Biosynthesis of complement C1 inhibitor by Hep G2 cells. Reactivity of different glycosylated forms of the inhibitor with C1s

A New Tool for Complement Research: In vitro Reconstituted Human Classical Complement Pathway

The complement, as part of the innate immune system, represents the first line of defense against Gram-negative bacteria invading the bloodstream. The complement system is a zymogen cascade that ultimately assemble into the so-called membrane attack complex (MAC), which lyses Gram-negative bacteria upon insertion into the outer membrane. Traditionally, serum has been used as complement source, for example to study the bactericidal activity of monoclonal antibodies or antibodies raised upon vaccination. Due to the significant donor to donor variability, as well as susceptibility of complement factors to handling and storage conditions, assay reproducibility using human serum is low. Moreover, the presence of pre-existing antibodies and antimicrobial compounds are confounding factors. To remove antibodies from human serum, we applied κ/λ-light chain specific affinity chromatography, however the method severely reduced the complement activity due to the depletion of complement components. Therefore, we attempted to reconstitute human complement-namely the alternative (rAP) and the classical (rCP) pathways-from purified complement factors. We found that adding C1-inhibitor to the mixture was essential to maintain a stable and functional C1 and thus to generate an active rCP. We further confirmed the functionality of the rCP by testing the complement-dependent bactericidal activity of a human monoclonal antibody, A1124 against an E. coli clinical isolate belonging to the ST131 clonal complex, and that of a polyclonal IVIg against a laboratory E. coli strain (MG1655) not expressing LPS O-antigen and capsule. Although the alternative pathway did not have any bactericidal activity by itself, it enhanced MAC deposition induced by rCP and increased the overall bactericidal activity against the ST131 E. coli strain. In conclusion, we report for the first time the successful in vitro reconstitution of the classical pathway of the human complement to establish a serum-free, complement dependent bactericidal assay. This system offers high level of standardization and could support the study of the complement in different research fields.

Hepatitis C virus NS3 serine protease interacts with the serpin C1 inhibitor

Both NS3 protein (1007-1657) and its protease moiety (NS3p, 1027-1207) were able to interact in vitro with C1 Inhibitor (C1Inh) to give a 95-kDa Mr C1Inh cleavage product similar to that obtained upon proteolysis by complement protease C1s. High-Mr reaction products were also detected after incubation of C1Inh with NS3 but not with NS3p; they correspond to ester-bonded complexes from their hydroxylamine lability. Similar reactivity of NS3 was observed upon incubation with alpha2-antiplasmin. Serpin cleavage was prevented by treatment of NS3 with synthetic serine protease inhibitors. This interaction between viral NS3 and host serpins suggests that NS3 is likely to be controlled by infected cell protease inhibitors.

Complement biosynthesis by mononuclear phagocytes

Mononuclear phagocytes are an important in vivo source of a wide range of complement components. They are able to rapidly up-regulate or down-regulate complement synthesis in response to many different pharmacological and biological stimuli. This ability is likely to make a significant contribution to maintaining host defences particularly in peripheral tissues. The important role of molecular biology in the study of complement biosynthesis by mononuclear phagocytes will be emphasised.

Fat-storing cells of the rat liver synthesize and secrete C1-esterase inhibitor; modulation by cytokines

During liver fibrogenesis, fat-storing cells transform into myofibroblast-like cells and produce increasing amounts of extracellular matrix proteins. Because fat-storing cells produce alpha 2-macroglobulin, an important serine protease inhibitor (serpin), we investigated whether fat-storing cells also synthesize C1-esterase inhibitor, another important serpin. C1-esterase inhibitor synthesis was studied in rat fat-storing cells at day 0, 3 and 7 after isolation by biosynthetic labeling, immunoprecipitation and sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Messenger RNA was examined by Northern-blot analysis. C1-esterase inhibitor gene expression and synthesis were detectable in freshly isolated fat-storing cells and increased distinctly during the time in culture. The cellular source of C1-esterase inhibitor in fat-storing cell cultures was also identified by in situ hybridization of cells at different times after isolation. By inhibition of the N-glycosylation using tunicamycin, rat C1-esterase inhibitor was identified as a glycoprotein. The time course of C1-esterase inhibitor secretion was determined by pulse-chase experiments. C1-esterase inhibitor synthesis was increased 6-fold to 10-fold by interferon-gamma. Specific messenger RNA levels were also raised distinctly by this cytokine. In contrast, interferon-alpha and dexamethasone did not alter C1-esterase inhibitor gene expression. Because C1-esterase inhibitor synthesis is increased by advancing culture time and by the inflammatory mediator interferon-gamma, we suggest that fat-storing cells may enhance the deposition of extracellular matrix proteins by inhibiting their degradation.

Effect of interferon-gamma on complement gene expression in different cell types

We have studied the expression of the complement components C2, C3, factor B, C1 inhibitor (C1-inh), C4-binding protein (C4-bp) and factor H in human peripheral blood monocytes, skin fibroblasts, umbilical vein endothelial cells (HUVEC) and the human hepatoma cell line G2 (Hep G2) in the absence and the presence of interferon-gamma (IFN-gamma). E.l.i.s.a. performed on culture fluids, run-on transcription assays, Northern blot and double-dilution dot-blot techniques confirmed that monocytes expressed all six components, whereas fibroblasts, HUVEC and HepG2 each expressed five of the six components. Fibroblasts and HUVEC did not synthesize C4-bp, and Hep G2 did not produce factor H. In addition to these differences, the synthesis rates of C3, C1-inh and factor H were not the same in all cell types. However, the synthesis rates of C2 and factor B were similar in all four cell types. The half-lives of the mRNAs were shorter in monocytes than in other cell types. Monocyte factor H mRNA had a half-life of 12 min in monocytes, compared with over 3 h in fibroblasts and HUVEC. The instability of factor H mRNA in monocytes may contribute to their low factor H secretion rate. IFN-gamma produced dose-dependent stimulation of C2, factor B, C1-inh, C4-bp and factor H synthesis by all cell types expressing these proteins, but decreased C3 synthesis in all four cell types. Cell-specific differences in the response to IFN-gamma were observed. The increased rates of transcription of the C1-inh and factor H genes in HUVEC were greater than in other cell types, while the increased rate of transcription of the C2, factor B and C1-inh genes in Hep G2 cells was less than in other cell types. IFN-gamma did not affect the stability of C3, factor H or C4 bp mRNAs, but increased the stability of factor B and C1-inh mRNAs and decreased the stability of C2 mRNA. Although these changes occurred in all four cell types studied, the half-life of C1-inh mRNA in monocytes was increased almost 4-fold, whereas the increases in the other cell types were less than 30%. These data show that the constitutive synthesis rates of complement components may vary in the different cell types. They also show that the degree of change in synthesis rates in response to IFN-gamma in each of the cell types often varies due to differences in transcriptional response, sometimes in association with changes in mRNA stability.

Dysfunctional C1 inhibitor Ta: deletion of Lys-251 results in acquisition of an N-glycosylation site

Hereditary angioneurotic edema is inherited as an autosomal dominant disorder and is characterized by potentially life-threatening episodic angioedema. In type II hereditary angioneurotic edema, a dysfunctional C1 inhibitor molecule is present together with low levels of normal C1 inhibitor. About 70% of these dysfunctional proteins result from reactive center (Arg-444) mutations. We describe the deletion of nucleotides encoding Lys-251 (AAG) in C1 inhibitor Ta, the dysfunctional C1 inhibitor from a family with type II hereditary angioneurotic edema. DNA sequence analysis was derived from clones obtained through polymerase chain reaction amplification of blood monocyte C1 inhibitor mRNA. As expected, clones with both normal and abnormal sequence were isolated. The deletion was verified by protein sequence analysis. These data, together with biochemical analysis of the protein and cell-free translation studies, suggest that this deletion, by altering the normal amino acid sequence from Asn-Lys-Ile-Ser to Asn-Ile-Ser, creates a new glycosylation site. The additional carbohydrate accounts for the larger size on SDS/PAGE and very likely interferes with protein function.

Expression of functional cell surface C1-inactivator by U937 cells

We have previously shown that the human monocyte-like cell line U937 synthesizes C1-INA and expresses cell surface C1-INA. In this report we provide evidence that this surface-expressed C1-INA is functionally active. Intact U937 cells demonstrated functional C1-INA activity in a hemolytic assay. This activity was blocked when the cells were incubated with monospecific antibody to C1-INA, and was not detectable in cell-free supernatants of U937 cells. SDS-PAGE analysis of radiolabeled U937 cell surface proteins purified by anti-C1-INA affinity chromatography revealed two distinct bands. One protein had a Mr of 105 kDa identical to plasma C1-INA, and the second had a Mr of 200 kDa. We were unable to determine the identity of the 200 kDa protein by Western blotting with anti-C1-INA. However, the possibility exists that this 200 kDa molecule may represent a C1-INA receptor, a dimeric form of C1-INA, or an unrelated cell surface protein with affinity for C1-INA. Furthermore, we show that treatment of U937 cells with phorbol ester resulted in an increase in the percentage of cells expressing surface C1-INA. These results suggest that U937 cells express functional cell surface C1-INA, which could function in vivo to protect these human tumor cells from lysis by host complement.

Proteolysis and deglycosylation of human C1 inhibitor. Effect on functional properties

The effects of proteolysis and deglycosylation on C1 inhibitor (C1Inh) were tested with respect to both its ability to form complexes with C1s and its capacity to block C1 autoactivation. Limited proteolysis of C1Inh by Staphylococcus aureus V8 proteinase, proline-specific endopeptidase or elastase generated a major high-Mr (approximately 86,000) fragment. In contrast with the fragment produced by elastase, which was inactive, the fragments resulting from V8 proteinase and proline-specific endopeptidase treatment retained activity. Deglycosylation with N-glycanase or O-glycanase, or both, had no major effect on the functional activity of C1Inh.

Molecular cloning of human C1 inhibitor: sequence homologies with alpha 1-antitrypsin and other members of the serpins superfamily

Genetic and acquired diseases in man show that the proteolytic activity of the complement component C1 is crucially regulated by C1 inhibitor (C1-INH), a plasma protein whose suspected relatedness to other serine proteinase inhibitors (serpins) contrasts with its atypically large size and high degree of glycosylation. Indeed we have found that the C1-INH polypeptide precursor synthesized in a cell-free system is a 64-kDa protein, hence it exceeds the length of the precursor forms of typical serpins. Seeking more conclusive sequence information and a probe for the structural locus, we isolated C1-INH cDNA clones from a library representing size-enriched human liver mRNA. Nucleotide sequence analysis of a clone covering the carboxyterminal half of C1-INH conclusively documents the relatedness of this protein with the serpins, and reveals 27% amino acid identity with alpha 1-antitrypsin.