The Paralogous Pairs of Genes Involved in Clavulanic Acid and Clavam Metabolite Biosynthesis Are Differently Regulated in Streptomyces clavuligerus

Abstract

Carboxyethylarginine synthase, encoded by the paralogous ceaS1 and ceaS2 genes, catalyzes the first reaction in the shared biosynthetic pathway leading to clavulanic acid and the other clavam metabolites in Streptomyces clavuligerus. The nutritional regulation of ceaS1 and ceaS2 expression was analyzed by reverse transcriptase PCR and by the use of the enhanced green fluorescent protein-encoding gene (egfp) as a reporter. ceaS1 was transcribed in complex soy medium only, whereas ceaS2 was transcribed in both soy and defined starch-asparagine (SA) media. The transcriptional start points of the two genes were also mapped to a C residue 98 bp upstream of ceaS1 and a G residue 51 bp upstream of the ceaS2 start codon by S1 nuclease protection and primer extension analyses. Furthermore, transcriptional mapping of the genes encoding the beta-lactam synthetase (bls1) and proclavaminate amidinohydrolase (pah1) isoenzymes from the paralogue gene cluster indicated that a single polycistronic transcript of ∼4.9 kb includes ceaS1, bls1, and pah1. The expression of ceaS1 and ceaS2 in a mutant strain defective in the regulatory protein CcaR was also examined. ceaS1 transcription was not affected in the ccaR mutant, whereas that of ceaS2 was greatly reduced compared to the wild-type strain. Overall, our results suggest that different mechanisms are involved in regulating the expression of ceaS1 and ceaS2, and presumably also of other paralogous genes that encode proteins involved in the early stages of clavulanic acid and clavam metabolite biosynthesis.

Members of the genus Streptomyces exhibit a complex life cycle that involves a hierarchy of regulatory genes controlling and coordinating antibiotic production and sporulation (14, 20). Streptomyces clavuligerus produces a number of β-lactam compounds, including cephamycin C, clavulanic acid, and at least four other known clavam metabolites (8, 12). Clavulanic acid and the other clavams differ from cephamycin C in that their bicyclic nucleus contains an oxygen atom instead of the sulfur atom found in the more conventional cephamycin-type antibiotics (6). Clavulanic acid is a clinically important inhibitor of β-lactamases, whereas the other clavam metabolites produced by S. clavuligerus show weak antibacterial and antifungal activities (35). The four clavam metabolites produced by S. clavuligerus are commonly referred to as the 5S clavams due to their stereochemistry, which differs from the 5R stereochemistry of clavulanic acid. The β-lactamase-inhibitory activity of clavulanic acid has been associated with this 5R stereochemistry (6). Since clavulanic acid is produced industrially by fermentation using S. clavuligerus, the regulation of clavulanic acid and 5S clavam biosynthesis is a point of great interest.

The biosynthetic pathway leading to clavulanic acid and the 5S clavams is partially shared, at least to the level of clavaminic acid (Fig. 1B) (10). However, in S. clavuligerus, the genes involved in the biosynthesis of clavulanic acid and the 5S clavams reside in three distinct gene clusters that are not physically linked (K. Tahlan, H. U. Park, and S. E. Jensen, unpublished data). The clavulanic acid gene cluster is situated immediately downstream of the cephamycin gene cluster, and together they form a larger gene cluster often referred to as the β-lactam supercluster (1, 13, 44). This clavulanic acid gene cluster contains genes encoding enzymes involved in the early shared stages of the clavulanic acid and 5S clavam pathway (the early genes), as well as genes encoding proteins involved in the later stages of clavulanic acid biosynthesis only (the late genes) (15, 17, 21, 24). The clavam gene cluster encompasses the gene encoding clavaminate synthase 1 (cas1), a paralogue of the cas2 gene from the clavulanic acid gene cluster and one of the genes involved in the biosynthesis of both clavulanic acid and the 5S clavams. In addition to cas1, the clavam cluster includes other genes involved exclusively in the biosynthesis of the 5S clavams and some genes of unknown function (26). The paralogue gene cluster contains paralogues of additional genes found in the clavulanic acid gene cluster that encode enzymes involved in the early stages of both clavulanic acid and 5S clavam biosynthesis (18, 42).

FIG. 1.

Early steps of clavulanic acid and 5S clavam metabolite biosynthesis in S. clavuligerus. (A) Genes involved in the early steps of clavulanic acid and 5S clavam biosynthesis from the clavam, clavulanic acid, and paralogue gene clusters. Genes flanking cas1 in the clavam cluster are shown only to provide context; they do not encode early enzymes. (B) Diagrammatic representation of the early steps of clavulanic acid and 5S clavam metabolite biosynthesis showing the enzymes and genes involved. The 5S clavams are shown as a family of metabolites; R represents the site of side chain modification giving rise to the different members.

The biosynthesis of both clavulanic acid and the 5S clavams initiates with the condensation of glyceraldehyde-3-phosphate and l-arginine to give N²-(2-carboxyethyl)arginine, catalyzed by the enzyme carboxyethylarginine synthase (CEAS) (Fig. 1B) (19). There are two copies of the gene encoding CEAS present in S. clavuligerus; ceaS2 is located in the clavulanic acid gene cluster, and ceaS1 is found in the paralogue gene cluster (Fig. 1A) (15, 33, 42). The next reaction in the pathway is catalyzed by the enzyme β-lactam synthetase (BLS) (3, 23), which is also encoded by two separate genes (bls1 and bls2) (15, 42). BLS activity closes the β-lactam ring to give deoxyguanidino proclavaminic acid (3, 23), which is then converted to guanidino proclavaminic acid by clavaminate synthase (CAS) (5). Two genes encode CAS in S. clavuligerus (22), cas1 from the clavam gene cluster and cas2 from the clavulanic gene cluster (22, 26). Guanidino proclavaminic acid is hydrolyzed to give proclavaminic acid by the action of proclavaminate amidinohydrolase (PAH) (46), which is encoded by pah1 and pah2 (1, 15, 18, 46). Next, in a series of two sequential reactions, CAS converts proclavaminic acid to clavaminic acid (4, 36). Clavaminic acid is thought to be the branch point of the pathway leading to clavulanic acid and the 5S clavams (10). The only other step known in the pathway beyond clavaminic acid is the reduction of clavaldehyde to clavulanic acid by the action of the enzyme clavulanic acid dehydrogenase (27), which is encoded by cad, a late gene from the clavulanic acid gene cluster (15). It is still not known how clavaminic acid is converted to clavaldehyde, nor have any of the steps specifically leading from clavaminic acid to the 5S clavams been elucidated (16).

The transcriptional activators CcaR and ClaR are known to regulate the expression of the clavulanic acid biosynthetic genes (16). The ccaR gene lies within the cephamycin biosynthetic gene cluster, and CcaR is the pathway-specific transcriptional regulator for cephamycin biosynthesis, as well as controlling expression of the claR gene from the clavulanic acid gene cluster (2, 31, 32). ClaR is the pathway-specific transcriptional regulator for clavulanic acid biosynthetic genes, but it affects only expression of the late genes for clavulanic acid biosynthesis (29). The early genes responsible for the steps shared with the 5S clavams are not regulated by ClaR (29). In this manner, CcaR coordinates the biosynthesis of cephamycin C, a β-lactam antibiotic, with clavulanic acid, a β-lactamase inhibitor (16).

The cas1 and cas2 paralogues can functionally replace each other and are known to be nutritionally regulated. The cas1 paralogue is expressed only when S. clavuligerus is grown on complex soy medium, whereas the cas2 paralogue is expressed during growth on both complex soy and defined SA media (30). Thus, a cas1 mutant can still produce clavulanic acid in both soy and SA media, because cas2 is expressed in both media. In contrast, a cas2 mutant produces clavulanic acid only when grown on soy medium, because cas1 is not expressed in SA medium (30). Similar phenotypes were also observed when mutants defective in each of the ceaS, bls, and pah paralogues were prepared and tested (15, 18, 42).

Since ceaS1 and ceaS2 encode proteins catalyzing the first reaction in the shared clavulanic acid and 5S clavam biosynthetic pathways, and since they are both located at the boundaries of their respective gene clusters (Fig. 1A), they seemed to be likely candidates for points of regulation. With this in mind, the nutritional regulation of ceaS1 and ceaS2 was examined at the transcriptional level to determine if they are regulated in a manner similar to cas1 and cas2. Previous studies have suggested that ClaR does not regulate ceaS2 transcription (29), but the effect of CcaR on ceaS1 and ceaS2 transcription is still not known. These effects were also examined and are discussed in the present study. Lastly, the transcriptional start points (TSPs) of ceaS1, bls1, pah1, and ceaS2 were also mapped.

MATERIALS AND METHODS

Bacterial strains, plasmids, media, and culture conditions.

The bacterial strains and plasmids used in this study are described in Table 1. Escherichia coli cultures were grown as described earlier (38), and cultures containing plasmids were supplemented with ampicillin (100 μg/ml), apramycin (50 μg/ml), chloramphenicol (25 μg/ml), kanamycin (50 μg/ml), or spectinomycin (100 μg/ml), as appropriate. S. clavuligerus cultures were maintained on either MYM (40) or ISP 4 medium, as described previously (42), and cultures containing plasmids were supplemented with apramycin (25 μg/ml) or thiostrepton (5 μg/ml). S. clavuligerus cultures for the isolation of chromosomal DNA were grown in trypticase soy broth supplemented with 1% starch (TSBS), and cultures for the isolation of exconjugants were grown in AS-1 medium supplemented with 10 mM MgCl₂, as described earlier (42). To prepare RNA from S. clavuligerus, spores were pregerminated for 4 h at 28°C in 2YT medium (38) and then used to inoculate SA or soy culture medium (30). RNA was isolated after 96 and 120 h of growth in soy medium and after 72 and 96 h of growth in SA medium. All Streptomyces liquid cultures were grown at 28°C on a rotary shaker at 250 rpm.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain, plasmid, or cosmid	Descriptiona	Reference or source
Bacillus sp. strain ATCC 27860	Indicator strain for alanyl clavam bioassay	34
K. pneumoniae ATCC 15380	Indicator strain for clavulanic acid bioassay	35
S. clavuligerus NRRL3585	Wild type	National Regional Research Laboratory (NRRL)
S. clavuligerus ΔccaR::tsrA	ccaR deletion mutant; ccaR replaced by tsr; Tsrr	2
S. clavuligerus C1G	Contains the egfp reporter gene fused to the ceaS1 promoter; integrated at the φC31 attB site in the chromosome	This study
S. clavuligerus C2G	Contains the egfp reporter gene fused to the ceaS2 promoter; integrated at the φC31 attB site in the chromosome	This study
Plasmids and cosmids
K6L2	pLAFR3-based cosmid clone spanning the clavulanic acid gene cluster	1
p2.8-18	pUC18 containing the 5′ end of ceaS1 and upstream region	42
p5.7	pUC118 containing blsI, pahI, and flanking sequences	42
pBB5.3A	pUC119 containing the 5.3-kb BamHI/BglII fragment encompassing pcbR and ceaS2 from pTZ40; Ampr	Jensen et al., unpublished
pCR2.1TOPO	E. coli phagemid cloning vector for rapid cloning of PCR fragments; Ampr Kanr	Invitrogen
pIJ8660	Promoterless EGFP reporter plasmid; free living in E. coli and integrates into the Streptomyces chromosome; Aprr	41
pIJ8660-ceaS2	EGFP reporter construct; pIJ8660 containing the 707-bp BamHI/KpnI fragment from pTOPO-ceaS2-8 encompassing the ceaS2 promoter region fused to the egfp gene	This study
pSET152	E. coli cloning vector; integrates into the Streptomyces chromosome; Aprr	NRRL
pSET-ceaS1	EGFP reporter construct; pSET152 containing the 2.37-kb BamHI/EcoRI fragment from pTO6-ceaS1 encompassing the ceaS1 promoter fused to the egfp gene	This study
pTO6	pIJ8660 with aac(3)IV replaced by aadA; Spcr	28
pTO6-ceaS1	EGFP reporter construct; pTO6 containing the 777-bp BamHI/KpnI fragment from pTOPO-ceaS1-4	This study
pTOPO-cea52-8	pCR2.1TOPO containing a 721-bp PCR fragment encompassing the ceaS2 promoter region	This study
pTOPO-ceaS1-4	pCR2.1TOPO containing a 784-bp PCR fragment encompassing the ceaS1 promoter region	This study
pTZ40	pTZ18R containing an ∼26-kb BamHI/PstI fragment from cosmid K6L2 spanning the clavulanic acid gene cluster; Ampr	Jensen et al., unpublished

^aAmp^r, ampicillin resistance; Apr^r, apramycin resistance; Kan^r, kanamycin resistance, Spc^r, spectinomycin resistance; Tsr^r, thiostrepton resistance.

DNA isolation, manipulation, and Southern analysis.

Routine manipulation of plasmid DNA isolated from E. coli cultures, including labeling of double-stranded DNA probes with [α-³²P]dCTP by nick translation and end labeling using [γ-³²P]dATP, was performed using standard procedures (38). DNA sequencing was carried out using the DYEnamic ET Terminator Cycle Sequencing kit (Amersham Pharmacia, Baie d'Urfe, Quebec, Canada) by the Molecular Biology Service Unit, University of Alberta. Manual sequencing of DNA was performed using the Thermo Sequenase Radiolabeled Terminator Cycle Sequencing kit (U.S. Biochemical) according to the manufacturer's directions. DNA fragments fractionated by agarose and polyacrylamide gel electrophoresis (PAGE) were isolated using the QIAquick Gel Extraction kit (Qiagen Inc.) and the crush-and-soak method (38), respectively. PCRs were carried out using the Expand high-fidelity PCR system (Roche) according to the manufacturer's instructions. Plasmid DNA was introduced into S. clavuligerus by intergeneric conjugation as described previously (42). Southern analysis of S. clavuligerus DNA was also carried out as described elsewhere (38), using the following wash conditions. After overnight incubation with labeled probes, the nylon membranes were washed twice for 15 min each time with 2× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)-0.1% sodium dodecyl sulfate (SDS) (38) at room temperature, once for 20 min with 1.0× SSC-0.1% SDS, and once for 20 min with 0.1× SSC-0.1% SDS. All incubations and washes were carried out at 65°C unless otherwise indicated.

RNA isolation, RT-PCR, S1 nuclease mapping, primer extension, and Northern blot analysis.

RNAs from wild-type S. clavuligerus and S. clavuligerus ΔccaR::tsrA were isolated using the modified Kirby procedure (20). Reverse transcriptase (RT)-PCR analysis of RNA was carried out using C. therm polymerase (Roche) for reverse transcription in two-step RT-PCR. All RT reactions were carried out at 62°C for 30 min according to the manufacturer's instructions with the following changes. The reactions were set up in a final volume of 10 μl using 0.5 μg of total RNA per reaction and 15.8 U of RNAguard RNase Inhibitor (Amersham). The reverse primers ceaS1-RT-Rev, ceaS2-RT-Rev, and CAN 122 (Table 2) were used to synthesize cDNA corresponding to the ceaS1, ceaS2, and hrdB transcripts, respectively. The PCRs were performed using 5 μl of the RT product from the reaction described above in a final volume of 50 μl, employing the Expand high-fidelity PCR system with buffer 2. The primer pairs ceaS1-RT-For plus ceaS1-RT-Rev and ceaS2-RT-For plus ceaS2-RT-Rev were used for PCR amplification of the ceaS1 and the ceaS2 RT products, respectively. Dimethyl sulfoxide (DMSO; 5% [vol/vol] final concentration) was used in these reactions with the following program: 94°C for 2 min, followed by 25 cycles of 94°C for 1 min, 62°C for 1 min, and 72°C for 1 min. The primers CAN 123 and CAN 122 were used to amplify the hrdB RT product by PCR using 10% (vol/vol) DMSO and the following program: 94°C for 2 min, followed by 25 cycles of 94°C for 1 min, 66°C for 1 min, and 72°C for 1 min. The RT-PCR products were analyzed by fractionation on 1.5% agarose Tris-borate-EDTA gels. The identities of the RT-PCR products were also verified by restriction analysis, followed by PAGE and by sequencing of both DNA strands (data not shown).

TABLE 2.

Oligonucleotide primers used in this study

Primer name	Sequence (5′→3′)a	Descriptionb
CAN 122	GGCCACCGCGACCTGCTGC	Reverse primer hrdB RT-PCR
CAN 123	CGGCCAAGCGCACCACTACC	Forward primer hrdB RT-PCR
ceaS1-S1-For	TCATGAATTCCGGTGGACGGAAGGGGACGG	Forward primer for ceaS1 S1 probe
ceaS1-PR-EX	ATGGCTTTCGCGGTCGTGGT	Reverse primer for ceaS1 S1 probe and primer extension
ceaS2-S1-For	TGGATCCGTCGCGAATCCAGGGAAGCCGAGC	Forward primer for ceaS2 S1 probe
ceaS2-PR-EX	GGGCGGTCGATACACGGG	Reverse primer for ceaS2 S1 probe and primer extension
bls1-S1-For	TCGGATTAATACCTCGCTGCTCGCCGCCCTCAC	Forward primer for bls1 S1 probe
bls1-S1-Rev	GGTCGGGGCCGGGCATGGTGAA	Reverse primer for bls1 S1 probe
pah1-S1-For	CCAGATTAATGCGGCGCGGACGGTGCAG	Forward primer for pah1 S1 and Northern probe
pah1-S1-Rev	CGGGGAGACGGCGGTGGACA	Reverse primer for pah1 S1 and Northern probe
pah1-UP-Rev	CGCGGCTGCCCCTCCCTC	Reverse primer for pah1 S1 probe
ceaS1-RT-For	GCGCAGTCCGAGTCGTAC	Forward primer ceaS1 RT-PCR
ceaS1-RT-Rev	TTGGCGGTGTAGGTGGTGAC	Reverse primer ceaS1 RT-PCR
ceaS2-RT-For	AGGCCGCGTCGATTCTCTTC	Forward primer ceaS2 RT-PCR
ceaS2-RT-Rev	CGGCGGGTTGGGGACGGT	Reverse primer ceaS2 RT-PCR
KTA-ceaS1-For	GCGGGATCCGGGCGGTCAGCACGGT	Forward primer for cloning ceaS1 promoter
KTA-ceaS1-Rev	CCGGGTACCAGGGTCGCGAAGCACG	Reverse primer for cloning ceaS1 promoter
KTA-ceaS2-For	AACCCCAGGATCCGAGCCCCACCGTCACG	Forward primer for cloning ceaS2 promoter
KTA-ceaS2-Rev	CGGCCGGGTACCCCAAACACCTTCCCCACAC	Reverse primer for cloning ceaS2 promoter

^aNonhomologous sequences incorporated into oligonucleotide primers are underlined, and engineered restriction sites are shown in boldface letters.

^bDetails of the function of each oligonucleotide primer are provided in Materials and Methods.

High-resolution S1 nuclease mapping was carried out using the sodium trichloroacetate method (20). All double-stranded DNA probes were prepared by PCR using custom primers. DMSO (5% [vol/vol]) was included in the reactions, with the following program: 94°C for 2 min, followed by 10 cycles of 94°C for 45 s, 60°C for 45 s, and 72°C for 45 s, and finally 15 cycles of 94°C for 45 s, 65°C for 45 s, and 72°C for 45 s. The primers ceaS1-S1-For and ceaS1-PR-EX, along with p2.8-18 as a template, were used to prepare the probe to map the ceaS1 TSP. The primers ceaS2-S1-For and ceaS2-PR-EX, along with the template plasmid pBB5.3A, were used to prepare the probe for mapping the TSP of ceaS2. The primers bls1-S1-For and bls1-S1-Rev, and pah1-S1-For and pah1-UP-rev, along with p5.7 as a template, were used to prepare probes for S1 protection analysis of bls1 and pah1, respectively.

Primer extension analysis was performed using C. therm polymerase for reverse transcription in two-step RT-PCR (Roche) according to the manufacturer's instructions with the following changes. Twenty-microliter reaction mixtures were set up using 5 pmol of the end-labeled reverse primers ceaS1-PR-EX and ceaS2-PR-EX (Table 2) and 40 U of RNaseOUT Recombinant RNase Inhibitor (Invitrogen), with the following program: extension at 60°C for 60 min and termination at 80°C for 10 min.

DNA sequencing ladders were prepared for size estimation using the reverse primers and the template plasmids used in the preparation of the S1 probes or in the primer extension reactions. Samples were separated on 6% denaturing polyacrylamide sequencing gels for analysis as described earlier (38).

Northern blot analysis was carried out using established techniques (30) with 40 μg of RNA isolated from wild-type S. clavuligerus grown on soy medium for 96 and 120 h. Molecular Weight Marker III (Roche) was run along with the RNA samples for size estimation. The primers pah1-S1-For and pah1-S1-Rev were used to generate a 288-bp probe by PCR using p5.7 as a template. This probe was used for both S1 nuclease protection analysis (data not shown) and Northern hybridization. Probe annealing and subsequent washes were carried out under the same high-stringency conditions used in Southern blot analysis.

Preparation of enhanced green fluorescent protein (EGFP) reporter constructs.

A 781-bp DNA fragment spanning the ceaS1 promoter region was amplified by PCR using p2.8-18 as a template and the primer pair KTA-ceaS1-For and KTA-ceaS1-Rev. Similarly, a 721-bp DNA fragment encompassing the ceaS2 promoter region was amplified by PCR using pBB5.3A as the template and the primers KTA-ceaS2-For and KTA-ceaS2-Rev. The PCR products were treated with Taq DNA polymerase before ligation to pCR2.1TOPO (Invitrogen) according to the manufacturer's instructions. This gave pTOPO-ceaS1-4 and pTOPO-ceaS2-8, which contain the ceaS1 and ceaS2 promoter regions in pCR2.1TOPO, respectively. The double-stranded DNA sequence of the inserts was obtained using universal primers to ensure that no mutations were introduced.

The ceaS2 promoter region from pTOPO-ceaS2-8 was isolated as a BamHI/KpnI fragment and ligated into the corresponding sites of pIJ8660 to give pIJ8660-ceaS2. pIJ8660-ceaS2 contains the ceaS2 promoter region fused to a promoterless egfp gene for use as a reporter of expression driven by the ceaS2 promoter.

For unexplained reasons, we were unable to subclone the ceaS1 promoter region into pIJ8660 and therefore used an alternative approach. The ceaS1 promoter region from pTOPO-ceaS1-4 was isolated as a BamHI/KpnI fragment and ligated into the corresponding sites of pTO6 to give pTO6-ceaS1. The entire cassette encompassing the ceaS1 promoter region fused to the egfp gene from pTO6-ceaS1 was isolated as a 2.37-kb BamHI/EcoRI fragment and introduced into the corresponding sites of pSET152 to give pSET-ceaS1, which served as the ceaS1 reporter construct.

The plasmids pSET-ceaS1 and pIJ8660-ceaS2 were introduced into wild-type S. clavuligerus by conjugation. Strains that had the plasmids integrated at the φC31 attB site in the chromosome were isolated based on apramycin resistance and were verified by Southern hybridization (data not shown).

Confocal microscopy.

S. clavuligerus cultures harboring egfp reporter constructs were grown in TSBS for 36 h. Five-hundred-microliter amounts of the TSBS cultures were used to inoculate 25 ml of either SA or soy medium. After 72 h of growth, 1-ml amounts of the cultures were harvested and washed once in acetonitrile and then twice in sterile distilled water. The washed mycelia were mounted in 40% (vol/vol) glycerol before observations were made under the microscope.

Confocal microscopy was carried out using a Leica DM IRB inverted microscope. An argon laser (50 to 52% attenuation) provided excitation at 488 nm. Fluorescence due to EGFP excitation was detected between 505 and 520 nm, and corresponding differential interference contrast images were also obtained.

Western analysis.

Five-milliliter amounts of S. clavuligerus cultures grown in soy and SA media were harvested and resuspended in 1 ml of lysing buffer (100 mM HEPES [pH 7.2], 0.5 mg of lysozyme/ml, 2× Complete EDTA free protease inhibitor cocktail [Roche]). The suspensions were incubated at 37°C for 30 min, the cell membranes were broken by ultrasonic disruption, and then the cell debris was removed by centrifugation. Samples of cell extracts (CFEs) containing 50 μg of total protein were separated by SDS-PAGE (12% gels) as described earlier (38). The proteins were transferred onto polyvinylidene difluoride membranes (Immobilon-P; Millipore) using a Bio-Rad Transblot apparatus. The BM Chemiluminescence Western Blotting Kit (Mouse/Rabbit) (Roche) was used to detect proteins in accordance with the manufacturer's instructions. The primary antibody used to detect EGFP was the commercially available BD Living Colors A.v. Peptide Antibody (BD Biosciences) and was used at 1/400 dilution.

HPLC analyses of culture filtrates, bioassays, and growth determination.

High-performance liquid chromatographic (HPLC) analysis of culture supernatants after imidazole derivatization was performed under previously described conditions (30). Bioassays were used to detect clavulanic acid and alanyl clavam production using Klebsiella pneumoniae ATCC 15380 and Bacillus sp. strain ATCC 27860, respectively, as the indicator organisms (18, 30). Growth of S. clavuligerus in fermentation medium was estimated by measuring the optical density of broken mycelia at 495 nm as described earlier (7).

RESULTS

Nutritional regulation of ceaS1 and ceaS2.

In previous studies, when ceaS1 and ceaS2 mutant strains were prepared individually, both mutant strains still retained some ability to produce both clavulanic acid and the 5S clavams, depending on the fermentation medium used (15, 33, 42). The ceaS1 mutant produced both clavulanic acid and the 5S clavams in soy medium but only clavulanic acid in SA medium. In contrast, the ceaS2 mutant produced small amounts of clavulanic acid and the 5S clavam metabolites in soy medium only, while no clavulanic acid or 5S clavam production was observed in cultures grown in SA medium (15, 42). Based on the observed phenotypes, ceaS1 was postulated to be expressed in soy medium only, whereas ceaS2 was expressed in both soy and SA media. To verify this hypothesis, we examined the effects of growth in soy and SA media on both ceaS1 and ceaS2 expression at the transcriptional level.

RT-PCR was used to detect ceaS1 and ceaS2 transcripts, using RNA isolated from wild-type S. clavuligerus grown on SA medium for 72 and 96 h and on soy medium for 96 and 120 h. When RNA isolated from SA cultures was subjected to analysis, ceaS1 transcripts were not detected, whereas the same samples showed the presence of ceaS2 transcripts (Fig. 2A). In similar analyses of RNA isolated from soy-grown cultures, both ceaS1 and ceaS2 transcripts were detected by RT-PCR (Fig. 2A). The expression of hrdB, which encodes a constitutively expressed sigma factor in Streptomyces, was monitored as a control, and hrdB transcripts were detected at similar levels in all samples tested (Fig. 2). HPLC analysis of culture supernatants showed the expected levels of clavulanic acid and 5S clavams in cultures grown in both media used for RNA isolation and analysis (data not shown).

FIG. 2.

Assessment of ceaS1 and ceaS2 transcript levels by RT-PCR. RNA samples isolated from various strains of S. clavuligerus were analyzed by RT-PCR using primers specific for ceaS1, ceaS2, and hrdB. (A) Analysis of RNAs from S. clavuligerus wild-type (WT) cultures grown on SA medium for 72 and 96 h and from cultures grown on soy medium for 96 and 120 h. (B) Analysis of RNAs from S. clavuligerus wild-type cultures grown on soy medium for 96 and 120 h and S. clavuligerus ΔccaR::tsrA mutant cultures grown on soy medium for 96 and 120 h.

The promoter regions of ceaS1 and ceaS2 (−501 to +215 and −551 to +95 bp relative to the putative start codons of ceaS1 and ceaS2, respectively) were subcloned in front of a promoterless egfp gene to give plasmids pSET-ceaS1 and pIJ8660-ceaS2, respectively. Fluorescence arising due to EGFP expression was used as a reporter to monitor transcription driven from the promoters. These promoter constructs were introduced into wild-type S. clavuligerus, where they integrated into the chromosome via the φC31 attB site. The integration of the reporter plasmids into the S. clavuligerus chromosome was confirmed by Southern hybridization using DNA probes specific for both egfp and the respective promoters (data not shown).

The S. clavuligerus reporter strains C1G and C2G (Table 1) were grown on soy and SA media for 72 h, after which mycelia were harvested and analyzed by confocal microscopy to examine fluorescence arising due to EGFP expression and excitation. Fluorescence was observed in all samples except in the C1G strain grown on SA medium (Fig. 3A). Cell extracts (CFEs) were also prepared from the same samples that were subjected to confocal microscopy, and the presence of EGFP in the CFEs was analyzed by Western blotting. A 27-kDa band corresponding to EGFP was observed in all of the samples except for the lane containing CFE from the C1G strain grown on SA medium (Fig. 3B).

FIG. 3.

Use of EGFP as a reporter to detect ceaS1 and ceaS2 promoter activities. (A) Mycelia from S. clavuligerus EGFP reporter strains grown on SA and soy media for 72 h were analyzed by confocal microscopy. Both differential interference contrast (odd-numbered) and fluorescence (even-numbered) images were obtained and are shown side by side. Images 1 and 2, ceaS1 reporter grown on SA medium; images 3 and 4, ceaS2 reporter grown on SA medium; images 5 and 6, ceaS1 reporter grown on soy medium; images 7 and 8, ceaS2 reporter grown on soy medium. (B) Detection of EGFP in cell extracts (CFEs) from S. clavuligerus EGFP reporter strains grown on soy and SA media for 72 h. Cell extracts from samples subjected to microscopic analysis were also analyzed by Western blotting using commercially available antibodies raised against EGFP. C1G, lanes containing CFEs from the ceaS1 reporter strain; C2G, lanes containing CFEs from the ceaS2 reporter strain. The different culture media used to grow the reporter strains are indicated.

Bioassays indicated that all of the strains used in microscopic and Western analyses produced the expected levels of both clavulanic acid and alanyl clavam, and growth assays indicated that all the strains grew at similar rates (data not shown).

Transcriptional mapping of ceaS1 and ceaS2 promoters.

S1 nuclease protection and primer extension analyses were used to map the TSPs of both ceaS1 and ceaS2. RNA for the analysis was isolated from wild-type S. clavuligerus grown on soy medium (for 96 and 120 h). S1 protection analysis, using a probe extending from −261 to +26 bp relative to the putative ceaS1 start codon, indicated that the ceaS1 TSP was located 98 to 99 bp upstream of the ceaS1 start codon (Fig. 4A). Primer extension analysis was also conducted using the reverse primer ceaS1-PR-EX (Fig. 4B), and the ceaS1 TSP was mapped to a C residue located 98 bp upstream of the ceaS1 ATG start codon (Fig. 4C).

FIG. 4.

Mapping of the ceaS1 TSP. (A and B) Sequencing ladders (lanes G, A, T, and C) were prepared using the reverse primer ceaS1-PR-EX, along with p2.8-18 as a template. (A) A 297-bp probe (−261 to +26 bp relative to the putative ceaS1 start codon plus 10 bp of nonhomologous sequence) was used in the S1 protection assay. Lanes 1 and 2, RNA from wild-type S. clavuligerus grown on soy medium for 96 and 120 h and subjected to analysis; lane P+S1, control lane with unprotected probe digested with S1 nuclease; lane P, undigested-probe control. (B) Primer extension analysis using the ceaS1-specific reverse primer, ceaS1-PR-EX. Lane 1, RNA isolated from wild-type S. clavuligerus grown on soy medium for 96 h and subjected to primer extension analysis. The most probable TSPs are marked by asterisks. (C) DNA sequence of the ceaS1 promoter region. The open arrow represents ceaS1, with the arrowhead indicating the orientation of the gene. The bent arrow indicates the ceaS1 TSP, and the respective −10 and −35 promoter regions are also shown.

A probe extending from −204 to +22 bp relative to the ceaS2 start codon was used in an S1 protection assay to map the ceaS2 TSP, which was located 51 to 52 bp upstream of the ceaS2 start codon (Fig. 5A). Primer extension analysis using the reverse primer ceaS2-PR-EX (Fig. 5B) confirmed that the ceaS2 transcript originated from a G residue located 51 bp upstream of the ceaS2 ATG start codon (Fig. 5C).

FIG. 5.

Mapping of the ceaS2 TSP. (A and B) DNA sequencing ladders (lanes G, A, T, and C) were prepared for size estimation using the reverse primer ceaS2-PR-EX and the template plasmid pBB5.3A. The most probable TSPs are marked by asterisks. (A) S1 nuclease protection analysis using a 236-bp probe (−204 to +22 bp relative to the ceaS2 start codon plus 10 bp of nonhomologous sequence) used to map the ceaS2 TSP. Lanes 1 and 2, RNA from wild-type S. clavuligerus grown on soy medium for 96 and 120 h and subjected to analysis; lane P+S1, control lane with unprotected probe digested with S1 nuclease; lane P, undigested-probe control. (B) Primer extension analysis using the ceaS2-specific reverse primer, ceaS2-PR-EX. Lane 1, RNA from wild-type S. clavuligerus subjected to primer extension analysis. (C) DNA sequence of the ceaS2 promoter region. The open arrow represents ceaS2, with the arrowhead indicating the orientation of the gene. The bent arrow indicates the most probable ceaS2 TSP, and possible heptameric repeats, which are recognized by SARPs, are shown in solid boxes. The −10 and −35 promoter regions are also indicated.

Mapping of the bls1 and pah1 transcripts.

Since bls2 and pah2 are transcribed as part of a larger polycistronic transcript that also includes ceaS2 and cas2 (30), the regions upstream of bls1 and pah1 were examined using S1 nuclease protection assays to determine if they were also expressed as part of a polycistronic message. The bls1 transcript was mapped using RNA isolated from wild-type S. clavuligerus grown on soy medium for 96 h, together with a probe extending from −137 to +16 bp relative to the proposed bls1 ATG start codon. Only full-length protection of the probe was observed, indicating that there was no individual TSP located in the 23-bp intergenic region between ceaS1 and bls1 (Fig. 6A).

FIG. 6.

Mapping of the polycistronic transcript including bls1 and pah1 using S1 nuclease protection analysis. (A and B) DNA sequencing ladders (lanes G, A, T, and C) were prepared for size estimation using the reverse primers and template plasmids used for probe preparation. Lane S, RNA isolated from wild-type S. clavuligerus grown on soy medium for 96 h and subjected to analysis; lane P+S1, control lane with unprotected probe digested with S1 nuclease; lane P, undigested-probe control. Bands observed due to probe-probe reannealing and full-length protection of the probes, minus the 3′ nonhomologous 10-nucleotide sequences, are indicated. Multiple bands due to degradation of the nonhomologous sequences were also observed above the full-length protected probe fragments. (A) A 163-bp probe (−137 to +16 bp relative to the proposed bls1 start codon plus 10 bp of nonhomologous sequence) was used to map the bls1 transcript. (B) A 116-bp probe (−257 to −151 bp upstream of the proposed pah1 start codon plus 10 bp of nonhomologous sequence) was used to map the pah1 transcript. (C) Diagrammatic representation of the ceaS1, bls1, and pah1 polycistronic transcript. The large arrows represent the genes, with the arrowheads indicating the direction of transcription. The line represents the rest of the S. clavuligerus chromosome. The upper arrow represents the mRNA transcript, and bars 1 and 2 represent the S1 probes used to map the bls1 and pah1 transcripts, respectively (the diagram is not to scale).

The intergenic region between bls1 and pah1 was examined by S1 nuclease protection assays using a probe extending from −257 to +21 bp relative to the proposed pah1 ATG codon, but once again, only full-length protection of the probe was observed (data not shown). Because the DNA sequence ladder was unclear in the region of the full-length protected probe, a second 116-bp probe (−257 to −151 bp upstream of the proposed pah1 start codon) was also used in an S1 protection assay. Again only full-length protection of the probe was observed (Fig. 6B), indicating that there was no promoter immediately upstream of pah1. Since the probes used for the S1 nuclease protection studies did not cover the entire 317-bp intergenic region between bls1 and pah1, it was still possible that a TSP might be found further upstream of pah1. Northern analysis of RNA isolated from wild-type S. clavuligerus grown on soy medium for 96 and 120 h was conducted to investigate this possibility. The 288-bp probe prepared for the first pah1 S1 protection assay was used as the pah1 Northern probe. Only a single large band of ∼4.9 kb hybridized to the probe (Fig. 7), indicating that pah1 does not have an individual promoter and that it is transcribed as part of a tricistronic operon together with bls1 and ceaS1.

FIG. 7.

Northern analysis of wild-type S. clavuligerus RNA using a pah1-specific probe. Lane 1, molecular size marker; lanes 2 and 3, RNA from wild-type S. clavuligerus grown on soy medium for 96 and 120 h, respectively.

Effects of CcaR on ceaS1 and ceaS2 transcription.

RNAs isolated from wild-type and ΔccaR::tsrA strains of S. clavuligerus grown on soy medium for 96 and 120 h were analyzed by RT-PCR to monitor the effect of CcaR status on ceaS1 and ceaS2 expression. The transcription of ceaS1 was comparable in both the wild-type and the ΔccaR::tsrA strains (Fig. 2B). When ceaS2 transcription was examined in the same RNA samples, almost no transcripts were detectable in the ΔccaR::tsrA mutant compared to the wild-type strain (Fig. 2B). The expression of hrdB was also monitored as a control and was found to be constant in all of the samples (Fig. 2B).

The levels of clavulanic acid and 5S clavam production in the wild-type and the ΔccaR::tsrA mutant cultures used to isolate RNA were also determined by HPLC analysis of culture supernatants. The wild-type strain produced the expected levels of clavulanic acid and the 5S clavams after 96 and 120 h of growth on soy medium. Under the same conditions, the ΔccaR::tsrA strain did not produce any detectable levels of clavulanic acid, whereas normal 5S clavam production was observed (data not shown).

DISCUSSION

Previous studies have shown that paralogous pairs of genes encode the enzymes involved in the early stages of the shared biosynthetic pathway leading to clavulanic acid and the 5S clavams (15, 18, 22, 42). In order to understand to what extent each of these sets of genes contributes to the production of clavulanic acid and the 5S clavams, the transcriptional regulation of the genes was examined.

The ceaS1 transcript originated from a single TSP located at a C residue 98 bp upstream from the start codon, but the ceaS1 promoter region showed little similarity to known Streptomyces promoters (Fig. 4C), a reflection of the large diversity found in Streptomyces promoter sequences (39). When the bls1 and the pah1 transcripts were analyzed by S1 nuclease protection assays, no individual TSP was detected for either of these genes and only full-length protection of the probes was observed. This suggested that neither bls1 nor pah1 has its own dedicated promoter and that the upstream ceaS1 promoter drives the transcription of these genes.

The large intergenic region of 317 bp that separates bls1 and pah1 was not fully covered by the probes used for S1 nuclease protection studies. Therefore, Northern analysis was used to ensure that any pah1 transcript originating from a promoter located further upstream was not missed. Since pah1 and pah2 show 72% end-to-end identity at the nucleotide level (18), the probe used in Northern analysis was carefully chosen to be specific for pah1. The only hybridization seen was to a band ∼4.9 kb in size, again indicating that pah1 was transcribed as part of a large polycistronic transcript (Fig. 7). This 4.9-kb transcript is postulated to include ceaS1, bls1, and pah1, as the predicted length of a transcript extending from the ceaS1 TSP to the stop codon of pah1 would be ∼4.7 kb. The next gene downstream of pah1 is oat1, which is oriented in the direction opposite to pah1 transcription, and therefore its presence on the 4.9-kb polycistronic transcript can be ruled out. In addition, computational analysis of the 126-bp intergenic region between pah1 and oat1 predicted the presence of considerable secondary structure, consisting of multiple stem-loops with a cumulative G of −101.4, which could function as a transcriptional terminator (data not shown).

The transcriptional arrangement of ceaS1, bls1, and pah1 is similar to that of their paralogous counterparts from the clavulanic acid gene cluster (Fig. 1A). The ceaS2, bls2, pah2, and cas2 genes from the clavulanic acid gene cluster are transcribed as a 5.3-kb polycistronic transcript. In addition, cas2 is also transcribed as a 1.2-kb monocistronic transcript (30). The most significant difference between the 5.3-kb transcript arising from the clavulanic acid gene cluster and the 4.9-kb transcript arising from the paralogue gene cluster is the absence of the cas1 coding sequence in the paralogue gene cluster. cas1 is located elsewhere on the S. clavuligerus chromosome and is expressed as a 1.4-kb monocistronic transcript (30).

The transcript comprising ceaS2 was also mapped and was also found to arise from a single TSP located 51 bp upstream of the ceaS2 start codon. As was the case for ceaS1, the proposed ceaS2 promoter region did not show any significant similarity to any known Streptomyces promoters. Since S1 nuclease and primer extension analyses were used to identify all of the TSPs described in this study, it should be noted that both of the analysis methods employed are predictive of the TSP, provided the mRNA is not processed.

The nutritional regulation of ceaS1 and ceaS2 transcription was examined using RT-PCR, which demonstrated that ceaS1 was transcribed in soy medium only and not in SA medium. Under the same conditions and using the same RNA preparations, ceaS2 was transcribed in both soy and SA media at comparable levels. The ceaS1 and ceaS2 promoter regions were also subcloned in front of a promoterless egfp gene, and EGFP expression was used as a reporter to monitor transcription driven by the respective promoters. Confocal microscopy was used to detect fluorescence due to EGFP expression and excitation, and the results confirmed that ceaS1 was expressed in soy medium only and not in SA medium, whereas ceaS2 was expressed in both media tested. The results obtained from confocal microscopy were confirmed by Western analysis, which indicated that the fluorescence observed in the samples was due to true EGFP expression. Since ceaS1, bls1, and pah1 are expressed only as a 4.9-kb polycistronic message, it can be inferred that bls1 and pah1 will show the same general trend of nutritional regulation as ceaS1. In combination, our results indicate that ceaS1, bls1, and pah1 are expressed in soy medium but not in SA medium, whereas ceaS2, bls2, pah2, and cas2 are expressed in both soy and SA media. This explains the clavulanic acid- and 5S clavam-producing phenotypes observed when mutants defective in these genes were prepared and tested in previous studies (15, 18, 42).

The ccaR gene from the cephamycin gene cluster encodes a pathway-specific transcriptional regulator that coordinates the production of both cephamycin C and clavulanic acid (2, 31). Its effect on clavulanic acid biosynthesis is exerted, at least in part, through activation of the expression of a second pathway-specific transcriptional regulator, ClaR, from the clavulanic acid gene cluster. Mutants defective in CcaR do not produce any cephamycin C, and clavulanic acid production is also knocked out (2, 31), as the transcription of claR is reduced to near zero in these mutants (32). ClaR positively regulates genes involved exclusively in the biosynthesis of clavulanic acid (29). Previous studies have shown that ceaS2 expression is not under the control of claR (29), but the detailed effects of CcaR on ceaS1 and ceaS2 expression are not known.

Our results indicate that ceaS1 transcription is unaffected in the ccaR mutant compared to the wild-type strain, whereas the transcription of ceaS2 is almost eliminated in the ccaR mutant. Therefore, CcaR controls the production of clavulanic acid through at least two routes, an indirect route mediated by ClaR and a second route that may involve CcaR directly regulating ceaS2 promoter activity.

CcaR belongs to a family of transcriptional regulators called the Streptomyces antibiotic regulatory proteins (SARPs), which bind to specific heptameric repeats and promote transcription (45). Imperfect heptameric repeats can be identified in the region upstream of ceaS2 (Fig. 5C), consistent with the idea that CcaR may bind directly to the ceaS2 promoter region to regulate transcription. Such a notion, however, has not yet been demonstrated experimentally, and it is also possible that CcaR exerts its effect indirectly through additional proteins (20a, 38a). Since claR is not expressed in the ccaR mutant, it can also be inferred that claR has no effect on ceaS1 transcription, which was unaffected in the ccaR mutant. Similarly, due to the polycistronic nature of the transcript including ceaS1, bls1, and pah1, it follows that neither bls1 nor pah1 is affected by CcaR or ClaR.

The reason why S. clavuligerus possesses two sets of genes encoding enzymes involved in the early stages of clavulanic acid and 5S clavam biosynthesis is still unclear. One suggestion is that this could be a strategy to increase precursor and metabolite flux through the shared part of the pathway by increasing the gene dosage. This is consistent with the observation that both sets of the paralogous genes are expressed in complex soy medium, where precursor availability and growth would support greater metabolite production levels than are possible on defined SA medium, where only the ceaS2-oat2 set of paralogues is expressed. In addition, the increased production of these secondary metabolites may be of greater advantage in complex medium to ward off competition, especially that posed by faster-growing organisms. In defined media, such as SA, the expression of only one set of paralogous genes may suffice to provide an adequate supply of precursors, given the lower levels of clavulanic acid and 5S clavam production.

Another explanation put forth is that the two sets of paralogous genes may belong to two separate pathways, with one leading to clavulanic acid and the other to the 5S clavams. Since the two pathways proceed through common early steps, a sharing of biosynthetic intermediates results, but the two pathways may be regulated differently. This is consistent with our observation that, although clavulanic acid production was knocked out in the ccaR mutant, the 5S clavams were still produced during growth on soy medium. The ccaR mutant shows ceaS1 transcription equivalent to that seen in the wild-type strain and still produces wild-type levels of 5S clavams, whereas ceaS2 transcription is almost absent and clavulanic acid production is lost. This suggests that ceaS1, bls1, and pah1 may be more closely associated with the production of 5S clavams via a CcaR-independent pathway, whereas ceaS2 and its related paralogues are associated with clavulanic acid production and are regulated by CcaR. This is an attractive hypothesis from the point of view that the producer organisms would be best served by coordinating production of a β-lactam antibiotic (cephamycin C) with production of a β-lactamase inhibitor (clavulanic acid) through the action of a common regulator (CcaR). In contrast, no apparent advantage would be gained by coordinating production of the 5S clavams with production of cephamycin C. Further investigation is required before any firm conclusions can be drawn. In previous studies (2), it was reported that production of both 5S clavam and clavulanic acid was lost in a ccaR mutant, whereas in our hands, mutation of ccaR had no effect on 5S clavam production. This inconsistency may be attributed to differences in the methodologies and growth media used for culture propagation in the two studies or to the extensive variability that has been observed in 5S clavam production profiles within this species (43).

The functional holoenzyme forms of CEAS2 (9), BLS2 (25), and PAH2 (11) have all been characterized structurally and shown to be oligomers. Since these proteins were overexpressed and purified from E. coli, only homo-oligomers were observed. It is reasonable to expect that when the corresponding homo-oligomeric forms of CEAS1, BLS1, and PAH1 are expressed and purified, they may have somewhat different activities or kinetic properties, just as was seen for the CAS1 and CAS2 monomers (37). These differences in activities may be important under the specific nutritional conditions in which each of these isoenzymes is expressed. It is also possible that within S. clavuligerus the two isozymic forms of each protein can form hetero-oligomers, which could provide another mechanism to modulate enzyme activity based on nutritional and precursor availability.

Studies are under way to investigate the interactions among the different isoenzymes to gain greater understanding of the roles of the paralogous genes involved in clavulanic acid and 5S clavam biosynthesis.