The JI PBL Intereron Source
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Data Supplement
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chtanova, T.
Right arrow Articles by Mackay, C. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chtanova, T.
Right arrow Articles by Mackay, C. R.
The Journal of Immunology, 2005, 175: 7837-7847.
Copyright © 2005 by The American Association of Immunologists

Identification of T Cell-Restricted Genes, and Signatures for Different T Cell Responses, Using a Comprehensive Collection of Microarray Datasets1

Tatyana Chtanova*,¶, Rebecca Newton*, Sue M. Liu*,¶, Lilach Weininger*, Timothy R. Young*, Diego G. Silva{dagger}, Francesco Bertoni{ddagger}, Andrea Rinaldi§, Stephane Chappaz§, Federica Sallusto§, Michael S. Rolph*,¶ and Charles R. Mackay2,*,¶

* Garvan Institute of Medical Research, Darlinghurst, Australia; {dagger} Australian National University Medical School, Canberra, Australia; {ddagger} Experimental Oncology, Oncology Institute of Southern Switzerland, Bellinzona, Switzerland; § Institute for Research in Biomedicine, Bellinzona, Switzerland; and Cooperative Research Center for Asthma, University of Sydney, Camperdown, Australia


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
We used a comprehensive collection of Affymetrix microarray datasets to ascertain which genes or molecules distinguish the known major subsets of human T cells. Our strategy allowed us to identify the genes expressed in most T cell subsets: TCR {alpha}{beta}+ and {gamma}{delta}+, three effector subsets (Th1, Th2, and T follicular helper cells), T central memory, T effector memory, activated T cells, and others. Our genechip dataset also allowed for identification of genes preferentially or exclusively expressed by T cells, compared with numerous non-T cell leukocyte subsets profiled. Cross-comparisons between microarray datasets revealed important features of certain subsets. For instance, blood {gamma}{delta} T cells expressed no unique gene transcripts, but did differ from {alpha}{beta} T cells in numerous genes that were down-regulated. Hierarchical clustering of all the genes differentially expressed between T cell subsets enabled the identification of precise signatures. Moreover, the different T cell subsets could be distinguished at the level of gene expression by a smaller subset of predictor genes, most of which have not previously been associated directly with any of the individual subsets. T cell activation had the greatest influence on gene regulation, whereas central and effector memory T cells displayed surprisingly similar gene expression profiles. Knowledge of the patterns of gene expression that underlie fundamental T cell activities, such as activation, various effector functions, and immunological memory, provide the basis for a better understanding of T cells and their role in immune defense.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Specialized subsets of effector T cells participate in different types of immune responses. Th1 cells produce IFN-{gamma} and protect against viral pathogens, whereas Th2 cells produce cytokines, such as IL-4 and IL-5, and protect against large extracellular parasites (1). The recently identified follicular B helper T cell (TFH)3 subset provides help to B cells in B cell follicles (2, 3). All these effector T cell subsets interact with other leukocyte types, including B cells, dendritic cells, and macrophages, as part of a coordinated response to antigenic challenge. In addition, other subsets of T cells provide immune protection at a more primitive level; {gamma}{delta} T cells and NK T cells contribute to early cellular immune responses, particularly innate immunity.

The specialized functions of leukocyte subsets are reflected in the differing patterns of molecules and genes they express. These molecules often underlie the unique function of the subsets they mark. For instance, CD3 defines T cells because it is an essential component of the TCR complex; CD4 is associated with MHC class II recognition, mostly for Th responses, whereas CD8 is associated with MHC class I recognition and cytotoxic responses, and B cells are defined by their Ig expression or production. Immunological research during much of the 1980–1990 era was directed at identifying and characterizing leukocyte markers, first through mAb production, and then through gene sequencing. This led to the discovery of most of the CD molecules that are used today to mark leukocyte subsets and define subset function.

Chemokine receptors and other molecules involved in cell migration have proved useful for defining leukocyte subsets, particularly stages of T cell differentiation and function. CXCR5 marks TFH cells and facilitates their migration to B cell follicles, where they provide help to B cells (2, 3). Th1 cells preferentially express CCR5 and CXCR3, whereas Th2 cells preferentially express CCR3, CCR4, and CRTh2 (4, 5, 6). Most human Th1 and Th2 clones maintain flexibility of cytokine gene expression and, when stimulated under opposite conditions, differentiate to Th0 capable of producing both IFN-{gamma} and IL-4 (7). CCR7 facilitates naive T cell homing to secondary lymphoid organs (8), but also distinguishes central memory (TCM) from effector memory (TEM) T cells (9, 10).

More recently, gene microarrays have proven to be a powerful tool for identifying novel gene expression patterns for various subsets of leukocytes (11). We and others have successfully used microarray technology to create gene expression profiles of effector and memory T cell subsets as well as other leukocyte subsets (12, 13, 14, 15). In this study we present a comprehensive analysis of most of the major subsets of human T cells and identify at the level of gene expression the distinguishing features associated with T cell differentiation to effector subsets, T cell activation, and memory T cell development. In addition, we identify T cell subset-specific gene expression signatures and describe a distinctive transcriptional profile for {gamma}{delta} T cells. Finally, the use of a comprehensive collection of genechip datasets, representing gene expression profiles for all the major human leukocyte subsets, has allowed us to identify numerous genes that are T cell restricted and presumably of relevance for T cell-specific functions.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
Generation of T cell and leukocyte subsets

The microarray experiments used in this study are summarized in Table I. Detailed descriptions of each microarray experiment are provided at <http://linkage.garvan.unsw.edu.au/public/microarrays/>. Generation of Th1 and Th2, TCM, TEM, and TFH cells has been described previously (15). Resting and activated Th0 and Th2 cells derived from peripheral blood naive T cells were generated as follows. Adult CD4+ naive T cells were isolated from PBMC using a combination of MACS and FACS sorting. Briefly, CD4+ T cells were positively selected with anti-CD4-coated MACS microbeads (Miltenyi Biotec). Cells were then stained with anti-CD45RA Abs, and naive CD45RA+ T cells were sorted on a FACSVantage (BD Biosciences). The cells were cultured in RPMI 1640 medium supplemented with 1% Glutamax, 1% sodium pyruvate, 1% nonessential amino acids, 50 µg/ml streptomycin/penicillin (Invitrogen Life Technologies), 5 x 10–5 M 2-ME, and 5% human serum. Single naive T cells were distributed in 96-well plates by FACSVantage sorting and were stimulated with 105 allogeneic irradiated (40 Gy) PBMC and 1 µg/ml PHA in IL-2-containing medium in the presence of 2 ng/ml rIL-4 and 1 µg/ml neutralizing anti-IL-12 Abs (R&D Systems). Clones were expanded in IL-2-containing medium. A cytokine production assay was performed after ~2 wk. After 8 wk, selected T cell clones were restimulated under the same Th2 conditions to generate Th2 clones or under opposite Th1 conditions (0.5 ng/ml rIL-12 and 0.5 µg/ml neutralizing anti-IL-4 Abs (R&D Systems)) to generate Th0 clones. Clones were expanded in IL-2-containing medium. A cytokine production assay was performed after 2 wk. RNA was extracted from resting and activated T cell clones after 4 wk using TRIzol (Invitrogen Life Technologies).


View this table:
[in this window]
[in a new window]
  		Table I. Microarray experiments used in this study

 
For cytokine production assays, T cells were stimulated with 2 x 10–7 M PMA and 1 µg/ml ionomycin (Sigma-Aldrich) for 4 h. Brefeldin A (10 µg/ml) was added for the last 2 h. Fluorochrome-labeled anti-IFN-{gamma} and anti-IL-4 Abs (BD Pharmingen) were used after fixation and permeabilization, performed using the Cytofix/Cytoperm kit (BD Pharmingen). T cells (107/condition) were left untreated or were activated for 4 h with 1 µg/ml anti-CD3 Ab (clone TR66) and 50 ng/ml phorbol dibutyrate (PdBu; Sigma-Aldrich).

Preparation of cRNA and genechip hybridizations

Total RNA was isolated from cells using the RNeasy Total RNA Isolation Kit (Qiagen) or TRIzol according to the manufacturer’s instructions. cRNA was prepared as previously described (16). Briefly, cDNA was specifically transcribed from 500 ng of mRNA using a poly-T nucleotide primer containing a T7 RNA polymerase promoter (Geneworks). Biotinylated, antisense target cRNA was subsequently synthesized by in vitro transcription using the BioArray High Yield RNA Transcript Labeling kit (Enzo Diagnostics). Fifteen micrograms of biotin-labeled target cRNA was then fragmented and used to prepare a hybridization mixture, which included probe array controls and blocking agents. Hybridization to U133A and B Affymetrix arrays was conducted for 16 h at 45°C and 60 rpm. After hybridization, washing and staining of the hybridized probe array were performed by an automated fluidics station, according to the manufacturer’s protocols. The stained probe array was scanned using the Agilent GeneArray Laser Scanner, and the resultant image was captured as a data image file.

Data analysis

Absolute levels of expression of genes were determined and scaled to 150 using algorithms in MicroArray Analysis suite 5.0 (MAS5) software (Affymetrix). The signal value represents the level of expression of a transcript. The signal log ratio is the change in expression level of a transcript expressed as the log2 ratio (a signal log2 ratio of 1 is equal to a fold change of 2). Annotations were extracted using the NetAffx analysis tool (<www.affymetrix.com>). All the microarray data presented in this study are publicly available from <http://linkage.garvan.unsw.edu.au/public/microarrays/>.

In-depth analyses and clustering of data were conducted using GeneSpring software (Silicon Genetics). After data transformation (to convert any negative value to 0.01), normalization was performed using a per chip 50th percentile normalization and per gene median normalization method. Genes that were consistently absent or below the noise level were excluded from analysis. To identify genes with statistically significant differences between T cell subsets or cell types, one-way ANOVA with a p value cutoff of 0.05 and the Benjamini and Hochberg false discovery rate as multiple testing correction were performed. The Student-Newman-Keuls post-hoc test was used to identify the specific groups in which significant differential expression occurred. Hierarchical clustering was performed on both genes and individual experiments, with Pearson correlation as a measure of similarity to group genes and samples with similar expression patterns. Data points were arranged in a hierarchy and were displayed in a phylogenetic tree of clusters of genes in a hierarchically ordered relationship. Branch lengths represent the degree of similarity between sets. Gene expression profiles that were similar across the experimental samples were clustered together. Principal component analysis using GeneSpring software was preformed using all genes that were expressed in at least two samples, to identify the components responsible for the greatest variability.

Data were imported as a Microsoft Excel file into Spotfire for graphical representation of gene expression patterns in TCM and TEM cells. Spotfire software was used to map genechip analysis results where fluorescence intensity detected on the genechips is represented by a color scale. Genes that showed a change of 2-fold or greater were considered differentially expressed.

Real-time PCR validation of gene expression

Total RNA was isolated from purified cells as described. RT was performed on 200 ng of total RNA per reaction using the Reverse-IT RTase Blend Kit (ABgene) according to the manufacturer’s instructions. After cDNA synthesis, semiquantitative real-time RT-PCR was performed with FastStart DNA Master SYBR Green I Reagent by LightCycler (Roche), as previously described (17). Primers were designed using Primer3 software (18). mRNA expression was normalized to GAPDH, and the fold change was calculated relative to Th1 gene expression.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
Comprehensive leukocyte subset profiling allows the identity of T cell-restricted genes

T cells differ from other leukocyte subsets in their phenotypic and functional properties, which is reflected by different gene expression patterns. We have produced a large dataset of genechip expression profiles for all the major leukocyte subsets, including various subsets of B cells, plasma cells, NK cells, eosinophils, neutrophils, basophils, dendritic cells, macrophages, mast cells, and all major effector and memory T cell subsets (Table I).

The first question we asked was whether this comprehensive dataset of genechips for various leukocyte types might allow us to identify T cell-specific transcripts. We applied one-way ANOVA to identify genes that most reliably discriminate between T cells and other leukocytes (Fig. 1A). We also used the publicly available SymAtlas database (<http://symatlas.gnf.org/SymAtlas/>) (19) to assess the expression of gene transcripts in many nonlymphoid tissues. We identified a large number of genes that were preferentially expressed by T cells compared with non-T cell leukocytes (>1200 probe sets on the Affymetrix U133 A+B arrays; p < 0.0001). The top ~100 genes most intimately associated with T cells are listed (Fig. 1A), and ~400 additional genes selectively expressed by T cells (p < 0.000001) are provided as supplementary information (supplementary Table I). As expected, some T cell-restricted genes were expressed outside of the hemopoietic lineage, whereas others were highly specific to T cells (Fig. 1A). Interestingly, we found that many genes were shared between T and NK cells, indicating that the origin and functional properties of these two leukocyte types are closely related. Importantly, many of the genes identified through this approach (Fig. 1A) were well-characterized, T cell-specific genes, such as CD3 chains, CD28, ICOS, CD40L, TCR, Lck, Zap70, and GATA3, which confirmed the validity of our approach. This gave greater credence and reliability to the numerous additional genes that have not yet been necessarily associated with T cell-selective roles. We selected several genes and confirmed their T cell-specific expression using real-time PCR (Fig. 1B). These genes included IL-32 (previously known as NK4), a recently identified inflammatory cytokine (20); BCL11b, a transcriptional repressor that controls T cell development (21, 22); EOMES, a paralogue of T-bet important during embryogenesis (23), but also able to regulate CD8+ effector function (24); and UBASH3A (also known as suppressor of TCR signaling-2), which is a regulator of signaling pathway involved in T cell activation. We also confirmed T cell-specific expression of two GTPase members of the immunity-associated protein (IMAP) (or immunity-associated nucleotide (IAN)) family GTPase, IMAP family member (GIMAP)5 (hIAN5 and IAN4L) and GIMAP7 (hINA7). Interestingly, we also noted preferential expression of GIMAP6 (hIAN2) by T cells (supplementary Table I). This family is relatively poorly characterized to date; however, members of this family seem to be highly expressed in the spleen and lymph nodes, and there is increasing evidence of their importance in T cell development and maturation (25, 26, 27). Fig. 1A shows that in addition to genes specifically expressed by T cells, there were also many genes that were specifically absent from T cells compared with other leukocyte subsets. Not surprisingly, these included several genes associated with Ag processing and presentation, endocytosis, complement receptors, and numerous other genes. It is beyond the scope of this study to discuss all the novel genes identified as T cell specific; however, the important point of this work is that bioinformatic strategies and large gene expression datasets provide a novel way of identifying T cell-biased molecules.



View larger version (68K):
[in this window]
[in a new window]
  		FIGURE 1. Identification of T cell-selective gene expression using comprehensive gene microarray profiling. A, cRNA was generated from numerous leukocyte types (Table I), and microarray gene expression profiles were generated using Affymetrix U133A+B genechips and analyzed using GeneSpring software. One-way ANOVA was used to identify the genes that were significantly different among the numerous T cell subsets and non-T cell leukocyte subsets. Approximately 100 genes most significantly (lowest p value) preferentially expressed in T cells are listed. SymAtlas (19 ) was used to assess the expression of these genes in nonhemopoietic tissues, and a summary is provided in the right column. B, Differential expression of genes identified from microarray analysis was validated using semiquantitative real-time PCR. RNA was isolated from purified populations of different human cell types, and relative mRNA expression of selected genes was determined. Lane 1, Th1 cells; lane 2, Th2 cells; lane 3, blood {alpha}{beta} T cells; lanes 4 and 5, eosinophils; lanes 6 and 7, neutrophils; lane 8, mast cells; lane 9, monocytes; lane 10, macrophages; lane 11, immature DC. EOMES, eomesodermin homologue; BCL11B, B cell CLL/lymphoma 11B; UBASH3A, ubiquitin associated and SH3 containing. GIMAP5, GTPase IMAP family member 5; GIMAP7, GTPase IMAP family member 7. C, Genes involved in TCR complex signaling and T cell costimulation were selected from the genes preferentially expressed by T cells and mapped using the software application GenMapp2. Genes that were preferentially expressed by T cells compared with all other leukocytes are shown in yellow boxes. Downward blue arrows signify that the gene was down-regulated after activation; upward red arrows indicate up-regulation. An asterisk next to the gene name indicates that multiple probes showed variable expression patterns. Gene expression was normalized at ~1 for all experiments, and the scale shown in this figure is the same for the following figures. The color scale indicates normalized expression levels, a value of 1 indicates normal expression, 0 indicates low expression, and values >1 indicate high expression.

 
One of the processes unique to T cells is TCR signaling. We found >150 TCR signaling or costimulation related genes that were significantly preferentially expressed in T cells (supplementary Table II). Selected genes involved in the development of the immunological synapse and downstream signaling in T cells are shown in Fig. 1C. The majority of genes involved in T cell signaling cascades were restricted to T cells, although some were expressed by other leukocytes. Interestingly, many of the T cell signaling molecules were expressed at a higher level in resting, rather than activated, T cells, which was noted in previous studies (28, 29) and could be linked to down-modulation of the TCR signaling apparatus after stimulation (30, 31).

Gene expression signatures of T cell subsets

After identification of all the genes that distinguish T cells from other leukocytes, we focused on the differences between the T cell subsets. Our dataset of genechip profiles contains a diverse array of T cell subsets, including resting and activated Th0 and Th2 cells; naive, central, and effector memory CD4+ T cells; CD8+ T cells; and {alpha}{beta} and {gamma}{delta} T cells (Table I). A one-way ANOVA with Student-Newman-Keuls post-hoc testing was used to identify the genes that distinguish particular subsets of T cells from all others. NK cells were also included in this study, because we noted previously that T cells share many genes with NK cells. Unsupervised clustering of these genes (Fig. 2A) revealed easily identifiable and distinct gene expression signatures that distinguished each subset. For instance, the TFH-specific signature consisted of well-known markers of this subset, such as CXCR5 and CXCL13, and many genes not previously associated with this subset (2, 3). Of note, gene expression signatures were not only characterized by highly preferentially expressed genes, but also by the genes whose expression was particularly low or absent in a particular subset, compared with others. This was especially true among closely related subsets, where few genes were expressed exclusively by a single subset. For instance, we identified few genes that were specific to TEM or TCM cells, and these memory T cell populations appeared to be highly related based on the limited differences in gene expression between the two. We have listed select genes from signature clusters, including genes up- and down-regulated in TFH cells (Fig. 2A, i and vi), down-regulated in {gamma}{delta} T cells (Fig. 2A, ii), and down-regulated in TCM (Fig. 2A, v) and TEM (Fig. 2A, vii) and others. Genes contributing to Th1 and Th2 signatures have been discussed previously (15). Importantly, gene expression profiling of the numerous T cell subsets allowed us to distinguish clear and discriminating signatures based on the presence and the absence of certain transcripts.



View larger version (68K):
[in this window]
[in a new window]
  		FIGURE 2. Distinct gene expression signatures characterize effector/memory T cell subsets. Gene expression profiles of different T cell subsets were analyzed using one-way ANOVA with Benjamini multiple testing correction and Student-Newman-Keuls post-hoc tests to identify genes with significant changes in expression among the subsets. A, The union of all genes that distinguish the various cell types was created by combining genes that differentiate a single subset from all others. Unsupervised clustering using Pearson similarity measure was used to arrange the genes (y-axis) and microarrays (x-axis). B, A selection of differentially expressed genes (highly significant) that reliably distinguish T cell subsets. One or two genes that most reliably (lowest p value) differentiate between one particular T cell subset and all the others were selected and combined into the predictor list. The classification value of this list was validated when the correct T cell subset was predicted using the k-nearest neighbors method (p < 0.05 for most subsets).

 
A signature for {gamma}{delta} T cells

A perplexing question in T cell biology is the relevance and functions of the {gamma}{delta} T cell subset (32). Certain transcriptional profiling approaches such as serial analysis of gene expression (33, 34) have identified preferentially expressed genes in the mouse; however, to date, sheep and cattle are the only species in which {gamma}{delta} T cells are known to express a distinctive marker, T19 (35). We sought to identify genes specific to human {gamma}{delta} T cells that might underlie a function unique to this enigmatic subset. Surprisingly, we found few genes expressed specifically by {gamma}{delta} T cells. A number of genes highly expressed by {gamma}{delta} T cells were also expressed by NK cells, in particular, a number of KIR genes (Fig. 2A, iv). The expression of these genes, however, was not confined to NK and {gamma}{delta} T cells, because some {alpha}{beta} effector T cells also expressed transcripts for these molecules. The gene expression signature of {gamma}{delta} T cells could be distinguished more easily by low or absent expression of a number of {alpha}{beta} T cell transcripts, rather than by the expression of any {gamma}{delta}-specific genes (Fig. 2A, ii). Even the {gamma}-chain of the TCR was expressed in some {alpha}{beta} T cells. Among the genes down-regulated in {gamma}{delta} T cells were several genes involved in pre-mRNA splicing processes, including transformer 2{beta}, suppressor of white apricot homologue 2 (splicing factors, arginine/serine rich 10 and 16, respectively), and pre-mRNA splicing factor 16 (the conserved motif Asp-Glu-Ala-Asp/H box polypeptide 38). The down-regulation of these splicing factors, compared with that in {alpha}{beta} T cells, may be due to {gamma}{delta} T cells not requiring extensive alternative splicing of the TCR. It should be noted that to date, only blood {gamma}{delta} T cells have been assessed, not those from epithelial or lymphoid tissues.

T cell subset predictor genes

As an extension of our signature gene analysis for the different T cell subsets, we next sought to identify small numbers of predictor genes that could reliably distinguish between all the different T cell subsets in this study. We selected one or two genes with the lowest p value from each subset that most reliably distinguished that particular T cell subset from all other T cells, creating a list of 13 genes (Fig. 2B). We found that using this small subset of genes it was possible to correctly identify the T cell type used in each microarray experiment using a k-nearest neighbors method. As few as 13 genes were sufficient to distinguish between 11 different T cell subsets and NK cells, demonstrating the remarkable power of this approach for cell type classification. Interestingly, the genes that made up the optimal predictor set were not necessarily well-known markers and included intracellular signaling molecule (such as phosphatases (DUSP4)) and obscure uncharacterized genes (such as BC006146 and AI766311, which do not show homology to any known gene families). These results demonstrate the utility of gene expression profiling for T cell type classification, similar to the way in which predictor genes have been used to distinguish cancer subtypes (36, 37, 38, 39). An important qualification, however, is that a much larger number of replicates will be required for a more reliable classification of unknown subsets of T cells, and probably the predictor set of genes will be refined with increased sample numbers for each subset and additional subsets, such as regulatory T cells. Nevertheless, our analysis demonstrates the potential of microarray dataset analysis for identification of unexpected molecules as biomarkers for T cell subset classification.

T cell activation induces an extensive transcriptional program

T cell subsets examined in this study represent many of the major paradigms of T cell biology, for instance, TCR {gamma}{delta} vs {alpha}{beta} T cells, Th1 vs Th2, and T cell activation and memory cell development. Principal component analysis was undertaken to identify the most distinctive expression patterns in all our T cell subsets. Principal component analysis involves a mathematical procedure that transforms a number of possibly correlated variables (in this case, gene expression data points) into a smaller number of uncorrelated variables, called principal components. The first three principal components, which accounted for more than half the variability in gene expression, clearly separated T cells based on the isolation method and activation state. Thus, T cell activation and culture were the two major influences on gene expression patterns in the T cell subsets in this study (Fig. 3). This was not totally surprising, because previous studies have documented the profound changes in T cell transcription upon activation (28).



View larger version (25K):
[in this window]
[in a new window]
  		FIGURE 3. Principal component analysis (PCA) identifies key parameters affecting gene expression patterns in T cell subsets. PCA was performed on all T cell arrays using GeneSpring software. Only probe sets that were expressed in at least two T cell arrays were used in the analysis. This analysis identified the predominant patterns or parameters responsible for most of the variability in the T cell gene expression profiles. The first three principal components, which accounted for over half the variability in gene expression, separated T cells based on the isolation method and activation state.

 
To identify transcriptional changes induced by T cell activation, we analyzed gene expression in Th0 and Th2 cells before and after activation in vitro. TCR engagement (anti-CD3 and PdBu treatment) was accompanied by a vast change in gene expression in Th0 and Th2 cells. The majority of genes affected by activation were regulated in a similar manner in both Th0 and Th2 cells (Fig. 4A). Activation induced an extensive change in gene expression, with >7000 genes significantly different between resting and activated T cells. A detailed description of many (or even some) of the individual genes is not possible in this study; however, global analyses that reflects changes in biological processed can be more informative, particularly when very large numbers of gene transcripts are regulated. We used gene ontology in the GeneSpring software program to identify the biological processes that were up-regulated in either resting or activated T cells (Fig. 4B). The p value for the random overlap allowed us to identify the processes that were over-represented in our gene list compared with what would be expected by chance. As expected, activated T cells up-regulated a number of processes associated with higher metabolic and effector activities, including genes involved in transcription and translation, and also cell-cell signaling and cell adhesion. Not surprisingly, genes involved in immune effector function were significantly regulated; for instance, a large number of cytokines and chemokines showed major changes in expression. Indeed, the transcript levels for almost all cytokines and chemokines were increased after activation (not shown). Interestingly, IL-10R was down-regulated after activation, suggesting that T cells are less susceptible to suppression after TCR engagement.



View larger version (56K):
[in this window]
[in a new window]
  		FIGURE 4. Gene expression profiles of resting and activated T cells. Th2 clones generated from naive CD4+ T cells were cultured under Th1- and Th2-polarizing conditions to generate Th0 and Th2 clones, respectively, and RNA was extracted from resting and activated T cell clones after 4 wk. T cells (107/condition) were left untreated (NA) or were activated (ACT) for 4 h with 1 µg/ml anti-CD3 Ab and 50 ng/ml PdBu. Gene expression of resting and activated T cells was analyzed using U133A+B Affymetrix microarrays (three different donors). A, Genes that showed significant changes in expression between resting and activated Th0 and Th2 cells were identified using one-way ANOVA. The numbers of genes regulated by activation in Th0, but not Th2; in Th2, but not Th0; and in both are shown by Venn diagram. B, The Gene Ontology biological process was created using GeneSpring software, and processes that showed a significant overlap with genes that increased or decreased after T cell activation are listed.

 
A number of processes involving proliferation, cell cycle, and apoptosis were regulated in both resting and activated cells. Resting T cells expressed higher levels of transcripts for genes involved in cytoskeleton organization, including dynein; myosin VA and IXB; capping protein; tubulin-specific chaperones c, e, and d; and others. Cytoskeleton rearrangements are an integral part of the T cell activation process (40) and are early events in the formation of the immunological synapse. Down-regulation of these transcripts in T cells stimulated for 4 h may serve as a feedback mechanism.

Gene expression in memory T cell subsets: TCM vs TEM

Gene signatures for T cell subsets are not necessarily informative for revealing all the differences between two particular subsets when a proportion of the genes is shared by numerous other T cell types. One interesting subdivision of the memory pool is the TCM vs TEM distinction defined by the presence or the absence of CCR7 (9). We compared TCM and TEM cells directly and identified differentially expressed genes using Affymetrix software. Surprisingly few genes were reproducibly differentially expressed between these two memory T cell populations (Fig. 5), suggesting that the phenotypes of these two subsets are quite similar. Predictably, TCM cells expressed higher levels of CCR7 and L-selectin, which are defining molecules for this subset and facilitate secondary lymphoid organ homing. TEM cells expressed a number of molecules more closely associated with effector function, including chemokines CCL4, CCL5, and XCL1 and CTL-related molecules such as granulysin and granzymes A and B (41). Because the CD45RO+CCR7– TEM subset might be quite heterogeneous (10), there may well be circulating Th1- and Th2-like effector cells and possibly also some NK T and {gamma}{delta} T cells in this population. Nevertheless, our genechip results suggest that TEM are somewhat related to and resemble effector T cells, although the distinction between TEM and TCM is also relatively small. Whether CD45RO+CCR7– TEM cells in tissues display a phenotype even closer to that of effector T cells is a likely possibility.



View larger version (41K):
[in this window]
[in a new window]
  		FIGURE 5. Genes differentially expressed by TCM and TEM cells. Genes that differed between TEM and TCM by >2-fold (as determined using MAS5.0 software) in two replicate experiments are shown. The color scale indicates the level of expression. The signal log ratio (SLR) is the log of fold change.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
In this study we used an extensive collection of Affymetrix genechip datasets, representing gene expression profiles of various leukocyte and T cell subsets. This dataset collection allowed us to identify T cell-specific gene transcription based on gene expression in the T cell subsets and absence in the non-T leukocyte subsets. Importantly, this signature comprised both well-characterized markers of T cell function and differentiation as well as a number of more obscure genes previously not associated with T cell biology. Several of these more obscure genes were validated by PCR, and we (15) and others (12) generally found a reasonable concordance between Affymetrix expression data and PCR results. Thus, we believe that our approach is valid for the identification of additional factors that are intimately involved in the biology of T cells. Of the numerous T cell-expressed genes that were absent from all other leukocyte types, many, in fact, were expressed in other tissues. For instance, some T cell specific genes (at least among the leukocytes) were also expressed in brain (approximately five genes; Fig. 1A); perhaps the genes serve the same function in both cell types, but, more likely, the T cell immune system may have simply adopted receptors and ligands from other systems. For instance, naive T cells use CXCR4 for cell migration; however, CXCR4 is a widely expressed molecule and in evolution predates the emergence of adaptive immune responses (42). One of the main aims of this study was to establish datasets and bioinformatic strategies for the identification of new T cell-restricted genes and molecules. The relevance of many of these genes for T cell responses must await studies with gene-deficient mice or other systems that suitably address molecular functions in T cells. Regardless, the datasets described in this study should focus the attention of T cell biologists to many poorly characterized genes that probably play important functional roles in particular T cell subsets.

An in-depth examination of gene expression profiles of numerous effector and memory T cell subsets revealed specific signatures that distinguished each subset. Interestingly, some subsets were more easily distinguished by the genes that were down-regulated or absent. We also showed that T cell subsets could be easily distinguished from each other through the use of a small number of predictor genes. Previous studies have demonstrated the application of gene expression profiling to distinguish subtypes of neoplastic disease (36, 37, 38, 39), and gene microarrays have been used for more precise diagnosis and subgrouping of cancers (43, 44, 45). Moreover, gene microarrays have proven useful for predicting disease outcomes, particularly for cancers (46, 47, 48, 49). Similarly, leukocyte-specific signatures can be used for classification of unknown cells or cancer cell types based on their gene expression patterns. Although leukocytes were previously characterized by their expression of cell surface markers, gene expression by microarrays provides a new basis for leukocyte subset classification. Precise signatures for many of the T cell subsets examined in this study must await further analyses with a greater numbers of genechips per subset. However, we established that the use of predictor signatures is applicable to T cell subset classification, and we expect that the predictor genes reported in this study will be refined to accurately distinguish various stages of T cell differentiation and function.

The {gamma}{delta} T cell subset remains a largely enigmatic population. Recent attempts to identify a specific transcriptional profile for {gamma}{delta} T cells (34, 50, 51, 52) revealed many interesting features of {gamma}{delta} T cells, such as their activated, yet resting, phenotype (34); however, a unique {gamma}{delta} molecular profile remains to be identified. Our study, which is one of the first to examine by microarray gene expression the profiles of circulating human {gamma}{delta} T cells, showed that {gamma}{delta} T cells shared many features with {alpha}{beta} T cells and NK cells, yet expressed few, if any, molecules that were {gamma}{delta} specific. This was surprising, because a {gamma}{delta}-specific marker, termed T19 (aka WC1), has been identified in sheep and cattle (35). This result was, however, in accordance with the results of a recent study that showed that {gamma}{delta} T cells in mice share many features with an unconventional subset of {alpha}{beta} T cells (33). Although we were unable to identify any {gamma}{delta}-specific markers, their gene expression signature did provide some interesting insights into the {gamma}{delta} T cell biology. Although {gamma}{delta} T cells have the potential for extensive receptor diversity, it appears that this potential is not fully realized (53). We found that, compared with conventional {alpha}{beta} T cells, {gamma}{delta} T cells showed reduced expression of several proteins involved in pre-mRNA splicing, which could explain the lack of TCR diversity in these cells. The failure to identify any {gamma}{delta}-specific transcripts confounds attempts to resolve the functional relevance of this subset. In accordance with mouse studies, our human expression data support the view that {alpha}{beta} and {gamma}{delta} subsets show overlapping molecular and functional profiles, and the reason why different vertebrates show varying levels of {gamma}{delta} T cells in blood or at epithelial surfaces, especially around birth, is still unanswered.

Memory T cells have been divided into effector and central subsets based on their expression of homing molecules as well as other attributes. TEM cells express tissue-homing receptors and provide immediate effector function after stimulation, whereas TCM cells express lymphoid-homing receptors and provide long-term systemic protection (9). We found surprisingly few differences between TCM and TEM cells isolated from human peripheral blood, at least at the level of gene expression. TEM did express higher levels of several molecules associated with T cell effector function, whereas TCM cells expressed, predictably, higher levels of genes encoding CCR7 and CD62L (L-selectin). The similar gene expression profiles of TEM and TCM suggest that these two memory subsets may serve largely similar functions, albeit in different locations. The functional distinction between TEM and TCM subsets is probably less clearly defined than previously thought, because TCM cells in mice are capable of providing immediate protection (54). In our study we examined gene expression profiles of circulating memory T cells; however, it is possible that TEM cells in tissues display a more distinctive phenotype, closer to that of an effector cell. Others have shown that TEM and TCM cells are heterogeneous (10) and include some Th1- and Th2-like cells as well as possibly some NK T and {gamma}{delta} T cells and skin- and gut-homing T cells. With recent advances in cRNA production techniques and multicolor cell sorting, isolation of even smaller subsets of TEM and TCM cells should be feasible, and this should provide additional insight into the nature of immunological memory in the T cell system.

One reservation regarding our approach is that not every conceivable leukocyte subset, at all stages of differentiation and stimulation, was included in our datasets. For instance, CD5 came up as a T cell-restricted gene, but is expressed by a particular subset of B cells, which have not yet been profiled by us. Another caveat is that certain T cell subsets have yet to be included in our analysis, such as T regulatory cells. It will also be interesting to profile Th1 or Th2 cells isolated directly from tissues, which currently is technically demanding because of cell numbers and uncertainty regarding the precise surface phenotype of Th1 or Th2 cells. Nevertheless, the potential of using large datasets for identification of interesting gene signatures is readily apparent from our study, and the potential exists for other interesting datasets to be combined with ours. The datasets used in this study are freely available for download from <http://linkage.garvan.unsw.edu.au/public/microarrays/> and will also be deposited in public databases. Because of the large numbers of genes identified in the various analyses described in this study, it would be impossible to validate all of these ourselves, and we urge investigators with an interest in a particular gene to validate, by PCR methods, genes of interest.

In conclusion, we have used an extensive dataset of leukocyte gene expression profiles to identify genes expressed selectively by T cells and to gain an understanding of the molecular changes that accompany T cell activation and differentiation. The identity of numerous novel genes selectively expressed by T cells will provide important insight into T cell function in general as well as the specialized roles of subsets of T cells. In addition, strategies used in this study for T cell and leukocyte microarray dataset analysis clearly identify interesting subset-restricted genes, which should provide important new insight into T cell functions.


   Acknowledgments
 
We thank Stuart Tangye, Kim Good, Mary Sisavanh, Sabine Zimmer, Melinda Frost, and Trina So for the use of microarray data, and Mark Hughes for help with data analysis. We also thank Ian Mackay for helpful suggestions.


   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 was supported by the Clinical Research Center for Asthma, the National Health and Medical Research Council, and the Swiss National Science Foundation (Grant 3100-101962). Back

2 Address correspondence and reprint requests to Dr. Charles R. Mackay, Garvan Institute of Medical Research, 384 Victoria Street, Darlinghurst, New South Wales 2010, Australia. E-mail address: c.mackay{at}garvan.org.au Back

3 Abbreviations used in this paper: TFH, follicular B helper T cell; EAT-2, EWS/FLI1 activated transcript-2; PdBu, phorbol dibutyrate; SH2, Src homology 2; TCM, T central memory; TEM, T effector memory; XLP, X-linked lymphoproliferative syndrome; IMAP, immunity-associated protein; IAN, immunity-associated nucleotide; GIMAP, GTPase, IMAP family member. Back

Received for publication January 21, 2005. Accepted for publication September 27, 2005.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 

  1. Mosmann, T. R., H. Cherwinski, M. W. Bond, M. A. Giedlin, R. L. Coffman. 1986. Two types of murine helper T cell clone. I. Definition according to profiles of lymphokine activities and secreted proteins. J. Immunol. 136: 2348-2357. [Abstract]
  2. Breitfeld, D., L. Ohl, E. Kremmer, J. Ellwart, F. Sallusto, M. Lipp, R. Forster. 2000. Follicular B helper T cells express CXC chemokine receptor 5, localize to B cell follicles, and support immunoglobulin production. J. Exp. Med. 192: 1545-1552. [Abstract/Free Full Text]
  3. Schaerli, P., K. Willimann, A. B. Lang, M. Lipp, P. Loetscher, B. Moser. 2000. CXC chemokine receptor 5 expression defines follicular homing T cells with B cell helper function. J. Exp. Med. 192: 1553-1562. [Abstract/Free Full Text]
  4. Sallusto, F., D. Lenig, C. R. Mackay, A. Lanzavecchia. 1998. Flexible programs of chemokine receptor expression on human polarized T helper 1 and 2 lymphocytes. J. Exp. Med. 187: 875-883. [Abstract/Free Full Text]
  5. Sallusto, F., A. Lanzavecchia, C. R. Mackay. 1998. Chemokines and chemokine receptors in T-cell priming and Th1/Th2-mediated responses. Immunol. Today 19: 568-574. [Medline]
  6. Qin, S., J. B. Rottman, P. Myers, N. Kassam, M. Weinblatt, M. Loetscher, A. E. Koch, B. Moser, C. R. Mackay. 1998. The chemokine receptors CXCR3 and CCR5 mark subsets of T cells associated with certain inflammatory reactions. J. Clin. Invest. 101: 746-754. [Medline]
  7. Messi, M., I. Giacchetto, K. Nagata, A. Lanzavecchia, G. Natoli, F. Sallusto. 2003. Memory and flexibility of cytokine gene expression as separable properties of human TH1 and TH2 lymphocytes. Nat. Immunol. 4: 78-86. [Medline]
  8. Forster, R., A. Schubel, D. Breitfeld, E. Kremmer, I. Renner-Muller, E. Wolf, M. Lipp. 1999. CCR7 coordinates the primary immune response by establishing functional microenvironments in secondary lymphoid organs. Cell 99: 23-33. [Medline]
  9. Sallusto, F., D. Lenig, R. Forster, M. Lipp, A. Lanzavecchia. 1999. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 401: 708-712. [Medline]
  10. Sallusto, F., J. Geginat, A. Lanzavecchia. 2004. Central memory and effector memory T cell subsets: function, generation, and maintenance. Annu. Rev. Immunol. 22: 745-763. [Medline]
  11. Shaffer, A. L., A. Rosenwald, E. M. Hurt, J. M. Giltnane, L. T. Lam, O. K. Pickeral, L. M. Staudt. 2001. Signatures of the immune response. Immunity 15: 375-385. [Medline]
  12. Rogge, L., E. Bianchi, M. Biffi, E. Bono, S. Y. Chang, H. Alexander, C. Santini, G. Ferrari, L. Sinigaglia, M. Seiler, et al 2000. Transcript imaging of the development of human T helper cells using oligonucleotide arrays. Nat. Genet. 25: 96-101. [Medline]
  13. Chtanova, T., R. A. Kemp, A. P. Sutherland, F. Ronchese, C. R. Mackay. 2001. Gene microarrays reveal extensive differential gene expression in both CD4+ and CD8+ type 1 and type 2 T cells. J. Immunol. 167: 3057-3063. [Abstract/Free Full Text]
  14. McHugh, R. S., M. J. Whitters, C. A. Piccirillo, D. A. Young, E. M. Shevach, M. Collins, M. C. Byrne. 2002. CD4+CD25+ immunoregulatory T cells: gene expression analysis reveals a functional role for the glucocorticoid-induced TNF receptor. Immunity 16: 311-323. [Medline]
  15. Chtanova, T., S. G. Tangye, R. Newton, N. Frank, M. R. Hodge, M. S. Rolph, C. R. Mackay. 2004. T follicular helper cells express a distinctive transcriptional profile, reflecting their role as non-Th1/Th2 effector cells that provide help for B cells. J. Immunol. 173: 68-78. [Abstract/Free Full Text]
  16. Baugh, L. R., A. A. Hill, E. L. Brown, C. P. Hunter. 2001. Quantitative analysis of mRNA amplification by in vitro transcription. Nucleic Acids Res. 29: E29[Medline]
  17. Ng, L. G., A. P. Sutherland, R. Newton, F. Qian, T. G. Cachero, M. L. Scott, J. S. Thompson, J. Wheway, T. Chtanova, J. Groom, et al 2004. B cell-activating factor belonging to the TNF family (BAFF)-R is the principal BAFF receptor facilitating BAFF costimulation of circulating T and B cells. J. Immunol. 173: 807-817. [Abstract/Free Full Text]
  18. Rozen, S., H. Skaletsky. 2000. Primer3 on the WWW for general users and for biologist programmers. Methods Mol. Biol. 132: 365-386. [Medline]
  19. Su, A. I., T. Wiltshire, S. Batalov, H. Lapp, K. A. Ching, D. Block, J. Zhang, R. Soden, M. Hayakawa, G. Kreiman, et al 2004. A gene atlas of the mouse and human protein-encoding transcriptomes. Proc. Natl. Acad. Sci. USA 101: 6062-6067. [Abstract/Free Full Text]
  20. Kim, S. H., S. Y. Han, T. Azam, D. Y. Yoon, C. A. Dinarello. 2005. Interleukin-32: a cytokine and inducer of TNF{alpha}. Immunity 22: 131-142. [Medline]
  21. Cismasiu, V. B., K. Adamo, J. Gecewicz, J. Duque, Q. Lin, D. Avram. 2005. BCL11B functionally associates with the NuRD complex in T lymphocytes to repress targeted promoter. Oncogene 24: 6753-6764. [Medline]
  22. Wakabayashi, Y., H. Watanabe, J. Inoue, N. Takeda, J. Sakata, Y. Mishima, J. Hitomi, T. Yamamoto, M. Utsuyama, O. Niwa, et al 2003. Bcl11b is required for differentiation and survival of {alpha}{beta} T lymphocytes. Nat. Immunol. 4: 533-539. [Medline]
  23. Russ, A. P., S. Wattler, W. H. Colledge, S. A. Aparicio, M. B. Carlton, J. J. Pearce, S. C. Barton, M. A. Surani, K. Ryan, M. C. Nehls, et al 2000. Eomesodermin is required for mouse trophoblast development and mesoderm formation. Nature 404: 95-99. [Medline]
  24. Pearce, E. L., A. C. Mullen, G. A. Martins, C. M. Krawczyk, A. S. Hutchins, V. P. Zediak, M. Banica, C. B. DiCioccio, D. A. Gross, C. A. Mao, et al 2003. Control of effector CD8+ T cell function by the transcription factor eomesodermin. Science 302: 1041-1043. [Abstract/Free Full Text]
  25. Krucken, J., R. M. Schroetel, I. U. Muller, N. Saidani, P. Marinovski, W. P. Benten, O. Stamm, F. Wunderlich. 2004. Comparative analysis of the human gimap gene cluster encoding a novel GTPase family. Gene 341: 291-304. [Medline]
  26. Poirier, G. M., G. Anderson, A. Huvar, P. C. Wagaman, J. Shuttleworth, E. Jenkinson, M. R. Jackson, P. A. Peterson, M. G. Erlander. 1999. Immune-associated nucleotide-1 (IAN-1) is a thymic selection marker and defines a novel gene family conserved in plants. J. Immunol. 163: 4960-4969. [Abstract/Free Full Text]
  27. Cambot, M., S. Aresta, B. Kahn-Perles, J. de Gunzburg, P. H. Romeo. 2002. Human immune associated nucleotide 1: a member of a new guanosine triphosphatase family expressed in resting T and B cells. Blood 99: 3293-3301. [Abstract/Free Full Text]
  28. Teague, T. K., D. Hildeman, R. M. Kedl, T. Mitchell, W. Rees, B. C. Schaefer, J. Bender, J. Kappler, P. Marrack. 1999. Activation changes the spectrum but not the diversity of genes expressed by T cells. Proc. Natl. Acad. Sci. USA 96: 12691-12696. [Abstract/Free Full Text]
  29. Diehn, M., A. A. Alizadeh, O. J. Rando, C. L. Liu, K. Stankunas, D. Botstein, G. R. Crabtree, P. O. Brown. 2002. Genomic expression programs and the integration of the CD28 costimulatory signal in T cell activation. Proc. Natl. Acad. Sci. USA 99: 11796-11801. [Abstract/Free Full Text]
  30. Salio, M., S. Valitutti, A. Lanzavecchia. 1997. Agonist-induced T cell receptor down-regulation: molecular requirements and dissociation from T cell activation. Eur. J. Immunol. 27: 1769-1773. [Medline]
  31. Valitutti, S., S. Muller, M. Salio, A. Lanzavecchia. 1997. Degradation of T cell receptor (TCR)-CD3-{zeta} complexes after antigenic stimulation. J. Exp. Med. 185: 1859-1864. [Abstract/Free Full Text]
  32. Hayday, A. C.. 2000. {gamma}{delta} cells: a right time and a right place for a conserved third way of protection. Annu. Rev. Immunol. 18: 975-1026. [Medline]
  33. Pennington, D. J., B. Silva-Santos, J. Shires, E. Theodoridis, C. Pollitt, E. L. Wise, R. E. Tigelaar, M. J. Owen, A. C. Hayday. 2003. The inter-relatedness and interdependence of mouse T cell receptor {gamma}{delta}+ and {alpha}{beta}+ cells. Nat. Immunol. 4: 991-998. [Medline]
  34. Shires, J., E. Theodoridis, A. C. Hayday. 2001. Biological insights into TCR{gamma}{delta}+ and TCR{alpha}{beta}+ intraepithelial lymphocytes provided by serial analysis of gene expression (SAGE). Immunity 15: 419-434. [Medline]
  35. Hein, W. R., C. R. Mackay. 1991. Prominence of {gamma}{delta} T cells in the ruminant immune system. Immunol. Today 12: 30-34. [Medline]
  36. Alizadeh, A. A., D. T. Ross, C. M. Perou, M. van de Rijn. 2001. Towards a novel classification of human malignancies based on gene expression patterns. J. Pathol. 195: 41-52. [Medline]
  37. Alizadeh, A. A., M. B. Eisen, R. E. Davis, C. Ma, I. S. Lossos, A. Rosenwald, J. C. Boldrick, H. Sabet, T. Tran, X. Yu, et al 2000. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature 403: 503-511. [Medline]
  38. Golub, T. R., D. K. Slonim, P. Tamayo, C. Huard, M. Gaasenbeek, J. P. Mesirov, H. Coller, M. L. Loh, J. R. Downing, M. A. Caligiuri, et al 1999. Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science 286: 531-537. [Abstract/Free Full Text]
  39. van ’t Veer, L. J., H. Dai, M. J. van de Vijver, Y. D. He, A. A. Hart, M. Mao, H. L. Peterse, K. van der Kooy, M. J. Marton, A. T. Witteveen, et al 2002. Gene expression profiling predicts clinical outcome of breast cancer. Nature 415: 530-536. [Medline]
  40. Cannon, J. L., J. K. Burkhardt. 2002. The regulation of actin remodeling during T-cell-APC conjugate formation. Immunol. Rev. 186: 90-99. [Medline]
  41. Zaunders, J. J., W. B. Dyer, B. Wang, M. L. Munier, M. Miranda-Saksena, R. Newton, J. Moore, C. R. Mackay, D. A. Cooper, N. K. Saksena, et al 2004. Identification of circulating antigen-specific CD4+ T lymphocytes with a CCR5+, cytotoxic phenotype in an HIV-1 long-term nonprogressor and in CMV infection. Blood 103: 2238-2247. [Abstract/Free Full Text]
  42. Huising, M. O., R. J. Stet, C. P. Kruiswijk, H. F. Savelkoul, B. M. Lidy Verburg-van Kemenade. 2003. Molecular evolution of CXC chemokines: extant CXC chemokines originate from the CNS. Trends Immunol. 24: 307-313. [Medline]
  43. Davis, R. E., L. M. Staudt. 2002. Molecular diagnosis of lymphoid malignancies by gene expression profiling. Curr. Opin. Hematol. 9: 333-338. [Medline]
  44. Rosenwald, A., G. Wright, W. C. Chan, J. M. Connors, E. Campo, R. I. Fisher, R. D. Gascoyne, H. K. Muller-Hermelink, E. B. Smeland, J. M. Giltnane, et al 2002. The use of molecular profiling to predict survival after chemotherapy for diffuse large-B-cell lymphoma. N. Engl. J. Med. 346: 1937-1947. [Abstract/Free Full Text]
  45. Staudt, L. M.. 2002. Gene expression profiling of lymphoid malignancies. Annu. Rev. Med. 53: 303-318. [Medline]
  46. Sturzebecher, S., K. P. Wandinger, A. Rosenwald, M. Sathyamoorthy, A. Tzou, P. Mattar, J. A. Frank, L. Staudt, R. Martin, H. F. McFarland. 2003. Expression profiling identifies responder and non-responder phenotypes to interferon-{beta} in multiple sclerosis. Brain 126: 1419-1429. [Abstract/Free Full Text]
  47. Dave, S. S., G. Wright, B. Tan, A. Rosenwald, R. D. Gascoyne, W. C. Chan, R. I. Fisher, R. M. Braziel, L. M. Rimsza, T. M. Grogan, et al 2004. Prediction of survival in follicular lymphoma based on molecular features of tumor-infiltrating immune cells. N. Engl. J. Med. 351: 2159-2169. [Abstract/Free Full Text]
  48. Wright, G., B. Tan, A. Rosenwald, E. H. Hurt, A. Wiestner, L. M. Staudt. 2003. A gene expression-based method to diagnose clinically distinct subgroups of diffuse large B cell lymphoma. Proc. Natl. Acad. Sci. USA 100: 9991-9996. [Abstract/Free Full Text]
  49. Wiestner, A., L. M. Staudt. 2003. Towards molecular diagnosis and targeted therapy of lymphoid malignancies. Semin. Hematol. 40: 296-307. [Medline]
  50. Hedges, J. F., D. Cockrell, L. Jackiw, N. Meissner, M. A. Jutila. 2003. Differential mRNA expression in circulating {gamma}{delta} T lymphocyte subsets defines unique tissue-specific functions. J. Leukocyte Biol. 73: 306-314. [Abstract/Free Full Text]
  51. Meissner, N., J. Radke, J. F. Hedges, M. White, M. Behnke, S. Bertolino, M. Abrahamsen, M. A. Jutila. 2003. Serial analysis of gene expression in circulating {gamma}{delta} T cell subsets defines distinct immunoregulatory phenotypes and unexpected gene expression profiles. J. Immunol. 170: 356-364. [Abstract/Free Full Text]
  52. Fahrer, A. M., Y. Konigshofer, E. M. Kerr, G. Ghandour, D. H. Mack, M. M. Davis, Y. H. Chien. 2001. Attributes of {gamma}{delta} intraepithelial lymphocytes as suggested by their transcriptional profile. Proc. Natl. Acad. Sci. USA 98: 10261-10266. [Abstract/Free Full Text]
  53. Carding, S. R., P. J. Egan. 2002. {gamma}{delta} T cells: functional plasticity and heterogeneity. Nat. Rev. Immunol. 2: 336-345. [Medline]
  54. Wherry, E. J., V. Teichgraber, T. C. Becker, D. Masopust, S. M. Kaech, R. Antia, U. H. von Andrian, R. Ahmed. 2003. Lineage relationship and protective immunity of memory CD8 T cell subsets. Nat. Immunol. 4: 225-234. [Medline]



This article has been cited by other articles:


Home page
 BloodHome page
N. C. Brembilla, J. Weber, D. Rimoldi, S. Pradervand, F. Schutz, G. Pantaleo, C. Ruegg, M. Quadroni, K. Harshman, and M.-A. Doucey
c-Cbl expression levels regulate the functional responses of human central and effector memory CD4 T cells
Blood, August 1, 2008; 112(3): 652 - 660.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
T. D. Manes and J. S. Pober
Antigen Presentation by Human Microvascular Endothelial Cells Triggers ICAM-1-Dependent Transendothelial Protrusion by, and Fractalkine-Dependent Transendothelial Migration of, Effector Memory CD4+ T Cells
J. Immunol., June 15, 2008; 180(12): 8386 - 8392.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Res.Home page
P. P. Piccaluga, C. Agostinelli, A. Califano, A. Carbone, L. Fantoni, S. Ferrari, A. Gazzola, A. Gloghini, S. Righi, M. Rossi, et al.
Gene Expression Analysis of Angioimmunoblastic Lymphoma Indicates Derivation from T Follicular Helper Cells and Vascular Endothelial Growth Factor Deregulation
Cancer Res., November 15, 2007; 67(22): 10703 - 10710.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
F. Sierro, C. Biben, L. Martinez-Munoz, M. Mellado, R. M. Ransohoff, M. Li, B. Woehl, H. Leung, J. Groom, M. Batten, et al.
Disrupted cardiac development but normal hematopoiesis in mice deficient in the second CXCL12/SDF-1 receptor, CXCR7
PNAS, September 11,&nbs