Mycobacterium Tuberculosis Activation Following Stimulation by T Cell +

Differential Gene Expression Identifies Novel Markers of CD4 + and CD8+ T Cell Activation Following Stimulation by Mycobacterium tuberculosis This information is current as of September 23, 2021. Jacqueline M. Cliff, Iryna N. J. Andrade, Rohit Mistry, Christopher L. Clayton, Mark G. Lennon, Alan P. Lewis, Ken Duncan, Pauline T. Lukey and Hazel M. Dockrell J Immunol 2004; 173:485-493; ; doi: 10.4049/jimmunol.173.1.485 Downloaded from http://www.jimmunol.org/content/173/1/485

Supplementary http://www.jimmunol.org/content/suppl/2004/06/17/173.1.485.DC1 http://www.jimmunol.org/ Material References This article cites 51 articles, 20 of which you can access for free at: http://www.jimmunol.org/content/173/1/485.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 23, 2021

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2004 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Differential Gene Expression Identiﬁes Novel Markers of CD4؉ and CD8؉ T Cell Activation Following Stimulation by Mycobacterium tuberculosis1

Jacqueline M. Cliff,2* Iryna N. J. Andrade,* Rohit Mistry,†‡ Christopher L. Clayton,† Mark G. Lennon,† Alan P. Lewis,† Ken Duncan,† Pauline T. Lukey,† and Hazel M. Dockrell*

T cell activation in response to antigenic stimulation is a complex process, involving changes in the expression level of a large number of genes. We have used cDNA array technology to characterize the differences in gene expression between human CD4؉ and CD8؉ T cells. PBMC from six healthy donors were stimulated with live Mycobacterium tuberculosis, and the gene expression profiles of each donor’s CD4؉ and CD8؉ T cells were analyzed separately. ANOVA revealed 518 genes that were consistently ؉ ؉ differentially expressed between CD4 and CD8 T cells. These differentially expressed genes include a combination of well- Downloaded from known, previously characterized genes with a range of biological functions and unknown in silico predicted hypothetical genes. Where possible, the novel genes have been characterized using bioinformatics, and putative transcription factors, signaling molecules, transmembrane, and secreted factors have been identified. A subset of these differentially expressed genes could be ex- ploited as markers of CD4؉ and CD8؉ T cell activation for use in vaccine trials. These observed differences in the gene expression profile of CD4؉ and CD8؉ T cells following activation by a human pathogen contribute to an increased understanding of T cell activation and differentiation and the roles these T cell subsets may play in immunity to infection. The Journal of Immunology, http://www.jimmunol.org/ 2004, 173: 485Ð493.

hen a T cell is stimulated by encountering an APC levels of thousands of genes simultaneously, and can be used to expressing its cognate Ag, many biological processes identify differentially expressed genes in different cell types or W occur, including metabolic activation, morphological tissues. They have been successfully used in the analysis of various changes, proliferation, and differentiation into distinct effector and aspects of T cell function, including T cell activation in vivo in memory cells. These processes are mediated in part by alterations response to staphylococcal enterotoxin B (2) and determination of in gene expression levels, and these changes are affected by co- the gene expression profile of suppressor T cells in mice (3, 4). In stimulation via cytokine receptors and accessory and adhesion human T cells, the genetic profiles of type 1 and 2 CD4ϩ and by guest on September 23, 2021 molecules, as well as the signals emanating from TCR triggering. CD8ϩ T cells have been characterized (5, 6), and the effects of CD4ϩ and CD8ϩ T cells are phenotypically similar, and both sub- signaling through the IL-2R (7) and members of the CD28 family sets respond to antigenic stimulation in a similar manner. Indeed, (8, 9) on TCR signaling have been determined. activated CD4ϩ and CD8ϩ T cells can both be induced to differ- To date, there has not been a systematic array-based analysis of entiate into either type 1 or 2 cells, secreting similar patterns of the differences in gene expression profiles of human CD4ϩ and cytokines (1). The main difference between the two T cell subsets CD8ϩ T cells following physiological Ag-specific stimulation. is their distinct roles in protective immunity, as effector CD8ϩ T This study addresses this issue by investigation of effector CD4ϩ cells recognize and destroy infected host cells by recognizing en- and CD8ϩ T cells specific for a relevant human pathogen, Myco- dogenous peptides presented by MHC class I, whereas effector bacterium tuberculosis, which leads to 8 million new cases of tu- CD4ϩ T cells recognize exogenous peptides presented via MHC berculosis (TB)3 disease and 2 million deaths each year (10). The class II and provide help to other immune cells. These differences efficacy of the currently available vaccine against M. tuberculosis, in effector CD4ϩ and CD8ϩ T cells must be reflected in differences Mycobacterium bovis bacillus Calmette-Guerin (BCG), has been in the gene expression profile of the two T cell types. variable in trials with ϳ80% observed protection against adult pul- High density cDNA and oligonucleotide microarray hybridiza- monary TB in the United Kingdom, but much lower protection tion technologies can be used to measure the mRNA expression demonstrated in large parts of the world, including SubSaharan Africa and Southeast Asia (11). The pivotal role of CD4ϩ T cells in protection against TB is clearly demonstrated by the suscepti- *Department of Infectious and Tropical Diseases, London School of Hygiene & Trop- † bility of HIV/AIDS patients to develop TB disease, and this cor- ical Medicine, London, United Kingdom; GlaxoSmithKline Research and Develop- ϩ ment, Stevenage, United Kingdom; and ‡Medical Research Council Centre for relates with the decline in CD4 T cell count. In mice, Ab-medi- ϩ Molecular & Cellular Studies, University of Stellenbosch, Cape Town, South Africa ated depletion of CD4 T cells results in rapid reactivation of a ϩ Received for publication November 18, 2003. Accepted for publication April persistent TB infection (12). Effector CD4 T cells are recruited 14, 2004. to infected lungs, where they secrete TNF-␣ and IFN-␥, leading to The costs of publication of this article were defrayed in part by the payment of page activation of the infected alveolar macrophages, enabling them to 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 funded by the GlaxoSmithKline Action TB Initiative. 2 Address correspondence and reprint requests to Dr. Jacqueline M. Cliff, Department 3 Abbreviations used in this paper: TB, tuberculosis; BCG, bacillus Calmette-Guerin; of Infectious and Tropical Diseases, London School of Hygiene & Tropical Medicine, DAP10, DNAX activation protein 10; ESTs, expressed sequence tags; MOI, multi- London, WC1E 7HT, U.K. E-mail address: [email protected] plicity of infection; PPD, purified protein derivative; qRT-PCR, quantitative RT-PCR.

ϩ kill the mycobacteria (13). A CD8 T cell response is also nec- [3H]thymidine (Amersham Biosciences, Bucks, U.K.) was added. Points ␤ shown represent the means of quadruplicate measurements for each donor. essary for immunity. 2-microglobulin-deficient mice, which lack ϩ functional CD8 T cells, are highly susceptible to M. tuberculosis IFN-␥ ELISA (14), and these cells appear to play a crucial role in the latent phase PBMC were incubated at 1 ϫ 106/ml and stimulated with PPD or M. bovis of infection (15). In human TB patients, healthy contacts, and ϩ BCG, and supernatants were collected on day 6. Quadruplicate superna- BCG-vaccinated donors, CD8 T cells reactive against a number tants were pooled and assayed in duplicate by sandwich ELISA, using a ϩ of mycobacterial Ags have been identified (16, 17). Such CD8 T matched Ab pair and rIFN-␥ standard purchased from BD PharMingen cells produce IFN-␥ and TNF-␣ (18) and are cytolytic for infected (San Diego, CA). ϩ monocyte-derived macrophages (19). Furthermore, these CD8 T Differential gene expression cells secrete granulysin, which, in conjunction with perforin, can Generation of nylon cDNA arrays, synthesis of 33P-labeled cDNA complex directly kill the M. tuberculosis bacillus within a macrophage (20). ϩ probes, and hybridization of complex probes to the arrays were performed, Cytotoxic CD4 T cells can also be generated by M. tuberculosis as previously described (26). cDNA arrays were synthesized from a nor- ϩ stimulation. However, the kinetics of CD4 T cell-mediated killing malized leukocyte cDNA library derived from five normal healthy donors, are different from those of CD8ϩ T cells, occurring after 16 h rather which was purchased from Invitrogen Life Technologies (Renfrewshire, than 4 h following antigenic stimulation in vitro (21). The control of U.K.). The library was normalized so that the redundancy was reduced, and therefore the representation of low level-expressed genes, such as cytokine cytolytic function in the two cell types is therefore different, and may and receptor genes, was enhanced (Invitrogen sales literature), allowing as be reflected in differences in the expression level of certain genes. many unique genes as possible to be analyzed. A total of 50,000 clones was M. tuberculosis exerts an immunosuppressive effect on host T picked from this library, and the inserts were amplified by PCR. The av- cell activation, which is mediated by the production of IL-10 and erage product length was 1.2 kb. These PCR products were spotted in Downloaded from duplicate across three nylon membranes. Similarly, the human sequence- TGF-␤ by infected macrophages (22). At diagnosis, TB patients ϩ verified arrays were produced from nearly 8,000 clones (Incyte Genomics, exhibit a depressed CD4 T cell proliferation and IFN-␥ secretion Palo Alto, CA), comprising ϳ6,000 characterized genes and 2,000 ex- response to purified protein derivative (PPD) stimulation in vitro pressed sequence tags (ESTs), spotted in duplicate. Before use, the DNA compared with BCG-vaccinated control subjects (23). CD8ϩ T loading on each array was checked by hybridization of [␥-33P]ATP-labeled cell responses are also reduced in TB patients, with a decrease in oligonucleotides, which were targeted to the vector sequence of the PCR products. the amount of IFN-␥ and TNF-␣ produced (24) and also in cyto- ␮ http://www.jimmunol.org/ ϩ RNA samples (5 g) were reverse transcribed using SuperScript II (In- lytic activity (25). Suppression of these functions of CD4 and vitrogen Life Technologies) and oligo(dT15)VN primer, in the presence of CD8ϩ T cells may therefore be to the pathogen’s advantage. [␣-33P]dCTP (Amersham Biosciences). Unincorporated nucleotides were In the present study, we used human cDNA arrays to charac- removed by purification on Sephadex G50 spin columns (Amersham Bio- sciences). Duplicate complex probes were synthesized from each RNA terize the differences in the gene expression profile of effector ϩ ϩ sample, and their quality was assessed by PAGE after denaturation. The CD4 and CD8 T cells, in response to M. tuberculosis infection complex probes were then hybridized to the arrays at 45¡C in DIG Easy of human PBMC cultures. We identified 518 genes that were sig- Hyb (Roche Diagnostics, East Sussex, U.K.) for 72 h, and the membranes nificantly differentially expressed, many of which have not been were washed, as described (27). Complex probe hybridization intensities characterized previously as involved in lymphocyte activation. were quantified using a Storm phosphor imager (Amersham Biosciences)

and in-house spot-finding software. by guest on September 23, 2021 Clones from the normalized leukocyte library were sequenced using a Materials and Methods pCMV sport vector-derived primer (ATTTAGGTGACACTATAGAAGG), Generation of CD4ϩ and CD8ϩ T cell samples using ABI Prism 3700 capillary sequencers with BigDye Terminator chem- istry (Applied Biosystems, Foster City, CA). Clones used on the human Buffy coats from healthy donors were purchased from the South Thames sequence-verified arrays were sequenced prior to membrane spotting. Blood Transfusion Service (London, U.K.), and PBMC were obtained by All data were analyzed on the log (base 10) scale, and the hybridization density centrifugation over Ficoll. The majority of the adult United King- intensities on each individual array were normalized, by dividing the in- dom population has received BCG vaccination. The PBMC were cultured tensity for each gene by the geometric mean for that array. All genes that at 1 ϫ 106/ml in complete medium (RPMI 1640 supplemented with 5% AB were not statistically significantly different from background were removed serum) stimulated with live M. tuberculosis strain H37Rv at a multiplicity from the analysis: this was nearly 25% of the normalized leukocyte library of infection (MOI) of 1:1 (M. tuberculosis:monocyte) for 6 days. Concur- arrays and nearly 50% of the human sequence-verified arrays. There were rently, autologous monocyte-derived macrophages were obtained by ad- four hybridization intensity values for each T cell sample from each donor, herence, and cultured in complete medium for 6 days. On day 6, one-half due to the duplicate reverse-transcription step and duplicate spots on each of the stimulated PBMC was collected by centrifugation and restimulated grid, allowing a determination of the variability in measurement for each by incubation with freshly infected (1:1 MOI) autologous monocyte-de- ϩ ϩ gene within each donor. ANOVA (28) was performed using SAS software rived macrophages. Sixteen hours later, CD4 and CD8 T cells were (version 8; SAS Institute, Cary, NC), to compare the intensities, within a positively selected from both the restimulated and the nonrestimulated donor, between CD4ϩ and CD8ϩ T cell samples for each gene separately PBMC samples, using Dynabeads (Dynal Biotech, Wirral, U.K.). In pilot across all six donors. Those genes with an overall p value Ͻ0.05 are re- studies using M. bovis BCG stimulation in place of M. tuberculosis, the ϩ ported. The geometric mean of the hybridization intensity for each gene for average purity of CD4 -selected T cells as determined by flow cytometry the six donor samples was determined, and used to calculate the fold dif- was 98.7% CD4ϩ, 0.4% CD8ϩ, and for CD8ϩ-selected T cells 97.8% ϩ ϩ ϩ ϩ ference between CD4 and CD8 T cells, for both the early and late CD8 and 1.2% CD4 (n ϭ 3). The purity of the T cell samples used in stimulation conditions. the differential gene expression study was not tested because of safety considerations. Real-time qRT-PCR validation Total RNA was extracted using RNAzol B (Biogenesis, Dorset, U.K.). The cDNA equivalent of 50 ng of total RNA was used in each reaction. Typically, RNA samples were obtained from ϳ3Ð6 ϫ 107 cells, resulting RNA was reverse transcribed using Superscript II, in duplicate for each in an average total RNA yield of 15 ␮g. Any DNA contamination in the sample. The PCR were performed in SYBR Green PCR Master Mix (Ap- RNA samples was removed using DNase I (Ambion, Cambs, U.K.) before plied Biosystems), containing 400 nM of each primer. These reactions use in differential gene expression or real-time quantitative RT-PCR (qRT- were performed by an ABI Prism 7700 Sequence Detection System, with PCR) experiments. The quality of RNA samples was checked by gel a 10-min 95¡C denaturation step, followed by 40 cycles of 95¡C, 15 s; electrophoresis. 60¡C, 1 min. The fluorescence emitted from the dsDNA PCR product was Proliferation assay measured at the end of each cycle. At the end of all the cycles, a dissoci- ation analysis was performed to ensure the presence of only one specific PBMC from each donor were incubated at 1 ϫ 106/ml in round-bottom 96- PCR product. The starting mRNA copy number was determined by com- well plates in complete medium. They were stimulated with M. tuberculosis parison with a standard curve. In parallel, PCR were performed on reverse- PPD (Statens Serum Institut, Copenhagen, Denmark), or with live M. bovis transcriptase-negative control samples, to ensure there was no genomic BCG (GlaxoEvans strain). For the last 16 h of the 7-day assay, 1 ␮Ci of DNA contamination. For each sample and for each gene, the results were The Journal of Immunology 487 normalized by dividing by the calculated ␤-actin copy number, to allow comparisons between samples. The PCR primers for each gene were: IFN-induced protein 35 kDa, CAGCAGGGCCTAGCAGTCTTC and CCTCCGCACTGTTGGTGAAC; brefeldin A-inhibited guanine-nucleotide exchange protein 2, TCCCAT GATTGGCTATAAACG and GGTGGCAGTTAGCCGTTACAAG; furin, TCCTGGTACCCAGCCATCTG and CTGTGCACCAACCCAGCATC; KIAA0100 protein, AGCACCTGGTCTGTTCGCTG and CTTGTTGC CTCTCTGGCCCT; FLJ38577 fis, TGCTATGGACTCTCCTGCCG and GGAAACCGCAAGTGAGAGGG; hypothetical protein LOC285381, TGAACATTTGGAATGAGCCCAG and TGTCATGTTATGGACTT GCTTCC; ␤-actin, GAGCTACGAGCTGCCTGACG and GTAGTTTCT GGATGCCACAGGACT. In silico analysis of differentially expressed genes The predicted full-length protein sequences obtained from the National Center for Biotechnology Information were used in the transmembrane hidden Markov model program based at the Center for Biological Se- quence Analysis (29) to predict whether the genes are likely to be transmembrane proteins. Signal peptides were identified using SignalP to determine whether proteins were likely to be secreted (30). Their intracellular

location was predicted using PSORT II (31). Previously characterized do- Downloaded from mains and motifs were identiﬁed using by screening the PROSITE (32), FIGURE 2. Proportions of functional classes of differentially expressed Pfam (33), PRINTS (34), and SMART (35) databases. genes. PBMC were infected with live M. tuberculosis for 6 days, followed by 16-h restimulation (early activation genes), or were incubated for 7 days Results with M. tuberculosis (late activation genes) before harvest. CD4ϩ and Mycobacteria-speciﬁc responses of donors CD8ϩ T cells were isolated from the PBMC, and total RNA was reverse transcribed to make 33P-labeled cDNA probes that were used to screen the CD4ϩ and CD8ϩ T cell samples were prepared from six healthy

cDNA arrays. All statistically significantly (p Ͻ 0.05, n ϭ 6) differentially http://www.jimmunol.org/ donors. Initially, the memory T cell responsiveness to mycobac- expressed genes at either time point were classified according to biological -Unknown: this category includes in silico predicted hypothet ,ء .terial Ags was tested, by incubation of PBMC with either live M. function bovis BCG or PPD from M. tuberculosis. Each donor responded in ical genes, as well as identified expressed genes with no functional an Ag-specific manner, as measured by both proliferation assay information. and IFN-␥ secretion (Fig. 1), reflecting memory T cell activation due to prior BCG vaccination of the donors. The magnitude of the responses varied considerably between donors, which is typical in BCG-vaccinated subjects (36). Identification of differentially expressed genes

PBMC from the six donors were incubated with live M. tubercu- by guest on September 23, 2021 losis for either 7 days to identify late activation genes, or 6 days, followed by restimulation for a further 16 h with freshly infected autologous monocyte-derived macrophages, to identify early activation genes. An early activation time point could not be generated by 16-h stimulation directly ex vivo, as the mycobacteria-specific memory T cell frequency is too low without prior clonal expansion in vitro. Array-based experiments only test genes present on the arrays, and most arrays only contain previously characterized genes and ESTs. So as to comprehensively cover the gene transcript profile of the activated T cells, we chose to screen a normalized leukocyte cDNA library. An in-house GlaxoSmithKline human sequence-verified cDNA array was also screened. Many genes were found to be differentially expressed by CD4ϩ and CD8ϩ T cells at both the early and late activation time points. As expected, there was considerable variation between donors. ANOVA was performed to identify genes that were consistently differentially expressed by T cells from all six donors. Fold differences between expression levels in CD4ϩ and CD8ϩ T cells were calculated from the geometric means of the hybridization intensities from the six donors for both the early and the late activation time points. A total of 518 genes was at least 1.3-fold more highly expressed in either CD4ϩ or CD8ϩ T cells at at least one time point, with a p value Ͻ0.05 (see Supplemental tables I and II for full lists of these genes).4 Functional categories were ascribed to these genes using a combination of literature searches, the Locuslink, Online Mendelian Inheritance in Man and Unigene databases at the National Center for Biotechnology Information FIGURE 1. Donor T cell responses to mycobacterial Ags. PBMC were (37), and the GeneCards database (available at http:// cultured with live M. bovis BCG at 1:1 MOI (mycobacterium:monocyte) or with 10 ␮g/ml PPD for 6 days. A, T cell proliferation was assayed by [3H]thymidine incorporation. B, IFN-␥ secretion was assayed by ELISA. 4 The on-line version of this article contains supplemental material. 488 NOVEL MARKERS OF CD4ϩ AND CD8ϩ T CELL ACTIVATION bioinformatics.weizmann.ac.il/cards).5 Genes involved in many different biological and molecular processes were identified using this cDNA array library screening methodology (Fig. 2). There were similar numbers of genes involved in some biological pathways, such as the immune system, signal transduction, and gene transcription, which were more highly expressed in either CD4ϩ or CD8ϩ T cells (Table I). Other types of genes, in partic- ular those involved in protein biosynthesis, metabolism, and those encoding structural proteins, tended to be more highly expressed in CD4ϩ T cells, especially at the late activation time point. This may reflect the higher proliferative rate of CD4ϩ T cells. The overall pattern of differential gene expression was different at the early and late activation time points (Supplemental tables I and II).4

Validation of cDNA array screening technique To confirm that the gene expression changes were valid, real-time qRT-PCR experiments were performed on biologically independent samples. Six genes identified as being differentially expressed by the cDNA array hybridization experiments were analyzed, and Downloaded from the expression patterns as measured by qRT-PCR of these genes were similar to those observed by cDNA array (Fig. 3). These six genes include genes more highly expressed in either CD4ϩ or CD8ϩ T cells, at either early (furin, hypothetical protein LOC285381) or late (ARFGEF2 (brefeldin A-inhibited guanine FIGURE 3. Real-time qRT-PCR validation of cDNA array hybridiza- ϩ ϩ nucleotide-exchange protein 2), KIAA0100, FLJ38577 fis, IFN- tion method. Biologically independent CD4 and CD8 T cell samples http://www.jimmunol.org/ ϭ induced protein 35) activation time points, and either previously were prepared from different donors for qRT-PCR (n 5) as for the cDNA ϭ characterized genes or unknown, hypothetical genes. These results array hybridization experiments (n 6), and expression levels of the in- dicated genes were measured by SYBR Green incorporation during PCR. clearly demonstrate the biological as well as the technical repro- ϩ Results are shown as the geometric mean fold difference between CD8 ducibility of our system. and CD4ϩ T cells following appropriate normalization, with error bars depicting the 95% confidence intervals. Kinetics of gene expression variance The 518 genes that were differentially expressed were clustered ϩ into six groups, based on their expression patterns in CD4 and ϩ ϩ genes were more highly expressed in CD4 T cells (335) than in by guest on September 23, 2021 CD8 T cells at the early and late activation time points (supple- ϩ mental figure 1 and supplemental table III).4 Some genes were CD8 T cells (184). ϩ ϩ consistently more highly expressed in either CD4 or CD8 T Differential expression of immune system genes cells, whereas for other genes the differences were only observed at either the early or the late activation time point. Overall, more Genes involved in various aspects of immune function were differentially expressed between CD4ϩ and CD8ϩ T cells (Fig. 4). These included secreted molecules, such as IL-16, which was ex- 5 M. Rebhan, V. Chalifa-Caspi, J. Prilusky, and D. Lancet. 1997. GeneCards: Ency- pressed more highly by CD8ϩ T cells, and cyclophilin A, which clopedia for Genes, Proteins and Diseases. Weizmann Institute of Science. Bioin- ϩ formatics Unit and Genome Center. Rehovot. was more highly expressed by CD4 T cells. Various receptor

Table I. Numbers of genes more highly expressed in CD4ϩ and CD8ϩ T cells, at early and late activation time pointsa

Early Activation Genes Late Activation Genes

Higher expression Higher expression Higher expression Higher expression Category in CD4ϩ T cells in CD8ϩ T cells in CD4ϩ T cells in CD8ϩ T cells

Immune 8 4 19 15 Signal transduction 12 12 19 11 Transcription factor 16 6 15 13 Nuclear - other 7 3 11 8 Intracellular trafﬁcking 9 6 7 5 Proteolysis 3 3 13 3 Protein biosynthesis 16 16 27 5 Metabolism 14 10 35 7 Hormone 0 0 1 4 Structural 3 0 13 1 Solute transporter 2 2 1 0 Channel 1 2 2 0 Unknown 44 12 23 24 ESTs 11 1 8 11 Total 146 77 194 107

a Experimental details are given in Fig. 2. The Journal of Immunology 489 Downloaded from http://www.jimmunol.org/ by guest on September 23, 2021

FIGURE 4. Immune system genes differentially expressed between CD4ϩ and CD8ϩ T cells. A, Genes more highly expressed in CD8ϩ than CD4ϩ T cells (p Ͻ 0.05) in cDNA array hybridization experiments. B, Genes more highly expressed in CD4ϩ than CD8ϩ T cells (p Ͻ 0.05) in cDNA array hybridization experiments. Results are shown as the fold difference in the geometric mean of the hybridization intensities following array normalization.

molecules were also differentially expressed, including the IL-7R were differentially expressed, including the vav 1 oncogene, which expressed by CD8ϩ T cells and CD99 expressed by CD4ϩ T cells. was up-regulated in CD8ϩ T cells, and IL-2-inducible T cell kinase, Interestingly, the chemokine receptors CCR5, CX3CR1, and the which was up-regulated in CD4ϩ T cells. Three MHC class II genes, Duffy blood group Ag were more highly expressed in CD4ϩ T cells DQ␤,DR␤, and DO␤, were all preferentially up-regulated by the than CD8ϩ T cells. Several proteins involved in signal transduction activated CD4ϩ T cells (Fig. 4). Data from the HSV grids showed that 490 NOVEL MARKERS OF CD4ϩ AND CD8ϩ T CELL ACTIVATION expression levels of perforin and granzyme A were similar for CD4ϩ macrophage phagolysosome. In addition, the monocyte/macro- and CD8ϩ T cells, and high levels of IFN-␥ mRNA were observed in phage APCs may receive signals from live M. tuberculosis, such as all T cell samples (data not shown). via Toll-like receptors, some of which may be M. tuberculosis specific, and this may affect the activation of both CD4ϩ and In silico characterization of unknown genes CD8ϩ T cells. M. tuberculosis infection also leads to immunosup- Of the genes found to be differentially expressed, 19% have no known pression and immunomodulation, including down-regulation of function, and most of these are in silico predicted genes based on MHC class II and induction of IL-10 and TGF-␤, which inhibit ϩ ϩ analysis of the genomic DNA sequence. A further 6% of the differ- CD4 and CD8 T cell activation (22, 38, 39). In this way, we entially expressed clones are ESTs that do not map to known genes believe our stimulation conditions more closely reflect the activa- (Fig. 2). To gain some insight into their potential role in CD4ϩ and tion of both CD4ϩ and CD8ϩ T cells within an infected lung CD8ϩ T cells, computer-based prediction tools were used. A selection granuloma than other culture systems. We were able to identify a of genes exhibiting different expression profiles were chosen, and the large number of genes that were differentially expressed between results of these predictions are summarized in Table II. CD4ϩ and CD8ϩ T cells, and these came from a variety of dif- From these predictions, we characterized three of the nine pro- ferent gene families, underlining the fact that immune function is teins as having transmembrane domains, the hypothetical proteins the result of interplay of a large number of biological processes. FLJ10936, DKFZp564A022, and FLJ11240, which showed differ- The recall response of memory T cells to in vitro stimulation is ent expression patterns. All three of these proteins were predicted highly variable between human donors. ANOVA was performed to be located in the endoplasmic reticulum. Another two proteins, between paired CD4ϩ and CD8ϩ effector T cells from individual hypothetical protein MGC2508 and hypothetical protein donors, to identify those genes that were consistently differentially Downloaded from LOC51255, are likely to be transcription factors, as they have high expressed. The time point at which gene expression levels are mea- probabilities of nuclear location and contain predicted zinc finger sured is clearly important. We observed large differences in the genes domains. These genes have opposing expression patterns. The hy- that were more highly expressed at either an early (16-h restimulation) pothetical protein MGC8407 was more highly expressed in CD4ϩ or late (7-day stimulation) time point, although over one-half of the T cells at both time points, and this protein contains a tyrosine genes showed a consistent differential gene expression pattern. kinase domain and is therefore possibly involved in signal trans- A low dynamic range in the fold difference in expression level http://www.jimmunol.org/ duction. Three genes were identified that are likely to encode se- was observed for a combination of reasons. The T cells were stim- creted proteins. These are 37-kDa leucine-rich repeat protein and ulated in an Ag-specific manner, so the collected samples con- hypothetical protein MGC10993, which are more highly expressed tained a mixed population of M. tuberculosis-specific activated in CD4ϩ T cells, and hypothetical gene supported by AL449243, cells and nonactivated resting T cells. mRNA from the nonacti- which was expressed more highly by CD8ϩ T cells. vated T cells will have reduced the signal observed derived from the activated cells, thereby dampening the observed fold differ- Discussion ences. Therefore, a 1.3-fold overall observed fold difference in the This is the first comprehensive description of the differences in the mixed cell populations reflects much greater variations occurring gene expression profile of human activated CD4ϩ and CD8ϩ ef- within the responder cell populations. Also, the extent of the donor by guest on September 23, 2021 fector T cells following Ag-specific stimulation. In this study, we variability means that for each gene, some donors will have much used live M. tuberculosis to generate effector T cells in PBMC higher fold differences between CD4ϩ and CD8ϩ T cells, while cultures from BCG-vaccinated donors. Live mycobacteria are others will have lower differences. As both CD4ϩ and CD8ϩ T known to elicit stronger CD8ϩ T cell activation than dead myco- cells are taken from each donor, they are compared using the vari- bacteria or soluble Ags (19), possibly by creating pores in the ability within donors. Within donor variability will be smaller than vacuole membrane, allowing Ags to enter the cytoplasm and between donor variability; thus, we have a more statistically pow- thereby enter the MHC class I presentation pathway. CD4ϩ T cells erful experiment, which enables us to detect smaller fold differ- are also strongly activated in this system, due to the binding of ences as being significant than one would expect if one had to use mycobacterial peptides to MHC class II molecules in the infected between donor variability.

Table II. In silico characterization of novel genes

Putative Reference Gene Name Sequence Length Secreted Location Domains Expression Pattern

Hyp Prota FLJ11240 NM_018368 540 No ERb (66.7%) 9-Transmembrane actin Up CD4ϩ early binding 37-kDa Leucine-rich NM_005824 313 Yes Up CD4ϩ early repeatProt Hyp Prot MGC10993 NM_030577 311 Yes Up CD4ϩ early Hyp Prot LOC51255 NM_016494 158 No Nuclear (73.9%) Zinc finger C3HC4 type Up CD4ϩ late Hyp Prot MGC8407 NM_024046 501 No Nuclear (65.2%) Tyrosine kinase Up CD4ϩ early ϩ late Hyp Prot MGC2508 NM_024327 170 No Nuclear (91.3%) 4-Zinc finger C2H2 type Up CD8ϩ early Hyp Prot supported by NM_033318 107 Yes Up CD8ϩ early AL449243 Hyp Prot FLJ10936 NM_018279 336 No ER (66.7%) 6-Transmembrane DUF92 Up CD8ϩ late Hyp Prot NM_030954 258 No ER (44.4%) 3-Transmembrane ring_2_zinc Up CD8ϩ early DKFZp564A022 finger ϩ late

a Hyp Prot, hypothetical protein. b ER, endoplasmic reticulum. The Journal of Immunology 491

Some genes that were found to be up-regulated in either CD4ϩ one donor. Granulysin was more highly expressed by CD8ϩ than or CD8ϩ T cells have been described previously. For example, the CD4ϩ T cells, with a geometric fold difference of 1.25. The IFN-␥ NK cell-activating receptor NKG2D is known to be expressed on gene was highly expressed in all T cell samples analyzed, and there CD8ϩ T cells, where it pairs with the adaptor molecule DNAX was a close correlation between the hybridization intensity in activation protein 10 (DAP10), thereby providing costimulation CD4ϩ T cells and the donor’s response to PPD in PBMC cultures. via a B7/CD28-independent pathway (40). In the present study, A better understanding of the protective CD4ϩ and CD8ϩ T cell both NKG2D and DAP10 were identified as being significantly immune response to intracellular infection, derived from this type more highly expressed in CD8ϩ T cells. The ligands for NKG2D, of array-based analysis, will facilitate the development of corre- MHC class I chain-related gene A and B, are expressed on APCs lates of protection to be used as surrogate markers in field trials of in response to infection or stress, and it is known that M. tuber- new vaccines. New prophylactic and immunotherapeutic TB vac- culosis induces MHC class I chain-related gene A expression on cines are urgently needed, due to the limited efficacy of M. bovis dendritic cells and epithelial cells (41). Also, our finding of en- BCG in some areas and because the length of antibiotic therapy hanced IL-7R expression on CD8ϩ T cells fits in with current makes effective TB treatment difficult for many low income coun- understanding of the role of IL-7 in driving memory CD8ϩ, but not tries. New vaccine candidates are being tested in animal models, CD4ϩ, T cell proliferation, and survival (42). Although expression including live attenuated M. tuberculosis, recombinant vaccines, of MHC class II molecules is constitutive on professional APCs, DNA vaccines, and protein subunit vaccines (49), but correlates of the expression on T cells is tightly regulated (43). Clearly, CD4ϩ protection must be identified that demonstrate the induction of pro- T cells up-regulated MHC class II expression more than CD8ϩ T tective immunity against M. tuberculosis and therefore the efficacy ϩ cells in response to M. tuberculosis stimulation. CD4 T cells also of the candidate vaccine before large-scale clinical trials can be Downloaded from up-regulated their expression of the chemokine receptors CCR5, initiated (50). The production of IFN-␥ in whole blood cultures in CX3CR1, and the Duffy group Ag more than CD8ϩ T cells; this response to mycobacterial Ags has been proposed as a possible would lead to differential homing of the T cells during an in vivo correlate, and has been used to demonstrate the difference between infection. Different requirements for the secretion of biologically immune responses in Malawi and the United Kingdom induced active IL-16 by CD4ϩ and CD8ϩ T cells have been described following BCG vaccination (36). However, in vaccine studies in ϩ previously, with CD4 T cells displaying slower kinetics and a cattle and mice, protection is not related to the magnitude of the http://www.jimmunol.org/ dependence on costimulation (44). In our system, we observed IFN-␥ response, suggesting that additional mechanisms must be increased IL-16 mRNA expression in the CD8ϩ T cells (Fig. 4); present or absent to provide protection against disease (50, 51). hence, our stimulation conditions may favor IL-16 production by Any correlates of protection identified in the present culture sys- CD8ϩ over CD4ϩ T cells. tem with M. tuberculosis are also likely to be applicable to vaccine In contrast, other genes were found to be differentially ex- development for viral infections, where both CD4ϩ and CD8ϩ T pressed, which had not previously been identified as being ex- cell immune responses are required. pressed in CD4ϩ or CD8ϩ T cells. For example, the serine endo- It is likely that a combination of assays, measuring effector protease furin is a widely expressed transmembrane protein CD4ϩ and CD8ϩ T cell function, will be required in conjunction located in various cellular compartments, including the trans- with in vitro analysis of the killing of mycobacteria (52) to assess by guest on September 23, 2021 Golgi network and endosomal compartments. It is known to be candidate vaccine efficacy. Ideally, ELISA-based assays based on expressed in endocrine and neuroendocrine cells in the immature secreted proteins would be developed. At the present time, 64 of secretory granules, but not the mature granules (45); it is possible the unknown genes identified in this study have full-length protein that its higher expression in activated CD8ϩ T cells is important reference sequences. Of these, 5 are predicted to be secreted pro- for the maturation of the cytolytic granules in these cells. We also teins, according to SignalP. In time, the remaining 38 unknown observed the enhanced expression of an IFN-inducible leucine zip- genes and the 30 ESTs will become fully annotated, allowing the per protein, IFN-induced protein 35 in CD8ϩ T cells compared prediction of more secreted proteins. Currently, CTL activity is with CD4ϩ T cells. This protein forms a heterodimer with the AP-1 usually measured using the 51Cr release assay, which is labor in- transcription factor (46), and may thereby participate in the regulation tensive, fairly insensitive, and involves radioactivity, and is thus of cellular responses to the IFN-␥ secreted in response to M. tuber- unsuitable for large-scale field trials. The identification of 48 pre- culosis. Interestingly, this gene was found to be up-regulated in Th1 viously uncharacterized genes that are more highly expressed in cells compared with Th2 cells in a different study (5). activated effector CD8ϩ T cells than CD4ϩ T cells provides many We had expected to find increased expression of cytolytic ef- new leads in the discovery of biological pathways unique to CD8ϩ fector molecules, such as perforin or granzymes, in the CD8ϩ T T cells and involved in their cytotoxic response. Eighty-six un- cells. There are several possible explanations as to why this was characterized genes were found to be more highly expressed in not the case. The statistical analysis of the data from nearly 60,000 CD4ϩ T cells, and some of these may also play a crucial role in cDNA clones, but only six separate donors, is complex, and it may immunity to intracellular infection. be that our cutoff limits were too stringent to identify these genes. This study has identified many genes that are up-regulated in Alternatively, the mRNA levels may not be that different, but pro- CD4ϩ T cells following stimulation with M. tuberculosis,bycom- tein levels may be controlled via another mechanism. Another ex- parison with CD8ϩ T cells, and vice versa. Some of these genes planation is that M. tuberculosis also induces the activation of may be important during the immune response to TB, as they were cytolytic CD4ϩ T cells (21, 47). Indeed, perforin mRNA was as- found to be differentially expressed in the blood of active TB pa- sayed using the human sequence-verified arrays, and high levels tients compared with Mantoux-positive latently infected people in were detected in both CD4ϩ and CD8ϩ samples (data not shown). a parallel study.6 The G protein-coupled receptor EDG6 was more Interestingly, Ksp37, which has previously been described as a highly expressed in CD4ϩ T cells than CD8ϩ T cells following in protein selectively produced by cytotoxic lymphocytes (48), was actually more highly expressed by CD4ϩ than CD8ϩ T cells in this system. As expected, the CD8␣ gene was more highly expressed in 6 R. Mistry, C. L. Clayton, J. M. Cliff, N. Beyers, Y. Mohamed, P. A. Wilson, D. M. ϩ ϩ Wallace, P. van Helden, K. Duncan, and P. T. Lukey. Gene expression patterns in CD8 than CD4 T cells; however, this failed to reach signifi- blood discriminate individuals with active, cured, recurrent or latent tuberculosis. cance due to variability in the replicate hybridization intensities for Submitted for publication. 492 NOVEL MARKERS OF CD4ϩ AND CD8ϩ T CELL ACTIVATION vitro stimulation with M. tuberculosis, while the Src homology 3 persistent tuberculosis despite continued expression of interferon ␥ and nitric domain kinase-binding protein 1 was more highly expressed in oxide synthase 2. J. Exp. Med. 192:347. ϩ 13. Flynn, J. L., and J. Chan. 2001. Immunology of tuberculosis. Annu. Rev. Immu- CD8 T cells; both of these proteins are involved in lymphocyte nol. 19:93. signal transduction. Both of these genes are much more highly 14. Flynn, J. L., M. M. Goldstein, K. J. Triebold, B. Koller, and B. R. Bloom. 1992. expressed in the blood of active TB patients than in latently in- Major histocompatibility complex class I-restricted T cells are required for resistance to Mycobacterium tuberculosis infection. Proc. Natl. Acad. Sci. USA fected controls, and this probably reflects the greater mycobacterial 89:12013. burden in these people, leading to a greater circulation of activated 15. van Pinxteren, L. A., J. P. Cassidy, B. H. Smedegaard, E. M. Agger, and T cells in the periphery. Conversely, some genes were more highly P. Andersen. 2000. Control of latent Mycobacterium tuberculosis infection is dependent on CD8 T cells. Eur. J. Immunol. 30:3689. expressed by latently infected people than in active TB patients, 16. Smith, S. M., R. Brookes, M. R. Klein, A. S. Malin, P. T. Lukey, A. S. King, and these genes may therefore be contributing to the immune pro- G. S. Ogg, A. V. Hill, and H. M. Dockrell. 2000. Human CD8ϩ CTL specific for tection. These include the oxysterol-binding protein 10, which was the mycobacterial major secreted antigen 85A. J. Immunol. 165:7088. ϩ 17. Lalvani, A., R. Brookes, R. J. Wilkinson, A. S. Malin, A. A. Pathan, P. Andersen, more highly expressed by CD4 T cells, and the signal transduc- H. Dockrell, G. Pasvol, and A. V. Hill. 1998. Human cytolytic and interferon tion molecule DAP10 and the lysosomal sialidase Neu1, which ␥-secreting CD8ϩ T lymphocytes specific for Mycobacterium tuberculosis. Proc. were more highly expressed by CD8ϩ T cells. Natl. Acad. Sci. USA 95:270. 18. Smith, S. M., A. S. Malin, T. Pauline, Lukey, S. E. Atkinson, J. Content, We now propose to develop some of these genes into surrogate K. Huygen, and H. M. Dockrell. 1999. Characterization of human Mycobacte- ϩ ϩ markers of CD4 and CD8 T cell activation. Likely candidates rium bovis bacille Calmette-Guerin-reactive CD8ϩ T cells. Infect. Immun. include those that were most highly differentially expressed, such 67:5223. 19. Turner, J., and H. M. Dockrell. 1996. Stimulation of human peripheral blood as the NKG2D receptor and its signal transduction partner DAP10. mononuclear cells with live Mycobacterium bovis BCG activates cytolytic CD8ϩ The extent of expression of these proteins in vitro in response to T cells in vitro. Immunology 87:339. Downloaded from mycobacterial stimulation might reflect the magnitude of the mem- 20. Stenger, S., D. A. Hanson, R. Teitelbaum, P. Dewan, K. R. Niazi, C. J. Froelich, ϩ T. Ganz, S. Thoma-Uszynski, A. Melian, C. Bogdan, et al. 1998. An antimicro- ory CD8 T cell response induced following vaccination. Also, bial activity of cytolytic T cells mediated by granulysin. Science 282:121. secreted proteins such as IL-16 and KSP37, and the putatively 21. Lorgat, F., M. M. Keraan, P. T. Lukey, and S. R. Ress. 1992. Evidence for in vivo secreted hypothetical proteins MGC10993 and AL449243 will be generation of cytotoxic T cells: PPD-stimulated lymphocytes from tuberculous pleural effusions demonstrate enhanced cytotoxicity with accelerated kinetics of analyzed to determine whether their production following in vitro induction. Am. Rev. Respir. Dis. 145:418. stimulation correlates with the induction of protective immunity by 22. Rojas, R. E., K. N. Balaji, A. Subramanian, and W. H. Boom. 1999. Regulation http://www.jimmunol.org/ ϩ ϩ ϩ vaccination. These surrogate markers might then be used in field of human CD4 ␣␤ T-cell-receptor-positive (TCR ) and ␥␦ TCR T-cell responses to Mycobacterium tuberculosis by interleukin-10 and transforming trials of new immunotherapeutic and prophylactic vaccines. growth factor ␤. Infect. Immun. 67:6461. 23. Hirsch, C. S., Z. Toossi, C. Othieno, J. L. Johnson, S. K. Schwander, Acknowledgments S. Robertson, R. S. Wallis, K. Edmonds, A. Okwera, R. Mugerwa, et al. 1999. Depressed T-cell interferon-␥ responses in pulmonary tuberculosis: analysis of We are grateful to Debbie Smith, Joanne Cox, and Paul Murdock for sci- underlying mechanisms and modulation with therapy. J. Infect. Dis. 180:2069. entific advice, and to Sara Atkinson for providing the M. tuberculosis.We 24. Smith, S. M., M. R. Klein, A. S. Malin, J. Sillah, K. P. McAdam, and thank the Transcriptome Analysis and Quantitative Expression Department H. M. Dockrell. 2002. Decreased IFN-␥ and increased IL-4 production by human ϩ and the Core Sequencing group at GlaxoSmithKline Medicines Research CD8 T cells in response to Mycobacterium tuberculosis in tuberculosis patients. Tuberculosis 82:7.

Centre (Stevenage, U.K.). by guest on September 23, 2021 25. Smith, S. M., M. R. Klein, A. S. Malin, J. Sillah, K. Huygen, P. Andersen, K. P. McAdam, and H. M. Dockrell. 2000. Human CD8ϩ T cells specific for References Mycobacterium tuberculosis secreted antigens in tuberculosis patients and 1. Croft, M., L. Carter, S. L. Swain, and R. W. Dutton. 1994. Generation of polar- healthy BCG-vaccinated controls in The Gambia. Infect. Immun. 68:7144. ized antigen-specific CD8 effector populations: reciprocal action of interleukin 26. Kittler, J. T., E. V. Grigorenko, C. Clayton, S. Y. Zhuang, S. C. Bundey, (IL)-4 and IL-12 in promoting type 2 versus type 1 cytokine profiles. J. Exp. Med. M. M. Trower, D. Wallace, R. Hampson, and S. Deadwyler. 2000. Large-scale 180:1715. analysis of gene expression changes during acute and chronic exposure to ␦9- 2. Teague, T. K., D. Hildeman, R. M. Kedl, T. Mitchell, W. Rees, B. C. Schaefer, THC in rats. Physiol. Genomics 3:175. J. Bender, J. Kappler, and P. Marrack. 1999. Activation changes the spectrum but 27. Cox, J. M., C. L. Clayton, T. Tomita, D. M. Wallace, P. A. Robinson, and not the diversity of genes expressed by T cells. Proc. Natl. Acad. Sci. USA J. E. Crabtree. 2001. cDNA array analysis of cag pathogenicity island-associated 96:12691. Helicobacter pylori epithelial cell response genes. Infect. Immun. 69:6970. 3. Gavin, M. A., S. R. Clarke, E. Negrou, A. Gallegos, and A. Rudensky. 2002. 28. Snedecor, G. W., and W. G. Cochran. 1989. Statistical Methods. Iowa State Homeostasis and anergy of CD4ϩCD25ϩ suppressor T cells in vivo. Nat. Im- University Press, Ames. munol. 3:33. 29. Moller, S., M. D. Croning, and R. Apweiler. 2001. Evaluation of methods for the 4. McHugh, R. S., M. J. Whitters, C. A. Piccirillo, D. A. Young, E. M. Shevach, prediction of membrane spanning regions. Bioinformatics 17:646. ϩ ϩ M. Collins, and M. C. Byrne. 2002. CD4 CD25 immunoregulatory T cells: 30. Nielsen, H., J. Engelbrecht, S. Brunak, and G. von Heijne. 1997. Identification of gene expression analysis reveals a functional role for the glucocorticoid-induced prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. TNF receptor. Immunity 16:311. Protein Eng. 10:1. 5. Rogge, L., E. Bianchi, M. Biffi, E. Bono, S. Y. Chang, H. Alexander, C. Santini, 31. Nakai, K., and M. Kanehisa. 1992. A knowledge base for predicting protein G. Ferrari, L. Sinigaglia, M. Seiler, et al. 2000. Transcript imaging of the devel- localization sites in eukaryotic cells. Genomics 14:897. opment of human T helper cells using oligonucleotide arrays. Nat. Genet. 25:96. 32. Falquet, L., M. Pagni, P. Bucher, N. Hulo, C. J. Sigrist, K. Hofmann, and 6. Chtanova, T., R. A. Kemp, A. P. Sutherland, F. Ronchese, and C. R. Mackay. A. Bairoch. 2002. The PROSITE database, its status in 2002. Nucleic Acids Res. 2001. Gene microarrays reveal extensive differential gene expression in both 30:235. CD4ϩ and CD8ϩ type 1 and type 2 T cells. J. Immunol. 167:3057. 33. Bateman, A., E. Birney, L. Cerruti, R. Durbin, L. Etwiller, S. R. Eddy, 7. Beadling, C., and K. A. Smith. 2002. DNA array analysis of interleukin-2-reg- S. Griffiths-Jones, K. L. Howe, M. Marshall, and E. L. Sonnhammer. 2002. The ulated immediate/early genes. Med. Immunol. 1:2. Pfam protein families database. Nucleic Acids Res. 30:276. 8. Diehn, M., A. A. Alizadeh, O. J. Rando, C. L. Liu, K. Stankunas, D. Botstein, G. R. Crabtree, and P. O. Brown. 2002. Genomic expression programs and the 34. Attwood, T. K., P. Bradley, D. R. Flower, A. Gaulton, N. Maudling, integration of the CD28 costimulatory signal in T cell activation. Proc. Natl. A. L. Mitchell, G. Moulton, A. Nordle, K. Paine, P. Taylor, et al. 2003. PRINTS Acad. Sci. USA 99:11796. and its automatic supplement, prePRINTS. Nucleic Acids Res. 31:400. 9. Riley, J. L., M. Mao, S. Kobayashi, M. Biery, J. Burchard, G. Cavet, 35. Letunic, I., L. Goodstadt, N. J. Dickens, T. Doerks, J. Schultz, R. Mott, B. P. Gregson, C. H. June, and P. S. Linsley. 2002. Modulation of TCR-induced F. Ciccarelli, R. R. Copley, C. P. Ponting, and P. Bork. 2002. Recent improve- transcriptional profiles by ligation of CD28, ICOS, and CTLA-4 receptors. Proc. ments to the SMART domain-based sequence annotation resource. Nucleic Acids Natl. Acad. Sci. USA 99:11790. Res. 30:242. 10. Dye, C., S. Scheele, P. Dolin, V. Pathania, and M. C. Raviglione. 1999. Con- 36. Black, G. F., R. E. Weir, S. Floyd, L. Bliss, D. K. Warndorff, A. C. Crampin, sensus statement: global burden of tuberculosis: estimated incidence, prevalence, B. Ngwira, L. Sichali, B. Nazareth, J. M. Blackwell, et al. 2002. BCG-induced and mortality by country: WHO Global Surveillance and Monitoring Project. increase in interferon-␥ response to mycobacterial antigens and efficacy of BCG J. Am. Med. Assoc. 282:677. vaccination in Malawi and the UK: two randomized controlled studies. Lancet 11. Fine, P. E. 1995. Variation in protection by BCG: implications of and for het- 359:1393. erologous immunity. Lancet 346:1339. 37. Pruitt, K. D., K. S. Katz, H. Sicotte, and D. R. Maglott. 2000. Introducing RefSeq 12. Scanga, C. A., V. P. Mohan, K. Yu, H. Joseph, K. Tanaka, J. Chan, and and LocusLink: curated human genome resources at the NCBI. Trends Genet. J. L. Flynn. 2000. Depletion of CD4ϩ T cells causes reactivation of murine 16:44. The Journal of Immunology 493

38. Turner, J., M. Gonzalez-Juarrero, D. L. Ellis, R. J. Basaraba, A. Kipnis, 46. Wang, X., L. M. Johansen, H. J. Tae, and E. J. Taparowsky. 1996. IFP 35 forms I. M. Orme, and A. M. Cooper. 2002. In vivo IL-10 production reactivates complexes with B-ATF, a member of the AP1 family of transcription factors. chronic pulmonary tuberculosis in C57BL/6 mice. J. Immunol. 169:6343. Biochem. Biophys. Res. Commun. 229:316. 39. Olobo, J. O., M. Geletu, A. Demissie, T. Eguale, K. Hiwot, G. Aderaye, and 47. Tan, J. S., D. H. Canaday, W. H. Boom, K. N. Balaji, S. K. Schwander, and S. Britton. 2001. Circulating TNF-␣, TGF-␤, and IL-10 in tuberculosis patients E. A. Rich. 1997. Human alveolar T lymphocyte responses to Mycobacterium and healthy contacts. Scand. J. Immunol. 53:85. tuberculosis antigens: role for CD4ϩ and CD8ϩ cytotoxic T cells and relative 40. Groh, V., R. Rhinehart, J. Randolph-Habecker, M. S. Topp, S. R. Riddell, and resistance of alveolar macrophages to lysis. J. Immunol. 159:290. T. Spies. 2001. Costimulation of CD8␣␤ T cells by NKG2D via engagement by 48. Ogawa, K., K. Tanaka, A. Ishii, Y. Nakamura, S. Kondo, K. Sugamura, MIC induced on virus-infected cells. Nat. Immunol. 2:255. S. Takano, M. Nakamura, and K. Nagata. 2001. A novel serum protein that is 41. Das, H., V. Groh, C. Kuijl, M. Sugita, C. T. Morita, T. Spies, and J. F. Bukowski. selectively produced by cytotoxic lymphocytes. J. Immunol. 166:6404. 2001. MICA engagement by human V␥2V␦2 T cells enhances their antigen- 49. Orme, I. M. 2001. The search for new vaccines against tuberculosis. J. Leukocyte dependent effector function. Immunity 15:83. Biol. 70:1. 42. Prlic, M., L. Lefrancois, and S. C. Jameson. 2002. Multiple choices: regulation of 50. Ellner, J. J., C. S. Hirsch, and C. C. Whalen. 2000. Correlates of protective memory CD8 T cell generation and homeostasis by interleukin (IL)-7 and IL-15. immunity to Mycobacterium tuberculosis in humans. Clin. Infect. Dis. 30 (Suppl. J. Exp. Med. 195:F49. 3):S279. 43. Barnaba, V., C. Watts, M. de Boer, P. Lane, and A. Lanzavecchia. 1994. Pro- 51. Skinner, M. A., A. J. Ramsay, G. S. Buchan, D. L. Keen, C. Ranasinghe, fessional presentation of antigen by activated human T cells. Eur. J. Immunol. L. Slobbe, D. M. Collins, G. W. de Lisle, and B. M. Buddle. 2003. A DNA 24:71. prime-live vaccine boost strategy in mice can augment IFN-␥ responses to my- 44. Wu, D. M., Y. Zhang, N. A. Parada, H. Kornfeld, J. Nicoll, D. M. Center, and cobacterial antigens but does not increase the protective efﬁcacy of two attenu- W. W. Cruikshank. 1999. Processing and release of IL-16 from CD4ϩ but not ated strains of Mycobacterium bovis against bovine tuberculosis. Immunology CD8ϩ T cells is activation dependent. J. Immunol. 162:1287. 108:548. 45. Dittie, A. S., L. Thomas, G. Thomas, and S. A. Tooze. 1997. Interaction of furin 52. Kampmann, B., P. O. Gaora, V. A. Snewin, M. P. Gares, D. B. Young, and in immature secretory granules from neuroendocrine cells with the AP-1 adaptor M. Levin. 2000. Evaluation of human antimycobacterial immunity using recom- complex is modulated by casein kinase II phosphorylation. EMBO J. 16:4859. binant reporter mycobacteria. J. Infect. Dis. 182:895. Downloaded from http://www.jimmunol.org/ by guest on September 23, 2021