• drug resistant mutant;
  • free radical;
  • peroxynitrite;
  • 23S rRNA


  1. Top of page

In the present study it has been shown that the reactive nitrogen species, peroxynitrite, can cause at least a 7.1-fold increase in the frequency of occurrence of drug-resistant mutants of Helicobacter pylori at a pathophysiological concentration (e.g. 1.0 μM) and in the presence of CLR. Furthermore, the CLR MIC of these resistant H. pylori strains increased by at least 250 times or higher in CLR susceptibility. In the 45 resistant strains, the modification of 23S rRNA A2142G was the predominant mutation (22/45), followed by A2143G (17/45) within the sequences of 23S rRNA. The other mutants were one each (1/45) in A2142T, and T2269G, and two each (2/45) in C2695G and T1944C, respectively. These results show that the inflammatory host reaction involving induction of reactive oxygen species (e.g. O·−2), and the inducible form of nitric oxide synthase, is a significant cause of mutation via peroxynitrite formation, particularly in drug-resistant bacterial strains.

List of Abbreviations: 

colony forming unit



E. coli

Escherichia coli

H. pylori

Helicobacter pylori


minimum inhibitory concentration




superoxide anion radical


nitric oxide



Generation of drug-resistant mutant pathogenic bacteria is a serious problem. In previous studies (1, 2), we have demonstrated greatly increased mutation frequency in a mouse model of influenza viral infection. During one 10–14 day disease-span of a mouse, a 6- to 7- fold increase in the occurrence of mutant virus compared with the non-diseased state was noted. This mutation is caused primarily by ONOO (1–4), a product of an endogenous reaction between O·2 and NO, which are generated by inflammatory leukocytes at sites of infection (1–7). ONOO is a highly oxidative chemical which functions in the rapid oxidation or cleavage of nucleic acids, proteins and lipids (8). It is also, however, a strong nitrating agent, and thus 3-nitrotyrosine (9) and concomitant formation of 8-oxoguanosine and more recently 8-nitroguanosine (5, 6) are now known to occur. Furthermore, we have shown that, given an adequate supply of oxygen and NADPH, 8-nitroguanosine will become a catalyst for the generation of O·2 in the presence of cytochrome P450 NADPH reductase, nitric oxide synthases, or cytochrome b5 reductase in a non-stoichiometric manner (10, 11), that is, similar to a propagation reaction of lipid peroxidation.

This result means that an inflammatory site is a “hot spot” of free radical generation. Also, we have published a more recent report showing that ONOO in pathophysiological concentration (1–8 μM) greatly increases the mutation frequency in Salmonella typhimurium, as revealed by a modified Ames test in which no cytosolic cytochrome P450 is required for activation (3). This ONOO-induced increase in mutation frequency is effectively suppressed by antioxidative agents or antidotes such as ebselen and uric acid, and, most effectively, by a phenolic natural product canolol (3).

In Japan, a large-scale study of the efficacy of CLR in eradicating H. pylori is currently under way. In the present study, we examined whether generation of drug-resistant mutants of H. pylori is facilitated by 1 μM ONOO in the presence of CLR, a macrolide antibiotic commonly used for eradication of H. pylori. Here, we report that the frequency of CLR resistant mutant H.pylori strains is indeed significantly increased, and that highly elevated MIC are found in the CLR- resistant mutants. We also determined the sites of DNA mutation based on 23SrRNA having 1402 bp, which represents a small fraction of the total chromosomal DNA of H. pylori (that consists of about 1600 Kbp).


  1. Top of page

Bacteria and culture conditions

H. pylori ATCC 43504 was obtained from the American Type Culture Collection (Manassas, VA, USA) and two clinical isolated strains from Kumamoto Chuo Hospital (Kumamoto, Japan). These bacteria were routinely grown in brucella broth (Becton Dickinson, Cockeysville, MD, USA) or on brucella agar plates supplemented with 10% fetal calf serum (Intergen, Purchase, NY, USA) under microaerobic conditions (i.e. 5% O2, 15% CO2, 80%N2) maintained in a GasPak jar (Becton Dickinson) with both hydrogen- and carbon dioxide- generating agents in addition to an oxygen-absorbing agent to result in 5% O2 (i.e. Helico Pak, Mitsubishi Gas Chemical, Tokyo, Japan). Broth-cultured bacteria were incubated at 37°C for 48 h under reciprocal shaking at 2 Hz in a water bath, and were then washed three times with 0.85% saline by centrifugation at 2000 g. This washing is required for experiments using ONOO exposure in a constant flux system, as described previously (3).

Treatment of H. pylori with ONOO

Authentic ONOO was synthesized from acidified NaNO2 and H2O2 by a quenched-flow method reported in the literature (12). Contaminated or remaining H2O2 was then decomposed by adding manganese dioxide. The concentration of ONOO was determined spectrophotometrically (13). The solution of ONOO was stored in alkaline solution pH of 11.0 at −80°C until use. The concentration of ONOO used in this study was usually 1.0 μM unless stated otherwise, which is pathophysiologically plausible. Because of the very short half-life of ONOO at acidic to neutral pH, (i.e. a few seconds at neutral pH), we used the constant flux method, in which the balance between influx–efflux (decomposition) maintains the ONOO concentration at 1.0 μM (3).

The bacterial suspension was adjusted to 109 CFU/ml with 0.5 M PBS (pH 7.6). An aliquot of 0.9 ml of this suspension was mixed gently in a 2-ml vial with a magnetic stirrer. ONOO solution was then infused into the vial for 10 min at a constant flow rate of 10 μl/min by using an automatic micro-syringe pump (Model ESP-64; Eicom, Kyoto, Japan). The concentration of ONOO in the reaction mixture was estimated by the dihydrorhodamine 123 oxidant assay method as described by Crow (14).

The ONOO treated bacterial suspension was then seeded onto 10 brucella-agar plates containing CLR at 0.2 μg/ml. The plates were incubated at 37°C for one week under microaerobic conditions as described above. The numbers of surviving bacteria were quantified by use of the colony-forming assay which usually required about one week.

CLR susceptibility test

CLR-resistant H. pylori strains, a standard wild-type reference strain ATCC 43504 and two wild-type strains of clinical isolates were tested for susceptibility to CLR by the agar dilution method, which is approved as the H. pylori susceptibility assay by the National Committee for Clinical Laboratory Standards of the Japanese Commission (15). Urease activity, cell motility, and morphology of the drug-resistant strains were confirmed before the CLR susceptibility test was undertaken.

Detection of mutants

Point mutations were identified by PCR amplification of portions of 23S rRNA and subsequent nucleotide sequencing of a 1402 bp PCR product. DNA from 45 CLR-resistant H. pylori isolates was extracted by use of an Easy-DNA kit (Invitrogen, Carlsbad, CA, USA). PCR amplifications were performed with primers derived from known sequences of the 23S rRNA gene: Heli-5 forward, 5′-AGTCGGGTCCTAAGCCGAG-3′ (positions 1445 to 1463; GenBank accession number U27270), and Heli-3 reverse, 5′-TTCCTGCTTAGATGCTTCAG-3′ (positions 2838 to 2854; GenBank accession number U27270). The PCR method involved 40 cycles consisting of a denaturation step at 94°C for 1 min, a primer annealing step at 50°C for 2 min, and an extension step at 72°C for 2 min, with a single first pre-incubation step at 94°C for 3 min and a single final extension step at 72°C for 7 min.

The PCR products obtained were then analyzed by electrophoresis with a 2% agarose gel in 0.05 M TAE buffer pH 8.2 run in parallel with a molecular mass marker: 100 bp DNA Ladder (New England Biolabs, Beverly, MA, USA) and stained with ethidium bromide. The PCR products were purified via the spin column QIAquick kit (QIAGEN GmbH, Hilden, Germany) and were sequenced by using the ABI PRISM Big Dye Terminator v1.1 cycle sequencing kit (Applied Biosystems, Foster, CA, USA) or DTCS Quick Start Master Mix (Beckman Coulter, CA, USA).

For sequencing primers, Heli-5, Heli-3, and the internal primer Heli-4 5′-ATGAGATGAGTATTCTAAGG-3′, (positions 1735 to 1748; GenBank accession number U27270) were used. DNA sequences were determined by using ABI PRISM 310 (Applied Biosystems), or CEQ 8000 Genetic Analysis System (Beckman Coulter).

Restriction analysis of PCR products

For rapid detection of mutation in the PCR products of 23S rRNA, these were digested with BsaI or MboII. For detection of the A2143G mutation, 5 U of BsaI (New England Biolabs) was added to each amplicon and digestions were performed for 14 h at 50°C. Identification of the A2142G mutation required incubation with 7.5 U of MboII (New England Biolabs) for 14 h at 37°C. The digested DNA patterns were analyzed by electrophoresis as described above.


  1. Top of page

Appearance of CLR-resistant mutants after treatment with ONOO

H. pylori was treated with ONOO under constant flux. Controls received no ONOO treatment. Exposure of H. pylori to ONOO for 10 min at a concentration of 1.0 μM killed about 50% of the organisms (Fig. 1a). H. pylori is very much vulnerable to the cytotoxicity of ONOO when compared with E. coli (Fig. 1b). It is considered that H. pylori is adapted to a microaerobic atmosphere, and thus is not equipped to survive against the insults of reactive oxygen species at high concentrations. It has also been reported that superoxide dismutase or catalase deficient Pseudomonas aeruginosa cannot grow as efficiently as wild type mutants (16). The pH values of these reaction mixtures before and after ONOO treatment (alkaline solution) were 7.57 and 7.96, respectively, and no killing or generation of mutant bacteria resulted from this slight pH shift alone.


Figure 1. Bactericidal effect of ONOO on H. pylori (a), and E. coli (b). The concentration of ONOO is maintained constant by use of constant flux (3). ONOO at a concentration of 1.0 μM killed about 50% of H. pylori while it was much higher for E. coli. The final concentrations of ONOO in the reaction mixtures were as follows: (▪) 0.5 μM; (•) 1.0 μM; (◆) 1.5 μM; and (▴) PBS (control) for H. pylori and (□) 1.0 μM; (⊚) 5.0 μM; (◊) 10.0 μM; and (▵) PBS (control) for E. coli. Data from three independent experiments are expressed as means ± SD.

Download figure to PowerPoint

The numbers of CLR-resistant colonies obtained were 45 in total: 35 after treatment with ONOO, and 10 without treatment (control) (Table 1). Isolated H. pylori CLR-resistant colonies showed negative results for gram staining but positive results for both bacterial motility and urease activity. There were 15 CLR-resistant mutants in the H. pylori ATCC strain, and 16 and 14 in the clinical isolates stains 1 and 2, respectively. These CLR-resistant colonies were numbered (Tables 1, 2). The frequency of appearance of CLR-resistant strains in H. pylori ATCC 43504 was 1 per 1.284 × 109 CFU after ONOO treatment compared with 1 per 9.946 × 109 CFU for the control. The frequency of appearance in the two clinical isolated strains was 1 per 0.498 × 109, 0.764 × 109 CFU after ONOO treatment compared with 1 per 4.587 × 109, 4.053 × 109 CFU for the control, respectively (Table 1). Thus, the relative mutation frequency of CLR-resistance in three strains after ONOO treatment was 7.1 times higher than that without ONOO treatment.

Table 1.  Emergence of CLR-resistant colonies after treatment with ONOO or PBS (control)a
Strains of H. pyloriColony formed (×107/ml)
Control1.0 μM ONOO exposure
Wild typeCLR-resistantWild typeCLR-resistant
  1. aTotal numbers of bacteria used per treatment were about 1 × 109 CFU/plate, with an average of 10 plates per experiment.

  2. Each agar plate contained 0.1 ml of bacterial suspension. These results are averages of 30 different repetitions. Upon treatment with ONOO at 1.0 μM, the number of colonies were reduced to about 50%.

(A) ATCC 43504
 Total colonies formed/30 experiments29843154112
   Mean mutation frequency1/994.71/128.4
   Relative mutation frequency17.7
(B) Clinical isolate strain No. 1
 Total colonies formed/30 experiments1376364713
   Mean mutation frequency1/458.71/49.8
   Relative mutation frequency19.2
(C) Clinical isolate strain No. 2
 Total colonies formed/30 experiments1618476410
   Mean mutation frequency1/404.51/76.4
   Relative mutation frequency15.3
(D) Sum of mutations in (A), (B), and (C)
 Total colonies formed/90 experiments597810295235
   Mean mutation frequency1/597.81/84.3
   Relative mutation frequency17.1
Table 2.  CLR susceptibility testa: MIC of mutant strains and wild type of H. pylori
Drug resistant mutant no.ATCC 43504Clinical isolate strain No. 1Clinical isolate strain No. 2
MIC (μg/ml)MIC (μg/ml)MIC (μg/ml)
  1. aCLR-resistant strains had MIC values more than 250 times higher than that of wild-type H. pylori.

 112.525 6.25
 22512.5 3.13
 725 6.2525
 92525 6.25
102512.5 6.25
122525 6.25
135012.5 6.25
1412.525 6.25
1525 6.25 
16 25 
Wild-type 0.05 0.025 0.025

The ONOO concentration used in this experiment (1.0 μM) is lower than that of a highly inflammatory region of infection, which could be as high as 10 μM. In addition, this in vitro medium did not contain any antioxidants such as glutathione and ascorbate, which do occur in vivo. At higher concentrations of ONOO, the increase in number of mutants would correlate with that seen for Sendai virus (1, 2), and Salmonella (3) in mouse models. We have previously shown that antioxidant content in the culture medium for Salmonella consumes about 2 μM equivalent of ONOO, and uric acid and α-tocopherol at 100 μM each reduce the mutation rate of Salmonella to 50% less than without antioxidants (3).

CLR susceptibility test and increased MIC in CLR-resistant strains

The MIC values of each CLR-resistant strain of H. pylori were determined (Table 2) and showed a marked increase from 0.05 μg/ml for wild-type H. pylori ATCC 43504 to 12.5 μg/ml or more, (i.e. a more than 250–500- fold increase), depending on the individual mutant strain. The MIC values of the other two clinical isolated strains increased after exposure to ONOO in similar proportions to the ATCC strain.

Restriction fragment length polymorphism analysis

DNA fragment patterns for the 45 CLR-resistant strains after restriction enzyme treatments with BsaI (A2143G mutation) or MboII (A2142G mutation) are shown in Figure 2. DNA of 17 CLR-resistant strains was cleaved with BsaI to yield bands near 700, 400, and 300 bp (lanes 1 and 2). Cleavage of DNA from 22 other CLR-resistant strains by MboII yielded a band near 700 bp (lane 3 and 4) as a dominant band; the DNA from the remaining strains was not cleaved by either BsaI or MboII. They were analyzed by DNA sequencer and the points of mutation were identified as described below.


Figure 2. Restriction fragment length polymorphism analysis. Lane 1, 2: CLR-resistant (A2143G), lane 3, 4: CLR-resistant (A2142G), lane 5: H. pylori ATCC 43504, Lane 6, 7: other mutation, Lane 8: standard marker. BsaI and Mbo II are endonucleases, the restriction enzymes cleaving DNAase.

Download figure to PowerPoint

DNA sequence analysis

DNA sequence analyses were performed to identify the mutation sites in the 45 strains. Data for 39 strains were consistent with results of the restriction fragment length polymorphism analysis of DNA mutation described above. The six strains that were not cleaved by either BsaI or MboII, were an A2142T, a T2269G, C2695G (2), T1944C (2) mutation. As shown in Table 3, when H. pylori strains were treated with PBS alone, three spontaneous mutations, A2142G, A2143G andT1944C were found. After treatment with ONOO, however, results were different: A2142G, A2143G, A2142T, T2269G, C2695G and T1944C mutations were found, where underlined denotes newly formed.

Table 3.  DNA sequence analyses for identification of sites of mutation in 45 CLR-resistant strains of H. pylori
TreatmentNo. of mutants isolatedSites of mutation
  1. aControl, 0.5 M phosphate buffer containing 0.15 M NaCl (pH 7.6).

  2. bONOO, in the same buffer as for the control.

  3. Two clinical isolates and one ATCC strain were subjected to this treatment (cf. Tables 1, 2).

None 4A2142G
(PBS control)a 5A2143G
1.0 μM ONOO18A2142G
(Constant flux)b12A2143G


  1. Top of page

The present series of experiments involving treatment of H. pylori strains with physiological levels of ONOO showed a significant increase in the occurrence of drug-resistant mutant strains. ONOO, one of the most reactive compounds produced by the host during an inflammatory response, is formed by an extremely rapid reaction between O·2 and NO. ONOO mediates cleavage of either DNA or RNA and nitration and hydroxylation of guanine (the preferred site). During these DNA modifications, guanine (guanosine) seems to be the most susceptible of the four bases, as tested in a cell-free system. In H. pylori, the preferred sites of mutation are 2142 and 2143. Despite the preference of ONOO for guanine, we found here that adenine was also a site of point mutation, that is, A to G and A to T alterations were found. These sites are the same as those reported previously (1, 2, 17), with the exception of the 2695 C to G, which is a new site. We have reported that ONOO not only nitrates guanine, but also cleaves to depurinate it, or leaves the phosphodiester linkage of deoxyribose (18). This process may lead to insertion of guanine during the DNA repair process. Furthermore, we have found that 8-nitroguanosine becomes an efficient catalyst of superoxide generation in the presence of NADPH and oxygen plus one of the redox-catalyzing enzymes: cytochrome NADPH P450 reductase, nitric oxide synthase (any isoform of inducible, neuronal, or endothelial origin), or cytochrome NADPH b5 reductase (10, 11). One mole of 8-nitroguanosine can generate a large excess of O·2 and can be converted to H2O2 by superoxide dismutase, with formation of hydroxyl radical from H2O2 catalyzed by transition metals, this hydroxyl radical being the most reactive of the reactive oxygen species. Our recent study of in vivo Sendai virus infection showed frequent transitions of A[RIGHTWARDS ARROW]G, G[RIGHTWARDS ARROW]A, U[RIGHTWARDS ARROW]C, and C[RIGHTWARDS ARROW]U, and the transversions A[RIGHTWARDS ARROW]U, A[RIGHTWARDS ARROW]C, and many others (1). Therefore, it is not surprising to find the A[RIGHTWARDS ARROW]G transition as well as the A[RIGHTWARDS ARROW]T or C[RIGHTWARDS ARROW]G transversion, since both A and G in Sendai virus, and 23Sr RNA are identical in chemical structure. In addition the present experiments conducted on 23Sr RNA would be expected to have the same chemical susceptibility as Sendai virus, and the effect of ONOO will be the same in either system. A to G may take place when an A site generates an apurinic site upon oxidation and cleaving off sugar residue, and mismatch replacement to G could have occurred. Further adenine content may be higher in the base composition, which favors higher probability of effects on A.

In the same context, highly sensitive DNA damage observed in H. pylori (Fig. 2) may not be applicable to the wild type aerobic microbes that are fully equipped against the insults of ROS. Further, in the case of single stranded RNA viruses, damage to the genome is more crucial than in double stranded genes, since there is no complementing counterpart during the repair process. The uniqueness of the present observation is that we examined the effect of ONOO on 23S rRNA, which is involved in CLR binding (17, 19, 20). Thus, chemical modification of this rRNA (CLR binding site) may be directly reflected as CLR resistance, and a high incidence of mutation may be expected as with the Sendai virus.

An inflammatory response thus generates reactive free radicals in vivo (5, 21–23), which is beneficial for eliminating infecting microbes. However, it also induces a higher occurrence of mutant bacteria, either drug resistant or with other functions for survival and adaptation, as these pathogens take advantage of free radical-induced mutant formation.


  1. Top of page
  • 1
    Yoshitake J., Akaike T., Akuta T., Tamura F., Ogura T., Esumi H., Maeda H. (2004) Nitric oxide as an endogenous mutagen for Sendai virus without antiviral activity. J Virol 78: 870919.
  • 2
    Akaike, T., Fujii S., Kato A., Yoshitake J., Miyamoto Y., Sawa T., Okamoto S., Suga M., Asakawa M., Nagai Y., Maeda H. (2000) Viral mutation accelerated by nitric oxide production during infection in vivo. FASEB J 14: 144754.
  • 3
    Kuwahara H., Kanazawa A., Wakamatu D., Morimura S., Kida K., Akaike T., Maeda H. (2004) Antioxidative and antimutagenic activities of 4-vinyl-2, 6-dimethoxyphenol (canolol) isolated from canola oil. J Agric Food Chem 52: 43807.
  • 4
    Kang J.M., Iovine N.M., Blaser M.J. (2006) A paradigm for direct stress-induced mutation in prokaryotes. FASEB J 20: 247685.
  • 5
    Akaike T., Suga M., Maeda H. (1998) Free radicals in viral pathogenesis: molecular mechanisms involving superoxide and NO. Proc Soc Exp Biol Med 217: 6473.
  • 6
    Akaike T., Okamoto S., Sawa T., Yoshitake J., Tamura F., Ichimori K., Miyazaki K., Sasamoto K., Maeda H. (2003) 8-Nitroguanosine formation in viral pneumonia and its implication for pathogenesis. Proc Natl Acad Sci USA 100: 68590.
  • 7
    Kuwahara H., Miyamoto Y., Akaike T., Kubota T., Sawa T., Okamoto S., Maeda H. (2000) Helicobacter pylori urease suppresses bactericidal activity of peroxynitrite via carbon dioxide production. Infect Immun 68: 437883.
  • 8
    Pryor W.A., Squadrito G.L. (1995) The chemistry of peroxynitrite: A product from the reaction of nitric oxide and superoxide. Am J Physiol 268: L699L722.
  • 9
    Beckman J.S. (2002) Protein tyrosine nitration and peroxynitrite. FASEB J 16: 1144.
  • 10
    Sawa T., Akaike T., Ichimori K., Akuta T., Kaneko K., Nakayama  , Stuehr J.D., Maeda H. (2003) Superoxide generation mediated by 8-nitroguanosine, a highly redox-active nucleic acid derivative. Biochem Biophys Res Commun 311: 3006.
  • 11
    Maeda H., Sawa T., Yubisui T., Akaike T. (1999) Free radical generation from heterocyclic amines by cytochrome b5 reductase in the presence of NADH. Cancer Lett 143: 11721.
  • 12
    Koppenol W.H., Kissner R., Beckman J.S. (1996) Synthesis of peroxynitrite: To go with the flow or on solid ground? Methods Enzymol 269: 296302.
  • 13
    Sawa T., Akaike T., Maeda H. (2000) Tyrosine nitration by peroxynitrite formed from nitric oxide and superoxide generated by xanthine oxidase. J Biol Chem 275: 3246774.
  • 14
    Crow J.P. (1997) Dichlorodihydrofluorescein and dihydrorhodamine 123 are sensitive indicators of peroxynitrite in vitro: Implications for intracellular measurement of reactive nitrogen and oxygen species. Nitric Oxide Biol Chem 1: 14557.
  • 15
    Osato M.S., Reddy R., Reddy S.G., Penland R.L., Graham D.Y. (2001) Comparison of the E test and the NCCLS-approved agar dilution method to detect metronidazole and clarithromycin resistant Helicobacter pylori. Int J Antimicrob Agents 17: 3944.
  • 16
    Hassett D.J., Schweizer H.P., Ohman D.E. (1995) Pseudomonas aeruginosa sodA and sodB mutants defective in manganese and iron-cofactored superoxide dismutase activity demonstrate the importance of the iron-cofactored form in aerobic metabolism. J Bact 177: 63307.
  • 17
    Van Doorn L.J., Glupczynski Y., G. Kusters J.  , Megraud F., Midolo P., Maggi-Solca N., Queiroz D.M., Nouhan N., Stet E., Quint W.G. (2001) Accurate prediction of acrolide resistance in Helicobacter pylori by a PCR line probe assay for detection of mutations in the 23S rRNA gene: multicenter validation study. Antimicrob Agents Chemother 45: 15004.
  • 18
    Kanazawa A., Sawa T., Akaike T., Maeda H. (2000) Formation of abasic sites in DNA by t-butyl peroxyl radicals: Implication for potent toxicity of lipid peroxyl radicals. Cancer Lett 156: 515.
  • 19
    Versalovic J., Osato M.S., Spakovsky K., Dore M.P., Reddy R., Stone G.G., Shortridge D., Flamm R.K., Tanaka S.K., Graham D.Y. (1997) Point mutations in the 23S rRNA gene of Helicobacter pylori associated with different levels of clarithromycin resistance. J Antimicrob Chemother 40: 2836.
  • 20
    Taylor D.E. (2000) Pathophysiology of antibiotic resistance: clarithromycin. Can J Gastroenterol 14: 8914.
  • 21
    Oda T., Akaike T., Hamamoto T., Suzuki F., Hirano T., Maeda H. (1989) Oxygen radicals in influenza-induced pathogenesis and treatment with pyran polymer-conjugated SOD. Science 26: 9746.
  • 22
    Maeda H., Akaike T. (1991) Oxygen free radicals as pathogenic molecules in viral diseases. Proc Soc Exp Biol Med 198: 7217.
  • 23
    Maeda H., Akaike T. (1998) Nitrite oxide and oxygen radicals in infection, inflammation, and cancer. Biochemistry (Moscow) 63: 85465.