-specific
cytotoxic IEL. In support of this idea are recent findings that DC in
the intestinal mucosa are indeed able to cross with their dendrites the
basal membrane to possibly gain access to the IEL (33, 34). However, alternative explanations, such as migration of
lamina propria CD8 T cells into the epithelial compartment following
specific activation, cannot be formally ruled out, although experiments
with parabiontic mice demonstrated only a minimal migration of
lymphocytes to and from the intestinal epithelium
(35).
An infection rate with LCMV of only 1–2% among DC and M may appear
low. However, considering that the maximal virus load of LCMV on day 4
postinfection rarely exceeds 107 infectious virus
particles/spleen (data not shown), one would expect a total number of
only 105-106 infected cells
when an average of 10–100 virus copies/infected cell are estimated.
Therefore the determined frequencies of 1–2% LCMV-infected DC and
M clearly indicate that these subsets represent a prominent fraction
of all LCMV-infected cells.
In conclusion, we demonstrate that IEL become activated to exert potent
virus-specific cytotoxicity after systemic infection with LCMV-WE.
CD8ß IEL not only represent the main IEL subset during the
inductive stage of the immune response in the IEL compartment, they are
also more potent in lysing Ag-primed targets than CD8 IEL. With
the appearance of the activated, cytotoxic IEL in situ, the number of
LCMV-infected cells in the intestinal mucosa rapidly decreases. The
functional relevance of these activated, virus-specific T cells in the
intraepithelial compartment in control and elimination of
virus-infected cells is further demonstrated by the observed dramatic
prevention of virus propagation in freshly LCMV-infected recipients
upon transfer of IEL from day 8 LCMV-infected donor mice. These
adoptive transfer experiments thus clearly demonstrate that ex vivo
isolated IEL are not generally prone to rapidly undergo apoptosis, but
can exert potent effector functions after re-encountering their
specific Ag in vivo.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Christoph Mueller, Division of Immunopathology, Institute of Pathology, Murtenstrasse 31, CH-3010 Bern, Switzerland. E-mail address:
3 Abbreviations used in this paper: IEL, intraepithelial lymphocytes; LCMV, lymphocytic choriomeningitis virus; GP, glycoprotein; IHC, immunohistochemistry; ISH, in situ hybridization; DC, dendritic cell; M, macrophage.
4 T. Brunner and C. Wasem. Submitted for publication.
Received for publication October 4, 1999. Accepted for publication December 7, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
/ß+ and /+ T cells in murine intestinal epithelium. Eur. J. Immunol. 20:1623.[Medline]
-producing activities of T cells in the mouse intestinal epithelium are strain dependent. Proc. Natl. Acad. Sci. USA 90:8204. T-cell antigen receptors. Proc. Natl. Acad. Sci. USA 90:8591./ß+ Thy-1+ gut intraepithelial lymphocytes, lytic for parasite-infected enterocytes. J. Immunol. 153:4596.[Abstract]
-dystroglycan as a receptor for lymphocytic choriomeningitis virus and Lassa fever virus. Science 282:2079.ß intestinal intraepithelial lymphocytes (IEL) are induced to express CD5 upon antigen-specific activation: CD5- and CD5+ CD8ß IEL do not represent separate T cell lineages. Eur. J. Immunol. 27:1756.[Medline]
/ cells. Annu. Rev. Immunol. 11:637.[Medline]
/ co-receptors by self-antigen in the murine gut. Proc. Natl. Acad. Sci. USA 89:5336./ß in the intestinal epithelium. J. Exp. Med. 178:1947./ß in the intestinal epithelium: evidence for distinct circulation patterns of gut- and thymus-derived T lymphocytes. J. Exp. Med. 176:187.This article has been cited by other articles:
|
|
 |
|
  A. Wakabayashi, Y. Nakagawa, M. Shimizu, K. Moriya, Y. Nishiyama, and H. Takahashi Suppression of an Already Established Tumor Growing through Activated Mucosal CTLs Induced by Oral Administration of Tumor Antigen with Cholera Toxin J. Immunol., March 15, 2008; 180(6): 4000 - 4010. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. M. Moretto, L. M. Weiss, C. L. Combe, and I. A. Khan IFN-{gamma}-Producing Dendritic Cells Are Important for Priming of Gut Intraepithelial Lymphocyte Response Against Intracellular Parasitic Infection J. Immunol., August 15, 2007; 179(4): 2485 - 2492. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Bharhani, J. S. Grewal, R. Peppler, C. Enockson, L. London, and S. D. London Comprehensive phenotypic analysis of the gut intra-epithelial lymphocyte compartment: perturbations induced by acute reovirus 1/L infection of the gastrointestinal tract Int. Immunol., April 1, 2007; 19(4): 567 - 579. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Cesson, K. Stirnemann, B. Robert, I. Luescher, T. Filleron, G. Corradin, J.-P. Mach, and A. Donda Active Antiviral T-Lymphocyte Response Can Be Redirected against Tumor Cells by Antitumor Antibody x MHC/Viral Peptide Conjugates Clin. Cancer Res., December 15, 2006; 12(24): 7422 - 7430. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C Mueller and A J Macpherson Layers of mutualism with commensal bacteria protect us from intestinal inflammation Gut, February 1, 2006; 55(2): 276 - 284. [Full Text] [PDF]
|
|
|
|
 |
|
 H. Mizubuchi, T. Yajima, N. Aoi, T. Tomita, and Y. Yoshikai Isomalto-Oligosaccharides Polarize Th1-Like Responses in Intestinal and Systemic Immunity in Mice J. Nutr., December 1, 2005; 135(12): 2857 - 2861. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. S. Bharhani, J. S. Grewal, M. J. Pilgrim, C. Enocksen, R. Peppler, L. London, and S. D. London Reovirus Serotype 1/Strain Lang-Stimulated Activation of Antigen-Specific T Lymphocytes in Peyer's Patches and Distal Gut-Mucosal Sites: Activation Status and Cytotoxic Mechanisms J. Immunol., March 15, 2005; 174(6): 3580 - 3589. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. L. Shacklett, C. A. Cox, M. F. Quigley, C. Kreis, N. H. Stollman, M. A. Jacobson, J. Andersson, J. K. Sandberg, and D. F. Nixon Abundant Expression of Granzyme A, but Not Perforin, in Granules of CD8+ T Cells in GALT: Implications for Immune Control of HIV-1 Infection J. Immunol., July 1, 2004; 173(1): 641 - 648. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Menge, I. Stamm, P. M. van Diemen, P. Sopp, G. Baljer, T. S. Wallis, and M. P. Stevens Phenotypic and functional characterization of intraepithelial lymphocytes in a bovine ligated intestinal loop model of enterohaemorrhagic Escherichia coli infection J. Med. Microbiol., June 1, 2004; 53(6): 573 - 579. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J-K Kim, M Takeuchi, and Y Yokota Impairment of intestinal intraepithelial lymphocytes in Id2 deficient mice Gut, April 1, 2004; 53(4): 480 - 486. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Saurer, I. Seibold, S. Rihs, C. Vallan, T. Dumrese, and C. Mueller Virus-Induced Activation of Self-Specific TCR{alpha}{beta} CD8{alpha}{alpha} Intraepithelial Lymphocytes Does Not Abolish Their Self-Tolerance in the Intestine J. Immunol., April 1, 2004; 172(7): 4176 - 4183. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Moretto, L. M. Weiss, and I. A. Khan Induction of a Rapid and Strong Antigen-Specific Intraepithelial Lymphocyte Response during Oral Encephalitozoon cuniculi Infection J. Immunol., April 1, 2004; 172(7): 4402 - 4409. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Fulton, J. Smith, C. Cunningham, and C. F. Cuff Influence of the Route of Infection on Development of T-Cell Receptor {beta}-Chain Repertoires of Reovirus-Specific Cytotoxic T Lymphocytes J. Virol., February 1, 2004; 78(3): 1582 - 1590. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Kinoshita, T. Hiroi, N. Ohta, S. Fukuyama, E. J. Park, and H. Kiyono Autocrine IL-15 Mediates Intestinal Epithelial Cell Death Via the Activation of Neighboring Intraepithelial NK Cells J. Immunol., December 1, 2002; 169(11): 6187 - 6192. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Hayday and J. L. Viney The Ins and Outs of Body Surface Immunology Science, October 6, 2000; 290(5489): 97 - 100. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), TCR
) were subsequently determined by FACS analysis.
B, Absolute numbers of isolated CD8+
splenocytes (
) were calculated using the relative frequencies in
A and the total numbers of isolated splenocytes, and
IEL, respectively. Data represent the mean (±SEM) of three individual
sets of experiments.
