Kidney International, Vol. 45 (1994), pp. 1562—1571

Cytokine-induced expression of leukemia inhibitory factor inrenal mesangial cells

ANDREA HARTNER, R. BERND STERZEL, NORBERT REINDL, GERTRUD M. HOCKE,GEORG H. FEY, and MARGARETE GOPPELT-STRUEBE

Medizinische Klinik IV und Inst 1w: für Genetik, Universitaet Erlangen-Nuernberg, Erlangen, Germany

Cytokine-induced expression of leukemia inhibitory factor in renalmesangial cells. Leukemia inhibitory factor (LIF) is a pleiotropiccytokine, which shares many characteristics with interleukin-6 (IL-6).Recent observations indicate a role for LIF in inflammatory processes.To examine the potential involvement of LIF in the regulation ofmesangial cell behavior, we studied LIF expression in early primarycultures of rat and human mesangial cells, as well as the response ofmesangial cells to exogenous LIF. Growing or growth-arrested ratmesangial cells constitutively expressed very low levels of LIF mRNA,barely detectable by Northern blot analysis. Strong induction of LIFmRNA expression was caused by cytokines, such as interleukin-l/3 (5ng/ml), tumor necrosis factor a (100 ng/ml) and PDGF (100 ng/ml), aswell as LPS (200 ng/ml). The induction was transient with a peak afterthree to five hours. Dexamethasone (0.1 LM) almost completely inhib-ited the induction of LIF. Weak induction of LIF mRNA was observedafter stimulation with basic fibroblast growth factor, endothelin andtransforming growth factor . In combination with IL-I p, TGF/3showed synergistic effects on LIF induction. LIF itself or IL-6 had noeffect on LIF mRNA expression. A similar induction pattern wasobserved for the expression of IL-6 mRNA. LIF protein was detectedby specific ELISA in the supernatants of human mesangial cellsstimulated with LPS or IL-l/3. In addition, we found that mesangialcells not only express LIF but they are also target cells for LIF,Recombinant LIF effectively induced transient expression of the im-mediate early genes, c-fos,jun-B and Egr-l in rat mesangial cells, witha maximum at 30 to 60 minutes. LIF was not mitogenic for mesangialcells. Our findings indicate that glomerular mesangial cells produce andreact to LIF. As a cytokine with autocrine potential, LIF may play aphysiological and/or pathophysiological role in the glomerulus, theexact nature and relevance of which remain to be explored.

Leukemia inhibitory factor (LIF) belongs to a group ofpleiotropic factors such as IL-6, G-CSF, IL-Il and oncostatinM, which exert multiple effects on a great variety of embryonic,developing or neoplastic cells of different origin [1—3]. LIF is apolypeptide with a molecular mass of 20 kDa. Native LIF ishighly glycosylated but its biological activity appears to beindependent of glycosylation [1, 4, 5]. While immature cellsseem to be important target cells for LIF, this cytokine has alsobeen shown to act on terminally differentiated cells such ascertain peripheral blood cells [3, 6], hepatocytes [7, 8], bonecells [9] or adipocytes [10]. The functional role of LIF in these

Received for publication August 18, 1993and in revised form January 7, 1994Accepted for publication January 7, 1994

© 1994 by the International Society of Nephrology

cell systems varies considerably, ranging from the inhibition oflipoprotein lipase [10] to the stimulation of calcium release frombone tissue [11, 12]. LIF is involved in the regulation of plateletformation, osteoblast and neuronal cell functions, calcium andlipid metabolism, tumor related cachexia and the production ofacute-phase proteins [11.

At present, the molecular mechanisms of signal transductionand the intracellular targets of LIF are largely unknown,although one endpoint of the cascade, the LIF responsiveelement, has recently been mapped [13]. The half-life of LIF inplasma is short (3 to 5 mm) and it is therefore considered alocally acting cytokine [1].

So far, most studies on the expression or action of LIF invitro have been performed with immature cells or cell lines.Expression of LIF in the adult organism is restricted to certainorgans or mature cell types [14]. LIF mRNA was detected inactivated normal T cells and certain T cell lines but not in Bcells or granulocytes [15]. Upon activation, LIF is released intothe supernatants of cultured human monocytes [16]. LIF isconstitutively expressed in several human mesenchymal cells(fibroblasts and endothelial cells) [15]. In these cells, expressionof LIF was found to be enhanced by the inflammatory cytokinesinterleukin I /3 (IL- 1/3) and tumor necrosis factor a (TNFa).Recently, LIF was shown to be expressed in human cartilageand synovium. Its expression in these tissues was stimulated byseveral cytokines and also by LIF itself [17].

Induction of LIF expression often occurs in parallel withinduction of interleukin 6 (IL-6) expression [14]. LIF and IL-6are induced by LPS, and glucocorticoids prevent this inductionin macrophages and hepatoma cells (Baffet and Fey, unpub-lished data). In the kidney, only IL-6 expression has beendemonstrated in glomeruli and primary cultures of renal mesan-gial cells [18—21]. Relatively high levels of LIF mRNA weredetected in neonatal rat kidney, but no comparison to adultkidney was made [22]. The development of Na-dependenthexose transport was found to be inhibited by LIF in the renalepithelial LLC-PK1 cell line, indicating a role for LIF in thedeveloping kidney [23]. The expression of IL-6 in mesangialcells supports the hypothesis that these cells might also expressLIF, and that LIF might be involved in the pathophysiology ofthe kidney as has been suggested for IL-6 [21, 24, 25].

Here we report that LIF is expressed in mesangial cells ofhuman and rat origin. Its expression is stimulated in a cellspecific manner by LPS and certain cytokines, such as IL-lp

1562

Hartner et a!: Leukemia inhibitory factor 1563

and platelet-derived growth factor (PDGF), while neither IL-6nor LIF itself are inducers of LIF in these cells. Similar to IL-6,LIF also activates mesangial cells in an autocrine manner toexpress immediate early genes, suggesting a potential role forthis cytokine in renal physiology and pathology.

Methods

Materials

CHO cells transfected with recombinant human LIF wereprovided by Genetics Institute (Cambridge, Massachussetts,USA). The supernatant of these cells was used as a source ofLIF activity. The concentration of LIF in these supernatantswas approximately 2 jsg/ml LIF protein as determined with acommercially available ELISA (R&D Systems, Minneapolis,Minnesota, USA), corresponding to 500 U/p! determined withthe FAO/HB3 assay [13]. For stimulation of mesangial cells orembryonic stem cells, LIF was used at a final concentration ofapproximately 2 ng/mI. LPS from E. coli, serotonin (5-HT) andcycloheximide were from Sigma (Deisenhofen, Germany), ba-sic fibroblast growth factor (bFGF), PDGF (BB) and humanIL-6 were from Boehringer (Mannheim, Germany), endothelinfrom Sigma, transforming growth factor p (TGFJ3) from Gibco-BRL (Eggenstein, Germany). TNFa with a specific activity of 8x 106 U/mg was provided by BASF/Knoll (Ludwigshafen,Germany), recombinant human IL-1f3 with a specific activity of2 x iO U/pg was provided by D. Borasci, Sclavo (Siena, Italy).3H-thymidine was purchased from Amersham Buchler (Braun-schweig, Germany). Embryonic stem cells were obtained fromV. Quaranta (La Jolla, California, USA) [26], buffalo rat livercells (BRL cells) were from American Type Culture Collection.

Cell culture

Rat mesangial cells were isolated and cultured as described[27]. The cells showed the typical smooth muscle cell-likemorphology and positive immunocytochemical staining for Thy1.1, smooth muscle cell actin and myosin. Cells were grown inDulbecco's Modified Eagle Medium (Gibco) containing 10%FCS (Gibco), 5 tg/ml insulin, 100 U/mI penicillin, 100 g/mlstreptomycin and 2 mi L-glutamine (Sigma) in a 95% air, 5%CO2 humidified atmosphere at 37°C. The cells were used forexperiments between passages 7 and 15. Subconfluent cultureswere serum-starved for four days with DMEM containing 0.5%FCS before stimulation.

Whole glomeruli were isolated as described [27] and trans-ferred into cell culture flasks (30,000 each) with DMEM con-taining 0.5% FCS. The isolated glomeruli were stimulated with10 g/ml LPS. After five hours glomeruli were washed twicewith PBS and homogenized in guanidinium isothiocyanate witha Dounce homogenizator. RNA was prepared and NorthernBlot analysis was performed as described below.

Human mesangial cells were isolated from histologicallynormal human kidneys as described by Radeke et a! [28]. Thecells stained positive for Thy 1.1 and smooth muscle cell actin.They were cultured in RPM! (Biochrom, Berlin, Germany) with20% FCS and 1% insulinltransferriniselen cocktail (Biochrom).Passages 3 to 6 were used for experiments and serum starvationwas carried out in RPM! containing 1% FCS for three days.

Determination of cell proliferation

Cell proliferation was estimated by 3H-thymidine uptake.Mesangial cells were seeded in 96-well microplates. At subcon-fluence, mesangial cells were growth-arrested in DMEM con-taining 0.5% FCS for four days. After adding agonists to themesangial cells, DNA synthesis was followed by labeling thecells with 3H-thymidine (2.5 CiIml) for the times indicated. Atthe end of the labeling period, mesangial cells were washedtwice with PBS, trypsinized with 0.05% trypsine and 0.02%EDTA, and harvested onto cellulose filters with an automatedcell harvester (Cambridge Technology, Watertown, Massachu-setts, USA). Finally, the filters were counted in a Beckman rackscintillation counter.

Determination of protein synthesis

Protein synthesis was measured by leucine incorporation.Mesangial cells were seeded in 24-well plates. At subconflu-ence, they were growth-arrested in DMEM containing 0.5%FCS for four days. Cells were then stimulated with the agonistsand leucine incorporation was determined similar as describedby Freshney [29].

Isolation of RNA and Northern blot analysis

Cellular RNA was isolated by the guanidinium isothiocyanatemethod as described [30]. Twenty micrograms RNA wereloaded on a 1% agarose/1 .8% formaldehyde gel, size fraction-ated and transferred to a nylon membrane (Hybond N, Amer-sham, Braunschweig, Germany) with 2OxSSC overnight.Ethidiumbromide staining of the gel and membrane confirmedequal loading and sufficient blotting of the RNA. The RNA wasfixed by baking for two hours at 80°C. The blots were probedwith a 300 bp rat LIF cDNA fragment (bp 5096 to 5306 of the 3.exon of rat LIF gene) purified with Geneclean Kit (Dianova,Hamburg, Germany) from plasmid pLIF-PBS after digestionwith PstI and EcoRI (Baffet and Fey, unpublished data).

For detection of IL-6 mRNA, a full length rat cDNA wasused (pRIL6C.94) [31]. The GAPDH hybridization control wascarried out with a 500 bp reverse transcribed fragment, ampli-fied from rat mRNA using the following PCR primers: senseprimer 5'-AATGCATCCTGCACCACCAA, antisense primer5'-GTCATTGAGAGCAATGCCAGC. cDNA probes for jun-B(mouse) and c-fos (mouse) were obtained from Dr. Bravo(Princeton) and for Egr-l (mouse) from Dr. Sukhatme (Boston).

mRNA sizes were evaluated by comparison with l8S and 28SrRNA. The sizes of the relevant mRNAs were: c-fos 2.3 kb,jun-B 2.1 kb, Egr-1 3.4 kb, GAPDH 1.4 kb, LIF 4.3 kb, and IL-62.4 kb and 1.2 kb.

DNA fragments were labeled with 32P-dCTP using a randompriming kit (Amersham). Prehybridization was performed at42°C for two hours with a buffer containing 5 x SSC, 3 M NaCl,0.3 M Na citrate, 0.1% SDS, SxDenhardt's solution, 50 mMNa2HPO4 (pH 6.5), 50% deionized formamide and 250 g/mldenatured salmon sperm DNA. Hybridization was carried outunder the same conditions for 12 to 24 hours. The membraneswere washed twice with 2xSSC/0.1% SDS at 40°C for 15minutes and twice with 0.2xSSC/0.1% SDS for 30 minutes atthe same temperature. Blots were then exposed to a KodakX-ray film with intensifying screen for varying time periods.

1564 Hartner et a!: Leukemia inhibitory factor

Autoradiographs of Northern blots were quantitated densito-metrically. mRNA signals were corrected for GAPDH mRNAexpression to confirm equal loading of RNA.

RT-PCR

Complementary DNA was synthesized using 0.5 jsg of totalRNA with oligo dT16 as downstream primer according to theprotocol of the Perkin Elmer Cetus GeneAmp RNA PCR Kit,using 1 mM 3-mercaptoethanol in the RT reaction. The cDNAwas then amplified with specific primers for human LIF (senseprimer 5'-GATGAGTCCAAGATAGAGAGC, and antisenseprimer 5'-CCTGACCCTAAGTTCTGC) provided by Dr.Schrell (Erlangen), and control primers for human cyclophilin(sense primer 5'-CATCTGCACTGCCAAGACTG, and an-tisense primer 5'-CTGCAATCCAGCTAGGCATG). PCR reac-tion was carried out according to the Perkin Elmer protocol for30 cycles with one minute at 94°C, one minute at 55°C and threeminutes at 72°C. The PCR products were size fractioned on an1.8% agarose gel and detected by ethidium bromide staining.(LIF: 516 base pair signal, cyclophilin: 326 base pair signal).Performing the RT-PCR without adding reverse transcriptase,there was no detectable signal, confirming that no DNA wasamplified.

Determination of LIF proteinThe concentration of LIF protein in the CHO cell superna-

tants was determined with a commercial ELISA (R&D Sys-tems). The same ELISA was used for the determination of LIFprotein in culture supernatants of human mesangial cells. Su-pernatants were concentrated 20-fold by overnight centrifuga-tion with CentriconlO (Amicon, Witten, Germany).

LIF bioassayBiological activity of LIF was measured using an embryonic

stem (ES) cell assay [321. These cells are known to proliferate inLIF-containing medium. In the absence of LIF, they developinto differentiated cells within two to three days. ES cells (l0cells per well) were grown in gelatin-coated 24-well plates inDMEM with 10% FCS, 1% glutamine, 1% mercaptoethanol and500 U/mI LIF (CHO supernatant). For use in the LIF assay, thecells were washed twice in PBS and cultured in the presence ofthe test samples (various concentrations of mesangial cellsupernatants) for 36 hours.

Results

Detection of LIF mRNA in rat mesangial cells and isolatedglomeruli

Rat mesangial cells were grown to subconfiuence in thepresence of 10% FCS. Resting cells were obtained after fourdays of culture in medium containing only 0.5% FCS. Underboth conditions, there were only trace amounts of LIF mRNAdetectable by Northern blot analysis (Fig. 1). Restimulation ofresting mesangial cells with 10% FCS led to the induction of the4.3 kb LIF mRNA. As a control, buffalo rat liver cells wereused which constitutively expressed small amounts of LIFmRNA [33]. After serum deprivation and restimulation forthree hours, these cells expressed LIF mRNA of the same size(Fig. 1). Stimulation of growing or resting mesangial cells withthe bacterial wall component LPS (10 j.gIm1) led to a strong

FCS - FCS LPS LPS -A R R R G G

Fig. 1. Expression of LIF mRNA in growing and growth-arrested ratmesangial cells. Growing (G) or growth arrested (R) mesangial cellswere incubated with LPS (10 g/ml) or FCS (10%), as indicated. As acontrol, growth arrested buffalo rat liver cells (BRL cells) were incu-bated with 10% FCS. After 3 hours, cells were harvested and RNAprepared as described in Methods. Northern blots were hybridized witha LIF cDNA probe. Equal loading of the gel was confinned by theethidium bromide staining of the 28S ribosomal RNA. The blot shownis representative for 3 similar experiments.

induction of LIF mRNA (Fig. 1). Resting mesangial cellsproved to be particularly sensitive to LPS: concentrations aslow as 2 ng/ml induced LIF mRNA to a degree similar to theinduction by jsg amounts of LPS (Fig. 2a). The cytokine IL-113(5 nglml) was found to be another strong stimulus of LIF mRNAinduction (Figs. 2b, 3). Small amounts of LIF mRNA were alsodetectable in freshly isolated glomeruli. Upon incubation withLPS (10 jsg/ml) for five hours, LIF mRNA expression wasalmost doubled (relative intensity of mRNA signals detected byautoradiography: 0.64 vs. 0.38 with GAPDH as reference; datanot shown).

Transient induction of LIF mRNA and IL-6 mRNAThe time kinetics of the induction of LIF mRNA were

determined in growth-arrested rat mesangial cells with 10%FCS, 5 jsg/ml LPS and 5 ng/ml IL-lf3 as stimuli (Fig. 3). TheFCS-induced mRNA expression was maximal after one to threehours, then declined and was barely detectable after eighthours. The kinetics of the induction by LPS and IL-1/3 werevery similar with a peak after five hours and a clear signal stillvisible after eight hours. After 20 hours, background levelswere obtained in all cases. Interestingly, IL-6 mRNA wasinduced by the same cytokines with the same kinetics as wasLIF mRNA (Fig. 3). Two species of IL-6 mRNA of 1.2 kb and2.4 kb were detected, as described before [34, 35].

Effects of cycloheximide and dexamethasone on LIF and IL-6mRNA induction

Inhibition of protein synthesis by cycloheximide in restingmesangial cells led to a strong induction of LIF and IL-6mRNA, comparable to the one caused by 10% FCS (Fig. 4).The glucocorticoid dexamethasone interfered with the IL-113-induced expression of LIF and IL-6 mRNA. For this type ofexperiment, the cells were preincubated with dexamethasone(l0 M) for two hours prior to the addition of IL-1/3 (Fig. 4).

BRLcells Mesangial cells

LIF .4.3kb

28S -5kb

e

Hartner et a!: Leukemia inhibitory factor 1565

IL-i 3 concentration, ng/mlFig. 2. Concentration-dependent induction of LIF mRNA by LPS andIL -1 13. Resting mesangial cells were incubated with LPS (a) or IL-l/3 (b)for 3 hours, as indicated. Autoradiographs of Northern blots werequantitated densitometrically. The LIF mRNA signal was corrected forGAPDH mRNA expression to confirm equal loading of RNA. The datapresented are representative for 2 independent experiments.

LIF mRNA concentrations were reduced to background valueswhile IL-6 mRNA expression was less sensitive to dexameth-asone.

Induction of LIF mRNA by various cytokinesA number of cytokines were tested for their ability to induce

LIF mRNA expression. Concentrations were chosen, whichhad been shown before to induce cellular responses (such asproliferation or prostanoid synthesis) in the rat mesangial cellsused. Cytokines with very different cellular effects on mesangialcells, namely IL-l/3 (5 ng/ml) or TNFa (100 ng/ml) and platelet-derived growth factor (PDGF, 100 nglml), were found to induceLIF mRNA (Fig. 5). A weak induction was also observed withbasic fibroblast growth factor (bFGF, 10 ng/ml), transforminggrowth factor (TGFI3, 4 ng/ml), endothelin (l0— M) and epider-mal growth factor (EGF, 60 ng/ml). When used in combinationwith IL-l)3, TGFJ3 showed a strong synergistic effect on LIF

inRNA expression (Fig. 6). LIF (200 to 1000 U/mi) was not ableto induce its own message, nor was LIF mRNA induced by IL-6(200 U/mi), or vice versa.

Detection of LIF mRNA in human mesangial cellsThe rat cDNA for LIF did not cross react with mRNA

obtained from human mesangiai cells. Therefore, RT-PCR wasused to show LIF mRNA expression. In unstimulated, growth-arrested human mesangial cells, very little LIF mRNA wasdetectable. Upon incubation with LPS or IL-1J3, there was asignificant increase in LIF mRNA reflected in an increase in theamount of PCR product (Fig. 7).

Induction of LIF protein releaseTo evaluate for secreted LIF activities, the supernatants of

stimulated rat mesangial cells were concentrated and intro-duced into an ES cell bioassay. These cells grow in the presenceof LIF but differentiate when LIF is withdrawn [36]. Themesangial cell supernatants did not support ES cell prolifera-tion, but rather induced a rapid differentiation of the targetcells, most likely due to a combination of different growthfactors which are known to be secreted by activated mesangialcells. Therefore, it was not possible to assess specific LIFbioactivity by this assay.

Since no antibody reacting with rat LIF was available to us,LIF protein was analyzed in the supernatants of human mesan-gial cells only. Within five hours LPS (10 g/ml) induced atwofold increase of LIF protein release: control 2.7 0.6pg/ml,LPS 5.3 1.7 pg/ml (means range of 2 independent isolatesdetermined in duplicates). In two additional supernatants ofother isolates of human mesangial cells, the release of LIF incontrols was at or below the detection limit of the ELISA (0.7pg/mI or less). Upon stimulation with IL-1f3 for 15 hours, large

10 amounts of LIF protein were released (5.0 0.7 pglml; meansrange of 2 independent isolates).

Induction of the expression of immediate early genes by LIFResting rat mesangial cells were incubated with human re-

combinant LIF (2 ng/ml). A transient induction of the immedi-ate early genes c-fos, Egr-l and jun-B was observed after 30 and60 minutes (Fig. 8). Similar responses of mesangial cells wereobtained after stimulation with IL-6 (data not shown). Incontrol experiments, mesangial cells were incubated with PBS/0.1% BSA. Such treatment of the cells did not lead to anyinduction of immediate early genes.

Effects of LIF on mesangial cell proliferationThe mitogenic potential of LIF on mesangial cells was

studied by incubating resting rat mesangial cells with humanrecombinant LIF (50 U/ml, 500 U/mI and 5000 U/mI) for 24, 48,72 and 96 hours. Cell proliferation was measured by 3H-thymidine incorporation for the last 24 hours. As shown inFigure 9, no increase of mesangial cell proliferation was in-duced by LIF at the concentrations tested. In combination withknown mesangial cell mitogens such as PDGF (25 ng/mi) and5-HT (1 SM), LIF caused a slight increase in 3H-thymidineincorporation after 24 hours. The PDGF-induced thymidineincorporation was enhanced 1.32 0.06-fold (means SD, N =5 independent experiments, P < 0.001). The interaction with5-HT was less significant (1.19 0.13, N = 5, P < 0.05). The

A1.6

1.4

1.1.?0.8

0.6a)

0.4

0.2

00.01

B

0.1 1 10 100 1000 10000

LPS concentration, ng/ml

>'(I)C

Ca)>(0a)

8

7

6

5

4

3

2

1

00.05 0.1 1

-1.2kb

Fig. 3. Time course of the induction of LIF and IL-6 mRNAs. Growth-arrested mesangial cells were incubated with FCS (10%), IL-l/3 (5 ng/ml)or LPS (5 gIml) for the times indicated. Northern blots were probed consecutively with LIF and IL-6 cDNAs. Equal loading was confirmed byethidium bromide fluorescence and hybridization with a GAPDH cDNA probe. Data shown are representative for 4 experiments.

GAPDH

Discussion

parallel in many cells and tissues. Only IL-6 has been known tobe expressed and produced by mesangial cells in culture, whileno information has been available about expression of LIF inmesangial cells or other kidney cells. Our results demonstratethat rat and human glomerular mesangial cells in culture ex-press LIF, In addition, LIF mRNA was also detectable in

- 4 3 kbfreshly isolated whole glomeruli. We found very little LIFmRNA expressed in resting mesangial cells and in mesangialcells growing in an unsynchronized fashion. LIF was tran-siently induced by various stimuli, including FCS, the bacterial

- 2.4 kb wall component LPS, inflammatory cytokines such as IL-113 orTNFa, or the mitogen PDGF. These diverse stimuli havepreviously been reported to induce IL-6 in mesangial cells[17—19], which was confirmed in our studies. In other organs,

- 1.2 kb such as liver [37], LIE and IL-6 have been shown to share manybut not all properties. Their receptors associate with the sametransducing protein, gpl3O [38]. In hepatocytes, the sameresponse element in the promoter of the a2-macroglobulin geneis activated by IL-6 and LIF, indicating that both use the samesignal transduction cascade [13]. Diverging functions of LIF- 1.4 kb and IL-6 have also been observed. As discussed below, IL-6has been implicated in glomerular pathology, while a functionalrole of LIF in the kidney remains to be elucidated. At present itis unclear to which extent these cytokines possess overlappingor differential characteristics in the kidney. LIF has beendiscussed to be involved in inflammatory reactions. Urinaryexcretion of LIF was found elevated during acute kidney graftrejection [39], but the origin of LIF production remainedunclear. LIF production was studied in detail in the inflamedjoint, where its synthesis was increased by IL-l/3 and TNFa[17, 40]. The present results demonstrate that these cytokinesare also able to induce LIF in mesangial cells, suggesting a rolefor LIF in glomerular inflammatory reactions, as has beendiscussed for IL-6 [24, 41].

LIF mRNA was barely detectable by Northern blot analysisin growth arrested or proliferating mesangial cells. Induction of

1566 Hartner et at: Leukemia inhibitory factor

FCS LPS IL-1J

LIF -4.3 kb

-2.4 kb

IL-6

xa)a+

(OX0 I C)u o a

2C00

LIF

IL-6

Fig. 4. Effect of cycloheximide and dexamethasone on LIF and IL-6mRNA expression. Growth-arrested mesangial cells were incubatedwith FCS (10%), cycloheximide (CHX, 50 Lg/ml), IL-lf3 (5 nglml) ordexamethasone (10—v 54) for 3 hours; control lane: cells incubated withPBS/0. 1% BSA. Northern blot analysis was performed with probes forLIF, IL-6 and GAPDH.

increase of thymidine incorporation disappeared after 48 hours.LIF by itself had no significant effect on total protein synthesisas determined by leucine incorporation. The PDGF-inducedincrease of leucine uptake was not augmented by LIF.

LIF is a polyfunctional cytokine which shares many charac-teristics with IL-6. Expression of LIF and IL-6 is regulated in

4 S

hr

5hr

8 hr

20 h

r

lhr

3hr

4 8h

r

2Ohr

1 hr

3 hr

5 hr

Bhr

20 h

r

a•*0

0

Hartner et a!: Leukemia inhibitory factor 1567

LIF

GAPDH

U U) (0(.:, 0w -J — -J

Ca)

..- 0U) U- U • D0 ( Z 0U- F- I— 0 w

LL

0

- 4.3 kb

-1.4kb

Fig. 5. Effect of various mesangial cell stimuli on LJF mRNA expression. Growth-arrested mesangial cells were incubated with IL-113 (5 ng/ml),bFGF (10 nglml), EGF (60 ng/ml), LPS (5 sgIml), IL-6 (200 U), LIF (500 U), TGFI3 (4 ng/ml), TNFa (100 ng/ml), endothelin (10—v M) or PDGF(100 ng/ml) for 3 hours; control lane: cells incubated with PBS/0. 1% BSA. Northern blot analysis was performed with probes for LIF and GAPDH.Data shown are representative for 2 independent experiments.

+•:- u.C

I—

(0. +LL (0 U..

2

(

GAPDH -1.4kb

Fig. 6. Synergistic induction of LIF mRNA by IL-1/3 and TGF/3.Growth-arrested mesangial cells were incubated with TGFI3 (5 ng/mI),IL-1/3 (5 ng/ml), or TGFI3 and IL-1/3 together for 3 hours, and withIL-113 for 3 hours after preincubation of 2 hours with TGF3. Controllane: cells incubated with PBS/0.l% BSA, Northern blot analysis wasperformed with probes for LIF and GAPDH.

LIF mRNA in mesangial cells was caused by LPS and a numberof cytokines with very different cellular functions in these cells:PDGF is a strong mitogen while LPS, IL-l$ and TNFa activatemesangial cells to secrete inflammatory mediators [42]. Induc-tion of LIF mRNA by LPS and IL-lf.3 also resulted in secretionof LIF protein, suggesting a biological significance. Another setof bioactive compounds, namely endothelin, EGF, bFGF andTGF/3, were only weak inducers of LIF mRNA expression. It isnot yet clear whether these changes in mRNA levels alsotranslated into significantly increased protein levels. The weakinduction was not due to insufficient concentrations of therespective stimuli, since in control experiments we had shownthat they were able to activate mesangial cells by enhancing therelease of lipid mediators or by induction of proliferation, at theconcentrations used for induction of LIF (data not shown). Astrong effect of TGF/3 on LIF expression was detected only

when mesangial cells were stimulated with the combination ofTGF/3 and IL-113. Though almost ineffective by itself, TGFJ3synergistically enhanced the ILl-p induced expression of LIF.Synergistic actions of cytokines are commonly observed andwere described for LIF induction in synovial fibroblasts [401.LIF mRNA was also induced by cycloheximide, suggesting arole for labile proteins in the regulation of LIF mRNA expres-sion and/or stability.

Neither IL-6 nor LIF itself was able to induce LIF mRNA,arguing against an autocrine induction of LIF expression inmesangial cells. Our data reveal that the inability to induce LIFexpression was not due to missing receptors of IL-6 or LIFsince both cytokines induced the expression of immediate earlygenes. These results on mesangial cells differ from recentobservations in cartilage cells. A strong autoinduction of LIFwas found in neonatal and adult chondrocytes [17]. In addition,LIP also induced IL-6 mRNA expression in these cells as wellas in human monocytes, synoviocytes and various cell lines[43].

Cell-specificity is also indicated by the observation thatcytokine-induced LIP expression varies considerably betweendifferent types of primary cultures: TGFI3 and bFGF are stronginducers of LIP expression in chondrocytes [17], but onlyweakly effective in synovial fibroblasts [40] and mesangial cells.PDGF, a strong mitogen for all three cell types, induces LIFmRNA in mesangial cells and chondrocytes, but not in synovialfibroblasts. The mechanisms responsible for this apparent cellspecificity of LIF expression and its potential biological signif-icance remain to be elucidated.

Once secreted, LIP could either act in an autocrine fashionon mesangial cells or on the neighboring cells, such as glomer-ular endothelial or epithelial cells. So far, no data are availableanalyzing effects of LIP on glomerular endothelial or epithelialcells. However, our results indicate an autocrine potential ofLIP in mesangial cells. LIF had no effect on total proteinsynthesis as determined by leucine incorporation, but specifi-cally induced a set of immediate early genes, c-fos, jun-B andEgr-1. Early response genes have also been shown to beinduced by LIF in sympathetic neurons [44], in Ml myeloidleukemia cells [45] and liver cells [37]. The induction of these

LIF -4.3kb

•$**Sww

WI..'

I- I- •- I.-- = .C .c

eq

GAPDH w • -1.4kb

a a a a - 326 bp

LU'

I-.-C -C0,

516 bp -

c-tos - 2.3 kb

-2.1 kb

-3.4kb

1568 Hartner et a!: Leukemia inhibitory factor

Fig. 7. Detection of LIP mRNA in humanmesangial cells. Growth-arrested humanmesangial cells were incubated with LPS orIL-l/3 for 3 hours. Control cells wereincubated with PBSIO. 1% BSA. PCR wasperformed with LIF and cyclophilin (cyc)primers in parallel. Left and right 4 lanesshow PCR products from two independentexperiments.

Fig. 8. induction of immediate early genes byLIP in rat mesangial cells. Growth-arrestedmesangial cells were incubated with LIF forthe times indicated. Northern blots wereanalyzed for Egr-l, c-fos, jun-B and GAPDHmRNA. The blots are representative for 3independent experiments.

genes could be a very early step in the signal transductionpathway which will then diverge to initiate different cellularresponses depending on the cell type. The biologic responseinduced by LIF in mesangial cells and its physiological orpathophysiological relevance are still unknown. Induction ofimmediate early genes, especially Egr- 1, has been discussed asan early event in the mitogenic signaling cascade in mesangialcells [46]. A correlation was found between the mitogenicactivity of vasoactive compounds and their ability to induceEgr-l expression. In our experiments, neither LIF nor IL-6were essential for mesangial cell proliferation. We were unableto induce proliferation by either stimulus in resting rat mesan-

gial cells during a time period of up to six days. These findingsare in accordance with studies in rabbit aorta smooth musclecells, in which only oncostatin M, but not LIF or IL-6, wasmitogenic [47]. In the present experiments, LIF only showed atransient effect on the PDGF-induced thymidine incorporation,but no significant co-mitogenic effect on cell proliferation.There are conflicting reports on the mitogenic potential of IL-6in mesangial cells [48]. Depending on the experimental design,IL-6 either promoted [24] or inhibited mesangial cell prolifera-tion [49]. There is indirect evidence that IL-6 plays a role inproliferative glomerulonephritis [24]; however, the precise in-volvement of this cytokine in the glomerular inflammatory

Experiment 1 Experiment 2

IL-1f3 Control LPS Control

LJ_ 0' Li.J C)

Li...J 0

Li....J 0

100000

Hartner et a!: Leukemia inhibitory factor 1569

— —. .—

(C1) 000IC) 0 0

24 hr 48 hr 72 hr 96 hr

Fig. 9. Effect of LIF on 3H-thymidine incorporation by rat mesangial cells. Resting mesangial cells were incubated with 50 U/mI, 500 U/mi and5000 U/mI LIF. DNA synthesis was followed by labeling the cells with 3H-thymidine from 0 to 24 hours, 24 to 48 hours, 48 to 72 hours and 72 to96 hours. Control cells were incubated with PBS. Data shown are means SD of 6 parallel wells and are representative for 3 experiments.

response is presently unclear. On the other hand, the LIF-induced expression of immediate early genes in mesangial cellsmay initiate biologic responses other than proliferation in thesecells, such as activation of synthesis and secretion of proteins.Recently, Richards et al [50] described a selective up-regulationof metalloproteinase inhibitor (TIMP-l) by LIF, IL-6 andespecially oncostatin M in fibroblasts. Similar effects of LIF onthe turnover of matrix proteins might also be relevant inmesangial cells.

Infiltrating monocytes or other bone marrow-derived cells arealso potential targets of LIF in the glomerulus. LIF has beenshown to mediate chemoattraction of eosinophils [51] and toinduce gene expression and protein synthesis of macrophagechemoattractant protein 1 (MCP-l) in chondrocytes [52]. Pre-liminary data from our laboratory suggest similar effects of LIFin mesangial cells (Hartner et al, unpublished observations).Thus, LIF may be involved in the recruitment of immunecompetent cells to the glomerulus once it has been induced byinflammatory stimuli, such as TNFa and IL-1/3, or by theexposure to bacterial products.

Acknowledgments

This work was supported by the Deutsche Forschungsgemeinschaft,SFB 263/B5, HI 291/5-2 and Hi 291/5-6. The authors are grateful to Dr.H. H. Radeke, Hannover, Germany, for providing additional superna-

tants of human mesangial cells, and to Dr. U. M. Schrell, Erlangen,Germany, for providing LIF primers for RT-PCR.

Reprint requests to M. Goppelt-Struebe, Medizinische K!inik IV,Universitaet Erlangen-Nuernberg, Loschgestrasse 8 1/2, 91054 Er-langen, Germany.

References

1. HILTON DJ: LIF: Lots of interesting functions. TIBS 17:72—77,1992

2. HILTON DJ, NICOLA NA, METCALF D: Distribution and bindingproperties of receptors for leukemia inhibitory factor, in Polyfunc-tional Cytokines: IL-6 and LIF, edited by CIBA-FOUNDATION,Chichester, John Wiley & Sons, 1992, pp 227—239

3. GODARD A, HEYMANN D, RAHER S, ANEGON I, PEYRAT M, LEMAUFF B, Moui&.y E, GREGOIRE M, VIRDEE K, SOULILLOU J,MOREAU F, JACQUES Y: High and low affinity receptors for humaninterleukin for DA cells/leukemia inhibitory factor on human cells.J Biol Chem 267:3214—3222, 1992

4. HILTON DJ, NICOLA NA, GOUGH NM, METCALF D: Resolutionand purification of three distinct factors produced by Krebs ascitescells which have differentiation-inducing activity on murine my-eloid leukemic cell lines. J Biol Chem 263:9238—92443, 1988

5. Wiusot. TA, METCALF D, GOUGH NM: Cross-species comparisonof the sequence of the leukemia inhibitory factor gene and itsprotein. Eur J Biochem 204:21—30, 1992

6. HILTON DJ, NIc0LA NA, METCALF D: Distribution and compari-son of receptors for leukemia inhibitory factor on murine hemopoi-etic and hepatic cells. J Cell Physiol 146:207—215, 1991

E

C000aCC

E>,

C,,

90000

80000

70000

60000

50000

40000

30000

20000

10000

0

II

Hflfl1 fl nnnl fl fln,flflH, 1n,

.1::; 0 —. - -CCO 0 0 0o U) 0 0u U) 0

—-J

— 0 - —. --.

a 0 0 0U) 0 0U IL U) 0— u-J-J

— 0 — •— •—C

U) 0 0 00 U) 0 0IL U) 0— u-J-J

1570 Hartner et a!: Leukemia inhibitory factor

7. HOCKE G, BAFFET G, Cul M, BRECHNER T, BARRY D, GOEL A,FLETCHER R, ABNEY C, HATTORI M, FEY GH: Transcriptionalcontrol of liver acute phase genes by interleukin-6 and leukemiainhibitory factor, in 42. Colloquium Mosbach: Molecular Aspectsof Inflammation 1991, pp 147—166

8. BAUMANN H, WONG G: Hepatocyte stimulating factor III sharesstructural and functional identity with leukemia inhibitory factor. JImmunol 143:1163—1167, 1989

9. ALLAN EH, HILTON DJ, BROWN MA, EVELY RS, YUMITA S,METCALF D, GOUGH NM, NG KW, NICOLA NA, MARTIN TJ:Osteoblasts display receptors for and response to leukemia inhibi-tory factor. J Cell Physiol 145:110—119, 1990

10. Mom M, YAMAGUCHI K, ABE K: Purification of a lipoproteinlipase-inhibiting protein produced by a melanoma cell line associ-ated with cancer cachexia. Biochem Biophys Res Commun 160:1085—1092, 1989

11. REID IR, LOWE C, CORNISH J, SKINNER JiM, HILTON DJ, WILL-soN TA, GEARING DP, MARTIN TS: Leukemia Inhibitory Factor: Anovel bone-active cytokine. Endocrinology 126:1416—1420, 1990

12. METCALF D, GEARING DP: Fatal syndrome in mice engrafted withcells producing high levels of the leukemia inhibitory factor. ProcNatl Acad Sci USA 86:5948—5952, 1989

13. HOCKE GM, Cul MZ, RIPPERGER JA, FEY GH: Regulation of therat a2-macroglobulin gene by interleukin 6 and leukemia inhibitoryfactor, in Acute Phase Proteins: Molecular Biology, Biochemistry,Clinical Applications, edited by KUSHNER I, BAUMANN H, MACK-IEWICZ A, CRC Press (in press)

14. GOUGH NM, WILLSON TA, STAHL J, BROWN MA: Molecularbiology of the leukemia inhibitory factor gene, in PolyfunctionalCytokines: IL-6 and LIF, edited by CIBA-FOUNDATION, Chich-ester, John Wiley & Sons, 1992, pp 24—34

15. LUBBERT M, MANTOVANI L, LINDEMANN A, MERTELSMANN R,HERRMANN F: Expression of leukemia inhibitory factor is regulatedin human mesenchymal cells. Leukemia 5:361—365, 1991

16. ANEGON I, MOREAU J-F, GODARD A, JACQUES Y, PEYRAT M-A,HALLET M-M, WONG G, SOULILLOU JP: Production of humaninterleukin for DA cells (HILDA)/leukemia inhibitory factor (LIF)by activated monocytes. Cell Immunol 130:50—65, 1990

17. LOTZ M, MOATS T, VILLIGER PM: Leukemia inhibitory factor isexpressed in cartilage and synovium and can contribute to thepathogenesis of arthritis. J Clin Invest 90:888—896, 1992

18. ABBOTT F, RYAN JJ, CESKA M, MATSUSHIMA K, SARRAF CE, REESAJ: Interleukin-l/3 stimulates human mesangial cells to synthesizeand release Interleukins-6 and -8. Kidney mt 40:597—605, 1991

19. COLEMAN DL, RUEF C: Interleukin-6: An autocrine regulator ofmesangial cell growth. Kidney Int 41:604—606, 1992

20. ZOJA C, WANG JM, BETrONI S, SIRONI M, RENZI D, CHIAFFARINOF, ABBOUD HE, VAN DAMME J, MANTOVANI A, REMUZZI G,RAMBALDI A: Interleukin-l/3 and tumor necrosis factor-a inducegene expression and production of leukocyte chemotactic factors,colony-stimulating factors and interleukin-6 in human mesangialcells. Am J Pathol 138:991—1003, 1991

21. IWANO M, D0HI K, HIRATA E, H0RII Y, SHIIKI H, ISHIKAWA H:Induction of interleukin-6 synthesis in mouse glomeruli and cul-tured mesangial cells. Nephron 62:58—65, 1992

22. PATTERSON PH, FANN Mi: Further studies of the distribution ofCDF/LIF mRNA, in Polyfunctional Cytokines: IL-6 and LIF,edited by CIBA-FOUNDATION, Chichester, John Wiley & Sons,1992, pp 125—135

23. TOMIDA M, YAMAMOTO-YAMAGUCHI Y, HozuMI M, HOLMES W,LOWE DG, GOEDDEL DV: Inhibition of development of Na-dependent hexose transport in renal epithelial LLC-PK1 cells bydifferentiation-stimulating factor for myeloid leukemic cells/leuke-mia inhibitory factor. FEBS Lett 268:261—264, 1990

24. H0RII Y, MURAGUCHI A, IwANo M, MATSUDA 1, HIRAYAMA T,YAMADA H, FuJII Y, D0HI K, ISHIKAWA H, OHMOTO Y,YOSHIZAKI K, HIRANO T, KIsHIMOro T: Involvement of IL-6 inmesangial proliferative glomerulonephritis. J Immunol 143:3949—3955, 1998

25. FUKATSU A, MATSUO 5, TAMAI H, SAKAMOTO N, MATSUDA T,HIRANO T: Distribution of Interleukin-6 in normal and diseasedhuman kidney. Lab Invest 65:61—66, 1991

26. COOPER HM, TAMURA RN, QUARANTA V: The major lamininreceptor of mouse embryonic stem cells is a novel isoform of thea6b1 integrin. J Cell Biol 115:843—845, 1991

27. SCHULZE-LOHOFF E, ZANNER S. OGILvIE A, STERZEL RB: Extra-cellular ATP stimulates proliferation of cultured mesangial cells viaP2-purinergic receptors. Am J Physiol 263:374—383, 1992

28. RADEKE HH, MEIER B, TOPLEY N, FL0EGE J, HABERMEHL GG,RESCH K: Interleukin 1 a and tumor necrosis factor a induce oxygenradical production in mesangial cells. Kidney mt 37:767—775, 1990

29. FRESHNEY RI: Culture of Animal Cells: A Manual of Basic Tech-nique, New York, Alan R. Liss, Inc., 1987, pp 235—236

30. CHOMCZYNSKI P, SACCHI N: Single-step method of RNA isolationby acid guanidium thiocyanate-phenol-chloroform extraction. AnalBiochem 162:156—159, 1987

31. NORTHEMANN W, BRACIAK T, HATrORI M, LEE F, FEY GH:Structure of the rat interleukin 6 gene and its expression inmacrophage-derived cells. J Biol Chem 264:16072—16082, 1989

32. SMITH AG: Culture and differentiation of embryonic stem cells. JTiss Cult Meth 13:89—94, 1991

33. SMITH AG, HOOPER ML: Buffalo rat liver cells produce a diffusibleactivity which inhibits the differentiation of murine embryonalcarcinoma and embryonic stem cells. Dev Biol 121(1): 1—9, 1987

34. RUEF C, BUDDE K, LACY J, NORTHEMANN W, BAUMANN M,STERZEL RB, COLEMAN D: Interleukin-6 is an autocrine growthfactor for mesangial cells. Kidney mt 38:249—257, 1990

35. NORTHEMANN W, HATTORI M, BAFFET 0, BRACIAK TA,FLETCHER RG, ABRAHAM U, GAULDIE J, BAUMANN M, FEY GH:Production of interleukin-6 by hepatoma cells. Mol Biol Med7:273—285, 1990

36. WILLIAMS RL, HILTON DJ, PEASE S, WILLSON TA, STEWART CL,GEARING DP, WAGNER EF, METCALF D, NICOLA NA, GOUGHNM: Myeloid leukemia inhibitory factor maintains the developmen-tal potential of embryonic stem cells. Nature 336:684-687, 1988

37. BAUMANN H, MARINKOVIC-PAJOVIC 5, WON K-A, JONES VE,CAMPOS SP, JAHREIS GP, MORELLA KK: The action of interleu-kin-6 and leukemia inhibitory factor on liver cells, in PolyfunctionalCytokines: IL-6 and LIF, edited by CIBA-FOUNDATION, Chich-ester, John Wiley & Sons, 1992, pp 100—115

38. I NY, NYE SH, BOULTON TG, DAVIS S, TAGA T, LI Y, BIRRENSi, YAsUKAWA K, KI5HIM0'ro T, ANDERSON DJ, STAHL N,YANCOPOULOS GD: CNTF and LIF act on neuronal cells via sharedsignaling pathways that involve the IL-6 signal transducing receptorcomponent gpl3O. Cell 69:1121—1132, 1992

39. TAUPIN JU, MOREL D, MOREAU iF, GUALDE N, POTAUX L,BEZIAN JH: HILDA/LIF urinary excretion during acute kidneyrejection. Transplantation 53:655—658, 1992

40. HAMILTON JA, WARING PM, FIL0NzI EL: Induction of leukemiainhibitory factor in human synovial fibroblasts by IL-6 and tumornecrosis factor-a. J Immunol 150:1496—1502, 1993

41. OHTA K, TAKANO N, SENO A, YACHIE A, MIYAWAKI T,YOKOYAMA H, TOMOSUGI N, KATO E, TANIGUCHI N: Detectionand clinical usefulness of urinary Interleukin-6 in the disease of thekidney and the urinary tract. Clin Nephrol 38:185—189, 1992

42. RADEKE HH, RESCH K: The inflammatory function of renal gb-merular mesangial cells and their interaction with the cellularimmune system. Clin Invest 70:825—842, 1992

43. VILLIGER PM, GENG Y, LOTZ M: Induction of cytokine expressionby leukemia inhibitory factor. J Clin Invest 91:1575—1581, 1993

44. YAMAMORI T: CDF/LIF selectively increases c-fos and jun-Btranscripts in sympathetic neurons. NeuroReport 2:173—176, 1991

45. LORD KA, ABDOLLAHI A, THOMAS SM, DEMARCO M, BRUGGE JS,HOFFMANN-LIEBERMANN B, LIEBERMANN DA: Leukemia inhibi-tory factor and interleukin-6 trigger the same immediate earlyresponse, including tyrosine phosphorylation, upon induction ofmyeloid leukemia differentiation. Mol Cell Biol 11:4371—4379, 1991

46. RUPPRECHT HD, DANN P, SUKHATME VP, STERZEL RB, COLEMANDL: Effect of vasoactive agents on induction of Egr-l in ratmesangial cells: Correlation with mitogenicity. Am J Physiol 263:F623—F636, 1992

47. GROVE RI, EBERHARDT C, ABID S, MAZZUCCO C, Liu J, KIENER

Hartner et a!: Leukemia inhibitory factor 1571

P, TODARO 0, SHOYAB M: Oncostatin M is a mitogen for rabbitvascular smooth muscle cells. Proc NatlAcad Sci USA 90:823—827,1993

48. FLOEGE J, TOPLEY N, HOPPE J, BARRETT TB, RESCH K: Mitogeniceffect of platelet-derived growth factor in human glomerularmesangial cells: Modulation and/or suppression by inflammatorycytokines. Cliii Exp immunol 86:334-341, 1991

49. IKEDA M, IKEDA U, OHARA T, KUSANO E, KANO S: RecombinantInterleukin-6 inhibits the growth of rat mesangial cells in culture.Am J Patho! 141:327—334, 1992

50. RICHARDS CD, SHOYAB M, BROWN TJ, GAULDIE J: Selectiveregulation of metalloproteinase inhibitor (TIMP-1) by oncostatin Min fibroblasts in culture. J immuno! 150:5596—5603, 1993

51. MOREAU J-F, BONNEVILLE M, GODARD A, GASCAN H, GRUART V,MOORE MA, SOULILLOU JP: Characterization of a factor producedby human T cell clones exhibiting eosinophil-activating and burst-promoting activities. J Immuno! 138:3844—3849, 1987

52. VILLIGER PM, TERKELTAUB R, LOTZ M: Monocyte chemoattrac-tant protein-i (MCP-l) expression in human articular cartilage. JClin invest 90:488—496, 1992