DNA Uptake During Genetic Transformation and the Growing

Proc. Nat. Acad. Sci. USA Vol. 68, No. 8, pp. 1848-1852, August 1971

DNA Uptake during Genetic Transformation and the Growing Zone of the Cell Envelope (pneumococcus/electron microscopy/choline/ethanolamine/autoradiography) ALEXANDER TOMASZ, EVA ZANATI, AND RICHARD ZIEGLER The Rockefeller University, New York, N.Y. 10021 Communicated by Norton D. Zinder, May 26, 1971

ABSTRACT Continuous cellular incorporation of cho- (10), determination of CF-receptor activity (11), and the line molecules is essential for DNA binding to compe- tests for the competence-specific agglutinin (12) were all tent pneumococci. The choline molecules are incorporated into the cell surface at the equatorial region of the performed by published procedures. The concentration of CF cocci. As a working hypothesis, it is proposed that DNA during the "activation" assay was usually 100-500 Units molecules enter these bacteria at the growing zone(s) of (13) per ml. Lysis sensitivity of pneumococci was determined the cell envelope. as follows: 0.9 ml of the culture was pipetted into tubes pre- 0.1 Ml in the pared with ml of phosphate buffer (1 M, pH 8) and 50 The presence of choline residues teichoic acid portion of a solution of deoxycholate. The preparation of Bacillus of the pneumococcal cell envelope (1) is essential for several 5% is subtilis DNA labeled with [3H]thymidine and radioautography physiological functions of this bacterium. Not only choline of cell-bound DNA were described (14). Pneumococcal DNA required for growth (2), but biosynthetic replacement of carrying the streptomycin-resistance marker and labeled with choline in the cell surface by structural analogs (such as [3H]thymidine was prepared from a streptomycin-resistant, ethanolamine) causes several striking physiological changes thymidine-requiring strain of pneumococcus (15). is a in cell in the bacteria (2-5); one such change defect divi- Reversible and irreversible binding (16) of radioactive DNA sion that prevents the physical separation of daughter cells was determined as follows: to a 3-ml culture of bacteria The of new choline molecules from one another. incorporation (about 107 viable cells/ml), 3H-labeled transforming DNA into the cell surface takes place at a distinct growing zone, was added (0.1 Mg; 106 dpm/ml). After incubation at 30°C of We now localized at the equatorial region pneumococci (6). for 20 min, the tube was chilled on ice; a small aliquot was describe evidence which suggests that the cell-surface changes removed into growth medium containing DNase (1 Mg/ml) that make pneumococci capable of absorbing and integrating and later scored for the number of transformants. The extracellular DNA are in some way coupled to the functioning rest of the suspension was centrifuged in the cold (10,000 X of the growing zone of the cell envelope. g, 10 min) and, after two washes with 5-ml portions of cold MATERIALS AND METHODS growth medium, a portion of the suspension was pipetted into Cultures of the R36A strain of Diplococcus pneumoniae cold 10% TCA and the precipitate was collected on a glass- were used in all experiments at a concentration of 107-108 fiber filter to determine "reversibly bound" DNA. The rest cells/ml. The composition of a chemically defined (Cd) (7) of the cell suspension was treated with 5-10 ug/ml of pan- and a semisynthetic (C) (8) medium was described earlier. creatic DNase and 10 Mul/ml of snake venom (Russell snake An enriched defined medium (Cden) was prepared by sup- venom, Worthington) at 37°C for 10 min; after two washings plementing Cd medium with all the amino acids present in with growth medium, a portion of the suspension was pipetted acid-hydrolyzed casein, at the same concentrations as in the into cold 10% TCA and the precipitate was collected onto C medium. "Choline-phase" and "ethanolamine-phase" glass-fiber filters to determine DNA bound in a nuclease- bacteria were grown in Cden medium supplemented with resistant ("irreversible") form. A Hitachi, Perkin-Elmer 5 MLg/ml of choline or 40 Mg/ml of ethanolamine, respectively. type HU-1lC-1 electron microscope was used at 75 kV. All Bacterial growth was monitored with a Coleman nephe- chemicals used were analytical grade commercial products. lometer. Cells were transferred from one medium to another RESULTS by filtration (Millipore filters, pore size 0.45 Mm). Radioactive Two types of experimental designs were used. In one of these, tracers were used at 5 Mg/ml and 1 MCi/ml concentrations, cells growing in growth medium containing choline ("choline- unless otherwise noted. The incorporation of radioactive phase cells") were transferred to an amino-alcohol-free tracers into cold trichloroacetic acid (TCA)-insoluble mate- medium or to a medium containing ethanolamine in place of rial was measured by collecting the samples on glass-fiber choline. In the second type of experiment, pneumococci that disks (Schleicher and Schuell); after drying, the disks were were growing in ethanolamine-containing medium for several counted in a toluene-based scintillator in a scintillation generations ("ethanolamine-phase cells") were transferred spectrometer (Unilux, Nuclear-Chicago). The preparation and back to choline-containing medium. assay of competence factor (CF) (9) and transforming DNA For the interpretation of such experiments, it was essential first to determine the effect of choline deprivation on the Abbreviations: TCA, trichloroacetic acid; CF, competence factor. general metabolism of pneumococci. It was found that for 1848 Downloaded by guest on October 7, 2021 Proc. Nat. Acad. Sci. USA 68 (1971) DNA Uptake and the Cell Envelope 1849 TABLE 1. Effect of choline deprivation on cellular response to CF

[32P] DNA bound (cpm/ml Transformants Tube Medium of culture) per ml A Complete 140 3.2 X 106 B No choline 2 <102 6IC C Choline readded 30 1.4 X 101 IU, 30 Ea 6 CI Control bacteria (A), choline-starved cells in choline-free 0c10 6 0 (n medium (B), and in medium resupplemented with choline (C) C 25 were treated with CF for 20 min (8 ml of cells to 1 ml of CF); 0 \ I , 5-ml portions of the cultures then received 0.1 ml of 32P-labeled D 6 transforming DNA, 2.5 of after 20 min of incuba- -oU.0 1g choline, and, a}) 03 ' 0 20 tion, DNase (10 usg/ml); after 15 min of incubation at 370C, E 7. small aliquots were tested for transformability, the rest of the z I I 15 b suspensions were washed four times with 5 ml of medium, then x 1-ml portions were filtered onto Millipore membranes, dried, and E tested for cell-bound radioactivity by scintillation counting. o02- 010203 0 06 110 about one generation time after the removal of choline from the medium, the cells grew; the rates of mass increase and 5 of DNA, RNA, and protein syntheses were normal. Depletion of the intracellular pools of choline compounds was completed during the first third of the period of growth in the absence of choline. In all experiments, the effect of choline Minutes deprivation on transformability was tested within this period of normal growth. FIG. 1. Decay of cellular response to CF during depletion of choline pools. Physiologically incompetent pneumococci were Choline deprivation experiments transferred from the low pH, Cden medium to fresh Cd.. medium A culture of physiologically incompetent bacteria growing in at pH 8, containing a limiting amount of choline (0.1 sg, 0.5 the enriched defined medium at low pH (6.8) was transferred A&Ci/ml). At frequent intervals after the transfer, 0.1-ml aliquots were removed to test incorporation of choline into cold TCA- to choline-free medium at the same pH. The culture was precipitable material (solid lines). 0.8-ml samples were pipetted halved; a control suspension was supplied with normal con- into tubes with 0.1 ml CF and, after 15 min of incubation, the bac- centration of choline (tube A), the second received no amino- teria were further diluted (0.1 ml into 1.0 ml) into C medium alcohol (tube B). After incubation at 370C for 30 min, the containing choline, transforming DNA, and subtilisin. Trans- pH of both cultures was adjusted to 8, and half of the suspen- formations were terminated after 10 min. Dashed lines indicate sion in tube B was transferred to a third test tube containing the number of transformants. A second "pulse" of choline (0.1 a normal concentration of choline (control for reversibility; ,pg; 0.5 ACi/ml) was added to the pH 8 culture after 40 min tube C). Each of the three cultures then received CF; after of incubation with the first pulse of choline. 20 min of treatment with CF, the level of competence was In another experiment of identical design, activatability and determined by incubation with transforming DNA. Table 1 binding of [3H] thymine-labeled transforming DNA was tested shows that (a) choline deprivation resulted in a virtual dis- during the incorporation of a single, nonradioactive choline pulse (0.1 Aig of choline/ml) at times indicated by A, B, C, and D. For appearance of "activatability" by CF, measured both as results, see Table 2. genetic transformation or as ability to bind radioactive DNA. (b) The inhibitory effect of choline deprivation was quickly reversible by the addition of choline to the same In another experiment, competent bacteria have been ex- medium in which the bacteria were starved. This result exposed to temporary choline starvation and/or ethanolamine cludes trivial artifacts such as "leakage" of proteolytic or treatment, subsequent to completion of DNA uptake (i.e., after nucleolytic enzymes into the medium by the choline-starved addition of DNase). No decrease in the number of transfor- cells. Fig. 1 shows that in a choline-limited culture, the pre- mants could be observed upon returning such cells to choline- cipitous decay in cellular response to CF closely follows in containing medium (streptomycin-resistant transformants per time the slow down and eventual stop of choline incorporation ml in culture without choline starvation: 1.2 X 107; after 30 into TCA-precipitable material. In the experiment of Fig. 1, min of choline starvation: 1.39 X 107; after 30 min of treat- one can see two consecutive choline pulses, each accompanied ment with ethanolamine 1.38 X 107). This experiment indi- by a burst of activatability. cates that, once it is absorbed by the cells, the fate of transforming DNA is no longer influenced by modification of the Shift of choline-phase cells to ethanolamine cell envelope. When physiologically incompetent bacteria growing in Cd The experiments described so far indicate that cellular medium containing choline were transferred to Cd medium competence for DNA uptake and transformation is inti- containing ethanolamine in place of choline, the cell's ability mately coupled to choline incorporation into the cell en- to respond to CF virtually disappeared within 40 min. velope. Since such incorporation occurs at distinct growing Downloaded by guest on October 7, 2021 1850 Microbiology: Tomasz et al. Proc. Nat. Acad. Sci. USA 68 (1971

c o To ,0 E

I- -- a, 31

CL I'

2' 1- i'

^ ,-

i Io-.'

1 e- . r FIG. 2. Reappearance of CF-induced competence during shift of ethanolamine cells to choline. Bacteria growing in enriched defined medium containing [1,2-'4C]ethanolamine, 1 uCi and 40 Mg/ml, pH 6.8, were diluted 10-fold into fresh medium at pH 8 with and without aminoalcohol. During 50 min of incubation in this medium, incorporation of residual ethanolamine into TCA-precipitable material was followed (dashed lines). Response to CF was determined by removing 0.8-ml aliquots to 30'C, adding 0.1 ml CF, and incubating for 20 min, transforming DNA and subtilisin were then added and the incubation was terminated FIG. 3. Removal of equatorial portions of the cell envelope 10 min later by the addition of DNase (horizontal bars indicate by autolysin treatment of choline-pulsed ethanolamine cells. the number of transformants). At the 55th min of the experi- Long arrows indicate sites of autolysin action. In one of the ment, 0.5 ,ug (and 5 ,uCi) per ml of ['H] choline was added to the electron micrographs (lower right) the two "cell-wall bands" are culture and the incorporation of choline into TCA-precipitable also visible (short arrows), indicating the boundaries of the newly material (solid line) and cellular response to CF were followed. formed cell surface (18). Frequent samples were also taken to determine lysis sensitivity: "Normal" lysis sensitivity (typical of choline-phase cells) ap- peared by the 12th min after choline addition (arrow). The insert cholate) of the culture, and (c) the reappearance of cellular shows the irreversible binding of ['H]DNA by ethanolamine- response to CF. This latter was assayed by incubating por- phase cells (growth in nonradioactive ethanolamine) that were tions of the culture with CF (20 min) followed by the addi- allowed to incorporate choline for 20 min, and were then ex- tion of subtilisin and transforming DNA (10 min). Fig. 2 posed to [3H]thymidine-labeled DNA (prepared from B. sub- shows that the incorporation of choline into TCA-precipitable tilis) in the presence of subtilisin, at 30'C, for 10 min. material started without delay; normal lysis sensitivity of the culture (i.e., 80-90% "clearing" by deoxycholate) was com- zones located at the cellular equator, experiments were de- plete by the 12th min after choline addition; within the first signed to provide further evidence for the involvement of 20 min after the reintroduction of choline the bacteria started these growing zones in DNA uptake. In these experiments, to regain their normal response to CF* (i.e., they could be we planned to exploit two properties of the pneumococcal induced to competence). While this rapid recovery of com- cell envelope: (a) the resistance of ethanolamine-containing petence required both choline and CF, choline addition alone cell walls to the hydrolytic action of the pneumococcal was sufficient for the return of normal lysis sensitivity. Actu- autolytic enzyme, and (b) the fact that cholinated regions of ally, the quantitation of the number of transformants in this the pneumococcal cell wall can be removed by externally supplied autolytic enzyme in properly sensitized bacteria. * Ethanolamine-treated cells growing in defined medium re- The following experiment was performed. Bacteria that were covered their responsiveness to CF more slowly, 80-90 min after grown for over 20 generations in defined medium, containing the readdition of choline. In the enriched defined medium used in [1,2-'4C]ethanolamine at pH 6.8, were transferred to fresh the present experiments, activatability was recovered much growth medium, lacking aminoalcohol, at pH 8. After cessa- faster. The only difference between the two media is in the amino-acid composition: Cd contains only eight amino acids, tion of further incorporation of radioactive ethanolamine into Cden is enriched by the addition of 18 reagent grade amino acids, TCA-precipitable material, [methyl-PH]choline was added to in relative proportions corresponding to the composition of the culture. At frequent intervals, samples were removed to casein hydrolysate and in concentrations identical to those used monitor: (a) incorporation of choline into macromolecular in the C medium. Recovery of lysis sensitivity and dechaining material, (b) the reappearance of lysis sensitivity (by deoxy- followed the same time course in both Cd and Cde. media. Downloaded by guest on October 7, 2021 Proc. Nat. Acad. Sci. USA 68 (1971) DNA Uptake and the Cell Envelope 1851

system yields an underestimate of the number of competent cells, since during the experiment most of the bacteria are still present as members of very long chains*. A more realistic assessment of the number of competent cells was obtained by %) CF + Receptor autoradiography of the cells capable of irreversible binding to tritiated DNA. The insert in Fig. 2 shows an autoradiogram of cells from an experiment of identical design to that of Fig. 2 except that nonradioactive aminoalcohols were used. The cells shown on the photograph were taken from an ethanolamine-phase culture that received choline and CF for 20 min, (I) CF (Receptor) followed by a 10-min exposure to [3H]DNA (prepared from Bacillus subtilis), and finally treated with DNase. The ex- DNA is evident. t inhibitors of tensive binding of protein and RNA synthesis 5-ml portions of the culture in Fig. 2 were removed at the time of maximal lysis sensitivity (12 min after the readdition of choline) and "sensitized" to the pneumococcal autolytic ( ( ) ) CF (Receptor) - Agglutinin enzyme, either by brief heating (5 min at 650C) or by re- suspending the cells in a solution containing 20% sucrose and * chol ine 0.3 mM 2-mercaptoethanol. After short (20 min) treatment deprival of such cells with pneumococcal autolysin (a crude preparation containing 109 cell equivalents of autolysin), the bacteria CF (Receptor) - Agglutinin were fixed with glutaraldehyde, then with OSO4, embedded in at functioning wall growing zone Epon, sectioned, and examined by electron microscopy for sites of autolysin action. Examination of a large number of sections showed that sites of wall damage were restricted to FIG. 4. Model for the stages in the acquisition of component the equatorial area of the cocci (Fig. 3). These areas must state in pneumococcus. Crossed arrows indicate sites of in- represent the regions of the cell envelope where choline re- hibitions. incorporation has occurred since the ethanolamine-containing which continuous functioning of the growing zone of the cell ("old") cell envelope is resistant to lysis, even when incorpora- wall might be needed for DNA binding in genetic transfor- tion of small amounts of choline into the cell wall makes the mation. It may be, for instance, that DNA binding requires bacterium as a whole sensitive to enzyme- or DOC-induced a lateral movement of the cell envelope, driven by a growing disintegration. (After 15 min of [3H]choline addition to the zone. An alternative interpretation, which we propose as a ['4C]ethanolamine-labeled culture, deoxycholate released 90% working hypothesis, is that the growing zone(s) of the cell of the incorporated choline in TCA-precipitable, but non- envelope represents the actual site of entry of DNA molecules sedimentable form, while 90% of the macromolecular etha- during genetic transformation. This growing zone has re- nolamine (representing "old" hemispheres of the cell en- cently been localized at the pneumococcal equator (6). In Fig. velope) could still be sedimented by a 5-min centrifugation at 3, this surface area seems to correspond to the lysis-sensitive 8000-10,000 X g.) band located at the region of cell division. DISCUSSION The second type of conclusion to be drawn from the ex- The essence of the findings described in this report is the experiments concerns the mechanism by which the presence or tremely rapid response of cellular competence to biochemical absence of choline-containing macromolecules at the growing changes that occur in the growing zone of the cell envelope. zone of the cell envelope influences cellular competence for We would like to draw two types of conclusions from these genetic transformation. It is clear that choline deprivation experiments. The first one concerns the topography of events does not affect bacteria that have completed DNA uptake, involved in genetic transformation. The experiments indicate and that choline starvation does not eliminate cellular binding that removal of choline from the growth medium of choline- of CF. In fact, such bacteria can even show the type of ag- phase cells causes an extremely rapid decay in the ability of glutination characteristic of competent pneumococci (Table these bacteria to respond to molecules of the CF and to de- 2). However, this binding of CF molecules may be "incorrect" velop the transformable state. Reincorporation of choline in some way and choline-deprived cells cannot bind radio- into the cell envelope of ethanolamine-phase cells also causes active DNA in either "reversible" or "irreversible" form. rapid return (within less than 10% of the generation time) The observations described in this paper indicate the im- of the cellular response to CF; at least 90% of the newly in- portance of an additional, until now unsuspected, structure corporated choline molecules form an autolysin-sensitive in the cellular uptake of DNA molecules: namely, the growing zone that could be localized by electron microscopy at the zone of the cell envelope. Operationally, one can now dis- coccal equator. If one assumes that these biological effects are tinguish several stages in the acquisition of the competent linked to the behavior of the majority (over 90%) of in- state in pneumococci and the following simple model in an corporated choline molecules, then the two types of experi- attempt to connect these stages in a temporal sequence ments together suggest that the area of the cell surface rele- (Fig. 4): vant for DNA uptake is near to or within the area where new (a) Molecules of the CF (produced endogenously by the teichoic acid molecules incorporate into the cell surface, i.e., culture, most of it in an apparently autocatalytic manner) the growing zone(s) of the cell envelope. penetrate the cell wall and combine with receptors located on Clearly, one can imagine several different mechanisms by the plasma membrane. (b) A new protein "agglutinin" is Downloaded by guest on October 7, 2021 1852 Microbiology: Tomasz et al. Proc. Nat. Acad. Sci. USA 68 (1971) TABLE 2. CF binding, agglutinin formation, DNA binding, and genetic transformation in cells with modified cell envelopes

Cell-bound ['H] DNA bound (cpm/ml) Aminoalcohol CF (units Nuclease Transformants in the medium per ml) Agglutinin Total resistant per ml Cells in choline phase choline 0 0 7 8 <102 Cells in choline phase + CF choline 2-400 ++++ 10795 7300 2.3 X 105 Choline-phase cells, 20 min after shift to aminoalcohol- free medium + CF none 50-70 +++ 47 16 <102 Choline-phase cells, in limiting concentration of choline (see Fig. 1) plus CF, at 0 min (A)* limiting - - 10827 7331 2.3 X 10' at 10 min (B)* limiting - - 1509 1038 3.4 X 104 at 20 min (C)* limiting - - 79 48 5.0 X 10' at 40 min (D)* limiting - - 53 48 <30

Physiologically incompetent bacteria (i.e., cells grown at low pH) were used at cell concentrations of 1-2 X 107 per ml. CF was used at 100 units/ml. * A, B, C, and D refer to the time points, as indicated in Fig. 1. (-): undetermined.

deposited on the membrane and in the space between the bacilli. Both of these observations suggest a relationship membrane and cell wall. (c) Bacteria "loaded" with receptor- between DNA entry sites and topographic areas of the cell bound CF and agglutinin will still not be able to bind (re- surface similar to the ones found essential for competence versibly or irreversibly) DNA or undergo genetic transfor- in the case of pneumococci. mation-if events a and b are not accompanied-or possibly These findings also give further weight to the suggestion followed in time-by the incorporation of choline molecules that uptake of DNA molecules in genetic transformation may into the growing zone of the cell wall. This latter observation exploit some structural and/or functional features of the suggests that the CF receptors may be located within the cell division apparatus of bacteria (14). same topographic area as the growing zone of the wall. This work was supported by grants from the U.S. Atomic At present one cannot decide whether the cellular compo- Energy Commission and the National Institutes of Health. nent needed for competence is a sheet of cholinated teichoic 1. Tomasz, A., Science, 157, 694 (1967). acid. Alternatively, the presence of cholinated molecules 2. Rane, L., and Y. Subborow, J. Bacteriol., 40, 695 (1940). and CF together might influence competence in some in- 3. Tomasz, A., Proc. Nat. Acad. Sci. USA, 59, 86 (1968). direct way, e.g. via modulating the activity of an autolytic 4. Tomasz, A., A. Albino, and E. Zanati, Nature, 227, 138 enzyme in the growing zone (3, 5). Such an autolytic enzyme (1970). could unmask some underlying structure, possibly the CF-re- 5. Mosser, J. L., and A. Tomasz, J. Biol. Chem., 245, 287 (1970). ceptor complex on the plasma membrane, and make the ap- 6. Briles-Barak, E., and A. Tomasz, J. Cell Biol., 47, 786 proach of DNA molecules to such sites possible. From the (1970). experiments reported here, it is clear, however, that the mere 7. Tomasz, A., Bacteriol. Proc., 29 (1964). presence of autolysin and autolysin-sensitive cell wall sub- 8. Tomasz, A., and R. D. Hotchkiss, Proc. Nat. Acad. Sci. is sufficient make the bac- USA, 57, 480 (1964). strate at the growing zone not to 9. Tomasz, A., and J. L. Mosser, Proc. Nat. Acad. Sci. USA, teria competent, since the loss of competence and "activatabil- 55, 58 (1966). ity" which results upon the transfer of cells from choline to 10. Hotchkiss, R. D., in Methods in Enzymology, ed. S. P. ethanolamine medium is not accompanied by a loss of lysis Colowick and N. 0. Kaplan (Academic Press, New York, sensitivity (unpublished results). Similarly, the rapid ap- 1957), vol. 3, p. 692. 11. Ziegler, R., and A. Tomasz, Biochem. Biophys. Res. Com- pearance of lysis sensitivity during reincorporation of choline mun., 41, 1342 (1970). into ethanolamine cells is only accompanied by development 12. Tomasz, A., and E. Zanati, J. Bacteriol., March (1970), in of competence provided that the bacteria are also exposed to press. CF subsequent to the readdition of choline. 13. Tomasz, A., J. Bacteriol., 101, 860 (1970). Ranhand and associates that localized 14. Javor, G., and A. Tomasz, Proc. Nat. Acad. Sci. USA, 60, Recently, reported 1216 (1968). autolysis at the equatorial zone of streptococci could be in- 15. Bean, B., and A. Tomasz, J. Bacteriol., June (1971), in duced under certain experimental conditions by the addition press. of streptococcal CF (17). Evidence available in another 16. Lerman, L. S., and L. J. Tolmach, Biochim. Biophys. Acta, strain of streptococci (18) makes it likely that this zone cor- 26, 68 (1957). 17. Ranhand, J. M., C. G. Leonard, and R. G. Cole, J. Bac- responds to the growing area of the cell wall. In the Bacillus teriol., 106, 257 (1971). subtilis transformation system, it was suggested that DNA 18. Shockman, G. D., H. M. Pooley, and J. S. Thompson, J. uptake sites may be localized at the tips and centers of the Bacteriol., 94, 1525 (1967). Downloaded by guest on October 7, 2021