Ann Hematol (2003) 82:496–499DOI 10.1007/s00277-003-0645-x

O R I G I N A L A R T I C L E

S. Dubey · S. Nityanand

Involvement of Fas and TNF pathways in the inductionof apoptosis of T cells by antithymocyte globulin

Received: 24 January 2003 / Accepted: 12 March 2003 / Published online: 29 May 2003� Springer-Verlag 2003

Abstract Antithymocyte globulin (ATG) is the treatmentof choice for those aplastic anemia patients who are notsuitable for bone marrow transplantation (BMT). ATG isalso used for the treatment of rejections in organtransplantation and as a conditioning regimen in BMT.Despite the proven efficacy of ATG in these areas, itsmechanism of action is not known. Profound T-celllymphopenia observed in vivo with ATG treatment issupposed to contribute to its therapeutic effect. We havepreviously shown that apoptosis is one of the mechanismsresponsible for ATG-induced lymphopenia. Our nextobjective was to investigate the effect of ATG onmodulation of Fas and TNF pathways, the two mainpathways of T-cell apoptosis. Maximum surface expres-sion of Fas on T cells was observed after 24 h at an ATGdose of 100 mg/ml; at this dose 88% of cells expressed Fasas compared to 26% of untreated cells. Surface expressionof FasL was found to peak after 24 h at an ATG dose of1000 mg/ml when 34% of cells were positive for FasL ascompared to 1.5% of untreated T cells. Tumor necrosisfactor (TNF)-a production was found to be maximumafter 6 h at 1000 mg/ml dose (20%) as measured byintracellular cytokine staining of T cells. TNF-a produc-tion was also measured by enzyme-linked immunosorbentassay (ELISA) in the supernatant of lymphocytes treatedwith ATG for 6 h. A dose-dependent increase in TNF-aproduction was found in these supernatants with a plateaubeing achieved at an ATG dose of 1000 �g/ml. Weconclude that ATG-induced apoptosis in T cells involvesboth Fas and TNF pathways and TNF-a is produced muchearlier than Fas and FasL expression.

Keywords Antithymocyte globulin (ATG) · Apoptosis ·Fas · FasL · TNF-a

Introduction

Antithymocyte globulin (ATG), a polyclonal mixture ofantibodies against lymphocyte cell surface antigens, iswidely used for the treatment of aplastic anemia (AA) [9]and as an immunosuppressant in organ transplantation [4]and bone marrow transplantation [1]. In spite of being aneffective immunosuppressant, the precise mechanism ofaction of ATG still remains unknown. Profound T-celllymphopenia is observed after ATG administration,indicating that ATG executes its immunomodulatoryeffect by elimination of T cells. The mechanism oflymphopenia has been attributed to several mechanismssuch as complement-dependent cytolysis, antibody-de-pendent cellular cytotoxicity as well as opsonization andsubsequent phagocytosis by macrophages. However, thereis no evidence for any of these mechanisms.

We have previously demonstrated that induction ofapoptosis of peripheral blood mononuclear cells may bean important mechanism for ATG-induced lymphopenia.ATG induces an apoptosis of peripheral blood mononu-clear cells (PBMCs) in vivo in AA patients undergoingtherapy as well as an in vitro apoptosis in normal PBMCs[3]. In vitro induction of apoptosis of PBMCs was dosedependent, a plateau being achieved at an ATG dose of1000 �g/ml.

The two main pathways involved in T-lymphocyteapoptosis are the Fas and the tumor necrosis factor (TNF)pathways. A study of the pathways involved in inductionof T-cell apoptosis by ATG is important for understand-ing the mechanism of ATG-induced T-cell lymphopenia.We therefore undertook this study to investigate whetherATG can modulate these two pathways and induce T-cellapoptosis.

S. Dubey · S. Nityanand ())Department of Immunology,Sanjay Gandhi Post Graduate Instituteof Medical Sciences (SGPGIMS),Raebareli Road, 226014 Lucknow, Indiae-mail: soniya@sgpgi.ac.inTel.: +91-522-2668700/268800/2668900Fax: +91-522-2668078/668017

Materials and methods

Culture of PBMCs from normal donors with different doses of ATG

PBMCs were isolated from heparinized blood of five normaldonors using the standard method. The PBMCs were cultured inRPMI-1640 (Gibco-BRL, Grand Island, N.Y., USA) supplementedwith 10% fetal calf serum (FCS), 2 mmol/l L-glutamine andantibiotics (penicillin 100 U/ml, streptomycin100 �g/ml) atconcentrations of 1�106/ml in 24-well culture plates (Costar,Cambridge, Mass., USA). Cultures were maintained in a humidatmosphere at 37�C containing 5% CO2 for the indicated time. Thecells were incubated with ATG (Lymphoglobuline, Pasteur Mer-rieux, Lyon, France) at different doses (10 �g/ml, 100 �g/ml,500 �g/ml, 1000 �g/ml) for 24 h for Fas and FasL expression and6 h for TNF-a production.

Fas and FasL expression

Surface staining for Fas and FasL expression was studied by dualcolor flow cytometry using the following panel of fluorochromeconjugated antibodies Fas (CD95)-FITC/CD3-PE and FasL(CD95L)-PE/CD3-FITC (Becton Dickinson, Mount View, Calif.,USA). Briefly, 1�106 cells were taken and washed once withphosphate-buffered solution (PBS). The cell pellet was stained with5 �l of anti-CD95-FITC for 30 min in the dark followed by 5 �l ofanti-CD3-PE antibody. The cell pellet was then washed once inPBS and resuspended in 500 �l of PBS for acquisition and analysis.Similarly, the staining for FasL was performed using anti-CD95L-PE and CD3-FITC antibodies. Cells were immediately analyzed ona flow cytometer (FACSCalibur, Becton Dickinson, FranklinLakes, N.J., USA) equipped with 488-nm argon laser as lightsource. For analysis first a dot plot was drawn between FSC andSSC parameters. The lymphocytes were gated on this plot and10,000 events in this gate were acquired for all the tubes. Quadrantswere set according to isotype control, and the total percentage of Tcells expressing Fas/FasL was calculated.

TNF-a production

TNF-a production was studied at single cell level by intracellularstaining by flow cytometry according to the method of Jung et al.[6]. Briefly, 1�106 cells/ml of cRPMI were incubated with either50 ng/ml phorbol myristyl acetate (PMA, Sigma Chemical, St.Louis, Mo., USA) and 1000 ng/ml ionomycin (Sigma Chemical, St.Louis, Mo., USA) as a positive control or with different doses ofATG for 6 h in each well of 24-well culture plates. Brefeldin A(10 �g/ml, Sigma Chemical, St. Louis, Mo., USA) was added as aprotein transport inhibitor in all the wells. After 6 h the cells werefixed in 1% paraformaldehyde and permeabilized using 0.1%saponin in PBS. The cell pellet was then stained with anti-TNF-a

FITC antibody for 45 min in the dark. The cells were washed oncein PBS and stained with anti-CD3-PE antibody for 30 min in thedark. The cell pellet was then resuspended in 500 �l PBS foracquisition and analysis.

TNF-a ELISA

A total of 1�106 PBMCs/ml cRPMI were incubated with ATG atdifferent doses or PMA (25 ng/ml) for 6 h. Supernatant washarvested from these cells and stored at �80�C for ELISA. TheTNF-a levels were estimated using the kit (R&D Systems,Minneapolis, Minn., USA). The kit protocol was followed. Threeseparate experiments were carried out from five donors for all theassays and the data are presented as mean€SD.

Results

Upregulation of Fas and FasL expressionon T cells treated with ATG

T cells exhibited an upregulation of Fas and FasL ontreatment with ATG. The time and dose kinetics studyshowed that expression of Fas was maximum when Tcells were incubated with ATG at a dose of 100 mg/ml for24 h. At this dose 88€6.2% T cells were positive for Fasexpression as compared to 26€5.3% of untreated cells.Increasing the ATG dose (500 mg/ml) did not result in anincrease in Fas (72%) expression. Representative flowcytometric dot plots of Fas expression on untreatedPBMCs and PBMCs treated with ATG 100 mg/ml areshown in Fig. 1a and b, respectively. Expression of FasLwas found to peak after 24 h at an ATG dose of 1000 mg/ml when 34€6.3% T cells were positive for FasLexpression as compared to1.5€1.1% in untreated T cells.FasL expression on T cells was 9€3% at 10 mg/ml and29€1.1% at 500 mg/ml ATG dose. Representative flowcytometric dot plots of FasL expression on untreatedPBMCs and on PBMCs treated with ATG 1000 mg/ml areshown in Fig. 2a and b, respectively.

Fig. 1. a, b Representative flowcytometric dot plots of Fas(CD95) expression on untreatedT cells (a) and lymphocytestreated with ATG dose of100 �g/ml (b). Values in theupper right quadrant representpercentage of T cell expressingCD95

497

Production of TNF-a by T cells treated with ATG

PBMCs treated with ATG for 6 h produced TNF-a, whichincreased in a dose-dependent manner. Maximum TNF-awas found in the supernatant of PBMCs treated with anATG dose of 1000 mg/ml (1510 pg/ml) as compared to316 pg/ml in supernatant of untreated PBMCs. A plateauwas achieved at an ATG dose of 1000 mg/ml and furtherincrease in the dose did not augment TNF production

(Fig. 3). Intracellular TNF-a was also found to attain apeak at an ATG dose of 1000 mg/ml when 20% of T cellsproduced TNF-a as compared to 0% in untreated PBMCs.At an ATG dose of 100 mg/ml and 500 mg/ml, 5% and15% of T cells were positive for TNF-a production,respectively. Representative flow cytometric dot plots ofintracellular TNF-a production by untreated T cells andcells treated with ATG 1000 mg/ml are shown in Fig. 4aand b, respectively.

Discussion

We have previously shown that ATG induces apoptosis oflymphocytes [3]. In this study we have demonstrated thatATG incubation leads to production of TNF-a andupregulation of Fas/FasL on T cells, which may be thepathways for T-cell apoptosis. Apoptosis can be triggeredby interaction of cell surface death receptors such as Faswith its specific ligand FasL, which stimulates anintracellular cascade that leads to induction of apoptosisin a wide variety of target cells [12]. TNF-a can also actas mediator of apoptosis in mature T lymphocytes [15].

We found a constitutive expression of Fas (26%) onuntreated T cells of normal donors. ATG incubation led toan upregulation of Fas expression after 24 h in a dose-dependent manner. It has been shown that Fas isconstitutively expressed on different cell types of hema-topoietic and nonhematopoietic tissues and is upregulatedafter activation [10]. Fas is a proapoptotic molecule, but

Fig. 2. a, b Representative flowcytometric dot plots of FasL(CD95L) expression on un-treated T cells (a) and lympho-cytes treated with ATG dose of1000 �g/ml (b). Values in theupper right quadrant representpercentage of T cell expressingCD95L

Fig. 3 Graph showing dose-dependent increase in secretion ofTNF-a (pg/ml) from the supernatants of PBMCs treated with ATGat different doses for 6 h

Fig. 4. a, b Flow cytometricdot plots of intracellular TNF-aproduction from untreated Tcells (a) and treated with ATGdose of 1000 mg/ml (b). Valuesin the upper right quadrantrepresent percentage of T cellspositive for TNF-a

498

the expression of Fas alone does not initiate an apoptoticcascade. Therefore, the expression of FasL on T cellsafter ATG incubation was also investigated. Unlike Fas,surface expression of FasL is more tightly controlledduring apoptosis and is more restricted and often requirescellular activation [14]. In accordance with these studies,we also found a very low level of expression of FasL onuntreated T cells (1.5%) as compared to cells which hadbeen treated with ATG 1000 mg/ml (34%) for 24 h.

A previous study by Genestier et al. has shown thatrabbit ATG at doses of 10 mg/ml and 100 mg/ml inducedFas and FasL mRNA and surface expression in nonacti-vated and preactivated peripheral blood lymphocytes [5]and found maximum expression 24 h after incubation.Our results also show similar expression of thesemolecules after 24 h, but a maximum at higher doses ofATG (1000 ug/ml) which was not studied so far.Genestier et al. also did not specify the percentage orthe cell type (T or B cells) which express Fas and FasLafter ATG incubation. We have shown by dual colorstaining the high level of expression of these moleculeson T cells after incubation with ATG.

In a previous study by Piaggio et al., a dose-dependentrelease of TNF-a was observed in supernatant of bonemarrow T cells treated with ATG for 72 h [13]. Theseauthors did not study TNF-a at earlier time points.Furthermore, there are no reports on intracellular produc-tion of TNF-a by T cells after incubation with ATG. Wehave shown TNF-a production from T cells at single celllevel as early as 6 h after ATG incubation. TNF-aproduction was observed on T cells much earlier than Fas/FasL expression. It is well established that production ofTNF-a leads to upregulation of Fas and FasL expression[8].

Whether both the pathways of apoptosis are operativeor TNF-a assists in induction of Fas/FasL which thenleads to apoptosis of T cells cannot be commented uponfrom the data in this study but needs further investigation.Maximum upregulation of FasL and maximum TNF-aproduction was observed at an ATG dose of 500–1000 mg/ml. This is significant because this is the average serumlevel reached during horse ATG treatment in AA [2]. Inour previous study we also have shown that in vitroinduction of apoptosis in PBMCs treated with ATGreached a plateau at a dose of 1000 mg/ml [3].

Crosslinking of lymphocyte cell surface antigens bycognate antibodies has been shown to induce apoptosis inT lymphocytes [7, 11]. Since ATG is a mixture ofantibodies against lymphocyte cell surface antigens, itmay be possible that persistent activation of T cells bythese antibodies may trigger an extrinsic signal ofapoptosis. In conclusion, this study suggests that ATG-induced apoptosis of T cells involves both Fas and TNFpathways.

Acknowledgements This work was supported by an intramuralgrant from SGPGIMS, Lucknow, India.

References

1. Bacigalupo A, Lamparelli T, Bruzzi P, Guidi S, AlessandrinoPE, Diartolomeo P, Oneto R, Bruno B, Barbanti M, Sacchi N,Van Lint MT, Bosi A (2001) Antithymocyte globulin for graft-versus-host disease prophylaxis in transplants from unrelateddonors: 2 randomized studies from Gruppo Italiano TrapiantiMidollo Osseo (GITMO). Blood 98:2942–2947

2. Bonnefoy B, Revillard JP (1996) Mechanism of immunosup-pression by antithymocyte globulins and OKT-3. J Heart LungTransplant 15:435–442

3. Dubey S, Srivastava A, Nityanand S (2002) Induction ofapoptosis of peripheral blood mononuclear cells by antithymo-cyte globulin (ATG) in aplastic anemia: an in vivo and in vitrostudy. Ann Hematol 81:249–253

4. Gaber OA, First M, Tesi RJ (1998) Results of double blindedrandomized, multicentre, phase III clinical trial of Thymoglob-uline versus Atgam in the treatment of acute graft rejectionepisodes after renal transplantation. Transplantation 66:29

5. Genestier L, Fournel S, Flacher M, Assossou O, Revillard JP,Bonnefoy-Berard N (1998) Induction of Fas (Apo-1, CD95)mediated apoptosis of activated lymphocytes by polyclonalantithymocyte globulins Blood 91:2360–2368

6. Jung T, Schauer U, Heusser C, Neumann C, Rieger C (1993)Detection of intracellular cytokines by flow cytometry. JImmunol Methods 159:197–207

7. Klaus SJ, Sidorento SP, Clark EA (1996) CD45 ligationinduces programmed cell death in B and T lymphocytes. JImmunol 156:2743–2746

8. Maciejewski J, Selleri C, Anderson S, Young NS (1995) Fasantigen expression on CD34+ human marrow cells is inducedby interferon-g and tumour necrosis factor-a and potentiateshematopoietic suppression in vitro. Blood 85:3183–3190

9. Marsh J, Schrezenmeir H, Marin P (1999) Prospectiverandomized multicenter study comparing cyclosporin aloneversus the combination of cyclosporin and antithymocyteglobulin for treatment of patients with nonsevere aplasticanemia: a report from the European Blood and MarrowTransplantation (EBMT) severe aplastic anemia working party.Blood 93:2191–2195

10. Miyawaki T, Uehara T, Nibu R, Tsuji T, Yachie A, Yonehara S,Taniguchi N (1992) Differential expression of apoptosis-relatedFas antigen on lymphocyte subpopulations in human peripheralblood. J Immunol 149:3753–3758

11. Mollereau B, Deckert M, Deas O, Rieux-Laucat F, Hersh F,Bernard A, Fisher A, Lynch DH, Charpentier B, Le Deist F,Senick A (1996) CD2 induced apoptosis in activated humanperipheral T cells. J Immunol 156:3184–3188

12. Nagata S, Golstein P (1995) The Fas death factor. Science267:1449–1456

13. Piaggio G, Podesta M, Pitto A, Pittaluga GB, Isaza A,Benvenuto F, Bruno B, Bacigalupo A (1998) ComparableTNF-a, IFN-g and GM-CSF production by purified normalmarrow CD3 cells in response to horse anti-lymphocyte andrabbit anti thymocyte globulin. Eur J Hematol 60:240–244

14. Suda T, Okazaki T, Naito Y, Yokota T, Arai N, Ozaki S, NakaoK, Nagata S (1995) Expression of the Fas ligand in cells of Tcell lineage J Immunol 154:3806–3813

15. Zheng L, Fisher G, Miller RE, Peschon J, Lynch DH, LenardoMJ (1995) Induction of apoptosis in mature T cells by tumournecrosis factor Nature 377:348–351

499