The aprX–lipA operon of Pseudomonas fluorescens B52: a molecular analysis of metalloprotease and lipase production

Rick G. Woods1, Michelle Burger1, Carie-Anne Beven1 and Ifor R. Beacham1

School of Health Science, Griffith University, PMB 50 Gold Coast Mail Centre, Gold Coast, Qld 4217, Australia1

Author for correspondence: Ifor R. Beacham. Tel: +61 7 5594 8185. Fax: +61 7 5594 8908. e-mail: i.beacham{at}mailbox.gu.edu.au


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Extracellular protease and lipase production by psychrotrophic strains of Pseudomonas fluorescens is repressed by iron and regulated by temperature. The regulation of protease and lipase has been investigated in P. fluorescens B52. Whereas lipase production is increased below the optimum growth temperature (‘low-temperature regulation’), protease production was relatively constant and only decreased above the optimum growth temperature. The genes encoding protease (aprX) and lipase (lipA) are encoded at opposite ends of a contiguous set of genes which also includes protease inhibitor, Type I secretion functions and two autotransporter proteins. Evidence is presented indicating that these genes constitute an operon, with a promoter adjacent to aprX which has been identified by S1 nuclease analysis. The regulation of aprX and lipA has been investigated at the RNA level and using lacZ fusion strains. Whereas the data are consistent with iron regulation at the transcriptional level, a lipA'–'lacZ fusion is not regulated by temperature, suggesting that temperature regulation is post-transcriptional or post-translational. The possibility of regulation at the level of mRNA decay is discussed.

Keywords: Protease, lipase, Pseudomonas fluorescens, promoter, gene regulation

Abbreviations: ECF, extracytoplasmic function

The GenBank accession numbers for the sequences reported in this paper are AF216700, AF216701 and AF216702.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The production of lipase and protease by psychrotrophic strains of Pseudomonas fluorescens has been studied for many years due to the importance of these enzyme activities in food spoilage at low temperatures (McKellar, 1989 ). Such studies have demonstrated that protease and lipase production are regulated by temperature and by iron levels (Andersson, 1980 ; McKay & Beacham, 1995 ; McKellar & Cholette, 1987 ; McKellar et al., 1987 ). It has been demonstrated that at least two strains of P. fluorescens (B52 and LS107d2) secrete only a single detectable lipase of approximately 52 kDa, since site-specific mutants are phenotypically lipase negative (McKay et al., 1995 ; and this study). The regulation of lipase production is of particular interest in that there is an inverse relationship between enzyme production and growth temperature in cultures growing in steady state (Andersson, 1980 ; Merieau et al., 1993 ). Furthermore, the lipase is stable following prolonged incubation at temperatures at which its production is not apparent, and is notably thermostable (Andersson, 1980 ; Andersson et al., 1979 ). It has therefore been proposed that lipase production is regulated resulting in increased enzyme synthesis at low temperature. This ‘low-temperature regulation’ is therefore in contrast to the increased production of enzyme synthesis at higher temperatures, ‘thermoregulation’, often associated with pathogenesis (e.g. Falconi et al., 1998 ). However, the mechanism of this form of temperature regulation, and its relationship to other regulatory mechanisms influencing lipase and protease synthesis, is unknown.

In this report we present a phenotypic and molecular study of the operon which encodes lipase and an analysis of its regulation by temperature.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains and growth conditions.
Escherichia coli strains TGI or S17.1{lambda}pir (Penfold & Pemberton, 1992 ) cells were used for routine plasmid transformations and maintained in Luria–Bertani (LB) medium containing the appropriate antibiotic (Sambrook et al., 1989 ): ampicillin (100 µg ml-1); kanamycin (50 µg ml-1) or chloramphenicol (50 µg ml-1). P. fluorescens strain LS107d2 has been described previously (Johnson et al., 1992 ). Lipase and protease activity profiles of P. fluorescens B52 (Richardson & Te Whaiti, 1978 ; Richardson, 1981 ) were performed on cultures grown in HEPES medium (2 mM HEPES, pH 7·0, 1 mM NH4Cl2, 0·1 mM CaCl2, 1 mM MgCl2 and 0·5% hydrolysed casein). The conditions for studying iron (FeCl3) regulation in HEPES basal medium were assessed. In the case of protease, repression occurred at a minimum concentration of 1 µM; however, lipase production showed a more complex pattern with a requirement for iron of 2 µM and repression by 5–20 µM (data not shown). For the determination of lipolytic and proteolytic phenotypes this HEPES media was solidified with 1·5% bacteriological agar and contained 1% (v/v) tributyrin or 2·5% (w/v) skimmed milk, respectively. Tributyrin media could be used for determination of the lipolytic phenotype since esterase activity is not apparent in B52, in contrast to strain LS107d2 (McKay et al., 1995 ; see also Fig. 3).



View larger version (66K):
[in this window]
[in a new window]
 
Fig. 3. Phenotypic and complementation analysis of P. fluorescens B52 and derived mutants. Individual colonies were patched onto HEPES agar media containing 1% tributyrin (a, d) or 2·5% skimmed milk (b, c), to determine their lipase and protease phenotypes, respectively, and incubated at room temperature for 4 d. (a, b) B52 (1), aprX'–'lacZ (2), aprX::pJPkan (3), and prtB::cam (4). (c, d) Complementation of strain B52aprX'–'lacZ with pBBR1-32. This plasmid contains the aprX–inh–aprDEF–prtA' genes cloned in the reverse orientation with respect to the vector promoter. A plasmid with the insert in the opposite orientation to pBBR1-32 gave identical results. Strains B52 and B52aprX'–'lacZ with and without pBBR1-32 were patched onto media containing skimmed milk (c) or tributyrin (d). (1), B52aprX'–'lacZ; (2), B52; (3), B52aprX'–'lacZ containing pBBR1-32.

 
DNA purification and sequencing.
Genomic DNA was isolated using a genomic purification column (Qiagen) or by the method of Ausubel et al. (1987) . Plasmid DNA was isolated by the alkaline lysis method (Sambrook et al., 1989 ). DNA cloning, ligations and transformations were essentially as described by Sambrook et al. (1989) . Nucleotide sequencing was carried out using the ABI PRISM Dye Terminator Cycle Sequencing Ready Reaction Kit with AmpliTaq DNA Polymerase FS and a model 377 automated sequencer (Applied Biosystems).

Cloning of aprX and upstream lipA sequence.
To isolate the metalloprotease gene from P. fluorescens B52, two primers, MU1 (5'-CGGAATTCACGAGATCGGCCATACC-3') and MD2 (5'-GCGGATCCGTAGAGGATGTCGTTGC-3'), were used that correspond to conserved regions between the metalloprotease genes from Serratia marcescens (Nakahama et al., 1986 ) and Pseudomonas aeruginosa (Duong et al., 1992 ). PCR using these primers and B52 genomic template gave a product of approximately 578 bp that was used as a probe to screen a B52 genomic ‘mini-library’ derived by PstI cleavage. In brief, B52 genomic DNA (5 µg) was digested with PstI, electrophoresed on a 0·8% Tris/borate/EDTA agarose gel and then probed with the 578 bp amplified fragment which revealed a 6 kb hybridizing fragment. DNA in this size range was extracted, ligated into pSK (Stratagene) and transformed into TG1. Recombinant colonies were then screened with the same probe and a recombinant with an insert size of 6 kb was isolated and called pRTMp11. To isolate a fragment containing upstream sequence of lipA from P. fluorescens LS107d2, pLAJ2 (containing the cloned LS107d2 lipA gene; Johnson et al., 1992 ) was digested with EcoRI/SalI to produce a 400 bp fragment. This was then labelled with [{alpha}-32P]dCTP using nick translation and used to probe an LS107d2 pHC79 cosmid library. A positive cosmid was identified (pHCB1) and found to contain a hybridizing 1·4 kb SalI fragment by Southern analysis (Sambrook et al., 1989 ). This SalI fragment was subcloned into pSK and designated pSWB1. Isolation of lipA upstream sequence from strain B52 was obtained by screening a ‘mini-library’ as described above, but using the 400 bp EcoRI–SalI fragment from pLAJ2. A recombinant with a 3 kb ApaI fragment was identified and designated pRW52. To isolate further upstream sequence, an ApaI–HindIII fragment from pRW52 (nt 0–451) was used to probe 5 µg HindIII-digested genomic B52 DNA. A recombinant with a ~6 kb HindIII insert was identified and designated pRW53.

Cloning of the secretion genes aprDEF.
To isolate the secretion genes aprDEF, a genomic library of P. fluorescens B52 was made using {lambda}Zap Express (Stratagene) and screened using an EcoRI–PstI labelled fragment from pRTMp11. A positive clone, designated p5232, with an insert size of ~3 kb was identified and sequence analysis confirmed the presence of the secretion genes aprDEF. To construct a single plasmid containing the aprX–inh–aprDEF genes, the 6 kb PstI insert from pRTMp11 (from within aprE to the upstream vector site) was cloned into PstI-cleaved p5232; the PstI digestion of p5232 removed the 2 kb fragment spanning the PstI site common to both plasmids (within aprE) and the upstream PstI site in the vector. Recombinants were identified in E. coli TG1 as protease-positive on skimmed milk media, and one was designated pRTM32. A 7 kb XhoI fragment from pRTM32, spanning aprX–inh–aprDEF to within prtA, was then cloned into alkaline phosphatase-treated SalI-cleaved pBBR1MCS (Kovach et al., 1994 ); recombinant clones, in each orientation with respect to the vector promoter, were confirmed by restriction analysis and used for complementation (see Results); one clone in the reverse orientation was designated pBBR1-32.

Transcription mapping.
The aprX transcriptional start point was mapped using S1 nuclease protection. A 274 bp DNA fragment was derived by PCR; two primers were used, PRTMA (5'-CGGATCCGTCACATTTTCGCTGCGG-3') and PRTMB1 (3'-CTGGCACCGCCGTTGGACCTAGG-GC-5'), which annealed at aprX positions 157 to -139 and 87 to 104, respectively (position 1 being the A of the aprX ATG start codon) and which incorporated BamHI sites (underlined). The amplified fragment was cloned into the BamHI site of pSK and called pRTMAB. To amplify a 32P-labelled single stranded antisense probe, 150 pmol primer PRTMS1 (3'-CATACTGGCACCGCCGTTGGA-5', nt 82–103) was used in a linear amplification using 2 µg pRTMAB, linearized at the SpeI site (located 5 bp 5' of the PRTMB1 BamHI site), as template. The PCR reaction, containing 62 µmol each of dATP, TTP and dGTP, 18 µmol dCTP, 3 µl [{alpha}-32P]dCTP (1014 Bq mmol-1; 370 Bq µl-1) and 2·5 U Taq DNA polymerase (Promega), was cycled under the following conditions: 96 °C for 40 s; 50 °C for 40 s and 72 °C for 90 s, for a total of 40 cycles. The probe was purified by electrophoresis on a 6% denaturing polyacrylamide gel, exposing to X-ray film and excising the appropriate band. The probe was eluted from the acrylamide gel slice by soaking in 300 µl elution buffer (0·5 M ammonium acetate, 1 mM EDTA and 0·2% SDS) overnight at 37 °C, ethanol precipitated and dissolved in 100 µl H2O (DEPC treated) and stored at -20 °C until required. Total RNA (5 µg), purified as described by Aiba et al. (1981 ), was mixed with the single-stranded antisense 32P-labelled probe, ethanol precipitated and resuspended in 10 µl hybridization buffer (80% deionized formamide, 100 mM sodium citrate, pH 6·4, 300 mM sodium acetate, pH 6·4, 1 mM EDTA). Following a denaturation step of 95  °C for 10 min, the nucleic acids were hybridized for 3 h at 37 °C. The annealed RNA–DNA mix was then made up to 200 µl with 1xS1 nuclease buffer (Promega), 50 U S1 nuclease was added and the mixture was incubated at 37 °C for 3 h. The digested products were ethanol precipitated and resuspended in 8 µl gel loading buffer (95% formamide, 20 mM EDTA, 0·05% bromophenol blue and 0·05% xylene cyanol FF). Typically, 2 µl was then electrophoresed on a 6% denaturing polyacrylamide gel alongside dideoxy sequencing reactions (Sequenase kit; USB) which incorporated [{alpha}-32P]dCTP. The gel was dried and visualized by autoradiography.

Quantitation of aprX mRNA.
To quantify the aprX message, RNA was subjected to S1 analysis (see above) and then electrophoresed on a 6% polyacrylamide denaturing gel. The intensity of each band was then measured using a Personal Molecular Imager FX (Bio-Rad) using storage phosphor screens (Kodak) and Quantity One version 4.03 software.

Lipase and protease assays.
A fluorimetric assay utilizing the substrate 4-methylumbelliferyl oleate (Roy, 1980 ; Stead, 1984 ) was used to measure lipase activity secreted during growth. Fluorescence was monitored in a luminescence spectrophotometer (2·5 nm slit widths; excitation wavelength 339 nm and emission wavelength 444 nm). The reaction mixture contained between 10 and 100 µl culture supernatant, 45 µl 0·1 mmol CaCl2 l-1, 300 µl 1 mmol 4-methylumbelliferyl oleate l-l made up to a total of 3 ml with 100 mmol Tris/HCl, pH 8·0, l-1. Units of activity are defined as pmol substrate hydrolysed min-1. The stability of lipase was measured essentially as described by Andersson (1980) except that kanamycin (100 µg ml-1) was used to prevent regrowth of cells. Protease assays were essentially by the method of Wassif et al. (1995) , utilizing the substrate azocasein. An azocasein (Sigma) solution (0·5%, w/v, in 2 mM CaCl2, 50 mM Tris/HCl, pH 8) was dissolved at 37 °C. Cultures to be assayed were pelleted by centrifugation and 50 µl supernatant was added to 450 µl azocasein solution. The reaction was incubated at 37 °C for 30 min and then terminated by the addition of 125 µl 15% trichloroacetic acid. The sample was then centrifuged for 10 min at 14000 g to pellet unhydrolysed azocasein and the A440 of the supernatant determined. Units of activity are defined as {Delta}A440 (30 min)-1. Assays were conducted in triplicate and varied by <2%.

Chromosomal insertional mutagenesis.
To construct an aprX insertion mutant, a 578 bp internal fragment of the aprX gene was generated by PCR using the primers MU1 and MD2, with MU1 incorporating an EcoRI restriction site and MD2 incorporating a BamHI site (see above), and P. fluorescens B52 genomic DNA as the template. The PCR product was then digested with BamHI/EcoRI and ligated into the multiple cloning site of the suicide vector pJP5603 (which encodes kanamycin resistance; Penfold & Pemberton, 1992 ). Following transformation into S17.1{lambda}pir the pJP5603 construct was transferred to P. fluorescens B52 by conjugation (Christopher & Franklin, 1985 ). Kanamycin-resistant exconjugates, derived via a single crossover, were patched onto solid media containing 2·5% skimmed milk to screen for a protease-negative phenotype. One protease-negative exconjugate was purified and designated B52aprX::pJPkan.

To construct a prtB insertion mutant, the 1910 bp ApaI–SalI fragment from pRW52 containing a HindIII site was ligated into pSK with a deleted HindIII site (pSK{Delta}H). The chloramphenicol resistance cassette of pHP45{Omega}-Cm (Fellay et al., 1987 ) was excised with HindIII and ligated into the HindIII site of the cloned pSK insert. The pSK insert containing the chloramphenicol cassette was excised with SalI/KpnI and ligated into pJP5603. Chloramphenicol-resistant exconjugates were then screened on solid media containing kanamycin. Kanamycin-sensitive colonies were then retained as putatively due to a double cross-over event (allele replacement) with loss of pJP5603 and designated B52prtB::cam. The identity of both aprX and prtB mutants were confirmed by restriction and Southern analysis of genomic DNA.

Chromosomal lacZ gene fusions.
Homologous recombination and the suicide vector pJP5608, encoding tetracycline resistance (R. J. Penfold & J. M. Pemberton, personal communication), were used to create translational lipA'–'lacZ or aprX'–'lacZ gene fusions in the chromosome of P. fluorescens B52. pJP5608 constructs were made which contained a promoterless lacZ–kan cassette (Szabo et al., 1992 ) flanked at each end by either lipA or aprX sequence. In the case of lipA, the flanking regions were made as two overlapping PCR products with the first PCR product spanning the ATG start site using the two sets of primers shown below. The PCR incorporated an in-frame BamHI site where the PCR products overlapped and EcoRI sites at the extreme ends. Following amplification, the PCR products were digested with BamHI, ligated together, digested with EcoRI and then ligated into the EcoRI site of pSK{Delta}B (pSK with a deleted BamHI site) and designated pSK{Delta}BlipA. The lacZ–kan cassette was then excised from pLK (Szabo et al., 1992 ) with BamHI and ligated into the BamHI site of pSK{Delta}BlipA. The lacZ–kan cassette flanked by the PCR products was then excised from pSK{Delta}BlipA using EcoRI, inserted into the EcoRI site of pJP5608 and called pJP5608lipA. Construct pJP5608lipA was then delivered to the recipient via conjugation (Christopher & Franklin, 1985 ). Exconjugates were selected on kanamycin and recombinants derived by a double crossover identified by screening for tetracycline sensitivity. One such strain was designated B52B.

In the case of aprX, a 770 bp SacII–HindIII fragment from pRTMp11, which encompassed 60 bp upstream of the aprX start codon was cloned into pSK. Two primer sets (shown below) were then used to amplify two overlapping PCR products. The PCR products were digested with BamHI, ligated together, digested with SacII/AccI (sites which were incorporated from the polylinker of pSK during amplification) and ligated into pSK{Delta}B. The lacZ–kan cassette was then ligated into the BamHI site. The resulting lacZ–kan cassette, flanked by the aprX PCR products was then excised using SacI/KpnI (located within the second PCR product), ligated into SacI/KpnI sites of pJP5608 and called pJP5608aprX. This construct was delivered to the recipient via conjugation, exconjugates selected on kanamycin and recombinants derived by a double crossover identified by screening for tetracycline sensitivity; one such recombinant was designated B52aprX'–'lacZ.

The primers used for these constructions were as follows (restriction sites are shown as a single underline for BamHI and double underline for EcoRI). lipA primers, set 1: B5290, 5'-CGGAATTCTGCATGCCAGCTCCCACCG-3', B5240, 5'-TGGCTCCCAAGGTTTTGCAACCTAGGCG-3'; set 2: LSWR2, 5'-GCGGATCCCGATGCCATGGCGATCACG-3', B52500, 5'-GAACTAGTCGCTAGACGAGCCTTAAGGT-3'. aprX primers, set 1: T3, 5'-ATTAACCCTCACTAAAG-3', PRTMB1, 5'-CTGGCACCGCCGTTGGACCTAGGGC-3'; set 2: T7, 5'-GATATCACTCAGCATAA-3', PRTMB2 5'-CGGGATCCAATGGCAAACCGTCCTT-3'.

The region encompassing the fusion junction was amplified from the genome and sequenced to confirm that both fusion types were in-frame. For the lipA'–'lacZ fusion, primer RWLZ50c, which annealed within lacZ, and primer D2300 (shown below), which annealed upstream of primer B5290, were used to amplify a region that overlapped the junction point. The PCR product was gel purified and sequenced. The aprX'–'lacZ fusion junction point was confirmed in the same way using primers RWLZ50c and PRTMA.

Primers used to amplify and sequence fusion junction points were as follows: RWLZ50c, 5'-ATTCAGGCTGCGCAACTGTT-3', D2300, 5'-AGGCGACGCCGCCGCGCTCAAAGGCCACGA-3', PRTMA, 5'-CGGATCCGTCACATTTTCGCTGCGG-3'.

ß-Galactosidase assays.
These were as described by Miller (1972) . Assays were performed in duplicate and varied by <1·5%. Cell extracts were made using the bacterial protein extraction reagent kit, B-PER (Pierce). The protein concentration of the extracts was determined by the method of Smith et al. (1985) .


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Regulation of lipase and protease by temperature and iron concentration
We have compared the profile of lipase production with that of protease with respect to temperature (Fig. 1). The production of maximal levels of lipase at temperatures below the optimum growth rate showed the ‘classical’ pattern (Andersson, 1980 ), with a level of activity ~sixfold greater at 10–17 °C than at 27 °C, confirming the ‘low-temperature regulation’ for strain B52. P. fluorescens B52 has an optimal growth temperature of 27 °C: this is higher than the strain used by Andersson (1980) which has an optimal growth temperature of 20 °C. This is not due to thermolability of lipase activity (see also Andersson, 1980 ), since we have determined that there is no detectable loss of lipase activity in culture supernatants when incubated at 30 °C for at least 30 h (see Methods; data not shown).



View larger version (23K):
[in this window]
[in a new window]
 
Fig. 1. Pseudomonas fluorescens B52 lipase activity profile (a) and protease activity profile (b). Cultures were grown aerobically in HEPES medium to late stationary phase and the generation time determined ({bullet}). This was plotted against the maximal lipase ({blacksquare}) and protease ({diamondsuit}) activity obtained during growth. Error bars represent the standard error between replicate cultures.

 
The production of protease in the same culture showed a different pattern; production was very similar at 17 and 27 °C, and only decreased when the growth rate declined above 27 °C (see also McKellar & Cholette, 1987 ). This differs from the pattern observed by Gugi et al. (1991 ) for P. fluorescens MF0 although, in this latter study, the difference in protease levels between 17 and 30 °C was only about twofold. We conclude that ‘low-temperature regulation’ in P. fluorescens B52 is exemplified by lipase production and that the decreased production of protease above the optimal growth temperature may be related to the decreased growth rate.

The apr–inh–prt–lip genes
The lipase-encoding gene (lipA) and its 5' flanking sequence was isolated from P. fluorescens B52 using the lipA gene from strain LS107d2 (Johnson et al., 1992 ) as a probe (see Methods). Sequencing of this region revealed an ORF immediately upstream with 58% amino acid sequence similarity and 51% identity to an autotransporter protein (SSP-H2) from S. marcescens. This ORF was designated prtB (see below and Fig. 2).



View larger version (10K):
[in this window]
[in a new window]
 
Fig. 2. Genetic organization of the aprX–lipA region of P. fluorescens B52. Heavy arrows indicate sequenced regions and those indicated with dotted lines have been submitted to the GenBank database as three contigs derived from pRW52, pRW53, p5232 and pRTMp11. The sizes of the inserts are indicated in square brackets. Complementary sequence information can be found in Kawai et al. (1999) ; Liao & McCallus (1998) and Tan & Miller (1992) . The intergenic regions are: aprX–inh, 48 bp; inh–aprD, 106 bp; aprD–E, 0; aprE–F, 6 bp; aprF–prtA, 30 bp; prtA–prtB, 115 bp; and prtB–lipA, 120 bp. Potential stem–loop structures are indicated by pins labelled A–E. These were predicted by MFOLD (Jaeger et al., 1990 ) using a temperature of 27 °C; the computed free energies of helix formation (kJ mol-1) of A–E are -105, -50·6, -90, -191·2 and -132·2.

 
The gene encoding the metalloprotease of strain B52 was isolated by DNA amplification from genomic DNA using primers based on conserved regions amongst similar enzymes from P. aeruginosa and S. marcescens (see Methods). The sequence is 95% identical to a subsequently isolated metalloprotease-encoding gene from a plant pathogenic strain of P. fluorescens, which contains the Type I secretion genes immediately downstream (Liao & McCallus, 1998 ). Further sequence analysis and PCR amplification showed that homologous secretion genes are also located downstream of the metalloprotease gene of P. fluorescens B52 and we have hence retained the designation aprX (alkaline protease) for the metalloprotease gene and aprDEF for the secretion genes (see Fig. 2). Subsequent sequencing downstream and adjacent to aprDEF identified a second prt-like gene (prtA), followed by prtB and lipA, indicating that these genes (aprX–inh–aprDEF–prtAB–lipA) are contiguous in the genome. Following the completion of this analysis, a similar gene arrangement in another strain (P. fluorescens 33) has been reported (Kawai et al., 1999 ). However, in P. fluorescens strain SIK W1, the equivalent gene arrangement (aprX–inh–aprDEF–lipA) does not include the prtAB genes (Ahn et al., 1999 ). The aprX–lipA region is summarized in Fig. 2. Also, in P. fluorescens LS107d2 (Johnson et al., 1992 ), sequence analysis of the 1·4 kb upstream of lipA did not reveal either prtAB or aprDEF and Southern analysis revealed that prtB is absent from the genome (data not shown).

The DNA clones encoding aprX and prtB were used to construct corresponding chromosomal insertion mutations (see Methods). The prtB::cam mutant was not detectably altered in its protease production on skimmed milk media but is completely lipase deficient (Fig. 3a, b). Enzymic assay of culture supernatant confirmed these observations (data not shown). We conclude, firstly, that prtB::cam has a polar effect on lipA expression and that the lipA promoter is upstream of prtB. Secondly, although, by these criteria, PrtB does not possess detectable protease activity it may have specificity for an unknown substrate; alternatively, it, and prtA, may be outer-membrane proteins, as suggested in the case of homologues in S. marcescens (Ohnishi et al., 1997 ), with a primary role other than proteolysis of exogenous substrates.

The aprX::pJPkan mutant, derived by a single cross-over event (see Methods) was phenotypically protease-negative on skimmed milk media and had visibly lower lipase activity (Fig. 3a, b). Direct assay of culture supernatant confirmed the complete absence of protease activity, and the production of 10% of the wild-type levels of lipase activity (data not shown). In contrast, an aprX'–'lacZ fusion strain generated by allele replacement (strain B52B; see below and Methods) had a complete absence of lipase activity, as well as being protease deficient (Fig. 3a, b). We conclude that mutation of aprX has a polar effect on the expression of lipA, consistent with the aprX–lipA region constituting an operon (see below). The low level of lipase activity in aprX::pJPkan is presumed to derive from a vector promoter; we therefore conclude that transcriptional readthrough of aprX to lipA does occur, which is also indicative of an operon, although some transcriptional attenuation is not excluded by this observation alone.

Evidence that the aprX–inh–aprDEF–prtAB–lipA genes constitute a single operon
The complete absence of lipase activity in the aprX'–'lacZ fusion strain, generated by allele replacement (strain B52B; see above and Methods), could be due to a polar effect of this mutation on expression of the Type I secretion genes (aprDEF). This assumes that lipase is secreted via the aprDEF-encoded secretion apparatus for which there is evidence in an E. coli host (Kawai et al., 1999 ). Alternatively (or in addition), the lipase-negative phenotype of aprX'–'lacZ may be due to a direct polar effect of this mutation on expression of lipA. This assumes that the aprX–lipA region constitutes a single operon with a single promoter upstream of aprX. To distinguish between these possibilities, we complemented the aprX'–'lacZ mutant strain with the secretion genes located on a plasmid. If the polar effect on lipase production is due solely to a secretion defect, then the complemented strain is predicted to be lipase proficient; on the other hand, if a single promoter upstream of aprX (the aprX promoter, see below) is responsible for transcription, inclusive of lipA, then the complemented strain should remain lipase-negative. Accordingly, a derivative of pBBR1MCS, a broad-host-range plasmid, containing aprX–inh–aprDEF was constructed (pBBR1-32; see Methods). When pBBR1-32 was transferred to B52aprX'–'lacZ, the resulting strain was protease-positive, consistent with the functionality of the secretion genes, and lipase-negative (Fig. 3c, d). An identical result was obtained when an aprD insertion mutant (which is also protease and lipase-negative; data not shown) was complemented with pBBR1-32. The secretion genes on pBBR1-32 also confer a protease-positive phenotype on E. coli (see Methods) and complement an aprD mutant, providing additional evidence for their functionality. We conclude that the aprX–inh–aprDEF–prtAB–lipA genes constitute a single unit of transcription.

The aprX promoter
The aprX gene and the 5' untranslated region was isolated and used to map the transcriptional initiation site using S1 nuclease analysis (see Methods and Fig. 4a). We presume that the triplet of protected fragments is most likely an artefact due to progressive digestion by the S1 nuclease, but could represent initiation at three consecutive nucleotides. Hence the larger fragment is suggested to represent the transcriptional start point, which is the same at 17, 27 and 31 °C (data not shown). Examination of the region immediately upstream of the transcriptional initiation site identified no immediately obvious sigma binding consensus sequences (Wosten, 1998 ). To aid in defining putative promoter elements, the promoter region, as defined by the region upstream of the transcriptional initiation site, was compared with the equivalent sequence of P. aeruginosa (Duong et al., 1992 ). Two regions of strong sequence similarity between these otherwise dissimilar sequences were identified, located around -10 and -35 nt from the transcriptional start point (Fig. 4b). As shown in Fig. 4(c), they bear some similarity to consensus sequences recognized by both primary and ECF (extracytoplasmic function) sigma factors (Wosten, 1998 ; Missiakis & Raina, 1998 ). With regard to the upstream promoter motif, there is evidence that this may be located at -25/27 (Sexton et al., 1996 ) and hence two possible motifs are indicated (Fig. 4c). We conclude that these regions are likely to be involved in promoter recognition, although it is not yet clear which type of {sigma}70 factor is involved.



View larger version (25K):
[in this window]
[in a new window]
 
Fig. 4. (a) Determination of the aprX transcriptional start point (underlined) by S1 nuclease analysis. Total RNA was prepared from a B52 culture grown in HEPES media, subjected to S1 nuclease analysis and electrophoresed on a 6% denaturing polyacrylamide gel alongside a dideoxy sequencing ladder. The transcriptional start point is indicated by an arrow. Lane 1, probe plus RNA digested with S1 nuclease; lane 2, undigested probe. (b) Alignment of the P. fluorescens B52 aprX promoter sequence with the equivalent region of P. aeruginosa. Identical bases are indicated by an asterisk below the sequence. The transcriptional start point (+1) of P. fluorescens B52 aprX is double underlined. Boxed regions with nucleotides in bold indicate regions of high homology between the aprX promoter of P. fluorescens B52 and the equivalent region of P. aeruginosa. (c) Alignment of the P. fluorescens B52 promoter sequence with primary {sigma}70 and ECF {sigma} factor binding consensus sequences (R=A or G); consensus (1), see Wosten (1998) ; consensus (2), see Sexton et al. (1996 ). Underlined bases within the consensus sequences are present in the aprX sequence of B52.

 
Regulation of aprX
Transcriptional regulation of aprX was investigated using nuclease S1 protection to quantify aprX mRNA levels (see Methods). The results clearly indicate that aprX mRNA levels are strongly repressed by iron (Fig. 5a). This is not surprising since it seems likely that transcriptional repression occurs (albeit perhaps indirectly) via the iron-sensing Fur repressor (see Discussion).



View larger version (15K):
[in this window]
[in a new window]
 
Fig. 5. Regulation of aprX. (a) aprX mRNA levels. Independent cultures of strain B52 grown in HEPES media ({bullet}) and in HEPES media supplemented with 20 µM iron ({blacktriangleup}) were incubated at 17 °C. AprX promoter activity was monitored during growth by S1 analysis and quantified by phosphorimaging. (b) Regulation of a translational aprX'–'lacZ fusion. Cultures of B52aprX'–'lacZ were grown in HEPES media ({bullet}) and in HEPES media supplemented with 20 µM iron ({blacktriangleup}).

 
A translational aprX'–'lacZ fusion within the chromosome was constructed by allele replacement (see Methods) and used to monitor aprX expression. Again strong repression of the fusion by iron was observed (Fig. 5b), which is consistent with the results of the mRNA analysis (Fig. 5a).

Regulation of a lipA'–'lacZ fusion
In view of the regulation of lipase by temperature as well as by iron, we have investigated the regulation of a chromosomal lipA'–'lacZ translational fusion, obtained by allele replacement (see Methods).

The results for the fusion strain, B52B, are presented together with those for lipase production in the parental strain, P. fluorescens B52, in parallel cultures grown at 17 °C and 27 °C (Fig. 6). In the case of cultures of B52 containing iron, the results (Fig. 6a, b) confirmed that lipase production is repressed by iron at both temperatures. ß-Galactosidase synthesis in the fusion strain B52B is also repressed by iron (Fig. 6a, b), consistent with repression at the mRNA level from the aprX promoter (see Fig. 5a); this result strongly suggests that the lipA'–'lacZ fusion is regulated normally.



View larger version (21K):
[in this window]
[in a new window]
 
Fig. 6. Regulation of a translational lipA'–'lacZ fusion. Cultures of P. fluorescens B52 and B52B (lipA'–'lacZ) were grown in parallel in HEPES media ({bullet}) and in HEPES media supplemented with 20 µM iron ({blacktriangleup}) and incubated at 17 °C (a) or 27 °C (b). ß-Galactosidase and lipase activity were measured during growth until late stationary phase.

 
The maximum levels of lipase, in the absence of repression, at 17 °C is sixfold higher than at 27 °C, demonstrating low-temperature regulation. However, ß-galactosidase synthesis in strain B52B is not significantly increased at low temperature (Fig. 6), indicating a mechanism of regulation other than at the level of transcriptional or translational initiation.


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The aprX promoter has been identified by transcription mapping and by comparison of the 5' untranslated region of aprX with that from a relatively unrelated species, P. aeruginosa (Fig. 4). The -10 and -27/35 regions show some similarity to those recognized by both primary and ECF sigma factors. It is possible that the PbrA sigma factor (Sexton et al., 1995 ) may be involved. PbrA shows significant sequence similarity to ECF sigma factors; it is involved in the regulation of protease production in a phytopathogenic strain of P. fluorescens and it is regulated by the iron responsive Fur repressor (Sexton et al., 1995 ; Venturi et al., 1995 ). This would be consistent with the apparent absence of Fur recognition sequences in the aprX promoter region (Fig. 4 and data not shown). We have so far been unable to isolate pbrA insertion mutants of B52 in order to test this hypothesis. A possible role for the homologue of PbrA clearly merits further investigation, as does further characterization of the aprX promoter region. Both the aprX and lipA fusions are regulated by iron levels, as are the levels of mRNA initiated from the aprX promoter, consistent with regulation at the level of transcription from the aprX promoter.

Although the lipA'–'lacZ fusion strain is transcribed and regulated by iron regulation, it is not up-regulated by low temperature (Fig. 6). This evidence suggests that temperature regulation is not at the level of transcriptional nor translational initiation. A post-transcriptional mechanism at the level of mRNA stability is possible, although it is not immediately evident why this is not reflected in a lipA'–'lacZ fusion, which involves neither deletion nor replacement of DNA. It is possible that the lacZ–kan cassette affects the stability of the transcript (see Goldenberg et al., 1996 ).

It has recently been reported that when lipase and the secretion genes from strain SIK W1 are independently expressed in E. coli, lipase production is abolished at 30 °C but is retained at 25 and 20 °C (Ahn et al., 1999 ). It was suggested that the function of the secretory apparatus is compromised at the higher temperatures due to fluidity, or other structural changes, in the inner membrane. Since this was not observed when the homologous genes from P. aeruginosa were used, it was hypothesized that this reflected an adaptation of the P. fluorescens AprDEF proteins in a psychrotroph. This hypothesis is contradicted by the fact that protease, also exported using AprDEF (data not shown), is not low-temperature regulated (Fig. 1). We suggest that these results may be explained as a response to a temperature which is above the optimum temperature for growth of P. fluorescens. That is, one or more of the AprDEF gene products are misfolded at such temperatures. It would clearly be desirable to compare the temperature profile (Andersson, 1980 ; and see Fig. 1) of lipase secretion in P. fluorescens, using secretion systems from P. aeruginosa and P. fluorescens.

An alternative model for low-temperature regulation proposes that the decay rate of the lipase-encoding segment of the polycistronic messenger varies with temperature. Differential segmental decay rates have been observed in the case of the puf operon, and furthermore, are subject to regulation by oxygen (Klug, 1991 ; Heck et al., 1996 ; Rauhut & Klug, 1999 ). Stem–loop structures are known to influence mRNA decay and a number of these have been noted within the intergenic regions of the B52 operon (see Fig. 2).

In summary, we present evidence that the contiguous genes aprX–inh–aprDEF–prtAB–lipA function as an operon, transcribed from the aprX promoter. The data are consistent with transcriptional regulation of the operon by iron but suggest that low temperature regulation of lipase is at the post-transcriptional level, possibly involving segmental regulation of mRNA decay rates.

Finally, our results bear on the question of whether AprX is solely responsible for casein degradation in milk. Insertion mutations in aprX are not proteolytic on skimmed milk media (Liao & McCallus, 1998 ; Fig. 3) but since such mutations are polar with respect to lipA and therefore also prtAB, then PrtA and/or PrtB could potentially be involved in casein degradation, although involvement of PrtB has been directly excluded (see Results). However, the aprX::pJPkan mutant can reasonably be expected to express prtAB, since the downstream lipA gene is expressed (see above and Results); since this strain is not proteolytic on skimmed milk media, we conclude that AprX is indeed solely involved.


   ACKNOWLEDGEMENTS
 
R.G.W. acknowledges the receipt of an Australian postgraduate award. This work was supported by the Australian Research Council and, in part, by the Dairy Research Development Corporation. R.G.W. and M.B. made an equal contribution to this work.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Ahn, J. H., Pan, J. G. & Rhee, J. S.(1999). Identification of the tliDEF transporter specific for lipase in Pseudomonas fluorescens SIK W1. J Bacteriol 181, 1847-1852.[Abstract/Free Full Text]

Aiba, H., Adhya, S. & de Crombrugghe, B.(1981). Evidence for two functional gal promoters in intact Escherichia coli cells. J Biol Chem 256, 11905-11910.[Abstract/Free Full Text]

Andersson, R. E.(1980). Microbial lipolysis at low temperatures. Appl Environ Microbiol 39, 36-40.[Medline]

Andersson, R. E., Hedlund, C. B & Jonsson, U.(1979). Thermal inactivation of a heat-resistant lipase produced by the psychrotrophic bacterium Pseudomonas fluorescens. J Dairy Sci 62, 361-367.[Medline]

Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A. & Struhl, K. (1987). Current Protocols in Molecular Biology. New York: Greene Publishing Associates and Wiley Interscience.

Christopher, F. & Franklin, H.(1985). Broad-host range cloning vectors for Gram-negative bacteria. In DNA Cloning, a Practical Approach , pp. 165-183. Edited by D. M. Glover & B. D. Hames. Oxford:IRL Press.

Duong, F., Lazdunski, A., Cami, B. & Murgier, M.(1992). Sequence of a cluster of genes controlling synthesis and secretion of alkaline protease in P. aeruginosa: relationship to other secretory pathways. Gene 121, 47-54.[Medline]

Falconi, M., Colonna, B., Prosseda, G., Micheli, G. & Gualerzi, C. O.(1998). Thermoregulation of Shigella and Escherichia coli EIEC pathogenicity: a temperature-dependent structural transition of DNA modulates accessibility of virF promoter to transcriptional repressor H-NS. EMBO J 17, 7033-7043.[Abstract/Free Full Text]

Fellay, R., Frey, J. & Krisch, H.(1987). Interposon mutagenesis of soil and water bacteria: a family of DNA fragments designed for in vitro insertional mutagenesis of gram-negative bacteria. Gene 52, 147-154.[Medline]

Goldenberg, D., Azar, I. & Oppenheim, A. B(1996). Differential mRNA stability of the cspA gene in the cold-shock response of Escherichia coli. Mol Microbiol 19, 241-248.[Medline]

Gugi, B., Orange, M., Hellio, F., Burini, J. F., Guillou, C., Leriche, F. & Guespin-Michel, J. F.(1991). Effect of growth temperature on several exported enzyme activities in the psychrotrophic bacterium Pseudomonas fluorescens. J Bacteriol 173, 3814-3820.[Medline]

Heck, C., Rothfuchs, R., Jäger, A., Rauhut, R. & Klug, G.(1996). Effect of the pufQ-pufB intercistronic region on puf mRNA stability in Rhodobacter capsulatus. Mol Microbiol 20, 1165-1178.[Medline]

Jaeger, J. A., Turner, D. H. & Zuker, M.(1990). Predicting optimal and suboptimal structures for RNA. Methods Enzymol 183, 281-306.[Medline]

Johnson, L. A., Beacham, I. R., MacRae, I. C. & Free, M. L.(1992). Degradation of triglycerides by a pseudomonad isolated from milk: molecular analysis of a lipase-encoding gene and its expression in Escherichia coli. Appl Environ Microbiol 58, 1776-1779.[Abstract]

Kawai, E., Idei, A., Kumura, H., Shimazaki, K., Akatsuka, H. & Omori, K.(1999). The ABC-exporter genes involved in the lipase secretion are clustered with the genes for lipase, alkaline protease, and serine protease homologues in Pseudomonas fluorescens no. 33. Biochim Biophys Acta 1446, 377-382.[Medline]

Klug, G.(1991). Endonucleolytic degradation of puf mRNA in Rhodobacter capsulatus is influenced by oxygen. Proc Natl Acad Sci USA 88, 1765-1769.[Abstract]

Kovach, M. E., Phillips, R. W., Elzer, P. H., Roop, R. M.II & Peterson, K. M.(1994). PBBR1MCS: a broad-host-range cloning vector. Bio/Technology 16, 800-802.

Liao, C.-H. & McCallus, D. E.(1998). Biochemical and genetic characterisation of an extracellular protease from Pseudomonas fluorescens CY091. Appl Environ Microbiol 64, 914-921.[Abstract/Free Full Text]

McKay, D. B. & Beacham, I. R.(1995). The effect of temperature on the degradation of triglycerides by a pseudomonad isolated from milk: free fatty acid accumulation as a balance between rates of triglyceride hydrolysis and fatty acid consumption. J Appl Bacteriol 79, 651-656.

McKay, D. B., Dieckelmann, M. & Beacham, I. R.(1995). Degradation of triglycerides by a pseudomonad isolated from milk: the roles of lipase and esterase studied using recombinant strains overproducing, or specifically deficient in these enzymes. J Appl Bacteriol 78, 216-223.[Medline]

McKellar, R. C(1989). Regulation and control of synthesis. In Enzymes of Psychrotrophs in Raw Food , pp. 153-172. Edited by R. C. McKellar. Boca Raton, FL:CRC Press.

McKellar, R. C. & Cholette, H.(1987). Effect of temperature shifts on extracellular proteinase-specific mRNA pools in Pseudomonas fluorescens B52. Appl Environ Microbiol 53, 1973-1976.[Medline]

McKellar, R. C., Shamsuzzaman, K., San Hose, C. & Cholette, H.(1987). Influence of iron(III) and pyoverdine on extracellular proteinase and lipase production by Pseudomonas fluorescens B52. Arch Microbiol 147, 225-230.[Medline]

Merieau, A., Gugi, B., Guespin-Michel, J. F. & Orange, N.(1993). Temperature regulation of lipase secretion by Pseudomonas fluorescens strain MFO. Appl Microbiol Biotechnol 39, 104-109.

Miller, J. H. (1972). Experiments in Molecular Genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Missiakis, D. & Raina, S.(1998). The extracytoplasmic function sigma factors: role and regulation. Mol Microbiol 28, 1059-1066.[Medline]

Nakahama, K., Yoshimura, K., Matumoto, R., Kikuchi, M., Lee, I. S., Hase, T. & Matsubara, H.(1986). Cloning and sequencing of Serratia protease gene. Nucleic Acids Res 14, 5843-5855.[Abstract]

Ohnishi, Y., Beppu, T. & Horinouchi, S.(1997). Two genes encoding serine protease homologues in Serratia marcescens and characterization of their products in Escherichia coli. J Biochem 121, 902-913.[Abstract]

Penfold, R. J. & Pemberton, J. M.(1992). An improved suicide vector for the construction of chromosomal insertion mutations in bacteria. Gene 118, 145-146.[Medline]

Rauhut, R. & Klug, G.(1999). mRNA degradation in bacteria. FEMS Microbiol Rev 23, 353-370.[Medline]

Richardson, B. C.(1981). The purification and characterization of a heat-stable protease from Pseudomonas fluorescens B52. NZ J Dairy Sci Technol 16, 195-207.

Richardson, B. C. & Te Whaiti, I. E.(1978). Partial characterization of heat-stable extracellular proteases of some psychrotrophic bacteria from raw milk. NZ J Dairy Sci Technol 13, 172-176.

Roy, R. N.(1980). Fluorimetric assay of the activity of extracellular lipases of Pseudomonas fluorescens and Serratia marcescens. J Appl Bacteriol 49, 265-271.[Medline]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sexton, R., Gill, P. R.Jr, Callanan, M. J., O’Sullivan, D. J., Dowling, D. N. & O’Gara, F.(1995). Iron-responsive gene expression in Pseudomonas fluorescens M114: cloning and characterisation of a transcription-activating factor, PbrA. Mol Microbiol 15, 297-306.[Medline]

Sexton, R., Gill, P. R.Jr, Dowling, D. N. & O’Gara, F.(1996). Transcriptional regulation of the iron-responsive sigma factor gene pbrA. Mol Gen Genet 250, 50-58.[Medline]

Smith, P. K., Krohn, R. I., Hermanson, G. T. & 7 other authors (1985). Measurement of protein using bicinchoninic acid. Anal Biochem 150, 76–85.[Medline]

Stead, D.(1984). Evaluation of a fluorometric assay on the lipases from strains of milk psychrophilic bacteria. J Dairy Res 51, 123-130.

Szabo, M., Maskell, D., Butler, P., Love, J. & Moxon, R.(1992). Use of chromosomal gene fusions to investigate the role of repetitive DNA in regulation of genes involved in lipopolysaccharide biosynthesis in Haemophilus influenzae. J Bacteriol 174, 7245-7252.[Abstract]

Tan, Y. & Miller, K. J.(1992). Cloning, expression and nucleotide sequence of a lipase gene from Pseudomonas fluorescens B52. Appl Environ Microbiol 58, 1402-1407.[Abstract]

Venturi, V., Weisbeek, P. & Koster, M.(1995). Gene regulation of siderophore-mediated iron acquisition in Pseudomonas: not only the Fur repressor. Mol Microbiol 17, 603-610.[Medline]

Wassif, C., Cheek, D. & Belas, R.(1995). Molecular analysis of a metalloprotease from Proteus mirabilis. J Bacteriol 177, 5790-5798.[Abstract]

Wosten, M. M.(1998). Eubacterial sigma-factors. FEMS Microbiol Rev 22, 127-150.[Medline]

Received 29 August 2000; revised 16 October 2000; accepted 30 October 2000.