Background
Hepatitis C virus (HCV) is classified in the
Hepacivirus genus within the
Flaviviridae family. The viral genome constitutes a 9.6-kb single-stranded positive-sense RNA with 5' and 3' noncoding regions and a long open reading frame encoding a polyprotein precursor of about 3,000 amino acids in length. The HCV polyprotein precursor is co- and post-translationally processed by cellular and viral proteases to yield 11 viral proteins [
1,
2]. The structural HCV proteins include the core protein and transmembrane glycoproteins, E1 and E2. The core region also encodes for an alternative open reading frame protein (ARFP) or F protein whose function is presently not known [
1,
3]. The region between the structural and non-structural genes encodes for an integral membrane cation channel protein p7 [
4] which is essential for virus production [
5]. HCV has six nonstructural proteins; NS2, NS3, NS4A, NS4B, NS5A and NS5B (see for reviews; [
2,
6]. NS2 is a cysteine protease responsible for an autoproteolytic NS2–NS3 cleavage and it requires the aminoterminal one-third of NS3 for its enzymatic activity. NS3 is a multifunctional protein with both serine protease and RNA helicase/NTPase activities and NS4A is as an essential cofactor for NS3 protease functions. Currently, there is little information of the function of NS4B protein, but it participates in the formation of a membranous web where HCV RNA replication is suggested take place [
6,
7]. NS5A is a phosphoprotein which takes part in virus particle formation and is involved in virus resistance against interferons [
8]. The NS5B protein encodes for an RNA-dependent RNA polymerase (RdRp), which is the central catalytic enzyme of the HCV replicase [
9,
10].
Generally, HCV is divided into six major genotypes (or clades) that can be further divided into several subtypes from A to L [
11,
12]. The amino acid sequences of the major HCV genotypes differ approximately 30% from each other [
11]. The geographical distribution of HCV genotypes is also diverse. The genotypes 1, 2 and 3 are found throughout the world whereas the distribution of the other genotypes is much more restricted; genotype 4 is found in the Middle East and Africa, genotype 5 in South Africa and genotype 6 in Southeast Asia [
11,
13]. In the United States less than 1% of HCV patients are infected with the HCV genotypes 4, 5 or 6 [
14]. However, the epidemiology of HCV infection is changing continuously, which is e.g. seen in a manner that the number of genotype 4 infected patients has increased in Europe as a consequence of increasing immigration and intravenous drug use during the last 15 years [
15]. The overall worldwide prevalence of HCV is approximately 3%. The highest HCV prevalence figures up to 10–20%, are found in Egypt where the genotype 4 is the most common one [
16]. The prevalence of HCV infection varies remarkably and for instance in different European countries it ranges from 0,1% to 4% [
15]. Acute HCV infection can be cleared spontaneously only in up to 15–30% of the cases, while usually the infection becomes chronic. Within 20 to 30 years chronic HCV infection can progress to cirrhosis in 20% of the patients leading to hepatocellular carcinoma roughly in yearly rate of 1–4%. Although the commercial methodology to detect HCV-specific RNA and antibody responses in patient sera has greatly advanced in recent years there is no detailed information of the immunogenicity of different HCV proteins in patients suffering from chronic HCV infection.
In the present work, we have described the expression and purification of nine different recombinant HCV proteins in insect cells and analyzed humoral immune response against each viral protein using Western blotting in patients suffering from chronic HCV infection of genotypes 1, 2, 3 or 4. We found that most of the 68 HCV RNA and antibody positive patient sera studied recognized the core, NS3, NS4B and NS5A proteins with high titers. Instead, only three sera recognised NS5B and none of the sera recognized NS2 protein. These results show that antibody responses to various HCV proteins show considerable qualitative and quantitative differences with certain proteins being highly immunogenic in practically all HCV-infected individuals while certain proteins such as NS2 and NS5B were virtually devoid of all immunogenic activity.
Methods
Cell culture
Monolayers and suspension cultures of
Spodoptera frugiperda Sf 9 cells were maintained in TNM-FH medium and 10% fetal calf serum (Integro, Zaandam, Netherlands) as described [
17].
Construction of expression plasmids for different HCV genes
Different HCV genes were amplified with PCR from pBRTM/HCV1-3011 [
18] carrying the HCV genotype 1b cDNA, and the PCR products were subcloned into the
Bam HI site of the pcDNA3.1(+)-FLAG plasmid under CMV promoter [
19]. The primers (Dako A/S, Glostrup), which were used to modify the 5' and 3'ends of core, E1, E2, NS2, NS3, NS4A, NS4B, NS5A, and NS5B genes have been described elsewhere [
20]. After partial sequencing, the HCV protein-coding cDNAs (core, E1, E2, NS2, NS3, NS4A, NS4B, NS5A, NS5B) were subcloned into the
Bam HI site of the pAcYM1 baculovirus transfer vector under the control of polyhedrin promoter [
17]. NS2 protein was expressed with a His-tag. To create recombinant HCV protein-expressing viruses pAcYM1 expression constructs were cotransfected with linearized baculovirus DNA using BaculoGold™ Transfection Kit (PharMingen, San Diego, CA) and recombinant viruses were obtained. All DNA manipulations were performed according to standard protocols.
In vitro translation of the HCV genes cloned into pcDNA3.1(+)-FLAG plasmid constructs was carried out with T7 Cap-Scribe and reticulocyte translation kit (Boehringer Mannheim GmbH, Mannheim, Germany). After translation, the samples were diluted in Laemmli sample buffer and analyzed by SDS-PAGE.
Production and purification of recombinant HCV proteins
Sf 9 cells were grown to confluence in plastic cell culture bottles (175 cm
2), infected with HCV core, E1, E2, NS2, NS3, NS4A, NS4B, NS5A, and NS5B expressing recombinant baculoviruses for 1 h and grown for 72 h to produce the recombinant proteins [
17]. The cells were collected by centrifugation at 1500 rpm for 10 min followed by washing with phosphate-buffered saline (PBS). The cells were processed further immediately or stored at -70°C. The cells were sonicated on ice, and concentrations of total cellular proteins were quantified with the Bio-Rad protein assay (Bio-Rad Laboratories, Richmond, CA). Expression of recombinant HCV proteins was verified with Coomassie Blue staining, metabolic labeling with [
35S]-methionine, and Western blotting.
5 mg of sonicated, cellular protein samples in Laemmli sample buffer was purified using preparative SDS-PAGE (Model 491 Prep Cell, Bio-Rad Laboratories). The recombinant HCV core, E1, E2, NS2, NS3, NS4A, NS4B, NS5A, and NS5B proteins were separated on 6 to 15% gradient SDS-PAGE. The sample fractions containing separated proteins from preparative SDS-PAGE were first lyophilized followed by resuspension into 0.5 ml of water. The purity and quantity of each sample fraction were verified with Coomassie Blue staining (compared to known standard protein) and with Western blotting, using specific immunosera. To reduce the amount of SDS in the lyophilized samples, each protein fraction was concentrated with Millipore protein concentration kit UFV5BCC25 (Millipore, Bedford, MA).
HCV antibodies
Primary antibodies used in Western analysis were rabbit anti-HCV core and NS5A [
20], mouse anti-FLAG M5 (for the detection of
in vitro translated HCV E1, E2, NS2, NS3, NS4A, NS4B, and NS5B proteins; Sigma Chemical Co., St. Louis, MO) and mouse anti-Penta-His (for the detection of 6xHis-NS2; Qiagen, Venlo, Netherlands). Secondary Abs were HRP-conjugated goat anti-rabbit and anti-mouse immunoglobulin (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) and HRP-conjugated goat anti-human IgG (H+L) (Vector Laboratories, Inc., Burlingame, CA).
HCV positive and negative human sera
Altogether 68 HCV RNA and antibody positive patients with various HCV genotypes were studied. Five of these patients were treated with interferon-α monotherapy [
21,
22] for 12 months in the case of genotype 1 infection and for 6 months in genotype 3 infection. The serum samples from these five patients were collected in the beginning and in the end of treatment and also 6 and/or 12 months after treatment. HCV antibodies were determined with commercial tests according to the manufacturer's instructions (Architect Anti-HCV, Abbott, Wiesbaden, Germany; Innotest HCV Ab IV, Innogenetics, Ghent, Belgium; Inno-LIA HCV Ab III update, Innogenetics, Ghent, Belgium). HCV RNA detection was performed by Cobas Amplicor HCV Test, Roche, and genotyping by Inno-LIPA, Innogenetics, Ghent, Belgium. Samples from 50 HCV antibody negative patients served as negative controls.
For safety reasons HCV RNA and antibody positive and HCV antibody negative human sera from patients were inactivated by heating the samples at 56°C for 1 h in the presence of 0.1% Triton X-100. To avoid repeated freezing and thawing an equal volume of 100% glycerol was added on inactivated sera and the 1:2 diluted serum specimens were stored at -20°C in a liquid form.
Detection of HCV antibodies in human sera by Western blotting using baculovirus-produced recombinant HCV proteins and commercial INNO-LIA Score test
To analyse humoral immune responses against HCV proteins serum specimens were obtained from 68 HCV RNA and antibody positive patients. For Western blot analysis 1 μg of each purified HCV protein was loaded onto two Novex pre-cast, preparative 10–20% Tris-glycine polyacrylamide gels (Invitrogen Corp., Carlsbad, CA). The core (21 kDa), NS2 (24 kDa), NS3 (68 kDa), NS4A (6 kDa), NS4B (29 kDa) and NS5A (49 kDa) proteins were loaded on one gel, and E1 (21 kDa), E2 (40 kDa) and NS5B (64 kDa) proteins to another gel. 10 μg of
Sf 9 cell extract was also loaded onto a separate gel as a control. Proteins separated on gels were transferred onto Immobilon-P membranes (polyvinylidine difluoride; Millipore) with an Isophor electrotransfer device (Hoefer Scientific Instruments, San Francisco, CA). The membranes were sliced and stained with HCV RNA and antibody positive patient sera (dilutions of 1:100, 1:500, 1:2500, 1:12500 and 1:62500) in PBS containing 5% nonfat milk at room temperature for 2 h. After washing with PBS, secondary peroxidase-conjugated anti-human IgG antibodies (Vector Laboratories, Burlingame, Inc., CA) were allowed to bind at room temperature for 1 h. After washing with PBS, the bands were visualized by 3-amino-9-ethylcarbazole (AEC) [
23] or the enhanced chemiluminescence system (ECL) (Amersham, Buckinghamshire, UK) as recommended by the manufacturer.
Also 50 HCV negative human sera were analysed as controls. 1 μg of each purified core, NS3, NS4B and NS5A HCV proteins was loaded onto 15% SDS-PAGE gels and blotted to Immobilon-P membranes. The membranes were sliced and stained with negative human sera diluted in 1:100 and 1:500.
These HCV RNA and antibody positive and HCV antibody negative human serum samples were also analysed with INNO-LIA™ * HCV Ab III update or INNO-LIA™ * HCV Score test according to the manufacturer's instructions (Innogenetics, Ghent, Belgium).
Discussion
In the present work, we have expressed recombinant HCV core, E1, E2, NS2, NS3, NS4A, NS4B, NS5A and NS5B proteins by baculovirus system in insect cells. This expression system was chosen in order for the recombinant proteins to undergo all possible posttranslational modifications such as glycosylation and phosphorylation. In addition, in contrast to proteins expressed in
E. coli, humans are likely to have very low or nonexisting antibodies against insect cell proteins that might be contaminating the recombinant protein preparations. The expression levels of individual HCV proteins were relatively high and they could be purified by preparative SDS-PAGE (Fig.
1). Some, but not all HCV genomes are also encoding protein F from an alternative reading frame of the core sequence. For unknown reasons we were not able to produce the F protein by baculovirus expression and therefore we could not include this protein in our analyses. We also did not express the small ion channel protein p7.
We used full-length recombinant HCV proteins from HCV genotype 1b to analyze antibody responses against individual viral proteins in patients suffering from chronic infection caused by HCV genotypes 1, 2, 3 or 4. Our analyzes revealed that all recombinant HCV proteins except that of NS2 were immunogenic in humans and there were no major differences in the magnitude of immune responses at least against the core and NS proteins between the different genotype infections. It was of interest that NS2 protein appears to completely lack immunogenicity in humans. This was unexpected, but yet we are confident with the results, since the sequence of NS2 expression construct was correct and monoclonal anti-NS2 antibodies readily detected the transiently expressed protein [
24]. This may indicate that in humans there may be proteases or other molecules homologous to NS2 leading to an inability of the host to recognize the NS2 protein as foreign. Further evidence that baculovirus expressed recombinant proteins of HCV genotype 1b are suitable for immunological analyses was obtained from the comparison of our Western blot analysis with the commercial INNO-LIA Score test, which is able to detect antibodies from genotypes 1–5. These methods showed a very good correlation in the case of anti-core, NS4A+B and NS5A antibody responses. This suggests that the baculovirus produced HCV proteins provide valuable and very specific research reagents for analyzing HCV-specific immune responses against HCV. However, the INNO-LIA Score test was more sensitive than the Western blot method in the case of anti-NS3 and E2 antibodies. The reason for this discrepancy is not known, but it may be that the relative amount of HCV antigens used in the INNO-LIA assay was higher that what we used in the Western blot analysis. In addition, the conformation of the recombinant proteins may also contribute to the results, since it is known that many antigenic epitopes in viral envelope glycoproteins like the E2 of HCV are likely conformational and thus these sites are not necessarily detected by antibodies in denatured proteins. By increasing the amount of viral antigens in Western blot analysis we could theoretically have been able to enhance the sensitivity of our analysis. However, the idea in our analysis was to systematically study the immunogenicity of various HCV proteins in order to reveal which viral proteins are the targets of humoral anti-HCV immune responses in humans. In order to detect the relative immunogenicity of different HCV proteins we used a similar amount of each purified protein in the assay. Based on this analysis we were able to determine qualitative and quantitative differences in host antibody responses to different HCV proteins, which had not been systematically studied before.
Previously, anti-HCV antibody responses have been analysed in acute and chronic phases of HCV infection [
25‐
28]. In the present study we focused on patients suffering from a chronic HCV infection and we found remarkable differences in the frequency of anti-HCV antibody responses as well as there was a lot of variation in antibody titers against individual HCV proteins (Fig.
3). We found out that 97% of the sera studied recognized the core protein in very high levels, whereas the other proteins such as the NS4B, NS3, NS5A and E2 were found to be immunogenic in 85% to 31% of the cases, respectively (Fig
3A). A study carried out by Chen and coworkers among 60 chronic HCV patients, revealed E2 antibodies in 98%, core in 97%, NS3 in 88%, NS5 in 68% and NS4 in 48% of the cases [
27]. As analyzed by EIA the highest antibody levels were observed against the core protein (ca. 1:5000), while the antibody responses against other viral proteins or peptides derived of them remained at a lower level [
27]. As a whole the results of the above study are concordant with the observations of the present study, except that our Western blot analysis gave up to 10-fold higher titers against the core proteins and several fold higher levels of specific antibody responses against other HCV proteins. Also Nikolaeva and coworkers observed the core protein to be highly immunogenic (antibody titers up to 1:40 000) while other HCV proteins were less important immunogens in chronic HCV patients [
25]. Direct comparisons of the frequencies and antibody levels to individual HCV proteins in different studies is very difficult, since the methods to produce and purify viral antigens vary and also the form of the assay to detect anti-HCV antibodies varies from one study to another. In our analysis we decided to use the full-length baculovirus-expressed HCV proteins and Western blot analysis in order to be sure of the specificity of the antibody responses to a given protein. One of the drawbacks of the assay is, however, that only antibodies against linear antigenic epitopes within the denatured proteins are being detected in Western blotting.
The prevalence of anti-HCV antibodies have been followed during the chronic phase of infection [
25‐
27]. When we analysed sera from five HCV RNA and antibody positive patients during a period of 18 to 25 month, the antibody levels against the major immunogenic proteins were found to remain relatively constant. However, in three patients there were some changes in anti-HCV antibody levels, namely a weak decrease in the core and NS-specific antibody levels during the follow-up was seen. Similar analysis by others [
27,
29] revealed very similar results with highly persistent antibody patterns. While in most cases anti-HCV antibodies remain at a constant level, there were some individuals whose antibody levels showed some fluctuation [
27].
Conclusion
We were able to produce nine structural and non-structural HCV proteins in high levels in Sf 9 insect cells. These purified recombinant HCV proteins were found to be suitable for analyzing the prevalence of antibodies against individual HCV proteins in human sera obtained from patients suffering from chronic HCV infection. Clearly the core, NS3, NS4B and NS5A represented the major antigenic proteins. By Western blotting antibody responses against the viral glycoproteins, E1 and E2 and the NS4A protein were found less frequently. Curiously, the recombinant NS5B protein was recognized only by three patient sera all of which were from patients infected with HCV genotype 3. It was of interest that NS2 protein, a viral cysteine protease was unable to mount humoral immune responses in our patients. These recombinant HCV proteins will also enable the analysis of cell-mediated immune responses in HCV infection as well as to study whether changes in anti-HCV antibody patterns have a prognostic value in patients suffering from chronic HCV infection.
Acknowledgements
We thank K. Lamminaho, H. Valtonen, M. Yliselä, M. Aaltonen, T. Westerlund, S. Sopanen, R. Tyni, and V. Mäkinen for excellent technical assistance. This study was funded in part by the Medical Research Council of the Academy of Finland, the Finnish Cancer Foundation and the Päivikki and Sakari Sohlberg Foundation.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
MS carried out some of the experiments and drafted the manuscript. KM participated in the design of the study and analysed the results. PP, RF, KN and KM constructed the expression vectors and produced and purified the recombinant HCV proteins and used these proteins to screen the patient sera for HCV antibodies. ML genotyped the HCV positive patient sera and provided the specimens for the study as well as participated in the design of the study. IJ initiated the study, participated in its design and coordination and helped to draft the manuscript. All authors have read and approved the final version of the manuscript.