thank Drs. Kevin Sullivan and Sandra. Christiansen for their helpful suggestions and Sandra Chambers for invaluable assistance in performing the experiments.
THE JOURNAL. OF B~LOGICAL. CHEMISTRY 0 1990 by The American Society for Biochemistry
Regulation Stimulating
Vol.
and Molecular Biology, Inc.
265, No. 21, Issue of July 25, pp. 12664-12670,199O Printed in U.S. A.
of the Hepatic Synthesis of Cl Inhibitor by the Hepatocyte Factors Interleukin 6 and Interferon y* (Received for publication,
Bruce From
L. ZurawS the Department
and Martin of Molecular
Lotz$ and Experimental
Medicine,
Cl inhibitor (ClINH), the major plasma inhibitor of activated Cl, kallikrein, and activated Hageman factor, may be an important factor in limiting inflammatory injury mediated by the complement and contact systems. ClINH is thought to be synthesized primarily in the liver; however, the regulators of hepatic Cl inhibitor synthesis are completely unknown. In this report, we analyze the regulation of ClINH synthesis by hepatocyte stimulating factors in human hepatoma cell lines and primary hepatocytes. Interleukin-6 and interferon y increase ClINH production in both hepatoma cells and hepatocytes. These cytokines stimulate de nouo synthesis of functional ClINH, acting at a pretranslational level as assessed by Northern blotting. The stimulatory effects of interleukin-6 and interferon y on ClINH synthesis are separate and are differentially modulated by interleukin1. These results establish that hepatic ClINH synthesis is regulated by hepatocyte stimulating factors and reveal novel interactions between these factors.
Proteins whose serum levels increase during inflammation are known as acute phase proteins (APP).’ This highly conserved response is thought to be involved in limiting the deleterious tissue injury that is a by-product of inflammation (1). Acute phase proteins are synthesized at least in part by hepatocytes, and their hepatic synthesis is regulated by soluble factors termed hepatocyte-stimulating factors (HSFs) (2). A number of different HSFs have been described, and it has become clear that multiple HSFs may act in a complex fashion to regulate the acute phase response (3). Cl inhibitor (ClINH), a member of the serpin superfamily, is the major inhibitor of the complement proteases Cir and Cis (4, 5) as well as the contact system proteases kallikrein (6-10) and activated Hageman factor (11-13). ClINH is syn*This research was supported in part by Grants HL39773, RR00833, and AR21175 from the National Institutes of Health. This is publication 5164BCR from the Research Institute of Scripps Clinic. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ Recipient of the Burroughs Wellcome Developing Investigator Award in ~mmunonharmacologv. To whom correspondence should be addressed: Department of MGlecular and Expegimental Medicine, Scrions Clinic & Research Foundation. 10666 N. TorreV Pines Rd., La &lla, CA 92037. § Investigator of the Arthritis Foundation. ‘The abbreviations used are: APP, acute phase proteins; HSF, hepatocyte-stimulating factors; IL, interleukin;.IFN, interferon; LPS, linonolvsaccharide: TNF, tumor necrosis factor; PMA, phorbol myristate acetate; ELISA, enzyme-linked immunosorbent assay; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; SDSPAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis. _ _ ”
November 27, 1989)
Research
Institute
of Scripps
Clinic,
La Jolla,
California
92037
thesized by diverse cell types, the major sources being hepatocytes (14), hepatoma cells (15, 16), monocytes (17-19), fibroblasts (20), and endothelial cells (21). In human monocytes, IFNr is the only mediator that is known to stimulate ClINH synthesis (22, 23). The regulation of hepatic ClINH synthesis is, however, completely unknown. Several other members of the serpin superfamily, including cul-proteinase inhibitor, cY1-antichymotrypsin, and angiotensinogen, are known to be APP, and ClINH has been suspected to be an APP (24, 25). This study was designed to define the factors that regulate ClINH synthesis in human hepatic cells. We now report that the synthesis of ClINH by human hepatoma cell lines and human and baboon primary hepatocytes is increased by the major mediator of the hepatic acute phase response, interleukin-6 (IL-6). IFNy was also found to stimulate hepatic synthesis of ClINH and other complement components, and it appears to act through a mechanism distinct from that of IL6. Co-addition of interleukin-1 (IL-l) resulted in either an increase or decrease of ClINH secretion depending on whether the IL-l was added with IFNr or IL-6. These results establish that hepatic ClINH synthesis is regulated by hepatocyte stimulating factors and contribute to our understanding of the regulation of the human acute phase response. EXPERIMENTAL
PROCEDURES
Cell Culture-HepGP and Hep3B lines were obtained from the ATCC. The Alexander PLC cell line was a gift from Dr. Frank Chisari (La Jolla, CA). These cell lines were maintained in 150-cm* flasks in RPM1 1640 supplemented with 10% fetal bovine serum, L-glutamine, and antibiotics (complete medium). At confluency, cells were trypsinized and transferred to 96-well flat-bottom tissue culture plates at lo5 cells/well or to 24-well tissue culture plates (5 x 10~ cells/well). All experiments were performed using the 96-well culture plates unless specifically stated otherwise. When these cultures had reached confluency, the medium was removed and fresh complete medium containing various stimuli was added. Culture supernatants were removed at the time points indicated and stored at -20 “C until analvzed for their APP content. The mono&e-like cell line U937 was obtained from the ATCC. Primary Hepatocytes-Hepatocytes were isolated from fresh liver samnles using a modification of a previous technique (26). The sample was-washed 3 times with saline, finely minced, and incubated for 3 h with 0.05% clostridial collagenase (Sipma, type IV). Single cell suspension was obtained by filtering the minced digested hepatic tissue through a nylon mesh and gentle pipetting. The dispersed cells were washed and hepatocytes separated from contaminating red blood cells and Kupffer cells b; 5 cycies of 1 x g sedimentation for 20 min each. The absence of Kunffer cells was documented by the lack of IL-6 secretion following stimulation with LPS.* The hepatocytes were resuspended in complete culture medium containing various stimuli and placed in 96-well flat-bottom tissue culture plates at 5 X lo4 cells/well. Cytokines, Antibodies, ana’ Other Materials-Recombinant human cytokines were kindly provided by the following investigators: IL-6
12664
’ M. Lotz,
and B. L. Zuraw,
unpublished
observation.
Regulation
of Hepatic Cl Inhibitor
(specific activity = 5 X lo6 units/mg) and rabbit antiserum to IL-6 (Drs. T. Hirano and T. Kishimoto, Osaka, Japan); interleukin-l/3 (IL-l@; specific activity = 4 X 106units/mg) and rabbit antiserum to IL-10 (Dr. C. Dinarello, Boston, MA); IFNy (specific activity = 2 X lO’units/mg), tumor necrosis factor a (TNFa; specific activity = 5.07 x lo7 units/mg), lymphotoxin, monoclonal anti-TNF (Dr. M. H. Shepard, Genentech Inc., San Francisco, CA); interferon 01 (IFNa) and interferon @l (IFNPl; Dr. P. Trown, Hoffman-LaRoche); and transforming growth factor /3 (Dr. David Loskutoff, La Jolla, CA). LPS, phorbol myristate acetate (PMA), and dexamethasone were purchased from Sigma. Interleukin-4 (IL-4) was purchased from Genzyme (Boston, MA). Rabbit anti-a,-antichymotrypsin Ig (A022) and rabbit anti-fibrinogen Ig (A080) were purchased from Dako Corp. (Santa Barbara, CA). Rabbit anti-haptoglobin (550311) was purchased from Calbiochem. Rabbit anti-al-proteinase inhibitor Ig (605002) was purchased from Boehringer Mannheim. Goat antialbumin IgG (001-03) was purchased from Atlantic Antibodies (Scarborough, ME). Anti-human Cl inhibitor, Cls, C2, C3, and C4 antisera were prepared by immunizing goats with purified proteins. The goat IgG fractions of these antibodies were prepared by precipitation with 40% ammonium sulfate. The anti-ClINH and anti-Cls antibodies were affinity-purified by adsorption on CNBr-Sepharose 4B conjugated with purified Cl inhibitor or Cls, respectively. The antisera were monospecific by Ouchterloney and immunoelectrophoresis analysis. The production of the murine monoclonal anti-Cl inhibitor antibody has been previously described (27). ClINH was purified from 400 ml of normal plasma with added inhibitors according to the method of Salvesen et al. (28). Cibacron Blue F3GA was obtained from Polysciences, Inc. (Warrington, PA) and coupled to Sepharose CL-6B (Pharmacia LKB Biotechnology Inc.). Cis was purified according to the method of Sim (29). Measurement of Proteins-ClINH in culture supernatant was measured by a double antibody sandwich-type ELISA exactly as previously reported (22). Culture supernatants were diluted 1:lO in phosphate-buffered saline prior to being added to the microtiter wells. C2 and C4 hemolytic titers were performed according to published procedures (30). Functional hemolytic complement titers are expressed as effective molecules per ml of undiluted culture supernatant. Antigenic levels of C2, C3, and C4 were measured by single radial immunodiffusion assay (31). Other APP were measured by rocket immunoelectrophoresis. Ten ~1 of undiluted culture supernatant or various dilutions of serum were used. Appropriate dilutions of reference serum and plasma (Binding Site, San Diego, CA) were included to standardize the assays. In Vitro f5S/Methionine Incorporation-Secreted ClINH was biosynthetically labeled with [?S]methionine and then immune-adsorbed and separated by 7.5% SDS-PAGE as described previously (22,32). The gels were dehydrated by washing three times in dimethyl sulfoxide, treated with 2,5diphenyloxazole/dimethyl sulfoxide (NEF964, Du Pont-New England Nuclear) for 3 h, washed in distilled water for 30 min, dried, and autoradiography performed. Northern Blotting for ClZNH mRNA-HepG2 cells were grown to confluency in 75-cm2 flasks (-2 x 10’ cells) and stimulated as described. Total RNA was harvested according to the method of Chirgwin et al. (33). The monolayers were lysed with guanidine isothiocyanate (Bethesda Research Laboratories) and separated on cesium chloride (Fisher) density gradients (yield -100 pg of RNA/ flask). Formaldehyde-denatured total RNA (4 pg of total RNA/lane) was separated on formaldehyde-containing agarose gels. The RNA was transferred to nitrocellulose, baked, and hybridized with a 32Plabeled synt,hetic oligonucleotide in 6 x SSC (15 M NaCl, 0.015 M sodium citrate) with 0.2 mg/ml yeast tRNA (Boehringer Mannheim), 0.1% SDS, 0.1% bovine serum albumin, 0.1% polyvinylpyrollidone, 0.1% Ficoll, and 0.05 M HEPES, pH 7.4. A 21-base oligonucleotide was synthesized to be complementary to nucleotides coding for ClINH amino acids 219-225 (34, 35). The filters were washed in 0.1 X SSC with 0.1% SDS at 37 “C for 30 min then 51 “C for 5 min, and ClINH mRNA was detected by autoradiography. Data Analysis-In experiments using 96-well culture plates, each sample was pooled from five identical culture wells. Analyses were performed at least in duplicate and often in quadruplicate. Results are expressed as a single mean value. Experiments were performed multiple times as indicated in the text, and representative experimental results shown. Statistical inferences were made using conservative techniques. The standard error of the mean was determined for each acute phase protein in control, IFNy-stimulated, and IL-6-stimulated cultures, and the largest standard error was utilized for the analysis. Sample sizes were 20 each for ClINH and 10 each for the other acute
Synthesis
phase proteins. differences that two-tailed test.
A critical z value of 1.96 was used to determine were significant at the 95% confidence level by a
RESULTS
IFNr
HepG2 CIINH Secretion-To of hepatic ClINH synthesis, we tested various mediators for their capacity to increase ClINH production in HepG2 cells, a widely used model for human hepatocyte function. HepG2 cells spontaneously secreted small amounts of ClINH into the culture supernatants. The addition of IFNr, the only inducer of ClINH synthesis in monocytes, stimulated a dose-dependent increase in ClINH secretion by HepG2 cells (Fig. 1). The effect of 1FN-y at earlier time points was examined in a separate experiment. However, ClINH secretion was not detectable before 12 h (see below). Viability of the HepG2 cells (checked by trypan blue dye exclusion) and proliferation (measured by [3H]thymidine incorporation) was not different in IFNy-stimulated compared to control cultures. Substitution of serum-free medium for complete medium did not change the IFNy- or IL6- (see below) stimulated ClINH secretion in short term HepG2 cultures. IFNPl and IFNa stimulation of HepG2 cells were checked in three separate experiments and did not stimulate ClINH secretion. The ClINH induced by IFNr was shown to be synthesized de nouo by incorporation of [35S]methionine (Fig. 2). In two separate experiments, biosynthetically labeled ClINH was detectable in IFNy-stimulated HepG2 cultures at 1 day (the first time point sampled) but could be seen only faintly by day 7 in unstimulated cultures. Although the intensity of the biosynthetic labeling of ClINH was maximal by 24 h, large amounts of secreted ClINH were not detected in the culture supernatant until at least 48 h. Therefore, most of the subsequent experiments employed a stimulation of 48-96 h or longer. Analysis of the molecular weight of the secreted elucidate
and IL-6 the
regulatory
Stimulate
control
FIG. 1. IFNy and IL-6 stimulate ClINH secretion by HepG2 cells. HepG2 cells (lo5 cells/well) in 96-well plates were cultured with media alone, IFNr, or IL-6 at the indicated doses. Cultures were terminated by harvesting the supernatants in sets of 5 wells which were pooled. Separate sets of 5 culture wells were terminated at each time point. The total cumulative ClINH secretion was determined by ELISA. Data are from one of three similar kinetic experiments.
Regulation
12666 Control Added
Cl%:
198K
-
110 K
-
-
of Hepatic Cl Inhibitor
Synthesis
IFNy +
+
FIG. 2. HepGL-synthesized ClINH forms SDS-PAGE-stable complexes with Cis. HepG2 cells (5 x 10” cells/well) in 24. well plates were cultured with medium (control) or IFNr (100 units/ ml) for 5 days, and then washed free of culture medium and pulsed with [““Slmethionine for 24 h. The supernatants were collected, centrifuged, and 250-~1 aliquots incubated with purified Cis (4 pg) or an equivalent volume of phosphate-buffered saline for 1 h at 37 “C. Each sample was then incubated with 40 ~1 of a 1:l suspension of anti-ClINH antibody conjugated to Sepharose 4B beads for 1 h at 23 “C to adsorb the ClINH from the supernatant. The adsorbed proteins were separated on 7.5% SDS-PAGE and detected by autoradiography. The expected molecular masses of native ClINH and Cis-ClINH complexes in SDS-PAGE are 110 and 198 kDa, respectively. The autoradiogram is from one of two similar experiments.
ClINH aft.er prolonged stimulation did not show any degradation, suggesting that destruction of the secreted ClINH was not a problem. The addition of cycloheximide (1 pg/ml) decreased IFNr (100 units/ml) stimulated ClINH secretion from 24 to 6 rig/ml and abolished ClINH secretion at 10 pg/ ml. The secreted ClINH was functional as shown by the ability of biosynthetically labeled ClINH (Fig. 2, lane 3) to form SDS-stable complexes with purified active Cis (lane 4). IL-6 has recently been identified as a hepatocyte-stimulating factor (36, 37). When added to HepG2 cells, purified recombinant human IL-6 (rIL-6) stimulated a dose-dependent increase in ClINH secretion (Fig. 1). Antibody to rIL-6 was able to completely abrogate the ClINH response to added rIL-6 (data not shown, repeated twice). We then assessed the ability of monocyte conditioned medium, which contains several different HSFs, to stimulate ClINH secretion from HepG2 cells. Peripheral blood monocytes were obtained from normal donors and cultured with or without LPS (100 rig/ml) for 24 h. U937, a monocyte-like cell line, was also cultured for 24 h with or without PMA (10 rig/ml). LPS and PMA do not induce ClINH synthesis from cultured monocytes or U937 cells (22), but they do induce secretion of HSFs including IL1, TNFo(, and IL-6 (38). The 24-h conditioned medium was added to confluent HepG2 cells in serial dilutions with or without antibody. The HepG2 supernatants were then collected after 3 days and analyzed for ClINH. There was no detectable ClINH in the monocyte or U937 conditioned medium. Fig. 3 shows that the conditioned medium from both LPS-stimulated monocytes and PMA-stimulated U937 cells increased ClINH synthesis from HepG2 cells. Antibody to IL-6 abrogated this response, whereas antisera to IL-l or antiTNFn had no discernable effect. IgG from nonimmunized rabbits also had no effect on ClINH synthesis. Thus it appears that native IL-6, secreted by LPS-stimulated cultured monocytes or PMA-stimulated U937 cells, was able to induce ClINH secretion in HepG2 cells. A number of other mediators that can act on hepatocytes were tested for their ability to stimulate ClINH secretion. The agents tested included: LPS (50-20,000 rig/ml, TNFcv (l1.000 rig/ml), IL-1B (lo-500 rig/ml), dexamethasone (1 PM),
Added 10condttfoned media No antibody conlrol Rb preimmune IgG Rb anti IL-? IgG MOanti TNF IgG Rb an,, IL-6 IgG
LPS Stim. Monocytes
PMA Stim. U937 Cells
Conditioned
Media Source
Frc. 3. Stimulated monocytes and U937 cells secrete IL-6 which increases ClINH secretion when added to HepG2 cells. Peripheral blood monocytes were isolated as described previously (22), stimulated with LPS (100 rig/ml) for 24 h, and the supernatant harvested. U937 cells were stimulated with PMA (10 rig/ml) and the 24-h supernatant collected. The conditioned medium was added at a 1:16 dilution in fresh culture media with or without added antibody to HepG2 cells (10” cells/well) in 96-well plates. The antibodies were added at 10 pg/ml. Culture supernatants were harvested at day 3 in sets of 5 wells which were pooled. Total cumulative ClINH was measured in the HepGP culture supernatant. There was no detectable ClINH in the monocyte or U937 24-h conditioned media. Data are from one of five similar experiments.
lymphotoxin (l-1,000 rig/ml), IL-4 (0.5-50 units/ml), and transforming growth factor fl (0.25-25 rig/ml). ClINH synthesis was not stimulated by any of these agents in at least three different experiments each. As a positive control, each experiment above also included stimulation with IL-6 and IFNy, both of which consistently stimulated ClINH secretion.
These
results
demonstrate
that
among
various
factors
that can potentially modulate hepatocyte function, only IL-6 and IFNy stimulated significant ClINH production. Effects of IL-6 and ZFN+y on Other Acute Phase ProteinsIL-6 significantly increased the secretion of all the positive APP studied (Table I). In contrast, IFNr at a dose that markedly increased ClINH secretion had no discernable effect on q-antichymotrypsin and fibrinogen secretion and actually decreased the secretion of haptoglobin and al-proteinase
inhibitor.
The
effect
of
IFNy
and
IL-6
on
HepG2
secretion of other complement proteins is shown in Table II. HepG2 cells are known to be capable of synthesizing Cls, C2, and C4 (15), and IFNr has previously been reported to stimulate C4 secretion by HepG2 cells (39). We found that IFNy and IL-6 increased both the hemolytic and antigenic C2 and C4 titers in the culture supernatants. The stimulation of C4 secretion
was
more
pronounced
than
that
of C2.
Anti-
genie levels of C3 were minimally stimulated by IL-6. Interactions Between HSFs and Other Hepatic ModulatorsHaving evaluated the effect of individual agents on ClINH synthesis, we studied interactions of the various HSFs and modulators (Table I). The addition of both IL-6 and IFNy resulted in an additive effect on ClINH secretion. Fig. 4 demonstrates this additive effect in more detail. HepGP cells were stimulated with varying doses of IFNy from 0 to 1000 units/ml. A constant dose of IL-6 (100 units/ml) was added at each dose of IFNr. The addition of IL-6 increased the ClINH secretion by a relatively constant amount (40 ng/ ml) irrespective of the IFNy dose, and this constant amount was approximately the same amount as induced by IL-6 itself. Fig. 4 also shows that IL-6 was able to stimulate this increased amount of ClINH secretion when the effect of IFNr by itself had plateaued (100-1000 units/ml), suggesting that IFNy and
Regulation of Hepatic Cl Inhibitor TABLE
Effect
I
of cytokineslfactors on CIINH and acute secretion in HepGP cells
Stimulation” Control IFNr (100 units/ml) IL-6 (100 units/ml) IL-1 (100 units/ml) DEX” (1 pg/ml) IL-6 + IL-1 IFNr + IL-1 IL-6 + DEX IFNr + DEX IL-6 + IL-1 + Dex IL-6 + IFNy
phase
protein
CIINHb
ACH’
HG’
FB’
API’
ALB’
5.5 58.6 28.4 8.1 5.2 22.7 104.5 41.9 75.0 30.5 93.7
1.3 1.3 4.0 1.8 1.3 4.5 1.5 4.9 1.1 4.2 2.7
2.3 NDd 5.1 2.8 ND 4.8 2.6 6.9 2.1 6.1 3.6
1.4 1.4 10.0 2.8 1.4 8.7 1.4 10.7 2.1 10.0 6.7
8.3 6.8 13.3 8.3 8.3 10.4 6.1 15.5 7.6 14.0 11.9
14.0 12.6 11.3 9.3 15.4 6.6 6.6 10.6 12.0 9.3 10.6
’ Confluent cultures of HepG2 cells in 96-well flat-bottom tissue culture plates were stimulated for 4 days as indicated. Each condition represents 5 wells pooled at the time of harvesting culture supernatants. Data shown are from one of two similar experiments. b Measured by ELISA; expressed as nanograms/ml culture supernatant. Differences in secretion were significant (p < 0.05) for changes of greater than 3.2 rig/ml. ’ cr,-Antichymotrypsin (ACH), haptoglobin (HG), fibrinogen (FB), cu,-proteinase inhibitor (API), and albumin (ALB) levels were measured by rocket immunoelectrophoresis and compared to standard serum or plasma pools; expressed as micrograms/ml culture supernatant. Differences in secretion were significant (p < 0.05) for changes of greater than 1.2 pg/ml for ai-antichymotrypsin, 0.9 rg/ml for haptoglobin, 0.9 rg/ml for fibrinogen, 0.9 yg/ml for a,-proteinase inhibitor, and 1.2 rg/ml for albumin. d Abbreviations: DEX, dexamethasone; ND, not detectable. TABLE
Complement component
component
Stimulus”
Hemolytic
c2
c3
c4
None IFNr IL-6 None IFNr IL-6 None IFN+y IL-6
1P
0
by HepG2 activity
Eff. Mol./mP 177 294 234 NT NT NT 9 41 18
0.1
1
TABLE
of IL-6
stimulated
3 5 4 27 29 37 12 36 24
n Confluent cultures of HepG2 cells in 24-well flat-bottom tissue culture plates were stimulated for 4 days as indicated. Each condition represents 5 wells pooled at the time of harvesting culture supernatants. Supernatants were concentrated 5 times prior to assay using centricon microconcentrators. Results are expressed per ml of unconcentrated culture supernatant. Data are from one of three similar experiments. * Abbreviations: Eff. Mol., effective molecules; NT, not tested.
IL-6 have separate and independent effects on HepG2 ClINH secretion. The direct effects of IL-6 and IFNy on ClINH production were significantly enhanced by dexamethasone, which did not stimulate ClINH by itself. A similar synergistic effect of IL6 and dexamethasone was seen for haptoglobin and ai-proteinase inhibitor. The combination of IL-6 and dexamethasone increased CY,-antichymotrypsin and fibrinogen secretion compared to IL-6 alone, but the difference did not reach significance. IL-l by itself did not significantly stimulate ClINH secretion, but in the presence of IL-6 or 1FN-r it showed interesting and divergent effects. IL-l and IFNr synergistically increased ClINH secretion. In contrast, IL-1 antagonized the effect of IL-6 on ClINH and fibrinogen. The inhibitory effect of IL-l on IL-6-stimulated ClINH secretion was maximal when the IL-l was added prior to or at the same time as the IL-6, and it was less pronounced when added after
1000
500
III
ClINH
secretion
by IL-1
% suppressionb
Timing of IL-1 addition”
Exut.
h -16 -4
1
Emt.
+4 +16
2
31’ 18 21 12 9
21 17 8
0
dml
100
effects of IL-6 and IFNr on HepG2 ClINH additive. HepG2 cells (lo6 cells/well) in 96-well plates with medium alone or stimulated as indicated. Culture were harvested at day 6 in sets of 5 wells which were cumulative ClINH was measured in the supernatants. one of four similar experiments.
Inhibition
cells Antigenic level
10
IFNy U/ml FIG. 4. The secretion are were cultured supernatants pooled. Total Data are from
II
secretion
x
12667
Synthesis
’ Relative to the addition of IL-6, which is called 0 h. b Calculated as ((ClINH in IL-6-stimulated control - ClINH in sample)/ClINH in IL-6-stimulated control) x 100. Confluent cultures of HepG2 cells in 96-well flat-bottom tissue culture plates were stimulated for 3 days in Expt. 1 and 2 days in expt. 2 after addition of IL-6. Each condition represents 5 wells pooled at the time of harvesting culture supernatants. ’ Differences of 10% or greater in suppression are significant (p < 0.05).
the IL-6 (Table III). The timing of IL-l addition did not change its interaction with IFNr. IFNr, IL-6, and IL-l all decreased albumin synthesis. HepG2 cells are thought to represent relatively well differentiated adult hepatocytes while Hep3B cells are similar to fetal hepatocytes (40). We therefore determined whether Hep3B cells also secrete ClINH when stimulated with IFNr or IL-6. Hep3B cells responded to stimulation with both cytokines by rapidly increasing ClINH secretion (Hep3B day 3: control, 9.6 rig/ml; IL-6 (100 units/ml), 36.9 rig/ml; 1FN-y (100 units/ml), 51.0 rig/ml. HepG2 day 3: control, 3.3 rig/ml; IL-6 (100 units/ml), 11.2 rig/ml; IFNy (100 units/ml), 31.8 rig/ml). Like HepG2 cells, Hep3B cells did not respond to LPS, IL-l, or TNFa. The Alexander PLC human hepatoma line constitutively secreted high levels of ClINH, unlike the other hepatoma lines. When stimulated with IL-6 or IFNr for 3 days, Alexander PLC cells secreted even more ClINH (control = 95 rig/ml; IL-6 (100 units/ml) = 116 rig/ml; IFNr (100 units/ml) = 136 rig/ml). CIINH
Secretion
from
Primary
Human
and
Baboon
Hepa-
tocytes-In the next set of experiments, we sought to validate the findings obtained with the hepatoma cell lines on primary hepatocytes. A fresh human liver biopsy sample (1 g) was obtained from the margin of an adrenal tumor resection. Primary human hepatocytes were isolated by collagenase treatment and differential sedimentation (see “Experimental
Regulation
12668
of
Hepatic
Procedures”). Because of the limited number of hepatocytes available, they were cultured at a low density (5 x 10’ cells/ well). Table IV shows that both IFNr and IL-6 stimulated small but discernable increases in the amount of secreted ClINH. LPS or IL-1 had no effect on ClINH secretion by the primary hepatocytes. In order to repeat and extent these findings, we also studied primary baboon hepatocytes which were more readily available. Baboon ClINH was recognized by the affinity purified goat anti-ClINH antibody but not the monoclonal antiClINH antibody. Therefore the affinity purified goat antiClINH was substituted for the monoclonal antibody in the following experiments. A liver wedge was removed from a baboon who was killed following cardiopulmonary bypass. Tissue blocks devoid of capsule were resected and hepatocytes were isolated and cultured as described above. Table IV shows that 1FN-y and IL-6 also stimulated ClINH secretion from these cells. The larger sample size of the baboon liver permitted an aliquot of the hepatocytes to be frozen. After thawing, the hepatocytes responded to cytokine stimulat.ion in a similar manner. The biosynthetic incorporation of [““S]methionine into primary hepatocyte secreted ClINH was also studied in a separate experiment. Fig. 5 shows that stimulation of the primary hepatocytes with 100 units of IFNy (lane 2) or 100 units of IL-6 (lane 3) resulted in a large increase in the
ClINH Dose
4-1 h Human
None IFN-, IL-6 LPS IL-l
,50 .iO 100 20
None IF!+ IL-S
100 units/ml 100 units/ml
0 1.07 0 0 0
units/ml unns/ml rig/ml rig/ml
“To measure secreted baboon using biotinylated affinity-purified antibody and streptavidin-conjugated cletectmg reagent. ” NT. not tested.
2.47 0.40
Day :1
Day 1
rI,~/nl/ 0 1.51
0 3.98
0.45 0 0
0.69 0 0
NT” N? NT
1.20 9.50 7.30
ClINH, the ELISA was modified by goat anti-ClINH as the second horseradish peroxidase as the
C
IFNy
IL-6
200 K-
116.3 K92.5 K66.2 K45K-
-
FIN:. 5. Primary baboon hepatocytes synthesize ClINH when stimulated with IFNr or JL-6. Baboon hepatocytes were cultured (5 X 10’ cells/well) with medium (Cl, IFNy (100 units/ml), or II,-6 (100 units/ml) for 4 days. and then washed free of culture medium and pulsed with [ “Slmethionme for 24 h. The supernatants were collected. centrifuged, and O.l-ml aliquots incubated with 40 ~1 of a 1:l suspension of affinity-purified goat anti-ClINH antibody conjugated to Sepharose 4R beads for 1 h at 23 “C to adsorb the ClINH. The adsorbed proteins were separated on 7.5% SDS-PAGE and detected by autoradiography.
Cl Inhibitor
Synthesis C IFNy
C
IL-6
28S18 S-
#j&i.
FI(;. 6. ClINH mRNA in HepG2 cells. HepGi! cells (-‘2 x 10’ cells) in 75.cm’ tlasks were cultured with media (Cl, IFNy (100 units/ ml), or II,-6 (100 units/ml) for 4 days. Total RNA was recovered by lysis with guanidine isothiocyanate and separated on cesium chloride density gradients. Four p g of total RNA/lane were separated on formaldehyde-agarose gels. transferred to nitrocellulose paper, baked, and hybridized with 2.8 rig/ml of .‘LP-Iat)eled synthetic oligonucleotide probe for ClINH (specific activity = 36 x 10” cpm/~gl. The nitrocellulose filter was washed at 51 “C and the autoradiogram exposed for 9 days. Data are from one of three similar experiments.
amount of [“S]methionine incorporated into ClINH, confirming the ELISA results. Regulation of Cl INH Synthesis Is Pre-translational-The mechanism of regulation of ClINH synthesis was studied by measuring levels of HepG2 ClINH-specific mRNA. Total RNA was extracted from HepG2 cells, separated on agarose gels, transferred to nitrocellulose, and probed with a “Plabeled synthetic oligonucleotide designed to be complementary to ClINH mRNA (34, 41). Northern blots of unstimulated (lanes I and 3), IFNy-stimulated (lane 2), and IL-6stimulated (lane 4) HepG2 total mRNA are shown in Fig. 6. A 2.1-kilobase band was detected in both the IFNr- and IL6-stimulated cultures. ClINH mRNA has previously been shown to migrate at an apparent size of 2.1 kilobases (41). The unstimulated cultures showed a very weak signal at the same position. Therefore stimulation of ClINH secretion by either 1FN-r or IL-6 is associated with an increase in the amount of ClINH mRNA. DISCUSSION
In this report we analyze the regulation of the hepatic synthesis of ClINH. The HSFs IL-6 and 1FN-r stimulate in vitro synthesis of ClINH by three different human hepatoma cell lines as well as primary human and baboon hepatocytes. HepG2 and Hep3B are hepatoma cell lines which are phenotypically similar to adult and fetal hepatocytes, respectively (40). Both hepatoma lines spontaneously secreted small amounts of ClINH and responded to IL-6 or IFNr by significantly increasing ClINH secretion. Alexander PLC is a hepatitis B surface antigen secreting human hepatoma cell line (42). These cells constitutively secreted high levels of ClINH, and IL-6 or IFNy further increased ClINH secretion. The IL-6- or IFNy-induced ClINH was shown by in uitro incorporation of [““Slmethionine to be synthesized de nouo and was functionally active as demonstrated by its ability to form SDS-stable complexes with Cis. The action of these cytokines was shown to be pre-translational by Northern blotting for ClINH mRNA. Analysis of the interactions between HSFs on the regulation of ClINH synthesis yielded several new observations. When IL-6 and IFNy were used in combination, an additive effect on ClINH secretion was noted. The combination of IL6 plus IFNr increased ClINH secretion above the level seen from IFNr alone, even when the dose of 1FN-y was raised to
Regulation of Hepatic Cl Inhibitor the point where its own ability to increase ClINH secretion had plateaued. In addition, antibody to IL-6 (which did abrogate the effects of IL-6) had no effect on IFNy-stimulated ClINH secretion. It is interesting to note that IL-6 has no stimulatory effect on ClINH secretion from monocytes (data not shown) despite the fact that human monocytes and the monocyte-like cell lines U937 and HL60 are rich in IL-6 receptors (43,44). Of particular interest were the interactions between IL-6 and IFNy with IL-l, which by itself did not stimulate ClINH secretion. In six separate experiments, when added before or at the same time as IL-6, IL-1 consistently decreased IL-6-stimulated ClINH secretion, similar to its effect on HSF-stimulated fibrinogen secretion (45). When added in combination with IFNr, however, IL-l enhanced ClINH secretion, regardless of the timing. Collectively, these findings demonstrate that ClINH can be induced via a least two distinct pathways that involve either IL-6 or IFNr. It has become apparent that not only are there multiple HSFs, but each HSF may have a different spectrum of stimulatory effects (3, 46). Furthermore our results indicate that the effects of one HSF may be dramatically altered by the presence of another HSF. Thus the profile of APP induction may be the product of the actions of each specific cytokine involved, and ultimately reflect both the initiating stimulus and the stage of the inflammatory response. IL-l and TNFa have previously been proposed as the primary mediators of the acute phase response. However, only a portion of the acute phase response can be directly attributed to these mediators (47-49), and they also fail to induce ClINH secretion. IL-6 has recently been identified as a major HSF (36, 37). IL-6 is a 21-26-kDa glycoprotein which regulates growth and differentiation of B and T lymphocytes (5054) and various other cell types in addition to being a HSF (55, 56). IL-6 is made by diverse cells including monocytes, fibroblasts, T-cells, bladder carcinoma cells, cardiac myxoma cells, and keratinocytes (37, 55, 57). IL-l and TNFa induce IL-6 synthesis from endothelial cells (58), synoviocytes (59), and even hepatic cells (60), suggesting that the effect of IL-l and TNFo( on the acute phase response could be mediated at least in part by IL-6 (37). Conversely, evidence is accumulating that IL-l can negatively regulate synthesis of some HSFinduced positive APP and positively affect some others (45). Little is known about IFNy as a HSF. IFNr has previously been shown to stimulate C4, C2, and factor B synthesis from hepatoma and transfected L cells (39,61). 1FN-y has also been reported to inhibit HepG2 synthesis of cu,-antichymotrypsin and haptoglobin while stimulating synthesis of a*-macroglobulin (62). We analyzed the spectrum of APP whose secretion could be stimulated by IFN7. Only a specific set of proteins, including ClINH and other components of the complement system, were stimulated by IFN7. The selectivity of IFNy is highlighted by the lack of response of the other APP belonging to the serpin family. IFNy does decrease albumin secretion, like other HSFs. In view of this and its ability to augment IL1 and TNF synthesis, IFNr has been considered a possible mediator of the acute phase response (63). In addition to the direct actions of IFNr on hepatoma cells as a HSF, IFNr has two other actions of relevance to the regulation of ClINH synthesis. First, 1FN-r potentiates IL-6 production from monocytes/macrophages.” Second, 1FN-y directly stimulates monocyte ClINH synthesis (22,23). In addition to human hepatoma cells, the regulation of ClINH synthesis was also studied in primary hepatocytes. Normal human and baboon hepatocytes were established in primary culture, stimulated with HSFs, and the secreted ‘I M. Lotz, unpublished
observation.
12669
Synthesis
ClINH determined. In contrast to the hepatoma cell lines, there was no detectable spontaneous ClINH secretion during a 4-day culture period. The addition of IL-6 or IFNr, however, stimulated measurable ClINH secretion as determined both by ELISA and metabolic labeling. We cannot totally exclude the possibility that the ClINH was synthesized by contaminating Kupffer cells, although LPS did not’stimulate IL-6 production as would have been expected from Kupffer cells.2 This suggests that IL-6 and IFNr may also regulate ClINH synthesis from hepatocytes in uiuo, and corroborates our hepatoma data. In conclusion, this study demonstrates that the hepatic synthesis of ClINH is regulated by IL-6 and IFNy. Each of these mediators act pre-translationally, and when added together show additive effects on ClINH synthesis. IL-l, which by itself has virtually no effect, has significant inhibiting or augmenting effects on ClINH secretion when added in combination with either IL-6 or IFNr, respectively. This is a novel interaction between different HSFs, and it establishes the value of studying ClINH synthesis as an aid to understanding the regulation of the acute phase response. The significance of ClINH as an APP is further apparent from its essential role in limiting complement and contact system activation. Finally, our demonstration that IL-6 and IFNr increase primary hepatocyte ClINH synthesis with the previous observation that increased IL-6 levels are elevated in the serum of patients during an acute phase response (64) strongly suggest that these mediators regulate ClINH synthesis in uiuo. Increased synthesis of ClINH as an acute phase reactant could contribute to the limitation of attacks of angioedema in ClINH deficient patients, and suggests the possibility of using these mediators for therapeutic intervention in ClINH deficiencies. Acknowledgments-We thank Drs. Kevin Sullivan and Sandra Christiansen for their helpful suggestions and Sandra Chambers for invaluable assistance in performing the experiments.
REFERENCES 1. Fey, G. H., and Fuller, G. M. (1987) Mol. Biol. Med. 4, 323-338 2. Ritchie, D. G., and Fuller, G. M. (1983) Ann. N. Y. Acad. Sci. 408,490-502 3. Baumann, H., Onorato, V., Gauldie, J., and Jahreis, G. P. (1987) J. Biol.
Chem.
262,9756-9768
4. Sim, R. B., Arloud, G. J., and Colomb, M. G. (1980) Biochim. Biophys. Acta 612, 433-449 5. Gigli, I., Ruddy, S., and Austen, K. F. (1968) J. Zmmunol. 100, 1154-1164
6. Landerman, H. E. (1962)
N. S., Webster, M. E., Becker, E. L., and Ratcliffe. J. Allergy
33,
330-341
7. Ratnoff. 0. D.. Penskv. J.. Oeston. D.. and Naff. G. B. (1969) J. Exp. Med. lb9,31~~33~ 8. Gigli, I., Mason, J. W., Colman, R. W., and Austen, K.F. (1970) J. Zmmunol. 104,574-581 9. Schapira, M., Scott, C. F., and Colman, R. W. (1982) J. Clin. Znuest.
69,462-468
10. van der Graaf, .F., Koedam, J. A., and Bouma, B. N. (1983) J. Clin. Invest. 71, 149-158 11. Schreiber, A. D., Kaplan, A. P., and Austen, K. F. (1973) J. Clin. Invest.
52,
1402-1409
12.
Forbes, C. D., Pensky, J., and Ratnoff, 0. D. (1970) J. Lab. Clin.
13.
Pixley, R. A., Schapira, M., and Colman, R. W. (1985) J. Biol.
Med. Chem.
‘76,809-815 260,1723-1729
14. Johnson, A. M., Alper, C. A., Rosen, F. S., and Craig, J. M. (1971) Science 173,553-554 15. Morris, K. M., Aden, D. P., Knowles, B. B., and Colten, H. R. (1982) J. Clin. Invest. 70,906-913 16. Prandini, M.-H., Reboul, A., and Colomb, M. G. (1986) Biochem. J. 237,93-98
17. Bensa, J. C., Reboul, A., and Colomb, M. G. (1983) Biochem. 216,385-392
J.
12670
Regulation of Hepatic Cl Inhibitor
18. Yeung-Laiwah, A. C., Jones, L., Hamilton, A. O., and Whaley, K. (1985) Biochem. J. 226, 199-205 19. Randazzo, B. P., Dattwyler, R. J., Kaplan, A. P., and Ghebrehiwet, B. (1985) J. Zmmunol. 135,1313-1319 20. Katz, Y., and Strunk, R. C. (1988) Arthritis Rheum. 31, 13651370 21. Schmaier, A. H., Murray, S. C., Heda, G. D., Farber, A., Kuo, A., McCrae, K., and Cines, D. B. (1989) J. Biol. Chem. 264,1817318179 22. Lotz, M., and Zuraw, B. L. (1987) J. Zmmunol. 139, 3382-3387 23. Hamilton, A. 0.. Jones, L.. Morrison. L.. and Whalev. “, K. (1987) , Biochem. J. 242, 809-815 24. Donaldson, V. H. (1966) J. Lab. Clin. Med. 68, 369-382 25. Astrup, J., Colstrup, H., and Frandsen, B. (1977) Acta Radiol. Ther. Biol. 16,394-400 26. Edge, S. B., Hoeg, J. M., Triche, T., Schneider, P. D., and Brewer, H. B.. Jr. (1986) J. Biol. Chem. 261.3800-3806 27. Zuraw, B., Takano, K., Jensen, F., Yeivington, M., and Curd, J. G. (1984) J. Allergy. Clin. Zmmunol. 73, 155A-155A 28. Salvesen, G. S., Catanese, J. J., Kress, L. F., and Travis, J. (1985) J. Biol. Chem. 260,2432-2436 29. Sim, R. (1981) Methods Enzymol. 80, 26-42 30. Muller-Eberhard, H. J. (1977) in Methods in Immunology and Immunochemistry (Williams, C. A., and Chase, M. W., eds) pp. 127-274, Academic Press, New York 31. Oudin, J., and Williams, C. A. (1971) in Methods in Immunology and Immunochemistry (Williams, C.A., and Chase, M.W., eds) pp. 213-224, Academic Press, New York 32. Zuraw, B. L., and Curd, J. G. (1986) J. Clin. Znuest. 78. 567-575 33. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J., and Rutter, W. J. (1979) Biochemistry 18.5294-5299 34. Bock, S.‘C., Skriver, K., Nielsen, E., Thogersen, H.-C., Wiman, B., Donaldson, V. H., Eddy, R. L., Marrinan, J., Radziejewska, E., Huber, R., Shows, T. B., and Magnusson, S. (1986) Biochemistry 25,4292-4301 35. Davis, A. E., III, Whitehead, A. S., Harrison, R. A., Dauphinais, A., Bruns, G. A. P., Cicardi, M., and Rosen, F. S. (1986) Proc. N&l. Acad. Sci. U. s. A. 83,3161-3165 36. Andus. T.. Geiger. T.. Hirano. T.. Northoff. H.. Ganter. U.. Bauer. J., Kishimo~o, ‘T.,‘and He&rich, P. (1987) FEBS ‘Lett. 221; 18-22 37. Gauldie, J., Richards, C., Harnish, D., Lansdorp, P., and Baumann, H. (1987) Proc. N&l. Acad. Sci. U. S. A. 84,7251-7255 38. Fuller, G. M., Bunzel, R. J., Woloski, B. M., and Naham, S.-U. (1987) Biochem. Biophys. Res. Commun. 144, 1003-1009 39. Miura, N., Prentice, H. L., Schneider, P. M., and Perlmutter, D. H. (1987) J. Biol. Chem. 262, 7298-7305 40. Oliviero, S., Morrone, G., and Cortese, R. (1987) EMBO J. 6, 1905-1912 41. Cicardi, M., Igarashi, T., Rosen, F. S., and Davis, A. E., III (1987) J. Clin. Incest. 79, 698-702 42. Macnab, G. M., Alexander, J. J., Lecatsas, G., Bey, E. M., and
Synthesis
Urbanowicz, J. M. (1976) Br. J. Cancer 34, 509-515 43. Bauer, J., Bauer, T. M., Kalb, T., Taga, T., Lengyel, G., Hirano, T., Kishimoto, T., Acs, G., Mayer, L., and Gerok, W. (1989) J. Exp. Med. 170, 1537-1549 44. Taga, T., Kawanishi, Y., Hardy, R. R., Hirano, T., and Kishimoto, T. (1987) J. Exp. Med. 166,967-981 45. Baumann, H., Richards, C., and Gauldie, J. (1987) J. Zmmunol.
139,4122-4128 46. Baumann, H., Hill, R. E., Sauder, D. N., and Jahreis, G. P. (1986) J Cell Biol. 102, 370-383 47. Gauldie, J., Sauder, D. N., McAdam, K. P. W. J., and Dinarello, C. A. (1987) Immunology 60. 203-207 48. Darlington, G. J., Wilson; D. R., and Lachman, L. B. (1986) J. Cell Biol. 103, 787-793 49. Le. P. T., and Mortensen. R. F. (1986) J. Zmmunal. 136. 25262533 50. Kawano, M., Hirano, T., Matsuda, T., Taga, T., Horii, Y., Iwato, K., Asaoku, H., Tang, B., Tanabe, O., Tanaka, H., Kuramoto, A., and Kishimoto, T. (1988) Nature 332,83-85 51. Tosato. G.. Seamon. K. B.. Goldman. N. D.. Seheal. P. B.. Mav. L. T.; Washington, G. C:, Jones, K. D., and PiYke,‘S. E. (1988) Science 239,502-504 52. Garman, R. D., Jacobs, K. A., Clark, S. C., and Raulet, D. H. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 7629-7633 53. Takai, Y., Wong, G. G., Clark, S. C., Burakoff, S. J., and Herrmann, S. H. (1988) J. Zmmunol. 140.508-512 54. Lotz, M., Jirik, F.,’ Kabouridis, P., Tsoukas, C., Hirano, T., Kishimoto, T., and Carson, D. A. (1988) J. Exp. Med. 167, 1253-1258 55. Sehgal, P. B., May, L. T., Tamm, I., and VilEek, J. (1987) Science 235, 731-732 56. Kishimoto, T. (1989) Blood 74, l-10 57. Hirano, T.1 Yusukawa, K., Harada, H., Taga, T., Watanabe, Y., Matsuda. T.. Kenshivunura. S.. Nakaiima. K.. Kavama. K.. Iwamatsu, A:, Tsunasawa, S:, Sakiyama, R.; Matsui, H., Tak: ahara, Y., Taniguchi, T., and Kishimoto, T. (1986) Nature 324,73-76 58. Jirik, F. R., Podor, T. J., Hirano, T., Kishimoto, T., Loskutoff, D. J., Carson, D. A., and Lotz, M. (1989) J. Zmmunol. 142, 144-147 59. Guerne, P.-A., Zuraw, B. L., Vaughan, J. H., Carson, D. A., and Lotz, M. (1989) J. Clin. Znuest. 83, 585-592 60. Lotz, M., Zuraw, B. L., Carson, D. A.. and Jirik, F. R. (1989) Ann. N. Y. Acad. Sci. 557,509-511 61. Perlmutter, D. H., Strunk, R. C., Goldberger, G., and Cole, F. S. (1986) Mol. Zmmunol. 23, 1263-1266 62. Magielska-Zero, D., Bereta, J., Czuba-Pelech, B., Pajdak, W., Gauldie. J.. and Koi. A. (1988) Biochem. Znt. 17. 17-23 63. Dowton, S. B., and C&en,‘H. R: (1988) Semin. Hematol. 25,8490 64. Nijsten, M. W. N., De Groot, E. R., Ten Duis, H. J., Klasen, H. J., Hack, C. E., and Aarden, L. A. (1987) Lancet ii, 921
