m-Xylene-Responsive Pu-PnifH Hybrid {sigma}54 Promoters That Overcome Physiological Control in Pseudomonas putida KT2442

The sequences surrounding the –12/–24 motif of them-xylene-responsive ${sigma}$ ⁵⁴ promoter Pu of the Pseudomonas putidaTOL plasmid pWW0 were replaced by various DNA segments of thesame size recruited from PnifH ${sigma}$ ⁵⁴ promoter variants known tohave various degrees of efficacy and affinity for ${sigma}$ ⁵⁴-RNA polymerase(RNAP). In order to have an accurate comparison of the outputin vivo of each of the hybrids, the resulting promoters wererecombined at the same location of the chromosome of P. putidaKT2442 with a tailored vector system. The promoters includedthe upstream activation sequence (UAS) for the cognate regulatorof the TOL system (XylR) fused to the –12/–24 regionof the wild-type PnifH and its higher ${sigma}$ ⁵⁴-RNAP affinity variantsPnifH049 and PnifH319. As a control, the downstream region ofthe glnAp2 promoter (lacking integration host factor) was fusedto the XylR UAS as well. When the induction patterns of thecorresponding lacZ fusion strains were compared in vivo, weobserved that promoters bearing the RNAP binding site of PnifH049and PnifH319 were not silenced during exponential growth, asis distinctly the case for the wild-type Pu promoter or forthe Pu-PnifH variant. Taken together, our results indicate thatthe promoter sequence(s) spanning the –12/–24 regionof Pu dictates the coupling of promoter output to growth conditions.

INTRODUCTION

Top

Introduction

Pseudomonas putida strains harboring the TOL plasmid pWW0 areable to grow on toluene, m-xylene, and p-xylene as the onlycarbon source because of a highly regulated pathway which rendersbenzoate or toluate from these aromatic substrates into Krebscycle intermediates (49, 53). Expression of the upper TOL operonfor bioconversion of toluene, m-xylene, and p-xylene into thecorresponding carboxylic acids is driven by the ${sigma}$ ⁵⁴-dependentpromoter Pu (27) (Fig. 1). In the presence of suitable aromaticeffectors (e.g., m-xylene), this promoter is activated at adistance by the XylR protein, a member of the prokaryotic enhancerbinding protein family of transcriptional regulators (32, 64),with the assistance of integration host factor (IHF), whichfacilitates the appropriate Pu geometry (21) and helps the recruitmentof the RNA polymerase to the promoter (2, 8, 38) (Fig. 1).

View larger version (17K):
[in this window]
[in a new window]
FIG. 1. Organization of the ${sigma}$ ⁵⁴ promoter Pu of P. putida plasmid pWW0. The distribution of relevant DNA sequences and their coordinates in respect to the transcription initiation site, as well as some important restriction sites, is shown. The region includes the UAS for XylR, the –12/–24 region recognized by ${sigma}$ ⁵⁴-RNAP, the IHF binding site located within the intervening region, and the adjacent UP-like elements (38) for docking of the ${sigma}$ ⁵⁴-RNAP. Protein sizes are symbolic. The upstream nucleoprotein complex may contain six or seven monomers of the regulator. The locations of the new EcoRI sites in pFH14 (–44) and pFH15 (–106) are indicated as well.

Pu activity in vivo not only requires the presence of XylR effectorsin the medium, it is also strongly dependent on the metabolicstatus of the cell. An excess of certain carbon sources (16,29) or rapid growth in rich medium inhibits the activity ofthe promoter in vivo even if the aromatic inducer is presentin the culture (15, 20, 22, 23, 31, 40, 53). At least four distinctelements appear to take part in such down-regulation. First,the presence of glucose and other carbohydrates (29) inhibitsPu activity through a process which involves the ptsN gene,encoding the IIA^Ntr protein of the phosphoenolpyruvate-sugarphosphotransferase system (16, 48). Second, rapid growth inrich medium (for instance, Luria-Bertani [LB] medium) restrainsthe performance or activity of the ${sigma}$ ⁵⁴ protein (15). This isrevealed by the fact that overproducing ${sigma}$ ⁵⁴ largely relievesphysiological inhibition of Pu. Third, intracellular concentrationsof the alarmone (p)ppGpp, a molecule involved in the stringentresponse (17), have a moderate stimulatory effect on the activityof Pu in vivo and in vitro (11). This outcome is mechanisticallyrelevant, as ppGpp appears to regulate sigma factor competitionfor the scarce core RNAP during stationary phase (33), and itfavors the entry of ${sigma}$ ⁵⁴ into the enzyme under conditions of aminoacid starvation (37). Finally, Pu (Fig. 1) is entirely dependenton IHF (7), which binds its target site only during stationaryphase in the Pu promoter (61). Other factors (for example, TurA)bind Pu as well, entering additional environmental inputs, suchas responsiveness to low temperatures (51).

Perhaps because of all the somewhat redundant mechanisms mentionedabove, it has been impossible so far to isolate a single P.putida mutant in which Pu could be altogether free of physiologicalcontrol. But is the contrary true as well (i.e., can Pu variantsdevoid of metabolic coregulation be generated)? Previous reportshave shown that changes in the promoter –12/–24sequences lead to variations in promoter performance in vitroand in vivo (18, 19, 55, 62, 63). In this work, we created anumber of hybrid promoters between the upstream activation sequence(UAS) and IHF sequence of Pu and the ${sigma}$ ⁵⁴-RNAP binding regionsfrom PnifH049 and PnifH319 promoters of Klebsiella pneumoniae(two variants of PnifH bearing nucleotide changes that increasethe activity of the promoter) (5, 50, 55) or from the glnAp₂promoter of Escherichia coli, which is independent of IHF forits activation (44). Our data show that Pu variants with such–12/–24 regions abolish growth phase-dependent controlof transcription, suggesting that physiological regulation ofthis promoter largely reflects the engagement of RNAP with the–12/–24 motif.

MATERIALS AND METHODS

Top

Materials and Methods

Strains and general procedures. The E. coli and P. putida strains and plasmids used in thiswork are listed in Table 1. The sequences and schemes of thehybrid promoters used are drawn in Fig. 2. The strains E. coliCC118 supF (35) and E. coli HB101(pRK600) were used as the hostfor pBK16 derivatives (Table 1) and as the helper for oriT-mediatedmobilization, respectively. Hybrid promoters were mobilizedinto strain P. putida KT2442 hom.fg. xylRS (35) and integratedin its chromosome as explained below. The strain derived fromP. putida KT2442 hom. fg. xylRS but bearing the wild-type promoterfusion Pu-lacZ was named P. putida SF05X. Equivalent P. putidastrains bearing other fusions were designated P. putida MR05X(Pu-PnifH-lacZ), P. putida SF02X (Pu-PnifH319-lacZ), P. putidaMR02X (Pu-PnifH049-lacZ), and P. putida SF03X Pu-PglnAp₂-lacZ.Recombinant DNA techniques were carried out according to publishedprotocols (54).

View this table:
[in this window]
[in a new window]
TABLE 1. Strains and plasmids used in this work

View larger version (56K):
[in this window]
[in a new window]
FIG. 2. Schematic representation of Pu variants and hybrid promoters. The reference wild-type Pu promoter is shown on top, along with an indication of functionally important segments and the restriction sites engineered for constructing the variants. A blowup of the sequence of the –12/–24 region spanning the essential GG-GC nucleotides is also shown. The other promoters (hybrids of Pu with various segments of PglnAp₂, PnifH, PnifH049, and PnifH319) are displayed with the reference UAS sequence (two quasipalindromic binding sites), which is shared by all of them, and a grey tone code to trace each of the segments to the correct donor of the sequence. Note the nucleotide changes within positions –17/–15 of promoters with ${sigma}$ ⁵⁴-RNAP binding variants PnifH049 and PnifH319.

Plasmid construction. The lacZ plasmid pBK16 (35) was used as the vector for the assemblyof the different hybrid promoters. This mobilizable plasmidbears a supF-supressible resistance to the streptomycin-spectinomycingene (aadR) and also a supF-supressible promoterless lacZ gene,in front of which the promoter of interest is cloned. Thesefeatures confer host supF⁺ E. coli ${Delta}$ lac strains (such as E. coliCC118 supF) resistance to streptomycin and lacZ⁺ phenotypesbut make the plasmid unstable in supF^o strains. The Pu-lacZdelivery plasmid pBK16Pu was constructed by cloning the 312-bpEcoRI/BamHI fragment from pEZ9 spanning the Pu promoter fromthe TOL plasmid pWW0 (21). Plasmids bearing hybrids betweenthe Pu promoter and downstream PnifH variants or glnAp₂ weremade as follows. First, Pu was cloned as a 312-bp EcoRI/BamHIfragment from pEZ9 into the site-directed mutagenesis vectorpCG2 (43), yielding pCG2 Pu. Novel EcoRI sites were enteredat positions –44 and –106 of the Pu promoter sequence(36), yielding plasmids pFH14 and pFH15, respectively (Fig.2). The source of the PnifH ${sigma}$ ⁵⁴-RNAP binding region was pMB1(5). This plasmid was subject to PCR with oligonucleotides 10(5'-ATGAATTCACAGGCACGGCT-3') and 11 (5'-GACGGGGATCCATGGTGACTTCT-3').This amplified an 88-bp EcoRI/BamHI DNA segment spanning the–44 to +33 region of PnifH, thus including the –24/–12motif of the promoter. This segment was cloned in pBK16, producingplasmid pMJ1. The EcoRI insert of pFH14 (spanning coordinates–207 to –44 of Pu, including the UAS and the IHFbinding site) was then ligated to EcoRI-digested pMJ1, and thecorrectly oriented insert gave rise to the Pu-PnifH-lacZ deliveryplasmid pMJ2 (Fig. 2). The source of the –12/–24region of PnifH319 was plasmid pJES366, which bears a 320-bpEcoRI/BamHI insert spanning the whole PnifH region of K. pneumoniaewith two C-to-T changes at positions –15 and –17(Fig. 2), which increase affinity for ${sigma}$ ⁵⁴-RNAP, all cloned ina pTZ18R vector (55). Site-directed mutagenesis (36) of thisplasmid with the oligonucleotide SF4 (5'-ATAAGAATGAATTCACAGGCACGGC-3')generated a new EcoRI site in position –44, producingthe plasmid pJE336-SF4. The 88-bp EcoRI/BamHI fragment of pJE336-SF4spanning coordinates –44 to +33 was then excised and clonedin the corresponding sites of pBK16, yielding pFH44A. The EcoRIinsert of pFH14 (spanning coordinates –207 to –44of Pu) was then ligated to the EcoRI-digested pFH44A, and thecorrectly oriented insert gave rise to the Pu-PnifH319-lacZdelivery plasmid pFH44 (Fig. 2). Similarly, a DNA segment spanningcoordinates –44 to +33 of the PnifH variant PnifH049,in which three C-to-T changes at positions –15 to –17had been made for increased affinity to ${sigma}$ ⁵⁴-RNAP, was generatedfrom plasmid pWC88049(5). This segment was assembled as beforein pBK16, along with the 168-bp EcoRI insert of pFH14, givingrise to the Pu-PnifH049-lacZ delivery plasmid pRF2. Finally,the UAS-less region –106 to +19 of the glnAp₂ promoterof E. coli was amplified from genomic DNA with oligonucleotidesgln1 (5'-CCCCCGAATTCCAACATTCAGATCGTGGTGC-3') and gln2 (3'-AAATGCCGCTGTGCCGGTTTCCACTGGCCCCC-5').The product was cloned as a 142-bp EcoRI-BstEII fragment inthe equivalent sites of pEZ9 (yielding pEZ9-PCR) and reclonedin pBK16 as a 240-bp EcoRI/BamHI insert, generating pFH43A.This plasmid was digested with EcoRI and ligated to the 100-bpEcoRI fragment of pFH15 (which spanned the UAS for XylR in Pu)(Fig. 2). The correctly oriented insert gave rise to the Pu-glnAp₂delivery plasmid pFH43 (Fig. 2). The plasmids used in the transcriptionassays (Table 1) were produced by cloning the promoter-bearinginserts of interest in vector pTE103, which adds a strong T7terminator downstream of the promoter under study (24). Allcloned inserts and DNA fragments were verified through automatedDNA sequencing in an Applied Biosystems device.

Growth and induction conditions. Bacteria were grown at 30°C in either rich LB medium (41)or M9 minimal medium (54) amended with 0.2% Casamino Acids and0.2% glucose, as indicated. When required, the media were supplementedwith 150 µg of ampicillin/ml, 50 µg of streptomycin/ml,50 µg of spectinomycin/ml, 30 µg of chloramphenicol/ml,or 50 µg of kanamycin/ml. Promoter activity in vivo wasmonitored in all cases by assaying the accumulation of ß-galactosidasein cells permeabilized with chloroform and sodium dodecyl sulfate,as described by Miller (41), under the conditions specifiedin each case. The ß-galactosidase activity valuesrepresent the averages of at least three independent experimentsin duplicate samples; standard deviations were <15%. Whereindicated, the cultures were exposed to saturating vapors ofthe upper TOL pathway inducer m-xylene (1).

Proteins and protein techniques. Purified factor ${sigma}$ ⁵⁴ and core RNAP from E. coli were the kindgift of B. Magasanik. Purified IHF protein was obtained fromH. Nash. XylR ${Delta}$ A was purified to apparent homogeneity by metalloaffinityof the His-tagged protein, as described by Pérez-Martínand de Lorenzo (46).

In vitro transcription assays. Plasmids used in the transcription assays (Table 1) were preparedwith the QIAGEN (Valencia, Calif.) plasmid purification system.Transcription assays were performed following previously publishedprocedures (18). Supercoiled DNA templates were used at 5 nMconcentration; 50-µl reaction mixtures were set at 37°Cin a buffer consisting of 50 mM Tris-HCl (pH 7.5), 50 mM KCl,10 mM MgCl₂, 0.1 mM bovine serum albumin, 10 mM dithiothreitol,and 1 mM EDTA. Unless otherwise indicated, each DNA templatewas premixed with 25 nM core RNAP, 100 nM ${sigma}$ ⁵⁴, 25 nM IHF, and100 nM XylR ${Delta}$ A. The DNA templates and the proteins were incubatedat 37°C with 4 mM ATP for 20 min to allow open-complex formation.Transcription was then initiated by adding a mixture of ATP,CTP, GTP (400 µM each), and UTP (5 mCi; 3,000/mmol). Insingle-round experiments, heparin (0.1 mg/ml) was added alongwith the nucleoside triphosphate mixture to prevent reinitiation.After the mixtures were incubated for 10 min at 37°C, thereactions were stopped with equal volumes of a solution containing50 mM EDTA, 350 mM NaCl, and 0.5 mg of carrier tRNA/ml. ThemRNA was then extracted, precipitated with ethanol, electrophoresedon a denaturing 7 M urea-4% polyacrylamide gel, and visualizedby autoradiography.

Mobilization and recombination of hybrid promoters in single-gene dosage. To generate P. putida strains harboring a monocopy fusion ofthe hybrid promoter Pu-PnifH, Pu-PnifH319, Pu-PnifH049, or Pu-PglnAp₂,or the wild-type Pu promoter to lacZ, E. coli CC118 supF harboringthe pBK16Pu (Pu-lacZ), pMJ2 (Pu-PnifH-lacZ), pRF2 (Pu-PnifH049-lacZ),pFH44 (Pu-PnifH319-lacZ), or pFH43 (Pu-PglnAp₂-lacZ) plasmidwas mobilized into the P. putida target strain KT2442 hom.fg.xylRS by tripartite mating using a filter technique with E.coli HB101(pRK600) as the helper strain (28). After 8 h of incubationat 30°C on LB plates, the cells were washed with 10 mM MgSO₄and plated on M9 citrate medium in the presence of streptomycin.The streptomycin-resistant exconjugants that arose by cointegrationof the hybrid fusions (Fig. 3) were regrown and screened forkanamycin-sensitive blue colonies in medium with X-Gal (5-bromo-4-chloro-3-indolyl-ß-D-galactopyranoside).This phenotype is the necessary result of a double homologousrecombination between the plasmid and the P. putida chromosomebearing the hom. fg. segment, as described by Kessler et al.(35) (Fig. 4).

View larger version (29K):
[in this window]
[in a new window]
FIG. 3. Formation of transcriptionally open complexes by Pu variants. The results of single-round transcription assays with 5 nM (each) supercoiled pTE103 derivatives inserted with the Pu hybrids (Table 1) are indicated in each case. The experiment was run as explained in Materials and Methods with 100 nM XylR ${Delta}$ A, 25 nM core RNAP, 100 nM ${sigma}$ ⁵⁴, and, where indicated (+), 25 nM IHF. The reactions were initiated by addition of the four nucleoside triphosphates in the presence of heparin. Under these conditions, Pu produces a transcript of 394 nucleotides (nt), Pu-PnifH and its derivatives produce a transcript of 336 nt, and Pu-PglnAp2 produces a transcript of 377 nt.

View larger version (35K):
[in this window]
[in a new window]
FIG. 4. Integration of Pu and its hybrid variants into the chromosome of P. putida KT2442 hom. fg. (A) The so-called homology fragment is stably inserted into the chromosome of P. putida KT2442 by means of a hybrid mini-Tn10 transposon (35). This fragment contains a selectable kanamycin resistance gene (aphA) flanked by an N-terminally truncated aadA gene (streptomycin-spectinomycin resistance) and a divergently oriented and also N-truncated lacZ gene. These DNA segments provide homology to cognate sequences in the lacZ transcriptional fusion vector pBK16 (B) containing the promoter of interest. RP4 oriT-mediated mobilization of pBK16 derivatives into P. putida KT2442 hom. fg. (which harbors the homology fragment) allows double recombination, leading to the transfer of the promoter into the chromosome. The double crossover is selected by streptomycin-spectinomycin resistance and blue color on X-Gal and further confirmed by the loss of kanamycin resistance (C). This event is facilitated by the presence of amber codons (TAG) at the ends of the addA and lacZ genes of pBK16. Some functionally important elements of the system are indicated. The resulting lacZ fusion is transcriptionally shielded upstream by an ${Omega}$ streptomycin-spectinomycin interposon and downstream by a strong T7 terminator (not shown). (D) Scheme of the second insert borne by the strain P. putida KT2442 hom. fg. xylRS. This is a mini-Tn5 Hg vector with a $~$ 2.5-kb segment of the pWW0 plasmid encoding the two regulators of the TOL system, xylR and xylS, in its natural divergently transcribed configuration.

RESULTS AND DISCUSSION

Top

Results and Discussion

Organization of hybrid promoters and validation of their functionality in vitro. The salient features of the promoters under scrutiny in thiswork are depicted in Fig. 2. They all share the binding sitesfor XylR (the UAS and adjacent sequences –205 to –106of the wild-type Pu promoter) placed at identical distancesupstream from the –12/–24 motif that is recognizedby the ${sigma}$ ⁵⁴-RNAP. The first hybrid promoter (Pu-PnifH) bears thewhole Pu upstream region (coordinates –205 to –44)spanning its native UAS and an IHF binding site fused to the–44 to +33 sequence of the PnifH promoter, which providesthe –12/–24 region. In its natural context, PnifHis also stimulated by IHF, and this effect was more apparentwhen the DNA template was linear (30). Since the distances betweenthe UAS, IHF, and –12/–24 motif are retained inPu-PnifH as in Pu, one can safely assume that the functionalarchitecture of the hybrid promoter is kept as well, and thusthat the only significant change affects the ${sigma}$ ⁵⁴-RNAP bindingsite. Two other hybrid promoters (Pu-PnifH049 and Pu-PnifH319)are identical to Pu-PnifH except for the sequences in the –15/–17region within the ${sigma}$ ⁵⁴-RNAP binding site. The nucleotides –17and –15 of ${sigma}$ ⁵⁴-dependent promoters are involved in modulatingthe recognition and binding of the polymerase to the wider –12/–24region (5, 6). The wild-type PnifH has a CCC in –17/–15,while PnifH319 and PnifH049 have TCT and TTT sequences, respectively,at the same positions (5, 45, 55) (Fig. 2). Both PnifH variantspossess superior promoter strength (30, 42, 55) and, unlikewild-type PnifH, they do not need IHF for transcription in vitrowith either supercoiled or linear DNA (55). The last hybridpromoter constructed, Pu-PglnAp₂, was the result of replacingthe region –106 to +19 of Pu (which includes the IHF siteand the –12/–24 region in this promoter) with thesame coordinates of glnAp₂. It is known that glnAp₂ lacks anIHF site and thus is totally independent of IHF for the initiationof transcription (44) in either supercoiled or linear templates(10). Furthermore, the sequence –17/–15 (TTT) ofglnAp₂ matches those present in strong ${sigma}$ ⁵⁴ promoters (4, 5, 42,62).

In order to examine whether hybrid promoters were functional,we performed single-round in vitro transcription assays usingas supercoiled DNA templates plasmids bearing each of the promoterscloned in vector pTE103 (see Materials and Methods) (Table 1)under conditions described previously (9, 46). The activatorprotein included in these assays was XylR ${Delta}$ A, a variant of thewild-type XylR with its N-terminal module (i.e., its A domain)deleted. This variant is constitutively active and can thuspromote transcription in the absence of any aromatic inducer(25, 46, 47). As shown in Fig. 3, all promoters were able toinitiate transcription, although their relative efficienciesvaried. As expected, Pu activity was absolutely dependent onIHF, a feature that is completely preserved in vivo (7). Thismay be due to the indispensable need for IHF in Pu for recruitmentof ${sigma}$ ⁵⁴-RNAP to the promoter (2, 9, 61). Unlike Pu, the Pu-PnifHhybrid could form a significant amount of transcript in theabsence of IHF, which was further stimulated by addition ofthe factor (however, dependence on IHF did increase on a lineartemplate [data not shown]). Similarly, Pu-PnifH049 and Pu-PnifH319initiated transcription efficiently on both supercoiled andlinear templates independently of IHF, although the factor hada detectable effect in Pu-PnifH319 (Fig. 3 and data not shown).Pu-PglnAp₂ also worked well in a fashion completely independentof IHF. These in vitro assays validated the capacities of thevarious hybrid constructs to bear transcription with the minimalset of components that suffice to activate Pu (46). However,since the assay system included mostly purified components fromE. coli, the relative amounts of transcripts under various conditionscannot be simply projected onto the situation in vivo in P.putida. We thus examined this issue in the native context, asexplained below.

Setup of a dependent genetic system to follow the activity of the Pu promoter in P. putida. In order to ensure the maintenance of the regulatory elementsacting on Pu in its native gene dose and stoichiometry, we employeda genetic system tailored for site-specific integration of lacZfusions into the chromosome of P. putida (35). This procedure(sketched in Fig. 4) allows a faithful comparison of the transcriptionaloutput of any given Pu promoter variant in vivo. The basis ofthis system is the presence in the target chromosome of a DNAfragment that provides sequence homology to the regions flankingthe lacZ fusion assembled in the delivery vector pBK16 (Table1). This system ensures that all promoters are inserted at thesame chromosomal location, in the same orientation, and shieldedfrom external transcriptional flows by an upstream ${Omega}$ elementand a downstream T7 terminator so that their activities canbe exactly compared. To examine whether this setup allowed thereproduction of the physiological control phenomenon of Pu ina fashion amenable to genetic scrutiny, we mobilized plasmidpBK16Pu toward P. putida strain KT2442 hom. fg. xylRS, whichbears in its chromosome insertions of the above-mentioned homologyfragment, as well as a DNA segment encoding both XylR and XylSproteins (Fig. 4D). Chromosomal recombination of the Pu-lacZfusion of pBK16Pu gave rise to strain P. putida SF05X, whichbears all the regulatory constituents of Pu in monocopy. Tovalidate the use of this strain as a reference, we ran the experimentshown in Fig. 5, in which ß-galactosidase accumulationof P. putida SF05X (Pu-lacZ xylR⁺) was followed during the growthcurve in rich medium in the presence or absence of saturatingm-xylene vapors. As shown in Fig. 5, even when P. putida SF05Xwas induced since early in culture, Pu remained basically silentuntil the cells reached an optical density at 600 nm of $~$ 1.2,when a distinct burst of transcription occurred. These datavalidated the use of the chromosomal integration system of Fig.4 as the preferred tool to judge the effect of replacing the ${sigma}$ ⁵⁴-RNAP binding region of Pu with others from various origins.

View larger version (34K):
[in this window]
[in a new window]
FIG. 5. Evolution of Pu activity during growth in rich medium. P. putida SF05X cells bearing all elements required for Pu regulation assembled in the chromosome by the procedure explained in the text (sketched on top) were grown overnight at 30°C in complete LB medium, diluted to an optical density at 600 nm (OD₆₀₀) of $~$ 0.05, and regrown under the same conditions in the presence or absence of saturating vapors of m-xylene. ß-Galactosidase levels were followed during growth as shown. Note that the promoter remained fully inhibited (as reflected by ß-Gal output) until the cultures entered stationary phase.

Physiological control of Pu promoter variants with diverse –12/–24 regions. In order to study the physiological regulation of the Pu hybridswith RNAP binding sequences recruited from PnifH (and its variants)and glnAp₂, we constructed P. putida strains completely identicalto P. putida SF05X (Pu-lacZ xylR⁺) except for the region downstreamof the UAS for XylR (Fig. 2). For this, we mobilized plasmidspMJ2 (Pu-PnifH-lacZ), pRF2 (Pu-PnifH049-lacZ), pFH44 (Pu-PnifH319-lacZ),and pFH43 (Pu-PglnAp₂-lacZ) toward P. putida KT2442 hom. fg.xylRS and forced the chromosomal recombination of the lacZ fusionsas before. These operations resulted in the strains P. putidaMR05X (Pu-PnifH-lacZ xylR⁺), P. putida MR02X (Pu-PnifH049-lacZxylR⁺), P. putida SF02X (Pu-PnifH319-lacZ xylR⁺), and SF03X(Pu-PglnAp₂-lacZ xylR⁺). Every strain was then subjected toan analysis of ß-galactosidase accumulation duringgrowth in the presence or absence of m-xylene vapors as describedabove.

The results in Fig. 6 show that each promoter behaved in a differentway in respect to both the induction pattern and the strengthof transcription. The fusion between the UAS and the IHF regionof Pu followed by the –12/–24 sequence of the wild-typePnifH promoter (Fig. 6a) acted in vivo in a fashion nearly identicalto that of the original Pu promoter (Fig. 5), i.e., there wasan evident repression of transcription during fast growth, followedby a timely boost of activity at the onset of stationary phase.The induction pattern for Pu-PnifH could largely reflect, asis the case for Pu (61), the growth phase-dependent occupationof the IHF site and the ensuing recruitment of the polymeraseto the –12/–24 region (38).

View larger version (29K):
[in this window]
[in a new window]
FIG. 6. In vivo performance of Pu variants. P. putida KT2442 hom. fg. xylRS strain derivatives bearing the chromosomal fusion Pu-PnifH-lacZ (a), Pu-PnifH049-lacZ (b), Pu-Pnif319-lacZ (c), or Pu-PglnAp₂-lacZ (d) were grown in LB medium at 30°C and assayed for ß-galactosidase activity (expressed in Miller units). The growth rates of all the strains were indistinguishable under all conditions tested. OD₆₀₀, optical density at 600 nm.

Given the sequence divergence between the –44/+33 regionsof Pu and PnifH (except the actual –12/–24 motif),the fact that silencing is preserved in the Pu-PnifH hybridrules out any influence of extra factors binding that regionin a sequence-specific manner. Otherwise, this result says nothingabout the role of the –12/–24 region in physiologicalcontrol. In contrast, comparison of the induction profiles ofPu-PnifH (Fig. 6a) with those of Pu-PnifH049 (Fig. 6b) and Pu-PnifH319(Fig. 6c) was informative. As mentioned above, the –12/–24regions of Pu-PnifH049 and Pu-PnifH319 differ from that of thewild-type PnifH by only a number of bases within the –15/–17coordinates that appear to increase their binding to the holoenzymeand facilitate the formation of an open complex in vitro (5,55). These two aspects are problematical to separate in vivofor any given sequence, because higher affinity does not translateautomatically into superior activity (62, 63). We prefer theoperative term "engagement" to describe the combination of affinityand ease of open-complex formation that is inherent in every–12/–24 region variant.

The one outstanding aspect of Fig. 6b is the lack of any significantinhibition of lacZ production at any growth stage. Unlike Pu(Fig. 5) and Pu-PnifH, the hybrid Pu-PnifH049 appears to respondto m-xylene as soon as the cells are exposed to the inducer.ß-Galactosidase accumulation then follows a steadyincrease that is perhaps greater at late growth stages (Fig.6b). In other words, the Pu-PnifH049 hybrid seems to be relievedfrom any physiological down-regulation while reaching ß-galactosidaselevels in the same range as those of the reference promotersPu (Fig. 5) and Pu-PnifH (Fig. 6a). Since the only differencebetween Pu-PnifH and Pu-PnifH049 is 3 bases within the –12/–24region that affect the affinity and the quality of the interactionwith ${sigma}$ ⁵⁴-RNAP, we argue that physiological control of Pu mayreflect the binding of the enzyme, the only step which can beregulated in vivo for any fixed DNA sequence. This picture isreinforced by the behavior of the related hybrid promoter Pu-PnifH319(Fig. 6c), which also appears to be free of physiological down-regulationwhile exhibiting somewhat lower transcriptional activity onthe whole. It thus appears that carrying such improved –12/–24sequences accounts completely for the release of any silencingof Pu during rapid growth.

Finally, we examined the induction pattern of the hybrid promoterbetween the UAS of Pu and the rest of the DNA sequence, allthe way to the –12/–24 motif, from the naturallyIHF-less promoter glnAp₂ (Fig. 2). As mentioned before, thispromoter forms a stable complex in vitro with the ${sigma}$ ⁵⁴-RNAP ofE. coli (44). On this basis, we examined the inducibility ofthe hybrid Pu-glnAp₂ (Fig. 2) in the same system employed invivo before. It should be noted that the range of the transcriptionaloutputs in this case (Fig. 6d) is within much lower ß-galactosidaseactivities. This is not unexpected, since the extensive sequenceexchange between Pu and glnAp₂ (Fig. 2) may flaw the geometryof the promoter, which, however, keeps the same distances andphasing between the UAS and the –12/–24 region asall the other promoters tested. While comparing absolute activitiesis, for that reason, not informative, Pu-glnAp₂ still showsa revealing induction profile (Fig. 6d). This consists of arapid (but relatively low) response to m-xylene from early ingrowth, which appears to be increased at the onset of stationaryphase.

C source (glucose) inhibition of Pu promoter variants. As mentioned above, the phenomenon that we refer to as physiologicalcontrol of Pu is the result of processing various environmentalconditions. One of them is the presence in the medium of somecarbon sources (in particular, glucose), which down-regulatesPu output in a fashion phenomenologically similar to catabolicrepression but mechanistically quite different (52). In fact,it has been possible to distinguish the effect of glucose onPu from other growth phase-related inputs either genetically(14) or by using a chemostat that fixes growth rates (22, 23).

Interestingly, Pu inhibition brought about by C sources is specificallycaused by carbohydrates metabolized through the Entner-Doudoroffpathway, such as glucose or gluconate, whereas organic acids,such as citrate or succinate, lack this negative influence (16,29). To examine whether the relief of Pu silencing under rapidgrowth in rich medium brought about by altering the ${sigma}$ ⁵⁴-RNAPbinding site could also overcome down-regulation by glucose,we ran the experiment shown in Fig. 7. In this case, inductionexperiments were done under conditions in which the growth ratesof all strains were identical because of the addition to theminimal medium of an excess of amino acids (16). Any differencein lacZ could thus be traced to the presence or absence of glucosein the medium. In this case, xylR⁺ P. putida strains bearingPu-lacZ, Pu-PnifH-lacZ, Pu-PnifH049-lacZ, Pu-PnifH319-lacZ,or Pu-PglnAp₂-lacZ fusions were induced with m-xylene for 15h, and the accumulation of ß-galactosidase was recorded.As shown in Fig. 7, the control Pu-lacZ fusion and its Pu-PnifHcounterpart, which are subject to intense physiological inhibition(Fig. 5 and 6a), behaved in the standard manner in this assay,namely, 30 to 60% reduction of transcriptional output in thepresence of glucose (16). On the other hand, the hybrids Pu-PnifH049and Pu-PnifH319 kept, and even went beyond (40 to 70% reduction),the standard response to the carbohydrate, despite being altogetherfree of growth phase inhibition (Fig. 6). Finally, the Pu-PglnAp₂fusion was basically blind to the presence of glucose in themedium. These results reveal the independence of the carboninhibition of Pu in respect to the mechanism(s) which releasesthe growth phase-dependent regulation (14).

View larger version (37K):
[in this window]
[in a new window]
FIG. 7. Glucose repression of transcriptional activities of Pu variants. P. putida KT2442 hom. fg. xylRS strain derivatives' lacZ fusions to the promoters indicated on top of the diagram bars were grown for 15 h at 30°C in M9 medium supplemented with 0.2% Casamino Acids to equal growth rates with (+) and without (–) 0.2% glucose (glu) in the presence of saturating vapors of m-xylene.

Conclusion. Pu activity in vivo is not just dependent on the regulator-promoterpair which suffices to cause transcription in vitro but alsoon the overall metabolic and energy status (12, 13, 53, 57).Such a physiological check of the Pu promoter probably involvesa number of mechanisms (8, 11, 13, 16, 29, 31). One aspect isthe role of the sequence –12/–24 in such physiologicalcontrol of Pu. It has been known for a long time that changesin the sequence bound by ${sigma}$ ⁵⁴-RNAP alter the performance of other ${sigma}$ ⁵⁴ promoters in vivo and in vitro (18, 19, 42, 55). In thiscontext, the main piece of information reported in this articleis that the silencing of the Pu promoter of the TOL plasmidwhen cells grow exponentially in rich medium (15) can be defeatedby exchanging the native –12/–24 region for an equivalentsequence of promoters known to have a different degree of engagementwith the polymerase.

How does this notion fit with the rest of the data availablefor Pu? This question finds a suitable context by comparingresults from Pu itself with those of the similar (but not identical) ${sigma}$ ⁵⁴ promoter Po, which drives the expression of a pathway fordegradation of dimethyl phenols when the cognate activator (akinto XylR), named DmpR, binds the corresponding aromatic effectors(37, 56-60). Growth phase control of Po has been attributedto the ability of the alarmone ppGpp to facilitate the accessof ${sigma}$ ⁵⁴ to the core RNAP during sigma factor competition at stationaryphase (33, 37). Such competition is critical in the case of ${sigma}$ ⁵⁴, given that only a few molecules of the factor ( $~$ 80) are presentin the cell at any growth stage (34). In this context, it ispossible that ${sigma}$ ⁵⁴ promoters with a better –12/–24region can function even at the low concentrations of ${sigma}$ ⁵⁴-RNAPavailable prior to stationary phase. In these instances (asseems to be the case for the Pu-PnifH049 and Pu-Pnif319 variants),the promoter may not be subject to any physiological inhibition,but it is active throughout the growth curve. However, it shouldbe noted that (as shown in Fig. 7) defeating the down-regulationof Pu due to growth phase effects does not imply the same forthe C source control of the same promoters. The mechanisms behindthese phenomena will be the subject of future investigations.

ACKNOWLEDGMENTS

We are indebted to B. Magasanik, H. Nash, M. Buck, and E. Santerofor sharing valuable proteins and plasmids. Inspiring discussionswith V. Shingler and I. Cases are gratefully acknowledged.

This work was supported in part by EU grants BIOCARTE, LINDANE,and ACCESS and by Project BIO2001-2274 of the Spanish CICYT.

FOOTNOTES

* Corresponding author. Mailing address: Centro Nacional de Biotecnología del CSIC, Campus de Cantoblanco, 28049 Madrid, Spain. Phone: 34 91 585 4536. Fax: 43 91 585 4506. E-mail: vdlorenzo{at}cnb.uam.es.

REFERENCES

Top