Kidney International, Vol. 50 (1996), PP. 1922—1927

Leukemia inhibitory factor ameliorates experimental anti-GBMAb glomerulonephritis

WINs0N W. TANG, MEIYING QI, GWYNETH Y. VAN, GRACE P. WARINER, and BABRU SAMAL

Departments of Pathology, Pharmacokinetics, and Computational Biology, Amgen Inc., Thousand Oakc, California, USA

Leukemia inhibitory factor ameliorates experimental anti-GBM Abglomerulonephritis. Leukemia inhibitoiy factor (LIF) is a pleiotropiccytokine that has been identified in acute and chronic inflammatoryconditions such as rheumatoid arthritis, sepsis, and renal allograft rejec-tion. We investigated the glomerular expression of LIF at 30 minutes, and3, 6, 9, 15 and 24 hours after administration of anti-GBM Ab (N = 3) bythe RNase protection assay. Control rats received rabbit sera and weresacrificed at 30 minutes, and 6 and 24 hours. LIF mRNA relative toGAPDH mRNA was detected at low levels within the glomeruli ofoccasional control rats. However, with the induction of anti-GBM Ab GN,there was a marked increase in LIF steady-state mRNA beginning at threehours which persisted through 24 hours. LIF mRNA was also detected incultured mesangial cells stimulated with TL-1J3, identifying this cell type asa potential glomerular source for this cytokine. To investigate the in vivoeffect of LIF, Lewis rats were continuously infused with recombinant (r)human (h) LIF (--0.5 ng/hr) or saline vehicle i.p. with ALZA osmoticpumps beginning at t = —24 hours (N = 8). All rats were injected withanti-GBM Ab intravenously at t = 0 (N = 16). LIF infusion decreased24-hour urinary protein excretion by 85% (17 15 vs. 114 37 mg/day,P = 0.0001) and was associated with a 60% decrease in glomerularmacrophage infiltration (0.8 0.2 vs. 2.0 0.6 ED-I cells/glom, P =0.0001). The administration of rhLIF did not affect the binding of theanti-GBM Ab to glomeruli. The beneficial effects of LIF were associatedwith a decrease in glomerular MCP-I (56%), IL-i (41%) and TNF (17%)steady state mRNA expression. The latter was associated with a 29%decrease in TNF-a protein expression within the glomerular lysate ofnephritic rats administered LIF when compared with control rats. Thesedata demonstrate a potential role for LIF in the therapy of anti-GBM AbGN.

ytokines are polypeptides that are important mediators ofinflammation by regulating the proliferation, differentiation, mo-bility and function of effector cells. Within the past decade, animportant role has been defined for many of these cytokines in thepathogenesis of inflammation [reviewed in 1—51. Through a com-plicated cascade which often feeds back onto itself, interleukin 1(IL-i) and tumor necrosis factor (TNF) induce the expression ofchemokines such as IL-8 and MCP-I [4], cell adhesion molecules[5], procoagulant activity [6] and collagenase secretion [7], andaugment leukocyte phagocytosis and cytotoxic activity [8]. More

Received for publication May 2, 1996and in revised form July 9, 1996Accepted for publication July 11, 1996

© 1996 by the International Society of Nephrology

recently, a number of cytokines with anti-inflammatory propertiesincluding the IL-i receptor antagonist (ra), IL-4, IL-6 and IL-iDhave been identified. For example, the latter three cytokinesinduce IL-Ira [9, 10], which in turn inhibits IL-I activity by bindingto its receptor [II], while IL-4 down-regulates prostaglandin, IL-I,TNF and interferon-y production [12, 13]. Similarly, IL-lU inhibitsIL-i, IL-8, TNF, and nitric oxide production by neutrophils(PMNs) and/or macrophages [14, 15].

Cytokines are believed to be important in the pathogenesis ofantiglomerular basement membrane (GBM) antibody (Ab) gb-merulonephritis (GN). The importance of pro-inflammatory cy-tokines in this disease is now well established. Therefore, neutral-ization of IL-i and TNF activities has been shown to decreasegbomerular neutrophil (PMN) and macrophage infiltration as wellas to attenuate proteinuria [16, 17]. More recently, there are datato suggest that anti-inflammatory cytokines may also be involved.For example, IL-Ira is found in the sera of rats with a-GBM AbUN, suggesting that mechanism(s) are activated to counteractglomerular inflammation while the pharmacologic administrationof IL-Ira has been reported to be beneficial [16]. Similarly,administration of IL-4 and IL-6 to rats with anti-GBM Ab GN wasreported to cause a 60 to 80% reduction in glomerular thrombiformation and urinary protein excretion [18].

Leukemia inhibitory factor (LIF) is a member of an unusualfamily of proteins known as the neuropoietic cytokines withpleiotropic properties which affects numerous cell types [19].Together with ciliary neurotrophic factor, oncostatin M, interleu-kins (IL) 6 and II, these cytokines bind to a unique a-receptorsubunit that associates with a common gpl3O receptor subunitthat initiates signal transduction [20]. The role of LIF in thepathogenesis of inflammation remains uncertain since both ben-eficial and detrimental effects have been ascribed to this cytokine.LIF has been demonstrated to decrease pulmonary PMN infiltra-tion following intratracheal LPS injection [21]. In addition, serumconcentrations of LIF are elevated in patients with septic shockand correlate with disease activity [22], while pretreatment withLIF protects against experimental E. coli induced sepsis [23, 24].However, this would appear to be inconsistent with another reportthat showed passive immunization of mice against LIF preventslethal endotoxemia [25]. Therefore, the present study was per-formed to determine the role of LIF in the pathogenesis ofanti-GBM Ab GN.

1922

Tang et al: LIF ameliorates a-GBMAb GN 1923

Methods

In Vitro studies

Rat mesangial cells were a generous gift from Dr. HannaAbboud (UTHSC, San Antonio, TX, USA) and had been previ-ously characterized in his laboratory. The mesangial cells wereplated onto 100 X 20 mm tissue culture plates in RPM! 1640 with10% fetal calf serum (FCS). When the cells had grown to 80%confluency, the cells were starved for 15 to 16 hours in serum freeRPMI 1640. IL-1/3 (1 ig/ml) or media (without FCS) were addedto the plates at time zero and the cells were harvested at 1, 4, 8and 24 hours. The preparation of total RNA from the ratmesangial cells is described below.

Experimental model

Normal rabbit serum (NRS) was obtained by bleeding naiveunimmunized rabbits while a-GBM Ab was prepared by immu-nizing New Zealand white rabbits with rat GBM as previouslydescribed [16]. Both the a-GBM Ab serum and NRS weredecomplemented and absorbed with Lewis rat erythrocytes.Twenty-seven Lewis rats (Charles River Lab, Wilmington, MA,USA) were injected with 0.5 ml of either a-GBM Ab (N = 18) orNRS (N = 9) intravenously. Three rats with a-GBM Ab GN wereeuthanized at the following time points: 30 minutes, and 3, 6, 9, 15and 24 hours. Control rats were euthanized at 30 minutes, 6 and24 hours.

The role of LIF in a-GBM Ab GN was investigated byadministering LIF to Lewis rats prior to the induction of nephritis.A different a-GBM Ab was used for these experiments than hadbeen used in the preceeding time course study. The preparationand characterization of recombinant human LIF (rhLIF) from asynthetic gene has been described previously [26]. The ability ofrhLIF to exert a biologic effect in rat tissues has been previouslydemonstrated [26]. Alzet microosmotic pumps (model 1003D;Aiza Corp., Palo Alto, CA, USA) loaded with rhLIF (N = 8) orsaline vehicle (N = 8) were implanted into the peritoneum ofLewis rats under isoflurane anesthesia at —24 hours. The 24-hourtime point was chosen so that the rats had sufficient time in whichto recover from the stress of surgery and anesthesia. The pumpswere calibrated to deliver 0.98 0.02 rd/hr which corresponds to—0.52 ng LIF/hr. All rats were administered a-GBM Ab (N 16)at t = 0 and a 24 hour urine was collected for determination ofurinary protein excretion. Proteinuria was quantitated by thesulfosalicylic acid method with rat albumin (Sigma Chemical Co.,St. Louis, MO, USA) serving as the standard. The rats weresacrificed at 24 hours and renal tissue was prepared for histologicevaluation. In addition, glomeruli were isolated by sieving andprepared for protein (N 6) or RNA (N = 10) analysis.

Morphologic studies

Renal tissue was fixed in 10% formalin for light microscopy andimmunohistochemistry. Histologic sections were stained with theperiodic acid-Schiff reaction and hematoxylin counterstain. Im-munohistochemistry was performed on deparaffinized sectionsthat were incubated with a murine monoclonal Ab to rat mono-cyte/macrophage, ED-i (Harlan Bioproducts for Science, Inc.,Indianapolis, IN, USA). Clone ED-i was detected using horse-radish peroxidase ABC reagents from BioTek (Santa Barbara,CA, USA) with DAB as the chromagen. Glomerular monocyte/macrophage infiltration was determined by counting the number

of cells expressing ED-i in 25 consecutive glomeruli with similarcross section diameters. Data are presented as the mean sum ofvariances.

Quantitation of steady state mRNA

Riboprobes for rat IL-i/3 [129 nucleotides (n)], TNF-a (184 n)and LIF (139 n) were prepared by RT-PCR from rat spleen orendotoxin stimulated macrophage RNA. Primers (24-mers) basedupon the rat LIF [27] and TNF-a cDNAs [28] and degenerateprimers based upon homologous regions within the cDNAs ofhuman and murine IL-i /3 [29, 30] were synthesized on an AppliedBiosystems automated DNA synthesizer. The sequences of the 5'and 3'-primers were:

LIF 5 '-primer: 5 -GCGCCATACGCCACCCGTGTCACG-3'3 '-primer: 5 '-TTATCCACGTflI'G'TTGGGAAATGGT-3'IL-i/3 5 '-primer: 5 '-AAAATACCTGTGGCcTFGGGCCTC-3'3 '-primer: 5' -(TC)TTTICCATCTflI'CYrCTTFGGGTA-3'TNF-a 5-primer: 5'-GTGGTGGTACCAGCAGATGGGGFG-3'3 '-primer: 5 '-TCAGCTCCCTCAGGGGTGTCCTFA-3'After reverse transcription, PCR was performed for 40 cycles

under the following conditions: denaturing at 94°C, annealing at54°C, and elongating at 72°C. The RT-PCR product was purifiedon agarose, ligated into pGEMT (Promega, Madison, WI, USA)and then sequenced with a Taq DyeDeoxy Terminator cyclesequencer kit (Applied Biosystems, Inc.) on an automated DNAsequenator (Applied Biosystems 373A). The sequences of theRT-PCR generated LIF and TNF-a were identical to the pub-lished rat cDNAs while IL-ip was 97% and 100% homologous tothe corresponding murine nucleic acid and deduced amino acidsequences, respectively [30].

The details of the PCR cloning of the rat monocyte chemoat-tractant protein-i (MCP-1) and GAPDH riboprobes and of theRNase protection assay have been described previously [31].Briefly, glomeruli (> 90% purity) from the kidneys of ratsadministered either a-GBM Ab (N = 18) or NRS (N = 9) in thetime course study and from rats in the subsequent LIF adminis-tration study (N = 10) were prepared by sieving. Glomerular andmesangial cell total RNA was isolated by acid phenol/chloroformextraction and frozen at —70°C until assayed. Anti-sense ribo-probes for IL-1/3, TNF-a, MCP-1, LIF and GAPDH were pre-pared by in vitro transcription with SP6 or T7 RNA polymerases(Ambion, Austin, TX, USA) and the incorporation of cx32P-UTP(Amersham Corp., Arlington Heights, IL, USA) at 37°C. Theriboprobes were purified on 6% polyacrylamide gel and theneluted with 0.5 M NH4OAc. Glomerular RNA (2 jsg) from thetime course study (18 anti-GBM Ab GN and 9 control rats) washybridized with I x i0 counts of 32P-UTP-labeled LIF andGAPDH riboprobes for 14 to 16 hours at 56°C while 2 rg of RNAfrom the LIF administration study (N = 6) was hybridized withIL-ij3, TNF-a, MCP-i and GAPDH under the same conditions.The unhybridized RNA was digested with RNase A and Ti(Ambion) at 30°C for one hour. The RNases were then digestedwith proteinase K (Boehringer Mannheim) at 37°C for 30 min-utes. After phenol/chloroform extraction and sodium acetate/ethanol precipitation, the samples were eleetrophoresed on 6%polyacrylamide gel. The radioactivity of the protected 32P-UTPlabeled riboprobes were quantitated with a Phosphorimager (Mo-lecular Dynamics, Sunnyvale, CA, USA).

eqe ee,q1924 Tang et al: LIF ameliorates a-GBM Ab GN

LI F

Quantitation of a-GBM Ab bindingTo insure that rhLIF did not interfere with the binding of the

a-GBM Ab, a-GBM Ab was eluted from the glomeruli ofnephritic rats that had been administered rhLIF (N = 3) or saline(N = 3) by sonification and quantitated with an ELISA. A 96-wellETA/RIA flat bottom plate (Costar Corp., Cambridge, MA, USA)was coated overnight with the glomerular lysate at 4°C. Afterblocking with 1% bovine serum albumin/5% sucrose in TRIS!EDTA, the plate was incubated with horseradish peroxidaseconjugated swine is-rabbit IgG (DAKO Inc., Carpinteria, CA,USA) at 25°C for one hour, washed with 0.02% Tween 20, andthen developed with hydrogen peroxide and 3,3',5,5'-tetramethyl-benzidine for five minutes. The reaction was quenched withphosphoric acid and the optical density at 450 nm was determinedon a Thermomax microplate reader (Molecular Devices, MenloPark, CA, USA). A reference curve was generated from knownconcentrations of afinity purified rabbit IgG (Jackson Immuno-Research, West Grove, PA, USA) which was assayed in parallel.Samples were assayed in triplicate and data are presented relativeto the protein concentration which was determined by a commer-cial colorimetric assay (Sigma).

Determination of TNF-cs

Glomerular lysate from rats with anti-GBM Ab GN that hadbeen administered either LIF (N = 3) or saline vehicle (N = 3)were assayed for TNF-a with a sandwich ELISA (BioSource Int'l,Camarillo, CA, USA) using rabbit and hamster polyclonal Abs torat TNF-a. Samples were assayed in triplicate and normalized forthe protein content present within each sample as describedabove.

Statfvtical methods

Data are presented as the mean SD or mean sum ofvariances. Statistical analyses were performed by comparing indi-vidual group means using the unpaired Student's t-test. A P < 0.05was considered significant.

Results

In vitro studies

Rat mesangial cells cultured in 0% or 10% FCS did not expressLIF mRNA at the time points studied. By comparison, there wasa rapid induction of LIF mRNA relative to GAPDH mRNA onehour after the addition of IL-lp (Fig. 1). The steady stateexpression of LIF mRNA relative to GAPDH mRNA remainedessentially unchanged through 24 hours.

Expression of LIE in a-GBM Ab GN

The glomerular pathology of the 27 rats in the time course study(0.5 to 24 hr) has been described in a previous communication[31]. Briefly, there was a diffuse infiltrative GN composed ofprimarily PMNs after induction of anti-GBM Ab GN that peakedat three hours and was replaced by a diffuse monocyte/macro-phage infiltration that peaked at 24 hours. There was a lowconstitutive expression of LIF mRNA within the glomeruli ofoccasional control rats that had been administered NRS at thethree time points examined. Similarly, rats with anti-GBM Ab GNat the 30-minute time point had minimal to no expression of LIFmRNA (Fig. 2). An increase in LIF mRNA was detected withinthe glomeruli of rats with anti-GBM Ab GN at three hours. Thesteady state expression of LIF mRNA relative to GAPDH mRNApeaked at three to six hours and declined thereafter, but was stillpresent albeit at lower levels by 24 hours (Fig. 3). Unfortunately,attempts to identify the LIF protein within either the mesangialcell conditioned media or glomerular lysate with monoclonal andpolyclonal Abs which recognizes human LIF were unsuccessful.

Administration of LIF ameliorates cs-GBM Ab GN

The role of LIF in the pathogenesis of anti-GBM Ab GN wasassessed by the continuous infusion of rhLIF. There was a 60%decline in glomerular macrophage infiltration (0.8 0.2 ED-Icells/glom) in rats infused with rhLIF when compared with ratsthat had received the saline vehicle (2.0 0.6 ED-i cells/glom,P = 0.0001). This was accompanied by a 85% decrease in 24-hoururinary protein excretion (17 15 vs. 114 37 mg/day, P =0.0001). The administration of rhLIF did not affect the binding ofthe s-GBM Ab (186.3 18.7 ng rabbit IgG/mg glom lysate) whencompared with rats that had received the saline vehicle (180.58.7 ng rabbit IgG/mg glom lysate).

To determine if the administration of rhLlF affected cytokineexpression, glomerular RNA was assessed for the presence ofIL-lp and TNF-a mRNA. The administration of rhLIF prior tothe induction of nephritis decreased the steady state expression ofIL-1f3 and TNF-a mRNAs (relative to GAPDH mRNA) whencompared with rats that had received saline. There was a 41%reduction in IL-1f3 steady state mRNA expression from 0.170.04 to 0.10 0.02 with rhLIF treatment (P = 0.009). Similarly,there was a 17% decline in TNF-cs steady state mRNA expression(0.10 0.01 vs. 0.12 0.01, P = 0.04). The decline in TNF-amRNA expression was confirmed at the protein level with anELISA. There was a 29% reduction in TNF-a protein expressionwithin the glomerular lysate of rats administered LIF (1717 264pg/mg glom lysate) relative to control rats (2415 154 pg/mg

GAPDH

o hr 0 hr 1 hr 1 hr 4 hr 4 hr 8 hr 8 hr 24 hr 24 hr

0% 1 0% (—) (+) (—) (+) (—) (+) (—) (+)FCS FCS

Fig. 1. Autoradiography of a RNase protectionassay demonstating increased mesangial cellexpression of LIF mRNA relative to GAPDHmRNA 1, 4, 8 and 24 hours after the addition ofIL-1/3 (+). In contrast, neither cells cultured inserum free media (0% FCS), 10% FCS, orvehicle (—) expressed LIF mRNA.

LI F

GAPDH

(+) (+) (+) (—) (—) (—) (+) (+) (+) (+) (+) (+) (—) (—)

30mm Shr Ghr

LIF —

GAPDH

6

(+) (÷)

9 hr

(+) (-1-) (i-) (+) (+) (—) (—) (—)

l5hr I 24hr

Tang et al: LIF ameliorates a-GBM Ab GN 1925

Fig. 2. Autoradiography of a RNase protection assay demonstrating increased LIF steady state mRNA expression relative to GAPDH mRNA 0.5, 3, 6, 9, 15and 24 hours after induction of a-GBM Ab GN (+). Occasional control rats that had received NRS (—) were sacrificed at 0.5, 6 and 24 hours andexpressed low levels of LIF mRNA.

0.00

Time, hours

Fig. 3. A temporal profile of LIF steady state mRNA expression relative toGAPDH mRNA in anti-GBMAb GN was determined by quantitation of theRNase protection assay with a Phosphorimager. Occasional control ratsexpressed LIF mRNA which was below the limit of quantitation while LIFmRNA was up-regulated in rats with anti-GBM Ab GN beginning threehours after induction of GN. Each point represents the mean SD of 3rats that were assayed individually.

glom lysate, P < 0.02). These changes in cytokinc expression wereassociated with a 56% decrease in the glomerular steady statemRNA expression of the C-C chemokinc, MCP-1, in rats treatedwith LIF (0.04 0.02 vs. 0.09 0.04, P = 0.02).

Discussion

These experiments demonstrate a potential role for LIF in thepathogenesis of heterologous phase anti-GBM Ab GN. There was

an increase in the steady state expression of LIF mRNA relativeto GAPDH mRNA within the glomeruli of nephritic rats threehours after induction of anti-GBM Ab GN. Although we wereunable to confirm the presence of the LIF protein with Abs thatrecognizes human LIF, it is likely that the mRNA was translatedsince complete dissociation of mRNA expression and proteintranslation is rare. In addition, these data are consistent withprevious studies that have demonstrated elevated levels of LIF inpatients with a variety of acute and chronic inflammations includ-ing sepsis [22], arthritis [32], and renal transplantation [33]. In thelatter group of patients, LIF was identified within the urine butnot in the serum of renal allograft recipients during episodes ofacute rejection. Similarly, LIF has been demonstrated within theserum [23, 25] and bronchoalveolar lavage fluid [21] of rodentsfollowing endotoxin instillation.

LIF exhibits a broad spectrum of biologic activities that seem-ingly appear to be unrelated [18]. Initially described as a factorwhich induces the differentiation of murine Ml leukemia cells,LIF also regulates differentiation, growth, metabolism, and in-flammation [181. It is produced by a variety of cells includingfibroblasts [34], monocytes/macrophages [35], osteoblasts [36], Tlymphocytes [37] and endothelial cells [38] in response to endo-toxin and inflammatory cytokines such as IL-i and TNF. Basedupon the data presented, glomerular mesangial cells may now beadded to this list. Mesangial cells have been shown to produce avariety of inflammatory cytokines including IL-i, IL-6 and TNF[39—41]. More recently, mesangial cells were reported to secreteIL-b that inhibited endotoxin induced TNF-a production [42].The ability of mesangial cells to produce IL-b and LIF suggestthat this cell type possesses the capability to autoregulate inflam-mation by producing pro- and anti-inflammatory cytokines.

The continuous infusion of rhLIF produced a 60% decline inglomerular monocyte/macrophage infiltration which was associ-ated with an 85% decrease in urinary protein excretion at 24hours. The ability of rhLIF to interact with rat tissues has beenpreviously demonstrated both in vitro and in vivo [21, 261. A

0.03

0.02

0.01

za:EI0a-

0za:E

LL-J

0 5 10 15 20 25

1926 Tang et al: LIF ameliorates a-GBMAb GN

continuous infusion of rhLIF was chosen due to the short half-lifeof this peptide (6 to 8 mm) with the kidneys being the major siteof clearance [43]. The beneficial effect of rhLIF was not related toalterations of the GN since rhLIF administration did not affectglomerular binding of the a-GBM Ab. The beneficial effect ofrhLIF was associated with a decrease in the expression of theinflammatory cytokines, IL-1/3 and TNF-a, and of the chemokine,MCP-1, within the glomeruli of nephritic rats. The latter chemo-kine has been shown to be an important mediator of glomerularmacrophage infiltration in anti-GBM Ab GN [44]. These resultsare similar to those reported by Ulich et al in which the intratra-cheal administration of rhLIF decreased TNF expression withinthe alveolar lavage fluid by 60% [211. In contrast, IL-4 decreasedalbuminuria by 60 to 80% in anti-GBM Ab GN but did not affectglomerular IL-i mRNA expression [18].

In conclusion, these data demonstrate rapid up-regulation ofLIF mRNA following the induction of anti-GBM Ab GN with themesangial cell as a potential site of LIF production. The pharma-cologic administration of rhLIF demonstrates that this cytokinefunctions to down-regulate inflammation within the glomerulus.rhLIF decreased glomerular macrophage infiltration by 60% 24hours after the induction of anti-GBM Ab GN, which wasassociated with an 85% reduction in proteinuria. The beneficialeffect of rhLIF was associated with decreased cytokine andchemokine production. The efficacy of rhLIF in experimentalanti-GBM Ab GN suggests a potential role for this cytokine in thetherapy of glomerular inflammation.

Acknowledgment

These data were presented in abstract form at the 29th Annual AmericanSociety of Nephrology Meeting, San Diego, California, USA, November5—8, 1995.

Reprint requests to Winson W Tang, M.D., Amgen Inc., 1840 DeHavillandDrive, MS124-2-A, Thousand Oaks, California 91320, USA.

References

1. MALKOVSKY M, SONDEL PM, STROBER W, DALGLEISH AG: Theinterleukins in acquired disease. Clin Exp Immunol 74:151—161, 1988

2. DINARELLO CA: Interleukin 1 and interleukin-1 antagonism. Blood77:1627—1652, 1991

3. BEUTLER B, CERAMI A: Cachectin: More than a tumor necrosis factor.N EngI J Med 316:379—285, 1987

4. MILLER MD, KRANGEL MS: Biology and biochemistry of the chemo-kines: A family of chemotactic and inflammatory cytokines. Crit RevImmunol 12:17—46, 1992

5. SPRINGER TA: Adhesion receptors of the immune system. Nature346:425—434, 1990

6. BEVILACQUA MP, POBER JS, MAJEAU RS, FIER5 W, COTRAN RS,GIMBRONE MA JR: Recombinant tumor necrosis factor inducesprocoagulant activity in cultured human endothelium: Characteriza-tion and comparison with the actions of interleukin-1. Proc NatlAcadSci USA 83:4533—4537, 1986

7. DAYER JM, BEUTLER B, CERAMI A: Cachetin/tumor necrosis factorstimulates collagenase and prostaglandin E2 production by humansynovial cells and dermal fibroblasts. J Exp Med 162:2163—2168, 1985

8. SHALABY MR, AGGARWAL BG, RINDERKNECHT E, SVEDERSKY LP,FINKLE BS, PAI.IADINO MA JR: Activation of human polymorphonu-clear neutrophil functions by interferon-y and tumor necrosis factor.J Immunol 135:2069—2073, 1985

9. JORDAN M, Ol-rERNESS IG, NG R, GESSNER A, ROLLINGHOFF M,BEUSCHER HU: Neutralization of endogenous TL-6 suppresses induc-tion of li-i receptor antagonist. J Immunol 154:4081—4090, 1995

10. KLINE JN, FISHER PA, MONICK MM, HUNNINGHAKE GW: Regulationof interleukin-1 receptor antagonist by Thi and Th2 cytokines. Am JPhysiol 269:L92—L98, 1995

11. EISENBERG SP, EVANS RJ, AREND WP, VERDERBER E, BREWER MT,HANNUM CH, THOMPSON RC: Primary structure and functionalexpression from complementary DNA of a human interleukin-lreceptor antagonist. Nature 343:341—346, 1990

12. HART PH, Vrvrl GF, BURGESS DR, WHIrrY GA, PIccoLl DS,HAMILTON JA: Potential antiinflammatory effects of interleukin 4:Suppression of human monocyte tumor necrosis a, interleukin 1, andprostaglandin E2. Proc NatlAcad Sci USA 86:3803—3807, 1989

13. PELEMAN R, Wu JC, FARGEAS C, DELE5PESSE G: Recombinantinterleukin 4 suppresses the production of interferon y by humanmononuclear cells. J Exp Med 170:1751—1756, 1989

14. DE WAAL MALEFYT R, ABRAMS RJ, BENNETT B, FI000R CU, DEVRIES JE: Interleukin 10 (IL-b) inhibits cytokine synthesis by humanmonocytes: An autoregulatory role of IL-lO produced by monocytes.JExp Med 174:1209—1220, 1991

15. CASTELLA MA, MEDA L, BONORA S, CESKA M, CONSTANTIN G:Interleukin 10 (IL-b) inhibits the release of proinflammatoly cyto-kines from human polymorphonuclear leukocytes: Evidence for anautocrine role of tumor necrosis factor and IL-1/3 in mediating theproduction of IL-8 triggered by lipopolysaccharide. J Exp Med 178:2207—2211, 1993

16. TANG WW, FENG L, VANNICE JL, WILSON CB: Interleukin-l receptorantagonist ameliorates experimental anti-glomerular basement anti-body-associated glomerulonephritis. J Clin Invest 93:273—279, 1994

17. HRUBY ZW, SHIROTA K, JOTHY S, LOWRY RP: Antiserum againsttumor necrosis factor-alpha and a protease inhibitor reduce immuneglomerutar injury. Kidney mt 40:43—51, 1991

18. REES AJ: Inhibition of pro-inflammatory cytokines in glomerulone-phritis. XHIth International Congress of Nephrology, Madrid, July 2—6,1995, p 73 (abstract)

19. METCALF D: The leukemia inhibitory factor. Int J Cell Cloning9:95—108, 1991

20. KISHIMOT0 T, AKIRA 5, NARAZAKI M, TAGA T: Interleukin-6 family ofcytokines and gpl3O. Blood 86:1243—1254, 1995

21. ULICH TR, FANN M-J, PATTERSON PH, WILLIAMS JH, SAMAL B, DELCASTILLO J, YIN 5, GUO K, REMICK DG: Intratracheal injection ofLPS and cytokines. V. LPS induces expression of hF and LIF inhibitsacute inflammation. Am J Physiol 267:L442—L446, 1994

22. WARING PM, WYCHERLEY K, CARY D, NIcol.A N, METCALF D:Leukemia inhibitory levels are elevated in septic shock and variousinflammatory body fluids. J Clin Invest 90:2031—2037, 1992

23. WARING PM, WARING U, BILLINGTON T, METCALF D: Leukemiainhibitory factor protects against experimental lethal Escherichia coliseptic shock in mice. Proc NatlAcad Sci USA 92:1337—1341, 1995

24. ALEXANDER HR, WONG GGH, DOHERTY GM, VENZON DJ, FRAKERDL, NORTON JA: Differentiation factor/leukemia inhibitory factorprotection against lethal endotoxemia in mice: Synergistic effect withinterleukin 1 and tumor necrosis factor. J Exp Med 175:1139—1142,1992

25. BLOCK MI, BERG M, MCNAMARA MJ, NORTON JA, FRAKER DL,ALEXANDER HR: Passive immunization of mice against D factorblocks lethality and cytokine release during endotoxemia. J Exp Med178:1085—1090, 1993

26. SAMAI, BB, ARAKAWA T, BOONE TC, JONES T, PRESTRELSKI SJ, NARHILO, WEN J, CRANDALL CA, POPE J, SUGGS 5: High level expression ofhuman leukemia inhibitory factor (LIF) from a synthetic gene inEscherichia coli and the physical and biological properties of theprotein. Biochim Biophys Acta 1260:27—34, 1995

27. YAMAMORI T, FUKADA K, AEBERSOLD RH, KORSCHING S, FANN M-J,PATTERSON PH: The cholinergic neuronal differentiation factor fromheart cells is identical to leukemia inhibitory factor. Science 246: 1412—1416, 1989

28. SHIIAI T, SHIMIzu N, HORIGUCHI S, ITO H: Cloning and expression inEscherichia coli of the gene for tumor necrosis factor. Agric Biol Chem53:1733—1736, 1989

29. MARCH CJ, MOSLEY B, LARSEN A, CERRETFI DP, BRAEDT U, PRICE V,GILLIS S, HENNEY CS, KRONHEIM SR, GRABSTEIN K, CONLON PJ,Ho TP, COSMAN D: Cloning, sequence and expression of twodistinct human interleukin-1 complementary DNAs. Nature 315:641—647, 1985

Tang et al: LIF ameliorates a-GBMAb GN 1927

30. TELFORD JL, MACCHIA G, MASSONE A, CARINCI V, PALLA F, MELLIM: The murine interleukin 1/3 gene: Structure and function. Nuc AcidRes 14:9955—9963, 1986

31. TANG WW,YIN S, WIYrWER AJ, Qi M: Chemokine gene expression inanti-glomerular basement membrane antibody glomerulonephritis.Am J Physiol 269:F323—F330, 1995

32. LUTZ 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

33. TAUPIN JL, MOREL D, MOREAU JF, GUALDE N, POTAUX L, BEZIANJ-H: HILDAJLIF urinary excretion during acute kidney rejection.Transplantation 53:655—658, 1992

34. TOMIDA M, YAMAMOTO-YAMAGUCHI Y, HOZUMI M: Purification of afactor inducing proliferation of mouse myeloid leukemia MI cellsfrom conditioned medium of mouse fibroblast L929 cells. J Biol Chem259:10978—10982, 1984

35. ANEGON I, MOREAU J-F, GODARD A, JACQUES Y, PEYRAT M-A,HALLET M-M, WONG G, SouI,ILL0U JP: Production of human inter-leukin for DA cells (HILDA)/leukemia inhibitory factor (LIF) byactivated m000cytes. Cell Immunol 130:50—65, 1990

36. IsHIMI Y, ABE E, uN CH, MIYAURA C, HONG MH, OsHIoA M,KUROSAWA H, YAMAGUCHI Y, TOMIDA M, HOzUMI M, SODA T:Leukemia inhibitory factor/differentiation-stimulating factor (LIF/D-Factor): Regulation of its production and possible role in bonemetabolism. J Cell Physiol 152:71—78, 1992

37. GEARING DP, GOUGH NM, KING JA, HILTON DJ, NICoIA NA,SIMPSON RJ, NICE EC, KELSO A, METCALF D: Molecular cloning and

expression of eDNA encoding a murine leukaemia inhibitory factor(LIF). EMBO J 6:3995—4002, 1987

38. GROSSET C, JAZWIEC B, TAUPIN J-L, LIu H, RICHARD S, MAHON F-X,REIFFERS J, MOREAU J-F, RIPOCHE J: In vitro biosynthesis of leukemiainhibitory factor/human interleukin for DA cells by human endothe-hal cells: Differential regulation by interleukin-lo and glucocorticoids.Blood 86:3763—3770, 1995

39. LovErr DH, SZAMEL M, RYAN JL, STERZEL RB, GEMSA D, RESCH K:lnterleukin-1 and the glomerular mesangium. I. Purification andcharacterization of mesangial cell-derived autogrowth factor. J Immu-nol 136:3700—3705, 1986

40. RUBE C, BUDDE K, LACY J, NORTHEMAN W, BAUMANN M, STERZELRB, COLEMAN DL: Interleukin 6 is an autocrine growth factor formesangial cells. Kidney mt 38:249—257, 1990

41. BAUD L, OUDIENT J-P, BENS M, NOE L, PERALDI M-N, RONDEAU E,ETIENNE J, ARDAILLOU R: Production of tumor necrosis factor by ratmesangial cells in response to bacterial lipopolysaccharide. Kidney mt35:1111—1118, 1989

42. FOUQUERAY B, PIIILIPPE C, BOUTARD V, KORNREICH A, MARCHANTA, PEREZ J, GOLDMAN M, BAUD L: Mesangial cell-derived Interleu-kin-lO modulates MC response to lipopolysaccharide. Am J Patholt47:176—182, 1995

43. HILTON DJ, NICOLA NA, WARING PM, METCALF D: Clearance andfate of leukemia-inhibitory factor (LIF) after injection into mice. JCell Physiol 148:430—439, 1991

44. TANG WW, 01 M, WARREN JS: The role of monocyte chemoattractantprotein-i (MCP-1) in anti-GBM Ab GN. Kidney mt 50:665—67t, 1996