| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Medicine, University of Pennsylvania, Philadelphia, PA 19104; and
Departments of Medicine and Microbiology/Immunology, University of North Carolina, Chapel Hill, NC 27599
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Earlier experiments by us and others have shown that autoimmune and cellular defects associated with gld expression are absent in gld recipients reconstituted with a mixture of bone marrow (BM) prepared from normal and gld donors (8, 9). Also, in vivo depletion of the normal T cells, but not the normal B cells, from such mixed chimeras restores the gld disease (10). These data suggest that T cells derived from the normal BM must be the FasL-expressing cells responsible for correcting the gld defect in these chimeras. This implies, then, that in the normal development of the immune system, FasL+ T cells play the same role. In the present study, we have extended our previous work by employing several mixed-BM chimeras designed to delineate the specific roles of FasL-competent CD8+ and CD4+ T cells in correcting the gld defect. We demonstrate that CD4+ and CD8+ T cell subsets originating from normal BM in mixed-BM chimeras are dissimilar in their contribution to the suppression of gld disease. Both T cell subsets mediate a suppressive effect on the development of serum autoantibodies, whereas the absence of the normal CD4 subset had much less effect on the suppression of the lymphoproliferation than did absence of the CD8 subset. We did not find evidence for a "veto cell" phenomenon in the effect normal CD8+ T cells had on the suppression of the lymphoproliferation.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
C57BL/6J (B6) and the congenic FasL-deficient B6Smn.C3H-Faslgld (B6gld), ß2-microglobulin (ß2m)-deficient C57BL/6J-B2mtm1Unc (B6ß2mKO), and C57BL/6J-Thy1aFaslgld (B6gldThy1.1) strains were originally obtained from The Jackson Laboratory (Bar Harbor, ME). B6 mice that are homozygous for the CD4 null mutation, C57BL/6-Cd4tm1 (B6CD4KO), were obtained from Dr. Dan Littman (Howard Hughes Medical Institute, San Francisco, CA). B6 mice that are homozygous for the CD8a null mutation, C57BL/6-Cd8atm1Mak (B6CD8KO) mice, were originally obtained from Dr. Tak W. Mak (University of Toronto, Toronto, Canada). All mouse strains were maintained in our breeding facility.
Chimeras
The method used for the preparation of mixed-BM chimeras has been described in detail elsewhere (11). Briefly, BM was harvested from 8- to 10-wk-old B6CD4KO, B6CD8KO, B6, and B6/gld donors and depleted of T cells with a mixture of cytotoxic mAb (anti-CD4, clone 172.4; anti-CD8, clone 31 M; and anti-Thy-1.2, clone Mmt1) in the presence of Low-Tox-M rabbit complement (Cedarlane Laboratories, Hornby, Ontario, Canada). The BM cell suspensions were mixed (1:1) and 107 cells injected via the tail vein into age- and sex-matched B6/gld recipients that had been lethally irradiated (split dose of 525 rads, 3 h apart). In the chimeras involving BM derived from ß2m-deficient mice, B6gld recipients were depleted of NK cells with i.p. injections of 200 µg of mAb PK136 (anti-NK1.1) on days -2 and -1 (12). This pretreatment prevents the rejection of BM cells originating from ß2m-deficient mice by radioresistant NK cells in ß2m-sufficient mice.
ELISAs
The serum concentrations of IgG2a antichromatin and RF (IgM anti-IgG2bb) were determined at the indicated time points after BM reconstitution by ELISA by previously described methods (11). Some results are reported in equivalent dilution factors (EDF) of standardized reference MRL/lpr sera, as previously defined by the formula: EDF = (dilution of standard reference sera which gives the equivalent OD of the test serum) x 106 (13).
Flow cytometry
Single-cell suspensions of all LN and spleens were made by passing them through cell strainers (Becton Dickinson, Franklin Lakes, NJ) in cold medium (RPMI 1640 with 15 mM HEPES, 5% FCS (HyClone, Logan, UT), 100 U/ml penicillin, and 100 mg/ml streptomycin (Life Technologies, Gaithersburg, MD)). RBC were lysed with NH4Cl, and cells were washed twice before counting by an automated cell counter (Coulter, Hialeah, FL). In 96-well microtiter plates, 1.5 x 106 cells/well were stained in cold medium containing 3% FCS and 0.1% NaN3. In general, first and second step reagents were incubated with cells on ice for 30 min. Cell surface immunofluorescence analysis of LN and spleen cells was performed by two- and three-color flow cytometry analysis with size gating on the lymphocyte population. At least 104 events were collected for each sample on a FACScan (Becton Dickinson, San Jose, CA) with Cytomation (Fort Collins, CO) or CellQuest (Becton Dickinson, San Jose, CA) data acquisition and software.
Reagents for immunofluorescence staining and flow cytometry
Reagents used for immunofluorescence staining included
anti-IgM (Bet-2-BNHS), anti-CD45R/B220-FITC (RA3-6B2;
PharMingen, San Diego, CA), anti-CD4 (172.4 overgrown supernatant;
RM4-5-PE, PharMingen), anti-CD8 (31 M overgrown supernatant;
53-6.72-PE, PharMingen), anti-Thy1.2 (MmT1 overgrown supernatant;
53-2-1-BNHS, PharMingen), 145-2C11 (hamster IgG anti-CD3; Dr. J.
Bluestone, University of Chicago, Chicago, IL), SAv-Cy-Chrome
(PharMingen), SAv-R-PE (Southern Biotechnology Associates, Birmingham,
AL), and FITC-conjugated anti-rat 
light chain (MAR-18.5)
(14).
Statistical analysis
Significant differences between experimental groups were measured using Student’s t test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our previous studies indicated that FasL-bearing T cells, but not
B cells, derived from normal BM suppressed the development of
autoimmunity and lymphoproliferation in lethally irradiated
B6gld mice reconstituted with a mixture of normal and
gld BM. In this study, we performed several mixed-BM
chimeras to elucidate which FasL-bearing T cell subset derived from the
normal BM was required for suppression. Mixed-BM chimeras were
generated by injecting into lethally irradiated B6gld
recipients a mixture of BM harvested from either B6CD4KO and
B6gld or from B6CD8KO and B6gld donors. B6CD4KO
mice are homozygous for the CD4 null mutation and thus lack
CD4+ T cells. B6CD8KO are homozygous for the
CD8
null mutation and thus lack CD8+ T cells.
With this approach, alternative cytolytic mechanisms for
lymphocyte-mediated immunoregulation, such as perforin-dependent lytic
mechanisms, are preserved in these chimeras. The effects of the
selective elimination of either the CD4+ or the
CD8+ T cell subsets from these mixed-BM chimeras
on the suppression of the gld syndrome were determined by
measuring the degree and phenotype of the lymphoproliferation and the
serum titers of autoantibodies (IgG2a anti-chromatin and IgM RF),
as described below. These data were compared with results from
B6gld recipients infused with normal BM along with
gld BM or with isologous BM.
A mixture of T cell-depleted BM from B6CD4KO mice and B6gld
mice was injected into lethally irradiated 6- to 8-wk-old
B6gld recipients in the first set of gld mixed-BM
chimeras (B6CD4KO-B6gld). Control mice were injected with
either a mixture of T cell-depleted BM from normal B6 mice and
B6gld mice (B6-B6gld) or with BM from
B6gld mice alone (B6gld-B6gld), as
outlined above. BM from B6CD8KO mice was used in a second set of
experiments with similar controls. LN lymphoproliferation of
gld mixed-BM chimeras was measured 5 months after
reconstitution. B6CD4KO-B6gld mixed-BM chimeras did not show
an increase in either organ size (data not shown) or in the number of
LN cells when compared with control B6-B6gld chimeras (Fig. 1
B). In contrast, their cell
numbers were significantly less than those seen in B6gld
recipients given isologous BM (p < 0.005). In
contrast, B6CD8KO-B6gld chimeras did exhibit an increase in
LN size (data not shown) with a corresponding 3-fold increase in LN
cell numbers when compared with identically prepared control
B6-B6gld chimeras (p < 0.05) (Fig. 2B). Despite this increase, LN
cell numbers in chimeras lacking the normal CD8+
T cell subset were still significantly less (p
< 0.05) than those measured in B6gld-B6gld
chimeras (Fig. 2B). This result was reproduced in three
separate experiments. Flow cytometric analysis of peripheral blood T
cells from allotypic B6CD8KO-B6gldThy1.1 and
B6-B6gldThy1.1 mixed-BM chimeras showed comparable T cell
chimerism in both groups. This indicated that the failure of the
B6CD8KO BM to suppress lymphoproliferation was not due to its inability
to engraft in the mixed chimeras (data not shown).
|
|
A and 3, A–C). In
similar experiments with B6CD8KO-B6gld chimeras, removal of
the normal CD8+ subset was associated with a
significant increase (p < 0.005) in the
percentage of DN T cells when compared with control B6-B6gld
chimeras (Fig. 2A) and made them indistinguishable from the
B6gld-B6gld controls (Fig. 3, D–F).
Absolute cell counts showed that the numbers of
CD4+, CD8+, and DN T cells
in B6CD4KO-B6gld chimeras were similar to those found in
B6-B6gld chimeras and significantly less than that
calculated for B6gld-B6gld mice (Fig. 1B). The low numbers of DN T cells in
B6CD4KO-B6gld chimeras contrasted with those observed in
B6CD8KO-B6gld chimeras lacking normal
CD8+ cells. B6CD8KO-B6gld chimeras had
a significant increase in DN T cells (p <
0.025, Fig. 2B) when compared with control chimeras;
however, the number of DN T cells in these chimeras was still fewer
than that determined for B6gld-B6gld chimeras.
CD4+ and CD8+ T cells and B
cells exhibited a similar pattern.
|
Autoantibody production analysis in mixed-BM chimeras
The presence of serum autoantibodies in mixed-BM chimeras was
determined 5 months after reconstitution by ELISA. Both IgG2a
anti-chromatin and IgM RF (anti-IgG2bb)
serum levels in B6CD4KO-B6gld chimeras were twice that
measured in control B6-B6gld chimeras (Fig. 4). However, production of these
autoantibodies was only restored to approximately half the amount
detected in control chimeras with gld disease
(B6gld-B6gld). B6CD8KO-B6gld chimeras
depleted of normal CD8+ T cells exhibited an
increase in IgG2a antichromatin but not in IgM RF
(anti-IgG2bb) serum autoantibodies. The
amount of IgG2a antichromatin detected was comparable to that measured
in B6gld-B6gld mice, whereas the serum level of
RF was comparable to the control B6CD8KO-B6gld chimeras
(Fig. 5).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
50%. It is possible that these lower numbers simply
reflect that B6CD8KO-B6gld mice received half as much
gld BM as B6gld-B6gld mice at the time
of reconstitution. The non-gld BM would not contribute much
to the DN subset because it could not provide
CD8+ cells that are the presumed precursors
(14, 17). Another possibility is that nonaberrant
CD8-/CD4- T cells,
normally seen in non-autoimmune mice, originating from the B6CD8KO BM
are capable of some degree of suppression. We know that gld
disease can be completely restored in mixed-BM chimeras after treatment
with an allotype-specific Ab directed against Thy-1 (10).
However, in the mixed-BM chimeras lacking normal
CD4+ T cells or CD8+ T
cells, the non-aberrant
CD8-/CD4- T cell
population would still be intact and thus be able to contribute to the
suppression of either autoantibody production or lymphadenopathy. In
support of this notion, certain functions normally attributed to a
particular T cell subset can be demonstrated for DN T cells, albeit to
a lesser degree. For example, reduced Th cell activity in mice
homozygous for the CD4 null mutation was shown to be mediated by DN T
cells in Fas-competent mice (18). We selected young gld mice as recipients to preclude the emergence of recipient FasL-competent BM-derived cells as a complicating variable in our analysis. In doing this, we cannot completely discount the effect that newly emerging FasL-competent BM elements may have on recipient radioresistant Fas-bearing elements in the BM or thymic microenvironment. This is a potentially confounding issue as there is evidence that FasL-deficient cells may overexpress Fas and be more susceptible to deletion (19). We believe that these concern are unwarranted as gld and normal mice recipients of gld BM are indistinguishable (9, 20).
The evidence that DN T cells derive from CD8+ cells is substantial. For example, a similar pattern of TCR Vß expression was observed for both CD8+ cells and DN T cells isolated from MRL/lpr mice (17). Also, in vivo anti-CD8 treatment causes a marked reduction in DN T cell numbers (21). More recently, B6ß2m-lpr mice, which lack normal MHC class I expression and are subsequently deficient in CD8 lineage cells, failed to generate large numbers of DN T cells (14). Thus, based on our current results, the same T cell subset that gives rise to the abnormal phenotype in Fas- or FasL-deficient mice also is primarily responsible for preventing its own expansion in normal mice through a Fas/FasL-dependent process. A viable explanation for such an observation would be the veto cell phenomenon. As seen in the results of the B6ß2mKO-B6gld mixed-BM chimeras; however, this is unlikely, because ß2m-deficient BM-derived elements were still capable of suppressing the lymphoproliferation and aberrant DN T cell formation. This is consistent with recent work which indicates that the veto cell phenomenon, while apoptotic, is not Fas dependent (22, 23).
The differential effect observed on autoantibody suppression is of note. The increase in serum autoantibodies that we saw in the mixed-BM chimeras lacking normal CD4+ T cells strongly supports the idea that normal CD4+ T cells help to remove autoreactive B cell clones by a cognate mechanism. Studies have shown that killing of Fas+ targets can be mediated by either a soluble form of FasL (noncognate) or by cell contact (cognate) with cells expressing FasL on their cell surface (4, 24). If the Fas-FasL interaction involving normal CD4+ T cells was noncognate in nature, it would seem likely that the absence of normal CD4+ T cells in the mixed-BM chimeras would have a nonspecific effect on the restoration of gld disease; that is, an increase in both DN T cells and autoantibodies, as opposed to a change in one or the other. However, this was not the case. The increase was only seen for autoantibody production and is therefore suggestive of a cognate interaction. Because CD4+ cells are restricted to Ag recognition in the context of class II MHC, only autoreactive B cells would be the target of this interaction as mouse T cells are devoid of MHC class II Ag expression. The demonstration that anergic HEL-specific B cells from soluble HEL/anti-HEL Ig double transgenic mice, but not naive B cells or anergic B cells from lpr homozygotes, are deleted by CD4+ HEL-specific T cells in a class II-restricted manner after adoptive transfer into irradiated soluble HEL transgenic recipients is consistent with our data (25). Similarly, in the mixed-BM chimeras depleted of normal CD8+ cells, the increase in serum levels of antichromatin, but not IgM RF, autoantibodies suggests that for at least some autoantibody specificities, autoreactive cells necessary for autoantibody formation are normally eliminated by CD8+ T cells. However, unlike the normal CD4+ T cell subset where the cell-to-cell interaction would be restricted to autoreactive B cells, it is more likely that cells on which the normal CD8+ T cell subset would exert its effects would be the autoreactive Th cells, perhaps by recognizing self-peptides in the context of class I Ags. Our observation that antichromatin levels were more effectively restored in the B6CD8KO mixed-BM chimeras argues that autoreactive Th cells, in addition to autoreactive B cells, are regulated by FasL-sufficient CD8+ T cells. The suppression of the abnormal DN phenotype by normal CD8+ subset may be by recognition of the Fas+ CD8 precursors directly by the FasL+ CD8+ regulatory cells, or perhaps by simultaneous recognition of a third cell by the killer and target cells.
In summary, our data show that while both normal CD4+ T cells and CD8+ T cells participate in the suppression of gld disease in our mixed-BM chimera model, they do so with different outcomes. The suppressive effect mediated by the normal CD4+ T cell subset is mainly directed at the development of autoantibodies, whereas the normal CD8+ T cell subset exhibits a broader effect by diminishing autoantibody production as well as lymphadenopathy. The variable effects of the two T cell subsets on gld disease development presumably relate to different MHC restriction during self Ag presentation. Exactly what recognition occurs by these regulatory CD4+ and CD8+ T cells during the normal in vivo control of the development of the immune system remains to be determined.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 This work was supported in part by National Institutes of Health Grants AR26574, AR40620, AR34156, and T32AR07416. P.L.C. was supported by Grant AR33887. At the time of their participation G.C.M. was a Postdoctoral Fellow of the Canadian Arthritis Society, and V.N.K. was a Postdoctoral Fellow of the Arthritis Foundation.
3 M.A.M. and G.C.M. contributed equally to this study.
4 Address correspondence and reprint requests to Robert A. Eisenberg, Division of Rheumatology, Department of Medicine, University of Pennsylvania, 504 Maloney, 3600 Spruce Street, Philadelphia, PA 19104-4283. E-mail address:
5 Abbreviations used in this paper: DN, double negative; LN, lymph nodes; BM, bone marrow; EDF, equivalent dilution factors; FasL, Fas ligand; ß2m, ß2-microglobulin.
Received for publication April 14, 1999. Accepted for publication July 7, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
3 domain of class I MHC. Science 252:1424.2 domain signals rapid Fas-independent cell death: a direct pathway for T cell-mediated killing of target cells?. J. Immunol. 160:4343.This article has been cited by other articles:
|
|
 |
|
  G. De Panfilis, A. Caruso, P. Sansoni, G. Pasolini, D. Semenza, and C. Torresani Identification of Fas-L-Expressing Apoptotic T Lymphocytes in Normal Human Peripheral Blood : In Vivo Suicide Am. J. Pathol., February 1, 2001; 158(2): 387 - 391. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
