Viral mutation accelerated by nitric oxide ... - The FASEB Journal

Download PDF
6 downloads 0 Views 382KB Size Report
Comment
860-0811, Japan; *Department of Viral Infection, Institute of Medical Science, University of Tokyo, ... trite greatly accelerates the mutation of Sendai virus. (SeV) ...
Viral mutation accelerated by nitric oxide production during infection in vivo TAKAAKI AKAIKE,1 SHIGEMOTO FUJII, ATSUSHI KATO,* JUN YOSHITAKE, YOICHI MIYAMOTO, TOMOHIRO SAWA, SHINICHIRO OKAMOTO, MORITAKA SUGA, MAKOTO ASAKAWA,† YOSHIYUKI NAGAI,* AND HIROSHI MAEDA Departments of Microbiology and Medicine I, Kumamoto University School of Medicine, Kumamoto 860-0811, Japan; *Department of Viral Infection, Institute of Medical Science, University of Tokyo, Tokyo 108-0071, Japan; and †DNAVEC Research Institute, Tsukuba, Japan

Nitric oxide (NO), superoxide (O2ⴚ), and their reaction product peroxynitrite (ONOOⴚ) are generated in excess during a host’s response against viral infection, and contribute to viral pathogenesis by promoting oxidative stress and tissue injury. Here we demonstrate that NO and peroxynitrite greatly accelerates the mutation of Sendai virus (SeV), a nonsegmented negative-strand RNA virus, by using green fluorescent protein (GFP) inserted into and expressed by a recombinant SeV (GFP-SeV) as an indicator for mutation. GFP-SeV mutation frequencies were much higher in the wild-type mice than in those lacking inducible NO synthase, suggesting that mutation of the virus in vivo is NO dependent. High levels of NO and NO-mediated oxidative stress were induced by GFP-SeV infection in the lung of the wild-type mice, but not in the iNOS-deficient mice, as evidenced by electron spin resonance spectroscopy and immunohistochemical analysis for nitrotyrosine formation as well as histopathological examination. Furthermore, peroxynitrite, an NOderived reactive nitrogen intermediate, enhanced viral mutation in vitro. These results indicate that the oxidative stress induced by NO produced during the natural course of viral infection increases mutation, expands the quasispecies spectrum, and facilitates evolution of RNA viruses.—Akaike, T., Fujii, S., Kato, A., Yoshitake, J., Miyamoto, Y., Sawa, T., Okamoto, S., Suga, M., Asakawa, M., Nagai, Y., Maeda, H. Viral mutation accelerated by nitric oxide production during infection in vivo. FASEB J. 14, 1447–1454 (2000)

ABSTRACT

Key Words: NO 䡠 peroxynitrite 䡠 oxidative stress

Overproduction of nitric oxide (NO) and superoxide (O2⫺) appears to be a common phenomenon in viral infections (1–5). The resultant reactive molecular species such as peroxynitrite (ONOO⫺) nonselectively affect the host’s cells and tissues (6, 7). Such host defense effectors are originally produced to injure the intruding pathogens, which will then 0892-6638/00/0014-1447/$02.25 © FASEB

suffer oxidative stress because of the host. Oxidative stress has been suggested to be involved in mutagenesis and hence carcinogenesis (8 –11). Also, RNA virus mutation was reported to be increased by chemical mutagens, including nitrous acid (HNO2) (12–15), although the degree of mutation appears to be slight (16). Therefore, it may be that mutagenesis of various viruses is occurring naturally in biological systems during infections as a result of host defense. Although involvement of NO-induced oxidative stress in viral pathogenesis is now well recognized (4, 5), the role of NO and NO-derived reactive nitrogen intermediates in viral mutation is only poorly understood. RNA viruses share high mutation rates ranging from 10⫺5 to 10⫺3 misincorporation/nucleotide site/round of copying because of the error-prone nature of their replicases. This high mutation rate accounts for their quasispecies nature, as they exist as a mixture of heterogeneous populations (17). Numerous methods exist for estimating viral mutation, including measurement of mutation frequencies of phenotypic variations such as temperaturesensitive growth, plaque morphology, host range, and pathogenicity. The mutations cannot be assessed precisely and quantitatively by these criteria, however, because the variant viruses often contain multiple base substitutions in different genes (18). Escape of a virus from a particular neutralizing monoclonal antibody occurs by a single base substitution, leading to a single codon change on the epitope. Thus, investigation of viral genome mutation has been based on selection of the escape mutants. The frequency of generation of escape mutants in vitro in cultured cells is reported to be ⬃10⫺4.5 for four different negative-strand RNA viruses, i.e., Sendai virus (SeV), vesicular stomatitis virus, Newcastle disease virus, and influenza A virus (19, 20). Nevertheless, the selection procedures are 1 Correspondence: Department of Microbiology, Kumamoto University School of Medicine, Kumamoto 860-0811, Japan. E-mail: [email protected]

1447

not fully developed because the frequencies fluctuate greatly, even within a given virus species, depending on the antibodies used for selection (18). Also, any such mutation analysis is not truly applicable to the estimation of mutation rates occurring in vivo in infected host organisms. Performing a green fluorescent protein (GFP) -indicated viral mutation assay with the use of a recombinant SeV constructed with GFP genome as a tool for estimating mutation frequency occurring in vivo, in the present study we verified for the first time that oxidative stress induced by high-output NO accelerates RNA virus mutation in vivo.

MATERIALS AND METHODS Construction of GFP-SeV The open reading frame (ORF) of GFP (EGFP; 720 bp, humanized codon; Clontech Laboratories, Inc., Palo Alto, Calif.) was inserted into plasmid pSeV18⫹b(⫹), which contained a positive-sense SeV antigenome immediately upstream of the N ORF (Fig. 1A) (21, 22). The final construct, pSeV18⫹bGFP(⫹), produced recombinant GFPSeV antigenome (16.25 kb). GFP-SeV was harvested from pSeV18⫹bGFP(⫹)-transfected LLCMK2 cells to obtain fully infectious particles and then was propagated in embryonated chicken eggs. Assessment of GFP-indicated mutation Mutation frequency of GFP-SeV was determined by measuring incidence of fluorescence-negative viral plaques among fluorescence-positive plaques formed by GFP-SeV replication as identifying the plaques by hemadsorption test. Specifically, viral plaques were formed on monolayers of CV-1 cells (a monkey kidney cell line) cultured in 6-well plates (35 mm diameter; Falcon, Oxnard, Calif.) overlaid with 0.4% agarose containing 0.5 ␮g/ml trypsin (Sigma, St. Louis, Mo.) in Dulbecco’s minimum essential medium (DMEM) (Life Technologies, Inc., Grand Island, N.Y.) ⫹ 0.2% bovine serum albumin (BSA) (Sigma), and the mutation was identified and quantified by hemadsorption and fluorescence. A hemadsorption test was performed by incubating the monolayer cells with 2% guinea pig erythrocyte suspension in DMEM (37°C; 30 min) after removal of the agarose overlay of the cell culture (4 days). Hemadsorption foci with clear contour were observed only in the viral plaques, when nonadsorbed erythrocytes were removed by washing with 10 mM phosphatebuffered saline (PBS; pH 7.4) (Fig. 1B, C). Simultaneously, GFP expression was detected via its strong fluorescence under microscopic observation with an inverted phase-contrast fluorescence microscope (Fig. 1B, C). Hemadsorption-positive foci were also detected by use of a color reaction: erythrocyte’s hemoglobin-catalyzed peroxidation of diaminobenzidine with 3.5% tert-butylhydroperoxide as substrate (23); hemadsorption foci show brown colors. In vivo study of GFP-SeV mutation GFP-SeV was administered to C57BL/6 (B6) and B6 iNOS⫺/⫺ mice (4 wk old; male) by inhalation of viral suspension (2). B6 inducible nitric oxide synthase (iNOS) knockout mice were produced in Jackson Laboratories (West 1448

Vol. 14

July 2000

Figure 1. Construct of GFP-SeV genome (A) and identification of GFP-SeV mutants (B, C). B) Phase-contrast inverted microscopy of GFP-SeV-infected CV-1 cells; light microscopy (left) and fluorescence microscopy (right) of the same field. After 2 days incubation of infected cells without trypsin, hemadsorption and fluorescence were observed. Erythrocyteadhered cells showed strong fluorescence except for the cell indicated by the arrow. Magnification, ⫻112. C) Plaque formation of SeV strains on CV-1 cell monolayers (cultured with 0.5 ␮g/ml trypsin for 4 days). Left panels show erythrocyte-adhered plaques observed by light-inverted microscopy; middle panels show the same plaque examined by fluorescence microscopy. Magnification, ⫻8. Viral plaques were detected by a hemoglobin (Hb) stain (right panels; bar, 1 mm). Grove, Pa.; 24). At various times after viral infection, the lung tissues were obtained and homogenized in ice-cold PBS (pH 7.4) as described earlier (2). The supernatant of homogenates prepared by centrifugation (10, 000 g for 10 min) was serially diluted and inoculated onto CV-1 cells to form viral plaques for GFP-indicated mutation assay. The mutation frequency of GFP-SeV isolated from the lung was measured after formation of viral plaques on CV-1 cell monolayers, as just mentioned. For accurate determination of mutation frequencies, the number of fluorescence-negative plaques was counted, with the cell culture plate having less than 200 plaques/well (35 mm diameter) after 4 days of culture. NO production in the GFP-SeV-infected lung tissue was identified directly by using electron spin resonance (ESR) spectroscopy. N-Dithiocarboxy(sarcosine) (DTCS)–Fe2⫹ complex, a spin trap for NO, was administered subcutaneously and the lung was subjected to X-band ESR spectroscopy (Bruker ESP 380E) at 110K (3). Immunohistochemistry for iNOS induction and

The FASEB Journal

AKAIKE ET AL.

nitrotyrosine formation in the lung was performed as described previously (3). Briefly, after the lung was fixed in 2% periodate/lysine/paraformaldehyde, 6 ␮m sections were prepared, followed by immunostaining with either an antiiNOS antibody (1:500; Santa Cruz Biotechnology, Santa Cruz, Calif.) or a polyclonal anti-nitrotyrosine antibody (1:500; Upstate Biotechnology, Lake Placid, N.Y.) by the indirect immunoperoxidase method, using diaminobenzidine as substrate. The lung sections were also stained with hematoxylin and eosin (H&E) for histopathological examination. Treatment of GFP-SeV with peroxynitrite GFP-SeV was treated in vitro with a constant-flux system of peroxynitrite (25) prepared by infusing 250 ␮M peroxynitrite solution (in 10 mM NaOH) at 120 ␮l per 30 s into 1.2 ml of 0.5 M PBS (pH 7.4) containing 1 ⫻ 107 PFU/ml GFP-SeV at room temperature, so that the concentration of peroxynitrite, which decays within a few seconds at neutral pH, was maintained at 0.8 ␮M. An aliquot (120 ␮l) of the reaction mixture was withdrawn every 30 s after initiation of the peroxynitrite infusion and inoculated onto CV-1 cells to form viral plaques so that the mutation frequency of GFP-SeV was determined as described above. Peroxynitrite used in this study was synthesized from acidified nitrite and hydrogen peroxide by use of a quenched flow reactor, as reported previously (25, 26), and its purity was ⬎ 90% with ⬍ 10% nitrite and ⬍ 0.1% hydrogen peroxide contaminations. During constant infusion of peroxynitrite, pH of the reaction mixture gradually increased up to pH 7.8 at 5 min after initiating infusion. Analyses for GFP and SeV virion expression The expression of GFP and GFP-SeV virion was examined by Western blotting. Each virus strain was inoculated onto CV-1 cell monolayers at a multiplicity of infection (moi) of 5 PFU/cell. After 2 days of culture with 0.5 ␮g/ml trypsin, the total protein (15 ␮g) extracted from the cells with 0.5% sodium dodecyl sulfate (SDS) in 50 mM Tris-HCl buffer (pH 7.4) containing a protease inhibitor mixture (2) was subjected to SDS-polyacrylamide gel electrophoresis (PAGE) with or without 8 M urea under reducing conditions. After transfer to Immobilon polyvinylidene difluoride membrane (Millipore, Bedford, Mass.), the protein was immunostained with anti-GFP polyclonal antibody (Clontech) or anti-SeV antiserum as described previously (21). Also, mRNA expression of GFP and each virion was tested by reverse transcriptase-polymerase chain reaction (RT-PCR). Specifically, after total RNA was extracted from the virus-infected CV-1 cells by a guanidine thiocyanate lysis method with Trizol reagent (Gibco, BRL, Grand Island, N.Y.), 0.3 ␮g of the RNA was subjected to RT-PCR, which was performed according to our protocol reported previously (27). The oligonucleotide primers used were: sense 22 mer, 5⬘-TGAGCAAGGGCGAGGAGCTGTT-3⬘; antisense 23 mer, 5⬘-TACAGCTCGTCCATGCCGAGAGT-3⬘ to obtain a 712 bp GFP cDNA fragment, and sense 24 mer, 5⬘-AAGATAGCTGGATCCCACGAATCG-3⬘; antisense 30 mer, 5⬘-AGGCTTTGATGAGCGCTATGTCTCTTTTGG-3⬘ to obtain a 334 bp F protein cDNA fragment, including a coding region for the amino acid sequence of trypsin cleavage site. Similarly, mRNA for glyceraldehyde-3phosphate dehydrogenase was examined as an internal control for mRNA expression in the cells with use of the primers as described elsewhere (27). Detection of chymotrypsin-sensitive SeV mutants Wild-type SeV was generated by reverse genetics without GFP insertion in a same manner as GFP-SeV and treated with bolus VIRAL MUTATION ACCELERATED BY NITRIC OXIDE

injections of peroxynitrite. GFP-SeV (1⫻107 PFU/ml) was reacted with peroxynitrite at room temperature by bolus additions of 10 ␮l of peroxynitrite solution (0.1 M NaOH) to the 1.0 ml of viral suspension (final concentration of peroxynitrite: 8.0 ␮M) in 0.5 M PBS (pH 7.4). The addition was repeated three times at 1 min intervals; 8 min after the last peroxynitrite addition, an aliquot of the reaction mixture was inoculated onto CV-1 cells, followed by plaque formation. Plaque was formed in agarose-overlaid CV-1 cell monolayers containing 2.0 ␮g/ml chymotrypsin (N␣-tosyl-l-lysyl chloromethyl ketone treated; free of trypsin; Sigma) in DMEM ⫹ 0.2% BSA. Cloning of GFP-SeV mutants and sequence analysis for GFP and F protein genomes GFP-SeV mutants were cloned by picking up a single plaque formed on the CV-1 cell monolayer. For cloning of spontaneously occurred mutants, 5000 –10,000 plaques of GFP-SeV without peroxynitrite treatment were produced at the density of 50 –100 plaques/well (35 mm diameter; Falcon), and fluorescence-negative plaques identified microscopically by hemadsorption test were isolated carefully and suspended in 100 ␮l DMEM. The plaque of GFP-SeV mutants generated after peroxynitrite treatment was isolated from the plaque formation plates (50 –100 plaques/well) showing apparently high mutation frequencies (⬎2%), usually obtained later than 2.0 min after initiating the constant flux of peroxynitrite (cf. Fig. 4). The mutant plaques of GFP-SeV generated by peroxynitrite was picked up in five different experiments of peroxynitrite treatment. The viral suspension prepared from a single plaque was then inoculated onto CV-1 cells in 12-well plates (22 mm diameter; Falcon). After 48 h culture in DMEM ⫹ 0.2% BSA in the presence of 0.5 ␮g/ml trypsin, total RNA extracted from the virus-infected cells was subjected to RT-PCR to obtain a cDNA fragment of GFP as mentioned above. Similarly, F protein cDNA was generated by RT-PCR from the viral clone of chymotrypsin-sensitive SeV mutant. Subsequently, cDNA fragments sequences for GFP and SeV F protein were determined by a dideoxy method using ALF Express (Pharmacia, Piscataway, N.J.) with a Cy5 AutoRead Sequencing Kit (Pharmacia) to verify the genetic basis of GFP and F protein mutations.

RESULTS NO-dependent GFP-SeV mutation occurring in vivo Because base substitutions occurring in the GFP in SeV, either synonymous or nonsynonymous, are largely neutral (silent) for SeV replication, loss of GFP fluorescence can be readily detected at the single-cell and plaque level with positive viral protein (hemagglutinin-neuraminidase) expression on the cell surfaces by hemadsorption, hence allowing for accurate and easy quantification of increased mutation frequency (Fig. 1B, C). It is also important that the present GFP-indicated mutation assay can be done with GFP-SeVs grown in mice, the natural host, with different genetic backgrounds. The effect of NO-induced oxidative stress on GFPSeV mutation was investigated in an in vivo system. A lethal dose (2 LD50) of GFP-SeV was administered to 1449

B6 mice or B6 iNOS⫺/⫺ mice by inhalation of viral aerosol. The GFP-SeV yield in the lung on days 2, 4, 6, and 7 after infection was isolated and the GFP mutation was measured. The GFP-SeV mutation frequency in wild-type B6 mice increased significantly as infection proceeded, whereas that in iNOS knockout mice was elevated only slightly (Fig. 2A). A mutation frequency almost six- to sevenfold higher was observed on days 4, 6, and 7 in wild-type B6 mice than in iNOS-deficient mice, although the viral growth in the lung became plateau at these time points (Fig. 2B). No significant difference between the two groups was found for virus production throughout the course of infection (Fig. 2B), indicating that enhanced mutation frequencies of GFP-SeV is not necessarily dependent on the cycles of viral replication. An ESR study showed that overproduction of NO was evident in the GFP-SeV-infected lung of the wild-type mice, and its time profile was correlated with that of the mutation frequency (Fig. 2C). In contrast, no appreciable NO production was observed in GFP-SeV-infected iNOS knockout mice. When GFP-SeV was replicated in CV-1 cells in culture, the mutation frequency increased per a single-step replication was estimated to be ⬃0.1%. If we assume that viral mutation was occurring in the lung cells at a same rate as in the CV-1 cells, excess generation of NO in wild-type B6 mice seems to enhance viral mutation to a great extent, similar to that achieved after more than 10 cycles of viral multiplication in cells without NO exposure in that the difference in mutation frequency (1.5%; on days 6 and 7) between wild-type mice and iNOS knockout mice was almost 15-fold higher than the frequency increased per single viral replication. NO-induced oxidative stress in GFP-SeV-induced pneumonia We reported earlier that overproduction of both NO and superoxide, possibly through peroxynitrite formation, contributes to the pathogenesis of influenza virus-induced pneumonia in mice (1–3). The contribution of high-output NO to influenza pathogenesis is also verified by Karupiah et al. (28) using mice lacking iNOS. Therefore, a similar pathogenetic mechanism involving NO-induced oxidative stress may operate in SeV-induced pneumonia. The immunohistochemical analysis showed that iNOS was expressed by inflammatory phagocytic cells such as exudate macrophages infiltrating alveoli and interstitial tissues of the GFP-SeV-infected lung in wildtype B6 mice (Fig. 3A). Similar distribution of strong immunostaining for nitrotyrosine was observed in the virus-infected lung of the wild-type mice (Fig. 3B), but nitrotyrosine formation was not identified significantly in iNOS-deficient mice (Fig. 3C). Thus, 1450

Vol. 14

July 2000

Figure 2. NO-dependent GFP-SeV mutation in GFP-SeVinduced pneumonia in mice. A) The mutation frequency of the GFP-SeV isolated from the lung of wild-type B6 and iNOS⫺/⫺ B6 mice was quantified by use of the GFP-indicated SeV mutation assay. The increased value in mutation frequency after infection is shown. Data are means ⫾ se (n⫽4; ⬎500 plaques counted/assay); *P ⬍ 0.05, **P ⬍ 0.01, between wild-type and knockout mice (t test). B) Virus yield in the lung of wild-type B6 and iNOS⫺/⫺ B6 mice. C) Amount of NO generated in the lung determined by ESR spin trapping. The inset shows the ESR spectrum of the NO-dithiocarboxy(sarcosine) (DTCS)–Fe complex produced in the lung (6 days after infection). Data are means ⫾ se (n⫽3).

it is quite likely that high production of NO in vivo accelerates RNA virus mutation through formation of reactive nitrogen oxides such as peroxynitrite. It is important that the inflammatory tissue injury of the lung infected with GFP-SeV was significantly reduced in the iNOS-deficient mice compared with wild-type

The FASEB Journal

AKAIKE ET AL.

Figure 3. Immunohistochemistry for iNOS expression and nitrotyrosine formation in mouse lungs infected with GFPSeV (A-C), and histopathological changes of GFP-SeV-infected lungs of wild-type and iNOS-deficient mice (D, E). The lung sections of wild-type mice infected with GFP-SeV (7 days postinfection) were stained with anti-iNOS antibody (A) or anti-nitrotyrosine antibody (B). C) Immunostaining for nitrotyrosine of the iNOS-deficient mouse lung infected with GFP-SeV (7 days postinfection). D, E) H&E stain of the virus-infected lungs (8 days postinfection) obtained from wild-type and iNOS-deficient mice, respectively. Magnification, ⫻100 (A--C) and ⫻60 (D, E).

mice (Fig. 3D, E), suggesting involvement of NOinduced oxidative stress in the pathogenesis of SeV pneumonia. Elevation of mutation frequency of SeV by peroxynitrite To further clarify the NO-dependent enhanced mutation, GFP-SeV was treated with peroxynitrite in vitro and its GFP mutation was tested after propagation on CV-1 cells monolayers (plaque-forming assay). Specifically, peroxynitrite-induced mutation was evaluated with a constant-flux system of peroxynitrite (0.8 ␮M). The mutation frequency was enhanced at 1.5 min after peroxynitrite infusion: a 2.6-fold increase (0.68⫾0.07%) compared with the background value (0.26⫾0.02%), when no appreciable virucidal effect of peroxynitrite was observed (Fig. 4). A time-dependent increase in mutation frequency was obtained, although the virus infectivity was reduced inversely. A surprisingly high mutation frequency was observed after a 4 min infusion period (Fig. 4, inset): 100% of the GFP-SeV became mutated. In contrast, the mutation frequency was not affected by the constant flux of decomposed peroxynitrite, which was prepared by incubation of peroxynitrite in neutral solution, indicating that remarkable elevation of GFP mutation frequency was induced by peroxynitrite, but not by hydrogen peroxide and nitrite contaminated in the peroxynitrite preparation if any. VIRAL MUTATION ACCELERATED BY NITRIC OXIDE

Specific proteolytic cleavage of SeV F protein by a trypsin-like enzyme is a prerequisite for the virus to be infectious (29). Altered protease-activated mutants can be selected by chymotrypsin (30). This conventional selection method was also used here to assess the effect of peroxynitrite on SeV mutation. SeV treated with peroxynitrite or vehicle was subjected to a plaque formation assay; in the agarose overlay containing chymotrypsin, as much as a 10fold increase in chymotrypsin-requiring plaque formation was found after peroxynitrite treatment (Fig. 5), although viral infectivity was reduced moderately (by 42%). The sequence analysis of F protein RNA showed that each chymotrypsin mutant isolated had an amino acid change at the P1 cleavage site of the F protein (Arg 3 Ile). This result suggests that peroxynitrite enhances mutation not only of the artificially inserted GFP genome, but also of the natural SeV genome. In a separate experiment, we examined the GFPSeV mutation treated with an NO donor propylamine NONOate (CH3N[N(O)NO]-(CH2)3NH2⫹ CH3), which releases spontaneously pure NO in solution at neutral pH (T1/2, 7.6 min) (31), for 30 min at 37°C under ambient condition. As a result, NO per se did not show any significant mutagenetic potential, even when the concentration of propylamine NONOate was increased up to 1 mM. According to the data shown in Fig. 2C, the level of NO production in the virus-infected lung was estimated to be less than 20 ␮M at maximum. Thus, we concluded that NO per se, even if it is converted to the oxidized intermediates such as NO2 and N2O3

Figure 4. Acceleration of SeV mutation by peroxynitrite (ONOO⫺). GFP-SeV was treated in a constant-flux peroxynitrite (0.8 ␮M) system, and the mutation frequency was determined by GFP-indicated mutation assay after plaque formation of GFP-SeV treated with peroxynitrite for various time periods. Values are means ⫾ se of four experiments (⬎500 plaques counted/assay); *P ⬍ 0.05, **P ⬍ 0.01 compared with controls (t-test). 1451

the same as that of peroxynitrite-induced mutations, except that the incidence of A 3 G substitution was higher in the latter than in the former and U 3 C transition was higher in the former than in the latter. Almost all these point mutations led to amino acid substitutions. Also, there are apparently hot spots of point mutations in both mutants. For example, A 3 G transition frequently occurred at nucleotide positions of 496, 530, 547, 601, and 623 (nucleotide number in viral cDNA), and these substitutions were observed with 10/19 clones of spontaneous mutants and 14/25 clones of peroxynitrite-induced mutants. This bias in the site of point mutations may be due to selection of the GFP-SeV mutants based on the phenotypic change, i.e., loss of fluorescence emission of GFP. Figure 5. Generation of chymotrypsin-activating SeV mutants by peroxynitrite treatment. Occurrence of chymotrypsin-sensitive mutants from SeV treated or untreated with peroxynitrite (8 ␮M ⫻ 3) was determined by the plaque-forming assay with chymotrypsin. Data are obtained from the three different experiments of peroxynitrite treatment. For each group (control and peroxynitrite-treated group), 2.5 ⫻ 106 PFU SeV was used to measure the number of plaques showing chymotrypsin-sensitive growth. Plaques formed by parent SeV with trypsin and those by chymotrypsin mutants with chymotrypsin are shown in the inset (Hb stain; bar, 1 cm). *P ⬍ 0.05 (chi-square test).

under ambient condition, cannot induce biological relevant GFP-SeV mutation. Characterization of various GFP-SeV mutants The GFP-SeV mutants were cloned (19 clones for control; 25 clones for peroxynitrite treatment), and GFP and mRNA expression of each clone was examined. Western blot analysis indicated that all nonfluorescence GFP variants possessed the same immunoreactivity and mobility as did the native GFP (Fig. 6). Mutant GFPs, however, showed slightly slower mobility than did native GFP on gels without urea, suggesting alteration of their primary structures. In contrast, expression of each virion protein was not impaired for all GFP-SeV mutants tested. GFP mRNA synthesis of all mutants was as efficient as that of the parent GFP-SeV, as revealed by RT-PCR (Fig. 6).

DISCUSSION The present study describes the potential role of NO and related reactive nitrogen species in acceleration of RNA viral mutation as a result of the host response against viral propagation in vivo. The infectioninduced enhancement of the mutation frequency was verified by using GFP-indicated SeV mutation assay developed in the present experiment. Of particular importance is our finding that NO-dependent viral mutation was clearly demonstrated with use of GFP-SeV-induced pneumonia in the wild-type and iNOS-deficient mice. Although the mechanism of point mutations generated by NO-induced oxidative stress still remains unclear, the mutation frequency of GFP-SeV was remarkably elevated by peroxynitrite, which appears most likely to be involved in the pathogenesis of SeV infection in mice, similar to other models of viral infections such as the influenza

Sequence analysis of GFP-SeV mutants GFP genes in various GFP-SeV clones were sequenced. Neither deletion nor insertion was found for any clone, and all mutations were point mutations. A 3 G transition (in viral negative sense) predominated over other transitions and transversions among 25 GFP-SeV mutants obtained after peroxynitrite treatment (Table 1). The overall substitution pattern found in the 19 peroxynitrite-untreated spontaneous mutant clones was essentially 1452

Vol. 14

July 2000

Figure 6. GFP and viral protein expression of various GFP-SeV mutants. A) Western blot analysis for GFP and SeV proteins produced in virus-infected cells. A recombinant GFP (Clontech) was the control. Upper panel: SDS-PAGE; middle panel: 8 M urea SDS-PAGE. Lower panel: Western blot for SeV proteins. B) RT-PCR analysis for GFP mRNA expression. Total RNA (0.3 ␮g) from virus-infected cells (2 days after infection; moi: 5 PFU/cell) was used for RT-PCR. G3PDH, glyceraldehyde-3-phosphate dehydrogenase. Two representative mutant clones of GFP-SeV were used for the analysis of GFP and its mRNA expression.

The FASEB Journal

AKAIKE ET AL.

TABLE 1. Mutation of GFP genes of mutant GFP-SeV Frequency (%) of base substitution

Transition A3G G3A U3C C3U Subtotal Transversion U3G U3A A3U C3A Subtotal Total

Spontaneous (19 clones)

ONOO⫺ treatment (25 clones)

57 (67.9) 2 (2.4) 18 (21.4) 4 (4.8) 81 (96.4)

78 (83.0) 2 (2.1) 6 (6.4) 4 (4.3) 90 (95.7)

1 (1.2) 1 (1.2) 0 1 (1.2) 3 (3.6) 84 (100)

1 (1.1) 0 1 (1.1) 2 (2.1) 4 (4.3) 94 (100)

virus pneumonia in mice and herpes simplex virus encephalitis in rats (3, 27). A previous study showed that human leukocytes producing superoxide exert a mutagenic potential on Salmonella typhimurium TA100 in vitro (32). Recent work by Gal and Wogan (33) demonstrated mutagenicity associated with NO in vivo by using lacZ gene-containing plasmid transgenic mice. It is therefore of paramount importance to investigate a link between NO-induced oxidative stress and viral mutation. Beck et al. (34) show that the pathogenicity of coxsackievirus B3 is strongly potentiated in vivo in mice fed a selenium-deficient diet. Intriguingly, an avirulent strain of the virus is converted to a potent cardiotoxic variant during infection in seleniumdepleted animals. The deficiency of selenium may result in an ineffective antioxidant system, e.g., low levels of glutathione peroxidase. These studies were extended to vitamin E- and glutathione peroxidasedeficient animals, and the results suggest that oxidative stress facilitates selection and generation of virulent mutants due to impairment of immunological clearance of the virus (35). Because peroxynitrite can permeate a biological phospholipid bilayer membrane efficiently (36, 37), enhanced mutation of GFP-SeV may be explained by the action of peroxynitrite through nitration, oxidation, or other modifications of the viral genome. Peroxynitrite causes mutagenesis of prokaryotic DNA (38, 39). Also, increased mutagenesis of the hprt gene of activated murine macrophages expressing iNOS was recently documented; this NO-associated mutation spectrum, however, did not differ from that arising spontaneously (40). Similarly, in our current study, the mutation profile of GFP-SeV mutants induced by peroxynitrite resembled that of spontaneously occurring mutants. However, peroxynitrite treatment in vitro preferentially elicits A 3 G transition in the GFP genome. Our preliminary VIRAL MUTATION ACCELERATED BY NITRIC OXIDE

study suggests that the GFP-SeV mutants recovered from the wild-type mice had a mutation pattern similar to that of peroxynitrite-induced mutants: the base substitutions found in six different clones of GFP-SeV mutants were A 3 G, 78%; U 3 C, 11%; C 3 U, 11%. Therefore, the mutational spectrum of GFP-SeV mutants seems not to be consistent with that of prokaryotic and eukaryotic DNA treated with peroxynitrite, in which G 3 T transversion is often obtained via depurination of nitrated deoxyguanosine (10, 38, 39). In a separate experiment, we found that nitro G formed in viral RNA by peroxynitrite is chemically more stable than that in DNA (data not shown). Thus, the G-specific point mutation might not occur in peroxynitrite mutagenesis of viral RNA. Because the SeV genome is a highly integrated complex of a single-stranded RNA and RNA polymerase as well as nucleoproteins, it is possible that peroxynitrite affects the integrity of the viral RNA–protein complex, leading to augmentation of the error-prone nature of the RNA replicase. In any case, more detailed experiments (e.g., analysis using a cell-free viral RNA replication system) may be needed to clarify the molecular events of peroxynitrite-induced RNA mutagenesis. It is believed that because of the absence of proofreading and repair functions of viral polymerases, RNA viruses exist as a mixture of heterogeneous genomic populations, termed quasispecies, and evolve rapidly under selective pressures. The present study provides direct evidence that hostderived oxidative stress promotes point mutations. This result expands heterogeneity and increases the number and repertories of variants from which a particular genotype will then evolve more readily under selective pressure. Induction of NO and peroxynitrite generation is a common phenomenon in virus-infected hosts. Therefore, our results and the concepts described here probably have great relevance to viral evolution in general, including the rapid generation of drug-resistant and neutralization escape variants as well as cell tropism-altered variants of human immunodeficiency viruses in vivo. We thank Ms. Judith B. Gandy for editorial work and Ms. Rie Yoshimoto for preparing the manuscript. This work was supported by a grant-in-aid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan, and grants from the Ministry of Health and Welfare of Japan.

REFERENCES 1.

Akaike, T., Ando, M., Oda, T., Doi, T., Ijiri, S., Araki, S., and Maeda, H. (1990) Dependence on O2⫺ generation by xanthine oxidase of pathogenesis of influenza virus infection in mice. J. Clin. Invest. 85, 739 –745 2. Oda, T., Akaike, T., Hamamoto, T., Suzuki, F., Hirano, T., and Maeda, H. (1989) Oxygen radicals in influenza-induced patho-

1453

3.

4. 5. 6.

7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

17. 18. 19. 20.

21.

22.

23.

1454

genesis and treatment with pyran polymer-conjugated SOD. Science 244, 974 –976 Akaike, T., Noguchi, Y., Ijiri, S., Setoguchi, K., Suga, M., Zheng, Y. M., Dietzschold, B., and Maeda, H. (1996) Pathogenesis of influenza virus-induced pneumonia: involvement of both nitric oxide and oxygen radicals. Proc. Natl. Acad. Sci. USA 93, 2448 – 2453 Akaike, T., Suga, M., and Maeda, H. (1998) Free radicals in viral pathogenesis: molecular mechanisms involving superoxide and NO. Proc. Soc. Exp. Biol. Med. 217, 64 –73 Schwartz, K. B. (1996) Oxidative stress during viral infection: a review. Free Rad. Biol. Med. 21, 641– 649 Beckman, J. S., Beckman, T. W., Chen, J., Marshall, P. A., and Freeman, B. A. (1990) Apparent hydroxyl radical production by peroxynitrite: implications for endothelial injury from nitric oxide and superoxide. Proc. Natl. Acad. Sci. USA 87, 1620 –1624 Rubbo, H., Darley-Usmar, V., and Freeman, B. A. (1996) Nitric oxide regulation of tissue free radical injury. Chem. Res. Toxicol. 9, 809 – 820 Vuillaume, M. (1987) Reduced oxygen species, mutation, induction and cancer initiation. Mutat. Res. 186, 43–72 Harris, C. C. (1991) Chemical and physical carcinogenesis: advances and perspectives for the 1990s. Cancer Res. 51, 5023s– 5044s Ohshima, H., and Bartsch, H. (1994) Chronic infections and inflammatory processes as cancer risk factors: possible role of nitric oxide in carcinogenesis. Mutat. Res. 305, 253–264 Liu, R. H., and Hotchkiss, H. (1995) Potential genotoxicity of chronically elevated nitric oxide: a review. Mutat. Res. 339, 73– 89 Tsugita, A., and Fraenkel-Conrat, H. (1962) The composition of proteins of chemically evoked mutants of TMV RNA. J. Mol. Biol. 4, 73– 82 Singer, B., and Fraenkel-Conrat, H. (1969) Mutagenicity of alkyl and nitroso-alkyl compounds acting on tobacco mosaic virus and its RNA. Virology 39, 395–399 Carp, R. I., and Koprowski, H. (1962) Mutation of type 3 poliovirus with nitrous acid. Virology 17, 99 –109 Granoff, A. (1961) Induction of Newcastle disease virus mutants with nitrous acid. Virology 13, 402– 408 Holland, J. J., Domingo, E., de la Torre, J. C., and Steinhauer, D. A. (1990) Mutation frequencies of defined single codon sites in vesicular stomatitis virus and poliovirus can be increased only slightly by chemical mutagenesis. J. Virol. 64, 3960 –3962 Domingo, E., Mene´ndez-Arias, L., and Holland, J. J. (1997) RNA virus fitness. Rev. Med. Virol. 7, 87–96 Smith, D. B., and Inglis, S. C. (1987) The mutation rate and variability of eukaryotic viruses: an analytical review. J. Gen. Virol. 68, 2729 –2740 Portner, A., Webster, R. G., and Bean, W. J. (1980) Similar frequencies of antigenic variants in Sendai, vesicular stomatitis, and influenza A viruses. Virology 104, 235–238 Nishikawa, K., Isomura, S., Suzuki, S., Watanabe, E., Hamaguchi, M., Yoshida, T., and Nagai, Y. (1983) Monoclonal antibodies of the HN glucoprotein of Newcastle disease virus. Biological characterization and use for strain comparisons. Virology 130, 318 –330 Kato, A., Sakai, Y., Shinoda, T., Kondo, T., Nakanishi, M., and Nagai, Y. (1996) Initiation of Sendai virus multiplication from transfected cDNA or RNA with negative or positive sense. Genes Cells 1, 569 –579 Hassan, M. K., Kato, A., Shinoda, T., Sakai, Y., Yu, D., and Nagai, Y. (1997) Creation of an infectious recombinant Sendai virus expressing the firefly luciferase gene from the 3⬘ proximal first locus. J. Gen. Virol. 78, 2813–2820 Akaike, T., Sato, K., Ijiri, S., Miyamoto, Y., Kohno, M., Ando, M., and Maeda, H. (1992) Bactericidal activity of alkyl peroxyl

Vol. 14

July 2000

24.

25.

26.

27. 28.

29.

30. 31.

32. 33. 34.

35. 36. 37. 38. 39. 40.

radicals generated by heme-iron-catalyzed decomposition of organic peroxides. Arch. Biochem. Biophys. 294, 55– 63 Laubach, V. E., Shesely, E. G., Smithies, O., and Sherman, P. A. (1995) Mice lacking inducible nitric oxide synthase are not resistant to lipopolysaccharide-induced death. Proc. Natl. Acad. Sci. USA 92, 10688 –10692 Okamoto, T., Akaike, T., Nagano, T., Miyajima, S., Suga, M., Ando, M., Ichimori, K., and Maeda, H. (1997) Activation of human neutrophil procollagenase by nitrogen dioxide and peroxynitrite: a novel mechanism for procollagenase activation involving nitric oxide. Arch. Biochem. Biophys. 342, 261– 274 Radi, R., Beckman, J. S., Bush, K. M., and Freeman B. A. (1991) Peroxynitrite oxidation of sulfhydryls. The cytotoxic potential of superoxide and nitric oxide. J. Biol. Chem. 266, 4244 – 4250 Fujii, S., Akaike, T., and Maeda, H. (1999) Role of nitric oxide in pathogenesis of herpes simplex virus encephalitis in rats. Virology 256, 203–212 Karupiah, G., Chen, J. H., Mahalingam, S., Nathan, C. F., and MacMicking, J. D. (1998) Rapid interferon ␥-dependent clearance of influenza A virus and protection from consolidating pneumonitis in nitric oxide synthase 2-deficient mice. J. Exp. Med. 188, 1541–1546 Homma, M., and Ohuchi, M. (1973) Trypsin action on the growth of Sendai virus in tissue culture cells. III. Structural difference of Sendai viruses grown in eggs and tissue culture cells. J. Virol. 12, 1457–1465 Scheid, A., and Choppin, P. W. (1976) Protease activation mutants of Sendai virus. Activation of biological properties by specific proteases. Virology 69, 265–277 Inoue, K., Akaike, T., Miyamoto, Y., Okamoto, T., Sawa, T., Otagiri, M., Suzuki, S., Yoshimura, T., and Maeda, H. (1999) Nitrosothiol formation catalyzed by ceruloplasmin: implication for cytoprotective mechanism in vivo. J. Biol. Chem. 274, 27069 – 27075 Weitzman, S. A., and Stossel, T. P. (1981) Mutation caused by human phagocytes. Science 212, 546 –547 Gal, A., and Wogan, G. N. (1996) Mutagenesis associated with nitric oxide production in transgenic SJL mice. Proc. Natl. Acad. Sci. USA 93, 15102–15107 Beck, M. A., Shi, Q., Morris, V. G., and Levander, O. A. (1995) Rapid genomic evolution of a non-virulent coxsackievirus B3 in selenium-deficient mice results in selection of identical virulent isolates. Nat. Med. 1, 433– 436 Beck, M. A., Esworthy, R. S., Ho, Y.-S., and Chu, F.-F. (1998) Glutathione peroxidase protects mice from viral-induced myocarditis. FASEB J. 12, 1143–1149 Marla, S. S., Lee, J., and Groves J. T. (1997) Peroxynitrite rapidly permeates phospholipid membranes. Proc. Natl. Acad. Sci. USA 94, 14243–14248 Denicola, A., Souza, J. M., and Radi, R. (1998) Diffusion of peroxynitrite across erythrocyte membranes. Proc. Natl. Acad. Sci. USA 95, 3566 –3571 Szabo´, C., and Ohshima, H. (1997) DNA damage induced by peroxynitrite: subsequent biological effects. Nitric Oxide 1, 373– 385 Juedes, M. J., and Wogan, G. N. (1996) Peroxynitrite-induced mutation spectra of pSP189 following replication in bacteria and in human cells. Mutat. Res. 349, 51– 61 Zhuang, J. C., Lin, C., Lin, D., and Wogan, G. N. (1998) Mutagenesis associated with nitric oxide production in macrophages. Proc. Natl. Acad. Sci. USA 95, 8286 – 8291

The FASEB Journal

Received for publication August 30, 1999. Revised for publication November 18, 1999.

AKAIKE ET AL.