| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Trudeau Institute, Saranac Lake, NY 12983
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have previously reported that following infection with BM5, the entire CD4+ T cell compartment changes phenotype within 6–8 wk, with the cells acquiring an activated/memory phenotype (L-selectinlow, CD45RBlow, CD69high, and CD44high) (7). CD4+ T cell numbers increase and there is direct evidence from in vivo bromodeoxyuridine (BrdU) incorporation studies that a large fraction of T cells divide (8). However, despite this polyclonal response, the CD4+ T cell compartment loses the ability to respond in vitro to polyclonal or antigenic stimuli as measured by proliferation or cytokine production (1, 7).
This dramatic activation of T cells is already detectable within 1–2 wk postinfection. Some B cell changes such as hypergammaglobulinemia and enhanced in vivo B cell proliferation also occur early during MAIDS. However, it is not until late in disease (around week 12) that the loss of B cell function and also the development of lymphomas become apparent (2, 6, 7), raising the possibility that the T cell changes precede those in the B cells and may in fact induce them.
Immediately following infection, B cells are the major target of the virus infection, but macrophages are also found to bear virus once infection is underway (9, 10). However, the number of infected B cells required for the characteristic loss in T cell function is low, and only small clusters of infected cells in the lymph nodes or spleens are detected at week 6 of infection when much of the loss of CD4+ T cell function is already apparent (9). Only a very few, if any, T cells are found to be infected at any time (9). This suggests that both the broad polyclonal effects of BM5 on virtually all CD4+ T cells and the polyclonal B cell activation are not the direct result of viral infection.
An intriguing feature of MAIDS is that disease, at least as determined from dramatic splenomegaly and hypergammaglobulinemia, does not develop in nude mice without T cells (11) or in animals which have been depleted of CD4+ T cells by in vivo Ab treatment (12). Animals depleted or devoid of B cells are also resistant to MAIDS (10, 13). MHC class II knockout (KO) animals, which lack CD4+ T cells, are also resistant to disease (14). One receptor-coreceptor interaction involved in most T-B interactions is that involving the CD40 ligand (CD40L) (gp39) on the T cell with CD40 on the B cell (15). Recent studies indicate that administration of Ab to gp39 (CD40L) blocks disease progression (16), further supporting the requirement for a T-B interaction in disease. Interactions of integrins with their coreceptors are also nearly universally involved in T cell responses. A requirement for LFA-1 (CD11a/CD18) interaction with its ligands (ICAM-1 or ICAM-2) is particularly well established as an adhesive interaction, but can also serve to costimulate CD4+ T cell responses (17). Ab to CD54 (ICAM-1) has recently been shown to slow down the progression of MAIDS (18), lending further support to the hypothesis that a T-B interaction, perhaps like those involved in induction of the Ab response, plays a key role in disease.
The mechanisms of MAIDS development and progression remain unclear.
There is growing evidence that neither specific Ag (19)
nor conventional superantigens are involved in
CD4+ T cell activation, which occurs (7, 20, 21), despite earlier suggestions of superantigen involvement
(14, 22). We found no detectable
V
selectivity in vivo but rather a completely
polyclonal CD4+ T cell response (7, 21). Furthermore, we showed that
V3-expressing transgenic mice were comparable
to nontransgenic wild-type mice in their susceptibility to MAIDS
(19), further supporting the absence of a requirement for
any epitope-specific TCR recognition.
A difficulty in interpreting earlier T cell depletion studies and studies of MHC class II KO mice was that, because all CD4+ T cells in such mice were deleted, the characteristic changes in T cells of the CD4 lineage could not be followed. Moreover, since such mice lack class II, introduced CD4+ T cells do not behave normally. Thus, although it was clear that splenomegaly and hypergammaglobulinemia did not develop, the stage in disease affected by the deficiency was not clear. Therefore, we chose to investigate MAIDS progression in a system where substantial numbers of CD4-like T cells persisted, but conventional CD4-class II interaction which contributes to T-B interaction is missing: the CD4 KO mouse. The CD4 KO mice contain a peripheral Thy-1+, TCR+, non-CD4+, non-CD8+, class II-restricted Ag-specific helper T population (23, 24), which apparently is selected because of the substantial avidity of the TCR for self-MHC in the absence of CD4 or CD4-like (LAG-3) coreceptor interactions (25). These double-negative (DN) CD4-/CD8- T cells found in the periphery have a substantial TCR repertoire as indicated by responses to some intense stimuli such as infectious organisms (23). However, the responses seen are of consistently lower magnitude than comparable responses in wild-type mice. Hence, we could investigate the effect of viral infection on the DN T cell population under conditions where reduced T-B interaction is expected due to the lack of the CD4 molecule interaction with class II on the B cells. At the same time, we could evaluate whether the dramatic changes in B cell function which are seen in MAIDS are induced when CD4 is absent.
We find that BM5 efficiently infects CD4 KO animals. The CD4-, CD8-, TCR+ lymphocyte population undergoes several of the changes associated with the normal course of MAIDS, including the loss of the ability to respond to in vitro stimulation as judged by proliferation or production of IL-2 following TCR stimulation. These T cell changes occur at a rate only slightly slower than in wild-type mice. However, the B cell population remains virtually unaffected. In vivo B cell proliferation and expansion, hypergammaglobulinemia, and later loss of B cell response manifested in splenomegaly are absent, as are dramatic shifts in the B cell phenotype that occur in wild-type animals with MAIDS that we report here. Hence, in CD4 KO animals, we observe the uncoupling of the induction of T cell immunodeficiency from the induction of B cell responses, suggesting there is a stepwise mechanism of disease in which interaction of CD4+ T cells with B cells is implicated in functional changes in the B cell compartment.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B6 mice were either obtained from The Jackson Laboratory (Bar Harbor, ME) or bred in our own facilities at University of California, San Diego or subsequently at the Trudeau Institute. Adult mice (6–8 wk of age) were injected i.p. with 0.5 ml of BM5 viral stock and compared with age-matched, uninjected control mice. Breeding pairs of CD4 KO animals were a kind gift from Dan Littman’s laboratory (24). They were generated on a 129 background and backcrossed nine times to B6 for use in our experiments. BM5 viral stocks were obtained as cell-free supernatants of chronically infected SC-1 cells (26).
Reagents
Medium used for all in vitro cultures was RPMI 1640 (Irvine Scientific, Santa Ana, CA) supplemented with 200 µg/ml penicillin, 200 µg/ml streptomycin, 4 mM L-glutamine, 10 mM HEPES, 5 x 10-5 M 2-ME, and 7.5% FCS (HyClone Laboratories, Logan, UT). BrdU was purchased from Sigma (St. Louis, MO).
Ab and recombinant cytokines
Cell lines were maintained for mAb production as previously described (27) and were used for cell depletions. These Ab include Ab to: Thy-1.2 (F7D5 and HO 13.14), CD8 (HO2.2), CD4 (RL172.4), class II (Iab/d; D3.137), and anti-HSA (J11D). Hybridomas secreting the Ab were obtained from the American Type Culture Collection (Manassas, VA) unless otherwise indicated, and Ab were partially purified from either culture supernatants or from ascites produced in BALB/c nu/nu mice. For IL-2 cytokine assays, 11B11 (anti-IL-4) was obtained as ascites from nude mice (28) and added to block the potential activity of IL-4 in the bioassay of proliferation of the NK.3 responsive line. Human rIL-2 was obtained from Cetus (Emeryville, CA) and rIL-4 from the X63.Ag8-653 line that was transfected with the murine cDNA for IL-4. Polyclonal rabbit anti-mouse IgM (Fab')2 was purchased from Cappel (West Chester, PA). Monoclonal rabbit anti-CD38 Ab (NIMR5) was kindly provided by Maureen Howard (Corixa, Redwood City, CA). Anti-CD40 Ab was used in the form of a concentrated tissue culture supernatant prepared from the 1C10 cell line.
In vivo BrdU labeling
BrdU (Sigma) was added to the drinking water of control and infected mice at a concentration of 1 mg/ml 5 days before harvesting spleens. The incorporation of BrdU was determined as described elsewhere (29). Splenocytes were labeled with PE-conjugated anti-B220 to detect B cells, washed, and fixed in ethanol and paraformaldehyde, treated with DNase, and stained for incorporated BrdU using FITC-conjugated anti-BrdU Ab (Sigma). Cells were analyzed immediately on a FACScan using LYSIS II software (Becton Dickinson, Mountain View, CA).
Preparation of CD4-/CD8- T cells and APC
Spleen cells were depleted of CD8+ T cells and B cells by treatment with anti-CD8, anti-J11D, and anti-class II Ab plus complement. Ab treatment was conducted on ice for 30 min followed by a combination of rabbit and guinea pig complement at 37°C for 45 min as previously described (19, 27). This treatment enriches for the CD4+ T cells in normal B6 animals and for the CD4-/CD8- T cells in the CD4 KO animals (typically 50% of the cells obtained). The absence of B cells and CD8+ T cells in the resulting populations was confirmed by FACS staining with anti-CD8 and anti-B220 Ab and purities typically ranged from 90 to 95%. For APC populations, spleen cells from uninfected B6 animals were depleted of T cells by treatment with two anti-CD8 and two anti-Thy1.2 Ab plus complement. The resulting cell population was treated with 50 µg/ml mitomycin C to prevent proliferation.
Preparation of B cells
Mice were sacrificed at various times after BM5 infection, and splenic B cells were purified by negative selection using magnetic microbeads (Miltenyi Biotec, Sunnyvale, CA). Briefly, spleens were removed, weighed, and gently teased apart in PBS containing 5% FCS. RBC were lysed by treatment with ammonium chloride. Cells (107/ml) were incubated on ice for 5 min with Fc block (CD16/CD32) and for another 15 min with a mixture of biotinylated Ab specific for CD4, CD8, CD3, CD11b, CD5, NK cells (DX-5), and granulocytes (GR-1). Abs were purchased from PharMingen (San Diego, CA). Cells were washed and then incubated on ice for 15 min with streptavidin-labeled magnetic microbeads. Cells were washed, filtered to remove debris, and then loaded onto magnetic columns in a volume of 5 ml of PBS/FCS. B cells were isolated by negative selection using a MACS system according to the manufacturer’s protocol (Miltenyi Biotec). Eluted cells were washed three times in complete tissue culture medium and used for flow cytometry or in vitro proliferation assays.
T cell proliferation and cytokine production
Both T cell proliferative responses and cytokine production to plate-bound anti-CD3 (10 µg/ml) with the addition of PMA (10 ng/ml) were determined. T-depleted splenocytes from uninfected B6 animals were used as APC. T cell proliferation was determined by mixing 1–2 x 105 (DN) T cells and 3–5 x 104 mitomycin C-treated APC in triplicate cultures of 200 µl. The T cell preparations were adjusted to reflect the percentage of T cells determined by FACS analysis, so comparable T cell numbers were added to cultures being compared. Proliferation was measured over 72 h and [3H]thymidine (NEN, Boston, MA) was added for the last 18 h. Cultures were harvested onto glass fiber filter paper using a multiple sample harvester, and bound radioactivity was determined. T cell IL-2 production was determined by testing supernatants of T cells stimulated as above. CD8 and B cell-depleted (DN) T cells were cultured at 1–2 x 106/ml with control APC at 5 x 105–2 x 106/ml in cultures of 1 or 2 ml in 24- or 48-well plates. Culture supernatants were harvested after 36 h. Supernatants were stored at -20°C before analysis of cytokine content. Bioassays for IL-2 and the method of data analysis have been previously described (28). Briefly, IL-2 was assayed using the NK indicator line, which is responsive to IL-2 as well as IL-4. To measure IL-2 content of the supernatants, 11B11 (anti-IL-4) was used to block IL-4 activity.
B cell proliferation assay
Purified B cells in complete tissue culture medium were aliquoted in triplicate (2 x 105 cells/well) into flat-bottom microtiter plates containing various stimuli or medium. After 48 h, cells were labeled with 0.2 µCi/well [3H]thymidine and 18 h later harvested onto filter mats and counted on a Wallac 1205 Betaplate counter (Wallac, Gaithersburg, MD).
Flow cytometry
All staining was done at 4°C in PBS with 1% BSA and 0.1% NaN3. Cells were preincubated for 5 min with Fc block (CD16/CD32) and then stained with PE-conjugated Ab directed at various cell surface molecules and FITC-labeled anti-B220 (PharMingen). Isotype controls were included to control for nonspecific staining and propidium iodide was added to allow gating on live cells. Flow cytometry was performed on a FACSCan (Becton Dickinson), and data were analyzed with CellQuest software unless stated otherwise.
Serum Ig determination
Groups of mice (n = 3) were bled individually at various times following BM5 infection and serum was collected. Serum Ig levels were measured by isotype-specific ELISA as previously described (30).
RT-PCR analysis of BM5def
RT-PCR analyses of BM5def and glucose-6-phosphate dehydrogenase (G6PD) transcripts in spleen cells from B6 or CD4 KO mice infected with BM5 murine leukemia virus mixture were performed. Isolated cDNA from 106 splenocytes was used for each PCR under the following conditions: 95°C for 1 min, 55°C for 2 min, and 72°C for 1 min for 40 cycles. The primers for amplification of the BM5def transcripts were 5'-CCTTTTCCTTTCTCGACACT-3' (sense) and 5'-ACCAGGGGGGGAATACCTCG-3' (antisense). The primers for amplification of the G6PD transcripts were 5'-GGTGACCTGGCCAAGAAGAT-3' (sense) and 5'-GCATTCATGTGGCTGTTGAG-3' (antisense). The 245-bp BM5def product and 260-bp G6PD product were visualized in an ethidium bromide-stained 2% agarose gel.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Establishment of infection in CD4 KO animals
Since the main targets of retrovirus infection are B cells, and to
some extent macrophages (9, 10), we considered it likely
that initial infection with the virus would be successful in the CD4 KO
host. Moreover, Jolicoeur and colleagues (31) reported
successful infection of nude and CD4 KO mice with a related virus
preparation. However, activation of B cells, dependent on
CD4+ T cells, could contribute to the rate or
extent of viral spread so that it was conceivable that CD4 KO mice
might never develop high viral titers. Thus, we evaluated the level of
virus infection in CD4 KO animals compared with B6-infected controls.
The development of MAIDS is completely dependent on the presence of a
defective virus, in the BM5 mixture, called
BM5def, which encodes a distinct gag
P12 region (32, 33). The viral mixture also contains a
helper virus and a mink cell focus-forming virus which help propagate
the defective disease-causing component (32). Therefore,
we compared BM5def transcripts detected by RT-PCR
of spleen cell RNA obtained from infected CD4 KO vs B6 animals to
estimate the extent of infection several weeks postinoculation with the
BM5 virus (Fig. 1
). RNA was prepared from
106 splenocytes from CD4 KO and B6 animals 6 wk
after infection and run through 40 rounds of PCR and products were
analyzed for the presence of BM5def transcripts.
The levels of transcript of the G6PD housekeeping gene were
also determined and used as a control. Both infected groups, B6 and CD4
KO clearly showed the presence of similar levels of
BM5def transcripts, whereas both uninfected
control groups were negative. Although subtle quantitative differences
in viral load would not be detected by this technique, this analysis
suggests that the absence of CD4 expression does not have a major
impact on viral load.
|
Interestingly, BM5-infected CD4 KO mice showed relatively little
splenomegaly (Fig. 2a) or
lymphadenopathy (data not shown) when compared with BM5-infected B6
animals. In a kinetic analysis, we measured the weight of spleens of
infected and uninfected CD4 KO and B6 animals. Only the BM5-infected
wild-type B6 mice showed the dramatic increase in spleen weight
associated with MAIDS disease with time (Fig. 2a). A very
modest, nonprogressing splenomegaly was detected in infected CD4 KO
mice, and even by week 14 the infected CD4 KO mice had spleens only
slightly larger than normal. Similar kinetics has been seen in four
sets of experiments.
|
b). The most
dramatic differences are seen between CD4 KO and wild-type mice. CD4 KO
mice differ from wild-type mice in the absence of
CD4+ T cells and a nearly compensatory increase
in DN T cells. In wild-type animals BM5 infection induces a relative
increase in CD4+ T cells (see also Ref.
7) as well as a dramatic total increase due to increased
lymphocyte recovery (data not shown), a rather dramatic reduction in
CD8+ T cells but no change in the undetectable
percentages of DN T cells, and little change in relative B cell
numbers. In the infected CD4 KO animals, the DN (CD4-like population)
stays the same or increases slightly, the CD8 population does not
decrease as much as that observed in wild-type B6 mice and B cells are
also unchanged. Thus, the MAIDS-induced changes in splenomegaly and
lymphocyte distribution seen in spleens of wild-type mice are largely
absent in CD4 KO mice. Functional changes in DN T cells from CD4 KO animals
Analysis of the cell surface phenotype of the DN T cell population from infected vs uninfected CD4 KO mice showed no increase in CD69 expression over the negligible levels in uninfected controls, no size increase as determined by increased scatter (16% blasts in uninfected vs 14% in infected), and the majority of DN cells retained CD45RBhigh expression (77% in uninfected vs 62% in infected; data not shown). Thus, there was much less evidence of activation compared with the results seen in CD4+ T cells in infected B6 mice (7) where a majority of CD4+ T cells in infected mice express CD69, express only low levels of CD45RB, and are enlarged.
The DN T cell population in CD4 KO mice shows most attributes of normal
CD4+ T cells in that they produce IL-2 after
activation and can help B cells produce Ab following priming with
specific Ag (23). Therefore, we analyzed the ability of
the DN T cell population to respond in vitro to TCR stimulation by
measuring their ability to proliferate (Fig. 3a) and synthesize IL-2 (Fig. 3b). Splenocytes were depleted of B and
CD8+ T cells and the resulting population
containing the CD4-like DN TCR+ T cell population
(usually <2% CD8+ cells) was stimulated using
plate-bound anti-CD3 plus PMA in the presence of mitomycin
C-treated APC derived from uninfected control B6 animals as a source of
costimulatory signals. Cells were stimulated for 3 days and then
assayed for DNA synthesis by addition of radiolabeled nucleotide.
Representative data demonstrating the proliferative response of cells
from infected and uninfected CD4 KO mice only is shown in Fig. 3a at week 10 after BM5 infection. DN T cells from infected
CD4 KO mice showed decreased proliferation in response to anti-CD3
stimulation compared with the DN population from uninfected control CD4
KO mice. Graded numbers of DN T cells from BM5-infected mice
incorporated significantly less radiolabel at all cell densities
tested. Similar results have been seen in two repeat experiments. At
this time of infection, CD4+ T cells from B6
MAIDS-infected animals showed little or no residual response to in
vitro stimulation as reported previously (7).
|
b shows that IL-2 production of
cells from infected mice is reduced by 65–90% at week 10 compared
with IL-2 values of uninfected control animals. By week 14, IL-2
production by the cells from infected CD4 KO mice was reduced by 98%
compared with control values (data not shown). By comparison, the
CD4+ T cells from uninfected B6 animals at a
final concentration of 5 x 105/ml produce
200–500 U/ml IL-2, and CD4+ T cells from
infected wild-type B6 animals do not produce any detectable IL-2 (<10
U/ml) at week 10 (7). This clearly demonstrates that in
infected CD4 KO animals the functional response of the DN T cell
population is lost and that such T cells become unresponsive to in
vitro stimulation at a rate and to an extent that is only slightly less
than that of CD4+ T cells in infected wild-type
B6 animals. DN T cells in CD4 KO mice did not produce detectable levels
of Th2 cytokines (IL-4, IL-5) either before or after infection of the
mice with MAIDS (data not shown). B cell functional changes in CD4 KO mice with MAIDS
In vivo proliferation.
Despite the lack of splenomegaly or accumulation of B cells in CD4 KO
mice infected with BM5, it was possible that B cells were responding
vigorously but then dying at a faster rate than in infected wild-type
mice. To analyze in vivo B cell proliferation during MAIDS in B6 and
CD4 KO animals, BrdU was added to the drinking water of animals
infected with BM5. We then surface labeled the splenic B cell pool with
PE-anti-B220 and counterstained with FITC-anti-BrdU to detect
cells that had divided during the 5-day in vivo labeling. The results
in Table I show the splenic BrdU staining
profile of B cells at 6 wk postinfection. After an initial lag of 
2
wk, in vivo B cell proliferation increased dramatically in B6 mice and
by 6 wk postinfection, 61% of B cells were BrdU positive. In CD4 KO
animals, however, there was little sign of increased in vivo B cell
proliferation as a result of BM5 infection (6-wk infected, 33%
BrdU+; control, 24%
BrdU+). Hence, in vivo proliferation of B cells
is much reduced in animals that do not have T cells expressing a
functional CD4 molecule. At 6 wk postinfection, DN T cells incorporated
more BrdU in infected mice than in uninfected CD4 KO mice, and they
expanded modestly (< 2-fold; data not shown).
|
a). To
examine the kinetics of BM5-induced B cell anergy, we performed the
same experiment with B cells isolated at different time points after
infection (Fig. 4b). The proliferative responses of B6 B
cells declined 50% at 5–10 wk after infection and then decreased
to barely detectable levels by week 14. Responses of B cells recovered
from infected CD4 KO mice declined only slightly at these time points
(Fig. 4b). Parallel studies with anti-CD38 or
anti-CD40 stimulation gave similar results (data not shown).
|
c). Increased levels of IgM, IgG1, and IgG3 were
detected in BM5-infected B6 mice, whereas in the CD4 KO-infected mice
all isotypes were present at lower levels. When comparing Ig levels in
infected CD4 KO mice vs uninfected controls, only IgG3 was increased,
and the total level was equivalent to that present in wild-type
uninfected B6 mice. A kinetic analysis of serum-isotype levels revealed
that B6-infected, but not CD4 KO-infected mice, were
hypergammaglobulinemic at weeks 3, 6, 10, and 14 (data not shown).
IgG2a and IgA were not detected in significant titers in either
infected or control mice (data not shown). Thus, the polyclonal B cell
response which leads to accumulation of serum Ig does not occur in CD4
KO mice with MAIDS.
Phenotypic changes.
Although the entire CD4+ T cell population in
infected mice progresses to a memory/anergic phenotype
(7), no comparable analysis of the B cell phenotype in
MAIDS has been performed. Therefore, we analyzed whether B cells
responding in MAIDS would undergo phenotypic changes and whether such
changes would occur in the CD4 KO mice with MAIDS. The phenotype of B
cells isolated from BM5-infected and uninfected B6 and CD4 KO mice was
analyzed by flow cytometry. At 10 wk postinfection (Fig. 5a), a significant fraction of
the B cells from B6 mice had acquired an "activated/memory"
phenotype characterized by an increase in cell size (elevated forward
scatter), elevated CD44 expression, down-regulation of L-selectin
(CD62L), and isotype switching from IgM to IgG. Other significant
changes were an increase in expression of costimulatory and adhesion
molecules CD80 (B7.1), CD86 (B7.2), CD11b (MAC-1), and CD54 (lCAM-l)
and a decrease in the levels of CD21 and CD23. The elevated expression
of CD43 in infected B6 mice suggests that a fraction of B cells are
undergoing terminal differentiation to form plasma cells, consistent
with the observed hypergammaglobulinemia. A kinetic analysis of the
expression of selected molecules (Fig. 5b) showed that the
most profound changes in phenotype occur during the later stages of
disease. In contrast to the marked changes in B6 mice induced by BM5
infection, the phenotype of B cells from infected and control CD4 KO
mice was very similar at all time points, with no evidence of
significant B cell activation or differentiation (Fig. 5). The dramatic
phenotypic shift in wild-type-infected mice involves the majority of
the B cell population, resulting in the differentiation of a
phenotypically heterogeneous population of activated B cells during the
progression of MAIDS.
|
Normally, wild-type B6 mice die within 15–20 wk of BM5 infection.
Although the causes of death are unclear, it has been suggested that
the enormous increase in the size of parathymic and tracheal lymph
nodes causes asphyxiation. There is also evidence for an increase in
opportunistic infections as a result of immunodeficiency
(34). We evaluated whether CD4 KO animals infected with
BM5, which have little in vivo lymphoproliferation, might be protected
from the lethal effects of MAIDS. Fig. 6
shows a comparison of lethality in B6 vs CD4 KO animals. In this
analysis all of the B6 mice had succumbed to the disease by week 16,
whereas the CD4 KO animals lived dramatically longer (30–40 wk). The
CD4 KO animals infected with BM5 have a shorter life span than their
uninfected controls, dying between 30 and 42 wk, whereas
uninfected B6 and CD4 KO animals have a life span of 78–104 wk.
Thus, the lack of CD4 expression on T cells causes dramatic decreases
in the aspects of MAIDS which are responsible for mortality.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CD4+ T cells are absolutely required for the induction and progression of the BM5-induced disease (11, 12). Thus, it was possible that mice lacking CD4+ T cells would be resistant to the induction of MAIDS. However, it has been reported that the CD4 KO animals contain a peripheral non-CD4+, non-CD8+, Thy-1+, TCR+, class II-restricted Ag-specific helper T population (23) that can mediate some helper T cell functions. These DN (CD4-/CD8-) T cells might be susceptible to viral-induced immunodeficiency and might be able to support the development of MAIDS, since Ag-specific activation of T cells does not appear necessary for induction of the MAIDS (7, 19, 20). However, in the absence of the CD4 molecule, the T cells may interact inefficiently with the B cell population, which is also required for disease (13, 14). To investigate whether the typical signs of T cell anergy and concurrent lymphoproliferation and B cell responses could be observed in the absence of optimal T-B interaction, we infected CD4 KO mice and followed the development of CD4- T cell immunodeficiency, splenomegaly, B cell activation/differentiation, and progression of disease to death.
The DN splenic T cell population from BM5-infected CD4 KO animals
became progressively anergic as assessed by measuring proliferation and
IL-2 secretion in response to in vitro stimuli (Fig. 3). The pace of
this development of anergy was somewhat slower than in wild-type mice,
but by 10 wk of infection, proliferation and IL-2 production by
non-CD4+, Thy-1+ T cells
was diminished dramatically compared with responses from cells of
uninfected CD4 KO animals.
In contrast, several other defining symptoms of the MAIDS in wild-type
B6 mice were nearly absent in CD4 KO mice infected for comparable
times. Increases in the spleen (weight) index, usually a result of B
cell and CD4+ T cell hyperproliferation as well
as accumulation of granulocytes, are much reduced in CD4 KO mice.
Spleens in BM5-infected CD4-deficient mice increased only 2- to 3-fold
in weight by 10–13 wk compared with the 9- to 13-fold increase seen in
wild-type mice at comparable times (Fig. 2). Similar differences in
lymph node size were also noted (data not shown). An analysis of the
distribution of lymphocyte subsets in spleens of infected wild-type and
CD4 KO animals shows two distinctions (Fig. 2b). In
wild-type mice infected with BM5, there is a relative expansion of
CD4+ T cells and decrease in
CD8+ T cells as MAIDS develops (1, 7). The CD4 KO animals show fewer marked changes in lymphocyte
subsets (Fig. 2b). The CD4-like DN population increases only
marginally and the CD8+ population does not
decrease at all in two of three animals tested at 6 wk postinfection.
The dramatic difference in splenomegaly and lymphadenopathy between
wild-type and CD4 KO animals supports a role for CD4-MHC class II
interaction in lymphoproliferation. Since B cells are present in normal
numbers in the CD4 KO mouse, they can serve as reservoirs for the virus
and establishment of infection is within normal range (Fig. 1 and see
Ref. 23).
We also examined changes in the B cell response in wild-type B6 mice
and animals deficient in the expression of the CD4 molecule. At 6 wk
postinfection, B cells from wild-type mice incorporated significantly
more BrdU, over a 5-day labeling period, than did B cells from control
uninfected or CD4 KO mice (Table I). These data suggest that in vivo B
cell proliferation in response to BM5 infection is dependent upon
expression of CD4 on T cells and by extension probably reflects
suboptimal cognate T-B interaction (Table I).
Analysis of the ex vivo B cell response clearly revealed an increasing
state of unresponsiveness in B6 mice, such that by 14 wk postinfection,
in vitro responses to anti-IgM were almost absent. In contrast, B
cells from CD4 KO mice were largely unaffected and responded only
marginally lower than uninfected controls (Fig. 4, a and
b). Similar results were obtained when B cells were
stimulated with anti-CD38 or anti-CD40 Ab (data not shown).
Although expression of the B cell receptor as well as CD38 and
CD40 were somewhat down-regulated in BM5-infected B6 mice, it is
unlikely that this alone would account for the observed
unresponsiveness since these molecules were still detected by FACS.
These results suggest that during the late stages of disease,
interactions between anergic T cells and B cells may render the B cells
unresponsive to stimulation by a mechanism dependent on the expression
of CD4.
Hypergammaglobulinemia is a characteristic feature associated with BM5 infection. At 10 wk, B6-infected, but not CD4 KO-infected, mice possessed significantly elevated levels of polyclonal IgM and IgG Ab. The dramatic increase in IgG3 levels seen in B6 mice (and to a much lesser extent in CD4 KO mice) may be regulated by IL-10 production. This cytokine has been shown to promote switching to IgG3 in LPS-activated B cells (35) and is produced in elevated amounts by splenic cells from MAIDS-infected mice (36). These results show that the absence of CD4 on T cells has the effect of down-regulating in vivo Ab responses during MAIDS infection, and furthermore that the interaction of CD4 with class II, on the B cells, provides signals required for isotype switching and terminal differentiation. Interestingly, the elevated production of serum Ig in B6-infected mice occurs concomitantly with progression to an anergic state within the CD4+ T cells. In vivo Ab responses may occur early following infection, at a time when the T cell compartment is still functional, and then persist in serum. Alternatively, there may be some T cell-independent component to the in vivo Ab response, although it is unclear why such a response would be more pronounced in B6 vs CD4 KO mice.
Although a number of studies have examined the effects of BM5 infection
on the phenotype of CD4+ T cells
(7), few, if any, reports have analyzed the phenotype of
responding B cells. Therefore, we compared the phenotype of splenic B
cells from BM5-infected B6 and CD4 KO mice and control uninfected mice
at 10 wk postinfection (Fig. 5). A significant fraction of the B cells
from B6-infected mice possessed a phenotype characteristic of
activated/memory cells or cells undergoing terminal differentiation.
Among the most notable phenotypic changes observed in these mice were
an increase in B cell size (forward scatter), CD44, CD69, and CD80 (B7)
expression and down-regulation of CD21, CD62L, and CD23. In addition,
many B cells had down-regulated expression of surface IgM and undergone
isotype switching to IgG. Interestingly, CD43, which is not expressed
on most naive follicular B cells but is present on cells undergoing
differentiation to plasma cells (37), was elevated. These
phenotypic changes associated with BM5 infection, which became more
pronounced during late infection, correlate well with the functional
phenotype (in vivo hyperproliferation, hypergammaglobulinemia) observed
in these mice. In contrast, B cells from CD4 KO-infected mice possessed
a phenotype, which was essentially indistinguishable from either B6 or
CD4 KO control (uninfected) mice.
As might be expected, infected CD4 KO mice survive much longer than
infected B6 animals (Fig. 6). This is consistent with the concept that
one of the causes of death in wild-type animals is the huge expansion
of cells within lymphoid organs. However, even infected CD4 KO animals
did not live as long as uninfected KO animals. This shows that
extensive proliferation of lymphocytes in the spleen and lymph nodes is
not necessary for some of the lethality associated with disease;
however, the disease lethality is much delayed in its absence. The
mortality seen long after infection might result, in part, from the
immunocompromised state of the mice due to the anergy in the T cell
compartment (34).
The mechanisms involved in the induction of CD4+
T cell expansion and subsequent anergy remain unknown. It is clear that
during disease progression T cell changes are detectable by 2–4 wk and
maximal by 6–8 wk. CD4+ T cells in infected
wild-type animals are induced to become large cells with an activated
phenotype (7, 8), and substantial incorporation of BrdU
occurs in the CD4+ T cell population. There is a
relative expansion of CD4+ T cells accompanied by
the increased susceptibility of the T cells to activation-induced cell
death (8). The magnitude of division suggested in the BrdU
experiments is far less than the increase in CD4+
T cell numbers, suggesting substantial apoptosis in vivo. The function
of CD4+ T cells assessed by ex vivo, mitogen- or
Ag-stimulated proliferation and IL-2 production, decreases over this
same time course. The factors or cell interactions in the infected host
responsible for the CD4+ T cell changes are not
Ag specific or V restricted (7, 19, 20), arguing
against any conventional Ag- or superantigen-induced effects. The
findings here in which lack of CD4 has only a small impact on induction
of anergy indicates that the mechanisms involved in anergy are not
highly dependent on the CD4 molecule, which argues against a mechanism
requiring TCR-CD4 interaction with MHC class II on an APC for this
phase of disease.
There is considerable indirect evidence that a T-B interaction contributes to the progression of the disease but it is nonetheless inconclusive. The evidence includes the fact that both CD4+ T cells (11, 12) and B cells (13) are required for disease development and progression. However, the lack of disease in B cell-deficient mice could either be a consequence of missing T-B interaction or it could be due to the fact that B cells are the major target of virus infection so that insufficient levels of virus to cause disease occur in the absence of B cells. On the other hand, class II KO mice, which have B cells, but lack the majority of CD4+ T cells, which are class II restricted, also are disease resistant (14). Infection of class II KO mice reconstituted with CD4+ T cells does not lead to development of disease, supporting a requirement for T-B interaction dependent on class II. However, the levels of CD4 reconstitution achieved were small (14) and not all aspects of disease were assessed. Recently, Green et al. (16) have shown that mice treated with Ab to CD40L fail to develop MAIDS in response to BM5 infection. The anti-CD40L treatment abolished splenomegaly and hypergammaglobulinemia. While supporting a need for either a cell expressing CD40L or for CD40L-CD40 interaction at some phase of disease, these observations also do not directly imply a requirement for T-B interaction, nor do they pinpoint when in disease progression it might be necessary. Studies of allochimeras, with MHC-mismatched T and B cells, indicated that if T and B cells do interact there does not have to be MHC-compatible interaction for induction of disease (38), and this coupled with the lack of indication of any TCR involvement argue for either a non-TCR-mediated set of events or for an interaction promoted without conventional peptide/MHC recognition.
Our working hypothesis is that an initial event in a BM5-infected B6 mouse is the interaction of the CD4+ T cell population with some sort of viral product or virally induced cellular product either in soluble form or on the infected B cells. We postulate that this initial interaction may lead to early CD4+ T cell division and induction of anergy. We further suggest that the CD4+ T cells in turn produce cell surface molecules such as costimulatory molecules or secreted cytokines that mediate a polyclonal activation of B cells and perhaps also further CD4+ T cell stimulation and proliferation. We argue that CD4-class II interaction is required for effectiveness of this polyclonal activation, which leads to splenomegaly and lymphadenopathy. Several studies have indicated that although initial activation of resting T and B cells usually requires signals from specific receptors, subsequent division can be driven by bystander cytokines. Recently, several reports have suggested that pathways other than those mediated through Ag receptors might sometimes activate T cells. Combinations of cytokines and inflammatory responses (39) have been reported to drive TCR-independent T cell response, and B cell proliferation can also be induced by anti-CD40 or anti-CD38 and cytokines (15). It is likely that the factors driving CD4+ T cells in MAIDS to proliferate and become anergic and those which stimulate B cell proliferation and hypergammaglobulinemia are in this class of non-Ag-receptor-dependent inducers of lymphocyte responses.
|
Footnotes |
|---|
3 Address correspondence and reprint requests to Dr. Susan L. Swain, Trudeau Institute, P.O. Box 59, Saranac Lake, NY 12983.
4 Abbreviations used in this paper: MAIDS, murine AIDS; DN, double negative; G6PD, glucose-6-phosphate dehydrogenase; BrdU, bromodeoxyuridine; KO, knockout; CD40L, CD40 ligand.
Received for publication November 22, 2000. Accepted for publication March 9, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
5 T cell antigen receptor. J. Immunol. 149:9.[Abstract]
. J. Exp. Med. 174:547.This article has been cited by other articles:
|
|
 |
|
  W. Li and W. R. Green The Role of CD4 T Cells in the Pathogenesis of Murine AIDS. J. Virol., June 1, 2006; 80(12): 5777 - 5789. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Knoetig, T. A. Torrey, Z. Naghashfar, T. McCarty, and H. C. Morse III CD19 Signaling Pathways Play a Major Role for Murine AIDS Induction and Progression J. Immunol., November 15, 2002; 169(10): 5607 - 5614. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. A. Green, W. J. Cook, A. H. Sharpe, and W. R. Green The CD154/CD40 Interaction Required for Retrovirus-Induced Murine Immunodeficiency Syndrome Is Not Mediated by Upregulation of the CD80/CD86 Costimulatory Molecules J. Virol., November 13, 2002; 76(24): 13106 - 13110. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. H. Yen, N. Lepak, and S. L. Swain Induction of CD4 T Cell Changes in Murine AIDS Is Dependent on Costimulation and Involves a Dysregulation of Homeostasis J. Immunol., July 15, 2002; 169(2): 722 - 731. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. L. Johnson and P. C. Sayles Deficient Humoral Responses Underlie Susceptibility to Toxoplasma gondii in CD4-Deficient Mice Infect. Immun., January 1, 2002; 70(1): 185 - 191. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
