Biosynthesis of C3 by human mesangial cells. Modulation by proinflammatory cytokines

Kidney International, VoL 47 (1995), pp. 829-836

CLINICAL INVESTIGATION

Biosynthesis of C3 by human mesangial cells. Modulation by

proinflammatory cytokinesVINCENZO MONTINARO, LEONARDO SERRA, SABRINA PERIssurrI, ELENA RANIERI, FRANCESCO TEDESCO,

and FRANCESCO PAOLO SCHENA

Division of Nephrology, University of Ban, "Istituto per l'Infanzia", Trieste and Istituto di Patologia Generale, Universitd di Trieste, Italy

Biosynthesis of C3 by human mesangial cells. Modulation by prom-flammatory cytokines. Deposits of complement (C) components are foundin the glomeruli of patients with various glomerulonephritides withoutdetectable immunoglobulins, thus suggesting a pathogenetic role of thelocally produced proteins of this system. In the present study, we haveexamined human mesangial cells (HMC) for their ability to secrete C3.Three different cell lines were examined and all showed a basal productionof C3, which was up-regulated following stimulation with IL-i/3. IL-6 hadno direct stimulatory effect on its own, but synergized with IL-i to inducean increased production of C3 in the culture supernatant and its relativeamount was confirmed by SDS-PAGE and immunoblot. Another agonistsuch as lipopolysaccharide was not able to induce any significant C3synthesis. Analysis of C3 HMC gene expression, performed by bothreverse transcription-polymerase chain reaction of isolated RNA andNorthern blot, confirmed the parallel increase of the specific transcriptunder IL-i /3 and IL-i/3 + IL-6 stimulation. From these data we concludethat production of C3 in the mesangium could have a pathophysiologicrelevance.

The complement (C) system has been implicated in the onset ofrenal damage in various forms of glomerulonephritides [1] and inexperimental models of immune glomerular injury [2]. Tissuedamage is mediated through the generation of C-derived prom-fiammatory products or is directly induced by the lytic terminal Ccomplex [1.

The C components responsible for renal tissue injury mostlyderive from the liver, since the bulk of circulating C proteins isproduced by hepatocytes [4, 5]. The activity of these cells has beenshown to be under regulatory controls of various cytokinesincluding IL-i, TNF and IL-6 [6, 7]. However, other cell typeshave been found to synthesize C components, thus providing anextrahepatic source of C that can also be modulated by cytokines[7].

The report by Falus et al [8] that the kidneys of mice treatedwith IL-i and LPS express the genes of C2, factor B and C3 andactively produce the corresponding proteins, indicates that thisorgan is one of the extrahepatic sites of C production. In addition,gene expression of components of the classical and alternative Cpathways observed in lupus-prone mice (NZB/W and MRLlpr/lpr) is related to the degree of mononuclear cell infiltration and

Received for publication July 1, 1993and in revised form October 31, 1994Accepted for publication October 31, 1994

© 1995 by the International Society of Nephrology

hence to the development of nephritis [9, 10], further emphasizingthe contribution of locally produced C components to tissuedamage.

The establishment of optimal culture conditions for renal cellshas allowed investigations on their ability to synthesize C compo-nents. Thus, proximal tubular epithelial cells have been shown tosecrete C4 and C3, particularly following stimulation by interfer-ou-y and IL-2 [11, 12] and glomerular epithelial cells alsosynthesize C proteins [13]. As to the mesangial cells, there is clearevidence that they release C8 binding protein [14] and decayaccelerating factor [15], two regulatory proteins that inhibit theactivity of the CSb-9 complex and C3 convertase, respectively [16,17].

The purpose of this investigation was to examine the ability ofhuman mesangial cells (HMC), obtained from different donors, tosecrete C3. The modulatory effect of cytokines on the productionof the proteins was also investigated.

Methods

Reagents

Standard cell culture reagents were obtained from differentsuppliers. Fetal bovine serum (FBS) was purchased from 1-lyClone(Logan, Utah, USA) and decomplemented for 30 minutes at 56°Cbefore use. RPMI, Hank's balanced salt solution (HBSS), bovineinsulin, L-glutamine, transferrin, sodium selenite, Hepes wereobtained from Sigma Chemical Co. (St. Louis, MO, USA).Penicillin and streptomycin were purchased from Gibco Labora-tories (Grand Island, NY, USA). Sodium pyruvate was purchasedfrom ICN Biomedicals (Costa Mesa, CA, USA), while non-essential amino acids were obtained from Seromed (Munich,Germany).

Guanidinium isothiocynate was obtained from BRL Life Tech-nologies (Gaithersburg, MD, USA), and Sarcosyl was from SigmaChemical Co.

Complement components and antisera

Purified C3, was purchased from Quinlan (La Jolla, CA, USA),which provided also monospecific goat IgG antibodies to humanC3.

Polyclonal rabbit anti-human fibronectin and laminin wereobtained from BRL Life Technologies, rabbit anti-desmin waspurchased from Sigma Chemical Co., anti-vimentin was obtainedfrom Polysciences (Warrington, PA, USA) and anti-factor VIII

829

830 Montinaro et a!: Biosynthesis of C3 by human mesangial cells

was provided by Boehringer Mannheim (Mannheim, Germany).Monoclonal anti-cytokeratin was purchased from Triton Bio-sciences (San Diego, CA, USA) and anti-macrophage antigen wasobtained from Dakopatts (Glostrup, Denmark).

Biotin-labeling of antibodiesGoat IgG to C3 was biotin-labeled according to a previously

reported procedure [18] with the only modifications that IgG wasdialyzed against 0.1 M sodium bicarbonate buffer pH 8.8 contain-ing 0.1 M NaC1 and was subsequently mixed with D-biotin-N-hydroxysuccinimide ester (Sigma Chemical Co.) at the ratio of 10to 1 (wt/wt).

CytokinesThe cytokines used in this study were recombinant human

IL-lp and IL-6. IL-lp was obtained through the courtesy of Prof.N. Semeraro (Institute of General Pathology, University of Ban,Ban, Italy). Specific activity (assessed by the murine thymocytemitogenicity assay) was 2.0 X i07 U/mg protein. IL-6 was pur-chased from Boehringer Mannheim.

Isolation of mesangial cellsNormal portions of renal cortex from kidneys removed for

carcinoma were used. Glomeruli were isolated by a standardsieving technique through graded mesh size screens. The lastpassage was performed onto a nylon 70 tm pore-size screen whichretained the glomeruli and allowed small tubular fragments andisolated cells to pass through. Glomeruli, decapsulated by forcedflow through a 19 gauge needle, were treated by gentle collage-nase type IV digestion (750 U/mi in HBSS) for 15 minutes at37°C. Thereafter, they were plated onto 10 cm Petri dishes inRPM! 1640 supplemented with 17% FBS, 2 mrvi L-glutamine,antibiotics (streptomycin, 100 g/ml, penicillin, 100 U/mI), 2 mMpyruvate, non-essential amino acids, insulin (5 jsglml), transferrin(5 ng/ml), sodium selenite (5 nglml) and 25 mrvi Hepes (completemedium). Outgrowth of mesangial cells was observed after four tofive days of culture, and homogeneous cultures were obtainedafter the first passage.

Characterization of cultured mesangial cells

Purity of cell cultures was checked by immunoperoxidasestaining with monoclonal (cytokeratin, macrophage antigen) andpolyclonal (factor VIII, desmin, fibronectin, laminin) antibodies.Briefly, HMC were grown to confluence onto glass slides, fixed informalin for 10 minutes, and after washing in PBS were incubatedfor 30 minutes with the specific antibody at the appropriatedilution. The slides were then washed in PBS and subsequentlyexposed to biotinylated secondary antibody (goat anti-mouse oranti-rabbit) for further 30 minutes. The bound biotin was allowedto react with avidin-biotin complex (Vectastain ABC kit, VectorLaboratories; Burlingame, CA, USA) and the enzymatic reactionwas developed using as substrate DAB and H202. The purity ofHMC was confirmed by its negative reaction with cytokeratin(epithelial), factor VIII (endothelial) or a membrane marker ofmonocytes, and, by contrast, for their positive reaction withantibodies to desmin, fibronectin, vimentin and laminin.

Immunohistochemistiy staining for monocytesTo further confirm the absence of contaminating monocytes-

macrophages within HMC cultures an immunochemistry tech-

nique based on the APAAP method was performed as previouslydescribed [19]. Briefly, HMC were grown onto plastic slides, atsubconfluence the medium was removed, the slides were washedthrice with Hank's balanced salt solution and, after air drying,were fixed by brief immersion in ice-cold acetone. Positive controlslides were prepared with peritoneal macrophages isolated bysequential centrifugation and plastic adhesion of the peritonealexchange fluid from a patient treated by CAPD. To maximize DRexpression, macrophages were stimulated for 24 hours with LPS(1 g/ml).

The primary antibody was a mouse monoclonal anti-humaninvariant DR antigen (Technogenetics, Cassina de' Pecchi, Mi-lano, Italy). Immobilized antibodies were detected by immunoal-kaline phosphatase (APAAP) with affinity purified rabbit anti-mouse immunoglobulin serum (Dakopatts) and APAAP complex(1:50 dilution). Alkaline phosphatase was developed by NewFuchsin (Sigma).

Cell culture experiments

Pure HMC cultures between passages 3 and 7 were allowed togrow to near confluence in complete medium, and then kept inresting conditions for 24 hours in a basal medium made of RPM!1640, 0.5% FBS, 2 mrvi L-glutamine, 2 m sodium pyruvate, poolof antibiotics and non-essential aminoacids. To exclude the pos-sible interference of proteins present in FBS on the C3 ELISA, insome experiments cells were also maintained, after confluence, ina basal medium composed as above with the substitution of FBS0.5% with BSA 0.1%.

Human recombinant IL-1/3 or IL-6 were added to HMC atconcentrations of 10, 100 or 1000 U/mI in a basal medium. Thecells were kept in culture for 72 hours and the supernatant washarvested afterward. The combined effect of IL-1J3 (10 U/mI) andIL-6 (1000 U/mi) was also evaluated.

Experiments were conducted to assess the eventual agonisticeffect of lipopolysaccharide (LPS) on C3 synthesis. For this reasonHMC were stimulated for 72 hours with LPS from E. coli atconcentration of 0.1 or 1 g/m1.

Finally, in some experiments with cytokines, the supernatantwas sampled at 24 hour intervals for the kinetic evaluation of C3production.

Measurement of C3 by ELISA

The levels of the C3 in the cell culture medium was measuredby a sandwich technique [18]. Briefly, the individual wells of a 96well flat bottomed microtiter plate (Maxisorb Nunc, MasciaBrunelli, Milano, Italy) were coated by an overnight incubationwith the purified IgG of the antiserum to C3 diluted 1:500 in0.1 M Na2HCO3 pH 9.6 (200 pd/well). After washing with PBScontaining 0.1% Tween 20 and blocking with PBS containing 1%BSA for one hour at 37°C, the bound antibodies were allowed toreact first with 200 j.Ll of cell culture supernatants for one hour at37°C, and then with the biotin-labeled specific antibodies for anadditional hour at 37°C, followed by alkaline phosphatase-labeledstreptavidin (Jackson Laboratories, West Grove, PA, USA) di-luted 1:8000. The enzymatic reaction was revealed using p-nitrophenyl phosphate (1 g/m1) as substrate and reading wasperformed with a Titertek Multiskan ELISA reader (Flow Labo-ratories, Milano, Italy) at 405 nm. Control culture medium servedas blank. Calibration curves were set up with increasing amountsof the purified component and sensitivity of the assay was 1 ng/ml.

'a '9,S.

.7S I.,•t.e

p.,

Ii ,a

nfl'VS u F ,• • Pg '. •0

— S

S

•1

S

a

SS

—

St. C,4*' I

1

e

S.

— S

a a

-

• a• •'

SS

'5 B a

4r

Montinaro et a!: Biosynthesis of C3 by human mesangial cells 831

Fig. 1. Immunoalkaline staining of HMC culture with an anti-DR antibody to exclude contamination of monocyte-macrophages. HMC line #3 grown ontoplastic slides or human peritoneal macrophages (positive control) were stained with a mouse monoclonal anti-DR (A, C) or an irrelevant mousemonoclonal antibody (B, D). The alkaline phosphatase was developed with New Fuchsin (red color). Reproduction of this figure in color was madepossible by a grant from Hoechst Italia Hospital Supply Department, Milan, Italy.

Hemolytic assay for C3The hemolytic activity of C3 secreted by mesangial cells was

evaluated by a test system that included 50 p.l of 1gM-sensitizedsheep erythrocytes (3 >< 108 cells/mi), 200 .tl of culture superna-tant and 7 p.l of human serum depleted of C3 by affinitychromatography (Sigma Chemical Co.). The mixture was incu-bated at 37°C for 30 minutes and the degree of lysis was measuredspectrophotometrically at 415 nm.

SDS-PAGE and immunoblot

The culture supernatants of mesangial cells either unstimulatedor treated with cytokines were analyzed by SDS-PAGE on a 7%gel under non-reducing conditions according to Laemmli [201.The samples were dialyzed against distilled water, lyophilized andresuspended in the sample buffer prior to loading on the gel. Aftercompletion of the run, the proteins were electrophoreticallytransferred onto the nitrocellulose membrane (Schleicher &Schuell Inc., Dassel, Germany) using a transblot semidry system(Hoefer Scientific Instruments, San Francisco, CA, USA) and thefree binding sites were blocked with an overnight incubation withPBS pH 7.4 containing 2% BSA. The nitrocellulose was thenexposed to goat IgG against C3 diluted 1:100 for one hour at 37°Cand subsequently to 1:1000 peroxidase-labeled rabbit antibodiesto goat IgU (Sigma Chemical Co.) for one additional hour. The

enhanced chemiluminescence (ECL) system purchased by Amer-sham (Amity, Milano, Italy) was employed to develop the enzy-matic reaction following the instructions of the manifacturer andthe chemiluminescence signal was detected by exposing thenitrocellulose sheet to X-Omat AR films (Kodak, Rochester, NY,USA).

Extraction of RNA

HMC pellets were treated with guanidinium denaturing solu-tion (4 M guanidinium isothiocyanate, 25 m sodium citrate, pH 7;0.5% Sarcosyl and 0.1 M 2-mercaptoethanol) and stored at —80°C.Total cellular RNA was extracted by the single step acid guani-dinium, phenol/chloroform extraction method [21]. Purity ofextracted RNA was checked by A2601280 reading and agarose-formaldehyde denaturing gel electrophoresis to evaluate theintegrity of 28S and 18S bands.

Analysis of C3 transcriptTwo micrograms of total RNA from each specimen were

reverse transcribed into cDNA by using the RNA-PCR kit(Perkin-Elmer, Norwalk, CT, USA) in presence of RNasin andoligo-dT as a universal primer for messenger RNAs. Reactionproceeded for 15 minutes at 42°C and was stopped by increasingtemperature at 99°C for five minutes. Then, the reaction mixture

832 Montinaro et al: Biosynthesis of C3 by human mesangial cells

160

140

120

C" 100U)

80

60

C 4020

0

1

0

Fig. 3. Effect of IL-I on the basal synthesis of C3 by HMC. Cells werestimulated with IL-113 at different concentration in basal medium for 72 hr.Symbols are cell lines: (—LI—) #1; (U-) #2; (-—D--) #3. Valuesrepresent mean se of three or four determinations. * P < 0.05, * * p <0.01 versus basal.

was supplemented with deoxynucleotides, Taq polymerase, andprimers specific for the human C3 gene (donated by Dr. A. Rifai,Providence, RI, USA). These primers were: downstream, CTT-GAGATCTGTACCAGGTACC, and upstream, CTrGATG-GTCTGCTCAATGGC, and included a 554 base pair long frag-ment spanning from positions 4216 to 4769 nucleotide residue ofthe eDNA [22]. Amplification was carried out in a termal cycler(Perkin-Elmer) for 30 cycles (denaturing for 2 minutes at 95°C,annealing for 2 minutes at 55°C, extension for 1 minute at 72°C).Analysis of amplified products was performed by 3:1 NuSieve:agarose (FMC BioProducts, Rockland, ME, USA) gel electro-phoresis in TAE buffer and staining with ethidium bromide, Theintensity of bands of the amplified product of the housekeepinggene GAPDH (primer downstream, TGGTATCGTGGAAGG-

Fig. 4. Kinetics of C3 production by HMC line #2 stimulated with IL-l/3compared to basal conditions (I). Sampling of supernatants was performedevery 24 hours. Each value represents the mean of three determinations.

0 10 100 1000

IL-6, U/miFig. 5. Effect of IL-6 on the basal synthesis of C3 by HMC. Symbols are celllines: (—LI—) #1; (---U---) #2; (— —. — —) #3. Cells were stimulated withIL-6 at the indicated concentration in basal medium for 72 hr. Values aremean SE of three determinations.

ACTCATGAC; primer upstream, ATTCGYFGTCATACCAG-GA; fragment length 450 bp) was also considered.

To further confirm the sytnhesis of the specific C3 transcript byHMC, total cellular RNA (30 jLg) was electrophoresed through1% agarose denaturing gel containing 2.2 M formaldehyde, thenblotted onto nylon membrane (Nytran N, Schleicher & Schuell)and hybridized with a specific probe for human C3. After washingwith a final stringency of 0.2x SSC, 0.1% SDS at 55°C for 15minutes, the membrane was exposed to a X-Omat AR Kodak film.

160 -

140 -

120 -

.i 100 -a)0

80

C 60'

40 -

20

(slope 1 76)

ope0.37)1 2 3

0' -

20 30 40 50 60 70 80

HMC line number

Fig. 2. Concentration of C3 in supernatants of three different HMC linescultured in basal medium for 72 hr. Values are mean SE of quadruplicatedeterminations.

5

4in

>0C,)

11)o2C

0U-

Time, hours

*

U)

,0a)>0a,ci)

a)0C

0U-

3

2

110 100

IL-i, U/mi

1000

1 2 a 4

—554 bp

—450 bp

Ca)

a)

a)Ca)Ca

C.)

•60U-

Fig. 7. Effect of FBS and inhibition of protein synthesis on C3 detection onsupernatants. HMC (line #3) were kept in basal condition with RPMImedium containing either FBS 0.5% or BSA 0.1% as protein carrier.Other conditions consisted in IL-113 100 U/mI with or without cyclohexi-mide 10 g/m1 in basal medium containing BSA 0.1%. All these cultureconditions were mantained for up to 72 hours.

The membrane was then stripped and rehybridized with aGAPDH-specific probe.

The C3 probe used was the pHLC3.11 clone [22] obtained fromthe American Type Culture Collection. The insert was removedby double Sal I/Cia I digestion, which yelded two fragments of 2.4and 1.8 kb for the presence of an internal Sal I site. The 1.8 kbfragment was purified from agarose gel using a kit based onmicrosilica beads (Geneclean, Bio 101, La Jolla, CA, USA), thenit was radiolabeled with 32P-dCFP (Amersham Italia, Milano,Italy) by the random primer method (Megaprime DNA labelingsystem, Amersham).

The purity of HMC used throughout this study was assessed bystandard immunoperoxidase techniques (Methods) to excludecontamination with glomerular epithelial or endothelial cells.Moreover, since bone marrow-derived cells may be residentwithin the glomerulus, we performed an immunoalkaline phos-phatase staining of HMC cultures with a monoclonal anti-DRantibody. As shown in Figure 1 DR-positive cells were completelyabsent in HMC cultures.

Three different HMC lines were tested for their ability tosecrete C3. Detectable levels of C3 were revealed in the superna-tant of cells cultured in the absence of cytokines, though theabsolute amounts varied with the different cell lines (Fig. 2). Thesame HMC line tested under different passage (3 to 7) showedsimilar results (data not shown).

The intra- and interassay variations of the ELISA for C3 were11% and 20%, respectively.

The secretion of C3 by HMC was up-regulated by IL-1/3, whichinduced the cells to release an amount of C3 two- to fivefoldhigher than that produced in basal conditions (Fig. 3). It isinteresting to note that the effect of the cytokine varied with theHMC lines and no clear relationship was found between the C3levels and the concentrations of IL-i j3. Thus, the cell lines #1 and#3 showed the maximal response at low concentration of IL-13(10 U/ml), whereas line #2 produced a significantly increasedamount of C3 only at higher concentration of IL-1/3. A kinetic

Montinaro et al: Biosynthesis of C3 by human mesangial cells 833

**5

4

3

2

0

*

Basal IL-i IL-6 IL-1+IL-6

Fig. 6. Effect of a combination of IL-1f3 (10 U/mi) and IL-6 (1000 U/mi) onC3production by HMC (line #1). Cytokine stimulation was carried out for72 hr, then supernatants were tested for C3 by ELISA. Values representmean of quadruplicate determinations. < 0.05, < 0.01 versus basalor IL-6.

Fig. 8. Analysis of C3 gene expression by HMC (line #1). Total RNAs werereverse transcribed into eDNA, then amplified by polymerase chainreaction using primers specific for C3 (A) or the housekeeping geneGAPDH (B). Amplified products were run in 3:1 NuSieve:agarose gel andstained with etidium bromide. Lane 1: molecular weight standard (cIX 174RF DNA Hae III); lane 2: basal; lane 3: IL-i (10 U/ml); lane 4: IL-6 (1000UIml); lane 5: IL-i (10 U/ml) and IL-6 (1000 U/ml). (bp = base pairs).

Results

6

5

4

3

2

1

0Basal Basal

(FBS 0.5%) (BSA 0.1%)IL-i IL-i + CHX

experiments were conducted in absence of FBS which mightinterfere with the C3 ELISA and others in presence of the proteinsynthesis inhibitor, cycloheximide. As shown in Figure 7, thepresence of FBS 0.5% in the medium did not appreciably modifythe C3 assay; moreover, inhibition of protein synthesis by cyclo-heximide blocked the secretion and release of C3 induced byIL-1f3 to an extent of 94%.

The culture supernatant of mesangial cells exhibited a detect-able, albeit low, C3 hemolytic activity that was comparable withthat of pooled human sera at dilutions varying between 1/150,000and 1/200,000.

The C3 specific gene expression was investigated by reversetranscription of isolated RNA and subsequent polymerase chainreaction of cDNA with specific primers. Low levels of the C3transcript were present in HMC under basal conditions, whereasthe message was up-regulated following stimulation with IL-1J3(Fig. 8). Analysis by Northern blot of total cellular RNA extractedfrom HMC, confirmed that the message was an intregral 5.2 kblong transcript (Fig. 9).

Finally, in accordance to the ELISA data and C3 gene expres-sion analysis, parallel findings were obtained when the C3 pro-duced by untreated or cytokine-stimulated HMC was analyzed bySDS-PAGE and Western blot (Fig. 10).

Discussion

Fig. 9. Northern blotting analysis of RNA estracted from HMC. RNA (30j.Lg/lane) from cell line #3 under basal conditions (lane 1) or stimulatedwith IL-i 10 U/mI (lane 2) or IL-i 10 U/mI and IL-6 1000 U/mI (lane 3)was electrophoresed through agarose gel, blotted onto nylon membraneand hybridized with the 32P-radiolabelled 1.8 kb fragment of insertobtained from pHC3.l1 clone. After exposure of an X-Omat film for oneweek (A), the membrane was stripped and rehybridized with a GAPDH-specific probe and exposed for 24 hours (B).

study of C3 secretion by the cell line #2 was performed in bothbasal conditions and after stimulation by the optimal amount ofIL-i (1000 U/ml). Samples of supernatants were collected at 24hour intervals and the level of C3 was assessed by ELISA. Asshown in Figure 4, the C3 synthetic rate, calculated by linearregression analysis of three-time points for each condition, was1.76 ng/106 cells/hr for the cells stimulated by IL-I, compared tothe basal value of 0.37 ng/106 cells/hr.

Stimulation of HMC by IL-6 did not induce an appreciablechange in the secretion of C3 by all three lines tested (Fig. 5).However, experiments performed on the HMC line #1 showedthat IL-6 added at the concentration of 1000 U/ml together withIL-i (10 U/ml) induced the release of an amount of C3 twofoldhigher than that obtained in the presence of IL-i alone, suggest-ing a synergistic effect of the two cytokines (Fig. 6).

Another stimulus tested was LPS. HMC line #3 cultured inpresence of LPS at concentration of 0.1 or i .tg/ml did not showany significant increase in the basal C3 synthesis (data not shown).

To confirm that the C3 detected by ELISA in the supernatantof cultures was really synthesized and secreted by HMC, some

The results of the present investigation provide evidence for thesecretion of C3 by HMC. This was demonstrated by: (i) immu-noenzymatic quantitation of C3 on the supernatants of cellcultures; (2) immunochemical demonstration of C3 in the culturemedium by SDS-PAGE and Western blot; and (3) expression ofspecific messenger RNA for C3 by reverse transcription-poly-merase chain reaction and Northern blot analysis. Our observa-tion extends previous reports by Rother, Hänsch and Rauterberg[14] and Shibata, Cosio and Birmingham [15], who showedproduction of regulators of C system by HMC.

A point to emphasize is that HMC spontaneously secrete C3even in basal conditions, though the extent of C production variedwith the different cell lines examined. Our findings are at variancewith those of Kaashoek et al [23], who were unable to revealsecretion of C3 by unstimulated cells, but did show the presenceof a C3 transcript. The reasons for the discrepant finding are notapparent, but they may be related to differences in cultureconditions. We have employed HMC cultured between passages 3and 7 and, under this experimental condition, the magnitude ofC3 secretion was essentially similar for each HMC line. A certaindegree of variation was observed in the absolute amount of C3secreted by the same cell line in separate experiments, but therelative amount of C3 produced by the individual cell lineremained fairly constant since the ratio of values was essentiallyunchanged. This finding suggests that the secretion of differentamounts of C3 is a feature of each cell line, and most likely reflectsa genetic background, a situation documented in inbreed mousestrains [8].

The stimulatory activity of IL-i/3 on C3 secretion by HMC isnot surprising, since this cytokine exerts a similar effect oncultured hepatocytes, which are stimulated to release a substantialamount of acute phase proteins [4, 24]. IL-la seems to have aneffect on HMC comparable to that of IL-113 (data not shown) inagreement with similar data reported by Kaashoek et a! [23].These findings are relevant in the pathophysiology of mesangial

834 Montinaro et al: Biosynthesis of C3 by human mesangial cells

1 2 3

A

28 S —

B

18S—

C

Fig. 10. Immunoblot analysis of C3 in theculture supernatant. The lyophilized samplescorresponding to 1 ml of the culture mediumwere run on 7% acrylamide gel under non-reducing conditions and blotted ontonitrocellulose membrane. A 5 ng of purified C3;B medium of unstimulated cells; C medium ofcells treated with IL-i (10 U/mI); D medium ofcells treated with IL-i (10 U/mI) and IL-6(1000 U/mI).

cells and may contribute to explain, at least in part, the prom-flammatory effect of IL-i in experimental models of variousglomerular diseases [25—31].

IL-6, the other cytokine tested, did not influence the spontane-ous production of C3. One possible explanation for this finding isthat HMC lack the receptor for this cytokine, although theincreased incorporation of 3H-thymidine by mesangial cellstreated with IL-6 does not support this possibility [32, 331.Furthermore, HMC respond to IL-6 with an increased secretionof C3 when the cytokine is added to the cell as a mixture with IL-i.A synergistic effect of the two cytokines was already reported in anexperimental model of IgA nephropathy in mice whereby theglomerular response to deposited IgA immune complexes and,more specifically, the intensity of glomerular C3 deposits and thelevel of C3 transcript strongly increased with a combined treat-ment with IL-i and IL-6 [31, 34]. On the other hand, the courseof the disease in this experimental model was not modified and, ifanything, was milder in mice treated with IL-6 alone. It istherefore possible that IL-6 may have a different effect on themesangium depending on whether or not an inflammatory processis established. The local presence of IL-i may favor the deleteri-ous effect of IL-6, which otherwise would be ineffective or evenprotective.

The role played by C components produced by HMC remainsto be elucidated. There is no doubt that these components canonly have a short range of action, given the low amount producedby the cells, and hence can only modulate pathogenetic mecha-nisms in a local fashion. Our finding that the C3 secreted bymesangial cells had a low hemolytic activity, though readilydetectable by ELISA, can in part be explained by the differentsensitivity of the two assays, but also by the fact that a substantialportion of C3 is enzymatically degraded and, consequently no

longer hemolytically active while being still measurable by ELISA.However, this does not exclude that the C3 fragments may exhibitproinflammatory functions more relevant to the pathologicalprocesses involving the glomeruli, though these functions aremore difficult to assess under our experimental conditions. Therecent observation by Lianos and Zanglis [35] that fragments ofC3, C3a and C3b, can stimulate rat mesangial cells to synthesizePGE2 and PAF would suggest that a vicious circle can beestablished whereby cytokines, C components, PAF and eico-sanoids, all produced locally and acting by autocrine or paracrinemodalities, contribute to tissue injury. The finding by the samegroup [36] that the local production of PAF is reduced indecomplemented animals in an experimental model of rat gb-merular immune injury further supports the importance of C inthe pathogenesis of the glomerular lesions. With regard to humandiseases, our finding may suggest that, at least in some istancescharacterized by glomerular deposition of C3 without detectableimniunoglobulins [37, 381, local production of C3 within mesan-gium could play a role.

In conclusion, HMC in culture express the specific C3 gene,produce detectable levels of mature C3 and release this protein inthe culture medium. Cytokines such as IL-i and IL-6 modulatethis response increasing the basal C3 production.

Acknowledgments

Part of this work has been presented at the "International Meeting onMolecular Approaches to Nephrology: Prospects in Diagnosis and Man-agement," held in Ban, Italy, March 19—21, 1992; and at the 25th AnnualMeeting of the American Society of Nephrobogy.

This work was partly supported by grants provided to F.P.S. by the CNR(Consiglio Nazionale delle Ricerche) Target Project on Biotechnologyand Bioinstrumentation (89.235.70, 90.115.70, 91.1253.70, 92.1272.70,93.1102.70); by the CNR Joint Program USA-Italy (90.1153.04, 91.466.04,

Montinaro et al: Biosynthesis of C3 by human mesangial cells 835

kDa

200 -3

92+

69 -9

836 Montinaro et al: Biosynthesis of C3 by human mesangial cells

92.769.04) and by the MURST (Ministero della Ricerca Scientifica eTecnologica) grants (91.1882, 92.7490, 93.5382). Grants obtained by FT.from CNR (92.02819.CTO4) and MURST (60 and 40%) are also gratefullyacknowledged. Dr. V. Montinaro was supported by a fellowship fromApulia Kidney Association. Reproduction of the color figure was madepossible by a grant from Hoechst Italia Hospital Supply Department,Milan, Italy. We thank Annalisa Brunaccini for technical support. Finally,we thank Dr. L. Gesualdo and T. Maffei for performing the immunoal-kaline phosphase technique.

Reprint requests to Prof F.P. Schena, Division of Nephrology, University ofBar, Polyclinic, P.zza G. Cesare, 11, 70124 Ban, Italy.

References

1. COUSER WG: Mediation of immune glomerular injury. J Am SocNephrol 1:13—19, 1990

2. COUSER WG, BAlout PJ, ADLER S: Complement and the directmediation of immune glomerular injury. Kidney mt 28:879—900, 1985

3. CYBULSKY AV, QUIGG RJ, SAL&NT DJ: The membrane attack com-plex in complement-mediated glomerular epithelial cell injury: For-mation and stability of C5b-9 and C5b-7 in rat membranous nephrop-athy.Jlmmunol 137:1511—1516, 1986

4. RAMADoRI G, SIPE JD, DINARELLO CA, MIZEL SB, COLTEN HR:Pretranslational modulation of acute phase hepatic protein synthesisby murine recombinant interleukin-1 (IL-i) and purified human IL-i.JExp Med 162:930—942, 1985

5. PERLMUTI-FER DH, GOLDBERGER G, DINARELLO CA, MIZEL SB,COLTEN HR: Regulation of class III major histocompatibility complexgene products by interleukin 1. Science 232:850—852, 1986

6. PERLMU1TER DH, COLTEN HR: Complement: Molecular genetics, InInflammation. Basic Principles and Clinical Correlates, edited by GAL-LIN JI, GOLDSTEIN TM, SNYDERMAN R, New York, Raven Press, 1988,p 75

7. Fus A, BEUSCHER HU, AUERBACFI HS, COLTEN HR: Hormonalregulation of complement biosynthesis in human cell lines. II. Up-regulation of the biosynthesis of complement C3, factor B and Clinhibitor by interleukin-6 and interleukin-1 in human hepatoma celllines. Mo! Immunol 27:197—201, 1990

8 Fus A, BEUSCHER HU, AUERBACH HS, COLTEN HR: Constitutiveand IL-i-regulated murine complement gene expression is strain andtissue specific. J Immunol 138:856—860, 1987

9. PASSWELL J, SCHREINER OF, NONAKA M, BEUSCHER HU, COLTENHR: Local extrahepatic expression of complement genes C3, factor B,C2, and C4 is increased in murine lupus nephritis. J Clin Invest82:1676—1684, 1988

10. PASSWELL JH, SCHREINER GF, WETSEL RA, COLTEN HR: Comple-ment gene expression in hepatic and extrahepatic tissues of NZB andNZBxW (Fl) mouse strains. Immunology 71:290—294, 1990

11. BROOLMANS RA, STEGMANN APA, v DORP WT, VAN DER ARK AAJ,VAN DER WOUDE FJ, VAN Es LA, DAMA MR: Interleukin 2 mediatesstimulation of complement C3 biosynthesis in human proximal tubularepithelial cells. J Clin Invest 88:379—384, 1991

12. BROOIMANS RA, SEELEN M, VAN Es L, DANA MR: Synthesis ofcomplement C4 in human proximal tubular epithelial cells (PTEC) isregulated by IFN-y. (abstract) Complement Inflamm 8:132, 1991

13. SACKS SH, ZHOU W, MARTIN J, CAMPBELL RD: Synthesis of the thirdand fourth components of human complement by glomerular podo-cytes and regulation by interferon-y. (abstract). J Am Soc Nephrol3:613, 1992

14, ROTHER K, HANSCH GM, RAUTERBERG EW: Complement in inflam-mation: induction of nephritides and progress to chronicity. mt ArchAlleigy App! Immunol 94:23—37, 1991

15. SHIBATA T, Costo FG, BIRMINGHAM DJ: Complement activationinduces the expression of decay accelerating factor on human mesang-ial cells. J Immunol 147:3901—3908, 1991

16. SCHONERMARK 5, RAUTERBERG EW, SHIN ML, LOi S, ROELCKE D,HANSCH GM: Homologous species restriction in lysis of humanerythrocytes: A membrane-derived protein with C8-binding capacityfunctions as an inhibitor. J Immunol 136:1772—1776, 1986

17. NICHOLSON-WELLER A, BURGE J, FEARON DT, WELLER PF, AUSTENKF: Isolation of a human erythrocyte membrane glycoprotein with

decay-accelerating activity for C3 convertases of the complementsystem. J Immunol 129:184—189, 1982

18. TEDESCO F, RONCELLI L, PETERSEN BH, AGNELLO V, DODETZ JM:Two distinct abnormalities in patients with C8a-y deficiency. Lowlevel of C8/3 chain and presence of disfunctional C8a-y subunit. J ClinInvest 86:884—888, 1990

19. GE5UALDO L, Di PAOLO 5, MILANI 5, PINZANI M, GRAPPONE C,RANIER! E, PANNARALE G, SCHENA FP: Expression of platelet-derivedgrowth factor receptors in normal and diseased human kidney. Animmunohistochemistiy and in situ hybridization study. J Clin Invest94:50—58, 1994

20. LAEMMLI UK: Cleavage of structural proteins during the assembly ofthe head of bacteriophage T4. Nature (London) 227:680—685, 1970

21. CHOMCZYNSKI P, SACCHI N: Single-step method of RNA isolation byacid guanidinium thiocyanate-phenol-chloroform extraction. AnalBiochem 162:156—159, 1987

22. DE BRUIJN MHL, FEY OH: Human complement C3: eDNA codingsequence and derived primary structure. Proc Natl Acad Sci USA82:708—712, 1985

23. KAASHOEK JGJ, VERHASSELT V, VAN DEN DOBBELSTEEN MEA, VANDER WOUDE FJ, VAN Es LA, DAHA MR: Effect of cytokines on theproduction of C3 and factor H by human mesangial cells (HMC).(abstract) Immunobiology 184:431, 1992

24. DINARELLO CA: Interleukin-1 and the pathogenesis of the acute-phase response. NEnglJMed 311:1413—1418, 1984

25. BOSWELL JM, Yui MA, ENDRES S, BURT DW, KELLEY VE: Novel andenhanced IL-i gene expression in autoimmune mice with lupus. JImmunol 141:118—124, 1988

26. BRENNAN DC, Yui MA, WUTHRICH RP, KELLEY VE: Tumor necrosisfactor and IL-i in New Zealand blacklwhite mice. Enhanced geneexpression and acceleration of renal injury. J Immunol 143:3470—3475,1989

27. ToMosuGI NI, CASHMAN SJ, HAY H, PUSEY CD, EVANS DJ, SHAw A,REES AJ: Modulation of antibody-mediated glomerular injury in vivoby bacterial lipopolysaccharide, tumor necrosis factor, and IL-i. JImmunol 142:3083—3090, 1989

28. CAMussi G, TEITA C, BussoLiNo F, Tiitw.o E, BRENTJENS 3, MoN-TRUCCHIO G, Ar'mis 0: Effect of leukocyte stimulation on rabbitimmune complex glomerulonephritis. Kidney Int 38:1047—1055, 1990

29. NOBLE B, REN K, TAVERNE J, DIPIRRO 3, VAN LIEw 3, DIJKSTRA C,JAN0SSY G, POULTER LW: Mononuclear cells in glomeruli andcytokines in urine reflect the severity of experimental proliferationimmune complex glomerulonephritis. Clin Exp Immunol 80:281—287,1990

30. TIPPING PG, Lowa MG, HOLDSWORTH SR: Glomerular interleukin 1production is dependent on macrophage infiltration in anti-GBMglomerulonephritis. Kidney mt 39:103—110, 1991

31. MONTINARO V, HEVEY K, AVENTAGGIATO L, FADDEN K, ESPARZA A,CHEN A, FINBLOOM DS, RIFA! A: Extrarenal cytokines modulate theglomerular response to IgA immune complexes. Kidney lot 42:341—353, 1992

32. HoRn Y, MURAGUCHI A, IWANO M, MATSUDA T, HIRAYAMA T,YAMADA H, FUJI! Y, DOHI K, I5HIKAWA H, OHMOTO Y, YOSHIZAKI K,HIRANO T, KI5HIM0T0 T: Involvement of IL-6 in mesangial prolifer-ative glomerulonephritis. J Immunol 143:3949—3955, 1989

33. RUEF C, BUDDE K, LACY J, NORTHEMANN W, BAUMANN M, STERZELRB, COLEMAN DL: Interleukin 6 is an autocrine growth factor formesangial cells. Kidney lot 38:249—257, 1990

34. MONTINARO V, Di CILLO M, PERI5SUTrI 5, SERRA L, TEDESCO F,RIFAI A, SCHENA FP: Modulation of renal production of C3 byproinflammatory cytokines. Kidney Int 43(Suppl 39):S37—S40, 1993

35. LIANOS EA, ZANGLIS A: Effects of complement activation on platelet-activating factor and eicosanoid synthesis in rat mesangial cells. J LabClin Med 120:459—464, 1992

36. Lis.i'os EA, ZANGLIS A: Glomerular platelet-activating factor levelsand origin in experimental glomerulonephritis. Kidney mt 37:763—770,1990

37. ORFILA C, PIERAGGI M-T, Suc J-M: Mesangial isolated C3 depositionin patients with recurrent or persistent hematuria. Lab Invest 43:1—8,1980

38. DO! T, KANATSU K, SUEHIRO F, NAGAI H, YOSHIDA H, HAMASHIMAY: Clinicopathological study of patients with mesangial isolated C3ddeposition in various glomerular diseases. Nephron 46:188—193, 1987