Kidney International, Vol. 61 (2002), pp. 1303–1311

CELL BIOLOGY – IMMUNOLOGY – PATHOLOGY

Transfection of tubule cells with Fas ligand causesleukocyte apoptosis

YIPING WANG, SHOUNAN YI, YUET-CHING TAY, XIMIN FENG, YANG WANG, LUKAS KAIRAITIS,and DAVID C.H. HARRIS

Department of Renal Medicine, The University of Sydney at Westmead Hospital, Sydney, New South Wales, Australia

Transfection of tubule cells with Fas ligand causes leukocyte Interstitial inflammation and fibrosis are prominentapoptosis. features of all types of chronic renal disease [1]. Although

Background. Since the Fas/Fas Ligand (FasL) interaction is remarkable progress has been made in the past ten yearsrecognized as a major pathway of apoptosis in immune cells,in understanding the pathogenesis of interstitial injury,we hypothesized that selective expression of FasL by tubular

cells (TC) may promote the resolution of interstitial inflamma- therapeutic strategies remain ineffective [2]. Recently,tion by inducing apoptosis of infiltrating immune cells. In this the role of individual mediators, especially cytokines,study, the effect of FasL transfection of rat TC on apoptosis has been explored extensively [3–5]. However, cytokineof leukocytes was examined.

therapy is ineffective or only partially effective in pre-Methods. Rat tubule cells (NRK52E) were transfected withventing interstitial injury, because of the complexities ofplasmids constructed using human and rat FasL (hFasL and

rFasL). The propensity of activated, transfected TC to undergo the redundant network of cytokines with overlappingapoptosis was examined. Similarly, the effects of FasL transfec- and/or opposing effects on interstitial target cells. Fortion on apoptosis of Jurkat cells and activated leukocytes were example, whereas interleukin-4 (IL-4) and transformingassessed directly following co-culture for 12 hours and in a cell

growth factor-� (TGF-�) inhibit the functions of macro-insert system intended to assess the effects of soluble FasL.phages and thereby limit interstitial inflammation, theyFas and FasL expression was assessed by flow cytometry and

apoptosis was examined using Annexin V staining and the also promote the proliferation of fibroblasts and thusTUNEL method. aggravate interstitial fibrosis [6–8].

Results. Expression of FasL in TC was increased after FasLFas is a receptor for initiating apoptosis and Fas ligandtransfection. Transfected TC showed no detectable increase in

(FasL) a mediator for the delivery of death signals. FasLapoptosis following lipopolysaccharide (LPS) or tumor necro-sis factor-� (TNF-�) activation. Jurkat cell apoptosis was in- induces apoptosis by binding to its receptor Fas. Thiscreased ninefold and eightfold after co-culture with TC trans- process has been recognized as the major pathway offected with hFasL and rFasL, respectively (67.0 � 12.1% and cell apoptosis for homeostasis of T cell numbers and60.1 � 8.8% vs. 6.7 � 1.8% with un-transfected TC, P �

maintenance of immunological privilege [9]. The Fas/0.01). Similarly, apoptosis of activated leukocytes was increasedFasL pathway has been utilized as a potential therapeuticfourfold by co-culture (26.8 � 4.9% vs. 6.7 � 2.0% with un-

transfected TC, P � 0.01). Leukocyte apoptosis also was in- approach for cancer and autoimmune disorders [10, 11].creased in an insert culture system (18.2 � 4.4% vs. 5.8 � 2.3% Attempts to resolve inflammation by manipulation ofwith un- transfected TC, P � 0.01). No increase of TC apoptosis

Fas/FasL interactions have been made in vivo in a murinewas detected in any of the co-culture experiments.asthma model and in vitro on endothelium to inhibitConclusion. Enhanced expression of FasL by TC is capable

of inducing apoptosis of activated leukocytes, without evidence leukocyte extravasation [12, 13]. However, the potentialfor increased susceptibility to apoptosis of the transfected cells efficacy of enhancing the Fas/FasL pathway to preventthemselves. This suggests a potential role for this approach in renal interstitial inflammation and fibrosis has not beenthe limitation and resolution of renal tubulointerstitial in-

tested. Human tubule cells (TC) expressed both Fas andflammation.FasL, but they appear to be resistant to Fas-mediatedapoptosis [14]. On the other hand, it has been reported

Key words: Fas/FasL, tubule cell, renal tubulointerstitial, cells inflam- that activated leukocytes and fibroblasts are sensitive tomation, cell death, chronic renal disease, fibroblasts. Fas-induced death [15, 16], and Fas-sensitized lympho-Received for publication June 30, 2000

cytes can be killed by tubular cells via the Fas-FasLand in revised form August 4, 2001Accepted for publication November 2, 2001 system [17]. This raises the interesting possibility that

the Fas-FasL interaction might be manipulated to inhibit 2002 by the International Society of Nephrology

1303

Wang et al: FasL cause leukocyte apoptosis1304

infiltrating inflammatory cells and fibroblasts in the inter- stimulated with 400 ng/mL recombinant human solublestitium in progressive chronic renal disease. The purpose FasL and 1 �g/mL enhancer (gifts of Dr. Jurg Tschopp,of this study was to examine whether TC could be trans- University of Lausanne, Switzerland) for 12 hours, andfected with FasL, and whether TC transfected with FasL TC with 2 �g/mL camptothecin (Sigma, Sydney, Austra-could induce apoptosis of leukocytes in vitro without lia). The specificity of FasL-induced apoptosis was con-causing TC apoptosis. firmed by adding a blocking antibody (10 �g/mL) against

rat FasL (Santa Cruz Biotechnology, Inc., Santa Cruz,CA, USA) to the cells in co-culture.METHODS

Insert co-culture. To study the effect of soluble FasLCells and cell cultureon leukocytes, confluent transfected TC were detached

For primary culture, TC were isolated and cultured with trypsin treatment and washed with 10% FCS me-from normal male Wistar rats using isopycnic centrifuga- dium. A total of 106 /mL TC in 5% FCS were transferredtion. Cells were grown in Dulbecco’s modified Eagle’s to 30-mm inserts (0.4 �m pore diameter; Millipore-CM,medium (DMEM) supplemented with epidermal growth Bedford, MA, USA). These inserts allow transfectedfactor (EGF; 10 ng/mL) and insulin (5 mg/mL) in a 5% TC to be co-cultured with leukocytes without physicalCO2/95% O2 environment [18, 19]. Experiments were

contact.commenced after cells had reached confluence, which

Construction. The protein expression vectors (pTar-was usually between five to six days after the isolationget) containing the cytomegalovirus (CMV) promoterprocedure. Leukocytes were separated from the spleenwere purchased from Promega (Promega, Australia). Toand peripheral blood of normal male Wistar rats by usingconstruct the rat (pTarget-rFasL) and human FasL-Lymphoprep (Pharmacia, Uppsala, Sweden) and centri-expression plasmid (pTarget-hFasL), mononuclear cellsfuged at 400 � g for 40 minutes. Cells were harvestedwere separated from rat peripheral blood or human pe-at the interface and washed twice in phosphate-bufferedripheral blood respectively. mRNA was extracted fromsaline (PBS). Leukocytes were cultured in RPMI me-mononuclear cells after stimulation with phytohemag-dium (Biosciences, NSW, Australia) for 24 hours, andglutinin (PHA; 10 ng/mL) and ConA (2 �g/mL) for 12then activated by ConA (2 �g/mL) for 12 hours. Jurkathours. DNA containing full coding region of the rat FasLcells, a human acute T-cell leukemia cell line, were growngene was amplified from rat mononuclear cells using thein RPMI medium supplemented with 10% fetal calf se-PCR primers 5�-CAAGACTGAGAGGAGGAAAC-3�rum (FCS). NRK52E tubular epithelial cells, of rat ori-

gin, were cultured in DMEM containing 5% FCS. All and 5�-TGGCATCCATGATAAAGAA-3�. Similarly,cells were cultured and maintained at 37�C in 5% CO2/ full coding region of human FasL was amplified from95%O2. NRK52E cells were used as TC in all experi- human mononuclear cells by using the polymerase chainments; TC in primary culture also were used in the exper- reaction (PCR) primers 5�-CCTCTACAGGACTGAGiment examining the effect of FasL on TC. AAGAAG-3� and 5�-CAACATTCTCGGTGCCTGT

Co-culture. To study the effect on TC of FasL, TC in AA-3�. PCR generated a 925 bp fragment of rat FasLprimary culture and transfected TC in which 80% of and a 963 bp of fragment of human FasL. The PCRcells expressed FasL were detached and co-cultured in products were cloned into pTarget vectors, and se-a ratio of 1:1. The mixed cells were cultured in FCS-free quenced to confirm that the inserts were vectored in themedium for 24 hours and then exposed to lipopolysac- correct direction.charide (LPS; 2.5 �g/mL) and tumor necrosis factor-� Transfections. NRK52E cells were seeded at a density(TNF-�; 3000 �/mL) for 48 hours. of 3 � 105 cells/per well into six-well cell culture plates

To study the effect of FasL on Jurkat cells and leuko-containing 1 mL DMEM without antibiotics. After 24

cytes, 106 transfected TC (effector cells) were incubatedhours, cells at 60% confluency were exposed to 1.5 �gwith Jurkat cells or leukocytes (target cells), with fivefoldof plasmids (pTarget-rFasL or pTarget-hFasL) and 4 �Lserial dilutions of target cells in six-well plates at varyingof lipofectamine (Life Technologies, Australia) per welleffector:target (E:T) ratios in 2 mL of RPMI mediumin DMEM without FCS. After five hours of transfection,for 12 hours at 37�C in 5% CO2. The suspended cellsFCS was added into the medium to a final concentration(Jurkat cells or leukocytes) were collected and adherentof 5%. Transfected TC were cultured until total conflu-cells (tubule cells) were detached for assessment. Afterence, and then G418 antibiotic (400 �g/mL) added intoJurkat cells were co-cultured with TC, the cell types inthe medium. Transfected cells were selected by culturesupernatant were identified using CD45 staining (Phar-with G418 for three weeks prior to commencing theMingen, San Diego, CA, USA). Of the suspended cells,experiments. Similarly, NRK52E cells were transfected98.1 � 1.4% expressed CD45 indicating that most sus-with plasmid (pTarget) without a FasL insert for use aspended cells were Jurkat cells and only a few were TC.

As a positive control for apoptosis, Jurkat cells were control in co-culture experiments.

Wang et al: FasL cause leukocyte apoptosis 1305

Fas and Fas ligand expression the target cells were resuspended in cold PBS on ice forFACS analysis. At least 10,000 target cells from eachFlow cytometry. After culture, suspended cells weresample were analyzed by FACS. Cells cultured alonecollected, and adherent cells were detached with 2.2and without Annexin V-FITC staining were used as ammol/L ethylenediaminetetraacetic acid (EDTA), 0.2%non Annexin V-FITC-binding control. The resultingbovine serum albumin (BSA) in PBS. Cells were washedfluorescence histograms were corrected for nonspecificand resuspended in with PBS containing 0.2% BSA. Tostaining by cumulative subtraction of the no Annexinstudy Fas expression, 5 � 105 cells from the same sampleV-FITC binding control to determine the percentage ofwere incubated with 1 �g/mL anti-Fas antibody (SantaAnnexin V-FITC-binding positive cells (apoptotic cells).Cruz Biotechnology) and control rabbit IgG for 30 min-Quantitation of cytotoxicity was determined by the per-utes at 4�C, followed by incubation with 1:100 FITC-centage of apoptotic target cells scored by FACS. Eachrabbit-anti-rabbit IgG for 30 minutes at 4�C. To studytest was performed in triplicate in three separate experi-intracellular Fas ligand expression, 5 � 105 cells werements.fixed in 2% glutaraldehyde/PBS, washed in PBS, and

In situ cell death detection (TUNEL method). Cellpermeabilized in PBS containing 0.1% Triton X-100,apoptosis was also assessed using in situ cell death detec-2% BSA and 1% goat serum for 30 minutes at roomtion kit, Fluorescein (Roche, Mannheim, Germany).temperature. Thereafter, cells were incubated with 1Briefly, after washing with PBS, cells were fixed in 4%mg/mL polyclonal anti-Fas ligand antibody in PBS con-paraformaldehyde solution for one hour and permeabil-taining 0.1% Triton X-100 and 0.2% BSA for 60 minutes,ized with permeabilization solution for two minutes onfollowed by incubation with 1:100 FITC-goat anti-rabbitice. Then, cells were incubated with TUNEL reactionIgG for 60 minutes. Control samples were stained withmixture for 60 minutes at 37�C. After washing, cells werean anti-FasL antibody (Santa Cruz Biotechnology) thatanalyzed by FACS.had been previously incubated with a control Fas ligand

The assays were evaluated in a side-to-side compari-peptide or with non-immune rabbit lgG. Cells were thenson. Apoptosis as detected by Annexin V and TUNEL

analyzed on a FACS.methods was highly correlated (r 0.9891). Annexin V

Dead cells and debris were excluded from analysis by staining is more sensitive than TUNEL, with the advan-selective gating based on anterior and right angle scatter. tage of being able to detect early stage apoptosis, which isAt least 10,000 events were collected for each sample, consistent with previously published results [21]. Furtherand data were displayed on a logarithmic scale of increas- apoptosis studies were therefore performed mostly usinging green-fluorescence intensity. Fluorescence of cells the Annexin V method.was calculated using LYSIS II software (Becton Dickin-son). Mean cell fluorescence obtained with the control Statistical analysisIgG was subtracted from that obtained with the anti-Fas The results are expressed as the mean � SD, andantibody. Data are expressed as the fold-increase over statistical significance of differences within the groupscontrol cells grown in 10% FCS RPMI. and among groups was determined by one way ANOVA.

Reverse transcription-polymerase chain reaction (RT- A P value less than 0.05 was considered significant.PCR). Total RNA was isolated from TC transfectedwith FasL by TRI reagent (Sigma Aldrich Pty Ltd, Aus-

RESULTStralia) and reverse-transcribed into cDNA by randompriming. Polymerase chain reaction (PCR) for FasL am- FasL expression of transfected tubule cellsplification was performed for 30 cycles with a two-step NRK52E cells were transfected with a construct ofamplification (94�C for 20 sec and 60�C for 80 sec), using rFasL or hFasL containing a CMV promoter. Afterthe following primers: 5�-GCAGCAGCCCGTGAAT transfection, FasL expression in TC transfected withTAC-3� and 5�-GCTTAGGGGCTGGCTGTT-3�. PCR FasL vectors was increased (32 � 3.8%), and furthergenerated a 612-bp nucleotide for rFasL cDNA and a increased after selection by the antibiotic G480. How-822bp for the housekeeping gene GAPDH. PCR prod- ever, Fas expression in TC transfected with FasL wasucts were analyzed on an ethidium bromide-stained 1.2% not increased (Fig. 1).agarose gel. FasL mRNA expression of transfected TC was also

Detection of apoptosis using Annexin V staining [20]. examined by RT-PCR. NRK52E cells transfected withThe collected target cells were washed in PBS, resus- rFasL showed high levels of FasL expression, as com-pended in Annexin binding buffer (Clontech, Palo Alto, pared to low levels of rFasL expression in untransfectedCA, USA), and stained with Annexin V-fluorescein iso- cells (Fig. 2).thiocyanate (FITC) for five minutes at room temperature The effect of exposure to LPS or TNF-� on apoptosisin the dark using the ApoAler Annexin V Apoptosis was examined in transfected TC by Annexin V binding.

There was no significant increase of apoptosis in TCKit (Clontech). The staining solution was removed and

Wang et al: FasL cause leukocyte apoptosis1306

Fig. 1. Efficiency of FasL transfection and se-lection by the antibiotic G480. (A) FasL and(B) Fas expression in (1) untransfected TC,(2) TC transfected with rFasL, and (3) TCtransfected with hFasL. FasL expression wasincreased, but Fas expression was not changedafter transfection and selection.

detection of apoptosis by Annexin V versus TUNELmethods (Fig. 5B). Apoptosis of Jurkat cells as assessedby Annexin V, was increased ninefold after co-culturewith TC transfected with hFasL (67 � 12.1% vs. 6.7 �1.8%, P � 0.01), and eightfold with TC transfected withrFasL (60.1 � 8.8% vs. 6.7 � 1.8%, P � 0.01; Fig. 6A).However, TC transfected with hFasL or rFasL did notshow increased apoptosis after co-culture with JurkatFig. 2. Reverse transcription-polymerase chain reaction (RT-PCR)

showing semiquantitative FasL mRNA levels relative to GAPDH cells (Fig. 6B).mRNA levels: low levels in untransfected TC (lanes 1 and 2) and higherlevels in TC transfected with rFasL (lanes 3 and 4). This figure is TC transfected with FasL induce leukocyte apoptosisrepresentative of experiments repeated three times.

Leukocytes from peripheral blood and spleen wereactivated by ConA (2 �g/mL) for 12 hours. After stimula-tion, 34.0 � 5.6% of leukocytes expressed Fas on their

transfected with FasL following exposure to LPS or cell surface (Fig. 4). Co-culture was used to examine theTNF-� for 48 hours (Fig. 3A). Similarly, exposure to effects of cell surface FasL of transfected tubule cells onLPS or TNF-� for 48 hours of transfected TC co-cultured leukocyte apoptosis. Apoptosis of activated leukocyteswith untransfected TC did not increase apoptosis in ei- was increased by coculture with TC transfected withther cell type (Fig. 3B), despite of an increase in FasL rFasL (26.8 � 4.9 vs. 6.7 � 2.0% with untransfected TC,expression by TC (Fig. 3C). Stimulation of FasL expres- and vs. 8.6% � 3.1% with TC transfected with pTargetsion [for example, with interferon- (IFN-)] on leuko- vector). Insert co-culture was used to examine the effectscytes or Jurkat cells did not increase their apoptosis (data of soluble FasL secreted from transfected tubule cellsnot shown). on leukocyte apoptosis. Increased apoptosis was found

in activated leukocytes in insert co-culture with TC trans-Tubule cells transfected with FasL induce fected with rFasL compared to both untransfected TCJurkat cell apoptosis and TC transfected with pTarget vector without rFasL

Jurkat cells are known to express Fas on their cell (18.2 � 4.4% vs. 5.8 � 2.3% with untransfected TC,surface [22]. The level of Fas expression was examined and vs. 8.2 � 3.6% with pTarget PC). The increase ofin Jurkat cells. 78.5 � 15% percent of Jurkat cells ex- leukocyte apoptosis was blocked by anti-FasL antibodypressed Fas (Fig. 4). After Jurkat cells were co-cultured in both co-culture and insert culture (Fig. 7). No in-with TC transfected with rFasL for 1, 6 and 12 hours, creased apoptosis was detected in TC transfected withapoptosis was examined by both Annexin V and TUNEL FasL after co-culture or cell-insert culture with activated

leukocytes.methods (Fig. 5A). There was a strong correlation in the

Wang et al: FasL cause leukocyte apoptosis 1307

Fig. 3. (A) Apoptosis of tubule cells (TC) transfected with rFasL orhFasL as assessed by Annexin V binding. There was no significantincrease of apoptosis in transfected TC following exposure to lipopoly-saccharide (LPS 2.5 �g/mL) or tumor necrosis factor-� (TNF-� 3000U/mL) for 48 hours). Symbols are: ( ) TC; (�) rFASL TC; ( ) bFasLTC. N 12 for each bar. (B) Apoptosis of TC in primary cultureand mixed cells from primary culture of TC ( ) and TC (NRK52E)transfected with rFasL (in ratio of 1:1; �) following exposure to LPS(2.5 �g/mL) or TNF-� (3000 U/mL) for 48 hours. N 12 for each bar.(C ) FasL expression of TC in primary culture following exposure toLPS (2.5 �g/mL), TNF-� (3000 U/mL) or PMA (10 �g) � ionomycin(500 ng/mL) for 48 hours. N 12 for each bar; *P � 0.01 and **P �0.001 vs. unstimulated cells.

DISCUSSION apoptosis of transfected or untransfected TC themselves,even when stimulated by LPS and TNF-�. However, TCAlthough the Fas/FasL interaction has been utilizedtransfected with FasL increased apoptosis of Jurkat cellsas a potential therapeutic approach for cancer and au-or activated leukocytes. None of the TC cells transfectedtoimmune disorders [10, 11], and for manipulatingwith FasL showed apoptosis in co-culture with eitherinflammation in other systems [12, 13], the utility of mod-Jurkat cells or leukocytes.ulating the Fas/FasL pathway to prevent renal inflam-

Fas, a 45-kD type I membrane protein, belongs to themation and fibrosis has not been tested previously. Intumor necrosis family, and is expressed in cells of mostthese studies, we demonstrated that tubular cell transfec-

tion with either rat FasL or human FasL had no effect on organs, including the thymus, liver, heart and kidney. In

Wang et al: FasL cause leukocyte apoptosis1308

Fig. 4. Expression of Fas in Jurkat cells andleukocytes. (A) Unstimulated leukocytes, (B)leukocytes after stimulation for 12 hours with2 �g/mL ConA, and (C) Jurkat cells withoutany stimulation. This figure is representativeof experiments repeated three times.

In these studies, an attempt has been made to manipu-late Fas/FasL interaction to eliminate leukocytes. Forthis to have any clinical utility, it is critical that the ef-fector cells (tubule cells) do not undergo apoptosis aftertransfection with FasL and during interaction betweenTC and leukocytes and fibroblasts. Fas expression hasbeen demonstrated in human renal tubular epithelialcells, but interaction between Fas and FasL did not in-duce apoptosis in TC [14], indicating that apoptosis of TCin that situation was Fas-independent. However, otherstudies have shown that LPS or TNF-� increased Fasexpression as well as apoptosis in TC [26, 27], suggestingthat Fas/FasL could be involved in the apoptosis of TC.Schelling et al found that baseline-expression of Fas byTC was insufficient for functional Fas-FasL interactionbetween TC [28]. However, induction of Fas expressionby cytokines was able to produce apoptosis in varioustubule cell lines [29]. The occurrence of apoptosis wasthought to be due to the interaction of TC Fas withmembrane FasL from adjacent TC [28]. In our presentstudy, TC transfected to a high level with FasL werestimulated with LPS or TNF-� for 48 hours, and showedno increased apoptosis. Further, TC in primary culturemixed with TC transfected with FasL and exposed toFig. 5. Comparison of assessment of apoptosis by Annexin V versus

TUNEL methods. (A) After Jurkat cells were co-cultured with TC LPS or TNF-� for 48 hours in a non-serum medium, alsotransfected with rFasL for 1, 6 and 12 hours, apoptosis was assessed showed no increased apoptosis in either cell type, evenby both Annexin V ( ) and TUNEL (�) methods. (B) There was a

though primary cultured TC could be stimulated to ex-strong correlation in the detection of apoptosis by Annexin V versusTUNEL method (y 0.8497x � 2.3593, R2 0.9891). press endogenous FasL. The discrepancy between these

data and other studies could be due to species differences(rat vs. mouse and human), cell origin (TC in primaryculture vs. cell lines) and culture conditions (with serum-contrast, FasL, a 40 kD type II membrane protein, iscontaining medium vs. serum-free medium). However,expressed in a much more restricted number of tissues,our results suggest highly that TC undergoing apoptosisincluding activated T cells and natural killer cells anddo not depend, at least in rats, on Fas/FasL.in immune privileged sites such as the eye and testis.

Interactions between TC and leukocytes are complex.Moreover, it has been demonstrated that FasL in certainA conventional paradigm is that the TC, expressing Fastissues such as testis or cornea plays an important roleand under the influence of activated T cells, could under-in creating immune privilege [23]. Induction of FasL ongo apoptosis by Fas/FasL- or perforin/granzyme B-basedT cells and T cell lines results in cell death by apoptosis.cytotoxicity in chronic renal disease or in renal allograftActivation of Fas via soluble FasL or anti-Fas antibodiesrejection [30, 31]. In our present study, the co-cultureresults in apoptotic cell death in both lymphoid and cer-of activated leukocytes with TC showed no increase intain nonlymphoid tissues [24]. FasL has been shown toapoptosis in either resting TC or TC transfected withprolong pancreatic islet allograft survival when ex-

pressed on co-transplanted syngeneic myoblasts [25]. FasL. These data demonstrate that under the conditions

Wang et al: FasL cause leukocyte apoptosis 1309

Fig. 6. Cellular apoptosis assessed by Annexin V binding after 12 hours co-culture of Jurkat cells with TC. (A) Apoptosis of Jurkat cells with (1)media only, (2) recombinant FasL, (3) untransfected TC, (4) TC transfected with hFasL, or (5) TC transfected with rFasL. (B) Apoptosis of TCwith (1) untransfected TC with media only, (2) with camptothecin (2 g/mL), (3) with Jurkat cells, (4) TC transfected with hFasL plus Jurkat cells,and (5) TC transfected with rFasL plus Jurkat cells. TC transfected with either hFasL or rFasL increased apoptosis of Jurkat cells (A) but notTC themselves (B). These figures are representative of experiments repeated five times.

described, TC are resistant to T cell-mediated apoptosis.On the other hand, baseline FasL expression in normalTC was also not capable of producing apoptosis of acti-vated T cells, via a Fas/FasL reaction. TC transfectedwith FasL increased Jurkat cell and leukocyte apoptosis.This finding demonstrated the possibility that TC trans-fected with FasL could act as effector cells to eliminateleukocytes in chronic diseases without themselves under-going self-destruction.

There are two types of FasL: membrane FasL andsoluble FasL. It has been demonstrated that soluble FasLis released from the cell surface of macrophages and Tcells under metalloproteinase digestion. Recently, solu-ble FasL was reported to be generated from various TClines as well [28]. TC with high level FasL expressioncould mediate apoptosis of Fas-expressing leukocytesand fibroblasts by direct contact in pathologic situations[32]. However, it is also important to explore the possibil-ity that TC could send death signals by soluble FasL todistant interstitial cells separated from TC by tubularbasement membrane, thickened in chronic renal dis-eases. In the present studies, TC transfected with FasLon one side of a membrane-separated chamber wereable to increase the apoptosis of leukocytes on anotherside of membrane, although the efficiency of this interac- Fig. 7. Apoptosis of leukocytes after 12 hours co-culture (�) or inserttion was less than that of direct contact. culture ( ) with either media only, untransfected TC, pTarget TC (TC

transfected with pTarget vector but no FasL insert), rFasL TC, or rFasLAlthough TC transfected with FasL were observed toTC � anti-FasL antibody. Increased apoptosis was seen in rFasL cellsdeliver death signals to leukocytes in vitro, the effects compared to each of the other cell types (*P � 0.01, N 12 for eachbar) and this increased apoptosis was blocked by anti-FasL antibody.of transfection of TC with FasL are likely to be complex

Wang et al: FasL cause leukocyte apoptosis1310

7. Branton MH, Kopp JB: TGF-beta and fibrosis. Microbes Infectin vivo. As an example, it was reported that increased1:1349–1365, 1999

FasL expression of transplanted pancreatic islets led to 8. Mattey DL: Interleukin-4 induces myofibroblast differentiation issynovial fibroblasts. Biochem Soc Trans 25:290S, 1997neutrophilic infiltration and rejection [33]. To avoid

9. Brunner T, Yoo NJ, Griffith TS, et al: Regulation of CD95 ligandFasL-initiated inflammatory responses, Li et al foundexpression: a key element in immune regulation? Behring Inst Mitt

that transfection to a certain level (by �10%) could 97:161–174, 199610. Springer ML, Kraft PE, Blau HM: Inhibition of solid tumorprolong survival of rat liver allografts without invoking

growth by Fas ligand-expressing myoblasts. Somat Cell Mol Genetan inflammatory response [34]. However, overexpression24:281–289, 1998

of FasL by high dose FasL transfection caused destruc- 11. Batteux F, Tourneur L, Trebeden H, et al: Gene therapy ofexperimental autoimmune thyroiditis by in vivo administration oftion of grafts by infiltrating neutrophils. Thus, control ofplasmid DNA coding for Fas ligand. J Immunol 162:603–608, 1999the level of FasL transfection to TC seems to be impor- 12. Tsuyuki S, Bertrand C, Erard F, et al: Activation of the Fas

tant for the success of any in vivo application of these receptor on lung eosinophils leads to apoptosis and the resolutionof eosinophilic inflammation of the airways. J Clin Invest 96:2924–observations. Recently, Nabel at al demonstrated that2931, 1995TGF-� was a critical factor in the inhibition of FasL- 13. Sata M, Walsh K: Cyclosporine downregulates Fas ligand expres-

induced neutrophil infiltration [35]. Differences in the sion on vascular endothelial cells: Implication for accelerated vas-culopathy by immunosuppressive therapy. Biochem Biophys Reseffect of FasL at distinct anatomic sites (induction ofCommun 263:430– 432, 1999a neutrophilic infiltrate) versus immune-privileged sites 14. Boonstra JG, Van Der Woude FJ, Wever PC, et al: Expression

(elimination of invading immune cells) were determined and function of Fas (CD95) on human renal tubular epithelialcells. J Am Soc Nephrol 8:1517–1524, 1997by different levels of TGF-�, rather than the amount

15. Genestier L, Fournel S, Flacher M, et al: Induction of Fasof FasL. Overexpression of TGF-� in tubular cells and (Apo-1, CD95)-mediated apoptosis of activated lymphocytes by

polyclonal antithymocyte globulins. Blood 91:2360–2368, 1998interstitial fibroblasts has been demonstrated in chronic16. Ortiz A, Lorz C, Gonzalez-Cuadrado S, et al: Cytokines andrenal diseases of all types [36, 37]. Thus, it is possible that

Fas regulate apoptosis in murine renal interstitial fibroblasts. J Ama microenvironment of augmented TGF-� expression in Soc Nephrol 8:1845–1854, 1997

17. Lorz C, Ortiz A, Justo P, et al: Proapoptotic Fas ligand is ex-chronic renal disease could permit TC transfected withpressed by normal kidney tubular epithelium and injured glomer-FasL to eliminate activated leukocytes or fibroblastsuli. J Am Soc Nephrol 11:1266–1277, 2000

without causing neutrophil infiltration. 18. Wang Y, Chen J, Chen L, et al: Induction of monocyte chemoat-tractant protein-1 in proximal tubule cells by urinary protein. JIn conclusion, enhanced expression of FasL in TC isAm Soc Nephrol 8:1537–1545, 1997capable of inducing apoptosis of activated leukocytes, 19. Chen L, Boadle RA, Harris DCH: Toxicity of holotransferrin

without increasing their own susceptibility to apoptosis. but not albumin in proximal tubule cells in primary culture. J AmSoc Nephrol 9:77–84, 1998This suggests a potential role for this approach in the

20. Vermes I, Haanen C, Steffens-Nakken H, Reutelingsperger C:limitation and resolution of interstitial inflammation. A novel assay for apoptosis. Flow cytometric detection of phospha-Further experiments are necessary to test whether TC tidylserine expression on early apoptotic cells using fluorescein

labelled Annexin V. J Immunol Methods 184:39–51, 1995transfected with FasL have the same ability to eliminate21. Aubry JP, Blaecke A, Lecoanet-Henchoz S, et al: Annexin V

the renal interstitial infiltration of activated leukocytes used for measuring apoptosis in the early events of cellular cytotox-in vivo. icity. Cytometry 37:197–204, 1999

22. Martinez-Lorenzo MJ, Alava MA, Anel A, et al: Release ofpreformed Fas ligand in soluble form is the major factor for activa-

ACKNOWLEDGMENT tion-induced death of Jurkat T cells. Immunology 89:511–517, 199623. Griffith TS, Brunner T, Fletcher SM, et al: Fas ligand-inducedThis study was supported by a grant from the Australia Kidney

apoptosis as a mechanism of immune privilege. Science 270:1189–Foundation.1192, 1995

24. Ishigami T, White CA, Pender MP: Soluble antigen therapy in-Reprint requests to Dr. Yiping Wang, Department of Renal Medicine,duces apoptosis of autoreactive T cells preferentially in the targetWestmead Hospital, Westmead, NSW 2145, Australia.organ rather than in the peripheral lymphoid organs. Eur J Immu-E-mail: Yiping_Wang@wsahs.nsw.gov.aunol 28:1626–1635, 1998

25. Kang SM, Hoffmann A, Le D, et al: Immune response and my-oblasts that express Fas ligand. Science 278:1322–1324, 1997REFERENCES

26. Ortiz-Arduan A, Danoff TM, Kalluri R, et al: Regulation of1. Nath KA: Tubulointerstitial changes as a major determinant in Fas and Fas ligand expression in cultured murine renal cells and

the progression of renal damage. Am J Kidney Dis 20:1–17, 1992 in the kidney during endotoxemia. Am J Physiol 271:F1193–F1201,2. Ter Wee PM, Donker AJ: Clinical strategies for arresting progres- 1996

sion of renal disease. Kidney Int 42(Suppl 38):S114–S120, 1992 27. Koide N, Narita K, Kato Y, et al: Expression of Fas and Fas3. Ou ZL, Natori Y, Natori Y: Gene expression of CC chemokines ligand on mouse renal tubular epithelial cells in the generalized

in experimental acute tubulointerstitial nephritis. J Lab Clin Med Schartzman reaction and its relationship to apoptosis. Infect Immun133:41– 47, 1999 67:4112– 4118, 1999

4. Kelly FJ, Anderson S, Thompson MM, et al: Acute and chronic 28. Schelling JR, Nkemere N, Kopp JB, Cleveland RP: Fas-depen-renal effects of recombinant human TGF-beta2 in the rat. J Am dent fratricidal apoptosis is a mechanism of tubular epithelial cellSoc Nephrol 10:1264–1273, 1999 deletion in chronic renal failure. Lab Invest 78:813–824, 1998

5. Rovin BH, Phan LT: Chemotactic factors and renal inflammation. 29. Khan S, Cleveland RP, Koch CJ, Schelling JR: Hypoxia inducesAm J Kidney Dis 31:1065–1084, 1998 renal tubular epithelial cell apoptosis in chronic renal disease. Lab

6. Kluth DC, Rees AJ: Inhibiting inflammatory cytokines. Semin Invest 79:1089–1099, 199930. Papadimitriou JC, Drachenberg CB, Hadley GA, et al: Interac-Nephrol 16:576–582, 1996

Wang et al: FasL cause leukocyte apoptosis 1311

tion between cytolytic T-cells and immobilized renal tubular epi- 34. Li XK, Okuyama T, Tamura A, et al: Prolonged survival of ratthelial cells: Evidence of polarity on both effector and target cells. liver allografts transfected with Fas ligand-expressing plasmid.J Submicrosc Cytol Pathol 29:379–386, 1997 Transplantation 66:1416–1423, 1998

31. Bailey NC, Kelly CJ: Nephritogenic T cells use granzyme C as 35. Chen JJ, Sun Y, Nabel GJ: Regulation of the proinflammatorya cytotoxic mediator. Eur J Immunol 27:2302–2309, 1997 effects of Fas ligand (CD95L). Science 282:1714–1717, 1998

32. Arrizabalaga P, Sole M, Quinto L, et al: Leukocyte infiltration 36. Ketteler M, Noble NA, Border WA: Increased expression ofand intercellular adhesion molecule-1-mediated cell interactionstransforming growth factor-beta in renal disease. Curr Opinin immunoglobulin A nephropathy. Arch Pathol Lab Med 122:817–Nephrol Hypertens 3:446– 452, 1994822, 1998

37. Tamaki K, Okuda S, Ando T, et al: TGF-beta 1 in glomerulosclero-33. Kang SM, Lin Z, Ascher NL, Stock PG: Fas ligand expression onsis and interstitial fibrosis of adriamycin nephropathy. Kidney Intislets as well as multiple cell lines results in accelerated neutrophilic

rejection. Transplant Proc 30:538, 1998 45:525–536, 1994