Vol. 146, No. 2JOURNAL OF BACTERIOLOGY, May 1981, p. 564-5700021-9193/81/050564-07$02.00/0

Transformation in Escherichia coli: Stages in the ProcessH. E. N. BERGMANS,* I. M. VAN DIE, AND W. P. M. HOEKSTRA

Department ofMolecular Cell Biology, State University of Utrecht, 3584 CH, Utrecht, The Netherlands

Received 30 October 1980/Accepted 11 February 1981

Transformation experiments with Escherichia coli recipient cells and linearchromosomal deoxyribonucleic acid (DNA) are reported. E. coli can be renderedcompetent for DNA uptake by a temperature shock (0°C -* 42'C 0°C) of therecipient cells in the presence of a high concentration of either Ca2+ or Mg2+ ions.Uptake of DNA into a deoxyribonuclease-resistant form, for which the presenceof Ca2" is essential, was possible during the temperature shock but appeared tooccur most readily after the heat shock during incubation at 0°C. When DNAwas added to cells that had been heat shocked in the presence of divalent cationsonly, DNA uptake also occurred. This suggests that competence induction anduptake may be regarded as separate stages. Under conditions used to inducecompetence, we observed an extensive release of periplasmic enzymes, probablyreflecting membrane damage induced during development of competence. Afterthe conversion of donor DNA into a deoxyribonuclease-resistant form, transform-ants could be selected. It appeared that incubation, before plating, of the trans-formation mixture in a medium containing high Ca2' and Mg2+ concentrationsand supplemented with all growth requirements increased the transformationfrequency. This incubation probably causes recovery of physiologically labilecells.

Uptake of free DNA by Escherichia coli canonly be achieved when the recipient cells havebeen made "competent," either by turning theminto spheroplasts (useful only in transfection, seereference 2) or by a heat shock in the presenceof Ca2" ions, as first reported by Mandel andHiga (14). Several modifications of the lattermethod are now widely used for transformationof E. coli with plasmid DNA (4, 7, 11, 12). It hasbeen shown that Ca2+-treated cells can be trans-formed also with chromosomal DNA (17, 29). Inthat case, a significant number of transformantswere obtained only when recipient cells wereused that lacked recBC nuclease activity andhad been rendered recombination proficient byan sbc mutation. The fact that Rec+ (recBCnuclease-proficient) cells yield very few trans-formants is probably due to effects that occurafter DNA adsorption and uptake. There is ge-netic evidence that partial degradation of donorchromosomal DNA by recBC nuclease is respon-sible for the reduction of transformation fre-quencies in Rec+ cells (9).The original protocols for transformation of

E. coli with chromosomal DNA (17, 29) essen-tially involved Ca2+ pretreatment of E. coli cellsat 0°C and heat shock of the cells in the presenceof Ca2" and purified donor DNA. Transformantswere subsequently selected on plates containingsodium citrate to bind excess Ca2+ ions.

We recently described a modified procedurethat yields a significantly higher transformationfrequency (20). The main modifications are: (i)no extensive pretreatment of the cells with Ca2",(ii) the presence of Mg2" ions as well as Ca2+during the whole transformation procedure, and(iii) the use of selective plates that contain Ca2"and Mg2+ ions, a low concentration of phosphate,and no sodium citrate. It is not clear why thesemodifications increase the transformation fre-quency. In this report we describe a series ofexperiments that give more insight into the com-plicated E. coli transformation system.

MATERIALS AND METHODSBacteria. All strains were E. coli K-12 derivatives.

PC0031 is a prototrophic strain. AM1095 (leu pdxAara car thi his trp recB21 recC22 sbcB15) is a trans-formable strain, described previously (10). AM1283 isa derivative of AM1170 (AM1095 rpoB) carrying plas-mid Rldrd-19 (Apr) (1).Media and buffers. Phosphate-based minimal

salts medium was as described previously (30). In theTris-based minimal salts medium, phosphate bufferwas replaced by Tris-hydrochloride buffer (1.2 x 10-'M, pH 7.0). In morpholinopropanesulfonic acid(MOPS)-based minimal salts medium, the phosphatebuffer was replaced by MOPS buffer (1.2 x 10-' M,pH 7.0). Media were solidified with 1.6% agar.

Growth requirements were added in the followingconcentrations (milligrams per liter): DL-leucine, 40;DL-arginine, 40; uracil, 20; pyridoxine, 10; DL-trypto-

564

on June 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

TRANSFORMATION IN E. COLI 565

phan, 40; thiamine, 5; L-histidine, 20; glucose, 2,000. InMOPS- or Tris-based medium, phosphate was added(final concentration, 2 x 10-4 M).

For induction of ,B-galactosidase, isopropyl-f-thio-galactoside, in a final concentration of 3 x 10-4 M, wasadded to the medium.DNA isolation. Donor DNA was isolated from

strain PC0031 as described by Cosloy and Oishi (5).Routinely isolated DNA has a molecular weight ofabout 4 x 107, as inferred from sedimentation onsucrose gradients with X DNA as a reference.Transformation procedure. The transformation

procedure used is a modification of the procedure ofReijnders et al. (20). An overnight culture of therecipient cells in phosphate-buffered minimal saltsmedium was diluted 1:20 into fresh minimal mediumand incubated with shaking at 370C until the celldensity measured in a Klett-Summerson photometer(650-nm filter) reached a value of 30. The culture waschilled, kept at 0°C for 15 mm, washed once with 0.25volume of 10 mM NaCl, and suspended in 0.05 volumeof 20 mM HEPES (N-2-hydroxyethylpiperazine-N'-ethanesulfonic acid) buffer (pH 6.0) at 00C. A portion(0.3 ml) of the cell suspension (10W viable cells) wasmixed with 0.1 ml of donor DNA (final DNA concen-tration, 60 ,ug/ml), after which 0.1 ml of a solutioncontaining 150 mM CaCl2 and 130 mM MgCl2 wasadded. The mixture was then carefully mixed on aVortex mixer and incubated for 10 min at 0°C, followedby 6 min at 420C and 60 min (or a different periodwhen indicated) at 00C. When indicated, uptake ofdonor DNA was terminated by incubating the mixturewith 400 ug of pancreatic DNase I (EC 3.1.21.1; Boeh-ringer Mannheim Corp.) per ml for 30 min at 00C. Insome experiments, the transformation mixture wasfurther incubated for 10 min at 370C, or the transfor-mation mixture was diluted 1:10 into MOPS-bufferedminimal medium containing growth requirements andwith CaC12 and MgCl2 added as indicated. Portions ofthe transformation mixture were plated on selectiveTris minimal medium supplemented with 0.01 MCaC12 and 0.01 M MgCl2. In all experiments, Leu+transformants were selected. The leu marker usedreverts at a rate yielding less than 10-8 revertants perviable cell. For each sample from which transformantswere selected, the number of viable cells was deter-mined by plating suitable dilutions on tryptone agarplates. During several steps in the transformationprocedure, there is a loss of viable cells. During theheat shock in the presence of Ca2' and Mg2' ions thereis a 25% decrease. Incubation after the heat shock for90 min at 00C causes another 30% decrease. Transfor-mation frequency was expressed as numbers of trans-formants per viable recipient cell.,-Lactamase assay. Suspensions of AM1283

[AM1095(Rldrd-19 Apr)] at the same cell density asused in our transformation procedure contain a veryhigh level of ,B-lactamase activity. For an assay using33-pl samples and an incubation period of 5 min (seebelow), the cell suspension should be diluted 10-foldso as to contain a measurable enzyme activity. As wedid not want to include additional dilution steps in theassay, but wanted to keep the cell density during theassays similar to that used in transformation experi-ments, we assayed /8-lactamase activities in AM1283-

AM1095 mixtures (1:10 ratio).Cultures of AM1283 and AM1095 were grown sep-

arately, harvested, suspended in HEPES buffer in thesame manner as in the transformation procedure andthen mixed. Part of the mixed-cell suspension wasconverted into spheroplasts by a modification (13) ofthe EDTA-lysozyme method of Osborn et al. (18). ,8-Lactamase was assayed as described by Nikaido et al.(16). The reaction mixture (1 ml) contained 10 mMsodium phosphate buffer (pH 7.0), 0.8 mM ampicillin(Enzyfarm), and 33-pl samples from suspensions ofwhole cells, supernatants of these after centrifugation,or spheroplasts. The reaction occurred for 5 min at28°C and was stopped by the addition of 1 ml of asolution containing 1 M acetic acid, 0.2 M sodiumtungstate, 0.2% soluble starch, 0.18 M I2, and 7.2 MKI. After 20 min of incubation at 28°C, the absorbanceat 623 nm was measured in a Bausch & Lomb Spec-tronic 700 spectrophotometer. A reaction mixture towhich the "stopping" solution was added before thetest sample served as a control for each incubation.

,B-Lactamase activities are expressed in arbitraryunits, being the difference in extinction at 623 nmbetween the control and test incubations.

P-Galactosidase assay. ,-Galactosidase was as-sayed in toluene-treated samples of suspensions ofwhole cells or directly in samples of supernatants aftercentrifugation, essentially as described by Miller (15).

RESULTSIn our basic transfonnation procedure, saline-

washed cells from an early-logarithmic cultureof AM1095, a multiauxotrophic recB21 recC22sbcB15 strain, were mixed at 00C with DNAisolated from a prototrophic strain; CaC12 andMgCl2 were added, and the mixture was shockedat 42°C for 6 min, followed by a 15-min incuba-tion at 00C. We observed that the frequency ofLeu+ transformants was enhanced by a factor of10 when the transfornation mixture was platedon selective Tris-based minimal salts mediumcontaining 10mM Ca2" and 10mM Mg2" insteadof on conventional phosphate-based minimalmedium. Since in this case uptake of DNA wasnot terminated with DNase, the observationcould mean that transfornation proceeds ormayeven start after the heat shock. Figure 1 and 2show initial experiments which further elucidatethe processes after the heat shock. In Fig. 1, thefrequency of transformants resistant to DNasetreatment is plotted versus the incubation timeat 00C after the heat shock. The DNase treat-ment if applied before the heat shock was suffi-cient to reduce the transformation frequency toless than lo-7 transformants per recipient cell.The results presented in Fig. 1 suggest thatduring incubation at 00C, DNA was still con-verted into a DNase-resistant form, presumablyby uptake of DNA into recipient cells renderedcompetent by the heat shock. Incubation of thetransfornation mixture at 00C for additional

VOL. 146, 1981

on June 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

566 BERGMANS, VAN DIE, AND HOEKSTRA

Lo 5go

,4w2-

0

C-)c3co

4-' 60 120 iomin inc. at 0°C

FIG. 1. Frequency of transformants resistant toDNase, at various times after the heat shock. CeUs ofAM1095 were cultivated, washed, and suspended inHEPES buffer (0°C); DNA, Ca2+, and Mg2+ wereadded as described in the text. The mixture wasincubated at 420C for 6 min and then futher incu-bated at 0°C. Time zero is the time of transfer from42 to 0°C. At times indicated, portions of the traits-formation mixture were incubated with DNase (30min, 0°C), after which the frequency of Leu4 trans-formants per viable cell was determined.

periods longer than 180 min (up to 24 h) elicitedno consistent increase in the yield of transform-ants. For practical purposes, we used an incu-bation period of 60 min at 00C in further exper-iments.

Ifone postulates that transformation proceedson suitable selective plates, it follows that plat-ing ofa transformation mixture implies one moreheat shock, since the mixture kept at 00C is thenincubated at 370C. Figure 2 illustrates the effectof a deliberately applied second heat shock 60min after the first heat shock and before plating.This second heat shock was applied to mixturesthat either had or had not been treated withDNase. In both cases, the second heat shock hada stimulating effect.We cannot exclude stimulation of DNA up-

take by the second heat shock when DNasetreatment was omitted; however, the stimulatingeffect observed in the experiment with DNasetreatment shows that there was at least an ad-ditional effect on some process(es) after the con-version ofDNA to a DNase-resistant form. Thestimulating effect was rather small but repro-ducible. In seven separate experiments, we ob-served such an effect. The average stimulationfactor was 2.2 + 0.6.Aspects of DNA uptake. If DNA is taken

up by competent cells in the period after theheat shock, it should be possible to obtain trans-formants when DNA is added to cells after theheat shock. The results presented in Table 1(lines A and B) prove that indeed transformants

were formed when DNA was added after theheat shock. The high transformation frequenciesobtained when DNA was added after the heatshock clearly show that cells were rendered com-petent for DNA uptake by a heat shock in theabsence of DNA. Lines C through I of Table 1show further aspects of competence develop-ment. Without a heat shock, there was no de-tectable amount of transformants (line C). Thepresence of either Mg2e or Ca2` ions was re-quired during the heat shock (lines E and F),since a heat shock in the absence of divalentcations was not effective (line D). From theresults presented in lines F and G, it appearsthat Ca`+ should at least be present during thepost-heat shock period. Finally, the results pre-sented in lines A, B, H, and I show that the cellswere better fit for DNA uptake when a coldshock was applied after the heat shock.Effect of the competence regimen on

membrane integrity. What could be the roleof the temperature shock in the presence ofdivalent cations for the development of compe-tence? One possibility is that such a conditionaffects the integrity of the inner membrane orouter membrane or both of E. coli. We testedthe effect of our competence regimen on outer

40,

20-

40-

20

E540_20 DNase._ s~~~~~~~~~

20OJ

transf.frequency11 x 105

4 x 105

K 3 x 10 5

1 x10DNaseI

30 60 90 120min.

FIG. 2. Effect of a second incubation at high tem-perature on transformation frequency. The figurshows graphic displays of the transformation proce-dures followed. Two transformation mixtures, com-posed as described in the text, were heat shocked andsubsequently incubated at 0°C for 60 min. In theexperiment diagrammed in the upper two graphs, onemixture was split and then either brought to 37°Cfor10 min or further incubated at 0°C. In the experimentpresented in the lower two graphs, DNase was addedto the other mixture, foUowed by a 30-min incubationat 0°C. The mixture was split and then incubatedeither at 370C for ID min or at 0°C. After the incu-bations, the frequency of Leu4 transformants perviable ceU was determined.

J. BACTERIOL.

on June 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

VOL. 146, 1981

membrane integrity by assaying the apparentactivity of /3-lactamase in suspensions ofAM1283, a derivative of AM1095 carrying plas-mid Rldrd-19 coding for a periplasmic /3-lacta-mase. Degradation of ampicillin by samples ofthe shocked cell suspension was assayed as ameasure for ,8-lactamase activity. As the perme-ability of the outer membrane is the rate-deter-mininmg step in the ft-lactamase assay of wholecells (16, 28), this assay can be used to measurechanges in outer membrane permeability. How-ever, if leakage of periplasmic enzymes, includ-ing,B-lactamase occurs, this would obviouslyenhance the apparent ,B-lactamase activity. Wechecked for leakage of ,B-lactamase by assayingthe ,t-lactamase activity in the supernatant aftercentrifugation of temperature-shocked cell sus-pensions. Table 2 shows the dramatic increasein fl-lactamase activity in cell suspensions thatwere treated by a cold -s heat -s cold shock in

the presence of divalent cations compared with

TRANSFORMATION IN E. COLI 567

celLs that were not shocked or were shocked onlyfrom 0 to 420C. Shortly after the temperatureshock, the increase was mainly due to increasedpermeability of the outer membrane, but after60 mim of further incubation at 00C, the totalincrease in apparent 86-lactamase activity wasdue to enzyme that had leaked out of the peri-plasm.

Apparently, major damage occurred to theouter membrane. The apparent fl-lactamase ac-tivity assayed 60 min after the cold-heat-coldshock amounted to 31% (ranging from 30 to 36%in different experiments) of the total Bi-lacta-mase activity (assayed in samples of the samesuspension but after the cells were convertedinto spheroplasts). Leakage of cytoplasmic en-zymes did not occur under these circumstances,as the f8-galactosidase activity in the superna-tants of the suspensions of isopropyl-ft-D-thio-galactopyranoside-induced cells varied from 0.5to 1% of the total activity (assayed in toluene-

TABLE 1. Influence of various competence regimens on transformation frequencySpecification of the transformation procedurea

Temp treatment' Addition, relative of shock, of: Relative transform-(OC) DNAd ca2+d Mge+ d

(A) 0 42 0 Before Before Before 1.0(B) 0 - 42 - 0 After Before Before 1.1(C) 0 - 0-s 0 Before Before Before <0.001(D) 0 42 0 Before After After <0.001(E) 0 42 0 After Before After 1.0(F) 0-. 42-O After After Before 0.6(G) 0 - 42-s 0 After None Before 0.001(H) 0 42 - 37 Before Before Before 0.1(I) 0 - 42 37 After Before Before 0.01

aDetails of the basic transformation procedure (line A) are presented in the text.b The frequencies are based on at least two different experiments. The absolute frequency in (A) amountedto 4 x 10-5.

'Cells kept at 00C were incubated at 42°C for 6 min and then brought at 00C or 370C as indicated. After 60min at the latter temperature, DNase was added, and the mixture was incubated for another 30-min period.d DNA, Ca2+, and Mg2e were added either before the transfer to 42°C or 3 min after the transfer from 42°Cto the lower temperature inldicated in the text.

TABLE 2. Influence ofheat shock on apparent /3-lactamase activity in suspensions ofAM12830,8-Lactamase activityb

Treatment t 0 rminC t = 60 mill

Cell suspension Supernatant Cell suspension Supernatant10 min, 000 - 6 min, 4200C 60 min, OOC 0.786 0.289 1.185 1.00510 min, O0C 6 min, 42°C - 60 min, 370C 0.111 0.043 0.136 0.03316 min, 00C 60 min, o0C 0.116 0.028 0.126 0.038

a Suspensions (in HEPES buffer) of AM1283 cells mixed with AM1095 cells were prepared as described inthe text; 0.4 ml of suspension was niixed with 0.1 ml of CaCl2 or MgCl2 (final concentrations, 30 and 26 mM,respectively) and heat shocked as indicated.

Expressed in arbitrary units, as described in the text.ct - 0, Time of chilling after incubation at 420C (or the comparable time if no heat shock was applied).

on June 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

568 BERGMANS, VAN DIE, AND HOEKSTRA

treated cells) irrespective of shock treatment.Leakage of fl-lactamase depended on the pres-ence of either Ca2" or Mg2e ions or a mixture ofthese (Table 3). In the absence of divalent cat-ions, only minor leakage occurred. Leakage ofperiplasmic endonuclease was also observed (de-tails not presented), which shows that the effectof the competence regimen is not restricted to,f-lactamase.

Thus, there is a clear, although not perfect,parallel between the effect of the competenceregimen on transformation frequency and onleakage of periplasmic enzymes.Processeii after DNA uptake. Figure 2 pre-

sents an initial experiment in which the increaseof transformation frequency after a second heatshock was demonstrated. The increase observedafter a second heat shock in the case in whichDNase treatment had been applied, implies thatprocesses occurring after DNA adsorption anduptake are stimulated. Figure 3 gives the resultsof an experiment in which the parameters ofpost-DNase incubation were examined in detail.Transformation was performed as described

in Table 1, line A. After DNase treatment, por-tions of the transformation mixture were diluted1:10 into MOPS-buffered minimal medium sup-plemented with growth requirements or into thesame medium additionally supplemented with10 mM CaCl2 and 10 mM MgCl2. After dilution,the mixture was further incubated at 370C. Dur-ing the incubation at 370C in the presence ofCa2' and Mg2e ions, the number of viable cellsremained constant, but the number of trans-formants increased by a factor of 2 to 3, resultingin a rise in the transformation frequency. Duringthe incubation at 370C in the absence of Ca2"and Mg2e ions, the number of viable cells dou-bled, whereas the number of transformants re-

TABLE 3. Influence of divalent cations on apparentf,-lactamase activity in suspensions ofAM12831

,B-Lactamase activityb

t =0 minc t = 60 minDivalent cations

Cell Sue- Celle- Super- sen Super-swe-natant u8e-natant

30 mM CaCl2 + 26 0.428 0.132 0.786 0.548mM MgC12

None 0.218 0.008 0.334 0.08656 mM CaCl2 0.508 0.136 0.808 0.60256 mM MgCl2 0.368 0.132 0.672 0.490

a Cell suspensions (in HEPES buffer) of AM1283 mixedwith AM1095 were prepared as described in the text; 0.4-mlamounts of cells were mixed with 0.1 ml of CaCl2 or MgCl2 or0.1 ml of both to the final concentrations indicated and heatshocked (10 min 0°C - 6 min 42C-- 60 min 0°C).

b Expressed in arbitrary units, as described in the text.c t -0, Time of chilling after incubation at 42°C.

LO 5-

c)3m 370cr 2

X ~~3'0 60 90 120min. inc.at 37 °C

FIG. 3. Effect ofpostincubation at high tempera-ture on transformation frequency. A transformationmixture, composed as described in the text, was heatshocked, incubated at 0°C for 60 min, and thentreated with DNase (30 min, 0°C). Portions of thismixture were diluted into MOPS buffered minimalmedium that was not supplemented or was supple-mented with 10 mM CaCl2 and 10 mM MgCl2. Timezero is the time of dilution into MOPS-buffered me-dium. At times indicated, the number ofLeu+ trans-formants in samples of the diluted mixtures wasassayed either on Tris-buffered selective plates con-taining 10 mM CaCl2 and 10 mM MgCl2 or on Tris-buffered plates without additional divalent cations.Transformation frequencies are expressed as Leu+transformants observed per viable cell. (-@*) In-cubation with added Ca2' and Mg2+; transformantswere plated on medium with Ca2' and Mg2+.(v--) Incubation with Ca2+ and Mg2+; trans-formants were plated on medium without Ca2+ andMg2+. (A A) Incubation without added Ca2+ andMg2+; transformants were plated on medium withCa2+ and Mg2+. (A---A) Incubation without Ca2+and Mg2+; transformants were plated on mediumwithout Ca2+ and Mg2+.

mained constant, resulting in a decrease in trans-formation frequency. When one of the growthrequirements (leucine) was omitted, we observedno increase of transformation frequency (datanot shown). After a 120-min incubation of thetransformation mixture in medium with Ca2+and Mg2+, it made no difference whether or notthe plating medium contained these ions.The second heat shock and the postincubation

period probably serve as a recovery stage, givingthe transformed cells an optimal chance to sur-vive.

DISCUSSIONThe main conclusion of our study is that it is

possible to render E. coli cells competent forDNA uptake in the absence of DNA by trans-ferring them from 0 to 420C, preferentially fol-lowed by a cold shock. This development ofcompetence is dependent on the presence ofCa2+ or Mg2+ ions. DNA subsequently added tothe competent cells is taken up into a DNase-

J. BACTERIOL.

on June 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

TRANSFORMATION IN E. COLI 569

resistant form during an extended period at 00C.The presence of Ca2" ions is essential duringDNA uptake. This fits with the ideas about therole of Ca2" ions in the transformation process(20-22, 24).

It should be mentioned that homogeneousdilution ofDNA into a solution containing Ca2"ions, e.g., addition ofDNA to Ca2+-heat-shockedcells, is difficult. We observed a tendency ofDNA molecules to "clump" in the presence ofCa2" and in the absence of cells. In our experi-ments, care was taken to ensure quick and thor-ough mixing ofthe transformation mixture. Onlyunder these conditions can reproducible resultsbe obtained.

Separation of a competence induction stageand a DNA uptake stage can be achieved by theaddition of DNA after the heat shock. Suchseparation might be advantageous to study de-tailed questions about E. coli transformation,e.g., the role of divalent cations in the inductionof competence and in DNA adsorption and up-take (cf. the hypothesis put forward by Grinius[8]).

In experiments in which DNA is added beforethe heat shock, the separation of competenceinduction and DNA uptake is not complete. Intransformation with plasmid DNA, uptake ofDNA occurs during the heat shock (O -. 4200)(11). Preliminary experiments in our transfor-mation system showed that uptake of chromo-somal DNA at a low concentration (<5 ,ug/ml),such as is typically used in plasmid transforma-tion, could not be enhanced by incubation at00C after the heat shock. A similar small amountof DNA was taken up during the heat shockwhen DNA was added at the high concentrationused in the experiment described here. In thiscase, however, most conversion of DNA into aDNase-resistant form occurred during incuba-tion at 00C after the heat shock (Table 1, linesA and H). During the induction of competence,cells are subjected to temperature changes andtreated with divalent cations. Such treatmentsmay cause changes in lipid conformation andinfluence the functional role of lipids in thebiological membrane (6, 19, 26, 27). The combi-nation of temperature changes and a high con-centration of either Ca2" or Mg2e could lead tospecific conformational changes in membranes,allowing DNA uptake. Whether other cationsthat are reported to be effective in transforma-tion, such as Ba2" (25), can replace Ca2" or Mg2ehas not been tested in our experiments.We observed leakage of periplasmic enzymes

during competence induction, probably as a con-sequence of outer membrane damage. There isa remarkable parallel between competence in-duction and leakage of periplasmic enzymes.

The results of ,B-lactamase assays showed thatabout 30% of the total f,-lactamase activity wasset free from the periplasm. This implies eitherthat each cell is rendered partially leaky or thatabout 30% of the celis release all f8-lactamaseactivity into the medium, whereas the rest stillcontain all ,8-lactamase activity. The parallelismbetween competence induction and leakage ofperiplasmic enzymes could be very indirect. If30% of the cells release all f,-lactamase activity,they may have been severely damaged by thecompetence regimen. The competent-cell frac-tion could include cells in which damage to theouter membrane is more subtle than in the cellsthat leak periplasmic enzymes. The competent-cell fraction is probably much smaller than 30%;from cotransformation data oftwo different plas-mids (11) or plasmid DNA and chromosomalDNA (3), a competent fraction ofabout 1% couldbe inferred. We observed that periplasmic en-donuclease also leaked from competent cells. Asthe results (not shown) of transformation of anendonuclease-deficient strain were similar to theresults shown for the endonuclease-proficientstrain, the presence or absence of endonucleaseapparently does not seem to influence the trans-formation process. Although we favor the ideathat outer membrane disruption is essential forinduction of competence in E. coli, it is quiteclear that the outer membrane is not the onlybarrier for uptake of chromosomal DNA. Inexperiments in which revertible E. coli sphero-plasts were used for transformation with chro-mosomal DNA, the transformation efficiencieswere similar to the (relatively low) frequenciesobtained with Ca2"-treated cells (23). If the cellsthat efficiently take up DNA have suffered somemembrane damage, it could be expected thatthey are more fragile than normal cells. We haveshown that postincubation ofthe transformationmixture at 370C in the presence of Ca2' andMg2e ions enhances the transformation fre-quency. This strongly suggests, as stated byKushner (12), that some cells that have takenup DNA are not yet ready for outgrowth onselective solid medium, especially when the me-dium does not contain Ca2' and Mg2" ions.We propose that in our transformation system

three stages can be distinguished: (i) inductionof competence by a heat shock followed by acold shock; during this stage, limited amounts ofDNA can be taken up; (ii) DNA uptake into aDNase-resistant form which can start or proceedduring further incubation at 00°; (iii) processesafterDNA uptake (tentatively considered recov-ery processes).

ACKNOWLEDGMENTSThis work was supported by the Netherlands Foundation

for Chemical Research (SON) with financial aid from the

VOL. 146, 1981

on June 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

570 BERGMANS, VAN DIE, AND HOEKSTRA

Netherlands Organization for the Advancement of Pure Re-search (ZWO).

E. J. J. Lugtenberg is acknowledged for advice about mea-surement of periplasmic leakage.

LITERATURE CITED

1. Beard, J. P., T. G. B. Howe, and M. H. Richmond.1972. Purification of sex pili from Escherichia colicarrying a derepresed F-like R-factor. J. Bacteriol.111:814-820.

2. Benzinger, R. 1978. Transfection of Enterobacteriaceaeand its applications. Microbiol. Rev. 42:194-236.

3. Bergman, H. E. N., D. H. Kooijman, and W. P. M.Hoekstra. 1980. Cotransformation of linear chromo-somal DNA and plasmid DNA in Escherichia coli.FEMS Microbiol. Lett. 9:211-214.

4. Cohen, S. N., A. C. Y. Chang, and L Hsu. 1972. Non-chromosomal antibiotic resistance in bacteria. Genetictransformation of Escherichia coli by R-factor DNA.Proc. Natl. Acad. Sci. U.S.A. 69:2110-2114.

5. Cosloy, S. D., and ML Oishi. 1973. The nature of thetransformation process in Escherichia coli K12. Mol.Gen. Genet. 124:1-10.

6. Cullis, P. R., and B. de Krudjff. 1979. Lipid polymor-phism and the functional roles of lipids in biologicalmembranes. Biochim. Biophys. Acta 559:399-420.

7. Dagert, M., and S. D. Ehrlich. 1979. Prolonged incuba-tion in calcium chloride improves the competence ofEscherichia coli cells. Gene 6:23-28.

8. Grlnus, L. 1980. Nucleic acid trnsport driven by iongradient across cell membrane. FEBS Lett. 113:1-9.

9. Hoekstra, W. P. IL, J. E. N. Bergmans, and E. MLZuidweg. 1980. Role of recBC nuclease in Eacherichiacoli transformation. J. Bacteriol. 143:1031-1032.

10. Hoekstra, W. P. ML, P. G. de Haan, J. E. N. BergmanD,and E. IL Zuidweg. 1976. Transformation in E. coliK12: relation of linkage to distance between markers.Mol. Gen. Genet. 145:109-110.

11. Humphreys, G. O., A. Weston, M. G. M. Brown, andJ. R. Saunders 1979. Plasmid transformation in Esch-erichia coli, p. 254-279. S. W. Glover and L. 0. Butler(ed.), Transformation 1978. Cotswold Press, Oxford.

12. Kushner, S. R. 1978. An improved method for transfor-mation of Eacherichia coli with col El derived plas-mids, p. 17-23. In H. W. Boyer and S. Nicosia (ed.),Genetic engineering. Elsevier/North-Holland Biomed-ical Pros, AmsterdanL

13. Lugtenberg, E. J. J., and R. Peters. 1976. Distributionof lipids in cytoplasmic and outer membranes of Esch-erichia coli K12. Biochim. Biophys. Acta 441:38-47.

14. Mandel, IL, and A. Higa. 1970. Calcium-dependent bac-teriophage DNA infection. J. Mol. Biol. 53:159-162.

15. Miller, J. H. 1972. Experiments in molecular genetics, p.352-355. Cold Spring Harbor Laboratory, Cold SpringHarbor, N.Y.

16. Nikaido, H., S. A. Song, L. Shaltiel, and M. Nurminen.1977. Outer membrane of Salmonella. XIV. Reducedtranamembrane diffusion rates in porin-deficient mu-

tants Biochem. Biophys. Res. Commun. 76:324-330.17. Oishl, M., and S. D. Cosloy. 1972. The genetic and

biochemical basis ofthe transformability ofEscherichiacoli K12. Biochem. Biophys. Res. Commun. 49:1568-1572.

18. Osborn, J., J. E. Gander, E. Parisi, and J. Carson.1972. Mechanism of assembly of the outer membraneof Salmonella typhimurium. Isolation and characteri-zation of cytoplasmic and outer membranes. J. Biol.Chem. 247:3962-3972.

19. Papahadjopulos, D., W. J. Vail, C. Newton, S. Nir, K.Jacobson, G. Poste, and R. Lazo. 1977. Studies onmembrane fusion. Im. The role of calcium inducedphase changes. Biochim. Biophys. Acta 465:579-598.

20. Reibnders, B., W. P. M. Hoekstra, J. E. N. Bergmans,and I. M. van Die. 1979. Transformation in Esche-richia coli, p. 113-121. In S. W. Glover and L. 0. Butler(ed.), Transformation 1978. Cotswold Press, Oxford.

21. Sabelnlkov, A. G., A. V. Ardeeva, and B. N. Ilyash-eno. 1975. Enhanced uptake of donor DNA by Ca2"treated Escherichia coil cells. Mol. Gen. Genet. 138:351-58.

22. Sabeildkov, A. G., and I. V. Domaradsay. 1979. Protonconductor vs cold in induction of Ca"+ dependent com-petence in Escherichia coli. Mol. Gen. Genet. 172:313-317.

23. Suzuki, M., and A. A. Szalay. 1979. Bacterial trandor-mation using temperature-sensitive mutants deficientin peptidoglycan synthesis. Methods Enzymol. 68:331-342.

24. Taketo, A. 1974. Sensitivity of Escherichia coli to viralnucleic acid. VIII. Idiosyncrasy of Cas2 dependent com-petence for DNA. J. Biochem. 75:895-904.

25. Taketo, A. 1975. Sensitivity of Escherichia coli to viralnucleic acid. X. Ba2 induced competence for transfect-ing DNA. Z. Naturforsch 30b:520-522.

26. Van Alphen, L, A. Verkleij, E. Burnell, and B. Lug-tenberg. 1980. 31P nuclear magnetic resonance andfreeze-fracture electron microscopy studies on Esche-richia coli. II. Lipopolysaccharide and lipopolysaccha-ride-phospholipids complexes. Biochim. Biophys. Acta597:502-517.

27. Van Alphen, L, A. Verkley, J. Leunissen-Bijvelt,and B. Lugtenberg. 1978. Architecture of the outermembrane ofEscherichia coli. Im. Protein-lipopolysac-charide complexes in intramembraneous particles. J.Bacteriol. 134:1089-1098.

28. Van Alphen, W., R. van Boxtel, N. van Selm, and B.Iugtenberg. 1978. Pores in the outer membrane ofEscherichia coli K12. Involvement of proteins b and cin the permeation of cephaloridine and ampicillin.FEMS Lett. 3:103-106.

29. Wackernagel, W. 1973. Genetic transformation in E.coi: the inhibitory role of the recBC DNAase. Biochem.Biophys. Res. Commun. 51:306-311.

30. Winkler, K. C., and P. G. de Haan. 1948. On the actionof sulfanilamide. XII. A set of non-competitive sulfanil-amide antagonists for Escherichia coil. Arch. Biochem.18:97-107.

J. BACTERIOL.

on June 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from