INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The exopolysaccharide alginate is a copolymer of O-acetylated -1,4-linked D-mannuronic acid and its C-5 epimer, L-guluronic acid (46). Alginate biosynthesis has been extensively studied in Pseudomonas aeruginosa, where it functions as a major virulence factor in strains infecting the lungs of cystic fibrosis patients (45). In P. aeruginosa, genes that encode the biosynthesis and regulation of alginate map to four chromosomal locations. With the exception of algC, which is located at 10 min, the structural genes are clustered within an 18-kb region located at 34 min (18, 46). Structural genes that have been characterized in this region include algA, which encodes a bifunctional enzyme which functions as a phosphomannose isomerase and a GDP-mannose pyrophosphorylase (54); algG, which encodes a C-5 epimerase (7); algF, algI, and algJ, which are involved in acetylation of the alginate polymer (16, 17, 55); and algD, which encodes GDP-mannose dehydrogenase (11). This region also contains algE and algK, which encode proteins with putative roles in polymer export and synthesis, respectively (1, 9, 22), and algL, which encodes alginate lyase (6, 49). Other genes which map within this region include alg44, alg8, and algX (alg60) (33, 41, 60); however, the functional role of the proteins encoded by these genes remains unclear. Chitnis and Ohman (8) postulated that the alginate biosynthetic gene cluster in P. aeruginosa is organized as an operon with transcription initiating at the algD promoter.

A region mapping at 68 min on the P. aeruginosa chromosome harbors a gene cluster consisting of algT (algU), mucA, mucB (algN), mucC, and mucD. These genes modulate the conversion to constitutive alginate production; at the head of this regulatory hierarchy is algT (algU). The alternative sigma factor encoded by algT, ²², is required for transcription of algD, algT, and algR1 (21, 51). mucA is a negative regulator of algT transcription and encodes an antisigma factor with affinity for ²² (52, 62). Mutations in mucA inactivate the MucA protein and result in the Alg⁺ phenotype; however, these mutations are unstable and spontaneous reversion to the Alg phenotype often occurs due to suppressor mutations in algT (14, 50, 52). The remaining muc genes also modulate the expression of algT and have been described elsewhere (19, 34, 52, 62).

Other genes controlling the regulation of alginate production include algR1 (algR), algR2 (algQ), algR3 (algP), and algB (20, 53). AlgR1 functions as a response regulator member of the two-component signal transduction system and binds to multiple sites upstream of algC and algD (12, 24, 39, 65). Both the algD and algR1 promoters show a consensus sequence at the 35/10 region which is consistent with recognition by ²², suggesting that an RNA polymerase-²² complex binds to both promoters and positively regulates transcription (51).

Like P. aeruginosa, phytopathogenic strains of P. syringae are normally nonmucoid in vitro. Kidambi et al. (28) previously showed that exposure to copper ions stimulated alginate production in selected strains of P. syringae. Furthermore, an indigenous plasmid designated pPSR12 conferred constitutive alginate production to P. syringae pv. syringae FF5. pPSR12 does not contain homologs of the biosynthetic or regulatory genes which control alginate production in P. aeruginosa; instead this plasmid presumably contains regulatory genes which remain uncharacterized (28). Mutagenesis of FF5(pPSR12) with Tn5 resulted in the isolation of alginate-defective (Alg) mutants, including FF5.31 and FF5.32 (28). The Tn5 insertion in FF5.31 was located in algL, which encodes alginate lyase. Alginate production in FF5.31 was restored by pSK2, a cosmid clone containing homologues of algD, alg8, alg44, algG, algX, algL, algF, and algA. The order and arrangement of the alginate structural gene cluster were virtually identical to those previously described for P. aeruginosa. Complementation analyses, however, indicated that the structural gene clusters in P. aeruginosa and P. syringae were not functionally interchangeable when expressed from their native promoters (44).

In the present study, the Alg mutant FF5.32 was shown to contain a Tn5 insertion in algR1. Unlike P. aeruginosa, expression from the P. syringae algD promoter (PsalgD) did not require a functional copy of algR1. Nucleotide sequence analysis indicated that PsalgD did not contain recognizable AlgR1 binding sites, which helps explain the differential regulation of alginate gene expression in P. aeruginosa and P. syringae.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains, plasmids, and media. Table 1 lists the bacterial strains and plasmids used in the present study. Pseudomonas spp. were routinely maintained at 28°C on King's medium B (29), mannitol-glutamate (MG) medium (25), or MG medium supplemented with yeast extract at 0.25 g/liter (MGY); Escherichia coli strains were grown on Luria-Bertani (LB) medium (36) at 37°C. Antibiotics were added to the media at the following concentrations: ampicillin, 100 µg/ml; tetracycline, 25 µg/ml; kanamycin, 25 µg/ml; spectinomycin, 25 µg/ml; streptomycin, 25 µg/ml; piperacillin, 250 µg/ml; and chloramphenicol, 25 µg/ml.

Molecular genetic techniques. Plasmid DNA was isolated from Pseudomonas spp. by alkali lysis (48). Restriction enzyme digests, agarose gel electrophoresis, Southern transfers, and isolation of DNA fragments from agarose gels were performed by standard methods (48). Genomic DNA was isolated from P. syringae by established procedures (56), and a total genomic library of FF5.32 was constructed in pRK7813 as described previously (2). Clones were mobilized into nonmucoid recipient strains by using a triparental mating procedure and the mobilizer plasmid pRK2013 (4).

Isolation and quantitation of alginate. Selected strains were inoculated by dilution streaking to MGY agar (three plates per strain) and incubated at 28°C for 72 h. Each plate was handled separately for quantification of alginate. The cells were washed from each plate and resuspended in 0.9% NaCl. Removal of cellular material from the mucoid growth and estimation of the alginate content and total cellular protein were performed as described previously (35). Alginic acid from seaweed (Macrocystis pyrifera; Sigma Chemical Co., St. Louis, Mo.) was used as a standard in these experiments. Mean values of three replicate determinations were expressed as micrograms of alginate per milligram of protein.

Construction of transcriptional fusions. PsalgD was initially cloned in pCR2.1 as a 2.7-kb PCR product. Plasmid pSK2 was used as template, and the following oligonucleotides were used as primers: forward primer, 5' TGGTGCTGGAAATATCCACACC (located 100 bp downstream of the presumed translational start site of algD [P1 in Fig. 1A]); and reverse primer, 5' AATTCTGCCAGTCCAGCCACTGAC (P2 in Fig. 1A). Following amplification of the 2.7-kb PCR product, ligation in pCR2.1, and transformation into E. coli DH5, plasmid pAPD was recovered. The promoter probe construct, pBBR.Gus, which contains a promoterless glucuronidase gene (uidA) downstream of the polylinker in pBBR1MCS (43), was used to more precisely define the promoter region upstream of algD. pAPD was digested with HindIII and EcoRV, and the 2.7-kb insert was isolated, end-filled with Klenow, and ligated into pBBR.Gus. Transformants were selected on LB agar containing 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (X-Gal) and chloramphenicol, and pAPDP was found to contain the algD::uidA fusion in the transcriptionally active orientation.

View larger version (18K): [in this window] [in a new window]   FIG. 1.   (A) Physical and functional map of the alginate structural gene cluster in Pseudomonas syringae pv. syringae FF5. The arrows within each open reading frame indicate the direction of translation. The locations of the primers (P1 and P2) used to amplify the algD promoter region are indicated. Abbreviations: F, algF; 44, alg44. (B) Expanded view of the region amplified with primers P1 and P2. The location and orientation of the coding region for algD are shown (horizontal arrow). The black boxes flanking the EcoRI site indicate the consensus sequence recognized by AlgT (22). The location and orientation of the algD::uidA transcriptional fusions are indicated; GUS activity is shown in the column adjacent to each construct. Values followed by the same letter were not significantly different (P = 0.01). Abbreviations: E, EcoRI; H, HindIII; V, EcoRV.

GUS assays. Transcriptional activity was initially screened by spotting bacterial suspensions (absorbance at 600 nm of 0.1) on MG agar medium amended with spectinomycin and 20 µg of X-Gluc (5-bromo-4-chloro-3-indolylglucuronide) per ml; the plates were then incubated at 28°C for 24 to 72 h. Glucuronidase (GUS) activity was quantified by fluorometric analysis of cells grown for 18 to 20 h in 3 ml of MG medium. Fluorescence was monitored with a Fluoroscan II version 4.0 microplate reader (ICN Biomedicals, Inc., Costa Mesa, Calif.) in 96-well microtiter plates. GUS activity was expressed in units per milligram of protein, with 1 U being equivalent to 1 nmol of methylumbelliferone formed per min. Values presented for GUS activity represent the average of three replicates per experiment. When significant differences in GUS activity were detected, the experiment was repeated.

DNA sequencing and analysis. Nucleotide sequencing reactions were performed by the dideoxynucleotide method with AmpliTaq DNA polymerase (Perkin-Elmer, Foster City, Calif.). Automated DNA sequencing was performed with an ABI 373A apparatus and the ABI PRISM Dye Primer cycle-sequencing kit (Perkin-Elmer). Automated sequencing was provided by the Oklahoma State University Recombinant DNA/Protein Resource Facility. The Tn5 insertion in FF5.32 was localized by sequencing the DNA flanking the transposon by using the oligonucleotide 5' GGTTCCGTTCAGGACGCTAC, which is derived from the border region of IS50. Sequence data were aligned and homology searches were executed by using the University of Wisconsin Genetics Computer Group (UWGCG) sequence analysis package, version 9.0. Sequences associated with ²² and AlgR1 binding were located by using the MOTIFS program included with the UWGCG software.

Nucleotide sequence accession numbers. The nucleotide sequences described in this study were deposited in GenBank under accession no. AF131199 (fimS-algR1-hemC) and AF131068 (PsalgD).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Location of Tn5 insertion in FF5.32. A genomic library of FF5.32 was constructed in pRK7813, and a clone containing the Tn5 insertion from FF5.32 was recovered and designated pAP32. The internal BamHI site in Tn5 and 2.5 kb of FF5.32 DNA were cloned from pAP32 into pBluescript SK(+), resulting in a clone named pAP32.1 (Fig. 2). A primer specific for the border region of IS50 was used to sequence approximately 300 bp of FF5.32 DNA flanking the Tn5 insertion site. This sequence showed 76% nucleotide identity to algR1 from P. aeruginosa, and the Tn5 insertion was located at nucleotide 51 of algR1 from P. aeruginosa (12).

View larger version (25K): [in this window] [in a new window]   FIG. 2.   Constructs used for the cloning and sequencing of algR1 from P. syringae pv. syringae FF5. pAP32.1 is a subclone containing Tn5 (shaded region) and flanking DNA from P. syringae pv. syringae FF5.32 (hatched region). The HindIII-EcoRI fragment in pAP32.1 was used as a probe for algR1 in the current study. pMF6.1 and pMF6.2 are subclones derived from pMF6, a cosmid which complemented FF5.32 for alginate production. The 2.0-kb PstI fragment in pMF6.2 was sequenced on both strands and shown to contain an intact copy of algR1. Abbreviations: B, BamHI; E, EcoRI; H, HindIII; P, PstI.

Cloning of algR1 from P. syringae. A genomic library of P. syringae pv. syringae FF5(pPSR12) was previously constructed in pRK7813 (44). In the current study, the 2.3-kb HindIII-EcoRI fragment from pAP32.1 (Fig. 2) was used to screen the library for clones containing the complete algR1 coding region. Seven cosmid clones hybridized with the probe; two clones designated pMF4 and pMF6 were chosen for further study and contained a 2.7-kb EcoRI fragment which hybridized with the probe. This fragment was subcloned from pMF6 in pBluescript SK(+), resulting in pMF6.1 (Fig. 2). Sequence information for pMF6.1 was generated with the T7 and T3 primers and indicated that this fragment contained DNA homologous to argH, fimS, and algR1. In previous studies, the fimS gene showed relatedness to sensor kinases of two-component systems and mapped immediately upstream of algR1 in P. aeruginosa (61). It is important to note that fimS, which was also named algZ (63), is distinct from the algZ described by Baynham and Wozniak (3). To avoid further confusion in nomenclature, the name "fimS" will be used hereafter to describe the sensor kinase which maps adjacent to algR1. In P. syringae, argH, which encodes arginosuccinate lyase, mapped adjacent to fimS; in P. aeruginosa, argH was divergently transcribed with respect to both fimS and algR1 (37, 63). Sequence analysis of pMF6.1 indicated that this arrangement is conserved in P. syringae (Fig. 2).

Complementation experiments. pMF4 and pMF6, the cosmid clones containing argH, fimS, algR1, and hemC, were evaluated for their ability to complement P. syringae pv. syringae FF5.32 for alginate production. Transconjugants of FF5.32 containing pMF4 or pMF6 were visibly mucoid and produced significantly more alginate than the mutant FF5.32 did (Table 2). Since Tn5 frequently causes polar mutations on downstream genes, the 2.0-kb PstI fragment in pMF6.2 was used to investigate whether the Alg phenotype in FF5.32 was caused by the mutation in algR1. pMF6.2 contains an intact copy of algR1 with the cognate ²² recognition site and truncated copies of fimS and hemC (Fig. 2). The 2.0-kb PstI fragment in pMF6.2 was subcloned in pRK415 to form pMF6.21 and pMF6.22, which contain algR1 in the transcriptionally active and inactive orientations with respect to the lac promoter (Table 1). Both pMF6.21 and pMF6.22 restored alginate production to FF5.32 (Table 2), indicating that the Alg phenotype of FF5.32 was caused by the Tn5 insertion in algR1. FF5.32 was complemented with both clones irrespective of the orientation of the lac promoter and without the addition of isopropyl--D-thiogalactopyranoside (IPTG), indicating that a functional promoter for algR1 was present on the 2.0-kb PstI fragment. To further confirm that FF5.32 was indeed an algR1 mutant, we investigated whether this mutant could be complemented by algR1 from P. aeruginosa. Plasmid pAD1039, which contains algR1 from P. aeruginosa (Table 1), complemented FF5.32 and restored alginate production in the mutant to a level equivalent to FF5(pPSR12) (data not shown).

Expression of the PsalgD promoter does not require AlgR1. In P. aeruginosa, AlgR1 is required for expression of the algD promoter (PalgD) and has been shown to bind PalgD at several conserved sites (24, 39). A portion of PsalgD was previously cloned as a 1-kb fragment in the promoter probe vector, pRG960sd, creating pSK3 (PsalgD::uidA; transcriptionally active orientation) and pSK4 (uidA::PsalgD; transcriptionally inactive) (44). In the present study, we investigated whether PsalgD was transcriptionally active in FF5.32, the algR1 mutant. GUS activities in FF5(pPSR12) and FF5.32(pSK3) were not significantly different (Table 3), indicating that a functional copy of algR1 was not required for transcription of algD in P. syringae.

Analysis of the PsalgD promoter. To more fully characterize the minimum sequence necessary for algD expression in P. syringae, we constructed a series of deletions from the 5' (EcoRV) end of the PsalgD promoter (Fig. 1B). A new construct, pAPDP (Fig. 1B), was designed for this purpose since the pBBR.Gus polylinker was more amenable to deletion analysis than was the multicloning site in pRG960sd, the vector used for construction of pSK3. Two deletion derivatives of pAPDP, pAPDP15 and pAPDP23, proved useful for delineating the algD promoter region; sequence analysis indicated that these two constructs lacked 1.5 and 2.3 kb of DNA downstream of the EcoRV site, respectively. FF5(pPSR12, pAPDP15) (Fig. 1B) retained the full level of GUS activity exhibited by FF5(pPSR12, pAPDP) (Fig. 1B), suggesting that the 1.5-kb region downstream of the EcoRV site was dispensable for promoter activity. However, GUS activity in FF5(pPSR12, pAPDP23) was 3.8-fold lower than in FF5(pPSR12, pAPDP15), demonstrating that deletion of an additional 0.8 kb from the 5' end of pAPDP15 virtually eliminated PsalgD promoter activity (Fig. 1B).

View larger version (80K): [in this window] [in a new window]   FIG. 3.   Alignment of the algD promoter sequences from P. syringae pv. syringae FF5 (Ps algD) and P. aeruginosa (Pa algD). The P. aeruginosa algD promoter was previously reported (24, 39); the nucleotides for this sequence are shown on the left, with +1 (asterisk) corresponding to the transcriptional start site. Nucleotides for the P. syringae pv. syringae algD promoter are shown on the right. The EcoRI site in the P. syringae sequence corresponds to the left border of EcoRI fragment 5 in Fig. 1A. Gaps (--) were used to maximize the alignment, and identical bases are shaded. The AlgR1 binding sites (ABS) in the P. aeruginosa algD promoter are shown in bold and double-underlined. The 22 recognition sequence in both species is indicated in bold and single-underlined. The algD translational start site and coding region are shown in bold (algD-).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The AlgR1 mutant characterized in the present study, FF5.32, was previously shown to be completely defective in alginate synthesis (28), thereby demonstrating that AlgR1 is absolutely required for alginate production in P. syringae. However, the role of AlgR1 in P. syringae is unclear, since this protein is not required for algD expression; it remains possible that AlgR1 is required for transcriptional activation of algC in P. syringae, which is true in P. aeruginosa (65). Alternatively, AlgR1 may function differently in P. syringae, perhaps as part of a signal transduction cascade which controls alginate production. A complex regulatory network for alginate synthesis in P. syringae seems plausible, since plasmid-encoded regulatory genes are known to mediate the constitutive production of alginate in the P. syringae strains which harbor them (28).

The organization of the region flanking AlgR1 is conserved in both P. aeruginosa and P. syringae (argH-fimS-algR1-hemC). In both species, the ²² recognition site preceding algR1 is located within the 3' end of fimS (63). FimS shows relatedness to the histidine protein kinases which function as environmental sensors, and both AlgR1 and FimS are required for twitching motility in P. aeruginosa, a process mediated by type IV pili. Although type IV pili have been identified in P. syringae (47), our efforts to demonstrate twitching motility in P. syringae pv. syringae FF5 were completely unsuccessful; therefore, the involvement of AlgR1 in twitching motility in P. syringae remains unclear. It has also been proposed that FimS may function as the cognate sensor kinase for AlgR1, but the exact role of FimS in alginate production remains unclear (61, 63). Interestingly, phosphorylation of AlgR1 was not required for alginate production in P. aeruginosa (32).

Sequence analysis of the algR1 and algD upstream regions in P. syringae revealed the presence of ²² recognition sites (Fig. 3). The ²² recognition site identified in the algR1 upstream region was identical to that identified in P. aeruginosa, whereas the ²² recognition sequence in PsalgD differed from the corresponding sequence in P. aeruginosa by a single nucleotide (51). Although the transcriptional start sites for algR1 and algD were not identified in P. syringae, the positions of the ²² recognition sites relative to the translational start site are conserved in both species. The conservation of ²² recognition sequences upstream of algR1 and algD strongly suggests that transcriptional activation of these genes requires a functional copy of algT. An algT homologue in P. syringae has recently been identified, and the role of algT in the transcriptional activation of algD and algR1 in P. syringae is under investigation (27).

The percent nucleotide identity in the algD coding region of P. syringae pv. syringae and P. aeruginosa ranged from 80 to 90% (Fig. 3 and data not shown); however, upstream of the translational start site, the relatedness between the two species diverged and nucleotide identity decreased to approximately 20% (Fig. 3). This divergence is consistent with the absence of specific sequences in PsalgD which are known to be involved in transcriptional activation of algD in P. aeruginosa. These include the consensus sequences for binding AlgR1 (24), integration host factor (38), and cyclic AMP receptor protein (13). Although some signals for activation of the algD promoter are conserved in P. aeruginosa and P. syringae (5, 31, 44), the algD promoter in P. syringae is stimulated by exposure to copper ions (44) and does not require a functional copy of AlgR1 for transcriptional activation. Recently, Yu et al. (64) provided the first genetic evidence for the role of alginate in the virulence and epiphytic fitness of P. syringae. Consequently, the differential regulation of algD expression in P. syringae and CF isolates of P. aeruginosa and the marked divergence in their algD promoter regions probably reflect their adaptation to plant and human hosts, respectively.

It remains possible that some unknown regulatory protein binds to PsalgD and that this regulator recognizes different signals (such as copper) and activates the algD promoter in P. syringae. Perhaps this putative DNA binding protein was recruited during the evolutionary divergence of P. aeruginosa and P. syringae to accommodate a different signal and perhaps another activator. The algD::uidA transcriptional fusion described in the present investigation could be used to screen for mutants lacking the unknown activator. Such experiments are under way and will probably reveal additional differences in the regulation of alginate biosynthesis in human and phytopathogenic bacteria.