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

This Article Abstract Full Text (PDF) A correction has been published 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 Denning, T. L. Articles by Kronenberg, M. Search for Related Content PubMed PubMed Citation Articles by Denning, T. L. Articles by Kronenberg, M.

Mouse TCR⁺CD8 Intraepithelial Lymphocytes Express Genes That Down-Regulate Their Antigen Reactivity and Suppress Immune Responses^1,2

Timothy L. Denning^3,*, Steve Granger, Daniel Mucida^*, Ryan Graddy^*, Georges Leclercq, Weiguo Zhang, Karen Honey^¶, Jeffrey P. Rasmussen^¶, Hilde Cheroutre^*, Alexander Y. Rudensky^¶ and Mitchell Kronenberg^3,*

^* Division of Developmental Immunology, La Jolla Institute for Allergy and Immunology, and Gemini Science. La Jolla, CA 92037; Department of Clinical Chemistry, Microbiology, and Immunology, University of Ghent, University Hospital, Ghent, Belgium; Department of Immunology, Duke University Medical Center, Durham, NC 27710; and ^¶ Howard Hughes Medical Institute, University of Washington, Seattle, WA 98195

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Mouse small intestine intraepithelial lymphocytes (IEL) that express TCR and CD8 homodimers are an enigmatic T cell subset, as their specificity and in vivo function remain to be defined. To gain insight into the nature of these cells, we performed global gene expression profiling using microarray analysis combined with real-time quantitative PCR and flow cytometry. Using these methods, TCR⁺CD8 IEL were compared with their TCR⁺CD8⁺ and TCR⁺ counterparts. Interestingly, TCR⁺CD8 IEL were found to preferentially express genes that would be expected to down-modulate their reactivity. They have a unique expression pattern of members of the Ly49 family of NK receptors and tend to express inhibitory receptors, along with some activating receptors. The signaling machinery of both TCR⁺CD8 and TCR⁺ IEL is constructed differently than other IEL and peripheral T cells, as evidenced by their low-level expression of the linker for activation of T cells and high expression of the non-T cell activation linker, which suppresses T cell activation. The TCR⁺CD8 IEL subset also has increased expression of genes that could be involved in immune regulation, including TGF-₃ and lymphocyte activation gene-3. Collectively, these data underscore the fact that, while TCR⁺CD8 IEL resemble TCR⁺ IEL, they are a unique population of cells with regulated Ag reactivity that could have regulatory function.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Mouse intestinal intraepithelial lymphocytes (IEL)⁴ are predominantly CD8⁺ T cells (reviewed in Ref. 1), but they can be divided into several subtypes based on the expression of TCR isotypes and coreceptors (1, 2). One population of cells expresses the TCR and CD8, the heterodimeric form of CD8 typically found on CD8⁺ T cells in the spleen, lymph node, and other sites. These cells, hereafter referred to as TCR⁺CD8 IEL, are believed to be mainly conventional activated/memory CD8⁺ T cells that have encountered Ag in the periphery (3). Perhaps as a result of contact with dendritic cells (DC) in the mesenteric lymph nodes and Peyer’s patches (4, 5, 6), they have been instructed to migrate to the intestine and adapt to the microenvironment there. Like other IEL, they express ₇-containing integrins (7), and they may coexpress CD8 homodimers along with CD8.

There are two additional populations of mouse IEL that do not express either of the well-defined CD4 or CD8 TCR coreceptors. These lymphocytes express CD8 exclusively, which does not function as an effective TCR coreceptor (8, 9, 10, 11, 12), perhaps in part because it does not localize to lipid rafts (13). These T cells include the TCR⁺ IEL and a population of TCR ⁺ IEL that express CD8 only. These two subpopulations are unconventional T cells that share the expression of a number of genes while lacking expression of several molecules found on most other T lymphocytes, such as CD2 (14), CD28 (15), and CD90 (16). The specificity and function of both of these IEL subpopulations remain incompletely characterized, but through their production of keratinocyte growth factor, TCR⁺ IEL may participate in resolving inflammatory lesions in the intestine (17).

TCR⁺CD8 IEL require ₂m expression, but they do not require expression of the classical class I molecules K and D for their development and/or homeostasis (18, 19, 20). These TCR⁺CD8 IEL may include cells reactive with either classical class I molecules, nonclassical class I molecules, or even class II molecules. The results from analysis of their V repertoire in wild-type mice (21) and the analysis of TCR-transgenic mice (22, 23, 24) both support the concept that TCR⁺CD8 IEL are self-reactive. Recent evidence suggests that these cells develop from a unique subset of double-positive intermediates in the thymus (25, 26), and the evidence indicates that exposure to self-agonists there imparts their distinct phenotype (26, 27, 28). An agonist-driven positive selection process is a feature they share with at least two other specialized T lymphocyte subpopulations, Foxp3⁺ regulatory T cells (29) and NKT cells, with an invariant V14 (V14i) TCR (30).

The function of TCR⁺CD8 IEL has not been well characterized, but in one study, they could prevent colitis induced by the transfer of CD4⁺CD45RB^high T cells to immune-deficient recipients, whereas TCR⁺CD8 IEL and TCR⁺ IEL were incapable of doing so (31). Colitis prevention required TCR-mediated activation of the TCR⁺CD8 IEL by self-Ag in the recipients, and cells derived from IL-10^–/– mice were not effective, although the TCR⁺CD8 did not produce IL-10 when stimulated in vitro. In a second study, TCR-transgenic TCR⁺CD8 IEL were shown to express increased levels of TGF-₁ mRNA, with relatively high constitutive levels that were increased only modestly after activation with cognate Ag (32). Collectively, these two studies suggest that TCR⁺CD8 IEL could have a regulatory function in the mucosal immune system, but the means by which they might regulate immune responses remain undefined.

To better understand the function of TCR⁺CD8 IEL, we have conducted a microarray analysis of gene expression by these cells. Our findings highlight unique properties of these cells not appreciated previously, including the expression of an atypical pattern of NK receptors, several of which are inhibitory, as well as the expression of an unusual signaling apparatus and a novel expression pattern of molecules that could be involved in mucosal immune regulation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Mice

C57BL/6J mice were obtained from The Jackson Laboratory. Mice were maintained under specific pathogen-free conditions in the vivarium of the La Jolla Institute for Allergy and Immunology, and all animal studies were approved by the Institutional Animal Care and Use Committee.

Isolation of IEL

Mucosal lymphocytes were isolated as described previously (33). Briefly, small intestines were removed and carefully cleaned from their mesentery, Peyer’s patches were excised, and intestines were opened longitudinally and washed of fecal contents. Intestines were then cut into 0.5-cm pieces, transferred into 50-ml conical tubes, and shaken at 250 rpm for 30 min at 37°C in Hanks’-5 (Ca/Mg-free HBSS plus 5% FBS) with 1 mM EDTA and 1 mM DTT (Sigma-Aldrich). The cell suspensions were passed through a strainer, then over glass wool columns, and subsequently pelleted by centrifugation at 1200 rpm. The cell pellets were resuspended in 44% Percoll (Pharmacia), underlain with 70% Percoll, and centrifuged at 2500 rpm for 25 min. IEL from the 44/70% interface were collected, washed, and resuspended in complete RPMI 1640 medium supplemented with 10% FBS.

Cell sorting of IEL populations

IEL suspensions were stained with FITC-labeled anti-CD4, PE-labeled anti-CD8, PE-Cy5-labeled anti-TCR, and allophycocyanin-labeled anti-CD8 to sort TCR⁺CD8⁺CD8⁺CD4^– (TCR⁺CD8 IEL) or TCR⁺CD8⁺CD8^–CD4^– (TCR⁺CD8 IEL) cells. To isolate TCR IEL in gene array experiment 1, cells were stained with PE-labeled anti-TCR and PE-Cy5-labeled anti-TCR to sort TCR^+– IEL. In gene array experiment 2 and all other experiments, TCR⁺TCR^–CD8⁺CD8^–CD4^– IEL were sorted using FITC-labeled anti-CD4 and anti-CD8, PE-labeled anti-TCR, PE-Cy5-labeled anti-TCR, and allophycocyanin-labeled anti-CD8. All Abs were purchased from BD Pharmingen. IEL were sorted using a FACSVantage cell sorter at La Jolla Institute for Allergy and Immunology. The purity of sorted IEL subsets was analyzed postsorting and was always >97%.

Microarray analysis

RNA was generated (RNeasy; Qiagen) on two occasions from sorted cells from a pool of 20 C57BL/6 mice (8–12 wk old) following cell sorting, and the quality was assessed using the Agilent Bioanalyzer 2100. Only RNA displaying minimal degradation and clear 18S/28S ribosomal bands was included. Preparation of the biotinylated RNA probes was done as described (Affymetrix GeneChip Expression Analysis Manual), except Thermoscript reverse transcriptase (Invitrogen Life Technologies) was used for first-strand cDNA synthesis (48°C at 10 min, 52°C at 40 min). In vitro transcription was done with ribonucleotide triphosphates and with 2500 U/L T7 RNA polymerase (Epicentre Technologies), according to the manufacturer’s protocol. Biotinylated RNA was generated with BioArray High Yield reagents (Enzo Diagnostics), according to the manufacturer’s protocol. Samples were hybridized to Affymetrix mouse U74Av2 chips and read at the Howard Hughes Medical Institute facility at the University of Texas Southwestern Medical Center (experiment 1) or at the University of California, Irvine (experiment 2). The data were processed and normalized using the robust multi-array average algorithm from the BioConductor software. These text files were produced by the Affymetrix software (MAS 5.0) using global scaling to a target signal of 500. Comparative analysis between expression profiles for each sample was also conducted using GeneSpring software version 6.0 (Silicon Genetics). In these analyses, the cross gene error model for deviation from 1.0 was active. To enable chip-to-chip comparisons, gene expression data were normalized to the 50th percentile of all values on that chip. To assess fold change, the data for a given gene were normalized to the median expression level of that gene across all samples. The data sets were assigned to the three groups (corresponding to TCR⁺CD8 IEL), (TCR⁺CD8 IEL), and (TCR⁺ IEL), and the expression profiles of the three groups were compared using Student’s t tests and multiple testing corrections to identify genes that were differentially expressed. There were two separate RNA preparations, and each was used to probe one set of chips. Hence, the experimental results were interpreted as the average of two biological replicates of material from 20 pooled mice in each case. A difference of 2-fold and a Student’s t test p < 0.05 was applied to select up-regulated genes. All gene array data files described in this manuscript have been deposited in the Gene Expression Omnibus database (www.ncbi.nlm.nih.gov/geo/) under accession number GSE5355 and are freely available upon acceptance of this manuscript.

Real-time PCR

Total RNA was isolated using the Qiagen RNeasy Mini Kit, according to the manufacturer’s protocol with on-column DNase digestion using the RNase-Free DNase set. cDNA was generated using the Superscript First-Strand Synthesis System for RT-PCR and random hexamer primers (Invitrogen Life Technologies), according to the manufacturer’s protocol. cDNA was used as a template for quantitative real-time PCR using SYBR Green Master Mix (Applied Biosystems), and gene-specific primers are listed below. PCR and analysis was performed using a 5700 GeneAmp Sequence Detector (ABI Prism). Gene expression was calculated relative to GAPDH: Gapdh forward, TGGCAAAGTGGAGATTGTTGCC, and reverse, AAGATGGTGATGGGCTTCCCG; TGF-₁ forward, ACCATGCCAACTTCTGTCTG, and reverse, CGGGTTGTGTTGGTTGTAGA; TGF-₂ forward, TGGTGGAAGCTAGGAGAAGC, and reverse, CCTTAACCCCTGTGGAACAA; TGF-₃ forward, GGAAATGGGTCCACGAACCTA, and reverse, TCCAAGCACCGTGCTATGG; IL-10 forward, CCCTTTGCTATGGTGTCCTT, and reverse TGGTTTCTCTTCCCAAG ACC; Fgl2 forward, TGGGAACTGTGGGCTCTATT, and reverse, CGGACACCTTTGTATTTCTGG; LAT forward, AGACGAAGGAGAAGAGGAAGG, and reverse, ACCCCAGCAAGTCCAGTTT; LAB/NTAL forward, CCCTCACCCTCAGCCTTAC, and reverse, AGCAGCAATAATCCCGACA; Nkrp1a forward, CCCTGATTGGGATGAGTGT, and reverse, TGTGAAAGCCAGTCTTGTG; Dap12 forward, CTGGTGTACTGGCTGGGATT, and reverse, TGCCTCTGTGTGTTGAGGT; 2B4 forward, CTGTCCTGTGGTGATGTTGG, and reverse, TCTGTGTGAGCCCTGTTCTG; and CD94 forward, TCAACACCTTCTCCAACCA, and reverse, CTGATGCCCAAACCCACTT.

Flow cytometric analysis of lymphocytes

Isolated IEL were resuspended in PBS staining buffer containing 2% BSA and 0.02% NaN₃. After preincubation for 15 min at 4°C with the blocking 2.4G2 anti-FcRIII/II mAb, the cells were stained at 4°C for 30 min with labeled mAb. Samples were then washed twice in PBS-staining buffer. The samples were immediately analyzed, or they were fixed in PBS containing 1% paraformaldehyde and 0.02% NaN₃ and stored at 4°C. mAbs used for analysis included FITC-labeled anti-Ly49A (clone A1), anti-Ly49C/I (clone 5E6), anti-Ly49I (clone YLI-90), anti-Ly49D (clone 4E5), anti-Ly49G2 (clone 4D11), and anti-rat IgG2a (clone RG7/1.30), PE-labeled anti-TCR (clone H57), anti-TCR (GL3), and anti-CD223 (lymphocyte activation gene-3 (LAG-3); clone C9B7W), PerCP-labeled anti-CD8 (clone 53-6.72) (all from BD Pharmingen), anti-Ly49F (clone HBF-719; Southern Biotechnology Associates), anti-Ly49E/C (clone 4D12) (34), and allophycocyanin-labeled anti-CD8 (clone CT-CD8; eBioscience). Specific Ly49E staining was confirmed by the lack of reactivity to Ly49C/I mAb (clone 5E6). V14i NKT cells were detected using CD1d tetramers loaded with the glycolipid -galactosylceramide (35). Flow cytometric analysis was performed on a BD Biosciences FACSCalibur flow cytometer at the La Jolla Institute for Allergy and Immunology. For intracellular detection of linker for activation of T cells (LAT), cells were first stained for surface Ags, then fixed and permeabilized using the Cytofix/Cytoperm Kit, according to the manufacturer’s instructions (BD Pharmingen), before being stained with polyclonal rabbit anti-LAT IgG (Upstate Cell Signaling Solutions), followed by allophycocyanin-labeled anti-rabbit IgG (Jackson ImmunoResearch Laboratories).

Western blot analysis/Immunoprecipitation

A total of 1 x 10⁷ cells (total IEL, splenic CD8⁺, or splenic CD19⁺) or 0.2 x 10⁷ cells from sorted IEL subsets was lysed in 500 µl of ice-cold radioimmunoprecipitation assay lysis buffer (1% Triton X-100, 0.5% sodium deoxycholic acid, 0.1% SDS, 25 mM Tris-Cl (pH 7.6), 150 mM NaCl, and 5 mM EDTA). B lymphocytes were selected from the spleen using CD19-coupled magnetic beads (Miltenyi Biotec), and CD8⁺ splenic T cells were positively selected using CD8-coupled magnetic beads. Lysates from these cells were subjected to immunoprecipitation with protein A beads cross-linked with antisera recognizing linker for activation of B cells (LAB), which is also known as non-T cell activation linker (NTAL) (36). For Western blotting, samples were separated by SDS-PAGE and transferred to nitrocellulose membranes. After incubation with polyclonal rabbit anti-mouse LAB/NTAL Ab, nitrocellulose membranes were washed three times and probed with anti-rabbit IgG conjugated to HRP (Kirkegaard & Perry Laboratories). Membranes were then visualized with the ECL Detection Reagent (Amersham Biosciences).

	Results

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Microarray analysis of IEL subsets

To analyze the differential gene expression profiles of mouse small intestine IEL, total RNA was extracted from sorted IEL subpopulations. The isolation was repeated, and independent hybridizations to Affymetrix Mu74Av2 chips were performed for each isolated subpopulation. The validity of this comparison was indicated by the detection of genes known to encode proteins differentially expressed by one of the IEL subpopulations. For example, when compared with the TCR⁺CD8 IEL, the TCR⁺CD8 IEL were enriched not only for the expression of CD8 mRNA but also for the expression of CD2, CD5, CD28, and CD90 (Fig. 1 and Table I), all of which were known previously to be expressed in greater amounts by TCR⁺CD8 IEL (37). Similarly, consistent with previous data (38), FcRI-chain mRNA was enriched in TCR⁺CD8 IEL, and the TCR and TCR transcripts were enriched in the populations expressing the respective TCR proteins.

View larger version (39K): [in this window] [in a new window]   FIGURE 1. Gene expression profiling was performed on sorted IEL subsets. Scatter plots displaying values of normalized signal intensities for selected genes are shown as a distance from the diagonal. Spots indicate the average values for each gene from two independent chip (Affymetrix Mu74Av2) hybridizations. The center diagonal indicates no change and the outer lines indicate a 2-fold expression difference. Green = genes encoding NK receptors known or believed to be inhibitory; red = genes encoding NK receptors that under some circumstances can be activating; white = validation controls, genes encoding proteins known to be differentially expressed in IEL subpopulations; blue = genes encoding molecules involved in proximal events in surface receptor signaling; and yellow = genes encoding immunoregulatory molecules.

TCR⁺CD8 IEL are enriched for the expression of NK receptors

Many of the genes preferentially expressed in TCR⁺CD8 IEL were NK receptor genes or genes acting downstream of NK receptors (Fig. 1). These NK cell-related genes, and selected other genes related to immune function, are listed in Table I. To confirm the validity of the microarray analysis results, we performed quantitative real-time PCR for NKR-P1A, DNAX-activating protein of 12 kDa (DAP12), 2B4, and CD94. As shown in Fig. 2A, in comparison to TCR⁺CD8 IEL, TCR⁺CD8 IEL were enriched (>2.4-fold) for the expression of each of these transcripts based on the results from the microarray analysis. The data obtained from real-time PCR and the microarray analyses were consistent, although the magnitude of the difference in mRNA expression was greater with PCR. Transcripts for the receptor NKR-P1A and the adaptor DAP12 showed 19- and 245-fold increases by real-time PCR, respectively, when TCR⁺CD8 IEL were compared with TCR⁺CD8 IEL (/). Although mRNA-encoding NKR-P1A (Klrb 1a in Table I) was highly enriched in TCR⁺CD8 IEL, mRNA encoding NKR-P1C (Klrb 1c) was not highly enriched, and in agreement with a previous report (39), very few mouse IEL of any phenotype express NK1.1 (data not shown), which is an allelic form of the klrb 1c gene. CD94 was 6.4-fold enriched and the 2B4 receptor (CD244) was increased 2.7-fold. As in the microarray analysis, when TCR⁺CD8 IEL were compared with TCR⁺ IEL (/) using real-time PCR, mRNA for the same set of NK-related proteins was enriched in the TCR⁺CD8, albeit to a lesser extent than when compared with the TCR⁺CD8 IEL.

View larger version (21K): [in this window] [in a new window]   FIGURE 2. TCR+CD8 IEL are enriched for the expression of NK receptor mRNAs. A, RNA was isolated from sorted IEL subpopulations, and expression of various NK cell-related genes was analyzed by quantitative real-time PCR comparing TCR+CD8 IEL () either to TCR IEL () or to TCR+CD8 IEL (). Values are expressed relative to GAPDH as an internal standard. Representative data are shown for one of three independent experiments. Numbers in parentheses indicate the fold change observed in the microarray analysis. B, Gene expression of Ly49 NK receptors determined by microarray analysis comparing TCR+CD8 IEL either to TCR IEL (/) or to TCR+CD8 IEL (/).

View larger version (18K): [in this window] [in a new window]   FIGURE 3. TCR+CD8 IEL have increased expression of inhibitory Ly49 family NK receptors. A, IEL were stained with Abs against Ly49A, E/C, F, or G2 or a mixture of these Abs. Ly49 expression by TCR+CD8 IEL (), TCR+CD8 IEL (), and TCR+CD8 IEL () is shown. Error bars indicate SD. B, TCR+CD8 IEL (), TCR+CD8 IEL (), and TCR+CD8 IEL () (upper panels) or splenic V14i NKT cells, CD8+ T cells, and NK cells (lower panels) were stained with a mixture of anti-Ly49 Abs. Percentage of positive cells are indicated above the gate in each histogram. C, Ly49E (upper panels) and Ly49F (lower panels) protein expression was assessed on TCR+CD8 IEL (), TCR+CD8 IEL (), as well as splenic NK cells gated as NK1.1+TCR– cells. The average of three independent experiments is shown, expressed as the percentage of positive cells.

A significant fraction of TCR⁺CD8 IEL also expressed Ly49F (17.9% positive cells; Fig. 3C, lower panel), compared with only 2.6% positive TCR⁺ IEL and 6.3% positive splenic NK cells. The populations of Ly49E⁺ and Ly49F⁺TCR⁺CD8 IEL were mostly nonoverlapping, with the percentage of Ly49E/F double-expressing cells approximately equal to the product of the frequency of the cells expressing either molecule (data not shown). Therefore, Ly49E and/or Ly49F expression accounted for the majority of the positively staining TCR⁺CD8 IEL using the anti-Ly49 mixture of Abs. In conclusion, TCR⁺CD8 IEL express a unique subset of Ly49 receptors compared with either NK cells or V14i NKT cells.

Many of the NK receptors showing increased expression on TCR⁺CD8 IEL are inhibitory. In the Ly49 family, multiple inhibitory receptors are expressed, but the activating receptor Ly49D is not expressed at the protein level (data not shown). Some of the NK-related transcripts enriched in TCR⁺CD8 IEL can have an activating or inhibiting function (41, 42). CD94 can form heterodimers that are either activating or inhibiting (43), and in NK cells, 2B4 is inhibitory when coupled to the Src homology 2 domain-containing adaptor Ewing’s sarcoma/FLI1-activated transcript 2 (EAT2) (SH2D1b) and activating when coupled to the homologous adaptor signaling lymphocytic activation molecule-associated protein (SAP) (SH2D1a) (44). Interestingly, although all IEL subsets express 2B4 protein (data not shown), we found increased EAT2 and decreased SAP mRNA in TCR⁺CD8 IEL (Fig. 1 and Table I). DAP12 is well known as an activating receptor, but recent work shows it can also inhibit signals from TLR and FcR (45). Unambiguous exceptions to this pattern of preferential expression of inhibitory NK receptors and adaptors is the increased amount of mRNA for the activating receptors Ly49H (41) and NKR-P1A, which may be activating because of the arginine in its cytoplasmic domain. NKG2D mRNA was also increased, but the amount of NKG2D mRNA is very low (Fig. 1).

IEL expression of Wbscr5/NTAL/LAB

The TCR-signaling apparatus of CD8 IEL is known to be different from that in conventional T lymphocytes because IEL use the FcRI -chain in addition to TCR (38). Consistent with this, we found an increased amount of mRNA encoding the FcRI -chain in CD8 IEL (Table I). We also found an increased amount of RNA encoding the Lyn src family kinase (Fig. 1 and Table I) in CD8 IEL, and Lyn can act downstream of the FcRI -chain (46). One of the more highly differentially expressed genes involving signal transduction comparing TCR⁺CD8 and TCR⁺CD8 IEL was Wbscr5 now known as Lat2. This gene encodes the mouse homolog of the widely expressed human gene and is termed NTAL or LAB (47, 48). NTAL/LAB is a structural homolog of LAT expressed by B, NK, and mast cells (49, 50), and it becomes rapidly tyrosine phosphorylated upon activation. Interestingly, unlike LAT, NTAL/LAB does not associate with phospholipase C1 (51). NTAL/LAB binds to some of the same adaptors as LAT, but therefore, it may not cause a complete activation event. Consistent with this, recent evidence indicates that NTAL/LAB is also expressed by activated T cells and that it is involved in down regulating the T cell immune response and mice with a T cell-specific conditional deletion of Lat2 developed an autoimmune syndrome (52).

As shown in Fig. 4A, TCR⁺CD8 and TCR⁺CD8 IEL express very low levels of LAT mRNA, whereas splenic B cells are LAT mRNA negative. TCR⁺CD8 IEL express approximately half as much LAT mRNA as splenic T cells. An opposite pattern of expression was detected for NTAL/LAB. TCR⁺CD8 and TCR⁺CD8 IEL expressed amounts of NTAL/LAB mRNA nearly identical to the splenic B cell-positive control, whereas TCR⁺CD8 IEL expressed much lower amounts and splenic CD8 T cells were NTAL/LAB negative. To confirm these data at the protein level, we performed intracellular staining and flow cytometry. As shown in Fig. 4B, LAT protein expression can be detected in splenic T lymphocytes but not in B cells. Furthermore, TCR⁺CD8 IEL expressed significantly lower amounts of LAT protein than did TCR⁺CD8 IEL. NTAL/LAB Abs did not function well for intracellular staining (data not shown), and therefore, we performed immunoprecipitation and Western blotting with a different antiserum to detect expression. Cells in total IEL preparations expressed high levels of NTAL/LAB, similar to what was detected with equal numbers of splenic B cells, whereas splenic T cells are virtually devoid of NTAL/LAB protein (Fig. 4C). To further clarify which IEL subsets express NTAL/LAB, TCR⁺CD8, TCR⁺CD8, and TCR⁺CD8 IEL were sorted by flow cytometry, and NTAL/LAB was immunoprecipitated with a polyclonal antiserum and subjected to Western blot analysis. As shown in Fig. 4C, lower panel, TCR⁺CD8 and TCR⁺CD8 IEL expressed abundant NTAL/LAB protein.

View larger version (26K): [in this window] [in a new window]   FIGURE 4. IEL express NTAL/LAB. A, RNA was extracted from sorted IEL populations, splenic CD8+ T cells, or splenic CD19+ B cells, and quantitative real-time PCR was performed for LAT (upper left panel) and NTAL/LAB (upper right panel) gene expression. Values are expressed relative to GAPDH as an internal standard. Representative data are shown from one of three independent experiments. B, Intracellular LAT expression analyzed by flow cytometry. Left, LAT expression shown for splenic B cells (black histogram) and T cells (gray histogram) was determined by pregating on CD19+ or CD8+ cells, respectively. Right, LAT expression on IEL subsets was determined by pregating on TCR+CD8+– (; black histogram) or TCR+CD8++ (; gray histogram). Intracellular secondary Ab staining alone is shown as a dashed line in each panel. C, NTAL/LAB protein expression was analyzed in splenic B and T cells and total IEL (upper panel) or sorted IEL subsets (lower panel) by immunoprecipitation and Western blotting as described in Materials and Methods.

It is not known how TCR⁺CD8 IEL prevent colitis in the T cell transfer model, but we found that Foxp3 mRNA expression was very low in the IEL subsets we tested in the microarray and by real-time PCR (data not shown). We also examined the ability of the three major subsets of mouse small intestine IEL to express constitutively mRNA for immunosuppressive cytokines and molecules that could be involved in immune regulation. Based on the microarray results, TCR⁺CD8 IEL were enriched for TGF-₃ mRNA, to a lesser extent, for TGF-₁ and not at all for IL-10 mRNA when analyzed by real-time PCR ex vivo (Fig. 1 and Table I). They were also enriched for mRNA encoding fibrinogen-like protein-2 (fgl2), which has been reported to have immune suppressive activity through inhibition of DC maturation (53). Additionally, TCR⁺CD8 IEL preferentially expressed LAG-3, a CD4-related molecule that binds MHC class II with high affinity. CD25⁺ regulatory T cells express increased amounts of LAG-3, and there is evidence indicating it contributes to their suppressive activity (54, 55).

These microarray results were confirmed by real-time PCR analysis. As shown in Fig. 5A, all IEL subpopulations expressed detectable levels of TGF-₁ and TGF-₃ mRNA when analyzed by PCR ex vivo, whereas TGF-₂ was not detected (data not shown). Interestingly, consistent with the microarray data, TCR⁺CD8 IEL expressed a greatly increased amount of TGF-₃ mRNA and slightly increased mRNA for TGF-₁. In contrast to IEL, splenic CD8⁺ T cells did not express detectable TGF-₃ mRNA, yet they did express an amount of TGF-₁ mRNA comparable to the IEL subpopulations. fgl2 mRNA was highly expressed in TCR⁺CD8 IEL as determined by real-time PCR. Furthermore, flow cytometry analysis verified that CD8 IEL expressed significantly higher levels of LAG-3 protein when compared with splenic CD8⁺ T cells or CD8 IEL (Fig. 5B). Therefore, it is possible that TCR⁺CD8 IEL could play a role in immune regulation in the intestine, either through a unique pattern of TGF- isoform expression or perhaps through a novel mechanism involving fgl2 and/or LAG-3.

View larger version (18K): [in this window] [in a new window]   FIGURE 5. IEL express immunoregulatory molecules. A, RNA was extracted from sorted IEL populations or splenic CD8 T cells, and quantitative real-time PCR was performed for TGF-1, TGF-3, IL-10, and fgl2 gene expression. Values are expressed relative to GAPDH as an internal standard. Representative data are shown from one of more than three separate experiments. B, LAG-3 expression. Left, LAG-3 expression shown for gated splenic CD8+ T cells (gray line) and CD8+ IEL (black line). Right, LAG-3 expression on IEL subsets was determined by pregating on CD8++ (; gray line) or CD8+– cells (; black line). The isotype control Ab staining is shown as a dashed line in each panel.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

Perhaps the most striking finding to emerge from this study is that TCR⁺CD8 IEL express a number of genes that should function to down-regulate their reactivity. This is consistent with the results from other studies indicating that TCR⁺CD8 IEL are self-reactive T cells that have been positively selected by self-agonists in the thymus (21, 22, 23, 24, 26). It is notable that a relatively high frequency of TCR⁺CD8 IEL expresses NK receptors, predominantly inhibitory receptors. Although very few of these lymphocytes express the activating receptor NK1.1, 40% express one or more Ly49 family members, in addition to their expression of other NK receptors such as 2B4. Because of this, TCR⁺CD8 IEL may be considered a unique type of NKT cell, defined broadly in this context as lymphocytes that coexpress NK receptors and a TCR, without specific reference to the CD1d-dependent V14i NKT cells. Consistent with our results, a previous report (39) also found that TCR⁺CD8 IEL were distinguished from both TCR⁺CD8 and TCR⁺ IEL by an increased frequency of cells expressing several NK receptors. In that study, the TCR⁺CD8 IEL also exhibited NK-type cytotoxicity following cytokine activation. In this study, we demonstrate for the first time that the repertoire of Ly49 receptor expression by TCR⁺CD8 IEL is different from several other populations, including NK cells and V14i NKT cells.

An outstanding feature of the pattern of NK receptor expression is the relatively high frequency of cells expressing two orphan members of the Ly49 family, Ly49E and Ly49F. Although neither of these receptors is capable of reacting to polymorphic MHC class I molecules (57), nonclassical class I molecules expressed by intestinal epithelial cells are potential ligands. Using tetramers of the thymus leukemia Ag, which is highly expressed by small intestine epithelium (58), or CD1d tetramers, we could not stain transfectants expressing Ly49E or Ly49F (data not shown). It remains possible, however, that some other nonclassical class I molecule, or another type of molecule such as a lectin (59) or E-cadherin (60), is a ligand for Ly49E and Ly49F. Although the NK receptor KLRG1 binds E-cadherin, which is highly expressed by intestinal epithelial cells, consistent with the microarray data, KRLG1 protein is not highly expressed by IEL (61).

Consistent with the possibility of controlled self-reactivity, the Ly49 family receptors expressed by TCR⁺CD8 IEL, such as Ly49A and Ly49G2, are predominantly inhibitory. Ly49E and Ly49F have ITIM as part of their cytoplasmic tails, and expression of the activating receptor Ly49D could not be detected. We speculate that the expression of inhibitory NK receptors may help to keep TCR⁺CD8 IEL self-reactivity in check. The inhibitory functions of Ly49E and Ly49F have not been unequivocally demonstrated, however, and in addition to its ITIM, Ly49E has a charged amino acid in its transmembrane domain, a characteristic of activating receptors (62). Furthermore, TCR⁺CD8 IEL also express an increased amount of transcripts for the known activating receptor Ly49H and for NKR-P1A. Therefore, as with NK cells and V14i NKT cells that express NKR-P1C, TCR⁺CD8 IEL express a mixture of inhibitory and activating NK receptors, although inhibitory receptors are more prevalent.

The pattern of expression of molecules involved in signal transduction is concordant with the possibility that TCR⁺CD8 IEL are in a state of partially inhibited reactivity. Prominent is the expression of the adaptor NTAL/LAB, which does not activate phospholipase C1 (51), although its LAT homolog does. Interestingly, mast cells are hyperresponsive to FcRI stimulation in mice deficient for NTAL/LAB, indicative of an inhibitory function (36, 63), and NTAL/LAB also plays a role in inhibiting T cell reactivity and autoimmunity (52). Additionally, TCR⁺CD8 IEL preferentially express mRNA encoding the inhibitory EAT2 adaptor for 2B4 in place of SAP (44), and previous studies have shown that TCR⁺CD8 IEL are relatively anergic following TCR engagement (64, 65, 66, 67). A similar mechanism of agonist-mediated selection and modulation of self reactivity by inhibitory receptors has been proposed for V14i NKT cells, which likewise predominantly express inhibitory Ly49 NK receptors and do not express Ly49D (68). In fact, over expression of Ly49D in transgenic mice led to the absence of V14i NKT cells (69), presumably due to self-tolerance mechanisms.

In TCR-transgenic mice, the addition of peptide self-agonists to fetal thymus organ cultures or reaggregation cultures led to the differentiation of CD8 single-positive T lymphocytes (28, 70, 71). The CD8-expressing cells in these organ cultures are putative precursors of TCR⁺CD8 IEL, and in fact, these cells express ₇ integrins associated with mucosal T cells and they populate the intestine epithelium upon cell transfer (72). An analysis of the gene expression profile of CD8 thymocytes generated in reaggregation cultures indicated that they had induced expression of a number of NK cell-related genes, but in marked contrast to our results, the increased NK cell-related genes were specifically those associated with activation. Therefore, it is possible that the acquisition of inhibitory receptor expression is a postthymic event in the ontogeny of TCR⁺CD8 IEL, similar to the increased expression of CD8 and the down-regulation of CD5 expression by these cells (26). Interestingly, NK receptor expression is also a late or postthymic event in the ontogeny of V14i NKT cells (73, 74).

A second striking feature of our results is the expression by TCR⁺CD8 IEL of genes that could be involved in immune regulation. TCR⁺CD8 IEL from the small intestine were reported to prevent colitis induced by the transfer of CD4⁺CD45RB^high T cells to immune-deficient mice (31), but the means by which TCR⁺CD8 IEL might prevent colitis and regulate normal mucosal immune responses have not been elucidated. In another investigation (32), these IEL constitutively synthesized some TGF-₁ mRNA, but this was not augmented after Ag-specific stimulation. We found that the constitutive amount of TGF-₁ mRNA TCR⁺CD8 IEL was only slightly greater than the amount in other IEL subpopulations or CD8⁺ spleen cells, but we did find increased expression of TGF-₃ mRNA. Although the different TGF- isoforms bind to the same receptor complex, they are known to have different functional properties. TGF-₃ enhanced intestinal epithelial healing in an in vitro model system (75), and interestingly, unlike other TGF- isoforms, TGF-₃ did not promote expression of tissue inhibitor of metalloproteinase-1 (76). Therefore, it should not contribute as much as other TGF- isoforms to fibrosis. Consistent with this, in a rat model of wound healing in the skin, TGF-₃ uniquely prevented scarring, whereas the other isoforms had an opposite effect (77). Collectively, these data suggest that TGF-₃ could have beneficial effects on the intestine epithelium distinct from the other TGF- isoforms, although the important targets of this cytokine could include other cell types as well.

We also found increased levels of mRNA encoding the novel immune coagulant fgl2/fibroleukin expressed by TCR⁺CD8 IEL. Elevated levels also have been found in CD25⁺ T regulatory cells (78), and it has been reported that fgl2 can inhibit T cell proliferation and DC maturation (53), although this was not observed in another study (79). Finally, we found increased levels of the MHC class II-binding LAG-3 protein expressed in all CD8 single-positive IEL, including the TCR⁺CD8 cells and TCR⁺ IEL. This protein has been reported to be important for the normal function of CD25⁺ regulatory T cells (55). Not only is MHC class II expressed by DC, but it can be expressed by intestinal epithelial cells, especially under inflammatory conditions (80), providing targets for the LAG-3 protein expressed by IEL.

In summary, our analysis has uncovered novel features of the gene expression profile of TCR⁺CD8 IEL and of the total population of CD8 single-positive IEL, which includes the TCR⁺ IEL. Based on the unique pattern of expression of NK receptors and signaling molecules, it appears that TCR⁺CD8 IEL are in a state of partially suppressed or controlled reactivity particular to this subset, perhaps reflecting the self-reactivity of these cells. Additionally, our results give some insight into potential means by which TCR⁺CD8 IEL might prevent inflammation. Although LAG-3 expression is not likely to be sufficient, given the expression of this protein by both TCR⁺ and TCR⁺CD8 IEL, coexpression of other molecules such as TGF-₃ and fgl2 by TCR⁺CD8 IEL may allow these cells to be the most potent regulators in the epithelium.

	Acknowledgments

 
We thank Samuel Connell for assistance with cell sorting; Chris Lena for purifying 4D12 Ab; Dr. Stephane Sidobre and Lise Sidobre for CD1d tetramers; and Drs. Gisen Kim, Denise Gangadharan, and Marcos Steinberg for helpful discussions.

	Disclosures

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
The authors have no financial conflict of interest.

	Footnotes

 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

¹ This work is supported by National Institutes of Health Grants P01 DK46763 and R01 AI 65016 (to M.K). T.L.D. is a recipient of a Jeane B. Kempner Scholarship from the University of Texas Medical Branch at Galveston.

² This is Publication Number 781 from the La Jolla Institute for Allergy and Immunology.

³ Address correspondence and reprint requests to Dr. Mitchell Kronenberg, Division of Developmental Immunology, La Jolla Institute for Allergy and Immunology, 9420 Athena Circle, La Jolla, CA 92037; E-mail address: mitch{at}liai.org or Dr. Timothy L. Denning, Emory Vaccine Center, Atlanta, GA 30329; E-mail address: tdennin{at}emory.edu

⁴ Abbreviations used in this paper: IEL, intestine intraepithelial lymphocyte; DAP12, DNAX-activating protein of 12 kDa; DC, dendritic cell; EAT2, Ewing’s sarcoma/FLI1-activated transcript 2; fgl2, fibrinogen-like protein-2; LAB, linker for activation of B cell; LAG-3, lymphocyte activation gene-3; LAT, linker for activation of T cell; NTAL, non-T cell activation linker; SAP, signaling lymphocytic activation molecule-associated protein; V14i, invariant V14.

Received for publication October 5, 2006. Accepted for publication January 8, 2007.

	References

Top