promoters of the broad host range plasmid rk2: analysis of ... · promoters of the broad host range...

Copyright 0 1992 by the Genetics Society of America

Promoters of the Broad Host Range Plasmid RK2: Analysis of Transcription (Initiation) in Five Species of Gram-Negative Bacteria

Alan Greener,' Samuel M. Lehman and Donald R. Helinski

Department of Biology and Center for Molecular Genetics, University of Calgornia at San Diego, La Jolla, Calgornia 92093 Manuscript received February 12, 1991

Accepted for publication September 14, 199 1

ABSTRACT A broad host range cloning vector was constructed, suitable for monitoring promoter activity in

diverse Gram-negative bacteria. This vector, derived from plasmid RSFl 010, utilized the firefly luciferase gene as the reporter, since the assay for its bioluminescent product is sensitive, and measurements can be made without background from the host. Twelve DNA fragments with promoter activity were obtained from broad host range plasmid RK2 and inserted into the RSFlO 10 derived vector. The relative luciferase activities were determined for these fragments in five species of Gram- negative bacteria. In addition, four promoters were analyzed by primer extension to locate transcriptional start sites in each host. The results show that several of the promoters vary substantially in relative strengths or utilize different transcriptional start sites in different bacteria. Other promoters exhibited similar activities and identical start sites in the five hosts examined.

P LASMID RK2 is a 60-kilobase, self-transmissible member of the IncP group of plasmids, whose

hallmark is their broad host range (for a recent review, see SMITH and THOMAS 1989). These plasmids are capable of transfer, replication, and maintenance in most genera of Gram-negative bacteria, Bacteroides and Myxococcus representing the only known exceptions (BRETON, JAOUA and GUESPIN-MICHEL 1985; GUINEY, HASEGAWA and DAVIS 1984).

T o be propagated or selected in such a wide variety of hosts, the essential genes of the plasmid must be expressed. These include genes required for replication, stable maintenance, conjugal transfer and antibiotic resistance. RK2 also encodes a group of proteins (kil) (FIGURSKI et al. 1982) which are lethal to cells that have lost the plasmid (for a recent review, see THOMAS and HELINSKI 1989). The transcriptional repressors (kor) of these genes are therefore also essential for the propagation of the plasmid. The mechanisms utilized by the IncP and other broad host range plasmids to ensure controlled expression of their genes in a wide-range of bacteria are not known. This question is of importance to our understanding of the properties of broad host range plasmids since transcription initiation factors are not universally conserved among the Gram-negative bacteria (THOMAS and FRANKLIN 1989).

T o begin studying the basis of this versatility in gene expression, promoter cloning vectors were constructed that replicate in diverse Gram-negative bacteria and contain a reporter gene whose product can

' Present address: Stratagene, Inc., 11099 North Torrey Pines Road, La Jolla, California 92037.

Genetics 190: 27-36 (January, 1992)

readily be quantified in the various cellular environments. Specifically, the firefly luciferase gene, lacking a promoter, was inserted downstream of a multicloning site within a derivative of the broad host range plasmid RSFlO 10, a member of the IncQ plasmid incompatibility group. The bioluminescent product of the firefly luciferase gene in this transcriptional fusion vector can be sensitively and accurately assayed with no background from the host.

Using this system, we cloned 12 DNA fragments with promoter activity from the broad host range plasmid RK2 and measured their activities in five species of Gram-negative bacteria. We also identified the locations of the promoter fragments on the RK2 physical map. Our results indicate that the strengths of these promoters vary widely when measured in five different Gram-negative bacteria. Also, no consistent pattern was observed when comparing relative activities of each promoter within the same host bacterium. Furthermore, we have analyzed the transcriptional start sites for four of these promoters and have found that their initiation sites in the different hosts are not always identical. Based on these results, we conclude that there is no single universal promoter or adaptive mechanism for expression in diverse hosts. Rather, each promoter may have evolved its own unique adaptive strategy.

MATERIALS AND METHODS

Bacterial strains, plasmids, media and microbiological procedures: The Escherichia coli strains and plasmids used in this study are described in the text. Pseudomonas aeruginosa PA0 1 16 1, Acinetobacter calcoaceticus BD4 1 3, Rhizobium meliloti 102F34nal and Agrobacterium tumefaciens A 18 are

28 A. Greener, S. M. Lehman and D. R. Helinski

standard laboratory strains. Standard protocols and recipes were used for transformation (MANIATIS, FRITSCH and SAM- BROOK 1982), plasmid DNA extraction (BIRNBOIM and DOLY 1979), conjugal matings (DITTA et al. 1980) and media (SCHMIDHAUSER and HELINSKI 1985). Construction of pAL4000 is described in the legend to Figure 1. pAL37 is a tetracycline-sensitive deletion derivative of RK2. pAL200 is essentially identical to pAL4000 except that it contains a promoterless chloramphenicol resistance (cat) gene (from Tn9) in place of the luciferase gene, luc.

Biochemical procedures, RNA isolation and determi- nation of transcriptional start sites: Restriction endonuclease digestions, ligations, and calf-intestinal alkaline phosphatase digestions were performed according to the manufac- turer's specifications. Procedures for nick translation and RNA-DNA dot blot hybridization (SAMBROOK, FRITSCH and MANIATIS 1989) have been described. Total RNA was isolated as described by ZHU and KAPLAN (1985), except for A. calcoaceticus, in which case lysozyme at a final concentra- tion of 500 pg/ml was added at 37" and incubated for 30 min prior to sodium dodecyl sulfate-pronase addition. Primer extensions were performed using AMV reverse transcriptase (Boehringer Mannheim) (VIRTS et al. 1988). A 24- base oligonucleotide, complementary to the XbaI-Hind111 segment of the pUC19 polylinker, was chemically synthesized and labeled at the 5' end with [y-"PIATP or [y"'S] ATP and polynucleotide kinase for use as the primer in the reverse transcription reactions. The same oligomer, unla- beled, served as the primer for DNA sequencing reactions performed by the dideoxynucleotide chain termination procedure (SANGER, NICKLEN and COULSON 1977).

Luciferase assays: Luciferase assays were performed on logarithmically growing cultures of bacteria essentially as described by WOOD and DELUCA (1987) and SUBRAMANI and DE LUCA (1 988). Cells (900 pI) (approximately 1 to 3 X 10' cells/ml), grown at 30°, were added to 100 PI of 1OX sonication buffer (50% glycerol, 1 M K2HP04, 10 mg/ml BSA, 20 mM EDTA, pH 7.4) and immediately chilled on ice. The samples were subjected to two cycles of 24 sec of sonication, separated by 5 min for cooling. Samples of 20 pl of the sonicate were then mixed with 200 pl of 10 mM MgC12, 3.0 mM ATP, 25 mM glycylglycine, pH 7.8, in a disposable cuvette designed for the Monolight 2001 luminometer, both of which were purchased from Analytical Luminescence Laboratories of San Diego, California. After insertion of the cuvette into the luminometer a 100-pl solution of 400 p M D-hCiferin (Analytical Luminescence) in 25 mM glycylglycine, pH 7.8, was injected, and the peak light emission recorded. A standard curve relating light units to luciferase was determined using purified luciferase enzyme (gift of D. VELLUM).

For whole cell (plate) assays, the bacteria were grown on a nitrocellulose filter placed on solid agar media for 18-36 hr at 30". The filters were then soaked in a solution of 1 mM luciferin, 100 mM Na-citrate (pH 5.5) for 10 min in the dark, blotted dry and exposed to X-ray film for 5-10 min.

RESULTS

Construction and properties of a broad host range cloning vector containing luciferase as a reporter gene: For the analysis of the relative strengths of promoters of the plasmid RK2 in diverse bacteria a broad host range cloning vector derivative of the IncQ plasmid RSFlOlO was constructed that contained a promoterless luciferase reporter gene whose product

E- pAL4000

FIGURE 1 .-Construction of the broad host range promoter cloning vector pAL4000. K, P, E, B, Bg, S and X refer to sites for restriction endonucleasesKpn1, PstI, EcoRI, BamHI, BglII, Sac1 and Xbal, respectively. The plasmid was constructed as follows: (1) The luciferase gene (luc) was obtained as a HindIII-SmaI fragment from pKW102 (K. WOOD, unpublished results) and inserted into Hind111 and EcoRV digested PAL1 3, a Tet5 derivative of pBR322 generated by filling in the BamHI site, and inserting an &base pair BglII linker (5' CAGATCTG) at the filled in Sal1 site. PAL1 3, carrying luc, was designated pAL14. (2) An EcoRI fragment from PAD9 carrying a transcriptional terminator (t) was cloned into EcoRI digested pUC 19 to produce pAL1. (3) The tetracycline resistance gene ( t e t ) was obtained as a XholI-StuI fragment, from pTJS37 (SCHMIDHAUSER and HELINSKI 1985), EcoRI linkers were added and inserted into a partial EcoR1 digest of pALl to construct pAL6. (4) A complete Hind111 and partial EcoRI digest of pAL6 was carried out to isolate the fragment carrying TET, t, and pUCI9 polylinker (MCS). The fragment was inserted into pAL14 to construct pAL23. (5) The translation termination fragment (T) (5' CTAGCTAGCTAG3') was inserted into pAL23 that was digested with Hind111 and filled in to construct pAL27. (6) A complete BglIl and partial EcoRl digest of pAL27 yielded the segment containing tet, t, MCS, T, luc and a region of plasmid pBR322 (coordinates of 185-65 1 bp). This was ligated to an RSFlOlO derivative deleted of the 0.8-kb Psi1 fragment with EcoRI and BglII linkers added to the PstI and EcoRI termini, respectively (BAGDASARIAN et al. 1981).

is readily measured in diverse bacteria [e.g. E. coli, P. aeruginosa, and R. meliloti (DEWET et a l . 1985, PALO- MARES, DELUCA and HELINSKI 1989)] and upstream restriction sites for making transcriptional fusions.

T h e basic plasmid construct, designated pAL4000, has several important features (Figure 1). (1) T h e RSF 10 10 replicon can be stably maintained in diverse species. Although RSFlO 10 (and thus, pAL4000) is not self transmissible, it is mobilizable by the transfer functions of plasmid RK2 (DITTA et al . 1980). (2) T h e tetracycline resistance gene ( M A ) of RK2 was included in pAL4000, since this gene can be selected in diverse Gram-negative bacteria (SCHMIDHAUSER 1987). (3) pAL4000 contains the multiple cloning site of pUc 19 upstream of the reporter gene. (4) The E. coli rpoC rho-independent transcriptional terminator is upstream of the multicloning site to reduce transcription

Broad Host Range Promoters 29

readthrough into luc. (5) A 12-bp linker containing translational stop signals in each reading frame is between the multicloning site and the luc open reading frame (ORF) in pAL4000 to eliminate translational fusions to luc.

An analogous pBR322-based plasmid, pAL20 (Fig- ure 3), was also constructed. This vector carried the chloramphenicol resistance (cat) reporter gene instead of luc to be able to select weak promoter signals directly. It contains the pUCl9 polylinker in the same orientation as in pAL4000, so that transfer of a cloned DNA fragment from it to pAL4000 can be a single step cloning procedure.

Luciferase as a quantitative reporter of promoter activity in different Gram-negative bacteria: It was important to validate the use of the luciferase enzyme as a quantitative measure of promoter activity. A standard curve of enzyme vs. light units demonstrated that the assay was linear from 3.0 pg to 1.5 pg of protein, which corresponded to 3 X lo6 to 1.5 X lo9 molecules of luciferase (data not shown). The vector pAL4000 without a promoter fragment insert produced 800-1600 light units per lo6 cells depending on the host, corresponding to 100-200 molecules of luciferase per E. coli cell (5-10 molecules of luciferase per plasmid copy). A sonicate prepared from plasmid free E. coli gave a value of about 60 light units/106 cells. This instrumentation background noise was sub- tracted from all values obtained.

The firefly luc gene product can be quantified by enzyme activity in many prokaryotic and eukaryotic organisms (reviewed by SUBRAMANI and DELUCA 1988), including E. coli (DEWET et al. 1985), P. aeruginosa, and R. meliloti (PALOMARES, DELUCA and HEL- INSKI 1989). We inserted known E. coli promoters into pAL4000 [i.e., tac, lac, KanR (Tn5)] and found that bioluminescence, due to luciferase could also be measured reproducibly in A. calcoaceticus and A. tumefaciens. Comparison of luc-specific RNA, determined by dot blot hybridization, to bioluminescence due to pAL4000 derivatives carrying various promoter inserts also confirmed that within a particular host or- ganism the luciferase activity generally was propor- tional to the amount of luciferase RNA (data not shown). Since the efficiency of post-transcriptional events is likely to be species-specific, no attempt was made to quantitatively compare data between the different bacterial species.

Figure 2 illustrates that simple plate (colony) assays of luciferase activity that complement quantitative assays are also available. In this experiment, bacteria harboring pAL4000 with different RK2 promoters (see below) were streaked onto nitrocellulose filters overlaid on petri dishes. The filters were treated with luciferin (MATERIALS AND METHODS) and exposed to X-ray film. The relative intensities of luciferase in the

five species were consistent with the quantitative sonication assays carried out on the bacteria harboring the various pAL4000 derivatives (Table 1).

Cloning and selection of RK2 fragments with promoter activity into pAL4000: Since certain regions of RK2 have been sequenced, restriction fragments containing known promoters were inserted into a promoterless cat vector, pAL20, by direct cloning (Figure 3). For promoters not yet characterized, a Sau3A digest of RK2 DNA was ligated to BamHI digested pAL20, and the ligated DNAs were used to transform E. coli HBlOl to chloramphenicol resistance (4 pg/ml). Twenty-four transformants expressing chloramphenicol resistance were obtained and further analyzed. Nine, distinguished by the size of the Sau3A insertion and/or levels of chloramphenicol resistance, were chosen for further analysis.

Mapping RK2 promoters by cointegrate formation: T o map the cloned fragments relative to the RK2 DNA molecule, cointegrates between a plasmid carrying the insert and a tetracycline sensitive derivative (pAL37) of RK2 were selected, and characterized as follows. E. coli strain TB1 (polA+, Ret+, Nals) harboring the self-transmissible plasmid pAL37 was transformed with DNA of each of the pAL20-derived plasmids. Transformants were mated with plasmid- free C2 1 1 ONalR ( polA-, Ret+) and NalR, TetR, KanR exconjugants were selected. Because pAL2O can not replicate in polA (DNA polymerase 1 deficient) bacteria (KINGSBURY and HELINSKI 1973), this selection favors cointegrates between pAL37 and the pAL20 derivated plasmids that occasionally should be formed by recombination between homologous segments in the TB1 (Rec+) host. There should only be two regions of homology between the two parental plasmids: in the P-lactamase (bla) genes of RK2 and pBR322 (HEFFRON et al. 1979), and the segments from RK2 cloned in pAL20. Cointegrates formed by recombination in bla would have nearly identical restriction digest patterns, differing only by the size of the particular promoter fragment of RK2. Cointegrates formed by recombination within RK2 sequences would yield a unique restriction pattern for each individual promoter fragment inserted into pAL2O (see Figure 4). In addition, the pattern obtained would indicate the direction of transcription.

Using this conjugal mating/cointegrate analysis technique, exconjugants were selected in C2 1 10NalR; their frequency was approximately 1 03-1 04-fold lower than exconjugants obtained when pAL37 and pAL200, a TetR RSFlO 10 derivative capable of au- tonomous replication in the PolA recipient, were used. DNA from these cointegrates was analyzed by restriction endonuclease digestion and for each Sau3A insert two restriction patterns were observed (only one pattern was observed for the cointegrate containing

30

A

A. Greener, S . M. Lehman and D. R. Helinski

B "

C

D .. . "

!

E

!

FIGURE 2.-Whole cell (plate) luciferase assays of pAL4000 containing different RK2 promoters in five species of Gram-negative bacteria. See MATERIALS AND METHODS for details of the assay procedure and text for description of promoter designations. Plasmid PAL1000 was an RSFlOlO derivative lacking the luciferase gene. (A) E. coli HBlOl: (1) pAL4000-87, (2) pAL1000, (3) pAL4000, (4) pAL4000-Rl1, (5) pAL4000-BK10, (6) pAL4000-223. (B) P. aeruginosa PA01161: (1) pAL4000-223, (2) pAL4000-87, (3) pAL1000, (4) pAL.4000, (5) pAL4000-R11, (6) no plasmid. (C) A. calcoaceticus BD413: (1) pAL4000-R11, (2) pAL4000-BK10, (3) pAL4000-223, (4) pAL4000-87, (5) pAL1000, (6) pAL4000. (D) R. mefiloti 102F34naf: (1) pAL4000-RI1, (2) pAL4000-BK10, (3) pAL4000-223, (4) pAL4000, (5) pAL1000, (6) pAL4000-87. (E) A. tumefaciens A18: (1) pAL4000-R11, (2) no plasmid, (3) pAL4000-223, (4) pAL4000-87, (5) pAL1000, (6) pAL4000.

TABLE I

Activity of RK2 promoters in five species of Gram-negative bacteria

Promoter Promoter designation (location) E. coli P. acmginosa A. calcoaccticus R. meliloti A. tumefaciens

pAL4000 None 1 .O 1 .O 1 .O 1 .o 1 .o 4007 tetA 8.0 11.4 6.5 4.0 6.4

R11 kcrA 1 23.0 10.2 10.7 11.3 20.8 R15 56.9 CCW 6.8 27.2 4.6 2.1 6.6 BK3 45.8 CCW 1.2 6.2 0.8 0.5 1.5 BK5 16.3 CW 1.3 1.7 1.1 3.5 5.8 BKlO korA 11.8 9.6 7.8 5.0 7.0 BLI EwA 9.4 11.2 2.2 2.1 6.6 220 48.0 CCW 1.2 11.8 0.8 0.8 2.1 223 Tral P1.2 1.6 4.9 1.1 2.3 3.7 87 oriT PB 1 2.7 10.4 1.7 1.6 2.3 88 oriT PL1 3.1 9.3 2.1 8.9 11.6

R9 tnpA 4.3 2.0 2.1 1.1 2.4

Activities are presented as a ratio of the number of light units from cells harboring the plasmid of interest to the number of light units expressed by pAL4000 in a given host bacterium. See MATERIALS AND METHODS for assay conditions. CCW and CW refer to counter clockwise and clockwise direction of promoter activity. Fach value represents an average of five independent trials; the standard deviation was less than 10% for each value obtained.

pAL20 that does not contain a Sau3A insert as ex- tion digestion patterns for eight independently iso- pected). Figure 4A, lanes 1-4 and 7-1 0, shows restric- lated cointegrates. One of the two patterns was similar


Sau3A P , 3 Sau3A I I RK2SauBA

FIGURE 3.-Cloning of RK2 fragments containing promoters into pAL4000. RK2 DNA was digested with Sau3AI and ligated to BamHIdigested and calf intestinal alkaline phosphatase treated pAL20 DNA and transformed into HBlOl selecting for chloramphenicol resistance (4 pglml). Plasmid pUCD2B (D. GUINEY, per- sonal communication), which contains an oriT RK2 segment inserted into pUC7, was digested with Hind111 then filled in and ligated to an 8-bp BamHI linker. This DNA was digested with BamH1, ligated to Bamdigested pAL2O and transformed into HBlO1. Orientation of the oriT fragment was determined by restriction mapping. Plasmid pAL6 (see Figure 1 legend) was digested with NaeI, filled in, and BamHI linkers were added prior to BamHI digestion and insertion into pAL20 as above. Each pAL20 derivative containing an insert of interest was digested with KpnI and PstI and inserted into pAL4000 that was digested with KpnI and PstI. Transformants were obtained in E. coli strain HBlOl. The letter designations for restriction enzymes are the same as in the legend to Figure 1 .

for each insert and to that obtained for pAL2O. Diges- tion with ApaI, ClaI and XhoI restriction enzymes showed that the site of recombination was within the segment of homology present in their bla genes (Fig- ure 4). Cointegrate DNAs formed between pAL37 and pAL2O-R15 (Figure 4B, lanes 1-3) or, for example, pAL37 and pAL20-BK3 (Figure 4C, lanes 1- 2) also yielded a unique pattern for each promoter. Analysis of these fragment sizes enabled us to map the site of homologous recombination within pAL37 and the relative orientation of the two plasmids to distin- guish the direction of transcription within the parent RK2 plasmid. For example, a unique pattern was observed after ApaI, ClaI and XhoI digestion of cointegrates between pAL37 and pAL20-220 (Figure 4B, lanes 6-8) which indicated recombination between

the promoter segment carried on pAL20-220 and its corresponding region of pAL37. Similar analyses of pAL37 and the remaining Sau3A inserts into pAL2O were performed to yield the data summarized in Fig- ure 5. Some of these insertions (R1 1, BKlO, R9, BLl, 223) corresponded to previously identified RK2 promoters, as subsequently confirmed by either DNA sequencing or restriction mapping. Others (BK3, R15, BK5, 220), however, mapped to areas of RK2 which have not been extensively studied and, therefore, only their relative coordinates on the RK2 map are given.

Promoter activity in five Gram-negative species: Twelve RK2 promoter fragments cloned into pAL2O were isolated as KpnI-PstI restriction fragments and inserted into pAL4000. These plasmids were conju- gally transferred into P. aeruginosa, A. calcoaceticus, R. meliloti and A. tumefaciens, using the ColEl-derived helper plasmid pRK2013 (DITTA et al. 1980). The results of luciferase assays are shown in Table 1 and are presented as a ratio of the number of light units obtained for the plasmid of interest compared with the pAL4000 vector in each individual host. It is clear that the strengths of the different promoter segments vary widely within a particular host and that several of the individual promoter segments direct variable levels of luciferase when comparing their relative activities across species lines. Due to post-transcriptional effects that are likely to differ from host to host, the numbers obtained are likely to be valid only when comparisons are carried out within a particular species. For interspecies comparisons, only general trends are likely to be significant.

Three promoters [ R l l (kcrAI), BKlO (KorA) and 4007 (tetA)] exhibited moderate to strong activity in all five hosts examined. Three promoters [BK3, 220, 4087 (oriT PR1)] showed moderate to strong activity only in P. aeruginosa; the activity was either weak or below the vector background in the other bacteria. Two promoters [4088 (oriT PL1) and 223 (Tral PL2)] exhibited moderate or strong activity in P. aeruginosa, R. meliloti and A. tumefaciens, but relatively weak activity in E. coli and A. calcoaceticus. One promoter (R15) was moderate to very strong in all hosts except R. meliloti. The BK5 promoter, on the other hand, showed moderate strength in R. meliloti and A. tumefaciens only, while the BLl promoter exhibited moderate to strong activity in E. coli, P. aeruginosa and A, tumefaciens, and substantially less activity in A. calcoaceticus and R. meliloti. Finally, the R9 (tnpA) promoter was weak to moderate in strength in all the hosts tested.

Identification of transcriptional starts: Since it has been shown that the sequences which constitute promoters in one species are not necessarily identical to those which constitute promoters in another (see THOMAS and FRANKLIN 1989), we were interested in

A

A. Greener, S. M. Lehman and D. R. Helinski

C

B

1 2 3 4 5 6 7 8 1 2 3 '

FIGURE 4.-Kestriction ~ .ndonucI~' ;~~t . digestion of pAL37-pAL20 cointegrate plasmid DNAs. DNA w a s isolated from 250 tnl cells by standard CsCI-ethidiun1 bronlide equilibrium gradient centrifugation and digested with the enzymes Apal, Clal and X h o l . Fach lane containing ;I cointegrate corresponds to an independenrly isolated plasmid. (A): Lanes (1-4) pAL37-pAL20; lane (5) pALS7: lane (6) A Hind111 size standards: lanes (7-10) pAL37-pA1,'LO. (R): Lanes (1-3) pAL37-pAL20-R1.5; lane (4) pAL37: lane (.i) A Hindlll size standards: lanes (6-8) ~~Al.S7-pAl.20-220. (C): Lanes ( 1 , 2) pAL37-pAL20BKS; lane (3) pA1.37: lane (4) A Hindlll size standards.

B S R l 1

kcrA 1

lnpR

RK2 BK3

k t8 kllD

KanR

BL 1 z-

30

FIGURE 5.-Location o f the isolated RK2 segments with promoter activity i n R K 2 . T h e location and direction o f transcription o f the isolated RK2 promoter fragments are indicated on the outside of the plasmid map.

determining the transcription initiation sites for the cloned RK'L promoters in the different hosts. Four promoter fragments (kcrA1, korA, oriT PR1 and Tral P1.2) were selected for primer extension analysis to identify transcription start sites. Our results for these four promoters indicated that start sites are constant among species with some fragments but are variable for others. Promoter R11 (kcrAl ) initiated transcription from the identical three bases in the five bacteria studied (Figure 6A,i); these start sites corresponded to those previously published for E. coli (THOMAS et al. 1988).

The results with the BK 10 (korA) promoter showed that transcription was also initiated from the same site in all five hosts (Figure 6A,ii), which coincided with the start site determined in E. coli (THOMAS and SMITH 1986). However, for A. tumefaciens, three smaller species of transcripts additionally were observed, located 44, 47 and 48 bases downstream from the conserved start. These are either the result of the activity of an additional promoter unique for A. tumefaciens, or processing sites for the transcript which initiates upstream.

The results with the oriT pR1 promoter (4087) demonstrated that it also contained an initiation site conserved among all five hosts (Figure 6B,iii). How- ever, a unique start site was observed for A. tumefaciens, 29-30 bases upstream. Since this was upstream of the conserved initiation site, it suggested a unique promoter sequence for A. tumefaciens in this region. The A. tumefaciens reverse transcription product which comigrated with those of the other hosts indi- cates an alternative promoter, or may result from RNA processing of the transcript initiated from the upstream promoter region. In A. calcoaceticus, two faster migrating species, located 8 1 and 101 bases downstream, respectively, were detected. Each of these bands was of approximately equivalent intensity as the one from the conserved upstream site and by a similar argument could represent either unique A. calcoaceticus promoter(s) or a processing product.

With the Tra l PL2 promoter transcription started from one site in E. coli, P. aeruginosa and A. calcoaceticus (Figure 6B,iv), and from a second site (32-33 bases upstream) in R. meliloti and A. tumefaciens. A secondary start or processing site for R. meliloti coincided with the initiation sites for the other three hosts;


-35 -10 pir-wT: AAAAAITGACRXCATGITAITGGCGITAAGATATACAGAAT

I pir -6 AAAAAITGAATCWATCITAITGGCGITAAGATATACAGAAT

A I ii " -

1 2 3 4 5 6 7 8 9

B

1 1 3 . 5 L I B 9

Luciferase Activities

E. coli P. orruginosa A. cdcooceticru R. nulilofi A . Nmrfocicns

pir-WT 16.2 2.9 8.7 8.9 16.4

pir-6 7.8 I .4 4.3 4.9 10.8

FIGURE 7.-Sequence of the pir-wt and pir-6 promoters and their activities relative to pAL4000 in the five species of Gram-negative bacteria. The location of the mutation is indicated by the arrow.

its intensity was considerably weaker than the upstream site. While the signals for the Tral P L ~ promoter were somewhat less intense than for the other three promoters, they were reproducible.

A mutation affecting promoter activity in all five hosts: The pir promoter of the narrow host range plasmid R6K directs transcription of the essential replication initiation protein ?r (STALKER, KOLTER and HELINSKI 1982). We had isolated a promoter-down mutation pir-6, due to a C + A transversion in the -35 region (A. GREENER, unpublished results) (Figure 7). T o ascertain whether the pir promoter was functional in the other hosts and whether the change in the -35 region affected its activity to the same degree in each of them, both the pir-WT and fir-6 promoters were inserted into pAL4000. The resulting plasmids, pAL4001 and pAL4002, respectively, were transferred into the different bacteria and luciferase was assayed (Figure 7).

The pir promoter, previously shown to be relatively strong in E. coli (SHAFFERMAN et al. 1982), displayed high activity in other hosts except P. aeruginosa. The pir-6 mutation, which resulted in an approximately 2- 3-fold weaker promoter in E. coli both in vivo and in

FIGURE B.-Analysis of transcriptional start sites. Primer extension analysis of the RNA start sites was carried out for the kcrA1 [A;(i)], korA, [A;(ii)], oriTpR1 [B;(iii)], and Tral pL2 [B;(iv)] promoters. For each: Lanes 1-5: primer extension reactions on RNA isolated from E. coli, P. aeruginosa, A. calcoaceticus, R. meliloti and A. tumefaciens, respectively; a '*P- or "S-24-base oligonucleotide complementary to the XbaI-Hind111 segment of the pUCI9 polylinker in pAL2O derivatives carrying the promoter fragments was used as the primer which was extended using 20 units of A M V reverse transcriptase. Lanes 6-9: GATC DNA sequencing reactions; the sites where RNA transcripts initiate or are processed are indicated as follows: all hosts; E. coli, P. aeruginosa, A. culcoacetincs, R. meliloti only; 4 A. tumefaciens only: A. calcoaceticus only; R. meliloti and A. tumefaciens only. The consensus -35 and -10 hexamers corresponding to the E. coli promoter are designated.


vitro (A. GREENER, unpublished results) was also weaker in the other hosts. This suggested that the -35 hexamer of the pir promoter is involved in transcription initiation in all five bacteria examined.

DISCUSSION

Plasmids which have evolved a broad host range have a number of important adaptations that ensure their stable propagation in the wide variety of cellular environments. One of these adaptations is the ability to carry out the controlled expression of its genes in different environments. RNA polymerases from a number of eubacterial species have been purified and found to be structurally related; the core enzymes have a common organization (a&3’) (MCCLURE 1985; GAO and GUSSIN 1991). The sigma factors of these polymerases, however, which are responsible for the specific selection of transcriptional start sites, vary widely (THOMAS and FRANKLIN 1989). Beside the heterogeneity in size and primary amino acid sequence of the CT factors, the specific bases with which they direct their cognate core polymerases to interact are not identical. Examples demonstrating that a promoter from one Gram-negative bacterium is often nonfunctional in another have been documented (reviewed in THOMAS and FRANKLIN 1989). However, some promoters examined have exhibited broad host range initiation activity (e.g., the tac promoter; FREY, MUDD and KRISCH 1988) and a previous study with the trfA promoter in RK2 demonstrated that its start site was conserved between E. coli and two species of Pseudomonads (PINKNEY et al . 1986).

We have developed a series of cloning and luc reporter gene expression vectors suitable for selection and quantitation of promoter activity in diverse Gram- negative bacteria. Using these vectors, we analyzed promoters from the broad host range plasmid RK2 to determine how transcription in different hosts is ac- complished and compared the relative strengths of twelve RK2 promoters within a particular host bacterium.

The data presented in Table 1 indicate that no universal rules concerning relative activities of promoters from broad host range promoters can be drawn. A wide variation in their relative activities within each host is observed. Certain regions which act as moderate to strong promoters in one Gram- negative bacterium do not do so in another. It is possible that either the gene products or the RNA synthesized from these variable promoter elements may play a different role in the RK2 broad host range properties and/or are regulated differentially depending upon the host. It is also possible that some of the promoters examined in this study are subject to up- or down-regulation by other genes within the intact RK2 plasmid.

TWO promoters, analyzed by luciferase activity and by primer extension mapping, were highly expressed in all hosts studied. The promoter designated R1 1, which was mapped to the kcrAl gene by cointegrate formation and confirmed by DNA sequencing, is responsible for producing the KcrA1 protein. It is not known what role this product plays in the RK2 life cycle. However, its promoter structure is recognized similarly in all five hosts investigated on the basis of the identity of the transcriptional start sites (Figure 6A). The identification of the start sites here coincides with a previous study of the kcrA1 promoter in E. coli reported by THOMAS et al . 1988.

A second promoter directing substantial luciferase activity in all hosts tested was the korA promoter, responsible for transcribing the korA, incC and korB genes (THOMAS and SMITH 1986). The KorA and KorB products have been implicated in the control of RK2 copy number via transcriptional repression of the RK2-specified replication gene t$A (SHINGLER and THOMAS 1984). In addition, they prevent synthe- sis of a number of host lethal gene products (KiIA, KiIB, KilC, KiID) by binding to their respective promoters (reviewed in THOMAS and HELINSKI 1989). The incC gene product whose coding region overlaps korA (8 bases upstream and produced out of frame) has been implicated in RK2-specified incompatibility (MEYER and HINDS 1982; THOMAS and SMITH 1986). The substantial activity of the korA promoter in all hosts is, therefore, not surprising. However, although the transcriptional start sites were identical for all five hosts, an alternative initiation region, representing either a unique promoter or an RNA processing site and located 44-48 bases downstream, was observed for A. tumefaciens. This apparent start downstream appeared relatively strong and could represent the primary start site in vivo in A . tumefaciens. Regardless of the mechanism, the resulting 5’ end of the korA mRNA, unique for A. tumefaciens, is not likely to specify the incC product since it lacks the ribosome binding site. It is possible that an incC product is not required in A. tumefaciens.

The remaining two promoters studied by primer extension (Tral pL2 and oriT pR1) exhibited interspecies differences in start site utilization and/or RNA processing. The oriT pR1 promoter, which controls transcription for 15-kDa and 26-kDa proteins that are involved in conjugal transfer (GUINEY and LANKA 1989), had identical starts in all five hosts. However, a secondary start site unique to A. tumefaciens was present 29-30 bases upstream of this consensus. In addition, two smaller species of transcripts were observed for A. calcoaceticus starting at 8 1 and 10 1 bases downstream respectively. The downstream sites may correspond to alternative initiation sites or RNA proc-


essing in A. calcoaceticus and their presence may have regulatory significance for this host.

The pL2 promoter, which may be responsible for transcription of the primase operon and expression of additional proteins involved in conjugal transfer (FURSTE et al . 1989), exhibited identical starts among all hosts except A. tumefaciens; for this bacterium, a distinct initiation site was located 43-44 bases upstream. This upstream start site was also observed for R. meliloti and appeared to be utilized in that host considerably more frequently than the downstream site common for the other three hosts.

It is interesting to note that for the four promoters analyzed by primer extension, all exhibit identical start sites in E. coli, P. aeruginosa and A. calcoaceticus. Although substantial sequence information is not yet available for determining a consensus promoter sequence for P. aeruginosa and A. calcoaceticus, based on these data, it may be similar to the E. coli consensus.

For the remaining eight promoters that were not analyzed by primer extension, each appears somewhat unique with regard to its relative activities in the various Gram-negative bacteria tested. Preliminary results suggested that with one or two exceptions, a given promoter had maximal activity in P. aeruginosa. This is not surprising since P. aeruginosa was the host from which plasmid RK2 originally was isolated. The one notable exception, BK5, is located downstream of the RK2 replication protein(s) gene trfA, and directs transcription of the gene in an antisense orientation.

An interesting observation was also made concerning the activities of a wild-type and mutant promoter isolated from the narrow host range (E. coli) plasmid R6K. Even though the relative strength of the fiir promoter varies, the mutation, a C + A transversion in the -35 region (Figure 7) resulted in an approximately twofold reduction of luciferase activity in the five hosts examined. These data suggested that this particular base and the adjoining bases are involved in the initiation of transcription in all five organisms.

There is a great deal which remains to be learned concerning how the plasmid RK2 ensures its broad host range transcription and achieves the balanced gene expression required for its replication and stable maintenance in the variety of Gram-negative hosts. The results presented here represent an initial attempt at studying some of the parameters of transcription from a broad host range plasmid in Gram-negative bacteria other than E. coli. The plasmids designed for this study may be useful in identifying and determining the primary sequence of different classes of functional promoters in a variety of gram-negative bacterial species. The promoter fragments isolated and partially characterized in this study may contain members of some of these different classes. The use of different classes of promoters with differential ac-

tivity in a range of bacteria may be one of the strate- gies employed by the broad host range plasmid RK2 to ensure the appropriate level of gene expression for conjugal transfer, stable maintenance and antibiotic resistance in diverse genetic environments.

We thank R. DURLAND, F. FANG, G. DITTA, G. KASSAVETIS, E. VIRTS and E. LANKA for helpful discussions, B. KITTELL, M. BO" and A. TOUKDARIAN for critical evaluation of the manuscript, K. CARINE for artwork, K. WOOD and D. VELLUM for advice on performing luciferase assays and M. LEVITT for preparation of the manuscript. This work was supported by Public Health Service grant AI-07194 from National Institutes of Allergy and Infectious Diseases.

LITERATURE CITED BAGDASARIAN, M., R. LURZ, B. R~CKERT, F. C. H. FRANKLIN, M.

M. BAGDASARIAN, J. FREY and K. N. TIMMIS, 1981 Specific- purpose plasmid cloning vectors. 11. Broad host range, high copy number, RSF1010-derived vectors, and a host-vector system for gene cloning in Pseudomonas. Gene 16: 237-247.

BRETON, A. M., S. JAOUA and J. GUESPIN-MICHEL, 1985 Transfer of Plasmid RP4 to Myxococcus xanthus and evidence for its integration into the chromosome. J. Bacteriol. 161: 523-528.

BIRNBOIM, H. C., and J. DOLY, 1979 A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7: 1513-1521.

DEWET, J. R., K. V. WOOD, D. R. HELINSKI and M. DELUCA, 1985 Cloning of firefly luciferase cDNA and expression of active luciferase in E. coli. Proc. Natl. Acad. Sci. USA 82:

DITTA, G., S. STANFIELD, D. CORBIN and D. R. HELINSKI, 1980 Broad-host-range DNA cloning systems for Gram-negative bacteria. Construction of a gene back of Rhizobium meliloti. Proc. Natl. Acad. Sci. USA 77: 7347-7351.

FIGURSKI, D. H., R. F. POHLMAN, D. H. BECHHOFER, A. S. PRINCE and C. A. KELTON, 1982 Broad host range plasmid RK2 encodes multiple kil genes potentially lethal to Escherichia coli host cells. Proc. Natl. Acad. Sci. USA 7 9 1935-1939.

FREY, J., E. A. MUDD and H. M. KRISCH, 1988 A bacteriophage T 4 expression cassette that functions efficiently in a wide range of gram-negative bacteria. Gene 62: 237-247.

FURSTE, J. P., W. PANSEGRAU, G. ZIEGELIN, M. KROGER and E. LANKA, 1989 Conjugative transfer of promiscuous IncP plasmids: interaction of plasmid-encoded products with the transfer origin. Proc. Natl. Acad. Sci. USA 86: 1771-1775.

GAO, J., and G. N. GUSSIN, 1991 RNA Polymerase from P. aeruginosa and P. syringae respond to E. coli activator proteins. J. Bacteriol. 173: 394-397.

GUINEY, D. G., P. HASEGAWA and C. E. DAVIS, 1984 Plasmid transfer from Escherichia coli to Bacteroidesfragilis: differential expression of antibiotic resistance phenotypes. Proc. Natl. Acad. Sci. USA 81: 7203-7206.

GUINEY, D. G., and E. LANKA, 1989 Conjugative transfer of IncP plasmids, pp. 27-56 in Promiscuous Plasmids of Gram Negative Bacteria, edited by C. M. THOMAS. Academic Press, London.

HEFFRON, F., B. J. MCCARTHY, H. OHTSUBO and E. OHTSUBO, 1979 DNA sequence analysis of the transposon Tn3: three genes and three sites involved in transposition of Tn3. Cell 18:

KINGSBURY, D. T., and D. R. HELINSKI, 1973 Temperature-sensitive mutants for the replication of plasmids in Escherichia coli: requirements for deoxyribonucleic and polymerase I in the replication of the plasmid ColEl. J. Bacteriol. 114 1 116-1 124.

MANIATIS, T., E. F. FRITSCH and J. SAMBROOK, 1982 Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

7870-7873.

1153-1 163.


MCCLURE, W. R., 1985 Mechanism and control of transcription initiation in prokaryotes. Annu. Rev. Biochem. 54: 171-204.

MEYER, R., and M. HINDS, 1982 Multiple mechanisms for expression of incompatibility by broad host range plasmid RK2. J. Bacteriol. 152: 1078-1090.

PALOMARES, A., M. A. DELUCA and D. R. HELINSKI, 1989 Firefly luciferase as a reporter enzyme for measuring gene expression in vegetative and symbiotic Rhizobium meliloti and other Gram- negative bacteria. Gene 81: 55-64.

PINKNEY, M., B. D. M. THEOPHILUS, S. R. WARNE, W. C. A. TACON and C. M. THOMAS, 1987 Analysis of transcription from the trfA promoter of broad-host-range plasmid RK2 in Escherichia coli, Pseudomonas putida, and Pseudomonas aeruginosa. Plasmid 17: 222-232.

SAMBROOK, J., E. F. FRITSCH and T. MANIATIS, 1989 Molecular Cloning: A Laboratory Manual, Ed 2. Cold Spring Harbor Lab- oratory, Cold Spring Harbor, N.Y.

SANGER, F., S. NICKLEN and A. R. COULSON, 1977 DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci.

SCHMIDHAUSER, T. J., 1987 Ph.D. thesis. Identification of plasmid functions essential to the broad host range replication and maintenance of the IncP-1 plasmid RK2. University of Califor- nia, San Diego.

SCHMIDHAUSER, T. J., and D. R. HELINSKI, 1985 Regions of broad host range plasmid RK2 involved in replication and stable maintenance in nine species of Gram-negative bacteria. J. Bac- teriol. 164: 446-455.

SHAFFERMAN, A., R. KOLTER, D. STALKER and D. R. HELINSKI, 1982 Plasmid DNA replication. 111. Regulatory properties of the r initiation protein. J. Mol. Biol. 161: 57-76.

SHINGLER, V., and C. M. THOMAS, 1984 Transcription in the trfA region of broad host range plasmid RK2 is regulated by trfA and korB. Mol. Gen. Genet. 195: 523-529.

SMITH, C. A,, and C. M. THOMAS, 1989 Relationships and Evo- lution if IncP plasmids, pp. 57-77 in Promiscuous Plasmids of Gram Negative Bacteria, edited by C. M. THOMAS. Academic Press, London.

STALKER, D. M., R. KOLTER and D. R. HELINSKI, 1982 Plasmid

USA 74: 5463-5467.

R6K DNA Replication I. Complete nucleotide sequence of an autonomously replicating segment. J. Mol. Biol. 161: 33-43.

SUBRAMANI, S., and M. DELUCA, 1988 Applications of the firefly luciferase as a reporter gene, pp. 75-91 in Genetic Engineering, Principles and Methods, Vol. 10, edited by J. K. SETLOW. Plenum Press, New York.

THOMAS, C. M., and F. C. H. FRANKLIN, 1989 Gene expression signals in Gram-negative bacteria, pp. 229-246 in Promiscuous Plasmids of Gram Negative Bacteria, edited by C. M. THOMAS. Academic Press, London.

THOMAS, C. M., and D. R. HELINSKI, 1989 Vegetative replication and stable inheritance of IncP plasmids, pp. 1-25 in Promiscu- ous Plasmids of Gram Negative Bacteria, edited by C. M. THOMAS. Academic Press, London.

THOMAS, C. M., and C. A. SMITH, 1986 The trfB region of broad host range plasmid RK2: the nucleotide sequence reveals incC and key regulatory gene trjBlkorAlkorD as overlapping genes. Nucleic Acids. Res. 1 4 4453-4469.

THOMAS, C. M., and C. A. SMITH, 1987 Incompatibility group P plasmids: genetics, evolution, and use in genetic manipulation. Annu. Rev. Microbiol. 41: 77-101.

THOMAS, C. M., J. P. IBBOTSON, N. WANG, C. M. SMITH, R. TIPPING and N. M. LOADER, 1988 Gene regulation on broad host range plasmid RK2: identification of three novel operons whose transcription is repressed by both KorA and KorC. Nucleic Acids. Res. 1 6 5345-5359.

VIRTS, E. L., S. W. STANFIELD, D. R. HELINSKI and G. S. DITTA, 1988 Common regulatory elements control symbiotic and microaerobic induction of niJA in Rhizobium meliloti. Proc. Natl. Acad. Sci. USA 85: 3062-3065.

WOOD, K. V., and M. DELUCA, 1987 Photographic detection of luminescence in Escherichia coli containing the gene for firefly luciferase. Anal. Biochem. 161: 501-507.

ZHU, Y . S., and S. KAPLAN, 1985 Effects of light, oxygen, and substrates on steady-state levels of mRNA coding for ribulose- 1,5, biphosphate carboxylase and light-harvesting and reaction center polypeptides in Rhodopseudomonas sphaeroides. J. Bacte- rial. 162: 925-932.

Communicating editor: D. E. BERG

promoters of the broad host range plasmid rk2: analysis of ... · promoters of the broad host range...

Documents