![[Feedback]](/icons/banner/feedbackACT.gif)
![[For Subscribers]](/icons/banner/subscriberACT.gif)
![[Archive]](/icons/banner/archiveACT.gif)
![[Search]](/icons/banner/searchACT.gif)
![[Contents]](/icons/banner/tocACT.gif)
Previous Article | Next Article 
Applied and Environmental Microbiology, December 2006, p. 7455-7459, Vol. 72, No. 12
0099-2240/06/$08.00+0 doi:10.1128/AEM.00761-06
Copyright © 2006, American Society for Microbiology. All Rights Reserved.
New Rapid and Simple Methods for Detection of Bacteria and Determination of Their Antibiotic Susceptibility by Using Phage Mutants
Nirit Ulitzur1 and
Shimon Ulitzur2*
CheckLight Ltd., Qiryat Tiv'on 36000, Israel,1
Department of Food Engineering and Biotechnology, The Technion Institute, Haifa 32000, Israel2
Received 1 April 2006/
Accepted 17 September 2006
 |
ABSTRACT
|
|---|
Three new methods applying a novel approach for rapid and simple detection of specific bacteria, based on plaque formation as the end point of the phage lytic cycle, are described. Different procedures were designed to ensure that the resulting plaques were derived only from infected target bacteria ("infectious centers"). (i) A pair of amber mutants that cannot form plaques at concentrations lower than their reversion rate underwent complementation in the tested bacteria; the number of plaques formed was proportional to the concentration of the bacteria that were coinfected by these phage mutants. (ii) UV-irradiated phages were recovered by photoreactivation and/or SOS repair mediated by target bacteria and plated on a recA uvrA bacterial lawn in the dark to avoid recovery of noninfecting phages. (iii) Pairs of temperature-sensitive mutants were allowed to coinfect their target bacteria at the permissive temperature, followed by incubation of the plates at the restrictive temperature to avoid phage infection of the host cells. This method allowed the omission of centrifuging and washing the infected cells. Only phages that recovered by recombination or complementation were able to form plaques. The detection limit was 1 to 10 living Salmonella or Escherichia coli O157 cells after 3 to 5 h. The antibiotic susceptibility of the target bacteria could also be determined in each of these procedures by preincubating the target bacteria with antibiotic prior to phage infection. Bacteria sensitive to the antibiotic lost the ability to form infectious centers.
|
INTRODUCTION
|
|---|
Due to the demand to detect as little as a single bacterium in a sample, determining levels of pathogenic bacteria in food, water, and other consumables requires at least 24 h by standard methods. In clinical microbiology, the main aim is to determine the antibiotic susceptibilities of pathogenic bacteria, a task that also requires at least 24 h.
Phage typing is a well-known microbiological method for bacterial identification, and specific phages have also been applied for rapid detection of bacteria (1, 6). Ulitzur and Kuhn (12) were the first to use recombinant phages for rapid detection of specific bacteria, together with the determination of their antibiotic susceptibilities. lux genes were introduced into the genome of a phage. The phage lacked the intracellular machinery necessary for light production and thus remained dark. Upon infection of a host, the phage genes, including the additional lux genes, were expressed within 1 h of infection, and the host cells became luminescent. Later, other reporter genes (ice nucleation [13] and green fluorescent protein [3]) were also used. Another approach, omitting the necessity to genetically modify phage, was suggested by Stewart et al. (8): specific phages infect the bacteria in question, most of the noninfecting phages are then removed or killed using pomegranate rind extract and ferrous sulfate, and the phage-infected bacteria are detected by their ability to form plaques or to initiate an observable signal in host cells. The detection limit was reported to be around 40 bacteria per ml within 4 h. The relatively low sensitivity stemmed from the facts that absolutely all of the noninfecting phages should be removed in the second stage and that the drastic means required to accomplish that affect the ability of the infected bacteria to support completion of the phage lytic cycle.
Our approach bypassed the need to remove all of the noninfecting phages from the test system by using the target bacteria to repair phage mutants that otherwise were unable to form plaques under the specified conditions. Three different principles were applied for phage recovery. In the first version, the bacteria were infected by a pair of complementary amber mutants that could not form plaques on the bacterial lawn at concentrations lower than their reversion rates. The plaques that did originated only from phages that were recovered either by complementation or recombination between the infected mutants. The second version harnessed the DNA repair system of the infected cell, enabling UV-irradiated mutants to complete their lytic cycles. Finally, the use of a complementing pair of temperature-sensitive (ts) mutants is presented. Using these methods, as few as 2 to 10 Salmonella enterica serovar Typhimurium or Eschericia coli O157 bacteria were detected during a 3- to 5-h procedure. To determine the antibiotic susceptibility of the target bacteria, the drugs were simply added at the first step of either procedure. Inhibition of cell wall, protein, or DNA synthesis led to a dramatic decrease in the number of plaques in comparison with the control. A patent application for the process has been filed (11).
|
MATERIALS AND METHODS
|
|---|
Growth media.
Growth media were as follows: LB medium (10 g tryptone, 5 g yeast extract, and 5 g NaCl in 1 liter distilled water); FT (10 g tryptone, 5 g NaCl, 2 g maltose, and 10 ml 1 M MgSO4 in 1 liter distilled water); solid agar (15 g agar in 1 liter LB or FT); and soft agar (7.5 g agar in 1 liter LB or FT).
Antibiotics.
Chloramphenicol, tetracycline, ampicillin, kanamycin, and vancomycin were purchased from Sigma Chemicals.
Phage and bacterial strains. (i) Phages.
The phages used were Felix-O1 (F-O1), specific for S. enterica serovar Typhimurium (4); AR-1, specific for E. coli O157:H7 (6); and OE, which was isolated from a chicken house and found to form plaques on various E. coli strains.
(ii) Bacterial strains.
E. coli MC4100 (ATCC 35695), S. enterica serovar Typhimurium LT2, S. enterica serovar Typhimurium recA-deficient strain 1970, and E. coli recA12 AB 2462 (10) were used. The last strain was used to avoid the repair of unwashed phages; cells that were infected by two phages were recovered by complementation.
Generation of amber phage mutants.
F-O1 amber mutants N6 and N14 were prepared as described by Kuhn et al. (5) with minor modifications. Mutations were induced by nitrosoguanidine, and the desired mutants were selected by their ability to grow on a sup+ host strain, but not on the wild-type strain. Most of the selected amber mutants exhibited a reversion rate of about 10–5 to 10–6. The phage mutants were propagated on sup+ or sup mutant (supF or supE) host cells to yield a phage stock of 1010 to 1011 PFU per ml. The selected F-O1 amber mutants were complementary to each other and showed much higher titers upon double infection of their host than after a single infection at the same phage concentration. Double infection of excess Salmonella cells by these amber mutants resulted in recovery of most of the coinfected phages (Fig. 1).
View larger version (17K):
[in this window]
[in a new window]
|
FIG. 1. Detection of S. enterica serovar Typhimurium using amber mutant phages. A mixture of two amber mutants, N6 and N14, of the phage F-O1 at concentrations of 5 x 108 PFU/ml each, was incubated with increasing concentrations of S. enterica serovar Typhimurium LT2 cells in 1 ml of LB medium. After 25 min of preincubation, the cultures were washed twice in sterile LB medium to reduce the extracellular noninfecting phage concentration. The final pellet was suspended in 0.5 ml LB containing 1 x 109 PFU of S. enterica serovar Typhimurium LT2 (bacterial lawn) and added to 3.5 ml of molten (45°C) LB soft agar. The mix was vortexed and poured over petri dishes containing sterile LA medium. Plaques were counted after a 3-h incubation at 37°C. No plaques were produced in controls in which bacteria were not present during the preincubation.
|
|
Generation of phage mutants by UV irradiation.
Two milliliters of phage (1010 PFU/ml) in 0.1% Triton X-100-saline (Triton spreads the bacteria evenly over the plate) was placed in an empty sterile petri dish and irradiated for 65 s using a bactericidal UV lamp (65 cm from the plate). The irradiated phage showed a titer of 108 PFU/ml on the E. coli MC4100 wild-type strain that was exposed to sunlight 30 min after infection (104 lux) and a titer of about 104 PFU/ml on E. coli recA12 AB 2462 mutant cells that were incubated in the dark.
Generation of temperature-sensitive mutants.
UV irradiation was used to generate temperature-sensitive AR1 (AR1ts) phage mutants that were selected by their ability to form plaques at 37°C, but not at 42°C. The mutants were propagated on their host at 30°C. AR1 mutants 15 and 16 were chosen as a complementary pair on the basis of their abilities to show much higher titers when coinfecting their host than with a single infection. When bacteria were infected at 42°C and then washed several times and incubated at 37°C, very low phage titers were measured.
Phage-mediated bacterial determination: general procedure.
Cultures of target bacteria grown overnight (16 to 18 h at 37°C) were serially diluted in sterile Eppendorf tubes. A stock mixture of phage mutants at a final concentration of about 2 x 1011 to 4 x 1011 PFU/ml was diluted 200-fold in each tested sample to yield a final phage concentration of 1 x 109 to 2 x109 PFU/ml for each mutant. After 25 to 45 min of incubation at the indicated temperature, the suspension was washed three times with LB medium in a microcentrifuge. The washed pellet of infectious centers was then resuspended with 1 ml of culture of the proper bacterial host (at a concentration of 109 cells per sample), mixed with soft agar at 45°C, and poured on LB agar plates to form a lawn. After the agar solidified, the plates were incubated at 37°C (or 45°C, as indicated) until plaques were visible, normally after 3 to 5 h. Several control sets of the untreated cells were also made to ensure that the bacteria were not contaminated by phages, and the sets were run in duplicate: phage development in one set was avoided by omitting the target bacteria from the preincubation step. Another control set was preincubated for 30 min with kanamycin (35 µg/ml) and tetracycline (25 µg/ml) prior to the addition of the phages. A control that omitted the mutated phages was designed to exclude the possibility that the tested sample contained phages that could form plaques on the bacterial lawn. The number of bacteria in the tested sample was verified by serial dilutions in saline and plating on Luria agar (LA) plates for 24 h at 37°C.
Iron sulfate procedure.
The iron sulfate procedure was first suggested by Stewart et al. (8) and was modified as follows: 20 µl of the bacteriophage (1011 PFU/ml) and 20 µl of the appropriate dilution of bacteria (109 CFU/ml) were placed in a sterile Eppendorf microcentrifuge tube. After 25 min of incubation at room temperature, 0.5 ml of the bacterium-phage suspension was mixed with 0.1 ml of freshly made 50 mM FeSO4 · 7H2O at pH 2.9. After 5 min of incubation at room temperature, 3.5 ml of molten (45°C) LB soft agar was added. The mixture was vortexed and poured over petri dishes containing sterile LA medium. We found that this treatment could successfully replace the centrifugation step and thus further simplify the test procedure.
|
RESULTS AND DISCUSSION
|
|---|
The main challenge in utilizing phage for detecting specific bacteria is the removal of noninfecting phages from the assay mixture. Mutant phages at concentrations below their reversion rates cannot form plaques on their bacterial hosts unless they are repaired by recombination or complementation in the target cells. Based on this fact, three principles were developed for phage recovery. The first method was based on using a pair of complementing phage mutants. When two different mutants of a given phage coinfect their bacterial host, completion of the infection cycle may be achieved in one of the following ways: (i) a reversion event may result in back mutation of the mutated gene in one of the infecting phages; (ii) if the mutations are not in the same gene locus, a recombination event may occur between the two phage DNAs, resulting in restoration of the wild-type genotype in one phage; or (iii) the two phages can complete their infection cycles by complementation, i.e., each phage "provides" the essential gene product missing in the other. While the first event occurs at very low frequencies (
10–7), the second event may yield a reversion rate of about 10–2 to 10–3 under certain conditions (such as recA activation). Complementation, however, is a much more common event and may occur whenever two complementing phages are introduced into the same host cell. The phenotypically rescued single phages will not be able to complete another lytic cycle unless they coinfect the next bacterial host together with another phenotypically rescued phage or if two different phage mutants are recombined to create an infective phage target. On solid medium, where phage and bacterial-cell diffusion is limited, this event occurs repeatedly until a visible plaque is formed. Below, we describe three different methods that apply this principle for rapid detection of bacteria and determination of their antibiotic susceptibilities.
Utilizing amber mutants.
The most suitable and reproducible mutation for this procedure is a nonsense amber mutation. Amber mutations are readily obtained after nitrosoguanidine mutagenesis. F-O1 is a phage that lyses almost all Salmonella strains, and it has been widely used as a diagnostic tool for the genus (1, 5). Most amber mutants have a reversion rate of about 10–6 to 10–5. (One may prepare a double amber mutant of a given phage whose reversion rate is about 10–10 [not shown].) For the purpose of this method, a complementing pair of amber mutants of F-O1 (N14 and N6) was selected. A stock mixture of these mutants, both at a final concentration of about 2 x 1011 to 4 x 1011 PFU/ml, was diluted 200-fold in the tested sample to yield a final phage concentration of 1 x 109 to 2 x 109 PFU/ml (as predetermined on a sup+ host). Using such a high phage concentration ensured that all host cells in the sample (when present in a concentration range of 100 to 107 cells/ml) were infected by at least one phage particle of each mutant during the 25 min of preincubation. At this stage, and well before the recovered phage completed its lytic cycle, the suspension was either washed with LB medium (three spins in a microcentrifuge were enough to reduce the phage concentration below 104 PFU/ml) or treated with FeSO4 solution. The pellet was then resuspended with a culture of the proper bacterial host (at a concentration of 109 cells per sample), mixed with soft agar at 45°C, and poured on LA plates to form a lawn. The number of resulting plaques after 3 to 5 h at 37°C reflected the number of bacteria in the tested sample. When the number of free phages introduced into the top agar was below 2 x 105 to 3 x 105, no plaques were formed. This could be attributed to the low reversion rate of the amber mutants (10–6 to 10–7) and to the low probability that two amber mutants would simultaneously infect a given host. Moreover, it is also very likely that when a given host cell was infected by only one phage mutant, and the infection by the second phage occurred only after one or two cell divisions, no complementation events could have taken place, as the DNA of the first phage was already diluted between the offspring of the infected host cells. Figure 1 depicts the numbers of plaques formed after various concentrations of S. enterica serovar Typhimurium were preincubated with a mixture of N6 and N14. A good correlation was found between the number of bacteria in the tested sample (from 5 cells up to 106 cells) and the number of plaques formed in both test procedures.
Utilizing temperature-sensitive mutants.
To bypass the need to wash the infected cell pellet, one may use ts phage mutants. ts phages are incapable of infecting or lysing the target cells at the restrictive temperature. The preferred mutants for this application are those in which the attachment of the phage to the bacterial host does not occur at the restrictive temperature. Complementation events after two different ts mutant phages coinfect a target cell enable one or both of them to form phage particles with active attachment capabilities. Repeated infection of the bacterial lawn by complementary phages may ultimately result in the appearance of turbid plaques. However, when a successful recombination event occurs between the members of the coinfecting pair, the resulting recombinant phage and its offspring will form clear plaques at the restrictive temperature. The unmodified phages will remain incapable of attaching to the host cell. Figure 2 shows an example in which two complementing temperature-sensitive phage mutants at the restrictive temperature (the E. coli O157:H7-specific phage AR1ts mutants 15 and 16) were mixed with a sample containing different concentrations of target cells. After 35 min of incubation at the permissive temperature (37°C), during which no phage burst occurred, the suspension was mixed with the host cells and soft agar and transferred to the restrictive temperature (42°C) for 3 h. The number of subsequently formed plaques reflected the number of E. coli O157 cells in the tested sample. As little as one E. coli O157 cell could be detected, and a linear correlation was found between a wide range of cell concentrations and plaque numbers. The added advantages of this version are rapidity and simplicity. Here, the excess of noninfective phages was not removed, since no plaques of AR1ts mutants appeared at 42°C, even when 108 PFU of either AR1 mutant 15 or 16 was spread on a lawn.
View larger version (17K):
[in this window]
[in a new window]
|
FIG. 2. Detection of E. coli O157:H7 using the temperature-sensitive phage mutant AR1. E. coli O157:H7 culture in LB medium was diluted in LB to yield a concentration ranging from 102 to 105 PFU/ml and was incubated with 0.1 ml of AR1ts mutants 15 and 16 (each at a final concentration of 2 x 108 PFU/ml) for 35 min at 37°C. The culture was then mixed with 3.5 ml of molten soft agar containing 0.5 ml E. coli O157:H7 (optical density at 600 nm, 0.5) at 50°C and poured over LA plates. The plaques were scored after 3 h of incubation at 42°C.
|
|
Utilizing UV-irradiated mutants.
Another way of differentiating between infecting and noninfecting phage is to harness the host repair system to repair mutations in the phage DNA. Toward that aim, OE phage (E. coli-specific) stock was irradiated using a bactericidal UV lamp to yield 0.01% survival. Under the conditions used, most (>99.99%) phage particles were mutated by the radiation. Samples of E. coli MC4100, at concentrations ranging from 102 to 105 cells/ml, were incubated with the irradiated phage in a covered petri dish exposed to bright sunlight (104 lux) for 30 min to allow photoreactivation, as well as excision and error-prone repair, of the mutated phages in the infected host cells. After two wash cycles with saline by standard microcentrifugation, the final pellets were mixed with a suspension of E. coli recA12 AB 2462 cells, poured over LA plates with soft agar, and incubated in the dark. Since the mutated phages could not be repaired by either the SOS system or photoreactivation of the mutated lawn cells, no plaques were developed on the lawn. In contrast, infected phages that had been repaired by the tested bacteria were able to form plaques. The recombination events that occurred between the mutant phages were very few, since the number of phages in the washed culture was around 103 PFU/ml. Statistically, such a low number could not support double infection of the bacterial lawn, and hence, any recombination events. Figure 3 summarizes such an experiment and shows the correlation between the number of plaques formed after incubation of E. coli MC4100 cells with the OE mutant (1 x 108 PFU/ml). Although the principle of this approach has been confirmed, this test appears to be considerably less sensitive than the other two tests (Fig. 1 and 2).
View larger version (17K):
[in this window]
[in a new window]
|
FIG. 3. Use of UV-irradiated phage to detect E. coli. Two milliliters of phage OE (1010 PFU/ml) in 0.1% Triton X-100-saline were placed in an empty sterile petri dish and irradiated for 65 s, using a bactericidal UV lamp (65 cm from the plate). Samples (0.4 ml) of E. coli MC4100 (grown in LB) at concentrations ranging from 101 to 105 cells/ml were incubated with 0.1 ml of irradiated phage in bright sunlight for 30 min. The tubes were spun for 3 min in a microcentrifuge, and the pellets were washed twice (3-min spin) with 1 ml LB each. The final pellets were resuspended in 0.2 ml of E. coli recA12 AB 2462 (optical density at 600 nm, 1) in 3.5 ml of soft agar and poured over LA plates. The plaques were scored after 4 h at 37°C.
|
|
Determination of antibiotic susceptibility.
The ability to determine the antibiotic susceptibilities of the tested bacteria could be implemented in each of the above-described methods. Toward that aim, each tested antibiotic drug was added at the preincubation step, 30 min before the bacteria were washed by centrifugation. Beta-lactam antibiotics (ampicillin and vancomycin), which affect cell wall synthesis, were expected to affect bacterial integrity and result in cell lysis, while antibiotics that inhibit DNA or protein synthesis (chloramphenicol, tetracycline, and kanamycin) were expected to inhibit phage propagation in the infected bacteria and consequently decrease the number of developed plaques in comparison with the control. Figure 4 shows the number of plaques formed by S. enterica serovar Typhimurium cells after preincubation with the F-O1 amber mutants N6 and N14 in the presence of ampicillin (50 µg/ml), vancomycin (50 µg/ml), kanamycin (30 µg/ml), tetracycline (15 µg/ml), and chloramphenicol (30 µg/ml). The presence of either kanamycin, chloramphenicol, or tetracycline led to a significant decrease in the number of plaques, reflecting the high sensitivity of the bacteria to these antibiotics. The bacteria were less sensitive to ampicillin and showed no sensitivity to vancomycin. This sensitivity profile agreed well with the results obtained in the standard disk test carried out in parallel.
View larger version (17K):
[in this window]
[in a new window]
|
FIG. 4. Determination of antibiotic susceptibility. The susceptibilities of S. enterica serovar Typhimurium LT2 to various antibiotics were examined by the method described in the legend to Fig. 1, in which 103 S. enterica serovar Typhimurium cells/ml were incubated for 30 min with the tested antibiotic drug, followed by the procedure described in the legend to Fig. 1. PFU were counted after 4 h. The susceptibility of the bacteria was measured in parallel using the standard disk assay, in which the zone of inhibition was determined (in millimeters) for each antibiotic drug (ampicillin, 5 mm; vancomycin, 2 mm; kanamycin, 12 mm; tetracycline, 16 mm; chloramphenicol, 16 mm).
|
|
In conclusion, we have presented here a new approach and three simple methods for rapid and specific determination of single cells of pathogenic bacteria, such as E. coli O157:H7 and S. enterica serovar Typhimurium, together with their antibiotic susceptibilities. One could apply this method to any given bacterial genus, as long as there is access to the appropriate natural phage.
|
FOOTNOTES
|
|---|
* Corresponding author. Mailing address: Department of Food Engineering and Biotechnology, The Technion Institute, Haifa 32000, Israel. Phone: 972 4 9930530. Fax: 972 4 9533176. E-mail: moni{at}tx.technion.ac.il. 
Published ahead of print on 22 September 2006.
|
REFERENCES
|
|---|
- Cherry, W. B., B. R. Davis, P. R. Edwards, and R. B. Hogan. 1954. A simple procedure for the identification of the genus Salmonella by means of a specific bacteriophage. J. Lab. Clin. Med. 44:51-55.[Medline]
- Fey, H., E. Burgi, A. Margadant, and E. Boller. 1978. An economic and rapid diagnostic procedure for the detection of Salmonella/Shigella using the polyvalent Salmonella phage O-1. Zentbl. Bakteriol. Orig. A 240:7-15.
- Funatsu, T., T. Taniyama, T. Tajima, H. Tadakuma, and H. Namiki. 2002. Rapid and sensitive detection method of a bacterium by using a GFP reporter phage. Microbiol. Immunol. 46:365-369.[Medline]
- Kallings, L. O. 1967. Sensitivity of various Salmonella strains to Felix 0-1 phage. Acta Pathol. Microbiol. Scand. 70:446-454.[Medline]
- Kuhn, J., M. Suissa, J. Wyse, I. Cohen, I. Weiser, S. Reznick, S. Lubinsky-Mink, G. Stewart, and S. Ulitzur. 2002. Detection of bacteria using foreign DNA: the development of a phage reagent for Salmonella. Int. J. Food Microbiol. 74:229-238.[CrossRef][Medline]
- Ronner, A. B., and D. O. Cliver. 1990. Isolation and characterization of a coliphage specific for Eschericia coli O157:H7. J. Food Protect. 53:944-947.
- Sechter, I., C. B. Gerichter, and D. Cahan. 1975. Method for detecting small numbers of Vibrio cholerae in very polluted substrates. Appl. Microbiol. 29:814-818.[Medline]
- Stewart, G. S. A. B., S. A. A. Jassim, S. P. Denyer, P. Newby, K. Linley, and V. K. Dhir. 1998. The specific and sensitive detection of bacterial pathogens within 4 h using phage amplification. J. Appl. Microbiol. 84:777-783.[CrossRef][Medline]
- Takahashi, T., Y. Nakakita, J. Watari, and K. Shinotsuka. 2000. Application of a bioluminescence method for the beer industry: sensitivity of MicroStar-RMDS for detecting beer-spoilage bacteria. Rapid Microbe Detection System. Biosci. Biotechnol. Biochem. 64:1032-1037.[CrossRef][Medline]
- Templin, A., L. Margossian, and A. J. Clark. 1978. Suppressibility of recA, recB, and recC mutations by nonsense suppressors. J. Bacteriol. 134:590-596.[Abstract/Free Full Text]
- Ulitzur, N. October 1998. Method for the detection of bacteria and a kit therefor. Israel patent application IL126697.
- Ulitzur, S., and J. Kuhn. 1987. Introduction of lux genes into bacteria, a new approach for specific determination of bacteria and their antibiotic susceptibility, p. 463-472. In J. Sclomerich, R. Andreesen, A. Kapp, M. Ernst, and W. G. Woods (ed.), Bioluminescence and chemiluminescence: new perspectives. Wiley Interscience, Bristol, United Kingdom.
- Wolber, P. K., and R. L. Green. 1990. Detection of bacteria by transduction of ice nucleation genes. Trends Biotechnol. 8:276-279.[CrossRef][Medline]
Applied and Environmental Microbiology, December 2006, p. 7455-7459, Vol. 72, No. 12
0099-2240/06/$08.00+0 doi:10.1128/AEM.00761-06
Copyright © 2006, American Society for Microbiology. All Rights Reserved.
Top
Abstract
Introduction
