Vol. 171, No. 5JOURNAL OF BACTERIOLOGY, May 1989, p. 2609-26130021-9193/89/052609-05$02.00/0Copyright © 1989, American Society for Microbiology

Chromosomal Transformation of Escherichia coli recD Strainswith Linearized Plasmidst

CHRIS B. RUSSELL,* DAVID S. THALER,t AND F. W. DAHLQUISTInstitute of Molecular Biology, University of Oregon, Eugene, Oregon 97403

Received 10 March 1988/Accepted 9 February 1989

Wild-type Escherichia coil are resistant to genetic transformation by purified linear DNA, probably in partbecause of exonuclease activity. We demonstrate that E. coli containing a recD mutation could be easilytransformed by linearized plasmids containing a selectable marker. The marker was transferred to thechromosome by homologous recombination, whereas plasmid markers not in the region of homology were lost.

In contrast to some other bacteria, Escherichia coli is notreadily transformable by linear pieces of DNA (34). This isdue in part to degradation of the incoming DNA by intrac-ellular exonucleases (13, 43). Oishi and Cosloy (8, 9, 26)showed that E. coli strains lacking exonuclease V (ExoV) byvirtue of recB or recC mutations can be transformed bylinear DNA if they carry as well the sbcB (and sbcC)mutations that restore recombination proficiency to the recBrecC mutant. Similarly, recD mutants also are lacking inexonuclease activity, but unlike the recB recC mutants arerobust and recombination proficient without the requirementof additional mutations (5).

Jasin and Schimmel (14) and Winans et al. (44) used recBrecC sbcB strains to transform with linearized plasmids andhomologously replace a chromosomal allele with an allelethat had been modified. Yamaguchi and Tomizawa (46) andGutterson and Koshland (12) used strains defective in polAto transfer genes from plasmids to the chromosome. Strainscarrying recB, recC, and sbcB grow poorly and generallycarry a mutation in sbcC (22). In addition, strains carryingpolA are not robust (15). These considerations make thegeneral use of such strains unattractive.

Using the E. coli chemotaxis system genes as a chromo-somal target, we show here that recD strains can be trans-formed with linear DNA. The availability of recD alleles thatare mutant by virtue of a mini-Tet insertion facilitates theconstruction of transformable derivatives of strains of inter-est by one-step bacteriophage-mediated transduction.

MATERIALS AND METHODSStrains, media, and chemicals. The bacterial strains, plas-

mids, and phages used are listed in Tables 1 and 2. LBmedium and Tryptone broth were as described previously(2). The lactose phenotypes of the bacterial strains weredetermined on agar plates containing 5-bromo-4-chloro-3-indoyl-p-D-galactopyranoside (X-Gal) at 20 ,ug/liter and iso-propyl-3-D-thiogalactopyranoside (IPTG) at 1 mM.Swarm assay for chemotaxis. Colonies were screened for

chemotaxis ability on semisolid agar plates. Che+ strainsform large circular swarms within a few h at 30°C; strainsthat are Che- but still motile form fuzzy diffuse colonies;and nonmotile strains form sharp colonies. Che+ revertants

* Corresponding author.t Dedicated to Franklin W. Stahl on the occasion of his 60th

birthday.t Present address: Department of Biology, University of Utah,

Salt Lake City, UT 84112.

form swarms that bulge out from one side of the smallercolonies, analogous to sectored colonies on colorimetricindicator plates.Methyl esterase assays. The activity of the CheB protein, a

methyl esterase whose activity is modulated in response toattractants and repellents, can be measured in vivo by theamount of radioactive methanol released from cells that havebeen incubated with methyl-labeled radioactive methionine(16, 18, 19, 37, 41, 42). Steady-state cumulative evolution ofmethanol was measured by the method of Toews and Adler(41) as described by Kehry et al. (16). Stimulus-dependentesterase activity was measured in the flow apparatus asdescribed by Kehry et al. (6, 16, 17).

Transfer of chromosomal DNA to plasmids. The method fortransferring chromosomal DNA to plasmids is described indetail in the accompanying paper (30). Briefly, plasmidpCR176 was introduced into a strain, resulting in streptomy-cin sensitivity (11). Revertants to streptomycin resistancethat still carried ampicillin resistance were selected. Plas-mids were harvested from these revertants, and their restric-tion maps were examined to determine the structure of theDNA in the region of interest.

Screen for RecD phenotype. Our screen for RecD relies onthe inability of Red- Gam- A phage lacking Chi sequences toform large plaques on wild-type bacterial lawns (36). Weused phage carrying the deletion b1453 and the substitutionimm" to provide the Red- Gam- phenotype and allowplating on lambda lysogens, respectively. On rec+ strains,Chio versions of our phage make smaller plaques than doChi' versions. Chi' and Chi' Red- Gam- X form indistin-guishable large plaques on recD strains (5, 33). Methods forplating X were as described by Arber et al. (2).

Mid-log-phase cultures of each candidate, grown in Tryp-tone broth supplemented with thiamine, MgSO4, and 0.2%maltose, were mixed for 10 min with about 100 PFU of eitherMMS1443 or its Chi' derivative, SMR46. They were thenplated on Trypticase agar (BBL Microbiology Systems,Cockeysville, Md.) (2). Plates were incubated overnight at37°C and examined for plaque size.

Construction of bacterial strains. Transduction with phageP1 was performed as described by Miller (24). RP487 is astandard chemotaxis wild-type strain (28). D63 is a A indlysogen of RP487. CR101 was made by transduction of D63to Tetr from NK5992. D308 resulted from transduction ofCR101 to Arg+ and screening for the RecD phenotype. D300was constructed by transducing D63 to Tetr with P1 grownon SG1039 and screening for white colonies on platescontaining IPTG and X-Gal. D310 was made by transducing

2609

on January 11, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

2610 RUSSELL ET AL.

XbaI

TABLE 2. Plasmids and phages

Strain Relevant genotype Sourceor phenotype

pMC1871 lacZ M. Casadaban (31)pUC4-KSAC Kanr Pharmaciaa (3)pNO1523 rpsL+ (Strs) bla Pharmaciaa (11)pCR67 tar tap cheR cheB blapCR70 tar tap cheRB Y blapCR73 tar tap cheRB YZ blapRSW220 cheB bla R. StewartpCR173 pCR73 cheBY::KanrpCR174 pCR73 cheB::/acZ::Kanr::

che Y374pCR175 pCR73 cheB: :lacZ: :Kanr::

che Y375pCR176 pCR73 cheB::rpsL::cheYSMR46 X b1453 imm21 Chi+D S. M. RosenbergMMS885 X b1453 imm21 c Chio M. M. Stahl

' Pharmacia, Inc., Piscataway, N.J.

Transformation with linear DNA. The E. coli chemotaxisgene cluster was used as the chromosomal target for trans-formation experiments. Figure 1 shows the genetic map ofthat region.The transforming DNA was a linearized plasmid contain-

ing the kanamycin resistance gene from Tn9O3 replacingparts of the cheB and che Y genes (Fig. 1). Plasmid pCR173contains unique XbaI and EcoRI sites on the E. coli DNAflanking the chemotaxis genes. However, there is an addi-tional EcoRI site in lacZ in plasmids pCR174 and pCR175.The plasmid was cut with XbaI, EcoRI, or both, and theextent of digestion was determined by gel electrophoresis.Strains were made competent by treatment with cold CaCl2as described by Maniatis et al. (23) and exposed to thetransforming DNA, exactly as if introducing a circularplasmid. Kanamycin-resistant transformants were selectedand screened for ampicillin resistance. Ampicillin-resistanttransformants were generally discarded as being the proba-ble result of transformation by uncut plasmid or religation of

TABLE 1. Bacterial strains

Strain Relevant genotype or phenotype Source

D63 RP487 (Che+ lacY eda) J. S. Parkinson (28)DPB271 recD::mini-Tet S. Cohen (4)SG1039 A(lacIZYA) U169 zai::TnJO proC S. GottesmanNK5992 argA::TnlO N. KlecknerA535 recD1009 argA+ A. Chaudhury (5)YK4136 flaI eda+ Y. KomedaD300 D63 A(lacIZYA) U169 Tet+ Pro- P1 transduction

from SG1039D310 D63 A(lacIZYA) U169 P1 transduction of

D300 from D63D301 D310 recDJ903::mini-Tet P1 transduction

from DPB271CR101 D63 argA::TnlO P1 transduction

from NK5992D308 CR101 recD1009 P1 transduction

from A535

D300 back to Pro' and screening for Tets. D301 was theresult of transducing D63 to Tetr and screening for RecD.Construction of the other strains is described below.

Construction of plasmids. DNA manipulations were per-formed as described by Maniatis et al. (23). The plasmidscontain DNA from various stretches of the E. coli mecheoperon (Fig. 1). Plasmid pRSW220 carries the cheB gene.Plasmid pCR67 carries the tar, tap, cheR, and cheB genes.

Plasmid pCR70 carries the tar, tap, cheR, cheB, and cheYgenes and a small part of the cheZ gene. Plasmid pCR73carries the entire operon. Plasmids pCR174 and pCR175were constructed from pCR73 by inserting the lacZ gene

from pMC1871 between the HindIll site in cheB and the Sallsite in cheY and inserting a 1.3-kilobase fragment frompUC4-KSAC containing the kanamycin resistance gene intothe BamHI site adjacent to the Sall site. Plasmid pCR173was constructed in the same process but lost the lacZ gene.

Plasmid pCR176, made from pCR173, contains the strepto-mycin sensitivity gene from pNO1523 in place of the kana-mycin resistance gene.

chromosome

pRSW220

pCR67

pCR70

pCR73

p C R 1 7 3

pCR 1 74

pCR 1 75

FIG. 1. Structures of plasmids. The top line shows the structure of the chromosome in the chemotaxis region; the next four lines showthe portions of the chemotaxis genes on plasmids used in complementation experiments. pCR173, pCR174, and pCR175 are the plasmids usedin linear transformations of the chromosome; all three have the general features shown for pCR173.

EcoRI

e da

H ndI I Is4S II

tar ta ch/eR cheg cheYch/eZ

or / b/a

che8-/acZ knLa

che8-/acZ kana

p ,,

kand -,1-

J. BACTERIOL.

on January 11, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

LINEAR TRANSFORMATION OF E. COLI recD 2611

TABLE 3. Results of transformation with linearized plasmids

No. of transformants withgiven phenotype

Expta Plasmid Enzyme(s)Kan' Amp-/ Che-/Kan' Amp

1 pCR173 XbaI, EcoRI 490 428/490 10/102 pCR173 XbaI 980 13/122 10/103 pCR173 EcORI 1,050 5/135 4/54 pCR174 XbaI, partial 4 2/4 1/2

EcoRI5 pCR174 XbaI 6 3/6 3/36 pCR175 XbaI, partial 1 1 1

EcoRI7 pCR175 XbaI 26 9/26 8/9a All experiments were performed with approximately 108 E. coli D308

(recD1009); 120 ng ofDNA was used in experiments 1 to 3, and 10 ng ofDNAwas used in experiments 4 to 7.

the plasmid in the bacterial cell. Ampicillin-sensitive trans-formants were screened for the ability to swarm on soft agarplates as a test for chemotaxis. Isolates that lost the ability toswarm were defined as transformed at the chemotaxis locus.

RESULTS

Selection of transformants. The results of one series oftransformations are shown in Table 3. Two different sorts oftransformations were attempted. In the first type, the, plas-mid was cut on both sides of selected kan DNA (Table 3,experiment 1), liberating a fragment that had only DNA thatwas homologous to the target (in this case, the chromosomalche locus) on both sides of kan and no plasmid replicationorigin. In the second type (experiments 2 and 3), the plasmidwas cut on just one side of kan, resulting in DNA with thekanamycin gene sandwiched between two regions of homol-ogy with the chromosome and a nonhomologous piece ofDNA at one end that contained the ampicillin resistancegene and origin of replication. Transformation was found tobe efficient in both cases, with mildly increased efficiencywhen both ends of the homologous DNA were cleaved. Overseveral experiments, the efficiency of transformation by thesingle-cut DNA was 10 to 80o that of the double-cut DNA.Transformation was also possible when a cleavage siteoutside the region of homology was chosen. These resultsare consistent with the model of RecBC as a travelingrecombinase entering double-strand ends, as discussed be-low, and somewhat different from the case in organisms inwhich the transforming DNA enters the cell as a singlestrand (33, 34).Chromosomal transformation to kanamycin resistance is

less easily demonstrated when the linear donor DINA piececontains the origin of replication. Under these conditions,the probability is increased for transformation by a recircu-larized plasmid or by one that was never linearized, andplasmids were always recoverable from kanamycin-resis-tant, Che+ transformants. None of several Che- isolatesappeared to contain plasmids.The ability to transform with plasmids cut at only one site

is useful when one does not have convenient restriction siteson both sides of the marker. An example is the pair ofplasmids pCR174 and pCR175. These plasmids contain aconvenient XbaI site to one side of the cheB-lacZ fusion, butthe second EcoRI site, in lacZ, makes use of EcoRI moredifficult for providing a DNA fragment with bacterial DNAon both sides of the dominant selectable marker. Transfor-

mations were performed with plasmids cut only at the XbaIsite and also with plasmids cut at XbaI and partially cut atEcoRI (Table 3, experiments 4 to 7).The ratio of ampicillin-sensitive transformants to the total

number of transformants varied from 20 to 100% for XbaI-EcoRI (two scissions)-cut pCR173 and from 2 to 80% forXbaI- or EcoRI (single scission)-cut pCR173 or pCR174. Thesame DNA samples gave similar results in transformationsinto either D308 (recD1009) or D301 (recD1903), whichimplied that the variability was in the DNA preparationrather than the cell preparation. The frequency of chromo-somal transformation to kan at the che locus with XbaI-EcoRI-cut pCR173 was approximately 10-4 relative to nor-mal transformation with the same quantity of uncut circularplasmid.The frequency of linear transformation decreased with

decreasing lengths of homologous DNA flanking the kana-mycin resistance marker. The frequency of transformationwith pCR173 cut with EagI (1,900-base-pair homology) (25)was less than 20% of the frequency with XbaI-cut plasmid(4,300-base-pair homology). The frequency of transforma-tion with BssHII-cut pCR173 (300-base-pair homology oneither side) was less than 2% that of XbaI-EcoRI-cut plas-mid.

Analysis of chemotaxis functions in the transforniants.Eight of the transformants from Table 3 were examined inmore detail to determine the structure of the chromosome inthe region around cheB and che Y. The eight chosen werefour each from transformations with pCR174 and withpCR175. Each set of four had three isolates from transfor-mation with XbaI-cut plasmid and one from plasmid cut withXbaI and partially cut with EcoRI. If the linear transforma-tion event is conversion of the bacterial chromosome to theallele carried on the linearized plasmid, the inserted lac andkan sequences should partially delete the cheB and che Ygenes while leaving intact the remainder of the mecheoperon containing the tar, tap, cheR, and cheZ genes (Fig.1). The candidates were examined for swimming behavior,ability to be complemented by a plasmid carrying the entiremeche operon, ability to be rescued to Che+ by recombina-tion with a plasmid carrying DNA that spans the insert,presence of a functional cheR gene, and absence of afunctional cheB gene.The che Y gene codes for a protein responsible for causing

clockwise rotation of the bacterial flagella (7, 20, 29, 45).Disruption of the cheYgene results in bacteria whose flagellaturn only counterclockwise; therefore, the bacteria alwaysswim without tumbling (20, 27, 45). All 8 candidates, as wellas 16 other presumed transformants, exhibited no tumblingbehavior, which supported the idea that the cheY deletionhad been transferred to the chromosome.The eight candidates were transformed to ampicillin resis-

tance with intact plasmids pRSW220 (which carries just thecheB gene), pCR67 (which carries the cheR and cheB genesand part of the cheYgene but does not overlap the right-handend of the insert), pCR70 (which carries cheR, cheB, andche Y and does overlap both ends of the insert), and pCR73(which expresses cheR, cheB, che Y, and cheZ). The resultsfor all eight candidates were the same. Plasmid pCR73complemented the chromosomal transformants to approxi-mately wild-type levels on swarm plates, whereas the otherthree did not. The small swarms formed by the strainscarrying pCR70 resembled those of a AcheRB YZ straincarrying pCR70. In this case, however, the small swarmsproduced fast-swarming flares, which then swarmed as wildtype. When these fast-swarming variants were tested on

VOL. 171, 1989

on January 11, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

2612 RUSSELL ET AL.

kanamycin plates, 37 of 40 had lost kanamycin resistance.These results suggest that the insertion in the chromosome ispolar, eliminating expression of the cheZ gene, but thatrecombination from pCR70 into the chromosome restoresthe original structure. Presumably, the 10% that retainedkanamycin resistance had transferred the resistance gene toa plasmid, either by reciprocal recombination or by priortransfer to another of the multiple copies of the plasmid. Noreversion to Che+ was observed in the presence of plasmidpCR67 or pRSW220, which indicated that no homologyexists for recombination downstream from the insert, as onewould predict.

Expression of the cheR and cheB genes can be assayedbiochemically. The products of these genes are a methyltransferase and a methyl esterase, respectively (32, 35, 37),which are responsible for transferring methyl groups fromS-adenosylmethionine to the chemotaxis signal transducer-receptor proteins and from the transducers-receptors intomethanol (41). The continuous evolution of methanol re-quires that both gene products be present, and radioactivemethanol is released as a result of incubation with [methyl-3H]methionine. The lesion introduced into the chromo-

some disrupts cheB but not cheR. Therefore, a plasmidcomplementing cheB should be required to restore methanolproduction without any necessity to complement cheR. Alltested candidates showed increased release of volatile radio-activity and stimulus-dependent esterase activity when car-rying pRS220, which provides just the CheB function. Theenhancement of methanol evolution as a result of addingCheB demonstrates both the absence of a functional cheBgene and the presence of a functional cheR gene.Chromosomal structure of the transformants. Phage P1

transduction experiments (24) gave similar results for each ofthe eight candidates examined above. The 100% cotransduc-tion of LacZ+ and Che- with kanamycin resistance (2,000 of2,000 and 200 of 200, respectively) demonstrated that theplasmid structure remained basically intact when transferredto the chromosome. The 40% cotransduction of Eda+ (85 of205) confirmed that integration took place at the homologousche region of the chromosome, which is approximately 1 minaway from eda on the E. coli chromosome.The local physical structure of the chromosome was

determined by selecting plasmids that had acquired thetransformed segment of the chromosome and showing thattheir restriction maps were identical to the original structureof the plasmids used for the linear transformation (data notshown). Successful acquisitions were found by selecting forloss of the streptomycin sensitivity conferred by a wild-typerpsL gene in pCR176 replacing the lacZ and kanamycinresistance inserts in the plasmids used for the originaltransformations. Homologous recombination with thechemotaxis region can replace the rpsL gene with thecorresponding DNA from the chromosome. In this case, thatDNA includes the kanamycin resistance gene and the lacZgene from plasmid pCR174 or pCR175. In every candidate,loss of the rpsL gene resulted in plasmids acquiring the lacZand kanamycin resistance genes and never reconstructed theoriginal chemotaxis region. The apparent identity of therestriction maps of these plasmids with those of the plasmidsused for the initial linear transformation demonstrated thatthe chromosome had not suffered any gross deletions in thisregion as a result of allele transfer to the chromosome.

DISCUSSIONThe recB, recC, and recD genes of E. coli code for three

polypeptides that make up a single complex protein (1, 21;

reviewed in reference 10) containing several enzymaticactivities, including a helicase (39) and ExoV (21). In wild-type E. coli, the RecBCD complex promotes homologousrecombination, and mutations in recB or recC result indecreased recombination activity and increased sensitivityto UV light. ExoV activity is missing in extracts from eitherrecB or recC mutants. However, one subunit required forExoV activity, named ax by Lieberman and Oishi, is noteliminated by mutations in either gene (21). Chaudhury andSmith (5) found nuclease-deficient mutations in the vicinityof the recBC genes, and it was later shown that the ExoVactivity can be eliminated by mutations in recD while leavingthe bacteria proficient for recombination (1, 4, 5). Thiscreates a situation similar to that found for the recBC sbcBsbcC background, also ExoV- and Rec+, which Oishi andCosloy used for transformation of E. coli with linear DNA(8, 9, 26). Our results show that strains carryingjust the recDmutation can also be transformed by linear DNA. Specula-tions on the mechanisms of recombination in these two kindsof strains have been offered elsewhere (39a, 40).On a pragmatic level, the results shown here demonstrate

that E. coli carrying mutations in recD can be used for strainconstruction with transformation by linear pieces ofDNA, inthis case linearized plasmids. The recD mutation does notnoticeably interfere with bacterial growth and can easily betransduced into working strains either through linkage withargA (1) or by using a recD::mini-Tet (4). If the presence ofthe recD mutation is undesirable, the gene altered by trans-formation can be P1 transduced to the original, rec+ strain.For practical use, the transferred information should in-

clude a selectable marker such as the kan insert in theexperiments described here. Another marker, which can beeasily screened for (such as the amp gene in our plasmid),should be located adjacent to the plasmid origin of replica-tion in order to score those transformation events that do notinvolve gene conversion in the chromosome. When scoringchromosomal events is difficult, it is also preferable that theplasmid be cleaved between the desired marker and theplasmid origin of replication. This places the cleavage on theopposite side of ori from the marker that is screened against,e.g., amp. Under these conditions, genetic informationresiding in a plasmid can be transferred to the chromosomeat acceptably high efficiency, providing E. coli geneticistsconvenient access to some of the techniques possible withyeasts and some other bacterial species.

ACKNOWLEDGMENTS

We thank Frank Stahl for useful discussions, for inciting thiscollaboration, and for redaction of the manuscript.

This work was supported by Public Health Service grant Al-17808from the National Institute of Allergy and Infectious Diseases toF. W. Dahlquist and by Public Health Service grant GM 33677 andNational Science Foundation grant PCM 8409843 to F. W. Stahl.

LITERATURE CITED1. Amundsen, S. K., A. F. Taylor, A. M. Chaudhury, and G. R.

Smith. 1986. recD: the gene for an essential third subunit ofexonuclease V. Proc. Natl. Acad. Sci. USA 83:5558-5562.

2. Arber, W., L. Enquist, B. Hohn, N. E. Murray, and K. Murray.1983. Experimental methods for use with Lambda, p. 433-466.In R. W. Hendrix, J. W. Roberts, F. W. Stahl, and R. A.Weisberg (ed.), Lambda II. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

3. Barany, F. 1985. Two-codon insertion mutagenesis of plasmidgenes by using single-stranded hexameric oligonucleotides.Proc. Natl. Acad. Sci. USA 82:4202-4206.

4. Biek, D. P., and S. N. Cohen. 1986. Identification and charac-

J. BACTERIOL.

on January 11, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

LINEAR TRANSFORMATION OF E. COLI recD 2613

terization of recD, a gene affecting plasmid maintenance andrecombination in Escherichia coli. J. Bacteriol. 167:594-603.

5. Chaudhury, A. M., and G. R. Smith. 1984. A new class ofEscherichia coli recBC mutants: implications for the role ofRecBC enzyme in homologous recombination. Proc. Natl.Acad. Sci. USA 81:7850-7854.

6. Chelsky, D., N. I. Gutterson, and D. E. Koshland, Jr. 1984. Adiffusion assay for detection and quantitation of methyl-ester-ified proteins on polyacrylamide gels. Anal. Biochem. 141:143-148.

7. Clegg, D. O., and D. E. Koshland, Jr. 1984. The role of asignaling protein in bacterial sensing: behavioral effects ofincreased gene expression. J. Bacteriol. 162:398-405.

8. Cosloy, S. D., and M. Oishi. 1973. The nature of the transfor-mation process in Escherichia coli K12. Mol. Gen. Genet.124:1-10.

9. Cosloy, S. D., and M. Oishi. 1973. Genetic transformation inEscherichia coli K12. Proc. Natl. Acad. Sci. USA 70:84-87.

10. Cox, M. M. 1987. Enzymes of general recombination. Annu.Rev. Biochem. 56:229-262.

11. Dean, D. 1981. A plasmid cloning vector for the direct selectionof strains carrying recombinant plasmids. Gene 15:99-102.

12. Gutterson, N. I., and D. E. Koshland, Jr. 1983. Replacement andamplication of bacterial genes with sequences altered in vitro.Proc. Natl. Acad. Sci. USA 80:4894-4898.

13. Hoekstra, W. P. M., J. E. N. Bergmans, and E. M. Zuidweg.1980. Role of recBC nuclease in Escherichia coli transforma-tion. J. Bacteriol. 143:1031-1032.

14. Jasin, M., and P. Schimmel. 1984. Deletion of an essential genein Escherichia coli by site-specific recombination with linearDNA fragments. J. Bacteriol. 159:783-786.

15. Joyce, C. M., and N. D. F. Grindley. 1984. Method for deter-mining whether a gene of Escherichia coli is essential: applica-tion to the polA gene. J. Bacteriol. 158:636-643.

16. Kehry, M. R., T. G. Doak, and F. W. Dahlquist. 1984. Stimulus-induced changes in methylesterase activity during chemotaxis inEscherichia coli. J. Biol. Chem. 259:11828-11835.

17. Kehry, M. R., T. G. Doak, and F. W. Dahlquist. 1985. Sensoryadaptation in bacterial chemotaxis: regulation of demethylation.J. Bacteriol. 163:983-990.

18. Kleene, S. J., M. L. Toews, and J. Adler. 1977. Isolation ofglutamic acid methyl ester from an Escherichia coli membraneprotein involved in chemotaxis. J. Biol. Chem. 252:3214-3218.

19. Kort, E. N., M. F. Goy, S. H. Larsen, and J. Alder. 1975.Methylation of a protein involved in bacterial chemotaxis. Proc.Natl. Acad. Sci. USA 72:3939-3943.

20. Kuo, S. C., and D. E. Koshland, Jr. 1987. Roles of cheY andcheZ gene products in controlling flagellar rotation in bacterialchemotaxis of Escherichia coli. J. Bacteriol. 169:1307-1314.

21. Lieberman, R. P., and M. Oishi. 1974. The recBC deoxyribonu-clease of Escherichia coli: isolation and characterization of thesubunit proteins and reconstitution of the enzyme. Proc. Natl.Acad. Sci. USA 71:4816-4820.

22. Lloyd, R. G., and C.-Buckman. 1985. Identification and geneticanalysis of sbcC mutations in commonly used recBC sbcBstrains of Escherichia coli K-12. J. Bacteriol. 164:836-844.

23. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

24. Miller, J. H. 1972. Experiments in molecular genetics. ColdSpring Harbor Laboratory, Cold Spring Harbor, N.Y.

25. Mutoh, N., and M. I. Simon. 1986. Nucleotide sequence corre-sponding to five chemotaxis genes in Escherichia coli. J. Bac-teriol. 165:161-166.

26. Oishi, M., and S. D. Cosloy. 1972. The genetic and biochemicalbasis of the transformability of Escherichia coli K12. Biochem.Biophys. Res. Commun. 49:1568-1572.

27. Parkinson, J. S. 1978. Complementation analysis and deletionmapping of Escherichia coli mutants defective in chemotaxis. J.Bacteriol. 135:45-53.

28. Parkinson, J. S., and P. T. Revello. 1978. Sensory adaptationmutants of E. coli. Cell 15:1221-1230.

29. Ravid, S., P. Matsumura, and M. Eisenbach. 1986. Restorationof flagellar clockwise rotation in bacterial envelopes by insertionof the chemotaxis protein CheY. Proc. Natl. Acad. Sci. USA83:7157-7161.

30. Russell, C. B., and F. W. Dahlquist. 1989. Exchange of chromo-somal and plasmid alleles in Escherichia coli by selection forloss of a dominant antibiotic sensitivity marker. J. Bacteriol.171:2614-2618.

31. Shipira, S. K., J. Chou, F. V. Richaud, and M. J. Casadaban.1983. New versatile plasmid vectors for expression of hybridproteins coded by gene fused to lacZ gene sequences encodingan enzymatically active carboxy-terminal portion of P-galactosi-dase. Gene 25:71-82.

32. Silverman, M., and M. Simon. 1977. Identification of polypep-tides necessary for chemotaxis in Escherichia coli. J. Bacteriol.130:1317-1325.

33. Smith, G. R., and F. W. Stahl. 1985. Homologous recombina-tion promoted by Chi sites and RecBC enzyme of Escherichiacoli. BioEssays 2:244-249.

34. Smith, H. O., D. B. Danner, and R. A. Deich. 1981. Genetictransformation. Annu. Rev. Biochem. 50:41-68.

35. Springer, W. R., and D. E. Koshland, Jr. 1977. Identification ofa protein methyltransferase as the cheR gene product in thebacterial sensing system. Proc. Natl. Acad. Sci. USA 74:533-537.

36. Stahl, F. W. 1979. Special sites in generalized recombination.Annu. Rev. Genet. 13:7-24.

37. Stock, J. B., and D. E. Koshland, Jr. 1978. A protein methyl-esterase in bacterial sensing. Proc. Natl. Acad. Sci. USA75:3659-3663.

38. Szostak, J. W., T. L. Orr-Weaver, R. J. Rothstein, and F. W.Stahl. 1983. The double-strand-break repair model for recombi-nation. Cell 33:25-35.

39. Taylor, A., and G. R. Smith. 1980. Unwinding and rewinding ofDNA by the recBC enzyme. Cell 22:447-457.

39a.Thaler, D. S., E. Sampson, I. Siddiqi, S. M. Rosenberg, F. W.Stahl, and M. Stahl. 1988. A hypothesis: Chi activation ofRecBCD enzyme involves removal of the RecD subunit, p.413-422. In E. Friedberg and P. Hanawalt (ed.), Mechanismsand consequences of DNA damage processing. Alan R. Liss,New York.

40. Thaler, D. S., and F. W. Stahl. 1988. DNA double chain breaksin recombination of phage A and of yeast. Annu. Rev. Genet.22:169-197.

41. Toews, M. L., and J. Adler. 1979. Methanol formation in vivofrom methylated chemotaxis proteins in Escherichia coli. J.Biol. Chem. 254:1761-1764.

42. Van Der Werf, P., and D. E. Koshland, Jr. 1977. Identification ofa y-glutamyl methyl ester in bacterial membrane protein in-volved in chemotaxis. J. Biol. Chem. 252:2793-2795.

43. Wackernagel, W. 1972. An improved spheroplast assay forX-DNA and the influence of the bacterial genotype on thetransfection rate. Virology 48:94-103.

44. Winans, S. C., S. J. EUledge, J. H. Krueger, and G. C. Walker.1985. Site-directed insertion and deletion mutagenesis withcloned fragments in Escherichia coli. J. Bacteriol. 161:1219-1221.

45. Wolfe, A. L., M. P. Conley, T. J. Kramer, and H. C. Berg. 1987.Reconstitution of signaling in bacterial chemotaxis. J. Bacteriol.169:1878-1885.

46. Yamaguchi, K., and J. Tomizawa. 1980. Establishment of Esch-erichia coli cells with an integrated high copy number plasmid.Mol. Gen. Genet. 178:525-533.

VOL. 171, 1989

on January 11, 2021 by guesthttp://jb.asm

.org/D

ownloaded from