JOURNAL oF BACERoLOY, Mar. 1975, p. 771-776 Copyright 0D 1975 American Society for Microbiology

Vol. 121, No. :3 Printed in UI.S.A.

Interactions Between Exogenous Deoxyribonucleic Acid and Membrane Vesicles Isolated from Competent and Noncompetent Bacillus subtilis HANS JOENJE,* WIL N. KONINGS, AND GERARD VENEMA Departments of Genetics* and Microbiology, Biological Centre, University of Groningen, Kerklaan 30, Haren (Gn.), The Netherlands

Received for publication 26 December 1974

Competent cultures of Bacillus subtilis 168 have been fractionated into a high-competent and a low-competent fraction by a large-scale separation technique. Membrane vesicles isolated from both cell fractions are equally active in the transport of L-glutamate. Both membrane vesicle preparations seem to have similar endo- and exonuclease activities. Also, both preparations are capable of binding deoxyribonucleic acid. However, especially at low deoxyribonucleic acid concentrations (1 ,g or less per ml), vesicles obtained from competent cells bind significantly more deoxyribonucleic acid (up to sixfold) than vesicles from noncompetent cells. In genetic transformation, several successive events have been recognized in the processing of donor deoxyribonucleic acid (DNA) after its addition to competent cultures of Bacillus subtilis (2, 4, 5). Immediately after the binding of a DNA molecule to a competent cell, the DNA is subjected to double-stranded nucleolytic cleavage and to exonucleolytic activity (5). Recently we showed that membrane vesicles isolated from B. subtilis seem to be able to carry out early steps of the transformation process, as observed in whole cells: DNA is bound to the vesicles and suffers endo- and exonucleolytic attack (7). However, these observations were made on membrane preparations from a mixed population of competent and noncompetent cells, because in maximally competent cultures the majority of the cells are not competent (17). This study was undertaken to establish whether differences exist in the interactions of DNA with membrane vesicles obtained from competent and noncompetent cells. This investigation required the development of a method allowing large-scale separation of competent and noncompetent cells. MATERIALS AND METHODS Strains. B. subtilis 1G-20 (trpC2) was used for the preparation of membrane vesicles. Strain 1G-22 (thy) was used for the isolation of 'H-labeled DNA; unlabeled DNA was isolated from strain OG-1 (prototrophic). Media. Spizizens (15) minimal salts plus glucose (0.5%), casein hydrolysate (200 Ag/ml), and auxotrophic requirements (20 gg/ml) (minimal growth

medium) was used for making overnight cultures. WB medium, used as a second growth medium, contained Spizizens (15) minimal salts plus glucose (0.5%), 5 mM MgSO4, and competence-stimulating amino acids (each 50 jg/ml), according to Wilson and Bott

(19).

DNA isolation. Unlabeled and 'H-labeled DNAs were isolated from strains OG-1 and 1G-22, respectively, as described previously (7). The specific radioactivity of the 'H-labeled DNA solution was approximately 1.4 x 105 counts/min per ug. Assay of L-glutamate transport activity in membrane vesicles. Transport of L-glutamate by membrane vesicles was measured in the presence or absence of reduced (by ascorbate) phenazine methosulfate, as described by Konings and Freese (10). The reaction mixtures (0.1 ml) contained membrane vesicles (ca. 400 jAg of protein per ml), 50 mM potassium phosphate (pH 7.0), 10 mM MgSO4, "4C-labeled L-glutamate (9.25 MM, specific activity 270 Ci/mol), and either 10 mM sodium ascorbate (pH 6.6) plus 10

gM phenazine methosulfate, or water. Separation of competent and noncompetent cells. The method described is a modification of the procedure reported by Cahn and Fox (3) and Hadden and Nester (6); they separated competent and noncompetent cells, which have different buoyant densities, by centrifugation of the cells through a stepwise density gradient of Renografin. An overnight culture in minimal growth medium was diluted in WB medium to give 2.5 liters of culture containing 5 x 107 colony-forming units per ml. After growth at 37 C for 3.5 h with vigorous aeration, the cells were harvested by centrifugation and resuspended in 210 ml of the culture supernatant fluid by means of a Potter-Elvehjem type homogenizer. The suspension was mixed with 125 ml of a 65% (wt/vol) Angiografin solution (Schering AG Berlin/Bergka-

771

772

JOENJE, KONINGS, AND VENEMA

men, Germany), and the refractive index was adjusted with either the culture supernatant or concentrated Angiografin to 1.374 (corresponding to a density of 1.140 g/cm3 at 20 C). Angiografin, like Renografin, is an aqueous solution of methyl glucamine N, N'-diacetyl-3,5-diamino-2,4,6-triiodobenzoate. Cellulose nitrate tubes of a fixed-angle rotor (type 30, Beckman Instruments, Palo Alto, Cal.) were filled with 27.5-ml portions of the mixture and, after a 2.5-ml cushion consisting of a mixture of 65% (wt/vol) Angiografin and WB medium (1:1) was injected underneath this layer, centrifugation was carried out at 20 C for 15 min at 60,000 x g (in the middle of the tube). The rotor was allowed to slow down without the use of the brake. Cells were collected from the top (T cells) and from the cushion (bottom or B cells) with a syringe after the tube wall was punctured. T and B cells were then washed by centrifugation with WB medium, then with 0.2 M MgSO4, and once more with WB medium and resuspended in 0.05 M potassium phosphate (pH 7.5) plus 10 mM MgSO4 to give about 109 colony-forming units per ml for the isolation of membrane vesicles. Samples of about 108 T and B bacteria were used to determine their transformabilities by assaying the trpC2' transformation frequency after a 20-min exposure to trpC2+ DNA at 37 C in WB medium. Isolation of membrane vesicles and the assay of DNA binding by membrane vesicles. Membrane vesicles were isolated and DNA binding was assayed as described previously (7). Determination of endonuclease activity in membrane vesicles. Endonucleolytic activity in membrane vesicles was determined by exposure of DNA to vesicles and subsequent determination of the DNA's rate of sedimentation through linear neutral sucrose gradients (5 to 20% [wt/voll sucrose in 0.15 M NaCl plus 0.015 M trisodium citrate). For this purpose, tritiated DNA (20 sg/ml) was exposed to membrane vesicles (300 ug of protein per ml) in the presence of 20 mM MgSO4 at 30 C for 30 min. Then ethylenediaminetetraacetic acid (50 mM) was added to stop further nuclease action. The membranes were dissolved by addition of 0.05% sodium dodecyl sulfate and autodigested (2.5 h, 42 C) Pronase (5 mg/ml); after incubation at 42 C for 2 h, an equal volume of 10% (wt/vol) sodium dodecyl sulfate was added. Amounts (0.1 ml) of the digest were layered on top of the sucrose gradients, and centrifugation was carried out at 40,000 rpm for 2 h at 20 C in an SW50L rotor of a Beckman L-2 preparative ultracentrifuge. The gradients were analyzed as described previously (7). Analysis of acid-soluble DNA degradation products. Membrane vesicles (840 gg of protein per ml) were incubated in 0.1 M potassium phosphate (pH 7.0) plus 20 mM MgSO4 with 3H-labeled DNA (33 gg/ml) at 30 C for 90 min. Acid-soluble degradation products were collected as described previously (7). Samples (0.1-ml) of acid-soluble material were layered on top of a Sephadex G-15 column (75 by 0.42 cm) together with 0.2 ml of a solution containing the following reference substances: blue dextran (0.5 mg), 5'-thymidylic acid (50 Ag), thymidine (50 Ag), and thymine (50 jg). The samples were eluted with 0.05 M

J B ACTERIOL .

ammonium acetate. Fifty fractions of 20 drops each were collected in 4 h. The positions of the void volume (blue dextran), 5'-thymidylic acid, thymidine, and thymine were recorded by ultraviolet spectrophotometry. The radioactivity of the fractions was determined as described previously (7). RESULTS

Large-scale separation of competent and noncompetent cells. Owing to the tendency of' cells to clump, satisfactory separation by the Cahn and Fox (3) and Hadden and Nester (6) method is obtained with small quantities of cells only. We have scaled-up the fractionation procedure by mixing a concentrated cell suspension with Angiografin to a density in between the densities of the competent and noncompetent cells. Subsequent centrif'ugation causes the competent (less dense) cells to float to the top of' the solution and noncompetent (more dense) cells to sediment to the bottom of the tube. The separation obtained is qualitatively similar to that of the conventional technique; however, many more cells can be fractionated. With a 360-ml fixed-angle rotor, cells from 2.5 liters of culture can be fractionated; with high-capacity zonal rotors, more than 10 liters of culture may be easily fractionated in this way. The general characteristics of both fractionation techniques are described in Fig. 1. In this paper, T and B cells refer to bacteria collected after centrifugation from the top and the bottom of the centrifuge tubes, respectively; the membrane vesicles are indicated accordingly. The transformability of the T cells was generally 50- to 200-f'old higher than that of the B cells. During incubation in WB medium at 37 C, this difference is maintained for at least 40 min; thereafter the B cells start to develop competence (Fig. 2), reducing the difference in transformability to 3.5-fold after 60 min of incubation. Functional and structural integrity of the membrane vesicles. The isolation procedure of membrane vesicles, employed in this investigation, results in membrane vesicles that have the same orientation as the cytoplasmic membrane, as has been demonstrated by f'reeze-etch electron microscopy (9) and by a study of the localization of several membrane-bound enzymes (8). A further indication for the functional and structural integrity of the vesicles is the occurrence of respiration-linked transport processes that require a functional coupling of the transport carrier with an intact respiratory chain. Uptake of L-glutamate in the presence of the electron donor system ascorbate-phenazine methosulfate was studied both in T and B vesicles (Fig. 3). Both types of vesicles transport

L-glutamate with a high initial rate, comparable with initial rates reported previously for membrane vesicles isolated from unfractionated cultures (7). BEFORE /DURING CENTRIFUGATION -

AFTER CENTRIFUGATION

centrifugal force

Htop A

773

DNA-MEMBRANE INTERACTIONS IN B. SUBTILIS

VOL. 121, 1975

cells

' eGz:41X C

II

s

bottom cells

renogrof in

mxture of cellstocel bottom cells and renogrotin C

FIG. 1. Schematic representation of the two cell fractionation methods described. Method A (Cahn and Fox [31, Hadden and Nester [6]), Suitable rotor types-swinging-bucket rotor; separation occurs in a plane (interface S); separation capacity limited by the internal tube diameter (maximum + 109 bacteria per cm2). Method B (this paper): Suitable rotor types-swinging-bucket, fixed-angle, zonal; separation occurs in the whole volume; separation capacity limited by the tube content (maximum i 3 x 10' bacteria per ml); arrows indicate the direction of movement of the two cell types. C, Cushion of high density.

Nuclease activities in T and B vesicles. Endonucleolytic activity was determined in T and B vesicles by sedimentation analysis of tritiated DNA after exposure to the vesicles. Exposure to B vesicles in the presence of excess ethylenediaminetetraacetic acid, which inhibits nuclease activity (7), served as a control. DNA is fragmented by T and B vesicles to approximately the same extent, although the fragmentation seems to be slightly more extensive and the fragments seem to be slightly more heterogeneous in the case of exposure to T vesicles (Fig. 4). These results indicate that no major difference exists in the endonuclease content of both types of vesicles. We have shown previously (7) that prolonged incubation of DNA with high concentrations of membrane vesicles leads to the appearance of acid-soluble DNA degradation products consisting partially of monomers; these monomers are indicative of exonuclease activity (7). T and B vesicles appear to produce similar amounts of acid-soluble products (data not shown). Analysis of these products by gel filtration over a column of Sephadex G-15, which separates monomers from oligomers and also separates thymidylic acid from its derivatives thymidine and thymine, reveals, however, that the composition of the acid-soluble material generated by T vesicles is different from that of the B vesicles (Fig. 5). The products generated by the T vesicles consist of 69.3% oligomers, less than 1%

0.100

0z

0~

cr as I0)

0 0.

I-

c E

QL05-

U)

L0

E

L-

A I

uX-IL-

0

T

20

I~ ~ ~ I

I

-

40 60 80 time (min)

C

I

100

FIG. 2. Competence development of T and B cells after fractionation. T and B cells were resuspended in WB medium and incubated at 37 C. After various intervals, samples were transformed for 10 min with 0.01 ug of DNA per ml and plated for trpC2+ transformants. Symbols: 0, T cells; 0, B cells.

time (min) FIG. 3. L-Glutamate transport in the presence ( ~) or absence (- -) of ascorbate-PMS by T vesicles (0) and B vesicles (x). -

J. BACTERIOL.

JOENJE, KONINGS, AND VENEMA

774

I

E

0. u

as a function of the DNA concentration. T vesicles are capable of binding more DNA than B vesicles in the whole range of DNA concentrations tested (Fig. 6). In both types of vesicles three saturation levels may be distinguished, as has been found also in membrane vesicles prepared from unfractionated competent cultures (7). The maximal difference in binding capacity (about sixfold) is observed at DNA concentrations up to 1 jig/ml (first saturation level); the difference decreases at higher DNA concentrations, being 2.5- and 1.7-fold at 10 and 70,ug of DNA per ml, respectively. These results suggest that, although the overall binding capacity of the T vesicles is greater than that of the B vesicles, the binding site with the highest affinity for DNA is relatively the most abundant in T vesicles.

DISCUSSION In an earlier paper (7) we showed that, in contrast to isolated cell walls, membrane vesicles isolated from unfractionated competent B. subtilis cultures, which are mixed populations of competent and noncompetent cells, are capable of binding DNA. Exposure of DNA to such membrane preparations also leads to endonu-

0 6 4

2

3D

0 24 16 8 fraction number FIG. 4. Sucrose gradient sedimentation profiles of DNA exposed to membrane vesicles. (A) DNA exposed to B vesicles in the presence of excess ethylenediaminetetraacetic acid for 30 min at 30 C (control). (B) DNA exposed to B vesicles in the presence of excess MgSO4 for 30 min at 30 C. (C) DNA exposed to T vesicles in the presence of excess MgSO4 for 30 min at 30 C. The arrow indicates the position of T7 DNA as determined in separate centrifugations. Sedimentation is from left to the right.

OD E

8-

Q

I

u

0

64-

5'-thymidylic acid and 9.1 and 20.8% thymidine and thymine, respectively. The acid-soluble material produced by the B vesicles contains 43.1% oligomers, 11.7% 5'-thymidylic acid, 44% thymidine, and 1.2% thymine. This indicates that the action of the enzymes 5'-nucleotidase and thymidine phosphorylase, which convert 5'-thymidylic acid into thymidine and thymidine into thymine, respectively, is more pronounced in the T vesicles than in the B vesicles, which seem to lack the thymidine phosphorylase activity almost completely. DNA binding capacity of T and B vesicles. DNA binding to T and B vesicles was measured

TMP

20%

a

thyrnne

I

thymidin.

.JnJTJEnr n I,WOD PU

30 20 10 froctin number

-

40

M

50

FIG. 5. Fractionation of acid-soluble DNA degradation products by gel filtration on Sephadex G-15. Products generated from DNA by exposure to B vesicles (B) and to T vesicles (T). The arrows indicate the peak fractions of the different reference substances. BD, Void volume determined with dextran blue; TMP, 5'-thymidylic acid.

VOL 121, 1975

DNA-MEMBRANE INTERACTIONS IN B. SUBTILIS

10

/

o

//

E z

CD

0.10.1

g 1.0

10 100 pg DNA/ml

FIG. 6. DNA binding to T vesicles (0) and B vesicles (0) as a function of the DNA concentration. The dashed line represents a first-order relationship between DNA concentration and the amount of bound DNA. Membrane concentrations amounted to 320 and 340 ug of protein per ml for T and B vesicles, respectively.

cleolytic and exonucleolytic breakdown of the DNA. It was questionable, however, whether these phenomena are restricted to the membrane of a competent cell exclusively, or whether they are also occurring in membranes from noncompetent cells. The results of the present experiments, in which membrane vesicles from the competent (T) and noncompetent (B) cell fractions have been compared, suggest that, qualitatively, the interactions between DNA and these membrane preparations are similar. However, we observed a lower DNAbinding capacity in the B vesicles than that in the T vesicles, especially at low DNA concentrations (Fig. 6). Three saturation levels seem to be distinguishable in these dose-response curves. These levels are suggestive of the existence of three DNA-binding sites with different affinities for DNA, being saturated at about 1, 10, and 70,Mg of DNA per ml. Similar observations have been made with membrane vesicle preparations from unfractionated cultures at a sevenfold higher protein concentration (7). The results suggest that the T vesicles contain more accessible DNA-binding sites per milligram of protein than the B vesicles, the high-affinity site (saturated at approximately 1 Mg of DNA per ml) being relatively the most abundant. It is conceivable that this particular binding site is involved in the transformation process, since

775

the transformation system of B. subtilis is also saturated at about 1 Mg of DNA per ml. In contrast to whole noncompetent cells, which are completely unable to bind DNA (3; H. Joenje, unpublished results), membrane vesicles from noncompetent cells do bind DNA. We entertain the possibility that the membranous sites of the noncompetent cells, which are potentially able to interact with DNA, are masked by the cell wall and that acquisition of the competent state requires a change in the cell wall permitting the DNA to interact with the binding sites. Such a model has been developed for B. subtilis (1) and Pneumococcus (13, 14,

16). Vermeulen and Venema (18) have presented evidence for the involvement of mesosomes in the primary fixation and transport of the donor DNA. This finding might suggest that the transformation-specific DNA-binding sites are associated with the mesosomes exclusively. Since they observed that competent cells contain more mesosomes than noncompetent cells, our results of the different DNA binding capacity in T and B vesicles might well be due to a higher proportion of mesosomal vesicles in preparations of T vesicles. ACKNOWLEDGMENTS We thank W. J. Feenstra for valuable comments on the manuscript. This investigation was carried out under the auspices of the Netherlands Foundation for Chemical Research (S.O.N.) and with financial aid from the Netherlands Organization for the Advancement of Pure Research (Z.W.O.). LITERATURE CITED 1. Akrigg, A., S. R. Ayad, and J. Blamire. 1969. Uptake of DNA by competent bacteria-a possible mechanism. J. Theor. Biol. 24:266-272. 2. Arwert, A., and G. Venema. 1973. Transformation in Bacillu.s subtilis. Fate ot newly introduced transtorming DNA. Mol. Gen. Genet. 123:185-198. 3. Cahn, F. H., and M. S. Fox. 1968. Fractionation of transformable bacteria from competent cultures of Bacillus subtilis on Renografin gradients. .J. Bacteriol.

95:867-875. 4. Davidoff-Abelson, R., and D. Dubnau. 1973. Kinetic analysis of the products of donor deoxyribonucleate in

transformed cells of Bacillus subtilis. J. Bacteriol. 116:154-162. 5. Dubnau, D., and C. Cirigliano. 1972. Fate of transforming DNA following uptake by competent Bacillus subtilis. III. Formation and properties of products isolated from transformed cells which are derived entirely from donor DNA. J. Mol. Biol. 64:9-29. 6. Hadden, C., and E. W. Nester. 1968. Purification of competent cells in the Bacillus subtilis transformation system. J. Bacteriol. 95:876-885. 7. Joenje, H., W. N. Konings, and G. Venema. 1974. Interactions between exogenous DNA and membrane vesicles isolated from Bacillus subtilis 168. .J. Bacteriol. 119:784-794.

776

JOENJE, KONINGS, AND VENEMA

8. Konings, W. N. 1975. Localization of membrane proteins in membrane vesicles of Bacillus subtilis. Arch. Biochem. Biophys., in press. 9. Konings, W. N., A. Bisschop, M. Veenhuis, and C. A. Vermeulen. 1973. New procedure for the isolation of membrane vesicles of Bacillus subtilis and an electron microscopy study of their ultrastructure. J. Bacteriol. 116:1456-1465. 10. Konings, W. N., and E. Freese. 1972. Amino acid transport in membrane vesicles of Bacillus subtilis. J. Biol. Chem. 247:2408-2418. 11. Kooistra, J., and G. Venema. 1971. Effect of temperature on the fate of donor DNA in transtormation of Haemophilus influenzae, p. 408-417. In L. G. H. Ledoux (ed.), Informative molecules in biological systems. North-Holland Publishing Co., Amsterdam. 12. Lacks, S., B. Greenberg, and K. Carlson. 1967. Fate of donor DNA in pneumococcal transtormation. J. Mol. Biol. 29:327-:347. 13. Seto, H., and A. Tomasz. 1974. Early steps in DNA binding and uptake during genetic transformation of pneumococci. Proc. Nat. Acad. Sci. UJ.S.A. 71:1493-1498. 14. Seto, H., and A. Tomasz. 1975. Protoplast formation and leakage of intramembrane components: induction by

15.

16.

17.

18.

J. BACTERIOL.

the competence activator substance of Pneumococcus. J. Bacteriol. 121:344-353. Spizizen, J. 1958. Transformation of biochemically deficient strains of Bacillus subtilis by deoxyribonucleate. Proc. Nat. Acad. Sci. U.S.A. 44:1072-1078. Tomasz, A. 1973. Interactions between cells and DNA molecules: the nature of the competent state, p. 322-352. In L. Leive (ed.), Bacterial membranes and walls. Marcel Dekker, New York. Vermeulen, C. A., and G. Venema. 1971. Autoradiographic estimation of competence and the relationship between competence and transformability in cultures of Bacillus subtilis. J. Gen. Microbiol. 69:239-252. Vermeulen, C. A., and G. Venema. 1974. Electron microscope and autoradiographic study of ultrastructural aspects of competence and deoxyribonucleic acid absorption in Bacillus subtilis: ultrastructure of competent and noncompetent cells and cellular changes during development of competence. J. Bacteriol.

118:334-341.

19. Wilson, G. A., and K. F. Bott. 1968. Nutritional tactors influencing the development of competence in the Bacillu.s subtilis transformation system. .J. Bacteriol.

95:1439-1449.