|
|
||||||||
CUTTING EDGE |


* CBR Institute for Biomedical Research and Department of Pediatrics, Harvard Medical School, Boston, MA 02115;
Juntendo University School of Medicine, Tokyo, Japan; and
Academic Medical Center, Amsterdam, The Netherlands
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
NKRs are not expressed by naive CD8 T cells (2, 5, 6). However following a viral or bacterial infection, a large proportion of activated CD8 T cells in the lymphoid organs express a variety of NKRs, including gp49B1 and CD94/NKG2A, that can regulate their functions (5, 6). Although the gut mucosal CD8 T cells differ in many respects from the peripheral CD8 T cells, including function and longevity, whether they also differ in NKR expression is not known (9, 10, 11, 12). Moreover, recent studies have highlighted tissue-specific differences in the development of mucosal immunity and suggest that the local milieu may play a role in the maintenance and function of lymphocytes (12, 13, 14). In this context, we have recently reported the presence of a novel gut-specific APC of nonhemapoietic origin that constitutively expresses the costimulatory molecule CD70 (13). In addition, a recent report showed that gut-specific dendritic cells (DCs) produce retinoic acid (RA) that is essential for imprinting mucosal T cells with gut-homing specificity (14). In the present study, we show that NKRs are differentially imprinted on Ag-specific CD8 T cells in the gut compared with the periphery and examine the role of tissue-specific factors that could contribute to these differences.
| Materials and Methods |
|---|
|
|
|---|
C57BL/6 mice were purchased from The Jackson Laboratory. B6;D2-Tg(TcrLCMV)327Sdz/JDvsJ (commonly named P14) mice were purchased from Taconic Farms. All mice were maintained under specific pathogen-free conditions and used at 610 wk of age.
Flow cytometric analysis
Abs to mouse NKR, CD8, and
4
7 were from BD Pharmingen. MHC Db/gp33 tetramers were obtained from Beckman Coulter. The CCR9 Ab was from R&D Systems. Gp49B1 expression was measured using the hamster IgG mAb H1.1 (15).
Viral, bacterial, and allogeneic tumor challenge
Mice were injected i.p. with 106 PFU vaccinia virus (WR strain), 104 CFU of Listeria monocytogenes (Lm) or 5 x 106 P815 or CD70-transfected P815 cells (16) and, at day 8 postinfection, their peritoneal exudate lymphocytes (PELs) and other organs were harvested.
Adoptive transfer
Naive CD8+ T cells were purified from the splenocytes of P14 mice by negative selection using the murine CD8 subset isolation kit (R&D Systems). Recipient mice were injected i.v. with 8 x 106 purified CD8 lymphocytes. At day 3 posttransfer, recipient mice were i.p. infected with 1 x 104 CFU of recombinant Lm encoding the lymphocytic choriomeningitis virus glycoprotein (rLmgp33; a gift of Dr. H. Shen, University of Pennsylvania School of Medicine, Philadelphia, PA) or with the gp3341 peptide (KAVYNFATC, 100 µg/mouse, BioSource International) in Freunds incomplete adjuvant and, at the indicated times postchallenge, their PELs and other organs were harvested. Isolation of lymphocytes from the intraepithelial lymphocyte (IEL) and lamina propria (LP) compartments was done as described in Ref. 13 .
Ab and RA treatment
Mice were given murine CD70 Ab (clone 3B9) (17) or hamster Ig (100 µg per i.p. injection) twice weekly. Other mice were given trans-RA (50 µl of 2 mg/ml in 16% DMSO per mouse; Sigma-Aldrich) or diluent (50 µl of 16% DMSO) twice weekly.
| Results and Discussion |
|---|
|
|
|---|
To initiate an immune response, subpopulations of CD8+ T cells may be differentially activated depending on the local environment (12, 13, 14). Because NKRs regulate T cell effector functions (2, 3, 4, 5, 6), we hypothesized that their expression may be also differentially tailored depending on the factors within the local milieu. We therefore compared the expression of various NKRs between activated CD8+ T cells in the gut and periphery. Naive T cells in the lymphoid organs in normal mice do not exhibit an activated phenotype neither do they express NKRs (2, 5, 6). However, the gut mucosal T cells from naive mice exhibit an activated phenotype (12, 13, 14). Thus we initially compared them with activated T cells that accumulate in the peritoneal cavity (PEL) after vaccinia virus infection (by the i.p. route) for NKR expression. In agreement with previous reports (2, 5, 6), a majority of CD8 T cells in the PEL expressed gp49B1 and NKG2A. However, they did not express 2B4 (Fig. 1, a and b). This NKR profile appears to be similar to activated CD8 T cells in the spleen (data not shown). In contrast, most of the gut mucosal CD8 T cells expressed gp49B1 and 2B4, but little or no NKG2A (Fig. 1a). In contrast, the Ly49 isoforms NK1.1 and DX5 were not expressed by CD8 T cells whether derived from the peritoneal cavity or gut mucosa (Fig. 1b). These results suggest that the "activated" resident CD8 T cells present in the gut mucosa of naive mice differ in the profile of NKR expression compared with the activated cells at the peripheral sites.
|
In the experiments described earlier, NKR expression was tested on total CD8 T cells and, thus, it could not be conclusively demonstrated that Ag-specific CD8 T cells diverge in the NKR profile at different tissue sites. We addressed this question using an adoptive transfer system. Purified naive CD8+ T cells from P14 transgenic mice that were uniformly negative for 2B4, NKG2A, and gp49B1 (Fig. 1c; column labeled "Input") were adoptively transferred to C57 recipients. Three days after transfer, the mice were injected with gp33 peptide in IFA and the expression of NKR on P14 tetramer+ cells was monitored after another 3 days in different organs using gp33 tetramers. Even under these conditions, whereas the majority of tetramer+ PELs uniformly expressed gp49B1 and NKG2A but not 2B4, the tetramer+ IELs highly expressed 2B4 and gp49B1 but exhibited little or no NKG2A (Fig. 1, c and d). Thus, NKG2A and 2B4 are differentially imprinted on CD8 T cells at the peripheral and intestinal tissue sites during the same antigenic challenge. One possibility for the physiological relevance of this observation is that the divergent NKR imprinting may contribute to some of the differences in the function and longevity of mucosal T cells as compared with peripheral T cells (9, 10, 11, 12). However, in this particular case the significance of replacing NKG2A with 2B4 on mucosal CD8 T cells is not immediately apparent. Although 2B4 was originally proposed as a stimulatory NKR, recent studies using knockout animals suggest that it essentially functions as an inhibitory molecule in NK cells (4, 8). Whether it suppresses or stimulates T cells is not yet clear. Further studies of mucosal immunity in the 2B4 knockout mice should clarify whether 2B4 expressed by mucosal T cells account for an increased function of mucosal T cells.
Blockade of CD70 costimulation attenuates 2B4 expression in mucosal CD8 
+ T cells
We considered the possibility that the local environment may play a role in shaping the NKR repertoire in T cells by favoring the expression of specific NKRs and suppressing others. A noteworthy feature of T cell activation is the multiplicity of costimulatory mechanisms, and it is conceivable that they might also play a role in regulating the NKR expression. In fact, costimulatory molecules appear to hold the key to determining the nature and magnitude of the T cell activation process (13, 18, 19, 20). In line with this view, we have recently identified a novel gut-specific APC of nonhemopoietic origin that constitutively expresses the CD70 costimulation molecule and showed that a blockade of the CD70 pathway attenuates the activation and expansion of mucosal CD8+ T cells (13). Thus, in an attempt to identify the mechanism leading to the differential expression of 2B4 and NKG2A, we tested the ability of CD70 costimulation to regulate this process. In the initial experiments, C57 mice were orally challenged with Lm and simultaneously i.p. injected with either control Ig or blocking CD70 Abs. As Fig. 2, ac show, the expression of 2B4 was drastically decreased in CD8
+ but not in CD8
+ IELs in CD70 Ab-treated animals. Because very small numbers of activated CD8 T cells could be found at the systemic site of PELs using this route, we could not conclusively prove that anti-CD70 Ab had no effect on NKR expression on the PEL CD8 T cells. To extend the experiment reported above and confirm whether the effect of anti-CD70 was specific to the expression of 2B4 or also to other NKRs, we repeated the CD70 Ab treatment in adoptive transfer experiments (Fig. 2d). Even in this setup, the CD70 Ab treatment did not change the NKR expression pattern of tetramer+ PELs (not shown). However, although CD70 Ab treatment significantly reduced the total number of tetramer+ cells in the IEL compartment (Fig. 2d) as reported earlier (13), it also substantially reduced 2B4 expression on the tetramer+ cells that were still present (Fig. 2d). In this context, one possibility for the effect of CD70 Ab on 2B4 expression is that this Ab may have interfered with CD40-dependent activation (19, 20). It was shown that in the absence of CD4 help, the CD70 expression on APCs plays a key role in CD40-dependent CD8 T cell activation (20). It is therefore possible that the CD4-helped CD8 cells are the ones that still migrate, accumulate, and express 2B4 in the anti-CD70-treated animals, whereas the CD70/CD40-dependent CD8 cells are not activated properly and, as a secondary effect, also fail to induce 2B4 in the gut environment.
|
To directly investigate whether CD70 costimulation is involved in 2B4 induction, we tested whether 2B4 expression by peripheral CD8 T cells can be enhanced by the provision of CD70 during stimulation. Mice were challenged with tumor P815 cells or CD70-transfected P815 cells and, 8 days later, PELs were harvested and examined for the expression of NKRs. As expected, most CD8+ PELs were positive for NKG2A and gp49B1 receptors (not shown). More importantly, deliberate CD70 costimulation enhanced dramatically the number of 2B4+ CD8+ PELs (
6-fold greater) (Fig. 2e). Thus, it appears that CD70 costimulation is necessary for the induction of 2B4 by CD8 T cells. Because gut-specific CD70+ APCs stimulate the mucosal T cells in situ (13), we suggest that tissue-specific APCs may be responsible for the differential induction of 2B4 by T cells. However, considering that mature DCs at systemic sites also express CD70 and can bypass CD4 T cell help (19, 20), the 2B4 induction may not be entirely mediated by gut-specific CD70+ APCs.
RA down-regulates NKG2A expression
Although our previous results could account for 2B4 induction by mucosal CD8 T cells, they failed to reveal how NKG2A is shut off in the mucosa. We considered the possibility that NKG2A may be actively suppressed in the gut microenvironment. It has been recently shown that RA produced by DCs in the intestinal lymphoid organs can alter specific gene expression in T cells to imprint gut-homing specificity (14). We therefore examined whether external provision of RA can alter NKG2A expression by peripherally activated CD8 T cells. In the initial experiments mice were infected i.p. with Lm, and on day 5 postinfection cells isolated from the peritoneal cavity (expressing NKG2A uniformly) were in vitro cultured with 20 ng/ml IL-2 in the presence or the absence of 100 nM trans-RA for 2 days. This concentration of RA was shown to be optimal for the expression of the gut homing molecules on activated T cells (14). As Fig. 3 a shows, a significant reduction in the number of NKG2A+ CD8+ cells was observed in the RA-treated samples compared with the controls, suggesting that NKG2A down-regulation can be achieved with RA treatment. To test whether RA treatment in vivo also suppresses NKG2A, we i.p. infected mice with Lm and simultaneously i.p injected them with RA or the diluent (DMSO). Although the CD8 T cell numbers and the profile of 2B4 expression was unaltered in both compartments, the number of NKG2A+ as well as CD94+ cells significantly decreased in the PEL of RA-treated animals (Fig. 3b and data not shown), showing that NKG2A/CD94 expression can be selectively down-regulated on CD8 T cells by RA treatment. We also confirmed this observation on activated splenocytes (data not shown) and Ag-specific P14 CD8 T cells (Fig. 3c). In these experiments and consistent with the previous report (14), we found that RA treatment up-regulated the expression of CCR9 and
4
7 on Ag-specific CD8+ T cells in the PEL (
5- to 14-fold greater percentage of positivity and
2 to 4-fold greater mean fluorescence intensity). More importantly, we found that RA treatment significantly attenuated the expression of NKG2A on Ag-specific CD8+ T cells in the PEL (Fig. 3c), suggesting that RA actively suppresses TCR-induced up-regulation of NKG2A expression. Moreover, we have investigated in vitro activation of P14 CD8 T cells with DCs from the spleen and mesenteric lymph node (MLN) in the presence or absence of 10 µM citral, a blocker of RA synthesis, (as described in Ref. 14). Activation with gp33-pulsed splenic DCs induced a robust induction of NKG2A expression on P14 CD8 T cells (
60%), and treatment with citral had no significant effect on the level of this expression. In contrast, activation with MLN-DC induced only a weak expression of NKG2A on activated CD8 (
18%), and treatment with citral enhanced this expression to a level that is similar to the one observed with splenic DCs.
|
| Acknowledgments |
|---|
| Disclosures |
|---|
|
|
|---|
| Footnotes |
|---|
1 This work was supported by National Institutes of Health Grant AI46566 and Senior Research Award from the Crohns and Colitis Foundation of America (to N.M.). ![]()
2 Address correspondence and reprint requests to Dr. Amale Laouar, CBR Institute for Biomedical Research, 800 Huntington Avenue, Boston. MA 02115. E-mail address: laouar{at}cbr.med.harvard.edu ![]()
3 Abbreviations used in this paper. NKR, natural killer receptor; DC, dendritic cell; IEL, intraepithelial lymphocyte; Lm, Listeria monocytogenes; LP, lamina propria; MLN, mesenteric lymph node; PEL, peritoneal exudate lymphocyte; RA, retinoic acid. ![]()
Received for publication December 8, 2005. Accepted for publication October 31, 2006.
| References |
|---|
|
|
|---|
responses of T cells and NK cells. J. Immunol. 170: 4095-4101. This article has been cited by other articles:
![]() |
N. S. Joshi and S. M. Kaech Effector CD8 T Cell Development: A Balancing Act between Memory Cell Potential and Terminal Differentiation J. Immunol., February 1, 2008; 180(3): 1309 - 1315. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |