Potential conflict of interest: Nothing to report.
Hepatitis C virus (HCV) is an important cause of chronic liver disease and is complicated by hepatocellular carcinoma (HCC). Mechanisms whereby the virus promotes cellular transformation are poorly understood. We hypothesized that the guanosine triphosphatase activity encoded in the HCV NS4B protein's nucleotide binding motif (NBM) might play a role in the transformation process. Here we report that NS4B can transform NIH-3T3 cells, leading to tumor formation in vivo. This transformation was independent of co-transfection with activated Ha-ras. Detailed analyses of NS4B mutants revealed that this transforming activity could be progressively inhibited and completely abrogated by increasing genetic impairment of the NS4B nucleotide binding motif. Conclusion: NS4B has in vitro and in vivo tumorigenic potential, and the NS4B transforming activity is indeed mediated by its NBM. Moreover, our results suggest that pharmacological inhibition of the latter might inhibit not only HCV replication but also the associated HCC. (HEPATOLOGY 2008.)
Chronic infection with the hepatitis C virus (HCV) is a major risk factor for the development of hepatocellular carcinoma (HCC). The incidence of HCC and the mortality rate associated with it are increasing dramatically.1, 2 Although chronic inflammation, fibrosis, and liver cell proliferation are considered the major factors contributing to the development of HCC, accumulating data support direct viral effects as well.3
HCV is a positive single-stranded RNA virus. Its 9.6-kb genome encodes a single ∼3000-amino-acid polyprotein, which is proteolytically processed into structural proteins, which are components of the mature virus, and nonstructural proteins, which are involved in replicating the viral genome.4 Several viral proteins including NS3, NS5A, and core have been implicated in cellular transformation.5–8 NS4B, a 27-kDa membrane protein, has been similarly shown by Park et al9 to transform NIH 3T3 cells.9 Transformation, however, only occurred when NS4B was co-transfected with the activated Ha-ras gene. The mechanism by which NS4B mediates its transformation potential remains unknown.
We have recently reported the identification of a nucleotide-binding motif (NBM) within NS4B and shown that this motif mediates both binding and hydrolysis of guanosine triphosphate (GTP) and HCV RNA replication.10 The NBMs of human oncogenes, such as ras, mediate their malignant transformation.11, 12 We thus hypothesized that the NBM of NS4B may similarly mediate its role in transformation. Here we report that NS4B of the Con 1 isolate transforms NIH3T3 cells independently of exogenous Ha-ras, the transformed cells lead to tumor formation in nude mice, and the NBM of NS4B mediates the latter's role in transformation, with exciting implications for novel therapies.
Standard recombinant DNA technology was used to construct and purify all plasmids. All regions that were amplified by polymerase chain reaction were analyzed by automated DNA sequencing. Plasmid DNAs were prepared from large-scale bacterial cultures and purified by a Maxiprep kit (Marligen Biosciences, Ijamsville, MD). Restriction enzymes were purchased from New England Biolabs (Ipswich, MA).
The plasmid pcDNA3.1-NS4B, which encodes the Con1 NS4B sequence, was generated by amplification of the NS4B gene from the Bart79I plasmid13 with primers containing BamH1 and EcoR1 restriction sites (primers 1 and 2; Table 1), digestion with BamH1 and EcoR1 and ligation into the corresponding site in pcDNA3.1 (Invitrogen, Carlsbad, CA). The mutations I131N, D228L, and F211A were introduced into this plasmid and into pcDNA-NS4B-GFP by site-directed mutagenesis [using primers 3-8; Table 1, and the QuikChange kit (Stratagene, La Jolla, CA)].
Table 1. Sequences of the Oligonucleotides Used in This Study
To create an NS4B homologous to the K isolate, we introduced 7 amino acid changes by using site-directed mutagenesis [using primers 9-18, Table 1, and the QuikChange kit (Strategene)] into the Con1 NS4B sequence within the pcDNA3.1-NS4B plasmid.
The plasmid pEJ6.614 encoding the Ha-ras gene was provided by Dr. R. Weinberg (Whitehead Institute, Cambridge, MA).
pUC19 was obtained from Promega Corporation (Madison, WI).
NIH 3T3 cells were a gift from Dr. J. M. White. A second clone was obtained from the ATCC (Manassas, VA). Cells were propagated in Dulbecco's modified Eagle's medium (Gibco/Invitrogen, Carlsbad, CA), supplemented with 10% calf serum (Colorado Serum Co., Denver, CO), 1% penicillin, 1% streptomycin, 1% L-glutamine (Gibco/Invitrogen, Carlsbad, CA), and maintained at 37°C in a humidified 10% CO2 incubator. Cells were discarded after 3 passages.
A mouse monoclonal antibody against NS4B was purchased from Virostat (Portland, ME). A rabbit polyclonal antibody against ras was purchased from Santa Cruz Biotechnology Inc (Santa Cruz, CA).
A standard transfection assay was performed essentially as previously described.9, 12, 15–17 Single stock NIH 3T3 cells were plated in 6-well plates (2.5 × 105 cells per well), grown for 24 hours, and then transfected with 0.5 μg of the pcDNA3.1-NS4B plasmid (wild-type or mutant versions), 0.5 μg pEJ6.6, or both, using Lipofectamine 2000 (Invitrogen, Carlsbad, CA). Empty pcDNA3.1 plasmid was co-transfected with pEJ6.6 samples as a source for neomycin resistance. The total amount of transfected DNA was kept constant at 2 μg per well by the addition of the empty vector, pUC19, as a carrier. Twenty-four hours after transfection, cells were split at a ratio of 1:20, transferred into 10-cm plates, and grown under G418 selection (400 μg/mL) (Invitrogen, Carlsbad, CA) for 2 weeks. G418-resistant colonies were stained with crystal violet, and colonies larger than 2 mm were counted using ImageJ analysis (NIH) of scanned plates. The morphology of G418-resistant colonies was analyzed microscopically. Experiments were repeated 2 to 4 times, each time with duplicates.
Transfection efficiency was determined using a secreted alkaline phosphatase (SEAP) reporter gene assay (Roche Applied Science, Indianapolis, IN). Relative SEAP activity was measured in the transfected media supernatants using a Berthold LB 96 V luminometer.
Establishment of Stable Clones.
At least 3 representative colonies were isolated from plates transfected with NS4B, Ha-ras, or both, and passaged 3 times in the presence of G418. These clones were used to study the transformed phenotype of the cells as well as for determining protein expression by western blot analysis.
Western Blot Analysis.
Cell lysates prepared from stable clones expressing wild-type NS4B, NS4B NBM mutants, Ha-ras, or both, were subjected to western blot analyses, essentially as described,18 using primary antibodies to either NS4B (Virostat, Portland, ME) or Ras (Santa Cruz Biotechnology Inc, Santa Cruz, CA) and actin (Sigma), followed by a corresponding pair of Alexa Fluor 680-coupled (Molecular Probes) and IRDye 800CW-coupled (Rockland Inc.) secondary antibodies. Proteins were visualized with a LI-COR infrared imager (Odyssey), and the bands were quantified with Odyssey version 1.2 infrared imaging software.
Growth Rate, Saturation Density, and Serum Dependence.
Standard methods were used essentially as described.15 Briefly, 105 cells of NS4B, ras, or empty pcDNA3.1 stable clones were seeded in a 10-cm dish in growth medium and incubated at 37°C. Cells were rinsed with phosphate-buffered saline, trypsinized, and counted daily for 7 days. Growth rates were determined from the slope of the logarithmic curve during exponential growth, and saturation densities were measured. Serum dependence was determined by performing this experiment in the presence of serum depletion (1% fetal bovine serum).
Soft Agar Assays.
These were performed essentially as described.15 In brief, cells (5000 or 20,000 of NS4B, ras, and empty pcDNA3.1 stable transfectants) were suspended in 0.25% agar mixture and overlayed onto 0.75% agar in 10-cm dishes. Cells were fed with growth medium weekly. After a 2-week incubation, colony morphology was assessed microscopically, and the number of colonies was counted after staining with crystal violet.
Tumorigenicity in Nude Mice.
NS4B stable transfectants (106 cells) were resuspended in 0.1 mL phosphate-buffered saline and injected subcutaneously into the flanks of 4-week-old to 6-week-old nude mice (Balb/c nude male, Taconic, Hudson, NY). Five clonal lines of NS4B cells were injected, each into 5 different mice. Three stable clones of ras transfectants were used as a positive control (each injected into 3-5 mice). Non-transfected NIH3T3 cells and neomycin vector alone stably transfected NIH3T3 cells were used as negative controls. Tumor size was measured biweekly with linear calipers, and cross-sectional area was calculated using the formula 3.14 × 0.5 × length × 0.5 × width—shown to correlate best with tumor mass when tumor size is <6 grams.19 Tumors were defined as a growth with a section area larger than 1.5 cm2. Mice were sacrificed before tumors reached a cross-sectional area of 2 cm2. All animals received humane care according to the criteria outlined in the Guide for the Care and Use of Laboratory Animals (NIH).
GTP Binding Assays.
These were performed essentially as described.10 Briefly, Huh-7 cells were transfected in duplicate with NS4B-GFP constructs harboring combinations of mutations in the A, B, and G elements of the NBM. Membrane fractions were prepared and aliquots incubated with 32P-labeled GTP—4-azidoanilide (an ultraviolet-photoactivatable, non-hydrolyzable GTP analog) (Affinity Labeling Technologies Inc.) followed by a brief pulse of ultraviolet irradiation, immunoprecipitation with a rabbit anti–green fluorescent protein antibody (Molecular Probes), sodium dodecyl sulfate polyacrylamide gel electrophoresis, transfer to nitrocellulose, western blotting with a mouse anti-green fluorescent protein primary antibody (Roche) and goat anti-mouse Alexa Fluor 680-coupled secondary antibody (Molecular Probes). Bound GTP was quantified by phosphorimager (Molecular Dynamics), and NS4B protein by LI-COR infrared scanner (Odyssey).
NS4B Transforms NIH 3T3 Cells Independently of Co-transfected Exogenous Ha-ras.
The K-isolate of genotype 1b NS4B was shown to transform NIH3T3 cells when co-expressed with the activated Ha-ras gene.9 To test the hypothesis that the Con1 isolate of genotype 1b NS4B13, 20 has a similar transformation potential, we constructed a plasmid termed pcDNA3.1-NS4B, which encodes the Con1 NS4B sequence. The plasmid pEJ6.614 encoding the Ha-ras gene was provided by Dr. R. Weinberg (Whitehead Institute, Cambridge, MA). These plasmids were then used in a standard transfection assay, essentially as described.9, 12, 15–17 As shown in Fig. 1A and B, a large number of colonies were detected in cells transfected with NS4B, Ha-ras, or both. Large colonies were present very rarely (<1 per 2 plates) in cells transfected with pcDNA3.1 vector encoding neomycin resistance only. When present, these were used for background subtraction. As expected, no colonies were seen in cells transfected with just the carrier vector pUC19. Transfection efficiency was determined using an SEAP reporter gene assay (Roche Applied Science, Indianapolis, IN). The transfection efficiency of the various constructs was quite comparable.
To exclude the possibility that the observed findings were attributable to the use of a specific NIH3T3 cell clone, the experiment was repeated using another NIH3T3 cell clone (ATCC, Manassas, VA). Although subtle differences in background did occur, the results described were not significantly different between the 2 tested clones.
Microscopic analysis of the morphology of the G418-resistant colonies (Fig. 1C) showed a multi-layered dense growth characteristic of a loss of density-dependent growth inhibition phenotype in the NS4B transfectants, similar to that observed in the Ha-ras transfectants. In contrast, pcDNA3.1 transfectants formed monolayers under the same conditions. This suggests that the large colonies detected in the NS4B transfectants are likely transformed. Thus, as shown by Park et al.,9 NS4B indeed transformed NIH3T3 cells when co-expressed with the Ha-ras gene. Surprisingly, however, NS4B was able to robustly transform NIH3T3 cells even in the absence of Ha-ras. Furthermore, co-transfection with Ha-ras (pEJ6.6) did not seem to increase the number of transformed foci induced by NS4B alone.
Expression of Transfected NS4B and Ha-ras in Stable Cell Lines.
To determine whether the transfected NS4B or Ha-ras were successfully expressed in transfected cells, we isolated at least 3 representative colonies from plates transfected with NS4B, Ha-ras, or both, and passaged them 3 times in the presence of G418. Cell lysates prepared from these stable clones were then subjected to western blot analyses using mouse anti-NS4B (Virostat, Portland, ME) and rabbit anti-ras (Santa Cruz Biotechnology Inc, Santa Cruz, CA) antibodies. The appropriate exogenously transfected gene(s) were present (Fig. 2A). The level of NS4B expression was comparable in the various clones tested (Fig. 2B). Expression of ras was only detected in cells transfected with activated Ha-ras (Fig. 2A). Although the anti-ras antibody was raised against a domain common to multiple ras isoforms, and in theory should detect endogenous ras proteins as well, only cells transfected with exogenous Ha-ras exhibited detectable level of ras protein. Presumably, this reflects the higher expression level of the transfected isoform, as has been observed by others.21 Moreover, the expression level of NS4B and ras in NS4B-ras double-transfected clones was not lower than their expression level in the NS4B or ras monotransfected clones (Fig. 2A). This suggests that the absence of synergy observed between ras and NS4B is not a result of a lower expression level of NS4B or ras in the double transfectants.
Transformed Phenotype of NS4B Clones In Vitro.
Confirmation that NS4B-transfected clones are indeed transformed was obtained using several phenotypic in vitro assays. Standard methods were used to determine the growth rate, saturation density, and serum dependence of the stable clones.15 As shown in Fig. 3A, the doubling time of NS4B clones was 15 ± 2 hours. This was comparable to the doubling time of ras clones (13.7 ± 2.5 hours) and significantly shorter than that of clones established by transfection of empty pcDNA3.1 (32 ± 2.5 hours). Moreover, the saturation density of NS4B clones was approximately 14-fold higher than that of clones transfected with the empty vector (10.2 × 106 cells versus 0.7 × 106) and comparable to that of ras-transformed clones (4.4 × 106). Furthermore, NS4B and ras clones—but not empty pcDNA3.1 transfectants—continued to readily proliferate in the face of serum depletion (1% fetal bovine serum) and demonstrated loss of density-dependent growth inhibition when grown to confluence (Fig. 3A).
To determine the anchorage-independent growth potential of the G418-resistant clones, soft agar assays were performed essentially as described.15 NS4B-transformed cells were able to form a large number of foci in soft agar (Fig. 3). When 5000 and 20,000 cells were plated, 153 ± 21 and 467 ± 43 foci, respectively, formed (Fig. 3A). Similar foci were formed with ras transfectants; however, no foci were observed with stable cell lines established from pcDNA3.1-transfected cells (Fig. 3B). Although growth in soft agar is not absolutely correlated with tumorigenic potential, this assay is the best in vitro correlate to in vivo growth potential.15
These data suggest that NS4B from the 1b con-1 isolate can transform NIH 3T3 cells and that these transformed cells demonstrate a classical transformed phenotype.
NS4B Clones Are Tumorigenic in Nude Mice.
To examine whether NS4B transfectants are able to form tumors in vivo, clonal lines of NS4B cells were injected into the flanks of nude mice, tumor size was measured biweekly, and cross-sectional area was calculated. Tumors were defined as a growth with a section area larger than 1.5 cm2. Stable clones of ras transfectants were used as a positive control, and non-transfected NIH3T3 cells and neomycin vector alone stably transfected NIH3T3 cells were used as negative controls. As shown in Table 2, 2 of the NS4B-transfected clones formed tumors with no latency period, appearing within 2 weeks after inoculation. At 2 weeks, 1 of the positive control Ha-ras–transfected clones similarly formed tumors, whereas none of the negative control Neo transfectants or non-transfected cells induced tumors. At 5 weeks after inoculation, all but 1 of the NS4B-transfected clones yielded tumors of considerable size around the inoculation site. Similarly, 2 of 3 Ras-transfected clones yielded tumors in the same time frame. In contrast, no mice inoculated with Neo alone transfected cells, and only one (presumably attributable to spontaneous transformation) of 13 mice inoculated with non-transfected NIH3T3 cells, had developed a tumor after 5 weeks. To the best of our knowledge this is the first report that NS4B has in vivo tumorigenic potential.
Table 2. NS4B-Transfected NIH 3T3 Cells Are Tumorigenic in Nude Mice
NOTE. NS4B, Ha-ras, and neomycin stable transfectants as well as non-tranfected NIH3T3 cells, were injected subcutaneously into the flanks of Balb/c nude mice. Tumors were measured bi-weekly, and cross-sectional area was calculated. A tumor was defined as a growth with a cross-sectional area larger than 150 mm2. Data at 2 weeks and 5 weeks after inoculation are documented. Data are the number of mice with tumors/total number of mice inoculated.
NS4B Transformation Potential Is Influenced by Genotype Subtype.
In contrast to Park et al.,9 we did not observe a requirement for co-transfection with Ha-ras to achieve NS4B-mediated transformation. Several possibilities exist to explain the apparent difference in the requirement for co-transfected Ha-ras. Subtle variations in subtype sequence within a given genotype have been shown to affect the transformation potential of HCV proteins.5, 7, 22 Although both of genotype 1b, the NS4B sequence of the con-1 isolate used by us differs in 7 amino acids from the K isolate used by Park et al.9, 23 To test the hypothesis that these sequence variations account for the different transforming phenotypes of the 2 HCV clones, we introduced these 7 amino acid changes by using site-directed mutagenesis into the Con1 isolate to create an NS4B homologous to the K isolate (Fig. 4A). The transformation potential of this construct was then tested by the focus formation assay in comparison with the Con1 isolate. Similarly to Park et al., there was an increase in the number of foci induced by the K isolate with Ha-ras co-transfection. This exceeded the number of foci induced by Ha-ras alone. The number of transformed foci induced by the K isolate, however, was significantly (approximately 5-fold) lower than the number induced by the Con1 NS4B (P = 0.0087 in Student t test) (Fig. 4B). This was not a result of differences in transfection efficiency (as monitored by SEAP reporter gene assay), cellular distribution, or level of expression (data not shown). Again, although co-transfection of activated Ha-ras gene did not change the number of foci induced by Con1 NS4B, it did increase the number of transformed foci generated by the K isolate by 4-fold, similar to the report by Park et al.9 Together, these results suggest that the described sequence variations between genotype 1b subtypes account for the difference in activated Ha-ras co-transfection requirement for efficient transformation.
The NS4B NBM Mediates Cellular Transformation.
The NBM of the human oncogene ras is known to mediate ras' role in transformation.11, 12 We therefore hypothesized that the NBM of NS4B may similarly mediate NS4B's role in transformation. To test this hypothesis, we first introduced the I131N mutation (IN) in the A-motif of the NS4B NBM—a mutation previously shown to significantly impair guanosine triphosphatase activity10—into the pcDNA3.1-NS4B vector by site-directed mutagenesis (Fig. 5A). The transformation potential of this mutant was then analyzed by standard transfection assay, as described previously. We detected a 2-fold reduction in the number of colonies formed in comparison with the wild-type (WT) NS4B construct (Fig. 5C, D). Although this reduction was found to be statistically significant (P < 0.001 by Student t test analysis), the IN mutation failed to completely inhibit transformation mediated by NS4B.
It is known from the solved crystal structures of several other GTP-binding G-proteins that additional elements of the NBM besides the A-motif participate in nucleotide binding. In particular, the B and G motifs are positioned together in space so as to mediate binding with the phosphates and guanine base, respectively, and contribute to guanosine triphosphatase function.24, 25 The importance of the corresponding elements in the NS4B NBM (Fig. 5A) was revealed by the dramatic impairment in HCV replication that results from point mutations in the NS4B B-(D228L) and G-(F211A) motifs (to be published elsewhere). Furthermore, to directly test the hypothesis that these B-elements and G-elements of the NS4B NBM (Fig. 5A) are involved in mediating nucleotide binding by NS4B, combinations of these mutations were tested for their GTP binding activity, as described.10 GTP binding was progressively inhibited and completely abrogated by increasing genetic impairment of the NS4B nucleotide binding motif (Fig. 5B). Similarly, these combinations of mutations had quite a dramatic effect on NS4B's transforming activity (Fig. 5C, D). The double mutant containing the IN and DL mutations decreased the number of transformed colonies by approximately 10-fold compared with WT NS4B. Essentially no transformed colonies were formed above the neomycin-transfected background when cells were transfected with the triple mutant harboring point mutations in the A-(IN), B-(DL), and G-(FA) motifs. These findings were neither a result of differences in expression levels (Fig. 5E) nor a result of differences in transfection efficiency (monitored by SEAP reporter gene assay) or an obvious gross effect of the mutations on intracellular distribution patterns (data not shown). Thus, impairment of NS4B transforming activity is proportional to the degree of disruption of the NBM, indicating the key role played by the NBM in mediating cellular transformation.
Together the above results demonstrate that NS4B can robustly transform NIH3T3 cells in the absence of exogenous Ha-ras. Moreover, the transforming activity of NS4B is mediated by its NBM. Thus, although our data confirm a transforming activity of NS4B described by others,9 our results extend their findings and differ in a significant way. In contrast to Park et al.,9 we did not observe a requirement for co-transfection with Ha-ras to achieve Con1 NS4B-mediated transformation. Our findings suggest that this difference in requirement for activated HA-ras co-transfection can be accounted for by the sequence variations between the Con 1 and K genotype 1b isolates.
Interestingly, none of the 7 amino acid mismatches between the K and the Con1 isolate are within the NBM. It is possible that these amino acid mismatches affect NS4B in a way that affects the 3-dimensional conformation of the NBM. Alternatively, we cannot exclude the possibility that although the NBM is necessary for mediating cellular transformation, other elements of NS4B also may play a role.
Several reports on other HCV proteins describe a potential for transformation.5–8 Similar to these, our data are subject to the same limitations resulting from overexpression of a single protein. It is difficult to know whether protein levels required to achieve transformation are met in individual cells within the context of a natural infection. Moreover, HCV-associated HCC typically occurs only after years of chronic viral infection, implying that viral factors alone are likely not sufficient for cellular transformation. Nevertheless, to the extent that viral factors are involved, our results suggest that NS4B may play an important role in the transformation process. In addition, in contrast to former reports on other HCV proteins describing a potential for transformation,5–8 our data on the NS4B NBM mutants are the first to shed light on the possible mechanism whereby an activity within the viral protein can mediate the observed transformation. Our findings demonstrate that mutating the domains that define the NBM impairs NS4B's role in cellular transformation.
HCC is a significant complication of HCV infection.26 HCV is the leading cause of HCC in the United States, and this is becoming an increasingly important problem as the cohort of HCV-infected individuals ages, with increasing length of infection and risk of developing HCC. Our findings that the NBM of NS4B not only mediates NS4B's role in viral replication10 but also mediates cellular transformation raise an exciting potential therapeutic implication. To the extent viral factors contribute to the increased risk of HCC, we hypothesize that inhibition of NS4B's NBM activity may not only suppress viral replication but also may decrease the rate of developing the associated HCC.
The authors thank Dr. Robert A. Weinberg for providing us with the pEJ6.6 vector.