Eur. J. Immunol. 1992. 22: 425-431

Philippe Lassalleo, Christian LaGrou., Yves DelnesteO, Josiane Sanceau., Jean Coll., GQrardTorpierO, Jeanne WietzerbinA, Dominique Stehelin., AndrC-Bernard Tonnelo and Andre CapronO C.J.F. INSERM 90-06 and Centre d’lmmunologie et de Biologie ParasitaireO, Unit6 Mixte INSERM U 167 - CNRS 624, Service d’oncologie Moleculaire., Unit6 INSERM U 186-CNRS 1160, Institut Pasteur, Lille and Service de Recherche sur les Interf6ronsA, Unit6 INSERM U 196, Institut Curie, Paris

Characterization of SV40-transfected human endothelial cells

425

Human endothelial cells transfected by SV40 T antigens: characterization and potential use as a source of normal endothelial factors* A putative role for the vascular endothelium as target for autoantibodies has been suggested in several autoimmune disorders and connective-tissue diseases. However, there are some difficulties linked to the use of cultured endothelial cells (EC) that limit considerably the extensive studies on the nature of endothelial target antigens involved. To overcome this problem, human EC, derived from umbilical veins, were transfected with recombinant plasmid pSVl which contained the early genes of simian virus SV40. These transfected cells, called EC-pSV1, are able to grow without E C growth supplement and demonstrate a population doubling time of about 50 h. Among the E C properties, EC-pSV1 retain intracellular content of angiotensin-converting enzyme activity, exhibit constitutive production of interleukin 6 and of a growth-promoting activity on early passage EV, express intercellular adhesion molecule 1 (ICAM-1) and its up-regulation by tumor necrosis factor a , but have lost the expression of factor VIII-related antigen. Moreover, EC-pSV1 express a 55-kDa antigen found on E C and human platelets, and presumably acting as an antibody target in some cases of non-allergic asthma. However, at the SO-SSth generation, morphological changes and altered growth behavior were visible. This work demonstrates that transfection of E C with SV40 T antigens mav be of interest. oarticularlv in areas of research including the stud; of E d targets involved in different human diseases. I

,

1 Introduction A putative role for the vascular endothelium as target for autoantibodies has been suggested in several autoimmune disorders. This was supported by the demonstration that an antibody binding to cultured human endothelial cells (EC) is present in serum from patients with progressive systemic sclerosis [l], systemic lupus erythematosous [2, 31, rheumatoid arthritis [4], primary hypoparathyroidism [S], Kawasaki’s disease [6], hemolytic-uremic syndrome [7], episodic angioedema and hypereosinophilia [8], and in severe asthma (Lassalle F’., submitted for publication). In trying to characterize the molecular targets of these autoantibodies, it has been demonstrated that a S5-kDa7 a 200-kDa, a 120-kDa and cytokine-induced cell surface antigens were specifically recognized by autoantibodies in severe asthma (Lassalle P., submitted for publication) , primary hypoparathyroidism [ 5 ] ,allergic granulomatosis and angiitis [9] and Kawasaki’s disease, respectively [lo].

All these important observations have been established with human EC in culture, usually derived from umbilical vein. However, there is a lack of knowledge about the exact nature of these target antigens which may be explained, in part, by the inadequacy of such cultured E C to serve as a source of large amounts of antigens for biochemistry. Indeed, early passage E C cultures have a number of limitations such as the small quantities of recovered cells dependent of the surface area of umbilical vein, the special conditioned medium required for culture linked to their poor ability to proliferate in vivo, and an uncontrollable phenotypic variability. The significance of these anti-EC antibodies remains, therefore, to be established. To have a cell line that produce enough amounts of antigens for biochemistry, we have obtained stable SV40-transfected human E C (EC-pSV1). Here, we describe our current results about their characteristics.

2 Materials and methods 2.1 Recombinant DNA plasmid [I 95221

*

This study was supported in part by INSERM C.J.F. 90-06, INSERM U 167 and CNRS ERA 624.

Correspondence: Philippc Lassalle, C. J. f . INSERM 90-06 and Ccntre d’Immunologie et de Biologie Parasitaire, Unite Mixte INSERM U 167-CNRS 624, Institut Pasteur, 1 rue du Professeur Calmettc, F-59019 Lillc, Francc Abbreviations: FVIII-R Ag: Factor VIII-related antigen ACE: Angiotensin-convcrting enzyme EC: Endothelial cells ECGS: Endothelial cells growth supplement

0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1992

The recombinant plasmids pSVl and pSVLT were a generous gift of Dr. M. Rassoulzadegan (INSERMU273, France). The recombinant plasmid pSVl was constructed by Benoist and Chambon [ 1I]. Briefly, pSVl containing the SV40 early genes was constructed by inserting the large Hpa UBam HI fragment from SV40 (0.73-0.15 map unit) into the Eco RI site of pBR322 by blunt-end ligation after repair of the extremities with DNA polymerase I. Upon transfection on eucaryotic cells, pSVl is able to produce the two early genes large T and small t [12]. The experiments were all carried out with the plasmid pSVLT.

+

0014-298019210202-0425$3.50 .2510

426

Eur. J. Immunol. 1992. 22: 425-431

F! Lassalle, C. LaGrou,Y. Delneste et al.

2.2 EC culture EC were obtained from human umbilical veins, according to the methods previously described with some modifications [13]. Briefly, EC were collected from umbilical cords treated by 0.2% (v/v) collagenase in HBSS (M.A. Bioproducts, Walkerville, MD). EC were isolated, suspended at approximately 1 x lo5 - 2 x lo5 cells/ml in RPMI 1640, 2 mM L-glutamine, 100 U/ml penicillin, 100 pg/ml streptomycin, supplemented with 20% FCS (Gibco BRL, CergyPontoise, France), 100 pg/ml heparin and 25 pglml Endothelial Cell Growth Supplement (ECGS, Sigma Chemical Co, St. Louis, MO). EC were then cultured in 35-mm diameter tissue culture wells and maintained in a humidified atmosphere of 5% C02 in air at 37 "C. The medium was changed twice a week and the cultures reached confluency in 5 or 7 days. EC were then detached by incubation with 5 mM EDTA in PBS for 30 min at 37"C, centrifuged at 500 X g for 10 min, resuspended in fresh medium and further cultured in 35-mm diameter tissue culture wells.The culture reached confluency within 3 to 5 days. Only the second- to fourth-passage EC cultures were used in these studies. Cell morphology and detection of factor VIIIrelated antigen (FVIII-R Ag) by indirect immunofluorescence were used to check the purity of culture.

2.3 Transfection Human EC plated in 35-mm diameter petri dishes, were transfected by pSVl by the calcium phosphate precipitation technique [14], and then cultured in RPMI 1640 containing 20% FCS, 25 pg/ml ECGS, and 100 pg/ml heparin for 2 weeks. Untransfected and transfected cells were then cultured in RPMI containing 10% FCS for another 2 weeks. At this time, colony formation was detected only in plates with transfected cells. Then medium was changed twice a week. At confluency, transfected cells were detached using trypsin-EDTA solution and were subcultured in 1: 3 ratio in 100-mm diameter petri dishes. Cell aliquots from each passage were stored in liquid nitrogen. 2.4 Cloning systems The efficiency of colony formation was determined by two methods: (a) EC-pSV1 were seeded at a density of 1 x lo3 cells per 60-mm petri dishes in medium containing 20% FCS, 30% EC-pSV1-conditioned medium, and 0.36% soft agar (Difco Laboratories, Detroit, MI) and observed for 1 month for colony formation; (b) EC-pSV1 were seeded at a limiting dilution of 0.5 cell/well in 96-well flat-bottom culture plates in medium containing 20% FCS and 30% EC-pSV1 conditioned medium.

2.5 Assay for cell proliferation Cell proliferation was performed by measuring [3H]thymidine incorporation and by a colorimetric assay for cellular growth based on the reduction of a tetrazolium salt 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide (MTT) which is strictly dependent of the number of viable cells, as previously described by Mosmann [15]. Early passage EC or EC-pSV1 were plated at lo4or 2 x lo4

cells/well on gelatin-coated 24-well culture plates in RPMI without FCS for 24 h at 37 "C. The medium was then removed and replaced by fresh RPMI, supplemented with FCS (1% to 20%0), or ECGS (25 pg/ml), or heparin (100 pglml), or ECGS plus heparin, or 20% EC- or EC-pSV1-conditioned medium. After 3 days, a 5-mg/ml solution of MTT in PBS was added at 1: 10 (v/v). After 4 h of incubation, the medium was removed and 300 pl of 0.04 N HC1-isopropanol solution was added to each well. After homogenization, a 2 0 0 4 aliquot of each well was transferred to a microtiter plate and read in an ELISA reader at 620 nm. All analysis were performed in duplicate. For [3H]thymidine incorporation, the procedure was identical except that lo3 or 2 x lo3 cells/well were plated in 96-well culture plates and that cells were incubated with the different additives for 24 h before [3H]thymidine incorporation. All analysis were performed in triplicate.

2.6 Assay of angiotensin-converting enzyme (ACE) activity ACE activity was evaluated by the method of Cushman et al. [16], modified by Lieberman [17] on non-ionic detergent cell lysates. Briefly, hippuryl-L-histidyl-L-leucine was employed as a substrate which measures the formation of free hippuric acid by the action of ACE. This end product was then extracted and quantitated on a spectrophotometer at 228 nm as a measure of enzyme activity. The results were expressed in unit/mg of total proteins contained in the sample. One unit of ACE activity is defined as nanomol of hippuric acid formed per min at 37°C under standard conditions.

2.7 Assay of IL 6 activity IL 6 was assayed described by Van Snick et al. 1181 by incubating IL 6-dependent hybridoma cells (mouse-mouse hybrid 7TD. 1) with serial dilutions of cell SN. After 4 days of incubation, the number of cells was evaluated by colorimetric determination of hexosaminidase levels. A dilution of recombinant IL 6 (1000 U/ml), produced by Dr. J. Wietzerbin (INSERM U 196, Paris, France), was used as an internal standard in all IL 6 assays.

2.8 Western blot analysis

Samples of EC were lysed in ice-cold buffer (10 mM Tris-HCI, pH 7.4, 150 mM NaCl, 10 mM MgC12 and 0.5% Nonidet P-40). The extracts were fractionated on a 10% polyacrylamide slab gel containing 0.1% SDS under reducing conditions according to the techniques of Laemmli [19]. Proteins were transfered onto nitrocellulose paper for Western blot analysis. The nitrocellulose strips were incubated in a quenching buffer (5% non-fat milk in PBS) for 30 min, followed by an overnight incubation at 4 "C with constant agitation in 1 : 100 diluted patient and control sera. After washing four times with PBS, bound antibody was detected by a 2-h incubation with horseradish peroxidase (HRP)-conjugated goat anti-human IgG (Tago, Burlingame, CA). The nitrocellulose strips were then rewashed, and developed with HRP color development

Eur. J. Immunol. 1992. 22: 425-431

Characterization of SV40-transfected human endothelial cells

reagent (Bio-Rad Laboratories, Richmond, CA). The similar procedure was used with human platelet extracts. 2.9 EC E L S A

The ELISA was performed on glutaraldehyde-fixed E C using the method described by Pober et al. [20], slightly modified. E C were subcultured at the second passage on gelatin-coated 96-microwell culture plates (Falcon, Oxnard, CA). Experiments were started at cell confluency. Recombinant human TNF-a (200 U/ml), and IFN-y, 1000U/ml were added for 24 h in E C cultures (these recombinant cytokines were kindly provided from Dr. J. Wietzerbin) EC were then washed twice with PBS, fixed for 10 min at 4 "C with 0.5% glutaraldehyde in PBS,washed four times with PBS containing lo-' M EDTA and 0.1% bovine serum albumin, and incubated for 1 h in this same medium. Fixed EC were then incubated for 1 h with 100 yl/well of the IgGl monoclonal antibody to ICAM-1 (Immunotech, Marseille Luminy, France) or the control IgGl monoclonal antibodies (a generous gift of Dr. J. Khalife, U 167, Lille, France) at the experimentally defined dose of 0.2 pg/ml. Cells were washed again four times, and incubated for another hour with 100 p1 of 1 : 5000 (v/v) diluted peroxidase-labeled anti-mouse IgG rabbit antiserum (Institut Pasteur Production, Courbevoie, France). After four additional washings, 100 yl of carbonate/hydrogen carbonate buffer containing H202 and o-phenylenediamine was added for 30 min, after which the reaction was stopped with 100 yl of 2 N H2SO4. Absorbance was read in a multiwell scanning spectrophotometer at 492 nm. All analyses were performed in duplicate.

ent that lead progressively to the loss of the cell line similarly to what has previously been described [21, 221. EC-pSV1 are unable to grow in a semi-solid medium, however, some clones were obtained using limiting dilution technique. Each clone showed an epithelioid morphology, but when the 50thgeneration was reached, and as described above, growth and morphological alterations were observed with each of these clones. Clones could not be obtained with normal E C which were carried out in parallel under limiting dilution. The growth properties of EC-pSV1 appeared to be independent of ECGS, heparin or a combination of these (Fig. l A ) , but dependent on FCS, and cell growth was detected for quantities of FCS as low as 1% (Fig. 1B). 3.2 Morphology At the sites of close contact of the bottom cell membrane with the substrate, typical dense plaques (Fig. 2C) of fibrillar material are present both in normal E C and in

-

I 2.10 Electron microscopy

427

0.0

'

I

-

I

1%

I

5%

Medium alone +Heparin

+ Heparin/ECGS

I

10%

I I

15%

I

20%

% of fetal calf serum added to the medium

EC and ECpSV1 were grown in 100-mm petri dishes, rinsed twice with cold HBSS and then fixed for 30 min with 1% glutaraldehyde in cacodylate buffer at 4 "C. The cells were washed with PBS. The cells were post-fixed in 1% osmium tetroxide for 60 min, dehydrated in acetone and embedded in araldite. Sections perpendicular to the mono layer culture cells, contrasted with lead hydroxide, were examined with a Philips EM420 electron microscope.

I

--tY

SV40-transfectedEndo. cells Untransfected Endo. cells

I

3 Results 3.1 Growth properties of ECpSV1 Control EC cultures did not survive more than a few number of passages in the absence of ECGS. After transfection, the majority of EC gradually degenerated. Only some of them survived and began to proliferate in the absence of ECGS. Transfected E C (ECpSV1) have continued to grow for more than 35 passages when diluted threefold at each passage. The cell doubling time was calculated to be close to 50 h in the presence of 10% FCS. An indirect immunofluorescence test revealed the presence of the large T antigen in the nuclei of EC-pSV1 with similarly to what is seen in the COS-1 cell line which served as control (data not shown). At a time corresponding to the 50-S5th cell generation, morphological changes with giant cell formation and an altered growth behavior were appar-

-.

0%

1%

5%

10%

15%

20%

% of fetal calf serum added to the medium

Figure I. Growth properties of EC-pSV1. Data are expressed as the mean k SEM of three separate experiments. (A) ECGS independence. ECGS and heparin (25 pgiml and 100 pglml, respectively) were incubated with various concentrations of FCS (0% t o 20%) on E C cultures and assayed for cell growth using the colorimetric assay (see Sect. 2.5).There is no significant difference between these different culture conditions (p > 0.05 Student's t-test). (B) FCS dependence of EC-pSV1 growth. Data demonstrated a significative effect as low as 1% of FCS (p < 0.01, Student's t-test). Following these results, 10% FCS was chosen as the standard FCS concentration for subcultures.

428

P. Lassalle, C. LaGrou,Y. Delneste et al.

Eur. J. Immunol. 1992.22: 425-431

Figure 2. Sections perpendicular to the monolayers of cul-

tured normal and transfected human EC. (A) Normal EC with Weibel-Palade bodies (arrows) surrounded by intermediate filaments (arrowheads). (B) Transfected EC showing interdigitations complex formed by the peripheral cytoplasm. Transfected EC with a typical dense plaque of fibrillar material (arrow).

EC-pSV1. Their basal location and their orientation in ECpSV1 indicate evidence of a certain degree of polarity. The EC-pSV1 revealed a complex pattern of cytoplasmic interdigitations with a certain loss of cell contact inhibition. (Fig. 2B). Cytoplasmic extensions sometimes enclosed a space with the adjacent cells. This bridging which occurs in cultured endothelium, was found with EC-pSV1. The relative richness in membrane-bound organelles, vesicular and tubular membranous structures, and free ribosomes of E C were equally observed in EC-pSV1. A unique organelle, the Weibel-Palade bodies, usually surrounded by intermediate filaments (Fig. 2A) are found exclusively in normal EC, but not in EC-pSV1.

3.3 Endothelial characteristics of ECpSV1

3.3.1 Presence of ACE and absence of FVIII-R Ag ACE is known to be produced essentially by megakaryocytes, macrophages and EC, and the large amounts of ACE found in EC are now considered as one of their identity markers. The detection of ACE activity was performed on the supernatants of non-anionic detergent lysates of E C and EC-pSV1. E C and EC-pSV1 were found to contain 2.3 f 0.6 U/mg and 2.7 f 0.9 U/mg, respectively. By contrast, the search for another endothelial marker such as FVIII-R Ag in EC-pSV1 by indirect immunofluorescence staining using both polyclonal and monoclonal antibodies provided from Behring Hoechst (Rueil-Malmaison, France) and Immunotech, respectively, was negative (data not shown). This can be related to the absence of WeibelPalade bodies on electron micrographs of EC-pSV1.

3.3.2 Production of an EC growth activity Medium conditioned by EC-pSV1 was tested for its ability to stimulate growth in early passage EC. Our results have

shown that supernatants from EC-pSV1 induced both [3H]thymidine incorporation (Fig. 3A) and increase of cell number, as determined by the colorimetric assay. This growth activity was dose dependent and appeared to be maximal at a 1: 5 (v/v) dilution of 72-h-cultured EC-pSVl supernatants. No such activity was detected either in EC-conditioned medium or in EC-pSVl fresh medium. The presence of this activity in EC-pSV1 supernatants appeared to be time dependent, with increasing activity after 24 h to 72 h of culture with EC-pSV1 (Fig. 3B). The granulocyte-macrophage colony stimulating factor (GM-CSF), released by E C upon various stimulations [23], is also known to exhibit growth activity on them [24]. The idea that a constitutive production of GM-CSF by EC-pSV1 might account for E C growth activity has already been tested. Co-incubation of 10 pg/well of neutralizing antiGM-CSF (1 pg inhibits 50 units of GM-CSF bioactivity) or equal amounts of unrelated IgGl murine monoclonal antibodies with EC-pSV1 supernatants on EC had no effect on E C growth promotion. So, these results do not agree with such a role. 3.3.3 Spontaneous release of IL 6

Previous reports have established that human umbilical vein E C synthesize IL 6, and that production of this cytokine can be augmented by IL 1, TNF-a, and bacterial lipopolysaccharide [25,26]. Here, we detected in the supernatants of SV40-transfected E C high amounts of IL 6 activity. This IL 6 production appeared to be time dependent increasing release from 24 h to 72 h of culture (Fig. 4). A strongly expressed IL 6 mRNA was found migrating as a single band of 1.4 kb similar in size to that found in human fibroblastic cell lines. The IL 6 molecule was specifically recovered in supernatants from SV40-transfected EC with a size of 27 kDa and 25 kDa. Thus, these results argue for a higher constitutive production rate of IL 6 in EC-pSVl than in EC.

Eur. J. Immunol. 1992. 22: 425-431

Characterization of SV40-transfected human endothelial cells

429

-0- Ecconditionnedmedium

*

Reshmedium

I

24 hours

48 hours

72 hours

Tlme culture of recovered supernatants

!

200

I

I

I

10%

20%

I

30%

.

% of conditionned medium added in cultures

Figure 4 . Kinetics of IL 6 release by EC and EC-pSV1. IL 6 was assayed by incubating IL 6-dependent hybridoma cells 7TD.1 with serial dilution of supernatants. Supernatants from E C and ECpSV1 cultures were recovered, centrifuged (500 x g, 10 min), and stored at - 20 "C until use. Data are mean f SD of three separate experiments.

--t. EC-pSVl conditionnedmedium

+ ECconditionnedmedim 0.4 0.5

N 0

d

03

p

Human endothelial cells transfected by SV40 T antigens: characterization and potential use as a source of normal endothelial factors.

A putative role for the vascular endothelium as target for autoantibodies has been suggested in several autoimmune disorders and connective-tissue dis...
825KB Sizes 0 Downloads 0 Views