Proc. Natl. Acad. Sci. USAVol. 81, pp. 1035-1039, February 1984Biochemistry

Generation of a Tn5 promoter probe and its use in the study of geneexpression in Caulobacter crescentus

(cysteine auxotrophs/flagellar mutants/Tn5 neomycin phosphotransferase II)

VIVIAN BELLOFATTO, LUCILLE SHAPIRO*, AND DAVID A. HODGSONtDepartment of Molecular Biology, Division of Biological Sciences, Albert Einstein College of Medicine, Bronx, NY 10461

Communicated by Frank Lilly, October 31, 1983

ABSTRACT A promoter probe, TnS-VB32, was con-structed and placed in a P group R plasmid containing bacte-riophage Mu sequences, allowing transfer of the transposon tobacteria such as Caulobacter, Rhizobium, and Agrobacteriumwithout retention of the plasmid. The probe carries an alteredTn5 transposon that allows detection of chromosomal promot-er regions by virtue of acquired kanamycin resistance. A frag-ment of DNA containing the neomycin phosphotransferase II(NPT II) gene from TnS, lacking its promoter region but re-taining its translation initiation signal, was inserted into a Tn5derivative that lacked the entire NPT II gene and a large por-tion of the ISSOL sequence while retaining its ability to trans-pose. This TnS derivative also contained the intact tetracyclineresistance-encoding region of the transposon TnlO. Transposi-tion of the TnS-VB32 promoter probe into the Caulobactercrescentus chromosome generated auxotrophic and motilitymutants and Southern blot analysis of DNA from these mu-tants showed TnS-VB32 sequences in random-sized chromo-somal restriction fragments. Transcriptional regulation by ex-ogenous cysteine of NPT II gene expression was demonstratedin a cysteine auxotroph generated by Tn5-VB32 insertional in-activation. NPT II synthesis, measured by agar plate assays ofkanamycin resistance and by immunoprecipitation of the NPTII protein, was repressed in the presence of cysteine and dere-pressed in its absence. Severalfla- mutants were also isolatedby TnS-VB32 mutagenesis and shown to confer kanamycinresistance. Insertions within temporally regulated genes, suchas those involved in flagellar biosynthesis and chemotaxis func-tions, can now be used directly to monitor transcriptional reg-ulation from Caulobacter promoter sequences.

The morphological changes that occur during the Caulo-bacter crescentus cell cycle are regulated both temporallyand spatially. The repertoire of temporally expressed genes,those involved directly or indirectly in cellular differentia-tion, has begun to be defined (1-5). Several differentially ex-pressed proteins have been shown to be dependent on RNAsynthesis (6, 7), and, for example, the transcript for flagellinmRNA can be detected only at specific times in the cell cycle(ref. 8; M. Purucker, personal communication). A key ques-tion is whether it is differential transcript synthesis or mRNAavailability that is responsible for these observations.To search for temporally controlled genes in C. crescentus

that are regulated at the level of transcription we have con-structed a promoter probe that, when inserted into thegenome, can place the neomycin phosphotransferase II(NPT II) gene of the transposon Tn5 under the control of C.crescentus promoter sequences. The NPT II gene product isresponsible for conferring resistance to various aminoglyco-side antibiotics, including neomycin and kanamycin (9). OurTnS derivative (Tn5-VB32) lacks the transcription promotersequences of the NPT II gene but retains the translation start

signals for this gene as well as the partial left end and com-plete right end of the IS50 sequences required for transposi-tion. In addition, Tn5-VB32 contains the tetracycline resis-tance (Tcr) gene from the transposon TnJO, which permitsselection of the promoter probe in C. crescentus.We have shown that Tn5-VB32 can be introduced into C.

crescentus on a P group R plasmid (pJB9JI) that containsbacteriophage Mu sequences. This plasmid is unstable in C.crescentus by virtue of an undetermined property of Mu(10). Chromosomal insertions of Tn5-VB32 appear to occurin a random manner, as is the case with wild-type Tn5 inser-tions (10). We describe here one TnS-VB32 insertion thatgenerated a kanamycin-resistant (Kmr) cysteine auxotroph.In this case, Tn5-VB32 had transposed into a gene involvedin cysteine biosynthesis in such a way as to place the expres-sion of the NPT II gene under the control of a cysteine-regu-lated promoter.To identify transcriptionally regulated regions of the

genome involved in flagellar biogenesis (a cell cycle stage-specific event), we have also isolated several strains thatlack a flagellum as a result of promoter-probe insertional in-activation. Two of these strains were Kmr and thus ex-pressed the NPT II gene from Caulobacter promoter se-quences.

MATERIALS AND METHODSMaterials. 14C-labeled reconstituted protein hydrolysate

(21 Ci/mmol; 1 Ci = 37 GBq; Schwarz mixture) was fromSchwarz/Mann. Kanamycin sulfate, gentamycin, and tri-methopnrm were from Sigma and tetracycline hydrochloridewas from Calbiochem. Spectinomycin hydrochloride wasfrom Upjohn. Pancreatic RNase and DNase I were fromWorthington and calf intestine alkaline phosphatase wasfrom Boehringer Mannheim. DNA polymerase I, DNA li-gase, and all restriction endonucleases were from BethesdaResearch Laboratories and [a-32P]dCTP (3000 Ci/mmol)was from Amersham.

Bacterial Strains and Genetic Manipulations. C. crescentusand Escherichia coli strains and plasmids used in this studyare listed in Table 1. C. crescentus wild-type CB15 and mu-tant strains were grown at 30°C in PYE medium (19) or inmodified minimal M2 glucose medium (2, 20). E. coli wild-type and mutant strains were grown in LB medium (21) or inM9 minimal media (22) supplemented with 40 ,ug of the ap-propriate amino acids per ml. Standard methods were usedto transform E. coli C600, HB101, and AEE410 (12). Gener-alized transductions using Caulobacter phage OCr30 wereperformed as described (23), except that the selection was ontetracycline "slant" plates. Fifty milliliters of PYE agar con-

Abbreviations: NPT II, neomycin phosphotransferase Il; Tc, tetra-cycline; Km, kanamycin; Gm, gentamycin: Sp. spectinomycin; r,resistant (resistance); S, sensitive (sensitivity).*To whom reprint requests should be addressed.tCurrent address: Department of Biochemistry, Stanford UniversitySchool of Medicine, Stanford, CA 94305.

1035

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

June

21,

202

1

1036 Biochemistry: Bellofatto et al.

Table 1. List of strains and plasmidsStrain

hsdR-, hsdM', thy-, thr-, leu-, pro-, lacY-argG6, metBi, pfka300::Mu, lacYl, galt6, galA-50, galC49, srlC49, xyl-7,

mtlA2, rpsL104 (Strr), recAl, tonA, txs-6, supE44F, F-hsdR-, hsdM-, recA13, supE44, proA2, leuB6, thi-l, ara-14, galK2,

xyl-5, mth-J, lacZ4, rpsL20 (Strr), X-, F-tyrA2, pyrD34, trp45, thyA33, thi-l, galK35, malAl, xyl-7, mtl-2, recAl,rpsL118 (Strr), F-

cys- (pJB9JI)pro- met- Nalr (pJB4JI)polAJ2 (ts), thyA36, deoC2, lacZS3, rha-S, rpsL151, (Strr), X-, F-KL202::Mu ctsDF456 (pJB4JI)5022::Mu c+HB101 (pVB32)C600 (pVB32)AEE399 (pJB9JI)AEE410 ColS Tcr Gmr SprDF456 (pPV2)

J. Sninsky; ref. 11P. Silverman

Ref. 12

P. Silverman

A. W. B. Johnston; ref. 13B. ElyRef. 14Mu et' lysogen1830 x DF456Mu lysogenTransformation of HB101Transformation of C6001799E x AEE399*Transformation of AEE410 with pVB32AEE420 x DF456

C. crescentus CB15AE5000SC451AE4001AE7000AE7001AE7002AE7003AE7004AE7005AE7006AE7007AE7009AE7010

Wild-typeproC104proCJ04, fadA101: :Tn5proCJ04, cys-501::Tn5-VB32proC104, cys-502::Tn5-VB32proCi04, phe-501::Tn5-VB32proCIO4, pur-501: :Tn5-VB32proC104, fla-501: :Tn5-VB32proCJ04, fla-502::TnS-VB32proCJ04, fla-503: :TnS-VB32proCJO4, fla-504::TnS-VB32proC104, zzz-501::TnS-VB32 (Tcr, Kms)proC104, zzz-502: :TnS-VB32 (Tcr, Kms)

This laboratoryB. Ely; ref. 15M. O'Connell; this laboratoryAEE431 x SC451AEE431 x SC451AEE431 x SC451AEE431 x SC451AEE431 x SC451AEE431 x SC451AEE431 x SC451AEE431 x SC451AEE431 x SC451AEE431 x SC451

ColEl: :Tn5ColEl: :TnS-341ColEl: :Tn5-VB32pPHlJI :MupJB9JI: :Tn5pPHlJI: :Mu: :TnS-VB32

R. C. Johnson; ref. 16R. C. Johnson; ref. 17This studyA. W. B. Johnston; ref. 13B. Ely; ref. 18This study

zzz = Insertion into unknown gene.*Conjugation.

taining 3 ,ug of tetracycline per ml was poured into a 15-cm(diameter) Petri dish and the dish then was set at a slant.After the agar had set, the dish was set horizontally and an-other 50 ml of PYE agar containing no antibiotic was added.This was done to ensure that the cells came into contact withthe tetracycline slowly so as to allow the induction of the Tcrgene. Colonies were picked well away from the slim crescentof confluent growth at the side of the dish containing theleast amount of antibiotic. Plasmid conjugation was carriedout as described by Ely (18).

Plasmids. pRZ102 and pRZ341 (Fig. 1) were obtained fromR. C. Johnson (16, 17). The newly constructed plasmid,pVB32, was isolated after transformation of E. coli HB101and selection of colonies on plates containing tetracycline(20 ug/ml). Plasmid pVB32 was then transformed into E.coli C600 and the modified plasmid DNA was used to trans-form the E. coli AEE410 (polAt') that contained pJB9JI, aself-transmissible P-type R plasmid encoding gentamycinresistance (Gmr), spectinomycin resistance (Spr) and con-taining the bacteriophage Mu (13). A Gmr, Tcr colony,AEE420, was mated with DF456 under selection for Gmr,Tcr derivatives. One of the rare transconjugants, AEE431,

was examined for the presence of a TnS-VB32 derivativeof pJB9JI (designated pPV2) by mating this strain withAEE147. One hundred percent cotransmissibility of Gmrand Tcr was found. The presence of the recAl allele in DF456coupled with the 100% cotransmissibility of the Gmr, Tcrphenotype in subsequent matings ruled out the possibility ofthe Tcr phenotype being derived from chromosome mobili-zation of strain AEE420 concomitant with pJB9JI movementand implied that the TnS-VB32 had transposed onto pJB9JIto form pPV2.

Transfer of TnS-VB32 into C. crescentus. pPV2 was trans-ferred from E. coli AEE431 to C. crescentus SC451 by con-jugation. The transferred pJB9JI vector containing Tn5-VB32 (pPV2) was not stable in C. crescentus because it con-tained Mu (10) and thus all Tcr, Gm-sensitive (Gms), spscolonies were presumed to be the result of insertion of Tn5-VB32 into the chromosome. This was confirmed for all colo-nies used in this study by transduction and Southern blotanalysis. All SC451 Tcr exconjugants were tested for theirlevel of resistance to kanamycin, potential auxotrophy, andmotility. Motility was assayed as the ability to form swarmson semisolid PYE plates (0.3% agar). Transposition frequen-

Genotype Source

E. coliC600DF456

HB101

KL262

1799E18305022AEE147AEE174AEE399AEE401AEE403AEE410AEE420AEE431

PlasmidpRZ102pRZ341pVB32pJB9JIpJB4JIpPV2

Proc. NatL Acad Sci. USA 81 (1984)

Dow

nloa

ded

by g

uest

on

June

21,

202

1

Proc. NatL Acad. Sci. USA 81 (1984) 1037

cy was measured as the number of Kmr or Tcr C. crescentuscolonies compared to the number of Kms or Tc' coloniesisolated after mating AEE174 or AEE431 with SC451.

RESULTSConstruction of TnS-VB32 and Its Introduction into C. cres-

centus. The plasmid pVB32 (Col EI::TnS-VB32), containinga promoterless Kmr gene, was constructed as described inMaterials and Methods and is shown in Fig. 1. The TnS-VB32 insert contains the most leftward 53 bases joined to themost rightward 19 bases of ISSOL. This is followed by anintact NPT II gene, lacking its promoter but beginning withits translation initiation sequences. The insert also containsthe Tcr gene of TnJO and the complete IS50R sequence. Ithas been shown in E. coli that only the most leftward 23bases of IS50L are essential for transposition (17).To introduce TnS-VB32 into C. crescentus, it was first

transposed onto a Mu-containing P-type R plasmid, pJB9JIcarried in E. coli, as described in Materials and Methods.The E. coli strain AEE431, carrying pPV2 (pJB9JI::Tn5-VB32), was then mated with C. crescentus SC451 and Tcrtransconjugants were selected. pJB9JI and its derivatives areunstable in C. crescentus presumably because Mu sequenceson these plasmids are not tolerated (10). Thus, Tcr, Gms, spstransconjugants contained TnS-VB32 inserted into the C.crescentus chromosome. The transposition frequency ofTnS-VB32 was found to be 1.6 x 10-8 compared to 1.6 xi-7 for wild-type TnS.Characterization of C. crescentus Strains Containing TnS-

VB32. Of 600 SC451 Tcr exconjugants, -1% was found to beauxotrophs and s1% had lost motility. A similar proportionof mutant phenotypes was obtained when wild-type Tn5 wasused for transposon mutagenesis (10). To ascertain whetherthe auxotrophic and motility mutants were in fact due to theinsertion of TnS-VB32 and not to spontaneous mutations,the Tcr marker was transduced into wild-type C. crescentus,AE5000, and the mutant phenotypes were scored. In all cas-es, Tcr, Kmr levels and the mutant phenotype cotransferred100% of the time. In addition, Southern blot analysis wasperformed to identify the presence of TnS-VB32 in the chro-mosome. BamHI digests of chromosomal DNAs isolatedfrom several of the mutants were probed with nick-translat-ed X::Tn5 DNA (12, 25). BamHI does not cut within TnS-VB32. A single band that differed in mobility from strain to

FIG. 1. Schematic of the construction of the Tn5-VB32 promoterprobe. E. coli plasmid pRZ341 (17) was digested with BamHI andSal I, treated with calf intestine alkaline phosphatase, and then ligat-ed to pRZ102 (16) DNA that had been treated with Bgl II and Sal I.pVB32 contains intact tetracycline resistance (tetA) and tetracyclinerepressor (tetR) genes from TnJO (24) and the NPT II gene with itstranslation start site but lacking its promoter sequence. The arrowindicates the direction of NPT II gene transcription. Black boxes areIS50 sequences; open boxes are portions of the unique sequences ofTn5. The zigzag line is 275 base pairs of pBR322, hatched boxes arethe Tcr-encoding gene, and the thin lines are colEl plasmid se-quences.

strain hybridized to the probe in all chromosome digests test-ed (data not shown).

Seventy-seven percent of the Tcr, Gms, sps exconjugantswere resistant to kanamycin at concentrations of 50 ,Ag/ml.Because TnS-VB32 does not contain its own NPT II promot-er but does contain its own translational start signal, thesetransconjugants were Kmr due to the presence of read-through transcription from a C. crescentus promoter. TheTcr exconjugants that were Kms (at 50 ,ug of kanamycin perml) were shown by Southern blot analysis to contain Tn5sequences. The observed Kms suggests that in these strainsthe TnS-VB32 sequences were not adjacent to C. crescentuspromoters. Therefore, the Tcr, Kms strains confirm thatTnS-VB32 lacks sequences that can be used for the tran-scription of the NPT II gene. One of the auxotrophic strains(AE7002, phe-501::Tn5-VB32) and two of the fla- strains(AE7005 and AE7007) are included in this group (Table 2).These three strains were also subjected to transductionalanalysis as were AE7009 and AE7010, two Kms strains withno detected auxotrophic requirements or motility dysfunc-tion (Table 1). Again, in all cases, Tcr, level of Kmr, and,where applicable, the genetic lesion cosegregated at a fre-quency of 100%.

Cysteine Concentrations Regulate Kmr and NPT II Synthe-sis in the cys Auxotroph AE7000. A Tcr cysteine auxotroph,AE7000, was isolated by TnS-VB32 mutagenesis and shownto be resistant to kanamycin. The mutant phenotype and theTcr and Kmr were shown to cotransduce. Because cysteinehas been shown to regulate the cysteine biosynthetic path-way in E. coli and Salmonella (26), we determined whetherC. crescentus cys auxotroph AE7000 carried the promoterprobe TnS-VB32 in a repressible gene of the cysteine biosyn-thetic pathway. Accordingly, AE7000 was supplementedwith various concentrations of cysteine and the level of Kmrwas determined in each case (Fig. 2). Widely varying con-centrations of cysteine did not affect the Km' of the parentstrain SC451. However, strain AE7000 exhibited changes inKmr when the levels of exogenous cysteine were varied (Fig.2). Low concentrations of cysteine resulted in high levels ofresistance to kanamycin. High concentrations of cysteine re-sulted in decreased levels of Kms, suggesting that cysteinerepressed the expression of the NPT II gene. Thiosulfatealso decreased levels of Kmr; however, higher concentra-tions of this supplement were required. Control experimentsusing the parent strain carrying a wild-type TnS showed thatcysteine concentrations did not affect the expression of theNPT II gene when it was under the control of its native pro-moter.To confirm that variations in cysteine concentrations were

directly affecting the expression of the NPT II gene responsi-

Table 2. Level of Kmr of strains carrying Tn5 and Tn5-VB32Strain Genotype Kmr level, ,ug/ml*

AE7000 cys-501::Tn5-VB32 1000AE7001 cys-502::Tn5-VB32 500AE7002 phe-501::TnS-VB32 20AE7003 pur-501: :Tn5-VB32 500AE7004 fla-501::Tn5-VB32t 1000AE7005 fla-502::Tn5-VB32t 20AE7006 fla-503::Tn5-VB32t 100AE7007 fla-504::Tn5-VB32t 20AE4001 fadAJOl::Tn5 2000SC451 proC104 10

All of these strains except AE4001 and SC451 are Tcr (3 ,g/ml),Gms (20 ,ug/ml), and sps (50 Ag/ml).*The highest kanamycin concentration that permitted growth onPYE plates.

tAll Fla- strains lacked a visible flagellum, as determined by elec-tron microscopy.

Biochemistry: Bellofatto et aL

Dow

nloa

ded

by g

uest

on

June

21,

202

1

1038 Biochemistry: Bellofatto et al.

SC451 AE4001 AE7000__ e-0.04 0.8 004 0.8

CYSTEINE (mM)004 08

FIG. 2. Agar plate assay for detection of Kmr. Cultures of C.crescentus SC451, AE4001, and AE7000 were grown on PYE agarplates supplemented with low or high concentrations of cysteine asindicated. Whatman 3MM paper filters, 0.5 cm in diameter, wereplaced on the plates and then spotted with 8 ul of kanamycin at 100mg/ml. The plates were incubated for 2 days at 30'C and all plateswere photographed at the same magnification. The dark ringsaround the filters reflect different extents of Kms.

ble for the observed Kmr, we measured the synthesis of theNPT II protein using anti-NPT II antisera (Fig. 3). StrainAE7000 was labeled with 14C-labeled reconstituted proteinhydrolysate under the conditions described in the legend toFig. 3. Immunoprecipitates of labeled cell extracts showedthat the presence of cysteine inhibited the production of theMr 25,000 NPT II protein, whereas its absence resulted inthe expression of the NPT II protein. Although thiosulf'atewas present at a concentration barely sufficient to fulfill thecysteine requirement of the cells in minimal media, it wasbelow the level required to repress the expression of the cys-teine gene.The cysteine-controlled expression of Kmr and NPT II

synthesis in a strain containing NPT II-encoded sequenceswithin a gene involved in cysteine biosynthesis demonstratesthat regulation of gene expression can easily be detectedwith the Tn5-VB32 probe.TnS-VB32 Insertion into Genes Involved in Flagellar Bio-

genesis. In our initial isolation of C. crescentus strains con-taining TnS-VB32, four nonmotile mutants were found,which were shown by electron microscopy to lack a flagel-lum. These mutants were of particular interest because fla-gellar genes are differentially expressed during defined peri-ods in the cell cycle. The levels of Kmr were determined foreach of the Tn5-VB32-induced Fla- mutants (Table 2).

+CYSTEINE +TH IOSULFATE -

_-NPT1

10 20 45 10 20 45 10 20 45TIME AFTER ADDITION OF SUPPLEMENT (MIN)

FIG. 3. Proteins immunoprecipitated by anti-NPT II antibodyfrom cell extracts of C. crescentus AE7000. Cultures initially grownin PYE containing 0.3% glucose, 0.5 mM CaCl,, and a high concen-tration of cysteine (0.7 mM) were washed and resuspended in mini-mal BMG medium containing the same supplement with 4 MCi of14C-labeled reconstituted protein hydrolysate per ml. After 25 min at30'C the culture was washed, divided into thirds, and resuspendedin minimal BMG medium with 4 ,uCi of "4C-labeled reconstitutedprotein hydrolysate containing either cysteine (0.7 mM), thiosulfate(0.8 mM), or no supplement, as indicated. These cultures were incu-bated at 30'C, aliquots (5 ml) were removed at the indicated times,and immunoprecipitates of cell extracts were prepared as described(27). Proteins were separated by electrophoresis through 12%NaDodSO4/polyacrylamide gels and visualized by autoradiography.The arrows indicate the Mr 25,000 NPT II protein (28).

There appeared to be three different levels of Kmr observed.AE7004 (fla-501) had a high level of Kmr comparable to thatseen with AE7000. This level of resistance is somewhat lessthan that observed in strains containing wild-type TnS.AE7005 (fla-502) and AE7007 (fla-504) were sensitive to lowconcentrations of kanamycin, as was AE7002 (phe-SOl),comparable to strains devoid of TnS (SC451). The Kms, Tcrphenotype is expected when transposition of TnS-VB32 sim-ply disrupts the gene by insertion in an orientation oppositethat of the normal direction of transcription. An intermediatelevel of Kmr was observed for strain AE7006 (fla-503). Thismay be the result of a weak promoter or may indicate thatthis Fla-related promoter is used only during a portion of thecell cycle.

DISCUSSION

Transposon mutagenesis has been effectively used to gener-ate and map mutations in C. crescentus (10, 15, 23). Thetransposition of TnS onto the C. crescentus chromosome isfacilitated by its introduction into the cell on a Mu-contain-ing P group R plasmid (10). TnS insertion mutations havebeen shown to be random (10) and have been used to isolatecell cycle-specific genes (2). We described here the construc-tion of a TnS-derived promoter probe carrying a promoter-less Kmr gene and an intact Tcr gene, TnS-VB32, that can beused (i) for the detection of regulated transcriptional startsites in the C. crescentus chromosome and (ii) to facilitatethe mapping and isolation of insertionally inactivated genestagged with the Tcr marker. TnS-VB32 was shown to trans-pose onto the C. crescentus chromosome when introducedon plasmid pPV2, which is derived from pJB9JI, in a similarmanner to the introduction of wild-type TnS. Because TnSon pJB4JI (derived from JB9JI) can be introduced into suchdiverse genera as Acinetobacter, Agrobacterium, Alcali-genes, Azospirillum, Erwinia, Pseudomonas, and Rhizobium(29), Tn5-VB32 could be similarly introduced into thesestrains to address questions of transcriptional regulation.Although the transposition frequency of Tn5-VB32 into C.

crescentus was found to be 1/10th as much as for wild-typeTnS, auxotrophic and motility mutants generated by TnS-VB32 insertion occurred at frequencies similar to thosefound with TnS. The Tcr phenotype, the auxotrophic or mo-tility phenotype, and the Kmr levels of all strains tested wereshown to cotransduce. Evidence that TnS-VB32 was insert-ed into different regions of the chromosome was provided bySouthern blot analysis, which showed the presence of a sin-gle TnS-VB32-containing fragment of a different size in eachmutant chromosome restriction digest. Therefore, TnS-VB32 appears to be randomly inserted into the genome.However, a high percentage (77%) of Kmr strains was ob-tained among the Tcr exconjugants, which could mean thatinsertion of the promoter probe is biased towards transcrip-tionally active regions of the chromosome. Clearly, the isola-tion of Kms/Tcr strains that had TnS-VB32 sequences intheir genomes shows that the probe does not contain a pro-moter for the NPT II gene that can be read by the C. crescen-tus RNA polymerase.The decreased transposition frequency of TnS-VB32 most

likely reflects the fact that TnS-VB32 is 1200 base pairs larg-er than TnS due to the presence of the TnWO Tcr gene. Thismight contribute to the reduced transposition frequency (30).However, TnS-VB32, like wild-type Tn5, seems to be stablymaintained in C. crescentus. The transduction of TnS-VB32chromosomal inserts was also found to be less efficient thanwild-type TNS inserts. This may be related to the finding thatthe TnWO Bgl II restriction fragment present in TnS-VB32contained the tetracycline repressor as well as the Tcr deter-minant (tetA) (25) and so cells may have been killed beforetetA was expressed. The use of tetracycline slant plates (de-

Proc. NatL Acad Sci. USA 81 (1984)

Dow

nloa

ded

by g

uest

on

June

21,

202

1

Proc. NatL. Acad. Sci. USA 81 (1984) 1039

scribed in Materials and Methods) increased the efficiencyof transduction.A cysteine auxotroph AE7000 (cys-501::Tn5-VB32) was

isolated in which the expression of the transposon-derivedNPT II gene occurred from a C. crescentus promoter and itsexpression was found to be regulated by exogenous cys-teine. Therefore, the insertion of Tn5-VB32 allows tran-scripts to initiate at a C. crescentus promoter, which thenread through the leftward 72 base pairs of ISSOL into theNPT II gene. A hybrid protein would not be generated fromthe insertion event because IS50 contains a stop codon in allthree reading frames within the first 30 bases (31). Further-more, Tn5-VB32 contains its own translation start signal forthe NPT II protein and immunoprecipitation data showedthat an intact Mr 25,000 NPT II protein was synthesized un-der conditions of cysteine starvation. In the absence of anal-ysis of the mRNA produced, we cannot rule out the possibil-ity that regulation of NPT II expression by cysteine ob-served in strain AE7000 is due to post-transcriptionalcontrol. It is possible that a hybrid transcript is formed be-tween C. crescentus cysteine biosynthesis-related mRNAand NPT II mRNA and that this transcript contains 5' se-quences that can be differentially translated depending onthe concentration of cysteine within the cell.

In E. coli and Salmonella the cysteine biosynthetic path-way has been shown to be repressed by cysteine and sulfideand to be derepressed by starvation for this compound (26).When Tn5-VB32 was inserted into a C. crescentus gene in-volved in cysteine biosynthesis, both Kmr and the produc-tion of the NPT II protein were repressed by exogenous cys-teine and derepressed in the absence of supplement. There-fore, the cysteine biosynthetic pathway of C. crescentusappears to be regulated, as it is in E. coli and Salmonella, bycysteine availability.The isolation of strains AE7004 (fla-501) and AE7006 (fla-

503), which have lost the ability to synthesize a flagellum dueto insertional inactivation by Tn5-VB32 but show varyinglevels of Kmr, is relevant to our studies of the control ofdifferentiation in Caulobacter. Several flagellar genes areknown to be differentially expressed and the availability ofthe Tn5-VB32 promoter probe now permits us to test direct-ly whether differential expression is transcriptionally regu-lated. This probe now permits measurement of transcriptionfrom flagellar promoters by pulse-labeling protein in syn-chronized cells at various times in the cell cycle and prepar-ing immunoprecipitates from cell extracts using anti-NPT IIantisera. Promoter probe-generated Fla mutants can also betransduced into strains containing other Fla mutations. Aregulatory cascade that controls the expression of flagellarand chemotaxis genes mapping to different regions of thechromosome has been demonstrated in E. coli (32) and re-cently in C. crescentus (33). Double flagellar mutants con-taining one fla gene with a Tn5-VB32 insertion will allow usto determine whether the hierarchy of control occurs at thelevel of transcription.

We thank Reid Johnson for sending us pRZ102 and pRZ341 priorto publication and for helpful discussions and Julian Davies for anti-NPT II antisera. We also thank Karen Hahnenberger for examina-

tion of the Fla mutants by electron microscopy and for the X Tn5DNA. This investigation was supported by grants from the NationalInstitutes of Health (GM 32506, GM 11301, and National CancerInstitute P30-CAI-13330).

1. Ohta, N., Chen, L. S. & Newton, A. (1982) Proc. Natl. Acad.Sci. USA 79, 4863-4867.

2. Purucker, M., Bryan, R., Amemiya, K., Ely, B. & Shapiro, L.(1982) Proc. Natl. Acad. Sci. USA 79, 6797-6801.

3. Milhausan, M., Gill, P. R., Parker, G. & Agabian, N. (1982)Proc. Natl. Acad. Sci. USA 79, 6847-6851.

4. Gill, P. R. & Agabian, N. (1983) J. Biol. Chem. 258, 7395-7401.

5. Shaw, P., Lopes Gomes, S., Sweeney, K., Ely, B. & Shapiro,L. (1983) Proc. Natl. Acad. Sci. USA 80, 5261-5265.

6. Newton, A. (1972) Proc. Natl. Acad. Sci. USA 69, 447-451.7. Agabian, N., Evinger, M. & Parker, E. (1979) J. Cell Biol. 81,

123-136.8. Milhausen, M. & Agabian, N. (1983) Nature (London) 302,

630-632.9. Berg, D. E., Jorgensen, R. & Davies, J. (1978) in Microbiology

1978, ed. Schlessinger, D. (Am. Soc. Microbiol., Washington,DC), p. 13.

10. Ely, B. & Croft, R. H. (1982) J. Bacteriol. 149, 620-625.11. Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular

Cloning (Cold Spring Harbor Laboratory, Cold Spring Harbor,NY), p. 51.

12. Davis, R. W., Botstein, D. & Roth, J. R. (1980) AdvancedBacterial Genetics (Cold Spring Harbor Laboratory, ColdSpring Harbor, NY), p. 70,

13. Beringer, J. E., Beynon, J. L., Buchanan-Wollastan, A. V. &Johnston, A. W. B. (1978) Nature (London) 276, 633-634.

14. Monk, M. & Kinvoss, J. (1972) J. Bacteriol. 109, 971-978.15. Barrett, J., Croft, R., Ferber, D., Gerardot, C., Schoenlein, P.

& Ely, B. (1982) J. Bacteriol. 151, 888-898.16. Jorgensen, P. A., Rothstein, S. & Reznikoff, W. S. (1979) Mol.

Gen. Genet. 177, 65-72.17. Johnson, R. C. & Reznikoff, W. S. (1983) Nature (London)

304, 280-282.18. Ely, B. (1979) Genetics 91, 371-380.19. Poindexter, J. S. (1964) Bacteriol. Rev. 28, 231-295.20. Contreras, I., Shapiro, L. & Henry, S. (1978) J. Bacteriol. 135,

1130-1136.21. Levine, M. (1957) Virology 3, 22-41.22. Miller, J. (1972) Experiments in Molecular Genetics (Cold

Spring Harbor Laboratory, Cold Spring Harbor, NY).23. Hodgson, D., Shaw, P., Letts, V., Henry, S. & Shapiro, L.

(1984) J. Bacteriol., in press.24. Bertrand, K. P., Postle, K., Wray, L. V. & Reznikoff, W. S.

(1983) Gene 23, 149-156.25. Rigby, P., Dieckmann, M., Rhodes, C. & Berg, P. (1977) J.

Mol. Biol. 113, 237-251.26. Umbarger, H. E. (1978) Annu. Rev. Biochem. 47, 533-607.27. Shapiro, L., Mansour, J., Shaw, P. & Henry, S. (1982) J. Mol.

Biol. 159, 303-322.28. Kleckner, N. E., Chan, R., Tye, B. K. & Botstein, D. (1975)

J. Mol. Biol. 97, 561-575.29. Berg, D. E. & Berg, C. M. (1983) Biotechnology 1, 417-435.30. Chandler, M., Clengel, M. & Golas, D. J. -(1982) J. Mol. Biol.

154, 229-243.31. Aurswald, E. A., Ludwig, G. & Schaller, H. (1980) Cold

Spring Harbor Symp. Quant. Biol. 45, 107-114.32. Komeda, Y. (1982) J. Bacteriol. 150, 16-26.33. Bryan, R., Purucker, M., Lopes Gomes, S., Alexander, W. &

Shapiro, L. (1984) Proc. Natl. Acad. Sci. USA, in press.

Biochemistry: Bellofatto et aL

Dow

nloa

ded

by g

uest

on

June

21,

202

1