Elastase is a major virulence factor in Pseudomonas aeruginosa that is believed to cause extensive tissue damage during infection in the human host. Elastase is secreted in non-mucoid P. aeruginosa. It is known that secretion of most virulence factors such as elastase, lipase, exotoxin A, etc., in P. aeruginosa is greatly reduced in alginate-secreting mucoid cells isolated from the lungs of cystic fibrosis (CF) patients. We have previously reported that in mucoid P. aeruginosaan intracellular protease cleaves the 16 kDa form of nucleoside diphosphate kinase (Ndk) to a truncated 12 kDa form. This smaller form is membrane associated and has been observed to form complexes with specific proteins to predominantly generate GTP, an important molecule in alginate synthesis. The main aim of this study was to purify and characterize this protease. The protease was purified by hydrophobic interaction chromatography of the crude extract of mucoid P. aeruginosa 8821, a CF isolate. Further analysis using a gelatin containing SDS–polyacrylamide gel detected the presence of a 103 kDa protease, which when boiled, migrated as a 33 kDa protein on a SDS–polyacrylamide gel. The first 10 amino acids from the N-terminus of the 33 kDa protease showed 100% identity to the mature form of elastase. An elastase-negative lasB ::Cm knock-out mutant in the mucoid 8821 background was constructed, and it showed a non-mucoid phenotype. This mutant showed the presence of only the 16 kDa form of Ndk both in the cytoplasm and membrane fractions. We present evidence for the retention of active elastase in the periplasm of mucoid P. aeruginosa and its role in the generation of the 12 kDa form of Ndk. Finally, we demonstrate that elastase, when overproduced in both mucoid and non-mucoid cells, stimulates alginate synthesis. This suggests that the genetic rearrangements that trigger mucoidy in P. aeruginosa also allow retention of elastase in the periplasm in an active oligomeric form that facilitates cleavage of 16 kDa Ndk to its 12 kDa form for the generation of GTP, required for alginate synthesis.
Pseudomonas aeruginosa is an opportunistic pathogen that produces a number of virulence factors, many of which, such as elastase, exotoxin A, lipase, etc., are secreted outside the cell ( Nicas and Iglewski, 1985 ; Galloway, 1991 ; Stanislavsky and Lam, 1997 ). The mechanism of secretion of elastase involves cleavage of the signal sequence when the prepro form of elastase passes from the cytoplasm to the periplasm and the separation of the propeptide from the mature elastase during its passage through the outer membrane so that only mature elastase is found outside the cell ( Galloway, 1991 ; Kessler et al., 1992 ). Although the propeptide has been shown to be non-covalently associated with elastase in the periplasm, leading to an inhibition of elastolytic activity, and the propeptide is believed to contribute to the secretion of mature elastase, acting as an intramolecular chaperon, a recent report describes the secretion of the propeptide as well ( McIver et al., 1995 ; Braun et al., 1996 ; 1998 ). It is not known whether elastase serves any function in the life cycle of P. aeruginosa or is simply secreted as a virulence factor during infection of the eukaryotic cells to cause tissue damage.
Another virulence factor in P. aeruginosa is alginate, a polysaccharide that encapsulates the infecting cells of P. aeruginosa in the lungs of cystic fibrosis (CF) patients. It is interesting to note that alginate synthesis is triggered primarily in the lung environment of the CF patients and only rarely during other infections, such as eye, burn or urinary tract infections (May and Chakrabarty, 1994a; Govan and Deretic, 1996).
An important consequence of alginate synthesis in the CF lung, leading to mucoidy in P. aeruginosa, is a greatly reduced level of secretion of other virulence factors such as elastase, exotoxin A, etc. (Ohman and Chakrabarty, 1982; Woods et al., 1991). Although the promoters of lasB and algD (Chitnis and Ohman, 1993) were initially shown to be inversely related in some P. aeruginosa strains (Mohr et al., 1990), more recent results report a positive correlation between algD transcript accumulation and both lasB and lasA transcript levels (Storey et al., 1997). Thus, the basis of the reduced secretion of elastase in mucoid cells that secrete copious amounts of alginate is unknown at present. As reduced secretion of virulence factors occurs in all mucoid strains studied so far, it is possible that the polysaccharide secretion machinery of P. aeruginosa directly interferes with the protein secretory apparatus.
Is there a physiological consequence or relationship of the reduced secretion of virulence factors such as elastase and alginate synthesis in mucoid P. aeruginosa cells? We previously speculated that reduced secretion of the virulence factors in the CF lung environment reduces tissue damage, thereby allowing P. aeruginosa to survive longer and cause chronic infections (Ohman and Chakrabarty, 1982). In this report we demonstrate that reduced secretion of elastase allows an accumulation of enzymatically active oligomeric form of elastase in the mucoid cell periplasm. This intracellular accumulation has important implications in the life cycle of the cells, namely that the intracellular elastase allows cleavage of the normally cytoplasmic 16 kDa form of nucleoside diphosphate kinase (Ndk) to a 12 kDa form that is known to associate with the membrane and forms complexes with other proteins such as pyruvate kinase (Pk), Pra, etc. (Shankar et al., 1996; Chakrabarty, 1998). Although cytoplasmic 16 kDa Ndk generates all the nucleoside triphosphates (NTPs), the 12 kDa Ndk complexes generate predominantly GTP (Chopade et al., 1997; Kim et al., 1998). GTP is an important nucleotide for synthesis of GDP-mannose, an obligatory intermediate in alginate synthesis (May et al., 1994). One molecule of GTP is consumed for every molecule of mannuronic acid in the alginate polymer, hence the cleavage of 16 kDa Ndk by elastase to generate the 12 kDa form ensures a steady supply of GTP for GDP-mannose synthesis. This may be an important mechanism devised by P. aeruginosa to allow a continuous supply of GTP for synthesis of alginate.
Purification and N-terminal sequencing of the protease that cleaves Ndk
The proteolytic activity that cleaves Ndk from a 16 kDa form to a 12 kDa form, both of which can undergo autophosphorylation, was purified from the whole-cell extract of mucoid P. aeruginosa 8821. After final purification on the TSK–phenyl hydrophobic interaction chromatography column, analysis of the proteolytic activity on SDS–PAGE revealed the presence of a protein migrating at molecular mass of 103 kDa (Fig. 1, lane 3). However, it is to be noted that these samples were not boiled before electrophoresis. Interestingly, on boiling, the protein migrated with a molecular mass of about 33 kDa (Fig. 1, lane 4). Similar observations have been reported with catalases of P. aeruginosa. Catalases have been observed to migrate in their active native form (228 kDa) when they are not boiled before electrophoresis (Brown et al., 1995). However, on boiling they migrate in their monomeric form (57 kDa) on SDS–PAGE. Similar behaviour was observed with the purified protease from mucoid P. aeruginosa 8821. N-terminal amino acid sequence analysis of both the high-molecular-weight form (103 kDa) and the low-molecular-weight form (33 kDa) was determined to be AEAGGPGGNQ and was found to be 100% identical to the mature form of elastase (Bever and Iglewski, 1988).
Purified elastase from P. aeruginosa PAO1 supernatant exhibits Ndk cleaving activity
To confirm our observations that the protease which cleaved Ndk is indeed elastase, we purified elastase (see Experimental procedures) from P. aeruginosa PAO1 cell-free growth medium. PAO1 is a non-mucoid strain that is known to secrete elastase in the medium. The cell-free supernatant (growth medium) was subjected to a 80% ammonium sulphate precipitation and the pellet was subjected to TSK–phenyl hydrophobic interaction chromatography. The purified protein fraction was tested for its ability to cleave the 16 kDa Ndk. As shown in Fig. 2 (lanes 2 and 3), both elastase and the purified protease cleaved the 16 kDa Ndk to a 12 kDa form. Prolonged incubation did not allow any further cleavage of the 12 kDa form.
Western blot analysis with anti-elastase antibody
To address the apparent discrepancy in the molecular mass of elastase (boiled versus non-boiled, Fig. 1), a Western blot analysis was performed. Both the protease and the elastase were subjected to SDS–PAGE. However, before loading, each sample was either not boiled or boiled after addition of sample buffer. As shown in Fig. 3, both the purified protease and the elastase cross-reacted to the anti-elastase antibody, suggesting that the 103 kDa elastase comprises 33 kDa monomers.
Zymographic analysis of elastase and protease
We were interested in finding out whether the higher molecular-mass form of elastase was enzymatically active. We therefore performed zymography by running purified protease and elastase on a 10% gelatin–SDS polyacrylamide gel. Zymography enables the detection of active proteases on a SDS–PAGE gel. Proteolytic activity can be seen as clear zones against a dark blue background. As demonstrated by the zymogram, the higher molecular-mass form of elastase was enzymatically active (Fig. 4A). Zymogram analysis was also performed under reducing conditions to test whether the oligomers were held by disulphide bonds. The addition of reducing agent (β-mercaptoethanol) did not cause conversion of the 103 kDa form of elastase into its monomers (Fig. 4B). This appears to indicate that the monomers are held together by a stronger interaction that could be disrupted only by boiling (Fig. 3, lanes 3 and 5). The higher molecular weight form of elastase has also been reported recently (O′Callaghan et al., 1996).
Elastase induces alginate synthesis in non-mucoid P. aeruginosa
To determine whether elastase is the only enzyme in P. aeruginosa that processes the 16 kDa Ndk, we constructed a lasB ::Cm knock-out mutant by allelic exchange in mucoid P. aeruginosa 8821. We could not detect the 12 kDa form of Ndk in the membrane fractions, which suggested that elastase is possibly the only enzyme that cleaves Ndk (Fig. 5, lane 6). We also observed that the knock-out mutant 8821 lasB ::Cm colonies were non-mucoid and produced extremely low amounts of alginate (Table 1). This raised the question of whether lasB is important for alginate synthesis. The lasB ::Cm mutant was then complemented by introducing plasmid pSAK5 harbouring the lasB gene. Although the colony did not show much mucoidy, alginate synthesis was restored (Table 1) and the 12 kDa form of Ndk was detected (data not shown). As introduction of plasmid copy of lasB could restore alginate biosynthesis in the non-mucoid lasB ::Cm mutant, we tested whether P. aeruginosa 8821NM, a non-mucoid segregant of mucoid strain 8821 and P. aeruginosa PAO1, a normally non-mucoid strain, could be made to synthesize alginate when supplied with the lasB gene on a high-copy plasmid. In both the strains tested, alginate was measured from cells grown on agar plates. As can be seen from the data in Table 1, introduction of the lasB gene enhanced alginate synthesis significantly compared with the non-mucoid parent strains or such strains harbouring vector plasmids as control. In 8821NM or PAO1, with or without the vector plasmid, alginate synthesis was negligible. Introduction of pSAK5 harbouring the lasB gene allowed restoration of alginate synthesis to the wild-type level. This reinforced the concept that elastase, encoded by the lasB gene, is important for alginate synthesis because not only its deficiency in mucoid strain 8821 (8821 lasB ::Cm) leads to a non-mucoid (non-producing) phenotype, but introduction of lasB on a multicopy plasmid (pSAK5) in either mucoid 8821 or non-mucoid 8821NM or PAO1 significantly enhanced alginate synthesis. Enhanced elastase production in non-mucoid strains such as 8821NM or PAO1, which normally secretes elastase completely, allows substantial retention of elastase in the periplasmic space (Table 1), which presumably contributes to the cleavage of 16 kDa Ndk to the 12 kDa form, which is necessary for preferential GTP synthesis (Chakrabarty, 1998).
Table 1. . Estimation of alginate and elastase activity in various constructs a . a. Alginate was estimated according to May and Chakrabarty (1994b ). These experiments were performed three times with equivalent results. b. Samples were assayed in duplicates and averaged. c. Elastase activity was measured both from cell-free supernatant (SUP) and the periplasmic extracts (PERI) by ECR assay ( Ohman et al., 1980 ). Elastolytic activity is expressed as OD 495 per mg protein per h ( Ohman et al., 1980 ).
One of the most interesting aspects of the present study is the role of elastase in alginate synthesis. A knock-out mutation in lasB, encoding elastase, in a mucoid cellular background renders it non-mucoid and unable to produce alginate. To see whether the insertion of the antibiotic cassette in the lasB gene may have a polar effect on a downstream gene involved in alginate synthesis, we introduced the lasB gene in trans in such a knock-out mutant. Introduction of the lasB gene alone restored alginate synthesis in such a mutant, demonstrating an important role of elastase in alginate synthesis. Indeed, introduction of additional copies of lasB in non-mucoid segregants derived from mucoid CF-isolate P. aeruginosa or even in normally non-mucoid P. aeruginosa PAO1 allowed significant alginate synthesis by such cells. It is to be noted that non-mucoid cells such as PAO1 or the non-mucoid segregants of mucoid CF isolates normally secrete the elastase to the outside medium and very little elastase is found inside the cell. In contrast, either mucoid cells or non-mucoid cells that harbour lasB-containing plasmids (and thereby hyperproduce elastase) retain substantial amount of elastase in their periplasm. Thus, there is a positive correlation between the retention of elastase in the periplasm and the synthesis of alginate by P. aeruginosa.
How may intracellular elastase contribute to alginate synthesis? In this report we demonstrate that elastase appears to be the only enzyme that is involved in the cleavage of the 16 kDa cytoplasmic Ndk to a 12 kDa form that has been shown to be associated with the membrane fractions of alginate producing mucoid cells (Shankar et al., 1996). We demonstrate in this paper that lasB mutants of mucoid cells harbour only the 16 kDa form of Ndk either in the cytoplasm or in the membrane (Fig. 5). We also have reported that only the 12 kDa form of Ndk, but not the 16 kDa form, can form efficient complexes with pyruvate kinase or Pra (Chopade et al., 1997; Kim et al., 1998). Such complexes of Ndk modulate the specificity of Ndk so that the complex generates predominantly GTP rather than all the NTPs (Chakrabarty, 1998). Alginate is a polysaccharide of mannuronic and guluronic acids, the latter is an epimer of mannuronic acid and is derived from it. A molecule of alginate may contain hundreds of mannuronate residues each of which comes from GDP-mannuronic acid, each of which in turn requires a molecule of mannose 1-phosphate and GTP (May et al., 1994). Thus, synthesis of each molecule of alginate involves consumption of hundreds of molecules of GTP. Once P. aeruginosa commits itself to make alginate, ostensibly for its own protection in the lungs of CF patients, it must devise a mechanism to ensure a steady supply of GTP (and mannose 1-phosphate). Thus, linking alginate synthesis with the intracellular retention of elastase, so that cleaved Ndk can generate the GTPs, is an important cellular mechanism that uses one virulence factor to generate another.
What is the mechanism in the mucoid cell that prevents elastase secretion to the outside and facilitates its retention in an oligomeric form in the periplasm? The propeptide, which contributes to elastase secretion (McIver et al., 1995; Braun et al., 1996), is either degraded partly in the periplasm or it may be secreted to the outside medium (Braun et al., 1998). We have not looked at the fate of the propeptide in mucoid cells. It is likely that the products of alginate biosynthetic genes such as algE, alg8, alg44, etc., which are normally silent in non-mucoid P. aeruginosa but are specifically activated in mucoid cells and which are believed to be involved in the polymerization and secretion of alginate (Chu et al., 1991; Maharaj et al., 1993), somehow interfere physically with the secretion of secretable proteins such as elastase or exotoxin A. The linkage, if any, between polysaccharide secretion and protein secretion will be an interesting topic to study in the future.
The triggering of alginate synthesis in non-mucoid segregants or in strain PAO1 by the introduction of additional copies of the lasB gene (Table 1) appears to indicate that lack of alginate synthesis by the non-mucoid cells is not only due to a low level of some critical gene products such as AlgT (Goldberg et al., 1993), but also due to a limiting level of precursors such as GTP. It is known that mucoid cells grow more slowly than non-mucoid cells, presumably through diversion of its NTP pools predominantly to GTP. It should be pointed out that non-mucoid segregant or PAO1 cells having extra copies of lasB, although capable of producing significant amount of alginate, nevertheless look less mucoid than CF isolates. This may be due to production of a more compact, high-molecular-weight alginate or more likely due to association of alginate on the cellular surface rather than its free secretion as the levels of channels such as AlgE (Chu et al., 1991) may be limiting in non-mucoid cells. It is to be noted that alginate-producing PAO1 cells harbouring extra copies of lasB still secrete substantial amount of elastase (Table 1), presumably because of less interference from a low level of alginate-secreting channels present in the non-mucoid background.
Our observations also raise an interesting question regarding the specific role of elastase in alginate synthesis. We report here that in P. aeruginosa, elastase is important for alginate synthesis; however, elastase secretion has not been reported in many other bacteria that are known to synthesize alginate under specific conditions. Alginate is synthesized by Azotobacter vinelandii (Campos et al., 1996; Rehm et al., 1996), but the elaboration of elastase by A. vinelandii has not been reported. Penaloza-Vazquez et al. (1997) have recently reported the triggering of alginate synthesis in the phytopathogen P. syringae by copper. P. syringae is not known to elaborate elastase. What mechanism may then be operative in providing the steady supply of GTP during alginate synthesis in P. syringae? Will introduction of the lasB gene in non-mucoid P. syringae, in absence or in presence of copper, allow intracellular elastase production and a triggering of alginate synthesis? Future attempts to evaluate whether cleavage of Ndk is an important event in predominant GTP synthesis, as it appears to be in P. aeruginosa, or whether such cleavage may be mediated by proteases other than elastase or whether there are other mechanisms that allow extensive GTP synthesis will provide important insights to the mechanism of NTP and alginate synthesis in other microorganisms as well.
An interesting aspect of alginate synthesis by P. aeruginosa is the fact that the bulk of alginate synthesis occurs during the late log/early stationary phase (Tatnell et al., 1993; Hassett, 1996). It now appears that this synthesis is modulated by the availability of GTP, which is provided by the membrane-associated complexes of 12 kDa Ndk with pyruvate kinase and Pra (Chopade et al., 1997). It is known that Ndk is 16 kDa in size and cytoplasmic during the log phase but is gradually converted to the 12 kDa membrane-associated form during early stationary phase when the cell density is high (Shankar et al., 1996). The pyruvate kinase and Pra are also detected in the membrane at the late log to early stationary phase (Chopade et al., 1997). We have shown that elastase appears to be the only enzyme in P. aeruginosa that cleaves the 16 kDa Ndk to the 12 kDa form as the lasB mutant of the mucoid strain 8821 shows only the presence of the 16 kDa form (Fig. 5). It is to be noted that elastase synthesis occurs only at high cell density and is controlled by quorum sensing through modulation of the lasR gene by homoserine lactones (Pearson et al., 1994; Brint and Ohman, 1995; Latifi et al., 1996). Thus, conversion of the 16 kDa Ndk to the 12 kDa form only occurs at high cell density when elastase production is triggered. The combined role of Ndk, which has been previously shown to be important for alginate synthesis (Sundin et al., 1996), and elastase therefore illustrates the important role of elastase in P. aeruginosa cell physiology.
Bacterial strains and plasmids are listed in Table 2. Escherichia coli and P. aeruginosa strains were maintained in Luria–Bertani (LB) and Pseudomonas isolation agar (PIA; Difco) media respectively; all strains were grown at 37°C. For plasmid maintenance in E. coli, ampicillin was used at a concentration of 100 μg ml−1. P. aeruginosa, with the appropriate plasmid, was grown in LB with chloramphenicol at 500 μg ml−1, or 500 μg ml−1 of carbenicillin.
Purification of the protease that cleaves the 16 kDa Ndk
P. aeruginosa 8821 was grown in batches of four 1 litre Luria broth with vigorous aeration at 37°C for 14–15 h. The cells were harvested by centrifugation at 4500 × g for 20 min in a refrigerated centrifuge. The cell pellet was washed with 0.9% NaCl and was resuspended in 3 vols of buffer A (50 mM Tris-HCl, pH 7.5; 10 mM MgCl 2 ; 1 mM dithiothreitol). The cells were lysed by sonication through 15 pulses of 15 s duration with a gap of 15 s between pulses using a sonicator (Branso Sonifier Model 450). The sonicated suspension was centrifuged at 10 000 × g for 20 min and the supernatant was transferred to a fresh tube. Ammonium sulphate was added to a final saturation of 45%, and the suspension was stirred on ice for 1 h. The pellet was collected by centrifugation and resuspended in 3 vols (w/v) of buffer A and dialysed for 10 h at 4°C against the same buffer. Ammonium sulphate was added to a final concentration of 1.0 M, and the suspension was loaded onto a phenyl–Sepharose hydrophobic interaction column at a flow rate of 1.0 ml min −1 . Samples were eluted on a 1.0–0 M ammonium sulphate gradient, and 2 ml fractions were collected. The fractions were assayed for Ndk cleavage activity as follows ( Kavanaugh-Black et al., 1994 ). Typically, in a 20 μl assay volume, 50 ng of purified 16 kDa Ndk was incubated with 5 μl of the column eluate for 1 h at 37°C in the presence of 1 μCi of [γ- 32 P]-ATP. The reaction was terminated by the addition of 5 μl of 5× stop buffer and was analysed by electrophoresis on a 15% SDS–PAGE gel. The dried gel was subsequently autoradiographed to look for autophosphorylated 16 kDa and/or 12 kDa Ndk. Elastase from the supernatant of P. aeruginosa PAO1 was purified by 80% ammonium sulphate precipitation. This was followed by the procedure described above.
Isolation of the periplasmic fraction
Mucoid P. aeruginosa 8821 cells were harvested from 14-h-old cultures by centrifugation at 5000 × g for 10 min and washed with buffer I (10 mM Tris-HCl, 30 mM MgCl2; pH 7.3) before further manipulation. To isolate the periplasmic fraction, the washed bacteria from 25 ml of culture solution were suspended in 1 ml of buffer I, and 15 μl of chloroform was added (Klotz and Hutcheson, 1992). After 15 min at 4°C, an additional 1 ml of ice-cold buffer I was added, and the solution was clarified by centrifugation at 10 000 × g for 10 min at 4°C. The resulting supernatant was used as the periplasmic fraction.
Elastase activity assays
Standard cultures of P. aeruginosa strains were used for elastase assay (Table 2). Elastase activity was quantified by a modification of a procedure described earlier (Ohman et al., 1980). To a 15 ml screw cap glass tube, 20 mg of elastin–Congo red (ECR) (Sigma) and 2 ml of reaction buffer (30 mM Tris-HCl; pH 7.5) were added. After equilibration to reaction at 37°C, 1 ml of culture supernatant or periplasmic fraction was added. The tubes were capped and incubated at 37°C with rapid shaking for 2 h. Elastin–Congo red was pelleted by centrifugation at 1400 × g for 10 min, and the optical density of the supernatant was measured at 495 nm. The culture supernatant was replaced with reaction buffer in the blank. The background absorbance at 495 nm was obtained by mixing culture supernatant or periplasmic fractions with the assay buffer in proportions described above without the substrate. The OD was determined using a Shimadzu Corporation spectrophotometer Model Biospec 1601.
The protease activity was detected by using a gelatin SDS–PAGE in-gel assay (Lee et al., 1995). Novex zymogram containing 10% Tris–glycine with 0.1% gelatin incorporated throughout the gel was used. Briefly, the protease samples were denatured in sample buffer containing SDS in the absence of reducing agent and were loaded onto the Novex zymogram precast gel. After electrophoresis, the gel was soaked in Novex renaturation buffer, followed by soaking in Novex developing buffer. The gel was fixed and stained in 0.5% solution of Coomassie brilliant blue and destained. Protease activity was detected as clear zones against a dark-blue background.
Polyacrylamide gel electrophoresis and immunoblotting
Approximately 0.5 μg of purified protein was loaded on a 15% SDS–PAGE according to the method of Laemmli (1970) and was transferred onto nitrocellulose membranes. The transfer was performed in a buffer containing Tris-glycine–methanol [25 mM Tris-HCl; pH 8.3, 192 mM glycine, 20% (v/v) methanol] at 0.3 A for 2 h. The nitrocellulose membrane was first treated with TBST (10 mM Tris-HCl; pH 8.0, 50 mM NaCl, 0.05% Tween 20) containing 5% skim milk at room temperature for 1 h. The membrane was incubated with polyclonal rabbit anti-elastase antibody (1: 1500 dilution) or with rabbit anti-Ndk antibody (1: 7000 dilution) in TBST at room temperature for 1 h. The blot was washed 3× with TBST for 10 min and was incubated with anti-rabbit IgG coupled to alkaline phosphatase (Sigma, according to manufacturer's suggestion) at a dilution of 1:7500 for 1 h. The blot was again washed three times with TBST and was developed in a solution containing NBT/BCIP (Sigma).
Construction of a lasB mutant in P. aeruginosa
The plasmid pRB1804 containing a 2.7 kb EcoRI–HindIII fragment harbouring the elastase gene (lasB ) was kindly provided by B. Iglewski (Bever and Iglewski, 1988). pRB1804 was transformed into E. coli DH5a, which was used as the recipient strain. A chloramphenicol cassette was introduced into the lasB gene by transposon mutagenesis using a mini-Tn5 that carried a chloramphenicol cassette. This transposon was maintained in E. coli S17-1λpir (Lorenzo and Timmis, 1994). The plasmid carrying the inactivated lasB (pSAK 6) was used to electroporate P. aeruginosa strain 8821 using the IBI electroporator and was selected for chloramphenicol resistance and carbenicillin sensitivity. The lasB ::Cm strain was confirmed for double cross-over event on the chromosome by Southern hybridizations (data not shown). This strain was also tested for elastase activity by ECR assay as described earlier.
Complementation of P. aeruginosa 8821 lasB::Cm mutant
The lasB gene was subcloned as a 2.7 kb HindIII–EcoRI fragment from pRB1804 into the compatible sites in pMMB67HE and was designated as pSAK5. The positive clones were verified by Southern hybridization. pSAK5 was introduced into P. aeruginosa strains by triparental matings (Darzins and Chakrabarty, 1984). The complemented P. aeruginosa strains were maintained on PIA medium containing carbenicilin.
Alginate isolation and quantification
The P. aeruginosa strains were grown overnight in LB with the appropriate antibiotics. An aliquot of 200 μl was spread on PIA plates. The plates were incubated at 37°C for 48 h, after which 0.9% NaCl solution was added to the plate. The bacteria were scraped from the plate surface using a sterile glass rod. After a brief vortexing, the solution was centrifuged at 18 000 rpm for 30 min and the supernatant was collected. After an additional centrifugation, 2 vols of 100% ethanol was added and the sample was stored at −70°C for 30 min The precipitated alginate was recovered by centrifugation at 18 000 rpm for 10 min. After three washing steps with 100% ethanol, the pellet was dried in vacuo. The dried pellet was suspended in sterile deionized water and analysed. The levels of uronic acids in the samples were determined by the colorimetric method (May and Chakrabarty, 1994b). Kelp alginate (Sigma) was used as standard.
Amino acid sequence determination
The N-terminal amino acid sequence of both the protease and the 12 kDa Ndk was performed by Dr Bob Lee at the Protein Research Laboratory, University of Illinois at Chicago, Chicago, IL, USA.
We thank Dr B. Iglewski for providing us with the plasmid pRB1804 and Dr D. R. Galloway for providing anti-elastase antibody. We thank Olga Zaborina for her help with the protein analysis. This work was supported by Public Health Service grant AI-16790-17 from the National Institutes of Health.