Circulating TCR gammadelta cells in the patients with systemic lupus erythematosus.

Systemic lupus erythematosus (SLE) is a disorder with a wide range of immunological abnormalities. The results of the studies undertaken in the last decade indicated that SLE pathogenesis was mainly connected with the breakdown of the activation control of B and T cells, generating humoral or cell-mediated responses against several self-antigens of affected cells. The last studies demonstrate that the role of gammadelta T lymphocytes in autoimmune diseases can be especially important. Flow cytometry techniques were used to investigate the number and percentage of TCR gammadelta T cells and their most frequent subtypes in peripheral blood of 32 patients with SLE and 16 healthy volunteers. We also correlated TCR gammadelta cells number with the level of T CD3+, T CD4+, T CD8+, and NK (CD16) cells (cytometric measurements) and SLE activity (on the basis of clinical investigations). Our studies were preliminary attempts to evaluate the role of that minor T cell subpopulation in SLE. Absolute numbers of cells expressing gammadelta TCR in most SLE blood specimens were significantly lower than in the control group (P<0.006). However, since the level of total T cell population was also decreased in the case of SLE, the mean values of the percentage gammadelta T cells of pan T lymphocytes were almost the same in both analysed populations (7.1% vs 6.3%, respectively). In contrast to Vdelta2+ and Vgamma9+ subtypes of pan gammadelta T cells, Vdelta3+ T cells number was higher in SLE patients (20 x 10 cells/microl) than in healthy control group (2 x 2 cells/microl) (P=0.001). However, we found no differences between the numbers of pan gammadelta T lymphocytes and studied their subtypes in the patients with active and inactive disease. These cell subpopulations were doubled in the treated patients with immunosuppressive agents in comparison with untreated ones; however, data were not statistically significant. Our study indicated that Vdelta3+ subtype of gammadelta T cells seems to be involved in SLE pathogenesis; however, we accept the idea that the autoimmunity does not develop from a single abnormality, but rather from a number of different events.


Introduction
Systemic lupus erythematosus (SLE) is a disorde r w ith a w ide range of immunologic al abnormalities. The disease is characterize d by B cell activation and formation of autoantibodies against nuclear, cytoplasmic and ce ll surface antige ns. 1 ,2 How e ver, increasing evide nce indicate s a critic al role of T cells, particular CD4 + cells in induc ing B ce ll hyperactivity. 3 On the other hand, insufficie nt suppressor ce ll ac tivity may be re sponsible for autoantibody overproduction. De fec tive concanavalin A-induce d suppre ssor ce ll function 4 ,5 and suppre ssor /c ytotox ic re sponse s to Epstein-Barr virus (EBV) 6 have be en demonstrate d in SLE patie nts. Most human mature T c ells ex press a b T-c ell rec eptors on the membrane (TCR a b ) in association w ith the signal transduc tion of CD3 complex . These TCR a b T ce lls are the ce ntral lymphoc ytes in the immune system. The y p rovide spe cific pathoge n re cognition and long-te rm me mory all w ithin the c ontex t of distinguishing foreign from se lf antige ns. 7 Subpopulation of T-ce ll, w hich ex presses TCR g d rec eptors (g d T ce lls) w as identifie d 15 ye ars ago 8 ,9 but its significance in immune re sponse and pathogenesis of diffe re nt dise ases is still poorly unde rstood. TCR g d rece ptors repre sent a disulp hide-linke d heterodimer composed of re arranged g and d chains homologous to the TCR a b and Ig he avy and light chains, w hich contribute to spec ific ity of T ce lls and B ce lls. 10 Human g d T cells range from 1% to 15% of peripheral blood lymp hocyte s and show a predile ction for the red pulp of the splee n and the gastrointe stinal tract. 1 1,1 2 g d T ce lls play a role in host epithe lial surfac e control and early stage e ngageme nt in immune response against viruse s, bacteria and parasites before the re cruitment of a b T ce lls. 7 The y are able to react w ith antigens both in a major histocompatibility c omp lex (MHC) -restricte d and MHC-unre stric te d fashion. 13 The pathologic re levance of g d c ells in human autoimmune dise ases is sugge ste d by their re ac tivity to highly conse rved stress proteins and by the ac cumulation of g d T ce lls in affec te d organs. 1 4 -17 The possible role of g d T cells in autoimmune dise ase is also raise d by their ability to recognise self antigens. Increase d perc entage s of TCR g d c ells have bee n found in the synovial fluids and synovia of patie nts w ith active rheumatoid arthritis. 1 8,1 9 A number of studies sugge st that g d T c ells play a role in the pathogene sis of systemic sclerosis. 2 0,21 Incre ased numbers of g d T c ells have bee n found in perivascular are as of the skin and bronchoalveolar lavage samples, espec ially in patients w ith rec ently diagnosed disease. Clonal ex pansion of TCR g d T c ells has been also re porte d in the p eriphe ry of patie nts w ith systemic lupus e rythematosus (SLE). 2 2 Furthe rmore , Volc-Platz er e t a l. described preferential ex pansion of Vg 2 /Vd 2 subset in lesions in chronic cutane ous lupus e rythe matosus. 23 In the pre sent study w e measured the number of total c irc ulating TCR g d ce lls and their subpopulations in 32 SLE patients and in 16 healthy volunte ers using monoclonal antibodies against pan-g d , Vd 2, Vd 3, Vg 9 chain regions and flow cytome try te chnique s. We also correlated the number of g d T ce lls w ith T CD3 + , T CD4 + , T CD8 + and NK (CD16) c ells and SLE activity.

Patients
A total of 32 unselec te d patients w ith SLE, 30 w omen and tw o men all fulfilling the 1982 re vise d criteria defined by the American Rheumatism Assoc iation (ARA) 24 w ere include d in our study. The ir me an age w as 43.7 years (range 22 -65 years). The mean duration of the disease w as 84.2 months (range 3 months to 28 ye ars). Te n patie nts had never be en tre ate d w ith steroids or any othe r immunosuppre ssive agents, 22 patients had been treated w ith prednisone and tw o of them w ith azatioprine for some time during the c ourse of their dise ase, but 12 of them had not been tre ate d for at least 4 w e eks be fore the g d T ce ll population analysis.
We include d both patie nts w ith active and inactive disease into the study. Dise ase ac tivity w as sc ore d during a visit to the outpatie nt clinic acc ording to the me thod de scribed by Liang e t a l. 25 Each p atie nt w as asse sse d on tw o separate oc casions, 2 -4 w ee ks apart. The system of Syste mic Lupus Activity Measure (SLAM) include s 24 clinical manifestations and e ight laboratory parameters. The max imum score in this system amounts to 84 points. In our group of patients, the points range d from 6 to 26. We c onside red the sc ore of 0-10 points indicative of inactive dise ase, and a score of over 10 p oints indicative of ac tive disease. This dec ision w as base d on our previous observations 26,27 that patie nts w ith score 10 had no clinical symptoms of ac tive dise ase such as photose nsitivity, fever, polyarthritis, se rositis, an elevate d erythroc yte sedimentation rate or a high antinuclear antibody level (ANA). A similar distinc tion betw e en active and inactive dise ase w as also performed by othe r authors. 2 8 Our group of p atie nts included 9 patients w ith inactive and 23 patients w ith ac tive disease. The clinic al and laboratory fe atures of SLE patients are pre se nted in Table 1. The c ontrol group of 16 he althy voluntee rs w as also studied. The y w ere 12 w ome n and 4 me n, aged from 37 to 56 ye ars (mean 48 ye ars). Each underw ent a through physic al e valuation by one of the authors (ER).

Laboratory tests
On the day of blood sampling for T c ells immunophenotyping, the follow ing laboratory parame te rs w ere analysed: c omp le te blood ce ll count (CBC), erythrocyte sedimentation rate , blood urea nitrogen and cre atinine levels, fibrinogen le ve l, partial thromboplastin time (PTT), live r func tion tests (GOT, GPT, bilirubin), immunoglobulins (IgG, IgA, IgM) and comple ment (C 3 , C 4 ), urine and cre atinine le ve ls, and anti-DNA antibodies. Chest X-rays and ECG w e re also e valuate d.

Immunophenotype analysis
Venous blood samples w ere c ollec te d at the time of clinical asse ssment into p yrogen-fre e tubes, containing anticoagulant (EDTA at a final c oncentration of 25 mM). Ge ne ral lymphoc yte immunophenotyping and TCR dive rsity analysis w ere performe d by standard tw o-colour immunofluore sce nce measureme nt. The de tails of the p roc edure are described e lsew here. 2 9 Briefly a combination of phyc oe rythrin (PE)conjugate d and fluoresc ein isothiocyanate (FITC)conjugate d monoclonal (MoAbs) w as use d. In polystyrene tube s 100 m l of w hole blood w ere dire ctly staine d w ith 10 m l app ropriate MoAbs in the dark at room te mperature. IgG1 isotype c ontrol antibody conjugate s w e re include d in order to e stablish the background f luoresc enc e. Afte r incubation for 30 min, the sample s w e re placed to Q-prep (Coulter) for lysis of erythrocytes and fix ation of nuclear c ells. At le ast 10,000 ce lls w ere then analyse d on a Coulte r Epics-XL flow c ytomete r (Coulte r, Hiale ah, FL, USA). Gate Check w as used to gate lymphoc yte population defined by FS/SS and anti CD14 and CD45RO MoAbs. Analysis w as pe rforme d using XLv2 softw are.

Monoclonal antibodies
The dire ct staining of cells by monoclonal antibodies (MoAbs) w as performed. We w ere able to use only comme rcially prepared MoAbs c onjugated w ith proper fluorochrome-PE-conjugate d UCHT1 (CD3 + , pan T), IMMU510 (all g d T c ells), FITC-conjugate d 13B8.2 (CD4 + T c ells), B9.11 (CD8 + T c ells), 3G8 (CD16 + , mainly NK and at some perc entage g d T ce lls), IMMU510 (all g d T ce lls), IMMU389 (Vd 2, spe cific domain of d chain TCR), P11.5B (Vd 3, spe cific domain of d chain TCR), IMMU360 (Vg 9, spe cific domain of g chain TCR), w ere all supplied by Immunote ch (A Coulter Company, USA). The anti-Vd 1 (Immunote ch) w as available as the plain proteins and w as not applied for staining in w hole blood sample s together w ith other MoAbs. It w as used for indire ct phenotyping of is olated lymphoc ytes. In these ex perime nts the goat/anti-mouse IgG (FITC and PE conjugated F(2ab9 ) 2 fragme nts) w as used for counte rstaining.

Statistical analysis
The analysis of the results indicate d that normal distribution w as obse rved for almost all studie d variables. We prese nte d our c alculations as a mean ± SD. The universal Mann-Whitne y U te st w as use d for dete rmination of diffe renc es in quantity of studie d ce ll populations. For p henotyp ic feature corre lation betw e en studied T and NK c ell populations, me asure d simultaneously in the same patient, w e used the Spe arman rank te st. Statistic ally characteristic changes w ere considered at P< 0.05.

Results
The g d TCR ex pre ssion on pe ripheral blood T lymphoc ytes from 32 SLE patients and 16 healthy donors has be en de te rmined using standard doublecolour immunofluorescence me asurement. Their clin- ical data are summarise d in Table 1. De tailed characte ristics of peripheral blood parame te rs, both for SLE patients and for the healthy donors, are show n in Table 2. The absolute numbe rs of WBC, pan T CD3 + , CD4 + , CD8 + ce lls and NK ce lls w e re significantly low er in SLE patients than in the control group. In contrast, the absolute numbe rs of cytotox ic /suppressor (CD8 + CD3 + ) and pan B (CD19 + ) c ells in both groups w ere similar.
The me an absolute value s and pe rc entages of g d T ce lls and their subpopulations in periphe ral blood of SLE patie nts and in healthy individuals are prese nte d in Table 3. The mean absolute number of p an g d T ce lls w as low er in SLE patients (40 ± 30 /m l) than in controls (67 ± 39 /m l) (p= 0.006 ). How ever, the perc entage of g d T c ells of p an T ce lls w as similar in both groups (7.1% ± 6.5% and 6.3% ± 3.9%, respe ctive ly ) (P=0.7 ). A comparable dec rease of c irc ulating Vd 2 and Vd 9 subtypes of pan g d T c ells in SLE patie nts (20 ± 10 c ells /m l and 20 ± 20 c ells /m l, respec tively ) in relation to normal controls (50 ± 41 c ells /m l and 49 ± 40 cells /m l, re spe ctively ) (P= 0.003 and 0.005 ) w as observe d. On the othe r hand, the absolute numbe r of Vd 3 + subtype s of g d ce lls w as signific antly higher in peripheral blood of SLE patients (20 ± 10 ce lls /m l) than in c ontrol donors (2 ± 21 c ells /m l) (P= 0.001 ). A similar diffe renc e w as note d in the perce ntage of this subpopulation in both groups (4.2% ± 5.7% vs 2% ± 0.1%, respec tively; P= 0.0003 ). Although the above calculated value s for Vd 3 + ce ll subtype s w e re diffe re nt for particular SLE patients (se e range s and SD in Table  3 ), w e found in patients blood spe cime ns a distinct positive c orre lation be tw een periphe ral blood absolute numbers of g d T cells and the number of Vd 3 + ce lls (R= 0.85, P= 0.00001) (Fig. 1 ). This ve ry high correlation c oe ffic ie nt for Vd 3 + lymphoc ytes provide d additional strong evidence, that this g d T subtype could play some role in SLE activity.
It should be noticed that g d T cell levels practic ally did not fluctuate if the analytical tests for partic ular patients w ere repe ated tw o or three times. How ever, the follow up of g d T lymphocytes changes during disease deve lopme nt w ere not done .
Data pre sented in Table 3 show e d that thre equarte rs of pan g d T ce lls in the blood of healthy individuals ex pre sse d Vd 2 and Vg 9 TCR chains. Although the above re sult w as obtained in se parate staining measure ments, w e c an univocally conclude that the subfrac tion Vd 2 /Vg 9 is the most frequent in normal blood, as state d in se ve ral w orks. 7 In SLE patients the proportion of Vd 2 and Vd 3 changed, but the usage of Vg 9 in TCR struc ture is still high. In both materials the p ercentage of the Vd 1 subpopulation w as low. As w e mentioned in Patients and Methods, the quantitie s of Vd 1 + w e re e valuated by the dire ct staining method. It is know n that there are disc repancie s of relative ce ll pe rc entages dete rmine d by dire ct and indirect stainin g proce dures. We observe d 2 ce lls /m l (0.4%) in SLE patients and 11 ce lls /m l (1.1%) in healthy donors w ith Vd 1 + phenotype . The se re sults confirmed data w hich could be c alc ulated from   The Vd 1 + fraction, w hich usually inc re ases as the response against tumor antigens, 10 ,13 did not play the important role in our mode l.
No differences in g d T cells proportion as w e ll as its Vd 2, Vd 3 and Vg 9 subpopulations w e re obse rved in patients w ith active and inactive disease (Table 4 ). How e ver, the absolute numbers of g d T ce lls and all three analysed subtypes of these ce lls w ere 2 times higher in patients treated w ith immunosuppre ssive drugs in comparison w ith untreated SLE patients. But the y w e re also not signific antly diffe re nt (Table 4 ).
Furthermore the absolute number g d T cells in SLE spe cime ns c orre late d positive ly w ith the total number of T CD3 + cells (R= 0.44; P= 0.02 ), but not w ith the number of T CD4 + ce lls (R= 0.31; P> 0.05 ) or NK c ells (R= 0.037; P> 0.05 ) (Fig. 2 ). It should be mentione d that double-c olour c ytometric tests of pan g d T cells did not allow us to de te c t CD8and CD4positive g d T lymphoc ytes. This obse rvation is considered in detail in the Discussion section.

Discussion
According to our data the absolute numbers of c ells ex pre sse d g d TCR in most SLE blood spec imens w e re significantly low er (~40%; P< 0.006 ) than in the control group. How eve r, since the le vel of total T ce ll fraction (measured as CD3 + c ells ) w as also de crease d in the c ase of SLE, the mean value s of the pe rcentage g d T ce lls of pan T lymphoc ytes w e re almost the same in both analysed populations (7.1% vs. 6.3%, respe ctively ). Similar obse rvations w e re made during the comparison of Vd 2 + and Vg 9 + subtypes of pan g d T ce lls.
The c omple te ly opp osite re sult w as obtaine d in the case Vd 3 + T ce lls. The very low amount of them w as identified in periphe ral blood of control donors (le ss 0.5% of pan T lymphocyte s ), but in spec imens from SLE patie nts their increase w as very signific ant and conc erning both the absolute number of c ells and the relative pe rc entage value (20 vs. 2 ce lls, P< 0.001 and 4.2% vs. 0.2%, P< 0.0003, in SLE and healthy donors, respective ly ). The diversity of the Vg region in TCR formation during dise ase de velopment se ems to be less important. The se parate proble m is the ex pression of c o-re pressors (CD4 and CD8 ) on studie d c ells. The y are very important for normal a b T ce ll function. We suspe cte d that our observations in the case of CD4 antigen (lack of ex pre ssion ) w ere correc t, but w e did not ex clude that CD8-negative staining w as false , c ause d, in part, by instrume nt limitations. We suspec t that p an g d T cells in c ontrol blood spe cime ns w e re CD8-negative; how e ve r, in peripheral blood of SLE patients at least some of pan g d T c ells w e akly ex pressed CD8 antigen, as an activation e ffec t of c ells involve d in autoimmune response (data not show n ).
Our observation conce rning Vd 3 + ce lls' ex pansion in SLE p atie nts is the first de monstration that this subse t of g d T lymphocyte s see ms to be involved in SLE pathogene sis. Pre viously p ublishe d data indicate d the incre ase of pan g d T cells in peripheral blood from patients w ith this disease. 3 0,31 Othe r w orks characte rised ex p anded subtypes as the polyclonal Vd 1 + and Vd 2 + or Vg 2 /Vd 2 + oligoclonal subsets . 23 The se discre pancie s probably arose as the re sult of diffe re nt features of analysed dise ase in partic ular groups of patients, and indicate d that changes of se ve ral parame te rs are involved in the dise ase de ve lopment How ever, our results are to some ex te nt unexpec te d, because most of the studies undertake n so far indicate that total le vels of pan g d T c ells (calc ulate d both as the ce ll number /m l and as the relative perc entage of pan T CD3+ lymphoc yte frac tion ) w ere significantly higher in p eriphe ral blood of patie nts w ith autoimmune diseases than in healthy donors. Janadi e t a l. 32 observed the ex p ansion CD4 + CD29 + g d T ce lls in the peripheral blood and synovial fluid of patients w ith rheumatoid arthritis (RA). The ele vation of pan g d T c ells in the same mate rials of RA patients w as stated by Ke ystone e t a l. 3 It w as also demon- Table 4. The frequency of g d T cell subtypes in peripheral blood of patients with SLE dependent on disease activity and immunosuppressive treatment (mean of cells number/m l ± SD and range in parentheses) strate d in many inve stigations that the involve ment of g d T lymphocytes in the pathoge nesis of autoimmune diseases w as most like ly, be cause the ir ex pansion w as alw ays signific ant. It has be en discusse d in the studies conc erning coeliac disease , 34 multiple scle rosis, 17 ,3 5 autoimmune thyroid diseases, 36 autoimmune liver disease 3 7 and syste mic scle rosis . 20 How ever, in the case s of the above-spec ifie d disease s, the ac cumulation of g d T c ells oc curred pre dominantly in pathologic ally changed tissues; their increase s in peripheral blood of the same patients w ere le ss e vident or e ven unnotice able . We fe el oblige d to say that our study of peripheral blood ce lls w as dic tated by the pragmatic goal to find a simple, highly standardised te st for some aspec ts of SLE diagnosis . Of course, w e knew that the ce ll representation in blood w as only the approx imate image of the ac tual situation in affe cte d tissue s, but the rec ip roc al circulation of T lymp hocyte s in an organism could allow the desc ription of re al changes in g d T ce lls, as it is observe d in tumor infiltratin g lymphoc ytes and peripheral circulation ce ll populations. 1 0 In the c ase of our SLE patients investigation, the decre ase of pan g d T ce lls numbe r (and similar e ve nts for Vd 2 + and Vd 9 + subtype s) partly resulte d from gene ral pan T lymphopenia, w hich w as more inte nsive than in other SLE studie s. Anothe r re ason for that phenomenon could be connec te d w ith the capacity of g d T lymphocytes to very strong infiltration and pathological damage of targe t tissues, as skin and kidne ys. Such ac cumulation, partic ularly Vg 2 /Vd 2 TCR ex pressed ce lls, w as obse rved in disease-damaged skin of patients w ith chronic cutane ous lupus e rythematosus 2 3 and SLE. 2 2 Probably in our case the Vd 3 + subfraction re sponde d to autologous immune antige ns. The pe rsistent tre atment of the most patients w ith glucoc ortic oids could be also responsible for the low er numbers of g d T ce lls in peripheral blood from SLE individuals than in spe cime ns from control donors. Se ve ral published data demonstrate d that long-c ontinue d immunosuppressive therapy de te rmined the disappearance of ex panded g d T c ell subset, in both targe t tissue s, and the peripheral blood of patients w ith polymyositis or othe r autoimmune dise ases. 3 0,3 8 Spinozzi e t a l. 3 1 demonstrate d data obtained from in v itro ex pe rime nts that all T lymphocytes bearing the g d TCR (isolate d both from SLE patie nts and he althy individuals) w ere susce ptible to dex ame thasone , and ste roid-induce d apoptosis w as basic mechanism responsible for ce lls death. The y also de monstrate d that 6 month glucocortic oids treatme nt normalise d the inc re ased SLE g d T c ell subfraction in blood, simultane ously w ith clinical remission of the disease symptoms.
In our studies the analyses of immunosuppre ssive tre atment influenc e on g d T c ell le ve l w ere not such univoc al. In all ce ll subtypes their amounts w ere tw o times highe r in the treate d p atie nts than in untre ated ones (but alw ays low er than in spe cime ns of control donors). It could be considered as a re construc tion proc ess of the initial g d T cells status, typical for normal blood. A relative ly low number of studied c ases can be re sponsible for statistical doubts. The re ason w hy it w as also observe d for Vd 3 + T c ells, w hic h see m to compose the unique subtype in healthy individuals' blood, is unclear. A simple look at the data pre sented in Table 4 may lead to the conclusion that the frequenc ie s of total g d T c ells and the ir partic ular subtype s in SLE patients w e re not generally dete rmine d by dise ase activity, e valuated acc ording to the method of Liang e t a l. 2 5 But statistically important, obje ctive data could probably be obtaine d during the analysis of affe cte d tissue s. We regis te re d the ce ll distribution in peripheral blood, w hich probably re flected only some functional te ndenc ie s occ urring in tissue s. The corre lation be tw ee n SLE activity and conce ntrations of some cytokines and their soluble rece ptors w as detected in our previous investigations. 26 ,27 Data from Table 2 and Figure 2 indic ate, that although absolute amounts of pan T CD3 + ce lls, T he lper CD4 + c ells and NK ce lls w e re significantly reduced in SLE patie nts, variations of T CD4 + and NK ce lls w e re not correlated w ith change s of pan g d T ce lls. In such case the influenc e (if any) of compare d ce ll populations on SLE de velop ment is indepe nde nt.
At pre sent the re is little know n about the function of g d T cells in SLE pathogene sis. The re ce nt ex pe riments confirm that the y really play a significant role in autoimmunity re gulation in v ivo , but some authors published c ontrary data, suggesting g d T lymphoc ytes' re sponsibility for dow n-regulation of autoimmune dise ases. 5 , 3 6 The w ork of Pe ng e t a l. 3 9 w as the first demonstration in vivo that mice w ith g d T c ells de ficie ncy (TCRd -/-MRL/lpr) developed a significantly more severe lupus-like disease and their mortality w as tw ice that in normal MRL/lpr mic e. How e ver, g d T lymphoc ytes c ause d also dow nregulation of the a b T c ell response to infec tion and thus the y c ould intensify the autoaggressive te stis inflammation e voked by Lis te ria infe c tion. 40 Additionally it w as show n that g d T ce lls promote d the B ce ll me diate d autoimmunity. 4 1 All the above obse rvations may indic ate that different subtypes (or distinct clones) of g d T lymphocytes realise se parate, some times e ven opposite functions in induction of autoreactive immune re sponse s, for ex ample in SLE. It should be note d that their function is re alise d ve ry e arly in ontoge ny, and further disturbance s of se ve ral me tabolic pathw ays can change the final effec ts. We belie ve that the indicated obse rvations do not reflec t epiphe nomenons, but univoc al evide nce of it has to be verifie d by additional ex pe riments. We have initiate d our future studies base d on more c ase s, include d cytome tric analysis w ith c ells isolate d from affec te d tissues and proper c ontrols. We have also ex tended our studies for molec ular te sts (RT-PCR de te c tion of g d TCR discre t, minor subpopulations, charac te risation of their ge nomic structure and dete rmination of clonality of ex pande d g d T c ells subtype s). Our study indicated that Vd 3 + subtype of g d T ce lls se ems to be involved in SLE pathoge nesis ; how e ve r, w e ac ce pt the view that the autoimmunity does not de ve lop from single abnormality, but rathe r from a numbe r of different events.