SEARCH

SEARCH BY CITATION

Keywords:

  • Gamma-delta;
  • Lymphocyte;
  • Lymphoma;
  • Microarray;
  • Phosphoantigen

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References
  9. Supporting Information

Global transcriptional technologies have revolutionised the study of lymphoid cell populations, but human γδ T lymphocytes specific for phosphoantigens remain far less deeply characterised by these methods despite the great therapeutic potential of these cells. Here we analyse the transcriptome of circulating TCRVγ+ γδ T cells isolated from healthy individuals, and their relation with those from other lymphoid cell subsets. We report that the gene signature of phosphoantigen-specific TCRVγ+ γδ T cells is a hybrid of those from αβ T and NK cells, with more ‘NK-cell’ genes than αβ T cells have and more ‘T-cell’ genes than NK cells. The expression profile of TCRVγ+ γδ T cells stimulated with phosphoantigen recapitulates their immediate physiological functions: Th1 cytokine, chemokine and cytotoxic activities reflect their high mitotic activity at later time points and do not indicate antigen-presenting functions. Finally, such hallmarks make the transcriptome of γδ T cells, whether resting or clonally expanding, clearly distinctive from that of NK/T or peripheral T-cell lymphomas of the γδ subtype.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References
  9. Supporting Information

Although the biology of αβ T lymphocytes is reasonably well understood, that of γδ T lymphocytes remains far less characterised, with development, reactivity and functions that relate to both innate and adaptive immunity. These lymphoid cells express both rearranged antigen receptors and non-rearranged receptors for self-HLA class I or stress signals. In human and non-human primates, the peripheral repertoire of the TCR is markedly biased, with most blood γδ T cells expressing TCRVγ9Vδ2 specific for phosphorylated isoprenoïds (the so-called phosphoantigens) produced by the microbial DOXP and eukaryote mevalonate pathways. Within 4–6 h after activation, TCRVγ9Vδ2+ γδ T cells release the pro-inflammatory chemokines MIP-1α, MIP-1β and RANTES, secrete Th1 cytokines IFN-γ and TNF-α and kill cancer cells through granzyme, perforin, FasL and TRAIL. In addition, TCRVγ9Vγ2+ γδ T cells can take various patterns of differentiation according to the tissue and physiological contexts. These lymphocytes were characterised as Th0 cells 1, Th1 cells 2, T follicular helper cells 3, terminally differentiated Ra cells 4, Treg cells 5, Th17 cells 6 and even as phagocytes 7 or antigen-presenting cells 8. Furthermore, the cytolytic activity of human γδ T lymphocytes may result from activation cascades driven by TCR, NKG2D, CD16, CD160, KIR2DS1, KIR2DS2 and NKp44 and under the negative regulation by inhibitory receptors CD94/NKG2A, KIR2DL1, KIR2DL2 and KIR2DL3 9. Thus, TCRVγ9Vδ2+ γδ T cells exert NK-like cytotoxic responses against HLA class I-deficient cell targets in addition to T-cell-type cytotoxic responses driven by antigen activation.

Together, these features make TCRVγ9V2+ γδ T lymphocytes versatile and attractive candidates for new cancer immunotherapies 10–12. For these reasons, synthetic analogues of natural phosphoantigens such as bromohydrin pyrophosphate (BrHPP) have been produced and the immune functions mediated by phosphoantigen-activated γδ T cells appear promising 13, 14. Preclinical and clinical studies with BrHPP have shown that phosphoantigens induce a potent and rapid response of the TCRVγ9Vδ2+ γδ T lymphocytes in humans and in various species of non-human primate models 15, 16. The unusual behaviour and high versatility of TCRVγ9Vδ2+ γδ T lymphocytes raise several questions on their biology, however. For the design of γδ T-cell-based vaccines, are these cells more related to adaptive αβ T lymphocytes or to NK cells? From a global pharmacological perspective, what is the transcriptional signature of their activation by phosphoantigens? From a physiopathological standpoint, what differs between freshly activated, normal TCRVγ9Vδ2+ γδ T cells, primary cultured cell lines of normal TCRVγ9Vδ2+ γδ T cells and peripheral T-cell lymphomas (PTCLs) of γδ subtype?

To clarify these issues here we report the transcriptomes of highly purified TCRVγ9Vδ2+ γδ T cells from healthy individuals and analyse them relative to those of other subsets of human lymphoid cells. These data sets have been deposited at NCBI GEO repository and are freely downloadable under accession number GSE27291.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References
  9. Supporting Information

The transcriptomes of normal TCRVγ9+ γδ T cells cluster together and next to αβ T and NK cells

The expression profiles of the 12 TCRVγ9+ γδ T-cell samples were compared with those of normal B, αβ and γδ T, NK cells and of peripheral T and NKT cell lymphomas (a total of 90 entire transcriptomes obtained on the same platform). The unsupervised clustering of these transcriptomes formed three main clusters based on differential expression of several thousands of genes. These three clusters were correlated quite well to the cell lineage (B, T, NK cells, R2=0.68) and the normal versus cancer status of cell samples (R2=0.84). Further subdivision of the clusters I–III created nine groups that encompassed three groups of B cells composing the previous cluster I, five groups in the T and NK cell cluster II, while cluster III comprised two lymphoma groups: PTCLs and NKT cell lymphomas (NKTCLs).

Cluster I comprised all the normal B-cell samples, the cluster II encompassed both T and NK cells from normal samples and established cell lines, whereas cluster III encompassed the freshly collected samples of PTCLs and NKTCLs. The T and NK cell cluster II encompassed two branches comprising the cultured cell lines on the first and freshly isolated cells on the other, which subdivided into two subgroups of T cells in one branch and all NK cells plus the CD8+ T cells in the other branch.

In cluster II, the freshly isolated, resting γδ T cells and their counterparts obtained 6 h after activation with the BrHPP phosphoantigen segregated with freshly isolated αβ CD4+ T cells. After 7 days of activation with BrHPP, however, the TCRVγ9+ γδ T cells clustered primarily with ‘normal γδ T cells’ (cluster 7) that were established by 2 weeks culture with zoledronate 17 and then with established cell lines of γδ T cells and NK cells (cluster 8). This clustering most likely reflects the mode of action of zoledronate on peripheral blood mononuclear cells, since this aminobisphosphonate is a phosphoantigen-inducing γδ T-cell agonist targeting the same TCRVγ9+ lymphocytes 18. Of note, additional transcriptomes of freshly isolated monocytes or differentiated macrophages produced in our laboratory 19 clustered outside of the whole lymphoid cell panel involved in this study (data not shown).

Cluster III was essentially composed of the primary PTCLs on the one hand and of a more heterogeneous subgroup comprising both cytotoxic αβ and γδ PTCLs plus the NKTCLs. This sub-clustering we obtained here with entire transcriptomes matched perfectly to the recently depicted clustering of PTCLs and extranodal NK/T-cell lymphomas of nasal type based on 762 differentially expressed genes 20 (Fig. 1).

thumbnail image

Figure 1. Unsupervised hierarchical clustering of 90 transcriptomes (20 606 genes) from γδ T cells and other lymphoid cell subtypes. Shown are 26 samples of normal B cells of naïve, memory, centroblastic and centrocytic types 52, 53, 6 samples of αβ T cells of CD4+ and CD8+ types 17, 54, 55, 13 samples of NK cells (this study and 55), 14 samples of TCRVγ9+ γδ T cells (this study and 17), 6 EBV+ γδ T-cell lines from EBV-infected patients 17, 16 peripheral T-cell lymphomas (PTCLs) 21 and 9 NKT cell lymphomas (primary tumours and cell lines) 20.

Download figure to PowerPoint

Therefore, the gene signature of normal TCRVγ9+ γδ T lymphocytes overlaps those from both αβ cells and NK cells.

Genes differentially expressed in γδ TCRVγ9+ cells relative to αβ T cells and to NK cells

Although in the hierarchical clustering, the normal γδ TCRVγ9+ cells were most closely related to normal αβ T and NK cells, there were clear differences in their respective gene expression patterns (Fig. 2). Seven thousand eight hundred forty-four genes were differentially (p<0.05) expressed by γδ TCRVγ9+ and αβ T cells, with 3379 genes up-regulated and 4465 down-regulated by the γδ TCRVγ9+ cells, while 11 264 genes were differentially expressed by γδ TCRVγ9+ and NK cells, with 5617 genes up-regulated and 5646 down-regulated by the γδ TCRVγ9+ T cells. Representative genes up-regulated by the γδ TCRVγ9+ cells relative to αβ T lymphocytes (as a whole) comprised genes encoding for the cell surface receptors γδ TCR, CD94 and NKG2D, which are well-known phenotypic markers of this lineage (Table 1). Indeed, although NKG2D is expressed on most γδ TCRVγ9+ cells, it is also found on a substantial proportion of CD8+ αβ T cells but barely on the CD4+ T cells (unless activated).

thumbnail image

Figure 2. Differential transcriptomes from freshly isolated B, αβ T, γδ T and NK cells. The most representative genes up-regulated by the γδ T cells relative to either αβ T cells or NK cells are listed by enriched gene ontology terms.

Download figure to PowerPoint

Table 1. Representative genes up-regulated in TCRVγ9+ γδ T cells
Gene symbolNamep-ValueFold change
Relative to αβ T cells
TRA, TRDTCR α and TCR δ locus4×10−71.56
TRGC2, TRGV2TCR γ locus0.0051.43
KLRK1NKG2D0.0161.57
KLRC1, KLRC2NKG2A, NKG2C7×10−52.73
KLRC3NKG2E0.0021.68
KLRC4NKG2F0.0291.57
KLRD1CD940.0041.59
KLRF1NKp800.0121.44
CD160NK cell activating/inhibitory receptor (By55)0.0011.50
KIR3DL2Killer cell Ig-like receptor, three domains, long cytoplasmic tail, 20.0001.38
NKG7GMP-170.0171.63
GZMBGranzyme B0.0141.66
GNLYGranulysin0.0121.56
TNFSF14LIGHT, lymphotoxin γ1×10−61.58
FASLGFas ligand3×10−61.48
IL18RAPInterleukin 18 receptor accessory protein2×10−61.66
KSP37Killer-specific secretory protein of 37 kDa0.0401.63
XCL1Lymphotactin, SCM-1β0.0081.69
XCL2Chemokine (C motif) ligand 20.0101.58
CCL3MIP-1α0.0091.56
CCL4MIP-1β0.0021.63
CCL5RANTES0.0211.47
TBX21Tbet transcription factor0.011.42
Relative to NK cells
CD28CD28 cell surface marker1.0×10−81.59
CD3DCD3δ molecule (CD3-TCR complex)6.4×10−61.24
TRAT1TCR-associated transmembrane adaptor 1, TRIM13.0×10−71.57
ICOSInducible T-cell costimulator3.0×10−81.36
CD5CD5 molecule2.0×10−71.50
LAG3Lymphocyte-activation gene 31.1×10−51.45
TNFRSF25DR3, LARD, APO-33.6×10−91.45
TNFSF8CD153 marker (CD30 ligand)2.2×10−81.36
TNFRSF25DR3, APO-3, LARD9.9×10−101.38
IL7RInterleukin 7 receptor1.7×10−51.38
IL21RInterleukin 21 receptor10−41.37
CXCR6Chemokine (C-X-C motif) receptor 61.6×10−101.62
CCL20MIP-3α0.0091.41
DMNDesmuslin (intermediate filament protein)8.2×10−81.87
PERPTP53 apoptosis effector1.6×10−81.60
PBX4Pre-B-cell leukaemia transcription factor 41.6×10−61.45

The 500 genes most significantly up-regulated by normal TCRVγ9 γδ T cells relative to αβ T cells were significantly enriched in genes from several related functional pathways from KEGG, Biocarta and GSEA C2 databases. These pathways comprised cytokine–cytokine receptor interactions (28 genes, p=8×10−12), NK cell-mediated cytotoxicity (16 genes, p=5.8×10−8), Jak-STAT signalling (12 genes, p=2×10−4) and Th1 cell functions (20 genes, p=10−9, geneset GS11 defined de Leval, Rogge and Chtanova) 21.

On the other hand, representative genes up-regulated by normal γδ T cells relative to NK cells comprised T-cell specifying genes such as CD3D and TRAT, which encore for the TCR-associated CD3δ molecule and the TCR-associated trans-membrane adaptor TRIM1, respectively (Table 1). The 500 genes most significantly up-regulated by γδ T cells relative to NK cells comprised genes encoding for cytokine–cytokine receptor interactions (15 genes, p=7.8×10−4, KEGG), MAPK signalling (12 genes, p<10−2, KEGG), NKT pathway (6 genes, p<10−4, Biocarta) and TCR/CTLA4 pathway (5 genes, p<10−4, KEGG/Biocarta, respectively). These results cannot be accounted for by differential culture conditions of the T-cell and NK-cell samples, respectively, since both cell types comprise both resting un-stimulated cells and IL-2-activated cells. Hence, the TCRVγ9 γδ T cells do not have their own unique signature (Fig. 2), since they merge hallmarks of T cells and NK-cells, such as more ‘NK cell’ genes than αβ T cells and more ‘T-cell’ genes than NK cells.

These results were validated by measuring with RT-qPCR the mRNA expression levels of representative genes from of TCRVγ9 γδ T cells as compared with those of purified αβ T cells and NK cells (n=6 independent samples). The expression levels of the KLRK1, KLRC1, KLRC2, KLRC3, KLRC4 and KLRD1 genes were significantly (p<0.05) over-expressed by 34-, 58- 66-, 11-, 19- and 66-fold, respectively, by the TCRVγ9 γδ T cells relative to αβ T cells while they over-expressed the CCL3, CCL4 and CCL5 genes by 3-, 6- and 6-fold, respectively. Likewise, the expression levels of the CD28, CD3, ICOS and PERP genes were significantly (p<0.05) over-expressed by 7-, 35-, 3- and 12-fold, respectively, by the TCRVγ9 γδ T cells relative to NK cells (Fig. 3A). A discrepancy between gene expression and phenotypes of killer Ig-type receptors has been reported, however 22, so we asked whether these mRNA expression patterns correlated to their surface phenotype by immunostaining and flow cytometry. For the representative set of cell surface antigens CD3ε, CD4, CD8α, CD8β, CD16, CD32, CD64, CD85j, CD94, CTLA4, KIR2DL1, KIR3DL1, KIR2DL3, NKp30, NKp44, NKp46, NKG2A, NKG2D and PD-1, the mRNA level matched with the cell surface phenotype of control TCRVγ9+ γδ T cells (Fig. 3B).

thumbnail image

Figure 3. Validation of transcriptome data. (A) RT-qPCR of representative genes. Result shows the means and SD (n=6 replicates from independent donors) for the expression of the indicated genes relative to GAPDH in the specified cell samples. *p<0.05, one-way paired Student's t-test. (B) Plot of mean of arbitrary units of mRNA expression (microarray-based) versus mean of fluorescence intensity (FACS-based) for cell surface expression of phenotypic markers on the resting control TCRVγ9 γδ T cells isolated from (n>10) healthy individuals. (C) Divergent changes of cell surface phenotype and mRNA expression (microarray-based) in phosphoantigen-activated TCRVγ9+ γδ T cells for TCR, CD3ε, NKG2D, CD94, KIRs, NCRs and other markers of these cells.

Download figure to PowerPoint

This hybrid gene expression profile positions normal human TCRVγ9+ γδ T lymphocytes at the interface of adaptive T lymphocytes and innate NK cells, in good agreement with their physiological development and functions 23.

Molecular signature of phosphoantigen-activated γδ T cells

The transcription profiles of the four human γδ T-cell samples after activation by BrHPP phosphoantigen and culture with IL-2 were compared with those from the same samples before activation. Labelling the purified γδ T-cell samples for intracellular Phospho-ZAP70 and flow cytometry indicated that the cell isolation procedure did not activate even a minor rate of the cells prior to stimulation with BrHPP and IL-2, in line with the baseline level of EGR gene expression in the control samples. From the 15 506 genes which expression was detectable in all samples, a total of 2573 genes were significantly modulated (p<0.01) 6 h after activation, including 1583 over-expressed genes and 990 down-regulated genes. Six hours after activation, the most over-expressed genes were related to early activation (e.g. EGR2, EGR3) and function of T lymphocytes (e.g. IFNG, LIF, TNFA, LAMP3). At this time point, these cells secreted IFN-γ MIP-1α and MIP-1β accordingly 24. All the genes over-expressed at this early time point corresponded to functional categories from the KEGG's pathways defined as TCR signalling (14 genes, p<0.0009), NK cell cytotoxicity (15 genes, p<0.02), Jak-Stat signalling (21 genes, p<0.0002), as well as purine and pyrimidine metabolisms (21 and 14 genes, respectively, p<0.0009 and <0.0005, respectively; Table 2).

Table 2. Representative genes up-regulated in phosphoantigen-activated TCRVγ9+ γδ T lymphocytes
Gene symbolNamep-ValueFold changea)
  • a)

    a) Over resting cells.

6 h after activation
IFNGInterferon γ5.9×10−342
LIFLeukaemia inhibitory factor2.6×10−331
EGR2Early growth response 26.4×10−420
EGR3Early growth response 31.5×10−328
LAMP3Lysosomal-associated membrane protein 38.0×10−324
CCL3MIP-1α6.3×10−418
CCL4MIP-1β4.9×10−38
IL32Interleukin 328.0×10−318
PTGER2Prostaglandin E receptor 2 (subtype EP2)1.5×10−313
CARD8Apoptotic protein NDPP1 (CARDINAL)3.2×10−38
TNFRSF9CD137 marker (accessory molecule 4-1BB)7.0×10−319
DPP4CD26 marker (adenosine deaminase complexing protein 2)3.4×10−37
CD96CD96 molecule (Tactile)2.1×10−35
RAB23RAS oncogene family member1.5×10−35
CDK5Cyclin-dependent kinase 56.8×10−35
BCL2B-cell lymphoma 24.9×10−45
BTN3A2Butyrophilin subfamily 3 member A24.8×10−35
7 days after activation
GINS1GINS complex subunit 1 (Psf1 homolog, actively cycling cell marker)4.8×10−3761
FOXM1G2/M transition-regulating transcription factor3.6×10−5465
HMMRHyaluronan-mediated motility receptor (RHAMM)2.9×10−4436
IL32Interleukin 326.5×10−4280
CEP55Centrosomal protein 55 kDa1.6×10−3251
CENPKCentromere protein K3.9×10−6204
CDC20CDC20 cell division cycle 20 homolog2.8×10−4195
CDC45LCDC45 cell division cycle 45-like1.5×10−3172
SPBC25Spindle pole body component 25 homolog3.1×10−4126
AURKBAurora kinase B1.7×10−3119
CDC2Cell division cycle 2, G1 to S and G2 to M1.6×10−3116
CENPMCentromere protein M7.2×10−398
CDCA8Cell division cycle associated 89.6×10−593
KIF15Kinesin family member 156.6×10−386
SDC4Syndecan 4 (amphiglycan, ryudocan)6.9×10−365
PYCARDPYD and CARD domain containing1.2×10−363
TOP2ATopoisomerase (DNA) II α 170 kDa4.2×10−360
CENPHCentromere protein H9.8×10−358
CCNB2Cyclin B22.6×10−356
KIF2CKinesin family member 2C8.0×10−446
CCNB1Cyclin B11.5×10−445
TYMSThymidylate synthetase3.3×10−344
CCR2Chemokine (C–C motif) receptor 24.5×10−343
KNTC2Kinetochore-associated 21.7×10−443
NUSAP1Nucleolar and spindle-associated protein 19.1×10−541

Seven days after activation and culture, the expression of a total of 3081 genes was significantly modified including 1835 over-expressed genes and 1246 down-regulated genes. At this time point, most of the over-expressed genes reflected segregation of chromosomes (e.g. KIF15, CENPK, CENPM, CEP55, SPBC25) and cell cycle (AURKB, CDCA8, CDC2, CDC20, CDC45L). The other up-regulated genes corresponded to KEGG's functional categories of purine and pyrimidine metabolisms (49 and 33 genes, p<10−18 and <10−15, respectively), regulation of actin cytoskeleton (41 genes, p<10−4), cell cycle (49 genes, p<10−40), focal adhesion (31 genes, p=0.037), MAPK signalling (57 genes, p<10−8), Wnt signalling (28 genes, p<0.003) and Jak-Stat signalling (28 genes, p<0.004; Table 2).

These results were confirmed by RT-qPCR of mRNA for representative genes from TCRVγ9 γδ T cells, by comparing resting and activated TCRVγ9 γδ T cells (n=6 independent samples). After 6 h of activation, the expression levels of TNF-α, IFN-γ, LIF, EGR2, CCL3 and CCL4 genes were significantly (p<0.05) over-expressed by 118-, 433-, 14-, 368-, 230- and 39-fold, respectively, whereas those of CCL5 and IL-32 were not changed. After 7 days of activation, the expression levels of IL-32, CDC20 and TOP2A genes were significantly (p<0.05) over-expressed by 5-, 17- and 43-fold, respectively, whereas FOXM1 and CCNB1 were almost unchanged (Fig. 3A). Some markers such as CTLA-4 showed good consistency for mRNA and cell surface protein along activation, as both were low in resting controls and strongly increased by activation (Fig. 3B and C). However, the correlate of mRNA expression with cell surface phenotype was more generally lost with activated TCRVγ9+ γδ T cells. KLRK1/NKG2D mRNA expression increased 6 h after activation but returned to baseline by day 7, whereas cell surface NKG2D was strongly increased at this time point. The mRNA expression of the CD28 gene was unchanged by day 7 while CD28 protein decreased at the cell surface, possibly reflecting effector memory maturation. The mRNA for PDCD1 decreased by day 7 while the encoded PD-1 protein strongly increased at the cell membrane (Fig. 3C), as recently reported 25. Hence, although the mRNA and phenotype of resting TCRVγ9 γδ T cells were rather matched, activation with phosphoantigen introduced kinetic changes, which generally dissociated this match. The recurrent gene signatures of cytokine/cytokine receptor, JAK/STAT and MAPK pathways reflected the IL-2-dependent conditions of γδ T cell activation with BrHPP. Of note, however, none of the functional associations of the genes up-regulated by activation corresponded to antigen presentation.

These experiments demonstrated that the molecular signature of γδ T cells activated by phosphoantigens corresponds primarily to cytolytic, cytokine and chemokine activities and further to clonal expansion but not to the professional antigen-presenting functions observed in some culture conditions 8, 26, 27.

Transcriptomes of TCRVγ9+ γδ T-cell lines from healthy individuals or lymphoma patients

It was important to determine whether the molecular signature of such cells from EBV-infected patients, which comprise not only BZLF1 but also FGF14, PDCD4, CDK2, HSP90, IL12A and TNFRSF10D 28, did overlap that of normal cells strongly activated by more physiological agonists. The transcription profiles of TCRVγ9+ γδ T-cell lines established from healthy donors (cluster 7) were thus compared with those of cell lines derived from patients with EBV-positive TCRVγ9+ γδ T lymphoproliferative disorders 17 and with the related cell lines SNK6 and SNT7 20 (cluster 8). The TCRVγ9+ γδ T-cell lines from healthy individuals had up-regulated genes responsible for TCR signalling pathway (p<10−8, KEGG), chemokine and cytokine signalling (p<10−5, KEGG and Biocarta), cytotoxicity (p<10−9, KEGG and Biocarta) and CTLA4 pathways (p<5×10−4, Biocarta), cell development, growth and apoptosis (p<10−5, GO Biological Process). On the other hand, there was no gene signature overlap with cell lines from lymphoma patients which over-expressed genes that were rather devoted to the metabolic needs of highly proliferating cells, such as DNA and RNA metabolisms (p<10−6, KEGG and GO Biological Process), oxidative phosphorylation (p<10−6, KEGG and GO Biological Process), transcription (p<2×10−6, KEGG and GO Biological Process) and cell proliferation (p<2×10−8, KEGG and GO Biological Process; Table 3).

Table 3. Representative genes up-regulated by activated cell lines from healthy donors and lymphoma patients
Gene symbolNamep-ValueFold changea)
  • a)

    a) Fold change of log2-transformed, normalized data.

Up-regulated by cell lines from healthy donors (cluster 7)
EGFL6EGF-like-domain protein 6 (promotes matrix assembly)1.1×10−52.30
GIMAP7Immunity-associated GTPase family member 72.7×0−82.29
GIMAP8Immunity-associated GTPase family member 81.6×10−102.00
GIMAP1Immunity-associated GTPase family member 11.9×10−91.81
GIMAP4Immunity-associated GTPase family member 410−41.77
KLRB1NKRP1A (CD161)5.8×10−52.27
KLRC3NKG2E4.6×10−61.78
KLRK1NKG2D10−41.66
KLRC4NKG2F2.9×10−41.65
KLRC1NKG2A0.0021.65
CCR2Chemokine (C–C motif) receptor 24.0×10−42.12
CXCR6Chemokine (C–X–C motif) receptor 62.4×10−51.99
CX3CR1Chemokine (C–X3–C motif) receptor 110−41.97
CD300ACD300a, CMRF35H molecule (NK inhibitory receptor p60)2.0×10−31.73
CD69CD69 molecule (lymphocyte activation marker)1.6×10−61.73
GNLYGranulysin5.8×10−31.77
CRTAMCytotoxic and regulatory T-cell molecule3.1×10−31.66
AMIGO2Adhesion molecule with Ig-like domain 22.0×10−112.15
PERPTP53 apoptosis effector4.7×10−52.07
P4HA2Proline 4-hydroxylase, α polypeptide II9.7×10−41.65
GSPT2G1-to-S phase transition 23.5×10−112.07
SIRPGSignal-regulatory protein γ (CD47 molecule)5.8×10−61.68
Up-regulated by cell lines from lymphoma patients (cluster 8)
DMDDystrophin (cytoskeleton anchoring to plasma membrane)4.8×10−62.11
LPHN2Latrophilin 2 (exocytosis-regulating receptor)10−32.00
MYO3BMyosin IIIB4.8×10−41.90
MMP12Matrix metallopeptidase 12 (macrophage elastase)0.011.86
IL-9Interleukin 9 (IL-2/IL-4-independent T-cell growth factor)0.021.85
DDX4ATP-dependent RNA helicase5.3×10−41.78
SOX2OTSOX2 overlapping transcript (non-coding RNA)0.011.69
RBPMSRNA-binding protein with multiple splicing4.8×10−41.63
TCF4Transcription factor 47.0×10−41.60
CCR7Chemokine (C–C motif) receptor 78.9×10−51.55
POU2F3POU domain. Class 2 transcription factor 30.041.55
ENPP2Ectonucleotide pyrophosphatase (autotaxin)0.011.48

PTCLs of the γδ lineage are rare entities with yet poorly defined oncogenic pathways and biological abnormalities relative to their normal γδ T-cell counterparts 20, 29. Of the 16 PTCLs and 7 NKTCL transcriptomes downloaded for the present meta-analysis, one of these (GD-PTCL_2.GD) corresponded to a lymphoma that was clinically identified as γδ PTCLs 20. This enabled us to compare it by principal component analysis (PCA) with that of the 12 resting and activated γδ T cells. This method determined 38 genes over-expressed by the γδ PTCLs which contributed to discriminate it from the healthy γδ T cells. These genes (Table 4) were involved in the complement pathways (p<10−4, KEGG and Biocarta), cytokine–cytokine receptor signalling (p<2×10−7, KEGG and Biocarta), TLR signalling (p<10−7, KEGG), chemokine–chemokine receptor and G-protein-coupled signallings (p<10−11, from GSEA's C5). In addition, this gene signature corresponded to the genomic region chr4q21 (p<10−6, from GSEA's C2 positional gene set collection). Accordingly, the CXCL9, CXCL10, CXCL11, CXCL13 and IgJ genes that are over-expressed by the γδ PTCLs are all located on chr4q21, so their expression pattern might reflect either genomic amplifications in this γδ PTCL, as recently depicted for unspecified PTCLs and adult T-cell leukaemia/lymphomas 30 or a higher transcriptional activity of this region.

Table 4. Representative genes preferentially expressed by the γδ PTCL relative to normal γδ T cells from healthy donors
Gene symbolNameFold change (γδ PTCLs versus normal γδ T cells)a)
  • a)

    a) Fold change of log2-transformed, normalized data from γδ PTCLs relative to normal γδ T cells.

ADAMDEC1ADAM-like decysin 12,7
C1QAComplement component 1, q sub-component, A chain2,4
C1QBComplement component 1, q sub-component, B chain2,4
C1QCComplement component 1, q sub-component, C chain2,5
CCL19MIP-3β2,4
CCL216Ckine2,0
CD14CD14 marker2,4
CD160NK cell activating/inhibitory receptor (By55)1,8
CD163Scavenger receptor CD1632,2
CST3Cystatin C1,8
CXCL10IP-101,8
CXCL11I-TAC2,4
CXCL13BCA-13,2
CXCL9MIG2,2
ENPP2Ectonucleotide pyrophosphatase/phosphodiesterase 22,3
GABBR1GABA-B receptor, 12,2
GPNMBTransmembrane glycoprotein B2,5
IGHIg heavy locus2,3
IGJIg J linker (for Ig α and Igμ)2,2
IGKCC region of Ig κ chain1,8
IGLIg λ chain1,8
IGLCC region of Ig λ chain2,0
LILRB1CD85, ILT2, LIR11,3
LUMLumican3,0
LY96TLR4-associated MD-2 protein2,3
LYZLysozyme2,2
SERPING1Serpin peptidase inhibitor, clade G member 12,2
SPARCL1SPARC-like 1, hevin2,5
TKTL1Transketolase-like 11,9

These results indicate that γδ T-cell lines from NKTLs and PTCLs had molecular signatures of higher metabolic and proliferative activities than in primary γδ cell lines from healthy individuals, whereas the profile of γδ PTCLs suggested higher chemokine and GPCR gene expression in this malignancy.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References
  9. Supporting Information

In this study, we show that the gene signature of healthy phosphoantigen-specific human γδ T cells mitigates T and NK-like patterns which reflect both of their cytolytic, pro-inflammatory and proliferative activities. With activated, established TCRVγ9+ γδ T-cell lines derived from healthy donors, however, these profiles were related to – but distinctive from – those derived from lymphoma patients, which showed exacerbated metabolic and mitotic pathways as well as increased cytokine, chemokine, TLR and GPCR signalling pathway genes.

Patterns of gene expression are powerful tools to improve our understanding of the biology of the human γδ T lymphocytes, but few studies are currently available on this topic. The first transcriptome studies of murine intra-epithelial γδ T lymphocytes failed to differentiate these cells from their αβ counterparts but concluded show an ‘activated yet resting’ profile 31, 32. Further studies on calf blood γδ T lymphocytes, exploiting the cross-reactivity of bovine genes with Affymetrix human micro-arrays, depicted a γδ T-cell signature hybrid of B and myeloid cells 33, 34 and proposed γδ T cells as ancestral to the other lymphoid subsets 35, 36. An Affymetrix platform-based study of human αβ and γδ T cells purified from PBMCs by microbeads failed to identify γδ T-cell specific transcripts and concluded that αβ and γδ T cells have fully overlapping profiles 37. Nevertheless, the transcriptional profiling (using custom microarrays) of established primary cell lines of TCRVγ9+ γδ T cells stimulated with isopentenyl pyrophosphate depicted induction of Fyn-binding protein, immediate early response 3, proinflammatory chemokines, IFN-γ, TNF-α, lymphotoxin α and CD25 38. Further, Hver 2.1.1 microarray-based study of TCRVγ9+ γδ T cells stimulated with the microbial (E)-4-hydroxy-3-methyl-but-2-enyl-pyrophosphate (HDMAPP) phosphoantigen in the presence of IL-2, IL-4 or IL-21 identified Th1, Th2 cells or follicular dendritic cell (FDC)-like polarisations of the TCRVγ9+ γδ T cells, respectively 39. These studies, however, failed to detect up-regulation of the genes (CD40, CD80, CD86 and HLA-DR) reportedly expressed by TCRVγ9+ γδ T antigen-presenting cells 8. Finally, a recent Affymetrix (HG U133 plus 2.0) microarray-based microarray study of human γδ T cells from cytomegalovirus-infected newborns characterised the signature of TCRVγ8/Vδ1+ cells 40. These lymphocytes up-regulate the expression of NK receptors (both activating and inhibitory), cytolytic mediators and pro-inflammatory chemokines and cytokines. Despite their use of different platforms which generally underestimate the differences between gene expression levels, these and the present study are all consistent in showing that in γδ T cells, the expression levels of mRNA transcripts correlate quite well to presence of cell surface antigens.

The issue of the γδ T cells' relatedness to either B cells, T cells, NK cells or even myeloid-like antigen-presenting cells has been raised in previous studies, but the clustering analysis presented here unambiguously locates these lymphocytes at the interface of T and NK cells. The γδ T and NK cell lineages do share some developmental programming, possibly driven by chronic stimulation 41. Nevertheless, the T-cell transcription factor BCL11b was expressed at the same level in TCRVγ9+ γδ and αβ T lymphocytes, which were significantly higher than in NK cells (p=0.001), in line with its repressive activity for NK cell-associated genes 42. That γδ T cells expressed more T-cell genes than NK cells and more NK-cell genes than T cells had been reported 37, but this mitigated profile was actually resolved temporally upon activation. Although the molecular signature of resting and early activated TCRVγ9+ γδ T cells were more closely like those of CD4+ T cells, those of established and activated TCRVγ9+ cell lines were mostly like those of NK cells, confirming the functional pleiotropy of cells previously demonstrated for human cells within different conditions 39, 43. This is also reminiscent of murine γδ T cells, where TCRVγ1+ T lymphocytes with memory-activated phenotype strongly resemble to bona fide NK cells and are referred to as NK-like γδ T cells 41, 44. In addition, phosphoantigen-activated TCRVγ9+ γδ T cells up-regulated the expression of IL-21R, a receptor which further interaction with IL-21 enables pre-committed γδ T cells to irreversibly trigger their cytotoxic, Th1 and proliferative programming 45, as well as the expression of lymphoid-homing and germinal centre reactions molecules 39. Within the TCRVγ9+ cell samples analysed here, no signature for differentiation of Th17 γδ T cells such as RORγT, IL-17, IL-23 were observed, as with TCRVγ8/Vδ1+ cells 40.

The phosphoantigen-activated TCRVγ9+ γδ T cells consistently up-regulated the expression of the most of the well-depicted Th1 cytokines, and of several classes of inhibitory molecules such as CTLA4, PD-1 and NK-cell receptors as reported elsewhere 37. In addition, the expression of BTN3A2 that encodes the inhibitory butyrophilin 3 46 was increased by activation. Furthermore, gene expression of PTGER2 and the corresponding cell surface phenotype for the EP2 subtype of prostaglandin E2 receptor were strongly up-regulated by activated TCRVγ9+ γδ T cells. Upon binding of PGE2, this receptor mediates a potent AMPc/PKA-dependent blockade of activation 47–49. Neither of the cell-suppressive cytokine TGF-β and its receptors or co-receptors 24 were up-regulated by activation, however. Hence, the gene signature of phosphoantigen-stimulated cells comprised Th1 cytokines but also a whole set of potent negative regulators of activation presumably acting as a normal physiological regulation. In this regard, the gene signature differentiating healthy from pathological γδ T cells was informative despite the low number of relevant transcriptomes available. The genes over-expressed by such cancer cells rather reflected the metabolisms and signalling pathways associated with their high proliferation. Not only the signature of these constitutively activated cancer γδ T cells did not overlap with that of healthy activated γδ T cells, but they lacked the regulatory loops found in healthy antigen-activated cells. Likewise, altered patterns of NK cell receptor expression have recently been reported for non-B-cell lymphomas 17, 50, 51.

In conclusion, this study depicted a hybrid molecular signature for phosphoantigen-specific human TCRVγ9+ γδ T cells, as a blend of the NK cell and αβ T-cell signatures. Future studies using these tools will now aim at characterising the pathways induced in TCRVγ9+ γδ T cells lacking these functional activities, such as those encountered in cancer patients.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References
  9. Supporting Information

Cells and samples

Whole PBMCs were isolated from four healthy donors (Etablissement Français du Sang, Toulouse, France) after Ficoll–Hypaque density centrifugation. TCRVγ9+ cells were purified (>98%) from PBMC by cell sorting using clone IMMU510 (Beckman-Coulter-Immunotech, Marseille, France). Control flow cytometry for intracellular phospho-ZAP70 (Y319) and P-ERK1/2 [using, respectively, Ax647-conjugated anti-phospho-ZAP70 and PE-conjugated phospho-(T202/Y204) ERK1/2 Abs (BD Pharmingen)] indicated this procedure did not activate the purified cells, unless additional stimuli were provided. These γδ T cell samples were purified either before PBMC treatment: ‘γδ TCRVγ9+ control’ samples (n=4), or 6 h after stimulation with the TCRVγ9+-specific agonist BrHPP 13 (500 nM): ‘γδ TCRVγ9+ act 6 h’ samples (n=4), or 7 days after PAg stimulation and culture with IL-2 (100 IU/mL): ‘γδ TCRVγ9+ act 7-day’ samples (n=4). Cells were cultured in vitro as described 24. Freshly purified CD3CD56+ NK cells (>90%, as checked by flow cytometry) were obtained from PBMC by positive selection from magnetic beads (Miltenyi Biotec, Auburn, CA, USA) according to manufacturer's instructions. NK cells were then activated with 100 U/mL IL-2 in complete RPMI medium for 2 days before RNA extraction.

Reagents and flow cytometry

Flow cytometry for cell surface phenotype of γδ T- and NK-cell samples was done as depicted using LSR-II and analysis with the FACS Diva 6.0 (BD Biosciences) or FlowJo (Treestar) softwares 24.

Microarray procedures

Total RNA from the specified human γδ TCRVγ9+ or NK cell samples was isolated using TRIzol™ Reagent (Invitrogen Life Technologies, Paisley, UK). The quality of RNA was assessed with Agilent 2100 Bioanalyser (Agilent Technologies, Palo Alto, CA, USA) after denaturation at 70°C for 2 min. Microarray analyses were performed using 1–3 μg total RNA as the starting material from human cells, amplified and labelled following the one-Cycle Target Labeling protocol (Affymetrix, Santa Clara, CA, USA). The labelled complementary RNA (cRNA) from these samples was then fragmented and hybridised to Affymetrix GeneChip arrays HG-U133 plus 2.0. The chips were then washed, scanned and analysed with GeneChip Operating Software (Version 1.1, Affymetrix) at the Microarray Core Facility of the Institute of Research on Biotherapy, CHRU-INSERM-UM1 Montpellier (http://irb.chu-montpellier.fr/). Microarray data and procedures were deposited at NCBI GEO data set under accession number GSE27291.

Gene expression analysis

The raw data (Affymetrix CEL files) were produced using HG U133-Plus 2.0 platform for the above-depicted 12 samples of highly purified TCRVγ9+ γδ cells (>98% purity) and 9 samples of highly purified, IL-2-activated NK cells (CD3CD56+ cells >98%). For comparison purposes, additional raw data files obtained on the same platform were downloaded from the NCBI repository GEO database and Array Express database. These comprised 26 normal B-cell samples: 18 from GSE12195 52 and 8 from GSE15271 53, 6 samples of normal αβ T CD4+ and CD8+ T cells from GSE15659 54, GSE8059 55 and GSE13906 17, 4 samples of normal NK cells with and without IL-2 from GSE8059 55, 2 samples of normal TCRVγ9+ γδ T-cell lines expanded by 2 weeks of in vitro culture with zoledronate plus IL-2 from GSE13906 17, 6 samples of malignant γδ T cells from GSE13906 17 which comprised 2 nasal lymphoma and 4 EBV+γδ T-cell lines from EBV-infected patients, 16 PTCLs 21, 7 NKTCLs and 2 NKTCL cell lines 20, both downloaded from Array Express (http://www.ebi.ac.uk/arrayexpress) under accession number E-TABM-791. The raw data from these 90 samples were normalised in batch by the RMA software and the 54 676 probe sets were then reduced to a total of 20 606 genes (HUGO symbols) by using the GSEA collapse function set on maximal probe mode (GSEA, http://www.broadinstitute.org/gsea).

Data mining

After log (base 2) transformation, normalisation and collapse, hierarchical clustering of the 90 transcriptomes was based on the 20 606 genes, the Euclidean distance between two transcriptomes and agglomeration by the Ward's criterion 56. Implementations were done using the ‘dist’ and ‘hclust’ functions in R (http://www.R-project.org). The differential transcriptomes from freshly isolated human B, T and NK cells shown in Fig. 2 were represented using dChip (http://www.dchip.org). To identify genes over-expressed by the PTCL_2.GD 20 relative to 12 samples of freshly isolated and purified normal γδ T cells, PCA was done with the ‘prcomp’ and ‘biplot’ functions in R. The sub-space determined by PCA captures the highest amount of the total data set's variability which biplot summarises the relationships between genes and samples 57. Genes differentially expressed between two groups of samples were defined using ANOVA or one-way Student's t-tests whenever appropriate by using the SigmaStat 12.0 software (Systat Software, Chicago, IL, USA). Text files were generated from the gene lists with one gene name per line; these text files were then uploaded in Autocompare. More than 5000 genes reference lists based on GSEA (http://www.broadinstitute.org/gsea/) pathways and 162 protein lists based on PANTHER pathways (http://www.pantherdb.org/pathway/) were collected. The differentially expressed gene subsets were analysed for enrichment in functionally related genes among lists downloaded from the gene sets collection. Selective enrichment analysis was computed with the Autocompare freeware that we developed from nwCompare 58 based on one-sided hypergeometric, Bonferroni and Holm tests. Autocompare was developed using the Perl programming language (Perl v5.10.1, http://www.perl.org/) and the R statistical programming language under the Linux operating system (ubuntu 10.04, http://www.ubuntu.com/). Autocompare is available for Linux and Windows (http://www.ifr150.toulouse.inserm.fn/en/article.asp?id=264) and runs on any operating system with Perl, either as a command line tool or with a graphical interface. As input, it takes any proteomic/genomic data files and performs strings comparisons by line, with any string including protein names, accession numbers or gene chip probesets.

RT-qPCR

Specific genes from αβ T cells, NK cells or TCRVγ+ γδ T cells were selected for verification with RT-qPCR. Preference was given to specific functions of each cell type, i.e. cytotoxicity, cytokine/receptor and proliferation. Briefly, 500 ng of total RNA was reverse transcribed using SuperScript™ III Reverse Transcriptase (Invitrogen, Carlsbad, CA, USA). Gene-specific primers (Supporting Information S1) were used for qPCR on the LightCycler® 480 Real-Time PCR System (Roche Applied Science, Mannheim, Germany). The quantification of each gene expression was performed using LightCycler® 480 SYBR Green I Master with 4 μL of RT product (1:5 dilution) and primers (3 μM). A cycle threshold (Ct) was assigned at the beginning of the logarithmic phase of PCR amplification, and the difference in the Ct values of the control and experimental samples were used to determine the relative expression of the gene in each sample. GAPDH was used for normalisation as its expression did not significantly change along the different real-time PCR experiments. Statistical analysis was performed with α=0.05 in Student's and Mann–Whitney's tests whenever appropriate by using SigmaPlot 12.0 software.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References
  9. Supporting Information

This work was supported in part by institutional grants from the Institut National de la Santé et de la Recherche Médicale (INSERM), Université de Toulouse, Centre National de la Recherche Scientifique and by contracts RITUXOP (PAIR LYMPHOME), V9V2TER and TUMOSTRESS from Institut National du Cancer. We thank Véronique Pantesco (Microarray Core Facility of the Institute of Research on Biotherapy, CHRU-INSERM-UM1 Montpellier) for the Microarray Core Facility, Philippe Gaulard, Marion Travert, Laurence de Leval and Aurélien de Reynies for genesets from NK/T-cell lymphoma transcriptomes. We are grateful to Innate Pharma for clinical grade batches of BrHPP and Sanofi (Toulouse, France) for recombinant hIL-2.

Conflict of interest:

The authors declare no financial or commercial conflict of interest.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References
  9. Supporting Information
  • 1
    Sireci, G., Champagne, E., Fournie, J. J., Dieli, F. and Salerno, A., Patterns of phosphoantigen stimulation of human Vgamma9/Vdelta2 T cell clones include Th0 cytokines. Hum. Immunol. 1997. 58: 7082.
  • 2
    Boullier, S., Paoquet, Y., Debord, T., Fournie, J. J. and Gougeon, M. L., Regulation by cytokines (IL-12, IL-15, IL-4 and IL-10) of the Vgamma9Vdelta2T cell response to mycobacterial phosphoantigens in responder and anergic HIV-infected persons. Eur. J. Immunol. 1999. 29: 9099.
  • 3
    Caccamo, N., Battistini, L., Bonneville, M., Poccia, F., Fournie, J. J., Meraviglia, S., Borsellino, G. et al., CXCR5 identifies a subset of Vgamma9Vdelta2T cells which secrete IL-4 and IL-10 and help B cells for antibody production. J. Immunol. 2006. 177: 52905295.
  • 4
    Angelini, D. F., Borsellino, G., Poupot, M., Diamantini, A., Poupot, R., Bernardi, G., Poccia, F. et al., FcgammaRIII discriminates between two subsets of Vgamma9Vdelta2 effector cells with different responses and activation pathways. Blood 2004. 104: 18011807.
  • 5
    Casetti, R., Agrati, C., Wallace, M., Sacchi, A., Martini, F., Martino, A., Rinaldi, A. et al., Cutting edge: TGF-beta1 and IL-15 Induce FOXP3+ gammadelta regulatory T cells in the presence of antigen stimulation. J. Immunol. 2009. 183: 35743577.
  • 6
    Ness-Schwickerath, K. J., Jin, C. and Morita, C. T., Cytokine requirements for the differentiation and expansion of IL-17A- and IL-22-producing human Vgamma2Vdelta2T cells. J. Immunol. 2010. 184: 72687280.
  • 7
    Wu, Y., Wu, W., Wong, W. M., Ward, E., Thrasher, A. J., Goldblatt, D., Osman, M. et al., Human gamma delta T cells: a lymphoid lineage cell capable of professional phagocytosis. J. Immunol. 2009. 183: 56225629.
  • 8
    Brandes, M., Willimann, K. and Moser, B., Professional antigen-presentation function by human gammadelta T Cells. Science 2005. 309: 264268.
  • 9
    von Lilienfeld-Toal, M., Nattermann, J., Feldmann, G., Sievers, E., Frank, S., Strehl, J. and Schmidt-Wolf, I. G., Activated gammadelta T cells express the natural cytotoxicity receptor natural killer p 44 and show cytotoxic activity against myeloma cells. Clin. Exp. Immunol. 2006. 144: 528533.
  • 10
    Kunzmann, V. and Wilhelm, M., Anti-lymphoma effect of gammadelta T cells. Leuk. Lymphoma 2005. 46: 671680.
  • 11
    Bonneville, M. and Scotet, E., Human Vgamma9Vdelta2T cells: promising new leads for immunotherapy of infections and tumors. Curr. Opin. Immunol. 2006. 18: 539546.
  • 12
    Kabelitz, D., Wesch, D. and He, W., Perspectives of gammadelta T cells in tumor immunology. Cancer Res. 2007. 67: 58.
  • 13
    Espinosa, E., Belmant, C., Pont, F., Luciani, B., Poupot, R., Romagne, F., Brailly, H. et al., Chemical synthesis and biological activity of bromohydrin pyrophosphate, a potent stimulator of human gamma delta T cells. J. Biol. Chem. 2001. 276: 1833718344.
  • 14
    Sicard, H., Al Saati, T., Delsol, G. and Fournie, J. J., Synthetic phosphoantigens enhance human Vgamma9Vdelta2T lymphocytes killing of non-Hodgkin's B lymphoma. Mol. Med. 2001. 7: 711722.
  • 15
    Sicard, H., Ingoure, S., Luciani, B., Serraz, C., Fournie, J. J., Bonneville, M., Tiollier, J. et al., In vivo immunomanipulation of V gamma 9V delta 2T cells with a synthetic phosphoantigen in a preclinical nonhuman primate model. J. Immunol. 2005. 175: 54715480.
  • 16
    Chargui, J., Combaret, V., Scaglione, V., Iacono, I., Peri, V., Valteau-Couanet, D., Dubrel, M. et al., Bromohydrin pyrophosphate-stimulated V gamma 9delta2 T cells expanded ex vivo from patients with poor-prognosis neuroblastoma lyse autologous primary tumor cells. J. Immunother. 2010. 33: 591598.
  • 17
    Zhang, Y., Ohyashiki, J. H., Shimizu, N. and Ohyashiki, K., Aberrant expression of NK cell receptors in Epstein–Barr virus-positive gammadelta T-cell lymphoproliferative disorders. Hematology 2010. 15: 4347.
  • 18
    Kunzmann, V., Bauer, E., Feurle, J., Weissinger, F., Tony, H. P. and Wilhelm, M., Stimulation of gammadelta T cells by aminobisphosphonates and induction of antiplasma cell activity in multiple myeloma. Blood 2000. 96: 384392.
  • 19
    Fruchon, S., Poupot, M., Martinet, L., Turrin, C. O., Majoral, J. P., Fournie, J. J., Caminade, A. M. et al., Anti-inflammatory and immunosuppressive activation of human monocytes by a bioactive dendrimer. J. Leukoc. Biol. 2009. 85: 553562.
  • 20
    Huang, Y., de Reynies, A., de Leval, L., Ghazi, B., Martin-Garcia, N., Travert, M., Bosq, J. et al., Gene expression profiling identifies emerging oncogenic pathways operating in extranodal NK/T-cell lymphoma, nasal type. Blood 2010. 115: 12261237.
  • 21
    de Leval, L., Rickman, D. S., Thielen, C., Reynies, A., Huang, Y. L., Delsol, G., Lamant, L. et al., The gene expression profile of nodal peripheral T-cell lymphoma demonstrates a molecular link between angioimmunoblastic T-cell lymphoma (AITL) and follicular helper T (TFH) cells. Blood 2007. 109: 49524963.
  • 22
    Leung, W., Iyengar, R., Triplett, B., Turner, V., Behm, F. G., Holladay, M. S., Houston, J. et al., Comparison of killer Ig-like receptor genotyping and phenotyping for selection of allogeneic blood stem cell donors. J. Immunol. 2005. 174: 65406545.
  • 23
    Bonneville, M., O'Brien, R. L. and Born, W. K., Gammadelta T cell effector functions: a blend of innate programming and acquired plasticity. Nat. Rev. Immunol. 2010. 10: 467478.
  • 24
    Capietto, A. H., Martinet, L., Cendron, D., Fruchon, S., Pont, F. and Fournie, J. J., Phosphoantigens overcome human TCRVγ9+ gamma delta cell immunosuppression by TGF-beta: relevance for cancer immunotherapy. J. Immunol. 2010. 184: 66806687.
  • 25
    Iwasaki, M., Tanaka, Y., Kobayashi, H., Murata-Hirai, K., Miyabe, H., Sugie, T., Toi, M. et al., Expression and function of PD-1 in human gammadelta T cells that recognize phosphoantigens. Eur. J. Immunol. 2011. 41: 345355.
  • 26
    Brandes, M., Willimann, K., Bioley, G., Levy, N., Eberl, M., Luo, M., Tampe, R. et al., Cross-presenting human gammadelta T cells induce robust CD8+ alphabeta T cell responses. Proc. Natl. Acad. Sci. USA 2009. 106: 23072312.
  • 27
    Meuter, S., Eberl, M. and Moser, B., Prolonged antigen survival and cytosolic export in cross-presenting human gammadelta T cells. Proc. Natl. Acad. Sci. USA 2010. 107: 87308735.
  • 28
    Zhang, Y., Ohyashiki, J. H., Takaku, T., Shimizu, N. and Ohyashiki, K., Transcriptional profiling of Epstein–Barr virus (EBV) genes and host cellular genes in nasal NK/T-cell lymphoma and chronic active EBV infection. Br. J. Cancer 2006. 94: 599608.
  • 29
    de Leval, L., Bisig, B., Thielen, C., Boniver, J. and Gaulard, P., Molecular classification of T-cell lymphomas. Crit. Rev. Oncol. Hematol. 2009. 72: 125143.
  • 30
    Nakagawa, M., Nakagawa-Oshiro, A., Karnan, S., Tagawa, H., Utsunomiya, A., Nakamura, S. et al., Array comparative genomic hybridization analysis of PTCL-U reveals a distinct subgroup with genetic alterations similar to lymphoma-type adult T-cell leukemia/lymphoma. Clin. Cancer Res. 2009. 15: 3038.
  • 31
    Fahrer, A. M., Konigshofer, Y., Kerr, E. M., Ghandour, G., Mack, D. H., Davis, M. M. and Chien, Y. H., Attributes of gammadelta intraepithelial lymphocytes as suggested by their transcriptional profile. Proc. Natl. Acad. Sci. USA 2001. 98: 1026110266.
  • 32
    Shires, J., Theodoridis, E. and Hayday, A. C., Biological insights into TCRgammadelta+ and TCRalphabeta+ intraepithelial lymphocytes provided by serial analysis of gene expression (SAGE). Immunity 2001. 15: 419434.
  • 33
    Hedges, J. F., Cockrell, D., Jackiw, L., Meissner, N. and Jutila, M. A., Differential mRNA expression in circulating gammadelta T lymphocyte subsets defines unique tissue-specific functions. J. Leukoc. Biol. 2003. 73: 306314.
  • 34
    Meissner, N., Radke, J., Hedges, J. F., White, M., Behnke, M., Bertolino, S., Abrahamsen, M. et al., Serial analysis of gene expression in circulating gamma delta T cell subsets defines distinct immunoregulatory phenotypes and unexpected gene expression profiles. J. Immunol. 2003. 170: 356364.
  • 35
    Richards, M. H. and Nelson, J. L., The evolution of vertebrate antigen receptors: a phylogenetic approach. Mol. Biol. Evol. 2000. 17: 146155.
  • 36
    Hayday, A. C., [gamma][delta] cells: a right time and a right place for a conserved third way of protection. Annu. Rev. Immunol. 2000. 18: 9751026.
  • 37
    Chtanova, T., Newton, R., Liu, S. M., Weininger, L., Young, T. R., Silva, D. G., Bertoni, F. et al., Identification of T cell-restricted genes, and signatures for different T cell responses, using a comprehensive collection of microarray datasets. J. Immunol. 2005. 175: 78377847.
  • 38
    Chen, L., Cencioni, M. T., Angelini, D. F., Borsellino, G., Battistini, L. and Brosnan, C. F., Transcriptional profiling of gamma delta T cells identifies a role for vitamin D in the immunoregulation of the V gamma 9V delta 2 response to phosphate-containing ligands. J. Immunol. 2005. 174: 61446152.
  • 39
    Vermijlen, D., Ellis, P., Langford, C., Klein, A., Engel, R., Willimann, K., Jomaa, H. et al., Distinct cytokine-driven responses of activated blood gammadelta T cells: insights into unconventional T cell pleiotropy. J. Immunol. 2007. 178: 43044314.
  • 40
    Vermijlen, D., Brouwer, M., Donner, C., Liesnard, C., Tackoen, M., Van Rysselberge, M., Twite, N. et al., Human cytomegalovirus elicits fetal gammadelta T cell responses in utero. J. Exp. Med. 2010. 207: 807821.
  • 41
    Stewart, C. A., Walzer, T., Robbins, S. H., Malissen, B., Vivier, E. and Prinz, I., Germ-line and rearranged Tcrd transcription distinguish bona fide NK cells and NK-like gammadelta T cells. Eur. J. Immunol. 2007. 37: 14421452.
  • 42
    Li, L., Leid, M. and Rothenberg, E. V., An early T cell lineage commitment checkpoint dependent on the transcription factor Bcl11b. Science 2010. 329: 8993.
  • 43
    Caccamo, N., Meraviglia, S., Ferlazzo, V., Angelini, D., Borsellino, G., Poccia, F., Battistini, L. et al., Differential requirements for antigen or homeostatic cytokines for proliferation and differentiation of human Vgamma9Vdelta2 naive, memory and effector T cell subsets. Eur. J. Immunol. 2005. 35: 17641772.
  • 44
    Huang, Y., Jin, N., Roark, C. L., Aydintug, M. K., Wands, J. M., Huang, H., O'Brien, R. L. et al., The influence of IgE-enhancing and IgE-suppressive gammadelta T cells changes with exposure to inhaled ovalbumin. J. Immunol. 2009. 183: 849855.
  • 45
    Thedrez, A., Harly, C., Morice, A., Salot, S., Bonneville, M. and Scotet, E., IL-21-mediated potentiation of antitumor cytolytic and proinflammatory responses of human V gamma 9V delta 2T cells for adoptive immunotherapy. J. Immunol. 2009. 182: 34233431.
  • 46
    Yamashiro, H., Yoshizaki, S., Tadaki, T., Egawa, K. and Seo, N., Stimulation of human butyrophilin 3 molecules results in negative regulation of cellular immunity. J. Leukoc. Biol. 2010. 88: 757767.
  • 47
    Martinet, L., Jean, C., Dietrich, G., Fournie, J. J. and Poupot, R., PGE2 inhibits natural killer and gamma delta T cell cytotoxicity triggered by NKR and TCR through a cAMP-mediated PKA type I-dependent signaling. Biochem. Pharmacol. 2010. 80: 838845.
  • 48
    Martinet, L., Fleury-Cappellesso, S., Gadelorge, M., Dietrich, G., Bourin, P., Fournie, J. J. and Poupot, R., A regulatory cross-talk between Vgamma9Vdelta2T lymphocytes and mesenchymal stem cells. Eur. J. Immunol. 2009. 39: 752762.
  • 49
    Prigione, I., Benvenuto, F., Bocca, P., Battistini, L., Uccelli, A. and Pistoia, V., Reciprocal interactions between human mesenchymal stem cells and gammadelta T cells or invariant natural killer T cells. Stem Cells 2009. 27: 693702.
  • 50
    Lin, C. W., Chen, Y. H., Chuang, Y. C., Liu, T. Y. and Hsu, S. M., CD94 transcripts imply a better prognosis in nasal-type extranodal NK/T-cell lymphoma. Blood 2003. 102: 26232631.
  • 51
    Miyazaki, K., Yamaguchi, M., Imai, H., Kobayashi, T., Tamaru, S., Nishii, K., Yuda, M. et al., Gene expression profiling of peripheral T-cell lymphoma including gammadelta T-cell lymphoma. Blood 2009. 113: 10711074.
  • 52
    Compagno, M., Lim, W. K., Grunn, A., Nandula, S. V., Brahmachary, M., Shen, Q., Bertoni, F. et al., Mutations of multiple genes cause deregulation of NF-kappaB in diffuse large B-cell lymphoma. Nature 2009. 459: 717721.
  • 53
    Caron, G., Le Gallou, S., Lamy, T., Tarte, K. and Fest, T., CXCR4 expression functionally discriminates centroblasts versus centrocytes within human germinal center B cells. J. Immunol. 2009. 182: 75957602.
  • 54
    Miyara, M., Yoshioka, Y., Kitoh, A., Shima, T., Wing, K., Niwa, A., Parizot, C. et al., Functional delineation and differentiation dynamics of human CD4+ T cells expressing the FoxP3 transcription factor. Immunity 2009. 30: 899911.
  • 55
    Dybkaer, K., Iqbal, J., Zhou, G., Geng, H., Xiao, L., Schmitz, A., d'Amore, F. et al., Genome wide transcriptional analysis of resting and IL2 activated human natural killer cells: gene expression signatures indicative of novel molecular signaling pathways. BMC Genomics 2007. 8: 230.
  • 56
    Lebart, L., Morineau, A. and Warwick, K. M., Multivariate Descriptive Statistical Analysis (Correspondence Analysis and Related Techniques for Large Matrices), Chichester John Wiley & Sons, INC., 605 Third Ave., NY, USA, 1984. pp. 304.
  • 57
    Chapman, S., Schenk, P., Kazan, K. and Manners, J., Using biplots to interpret gene expression patterns in plants. Bioinformatics (Oxford, England) 2002. 18: 202204.
  • 58
    Pont, F. and Fournie, J. J., Sorting protein lists with nwCompare: a simple and fast algorithm for n-way comparison of proteomic data files. Proteomics 2010. 10: 10911094.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. References
  9. Supporting Information

Detailed facts of importance to specialist readers are published as ”Supporting Information”. Such documents are peer-reviewed, but not copy-edited or typeset. They are made available as submitted by the authors.

FilenameFormatSizeDescription
eji_201141870_sm_SupplInfo.pdf468KSupplInfo

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.