Angiotensin-converting enzyme is a GPI-anchored protein releasing factor crucial for fertilization

The angiotensin-converting enzyme (ACE) is a key regulator of blood pressure. It is known to cleave small peptides, such as angiotensin I and bradykinin and changes their biological activities, leading to upregulation of blood pressure. Here we describe a new activity for ACE: a glycosylphosphatidylinositol (GPI)-anchored protein releasing activity (GPIase activity). Unlike its peptidase activity, GPIase activity is weakly inhibited by the tightly binding ACE inhibitor and not inactivated by substitutions of core amino acid residues for the peptidase activity, suggesting that the active site elements for GPIase differ from those for peptidase activity. ACE shed various GPI-anchored proteins from the cell surface, and the process was accelerated by the lipid raft disruptor filipin. The released products carried portions of the GPI anchor, indicating cleavage within the GPI moiety. Further analysis by high-performance liquid chromatography-mass spectrometry predicted the cleavage site at the mannose-mannose linkage. GPI-anchored proteins such as TESP5 and PH-20 were released from the sperm membrane of wild-type mice but not in Ace knockout sperm in vivo. Moreover, peptidase-inactivated E414D mutant ACE and also PI-PLC rescued the egg-binding deficiency of Ace knockout sperms, implying that ACE plays a crucial role in fertilization through this activity.     

Implications of lipid moiety in oligomerization and immunoreactivities of GPI-anchored proteins

Glycosylphosphatidylinositol (GPI) enriches GPI-anchored proteins (GPI-AP) in lipid rafts by intimate interaction of its lipid moiety with sphingolipids and cholesterol. In addition to such lipid-lipid interactions, it has been reported that GPI may interact with protein moiety linked to GPI and affect protein conformations because GPI delipidation reduced immunoreactivities of protein. Here, we report that GPI-APs that have not undergone fatty acid remodeling exhibit reduced immunoreactivities in Western blotting, similar to delipidated proteins, compared with normal remodeled GPI-APs. In contrast, immunostaining in flow cytometry and immunoprecipitation did not show significant differences between remodeled and unremodeled GPI-APs. Moreover, detection with premixed primary/secondary antibody complexes or Fab fragments eliminated this difference in Western blotting. These results indicate that normally remodeled GPI enhanced oligomerization of GPI-APs and that inefficient oligomerization of unremodeled GPI-APs was responsible for reduced immunoreactivities. Moreover, the reduction in immunoreactivities of delipidated GPI-APs was most likely caused by the same effect. Finally, by chemical cross-linking of surface proteins in living cells and cell killing assay using a pore-forming bacterial toxin, we showed that enhanced oligomerization by GPI-remodeling occurs under a physiological membrane environment. Thus, this study clarifies the significance of GPI fatty acid remodeling in oligomerization of GPI-APs and provides useful information for technical studies of these cell components.     

Vaccination against babesiosis using recombinant GPI-anchored proteins

The increase in human babesiosis is of major concern to health authorities. In the USA, most of these cases are due to infections with Babesia microti, whereas in Europe B. divergens is the major cause of clinical disease in humans. Here we review the immunological and biological literature of glycosylphosphatidylinositol (GPI)-anchored merozoite proteins of human Babesia parasites with emphasis on their role in immunity, and provide some new bioinformatical information on B. microti GPI-Anchored Proteins (GPI-AP). Cattle can be vaccinated with soluble parasite antigens (SPA) of Babesia divergens that are released by the parasite during proliferation. The major component in SPA preparations appeared to be a 37?kDa merozoite surface protein that is anchored in the merozoite membrane by a GPI anchor. Animals could be protected by vaccination with the recombinant 37?kDa protein expressed in Escherichia coli, provided the protein had a hydrophobic terminal sequence. Based on this knowledge, a recombinant vaccine was developed against Babesia canis infection in dogs, successfully. In order to identify similar GPI-AP in B. microti, the genome was analysed. Here it is shown that B. microti encodes all proteins necessary for GPI assembly and its subsequent protein transfer. In addition, in total 21 genes encoding for GPI-AP were detected, some of which reacted particularly strongly with sera from B. microti-infected human patients. Reactivity of antibodies with GPI-anchored merozoite proteins appears to be dependent on the structural conformation of the molecule. It is suggested that the three-dimensional structure of the protein that is anchored in the membrane is different from that of the protein that has been shed from the merozoite surface. The significance of this protein’s dynamics in parasite biology and immune evasion is discussed. Finally, we discuss developments in tick and Babesia vaccine research, and the role such vaccines could play in the control of human babesiosis.     

Testicular Angiotensin-converting enzyme with different glycan modification: characterization on glycosylphosphatidylinositol-anchored protein releasing and dipeptidase activities

We have previously found that the angiotensin-converting enzyme (ACE) carries GPI-anchored protein releasing activity (GPIase) as well as dipeptidase activity. Testicular ACE (tACE), the male germinal specific isozyme, plays a crucial role in male fertilization. The amino-terminal region of this isozyme is different from that of somatic isozyme (sACE) and contains potential O-linked glycosylation sites. By multiple mutagenesis after an in silico prediction, amino acid residues acquiring O-glycans were assigned. Both GPIase and dipeptidase activities were compared between O-glycan null mutant and wild-type molecules, but no differences were found. Furthermore, the wild-type tACE was produced in two different cells (COS7 and CHO) and its activities compared. The GPIase activity, but not dipeptidase, was apparently higher for CHO-derived molecule than COS7. Sensitivity to neuraminidase and O-glycosidase digestions and the profile of glycosylation were quite different between these two molecules. Moreover, serial digestions with neuraminidase and O-glycosidase have no influence on GPIase activity of both molecules, suggesting that the sialylation and the presence of O-glycan has no influence on tACE enzyme activities, while the set of glycans modulate GPIase activity.     

Gel chromatographic characterization of the hydrophobic interaction of glycosylphosphatidylinositol-alkaline phosphatase with detergents

The interaction of a glycosylphosphatidylinositol (GPI) protein with different detergents was studied for the first time with a purified protein. Four differently hydrophobic fractions of GPI-alkaline phosphatase (GPI-AP) from calf intestine were used as model proteins. The mode of interaction was determined by investigating (i) the self-aggregation behaviour of the GPI-AP fractions, (ii) the interference of detergents with GPI-AP binding to octyl-Sepharose, and (iii) the elution of GPI-AP bound to octyl-Sepharose. It was shown that polyoxyethylene-type detergents surprisingly interact much stronger than n-octylglucoside with GPI-AP, which is in contrast to the known behaviour of GPI-proteins in natural membranes. Gel filtration chromatography of Triton X-100 at concentrations above the critical micellar concentration yields three different micelle species with apparent molecular weights of about 166, 54, and 16 kDa. GPI-AP fraction II, which is shown to bear only one anchor per dimer, does not bind to any of these micelles. We demonstrate that a complex is formed containing about 150 Triton X-100 molecules and about 4700 molecules of water per molecule of GPI-AP dimer. The experimental findings are in accordance with a simple geometrical model based on the physical data of fatty acids and the arrangement, mean size, and shape of Triton X-100 molecules.     

Inherited glycosylphosphatidyl inositol deficiency: A treatable CDG

Tethering to cell membrane through attachment to the complex glycolipid anchor glycosylphosphatidyl inositol (GPI) is a mode of protein expression highly conserved in eukaryotes. The evolutionary purpose of such an elaborate way of expressing proteins is not clear and neither is the functional role of GPI itself. GPI-anchored proteins (GPI-AP) serve a variety of functions that include adhesion, receptors, signal transduction and complement activation. GPI biosynthesis, a process that is accomplished in at least 9 steps and involves several proteins, some with enzymatic activity, would be expected to be a fertile ground for development of inherited, autosomal recessive disorders. However, until recently, paroxysmal nocturnal haemoglobinuria, a rare haematological disorder caused by somatic mutations in the X-linked PIGA gene, was the only genetic disorder affecting GPI biosynthesis. Here we review the clinical spectrum, biochemical defect and genetic pathogenesis of inherited GPI deficiency, the first described form of inherited, autosomal recessive disorder of GPI biosynthesis and outline the molecular basis of targeted therapy for this condition.     

Ultrastructural cytochemical and autoradiographic demonstration of nonspecific esterase(s) in guinea pig basophils

We used ultrastructural autoradiographic and cytochemical methods to localize esterase activities in unstimulated guinea pig basophils and in basophils undergoing degranulation or recovery from degranulation. We used tritium-labeled diisopropylfluorophosphate (DFP) as a probe for serine enzymes and localized this probe by ultrastructural autoradiography to cytoplasmic granules of immature or mature unstimulated basophils, as well as to granules released by degranulating basophils. Ultrastructural cytochemistry using alpha naphthyl acetate (ANA) as substrate localized nonspecific esterase activity to extruded granules, either within the interiors of degranulation sacs or within granules completely separated from degranulating basophils. Extruded granules retained their esterase activity for as long as 24 hr after antigen-induced degranulation. The plasma membranes of unstimulated or degranulating basophils, as well as of basophils recovering from degranulation, displayed prominent cell surface ANA esterase ectoenzyme activity. Lipid bodies, organelles present in the cytoplasm of both control and recovering basophils, were also alpha naphthyl acetate esterase (ANAE)-positive. Thus, cytochemical and autoradiographic techniques localized esterase and/or [3H]-DFP-binding activities to cytoplasmic granules, lipid bodies, and cell surface of basophils, and these enzyme activities persisted during both degranulation and recovery from degranulation.     

Release of glycosylphosphatidylinositol-anchored carboxypeptidase M by phosphatidylinositol-specific phospholipase C upregulates enzyme synthesis

Carboxypeptidase M (CPM), a glycosylphosphatidylinositol (GPI)-anchored membrane protein, remained at a constant level in confluent Madin Darby canine kidney (MDCK) cells but was continually released into the medium in soluble form. The released CPM contained ethanolamine, indicating liberation by a phospholipase. Treatment of MDCK cells with 0.01 U/ml phosphatidylinositol-specific phospholipase C for 6 h led to a 5.5-fold increase in soluble CPM, yet the activity in cells remained constant, resulting in a 30% increase in total activity. The increase was due to new protein synthesis as evidenced by inhibition with 0.2 microM cycloheximide and a 63% increase in [35S]methionine incorporation into newly synthesized CPM. MDCK cells treated with 1-alkyl-2-acyl-glycerol, the diglyceride component of mammalian glycosylphosphatidylinositol anchors, exhibited a 36% increase in CPM activity, but diacylglycerols or phorbol esters were ineffective. Thus, release of GPI-anchored CPM can generate a diglyceride signal to replenish and maintain constant levels on the cell surface.     

Sorting of GPI-anchored proteins into ER exit sites by p24 proteins is dependent on remodeled GPI

Glycosylphosphatidylinositol (GPI) anchoring of proteins is a posttranslational modification occurring in the endoplasmic reticulum (ER). After GPI attachment, proteins are transported by coat protein complex II (COPII)-coated vesicles from the ER. Because GPI-anchored proteins (GPI-APs) are localized in the lumen, they cannot interact with cytosolic COPII components directly. Receptors that link GPI-APs to COPII are thought to be involved in efficient packaging of GPI-APs into vesicles; however, mechanisms of GPI-AP sorting are not well understood. Here we describe two remodeling reactions for GPI anchors, mediated by PGAP1 and PGAP5, which were required for sorting of GPI-APs to ER exit sites. The p24 family of proteins recognized the remodeled GPI-APs and sorted them into COPII vesicles. Association of p24 proteins with GPI-APs was pH dependent, which suggests that they bind in the ER and dissociate in post-ER acidic compartments. Our results indicate that p24 complexes act as cargo receptors for correctly remodeled GPI-APs to be sorted into COPII vesicles.     

Interaction of phosphoinositolglycan(-peptides) with plasma membrane lipid rafts of rat adipocytes

Insulin receptor-independent activation of the insulin signal transduction cascade in insulin-responsive target cells by phosphoinositolglycans (PIG) and PIG-peptides (PIG-P) is accompanied by redistribution of glycosylphosphatidylinositol (GPI)-anchored plasma membrane proteins (GPI proteins) and dually acylated nonreceptor tyrosine kinases from detergent/carbonate-resistant glycolipid-enriched plasma membrane raft domains of high-cholesterol content (hcDIGs) to rafts of lower cholesterol content (lcDIGs). Here we studied the nature and localization of the primary target of PIG(-P) in isolated rat adipocytes. Radiolabeled PIG-P (Tyr-Cys-Asn-NH-(CH(2))(2)-O-PO(OH)O-6Manalpha1(Manalpha1-2)-2Manalpha1-6Manalpha1-4GluN1-6Ino-1,2-(cyclic)-phosphate) prepared by chemical synthesis or a radiolabeled lipolytically cleaved GPI protein from Saccharomyces cerevisiae, which harbors the PIG-P moiety, bind to isolated hcDIGs but not to lcDIGs. Binding is saturable and abolished by pretreatment of intact adipocytes with trypsin followed by NaCl or with N-ethylmaleimide, indicating specific interaction of PIG-P with a cell surface protein. A 115-kDa polypeptide released from the cell surface by the trypsin/NaCl-treatment is labeled by [(14)C]N-ethylmaleimide. The labeling is diminished upon incubation of adipocytes with PIG-P which can be explained by direct binding of PIG-P to the 115-kDa protein and concomitant loss of its accessibility to N-ethylmaleimide. Binding of PIG-P to hcDIGs is considerably increased after pretreatment of adipocytes with (glycosyl)phosphatidylinositol-specific phospholipases compatible with lipolytic removal of endogenous ligands, such as GPI proteins/lipids. These data demonstrate that in rat adipocytes synthetic PIG(-P) as well as lipolytically cleaved GPI proteins interact specifically with hcDIGs. The interaction depends on the presence of a trypsin/NaCl/NEM-sensitive 115-kDa protein located at hcDIGs which thus represents a candidate for a binding protein for exogenous insulin-mimetic PIG(-P) and possibly endogenous GPI proteins/lipids.     

The function of GPI-anchored proteins in T cell development, activation and regulation of homeostasis

Many glycosylphosphatidyl-inositol-anchored proteins (GPI-AP) are expressed on T lymphocytes. Ligand or mAb-mediated aggregation of all GPI-AP tested to date results in the initiation of signal transduction pathways via the activation of src family protein tyrosine kinases. Src family kinases co-localise with GPI-AP in specialised sub-domains of the plasma membrane, referred to as detergent insoluble membrane microdomains (DIGS), which are thought to function as signalling platforms. GPI-AP may play a role in the regulation of T cell clonal expansion and effector functions at multiple levels, including the initiation of T cell activation through the antigen receptor complex, the regulation of ongoing responses supported by persisting antigen, as well as proliferative responses to the major T cell growth factor, IL-2. Evidence supporting the role of GPI-AP in the regulation of T cell development, activation and homeostasis is discussed, as well as insights provided by studies in humans and mice lacking GPI-AP.     

The antibacterial activity of peptides derived from human beta-2 glycoprotein I is inhibited by protein H and M1 protein from Streptococcus pyogenes

During the last years, the importance of antibacterial peptides has attracted considerable attention. We report here that peptides derived from the fifth domain of beta-2 glycoprotein I (beta(2)GPI), a human heparin binding plasma protein, have antibacterial activities against Gram-positive and Gram-negative bacteria. Streptococcus pyogenes, an important human pathogen that can survive and grow in human blood, has developed mechanisms to escape the attack by these peptides. Thus, protein H and M1 protein, two surface proteins of the highly pathogenic S. pyogenes AP1 strain, bind full-length beta(2)GPI and thereby prevent the processing of beta(2)GPI by proteases from polymorphonuclear neutrophils (PMNs) into antibacterial peptides. In addition, protein H and M1 protein, released from the bacterial cell wall by PMN-derived proteases, bind to, and inhibit the activity of, beta(2)GPI-derived antibacterial peptides. Taken together, the data suggest that the interaction between the streptococcal proteins and beta(2)GPI or beta(2)GPI-derived peptides presents a novel mechanism to resist an antibacterial attack by beta(2)GPI-cleavage products.     

Partial isolation from intact cells of a cell surface-exposed lysophosphatidylinositol-phospholipase C

A novel cell surface phosphoinositide-cleaving phospholipase C (ecto-PLC) activity was isolated from cultured cells by exploiting its presumed external exposure. Biotinylation of intact cells followed by solubilization of the biotinylated proteins from a membrane fraction and recovery onto immobilized-avidin beads, allowed assay of this cell surface enzyme activity apart from the background of the substantial family of intracellular PLCs. Several cell lines of differing ecto-PLC expression were examined as well as cells stably transfected to overexpress the glycosylphosphatidylinositol (GPI)-anchored protein human placental alkaline phosphatase (PLAP) as a cell surface enzyme marker. The resulting bead preparations from ecto-PLC positive cells possessed calcium-dependent PLC activity with preference for lysophosphatidylinositol (lysoPI) rather than phosphatidylinositol (PI). The function of ecto-PLC of intact cells evidently is not to release GPI-anchored proteins at the cell surface, as no detectable Ca²⁺-dependent release of overexpressed PLAP from ecto-PLC-positive cells was observed. To investigate the cell surface linkage of the ecto-PLC itself, intact cells were treated with bacterial PI-PLC to cleave simple GPI anchors, but no decrease in ecto-PLC activity was observed. High ionic strength washes of biotinylated membranes prior to the generation of bead preparations did not substantially reduce the lysoPI-PLC activity. The results verify that the ecto-PLC is truly cell surface-exposed, and unlike other members of the PLC family that are thought to be peripheral membrane proteins, this novel lysoPI-PLC is most likely a true membrane protein. J. Cell. Biochem. 65:550–564. © 1997 Wiley-Liss Inc.     

ARF1 is directly involved in dynamin-independent endocytosis

Endocytosis of glycosylphosphatidyl inositol (GPI)-anchored proteins (GPI-APs) and the fluid phase takes place primarily through a dynamin- and clathrin-independent, Cdc42-regulated pinocytic mechanism. This mechanism is mediated by primary carriers called clathrin-independent carriers (CLICs), which fuse to form tubular early endocytic compartments called GPI-AP enriched endosomal compartments (GEECs). Here, we show that reduction in activity or levels of ARF1 specifically inhibits GPI-AP and fluid-phase endocytosis without affecting other clathrin-dependent or independent endocytic pathways. ARF1 is activated at distinct sites on the plasma membrane, and by the recruitment of RhoGAP domain-containing protein, ARHGAP10, to the plasma membrane, modulates cell-surface Cdc42 dynamics. This results in the coupling of ARF1 and Cdc42 activity to regulate endocytosis at the plasma membrane. These findings provide a molecular basis for a crosstalk of endocytosis with secretion by the sharing of a key regulator of secretory traffic, ARF1.     

Variations in the amount of glucose phosphate isomerase in lymphocytes and erythrocytes from A/J and C57BL/6J mice

In a comparative study of A/J (Gpi-1a) and C57BL/6J (Gpi-1b) mice, we observed that erythrocytes of A/J mice exhibited significantly higher glucose phosphate isomerase (GPI) activity compared to erythrocytes of C57BL/6J mice on a per cell, per gram of protein, or per gram of hemoglobin basis. Higher GPI activity per cell was detected for peripheral blood lymphocytes of A/J compared to C57BL/6J mice. (A/J X C57BL/6J)F1 mice expressed erythrocyte and peripheral blood lymphocyte GPI activities intermediate to those of the parental mouse strains. The GPI activities of spleen lymphocytes from A/J, C57BL/6J, or (A/J X C57BL/6J)F1 mice were not significantly different from each other. The higher activity in the A/J mice could be due to GPI of a higher catalytic rate or to the presence of more GPI molecules. In order to distinguish these two possibilities, GPI was purified to homogeneity from both strains of mice. The specific activities (activity per milligram of protein) of the purified enzymes from the two strains were found to be similar, indicating that GPI from the A/J strain was not a more active enzyme. Antibody to the purified enzymes was prepared and used in an enzyme-linked immunosorbent assay (ELISA) to compare the relative amounts of enzyme molecules in cells of A/J and C57BL/6J mice. Results of the ELISA tests on peripheral blood lymphocytes indicated that A/J mice contain more molecules of GPI per cell and, therefore, have a higher GPI activity than C57BL/6J mice.     

Expression of bloodstream variant surface glycoproteins in procyclic stage Trypanosoma brucei: role of GPI anchors in secretion.

Using transformed procyclic trypanosomes, the synthesis, intracellular transport and secretion of wild-type and mutant variant surface glycoprotein (VSG) is characterized. We find no impediment to the expression of this bloodstream stage protein in insect stage cells. VSG receives a procyclic-type phosphatidylinositol-specific phospholipase C-resistant glycosyl phosphatidylinositol (GPI) anchor, dimerizes and is N-glycosylated. It is transported to the plasma membrane with rapid kinetics (t(1/2) approximately 1 h) and then released by a cell surface zinc-dependent metalloendoprotease activity, a possible homolog of leishmanial gp63. Deletion of the C-terminal GPI addition signal generates a soluble form of VSG that is exported with greatly reduced kinetics (t(1/2) approximately 5 h). Fusion of the procyclic acidic repetitive protein (PARP) GPI anchor signal to the C-terminus of the truncated VSG reporter restores both GPI addition and transport competence, suggesting that GPI anchors play a critical role in the folding and/or forward transport of newly synthesized VSG. The VSG-PARP fusion is also processed near the C-terminus by events that do not involve N-linked oligosaccharides and which are consistent with GPI side chain modification. This unexpected result suggests that GPI processing may be influenced by adjacent peptide sequence or conformation.     

Catabolism of 4-hydroxy-2-trans-nonenal by THP1 monocytes/macrophages and inactivation of carboxylesterases by this lipid electrophile

Oxidative stress in cells and tissues leads to the formation of an assortment of lipid electrophiles, such as the quantitatively important 4-hydroxy-2-trans-nonenal (HNE). Although this cytotoxic aldehyde is atherogenic the mechanisms involved are unclear. We hypothesize that elevated HNE levels can directly inactivate esterase and lipase activities in macrophages via protein adduction, thus generating a biochemical lesion that accelerates foam cell formation and subsequent atherosclerosis. In the present study we examined the effects of HNE treatment on esterase and lipase activities in human THP1 monocytes/macrophages at various physiological scales (i.e., pure recombinant enzymes, cell lysate, and intact living cells). The hydrolytic activities of bacterial and human carboxylesterase enzymes (pnbCE and CES1, respectively) were inactivated by HNE in vitro in a time- and concentration-dependent manner. In addition, so were the hydrolytic activities of THP1 cell lysates and intact THP1 monocytes and macrophages. A single lysine residue (Lys105) in recombinant CES1 was modified by HNE via a Michael addition reaction, whereas the lone reduced cysteine residue (Cys389) was found unmodified. The lipolytic activity of cell lysates and intact cells was more sensitive to the inhibitory effects of HNE than the esterolytic activity. Moreover, immunoblotting analysis using HNE antibodies confirmed that several cellular proteins were adducted by HNE following treatment of intact THP1 monocytes, albeit at relatively high HNE concentrations (>50 μM). Unexpectedly, in contrast to CES1, the treatment of a recombinant human CES2 with HNE enhanced its enzymatic activity ∼3-fold compared to untreated enzyme. In addition, THP1 monocytes/macrophages can efficiently metabolize HNE, and glutathione conjugation of HNE is responsible for ∼43% of its catabolism. The functional importance of HNE-mediated inactivation of cellular hydrolytic enzymes with respect to atherogenesis remains obscure, although this study has taken a first step toward addressing this important issue by examining the potential of HNE to inhibit this biochemical activity in a human monocyte/macrophage cell line.     

Activity-based Protein Profiling Identifies a Host Enzyme,Carboxylesterase 1, Which Is Differentially Active during Hepatitis C Virus Replication

Hepatitis C virus (HCV) relies on many interactions with host cell proteins for propagation. Successful HCV infection also requires enzymatic activity of host cell enzymes for key post-translational modifications. To identify such enzymes, we have applied activity-based protein profiling to examine the activity of serine hydrolases during HCV replication. Profiling of hydrolases in Huh7 cells replicating HCV identified CES1 (carboxylesterase 1) as a differentially active enzyme. CES1 is an endogenous liver protein involved in processing of triglycerides and cholesterol. We observe that CES1 expression and activity were altered in the presence of HCV. The knockdown of CES1 with siRNA resulted in lower levels of HCV replication, and up-regulation of CES1 was observed to favor HCV propagation, implying an important role for this host cell protein. Experiments in HCV JFH1-infected cells suggest that CES1 facilitates HCV release because less intracellular HCV core protein was observed, whereas HCV titers remained high. CES1 activity was observed to increase the size and density of lipid droplets, which are necessary for the maturation of very low density lipoproteins, one of the likely vehicles for HCV release. In transgenic mice containing human-mouse chimeric livers, HCV infection also correlates with higher levels of endogenous CES1, providing further evidence that CES1 has an important role in HCV propagation.