Genomic expression profiling of TNF-α-treated BDC2.5 diabetogenic CD4⁺ T cells

Abstract

TNF-α plays an important role in immune regulation, inflammation, and autoimmunity. Chronic TNF exposure has been shown to down-modulate T cell responses. In a mouse T cell hybridoma model, TNF attenuated T cell receptor (TCR) signaling. We have confirmed that chronic TNF and anti-TNF exposure suppressed and increased T cell responses, respectively. In adult TCR (BDC2.5) transgenic nonobese diabetic mice, DNA microarray analysis of global gene expression in BDC2.5 CD4⁺ T cells in response to chronic TNF or anti-TNF exposure showed that genes involved in functional categories including T cell signaling, cell cycle, proliferation, ubiquitination, cytokine synthesis, calcium signaling, and apoptosis were modulated. Genes such as ubiquitin family genes, cytokine inducible Src homology 2-containing genes, cyclin-dependent kinase inhibitors p21, p57, calmodulin family genes (calmodulin-1, -2, and -3) and calcium channel voltage-dependent, N type α1B subunit (CaV2.2) were induced by TNF, whereas Vav2, Rho GTPase-activating protein, calcium channel voltage-dependent, L type α1C subunit (CaV1.2), IL-1 receptor-associated kinase-1 and -2, and IL enhancer binding factor 3 were reduced by TNF. Genes such as CaV1.2 and proliferating cell nuclear antigen, repressed by TNF, were induced by anti-TNF treatment. Further, we showed that chronic TNF exposure impaired NF-κB and adaptor protein 1 transactivation activity, leading to T cell unresponsiveness. Thus, our results present a detailed picture of transcriptional programs affected by chronic TNF exposure and provide candidate target genes that may function to mediate TNF-induced T cell unresponsiveness.

Tumor necrosis factor (TNF) is a multipotent inflammatory cytokine that plays an important role in the pathogenesis of diseases caused by autoimmune T cell responses, not only as a stimulatory mediator (1–4), but also as an inhibitor in certain settings (5–7). Excess TNF has been associated with rheumatoid arthritis (RA) and other forms of inflammatory autoimmune diseases. Infliximab was the first anti-TNF blocker used to treat RA, juvenile RA, ankylosing spondylitis, and psoriatic arthritis. Humanized anti-TNF antibodies have been reported to be of benefit in the treatment of the above diseases as well as psoriasis and Crohn's disease. Overall, it is evident that anti-TNF therapy results in a moderate to dramatic decrease in RA symptoms, and in some cases a near complete remission, although the disease recurs relatively promptly after cessation of anti-TNF therapy (3, 8).

Signals through the TCR are propagated by several molecules. TCR ligation results in phosphorylation of immunoreceptor tyrosine-based activation motifs (ITAMs) in CD3 polypeptides through the Src kinases lck and fyn (9). The phosphorylated ITAMs recruit the tyrosine kinase ζ-associated protein of 70 kDa, (ZAP-70), which is then phosphorylated and activated by a Src family kinase. This, in turn, leads to phosphorylation by ZAP-70 of the transmembrane adapter protein linker for activation of T cells (LAT) (10). Phosphorylated LAT induces the formation of a multimolecular complex including VAV, NCK, GADS, phospholipase Cγ (PLCγ), PI3K, CBL, SLAP-130, and GRB2/SOS/RAS (11). The end result of the entire signaling cascade is to trigger specific key events: a sustained rise in intracellular Ca⁺² and sustained activation of the Ras/MAPK, IKK, and Akt pathways, which regulate activation of several transcription factors, including NF-AT, adaptor protein 1, and NF-κB (12).

Several lines of evidence suggest that the presence of TNF affects T cell responsiveness and TCR signal transduction. These studies have documented a decrease in T cell proliferation, cytokine production and calcium flux in normal T cells after chronic exposure to TNF in both in vitro and in vivo studies (6, 13–15). In vivo the suppressive effect of endogenous TNF could be inhibited by anti-TNF mAb injections in mouse models (14) and in patients with RA (2). As demonstrated by Isomaki et al. (16) T cell hybridomas cultured in vitro in the presence of nontoxic levels of TNF have decreased phosphorylation in the TCRζ chain, CD3ε, and ZAP 70. However, TCRζ reconstitution failed to restore T cell responses after long-term TNF treatment, indicating that other mechanisms are also likely to be involved (17).

We have examined the effects of TNF and anti-TNF in BDC2.5 TCR transgenic (tg) mice. We used cDNA microarrays to analyze global transcriptional alterations resulting from TNF treatment on TCR signaling pathways. We have identified several genes relevant for T cell activation pathways that are up-regulated, such as cytotoxic T lymphocyte antigen-4 (CTLA-4), lymphocyte-specific protein tyrosine kinase (Lck), RAS p21 protein activator 1, and, calmodulin-1, -2, and -3 in TNF-treated animals, whereas Vav2 and PI3K were down-regulated in the TCR signaling pathway. Furthermore, some important genes involved in cytokine inducible Src homology 2 (SH2)-containing protein (CIS), calcium channel, and protein ubiquitination pathways were up-regulated and will be discussed below. These findings provide a better understanding of the mechanisms by which TNF causes T cell unresponsiveness. These results may also be relevant for the development of drugs for autoimmune disease therapy in the future.

Results

Effect of Chronic TNF Exposure on Activated T Cells In Vitro in BDC2.5 Tg Mice.

Previously, we have reported that chronic exposure to TNF-α resulted in a decrease in T cell proliferation, cytokine production, and calcium flux in HNT TCR tg T cells (14). To study this observation in a diabetic animal model we used BDC2.5 TCR tg mice after chronic exposure to TNF. Both in vitro and in vivo analyses were performed according to the protocol outlined by Cope et al. (14). Repeated exposure of BDC2.5 tg T cells to TNF for up to 11 days led to marked suppression of T cell responses after restimulation with 1 μg/ml and 0.1 μg/ml of 1047−7 peptide plus fresh splenic antigen-presenting cells (APCs) [supporting information (SI) Fig. S1 A and B]. Suppressive effects in vitro were dose dependent; concentrations between 2 and 10 ng/ml TNF appeared sufficient to suppress T cell responses. IL-2 levels in this proliferation assay decreased, indicating that TNF inhibits IL-2 production in a dose-dependent manner (Fig. S1C).

Prolonged TNF Exposure in Vivo Suppresses T Cell Responses in BDC2.5 Tg Mice.

BDC 2.5 tg mice (8–12 weeks old) were treated with alternate day injections of 3 μg TNF or PBS i.p. After 3 weeks of treatment with this dose of TNF, the mice had no change in cell numbers in LNs, nor any clinical symptoms such as weight loss (data not shown). However, in pooled LN and splenic T cells, chronic TNF exposure in vivo suppressed 1047-7 peptide-specific T cell responses 64% and 52%, respectively (Fig. 1 A and B).

Fig. 1.

The effect of 3-week TNF and anti-TNF exposure in vivo on T cell proliferation in BDC2.5 tg mice. Eight- to 12-week-old NOD.BDC2.5 tg mice were injected i.p. with PBS or 3 μg of murine TNF or 100 μg of anti-TNF on alternate days for 3 weeks before study. The proliferative responses of LNs (A and C) or splenic (sp) T cells (B and D) from a pair of tg littermates treated with PBS or control IgG (gray bars) or TNF or anti-TNF (black bars) were compared after stimulation with various concentration of 1047-7 peptide as indicated. Data are expressed as mean cpm ± SEM from triplicates.

Prolonged Anti-TNF Exposure in Vivo Increases T Cell Responses in BDC2.5 Tg Mice.

Pairs of tg littermates were injected with 100 μg of anti-TNF or control hamster Ig every other day for 3 weeks, and proliferative responses to 1047-7 peptide were determined. Increased T cell responses were observed after stimulation with a wide range of peptide concentrations (0.01, 0.1, 1, 2, and 5 μg/ml) (Fig. 1 C and D). These results indicate that regulation of T cell function by endogenous TNF in vivo can be reversed by anti-TNF.

Overview of Gene Expression Patterns.

To elucidate the mechanisms underlying the suppressive effect of TNF on T cells, we extracted RNA from pancreatic draining lymph node (LN) CD4⁺ T cells that were isolated from NOD.BDC 2.5 mice treated in vivo with TNF or PBS and subjected to mimotope stimulation for 24 h. RNA was amplified, reverse-transcribed, and hybridized in mouse cDNA microarrays. The resulting data were analyzed by the GABRIEL (Genetic Analysis By Rules Incorporating Expert Logic) system, a knowledge-based system of computer algorithms (18). Using a pattern-based rule that identifies genes whose mean intensity (over six hybridizations) value in green to red fluorescence intensity ratios exceeds a defined threshold, we found 2,173 genes that were up-regulated [false discovery rate (FDR) = 0.021] and 1,326 genes that were down-regulated (FDR = 0.009) at least 2-fold by TNF (Fig. 2).

Fig. 2.

Global overview of the gene expression program of BDC CD4⁺ T cells from mice exposed to TNF. Array elements that were induced (red) or repressed (green) >2-fold using GABRIEL pattern-based rule are shown. A total of 2,173 and 1,326 genes were up- and down-regulated, respectively, in the TNF-treated animals. The data were visualized and analyzed with Tree View. Rows represent genes (unique cDNA elements), and columns represent six repeated experiments. The detailed information for each gene is listed in Dataset S1.

Genes Altered in the T Cell Activation Pathway.

Given the fact that TNF has a broad range of effects on many cell types, it is conceivable that there are multiple transcriptional changes caused by TNF. Therefore, we analyzed several signaling pathways and summarize the results in Tables 1 and 2. We found a number of genes whose products are involved in T cell signal transduction that were repressed in TNF-treated T cells, such as phosphatidylinositol 3-kinase α polypeptide (PIK3ca), Vav2 (Table 2); and lymphocyte protein tyrosine kinase (Lck), CD28, ras 21 protein activator-1 (Rasa1) and -2 (Rasa2) that contains Ras GTPase activator activity, nuclear factor of kappa light chain gene enhancer in B cells inhibitor, α (IκBα), and calmodulin 1, -2, and –3, which were induced in TNF-treated mice (Table 1). Interestingly, CTLA-4 (CD152), reported to shut down T cell activation, culminating in T cell death or the induction of anergy (19), was induced in TNF-treated mice (Table 1).

Table 1.

Genes up-regulated by TNF-α in NOD BDC2.5 T cells

Pathway	Fold of change
T cell activation pathway
IκBα	2.05
CD28 molecule	2.14
Calmodulin 3	2.17
Lymphocyte-specific protein tyrosine kinase	2.24
RAS p21 protein activator 1	2.24
MAPK8	2.3
Calmodulin 2	2.38
Cytotoxic T lymphocyte-associated protein 4	2.54
NF-κB (p49/p100)	2.54
Neuroblastoma RAS viral (v-ras) oncogene homolog	3.63
Calmodulin 1	4.07
NF-κB (p105)	4.46
Toll-like receptor
Nuclear factor of κ light polypeptide gene enhancer in B cells inhibitor, α	2.05
MAPK 8	2.30
Nuclear factor of κ light polypeptide gene enhancer in B cells 2 (p49/p100)	2.55
MAPK14	3.28
Nuclear factor of κ light polypeptide gene enhancer in B cells 1 (p105)	4.46
JakStat pathway
Signal transducer and activator of transcription 5B	2.19
v-akt murine thymoma viral oncogene homolog 2	2.20
Cyclin-dependent kinase inhibitor 1A (p21, Cip1)	2.21
Signal transducer and activator of transcription 3	2.31
Protein tyrosine phosphatase, nonreceptor type 11 (Noonan syndrome 1)	2.79
Cytokine inducible SH2-containing protein	3.02
Neuroblastoma RAS viral (v-ras) oncogene homolog	3.63
Calcium pathway
Histone deacetylase 2	2.02
RAP1A, member of RAS oncogene family	2.10
Inositol 1,4,5-triphosphate receptor, type 3	2.13
Calmodulin 3 (phosphorylase kinase, delta)	2.17
Tropomyosin 1 (α)	2.20
RAP1B, member of RAS oncogene family	2.23
Protein kinase, cAMP-dependent, catalytic, α	2.29
Calmodulin 2 (phosphorylase kinase, delta)	2.38
ATPase, Ca++ transporting, cardiac muscle, slow twitch 2	2.80
Protein kinase, cAMP-dependent, type I, α (tissue-specific extinguisher 1)	3.03
Tropomyosin 3	3.83
Protein kinase, cAMP-dependent, regulatory, type II, α	3.92
Calmodulin 1 (phosphorylase kinase, delta)	4.07
Myosin, light polypeptide 6, alkali, smooth muscle and nonmuscle	4.52
Calcium channel, voltage-dependent, N type, α1B subunit	4.57
Protein ubiquitination pathway
Proteasome (prosome, macropain) 26S subunit, non-ATPase, 5	2.00
Ubiquitin specific peptidase 16	2.07
Ubiquitin-conjugating enzyme E2R 2	2.21
Heat shock protein 90-kDa α (cytosolic), class A member 1	2.22
Ubiquitin-conjugating enzyme E2B (RAD6 homolog)	2.22
Ubiquitin specific peptidase 14 (tRNA-guanine transglycosylase)	2.23
Ubiquitin specific peptidase 10	2.36
Ubiquitin-conjugating enzyme E2G 2 (UBC7 homolog, yeast)	2.36
WW domain containing E3 ubiquitin protein ligase 2	2.44
Ubiquitin-specific peptidase 12	2.49
Ubiquitin B	4.23

Table 2.

Genes down-regulated by TNF-α in NOD BDC2.5 T cells

Pathway	Fold of change
T cell activation pathway
Vav2 oncogene	−3.24
Phosphoinositide-3-kinase, catalytic, a polypeptide	−2.2
Toll-like receptor
IL1 receptor-associated kinase 1	−3.02
IL-1 receptor-associated kinase 2	−2.19
JakStat pathway
Phosphoinositide-3-kinase, catalytic, α polypeptide	−2.18
Calcium pathway
Calcium channel, voltage-dependent, L type, α 1C subunit	−2.24
Protein ubiquitination pathway
Ubiquitin carboxyl-terminal hydrolase L5	−2.54
Ubiquitin-conjugating enzyme E2C	−2.47
Baculoviral IAP repeat-containing 6 (apollon)	−2.37
Ubiquitin-specific peptidase 47	−2.37
Neural precursor cell expressed, developmentally down-regulated 4	−2.23
Ubiquitin-conjugating enzyme E2E 1 (UBC4/5 homolog, yeast)	−2.01

Genes Altered in the Toll-Like Receptor Pathway.

Chronic repeated exposure to TNF reduced expression of IL-1 receptor-associated kinase-1 and -2 (IRAK-1 and -2) and IL enhancer binding factor 3 (IEBF3) (Table 2). In contrast, MAPK-8 and MAPK-14 were induced by TNF treatment, indicating that TNF, reported to function as a TLR agonist, has diverse effects on innate immune responses.

Genes Altered in the Jak–Stat Pathway.

Activated T cells enter the cell cycle, proliferate and can eventually become memory cells. We found that chronic TNF exposure induced cyclin-dependent kinase inhibitors p21 and p57, whereas anti-TNF treatment repressed p21 gene expression (Table 1). Another family of suppressor of cytokine signaling (SOCS) genes, such as cytokine-inducible SH2-containing protein (CISH), induced by signal transducer and activator of transcription (STATs), which inhibit signaling through various negative-feedback mechanisms, were markedly induced in TNF-treated mice (Table 1). We would expect that up-regulation of CISH would normally result in down-regulation of STAT3 and STAT5. However, the v-akt murine thymoma viral oncogene homolog 2 (AKT2), STAT3, and STAT5B factor were increased after chronic TNF exposure.

Genes Altered in the Protein Ubiquitination Pathway.

Many of the genes whose products are involved in the ubiquitin-mediated protein degradation pathway were induced in TNF-treated mice. These genes include ubiquitin B and C, ubiquitin-specific protease-10, -12, -14, and -16, ubiquitin-conjugating enzymes E2D3, E2G2, and E2R2, ubiquitin-activating enzyme, and WW domain-containing E3 ubiquitin protein ligase 2 (Wwp2) (Table 1). Further, ubiquitin ligase E3A was repressed in anti-TNF-treated mice (data not shown).

Genes Altered in the Calcium Signaling Pathway.

Among the genes regulated by chronic TNF exposure in the microarray analysis, the most dramatic are calcium-related genes. Prominent among this group of induced genes were calmodulin-1, -2, and -3 and CaV2.2, whereas CaV1.2 was repressed after 3 weeks TNF exposure (Tables 1 and 2).

Quantitative PCR Validation.

To confirm the results obtained from the microarray analysis, we performed quantitative PCR analysis for five genes (CTLA-4, CISH, Wwp2, calmodulin-1, and CaV1.2). As shown in Fig. 3 and Tables 1 and 2, expression patterns for all of the tested genes correlated strongly with the microarray data.

Fig. 3.

Some of the up- or down-regulated genes were validated by quantitative real-time RT-PCR. The data were expressed as expression intensity relative to PBS-treated mice (n = 10). CTLA-4, calmodulin (CALM), CISH, Wwp2, calcium channel, voltage-dependent, N type, α1B subunit (Cncna1b), and PI3Kca are shown. Data are expressed as mean ± STD from triplicates.

Chronic TNF Treatment Reduced AP-1 and NF-κB activity.

Because we observed that NF-κB and JNK gene expression was induced after chronic TNF treatment at the transcriptional level, we tested whether these transcriptional changes reflect the translocation of AP-1 and NF-κB by chronic TNF exposure. We used a quantitative ELISA-based DNA binding assay, in which oligonucleotides corresponding to the response elements for NF-κB and AP-1 were incubated with nuclear extracts prepared from primary BDC2.5 CD4⁺ T cells with or without TCR (α-CD3 plus α-CD28 or mimotope) stimulation. The binding of each transcription factor was distinguished from the binding of others by a transcription factor-specific antibody. As shown in Fig. 4, TCR activation leads to increased binding activity of c-Fos, c-Jun, p65, and p50. Chronic TNF treatment significantly abolished the binding activity of c-Fos, c-Jun, p65, and p50 (Fig. 4).

Fig. 4.

Chronic TNF treatment significantly decreased activation of p65, p50, and cRel (A) and c-Jun and c-Fos (B) after TCR stimulation. BDC2.5 CD4⁺ T cells were isolated from mice treated with TNF for 3 weeks. Competitor oligonucleotide controls were included for every sample harvested at 3 h to demonstrate the specificity of the assay for each transcription factor indicated. 1, No treatment; 2, α-CD3+ α-CD28; 3, 1047-7; 4, 1047+ competitor.

Discussion

Here, we report characterization of the BDC2.5 CD4⁺ T cells transcriptional profile of genes influenced by TNF exposure. We observed decreased T cell responses in vivo and in vitro as described (14). Because TNF is a multipotent cytokine, many transcriptional changes were seen with TNF. These will only be selectively discussed here.

The mechanisms whereby chronic TNF effects converge to down-regulate T cell signaling are not clear. Based on our gene expression profiling, it appears that TNF has affected several genes that lie either on upstream or downstream pathways of TCR signaling. The key question is: could all of the genes, Lck, Vav2, PI3Kca, PI4Kca, Ras 21 protein activator -1 and -2, CIS, calmoudulin, and calcium channel genes in concert, cause the observed decrease in TCR signaling? To consider how TNF signaling converges on TCR signaling, we will briefly consider the roles of TCR signaling genes and their possible roles in attenuation of TCR signaling. Finally, we hypothesize a model by which TNF causes T cell unresponsiveness.

The observed up-regulation of TCR components after chronic TNF exposure is consistent with published data showing rapid internalization and degradation of the TCR upon stimulation (20). Diehn et al. (21) showed that PLC, LAT, LCK, TRIM, and CD3 ζ were repressed on T cell activation. The increased expression of LCK after TNF might thus reflect a slower transition from resting to activated T cells compared with PBS-treated T cells. Phosphorylated LAT binds phospholipase Cγ (PLCγ), which hydrolyzes phosphatidyl inositol 4,5-biphosphate [ptdins (4, 5)p2] to yield the critical second messangers inositol triphosphate (IP3) and diacylglycerol (DAG). We observed a down-regulation of PIK3ca (catalytic subunit of PIK3) and PIK4ca (Table 2). PIK3ca (p110) phosphorylates ptdins, ptdins4p, and ptdins (4,5)p2 with a preference for ptdins (4,5)p2, whereas PIK4ca acts on phosphatidylinositol (pi) in the first committed step in the production of the second messenger inositol-1,4,5,-triphosphate. The down-regulation of PI3Kca and PI4Kca suggest that TNF has a direct effect on the production of ptdins (3,4,5)p3 and the second messanger IP3.

Key steps during T cell activation are the activation of Ras and Rho family GTPase, which are also important targets for the products of PI3K. Activation of these molecules are stimulated by guanine nucleotide exchange factor (GEF). This multiprotein complex nucleates many of the subsequent events of productive TCR signaling, transmitting the TCR signal by binding a variety of proteins, one of which is Vav. Vav is a GEF for small GTPases (22). Intriguingly, Vav2, a GEF for the Rho family of Ras-related GTPases, was down-regulated after chronic TNF treatment. Vav is important in T cell activation (23), is required for synapse formation, and is necessary for Ca²⁺ flux and activation of extracellular signal-regulated kinase (ERK) (23–27). Holsinger et al. (28) showed that vav−/− T cells were deficient in IL-2 production and proliferation, and the peak of Ca²⁺ mobilization was reduced, although of normal duration. Our observed down-regulation of Vav2, Rho GTPase-activating protein-6 and –19, suggests that TNF treatment changed the expression of Ras/Rho GTPase, which in turn plays a down-regulatory role in TCR signaling (Table 2).

We propose at least two ways in which increased calmodulin may alter T cell responsiveness, and these mechanisms are not necessarily mutually exclusive (1). Increased calmodulin modulates TCR signaling through binding to an active isoform of calcium/calmodulin-dependent protein kinase II (CaMKII). Recent studies have shown that CaMKII antagonizes TCR signaling and impairs positive selection of T cells in mice (29). Lin et al. (30) also showed that CaMKIIβ'e, the splice variant of CaMKII is expressed in mouse T cells and the expression of the active form resulted in very similar changes in activated CD8 T cells, including enhanced cytotoxic activity, diminished IL-2 production, and the induction of unresponsiveness. Optimal T cell activation requires Ca²⁺-calcineurin-NFAT signaling in concert with AP-1, NF-κB, and Ras/MAPK activation. In the absence of AP-1 and NF-κB, NFAT imposes an opposing genetic program that leads to T cell unresponsiveness (31, 32). The Ca²⁺-regulated transcription factor NFAT has an integral role in both activation and anergy induction depending on whether NFAT cooperates with the transcription factor AP-1 (Fos/Jun). The integration of TCR and CD28 signaling has been proposed to occur by synergistic activation of NFATc and AP-1. According to a recent model proposed by Macian et al. (32), TCR signaling in the absence of costimulation leads to activation of NFATc but not AP-1. NFATc transcriptional activity in the absence of its transcriptional partner AP-1 would then cause a Ca²⁺-induced unresponsiveness status in T cells.

Indeed, we observed that AP-1 and NF-κB binding activity were significantly reduced after chronic TNF exposure (Fig. 4). We have also found that chronic 5-day exposure of Jurkat T cells to nontoxic levels of TNF results in a 90% reduction in AP-1 and NF-κB activation after TCR stimulation (S. Munson and H.O.M., unpublished data). Although c-Rel is crucial for TCR-dependent proliferation and IL-2 gene expression we did not observe enhanced c-Rel binding after TCR stimulation in primary BDC CD4⁺ T cells (Fig. 4). Our data are consistent with the report by Banerjee et al. (33) that cytokine-mediated priming of naive T cells requires the NF-κB family member c-Rel. However, in primary CD4⁺ T cells, IL-2 and IFN-γ gene induction was c-Rel independent. In addition, a recent study showed that one of the mechanisms by which calcineurin-NFATc can induce a state of unresponsiveness in T cells is by induction of E3 ubiquitin ligases including Itch, Cbl-b, and GRAIL, which in turn cause selective degradation of crucial signaling molecules such as protein kinase C and PLC (31, 34). We observed that several genes involved in protein ubiquitination, including E3 ubiqutin ligases, ubiquitin-conjugating enzymes, activating enzymes, and proteasomes were up-regulated in TNF-treated BDC2.5 CD4 T cells, suggesting that this might be one of the mechanisms by which TNF can induce a state of unresponsiveness in T cells.

We have discussed the alteration of signaling via the TNF receptor on TCR signaling above. However, we cannot exclude the fact that chronic TNF stimulation will exert multiple controls governing NF-κB activation and nuclear location independent of TCR signaling (Tables 1 and 2 and data not shown) that ultimately may influence TCR signaling.

In summary, we have shown that several families of T cell genes were regulated after chronic TNF treatment. The identification of T cell activation pathway genes, CISH, protein ubiquitination pathway genes, and Ca²⁺-regulated genes may provide candidate target genes for studying the mechanisms underlying TNF-mediated suppression of T cell responses in TCR tg mice. Information generated from these studies may be the basis for future experiments and may be highly relevant for the development of novel therapies for the treatment of autoimmune inflammatory diseases.

Materials and Methods

Mice.

NOD BDC2.5 TCR tg mice were the kind gift of Diane Mathis (Joslin Diabetes Center, Harvard University, Cambridge, MA). The BDC2.5 tg mice express a Vα1 Vβ4 TCR that recognizes an unknown β cell antigen presented by I-Ag7 (35). All mice were housed under barrier conditions in Stanford University animal facilities. All animal studies have been approved by Stanford University's Administrative Panel of Laboratory Animal Care.

Tg T Cells, Chronic TNF Treatment, and Proliferation Assays.

Eight- to 12-week-old female BDC2.5 TCR tg mice were treated with TNF or anti-TNF on alternate days for 3 weeks and then killed for isolation of LN and/or spleen T cell suspensions. Chronic TNF treatment in vitro was performed as described (14). Axillary, brachial, inguinal, and popliteal LNs were used in all experiments and plated at a concentration of 5 × 10⁵ cells per well in complete medium (RPMI medium 1640, 10% FCS, 20 mM l-glutamine, 1 mM sodium pyruvate, 10 mM Hepes, and antibiotics) and 20 units/ml recombinant murine IL-2 and stimulated with the indicated concentrations of the cross-reactive mimotope 1047-7 (YVAPVWVRME). After two or three cycles of stimulation with 2 μg/ml peptide plus irradiated (3,000 rad) splenic APCs, CD4⁺ T cells were purified with specific mAbs coupled to magnetic beads, according to the manufacturer's instructions (Miltenyi Biotec). Proliferation was determined in round-bottomed 96-well plates in triplicate by [^3H]thymidine incorporation during the last 16–18 h of 72-h assays.

Cytokines, mAbs, and Cytokine Assays.

Recombinant murine TNF was a gift from Centocor, and anti-TNF (clone TN3.19.12, a hamster IgG mAb) was a generous gift from Robert Schreiber (Washington University, St. Louis) (36). Hamster IgG (Pierce Biotechnology) is an Ig control for the anti-TNF. Murine TNF, anti-TNF, or control antibodies were diluted in PBS and injected i.p. in a final volume of 100 μl as described (14). At day 24 of treatment, suspensions of LNs and/or splenocytes were stimulated with the indicated concentration of the mimotope [1047-7 peptide (YVAPVWVRME)] at a density of 7–8 × 10⁶ cells/ml, in a final volume of 1 ml in 24-well plates. After prolonged in vitro culture, supernatants were harvested at the indicated times and stored at −80°C until assayed. Cytokine immunoassays were performed as described (14).

Microarray Procedure and Data Analysis.

BDC2.5 TCR tg mice were treated as described above, and total RNA was purified with TRIzol reagents (Invitrogen) and amplified by using a T7 polymerase-based method according to the manufacturer's protocol, MessageAmp aRNA Kit (Ambion). Briefly, RNA from PBS-treated T cells was reverse-transcribed into Cy3-labeled cDNA, which was cohybridized in the microarrays with Cy5-labeled cDNA derived from TNF-treated cells. Mouse cDNA microarrays were obtained from the Stanford Functional Genomics Facility and included 42 000 murine cDNAs [(49% full-length clones from the RIKEN collection (www.gsc.riken.go.jp/e/FANTOM), 36% from the NIA 15K clone set (http://lgsun.grc.nia.nih.gov/cDNA/15k.html), 3% miscellaneous custom-ordered clones, 2.3% from the Brain Molecular Anatomy Project (http://trans.nih.gov/bmap/resources/resources.htm) clone set, 2.3% Research Genetics sequence-verified clones (Invitrogen)] and the rest from private libraries, together representing ≈14,000 different mapped genes. Hybridization, scanning, and analysis using the Stanford Microarray Database (37) were performed as described (38). Briefly, six independent hybridizations were carried out, and data were analyzed by the rule-based computer program GABRIEL (18) and then displayed in color presentation by Tree View (39). The pathway genes were analyzed by Ingenuity Pathways Analysis Software.

Real-Time PCR.

Quantitative real-time RT-PCR was used for validation of gene expression changes observed in microarrays. First-strand cDNA synthesis was carried out by using 2 μg total RNA and SuperScript II (Invitrogen), following the manufacturer's instructions. Quantitative real-time PCR of randomly selected genes was performed by using the Bio-Rad iCycler Real-Time PCR Detection System and iQ SYBR Green Supermix Kit. Five percent of the first-strand reaction was used as the DNA template. Real-time PCR conditions were as follows: 94°C for 10 min, 40 cycles of 94°C for 30 s, 62°C for 30 s, and 72°C for 30 s. In all experiments, samples were run in triplicate, and mean threshold cycle (C_T) values were calculated. Quantification of a given gene is expressed as fold induction over control (PBS-treated sample). Relative quantitation of gene expression was calculated according to the comparative C_T method (40). The oligonucleotide primer pairs can be found in SI Text.

Analysis of AP-1 and NF-κB DNA Binding Activities.

An ELISA-based detection kit (Mercury TransFactor Kit; BD Biosciences Clontech) was used to detect transcription factor AP-1 and NF-κB binding. CD4⁺ T cells were purified from mice treated with TNF or PBS and stimulated with mimotope 1047-7 as described above. Nuclear extracts were tested in a 96-well format plate with oligonucleotides containing the consensus binding sequences for each transcription factor coated on the well. Competitor oligonucleotides were added in control wells to demonstrate the specificity of the DNA–protein interactions. Bound transcription factors were detected by specific primary antibodies against AP-1 and NF-κB and HRP-conjugated secondary antibodies. The absorbance of enzyme product was measured with a microplate reader.