You are on page 1of 9

JOURNAL OF BACTERIOLOGY, Sept. 1967, p.

562-570

Copyright 1967 American Society for Microbiology

Vol. 94, No. 3 Printed In U.S.A.

Development of Competence in the Bacillus subtilis Transformation System


KENNETH F. BOTT AND GARY A. WILSON Department of Microbiology, University of Chicago, Chicago, Illinois

Received for publication 7 June 1967

Competence in Bacillus subtilis, assayed by the ability of cells to be transformed with bacterial deoxyribonucleic acid (DNA) or transfected by phage DNA, has been shown to occur in a single semisynthetic medium with peak activity occurring 3 hr after the cessation of logarithmic growth. No step-down conditions or culture manipulations were necessary for routine transfection of 1 % of the population. The results demonstrate that bacteriophage DNA is a valid assay for studying the development of competence in B. subtilis. Predictions of workers using transforming bacterial DNA, who have suggested that competence in B. subtilis is associated with a specific phase of growth, are substantiated. The peak of competence is not affected by marked differences in the rate of growth during the logarithmic phase. The effect on development of competence by this procedure of some components (including casein hydrolysate, tryptophan, and histidine) which were routinely included in the transformation medium by other investigators has been determined by use of infectious phage DNA as an assay. We have demonstrated that tryptophan, as well as histidine, increases the transformation frequency-even in strains which do not have auxotrophic demands for these components. Glutamic acid and alanine depress optimal levels of transfection.
Competence refers to the ability of a bacterial cell to bind irreversibly deoxyribonucleic acid (DNA) of high molecular weight in such a manner that it becomes resistant to deoxyribonuclease; it is the first major step in the transformation process. Later steps involve integration or recombination with the recipient genome, which is followed by replication, segregation, and expression of the new genetic information. Although competence can be subdivided into finer stages (reversible and irreversible), there appears to be no specificity in this step which enables a bacterial cell to discriminate against heterologous DNA (6, 11, 13). It is generally agreed that the ability to develop competence is genetically controlled-possibly by specifying the nature of receptor sites on the cell surface (13, 26). Specific studies of the competence process have been hampered by difficulties in procedure, which require that recombination and expression of the transforming DNA occur before the transformants can be recognized. These are especially critical in evaluating the state of competence in Bacillus subtilis, since only a relatively small percentage of the recipient population (1 to 4%) has ever been transformed for any single marker, even though the whole population has the genetic capability to become competent.
Several workers have noted the appearance of competent cells of B. subtilis near the end of exponential growth in a minimal medium (although not in complex medium) and the inability of some asporogenic mutants to be transformed. They have suggested that this reflects an association with the process of sporulation (17, 20, 24, 25). However, the data are not yet sufficient to establish a quantitative association (13). Spizizen and co-workers (2, 25) described an efficient method of obtaining competent cultures which involves a "step-down" to a growth limit-

ing medium after 4 or 5 hr of normal growth. When competence is induced by this method, one cannot be certain whether it is associated with normal growth or whether it is a type of unbalanced growth that requires "step-down" conditions for its most efficient expression. In this report, we present evidence from studies on the uptake of infectious phage DNA that the appearance of competence can be identified with a specific stage of the growth cycle. In addition, we show that there is little, if any, difference in the development of competence for phage DNA or bacterial DNA, since conditions which alter competence development for the bacterial DNA system have similar effects when competence is assayed by use of infectious phage DNA. The

562

VOL. 94, 1967

COMPETENCE IN B. SUBTILIS

563

level of competence obtained under our conditions of growth is usually 1%, and is reproducible within a twofold range for several of the commonly used recipient strains. Other strains have been transfected at the same stage of growth with similar reproducibility, but at lower frequencies. The results suggest that subsequent modifications which can shorten the doubling time of the logarithmic culture, in the absence of inhibitory components, may cause the period of competence to be shorter by virtue of its ability to "synchronize" the events which occur in stationary phase. Several aspects of the technique are significant. First, this more thoroughly defined regimen for obtaining competence enchances the likelihood that normal strains, and mutants with abnormal growth characteristics, can be optimally transformed (or transfected). Second, it provides assurance that each transformation experiment is being performed when the strain is maximally competent, and it allows unambiguous interpretation of the effects of environmental or metabolic alterations. The use of an infectious phage DNA assay enables one to estimate the competence for uptake of biologically active DNA apparently without the subsequent requirement of integration.
MATERIALS AND METHODS Definitions and abbreviations. "Synchrony" refers to the degree of uniformity with which events occur in the stationary phase, not to simultaneous division of cells, as is the conventional use. "Transfection" is the process by which a cell produces bacteriophage after uptake of DNA isolated from bacteriophage particles. The DNA which brings about this process is referred to as "infectious DNA." "Transfectants" are cells which have been induced to produce phage by this technique. The abbreviations used are as follows: ind, indole (tryptophan); thiy, thymine; met, methionine; ilv, isoleucine; leu, leucine; uvr, the ability to repair

lesions induced by ultraviolet light. A superscript minus following one of these abbreviations refers to the inability of a cell to grow in a synthetic medium lacking this component, or in the case of uvr the inability to repair the damage induced by ultraviolet light. A superscript plus following the abbreviation indicates the ability of a cell to grow in a synthetic medium lacking this component. Bacterial anid phage cultures. B. subtilis strain 168 i,id- thy- was obtained from F. Rothman. Strains 168 uvr-, 168 uvr-thy-, and the 168T+ prototroph for
donor DNA were obtained from B. Strauss. The 168 uvr- thy- is a transformant of 168 int thy- which is defective in its ability to repair the damage induced

by ultraviolet light. The acridine orange induced mutants (166 AO and 26 AO) were derived from 168 ind thy- (4). They are known to be defective in their ability to sporulate, and they grow at a slower rate than is normal on the conventional minimal

salts medium of Spizizen (2). B. subtilis Mu8u5ul (leu- ilv- met-, abbreviated here as Ul) was obtained from N. Sueoka. Bacteriophage 029, its host for production of phage lysates (HW+), and a streptomycin-resistant derivative of it, HW+Sr, were obtained from B. Reilly. Both of these strains require tryptophan for growth. DNA. DNA of bacteriophage +29 was obtained by extracting the purified, concentrated phage (3) twice with phenol-saturated buffer, according to the procedure of Reilly and Spizizen (10). Reagentgrade phenol was distilled before use. Bacterial DNA was extracted from early stationary-phase cultures of 168T+ grown in Difco Antibiotic Assay medium by use of a modification of the method of Saito and Miura (12) in which the phenol was removed by dialysis against several changes of buffer containing 0.1 M phosphate (pH 7.4) and 1 M NaCl. Before use, all preparations of DNA were equilibrated to, and stored in, standard saline citrate (0.15 M NaCl + 0.015 M trisodium citrate, pH 7). Growth of recipient cultures. Cultures were grown by use of a single medium which was identical to the growth medium of Anagnostopoulos and Spizizen (2) except for the addition of extra magnesium to the level which they routinely used in their transformation medium (i.e., 0.072% magnesium instead of 0.01%). Our medium had the following composition: 0.5% glucose; 1.4% K2HPO4; 0.6% KH2PO4; 0.072% anhydrous MgSO4; 0.2% (NH4)2SO4; 0.19%Na citrate- 2H20; 0.02% acid-hydrolyzed casein (Nutritional Biochemicals Corp., Cleveland, Ohio); 50,ug/ml of tryptophan and 50 ,g/ml of histidine; plus 50 ,ug/ml of any other amino acid required by the auxotrophic recipient. The medium was prepared completely (except for magnesium) in glass-distilled water at a 10-fold concentration, filter-sterilized by use of a prewashed 0.45-,u membrane filter (Millipore Corp., Bedford, Mass.), dispensed into sterile tubes in 10-ml amounts, and then frozen until use. Just before use, a tube was thawed, and the contents were added to 85 ml of sterile glass-distilled water in a I-liter Erlenmeyer flask with 5.0 ml of 0.12 M MgSO4. Experiments were begun with an overnight culture (37 C) of the recipient in Antibiotic Assay medium which had been inoculated with a loopful of growth from a Difco Tryptose Blood Agar Base plate. No difference was noted between overnight cultures grown 12, 16, 20, or 24 hr, although 12 to 14 hr was most routinely used. The overnight culture was centrifuged, resuspended in an equal volume of minimal salts solution, and added to the growth medium to give an optical density at 500 m,u of 0.1 (approximately 2 X 107 colony formers/ml) with a Bausch & Lomb Spectronic-20 colorimeter. Usually, 4.5 ml was required. Growth was monitored by following the changes in optical density at 500 m,. As noted by Young and Spizizen (25), plates of the recipient culture which were 3 to 4 days old gave rise to cultures having higher transformation frequencies than freshly subcultured growth. In fact, the best results were obtained with cultures which had been kept at room temperature for 2 to 3 weeks.

564

BOTT AND WILSON

J. BACTERIOL.
.

Transformation. The transformation mixture contained 0.1 ml of DNA (6 to 10 ,ug) and 0.9 ml of competent bacteria in a 150-mm test tube. The mixture was rapidly agitated at a 450 angle on a reciprocal shaker[1Li-inch (2.9-cm) stroke, 450 strokes/ min] at 37 C. After 30 min, 100 ,g (0.1 ml) of deoxyribonuclease was added, and the shaking was continued for 15 min. The necessary dilutions were then prepared with the use of a twice-concentrated solution (21.5 g/liter) of Davis Minimal medium without glucose (Difco) as diluent. Hereafter, this will be referred to as Davis salts. If bacterial DNA (from 168T+) was used as donor, 0.1 ml of dilutions from the transformation mixture was plated on the minimal medium of Spizizen (2), which contained all essential amino acids for growth of the auxotroph, except the one being tested by transformation. When phage DNA was used, the plating was by the soft agar technique of Adams (1) with the use of Tryptose Blood Agar Base plates and the peptone-NaCl softagar overlay of Reilly and Spizizen (10). A 0.1-ml amount of the transfection mixture at the desired dilution was added to 2.5 ml of warm soft agar; then 250 ,ug (0.1 ml) of streptomycin and bacterial strain HW+Sr as an indicator were added. The use of HW+Sr as indicator has several functions. It provides the necessary lawn of indicator at high dilutions of the transfection mixture. It is not transfectable by DNA isolated from bacteriophage 029 (or transformable by bacterial DNA), and, as such, is a double check that any infectious centers must have arisen from phage synthesis in competent cells. The presence of streptomycin inhibits development in the lawn of recipient colonies which might produce a bacteriocin that could be confused with plaques (B. Reilly, personal communication). Controls for both the phage transfection and bacterial transforming activity were performed by incubating the DNA with 100 jg of deoxyribonuclease in the presence of 0.0012 M magnesium sulfate for 15 min before addition of the competent cells. Transformation frequency refers to the number of transformants or transfectants per milliliter times 100, divided by viable cells per milliliter (determined on Tryptose Blood Agar Base medium at the time DNA was added).

fi-

0-%

-J

I-4c

U)

U)

U.

49

RESULTS Under conditions of growth in the medium described, the peak of competence has always been detected at one specific point in the stationary phase of growth. We have used the terminology of Schaeffer et al. (14) to designate the beginning of stationary phase (To) as the point at which the optical density curve departs from the straight line of logarithmic growth. Competence is maximal 3 hr later, T3 (Fig. 1). The appearance of this peak has been verified using as recipients strains Ut, 168 uvr-, 168 ind- thy-, and derivatives of them. Recent experiments suggest that the peak of competence can be shifted away from T3 (in both directions) by changing some of the compo-

HOURS INCUBATION FIG. 1. Transfection and transformation of U1 with DNA from bacteriophage k29 and 168T+. At the indicated intervals after To, 0.9 ml of the bacterial culture (UJ) was withdrawn and added to 0.1 ml of the phage DNA. Another 0.9 ml was mixed with 0.1 ml of DNA from 168T+ and isoleucine+ transformants recovered on minimal medium plus leucine and methionine. A third sample was diluted appropriately and plated on complex medium to determine total cell viability. Transformation procedures were as described in Materials and Methods. The turbidimetric doubling time during the log phase was- 48 min. (Closed circles represent optical density; open circles, isoleucine+ transformants; starred circles, phage DNA transfectants).

nents of the growth medium. These alterations can result in a decreased efficiency of competence. Figure 1 illustrates that maximal competence is attained simultaneously for bacterial DNA (ilv+ transformants) and infectious 429 DNA. This confirms the observations of Reilly and Spizizen (10) that the peak of transfection coincides with that for transformation with the use of a different bacterial marker, and substantiates the usefulness of phage DNA to assay competence. We have consistently observed the peak of transfection to be sharper than that for transformation under our experimental conditions. This difference is not apparent in the data of Reilly and

VOL. 94,7 1967

COMPETENCE IN B. SUBTILIS

565

Spizizen (10). We suggest that it may reflect a discrimination between the two types of DNA. Whether the broadness of the transformation peak can be attributed to size heterogeneity of the bacterial DNA, or to some other difference, has

creasing competence, the effect is not real; subsequent experiments have indicated that the final population size is not significantly affected by this concentration of tryptophan or histidine. Similarly, the growth curves have never varied in a not yet been determined. However, we have noted manner that would suggest that these amino that the number of transfectants is increased by acids were inhibiting growth. The importance of magnesium to a slightly greater extent than the histidine was not unexpected, since Anagnostobacterial transformants (unpublished data). poulos and Spizizen (2) indicated that one of the Whether or not the two observations are related functions of casein hydrolysate was to supply is also unknown. histidine. They proposed that it acted as a chelatExperiments illustrated in Fig. 2 have verified ing agent for divalent cations and noted that its that competence is a transitory stage in the growth effects could be mimicked by a, a-bipyridyl. Since of the population, which always occurs 3 hr after no alternative explanation has been proposed for To and is not dependent upon the initial growth the effects of histidine, we attempted to test the rate of the culture. Part A illustrates that when the theory of chelation more directly. The complete logarithmic growth rate is reduced by some nu- growth medium was extracted with dithizone actrient limitation (in this case, phosphate) the cording to the procedure of v. Hofsten (7). This peak is still at T3. Parts B and C show the develop- extraction method has been shown to remove ment of competence in slow-growing mutants copper and other heavy metal ions to at least a that had been induced by acridine orange (4). level of 0.5 X 10- 6 M. In this experiment, 100 ml The 168 uvr- thyj strain in part D is very poorly of complete medium containing tryphophan and transformable by conventional methods. This histidine, without magnesium, was shaken with graph indicates that one reason is probably its an equal volume of carbon tetrachloride containslower growth rate, since the 4 hr plus 90 min ing 6 mg of dithizone per liter. The extraction regimen would not give it time to reach the proper was repeated five times (no further color change point in the stationary phase. These results are of the solvent layer); then excess solvent was substantiated by experiments of Okubo and evaporated with a stream of filtered air. The Romig (9), who showed that another slow-grow- MgSO4 was prepared as a concentrated solution ing mutant of B. subtilis (MC-1) was made more in deionized water and diluted appropriately into highly competent by conventional methods with the extracted medium. The medium was filtera prolonged growth period. During the experi- sterilized by use of a prewashed membrane filter ment illustrated by part E, the culture growth was of 0.45 ,u porosity. The control consisted of a unusually rapid for reasons not understood; this similar portion of medium which was extracted may have affected the "synchrony" of stationary- the same number of times with carbon tetrachlophase activities in such a manner that more of the ride having no dithizone. Magnesium sulfate cells became competent at the same time. One can from the same solution was then added to this see that the peak of transfection was unusually medium; evaporation of carbon tetrachloride and sharp under these conditions. Despite the varia- sterilization procedures were the same. If ions inhibitory to DNA uptake were being tions in growth rate or the strains used here (Fig. 2), the peak of competence was always detected complexed by histidine in the medium, we would at or near T3. have expected the dithizone treatment to increase Effects of tryptophan and histidine. We have the transfection frequency due to removal of these tested some of the constituents of the growth inhibitors. This effect was not noted, although the medium to define their importance, and to verify growth rate and total amount of growth on the that their effect on competence for phage DNA two media were identical, indicating that growthwas the same as that for bacterial DNA. Table 1 supporting ability had not been altered. The dishows the effects of withholding tryptophan or thizone treatment reduced the number of transhistidine from the growth medium of strain Ul, fectants to 11% of the value obtained in its aban auxotroph which does not require either of sence (i.e., 2.0 X 106 transfectants/ml with CCl4 these amino acids. Deprivation of either of these alone; 2.2 x 105/ml with CC14 plus dithizone). components did not shift the peak of competence Essentially the same results were obtained when away from T3. It can be seen that both tryptophan the medium was extracted with chloroform conand histidine are essential for maximal compe- taining the dithizone. These results may be intence, although a low level of transformation terpreted to mean that the inhibitory component does occur in their absence. Although it appears which is being complexed by histidine is not only from these data that omission of tryptophan and a heavy metal ion, as suggested by Anagnoshistidine stimulates a high cell viability while de- topoulos and Spizizen (2), but some other com-

566

BOTT AND WILSON

J. BACTERIOL.

A
0-5
11
n z

C
*7. -01. I

0*5-

/,
/ 0I

I*S x

0300

f'p

3\

0
x

I.a
z

, z

0*-i

I1-

I.z 4

: z

I-C

D
2 HOURS

,4

6 8 HNCU8ATION

HOURS

INCUBATION

'r
07 .

0-r
1--

z
CM 0

cb 0-3- 4 I-

-i

4c
I.0

0-

0,
0
0

*10e

2
x

007

a~~~~~~0%

5 -, a
3
U.
z 4

HOURS

10

12

INCUBATION

HOURS

INCUBATION

FIG. 2. Transfection of strains which grow at rates different from those of the typical recipient cultures. At each indicatedinterval, 0.1 mlof 429DNA wasaddedto 0.9ml of culture according to procedure described in Materials and Methods (0, transfectants; 0, optical density). (A) Normal Ul recipient growing in a minimal mediwn like that described in Materials and Methods except that the phosphate component was provided by 0.5 g of K2HPO4. Final pH was adjusted to 7.0 with I N HCI. The turbidimetric doublingtime was 72 min. (B) Strain 166 AO growing in the minimal medium described in Materials and Methods. Turbidimetric doubling time, 58.5 min. (C) Strain 26 AO in minimal medium. Turbidimetric doubling time, 72.5 min. (D) Strain 168 uvr, thy, in minimal medium Turbidimetric doubling time, 114 min. (E) Strain Ul (a normal recipient) in minimal medium. For unknown reasons, this culture grew unusually fast. The turbidimetric doubling time of 37 min illustrated here is to be compared with the normal pattern illustrated in Fig. 1.

VOL. 94, 1967


TABLE
1.

COMPETENCE IN B. SUBTILIS

567

from

the

Efrects of witholdintg certain conistituents growth medium of Ul on the transTransfor-

TABLE 2. Effects of glutamic acid oni transfection at the peak of competence


Concn Recipient strain of glutamic
Transfec-

fection at the peak of competence (T3)


Transfectants Nutrient removed /ml from complete mediuma

Viable cells/ml

mation
frequency

tants/ml

Vliable cells/mi

Transforma-

fore

quency

None

8.7 3.3
1.4

X X

105

1.3 3.9 3.3

X 108

.65

U1

mg/ml
........... ...........

Tryptophan
Histidine

105
105

X 108
X

.084 .042

U1

108

Ul ...........

0.05 2

2.9 X 106 4.07 X 108 2.1 X 105 5.49 X 108 7.6 X 103 5.23 X 108

.73 .038 .0015

Histidine and tryptophan .. ............ 2.2 X 104


Casein

3.7 X 108
5.6 X 10'

.006
.32

hydrolysate.....11.8

105

168 uvr ......... 0 168 uvr- (low

9.4 X 105 1.7

X 108

.57
.00038 .00037

The minimal medium described in Materials and Methods was complete except for the total removal of those components indicated in the left column. All values are those observed at Ts, the peak of competence.
a

4.0 X 102 1.1 Mg) ......... 2 . 8.0 X 102 168 uvr-. ........ 2.2 2

X 108 X 108
X 108 X 108

26 AO ......... 0 26 AO .........1 2

8.5

40

104 4.0 4.0

.021 .000010

ponent of the casein hydrolysate, or a by-product of cell growth. Actually, we favor the former alternative because of reports from various laboratories that the particular batch of casein hydrolysate is extremely important in achieving good transformation frequencies. Dithizone treatment might not be expected to permit the same degree of competence in both systems. A metabolic by-product would be more likely to inhibit transfection in our system, since it is not diluted out as it could be if a step-down procedure

a Final concentration of neutralized glutamic acid solution which was added aseptically to minimal growth medium at the beginning of growth. b Final magnesium sulfate concentration was 100 pg/ml instead of the usual 720 ,g/ml.

employed. The requirement of tryptophan for maximal competence of a strain which does not require tryptophan for growth was unexpected and is still not understood, but may reflect a function of tryptophan that has not yet been recognized. Effect of glutamic acid and alanine on transfection. The acridine orange-induced mutants of B. subtilis are restored to the parental growth rate by the presence of 2 mg of glutamic acid per ml (unpublished data). Attempts to transfect these strains first used a medium which had been modified to contain 2 g of glutamic acid per liter. Under these conditions, none of the mutants was ever transfected. Subsequent control experiments with 168 uvr- and Ul showed that normal levels of competence were not produced in the presence of glutamic acid. In fact, 50 ,g/ml was shown to reduce markedly the frequency of transfection. The results of these experiments are shown in Table 2. It does not appear that this inhibition is due to a general stimulation of growth. At 50 ,g/ml, little or no effect on growth rate was detected in the 168 uvr- and Ul recipients. The peak of maximal competence, although lowered, was still at T3. The presence of 2 mg of glutamic acid per ml did not affect the logarithmic growth rate of these recipients (unlike its effect on the acridine orange-induced mutants), but
were

it did slightly extend the logarithmic growth phase. Periodic samples of this culture tested for competence indicated that glutamic acid was indeed inhibitory and did not act to shift the position of the competence peak to a later hour. Alanine was also shown to be inhibitory when, in similar experiments, it was added to the casein hydrolysate medium. It, too, slightly increased the population size, and caused a decreased transfection frequency at the peak of competence (T3). In one experiment of this type, 50 ,ug/ml of added alanine reduced the number of transfectants per ml, with strain Ul, from 1.1 x 106 (in the culture lacking alanine) to 4.5 x 105 in the culture grown in the presence of alanine. These results suggest that several components of casein hydrolysate are acting to keep the level of competence low, whereas other components may be stimulatory.
DISCUSSION These experiments have illustrated that, when B. subtilis is cultured in a single growth medium containing 0.072% magnesium, a peak of competence appears 3 hr after the cessation of logarithmic growth (T3). The increased population size which accompanies growth in a single medium provides a larger number of competent cells than can be obtained by the "step-down" method previously used. The levels of competence

obtained by this procedure are as good as or better than we were able to achieve using the method of Anagnostopoulos and Spizizen (2); results were also more reproducible. A direct comparison of the two methods is difficult, since

568

BOT-r AND WILSON

J. BACTERIOL.

in our hands the 4 hr plus 90 min regimen has been highly variable even when used in duplicate or triplicate experiments on the same day. The frequency of competence with Spizizen's method yielded from 0.05 to 0.5% in the majority of our trials. It does appear that under the conditions described (24) the maximal frequency of competence (1 to 4%) that has been obtained by the method of Spizizen and co-workers makes that method slightly more efficient, but we have never been able to achieve quite that level of competence in many attempts to duplicate their efficiency. Perhaps some slight difference, such as recipient strain, amount of inoculum, the water used to make up the medium, or shaker speed, would be enough to affect the final level of competence. We believe that these variables can now all be tested systematically by use of the method outlined in this paper. Three different recipient strains have been transfected by our procedure at a frequency of approximately 1% (within a factor of 2) for more than 25 transfection experiments. All of the cultures examined were shown to develop a peak at T3. It is of interest that those cultures which grow at a rapid rate [similar to, or faster than, the conventionally used recipients (Fig. 2)] seem to have the sharpest peak of transfection. It is tempting to suggest that these conditions may be providing a better opportunity for a more "synchronous" stationary phase in which a larger percentage of the population would reach competence at the observed peak . A peak of competence is still detected at T3 when the growth rate is reduced because of phosphate limitation or because of ancillary growth requirements (Fig. 2). The appearance of the same pattern of competence, despite an altered growth rate in logarithmic phase, suggests that metabolic activity due to these alterations has little effect on timing of the development of competence. Our studies have illustrated that tryptophan and histidine are both necessary for maximal competence under the conditions described, even in a strain (Ul) which does not require them to satisfy auxotrophic deficiencies. More recent experiments (unpublished data) have revealed that the peak of competence can also be observed at T3 when cells are grown in the absence of casein hydrolysate, tryptophan, and histidine, i.e., a completely synthetic medium. In the absence of all of these supplements, the maximal frequency is roughly 30% of the value obtained in their presence. These conditions also prolong the duration of logarithmic growth. Taken together, these results suggest that competence is indeed associated with a specific aspect of stationary phase; it appears that some

constituents of casein hydrolysate are stimulatory, but under the conditions outlined above they must function in the presence of inhibitory substances which are also present. Certainly, a detailed analysis of the casein hydrolysate component is essential before further experiments can be effectively evaluated. Another observation from the current studies is that the addition of some components does alter the overall growth rate (as reflected by turbidimetric doubling time), but not competence. These results suggest that experiments of some workers who have noted stimulatory or inhibitory effects of certain constituents on competence development may in fact have served only to alter the growth rate sufficiently to make the routine 4 hr plus 90 min regimen slightly out of phase with the peak of competence. That is, an evaluation based on transformation at a single point may have resulted in submaximal levels and thus introduced some inaccuracy into their interpretations. Special note should be made of the observations that glutamic acid and alanine inhibit transfection. This is consistent with the following. (i) No reports have appeared in the literature describing transformation of an alanine-less or glutamicless locus; furthermore, Spizizen et al. (18) reported that a group of their mutants requiring glutamate are defective in competence. (ii) Aconitase-less mutants which can only grow in the presence of glutamic acid are not transformable (19). (iii) Thorne and Stull (21) found that glutamic acid inhibits transformation of B. licheniformis. (iv) Several investigators (8, 15, 16) have suggested that alanine or glutamic acid might alter the maximal number of transformants in the B. subtilis system using bacterial DNA. (v) In the past, it has been well known to those working routinely with the B. subtilis transformation system that certain batches of casein hydrolysate support the development of good competence, whereas other batches from the same manufacturer are completely useless. Attempts to replace this constituent with yeast extract usually lower transformation frequency (18). Glutamic acid and alanine are major components of the mucopeptide chain in cell walls of a transformable strain of B. subtilis (22, 23, 27). Young (23) has proposed that hydrolysis of this cross-linked peptide could permit expansion of the rigid cell wall. We suggest that, in the absence of excessive glutamic acid or alanine (or both), such an expansion might occur and be enough to accommodate the binding of DNA at a pertinent site on the surface of the cytoplasmic membrane. If an excess of these two amino acids were provided, then competence

VOL. 94, 1967

COMPETENCE IN B. SUBTILIS

569

might be inhibited by a strengthening of the cell wall. Because of the timing of the appearance of competent cells, it is not surprising that suggestions of an association between competence and sporulation have been made (16, 20). However, the observations that competence will not develop in medium which is conducive to sporulation, and that transformed cells do not sporulate before giving rise to transformants [the medium usually employed does not have the necessary cofactors for sporulation (24)] indicate that competence probably represents a stage before the irreversible commitment to sporulation. We would expect it to appear after the induction of tricarboxylic acid cycle activity (5), since Spizizen et al. (18) have suggested that a-picolinic acid (an inhibitor of tricarboxylic acid activity) inhibits the development of competence. We suggest that the use of infectious phage 4)29 DNA to assay competence by this method permits more precision in the study of exogenous factors affecting competence than some other methods, as well as being potentially useful for studies on competence per se. Some of the specific advantages include the following: The entire 429 genome is small (approximately the same size as the average piece of bacterial DNA which is isolated, 1.0 x 107 daltons versus 6 X 106 to 1 X 107 for bacterial DNA), thereby providing a diminished chance of shearing during isolation and handling. This phage DNA is easy to isolate and has been shown by Reilly and Spizizen (10) to give a linear response of infectivity with increasing DNA concentration. It is also likely that phage DNA does not have to integrate or recombine with the recipient genome to be expressed (although this fact has not been verified experimentally). This would eliminate the differential probability of integration or variation due to marker reversion which has been associated with specific bacterial markers. It would also facilitate the study of competence in strains having no auxotrophic requirements. If integration is not necessary, phage DNA should serve as a direct assay of every competent cell in the population and thus act as a much better indicator of competence than bacterial DNA. In this respect, we should point out that the level of competence obtained with phage DNA should be greater than that indicated by use of any single bacterial marker, where selection of one class of transformants is a prerequisite for recording competence. In the phage system, all of the DNA molecules which have successfully entered the cells should be scored as transfectants. As seen in Fig. 1, this has not been observed in our experiments. It also does not appear to be evident

in the experiments of Reilly and Spizizen (10). We cannot presently offer an explanation for this discrepancy. One possibility is that some form of DNA restriction results, since the phage are grown on a host which is different from the transformable cells, and is assayed on a strain different from the competent cells. Another finding which may bear significance in this interpretation is that of Reilly and Spizizen, who demonstrated that although the number of transfectants with 4)29 DNA is directly proportional to the DNA concentration, an average of 10,000 to 20,000 phage equivalents is necessary for the expression of each plaque. This may reflect some form of DNA degradation within competent cells, since the data of Anderson et al. (3), which we have confirmed (unpublished data), indicates that the isolated 429 DNA is homogenous in size (5.8 u) and does not show that a substantial portion of the molecules are sheared during isolation.
ACKNOWLEDGMENTS We gratefully acknowledge the technical assistance of Sheila Prachand and the suggestions of B. Strauss during preparation of the manuscript. This research was supported by grant GB 4291 from the National Science Foundation, by institutional grant 41-G from the American Cancer Society to the University of Chicago, and by Public Health Service Training Grant

2-TO1-GM-0603.
LITERATURE CITED ADAMS, M. 1959. Bacteriophages. Interscience Publishers, Inc., New York. ANAGNOSTOPOULOS, C., AND J. SPIZIZEN. 1961. Requirements for transformation in Bacillus subtilis. J. Bacteriol. 81:741-746. ANDERSON, D. L., D. D. HICKMAN, AND B. E. REILLY. 1966. Structure of Bacillus subtilis bacteriophage c29 and the length of 029 deoxyribonucleic acid. J. Bacteriol. 91:2081-2089. Borr, K. F., AND R. DAvIDoFF-ABELSON. 1966. Altered sporulation and respiratory patterns in mutants of Bacillus subtilis induced by acridine orange. J. Bacteriol. 92:229-240. HALVORSON, H. 1965. Sequential expression of biochemical events during intracellular differentiation. Symp. Soc. Gen. Microbiol. 15:343368. HAYES, W. 1964. The genetics of bacteria and their viruses. John Wiley & Sons, New York. HOFSTEN, B. v. 1962. The effect of copper on the growth of Escherichia coli. Exptl. Cell Res. 26:606-607. KAMMEN, H. O., R. J. WOJNAR, AND E. S. CANELLAKIS. 1966. Transformation in Bacillus subtilis. II. The development and maintenance of the competent state. Biochim. Biophys. Acta 123: 56-65.

1.
2.

3.

4.

5.

6.
7.
8.

570

BOTT AND WILSON

J. BAcr1ERioL.

9. OKuBo, S., AD W. R. ROMIG. 1966. Impaired transformability of Bacillus subtilis mutant sensitive to mitomycin C and ultraviolet radiation. J. Mol. Biol. 15:440-454. 10. RELuY, B. E., AND J. SPIZIZEN. 1965. Bacteriophage deoxyribonucleate infection of competent Bacillus subtilis. J. Bacteriol. 89:782-790. 11. RomivG, W. R. 1962. Infection of Bacillus subtilis with phenol extracted bacteriophages. Virology 16:452-459. 12. SArro, H., AND K. MIURA. 1963. Preparation of transforming deoxyribonucleic acid by phenol treatment. Biochim. Biophys. Acta 72:619-629. 13. SCHAEFFER, P. 1964. Transformation, p. 87-153. In I. C. Gunsalus and R. Y. Stanier [ed.], The bacteria, vol. 5, Heredity. Academic Press, Inc., New York. 14. SCHAEFFER, P., H. IoNEsco, A. RYTER, AND G. BALASSA. 1965. La sporulation de Bacillus subtilis; etude genetique et physiologique. Colloq. Intern. Centre Natl. Rech. Sci. Mecanismes R6gulation, MarseiUe, 1963, p. 553-563. 15. SpuzzNs, J. 1959. Genetic activity of deoxyribonucleic acid in the reconstitution of biosynthetic pathways. Federation Proc. 18:957-965. 16. SPuZzN, J. 1961. Studies of transformation of sporulating characters, p. 142-148. In H. 0. Halvorson [ed.], Spores II, Burgess Publishing Co., Minneapolis. 17. SPLzzEN, J. 1965. Analysis of asporogenic mutants in Bacillus subtilis by genetic transformation, p. 125-137. In L. L. Campbell and H. 0. Halvorson [ed.], Spores III. American Society for Microbiology, Ann Arbor, Mich. 18. SPIzIzEN, J., B. E. REILLY, AND A. H. EVANS. 1966.

19.

20.
21.

22.

23.

24.
25.

26.

27.

Microbial transformation and transfection. Ann. Rev. Microbiol. 20:371-400. SZULMAJSTER, J., AND R. HANSON. 1965. Physiological control of sporulation in Bacillus subtilis, p. 162-173. In L. L. Campbell and H. 0. Halvorson [ed.], Spores III. American Society for Microbiology, Ann Arbor, Mich. TAKAHASHI, I. 1961. Genetic transduction in Bacillus subtilis. Biochem. Biophys. Res. Commun. 5:171-175. THORNE, C., AND H. B. STULL. 1966. Factors affecting transformation of Bacillus licheniformis. J. Bacteriol. 91:1012-1020. YOUNG, F. E. 1965. Variation in the chemical composition of the cell walls of Bacillus subtilis during growth in different media. Nature 207: 104-105. YOUNG, F. E. 1966. Autolytic enzyme associated with cell walls of Bacillus subtilis. J. Biol. Chem. 241:3462-3467. YOUNG, F. E. 1967. Competence in Bacillus subtilis transformation system. Nature 213:773775. YOUNG, F. E., AND J. SPiZIZEN. 1961. Physiological and genetic factors affecting transformation of Bacillus subtilis. J. Bacteriol. 81:823-829. YOUNG, F. E., AND J. SPizIzEN. 1963. Incorporation of deoxyribonucleic acid in the Bacillus subtilis transformation system. J. Bacteriol. 86:392-400. YOUNG, F. E., J. SPIZIZEN, AND I. CRAWFORD. 1963. Biochemical aspects of competence in the Bacillus subtilis transformation system. I. Chemical composition of cell walls. J. Biol. Chem. 238:3119-3125.