Sensitization by interleukin-6 of rat hepatocytes to tumor necrosis factor α-induced apoptosis

Journal of Hepatology 38 (2003) 728–735 www.elsevier.com/locate/jhep

Sensitization by interleukin-6 of rat hepatocytes to tumor necrosis factor a-induced apoptosis Ulrike Bo¨er †, Alexandra Fennekohl †, Gerhard P. Pu¨schel* Institut fu¨r Erna¨hrungswissenschaft, Universita¨t Potsdam, Abt. Biochemie der Erna¨hrung, Arthur-Scheunert-Allee 114-116, D-14558 Bergholz-Rehbru¨cke, Germany

Background/Aims: Tumor necrosis factor (TNF) elicits hepatocyte apoptosis in toxic liver injury and is also central in hepatocyte proliferation after partial hepatectomy. In both circumstances interleukin (IL)-6 levels are also elevated. In mouse liver IL-6 attenuated Fas receptor-mediated apoptosis indicating its interference with pro-apoptotic signal chains. It was, therefore, the aim to examine the modulation by IL-6 of TNFa-induced apoptosis in rat hepatocytes. Methods: Primary rat hepatocytes were treated with IL-6 prior to induction of apoptosis with TNFa/actinomycin D or anti-Fas antibody M-20. Apoptosis was detected by determination of caspase-3 activation and bisbenzimide staining of condensed nuclei. Expression of TNFa receptors was analyzed by semi-quantitative polymerase chain reaction and ligand binding studies with [ 125I]-TNFa. Results: IL-6 treatment doubled TNFa/actinomycin D-induced caspase-3 activity and significantly enhanced chromatin condensation. By contrast IL-6 inhibited Fas-induced increase in caspase-3 activity by 45% and significantly reduced chromatin condensation. IL-6 increased the mRNA level of TNF-R1 1.35-fold and augmented cell surface binding of [ 125I]-TNFa 3-fold. The latter and TNFa-mediated caspase activation was attenuated by prostaglandin E2. Conclusions: IL-6 – in contrast to its anti-apoptotic modulation of the Fas-induced pathway – exerted a pro-apoptotic effect on the TNFa/actinomycin D-induced apoptosis by increasing the number of TNF-R on hepatocytes. q 2003 European Association for the Study of the Liver. Published by Elsevier Science B.V. All rights reserved. Keywords: Prostaglandin E2; Ligand binding; Receptor

1. Introduction Tumor necrosis factor a (TNFa) is a pleiotropic cytokine which in liver depending on the general setting can confer two opposing effects: It can elicit hepatocyte apoptosis in a number of toxic liver injury models [1–6] but its action is also required to prime hepatocytes for subsequent growth factor-stimulated cell division in liver regeneration [5,7,8]. TNFa can signal through two receptors, TNF-R1 (55 kDa) and TNF-R2 (75 kDa) [4] both of which are expressed on mouse hepatocytes [9]. Activation of TNF-R1 can lead to apoptosis via the adaptor molecules TNF-R1-associated death domain (TRADD) and Fas-associated death domain Received 2 July 2002; received in revised form 4 March 2002; accepted 7 February 2003 * Corresponding author. Tel.: 149-33200-88-694; fax: 149-33200-88541. E-mail address: [email protected] (G.P. Pu¨schel). † Both authors contributed equally to this work.

(FADD) leading to the activation of procaspase-8 terminally resulting in an activation of the effector caspase-3 [4]. In hepatocytes this pro-apoptotic stimulus is counterbalanced by signaling through TRADD, receptor interacting protein 1 and TNF receptor-associated factor (TRAF) 2 which lead to an activation of nuclear factor (NF) kB and Jun-N-terminal kinase and thereby ultimately to the de novo synthesis of anti-apoptotic proteins and to the synthesis of proteins that allow the hepatocyte to leave the Go phase and to reenter the cell cycle [4]. This latter anti-apoptotic signal chain can also be elicited via the TNF-R2 and the adaptor molecules TRAF1 and TRAF2. If NFkB activation or NFkB-dependent synthesis of mRNAs or anti-apoptotic proteins is inhibited, TNFa causes hepatocyte apoptosis because the TNFdependent pro-apoptotic signal cascade is still active. Distal of the adaptor molecule FADD, TNFa shares the pro-apoptotic signal cascade with the Apo1/Fas/CD95 signal chain [10]. The Fas-induced apoptosis has recently been shown to be inhibited by interleukin (IL)-6 [11]. Intra-

0168-8278/03/$30.00 q 2003 European Association for the Study of the Liver. Published by Elsevier Science B.V. All rights reserved. doi:10.1016/S 0168-82 78(03)00090-4

U. Bo¨ er et al. / Journal of Hepatology 38 (2003) 728–735

hepatic IL-6 production is enhanced both in liver regeneration and models of toxic liver injury [12,13]. The modulation by IL-6 of hepatocyte functions has recently been found to be modulated by prostaglandin E2 which in many situations is released concomitantly with IL-6 from nonparenchymal liver cells [14]. It, therefore, was the purpose of the current investigation to determine, whether and how IL-6 influences TNFa-induced apoptosis in primary hepatocytes and whether such a modulation is further regulated by prostaglandin E2. It was found, that IL-6 enhances TNFainduced apoptosis in hepatocytes probably via increasing cell surface expression of the TNFa receptor(s) and that this increase could partially be reversed by prostaglandin E2. 2. Materials and methods 2.1. Chemicals The source of chemicals is indicated throughout the text. Primers were custom-synthesized by MWG-Biotech (Heidelberg, Germany).

2.2. Animals Male Wistar rats (200–300 g) were purchased from Charles River (Sulzfeld, Germany) and kept on a 12 h day/night rhythm (light from 07:00 to 19:00 h) with free access to water and the standard rat diet 1326 (altromin, Gesellschaft fu¨r Tierhaltung, Lage, Germany). Treatment of the animals followed the German Law on the Protection of Animals and was performed with permission of the state animal welfare committee.

2.3. Hepatocyte preparation Rats were anaesthetized by intraperitoneal injection of Narcoren (100 ml/ 100 g). Hepatocytes were isolated basically according to Meredith without the use of collagenase [15] with minor modifications [16]. Detritus and nonparenchymal cells were removed by washing steps at 20 £ g and viable hepatocytes were further purified by a 58% Percoll gradient (Pharmacia, Freiburg, Germany). Purity and viability of hepatocytes as identified on the basis of their typical light microscopic appearance was almost 100%.

2.4. Hepatocyte culture Hepatocytes were plated on uncoated plastic 3.5-cm diameter tissue culture plates (1.3 £ 10 6 cells/plate) in M199 (Gibco-BRL, Eggenstein, Germany) containing 0.5 nM insulin, 100 nM dexamethasone (both Sigma-Aldrich, Deisenhofen, Germany), 10 mg/ml penicillin/streptomycin (Gibco-BRL, Eggenstein, Germany) and for the first 4 h of culture 4%

729

newborn calf serum (v/v) (Biochrom, Berlin, Germany). Culture was then continued in serum free medium for 26 h prior to the experiments.

2.5. Polymerase chain reaction (PCR) analysis of rat TNF receptor mRNA After a total of 30 h of culture IL-6 (100 ng/ml recombinant human, Strathmann Biotech GmbH, Hannover, Germany) was added and the culture was then continued for 16 h. Total RNA was isolated by RNeasy Kit (Qiagen, Hilden, Germany). cDNA was synthesized from 5 mg total RNA by oligo(dT)-primed reverse transcription with Superscript II (GibcoBRL, Eggenstein, Germany) according to the instructions of the manufacturer. All PCRs were repeated at least three times with cDNAs from different hepatocyte preparations. PCRs were carried out in a 50 ml reaction mix containing 30 pmol of both forward and reverse primer (Table 1), 200 mM of each deoxyribonucleotide triphosphate, 3% (v/v) DMSO, cDNA dilutions, 5 ml of PCR 10 £ buffer containing 1.5 mM MgCl2 and 0.5 U Powerscript-polymerase (PAN, Aidenbach, Germany). To amplify TNFR1 cDNA the following PCR program was applied: 3 min 958C, 35 £ (1 min 958C, 1 min 608C, 2 min 728C), 10 min 728C. TNF-R2 cDNA was amplified by 3 min 958C, 35 £ (1 min 958C, 1 min 588C, 2 min 728C), 10 min 728C. The quantity and integrity of the cDNAs were checked by PCR for b-actin using the following program: 3 min 958C, 35 £ (1 min 958C, 1 min 608C, 2 min 728C) [17]. The same cDNA preparations were compared in subsequent PCR reactions for b-actin and both TNF receptors. The cDNA stocks were diluted 1:100 and further diluted 1:4 and 1:16 in order to be in the linear range of the PCR. Products were separated electrophoretically on 2% (w/v) agarose gels and visualized by staining with ethidium bromide. Bands were analyzed densitometrically using the BioRad ChemiDoc system. The intensity of the bands of TNF-R1 and TNF-R2 were compared with the intensity of the corresponding b-actin bands.

2.6. Determination of caspase-3 activity in apoptotic hepatocytes After 30 h of culture hepatocytes were prestimulated with 100 ng/ml IL-6 for additional 16 h. For the determination of the IL-6 dose dependence 0.25–250 ng/ml IL-6 was added for 16 h and for the determination of the IL-6 time dependence 100 ng/ml IL-6 was added for 1, 2, 4, 8, 13 and 16 h. Then apoptosis was induced by adding 333 nM actinomycin D (SigmaAldrich, Deisenhofen, Germany) and 100 ng/ml TNFa (recombinant human, Strathmann Biotech GmbH, Hannover, Germany) for 9 h or alternatively with 1–2 mg/ml M-20 anti-Fas polyclonal antiserum (Santa Cruz, USA) for 24 h. For TNFa dose dependence experiments apoptosis was induced after 16 h of prestimulation with 100 ng/ml IL-6 by adding 333 nM actinomycin D and 0.1 ng/ml to 100 ng/ml TNFa. Caspase-3 activity was determined by measuring proteolytic cleavage of the fluorogenic specific substrate Asp-Glu-Val-Asp-7-amino-4-methylcoumarin (Biomol, Hamburg, Germany): The medium was removed and cells were washed two times with phosphate-buffered saline (137 mM NaCl, 2.7 mM KCl, 1.5 mM KH2PO4, 8 mM Na2HPO4, pH 7.4) and lysed in 200 ml

Table 1 Sequence and location of the PCR primers used to amplify the cDNA of b-actin, TNF-R1 and TNF-R2 a Primer name

Sequence (5 0 -3 0 )

Position

b-actin-f b-actin-r TNF-R1-f TNF-R1-r TNF-R2-f TNF-R2-r

GATATCGCTGCGCTCGTCGTC CCTCGGGGCATCGGAACC AGGCCCAGGGTCTACTCCATCATT CACGCCATCCACCACAGCATACAG GCCCCTGGGACGTTCTCTGA AGGCACCATGGTTTCTCGTTGTAG

Acc.-No. J00691; pos. 1251–1271 Acc.-No. J00691; pos. 2553–2570 (reverse) Acc.-No. NM_013091; pos. 955–978 Acc.-No. NM_013091; pos 1356–1333 (reverse) Acc.-No. AF142499; pos. 55–74 Acc.-No. AF142499; pos. 462–439 (reverse)

a The locations given are the sequence positions in the data files retrieved from GenBank under the accession number indicated. The reverse primers are the complementary sequences to the indicated positions. Primers are shown in the 5 0 to 3 0 direction.

730

U. Bo¨er et al. / Journal of Hepatology 38 (2003) 728–735

lysis buffer containing 25 mM HEPES pH 7.5, 5 mM MgCl2, 1 mM EGTA, 0.1% (v/v) Triton X 100 and as protease inhibitors 2 mM pefabloc (Biomol, Hamburg, Germany), 10 mg/ml leupeptin (Biomol, Hamburg, Germany) and 10 mg/ml trypsin inhibitor (Biomol, Hamburg, Germany). Lysates were spinned at 12 000 £ g and supernatants were diluted 1:10 in assay buffer (50 mM HEPES pH 7.5, 1% (w/v) sucrose, 0.1% (w/v) CHAPS, 10 mM DTT and 50 mM casapse-3 specific substrate DEVD-AMC). Samples were incubated at 378C for 30 min and fluorescence was quantified at 390 nm excitation and 460 nm emission with a fluorescence plate reader (Fluoroskan Ascent FL, Dynex Hybaid Labsystems, Frankfurt a.M., Germany). Protein content of each sample was determined according to Bradford et al. [18].

2.7. Determination of apoptotic cells by nuclear staining with Hoechst 33342 and propidium iodide of hepatocytes without prior fixation After 30 h of culture hepatocytes were prestimulated for 16 h with 100 ng/ml IL-6, then apoptosis was induced by adding 333 nM actinomycin D and 100 ng/ml TNFa for 9 h. Afterwards the culture medium was replaced with medium containing 1 mg/ml of the membrane-permeable fluorochrome Hoechst 33342 and 10 mg/ml of the DNA-binding fluorochrome propidium iodide which is impermeable to the normal plasma membrane but stains the nuclei of necrotic and late apoptotic cells. Cells were incubated for 15 min at 378C and thereafter immediately assessed using a Nikon fluorescent microscope with digital camera (dxm 1200, Nikon, Japan) with excitation wavelength of 330–380 nm. Nuclear staining and morphology were documented photographically and evaluated by digital image analysis.

2.8. Determination of nuclear apoptotic morphology by bisbenzimid staining Hepatocytes were treated as described for caspase-3 assay but instead of determining the enzyme activity a morphological evaluation of apoptotic cells was performed by staining with bisbenzimid trichloride (SigmaAldrich, Deisenhofen, Germany): Hepatocytes were fixed with 4% (w/v) paraformaldehyde (Sigma-Aldrich, Deisenhofen, Germany) for 30 min, washed with distilled water and stained with 8 mg/ml bisbenzimid for 1 min. After washing with distilled water nuclear morphology was determined using a Nikon fluorescent microscope with digital camera (dxm 1200, Nikon, Japan). Cells whose nuclei exhibited light blue bright staining and condensed chromatin were designated as apoptotic, those with white bright staining and nuclear fragments as strongly fragmented. For each sample, 900–1200 nuclei were counted and number of apoptotic cells was expressed as percentage.

2.9. [ 125I]-TNFa -binding on cultured hepatocytes For surface ligand binding assays with cultured hepatocytes cells were cultured 30 h before adding 100 ng/ml IL-6 ^ 10 mM prostaglandin E2 (PGE2) for 16 h, washed twice with ice cold incubation buffer containing 15 mM HEPES pH 7.4, 140 mM NaCl, 4.7 mM KCl, 2.2 mM CaCl2, 1.2 mM KH2PO4 and 11 mM glucose and than incubated for 1 h at 48C with 100 pM [ 125I]-TNFa (Amersham Pharmacia Biotech, Freiburg, Germany). Nonspecific binding was determined in the presence of 10 nM TNFa. Culture plates were washed four times with ice cold incubation buffer and scraped into incubation buffer. Radioactivity was determined in a g-counter (Packard, Frankfurt a. M., Germany). To determine the effect of a delayed PGE2 administration hepatocytes were incubated for 16 h in presence of 100 ng/ ml IL-6 and culture was then continued for another 8 h in presence of 100 ng/ml IL-6 ^ 10 mM PGE2. The binding assay was performed as described above.

3. Results 3.1. Enhancement by IL-6 of TNFa -induced caspase-3 activity in hepatocytes In rat hepatocyte cultures maintained under serum free conditions for a total of 55–70 h a low basal caspase-3 activity was detectable probably corresponding to a low frequency of spontaneous apoptotic events in theses cultures (Fig. 1). This basal caspase activity was not modulated by prior incubation of hepatocyte cultures with IL-6 (100 ng/ ml, 16 h). There was an about 2-fold increase in caspase-3 activity if hepatocytes were cultured for the last 24 h in the presence of an anti-Fas antibody (Fig. 1). As reported previously [11] this anti-Fas antibody-induced caspase-3 activation was completely blocked by prior incubation of hepatocytes with IL-6. Neither TNFa nor actinomycin D alone increased caspase-3 activity in hepatocyte cultures significantly (Fig. 1). By contrast the combination of actinomycin D with TNFa led to a 3-fold increase in caspase-3 activity in hepatocyte cultures. In contrast to the anti-Fas antibody-induced caspase activity the TNFa/actinomycin D-induced activation of caspase-3 was further enhanced by prior incubation of the hepatocytes with IL-6 (Fig. 1). The caspase-3 activity was dependent on the TNFa concentration (Fig. 2). For any concentration of TNFa the TNFa/actinomycin D-dependent increase in caspase-3 activity was higher after pretreatment of cultured hepatocytes with 100 ng/ml IL-6 for 16 h than in controls. The EC50, which due to the low maximal activity could not be determined for control cells with certainty, was in the range of 10 ng/ml TNFa and appeared not to be affected by prior stimulation with IL-6. The augmentation by IL-6 of the TNFa/actinomycin Dinduced activation of caspase-3 was tested with regard to its time and dose dependence. The dose response curve was sigmoidal with an EC50 of 3.5 ng/ml (160 pM, Fig. 3). IL6 enhanced TNFa/actinomycin D-induced activation of caspase-3 as early as 1 h after administration. The IL-6dependent enhancement continued to increase slightly hyperbolically up to 16 h, the last time point taken (Fig. 4). The TNFa/actinomycin D-induced activation of caspase-3 was reduced by PGE2 in IL-6-pretreated hepatocytes but not in hepatocytes cultured in absence of IL-6 (Fig. 5). Thus, the enhancement by IL-6 of the TNFa/actinomycin D-induced activation of caspase-3 was partially antagonized by PGE2. 3.2. Enhancement by IL-6 of TNFa -induced chromatin condensation Apoptotic cells can be identified by staining the condensed chromatin of their nuclei with the fluorescent dye bisbenzimide. Apoptotic cells were counted and were ranked into two groups according to the intensity of nuclear staining and fragmentation: group 1 were cells with begin-

U. Bo¨ er et al. / Journal of Hepatology 38 (2003) 728–735

731

membranes to discriminate necrotic cells [19–22]. Digital image analysis was performed and those cells that showed nuclear fluorescence above the background threshold value and no propidium idodide staining were counted as apoptotic cells (Table 2). TNFa/actinomycin-D treatment increased the number of apoptotic cells 2-fold over basal in control hepatocytes. By contrast, TNFa/actinomycin-D treatment increased the number of apoptotic cells 3-fold over basal in IL-6-pretreated hepatocytes. The number of apoptotic cells in IL-6-treated hepatocytes was significantly higher than in control cells. 3.3. Increase by IL-6 of TNFa receptor expression on the hepatocyte surface

Fig. 1. Increase by IL-6 in TNFa-induced caspase-3 activity. Primary rat hepatocytes were prepared as described under Section 2, cultivated for 24 h and prestimulated with 100 ng/ml IL-6 for 16 h. Apoptosis was induced by 333 nM actinomycin D and 100 ng/ml TNFa for 9 h or 2 mg/ ml anti-Fas polyclonal antiserum for 24 h. Apoptosis was determined as cleavage of the caspase-3 specific fluorogenic substrate DEVD-AMC and fluorescence was quantified using a microtiter plate fluorescence reader. Values are means ^ SEM of three to five independent experiments. Statistics: Student’s t-test for unpaired samples.

ning chromatin condensation, appearing as light blue fluorescent nuclei with white dots and group 2 were cells with advanced chromatin fragmentation, appearing as bright white dots scattered throughout the cell (Fig. 6A). Both types of cells were counted and expressed as percent of total cell number. In untreated cell cultures about 5% of the cells showed signs of beginning apoptosis; about 1% of the hepatocytes had strongly fragmented nuclei (Fig. 6B). IL-6 treatment did not influence this basic apoptotic rate. The number of hepatocytes with condensed chromatin or strongly fragmented nuclei was doubled by incubation with the anti-Fas antibody; preincubation with IL-6 almost completely abolished the anti-Fas antibody-induced chromatin condensation and nuclear fragmentation (Fig. 6B). Roughly 23% of the hepatocytes in TNFa/actinomycin Dtreated cultures had condensed chromatin, 4% had strongly fragmented nuclei. The number of cells with chromatin condensation was slightly increased by IL-6 whereas the number of fragmented nuclei was doubled (Fig. 6B). Thus, the results of the morphological analysis are in line with the results of the caspase-3 assay and indicate, that IL-6 attenuated the anti-Fas antibody-induced apoptosis while it enhanced the TNFa/actinomycin D-induced apoptosis. To further corroborate the finding that IL-6 increases the TNFa/actinomycin D-induced apoptosis an additional assay was performed using the cell permeable dye Hoechst 33342 to stain apoptotic nuclei in fresh cells without prior fixation and propidium iodide which only permeates leaky plasma

To elucidate a possible mechanism by which IL-6 could enhance TNFa-induced apoptosis, the IL-6-dependent modulation of expression of TNF-R1 and TNF-R2 was examined. Densitometrical analysis of semi quantitative reverse transcriptase-PCRs with receptor subtype specific primer pairs revealed a slight but significant increase of TNF-R1 mRNA but not TNF-R2 mRNA in IL-6 hepatocytes (Fig. 7). Bearing in mind the limitation of the technique, the results have to be interpreted with very much caution. To corroborate the findings cell surface expression of TNF-R was further characterized by ligand binding studies with [ 125I]-TNFa on whole cells, which, however, do not allow to discriminate between TNF-R1 and TNF-R2. Binding was performed at 48C to prevent internalization.

Fig. 2. TNFa-dependent increase in caspase-3 activity. Primary rat hepatocytes were prepared as described under Section 2, cultivated for 24 h and incubated with or without 100 ng/ml IL-6 for additional 16 h. Then apoptosis was induced by 333 nM actinomycin D and 0.1, 1, 2.5, 5, 10, 25, 50 and 100 ng/ml TNFa for 9 h. Apoptosis was determined as cleavage of the caspase-3 specific fluorogenic substrate DEVD-AMC and fluorescence quantified using a microtiter plate fluorescence reader. Values are means ^ SEM of three to four independent experiments.

732

U. Bo¨ er et al. / Journal of Hepatology 38 (2003) 728–735

Fig. 3. Dose dependence of the IL-6-mediated increase in TNFainduced caspase-3 activity. Primary rat hepatocytes were prepared as described under Section 2, cultivated for 24 h and prestimulated with 0.25–250 ng/ml IL-6 for 16 h. Apoptosis was induced by 333 nM actinomycin D and 100 ng/ml TNFa for 9 h. Apoptosis was determined as cleavage of the caspase-3 specific fluorogenic substrate DEVD-AMC and fluorescence quantified using a microtiter plate fluorescence reader. Values are means ^ SEM of five independent experiments, the maximum IL-6 induced increase in caspase activity over the respective control in the same experiment was set 100%.

Fig. 5. Inhibition by PGE2 of the IL-6-mediated increase in TNFainduced caspase-3 activity. Primary rat hepatocytes were prepared as described under Section 2, cultivated for 24 h and prestimulated with 100 ng/ml IL-6 for 16 h. Apoptosis was induced by 333 nM actinomycin D and 100 ng/ml TNFa for 9 h in presence or absence of 10 mM PGE2. Apoptosis was determined as cleavage of the caspase-3 specific fluorogenic substrate DEVD-AMC and fluorescence was quantified using a microtiter plate fluorescence reader. Values are means ^ SEM of three to five independent experiments. TNFa/actinomycin D-induced caspase-3 activity without IL-6 preincubation in the respective experiments was set 100%. Statistics: Student’s t-test for unpaired samples; a: P , 0.05.

The number of binding sites determined in this way on intact hepatocytes was increased 2.5-fold by IL-6 treatment (Fig. 8A). Thus, the enhancement by IL-6 of TNFa-induced apoptosis might reflect a sensitization of the hepatocyte towards TNFa by increasing the number of TNFa-R on the hepatocyte cell surface. The increase by IL-6 of TNFa cell surface binding was almost abolished if PGE2 was added simultaneously with IL-6 (Fig. 8A). If, as in the experiments for the caspase-3 assay, IL-6 was added first Table 2 Determination of the percentage of apoptotic cells in control and IL-6pretreated primary rat hepatocytes a

Basal TNFa Actinomycin D TNFa 1 actinomycin D Fig. 4. Time dependence of the IL-6-mediated increase in TNFainduced caspase-3 activity. Primary rat hepatocytes were prepared as described under Section 2, cultivated for 24 h and prestimulated with 100 ng/ml IL-6 for 1, 2, 4, 8, 13 and 16 h. Apoptosis was induced by 333 nM actinomycin D and 100 ng/ml TNFa for 9 h. Apoptosis was determined as cleavage of the caspase-3 specific fluorogenic substrate DEVD-AMC and fluorescence quantified using a microtiter plate fluorescence reader. Values are means ^ SEM of five independent experiments. TNFa/actinomycin D-induced caspase-3 activity without IL-6 preincubation in the respective experiments was set 100%.

Control

IL-6

8.09 ^ 2.04 9.77 ^ 1.61 9.86 ^ 1.08 17.18 ^ 3.54

8.29 ^ 2.12 8.92 ^ 2.56 10.35 ^ 1.56 24.47 ^ 5.12*

a Primary rat hepatocytes were prepared as described under Section 2, cultivated for 24 h and prestimulated with 100 ng/ml IL-6 for 16 h. Apoptosis was induced by adding 333 nM actinomycin D and 100 ng/ml TNFa for 9 h. Apoptosis was determined by staining with Hoechst 33342 and propidium iodide of fresh cells without prior fixation and digital image analysis. Hoechst 33342-positive cells were counted and expressed as % of total cell number. Values are means ^ SEM of six experiments from three independent cell preparations. Statistics: Student’s t-test for unpaired samples; *: P , 0.05 IL-6 versus control of TNFa.

U. Bo¨ er et al. / Journal of Hepatology 38 (2003) 728–735

733

transcription during TNFa application, hepatocytes undergo TNFa-induced apoptosis. Apoptosis was quantified in hepatocyte cultures by two different approaches: (1) caspase-3 activity was measured. Caspase-3 is an effector caspase which is activated by upstream caspases and cleaves inhibitor of caspase-activated DNase (ICAD). Free CAD can then degrade DNA [10]. Therefore caspase-3 activation is a reliable indicator of the apoptotic status of a cell since activation of the initiator caspases must have been preceded and cell death follows immediately after DNA cleavage. Measurement of caspase-3 activity was suitable for determining both the dose-response curve and the time course of the IL6 effect on the TNFa/actinomycin D-induced apoptosis. (2) The number of apoptotic cells and the degree of nuclear fragmentation was determined by bisbenzimide staining. 4.1. Sensitization by IL-6 of hepatocytes towards TNFa / actinomycin D-induced apoptosis

Fig. 6. Increase by IL-6 in TNFa-induced chromatin condensation. Primary rat hepatocytes were prepared as described under Section 2, cultivated for 24 h and prestimulated with 100 ng/ml IL-6 for 16 h. Apoptosis was induced by 333 nM actinomycin D and 100 ng/ml TNFa for 9 h or 2 mg/ml anti-Fas polyclonal antiserum for 24 h. Apoptosis was determined as chromatin condensation visualized by bisbenzimide staining. Cells with condensed chromatin (arrowheads) and strongly fragmented nuclei (arrows) (A) were counted separately; and expressed as % of total cell number (B). Values are means ^ SEM of three independent experiments, at least 1000 cells per assay were counted. Statistics: Student’s t-test for unpaired samples; a: P , 0.05.

Both experimental approaches showed that IL-6 pretreatment sensitized hepatocytes towards TNFa/actinomycin Dinduced apoptosis (Figs. 1 and 6, Table 2). The sensitization was fast and sustained (Fig. 4) and occurred at physiologically relevant IL-6 concentrations (Fig. 3), the EC50 was in the range of the Kd of 500 nM that was reported previously for the IL-6 receptor [27]. The bisbenzimide staining indicated, that the sensitivity of the individual cell rather than the number of susceptible cells was increased by IL-6: The total number of cells showing TNFa/actinomycin D-

and PGE2 was added after 16 h and culture was continued for another 8 h in presence of PGE2 and IL-6, the IL-6induced increase in TNF-R cell surface expression was partially reverted by PGE2 (Fig. 8B). This indicated that the attenuation by PGE2 of the IL-6-induced increase in TNFa/actinomycin D-dependent caspase-3 activation might be due to the inhibition of IL-6-dependent increase in TNF-R cell surface expression. 4. Discussion TNFa is critically involved in hepatocyte apoptosis in various models of liver injury as well as in initiation of hepatocyte proliferation in liver regeneration. It is released upon various stimuli from resident liver macrophages [23– 25], bile duct epithelia, intrahepatic vascular endothelial cells [26], or immigrating leukocytes. It acts onto the hepatocyte via the cell surface receptors TNF-R1 and TNF-R2. The current study showed, that pretreatment of hepatocytes with IL-6 sensitized hepatocytes to TNFa. With the experimental conditions chosen, i.e. inhibition of

Fig. 7. Induction by IL-6 of TNF-R1 mRNA. Primary rat hepatocytes were prepared as described under Section 2, cultivated for 24 h and prestimulated with 100 ng/ml IL-6 for 16 h. RNA was isolated and cDNA was synthesized as described. PCR was performed using specific primer pairs. PCR bands were quantified densitometrically and normalized to b-actin bands. Values are means ^ SEM of three independent experiments. Statistics: Student’s t-test for unpaired samples; a: P , 0.05.

734

U. Bo¨ er et al. / Journal of Hepatology 38 (2003) 728–735

of TNFa receptors in liver plasma membranes has been postulated previously [28]. These latent receptors were made accessible to ligand either by mild detergent treatment of the membranes or by homologous upregulation with TNFa. By contrast, an IL-6-induced increase in TNF-R expression in HepG2 cells was inhibited by actinomycin D, inferring a transcriptional level of control [29]. However, TNF-R mRNA was not determined in this latter study. 4.3. Inhibition by PGE2 of the IL-6-induced increase in TNF-R expression on hepatocytes

Fig. 8. Induction by IL-6 of TNFa binding sites on hepatocytes and inhibition by PGE2. Primary rat hepatocytes were prepared as described under Section 2, cultivated for 24 h and (A) prestimulated with 100 ng/ml IL-6 ^ 10 mM PGE2 for 16 h. Culture plates were washed twice with ice-cold incubation buffer and then incubated for 1 h at 48C with 100 pM [ 125I]-TNFa ^ 10 nM unlabeled TNFa to determine unspecific binding. Culture plates were washed four times with ice-cold incubation buffer and surface-bound radioactivity was determined. Values are means ^ SEM of three to six independent experiments. (B) Cells were prepared and cultivated as described for A, however, PGE2 was added after 16 h of prestimulation with IL-6 and culture was then continued for another 8 h in presence of IL-6 ^ 10 mM PGE2 before the binding assay was performed as described for A. Statistics: Student’s t-test for unpaired samples; a: P , 0.05.

induced beginning chromatin condensation remained constant while the number of TNFa/actinomycin D-dependently strongly condensed nuclei increased significantly after prior incubation with IL-6. Similarly, the number of apoptotic cells determined by digital image analysis increased (Table 2). In line with previous findings by Kovalovich et al. [11] the Apo1/Fas/CD95-induced apoptosis was attenuated by IL-6. Since TNFa and Fas share a common intracellular signal chain distal of the adaptor protein FADD, the sensitization of hepatocytes to a TNFa-induced apoptosis most likely occurred at a site proximal to FADD. The apparent increase in TNFa receptor number on the hepatocyte surface offers a plausible explanation. 4.2. Increase by IL-6 of TNF-R on hepatocytes The sensitization was accompanied by an increase in TNF-R1 mRNA (Fig. 7) and, more importantly, an increase in the number of TNFa binding sites on the hepatocyte surface (Fig. 8A). Although there was a slight but significant increase in TNF-R1 mRNA after IL-6 treatment, this small increase in steady state mRNA levels appears not to be sufficient to account for the 3-fold increase in TNFa binding, indicating that there might be an (additional) posttranscriptional level of regulation of TNF-R cell surface expression. This could for example be the mobilization of a latent pool of TNF-R. The existence of such a latent pool

The IL-6-dependent increase in TNFa/actinomycin Dinduced caspase-3 activation was partially reduced if PGE2 was added simultaneously with TNFa/actinomycin D. Previously it had been shown, that IL-6 can induce the de novo expression of the Gs-coupled prostaglandin E2 receptors EP2-R and EP4-R thereby establishing a feedback inhibition loop by which PGE2 can inhibit the IL-6-induced a2-macroglobulin expression in hepatocytes [14]. It seemed, therefore, plausible that PGE2 by a similar mechanism might inhibit the IL-6- induced increase in TNF-R expression and hence the increase in sensitivity of the hepatocyte towards TNFa/actinomycin D-induced apoptosis. The current findings support this assumption: PGE2 abolished IL-6-induced increase TNFa cell surface binding when administered simultaneously with IL-6 (Fig. 8A). In addition, PGE2 partially reverted the IL-6-induced increase in TNFa cell surface binding when added after IL-6 (Fig. 8B). 4.4. The IL-6-dependent sensitization of hepatocytes towards TNFa might not be restricted to its pro-apoptotic effect It is important to keep in mind that the apoptosis induced by TNFa in the current study depended on the highly artificial setting of a simultaneous inhibition by actinomycin D of transcription. In absence of transcriptional inhibitors TNFa does not induce hepatocyte apoptosis but rather is essential to allow growth factor-dependent hepatocyte proliferation [8]. Therefore, even though only apoptosis was determined in this study, IL-6 treatment will probably result in a general sensitization of the hepatocyte to TNFa if the sensitization is due to the increase in cell surface TNF-R number. Hence, in a situation where the transcriptiondependent anti-apoptotic/cell cycle-promoting branch of the TNF-R signal chain is not artificially interrupted, IL-6 might promote the priming of hepatocytes by TNFa that appears to be the prerequisite for growth factor-dependent hepatocyte cell division [8]. Acknowledgements The excellent technical assistance of Manuela Kuna is gratefully acknowledged. This work was supported by the Deutsche Forschungsgemeinschaft through the Sonder-

U. Bo¨ er et al. / Journal of Hepatology 38 (2003) 728–735

forschungsbereich 366, Teilprojekt A16, Sonderforschungsbereich 402, Teilprojekt B6 and the Fonds der Chemischen Industrie.

References [1] Leist M, Gantner F, Naumann H, Bluethmann H, Vogt K, BrigeliusFlohe R, et al. Tumor necrosis factor-induced apoptosis during the poisoning of mice with hepatotoxins. Gastroenterology 1997;112:923–934. [2] Czaja MJ, Xu J, Alt E. Prevention of carbon tetrachloride-induced rat liver injury by soluble tumor necrosis factor receptor. Gastroenterology 1995;108:1849–1854. [3] Gantner F, Leist M, Jilg S, Germann PG, Freudenberg MA, Tiegs G. Tumor necrosis factor-induced hepatic DNA fragmentation as an early marker of T cell-dependent liver injury in mice. Gastroenterology 1995;109:166–176. [4] Bradham CA, Plumpe J, Manns MP, Brenner DA, Trautwein C. Mechanisms of hepatic toxicity. I. TNF-induced liver injury. Am J Physiol 1998;275:G387–G392. [5] Diehl AM. Cytokine regulation of liver injury and repair. Immunol Rev 2000;174:160–171. [6] Kanzler S, Galle PR. Apoptosis and the liver. Semin Cancer Biol 2000;10:173–184. [7] Diehl AM, Rai RM. Liver regeneration 3: Regulation of signal transduction during liver regeneration. FASEB J 1996;10:215–227. [8] Fausto N. Liver regeneration. J Hepatol 2000;32:19–31. [9] Adamson GM, Billings RE. Tumor necrosis factor: receptor binding and expression of receptors in cultured mouse hepatocytes. J Pharmacol Exp Ther 1994;269:367–373. [10] Medema JP, Scaffidi C, Kischkel FC, Shevchenko A, Mann M, Krammer PH, et al. FLICE is activated by association with the CD95 deathinducing signaling complex (DISC). EMBO J 1997;16:2794–2804. [11] Kovalovich K, Li W, DeAngelis R, Greenbaum LE, Ciliberto G, Taub R. Interleukin-6 protects against Fas-mediated death by establishing a critical level of anti-apoptotic hepatic proteins FLIP, Bcl-2, and BclxL. J Biol Chem 2001;276:26605–26613. [12] Iwai M, Cui TX, Kitamura H, Saito M, Shimazu T. Increased secretion of tumour necrosis factor and interleukin 6 from isolated, perfused liver of rats after partial hepatectomy. Cytokine 2001;13:60–64. [13] Fausto N, Laird AD, Webber EM. Liver regeneration. 2. Role of growth factors and cytokines in hepatic regeneration. FASEB J 1995;9:1527–1536. [14] Fennekohl A, Lucas M, Pu¨ schel GP. Induction by interleukin 6 of G(s)-coupled prostaglandin E(2) receptors in rat hepatocytes mediating a prostaglandin E(2)-dependent inhibition of the hepatocyte’s acute phase response. Hepatology 2000;31:1128–1134. [15] Meredith MJ. Rat hepatocytes prepared without collagenase:

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

735

prolonged retention of differentiated characteristics in culture. Cell Biol Toxicol 1988;4:405–425. Pu¨ schel GP, Kirchner C, Schro¨ der A, Jungermann K. Glycogenolytic and antiglycogenolytic prostaglandin E2 actions in rat hepatocytes are mediated via different signalling pathways. Eur J Biochem 1993;218:1083–1089. Fennekohl A, Schieferdecker HL, Jungermann K, Pu¨ schel GP. Differential expression of prostanoid receptors in hepatocytes, Kupffer cells, sinusoidal endothelial cells and stellate cells of rat liver. J Hepatol 1999;30:38–47. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976;72:248–254. Kurose I, Higuchi H, Miura S, Saito H, Watanabe N, Hokari R, et al. Oxidative stress-mediated apoptosis of hepatocytes exposed to acute ethanol intoxication. Hepatology 1997;25:368–378. Rauen U, Polzar B, Stephan H, Mannherz HG, de Groot H. Coldinduced apoptosis in cultured hepatocytes and liver endothelial cells: mediation by reactive oxygen species. FASEB J 1999;13:155–168. Gressner AM, Lahme B, Mannherz HG, Polzar B. TGF-betamediated hepatocellular apoptosis by rat and human hepatoma cells and primary rat hepatocytes. J Hepatol 1997;26:1079–1092. Dive C, Gregory CD, Phipps DJ, Evans DL, Milner AE, Wyllie AH. Analysis and discrimination of necrosis and apoptosis (programmed cell death) by multiparameter flow cytometry. Biochim Biophys Acta 1992;1133:275–285. Reinstein LJ, Lichtman SN, Currin RT, Wang J, Thurman RG, Lemasters JJ. Suppression of lipopolysaccharide-stimulated release of tumor necrosis factor by adenosine: evidence for A2 receptors on rat Kupffer cells. Hepatology 1994;19:1445–1452. Dieter P, Hempel U, Kamionka S, Kolada A, Malessa B, Fitzke E, et al. Prostaglandin E2 affects differently the release of inflammatory mediators from resident macrophages by LPS and muramyl tripeptides. Mediators Inflamm 1999;8:295–303. Dieter P, Ambs P, Fitzke E, Schwende H. Lipopolysaccharide and liposome-encapsulated MTP-PE-induced cytotoxicity and release of eicosanoids, tumor necrosis factor- alpha and nitric oxide in liver macrophages. Adv Exp Med Biol 1997;407:485–490. Loffreda S, Rai R, Yang SQ, Lin HZ, Diehl AM. Bile ducts and portal and central veins are major producers of tumor necrosis factor alpha in regenerating rat liver. Gastroenterology 1997;112:2089–2098. Zohlnhofer D, Graeve L, Rose-John S, Schooltink H, Dittrich E, Heinrich PC. The hepatic interleukin-6 receptor. Down-regulation of the interleukin-6 binding subunit (gp80) by its ligand. FEBS Lett 1992;306:219–222. Han HM, Kolhatkar AA, Marino MW, Manchester KM, Donner DB. Identification, characterization, and homologous up-regulation of latent (cryptic) receptors for tumor necrosis factor-alpha in rat liver plasma membranes. J Biol Chem 1990;265:18590–18594. Van Bladel S, Libert C, Fiers W. Interleukin-6 enhances the expression of tumor necrosis factor receptors on hepatoma cells and hepatocytes. Cytokine 1991;3:149–154.