HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chen, Y.-T. Articles by Kung, J. T. Search for Related Content PubMed PubMed Citation Articles by Chen, Y.-T. Articles by Kung, J. T.

CD1d-Independent Developmental Acquisition of Prompt IL-4 Gene Inducibility in Thymus CD161(NK1)^–CD44^lowCD4⁺CD8^– T Cells Is Associated with Complementarity Determining Region 3-Diverse and Biased V2/V7/V8/V3.2 T Cell Receptor Usage¹

Yi-Ting Chen^* and John T. Kung^2,

^* Graduate Institute of Immunology, College of Medicine, National Taiwan University, Taipei, Taiwan; and Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Among Ag-inexperienced naive T cells, the CD1d-restricted NKT cell that uses invariant TCR--chain is the most widely studied cell capable of prompt IL-4 inducibility. We show in this study that thymus CD161^–CD44^lowCD4⁺CD8^– T cells promptly produce IL-4 upon TCR stimulation, a response that displays biased V(2/7/8) and V3.2 TCR usage. The association of V family bias and IL-4 inducibility in thymus CD161^–CD44^lowCD4⁺CD8^– T cells is found for B6, B10, BALB/c, CBA, B10.A(4R), and ICR mouse strains. Despite reduced IL-4 inducibility, there is a similarly biased V(2/7/8) TCR usage by IL-4 inducibility⁺ spleen CD161^–CD44^lowCD4⁺CD8^– T cells. Removal of -galacotosylceramide/CD1d-binding cells from CD161^–CD44^lowCD4⁺CD8^– thymocytes does not significantly affect their IL-4 inducibility. The development of thymus CD161^–CD44^lowCD4⁺CD8^– T cells endowed with IL-4 inducibility and their associated use of V(2/7/8) are ₂-microglobulin-, CD1d-, and p59^fyn-independent. Thymus CD161^–CD44^lowCD4⁺CD8^– T cells produce low and no IFN- inducibility in response to TCR stimulation and to IL-12 + IL-18, respectively, and they express diverse complementarity determining region 3 sequences for both TCR-- and --chains. Taken together, these results demonstrate the existence of a NKT cell distinct, TCR-repertoire diverse naive CD4⁺ T cell subset capable of prompt IL-4 inducibility. This subset has the potential to participate in immune response to a relatively large number of Ags. The more prevalent nature of this unique T cell subset in the thymus than the periphery implies roles it might play in intrathymic T cell development and may provide a framework upon which mechanisms of developmentally regulated IL-4 gene inducibility can be studied.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Interleukin-4 promotes Ag-dependent differentiation of naive T cells into Th2 and Tc2 effector cells that produce large amounts of IL-4 as well as IL-5, IL-6, IL-9, IL-10, and IL-13 (1, 2, 3). The requirement for IL-4 to make IL-4 is an example of the chicken-and-egg dilemma, and the initial source of IL-4 required for Th2 development has been the subject of intensive investigation. Cells capable of prompt IL-4 production in an IL-4-independent fashion are good candidates for the source of IL-4 required for Th2 development. Because the CD1d-restricted NKT cell responds to TCR stimulation by prompt IL-4 production, it was originally thought to play a major role in the initiation of IgE response (4). However, the ability to mount IgE responses in mice deficient in NKT cells clearly indicate that there are NKT cell-independent pathways (4, 5, 6). We had previously reported a small subset of CD161^– CD44^lowCD4⁺CD8^– thymocytes that is capable of TCR-stimulated prompt and stat6-independent IL-4 inducibility (7). Because its CD161^– CD44^low expression is distinct from the CD161⁺CD44^high NKT cells, we suggested that it may represent another lineage of T cells that may provide IL-4 during early immune response. Since then, a more detailed intrathymic development pathway of NKT cells has been elucidated and whereas the majority of mature NKT cells are indeed CD161⁺CD44^high, most of them develop from a minor population of CD161^–CD44^low precursors (8, 9), thus raising the possibility that the IL-4 inducibility response we have observed for CD161^– CD44^lowCD4⁺CD8^– thymocytes may be coming from NKT precursor cells rather than from another T cell lineage as we had suggested before. Because of the potential significance of finding a NKT-independent T cell lineage that is capable of TCR-stimulated rapid IL-4 inducibility, we performed additional experiments aimed at addressing the relatedness between the CD161^–CD44^lowCD4⁺CD8^– T cells that we have reported before (7) and NKT cells. In this study, we show that whereas the CD161^–CD44^lowCD4⁺CD8^– thymocytes we are working with express similar V(2/7/8) bias as do NKT cells, they are distinct from NKT cells in that their development is CD1d/₂-microglobulin (₂m)³/p59^fyn-independent and that they express highly diverse complementarity determining region 3 (CDR3) sequences in both TCR-- and TCR--chain.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Mice

Breeders of the MHC class II (MHC-II)-restricted AND TCR-transgenic (Tg) mice (10) were provided by Dr. S. Hedrick (University of California San Diego, San Diego, CA). C57BL/10ScN (B10), B10.A, and B10.A(4R) breeders were originally obtained from the Division of Research Service, National Cancer Institute (NCI), National Institutes of Health (Frederick, MD). Breeders for BALB/cJ, C57BL/6J, CBA/CaJ, p59^fyn-null on C57BL/6 background (11), and CD1d-null on BALB/c background (5) mice were obtained from The Jackson Laboratory. ICR mice were obtained from the National Laboratory of Animal Breeding and Research Center (Taipei, Taiwan). Stat6-null mice on a mixed B6/129 background (12) and IL-4R-null mice on BALB/c background (13) were kindly provided by C. Watson and Dr. W. E. Paul (Laboratory of Immunology, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD). Breeders for ₂m-null (14) and MHC-II-null (15) mice on C57BL/6 background were kindly provided by Dr. B. J. Fowlkes (National Institutes of Health, Bethesda, MD). B10.TL mice are B10 congenic mice that express Thy-1^a and CD8^a alleles (16). MHC-II-restricted DO11.10 TCR Tg mice on BALB/c background (17) were obtained from the Laboratory Animal Center (College of Medicine, National Taiwan University, Taipei, Taiwan). Mice used were 30–45 days of age. Unless otherwise indicated, all mice used were bred and housed under specific pathogen-free conditions at the Institute of Molecular Biology Animal Facility (Academia Sinica, Taipei, Taiwan).

T cell isolation and activation

Thymocytes were twice depleted of CD8⁺ cells by adherence to culture plates coated with anti-CD8 mAb by a modified procedure used to deplete spleen CD8⁺ T cells (7, 16). Spleen CD4⁺ T cells from B10.TL mice were isolated as described previously (16). Two-cycle CD8-depleted B10 thymocytes were stained by using indicated combinations of appropriately conjugated reagents from the following: FITC-anti-V2 (18), FITC- or Cy5-anti-V3 (19), FITC-anti-V5 (20), FITC-anti-V6 (21), FITC-anti-V7 (22), FITC- or Cy5-anti-V8 (23), FITC-anti-V11 (24), FITC-anti-V2 (25), FITC- or AF647-anti-V3.2 (26), FITC-anti-V8 (18), FITC-anti-V11 (27), FITC-, biotin-, PE-, or AF680-anti-CD161 (28), Cy5- or AF680-anti-CD44 (29), Texas Red (TR)-anti-CD4 (30), PE-anti-pan-NK (clone DX5; BD Biosciences), A405- or biotin-anti-CD8 (31), streptavidin-R-PE (Phycoprobe R-PE Streptavidin; Biomedia). DimerX (CD1d:Ig; BD Biosciences) loaded with -galacotosylceramide (-GalCer; Pharmaceutical Research Laboratories, Kirin Brewery) was detected by, PE-anti-mouse IgG1 (clone A85-1; BD Biosciences). All staining reactions were performed with the addition of 2.4G2 anti-FcR mAb (32) to block FcR-mediated binding of fluorochrome-labeled mAbs. The stained cells were subjected to electronic cell sorting (FACStarPlus equipped with dual lasers with 488 nm and 595 nm excitations or Digital FACSVantage SE fluorescence-activated cell sorter equipped with three lasers with 405 nm, 488 nm, and 595 nm excitation; BD Biosciences). The panel of markers that were used to sort responding T cells depended on the nature of responding T cells and on whether the mouse strains used were CD161⁺. Responding T cells isolated by electronic cell sorting were one of CD161^–CD44^lowCD4⁺CD8^–, DX5(Pan-NK)^–CD44^lowCD4⁺CD8^–, CD44^lowCD4⁺CD8^–,or CD161⁺CD44^highCD4⁺CD8^– phenotypes. The exact combination of CD161/CD44/DX5 and TCR- family/CD4/CD8 marker expression used as criteria to sort out responding T cells is as given in figure legends and table footnotes. The purity of sorted T cells was always re-analyzed and was always >98%.

All of the T cell activation cultures were set up in Mishell-Dutton medium (33) containing 5% FCS (HyClone), 50 mM HEPES, and 5 x 10^–5 M 2-ME in standard 96-well microculture plates (0.1 ml final volume/well). Mitogenic activation cultures were set up with T cells (2 x 10⁴ cells/well), 100 ng/ml each of anti-CD3 mAb (34) and anti-CD28 mAb (35), together with mitomycin (10 µg/ml)-treated B cell blasts (10⁵ cells/well) that had been generated by culturing T cell-depleted C57BL/6 spleen cells (36) in the presence of 2 µg/ml LPS (List Biological Laboratoies) for 2 days. Where indicated, T cells (7–10 x 10⁴ cells/well) were stimulated in culture wells that had been previously coated with anti-CD3 (10 µg/ml) + anti-CD28 (10 µg/ml) mAbs as described previously (7). To activate AND TCR Tg CD4⁺ T cells (2 x 10⁴ cells/well) in an Ag-specific manner, pigeon cytochrome c (PCC) 88-104 peptide (KAERADLIAYLKQATAK at 10 µM; kindly provided by Dr. S. Hedrick, University of California San Diego, San Diego, CA) was added and presented by B cell blasts (10⁵ cells/well) from I-E^k-expressing B10.A mice. To study -GalCer reactivity, it was added at a final concentration of 2 µg/ml to cultures of responding T cells (2 x 10⁴ cells/well) containing CD1d^high spleen cells (750R irradiated, 10⁵ cells/well). CD1d^high spleen cells were sorted out from Thy-1-, CD4-, and CD8-depleted spleen cells (36) that had been stained with FITC-anti-CD1 mAb (clone 1B1; BD Biosciences). To study cytokine-mediated IFN- production response, recombinant mouse IL-12 (PeproTech) and IL-18 (BioSource International) were both added at 10 ng/ml to cultures of indicated responding T cells.

Purification and fluorochrome conjugation of mAbs

Purified mAbs were obtained from hybridoma culture supernatant by affinity chromatography. Sepharose 4B beads conjugated with MAR18.5 mouse anti-rat Ig--chain mAb (37) were used to purify 3.155, 53-6.7, GK1.5, RL172, IM7, B20.6, TR310, RR3-15, RR3-16, RR8-1, and 30H12 mAbs; protein A-Sepharose columns were used to purify PK136, 500A.A2, KJ25, MR9-4, and F23.1 mAbs as described previously (38). RR4-7 was purified by protein G-Sepharose affinity chromatography. Anti-CD28 mAb was similarly purified except that a polyvalent goat anti-hamster Ig-conjugated Sepharose 4B column was used. FITC and biotin conjugation was performed as described previously (39). Cy5 (Amersham Biosciences), TR and Alexa Fluor 405 (Molecular Probes) conjugations were made according to the manufacturer’s instructions. All fluorochrome-conjugated mAbs were subjected to specificity testing, and all mAbs used in this study were highly specific because >98% of the observed fluorescence was inhibited by relevant unlabeled mAbs but was unaffected by irrelevant control mAbs.

IL-4 and IL-2 bioassay

Culture supernatants from activated T cells were collected and frozen at –70°C until the day of assay. Serially titrated (2- to 5-fold) culture supernatant samples were assayed for IL-2 and IL-4 using CTLL (40) and CT.4S indicator cells (41), respectively, according to previously described procedures (16). One unit of IL-4 (IL-2) was defined as the amount that induced half maximal proliferation as assessed by [³H]thymidine incorporation of CT.4S (CTLL) indicator cells. Because IL-4 and IL-2 bioassays were performed in microwells containing 0.1 ml of culture medium, 1 U of IL-4 and IL-2 under our assay condition is thus equivalent to 10 U/ml. The lower limit of detection was arbitrarily defined by 3 SD above the mean of [³H]thymidine incorporated by CT.4S (CTLL) cells cultured in medium alone. IL-4 and IL-2 detected by the CT.4S and CTLL bioassays, respectively, were verified by addition of 11B11 anti-IL-4 mAb (42) and S4B6 anti-IL-2 mAb (43). Using rIL-4 (provided by Dr. William E. Paul, National Institutes of Health, Bethesda, MD) as a reference standard, 1 ng/ml IL-4 was found to be equivalent to 2,000 U/ml in the CT.4S assay as performed in our laboratory. A Biological Response modifiers Programme (BRMP) IL-2 reference standard (provided by Dr. Craig Reynolds, Biologic Resources Branch, BRMP, NCI, Frederick, MD) was assayed using the CTLL subline being carried in our laboratory. One unit as defined in our assay was equivalent to 0.229 pg of IL-2 as provided by BRMP (1 ng of BRMP IL-2 was equivalent to 4370 U of IL-2). All IL-2 and IL-4 activities were normalized and are shown in amounts equivalent to the reference IL-2 and IL-4 standards as described above.

IFN-, IL-5, IL-10, and IL-13 detection by ELISA

Culture supernatants from activated T cells were collected and frozen at –70°C until assayed. Briefly, ELISA plates were first coated with R4-6A2 anti-IFN- mAb (44), TFRK-5 anti-IL-5 mAb (45), anti-IL-10 polyclonal Ab, and anti-IL-13 polyclonal Ab. (R&D Systems). Graded amounts of culture supernatant were then added to allow capture of IFN-, IL-5, IL-10, and IL-13, which were then detected by biotin-conjugated XMG1.2 anti-IFN- mAb (BD Biosciences), TRFK-4 anti-IL-5 mAb, anti-IL-10 polyclonal Ab, and IL-13 polyclonal Ab. Biotin groups were then detected by HRP-streptavidin (Pierce), followed by addition of HRP substrate ABTS (Sigma-Aldrich) and absorbance reading at 405 nm. The sources of capture and detecting Abs were as follows: IL-5, purified in our own laboratory; IFN- (BD Biosciences); IL-10 and IL-13 (R&D Systems). Recombinant IFN-, IL-5, IL-10, and IL-13 standard curves ranging from 0.024 to 12.5, 0.01 to 5, 0.01 to 5, and 0.04 to 10 ng/ml, respectively, were established in every assay performed. Recombinant IFN- was obtained from Genzyme; IL-5, IL-10, and IL-13 were obtained from R&D Systems.

IL-2 and IL-4 gene expression by competitive RT-PCR

Expression of IL-2 and IL-4 mRNA was analyzed by competitive RT-PCR (46), according to a modified 2-stage procedure we had established previously (7, 46). The amount of competitive DNA at the equivalence point was arbitrarily normalized to the amount of wild-type mRNA of 500 activated T cells. Relative cytokine mRNA expression was then determined by taking the ratio of the cell number-normalized cytokine equivalence to control -actin equivalence.

IL-12R1, IL-12R2, IFN-R1, IFN-R2, IL-4, GATA-3, c-Maf, and JunB gene expression by real-time RT-PCR

Gene expression levels of IL-12R1, IL-12R2, IFN-R1, IFN-R2, IL-4, GATA-3, c-Maf, JunB, and GAPDH were analyzed by real-time quantitative PCR (LightCycler; Roche Diagnostic Systems). RNA extraction and reverse transcription were performed as described previously (7). All RNA samples were subjected to DNase I treatment (Zymo Research) before reverse transcription. Real-time PCR were performed in duplicates using the Fast-Start DNA master SYBR Green system (Roche Diagnostic Systems) and appropriate primers (Table I). All results were normalized against GAPDH. For each and every experiment, relevant standard curves were established for the mRNAs being assayed. Standard curves were constructed by performing real-time PCR using 10-fold serially titrated relevant PCR products that covered a 6- to 7-log concentration range. Results of gene expression were acceptable only when the correlation coefficients (r²) for the standard curves were 0.99 or greater. For all experiments, the specificity of the reaction products was confirmed by melting profile analysis. To obtain the melting profile, the reactions were held at 65°C for 15 s and then heated to 95°C at a rate of 0.1°C/s while measuring the emitted fluorescence. To insure that the correct sizes of DNA fragments had been amplified, spot-checking by agarose electrophoresis was performed at the end of real-time PCR runs. Each real-time PCR started with an initial 10-min denaturation at 95°C, followed by 45 cycles of PCR. The conditions of PCR were as follows: IL-12R1, 10 s at 95°C, 5 s at 58°C, and 6 s at 72°C; IL-12R2, 10 s at 95°C, 5 s at 61°C, and 6 s at 72°C; IFN-R1 and IFN-R2, 10 s at 95°C, 5 s at 60°C, and 6 s at 72°C; GATA-3, 10 s at 95°C, 5 s at 67°C, and 11 s at 72°C; c-Maf, 10 s at 95°C, 5 s at 65°C, and 7 s at 72°C; JunB, 10 s at 95°C, 5 s at 63°C, and 10 s at 72°C; IL-4, 10 s at 95°C, 5s at 56°C, and 8 s at 72°C; GAPDH, 10 s at 95°C, 5s at 58°C, and 11 s at 72°C.

View this table: [in this window] [in a new window]   Table I. Primer sequences used for nested PCR amplification of V3.2 and V8 CDR3, and for real-time PCRa

The relatively low number of V3.2⁺V8⁺CD161^–CD4⁺CD8^– T cells that can be obtained by cell sorting is insufficient to achieve the cell density required for proper activation. To insure proper activation, B10 (Thy-1^bCD8^b) V3.2⁺V8⁺CD161^–CD4⁺CD8^– thymocytes (8 x 10³ cells/well) were activated in the presence of V(2/7/8)^–CD161^–CD44^lowCD4⁺CD8^– spleen T cells (7 x 10⁴ cells/well) obtained from B10.TL (Thy-1^aCD8^a) congenic mice by plate-bound anti-CD3/CD28 for 2 days. Activated cells were stained with 30H12 FITC-anti-Thy-1.2 mAb (31) and deposited 1 cell/well (Terasaki plate; Robbins Scientific) to which 6 µl of lysis/reverse transcriptase buffer had been added, with the aid of the automated cell deposition unit attachment (FACStarPlus cell sorter; BD Bioisciences) as described previously (7). An aliquot of cDNA was used for IL-4-specific nested PCR as described (7). cDNA samples from 20 each of randomly picked IL-4 mRNA⁺ and IL-4 mRNA^– single cells were further subjected to nested PCR using primers specific for V3.2 and V8 (Table I). Identical conditions (40 cycles at 94°C for 30 s, at 55°C for 60 s, and at 72°C for 90 s) were used for all nested PCR amplifications.

TCR gene sequencing

DNA sequencing of RT-PCR amplified V3.2 and V8 CDR3 regions of randomly picked IL-4 mRNA⁺ and IL-4 mRNA^– single cells was performed using an ABI PRISM 377 DNA Sequencer (Applied Biosystems). D/J nomenclature is as described previously (47, 48, 49, 50). J nomenclature is in accordance with the international ImMunoGeneTics database (51), which can be conveniently accessed at the http://imgt.cnusc.fr:8104 website. The CDR3 sequences of TCR-- and TCR--chain were analyzed by the Pepplot program of Genetics Computer Group. Protein hydrophobicity was determined as described (52).

	Results

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Lack of IL-4 inducibility in V11⁺V3⁺CD161^–CD4⁺CD8^– thymocytes from I-E^k-restricted and TCR Tg mice

Our previous finding of TCR-stimulated prompt IL-4 gene inducibility in a small subpopulation of thymus CD161^– CD44^lowCD4⁺CD8^– T cells may be explained by one of two models. IL-4 inducibility is temporally controlled and is development stage-specific. Alternatively, IL-4 inducibility is restricted to a distinct lineage or subset of CD4⁺ T cells. To discriminate between these two models, we studied IL-4 inducibility using PCC-specific, I-E^k-restricted AND TCR Tg mice. The AND TCR is made up from V11 and V3 chains, and most of the positively selected CD4⁺ T cells expressed the V11⁺V3⁺CD4⁺CD8^– phenotype (Fig. 1A). The very small fraction (3%) of CD4⁺CD8^– T cells that did not coexpress Tg V11 and V3 chains most likely developed through positive selection of endogenously rearranged TCR- and/or - genes. Consistent with our previous observations, IL-4 mRNA was promptly induced within 4 h upon anti-CD3/CD28 stimulation of B10 CD161^–CD44^lowCD4⁺CD8^– thymocytes (Fig. 1B). Similar levels of IL-4 gene activation were seen at 24-h and 48-h time points. In marked contrast, anti-CD3/CD28 failed to induce IL-4 mRNA expression in AND CD161^–CD44^lowCD4⁺CD8^– thymocytes at all time points studied (Fig. 1B). Both B10 and AND CD161^–CD44^lowCD4⁺CD8^– thymocytes displayed similar IL-2 mRNA induction kinetics (Fig. 1B), indicating that the failure for IL-4 gene inducibility in AND CD161^–CD44^lowCD4⁺CD8^– thymocytes was not the result of a generalized defect in TCR-mediated signal transduction. Consistent with the mRNA results, AND V11⁺V3⁺CD161^–CD4⁺CD8^– thymocytes produced bioactive IL-2 but not IL-4 in response to anti-CD3/CD28 or to PCC peptide stimulation (Fig. 1C). AND CD161^–CD4⁺CD8^– thymocytes that did not coexpress the V11/V3 Tg TCR produced neither IL-2 nor IL-4 in response to PCC antigenic stimulation, but produced both IL-2 and IL-4 in response to anti-CD3/CD28 (Fig. 1C). Because the stages of intrathymic-positive and -negative selection processes are similar for developing T cells regardless of the Tg or endogenous origin of TCR, the failure of AND CD161^–CD4⁺CD8^– thymocytes to express IL-4 inducibility is inconsistent with the development stage-specific model of IL-4 gene inducibility. Extrapolating results from Tg mice, however, is always complicated by potential positional effects caused by the insertion of transgenes into the host genome. The small population of CD161^–CD4⁺CD8^– thymocytes from the AND Tg mice that did not coexpress Tg TCR and therefore expressed endogenously rearranged TCR, in contrast, produced significant levels of IL-4 upon CD3/CD28 stimulation. Finding of IL-4 inducibility in T cells that did not coexpress the V11 and V3 transgenes of the AND TCR but not in V11⁺V3⁺ subset among CD4⁺CD8^– T cells rules out position effects caused by transgene insertion because both of these subsets are genetically identical at the level of the genome.

View larger version (31K): [in this window] [in a new window]   FIGURE 1. Extremely low IL-4 production by CD4+CD8– thymocytes from AND TCR Tg mice. A, Thymocytes from AND TCR Tg mice were stained with FITC-anti-V11, biotinyl-anti-CD8, Cy5-anti-V3, TR-anti-CD4, followed by PE-streptavidin to detect the biotinyl group. Correlated V11/V3 expression within CD4+CD8– thymocytes is shown. B, Relative levels of IL-4 (solid symbols) and IL-2 (open symbols) mRNA expression at indicated time points after CD161–CD44lowCD4+CD8– thymocytes from AND TCR Tg mice (circle symbols) and B10 mice (square symbols) were stimulated by plate-bound anti-CD3/CD28. Relative IL-4/-actin ratios for 0-, 4-, 24-, and 48-h time points: AND, 0, 0.0019, 0.0152, and 0.0038, respectively; B10, 0.01, 0.66, 0.83 and 0.5, respectively. Relative IL-2/-actin ratios for 0-, 4-, 24-, and 48-h stimulated time points: AND, 0, 1.598, 1.997, and 1.598, respectively; B10, 0, 1.515, 2.13, and 1.89, respectively. C, Duplicate cultures of CD161–CD4+CD8–, V11+V3+CD161–CD4+CD8–, and (V11/V3)–CD161–CD4+CD8– thymocytes from AND TCR Tg mice and CD161–CD4+CD8– thymocytes from B10 mice were stimulated by anti-CD3/CD28 or by PCC peptide in the context of I-Ek as indicated. Average values of IL-4 () and IL-2 () bioactivities released into day 3 culture supernatant are shown. For simplicity, AND CD161–CD4+CD8– thymocytes that did not coexpress high levels of both V11/V3 were written as (V11/V3)– even though they contain cells that express V11 or V3 in combination with endogenously rearranged TCR- or --chains, respectively. Unstimulated T cell controls produced <0.01 ng/ml IL-2 and no detectable IL-4. Data shown in all panels were typical of one of the two independently performed experiments.

If indeed the distinct subset/lineage model is correct, the molecular and cellular basis responsible for IL-4 inducibility may be determined during intrathymic-positive selection, a process critically dependent on the specificity of TCR and may therefore display biased TCR family usage. To examine this possibility, CD161^–CD44^lowCD4⁺CD8^– thymocytes that expressed various V and V TCR families were examined for IL-4 inducibility. For B10 mice, V2⁺, V7⁺, V8⁺, and V3.2⁺ subsets of CD161^–CD44^lowCD4⁺CD8^– T cells expressed 2.6-, 3.2-, 2.8-, and 2.6-fold enhanced IL-4 inducibility, respectively (Fig. 2A). In contrast, highly significant 5.3- to 12-fold depletion of IL-4 inducibility was observed for CD161^–CD44^lowCD4⁺CD8^– thymocytes that expressed V3, V6, V2, V8, and V11. A moderate 2- and 2.8-fold depletion was observed for those that expressed V5 and V11, respectively. The highest IL-4 production response of 155 pg/ml by V7⁺ cells was 39 times that of the lowest response of 4 pg/ml by V3⁺ T cells. To address whether the TCR family associated bias in IL-4 inducibility was a general phenomenon, IL-2 and IL-4 production was examined in additional experiments (Fig. 2B). Again, biased IL-4 inducibility was observed for V2-, V7-, V8-, and V3.2-bearing CD161^–CD44^lowCD4⁺CD8^– thymocytes, with highly significant reduction in all other V/V families examined. However, the failure of certain T cell V families to produce IL-4 was not caused by inadequate levels of stimulation because they all expressed similar levels of IL-2 gene activation (Fig. 2B). We next examined biased TCR- family usage for BALB/c mice, a strain prone to mount Th2 immune responses (53, 54). Similar and significant 3.8-, 4.1-, and 4.6-fold enhanced IL-4 inducibility was observed for V2⁺, V7⁺, and V8⁺ subsets of CD44^lowCD4⁺CD8^– thymocytes, respectively. V6⁺, V2⁺, and V8⁺ subsets, in contrast, displayed 10.8-, 9.6-, and 3.6-fold depleted IL-4 inducibility (Fig. 2A). The highest IL-4 inducibility response of 792 pg/ml by V8⁺ T cells was 49.5 times that of the lowest response of 16 pg/ml by V6⁺ T cells. Association of IL-4 inducibility in CD161^–CD44^lowCD4⁺CD8^– thymus T cells with biased TCR- family usage strongly favors a distinct subset/lineage model and is inconsistent with the temporally controlled development stage-specific model.

View larger version (34K): [in this window] [in a new window]   FIGURE 2. Biased TCR usage by IL-4-producing CD161–CD44lowCD4+CD8– thymocytes. A, BALB/c thymocytes depleted of CD8+ cells were stained with Cy5-anti-CD44, TR-anti-CD4, biotin-anti-CD8, and FITC-conjugated TCR family-specific mAb as indicated, followed by detection of biotinyl group with PE-streptavidin. B10 thymocytes were stained with the same set of reagents, except that biotin-anti-CD161 mAb was added. CD44lowCD4+CD8– BALB/c thymocytes or CD161–CD44lowCD4+CD8– B10 thymocytes that expressed indicated TCR- or - families were stimulated in duplicate cultures by anti-CD3/CD28 in the presence of B cell blasts. Because the IL-4 response by BALB/c is significantly higher than that of B10, average values of IL-4 bioactivities released into day 3 culture media were normalized by arbitrarily setting responses of their respective total CD44lowCD4+CD8– thymocytes to 1. Data shown are typical of one of three independently performed experiments. B, Production of IL-4 () and IL-2 () by indicated V- or V-bearing CD161–CD44lowCD4+CD8– thymocytes from B10 mice were stimulated by anti-CD3/CD28 mAbs as above; average values of IL-2 and IL-4 bioactivities released into culture media by indicated TCR- and - families are shown (Vt, total T cells). Data shown are typical of one of three to five independently performed experiments. C, Vtotal (), V(2/7/8)+ ( ), and V(2/7/8)– () subsets of DX5–CD44lowCD4+CD8– thymocytes from a group of three BALB/c mice and DX5–CD161–CD44lowCD4+CD8– thymocytes from groups of three B10.A(4R), ICR, and CBA, and C57BL/6 mice were stimulated by anti-CD3/CD28 as above. Average IL-4 bioactivities released into culture supernatant from each of the groups of three mice are shown. Unstimulated T cell controls produced <0.01 ng/ml IL-2 and no detectable IL-4. The amount of IL-2 (ng/ml) produced were as follows: BALB/c, 14.4 ± 2.2; B10.A(4R), 27.7 ± 2.5; ICR, 32.7 ± 4.4; CBA, 23.3 ± 3.5; B6, 20.5 ± 2.0. D, CD44lowCD4+CD8– thymocytes as well as their V8+ and V8– subsets were obtained from three individual DO11.10 TCR Tg mice and stimulated with anti-CD3/CD28 in the presence of B cell blasts. V(2/7/8)+ and V(2/7/8)– subsets of CD44lowCD4+CD8– thymocytes from three individual BALB/c mice were also stimulated under identical conditions. Average values of IL-4 () and IL-2 () bioactivities released into day 3 culture supernatant from the indicated responding cell types are shown. Unstimulated T cell controls produced <0.01 ng/ml IL-2 and no detectable IL-4. E, V(2/7/8)+ and V(2/7/8)– subsets of CD44lowCD4+CD8– spleen cells from three individual BALB/c mice, and V(2/7/8)+ and V(2/7/8)– subsets of CD161–CD44lowCD4+CD8– spleen cells from three individual B10 mice were stimulated by anti-CD3/CD28 in triplicate cultures. Average values of IL-4 () and IL-2 () production are shown. Unstimulated T cell controls produced <0.01 ng/ml IL-2 and undetectable IL-4.

One possible interpretation of the association of IL-4 inducibility and biased usage of V(2/7/8) by both CD161^–CD44^lowCD4⁺CD8^– thymocytes and NKT cells is that IL-4 inducibility is determined by TCR- family usage rather than the lineage it belongs to. This possibility was tested by studying IL-4 inducibility in CD44^lowCD4⁺CD8^– thymocytes of the DO11.10 TCR (V8/V3.1) Tg mice (Fig. 2D). Consistent with results already presented, highly enriched and depleted IL-4 inducibility was respectively observed for V(2/7/8)⁺ and V(2/7/8)^– subsets of CD44^lowCD4⁺CD8^– thymocytes from BALB/c (wild-type control) mice. However, V8⁺CD44^lowCD4⁺CD8^– thymocytes from DO11.10 Tg mice showed slightly depressed rather than enriched IL-4 inducibility. Alterations in the levels of IL-4 inducibility did not result from inadequate stimulation because similar levels of IL-2 production was found for all subsets studied. Thus, the mere expression of V8 transgene was insufficient in conferring IL-4 inducibility.

Because immune response takes place in the periphery and not in the thymus, we examined IL-4 inducibility for CD44^lowCD4⁺CD8^– spleen cells (Fig. 2E). Although IL-4 production responses were readily detectable for CD44^lowCD4⁺CD8^– spleen cells of both B10 and BALB/c mice, they were 5- to 10-fold decreased in comparison to their thymic counterparts, consistent with our previously published results (7). In addition, highly significant enrichment and depletion of IL-4 inducibility responses were observed for V(2/7/8)⁺ and V(2/7/8)^– subsets for both B10 and BALB/c mice, respectively. Comparable levels of IL-2 production by all subsets studied indicate that the different IL-4 inducibility responses were not due to inadequate signaling through the TCR.

CD161^–CD44^lowCD4⁺CD8^– thymocytes do not respond to -GalCer

The shared V(2/7/8) bias between NKT cells and the CD161^–CD44^lowCD4⁺CD8^– thymocytes we are working with suggests relatedness between these phenotypically distinct T cell subsets. In addition, the presence of a small number of NKT cell precursors capable of high level IL-4 inducibility has been found within the CD161^–CD44^lowCD4⁺CD8^– subset (8, 9). These results prompted us to investigate the relative contribution of NKT precursors to the IL-4 inducibility response we have observed for CD161^–CD44^lowCD4⁺CD8^– thymocytes. B10 CD161^–CD44^lowCD4⁺CD8^– thymocytes were sorted into V(2/7/8)⁺ and V(2/7/8)^– subsets and subjected to stimulation by anti-CD3/CD28 or by -GalCer in the context of CD1d. For the V(2/7/8)⁺ subset, anti-CD3/CD28 stimulated readily detectable IL-2 and IL-4 production (Fig. 3A). Stimulation by -GalCer/CD1d, in contrast, did not result in either IL-2 or IL-4 production (Fig. 3B). Consistent with results already presented, the V(2/7/8)^– subset produced IL-2 but not IL-4 in response to TCR stimulation (Fig. 3A). As expected, NKT (CD161⁺CD44^high) cells responded to anti-CD3/CD28 as well as to -GalCer/CD1d by producing IL-2 and IL-4 (Fig. 3, A and B) (55, 56, 57). To further rule out the participation of NKT precursors in the IL-4 inducibility response we have observed for CD161^–CD44^lowCD4⁺CD8^– thymocytes, cells that express the invariant V14/J18 TCR were identified with -GalCer-loaded CD1d DimerX and removed by cell sorting. Clearly, removal of the invariant V14/J18 TCR⁺ T cells did not affect the ability of CD161^–CD44^lowCD4⁺CD8^– thymocytes to produce either IL-2 or IL-4 in response to anti-CD3/CD28 stimulation (Fig. 3C). Another unique feature of the NKT cell is its ability to produce IFN- in addition to IL-4. It is therefore of interest to study whether the CD161^–CD44^lowCD4⁺CD8^– thymocytes we are working with also make IFN-. Consistent with previously published results (58), NKT cells made a robust IFN- response to anti-CD3/CD28 stimulation and to Ag-independent effects of IL-12 + IL-18 (Fig. 3D). Stimulation by -GalCer in the context of CD1d also resulted in significant IFN- production, although at lower levels. In marked contrast, very low levels of 2.3 and 2.5 ng/ml IFN- production were respectively observed for anti-CD3/CD28-stimulated V(2/7/8)⁺ and V(2/7/8)^– subsets of CD161^–CD44^lowCD4⁺CD8^– thymocytes. In addition, IFN- production was undetectable for V(2/7/8)⁺ and V(2/7/8)^– subsets of CD161^–CD44^lowCD4⁺CD8^– thymocytes in response to IL-12 + IL-18 (Fig. 3D). The relative low level of TCR-stimulated IFN- inducibility in CD161^– CD44^lowCD4⁺CD8^– thymocytes is consistent with reports that show that T cells capable of high-level IFN- production are confined within the CD44^high subset (59).

View larger version (31K): [in this window] [in a new window]   FIGURE 3. V(2/7/8)+CD161–CD44lowCD4+CD8– thymocytes lack reactivity to GalCer-CD1d. A, Duplicate cultures of CD161+CD44highCD4+CD8–, V(2/7/8)+CD161–CD44lowCD4+CD8–, and V(2/7/8)–CD161–CD44lowCD4+CD8– thymocytes from B10 mice were stimulated by plate-bound anti-CD3/CD28 (7 x 104 cells/well) as in Fig. 1B. Average values of IL-2 () and IL-4 () bioactivities present in day 3 culture supernatant are shown. B, B10 CD161+CD44highCD4+CD8–, V(2/7/8)+CD161–CD44lowCD4+CD8–, and V(2/7/8)–CD161–CD44lowCD4+CD8– thymocytes (2 x 104 cells/well) were stimulated by -GalCer (2 µg/ml) presented by sorted CD1dhigh spleen cells (105 cells/well). Average values of IL-2 () and IL-4 () bioactivities present in day 4 culture supernatant are shown. C, B10 CD161–CD44lowCD4+CD8– thymocytes and its subset depleted of cells identified by -GalCer/CD1d DimerX were stimulated by anti-CD3/CD28 mAbs as in Fig. 1C. IL-2 () and IL-4 () bioactivities present in day 3 culture supernatant are shown. Data shown are typical of one of two independently performed experiments. Unstimulated T cell controls produced <0.01 ng/ml IL-2 and no detectable IL-4. D, B10 CD161+CD44highCD4+CD8– thymocytes and V(2/7/8)+ and V(2/7/8)– subsets of CD161–CD44lowCD4+CD8– thymocytes were stimulated by anti-CD3/CD28 (7 x 104 cells/well) and -GalCer (2 x 104 cells/well), as in A and B, respectively. These same T cell subsets were also cultured in media supplemented with IL-12 + IL-18. IFN- in culture supernatant from anti-CD3/CD28-, -GalCer-, and IL-12/IL-18-stimulated cells was collected 3, 4, and 3 days after stimulation, respectively, and displayed by solid, gray, and open bars, respectively. E, Expression of IL-12R1, IL-12R2, IFN-R1, and IFN-R2 was examined for B10 V(2/7/8)+CD161–CD44lowCD4+CD8–, V(2/7/8)–CD161–CD44lowCD4+CD8–, and CD161+CD44hiCD4+CD8– (NKT) thymocytes by real-time RT-PCR. Normalized levels of expression for V(2/7/8)+ ( ), V(2/7/8)– (), and NKT cells () are as follows: IL-12R1:GAPDH, 0.013, 0.008, and 0.108, respectively; IL-12R2:GAPDH, 0.0031, 0.0034, and 0.32, respectively; IFN-R1:GAPDH, 0.541, 0.379, and 2.074, respectively; IFN-R2:GAPDH, 0.28, 0.31, and 0.037, respectively. Data shown are typical of one of the two independently performed experiments.

Acquisition of the IL-4 inducibility phenotype in CD161^–CD44^lowCD4⁺CD8^– thymocytes is independent of ₂m, CD1d, p59^fyn, stat6, and IL-4R

TCR-stimulated prompt IL-4 inducibility is a well-characterized property of NKT cells (63) and of TCR repertoire-diverse conventional T cells that have undergone Th2/Tc2 differentiation (1, 2, 3). The development of NKT cells requires the MHC-I-like CD1d and is therefore highly deficient in ₂m- and CD1d-null mice (4, 64). In addition, NKT cell development has been shown to be severely affected in p59^fyn-null mice (65). CD161^–CD44^lowCD4⁺CD8^– thymocytes from ₂m- and CD1d-null mice not only produced IL-4 in response to anti-CD3/CD28 stimulation, they showed the same V2, V7, and V8 bias as did their wild-type controls (Fig. 4A). Because the IL-4 inducibility response was higher in CD161^–CD44^lowCD4⁺CD8^– thymocytes from CD1d-null than ₂m-null mice (Fig. 4A), a side-by-side comparison was made between CD1d- or ₂m-null mice with their respective wild-type BALB/c or B10 controls (Fig. 4B). Higher levels of IL-4 inducibility were again found for CD161^–CD44^lowCD4⁺CD8^– thymocytes from CD1d- than ₂m-null mice, and they both showed biased V(2/7/8) usage. This is most likely due to the different genetic background of these mice because similar differences were also observed in their respective BALB/c and B10 wild-type controls. For p59^fyn-null mice, CD161^–CD44^lowCD4⁺CD8^– thymocytes also produced IL-4 with V(2/7/8) bias in response to anti-CD3/CD28 stimulation (Fig. 4C).

View larger version (25K): [in this window] [in a new window]   FIGURE 4. IL-4 gene inducibility and biased TCR usage by CD161–CD44lowCD4+CD8– thymocytes is independent of 2m, CD1d, p59fyn, stat6, and IL-4R. Except for BALB/c, CD1d-null, and IL-4R-null strains, CD161–CD44lowCD4+CD8– thymocytes that expressed indicated TCR V families were isolated. For BALB/c, CD1d-null, and IL-4R-null strains, CD44lowCD4+CD8– thymocytes that expressed indicated TCR V families were isolated. Anti-CD3/CD28-stimulated IL-4 production was performed as in Fig. 1C and is shown as mean ± SD (n = number of mice). A, IL-4 production by CD161–CD44lowCD4+CD8– and CD44lowCD4+CD8– thymocytes and their V2+, V7+, V8+, and V(2/7/8)– subsets from 2m-null (n = 2) and CD1d-null (n = 2) mice, respectively. B, IL-4 production by CD161–CD44lowCD4+CD8– thymocytes (Vtotal), their V(2/7/8)+ and V(2/7/8)– subsets from 2m-null (n = 2) and B10 (n = 2) mice, and by CD44lowCD4+CD8– (Vtotal), their V(2/7/8)+ and V(2/7/8)– subsets from CD1d-null (n = 2) and BALB/c (n = 2) mice. C, IL-4 production by CD161–CD44lowCD4+CD8– thymocytes (Vtotal), their V(2/7/8)+ and V(2/7/8)– subsets from p59fyn-null (n = 3) and B10 (n = 3) mice. D, IL-4 production by Vtotal, V2+, V7+, V8+, and V(2/7/8)– subsets of CD161–CD44lowCD4+CD8– thymocytes from stat6-null mice (n = 3) and of CD44lowCD4+CD8– thymocytes from IL-4R-null mice (n = 3). Unstimulated T cell controls did not produce detectable IL-4. E, CD161–CD44lowCD4+CD8– thymocytes were isolated from MHC-II-null and B10 mice and stimulated by anti-CD3/CD28 as in Fig. 1C. Due to the scanty nature of CD4+ T cells, only a single activation culture was set up for each of the three individual MHC-II-null mice. Wild-type CD161–CD44lowCD4+CD8– thymocytes from a B10 mouse were set up in triplicate. IL-4 gene expression was determined by real-time PCR for total CD4+ T cells and their V(2/7/8)–, V(2/7/8)+, V3.2+ subsets that had been activated for 2 days. The low percentage of V3.2+ cells necessitated pooling of approximately one-third of activated cells from the three MHC-II-null mice, to obtain sufficient numbers of V3.2+ cells for analysis of IL-4 gene expression. Duplicate real-time PCR were performed for each RNA sample as described in Materials and Methods, and mean ± SD were calculated. To facilitate comparison, relative IL-4 expression levels for anti-CD3/CD28-stimulated total CD161–CD44lowCD4+CD8– thymocytes from MHC-II-null and from B10 mice were arbitrarily assigned a value of 1, against which IL-4 gene expression levels for V(2/7/8)– (), V(2/7/8)+ ( ), and V3.2+ () subsets were normalized. The actual IL-4:GAPDH ratios for anti-CD3/CD28-stimulated total CD161–CD44lowCD4+CD8– thymocytes were as follows: MHC-II-null mouse no. 1, 31.5; MHC-II-null mouse no. 2, 24.3; MHC-II-null mouse no. 3, 22.2; pooled MHC-II-null mice nos. 1–3, 26.4; B10, 0.25. Unstimulated T cells did not produce detectable IL-4.

Because CD161^–CD44^lowCD4⁺CD8^– thymocytes acquire IL-4 inducibility in the absence of CD1d, ₂m, and p59^fyn, they are most likely MHC-II-restricted CD4⁺ T cells. Because the few CD4⁺ T cells that develop in MHC-II-null background are CD1d-restricted (66) and cannot possibly be restricted by MHC II, analysis of biased TCR- family usage may shed light on the lineage origin of CD161^–CD44^lowCD4⁺CD8^– capable of IL-4 inducibility. We therefore examined biased V(2/7/8) and V3.2 family usage in CD161^–CD44^lowCD4⁺CD8^– thymocytes of MHC-II-null mice. Due to the scanty nature of CD4⁺ T cells in MHC-II-null mice, it was not possible to sort out sufficient numbers of CD4⁺ T cells to perform IL-4 inducibility in ways that had been done up to this point. Instead, real-time PCR analysis of IL-4 gene expression was studied. Consistent with results already shown, highly enriched IL-4 gene expression was seen for V(2/7/8)⁺ and V3.2⁺ subsets of CD161^–CD44^lowCD4⁺CD8^– thymocytes (Fig. 4E). V(2/7/8)^– subset, in contrast, showed highly depleted IL-4 inducibility. In marked contrast, neither biased V(2/7/8) nor V3.2 TCR family usage was correlated with IL-4 inducibility for CD161^–CD44^lowCD4⁺CD8^– thymocytes from MHC-II-null mice.

V(2/7/8)⁺CD161^– CD44^lowCD4⁺CD8^– thymocytes are also capable of TCR-stimulated IL-5, IL-10, and IL-13 production

The prompt IL-4 production by V(2/7/8)⁺CD161^–CD44^lowCD4⁺CD8^– thymocytes raises the question of whether other Th2 cytokines are also produced. V(2/7/8)⁺ and V(2/7/8)^– subsets of CD161^–CD44^lowCD4⁺CD8^– thymocytes were stimulated by anti-CD3/CD28, and secreted cytokines were measured (Fig. 5). Consistent with results already presented, highly enriched and depleted IL-4 production was observed for V(2/7/8)⁺ and V(2/7/8)^– subsets, respectively, of CD161^– CD44^lowCD4⁺CD8^– thymocytes (Fig. 5B). No bias in V(2/7/8) usage was found for IL-2 and IFN- responses (Fig. 5, A and F). For IL-5 and IL-10 responses, significant enrichment was observed for V(2/7/8)⁺CD161^–CD44^lowCD4⁺CD8^– thymocytes, whereas substantial depletion was found for V(2/7/8)^–CD161^–CD44^lowCD4⁺CD8^– thymocytes (Fig. 5, C and D). Although IL-13 production responses showed highly variable mouse-to-mouse variation, the general trend of enriched and depleted IL-13 inducibility in V(2/7/8)⁺ and V(2/7/8)^– subsets of CD161^–CD44^lowCD4⁺CD8^– thymocytes, respectively, was observed at the individual mouse level (Fig. 5E).

View larger version (25K): [in this window] [in a new window]   FIGURE 5. V(2/7/8)+ subset of CD161–CD44lowCD4+CD8– thymocytes are also enriched in TCR-stimulated IL-5, IL-10, and IL-13 inducibilities. CD161–CD44lowCD4+CD8– thymocytes and their V(2/7/8)+ and V(2/7/8)– subsets were obtained from each of four B10 mice and stimulated as in Fig. 1C. Day 3 culture supernatant was assayed for IL-2 (A), IL-4 (B), IL-5 (C), IL-10 (D), IL-13 (E), and IFN- (F). To facilitate tracking of cytokine production response at the level of individual mice, different symbols were used to denote each of the four mice. Horizontal bars were added to show average values of cytokine secretion. Unstimulated T cell controls produced <0.01 ng/ml IL-2 and no detectable IL-4, IL-5, IL-10, IL-13, and IFN-.

Because GATA-3, c-Maf, and JunB transcription factors have been shown to regulate IL-4 gene activation, we examined whether the prompt IL-4 inducibility in V(2/7/8)⁺CD161^–CD44^lowCD4⁺CD8^– thymocytes is due to increased endogenous expression of these transcription factors (Fig. 6). NKT cells were used as a control because they are known to undergo prompt IL-4 gene activation. Consistent with a role played by c-Maf in prompt IL-4 inducibility in NKT cells (67), a strikingly elevated c-Maf expression was observed for freshly isolated NKT cells than either V(2/7/8)⁺ or V(2/7/8)^– subsets of CD161^–CD44^lowCD4⁺CD8^– thymocytes (Fig. 6C). However, the level of c-Maf expression was only slightly (25%) higher in V(2/7/8)⁺ than the V(2/7/8)^– subset of CD161^–CD44^lowCD4⁺CD8^– thymocytes. The levels of GATA-3 and JunB expression were similar for V(2/7/8)^– and V(2/7/8)⁺ subsets. After TCR stimulation, GATA-3, c-Maf, and JunB all underwent highly significant down-regulation in the V(2/7/8)^– subset, but not the V(2/7/8)⁺ subset of CD161^–CD44^lowCD4⁺CD8^– thymocytes. Thus, the levels of GATA-3, c-Maf, and JunB expression for freshly isolated (unstimulated) V(2/7/8)⁺CD161^–CD44^lowCD4⁺CD8^– thymocytes were 0.9-, 1.2-, and 0.9-fold those found in V(2/7/8)^–CD161^–CD44^lowCD4⁺CD8^– thymocytes. By 20 h after TCR stimulation, GATA-3, c-Maf, and JunB expression levels in V(2/7/8)⁺CD161^–CD44^lowCD4⁺CD8^– thymocytes had changed to 5-, 4-, and 5-fold those observed for V(2/7/8)^–CD161^–CD44^lowCD4⁺ thymocytes.

View larger version (13K): [in this window] [in a new window]   FIGURE 6. GATA-3, c-Maf, and JunB mRNA expression by V(2/7/8)– and V(2/7/8)+ subsets of CD161–CD44lowCD4+CD8– thymocytes. V(2/7/8)+ and V(2/7/8)– subsets of CD161–CD44lowCD4+CD8– thymocytes were obtained from B10 mice and stimulated by plate-bound anti-CD3/CD28 as in Fig. 1B. IL-4, GATA-3, c-Maf, and JunB mRNA expression at indicated time points were analyzed by real-time PCR. Relative expression levels of IL-4 (A), GATA-3 (B), c-Maf (C), and JunB (D) for CD161+CD44highCD4+CD8– thymocytes (), V(2/7/8)– (), and V(2/7/8)+ ( ) subsets of CD161–CD44lowCD4+CD8– thymocytes that were unstimulated or stimulated for 4 and 20 h are shown. Data shown are typical of one of the two independently performed experiments.

The TCR used by the vast majority of NKT cells consists of the invariant V14-J18 paired with V8, V7, or V2 (68). This mostly invariant TCR--chain usage by NKT cells places severe constraints on the number of Ags they as a population can recognize. Ag-inexperienced naive CD4⁺ T cells that simultaneously possess properties of rapid TCR-stimulated IL-4 inducibility and a diverse TCR repertoire has not been described so far. Should such cells exist, significant roles can be expected of them in the regulation of immune responses. Because V3.2⁺V8⁺ CD161^–CD4⁺CD8^– had been shown to be highly enriched in IL-4 inducibility response, they were stimulated by anti-CD3/CD28 and CDR3 sequences for both TCR-- and --chains were determined at the single-cell level. Of the 324 single cells analyzed, 25% (81 ÷ 324) were IL-4 mRNA⁺. DNA sequencing of RT-PCR-amplified V3.2 and V8 CDR3 regions of randomly picked IL-4 mRNA⁺ and IL-4 mRNA^– single cells was performed. In contrast to the lack of N region nucleotide additions/deletions of V14-J18 invariant chains of NKT cells, IL-4-producing V3.2⁺V8⁺CD161^–CD4⁺CD8^– thymocytes displayed a high degree of variability in V3.2 as well as V8 CDR3 regions. TCR on NKT cells recognizes glycolipids in the context of CD1d (69). The hydrophobic tail of the glycolipid binds CD1d and the hydrophilic residues are recognized by TCR (56, 70). Among the 20 analyzed IL-4 mRNA⁺ cells, none expressed identical CDR3 sequences for either TCR-- or --chains, and all showed different usage of J-D/J pairs. No obvious difference in hydrophobicity and net charge was found for V and V CDR3 amino acid sequences from IL-4 mRNA^– vs IL-4 mRNA⁺ cells (Table II).

View this table: [in this window] [in a new window]   Table II. Deduced CDR3 amino acid sequences of IL-4 mRNA+ single V3.2+ V8+ CD161– CD4+ CD8– T cellsa

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

NKT cells were originally described as a subset of CD4⁺ T cells that expresses the NK1.1 Ag (CD161), hence the name NKT cells. The bulk of the studies performed in this study relied on the use of the CD44^highCD161⁺ phenotype to exclude NKT cells. Although these markers are convenient and generally acceptable mark