| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Pathology and Laboratory Medicine, University of Wisconsin Medical School, Madison, WI 53706; and
Department of Pediatrics, Jacobi Medical Center and Albert Einstein College of Medicine, Brooklyn, NY 10461
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
11/V
3 anti-pigeon cytochrome c) are
preferentially excluded from granulomas of BCG-infected AND mice, and
cells expressing secondary endemic V-chains are enriched among AND
cells homing to granulomas. Next, we addressed whether TCR
heterogeneity is required for effective granuloma formation. We
infected 5CC7/recombinase-activating gene 2-/- mice with
recombinant BCG that express pigeon cytochrome c peptide
in a mycobacterial 19-kDa bacterial surface lipoprotein. A
CD4+ T cell with a single specificity in the absence of
CD8+ T cells is sufficient to form granulomas and
adequately control bacteria. Our study shows that expanded monoclonal T
cell populations can be protective in mycobacterial
infection. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
T cells (10, 11), MHC class II (11, 12)
pathways, and Th1-type cytokine production (TNF- (13),
IFN-
(14, 15, 16), and IL-12 (17, 18, 19)) in
protective T lymphocyte/macrophage interactions
(20, 21, 22, 23, 24).
Although mycobacteria contain numerous B cell Ags recognized by the
humoral arm of the host immune response (25), the
protective role of B cells is secondary to T cell responses
(26). This study is focused on the role of TCR specificity
of CD4+ T cells in inducing protective
granulomatous responses using a murine model of chronic BCG infection.
We studied TCR diversity at the local granulomatous lesion and in the
spleen. Measurements of TCR V and J-chain usage show a broadly
based and diverse systemic immune response to mycobacterial infection.
However, after infection of AND TCR transgenic mice with BCG,
nontransgenic (Ag-specific) T cells were preferentially localized to
the granuloma, suggesting that selection plays a role in homing to or
retention in the inflammatory lesion. Other studies have also described
skews in the TCR repertoire at local inflammatory sites, suggesting
that the localization of T cells to lesions is not haphazard
(27). Given these observations, our central question in
this study was whether the TCR heterogeneity observed after BCG
infection is essential for protective granuloma formation, or whether
an Ag-specific T cell response is essential. The data presented suggest
that a single type of Ag-specific T cell can be sufficient for
protective granuloma formation in response to BCG infection.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In these studies we used C57BL/6, B10.BR, B10 recombinase-activating gene 2 (Rag2)-/-(H2d), and C3H Rag2-/-(H2k) strains of mice (The Jackson Laboratory, Bar Harbor, ME) and Taconic Farms, Warrington, PA). TCR transgenic mice used were AND (S. Hedrick, University of California, San Diego, CA) and 5CC7/Rag2-/-(H2k) (28) (Taconic Emerging Models Program). Animals were housed at the University of Wisconsin Medical School Animal Care Unit, which has American Association for Accreditation of Laboratory Animal Care accreditation and meets Public Health Service policy. All studies were approved by the University of Wisconsin Medical School’s animal care committee.
M. bovis BCG infections
The Pasteur strain of BCG (Staten Serum Institute) was grown in Middlebrook 7H9 supplemented with 0.05% Tween 80 and 10% oleic acid-dextrose-catalase supplement (Difco, Detroit, MI) and stored in frozen aliquots at -70°C. For infections, ampoules were thawed, and the inoculum was diluted in saline plus 0.05% Tween 80 and briefly exposed to sonic oscillation to obtain a single-cell suspension. Mice were infected ip with 1 x 107 CFU of BCG in 100 µl (14). The dose injected is not lethal in C57BL/6 mice and induces a chronic infection. Infection was verified by histology of liver tissue samples.
Histology
Small pieces of liver were fixed in 10% formalin, before being imbedded in paraffin for thin sectioning (8–10 µm). Hematoxylin-eosin (H&E) staining and Ziehl-Neelsen staining for acid fast bacteria were performed by the University of Wisconsin Department of Pathology’s Histopathology Service. Quantitative studies were performed by direct microscopic examination using an Olympus reticular eyepiece (New Hyde Park, NY) containing a 10 x 10 grid. Bacteria per lesion is the number of acid fast rods visible per granuloma at x1000 total magnification under oil. Data are presented as the mean ± SEM of a minimum of 30 counts/mouse liver section. The number of individual mice is indicated in each figure legend.
Isolation of splenocytes and granuloma-infiltrating cells
Isolation of granulomas was described previously (15, 29, 30). Spleens were removed aseptically from 8- to 16-wk-old mice, and viable cells were separated by centrifugation through Lympholyte M solution (Cedarlane Laboratories, Hornsby, Canada) as previously described (31). Single granulomas were isolated from the preparative suspension before dispersal with collagenase using a Pasteur pipette flame drawn to a finer tip under x10 magnification.
Flow cytometry and Abs
Splenocytes or granuloma cell suspensions were incubated for 30 min at 4°C with different labeled Abs at saturation, then washed and analyzed. Unlabeled 50 µg/ml anti-Fc receptor Ab (2.4G2) was used to block nonspecific binding of Fc receptors. Cell surface staining on 10,000 events was measured using a FACSCalibur (Becton Dickinson, Mountain View, CA) and analyzed using the CellQuest computer program (Power Macintosh version 3.0; Becton Dickinson).
Monoclonal Abs used were purified from hybridoma cell lines obtained
from the American Type Culture Collection (Manassas, VA) as indicated.
Hybridoma cells were cultured in HB-101 serum-free medium, and the Abs
were precipitated from supernatants by 45% saturated ammonium sulfate.
The purified Abs were labeled with biotin, FITC, or Cy5. Abs used
included those specific for murine CD4 (GK1.5), CD8 (53-6.7), MAC-1
(MI/70.15), CD44 (Pgp-1, IM 7.8.1), L-selectin (CD62L, Mel-14), LFA-1
(M17/4), and transferrin receptor (C2). The KJ25 hybridoma specific for
V3 was a gift from Philippa Marrack (University of Colorado Heath
Sciences Center, Denver, CO). Quantum Red-labeled anti- TCR
and CD4 and PE-labeled anti-V3 Abs were purchased from Sigma
(St. Louis, MO) and PE-labeled anti-MHC class II and FITC-labeled
anti-V4, V6, and V8.1+8.2 were purchased from PharMingen
(San Diego, CA).
Polymerase chain reactions
RT-PCR was performed as previously described (29) using published primers (32). Primers were titrated using positive (spleen cDNA) and negative (endothelial cell cDNA) templates.
Genetic manipulation of M. bovis BCG
A mycobacterial transformation vector based on pMV261
(33) was used for surface expression of a 19-kDa
lipoprotein-pigeon cytochrome c (PCC) epitope fusion protein
under control of a mycobacterial hsp60 promoter (Fig. 4
A). A sticky ended oligo encoding a 17-aa PCC epitope
(34, 35, 36) was cloned into the multiple cloning site in the
expression vector, and the insertion was confirmed by both restriction
digests and sequencing (University of Wisconsin Biotechnology Center
Sequencing Facility). BCG was transformed by electroporation of plasmid
DNA using published protocols (37). Transformed cells were
plated on 7H10 agar plates containing 50 µg/ml kanamycin sulfate and
incubated at 37°C for 2–3 wk. Kanamycin-resistant clones were
picked, and the presence of the PCC epitope in the transformant DNA was
confirmed by PCR using gene-specific primers and boiling minipreps of
candidates (Fig. 4B). Using this method, we did not detect
any loss or alteration of the transforming DNA in any of the
kanamycin-resistant transformants examined. PCC containing
transformants (Fig. 4B, lanes 1 and 2)
were clearly distinguished from those containing parental plasmids
(Fig. 4B, lane 3) by the increased size of the
specific PCR product. We screened antibiotic-resistant clones
containing plasmid DNA for those expressing the epitope using a T cell
proliferation assay (see Results) using splenocytes from
AND mice.
|
Splenocytes were seeded into 96-well plates at
106/well with medium alone, with 5 µg/ml
-CD3 Ab as a positive control, or with 100 µg/ml PCC (Sigma) to
measure Ag-specific activation. After 48–72 h, the extent of
stimulation was measured by flow cytometric analysis of harvested cells
for the expression of activation and cell cycling markers.
IFN- ELISA
Samples for cytokine analysis were collected from 1 x
106 liver granuloma cells or spleen cells (live
by trypan blue exclusion) seeded into 96-well plates in 0.2 ml of
complete medium, and stimulated with 5 µg/ml CD3 Ab. After 72
h, cell culture supernatants were harvested and stored at -70°C
until testing. Measurement of secreted IFN- was made by ELISA using
standard methods. Briefly, plates were coated with anti-IFN-
capture Ab (PharMingen, San Diego, CA). Serial 2-fold dilutions of
either murine rIFN- (Genzyme, Cambridge, MA) or test supernatants
were added to triplicate wells. Detection of bound cytokine used
biotinylated anti-mouse IFN- (PharMingen) followed by
streptavidin alkaline phosphatase (Molecular Probes, Eugene, OR). Wells
were developed with 50 µM 4-methylumbelliferyl phosphate (MUP;
Molecular Probes), and fluorescence intensity was measured with an
HTS7000 Bioassay reader (Perkin-Elmer, Foster City, CA). Units of
IFN- were calculated with reference to the rIFN- standard. All
data are expressed as the amount of secreted cytokine per 1 x
106 cells.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Fig. 1A shows the chronic
inflammatory response present in the liver from a C57BL/6 mouse after 6
wk of infection. H&E staining of thin liver sections showed that
granulomatous lesions are numerous and well formed. Ziehl-Neelsen
staining for acid-fast bacilli demonstrated that bacteria were present
in lesions only rarely and were never present outside of granulomas.
Hence, both bacterial growth in infected macrophage and widespread
dissemination are well controlled.
|
gene usage were made. Six weeks after injection of C57BL/6
mice with a bacterial dose leading to chronic infection, splenocytes
and isolated granuloma cell suspensions were stained with a panel of
TCR V allele-specific Abs (Fig. 2, A and B). Surface staining patterns of spleen
cells from infected and uninfected mice were largely indistinguishable
from each other and from staining of granuloma cells. Fig. 2A shows typical flow cytometric plots from which
quantitative data were derived (V8.1+8.2 staining is shown).
Staining eight infected mice and four uninfected mice individually
allowed us to calculate an average frequency of expression in the
CD4+ cell population for each V allele (Fig. 2B). Similar uniformity of V allele expression between
spleen, infected spleen, and granuloma cells was seen in the
CD8+ cell population (data not shown). These data
rule out the existence of any mycobacterial superantigen-mediated
shifts occurring in the measured V populations (38) and
argue against the presence of a few dominant T cells in granulomas and
for the presence of a diverse repertoire.
|
usage in TCR -chains.
Although our FACS analysis separated TCR usage into four categories
(V3, V4, V6, and V8.1+8.2), the analysis of J gene usage
divided each category into an additional 12 subsets. In Fig. 2
C, we show J gene segment usage in V4 chain RT-PCR
products from infected spleen or granuloma using nested PCR. The
presence of product bands in all lanes from both infected spleen
and granuloma indicates, similarly to the FACS analysis, that the
TCR repertoire in granulomas induced by BCG is diverse and encompasses
all J-chain genes examined. At the same time, the variation in the
size and intensity of individual bands also suggests that antigenic
specificity or selection in response to a large number Ags may also be
active at the inflammatory site.
The study of bulk granuloma populations does not address the
possibility that each granuloma has a very limited T cell pool that is
different in different granulomas. To address this question, we
isolated single granulomas. Although we do not know the T cell yield
from a single granuloma, we are able to recover 2,000-10,000
cells total from a granuloma after collagenase treatment and
dissociation. This would predict a yield of several hundred to a few
thousand T cells from each granuloma. Fig. 2D illustrates
the appearance of a single granuloma after isolation stained with H&E,
and the J gene usage of V4+ T cells in
seven single granulomas. Our data show that single granulomas have a
diverse T cell repertoire. This analysis divides the total
V4+ repertoire (3.0 ± 0.1% of
CD4+ granuloma-infiltrating cells, Fig. 2B; and 2.7 ± 0.25% of CD8+
granuloma-infiltrating cells, data not shown) into 12 groups using J
primers and clearly indicates that a single granuloma is not comprised
of a few T cells that expand locally, but of a large number of T cells
that locate from the periphery. It is also clear from differences in
the intensity and relative size of individual PCR products that each
single granuloma has a different T cell repertoire (for example,
compare the J2.4 products from single granulomas 5 and 6).
Accumulation of T cells into granulomas is influenced by Ag specificity and activation phenotype
Given the overall heterogeneity of the TCR repertoire generated in
response to BCG infection, we questioned whether T cell accumulation is
dependent upon TCR specificity. To address this question, we infected
TCR transgenic AND mice containing V11 and V3 transgenes that
recognize a 17-aa epitope of PCC in the context of
IEk class II molecules (34). In
naive animals, <4 mo old, and kept under clean, pathogen-free
conditions, the percentage of TCR transgene-expressing T cells varied
from 95–98% of the total T lymphocytes (Fig. 3A). AND T cells did not
recognize BCG as measured by the absence of increased activation marker
expression after 3 days of in vitro culture in the presence of BCG (see
Fig. 4C). The timing and
morphology of granuloma formation in AND mice in response to BCG
infection were indistinguishable from those in our nontransgenic
controls (data not shown).
|
A shows a typical flow cytometric profile in which the
ratio of transgenic to nontransgenic T cells in a lymphocyte gate of
total splenocytes from an AND mouse is 
25:1 (left
panel). After infection with wild-type BCG, that ratio falls to 3
or 4:1 in the spleen after 6 wk as the nontransgenic T cell population
expands (middle panel). Significantly, the ratio of
transgenic to nontransgenic T cells reverses in the
granuloma-infiltrating lymphocytes to 1:7 (right panel),
suggesting that TCR specificity plays a role in localization to
inflammatory sites.
We also characterized the activation status of granuloma-infiltrating T
cells gated into transgene-expressing and nontransgene-expressing
groups. Fig. 3B shows that in the infected spleen, a larger
fraction of V3-negative T cells (second panel) expresses
a CD44high phenotype than do the transgenic T
cells (first panel), which are very comparable to
CD44 staining of naive transgenic T cells (not shown). However, in
liver granuloma cells, >90% of the V3-negative T cells
(fourth panel) express a
CD44high phenotype, and the transgenic cells
found in the granuloma (third panel) also have a
greater level of CD44 expression than those found in the spleen
(first panel). Overall, Fig. 3 indicates that the
TCR repertoire that accumulates in BCG-induced granulomatous lesions is
regulated by Ag specificity, activation phenotype, or both and is not a
haphazard process. Significantly, either V3+
or V3- T cells that accumulate in the
granuloma are more activated than those found in the spleen. The level
of V3+ T cells that accumulate in granulomas
is low. Interestingly, and unlike the wild-type cells (data not shown),
we see T cells with a virgin (CD44low) phenotype
in the granulomas induced in TCR transgenic animals. This indicates
that virgin T cells can localize to granulomas and also suggests that
some T cells may be primed in the granuloma. Transgenic cells with an
activated phenotype (CD44high) in the granuloma
may be activated elsewhere via bystander mechanisms and home to the
inflammatory site without Ag specificity. Alternatively, because the
TCR transgene is not on a Rag-/- or SCID
background, TCR transgenic T cells that express a second V-chain
allele on the cell surface may be accumulating in the granuloma. In two
experiments using a subset of V-chain-specific Abs, we found that
the percentage of double-positive TCR-bearing T cells was higher in
granuloma-infiltrating T cells than in splenic T cells (Table I). Whether these double-positive cells
are responsible for some or all of the CD44high
transgenic T cells in granulomas (Fig. 3B) remains to be
determined, but their overrepresentation in the granuloma lends further
support to the role of Ag specificity in T cell accumulation in the
granuloma.
|
The studies outlined to date indicated that in granulomas,
infiltrating T cells have a diverse, but selected, TCR repertoire. The
next question we asked was how much of this large, diverse, regulated T
cell response is required for granuloma formation and function? Could
single monospecific TCR-bearing T cells confer protection against a BCG
infection? We constructed the following model system to answer this
question. A 17-aa PCC epitope was cloned into the carboxyl-terminal end
of the 19-kDa mycobacterial lipoprotein gene to target PCC to the
surface of the recombinant BCG (Fig. 4; see Materials and
Methods). The presence of PCC peptide in the 19-kDa membrane
lipoprotein in the transfected bacteria was first detected by PCR (Fig. 4B). Then, in vitro spleen cell activation and surface
phenotype assays were used to confirm the expression of PCC and its
recognition by AND TCR transgenic T cells (Fig. 4C). Culture
of transgenic T cells in medium alone served as a negative control for
background activation, while PCC-stimulated cultures showed the full
activation potential. The third row of graphs shows the stimulatory
effect of wild-type BCG. Finally, row 4 illustrates the effect of
recombinant BCG-PCC in a 2-day coculture with AND splenocytes. A shift
in the population toward large cells (forward scatter on panels of the
first column) and a larger proportion of V3+
TCR transgenic cells (second column, small square) are basic
indications of activation. Higher levels of LFA-1 (activation marker)
and transferrin receptor (cell cycle marker) on lipo-PCC
rBCG-stimulated cells indicate T cell recognition of the PCC epitope.
It is clear that the PCC epitope is expressed by this rBCG clone and
that it is recognized by AND transgenic T cells.
We infected 5CC7/Rag2-/- mice with lipo-PCC
rBCG. This strain of mouse lacks both endogenous B and T cells due to
its lack of Rag2. In these animals, every T cell expresses the 5CC7
transgenic TCR that recognizes the PCC epitope (28).
First, we examined the morphology of the infected liver in these
animals. Fig. 1B shows a comparison of thin liver sections
from B10.BR, B10 Rag2-/-, and
5CC7/Rag2-/- mice infected with lipo-PCC rBCG.
The left column of pictures are of H&E-stained sections at
x100 magnification to show the scope of the granulomatous response.
Arrowheads indicate the large perivascular lesions found in the B10
Rag2-/- infected mice in contrast to the
granulomatous response seen in the B10.BR and
5CC7/Rag2-/- infections. It is clear that the
single T cell in the 5CC7/Rag2-/- is sufficient
to induce granulomatous lesion formation. The middle
column of pictures in Fig. 1B contains
representative pictures of H&E-stained sections at x1000
magnifications to illustrate differences in the morphology of the
lesions and their component cells. B10 Rag2-/-
lesions typically lack defined borders and are dominated by
neutrophils, while granulomas in B10.BR and
5CC7/Rag2-/- infections contain the
inflammatory cells within distinct zones and protect the surrounding
tissue. The Ziehl-Neelsen staining at x1000 magnification in the
right column shows the numbers of acid-fast bacteria visible
within the different inflammatory lesions. Arrowheads point to
individual rods or groups of rods and are proportional to the number of
bacteria present. Numerous examples of bacteria or infected macrophage
with no surrounding inflammatory cells can be found in the B10
Rag2-/- infected animals. The morphology of
these sections indicates that a single T cell in the
5CC7/Rag2-/- is sufficient to control bacterial
replication in lesions.
We quantified our observations by directly counting the number of acid
fast bacteria at x1000 magnification in a thin liver section stained
by the Ziehl-Neelsen method. Fig. 5 shows
that the number of bacteria per granuloma or inflammatory lesion is 4-
to 5-fold higher in B10 Rag2-/- mice infected
with lipo-PCC rBCG than in similarly infected B10.BR or
5CC7/Rag2-/- mice. Thus, we conclude that the
presence of a single TCR specificity can control bacterial replication
within granulomas (p < 0.0001, by one-way
ANOVA).
|
3 T cells (boxed region; Fig. 6). The top row of Fig. 6
shows a representative nontreated Rag2-/-
infection, and the bottom row shows a
Rag2-/- rescued by the monospecific T cells.
One week after transfer, the formation of granulomas in comparison to
nontreated infected Rag2-/- mouse controls was
assessed. The results of this study were very similar to the native
infections shown in Fig. 1B. We clearly saw that granuloma
formation capable of controlling bacteria is initiated within 1 wk
after transfer of naive 5CC7 T cells to BCG-infected
Rag2-/- mice. Ziehl-Neelsen staining for
acid-fast bacteria is shown on the right side of Fig. 6. The
granuloma shown for the recipient mouse contains no bacteria, which
although nonrepresentative, illustrates the extent of bacterial control
possible at isolated lesion sites after only 1 wk of exposure to
Ag-specific T cells. Most other granulomas contained a few bacteria,
which is comparable to granulomas in wild-type animals (0.73 ±
0.18 vs 22.8 ± 4.0 bacteria/granuloma in B10
Rag2-/-; Fig. 5).
|
shows the flow cytometric analysis
of spleen cell suspensions and granuloma cell suspensions from B10.BR
and 5CC7/Rag2-/- mice infected with lipo-PCC
rBCG. Cells are analyzed from a CD4+ gate to
compare the bulk population of T cells from B10.BR to the transgenic
V3+ T cells from the
5CC7/Rag2-/- mice. B10.BR mice show an
accumulation of activated CD44high
L-selectinlow cells in both spleen and granuloma
cell populations consistent with that seen in C57BL/6 mice (data not
shown). 5CC7/Rag2-/- granuloma cell populations
also contain activated CD4+ T cells (31.5%).
Splenic T cells were less activated in
5CC7/Rag2-/- infected mice, possibly reflecting
the monospecific nature of the T cell-activating Ag, the restriction of
expressed PCC to the bacterial surface, or, alternatively, the more
extensive recruitment of activated cells to granulomas. Hence, a
monospecific T cell is sufficient to promote accumulation of activated
T cells in the granulomatous inflammatory site and is sufficient for
the formation of protective, well-formed granulomas observed in thin
liver sections (Fig. 1B).
|
activity. IFN- production
by activated T cells is an important means of promoting
anti-bacterial macrophage functions, and thus class II expression
is an excellent surrogate marker of biologically relevant IFN-
function. Fig. 8 shows that B10.BR and
5CC7/Rag2-/- mice infected with lipo-PCC rBCG
contain equivalent numbers of gated macrophage with elevated class II
expression in both splenocytes and granuloma cell populations, and that
granuloma cell populations contain 2 times as many macrophage with
elevated class II. In contrast, B10 Rag2-/-
mice infected with lipo-PCC rBCG contain reduced levels of class
II-expressing macrophage in spleen (2- to 3-fold less) and in cells
recovered after a liver lymphocyte preparation (7-fold less). The
production of IFN- by 5CC7 granuloma cells and spleen cells in
response to CD3 during in vitro culture was comparable to the
production by wild-type cells (ELISA measurements of 50.6 ng/ml from
1 x 106 5CC7 granuloma cells and 8.8 ng/ml
from 1 x 106 5CC7 spleen cells). Clearly,
mice containing T cells with a single Ag specificity are able to
activate macrophage both systemically and locally comparably to
parental strains of mice after BCG infection.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(24, 39), IFN- (16, 21, 22, 23, 24, 39), and cytokines that regulate IFN- levels (18, 19, 40, 41) are essential for both forming a protective structure
capable of walling off the inflammatory response and for killing the
sequestered bacteria. Protective granuloma formation requires CD40
ligand function on T cells, and CD40 ligand function is probably
related to the regulation of IFN- levels (42, 50). In
the present study we observed a large and diverse TCR repertoire in
liver granulomas (Fig. 2). This is true not only examining bulk
granuloma preparations, but also examining isolated single granulomas.
Granulomas may share T cells of similar specificities, a statement
needing further refined studies already underway, but each granuloma
also appears to have a diverse and unique TCR repertoire. The
heterogeneity is obvious at the level of J usage (Fig. 2D). These data seem to favor a model in which a large
number of peripherally activated T cells home to single lesions,
instead of a limited number of T cells expanding in situ.
The large T cell repertoire present in granulomas does not mean that
there is not a TCR-dependent selection in the accumulation of
granuloma-infiltrating T cells. AND TCR transgenic T cells that do not
recognize BCG (Fig. 3) are preferentially excluded from granulomas.
Additionally, among granuloma-infiltrating cells, the proportion of AND
cells that express additional endemic V-chains ranges from 1.5 to 9
times higher than the proportion in splenocytes (Table I). These
findings agree with data suggesting selection at local sites in
Leishmania granulomas (43). In those studies
Pingel and coworkers used altered peptide ligands to demonstrate that
the TCR repertoire of the T cell response to a single dominant Ag of
Leishmania is of limited diversity.
Our results established that in BCG-induced granulomas, T cells have a diverse, but selected, TCR repertoire. Our next experimental question was how much of the observed TCR heterogeneity in response to BCG infection is required for CD4+ T cell function? Put another way, can a single specificity CD4+ T cell population induce adequate protective granulomas after BCG infection? Our model system for this work consisted of infection of PCC-specific TCR transgenic mice on a Rag2-/- background with a recombinant BCG strain expressing a PCC T cell epitope in the bacterial 19-kDa surface lipoprotein.
Expression of recombinant Ag by infectious agents has been analyzed
using viruses, parasites, and bacteria to model autoimmune disease and
immune responses to infection. CD8+ TCR
transgenic T cells are unresponsive to expression of lymphocytic
choriomeningitis virus (LCMV) glycoprotein (gp) by pancreatic
islet cells in LCMV-gp TCR transgenic mice. Infection by LCMV or
recombinant gp-expressing vaccinia virus is required to directly
activate sufficient numbers of previously self-tolerant T cells for
disease induction (44, 45). Reiner and coworkers showed
that TCR transgenic mice with a single T cell repertoire
specific for Leishmania major LACK Ag were able to establish
substantial, albeit incomplete, control of experimental L.
major footpad infections (46). This was consistent
with the ability of altered peptide ligands to confer a healer
phenotype on susceptible strains of mice by modulating the interaction
of LACK-specific TCR with cognate I-Ad molecules (43).
Using the 19-kDa lipoprotein-PCC rBCG infection of PCC-specific TCR
transgenic mouse model we examined the function of cells at local
inflammatory sites in the liver and spleen and assessed both T cell and
macrophage surface phenotypes locally (Figs. 7 and 8). Histopathology
demonstrated the development of protective granulomatous lesions able
to contain and eliminate bacteria (Figs. 1, 5, and 6). In placing our
study among the spectrum of examples, the 5CC7 TCR transgenic T cells
are more activated than tolerant, in that they are able to induce a
fully protective local inflammatory response. T cell surface staining
consistently indicated the presence of activated T cells at the local
inflammatory site (Fig. 7), varying from 30–60% over four
experiments. Isolated granuloma 5CC7 cell preparations are responsive
to CD3 in culture and produce IFN-, suggesting that at least a
fraction of the T cells primarily responsible for IFN- production
are not anergic. Thus, a single TCR:Ag interaction is able to activate
T cells for localization to an inflammatory site (Fig. 7) and confer
adequate levels of protective inflammation. That protection consists of
IFN--activated macrophage (Fig. 8), formation of granulomas (Figs. 1 and 6), and eradication of bacteria (Fig. 5). The activated phenotype
of T cells in this model is likely to result from the systemic and
chronic infection by BCG, so that lipoprotein-PCC Ag is continuously
available at a low concentration to the immune system.
We considered whether part of the response of 5CC7 mice to the lipo-PCC
rBCG strain might be specified by the mycobacterial 19-kDa lipoprotein
fusion partner. Abou-Zeid and coworkers found that immunization of mice
with recombinant Mycobacterium vaccae expressing the
Mycobacterium tuberculosis 19-kDa Ag resulted in lower
levels of protection than immunization by recombinant M.
vaccae transformed with vector alone despite eliciting a strong
Th1 immune response (47). This was assessed using both
high dose lethal challenge and low dose aerosol challenge with virulent
M. tuberculosis. This evidence suggests that the 19-kDa
lipoprotein can exert a systemic immunosuppressive effect. However, one
component of the strong Th1 immune response was a high level of IFN-
production from splenocytes cultured in vitro after immunization. In
this context, the local sequestration of mycobacteria in the liver
granulomas could increase the local production of IFN- leading to
enhanced killing of bacteria by macrophage. Because BCG also expresses
a 19-kDa lipoprotein on its cell surface with sequence identity to the
M. tuberculosis 19-kDa Ag (48), our working
assumption was that the 19-kDa fusion partner for PCC epitope would be
immunologically neutral. However, we cannot exclude the possibility
that the recombinant PCC fusion may be expressed at high enough levels
on the mycobacterial cell surface to either inhibit or enhance host
immune responses.
Even if lipoprotein-mediated inhibition of host immune responses is
present, infection of 5CC7/Rag2-/- mice by
lipo-PCC rBCG leads to activation of T cells and formation of
protective granulomas. We have overwhelming data that a population of T
cells with a single TCR specificity can induce protective granulomas.
The histology is strikingly different when comparing the B10
Rag2-/- infection and the
5CC7/Rag2-/- infection at 6 wk (Fig. 1). B10
Rag2-/- infection by BCG leads to the formation
of huge perivascular lesions dominated by neutrophils and dying cells
and containing very large numbers of acid fast bacteria. The presence
of the single Ag-specific TCR-bearing T cell in the 5CC7 mouse allows
the formation of organized granulomatous inflammation, which controls
the numbers of visible bacteria per lesion to the level seen in B10
mice containing a broad spectrum of TCR-bearing T cells (Figs. 1 and 5). The ability of 5CC7 T cells to initiate granuloma formation and
control of bacteria only 1 wk after adoptive transfer into infected B10
Rag2-/- mice (Fig. 6) also indicates that the
difference in phenotype is specific to T cells. The acquisition of an
activated phenotype in granuloma-infiltrating cells (Fig. 7) and the
up-regulation of class II expression on macrophage (Fig. 8) indicate
the presence of local protective IFN- and are a functional
demonstration that protection can be mediated by a single specific TCR.
This occurs despite the absence of CD8+ T cells
and is consistent with the partial protection observed after BCG
infection of 2-microglobulin-deficient mice
(49) Moreover, the low bacterial numbers in the granulomas
indicate that IFN- levels are sufficient to induce mycobacterial
killing
Our study points toward the conclusion that a single TCR specificity
can provide protective granuloma formation. However, it does not
address the question of whether the inducing Ag must be present on the
pathogen or in the granuloma. In fact, when we performed control
experiments in which 5CC7/Rag2-/- mice were
infected with wild-type BCG, we saw liver granuloma formation similar
to that in the rBCG-PCC infection of
5CC7/Rag2-/- mice shown in Fig. 1, but in which
the bacterial load was less well controlled. The bacterial load was
consistently statistically intermediate between the rBCG-PCC infection
of 5CC7/Rag2-/- mice and that of B10
Rag2-/- mice (mean ± SEM, 3.6 ± 0.4
vs 7.8 ± 1.1 bacteria/granuloma for rBCG-PCC vs wild-type BCG
infections, respectively; by Student’s t test,
p < 0.05). Overall, our data indicate a less black and
white picture in which nonbacterial specific T cells can also promote
granuloma formation, but the presence of an Ag recognized by the
monoclonal T cell on the granuloma-inducing pathogen provides better
protection. The issue is further complicated by the absence of
granuloma formation or protection (a Rag-/-
phenotype) we observed after wild-type BCG infection of other TCR
transgenic/Rag-/- strains of mice. A detailed
analysis of these data is currently underway in our laboratory. Our
working hypothesis is that T cells activated outside the granuloma even
without granulomatous specificity can induce granuloma formation and
confer partial protection.
These studies suggest the possibility that expanded monoclonal Mycobacterium-specific T cells may be protective for immunodeficient patients. We will extend this work to the analysis of T cell expansion and specificity in single granulomas and the use of Ag to modulate protective granulomas at chronic infection stages. This model using monoclonal T cells will be helpful in elucidating the mechanisms by which T cells initiate, regulate, and organize granulomatous immune responses. Proof of the principle that expanded monoclonal Mycobacterium-specific T cells can be therapeutic may open the possibility of manipulating the immune response clinically in persons already infected with chronic inflammatory agents, including M. tuberculosis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Laura Hogan, Room 5580 MSC, 1300 University Avenue, Madison, WI 53706.
3 Abbreviations used in this paper: BCG, Mycobacterium bovis bacillus Calmette-Guérin; PCC, pigeon cytochrome c; H&E, hematoxylin and eosin; Rag2, recombinase-activating gene 2; gp, glycoprotein; LCMV, lymphocytic choriomeningitis virus.
Received for publication February 13, 2001. Accepted for publication March 2, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
/ and /
T lymphocytes to immunity against Mycobacterium bovis bacillus Calmette Guerin: studies with T cell receptor-deficient mutant mice. Eur. J. Immunol. 25:838.[Medline]
is required in the protective immune response against Mycobacterium tuberculosis in mice. Immunity 2:561.[Medline]
in resistance to Mycobacterium tuberculosis infection. J. Exp. Med. 178:2249. genes. Science 259:1739. in host defense against mycobacteria and salmonella in mice and men. Curr. Opin. Immunol. 11:346.[Medline]
gene-disrupted mice. J. Exp. Med. 178:2243. receptor have profoundly altered responses to infection with bacillus Calmette-Guerin and subsequent challenge with lipopolysaccharide. J. Exp. Med. 178:1435.-deficient mice infected with mycobacteria. Blood 91:2914. T cells expressing different V delta genes are recruited into schistosome-induced liver granulomas. J. Immunol. 155:275.[Abstract]
chain of the T-cell antigen receptor. Science 239:1541. interferon production in human tuberculosis. Infect. Immun. 68:3002.. J. Immunol. 150:5185.[Abstract]
T cell repertoire. J. Immunol. 160:884. interferon and CD40 ligand but does not require CD28. Infect. Immun. 69:2596.This article has been cited by other articles:
|
|
 |
|
  L. H. Hogan, E. Heninger, R. A. Elsner, H. A. Vonderheid, P. Hulseberg, D. Co, and M. Sandor Requirements for CD4+ T cell levels in acute Mycobacterium bovis strain bacille Calmette Guerin (BCG)-induced granulomas differ for optimal mycobacterial control versus granuloma formation Int. Immunol., May 1, 2007; 19(5): 627 - 633. [Abstract] [Full Text] [PDF]
|
), infected spleen (
), and granuloma cells (
)
are shown for each of four V
