| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Microbiology, Immunology and Parasitology, School of Dentistry, Louisiana State University Health Sciences Center, New Orleans, LA 70119, USA
2 Department of Endodontics, School of Dentistry, Louisiana State University Health Sciences Center, New Orleans, LA 70119, USA
3 Department of Biochemistry and Molecular Biology, University of Georgia, Athens, GA 30602, USA
4 Department of Biology, University of South Florida, Tampa, FL 33620, USA
5 University of Maryland Biotechnology Institute, Center for Advanced Research in Biotechnology, Rockville, MD 20850, USA
6 Department of Periodontics, School of Dentistry, Louisiana State University Health Sciences Center, New Orleans, LA 70119, USA
Correspondence
John A. Mayo
jmayo{at}uga.edu
Received 6 November 2007
Accepted 5 February 2008
Abbreviations: AAPV, alanyl-alanyl-prolyl-valyl; BODIPY, 4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-propionic acid; DFP, di-isopropylfluorophosphate; E-64, L-trans-epoxysuccinyl-leucylamide-(4-guanidino)-butane; FL, phenylalanyl-leucyl; FLE, phenylalanyl-leucyl-glutamyl; FPA, phenylalanyl-prolyl-alanyl; GGL, glycyl-glycyl-leucyl; HPA, Hide Powder Azure; K, lysyl; L, leucyl; LTR, leucyl-threonyl-arginyl; MMP-1, matrix metalloproteinase 1 (tissue collagenase); OP, o-phenanthrolene; pNA, p-nitroanilide; R, arginyl; TLCK, tosyl lysyl chloromethyl ketone; TPCK, tosyl phenylalanyl chloromethyl ketone; TPPI, tripeptidyl-peptidase I; VLK, valyl-leucyl-lysyl.
Present address: Oral & Maxillofacial Surgery Associates of Eau Claire, 1120 Oak Ridge Drive, Eau Claire, WI 54701, USA. 
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Previously, we reported the isolation under aseptic surgical conditions of Bacillus spp. from periodontal pockets associated with severe marginal periodontitis and from root canals of teeth with necrotic pulps/apical periodontitis (Johnson et al., 1999). Bacillus species isolated from human oral samples have often been regarded as transient microflora (Marsh & Martin, 1992) or as external contaminants due to careless handling of endodontic samples (Dahlen & Moller, 1992). For these reasons, the genus Bacillus has received relatively little attention in oral microbiology. However, Bacillus species as suspected pathogens in dental plaque or oral lesions have been reported in a survey of collagenolytic activity of bacteria in dental plaque clinically associated with gingivitis or periodontitis (Soderling & Paunio, 1981), from cases of apical periodontitis (Molander et al., 1998) and from marginal and apical periodontitis (Helgason et al., 2000). In the latter report it was suggested that virulent clones of Bacillus cereus and Bacillus thuringiensis might play a role in oral disease (Helgason et al., 2000).
It is apparent that Bacillus spp. are emerging opportunistic pathogens in oral disease. Therefore, the purpose of this study was (i) to identify the previously isolated Bacillus strains to the species level, and (ii) to determine whether these strains exhibited extracellular proteolytic activities capable of degrading components of periodontal tissues. Such activities may function as virulence factors in endodontic and periodontal infections.
|
METHODS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Putative Bacillus spp. isolated previously were endospore-forming Gram-positive rods (Johnson et al., 1999). In the present study, Bacillus isolates were identified by confirmation of Gram stain, cellular morphology and endospore formation, and by biochemical and carbohydrate fermentation patterns using API 20E and API 50 CHB kits used according to the manufacturer's instructions (bioMérieux). In addition, the identity of one isolate (BJ0055) was confirmed by sequencing the 16S rRNA gene. Genomic DNA was prepared using a Qiagen DNeasy kit according to the manufacturer's instructions, with lysozyme serving as the lytic agent. This was then used as template for an AccumPrime PFX (Invitrogen) PCR reaction with the primer pair OL398 (5'-TCCTACGGGAGGCAGCAGT-3') and OL399 (5'-GGACTACCAGGGTATCTAATCCTGTT-3'), described originally by Nadkarni et al. (2002), resulting in a product of approximately 550 bp. This fragment was purified by gel extraction using a Qiaquick gel extraction kit (Qiagen), according to the manufacturer's protocol, before being cloned into pCR2.1 using a TOPO TA cloning kit (Invitrogen), creating construct pLES226. The cloned insert was then sequenced (University of Georgia Integrated Biotech Laboratories) using the M13 forward and reverse primers, the target sequences of which are located immediately upstream and downstream, respectively, of the pCR2.1 multiple cloning site. The sequence was subjected to BLAST analysis (www.ncbi.nlm.nih.gov).
Cultures were grown in tryptic soy broth with glucose (TSBG; Difco) or in a semi-defined medium (modified FMC; MFMC) based on that of Terleckyj et al. (1975) and containing (per litre): 20 g D-glucose, 6 g NaC2H3O2.3H2O, 600 mg ammonium sulfate, 10 mg NaCl, 440 mg KH2PO4, 300 mg K2HPO4, 3.15 g Na2HPO4, 2.05 g NaH2PO4.H2O, 225 mg NaC6H5O7.2H2O, 200 mg L-cysteine, 2.5 g Na2CO3, 5.0 g acid-hydrolysed casein, 1.0 g enzymic digest of casein, 1.0 g yeast extract, 10 mg FeSO4.7H2O, 20 mg MgSO4.7H2O, 10 mg MnSO4, 35 mg adenine, 27 mg guanine, and 30 mg uracil. The pH was adjusted to 7.8 and the medium was filter-sterilized. In preliminary experiments, cultures were grown for 24–48 h at 37 °C aerobically with shaking, and in 5–10 % CO2 in air (static) or anaerobically (static).
Detection and preliminary characterization of extracellular proteolytic activities. Growth and the presence of extracellular proteolytic activity were observed in both TSBG and MFMC and under all growth conditions, with the greatest proteolytic activity in shaken aerobic cultures (data not shown). Subsequent experiments were all done with shaken aerobic cultures. Unconcentrated culture supernatants of all strains (grown in TSBG) were screened for the presence of extracellular proteolytic activity essentially as described by Makinen & Makinen (1987), using the general proteinase substrate Hide Powder Azure (HPA; Calbiochem) at 10 mg ml–1 (Rinderknecht et al., 1968). Complete digestion of HPA (1 mg ml–1) yielded an A595 of 0.30 (data not shown). Strain BJ0055 showed the greatest proteolytic activity and was selected for further characterization. Bacillus pumilus BJ0055 was grown in MFMC and culture fluids were collected by centrifugation. The culture fluids were then passed through membrane filters (0.45 µm pore size) and concentrated 300-fold by ultrafiltration (3 kD cut-off).
Concentrated culture fluids were screened for proteolytic activity against a variety of fluorogenic and chromogenic substrates. Fluorescein-labelled type IV collagen and BODIPY (4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-propionic acid)-conjugated casein and elastin were obtained from Molecular Probes and were assayed according to the manufacturer's instructions (www.probes.invitrogen.com) using a Gemini EM microplate fluorometer (Molecular Devices) at 37 °C. Briefly, enzyme and buffer (50 mM Tris/10 mM NaCl, pH 7.9) were pre-incubated in microplate wells for 10 min at 37 °C, after which the reaction was started by the addition of substrate. Data were collected in the kinetics mode. Instrument settings were excitation 495 nm and emission 515 nm for collagen and excitation 485 nm, emission 514 nm and cut-off 495 nm for casein and elastin.
Chromogenic substrates used were amino acid- or peptide-p-nitroanilides (pNAs) obtained from Bachem or Sigma. Substrates with blocked amino-termini included R (arginyl), FL (phenylalanyl-leucyl), FLE (phenylalanyl-leucyl-glutamyl), GGL (glycyl-glycyl-leucyl) and AAPV (alanyl-alanyl-prolyl-valyl). Substrates with free amino-termini included K (lysyl), L (leucyl), FPA (phenylalanyl-prolyl-alanyl), VLK (valyl-leucyl-lysyl), and LTR (leucyl-threonyl-arginyl).
Assays for chromogenic substrates were performed as described for fluorogenic substrates but using a SpectraMax Plus microplate spectrophotometer (Molecular Devices) at 37 °C. Reactions were started by addition of substrate (0.2 mM final concentration), and absorbance at 405 nm (A405) was recorded.
Types of proteolytic activities were assessed by use of class-specific inhibitors (Beynon & Salvesen, 1989). Inhibitors were obtained from Bachem, Boehringer Mannheim or Sigma. The inhibitors and their target proteinase classes are shown in Tables 1
and 2. Assays were run as described for fluorogenic and chromogenic substrates, with enzyme and inhibitor pre-incubated in buffer before addition of substrate. In preliminary experiments it was observed that methanol, the solvent for o-phenanthrolene (OP), by itself inhibited proteolytic activity (data not shown). All OP results were corrected for this effect by use of appropriate controls.
|
|
|
RESULTS AND DISCUSSION |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Bacillus anthracis has long been known to be a major human pathogen (Xu & Frucht, 2007). Other members of the genus are regarded as opportunistic pathogens causing endopthalmitis after penetrating eye injuries, food poisoning and, in compromised patients, bacteraemia, septicaemia, endocarditis, and central nervous system or respiratory infections (Callegan et al., 1999; Ewald et al., 2006; Granum & Lund, 1997; Banerjee et al., 1988; Drobniewski, 1993; Tuazon et al., 1979). B. pumilus can also be a systemic pathogen, causing bacteraemia in neutropenic patients and producing an allergen involved in allergic pneumonitis in machinists (Banerjee et al., 1988; Bernstein et al., 1995; Tuazon et al., 1979). Additionally, B. pumilus has been found in the intestinal flora of humans (Vann et al., 1976) and B. pumilus strains have been shown to produce exotoxins that are cytotoxic for cultured mammalian cells (Brophy & Knoop, 1982; Hoult & Tuxford, 1991). Furthermore, this organism can produce clindamycin-associated enterocolitis in guinea-pigs (Brophy & Knoop, 1982).
Members of the genus Bacillus as oral pathogens have received little attention, having been regarded as contaminants when isolated from endodontic or other oral samples (Dahlen & Moller, 1992; Marsh & Martin, 1992). However, in more recent reports of Bacillus isolates from marginal and periapical sites, carefully controlled clinical and sampling procedures made such contamination unlikely (Johnson et al., 1999; Sunde et al., 2000, 2002). Thus it is probable that such Bacillus isolates are in fact part of the microbiota of the lesion. It has been postulated that the presence of unusual organisms in such lesions suggests an aetiological role (Helgason et al., 2000), and a number of Bacillus isolates from periodontal and endodontic lesions have been reported. These include 12 Bacillus isolates from marginal and periapical periodontitis (Johnson et al., 1999), one Bacillus sp. isolated from a series of endodontic samples (Molander et al., 1998), seven Bacillus spp. including B. pumilus from another series (Sunde et al., 2002), and 20 B. cereus and B. thuringiensis isolates from periodontal and periapical lesions (Helgason et al., 2000). In an earlier study, 15 % of apical periodontitis samples contained B. pumilus, with numbers as high as approximately 0.7 % of the total viable count (Johnson et al., 1999). We propose that members of the genus Bacillus that were isolated carefully from oral disease sites are legitimate isolates and potential oral pathogens.
Detection and characterization of extracellular proteolytic activities
Cell-free culture fluids from all 12 oral strains of B. pumilus exhibited proteolytic activity that degraded the general proteinase substrate HPA. The mean rate was 3.88 mg HPA ml–1 h–1 (range 1.03–5.30) (data not shown). The highest activity was seen with strain BJ0055, which was used for all subsequent experiments.
Extracellular proteolytic activities in concentrated culture fluids of B. pumilus BJ0055 degraded various protein (Table 1
) and peptide (Table 2) substrates. In Table 1, caseinolytic activity was inhibited by di-isopropylfluorophosphate (DFP) and therefore is a serine proteinase, whereas elastinolytic and collagenolytic activities involve chymotrypsin-like serine proteinases that are inhibited by both DFP and chymostatin. This was confirmed for elastinolytic activity, which was also inhibited by tosyl phenylalanyl chloromethyl ketone (TPCK; Table 1); collagenolytic activity was not tested with TPCK. No other classes of caseinolytic, elastinolytic or collagenolytic activity were observed.
Table 2 shows the hydrolysis of various pNA-conjugated peptides that were used as convenient proteinase or peptidase substrates. Activities representing elastase-like (AAPV-pNA-hydrolysing) and chymotrypsin-like (GGL-pNA-hydrolysing) enzymes, glutamyl endopeptidase (FLE-pNA-hydrolysing), and alanyl tripeptidyl peptidase (FPA-pNA-hydrolysing) were observed. No leucine peptidase or trypsin-like activities were detected, as indicated by the failure to hydrolyse the following relevant peptide-pNA substrates: L (leucine aminopeptidase-like); FL (leucine endopeptidase-like); K, VLK, R and LTR (trypsin-like enzymes). The absence of trypsin-like activities was further indicated by the lack of inhibition by tosyl lysyl chloromethyl ketone (TLCK; Tables 1 and 2). In addition, no aminopeptidases other than alanyl tripeptidyl peptidase were detected. No pNA substrates with a free amino-terminus, other than FPA, were hydrolysed.
Glutamyl endopeptidase activity was inhibited by DFP and therefore is a serine proteinase, whereas elastase- and GGL-pNA-hydrolysing activities are chymotrypsin-like serine proteinases that are inhibited by DFP, chymostatin and TPCK. This inhibition by DFP, chymostatin and TPCK was virtually complete for both activities, as was the DFP inhibition of glutamyl endopeptidase. The GGL-pNA-hydrolysing chymotrypsin-like activity may be metal activated or stabilized as it was slightly inhibited by 10 mM OP. No other classes of elastase-like, chymotrypsin-like- or glutamyl endopeptidase activities were observed. In contrast to results for elastase- and chymotrypsin-like activities and glutamyl endopeptidase, inhibition of alanyl tripeptidyl peptidase showed both chymotrypsin-like serine- and metallo-proteinase characteristics with incomplete inhibition by DFP, chymostatin, TPCK or OP.
Gingival crevicular fluid is approximately neutral in pH in good oral health and becomes more alkaline (pH 8.5) as periodontitis develops (Bickel & Cimasoni, 1985). Accordingly, it was of interest to examine the modulation of selected activities by pH. The effects of pH on elastinolytic and alanyl tripeptidyl peptidase activities are shown in Fig. 1. It is important to note that both substrates were stable under all conditions of pH, temperature, time and assay buffer used in these experiments (data not shown). Furthermore, fluorescence of the elastin substrate is insensitive to pH as the BODIPY reporter contains no ionizable groups (Jones et al., 1997). Thus, the results shown in Figs 1 and 2 are not attributable to pH effects on substrates.
|
|
) peaked at pH 9 with substantial activities at pH 8 and 10, consistent with neutral/alkaline proteinase activity. The pH profile for alanyl tripeptidyl peptidase activity (Fig. 1b) was broad, peaking at pH 8 and 9 with substantial activity from pH 7 to 11. This may represent a mixture of activities. Acid proteinase activities were not detected for either substrate. These results suggest that the elastinolytic and alanyl tripeptidyl peptidase activities could be active in situ at the elevated pH of periodontitis lesions, as high as 8.5 (Bickel & Cimasoni, 1985). Activity at pH 11 is unlikely to be pathophysiologically important in periodontitis. However, it probably does represent part of the array of proteinases exhibited by B. pumilus, some of which retain activity even at pH 12 and 13 (Kumar, 2002; Miyaji et al., 2006).
The broad pH activity profile of alanyl tripeptidyl peptidase (FPA-pNA hydrolysing) and its partial inhibition by serine- and metallo-class inhibitors suggested a mixture of activities. This concept was investigated using mixed inhibitor studies at pH 9 and 11. The results for inhibition of alanyl tripeptidyl peptidase activity by mixtures of DFP and either OP or EDTA at pH 9 and 11 are shown in Fig. 2a and b, respectively. Inhibition by mixtures of DFP and either OP or EDTA was greater than that for any of these inhibitors alone, at both pH values. However, 30–40 % of total activity was still resistant even to these combinations of inhibitors.
Bacterial proteinases, lipases and haemolysins are considered to be virulence factors (Drobniewski, 1993; Travis et al., 1995, 1997), and members of the genus Bacillus are noted for the production of extracellular forms of such activities (Gupta et al., 2002; Logan & Berkeley, 1984; Priest, 1985). In view of the proteolytic properties of the B. pumilus strains as described in the present study, proteinases are of particular interest. The potential of proteinases as virulence factors in periodontitis and in systemic disease has been recognized (Travis et al., 1995, 1997; Xu & Frucht, 2007), with sites of action in the host including cytokines, collagen, clotting factors, immunoglobulins, extracellular matrix proteins, cell-surface receptors and disruption of signal transduction.
The B. pumilus strains isolated in this study exhibited an extracellular activity that extensively degraded the general proteinase substrate HPA (data not shown). An extracellular proteinase from an oral strain of B. cereus has been purified and characterized as a true metalloenzyme collagenase similar to that produced by Clostridium histolyticum (Makinen & Makinen, 1987; Soderling & Paunio, 1981). Such activities have been implicated as virulence factors contributing to the initiation and progression of endodontic and periodontal lesions (Johnson et al., 1999; Travis et al., 1995, 1997).
As exemplified by strain BJ0055, oral isolates of B. pumilus also exhibit extracellular proteolytic activities that result in the degradation of a number of protein and peptide substrates. The results shown in Fig. 1 illustrate neutral and alkaline activities degrading elastin and FPA-pNA. Collagen and elastin, as well as the general substrate casein, were digested at neutral pH (Table 1) and collagenolysis and elastinolysis are part of the tissue destruction seen in marginal and periapical periodontitis. In addition to these activities, chymotrypsin-, glutamyl endopeptidase- and alanyl tripeptidyl peptidase-like activities were found (Table 2). While extracellular subtilisin- and glutamyl endopeptidase-like activities are typical of members of the genus Bacillus (Gupta et al., 2002; Kakudo et al., 1992; Leshchinskaya et al., 1997; Priest, 1985), chymotrypsin-like activity in the genus Bacillus has been observed infrequently (Aoyama et al., 2000; Kato et al. 1992). Furthermore, when observed it has been described as subtilisin-like rather than chymotrypsin-like. For example, Aoyama et al. (2000) mischaracterized a soy-milk-coagulating enzyme from B. pumilus as ‘subtilisin-like’ despite its complete inhibition by chymostatin. Similarly, a chymostatin-inhibited serine proteinase from Bacillus subtilis was also described as subtilisin-like (Kato et al., 1992). Chymostatin is not known to be an inhibitor of subtilisin (Ballinger & Wells, 1998), but does target chymotrypsin-like serine proteinases (Beynon & Salvesen, 1989).
Furthermore, alanyl tripeptidyl peptidase-like activity is an unexpected finding. This activity, also known as tripeptidyl-peptidase I (TPPI) (Golabek & Kida, 2006; McDonald, 1998), is brought about by an aminopeptidase that can act on tripeptides with free amino-termini including FPA-pNA (Rawlings et al., 2006). It is found in mammalian neuronal tissue, where enzyme deficiency leads to neurodegenerative disease (Golabek & Kida, 2006). Whereas the mammalian activity is due to an acidic peptidase (Golabek & Kida, 2006; McDonald, 1998), the B. pumilus activity has neutral/alkaline characteristics (Figs 1 and 2). With the exception of the alanyl tripeptidyl peptidase-like activity, the activities reported in Tables 1 and 2 were serine- or chymotrypsin-like serine-class proteolytic activities. Such extracellular proteinases have been observed previously in the genus Bacillus (Gupta et al., 2002; Kato et al., 1992; Leshchinskaya et al., 1997; Priest, 1985), including B. pumilus (Aoyama et al., 2000; Fabian, 1970; Kumar, 2002; Miyaji et al., 2006; Tran-Chau & Urbanek, 1974).
In contrast to these typical Bacillus extracellular proteinases, approximately 40 % of the alanyl tripeptidyl peptidase-like activity was resistant to high concentrations of serine or metallo-proteinase inhibitors, alone (Table 2) or in combination (Fig. 2). Furthermore, this activity was totally resistant to class inhibitors of cysteine and aspartic proteinases and alanyl aminopeptidases (Table 2). Thus, this activity is partially or totally resistant to inhibitors of the four major classes of proteinases and to an inhibitor of a highly relevant group of aminopeptidases.
This is a puzzling result as typical Bacillus extracellular neutral and alkaline proteinases belong to serine or metallo classes (Gupta et al., 2002; Miyaji et al., 2006; Priest, 1985). Mammalian TPPI is also resistant to inhibition by DFP and belongs to the sedolisin family of serine-carboxypeptidases with greatest activity at acid pH (Golabek & Kida, 2006; Kakudo et al., 1992). However, the B. pumilus activity was virtually devoid of activity at pH values below 6 (Fig. 1b). Furthermore, the B. pumilus activity was not inhibited by bestatin, a classical inhibitor of alanyl aminopeptidases (Table 2; Beynon & Salvesen, 1989). Thus, the bestatin–DFP–EDTA–OP-resistant portion of the B. pumilus extracellular alanyl tripeptidyl peptidase activity could represent a novel enzyme, although the total activity probably consists of a mixture of activities with differing pH optima and inhibitor sensitivities. However, it should be noted that all proteolytic activities described here were observed in crude culture fluid concentrates. Proper and complete characterization of these activities will require that the individual enzymes be purified.
To summarize, these results suggest that B. pumilus BJ0055 exhibits at least three extracellular proteolytic activities, a caseinolytic serine proteinase, an elastin-degrading chymotrypsin-like enzyme and an alanyl tripeptidyl peptidase-like activity. Other activities may be related to these. The DFP-sensitive, chymostatin- and TPCK-resistant caseinolytic activity may be subtilisin-like (Ballinger & Wells, 1998; Nonaka et al., 2004), as also may the glutamyl endopeptidase-like activity. Its inhibitor profile is similar to that of the caseinolytic activity, and an analogous enzyme from Bacillus intermedius uses both FLE-pNA and casein as substrates (Leshchinskaya et al., 1997). The chymotrypsin-like elastinolytic activity has inhibitor sensitivities and a pH profile in common with a soybean-milk-coagulating activity from B. pumilus that was described by Aoyama et al. (2000); the AAPV- and GGL-pNA-degrading activities also share this inhibitor sensitivity profile. An alkaline protease purified from B. pumilus by Kumar (2002) had a broad pH profile with highest activity at pH 9–12, attacked di-, tri- and tetra-peptide-pNA substrates, required a free amino-terminus on the peptide, and preferred alanine or phenylalanine in the P1 position of the peptide. These properties are similar to those of the alanyl tripeptidyl peptidase-like activity in the present study.
Possible interactions of proteinases and peptidases from oral bacteria and host in the pathogenesis of periodontitis
Collagen and elastin are major components of the structure of the periodontium (Chavrier, 1990; Crouch & Bornstein, 1978; Rosenbloom et al., 1993), and destruction of periodontal tissue is one of the obvious manifestations of periodontitis. Fig. 3 shows a scheme in which the proteolytic activities of oral bacteria including B. pumilus might act in concert with host proteinases and peptidases to degrade collagen and elastin. The roles of tissue collagenase (matrix metalloproteinase 1; MMP-1) and neutrophil elastase in periodontal disease are well known (Eley & Cox, 1992; Harrington, 1996; Loos & Tjoa, 2005). Furthermore, native collagen and elastase have non-fibrillar or less ordered regions that can be accessible to bacterial collagenolytic and elastinolytic activities that are not true collagenases or elastases (Harrington, 1996). These activities could produce nicks or gaps that might be access points, or larger peptides that might be substrates, for other bacterial activities such as the GGL-pNA-hydrolysing- and glutamyl endopeptidase-like activities observed here. Products of these reactions in turn could serve as substrates for alanyl tripeptidyl peptidase and other bacterial and host peptidases, ultimately reducing collagen and elastin to amino acids and small peptides.
|
|
ACKNOWLEDGEMENTS |
|---|
|
REFERENCES |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |
