JOURNAL OF BACTEIUOLOGY, Mar. 1976, p. 1080-1087 Copyright 0 1976 American Society for Microbiology
Vol. 125, No. 3
Printed in U.SA.
Physiological Factors Affecting Transformation of Azotobacter vinelandiil W. J. PAGE* AND H. L. SADOFF Department of Microbiology and Public Health, Michigan State University, East Lansing, Michigan 48824 Received for publication 17 November 1975
Cells of Azotobacter vinelandii (ATCC 12837) can be transformed by exogenous deoxyribonucleic acid towards the end of exponential growth. Transformation occurs at very low frequencies when the deoxyribonucleic acid is purified or when the transformation is carried out in liquid medium. Optimal transformation occurs on plates of Burk nitrogen-free glucose medium containing either high phosphate (10 mM) or low calcium (0 to 0.29 mM) content. Higher levels of calcium are inhibitory, whereas magnesium ions are essential for transformation and growth. Extracellular polymer and capsule are increasingly inhibitory to transformation and are most abundant when the calcium content of the medium is high. Transformation is optimal at pH 7.0 to 7.1 and at 30 C, conditions which also coincide with minimal extracellular polymer production. Nonencapsulated strains are excellent transformation recipients. Glycine-induced pleomorphism reduces the transformation frequency and the degree of inhibition is dependent on the phosphate concentration ofthe medium. Rifampin resistance and shifts from adenine, hypoxanthine, uracil, and nitrogenase auxotrophy to prototrophy can be achieved. Although single marker transfer is always greater than double marker transfer, the data suggest that rifampin resistance is linked to hypoxanthine, adenine and uracil prototrophy at intervals of increasing distance. Rifampin resistance did not appear to be linked to nitrogenase. Azotobacter vinelandii is a gram-negative, nitrogen-fixing soil organism which differentiates into cysts upon transfer of a culture from growth on glucose to 3-hydroxybutyrate (20). Strains blocked at various stages of this process would be useful in studying encystment, but chemically induced mutants, other than those with a nitrogen-fixation deficiency (35), are difficult to isolate (23, 31). The basis for this difficulty is unknown, but the possibility of genetic redundancy exists, since A. uinelandii has an estimated 3 x 10-14 g of deoxyribonucleic acid (DNA) per mononucleate cell, or approximately 10 times that of Escherichia coli (30).
vinelandii using several stable drug resistance and auxotrophic markers and have established that genetic transfer does occur. As nothing was known about the development of competence or environmental factors influencing transformation, we have examined each of the major constituents of the Burk nitrogen-free medium for its effect on transformation. A preliminary report of these results was communicated recently (W. J. Page and H. L. Sadoff, Abstr. Annu. Meet. Am. Soc. Microbiol. 1975, H40, p. 102). Transformation in A. vinelandii shares characteristics with other established bacterial systems, but has some major differGenetic transfer in A. vinelandii would also ences. provide an obvious advantage in the studies of MATERIALS AND METHODS encystment or nitrogen fixation. Interspecific transformation ofAzotobacter and intergeneric Strain and growth conditions. A. vinelandii transformation of Azotobacter and Rhizobium (ATCC 12837) was used throughout these studies. have been reported (33, 34), but no details of cell Strains auxotrophic for uracil (ura-21), adenine competence were presented and the genetic (ade-15), and hypoxanthine (hyp-18), deficient in markers were of poor reliability. A recent at- nitrogen fixation (nif-5), and resistant to rifampin (rifrll3), were derived from the parent strain by Ntempt to transform Azotobacter by published methyl-N'-nitro-N-nitrosoguanidine (NTG) mutaprocedures was unsuccessful (29). and nif (UWI) genesis (1). Prototrophic (UW) We have investigated transformation in A. strains of A. vinelandii OP were obtained from W.
'Journal article no. 7428 from the Michigan Agricultural Experiment Station.
Brill, University of Wisconsin, and have been described previously (35). 1080
VOL. 125, 1976
Prototrophic cells were grown routinely in Burk nitrogen-free medium composed of 0.81 mM MgSO4, 0.58 mM CaSO4, 18 ,uM FeSO4, and 1.0 ,uM NaMoO4 in 5 mM potassium phosphate buffer, pH 7.1, and containing 1% glucose (Burk medium). Burk buffer is Burk medium without glucose. Strains ura-21, ade-15, and hyp-18 were grown on media supplemented with 25 ,ug of the appropriate nucleotide per ml. Growth conditions for nift strains have been described previously (35). Medium for petri plates was solidified with 1.8% agar and all cultures, except where noted, were incubated at 30 C. DNA preparation and transformation procedure. DNA was prepared by lysis of donor cells in 15 mM saline-15 mM sodium citrate buffer, pH 7.0, containing 0.05% sodium dodecyl sulfate, at 60 C for 60 min, and then assayed according to the method of Burton (5). When cool, this crude DNA was used directly in transformation studies (9). On a petri plate containing the appropriate complete medium, approximately 2 x 107 active recipient cells (50 ,ul) were mixed with 1 to 1.5 jsg of crude DNA (5 ILI) over an area of approximately 2 cm2. Each mixture was repeated in triplicate on one petri plate. After 24 h of incubation, the entire resultant area of growth was scraped aseptically from the agar surface into a 5-ml sterile Burk buffer blank. The cell suspension was then diluted as necessary and plated on selective medium to detect transformant cells. In routine experiments, A. vinelandii ATCC 12837 cells were transformed with rifl13 DNA. Transformants were selected on Burk medium containing 10 utg of rifampin per ml and were scored after 4 days of incubatior. at 30 C. Recipient cells also were grown in 50 ml of liquid Burk medium in a 250 ml side-arm flask on a shaking water-bath incubator (New Brunwick Scientific Co., New Brunswick, N. J.). Turbidity (as absorbance at 620 nm) was monitored throughout the growth period. At time intervals, 50-pl cells were sampled and assayed for their transforming ability, but in this case 50 j1A of deoxyribonuclease (DNase, 400 ,ug/ml in sterile Burk buffer) was added after 30 min of incubation. After 24 h of total incubation, the cells were screened for transformants. Each crude DNA preparation was tested for sterility by plating on Burk medium and each recipient strain without added DNA was screened to determine the frequency of spontaneous revertants. Transformation frequency was calculated as the transformant count divided by the total cells plated on selective media. Total cells were estimated directly from plate counts or by comparison of turbidities of suspensions to a standard curve. Each transformation frequency was calculated as the average of three independent transformations. Capsule size and growth determination. Late exponential phase cells were separated from vegetative slime by centrifugation at 10,500 x g for 30 min, 20 C. The soluble slime was decanted from the celis and the capsular material that remained with the cell pellet was solubilized by suspension of cells in 5 mM ethylenediaminetetraacetic acid (27). The solu' bilized capsule and vegetative slime were precipitated with three volumes of cold 95% ethanol and
TRANSFORMATION OF A. VINELANDII
1081
dry weights of the fibrous capsular or slime precipitates and the decapsulated cells were estimated after drying at 90 C for 18 h. All determinations were done in duplicate.
RESULTS Time of maximal competence. As shown in Fig. 1A, the growth curve of A. vinelandii was biphasic, with rapid exponential growth occurring from 2 to 8 h, followed by a slower growth phase from 8 to 21 h, and then a stationary phase. Microscopically, the cells in the first phase were rapidly dividing and motile, whereas cells in the second phase were rounded, nondividing, encapsulated, nonmotile, and filled with phase-bright granules of poly-,8-hydroxybutyrate. Increasing numbers of cysts were formed as the second and stationary phases progressed. Cells were sampled
.8 EE.6
0
(DA 0
.2
U 3
z
_ I 0 Dx (_ 0
0
12 24 18 HOURS GROWTH FIG. 1. Growth (A) and development of competence (B) of A. vinelandii in Burk medium, pH 7.1,
6
30 C. The circled area of Fig. IA corresponds to the time of maximal competence. Donor DNA (1.5 pg) from strain rifl 13 was mixed with recipient cells on petri plates ofgrowth medium. After 30 min, 20 pg of DNase was added to the mixture. After 24 h of incubation, transformants were selected on Burk medium
containing 10 pg ofrifampin per ml. Transformation frequency was calculated as transformants per total number of cells plated on selective medium.
1082
J. BACTERIOL.
PAGE AND SADOFF
throughout this growth cycle and screened for transformation competence with crude rifl13 DNA. As shown in Fig. 1B, A. vinelandii was most competent during a short time late in the exponential phase (area circled in Fig. 1A). Before this time, transformants were not detected. Transformation competence decreased rapidly as cells began the second growth phase and was not detected in stationary phase cultures. Transformation was blocked either by the immediate addition of DNase (20 ,ug) to the cell plus DNA mixture or pretreatment of donor DNA with DNase. Transformation frequencies were very low (0.29 mM resulted in decreased transformation frequencies. Calcium levels of 2.32 mM and 2.90 mM reduced transformation frequencies to 3.1 x 10-7 and 6.6 x 10-8, respectively. Magnesium ions were essential for growth and transformation and increasing their concentration to >0.81 mM increased the transformation frequency. Concentrations greater than 1.62 mM did not enhance transformation further. These results suggested that magnesium was functional in the transformation process whereas calcium was not. The calcium content of the medium determined the capsule mass of the recipient cells
TRANSFORMATION OF A. VINELANDII
VOL. 125, 1976
(late exponential) (Table 2). Growth of cells in excess calcium ion (1.16 mM) resulted in more abundant capsular and total polymer synthesis. Low levels of calcium ion resulted in the production of a small amount of capsule and a large amount of soluble slime polymer. There was an inverse relationship between the amounts of capsular and soluble polymer of cells grown in Burk medium with 0.58 mM and 0.058 mM calcium. The capsule mass from cells grown in Burk medium containing 1.62 mM or 0.81 mM magnesium ion was the same when the calcium ion content ofboth these media was constant. The twofold increase in magnesium ion content, however, resulted in the same stimulation of transformation as a twofold decrease in calcium ion content (Fig. 3) without a corresponding decrease in capsule size (Table 2). The capsule of cells grown in 10 mM phosphate-buffered Burk medium had a 2.5-fold greater mass than the capsule of cells from 5 mM phosphate-buffered medium containing 0.058 mM calcium, and had approximately equal mass as cells from 0.29 mM calcium medium (Tables 1 and 2). The transformation frequency of cells from the 10 mM phosphatebuffered medium, however, was sixfold greater
1083
than in the low calcium media (Table 1 and Fig. 3). This result suggested that phosphate stimulated transformation by means other than simple reduction of capsule mass or chelation of calcium ions. Effect of pH and temperature on transformation frequency. Transformation was optimal at an initial pH 7.0 to 7.1 (Fig. 4A). Transformation occurred at much lower frequencies below pH 7.0 and above pH 7.2. Transformation also exhibited a sharp temperature optimum at 30 C, with frequencies decreasing approximately sixfold at 26 and 33 C (Fig. 4B). Because the plate method of transformation required recipient cell growth, it was important to determine if these sharp pH and temperature optima merely reflected growth optima. As shown in Fig. 5A, cell growth was optimal at pH 7.0, but the effect of pH on total mass and cell mass was not as narrow as the transformation optimum. Similarly, the effect of temperature on total and cell mass was not as limited as the transformation optimum (Fig. 5B). The pH and temperature optima also could not be attributed to minimal capsule mass at pH 7.0 and 30 C (Fig. 5C and D), although capsule mass of the late exponential cells was small under these conditions. The production of total polymer was at a TABLE 2. Effect of calcium and magnesium ions on capsule size Medium compositiona Mg (mM) Ca (mM)
Polymer Capsule
(jtg/mg dry wt) Slime
Total
925.0 874.2 50.8 0.58 0.81 57.2 940.4 883.2 0.81 0.058 912.4 174.9 737.5 0.81 0.29 926.8 809.9 116.9 1.62 0.58 27.7 1552.6 1.16 1524.9 0.81 a Other medium components as in Burk medium.
U z
Lw L
a
w LL x (0
0 0
_
(
0.5
1.0
1.5
20TOID
mM FIG. 3. Effect of magnesium and calcium ion concentration on transformation. The Burk medium contained either 0.58 mM calcium sulfate and variable magnesium sulfate (a) or 0.81 mM magnesium sulfate and variable calcium sulfate (0). The dashed line indicates growth limiting conditions. The transformation procedure was as reported in Fig. 2.
INITIAL
pH
INCUBATION
TEMPERATURE
FIG. 4. The effect of initial pH (A) and incubation temperature (B) on transformation frequency. Burk medium was buffered with 10 mM phosphate. The transformation procedure was as reported in Fig. 2.
1084
PAGE AND SADOFF
J. BACTERIOL.
o0
5 10L
, -I
I,
37
\ C 5 1. U
Vol. 125, No. 3
Printed in U.SA.
Physiological Factors Affecting Transformation of Azotobacter vinelandiil W. J. PAGE* AND H. L. SADOFF Department of Microbiology and Public Health, Michigan State University, East Lansing, Michigan 48824 Received for publication 17 November 1975
Cells of Azotobacter vinelandii (ATCC 12837) can be transformed by exogenous deoxyribonucleic acid towards the end of exponential growth. Transformation occurs at very low frequencies when the deoxyribonucleic acid is purified or when the transformation is carried out in liquid medium. Optimal transformation occurs on plates of Burk nitrogen-free glucose medium containing either high phosphate (10 mM) or low calcium (0 to 0.29 mM) content. Higher levels of calcium are inhibitory, whereas magnesium ions are essential for transformation and growth. Extracellular polymer and capsule are increasingly inhibitory to transformation and are most abundant when the calcium content of the medium is high. Transformation is optimal at pH 7.0 to 7.1 and at 30 C, conditions which also coincide with minimal extracellular polymer production. Nonencapsulated strains are excellent transformation recipients. Glycine-induced pleomorphism reduces the transformation frequency and the degree of inhibition is dependent on the phosphate concentration ofthe medium. Rifampin resistance and shifts from adenine, hypoxanthine, uracil, and nitrogenase auxotrophy to prototrophy can be achieved. Although single marker transfer is always greater than double marker transfer, the data suggest that rifampin resistance is linked to hypoxanthine, adenine and uracil prototrophy at intervals of increasing distance. Rifampin resistance did not appear to be linked to nitrogenase. Azotobacter vinelandii is a gram-negative, nitrogen-fixing soil organism which differentiates into cysts upon transfer of a culture from growth on glucose to 3-hydroxybutyrate (20). Strains blocked at various stages of this process would be useful in studying encystment, but chemically induced mutants, other than those with a nitrogen-fixation deficiency (35), are difficult to isolate (23, 31). The basis for this difficulty is unknown, but the possibility of genetic redundancy exists, since A. uinelandii has an estimated 3 x 10-14 g of deoxyribonucleic acid (DNA) per mononucleate cell, or approximately 10 times that of Escherichia coli (30).
vinelandii using several stable drug resistance and auxotrophic markers and have established that genetic transfer does occur. As nothing was known about the development of competence or environmental factors influencing transformation, we have examined each of the major constituents of the Burk nitrogen-free medium for its effect on transformation. A preliminary report of these results was communicated recently (W. J. Page and H. L. Sadoff, Abstr. Annu. Meet. Am. Soc. Microbiol. 1975, H40, p. 102). Transformation in A. vinelandii shares characteristics with other established bacterial systems, but has some major differGenetic transfer in A. vinelandii would also ences. provide an obvious advantage in the studies of MATERIALS AND METHODS encystment or nitrogen fixation. Interspecific transformation ofAzotobacter and intergeneric Strain and growth conditions. A. vinelandii transformation of Azotobacter and Rhizobium (ATCC 12837) was used throughout these studies. have been reported (33, 34), but no details of cell Strains auxotrophic for uracil (ura-21), adenine competence were presented and the genetic (ade-15), and hypoxanthine (hyp-18), deficient in markers were of poor reliability. A recent at- nitrogen fixation (nif-5), and resistant to rifampin (rifrll3), were derived from the parent strain by Ntempt to transform Azotobacter by published methyl-N'-nitro-N-nitrosoguanidine (NTG) mutaprocedures was unsuccessful (29). and nif (UWI) genesis (1). Prototrophic (UW) We have investigated transformation in A. strains of A. vinelandii OP were obtained from W.
'Journal article no. 7428 from the Michigan Agricultural Experiment Station.
Brill, University of Wisconsin, and have been described previously (35). 1080
VOL. 125, 1976
Prototrophic cells were grown routinely in Burk nitrogen-free medium composed of 0.81 mM MgSO4, 0.58 mM CaSO4, 18 ,uM FeSO4, and 1.0 ,uM NaMoO4 in 5 mM potassium phosphate buffer, pH 7.1, and containing 1% glucose (Burk medium). Burk buffer is Burk medium without glucose. Strains ura-21, ade-15, and hyp-18 were grown on media supplemented with 25 ,ug of the appropriate nucleotide per ml. Growth conditions for nift strains have been described previously (35). Medium for petri plates was solidified with 1.8% agar and all cultures, except where noted, were incubated at 30 C. DNA preparation and transformation procedure. DNA was prepared by lysis of donor cells in 15 mM saline-15 mM sodium citrate buffer, pH 7.0, containing 0.05% sodium dodecyl sulfate, at 60 C for 60 min, and then assayed according to the method of Burton (5). When cool, this crude DNA was used directly in transformation studies (9). On a petri plate containing the appropriate complete medium, approximately 2 x 107 active recipient cells (50 ,ul) were mixed with 1 to 1.5 jsg of crude DNA (5 ILI) over an area of approximately 2 cm2. Each mixture was repeated in triplicate on one petri plate. After 24 h of incubation, the entire resultant area of growth was scraped aseptically from the agar surface into a 5-ml sterile Burk buffer blank. The cell suspension was then diluted as necessary and plated on selective medium to detect transformant cells. In routine experiments, A. vinelandii ATCC 12837 cells were transformed with rifl13 DNA. Transformants were selected on Burk medium containing 10 utg of rifampin per ml and were scored after 4 days of incubatior. at 30 C. Recipient cells also were grown in 50 ml of liquid Burk medium in a 250 ml side-arm flask on a shaking water-bath incubator (New Brunwick Scientific Co., New Brunswick, N. J.). Turbidity (as absorbance at 620 nm) was monitored throughout the growth period. At time intervals, 50-pl cells were sampled and assayed for their transforming ability, but in this case 50 j1A of deoxyribonuclease (DNase, 400 ,ug/ml in sterile Burk buffer) was added after 30 min of incubation. After 24 h of total incubation, the cells were screened for transformants. Each crude DNA preparation was tested for sterility by plating on Burk medium and each recipient strain without added DNA was screened to determine the frequency of spontaneous revertants. Transformation frequency was calculated as the transformant count divided by the total cells plated on selective media. Total cells were estimated directly from plate counts or by comparison of turbidities of suspensions to a standard curve. Each transformation frequency was calculated as the average of three independent transformations. Capsule size and growth determination. Late exponential phase cells were separated from vegetative slime by centrifugation at 10,500 x g for 30 min, 20 C. The soluble slime was decanted from the celis and the capsular material that remained with the cell pellet was solubilized by suspension of cells in 5 mM ethylenediaminetetraacetic acid (27). The solu' bilized capsule and vegetative slime were precipitated with three volumes of cold 95% ethanol and
TRANSFORMATION OF A. VINELANDII
1081
dry weights of the fibrous capsular or slime precipitates and the decapsulated cells were estimated after drying at 90 C for 18 h. All determinations were done in duplicate.
RESULTS Time of maximal competence. As shown in Fig. 1A, the growth curve of A. vinelandii was biphasic, with rapid exponential growth occurring from 2 to 8 h, followed by a slower growth phase from 8 to 21 h, and then a stationary phase. Microscopically, the cells in the first phase were rapidly dividing and motile, whereas cells in the second phase were rounded, nondividing, encapsulated, nonmotile, and filled with phase-bright granules of poly-,8-hydroxybutyrate. Increasing numbers of cysts were formed as the second and stationary phases progressed. Cells were sampled
.8 EE.6
0
(DA 0
.2
U 3
z
_ I 0 Dx (_ 0
0
12 24 18 HOURS GROWTH FIG. 1. Growth (A) and development of competence (B) of A. vinelandii in Burk medium, pH 7.1,
6
30 C. The circled area of Fig. IA corresponds to the time of maximal competence. Donor DNA (1.5 pg) from strain rifl 13 was mixed with recipient cells on petri plates ofgrowth medium. After 30 min, 20 pg of DNase was added to the mixture. After 24 h of incubation, transformants were selected on Burk medium
containing 10 pg ofrifampin per ml. Transformation frequency was calculated as transformants per total number of cells plated on selective medium.
1082
J. BACTERIOL.
PAGE AND SADOFF
throughout this growth cycle and screened for transformation competence with crude rifl13 DNA. As shown in Fig. 1B, A. vinelandii was most competent during a short time late in the exponential phase (area circled in Fig. 1A). Before this time, transformants were not detected. Transformation competence decreased rapidly as cells began the second growth phase and was not detected in stationary phase cultures. Transformation was blocked either by the immediate addition of DNase (20 ,ug) to the cell plus DNA mixture or pretreatment of donor DNA with DNase. Transformation frequencies were very low (0.29 mM resulted in decreased transformation frequencies. Calcium levels of 2.32 mM and 2.90 mM reduced transformation frequencies to 3.1 x 10-7 and 6.6 x 10-8, respectively. Magnesium ions were essential for growth and transformation and increasing their concentration to >0.81 mM increased the transformation frequency. Concentrations greater than 1.62 mM did not enhance transformation further. These results suggested that magnesium was functional in the transformation process whereas calcium was not. The calcium content of the medium determined the capsule mass of the recipient cells
TRANSFORMATION OF A. VINELANDII
VOL. 125, 1976
(late exponential) (Table 2). Growth of cells in excess calcium ion (1.16 mM) resulted in more abundant capsular and total polymer synthesis. Low levels of calcium ion resulted in the production of a small amount of capsule and a large amount of soluble slime polymer. There was an inverse relationship between the amounts of capsular and soluble polymer of cells grown in Burk medium with 0.58 mM and 0.058 mM calcium. The capsule mass from cells grown in Burk medium containing 1.62 mM or 0.81 mM magnesium ion was the same when the calcium ion content ofboth these media was constant. The twofold increase in magnesium ion content, however, resulted in the same stimulation of transformation as a twofold decrease in calcium ion content (Fig. 3) without a corresponding decrease in capsule size (Table 2). The capsule of cells grown in 10 mM phosphate-buffered Burk medium had a 2.5-fold greater mass than the capsule of cells from 5 mM phosphate-buffered medium containing 0.058 mM calcium, and had approximately equal mass as cells from 0.29 mM calcium medium (Tables 1 and 2). The transformation frequency of cells from the 10 mM phosphatebuffered medium, however, was sixfold greater
1083
than in the low calcium media (Table 1 and Fig. 3). This result suggested that phosphate stimulated transformation by means other than simple reduction of capsule mass or chelation of calcium ions. Effect of pH and temperature on transformation frequency. Transformation was optimal at an initial pH 7.0 to 7.1 (Fig. 4A). Transformation occurred at much lower frequencies below pH 7.0 and above pH 7.2. Transformation also exhibited a sharp temperature optimum at 30 C, with frequencies decreasing approximately sixfold at 26 and 33 C (Fig. 4B). Because the plate method of transformation required recipient cell growth, it was important to determine if these sharp pH and temperature optima merely reflected growth optima. As shown in Fig. 5A, cell growth was optimal at pH 7.0, but the effect of pH on total mass and cell mass was not as narrow as the transformation optimum. Similarly, the effect of temperature on total and cell mass was not as limited as the transformation optimum (Fig. 5B). The pH and temperature optima also could not be attributed to minimal capsule mass at pH 7.0 and 30 C (Fig. 5C and D), although capsule mass of the late exponential cells was small under these conditions. The production of total polymer was at a TABLE 2. Effect of calcium and magnesium ions on capsule size Medium compositiona Mg (mM) Ca (mM)
Polymer Capsule
(jtg/mg dry wt) Slime
Total
925.0 874.2 50.8 0.58 0.81 57.2 940.4 883.2 0.81 0.058 912.4 174.9 737.5 0.81 0.29 926.8 809.9 116.9 1.62 0.58 27.7 1552.6 1.16 1524.9 0.81 a Other medium components as in Burk medium.
U z
Lw L
a
w LL x (0
0 0
_
(
0.5
1.0
1.5
20TOID
mM FIG. 3. Effect of magnesium and calcium ion concentration on transformation. The Burk medium contained either 0.58 mM calcium sulfate and variable magnesium sulfate (a) or 0.81 mM magnesium sulfate and variable calcium sulfate (0). The dashed line indicates growth limiting conditions. The transformation procedure was as reported in Fig. 2.
INITIAL
pH
INCUBATION
TEMPERATURE
FIG. 4. The effect of initial pH (A) and incubation temperature (B) on transformation frequency. Burk medium was buffered with 10 mM phosphate. The transformation procedure was as reported in Fig. 2.
1084
PAGE AND SADOFF
J. BACTERIOL.
o0
5 10L
, -I
I,
37
\ C 5 1. U
