| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
24-JQ TCR Supports the Development of CD1d-Dependent NK1.1+ and NK1.1- T Cells in Transgenic Mice 1
* Ludwig Institute for Cancer Research, Lausanne Branch, University of Lausanne, Epalinges, Switzerland;
Cancer Immunotherapy and Gene Therapy Program, DIBIT, H. San Raffaele Scientific Institute, Milan, Italy;
La Jolla Institute for Allergy and Immunology, La Jolla, CA 92121; and
Swiss Institute for Experimental Cancer Research, Epalinges, Switzerland
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
24-JQ and V14-J18 rearrangements in humans and mice, respectively, and are thus defined as inv. NKT cells. Because human inv. NKT cells recognize mouse CD1d in vitro, we wondered whether a human inv. V24 TCR could be selected in vivo by mouse ligands presented by CD1d, thereby supporting the development of inv. NKT cells in mice. Therefore, we generated transgenic (Tg) mice expressing the human inv. V24-JQ TCR chain in all T cells. The expression of the human inv. V24 TCR in TCR C-/- mice indeed rescues the development of inv. NKT cells, which home preferentially to the liver and respond to the CD1d-restricted ligand -galactosylceramide (-GalCer). However, unlike inv. NKT cells from non-Tg mice, the majority of NKT cells in V24 Tg mice display a double-negative phenotype, as well as a significant increase in TCR V
7 and a corresponding decrease in TCR V8.2 use. Despite the forced expression of the human CD1d-restricted TCR in C-/- mice, staining with mCD1d--GalCer tetramers reveals that the absolute numbers of peripheral CD1d-dependent T lymphocytes increase at most by 2-fold. This increase is accounted for mainly by an increased fraction of NK1.1- T cells that bind CD1d--GalCer tetramers. These findings indicate that human inv. V24 TCR supports the development of CD1d-dependent lymphocytes in mice, and argue for a tight homeostatic control on the total number of inv. NKT cells. Thus, human inv. V24 TCR-expressing mice are a valuable model to study different aspects of the inv. NKT cell subset. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
TCR with the NK receptor NK1.1/NKRP1 (CD161) (1, 2, 3). A sizable fraction of NKT cells express a very conserved TCR, made up by the invariant (inv.)5 V14-J18 TCR rearrangement in mice or its homolog inv. V24-JQ TCR in humans, and are thus called inv. NKT cells (4, 5, 6). The inv. TCR V chains from humans and mice pair with a restricted number of junctionally diverse TCR V chains, namely V11 in humans and V8.2, V7, and V2 in mice, giving rise to a semi-inv. TCR repertoire (4, 5, 6, 7, 8). In mice, the majority of inv. NKT cells display a CD4+ phenotype (80%), whereas a minority are double negative (DN) (20%) (2, 3, 9). inv. NKT cells also have characteristic tissue distribution in mice, being selectively represented in the thymus and liver and relatively rare in secondary lymphoid organs and bone marrow (BM), which are targets for non-inv. NKT cell homing (2, 9, 10).
inv. NKT cells display a constitutively activated phenotype and are characterized by the prompt and massive production of diverse cytokines upon primary activation (2, 11). Owing to these functional characteristics, inv. NKT cells are regarded as regulatory T cells, and indeed the selective manipulation of inv. NKT cells leads to the control of autoimmune diseases and to the enhancement of pathogen- and tumor-specific immune responses (12, 13, 14, 15, 16, 17).
The semi-inv. TCR from both species recognizes CD1d (18, 19), a member of the CD1 family of Ag-presenting molecules that is so conserved between humans and mice that in vitro human and mouse inv. NKT cell clones recognize APC expressing either mouse or human CD1d, respectively (20). CD1d displays close structural similarities with MHC class I molecules; however, its Ag binding groove is very hydrophobic, compatible with its presenting functions for Ags with lipid compositions (21). Although the structure of the natural CD1d-associated ligand recognized by inv. NKT cells is not yet known, both the inv. V14 and inv. V24 TCRs display exquisite specificity for -galactosylceramide (-GalCer), a glycolipid Ag derived from marine sponges that binds to and is presented by CD1d (20, 22, 23, 24).
The structural conservation between the human and mouse semi-inv. TCR and CD1d raises the question of whether endogenous mouse CD1d-restricted ligand(s) would select the inv. V24 TCR, if expressed by mouse T cells. Thus, we generated a transgenic (Tg) mouse model expressing the human inv. V24-JQ TCR and characterized the CD1d-dependent inv. NKT cells present in these mice. We show here that Tg human inv. V24-JQ TCR replaces mouse inv. V14-J18 in supporting the development of -GalCer-reactive and CD1d-dependent NK1.1+ and NK1.1- T cells in mice.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To generate human inv. V24 Tg mice, the V24-JQ gene segment was PCR-amplified from the genomic DNA derived from the human inv. NKT cell clone CO9 (25) using the following oligonucleotides: uphV24-5'-GAAGTCAACTTGCGGCCGCAGATCTCT-3' and dw-hJQ(mC)-5'-CTGGGTTCTGGATGTCAGGCCAGACAGTCA-3', containing an overhanging nucleotide sequence complementary to mouse TCR C cDNA.
The V24-JQ gene segment was fused by PCR with a PCR fragment containing the mTCR C cDNA, amplified with the following oligonucleotides: up-mC (hJQ)-5'-TGACTGTCTGGCCTGACATCCAGAACCCAG-3', containing an overhanging nucleotide sequence complementary to the human JQ gene segment, and dw-mC-5'-CGGGATCCTCAACTGGACCACAGCCTCAG-3'.
The fused hV24-JQ-mC DNA segment, displaying the correct sequence, was cloned into pREP7 expression vector and cotransfected with a functional V8.2-mC TCR cDNA into the 58- mouse T cell hybridoma line, confirming that the V24 chimeric chain could properly pair and be expressed on the cell surface together with an endogenous mouse V-chain (data not shown).
To generate the Tg TCR mice, the hV24-JQ-mC cDNA was cloned into the human CD2 enhancer-promoter-based vector (26). The fragment containing the human CD2 transcription control regions and the inserted cDNAs was excised from the plasmid backbone by NotI-SalI double digestion and microinjected into C57BL/6 x DBA/2 fertilized eggs. Transgenic founders were screened by flow cytometry analysis of PBLs with anti-V24-specific mAb. C57BL/6 mice were obtained from Charles River Breeding Laboratories (Calco, Italy) and Harlan Breeders (Zeist, The Netherlands). TCR C-/- mice (27), backcrossed seven generations onto C57BL/6 mice, were obtained from Centre de Distribution, Typage and Archivage animal–Centre National de la Recherche Scientifique (Orleans, France). J18-/- mice (kindly provided by Prof. M. Taniguchi, Chiba University School of Medicine, Chiba, Japan) were backcrossed eight times onto C57BL/6 mice (16). All mice were kept in specific pathogen-free conditions and used at 8–12 wk of age.
Cell preparations
Single-cell suspensions were prepared from the liver, spleen, thymus, and BM. Total liver cells were resuspended in a 40% isotonic Percoll solution (Pharmacia Biotech, Uppsala, Sweden) and underlaid with an 80% isotonic Percoll solution. Centrifugation for 20 min at 1000 x g isolated the mononuclear cells at the 40–80% interface. Cells were washed twice with PBS containing 2% FCS. Spleen and BM (femur) cells were resuspended in DMEM supplemented with 5% FCS (DF) and loaded onto 10-ml nylon wool columns that had been preincubated overnight at 37°C with DF. The columns were incubated for 45 min at 37°C, and cells depleted of B cells and monocytes were harvested by washing the columns with 20 ml of DF. For some staining experiments, thymocytes were resuspended in PBS containing 2% FCS together with a 1/10 dilution of J11d or B2A2 (anti-heat-stable Ag) hybridoma culture supernatants. After an incubation of 45 min at 4°C, the cells were washed and incubated for another 45 min at 37°C with an appropriate dilution of rabbit complement. The live mature (heat-stable Ag-) thymocytes were isolated and washed twice.
Flow cytometry
Cells were preincubated with 2.4G2 culture supernatant to block FcRs, then washed and incubated with the indicated mAb conjugates for 15 min at 4°C in a total volume of 50 µl of PBS containing 2% FCS. Cells were washed and, if required, incubated with streptavidin conjugates for 10 min at 4°C. After another wash, cells were resuspended in PBS containing 2% FCS and analyzed on a FACSCalibur flow cytometer (BD Biosciences, Mountain View, CA). The following mAbs were purchased from BD PharMingen (San Diego, CA): FITC-, Cy-Chrome-, APC-conjugated anti-TCR (H57-597); PE- or biotin-conjugated anti-NK1.1 (PK136); FITC- or PE-conjugated anti-CD4 (H129.19) and anti-CD8 (53-6.7); PE-conjugated anti-CD25 (PC61); FITC-conjugated anti-V11 (RR3-15); biotin-conjugated anti-V3 (KJ25), V 4 (KT4), and V 17a (KJ23); PE-conjugated anti-CD122 (TM-1); and PE-conjugated anti-CD69 (H1.2F3). FITC-conjugated anti-CD62L (Mel-14) was from Caltag Laboratories (Burlingame, CA). FITC-conjugated anti-V2 (B20.6), V6 (44-22), V8.1/8.2/8.3 (F23.1), V8.1/8.2 (KJ16), V8.2 (F23-2), V9 (MR10-2), V10b (B21.5), and anti-CD44 (Pgp-1); biotin-conjugated anti-V5.1/5.2 (MR9-4), V5.1 (MR9-8), V7 (TR310), and V 12 (MR11-1); and anti-human V24 (C15) were purified and conjugated in the laboratories in Lausanne and Milan. PE- and APC-conjugated streptavidin was from Molecular Probes Europe (Leiden, The Netherlands). mCD1d-GalCer tetramers were produced and used as described (28). For competition experiments between CD1d-GalCer tetramers and anti-V24 mAb, purified lymphocytes from the thymus and liver were first incubated for 10 min at 20°C with 20 µg/ml purified unconjugated anti-V24 mAb (established as the concentration of unconjugated anti-V24 mAb that completely inhibits the binding of biotin-conjugated anti-V24 mAb), followed by streptavidin-APC. After washing, cells were stained with CD1d--GalCer tetramers and anti-CD4 mAb.
Proliferation and cytokine production assays
All cell cultures were performed in DMEM containing 5% FCS and 2-ME at 37°C in a 5% CO2 atmosphere. For proliferation assays, purified T lymphocytes from the spleen were depleted of CD8+ T cells by incubating with anti-CD8 mAb followed by rabbit complement. Next, 2 x 105 CD8-depleted splenic T lymphocytes were seeded in flat-bottom 96-well plates with increasing amounts of total C57BL/6 spleen cells, preincubated for 1 h at 37°C with 50 ng/ml -GalCer (Kirin Brewery, Gunma, Japan), and irradiated at 104 rad. Human recombinant IL-2 (30 U/ml) was added to the culture 24 h later, followed by [3H]thymidine on day 3. After 16 h, the cultures were harvested and counted.
To determine cytokine production by V24 Tg NKT cells, 2 x 105 purified intrahepatic lymphocytes were mixed in flat-bottom 96-well plates with increasing numbers of RMA-S cells, engineered or not with mouse CD1d, preincubated for 1 h at 37°C with 50 ng/ml -GalCer, and treated with 50 µg/ml mitomycin C for 40 min at 37°C. After 48 h, culture supernatants were collected, and their concentrations of IFN-
and IL-4 were measured by ELISA using combinations of specific mAbs (BD PharMingen).
Heteroduplex analysis
The PCR-heteroduplex analysis was performed according to published protocols (29). Briefly, total RNA was extracted from NK1.1+CD3+ and NK1.1-CD3+ sorted cells from different organs, reverse-transcribed into cDNA, and amplified by PCR using the following V8.2- and C-specific oligonucleotides: up-mV8.2-5'-AAAGGTGACATTGAGCTGTAAT-3' and dw-mC-5'-TTGGGTGGAGTCACATTTCTCA-3'. The PCR products were subjected to heteroduplex formation in the presence of 5 µg of a carrier DNA as described (29). The heteroduplex was separated on 12% acrylamide gels, electroblotted onto Hybond filters (Amersham Pharmacia Biotech, Piscataway, NJ), and hybridized with a C-specific oligonucleotide annealing only to the C region of the carrier DNA (5'-TTGATGGCTCAAACAAGGAGACC-3').
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
24 TCR chain by mouse T cells
Human inv. V24 Tg mice were produced using a CD2 cassette for T cell-specific expression (26). Five different founders were identified by flow cytometry analysis with human V24-specific mAb on PBLs (not shown). The human inv. V24 TCR chain was expressed at high frequency in the PBLs of four of five founder mice (average 72.9 ± 5.5% TCR+V24+ cells, referred to hereafter as V24high Tg), whereas founder line 4 (V24low Tg) had 4-fold fewer TCR+V24+ PBLs. NK1.1 was expressed at similar levels in all founder lines. One V24high and the V24low Tg lines were chosen for further investigation.
Normal frequency and peculiar DN phenotype of NKT cells from V24 Tg mice
First we assessed the effects of the human inv. V24 TCR expression on the development of mouse NKT and T cells. Thymi from V24high Tg mice were reduced (3- to 4-fold) in size and cellularity (on average Tg 38.5 x 106, non-Tg 126 x 106), and contained a marked expansion of DN cells with a concomitant reduction of double-positive (DP) cells (Fig. 1A). CD4+ and CD8+ subsets, in contrast, were equally represented in V24high Tg and non-Tg animals (Fig. 1A). DN, CD4+, and CD8+ subsets from V24high Tg mice contained significant proportions of V24+ cells. In particular, DN cells from V24+ Tg mice included up to 45% V24+ cells, at variance with DN thymocytes from C57BL/6 mice, which contained far fewer TCR+cells (Fig. 1A).
|
+) were not expanded in V24high Tg mice. Their absolute numbers were actually significantly decreased due to their reduced thymic size. Moreover, in marked contrast to non-Tg NKT cells, thymic NKT cells from V24high Tg mice displayed an altered phenotype, being mostly DN. Thymic NKT cells from V24low Tg mice, which harbored 
3-fold fewer TCR+V24+ thymocytes than the V24high, were both CD4+ and DN (Fig. 2B), similar to thymic NKT cells from C57BL/6 mice.
|
24high Tg mice, the organ where CD1d-dependent inv. NKT cells preferentially accumulate (2, 9, 10). In contrast to the thymus, the frequency and absolute numbers of hepatic NKT cells were comparable in non-Tg and V24high Tg mice, and even slightly increased in the latter (Fig. 2A). Again, most hepatic NKT cells in both V24high and V24low Tg mice displayed a DN phenotype, in contrast to non-Tg hepatic NKT cells, which are mainly CD4+ (Fig. 2B).
Altogether, these results showed that the human inv. V24 TCR chain is expressed efficiently although prematurely by mouse thymocytes and does not lead to the expansion of NKT cells either in the thymus or in the liver. Moreover, the premature expression of the V24 transgene appears to interfere with the developmental acquisition of the CD4 coreceptor.
Human inv. V24 TCR is sufficient for NKT cell development in the absence of endogenous TCR -chain
V24high Tg mice were bred with TCR C-/- mice to exclude the development of T cells expressing endogenous TCR chains. V24high Tg C-/- thymi were four to five times smaller than non-Tg C+ thymi (Table I). CD4/CD8 coreceptor staining showed that the DN subset accounted for up to 50% of total thymic cells (Fig. 1B), although DN thymocytes were in absolute numbers only twice as frequent as in control mice (data not shown). Among the DN thymocyte populations, the proportion of TCR+ cells within CD44-CD25- (DN4) thymocytes was markedly increased in Tg vs non-Tg mice (Fig. 1B), indicating the premature expression of the Tg TCR. Both CD4+ and CD8+ thymocytes from V24high Tg C-/- mice expressed lower TCR levels compared with C57BL6 mice (Fig. 1B). Furthermore, CD4+ thymocytes contained a substantial fraction of TCR- cells. TCR 
T cells were also present among TCR- cells, at a frequency somewhat higher than that of T cells in non-Tg thymi (5% vs 1%, respectively; data not shown).
|
chains, the frequency of thymic NKT cells in V24high Tg C-/- mice was not increased compared with non-Tg mice. The absolute numbers of NKT cells present in V24high Tg C-/- thymi were actually reduced, due to their reduced cellularity (Table I). However, in contrast to the thymus, NKT cell frequencies and absolute numbers in the liver, spleen and BM from V24high Tg Ca-/- mice were comparable to those of non-Tg mice (Table I). Unlike non-Tg mice, in which both CD4+ (in the thymus and liver) and CD8+ (in the spleen and BM, comprising NKT cells not expressing the inv. V14-J18 chain) NKT cells accumulate (Table II), the majority of NKT cells from V24high Tg C-/- mice displayed a DN phenotype (Table II). Moreover, splenic NKT cells from V24high Tg C-/- mice expressed an activated/memory phenotype (CD62L-, CD122+, and 60% CD69+ on average; data not shown) comparable to that of NKT cells from non-Tg mice. With the notable exception of the liver, there was a marked reduction of T cells (TCR+NK1.1-) in V24high Tg C-/- mice (Table I), particularly in the spleen and BM. Also, the majority of T cells developing in V24high Tg C-/- mice displayed a DN phenotype (Table II). Very similar results were obtained in V24high Tg mice crossed with J18-/- mice, which selectively lack NKT cells expressing mouse inv. V14 chains (not shown).
|
24 suffices for the development of normal numbers of peripheral NKT cells in the absence of endogenous TCR chains, while it only poorly supports the development of T cells in all organs analyzed except for the liver. Furthermore, NKT cells in V24high Tg C-/- mice maintain the DN phenotype.
Limited increase in the number of -GalCer specific, CD1d-dependent T lymphocytes in human inv. V24high Tg C-/- mice
In the absence of increased numbers of NKT cells, we measured the total number of -GalCer-specific, CD1d-dependent T cells in V24high Tg C+ mice by staining cell suspensions prepared from thymus, liver, spleen, and BM with mCD1d tetramers, loaded or not with the CD1d ligand -GalCer. Fig. 3 shows the results obtained with CD1d tetramers loaded with -GalCer (staining with empty tetramers gave background fluorescence; data not shown). As expected, in both V24high Tg C-/- and non-Tg mice most NKT cells from the thymus and liver bound mCD1d--GalCer tetramers (Fig. 3). However, unlike non-Tg mice, the majority of NKT cells from the spleen and BM of V24high Tg C-/- mice were also stained by mCD1d--GalCer tetramers (Fig. 3). V24high Tg C-/- mice also harbored in the thymus and, to a much greater extent, the liver substantially more TCR+NK1.1- T cells stained by mCD1d-aGalCer tetramers, leading to a 2-fold overall increase in both the percentage and the absolute numbers of CD1d--GalCer-specific T lymphocytes in peripheral tissues of V24high Tg C-/- mice compared with non-Tg mice (Fig. 3 and Table I). Only in the thymus of V24high Tg C-/- mice did the absolute numbers of TCR+ cells specific for mCD1d--GalCer tetramers remain low, due to the reduced total thymic cell numbers (Table I). Collectively, therefore, these data indicate that -GalCer-reactive, CD1d-dependent T lymphocytes underwent a peripheral expansion in V24high Tg C-/- mice, although to a rather limited extent.
|
24high Tg C-/- cells to respond to the CD1d-restricted ligand -GalCer in vitro. As shown in Fig. 4A–C, splenic lymphocytes from V24high Tg C-/- mice proliferated while intrahepatic lymphocytes secreted both IFN- and IL-4 in response to -GalCer in a manner comparable with that of non-Tg mice, indicating that the hybrid human/mouse inv. TCR was functional.
|
24high Tg C-/- mice display a significantly increased use of TCR V7 gene
inv. NKT cells in mice express a restricted set of TCR V chains. V8.2 is the most frequently used, followed by V7 and V2 (7). In human inv. NKT cells, V24 is paired almost exclusively with V11, the V8.2 homolog (6, 30). Hence, we set out to determine whether this restricted TCR V repertoire was maintained in human inv. V24 TCR-expressing cells. NK1.1+ and NK1.1- T cells obtained from the thymus, liver, spleen, and BM of V24high Tg C-/- mice were stained with an extended panel of anti-TCR V-specific mAbs. We found that V8.2 and V7 were indeed the two most frequently used TCR chains in NKT cells from V24high Tg C-/- mice, from a panel that included V3, 4, 5.1, 5.2, 6, 9, 10, 11, 12, and 17 (data not shown). However, in the thymus and liver, V24high Tg C-/- mice harbored markedly fewer TCR V8.2-expressing NK1.1+ cells than did non-Tg mice (Fig. 5A). In contrast, TCR V7 was used much more frequently by Tg NKT cells (Fig. 5A). This skewed V7 use was already evident in the thymus of V24high Tg C-/- mice and was maintained in all the other organs analyzed, suggesting that the biased repertoire is already selected in the thymus. Interestingly, preferential use of TCR V7 was also observed in both thymic and intrahepatic NK1.1- T cells of V24high Tg C-/- mice, (Fig. 5B), whereas TCR V8.2 use was selectively increased only in intrahepatic NK1.1- T cells of Tg animals. These findings are consistent with the increased intrahepatic population of NK1.1- T cells binding CD1d--GalCer tetramers present in V24high Tg C-/- animals.
|
24 and mouse V8.2 TCR chains might reduce the size of the V8.2+ population of Tg T cells, we determined the junctional heterogeneity of V8.2 TCR chains expressed by thymic and intrahepatic V24high Tg C-/- NK1.1+ and NK1.1- T cells. As shown in Fig. 5C, the heteroduplex analysis performed on sorted NKT and T cells from either V24high Tg C-/- or non-Tg mice shows comparable heterogeneity in the clonal composition of TCR V8.2+-expressing cells. Hence, expression of human inv. V24 TCR in mouse NKT cells does not significantly bias their TCR repertoire.
Reciprocal exclusion of endogenous V14 and Tg V24 TCR expression in NKT cells
Both the homologous human V24 and mouse V14 inv. TCRs serve to select CD1d-dependent NKT cells. Therefore, we asked what the relative contribution of each inv. TCR chain in the selection of NKT cells would be in a competitive situation in which both chains are expressed by developing lymphocytes. For this purpose, we used V24low Tg line, which permits higher expression of endogenous TCR V chains. The percentages of NKT cells in the thymus and liver from the V24low Tg line were comparable to those found in non-Tg animals and the V24high Tg line; however, their phenotype differed according to their anatomic distribution (Fig. 2B). In the thymus of V24low Tg mice NKT cells were mostly CD4+, while in the liver they were mostly DN. In the V24low Tg line, staining with anti-V24 plus anti-V8.2 mAbs on gated NKT cells confirmed that most (70%) thymic NKT cells were V24- (Fig. 6). In contrast, the majority (>80%) of NKT cells found in the liver of V24low Tg mice were V24+ (Fig. 6). To reveal the expression of endogenous V14 by NKT cells from V24low Tg mice, thymus- and liver-derived lymphocytes were stained with mCD1d--GalCer tetramers and anti-CD4 mAb, after blocking tetramer binding to V24 TCR with saturating quantities of unlabeled anti-V24 mAb. As shown in Fig. 6, the addition of anti-V24 blocking Ab did not inhibit mCD1d--GalCer tetramer binding to thymic CD4+ NKT cells, although it modestly inhibited binding to thymic CD4- NKT cells. In contrast, unlabeled anti-V24 mAb almost completely inhibited mCD1d-GalCer tetramer binding to CD4- NKT cells in the liver and significantly reduced the percentage of tetramer-positive CD4+ NKT cells. The endogenous inv. V14 TCR chain was expressed mainly by CD4+ NKT cells, which remained mostly confined in the thymus. In contrast, the Tg human inv. V24 chains were expressed mostly by DN NKT cells, which preferentially accumulated in the liver. Moreover, the competition experiment showed that NKT cells coexpressing both endogenous and Tg V chains were infrequent in the liver, suggesting the reciprocal exclusion of endogenous vs Tg inv. V-chain expression on NKT cells.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
24-JQ TCR replaces the murine inv. V14-J18 TCR homolog in C-/- mice, supporting the development of functional CD1d-dependent NK1.1+ and NK1.1- T cells. These data confirm and extend previous in vitro findings demonstrating the remarkable and unique conservation of CD1d recognition by human and mouse inv. NKT cells (20), and reinforce the concept that human and mouse inv. TCR must recognize structurally related, CD1d-restricted ligand(s) in both species.
NKT cells from V24high Tg C-/- mice displayed mostly a DN phenotype, at variance with NKT cells from non-Tg mice, which are typically 80% CD4+ and 20% DN (2, 10, 11). The predominance of the DN phenotype within Tg NKT cells is most likely dependent on the premature expression of the human Tg TCR, because the human CD2 promoter is used to drive its expression (26). In fact, we found a high frequency of Tg thymocytes expressing surface TCR already at the DN4 stage, in contrast to non-Tg mice. In support of this hypothesis, thymocytes from the V24low Tg line harbor significantly fewer V24-expressing cells, displaying a nearly normal thymic subset distribution. Therefore, these data establish a direct relation between the level of V24 expression and the alteration of thymic development. Early Tg expression of the murine inv. V14-J18 TCR chain, a Tg model similar to ours, also causes a relative expansion of DN NKT cells at the expense of CD4+ NKT cells (31), although not so pronounced as in the inv. V24 Tg model. Moreover, premature expression of Tg TCR chains, derived from MHC-restricted T cells, also impairs thymocyte expansion and blocks development at the DN stage (32, 33). Thus, human inv. V24-JQ TCR may pair with murine TCR chains already present in DN thymocytes, thereby allowing their premature selection into the NKT cell lineage and bypassing the DP and SP stages (34).
A critical question concerning the forced expression and function of the CD1d-restricted human inv. V24 TCR chain in C-/- mice is whether it would favor the development of CD1d-dependent T cells, leading to the expansion of this compartment at the expense of MHC-dependent conventional T cells. Staining with CD1d--GalCer tetramers reveals that this is the case, although to a somewhat limited extent: a barely 2-fold increase in the absolute numbers of CD1d-dependent T cells is in fact present in peripheral organs of V24high Tg C-/- mice compared with non-Tg controls. In contrast, V24high C-/- mice display a marked reduction in both the frequency and absolute numbers of NK1.1- T cells in the spleen and BM, which is likely to represent a reduction in MHC-dependent conventional T cells.
Interestingly, the increase in CD1d--GalCer-specific T cells found in V24high Tg C-/- mice is not homogeneous in terms of NK1.1 expression and anatomical distribution. CD1d--GalCer-specific T cells fill up the NK1.1+ compartment in the spleen and BM, leading to a modest 1.5- to 3-fold net increase in absolute numbers, whereas they fill up the NK1.1- compartment in the thymus and liver, increasing 10 times in absolute numbers. This peculiar distribution also explains the overall normal frequency and absolute numbers of total NKT cells found in peripheral organs of V24high Tg C-/- mice and highlights the necessity of using CD1d--GalCer tetramers to reveal the expansion of inv. CD1d-dependent T cells. From these data, it would appear that -GalCer-reactive, CD1d-dependent NK1.1+ NKT cells are placed under a tight homeostatic control on their total numbers in the thymus as well as in the periphery, because even the forced expression of the V24 TCR chain in the absence of other competitor TCR cannot increase their numbers. By contrast, CD1d-dependent NK1.1- NKT cells appear to escape this homeostatic control and expand markedly, particularly in the thymus and liver of V24high Tg C-/- mice. In this respect, it has recently been shown that inv. V14 NKT cells develop through an NK1.1- precursor, which acquires the NK1.1 receptor following an activation/differentiation step occurring in both the thymus and the liver (35, 36, 37). According to this model, NK1.1- NKT cell precursors might be allowed to accumulate in the thymus and liver of human V24high Tg C-/- mice, while more mature NK1.1+ inv. NKT cells would be much more constrained by homeostatic mechanisms. Whether this apparent differential in the homeostatic control of NK1.1- vs NK1.1+ NKT cells depends on the expression of the human inv. V24 TCR or is a general feature of the CD1d-dependent NKT cell subset is unclear. It has been reported that Tg expression of the mouse inv. V14 TCR homolog apparently leads to a substantial expansion (a 7- to 10-fold increase) of splenic NK1.1+ NKT cells, all specific for CD1d--GalCer tetramers, suggesting that in some circumstances the NK1.1+ NKT cell compartment can enlarge (31, 38). The expansion of the NKT compartment due to Tg expression of the mouse inv. V14-J18 rearrangement has also been detected following the injection of the Tg construct directly into nonobese diabetic mice (39, 40), indicating that the phenomenon is independent of the mouse’s genetic background. Therefore, it is possible that differences in the timing of expression between the Tg human inv. V24-JQ and the mouse inv. V14-J18 TCR, or in the affinity between the two TCRs for CD1d bound with endogenous ligands, may play a role in controlling the expansion of immature and mature NKT cell compartments in the two Tg models. Expansion of NK1.1+ T cells also occurs in Tg mice expressing the diverse CD1d-specific V3.2/V9 TCR rearrangements, which do not recognize -GalCer (41). However, in this case the coexpression of both and TCR chains is required to generate the CD1d-restricted specificity and the consequent NKT cell expansion documented in Tg mice, suggesting a substantial difference between the inv. and the diverse CD1-dependent TCR V chains in their ability to generate the peripheral NKT cell repertoire.
The selective affinity of inv. V24 TCR for CD1d is also reflected in its preferential pairing with mouse TCR V8.2 and 7, which is analogous to the TCR V gene pairing of the endogenous mouse inv. V14 TCR (7). Interestingly, however, the human inv. V24-JQ TCR chain pairs significantly more frequently with TCR V7 and less often with V8.2, compared with the mouse inv. V14 TCR homolog. In human NKT cells, V24-JQ TCR pairs almost exclusively with V11 (6, 30, 42), which shares comparable homology with mouse V8.2 (52%) and V7 (47%) (43). There are no evident structural differences between human inv. V24-JQ and mouse inv. V14-J18 TCR that would justify a preferential pairing of the human TCR with mouse V7 TCR compared with V8.2 (43), although future studies of swapped human/mouse TCR V chains, coexpressed with the human inv. V24 TCR, will be needed to address this issue. Therefore, we hypothesize that the altered V7/V8.2 ratios in the V24 Tg mice may reflect an increase in either positive selection of V7 or negative selection of V8.2 when paired with V24, although evidence for negative selection has been provided only for inv. NKT cells expressing CD8 (2).
Although human inv. V24 TCR displays a reduced preference for mouse V8.2 compared with mouse inv. V14 TCR, the clonal heterogeneity within V8.2+ NKT cells of Tg mice is comparable to that of V8.2+ NKT cells from non-Tg mice, confirming that there are no structural constraints in the pairing that could affect the clonal diversity of V24+V8.2+ NKT cells.
Despite the clear bias toward the generation of the -GalCer-specific, CD1d-restricted TCR repertoire, a sizable fraction of NK1.1- T cells in V24high Tg C-/- mice do not bind CD1d--GalCer tetramers. This could be due to the selection and, possibly, expansion of an inv. V24+ CD1d-restricted subset not specific for -GalCer, analogous to the Ag specificity displayed by some mouse inv. NKT cell hybridomas (44). Alternatively (and these are not mutually exclusive propositions), we cannot discount a possible interaction between human inv. V24 TCR and mouse MHC molecules, albeit at low efficiency. Consistent with these two hypotheses, inv. V24 can pair with TCR V other than the CD1d-specific V8.2, 7, and 2 (data not shown).
Given the similarity between the human and the mouse inv. V TCR chains, we sought to compare their developmental potential in V24low Tg mice, where both chains are expressed by developing inv. NKT cells. The competitive binding experiments, performed with mCD1d--GalCer tetramers in the presence of blocking anti-V24 mAb, show that in fact CD4+ NKT cells specific for mCD1d--GalCer tetramers exist in large numbers in the thymi of the V24low Tg line and ought to express the endogenous inv. V14 TCR. Nevertheless, these cells are poorly exported to the liver, where they are vastly outnumbered by DN CD1d--GalCer-binding cells that mostly express only the human inv. V24 TCR. These data indicate that in V24low Tg mice, CD1d-dependent cells expressing the endogenous inv. V14 TCR (with or without the Tg human inv. V24) stay confined in the thymus and do not emigrate efficiently into the periphery. Because NKT cells expressing a CD1d-restricted human inv. V24 TCR presumably mature in the thymus earlier than NKT cells expressing the endogenous inv. V14 TCR, they may leave the thymus prematurely to colonize preferentially the liver, where they block the export of more recently differentiated inv. V14+ NKT cells. This model suggests that the peripheral NKT cell compartment may not be replenished rapidly with cells newly generated from the thymus, thus arguing in favor of either a slow turnover of mature NKT cells in peripheral organs, or a maintenance of the peripheral NKT cell compartment independent of thymic output (45).
Altogether, the data obtained with V24high Tg C-/- mice demonstrate the conservation in vivo of the CD1d-dependent pathway of the NKT cell generation system. Therefore, these Tg mice represent a valuable model to address future questions about NKT cells, especially those concerning homeostasis and function.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: Unité Mixte de Recherche-Centre National de la Recherche Scientifique 5540, Immunologie, Université Bordeaux 2, 146, Rue Léo Saignat B.P.14, 33076, Bordeaux Cédex, France.
3 Current address: Department of Pathology, Washington University School of Medicine, St. Louis, MO 63110.
4 Address correspondence and reprint requests to Dr. Giulia Casorati, Experimental Immunology Unit, Cancer Immunotherapy and Gene Therapy Program, DIBIT, H. San Raffaele Scientific Institute, Via Olgettina 58, I-20132, Milan, Italy. E-mail address: casorati.giulia{at}hsr.it; or to Dr. H. Robson MacDonald, Ludwig Institute for Cancer Research, Ch. des Boveresses 155, CH-1066, Epalinges s/Lausanne, Switzerland. E-mail address: hughrobson.macdonald{at}isrec.unil.ch
5 Abbreviations used in this paper: inv., invariant; DN, double negative; BM, bone marrow; -GalCer, -galactosylceramide; Tg, transgenic; DP, double positive.
Received for publication August 12, 2002. Accepted for publication December 18, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
/+ cells: new clues to their origin, specificity, and function. J. Exp. Med. 182:633./ T cells demonstrates preferential use of several V genes and an invariant TCR-chain. J. Exp. Med. 178:1.-chain is used by a unique subset of major histocompatibility complex class I-specific CD4+ and CD4-8- T cells in mice and humans. J. Exp. Med. 180:1097.24-JQ/V11 T cell receptor is expressed in all individuals by clonally expanded CD4-8- T cells. J. Exp. Med. 180:1171. requirement for the development of NK1.1+ T cell receptor-/+ cells in mouse liver. J. Exp. Med. 183:1277. repertoire and small clone size. Proc. Natl. Acad. Sci. USA 98:12636.-galactosylceramide prevents autoimmune diabetes in nonobese diabetic mice. Nat. Med. 7:1052.[Medline]
-galactosylceramide treatment prevents the onset and recurrence of autoimmune type 1 diabetes. Nat. Med. 7:1057.[Medline]
-galactosylceramide enhances protective immunity induced by malaria vaccines. J. Exp. Med. 195:617.14 Nkt cells in Il-12-mediated rejection of tumors. Science 278:1623.-galactosylceramide. J. Immunol. 163:2387.24+ CD4-CD8- T cells. J. Exp. Med. 186:109.-galactosylceramide by natural killer T cells is highly conserved through mammalian evolution. J. Exp. Med. 188:1521.14 NKT cells by glycosylceramides. Science 278:1626.-galactosylceramide specifically stimulates V14+ NK T lymphocytes. J. Immunol. 161:3271./ CD4-8- subset. J. Exp. Med. 177:1763.-expressing cells. Science 256:1448.
0.05). C, Clonal heterogeneity in TCR V