Lack of transglutaminase 2 diminished T-cell responses in mice

Authors


Summary

Transglutaminase 2 (TG2) has been reported to play a role in dendritic cell activation and B-cell differentiation after immunization. Its presence and role in T cells, however, has not been explored. In the present study, we determined the expression of TG2 on mouse T cells, and evaluated its role by comparing the behaviours of wild-type and TG2−/− T cells after activation. In our results, naive T cells minimally expressed TG2, expression of which was increased after activation. T-cell proliferation, expression of activation markers such as CD69 and CD25, and secretions of interleukin-2 and interferon-γ were suppressed in the absence of TG2, presumably due, in part, to diminished nuclear factor-κB activation. These effects on T cells seemed to be reflected in the in vivo immune response, the contact hypersensitivity reaction elicited by 2,4-dinitro-1-fluorobenzene, with lowered peak responses in the TG2−/− mice. When splenic T cells from mice immunized with tumour lysate-loaded wild-type dendritic cells were re-challenged ex vivo with the same antigen, the profile of surface markers including CD44, CD62L, and CD127 strongly indicated lesser generation of memory CD8+ T cells in TG2−/− mice. In the TG2−/− CD8+ T cells, moreover, Eomes expression was markedly decreased. These results indicate possible roles of TG2 in CD8+ T-cell activation and CD8+ memory T-cell generation.

Abbreviations
Blimp-1

B-lymphocyte-induced maturation protein-1

BM-DCs

bone marrow-derived dendritic cells

CFSE

carboxyfluorescein succinimidyl ester

CHS

contact hypersensitivity

Con A

concanavalin A

CPM

count per minute

DAPI

4′,6-diamidino-2-phenylindole

DCs

dendritic cells

DIC

differential interference contrast

DNFB

2,4-dinitro-1-fluorobenzene

GM-CSF

granulocyte–macrophage colony-stimulating factor

LPS

lipopolysaccharides

MLR

mixed lymphocyte reactions

PMA/I

phorbol myristate acetate/ionomycin

rmIL-2

recombinant mouse interleukin-2

TG2

transglutaminase 2

WT

wild-type

Introduction

Transglutaminase 2 (tissue transglutaminase; TG2) is a ubiquitous enzyme that is present not only inside the cell but also on the cell surface and even in the extracellular matrix. It modifies the structure and function of its substrate proteins by deamidation, transamidation, and by cross-linking glutamine and lysine residues, so exerting many effects on behaviours such as apoptosis and survival, adhesion and migration, signal transduction, in a variety of cells.[1] These effects modify biological processes, one of which is inflammation. Indeed, TG2 has been associated with inflammatory afflictions such as coeliac disease,[2] cystic fibrosis,[3] autoimmune inflammation,[4] and others. The possible mechanisms through which it is involved in inflammation are diverse. TG2 activates nuclear factor-κB (NF-κB) in murine macrophages by degrading IκB and letting these cells secrete inflammatory cytokines and chemokines.[5] Extracellular TG2 is known to convert latent transforming growth factor-β to its active form,[6] so initiating inflammation. In cystic fibrosis, TG2 polymerizes and so inactivates peroxisome proliferator-activated receptor-γ, which is a negative regulator of inflammation, in monocytes and macrophages.[2] Neutrophils in the absence of TG2 show impaired extravasation, attenuated chemotaxis, and diminished formation of superoxide anion,[7] all of which imply decreased inflammation.

In addition to inflammation, there are other clues to the possible involvement of TG2 in adaptive immunity, via its action on dendritic cells (DCs), T cells and B cells, which are major actors in adaptive immune responses. TG2 is expressed on the surface of human monocyte-derived DCs, the expression of which is increased upon lipopolysaccharide (LPS) stimulation.[8] Its ablation in these cells deteriorates cell maturation with respect to the up-regulation of surface molecules (MHC I, CD80, and CD86) and cytokine secretion [interleukin-10 (IL-10), IL-12]. Also, upon LPS stimulation, TG2−/− DCs show impaired up-regulation of MHC II and CD86.[9] These results suggest a modulatory role of TG2 in DC activation and function. In activated B cells, TG2 is suggested to be involved in the expression of activation-induced cytidine deaminase and B-lymphocyte-induced maturation protein-1 (Blimp-1), and thereby also in the modulation of the extent of humoral immune responses.[10] The presence of TG2 in activated T cells has been reported of HIV-infected patients.[11, 12] However, the presence of TG2 in T cells from healthy individuals or from experimental animals has not yet been determined; even the role of TG2 in T cells has scarcely been studied. There was one report that treatment of CD8+ T cells with 6B9 antibody, a known anti-TG2 antibody, inhibited their migration;[13] but it turned out that this antibody does not recognize TG2 but rather CD44.[14] Hence, the role of TG2 in T cells has remained obscure. Recently, TG2, in a mouse experimental autoimmune encephalomyelitis model, was reported to exacerbate the pathogenesis and progress of the disease, which effects were partly T-cell intrinsic.[4] Still, the exact relationship between TG2 and T cells awaits thorough investigation.

In the present study, we evaluated the presence and putative role of TG2 in mouse T cells by comparison of the behaviours of cell samples from wild-type (WT) and TG2−/− mice.

Materials and methods

Mice

TG2−/− mice obtained from De Laurenzi and Melino,[15] and WT C57BL/6J mice purchased from the Jackson Laboratory were maintained at the animal facility of Seoul National University College of Medicine. Male mice 7–12 weeks old were used in the experiments, which were performed with the approval of the Institutional Animal Care and Use Committee of Seoul National University (permission #: SNU-121004-1).

Isolation and activation of T cells

Mouse splenic T cells were isolated by negative selection using a mouse Pan T cell Isolation Kit (Miltenyi Biotech, Bergisch Gladbach, Germany) or a mouse CD8+ T-cell Isolation Kit (Miltenyi Biotech) according to the manufacturer's instructions. They were cultured in RPMI-1640 medium (Welgene, Seoul, Korea) with 10% (volume/volume) fetal bovine serum, 100 U/ml penicillin and 100 mg/ml streptomycin (Gibco, Carlsbad, CA). T cells were activated with anti-CD3ε (1 μg/ml, clone 1.45-2C11; BD Biosciences, San Diego, CA) and anti-CD28 (2 μg/ml, clone 37.51; BD Biosciences) antibodies, concanavalin A (Con A; 5 μg/ml; Sigma-Aldrich, St Louis, MO), or PMA (10 ng/ml; Sigma-Aldrich) and ionomycin (I; 50 ng/ml; Sigma-Aldrich). For prolonged culturing of isolated CD8+ T cells, they were seeded in a six-well plate at a density of 1 × 106 cells/ml, and activated with anti-CD3 and anti-CD28 antibodies for 48 hr. Then, the cells were re-plated at a 5 × 105 cells/ml in fresh media supplemented with 100 U/ml recombinant murine IL-2 every 24 hr.

Generation of bone marrow-derived DCs

Bone marrow-derived DCs were prepared as previously described.[16] Briefly, bone marrow cells were obtained by flushing the tibia of mice with PBS (pH 7·4). After red blood cell lysis, the cells were washed with PBS and seeded in a six-well plate at a density of 1 × 106 cells per well in a final volume of 2 ml. They were cultured at 37° in a humidified 5% CO2 atmosphere. Two mililitres of fresh medium was added on day 3, and half of the medium was replaced except IL-4 on day 6. The cells were harvested and ready for experimental use on day 7. The purity of the CD11c+ DCs, as confirmed by flow cytometric analysis, was over 90%, (data not shown).

Mixed lymphocyte reactions

Mixed lymphocyte reactions (MLR) were conducted as previously described.[16] Briefly, bone marrow-derived DCs were pre-treated with 25 μg/ml mitomycin C (Sigma-Aldrich) for 30 min at 37°, after which they were co-cultured with splenic T cells at a ratio of 1 : 80 (DC : T) in a U-bottom 96-well plate (NUNC, Copenhagen, Denmark). For syngeneic MLR, the DCs were stimulated with 1 μg/ml LPS (Sigma-Aldrich) for 12 hr, washed with PBS, and co-cultured with T cells. When needed, DCs were loaded, before MLR, with B16F10 tumour cell lysate (100 μg protein/well), or with dinitrobenzene sulphonic acid (50 mg/ml, Sigma-Aldrich) for 24 hr.

DC vaccination

B16F10 tumour cell lysates were prepared by repeated freezing and thawing, and kept at −80° until used. The lysate (100 μg protein/well) was added to the DC culture on day 6 for 24 hr, after which the DCs were harvested, washed and re-suspended in PBS at a cell density of 1 × 107 cells/ml. On days 0, 5 and 10, each mouse was intraperitoneally injected with 1 × 106 DCs.

T-cell proliferation assay

To evaluate cell proliferation, 1 μCi/well [3H]thymidine (Amersham Pharmacia Biotech, Oslo, Norway) was added to the culture media for 18 hr, after which the cells were harvested using an INOTECH cell harvester (Inotech, Dietikon, Switzerland) and their radioactivities were measured using a scintillation β-counter (MicroBeta Trilux, Perkin Elmer, Waltham, MA).

Naive T cells were suspended at 1 × 106 cells/ml and labelled with 10 μm carboxyfluorescein succinimidyl ester (CFSE; Molecular Probes, Eugene, OR) for 10 min at 37°. The reaction was stopped by addition of an equal volume of fetal bovine serum (Gibco) for 2 min at room temperature. The CFSE-labelled cells were washed and then re-suspended in the complete medium, activated, incubated for 3 or 6 days, and analysed on a FACSCalibur (Becton Dickinson, Lincoln Park, NJ). The data obtained were processed using Winmdi software.

Reverse transcription polymerase chain reaction

Total RNA was extracted from T cells using Trizol (Invitrogen, Carlsbad, CA). From the total RNA (1 μg), cDNA was generated using a reverse transcription (RT-) PCR Kit (iNtRON Biotechnology, Seoul, Korea), and amplified for TG2, Blimp-1, B-cell CLL/lymphoma 6 (Bcl-6), T-bet, Eomes or β-actin. The following primer pairs were employed: 5′-AGGACATCAACCTGACCCTG-3′ (forward primer for TG2), 5′-TTAGGCCGGGCCGATGATAAC-3′ (reverse primer for TG2), 5′-TGACTTTGTGGACAGAGGCCGAGT-3′ (forward primer for Blimp-1), 5′- CTGTTGTTGGCAGCATACTTGAAA-3′ (reverse primer for Blimp-1),[17] 5′-CGGGACATCTTGACGGAC-3′ (forward primer for Bcl-6), 5′-CAGGGTGATTTCAGGATATA-3′ (reverse primer for Bcl-6),[17] 5′-GTTCCCATTCCTGTCCTTC-3′ (forward primer for T-bet), 5′-CCTTGTTGTTGGTGAGCT-3′ (reverse primer for T-bet) Tayade et al.[18], 5′-GCCTACCAAAACACGGATA-3′ (forward primer for Eomes), 5′- TCTGTTGGGGTGAGAGGAG -3′ (reverse primer for Eomes), 5′-AACCCTAAGGCCAACCGTGAAAAG-3′ (forward primer for β-actin), 5′-GCAGGATGGCGTGAGGGAGAG-3′ (reverse primer for β-actin).

For β-actin PCR, each cycle consisted of denaturation at 95° for 30 seconds, annealing at 58° for 30 seconds and amplification at 72° for 30 seconds. For TG2 PCR, the cycle consisted of denaturation at 95° for 40 seconds, annealing at 55° for 40 seconds and amplification at 72° for 40 seconds. The PCR products after 35 to 40 cycles were analysed by 2% agarose gel electrophoresis and subjected to densitometric analysis using Quantity One software (Bio-Rad Laboratories, Hercules, CA).

Western blot analysis of TG2

T cells were harvested in radioimmunoprecipitation assay buffer. After 30 min on ice, cellular debris was removed by centrifugation (10 min, 4°). Protein concentrations were determined by a bicinchoninic acid protein assay method. Total protein was resolved by SDS–PAGE, and then transferred to PVDF membranes (Millipore Corporation, Billerica, MA). After blocking with 5% skimmed milk, the membranes were probed with anti-TG2 (1 : 500 dilutions; Thermo Scientific, Fremont, CA) or anti-β-actin (1 : 4000 dilutions; Sigma-Aldrich) antibody. Blots were developed using an enhanced chemiluminescence Western blotting detection system (GE Healthcare, Chalfont St Giles, UK).

Flow cytometric analysis

T cells were stained with various antibodies; anti-CD4-allophycocyanin or -FITC (clone RM4-5), anti-CD8-peridinin chlorophyll protein or -phycoerythrin (clone 53-6.7), anti-CD69-FITC (clone H1.2F3), anti-CD25-phycoerythrin (clone 7D4), anti-CD127-FITC (clone 1M7.8.1), anti-CD44-phycoerythrin (clone A7R34, all from BD Biosciences), and anti-CD62L-allophycocyanin (clone MEL-14, eBioscience, San Diego, CA). Intracellular staining was carried out with anti-interferon-γ (IFN-γ) -FITC (clone XMG1.2, BD Biosciences) after permeabilization using the BD Cytofix/Cytoperm Kit (BD Biosciences) according to the manufacturer's instructions. Cell surface TG2 staining of T cells was carried out in the same way as immunofluorescence staining except that it was performed in a FACS tube (BD Falcon, BD Biosciences, San Jose, CA).

Immunofluorescence staining

T cells were activated for 2 days on poly-l-lysine (0·01%, Sigma-Aldrich) coated slides, fixed in 4% paraformaldehyde for 10 min at room temperature, blocked with endogenous blocking reagent (M.O.M kit; Vector Labs, Burlingame, CA) in the diluents (M.O.M kit; Vector Labs) for 1 hr, and incubated with anti-TG2 antibody TG100 (Thermo Scientific) for 1 hr. Anti-mouse immunoglobulin-Alexa Fluor 488 (Molecular Probes) was applied as a secondary antibody for 30 min, and observed under a confocal fluorescence microscope (Olympus, Tokyo, Japan) and FV10-ASW 2.0 Viewer (Olympus).

ELISA

Cell culture supernatants were analysed for their IFN-γ and IL-2 levels (Invitrogen) according to the manufacturer's instructions. Intracellular TG2 activity was evaluated by assay of 5-(biotinamido)-pentylamine (Thermo Scientific) incorporation into cellular proteins as previously described.[19]

Induction of contact hypersensitivity reaction

After obtaining ear thickness measurements in untreated mice as a baseline, contact hypersensitivity (CHS) was induced as previously described.[20] Briefly, mice were sensitized on days 0 and 1 by application of 20 μl of 0·5% 2,4-dinitro-1-fluorobenzene (DNFB, Sigma-Aldrich) in olive oil : acetone (1 : 4 volume/volume) to the shaved back. On day 40, 20 μl of 0·2% DNFB in olive oil : acetone (1 : 4 volume/volume) was applied to the left pinna. The pinna thickness was measured with a constant-loading dial micrometer (Mitutoyo, Kawasaki, Japan) every 24 hr until ear swelling subsided. The per cent pinna thickness was calculated as follows: % pinna thickness = [(thickness after sensitization − thickness before sensitization)/(thickness before sensitization)] × 100.

Statistical analysis

Data were expressed as mean ± SD. The statistical significance between the groups was analysed by non-parametric, Mann–Whitney U-test using spss software version 11 (SPSS Inc., Chicago, IL). A value of P < 0·05 for statistical significance was set.

Results

TG2 expression was increased in activated T cells

We first determined the presence of TG2 on naive and activated T cells. To this end, mouse splenic T cells were isolated, stimulated with anti-CD3 and anti-CD28 antibodies, and examined for expression of TG2, both in mRNA (Fig. 1a) and protein (Fig. 1b) levels. The naive T cells minimally expressed TG2, and the expression was increased five- to sixfold after stimulation. This was confirmed by immunohistochemical staining, which was minimal in the case of naive T cells, but maximal in the case of 48 hr-stimulated T cells, both in the cytoplasm and on the cell surface (Fig. 1c). The surface expression of TG2 was confirmed, this time by flow cytometric analysis results showing higher mean fluorescence intensity in the activated T cells (Fig 1d). And once again, the intracellular TG2 enzyme activity was much higher in the activated T cells than in the naive ones (Fig. 1e).

Figure 1.

Expression of transglutaminase 2 (TG2) in naive and activated T cells. Wild-type (WT) mouse splenic T cells were isolated and stimulated with anti-CD3 and anti-CD28 antibodies. After 24 or 48 hr, they were subjected to (a) RT-PCR and (b) Western blot analysis to determine the presence of TG2. For each result, a densitometric analysis was performed, the results of which are shown on the right side of each panel. (c) Immunofluorescence-staining for TG2 was minimal in the naive T cells (upper panels), but strongly positive in the 48 hr-stimulated T cells, both in the cytoplasm and on the cell surface (lower panels). (d) Using the same antibodies, naive and 48 hr-activated T cells were stained without permeabilization, and then subjected to flow cytometric analysis. (e) Intracellular TG2 enzyme activity was determined as described in the 'Materials and methods'. HeLa cells were used as a positive control. All of the experiments were repeated more than three times; the results depicted are representative. DAPI, 4′,6-diamidino-2-phenylindole; DIC, differential interference contrast.

In vitro T-cell proliferation was decreased in the absence of TG2

The presence of TG2 in T cells suggests a possible role for this enzyme in T-cell development. Hence, we first determined whether there were, between WT and TG2−/− mice, differences in the frequencies of T-cell subsets in their spleen, and found none (see Supporting information, Fig. S1). Enhanced expression of TG2 in activated T cells suggested a possible role of this enzyme in T-cell activation. To this end, we stimulated T cells from WT and TG2−/− mice with Con A, anti-CD3 and anti-CD28 antibodies, or PMA/I for 2 days, and performed a [3H]thymidine incorporation assay (Fig. 2a). According to the observed effects, for all of the stimulants, cell proliferation was decreased in the TG2−/− T cells relative to the WT ones. After the cells were loaded with CFSE and stimulated with anti-CD3 and anti-CD28 antibodies for 3 days (Fig. 2b), moreover, the number of cell cycles was smaller in the TG2−/− T cells, in both the CD4+ and CD8+ subsets. This tendency was more pronounced on day 6 (see Supporting information, Fig. S2). The same results, furthermore, were obtained when the T cells were co-cultured with allogeneic (Fig. 2c) or syngeneic (Fig. 2d) DCs.

Figure 2.

In vitro proliferation of wild-type (WT) and transglutaminase 2 deficient (TG2−/−) T cells with various stimulants. (a) Mouse splenic T cells were isolated and stimulated for 48 hr, [3H]thymidine was added to the culture medium for an additional 18 hr, and the counts per minute (CPM) were measured. (b) T cells were loaded with CFSE and activated with anti-CD3 and anti-CD28 antibodies for 72 hr, and then stained with anti-CD4-allophycocyanin and anti-CD8-peridinin chlorophyll protein (PerCP) antibodies. They were gated based on CD4 or CD8 staining and analysed for their CFSE profiles, respectively. (c) T cells were co-cultured with dendritic cells (DCs) from BALB/c mice for 48 hr, and subjected to a [3H]thymidine incorporation assay. (d) T cells were co-cultured with syngeneic DCs pre-stimulated with lipopolysaccharide. All of the experiments were repeated more than three times; the results depicted are representative. *< 0·05; WT, wild-type; DC, dendritic cells; PMA/I, phorbol myristate acetate/ionomycin; CFSE, carboxyfluorescein succinimidyl ester.

Expression of activation markers was decreased in the absence of TG2

We next examined the expression levels of the early activation markers, CD69 and CD25, the expression of which is critical for the proliferation of activated T cells in the initial stage.[21, 22] Cells were stimulated with anti-CD3 and anti-CD28 antibodies for 3 days, stained, and subjected to flow cytometric analysis (Fig 3a). The frequency of the CD69+ CD25+ T cells was lower in the TG2−/− T cells, especially in the CD8+ subset (61·0% versus 43·8%). Additionally, when T cells were co-cultured with allogeneic DCs for 72 hr (Fig. 3b), a reduced frequency of CD69+ CD25+ cells was once again observed in TG2−/− T cells. As another criterion of T-cell activation, cytokine secretion in the culture supernatants of antibody-stimulated T cells was determined by ELISA, and it was found that the concentrations of IL-2 and IFN-γ were both lower in the TG2−/− T cells (Fig. 3c). These results together suggest, once again, a probable TG2 role in T-cell activation.

Figure 3.

Expression of activation markers upon in vitro stimulation. (a) Splenic T cells from wild-type (WT) and transglutaminase 2-deficient (TG2−/−) mice were isolated, stimulated with anti-CD3 and anti-CD28 antibodies for 3 days, stained for CD4, CD8, CD69 and CD25, and subjected to flow cytometric analysis. (b) In addition, T cells were co-cultured with allogeneic bone marrow-derived dendritic cells for 72 hr, stained, and processed in the same way. (c) The concentrations of interleukin-2 (IL-2) and interferon-γ (IFN-γ) in the experiment A supernatants were determined by ELISA. (d) For the evaluation of nuclear factor (NF) -κB activation, splenic T cells were stimulated with antibodies for 6 hr, harvested, and their relative amounts of nuclear NF-κB were estimated using commercial kits. All of the experiments were repeated more than three times; the results depicted are representative. *< 0·05; **< 0·01; WT, wild-type; DC, dendritic cells.

Because cell proliferation,[23] CD69 and CD25 expression,[24, 25] and IL-2 secretion[26] following T-cell activation are under the influence of NF-κB activation, and TG2 has been reported to enhance NF-κB activation in various murine cells,[5, 27, 28] we speculated as to whether the absence of TG2 in activated T cells leads to diminished NF-κB activation. To find the answer, splenic T cells were activated with anti-CD3 and anti-CD28 antibodies for 6 hr, harvested and the relative amounts of nuclear NF-κB were estimated using commercial kits. As expected, the level of NF-κB in TG2−/− T cells was lower than that in WT T cells (Fig. 3d).

Intensity of CHS reaction to DNFB was attenuated in TG2−/− mice

We next considered whether a decreased TG2−/− T-cell response, especially that of CD8+ T cells, affected in vivo immune responses. We provoked a CHS reaction using DNFB as an antigen, because this type of response is known to be largely dependent on CD8+ T cells;[29] to which end, mice were sensitized and challenged with DNFB. As the plot in Fig. 4(a) makes clear, the reaction intensity, measured as the increase in pinna thickness, was significantly lower in the TG2−/− mice on the peak days, days 3 and 4 (= 0·026 and 0·014, respectively). To confirm that the reduced CHS reaction in the TG2−/− T cells was associated with a reduced T-cell response in these mice, splenic T cells were isolated 3 weeks after the CHS response and then re-stimulated with DCs loaded with dinitrobenzene sulphonic acid, the water-soluble analogue of DNFB. In the results, not only the proliferation (Fig. 4b) but also the frequency (Fig. 4c) of the IFN-γ-secreting CD8+ cells was decreased in the T cells from the TG2−/− mice. Correspondingly, the concentration of IFN-γ in the TG2−/− T cells' culture supernatant was lower than that in the WT T cells (Fig. 4d).

Figure 4.

Contact hypersensitivity reaction. Mice were sensitized on days 0 and 1 with 2,4-dinitro-1-fluorobenzene (DNFB). (a) Forty days later, the mice were challenged with DNFB on the surface of the left pinna. The pinna thickness was measured every 24 hr until the swelling subsided. The percentage increase of pinna thickness was calculated and depicted in the graph. (b) Twenty-one days after challenge, splenic T cells were isolated and re-stimulated with dinitrobenzene sulphonic acid-loaded DCs for 24 hr, and a [3H]thymidine uptake assay was performed. (c) The re-stimulated T cells were stained for CD8 and intracellular interferon-γ (IFN-γ) and then subjected to flow cytometric analysis. (d) The concentrations of IFN-γ in the culture supernatants were determined by ELISA. All of the experiments were repeated twice and the representative one is shown. *< 0·05; **< 0·01; WT, wild-type.

In vivo T-cell response was suppressed in the absence of TG2

Because DCs are involved in the process of the CHS response, and the absence of TG2 in DCs affect cell behaviours,[8, 9] the above results should be regarded as a summation of the effects of the absence of TG2 both in DCs and T cells. To rule out the effect of DCs, WT and TG2−/− mice were immunized with tumour-lysate-loaded WT bone-marrow-derived DCs. Thirty days later, splenic T cells were isolated, re-stimulated with DCs loaded with the same antigen for 72 hr to stimulate antigen-specific T cells, and a [3H]thymidine incorporation assay was performed. At the same time, the re-stimulated cells were stained for CD8, CD44, CD62L and CD127, and subjected to flow cytometric analysis. For comparison, T cells from un-immunized WT and TG2−/− mice were treated in the same manner.

In the results, 30 days after DC vaccination, splenic T cells from the TG2−/− mice showed less proliferation than those from the WT mice (Fig. 5a), as was shown in Fig. 4(b). As for the surface marker expression, in vitro-stimulated CD8+ T cells from the un-immunized mice (Fig. 5b) showed overall down-regulation of CD127 and increased frequencies of CD44high CD62Llow effector CD8+ T cells relative to the un-stimulated T cells from the same mice (see Supporting information, Fig. S3). However, there was no remarkable cell fraction difference between the in vitro-stimulated WT and TG2−/− CD8+ T cells (Fig. 5b). When the same analysis was performed on T cells 30 days after vaccination (Fig. 5c), the frequencies of the CD44high CD62Llow CD8+ T cells, which, at this time, were expected to comprise both recently activated effector cells and effector memory cells, were much higher in both the WT and TG2−/− T cells compared with their partners in Fig. 5(b), suggesting a build-up of memory responses by DC vaccination. Among the WT and TG2−/− T cells from the vaccinated mice, the TG2−/− T-cell frequency was 50% lower than that of the WT T cells. The frequency of CD127+ cells, the memory cell marker, with the CD62Llow or CD44high phenotypes was also lower in TG2−/− CD8+ T cells.

Figure 5.

T-cell responses after dendritic cell (DC) vaccination. Mice were immunized with wild-type ( WT) bone marrow-derived dendritic cells (DCs) loaded with tumour lysate. (a) Thirty days later, splenic T cells were isolated and re-stimulated with lysate-loaded DCs for 72 hr, and a [3H]thymidine uptake assay was performed. T cells from (b) the un-immunized mice and (c) the mice immunized 30 days earlier were isolated and stimulated with antigen-loaded DCs for 72 hr. Then, the cells were stained for CD8, CD44, CD62L, and CD127, gated on the basis of CD8, and were subjected to flow cytometric analysis to determine the cell fractions of CD62Llow CD44high and CD62Llow CD127+ cells. All of the experiments were repeated more than three times; the results depicted are representative. WT, wild-type, *< 0·05.

Eomes expression was markedly decreased in activated TG2−/− T cells

Next, we explored the expression of transcription factors related to effector and memory T-cell differentiation including Blimp-1, Bcl-6, T-bet, and Eomes.[30] To that end, naive splenic CD8+ T cells were isolated, stimulated with anti-CD3 and anti-CD28 antibodies, and cultured in the presence of recombinant mouse IL-2. On the appointed days, cells were harvested and RT-PCR was performed for each molecule. Figure 6 depicts the results with densitometric analysis. As can be seen, the expression levels of the molecules were generally lower in the TG2−/− T cells on all of the post-stimulation days. Of particular note were the late appearance of Eomes and its marked decrease in the TG2−/− T cells.

Figure 6.

Expression of molecules related to T-cell differentiation. Splenic CD8+ T cells were isolated, stimulated with anti-CD3 and anti-CD28 antibodies for 48 hr, re-plated at a density of 5 × 105 cells/ml of culture medium, and maintained for 8 days with every day recombinant murine, interleukin 2 (IL-2) supplementation. (a) The cells were harvested on the days indicated, total RNA was extracted, and gene amplification was performed using appropriate primer sets. (b) The results are depicted as densitometric values. This experiment was repeated twice; the representative one is shown. D, day; WT, wild-type.

Discussion

Expression of TG2 in T cells has been suggested but not, with the exception of HIV-infected patients,[11, 12] reported. In the present study, we found in mouse naive T cells both minimal TG2 expression and its increase following cell activation. The expression was present not only on the surface but also in the cytosol, as is usual in many cell types.[1]

The increased expression of TG2 following activation could be a physiological response to a type of cellular stress,[31] namely the ‘activation’, whereby cellular metabolism changes abruptly,[32] cells are driven to die,[33] and intracellular reactive oxygen species increase.[34] However, our data indicate that the activation events including cell proliferation, CD69 and CD25 expression, and IL-2 and IFN-γ secretion were suppressed in the absence of TG2 (Fig. 3). This could be the result of increased susceptibility of activated T cells to apoptosis in the absence of TG2, because this enzyme has been reported to be anti-apoptotic in some kinds of cells.[35, 36] However, when we analysed the frequency of dead cells 3 days after in vitro activation with anti-CD3 and anti-CD28 antibodies, there was no difference of cell death ratio of total T cells, CD8+ T cells, or CD4+ T cells between WT and TG2−/− T cells (see Supporting information, Fig. S4). Hence, the diminished levels of activation events observed in this study preferentially suggest a definite role of this enzyme in T-cell activation.

These events are known to be affected by NF-κB activation.[23-26] Because the lack of TG2 lowered NF-κB activation in mouse T cells (Fig. 3d), as has been found in other cells such as murine peritoneal macrophages and embryonic fibroblasts,[5, 27, 28] we can assume that the lowered NF-κB activation contributed to these phenomena, at least in part, if not all. Other than via NF-κB activation, TG2 also regulates gene expression directly or indirectly, by modifying other transcriptional factors and related proteins such as Akt1, p53, RhoA, Rac1, and many others, using its enzymatic and non-enzymatic activities in various kinds of cells.[1] For example, in TG2-expressing ovarian cancer cell line, SKOV3, this enzyme recruits c-Src, which phosphorylates β-catenin, resulting in an increase of nuclear β-catenin concentration and subsequent T-cell factor/lymphoid enhancer-binding factor 1-mediated gene activation.[37] The same group also observed in several metastatic cell lines that TG2 promoted degradation of protein phosphatase 2. This shifted the balance between phosphorylation and dephosphorylation of cAMP-response element-binding protein (CREB) towards phosphorylation, and so augmented the activity of CREB, which in turn increased the expression of matrix metalloproteinase-2.[38] Besides, many other proteins are target substrates of TG2.[39] This opens the possibility that TG2 contributes to the activation of T cells in a variety of ways. TG2 is also involved in the intracellular Ca2+ flow. Ca2+-activated TG2 increases Ca2+ release from the endoplasmic reticulum via cross-linking of RAP1GDS1, a guanine exchange factor, in TG2-transfected Jurkat T cells, amplifying Ca2+ signal.[40] Hence, physiological amounts of TG2 in activating T cells could increase intracellular Ca2+ concentration, which would augment cell activation. All these putative mechanisms of signal modulation by TG2 in T cells have not been tested. If increased intracellular TG2 activity in activated T cells (Fig. 1e) helps signal pathway-associated molecules, which is not known, the absence of this enzyme would attenuate T-cell activation, as observed in this study. For now, we have no idea which mechanism is related to which activation event, and to what extent. To make it clear, more sophisticated studies are needed.

Effects of TG2 on human DCs have been reported, when the cells are activated.[9] Specifically, in the presence of TG2 inhibitor, KCC009, human monocyte-derived DCs down-regulated the expression of CD80 and CD86, and secreted less IL-10 and IL-12. Hence, the KCC009-treated DCs suppressed the T-cell secretion of IFN-γ. Also, in B cells, TG2 expression was increased with activation, and the absence of this enzyme led to over-expression of cytidine deaminase and Blimp-1, suggesting a regulatory role of this enzyme in B-cell activation.[10] A recent study, furthermore, suggested a Th17- and Th1-promoting role for TG2 in mouse CD4+ T cells.[4] Overall, our results, together with those in the recent literatures, indicate a certain TG2 role in adaptive immune responses.

Actually, when we elicited a CHS reaction in WT and TG2−/− mice to evaluate the effect of TG2 absence on the in vivo immune response, the magnitude of the peak response on days 3 and 4 was lower in the TG2−/− mice (Fig. 4a). On the one hand, these results, because the reaction largely depends on CD8+ memory T cells,[29] can be regarded as reflecting the effect of TG2 absence on CD8+ T cells. On the other hand, these results do not exclusively reflect the effect of TG2 absence in CD8+ T cells because DCs also are involved in this type of response, and the absence of TG2 on DCs would affect the whole response.

Hence, in our next experiment, we immunized mice with tumour lysate as loaded on WT bone-marrow- derived DCs, not with the tumour lysate itself, in order to rule out the possible effect of TG2−/− DCs in TG2−/− mice. When the splenic T cells were isolated 30 days after DC vaccination and stimulated ex vivo with tumour lysate-loaded DCs, it was found that TG2−/− T-cell proliferation was decreased relative to the WT T cells (Fig. 5a), which also had been observed of the DNFB-committed T cells in the CHS experiment (Fig 4b). In both cases, the splenic T cells should have contained the antigen-specific memory T cells. However, it was not clear whether the suppressed antigen-specific proliferation of TG2−/− T cells reflected a reduced number of antigen-specific memory cells in this population. Even though the memory T cells were lowered in their activation threshold,[41] we have no idea whether the observed low responsiveness of naive TG2−/− T cells compared with naive WT T cells was overcome as they became memory cells. Instead, we assumed a reduced pool of memory cells in TG2−/− mice after antigen commitment, based on a previous report that decreased primary burst size resulted in decreased generation of memory cells,[42] as well as the present surface marker analysis (Fig. 5b,c). When T cells from naive mice, both WT and TG2−/−, were ex vivo-stimulated with tumour lysate-loaded DCs, CD8+ T cells from both the WT and TG2−/− mice showed similar fractions of CD62Llow CD44high cells, or CD62Llow CD127+ cells, which at that point could be regarded as effector cells.[43] These cell fractions markedly increased upon antigen-specific stimulation of T cells from committed mice (Fig. 5c), relative to those from naive mice (Fig. 5b), suggesting the participation of memory CD8+ T cells in eliciting such cell fractions. It should be noted that the extent of this increase was more prominent in CD8+ T cells from WT mice than in those from TG2−/− mice, suggesting a diminished generation of CD8+ memory T cells in TG2−/− mice. Nonetheless, these results are circumstantial; additional studies using antigen-specific tetramer are needed to make a clear determination of whether TG2 absence led to decreased generation of memory CD8+ T cells.

We examined the expression of transcription factors related to T-cell activation and memory cell differentiation, after in vitro activation (Fig. 6).

Blimp-1 drives primary effector T-cell generation,[44] while Bcl-6 is important for the generation and maintenance of memory CD8+ T cells.[45] These two molecules are known to react reciprocally. Meanwhile, T-box-containing transcription factors, T-bet and Eomes are also related to the development of CD8+ effector and memory T cells, respectively.[46, 47] The overall expression of these transcription factors, in the present study, was decreased in TG2−/− T cells. These results seem to reflect the decreased fraction of activated T cells in the population. Of note was the prominent decrease of Eomes in the TG2−/− T cells. Eomes is a key transcription factor of CD8+ T cells for primary clonal expansion and re-expansion after challenge.[48] Hence, the decreased secondary CD8+ T-cell response observed in the absence of TG2 could, in part, be mediated by decreased expression of Eomes late in the primary expansion.

In conclusion, we posit that TG2 is involved in T-cell responses, especially in those of CD8+ T cells, and presumably in both the primary and secondary responses. The roles of TG2 in coeliac disease pathogenesis have been much studied. One of these roles is deamidation of gliadin peptide which, in this state, binds more strongly to HLA-DQ2 and HLA-DQ8 molecules on antigen-presenting cells.[49] These complexes then stimulate gluten-specific CD4+ T cells more efficiently, which initiates events leading to coeliac disease.[50] However, the roles of TG2 in T-cell activation per se have not been considered. The elucidation of TG2 involvement in T-cell activation will prove helpful in understanding the basic mechanisms of related diseases.

Acknowledgements

JHK and JMH performed the experiments, EMJ took care of and supplied the experimental animals, IGK and YIH designed the experiment, WJL, HRK, and JSK participated in the follow up and discussion about the experimental data, JHK and YIH wrote the manuscript. This work was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean Government (2010-0010184).

Disclosures

The authors declare that they have no financial and commercial conflicts of interest.

Ancillary