Effects of Combination of Different −10 Hexamers and Downstream Sequences on Stationary-Phase-Specific Sigma Factor ς^S-Dependent Transcription in Pseudomonas putida

Abstract

The main sigma factor activating gene expression, necessary in stationary phase and under stress conditions, is ς^S. In contrast to other minor sigma factors, RNA polymerase holoenzyme containing ς^S (Eς^S) recognizes a number of promoters which are also recognized by that containing ς⁷⁰ (Eς⁷⁰). We have previously shown that transposon Tn4652 can activate silent genes in starving Pseudomonas putida cells by creating fusion promoters during transposition. The sequence of the fusion promoters is similar to the ς⁷⁰-specific promoter consensus. The −10 hexameric sequence and the sequence downstream from the −10 element differ among these promoters. We found that transcription from the fusion promoters is stationary phase specific. Based on in vivo experiments carried out with wild-type and rpoS-deficient mutant P. putida, the effect of ς^S on transcription from the fusion promoters was established only in some of these promoters. The importance of the sequence of the −10 hexamer has been pointed out in several published papers, but there is no information about whether the sequences downstream from the −10 element can affect ς^S-dependent transcription. Combination of the −10 hexameric sequences and downstream sequences of different fusion promoters revealed that ς^S-specific transcription from these promoters is not determined by the −10 hexameric sequence only. The results obtained in this study indicate that the sequence of the −10 element influences ς^S-specific transcription in concert with the sequence downstream from the −10 box.

In their natural environment, most bacteria are challenged by widely changing nutrient availability and by exposure to various forms of physical stress (temperature shock, oxidative stress, etc.). When starvation or various other stress factors cause a reduction or cessation of growth, many genes are shut down while others are induced to help the cells to survive. One way to modulate gene expression is replacement of the main sigma factor with an alternative sigma factor that recognizes a specific promoter of the stimulus response gene (reviewed in reference 19). In Escherichia coli, ς^S regulates the expression of more than 100 genes involved in cell survival in the stationary phase and in response to different stresses (reviewed in references 14 and 15). The rpoS gene encoding ς^S has been described also for nonenteric bacteria, e.g., fluorescent pseudomonads (21, 36, 37, 40). Despite clearly different physiological roles, ς^S is similar to the major sigma factor ς⁷⁰ in terms of structure and molecular function (7, 26, 42). No clear differences between ς⁷⁰- and ς^S-dependent promoters are apparent. A compilation of ς^S-dependent promoters deduced a −10 consensus sequence, CTATACT, that is slightly different from the typical TATAAT sequence recognized by ς⁷⁰ (11). However, the binding patterns of Eς^S and Eς⁷⁰ revealed by DNase I protection experiments are not completely identical (29, 41). Eς^S appears to be less dependent on contacts in the −35 region (7, 17, 41). The activity of Eς^S and Eς⁷⁰ is differentially influenced by salt concentrations and by the degree of negative supercoiling of the DNA template (2, 10, 23). Additionally, a number of global regulators and histone-like proteins, such as H-NS, Lrp, CRP, IHF, and Fis, are involved in determination of sigma factor specificity (29; see also references 14 and 15 for a review and references cited therein).

We have previously shown the generation of constitutively expressed promoters in starving population of Pseudomonas putida cells as a result of base substitutions and deletions and insertion of Tn4652 (20, 33). These promoters, containing a sequence similar to the ς⁷⁰-specific promoter consensus, activated the transcription of phenol degradation genes pheBA (which encode catechol 1,2-dioxygenase and phenol monooxygenase, respectively) and enabled bacteria to utilize phenol as a growth substrate. The fusion promoters were created at the junction of the sequence of the Tn4652 inverted repeats (provides a −35 hexamer) and the target DNA upstream of the phenol monooxygenase gene pheA (provides the −10 hexamer of the promoter) (20, 33) (Fig. 1). Thus, the sequences of the −10 hexamer and the downstream region of the promoters are different, depending on the site of the transposon insertion. In this study, we have shown that the level of transcription from the six different fusion promoters studied depends on the growth phase of the P. putida cells, being in all cases higher in stationary phase. The positive effect of ς^S on transcription was detected in the case of three fusion promoters. Although the importance of the sequence of the −10 hexamer has been pointed out in several reports (see references cited above), the role of the downstream sequences in ς^S-dependent transcription has not been reported. Analysis of the −10 hexameric sequence replacement mutant forms of the fusion promoters constructed in this study indicated that not only the −10 hexameric sequence but also the sequence downstream from the −10 hexamer is important for ς^S-dependent transcription.

FIG. 1.

Nucleotide sequences of fusion promoters PLA1, PRA1, PRA2, PRA3, PRA4, and PRA7 created at the junction of the sequence of the Tn4652 inverted repeats and the target DNA upstream of the phenol monooxygenase gene pheA (33). The −35 hexameric sequence TTGCCT (boxed) of the fusion promoters originated from either the right (PRA1 to PRA7) or the left (PLA1) inverted repeat of transposon Tn4652. The sequence of the −10 hexamer (boxed) and the downstream region of the fusion promoters is different, depending on the site of the transposon insertion. The ς⁷⁰-specific promoter consensus sequences of the −35 and −10 hexamers are shown above the sequences of the fusion promoters.

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth conditions.

The bacterial strains and plasmids used in this study are listed in Table 1. E. coli TG1 (6) was used for the DNA cloning procedures. Exponential- and stationary-phase cultures of P. putida PaW85 (3) and derivative strains PKS54 and PKSRpoS (this work) were used for enzyme assays. We incubated E. coli at 37°C and P. putida at 30°C. E. coli was transformed with plasmid DNA as described by Hanahan (13). P. putida was electrotransformed as described by Sharma and Schimke (38). Bacteria were grown on Luria-Bertani (LB) medium (30). Antibiotics were added at the following final concentrations: ampicillin at 100 μg/ml for E. coli; carbenicillin at 1,500 μg/ml, kanamycin at 50 μg/ml, and tetracycline at 10 μg/ml for P. putida.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Genotype or construction	Source or reference
Strains
E. coli
TG1	supE hsdΔ5 thi Δ(lac-proAB) F′ (traD36 proAB+ lacIqlacZΔM15)	6
BL21(DE3)	hsdS gal (λcIts857 ind1 Sam7 nin5 lacUV5-T7 gene 1	39
S17-1 λpir	Tpr SmrrecA thi pro (r− m+) RP4::2 Tc::Mu::Km Tn7 λpir	30
C118 λpir	Δ(ara-leu) araD ΔlacX74 galE galK phoA20 thi-1 rpsE rpoB argE(Am) recA1 λpir phage lysogen	16
P. putida
PaW85	Tn4652	3
PKS54	Tn4652 rpoS::Kmr	This work
PKSRpoS	P. putida PKS54 rpoS under control of Ptac promoter and lacIq repressor; Tcr
Plasmidsa
pBluescript KS(+)	Cloning vector (Apr)	Stratagene
pBlcrpoS-I	pBluescript KS(+) containing 1-kb PCR-amplified rpoS cloned into EcoRV site
pUC4K	Cloning vector containing Kmr gene from transposon Tn903 (Apr Kmr)	34, 46
pBlcrpoS-Kmr	2.5-kb rpoS-Kmr sequence containing XbaI-EcoRI fragment from pUC4K cloned into Eco72I site in rpoS gene	This work
pUTmini.Tn5 luxAB	Delivery plasmid for mini-Tn5 luxAB (Apr Tetr)	8
pUTrpoS-Kmr	2.5-kb rpoS-Kmr sequence-containing XbaI-EcoRI fragment from pBlcrpoS-Kmr cloned into pUTmini-Tn5 luxAB	This work
pBlcrpoS-II	pBluescript KS(+) containing 900-bp PCR-amplified rpoS cloned into EcoRV site	This work
pET24d	Protein expression vector (Kmr)	Stratagene
pET24d-rpoS	pET24d containing rpoS gene from pBlcrpoS-II inserted into NcoI and HindIII sites	This work
pKTlacZ	Cloning vector (Apr)	18
pBRlacItac	Ptac promoter and lacIq repressor in 2.2-kb NruI-EcoRI fragment from plasmid pMMB208 cloned into EcoRV-EcoRI-cleaved pBR322	—b
pMir61	rpoS gene of P. putida KT2440 in pUN19Ø	36
pBRlacItac-rpoS	rpoS in 1-kb XhoI-HindIII fragment from plasmid pMir61 cloned into SalI-SmaI-cleaved pBRlacItac	This work
pUC18Not	pUC18 with NotI restriction site in multicloning region (Kmr)	16
pUCptac-rpoS	3.2-kb rpoS expression cassette Ptac-rpoS-lacIq from pBRlacItac-rpoS inserted into BamHI-KpnI-cleaved pUC18Not	This work
pUTptac-rpoS	Ptac-rpoS-lacIq from pUCptac-rpoS inserted into NotI-cleaved pUTmini-Tn5 luxAB	This work

^aThe strategy of cloning of the fusion promoters and their mutant forms into promoter probe vector pKTlacZ is shown in Materials and Methods and in Fig. 1 and 2. 

^b—, Hõrak and Kivisaar, submitted. 

Cloning of fusion promoters into a promoter probe vector and construction of promoter mutants.

For amplification of DNA fragments containing fusion promoters PLA1, PRA1, PRA2, PRA3, PRA4, and PRA7 cloned into plasmid pEST1332 upstream of the phenol degradation genes pheBA (33), two oligonucleotides, PAYC32 (5′-CTCGACCTTTGAGCCAAATG-3′) and AB 5′-GGTATGCTTGGCAGTCGT-3′), complementary to sequences upstream and downstream of the Ecl136II and ClaI cloning sites in plasmid pEST1332, respectively, were used. PCR amplification products were cleaved with Ecl136II and ClaI and cloned into plasmid pBluescript KS(+) restricted with SmaI and ClaI. Subsequently, the promoters were inserted with BamHI- and XhoI-generated ends into promoter probe vector pKTlacZ (18).

Mutant promoters L1-R1, L1-R2, L1-R3, L1-R4, and L1-R7 were constructed by substituting the −10 hexamer of promoter PLA1 for the −10 hexamer of promoters PRA1, PRA2, PRA3, PRA4, and PRA7, respectively. Mutant promoters R1-L1, R4-L1, and R7-L1 were constructed by replacing the −10 hexamer of promoters PRA1, PRA4, and PRA7, respectively, with the −10 hexamer of promoter PLA1. Sequences of the mutant promoters are shown in Fig. 2. For site-directed mutagenesis and amplification of the fusion promoters from pEST1332, oligonucleotide PAYC32 and oligonucleotides containing the specific substitutions were used. The mutating primers were partially complementary to the sequence at nucleotides (nt) −21 to +7 relative to the transcriptional start site of the wild-type (wt) promoters. They carried the specific changes within the −10 hexameric sequence, and the ClaI site was designed 10 to 12 nt downstream of the −10 hexamer. L1Cla was constructed as a control by designing a ClaI restriction site 12 nt downstream of the −10 hexamer of PLA1 (Fig. 2). The amplified PCR products were cleaved with Ecl136II and ClaI and cloned into pBluescript KS(+). Thereafter, the mutated sequences were inserted upstream of the lacZ reporter gene into plasmid pKTlacZ by using the BamHI- and XhoI-generated ends. All introduced base substitutions were verified by dideoxy sequencing with a Sequenase version 2.0 kit (Amersham).

FIG. 2.

Combination of −10 hexamers and downstream sequences of the fusion promoters. Mutants L1-R1, L1-R2, L1-R3, and L1-R7 were constructed by replacement of the −10 hexamer of promoter PLA1 with the −10 hexamer of promoters PRA1, PRA2, PRA3, PRA4, and PRA7, respectively. In mutants R1-L1, R4-L1, and R7-L1, the −10 hexamers of PRA1, PRA4, and PRA7, respectively, were replaced with the −10 hexamer of promoter PLA1. The −10 hexameric sequence is boxed and marked in bold. All of the mutant promoters carry the ClaI restriction site (marked in bold and italic) 10 to 12 nt downstream of the −10 hexamer for cloning of the promoter sequences into promoter probe vector pKTlacZ (19). L1Cla was constructed as a control by designing a ClaI restriction site 12 nt downstream of the −10 hexamer of PLA1.

Construction of P. putida PaW85 rpoS knockout mutant PKS54.

P. putida rpoS knockout mutant PKS54 was constructed as a derivative of PaW85 by interrupting the rpoS gene with a kanamycin resistance-encoding gene (Km^r) cloned from transposon Tn903 (34) in plasmid pUC4K (46). The oligonucleotides used for amplification of the rpoS gene of P. putida PaW85 were designed on the basis of the published sequence of the rpoS gene of P. putida KT2440 (36). Two oligonucleotides, PprpoSall, (5′-AAAGCTTCCCCTTGCCGGGTGTGTAGAGGA-3′) and PprpoSyll (5′-CAAGCGCTGCCAGGGAGAAA-3′), complementary to the sequence 159 nt downstream of the TAG stop codon and 68 nt upstream of the ATG initiator codon of the rpoS gene of P. putida KT2440, respectively, were used. The PCR-generated DNA fragment containing the rpoS gene was subcloned into pBluescript KS(+) cleaved with EcoRV (pBlcrpoS-I in Table 1). The EcoRI-generated ends of the DNA fragment containing the Km^r from plasmid pUC4K were filled with Klenow fragment, and the blunt-ended DNA segment was inserted into the Eco72I-cleaved rpoS gene. The resulting rpoS-Km^r sequence from pBlcrpoS-Km^r was inserted into conjugative plasmid pUTmini-Tn5 luxAB (9) by using XbaI and EcoRI sites, and pUTrpoS-Km^r was selected in E. coli CC118 λpir (16). The interrupted rpoS gene was inserted into the chromosome of PaW85 by homologous recombination using E. coli S17 λpir (31) as the recipient strain. P. putida rpoS knockout mutant PKS54 was selected at 30°C on glucose-kanamycin plates and verified with immunoblotting of P. putida RpoS (see below).

Construction of P. putida rpoS complementation strain PKS54.

For construction of rpoS complementation strain PKS54, the rpoS gene was cloned from plasmid pMir61 (36) by using XhoI- and HindIII-generated ends into the vector pBRlacItac (R. Hõrak and M. Kivisaar, submitted for publication) cleaved with SalI and SmaI (pBRlacItac-rpoS in Table 1). The rpoS expression cassette Ptac-rpoS-lacI^q was inserted into pUC18Not (16) using BamHI- and KpnI-generated ends. Thereafter, the hybrid sequence from pUCptac-rpoS was inserted into the NotI site of pUTmini-Tn5 luxAB (8). The resulting construct, pUTptac-rpoS, was selected in E. coli CC118 λpir (16). The Ptac-rpoS-lacI^q cassette was inserted into the chromosome of P. putida PKS54 by random insertion using E. coli S17 λpir as the recipient strain. P. putida PKSRpoS was selected at 30°C on glucose-kanamycin-tetracycline plates. The expression of RpoS in PKSRpoS was verified with Western blot analysis using polyclonal antibodies against P. putida RpoS (data not shown).

Cloning, overexpression, and purification of RpoS of P. putida PaW85.

For amplification and cloning of the rpoS gene of P. putida PaW85, oligonucleotides RpoSCNco (5′-TCCCATGG[NcoI]CTCTCAGTAAAGAAGTGCCC-3′; complementary to the sequence 21 nt downstream of the TAG stop codon of rpoS of P. putida KT2440) and RpoSCHind (5′-GCAAGCTT[HindIII]CTGGAACAATGACTCGCTGGT-3′; complementary to the sequence 20 nt upstream of the ATG initiator codon of rpoS of P. putida KT2440) were used. A PCR-generated DNA fragment containing the rpoS sequence was subcloned into pBluescript KS(+) cleaved with EcoRV to obtain pBlcrpoS-II. Thereafter, the DNA fragment containing the rpoS gene from pBlcrpoS-II was inserted into pET24d (Stratagene) with NcoI- and HindIII-cleaved ends to generate a His tag at the C terminus of RpoS (pET24d-rpoS in Table 1).

To obtain soluble RpoS-His protein, E. coli BL21(DE3) (39) carrying pET24d-rpoS was grown overnight at 37°C in 20 ml of M9 minimal medium (1). Subsequently, the culture was diluted into 500 ml of fresh M9 medium and the bacteria were grown at 37°C until the optical density of the culture at 590 nm reached about 0.6. For the expression of RpoS-His, the culture was incubated at 20°C for 0.5 h and then induced for 4 h at 20°C by adding isopropyl-β-d-thiogalactopyranoside (IPTG; final concentration, 0.5 mM). Cells were pelleted and sonicated in buffer A (100 mM NaH₂PO₄, 1 M NaCl, pH 8.0). The cell lysate was centrifuged at 5,000 × g for 10 min. Supernatant was loaded onto an Ni²⁺-iminodiacetic acid-activated chelating Sepharose 6B column previously equilibrated with 5 volumes of buffer A. After 4 h of incubation at 10°C, the loaded column was washed three times with 2 volumes of buffer A supplemented with 10% glycerol (pH 6.2). The purified His-RpoS was eluted three times with 2 volumes of buffer A supplemented with 10% glycerol and 1 M imidazole (pH 6.2). The imidazole and excess salt were removed by dialyzing the eluate against 1× phosphate-buffered saline, and the purified protein was stored at −20°C. The purified RpoS was used for production of mouse anti-RpoS polyclonal antibody for the immunoblotting assay.

Preparation of cell lysates and immunoblotting of P. putida RpoS.

Cell lysates of P. putida PaW85, PKS54, and PKS54RpoS were prepared from 50-ml stationary-phase LB medium cultures and from 200-ml exponential-phase LB medium cultures. Cells were pelleted and sonicated in 500 μl of 100 mM phosphate buffer (87 mM Na₂HPO₄, 13 mM KH₂PO₄, pH 7.5). The protein concentration in cleared lysates was estimated as described by Bradford (5). Equal amounts (30 μg) of total protein were used for a Western immunoblotting assay. Proteins were separated by sodium dodecyl sulfate–10% polyacrylamide gel electrophoresis and transferred to nitrocellulose membrane (BA85; Schleicher & Schuell). For Western blotting, the membrane was probed with mouse anti-RpoS polyclonal serum diluted 1:250, followed by alkaline phosphatase-conjugated coat anti-mouse immunoglobulin G (LabAS Ltd., Tartu, Estonia) diluted 1:5,000. The blots were developed with 5-bromo-4-chloro-3-indolylphosphate–nitroblue tetrazolium.

β-Gal assay.

Exponential- and stationary-phase cells of P. putida PaW85 and its derivatives harboring different fusion promoter constructs were used for a β-galactosidase (β-Gal) assay performed as specified by Miller (30). The protein concentration in crude lysates was measured by the method of Bradford (5). The starting density of the overnight culture for 50 ml of LB medium at 580 nm was 0.02 (∼10⁶ cells/ml). The samples for the β-Gal assay were collected at 4, 12, 24, 36, and 48 h.

RESULTS AND DISCUSSION

Study of the effect of ς^S on transcription from fusion promoters.

Promoters were created during transposition of Tn4652 in stationary-phase cells of P. putida as fusions between the −35 hexamer provided by the terminal inverted repeats of Tn4652 and the −10 hexamers in the target DNA (33). Five of the six different fusion promoters identified were generated at the junctions of the right terminus of the transposon and target DNA (promoters PRA1, PRA2, PRA3, PRA4, and PRA7), and in only one particular case (promoter PLA1) was the −35 hexamer provided by the left end of Tn4652 (33) (Fig. 1). In this study, we cloned the sequences containing different fusion promoters upstream of the reporter gene lacZ in plasmid pKTLacZ (see Materials and Methods) and studied the effect of the growth phase of bacteria on transcription from these promoters. P. putida PaW85 cells carrying lacZ transcriptional fusions were grown on LB medium, and β-Gal activity was measured in both exponentially growing cells (sampled at 4 h) and stationary-phase cells (sampled at 12, 24, 36, and 48 h). The growth curve of bacteria is shown in Fig. 3A. Data presented in Table 2 demonstrate that the level of expression of β-Gal activity was remarkably elevated in stationary-phase cells in all cases and it remained high in deep-stationary-phase cultures during the next 2 days studied. This indicated that the fusion promoters are certainly stationary phase specific. Previously we have shown that transcription from the fusion promoters PRA1 and PLA1, containing sequences of either the right or left end of Tn4652, respectively, is modulated by IHF and that the positive effect of IHF becomes apparent in stationary-phase cells of P. putida (43). However, the fusion promoters cloned into the pKTLacZ reporter plasmid lacked functional IHF binding sites but were still expressed at an increased level in stationary-phase cells.

FIG. 3.

(A) Growth curve of P. putida PaW85 cells grown in LB medium. The growth rate of strains PKS54 and PKSRpoS is similar to that of PaW85. (B) Study of the amounts of RpoS in P. putida PaW85 and PKS54 cells by Western blot analysis with P. putida anti-RpoS polyclonal antibodies. The cell lysates used were prepared from P. putida PaW85 cells sampled at h 3, 4, 5, 6, 12, and 24 (lanes 1 to 6). PKS54 cells were sampled at h 24 (lane 7). Purified RpoS-His (lane 8) served as a control. A 30-μg portion of the total protein from cell lysates and 0.6 μg of RpoS-His were loaded onto the gel. OD580, optical density at 580 nm.

TABLE 2.

Study of the expression of fusion promoters in P. putida strain PaW85 and its ς^S-deficient derivative PKS54

Strain	Promoter	Mean β-Gal sp acta ± SD (wt/rpoSb)
		4 h	12 h	24 h	36 h	48 h
rpoSWt	PLA1	10.4 ± 0.5 (0.8)	166.2 ± 2.9 (1.6)	309.0 ± 129 (2.3)	708.1 ± 8.4 (3.0)	436.7 ± 3.9 (1.6)
rpoS	PLA1	12.9 ± 7.4	105.0 ± 9.1	133.8 ± 14.8	238.0 ± 6.0	268.9 ± 19.1
Wt	PRA1	100.8 ± 2.4 (0.9)	357.4 ± 122.4 (0.8)	406.6 ± 9.6 (0.8)	733.0 ± 22.1 (0.8)	591.3 ± 42.9 (0.8)
rpoS	PRA1	106.2 ± 2.9	432.8 ± 25.8	518.8 ± 31.9	888.5 ± 70.3	774.0 ± 42.4
Wt	PRA2	19.2 ± 3.6 (1.0)	114.4 ± 7.22 (0.9)	191.5 ± 12.3 (1.2)	161.6 ± 4.5 (1.0)	159.2 ± 1.3 (0.9)
rpoS	PRA2	19.5 ± 4.1	120.8 ± 10.0	154.0 ± 6.2	157.0 ± 10.2	176.4 ± 6.1
Wt	PRA3	222.2 ± 18.6 (1.2)	460.8 ± 2.8 (1.0)	1,445.2 ± 129 (1.1)	1,716.5 ± 141.0 (0.9)	1,392.4 ± 126.0 (0.9)
rpoS	PRA3	181.1 ± 4.9	475.1 ± 17.3	1,369.7 ± 124	1,837.8 ± 88.0	1,557.3 ± 102.0
Wt	PRA4	98.7 ± 6.5 (1.0)	996.5 ± 54.6 (3.0)	1,948 ± 27.2 (4.8)	2,390.0 ± 285.0 (3.8)	2,013.0 ± 272.0 (3.1)
rpoS	PRA4	103.6 ± 9.1	333.1 ± 15.6	408.6 ± 39.8	627.4 ± 22.4	647.5 ± 63.2
Wt	PRA7	97.7 ± 21.1 (1.3)	488.3 ± 146.2 (0.9)	2,500.1 ± 130.6 (2.5)	3,024.8 ± 29.9 (2.6)	2,873.6 ± 592.4 (2.6)
rpoS	PRA7	75.2 ± 21.2	549.1 ± 145.1	1,000.6 ± 38.8	1,167.4 ± 15.8	1,121.3 ± 24.1

^aβ-Gal specific activity (nanomoles of o-nitrophenol formed per minute per milligram of protein) was measured in P. putida PaW85 (wt) and PKS54 (rpoS) cells grown for 4, 12, 24, 36, and 48 h. Data of at least three independent experiments are presented. 

^bThe effect of ς^S on transcription from the fusion promoters is expressed as the ratio of β-Gal activity in PaW85 cells to the β-Gal activity in PKS54 cells (wt/rpoS). 

The most important regulator of stationary-phase-induced genes is sigma factor ς^S, encoded by rpoS, which directs the expression of nearly 100 genes in E. coli (25, 45) and is considered the second principal sigma factor (42). The requirement of ς^S for stimulation of transcription has also been described in the case of genes originating from P. putida (27, 28, 32). The gene homologous to the rpoS gene of E. coli has been cloned from P. putida by complementation of the rpoS-deficient E. coli strain (36). In order to test whether functional ς^S would be responsible for the increased level of transcription from the fusion promoters in stationary-phase cells of P. putida, we constructed an rpoS knockout mutant of P. putida. P. putida PaW85 rpoS knockout mutant PKS54 was obtained by homologous recombination between the P. putida original chromosomal rpoS gene and the rpoS gene interrupted by a kanamycin resistance gene. P. putida ς^S was overexpressed using protein expression vector pET24d and purified to obtain polyclonal antibodies against it (for details, see Materials and Methods). The absence of ς^S in the mutant strain was verified by Western blot analysis (Fig. 3B). In E. coli, an increase in the cellular ς^S concentration during entry into stationary phase has been observed (12, 24, 42). Western blot analysis of the intracellular ς^S content of P. putida (Fig. 3B) also demonstrated that a low level of this sigma factor is detectable even in actively growing cells (h 3 and 4) and that the intracellular level of ς^S increases in bacteria sampled from the late-exponential-phase culture (h 5 and 6).

We compared the expression of the fusion promoters in wt strain PaW85 and in its ς^S-deficient derivative PKS54 (Table 2). The effect of ς^S became evident only in the case of three fusion promoters: the level of transcription from promoters PLA1, PRA4, and PRA7 was approximately three times lower in strain PKS54 than in the wt strain. The maximal difference appeared in a deep-stationary-phase culture (cells sampled at 24 h and later). At the same time, the level of transcription from fusion promoters PRA1, PRA2, and PRA3 did not depend on the presence or absence of functional ς^S in the cells. In order to test whether the full level of transcription from fusion promoters PLA1, PRA4, and PRA7 would be restored by complementation of the ς^S-deficient mutant with the functional rpoS gene, strain PKSRpoS was constructed. In this strain, the rpoS gene was introduced into the chromosome of PKS54 under control of the Ptac promoter and the lacI^q repressor (Materials and Methods). As a result of complementation, the level of expression of fusion promoters PLA1, PRA4, and PRA7 in PKSRpoS was approximately the same as that estimated in the wt strain (data not shown).

As shown in Table 2, the stationary-phase-induced transcription from the fusion promoters cannot be explained as the effect of ς^S only. The ς^S-independent increase in transcription became evident with all of the fusion promoters. Gel mobility shift experiments with Tn4652 ends and crude lysate of P. putida have revealed that in addition to IHF, some other proteins form specific complexes with the ends of Tn4652 (18, 43). The amount of the protein-DNA complexes detected depends on the growth phase of the bacteria used for the preparation of cell extracts (43). It is therefore possible that a protein(s) which binds termini of Tn4652 could sequester transcription from the fusion promoters in a growth phase-dependent manner. Experiments intended to identify these factors are currently in progress.

Study of the effect of combination of the −10 hexameric sequences and downstream sequences of the fusion promoters on transcription in ς^S-deficient P. putida.

Notably, the fusion promoters providing a lower level of reporter gene expression in the rpoS mutant strain differed from the others only by the −10 hexameric sequence and by the sequence downstream from that. There was only one mismatch between the right and left inverted repeat sequences, making the spacer region of the fusion promoters created from either the right or the left end of the transposon different by 1 nt (Fig. 1). Thus, our results indicated that the sequence of the −10 element could be important for ς^S-dependent transcription. The same has also been concluded in other published reports (17, 22, 41). A compilation of ς^S-dependent promoters deduced a −10 consensus sequence, CTATACT (11). Fusion promoters PLA1, PRA4, and PRA7, that were expressed at a decreased level in the rpoS-deficient background, contained −10 hexamers TATACT, TAAACT, and TATAAT, respectively, that were preceded by a C nucleotide. At the same time, the sequence of the −10 element of ς^S-independent promoter PRA1 differed in two position from the −10 hexamer of PLA1 (sequence TATCAT instead of TATACT).

To confirm whether the ς^S-dependent transcription from promoters PLA1, PRA4, and PRA7 would be influenced by the specific sequence of the −10 hexamer, the identical downstream sequence was constructed for all of the different fusion promoters studied. For that purpose, the −10 sequence of PLA1 (TATACT) was replaced with the sequences of the −10 hexamers of the other fusion promoters (Fig. 2). The sequence of PLA1 was chosen because its −10 hexamer is identical to the fic promoter −10 element (42, 44). ς^S-dependent transcription initiation from the fic promoter requires the −10 hexamer TATACT but does not need a specific sequence in the −35 region (17). The resulting PLA1 −10 substitution mutants L1-R1, L1-R2, L1-R3, L1-R4, and L1-R7 contained 12 nt of the downstream sequence of PLA1 and an artificial ClaI site for cloning of these promoters into the pKTLacZ vector. The same strategy was used to subclone PLA1 upstream of the lacZ gene to obtain L1Cla. We also replaced the −10 hexamer of promoters PRA1, PRA4, and PRA7 with the −10 sequence of PLA1 (see R1-L1, R4-L1, and R7-L1 in Fig. 2). β-Gal activities measured in the wt strain and in its rpoS-deficient derivative PKS54 carrying different −10 substitution mutants revealed unexpected results (Table 3). Although the expression of fusion promoters PRA1 and PRA3 was not influenced by ς^S in the original location of these sequences, a two- to fourfold positive effect of ς^S was still apparent when the −10 hexameric sequence of PLA1 was replaced with −10 hexamers of PRA1 and PRA3. At the same time, only a very mild effect, if any, was observed in the case of L1-R7 despite the fact that both PLA1 and PRA7 separately exhibited decreased expression in the rpoS mutant strain. No effect of the presence of functional ς^S was detected in the case of L1-R2. A mild effect, not exceeding 1.5 times, became evident in R1-L1 and R7-L1 (the −10 hexamers of PRA1 and PRA7 were replaced with the −10 hexamer of PLA1, respectively), and an approximately twofold effect was revealed when the −10 element of PRA4 was changed to that of PLA1 (see R4-L1).

TABLE 3.

Study of the expression of mutant fusion promoters in P. putida strain PaW85 and its ς^S-deficient derivative PKS54

Strain	Promoter	Mean β-Gal sp acta ± SD (wt/rpoS)b
		4 h	12 h	24 h	36 h	48 h
Wt	L1Cla	24.4 ± 1.9 (1.1)	196.6 ± 20.1 (1.4)	236.8 ± 8.3 (1.9)	283.8 ± 14.8 (1.9)	279.8 ± 8.9 (2.0)
rpoS	L1Cla	21.0 ± 0.5	140.3 ± 7.2	126.1 ± 3.2	146.4 ± 4.3	139.1 ± 0.4
Wt	L1-R1	5.1 ± 1.2 (0.8)	62.5 ± 5.9 (2.2)	168.3 ± 7.2 (3.6)	287.2 ± 4.6 (4.8)	202.1 ± 44.3 (2.4)
rpoS	L1-R1	6.1 ± 1.3	28.3 ± 2.1	46.7 ± 5.4	59.7 ± 6.1	83.3 ± 15.2
Wt	L1-R2	5.8 ± 0.3 (0.6)	53.9 ± 8.4 (1.0)	94.1 ± 16.4 (0.9)	153.2 ± 19.0 (1.2)	123.8 ± 12.7 (1.1)
rpoS	L1-R2	9.2 ± 1.0	55.1 ± 5.2	109.3 ± 28.2	130.4 ± 15.6	112.9 ± 21.3
Wt	L1-R3	13.8 ± 2.9 (1.8)	43.3 ± 1.2 (1.8)	95.0 ± 1.8 (2.2)	97.6 ± 2.7 (2.0)	100.8 ± 3.9 (2.2)
rpoS	L1-R3	7.8 ± 0.3	24.1 ± 1.7	43.5 ± 2.4	49.8 ± 4.1	46.1 ± 4.8
Wt	L1-R4	3.4 ± 0.6 (1.1)	26.0 ± 0.8 (1.8)	52.8 ± 2.9 (1.8)	114.8 ± 15.3 (2.7)	84.7 ± 4.8 (2.1)
rpoS	L1-R4	3.0 ± 0.1	14.1 ± 2.7	29.7 ± 7.6	42.5 ± 9.9	39.6 ± 13.2
Wt	L1-R7	31.6 ± 1.6 (0.8)	126.5 ± 7.9 (1.0)	242.3 ± 29.0 (1.1)	442.2 ± 13.3 (1.2)	358.5 ± 15.8 (1.3)
rpoS	L1-R7	38.1 ± 10.8	122.7 ± 7.7	222.3 ± 54.8	364.2 ± 9.6	282.2 ± 7.3
Wt	R1-L1	118.4 ± 24.4 (1.2)	874.3 ± 288.4 (1.3)	2,358.4 ± 255.0 (0.9)	1,417.9 ± 208.8 (1.4)	1,344.3 ± 82.1 (1.4)
rpoS	R1-L1	100.6 ± 19.4	659.5 ± 121.7	2,484.0 ± 274.8	992.7 ± 65.0	931.8 ± 107.3
Wt	R4-L1	291.4 ± 27.3 (1.7)	1,160.6 ± 101.9 (1.6)	2,433.1 ± 161.0 (2.0)	2,715.1 ± 116.9 (1.8)	2,484.7 ± 120.0 (1.4)
rpoS	R4-L1	175.8 ± 9.8	737.7 ± 59.1	1,214.2 ± 72.6	1,537.7 ± 152.7	1,825.9 ± 95.0
Wt	R7-L1	60.8 ± 2.5 (0.9)	319.8 ± 22.0 (1.1)	914.1 ± 69.3 (1.4)	954.1 ± 42.1 (1.3)	1,167.8 ± 99.0 (1.2)
rpoS	R7-L1	66.2 ± 9.8	286.7 ± 13.2	647.1 ± 17.7	714.4 ± 31.9	987.7 ± 48.8

^aβ-Gal specific activity (nanomoles of o-nitrophenol formed per minute per milligram of protein) was measured in P. putida PaW85 (wt) and PKS54 (rpoS) cells grown for 4, 12, 24, 36, and 48 h. Data of at least three independent experiments are presented. 

^bThe effect of ς^S on transcription from the promoters is expressed as the ratio of β-Gal activity in PaW85 cells to the β-Gal activity in PKS54 cells (wt/rpoS). 

To summarize the results obtained in this study, we can conclude that neither the sequence of the −10 hexamer of the fusion promoters nor the sequence downstream of the −10 element can separately affect ς^S-dependent transcription. Rather, the sequence of the −10 hexamer in concert with the sequence downstream from the −10 element influences ς^S-dependent transcription from the fusion promoters. Based on in vivo experiments carried out in this study, we cannot elucidate the exact mechanisms of interaction of the Eς^S holoenzyme with downstream sequences of promoters during transcription initiation; i.e., it is difficult to say whether the downstream sequences could somehow be involved in Eς^S-specific promoter recognition. Transcription initiation involves two major steps, formation of a closed complex which, in a second step, isomerizes to an open complex (47). Until now, in contrast to Eς⁷⁰, only a few biochemical studies on the stationary-phase RNA polymerase have been performed. Comparative investigations have revealed that for specific DNA-protein binding, the Eς^S holoenzyme interacts with a smaller part of the promoter than Eς⁷⁰ and recognizes the sequence in the −10 region (17, 41). Recently, the interaction of ς^S and ς⁷⁰ in RNA polymerase-promoter open complexes was analyzed using FeBABE nuclease (4, 7, 35). The two holoenzymes revealed similar cutting patterns, but the cutting pattern of Eς^S extended toward the downstream part of the promoter, around +10 and +20 (7). Therefore, taking together the published data and the results presented in this study, we propose that the possible role of downstream sequences in ς^S-dependent transcription needs more consideration.