APPLIED AND ENVIRONMENTAL MICROBIOLOGY, June 1990, p. 1981-1983
Vol. 56, No. 6
0099-2240/90/061981-03$02.00/0 Copyright © 1990, American Society for Microbiology
Stability of Antibiotics under Growth Conditions for Thermophilic Anaerobes RALPH
PETERANDERL,l EMMETT B. SHOTTS,
JR. ,2
AND
JURGEN WIEGELl* Departments of Microbiology' and Medical Microbiology,2 University of Georgia, Athens, Georgia 30602 Received 10 July 1989/Accepted 21 March 1990
It
was
shown that the inhibitory effect of kanamycin and streptomycin in
a
growing culture of Clostridium
thermohydrosulfuricum JW 102 is of limited duration. To screen a large number of antibiotics, their stability during incubation under the growth conditions of thermophilic clostridia was determined at 72 and 50°C by using a 0.2% yeast extract-amended prereduced mineral medium with a pH of 7.3 or 5.0. Half-lives were determined in a modified MIC test with Escherichia coli, Staphylococcus aureus, and Bacillus megaterium as indicator strains. All compounds tested were similar at the two temperatures or more stable at 50 than at 72°C. The half-life (t012) at pH 7.3 and 72°C ranged from 3.3 h (k = 7.26 day-', where k [degradation constant] = 1/t1/2) for ampicillin to no detectable loss of activity for kanamycin, neomycin, and other antibiotics. Apparently some compounds (e.g., lasalocid and neomycin) became more potent during incubation (k > 0). A change to pH 5.0 caused some compounds to become more labile (e.g., kanamycin) and others (e.g., streptomycin) to become more stable than at pH 7.3.
For the development of a genetic system in thermophilic anaerobes, it seems promising to use antibiotic resistance as a genetic marker, similar to the approach taken with many mesophiles. Preliminary experiments have shown, however, that the instability of some antibiotics at elevated temperatures might cause problems. Although the shelf lives of most antibiotics were known, it was in general not known to what degree they are stable as dilute solutions in growth medium at elevated temperatures and low redox potentials. Preliminary results were available for only a few compounds (2). Therefore, we decided to screen the stabilities (i.e., degradation constant [k] and half-life [t1/2], where k = lt1/2) of 19 antibiotics (purchased from Sigma Chemical Co., St. Louis, Mo.) which seemed useful as selective agents for organisms like Clostridium thermohydrosulfuricum (minimum and maximum growth temperatures, 35 and 77°C, respectively) (3). The stabilities were determined at 50 and 72°C (frequently used growth temperatures) and at pHs 7.3 and 5.0 (approximate starting and final pHs of C. thermohydrosulfuricum cultures). The antibiotics were solubilized in prereduced medium (0.5 g of Na2S and 0.5 g of cysteine per liter) suitable for culturing C. thermohydrosulfuricum and similar thermophiles (3). When necessary, ethanol was added to solubilize hydrophobic compounds. The solutions were prepared and filter sterilized in an anaerobic chamber and were then incubated at 50 and 72°C. Samples were taken at 0, 12, 24, 48, and 72 h from solutions incubated at 72°C and at 0, 24, 48, 96, and 120 h from solutions incubated at 50°C. The residual activities of the antibiotics were tested in a modified MIC test. This test was conducted individually for each antibiotic, and samples were taken at the times indicated above; separate tests were used for the indicator organisms Escherichia coli ATCC 25922, Staphylococcus *
aureus ATCC 25923 (1), and Bacillus megaterium UGA 1001 (representative of gram-positive bacteria). The antibiotic samples were serially diluted in MuellerHinton broth (Difco Laboratories, Detroit, Mich.) which had been preinoculated with overnight cultures of each of the strains to a cell density of 2 x 105/ml. Twofold dilutions were made with 96-well microtiter plates (U-shaped wells, presterilized: Fisher Scientific Co. Pittsburgh, Pa). The microtiter plates were incubated at 37°C for 48 h. Thereafter, they were stored at 4°C until the plates from all time points of one experiment could be read. This procedure was necessary to read various levels of inhibition within one experiment consistently, since the activities of some of the antibiotics lead to a gradual inhibition over a wide range of concentrations. Representative samples are shown in Fig. 1. The k values are listed in Table 1 in order of decreasing stability at 72°C and pH 7.3. In this test, antibiotics commonly used for the selection of resistant mutants of mesophiles, such as penicillin G or tetracycline, lost their potency rapidly under the given conditions (t1/2 between 5.6 and 15.2 h), indicating a need for precaution if these antibiotics are to be used for work with thermophilic anaerobes. Neomycin, kanamycin, and lasalocid were sufficiently stable at neutral pH to guarantee suppression of sensitive strains even during prolonged incubation. The relatively small pH change from 7.3 to 5.0 had a strong effect on the stabilities of some compounds: the
stabilities of kanamycin, neomycin, lasalocid, monensin, and erythromycin were lower at pH 5.0 than at pH 7.3, while the reverse was true for streptomycin, polymyxin, bacitracin, and novobiocin (Table 1). The potencies of some compounds (e.g., lasalocid and neomycin at pH 7.3 and trimethoprim and metronidazole at pH 5.0) seemed to increase during incubation leading to a positive k (i.e., with increasing incubation time, dilutions higher than those at time zero inhibited the growth of the test organisms). This indicates either that the materials contained inactive forms which became activated or that breakdown
Corresponding author. 1981
APPL. ENVIRON. MICROBIOL.
NOTES
1982
a
FIG. 1. Loss of antibacterial activity of chloramphenicol and penicillin G under different conditions. It was assumed that the decay followed first-order kinetics, that the degradation products were without effect, and that the activity (I) was proportional to the concentration. Then the decay should follow the following model: I, = IO x 2-'", where k is the slope in the semilogarithmic plot, the degradation constant, and the inverse of t1I2. I is plotted as the log2 of the ratio between activity at time t (I,) and activity at time 0 (IO) versus time (in hours). (A) Loss of activity of chloramphenicol due to incubation at 72°C and effects of pH 7.3 (E1) and pH 5.0 (*). (B) Loss of activity of penicillin G at pH 7.3 and effects of incubation at 72°C (El) and 50°C (*).
1-
00 On
04
0 OP
*j
0-
-1
n
Rsn
M
4n
'n
products of some compounds were more potent than the parent compound. For instance, polymyxin became inactivated at pH 7.3 and 72°C with a t1l2 of about 26 h (k = -0.92) and was nearly stable (k = -0.01) at pH 7.3 and 50°C, but at pH 5.0 and 50°C, its activity doubled in about 30 h. Experiments showed that kanamycin is able to inhibit the growth of C. thermohydrosulfuricum JW 102 (3) in modified reinforced clostridial medium (10 g of tryptone, 3 g of yeast extract, 2.5 g of glucose, and 0.5 g of cysteine per liter plus 10 mM phosphate buffer [pH 7.3, 72°C]) for 25 h at a concentration of 12.5 ,ug/ml (Fig. 2b). That time should in most cases be sufficient for the selection between resistant and susceptible bacteria. In contrast, streptomycin under the same conditions completely lost its antimicrobial activity at a concentration of 16 ,ug/ml within 18 h and is therefore of limited use as a selective agent for slow-growing sporeform-
ti me (h)
b 0 o p-
0
20
40
60
time (h)
80
100
ing thermophiles (Fig. 2a).
The extent of growth suppression, as measured by increase of optical density at 600 nm, was dependent on the concentration of streptomycin. When a concentration of streptomycin from a different batch which inhibited increase in optical density for 20 h was used, viable counts were
TABLE 1. Degradation constants for a variety of antibiotics incubated in anaerobic medium Avg k value
Antibiotic
(SD)W
at:
pH 7.3
pH 5.0
72°C
500C
72°C
500C
+0.57 (0.12) +0.15 (0.05) +0.13 (0.06) +0.05 (0.03) 0 (0)
ND ND ND ND ND
-0.16 (0.02) -0.15 (0.02) -0.37 (3.22) ND -0.55 (0.05)
0 (0) -0.23 (0.02) +0.15 (0.04) +1 (0) -0.06 (0.01)
Chloramphenicol
0 (0) -0.04 (0.03) -0.10 (0.02) -0.31 (0.08) -0.59 (0.06)
ND ND ND -0.75 (0.08) +0.22 (0.04)
+0.17 (0.03) +0.09 (0.01) +0.06 (0.01) -1.56 (0.07) 0.00 (0.02)
ND ND ND -0.75 (0.08) ND
Novobiocin Polymyxin Bacitracin Streptomycin Vancomycin
-0.90 (0.07) -0.92 (0.05) -0.93 (0.09) -1.34 (0.06) -1.50 (0.12)
-0.18 +0.25 -0.08 -0.45 -0.04
-0.04 (0.05) +0.01 (0.04) -0.02 (0.09) 0 (0) -1 (0)
-0.19 (0.03) +0.81 (0.15) ND ND -0.04 (0.01)
Penicillin G Tetracycline Ampicillin
-1.58 (0.12) -4.27 (0.15) -7.26 (0.81)
-0.82 (0.09) -0.30 (0.03) -1.56 (0.17)
Lasalocid Neomycin Monensin
Cycloheximide Kanamycin Trimethoprim Metronidazole Amphotericin Erythromycin
(0.03) (0.05) (0.01) (0.10) (0.01)
-3.33 (0.33) -3.48 (0.07) -3.79 (0.04)
-0.43 (0.04) -0.38 (0.03) -0.41 (0.03)
a Determined after outliers were removed by means of a winsorized t statistic (4). The change of activity over time was determined separately for all possible time intervals in each experiment. Negative values indicate a loss of potency and positive values indicate an increase of potency of the antibiotic during incubation; ND indicates that the compound did not show observable decay at 72°C and was therefore not tested at the same pH at 50°C.
NOTES
VOL. 56, 1990
a
4 ________ 3
C Co
2
0
-
o
%O
PM
0 T
0
b
10
30 20 time (h)
40
. 50
5
1983
checked over time by plating the cells on modified reinforced clostridial medium solidified with 0.8% agar. Following an initial doubling, the viable counts decreased by more than 1 order of magnitude in the first 12 h. After that time, however, the number of CFU started to rise again. This indicated that streptomycin exhibits an incomplete bactericidal effect on C. thermohydrosulfuricum. The results for streptomycin and kanamycin relate well to the effects of these antibiotics in the inhibition experiments with C. thermohydrosulfuricum. However, in an experiment similar to the latter one, penicillin G (data not shown) inhibited growth completely for more than 50 h at the tested concentrations, which ranged from 60 to 1,800 U/ml. This suggests that the mode of action of a specific antibiotic for a given thermophile might overcome the limitations that are due to its thermolability. However, even for such compounds, the half-life values reported here should be considered when experiments are being designed. This is especially true if spore-containing or slowly growing cultures are incubated for prolonged times.
4 0 C %o
0 o C
The research was supported by grant DEFG 09-86 ER 136/4 to J.W. from the Department of Energy. Initial experiments were supported by a grant from the office of the Vice President for Research of the University of Georgia to J.W. and E.B. We thank H. Morgan, University of Waikato, Hamilton, New Zealand, for some preliminary results on this subject (2).
3
2
C p-
1
LITERATURE CITED 0
50 40 20 30 time (h) FIG. 2. Loss of antibacterial activity of kanamy cin and streptomycin, respectively, during incubation at 72°C in reirnforced clostridial medium. C. thermohydrosulfuricum JW 102 Mvas used as the indicator organism. Its growth was measured as inccrease in optical density at 600 nm (OD). The log2 of the ratio of the optical density at times t and zero is plotted. (A) Inhibition by:streptomycin at concentrations of 1 ([3), 2 (*), 4 (O), 8 (O), 16 (O), a]nd 32 (O) ,ug/ml. (B) Inhibition by kanamycin at concentrations of a).8 (El), 1.6 (), 3.2 (U), 6.25 (O), and 12.5 (-) ,ug/ml. 0
10
Thornsberry, C. 1985. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically: approved standard. Publication M7-A. National Committee for Clinical Standards, Villanova, Pa. 2. Wiegel, J. 1986. Methods for isolation and study of anaerobic thermophiles, p. 17-37. In T. Brock (ed.), Thermophiles: general, molecular and applied microbiology. John Wiley & Sons. New York. 3. Wiegel, J., L. G. Ljungdahl, and J. R. Rawson. 1979. Isolation from soil and properties of the extreme thermophile Clostridium thermohydrosulfuricum. J. Bacteriol. 139:800-810. 4. Winer, B. J. 1971. Statistical principles in experimental design, 2nd ed., p. 51-54. McGraw-Hill Book Co., New York. 1.
Vol. 56, No. 6
0099-2240/90/061981-03$02.00/0 Copyright © 1990, American Society for Microbiology
Stability of Antibiotics under Growth Conditions for Thermophilic Anaerobes RALPH
PETERANDERL,l EMMETT B. SHOTTS,
JR. ,2
AND
JURGEN WIEGELl* Departments of Microbiology' and Medical Microbiology,2 University of Georgia, Athens, Georgia 30602 Received 10 July 1989/Accepted 21 March 1990
It
was
shown that the inhibitory effect of kanamycin and streptomycin in
a
growing culture of Clostridium
thermohydrosulfuricum JW 102 is of limited duration. To screen a large number of antibiotics, their stability during incubation under the growth conditions of thermophilic clostridia was determined at 72 and 50°C by using a 0.2% yeast extract-amended prereduced mineral medium with a pH of 7.3 or 5.0. Half-lives were determined in a modified MIC test with Escherichia coli, Staphylococcus aureus, and Bacillus megaterium as indicator strains. All compounds tested were similar at the two temperatures or more stable at 50 than at 72°C. The half-life (t012) at pH 7.3 and 72°C ranged from 3.3 h (k = 7.26 day-', where k [degradation constant] = 1/t1/2) for ampicillin to no detectable loss of activity for kanamycin, neomycin, and other antibiotics. Apparently some compounds (e.g., lasalocid and neomycin) became more potent during incubation (k > 0). A change to pH 5.0 caused some compounds to become more labile (e.g., kanamycin) and others (e.g., streptomycin) to become more stable than at pH 7.3.
For the development of a genetic system in thermophilic anaerobes, it seems promising to use antibiotic resistance as a genetic marker, similar to the approach taken with many mesophiles. Preliminary experiments have shown, however, that the instability of some antibiotics at elevated temperatures might cause problems. Although the shelf lives of most antibiotics were known, it was in general not known to what degree they are stable as dilute solutions in growth medium at elevated temperatures and low redox potentials. Preliminary results were available for only a few compounds (2). Therefore, we decided to screen the stabilities (i.e., degradation constant [k] and half-life [t1/2], where k = lt1/2) of 19 antibiotics (purchased from Sigma Chemical Co., St. Louis, Mo.) which seemed useful as selective agents for organisms like Clostridium thermohydrosulfuricum (minimum and maximum growth temperatures, 35 and 77°C, respectively) (3). The stabilities were determined at 50 and 72°C (frequently used growth temperatures) and at pHs 7.3 and 5.0 (approximate starting and final pHs of C. thermohydrosulfuricum cultures). The antibiotics were solubilized in prereduced medium (0.5 g of Na2S and 0.5 g of cysteine per liter) suitable for culturing C. thermohydrosulfuricum and similar thermophiles (3). When necessary, ethanol was added to solubilize hydrophobic compounds. The solutions were prepared and filter sterilized in an anaerobic chamber and were then incubated at 50 and 72°C. Samples were taken at 0, 12, 24, 48, and 72 h from solutions incubated at 72°C and at 0, 24, 48, 96, and 120 h from solutions incubated at 50°C. The residual activities of the antibiotics were tested in a modified MIC test. This test was conducted individually for each antibiotic, and samples were taken at the times indicated above; separate tests were used for the indicator organisms Escherichia coli ATCC 25922, Staphylococcus *
aureus ATCC 25923 (1), and Bacillus megaterium UGA 1001 (representative of gram-positive bacteria). The antibiotic samples were serially diluted in MuellerHinton broth (Difco Laboratories, Detroit, Mich.) which had been preinoculated with overnight cultures of each of the strains to a cell density of 2 x 105/ml. Twofold dilutions were made with 96-well microtiter plates (U-shaped wells, presterilized: Fisher Scientific Co. Pittsburgh, Pa). The microtiter plates were incubated at 37°C for 48 h. Thereafter, they were stored at 4°C until the plates from all time points of one experiment could be read. This procedure was necessary to read various levels of inhibition within one experiment consistently, since the activities of some of the antibiotics lead to a gradual inhibition over a wide range of concentrations. Representative samples are shown in Fig. 1. The k values are listed in Table 1 in order of decreasing stability at 72°C and pH 7.3. In this test, antibiotics commonly used for the selection of resistant mutants of mesophiles, such as penicillin G or tetracycline, lost their potency rapidly under the given conditions (t1/2 between 5.6 and 15.2 h), indicating a need for precaution if these antibiotics are to be used for work with thermophilic anaerobes. Neomycin, kanamycin, and lasalocid were sufficiently stable at neutral pH to guarantee suppression of sensitive strains even during prolonged incubation. The relatively small pH change from 7.3 to 5.0 had a strong effect on the stabilities of some compounds: the
stabilities of kanamycin, neomycin, lasalocid, monensin, and erythromycin were lower at pH 5.0 than at pH 7.3, while the reverse was true for streptomycin, polymyxin, bacitracin, and novobiocin (Table 1). The potencies of some compounds (e.g., lasalocid and neomycin at pH 7.3 and trimethoprim and metronidazole at pH 5.0) seemed to increase during incubation leading to a positive k (i.e., with increasing incubation time, dilutions higher than those at time zero inhibited the growth of the test organisms). This indicates either that the materials contained inactive forms which became activated or that breakdown
Corresponding author. 1981
APPL. ENVIRON. MICROBIOL.
NOTES
1982
a
FIG. 1. Loss of antibacterial activity of chloramphenicol and penicillin G under different conditions. It was assumed that the decay followed first-order kinetics, that the degradation products were without effect, and that the activity (I) was proportional to the concentration. Then the decay should follow the following model: I, = IO x 2-'", where k is the slope in the semilogarithmic plot, the degradation constant, and the inverse of t1I2. I is plotted as the log2 of the ratio between activity at time t (I,) and activity at time 0 (IO) versus time (in hours). (A) Loss of activity of chloramphenicol due to incubation at 72°C and effects of pH 7.3 (E1) and pH 5.0 (*). (B) Loss of activity of penicillin G at pH 7.3 and effects of incubation at 72°C (El) and 50°C (*).
1-
00 On
04
0 OP
*j
0-
-1
n
Rsn
M
4n
'n
products of some compounds were more potent than the parent compound. For instance, polymyxin became inactivated at pH 7.3 and 72°C with a t1l2 of about 26 h (k = -0.92) and was nearly stable (k = -0.01) at pH 7.3 and 50°C, but at pH 5.0 and 50°C, its activity doubled in about 30 h. Experiments showed that kanamycin is able to inhibit the growth of C. thermohydrosulfuricum JW 102 (3) in modified reinforced clostridial medium (10 g of tryptone, 3 g of yeast extract, 2.5 g of glucose, and 0.5 g of cysteine per liter plus 10 mM phosphate buffer [pH 7.3, 72°C]) for 25 h at a concentration of 12.5 ,ug/ml (Fig. 2b). That time should in most cases be sufficient for the selection between resistant and susceptible bacteria. In contrast, streptomycin under the same conditions completely lost its antimicrobial activity at a concentration of 16 ,ug/ml within 18 h and is therefore of limited use as a selective agent for slow-growing sporeform-
ti me (h)
b 0 o p-
0
20
40
60
time (h)
80
100
ing thermophiles (Fig. 2a).
The extent of growth suppression, as measured by increase of optical density at 600 nm, was dependent on the concentration of streptomycin. When a concentration of streptomycin from a different batch which inhibited increase in optical density for 20 h was used, viable counts were
TABLE 1. Degradation constants for a variety of antibiotics incubated in anaerobic medium Avg k value
Antibiotic
(SD)W
at:
pH 7.3
pH 5.0
72°C
500C
72°C
500C
+0.57 (0.12) +0.15 (0.05) +0.13 (0.06) +0.05 (0.03) 0 (0)
ND ND ND ND ND
-0.16 (0.02) -0.15 (0.02) -0.37 (3.22) ND -0.55 (0.05)
0 (0) -0.23 (0.02) +0.15 (0.04) +1 (0) -0.06 (0.01)
Chloramphenicol
0 (0) -0.04 (0.03) -0.10 (0.02) -0.31 (0.08) -0.59 (0.06)
ND ND ND -0.75 (0.08) +0.22 (0.04)
+0.17 (0.03) +0.09 (0.01) +0.06 (0.01) -1.56 (0.07) 0.00 (0.02)
ND ND ND -0.75 (0.08) ND
Novobiocin Polymyxin Bacitracin Streptomycin Vancomycin
-0.90 (0.07) -0.92 (0.05) -0.93 (0.09) -1.34 (0.06) -1.50 (0.12)
-0.18 +0.25 -0.08 -0.45 -0.04
-0.04 (0.05) +0.01 (0.04) -0.02 (0.09) 0 (0) -1 (0)
-0.19 (0.03) +0.81 (0.15) ND ND -0.04 (0.01)
Penicillin G Tetracycline Ampicillin
-1.58 (0.12) -4.27 (0.15) -7.26 (0.81)
-0.82 (0.09) -0.30 (0.03) -1.56 (0.17)
Lasalocid Neomycin Monensin
Cycloheximide Kanamycin Trimethoprim Metronidazole Amphotericin Erythromycin
(0.03) (0.05) (0.01) (0.10) (0.01)
-3.33 (0.33) -3.48 (0.07) -3.79 (0.04)
-0.43 (0.04) -0.38 (0.03) -0.41 (0.03)
a Determined after outliers were removed by means of a winsorized t statistic (4). The change of activity over time was determined separately for all possible time intervals in each experiment. Negative values indicate a loss of potency and positive values indicate an increase of potency of the antibiotic during incubation; ND indicates that the compound did not show observable decay at 72°C and was therefore not tested at the same pH at 50°C.
NOTES
VOL. 56, 1990
a
4 ________ 3
C Co
2
0
-
o
%O
PM
0 T
0
b
10
30 20 time (h)
40
. 50
5
1983
checked over time by plating the cells on modified reinforced clostridial medium solidified with 0.8% agar. Following an initial doubling, the viable counts decreased by more than 1 order of magnitude in the first 12 h. After that time, however, the number of CFU started to rise again. This indicated that streptomycin exhibits an incomplete bactericidal effect on C. thermohydrosulfuricum. The results for streptomycin and kanamycin relate well to the effects of these antibiotics in the inhibition experiments with C. thermohydrosulfuricum. However, in an experiment similar to the latter one, penicillin G (data not shown) inhibited growth completely for more than 50 h at the tested concentrations, which ranged from 60 to 1,800 U/ml. This suggests that the mode of action of a specific antibiotic for a given thermophile might overcome the limitations that are due to its thermolability. However, even for such compounds, the half-life values reported here should be considered when experiments are being designed. This is especially true if spore-containing or slowly growing cultures are incubated for prolonged times.
4 0 C %o
0 o C
The research was supported by grant DEFG 09-86 ER 136/4 to J.W. from the Department of Energy. Initial experiments were supported by a grant from the office of the Vice President for Research of the University of Georgia to J.W. and E.B. We thank H. Morgan, University of Waikato, Hamilton, New Zealand, for some preliminary results on this subject (2).
3
2
C p-
1
LITERATURE CITED 0
50 40 20 30 time (h) FIG. 2. Loss of antibacterial activity of kanamy cin and streptomycin, respectively, during incubation at 72°C in reirnforced clostridial medium. C. thermohydrosulfuricum JW 102 Mvas used as the indicator organism. Its growth was measured as inccrease in optical density at 600 nm (OD). The log2 of the ratio of the optical density at times t and zero is plotted. (A) Inhibition by:streptomycin at concentrations of 1 ([3), 2 (*), 4 (O), 8 (O), 16 (O), a]nd 32 (O) ,ug/ml. (B) Inhibition by kanamycin at concentrations of a).8 (El), 1.6 (), 3.2 (U), 6.25 (O), and 12.5 (-) ,ug/ml. 0
10
Thornsberry, C. 1985. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically: approved standard. Publication M7-A. National Committee for Clinical Standards, Villanova, Pa. 2. Wiegel, J. 1986. Methods for isolation and study of anaerobic thermophiles, p. 17-37. In T. Brock (ed.), Thermophiles: general, molecular and applied microbiology. John Wiley & Sons. New York. 3. Wiegel, J., L. G. Ljungdahl, and J. R. Rawson. 1979. Isolation from soil and properties of the extreme thermophile Clostridium thermohydrosulfuricum. J. Bacteriol. 139:800-810. 4. Winer, B. J. 1971. Statistical principles in experimental design, 2nd ed., p. 51-54. McGraw-Hill Book Co., New York. 1.
