Acute-phase response of human hepatocytes: regulation of acute-phase protein synthesis by interleukin-6

Extent of the acute phase response in fulminant hepatic failure

The extent of the acute phase response and the relation between acute phase proteins and cytokines in plasma was investigated in 50 patients with fulminant hepatic failure. On admission, C reactive protein was significantly higher in fulminant hepatic failure (median: 12.4 micrograms/ml, range:0.2-112 micrograms/ml) than in 20 controls (median: 0.8 microgram/ml, range: 0.3-2.9 micrograms/ml, p < 0.001). Serial measurements showed that plasma C reactive protein increased daily after admission until day 5, the end of the study period. alpha 1-Antitrypsin (AAT) (median: 69.1%, range: 27.5-124%) and fibrinogen (median: 1.10 g/l, range: 0-2.82 g/l) were significantly lower in fulminant hepatic failure on admission than in controls (AAT: median: 126%, range: 75.4-149%; fibrinogen: median 2.48 g/l, range: 1.82-3.39 g/l, p < 0.001) and did not change subsequently. Both AAT and fibrinogen were maintained at significantly higher concentrations in survivors than in those who did not. Bacterial infection occurred in 23 patients during the course of fulminant hepatic failure, but did not influence the concentrations of these three proteins. Interleukin 6 was significantly higher in fulminant hepatic failure (median: 21.2 pg/ml, range: 0-871 pg/ml) than in controls (median: 2.4 pg/ml, range: 1.5-8.2 pg/ml, p < 0.001). There was a significant correlation between interleukin 6 and the C reactive protein concentrations in patients with viral hepatitis on admission and in all patients 48 hours later, consistent with other evidence that interleukin 6 stimulates synthesis of this acute phase protein.

Interleukin-6: an early marker of bacterial infection in decompensated cirrhosis

Fifty-seven patients with decompensated cirrhosis were studied prospectively to assess the sensitivity and specificity of early clinical or biological signs of bacterial infection. Among them, 19 had proven infection on admission (7 spontaneous bacterial peritonitis, 5 bacteraemia, 3 urinary tract infections, 2 pneumonia, 1 dental abscess and 1 cholangitis). Fever, polymorphonuclear cell count, fibrinogen and C-reactive protein levels were found to be of little or no help in diagnosing bacterial infection on admission. Interleukin-6 plasma levels were, however, significantly different between infected (median: 1386 pg/ml, range: 237-20,000) and non-infected patients (median: 34 pg/ml, range: 0-4500, p < 0.00001). Levels above 200 pg/ml were always found in infected patients, giving a sensitivity of 100% and a specificity of 74%. C-reactive protein correlated weakly with interleukin-6 levels, indicating a defective acute-phase response in cirrhosis. Tumor necrosis factor alpha plasma levels were less sensitive (95%) and specific (68%) for the diagnosis of bacterial infection at a threshold of 50 pg/ml, but were more closely related to a poor patient outcome. In decompensated cirrhosis, interleukin-6 plasma levels on admission provided the most sensitive and specific tool for the diagnosis of bacterial infection.

Determination of haptoglobin expression in IL-6 treated HepG2 cells by ELISA and by RNA hybridization--evaluation of a quantitative method to measure IL-6

Interleukin-6 (IL-6) is known to be an important modulator of acute phase (AP) protein expression in hepatocytes both in vivo and in vitro. In the present study the inducing activity of IL-6 on the expression of the AP protein haptoglobin (HP) by the human hepatoma cell line HepG2, has been evaluated. HP mRNA inducibility was analysed by Northern and slot-blot hybridization, while HP protein was detected by means of an ELISA procedure. A dose-response relationship from 0.3 to 4.8 ng/ml of a human recombinant IL-6 preparation derived from a Chinese hamster ovary (CHO) cell line was observed after 48 h of treatment. Comparable results were obtained by analysing both HP mRNA expression and HP protein secretion. Detectable induction of HP protein secretion was observed with as little as 25 pg/ml of IL-6. The effect of IL-6 was potentiated by dexamethasone, while an inhibition on HP mRNA inducibility could be prevented by lowering the foetal calf serum (FCS) concentration to 1%. Preliminary data indicate that neither IL-1 beta nor TNF-alpha were able to induce significantly HP mRNA expression and protein secretion. The activity ratio between two IL-6 preparations (from CHO and E. coli cells) obtained with a conventional IL-6 bioassay (i.e., T1165 cell growth assay) was comparable to that obtained in the induction of HP expression. The nominal specific activity of the CHO-derived IL-6 was two to three times higher with both responses.

Regulation of hepatocyte albumin and alpha 1-acid glycoprotein secretion by monokines, dexamethasone, and nitric oxide synthase pathway: significance of activated liver nonparenchymal cells

To clarify the mechanism involved in regulating the secretion of albumin and alpha 1-acid glycoprotein by rat hepatocytes, we studied hepatocyte culture and cocultures of hepatocyte and liver nonparenchymal cells. The secretion of alpha 1-acid glycoprotein by hepatocytes was stimulated and that of albumin was inhibited by combinations of dexamethasone and monokines, especially by dexamethasone and interleukin-6. The secretion of these proteins was equally inhibited during stimulation by lipopolysaccharide in cocultures. The inhibitory effect of sinusoidal endothelial cells was smaller than that of Kupffer cells. This inhibition was partially abolished by blocking the nitric oxide synthase pathway in cocultured cells and was completely abolished by dexamethasone. In conclusion, the secretion of albumin and alpha 1-acid glycoprotein by hepatocytes was regulated by monokines, dexamethasone, and the inducible nitric oxide synthase pathway in hepatocytes and liver nonparenchymal cells in vitro.

Value of serum C-reactive protein measurement in the detection of hepatocellular carcinoma superimposed on liver cirrhosis

We investigated whether, in Italian patients, C-reactive protein (CRP) determination could be considered a useful adjunct, complementary to alpha 1-fetoprotein, in the detection of liver cancer. CRP was determined by particle-enhanced nephelometry in 171 subjects (102 male, 69 female). Fifty-five patients had mild chronic liver disease (CLD), 45 cirrhosis (CIR), 38 hepatocellular carcinoma (HCC); 33 subjects were healthy controls. Patients with HCC and CIR had higher CRP levels (P < 0.05) than those found in patients with CLD and controls. CRP higher than 5 mg/l was found in 30/38 (78.9%) patients with HCC, 28/45 (62.2%) patients with CIR, 16/55 (29.1%) patients with CLD (chi 2 56.0, P < 0.0001). Sensitivity, specificity and diagnostic accuracy of CRP in diagnosing HCC with respect to CLD+CIR were: 78.9%, 56.0% and 34.9%. However, when considered only in the subgroup of patients with alpha 1-fetoprotein below or equalling 30 ng/ml, they were 50.0%, 54.3% and 4.3% respectively. In conclusion, CRP concentration is frequently elevated in patients with HCC, however, it does not seem to improve the ability of alpha 1-fetoprotein to discriminate HCC from CIR.

Serum immunoreactive interleukin-6 and C-reactive protein levels in patients with multiple myeloma at diagnosis

Serum bioactive but not immunoreactive interleukin-6 (IL-6), and serum C-reactive protein (CRP), have been reported to be of prognostic significance in multiple myeloma (MM). We measured serum immunoreactive IL-6 by a sensitive enzyme-linked immunosorbent assay in 30 MM patients at diagnosis. In 30% of the patients serum immunoreactive IL-6 exceeded the upper reference limit. The concentrations of CRP and IL-6 showed a linear association. Logarithmically transformed IL-6, CRP and beta 2-microglobulin were significant variables by univariate survival analysis; by multivariate analysis CRP was a slightly stronger prognostic factor than IL-6 and the only one of independent prognostic significance.

Interleukin-6 does not activate protein breakdown in rat skeletal muscle

The addition of recombinant human interleukin-6 (rIL-6) to either soleus or extensor digitorum longus (EDL) muscle preparations did not affect the rate of protein breakdown as measured by the rate of tyrosine released to the medium. In addition, the presence of the cytokine did not influence either the rate of protein synthesis or that of alpha-(methyl)-aminoisobutyric acid (MeAIB) uptake by the muscle preparations. It is concluded that IL-6 is not the mediator in activating muscle protein turnover during sepsis in the rat.

Interleukin 4 inhibits the production of some acute-phase proteins by human hepatocytes in primary culture

Interleukin 4 (IL4) has been shown to exhibit anti-inflammatory effects by inhibiting the secretion by monocytes of proinflammatory cytokines such as interleukin 1 (IL1), interleukin 6 (IL6), and tumor necrosis factor (TNF) and by inducing the secretion of the IL1 receptor antagonist. We investigated the role of this cytokine on the production of acute-phase proteins in primary human hepatocyte cultures. Cells were exposed to either IL4 and/or IL6, the most potent mediator of hepatic acute phase proteins. IL4 led to decreased production of haptoglobin, C-reactive protein and albumin while alpha 1-antitrypsin and fibrinogen remained unaffected. These inhibitory effects of IL4 were also observed at the mRNA level. In addition, IL4 inhibited the IL6-induced production of haptoglobin although it had no effect on the induced C-reactive protein and fibrinogen. Our results demonstrate that IL4 can affect the production of a subset of acute-phase proteins by human hepatocytes and can antagonize some of the effects of IL6. These observations reinforce the notion that IL4 can be considered as an anti-inflammatory cytokine.

Profiles of cytokines (TNF alpha and IL-6) and acute phase proteins (CRP and alpha 1AG) related to the disease course in patients with systemic lupus erythematosus

Tumor necrosis factor alpha (TNF alpha) and interleukin-6 (IL-6) play a main role in inducing acute phase protein production by hepatocytes. This study describes the serum levels of TNF alpha and IL-6 in relation to serum levels of C-reactive protein (CRP) and alpha 1-acid glycoprotein (alpha 1AG) in three systemic lupus erythematosus (SLE) patients. Disease courses of these patients were divided in a total of 19 clinical periods, according to the clinical symptoms and interleukin profiles. Significantly elevated TNF alpha levels were found in all but three of the defined periods, without being associated with disease activity. In only four of the defined periods elevated TNF alpha were observed combined with elevated IL-6 and CRP levels. Two of these periods coincided with minor symptoms of SLE, one with an exacerbation and the other one with a systemic infection while SLE activity was low. All other periods showed varying combinations of elevated TNF alpha and/or IL-6 levels being followed or not by elevated CRP levels. Significantly raised alpha 1AG levels were measured in all clinical periods. In most of the observed periods a dissociation was found between TNF alpha and IL-6 and also between the different cytokine (TNF alpha and IL-6) levels and acute phase protein (CRP and alpha 1AG) levels. These data could not be explained by differences in disease course or influences of medication. We conclude that more factors other than TNF alpha and IL-6 must play a role in the regulatory pathway of the acute phase response in SLE.(ABSTRACT TRUNCATED AT 250 WORDS)

Overall interleukin-6 production exceeds 7 mg/day in multiple myeloma complicated by sepsis

We previously reported that injection of anti-IL-6 monoclonal antibody (mAb) in a patient with multiple myeloma (MM) induced the circulation of high amounts of IL-6 in the form of IL-6/anti-IL-6 monomeric complexes. This made it possible to estimate overall daily IL-6 production in patients in vivo, which had not been achieved in animals or humans before. In this study, estimations are given for a patient with MM who developed Escherichia coli sepsis during anti-IL-6 mAb. During the first 12 days, the overall IL-6 production was estimated at 1.5 to 2.0 micrograms/day. On day 13, serum IL-6 concentration, in the form of IL-6/anti-IL-6 complexes, increased 1000-fold and was 1.7 x 10(6) pg/ml, in relation with the development of E. coli sepsis. Overall IL-6 production was estimated to be greater than 7 mg/day, i.e. 3500 times higher than before sepsis. Serum IL-6 levels in the form of monomeric immune complexes remained very high for 20 days after sepsis indicating the persistence of very high overall IL-6 production (100 to 3500-fold greater than pre-sepsis production). This study demonstrates a considerable and persistent potential for IL-6 production in this patient during and after sepsis.

Human inter-alpha-inhibitor family in inflammation: simultaneous synthesis of positive and negative acute-phase proteins

The inter-alpha-inhibitor (I alpha I) family encompasses four plasma proteins, namely free bikunin as well as I alpha I, pre-alpha-inhibitor (P alpha I) and inter-alpha-like inhibitor (I alpha LI). Each of the last three proteins is a distinct assembly of one bikunin chain with one or more unique heavy (H) chains designated H1, H2 and H3. The three H chains and the bikunin chain are encoded by four distinct mRNAs. These molecules and chains, as well as the corresponding mRNAs, were quantified in sera and liver biopsies from a series of patients with or without mild or severe acute infection. The decrease or increase observed for a given molecule or chain in the serum was in agreement with a similar change in the corresponding liver mRNA. In acute inflammation the H2 and bikunin chains are down-regulated and the relevant molecules (I alpha I, I alpha LI) behave as negative acute-phase proteins, whereas the H3 chain is up-regulated and the corresponding P alpha I molecule is a positive acute-phase protein. Also, P alpha I displays a higher-than-usual M(r); this is probably due to ligand binding. The H1 gene does not seem to be affected by the inflammatory condition. The quantitative changes in RNA levels seen in vivo were confirmed in vitro in the human hepatoma Hep3B cell line prior to or after induction with the acute-phase mediators interleukin-1 and/or -6. These results provide the first example in humans of positive and negative acute-phase proteins that are encoded by evolutionary related genes.

The effect of interleukin 1 alpha on acetaminophen-induced hepatotoxicity

The protective effect of interleukin 1 alpha (IL-1 alpha) in mice with acetaminophen (AAP)-induced hepatitis was investigated. IL-1 alpha had a significant protective effect if given 2 or more hours (up to 24 hours) before AAP; it significantly reduced mortality of mice and decreased serum transaminase level. The maximal effect was obtained with the dose of 1000 U (166 ng/kg) IL-1 alpha. Pretreatment with IL-1 significantly increased the synthesis of prostaglandin E2 (PGE2) in samples of liver tissue from AAP-treated mice, but had no effect on the synthesis of leukotriene C4 (LTC4). Pretreatment with indomethacin (IMC) did not abrogate significantly the protective effect of IL-1. Thus, the hepatoprotective effect of IL-1 alpha can not be entirely explained by the stimulation of prostaglandin (PG) synthesis.

Immune modulation of adhesion molecules ICAM-1 (CD54) and LFA-3 (CD58) in human hepatocytic cell lines

Interactions between leukocytes such as T cells and accessory or target cells are promoted by adhesion molecules, in particular intercellular adhesion molecule-1 (ICAM-1) and lymphocyte function-associated antigen-3 (LFA-3). Hepatocytes are usually negative for these surface membrane proteins which, however, may be up-regulated in inflammatory processes within the liver. Because the regulatory signals for, and tissue distribution of, these adhesion molecules vary among different tissues, expression of ICAM-1 and LFA-3 was studied in the Hep-G2 and SK-Hep-1 human hepatocytic cell lines in vitro. Low, constitutive membrane expression of the two molecules was detected in both cell lines. ICAM-1, but not LFA-3, was rapidly up-regulated by interleukin-1 (IL-1), tumor necrosis factor-alpha (TNF), and to some extent by interferon-gamma (IFN) and IL-6, whereas IL-4 had variable low effects, if any. Considerable synergism on ICAM-1 protein levels was observed after stimulation with TNF, IL-1, and IFN, whereas co-incubation with actinomycin D abolished these effects. ICAM-1 mRNA levels increased 16-20 times after cytokine incubation. Our data indicated that hepatocytes share the regulatory pathways for ICAM-1 described for several other cell types. Absence of these molecules in vivo may reflect a dominance of negative modulation signals in the normal liver, which might also explain the low levels of HLA class I molecules.

Inflammation-induced expression of sialyl Lewis X-containing glycan structures on alpha 1-acid glycoprotein (orosomucoid) in human sera

The glycosylation of the acute phase glycoprotein alpha 1-acid glycoprotein (AGP) in human sera is subject to marked changes during acute inflammation as a result of the cytokine-induced hepatic acute phase reaction. The changes described thus far comprise alterations in the type of branching of the carbohydrate structures as revealed by increased reactivity of AGP with concanavalin A. We now report on acute inflammation-induced increases in alpha 1-->3-fucosylated AGP molecules, as detected by the reactivity of AGP towards the fucose-binding Aleuria aurantia lectin (AAL) in crossed affino-immunoelectrophoresis of human sera. Laparotomy of women, for the removal of benign tumors of the uterus, was used as a model for the development of the hepatic acute phase response. Hugh increases were detected in the amounts of strongly AAL-reactive fractions of AGP, presumably containing three or more fucosylated N-acetyllactosamine units. At least part of these Lewis X-type glycans (Gal beta 1-->[Fuc alpha 1-->3]GlcNAc-R) appeared to be substituted also with an alpha 2-->3-linked sialic acid residue. This was revealed by the laparotomy-induced abundant staining of AGP with an antisialyl Lewis X monoclonal antibody (CSLEX-1) on blots of sodium dodecyl sulfate-polyacrylamide gels containing AGP isolated from the sera of a patient at various days after operation. It is concluded that acute inflammation induces a strong increase in sialyl Lewis X-substituted AGP molecules that persists at a high level throughout the inflammatory period. We postulate that these changes represent a physiological feedback response on the interaction between leukocytes and inflamed endothelium, which is mediated via sialylated Lewis X structures and the selectin endothelial-leukocyte adhesion molecule 1.

The C-reactive protein responses in HIV-infected patients with pneumonia

The acute phase C-reactive protein (CRP) was measured in serum of HIV-infected patients suffering from Pneumocystis carinii pneumonia (PCP) (32 patients), bacterial pneumonia (10 patients), and in 19 immunocompetent patients with bacterial pneumonia. The HIV-infected patients with bacterial pneumonia had a significantly lower CRP level than the immunocompetent patients (50% versus 95% had an s-CRP level > 80 mg/l). No significant difference was found in the CRP response to P. carinii or bacteria in HIV-infected patients with pneumonia due to these microorganisms (20% versus 50% had s-CRP > 80 mg/l). In the group of PCP patients, a significantly lower CRP level was found in those with CD4 positive lymphocyte counts below 50 x 10(6)/l. There was no correlation between the CRP response and the severity of the PCP as estimated by the degree of hypoxia. We conclude that the CRP level cannot be used to discriminate between PCP and bacterial pneumonia in HIV-infected patients.

High amounts of circulating interleukin (IL)-6 in the form of monomeric immune complexes during anti-IL-6 therapy. Towards a new methodology for measuring overall cytokine production in human in vivo

A patient with plasma cell leukemia was treated with anti-interleukin (IL)-6 monoclonal antibodies (mAb) for 2 months. Using chromatography on protein A-Sepharose, anti-murine-IgG-Sepharose, anti-IL-6-mAb-Sepharose and gel filtration at pH 2.3, we have demonstrated that the anti-IL-6 mAb, by preventing the binding of IL-6 to its cell membrane receptor and its renal elimination, has induced huge amounts of IL-6 to circulate in the form of monomeric immune complexes. By using this observation, we have developed a mathematical modelling that allows the determination of the overall daily production of IL-6 in this patient, which was in the range of 15 micrograms per day. Overall in vivo production of cytokines has never been evaluated in animals or in humans before.

Biosynthesis and intracellular transport of alpha-2,6-sialyltransferase in rat hepatoma cells

We investigated biosynthesis, intracellular transport and release of beta-galactoside alpha-2,6-sialyltransferase in a dexamethasone-inducible rat hepatoma cell line. Confluent cells were induced by 10 microM dexamethasone for 24 h, and metabolically labelled with [35S]methionine/cysteine, followed by immunoprecipitation of sialyltransferase and electrophoretic/fluorographic analysis. The 35S-labelled enzyme was synthesized as a 46-kDa precursor, converted to an intermediate 47-kDa form after 1 h, and gradually to a mature form of 48 kDa within the following 3 h. By means of either tunicamycin inhibition of N-glycosylation or cleavage of N-glycans from isolated sialyltransferase using N-glycosidase F, the sizes of the precursor and the mature form were reduced to 41 kDa and 43 kDa, respectively. After a 4-h chase, treatment with endoglycosidase H revealed two distinct molecular forms of sialyltransferase, bearing either two N-acetyllactosamine-type or one oligomannose-type and one N-acetyllactosamine-type N-linked sugar chain. In addition, sialyltransferase became sensitive to neuraminidase digestion after a 4-h chase. The half-life of intracellular [35S]sialyltransferase was estimated at 3 h. A soluble form was detectable in the supernatant, 2 h after the pulse. Only 12% of the initially labelled sialyltransferase was found in the medium after 12 h, while 73% of the enzyme was degraded intracellularly. To characterize a possible intracellular degradation site, we studied intracellular transport in the presence of either secretion-blocking or acidotropic agents or protease inhibitors. Degradation was significantly delayed by all treatments. Our results show that sialyltransferase follows the secretory pathway as a membrane protein and is retained at a late Golgi stage. We suggest that the bulk of sialyltransferase in rat hepatoma cells is diverted to a post-Golgi degradation pathway. This route contrasts with the post-Golgi trafficking of beta-1,4-galactosyltransferase in HeLa cells, which is constitutively secreted [Strous, G. J. A. M. & Berger, E. G. (1982) J. Biol. Chem. 257, 7623-7628].