Presence of Copper- and Zinc-Containing Superoxide Dismutase in Commen
http://www.100md.com
《微生物临床杂志》
Department of Paediatrics, Division of Medicine, Imperial College, St. Mary's Campus, London W2 1PG, United Kingdom
Departments of Microbiology and Medicine, University of Buffalo, State University of New York, and VA Western New York Healthcare System, Buffalo, New York
ABSTRACT
Respiratory isolates of Haemophilus haemolyticus are regularly misclassified as nontypeable (NT) Haemophilus influenzae due to an aberrant hemolytic reaction on blood agar, with implications for treatment. The presence of sodC or its cognate protein, copper-zinc superoxide dismutase, can distinguish respiratory isolates of H. haemolyticus from NT H. influenzae with 100% accuracy.
TEXT
Infection with nontypeable (NT) Haemophilus influenzae is the most common cause of exacerbations of chronic obstructive pulmonary disease (COPD), characterized by the increased production of mucopurulent sputum and airway obstruction, and is implicated in disease progression (25). The mechanisms underlying the association between NT H. influenzae acquisition and exacerbations are poorly understood (5, 18). In a prospective study of COPD patients, Murphy et al. (19) noted that selected sputum isolates of apparent NT H. influenzae had an altered phenotype, including slower growth in broth culture and a tendency to form aggregates. On the basis of genetic analyses (ribosomal DNA sequencing, multilocus sequence typing, DNA-DNA hybridization, and sequencing of the gene encoding the outer membrane protein P6), the variant isolates were proven to be H. haemolyticus, a bacterium that had not previously been considered a constituent of the respiratory tract flora. By conventional criteria, 102 of 258 (39.5%) and 12 of 44 (27.3%) sputum and nasopharyngeal isolates, respectively, had been misclassified (19). In routine analysis in the clinical laboratory, hemolysis on blood agar is the only characteristic that distinguishes NT H. influenzae from H. haemolyticus (4, 9, 10). NT H. influenzae is the most common bacterial cause of exacerbations of COPD, whereas H. haemolyticus is a commensal and does not cause exacerbations. Because most authorities recommend antibiotic therapy for moderate to severe exacerbations of COPD, a rapid and accurate test that discriminates between H. influenzae and H. haemolyticus will lead to more rational antibiotic therapy for adults with COPD and reduce the unnecessary use of antibiotics. In the work described here, we demonstrate that the presence of sodC (or the copper- and zinc-cofactored superoxide dismutase [CuZnSOD] that it encodes) can reliably distinguish H. haemolyticus from NT H. influenzae.
Bacterial CuZnSODs are enzymes found in the periplasms of various gram-negative bacteria and, in the case of pathogens, are thought to have a role in the protection of organisms against host defense-derived free-radical-mediated damage (1, 2, 6, 11, 12, 15, 27). In extensive studies, we have established that capsulated or noncapsulated H. influenzae sensu stricto strains do not produce active CuZnSOD (11, 12, 14, 16). The population of capsulated H. influenzae strains falls into two widely separated phylogenetic divisions, and while strains segregating to phylogenetic division II (some serotype a and b strains and serotype f strains) plus serotype e strains (distantly related to phylogenetic division II) possess the sodC gene, they do not produce active enzyme (11, 16), probably due to a mutation that converts an active-site histidine to tyrosine (11). Among a collection of 45 NT H. influenzae isolates (20), characterized by multilocus enzyme electrophoresis, comprising electrophoretic types 11 to 13, 26 to 27, 29 to 32, 35 to 45, 49 to 51, 53 to 55, 57 to 61, and 63 to 76, neither the sodC gene nor the CuZnSOD protein was detected (16). The only NT H. influenzae strains that have been shown to produce CuZnSOD are the cryptic genospecies biotype IV strains responsible for urogenital and neonatal infections (16). The cryptic biotype IV strains are genetically distinct, although they are related to H. influenzae and H. haemolyticus (21-23); but they have not been isolated from the respiratory tract. The cryptic genospecies biotype IV strains can be distinguished from H. haemolyticus by a positive ornithine decarboxylase reaction. In a previous study we described the presence of a sodC gene and active CuZnSOD in a single strain of H. haemolyticus (14). We now hypothesize that the presence of the sodC gene and/or CuZnSOD could differentiate H. haemolyticus from NT H. influenzae isolates obtained from the respiratory tract.
Initially, a collection of NT H. influenzae (n = 10) and H. haemolyticus (n = 10) isolates that formed part of the study describing the misclassification of the two bacteria (19) was investigated. The Haemophilus isolates used in phase 1 of this study, which, for clarity, are given a short name in parentheses, were 1P26H (TIM1), 3P5H (TIM2), 3P16H1 (TIM3), 3P18H1 (TIM4), 5P1H (TIM5), 6P18H1 (TIM6), 7P49H1 (TIM7), 11P18H1 (TIM8), 14P41H1 (TIM9), 14P38H1 (TIM10), 24P19H1 (TIM11), 27P25H (TIM12), 48P38H1 (TIM13), 49P5H1 (TIM14), 55P13H1 (TIM15), 58P6H1 (TIM16), 65P13H1 (TIM17), 67P13H1 (TIM18), 70P28H1 (TIM19), and 73P18H1 (TIM20). This phase of the study was blind, with only one of the authors (T.F.M.) knowing the species designation. Strains used as positive controls were H. haemolyticus NCTC 10839, biotype IV cryptic genospecies H. influenzae strain 16N (16), Neisseria meningitidis MC58 (27), and Actinobacillus pleuropneumoniae 4074 ATCC 27088 (12, 13, 26); H. influenzae Rd (7, 11) was used as the negative control. A. pleuropneumoniae 4074 and Haemophilus isolates were grown in brain heart infusion (BHI) broth supplemented with 1 μg/ml NAD and 10 μg/ml hemin at 37°C on an orbital shaker (180 rpm) or on BHI agar plates supplemented with Levinthal's base at 37°C in 5% carbon dioxide (11). N. meningitidis was cultured as described previously (27). Chromosomal DNA was prepared from the plate-grown organisms (17), and standard methods were used for restriction endonuclease digestion and Southern blotting, with washing to 80% stringency (24). PCR amplification of genomic DNA with the degenerate primers 3'UNIVSOD and 5'UNIVSOD, designed to amplify a 300-bp conserved region of prokaryotic sodC genes, was performed as described previously (12), except that an annealing temperature of 44°C was used.
Use of the degenerate oligonucleotide primers and genomic DNA as the template resulted, as expected, in the PCR amplification of a number of fragments for each isolate (Fig. 1). However, for 10 isolates (TIM1, TIM2, TIM4, TIM8, TIM10, TIM12, TIM16, TIM17, TIM19, and TIM20) there was a band of 300 bp, indicative of the presence of a sodC gene. The strong 300-bp product was distinct from a fainter band of 280 bp present in isolates TIM3, TIM5, TIM6, TIM9, TIM14, and TIM15. The positive controls (H. haemolyticus NCTC 10839, H. influenzae 16N, and N. meningitidis MC58) amplified a corresponding 300-bp PCR product, while the negative control (H. influenzae Rd) did not. To confirm the presence or absence of sodC, as indicated by the PCR results, Southern blots were probed with a digoxigenin (DIG)-labeled 509-bp fragment from the sodC gene of H. parainfluenzae 1391 (12). The H. parainfluenzae internal sodC probe was labeled with a PCR DIG probe synthesis kit (Roche) by using oligonucleotides ATGATGAAAATGAAAACT and TGATCTGAGTGGTTATC with plasmid pJSK134 (12) as the DNA template. The H. parainfluenzae sodC probe hybridized to the EcoRI restriction fragment-digested genomic DNA of all the Haemophilus isolates when a 300-bp fragment had been amplified by PCR but not to the DNA of those isolates from which this band was absent (Fig. 2). The sodC probe hybridized to the positive controls (H. haemolyticus NCTC 10839 and H. influenzae biotype IV) but not the negative control (H. influenzae Rd). There were variations in the sizes of the EcoRI fragments that were hybridized by the sodC probe, indicative of genetic heterogeneity between such isolates.
Next we determined whether the CuZnSOD protein was produced by Haemophilus isolates. Whole-cell extracts were prepared as described previously (14, 15), except that the FastPrep FP120 instrument (ThermoElectron) instead of sonication was used to disrupt the cells. The settings used were two runs at setting 6.0 for 40 s each, with each run separated by a 10-min cooling step at 4°C. Prior experiments had determined that this resulted in optimal cell breakage. The protein concentration was measured on an ND-1000 spectrophotometer (Nanodrop Technologies). Whole-cell extracts (12.5 μg) were spotted onto Hybond-C (GE Healthcare) membranes; and their reactivities with mouse monoclonal antibody HD1, which recognizes Haemophilus, A. pleuropneumoniae, and meningococcal CuZnSODs, were determined. The membranes were blocked for 30 min with 5% skim milk in phosphate-buffered saline (PBS) containing 0.3% Tween 20 (PBS-T). Primary antibody (HD1) was used at a 1:500 dilution of hybridoma supernatant (with incubation for 60 min), and secondary antibody (rabbit anti-mouse immunoglobulin-horseradish peroxidase conjugate; Dako) was used at a 1:10,000 dilution for 30 min. PBS-T was used for all washes except the final wash, which was done in PBS alone. The dot blots were developed with ECL-Plus (GE Healthcare), according to the manufacturer's instructions. Again, there was a perfect correlation between those isolates that possessed sodC, as determined by PCR or Southern blotting, and the presence of CuZnSOD, as indicated by the recognition by HD1 (Fig. 3).
Finally, we determined whether isolates possessed active CuZnSOD. This was achieved by separation of whole-cell extracts by isoelectric focusing (IEF) with pH 3 to 10 Ready gels (Bio-Rad) and SOD activity staining, coupled with differential inhibition, as described previously (15). The copper chelator diethyl dithiocarbamic acid (DEDC, 10 mM) was used as the inhibitor of CuZnSOD activity (3, 15). SOD gels demonstrating the presence of CuZnSOD activity in eight isolates (TIM1, TIM2, TIM4, TIM8, TIM10, TIM12, TIM16, and TIM17, also representative of TIM19 and TIM20), an example of a single strain (TIM18) representative of nine other strains (TIM3, TIM5 to TIM7, TIM9, TIM11, and TIM13 to TIM15), plus the positive control A. pleuropneumoniae 4074, are shown in Fig. 4. In the eight isolates containing CuZnSOD activity (TIM1, TIM2, TIM4, TIM8, TIM10, TIM12, TIM16, and TIM17; Fig. 4, lanes 1 to 8, respectively), two achromatic bands (arrowed) that disappear with DEDC are clearly visible. In contrast, the single band of SOD activity seen in TIM18 was unaffected by DEDC treatment. The positive control (A. pleuropneumoniae 4074) showed three bands of activity, two of which were DEDC inhibitable. Active CuZnSOD (DEDC inhibitable) was found only in those Haemophilus isolates that (i) contained a sodC gene, as indicated by the presence of the 300-bp PCR product by using 3'- and 5'UNIVSODs as primers or hybridization of the H. parainfluenzae sodC probe in Southern blotting experiments, and (ii) produced the CuZnSOD protein, as determined by reactivity with monoclonal antibody HD1. The presence of two bands with CuZnSOD activity in the IEF gels is probably due to the fact that H. haemolyticus SodC is dimeric. HD1 does not recognize monomeric CuZnSODs from Salmonella enterica serovar Typhimurium or Escherichia coli. The monoclonal antibody does, however, recognize dimeric SODs, such as that from A. pleuropneumoniae 4074 (8), which also showed two bands on the IEF gel (Fig. 4). In Western blots, HD1 recognized a protein of approximately 50 kDa in all the H. haemolyticus isolates, i.e., TIM1, TIM2, TIM4, TIM8, TIM10, TIM12, TIM16, TIM17, TIM19, and TIM20, when there was reactivity with HD1 (data not shown). Given the presence of the 300-bp band upon PCR with 3'- and 5'UNIVSOD primers, designed to amplify a highly conserved C-terminal region of sodC genes, it is probable that the structure of the full-length protein of H. haemolyticus is similar to those established for other prokaryotic CuZnSODs. Typically, the molecular mass of the predicted encoded proteins for known monomeric CuZnSODs or a single subunit of dimeric SODs is 17 to 25 kDa (2, 15). Thus, the available evidence suggests that the H. haemolyticus CuZnSOD is dimeric, although this awaits definitive confirmation.
these results with blinded isolates, isolates TIM1, TIM2, TIM4, TIM8, TIM10, TIM12, TIM16, TIM17, TIM19, and TIM20 were predicted to be H. haemolyticus and the remainder were predicted to be NT H. influenzae. When the code was broken, all of the predictions were found to be correct. To confirm the usefulness of sodC/CuZnSOD as a means of distinguishing between H. haemolyticus and NT H. influenzae, a further 20 strains were analyzed in a blinded manner. These were strains 1P67H1 (TIM21), 3P60H1 (TIM22), 5P54H1 (TIM23), 6P23H3 (TIM24), 7P24H1 (TIM25), 12P56H1 (TIM26), 13P13H5 (TIM27), 18P25H1 (TIM28), 19P49H1 (TIM29), 24P21H1 (TIM30), 26P1H1 (TIM31), 28P15H2 (TIM32), 31P13H1 (TIM33), 32P3H1 (TIM34), 34P4H6 (TIM35), 39P8H1 (TIM36), 43P2H1 (TIM37), 45P9H1 (TIM38), 47P68H1 (TIM39), and 50P5H1 (TIM40). In view of the 100% correlation between the results of PCR, Southern blotting, dot blotting, and CuZnSOD activity in the initial phase, only PCR and dot blotting with monoclonal antibody HD1 were performed. Isolates TIM21, TIM22, TIM24, TIM25, TIM27, TIM28, TIM30, TIM32, TIM34, and TIM36 were predicted to be H. haemolyticus, and the remainder were predicted to be NT H. influenzae. When the code was broken, 100% predictive accuracy was again achieved.
We conclude that the presence of a sodC gene or CuZnSOD protein can reliably differentiate between NT H. influenzae and H. haemolyticus strains isolated from the respiratory tract. A test based on the CuZnSOD activity of single bacterial colonies on agar plates or in broth based on a chromogenic reaction would be ideal; but this is challenging, as current methods do not distinguish between the three major bacterial forms of SOD (MnSOD, FeSOD, and CuZnSOD) and a reliable inhibitor for prokaryotic MnSODs is lacking. Nevertheless, while we await appropriate developments for simple rapid tests in the field of SOD assays, the procedures described herein can clearly distinguish H. haemolyticus from NT H. influenzae. A test that combines monoclonal antibodies HD1 (which recognized the CuZnSOD of H. haemolyticus but not that of NT H. influenzae) and 7F3 (which recognized the P6 protein of NT H. influenzae but not that of H. haemolyticus) (19) has some merit. Alternatively, a multiplex PCR based on H. haemolyticus sodC and a gene unique to NT H. influenzae might be developed. It should be noted that in a previous study representative strains of H. parahaemolyticus and H. paraphrohaemolyticus produced active CuZnSOD (15). Therefore, the presence of CuZnSOD cannot be used to distinguish H. parahaemolyticus or H. paraphrohaemolyticus from H. haemolyticus. We speculate that the CuZnSODs produced by both H. parahaemolyticus and H. paraphrohaemolyticus would be recognized by monoclonal antibody HD1, although this remains to be confirmed.
The finding that H. haemolyticus, a relatively newly recognized commensal of the lower respiratory tract, produces active CuZnSOD, while NT H. influenzae (which is associated with new exacerbations in COPD patients) does not suggests that SodC is not an important virulence factor, at least in the context of the lower respiratory tract. It is not known whether H. haemolyticus produces CuZnSOD in vivo, and its role in host interactive biology, if any, remains to be determined.
In summary we have described a marker (sodC and/or CuZnSOD) that differentiates H. haemolyticus from NT H. influenzae. While the methods used in the present study would be too time-consuming to become part of a routine laboratory protocol, they pave the way forward for the development of a simple rapid test.
ACKNOWLEDGMENTS
This work was supported by awards from The Wellcome Trust and BBSRC (to P.R.L. and J.S.K.) and by NIH grant AI 19641 and the U.S. Department of Veteran Affairs (to T.F.M.).
FOOTNOTES
Corresponding author. Mailing address: Department of Paediatrics, Division of Medicine, Imperial College, St Mary's Campus, London W2 1PG, United Kingdom. Phone: 44 (0)2075943359. Fax: 44 (0)2075943984. E-mail: p.langford@imperial.ac.uk.
Published ahead of print on 20 September 2006.
W.W.M.F. and C.A.O. contributed equally to the work.
REFERENCES
Ammendola, S., M. Ajello, P. Pasquali, J. S. Kroll, P. R. Langford, G. Rotilio, P. Valenti, and A. Battistoni. 2005. Differential contribution of sodC1 and sodC2 to intracellular survival and pathogenicity of Salmonella enterica serovar Choleraesuis. Microbes Infect. 7:698-707.
Battistoni, A. 2003. Role of prokaryotic Cu,Zn superoxide dismutase in pathogenesis. Biochem. Soc. Trans. 31:1326-1329.
Benov, L. T., and I. Fridovich. 1994. Escherichia coli expresses a copper- and zinc-containing superoxide dismutase. J. Biol. Chem. 269:25310-25314.
Collins, C. H., P. M. Lyne, J. M. Grange, and J. O. Falkingham. 2004. Collins and Lyne's microbiological methods, 8th ed. Arnold, London, United Kingdom.
Eldika, N., and S. Sethi. 2006. Role of nontypeable Haemophilus influenzae in exacerbations and progression of chronic obstructive pulmonary disease. Curr. Opin. Pulm. Med. 12:118-124.
Farrant, J. L., A. Sansone, J. R. Canvin, M. J. Pallen, P. R. Langford, T. S. Wallis, G. Dougan, and J. S. Kroll. 1997. Bacterial copper- and zinc-cofactored superoxide dismutase contributes to the pathogenesis of systemic salmonellosis. Mol. Microbiol. 25:785-796.
Fleischmann, R. D., M. D. Adams, O. White, R. A. Clayton, E. F. Kirkness, A. R. Kerlavage, C. J. Bult, J. F. Tomb, B. A. Dougherty, J. M. Merrick, et al. 1995. Whole-genome random sequencing and assembly of Haemophilus influenzae Rd. Science 269:496-512.
Forest, K. T., P. R. Langford, J. S. Kroll, and E. D. Getzoff. 2000. Cu,Zn superoxide dismutase structure from a microbial pathogen establishes a class with a conserved dimer interface. J. Mol. Biol. 296:145-153.
Holt, J. G., N. R. Krieg, P. H. A. Sneath, H. Staley, and S. T. Williams. 1994. Bergey's manual of determinative bacteriology, 9th ed. The Williams & Wilkins Co., Baltimore, Md.
Kilian, M. 2003. Haemophilus, p. 623-635. In P. R. Murray, E. J. Baron, J. H. Jorgensen, M. A. Pfaller, and R. H. Yolken (ed.), Manual of clinical microbiology, 8th ed. ASM Press, Washington, D.C.
Kroll, J. S., P. R. Langford, and B. M. Loynds. 1991. Copper-zinc superoxide dismutase of Haemophilus influenzae and H. parainfluenzae. J. Bacteriol. 173:7449-7457.
Kroll, J. S., P. R. Langford, K. E. Wilks, and A. D. Keil. 1995. Bacterial [Cu, Zn]-superoxide dismutase: phylogenetically distinct from the eukaryotic enzyme, and not so rare after all! Microbiology 141:2271-2279.
Langford, P. R., B. M. Loynds, and J. S. Kroll. 1996. Cloning and molecular characterization of Cu,Zn superoxide dismutase from Actinobacillus pleuropneumoniae. Infect. Immun. 64:5035-5041.
Langford, P. R., B. M. Loynds, and J. S. Kroll. 1992. Copper-zinc superoxide dismutase in Haemophilus species. J. Gen. Microbiol. 138:517-522.
Langford, P. R., A. Sansone, P. Valenti, A. Battistoni, and J. S. Kroll. 2002. Bacterial superoxide dismutase and virulence. Methods Enzymol. 349:155-166.
Langford, P. R., B. J. Sheehan, T. Shaikh, and J. S. Kroll. 2002. Active copper- and zinc-containing superoxide dismutase in the cryptic genospecies of Haemophilus causing urogenital and neonatal infections discriminates them from Haemophilus influenzae sensu stricto. J. Clin. Microbiol. 40:268-270.
Moxon, E. R., R. A. Deich, and C. Connelly. 1984. Cloning of chromosomal DNA from Haemophilus influenzae. Its use for studying the expression of type b capsule and virulence. J. Clin. Investig. 73:298-306.
Murphy, T. F. 2006. The role of bacteria in airway inflammation in exacerbations of chronic obstructive pulmonary disease. Curr. Opin. Infect. Dis. 19:225-230.
Murphy, T. F., A. L. Brauer, S. Sethi, M. Kilian, X. Cai, and A. J. Lesse. Haemophilus haemolyticus: a human respiratory tract commensal to be distinguished from Haemophilus influenzae. J. Infect. Dis., in press.
Musser, J. M., S. J. Barenkamp, D. M. Granoff, and R. K. Selander. 1986. Genetic relationships of serologically nontypable and serotype b strains of Haemophilus influenzae. Infect. Immun. 52:183-191.
Quentin, R., A. Goudeau, R. J. Wallace, Jr., A. L. Smith, R. K. Selander, and J. M. Musser. 1990. Urogenital, maternal and neonatal isolates of Haemophilus influenzae: identification of unusually virulent serologically non-typable clone families and evidence for a new Haemophilus species. J. Gen. Microbiol. 136:1203-1209.
Quentin, R., C. Martin, J. M. Musser, N. Pasquier-Picard, and A. Goudeau. 1993. Genetic characterization of a cryptic genospecies of Haemophilus causing urogenital and neonatal infections. J. Clin. Microbiol. 31:1111-1116.
Quentin, R., J. M. Musser, M. Mellouet, P. Y. Sizaret, R. K. Selander, and A. Goudeau. 1989. Typing of urogenital, maternal, and neonatal isolates of Haemophilus influenzae and Haemophilus parainfluenzae in correlation with clinical source of isolation and evidence for a genital specificity of H. influenzae biotype IV. J. Clin. Microbiol. 27:2286-2294.
Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Sethi, S., N. Evans, B. J. Grant, and T. F. Murphy. 2002. New strains of bacteria and exacerbations of chronic obstructive pulmonary disease. N. Engl. J. Med. 347:465-471.
Sheehan, B. J., P. R. Langford, A. N. Rycroft, and J. S. Kroll. 2000. [Cu, Zn]-superoxide dismutase mutants of the swine pathogen Actinobacillus pleuropneumoniae are unattenuated in infections of the natural host. Infect. Immun. 68:4778-4781.
Wilks, K. E., K. L. Dunn, J. L. Farrant, K. M. Reddin, A. R. Gorringe, P. R. Langford, and J. S. Kroll. 1998. Periplasmic superoxide dismutase in meningococcal pathogenicity. Infect. Immun. 66:213-217.(W. W. Mary Fung, Clíona A. O'Dwyer, Suni)
Departments of Microbiology and Medicine, University of Buffalo, State University of New York, and VA Western New York Healthcare System, Buffalo, New York
ABSTRACT
Respiratory isolates of Haemophilus haemolyticus are regularly misclassified as nontypeable (NT) Haemophilus influenzae due to an aberrant hemolytic reaction on blood agar, with implications for treatment. The presence of sodC or its cognate protein, copper-zinc superoxide dismutase, can distinguish respiratory isolates of H. haemolyticus from NT H. influenzae with 100% accuracy.
TEXT
Infection with nontypeable (NT) Haemophilus influenzae is the most common cause of exacerbations of chronic obstructive pulmonary disease (COPD), characterized by the increased production of mucopurulent sputum and airway obstruction, and is implicated in disease progression (25). The mechanisms underlying the association between NT H. influenzae acquisition and exacerbations are poorly understood (5, 18). In a prospective study of COPD patients, Murphy et al. (19) noted that selected sputum isolates of apparent NT H. influenzae had an altered phenotype, including slower growth in broth culture and a tendency to form aggregates. On the basis of genetic analyses (ribosomal DNA sequencing, multilocus sequence typing, DNA-DNA hybridization, and sequencing of the gene encoding the outer membrane protein P6), the variant isolates were proven to be H. haemolyticus, a bacterium that had not previously been considered a constituent of the respiratory tract flora. By conventional criteria, 102 of 258 (39.5%) and 12 of 44 (27.3%) sputum and nasopharyngeal isolates, respectively, had been misclassified (19). In routine analysis in the clinical laboratory, hemolysis on blood agar is the only characteristic that distinguishes NT H. influenzae from H. haemolyticus (4, 9, 10). NT H. influenzae is the most common bacterial cause of exacerbations of COPD, whereas H. haemolyticus is a commensal and does not cause exacerbations. Because most authorities recommend antibiotic therapy for moderate to severe exacerbations of COPD, a rapid and accurate test that discriminates between H. influenzae and H. haemolyticus will lead to more rational antibiotic therapy for adults with COPD and reduce the unnecessary use of antibiotics. In the work described here, we demonstrate that the presence of sodC (or the copper- and zinc-cofactored superoxide dismutase [CuZnSOD] that it encodes) can reliably distinguish H. haemolyticus from NT H. influenzae.
Bacterial CuZnSODs are enzymes found in the periplasms of various gram-negative bacteria and, in the case of pathogens, are thought to have a role in the protection of organisms against host defense-derived free-radical-mediated damage (1, 2, 6, 11, 12, 15, 27). In extensive studies, we have established that capsulated or noncapsulated H. influenzae sensu stricto strains do not produce active CuZnSOD (11, 12, 14, 16). The population of capsulated H. influenzae strains falls into two widely separated phylogenetic divisions, and while strains segregating to phylogenetic division II (some serotype a and b strains and serotype f strains) plus serotype e strains (distantly related to phylogenetic division II) possess the sodC gene, they do not produce active enzyme (11, 16), probably due to a mutation that converts an active-site histidine to tyrosine (11). Among a collection of 45 NT H. influenzae isolates (20), characterized by multilocus enzyme electrophoresis, comprising electrophoretic types 11 to 13, 26 to 27, 29 to 32, 35 to 45, 49 to 51, 53 to 55, 57 to 61, and 63 to 76, neither the sodC gene nor the CuZnSOD protein was detected (16). The only NT H. influenzae strains that have been shown to produce CuZnSOD are the cryptic genospecies biotype IV strains responsible for urogenital and neonatal infections (16). The cryptic biotype IV strains are genetically distinct, although they are related to H. influenzae and H. haemolyticus (21-23); but they have not been isolated from the respiratory tract. The cryptic genospecies biotype IV strains can be distinguished from H. haemolyticus by a positive ornithine decarboxylase reaction. In a previous study we described the presence of a sodC gene and active CuZnSOD in a single strain of H. haemolyticus (14). We now hypothesize that the presence of the sodC gene and/or CuZnSOD could differentiate H. haemolyticus from NT H. influenzae isolates obtained from the respiratory tract.
Initially, a collection of NT H. influenzae (n = 10) and H. haemolyticus (n = 10) isolates that formed part of the study describing the misclassification of the two bacteria (19) was investigated. The Haemophilus isolates used in phase 1 of this study, which, for clarity, are given a short name in parentheses, were 1P26H (TIM1), 3P5H (TIM2), 3P16H1 (TIM3), 3P18H1 (TIM4), 5P1H (TIM5), 6P18H1 (TIM6), 7P49H1 (TIM7), 11P18H1 (TIM8), 14P41H1 (TIM9), 14P38H1 (TIM10), 24P19H1 (TIM11), 27P25H (TIM12), 48P38H1 (TIM13), 49P5H1 (TIM14), 55P13H1 (TIM15), 58P6H1 (TIM16), 65P13H1 (TIM17), 67P13H1 (TIM18), 70P28H1 (TIM19), and 73P18H1 (TIM20). This phase of the study was blind, with only one of the authors (T.F.M.) knowing the species designation. Strains used as positive controls were H. haemolyticus NCTC 10839, biotype IV cryptic genospecies H. influenzae strain 16N (16), Neisseria meningitidis MC58 (27), and Actinobacillus pleuropneumoniae 4074 ATCC 27088 (12, 13, 26); H. influenzae Rd (7, 11) was used as the negative control. A. pleuropneumoniae 4074 and Haemophilus isolates were grown in brain heart infusion (BHI) broth supplemented with 1 μg/ml NAD and 10 μg/ml hemin at 37°C on an orbital shaker (180 rpm) or on BHI agar plates supplemented with Levinthal's base at 37°C in 5% carbon dioxide (11). N. meningitidis was cultured as described previously (27). Chromosomal DNA was prepared from the plate-grown organisms (17), and standard methods were used for restriction endonuclease digestion and Southern blotting, with washing to 80% stringency (24). PCR amplification of genomic DNA with the degenerate primers 3'UNIVSOD and 5'UNIVSOD, designed to amplify a 300-bp conserved region of prokaryotic sodC genes, was performed as described previously (12), except that an annealing temperature of 44°C was used.
Use of the degenerate oligonucleotide primers and genomic DNA as the template resulted, as expected, in the PCR amplification of a number of fragments for each isolate (Fig. 1). However, for 10 isolates (TIM1, TIM2, TIM4, TIM8, TIM10, TIM12, TIM16, TIM17, TIM19, and TIM20) there was a band of 300 bp, indicative of the presence of a sodC gene. The strong 300-bp product was distinct from a fainter band of 280 bp present in isolates TIM3, TIM5, TIM6, TIM9, TIM14, and TIM15. The positive controls (H. haemolyticus NCTC 10839, H. influenzae 16N, and N. meningitidis MC58) amplified a corresponding 300-bp PCR product, while the negative control (H. influenzae Rd) did not. To confirm the presence or absence of sodC, as indicated by the PCR results, Southern blots were probed with a digoxigenin (DIG)-labeled 509-bp fragment from the sodC gene of H. parainfluenzae 1391 (12). The H. parainfluenzae internal sodC probe was labeled with a PCR DIG probe synthesis kit (Roche) by using oligonucleotides ATGATGAAAATGAAAACT and TGATCTGAGTGGTTATC with plasmid pJSK134 (12) as the DNA template. The H. parainfluenzae sodC probe hybridized to the EcoRI restriction fragment-digested genomic DNA of all the Haemophilus isolates when a 300-bp fragment had been amplified by PCR but not to the DNA of those isolates from which this band was absent (Fig. 2). The sodC probe hybridized to the positive controls (H. haemolyticus NCTC 10839 and H. influenzae biotype IV) but not the negative control (H. influenzae Rd). There were variations in the sizes of the EcoRI fragments that were hybridized by the sodC probe, indicative of genetic heterogeneity between such isolates.
Next we determined whether the CuZnSOD protein was produced by Haemophilus isolates. Whole-cell extracts were prepared as described previously (14, 15), except that the FastPrep FP120 instrument (ThermoElectron) instead of sonication was used to disrupt the cells. The settings used were two runs at setting 6.0 for 40 s each, with each run separated by a 10-min cooling step at 4°C. Prior experiments had determined that this resulted in optimal cell breakage. The protein concentration was measured on an ND-1000 spectrophotometer (Nanodrop Technologies). Whole-cell extracts (12.5 μg) were spotted onto Hybond-C (GE Healthcare) membranes; and their reactivities with mouse monoclonal antibody HD1, which recognizes Haemophilus, A. pleuropneumoniae, and meningococcal CuZnSODs, were determined. The membranes were blocked for 30 min with 5% skim milk in phosphate-buffered saline (PBS) containing 0.3% Tween 20 (PBS-T). Primary antibody (HD1) was used at a 1:500 dilution of hybridoma supernatant (with incubation for 60 min), and secondary antibody (rabbit anti-mouse immunoglobulin-horseradish peroxidase conjugate; Dako) was used at a 1:10,000 dilution for 30 min. PBS-T was used for all washes except the final wash, which was done in PBS alone. The dot blots were developed with ECL-Plus (GE Healthcare), according to the manufacturer's instructions. Again, there was a perfect correlation between those isolates that possessed sodC, as determined by PCR or Southern blotting, and the presence of CuZnSOD, as indicated by the recognition by HD1 (Fig. 3).
Finally, we determined whether isolates possessed active CuZnSOD. This was achieved by separation of whole-cell extracts by isoelectric focusing (IEF) with pH 3 to 10 Ready gels (Bio-Rad) and SOD activity staining, coupled with differential inhibition, as described previously (15). The copper chelator diethyl dithiocarbamic acid (DEDC, 10 mM) was used as the inhibitor of CuZnSOD activity (3, 15). SOD gels demonstrating the presence of CuZnSOD activity in eight isolates (TIM1, TIM2, TIM4, TIM8, TIM10, TIM12, TIM16, and TIM17, also representative of TIM19 and TIM20), an example of a single strain (TIM18) representative of nine other strains (TIM3, TIM5 to TIM7, TIM9, TIM11, and TIM13 to TIM15), plus the positive control A. pleuropneumoniae 4074, are shown in Fig. 4. In the eight isolates containing CuZnSOD activity (TIM1, TIM2, TIM4, TIM8, TIM10, TIM12, TIM16, and TIM17; Fig. 4, lanes 1 to 8, respectively), two achromatic bands (arrowed) that disappear with DEDC are clearly visible. In contrast, the single band of SOD activity seen in TIM18 was unaffected by DEDC treatment. The positive control (A. pleuropneumoniae 4074) showed three bands of activity, two of which were DEDC inhibitable. Active CuZnSOD (DEDC inhibitable) was found only in those Haemophilus isolates that (i) contained a sodC gene, as indicated by the presence of the 300-bp PCR product by using 3'- and 5'UNIVSODs as primers or hybridization of the H. parainfluenzae sodC probe in Southern blotting experiments, and (ii) produced the CuZnSOD protein, as determined by reactivity with monoclonal antibody HD1. The presence of two bands with CuZnSOD activity in the IEF gels is probably due to the fact that H. haemolyticus SodC is dimeric. HD1 does not recognize monomeric CuZnSODs from Salmonella enterica serovar Typhimurium or Escherichia coli. The monoclonal antibody does, however, recognize dimeric SODs, such as that from A. pleuropneumoniae 4074 (8), which also showed two bands on the IEF gel (Fig. 4). In Western blots, HD1 recognized a protein of approximately 50 kDa in all the H. haemolyticus isolates, i.e., TIM1, TIM2, TIM4, TIM8, TIM10, TIM12, TIM16, TIM17, TIM19, and TIM20, when there was reactivity with HD1 (data not shown). Given the presence of the 300-bp band upon PCR with 3'- and 5'UNIVSOD primers, designed to amplify a highly conserved C-terminal region of sodC genes, it is probable that the structure of the full-length protein of H. haemolyticus is similar to those established for other prokaryotic CuZnSODs. Typically, the molecular mass of the predicted encoded proteins for known monomeric CuZnSODs or a single subunit of dimeric SODs is 17 to 25 kDa (2, 15). Thus, the available evidence suggests that the H. haemolyticus CuZnSOD is dimeric, although this awaits definitive confirmation.
these results with blinded isolates, isolates TIM1, TIM2, TIM4, TIM8, TIM10, TIM12, TIM16, TIM17, TIM19, and TIM20 were predicted to be H. haemolyticus and the remainder were predicted to be NT H. influenzae. When the code was broken, all of the predictions were found to be correct. To confirm the usefulness of sodC/CuZnSOD as a means of distinguishing between H. haemolyticus and NT H. influenzae, a further 20 strains were analyzed in a blinded manner. These were strains 1P67H1 (TIM21), 3P60H1 (TIM22), 5P54H1 (TIM23), 6P23H3 (TIM24), 7P24H1 (TIM25), 12P56H1 (TIM26), 13P13H5 (TIM27), 18P25H1 (TIM28), 19P49H1 (TIM29), 24P21H1 (TIM30), 26P1H1 (TIM31), 28P15H2 (TIM32), 31P13H1 (TIM33), 32P3H1 (TIM34), 34P4H6 (TIM35), 39P8H1 (TIM36), 43P2H1 (TIM37), 45P9H1 (TIM38), 47P68H1 (TIM39), and 50P5H1 (TIM40). In view of the 100% correlation between the results of PCR, Southern blotting, dot blotting, and CuZnSOD activity in the initial phase, only PCR and dot blotting with monoclonal antibody HD1 were performed. Isolates TIM21, TIM22, TIM24, TIM25, TIM27, TIM28, TIM30, TIM32, TIM34, and TIM36 were predicted to be H. haemolyticus, and the remainder were predicted to be NT H. influenzae. When the code was broken, 100% predictive accuracy was again achieved.
We conclude that the presence of a sodC gene or CuZnSOD protein can reliably differentiate between NT H. influenzae and H. haemolyticus strains isolated from the respiratory tract. A test based on the CuZnSOD activity of single bacterial colonies on agar plates or in broth based on a chromogenic reaction would be ideal; but this is challenging, as current methods do not distinguish between the three major bacterial forms of SOD (MnSOD, FeSOD, and CuZnSOD) and a reliable inhibitor for prokaryotic MnSODs is lacking. Nevertheless, while we await appropriate developments for simple rapid tests in the field of SOD assays, the procedures described herein can clearly distinguish H. haemolyticus from NT H. influenzae. A test that combines monoclonal antibodies HD1 (which recognized the CuZnSOD of H. haemolyticus but not that of NT H. influenzae) and 7F3 (which recognized the P6 protein of NT H. influenzae but not that of H. haemolyticus) (19) has some merit. Alternatively, a multiplex PCR based on H. haemolyticus sodC and a gene unique to NT H. influenzae might be developed. It should be noted that in a previous study representative strains of H. parahaemolyticus and H. paraphrohaemolyticus produced active CuZnSOD (15). Therefore, the presence of CuZnSOD cannot be used to distinguish H. parahaemolyticus or H. paraphrohaemolyticus from H. haemolyticus. We speculate that the CuZnSODs produced by both H. parahaemolyticus and H. paraphrohaemolyticus would be recognized by monoclonal antibody HD1, although this remains to be confirmed.
The finding that H. haemolyticus, a relatively newly recognized commensal of the lower respiratory tract, produces active CuZnSOD, while NT H. influenzae (which is associated with new exacerbations in COPD patients) does not suggests that SodC is not an important virulence factor, at least in the context of the lower respiratory tract. It is not known whether H. haemolyticus produces CuZnSOD in vivo, and its role in host interactive biology, if any, remains to be determined.
In summary we have described a marker (sodC and/or CuZnSOD) that differentiates H. haemolyticus from NT H. influenzae. While the methods used in the present study would be too time-consuming to become part of a routine laboratory protocol, they pave the way forward for the development of a simple rapid test.
ACKNOWLEDGMENTS
This work was supported by awards from The Wellcome Trust and BBSRC (to P.R.L. and J.S.K.) and by NIH grant AI 19641 and the U.S. Department of Veteran Affairs (to T.F.M.).
FOOTNOTES
Corresponding author. Mailing address: Department of Paediatrics, Division of Medicine, Imperial College, St Mary's Campus, London W2 1PG, United Kingdom. Phone: 44 (0)2075943359. Fax: 44 (0)2075943984. E-mail: p.langford@imperial.ac.uk.
Published ahead of print on 20 September 2006.
W.W.M.F. and C.A.O. contributed equally to the work.
REFERENCES
Ammendola, S., M. Ajello, P. Pasquali, J. S. Kroll, P. R. Langford, G. Rotilio, P. Valenti, and A. Battistoni. 2005. Differential contribution of sodC1 and sodC2 to intracellular survival and pathogenicity of Salmonella enterica serovar Choleraesuis. Microbes Infect. 7:698-707.
Battistoni, A. 2003. Role of prokaryotic Cu,Zn superoxide dismutase in pathogenesis. Biochem. Soc. Trans. 31:1326-1329.
Benov, L. T., and I. Fridovich. 1994. Escherichia coli expresses a copper- and zinc-containing superoxide dismutase. J. Biol. Chem. 269:25310-25314.
Collins, C. H., P. M. Lyne, J. M. Grange, and J. O. Falkingham. 2004. Collins and Lyne's microbiological methods, 8th ed. Arnold, London, United Kingdom.
Eldika, N., and S. Sethi. 2006. Role of nontypeable Haemophilus influenzae in exacerbations and progression of chronic obstructive pulmonary disease. Curr. Opin. Pulm. Med. 12:118-124.
Farrant, J. L., A. Sansone, J. R. Canvin, M. J. Pallen, P. R. Langford, T. S. Wallis, G. Dougan, and J. S. Kroll. 1997. Bacterial copper- and zinc-cofactored superoxide dismutase contributes to the pathogenesis of systemic salmonellosis. Mol. Microbiol. 25:785-796.
Fleischmann, R. D., M. D. Adams, O. White, R. A. Clayton, E. F. Kirkness, A. R. Kerlavage, C. J. Bult, J. F. Tomb, B. A. Dougherty, J. M. Merrick, et al. 1995. Whole-genome random sequencing and assembly of Haemophilus influenzae Rd. Science 269:496-512.
Forest, K. T., P. R. Langford, J. S. Kroll, and E. D. Getzoff. 2000. Cu,Zn superoxide dismutase structure from a microbial pathogen establishes a class with a conserved dimer interface. J. Mol. Biol. 296:145-153.
Holt, J. G., N. R. Krieg, P. H. A. Sneath, H. Staley, and S. T. Williams. 1994. Bergey's manual of determinative bacteriology, 9th ed. The Williams & Wilkins Co., Baltimore, Md.
Kilian, M. 2003. Haemophilus, p. 623-635. In P. R. Murray, E. J. Baron, J. H. Jorgensen, M. A. Pfaller, and R. H. Yolken (ed.), Manual of clinical microbiology, 8th ed. ASM Press, Washington, D.C.
Kroll, J. S., P. R. Langford, and B. M. Loynds. 1991. Copper-zinc superoxide dismutase of Haemophilus influenzae and H. parainfluenzae. J. Bacteriol. 173:7449-7457.
Kroll, J. S., P. R. Langford, K. E. Wilks, and A. D. Keil. 1995. Bacterial [Cu, Zn]-superoxide dismutase: phylogenetically distinct from the eukaryotic enzyme, and not so rare after all! Microbiology 141:2271-2279.
Langford, P. R., B. M. Loynds, and J. S. Kroll. 1996. Cloning and molecular characterization of Cu,Zn superoxide dismutase from Actinobacillus pleuropneumoniae. Infect. Immun. 64:5035-5041.
Langford, P. R., B. M. Loynds, and J. S. Kroll. 1992. Copper-zinc superoxide dismutase in Haemophilus species. J. Gen. Microbiol. 138:517-522.
Langford, P. R., A. Sansone, P. Valenti, A. Battistoni, and J. S. Kroll. 2002. Bacterial superoxide dismutase and virulence. Methods Enzymol. 349:155-166.
Langford, P. R., B. J. Sheehan, T. Shaikh, and J. S. Kroll. 2002. Active copper- and zinc-containing superoxide dismutase in the cryptic genospecies of Haemophilus causing urogenital and neonatal infections discriminates them from Haemophilus influenzae sensu stricto. J. Clin. Microbiol. 40:268-270.
Moxon, E. R., R. A. Deich, and C. Connelly. 1984. Cloning of chromosomal DNA from Haemophilus influenzae. Its use for studying the expression of type b capsule and virulence. J. Clin. Investig. 73:298-306.
Murphy, T. F. 2006. The role of bacteria in airway inflammation in exacerbations of chronic obstructive pulmonary disease. Curr. Opin. Infect. Dis. 19:225-230.
Murphy, T. F., A. L. Brauer, S. Sethi, M. Kilian, X. Cai, and A. J. Lesse. Haemophilus haemolyticus: a human respiratory tract commensal to be distinguished from Haemophilus influenzae. J. Infect. Dis., in press.
Musser, J. M., S. J. Barenkamp, D. M. Granoff, and R. K. Selander. 1986. Genetic relationships of serologically nontypable and serotype b strains of Haemophilus influenzae. Infect. Immun. 52:183-191.
Quentin, R., A. Goudeau, R. J. Wallace, Jr., A. L. Smith, R. K. Selander, and J. M. Musser. 1990. Urogenital, maternal and neonatal isolates of Haemophilus influenzae: identification of unusually virulent serologically non-typable clone families and evidence for a new Haemophilus species. J. Gen. Microbiol. 136:1203-1209.
Quentin, R., C. Martin, J. M. Musser, N. Pasquier-Picard, and A. Goudeau. 1993. Genetic characterization of a cryptic genospecies of Haemophilus causing urogenital and neonatal infections. J. Clin. Microbiol. 31:1111-1116.
Quentin, R., J. M. Musser, M. Mellouet, P. Y. Sizaret, R. K. Selander, and A. Goudeau. 1989. Typing of urogenital, maternal, and neonatal isolates of Haemophilus influenzae and Haemophilus parainfluenzae in correlation with clinical source of isolation and evidence for a genital specificity of H. influenzae biotype IV. J. Clin. Microbiol. 27:2286-2294.
Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Sethi, S., N. Evans, B. J. Grant, and T. F. Murphy. 2002. New strains of bacteria and exacerbations of chronic obstructive pulmonary disease. N. Engl. J. Med. 347:465-471.
Sheehan, B. J., P. R. Langford, A. N. Rycroft, and J. S. Kroll. 2000. [Cu, Zn]-superoxide dismutase mutants of the swine pathogen Actinobacillus pleuropneumoniae are unattenuated in infections of the natural host. Infect. Immun. 68:4778-4781.
Wilks, K. E., K. L. Dunn, J. L. Farrant, K. M. Reddin, A. R. Gorringe, P. R. Langford, and J. S. Kroll. 1998. Periplasmic superoxide dismutase in meningococcal pathogenicity. Infect. Immun. 66:213-217.(W. W. Mary Fung, Clíona A. O'Dwyer, Suni)