Dexamethasone modulates complement secretion by monocytes

Eur. J. Immunol. 1992. 22: 909-915

Claudie LemercierA, Nathalie Julenvo, Muriel CoulpierA, H61;ne DaucheP, Denyse OzanneA, Marc FontaineA and Jean RipocheA INSERM Unit6 7gA, Bois-Guillaume and Institut fur Hygienev, Innsbruck

Differential modulation by glucocorticoids of alternative complement protein secretion in cells of the monocyte/macrophage lineage* The effect of the synthetic glucocorticoid dexamethasone (DXM) on the secretion by human monocytes of alternative complement proteins C3, factor B and factor H was investigated. Results indicated that DXM modulates this secretion in a direction which would be consistent with its anti-inflammatory properties. DXM, at therapeutic concentrations, had a suppressive effect on C3 and factor B secretion and a stimulatory effect on factor H secretion by monocytes. This differential modulation on C3, factor B and factor H secretion was similar in mature macrophages. Together with previous studies showing that DXM had a suppressive effect on C3 and factor B secretion and a stimulatory effect on factor H secretion by human endothelial cells, our results indicate that DXM appears to have the general property of regulating local production of complement components so as to control complement activation.

1 Introduction The complement system occupies a central place in immunity. I t is essential t o the defense against infection, the handling of immune complexes and probably to the regulation of the immune response [l-31. There has been interest in the local production of complement proteins mainly because this local production may represent the primary source of complement components in the first stages of inflammation before the increase in vascular permeability. An abnormality in the local production of complement proteins may be involved in the pathophysiology of inflammatory disorders [4]. Elucidating the regulation of this local secretion of complement proteins appears, therefore, to be crucial to the understanding of the molecular basis of the inflammatory reaction. Monocytes and mature macrophages have been extensively studied as a source of local complement protein secretion. They have been shown to secrete C1 subcomponents, C4, C2, C3, C5 and all the components of the alternative pathway and of the membrane-attack complex.They also express a number of regulatory proteins, factor H, factor I, C4bp, and several membrane-bound complement regulatory proteins CR1, membrane cofactor protein and decay-accelerating factor (MCP and DAF; see [5-111). In view of the anti-inflammatory action of glucocorticoids, we have analyzed the effect of a synthetic glucocorticoid, dexamethasone (DXM), on the secretion of the alternative complement proteins C3, factor B and factor H by monocytes and in vitro derived macrophages. Results presented

[I 96321 ::

909

This work supported by INSERM, The Fondation pour la Recherche MCdicale and the University of Rouen. N. Julen is a fellow of the Association de Recherche sur la Polyarthrite.

Correspondence: Claudie Lemercier, INSERM Unite 78, B.P.73, F-76233 Bois-Guillaumc Cedex, France Abbreviations: DXM: Dexamethasone MNC: Mononuclear cells

0 VCH Verlagsgesellschaft mbH. D-6940 Weinheim, 1992

in this report indicate that DXM modulates this secretion in a direction which would decrease complement activation and, therefore, is consistent with the anti-inflammatory properties of this synthetic glucocorticoid. The secretion of C3, factor B and factor H by the undifferentiated monocytic cell lines U-937 and THPl and the modulation of this secretion by DXM were also examined.

2 Materials and methods 2.1 Cytokine preparations and other reagents Human rIFN-y was a gift from Roussel-Uclaf, Paris, France. DXM, sodium phosphate salt (sterile apyrogenic solution for human therapeutic use) was from Merck, Sharp and Dohme-Chibret, Paris, France. RPMI 1640, fetal calf serum and antibiotics were purchased from BoehringerMannheim (Mannheim, FRG). Endotoxin-free Dulbecco’s PBS and the phosphatase substrate p-nitrophenyl phosphate disodium salt were purchased from Sigma (St. Louis, MO). Lymphoprep (e = 1.077, 300 mOsm) was from Nycomed, Oslo, Norway. Immobilized protein G was from Pierce, Beijerland,The Netherlands, and Triton X-100 was from Aldrich-Chemie, Steinheim, FRG. 2.2 Proteins and antisera Factor H and C3 were purified from human plasma as described previously [12]. An additional purification step was performed for factor H by HPLC on a Spherogel TSK column (phenyl5TW, Beckman, Palo Alto, CA). Factor B was purified from human plasma as described [ 131. Purity of the proteins was assessed by SDS-PAGE and Coomassie blue staining. Polyspecific and monospecific polyclonal anti-factor H and anti-C3 antibodies were prepared as described [14]. Monospecific anti-factor B antibodies were from Atlantic Antibodies, Scarborough, ME. Antibodies were conjugated with alkaline phosphatase (Sigma) by a one-step procedure as described [ 151. Monoclonal MRC OX23 and MRC OX24 anti-factor H antibodies were a kind gift from Dr. R. B. Sim, Immunochemistry Unit, Department of Biochemistry, Oxford University. 0014-2980192/0404-0909$3S O

+ ,2510

910

C. Lernercier, N. Julen, M. Coulpier et a1

Eur. J. Immunol. 1992. 22: 909-915

2.3 Cells

2.6 Northern blot analysis

The human histiomonocytic cell line U-937 was from ATCC (Rockville, MD). The monoblastic cell lineTHPl was a gift from Prof. M. G. Colomb (INSERM U238, Grenoble, France). U-937 and THPl cells were cultivated in RPMI 1640 medium supplemented with 10% FCS (decomplemented 30 min at 56"C), L-glutamine (2 mM final concentration), 100 IU/ml penicillin and 100 pglml streptomycin, this being referred to as standard medium. Cells were grown at 37°C in a 5% C02/95% air atmosphere.

FWA extraction and electrophoresis were performed as described [17]. Double-stranded cDNA probes were as follows: B38.1 encoding the first N-terminal third of factor H and recognizing, in addition to the 4.3-kb fulllength message, a 1.8-kb message encoding a truncated form of factor H containing the N-terminal region of the molecule [HI, pC3.11 encoding the C-terminal (900/) of human C3 [19], a gift from G. Fey (Research Institute of Scripps Clinic, La Jolla, CA), factor B probe, pFB-3b, encoding the near complete sequence of the Ba fragment of factor B, a gift from Duncan Campbell (MRC Immunochemistry Unit, Oxford, GB). Probes were labeled with 32P by random priming (Boehringer-Mannheim). Autoradiographs were scanned using a Gelman DCD16 densitometer (Aubervilliers, France).

Human monocytes were prepared from the discarded buffy coats of blood from healthy donors (Regional Blood Transfusion Center, Bois-Guillaume, France). Buffy coats were first depleted in erythrocytes by centrifugation at 600 x g for 10 min. The mononuclear cells (MNC) were then isolated by density gradient on a Lymphoprep cushion. After centrifugation (800 X g, 15 min, 20"C), the MNC were collected and washed four times at low gravity (200 x g, 10 min, 20°C) with PBS to remove platelets. MNC were resuspended at a final density of 7 X lo6 celldm1 in standard medium. The monocytes were purified by adherence to tissue culture plastic (Falcon flasks 75 cm2 or 6-well tissue culture dishes, Costar, Cambridge, MA) in the presence of 10% FCS for 1h at 37°C followed by three vigorous washes with PBS to remove the nonadherent cells. The monocytes were characterized by May-Griinwald Giemsa staining and peroxidase, specific and nonspecific esterase staining using standard protocols. In vitro derived macrophages were obtained by allowing monocytes to differentiate in vitro for 10 days in the flasks with medium changes every 3 days. In some experiments, monocytes were cultivated in Teflon flasks to prevent adherence [16]; to this end, the cells were gently detached after the adhesion step by scraping in ice-cold PBS, pooled, washed twice with PBS at 200 X g, checked for their viability by Trypan blue exclusion and resuspended in standard medium in Teflon flasks where they were grown for the indicated time. 2.4 Experimental procedures

Cells were incubated in standard medium containing the appropriate stimulus during 24, 48 or 72 h. Working concentrations were 200IU/ml for IFN-y and lop6 to lopgM for DXM. At indicated time intervals, the supernatants were decanted, added with proteinase inhibitors [5 mM aprotinin, 50 pg/ml soybean trypsin inhibitor (SBTI), 1 pg/ml pepstatin A , 2 mM PMSF final concentrations (all from Sigma)] and frozen immediately at -70 "C until being assayed. Monocytes and macrophages were counted after detaching the cells by gentle scraping in ice-cold PBS. 2.5 ELISA

Assays were performed exactly as described previously [14]. For each ELISA, a standard curve was established using dilutions of a pool of 200 normal human sera and a known amount of the appropriate purified protein. Concentrations were determined from the standard curves. Results were analyzed using the paired two-tailed t-test.

2.7 Cell surface labeling and immunoprecipitation

Monocytes were prepared as described in the previous section. Twenty million cells were resuspended in 20 ml of standard medium in Teflon flask and incubated for 72 h in the presence of the appropriate stimulus. At the end of the incubation period, cells were washed three times with PBS, checked for their viability by Trypan blue exclusion and surface labeled with 1 mCi = 37 MBq lZ5I(Amersham Int., Amersham, GB) by the Iodogen method, as described [20]. The surface-labeled cells were washed twice with ice-cold PBS containing 5% (v/v) FCS, layered on ice-cold FCS and centrifuged for 10 min at 400 x g. The cell pellet was then lysed for 20 min on ice withTriton X-100,1.5% (v/v) in PBS containing 5 mM PMSE 0.32 IU aprotinin, 1 pg/ml pepstatin and 50 yg/ml SBTI, final concentrations. The lysate was cleared from insoluble material by centrifugation for 20 min at 10 000 x g at 4"C, and the clear lysate was precleared by an overnight incubation with 100 pl, packed volume, of protein G (Pierce, Cheshire, GB) under agitation at 4°C. Protein G was removed by centrifugation and the precleared lysate frozen at -70°C until further use. Immunoprecipitations were performed by adding 50 pg of the indicated antibody to 500 p1 of precleared lysate followed by an overnight incubation at 4°C under gentle agitation. At the end of the incubation period, 100 pl of packed protein G-Sepharose was added and incubation carried on for 2 h at 4°C. The beads were pelleted down, and washed five times in lysis buffer containing 0.5 M NaCl and 0.1% (w/v) SDS at room temperature. The bound antigens were eluted by boiling for 4 min in denaturing buffer [8 Murea, 2% (w/v) SDS, 0.2 M NaHZP04, pH 71 and analyzed by SDS-PAGE according to Laemmli [21].

2.8 Western blot analysis Approximately 20 x lo6 monocytes were prepared as described in the previous section. After the adhesion step, the cells were washed three times with PBS and incubated for 72 h in the presence of the indicated stimulus. At the end of the incubation period, the supernatants were collected and proteinase inhibitors were added (PMSF 5 mM, SBTI 50 pg/ml, pepstatin A 1 pg/ml, leupeptin 1 pg/ml, final concentrations). For Western blot analysis, 10-20 ml of the supernatants and 25 yl of an equal mixture (v/v) of

Dexamethasone modulates complement secretion by monocytes

Eur. J. Immunol. 1992. 22:909-915

MRCOX23 and MRCOX24 ascites fluid were incubated overnight at 4 "C.The immune complexes were isolated by chromatography on a column of Avid-AL (Bioprobe International Inc., Tustin, CA) according to the manufacturer's instructions. The isolated antigen-antibody complexes were concentrated, boiled for 4 min in denaturing buffer [ 0 . 2 M Tris, 8 M urea, 2% (wh) SDS, pH8.01, subjected to SDS-PAGE according to Laemmli [21] and transferred onto Immobilon P (Millipore, Bedford, MA) as described [22]. Blots were developed using monospecific rabbit polyclonal anti-factor H antibody and peroxidaseconjugated second antibody using standard procedures [23]. Blocking buffer was 5% dried milk in PBS.

3 Results 3.1 Effect of DXM on alternative complement protein biosynthesis by human monocytes: the association of IFN-y and DXM stimulates factor H secretion

The modulatory role of the synthetic glucocorticoid DXM on the biosynthesis of complement control protein factor H by monocytes was first examined. Monocytes were incubated for 72 h in standard medium containing IFN-y or DXM, or both stimuli. The secretion of factorH was quantitated by ELISA on cell supernatants as described in Sect. 2.5. In agreement with previous studies [S] our results show that there is a basal secretion of factor H by monocytes (about 5 ngi106cellsi72 h) which is quantitatively not important compared for instance with the basal

911

secretion of factor H by human endothelial cells [14]. IFN-y, used at 200 IU/ml, had a consistent but moderate stimulating effect (p < 0.2, not significant) on factor H secretion (Fig. 1A). DXM, at a concentration of M, increased significantly 0, < 0.002) factor H production by monocytes. The effects of IFN-y and DXM appear to be additive 0, < O.O2).This DXM + IFN-y-induced increase of factor H secretion by monocytes could also be evidenced using Western blot analysis with monoclonal anti-factor H antibodies (Fig. 1B). Fig. 1B also shows that factorH secreted by monocytes has the same M, as factor H purified from plasma. Immunoprecipitation of surface-labeled cell lysates with a monospecific polyclonal anti-factor H antibody and subsequent analysis on SDS-PAGE under nonreducing conditions showed the presence of a 155-kDa species co-migrating with control purified factor H. The intensity of this band was strongly reinforced (Fig., 1C) in IFN-y + DXM-stimulated monocytes. Another 70-kDa species whose expression was also strongly reinforced upon stimulation with IFN-y and DXM, could be observed in these experiments. This material could also be detected after precipitation of the labeled cell lysates with irrelevant antibodies (not shown). On the basis that this 70-kDa species was precipitated by antibody and its expression reinforced upon stimulation by IFN-y + DXM, we assumed that it was the monocyte Fcy receptor. Fcy receptor expression on monocytes has been shown to be strongly increased by stimulation with the association IFN-y + DXM [24]. Finally, in Northern blot experiments, we were not able to detect factor H-specific messages in monocytes using 5 to 15 pg of total RNA from control or DXM + IFN-y-stimulated cells.

kDa

-1 55 kDa - 70

- 155

1

2

3

4

1

2

3

1

2

Figure I . (A) Secretion of complement factor H by monocytes. Effects of IFN-y and DXM. Monocytes were stimulated with the indicated stimuli during 72 h and the concentration of complement factor H measured by ELISA as described in Sect. 2.5.Values are the mean k SEM for seven independent cultures. Final concentrations were 200 IU/ml for IFN-y and M for DXM.Track 1: control cells, track 2: IFN-y-stimulated cells, track 3: DXM-stimulated cells, track 4: IFN-y plus DXM-stimulated cells. (B) Secretion of complement factor H by monocytes. Westcrn blot analysis. Monocytes were stimulated for 72 h with both DXM and IFN-y, lop6M and 200 IU/ml, rcspectively, final concentrations. Cell supernatants of control and stimulated cells were analyzed for the presence of complement factor H by Western blotting as dcscribed in Sect. 2.8. Track 1: control cells, track 2: DXM plus IFN-y-stimulated cells, track 3: control purified factor H (2.50 ng). (C) Expression of complement factor H by monocytes. Cell surface labeling experiments. Monocytes were grown for 72 h in Teflon flasks in the presence or absence of both DXM and IFN-y M and 200 IU/ml final concentrations, respectively). The cells werc labeled with 1251and cell extracts analyzed for the presence of factor H by immunoprecipitation and SDS-PAGE under nonreducing conditions, as described in Sect. 2.8. Track 1: control cells, track 2: IFN-y plus DXM-stimulated cells.

012

C. Leniercicr, N. J u k u , M. Coulpier et al.

Eur. J. Immunol. 1992. 22: 909-915

3.2 DXM has a suppressive effect on C3 and factor B biosynthesis by human monocytes The effects of IFN-y or DXM or both stimuli on C3 and factor B secretion by human monocytes were next investigated (Fig. 2A). C3 and factor B are constitutively secreted by monocytes in agreement with previous studies [S].When monocytes were stimulated during 72 h by IFN-y at a concentration of 200 IU/ml, there was a differential effect on C3 and factor B secretion. IFN-y had a suppressive effect on the secretion of C3 (p < 0.05). In contrast, IFN-y had a stimulatory effect (p < 0.004) on factor B secretion by monocytes. These results are in agreement with recent studies by D. Lappin et al. [25]. When monocytes were cultivated in the presence of DXM at a concentration of lop6M, there was a consistent inhibition of C3 and factor B secretion 0) < 0.05). When cells are grown in the presence of both DXM and IFN-y, we observed that the inhibitory effects of IFN-y and DXM were additive on C3 secretion and that the inhibitory effect of DXM counteracted the stimulatory effect of IFN-y on factor B secretion $f < 0.01 vs. IFN-y alone). Dose-dependent experiments indicated that concentration of lop7was optimal for the suppressive effect of DXM on C3 secretion by human monocytes (not shown). Kinetic Northern blot experiments were performed in which monocytes were cultured in the presence of TFN-y, DXM, or both stimuli, and total RNA was prepared at 24,48 and 72 h. Results showed that DXM had

a marked suppressive effect on factor B mRNA expression (Fig. 2B). However, this effect was only apparent after 48 h of stimulation. At 24 h, there was, instead, threefold increase in factor B mRNA. After 48 h of stimulation with DXM, the level of factor B mRNA started to decline with a sharp fall at 72 h (Fig. 2B). IFN-y alone had a stimulatory effect on factor B mRNA expression and DXM moderately counteracted the IFN-y-induced increase in factor B mRNA. DXM had little effect on the expression of C3 mRNA (Fig. 2B). One paradoxal result was the enhancing effect of the association DXM and IFN-y which increased C3 mRNA level by more than twofold relatively to control after 72 h of stimulation. This effect was apparent after 48 h of stimulation (not shown) and was in striking contrast with the net effect of the association DXM+IFN-y on C3 protein secretion (Fig. 2A). Long-term stimulation of monocytes

24h 1

2

3

72h 4

1

2

3

A B

100

C

50

n 2

3

4

1

2

3

4

Figurc2. (A) Secretion of complement C3 and factor B by monocytcs. Effects of IFN-y and DXM. Monocytes were stimulated with the indicated stimuli during 72 h and the concentrations of complement C3 and factor B measured by ELISA as described in Sect. 2.5. Values are the mean k SEM for four independent experiments. Working final concentrations were 200 IU/ml for IFN-y and lo-" M for DXM. Track 1: control cells, track 2: IFN-y-stimulated cells, track 3: DXM-stimulated cells, track 4: IFN-y plus DXM-stimulated cells. (B) Differential modulation of C3 and factor B by IFN-y and DXM. Northern blot studies. Monocytes were stimulated with I F N y (200 IU/ml) or DXM (lo-" M) or both stimuli, and total RNA prepared as described in Sect. 2.6. Total RNA ( 5 pg for each track) was run under denaturing conditions, the gel was stained with ethidium bromide and blotted onto nylon. The filters werc hybridized with a C3-specific cDNA probe (panel A) or factor B-specific cDNA probe (panel B). In panel C are shown the ethidium bromide-stained gels and, in panel D, the results of the densitometric analysis of the autoradiographies. Track 1: control cells, track 2: IFN-y-stimulated cells, track 3: DXM-stimulated cells, track 4: IFN-y plus DXM-stimulated cells.

D

C3 72h

C3 24h 1.5 r

z

E

1.o

0.5

cc

0 a,

u c

0.0 - ._

1 2 3 4

1 2 3 4

f B 24h

fB 72h

RI -U

c 1 n U

a,

> -4

5r

1.5

r

1 .o

RI

c

a,

a

0.5 0 .o

1 2 3 4

1 2 3 4

4

Dexamethasone modulates complement secretion by monocytes

Eur. J. Immunol. 1992. 22: 909-915

by IFN-y indicated that, after 72 h of stimulation, IFN-y increased the level of C3 mRNA whereas it had a suppressive effect upon short-term stimulation (Fig. 2B). This long-term enhancing effect of IFN-y was again in contrast with the observed diminution in C3 protein secretion (Fig. 2A). All these experiments were performed three times with monocytes coming from different donors.

r

913

6

N

h \

0

8 4

-

'g

3.3 The modulation of C3, factor B and factor H biosynthesis by DXM is not influenced by the in vitro maturation stage of monocytes To analyze whether DXM modulated complement alternative protein secretion by monocytes at different maturation stages, we performed similar studies with in vitro derived macrophages. I n vitro derived macrophages prepared as described in Sect. 2.3. were cultured for 72 h in the presence of IFN-y, DXM, or both stimuli, and the secretion of C3, factor B and factor H was quantitated by ELISA. Results showed that the modulatory effects of DXM on C3 and factor B secretion were similar for differentiated macrophages and monocytes. As shown in Fig. 3, C3 and factor B secretions were consistently diminished when the cells were cultured in the presence of DXM at a concentration of lop6M (p < 0.01 and p < 0.02, respectively). IFN-y alone had a suppressive effect on the C3 secretion by mature macrophages (p < 0.03),as for monocytes, and growing the cells in the presence of both DXM and IFN-y resulted in a marked decrease in C3 secretion. DXM also counteracted the observed stimulatory effect of IFN-y on factor B production by mature macrophages (p < 0.1 vs. IFN-y alone). In contrast, there was a moderate stimulatory effect of DXM on factor H secretion by mature macrophages and a synergistic effect between IFN-y and DXM to increase factor H secretion (Fig. 3). 3.4 Study of the modulation of C3 and factor H biosynthesis by DXM and IFN-y in the immature monocytic cell lines U-937 and THPl

We next investigated the immature monocytic cell lines U-937 and THP1. These cells undergo maturation and

1

-

m

=

2

0 1

2

3

4

Figure 4. Secretion of C3 by the U-937 and THPl monocytic cell lines. Modulation by IFN-y and DXM. U-937 and THPl cells were grown as described in Sect. 2.3 and stimulated for 72 h with IFN-y M) or both stimuli. The secretion of C3 (200 IU/ml), DXM was quantitated by ELISA.Values are the mean SEM (n = 6 for U-937 and n = 4 for THP1). Track 1: control cells, track 2: IFN-y-stimulated cells, track 3: DXM-stimulated cells, track 4: I F N y plus DXM-stimulated cells.

*

functional differentiation into macrophage-like cells in the presence of PMA or lymphokines [26, 271. U-937 were found to secrete factor H (1-3 ng/106 cells/72 h), C3 (2-4 ng/lO6 cells172 h) in agreement with other reports [28-30]. No modulation of factor H secretion could be found either with IF N y or DXM (not shown). In contrast to what is observed with monocytes, C3 secretion was increased after 72 h stimulation with IFN-y and DXM IFN-y had a synergistic effect resulting in a fivefold increase in C3 secretion over control (p < 0.01) (Fig. 4A). No factor B secretion by U-937 could be detected. This could mean a very low level of secretion by these cells ~311.

+

The THPl monocytic cell line constitutively secretes factor H and C3. Factor H secretion is quantitatively of low importance (1-3 ng/106 cells/72 h) and no modulation of this secretion by DXM could be seen (not shown). As for U-937 and in contrast to what is seen with more mature cell monocytes and macrophages C3 secretion is increased by IF N y (p